Radiosynthesis of 1′-[18F]fluoroethyl-β-D-lactose ([18F]-FEL) for early detection of pancreatic carcinomas with PET

Article abstract of DOI:10.1002/jlcr.1855

Introduction: The hepatocellular carcinoma-intestine-pancreas and pancreatitis-associated proteins, also known as lactose-binding protein, is upregulated in peritumoral pancreatic tissue. Previously, we reported ethyl-β-D-galactopyranosyl-(1,4′)-2′-deoxy-

Full text of DOI:10.1002/jlcr.1855

Research Article
Received 25 August 2010,
Revised 19 October 2010,
Accepted 20 October 2010
Published online 25 March 2011 in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/jlcr.1855
Radiosynthesis of 10-[18F]fluoroethyl-b-D-
lactose ([18F]-FEL) for early detection of
pancreatic carcinomas with PET
Nashaat Turkman,y Ashutosh Pal,y William P. Tong, Juri G. Gelovani, and
Ã
Mian M. Alauddin
Introduction: The hepatocellular carcinoma–intestine–pancreas and pancreatitis-associated proteins, also known as lactose-
binding protein, is upregulated in peritumoral pancreatic tissue. Previously, we reported ethyl-b-D-galactopyranosyl-(1,40)-
20-deoxy-20-[18F]fluoro-b-D-glucopyranoside (Et-[18F]-FDL),
a radiofluorinated lactose analog for positron emission
tomography (PET) of small pancreatic carcinomas in mice. However, synthesis of the precursor for Et-[18F]-FDL involves
11 steps, which is quite lengthy, and produces overall low yields. Here, we report on synthesis and radiolabeling of another
analog of lactose, the 10-[18F]fluoroethyl-b-D-lactose for PET imaging of pancreatic carcinomas. Methods: Two precursor
compounds, 10-bromoethyl-20,30,60,2,3,4,6-hepta-O-acetyl-b-D-lactose 4, and 10-p-toluenesulfonylethyl-20,30,60,2,3,4,6-hepta-
O-acetyl-b-D-lactose 5, were synthesized in two and three steps, respectively; then, cold fluorination and radiofluorination
of these precursors were performed. The reaction mixture was passed through a silica gel Sep-pack cartridge, eluted with
EtOAc, and the 10-[18F]fluoroethyl-20,30,60,2,3,4,6-hepta-O-acetyl-b-D-lactose ([18F]-6) purified by HPLC. After hydrolysis of
the protecting groups, the 10-[18F]fluoroethyl-b-D-lactose [18F]-7 was neutralized, diluted with saline, filtered through a
sterile Millipore filter, and analyzed by radio-TLC. Results: The average decay-corrected radiochemical yield was 9% (n = 7)
with499% radiochemical purity and specific activity of 55.5 GBq/lmol. Conclusion:
A new analog of lactose,
10-[18F]fluoroethyl-b-D-lactose, has been synthesized in good yields, with high purity and high specific activity suitable for
PET imaging of early pancreatic carcinomas.
Keywords: fluorine-18; lactose; HIP/PAP; pancreatic carcinoma; PET
than 130 fold in peritumoral pancreatic acinar cells, as compared
to normal pancreas during early stages of cancer develop-
Introduction
ment.12–15 Also, HIP/PAP is overexpressed in hepatocellular
Pancreatic cancer is a deadly disease, although the numbers of
carcinomas by the tumor cells.16–18 The level of HIP/PAP
patients are much less compared with other cancers. The
estimated number of new cases of pancreatic cancer in the
United States in 2010 is more than 43000 and an estimation of
deaths is 36000.1 Early diagnosis of pancreatic carcinoma using
highly sensitive diagnostic imaging methods could save the lives
of many thousands of patients, because early detection is likely to
increase resectability rates. [18F]-2-Fluorodeoxy-D-glucose ([18F]-
FDG) is the conventional positron emission tomography (PET)
imaging agent that is routinely used for early detection of many
types of cancer.2–6 However, [18F]-FDG PET is not suitable for
pancreatic cancer or hepatocellular cancer, because of the low
[18F]-FDG accumulation in early primary lesions of these tumor
types.7–9 Also, the resolution and sensitivity of 18F-FDG/PET
imaging is 5 mm, which is larger than the size of early pancreatic
or hepatocellular carcinomas (2–3mm) that are not detectable by
CT or MRI.10 The poor prognosis is due to 80–90% of patients
having unresectable disease at the time of diagnosis.11 Therefore,
there is a need of a sensitive and specific PET imaging agent
expression continues to increase with the increasing tumor
mass.19 Therefore, HIP/PAP is a promising target for PET imaging
of pancreatic carcinoma and hepatocellular carcinoma.15 HIP/
PAP is a 16 kD secreted protein, which belongs to a family of
proteins that contain a C-type lectin-like domain,13, 19 that binds
carbohydrates and also known as ‘lactose-binding protein’.20
Previous biochemical studies demonstrated that glutathione-
S-transferase purified recombinant HIP/PAP has a very high
affinity to D-lactose but it binds insignificantly to a-D-glucopyro-
anose, a-L-fucose, a-D-galactopyranose, and N-acetyl-b-D-gluco-
samine.21 Therefore, radiolabeled D-lactose analogs, such as
[18F]-labeled D-lactose or its derivatives, should be highly useful
Department of Experimental Diagnostic Imaging, Center for Advanced
Biomedical Imaging Research (CABIR), The University of Texas MD Anderson
Cancer Center, Houston, TX 77030, USA
suitable for early detection of pancreatic and hepatocellular *Correspondence to: Mian M. Alauddin, 1515 Holcombe Blvd. T8.3895, Box 059,
Houston, TX 77030, USA.
E-mail: alauddin@mdanderson.org
carcinoma.
The hepatocellular carcinoma–intestine–pancreas/pancreatitis-
associated protein (HIP/PAP) gene is overexpressed by more yNashaat Turkman and Ashutosh Pal contributed equally to this work.
J. Label Compd. Radiopharm 2011, 54 233–238
Copyright r 2011 John Wiley & Sons, Ltd.

N. Turkman et al.
for detection of early pancreatic and hepatocellular carcinomas O-acetyl-b-D-lactose (10 g, 14.72 mmol) in anhydrous dichloro-
by PET.
methane (20 mL) was treated with 30% HBr in acetic acid
Two different [18F]-labeled lactose derivatives have been (20 mL) at room temperature for 30 min. The solution was
previously reported: b-O-D-galactopyranosyl-(1,40)-20-[18F]fluoro-20- cooled in an ice bath, diluted with dichloromethane (CH2Cl2,
deoxy-D-glucopyranose ([18F]-FDL)22 and Et-b-D-galactopyranosyl- 50 mL), washed with ice-cold water (2 ꢀ 5 mL) and ice-cold
(1,40)-20-deoxy-20-[18F]fluoroethyl-D-glucopyranose (Et-[18F]-FDL).23 saturated bicarbonate (2 ꢀ 25 mL), dried under Na2SO4 and
Biodistribution studies of [18F]-FDL in normal mice have demon- evaporated to give a colorless oil. The crude product was stirred
strated that [18F]-FDL did not accumulate in any of the major with diethyl ether (50 mL) to give a white crystalline solid, which
1
organs and that it was rapidly cleared from the circulation by was filtered to give 2 (8.5 g) in 82% yield. m. p. 139–1421C. H
renal clearance and appeared in urine as the non-metabolized NMR (CDCl3) d: 6.53 (d, J = 3.6 Hz, 1H), 5.56 (t, J = 9.6 Hz, 1H), 5.36
parent compound.22 Recently, we reported a high-yield radio- (d, J = 3.0 Hz, 1H), 5.13 (dd, J = 7.8, 7.8 Hz, 1H), 4.97 (dd, J = 3.0,
synthesis of Et-[18F]-FDL24 and its efficacy for detecting early 3.6 Hz, 1H), 4.77 (dd, J = 3.6, 4.2 Hz, 1H), 4.49–4.53 (m, 2H),
microscopic pancreatic carcinoma lesions in a bioluminescent 4.07–4.23 (m, 4H), 3.85–3.92 (m, 2H), 2.10 (s, 3H), 2.14 (s, 3H),
variant of an orthotopic pancreatic carcinoma xenograft model in 2.10 (s, 3H), 2.08 (s, 3H), 2.07 (s, 3H), 2.06 (s, 3H), 1.98 (s, 3H). 13
C
mice using microPET/CT.25 MicroPET/CT results from our study NMR (CDCl3) d: 170.38, 170.21, 170.17, 170.11, 170.01, 169.26,
suggested that Et-[18F]-FDL is an excellent agent for PET imaging 168.99, 100.84, 86.40, 75.00, 72.99, 71.02, 70.88, 70.81, 69.62,
of HIP/PAP expression in pancreatic carcinomas in animals.25 69.04, 66.62, 61.06, 60.89, 20.84, 20.82, 20.69, 20.69, 20.69, 20.89,
However, synthesis of the precursor for Et-[18F]-FDL involves 11 20.53. MS (ESI): calculated for C26H35BrO17 698.10, found 716.40,
steps, which is quite lengthy and produces overall low yields. As and 718.40 (M1NH4).
binding of lactose with HIP/PAP involves only 3-hydroxyl and
4-hydroxyl groups of the galactose unit,19 modification at the Preparation of 10-hydroxyethyl-20,30,60,2,3,4,6-hepta-O-
acetyl-b-D-lactose: 3
1-position of the glucose moiety of D-lactose should not change
the binding efficiency. With this concept in mind, we have
synthesized a new derivative of D-lactose and radiolabeled it with
[18F]. In this paper, we report the radiosynthesis of a new analog
of lactose, 10-[18F]fluoroethyl-b-D-lactose ([18F]-FEL) for detection
of hepatocellular and pancreatic carcinomas by PET.
A mixture of 2 (2.0 g, 2.86 mmol) and ethylene glycol (5 mL) in
dry acetonitrile (MeCN, 20 mL) containing activated molecular
sieves (4 A1, 5 g) was stirred at room temperature under N2 for
15 min. Silver p-toluenesulfonate (AgOTs, 1.6 g, 5.8 mmol) was
added and the reaction mixture was stirred for 1 h at room
temperature. The reaction mixture was diluted with MeCN
(5 mL), filtered through a celite pad and the filtrate was
evaporated. The product was dissolved in ethyl acetate (EtOAc,
30 mL), washed with water (3 ꢀ 30 mL) and the aqueous
Experimental
Reagents and instrumentation
All reagents and solvents were purchased from Aldrich Chemical washing was back extracted with EtOAc (30 mL). The combined
Co. (Milwaukee, WI), and used without further purification. Solid- EtOAc extract was dried over anhydrous Na2SO4 and concen-
phase extraction cartridges (Sep-Pak, silica gel, 900 mg) were trated under vacuum. The crude product was purified by flash
purchased from Alltech Associates (Deerfield, IL). Thin layer column chromatography on a silica gel column, eluted with
chromatography (TLC) was performed on pre-coated Kieselgel 60–80% EtOAc in hexane as a gradient elution. Solvent was
1
60 F254 (Merck, Darmstadt, Germany) glass plates. NMR spectral evaporated to afford 3 (1.4 g) in 72% yield as a white solid. H
data were collected at the University of Texas M. D. Anderson NMR (CDCl3) d: 5.35 (d, J = 3.6 Hz, 1H), 5.21 (t, J = 9.6 Hz, 1H), 5.12
Cancer Center using either a Bruker 300 or a 600 MHz spectro- (dd, J = 7.8, 7.8 Hz, 1H), 4.96 (dd, J = 3.0, 3.6 Hz, 1H), 4.92 (dd,
meter. Proton, 13C, and 19F NMR spectra were recorded using J = 7.8, 7.8 Hz, 1H), 4.48–4.56 (m, 3H), 4.05–4.17 (m, 4H),
tetramethylsilane as an internal reference or hexafluorobenzene 3.86–3.91 (m, 2H), 3.66–3.81 (m, 5H), 2.16 (s, 3H), 2.13 (s, 3H),
as an external reference. High-resolution mass spectra (MS) were 2.07 (s, 3H), 2.06 (s, 6H), 2.05 (s, 3H), 1.97 (s, 3H). 13C NMR (CDCl3)
obtained on a Bruker BioTOF II mass spectrometer at the d: 170.42, 170.38, 170.16, 170.09, 169.78, 169.73, 169.12, 101.33,
University of Minnesota using electrospray ionization (ESI).
101.11, 76.40, 73.54, 72.78, 72.65, 70.68, 70.96, 70.74, 69.09,
High Performance Liquid Chromatography (HPLC) was per- 66.60, 62.07, 62.04, 60.80, 20.80, 20.77, 20.74, 20.66, 20.66, 20.64,
formed with an 1100 series pump, (Agilent Technologies, Stuttgart, 20.53, 20.53. High-resolution MS: calculated for C28H40O19
Germany), with a built-in UV detector operated at 210 and 254 nm, 703.2056, found 703.2107 (M1Na); calculated 698.2502, found
and a radioactivity detector with a single-channel analyzer (Bioscan, 698.2593 (M1NH4).
Washington, DC) and a semipreparative C18 reverse-phase column
(Alltech, Econosil, 10 ꢀ 250 mm) and an analytical C18 column Preparation of 10-bromoethyl-20,30,60,2,3,4,6-hepta-O-acetyl-
b-D-lactose: 4
(Alltech, Econosil, 4.6 ꢀ 250 mm). An acetonitrile/water solvent
system (45% MeCN/H2O) was used for purification of the
radiolabeled product at a flow of 4 mL/min. Quality control analyses
were performed on an analytical HPLC column with 55% MeCN/
H2O solvents at a flow of 1 mL/min. Radio-TLC was performed on an
AR-2000 Imaging Scanner (Bioscan, Washington, DC).
10-Bromoethyl-20,30,60,2,3,4,6-hepta-O-acetyl-b-D-lactose
4 was
prepared from 2 as described for the preparation of 3. Briefly,
compound 2 (1.0 g 1.43 mmol) and 2-bromoethanol (2.0 mL) in
MeCN (12 mL) was stirred under argon in the presence of
molecular sieves (2 g, 4 A1) and Ag-OTs (0.8 g) for 1 h at room
temperature. The reaction mixture was diluted with MeCN
(5 mL), filtered through a celite pad and the filtrate evaporated.
The product was dissolved in EtOAc (20 mL), washed with water
Preparation of 10-bromo-10-deoxy-20,30,60,2,3,4,6-hepta-O-acetyl-
a-D-lactose: 2
This compound was prepared according to a previously (3 ꢀ 20 mL), and the aqueous washing back-extracted once with
published method.26 Briefly, a solution of 10,20,30,60,2,3,4,6-octa- EtOAc (15 mL). The combined EtOAc extract was dried over
www.jlcr.org
Copyright r 2011 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2011, 54 233–238

N. Turkman et al.
anhydrous Na2SO4 and concentrated under vacuum. The crude 3.94–4.04 (m, 1H), 3.78–3.91 (m, 3H), 3.60–3.64 (m, 1H), 2.16 (s,
product was purified by flash chromatography on a silica gel 3H), 2.13 (s, 3H), 2.07 (s, 3H), 2.05 (s, 9H), 1.97 (s, 3H). 13C NMR
column and eluted with 60–80% EtOAc in hexane as a gradient (CDCl3) d: 170.39, 170.39, 170.18, 170.09, 169.77, 169.77, 169.09,
elution. Solvent was evaporated to afford 4 (0.53 g) in 50% yield 101.07, 100.73, 83.05, 81.92, 76.18, 72.69, 71.53, 70.99, 70.69,
1
as a white solid. H NMR (CDCl3) d: 5.35 (dd, J = 3.6, 1.2 Hz, 1H), 69.10, 68.82, 61.90, 60.81, 20.87, 20.82, 20.66(4C), 20.53. 19F NMR
5.21 (t, J = 9.0 Hz, 1H), 5.12 (dd, J = 7.8, 7.8 Hz, 1H), 4.96 (dd, (coupled) d: ꢁ224.22 (m). High-resolution MS: calculated for
J = 3.0, 3.6 Hz, 1H), 4.92 (dd, J = 7.8, 7.8 Hz, 1H), 4.48–4.55 (m, 3H), C28H39FO18 705.2013, found 705.2032 (M1Na).
4.06–4.16 (m, 4H), 3.86–3.89 (m, 1H), 3.78–3.83 (m, 2H), 3.61–3.64
Method 2: Compound 6 prepared from 2 (method 1) is not
(m, 1H), 3.41–3.48 (m, 2H), 2.16 (s, 3H), 2.13 (s, 3H), 2.07 (s, 6H), suitable for radiosynthesis; therefore, an alternative method was
2.052 (s, 3H), 2.05 (s, 3H), 1.97 (s, 3H). 13C NMR (CDCl3) d: 170.37, developed. In this method 6 was prepared either from
170.37, 170.17, 170.10, 169.76, 169.74, 169.09, 101.10, 100.79, compound 4 or 5 by fluorination with Bu4NF; a representative
76.19, 72.76, 72.59, 71.41, 70.98, 70.71, 69.81, 69.10, 66.59, 61.85, preparation from 4 is described here. Compound 4 (10 mg,
60.80, 29.86, 20.89, 20.82, 20.80, 20.67 (3C), 20.54. High- 13.5 mmol) was dissolved in MeCN (1.0 mL) in a v-vial. To this
resolution MS: calculated for C28H39BrO18 765.1212, found solution, 40 mL of Bu4NF in tetrahydrofuran (1 M solution, 3 eq)
765.1212 (M1Na); 760.1658, found 760.1673 (M1NH4).
was added, and the reaction mixture was heated at 1001C for
20 min. The reaction mixture was cooled to room temperature,
Preparation of 10-p-toluenesulfonylethyl-20,30,60,2,3,4,6- filtered through a small silica gel column and eluted with EtOAc
hepta-O-acetyl-b-D-lactose: 5
(5 mL). Solvent was evaporated; the crude product was re-
dissolved in MeCN and analyzed by HPLC, which showed the
product to be 40% in the crude mixture. 19F NMR spectrum of
this crude product showed a peak at ꢁ224.22 ppm, which was
identical with the pure product as described above in method 1.
10-Hydroxyethyl-20,30,60,2,3,4,6-hepta-O-acetyl-b-D-lactose 3 (1.0 g,
1.5mmol) was dissolved in dichloromethane (20 mL). Triethyla-
mine (0.5 mL, 3.6 mmol) and p-toluenesulfonic anhydride (1.0 g,
3.0 mmol) were added, and the solution was stirred at room
temperature for 5 h. The solvent was evaporated and the crude
product purified by flash chromatography on a silica gel column
Preparation of 10-fluoroethyl-b-D-lactose: 7
using 60–70% EtOAc in hexane as a gradient elution to yield 5 A solution of 10-fluoroethyl-20,30,60,2,3,4,6-hepta-O-acetyl-b-D-
(0.98 g) in 50% yield as a white solid. 1H NMR (CDCl3) d: 7.78 lactose 6 (200 mg, 0.29 mmol) in absolute methanol (10 mL)
(d, J = 7.78 Hz, 2H), 7.36 (d, J = 7.78 Hz, 2H), 5.35 (d, J = 3.6 Hz, 1H), was diluted with 0.1 M sodium methoxide in absolute methanol
5.17 (t, J = 9.0 Hz, 1H), 5.10 (dd, J = 7.8, 7.8 Hz, 1H), 4.95 (6 mL). The solution was stirred at room temperature for 1 h and
(dd, J = 3.0, 3.6 Hz, 1H), 4.86 (dd, J = 7.8, 7.8 Hz, 1H), 4.46–4.50 then quenched with ion-exchange resin (Amberlite IRC-50,
(m, 3H), 4.06–4.17 (m, 5H), 3.95–3.99 (m, 1H), 3.86–3.99 (m, 1H), 200 mg). The resin was removed by filtration, the resin washed
3.74–3.81 (m, 2H), 3.57–3.61 (m, 1H), 2.46 (s, 3H), 2.16 (s, 3H), with methanol (5 mL), and the combined filtrate evaporated to
2.12 (s, 3H), 2.06 (s, 3H), 2.05 (s, 6H), 2.04(s, 3H), 1.97 (s, 3H). 13C give 7 as an off-white solid (108 mg) in 95% yield. 1H NMR (D2O)
NMR (CDCl3) d: 170.38, 170.38, 170.17, 170.09, 169.79, 169.72, d: 4.47 (dt, 2H, JH-F = 47.4 Hz, JH–H = 3.6 Hz), 4.35 (d, 1H, J = 8.4 Hz),
169.09, 145.02, 132.77, 129.95, 127.93, 101.07, 100.03, 76.09, 4.24 (d, J = 7.8 Hz, 1H), 3.90–3.98 (m, 1H), 3.71–3.81 (m, 3H),
72.71, 72.62, 71.34, 70.98, 70.68, 69.09, 68.49, 67.09, 66.59, 61.86, 3.51–3.62 (m, 4H), 3.44–3.48 (m, 3H), 3.40 (m, 1H), 3.32–3.36 (m,
60.79, 21.66, 20.86, 20.81, 20.67, 20.67, 20.67, 20.65, 20.53. High- 1H), 3.13–3.17 (m, 1H). 13C NMR (D2O) d: 102.90, 102.15, 83.82,
resolution MS: calculated for C35H46O21S 857.2145, found 82.74, 78.25, 75.32, 74.78, 74.28, 72.73, 72.47, 70.91, 69.25, 69.13,
457.2182 (M1Na).
68.50, 60.99, 59.99. 19F NMR (coupled) d: ꢁ223.19 (m). High-
resolution MS: calculated for C14H25FO11 411.1273, found
Preparation of 10-fluoroethyl-20,30,60,2,3,4,6-hepta-O-acetyl- 411.1255 (M1Na).
b-D-lactose: 6
Radiosynthesis of 10-[18F]fluoroethyl-b-D-lactose: [18F]-7
This compound was prepared using two different methods as
described below.
The aqueous [18F]fluoride produced from the cyclotron by the
Method 1: 10-Fluoroethyl-20,30,60,2,3,4,6-hepta-O-acetyl-b-D- reaction of 18O(p, n)[18F] was trapped on an ion-exchange
lactose 6 was prepared from 2 as described for the preparation cartridge (Chromafix 30-PS-HCO3, ABX) and eluted with an
of 3 and 4. Briefly, a mixture of 2 (1.0 g, 1.43 mmol) and aqueous solution of K2CO3 (2.75 mg/mL) into a V-vial containing
2-fluoroethanol (2.5 mL) in dry MeCN (10 mL) containing kryptofix [2.2.2] solution (12.0 mg/mL) in acetonitrile. Water was
activated molecular sieves (4 A1, 2 g) and Ag-OTs (0.8 g) was removed by an azeotropic evaporation at 901C with acetonitrile
stirred at room temperature under N2 for 1 h. The reaction (1.0 mL) under a stream of argon. A solution of 4 or 5 (5–6 mg) in
mixture was diluted with MeCN (5 mL), filtered through a celite acetonitrile (0.5 mL) was added to the dry K18F/kryptofix [2.2.2].
pad, and the filtrate evaporated. The product was dissolved in The reaction mixture was heated at 1001C for 20 min. The crude
ethyl acetate (20 mL), washed with water (3 ꢀ 20 mL), and the reaction mixture was passed through a silica Sep-Pak cartridge
aqueous washing back-extracted once with EtOAc (15 mL). The followed by elution with two portions of ethyl acetate (2.5 mL,
combined EtOAc extract was dried over anhydrous Na2SO4 and total) which was evaporated at 801C under a stream of argon.
concentrated under vacuum. The crude product was purified by The residue was dissolved in 60% acetonitrile/water (1.0 mL) and
flash chromatography on a silica gel column and eluted with injected onto the HPLC connected with a semipreperative C18
60–80% EtOAc in hexane as a gradient elution. Solvent was column. The product [18F]-6 was eluted with 45% acetonitrile/
1
evaporated to afford 6 (0.49 g) in 50% yield as a white solid. H water at a flow of 4 mL/min. The appropriate fraction (radio-
NMR (CDCl3) d: 5.35 (d, J = 3.0 Hz, 1H), 5.21 (t, J = 9.0 Hz, 1H), 5.10 active) was collected between 16.5 and 19.0 min, and an aliquot
(dd, J = 8.4, 7.8 Hz, 1H), 4.96 (dd, J = 3.6, 3.6 Hz, 1H), 4.92 of the product [18F]-6 was analyzed on an analytical HPLC
(dd, J = 8.4, 7.8 Hz, 1H), 4.47–4.58 (m, 5H), 4.07–4.16 (m, 4H), column to verify its identity and purity by coinjection with the
J. Label Compd. Radiopharm 2011, 54 233–238
Copyright r 2011 John Wiley & Sons, Ltd.
www.jlcr.org

N. Turkman et al.
nonradioactive authentic sample 6. The solvent from the rest of compound, such as 3, 4, and 6. Compound 6 is a 10-fluoroethyl-
the product [18F]-6 was evaporated under reduced pressure, 20,30,60,2,3,4,6-hepta-O-acetyl-b-D-lactose derivative, which pro-
and the product was dissolved in CH2Cl2, transferred to a V-vial, duced the desired FEL 7 upon hydrolysis. Although this
and the solvent evaporated again. The residue was dissolved in methodology is not suitable for radiosynthesis, 6 obtained by
methanol (0.4 mL), 0.5 M NaOMe/MeOH solution (0.1 mL) was this methodology can be used as the standard compound
added, and the mixture was heated at 801C for 5 min, when for HPLC analysis. Therefore, we prepared 6 from 2 as a
HPLC showed complete hydrolysis of [18F]-6. Solvent was direct method to achieve the standard compound with full
evaporated, the residue, [18F]-7, was neutralized with HCl, characterization by 1H, 13C, and 19F-NMR spectroscopy and high-
diluted with saline and finally filtered through a Millipore filter. resolution mass spectrometry. Reaction of 2 with 2-fluoro-
Compound 7 ([18F]-FEL) was analyzed by radio-TLC, which ethanol was performed in the presence of AgOTs and molecular
showed499% pure product with same Rf value as the authentic sieves, the latter of which act as a drying agent. AgOTs was used
compound.
to accelerate the reaction for production of the corresponding
compounds 3, 4, and 6. The isolated yield of 6 in this
1
methodology was moderate, 50%. The H NMR spectrum of 6
Results and discussion
was quite complex in the region where the CH2-F protons
Scheme 1 describes the synthesis of 10-fluoroethyl-b-D-lactose resonate. These two protons were observed at 4.54 ppm with a
(FEL) 7, and Scheme 2 describes the radiosynthesis of 10- typical germinal H-F coupling constant of 47.4 Hz. However, a
[18F]fluoroethyl-b-D-lactose ([18F]-FEL) [18F]-7. The lactose deri- few other protons were also resonating with similar chemical
1
vative 10-bromo-10-deoxy-20,30,60,2,3,4,6-hepta-O-acetyl-a-D-lac- shifts, therefore complicating the interpretation of the H NMR
tose 2 was synthesized from peracetyl lactose 1 using a spectrum. The 19F NMR spectrum (coupled) was clear, showing a
method described previously.26 The chemical yield of 2 was peak at ꢁ224.22 ppm as a multiplet, which supports the
quite high (82%), and the 1H NMR spectrum was consistent with presence of CH2-F.
that reported in the literature.26 Compound 2 was the key
Reaction of ethylene glycol with 2 in the presence of
intermediate that could be used to produce more than one AgOTs and molecular sieves produced 10-hydroxyoethyl-
OAc
OAc
OAc
AcO
AcO
AcO
AcO
AcO
AcO
OAc
O
OAc
O
OAc
O
AcO
O
AcO
O
AcO
O
OH
O
O
O
OAc
Br
O
AgOTs
OAc
OAc
OAc
HBr/AcOH
OH
HO
OAc
OAc
OAc
3
2
1
Br
HO
F
HO
(Ts)2O, Et3N
AgOTs
OAc
AgOTs
AcO
AcO
OAc
OAc
AcO
AcO
AcO
OAc
O
AcO
OTs
O
OAc
O
OAc
AcO
O
Br
O
AcO
O
F
O
O
O
O
O
OAc
AcO
OAc
O
OAc
MeCN, F-
MeCN, F-
OAc
OAc
OAc
5
4
6
OH
OAc
HO
AcO
AcO
OH
OAc
O
HO
F
O
AcO
O
F
O
O
O
O
HO
O
OH
OAc
MeOH/NaOMe
OH
OAc
7
6
Scheme 1. Scheme for synthesis of 10-fluoroethyl-b-D-lactose 7.
OH
HO
OAc
AcO
OAc
AcO
AcO
OH
HO
AcO
18F
OAc
O
18F
O
OAc
O
R
O
AcO
O
O
O
O
O
HO
O
O
AcO
O
OH
OAc
OAc
OH
OAc
MeCN, K18F/Kryptofix
OAc
MeOH/NaOMe
6
7
4 R= Br
5 R=OTs
Scheme 2. Scheme for radiosynthesis of 10-[18F]fluoroethyl-b-D-lactose [18F]-7.
www.jlcr.org
Copyright r 2011 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2011, 54 233–238

N. Turkman et al.
Figure 1. HPLC purification of 10-[18F]fluoroethyl-20,30,60,2,3,4,6-hepta-O-acetyl-b-D-lactose, [18F]-6: Semipreparative C18 Column; 45% MeCN and 55% H2O; flow 4.0 mL/min.
20,30,60,2,3,4,6-hepta-O-acetyl-b-D-lactose 3 in 72% yield. This tosylate 5 at around 1001C for 20 min may be the ideal reaction
reaction was also enhanced and the yield improved by the condition for routine production.
presence of AgOTs. Similarly, 10-bromoethyl-20,30,60,2,3,4,6-
After solvent evaporation, the crude product [18F]-6 was
hepta-O-acetyl-b-D-lactose 4 was prepared from 2 by reacting purified by HPLC. Figure 1 represents a HPLC chromatogram for
with bromoethanol, and obtained in 50% yield. Both com- purification of 10-[18F]fluoroethyl-20,30,60,2,3,4,6-hepta-O-acetyl-
pounds 3 and 4 were characterized by 1H and 13C NMR b-D-lactose 18F-6. An aliquot of [18F]-6 was analyzed by HPLC on
spectroscopy and high-resolution mass spectrometry. The an analytical column, which showed a single radioactive peak
1-hydroxyethyl-b-lactose 3 was converted to the corresponding co-eluted with an authentic nonradioactive standard (Figure 2).
tosyl derivative 5 by treatment with tosyl anhydride and triethyl After removal of the HPLC solvent under vacuum, the product
amine in CH2Cl2. The chemical yield in this step was 50%. The 18F-6 was hydrolyzed with NaOMe to obtain the final product
product was fully characterized by 1H NMR spectroscopy and [18F]-7 in quantitative yield. Hydrolysis was monitored by HPLC,
high-resolution mass spectrometry. Both the bromoethyl lactose which showed that in 5 min the hydrolysis was complete. After
4 and the tosyl-ethyl lactose 5 were used as precursors solvent evaporation from [18F]-7, the product was neutralized
for radiosynthesis of the peracetyl-10-fluoroethyl-b-lactose 6. with 1 M HCl solution, diluted with saline, and finally filtered
Reaction of 4 and 5 with n-Bu4NF at 80–1001C for 20 min through a Millipore filter. Radio-TLC of the product [18F]-7
produced 6 in 40–42% yields. After deacetylation of peracetyl-10- showed a single product with an Rf value of 0.82, consistent with
fluoroethyl-b-lactose 6 by reaction with 0.5 M sodium methox- that of the non-radioactive compound (Figure 3). The Rf value
ide (0.1 mL) in methanol (0.4 mL) for 10 min at 801C, was similar to that of the previously reported compound.24 The
10-fluoroethyl-b-lactose 7 was produced in quantitative yield.
overall radiochemical yield was 7–11% with an average of 9%
For radiosynthesis of 7, both precursors 4 and 5 were (d. c., n = 7). The radiochemical purity was499% with specific
radiofluorinated with K18F/kryptofix [2.2.2.] in dry MeCN activity of 55.5 GBq/mmol at the end of synthesis. The synthesis
(Scheme 2) at several temperatures. Compound 4 was radio- time was 100–105 min from the end of bombardment.
fluorinated at temperatures of 85–100oC; and 5 was radio-
fluorinated at temperatures of 90–110oC. The crude product
18F-6 was passed through a silica gel cartridge to remove
unreacted fluoride then eluted with EtOAc, and solvent was
Conclusion
A new [18F]-labeled analog of lactose, 10-[18F]fluoroethyl-b-D-
evaporated under a stream of argon. The precursor 4 produced
18F-6 in 15, 19, and 21% crude yields at 85, 90, and 1001C,
respectively; and the precursor 5 produced 18F-6 in 17, 31, and
17% crude yields at 90, 100, and 1101C, respectively. In the
radiofluorination step, the tosylate precursor 5 appeared to be
slightly better precursor with an average crude yield of 21%,
whereas that from the bromo-precursor 4 was 18%. Thus,
lactose, has been synthesized in good yields with high purity
and high specific activity. Unlike synthesis of similar precursor
compounds, our method uses less number of steps and thus is
more practical for rapid, high-volume production. The com-
pound [18F]-7 may be useful for early detection of pancreatic
carcinomas by PET imaging and should be tested in vivo in
tumor-bearing mice.
J. Label Compd. Radiopharm 2011, 54 233–238
Copyright r 2011 John Wiley & Sons, Ltd.
www.jlcr.org

N. Turkman et al.
Figure 2. HPLC chromatogram of 10-[18F]fluoroethyl-20,30,60,2,3,4,6-hepta-O-acetyl-b-D-lactose [18F]-6, co-injected with standard 10-fluoroethyl-20,30,60,2,3,4,6-hepta-O-
acetyl-b-D-lactose 6: Analytical C18 Column; 55% MeCN and 45% H2O; flow 1.0 mL/min.
Lactose - LACTOSE_TS.R001
[6] F. U. Chowdhury, N. Shah, A. F. Scarsbrook, K. M. Bradley, Postgrad
Med. J. 2010, 86, 174–182.
25000
20000
15000
10000
5000
0
[7] D. Delbeke, W. H. Martin, Surg. Oncol. Clin. N. Am. 2010, 19(2),
235–254.
[8] S. Tang, G. Huang, J. Liu, T. Liu, L. Treven, S. Song, C. Zhang,
L. Pan, T. Zhang, Eur. J. Radiol. 2009 [Epub ahead of print].
DOI: 10.1016/j.ejrad.2009.09.026.
[9] R. Hustinx, Radiol. Clin. North Am. 2004, 42(6), 1123–1139.
[10] F. Pakzad, A. M. Groves, P. J. Ell, Semin. Nucl. Med. 2006, 36, 248–256.
[11] G. J. Poston, R. C. Williamson, Compr. Ther. 1990, 16, 36–42.
[12] C. Lasserre, L. Christa, M-T Simmon, P. vernier, C. Brechot, Cancer
Res. 1992, 52, 5089–5095.
[13] C. Lasserre, C. Colnot, C. Brechot, F. Poirier, Am. J. Pathol. 1999,
154, 1601–1610.
[14] N. Fukushima, J. Koopmann, N. Sato, N. Prasad, R. Carvalho,
S. D. Leach, R. H. Hruban, M. Goggins, Mod. Pathol. 2005, 18, 779–787.
[15] C. Rosty, L. Christa, C. S. Kuzdzal, W. M. Balwin, M. L. Zahurak,
F. Carnot, D. W. Cham, M. Canto, K. D. Lillemoe, J. L. Cameron,
C. J. Yeo, R. H. Hruban, M. Goggins, Cancer Res. 2002, 62, 1868–1875.
[16] F. Demaugre, Y. Philippe, S. Sar, B. Pileire, L. Christa, C. Lasserre,
C. Brechot, Eur. J. Biochem. 2004, 271, 3812–3820.
[17] M. Cervello, L. Giannitrapani, M. La Rosa, M. Notarbartolo,
N. D’Alessandro, L. Virruso, J. L. Iovanna, G. Montalto, Ann. N Y
Acad. Sci. 2002, 963, 53–58.
0
50
100
150
200
Position (mm)
Figure 3. Radio-TLC of 10-[18F]fluoroethyl-b-D-lactose [18F]-7: Developing solvent;
MeOH:H2O (95:5).
Acknowledgements
[18] C. Cavard, B. Terris, G. Grimber, L. Christa, V. Audard, B. Radenen-
Bussiere, M. T. Simon, C. A. Renard, M. A. Buendia, C. Perret,
Oncogene 2006, 25, 599–608.
[19] K. Drickamer, Curr. Opin. Struct. Biol. 1999, 9, 585–590.
[20] J. L. Iovanna, J. C Dagorn, Biochim Biophys Acta. 2005, 1723, 8–18.
[21] L. Christa, M. Felin, O. Morali, M. T. Simon, C. Lasserre, C. Brechot,
A. P. Seve, FEBS Lett. 1994, 337, 114–118.
This work was supported by the Small Animal Imaging Core
facility Grant, 1 U24 CA126577 01, NIH/NCI; and the NCI CCSG
Core Grant CA 106672.
[22] G. Bormans, A. Verbrugen, J. Label. Compd. Radiopharm. 2001, 44,
417–423.
[23] M. Yun, K. H. Chun, J. E. Shin, J. Oh, Bull. Korean Chem. Soc. 2002,
23, 177–178.
References
[24] Y. Ying, P. Ghosh, L. Guo, A. Pal, U. Mukhapadhyay, Z. Peng,
H. H. Yeh, S. Bertolini, L. G. Flores, D. Young, A. Volgin, S.
Soghomonyan, W. Bornmann, C. Logsdon, M. M. Alauddin, J. G.
Gelovani, Mol. Imaging Biol. 2010; DOI: 10.1007/s11307-010-0408-8.
[25] L. Flores, S. Bertolini, H. H. Yeh, D. Young, U. Mukhapadhyay,
A. Pal, Y. Ying, A. Volgin, A. Shavrin, S. Soghomonyan, W. Tong, W.
G. Bornmann, M. M. Alauddin, C. Logsdon, J. G. Gelovani,
PLoSONE 2009, 4(11):e7977.
[1] National Cancer Institute N, U.S. http://www.cancer.gov/cancer-
topics/types/pancreatic (2010) Pancreatic Cancer (Online).
[2] R. Kubota, K. Kubota, S. Yamada, J. Nucl. Med. 1994, 35, 104–112.
[3] R. J. Hicks, V. Kalff, M. P. MacmManus, J. Nucl. Med. 2001, 42,
1596–1604.
[4] J. Czernin, FDG-PET in breast cancer—a different view of its
clinical usefulness. Mol. Imag. Biol. 2002, 4, 35–45.
[5] J.-S. Ryua1, N. C. Choib, A. J. Fischmana, T. J. Lynchc,
D. J. Mathisend, Lung Cancer 2002, 35, 179–187.
[26] O. T. Schmidt, Methods Carbohydr. Chem. 1962, 1, 349–352.
www.jlcr.org
Copyright r 2011 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2011, 54 233–238