An efficient and practical radiosynthesis of [11C]temozolomide

Article abstract of DOI:10.1021/ol302791x

Temozolomide (TMZ) is a prodrug for an alkylating agent used for the treatment of malignant brain tumors. A positron emitting version, [ ¹¹C]TMZ, has been utilized to help elucidate the mechanism and biodistribution of TMZ. Challenges in [¹¹C]TMZ synthesis and reformulation make it difficult for routine production. A highly reproducible one-pot radiosynthesis of [¹¹C]TMZ with a radiochemical yield of 17 ± 5% and ≥97% radiochemical purity is reported.

Full text of DOI:10.1021/ol302791x

ORGANIC
LETTERS
2012
Vol. 14, No. 23
5872–5875
An Efficient and Practical Radiosynthesis
of [11C]Temozolomide
Christian K. Moseley, Stephen M. Carlin, Ramesh Neelamegam, and Jacob M. Hooker*
Athinoula A. Martinos Center for Biomedical Imaging, Department of Radiology,
Massachusetts General Hospital, Harvard Medical School, Charlestown,
Massachusetts 02129, United States
hooker@nmr.mgh.harvard.edu
Received October 10, 2012
ABSTRACT
Temozolomide (TMZ) is a prodrug for an alkylating agent used for the treatment of malignant brain tumors. A positron emitting version, [11C]TMZ,
has been utilized to help elucidate the mechanism and biodistribution of TMZ. Challenges in [11C]TMZ synthesis and reformulation make it difficult
for routine production. A highly reproducible one-pot radiosynthesis of [11C]TMZ with a radiochemical yield of 17 ( 5% and g97% radiochemical
purity is reported.
Chemotherapeutic treatment for advanced cancers
persists as an area of intense research exploration. Exist-
ing treatment options, including surgery, radiation, and
chemotherapy, seldom result in even one year of median
improved survival for aggressive tumors such as malignant
glioma.1,2 Temozolomide (TMZ) is the prodrug to an anti-
neoplastic alkylating agent that has drawn interest among
the research community due to growing evidence of its
broad range of chemotherapeutic activity in both in vitro
and preclinical in vivo studies. TMZ displays activity against
tumors which have resistance to other antitumor agents
and shows potential to improve patient survival outcomes
when used in conjunction with radiation and antiangio-
genic therapies.1,3 TMZ was approved for clinical use in
the United States in 1999 and has since proven to be
effective as an oral treatment for malignant glioma, malig-
nant metastatic melanoma, and other adult central ner-
vous system tumors.
TMZ, de novo and acquired resistance to available chemo-
therapeutics, as well as complications from toxicity,
prevented any single drug from attaining much more than
a modest increase in overall survival rate.2
In comparison to other chemotherapies, TMZ is well-
tolerated by patients and elicits less severe complications
due to nonspecific toxicity. Its off-target toxicity is gen-
erally considered mild-to-moderate, and side effects are
predictable and easily managed in both adult and pediatric
patients, usually with the aid of antiemetics.2 Patients who
receive TMZ treatment also show improved Health-
Related Quality-Of-Life (HR-QOL) scores;a common
metric for cancer treatment efficacy assessed alongside side
effects;relative to similar antineoplastic alkylating agents.3
TMZ has physiochemical properties that account for its
broad spectrum of efficacy and low toxicity. TMZ dis-
tributes widely throughout all tissues and penetrates the
central nervous system, crossing the bloodꢀbrain barrier.
Its pharmacokinetic profile is predictably described by a
one-compartment open model, with dose-dependent linear
increases in both Area Under the Curve (AUC) and peak
plasma concentration (Cmax) values. TMZ is stable in
acidic conditions, and its oral bioavailability is approxi-
mately 100%. Following absorption into the intestine,
TMZ undergoes spontaneous nonenzymatic, pH-dependent
hydrolysis to generate 5-(3-methyl-(triazen-1-yl)-imidazole)-
4-carboxamide 2 (MTIC), the pharmacologically active
TMZ is currently the standard of care for all newly
diagnosed glioblastoma.4 Prior to the development of
(1) Johansson, F.; Ekman, S.; Blomquist, E.; Henriksson, R.;
€
Bergstrom, S.; Bergqvist, M. Anticancer Res. 2012, 32, 4001–4006.
(2) Friedman, H. S.; Kerby, T.; Calvert, H. Cancer Res. 2000, 6,
2585–2597.
(3) Darkes, M. J. M.; Plosker, G. L.; Jarvis, B. Am. J. Cancer 2002, 1,
55–80.
(4) Surawicz, T. S.; Davis, F.; Freels, S.; Laws, E. R.; Menck, H. R.
J. Neurooncol. 1998, 40, 151–160.
r
10.1021/ol302791x
Published on Web 11/14/2012
2012 American Chemical Society

Scheme 1. Hydrolysis of TMZ to MTIC
Scheme 2. Synthesis of [11C]TMZ via Cycloaddition of 8
and 9
[11C]MIC is not routinely produced at most PET radio-
tracer facilities. In conventional chemistry, the cycloaddi-
tion of a diazoimidazole precursor and methyl isocyanate
is slow, with the fastest reported syntheses taking as long
as one day using a >150-fold excess of MIC.11 The long
reaction time and high molar ratio of MIC are incompa-
tible with the short half-life of carbon-11 (t1/2 = 20.4 min)
and the substoichiometric amounts of [11C]MIC utilized in
radiochemistry. Illustrating this point, the Brown group
reported a 70% drop in chemical yield when optimizing
synthesis parameters for radiochemistry by raising the
reaction temperature and shortening the reaction time.
In addition, [11C]MIC is synthesized from [11C]CH3I 7
using a heated silver cyanate column; this added step
decreases the final radiochemical yield due to losses in-
curred during chemical transformation and radioactive
decay. In order to simplify the synthetic parameters and
reduce the total time required to synthesize [11C]TMZ, we
sought a new approach to circumvent the use of MIC.
Given the option to pursue either [3-N-11C-methyl]TMZ
or [4-11C-carbonyl]TMZ as our target tracer, we chose the
former for two reasons. First, the 3-N-methyl carbon is
incorporated into guanine residues during DNA methyla-
tion as demonstrated by Saleem et al.12 Radiolabeling
at the 3-N position is critical to [3-N-11C-methyl]TMZ func-
tion as a tumor imaging agent; synthesizing [4-11C-carbonyl]-
TMZ results in the loss of the carbon-11 radionuclide as
expelled [11C]CO2 6.12 Second, 3-N-methylation expands
the synthetic possibilities to include SN2 methylation of the
TMZ desmethyl analogue, nortemozolomide (norTMZ)
14, with [11C]CH3I.
alkylating agent. This reaction occurs through a base-
catalyzed nucleophilic attack by water. MTIC then
undergoes further hydrolysis in the presence of acid
into 5-aminoimidazole-4-carboxamide 3 (AIC) and a
methyldiazonium cation 4 (Scheme 1). This cation irrever-
sibly binds to DNA nucleotides via nucleophilic attack by
guanine residues, resulting in DNA alkylation. Accumula-
tion of methylated guanine residues leads to breaks in the
daughter DNA strand, causing cell cycle arrest and cellular
apoptosis.1,3
The mechanism of TMZ is well understood, making
[11C]TMZ a useful radiopharmaceutical to observe tumor
cells via transfer of the 3-N-11C-methyl group to tumor
cell DNA in vivo. Also, the tissue distribution and broad-
spectrum antitumor properties of TMZ provide utility
as a diagnostic and prognostic agent.5 Positron Emission
Tomography (PET) imaging has been used with [11C]TMZ
to confirm the drug’s therapeutic action and evaluate its
metabolic activity, pharmacokinetics, and biodistribu-
tion.6 However, challenges with reproducibly synthesizing
[11C]TMZ, as well as with formulating the final pro-
duct in an injectable solution that maintains its stability,
have limited its accessibility. The first radiochemical route
to [11C]TMZ 10, published by Brown et al.7 (Scheme 2),
utilized conventional synthetic chemistry methods initially
described by Wang et al.8 Alternative approaches to a
cycloaddition with MTIC 2 have been attempted using
1,10-carbonyldiimidazole, 4-nitrophenyl chloroformate,
and chloroformic acid trichloromethyl ester without
success.10 Although the Brown group successfully radio-
labeled [3-N-11C-methyl]TMZ7 10 using diazoimidazole
9 and [11C]MIC 8,9 a simpler synthetic method would
benefit other researchers hoping to access [11C]TMZ.
In order to access the 3-N-methylation route to [3-N-11C-
methyl]TMZ, we first synthesized norTMZ 14 according
to a published patent13 (Scheme 3). (Methods and
intermediates for the synthesis of 4-oxo-3,4-dihydro-
imidazo [5,1-d] [1,2,3,5] tetrazines. WO 2011107726,
2011. The authors of the patent state that the compound
(5) Neidle, S.; Thurston, D. E. Nat. Rev. Cancer 2005, 5, 285–296.
(6) Brock, C.; Matthews, J.; Brown, G.; Luthra, S.; Brady, F.;
Newlands, E.; Price, P. The Kinetic Behavior of Temozolomide in Man;
ASCO Annual Meeting; Philadelphia, PA, May 18ꢀ21, 1996; p 475.
(7) Brown, G. D.; Luthra, S. K.; Brock, C. S.; Stevens, M. F. G.;
Price, P. M.; Brady, F. J. Med. Chem. 2002, 45, 5448–5457.
(8) Wang, Y.; Stevens, M. F. G. Bioorg. Med. Chem. Lett. 1996, 6,
185–188.
(10) Wang, Y.; Stevens, M. F. G.; Chan, T.; DiBenedetto, D.; Ding,
Z.; Gala, D.; Hou, D.; Kugelman, M.; Leong, W.; Kuo, S. J. Org. Chem.
1997, 62, 7288–7294.
(11) Stevens, M. F. G.; Hickman, J. A.; Stone, R.; Gibson, N. W.;
Baig, G. U.; Lunt, E.; Newton, C. G. J. Med. Chem. 1984, 27, 196–201.
(12) Saleem, A.; Brown, G. D.; Brady, F.; Aboagye, E. O.; Osman, S.;
Luthra, S. K.; Ranicar, A. S. O.; Brock, C. S.; Stevens, M. F. G.;
Newlands, E. Cancer Res. 2003, 63, 2409.
(9) Brown, G. D.; Henderson, D.; Steel, C.; Luthra, S.; Price, P. M.;
Brady, F. Nucl. Med. Biol. 2001, 28, 991–998.
(13) Hummersone, M. G.; Cousin, D. WO 2011107726, 2011.
Org. Lett., Vol. 14, No. 23, 2012
5873

Scheme 3. Synthesis of Nortemozolomide and
[11C]Temozolomide
Figure 1. 13C NMR spectra for methylation of norTMZ with
[13C]CH3I. 13C NMR spectra (500 MHz, DMF-d7 peaks at
30 and 35 ppm) of (a) reference standard temozolomide (TMZ, 1);
the peak corresponding to 3-N-methyl appears at 36.1 ppm;
(b) reaction using >1.0 equiv of NaH (60% w/w) with 13CH3I at
0 °C, tR = 5 min; primary amide methylation at the 8-N position
corresponding to 25.5 ppm (pointed as arrow) was observed;
(c) using 1 equiv of NaH (60% w/w) with 13CH3I at 0 °C, tR
=
reported by Wang et al. (Bioorg. Med. Chem. Lett. 1996, 6,
2, 185ꢀ188) was not properly characterized and was thus
misinterpreted to be nortemozolomide.) Diazotization of
aminoimidazole 3 afforded diazoimidazole 9.14 In parallel,
N-(tert-butoxycarbonyl) glycine 11 was treated with triethyl-
amine and ethylchloroformate to give 12. Compound 12 was
immediately treated upon isolation with 9 to furnish com-
pound 13 at a total yield of 44% starting from compound 9.
Deprotection of 13 with 3 N HCl produced 14 as a pink solid
that is stable for at least six months when stored at 2ꢀ4 °C.
We anticipated two methylation sites on the norTMZ 14
molecule, at the 3-N heterocyclic position and the 8-N
primary carboxamide position. We expected 3-N-methyla-
tion to proceed through an anion that is more stable than
the primary amide anion produced by 8-N amide depro-
tonation. Synthesis conditions for 3-N-methylation of
norTMZ were screened using substoichiometric [13C]CH3I
in order to preoptimize for radiochemistry with [11C]CH3I.
We tested several bases including cesium carbonate, potas-
sium tert-butoxide, and NaH for their ability to selectively
deprotonate either the 3-N or8-N position ofnorTMZ. We
found that both cesium carbonate and potassium tert-
butoxide preferentially deprotonated the 8-N amide posi-
tion, resulting in undesirable 8-N-methylation product 15.
Although deprotonation can also occur at the 8-N
amide position with NaH, we found NaH most effectively
deprotonated the 3-N carboxamide. The amount of NaH
5 min; [3-N-13C-methyl]temozolomide formation at 36.1 ppm
and 8-N-methylated product formation at 26.4 ppm (specified by
arrow) were observed; (d) using <1.0 equiv of NaH (60% w/w)
with 13CH3I at 0 °C, tR = 5 min; [3-N-13C-methyl]temozolomide
formation at 35.9 ppm was observed. Excess [13C]CH3I observed
at ꢀ23.5 ppm.
used is critical in favoring methylation at the 3-N position
over methylation at the 8-N position. An excess of base is
often used in radiochemistry, and under these conditions
we observed only deprotonation at the 8-N carboxamide
position of 14. Use of 1 equiv of NaH at the same tempera-
ture resulted in the formation of the 8-N-[13C]methylated
product 15 and 3-N-methyl 16 in a 1:1 ratio. Selective 3-N-
methylation occurred using <1 equiv of NaH in the pre-
sence of [13C]CH3I at 45 °C in DMF to furnish [3-N-13C-
methyl]TMZ 16 as the major product (Figure 1). The
respective positions of 13C-labeling were confirmed by
13C NMR (Figure 1).
The optimized reaction conditions were subsequently
applied to carbon-11 radiochemistry. [11C]CO2 was pro-
duced through the 14N(p,R)11C nuclear reaction in an
Eclipse 11-MeV cyclotron (Siemens) and was reduced
to [11C]CH3I using a TRACERlab FX-MeI unit (GE
Healthcare). [11C]CH3I was trapped in a TRACERlab
FX-M reactor (GE Healthcare) preloaded with a solu-
tion containing excess norTMZ 14 and 0.7 equiv of NaH
(60% w/w dispersion in mineral oil) in dry DMF that
had been stirred at ꢀ5 °C for 1 min prior to trapping. The
(14) Shealy, Y. F.; Struck, R. F.; Holum, L. E. E. B.; Montgomery,
J. A. J. Org. Chem. 1961, 26, 2396–2401.
5874
Org. Lett., Vol. 14, No. 23, 2012

solution washeatedto45°C for 5 min, then cooled to room
temperature, and quenched with 0.5% aqueous AcOH.
The reaction mixture was purified by reversed-phase semi-
preparative HPLC and eluted with a mobile phase of 0.5%
AcOH in H2O/EtOH (95:5). The desired fraction was
collected, and aliquots were used to establish the chemical
and radiochemical purity by analytical HPLC. The iden-
tity of the product was confirmed by analytical HPLC with
additional coinjection of the TMZ reference standard
(Figure 2AꢀC) and withradio-thin-layer chromatography
(Figure 2D). Theaverage radiochemical yield was 17( 5%
(decay-corrected to trapped [11C]CH3I; n = 6). Chemical
and radiochemical purities were g97% in all instances.
Specific activity measurements by HPLC reached the limit
of detection but were at least 3 Ci/μmol at the end of syn-
thesis. The average time required for the synthesis from the
end of cyclotron bombardment to the end of synthesis and
purification was 30 min;an improvement over the pre-
viously reported synthesis time of 47 min using [11C]MIC.7
We observed the chemical hydrolysis of [3-N-11C-
methyl]TMZ in the final formulation according to analy-
tical HPLC, corroborating previous reports that the com-
pound decomposes at physiological pH.15 Our purification
and preparative method was designed to mitigate this
issue; we monitored the stability of the radiopharmaceu-
tical in an undiluted semipreparative HPLC fraction and
found it to be stable for at least 75 min in the acidic HPLC
mobile phase. The mobile phase is composed of reagents
that fall within USP safety limits for human injection; as
such, buffering and salinating the mobile phase in accor-
dance with these standards would afford a solution of
[11C]TMZ 10 that could immediately be filtered for injec-
tion without reformulation.
Figure 2. [11C]Temozolomide chromatography and radio TLC.
(A) Black and red lines correspond to UV (254 nm) and radio-
activity, respectively. Product peak collected between 5.2 and
7.0 min. (B) Aliquots of purified [11C]TMZ solution mixed with
TMZ reference standard were analyzed by analytical HPLC
(Column: Agilent Eclipse XDB-C18, 150 mm ꢁ 4.6 mm; mobile
phase: 0.1% TFA in H2O/CH3CN, 97:3; flow rate: 2 mL/min).
(C) UV spectra of [11C]TMZ (red) and coinjection of [11C]TMZ
solution mixed with TMZ reference standard (black). (D) Radio
TLC of purified [11C]TMZ (mobile phase: DCM/MeOH, 85:15).
analogue, norTMZ 14, using [11C]CH3I 7 to produce
[3-N-11C-methyl]TMZ10 in high chemical and radiochemical
purity. Our synthesis is practical, efficient, and reproduci-
ble and can be applied for oncological PET imaging in
humans.
TMZ is a front-line treatment for various types of
cancer, and its application asa radiopharmaceutical makes
it a novel ligand for imaging tumor cells in vivo. In this
paper, we report a direct methylation of its 3-N-desmethyl
Supporting Information Available. The experimental
procedure and compound characterization data. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(15) Denny, B. J.; Wheelhouse, R. T.; Stevens, M. F. G.; Tsang,
L. L. H.; Slack, J. A. Biochemistry 1994, 33, 9045–9051.
The authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 23, 2012
5875