Combination Therapy Enhances the Inhibition of Tumor Growth with the Fully Human Anti–Type 1 Insulin-Like Growth Factor Receptor Monoclonal Antibody CP-751,871

INTRODUCTION

The insulin-like growth factors IGF-I and IGF-II play important roles in cellular physiology, especially in the regulation of cell growth, differentiation, and survival. Both IGFs act by binding to the type 1 IGF receptor (IGF-1R) and thereby activating its tyrosine kinase activity. Upon activation, the IGF-1R transduces signals through a variety of substrate effector proteins, including insulin receptor substrate 1 (IRS-1) and Shc. Extensive studies (1, 2) have elucidated the mechanism by which IGF-1R signals ultimately activate the phosphatidylinositol-3 kinase pathway and protein kinase B (PKB/Akt) and indicate that cell survival is enhanced by IGFs.

IGF-1R is also crucial in tumor biology and malignant transformation. Fibroblasts lacking IGF-1R because of targeted disruption cannot be transformed by most oncogenes, but susceptibility to transformation is restored if IGF-1R expression is provided by transfection (3). Numerous studies have linked higher plasma concentrations of IGF-I with increased cancer risk (1), especially increased risk of prostate and colorectal cancer (4, 5). Also, many tumors and tumor cell lines have increased expression of IGF-I or IGF-1R (2). In addition to a role in transformation of cells, continued signals from IGF-1R are important for tumor cell proliferation and survival. Interruption of those signals by various means produce antiproliferative effects in cell culture and antitumor activity in tumor xenograft models. Down-regulation of IGF-1R by antisense inhibits growth of melanoma cells in soft agar and growth of these cells as xenografts in athymic mice (6). Similar results were found for glioblastoma cells (7) and breast cells (8). Blocking IGF-1R signals with antagonistic IGF-I peptide analogues inhibit the growth of prostate cancer cell lines (9). Expression of a dominant-negative form of IGF-1R, introduced by transfection into HT29 human colon cancer cells, virtually eliminated growth of these cells in soft agar and greatly slowed growth of the cognate tumors in athymic mice (10). A secreted truncated (extracellular region) IGF-1R was shown to block IGF-1R autophosphorylation, induce tumor cell apoptosis, and inhibit the formation of tumors from C6 rat glioblastoma cells (11). Thus, multiple independent approaches indicate that inhibition of IGF-1R signals may be an effective approach to cancer therapy.

Monoclonal antibodies that prevent IGF-1 binding to IGF-1R have also been used to inhibit IGF-1R function. One such antibody, the murine monoclonal antibody αIR-3, identified almost 20 years ago, inhibits binding of IGF-I to MDA-231 human breast cancer cells and inhibits the growth of these cells in soft agar and as tumor xenografts (12–14). αIR-3 also had antitumor activity against multiple tumor xenografts (15–18). These experiments support the applicability of anti-IGF-1R antibodies in chemotherapy, but the expected human anti-mouse antibody response to αIR-3 would limit the use of this agent in the clinic.

Human monoclonal antibodies can be generated by use of the “XenoMouse,” in which the human immunoglobulin loci was transplanted into the genome of mice with an inactivated endogenous immunoglobulin locus (19). This approach has been successfully used to generate human therapeutic antibodies (e.g., an antibody directed against the epidermal growth factor receptor; ref. 20). We report here the discovery and characterization of a human monoclonal antibody that binds the human IGF-1R with high affinity and specificity, blocks IGF-I binding, and prevents activation of IGF-1R. This antibody induces a depletion of cellular IGF-1R protein, inhibits tumor growth as a single agent and enhances the efficacy of other cancer agents in human tumor xenograft models.

MATERIALS AND METHODS

CP-751,871. Abgenix's XenoMouse technology (19, 20) was used to produce fully human IgG2 antibodies from mice immunized with human IGF1R extracellular domain. B cells from immunized XenoMouse were then fused with a mouse myeloma cell line (P3X63-Ag8.653) to produce hybridoma lines that could be propagated in bioreactors.

Cell Lines. The Colo-205 human colorectal carcinoma cell line, MCF7 breast carcinoma, and RPMI multiple myeloma cell lines were obtained from the American Type Culture Collection (Manassas, VA) repository. The 3T3-IGF-1R cells are 3T3 cells transfected with full length human IGF-1R (kindly provided by Dr. Renato Baserga, Kimmel Cancer Center at Thomas Jefferson University, Philadelphia, PA). Colo-205 cells were maintained using DMEM (Invitrogen, Carlsbad, CA) supplemented with 10% heat inactivated FSC (Invitrogen), 2 mmol/L l-glutamine (Invitrogen), and penicillin-streptomycin (Invitrogen). 3T3-IGF-1R cells were maintained using the same medium as Colo-205, supplemented with 0.5 mg/mL Geneticin (Invitrogen). MCF7 cells were maintained using DMEM F12 (Invitrogen) supplemented with 10% heat inactivated FCS, 2 mmol/L l-glutamine, and penicillin-streptomycin.

[¹²⁵I]-IGF-1 Cell Binding Assay. 3T3-IGF-1R cells were plated in 96-well plates, in full medium. Following overnight incubation, the cells were changed to serum-free DMEM and incubated for 1 hour. CP-751,871 was serially diluted in serum-free DMEM. Medium was decanted from the cell plates, and 50 μL of each antibody dilution added to the cell plate. The plate was incubated at 37°C for 10 minutes. [¹²⁵I]-IGF-I (NEN, Wilmington, DE) was diluted in serum-free DMEM to 1 μCi/mL. Fifty microliters were added per well of the cell plate. Addition of 100 ng/mL cold IGF-1 (R&D Systems, Minneapolis, MN) was used as control for background. The cell plate was incubated at 37°C for 10 minutes. Medium was decanted and the plates washed twice with serum-free DMEM. The cells were lysed in 0.1 N NaOH + 0.1% SDS, then transferred to a scintillation plate, mixed with OptiPhase Supermix liquid scintillant (Wallac, Shelton, CT), and cell-associated [¹²⁵I]-IGF-1 quantitated on a Wallac Micro-Beta counter.

IGF-1-R Autophosphorylation. Capture ELISA plates were prepared by incubating Reacti-Bind Goat anti-rabbit plates (Pierce, Rockford, IL) with anti-IGF-1R antibody (Santa Cruz Biotech, Santa Cruz, CA), diluted in 5% bovine serum albumin (Sigma, St. Louis, MO)/TBST [50 mmol/L Tris (pH 8.0), 138 mmol/L NaCl, 2.7 mmol/L KCl, and 0.1% Tween 20]. Plates were incubated while shaking at room temperature for 1 hour. Plates were washed with TBST before addition of the cell lysates.

3T3-IGF-1R cells were plated in 96-well plates in growth medium (see above) and incubated overnight at 37°C, 5% CO₂. Media was decanted from the cell plates and 100 μL fresh growth medium added. CP-751,871 to the indicated final concentration was added to the cell plate (all samples were in triplicate). The plates were incubated at 37°C for 1 hour. Cells were stimulated by addition of 100 ng/mL IGF-1 (final concentration in growth medium) and incubated at room temperature for 10 minutes. Media was decanted from the plates. Cells were lysed in 50 μL lysis buffer [50 mmol/L HEPES (pH 7.4), 10 mmol/L EDTA, 150 mmol/L NaCl, 1.5 mmol/L MgCl₂, 1% Triton X-100, 1.6 mmol/L NaVO₄, and 1% glycerol], and lysates transferred to dilution plates; 100 μL dilution buffer [50 mmol/L HEPES (pH 7.4) and 1.6 mmol/L NaVO₄] were added to each lysate, mixed, and 100 μL of each diluted lysate were transferred to the ELISA capture plate.

The ELISA plates were incubated shaking gently at room temperature for 2 hours, washed with wash buffer, then treated with HRP-PY20 (BD Transduction Labs, Lexington, KY) diluted in bovine serum albumin/TBST and incubated at room temperature for 30 minutes with shaking. The ELISA plates were washed, color developed by addition of 3,3′,5,5′ tetramethylbenzidine substrate (Calbiochem, San Diego, CA), and stopped with 0.9 mol/LH₂SO₄. IGF-1R phosphotyrosine signal was quantitated by measurement at absorbance 450 nm.

IGF-1R Confocal Analysis. 3T3IGF-1R cells were plated onto 8-well no. 1.5 cover glass chamber slides in a phenol red–free cell culture medium and grown to 80% confluence. LysoTracker Red (1 mmol/L stock solution, Molecular Probes, Eugene, OR) was diluted to 50 nmol/L in each chamber for a period of 2 hours. CP-751,871 (1 μg/mL final culture chamber concentration) was added in a time course fashion to track IGF-1R internalization. At the conclusion of the time course incubation, culture medium was aspirated and cells were quickly washed with fresh warm culture medium. The washed cells were again aspirated and fixed with ice-cold methanol for 20 minutes. Fixed cells were washed with PBS at room temperature, and nonspecific binding was blocked using 1% normal goat serum for 20 minutes at 37°C. CP-751,871 was detected in these cells using a goat anti-human secondary antibody conjugated to Alexa488 (1:100 dilution from 2 mg/mL stock, Molecular Probes) for 60 minutes at 37°C. Cells were again washed thrice in PBS. Coverslips were mounted with VectaShield anti-fade medium containing the DNA dye 4′,6-diamidino-2-phenylindole and sealed with nail polish. Digital images were collected sequentially using the Leica SP laser-scanning confocal microscope using appropriate lasers and spectrophotometer setting matched to the fluorochromes.

Western Blotting. Cells were lysed in lysis buffer (as above) cleared by centrifugation, boiled in Laemmli buffer and the proteins separated on 4-12% Tris/Bis gels (Invitrogen). Western blots were done with anti-IGF1R antibody sc-713 (Santa Cruz). PY54 (OSIP), phospho-AKT (Cell Signaling, Santa Cruz, CA), or actin monoclonal (Sigma) as a loading control, with appropriate secondary antibodies.

Hybrid IGF-1R/IR Detection. IGF-1R and IR were immunprecipitated from MCF7 lysates with either CP-751,871 or a polyclonal rabbit anti-IR antibody PP5 (21). Immune complexes were washed and boiled in SDS dissociation buffer. Denatured complexes were divided and analyzed by Western blot probing for IGF-1R (Santa Cruz) or IR (PP5). In some cases, MCF7 cells were pretreated with CP-751,871 (1 μg/mL) for 15 hours in complete medium under normal growth conditions.

CP-751,871 Quantitation. Serum concentrations of CP-751,871 were determined by an ELISA method. Serum samples (15-20 μL) were diluted in lysis buffer (described above), to be within the linear range of the assay (3-30 ng/mL in assay plate). CP-751,871 was diluted similarly for the assay standards. The immunosorbent assay plates (Costar, Cambridge, MA) were coated with recombinant human IGF-1R extracellular domain (ECD, R&D Systems), and then blocked with 5% bovine serum albumin/TBST. The diluted samples (50 μL) were added to the plates with an equal volume (50 μL) of 10% milk/TBST and incubated with shaking at room temperature for 1 to 2 hours, washed in TBST, and probed with horseradish peroxidase–conjugated anti-human IgG antibody (Jackson ImmunoResearch, West Grove, PA) for 30 minutes. The plates were developed by color reaction with TMB substrate (see above) and stopped with 0.9 mol/L H₂SO₄. Absorbance was determined at a wavelength of 450 nm. The CP-751,871 standard dilutions were used to construct a standard curve. Serum concentrations of CP-751,871 were calculated using linear least square regression with uniform weighting.

Animal Care and Use. Three to 4-week-old female athymic mice (CD-1 nu/nu) were used for tumor xenografts. Mice were obtained from Charles River Laboratories (Wilmington, MA) and were housed in specific pathogen-free conditions according to the guidelines of the Association for Assessment and Accreditation of Laboratory Animal Care. All studies were carried out under approved institutional experimental animal care and use protocols. During these studies, animals were provided pelleted food and water ad libitum. Sentinel mice were monitored routinely at 4-week intervals by serologic assays and were found to be free of exposure to the following agents: murine hepatitis virus, Senai virus, pneumonia virus of mice, minute virus of mice, mouse poliovirus, type 3 Reovirus, Mycoplasma pulmonis, mouse parvovirus, and epizootic diarrhea of infant mice. In addition, the sentinels were monitored on a quarterly basis for lymphocytic choriomeningitis virus, mouse adenovirus, ectromelia, mouse pneumonitis, and polyomavirus. For all studies, the mice were allowed to acclimate for 3 days after receipt of shipment; test animals were randomized before commencement of treatments.

Ex vivo Studies. For tumor implantation, cells (3T3-IGF-1R) were grown in vitro, harvested with Trypsin-EDTA, washed 1X with PBS and resuspended in Matrigel (BD Biosciences, Bedford, MA) to a final concentration of 2.5 × 10⁷ cells/mL. Athymic mice were injected with 0.2 mL of this suspension s.c. in the flank region. Once tumor sizes reached 400 to 500 mm³, the animals were dosed with either CP-751,871 or vehicle (i.p.). CP-751,871 was prepared in Dulbecco's phosphate-buffer saline (Invitrogen). At 24 hours post-dose, animals were euthanized according to Pfizer IACUC guidelines. Tumors were removed and immediately flash frozen in liquid nitrogen. Blood was collected into serum separation tubes and kept on ice. Tumors were stored at −80°C until time of analysis. Serum samples were stored at 4°C until time of analysis.

Tumor extract preparation. 3T3-IGF-1R tumors were homogenized (1 mL lysis buffer supplemented with protease inhibitor cocktail tablets; Roche, Indianapolis, IN; per 100 mg of tumor weight) on ice. The homogenates were centrifuged for clarification at 4°. Each lysate was analyzed for total protein content using the bicinchoninic acid Protein Assay kit (Pierce).

Ex vivo ELISA. Reacti-Bind goat anti-Rabbit coated plates were blocked for 30 minutes by shaking at room temperature with SuperBlock (Pierce). Blocker was aspirated and plates incubated with rabbit polyclonal anti-IGF-1R (Santa Cruz Biotechnology) diluted in SuperBlock. The plates were washed with TBST. Tumor extracts were added to plates and incubated for 1 hour at room temperature. Plates were washed with TBST and incubated with biotinylated goat anti-IGF-1R (R&D Systems) for 1 hour. Plates were washed and probed with streptavidin horseradish peroxidase (Zymed, South San Francisco, CA) for 30 minutes. Plates were washed with TBST and developed by addition of TMB. Signal was stopped by addition of 0.9 mol/L H₂SO₄. Quantitation of signal was measured by absorbance 450 nm. All readings were normalized to the total protein content of each sample.

Tumor Growth Inhibitory Experiments. Treatment was initiated when all mice in the experiment had established tumors of 200 to 250 mm³. CP-751,871 was prepared in PBS. Vehicle control groups were dosed with PBS alone. The general health of the mice was monitored and body weight recorded at the time of tumor measurement. Tumor dimensions (volume = length × width² × 0.5) were recorded thrice/wk, starting with the first day of treatment. Statistical analysis by Student's t test was used to compare treatments. Animals were euthanized according to IACUC guidelines. Antitumor efficacy was measured as the percent tumor growth inhibition.

Monkey Pharmacokinetic Studies. CP-751,871, prepared in acetate formulation buffer [20 mmol/L Na acetate, 140 mmol/L NaCl, and 2 mg/mL polysorbate 80 (pH 5.5)], was given i.v. via cephalic vein to cynomolgus monkeys (1 per sex per dose) naive to CP-751,871 and other monoclonal antibodies at a bolus dose of 3, 30, and 100 mg/kg. Blood (1 mL) was collected from the femoral vein into microtainer serum separator tubes at 0.08, 2, 6, 24, 48, 72, and 168 hours post-dose and then weekly thereafter until the serum concentration of CP-751,871 reached the lower limit of quantitation (0.3 μg/mL) or day 70. Serum was prepared from whole blood by centrifugation, transferred to clean microtubes, and stored at 4°C before analysis. Serum concentrations of CP-751,871 were determined by the ELISA method as described for analysis of mouse serum samples.

Pharmacokinetic Analysis. The pharmacokinetic variables of CP-751,871 were calculated by fitting the data into a two-compartmental model using WinNonlin 3.2 (Pharsight, Corporation, Mountain View, CA): distribution and elimination coefficients and rate constants (A, B, α, and β, respectively) were directly calculated from the program, the distribution and elimination half-lives were calculated as 0.693 / α and 0.693 / β, respectively, and the other pharmacokinetic variables (AUC, CL, and V_dss) were obtained from standard formulas.

Human Lymphocyte Treatment. The expression of IGF-1R on human lymphocytes was tested using samples of five subjects and a flow cytometry assay that employs 1H7 (mouse IgG1, κ; BD-PharMingen, San Diego, CA), an anti-IGF-1R monoclonal antibody conjugated with R-Phycoerythrin, as an analytic reagent. 1H7 effectively detected the expression of IGF-1R in human lymphocytes. Immunophenotyping of T helper and cytotoxic cells, B cells, monocytes and granulocytes using CD3, CD4, CD8, CD19, CD15 markers and forward scatter channel and side scatter channel characteristics was also done to determine cell type specific levels of anti-IGF-1R signal. Range of expression was found to be ∼300 to 500 molecules of equivalent soluble fluorochrome (MESF) for T-helper cells, 250 to 450 MESFs for T cytotoxic cells, 200 to 500 MESFs for B cells, 500 to 600 MESFs for monocytes, and 1,500 to 2,300 MESFs for granulocytes. The ratio of positive cells (cells with detectable IGF-1R expression) was 5% to 10% for lymphocytes, 30% for monocytes, and 90% for granulocytes. Incubation of the blood samples for 24 hours in the presence of CP-751,871 resulted in a concentration-dependent down-regulation of IGF-1R expression; 10⁶ leucocytes were incubated with 10 μL of each analytic antibody per test for 20 minutes at 4°C, washed twice, and acquired using a Becton Dickinson FACSort with CELL Quest v3.1 software. A minimum of 50,000 total cells were analyzed. Cells were identified using a sequential gating strategy, assessing CD8, CD4 or CD19 and IGF1R (Becton Dickinson, Sunnyvale, CA) fluorescence, as well as forward and side light-scatter characteristics). MESF standards were spectrally matched with the fluorochrome used for the analysis.

CP-751,871 was tested for analytic interference with the monoclonal antibody 1H7 used in the determination of surface IGF-1R expression. Following RBC lysis with fluorescence-activated cell sorter lysis buffer (Becton Dickinson, Franklin Lakes, NJ), triplicate aliquots of 20 μL per 10⁶ nucleated cells were added 5 μg/mL of CP-751,871 or PBS and flow cytometry analysis of IGF1R with 1H7 antibody done using a Becton Dickinson FACSCalibur. The percentage of granulocytes stained for IGF-1R in the control (PBS) and CP-751,871 samples were 94.7 ± 2.5% and 91.3 ± 3.2%, respectively. Thus, no significant analytic interference of CP-751,871 on 1H7 was detected.

RESULTS

Characterization of a Fully Human Antibody to Human IGF-1R, Designated CP-751,871. CP-751,871 is a human anti-IGF-1R antibody produced by a hybridoma which was derived by immunization of XenoMice with the extracellular domain of IGF-1R. The properties of this antibody are summarized in Table 1. Sequence analysis determined that CP-751,871 consists of a DP47 heavy chain and A30 light chain. Biacore analysis with immobilized CP-751,871 showed an affinity of 1.5 × 10^-9 mol/L for the human IGF-1R ECD (extracellular domain). CP-751,871 binds to IGF-1R expressed on human, cynomolgus, and rhesus cells with K_d values of 0.6, 0.5, and 0.6 nmol/L, respectively. Immunoprecipitation of IGF-1R by CP-751,871 followed by Western blot analysis showed that CP-751,871 did not bind the rat, dog, rabbit, and marmoset IGF-1R (data not shown). CP-751,871 blocks the binding of iodinated IGF-I to NIH3T3 cells that overexpress the IGF-I receptor (NIH3T3/IGF-1R) with an IC₅₀ of 1.8 nmol/L (0.28 μg/mL; Fig. 1A). Similarly, CP-751,871 inhibited IGF-I induced autophosphorylation of IGF-1R with an IC₅₀ of 0.42 nmol/L (0.08 μg/mL; Fig. 1B), and also blocked IGF-II-induced receptor autophosphorylation (Fig. 2A) CP-751,871 did not show any agonistic activity upon addition to NIH3T3/IGF-1R cells as evaluated by receptor autophosphorylation on Western blots (Fig. 2A). In general, data generated by a variety of assays show a high affinity interaction of CP-751,871 with IGF-1R that antagonizes the binding of IGF-I.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1

CP-751,871 inhibits ¹²⁵I-IGF-I binding and IGF-1R autophosphorylation. A, 3T3/IGF-1R cells were treated with CP-751,871 for 10 minutes and IGF-I binding evaluated as described in Materials and Methods. Points, mean values are plotted (n = 3); bars, ±SE. B, 3T3/IGF-1R cells were incubated for 1 hour at 37° with CP-751,871 followed by stimulation with IGF-I (100 ng/mL) for 10 minutes. Cells were lysed, IGF-1R was captured, and the level of phospho-IGF-1R determined by an ELISA using an horseradish peroxidase–labeled anti-phosphotyrosine antibody as described in Materials and Methods. Points, mean values (n = 3); bars, ±SE.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2

CP-751,871 induces IGF-1R internalization and down-regulation. A, duplicate cultures of 3T3/IGF-1R cells were treated with CP-751,871 (1 μg/mL) or vehicle in complete medium at 37°C for either 1 minute (upper) or 24 hours (bottom). IGF-I or IGF-II (100 ng/mL) was then added as indicated for 10 minutes. Lysates were prepared and IGF-1R, phospho-IGF-1R, phospho-Akt, and actin levels were estimated by Western blotting as described in Materials and Methods. Two additional experiments gave similar results. B, MCF7 and 3T3/IGF-1R cells were treated with CP-751,871 (1 μg/mL) in complete medium at 37°C for designated time periods. Cells were lysed and the level of IGF-1R in total lysates determined by immunoblot analysis. Values below the lanes derived from enhanced chemiluminescence emission analysis. C, 3T3/IGF-1R cells were treated with CP-751,871 (1 μg/mL), fixed, and stained with a goat anti-human Alexa488 secondary antibody. The localization of CP-751,871 was monitored by confocal microscopy. Red, lysosomal staining; blue, DNA; green, anti-human IgG.

CP-751,871 Induces the Down-Regulation of IGF-1R In vitro. CP-751,871 blocked IGF-I- or IGF-II-induced autophosphorylation of the IGF-1R in NIH3T3/IGF-1R cells very rapidly upon addition of antibody, with substantial but incomplete inhibition when antibody was added only 1 minute before addition of activating ligand (Fig. 2A). IGF-I or IGF-II addition resulted in increased levels of phospho-akt, a known downstream mediator of the IGF-I pathway, and this activation was also blocked by addition of CP-751,871 for only 1 minute before IGF-I addition (Fig. 2A). At these early times, no down-regulation of IGF-1R occurred. In contrast, after 24 hours of exposure to CP-751,871, marked loss of cellular IGF-1R accompanied the block of receptor autophosphorylation and akt activation upon IGF-I stimulation (Fig. 2A). Thus, CP-751,871 produces an immediate inhibition of IGF-1R activation followed by a delayed loss of the receptor.

The time course of IGF-1R loss on CP-751,871 treatment was examined in more detail. Treatment of tumor cells in culture with CP-751,871 at 1 μg/mL resulted in a time-dependent decrease in total cellular IGF-1R, which was maximal between 3 and 4 hours as determined by Western blot analysis (Fig. 2B). CP-751,871-induced internalization of IGF-1R could be monitored indirectly by following the localization of CP-751,871 over time by confocal microscopy (Fig. 2C). NIH 3T3/IGF-1R cells were exposed to CP-751,871 for different time periods, fixed, and stained with an anti-human Alexa488 conjugated antibody. Binding of CP-751,871 to the cell surface was detected within 15 minutes, and the complex translocated to the cytoplasmic compartment within 1 hour after addition of antibody. CP-751,871 also induced the down-regulation of IGF-1R protein on MCF7 breast carcinoma cells and multiple myeloma cells (RPMI 8226) in vitro.

CP-751,871 Recognizes the IGF-1R/IR Heterodimer Complex. IGF-1R and insulin receptors have been shown to associate in both homodimer and heterodimer complexes in multiple cell and tissue types (22, 23). The ability of CP-751,871 to recognize the heterodimer complex was investigated using MCF7 cells that express both IR and IGF-1R in homodimer and heterodimer complexes (24). IGF-1R complexes from cell lysates were immunoprecipitated by CP-751,781 (CP-751,871 does not cross react with IR; data not shown). Immunoblot analysis of these IGF-1R immune complexes with an anti-insulin receptor antibody showed the presence of the IR (data not shown). Conversely, immunoprecipitation with an insulin receptor antibody and immunoblotting with an IGF-1R antibody also revealed the IGF-1R/IR complex (data not shown). These data show that CP-751,871 recognizes an IR/IGF-1R heterodimer complex, and treatment of the MCF7 cells with CP-751,871 results in the down-regulation of both IGF-1R homodimers and heterodimers of IGF1-R with insulin receptor.

CP-751,871 Induces the Down-Regulation of IGF-1R In vivo. Athymic mice bearing NIH3T3/IGF-1R tumors were used to investigate the biochemical activities of CP-751,871 in vivo; i.p. injection of CP-751,871 resulted in a serum C_max between 12 and 24 hours. At 24 hours, there was a dose-dependent reduction of IGF-1R protein in tumors, with 50% reduction observed at a serum concentration of 15 μg/mL (Fig. 3A). A time course analysis relating CP-751,871 serum concentrations with IGF-1R tumor levels showed strong relationship between antibody levels with IGF-1R down-regulation (Fig. 3B). An injected dose of 125 μg of CP-751,871 resulted in serum levels of approximately 40 μg/mL CP-751,871 and >90% down-regulation of IGF-1R from the tumor at 24 hours. The serum antibody concentrations drop to 20 μg/mL at 48 to 72 hours and IGF-1R levels rebound to 70% of control levels. CP-751,871 serum concentrations and IGF-1R reduction are maintained at this level throughout the remainder of the experiment (7 days). The half-life of CP-751,871 in an athymic mouse was determined to be 4 to 6 days by longer-term studies (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3

CP-751,871 down-regulates tumor associated IGF-1R. CP-751,871 was administered i.p. to female athymic (nu/nu) mice bearing NIH-3T3 tumors (400-500 mm³) overexpressing human IGF-1R. A, doses of 31 to 125 μg per mouse were given and tumors were excised 24 hours post-treatment (C_max). B, a dose of 125 μg was given and tumors excised at the indicated times. Blood samples were taken at time of tumor excision, and the level of tumor-associated IGF-1R and CP-751,871 serum levels were determined by capture ELISA as described in Materials and Methods. Points, mean values (n = 4 animals); bars, ±SE.

Pharmacokinetic Properties of CP-751,871 in Cynomolgus Monkeys. IGF-1R has been shown to be expressed on many tissues, thus binding to tissue receptors could influence both the clearance and exposure of CP-751,871. Due to the lack of binding of CP-751,871 to mouse IGF-1R, the pharmacokinetics of CP-751,871 were characterized in cynomolgus monkeys where the affinity of CP-751,871 to cynomolgus IGF-1R is similar to the human receptor. Following i.v. bolus administration, serum concentrations of CP-751,871 declined biphasically as presented in Fig. 4. The decline of serum concentration of CP-751,871 seemed to be parallel among the doses of 3, 30, and 100 mg/kg, which suggested that saturation of total IGF-1R binding sites has been achieved at the lowest dose, 3 mg/kg. The systemic clearance of CP-751,871 in cynomolgus monkeys were characterized to be low (0.0065-0.0116 mL per min per kg) and the elimination half-lives were ∼6 days in the monkeys. The observed maximal serum concentration of CP-751,871 following i.v. injections of 3, 30, and 100 mg/kg doses reached 0.1 to 2 mg/mL. The elimination of CP-751,871 in cynomolgus monkeys may have been influenced by a PAHA (primate anti-human antibody) response. Direct measurement of primate anti-CP-751,871 antibodies in these monkeys were not conducted for this study, but when evaluated in subsequent studies total and neutralizing anti-CP-751,871 antibodies were detected in the monkeys treated with CP-751,871 (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4

Serum concentration versus time profiles of CP-751,871 in cynomolgus monkeys following intravenous administration. The concentration of CP-751,871 in monkey serum was determined using a capture ELISA as described in Materials and Methods.

CP-751,871 Inhibits Tumor Growth. CP-751,871 was tested for the ability to inhibit tumor growth, both as a single agent and in combination with cytotoxic chemotherapy. As a single agent, a single dose of CP-751,871 produced dose-dependent inhibition of NIH3T3/IGF-R tumor growth. A dose of 125 μg produced ∼90% IGF-1R down-regulation in tumors at C_max of 24 hours (Fig. 3A), and 90% tumor growth inhibition by day 10 (Fig. 5A). Tumor growth inhibition studies were extended to human tumor xenograft models. As a single agent, CP-751,871 inhibited the growth of s.c. xenografts derived from colon (Colo-205, Fig. 6A), breast (MCF7, Fig. 7), and lung (H460) cancer cell lines (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 5

CP-751,871 inhibits tumor growth as a single agent and in combination with Adriamycin. 3T3/IGF-1R tumor-bearing mice were injected with a single i.p. dose of CP-751,871 (A), a single dose of Adriamycin (50% maximal tolerated dose = 7.5 mg/kg i.v.) or both in combination (B). Points, mean tumor volumes (n = 4 mice per group); bars, ±SE. CP-751,871 alone (125 μg) significantly inhibited tumor growth (P < 0.01, day 16, Student's t test), as did higher doses (250, 500, and 1,000 μg; data not shown). Treatment with Adriamycin alone and all combinations with Adriamycin significantly inhibited tumor growth at day 16 (P < 0.01). Combination of Adriamycin and antibody at 62.5 or125 μg produced significantly greater inhibition of tumor growth than Adriamycin alone at day 16 (P < 0.05). C, administration of CP-751,871 and Adriamycin on a weekly schedule prolongs tumor growth inhibition. Athymic mice bearing 3T3/IGF-1R tumors were dosed with CP-751,871 (i.p.) or Adriamycin (7.5 mg/kg i.v.) once a week as single agents, or in combination starting at day 1. Points, mean tumor volumes (n = 5 mice per group); bars, ±SE. On day 12, all treated groups had significantly smaller tumors than vehicle (P < 0.01, Student's t test); and on both days 12 and 21, tumors in all groups with antibody combined with Adriamycin were significantly smaller than in the group treated with Adriamycin alone (P < 0.01).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 6

CP-751,871 inhibits Colo-205 tumor growth as a single agent and in combination with 5-FU. Colo-205 tumor-bearing mice were injected with a single dose of CP-751,871 IP (A), a single dose of 5-FU (100 mg/kg, i.v.) or in combination on day 1 (B). Points, mean tumor volumes (n = 4 mice per group); bars, ±SE. CP-751,871 significantly inhibited tumor growth relative to vehicle controls evaluated on day 20 at both doses (P < 0.05, Student's t test). Treatment with 5-FU also significantly inhibited tumor growth, but addition of either dose of CP-751,871 to the 5-FU produced significantly greater tumor inhibition than 5-FU alone (P < 0.05). C, administration of CP-751,871 and 5-FU on a weekly schedule prolongs Colo-205 tumor growth inhibition. Athymic mice bearing Colo-205 tumors were dosed with CP-751,871 (i.p.) or 5-FU (100 mg/kg i.v.) once a week as single agents, or in combination starting at day 1. Points, mean tumor volumes (n = 5 mice per group); bars, ±SE. CP-751,871 alone, 5-FU, or the combination produced significant inhibition of tumor growth (P < 0.05). The combination of antibody CP-751,871 and 5-FU was produced significantly greater inhibition of tumor growth than 5-FU alone (P < 0.05).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 7

CP-751,871 inhibits MCF7 tumor growth as a single agent and in combination with tamoxifen. MCF7 tumor-bearing mice were treated with a single dose of CP-751,871 (i.p.), tamoxifen (3 mg/kg q2d × 5), or both agents together. Points, mean tumor volumes (n = 4 mice per group); bars, ±SE. The combination of CP-751,871 and tamoxifen produced significant inhibition of tumor growth relative to vehicle controls (P < 0.01, Student's t-test) and the inhibition of the combination was also significantly greater than that of either agent alone (P <0.05).

These results show the antitumor efficacy of CP-751,871 as a single agent; however, combinations of this antibody with other standard chemotherapeutic antitumor agents provided a substantial improvement in tumor growth inhibition relative to that achieved with the single agents. Administration of single doses of CP-751,871 with Adriamycin (half maximal tolerated dose) to animals bearing staged NIH3T3/IGF-1R tumors resulted in enhanced efficacy as compared with either agent alone (Fig. 5B). Repeated weekly administration of CP-751,871 or Adriamycin as single agents inhibited tumor growth for only 12 to 14 days, at which time the tumors assumed a growth rate similar to control tumors (Fig. 5C). In contrast, combination of both agents inhibited tumor growth almost completely throughout the 28 days of dosing period.

Combinations of CP-751,871 were also extended to other tumors and other cytotoxic agents. Combining a single dose of CP-751,871 with 5-fluorouracil (5-FU, 100 mg/kg i.v., 50% maximal tolerated dose) in animals bearing Colo-205 tumors resulted in enhanced efficacy relative to either agent alone (Fig. 6B). Extension of these dosing combinations to a weekly schedule resulted in initial minor regression and 100% inhibition of tumor growth when both drugs were used in combination, which was superior to the single agent activities of either agent alone (Fig. 6C).

Combination of CP-751,871 with a more targeted therapeutic agent was also investigated. Tamoxifen is an anti-estrogen and is a front line therapy for estrogen receptor–positive tumors. MCF7 breast carcinoma cells are estrogen receptor positive, dependent upon estrogen for growth in athymic mice, and responsive to anti-estrogens such as tamoxifen (24). Tamoxifen was combined with CP-751,871 in the treatment of MCF7 tumors and showed an enhanced antitumor growth activity when used in combination versus either agent alone (Fig. 7). These data show that CP-751,871 can inhibit tumor growth as a single agent, and enhances the efficacy of both cytotoxic and targeted cancer therapies.

CP-751-871 Biomarker Activity. Measurements of IGF-1R depletion in patients would be a valuable biomarker to monitor the biochemical activity of an antibody directed against the IGF-1R. Ex vivo treatment of blood with CP-751,871 resulted in the down-regulation of IGF-1R on human peripheral blood mononuclear cells as monitored by flow cytometry (Fig. 8). A rapid dose-dependent depletion was observed with an IC₅₀ of 40 to 50 ng/mL. This concentration is similar to the IC₅₀ of CP-751,871 for in vitro inhibition of IGF-1 binding (180 ng/mL) and IGF-1-induced receptor autophosphorylation (80 ng/mL) but is far lower than the plasma concentrations needed to down-regulate IGF-1R in tumors (EC₅₀ 15 μg/mL) presumably due to poor tumor penetration.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 8

CP-751,871 induced the down-regulation of IGF-1R on PBMCs derived from healthy volunteers. Expression of surface IGF-1R in T helper (CD3/CD4), cytotoxic cells (CD3/CD8), and B cells (CD19) after incubation in whole blood in the presence of CP-751,871 was analyzed by flow cytometry as indicated in Materials and Methods. Points, mean values from five individuals.

DISCUSSION

The IGF signaling pathway has been shown to be important for mitogenesis, survival, and tumorigenesis and may offer an attractive target for cancer chemotherapy. We report here a high-affinity human anti-IGF-1R antibody that inhibits the function of IGF-1R and has antitumor activity as a single agent or in combination with other cancer drugs.

CP-751,871 is of the IgG2 subtype and was specifically chosen because that subtype is a poor activator of cellular immune modulators. Therefore, any tumor growth inhibition seen in vivo following antibody treatment is likely attributed to the blockade of IGF-1R function and not to an antibody-dependent cell cytotoxic (ADCC) or complement activity. Subsequent to the preparation of this article, Burtrum et al. (25) and Maloney et al. (26) have reported the creation of distinct anti-IGF-1R antibodies of the IgG1 isotype. Both of these antibodies show similar activities such as IGF-1 antagonism, IGF-1R internalization and single agent antitumor activity as compared with CP-751,871. The major distinguishing feature between these antibodies and CP-751,871 is the use of an IgG1 Fc domain versus an IgG2 for CP-751,871. These two isotypes differ in that IgG2s have a longer half-life in humans, and the IgG1 isotype in humans is capable of activating immune cell effector functions (antibody-dependent cell cytotoxic) but the IgG2 isotype is inefficient in this process. Whereas it remains to be determined whether efficacy in humans is linked to this immune effector activity, the isotype differences may also impact the safety profile of these candidates. Furthermore, given the antibody-induced IGF-1R down-regulation, antigen density may be insufficient to trigger an antibody-dependent cell cytotoxic response.

The data presented here show that CP-751,871 is able to bind to IGF-1R with high affinity and inhibit its function both by inhibiting IGF-I binding and by inducing IGF-1R down-regulation in vitro and in tumor xenografts in vivo. Down-regulation of surface receptors upon antibody binding has been frequently observed, for example with antibodies to HER2 (27) as well as antibodies to IGF1R (25, 26). IGF-1R down-regulation by CP-751,871 was proportional to the serum concentrations and was maintained for >7 days after a single dose in athymic mice. The concentration of CP-751,871 in the serum required for maximal inhibition of tumor growth as a single agent (40 μg/mL) corresponded to ∼90% down-regulation of IGF-1R at day 1. Serum concentrations decreased to 20 μg/mL at day 7, which corresponded to 30% reduction of IGF-1R. These data provided a basis for estimating the efficacious concentrations of CP-751,871 required for the treatment of solid tumors as a single agent.

Survival signals from IGF-1R can reduce the effectiveness of chemotherapy and various approaches for inhibition of IGF-1R have been shown to increase the activity of cytotoxic drugs (28–30). Maloney et al. have reported that a mouse monoclonal anti-IGF-1R antibody inhibited tumor growth as a single agent and enhanced the activity of gemcitabine in a pancreatic tumor model (26). We have extended these findings by use of a fully human anti-IGF-1R antibody in combination with Adriamycin, 5-fluorouracil, and tamoxifen. In all cases, submaximum tolerated doses of chemotherapy in combination with CP-751,871 enhanced efficacy and produced a more durable response. Similarly, dose escalation studies in the NIH3T3/IGF-1R tumor model showed that doses of CP-751,871 that were not efficacious as a single agent, could substantially enhance the antitumor activity of Adriamycin. These data suggest that usage of sub–maximal tolerated dose of both agents in combination may reduce high-dose associated toxicities without sacrificing clinical efficacy.

IGF-1R has been shown to be expressed on many tissues, and that widespread expression could present an obstacle to generating a therapeutic antibody, in that saturation of the target is either not possible (nonlinear PK) or requires large amounts of antibody. However, our studies in the cynomolgus monkeys, indicated a nearly linear increase in antibody concentrations with doses of 3, 30, and 100 mg/kg. Thus, depletion of anti-IGF1R by binding to normal tissues does not preclude reasonably low projected human doses.

A trend in drug development today is to shorten the time required to make critical decisions regarding the continued development of clinical candidates. The use of biomarkers can prove that the drug is inhibiting its intended target at clinically obtainable concentrations and allow comparison with those required for biochemical effects in preclinical models. One such model for CP-751,871 is to measure the down-regulation of IGF-1R on peripheral blood lymphocytes, which can be readily obtained from both human volunteers and cancer patients. Ex vivo treatment of human blood with CP-751,871 resulted in a dose-dependent down-regulation of IGF-1R at concentrations which are in line with the in vitro inhibition of IGF-1R autophosphorylation and ligand binding. Whereas the measurement of IGF-1R receptor down-regulation in circulating cells could provide evidence in clinical trials for functional biochemical modulation, it may not reflect the concentrations required to impact the growth of solid tumors due to inefficient penetration of macromolecules into solid tumors. Nonetheless, our results with this assay suggest that it may be useful for establishing proof of biochemical mechanism in conjunction with clinical trials of this agent.

CP-751,871 is a human monoclonal anti-IGF-1R antibody that can block the function of IGF-1R through multiple mechanisms. The activity showed in both in vitro and in vivo assays suggests IGF-1R provides a critical survival signal in multiple tumor types. These data together with the observations that IGF-1R is an indicator of poor prognosis, is up-regulated in human tumors, and promotes tumor cell survival, suggest that blockade of IGF-1R will successfully inhibit the progression of multiple tumor types in cancer patients. This antibody is currently in phase I clinical trials.