Nuclear Medicine and Biology 41 (2014) e36–e43

Contents lists available at ScienceDirect

Nuclear Medicine and Biology

journal homepage: www.elsevier.com/locate/nucmedbio

Optimization of reaction conditions for the radiolabeling of

DOTA and DOTA-peptide with 44m/44Sc and experimental evidence
of the feasibility of an in vivo PET generator
S. Huclier-Markai a, b,⁎, R. Kerdjoudj a, C. Alliot b, c, A.C. Bonraisin b, N. Michel b, F. Haddad a, b, J. Barbet b, c
a
Laboratoire Subatech, UMR 6457, Ecole des Mines de Nantes/CNRS/IN2P3/Université de Nantes, 4 Rue A. Kastler, BP 20722, F-44307 Nantes Cedex 3, France
b
ARRONAX GIP, 1 rue Aronax, F-44817 Nantes Cedex, France
c
CRCNA, Inserm/CNRS/Université de Nantes, 8 quai Moncousu, 44007 Nantes Cedex 1, France

a r t i c l e i n f o a b s t r a c t

Article history: Introduction: Among the number of generator systems providing radionuclides with decay parameters
Received 28 September 2013 promising for imaging and treatment applications, there is the 44Ti (T1/2 = 60 years)/ 44Sc (T1/2 = 3.97 h)
Received in revised form 28 October 2013 generator. This generator provides a longer-lived daughter for extended PET/CT measurements compared to
Accepted 11 November 2013 the chemically similar system 68Ge/ 68Ga. Scandium also exists as 47Sc, a potential therapeutic radionuclide. It
is possible to produce 44Sc in a cyclotron using, for example, the 44Ca (d, n) 44Sc nuclear reaction. In that case,
Keywords:
the isomeric state 44mSc (T1/2 = 58.6 h) is co-produced and may be used as an in vivo 44mSc/ 44Sc generator.
In vivo generator
44m
Sc
The aim of this study is to evaluate the feasibility of this in vivo 44mSc/ 44Sc generator and to demonstrate that
44
Sc the daughter radionuclide stays inside the chelator after decay of the parent radionuclide. Indeed, the
DOTATATE physico-chemical process occurring after the primary radioactive decay (EC, IT, Auger electron …) has
PET prevented in many cases the use of in-vivo generator, because of the post-effect as described in the literature.
Methods: The DOTA macrocyclic ligand forms stable complexes with many cations and has been shown to be
the most suitable chelating moiety for scandium. Initially, the radiolabeling of DOTA and a DOTA-peptide
(DOTATATE) with Sc was performed and optimized as a function of time, pH, metal-to-ligand ratio and
temperature. Next, the physico-chemical processes that could occur after the decay (post-effect) were
studied. 44mSc(III)-labeled DOTA-peptide was quantitatively adsorbed on a solid phase matrix through a
hydrophobic interaction. Elutions were then performed at regular time intervals using a DTPA solution at
various concentrations. Finally, the radiolabelled complex stability was studied in serum.
Results: Radiolabeling yields ranged from 90% to 99% for metal-to-ligand ratio ranging from 1:10 to 1:500 for
DOTA or DOTATATE respectively. The optimum physico-chemical parameters were pH = 4–6, t = 20 min,
T = 70 °C. Then, the 44mSc-DOTATATE complex, radiolabeled at 98%, was adsorbed through a hydrophobic
interaction to a solid phase. Unlabeled scandium was completely eluted from the column whereas the
Sc-DOTATATE complex was 100% retained. The release of 44Sc from the complex due to decay was less than 1%
over 2 periods of 44mSc, independent of the DTPA concentration used for elution. 44mSc/ 44Sc-DOTATATE was
stable in serum over 72 h.
Conclusions: The results indicate that the decay of 44mSc to 44Sc does not affect the integrity of the radiolabeled
compound. Thus the 44mSc/ 44Sc generator is chemically valid and stable in serum. It could be used for PET
imaging as an in-vivo generator increasing the life time of the scandium and allowing the use of antibody as
labelled compound. Further in-vivo biological evaluations should complete this work.
© 2014 Elsevier Inc. All rights reserved.

1. Introduction
Cancer represents an important research axis all around the world.
The current treatments rest on surgery, chemotherapy or external
irradiation. Targeted radionuclide therapy, of which radioimmu-
notherapy (RIT) is an example, is a particularly innovative technique
for cancer therapy complementary to the existing therapies. It is based
⁎ Corresponding author. Tel.: +33 2 51 85 85 37; fax: +33 2 51 85 84 52.
E-mail address: huclier@subatech.in2p3.fr (S. Huclier-Markai).
on the use of a specific vector (e.g. antibody or peptide) that targets
the cell to destroy, labeled by a radioactive isotope. This radionuclide
can be an alpha emitter, a beta emitter or an emitter of Auger
electrons. The challenge consists in delivering the radioactive
molecules to cellular targets and to find a compromise between
toxicity to healthy tissues and antitumor effect. Present targeted
radiotherapy is used in clinics with 131I-iodide [1], 131I-MIBG [2,3],
131
I-tositumomab in Bexxar® or 90Y in Zevalin®. Recent develop-
ments reflect a clear trend towards a patient-based individualized
approach of treating disease, including the precise planning and

0969-8051/$ – see front matter © 2014 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.nucmedbio.2013.11.004
 S. Huclier-Markai et al. / Nuclear Medicine and Biology 41 (2014) e36–e43
accurate control of a given therapeutic protocol for an individual
patient [4]. This can be achieved by using a pair of isotopes one
dedicated to imaging and the other for treatment. By imaging the
patient pharmacokinetics and dosimetry information are obtained
allowing fine tuning of the injected activity of the isotope for therapy.
This is the underlying principle of Theranostic Radiopharmaceuticals
which will result in personalized treatment [5]. Among all the
emitters under consideration in this field, scandium is an innovative
element for Nuclear Medicine, which possesses an isotope suitable for
imaging by Positron Emission Tomography (PET) (Scandium 44) and
for targeted radionuclide therapy (Scandium 47). 47Sc is a β − emitter
which has physical properties close to 67Cu or 177Lu, which makes it a
promising candidate for targeted radionuclide therapy with appro-
priate properties [6–8]. Application of 47Sc, as an alternative
radionuclide to 177Lu, was proposed in earlier work [7,9]. As a
trivalent element, scandium should behave like 177Lu. The relatively
new 44Sc is a new option for the development of innovative PET
imaging probes. 47Sc which can be used for targeted radionuclide
therapy opens the field of Sc-based vectors from diagnosis to
therapy and gives a great opportunity for dosimetric calculations.
47
Sc is a low energy β − emitter with a 3.35-day half-life, and shows
a primary γ-ray at 159 keV, which is suitable for SPECT and
planar imaging. The advantage of 47Sc production compared to that
of 177Lu, is relatively easy isolation of the radionuclide from the target,
but the disadvantage is the smaller cross-section of the nuclear
reaction [10]. Scandium also displays favorable chemistry for
conjugation to MAb-chelate systems. It is chemically similar to 90Y,
and the same ligands developed for 90Y [5] can be used for chelating
47
Sc. Finally, even if 67Cu has been suggested for therapy, its
availability stays limited and the pair isotope 64Cu (T1/2 = 12.7 h)
can only be labeled with peptides.
Compared to gallium-68 or fluorine-18, the longer half-life of 44Sc
may allow for PET imaging of larger peptides and antibody fragments
that are currently limited due to the short half-lives of commonly used
PET radionuclides. Associated with targeted radionuclide therapy is
the necessity to perform dosimetry. This can be done using the same
compound labeled with a β + emitter of the same element. In this case,
a PET image can be obtained and dosimetry can be assessed prior to
the treatment in order to adjust the injected dose to minimize toxicity
and maximize patient response.
44
Fig. 1. Decay scheme of Sc (source: http://www.nndc.bnl.gov/nudat2).
e37
As stated before, 44Sc (T1/2 = 3.97 h) can be used in association
with 47Sc. However, for a radionuclide pair (one for PET imaging, the
other for therapy), it is better if their half-lives are comparable. This
can be achieved by using 44mSc (T1/2 = 58.6 h), an isomeric state of
44
Sc, which decays in vivo to 44Sc, thus 44mSc/ 44Sc generator could
allow for longer pharmacokinetic studies. This can be very important
when dealing with MAbs. The idea is to bind the “parent” radionuclide
to the vector to provide an in-vivo generator of the “daughter”
radionuclide. Indeed, 44mSc and 47Sc have similar half-lives (2.44 vs
3.35 days) which allow monitoring metabolic pathways over longer
periods of time. This could be an advantage compared to the use of
44
Sc. In addition, 44mSc is located at an energy level of 271 keV above
the ground state. It decays through β + emission (1.2%) to 44Ca and by
internal transition (98.8%) to the ground state ( 44Sc). Associated to
the internal transition, only a small secondary emission, mainly Auger
emission, is observed (2.74%). The photon emission induces some
recoil of the nuclei. The recoil energy is small and can be estimated
using the momentum conservation law to be only 0.89 eV. The ratios
of 44mSc in 44Sc are respectively 2% in activity at EOB. 5% of the 44Sc
decays by electron capture and the remaining 95% of the nuclide
decays by positron emission with a positron Emax of 1.472 MeV and
the positron emission at 1.457 MeV is the one that would be
monitored (see Fig. 1).
The energetic properties of these two radionuclides and the fact
that they are the same chemical element suggest that the 44mSc/ 44Sc
should be a good candidate for an in vivo generator. Indeed, there will
be no change in oxidation state, little recoil that could lead to the
escape of the isotope from the molecule [11], and little radiolysis
effects coming from the secondary emission, at variance with
previously considered in-vivo generators [12,13].
Finally, another application of scandium is the development of a
new nuclear medical imaging technique based on the measurement of
the emitter location in three dimensions on an event by event basis
using β +/γ emitters, as proposed by Grignon et al. [14]. Such
measurements could be realized with radionuclides which emit a γ-
ray quasi-simultaneously with the β + decay, like 44Sc. The aim of this
technique is to reconstruct the intersection of the classical line of
response of the two 511 keV photons (obtained with a standard PET
camera) with the incident direction of the third γ-ray. The emission
angle measurement of this additional γ-ray involves the use of a
e38 S. Huclier-Markai et al. / Nuclear Medicine and Biology 41 (2014) e36–e43
Compton telescope utilizing a new generation of cameras based on a
liquid xenon (LXe) time projection chamber. A spatial resolution of
2.3 mm has been determined with an injected activity of 0.5 MBq for
a 44Sc point source emitter [9].
Recently, the interest in scandium chemistry has increased
demonstrated by the number of papers about the 44Ti/44Sc genera-
tor [15–17], 44Sc [18], natSc [19], 46Sc [20–22] or 47Sc [10]. Our
team has previously developed a production and purification of
44m
Sc/44Sc from 44Ca (d, 2n) [23]. The present work is focused on
the experimental evidence of the feasibility of this new type of in vivo
generator. Notably, a complete study on the “post-effect” on 44mSc/44Sc
DOTA-like complexes is given. Furthermore, the biological stability of
44m
Sc/44Sc radiolabeled complexes is evaluated in this paper.
2. Materials and methods
2.1. Reagents and instrumentation
DOTATATE (DOTA-[Tyr3, Thr8]-octreotide) was purchased from
piChem (Austria). The bifunctional chelator 1,4,7,10-tetraazacyclodo-
decane-N,N’,N”,N”’-1,4,7,10-tertaacetic acid (DOTA) is conjugated to
the N-terminus of the 8 amino-acid peptide octreotide. The peptide
(250 μg) was dissolved in 250 μL of pure water (stock solution at
1 mg/mL). Hydrochloric acid, methanol, ethanol, and ammonia were
of analytical grade. DTPA (diethylenetriamine pentaacetic acid, Fluka)
and DOTA (1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid,
Macrocyclics Inc.) were of reagent grade. All other chemicals were of
pure reagent grade and used as received unless otherwise specified.
Deionized Milli-Q water (18.2 MΩ.cm, Millipore) was used in all
reactions. For some experiments, 46Sc(III) was used. It was provided
by Perkin Elmer (Specific Activity (S.A.) = 4.3.10 3 Ci/mol) or by ILL
(France). The radionuclidic purity was approximately 99% and was
checked by high-resolution gamma ray spectrometry. It was found
that the amount of metal ion brought by the tracer was less than
10 −10 M. The 44mSc/ 44Sc mixture was produced at the Arronax
cyclotron according to a procedure described elsewhere [23].
Rat serum was provided by Sigma (France) and aliquots of 450 μL
were prepared from this stock solution in a screw cap V-bottom
vial. The aliquots were kept frozen until use. Before the addition of
50 μL of Sc-Ligand aliquot, the serum aliquot was placed in a water
bath at 37 °C.
2.1.1. Mass Spectrometry
In order to achieve the analysis of DOTATATE, mass spectrometry
with electrospray ionization mass spectrometry in negative ion mode
(ESI(+)) was employed. ESI(+) has been proven to be the most
appropriate mode for the analysis of anionic compounds. An LC-MS-
IT-TOF (Shimadzu) mass spectrometer available at the Arronax
facility was employed. The samples analysed by direct infusion were
introduced into the source with a syringe pump at a flow rate of
5 μL/min. For mass calibration checks, a calibration solution of Na-TFA
(Shimadzu) was used. Nitrogen was employed both as drying and
nebulization gas. The ESI(+)-MS spectra were recorded with a source
temperature of 200 °C, a cone voltage of + 25 V and a capillary
voltage of ± 3.5 kV. The spectra were the sum of 3 micro scans for an
ion trap opened at maximum during 100 ms. Spectra were recorded
in the m/z range 500–2000.
2.1.2. Radioactive measurements
The solutions of 44m/44Sc were counted on a γ-spectrometer
(Ortec-Ametek, France). The detector was previously calibrated in the
same geometry using standard from LEA CERCA (France). The energy
rays are 271.1 keV and 1157 keV for 44mSc and 44Sc respectively with
a branching ratio of 98.8% and 99.9%. The 46Sc solutions were
measured by liquid scintillation using a Quantulus (Perkin Elmer)
Liquid Scintillation analyzer with Ultima Gold AB scintillation cocktail.
44m/44
2.2. Optimization of the labeling of DOTA and DOTATATE with Sc:
physico-chemical study
44m/44
Sc was obtained in 200 μL of 0.1 M HCl with an average
batch activity of approx. 30 MBq/mL. Labeling of DOTA and DOTA-
TATE was performed by mixing different volumes of DOTA or
DOTATATE stock solutions with the 44m/44Sc eluate in a 2 mL screw-
Cap Wheaton V-bottom vial. The solution was placed in a bath for the
required time and then cooled for 10 min at room temperature. To
test the radiolabeling yield, a radio-TLC was performed by spotting
3 μl onto a TLC Flex Plate (silica gel 60A, F-254, 200 μm, Merck) and
eluted with a developing solution of 25% aq. NH3/H2O/MeOH, 2/1/1
(v/v). The resulting TLC plate was counted for 20 min on an
autoradiographic system and associated software (Cyclone, Perkin
Elmer). The Rf values for unlabelled 44m/44Sc 3+, 44mSc/ 44Sc-DOTA,
44m
Sc/ 44Sc-DOTATATE conjugate were determined. Under these
conditions, the unchelated Sc 3+ has an Rf = 0 whereas 44mSc/ 44Sc-
DOTA, 44m/44Sc-DOTATATE moved with the front of the developing
solution (Rf ≅ 0.9 for both).
The study was performed by varying the temperature, the time,
the metal-to-ligand ratio, as well as the pH of the reaction solution, in
order to optimize the radiolabeling. Optimization of each parameter
corresponds at least to triplicate experiments.
The effect of DOTA or DOTATATE precursor amounts on the
labeling yield was analyzed in solutions at pH = 4. The metal to
ligand or peptide ratio was varied from 1:1 to 1:1000. Solutions were
heated for 1 h at 90 °C as described elsewhere [20,21]. The influence
of the temperature (at 25 °C, 40 °C, 70 °C and 90 °C) was investigated
at a fixed pH value of 4. For all these experiments, the radiolabeling
yield was monitored for times ranging from 5 min up to 1 h. The effect
of pH on the 44mSc/ 44Sc-DOTA or 44m/44Sc-DOTATATE formation was
determined between pH 2 to 7, with respective addition of acetate
buffer for pH ranging from 4 to 6, and HEPES buffer for pH 7. The
solutions were heated at 90 °C for 30 min. The pH of the solutions was
measured before and after the heating to ensure its stability.
44m/44
2.3. Feasibility of the Sc generator
For evaluating the “post-effect” on 44mSc/ 44Sc generator, to check
that there is no release of the metal from the DOTA chelating moiety
after the decay, a C-18 cartridge Phenomenex Strata-X Tube, 30 mg
(Phenomenex, France) was used. The cartridge was first activated with
5 mL of EtOH and rinsed with pure water. It was preliminary checked
that uncomplexed scandium Sc 3+ was not retained on the cartridge
and was simply eluted with water. In addition, it was checked that for a
concentration of free DOTATATE similar to the one used for the
radiolabeling, the peptide was entirely retained by hydrophobic
interaction onto the C-18 Strata-X cartridge and no peptide was eluted
by water. For this aim, solutions of 46ScCl3 (A = 400 kBq) and
DOTATATE (c = 5.10−7 M) in water were loaded onto the cartridge.
A detailed discussion on these preliminary experiments is given in the
results and discussion section of this paper.
The 44mSc/ 44Sc-DOTATATE complex was prepared in aqueous
solution according to the optimized procedure described in section
2.2 above. The radiolabeling yield was about 98 % meaning that there
was an excess of DOTATATE unlabeled and there was 2% of Sc 3+ in
solution. About 350 kBq of 44mSc/ 44Sc-DOTATATE was loaded onto
the C-18 Strata-X cartridge. The cartridge was then washed with 5 mL
of water in order to remove the 2% of uncomplexed scandium. Then,
the cartridge itself was eluted using increasing concentrations of
DTPA solutions (10 −7 M to 10 −2 M) and as a function of time (i.e.
0 min, 1 h, 4 h, 24 h, 48 h and 72 h). The decay could induce a release
of scandium which could be complexed by DOTATATE in excess on the
cartridge. If this occurs, there would be no possibility to measure Sc in
the eluate fraction. But the kinetic of complexation of Sc with
DOTATATE is quite low at room temperature. In contrast, Sc-DTPA has
 S. Huclier-Markai et al. / Nuclear Medicine and Biology 41 (2014) e36–e43 e39

been shown to present the same order of magnitude of the stability A

constant as Sc-DOTA but with a faster kinetics of complexation at 120
room temperature. Consequently, this experiment would allow for a
100
quick complexation of scandium by DTPA that could be released from
DOTATATE induced from the 44mSc/ 44Sc transition, and that could be
80

yield (%)
measured in the eluate.
For ensuring the validity of this experimental set-up, the
60
experiment was performed in the exact same conditions as described
above by using 46Sc-DOTATATE loaded onto the C-18 Strata-X 40
cartridge and eluted with the same increasing concentrations of
DTPA. The inlet and the outlet solutions were analyzed by gamma 20
spectrometry and by TLC. It was observed that neither the Sc-
DOTATATE complex nor the Sc-DTPA complex was eluted from the 0
cartridge. Thus, the experiment could determine the potential release 0 20 40 60 80
of 44Sc from DOTATATE. time (min)
Finally, in order to get the balance of activity (loaded, eluted), the
column was washed with pure ethanol, breaking the hydrophobic T=25°C T=40°C T=70°C T=90°C
interaction between DOTATATE and the cartridge, and the resulting
solution was counted on a γ-spectrometer. B
120

2.4. Stability study

100
44m
Sc/ 44Sc-DOTA and 44mSc/ 44Sc-DOTATATE as obtained previ-
80
ously were used for stability studies. An aliquot of 450 μL of rat serum

yield (%)
was incubated at 37 °C for 1 h. 50 μL of the Sc-L radiolabeled solution
60
was then added to serum and incubated at 37 °C. Samples were
analyzed by radio-TLC by spotting 3 μL of the corresponding solution
40
onto a TLC flex plate every hour during the first 4 h and then once a
day for a week. The TLC plate was developed in a solution 25% aq. NH3/
20
H2O/MeOH, 2/1/1 (v/v). The resulting TLC plate was counted for
20 min on a Cyclone counter (Perkin Elmer).
0
0 5 10 15 20 25 30 35
3. Results and discussion time (min)
44m/44 T=25°C T=40°C T=70°C T=90°C
3.1. Labeling of DOTA and DOTATATE with Sc
Fig. 2. Formation of (A) 44Sc-DOTA and (B) 44Sc-DOTATATE (triangle) at pH 4.0 as a
The macrocyclic chelator DOTA, either free or coupled to targeting function of temperature. Experimental conditions M:L = 1500; pH = 4, monitored for
molecules, forms very stable complexes with many trivalent 1 h (represented only out to 30 min).
radionuclides used for molecular imaging or treatment, like 68Ga,
90
Y, 225Ac etc. [17,24,25], but also with non-radioactive Gd(III)
for magnetic resonance imaging (MRI) [26,27]. For Sc(III), the stability heating. With peptides, the temperature parameter is less important
constant values with the typical ligands EDTA, DTPA, DOTA, compared to antibodies and the optimum radiolabeling could be
NOTA, and TETA were determined with the following order reached by using a temperature of 90 °C without affecting their
TETA b NOTA b EDTA b DTPA b DOTA [21]. Sc(III) complexes are biological efficacy as shown for 111In, 90Y, 177Lu or 68Ga for instance
formed immediately with EDTA, DTPA and NOTA. For DOTA, [28–31].
complexation takes several hours to several days and the complex Therefore, for studying the other parameters and in order to reach
formation generally requires heating at high temperatures in contrast a fast kinetics, the temperature was held constant at 90 °C, at a fixed
to open-chain analogues. pH of 4. The overall 44mSc/ 44Sc radiolabeling yield for DOTA and
This work examines the influence of the different parameters on DOTATATE was then studied at different metal-to-ligand molar ratios
radiolabeling of DOTA and DOTATATE, with 44m/44Sc. The chemical and results are given in Table 1. As expected decreasing the metal-to-
feasibility of 44m/44Sc as a potential in vivo generator was evaluated by ligand ratio increased the radiolabeling yield. Nonetheless, for in vivo
obtaining the maximum radiolabeling yield prior to any loading of the
complex on the C-18 cartridge and by monitoring the eluate.
The influence of the temperature (25–90 °C) on the formation of Table 1
44m
Sc/ 44Sc-DOTA and 44mSc/ 44Sc-DOTATATE conjugate was deter- Labeling yields of 44mSc/44Sc–DOTA and 44m
Sc/44Sc–DOTATATE complexes at different
mined at a fixed metal to ligand ratio of 1:500, in a total volume of metal:ligand molar ratios.
500 μL and at a pH = 4. This metal-to-ligand ratio corresponds to a Ratio DOTA DOTATATE
total of 3.10 −12 mol of either DOTA or peptide. The results are M: L
Radiolabelling yield (%) Radiolabelling yield (%)
presented in Fig. 2.
At room temperature (i.e. 25 °C), for 1 h of contact, the 1:10 2.6 ± 0.5 4.7 ± 0.5
1:50 16.4 ± 0.8 18.3 ± 0.9
radiolabeling yield was about 45% and 60% for 44mSc/ 44Sc-DOTA and
44m 1:100 26.9 ± 1.4 51.3 ± 1.6
Sc/ 44Sc-DOTATATE respectively. Heating at 40 °C achieved more 1:500 78.6 ± 1.9 96.3 ± 2.8
than 80% of the 44mSc/ 44Sc-DOTA and 44mSc/ 44Sc-DOTATATE complex 1:1000 86.7 ± 2.5 96.7 ± 2.8
after 30 min of heating, reaching a plateau at 1 h. The yields reached Experimental conditions: 1 h, 90 °C, pH = 4. The radiolabeling yields correspond to
85% and up to 95% at temperatures of 55 °C, 70°–90 °C respectively. the average value determined on triplicate experiments and are given with an
The best radiolabeling yield was obtained at 70 °C after 20 min of uncertainty of 5%.
e40 S. Huclier-Markai et al. / Nuclear Medicine and Biology 41 (2014) e36–e43

imaging (especially in small animals), the total amount of peptide
should be as low as possible. Thus, for DOTATATE, the highest
radiolabeling yields (i.e. N 95%) were obtained at ratio over 1:500.
These conditions represent 3 pmol of peptide and the specific activity
is about 8 MBq/nmol of peptide. This value could be improved with
increasing activities. This was not the aim of this work.
Labeling of DOTA and DOTA-conjugated derivatives with medi-
cally useful radionuclides is generally performed from pH = 2 to 6
[15,32–34], sometimes higher depending of the radionuclide consid-
ered. Scandium belongs to the transition metals group and its
complex formation is strongly dependent on the pH of the aqueous
solution. Therefore, pH is an important parameter during the
radiolabeling optimization process. Since it is a trivalent element
with a lanthanide-like behavior, a range between 2 and 6 for the pH
values was considered. It was decided to keep the temperature
constant at 90 °C, and the metal-to-ligand ratio fixed at 1:500. As
previously observed with other radionuclides, the highest reaction
yield was obtained when the pH was kept between 4 and 5.5 (Fig. 3).
No significant difference was observed for the radiolabeling yield in
this range. The yields were 92%–94% for DOTA and 92%–95% for
A 110
100
90
yield (%)
DOTATATE in this range. Increasing the pH up to 6 caused some
decrease in labeling yields due to the hydrolysis of 44Sc(III). Acidifying
the solution below pH 2 resulted in an extreme decrease in labeling
yields, reaching approximately 40%, probably due to the protonation
of the DOTA chelator. For further radiolabeling, a pH between 4 and
5.5 seemed to be appropriate for 44mSc/ 44Sc with DOTA-based ligands.
Finally, from all previous parameters studied (Figs. 1 and 2;
Table 1), the influence of time on the radiolabeling yield was
scrutinized. Whichever parameter studied (pH, temperature, metal-
to-ligand ratio), it was observed that after 20 min, the radiolabeling
yield reached a plateau, either for DOTA or DOTATATE.
In conclusion, from this systematic study, the radiolabeling of
44m/44
Sc on DOTA and its peptide derivative DOTATATE, has been
optimized. The most suitable conditions have been identified as a
metal-to-ligand ratio of 1:500 in acetate buffer at pH = 4 and heating
the mixture at 90 °C for 20 min provided the best labeling yields N 94%
and N 98%, respectively. These values are in agreement with previous
studies performed on 44Sc-DOTATOC and 44Sc-DOTATATE with 44Sc
eluated from 44Ti generator [17,35].
3.2. Experimental concept for testing the feasibility of the
44m/44
Sc generator
The recoil effect in β-decay in vivo generators has been studied in
the literature, notably in the case of 140Nd/ 140Pr [12], 166Dy/ 166Ho [13]
or 103Pd/ 103mRh [11,36]. Nonetheless, these last authors did some case
calculations in order to evaluate if the recoil of the 103mRh out of the
carrier molecule could occur. 103Pd decays by electron capture but a
number of Auger electrons and gammas could also contribute to the
post-effect of the chelating agent. The same was done for 166Dy/ 166Ho

80
but in the case of 140Nd/ 140Pr, the chemical fate of 140Pr in aqueous
solutions was experimentally studied and a 140Nd/ 140Pr radionuclide
70
generator was conceived by use of a DOTATOC peptide absorbed on
a solid phase [12].
60 In the perspective of medical applications, the Sc(III)-DOTA
complex has been shown to be stable over several days in the
50 presence of a bone mimic and in rat serum [21]. DOTA is also a good
chelator of Ho and Nd but against all expectations, it was evidenced
40 that 166Ho, from 166Dy/ 166Ho in vivo generator was released
0 10 20 30 40 50 60 70 from DOTA [13]. It was evidenced that this was not due to the low
time (min) recoil energy (Q = 0.486 MeV) but to a “post-effect” as shown
pH=2 pH=3 pH=4 pH=5 pH=6 previously for the 140Nd/ 140Pr generator with an even lower recoil
energy (Q = 0.222 MeV) [12]. The post-effect is attributed to the
physico-chemical processes occurring after the primary radioactive
B 110 decay (EC, IT, Auger electron …). From published data on 140Nd/ 140Pr,
DOTA-conjugated peptides (such as DOTATOC or DOTATATE) can be
100
adsorbed onto a solid reversed phase from aqueous solutions with
90 high distribution coefficient, while free lanthanide cations or simple
lanthanide complexes can be eluted [12]. If the daughter nuclide 44Sc
80 is generated as a free ionic species, it will be easily separated using
small volumes of solutions (due to high stability constant and fast
yield (%)

70 kinetics of complexation with DTPA). The peptide is bound to the C-18

cartridge. Since scandium has a lanthanide-like chemical behavior and
60
due to the high thermodynamic stability and the kinetic inertness of
50 Sc-DOTA type complexes [20,21], the release of the longer lived
parent radionuclide 44mSc is inhibited. In contrast, the shorter-lived
40 radionuclide 44Sc could be released due to the post-effect. Based on
these elements, the chemical fate of 44Sc in aqueous solution was
30 studied by separation of the different chemical forms of the parent
and the daughter radionuclides.
20
0 10 20 30 40 50 60 70 It was first ensured that the C-18 cartridge did not retain scandium
time (min) as a cationic form Sc 3+. For this aim, a 46ScCl3 solution was loaded
onto the column and in a second experiment a 44m/44Sc solution was
pH=2 pH=3 pH=4 pH=5 pH=6
loaded. Columns were eluted with water. It was determined that 70%
Fig. 3. Formation of 44Sc-DOTA (A) and 44Sc-DOTATATE (B) as a function of pH. of Sc 3+ was directly eluted and the remaining 30% eluted by washing
Experimental conditions: 1 h, 90 °C, M:L = 1:500. the cartridge with 5 mL of water. The 46ScCl3 solution was set at a pH
 S. Huclier-Markai et al. / Nuclear Medicine and Biology 41 (2014) e36–e43 e41

A
Inten.(x10,000,000)

1.5 479.2057
[M] 2+

1.0

718.3063
0.5

242.2841
243.2879 658.7778 1435.7100
0.0
250 500 750 1000 1250 1500 1750 m/z

B
Inten.(x1,000,000)

242.2852
2.0

1.0

243.2888
371.1023 519.1441 872.7412
0.0
250 500 750 1000 1250 1500 1750 m/z

Fig. 4. ESI(+)-IT-TOF-MS spectra (A) of the DOTATATE solution before elution on C18 Strata X cartridge and (B) of the post-column eluate (similar to the spectrum of pure Milli Q water).
The signal intensities are given in arbitrary units.

A
6,00%
%of release activity with regards to the

5,00%
radiolableing yield

4,00%

3,00%

2,00%

1,00%

0,00%
0 1 2 3 4 5 6 7 8 9 10 11
volume DTPA (mL)

Série1 Série2

Sc-DTPA

Rf=0.6 Sc-DOTATATE

Rf=0.3

46 44m/44
Fig. 5. (A) Elution profiles of uncomplexed Sc from the C18 Strata X column Sc-DOTATATE and Sc-DOTATATE. (B) TLC of Sc-DOTATATE and Sc-DTPA.
e42 S. Huclier-Markai et al. / Nuclear Medicine and Biology 41 (2014) e36–e43
of 4 to 7 and the same experiment performed in order to determine if
the hydrolyzed species of scandium could be retained by the column.
Again, it was observed that regardless of the initial pH all of the
46
ScCl3 solution, Sc 3+ and hydroxo species [Sc(OH)] 2+, [Sc(OH)2] + or
[Sc(OH)3] present, were eluted from the column.
In a second experiment, free DOTATATE solution was loaded onto
the cartridge and was eluted with DTPA solutions at increasing
concentrations. The initial solution and the eluted fraction were
analyzed by ESI(+)-IT-TOF mass spectrometry. The spectra are given
in Fig. 4 A and B respectively. As shown, DOTATATE was entirely
retained by the cartridge.
Finally, the 46Sc-DOTATATE, radiolabeled at 97 %, was loaded onto
the C18 Strata X column and eluted with water and DTPA solutions at
increasing concentrations. The resulting solutions were analyzed by
liquid scintillation. In this case, the decay did not lead to metal release.
The uncomplexed Sc was eluted from the cartridge by washing with
water. The elution profile is given in Fig. 5 A. It was further verified
that no 46Sc-DTPA, neither free 46Sc nor 46Sc-DOTATATE was eluted
from the cartridge by TLC analysis as shown in Fig. 5B. The same
experiment was realized with the 44m/44Sc-DOTATATE radiolabeled
complex and the corresponding elution profile is given in Fig. 5A.
From this, it was demonstrated, as expected, that the behaviors of 46Sc
and 44m/44Sc are similar. The unlabeled fraction of scandium was
totally eluted from the cartridge after 5 mL of water. Then, the
cartridge was washed with DTPA solutions at different concentrations
and the process was repeated at various times. It means that the
stability of the system was evaluated.
The elution yield was calculated by taking into consideration the
decrease of 44mSc activity in the eluate as followed, according to
Zhernosekov et al. [12]:
 
44
A Sc ¼ x:Aact44mSc :expð−l44mSc :tÞ ð1Þ
Where Aact is the actual (at elution time) activity of 44mSc adsorbed
on the column and ξ is the fraction of the eluted 44mSc. The cartridge
was eluted at various times represented by t. This value represents the
elution yield of the radionuclide generator. The activity calculated is
given in Fig. 6. It was assumed that the two isotopes of scandium have
reached secular equilibrium before starting the elution process. This
calculated value represents the activity of 44Sc that would be eluted
from the cartridge by assuming that 100% of the 44mSc loaded onto
that the column decays releasing 44Sc from the chelating moiety.
These values are confronted to the experimental activities that were
measured in the eluate fraction at various times. As shown in Fig. 6, at
14000
12000
10000
Activity of Sc-44
A (Sc44) th
8000
A (Sc44) exp
6000
4000
2000
0 4. Conclusions
0 50 100 150 200
time (hours)
Fig. 6. Decay of 44m
Sc after an elution of the generator. The line represents the activity
calculated considering a complete release of 44Sc after 44mSc decay and the dots are the
experimental activities measured post-column in the DTPA fractions at various times.
serum stability
120
dotatate
100
80
yield (%)
60
40
20
0
0 1 3 5 20 22 24 48 72 96
time(hr)
44m/44
Fig. 7. Stability of Sc(III)-DOTATATE in rat serum.
all times considered, the measured activities in the eluated fraction
were under the detection limit of the γ-detector, indicating there was
no release of scandium. Consequently, there was no release from the
chelating moiety induced by the 44mSc to 44Sc decay. The presented
experimental generator design allows thus to consider 44m/44Sc as a
in vivo PET generator due to its stability observed over several half-lives.
In our case, the in-vivo generator is feasible as compared to the other
studied in-vivo generators [12,13] mainly because the recoil energy is
low for the emission of Auger electron. Nonetheless, the proposed
generator offers an appropriate biological stability to be further
considered for medical applications.
3.3. Challenge studies for testing the biological feasibility of the
44m/44
Sc generator
For RIT applications, metal radionuclides must remain associated
with the targeting chelate-protein to minimize the toxicity derived
from the dissociation and/or unsuitable biodistribution. Thus, a
critical point is to study the reversibility of the equilibrium. One of
the most informative stability assays that can be readily performed in
the laboratory is a measurement of the stability of a metal ion complex
in blood serum [37]. Whilst serum stability can be seen as a
benchmark of in vivo stability in the extracellular environment
(i.e. during blood perfusion), it cannot be relied on alone as it does
not fully mimic the intracellular conditions that might be encountered.
However, serum stability studies provide information on possible
pathways by which the radiopharmaceutical in question can become
demetallated. These studies are often informative and practical at
this stage before performing in vivo studies of the complexes. The
stabilities of 44m/44Sc-DOTATATE with rat serum were monitored as a
function of time and are presented in Fig. 7. It was observed that the
complex is very stable over several hours with regard to serum. This
result led us to the conclusion that the 44m/44Sc generator with
radiolabeled peptides would be stable in biological medium as already
observed in the case of 44Sc-DOTATOC with 44Sc provided from 44Ti
[17,38]. Work was done to show that high labeling yields (N98%) can be
obtained with DOTATOC and that it remains stable out to at least 25 h
even when challenged by ligands such as EDTA and DTPA. Additionally,
a patient was injected with 32 MBq (0.86 mCi), and a PET/CT scan was
taken at 19 h post-injection [5,39].
This study examined the chemico-physical parameters on the
radiolabeling of 44m/44Sc with DOTA and DOTATATE. The most suitable
conditions were identified as a metal-to-ligand ratio of 1:500 in
acetate buffer at pH = 4 and heating the mixture at 90 °C for 20 min
 S. Huclier-Markai et al. / Nuclear Medicine and Biology 41 (2014) e36–e43
provided the best labeling yields N94% and N98%, respectively. In
addition, it was evidenced that there was no chemical bond break due
to the decay of 44mSc to 44Sc or to post-effects. We have thus
demonstrated that 44m/44Sc is a promising in vivo generator as a PET
imaging agent for being coupled to proteins or even better to
antibodies. In this last case, further improvement of the radiolabelling
is foreseen (lower temperature). The study of the stability in the
presence of rat serum, indicated that 44m/44Sc-DOTATATE was stable
over several days. Nonetheless, careful biological studies in animal
models still remain the most appropriate method for evaluating the
in vivo inertness of radiometal-labeled chelates prior to clinical trials.
Acknowledgments
The authors acknowledge Mr Ulli Koester (Institut Laue Langevin
ILL, Grenoble, France) for providing 46Sc. Pr C.S. Cutler (MURR,
University of Missouri, USA) is acknowledged for the fruitful
discussions that have contributed to enhance this work.
The ARRONAX cyclotron is a project promoted by the Regional
Council of Pays de la Loire financed by local authorities, the French
government and the European Union. This work was supported by
grants from the Pays de la Loire Council "Nucléaire pour la Santé"
(NucSan) and from the French National Agency for Research called
“Investissements d’Avenir” n°ANR-11-LABX-0018-01.
References
[1] Bender JM, Dworkin I. Iodine-131 as an oncologic agent. J Nucl Sci Technol
1993;21:140–50.
[2] Levington VJ, Zivanoic MA, Ackery DM. [131I]-MIBG therapy for malignant
pheochromocytoma. Nucl Med Commun 1991;12:257–9.
[3] Naeem M, Home T, Hawkins LA. Follow up of patients post therapy with [131I]-
MIBG. Nucl Med Commun 1989;10:225–8.
[4] Roesch F, Baum RP. Generator-based PET radiopharmaceuticals for molecular
imaging of tumours: on the way to THERANOSTICS. Dalton Trans 2011 Jun 21;40
(23):6104–11.
[5] Baum RP, Kulkarni HR. THERANOSTICS: from molecular imaging using Ga-68
labeled tracers and PET/CT to personalized radionuclide therapy — the Bad Berka
experience. Theranostics 2012;2(5):437–47.
[6] Mausner LF, Kolsky KL, Joshi V, Srivastava SC. Radionuclide development at BNL
for nuclear medicine therapy. Appl Radiat Isot 1998;49:285–92.
[7] Mausner LF, Joshi V, Kolsky KL, Meinken GE, Mease RC, Sweet MP, et al. Evaluation
of chelating agents for radioimmunotherapy with scandium-47. J Nucl Med
1995;36:104–10.
[8] Kolsky KL, Pietrelli L, Mausner LF, Srivastava SC. Radiochemical purification of no-
carrier-added scandium-47 for radioimmunotherapy. J Nucl Med 1995;32:
945–50.
[9] Bartos B, Majkowska A, Krajewski S, Bilewicz A. New separation method of no-
carrier-added 47Sc from titanium targets. Radiochim Acta 2012;100:457–63.
[10] Polosak M, Kasperek A, Krajewski S, Bilewicz A. Stability of 47Sc-complexes with
acyclic polyamino-polycarboxylate ligands. J Radioanal Nucl Chem 2013;295:
1867–72.
[11] van Rooyen J, Szucs Z, Zeevaar JR. A possible in vivo generator 103Pd/103mRh—recoil
considerations. Appl Radiat Isot 2008;66:1346–9.
[12] Zhernosekov KP, Filosofov DV, Qaim SM, Rösch F. A 140Nd/140Pr radionuclide
generator based on physic-chemical transitions in 140Pr complexes after electron
capture decay of 140Nd-DOTA. Radiochim Acta 2007;95:319–27.
[13] Zeevart JR, Szucs Z, Tkacs S, Jarvis N, Jansen D. Recoil and conversion electron
considerations of the 166Dy/166Ho. Radiochim Acta 2011;99:1–5.
[14] Grignon C, Barbet J, Bardiès M, Carlier T, Chatal JF, Couturier O, et al. Nuclear
medical imaging using β+/γ coincidences from 44Sc radionuclide with liquid
xenon as detection medium. Nucl Inst Methods Phys Res A 2007;571(1–2):
142–53.
e43
[15] Filosofov DV, Loktionova NS, Rösch F. A 44Ti/44Sc radionuclide generator for
potential application of 44Sc-based PET-radiopharmaceuticals. Radiochim Acta
2010;98:149–56.
[16] Pruzynsky M, Loktionova NS, Filosofov DV, Rösch F. Processing of generator
produced 44Sc for medical application — radiolabelling of DOTATOC with 44Sc.
J Label Compd Radiopharm 2009;52:S490.
[17] Pruzynsky M, Majkowska A, Loktionova NS, Rösch F. Radiolabelling of DOTATOC
and DOTATATE with the longer-lived, generator-derived trivalent positron
emitter 44Sc. Appl Radiat Isot 2012;70(6):974–9.
[18] Severin GW, Engle JW, Valdovinos HF, Barnhart TE, Nickles RJ. Cyclotron produced
44g
Sc from natural calcium. Appl Radiat Isot 2012;70(8):1526–30.
[19] Koumariou E, Pawlak D, Korsak A, Mikolajczak R. Comparison of receptor
affinity of natSc-DOTA-TATE versus natGa-DOTA-TATE. Nucl Med Rev 2011;14
(2):85–9.
[20] Majkowska-Pilip A, Bilewicz A. Macrocyclic complexes of scandium radionuclides
as precursors for diagnostic and therapeutic radiopharmaceuticals. J Inorg
Biochem 2011;105:313–20.
[21] Huclier-Markai S, Sabatie A, Ribet S, Kubicek V, Paris M, Vidaud C, et al. Chemical
and biological evaluation of scandium(III)-polyaminopolycarboxylate complexes
as potential PET agent and radiopharmaceutical. Radiochim Acta 2011;99:653–62.
[22] Moghaddam-Banaem L, Jalilian AR, Pourjavid MR, Radfar E, Bahrami-Samani A,
Yavari K, et al. Development of a radioscandium immunoconjugate for radio-
immunotherapy. Radiochim Acta 2012;100:215–21.
[23] Alliot C, Kerdjoudj R, Mokili M, Michel N, Haddad F, Huclier-Markai S, et al.
Processing of a 44m/44Sc generator for potential medical applications: production
and chemistry aspects; 2013 [Submitted to Appl. Rad., Isot.].
[24] McDevitt MR, Ma D, Simon J, Frank RK, Scheinberg DA. Design and synthesis of
225
Ac radioimmunopharmaceuticals. Appl Radiat Isot 2002;57:841–7.
[25] Mohsin H, Fitzsimmons J, Shelton T, Hoffman TJ, Cutler CS, Lewis MR, et al.
Preparation and biological evaluation of 111In-, 177Lu- and 90Y-labeled DOTA
analogues conjugated to B72.3. Nucl Med Biol 2007;34:493–502.
[26] Meyer D, Schaefer M, Bonnemain B. Gd-DOTA, a potential MRI contrast agent.
Current status of physicochemical knowledge. Invest Radiol 1988;23:S232–5.
[27] Oudkerk M, Sijens PE, van Beek EJ, Kuijpers TJ. Safety and efficacy of Dotarem
(Gd-DOTA) versus Magnevist (Gd-DTPA) in magnetic resonance imaging of the
central nervous system. Invest Radiol 1995;30:75–8.
[28] Frampas E, Maurel M, Remaud-Le Saec P, Mauxion T, Faivre-Chauvet A,
Davodeau F, et al. Pretargeted radioimmunotherapy of colorectal cancer
metastases: models and pharmacokinetics predict influence of the physical
and radiochemical properties of the radionuclide. Eur J Nucl Med Mol Imag
2011;38:2153–64.
[29] Schoffelen R, Sharkey RM, Goldenberg DM, Franssen G, McBride WJ, Rossi EA, et al.
Pretargeted immuno-positron emission tomography imaging of carcinoembryo-
nic antigen-expressing tumors with a bispecific antibody and a 68Ga and 18 F-
Labeled hapten peptide in mice with human tumor xenografts. Mol Cancer Ther
2010;9:1019–27.
[30] Gold DA, Goldenberg DM, Karacay H, Rossy EA, Chang CH, Cardillo TM, et al. A
novel bispecific trivalent antibody construct for targeting pancreatic carcinoma.
Cancer Res 2008;68:4819–26.
[31] Green J, Pagel JM, Pantelias A, Hedin N, Lin Y, Wilbur DS, et al. Pretargeted
radioimmunotherapy for B-cell lymphomas. Clin Cancer Res 2007;13:5598–603.
[32] Loncin MF, Desreux JF, Merciny E. Coordination of lanthanides by two polyamino
polycarboxylic macrocycles: formation of highly stable lanthanide complexes.
Inorg Chem 1986;25:2646–9.
[33] Moi MK, Meares CF, DeNardo SJ. The peptide way of macrocyclic bifunctional
chelating agents. Synthesis of 2-(p-nitrobenzyl)-1,4,7,10-tetracyclododecane-N,
N', N", N"', N""-tetraacetic acid, and study of its yttrium (III) complex. J Am Chem
Soc 1988;110:6266–71.
[34] Izatt R, Pawlak K, Bradshaw SJ, Bruening RL. Thermodynamic and kinetic data for
macrocycle interaction with cations, anions, and neutral molecules. Chem Rev
1995;95:2529–33.
[35] Loktionova N, Filosofov D, Pruszynski M, Roesch F. Design and performance of a
novel 5 mCi 44Ti/44Sc-radionuclide generator. J Nucl Med 2010;51(S2):1467–73.
[36] Szucs Z, van Rooyen J, Zeevart JR. Recoil effect on β-decaying in vivo generators,
interpreted for 103Pd/103mRh. Appl Radiol Isot 2009;67:1401–4.
[37] Cole WC, Denardo SJ, Meares CF, McCall MJ, Denardo GL, Epstein AL, et al.
Comparative serum stability of radiochelates for antibody radiopharmaceuticals.
J Nucl Med 1987;28(1):83.
[38] Koumariou E, Pawlak D, Korsak A, Mikolajczak R. Comparison of receptor affinity
of natSc-DOTAT-TATE versus natGa-DOTA-TATE. Nucl Med Rev 2011;14(2):85–9.
[39] Cutler CS, Hennekens H, Nebiat S, Huclier-Markai S, Jurisson SS. Combined
radionuclide imaging and therapy. Chem Rev 2013;113(2):858–83.