Alternative synthesis for the preparation of 16α-[ 18F]fluoroestradiol

Article abstract of DOI:10.1002/jlcr.3076

We have developed a new precursor, 3,17β-O-bis(methoxymethyl)- 16β-O-p-nitrobenzenesulfonylestriol (14c) of 16α-[ ¹⁸F]fluoroestradiol ([¹⁸F]FES). Although we could not selectively protect the C17 alcohol in the presence of the C16 al

Full text of DOI:10.1002/jlcr.3076

Research Article
Received 13 February 2013,
Revised 16 May 2013,
Accepted 19 May 2013
Published online 9 July 2013 in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/jlcr.3076
Alternative synthesis for the preparation
18
of 16a-[ F]fluoroestradiol
Hee Seup Kil,c Han Yang Cho,a Sang Ju Lee,a,b Seung Jun Oh,b
and Dae Yoon Chia,c
*
We have developed a new precursor, 3,17b-O-bis(methoxymethyl)-16b-O-p-nitrobenzenesulfonylestriol (14c) of 16a-[18F]
fluoroestradiol ([18F]FES). Although we could not selectively protect the C17 alcohol in the presence of the C16 alcohol,
we were able to prepare and chromatographically isolate the desired C16 TBDMS, C17,C3-dimethoxymethyl (diMOM)
protected estriol derivative and convert into the ultimate fluorination precursor. The MOM protective group proved to be
more quickly removed than the cyclic sulfate group. The di-MOM protective precursor at the C3 and C17 alcohols instead
of a cyclic sulfate group shortened hydrolysis time. We prepared three different sulfonate precursors at C16 alcohol. After
checking their reactivity in the [18F]fluorination step and considering the stability of the precursors, we obtained the best
results with nosylate precursor 14c.
Keywords: 16a-[18F]Fluoroestradiol; [18F]FES; imaging of estrogen receptor; PET; radiopharmaceutical
azeotropic distillation in an open system to increase concentration
Introduction
of reagents. This hydrolysis method cannot be conveniently
16a-[18F]Fluoroestradiol ([18F]FES, Figure 1) is a promising
positron emission tomography (PET) radiopharmaceutical for
imaging estrogen receptors (ER) in breast cancer.1–7 The
identification of the ER status (positive or negative) is very
important in treating breast cancers. In clinical practice, in vitro
ER assays are usually used to predict tumor response to
hormonal therapy and patient prognosis.8–11 The high binding
affinity of [18F]FES for ER provides a mechanism for the selective
accumulation within ER positive breast cancers. PET imaging of
ER is a noninvasive alternative to in vitro assays.4–6,12–14 In this
aspect, the efficient preparation of [18F]FES is very important.
Two synthetic methods for 16a-[18F]fluoroestradiol have been
reported (Scheme 1).
applied to an automatic module for routine production.25 In this
paper, we report a new synthetic pathway that is applicable to
the automated production of [18F]FES.
Experimental
General procedures
All chemicals were purchased from commercial suppliers (Sigma-Aldrich,
TCI, Acros) and used without further purification. Analytical TLC was
carried out on a pre-coated plate (Merck, silica gel 60 F254). Flash column
chromatography was performed with a silica gel (Merck, 230–400 mesh,
ASTM). 1H and 13C NMR spectra were recorded on a Varian 400-MR
(400 MHz, Agilent Technologies, USA) spectrometer at ambient temperature.
16a-[18F]Fluoroestradiol was first developed by the
Katzenellenbogen group in 1984 from 3,16b-bistrifluoromethanesulfonate
of 16b-hydroxyestrone precursor 2.15,16 After [18F]fluorination, a
reduction step using LiAlH4 is required. In this reduction step,
17a-epimeric alcohol is produced as a byproduct and must be
separated.17–20 A second method was developed by Lim et al.21
to preclude formation of 17-epimers using 3-methoxymethyl-
l6b,17b-epistriol-O-cyclic sulfate (3). As this procedure does not
lead to epimerization under radiolabeling conditions, this route
is most commonly used among the several available precursors.
Romer et al., Liang et al., Knott et al., and our group also reported
manual syntheses and automatic production of [18F]FES using the
cyclic sulfate precursor 3.22–25 [18F]FES was prepared in two steps
– fluorination and acid hydrolysis of the sulfate. Although the
fluorination step resulted in high radiolabeling yield, some
difficulties were experienced in the acid hydrolysis by triple
azeotropic distillation methods using 2 N HCl and acetonitrile. In
closing reaction vial with 2 N HCl, the hydrolysis did not occur. It
took more than 30 min by evaporation of solvent with triple
3,16b-O-Bis(benzyloxycarbonyl)estrone (9)
16b-Hydroxyestrone26 (7, 1.00 g, 3.49 mmol) was dissolved in dry THF
(10 mL). TMEDA (0.63 mL, 4.20 mmol) was added, and the solution was
cooled to 0 ꢀC. Benzyl chloroformate (1.09 mL, 7.64 mmol) was added,
and the cooled mixture was stirred for 2 h. The reaction was quenched
aDepartment of Chemistry, Sogang University, 35 Baekbeomro Mapogu, Seoul
121-742, Korea
bDepartment of Nuclear Medicine, Asan Medical Center, University of Ulsan
College of Medicine, 388-1 Pungnap-2-dong Songpagu, Seoul 138-736, Korea
cResearch Institute of Labeling, FutureChem Co. Ltd., 388-1 Pungnap-2-dong,
Songpagu, Seoul 138-736, Korea
*Correspondence to: Prof. Dae Yoon Chi, Department of Chemistry, Sogang
University, 35 Baekbeomro Mapogu, Seoul 121-742, Korea.
E-mail: dychi@sogang.ac.kr
J. Label Compd. Radiopharm 2013, 56 619–626
Copyright © 2013 John Wiley & Sons, Ltd.

H. S. Kil et al.
mixture was heated at 60 ꢀC for 5 h. Ethyl acetate (50 mL) was added, and
the solution was washed with saturated NaHCO3 (40mL). The organic layer
was dried over Na2SO4 and evaporated to dryness in vacuo. Purification by
flash column chromatography (5% ethyl acetate/hexane) afforded products
12a (1.2g, 51% from 4) and 12b (0.51g, 22% from 4) as a yellowish oil.
(12a): 1H NMR (400 MHz, CDCl3) d 0.04 (s, 3H), 0.05 (s, 3H), 0.89 (s, 9H), 0.94 (s,
3H), 1.23–1.38 (m, 3H), 1.44–1.58 (m, 2H), 1.82–1.91 (m, 1H), 1.98 (dt, J = 12.4,
3.2 Hz, 1H), 2.14–2.32 (m, 4H), 2.77–2.93 (m, 2H), 3.34–3.42 (m, 1H), 3.40
(s, 3H), 3.47 (s, 3H), 4.24–4.30 (m, 1H), 4.67 (ABq, J = 6.6Hz, 2H), 5.14 (s, 2H),
6.76 (d, J = 2.8Hz, 1H), 6.82 (dd, J = 8.4, 2.8Hz, 1H), 7.20 (d, J= 8.8Hz, 1H); 13
C
NMR (100MHz, CDCl3) d À4.7, À4.0, 13.1, 18.3, 26.1, 26.3, 27.6, 30.1, 37.1,
38.0, 38.3, 42.7, 44.5, 47.3, 55.4, 56.2, 69.5, 85.3, 94.7, 95.2, 113.9, 116.4, 126.5,
134.1, 138.1, 155.2; MS FAB+ m/z 491.3 (M + H+). HRMS FAB+ calcd for
C28H47O5Si (M + H+) 491.3193 (M+ H+), found, 491.3184. (12b): 1H NMR
(200MHz, CDCl3) d 0.08 (s, 3H), 0.09 (s, 3H), 0.88 (s, 3H), 0.93 (s, 9H), 1.12–1.64
(m, 5H), 1.80–1.96 (m, 2H), 2.10–2.37 (m, 4H), 2.80–2.91 (m, 2H), 3.37 (s, 3H),
3.42–3.55 (m, 1H), 3.47 (s, 3H), 3.90–4.04 (m, 1H), 4.67 (s, 2H), 5.14 (s, 2H),
6.77 (d, J = 2.6Hz, 1H), 6.83 (dd, J = 8.5, 2.5Hz, 1H), 7.20 (d, J= 8.4Hz, 1H); 13C
NMR (50 MHz, CDCl3) d À4.4, À4.3, 12.2, 18.4, 26.0, 26.2, 27.4, 29.9, 34.0, 37.7,
38.3, 43.3, 44.3, 46.6, 55.6, 56.0, 76.8, 82.7, 94.5, 97.0, 113.8, 116.3, 126.4,
134.0, 138.1, 155.1; MS FAB+ m/z 491.3 (M + H+). HRMS FAB+ calcd for
Scheme 1. Two synthetic methods for 16a-[18F]fluoroestradiol.
Figure 1. Structure of 16a-[18F]fluoroestradiol ([18F]FES, [18F]1).
with 10 mL of water. The reaction mixture was extracted with
dichloromethane (3Â 20 mL). The organic layer was dried over Na2SO4
and evaporated to dryness in vacuo. Purification by flash column
chromatography (20% ethyl acetate/hexane) afforded product 9 (1.6 g,
81%) as a foamy colorless solid: 1H NMR (400MHz, CDCl3) d 0.86 (s, 3H),
1.43–1.87 (m, 6H), 1.93–2.05 (m, 2H), 2.35–2.49 (m, 3H), 2.85–2.92
(m, 2H), 4.92 (s, 1H), 5.22 (s, 2H), 5.25 (s, 2H), 6.90 (d, J= 2.4Hz, 1H), 6.95
(dd, J= 8.4, 2.4Hz, 1H), 7.27 (d, J= 8.4Hz, 1H), 7.33–7.46 (m, 10H); 13C NMR
(100MHz, CDCl3) d 12.2, 25.6, 27.3, 29.3, 35.6, 36.2, 37.1, 41.8, 43.8, 44.1,
70.1, 70.2, 88.6, 118.2, 121.0, 126.1, 128.2, 128.4, 128.46, 128.52, 128.6,
134.7, 134.8, 137.1, 137.7, 149.0, 153.7, 154.7, 209.3; MS FAB+ m/z 555.2
(M+ H+). HRMS FAB+ calcd for C34H35O7 (M+ H+) 555.2383, found 555.2379.
C
28H47O5Si (M + H+) 491.3193, found 491.3182.
3,17b-O-Bis(methoxymethyl)-16b-estriol (13)
3,17b-O-Bis(methoxymethyl)-16b-O-t-butyldimethylsilylestriol (12a,
0.640 g, 1.30 mmol) was dissolved in THF (5 mL). Tetrabutylammonium
fluoride (TBAF) hydrate (0.44 g, 1.69 mmol) was added, and the reaction
mixture was stirred for 18 h at 80 ꢀC. Dichloromethane (20 mL) was added,
and the solution was washed with water (20 mL). The organic layer was dried
over Na2SO4 and evaporated to dryness in vacuo. Purification by flash
column chromatography (30% ethyl acetate/hexane) afforded product 13
(0.48 g, 98%) as a yellowish oil: 1H NMR (400 MHz, CDCl3) d 0.92 (s, 3H),
1.02–1.11 (m, 1H), 1.28–1.46 (m, 3H), 1.47–1.69 (m, 2H), 1.86–1.93 (m, 1H),
1.98 (dt, J= 12.0, 3.2 Hz, 1H), 2.16–2.24 (m, 1H), 2.25–2.32 (m, 2H), 2.75–2.81
(m, 1H), 2.82–2.90 (m,. 2H), 3.41 (d, J= 7.2 Hz, 1H), 3.44 (s, 3H), 4.47 (s, 3H),
4.18–4.24 (m, 1H), 4.75 (q, J= 6.4 Hz, 2H), 5.14 (s, 2H), 6.77 (d, J= 2.8 Hz, 1H),
6.83 (dd, J= 8.4, 2.8 Hz, 1H), 7.19 (d, J= 8.4 Hz, 1H); 13C NMR (100 MHz, CDCl3)
d 13.1, 26.3, 27.6, 30.0, 34.7, 37.9, 38.3, 42.9, 44.4, 47.2, 56.0, 56.2, 69.6, 87.6,
94.7, 97.2, 113.9, 116.4, 126.4, 133.9, 138.1, 155.2; MS FAB+ m/z 377.2
(M + H+). HRMS FAB+ calcd for C22H33O5 (M + H+) 377.2328; found 377.2318.
3-O-Benzyloxycarbonyl-16b,17b-estriol (10)
3,16b-O-Bis(benzyloxycarbonyl)estrone (9, 100 mg, 0.18 mmol) was
dissolved in dry THF (4 mL), followed by the addition of lithium tri-t-
butoxyalumminium hydride (0.56 mL, 0.595 mmol) under an argon
atmosphere. The reaction mixture was stirred for 1 h at room
temperature. The reaction was quenched with ethyl acetate followed
by the addition of 1N HCl (2.5 mL) and ice water (5 mL). The aqueous
layer was extracted with ethyl acetate (3Â 10 mL). The organic extracts
were dried over Na2SO4 and evaporated to dryness in vacuo. Purification
by flash column chromatography (25% ethyl acetate/hexane) afforded
product 10 (50mg, 66%) as a foamy colorless solid: 1H NMR (400 MHz,
CDCl3) d 0.83 (s, 3H), 0.96–1.07 (m, 1H), 1.22–1.40 (m, 3H), 1.45–1.60 (m,
2H), 1.84–1.94 (m, 1H), 1.99 (dt, J = 12.4, 3.2 Hz, 1H) 2.35–2.17 (m, 3H), 2.58
(bs, 2H), 2.79–2.95 (m, 2H), 3.45 (d, J = 7.6Hz, 1H), 4.20 (td, J = 12.6, 7.6Hz,
1H), 5.25 (s, 2H), 6.87 (d, J= 2.4Hz, 1H), 6.92 (dd, J= 8.4, 2.8Hz, 1H), 7.27
(d, J = 8.4Hz, 1H), 7.32–7.46 (m, 5H); 13C NMR (100 MHz, CDCl3) d 12.1,
26.1, 27.4, 29.7, 35.0, 37.5, 38.0, 43.0, 44.5, 46.7, 70.3, 70.5, 81.0, 118.3,
121.2, 126.6, 128.8, 128.9, 129.0, 135.0, 138.36, 138.44, 149.1, 154.2; MS (EI
+) m/z 404 (M+-H2O). HRMS EI calcd for C26H28O4 (M+-H2O) 404.1977, found
404.1988.
3,17b-O-Bis(methoxymethyl)-16b-O-methanesulfonylestriol (14a)
3,17b-O-Bis(methoxymethyl)-16b-estriol (13, 51 mg, 0.13 mmol) was
dissolved in pyridine (0.5 mL) and cooled in an ice bath under an argon
atmosphere. Methanesulfonyl anhydride (47 mg, 0.27 mmol) in pyridine
(0.5 mL) was added, and the reaction mixture was stirred for 2 h at 0 ꢀC.
Ethyl acetate (10 mL) was added, and the solution was washed with 1N
HCl (2Â 10 mL) and saturated NaHCO3 (10 mL). The organic layer was
dried over Na2SO4 and evaporated to dryness in vacuo. Purification by
flash column chromatography (30% ethyl acetate/hexane) afforded
product 14a (58 mg, 95%) as a colorless oil: 1H NMR (400 MHz, CDCl3) d
0.92 (s, 3H), 1.03–1.12 (m, 1H), 1.28–1.42 (m, 2H), 1.47–1.58 (m, 2H), 1.70
(td, J = 13.6, 4.4 Hz, 1H), 1.81–1.89 (m, 1H), 1.99 (dt, J = 3.2, 2.4 Hz, 1H),
2.21 (td, J = 11.2, 4.4 Hz, 1H), 2.26–2.34 (m, 1H), 2.40–2.49 (m, 1H), 2.82–2.88
(m, 2H), 3.04 (s, 3H), 3.42 (s, 3H), 3.46 (s, 3H), 3.58 (d, J = 7.2 Hz, 1H), 4.71
(ABq, J = 6.6, Hz, 2H), 5.13 (s, 2H), 5.06–5.18 (m, 1H), 6.76 (d, J= 2.8 Hz, 1H),
6.82 (dd, J= 8.4, 2.8Hz, 1H), 7.18 (d, J = 8.8Hz, 1H); 13C NMR (100 MHz, CDCl3)
d 12.6, 26.2, 27.5, 29.9, 33.6, 37.4, 38.1, 39.0, 42.9, 44.3, 47.2, 56.0, 56.2, 78.7,
84.6, 94.7, 96.3, 114.0, 116.5, 126.4, 133.5, 137.9, 155.3; MS FAB+ m/z 455.2
(M+ H+). HRMS FAB+ calcdforC23H35O7S(M+ H+) 455.2103; found 455.2117.
3,17b-O-Bis(methoxymethyl)-16b-O-t-butyldimethylsilylestriol (12a)
273-O-Methoxymethyl-16b,17b-estriol (4, 1.59 g, 4.78 mmol), N,N-
dimethylaminopyridine (0.580 g, 4.78 mmol), and t-butyldimethylsilyl chloride
(0.870 g, 5.74 mmol) were dissolved in anhydrous dimethylformamide (DMF)
(16 mL). Triethylamine (3.4 mL, 24.0 mmol) was added, and the reaction
mixture was stirred for 24 h at room temperature. Ethyl acetate (40 mL) was
added and the solution was washed with 1N HCl (2Â 50 mL) and saturated
NaHCO3 (50 mL). The organic layer was dried over Na2SO4 and evaporated
3,17b-O-Bis(methoxymethyl)-16b-O-p-toluenesulfonylestriol (14b)
to dryness in vacuo. Purification by flash column chromatography (10% ethyl 3,17b-O-Bis(methoxymethyl)-16b-O-p-toluenesulfonylestriol was
acetate/hexane) afforded 11a with undesired byproduct 11b (1.96 g, 92%, as
a mixture of 11a and 11b). The mixture was dissolved in anhydrous DMF
(20 mL), followed by the addition of N,N-diisopropylethylamine (3.96 mL,
23.9 mmol) and methoxymethyl chloride (1.82 mL, 23.9 mmol). The reaction
prepared by similar method for 14a: 87% yield as a colorless oil; 1H NMR
(400 MHz, CDCl3) d 0.90 (s, 3H), 0.96–1.06 (m, 1H), 1.26–1.37 (m, 2H), 1.43–
1.53 (m, 2H), 1.54–1.63 (m, 1H), 1.74–1.82 (m, 1H), 1.96 (dt, J= 12.0, 3.6 Hz,
1H), 2.13–2.33 (m, 3H), 2.45 (s, 3H), 2.79–2.87 (m, 2H), 3.33 (s, 3H), 3.46 (s,
www.jlcr.org
Copyright © 2013 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2013, 56 619–626

H. S. Kil et al.
3H), 3.51 (d, J= 7.2 Hz, 1H), 4.51 (s, 2H), 4.92–4.98 (m, 1H), 5.13 (s, 2H), 6.76 (d, (2N, 250mL) was added for neutralization, and 200 mL of citrate buffer was
J= 2.8 Hz, 1H), 6.82 (dd, J= 8.4, 2.8 Hz, 1H), 7.17 (d, J= 8.4 Hz, 1H), 7.33 (d,
added. This solution was injected into the high-performance liquid
J = 8.4 Hz, 2H), 7.81 (d, J = 8.4 Hz, 2H); 13C NMR (100 MHz, CDCl3) d chromatography (HPLC) system for purification. HPLC purification was
12.6, 22.0, 26.2, 27.5, 29.9, 33.7, 37.3, 38.0, 42.9, 44.3, 47.2, 55.7,
56.2, 78.3, 78.4, 83.9, 94.7, 95.6, 104.0, 114.0, 116.4, 126.4, 127.9,
129.9, 133.6, 134.6, 137.9, 144.6, 155.3; MS FAB+ m/z 531.2 (M + H+).
HRMS FAB+ calcd for C29H39O7S (M + H+) 531.2416; found 531.2394.
conducted with a Nucleosil 100-7 C18 250Â 16 semi-prep column
(Macherey-Nagel) eluted at 5 mL/min with 30% EtOH:40% water:30%
acetonitrile. Purified 16a-[18F]fluoroestradiol ([18F]1) was eluted. We
obtained a decay-corrected radiochemical yield (RCY) of 19–24% (n = 3)
after HPLC purification for manual synthesis. Total synthesis time, including
purification time, was about 36 min.
3,17b-O-Bis(methoxymethyl)-16b-O-p-nitrobenzenesulfonylestriol
(14c)
3,17b-O-Bis(methoxymethyl)-16b-O-p-nitrobenzenesulfonylestriol
was prepared by similar method for 14a: 94% as a colorless oil: 1H
NMR (400 MHz, CDCl3) d 0.88 (s, 3H), 0.99–1.08 (m, 1H), 1.27–1.38
(m, 2H), 1.43–1.53 (m, 2H), 1.54–1.64 (m, 1H), 1.74–1.82 (m, 1H),
1.96 (dt, J = 12.4, 3.6 Hz, 1H), 2.17 (td, J = 11.6, 4.4 Hz, 1H), 2.26–2.36
(m, 2H), 2.78–2.86 (m, 2H), 3.33 (s, 3H), 3.46 (s, 3H), 3.54 (d, J = 7.2 Hz,
1H), 4.54 (ABq, J = 6.4 Hz, 2H), 5.07–5.18 (m, 1H), 5.13 (s, 2H), 6.75 (d,
J = 2.4 Hz, 1H), 6.81 (dd, J = 8.8, 2.4 Hz, 1H), 7.16 (d, J = 8.8, 1H), 8.12
(d, J = 8.8 Hz, 2H), 8.38 (d, J = 8.8 Hz, 2H); 13C NMR (100 MHz, CDCl3) d
12.5, 26.0, 27.4, 29.7, 33.4, 37.2, 37.9, 42.8, 44.2, 47.1, 55.8, 56.1, 80.5,
84.2, 94.7, 96.0, 114.1, 116.5, 124.5, 126.5, 129.2, 133.5, 137.9, 143.6,
150.7, 155.4; MS FAB+ m/z 562.2 (M + H+). HRMS FAB+ calcd for
Results and discussion
Although we have used the second synthetic route from cyclic
sulfate precursor 3 using an automatic module, we faced two
difficulties in routine production. One is an HPLC separation
problem, and the other is an increase of preparation time from
three repeated processes in the hydrolysis step of the sulfate.
These two difficulties motivated us to develop a new preparation
route of [18F]FES. The basic idea of the new strategy is shown as
a retrosynthetic route in Scheme 2. If we can selectively protect
C3 and C17b-alcohol groups from 16b-epistriol, we can activate
the 16b-alcohol group by suitable leaving group such as
sulfonate to introduce [18F]fluoride. The protective groups on
C3 and C17b should be easily removed after the labeling step.
We tried to introduce a more convenient protective group on
C17 alcohol. As protective groups, benzyl, benzylcarbonate
(Cbz), and methoxymethyl (MOM) groups were chosen.
C
28H36O9NS (M + H+) 562.2111; found 562.2087.
16a-Fluoroestradiol (1)
3,17b-O-Bis(methoxymethyl)-16b-O-p-nitrobenzenesulfonylestriol 14c
(11 mg, 0.02 mmol) was dissolved in acetonitrile (0.1 mL) followed by
the addition of tert-amyl alcohol (0.5 mL) and n-tetrabutylammonium
fluoride (26 mg, 0.1 mmol). The reaction mixture was stirred for 1 h at
120 ꢀC. The reaction solvents were evaporated to dryness in vacuo.
Aqueous hydrochloric acid (2N, 1 mL) was added, and the reaction
mixture was heated at 120 ꢀC for 15 min. Dichloromethane (2 mL)
was added, and the solution was washed with water (2 mL). The
organic layer was dried over Na2SO4 and evaporated to dryness in
vacuo. Purification by flash column chromatography (3% methanol/
dichloromethane) afforded product 1 (3.0mg, 52%) as a white solid: 1H
NMR (400 MHz, CDCl3) d 0.80 (s, 3H), 1.34–1.49 (m, 2H), 1.58 (s, 2H), 1.60–
1.68 (m, 1H), 1.81–1.97 (m, 5H), 2.20–2.34 (m, 2H), 2.79–2.86 (m, 2H), 3.86 (d,
J = 28.8 Hz, 1H), 4.68 (s, 1H), 4.86–4.93 and 5.00–5.06 (dm, J = 52.0Hz, 1H),
6.56 (d, J = 2.4Hz, 1H), 6.63 (dd, J= 8.4, 2.8Hz, 1H), 7.14 (d, J= 8.4Hz, 1H);
13C NMR (100 MHz, CD3OD) d 13.0, 27.3, 28.7, 30.8, 32.8 (d, J = 23.5 Hz, C15),
38.0, 40.0, 45.3, 45.4, 49.2, 88.6 (d, J = 22.0Hz, C17), 101.5 (d, J = 177.3Hz,
C16-F), 113.8, 116.1, 127.1, 132.3, 138.7, 156.0; Registry No. 92817-10-2.
Synthetic trials of the selective protection of C16 alcohol or
C17 alcohol
Our first attempt to synthesize selective protective estriol of either
C16 alcohol or C17 alcohol of 16b-epistriol is shown in Scheme 3.
Stein et al. reported a selective formyl protecting method of C16
alcohol in dl-13-ethylgona-1,3,5(10)-triene-3,16b,17b-triol under a
USA patent in 1972.28 We planned to introduce the selective
formylation of C16 alcohol and a MOM protecting group of C17b
alcohol in 3-O-methoxymethyl-16b,17b-estriol (4). Unfortunately,
the formylation reaction of the diol 4 gave a 1:1 mixture of
formylated products of C16b alcohol and C17b alcohol, which
was confirmed by NMR spectroscopy.
To make 3-O-protected-16b-protected-17b-estriol, we also
attempted to protect C16b alcohol from 3-O-protected-16b-
hydroxyestrone, followed by reduction of the ketone group.
The starting material, 3-O-protected-16b-hydroxyestrone, was
prepared from 3,16b-dihydroxyestrone (7), which was obtained
from 3,16b-O-diacetylestrone (6) by treatment with K2CO3 in
good yield.27 Dibenzylation of 3,16b-dihydroxyestrone (7) was
carried out with excess amounts (10 equiv) of benzyl bromide
and K2CO3 in acetone at 50 ꢀC overnight. Unfortunately, we
obtained only monobenzylation product 8 of C3 alcohol in
3,16b-dihydroxyestrone (7). This result is most likely because of
16a-[18F]Fluoroestradiol
[
18F]FÀ/H128O was trapped on a Chromafix PS-HCO3 cartridge (Machery-
Nagel, Germany), and [18F]FÀ was eluted with 600mL of stock solution
(0.1M potassium mesylate (KOMs) in H2O (250 mL) and 22mg of K222 in
methanol (1.0mL)). The radioactivity was dried with 1 mLÂ 3 times of
acetonitrile by the azeotropic-distillation method. After complete drying,
the nosylate precursor (14c, 2 mg, 3.67 mmol) in 100 mL of acetonitrile and
500 mL of t-amyl alcohol were added to the reactor vial. [18F]Fluorination
was carried out at 120 ꢀC for 20 min. The reaction mixture was evaporated
and dissolved in acetonitrile (100mL). Aqueous hydrochloric acid (1N,
500 mL) was added to the reactor vial, and then the reactor was heated at
120 ꢀC for 3min for hydrolysis. After cooling, aqueous sodium hydroxide low reactivity of the secondary C16b alcohol position. To
Scheme 2. Retrosynthetic route.
J. Label Compd. Radiopharm 2013, 56 619–626
Copyright © 2013 John Wiley & Sons, Ltd.
www.jlcr.org

H. S. Kil et al.
Scheme 3. Failed reactions of selective protection at C17 position.
overcome the low reactivity of the benzyl group (alkylation (estriol) in 1987.30 We conducted the selective silylation reaction
reaction), we introduced the Cbz group (acylation reaction). This using various bulky silyl ethers. 3-O-Methoxymethyl-16b-O-t-
group could also readily be removed by hydrogenation using butyldiphenylsilylestriol was prepared from 3-O-methoxymethyl-
Pd/C. 3,16b-Dibenzyloxycarbonylestrone (di-Cbz-estrone, 9) was 16b,17b-estriol (4) and t-butyldiphenylsilyl chloride (TBDPSCl)
prepared in 81% yield from 3,16b-dihydroxyestrone (7) and in good yield. Unfortunately, the 17b-methoxymethyl derivative
benzyl chloroformate (2.2 equiv). Reduction of the C17 keto of 3-O-methoxymethyl-16b-O-t-butyldiphenylsilylestriol
group of di-Cbz-estrone (9) with lithium tri-t-butoxy-aluminum could not be prepared from 3-O-methoxymethyl-16b-t-
hydride (Li(t-BuO)3AlH) provided 3-O-benzyloxycarbonyl-16b,17b- butyldiphenylsilylestriol. The 17b alcohol was located in the
estriol (10), an unexpected product. Generally, the Cbz protected steric hindered position, and hence, alkylation at 17b alcohol
functionality is stable under the Li(t-BuO)3AlH hydride reaction using MOM chloride did not proceed well.
conditions.29
We also tried the selective silylation reaction using the t-
Laurent et al. reported a selective silyl ether protection butyldimethylsilyl (TBDMS) group, which is less bulky than the
method of C16 alcohol in 1,3,5(10)-estratriene-3,16a,17b-diol TBDPS group. The introduction of the TBDMS group could be
Scheme 4. Optimized synthetic route for 16a-fluoroestradiol (1).
www.jlcr.org
Copyright © 2013 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2013, 56 619–626

H. S. Kil et al.
Table 1. [18F]Fluorination using the FES nosylate precursor 14c under various conditions
R-TLC yield (%)b
18F in
solution
(%) at 20
minc
Base amount
RCY
(%)
Entry
Solvent(s)a
Base
TBAOH
TBAOH
CsOH
CsOH
KOH
(mmol)
10 min
20 min
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
t-amyl alcohol
t-amyl alcohol
t-amyl alcohol
t-amyl alcohol
t-amyl alcohol
t-amyl alcohol
t-amyl alcohol
t-amyl alcohol
t-amyl alcohol
t-amyl alcohol
t-amyl alcohol
t-amyl alcohol
acetonitrile
4 mL (6)
8 mL (12)
1.3 mL (6)
31.3
17.7
9.9
51.4
21.9
16.7
0.8
50.3
15.8
25.8
5.8
28.0
10.7
23.1
0
45.3
32.0
34.5
56.5
80.0
92.1
90.9
91.8
84.7
95.3
94.7
96.6
95.0
95.8
95.4
96.6
60.4
78.2
66.7
78.0
41.1
20.2
15.2
0.7
42.6
15.1
24.4
5.6
26.2
10.3
22.0
0.0
27.4
25.0
23.0
44.1
2.6 mL (12)
0.35 mg (6)
0.70 mg (12)
4.7 mL (6)
9.4 mL (12)
0.62 mg (6)
1.24 mg (12)
0.85 mg (6)
1.70 mg (12)
4 mL (6)
0.5
37.1
14.8
21.7
4.8
28.1
10.4
22.3
1.3
47.6
26.5
25.4
23.4
KOH
TBAHCO3
TBAHCO3
KHCO3
KHCO3
K2CO3
K2CO3
TBAOH
KOH
acetonitrile
0.35 mg (6)
2 mL (3)
t-amyl alcohol + acetonitrilec
t-amyl alcohol
TBAOH
TBAOH
2 mL (3)
FES, fluoroestradiol; RCY, radiochemical yield.
aNosylate precursor 14c (2 mg) in acetonitrile (100 mL) was added to the reactor, and then t-amyl alcohol (500 mL) was added.
bRadio-TLC yield, 30% ethyl acetate/hexane elution.
ct-Amyl alcohol (300 mL) and acetonitrile (300 mL) were used.
accomplished by the same method used to introduce the TBDPS month, we optimized fluorination reaction using nosylate
group. Although the silylated product was formed in high yield precursor. The nosylate precursor 14c was reacted with TBAF in
(92%) and shown as a single spot on TLC, it proved to be a acetonitrile and t-amyl alcohol. After removal of the solvents,
mixture of regioisomers based on 1H NMR, and HPLC analysis the resulting MOM-protected fluoroestradiol was hydrolyzed
data (11a:11b = 7:3). These mixtures could not be separated by under acidic conditions (2N HCl) to give the non-radioactive
normal phase silica gel chromatography. However, we found 16a-fluoroestradiol (1). Scheme 4 is an optimized synthetic route
that MOM-protected 12a and 12b derivatives, prepared from a for 16a-fluoroestradiol (1).
mixture of 11a and 11b, could be separated by silica gel column
chromatography to provide 12a and 12b in 51% and 22% yield,
respectively, from 4. After the preparation of 3,17b-O-bis
(methoxymethyl)-16b-O-t-butyldimethylsilylestriol (12a), the
16b-TBDMS group was removed by treatment of TBAF in
refluxing THF for 18 h.
3,17b-O-Bis(methoxymethyl)-16b-estriol (13) was synthesized
as a key intermediate. From this intermediate, three precursors
can be synthesized with methanesulfonyl anhydride (for 14a)
or p-toluenesulfonyl chloride (for 14b) or p-nitrobenzenesulfonyl
chloride (for 14c).
Using the nosylate precursor 14c, we tested the fluorination
step and hydrolysis reaction step using fluorine-19. The cold
and hot fluorination adapted our fluorination method31 using
protic solvents to the new FES precursor. Among the three
precursors, as we expect, the nosylate precursor gave best result.
As it is quite stable at a room temperature for more than a
Figure 2. Typical radio thin-layer chromatography chromatogram of [18F]15 using
30% ethyl acetate/hexane as an elution solvent.
J. Label Compd. Radiopharm 2013, 56 619–626
Copyright © 2013 John Wiley & Sons, Ltd.
www.jlcr.org

H. S. Kil et al.
Scheme 5. Synthesis of [18F]FES using Nosylate precursor 14c.
Figure 3. High-performance liquid chromatography chromatogram of 16a-[18F]fluoroestradiol ([18F]1) at the end of manual synthesis. The upper radioactive (RA)
chromatogram shows a simple peak at retention times of 13–14 min corresponding to 16a-[18F]fluoroestradiol ([18F]1) with a Nucleosil 100-7 C18 250Â 16mm semi-prep
column (Macherey-Nagel, Germany) eluted at 5 mL/min with 30% EtOH:40% water:30% acetonitrile.
Optimization of the [18F]fluorination conditions
of t-amyl alcohol were added to the reactor vial. Also, acetonitrile
and 50% acetonitrile in t-amyl alcohol as a mixing reaction solvent
Generally, an aliphatic nucleophilic substitution (SN2) reaction is
was tested to check the solvent effect. A crude radio-TLC
used with aprotic polar solvents such as acetonitrile, DMF, and
chromatogram of the reaction mixture is shown in Figure 2.
dimethyl sulfoxide. Recently, our group developed an efficient
[18F]Fluorination was first performed according to a previously
fluorination method for the aliphatic SN2 reaction with various
reported labeling method.31 RCY was measured by multiplying
tertiary alcohol solvents.31 In this novel synthetic method
the relative integration of a radio TLC scan and the percentage of
system, the polar protic solvent enhances the nucleophilicity of
the radioactivity in the reaction solution. For [18F]fluorination,
the fluorine ions. We found a dramatic increase of the RCY for
the [18F]fluorination reaction. In this report, we developed a
100 mL of acetonitrile and 500 mL of t-amyl alcohol were used as
solvents with 4 mL of TBAOH (6mmol) as an additional base, giving
new nosylate precursor 14c using a t-amyl alcohol solvent
system for the [18F]fluorination reaction.
41% of RCY (entry 1). When the amount of base (12 mmol) was
Table 1 shows the results of the [18F]fluorination using the
increased, the yield was lower (entry 2). We found that the optimal
amount of TBAOH was 2 mL (3mmol, entry 16) to 4 mL (6mmol). We
nosylate precursor 14c with various bases, in different amounts
of base and in reaction solvents. To optimize the [18F]fluorination
conditions, several factors should be considered. We fixed the
following reaction conditions: precursor amount, 2 mg/3.67mmol;
also checked other kinds of bases (e.g., CsOH, KOH, TBAHCO3,
KHCO3 and K2CO3). However, we could not obtain good results
except for KOH (entry 5). From a comparison of solvents in [18F]
reaction temperature, 120ꢀC; reaction volume, 600mL. [18F]
fluorination using 4 mL of TBAOH (6mmol, entry 1 vs 13) and
0.35 mg of KOH (6mmol, entry 5 vs 14), we obtained higher yields
Fluorination yield was determined by radio thin-layer
chromatography (radio-TLC). [18F]Fluorination was monitored by
in the mixture of acetonitrile (100 mL) and t-amyl alcohol (500 mL).
radio-TLC at 10 min and 20 min. After 20 min of reaction, we
Preparation of 16a-[18F]fluoroestradiol ([18F]1)
obtained the results by radio-TLC yield after 20min to determine
the incorporation of [18F]fluoride as well as by calibration of the
activity in the vial.
[18F]Fluorination was carried out using 2 mg of nosylate precursor
To optimize [18F]fluorination with various reaction conditions, 14c and [18F]fluoride (7.4GBq) in 100 mL acetonitrile and 500 mL of
[18F]fluorination was performed as follows: [18F]FÀ/H128O (85.1– t-amyl alcohol at 120 ꢀC for 20 min, as shown in Scheme 5. The
125.8MBq) was trapped in a Chromafix PS-HCO3 cartridge reaction solvent was evaporated, and the residue was then
(Machery-Nagel, Germany), and [18F]fluoride was eluted with dissolved in acetonitrile (100 mL). Aqueous hydrochloric acid (1N,
600 mL of stock solution (0.1M KOMs in H2O (250 mL) and 22mg 500 mL) was added to the reactor vial and then heated at 120 ꢀC
of K222 in methanol (1.0mL)). We added additional base (e.g., for 3 min for hydrolysis. After cooling the reaction mixture,
TBAOH, CsOH, KOH, TBAHCO3, KHCO3 or K2CO3 with 3, 6, or aqueous sodium hydroxide (2N, 250 mL) was added for
12mmol) into the reactor vial to test the base effect. The neutralization, followed by citrate buffer (200mL). This solution
radioactivity was then dried with 1 mLÂ 3 times of acetonitrile with was injected onto an HPLC system for purification. HPLC
heating under a stream of nitrogen. After drying, 2 mg (3.67 mmol) purification was conducted with a Nucleosil 100-7 C18 250
of FES nosylate precursor 14c in 100 mL of acetonitrile and 500mL Â 16mm semi-prep column (Macherey-Nagel, Germany) eluted
www.jlcr.org
Copyright © 2013 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2013, 56 619–626

H. S. Kil et al.
fluorination with the use of the new precursor. We found that t-amyl
alcohol showed higher yield than acetonitrile, and use of TBAOH
(3 mmol) base was the best condition.
Conflict of Interest
The authors did not report any conflict of interest.
Acknowledgments
This research was supported by the Converging Research
Program (2012 K001486) through the Ministry of Education,
Science and Technology. High resolution mass spectra were
carried out at the Korea Basic Science Institute (Daegu, Korea).
References
[1] M. A. Mintun, M. J. Welch, B. A. Siegel, C. J. Mathias, J. W. Brodack,
A. H. McGuire, J. A. Katzenellenbogen, Radiology 1988, 169,
45–48.
[2] F. Dehdashti, J. E. Mortimer, B. A. Siegel, L. K. Griffeth, T. J. Bonasera,
M. J. Fusselman, D. D. Detert, P. D. Cutler, J. A. Katzenellenbogen, M.
J. Welch, J. Nucl. Med. 1995, 36, 1766–1774.
[3] A. H. McGuire, F. Dehdashti, B. A. Siegel, A. P. Lyss, J. W. Brodack, C. J.
Mathias, M. A. Mintun, J. A. Katzenellenbogen, M. J. Welch, J. Nucl.
Med. 1991, 32, 1526–1531.
[4] J. A. Katzenellenbogen, M. J. Welch, F. Dehdashti, Anticancer Res.
1997, 17, 1573–1576.
[5] D. A. Mankoff, L. M. Peterson, T. J. Tewson, J. M. Link, J. R. Gralow, M.
M. Graham, K. A. Krohn, J. Nucl. Med. 2001, 42, 679–684.
[6] P. Kumar, J. Mercer, C. Doerkson, K. Tonkin, A. J. B. McEwan, J. Pharm.
Pharm. Sci. 2007, 10, 256s–265s.
[7] L. Sundararajan, H. M. Linden, J. M. Link, K. A. Krohn, D. A. Mankoff,
Semin. Nucl. Med. 2007, 37, 470–476.
[8] W. L. McGuire, K. B. Horwitz, O. H. Pearson, A. Segaloff, Cancer 1977,
39, 2934–2947.
[9] C. A. Bertelsen, A. E. Guiliano, D. H. Kern, B. D. Mann, D. J. Roe, D. L.
Morton, J. Surg. Res. 1984, 37, 257–263.
[10] L. Vollenweider-Zeragui, G. Barrelet, Y. Wong, T. Lemarchand-Béraud,
F. Gomez, Cancer 1986, 57, 1171–1180.
[11] G. M. Clarck, G. W. Sledge, Jr., C. K. Osborne, W. L. McGuire, J. Clin.
Oncol. 1987, 5, 55–61.
Figure 4. High-performance liquid chromatography chromatogram of purified
16a-[18F]fluoroestradiol ([18F]1) with
a
Luna C18(2) 250Â 4.6 mm, 1 mL/min,
water/MeCN = 40:60, 280 nm: (A) and (B) Radio and UV chromatogram of quality
control of purified [18F]1; (C) and (D) radio and UV chromatogram of purified [18F]
1 with coinjection of a cold authentic sample.
[12] J. P. Van Netten, J. B. Armstrong, S. S. Carlyle, N. L. Goodchild, I. G.
Thornton, M. L. Brigden, P. Coy, C. Fletcher, Eur. J. Cancer Clin. Oncol.
1988, 24, 1885–1889.
[13] D. F. Heiman, J. A. Katzenellenbogen, K. E. Carlson, J. E. Lloyd. In vivo
and in vitro steroid receptor assays in the design of estrogen
pharmaceuticals. In Receptor Binding Radiotracers (Ed.: W. C.
Eckelman), CRC Press, Boca Raton, FL, 1982, pp. 93–126.
[14] C. H. Cummins, Steroids 1993, 58, 245–259.
[15] D. O. Kiesewetter, M. R. Kilbourn, S. W. Landvatter, D. F. Heiman,
J. A. Katzenellenbogen, M. J. Welch, J. Nucl. Med. 1984, 25,
1212–1221.
[16] D. O. Kiesewetter, J. A. Katzenellenbogen, M. R. Kilbourn, M. J. Welch,
J. Org. Chem. 1984, 49, 4900–4905.
[17] J. W. Brodack, M. R. Kilbourn, M. J. Welch, J. A. Katzenellenbogen,
J. Nucl. Med. 1986, 27, 714–721.
at 5 mL/min with 30% EtOH:40% water:30% acetonitrile. In the
hydrolysis step, we could not find product-like impurities in the
HPLC chromatogram (Figure 3). Collected 16a-[18F]fluoroestradiol
([18F]1) was analyzed by injection to HPLC (Luna C18(2) 250
 4.6mm, 1 mL/min, water/MeCN= 40:60, 280 nm), and the
obtained HPLC chromatogram is shown in Figure 4. We obtained
purified 16a-[18F]fluoroestradiol ([18F]1) with decay-corrected RCY
of 19-24% (n = 3) after HPLC purification for manual synthesis. Total
synthesis time including purification time was about 40 min. The
specific activity of [18F]1 was 84.2 GBq/mmol.
[18] J. W. Brodack, M. R. Kilbourn, M. J. Welch, J. A. Katzenellenbogen,
J. Rad. Appl. Instrum. [A]. 1986, 37, 217–221.
[19] J. W. Brodack, M. R. Kilbourn, M. J. Welch, J. Rad. Appl. Instrum. [A]
1988, 39, 689–698.
[20] J. L. Lim, L. Zheng, M. S. Berridge, T. J. Tewson, Nucl. Med. Biol. 1996,
23, 911–915.
[21] J. Lim, M. S. Berridge, T. J. Tewson, J. Label. Compd. Radiopharm.
1994, 35, 176–177.
[22] J. Romer, F. Fuchtner, J. Steinbach, B. Johannsen, Nucl. Med. Biol.
1999, 26, 473–479.
[23] S. Liang, X. L. Lan, Y. X. Zhang, X. M. Xu, B. Li, Nucl. Med. Commun.
2012, 33, 29–33.
[24] K. E. Knott, D. Gratz, S. Huebner, S. Juttler, C. Zanki, M. Muller, J. Label.
Compd. Radiopharm. 2011, 54, 749–753.
Conclusion
We described a new synthesis of 16a-[18F]fluoroestradiol ([18F]1)
using nosylate precursor 14c as an alternative precursor. An efficient
synthetic route for the preparation of nosylate precursor 14c was
also achieved. Using this nosylate precursor 14c, [18F]fluorination
and hydrolysis steps are suitable for an automatic chemistry module.
We also evaluated the effects of various bases and solvents on [18F]
J. Label Compd. Radiopharm 2013, 56 619–626
Copyright © 2013 John Wiley & Sons, Ltd.
www.jlcr.org

H. S. Kil et al.
[25] S. J. Oh, D. Y. Chi, C. Mosdzianowski, H. S. Kil, J. S. Ryu, D. H. Moon, [29] P. G. M. Wuts, T. W. Greene. In Greene’s Protective Groups in Organic
Appl. Radiat. Isot. 2007, 65, 676–681.
[26] J. Romer, J. Steinbach, H. Kasch, Appl. Radiat. Isot. 1996, 47,
Synthesis, 4th edition. Wiley-Interscience, Hoboken, NJ, 2007, p 999
reactivity chart 2.
395–399.
[30] H. Laurent, D. Bittler, S. Beier, W. Elger. 3,17b-Estriol diesters, method
of their use, and pharmaceutical preparations containing them. USP
4,681,875, 1987.
[27] M. Numazawa, M. Shelangouski, M. Nakakoshi, Steroids 2001, 66,
743–748.
[28] R. P. Stein, Conshohocken, R. C. Smith, Jr., King of Prussia, S. Herchel, [31] D. W. Kim, D. S. Ahn, Y. H. Oh, S. Y. Lee, H. S. Kil, S. J. Oh, S. J. Lee, J. S.
M. Bryn. Preparation of steroidal formate esters. USP 3,647,784,
Kim, J. S. Ryu, D. H. Moon, D. Y. Chi, J. Am. Chem. Soc. 2006, 128,
1972.
16394–16397.
www.jlcr.org
Copyright © 2013 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2013, 56 619–626