Towards Complete Tumor Resection: Novel Dual-Modality Probes for Improved Image-Guided Surgery of GRPR-Expressing Prostate Cancer

Nuclear and optical dual-modality probes can be of great assistance in prostate cancer localization, providing the means for both preoperative nuclear imaging and intraoperative surgical guidance. We developed a series of probes based on the backbone of the established GRPR-targeting radiotracer NeoB. The inverse electron demand of the Diels–Alder reaction was used to integrate the sulfo-cyanine 5 dye. Indium-111 radiolabeling, stability studies and a competition binding assay were carried out. Pilot biodistribution and imaging studies were performed in PC-3 tumor-bearing mice, using the best two dual-labeled probes. The dual-modality probes were radiolabeled with a high yield (>92%), were proven to be hydrophilic and demonstrated high stability in mouse serum (>94% intact labeled ligand at 4 h). The binding affinity for the GRPR was in the nanomolar range (21.9–118.7 nM). SPECT/CT images at 2 h p.i. clearly visualized the tumor xenograft and biodistribution studies, after scanning confirmed the high tumor uptake (8.47 ± 0.46%ID/g and 6.90 ± 0.81%ID/g for probe [111In]In-12 and [111In]In-15, respectively). Receptor specificity was illustrated with blocking studies, and co-localization of the radioactive and fluorescent signal was verified by ex vivo fluorescent imaging. Although optimal tumor-to-blood and tumor-to-kidney ratios might not yet have been reached due to the prolonged blood circulation, our probes are promising candidates for the preoperative and intraoperative visualization of GRPR-positive prostate cancer.


Introduction
Worldwide, prostate cancer (PCa) is the second most frequently diagnosed cancer among men, with about 1.4 million new cases in 2020 alone [1]. The surgical removal of the prostate gland, in whole or in part, combined with pelvic lymph node dissection is one of the most widely used treatment options to cure localized PCa [2]. Although successful in many cases, the recurrence rate after radical prostatectomy is still as high as 20-40% [3]. One of the indicators for an increased risk of relapse is the observation of positive surgical margins [4]. As well as the multifocal nature of many primary prostate tumors, the need to maintain physiological functions, such as potency and continence, constitutes an additional challenge for surgeons [5,6]. Such a nerve-sparing surgical approach is a complex procedure and may therefore come at the expense of complete tumor eradication [7].
To reduce recurrence and, therefore, the need for secondary treatment, new imaging techniques that provide accurate surgical guidance may offer a solution. Over the years, fluorescence-guided surgery has emerged as a powerful tool for real-time intraoperative imaging in the surgical field [8]. To enable the precise removal of cancer tissue, receptortargeting agents coupled to a fluorescent dye can lead to improved tumor localization [8]. Fluorescent tumor-targeting tracers have a high spatial resolution but are limited by their low tissue penetration. This is where nuclear medicine can play a pivotal role, as a radioisotope can be used for non-invasive preoperative nuclear imaging that supports surgical planning, as well as for the determination of the approximate localization of deeper lesions intraoperatively, with high sensitivity [9]. The benefits of the complementary use of a radioactive and fluorescent signal for image-guided surgery have led to an increased interest in the development of nuclear and optical dual-modality probes [10,11].
In nuclear medicine, recent advances in PCa-targeting radiotracers have provided a range of promising vectors for tumor targeting. One of the aberrantly overexpressed targets in PCa is the gastrin-releasing peptide receptor (GRPR) [12,13]. Imaging studies with GRPRtargeted radiotracers have demonstrated high tumor uptake and excellent visualization of tumor lesions in cancer patients [14][15][16][17]. NeoB (formerly known as NeoBOMB1) is one such established radiotracer, with a high binding affinity for GRPR, and is favorable for in vivo pharmacokinetics [18,19]. NeoB is therefore an excellent molecule to serve as a basis for the development of a dual-modality probe.
In this study, we integrated the sulfo-cyanine 5 fluorescent (sCy5) dye into the DOTAcoupled GRPR antagonist NeoB, using the inverse electron-demand Diels-Alder reaction (IEDDA) [20,21]. A tetrazine (Tz) moiety was coupled to the fluorescent dye and a transcyclooctene (TCO) group was incorporated to the backbone of NeoB via an additional lysine residue. A linker (pADA or PEG 4 ) was introduced between the binding domain and the DOTA chelator, in combination with or without a PEG 4 linker between the TCO group and the lysine. The methodological approach taken in this research resulted in a panel of two single-and four dual-modality probes. After the development of these compounds, we assessed the effect of the dye attachment on the stability and affinity for GRPR in vitro and the tumor-targeting capability and biodistribution in vivo. With this study, we aim to demonstrate the potential of the novel GRPR-targeting dual-modality probes for preoperative and intraoperative PCa visualization.

TCO
Once the peptide sequence was completed, the Boc-NH-pADA-OH linker was coupled to the peptide (Scheme 1). Coupling of the linker was carried out by the treatment of 1 with Boc-NH-pADA-OH (2 equiv.), a mixture of HBTU/Oxyma Pure (3.9 and 4 equiv., respectively) and DIPEA (8 equiv.). The beads were shaken for 2 h at rt, then they were washed thrice with DMF. Subsequently, the peptide 2a was cleaved from the resin using a cleavage cocktail of 1,1,1,3,3,3-hexafluoro-2-propanol/dichloromethane (HFIP/DCM) (2 mL, v:v = 20:80). The beads were mixed for 1 h at rt and washed twice with the cleavage cocktail, then the liquid phase was collected in a round-bottomed flask. The solvent was removed using a rotary evaporator. The resulting peptide was precipitated using cold diethyl ether and collected by centrifugation. After cleavage, the coupling of 4-amino-2,6-dimethylheptane (2.5 equiv.) on the C-terminus was performed using benzotriazole-1-yl-oxy-tris-pyrrolidino-phosphonium hexafluorophosphate (PyBOP) (2.5 equiv.) and DIPEA (5 equiv.) in DMF. The reaction mixture was stirred for 1 h at rt, the solvent was removed under a vacuum and the product was collected by precipitation in cold diethyl ether. Global deprotection of the peptide was performed by treatment of the peptide with trifluoroacetic acid/water/triisopropyl silane (TFA/H 2 O/TIS) (2 mL, v:v:v = 95:2.5:2.5) for 1 h at rt. Later, the TFA was removed using a gentle air stream, and the resulting crude product 3a was washed with cold diethyl ether and collected by centrifugation. The 3a was purified with a Sep-Pak C18 35 cc Vac cartridge (10 g) (Waters, Etten-Leur, The Netherlands). The column was pre-conditioned with methanol (100 mL) and H 2 O (200 mL). The peptide was subsequently loaded onto the column and washed with water (200 mL) until the pH of the eluate became neutral. Then, 3a was eluted using a mixture of H 2 O/ACN (v:v = 1:1, 4 × 20 mL), followed by two fractions of 20 mL ACN. The fractions containing the product were combined and lyophilized for further experiments. The final product 4a was prepared by adding trans-cyclooctene-N-hydroxysuccinimide ester (TCO-NHS ester, 3 equiv.), triethylamine (10 equiv.) and H 2 O/ACN (2 mL, v:v = 1:1). The reaction was stirred for 3 h at rt. The crude compound was purified by the semi-preparative HPLC to provide 4a as a white solid (8.6 mg, 2.0% yield). Analytical HPLC retention time of 4a: t R = 18.

TCO-PEG 4 -fQWAVGH-NHCH[CH 2 CH(CH 3 ) 2 ] 2 (4b)
Compound 4b was synthesized according to the protocol previously described for 4a, with Boc-NH-PEG 3 -COOH (PEG 4 ) as a linker (Scheme 1). The crude product was purified by semi-preparative HPLC to yield 4b as a white solid (11.7  The first steps of the synthesis were performed from intermediate 1, as described in the protocol above. The linker was protected with a Fmoc group instead of a Boc protecting group on the N-terminal position. Deprotection of Fmoc was carried out using 20% piperidine in DMF. The coupling of Fmoc-L-Lys(Boc)-OH was achieved with 4 equivalents of amino acid, HBTU/Oxyma Pure (3.9 and 4 equiv., respectively) and DIPEA (8 equiv.). The beads were shaken for 1 h at rt. After coupling, the beads were washed with DMF (3 × 1 mL), and Fmoc deprotection was performed as reported above. The coupling of 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid tris tert-butyl ester (DOTA-tris(tBu) ester, 3 equiv.) was realized in the presence of PyBOP (3 equiv.), DIPEA (6 equiv.) and DMF (3 mL) (Scheme 2). The reagents and the resin were mixed for 2 h at rt. The beads were washed with DMF (3 × 1 mL), followed by the cleavage of the peptide from the solid support, the coupling of 4-amino-2, 6-dimethylheptane on histidine, global deprotection, C18 Sep-Pak purification and conjugation of the TCO-NHS ester, as described above.

DOTA-L(TCO)-PEG 4 -fQWAVGH-NHCH[CH 2 CH(CH 3 ) 2 ] 2 (9b)
After the conjugation of TCO, the crude product was purified by semi-preparative HPLC to obtain 9b as a white solid ( TCO-PEG 4 NHS ester was coupled to 8a to give 9c, following the same protocol described above (Scheme 2). Compound 9c was purified by semi-preparative HPLC and was obtained as a white solid (17.

Tetrazine-Sulfo Cyanine 5 (Tz-sCy5)
The synthesis of Tz-Cy5 was performed, following the protocol described by Sasmal and coworkers [22]. The characteristics of the compound are provided in the Supplementary Materials.

Synthesis of 10
Compound 10 was prepared by reacting 4a (2.01 mg, 1.49 µmol) and Tz-sCy5 (1.33 mg, 1.1 equiv.). Both starting materials were dissolved in H 2 O/ACN (v:v = 1:1) and incubated at 37 • C for 10 min in an Eppendorf tube protected from light. The reaction mixture was purified by the semi-preparative HPLC to give 10 as a blue solid (  cool down for 5 min and diethylenetriaminepentaacetic acid (DTPA) (5 µL) was added to complex free 111 In. Then, the radiochemical yield of the 111 In-labeled peptides was determined using radio-HPLC.

Determination of the Distribution Coefficients (LogD 7.4 )
Distribution coefficients (LogD 7.4 ) for the 111 In-labeled compounds were determined by a shake-flask method. The experiments were performed in triplicate for each radioligand. A sample containing the radioligand (0.5-2.0 MBq) was dissolved in 1 mL solution of phosphate-buffered saline (0.01 M, pH 7.4) and n-octanol (v:v = 1:1). The vials were vortexed vigorously and then centrifuged at 10,000 rpm for 3 min, for phase separation. Samples (10 µL) of each phase were taken out and analyzed using a gamma counter. LogD 7.4 values were calculated, using the following equation: LogD 7.4 = log {(counts in n-octanol phase)/(counts in PBS phase)}.
The affinity of all six probes for GRPR was determined using a competition binding assay. PC-3 cells were seeded in 12-well plates (2.5 × 10 5 cells/well) one day prior to the experiment. The next day, cells were washed with warm Dulbecco's phosphate-buffered saline (PBS) (Gibco). Subsequently, cells were incubated for 1 h at 37 • C with 0.5 mL incubation medium (Ham's F-12K, 20 mM HEPES, 1% BSA, pH 7.4) containing 10 −9 M [ 111 In]In-NeoB, in the presence or absence of increasing concentrations (10 −12 to 10 −6 M) of one of the six unlabeled probes or NeoB (positive control). After incubation, the medium was removed and cells were washed twice with cold PBS. To determine the amount of activity that was taken up by the cells, cells were lysed using 1 M NaOH for > 20 min at rt, then collected and measured in a γ-counter. To determine the amount of radioactivity added per well, samples of the incubation medium containing 10 −9 M [ 111 In]In-NeoB were also measured. The results are expressed as the percentage of added dose (%AD) and were normalized to the uptake of [ 111 In]In-NeoB (in the absence of cold compound). Data represent the mean ± standard deviation (SD) of triplicate wells.

Animal Model
Seven-week-old male Balb/c nu/nu-specific and opportunistic pathogen-free (SOPF) mice (Janvier Labs, Le Genest-Saint-Isle, France) were housed in individually ventilated cages, with 4 mice per cage. Upon arrival, mice were acclimated for 1 week, with access to food and water ad libitum. Mice were subcutaneously inoculated on the right shoulder with PC-3 cells (5 × 10 6 cells suspended in 100 µL of 1/3 Matrigel (Corning Inc., Corning, NY, USA) and 2/3 Hank's balanced salt solution (Gibco). PC-3 xenografts were allowed to grow for 3 weeks. Tumor sizes were 391 ± 173 mm 3 at the start of the studies. All animal experiments were approved by the Animal Welfare Committee of the Erasmus MC and were conducted in agreement with institutional guidelines (license number: AVD101002017867, 28 September 2017). . Whole-body SPECT images (transaxial field of view (FOV) 54 mm) were acquired over 30 min using a spiral scan in normal scan mode, in list-mode acquisition. This was followed by a whole-body CT scan within 5 min, with the following imaging settings: full angle scan, angle step 0.75 degrees, normal scan mode, 50 kV tube voltage, 0.21 mA tube current, 500 µm aluminum filter. Reconstruction of the SPECT images was performed using the similarity-regulated OSEM method and MLEM method (MILabs Rec 11.00 software, MILabs B.V., Houten, The Netherlands) performing 9 and 128 iterations, respectively, at 0.8 mm 3 resolution, using 173 keV ± 10% and 247 keV ± 10% energy windows for indium-111. Two adjacent background windows per photo peak were used for triple-energy window scatter and crosstalk correction. Reconstructed volumes of SPECT scans were post-filtered with an isotropic 3-dimensional Gaussian filter of 1 mm full width, at half-maximum. The CT and registered, attenuation-corrected SPECT images were analyzed using PMOD (PMOD 3.9, Zurich, Switzerland) and quantification was performed by placing volumes of interest (VOIs) around the tumors and kidneys. An Eppendorf tube filled with a solution of indium-111 of a known activity was measured to determine the calibration factor. The total activity measured in the VOI was divided by the volume of all VOI pixel values and multiplied by the calibration factor to obtain the percentage of injected activity per volume unit (%IA/mL).

Ex Vivo Biodistribution Studies and Optical Imaging
To determine the biodistribution of the compounds after imaging (~3 h p.i.), blood was collected via cardiac puncture under isoflurane/O 2 anesthesia, after which the mice were sacrificed. The tumor and organs of interest (prostate, pancreas, spleen, liver, GI tract (stomach, small intestine, cecum, large intestine), kidneys, lungs, heart, muscle, bone, and brain) were excised, washed in PBS and blotted dry. The stomach, intestines and cecum were emptied of their contents. The tumor was cut in half; one half was freshfrozen for further analysis and the other half was collected for ex vivo optical imaging and radioactivity measurements. After imaging, the blood, tumor, and relevant organs were weighed and measured in a γ-counter. To determine the total injected radioactivity per animal, samples of the injected solutions were measured as well. The percentage of injected dose per gram (%ID/g) was determined for each tissue sample and corrected for both the injected volume and %ID present at the injection site (the tail). Because the low weight of the prostate limited accurate organ weight measurements, the average prostate weight of all animals was used.
Before gamma counter measurements, the tumor half, pancreas, kidneys, lungs, small intestine, large intestine, liver, muscle, and bone were placed in two petri dishes and ex vivo optical imaging was performed with the IVIS Spectrum system (Perkin Elmer, Waltham, MA, USA) using the following settings for all measurements: FOV 12.6 cm, medium binning, f2, 0.5 s exposure with an excitation/emission filter of 640 nm/680 nm. Living Image version 4.5.2 software (Perkin Elmer) was used to perform data analysis by drawing a region of interest around the organ/tissue to quantify the radiant efficiency {(photons/second/cm 2 /steradian)/(µW/cm 2 )}.

Statistical Analysis
Statistical analysis was performed using GraphPad Prism version 5.01 (GraphPad Software Inc., San Diego, CA, USA). The half-maximal inhibitory concentration (IC 50 ) values were estimated by the log (inhibitor) vs. the normalized response fitting routine. To compare the uptake values of the two probes studied in vivo, an unpaired t-test was used with the significance levels set at 5%. The results are represented as mean ± standard deviation (SD).

Chemistry and Radiochemistry
Synthesis of the NeoB analogs was performed following solid-phase peptide synthesis (SPPS) protocols, using a standard Fmoc strategy. Coupling of the commercially available amino acids and linkers containing a Fmoc or a Boc protecting group was carried out with conventional coupling reagents, HBTU and Oxyma Pure (Scheme 1). Cleavage of the peptides from the solid support was achieved via treatment with a solution containing HFIP and DCM. Briefly, 4-Amino-2, 6-dimethylheptane was coupled on the C-terminal histidine by using PyBOP under basic conditions. Subsequent removal of the protecting groups from the compounds was achieved using a cocktail of TFA/H 2 O/TIS. Next, the peptides were purified by a C18 Sep-Pak cartridge, using H 2 O/ACN. Finally, compounds 4a and 4b were obtained in 2.0 and 2.7% yields, respectively, after coupling of the TCO-NHS ester under basic conditions and the purification of the products by semi-preparative HPLC.
Synthesis of the compounds 9a, 9b, 9c and 9d (Scheme 2) was achieved following a similar synthetic approach. Briefly, the Fmoc-protected linkers (pADA and PEG 4 ) were successfully coupled to the peptide sequence, followed by the removal of the Fmoc-protecting group with piperidine. The subsequent coupling of the lysine residue and DOTA chelator provided intermediates 7a and 7b. Cleavage of the peptides from the solid support, the coupling of 4-amino-2, 6-dimethylheptane on histidine, deprotection of the remaining protecting groups, and purification of the crude compounds by a C18 Sep-Pak cartridge afforded compounds 8a and 8b. The final peptides were obtained by coupling TCO-NHS ester or TCO-PEG 4 NHS ester to the free amine on the side chain of the lysine residue. Compounds 9a, 9b, 9c and 9d were purified by semi-preparative HPLC and obtained in 1.2, 0.6, 0.7 and 0.7% yields, respectively.
Tz-sCy5 was obtained in 63% yield in a one-step synthesis under basic conditions [22]. With the six final NeoB analogs and Tz-sCy5 in hand, we prepared two mono-and four dual-modality probes (Figure 1) via an IEDDA click reaction. The reaction was performed under mild conditions at 37 • C. The final compounds 10, 11, 12, 13, 14 and 15 were obtained in 63, 46, 23, 29, 79 and 46% yields, respectively, after semi-preparative HPLC purification.
Labeling of the final dual-modality probes was performed with 111 InCl 3 using a mixture of sodium acetate and Gz/Asc as scavengers to prevent radiolysis. The labeling efficiency was monitored by iTLC and radio HPLC (

Binding Affinity Assays
[ 111 In]In-NeoB was displaced from the GRPR binding site in PC-3 cells by the newly synthesized NeoB analogs. The results presented in Table 2 indicate that the binding affinity of the probes is 4.5 to 24.6-fold (21.9-118.7 nM) lower than the binding affinity of the parent peptide NeoB (4.8 nM) ( Figure S1). The two dual-modality probes with the highest binding affinity (i.e., 12 and 15) were selected for pilot in vivo pharmacokinetic characterization.  (Figure 2A). The uptake was GRPR-specific, since reduced tumor uptake was observed when the GRPR was blocked using an excess of unlabeled NeoB.

Biodistribution Studies
The results of the ex vivo biodistribution of [ 111 In]In-12 and [ 111 In]In-15 in PC-3xenograft Balb/c nu/nu mice after SPECT/CT scanning are depicted in Figure 3 and Table  S1. In agreement with the in vivo measurements, high radioactivity uptake was observed in the tumor with both probes (8.47 ± 0.46%ID/g and 6.90 ± 0.81%ID/g for probe [ 111 In]In-12 and [ 111 In]In-15, respectively). In addition, the probes highly accumulated in the GRPRexpressing pancreas (16.55 ± 1.41%ID/g for [ 111 In]In-12; 9.72 ± 1.63%ID/g for [ 111 In]In-15). Both probes have similar tumor-to-muscle ratios, but the higher pancreas uptake of probe [ 111 In]In-12 resulted in a significantly lower tumor-to-pancreas ratio (p = 0.0131). In lowlevel GRPR-expressing organs, such as the GI tract and the prostate gland, moderate uptake was observed. Co-injection with an excess of unlabeled NeoB resulted in a strongly decreased uptake level of the probes in the GRPR-positive tumor and organs, indicating receptor specificity. Furthermore, a relatively high radioactivity uptake was seen in the lungs (6.51 ± 1.15%ID/g for [ 111 In]In-12; 6.77 ± 1.73%ID/g for [ 111 In]In-15). This uptake was not GRPR-specific since a 2-3-fold higher uptake in this organ was also observed in animals that were coinjected with an excess of unlabeled NeoB (16.14%ID/g and 18.95%ID/g for [ 111 In]In-12 and [ 111 In]In-15, respectively). Next to the lungs, unexpected high radioactivity levels were noticed in the liver (15.35 ± 0.43%ID/g for [ 111 In]In-12; 8.87 ± 0.65%ID/g for [ 111 In]In-15).

Ex Vivo Optical Imaging
A subset of organs and the tumor were also imaged by ex vivo fluorescence imaging, to confirm the localization of the dye (Figure 4). Here, the results showed that the fluorescence intensities of the organs generally followed the same trend as the radioactivity uptake. As can be deduced from Figure 4,

Discussion
Image-guided surgery of PCa can greatly aid surgeons to resect tumor tissues completely. Extensive research has been carried out on probes targeting the prostate-specific membrane antigen (PSMA), but it has been suggested that GRPR-targeted imaging probes may be of great value for PSMA-negative tumor lesions [23]. Moreover, the overexpression of GRPR typically occurs at an early stage of the disease, while PSMA is often associated with late-stage disease. Considering that image-guided surgery is particularly suitable for primary tumors, dual-modality probes targeting GRPR would be very attractive. NeoB exhibits a very high GRPR affinity and was introduced in the literature as a very promising peptide for GRPR-mediated radionuclide imaging and therapy [18,24]. Therefore, the chemical design of the dual-labeled probes is based on the amino acid sequence of the parent peptide NeoB. We herein described the synthesis of the new library of NeoB analogs with two linkers, the pADA and PEG 4 linkers. Those two linkers provide different physicochemical properties to the peptides, such as hydrophilicity, rigidity and spacing. In the chemical structure of compounds 4a and 4b, the original DOTA chelator was replaced by a TCO moiety, in order to preserve a chemical structure closer to that of the parent peptide. To add a TCO moiety to the original NeoB, a lysine residue was introduced between the peptide sequence and the DOTA chelator. This method has already been used by Li et al. and Zhang et al. for the insertion of a fluorescent dye into peptides [25,26]. Purification of the peptides by a C18 Sep-Pak cartridge before the coupling of TCO was implemented to remove the excess of TFA remaining from the removal of the protecting groups in the previous step. In fact, La-Venia and coworkers demonstrated that TCO is sensitive to acidic conditions and results in its change of conformation from TCO to CCO, the latter being less reactive toward Tz [27]. The final dual-labeled probes, containing a DOTA chelator and a fluorescent dye, were obtained via an IEDDA click reaction between the TCO coupled to the peptides and Tz coupled to the fluorescent dye, sCy5. The IEDDA reaction was chosen because of its fast kinetics, irreversibility, and stability of the generated product. The sCy5 dye was employed due to its strong fluorescence intensity and extinction coefficient, and excellent brightness [28,29]. Fluorescent dyes can show instability in aqueous solutions and a low pH. However, IEDDA provides the opportunity to synthesize the dual-labeled probes after obtaining the radiolabeled peptides containing the TCO moiety, preventing such instability from occurring. Another advantage of the click reaction is that it offers the possibility of synthesizing a large library of compounds in a single step.
The radiolabeling of the dual-labeled probes with 111 InCl 3 was successfully achieved with high RCYs. The negative LogD 7.4 values were obtained for all the radiolabeled peptides. Compared to their radiolabeled NeoB analog, [ 68 Ga]Ga-NeoB (LogD 7.4 = −0.88 ± 0.02) [30], the introduction of the fluorescent dye to the peptides increased their hydrophilicity; therefore, negative LogD 7.4 values were obtained for all the radiolabeled peptides. All radiolabeled compounds were stable in mouse serum, demonstrating their inertness toward peptidase digestion. However, the dual-labeled probes turned out to be less stable in PBS, demonstrating their sensitivity toward radiolysis [31].
The replacement of the DOTA chelator by a TCO moiety in compound 10 impaired the binding affinity of the peptide toward the receptor, probably due to a change in the conformation of the molecule. However, the introduction of a PEG 4 linker instead of a pADA linker did not substantially hamper the binding affinity of 11 toward GRPR, probably due to the flexibility offered by the linker. The introduction of two linkers in the four duallabeled probes, 12, 13, 14 and 15, induced changes in the conformation of the molecules, leading to a lower affinity of the compounds toward GRPR in comparison to the parent peptide, NeoB.
The results of the pilot in vivo study showed that probes [ 111 In]In-12 and [ 111 In]In-15 also possess good in vivo binding properties, depicted by a high tumor accumulation and a clear delineation of the xenograft on the SPECT scans. When compared to previous studies with NeoB [18,24], the fluorescent dye component of the probes noticeably influences the in vivo pharmacokinetic profile. The dual-modality probes have a prolonged blood circulation time, as high levels of radioactivity in the blood were observed at~3 h postinjection. Hence, it could conceivably be hypothesized that the optimal tumor uptake and tumor-to-background levels may not have been reached in the current study. The observed uptake levels in the excretory organs (i.e., the liver and kidneys), compared to the biodistribution profile of the parent peptide NeoB, can also be attributed to the longer blood circulation time. Further work is needed to see if this can be improved, for example, by performing biodistribution and imaging studies at later time points.
The elevated radioactivity levels observed in the liver might be indicative of both renal and hepatobiliary routes of elimination. The elimination through hepatobiliary excretion was unexpected because this is unusual for probes with a non-lipophilic character [32]. However, TCO-Tz conjugates have a tendency to accumulate in hepatobiliary organs [33,34]. Furthermore, the involvement of this excretion pathway can possibly be attributed to aggregate formation due to increased "stickiness" (i.e., cohesive forces) of the dual-modality probes. This process could also be the cause of the unexpected uptake in the lungs, as the presence of aggregates in this organ can lead to capillary blockages, especially when high peptide amounts are injected, as was the case for the blocked animals [35]. Additional studies are needed that focus on mass optimization and practical and chemical implementations to reduce stickiness.
For a high image contrast on the preoperative scan and good visual inspection of the surgical field intraoperatively, it is crucial to have a high tumor-to-background ratio. As a rule of thumb, a ratio of 2 is reported in the literature [9]. In our study, the tumor-to-muscle ratios of the probes were found to be 9.04 ± 1.74 and 6.27 ± 0.44 for probe [ 111 In]In-12 and [ 111 In]In-15, respectively. However, the tumor-to-blood ratio was not favorable. In addition, sufficient washout from the bladder (and thus, indirectly, the kidneys) is especially important in the image-guided surgery of PCa, because it is close to the surgical site [36]. A later imaging time point will most likely lead to reduced background signals in the blood and excretory organs.
To our knowledge, this is the second study, next to the investigations by Zhang et al. [26], about the in vivo evaluation of GRPR-targeted optical and nuclear dual-modality probes for PCa. The tumor uptake at~2 h reported in our study (8.47 ± 0.46%ID/g and 6.90 ± 0.81%ID/g for probe [ 111 In]In-12 and [ 111 In]In-15, respectively) was higher than the value that Zhang et al. obtained at~1 h (5.50 ± 1.03%ID/g) for their GRPR-directed dualmodality probe, based on the RM2 backbone. However, they conducted a biodistribution study following a PET scan 1 h after injection. At this time point, their nuclear scan did benefit from a low background, due to a higher tumor-to-blood ratio compared to our designed probes, again arguing for optimization of the imaging time point in the future.
Further increasing the signal specificity is important for accurate intraoperative delineation of the target region and, more specifically, to distinguish between normal and tumor tissue [37]. Here, the specificity of our probes for the GRPR as a tumor target was demonstrated by a pilot blocking study using a single animal. Co-injection of an excess of unlabeled NeoB led to a decreased uptake in the tumor and the GRPR-expressing organs, such as the pancreas [38]. A reduction in activity levels in non-GRPR-expressing organs, such as the kidneys, on blocking might be due to a lower amount of activity injected in those animals. Despite the use of kolliphor as a surfactant in the solvent for the injections, this measure was not enough to compensate for the cohesive and adhesive properties of NeoB and the probes [18].
Another important finding was that the ex vivo fluorescence imaging confirmed the co-localization of the fluorescent and radioactive signal. This nicely illustrates the benefit of using a dual-modality probe with the same pharmacokinetics for 2 different purposes: pre-and intraoperative guidance. It is difficult to compare our measurements to previous studies with fluorescent agents targeting the GRPR because there is a potential for bias from the injected mass [39][40][41][42][43]. The injected amount can influence uptake, especially when receptor saturation levels are not yet reached. Nuclear imaging techniques have a slightly higher sensitivity than optical imaging. This means that in general, nmol amounts must be administered to allow fluorescence detection, while pmol amounts are often sufficient for nuclear detection. A further study with more focus on the optimal mass and specific labeling activity for both modalities is therefore suggested.
Since our study accommodated a pilot in vivo evaluation, future work should include a late-uptake analysis for all probes with a larger sample size for the blocked groups. Our study is supported by quantitative data obtained using various methods. Due to the image resolution and the fact that regions were drawn manually, the uptake values that were quantified by volume and count measurements from the SPECT/CT scans are less accurate than those obtained from the ex vivo biodistribution study. However, the ratios between organs calculated using both methods correlate well. Quantification of ex vivo optical data suffers from light attenuation. The differences between the organs relative to each other are therefore smaller overall, as is the difference between the two probes. The differences are therefore more evident from the radioactivity uptake levels measured in the ex vivo biodistribution study.
This study demonstrated the successful development and initial characterization of four promising dual-modality probes for the preoperative imaging and image-guided surgery of GRPR-positive PCa. Despite its exploratory nature, this study provides valuable insights into the influence of the incorporation of the sCy5 dye into the radiotracer NeoB on its binding affinity and pharmacokinetic properties. Although uptake was seen in the liver, lungs and pancreas, their location is not near the prostate and will therefore not interfere with the image-guided surgery. The prolonged blood circulation and high renal uptake require further evaluation of the optimal timing for imaging. Moreover, further in vivo preclinical evaluation with all four dual-modality probes will be performed to select the best probe for clinical translation. Overall, this study reinforces the idea that multimodal probes have very interesting properties to advance the field of image-guided surgery.