What is the Best Radionuclide for Immuno-PET of Multiple Myeloma? A Comparison Study Between 89Zr- and 64Cu-Labeled Anti-CD138 in a Preclinical Syngeneic Model

Although positron emission tomography (PET) imaging with 18-Fluorodeoxyglucose (18F-FDG) is a promising technique in multiple myeloma (MM), the development of other radiopharmaceuticals seems relevant. CD138 is currently used as a standard marker for the identification of myeloma cells and could be used in phenotype tumor imaging. In this study, we used an anti-CD138 murine antibody (9E7.4) radiolabeled with copper-64 (64Cu) or zirconium-89 (89Zr) and compared them in a syngeneic mouse model to select the optimal tracers for MM PET imaging. Then, 9E7.4 was conjugated to TE2A-benzyl isothiocyanate (TE2A) and desferrioxamine (DFO) chelators for 64Cu and 89Zr labeling, respectively. 64Cu-TE2A-9E7.4 and 89Zr-DFO-9E7.4 antibodies were evaluated by PET imaging and biodistribution studies in C57BL/KaLwRij mice bearing either 5T33-MM subcutaneous tumors or bone lesions and were compared to 18F-FDG-PET imaging. In biodistribution and PET studies, 64Cu-TE2A-9E7.4 and 89Zr-DFO-9E7.4 displayed comparable good tumor uptake of subcutaneous tumors. On the bone lesions, PET imaging with 64Cu-TE2A-9E7.4 and 89Zr-DFO-9E7.4 showed higher uptake than with 18F-FDG-PET. Comparison of both 9E7.4 conjugates revealed higher nonspecific bone uptakes of 89Zr-DFO-9E7.4 than 64Cu-TE2A-9E7.4. Because of free 89Zr’s tropism for bone when using 89Zr-anti-CD138, 64Cu-anti-CD138 antibody had the most optimal tumor-to-nontarget tissue ratios for translation into humans as a specific new imaging radiopharmaceutical agent in MM.


Introduction
Over the last 30 years, major advances have been made with regard to the management of multiple myeloma (MM) [1,2]. These improvements have occurred along with our evolving understanding of this malignancy [3]. Multi-clonal heterogeneity of MM still remains one of the main challenges in developing effective therapeutic strategies [4]. Immunologic approaches represent an attractive solution to address this issue for treatment [5], however also for imaging [6] in the context of theranostic approaches. Indeed, the combination of positron emission tomography (PET) with monoclonal antibodies (mAbs) enables the realisation of a specific imaging called immuno-PET [7]. Among the interesting targets, CD138 or syndecan-1 is a cell surface proteoglycan that plays a critical role in the interaction between MM cells and their microenvironment [8,9]. This antigen is currently used as a standard marker for the identification and purification of MM cells in daily practice. Anti-CD138 immuno-PET thus has the potential to improve MM imaging, especially regarding lesions with low metabolic activity [10,11]. Moreover, it could also be considered as a companion agent for currently developed therapies targeting CD138 [12,13]. In the past several years, our group has also proven that radioimmunotherapy (RIT) combining anti-CD138 mAb and alpha-emitters radionuclides is effective in an immuno-competent preclinical MM model and is feasible in humans [14,15].
For immuno-PET, the choice of the radionuclides remains a fundamental question [7]. Combining appropriate half-lives for mAbs biodistribution and favorable emission properties for imaging, Copper-64 ( 64 Cu) and Zirconium-89 ( 89 Zr) have monopolized much of the researchers' attention during the last decade with an advantage for the second one in terms of number of studies [16,17]. However, reported release of 89 Zr from the imaging probe may represent a drawback for bone lesions imaging [18] and therefore for MM assessment. In this work, we report the preclinical evaluation of a novel PET imaging agent based on the 89 Zr-labeled anti-mouse syndecan-1 mAb (9E7.4, IgG2a κ isotype) [19] in a subcutaneous model and a bone marrow disseminated MM model using desferrioxamine B (DFO) as chelator. This is compared to 18 F-FDG-PET and bioluminescence imaging as gold standards and to 89 Zr-oxalate imaging as a control of potential 89 Zr release by the chelator agent. Furthermore, given our previous experience using 64 Cu [11] and to establish the optimal radiolabeled 9E7.4 mAb for immuno-PET, biodistribution and PET imaging in vivo of 89 Zr-and 64 Cu-mAb conjugates have been compared with emphasis on bone uptake.

Results
To evaluate and select the optimal radiolabeled 9E7.4 mAb for immuno-PET of MM lesions in bones, we have generated two radio-immunoconjugates PET tracers ( 89 Zr-DFO-9E7.4 and 64 Cu-TE2A-9E7.4). Biodistribution and imaging studies were performed. This report focuses on PET imaging with 89 Zr-DFO-9E7. 4 and follows a recent published work on 64 Cu-TE2A-9E7.4 [11]. However, comparison between both in a disseminated model is presented.

Ex Vivo Biodistribution Experiments
Ex vivo biodistribution at 24 h and 72 h post-injection (PI) results are presented in Figure 1 and Table 1. On the study conducted 24 h after administration of 89 Zr-DFO-9E7.4 ( Figure 1A,B) in a subcutaneous model of MM, the tracer displayed correct accumulation in the tumors which decreased at 72 h PI. 89 Zr-DFO-9E7.4 showed significant blood clearance from 24 h PI to 72 h PI, resulting in increased tumor-to-blood ratios. The radio-immunoconjugate also showed relative high uptakes of 89 Zr-DFO-9E7.4 in several normal organs such as liver, spleen and guts. Low muscle uptake was found at both 24 h and 72 h PI. All other organs displayed activity concentrations of 5 %ID/g or less at 24 h with decreasing activity at 72 h PI. Only flat bones and femurs showed increasing activity between 24 h and 72 h PI. As a control of potential 89 Zr release, biodistribution experiments at 24 h PI of 89 Zr-oxalate were performed ( Figure 1C,D). These showed high activity in the bones and relative high uptakes in several normal organs including blood and tumors resulting in very poor tumor-to-blood ratios.  Table 1. Biodistribution results and organ-to-blood ratios of 89 Zr-DFO-9E7.4, 89 Zr-oxalate in tumor-bearing mice. Ex vivo biodistribution results and organ-to-blood ratios of 89 Zr-DFO-9E7.4 at 24 h and 72 h post-injection (PI), in the subcutaneous tumor model (n = 3 for each group). Ex vivo biodistribution results and organ-to-blood ratios of 89 Zr-oxalate at 24 h PI (n = 3). Values are expressed in percentage of the injected radioactive dose per gram of tissue (%ID/g) and presented as mean +/− SD.  Figure 2A) while presenting very modest accumulation with 89 Zr-oxalate ( Figure 2C), with 18 F-FDG-PET being used as control imaging ( Figure 2B). Figure 3 illustrates the progressive decreasing contrast of SC tumors at 24 h, 48 h and 72 h PI and the conspicuous increasing bones uptake (predominant on coxo-femoral joints, shoulders and knees). High liver, spleen and intestines uptakes were observed at all the time points. Because of this 89 Zr accumulation in bones, as early as 48 h PI, 24 h PI time-point seems the optimal one for immuno-PET imaging with this radio-immunoconjugate.

PET Imaging of Disseminated Disease
To establish a model of disseminated disease, mice were injected intravenously via a tail vein and the distribution was assessed using bioluminescence. Mice injected IV developed lesions in the skull, spine, sacrum, iliac wings and members (Table S1).
PET imaging with 89 Zr-DFO-9E7.4 was performed 24 h, 48 h and 72 h PI ( Figure 4). Besides the physiological uptakes also observed in the SC model, bone metastases were easily distinguished with excellent tumor-to-background ratios. Table S1 shows the lesion territories for each imaging method. Imaging with 89 Zr-DFO-9E7.4 was able to detect all the lesions observed with bioluminescence imaging.
High contrast was observed at all time-points yet, as observed in the SC model, PET images that were realized at 24 h PI seemed less hampered by aspecific bone uptakes ( Figure 4).

Focus on Bones and Bone Marrows
A representative group of mice with femur lesions imaged at 24 h and 48 h PI were studied in more detail regarding bone uptake of 89 Zr ( Figure 5). Both femurs of the mice were harvested and digital autoradiography acquisitions were performed. An accumulation of activity was observed on PET and autoradiography images in mineralized constituents of the bones (compact bone and epiphysis) with similar uptakes between 89 Zr-DFO-9E7.4 at 48 h PI and 89 Zr-oxalate. This apparent strong affinity of 89 Zr for the bones and joints resulted in a hampered contrast for visualization of the MM lesions in the bone marrow cellular compartment observed at 48 h PI compared to the images realized at 24 h PI ( Figure 5).   Figure 6.B,C,E,F) showed multiple osseous uptakes corresponding to MM lesions as assessed by bioluminescence imaging, used as the gold standard. Yet symmetric joints' uptake (shoulders and knees) was also observed on PET imaging of Mouse 17, corresponding to false-positive osseous foci due to free 89 Zr.

In the Same Mouse
To determine the best radio-immunoconjugate for CD138 immuno-PET imaging in MM bone lesions, 89 Zr-labeled and 64 Cu-labeled 9E7.4 were compared. 64 Cu-TE2A-9E7.4 first and 89 Zr-DFO-9E7.4 seven days after were used for serial imaging (Table S1) and high-contrast images showing specific tumor uptakes were obtained (Figure 7). PET images were collected at 24 h PI as optimal contrast was observed at this time point for both tracers. Both tracers were able to detect all the lesions observed with bioluminescence imaging. 64 Cu-TE2A-9E7.4 was also able to detect skull infringement of the Mouse 18 ( Figure 7E), undistinguishable on 18 F-FDG-PET images ( Figure 7B). Higher tumor-to-normal tissue contrast was observed on 89 Zr-DFO-9E7.4, probably due to the seven-day interval between both immuno-PET imaging and the continuous tumor progression, as assessed by the skull infringement's visualization on the second 18 F-FDG-PET images ( Figure 7F).

Discussion
In recent years, immuno-PET established itself as a promising tool for personalized medicine in the context of multimodality treatment strategies [7]. In this context, the favorable properties of 89 Zr for mAbs imaging have resulted in the growing interest and use of this isotope [17]. Given our past experiences with the anti-CD138 mAb 9E7.4 labeled with 64 Cu [11], we evaluated in this present work 89 Zr as an alternative radiolabel for proper imaging of MM tumors with 9E7.4. This study showed that 89 Zr-DFO-9E7.4 binds effectively to CD138 tumors and allows MM imaging in a syngeneic mouse model (Figures 2-4 and Figure 7). The radiotracer displayed good targeting properties, enabling high-contrast imaging as early as 24 h PI. The images showed excellent tumor to background ratios and although the contrast decreased at 48 h PI and 72 h PI, the tumors were clearly visible with 89 Zr-DFO-9E7.4. The biodistribution data agreed well with the small animal PET results and showed 89 Zr-DFO-9E7.4 peak uptake in the MM tumors at 24 h PI (12.47 ± 4.77 %ID/g) which decreased over time. Overall, 89 Zr-DFO-9E7.4 presented markedly lower accumulation in non-target tissues with decreasing activity at 72 h PI, except for the liver, which displayed the second highest uptake at 24 h PI (13.92 ± 1.36 %ID/g) and showed longer tracer residency times (Figure 1). Only flat bones (2.97 ± 1.07 %ID/g at 24 h PI and 3.43 ± 1.12 %ID/g at 72 h PI) and femurs (3.22 ± 1.38 %ID/g at 24 h PI and 3.30 ± 0.70 %ID/g at 72 h PI) showed increasing activity between 24 h and 72 h PI (Figure 1). Such observations were not noted with the same antibody radiolabeled with 64Cu (data not published, observed between 24 h and 48 h PI) and were consistent with some preclinical studies describing the known in vivo gradual transchelation of 89 Zr over time [18].
Indeed, to date, DFO is the most common chelator used in 89 Zr-labeled radioimmuno-conjugates' studies [17,20]. However, despite good stability of the 89 Zr-DFO complex over a short period of time in preclinical studies, prolonged in vivo circulation times result in substantial release of 89 Zr from DFO and increase of 89 Zr uptake in the bones [18,[21][22][23][24]. As seen by PET images, 89 Zr bone depositions were markedly observed at the epiphyses of humerus, tibia and femur bones at 48 h and 72 h PI (Figures 3 and 4), however also as early as 24 h PI. Due to a strong affinity for phosphate, 89 Zr is steadily incorporated to hydroxyapatite, phosphates constituents of bones and particularly in the epiphyses where more active bone formation takes place [18,25]. This is illustrated by the bone dissection, PET and autoradiography imaging (Figure 4) where clear visual separation of the marrow cellular compartment with intra-medullary lesions from the mineralized bone seemed less evident at 72 h PI than 24 h PI. PET images with 89 Zr-DFO-9E7.4 slowly yet surely "turned into" PET images with 89 Zr-oxalate over time (Figures 3-5). Yet, stability of the isotopes after chelation with conjugates and fate of free radionuclides are fundamental questions for isotope selection in the design of a radiotracer. Indeed, dissociation of the 89 Zr-DFO complex could reduce the efficacy of an immuno-PET probe as an effective tool for bone lesions imaging as non-specific binding of free 89 Zr at sites of osseous turnover could induce false-positive osseous foci. This potential drawback is illustrated in Figure 6 as multiple epiphyses false-positive osseous foci are observed in Mouse 17.
Thus, although 89 Zr may hold great potential, this isotope is currently limited for both pre-clinical studies and clinical transition. In the oncological mouse model, many metastases occur in the metaphyses and epiphyses of the long bones. These sites conjugate active bone remodeling and high blood flow with fenestrated sinusoids which may predispose to tumor cells embolization and tumors growth in rodents [26,27]. Yet, with the "bone-seeking" nature of 89 Zr making these localizations preferential sites for bone accumulation too, this could directly impact immuno-PET imaging and its use as a therapeutic planning companion. Indeed, multiple preclinical studies reported the use of 89 Zr-labeled mAbs with elevated uptakes in the bones which (although not always described) clearly altered PET images' interpretations [22][23][24]28,29]. These results are also in agreement with precedent findings of the use of immuno-PET as a scouting procedure before radioimmunotherapy (RIT) with large disparity of the distribution of 89 Zr-conjugates and RIT conjugates [30][31][32]. Not to mention the fact that these preclinical studies were realized in human tumor xenografts models which usually overestimate tumor-uptake as being the sole source of antigen expression as opposed to the syngeneic mouse model.
In regards to clinical translation, 89 Zr-labeled conjugates have shown minimal uptakes in bones in most of the few human studies found in the literature. Significantly slower bone turnover is indeed observed in comparison to rodents and 89 Zr release is not expected to hinder the use of this radionuclide in patients [33][34][35][36]. Yet, recent works reporting a high level of false-positive suspicious 89 Zr-trastuzumab-avid foci in patients with breast cancer [37,38] again caused concern for the translation of this radionuclide and chelator couple.
In this study, we have also opposed 64 Cu-TE2A-9E7.4 and 89 Zr-DFO-9E7.4 with the purpose of choosing the best tracer. As seen previously, free 89 Zr and 64 Cu from unstable chelates are known to accumulate in the bone and liver, respectively. Yet, our data showed similar uptakes of 89 Zr-labeled and 64 Cu-labeled 9E7.4 in the liver (13.92 ± 1.36 %ID/g versus 9.04 ± 0.36 %ID/g, respectively, at 24 h PI). At the same time point, bone uptakes of 89 Zr-DFO-9E7.4 were higher than 64 Cu-TE2A-9E7.4 (3.1 ± 1.15 versus 1.48 ± 0.29, respectively, at 24 h PI; p = 0.0061; non-parametric test) while the opposite was observed for tumor-to-blood ratio (1.42 ± 0.24 versus 4.08 ± 1.09, respectively, at 24 h PI; p = 0.0391; non-parametric test). Weighing these factors alone, 64 Cu-labeled 9E7.4 appears as the best tracer for immuno-PET imaging. This conclusion is in agreement with the only other study which directly compared these two radionuclides in a preclinical model [39], however also with two recent works depicting the development of two immuno-PET tracers using CD38-directed human antibody daratumumab and 89 Zr [40] and 64 Cu [41], respectively, in xenograft MM mice model. Even if different methodologies were applied, the 64 Cu-labeled radioconjugate seems to be the better choice. The development of better chelator agent for 89 Zr as described in a recent review by Heskamp et al. [42] might change the face of the ranking in the future.
Consistent preclinical and clinical studies have been performed showing the potential for proper estimation of the probe biodistribution of immuno-PET before RIT [7]. Indeed, by translating tumor-to-background ratios into potential absorbed radiation doses, this approach allows for improved optimal dosing for personalized medicine in the context of multimodality treatment strategies. Moreover, 64 Cu with the beta-emitting 67 Cu provides an interesting theranostic pair with easy switch between diagnostic and therapeutic applications [16,43]. Up until now, due to its difficult production process, research on 67 Cu is still restricted [16]. Nonetheless, the feasibility of anti-CD138 RIT was reported by our team with encouraging dosimetry results. CD138 targeting with a mAb coupled to a radionuclide emitting alpha particles also represents a potential new therapeutic option for MM and the use of alpha emitters with longer half-lives, such as 211At (7.2 h), should be evaluated in the clinic.

Cell Lines and Cultures
The 5T33 murine MM cell line that was used in this study was kindly provided by Dr. Harvey Turner (Nuclear Medicine Service, Fremantle Hospital, Western Australia) with the permission of Dr. Jiri Radl (TNO Institute, Leiden, Netherlands) [44]. Cells were transfected with luciferase cDNA as previously described [45]. Then, 5T33-Luc(+) were cultured in RPMI1640 medium (Gibco, Saint Aubin, France) containing 2 mM L-glutamine and 10% heat-inactivated fetal calf serum (PAA Laboratories/GE Healthcare Europe GmbH) at 37 • C, 5% CO 2 , 95% humidity.

Preparation of 89 Zr Oxalate
The supplied 89 Zr-oxalate was neutralized with Na 2 CO 3 (2 M) and diluted with saline (0.9% NaCl) to give a final oxalate concentration of 10 mM. To establish the optimal radiolabeled 9E7.4 mAb for bone lesions immuno-PET in MM, a well-known experimental disseminated model was generated: 1.10 6 5T33-Luc(+) cells were suspended in 100 µL of PBS and injected via the tail vein into seven mice 34 days before the first PET images. Mice were monitored for bone marrow lesions by bioluminescence imaging over 33 days.

Bioluminescence Imaging
Mice were serially imaged using 2D bioluminescence imaging as previously described [14] to locate tumor progression.
The mice were anesthetized with intraperitoneal injection of 100 µL/10g of an anesthetic solution (consisting of 1 mL ketamine at 100 mg/mL (Panpharma); 0.5 mL xylazine at 20 mg/mL (Bayer); and 8.5 mL PBS). Mice were injected intraperitoneally with 100 µL of luciferin (Interchim, 12 mg/mL) 5 min before being imaged. Mice were imaged in ventral and dorsal positions using a Photon IMAGER ™ (Biospace Lab, Paris, France) for 30 s. The images were analyzed using the M3Vision ™ software (Biospace Lab, Paris, France). One mouse (Mouse 20) with no graft was imaged with 89 Zr-Oxalate as control imaging for this radionuclide.
For 18 F-FDG-PET imaging, mice were fasted overnight (6 h to 12 h) with free access to water. Mice were warmed for at least one hour, anesthetized with inhaled isoflurane 2.5% and intravenously injected with 10 MBq of 18 F-FDG in a volume of 100 µL through the lateral tail vein. Mice were maintained under anesthesia for a 1 h uptake period and then scanned (350-650 keV energy window, 20 min listmode acquisition, 3D rebinning followed by OSEM-MAP reconstruction) on a multi-modality preclinical imaging system (Inveon™, Siemens Healthcare, Erlangen, Germany). CT acquisitions (80 kV, 0.5 mA) were also performed immediately before the PET imaging. The reconstructed PET images were analyzed using Inveon Research Workplace (Siemens Healthcare).
For 89 Zr PET studies, similar procedures were followed 24 h post-18 F-FDG-PET imaging, except that no fasting was performed and imaging occurred at 24 h, 48 h and 72 h post-injection (PI). Considering the long physical half-life of 89 Zr and the pharmacokinetic profile of antibody-based radiotracers [11], time points before 24 h were not realized for longitudinal PET imaging studies. Each mouse was intravenously injected with 5 MBq of radiotracer ( 89 Zr-DFO-9E7.4 or 89 Zr-Oxalate) in a volume of 100 µL via the lateral tail vein. According to the 89 Zr decay, the specific activity at the injection time was between 116 MBq/mg and 180 MBq/mg for 89 Zr-DFO-9E7.4.
For 64 Cu PET studies, similar procedures were followed and imaging occurred at 24 h PI. Mice were intravenously injected with 10 MBq of 64 Cu-TE2A-9E7.4 in a volume of 100 µL via the lateral tail vein. According to the 64 Cu decay, the specific activity at the injection time was between 140 MBq/mg and 170 MBq/mg.

Biodistribution Study
Tracer biodistribution studies were carried out in all the SC-tumor-bearing mice after PET imaging (n = 3 for each group): at 24 h and 72 h PI for 89 Zr-DFO-9E7.4 and at 24 h PI for 89 Zr-oxalate. Tumor, blood and other selected tissues (liver, kidney, gut, lungs, muscle, spleen, skin, brain, heart, flat bone, femur and stomach) were dissected, weighed and counted on a calibrated and normalized gamma-counter. For each organ, the percentage of injected dose per gram (%ID/g) was calculated. The organ to blood ratios were also compared.

Digital Autoradiography
Femurs of Mice 14, 15 and 20 were removed, fast-frozen in cold 2-methylbutane solution, embedded in optimal-cutting temperature compound and cut into 10 µm sections using a cryomicrotome (CM3050 Leica Biosystems®). Sections were mounted on Superfrost™ slides and digital autoradiography images were obtained on a BeaQuant -Real-time autoradiography (Ai4R, Nantes, France). Image analysis was performed on the dedicated software Beamage ® (version 1.0, Ai4R, Nantes, France).

Statistical Analysis
Statistical analysis was performed using GraphPad Prism version 7.00. Differences in uptake were tested for significance using the non-parametric Mann-Whitney test for two groups. A p value below 0.05 was considered significant.

Conclusions
In this study, we synthetized 64 Cu-and 89 Zr-labeled anti-CD138 antibodies that were able to detect subcutaneous MM tumors and bone marrow lesions with high sensitivity, outperforming 18 F-FDG-PET in this preclinical model. In a theranostic approach, the stability issue and "bone-seeking" nature of 89 Zr could directly impact immuno-PET imaging and its use as a therapeutic planning companion.

Acknowledgments:
The authors would like to thank the Radioactivity Technical Platform, and especially Catherine Maurel and Séverine Lambot, the UTE IRS-UN, and particularly Sylvia Lambot, and the Cellular and Tissular Imaging Core Facility of Nantes University (MicroPICell), and specifically Myriam Robard and Stephanie Blandin, for their help.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of the data; in the writing of the manuscript, or in the decision to publish the results.