Preclinical Comparison of the 64Cu- and 68Ga-Labeled GRPR-Targeted Compounds RM2 and AMTG, as Well as First-in-Humans [68Ga]Ga-AMTG PET/CT

Visual Abstract

Al though prostate cancer (PCa) is associated with a high morbidity and mortality in metastasized castration-resistant PCa (mCRPC) (1), treatment has recently made some progress due to approval by the U.S. Food and Drug Administration and the European Medicines Agency of radioligand therapy (RLT) targeting the prostate-specific membrane antigen (PSMA) as a third-line therapy (2,3).However, in approximately 10%-20% of patients with recurrent PCa, a sufficiently high PSMA expression either is not present in PSMA-targeted PET imaging (4,5) or is lost in the course of subsequent treatment.Indeed, loss of PSMA expression may reflect aggressive transdifferentiation, a resistance mechanism to currently available standard therapies (6,7).In these patients, a PCa-atypical metastatic pattern is frequently observed (especially with visceral metastasis), whereas the classic adenocarcinoma features are often lost.Furthermore, PCa is known to be highly heterogeneous (8), which is why alternative compounds for imaging and RLT of PCa are required.
To confirm a sufficient GRPR expression on PCa cells before GRPR RLT, imaging of preferably PSMA-negative PCa patients is required first, which is why, with regard to a potential RLT using [ 177 Lu]Lu-AMTG, it would be advantageous to have a pendant for PET imaging.Due to the presence of a DOTA chelator, the use of the established theranostic pair 68 Ga and 177 Lu is well feasible.Besides 68 Ga, 64 Cu has recently emerged as an interesting alternative for PET imaging because of its longer half-life (12.7 h) and positron energy (E b1,max , 653 keV), which is similar to the positron energy of 18 F and thus enables a high spatial resolution in PET despite its low positron decay probability of approximately 17% (20).
Hence, this study aimed to elucidate whether the AMTG peptide, originally designed for RLT, could also be used for PET imaging when labeled with either 64 Cu or 68 Ga.A comparative preclinical evaluation on [ 64 Cu]Cu-/[ 68 Ga]Ga-AMTG and [ 64 Cu]Cu-/[ 68 Ga]Ga-RM2 encompassed the determination of GRPR affinity (half-maximal inhibitory concentration [IC 50 ]) and GRPR-mediated internalization on PC-3 cells, lipophilicity (as evaluated by n-octanol/phosphatebuffered saline solution distribution coefficient distribution coefficients at pH 7.4 [logD 7.4 ]), and biodistribution in PC-3 tumor-bearing mice.Moreover, we selected [ 68 Ga]Ga-AMTG for clinical translation in a first-in-humans PET/CT examination in a patient with mCRPC.

Chemical Synthesis and Labeling Procedures
A detailed description of the precursor synthesis is provided in the supplemental materials (available at http://jnm.snmjournals.org).Purification was accomplished via reversed-phase high-performance liquid chromatography (HPLC).

64
Cu and 68 Ga labeling was performed in analogy to an established protocol for 177 Lu labeling (19).A detailed description of the labeling procedures is provided in the supplemental materials.[ 64 Cu]CuCl 2 was purchased from DSD-Pharma GmbH.[ 68 Ga]GaCl 3 was provided by ITM Isotope Technologies Munich SE.The radiolabeled reference, 3-[ 125 I]I-Tyr 6 -MJ9 (Supplemental Fig. 1), was prepared according to reported procedures (19,21).Characterization of all GRPR ligands is provided in Supplemental Figures 2-4.
The synthesis of [ 68 Ga]Ga-AMTG according to good manufacturing practices for human PET/CT studies was performed using a good radiopharmaceutical practice module (Scintomics GmbH) while using an SC-01 gallium peptide labeling kit (ABX).[ 68 Ga]GaCl 3 was obtained from a GalliaPharm generator (Eckert & Ziegler) and was trapped on a PS-H 1 cartridge (ABX), which was eluted by a sodium chloride solution.The eluate was transferred in the reactor containing the AMTG precursor and the 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) buffer.The solution was heated and afterward transferred to a Sep-Pak C18 Light cartridge (Waters) for purification.After washing with water, the cartridge was eluted with ethanol and the solution was diluted with phosphate-buffered saline.The Cathivex-GV (Merck KGaA) was used as a sterile filter after the synthesis.Quality control included an instant thinlayer chromatography silica gel scan (NH 4 OAc/MeOH; Agilent), as well as an HPLC measurement against the corresponding reference compound, [ nat Ga]Ga-AMTG.Compliance with the HEPES limit was determined by a spot test.Furthermore, a sterile filter integrity test, a limulus amebocyte lysate test, and a postapplication sterility test were performed.The ethanol concentration was measured by gas chromatography analysis.

In Vitro Experiments
All in vitro experiments (IC 50 and logD 7.4 ) were performed according to a previously published procedure (19).A detailed description is provided in the supplemental materials.

In Vivo Experiments
Animal Experiments.All animal experiments were conducted according to a previously published protocol (19) and in accordance with general animal welfare regulations in Germany (German animal protection act, in the edition of the announcement dated May 18, 2006, as amended by article 280 of June 19, 2020, approval ROB-55.2-1-2532.Vet_02-18-109 by the General Administration of Upper Bavaria) and the institutional guidelines for the care and use of animals.Exclusion criteria for animals from an experiment were weight loss of more than 20%, a tumor size of more than 1,500 mm 3 , ulceration of the tumor, respiratory distress, or a change in behavior.None of these criteria applied to any animal from the experiment.Neither randomization nor masking was applied in the allocation of the experiments.The health status of the animals was specific pathogen-free according to the recommendation of the Federation of European Laboratory Animal Science Associations.The study was performed in compliance with the ARRIVE guidelines (Animal Research: Reporting of In Vivo Experiments).
Acquired data were statistically analyzed via a Student t-test via Excel (Microsoft Corp.) and OriginPro software (version 9.7) from OriginLab Corp. Acquired P values of less than 0.05 were considered statistically significant.
PET/CT in Patient.[ 68 Ga]Ga-AMTG was applied as an individual medical diagnostic test in a 72-y-old patient with advanced-stage mCRPC for whom no other diagnostic or therapeutic options were available.This use is allowed by the German Medical Act ( §13 2b Arzneimittelgesetz), which waives the need for institutional review board approval.The legal and ethical compliance of this approach has recently been reviewed by the local ethics committee in the context of requesting approval for retrospective evaluation of therapy data obtained in this way (Ethics Committee at Rostock University, file no.A 2018-0240).The patient gave written informed consent after receiving comprehensive medical information from a board-certified nuclear medicine physician.The anonymized analyses were performed in accordance with the Declaration of Helsinki and its later amendments and with the legal considerations of clinical guidelines.
A detailed description of the patient's history is provided in the supplemental materials.The patient underwent [ 68 Ga]Ga-AMTG wholebody PET/CT using a Gemini TF 16 (Philips Healthcare) at 75 min after injection of 173 MBq of [ 68 Ga]Ga-AMTG.Whole-body CT imaging was performed as auxiliary CT (120 kVp, 40 mAs).PET datasets were reconstructed using the blob ordered-subsets time-offlight protocol (3 iterations, 31 subsets), corrected for randoms, scatter, decay, and attenuation (using whole-body auxiliary CT).

Synthesis and Radiolabeling
Manual synthesis of RM2 and AMTG yielded 14% and 12% labeling precursor, respectively, after purification by reversed-phase HPLC (chemical purity .98%, determined by reversed-phase HPLC at l 5 220 nm).Complexation of all ligands with a 2.5-fold excess of [ nat Ga]Ga(NO 3 ) 3 and [ nat Cu]CuCl 2 each resulted in quantitative yields.The radiolabeled reference, 3-[ 125 I]I-Tyr 6 -MJ9, was produced in a radiochemical yield and purity of 29% and more than 98%, respectively, after reversed-phase HPLC purification.Both 64 Cu-and 68 Ga-labeling was performed manually, each resulting in radiochemical yields and purities of more than 98% as well as molar activities of 55 6 4 GBq/mmol (decay-corrected) and 35 6 3 GBq/mmol (decay-corrected).Both 64 Cu-and 68 Ga-labeled RM2, as well as AMTG, were used without further purification.
The synthesized batch used for the patient yielded 409 MBq (56% non-decay-corrected).All specifications were fulfilled.The pH of the 16 mL of solution was 7. The reference compound and the radiolabeled product displayed the same HPLC retention times.The radiochemical purity determined by HPLC was 98.7%, and the content of unbound [ 68 Ga]Ga-species was less than 0.3%.Thin-layer chromatography measurement was in line with less than 0.5% of unbound [ 68 Ga]Ga 31 .

In Vitro Characterization
The nat Cu-and nat Ga-labeled compounds exhibited high GRPR affinity on PC-3 cells, with IC 50 values in the range of 1.5-4.0nM (Fig. 1A; Supplemental Table 1; Supplemental Fig. 5).Although overall internalization was low for all GRPR ligands, the more affine 68 Ga-labeled compounds were internalized significantly higher within 1 h by PC-3 cells than were their 64 Cu-labeled counterparts (P , 0.04, Fig. 1B).It was evident for all analogs that a higher GRPR affinity led to increased internalized (Fig. 2A), as well as receptor-bound noninternalized, fractions (Fig. 2B), which is why the ratio of receptor-bound to internalized fraction was nearly constant, regardless of their GRPR affinity (Fig. 2C).LogD 7.4 was similar for all compounds (Fig. 1C).However, the 68 Ga-labeled ligands exhibited significantly lower lipophilicity than their 64 Cu-labeled analogs (P , 0.01).

In Vivo Characterization
In vivo studies on PC-3 tumor-bearing mice at 1 h after injection (Fig. 3; Supplemental Table 2) revealed favorable biodistribution profiles with similarly low off-target accumulation for 64 Cu-and 68 Ga-labeled RM2 and AMTG in most organs.Although the highest uptake values were determined for all derivatives in the tumor (14.1%-15.1% injected dose per gram [%ID/g]) and the pancreas (12.6-30.7 %ID/g), significantly increased accumulation was observed for the 64 Cu-labeled analogs in the heart and the lung as compared with their 68 Ga-labeled counterparts (P , 0.003).Moreover, activity levels in the liver were distinctly enhanced for the 64 Cu-labeled compounds (P , 0.001).Uptake values were elevated in the pancreas for the 68 Ga-labeled analogs as compared with the 64 Cu-labeled analogs (P , 0.05).Accumulation in the adrenals was significantly higher for [ 64 Cu]Cu-/ [ 68 Ga]Ga-AMTG than for [ 64 Cu]Cu-/[ 68 Ga]Ga-RM2 (P , 0.03) yet was on a very low level (1.9-2.8 vs. 1.0-1.1 %ID/g).
In general, tumor-to-background (T/B) ratios for the 68 Galabeled compounds were higher than for the 64 Cu-labeled derivatives (Fig. 4; Supplemental Table 3).Both [ 64 Cu]Cu-AMTG and [ 64 Cu]Cu-RM2 showed similar T/B ratios, except in the adrenals and kidneys, in which the ratio for the latter was slightly higher.Although [ 68 Ga]Ga-AMTG displayed enhanced T/B ratios in the muscle and the bone, [ 68 Ga]Ga-RM2 demonstrated higher T/B ratios in the spleen, the liver, and the adrenals.Because overall biodistribution patterns were comparable for both 68 Ga-labeled GRPR ligands, and [ 68 Ga]Ga-RM2 had already been applied in clinical studies, we selected [ 68 Ga]Ga-AMTG for PET imaging in a first-in-humans application.

Proof-of-Concept Study in Patient
[ 68 Ga]Ga-AMTG PET showed a favorable biodistribution, with uptake being highest in tumor lesions and the pancreas.Besides the bladder (because of excretion), no significant activity levels were found in other organs.One month previously, the patient had undergone [ 18 F]F-PSMA-1007 PET/CT, which showed only 1 subphrenic lesion with low [ 18 F]F-PSMA-1007 uptake (Fig. 5A; Supplemental Fig. 7).On [ 68 Ga]Ga-AMTG PET/CT, multiple lesions with intense focal uptake could be detected in the peritoneum, the subphrenic area adjacent to the liver, and between the left internal and external iliac arteries; these [ 68 Ga]Ga-AMTG-positive findings corresponded to soft-tissue lesions that were visualized on CT (Fig. 5B; Supplemental Fig. 7).

DISCUSSION
Based on the need for treatment options for PSMA-negative PCa patients, alternative targets such as the GRPR may become more relevant.We recently developed [ 177 Lu]Lu-AMTG, an RM2 derivative with noticeably increased metabolic stability in vivo (19), initially for an improved GRPR RLT.Because we additionally wanted to explore its potential for PET imaging, we performed a preclinical study on [ 64 Cu]Cu-/[ 68 Ga]Ga-AMTG and a first-in-humans application using [ 68 Ga]Ga-AMTG.Synthesis of the precursors was easily accessible via solid-phase peptide synthesis, and complexation with nat/64 Cu and nat/68 Ga proceeded quantitatively.The nat Cu-and nat Ga-labeled compounds were not purified before affinity studies, as we could show in a previous study that an excess of ions, such as Lu 31 , did not have any influence on GRPR affinity (19).Because of their structural similarity, all 4 compounds revealed comparably high GRPR affinity (IC 50 , 1.5-4.0nM), which met or even surpassed the values of their nat/177 Lu-labeled counterparts (IC 50 , 3.0-3.5 nM) (19).The high receptor-bound fractions (79%-84% of cell-associated activity) found for all 4 analogs corroborated well with the values determined for [ 177 Lu]Lu-AMTG/RM2 (19) and are generally expected for antagonists (22).The logD 7.4 of the 4 compounds was in a range (22.6 to 22.2) comparable to that of the previously published 177 Lu-labeled analogs (19).
Because we could show that a change in the radionuclide ( 64 Cu and 68 Ga instead of 177 Lu) had minimal impact on in vitro parameters, we expected in vivo properties similar to those of the previously reported [ 177 Lu]Lu-AMTG/RM2 (19).Indeed, the overall biodistribution profiles of [ 64 Cu]Cu-/[ 68 Ga]Ga-AMTG and [ 64 Cu]Cu-/[ 68 Ga]Ga-RM2 at 1 h after injection displayed high activity levels in the tumor and pancreas, whereas further off-target accumulation was either low or cleared rapidly within the first hour after injection.An exception was the liver, which revealed increased activity levels for the 64 Cu-labeled compounds as compared with the 68 Ga-labeled compounds.This increase was expected, since there are reports on the insufficient in vivo stability of the Cu-DOTA chelate (23)(24)(25), which led to a significant removal of 64 Cu from the DOTA chelator via transchelation to the superoxide dismutase in the liver and storage of free 64 Cu 21 ions in hepatocytes by metallothionein 1 and 2 (26).The instability of the 64 Cu-DOTA chelate could also explain the generally decreased T/B ratios of the 64 Cu-labeled ligands as compared with the 68 Galabeled ligands because of the thus higher off-target accumulation in most organs at 1 h after injection.
However, the high in vivo instability of the Cu-DOTA chelate in mice cannot be directly transferred to the human situation, because metabolism between these species is noticeably different.For example, high activity levels (6 %ID/g) were found in the murine liver at 1 h after injection of [ 64 Cu]Cu-DOTATATE (25), whereas in humans slightly elevated activity levels observed in the liver did not hamper the detection of a variety of liver metastases (27)(28)(29).Dosimetry studies on [ 64 Cu]Cu-DOTATATE in 5 patients revealed that the dose delivered to the liver was similar to that delivered to the kidneys (0.161 vs. 0.139 Gy/GBq) (27), and this dose is indeed 2-to 4-fold higher than for [ 68 Ga]Ga-DOTATATE (0.045 and 0.092 Sv/GBq, respectively) but does not affect its clinical use (30).
[ 68 Ga]Ga-AMTG (100 pmol, CB17-SCID mice) displayed a similar biodistribution profile (Supplemental Fig. 6) and thus comparable T/B ratios in most organs to the clinical standard for PET imaging of GRPRexpressing malignancies, [ 68 Ga]Ga-RM2 (10 pmol, NMRI nu/nu-mice, data taken from Mansi et al. (31)).The comparison of these 2 tracers is a limitation of this study, as biodistribution studies of [ 68 Ga]Ga-RM2 have been performed by another group using different mouse models and precursor amounts than we did for [ 68 Ga]Ga-AMTG.Because preclinical data, particularly overall pharmacokinetics, were mainly comparable for [ 64 Cu]Cu-/[ 68 Ga]Ga-AMTG and [ 64 Cu] Cu-/[ 68 Ga]Ga-RM2, both AMTG derivatives appear to be promising candidates for clinical translation.Although [ 68 Ga]Ga-AMTG  was selected for a first-in-humans investigation in a PCa patient, 64 Cu-labeled compounds might be superior to 68 Ga-labeled compounds in the future because of the longer half-life of 64 Cu (12.7 vs. 67.6 min), which would enable PET imaging over an extended time span.In addition, its lower positron energy (E b1,max , 653 vs. 1,899 keV) allows for a higher resolution in PET (20).However, it remains to be seen whether the low positron decay probability of 64 Cu generates some drawbacks for PET imaging, such as with regard to counting statistics.Moreover, the current availability of 68 Ga (generator) is a significant advantage over 64 Cu (cyclotron).For a first-in-humans application, good-manufacturing-practice synthesis of [ 68 Ga]Ga-AMTG was achieved within 36 min, yielded 409 MBq (56% non-decay-corrected), and fulfilled all specifications (clear and particle-free solution, pH 7, radiochemical purity .98%, unbound [ 68 Ga]Ga 31 # 0.5%, HEPES limit, ethanol concentration, endotoxin limit, filter integrity).
[ 68 Ga]Ga-AMTG revealed a favorable biodistribution in a PET/CT scan of an mCRPC patient at 75 min after injection, with high uptake in tumor lesions and the pancreas, whereas further off-target accumulation was low.Moreover, [ 68 Ga]Ga-AMTG PET/CT revealed distinctly more lesions and increased SUVs than [ 18 F]F-PSMA-1007 in this mCRPC patient.An epithelialmesenchymal transition is expected in PCa after several treatment lines including PSMA-targeted RLT (32), can lead to a loss of PSMA expression, and thus requires alternative treatment options.For such patients, GRPR-targeted compounds could offer an alternative option for imaging and RLT.The feasibility of [ 68 Ga]Ga-AMTG for such cases has been shown for only 1 patient to date-a limitation of this study and why further clinical evaluation has to be performed to confirm this promising preliminary result.Because [ 68 Ga]Ga-AMTG PET did not display noticeable uptake in organs other than the pancreas, and because pancreatic clearance is known to occur within the first hours after injection, this could open new possibilities for treatment of PCa, provided that GRPR expression is sufficient.Moreover, unlike PSMA inhibitors, GRPR-targeted compounds do not accumulate in the salivary glands and the kidneys, among others, which is why the use of 90 Y or a-particle-emitting radionuclides likely causes less severe side effects.

CONCLUSION
Both [ 64 Cu]Cu-and [ 68 Ga]Ga-AMTG revealed excellent preclinical data and might be valuable tools for PET imaging of PSMA-negative PCa in progressed disease stages, as well as in other cancer types such as breast cancer and glioblastoma multiforme.A first-in-humans examination in an mCRPC patient displayed favorable biodistribution patterns and did not show any biosafety issues or significant differences from the clinically established reference, [ 68 Ga]Ga-RM2.[ 68 Ga]Ga-AMTG PET/CT identified noticeably more lesions than [ 18 F]F-PSMA-1007, likely due to an aggressive transdifferentiation of PCa and, thus, limited PSMA expression.Because an enhanced in vivo stability was observed in previous studies for the AMTG peptide, an improved RLT might be achievable in these patients, rendering this peptide a valuable theranostic tool.Further patient studies will elucidate whether these promising results are reflected on a broader scale.

DISCLOSURE
A patent application on modified GRPR-targeted ligands including AMTG, with Thomas G€ unther and Hans-J€ urgen Wester as inventors, has been filed.Hans-J€ urgen Wester is founder and shareholder of Scintomics GmbH.Bernd Joachim Krause is an advisor for Terumo, Rotop, AAA/Novartis, PSI CRO, ITM, Bayer, and Janssen; receives third-party funding from AAA/Novartis, AMGEN, and Eisai; receives travel support from AAA/Novartis; and receives royalties from AAA/Novartis, Bayer, and Janssen.No other potential conflict of interest relevant to this article was reported.

KEY POINTS
QUESTION: Is the metabolically more stable peptide AMTG (as compared with RM2)-particularly designed for an improved RLT of GRPR-expression malignancies-also an option for imaging applying 64 Cu or 68 Ga?

IMPLICATIONS FOR PATIENT CARE:
Although the clinical value of [ 64 Cu]Cu-and [ 68 Ga]Ga-AMTG (and [ 177 Lu]Lu-AMTG) has to be further elucidated, this study might pave the way for clinical use of an improved theranostic peptide and, thus, patient care.