Abstract
1053
Objectives: CD33 antigen is expressed on the blast cells of most acute myeloid leukemias (AML) and has been used as a therapeutic target. The goal of this study was to develop an effective 89Zr-labeled anti-CD33 monoclonal antibody for PET imaging CD33 expression in mouse models of AML.
Methods: Commercially available CD33 antibody (Lintuzumab, Creative Biolabs, Shirley, NY) was derivatized with with p-isothiocyanatobenzyl-desferrioxamine (p-SCN-Bz-DFO) from Macrocyclics, Inc (Dallas, TX) according to previous methods.1-3 pSCN-Bz-DFO in DMSO was added to a solution of 2 mg of CD33 antibody in 200 µl of 0.1 M sodium bicarbonate pH 9.0 in a 10:1 molar ratio respectively. The solution was incubated in a shaker incubator for 60 min at 350 rpm at 37oC. The resulting CD33-DFO products were isolated using size exclusion chromatography (SEC) with a PD10 cartridge with PBS as the eluent. The fractions containing antibody were combined and concentrated down to a volume of less than 200 µl using VWR centrifugal filters with a MWCO of 10k. The resulting antibody concentration was estimated using UV-VIS spectroscopy. Lintuzumab was also conjugated using similar methods with a bifunctional isothiocyanate octadentate variant of the desferrioxamine chelator pSCN-Bz-DFO* which includes an additional hydroxamate group for improved Zr-89 stability4. The antibody chelators were radiolabeled with approximately 5 mCi (185 MBq) of 89Zr(oxalate) (obtained from Washington University in St. Louis) neutralized with 1.0 M sodium bicarbonate and buffered to pH 7.2 with PBS, and incubated at room temperature for 60 min. The radiolabeled antibody was isolated by SEC using a PD10 cartridge. The CD33 antibody was characterized by radio-TLC. Radiostability of the 89Zr was conducted by incubating the 89Zr-CD33 antibody in the presence of mouse serum and measured for 3 days. Two million CD33-positive HL-60 AML cells and 3 million CD33-negative Raji lymphoma cells were injected into either flanks of an NSG mouse. Imaging was performed using a NanoScan PET/MRI scanner (Mediso, Budapest, Hungary). FDG-PET/MRI imaging was performed at approximately 2 weeks post-implantation to confirm growth of AML and lymphoma tumors. Approximately 0.2 mCi of 89Zr-labeled anti-CD33 antibody was injected immediately after the initiation of a 1 hour dynamic PET scan. At the conclusion of the PET scan, a T1 3D GRE MRI sequence was acquired. At 72-96 hours post-injection CT, MRI, and PET imaging was repeated using a 30 minute PET acquisition. Image analysis was performed using VivoQuantTM (inviCRO, Boston, MA) to calculate Standardize Uptake Values of tumor VOIs. The same procedure was used on a 89Zr-labeled non-specific IgG control antibody to verify specificity of the CD33 antibody. Biodistribution of the 89Zr-labeled antibodies in organs and tissues were also confirmed ex-vivo using a Perkin-Elmer Packard Cobra II Autogamma (Akron, OH).
Results: The 89Zr to CD33-DFO and CD33-DFO* radiolabeling had a radiochemical yield of 58.4 + 19.7% (n=5) and 41.5 + 15.7% (n=2), respectively. A radiochemical purity of > 99.0% was measured by Radio-TLC. 89Zr-anti-CD33 antibody demonstrated minimal loss in mouse serum over 72 hrs in vitro. PET imaging with 89Zr-anti-CD33 demonstrated high uptake in HL-60 CD33-positive tumors while the specificity was confirmed with minimal uptake using 89Zr-IgG. Similarly, the Raji CD33-negative tumor did not have significant 89Zr-anti-CD33 uptake. As expected, liver was the major source of non-specific background uptake in addition to spleen.
Conclusions: The 89Zr-anti-CD33 antibody was successfully radiolabeled using both DFO and DFO* demonstrating specific targeting for AML tumors with negligible binding to the CD33-negative tumors. Further evaluation is needed to compare whether the use of DFO* vs. DFO chelator improves stability. PET imaging with 89Zr-anti-CD33 could serve as a diagnostic companion with CD33-targeted antibody drug conjugates such as Gemtuzumab Ozogamicin.