The clinical efficacy of the first approved alpha pharmaceutical, Xofigo (radium-223 dichloride, 223RaCl2), has stimulated significant interest in the development of new alpha-particle emitting drugs in oncology. Unlike radium-223 (223Ra), the parent radionuclide thorium-227 (227Th) is able to form highly stable chelator complexes and is therefore amenable to targeted radioimmunotherapy. We describe the preparation and use of a CD33-targeted thorium-227 conjugate (CD33-TTC), which binds to the sialic acid receptor CD33 for the treatment of acute myeloid leukemia (AML). A chelator was conjugated to the CD33-targeting antibody lintuzumab via amide bonds, enabling radiolabeling with the alpha-emitter 227Th. The CD33-TTC induced in vitro cytotoxicity on CD33-positive cells, independent of multiple drug resistance (MDR) phenotype. After exposure to CD33-TTC, cells accumulated DNA double-strand breaks and were arrested in the G2 phase of the cell cycle. In vivo, the CD33-TTC demonstrated antitumor activity in a subcutaneous xenograft mouse model using HL-60 cells at a single dose regimen. Dose-dependent significant survival benefit was further demonstrated in a disseminated mouse tumor model after single dose injection or administered as a fractionated dose. The data presented support the further development of the CD33-TTC as a novel alpha pharmaceutical for the treatment of AML. Mol Cancer Ther; 15(10); 2422–31. ©2016 AACR.
Xofigo (radium-223 dichloride, 223RaCl2) is the first in-class alpha pharmaceutical approved for the treatment of metastatic castration-resistant prostate cancer (1). The inherent bone-seeking properties of radium-223 (223Ra; see ref. 2) allows for the effective delivery to bone metastases. The lack of suitable complexing agents for 223Ra limits its use for tumor-targeting moieties, such as antibodies, and hence broader utility in oncology (3). In contrast, thorium-227 (227Th), the parent radionuclide of 223Ra, can be readily complexed by octadentate chelates of the 3,2 hydroxypyridinone (3,2-HOPO) class. These 3,2-HOPO chelates, complexed with 227Th, can further be attached to tumor-targeting antibodies, referred herein as targeted thorium-227 conjugates (TTC). TTCs have the potential to deliver the high potency of the alpha-particle to a multitude of different cancer types, including those of the lympho-hematopoietic system.
Acute myeloid leukemia (AML) is a rare, highly malignant neoplasm with a poor survival prognosis and results in a high number of cancer-related deaths. The age-adjusted incidence rate of AML in the United States in the years 1975 to 2003 was 3.4 per 100,000 persons and the disease accounted for about 25% of all leukemia in the Western world (4). Despite the high unmet medical need, the pace of new drug approvals for AML in recent years has remained almost at a standstill. The antibody–drug conjugate (ADC) gemtuzumab ozogamicin (GO; Mylotarg) targeting the CD33-positive leukemic blasts, initially approved by the FDA in 2001 was later withdrawn from the market in 2009 due to lack of prespecified overall improvement in outcome, along with drug-related toxicities (5, 6). Although many strategies are currently being assessed which attempt to target the genetic heterogeneity of AML, including hypomethylating agents, fms-related kinase inhibitors, and many combination therapies, there is still need for new effective drugs for the treatment of AML (7).
Recently, preclinical and clinical development has focused on two new antibody drug conjugates (ADC), SGNCD33A (8) and IMGN779 (9), both currently being tested in clinic trials. Furthermore, AMG330, a CD33/CD3-directed bispecific T-cell engager antibody (BiTE; refs. 10–13), has entered clinical testing and chimeric antigen receptors T-cell approaches (CAR-T) are additionally feeding the pipeline for the treatment of AML (14, 15). All of these immunotherapeutic agents are targeting the sialoadhesin receptor CD33 (Siglec-3), a 67-kDa protein, expressed on leukemic blasts of AML patients (16). Lintuzumab, a humanized anti-CD33 IgG1 antibody (17, 18), although well-tolerated in AML patients, showed only modest activity even at doses capable of completely saturating CD33-binding sites (19). Furthermore, AVE9633, an anti-CD33 antibody coupled to a maytansinoid derivative, and an immunotoxin conjugate, that is, HuM-195/rGel, lintuzumab linked to gelonin, were evaluated in clinical phase I trials and showed only modest activity (20, 21).
An alternative to antibody–drug conjugates or immunotoxins is radioimmunotherapy which capitalizes on the radiosensitivity of AML (22, 23). Radioimmunotherapy has the potential added benefit of overcoming chemoresistance induced by multi-drug resistance pumps (24, 25). As such, lintuzumab has been evaluated as radioimmunoconjugate with the beta-emitter iodine-131 (131I) and with the alpha-emitters bismuth-213 (213Bi) and actinium-225 (225Ac) (26). Early clinical studies using 131I-lintuzumab demonstrated the clinical potential of radioimmunotherapy in AML (27), but also highlighted limitations of this approach due to radiolysis of the drug product and off-target toxicity caused by the long range of γ-radiation. Off-target toxicity can potentially be reduced by using high-energy, short-range alpha-emitters (range in tissue of approximate 50–80 μm) which are characterized by their high linear energy transfer (LET; ref. 28). Even though 213Bi-lintuzumab, when used as single agent or in combination with conventional chemotherapy, demonstrated acceptable safety and antileukemic activity (29, 30), its use in clinic is challenging due to the short half-life of 213Bi (46 minutes). In contrast, 225Ac has a half-life of 10 days and 225Ac-lintuzumab (Actimab A) and demonstrated antileukemic activity as a single agent or in combination with conventional chemotherapy in clinical trials at doses of ≤111 kBq/kg (31, 32).
We present the development of a CD33-targeted thorium-227 conjugate (CD33-TTC), comprising the CD33 monoclonal antibody lintuzumab, an octadentate 3,2-HOPO chelate (33), conjugated to the antibody allowing radiolabeling with the alpha-emitter thorium-227. Thorium-227 has a half-life of 18.7 days and is purified from an actinium-227 generator. In the decay chain of thorium-227, a total of five high-energy daughter alpha-particles (223Ra, 219Rn, 215Po, 211Bi, and 211Po) and two beta decays (211Pb and 207Tl) are generated, with 211Pb as the final stable end product (ref. 34; Supplementary Fig. S1). The presented preclinical data support the further development of this new class of targeted alpha pharmaceutical for the treatment of AML.
Materials and Methods
HL-60, KG-1, and Ramos cells were obtained from ATCC. All cell lines were authenticated using by the manufacturer. The cell lines were purchased within the period of 2008–2015, characterized by the vendor (PCR fingerprinting), and no further cell line authentication was conducted by the authors. Cells were maintained in an incubator of 5% CO2 at 37°C. HL-60 and KG-1 cells were cultured in IMDM with 20 % FCS and 1% penicillin and 1% streptomycin. Ramos cells were cultured in RPMI1640 with 10% FCS and 1% penicillin and 1% streptomycin.
Expression, purification, and conjugation of the CD33 antibody–chelator conjugate
The gene encoding the CD33 antibody lintuzumab was synthesized (GeneArt/Life Technologies) and cloned into a plasmid enabling expression in CHO-K1 cells. The CD33 antibody was harvested from the cell culture and purified via an affinity MAbSelectSure column (GE Healthcare) followed an anion exchange chromatography column (QFF Sepharose; GE Healthcare) and a cation exchange chromatography column (Poros XS; Life Technologies) to remove high-molecular weight fractions and in-process impurities. The antibody was filtered through a 0.22-μm filter (Millipore), formulated in PBS, pH 7.2 and stored at −64°C. All work was conducted by Cobra Biologics (Sweden).
The synthesis of the 3,2-HOPO chelator is described elsewhere (33) and was conducted by Synthetica (Norway).
Conjugation of the 3,2-HOPO chelator to lysine residues within the CD33 antibody and an isotype control was achieved through in situ activation of the chelator using EDC/NHS chemistry and subsequent incubation with the antibody in PBS, pH 7.0, for 60 minutes at room temperature. The resulting antibody–chelator conjugates were purified by size-exclusion chromatography (SEC; HiLoad 16/600 Superdex 200; GE Healthcare) and stored at −20°C.
The chelator to antibody ratio (CAR) was determined by SEC (TSKgel SUPER SW 3000 column; Tosoh) by monitoring the absorbance of the CD33 antibody–chelator conjugate at 280 and 335 nm. The CAR value was calculated using the following formula: CAR = extinction coefficient of mAb (335 nm) − ratio × extinction coefficient of mAb (280 nm)/ratio × extinction coefficient of chelator (280 nm) − extinction coefficient of chelator (335 nm).
Radiolabeling and characterization of the CD33-TTC
Thorium-227 was purified from an actinium-227 generator as described previously (35). Two hundred and fifty micrograms of CD33 antibody–chelator conjugate were mixed with 227Th-activities, ranging from 0.2 to 2.5 Mbq and incubated at room temperature for 60 minutes. Radiochemical purity (RCP) was determined by instant thin-layer chromatography (iTLC). For radiostability testing, the CD33-TTC was analyzed over the course of 48 hours by iTLC and SEC.
The immunoreactive fraction (IRF) was determined after Lindmo (36). Recombinant human CD33 (Novus Biologicals) or BSA (Sigma) was coated to tosyl-activated magnetic beads (Dynabeads M-280; Life-Technologies). Fifty becquerel of CD33-TTC/sample [measured on an HPGe counter (GEM; Perkin Elmer)] were incubated for 1.5 hours at 37°C on a titration of 1.5 to 25 million beads of either CD33- or BSA-coated beads. Beads were sorted on a magnetic rack and the radioactivity in the supernatant and on the bead pellets was determined. The IRF was calculated as the fraction of activity bound to CD33-coated beads subtracting the unspecific activity measured on BSA-coated beads.
Internalization of the CD33-TTC was compared with an isotype control-TTC as described previously (37). HL-60 cells (400,000/mL) were incubated in presence of 1 μg of CD33-TTC or isotype control-TTC (radiolabeled at a specific activity of 20 kBq/μg) at 37°C with 5% CO2. Cells were pelleted by centrifugation at different timepoints (30, 60, 120, and 240 minutes, and 6 and 24 hours). The radioactivity in the cell pellets was recorded using a gamma counter (Wizard/Perkin Elmer) and counts per minute were plotted against time to monitor internalization.
Binding potency to CD33 by ELISA and FACS analysis
Recombinant human CD33 (Novus Biologicals) was coated to ELISA plates (1 μg/mL; NUNC/Maxisorp). Wells were blocked with skim milk [4 % (w/v)]. CD33 antibody, CD33 antibody–chelator conjugate, an isotype antibody–chelator conjugate, and CD33-TTC (radiolabeled at a specific activity of 50 kBq/μg) were titrated (1:3; 100 μg/mL) on the CD33-coated ELISA plate. Unbound samples were washed off and bound samples were visualized using horseradish peroxidase–labeled anti-human kappa-HRP antibody (Southern Biotech), followed by incubation with 1-step TMB blotting solution (Thermo Scientific Pierce). The reaction was stopped by addition of 1 mol/L HCl. The absorbance was measured at 450 nm in a plate reader (Perkin Elmer). EC50 values were calculated using GraphPad Software.
CD33-expresisng HL-60 cells (200,000 cells/well) were incubated in presence of serial dilutions (1:3; 100 μg/mL) of CD33 antibody, CD33 antibody–chelator conjugate or an isotype antibody-chelator conjugate. Unbound samples were washed off and bound samples were detected using PE-labeled anti-human IgG (Biolegend). Mean fluorescence intensities (MFI) were recorded using a Quanta SC MPL machine (Beckman Coulter). EC50 values were calculated using GraphPad software.
Measurement of cell viability and cell-cycle arrest
CD33 receptor densities on HL-60, KG-1, and Ramos cells were determined using QiFiKit (Dako). Cells were seeded into 24-well culture plates (200,000 cells/mL). CD33-TTC, an isotype control-TTC (both radiolabeled at a specific activity of 50 kBq/μg) and CD33 antibody–chelator conjugate were incubated on the cells for 4 hours at 37°C at activities ranging from 0 to 20 kBq/mL. Cells were washed and reseeded into a new 24-well culture plate. At different timepoints, cells were harvested and the viability was measured using CellTiterGlo kit (Promega). The cell viability was expressed in percent by normalization to cells grown in medium only. DNA double-strand breaks were determined using AlexaFluor 647–labeled anti-phospho-Histone H2A.X antibody (Cell Signaling Technology). Cells were costained with an AlexaFluor 488–labeled anti-cleaved caspase-3 antibody (Cell Signaling Technology) to detect no-apoptotic cells. Cell-cycle analysis was analyzed using the apoptosis, DNA damage, and cell proliferation kit (ADDCP; BD Biosciences). FACS analysis was performed using a Quanta SC MPL instrument (Beckman Coulter).
All animal models were conducted at Pipeline Biotech and were in accordance with the Danish animal welfare law, approved by local authorities. In all studies, animals received an intraperitoneal (i.p.) injection of an unrelated murine IgG2a antibody (200 μg/animal; UPC10; Sigma) 24 hours prior treatment to block unspecific spleen uptake (38).
In vivo biodistribution
Five million HL-60 cells, suspended in 0.1 mL 50% Matrigel (BD Biosciences), were inoculated subcutaneously into 6 mice per group (female/NMRI nu/nu mice/Taconic, Europe). At an average tumor volume of 200–300 mm3, animals were administered with a single intravenous injection of CD33-TTC (500 kBq/kg; protein dose of 0.36 mg/kg). Tumors and organs were harvested at t = 24 hours, t = 4 days, and t = 7 days. Radioactivities were counted using an HPGe detector linked to an autosampler (Gamma Data) at Bayer AS. Thorium-227 and 223Ra activities were decay corrected to the timepoint of injection and plotted in Bq/g.
In vivo efficacy models
Five million HL-60 cells, suspended in 0.1 mL 50% Matrigel (BD Biosciences), were inoculated subcutaneously into 10 mice per group (female/NMRI nu/nu mice/Taconic). At an average tumor volume of 70–100 mm3, animals received either a single intravenous injection of CD33-TTC or isotype control-TTC (radioactive dose of 700 kBq/kg; protein dose of 0.36 mg/kg), CD33 antibody–chelator conjugate (0.36 mg/kg) or vehicle. Body weights were measured biweekly. Tumor growth was calculated using the formula: V = 0.5 × (length + width)2. Animals were sacrificed by cervical dislocation upon reaching the humane endpoint (tumor volume ≥ 1,500 mm3; body weight loss ≥ 20%).
In the disseminated tumor model, female C.B-17 SCID mice (Taconic) were injected intravenously with 5,000,000 HL-60 cells. At study day 5, animals received a single intravenous injection of CD33-TTC at radioactivity doses of 50, 150, or 300 kBq/kg (protein dose of 0.04 mg/kg). An additional treatment group received a second intravenous injection of CD33-TTC at a dose of 150 kBq/kg (day 19). Control groups received a single intravenous injection of an isotype control-TTC (radioactivity dose of 300 kBq/kg; 0.04 mg/kg), CD33 antibody–chelator conjugate (0.04 mg/kg), or vehicle. White and red blood cell counts were determined (Vet ABC analyzer) at study day -1 (vehicle), day 67, and at day 123 (end of study). Animals were sacrificed by cervical dislocation upon reaching humane endpoints defined by clinical signs of disease including palpable tumors, scruffy fur, weight loss, and hind leg paralysis. Diseased organ samples were isolated and analyzed for CD33 (Novacastra) and CD45 (Dako) expression by IHC on paraffin-embedded tissue sections (TISSUE-TEK/ Miles Scientific) at Micromorph (Sweden). Survival plots were generated using GraphPad software. Statistical analysis was performed using a log-rank (Mantel–Cox) test and was considered to be significant at P < 0.05.
Preparation and characterization of the CD33-TTC
The cDNA sequence of the CD33 antibody lintuzumab was derived from the literature (18). Standard molecular biology techniques were used, enabling recombinant expression in CHO-K1 cells and subsequent purification as described previously. In parallel, the 3,2-HOPO chelator was synthesized (33), capable of binding the radionuclide 227Th at one end, and carrying a reactive NHS-group at the other end to enable conjugation of the chelator to free lysine residues in the CD33 antibody, thereby forming stable amide bonds. A schematic presentation of the resulting CD33-TTC is presented in Fig. 1A. A chelator to antibody ratio (CAR) of 0.8 chelator molecules per antibody was determined by size-exclusion chromatography by monitoring the UV signal at 280 nm for protein absorbance and at 335 nm for chelator absorbance in parallel.
The radiolabeling properties of the CD33 antibody–chelator conjugate were studied in more detail. The CD33 antibody–chelator conjugate was incubated with 227Th at room temperature for ≥60 minutes. The radiochemical purity (RCP), defined as the amount of bound 227Th in the final CD33-TTC product, was determined by iTLC to be consistently ≥95%. Furthermore, the CD33-TTC demonstrated high RCP and only slight increase in high-molecular weight fractions when stored for 48 hours at room temperature (Fig. 1B), enabling in vivo applications.
The binding properties of the CD33 antibody, the CD33 antibody–chelator conjugate, and the CD33-TTC were subsequently tested. No significant loss in binding affinity to recombinant human CD33 due to conjugation or radiolabeling was observed by ELISA analysis (Fig. 2A). Similarly, no loss in binding affinity to CD33 expressed on HL-60 cells was observed by FACS analysis (Supplementary Fig. S2). Determination of the immunoreactive fraction (IRF) for CD33-TTC demonstrated that ≥95% of the radiolabeled CD33-TTC bound to CD33 molecules in a bead-based assay (Fig. 2B). Furthermore, as described in the literature (17), the lintuzumab-based CD33-TTC was found to internalize at a 3.4-fold higher rate than an isotype control-TTC (Fig. 2C).
In vitro cytotoxicity of the CD33-TTC
The in vitro cytotoxicity of the CD33-TTC was tested on the CD33-positive cell lines HL-60 and KG-1. These two cell lines were determined to express approximately 16,000 and 7,800 receptors/cell, respectively, and differ in their multidrug efflux pump status (HL-60, MDR-negative; KG-1, MDR-positive). Exposure of cells to CD33-TTC demonstrated dose-dependent reduction of cell viability on both cell lines (Fig. 3A and B). The reduction in viability was specific to CD33-TTC as an isotype control-TTC did not decrease cell viability. In addition, CD33 antibody–chelator conjugate did not reduce cell viability, nor did treatment of CD33-negative cells Ramos with CD33-TTC (Fig. 3C), thus demonstrating the targeting specificity of the CD33-TTC.
We next investigated the mode of action of the CD33-TTC. Thorium-227 has been shown to induce DNA double-strand breaks (DSB) when complexed to other targeting moieties such as rituximab (39). Therefore, HL-60 cells were exposed to isotype control-TTC or CD33-TTC (activities of 20 kBq/mL). Induction of DNA DSBs in nonapoptotic cells, gated using an anti-caspase3 antibody, was measured by detection of phosphorylated H2AX (γ-H2AX) by FACS. Approximately 2% of cells cultured in medium stained positive for γ-H2AX, whereas approximately 10% of cells exposed to isotype control-TTC and approximately 40% of cells exposed to CD33-TTC stained positive for γ-H2AX (Fig. 3D).
Next, we evaluated the effect on the cell cycle after exposure of HL-60 cells to CD33-TTC by flow cytometry analysis. As presented in Fig. 3E, cells were found to enter the cell-cycle G2 phase within 24 hours after exposure to CD33-TTC, which further increased 48 hours after exposure to CD33-TTC in comparison with cells cultured in medium. These results indicate that CD33-TTC induces G2 cell-cycle arrest.
In vivo biodistribution of the CD33-TTC
The biodistribution of the CD33-TTC was evaluated in a subcutaneous HL-60 xenograft mouse model. An increase over time of 227Th activity in the tumor was observed with an accumulated activity of approximately 1,500 Bq/g (corresponding to 22% of injected dose 227Th per gram) on day 7 (Fig. 4A). The accumulation in the tumor was the reciprocal of the decrease of 227Th in the blood and an activity of approximately 300 Bq/g of 227Th per mL blood (corresponding to 5% of injected dose227Th per gram) was determined on day 7. No major accumulation in any other organs was observed, which can be attributed to the lack of cross-reactivity of lintuzumab to murine CD33. The 223Ra activity from decaying 227Th was analyzed in parallel (Fig. 4B). A gradual increase of the 223Ra activity in the tumor was observed with an accumulated activity of approximately 300 Bq/g on day 7. Similarly, increase of 223Ra activity at a count of approximately 300 Bq/g was detected in the femur. In summary, high tumor accumulation of the CD33-TTC was demonstrated.
In vivo antitumor activity of the CD33-TTC
The antitumor activity of the CD33-TTC was evaluated in vivo in a subcutaneous HL-60 xenograft model. At the administered dose of 700 kBq/kg, complete tumor regression was observed, lasting for 20 days. Fifteen of 18 animals were free of any palpable tumors at the end of the study (Fig. 5A). In contrast, an isotype control-TTC administered at an equal radioactive dose showed no antitumor activity. Furthermore, tumor growth inhibition was dependent on the presence of 227Th, as CD33 antibody–chelator conjugate lacked in vivo efficacy. No loss in body weight ≥10% was observed during the course of the study, demonstrating that the dose was well tolerated (Fig. 5B) which is in agreement with previous published data (40).
To further substantiate in vivo activity, CD33-TTC was evaluated in a disseminated HL-60 model. CD33-TTC was administered at radioactivity activities as indicated in Fig. 6. Dose-dependent efficacy of the CD33-TTC was observed (Fig. 6A) and median survival times (MST) of 90 days, 115.5 days, or undefined for animals receiving 50, 150, and 300 kBq/kg were calculated. In addition, animals receiving a fractionated dose of 2 × 150 kBq/kg of CD33-TTC at an interim of two weeks had similar MST as a single dose of 300 kBq/kg of CD33-TTC. In contrast, MST of 55, 56, and 57 days for animals receiving vehicle, CD33 antibody–chelator conjugate, or an isotype control-TTC, respectively, were determined. All doses were well tolerated as judged by body weight measurements (Fig. 6B) and clinical blood chemistry. As presented, white and red blood cell counts measured at study day 67 (3.5 half-lives of 227Th) and 123 (6.5 half-lives of 227Th) were comparable with levels measured in animals treated with vehicle at study day −1 (Supplementary Fig. S3B). This is in agreement with previous published data (40), which demonstrate that 227Th induces reversible myelosuppression.
Immunohistochemical analysis on diseased animals terminated at the humane endpoint revealed HL-60 positive tumors in kidneys and bone marrow (Supplementary Fig. S4) which were not present in animals treated with CD33-TTC, alive at the end of the study.
There is an unmet medical need for the treatment of AML and CD33 is a validated target despite rather discouraging results from clinical trials in the past, including the withdrawal of Mylotarg. As such, SGN-CD33A is an ADC consisting of a humanized CD33 antibody that is armed with DNA cross-linking pyrrolobenzodiazepine dimers (PBD) which are released via a protease-cleavable linker. SGNCD33A demonstrated promising preclinical activity (8) and is currently being investigated in clinical trials. Similarly, IMGN779 (9) makes use of a new class of DNA alkylating agents called indolinobenzodiazepine pseudodimers (termed IGN) and has recently entered clinical phase I. Another strategy being pursued includes the development of bispecific T-cell engaging molecules including BiTE AMG330 (10–13) and a recently developed tetravalent TandAb CD33/CD3 molecule (41). The development of BiTEs for the treatment of hematopoietic disorders might be promising as exemplified by blinatumomab, a CD19-CD3 BiTE, approved for the treatment of acute lymphoblastic leukemia (42).
The current study describes the development of a CD33-targeted thorium-227 conjugate (CD33-TTC), enabling the targeted delivery of 227Th to CD33-positive cells. The data presented include detailed in vitro characterization and demonstrate robust in vivo antitumor activity of the CD33-TTC in subcutaneous and disseminated mouse models of AML.
In contrast to ADCs and T-cell engaging molecules, the efficacy of the CD33-TTC relies only on binding of the targeting moiety to the antigen, thereby delivering a toxic dose of radiation to the cells. With the purpose to not impair the initial binding event during manufacturing of the CD33-TTC, we have developed a conjugation and radiolabeling process that is performed at neutral pH and ambient temperatures. As demonstrated by ELISA, FACS, and IRF experiments, neither the conjugation nor the radiolabeling process impaired the binding properties to CD33 and the RCP was repeatedly ≥95%. In contrast, antibodies have been previously radiolabeled with 227Th using p-SCN-Bn-DOTA, requiring conjugation of p-SCN-Bn-DOTA at a basic pH (pH 9) and subsequent radiolabeling with 227Th for a minimum of 2 days at room temperature or at shorter periods at temperatures ≥60°C (43). Both of these steps are rather harsh with the increased likelihood of impairing the biological activity.
The mode of action (MoA) of the CD33-TTC was demonstrated to induce DNA double-strand breaks in cells. Furthermore, cells entered G2 cell-cycle arrest and potentially apoptosis as observed by the detection of apoptotic bodies using microscopy (data not shown). This is in agreement with other studies using alpha-emitters as described for 213Bi-labeled radioimmunoconjugates (44, 45). Furthermore, the CD33-TTC demonstrated reduction of in vitro viability on CD33-positive cell lines independent on the MDR status, potentially being an advantage in AML patients which have developed resistance to standard-of-care treatment (46).
The observed reduction of viability in in vitro assays correlated with in vivo efficacy in a subcutaneous xenograft model. Treatment with a single dose of CD33-TTC stopped tumor growth as well as caused tumor regression in contrast to an isotype control-TTC. These findings were supported by the biodistribution of the CD33-TTC where a gradual increase of 227Th activity in the tumor was observed. Similar, a gradual increase of 223Ra activity was measured in femur. This accumulation in the femur can be explained by 227Th decaying to 223Ra, which is detached from the CD33-TTC and subsequently incorporated into the core bone structure (2, 47).
In addition to the subcutaneous model, we demonstrated efficacy of the CD33-TTC in a disseminated mouse model of AML, resembling more closely human AML disease. Dose dependency of the CD33-TTC was observed, resulting in prolonged median survival in comparison with respective control groups. The disseminated model was performed in SCID mice due to the biology of the model. However, SCID mice are reported to be more sensitive to ionizing radiation due to their reduced efficiency in repairing DNA double-strand breaks (48, 49). Furthermore, it has been previously described that the no-observed-adverse-effect-level (NOAEL) for 227Th-radioimmunoconjugates is in between 200 and 300 kBq/kg (40) and the maximum tolerated activity ranges from 600 to 1,000 kBq/kg dependent on the mouse strain (Balb/C and nude). Therefore, the maximum applied activity in the disseminated model was 300 kBq/kg (SCID) and 700 kBq/kg in the subcutaneous model (nude). White and red blood cell counts of animals treated with CD33-TTC in the disseminated model had comparable levels as vehicle-treated animals and no other adverse signs of toxicity (loss of body weight) were observed. Similarly, the activity of 700 kBq/kg in nude mice (subcutaneous model) was well tolerated as judged by body weight measurements and is in agreement with previously published work (40). However, caution should be taken during a potential clinical dose escalation and myelosuppression should be monitored carefully.
Taken together, our results support the development of CD33-TTC for the treatment of AML. The optimized chelator chemistry allows conjugation and radiolabeling reactions under mild conditions. Thorium-227 has a half-life of 18.7 and can be produced in commercial quantities from an 227Ac-generator. These features reduce logistical hurdles and can be adapted to other targeting moieties. As such, we have recently initiated a clinical trial to assess the safety and tolerability of a CD22-targeted thorium-227 conjugate (BAY1862864) in relapsed or refractory CD22-positive non-Hodgkin lymphoma (clinical trial number NCT02581878).
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Conception and design: U.B. Hagemann, K. Wickstroem, A.O. Shea, J. Karlsson, R.M. Bjerke, A.S. Cuthbertson
Development of methodology: U.B. Hagemann, K. Wickstroem, E. Wang, A.O. Shea, J. Karlsson, O.B. Ryan, A.S. Cuthbertson
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): K. Wickstroem, E. Wang, A.O. Shea
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): U.B. Hagemann, K. Wickstroem, E. Wang, A.O. Shea, K. Sponheim, J. Karlsson, R.M. Bjerke, A.S. Cuthbertson
Writing, review, and/or revision of the manuscript: U.B. Hagemann, K. Wickstroem, E. Wang, A.O. Shea, J. Karlsson, R.M. Bjerke, A.S. Cuthbertson
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): U.B. Hagemann, K. Wickstroem, E. Wang, J. Karlsson, R.M. Bjerke, O.B. Ryan
Study supervision: U.B. Hagemann, E. Wang, A.O. Shea, R.M. Bjerke, O.B. Ryan, A.S. Cuthbertson
We thank Lars Abrahamsen for scientific advice and contributions. We thank Cobra Biologics (Sweden) for manufacturing of the CD33 antibody, Synthetica for synthesis of the chelator, Pipeline Biotech for conducting the animal experiments, and Micromorph for assisting with immunohistochemistry.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Note: Supplementary data for this article are available at Molecular Cancer Therapeutics Online (http://mct.aacrjournals.org/).
- Received April 29, 2016.
- Revision received July 31, 2016.
- Accepted August 2, 2016.
- ©2016 American Association for Cancer Research.