64Cu-PSMA-BCH: A New Radiotracer for Delayed PET Imaging of Prostate Cancer

Teli Liu Peking University Cancer Hospital: Beijing Cancer Hospital Chen Liu Peking University Cancer Hospital: Beijing Cancer Hospital Zhongyi Zhang Peking University Cancer Hospital: Beijing Cancer Hospital Ning Zhang Peking University Cancer Hospital: Beijing Cancer Hospital Xiaoyi Guo Peking University Cancer Hospital: Beijing Cancer Hospital Lei Xia Peking University Cancer Hospital: Beijing Cancer Hospital Jinquan Jiang Peking University Cancer Hospital: Beijing Cancer Hospital Qing Xie Peking University Cancer Hospital: Beijing Cancer Hospital Kun Yan Peking University Cancer Hospital: Beijing Cancer Hospital Steven P. Rowe Johns Hopkins Medical Institutions: Johns Hopkins Medicine Hua Zhu Peking University Cancer Hospital: Beijing Cancer Hospital Zhi Yang (  pekyz@163.com ) Peking University Cancer Hospital &amp; Institute https://orcid.org/0000-0003-2084-5193


Introduction
Prostate cancer (PCa) is common among men. PSMA PET/CT has shown high sensitivity and speci city in identifying sites of PCa, and some studies have demonstrated that PSMA-based PET/CT and PET/MR can give accurate location of tumor lesions in primary PCa superior to MRI [1,2], which offers the possibility that PSMA PET images prior to biopsy giving a simpler criteria for targeted biopsy. Our group had tried 68 Ga-PSMA PET-ultrasound fusion targeted biopsies, which improved the detection rate of PCa and decrease the possibility of repeated biopsy [3]. Unfortunately, limited by the short half-life of 68 Ga, the images were obtained at 1 h post injection (p.i.) with high radioactivity in bladder, which lead the low contrast between the PCa and background.
Radiopharmaceuticals labeled with radionuclides having longer half-life can be used in delayed imaging, potentially leading to higher tumor-to-background contrast. Because of the internalization of PSMA radiotracers, the contrast in tumor lesions increases at longer time points after injection, with the radioactivity in non-target organs signi cantly decreasing at later time points. Among such radionuclides, copper-64 ( 64 Cu) was deemed to be a good choice due to its moderate half-life (12.7 h) and high resolution [4]. 64 Cu labelled PSMA-targeting probes, 64 Cu-PSMA-617 [5][6][7][8][9], 64 Cu-DOTA-scFv-anti-PSMA [10] and 64 Cu-CA003 [11] have been reported. 64 Cu-PSMA-617 showed high diagnostic accuracy for primary LN staging [5,6]. But the preclinical and in-man studies of 64 Cu-PSMA-617 demonstrated high uptake in liver and the in vivo stability showed most of 64 Cu-PSMA-617 has been dissociated within 2 hours [9,8,7]. This was possibly due to endogenous proteins involved in copper metabolism transchelating 64 Cu from DOTA [12][13][14]. As human copper transporter 1 is expressed in most tumors, 64 CuCl 2 itself can be used for the diagnosis of PCa [15]. The dissociation of 64 Cu from 64 Cu-PSMA-617 in vivo not only may lead extra radiation exposure of liver but also could affect the detection of some liver metastases. Compared with DOTA, triaza macrocyclic compounds, such as NOTA and its derivatives, have been reported to chelate 64 Cu with higher in vivo stability [16,17]. In order to obtain high quality images for PCa imaging and give delayed images for improved detection of sites of disease, we prepared a 64 Cu labeled radiotracer, 64 Cu-PSMA-BCH, with a NOTA-conjugated precursor. The in vitro and in vivo studies were performed to evaluate the stability, PSMA speci city, radiation safety and tumor targeting of 64 Cu-PSMA-BCH.

Materials And Methods
General 64 CuCl 2 was obtained from the department of Nuclear Medicine, Peking University of Cancer Hospital. All chemicals, reagents and solvents were purchased commercially without further puri cation. Sep-Pak C18-Light cartridges were purchased from Waters. The product was analyzed by reversed-phase high performance liquid chromatography (RP-HPLC; Eclipse Plus C18, 4.5×250 mm, 5μm; Agilent) performed using a linear A-B gradient (15%-60% of B in 15 min) with a ow of 1 mL/min. Solvents were 0.1% aqueous TFA (A) and 0.1% TFA in acetonitrile (B). The HPLC system was equipped with UV and γ detectors. UV absorbance was measured at 220 nm. Micro-PET was performed on Super Argus PET (Sedecal, Spain). PET/CT scans were obtained on a Biograph mCT Flow 64 scanner (Siemens, Erlangen, Germany) with unenhanced low-dose CT. PSMA (+) 22Rv1 and PSMA (-) PC-3 cell lines were obtained from China Cell Line Resource.
Cell culture and animal models Human prostate cancer cell lines 22Rv1 and PC-3 were cultured and the tumor models were established as previously reported [18]. All animal experiments were conducted in accordance with the guidelines approved by Peking University Cancer Hospital Animal Care and Use Committee.
After cooling to room temperature, 64 Cu-PSMA-BCH was puri ed by C18 Light Cartridge and obtained as shown in supporting information. 64 Cu-PSMA-BCH was diluted with saline for further studies and analyzed by radio-HPLC for radiochemical purity, checked for pH value and sterility tested.

Partition coe cient
The partition coe cient of 64 Cu-PSMA-BCH was studied in the PBS (0.1 M, pH 7.4)-octanol system (supporting information) and the value was presented as log P±SD, P was calculated as below: P= (average of CPM in octanol / average of CPM in PBS).

In vitro stability
The in vitro stability of 64 Cu-PSMA-BCH (3.7 MBq) was tested in solution of saline and 5% HSA at 37 ℃ till 36 h and analyzed by radio-HPLC (see supporting information).

Biodistribution
Biodistribution of 64 Cu-PSMA-BCH in normal BALB/c male mice was performed (see supporting information) and results were expressed as the percent of injected dose per gram (ID%/g).
Micro-PET imaging, biopsy and histology study in tumor model mouse 18.7 MBq of 64 Cu-PSMA-BCH were injected into mice bearing 22Rv1 and PC-3 vial a tail vein. At 3, 12 and 20 h p.i., the mice were anaesthetized and performed micro-PET imaging on Super Argus PET (Sedecal, Spain) acquired with 80 mm diameter Transaxial FOV, OSEM 3D reconstruction algorithms with attenuation and random corrections. Finally, the images were displayed by MMWKS Super Argus. The milicounts/sec values of ROI (regions of interest) over tumor, kidneys and liver were collected.
After imaging at 20 h p.i., the mouse was anaesthetized with 3% (v/v) and sacri ced. Then puncture in tumor lesions was performed. The samples were immobilized in 10% neutral formaldehyde xative and then performed radioautography to further investigate the possibility of methodology.

Radioautography and immunohistochemistry
The samples were stored on slides with 10% neutral formaldehyde xative and exposed on a phosphorus plate (Perkin-Elmer, USA) for 12 h. The plate was scanned using a phosphor imaging system (Cyclone, Packard) to obtain the images. The slides were prepared and analyzed for PSMA expression by immunochemistry as described previously [7].

PET/CT imaging and analysis
With the approval of Ethics Committee of Beijing Cancer Hospital (No. 2017KT97), four patients (age 77.25±5.36, range 68-81, PSA 35.40±31.05 ng/mL, range 8.5-86.15; Gleason score 8.25±0.43, range 8-9) with suspected prostate cancer, who were clinically appropriate for biopsies, were included in this study (Table 1). Three patients without metastasis were performed whole-body PET/CT scans at 1 and 24 h p.i. and pelvic cavity scans at 4h p.i.. One patient with multiple bone and lymph nodes metastases performed whole-body PET/CT scans at 1, 4 and 24 h p.i.. Imaging was performed and all images were read by 2 experienced nuclear medicine specialists, the SUVmean, radioactivity concentration (Bq/mm 3 ) and volume (mm 3 ) of each organ and SUVmax of tumor lesions were obtained as literature.as previously reported [18].

Absorbed dosimetry
The radioactivity concentration (Bq/mm 3 ) and volume (mm 3 ) were used for calculating ID% of each organ. And the data was used for estimating human organ radiation dosimetry. OLINDA/EXM 2.0 software (Hermes Medical Solution, Sweden) was used with adult male model without special kinetics.

PET-ultrasound guided targeted biopsy
After imaging at 24 h p.i., the images were reconstructed and analyzed, then the patients were performed classic ultrasound-guided biopsies with visual fusion of PET images according to PROMISE criteria (version 1.0) as reported [19,3].

Results
Radiochemistry and quality control 64 Cu-PSMA-BCH was prepared with the yield over 95% and the radiochemical purity over 99% analyzed by radio-HPLC. The retention time of 64 Cu-PSMA-BCH and 64 Cu 2+ were 9.34 min and 3.46 min, respectively.
The speci c activity of 64 Cu-PSMA-BCH was 14.3±1.91 GBq/μmol. After diluted with saline, the quality control of 64 Cu-PSMA-BCH was performed, and the result was shown in Table 2.

Partition coe cient and in vitro stability
The log P value of 64 Cu-PSMA-BCH was calculated as -2.46±0.11, indicating 64 Cu-PSMA-BCH was hydrophilic.

Micro-PET imaging
Micro-PET imaging was performed on a male BALB/c nude mouse bearing 22Rv1 and PC-3 tumors. 22Rv1 tumor, kidneys, bladder and liver were clearly visualized, while the PC-3 tumors cannot be observed ( Figure 2a). The radioactivity accumulation in liver and 22Rv1 tumor was lightly increased within 20 h, while the uptake in kidney was decreased and the kidneys were nearly invisible at 12 h and 20 h p.i.. At 12 h p.i., only 22Rv1 tumor and liver were observed. The ROIs of kidney, liver, 22Rv1 and PC-3 tumors were measured. The 22Rv1/PC-3 tumors and 22Rv1/kidney ratios were increased with time (from 3.92 to 20.30 and from 0.2 to 2.99, respectively) between 3 h and 12 h p.i. (Figure 2b). Compared with reported data of 68 Ga-PSMA-617, the radioactivity of 64 Cu-PSMA-BCH in liver is lower, which was coincidence with the biodistribution result in normal BALB/c mice.

PET/CT imaging
Four patients who were suspected of having PCa were included in this study and conducted 64 Cu-PSMA-BCH PET/CT imaging. At 1 h post injection, bladder, kidneys, lacrimal glands, parotid glands, submandibular glands, liver, proximal small bowel and tumor lesion were clearly observed, which was coincidence with other PSMA radioligands (Figure 3a, Table S2). The uptake in liver was increased and the SUVmean was increased from 4.38 to 5.31 between 1 h and 24 h p.i.. Though kidneys, lacrimal glands, parotid glands and submandibular glands were visible at 24 h p.i., the SUVmean values were decreased. The SUVmax of tumor lesions in three PCa patients without metastases were increased between 1 h and 24 h p.i. (Figure 3b), as well as the tumor-to-background contrasts ( Figure S2).
In this study, one PCa patient with multiple metastases were included. At 4 h and 24 h p.i., more tumor lesions were observed and the SUVmax values were higher than that at 1 h p.i. (Figure 4).

Radiation dosimetry estimates
With the biodistribution data of 64 Cu-PSMA-BCH patients with primary prostate cancer, the human organ radiation dosimetry of 64 Cu-PSMA-BCH was calculated using OLINDA/EXM 2.0 software package. The estimated dosimetry was shown in Table 3. Gallbladder wall was the most critical organ (2.04E+00 mGy/MBq), followed by liver (1.45E-02 mGy/MBq) and kidneys (9.47E-03 mGy/MBq). The effective dose was 0.0292 mSv/MBq, which means the effective dose of a patient was 4.3 mSv when injected with 4 mCi of 64 Cu-PSMA-BCH. The whole-body effective dose of 18 F-FDG that International Commission on Radiological Protection's (ICRP) is 0.019 mSv/MBq, equating to an effective dose for a 70 kg adult man (4.9 mSv). The effective dose of 64 Cu-PSMA-BCH therefore is comparable to 18 F-FDG for a male.

PET-ultrasound guided targeted biopsy and histology
After imaging at 4 h p.i., patients with suspected primary prostate cancer underwent PSMA-targeted PETultrasound fusion guided biopsies. Figure 5a showed the 64 Cu-PSMA-BCH PET/CT images of one patient (age 80-y, PSA 31.6 ng/mL) at 24 h p.i.. First two targeted scores were kept in 10% neutral formaldehyde xative and exposed to a screen (Figure 5c). The radioautography images clearly displayed the morphology of corresponding tissues (Figure 5d). The HE (hematoxylin-eosin) and PSMA immunohistochemical staining of targeted biopsies showed Gleason Score of 4+4 with high PSMA expression (100%) in enchylema (Figures 5e and 5f).

Discussion
PCa is widely studied as its high incidence and mortality. PSMA is overexpressed in PCa with high speci city, and has become an important target for the diagnosis and therapy of PCa. Multiple radiotracers targeting PSMA have been developed and PSMA PET/CT showed high sensitivity and speci city for the detection of PCa. In this study, we prepared a tracer, 64 Cu-PSMA-BCH, labeled with copper-64, a radionuclide with longer half-life and allows high resolution imaging. Among the reported 64 Cu labeled PSMA tracers, most of them were chelating with 64 Cu by bifunctional conjugator DOTA. 64 Cu-DOTA complexes was considered to have relative low in vivo stability, especially under acidic condition, which commonly lead to 64 Cu ion transchelation to serum protein. [20] Compared with DOTA, NOTA has higher a nity to copper-64 and the resulting complexes are more stable in vivo with less chance of transchelation by copper metabolism. [16] Compared with 64 Cu-PSMA-617, a DOTA-conjugated tracer, 64 Cu-PSMA-BCH was more stable in vivo with lower radioactivity accumulation in liver. While the uptake of 64 Cu-PSMA-BCH in kidneys is higher than other organs, which was different from the 64 Cu labeled DOTA-conjugated complexes. Indicating the stability of 64 Cu-PSMA-BCH in vivo. 64 Cu-PSMA-BCH accumulated in kidneys and excreted through kidneys into urine with high bladder uptake within the rst few hours. It can speci cally accumulate in PSMA(+) 22Rv1 cells and tumor, which can be blocked by ZJ-43, a potent PSMA inhibitor commonly applied for blocking study [21], while Some studies showed that delayed PSMA PET/CT imaging can increase the uptake in tumor and may detect more tumor lesions with small volume or low PSMA expression [22][23][24]. Because of the long halflife of 64 Cu, increased uptake and contrast of 64 Cu-PSMA-BCH, delayed PET/CT imaging were allowed. 64 Cu-PSMA-BCH-based delayed imaging holds the potential for detecting more small tumor lesions or some occult recurrent metastasis.
Not only for detecting tumors, delayed PSMA PET/CT imaging of 64 Cu-PSMA-BCH may have more application in clinic. Our group had reported a study that performed PSMA PET/CT prior to biopsy and then PSMA PET-ultrasound fusion targeted biopsies, which improved the detection rate and decrease possibility of repeated biopsies [3]. As shown in Figures 4 and 6