Reliable radiosynthesis of 4-[10B]borono-2-[18F]fluoro-l-phenylalanine with quality assurance for boron neutron capture therapy-oriented diagnosis
Reliable radiosynthesis of 4-[10B]borono-2-[18F]fluoro-l -phenylalanine with quality assurance for boron neutron capture therapy-oriented diagnosis
Kiichi Ishiwata 0 1 2 4
Ryoichi Ebinuma 0 1 2 4
Chuichi Watanabe 0 1 2 4
Kunpei Hayashi 0 1 2 4
Jun Toyohara 0 1 2 4
0 Frontier Laboratories , Koriyama , Japan
1 Research Team for Neuroimaging, Tokyo Metropolitan Institute of Gerontology , Tokyo , Japan
2 Department of Biofunctional Imaging, Fukushima Medical University , Fukushima , Japan
3 Kiichi Ishiwata
4 SHI Accelerator Service , Tokyo , Japan
Objective The aim of this study was to establish a reliable and routine method for the preparation of 4-[10B]borono-2[18F]fluoro-l -phenylalanine (l -[18F]FBPA) for boron neutron capture therapy-oriented diagnosis using positron emission tomography. Methods To produce l -[18F]FBPA by electrophilic fluorination of 4-[10B]borono-l -phenylalanine (l -BPA) with [18F]acetylhypofluorite ([18F]AcOF) via [18F]F2 derived from the 20Ne(d,α)18F nuclear reaction, several preparation parameters and characteristics of l -[18F]FBPA were investigated, including: pre-irradiation for [18F]F2 production, the carrier F2 content in the Ne target, l -BPA-to-F2 ratios, separation with high-performance liquid chromatography (HPLC) using 10 different eluents, enantiomeric purity, and residual trifluoroacetic acid used as the reaction solvent by gas chromatography-mass spectrometry. Results The activity yields and molar activities of l -[18F]FBPA (n = 38) were 1200 ± 160 MBq and 46-113 GBq/mmol, respectively, after deuteron-irradiation for 2 h. Two 5 min pre-irradiations prior to [18F]F2 production for 18F-labeling were preferable. For l -[18F]FBPA synthesis, 0.15-0.2% of carrier F2 in Ne and l -BPA-to-F2 ratios > 2 were preferable. HPLC separations with five of the 10 eluents provided injectable l -[18F]FBPA without any further formulation processing, which resulted in a synthesis time of 32 min. Among the five eluents, 1 mM phosphate-buffered saline was the eluent of choice. The l -[18F]FBPA injection was sterile and pyrogen-free, and contained very small amounts of D-enantiomer (< 0.1% of l -[18F] FBPA), l -BPA (< 1% of l -FBPA), and trifluoroacetic acid (<0.5 ppm). Conclusions l -[18F]FBPA injection was reliably prepared by the electrophilic fluorination of l -BPA with [18F]AcOF followed by HPLC separation with 1 mM phosphate-buffered saline.
l -[18F]FBPA; [18F]F2 production; Quality control; PET; BNCT
Institute of Cyclotron and Drug Discovery Research,
Southern TOHOKU Research Institute for Neuroscience,
7-115 Yatsuyamada, Koriyama 963-8052, Japan
4-[10B]borono-2-[18F]fluoro-l -phenylalanine (l -[18F]FBPA)
was developed in 1991 as a probe for positron emission
tomography (PET) to evaluate in vivo 4-[10B]borono-l
-phenylalanine (l -BPA) used in boron neutron capture therapy
(BNCT) for patients with malignant tumors [
]. Based on
several basic studies that verified the usefulness of l -[18F]
], l -[18F]FBPA PET has been clinically applied
for this purpose [
] and expanded in limited numbers of
PET facilities over the past 20 years [
], mainly because
BNCT is performed using a nuclear reactor for neutron
irradiation. Recently, a cyclotron that acts as an
epithermalneutron source for BNCT and that can be installed in the
hospitals has been developed [
]. Phase I and II clinical
trials of l -BPA BNCT using this cyclotron are in progress
at two institutes in Japan, including the Southern Tohoku
BNCT Research Center at the Southern TOHOKU Research
Institute for Neuroscience. Therefore, the importance of
l -[18F]FBPA PET is increasing, and further basic studies
on the characterization of l -[18F]FBPA have been reported
in recent years [
l -[18F]FBPA has been synthesized by the electrophilic
fluorination of l -BPA with carrier-added [18F]F2 or [18F]
acetylhypofluorite ([18F]AcOF) produced by three different
routes. [18F]F2 was originally produced by deuteron
irradiation of carrier F2-containing Ne, termed the 20Ne(d,α)18F
nuclear reaction [
]. However, the activity yields of [18F]
F2 and the resultant l -[18F]FBPA were low. For example,
Wang et al. prepared 444–518 MBq of l -[18F]FBPA from
5.55 GBq of [18F]F2 after deuteron irradiation for 2 h [
The molar activity of l -[18F]FBPA was also low because of
carrier F2: 30–60 MBq/µmol [
]. The second and third routes
used the 18O(p,n)18F nuclear reaction for carrier-added [18F]
F2 production. In the former, [18F]F2 was produced by proton
irradiation of highly enriched [18O]O2 gas followed by a
second proton irradiation step for the release of [18F]F2 [
In the latter, [18F]fluoride produced by proton irradiation of
[18O]H2O was converted to [18F]F2 via [18F]fluoromethane
. The 18O(p,n)18F reaction can produce potentially large
amounts of 18F compared with the 20Ne(d,α)18F reaction;
therefore, the activity yields of l -[18F]FBPA synthesized
via the second route (2 GBq [
] to 5.3 GBq [
]), and the
molar activity (257 MBq/µmol [
]) has been improved. The
third route was especially devised to give less carrier-added
[18F]F2. Consequently, the molar activity of l -[18F]FBPA
was the highest (3700 MBq/µmol), but the activity yields
have not been clearly described [
In the future, the synthesis of l -[18F]FBPA by
nucleophilic fluorination using no-carrier-added [18F]fluoride will
be developed to obtain higher activity yields and higher
molar activities of l -[18F]FBPA, as the synthesis of
2-deoxy2-[18F]fluoro-d -glucose has progressed from the method
using [18F]F2 to that using no-carrier-added [18F]fluoride.
However, at present such radiosynthesis is still under
development, although a preliminary synthesis was reported
At the present stage of l -[18F]FBPA PET for l -BPA
BNCT-oriented diagnosis, l -[18F]FBPA PET is applied to
only a few patients per l -[18F]FBPA preparation and not to
mass screening; therefore, a steady and reliable synthesis
of l -[18F]FBPA is required. The molar activity of l -[18F]
FBPA is not a critical issue. Among the three methods of
l -[18F]FBPA synthesis described previously, Ne-derived
[18F]F2 is produced simply and cost-effectively compared
with 18O-derived [18F]F2. Therefore, the original method of
l -[18F]FBPA synthesis has been adapted clinically to date;
however, detailed procedures, the optimization of each
process, and technical knowhow have not been described
sufficiently in previous reports on the three methods [
]. In the present study, we aimed to elaborate on
the original method from the viewpoint of routine clinical
For this purpose, steady production of [18F]F2 is first
essential. It is well known empirically that a short
preirradiation step is essential before the main irradiation for
18F-labeling; however, systematic studies on [18F]F2
production have not been reported. A larger amount of carrier F2 in
Ne target, for example 0.5% F2, has a benefit for the steady
production of [18F]F2 but provides low molar activity, and
the stoichiometric relationship in the fluorination of l -BPA
with [18F]F2/[18F]AcOF should be considered carefully. For
electrophilic fluorination of l -BPA, the original method
used [18F]AcOF due to its higher selectivity than [18F]F2
], whereas the second and third methods employed [18F]F2
probably to avoid activity loss of activity during the
conversion process from [18F]F2 to [18F]AcOF [
the formulation of l -[18F]FBPA, the most popular
preparation method is purification by high-performance liquid
chromatography (HPLC) using a reversed-phase column
with 0.1% aqueous AcOH as the mobile phase followed by
evaporation of the l -[18F]FBPA fraction and re-dissolution
in physiological saline [
]. To avoid this time-consuming
evaporation process, Vähätalo et al. separated l -[18F]FBPA
by HPLC using physiological saline containing 1–2% EtOH
and 0.01% AcOH as the eluent, and the l -[18F]FBPA fraction
was used directly for injection in clinical studies [
however, the pH of this injection was not described, although it
appeared to be below 4. Prior to this, Ishiwata et al. proposed
HPLC separation with physiological saline alone without
clinical use [
In the present study, we investigated (1) the importance
of pre-irradiation for [18F]F2 production with an appropriate
F2 carrier, (2) steady production of l -[18F]FBPA in relation
to F2 content and l -BPA, (3) HPLC separation methods to
provide injectable l -[18F]FBPA without any further
formulation processing, (4) the optical purity of l -[18F]FBPA, and
(5) analysis of residual trifluoroacetic acid (TFA) used as
a solvent in radiosynthetic preparation of l -[18F]FBPA. To
the best of our knowledge, no report on points (4) and (5)
has been published previously. Findings for the three other
points would also provide useful information for the
radiosynthesis of l -[18F]FBPA using 18O-derived [18F]F2.
Labeled compounds and related terms are expressed
according to the International Consensus Radiochemistry
Nomenclature Guidelines recently recommended by an
international Working Group on ‘Nomenclature in
Radiopharmaceutical Chemistry and related areas’ [
Materials and methods
l -BPA was purchased from Sigma-Aldrich Chemical (St
Louis, MO). l -BPA, l -FBPA, and d -FBPA were kindly
supplied by Stella Pharma (Osaka, Japan). 2-, 3-, and
4-fluoro-d ,l -phenylalanine (2-, 3-, and 4-FPhe,
respectively) were purchased from Tokyo Chemical Industry
(Tokyo, Japan). Normal saline (500 and 1000 mL
plastic bag), distilled water (20 ml plastic ampule and 500 ml
plastic bottle) for injection, and sodium phosphate
corrective injection 0.5 mmol/ml (pH 6.5, 20 ml plastic ampule)
were purchased from Otsuka Pharmaceutical (Tokyo,
Japan). Other chemical reagents were obtained from
Production of [18F]F2
An 18 MeV cyclotron (CYPRIS HM-18, 18 MeV protons
and 9 MeV deuterons, Sumitomo Heavy Industries, Tokyo,
Japan) was employed. Elemental [18F]F2 was produced via
the 20Ne(d,α)18F reaction in Ne containing F2 in a
cylindrical target chamber [30 mm inner diameter (i.d.) and
242 mm length] made of aluminum. The incident deuteron
energy was 7.9 MeV. All deuteron irradiation processes
were performed at a fixed current of 20 µA.
Conditioning production of [18F]F2
To determine a suitable pre-irradiation protocol before
the main [18F]F2 production for 18F-labeling, three or
four successive 5 min irradiations were conducted within
1–39 day intervals. The content of F2 in Ne was set in the
same range in each experiment: 0.1% (v/v) (n = 7), 0.15%
(n = 7), and 0.2% (n = 5) by mixing 5% F2-containing Ne
and pure Ne gases. The final pressure was set at 380 kPa.
However, the actual F2 concentrations calculated from the
pressure were determined to have certain ranges. After
the end of the 5 min irradiation, the [18F]F2 produced was
recovered with a maximum flow rate by the target pressure
and absorbed into a tandem column of soda lime (No.1,
Wako Pure Chemicals, Osaka, Japan, 9 mm i.d. × 80 mm
length) and activated charcoal (Granular, Wako Pure
Chemicals, 9 mm i.d. × 80 mm length). The average flow
rates from maximum pressure (ca. 380 kPa) to 100 kPa
were in range of 756–845 ml/min. The activity absorbed
in the tandem columns was estimated as the total activity
recovered from [18F]F2 production, and corrected for decay
to the end of cyclotron bombardment (EOB).
Synthesis of l ‑[ 18F]FBPA
l -[18F]FBPA (total 38 runs) was prepared by electrophilic
fluorination with [18F]AcOF using a multipurpose
synthesizer (CFN-MPS200, Sumitomo Heavy Industries) using
a method that was a slightly modified method from
previous reports [
]. Two conditioning 5 min irradiations
were performed, and a main irradiation for 18F-labeling
was performed for 90–156 min (123 ± 17 min). In all three
irradiations, the content of F2 in Ne was set at the same
percentage of 0.10–0.30% (30–89 µmol: 0.10%, n = 3;
0.15%, n = 8; 0.20%, n = 23; 0.25%, n = 2; and 0.30%,
n = 2). The [18F]F2 produced was passed through a column
containing sodium acetate trihydrate or sodium acetate
anhydrous (4 mm i.d. × 40 mm length), and the resultant
[18F]AcOF was bubbled into 4 ml TFA containing l -BPA
at room temperature with maximal flow rates: 426± 68 ml/
min from the target pressure (ca. 380 kPa) to 100 kPa. To
examine the effect of the l -BPA-to-F2 ratios on l -[18F]
FBPA synthesis, the amount of l -BPA was varied in the
range 14.2–33.3 mg (68–160 µmol: 14.2–14.8 mg, n = 2;
18.5 mg, n = 1; 25.1–25.3 mg, n = 3; and 29.3–33.3 mg,
n = 32). The total (100%) recovered from [18F]F2
production was estimated as the summed activities of the sodium
acetate column and the TFA solution. The radioactivity
sensor using to monitor a reaction vial containing the
TFA solution was calibrated according to the standard
18F-activity measured with a dose calibrator (CRC-15
PET, Capintec, Florham Park, NJ, USA).
The TFA solution was heated to 120 °C, and the TFA
was removed using a 200 ml/min N2 flow. The residue was
dissolved in 2 ml of the eluent used for preparative HPLC
(described below). The solution was applied to HPLC
separation, and the fraction with l -[18F]FBPA was obtained.
The volumes of the l -[18F]FBPA fractions were estimated
by weight (1.0 g = 1.0 ml), and the pH was measured using
a pH meter (Laqua act, Horiba Scientific, Tokyo, Japan).
In four of 38 runs, the l -[18F]FBPA fraction was collected
through a 0.22 µm membrane filter (SLGVJ33RS, Merck
Millipore, Darmstadt, Germany) for clinical purposes, and
the sterility and apyrogenicity were examined. Filter
integrity (> 150 kPa) was evaluated using a Millex/Sterivex
integrity tester (Merck Millipore). The activity yields of l -[18F]
FBPA at the end of synthesis (EOS) were normalized with
respect to those produced by irradiation for 120 min.
HPLC separation of l ‑[ 18F]FBPA
The column used for HPLC separation of l -[18F]FBPA was
YMC-Pack ODS-A (S-5 µm, 20 nm, 20 mm i.d. × 150 mm
length, YMC, Kyoto, Japan) with a guard cartridge ODS-A
(S-5 µm, 12 nm, 20 mm i.d. × 10 mm, YMC). The 10
different mobile phases investigated are summarized in Table 1
Data are average ± standard deviation
al -[18F]FBPA obtained at the end of synthesis was normalized to that produced by 120-min irradiation
bRadiochemical purity (RCP) was determined based on HPLC analysis
cContamination (moles) of 4-[10B]borono-l -phenylalanine (l -BPA) is expressed as a percentage against the mass of l -FBPA.
dn = 2
(eluents 1–10). Eluents of 10 and 5 mM phosphate-buffered
saline (PBS) were prepared by mixing normal saline, sodium
phosphate corrective injection 0.5 mmol/ml (pH 6.5), and
distilled water to an isotonic ion strength of 0.15 mEq/ml.
Eluent containing 1 mM PBS was prepared by adding 1/500
volume of sodium phosphate corrective injection 0.5 mmol/
ml (pH 6.5) into normal saline. The flow rate was 10 ml/min,
and the elution profile was monitored using an ultraviolet
(UV, 260 nm) detector (UV 2715 Plus, Jasco, Tokyo Japan)
and a radioactivity monitor (UG-PD1A, Universal Giken,
Odawara, Japan). First, in the separation with eluent 1, based
on previous reports [
], a major radioactive peak, and
later other minor radioactive peaks and shoulder
components were fractionated, l -[18F]FBPA and three by-products
of 2-, 3-, and 4-[18F]fluoro-l -phenylalanine were identified
by comparison of their retention times with those of the
authentic compounds (enantiomeric mixtures in the case of
fluorophenylalanines) in the HPLC analysis described below.
HPLC analysis of l ‑[ 18F]FBPA
The column used was YMC-UltraHT Pro C18 S-2 µm
(3.0 mm i.d. × 100 mm length, YMC). Four different
mobile phases were investigated: (a) 50 mM AcOH/50 mM
AcONH4 (1/1), (b) 50 mM NaH2PO4, (c) 0.1% AcOH, and
(d) 0.8% AcOH containing 1 mM ethylenediaminetetraacetic
acid (EDTA) and 1 mM sodium octylsulfate [similar eluents
as described in Refs. 1, 23, 25] at a flow rate of 0.5 ml/
min at 20–21 °C, and the elution profiled was monitored
using a UV detector at 280 nm (SPD-20A Prominence UV/
VIS detector, Shimadzu, Tokyo, Japan) and a radioactivity
monitor (US-3000, Universal Giken). The retention times of
l -BPA and l -FBPA were 6.0 and 8.5 min, 6.1 and 8.7 min,
5.5 and 7.7 min, and 4.4 and 6.1 min with eluents a, b, c,
and d, respectively. The retention times of 2-, 3-, and 4-FPhe
were 8.5, 12.5, and 12.7 min, and 12.5, 14.1, and 19.9 min
with eluents a and d, respectively.
Optical purity of l ‑[ 18F]FBPA
A Crownpak CR (+) (4.0 mm i.d. × 150 mm length, Daicel,
Tokyo, Japan) column was used with a mobile phase of
HClO4 (pH 2.0) at a flow rate of 1.0 ml/min at 20–21 °C.
The retention times of d -FBPA and l -FBPA were 6.5 and
9.9 min, respectively.
Measurement of TFA in l ‑[ 18F]FBPA
Gas chromatography-mass spectrometry (GC-MS) was
applied for the analysis of residual TFA in eight l -[18F]
FBPA preparations: eluents 1, n = 1; 4, n = 5; 9, n = 1; and
10, n = 1. 0.1 ml of concentrated H2SO4/MeOH (4/1) was
added to a 0.5 ml l -[18F]FBPA sample, and the mixture was
shaken vigorously for 30 s. 0.4 ml of CH2Cl2 was then added
to the mixture followed by a 15 s extraction with methyl
trifluoroacetate. The CH2Cl2 phase solution (1 µl) was applied
to GC–MS analysis 1 min after the end of extraction.
Standard aqueous TFA (0.1–100 ppm) was also treated in the
same way, and a calibration curve of methyl trifluoroacetate
A quadruple mass spectrometer (5975C, Agilent
Technologies, Santa Clara, CA) in conjunction with a gas
chromatograph (5890GC, Agilent Technologies) was used
with a deactivated metal capillary column (0.25 mm i.d. ×
30 m) with 1 µm film thickness of Ultra ALLOY-CW
(Frontier Laboratories, Koriyama, Japan). The oven temperature
was maintained at 40 °C for 3 min and then increased to
150 °C at 100°C/min and held there for 3 min. The split/
splitless injector and the transfer line were kept at 200 °C.
The flow rate of He carrier gas was 1.5 ml/min.
Results and discussion
Conditioning production of [18F]F2
Figure 1 shows a plot of the activity yields of [18F]F2
after each of three or four successive 5 min irradiations
in experiments for the conditioning production of [18F]F2
together with the yields in each of two conditioning 5 min
irradiations for the l -[18F]FBPA synthesis as a function of
the intervals of each irradiation day. The calculated actual
F2 concentrations for 0.10%, 0.15%, and 0.20% F2/Ne targets
were 0.11 ± 0.01% (range 0.08–0.14%, n = 27), 0.15 ± 0.02%
(0.10–0.19%, n = 38), and 0.20 ± 0.02% (0.17–0.25%,
n = 65), respectively. The activity yields were variable after
the first irradiation regardless of the F2 content or interval
days. For 0.10% and 0.15% F2/Ne targets, the third
irradiations produced almost steady-state yields. With 0.20% F2/
Ne, the yields reached steady state with the second
irradiation. It is noted that steady-state yields are useful from
a practical perspective but did not mean the quantitative
recovery of [18F]F2 produced with each irradiation even if a
higher concentration of F2 was employed. The relationship
Fig. 1 Effect of irradiation times and irradiation interval on the
production of [18F]F2. a–c Show the activity yields (GBq) of [18F]F2 for
5 min irradiation of Ne containing 0.1%, 0.15%, and 0.2% F2,
respectively, and –(1), -(2), -(3), and (4) indicate the first, second, third, and
fourth 5 min irradiations, respectively
Fig. 2 Effects of reaction conditions on the production of
[18F]acetylhypofluorite ([18F]AcOF) and 4-[10B]borono-2-[18F]fluoro-l
-phenylalanine (l -[18F]FBPA). a Relationship between F2 content (%) in Ne
and activity yield of [18F]AcOF (GBq). b Relationship between F2
content (%) in Ne and activity yield of l -[18F]FBPA (GBq). c
Relationship between 4-[10B]borono-l -phenylalanine (l -BPA)-to-F2 ratio
(molar ratio) and radiochemical yield (RCY) of l -[18F]FBPA (%).
d Relationship between F2 (mole) in Ne and F2 (mole) recovered
as [18F]F2. The amount (mole) of [18F]F2 was calculated from the
molar activity of l -[18F]FBPA and the total activity (summed activity
absorbed in a sodium acetate column and recovered in a reaction vial
between the molar mass of F2 set and that of recovered is
shown later (Fig. 2d).
In an early study on [18F]F2 production, a carrier
concentration of 0.1% F2 produced almost 95% of the
theoretical yield of [18F]F2, and a part of [18F]F2 was adsorbed on
a stainless steel tube in recovery from the target chamber
]. The present study demonstrated the clear requirement
of pre-irradiation in a range of 0.1–0.2% F2, and suggested
that two pre-conditioning irradiations with these F2
concentrations were preferable for steady [18F]F2 production. It was
also suggested that mixing 5% F2-containg Ne and pure Ne
gases made accurate setting of the F2 concentration difficult.
Synthesis of l ‑[ 18F]FBPA using [18F]AcOF
The total activity of [18F]F2 (summed activities absorbed
in a sodium acetate column and recovered in a reaction vial
as [18F]AcOF) that was normalized as those produced by
irradiation for 120 min, was 12.3 ± 1.9 GBq (n = 38). The
activity yields of [18F]AcOF trapped in the reaction vial
tended to increase with increasing the F2 percentage in Ne,
and ≥ 0.15% F2 was preferable (Fig. 2a). The flow rates of
[18F]F2 passing through a sodium acetate trihydrate column
were slightly higher than those passing through a sodium
acetate anhydrous column: 460 ± 89 ml/min (n = 23) vs.
363 ± 55 ml/min (n = 11, 4 data missing), and the
respective radiochemical yields (RCYs) of [18F]AcOF were
38.8 ± 2.7% (n = 23) and 35.5 ± 1.9% (n = 11, 4 data missing)
based on the total activity of [18F]F2 recovered. The
difference in RCYs between the two cases was not large, but two
other experiments using a sodium acetate anhydrous column
with flow rates of 252 and 319 ml/min produced very low
[18F]AcOF RCYs of 10.9 and 15.9%, respectively. These
results suggested that low flow rates of [18F]F2 decreased the
recovery of [18F]AcOF, and that sodium acetate trihydrate
with larger particle sizes would be preferable compared with
sodium acetate anhydrous with smaller particle sizes.
The activity yields of l -[18F]FBPA were variable and did
not tend to increase with the F2 percentage in Ne (Fig. 2b).
The averaged activity yield was 1200 ± 160 MBq (n = 38)
at 31.6 ± 1.7 min from the EOB. The RCY of l -[18F]
FBPA based on [18F]AcOF trapped in the reaction vial was
33.1 ± 3.8% (n = 38). The RCY increased with
increasing l -BPA-to-F2 ratios (Fig. 2c), which suggested that the
increased [18F]AcOF relative to l -BPA further fluorinated
to produce 18F-difluorinated l -BPA as observed in the
electrophilic fluorination of l -3-(hydroxy-4-pivaloyloxyphenyl)
alanine with [18F]AcOF [
] and/or degraded l -[18F]FBPA.
The preferred l -BPA-to-F2 ratio is > 2 for the steady
production of l -[18F]FBPA. It is emphasized that the synthesis time
(32 min) was the shortest compared with those in previous
reports: 50 min [
], 72 min [
], 80 min [
], 88 min [
and 110 min [
As previously described [
], three by-products of 2-,
3-, and 4-[18F]fluoro-l -phenylalanine were tentatively
identified by comparison with the retention times of authentic
samples. Although baseline separation of 3- and 4-[18F]
fluoro-l -phenylalanine could not be performed in all HPLC
separations investigated, the relative amounts were in the
order of 3- > 4- > 2-isomer (Fig. 3), and the summed RCYs
of the three were constant at 11.6 ± 1.3% (n = 37).
Electrophilic fluorination at the aromatic carbon 4 could explain
undesired deboronation that leads to the 4-isomer;
however, there is no known mechanism that produces the 2- and
3-isomers. Coenen et al. reported that the fluorination of
l -phenylalanine in TFA with [18F]F2 produced 2- (72.5%),
3(13.9%), and 4-[18F]fluoro-l -phenylalanine (13.6%) . It
is unlikely that l -phenylalanine produced after deboronation
260 nm) absorbance, respectively. The unit of the vertical axis are
the millvolt output of the UV detector. Elution positions of 4-[10B]
borono-l -phenylalanine (l -BPA), l -FBPA, and 2-, 3-, and
4-fluorod ,l -phenylalanine (2-, 3-, and 4-FPhe, respectively) are indicated
of l -BPA was fluorinated. We tried further identification of
these byproducts by GC-MS as described for the
determination of TFA using a deactivated metal capillary column
(0.25 mm i.d. × 30 m) with 1 µm film thickness of Ultra
ALLOY-1 (polydimethylsiloxane) (Frontier Laboratories),
but could neither identify nor deny three byproducts to be
2-, 3-, and 4-[18F]fluoro-l -phenylalanine.
The molar activities of l -[18F]FBPA synthesized using
0.1%, 0.15%, 0.2% 0.25%, and 0.3% F2 were 103.5 ± 9.5
(n = 3), 86.1 ± 24.4 (n = 8), 69.0 ± 7.3 (n = 23), 66.0 (n = 2),
and 50.3 GBq/mmol (n = 2), respectively. It is
reasonable that lower carrier F2 contents in [18F]F2 production
resulted in higher molar activities of l -[18F]FBPA;
however, no linear relationship was found between the F2
content and molar activity. Figure 2d shows the molar amounts
of recovered [18F]F2 that were calculated from the molar
activity of l -[18F]FBPA and the total activity. Large
differences observed between the amounts of F2 set in Ne and
the recovered F2 indicated that the carrier F2 added could
not be recovered constantly, even after two pre-conditioning
Discussion of the differences between the present and
previous studies on l -[18F]FBPA synthesis using Ne-derived
[18F]AcOF is difficult because the detailed reaction
conditions were not described in the previous reports. However,
the present activity yields (1200 ± 160 MBq) were much
higher than in previous reports: 444–518 MBq (n = 10) [
and 750 ± 250 MBq (n = 8) [33 figures not shown], and the
molar activities were higher than those in some reports [
] but lower than that reported in [
] (130 GBq/mmol).
Fig. 3 shows HPLC separation patterns with 10 different
eluents. Eluent 1 0.1% AcOH, a standard mobile phase used
previously, gave a baseline separation (Fig. 3a); however,
lower 0.01% AcOH (Fig. 3b) was preferable. Physiological
saline (Fig. 3c) showed leading peaks of l -BPA and l -[18F]
FBPA, but did not show baseline separation. The addition of
AcOH to saline (Fig. 3d) improved the separation slightly.
Further addition of EtOH (Fig. 3e) caused slightly faster
elution but without improved separation. The addition of
sodium phosphate corrective injection 0.5 mmol/ml (pH
6.5) to saline with/without AcOH improved the separation
(Fig. 3f, h, i). Lower sodium phosphate eluted l -BPA more
broadly, but l -[18F]FBPA was separated as an apparently
single peak (Fig. 3g, j).
The characteristics of the 10 l -[18F]FBPA preparations
are summarized in Table 1. First, the radiochemical purities
(RCPs) of the four eluents for HPLC analysis were
compared for five l -[18F]FBPA preparations. In analyses with
eluents (a) 50 mM AcOH/50 mM AcONH4 (1/1) (Fig. 4a),
(b) 50 mM NaH2PO4, (c) 0.1% AcOH, and (d) 0.8% AcOH
containing 1 mM EDTA and 1 mM sodium octylsulfate,
the RCPs were 97.7 ± 1.3%, 97.7 ± 1.4%, 98.6 ± 0.6%, and
98.4 ± 0.6%, respectively. Analyses with eluents (a) and (b)
were similar, and preferable compared to eluents (c) and
(d). Therefore, all subsequent analyses were conducted with
The RCPs were over 97% for all 10 l -[18F]FBPA
preparations. The pH was below 4.0 in four of the 10 preparations.
The volumes of l -[18F]FBPA separated with saline with/
without 0.01% AcOH were over 20 ml, and the addition of
sodium phosphate corrective injection reduced the volumes.
Baseline separation could not be performed using eluents
3–5 and 1–2% contamination levels of l -BPA were obtained.
Contamination levels of l -BPA in five other preparations
with eluents 6–10 were very low. These five preparations
were thus considered as acceptable for intravenous
injection without any further processing. The l -[18F]FBPA eluted
with 0.1% AcOH (eluent 1) was previously used directly in a
clinical study after the addition of 25% ascorbic acid
injection and 10% sodium chloride injection to manage the pH
and ion strength, respectively [
]. Comparison of activity
yields and molar activities in the 10 preparations shown in
Retention time (min)
Retention time (min)
Fig. 4 HPLC chromatograms of 4-[10B]borono-2-[18F]fluoro-l
-phenylalanine (l -[18F]FBPA) analysis on a YMC-UltraHT Pro C18
S-2 µm and b Crownpak CR (+) columns. Red and blue lines show
the elution profiles for radioactivity and ultraviolet (UV, 277 nm)
absorbance, respectively. The units of the vertical axis are millivolt
output of the UV detector. a In the stability test for 4 h, a
radiochemical impurity appeared at the retention time indicated by the arrow. b
The inset emphasizes the low levels of the chromatograms and the
elution position of d -FBPA is indicated by the arrow
Data are average ± standard deviation
aThe time for the first analysis after the end of synthesis was defined as “0 h”, and then analyses were performed successively at approximately
the indicated intervals until approximately 4 h
Table 2 may not be significant, because the F2 contents were
From these results and the stability of l -[18F]FBPA
described later, eluent 10 was selected for routine clinical
use because of the small amount of one additive in
physiological saline and ease of eluent preparation. The
clinical injection volumes of l -[18F]FBPA separated with
eluent 10 (1120 MBq/13.2 ml, Table 2), were expected to be
2.6–3.9 ml/60 kg when the radioactive injection doses in
l -[18F]FBPA PET were 3.7–5.55 MBq/kg [
], and one
l -[18F]FBPA preparation can be used for 2 subjects with
one PET scanner or 3–4 subjects with two PET scanners. In
four l -[18F]FBPA preparations separated with eluent 10 and
collected through a 0.22 µm membrane filter, sterility and
apyrogenicity (< 0.0029 EU/ml) of the l -[18F]FBPA
injection and filter integrity (≥330 kPa) were confirmed. It is
noted that no radionuclidic impurities were found in l -[18F]
FBPA prepared using the present method .
In previous research to improve the activity yields or the
molar activity of l -[18F]FBPA, l -BPA was fluorinated with
[18F]F2 produced by the 18O(p,n)18F reaction [
also fluorinated l -BPA with [18F]F2 and separated l -[18F]
FBPA by HPLC with eluents 1, 6, and 8; however, the RCPs
(n = 6) were lower than those of [18F]AcOF-derived l -[18F]
FBPA separated using the same eluents. The higher
reactivity of [18F]F2 than [18F]AcOF may result in more side
reactions. Therefore, no further investigation was conducted for
the synthesis using [18F]F2.
The stabilities of five l -[18F]FBPA preparations are
summarized in Table 2. The l -[18F]FBPA separated by eluent 7 was
the most stable over 4 h after EOS. In the four other
preparations the RCPs decreased gradually but were maintained at
over 96% for 4 h. Low pH with eluent 7 may contribute to
the stability of l -[18F]FBPA; however, these five
preparations without stabilizers such as EtOH [
] and ascorbate
] were suitable for routine clinical use for at least 4 h.
Preliminarily we found that the ascorbate was not effective
for stability. In two l -[18F]FBPA preparations separated
with 0.01% AcOH saline and 5 mM PBS, the addition of
ascorbate injection (final concentration of 10 mg/ml) slightly
decreased the RCPs from 98.4% and 97.7–93.2% and 91.3%,
respectively, by approximately 4 h.
No signal associated with the d -isomer was found in l -[18F]
FBPA preparations, probably because of the low activity
concentrations (Table 1, 47–107 MBq/mL) and the
sensitivity of the US-3000 radioactivity detector. However, in the
l -[18F]FBPA samples concentrated by evaporation or only
the peak fraction of l -[18F]FBPA from HPLC separation
(270–970 MBq/ml), a very small amount of the d -isomer
was present (Fig. 4b): < 0.1% (0.09 ± 0.04%, n = 4)
compared with the l -isomer. The UV peak that corresponds to
this activity peak was increased by the addition of small
amounts of standard d -FBPA into the samples. It was noted
that a temperature at 20–21 °C was critical in this analysis
because d -[18F]FBPA was not separated from the radioactive
impurities at higher temperature (25 °C).
A UV peak corresponding to d -FBPA was evident and
the UV detection sensitivity was higher than that of
radioactivity. However, UV signals could not be used to
evaluate d -[18F]FBPA, because the retention times of authentic
d -FBPA and l -BPA coincided. No UV peak that
corresponded to the d -isomer was observed for the starting
compound l -BPA (both from Sigma-Aldrich and Stella Pharma);
therefore, we considered that epimerization of l -[18F]FBPA
might occur in TFA solution in the radiosynthesis process,
as l -amino acids such as l -phenylalanine were epimerized
in acetic acid [
TFA analysis by GC‑MS
The residual TFA in l -[18F]FBPA preparations was analyzed
by GC-MS after methylation. The retention time of methyl
trifluoroacetate in the GC stage was 2.5 min. Three ions
were monitored in the SIM-EI+ mode of operation: m/z 59,
69, and 99. These are the most characteristic ions in the
mass spectrum of methyl trifluoroacetate [
]. Because the
signal at m/z 69 was much stronger than those at m/z 59 and
99; therefore, only the area of the m/z 69 peak was used for
quantitative analysis. Methyl trifluoroacetate was degraded
gradually by approximately 10% after 20 min from the end
of extraction; therefore, the GC-MS analysis was started
just 1 min after the end of extraction. The detection limit of
methyl trifluoroacetate was determined to be 0.5 ppm
[signal-to-nosise ratio (S/N) = 12], and the S/N of the 0.1 ppm
standard sample was 8. In eight l -[18F]FBPA preparations
the residual TFA was less than 0.5 ppm: 0.2 ± 0.1 ppm
(range, 0.0–0.3 ppm) without evaporation processing for
preparing the l -[18F]FBPA injection. It is noted that the
LD50 values are 200 mg/kg in rats (oral administration) and
1200 mg/kg in mice (intravenous injection) (Hazardous
Substances Data Bank, 2007: https://toxnet.nlm.nih.gov/cgi-bin/
Two 5 min pre-irradiations enabled the steady production
of [18F]F2 for 18F-labeling by electrophilic fluorination. To
achieve a high RCY of l -[18F]FBPA 0.15–0.2% carrier F2
in Ne and an l -BPA-to-F2 ratio > 2 were preferable. HPLC
separation using 1 mM PBS provided injectable l -[18F]
FBPA without any further formulation processing, which
resulted in a 32-min synthesis period from EOB. The l -[18F]
FBPA injection contained small amounts of d -enantiomer
(< 0.1% of l -[18F]FBPA), l -BPA (< 1% of l -FBPA), and
TFA (< 0.5 ppm).
Open Access This article is distributed under the terms of the
Creative Commons Attribution 4.0 International License (http://creativeco
mmons.org/licenses/by/4.0/), which permits unrestricted use,
distribution, and reproduction in any medium, provided you give appropriate
credit to the original author(s) and the source, provide a link to the
Creative Commons license, and indicate if changes were made.
1. Ishiwata K , Ido T , Mejia AA , Ichihashi M , Mishima Y. Synthesis and radiation dosimetry of 4-borono-2-[18F]fluoro-d ,l - phenylalanine: a target compound for PET and boron neutron capture therapy . Appl Radiat Isot . 1991 ; 42 : 325 - 8 . https://doi. org/10.1016/ 0883 - 2889 ( 91 ) 90133 -L.
2. Ishiwata K , Ido T , Kawamura M , Kubota K , Ichihashi M , Mishima Y. 4 -Borono-2-[18F] fluoro -d ,l -phenylalanine as a target compound for boron neutron capture therapy: tumor imaging potential with positron emission tomography . Nucl Med Biol . 1991 ; 18 : 745 - 51 . https://doi.org/10.1016/ 0883 - 2897 ( 91 ) 90013 - B .
3. Ishiwata K , Ido T , Honda C , Kawamura M , Ichihashi M , Mishima Y. 4 -Borono-2-[18F] fluoro -d ,l -phenylalanine: a possible tracer for melanoma diagnosis with PET . Nucl Med Biol . 1992 ; 19 : 311 - 8 . https://doi.org/10.1016/ 0883 - 2897 ( 92 ) 90116 -G.
4. Ishiwata K , Shiono M , Kubota K , Yoshino K , Hatazawa J , Ido T , Honda C , et al. A unique iv vivo assessment of 4-[10B]boronol -phenylalanine in tumour tissues for boron neutron capture therapy of malignant melanoma using positron emission tomography and 4-borono-2-[18F]fluoro-l -phenylalanine . Melanoma Res . 1992 ; 2 : 171 - 9 .
5. Kubota R , Yamada S , Ishiwata K , Tada M , Ido T , Kubota K. Cellular accumulation of 18F-labelled boronophenylalanine depending on DNA synthesis and melanin incorporation: a double-tracer microautoradiographic study of B16 melanomas in vivo . Br J Cancer . 1993 ; 67 : 701 - 5 .
6. Imahori Y , Ueda S , Ohmori Y , Yoshino E , Ono K , Kobayashi T , et al. A basic concept for PET-BNCT system . In: Mishima Y (ed) Cancer neutron capture therapy . New York: Plenum Press; 1996 . pp. 691 - 6 .
7. Ueda S , Imahori Y , Ohmori Y , Yoshino E , Ono K , Kobayashi T , et al. Positron emission tomography and boron neutron capture therapy system to the patient with malignant brain tumor: the first clinical trial using 10B-BPA . In: Mishima Y, editor. Cancer neutron capture therapy . New York: Plenum Press; 1996 . pp. 823 - 7 .
8. Mishima Y , Imahori Y , Honda C , Hiratsuka J , Ueda S , Ido T. In vivo diagnosis of human malignant melanoma with positron emission tomography using specific melanoma-seeking 18F-DOPA analogue . J Neuro Oncol . 1997 ; 33 : 163 - 9 .
9. Imahori Y , Ueda S , Ohmori Y , Kusuki T , Ono K , Fujii R , et al. Fluorine-18 -labeled fluoroboronophenylalanine PET in patients with glioma . J Nucl Med . 1998 ; 39 : 325 - 3 .
10. Miyatake S , Kawabata S , Hiramatsu R , Kuroiwa T , Suzuki M , Kondo N , et al. Boron neutron capture therapy for malignant brain tumors . Neurol Med Chir (Tokyo). 2016 ; 56 : 361 - 71 . https://doi. org/10.2176/nmc.ra.2015- 0297 .
11. Evangelista L , Jori G , Martini D , Sotti G . Boron neutron capture therapy and 18F-labelled borophenylalanine positron emission tomography: A critical and clinical overview of the literature . Appl Radiat Isot . 2013 ; 74 : 91 - 101 . https://doi.org/10.1016/j.aprad iso. 2013 . 01 .001.
12. Nariai T , Ishiwata K. Analysis and imaging: PET . In: Sauerwein WAG , Wittig A , Moss R , Nakagawa Y , editors. Neutron capture therapy, principles and applications . Berlin: Springer; 2012 . pp. 201 - 12 . https://doi.org/10.1007/978-3- 642 -31334-9_ 11 .
13. Wittig A , Michel J , Moss RL , Stecher-Rasmussen F , Arlinghaus HF , Bendel P , Mauri PL , Altieri S , Hilger R , Salvadori PA , Menichetti L , Zamenhof R , Sauerwein WA . Boron analysis and boron imaging in biological materials for boron neutron capture therapy (BNCT) . Crit Rev Oncol Hematol . 2008 ; 68 : 66 - 90 . https ://doi.org/10.1016/j.critrevonc. 2008 . 03 .004.
14. Tanaka H , Sakurai Y , Suzuki M , Masunaga S , Kinashi Y , Kashino G , et al. Characteristics comparison between a cyclotron-based neutron source and KUR-HWNIF for boron neutron capture therapy . Nucl Instrum Methods Phys Res B . 2009 ; 267 : 1970 - 7 . https://doi.org/10.1016/j.nimb. 2009 . 03 .095.
15. Tanaka H , Sakurai Y , Suzuki M , Masunaga S , Mitsumoto T , Fujita K , et al. Experimental verification of beam characteristics for cyclotron-based epithermal neutron source (C-BENS) . Appl Radiat Isot . 2011 ; 69 : 1642 - 5 . https://doi.org/10.1016/j. apradiso. 2011 . 03 .020.
16. Yoshimoto M , Kurihara H , Honda N , Kawai K , Ohe K , Fujii H , et al. Predominant contribution of L-type amino acid transporter to 4-borono-2-18F-fluoro-phenylalanine uptake in human glioblastoma cells . Nucl Med Biol . 2013 ; 40 : 625 - 9 . https://doi. org/10.1016/j.nucmedbio. 2013 . 02 .010.
17. Hanaoka K , Watabe T , Naka S , Kanai Y , Ikeda H , Horitsugi G , et al. FBPA PET in boron neutron capture therapy for cancer: prediction of 10B concentration in the tumor and normal tissue in a rat xenograft model . EJNMMI Res . 2014 ; 4 : 70 . https://doi. org/10.1186/s13550-014-0070-2.
18. Yang FY , Chang WY , Li JJ , Wang HE , Chen JC , Chang CW . Pharmacokinetic analysis and uptake of 18F-FBPA-Fr after ultrasound-induced blood-brain barrier disruption for potential enhancement of boron delivery for neutron capture therapy . J Nucl Med . 2014 ; 55 : 616 - 21 . https://doi.org/10.2967/jnume d. 113 .125716.
19. Watanabe T , Hattori Y , Ohta Y , Ishimura M , Nakagawa Y , Sanada Y , et al. Comparison of the pharmacokinetics between l -BPA and l -FBPA using the same administration dose and protocol: a validation study for the theranostic approach using [18F]-l -FBPA positron emission tomography in boron neutron capture therapy . BMC Cancer . 2016 ; 16 : 859 . https://doi. org/10.1186/s12885-016-2913-x.
20. Wingelhofer B , Kreis K , Mairinger S , Muchitsch V , Stanek J , Wanek T , et al. Preloading with l -BPA, l -tyrosine and l -DOPA enhances the uptake of [18F]FBPA in human and mouse tumour cell lines . Appl Radiat Isot . 2016 ; 118 : 67 - 72 . https://doi. org/10.1016/j.apradiso. 2016 . 08 .026.
21. Watabe T , Hanaoka K , Naka S , Kanai Y , Ikeda H , Aoki M , et al. Practical calculation method to estimate the absolute boron concentration in tissues using 18F-FBPA PET . Ann Nucl Med . 2017 ; 31 : 481 - 5 . https://doi.org/10.1007/s12149-017-1172-5.
22. Grunewald C , Sauberer M , Filip T , Wanek T , Stanek J , Mairinger S , et al. On the applicability of [18F]FBPA to predict l -BPA concentration after amino acid preloading in HuH-7 liver tumor model and the implication for liver boron neutron capture therapy . Nucl Med Biol . 2017 ; 44 : 83 - 9 . https://doi. org/10.1016/j.nucmedbio. 2016 . 08 .012.
23. Watabe T , Ikeda H , Nagamori S , Wiriyasermkul P , Tanaka Y , Naka S , et al. 18F-FBPA as a tumor-specific probe of L-type amino acid transporter 1 (LAT1): a comparison study with 18F-FDG and 11C-Methionine PET . Eur J Nucl Med Mol Imaging . 2017 ; 44 : 321 - 31 . https ://doi.org/10.1007/s0025 9- 016 -3487-1.
24. Wang HE , Liao AH , Deng WP , Chang PF , Chen JC , Chen FD , et al. Evaluation of 4-borono-2-18F-fluoro-l -phenylalanine-fructose as a probe for boron neutron capture therapy in a glioma bearing rat model . J Nucl Med . 2004 ; 45 : 302 - 8 .
25. Kabalka GW , Smith GT , Dyke JP , Reid WS , Longford CPD , Roberts TG , et al. Evaluation of fluorine-18-BPA-fructose for boron neutron capture treatment planning . J Nucl Med . 1997 ; 38 : 1762 - 7 .
26. Mairinger S , Stanek J , Wanek T , Langer O , Kuntner C . Automated electrophilic radiosynthesis of [18F]FBPA using a modified nucleophilic GE TRACERlab FXFDG . Appl Radiat Isot. 2015 ; 104 : 124 - 7 . https://doi.org/10.1016/j.apradiso. 2015 . 06 .034.
27. Vähätalo JK , Eskola O , Bergman J , Forsback S , Lehikoinen P , Jääskeläinen J , et al. Synthesis of 4-dihydroxyboryl-2-[18F]fluorophenylalanine with relatively high-specific activity . J Label Compd Radiopharm . 2002 ; 45 : 697 - 704 . https://doi.org/10.1002/ jlcr.600.
28. Havu-Aurén K , Kiiski J , Lehtiö K , Eskola O , Kulvik M , Vuorinen V , et al. Uptake of 4-borono-2-[18F] fluoro-l -phenylalanine in sporadic and neurofibromatosis 2-related schwannoma and meningioma studied with PET . Eur J Nucl Med Mol Imagimg . 2007 ; 34 : 87 - 94 . https://doi.org/10.1007/s00259-006-0154-y.
29. Honda N , Yoshimoto M , Mizukawa Y , Osaki K , Kanai Y , Kurihara H , et al. Radiosynthesis of 2-[18F] fluoro -4 -borono-phenylaranine ([18F]FBPA) using copper mediated oxidative aromatic nucleophilic [18F]fluorination . J Label Compd Radiopharm . 2017 : 60 ( Suppl . 1): S512 .
30. Ishiwata K , Ishii S , Senda M. HPLC using physiological saline for the quality control of radiopharmaceuticals used in PET studies . Appl Radiat isot. 1993 ; 44 : 1119 - 24 . https://doi.org/10.1016/ 0969 - 8043 ( 93 ) 90116 - R .
31. Coenen HH , Gee AD , Adam M , Antoni G , Cutler CS , Fujibayashi Y , et al. Consensus nomenclature rules for radiopharmaceutical chemistry . Nucl Med Biol . 2017 ; 55 :v-xi. https://doi.org/10.1016/j. nucmedbio. 2017 . 09 .004.
32. Coenen HH , Gee AD , Adam M , Antoni G , Cutler CS , Fujibayashi Y , et al. Open letter to journal editors on: international consensus radiochemistry nomenclature guidelines . Ann Nucl Med . 2018 ; 32 : 236 - 8 . https://doi.org/10.1007/s12149-018-1238-z.
33. Sakata M , Oda K , Toyohara J , Ishii K , Nariai T , Ishiwata K. Direct comparison of radiation dosimetry of six PET tracers using human whole-body imaging and murine biodistribution studies . Ann Nucl Med . 2013 ; 27 ( 3 ): 285 - 96 . https://doi.org/10.1007/s1214 9- 013 -0685-9.
34. Ishiwata K , Hayashi K , Sakai M , Kawauchi S , Hasegawa H , Toyohara J . Determination of radionuclides and radiochemical impurities produced by in-house cyclotron irradiation and subsequent radiosynthesis of PET tracers . Ann Nucl Med . 2017 ; 31 : 84 - 92 . https://doi.org/10.1007/s12149-016-1134-3.
35. Casella V , Ido T , Wolf AP , Fowler JS , MacGegor RR , Ruth TJ . Anhydrous F-18 labeled elemental fluorine for radiopharmaceutical preparation . J Nuci Med . 1980 ; 21 : 750 - 7 .
36. Hatano K , Ishiwata K , Yanagisawa T . Co production of 2 , 6 -[18F] difluoroDOPA during electrophilic synthesis of 6-[18F]fluoro-l - DOPA. Nucl Med Biol . 1996 ; 23 : 101 - 3 .
37. Coenen HH , Franken K , Kling P , Stöcklin G . Direct electrophic radiofluorination of phenyl, tyrosine and dopa . Appl Radiat Isot . 1988 ; 39 : 1243 - 50 .
38. Ariyoshi Y , Shimahara M , Kimura Y , Ito Y , Shimahara T , Miyatake S , et al. Fluorine-18 -labeled boronophenylalanine positron emission tomography for oral cancers: Qualitative and quantitative analyses of malignant tumors and normal structures in oral and maxillofacial regions . Oncol Lett . 2011 ; 2 : 423 - 7 . https://doi. org/10.3892/ol. 2011 . 265 .
39. Tani H , Kurihara H , Hiroi K , Honda N , Yoshimoto M , Kono Y , et al. Correlation of 18F-BPA and 18F-FDG uptake in head and neck cancers . Radiother Oncol . 2014 ; 113 : 193 - 7 . https://doi. org/10.1016/j.radonc. 2014 . 11 .001.
40. Sato M , Tatsuno T , Matsuo H . Studies on the racemization of amino acids and their derivatives . IV. Structural relation of amino acids to their racemizability in acetic acid . Yakugaku Zasshi . 1970 ; 9 : 1160 - 3 . (Japanese).
41. McLafferty FW , Stauffer DB . The Wiley/NBS registry of mass spectral data . Vol. 1 . New York: Wiley-Interscience; 1989 .