ARTICLE IN PRESS
`
`Applied Radiation and Isotopes 66 (2008) 497–505
`
`www.elsevier.com/locate/apradiso
`
`Synthesis, stabilization and formulation of [177Lu]Lu-AMBA,
`a systemic radiotherapeutic agent for Gastrin Releasing
`Peptide receptor positive tumors
`
`
`, Aldo Cagnolini, Edmund Metcalfe, Natarajan Raju,
`Jianqing Chen, Karen E. Linder
`Michael F. Tweedle, Rolf E. Swenson
`
`Bracco Research USA Inc., 305 College Road East, Princeton, NJ 08540, USA
`
`Received 30 July 2007; received in revised form 30 October 2007; accepted 13 November 2007
`
`Abstract
`
`A robust formulation was developed for [177Lu]Lu-AMBA (177Lu-DO3A-CH2CO-G-[4-aminobenzoyl]-QWAVGHLM-NH2), a
`Bombesin-like agonist with high affinity for Gastrin Releasing Peptide (GRP) receptors. During optimization of labeling, the effect of
`several radiostabilizers was evaluated; a combination of selenomethionine and ascorbic acid showed superiority over other tested
`radiostabilizers. The resulting two-vial formulation maintains a radiochemical purity (RCP) of 490% for at least 2 days at room
`temperature. The method of stabilization should be useful for other methionine-containing peptide radiopharmaceuticals in diagnostic
`and therapeutic applications.
`r 2007 Elsevier Ltd. All rights reserved.
`
`Keywords: 177Lu-AMBA; Bombesin; Gastrin releasing peptide; Targeted radiotherapy; Radiolysis protection
`
`1. Introduction
`
`Interest in using radiolabeled bombesin derivatives as
`agents for diagnostic imaging and/or systemic radiotherapy
`of tumors (Smith et al., 2003; Zhang et al., 2004; Lantry et
`al., 2006 and references therein) has increased considerably
`because of the observation that Gastrin Releasing Peptide
`receptors (GRPr) are over-expressed in a variety of human
`tumor cells. Lantry et al. (2006) recently demonstrated that
`the [177Lu]Lu-labeled Gastrin Releasing Peptide (GRP)
`derivative known as [177Lu]Lu-AMBA (AMBA ¼ (DO3A-
`CH2CO-G-(4-aminobenzoyl)-QWAVGHLM-NH2)) binds
`with nanomolar affinity to GRP receptors; preclinical
`studies with this Lu-labeled compound demonstrated
`therapeutic efficacy in a GRPr positive PC-3 human
`[177Lu]Lu-
`prostate tumor-bearing nude mouse model.
`AMBA is now in clinical trials for the radiotherapeutic
`treatment of prostate cancer.
`
`
`
`Corresponding author. Tel.: +1 609 514 2416; fax: +1 609 514 2446.
`E-mail address: karen.linder@bru.bracco.com (K.E. Linder).
`
`0969-8043/$ - see front matter r 2007 Elsevier Ltd. All rights reserved.
`doi:10.1016/j.apradiso.2007.11.007
`
`Radiopharmaceuticals for systemic therapeutic applica-
`tions are designed to deliver a therapeutic dose of radiation
`to specific disease sites. The ionizing radiation (e.g., a- or
`b-particles) given off from such compounds can either
`damage cellular components in the target tissue directly, or
`d
`d
`d
`, H
`)
`, O2
`indirectly via the free radicals (e.g., OH
`formed by the interaction of ionizing radiation with water
`in the target tissue (Burton and Lipsky, 1957; Liu and
`Edwards, 2001; Liu et al., 2003; Pozzi and Zalutsky, 2005).
`However,
`the potentially destructive properties of a
`therapeutic radioisotope’s emissions are not limited to
`their cellular targets. Radiation-induced damage to the
`radiolabeled compound itself is one of the most challenging
`aspects in the development of a therapeutic radiopharma-
`ceutical. For peptides and proteins, Garrison (1987) has
`reported that radiation induced damage may include
`oxidation, hydroxylation, aggregation and/or bond scis-
`sion.
`Preliminary tests showed that [177Lu]Lu-AMBA was
`very radiosensitive;
`in the absence of radiostabilizers,
`degradation occurred both during and after radiolabeling.
`
`Evergeen Ex. 1029
`1 of 9
`
`

`

`498
`
`J. Chen et al. / Applied Radiation and Isotopes 66 (2008) 497–505
`
`ARTICLE IN PRESS
`
`In particular, the methionine residue of the targeting
`peptide was found to be readily oxidized to its methionine
`sulfoxide [Met(O)]
`form, one of
`the major radiolytic
`degradants of [177Lu]Lu-AMBA. Biological studies demon-
`strated that the targeting capability of [177Lu]Lu-AMBA
`was totally inactivated by this oxidization. Hence, use of a
`radiostabilizer or stabilizer combination in the [177Lu]Lu-
`AMBA formulation was an absolute requirement.
`The purpose of this study was to establish a robust
`formulation for [177Lu]Lu-AMBA for use in Phase I
`clinical trials. The effect of pH, ligand concentration, and
`reaction time were determined, and several stabilizers were
`evaluated to identify a formulation yielding and maintain-
`ing high radiochemical purity (RCP).
`
`2. List of abbreviations
`
`Potential radiostabilizers evaluated included (a) amino
`acids: glycine (Gly), methionine (Met), cysteine (Cys),
`cysteine ethyl ester (CEE), tryptophan (Trp), and histidine
`(His); (b) naturally occurring selenium compounds: sele-
`nomethionine (Se-Met) and selenocysteine (Se-Cys); (c)
`sulfur-containing
`reducing
`agents:
`2-mercaptoethanol
`(ME), dithiothreitol (DTT), and 1-pyrrolidinecarbodithioic
`acid (PDTC). The results were compared to those obtained
`with the commonly used radical scavengers such as
`ascorbic acid (AA), gentisic acid (GA), human serum
`albumin (HSA), and ethanol (EtOH).
`
`3. Materials and methods
`
`Glacial acetic acid (Ultrapure) and sodium acetate
`trihydrate (USP) were purchased from J.T. Baker. L-(+)-
`Selenomethionine (Se-Met) was obtained from Sabinsa
`Corp. Amino acids, ammonium sulfate, trifluoroacetic acid
`(TFA), acetonitrile and methanol were bought from EMD
`Chemicals, Inc. Bacteriostatic 0.9% Sodium Chloride
`Injection (USP) was purchased from Abbott Laboratories.
`s Ascorbic Acid Injection (USP) [containing
`ASCOR L500
`500 mg/mL Ascorbic acid and 0.025% (w/v) Edetate
`disodium] was obtained from McGuff Pharmaceuticals,
`[177Lu]LuCl3 in 0.05 N HCl was purchased from
`Inc.
`Missouri University Research Reactor (MURR). ITLC SG
`strips were from Pall Life Sciences. Deionized water was
`used for all solutions containing water, including HPLC
`mobile phases.
`
`3.1. Peptide synthesis
`
`AMBA
`
`[(DO3A-CH2CO-G-(4-aminobenzoyl)-
`QWAVGHLM-NH2), DO3A ¼ (1,4,7,10-tetraaza-4,7,10-
`
`tris(carboxymethyl)-cyclododecyl)-acetyl] was synthesized
`using solid phase peptide synthesis chemistry, as described
`by Lantry et al. (2006). The compound structure was
`confirmed by LC/MS, amino acid sequence and elemental
`analysis. The proposed chemical structure of its lutetium
`complex is shown in Fig. 1. An authentic sample of
`AMBA-Met(O), a mixture of
`the two unresolvable
`methionine oxide epimers of
`the AMBA ligand was
`prepared using the appropriate Met(O) containing pro-
`tected amino acids. This methionine oxidized compound
`mixture and its Lu complex [Lu-AMBA-Met(O)] were
`characterized by MS and HPLC.
`
`3.2. Radiochemistry
`
`3.2.1. Standard procedure for preparation of [177Lu]Lu-
`AMBA
`To a lead-shielded 7-mL vial containing 120 mg of
`AMBA and 1 mg Se-Met in 1 mL of 0.2 M (pH 4.8)
`sodium acetate
`(NaOAc)
`buffer,
`was
`added
`4.0770.37 GBq [177Lu]LuCl3 in 0.05 N HCl (radioconcen-
`tration 37 GBq/mL, specific activity 103.6–151.3 GBq/
`mmol). The mixture was heated at 100 1C in a heating block
`for 10 min. After cooling to ambient temperature in a
`water-bath for 2 min, the reaction solution was diluted by
`adding 4 mL of ascorbate dilution solution [a 9:1 mixture
`of Bacteriostatic 0.9% Sodium Chloride Injection USP and
`ASCOR L500s Ascorbic Acid Injection USP (final
`ascorbic acid concentration, 40 mg/mL)] yielding a final
`radioconcentration of 814 MBq/mL (22 mCi/mL). Any
`possible non-incorporated 177Lu remaining in the reaction
`solution was converted to [177Lu]Lu-EDTA by the EDTA
`contained in the Ascorbic Acid Injection. The radio-
`complex was then characterized by HPLC. In some cases,
`this reaction was performed at a 1/10th or 1/5th scale,
`maintaining the same concentrations as described above,
`but using a 2-mL reaction vial.
`
`3.2.2. Effect of buffer pH and ligand concentration
`For the effect of buffer pH on [177Lu]Lu-AMBA
`incorporation, studies were performed at 1/10th of the full
`scale formulation as described above, but using 0.2 M
`
`NH
`
`NH
`
`N
`
`S
`
`NH2
`
`O
`
`N H
`
`O
`
`NH
`
`O
`
`N H
`
`O
`
`NH
`
`O
`
`N H
`
`O
`
`NH
`
`O
`
`N H
`
`O
`
`NH
`
`O
`
`H2N
`
`O
`
`COO
`
`COO
`
`N
`
`N H
`
`O
`
`NH
`
`N
`
`Lu
`
`N
`
`N
`
`O
`
`COO
`
`Fig. 1. The proposed chemical structure of Lu-AMBA.
`
`Evergeen Ex. 1029
`2 of 9
`
`

`

`ARTICLE IN PRESS
`
`J. Chen et al. / Applied Radiation and Isotopes 66 (2008) 497–505
`
`499
`
`NaOAc buffers at pH values from 2.8 to 6.8. To study the
`[177Lu]Lu-AMBA was
`effect of
`ligand concentration,
`prepared using the standard procedure at 1/10th of the
`full scale formulation, using 7.5, 10, 12.5, or 15 mg of
`AMBA instead of 12 mg, thus providing AMBA:Lu ratios
`from 1.5:1 through 3:1. RCP and radiocolloid were assayed
`immediately after labeling.
`
`temperature and heating time on Met
`
`3.2.3. Effect of
`oxidation
`To determine if heating at 100 1C for 10 min causes Met
`oxidation in the absence of radiation, labeling reactions
`were performed as described for the standard procedure,
`but using 30 nmol of 175Lu (Lu2O3 dissolved in 5% HNO3)
`in the place of 177Lu. Two such reactions were performed,
`one containing 1 mg/mL Se-Met, the other containing no
`stabilizer. These reactions were examined by UV (A280)
`after 10 min at 100 1C to determine the amount of Lu-
`AMBA Met(O) formed. Studies on the effect of heating
`time were performed with [177Lu]Lu-AMBA prepared as
`described in the standard procedure, but reaction mixtures
`were heated for 8, 9, 10, 12 or 15 min.
`
`3.3. Evaluation and comparison of potential stabilizers
`
`3.3.1. Post-labeling stabilization of [177Lu]Lu-AMBA
`In these studies, the stabilizer was added immediately
`after radiolabeling; no stabilizer was present during 177Lu
`incorporation. Five amino acids were evaluated: Met, Cys,
`Trp, His, and Gly. For comparison, studies with the
`commonly used radiolysis protecting agents AA, GA,
`HSA, and EtOH were also performed. A 10 mg/mL
`solution of each potential stabilizer was prepared in
`0.05 M (pH 5.3) citrate buffer or 10% (v/v) for the EtOH.
`[177Lu]Lu-AMBA was prepared by treating 36 mg AMBA
`in 300 mL of 0.2 M (pH 4.8) NaOAc buffer with 0.5 GBq
`[177Lu]LuCl3. Immediately after labeling, 50 mL (83 MBq)
`of the [177Lu]Lu-AMBA solution was mixed with the
`stabilizer solution (100 mL) to yield a final radioactivity
`concentration of 0.55 GBq/mL and a final
`stabilizer
`concentration of 6.6 mg/mL or 6.6% (v/v) for the EtOH.
`[177Lu]Lu-AMBA solution mixed with
`An aliquot of
`0.05 M citrate buffer, pH 5.3 (100 mL) was used as a
`control. The samples were stored at room temperature
`(RT), and analyzed by HPLC at 0, 24 and 48 h to
`determine the RCP and percentage of the Met(O) form
`of [177Lu]Lu-AMBA.
`
`3.3.2. Evaluation of reducing agents as stabilizers
`The use of reducing agents as radioprotectants (specifi-
`cally, to prevent methionine sulfoxide formation) and their
`efficacy in reducing methionine sulfoxide residues to
`methionyl residues was evaluated. For radiostabilization
`studies, 1 mg (5 mg/mL) of Cys, CEE, or PDTC was added
`into a 1/5th scale formulation in the place of Se-Met. After
`labeling, 0.8 mL of Bacteriostatic 0.9% Sodium Chloride
`Injection (containing 1 mg/mL EDTA), without ASCORs
`
`Ascorbic Acid Injection, was added to dilute the reaction
`solution. RCP was determined at 0 and 24 h after the
`labeling.
`Cys, ME, and DTT were tested to determine their ability
`to reduce the methionine sulfoxide [Met(O)] residue in
`[177Lu]Lu-AMBA-Met(O), the primary radiodegradant of
`[177Lu]Lu-AMBA-Met(O) was pre-
`[177Lu]Lu-AMBA.
`pared by heating 12 mg AMBA-Met(O) ligand in 100 mL
`of 0.2 M (pH 4.8) NaOAc buffer with 55.5 MBq
`[177Lu]LuCl3 in the absence of any radiostabilizer. An
`aliquot (10 mL) of the [177Lu]Lu-AMBA-Met(O) solution
`was mixed with the reducing agent (90 mL) at a final
`concentration of 10 mg/mL. As controls, Met and Se-Met
`were used in the place of the reducing agents. The samples
`were analyzed by HPLC after storage at RT for 1 and 3
`days to determine the percentage of [177Lu]Lu-AMBA that
`formed via reduction of [177Lu]Lu-AMBA-Met(O).
`
`3.3.3. Evaluation of selenium compounds
`Two naturally occurring organoselenium compounds,
`Se-Met and Se-Cys were evaluated. The [177Lu]Lu-AMBA
`solution was prepared in a full scale formulation as
`described in the standard procedure above, using 1 mg of
`either Se-Met or Se-Cys as the stabilizer present during
`[177Lu]Lu-AMBA formulated using the same
`labeling.
`conditions without any stabilizer was tested as a control.
`Immediately after the labeling, the reaction solutions were
`diluted by addition of 4 mL of ascorbate dilution solution,
`and characterized by HPLC.
`
`3.3.4. Labeling of frozen formulations
`In this study, five stabilizers found to have efficacy in
`preliminary studies (Met, Se-Met, Cys, CEE, and PDTC)
`were further evaluated to determine their capacity for
`radiolysis protection after storage at 20 1C in a frozen
`formulation solution for 1 month. Stabilizer (1 mg/mL)
`and AMBA (120 mg/mL) were dissolved in 0.2 M (pH 4.8)
`NaOAc buffer under N2, and 1-mL aliquots of the solution
`dispensed into 7-mL vials. Vials were bubbled with N2 gas,
`crimp-sealed, and stored at 20 1C. After 1 month of
`storage, the vials were warmed to ambient temperature,
`and an aliquot (100 mL) of each solution was used to
`prepare [177Lu]Lu-AMBA (1/10th of the full scale for-
`mulation). After labeling, 0.4 mL of Bacteriostatic saline
`containing 1 mg/mL EDTA, without ASCORs Ascorbic
`Acid Injection, was added to dilute the reaction solution to
`a radioconcentration of 0.814 GBq/mL. RCP was deter-
`mined at 0 and 24 h after labeling.
`
`3.4. Effect of heating time and radioconcentration
`
`For the effect of heating time, [177Lu]Lu-AMBA was
`prepared as described in the standard procedure above, but
`the heating time was varied from 8 to 15 min. To determine
`the effect of radioconcentration, 177Lu-complexation was
`[177Lu]LuCl3
`performed using 5.55 GBq (150 mCi) of
`
`Evergeen Ex. 1029
`3 of 9
`
`

`

`ARTICLE IN PRESS
`
`500
`
`J. Chen et al. / Applied Radiation and Isotopes 66 (2008) 497–505
`
`instead of 4.07 GBq (110 mCi). RCP was determined after 0
`and 48 h at RT.
`
`4. Results
`
`4.1. HPLC analysis
`
`3.5. Analytical methods
`
`3.5.1. Instrumentation
`An Agilent Technologies quaternary 1100 Series HPLC
`equipped with an autosampler and variable wavelength
`detector was used. The radioactivity was monitored using a
`Canberra NaI detector (Model 802-2  2W) with high
`voltage power supply (Model 3102D), single channel
`analyzer (Model 2015A) and either a linear ratemeter
`(Model 2081) or a linear/logarithmic ratemeter (Model
`1481LA). The Model 1481LA ratemeter provided excellent
`from 3.7  10
`3 to 4.44 MBq of
`linearity (R240.99)
`injected radioactivity.
`
`3.5.2. HPLC analysis
`[177Lu]Lu-AMBA HPLC analysis was performed as
`follows: column: Zorbax Bonus-RP (5 mm, 80 A˚ pore size,
`250  4.6 mm, Agilent); column temperature: 37 1C; flow
`rate: 1.5 mL/min; mobile phases: (A ¼ H2O; B ¼ 30 mM
`ammonium sulfate with 0.1% TFA (v/v); C ¼ methanol;
`D ¼ acetonitrile). The HPLC gradient started at 30%A/
`60%B/5%C/5%D, ramped to 14%A/60%B/13%C/13%D
`over 5 min, and was held at this composition for 32 min.
`The
`retention times
`for unlabeled peptide
`ligand,
`[177Lu]Lu-AMBA and its methionine oxide analog
`[177Lu]Lu-AMBA-Met(O) were 24, 30, and 12 min,
`respectively. RCP was calculated as the percentage of the
`area present as [177Lu]Lu-AMBA relative to the total
`integrated area (all combined radioactive peaks plus any
`segments of elevated baseline). Percentage 177Lu incorpora-
`tion was defined as the percentage of activity in all
`combined radioactive peaks except that in the void volume
`of the column, relative to the total integrated area, where
`the radioactive peak in the void volume of the column
`(‘‘void peak’’) represented [177Lu]Lu-EDTA.
`
`3.5.3. Radiocolloid determination and recovery studies
`Radiocolloid was monitored using 2  10 cm silica gel
`(SG) thin layer chromatography (TLC) strips developed
`with acetone/0.9% saline (1:1). Radiocolloid remained at
`[177Lu]Lu-EDTA and [177Lu]Lu-AMBA mi-
`the origin;
`grated with the solvent front. The percentage radiocolloid
`was calculated as the percentage of the total applied
`radioactivity that remained at the origin. Recovery studies
`were performed by removing all product from the vial
`using a glass pipette and counting the radioactivity that
`could be removed relative to that which remained in the
`‘‘empty’’ vial and crimp-seal.
`
`3.5.4. Statistical analysis
`Statistical analysis was performed using the Student’s t-
`test for paired data with one-tailed distribution.
`
`4.1.1. Radiochromatogram of [177Lu]Lu-AMBA
`[177Lu]Lu-
`Preliminary labeling studies showed that
`AMBA prepared in acetate buffer at a radioconcentration
`of 814 MBq/mL (22 mCi/mL)
`in the absence of any
`radiostabilizers underwent significant radiolysis. Figs. 2a
`and b show typical radiotraces of unstabilized [177Lu]Lu-
`AMBA at 0 and 12 h post-formulation, respectively, when
`[177Lu]Lu-AMBA-Met(O),
`formed by
`stored at RT.
`oxidation of the methionyl residue of [177Lu]Lu-AMBA,
`was a significant degradant immediately after labeling
`(Fig. 2a), and RCP fell to o10% within 12 h. It was clear
`from these results that radiostabilization was needed.
`After significant study, an optimized ‘‘standard proce-
`dure’’ for radiolabeling was identified, wherein radiolabel-
`ing was performed in the presence of Se-Met to prevent
`radiolysis during 177Lu incorporation, followed by addition
`of ascorbic acid and bacteriostatic saline to provide good
`long-term post-labeling stability. A typical HPLC radio-
`chromatogram of [177Lu]Lu-AMBA prepared using the
`standard procedure is shown in Fig. 2c.
`
`4.2. [177Lu]Lu-AMBA formulation and radiolysis
`protection
`
`4.2.1. Effect of buffer pH and ligand concentration
`The effect of buffer pH is shown in Table 1. The
`percentage of 177Lu incorporation did not change sig-
`nificantly between pH 3.8 and 5.8, but at pH values of 2.8
`and 6.8, a decrease in the percentage of 177Lu incorpora-
`tion was seen. Furthermore, the amount of radiocolloid
`was significantly increased when the reaction pH reached
`6.8. Based on these results, 0.2 M NaOAc buffer at pH 4.8
`was selected for all subsequent [177Lu]Lu-AMBA labeling,
`unless otherwise indicated.
`Studies on the effect of ligand concentration (ligand-to-
`177Lu incorporation was
`metal
`ratios)
`showed that
`incomplete with an AMBA to Lu molar ratio of 1.5,
`showing a 5% void peak in the radiochromatogram.
`However, it was close to 100% at all higher molar ratios
`tested.
`
`4.2.2. Evaluation and comparison of potential stabilizers
`Three groups of potential stabilizers were tested by
`adding them before and/or after [177Lu]Lu-AMBA com-
`plexation.
`
`4.2.2.1. Amino acids. The RCP results obtained when
`amino acids and commonly used radical scavengers (AA,
`GA, HSA, and EtOH) were added (6.6 mg/mL)
`to
`[177Lu]Lu-AMBA (0.55 GBq/mL) after Lu incorporation
`are shown in Table 2. The percentage of the Met(O) form
`of [177Lu]Lu-AMBA is also listed. In these studies, no
`
`Evergeen Ex. 1029
`4 of 9
`
`

`

`ARTICLE IN PRESS
`
`J. Chen et al. / Applied Radiation and Isotopes 66 (2008) 497–505
`
`501
`
`Fig. 2. HPLC radiochromatograms of unstabilized [177Lu]Lu-AMBA at 0 (a) and 12 h (b) post-formulation; and (c) stabilized [177Lu]Lu-AMBA,
`expanded scale, at 12 h post-formulation. Four microliters of the [177Lu]Lu-AMBA formulation solution (3.26 MBq) were injected.
`
`Table 1
`Effect of buffer pH on [177Lu]Lu-AMBA incorporation and radiocolloid
`
`pH
`
`2.8
`3.8
`4.8
`5.8
`6.8
`
`Incorporation (%)
`
`Radiocolloid (%)
`
`97.4
`99.6
`99.7
`99.6
`96.7
`
`0.30
`0.32
`0.31
`0.34
`1.20
`
`Ascor solution was added, so the relative effect of each
`stabilizer could be tested.
`Of the amino acids tested in this study, Met was one of
`the most effective radioprotectants for prevention of the
`oxidation of the methionine residue in [177Lu]Lu-AMBA.
`Significant radioprotection was observed for both Cys and
`Trp, compared to that of His and Gly. Surprisingly, Trp
`
`prevented all damage except the oxidation of the Met
`residue; almost no other degradants except [177Lu]Lu-
`AMBA-Met(O) were observed during 2 days of storage at
`RT.
`AA and GA had significant stabilizing effects, but at the
`concentrations tested, none of the commonly used radio-
`protection agents evaluated (AA, GA, or HSA) supplied
`enough protection to inhibit all radiolysis of [177Lu]Lu-
`AMBA. It was found that EtOH could stabilize the
`radiocomplex, and sometimes benefited RCP and recovery
`when a trace amount of ligand was used.
`
`4.2.2.2. Organic selenium and sulfur containing compound-
`s. Table 3 lists the initial (t ¼ 0) RCP values obtained for
`[177Lu]Lu-AMBA radiolabeled in the presence or absence
`of Se-Met, Se-Cys, Cys, CEE and PDTC. At
`the
`concentrations tested, all compounds tested in this study
`
`Evergeen Ex. 1029
`5 of 9
`
`

`

`502
`
`J. Chen et al. / Applied Radiation and Isotopes 66 (2008) 497–505
`
`ARTICLE IN PRESS
`
`Table 2
`Effect of amino acids and other stabilizers on [177Lu]Lu-AMBA radiostabilitya
`
`Stabilizer
`
`24 h
`
`RCP (%)
`
`Met(O) form (%)
`
`48 h
`
`RCP (%)
`
`Met(O) form (%)
`
`7.572.5
`9.172.3
`Control
`Histidineb
`44.0
`4.6
`Glycineb
`24.2
`14.6
`74.473.6
`17.472.7
`Tryptophan
`2.470.6
`55.272.5
`Methionine
`14.570.8
`71.671.7
`Cysteine
`17.970.9
`72.174.6
`Gentisic acid
`5.570.3
`83.671.1
`Ascorbic acid
`17.870.3
`18.875.2
`HSA
`14.270.2
`52.572.8
`Ethanol
`Radiolabeled product [0.55 GBq/mL] stored at RT for 24 and 48 h (mean7S.D.) (n ¼ 3).
`aThe initial (t ¼ 0 h) RCP and percentage of Met(O) form were 92.970.5% and 1.570.2%, respectively.
`bn ¼ 1.
`
`0
`0
`0
`0.470.4
`33.573.2
`54.370.9
`40.276.5
`75.071.3
`2.571.7
`21.671.7
`
`0
`29.5
`13.8
`89.470.7
`2.270.4
`24.370.8
`45.671.3
`9.770.8
`11.672.2
`23.570.6
`
`0 h
`
`24 h
`
`100
`
`80
`
`60
`
`40
`
`20
`
`0
`
`RCP of 177Lu-AMBA
`
`Table 3
`Effect of selenium and sulfur compounds on initial RCP of [177Lu]Lu-
`AMBA formulation (mean7S.D.) (n ¼ 3)
`
`Stabilizer
`
`Se-Met
`
`Se-Cys
`Cys
`CEE
`PDTC
`Control
`an ¼ 1.
`
`Amount (mg)
`
`0.5
`1
`5
`10
`1
`2.5
`2.5
`2.5
`No stabilizer
`
`RCP (%)
`
`99.3a
`98.471.6
`97.1a
`95.0a
`98.471.2
`99.1a
`98.4a
`98.2a
`86.572.8
`
`could completely protect [177Lu]Lu-AMBA from radiolysis
`during complexation. Furthermore, it was found that by
`increasing the applied amount from 1 to 10 mg, Se-Met
`could fully protect [177Lu]Lu-AMBA from radiolysis for
`over 48 h at RT (data not shown). However, 177Lu
`incorporation into AMBA was adversely affected by large
`amounts of Se-Met; a significantly lower RCP with a larger
`radioactive ‘‘void volume’’ peak was observed when 10 mg
`of Se-Met was used.
`
`4.2.2.3. Reducing agents. Fig. 3 shows the RCP values
`obtained for [177Lu]Lu-AMBA that was radiolabeled in the
`presence of 5 mg/mL reducing agents and then stored at
`RT for 0 and 24 h. Of the tested reducing agents, the
`oxygen radical scavenger PDTC was the most effective as a
`radiostabilizer in freshly prepared formulations.
`An attempt was made to determine whether thiol-
`containing reductants might have utility in reducing the
`oxidized product [177Lu]Lu-AMBA-Met(O) back to the
`desired [177Lu]Lu-AMBA. Over 24 h at RT, Cys, ME, and
`[177Lu]Lu-
`DTT reduced 5.6%, 9.7% and 16.2% of
`AMBA-Met(O) to [177Lu]Lu-AMBA, suggesting a rela-
`tively modest effect at radioprotection. In contrast, no
`
`PDTC
`
`Cys
`
`CEE
`
`Fig. 3. Effect of reducing agents (5 mg/mL) on RCP of [177Lu]Lu-AMBA
`(%) after storage at RT for 0 and 24 h (mean7S.D.) (n ¼ 3).
`
`reduction of the Met(O) moiety was observed when using
`Met or Se-Met.
`
`-
`4.2.2.4. Evaluation of stabilizers in frozen formulations.
`Met, Se-Met, Cys, CEE, and PDTC were selected for
`further investigation to determine whether these stabilizers
`impacted [177Lu]Lu-AMBA incorporation and/or retained
`their stabilizing property in a stabilizer/AMBA ligand-
`containing solution that was stored at 20 1C for 1 month
`prior to radiolabeling. The results are listed in Table 4.
`High RCP values were obtained for all stabilized formula-
`tions at t ¼ 0, but RCP dropped significantly in all ‘‘single-
`stabilizer’’
`formulations after 24 h at RT. For frozen
`formulations containing Cys, Met and PDTC, the RCP
`after 24 h was significantly lower than had been observed
`with freshly prepared solutions. It is believed that this
`difference is primarily due to the higher amount of
`radioactivity used in the previously frozen kits. The studies
`on freshly prepared solutions (Table 2) were performed at a
`final radioconcentration of 15 mCi/mL (0.55 GBq/mL),
`
`Evergeen Ex. 1029
`6 of 9
`
`

`

`ARTICLE IN PRESS
`
`J. Chen et al. / Applied Radiation and Isotopes 66 (2008) 497–505
`
`503
`
`Table 4
`Behavior of frozen formulations for [177Lu]Lu-AMBA (0.814 GBq/mL)
`when stored at RT at 0 and 24 h post-labeling (% RCP) (mean7S.D.)
`(n ¼ 3)
`
`Table 5
`Effect of heating time at 100 1C on [177Lu]Lu-AMBA RCP, % of
`[177Lu]Lu-AMBA-Met(O), and % ‘‘void peak’’ ([177Lu]Lu-EDTA)
`
`Stabilizer
`
`RCP (%)
`
`0 h
`
`24 h
`
`Control (no stabilizer)
`PDTC
`Met
`Cys
`CEE
`Se-Met
`Se-Met/AAa
`po0.001, compared to Se-Met.
`
`aCombination of Se-Met present during 177Lu incorporation, followed
`by ascorbic acid addition after labeling.
`
`
`6.577.9
`
`1.371.4
`
`19.572.7
`
`26.673.9
`
`48.074.7
`72.072.6
`
`99.770.2
`
`
`87.976.1
`98.670.1
`95.972.2
`97.772.5
`97.371.1
`99.570.3
`99.770.2
`
`whereas the studies on frozen solutions (Table 4) were
`performed at a more demanding radioconcentration of
`22 mCi/mL (0.814 GBq/mL). However, we cannot discount
`the potential for some Cys oxidation during frozen storage.
`The PDTC-containing formulations formed an insoluble
`precipitate during storage at 20 1C that
`failed to
`redissolve. As a consequence, no protecting effect was
`observed for frozen kits containing PDTC at the 24-h time
`point after radiolabeling. This was in contrast to the results
`shown in Fig. 3, where freshly prepared PDTC (5 mg/mL)
`was used. The nature of the precipitate was not studied, but
`dithiocarbamates are known to be unstable in acidic
`aqueous solution (Hilder et al., 1998). Of the five tested
`stabilizers, Se-Met provided significantly higher efficacy in
`protecting [177Lu]Lu-AMBA from radiolysis than did the
`other agents (po0.001). Overall, best results were obtained
`when radiolabeling was performed in the presence of Se-
`Met (1 mg/mL) to prevent radiolysis during Lu incorpora-
`tion,
`followed by the addition of ascorbate dilution
`solution to provide excellent
`long-term post-labeling
`stability (Table 4).
`
`4.2.3. Heating time and radioconcentration
`For the optimized Lu-AMBA formulation, the effect of
`heating time on RCP, and percentages of void peak and
`[177Lu]Lu-AMBA-Met(O) are shown in Table 5. Incor-
`poration was incomplete with a heating time of 8 min, but
`RCP values 490% were found for heating times of
`9–15 min. Heating longer than 10 min caused RCP to fall
`due to increased formation of radiolytic decomposition
`[177Lu]Lu-AMBA-Met(O). Reaction
`products
`such as
`mixtures that were prepared with 175Lu instead of 177Lu
`did not show significant Met oxidation (Table 5). When
`[177Lu]Lu-AMBA was prepared with 5.55 GBq of [177Lu]-
`LuCl3 instead of 4.07 GBq an initial RCP of 93.3% was
`obtained, vs. 94.4% with 4.07 GBq of [177Lu]LuCl3. After
`48 h at RT, these kits showed RCP values of 90.1% and
`91.4%, respectively, showing that the formulation toler-
`
`Lu source
`
`Heating
`time (min)
`
`Lu-
`AMBA
`RCP (%)
`
`177Lu (1 mg Se-Met)
`
`177Lu (no Se-Met)
`175Lu (no Se-Met)
`175Lu (1 mg Se-Met)
`
`n.r., not reported.
`
`8
`9
`10
`15
`10
`10
`10
`
`89.6
`93.7
`92.4
`91.2
`67.5
`
`Void
`peak
`(%)
`
`5.0
`1.0
`0.3
`0.4
`n.r.
`
`Lu-
`AMBA-
`Met(O)
`(%)
`
`0.6
`0.6
`0.7
`0.9
`15.5
`0.34
`0.07
`
`ated the addition of 36% more radioactivity without a
`significant drop in RCP.
`Based on the results described above, a robust formula-
`tion method has been developed for [177Lu]Lu-AMBA at a
`radiotherapeutic dose level
`(4.07 GBq). By using an
`optimized combination of radiostabilizers, 177Lu incor-
`poration of greater than 98% (RCP 490%) could be
`obtained and maintained for at least 2 days, when stored at
`RT. Radioactivity recovery in this
`formulation was
`excellent [98.670.8%] (mean7S.D.) (n ¼ 23).
`
`5. Discussion
`
`A large number of studies have been carried out by
`others to determine methods to prevent radiolytic damage
`in radiopharmaceutical formulations, and several radical
`scavengers have been used for this purpose (Chen et al.,
`2005; Storey et al., 2001; McGill and Henriksen, 2004;
`Miller and De La Fourniere, 1995; Derosch et al., 1993;
`Wolfangel, 1992; Cyr and Pearson, 2002). Although studies
`by others have demonstrated that compounds such as AA,
`GA, and Met are effective radiostabilizers
`in some
`applications, the present studies indicate that as single
`agents, they are insufficient in fully protecting [177Lu]Lu-
`AMBA from radiolytic damage (Tables 2 and 4). For
`example, Cyr and Pearson (2002) demonstrated that
`radiopharmaceuticals could be stabilized by addition of a
`hydrophilic thioether, and that the amino acid Met was
`especially useful for this purpose. However, our studies
`showed that at RT, 80% of [177Lu]Lu-AMBA (814 MBq/
`mL) was destroyed by radiolysis within 24 h when Met was
`used as a stabilizer. The radioprotection efficiency of Met
`was found to be significantly lower than that of Se-Met
`(po0.001) (Table 4). Furthermore, when a large amount of
`Met (5 mg) was added into a [177Lu]Lu-AMBA formula-
`tion at 814 MBq/mL (full scale), almost a complete
`decomposition of the product occurred within 48 h (data
`not shown), confirming that Met alone was insufficient for
`[177Lu]Lu-AMBA protection.
`
`Evergeen Ex. 1029
`7 of 9
`
`

`

`504
`
`J. Chen et al. / Applied Radiation and Isotopes 66 (2008) 497–505
`
`ARTICLE IN PRESS
`
`The degree of radiolytic damage observed is expected to
`be dependent upon both the radioconcentration and the
`type of ionizing radiation (e.g., a-, b- and g-emitter) used.
`In addition, the effectiveness of a particular radiostabilizer
`also depends on the sensitivity of various functional groups
`in the targeting molecule itself. For example, tryptophan
`could efficiently prevent all radiolytic damage in [177Lu]Lu-
`AMBA except that of oxidation of methionine residues
`(Table 2) and might prove suitable for the stabilization of
`compounds that do not contain methionine.
`Oxidation of the methionyl residue in [177Lu]Lu-AMBA
`was found to be due almost entirely to radiolysis. Thermal
`degradation was minimal during the 100 1C heating step
`used for Lu incorporation. This was determined by
`studying the UV spectrum of Lu-AMBA formulations
`prepared using [175Lu]Lu-AMBA in the place of [177Lu]Lu-
`AMBA. When non-radioactive 175Lu and AMBA were
`heated either in the presence or absence of 1 mg/mL Se-
`Met, the UV traces after heating at 100 1C for 10 min
`showed very little [175Lu]Lu-AMBA-Met(O) formation.
`These results suggested that radiation-induced oxidation of
`the Met residue via free radicals formed from radiolysis of
`water, rather than reaction with oxygen or chemical/
`thermal
`instability was the key factor in Lu-AMBA-
`Met(O) formation.
`Of all the tested compounds, Se-Met proved the most
`effective in protecting [177Lu]Lu-AMBA from radiolytic
`destruction (Table 4). A relatively small amount (1 mg) was
`needed for
`this purpose;
`toxicology studies on the
`formulation were performed prior to human Phase I
`clinical trials to demonstrate that it was well tolerated.
`Selenium is an essential nutrient for the normal function of
`the immune system, and the toxicity of selenium (especially
`naturally occurring organic forms such as Se-Met)
`is
`relatively low. Se-Met has been widely used for protection
`against radiation- and chemical-induced carcinogenesis in
`both preclinical and clinical studies (Weiss et al., 1994;
`Klein, 2004; El-Bayoumy and Sinha, 2004; Schrauzer,
`2000; Kennedy et al., 2004). Furthermore,
`it has been
`found that selenium compounds can act as free radical
`scavengers and antioxidants, and they have also been used
`as protecting agents in external radiation therapy (Rafferty
`et al., 2002; Srinivasan et al., 1997). In irradiation studies
`with sulfur- and selenium-containing amino acids, seleno-
`methionine was noted to have good radical scavenging
`ability (Colombetti and Monti, 1972) when radicals were
`monitored by E.P.R. However, to the best of our knowl-
`edge, there is no report of selenium compounds being used
`as radiostabilizers in radiopharmaceutical development.
`In this study,
`it was observed that thiol-containing
`reducing agents, e.g., Cys, ME, and DTT, could partially
`reverse the oxidation of the Met residue in [177Lu]Lu-
`AMBA. This ability might be useful in preventing possible
`oxidation of methionine-containing compounds during
`long-term storage. Another observation in this study was
`that a high degree of radioprotection was demonstrated for
`freshly prepared PDTC solutions (Fig. 3). However, a
`
`precipitate formed in PDTC-containing formulations after
`1 month of storage at 20 1C, and showed no protection
`afterwards (Table 4).
`In addition to the optimization of radiostabilizers for the
`prevention of radiolytic damage to [177Lu]Lu-AMBA both
`the [177Lu]Lu-
`during and after 177Lu incorporation,
`AMBA formulation wa