In vitro and In vivo Characterization of ⁶⁴Cu-Labeled Abegrin^TM, a Humanized Monoclonal Antibody against Integrin α_vβ₃

Angiogenesis, the formation of new blood vessels from pre-existing vasculature, is a fundamental process occurring during tumor progression and it depends on the balance between proangiogenic molecules and antiangiogenic molecules (1, 2). Cancer cells spread throughout the body by metastasis (3). Interactions between vascular cells and the extracellular matrix (ECM) are involved in multiple steps of tumor angiogenesis and metastasis (4, 5). Integrins, a family of cell adhesion molecules composed of noncovalently associated α and β subunits, are involved in a wide range of cell-ECM and cell-cell interactions (4–6). Inhibition of α_v integrin activity by monoclonal antibodies (mAb), cyclic RGD peptide antagonists, and peptidomimetics has been shown to induce endothelial cell apoptosis, to inhibit angiogenesis, and to increase endothelial monolayer permeability (7, 8). The α_vβ₃ integrin, which recruits and activates matrix metalloproteinase-2 and plasmin to degrade the basement membrane and interstitial matrix (9), is significantly up-regulated on endothelium during angiogenesis but not in quiescent endothelium (4, 5, 7). Integrin α_vβ₃ is also expressed in melanoma, late-stage glioblastoma, ovarian, breast, and prostate cancer cells (10), where it can potentiate metastasis by facilitating cancer cell invasion and movement across blood vessels.

LM609, a mouse anti-human integrin α_vβ₃ mAb, which cross-reacts with α_vβ₃ originated from rabbits, chicken, and hamsters but not from mice and rats, was found to be able to immunoprecipitate integrin α_vβ₃ from M21 human melanoma cells (11). Recognizing the heterodimer as one entity, LM609 is much more specific and superior to other mAbs, which recognizes either the α_v or β₃ subunit. However, being a murine mAb, inefficient interactions with human immune effector cells and the short serum half-life (t_1/2) of LM609 caused by its immunogenicity in humans severely limits the therapeutic potential of LM609 when chronic administrations are needed (12). For these reasons, LM609 was humanized and later affinity matured as reported previously (13). The humanized version, named Vitaxin^TM or Vitaxin I (later named MEDI-523) was used in several phase I clinical trials (13–15). Radioimaging of tumor vasculature using this early version of Vitaxin^TM labeled with ^99mTc was unsuccessful due to the instability of the ^99mTc labeling in vivo (16). The affinity matured version, now called Abegrin^TM (also called Vitaxin^TM or MEDI-522), is also in clinical trials. In phase I studies of Abegrin, prolonged stable disease was observed in several patients with renal cell carcinoma and an effect on tumor perfusion was correlated with treatment, although no significant toxicity or immune response was noted at the dose levels tested (17). In 2003, MedImmune, Inc. (Gaithersburg, MD), who licensed Vitaxin^TM and designated it as Abegrin^TM for clinical development (18), announced the initiation of phase II clinical trials in prostate cancer and melanoma. Recently, the company announced the initiation of phase III for patients with metastatic melanoma based on promising results from their phase II trial with Abegrin^TM.

We and others have reported small molecule, peptide, protein, and antibody-based probes for fluorescence (19, 20), magnetic resonance (21, 22), ultrasound (23), single-photon emission computed tomography (24, 25), and positron emission tomography (PET; refs. 26–37) imaging of integrin α_vβ₃ expression in vivo. To date, most of integrin α_vβ₃ targeted PET studies have been focused on the radio labeling of arginine-glycine-aspartic acid (RGD) peptide antagonists of integrin α_vβ₃ due to their high-binding affinity (31, 33, 37, 38). [¹⁸F]Galacto-RGD has been tested in healthy volunteers and cancer patients (33, 35). Very recently, we reported that [¹⁸F]FRGD2 can be used to quantify tumor integrin α_vβ₃ expression level in vivo with static PET scans in xenograft tumor models (29, 37). Both peptide and antibody-based imaging can help in evaluating integrin α_vβ₃ expression level. RGD peptide-based probes are useful in planning and monitoring peptide-based therapeutics due to the similar pharmacokinetics, whereas antibody-based tracer is more relevant to antibody-based therapy. The relatively longer t_1/2 of ⁶⁴Cu [t_1/2 = 12.7 hours; β⁺ = 655 keV (17%); β⁻ = 573 keV (39%)] is well suited for the mAb labeling and imaging. To date, no imaging using Abegrin^TM has been reported. Here, we report the first ⁶⁴Cu-labeled Abegrin^TM through 1,4,7,10-tetra-azacylododecane N,N′,N″,N‴-tetraacetic (DOTA) chelator and the in vitro as well as in vivo characterizations of the resulting ⁶⁴Cu-DOTA-Abegrin^TM.

All commercially available chemical reagents were used without further purification. DOTA was purchased from Macrocyclics, Inc. (Dallas, TX). 1-Ethyl-3-[3-(dimethylamino)-propyl] carbodiimide (EDC), N-hydroxysulfonosuccinimide (SNHS), and Chelex 100 resin (50-100 mesh) were purchased from Sigma-Aldrich (St. Louis, MO). Water and all buffers were passed through Chelex 100 column before use in radiolabeling procedures to ensure that the aqueous buffer is heavy metal-free. The syringe filter, polyethersulfone membranes (pore size, 0.2 μm; diameter, 13 mm) were obtained from Nalge Nunc International (Rochester, NY). ¹²⁵I-Echistatin and PD-10 desalting column were purchased from GE Healthcare (Piscataway, NJ). Female athymic nude mice were supplied from Harlan (Indianapolis, IN) at 4 to 5 weeks of age. ⁶⁴Cu was obtained from Washington University (St. Louis, MO) and University of Wisconsin (Madison, WI). ⁶⁴Cu was produced using the ⁶⁴Ni(p,n)⁶⁴Cu nuclear reaction and supplied in high specific activity as ⁶⁴CuCl₂ in 0.1 mol/L HCl.

DOTA conjugation and radiolabeling. DOTA was activated by EDC and SNHS at pH 5.5 for 30 minutes with a molar ratio of 10:5:4 (DOTA/EDC/SNHS). Without purification, the DOTA-N-hydroxysulfosuccinimidyl (OSSu) was cooled to 4°C and added to Abegrin at five different molar ratios (DOTA-OSSu/Abegrin^TM of 20:1, 50:1, 100:1, 200:1, and 1,000:1). The reaction mixture was adjusted to pH 8.5 with 0.1 N of NaOH and allowed to incubate for overnight at 4°C. The DOTA-Abegrin conjugates were then purified by PD-10 column and concentrated by Centricon filter (Millipore, Bedford, MA), and the final concentration was measured based on UV absorbance at 280 nm using unconjugated Abegrin^TM of known concentrations as standard. ⁶⁴CuCl₂ (2 mCi) was diluted in 300 μL of 0.1 mol/L sodium acetate buffer (pH 6.5) and added to the DOTA-Abegrin^TM conjugates (25 μg DOTA-Abegrin^TM/mCi of ⁶⁴Cu). The reaction mixture was incubated for 1 hour at 40°C with constant shaking. The ⁶⁴Cu-DOTA-Abegrin^TM conjugates were then purified by PD-10 column using PBS as the mobile phase. The radioactive fractions containing ⁶⁴Cu-DOTA-Abegrin^TM was collected and passed through a 0.2-μm syringe filter for further in vitro and in vivo experiments.

Number of DOTA per Abegrin^TM and immunoreactivity. The average number of DOTA chelators per Abegrin^TM antibody was determined using a previously reported procedure with slight modifications (39). Briefly, a defined amount of nonradioactive CuCl₂ (80-fold excess of DOTA-Abegrin^TM) in 40 μL 0.1 N sodium acetate (NaOAc) buffer (pH 6.5) was added to 1.0 mCi ⁶⁴CuCl₂ in 20 μL 0.1 N NaOAc buffer, and 20 μg of each DOTA-Abegrin^TM conjugate in 90 μL 0.1 N NaOAc buffer were added to the above carrier-added ⁶⁴CuCl₂ solution. The reaction mixture was incubated with constant shaking at 40°C for 1 hour. The ⁶⁴Cu-DOTA-Abegrin^TM was purified using PD-10 column and the radiolabeling yield was calculated. The number of DOTA per Abegrin^TM was calculated using the following equation: number of DOTA per Abegrin^TM = moles (Cu²⁺) × yield/moles (DOTA-Abegrin^TM). The results were expressed as mean ± SD (n = 3). An isotype control IgG (against a bacterial antigen with no cross-reaction with human or mouse integrin α_vβ₃; supplied by MedImmune) was conjugated with DOTA under the same condition and the number of DOTA per IgG was also determined as described above.

The immunoreactivity of each ⁶⁴Cu-DOTA-Abegrin^TM conjugate (with different number of DOTA per Abegrin^TM) was determined by incubating with integrin α_vβ₃-positive U87MG cells in suspension culture under conditions of antigen excess (n = 3; ref. 40). The immunoreactivity was calculated as follows: immunoreactivity = bound activity on cells/total added activity × 100%.

Cell lines and cell integrin receptor assay. Four cell lines were used for in vitro and in vivo experiments. U87MG human glioblastoma, MDA-MB-435 human breast cancer carcinoma, and PC-3 human prostate adenocarcinoma cell lines were obtained from American Type Culture Collection (Manassas, VA). The GL-26 mouse glioblastoma cell line was kindly provided by Dr. Victor K. Tse (Department of Neurosurgery, Stanford University, Stanford, CA). All culture media were obtained from Invitrogen Corp. (Carlsbad, CA). U87MG cells were grown in DMEM (low glucose), MDA-MB-435 cells were grown in Leibovitz's L-15 Medium, PC-3 cells were grown in F-12K nutrient mixture (Kaighn's Modification), and GL-26 cells were grown in DMEM (high glucose). All cell lines were cultured in medium supplemented with 10% (v/v) fetal bovine serum at 37°C in a humidified atmosphere with 5% CO₂, except for MDA-MB-435, which was cultured without CO₂. The procedure of cell integrin receptor assay has been reported earlier (29, 41).

Flow cytometry. Cells were collected and washed with PBS and incubated with Abegrin^TM (20 μg/mL) in PBS supplemented with 1% bovine serum albumin (BSA) for 30 minutes at 4°C. After washing with PBS containing 1% BSA, the cells were incubated with FITC donkey anti-human IgG (1:100; Jackson ImmunoResearch Laboratories, Inc., West Grove, CA) for 30 minutes at 4°C. The cells were washed again, resuspended in PBS, and analyzed using a LSR model 1A analyzer (Becton Dickinson, Heidelberg, Germany) and FlowJo analysis software (Tree Star, Inc., Ashland, OR).

Animal models. Animal procedures were done according to a protocol approved by Stanford University Institutional Animal Care and Use Committee. The MDA-MB-435 breast cancer model was established by orthotopic injection of 5 × 10⁶ cells (in 50 μL PBS) into the left mammary fat pad. The U87MG, PC-3, and GL-26 tumor models were obtained by s.c. injection of the corresponding cells (5 × 10⁶ in 50 μL PBS) into the right front leg of the mice (male mice were used for PC-3 tumor). The mice were subjected to microPET imaging and biodistribution studies when the tumor volume reached 200 to 500 mm³ (2 weeks after inoculation for GL-26 tumor model; 3-4 weeks after inoculation for U87MG, MDA-MB-435, and PC-3 tumor models).

MicroPET imaging studies. PET imaging of tumor-bearing mice was done on a microPET R4 rodent model scanner (Concorde Microsystems, Knoxville, TN). The scanner has computer controlled vertical and horizontal bed motion, with an effective axial field of view (FOV) of 7.8 cm and transaxial FOV of 10 cm. The tumor-bearing mice were imaged in prone position in the microPET scanner. The mice were injected with 200 to 300 μCi ⁶⁴Cu-DOTA-Abegrin^TM via the tail vein, anesthetized with 2% isoflurane, and placed near the center of the FOV where the highest image resolution and sensitivity is available. Three- to five-minute static scans were done before 24-hour postinjection, whereas 10- to 20-minute static scans were carried out after 24-hour postinjection For each microPET scan, three-dimensional regions of interests (ROI) were drawn over the tumor, heart, liver, kidneys, and muscle on decay-corrected whole-body coronal images. The average radioactivity concentration (accumulation) within a tumor or an organ was obtained from mean pixel values within the ROI volume, which were converted to counts per milliliter per minute by using a conversion factor. Assuming a tissue density of 1 g/mL, the counts per milliliter per minute were converted to counts per gram per minute and then divided by the injected dose (ID) to obtain an imaging ROI-derived percent injected dose per gram of tissue (% ID/g) of tissue. A mouse bearing a MDA-MB-435 tumor was also imaged using ⁶⁴Cu-DOTA-Abegrin^TM coinjected with 2 mg of unconjugated Abegrin. As a control experiment, ⁶⁴Cu-DOTA-IgG (200-300 μCi) was injected into U87MG tumor-bearing mice and microPET scans were done as described above.

Biodistribution studies. Female nude mice bearing MDA-MB-435 tumors were injected with about 20 μCi ⁶⁴Cu-DOTA-Abegrin^TM. The mice were sacrificed and dissected at 18-, 44-, and 68-hour postinjection. Blood, tumor, major organs, and tissues were collected and wet weighed. The radioactivity in the tissues was measured using a gamma counter (Packard, Meriden, CT). The results were presented as %ID/g. For each mouse, the radioactivity of the tissue samples was calibrated against a known aliquot of the injectate and normalized to a body weight of 20 g. Values are presented as mean ± SD for a group of three animals.

Radiation dosimetry extrapolation to human. Estimated human dosimetry was calculated from microPET imaging results on Sprague-Dawley female rats (Harlan) injected with ⁶⁴Cu-DOTA-Abegrin^TM. The rats were scanned at two bed positions to cover the whole body and ROI analysis were carried out on major organs. Time-activity curves were generated from the mean values obtained in rats for each organ of interest. We then calculated source organ residence times for the human model by integrating a monoexponential fit to the experimental biodistribution data for major organs (heart, lung, liver, kidneys, and spleen) and the whole body. The source organ residence times obtained forthwith were used with a standard quantitation platform organ level internal dose assessment (Vanderbilt University, Nashville TN; ref. 42).

Immunofluorescence staining. Frozen tumor sections were warmed to room temperature, fixed with ice-cold acetone for 10 minutes, and dried in the air for 30 minutes. The sections were blocked with 10% donkey serum for 1 hour at room temperature. For CD31 and human integrin α_vβ₃ double staining, the sections were incubated with rat anti-mouse CD31 (1:100; BD Biosciences, San Jose, CA) and Abegrin^TM (100 μg/mL) for 1 hour at room temperature. After incubating with Cy3-conjugated donkey anti-mouse secondary antibody (1:200; Jackson ImmunoResearch Laboratories) and FITC-conjugated donkey anti-human secondary antibody (1:200), the tumor sections were examined under the microscope (Carl Zeiss Axiovert 200M, Carl Zeiss, Thornwood, NY). For CD31 and mouse integrin β₃ double staining, hamster anti-mouse β₃ (1:100) and FITC-conjugated goat anti-hamster secondary antibody (1:400; Jackson ImmunoResearch Laboratories) were used.

Statistical analysis. Quantitative data were expressed as mean ± SD. Means were compared using one-way ANOVA and Student's t test. Ps < 0.05 were considered statistically significant.

Number of DOTA per Abegrin^TM, radiolabeling yield, specific activity, and immunoreactivity. Five different reaction ratios were tested for DOTA-Abegrin^TM conjugation (Table 1). For the DOTA/Abegrin^TM ratio of 20:1, 50:1, 100:1, 200:1, and 1,000:1 used for conjugation reaction, the number of DOTA per Abegrin^TM was determined to be between 1.65 ± 0.32 and 38.53 ± 5.71. The reaction conditions and the amount of ⁶⁴Cu and each DOTA-Abegrin^TM conjugate used for the radiolabeling were quite similar. Because the radiolabeling yield varied from 5.20 ± 3.16% to 88.12 ± 6.98%, the specific activity of the resulting ⁶⁴Cu-DOTA-Abegrin^TM also varied from 2.11 ± 1.28 to 25.89 ± 2.05 mCi/mg Abegrin^TM. Although the number of DOTA per Abegrin^TM was quite different for the five conjugates, the immunoreactivity was similar and all of them were about 60% to 70%, suggesting that the accessible lysine residues were largely away from the complementarity-determining region. ⁶⁴Cu-DOTA-IgG was determined to have 24.08 ± 0.42 DOTA residues per IgG (n = 3) and the ⁶⁴Cu-labeling yield was about 85 % under similar conditions.

MicroPET studies of ⁶⁴Cu-DOTA-Abegrin^TM conjugates in U87MG tumor-bearing mice. As can be seen in Fig. 1A, the tumor uptake of ⁶⁴Cu-DOTA-Abegrin^TM (1,000:1) was significantly higher than ⁶⁴Cu-DOTA-Abegrin^TM (100:1; P < 0.05 at all time points examined). All five conjugates with different DOTA/Abegrin ratio exhibited good tumor contrast (data not shown for the other three conjugates). Note that the scale for ⁶⁴Cu-DOTA-Abegrin^TM (1,000:1) was 0 to 60 %ID/g and the scale for ⁶⁴Cu-DOTA-Abegrin^TM (100:1) was 0 to 25 %ID/g. MicroPET imaging of the same tumor model with ⁶⁴Cu-DOTA-IgG did not give significant tumor uptake above the background (P < 0.05 at all time points when compared with ⁶⁴Cu-DOTA-Abegrin^TM; 1,000:1). The U87MG tumor uptake of ⁶⁴Cu-DOTA-Abegrin^TM (1,000:1) increased with time and reached a plateau at about 24-hour postinjection. The tumor uptake was 12.50 ± 2.73, 39.88 ± 7.05, and 49.41 ± 4.54 %ID/g at 4-, 25-, and 71-hour postinjection, respectively (Fig. 1B). All five conjugates exhibited high uptake in the heart (due to blood pool activity) and liver (due to the nonspecific uptake of the tracer by the reticuloendothelial system) at early time points, whereas the tracer uptake in all the other organs was at the background level. No significant difference was observed between the five conjugates in either the uptake level of the tracer in the major organs or the washout pattern. As can be seen in Fig. 1C, the tracer uptake in both the heart and the liver dropped steadily over time. For the heart, the uptake was 29.08 ± 5.90, 13.89 ± 3.13, 8.45 ± 3.59, and 4.38 ± 0.82 %ID/g at 1-, 17-, 47-, and 71-hour postinjection, respectively, indicating that the serum t_1/2 of the tracer was about 12 to 24 hours in the mouse. For the liver, the uptake was 19.56 ± 4.43, 11.91 ± 1.68, 10.24 ± 2.05, and 8.51 ± 1.31 %ID/g at 1-, 17-, 47-, and 71-hour postinjection, respectively. The small fraction of ⁶⁴Cu dissociated from the DOTA chelator did not cause appreciable increase of liver uptake over the time course of this study. Based on the immunoreactivity and microPET imaging results, ⁶⁴Cu-DOTA-Abegrin^TM (1,000:1) was used for all the following studies and it will be referred to as ⁶⁴Cu-DOTA-Abegrin^TM in the following text.

Fluorescence-activated cell sorting analysis and cell integrin α_vβ₃ expression level. Fluorescence-activated cell sorting (FACS) analysis using Abegrin as the primary antibody against integrin α_vβ₃ clearly showed that U87MG, PC-3, and MDA-MB-435 cells are integrin α_vβ₃ positive (Fig. 2A). Although GL-26 cells are mouse integrin α_vβ₃ positive (where anti-mouse β₃ mAb was used as the primary antibody; Fig. 2A,, bottom, right), it is not recognized by Abegrin, which showed that Abegrin^TM is human integrin α_vβ₃ specific and does not cross-react with mouse α_vβ₃. Next, we quantified the α_vβ₃ integrin expression level of the four cell lines through competitive cell-binding assay using ¹²⁵I-echistatin as radioligand. The density of integrin α_vβ₃ on the cell surface follows the order U87MG > MDA-MB-435 > GL-26 ≈ PC-3 (Fig. 2B). Note that echistatin binds to integrin α_vβ₃ as one entity but not integrin α_v or integrin β₃ alone and it recognizes both human and mouse integrin α_vβ₃.

MicroPET imaging of tumor models with different integrin α_vβ₃ expression level. The localization of ⁶⁴Cu-DOTA-Abegrin^TM in all four tumor models was evaluated by multiple time point microPET imaging. Selected coronal images at different time points postinjection were shown in Fig. 3A. At 4-hour postinjection, the tumor was already visible. The tumor signal at this early time point may be due to both specific targeting and nonspecific targeting because of the enhanced permeability and retention (EPR) effect, as tumors have abnormal and leaky vasculature and lack lymphatic drainage. High tumor activity accumulation was observed as early as 17-hour postinjection The tumor signal of U87MG-, PC-3-, and MDA-MB-435-bearing mice increased over time (Fig. 3B), indicating specific binding between Abegrin and human integrin α_vβ₃. The tumor signal intensity was similar for MDA-MB-435 and PC-3 tumors, whereas the U87MG tumor uptake is much higher (Fig. 1B), agreeing with the cell integrin receptor level. The fact that the GL-26 tumor signal intensity reached a peak at 24-hour postinjection and steadily dropped afterwards is indicative of passive targeting due to the EPR effect, corroborating with the FACS analysis data because Abegrin does not cross-react with mouse integrin α_vβ₃. As all tumor models in this study are murine xenografts, the integrin α_vβ₃ expressed on the tumor vasculature is of mouse origin and is not recognized by Abegrin^TM. The uptake profile in the heart, liver, kidneys, and muscle were similar in all four tumor models.

Immunofluorescence staining. After the microPET studies and the radioactivity mostly decayed, the mice were sacrificed and frozen tumor sections were stained for CD31, human integrin α_vβ₃, and mouse integrin β₃. Only the images of U87MG and GL-26 tumors were shown. As can be seen from Fig. 4A, the CD31 and mouse β₃ staining colocalizes very well, indicating that the tumor vasculature expresses mouse integrin α_vβ₃. For the GL-26 tumor, both the tumor vasculature and the GL-26 tumor cell express mouse integrin α_vβ₃ as evidenced by the staining (anti-mouse β₃ antibody stains both the tumor cells and the tumor vasculature). When using Abegrin as the primary antibody, there is no colocalization with CD31 for the U87MG tumor because only the human tumor cells but not the mouse tumor vasculature are recognized by Abegrin^TM (Fig. 4B). For GL-26 tumor, no Abegrin staining was visible as neither the tumor vasculature nor the tumor cells express human integrin α_vβ₃. Ex vivo staining experiments again confirmed the human integrin α_vβ₃-specific binding of Abegrin.

Blocking experiment and biodistribution studies. Blocking studies were carried out in a MDA-MB-435 tumor-bearing mouse by coinjecting ⁶⁴Cu-DOTA-Abegrin^TM with 2 mg Abegrin^TM. Serial microPET imaging clearly showed much lower tumor uptake of the tracer under blocking conditions (Fig. 5A). Effective blocking showed the integrin α_vβ₃-specific binding of ⁶⁴Cu-DOTA-Abegrin^TM. Biodistribution of ⁶⁴Cu-DOTA-Abegrin^TM was also carried out in the MDA-MB-435 tumor model at 18-, 44-, and 68-hour postinjection (Fig. 5B). It can be seen that the major tracer clearance is through the liver and the spleen. All other organs had background level uptake of the tracer, providing good tumor contrast of the tracer. The MDA-MB-435 tumor uptake was 4.54 ± 0.39 %ID/g based on biodistribution studies and 5.63 ± 1.16 %ID/g based on microPET studies at 18-hour postinjection (P = 0.17). At 46-hour postinjection, the MDA-MB-435 tumor uptake was 5.40 ± 0.75 %ID/g based on biodistribution experiments and 7.04 ± 1.47 %ID/g based on microPET studies, respectively (P = 0.16). Comparing the quantification results of the tracer uptake between microPET and biodistribution studies, the uptake in all the other organs follow similar pattern, indicating that the quantification methods used for the noninvasive microPET study is a true reflection of the tracer uptake in vivo.

Radiation dosimetry. Human absorbed doses to normal organs from ⁶⁴Cu-DOTA-Abegrin^TM were estimated from microPET imaging data in Sprague-Dawley rats and presented in Table 2. The tracer uptake of different organs in rats is similar to that of the mice, except that the liver uptake in the rats appears to be more prominent and persistent (Fig. 5C). Except for the heart, liver, and spleen, all other organs exhibited background level of tracer uptake at all time points examined. There is no significant kidney uptake or renal excretion of the mAb-based tracer due to the high molecular weight (150 kDa). These results predict that the highest radiation-absorbed doses will be to the liver (0.33 ± 0.02 mGy/MBq) and spleen (0.21 ± 0.01 mGy/MBq). The whole-body absorbed dose was found to be 0.034 ± 0.001 mGy/MBq. Although the radiation doses to liver and spleen are higher than other ⁶⁴Cu-labeled mAb reported in the literature (43), the doses to other major organs (kidneys, lungs, intestines, etc.) are much lower.

This study shows that ⁶⁴Cu-labeled Abegrin^TM, a humanized mAb against human integrin α_vβ₃, exhibits high integrin α_vβ₃ specificity in vitro and in vivo. As Abegrin^TM is currently in phase I/II clinical trials and soon to enter phase III, PET imaging is critical in early and sensitive lesion detection, patient selection for clinical trials based on in vivo integrin expression quantification, better treatment monitoring and dose optimization, and elucidation of the mechanisms of treatment efficacy underlying integrin signaling.

We have optimized the DOTA/Abegrin^TM ratio for better in vivo targeting and maximized tumor uptake. As the equivalence of DOTA used for conjugation reaction increases, the number of DOTA residues per Abegrin^TM, the specific activity of the tracer, as well as the absolute tracer uptake in the tumor all increases. Because the difference in immumoreactivity is minimal between the five tracers, the conjugate with the most DOTA residues per Abegrin^TM is more suitable for future clinical translation. The fact that DOTA conjugation of Abegrin^TM does not deteriorate the receptor affinity of the antibody indicates that the accessible lysine residues are mostly located away from the antigen-binding site. High specific activity may also play a role here. It has been reported previously that a dendrimer-mAb conjugate with high specific activity exhibited higher tumor uptake (44). Although not tested, we envision that Abegrin coupled to dendrimeric DOTAs may give comparable or potentially even higher tumor uptake. Due to the relatively high molecular weight of the mAb (150 kDa), the tracer exhibited mainly hepatic clearance. Intact IgG is designed to remain in the bloodstream for many weeks, which is of considerable advantage when acting in the natural setting of removing foreign entities. The serum t_1/2 of ⁶⁴Cu-DOTA-Abegrin^TM in mouse is about 12 to 24 hours based on microPET studies, much shorter than the t_1/2 in humans (5-8 days; ref. 17), presumably due to the much faster metabolic rate of mouse compared with human. In this study, we followed the tracer uptake for about 72 hours (almost six t_1/2 of ⁶⁴Cu). Longer term monitoring may be necessary for future studies and ⁸⁹Zr with a t_1/2 of 79.3 hours can be used. It has been reported that ⁸⁹Zr imaging might be a good reflection of ⁹⁰Y distribution (45), which may shed light on future radioimmunotherapy studies using ⁹⁰Y-Abegrin^TM conjugates.

For antiangiogenic treatment using Abegrin alone or in combination with other treatment modalities, the long circulation t_1/2 is advantageous because less administration is needed to maintain serum levels of the antibody. The very high tracer uptake in the tumor (up to 49.41 ± 4.54 %ID/g in the U87MG tumor) along with minimal signal in the nontargeting organs warrants future radioimmunotherapy studies using DOTA-Abegrin^TM, as therapeutic radioisotopes (⁶⁷Cu, ¹⁷⁷Lu, ⁹⁰Y, etc.) can be incorporated using a similar strategy. Previous report has shown prominent expression of integrin α_vβ₃ by glioma cells and vasculature in cancer patients, especially in the periphery of high-grade gliomas (46). Systematic investigation of α_vβ₃ expression on tumor-associated vessels in cancer patients also revealed that a considerable number of colon, pancreas, lung, and breast carcinoma lesions have many α_vβ₃-expressing vessels that could be targets for anti-integrin α_vβ₃ therapy (47). Clinical translation of ⁶⁴Cu-DOTA-Abegrin^TM will be critical for the maximum benefit of Abegrin-based anticancer agents as imaging can provide a straightforward and convenient way to monitor the biological changes at the molecular level in vivo.

Abegrin is a humanized antibody, which recognizes only human but not murine integrin α_vβ₃. Both the microPET imaging studies and the immunofluorescence staining confirmed that Abegrin^TM does not bind to mouse tumor vasculature. Therefore, the difference in tumor uptake is a good reflection of the integrin α_vβ₃ expression level of different tumor cells. For GL-26 tumor, which only expresses murine integrin α_vβ₃, the tracer uptake is mostly nonspecific. As can be seen from Fig. 3B, the tracer uptake in the tumor reached maximum at about 24-hour postinjection and dropped steadily afterwards. This is characteristic of passive targeting due to the EPR effect of the tumor. For all the other tumors, the tracer uptake increased or remained steady over time, which is characteristic of specific targeting. In this study, the passive targeting of GL-26 is quite prominent and, at its peak, the uptake is even higher than integrin α_vβ₃-positive tumors MDA-MB-435 and PC-3. PET imaging of other tumors (e.g., C6 rat glioma) that are human/mouse integrin α_vβ₃ negative revealed a similar tracer uptake profile as the GL-26 tumor (data not shown), although the time point of maximum tumor uptake varied between these tumors because of the difference in vascular density and leakiness. The tracer uptake, tracer clearance, and the pharmacokinetics of ⁶⁴Cu-DOTA-Abegrin^TM are quite different from our previously reported RGD peptide-based imaging (29, 31, 48). The passive targeting of this antibody may provide a more general use in cancer therapy as most tumors have leaky vasculature, which may result in more uptake of Abegrin^TM-based agents and greater efficacy.

We realize that the pharmacokinetics of ⁶⁴Cu-DOTA-Abegrin^TM in the present study may not be a true mimicry of the clinical situation because Abegrin does not bind mouse integrin α_vβ₃. This may also be partially responsible for the faster tracer clearance. For future studies, other rodent models, such as hamsters or primates (Abegrin^TM cross-reacts with hamster and certain primate integrin α_vβ₃), may be used to obtain dosimetry instead of mice and rats. Transgenic mouse models that express both human and mouse integrin α_vβ₃ may also be relevant. Due to both the heterogeneity of the tumor itself and the weaker tissue penetration of the mAb-based tracer because of its large molecular weight, the tracer uptake in the tumor is also somewhat heterogeneous. As can be seen from Fig. 1A, the tracer uptake of the U87MG tumor is higher in the periphery region of the tumor and significantly lower in the center of the tumor, although ex vivo examination revealed that the tumor center was not necrotic.

In summary, we describe here the in vitro and in vivo characterization of ⁶⁴Cu-labeled Abegrin^TM, a humanized mAb against human integrin α_vβ₃, in various tumor models. No significant difference in immunoreactivity was found among the five ⁶⁴Cu-DOTA-Abegrin^TM conjugates tested. MicroPET imaging of ⁶⁴Cu-DOTA-Abegrin (1,000:1) showed U87MG tumor uptake as high as 49.41 ± 4.54 %ID/g at 71-hour postinjection. The success of human integrin α_vβ₃-specific tumor imaging using ⁶⁴Cu-DOTA-Abegrin^TM may be translated into the clinic to evaluate the pharmacokinetics, tumor targeting efficacy, dose optimization, and dose interval of Abegrin and Abegrin-based cancer therapeutics. A certain level of nonspecific targeting due to the leaky vasculature of the tumors may also have potential applications in radioimmunotherapy in general.

Grant support: MedImmune, National Institute of Biomedical Imaging and Bioengineering grant R21 EB001785, National Cancer Institute (NCI) grant R21 CA102123, NCI In Vivo Cellular Molecular Imaging Center grant P50 CA114747, NCI Small Animal Imaging Resource Program grant R24 CA93862, NCI Centers of Cancer Nanotechnology Excellence U54 grant 1U54CA119367-01, Department of Defense (DOD) Breast Cancer Research Program IDEA Award W81XWH-04-1-0697, DOD Ovarian Cancer Research Program Award OC050120, DOD Prostate Cancer Research Program New Investigator Award DAMD1717-03-1-0143, and Education and Research Foundation of the Society of Nuclear Medicine Benedict Cassen Postdoctoral Fellowship (W. Cai).

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.

We thank Ms. Pauline Chu and Dr. Xianzhong Zhang for their excellent technical support and the cyclotron teams of Washington University and University of Wisconsin for ⁶⁴Cu production.