Comparison of the radiolabeled PSMA-inhibitor 111In-PSMA-617 and the radiolabeled GRP-R antagonist 111In-RM2 in primary prostate cancer samples

Purpose Prostate-specific membrane antigen (PSMA) and gastrin-releasing peptide receptor (GRP-R) are expressed in prostate cancer and can be targeted with radiolabeled inhibitors and antagonists. Their performances for the initial characterization of prostatic tumors have been barely evaluated but never compared. We aimed to gather comparative preclinical data of the role of PSMA and GRP-R targeting in prostate cancer. Procedures We retrospectively studied 20 frozen prostatectomy samples with various metastatic risks of the D’Amico classification. Tissue samples were investigated by tissular microimaging using the radiolabeled PSMA inhibitor 111In-PSMA-617 and the radiolabeled GRP-R antagonist 111In-RM2. Bindings of the two radiopharmaceuticals were compared to histology and clinico-biological data (Gleason score, PSA values, metastatic risks). Results Binding of 111In-PSMA-617 was high whatever the metastatic risk (p = 0.665), Gleason score (p = 0.555), or PSA value (p = 0.404) while 111In-RM2 exhibited a significantly higher binding in the low metastatic risk group (p = 0.046), in the low PSA value group (p = 0.001), and in samples with Gleason 6 score (p = 0.006). Conclusion PSMA and GRP-R based imaging might have complementary performances for the initial characterization of prostatic tumors. Prospective clinical studies comparing the two tracers in this setting are needed.


Introduction
Prostate cancer is the most common cancer in men and the third cause of cancer deaths [1]. It is also a multifocal disease as cancerous cells may arise from different locations within the prostatic gland. Thus, prostate cancer is a combination of different cancerous cells with their own metastatic risks. Prostate cancer classification, prognosis, and management are today based on the two major cell populations (Gleason score). Beside primary staging which includes multi parametric pelvic magnetic resonance imaging (mpMRI), thoraco-abdomino-pelvic computed tomography (CT), and bone scintigraphy, only 18 F-Choline positron emission tomography/computed tomography (PET/CT) may be proposed to some patients with high metastatic risk but it has a low accuracy for detection of primary prostate cancer [2].
Attractive targets for a more specific and sensitive imaging of primary prostate cancer are the prostate-specific membrane antigen (PSMA) and the gastrin-releasing peptide receptor (GRP-R). They can be effectively targeted with radiolabeled inhibitors [3] and antagonists [4], respectively.
Prostate-specific membrane antigen (PSMA) is a type 2 glycoprotein expressed in secretory cells of prostatic epithelium. Several radiolabeled PSMA inhibitors have been developed for imaging ( 68 Ga-PSMA-11, 68 Ga-PSMA-617, score (GS) and PSA level [6] indicating a role for this imaging procedure in high-risk prostate cancer.
The gastrin-releasing peptide receptor (GRP-R) is a G-protein-coupled receptor of the bombesin receptor family [7] which can be targeted with radiolabeled antagonists such as 68 Ga-RM2 [8], 68 Ga-NeoBOMB1 [9], or 68 Ga-RM26 [10] for PET imaging. Contrarily to PSMA, GRP-R is overexpressed in low-risk prostate cancers (low Gleason score, low PSA value, and low tumor size) [11,12]. A study of initial staging of prostate cancer on 14 patients observed a detection rate of 83%, a sensitivity of 89%, and a specificity of 81% [8].
Although few pilot clinical studies targeting PSMA or GRP-R for initial staging of prostate cancer suggest a complementary role of these imaging procedures, there have never been compared in the same patients. Therefore, in this preclinical work, we aimed to compare PSMA and GRP-R expression on the same histological samples of prostate tumors using radiolabeled probes.

Patient characteristics
Twenty frozen samples of prostate cancer were available from the Department of Pathology of University Hospital of Toulouse, France. Patient samples were obtained after informed consent in accordance with the Declaration of Helsinki and stored at the "CRB Cancer des Hôpitaux de Toulouse (BB-0033-00014)" collection. According to the French law, CRB Cancer collection has been declared to the Ministry of Higher Education and Research (DC-2008-463) and obtained a transfer agreement (AC-2013-1955) after approbation by ethical committees (Conseil Scientifique du Centre de Ressources Biologiques). Clinical and biological annotations of the samples have been declared to CNIL (Comité National Informatique et Libertés). Sample characteristics' are presented in Table 1. No patient had received neoadjuvant hormone therapy or chemotherapy. For each case, five adjacent sections were used: one for Hematoxylin-Eosin-Saffron (HES) staining and four for high-resolution microimaging (one section per radiopharmaceutical for total binding and another one for non-specific binding). An experienced pathologist manually drew tumoral areas on the HES-stained section. All patients were classified according to their metastatic risk, following the D' Amico classification [13], using clinical and biochemical criteria including tumoral size, PSA value, and Gleason score.

High-resolution microimaging Binding assay
Protocol edited by Reubi and co-workers for binding assays was strictly adhered [14]. Frozen samples were kept at − 80°C. Three days before handling, samples were placed at − 20°C. The day of the experiment, samples were pre-incubated for 10 min at 37°C in Tris-HCl buffer at pH 7.4. A hydrophobic pen was used to surround the sample. In-RM2 (IC 50 = 9.3 ± 3.3 nM on the GRP-R) [15] or 10 nM (0.03-0.06 MBq) of 111 In-PSMA-617 (Kd = 5.4 ± 0.8 nM on the PSMA) [16] in Tris-HCl buffer at pH 8.2, 1% of BSA (Sigma®A2153), 40 μg/mL of bacitracin (Sigma®11,702), and 10 nM of MgCl 2 (Sigma®M8266) was applied (ethanol content for 111 In-RM2 was 0.003 ± 0.002% and 0.0009 ± 0.0008% for 111 In-PSMA-617). For non-specific binding, 1 μM of nat Ga-labeled RM2 or nat-Ga-PSMA-617 was added to determine non-specific binding ( nat In-RM2 and nat In-PSMA-617 were not available to us, we used nat Ga-RM2 and nat Ga-PSMA-617 which also bind with high affinity to GRP-R and PSMA, respectively). IC 50 of Ga-RM2 for GRP-R is below 0.1 nM and K i of Ga-PSMA-617 for PSMA is 6.40 ± 1.02 nM [17,18]. Samples were incubated at 37°C for 2 h. Afterward, samples were rinsed five times for 8 min in cold Tris-HCl buffer at pH 8.2 with 0.25% of BSA, two times for 8 min in cold Tris-HCl buffer at pH 8.2 without BSA and finally two times for 5 min in distilled water.

Tissular microimaging
Beta Imager-2000 (Biospace Lab) device was used to image and quantify radioactivity in the sample. Then, a Micro Imager (Biospace Lab) was used to obtain high-resolution images (radioactive and optical). Acquisition duration was about 20 h for the Beta Imager 2000 (4 × 10 6 counts) and 15 h for the Micro Imager.

Data analysis
Imaging analysis was performed as previously described [19].

Statistical analysis
Data, presented as the mean ± standard deviation (SD), were compared using non-parametric t test (Wilcoxon test) and non-parametric one-way ANOVA (Kruskal-Wallis test). Statistical analyses were performed using GraphPad software (v 6.01, San Diego, USA). p values < 0.05 were considered statistically significant.

High-resolution microimaging (HRMI) Qualitative analysis
Both radiopharmaceuticals were easily detectable, without excessive noise. As shown in Fig. 1, on samples from low metastatic risk tumors, discrimination between tumoral tissues and normal tissues was good with both 111 In-RM2 and 111 In-PSMA-617. On high metastatic risk samples, signal-to-noise ratio was higher with 111 In-PSMA-617 (Fig. 2).
All results are reported in Table 2 and resumed in Fig. 3.

Discussion
Several radiopharmaceuticals have been developed for accurate staging of prostate cancer. 11 C-Acetate, marking lipid metabolism, cannot reliably distinguish benign prostatic hyperplasia from prostate tumors [20]. Moreover, the radiolabeled amino-acid 18 F-FABC ( 18 F-Flucicovine) did not show good diagnostic performances for characterization of primary lesions [21]. Finally, 11 C/ 18 F-Choline, also marking lipid metabolism, showed lower sensitivity than mpMRI for primary detection of prostate cancer [22]. Thus, improvements in current molecular imaging of prostate cancer appear necessary for accurate characterization of primary prostate tumors. PSMA and GRP-R are differently overexpressed in prostate cancer, which raises hopes for molecular imaging of tumor lesions in the prostate gland. Few studies have investigated PSMA and GRP-R-based PET/CT imaging at initial staging, before radical prostatectomy. In a recent prospective study performed by Liu et al. on 50 newly diagnosed patients with prostate cancer candidates for radical prostatectomy, 68 Ga-PSMA-617 PET/ CT was able to detect 95% of primary tumors. The detection rate was similar to that of 68 Ga-PSMA-11 PET/ CT [23]. Another excellent work was performed by Touijer et al., in which authors prospectively investigate 68 Ga-RM2 PET/CT in 16 patients before radical prostatectomy. The performances of 68 Ga-RM2 PET/CT imaging did not significantly differ compared to mpMRI in terms of sensitivity, specificity, and accuracy. Moreover, 68 Ga-RM2 binding did not correlate with Gleason score [24]. To date, no intra-patient comparison of PSMA and GRP-R targeting at initial staging was reported. Therefore, the objective of this work was to investigate and compare in vitro the potential role of 111 In-PSMA-617 and 111 In-RM2 at the initial staging of prostate cancer.
Qualitative comparison of 111 In-PSMA-617 and 111 In-RM2 on our primary prostate cancer samples showed good detectability of both radiopharmaceuticals which is an essential quality for contrasted images in vivo. Then, we quantitatively compared 111 In-PSMA-617 and 111 In-RM2. When considering all metastatic risk groups together, 111 In-PSMA-617 binding was significantly  In-PSMA-617 is high in all samples with no significant differences between groups. Non-parametric one-way ANOVA (Kruskal-Wallis test). p < 0.05 was considered significant higher than 111 In-RM2 (p < 0.001). 111 In-PSMA-617 binding was high and no differences were seen according to Gleason score or pre-operative PSA values (Table 2). This high 111 In-PSMA-617 binding, whatever the characteristics of the sample, clearly reflects the ability of PSMA imaging to detect most prostate cancers [23] whatever their grade or risk [25]. Moreover, this high signal level may also be explained by a lower binding of 111 In-PSMA-617 to normal tissues (p = 0.0161), resulting in a higher TNR for 111 In-PSMA-617. An interesting result of our work is that 111 In-RM2 was able to discriminate low metastatic risk samples (p = 0.0046) and therefore low Gleason score samples (p = 0.0061) and samples with low PSA value (p = 0.0012). These results agree with the known high GRP-R expression in low-grade prostate cancer [11]. However, data from GRP-R immunohistochemistry and our results did not necessarily translate into parallel findings at patient imaging in pilot studies. For instance, the only two GRP-R imaging study, performed at the initial staging of prostate cancer, did not show any correlation (positive or negative) between SUV max on PET/CT and Gleason scores [10,24]. However, only 1/16 prostate cancers in the study by Touijer et al. and 2/17 in the study by Zhang et al were Gleason 6 [10,24]. Larger clinical studies are needed to elucidate the potential offered by GRP-R targeting at the initial staging of prostate cancer. Comparison with PSMA would also be helpful.
In intermediate and high-risk samples, 111 In-PSMA-617 binding was substantially higher than 111 In-RM2 binding, in agreement with the known expression of GRP-R which decreases in higher Gleason scores [11]. 111 In-PSMA-617 binding was also higher than 111 In-RM2 binding in patients for whom pre-surgical PSA value was > 10 ng/mL. These results agree with the known efficacy of PSMA imaging of intra-prostatic tumors in newly diagnosed high-risk prostate cancer patients [6].
Our results may have future clinical value. Prostate cancer patients with low metastatic risk are today not eligible for radical treatments anymore but rather to active surveillance or focal treatments [26]. Moreover, upgrading in Gleason score between biopsies and radical prostatectomy occurs in about 30% of patients [27]. Hence, an imaging procedure capable to discriminate "true" low metastatic risks would be required to schedule focal treatments in this group of patients and not under-treat patients that would in fact be at higher risk. Results of this work indicate that GRP-R targeting is the only imaging procedure amenable to discriminate low metastatic risk from higher risks. We suggest that GRP-R-based imaging may be first proposed in low metastatic risk patients for biopsy guidance and follow-up of active surveillance. Absence or low uptake at GRP-R imaging would suggest a disease of higher risk (or no disease).
In newly diagnosed prostate cancer patients with intermediate and/or high metastatic risk, PSMA-based imaging is obviously the imaging procedure of choice to characterize intra-prostatic tumors. PSMA-based guided biopsies, staging or radiation treatment planning is being explored in prospective studies [28][29][30].
We have translated GRP-R and PSMA-based imaging in our center. Patient candidates for radical prostatectomy benefit from sequential 68 Ga-PSMA-617 PET/CT and 68 Ga-RM2 PET/CT. Preliminary results would support our in vitro data presented in this article.
Limitation of our study is obviously the limited number of samples studied. Moreover, the clinical outcome of patient for whom samples have been used in this study is not known, and we could not assess the prognostic value of GRP-R-and/or PSMA-based imaging and therefore their role in the follow-up of patients.

Conclusion
In this work, we have compared GRP-R and PSMA expression in vitro on primary prostate cancer samples by means of 111 In-RM2 and 111 In-PSMA-617. Our results show that GRP-R and PSMA-based imaging may have a complimentary role to fully characterize prostate cancer disease, GRP-R being targeted in low metastatic risk patients while PSMA could be a valuable target in higher risks. Future prospective studies are warranted to confirm these data. Funding This study was funded by "La ligue contre le cancer de Gironde" and was achieved within the context of the Laboratory of Excellence TRAIL ANR-10-LABX-57.

Availability of data and materials
The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.
Authors' contributions RS performed experiments, analyzed the data, and wrote the manuscript. HDCG supervised the study and reviewed and approved the final manuscript. MY analyzed the biological samples and approved the final manuscript. MLQR collected the samples and obtained consents from the patients. NB and DV performed experiments and approved the final manuscript. EH supervised the study and reviewed and approved the final manuscript. PF was a recipient of the funding and approved the final manuscript. CM supervised the study, participated in experiments and funding acquisition, analyzed the data, and reviewed and approved the final manuscript. All authors read and approved the final manuscript.