Efficient [18F]AlF radiolabeling of ZHER3:8698 affibody molecule for imaging of HER3 positive tumors

The human epidermal growth factor receptor 3 (HER3) is over-expressed in several cancers, being linked to a more resistant phenotype and hence poor patient prognosis. Imaging HER3 is challenging owing to the modest receptor number (<50000 receptors/cell) in overexpressing cancer cells. Therefore, to image HER3 in vivo, high target affinity PET probes need to be developed. This work describes two different [F]AlF radiolabeling strategies of the ZHER3:8698 affibody molecule specifically targeting HER3. The one-pot radiolabeling of ZHER3:8698 performed at 100oC and using 1,4,7-triazanonane-1,4,7-triacetate (NOTA) as chelator resulted in radiolabeled products with variable purity attributed to the radioconjugate thermolysis. An alternative approach based on the inverse electron demand Diels-Alder (IEDDA) reaction between a novel tetrazine functionalized 1,4,7triazacyclononane-1,4-diacetate (NODA) chelator and the trans-cyclooctene (TCO) functionalized affibody molecule was also investigated. This method enabled the radiolabeling of the protein at room temperature. The [F]AlF-NOTA-ZHER3:8698 and [F]AlF-NODA-ZHER3:8698 conjugates showed a specific uptake at 1 h after injection in high HER3-expressing MCF-7 tumors of 4.36 ± 0.92 % ID/g and 4.96 ± 0.65 % ID/g respectively. The current results are encouraging for further investigation of [F]AlF-NOTA-ZHER3:8698 as a HER3 imaging agent.


INTRODUCTION
HER3 is a member of the human epidermal growth factor receptor (EGFR/HER) tyrosine kinase family. Its overexpression has been found in a wide variety of cancers including breast, ovary, lung, and head and neck cancer. 1 Multiple studies have highlighted HER3 as a 2 fundamental receptor involved in the establishment of malignancy, demonstrating that HER3 overexpression correlates with advanced disease stage and decreased overall survival. [2][3][4][5][6][7] Moreover, signaling by HER3 has recently been identified as a prominent mediator of tumor resistance to HER1 and HER2-targeted therapies. 1,[8][9][10][11] This has driven the development of HER3-targeting monoclonal antibodies (mAbs), some of which have already entered clinical trials. [12][13][14][15] However, the lack of suitable biomarkers to evaluate HER3 status impedes the effective patient selection for these therapies. Radionuclide-based molecular imaging of HER3 receptor expression could guide patient stratification for HER3 targeted therapies, and more importantly aid monitoring early tumor response to therapeutic intervention. Recently, anti-HER3 mAbs and F(ab') 2 fragments were radiolabeled with 64 Cu and 89 Zr for molecular imaging of HER3 expression. [16][17][18] Additionally, a 89 Zr radiolabeled mAb (GSK2849330) is currently under evaluation in patients with advanced solid tumors (clinical trial identifier: NCT12345174). However, clinical application of full-length radiolabeled antibodies as imaging agents is not optimal because of their long biological half-life and relatively poor tumor penetration which can affect image quality. By contrast, affibody molecules undergo rapid clearance as a result of their smaller size (ca 7 kDa), with elimination mainly through the kidneys, increasing the tumor-to-background ratio. Affibody molecules are non-toxic scaffold small proteins characterized by a high binding affinity (K d in the low-nanomolar to picomolar range) and high target specificity. Furthermore, the introduction of a unique cysteine residue at the C-terminus of the molecules allows the site-specific conjugation of chelators, dyes or radiolabeled groups to the protein. 19,20 The high affinity affibody molecule Z HER3:8698 has been previously selected and radiolabeled with the single photon emission computed tomography (SPECT) radioisotopes 99m Tc and 111 In in order to image HER3 expression in vivo. [21][22][23] To successfully visualize the receptor, the authors had to overcome a variety of challenges including the modest overexpression of the HER3 target (ca 50000 receptors/cell in high-expressing cell lines such as BT-474) and the natural high expression of HER3 in normal organs and tissues (such as intestine, salivary glands and lung). 22,24 Additionally, the same research group reported that the affibody molecule uptake in tumors having modest receptor overexpression is highly influenced by the injected protein dose. They identified 1 µg of radiolabeled protein to be the optimal quantity for HER3 imaging, since higher amounts, associated to lower specific activity, resulted in diminished tumor uptake. 25 his study describes the development of affibody-based agents suitable for 18 F positron emission tomography (PET) imaging of HER3-expression. Because of its half-life (109 minutes), the PET radionuclide fluorine-18 is more compatible with the pharmacokinetics of the affibody molecule than SPECT radioisotopes enabling the acquisition of high quality PET images at 1-3 hours post injection. The affibody molecule Z HER3:8698 was radiolabeled using the rapid and efficient aluminum 18 F-fluoride ([ 18 F]AlF) approach, based on the formation of a stable 18 F-aluminum bond, and consequent coordination by macrocyclic ligands. 26,27 The [ 18 F]AlF radiolabeling method can be performed in a single step (one-pot reaction) and it has been used to successfully radiolabel a variety of peptides and proteins including affibody molecules. [28][29][30][31][32][33][34][35] Two different radiolabeling strategies were investigated to establish the best protocol for a reproducible production of [ 18 F]AlF radiolabeled Z HER3:8698 suitable for imaging HER3 expression. Firstly, the synthesis reported in the literature using NOTA as the chelator for the [ 18 F]AlF complex 3 was followed. 36 Secondly, the inverse electron demand Diels-Alder (IEDDA) reaction to radiolabel the affibody molecule under mild reaction conditions in a two-step method was adopted. McBride et al. reported the radiolabeling of a heat-sensitive antibody Fab' fragment using the maleimide functionalized 1,4,7-triazacyclononane-1,4-diacetate chelator (NODA-MPAEM) which was mixed with Al 3+ and [ 18 F]Fand then conjugated to the reduced thiol group of the protein. 37 Kiesewetter et al. employed a similar maleimide-thiol conjugation approach to synthesize the affibody-based radioconjugate [ 18 F]FBEM-Z HER2:342 . A limitation of this approach is that an initial reduction of the thiol group of the terminal cysteine of the affibody molecule was required as part of the radiolabeling protocol. The reducing agent had to be removed by size exclusion chromatography prior to conjugation of the reduced affibody molecule with the maleimide functionalized prosthetic group; as a result, the radiolabeling reaction lasted more than an hour and required multiple purification steps. 38 The IEDDA reaction is an attractive alternative to the maleimide-thiol conjugation because it eliminates the need for the reduction step as part of the radiolabeling procedure. This approach, which is based on the reaction between a tetrazine and a strained cycloalkene, is chemospecific, proceeds in aqueous medium at room temperature, and forms stable products with negligible side products. 39 These characteristics have prompted an increased interest in IEDDA reactions for bioconjugation as well as pretargeting purposes. [39][40][41] To perform this synthetic methodology, the novel [ 18 F]AlF-3, containing the NODA chelator instead of NOTA, was prepared and subsequently reacted with the TCO-functionalized-affibody molecule at room temperature via an IEDDA reaction. This methodology enables the radiofluorination of heatsensitive compounds. The two different strategies described above were developed for the [ 18 F]AlF radiolabeling of the HER3 targeting affibody molecule Z HER3:8698 with a particular focus on protein stability and radiolabeling efficiency (expressed as specific activity of the final radiocojugates). The performance in vitro and in vivo of both radioligands has been further compared using high HER3-expressing MCF-7 human breast cancer cells and tumor xenografts.

RESULTS AND DISCUSSION
Synthesis of the affibody conjugates. When analyzed by RP-HPLC, the affibody molecule having a cysteine residue at its C-terminus (Z HER3:8698 -Cys) was observed to be present in solution mostly in a dimeric form due to the formation of an intermolecular disulfide bond ( Figure S1A). In order to proceed with the conjugation using the maleimide functionalized NOTA or TCO, the affibody dimer required reduction to its monomeric free sulfhydryl form ( Figure S1B). Different protocols described in the literature use an excess of a reducing agent such as DTT or TCEP-HCl which is then removed by gel permeation chromatography using either NAP-5 or Zeba columns just before proceeding to the conjugation reaction. 38,42 When such procedures were followed for the present study, the outcome was a substantial loss of affibody molecule because of the similarity between its molecular weight and the MWCO of the columns. The protein loss was found to be more prominent when low sample volumes and low protein concentrations were used. Moreover, the protein was usually recovered from size exclusion columns in a more diluted solution requiring the addition of a molar excess of maleimide functionalized compounds (i.e. MMA-NOTA and TCO-PEG 3 -Maleimide) which 4 had to be subsequently removed by gel permeation chromatography. Added to the loss of protein, this overall multi-step process was found to be time consuming and required the use of multiple size exclusion columns. To improve the product yield and simplify its purification, a one-pot conjugation reaction was developed using TCEP-HCl as the reducing agent, which was not removed from the mixture. A quantitative yield of the products NOTA-Z HER3:8698 and TCO-Z HER3:8698 in the shortest time possible was thus achieved, indicated by the disappearance of the starting material when analyzed by RP-HPLC (Figures S1, S2A and S3A). To attain extremely pure products for the radiolabeling reaction, a single final purification of the conjugates by semi-preparative RP-HPLC was performed ( Figures S2B  and S3B). Following this optimized synthetic method, the pure products were obtained in a ca 60% and ca 45% yield for NOTA-Z HER3:8698 and TCO-Z HER3:8698 conjugates respectively compared to the 15-30% yield obtained when conventional methods were utilized. All products were characterized by MALDI-mass spectrometry (Supporting Information).

Synthesis of NODA-tetrazine (3).
The design of the tetrazine-chelator construct was dictated by the hydrophilicity of the final affibody conjugate and the synthetic accessibility of the chosen tetrazine. Therefore, a small molecular weight tetrazine-chelator having the closest structure to MMA-NOTA was developed. The non-commercially available tetrazine 1 was synthesized in a simple one step process following a literature procedure. 43 The NODA macrocycle was selected because it showed a higher labeling efficiency compared to NOTA, believed to be caused by the absence of a destabilizing electron-withdrawing carbonyl adjacent to a coordinating nitrogen atom of the macrocycle. 41 The tosylated tetrazine 2 was reacted with NO2AtBu and subsequently deprotected using TFA yielding the NODAcontaining tetrazine 3 (Scheme 1). The final product was purified by RP-HPLC to ensure a high purity material for the subsequent radiolabeling ( Figure S4).  pot reaction at pH 4, 100ºC for 15 min using ethanol as organic co-solvent (50% v/v) as the addition of a hydrophilic organic solvent showed to increase the radiolabeled product yield. 32,33,36 The radiochemical conversion and the specific activity of the radioconjugate were 38.8 ± 5.8 % (non-decay corrected) and 0.8-1.5 MBq/µg (6.0-11.9 GBq/µmol) respectively ( Figure  S13). Mirroring Glaser et al., it was observed that optimal radiolabeling efficiencies were achieved by heating the [ 18 F]AlF-complexation reaction at 100°C in conjunction with the formation of a by-product assumed to originate from the decomposition of the radioconjugate ( Figure  1A). 33 Therefore, a procedure comprising of RP-HPLC followed by HLB-SPE purification was developed to produce [ 18 F]AlF-NOTA-Z HER3:8698 with a RCP > 98% verified by HPLC ( Figure 1B and Figure S5). To confirm its purity, the isolated product was analyzed using radio-SDS-PAGE and silver staining. Only one band corresponding to the molecular weight of the monomeric affibody molecule was detected on the gel autoradiography. No other bands which could indicate aggregation of the radioconjugate (high molecular weight compounds) or degradation (low molecular weight species) were observed ( Figure S15). 21   The alternative two-step approach first required the radiolabeling of the tetrazine-bearing NODA chelator (3) with the [ 18 F]AlF complex followed by the IEDDA reaction with the TCO-functionalized Z HER3:8698 affibody molecule (Scheme 2). Crucially, the mild conjugation conditions used in the IEDDA reaction can limit the formation of the by-product ( Figure 1A, product eluting at 4:22) caused by the exposure of the conjugate to the high temperatures required for radiolabeling step. [ 18 F]AlF-3 was prepared in high RCC (70-95%, calculated by HPLC and non-decay corrected) in a one pot reaction using ethanol as organic co-solvent (50% v/v). To confirm the hydrophilicity of [ 18 F]AlF-3, its distribution coefficient (logD) between PBS (pH 7.4) and n-octanol was determined (Supporting Information). The logD 7.4 was found to be -1.87 ± 0.01 which confirmed that the product possessed suitable hydrophilicity for further investigation. The radio-HPLC of [ 18 F]AlF-3 shows the presence of predominantly the two species [ 18 F]AlF-3A and [ 18 F]AlF-3B ( Figure 2). When analyzed immediately after the end of reaction, the mixture was composed of mainly the product  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60 eluting at 7 min 55 seconds ([ 18 F]AlF-3B) with a minimal amount of product having a retention time of 7 min 11 seconds ([ 18 F]AlF-3A). The [ 18 F]AlF-3B/A ratio (based on peak areas) was found to be generally in the range of 11-23. After 2 hours, the ratio decreased to as little as 0.26 indicating that [ 18 F]AlF-3B converted into [ 18 F]AlF-3A over time, although not completely. The same behavior was observed when the radiolabeling was carried out in buffer without ethanol ( Figure S6). Alternative organic co-solvents to ethanol (i.e. acetonitrile, tert-butanol) produced the same number of [ 18 F]AlF-3 forms which surprisingly did not convert over time ( Figure S7 and S8). A similar NODA-derivative, lacking the tetrazine moiety, showed a single radioactive product when radiolabeled and analyzed by RP-HPLC (data not shown) indicating the involvement of the tetrazine in the generation of the two species. This was further confirmed by the formation of a comparable set of products when a similar compound, but having a longer linker between the NODA chelator and the tetrazine moiety, was radiolabeled and analyzed by RP-HPLC (data not shown). The precise molecular structures for [ 18 F]AlF-3A and [ 18 F]AlF-3B are uncertain at present and will be the subject of further study. Non-purified [ 18 F]AlF-3 was reacted with lyophilized TCO-Z HER3:8698 in a 7:1 molar ratio at room temperature yielding the product almost quantitatively ( Figure S9). RP-HPLC followed by HLB-SPE were used to purify the final product [ 18 F]AlF-NODA-Z HER3:8698 from the excess of [ 18 F]AlF-3 and free 18 F-fluoride. As shown in Figure 3 and Figure S10, the IEDDA reaction yielded two radiolabeled products identified by HPLC.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  At this stage, an investigation on whether the formation of the IEDDA adducts was dependent on the species composition of [ 18 F]AlF-3 was carried out. [ 18 F]AlF-3A and [ 18 F]AlF-3B were hence collected separately by HPLC and, after confirming that no conversion was occurring once isolated ( Figure S11), they were incubated with an excess of TCO-Z HER3:8698 . HPLC analysis of the reaction mixtures showed that [ 18 F]AlF-3A did not react with TCO-Z HER3:8698 even after 4 hours ( Figure S12A). Conversely, [ 18 F]AlF-3B reacted with the TCO-affibody conjugate generating the previously described multiple peaks pattern ( Figure 3 and Figure S12B). The reason behind the inactivity of product [ 18 F]AlF-3A is under investigation. To reduce its effect on the yield of the final radioconjugate, it was necessary to either add the freshly prepared [ 18 F]AlF-3 to TCO-Z HER3:8698 promptly or synthesize [ 18 F]AlF-3 using an alternative organic co-solvent to ethanol to prevent species interconversion. The two peaks from the IEDDA adduct [ 18 F]AlF-NODA-Z HER3:8698 were collected separately by HPLC and investigated further. When the peak having a R t of 9 min (Peak 1, Figure 3) was reinjected, one single product with matching retention time to the one in the original separation was detected ( Figure S13A). Attempts to isolate the peak at 10 min 36 sec (Peak 2, Figure 3) were unsuccessful. Each trial resulted in a mixture of the two peaks suggesting that Peak 2 partially transforms into Peak 1 over time ( Figure S13B). Selvaraj et al suggested that by reacting with water, the IEDDA ligation product [ 18 F]AlF-NODA-Z HER3:8698 -Dihydropyridazine undergoes a series of structural rearrangements that ultimately leads to the stable aromatic [ 18 F]AlF-NODA-Z HER3:8698 -Pyridazine (Scheme 3). 44 Therefore, when analyzed by HPLC the radioconjugate is present as a combination of the aromatized [ 18 F]AlF-NODA-Z HER3:8698 -Pyridazine (Peak 1, Figure 3) and of [ 18 F]AlF-NODA-Z HER3:8698 -Dihydropyridazine (Peak 2, Figure 3). Furthermore, when [ 18 F]AlF-NODA-Z HER3:8698 was analyzed as a mixture by radio-SDS-PAGE only a single band with a comparable molecular weight to the Z HER3:8698 affibody was visible. This indicated that neither of the two products (Peak1 and Peak2) was the result of either aggregation or degradation of the protein ( Figure  S17). [ 18 F]AlF-NODA-Z HER3:8698 was isolated in high radiochemical purity exceeding 98% ( Figure 3) and the specific activity was in general higher than the NOTA radioconjugate but over a large range (0.7-2.3 MBq/µg; 5.5-18.4 GBq/µmol) ( Figure S14) which may be due to the inconsistent quantity of inert product present in [ 18 F]AlF-3. To investigate the impact of  In vitro stability studies. The stability of [ 18 F]AlF-NOTA-Z HER3:8698 and [ 18 F]AlF-NODA-Z HER3:8698 was determined by incubating the radioconjugates in mouse serum at 37°C. By HPLC analysis, it was found that 97.9 ± 0.5% of [ 18 F]AlF-NOTA-Z HER3:8698 and 91.5 ± 1.2% of [ 18 F]AlF-NODA-Z HER3:8698 remained intact after 1 h (Figures S15). In both cases, the release of 18 F (most probably in the form of the [ 18 F]AlF complex) was observed as the main product of decomposition of the radioligands with [ 18 F]AlF-NODA-Z HER3:8698 less stable to degradation. However, due to the short blood half-life of the affibody molecule, it is improbable that the breakdown of the radioconjugate would have a significant impact on its further in vivo validation. 45 Notably, the HPLC trace of [ 18 F]AlF-NODA-Z HER3:8698 ( Figure  S15B) shows the presence of two peaks (Peak 1 and Peak 2, Figure 3) corresponding to [ 18 F]AlF-NODA-Z HER3:8698 -Dihydropyridazine and [ 18 F]AlF-NODA-Z HER3:8698 -Pyridazine as previously described (Scheme 3). This confirms that the radioconjugate is present as a mixture in physiological conditions. A residual activity associated with the pelleted protein (20 ± 0.1% and 18.3 ± 2.3% for [ 18 F]AlF-NOTA-Z HER3:8698 and [ 18 F]AlF-NODA-Z HER3:8698 respectively) shows some non-specific affinity of the radiotracers towards the serum protein.
Having a less than 50% binding affinity value, both radioconjugates can be considered as low plasma protein binding compounds .

Radiosynthesis of [ 18 F]AlF-3 and IEDDA reaction with TCO-Z HER3:8698.
A mixture containing 3 in water (4 µL, 28 nmol), 2 mM AlCl 3 in 0.5 M sodium acetate pH 4 (14 µL, 18-28 nmol), non-purified 18 F-fluoride (460-500 MBq) and an equal volume of EtOH (or acetonitrile, or tert-butanol) was incubated at 100 °C for 15 min. Analytical RP-HPLC (Gradient D, Supporting Information): R t [ 18 F]AlF-3 = 7.1 and 7.5 min. Specific activity = 10.6-16.6 GBq/µmol (decay corrected to the end of reaction). The reaction solution containing [ 18 F]AlF-3 was quickly added to lyophilized TCO-Z HER3:8698 (3.5 nmol, 3 to TCO-Z HER3:8698 molar ratio 7:1) and the mixture was incubated at ambient temperature with mixing (800 rpm) for 17 min before purification by RP-HPLC using Gradient A (Supporting Information). The collected fraction containing the product was diluted with 0.1% aq TFA (3 mL) and loaded on an Oasis HLB-SPE cartridge. The trapped radioactivity was washed with 0.1% aq TFA (3 mL) and then eluted with 50% ethanol/water (v/v, 100 µl). The product was quantified by measuring the UV absorbance at 280 nm on a Nanodrop 2000. Synthesis time (from the beginning of the reaction) = ca 70 min. The RCC (non-decay corrected) was expressed as the mean of n = 12 experiments ± SD. Protein recovery = 34.0-54.8% In vitro serum stability assay. 34 The stability of [ 18 F]AlF-NOTA-Z HER3:8698 and [ 18 F]AlF-NODA-Z HER3:8698 , with respect to change in RCP and loss of radioactivity from the affibody molecule, was assessed by incubating the purified [ 18 F]AlF-radioconjugates (ca 4 MBq) in mouse serum (500 µL) in a Thermomixer at 37°C for 1 h (300 rpm). The mixture was then precipitated by addition of EtOH (300 µL) and centrifuged at 16000 × g for 2 min at 21°C. DMF (300 µL) was added to the supernatant which was then centrifuged at 16000 × g for 2 min at 21°C. The supernatant was acidified with 0.1% aq TFA (300 µL), filtered through a 0.2 µm Iso-Disc PVDF syringe filter and analyzed by RP-HPLC using Gradient A (Supporting Information). The radioactivity associated with the protein from the centrifuge spins was measured in a dose calibrator. Aqueous non-purified [ 18 F]Fluoride solution (ca 4 MBq) incubated in mouse serum (500 µL) was processed in the same way and used as control (to confirm the R t of free [ 18 F]Fluoride). The experiments were performed in triplicate. The data are expressed as the average of n = 3 measurements ± SEM.
In vitro binding affinity and specificity of uptake of the radioconjugates. The dissociation constants (K d ) of the 18 F-radiolabeled conjugates were assessed by a saturation binding assay using the high HER3-expressing MCF-7 cell line. The cells (5 × 10 5 ) were plated on 12-well plates and cultured overnight. After removing the growth medium, the cells were washed with PBS and incubated with increasing concentrations of the radioconjugates (final concentrations of 0.01, 0.1, 0.5, 1.0, 2.5, 3.5, and 5.0 nM, ca 0.05-37 kBq/well) diluted in non-supplemented DMEM (1 mL). Non-specific binding was determined by pre-incubating the cells with 100-fold molar excess of the non-radiolabeled affibody molecule for 10 min. After 1 h incubation at 4°C, the cells were rinsed twice with PBS prior to detachment with trypsin-EDTA. Afterwards, the cells from each well were transferred to scintillation vials and  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60 the total cell-associated radioactivity measured in a gamma counter. To estimate the K d , the specific binding was determined by subtracting the fraction of non-specific binding from the total binding. The data were plotted as the amount (nM) of bound vs free radioconjugate. The binding curve was fitted to a one-site receptor-binding model using GraphPad Prism 6.00. The specificity of binding of each radioconjugate was evaluated using MCF-7 and MDA-MB-231 cells (5 × 10 5 ) plated on 12-well plates 24 h prior to the experiment. The cells were pre-incubated with a 100-fold molar excess of non-radiolabeled affibody molecule for 10 min at 4°C, followed by incubation with the radioconjugate (1 nM, ca 5 kBq/well) for 1 h at 4°C. The cells were rinsed twice with PBS, trypsinized, and collected into scintillation vials. The radioactivity was assessed using a gamma counter. The specificity of binding was normalized to the maximum cell-associated radioactivity per experiment and presented as mean of n = 3 independent measurements ± SEM.

In vivo evaluation.
All experiments were performed in compliance with licenses issued under the UK Animals (Scientific Procedures) Act 1986 and following local ethical review. Studies were compliant with the United Kingdom National Cancer Research Institute Guidelines for Animal Welfare in Cancer Research. 49 Female NCr athymic mice (6-8 weekold) were subcutaneously injected on the shoulder with MCF-7 or MDA-MB-231 cells (7.5 × 10 6 /mouse) suspended in 30% Matrigel. For MCF-7 cells, 17β-estradiol pellets (0.72 mg, 90 days release) were implanted 48 h before cell inoculation, and remained in place until the end of the study. Tumors were allowed to grow for 3-4 weeks until reaching 100 mm 3 . PET/CT imaging studies were conducted using an Albira PET/SPECT/CT imaging system. Mice were administered the radioconjugate (1 µg in 100 µL of 0.9% sterile saline, 0.7-0.8 MBq/mouse) by intravenous tail vein injection and approximately 5 minutes prior to imaging were anesthetized using isoflurane/O 2 mixture (1.5-2.0 % v/v) and placed prone in the center of the scanner's field of view. Whole body PET static images were acquired 1 h post radioconjugate injection for the duration of 10 min with a 358 to 664 keV energy window, followed by CT acquisition. The image data were normalized to correct for PET non-uniformity, dead-time count losses, positron branching ratio, and physical decay to the time of injection. No attenuation or partial-volume averaging corrections were applied. The PET images were reconstructed using a MLEM algorithm (12 iterations) with a voxel size of 0.5 × 0.5 × 0.5 mm 3 . Whole body standard high resolution CT scans were performed with the X-ray tube setup at a voltage of 45 kV, current of 400 µA and 250 projections (1s per projection) and a voxel size of 0.5 × 0.5 × 0.5 mm 3 . The CT images were reconstructed using a FBP algorithm. Image analysis was performed using the PMOD software package. Immediately after the image data acquisition, the mice were euthanized by cervical dislocation for the biodistribution studies. The major organs/tissues were dissected, weighed, and the radioactivity was measured in a gamma counter. The percentage of the injected dose per gram of tissue (%ID/g) was determined for each organ/tissue. The data are expressed as the average of n = 3 mice ± SD.