Copper 64–labeled daratumumab as a PET/CT imaging tracer for multiple myeloma

Prognostic information to stratify multiple myeloma (MM) patients into distinct risk-based treatments is currently based on specific genetic and cytogenetic abnormalities and clinical characteristics.^1,2  However, the extensive heterogeneity in therapeutic response, even in patients sharing similar characteristics,^1,2  emphasizes the need for more accurate prognostic tools.

Positron emission tomography/computed tomography using fluorodeoxyglucose-18 (¹⁸F-FDG PET/CT) imaging in newly diagnosed MM patients correlates with progression-free survival.^3,4  18F-FDG PET/CT also predicts the progression rate from smoldering MM to active disease.⁵  Hence, most investigators now agree on the importance of using imaging as a prognostic criterion for MM.^6-8 

However, despite widespread use of ¹⁸F-FDG PET/CT, its ability to detect malignancies is dependent on metabolic activity. Lesions with a low metabolic rate may therefore be undetectable. In fact, low expression of hexokinase-2 in MM cells is associated with false-negative ¹⁸F-FDG PET/CT in MM.⁹  By binding to a tumor antigen rather than relying on the metabolic state of a malignancy, antibody-based approaches may allow imaging of cancer cells with low metabolic activity. In support, the chimeric fibril-reactive monoclonal antibody can be used in patients with light chain–associated amyloidosis to identify candidates for passive immunotherapy.^10,11 

Daratumumab (Dara), a human anti-CD38 immunoglobulin G₁ (ĸ subclass) antibody against the highly expressed plasma cell receptor CD38, has shown particularly efficacious clinical activity^12-16  and recently has been approved by the US Food and Drug Administration for the treatment of relapsed MM. Because almost all MM patients express CD38 on the surface of their cancer cells, we hypothesize that tracing MM cell dissemination by targeting CD38 could be a successful approach.

See the supplemental Methods section, available on the Blood Web site.

The conjugation of Dara to DOTA was successful, as was demonstrated by gel electrophoresis (supplemental Figure 1A-B); DOTA-Dara was labeled with copper 64 (⁶⁴Cu) efficiently, with no evidence of aggregates (supplemental Figure 1C-D). The immunoreactivity of ⁶⁴Cu-DOTA-Dara preparations was >95% (supplemental Figure 1E). ⁶⁴Cu-DOTA-Dara was extremely stable up to 48 hours in saline solution and in mouse serum (supplemental Figure 2A-B). CD38-positive MM.1S green fluorescent protein/luciferase-positive (GFP⁺/Luc⁺) and CD38-negative U266 Luc⁺ MM cells,¹⁷  which express CD38 antigen at the same levels as do the respective parental cell lines (supplemental Figure 3A-B), were used to assess ⁶⁴Cu-DOTA-Dara specificity in vivo. Mice were analyzed 7 and 21 days after MM.1S GFP⁺/Luc⁺ cell injection. Each animal was then injected with the radiolabeled antibody 24 hours after bioluminescence (BLI) assessment and imaged immediately (T₀), at 4 hours and at days 1 and 2 after the injection (Figure 1A; supplemental Figure 4A). Both BLI and ⁶⁴Cu-DOTA-Dara PET/CT imaging of the same mice suggested absence of MM bone engraftment in the mice imaged at day 7, but both methods detected signals at day 21 (Figure 1A). No positive bone signs were indicated in nonengrafted mice or in mice engrafted with U266 Luc⁺ cells (Figure 1B; supplemental Figure 4B). Lack of bone signaling was also observed in mice engrafted with MM.1S GFP⁺/Luc⁺ cells but imaged by a MM nonspecific radiolabeled antibody (⁶⁴Cu-trastuzumab-DOTA [⁶⁴Cu-trast-DOTA]) (Figure 1C), and when unlabeled Dara (cold) (50:1 excess) was used in combination with ⁶⁴Cu-DOTA-Dara (Figure 1D-E; supplemental Figure 4C). Ex vivo biodistribution studies at 24 hours verified the trends observed through PET analysis (Figure 1F). ⁶⁴Cu-DOTA-Dara was significantly detectable in MM.1S-engrafted mice (n = 4) in comparison with the tumor-free mice (n = 4) and with the mice injected with ⁶⁴Cu-trast-DOTA (n = 3) in the femurs, tibias, humeri, and scapola, known sites of MM cell invasion (Figure 1F). The same biodistribution trend observed in the control mice was observed in the organs of the mice injected with U266 cells or with cold unlabeled Dara (1:50) (Figure 1F). Higher signals were also observed in all organs made of bone tissue such as sternum, skull, and spinal column and in the lung (supplemental Figure 5A-B). ⁶⁴Cu-DOTA-Dara was also significantly detected in the spleen of MM.1S GFP⁺/Luc⁺–engrafted mice (Figure 1F; supplemental Figure 5A), which was also confirmed by flow cytometry (supplemental Figure 5B). In contrast, the majority of ⁶⁴Cu-DOTA-Dara remained in the blood of the control mice (P < .0001) and was not significantly different in the other organs (supplemental Figure 5A). To assess the sensitivity of the ⁶⁴Cu-Dara signal in comparison with the standard ¹⁸F-FDG PET-CT scan, we analyzed an independent group of animals bearing CD38⁺ tumors. At 21 days after cell injection, mice were first assessed by BLI and then injected with ¹⁸F-FDG (Figure 2A-B). The same group of mice were then injected with ⁶⁴Cu-DOTA-Dara the next day, and PET imaging was performed (Figure 2C; supplemental video A). BLI and ⁶⁴Cu-DOTA-Dara PET/CT revealed the presence of MM cells, whereas tumors were almost undetectable with the ¹⁸F-FDG PET/CT (Figure 2C-D). The ¹⁸F-FDG PET signals were mainly associated with the nonspecific uptake of glucose in leg muscles and the brain. The ¹⁸F-FDG bone signal was almost equal to the background signal of ¹⁸F-FDG in the heart, a contrast to the highly specific bone signal detected by ⁶⁴Cu-DOTA-Dara (Figure 2D; see also supplemental videos A and B).

To assess the ability of ⁶⁴Cu-DOTA-Dara to recognize MM cells in the bones when the BLI signal was minimal or completely absent, we used an independent group of mice for histology analysis and for immunohistochemical staining for the MM cell marker CD138. Step section of the bones show foci of MM cells, as is also assessed by CD138 staining in the right and left femurs, confirming the presence of MM cells in the location where both BLI and ⁶⁴Cu-DOTA-Dara signals were present (Figure 2E-G). We also detected MM cells in the lower vertebral body, which had no BLI signal (Figure 2E-G) and no cells in the humerus, which was also found negative by PET/CT scan (Figure 2E-G).

The relatively short half-life of ⁶⁴Cu (12.7 hours) may be ideal for patient safety and is sufficiently long for PET/CT imaging up to 48 hours after administration. We have used ⁶⁴Cu-labeled trastuzumab to safely detect human epidermal growth factor receptor 2 (HER2)–positive tissue in metastatic breast cancer (MBC) patients.^18 64Cu was safe, and by day 1 after injection, ⁶⁴Cu-trastuzumab-DOTA was able to detect MBC, its binding correlating with patient HER2 status. Although Dara has been conjugated with the radionuclide ⁸⁹zirconium (⁸⁹Zr),¹⁹  the tendency of ⁸⁹Zr to accumulate in the bones,^19-21  associated with a longer time requested for imaging in contrast to ⁶⁴Cu, may be a limitation in specifically detecting MM cell dissemination in the bones. Our data clearly show higher sensitivity of ⁶⁴Cu-Dara-DOTA not only in comparison with ¹⁸F-FDG signal but also in comparison with BLI imaging, an effect that was not seen when Dara was conjugated with ⁸⁹Zr.¹⁹  Because our studies show a high sensitivity of ⁶⁴Cu-Dara-DOTA in detecting MM cells, we believe that this imaging approach for MM may also be clinically relevant for detecting minimal residual disease (MRD). MRD, as detected by high throughput bone marrow DNA sequencing, is being considered as a surrogate endpoint in clinical trials,^22-24  because effective therapies necessitate 5 to 10 years of follow-up to detect differences in survival.^25,26  However, irrespective of its sensitivity, this technique is marrow based and cannot account for a nonhomogenous bone marrow infiltration and extramedullary involvement. In support of Dara as a unique tool for imaging in MM, our group developed a ⁶⁴Cu antibody (elotuzumab [Elo]) against the highly specific MM receptor CS1 (⁶⁴Cu-DOTA-Elo). Although ⁶⁴Cu-DOTA-Elo bound the CS1 antigen in vitro, it did not identify bone engraftment of CS1-positive MM cells in in vivo studies (data not shown). We have begun a ⁶⁴Cu-DOTA-Dara PET/CT imaging trial that has the potential to affect MM clinical cell detection.

The authors acknowledge Dori Triplett, Evelyn Flores, and Debbie Flood for administrative support. Research reported in this publication included work performed in the Center for Biomedicine and Genetics facility.

This work was supported by a Steven Gordon and Briskin Family Innovation Grant (F.P., D.C., J.S.). The Small Animal Imaging Center City of Hope (COH) core and Pathology core used for immunohistochemical staining were supported by the National Institutes of Health National Cancer Institute (P30CA033572) and by the COH Veterinary Pathology Program. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Contribution: E.C., S.H., J.K., S.R., J.S., N.C., D.C., A.K., and F.P. designed the research; E.C., J.C., M.M., E.K.P., D.V., S.V., J.K., P.Y., Y.K., and R.B. performed the research; E.C., J.F.S., J.M.P., G.M., J.S., D.C., and F.P. analyzed the data; J.F.S. and F.P. wrote the manuscript; and all authors reviewed the manuscript.

Conflict-of-interest disclosure: A.K. is a consultant for and on the speakers’ bureau of Janssen Pharmaceuticals. The remaining authors declare no competing financial interests.

Correspondence: Flavia Pichiorri, Department of Hematology and Hematopoietic Cell Transplantation, City of Hope, 1500 E Duarte Rd, Duarte, CA 91010; e-mail: fpichiorri@coh.org; and David Colcher, Department of Molecular Immunology, City of Hope, 1500 E Duarte Rd, Duarte, CA 91010; e-mail: dcolcher@coh.org.