The outcome for patients with lung cancer has not changed significantly for more than two decades. Several studies show that the overexpression of vascular endothelial growth factor (VEGF)/vascular permeability factor and epidermal growth factor (EGF) and their receptors correlates with the clinical outcome for lung cancer patients. However, clinical trials of agents that target either of these pathways alone have been disappointing. We hypothesize that targeting both the tumor and its vasculature by simultaneously blocking the VEGFR and EGFR pathways will improve the treatment of locoregional lung cancer. Human lung cancer specimens were first examined for the activation of VEGF receptor 2 (VEGFR2) and EGF receptor (EGFR) for tumor and tumor-associated endothelial cells, and both were found to be activated. The effects of ZD6474 (ZACTIMA), a small-molecule inhibitor of VEGFR2 and EGFR tyrosine kinases, were then studied in vitro using human lung cancer and microvascular endothelial cells. In vitro, ZD6474 inhibited EGFR, VEGFR2, mitogen-activated protein kinase and Akt phosphorylation, EGF- and VEGF-induced proliferation, and endothelial cell tube formation and also induced apoptosis. ZD6474 was further studied in vivo using an orthotopic mouse model of non–small cell lung cancer using NCI-H441 human lung adenocarcinoma cells. The inhibition of both VEGFR2 and EGFR signaling pathways by ZD6474 resulted in profound antiangiogenic, antivascular, and antitumor effects. These results provide a basis for the development of clinical strategies for the combination of selective protein tyrosine kinase inhibitors that block both EGFR and VEGFR signaling as part of the management of locally advanced lung cancer. [Mol Cancer Ther 2007;6(2):471–83]
Lung cancer is a major health problem worldwide and is the leading cause of cancer-related death for both men and women, with ∼169,000 deaths annually (1) in the United States alone. Non–small cell lung cancer (NSCLC) accounts for up to 80% of all lung cancer cases and patients typically present with advanced disease, and fewer than 50% will have surgically resectable lung cancer with the potential for cure (2). The prognosis of patients with advanced lung cancer remains poor, and recent studies show that conventional therapies may have reached a therapeutic plateau (3) as evidenced by the 5-year survival for NSCLC, which remains at 15%. Clearly, new strategies for the therapy of lung cancer are urgently needed, and the current therapeutic challenge is to optimize available nonoperative strategies by incorporating new agents into current therapeutic regimens. Strategies that target growth factor receptor tyrosine kinase signaling pathways involved in tumor angiogenesis, growth, and survival offer great potential, and two of the most widely studied targets are vascular endothelial growth factor receptor 2 (VEGFR2) and epidermal growth factor receptor (EGFR).
VEGF induces vascular permeability and plays a pivotal role in vasculogenesis, angiogenesis, and endothelial integrity and survival (4, 5). Several approaches to block VEGF activity are under evaluation, including anti-VEGF or anti-VEGFR antibodies and VEGFR tyrosine kinase inhibitors (5). Bevacizumab, a humanized antibody against VEGF, has shown great promise and improved survival for patients with metastatic colorectal or lung cancer when used in combination with chemotherapy (6, 7). However, although survival was improved in these studies, there remains an unmet need for improved efficacy, and clinical trials with VEGF inhibitors have been associated with thromboembolic and hemorrhagic complications (8).
Increased expression of EGFR is one of the earliest occurring and most consistently detected abnormalities in the bronchial epithelium of heavy smokers and is pronounced in 65% to 84% of all NSCLC (9). EGFR overexpression is associated with poor outcome for several malignancies; yet, the relationship between EGFR overexpression and NSCLC prognosis is controversial (10). Although the primary direct target of anti-EGFR agents is the tumor cell, they can also exert direct (11) or indirect (12) antiangiogenic and antivascular effects. Currently, several different strategies, including antibodies or small-molecule inhibitors, to target EGFR are in late-stage clinical development for the treatment of lung and other cancers. In phase III clinical trials of patients with advanced lung cancer, erlotinib, a small-molecule antagonist of EGFR, was associated with a modest improvement in survival when given alone after the failure of first- or second-line chemotherapy (13); yet, erlotinib with concurrent carboplatin and paclitaxel did not confer a survival advantage over chemotherapy alone in patients with previously untreated advanced NSCLC (14).
To study the influence of lung microenvironment upon response to therapy, we developed an orthotopic lung cancer model to recapitulate the local and regional growth pattern seen in lung cancer patients (15, 16). A single solid lung tumor progressed to a widespread and fatal thoracic process characterized by dissemination of lung cancer in the pleura and metastasis to intrathoracic and extrathoracic lymph nodes. The purpose of the present study was to evaluate and characterize the effects of inhibition of VEGFR2 and EGFR signaling by ZD6474, an orally available small-molecule receptor tyrosine kinase inhibitor (17, 18), on tumor growth, progression, angiogenesis, and vascularization of human lung cancer in the lungs of nude mice.
Several lines of evidence indicate that VEGF and EGF and other proangiogenic factors (i.e., basic fibroblast growth factor and interleukin-8) are critical for the integrity of the normal lung (19) and for lung cancer vascularization, growth, and progression (20). However, the relationship between VEGFR and EGFR overexpression and NSCLC prognosis is controversial (10). We therefore first examined the expression and activation of VEGFR2 and EGFR on both tumor cells and tumor-associated endothelial cells in human lung cancer specimens as potential therapeutic targets. The effects of EGFR blockade on cell cycle progression have been investigated in several human cell types (21). However, in the preclinical setting, the role of EGFR is typically studied using only cells that markedly overexpress this receptor and thus may not adequately reflect the variability of EGFR expression observed in human lung cancer specimens. Furthermore, although the activity of VEGF upon its receptors had been assumed to be restricted to endothelial cells, recent studies suggest that VEGFRs may also be expressed by non-endothelial cells, including lung cancer cells (22, 23), and may contribute to tumor cell motility (24) and invasion (25). However, the functional significance of VEGFR expression by non-endothelial cells has not been fully elucidated. To address these issues, we evaluated the functional role of VEGFR2 and EGFR signaling for tumor cells and endothelial cells in vitro using ZD6474. We then studied the effects of ZD6474 upon angiogenesis, vascularization, tumor growth, and metastasis on lung tumors growing orthotopically in mice in vivo.
Materials and Methods
NCI-H441, NCI-H358, A427, and Calu-3 human lung adenocarcinoma cells and H226, SK-MES-1, and SW900 human squamous cell carcinoma cells were obtained from the American Type Culture Collection (Manassas, VA). PC14-PE6 human adenocarcinoma cells were selected from the parental PC14 cell line for the ability to form pleural effusions in the mouse (26). All lung cancer cell lines were maintained in Eagle's MEM with the exception of NCI-H441 that was maintained in RPMI 1640. Murine lung microvascular endothelial cells (MLEC) were derived (27) from transgenic H-2Kb-tsA58 mice (CBA/ca X C57BL/10 hybrid; Charles River Laboratories, Wilmington, MA) and maintained in DMEM. All media were supplemented with 10% fetal bovine serum, sodium pyruvate, nonessential amino acids, l-glutamine, 2-fold vitamin solution, and penicillin-streptomycin (Flow Laboratories, Rockville, MD). Human umbilical vein endothelial cells (HUVEC; Cascade Biologics, Portland, OR) were grown on attachment factor-coated tissue culture flasks in MEM containing 10 ng/mL basic fibroblast growth factor and 15% fetal bovine serum. Human pulmonary artery endothelial cells (HPAEC; Cascade Biologics) were maintained in medium 200 supplemented with microvascular growth supplement and 5% fetal bovine serum. All cultures were maintained in a humidified incubator at 37°C (33°C for MLEC) under 5% CO2/95% air free of Mycoplasma and pathogenic murine viruses. HUVEC and HPAEC were used for experiments at passages 3 to 5, and other cells were used for experiments at passages <8 after recovery from frozen stocks.
Western blotting was done as previously described (11). Cultured cells treated with recombinant human EGF (Invitrogen, Carlsbad, CA), VEGF (R&D Systems, Minneapolis, MN), and/or ZD6474 (AstraZeneca Pharmaceuticals, Macclesfield, United Kingdom) or vehicle control were harvested and lysed. Total cell proteins were resolved by 7.5% SDS-PAGE, transferred to polyvinylidene difluoride membranes (90 V for 2 h), and probed with monoclonal antibodies directed against human EGFR (Upstate Biotechnology, Lake Placid, NY), phosphorylated EGFR (pEGFR; Tyr845; Cell Signaling, Beverly, MA), VEGFR2 (Santa Cruz Biotechnology, Santa Cruz, CA), pVEGFR2 (Tyr1045; Biosource International, Camarillo, CA), mitogen-activated protein kinase (MAPK), pMAPK (Zymed, South San Francisco, CA), Akt, pAkt (Cell Signaling), or α-actin (Sigma, St. Louis, MO). Protein bands were scanned by a densitometer, and the relative intensities were quantified using ImageQuant software (Molecular Dynamics, Sunnyvale, CA).
Cell Proliferation Assay
Cell proliferation was approximated by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide dye reduction method. HUVEC, HPAEC, MLEC, or H441 cells plated at 5 × 103 per well in triplicate in 96-well plates were incubated overnight at 37°C. The cells were then washed and incubated for 72 h with ZD6474 or vehicle control in fresh growth medium at 37°C. The number of viable cells was determined in the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay using an MR-5000 plate reader with absorbance set at 570 nm.
Cell Apoptosis Assay
Cells cultured in chamber polystyrene vessel tissue culture–treated glass slides (2 × 104; Becton Dickinson Labware, Franklin Lakes, NJ) were serum-starved in the absence or presence of human recombinant VEGF or EGF and ZD6474 or vehicle control for 48 h. The cells were centrifuged, fixed with 4% paraformaldehyde, and immunofluorescently stained using a terminal deoxynucleotidyl transferase–mediated nick-end labeling (TUNEL) apoptosis detection kit (Promega, Madison, WI) or for activated caspase-3 (BD Bioscience PharMingen, San Diego, CA) with modifications as described previously (28). For nuclear staining, cells were incubated with 1 μg/mL Hoechst at room temperature in PBS for 10 min and then rinsed thrice with PBS. A coverslide was mounted with a mixture of 90% glycerol and 10% PBS. Quantification of the apoptotic cells was expressed as the percentage of positive cells to the total number of cells in 10 random 0.039-mm2 fields at ×200 magnification.
Migration and Invasion Assay
Migration and invasion of HUVEC, HPAEC, MLEC, or H441 were assessed in Transwell cell culture chambers (Costar, Cambridge, MA). For the invasion assay, the lower surface of the filters was coated with 1 μg of fibronectin, and the upper surface was coated with 5 μg of Matrigel. Cells (1 × 105) were seeded into the upper compartment and incubated with ZD6474 in serum-free MEM containing 0.1% bovine serum albumin at 37°C for 3 to 4 h for migration and 6 to 8 h for invasion. The filters were then stained with crystal violet, and the cells in five ×200 magnification fields were counted.
Male athymic BALB/c nude mice (6–8 weeks old) were obtained from the Animal Production Area of the National Cancer Institute-Fredrick Cancer Research and Development Center in Frederick, MD. Mice were housed and maintained in specific pathogen-free conditions in accordance with the guidelines of the American Association for Laboratory Animal Care and all current regulations and standards of the U.S. Departments of Agriculture and of Health and Human Services and the NIH.
In vivo Gelfoam-Agarose Sponge Angiogenesis Assay
Male athymic BALB/c nude mice (6–8 weeks old) were anesthetized with sodium pentobarbital (50 mg/kg body weight) and placed in a supine position. Sterile absorbable Gelfoam sponges (Pharmacia & Upjohn, Peapack, NJ) were prepared as previously described (29). After cleansing the overlying skin with alcohol, a 5-mm midline incision was made, and s.c. pockets were created on both lateral aspects of the chest 2 to 3 cm away from the incision. A Gelfoam sponge was inserted into each pocket, and the skin wound was closed with two surgical metal clips. Treatment with ZD6474 or vehicle began on the day after Gelfoam implantation and was continued for 14 days, and the mice were then killed. Gelfoam sponges were recovered, washed with PBS, and snap-frozen with liquid nitrogen in optimal cutting temperature compound.
Miles Assay of Vascular Permeability
The Miles assay (30) was done as described (26) with minor modifications. Male athymic BALB/c nude mice (6–8 weeks old) without evidence of downy hair were treated once daily with ZD6474 or vehicle via oral gavage for 14 days (25 or 50 mg/kg) and were i.v. injected with 150 μL of 0.5% Evans blue dye (Sigma Chemical Co., St. Louis, MO) after the final dose of ZD6474 or vehicle. Mice were caged separately for the duration of the assay. Fifteen minutes after administration of dye, the dorsal skin of each mouse was injected i.d. with VEGF, EGF, H441 lung adenocarcinoma conditioned media, and PBS (20 μL volume) in a row. Thirty minutes after i.d. injection, mice were killed, and the skin was removed. Wheals (5 mm in diameter) corresponding to the site of sample injection were individually resected and incubated in 100 μL of formamide at 37°C for 48 h to extract Evans blue dye. The amount of Evans blue dye was determined by measuring the absorbance at 660 nm in a spectrophotometer (31).
Orthotopic Lung Adenocarcinoma Model
Male athymic BALB/c nude mice (6–8 weeks old) were anesthetized with sodium pentobarbital (50 mg/kg body weight) and placed in the right lateral decubitus position (15). The skin overlying the left chest wall in the mid-axillary line was prepped with alcohol and incised (3 mm), and the underlying chest wall and intercostal spaces were visualized. Logarithmically growing H441 cells (5 × 105; single-cell suspensions of over 90% viability as determined by trypan blue exclusion) in 50 μL containing 50 μg growth factor–reduced Matrigel (Becton Dickinson & Co., San Jose, CA) were injected into the left lateral thorax, at the lateral dorsal axillary line. After tumor injection, the mice were turned to the left lateral decubitus position and observed for 45 to 60 min until fully recovered. No anesthesia or surgery-related deaths occurred.
Therapeutic Experiments with ZD6474
H441 human lung adenocarcinoma cells were injected into the lungs of nude mice as described above. Five, 10, or 15 days after tumor cell implantation into the lungs, five mice were killed and autopsied at each of the three time points, and lung tumor burden and pleural effusion volume were assessed. The remaining mice were randomized into three groups (n = 8–10) and treated with ZD674 at doses of 25 or 50 mg/kg daily or vehicle control starting at each of the three time points (5, 10, or 15 days) after tumor cell implantation. ZD6474 was dissolved in vehicle containing 1% Tween 80 and administered via oral gavage once daily. Treatment was continued until control mice displayed signs of morbidity (typically 28–32 days after tumor cell implantation), at which point mice were killed and autopsied, and lung tumor weight, pleural effusion, and metastasis were evaluated. In a separate survival experiment, mice (n = 10 per group) were randomized to treatment with ZD6474 (50 mg/kg daily) or vehicle control starting 15 days after tumor cell implantation into the lungs. Mice were then closely monitored and killed when they became moribund, at which point the day was recorded. For all of the experiments, tumor and lung tissues were subsequently subjected to histologic and immunohistochemical analyses as described below.
Histology and Immunohistochemistry
Tumor and adjacent lung tissues from five or more mice for each group were frozen in optimal cutting temperature compound or fixed with formalin and embedded in paraffin and then sectioned and stained with H&E or immunostained. Immunostaining for proliferating cell nuclear antigen (DAKO, Copenhagen, Denmark); basic fibroblast growth factor, interleukin-8, EGF, transforming growth factor-α (TGF-α), and VEGF (all from Santa Cruz Biotechnology); EGFR (Zymed); activated-EGFR (Biosource International); VEGFR2 (Santa Cruz Biotechnology); activated VEGFR2 (Oncogene, Cambridge, MA); activated Akt and activated MAPK (Cell Signaling); and bromodeoxyuridine, caspase-3, and/or CD31/platelet/endothelial cell adhesion molecule 1 (PharMingen, San Diego, CA) were done using the methods indicated as described previously (11, 32). TUNEL was done using a commercial apoptosis detection kit (Promega) with modifications (32). Bromodeoxyuridine- or TUNEL-positive endothelial cells were counted in 10 high power fields at ×200 magnification, and the percentage of positive cells were calculated.
Quantification of Microvessel Density, Proliferating Cell Nuclear Antigen, TUNEL, Caspase-3, EGF, VEGF, EGFR, and VEGFR
For quantification of microvessel density in Gelfoam sponges or tumors, 10 random 0.159-mm2 fields at ×100 magnification were captured for each sponge or tumor after staining with anti-CD31 antibody, and microvessels were quantified using Scion software based on the NIH Image program for Macintosh (Scion Corp., Frederick, MD) according to the method described previously (29). The number of proliferating cell nuclear antigen–, TUNEL-, or caspase-3–positive cells was quantified in 10 random 0.159-mm2 fields at ×100 magnification. The immunohistochemical reaction intensity absorbance for EGF, VEGF, EGFR, and VEGFR in lung tumors was determined in 10 random 0.039-mm2 fields at ×200 magnification tumor tissues and was measured using the Optimas Image Analysis software. The samples were not counterstained so that the absorbance was attributable solely to the product of the immunohistochemical reaction. The receptor or cytoplasmic immunoreactivity was evaluated by computer-assisted image analysis and was expressed as percent of tumor cell expression to tumor-free adjacent tissue expression.
The data were analyzed by χ2 test (only for the incidence) and the Student's t test (two tailed).
VEGFR2 and EGFR Are Activated in Clinical Specimens of Human Lung Cancer
To study the role of VEGFR2 and EGFR in human lung cancer, we analyzed the activation of both receptors (Fig. 1) in 10 randomly selected human lung cancer specimens (adenocarcinoma and squamous cell carcinoma) obtained from the Surgical Pathology Laboratory at M. D. Anderson Cancer Center. Using double immunofluorescence staining techniques as described above, we characterized receptor activation for both tumor cells and tumor-associated endothelial cells. Five of 10 samples (50%) were positive for pVEGFR2/CD31, and 6 of 10 samples (60%) were positive for pEGFR/CD31. As shown in Fig. 1A and B, the co-localization of CD31 and pVEGFR2 (or CD31 and pEGFR) indicated the presence of activated VEGFR2 or EGFR not only on tumor cells but also on tumor-associated endothelial cells. Surprisingly, tumor cells expressed VEGFR2, suggesting that antagonists of VEGFR2 could have a direct antitumor effect. The 10 clinical specimens were also evaluated for activation of both VEGFR2 and EGFR. Seven (70%) samples stained positive for pVEGFR2; eight (80%) stained for pEGFR; and five (50%) stained for both pVEGFR2 and pEGFR, indicating that VEGFR2 and EGFR are activated independently in human lung cancer tissues (Fig. 1C). These data show that both VEGFR2 and EGFR are relevant targets for lung cancer.
Activation of VEGFR2 and EGFR in Human Lung Cancer Cells and Endothelial Cells In vitro
To study the role of EGFR and VEGFR in human lung cancer, we screened human lung adenocarcinoma (H441, PC14-PE6, A427, H358, and Calu-3), squamous cell carcinoma (H226, SW900, and SK-MES-1), and endothelial cells (HUVEC, HPAEC, and MLEC) for activation of VEGFR2 and EGFR by Western blotting using antibodies specific for pVEGFR2 and pEGFR. All cell lines were tested and determined to be wild type for the EGFR gene (data not shown). As shown in Fig. 2A, activation of VEGFR2 and EGFR was detected in H441, Calu-3, HUVEC, HPAEC, and MLEC in culture. The simultaneous presence of VEGFR and EGFR signaling pathways on both tumor cells and endothelial cells may therefore contribute to the malignancy of lung cancer.
VEGFR2 and EGFR Activation and Signaling for Human Lung Cancer Cells and Endothelial Cells In vitro
The activation of VEGFR2 and EGFR by both lung cancer and endothelial cells in human lung cancer suggests that ZD6474 and agents that target these receptors can directly affect both the tumor and endothelial components of lung cancer. To characterize the effects of ZD6474 upon VEGFR2 and EGFR tyrosine kinase activity, MLEC and H441 cells were pretreated with ZD6474 and briefly stimulated with VEGF or EGF. Western blotting showed that ZD6474 potently inhibited phosphorylation of VEGFR2 and EGFR in lung adenocarcinoma cells (H441), and in endothelial cells (MLEC) to a lesser degree, in a dose-dependent manner (Fig. 2B). VEGFR2 and EGFR blockade corresponded with decreased phosphorylation of downstream signaling intermediates Akt and MAPK. There were no significant changes in total proteins expression for either cell lines. Similar results were observed for HUVEC and HPAEC (data not shown).
Functional Effect of VEGFR2 and EGFR Signaling Blockade by ZD6474 Upon Growth, Survival, and Spread of Human Lung Cancer and Endothelial Cells In vitro
Our in vitro studies show that Akt and MAPK activation are suppressed by blocking the activation of VEGFR2 and EGFR using ZD6474. The activation of Akt and MAPK pathways is associated with cell survival, proliferation, migration, and invasion. We therefore examined the functional effects of ZD6474, a selective small-molecule inhibitor of VEGFR2 and EGFR tyrosine kinase phosphorylation, upon human lung cancer cell lines and endothelial cells in vitro. We investigated the direct effect of ZD6474 on migration, invasion, and proliferation of endothelial cells and tumor cells in vitro (Fig. 2C). ZD6474 markedly inhibited migration of HUVEC, HPAEC, MLEC, and H441 in a dose-dependent fashion with an IC50 of 8, 5.0, 1.25, and 0.3 μmol/L, respectively. ZD6474 also blocked invasion with an IC50 of 8.8, 4.6, 1.0, and 0.24 μmol/L for HUVEC, HPAEC, MLEC, and H441, respectively. Similar inhibitory effects were observed when ZD6474 was tested in a proliferation assay with an IC50 of 2.7, 1.9, 3.2, and 2.8 μmol/L for HUVEC, HPAEC, MLEC, and H441 cells, respectively.
To investigate the role of VEGFR2 and EGFR signaling on cell survival, we assessed the effect of ZD6474 on the induction of apoptosis in both endothelial cells and H441 lung adenocarcinoma cells. As shown in Fig. 2D, ZD6474 induced apoptosis of both endothelial cells and lung cancer cells in dose-dependent manner. The TUNEL assay revealed that starvation-induced apoptosis of MLEC or H441 was prevented by pretreatment with EGF or VEGF. Treatment with ZD6474 at 1.25 μmol/L blocked VEGFR2 and EGFR signaling and antagonized the antiapoptotic effects of EGF and VEGF in the TUNEL assay. These data were confirmed by analysis of the apoptosis-related protein caspase-3. As shown in Fig. 2D, there was a significant increase in the expression of activated caspase-3 in ZD6474-treated endothelial or tumor cells compared with those stimulated with EGF or VEGF alone.
Taken together, these in vitro studies show that Akt and MAPK activation can be significantly suppressed by blocking the activation of VEGFR2 and EGFR with ZD6474, and that this signal inhibition is associated with an inhibition of cell function, including proliferation, migration, invasion, and an induction of apoptosis, for both H441 lung adenocarcinoma and endothelial cells.
Blockade of VEGFR2 and EGFR Signaling Inhibits Angiogenesis and Vascular Permeability In vivo
To determine if blockade of VEGFR2 and EGFR signaling by ZD6474 would affect angiogenesis in vivo, we used a Gelfoam-agarose sponge assay (29) in which angiogenesis was induced by conditioned medium from H441 cells (CM), VEGF, or EGF stimulation compared with a PBS control. Treatment of mice with ZD6474 significantly reduced the microvessel density in all implanted sponges containing PBS, CM, VEGF, or EGF (Fig. 3A and B). To complement the angiogenesis assay, we also examined the effect of ZD6474 treatment upon vascular permeability in vivo using the Miles assay (30). Treatment of mice with ZD6474 decreased the vascular permeability induced by CM, VEGF, or EGF. This effect was particularly evident for VEGF-induced vascular permeability (Fig. 3C and D). Treatment with ZD6474 did not produce any obvious toxicity, including diarrhea or body weight loss.
Characterization of H441 Human Lung Adenocarcinoma Growth after Orthotopic Injection into the Mouse Lung
Based upon the observed activation of VEGFR2 and EGFR in lung cancer cell lines (Fig. 2A and B), we implanted H441 lung adenocarcinoma cells orthotopically into the mouse lung. Within 5 days of tumor cell injection, small but visible solitary lesions could be detected in the lungs of mice (Fig. 4A). H441 lung adenocarcinomas grew rapidly in the lung (Fig. 4A), and pleural effusions were detected within 15 days of tumor cell injection. Diffuse thoracic growth with extensive bilateral mediastinal, axillary, and neck lymphatic metastasis led to significant morbidity within 4 weeks of injection. The observed patterns of growth and dissemination of tumor in our lung orthotopic model recapitulates the local and regional growth pattern seen in lung cancer patients with solitary nodules progressing to diffuse disease involving pleural effusion with pleural and lymphatic metastasis.
Lung tumor tissues obtained from five mice at 5, 10, or 15 days after tumor cell injection were examined for expression and activation of VEGFR2, EGFR, and their ligands VEGF, EGF, and TGF-α compared with adjacent tumor-free lung. As shown in Fig. 4B, increased expression and activation of each receptor and their respective ligands was observed at an early stage of tumor progression (day 5 after injection). VEGFR2 and EGFR were activated for both tumor and tumor-associated endothelial cells (Fig. 4C) but not by endothelial cells of the vasculature in the lung tissue surrounding the tumors. The observed co-localization of pVEGFR and CD31 and/or pEGFR and CD31 in lung tumor tissues at each stage of tumor progression correlated to increased tumor and endothelial cell proliferation. Co-immunostaining for CD31 and bromodeoxyuridine incorporation (indicating proliferating cells) showed proliferation of tumor and tumor-associated endothelial cells (Fig. 4C and D).
Blockade of VEGFR2 and EGFR Signaling by ZD6474 Inhibits Growth and Metastasis of Orthotopic Human Lung Adenocarcinoma and Prolongs Survival in Mice
To study the role of VEGFR2 and EGFR signaling in the progressive growth of lung tumors, we evaluated the therapeutic effects of ZD6474 on tumor growth, angiogenesis, and metastasis in established H441 human lung adenocarcinomas growing orthotopically in the lungs of nude mice. The H441 cell line was selected for study because the cells express VEGF, EGF, TGF-α, VEGFR2, and EGFR and grew well in the lung after orthotopic injection. Mice were injected with H441 lung adenocarcinoma cells, and treatment with ZD6474 (25 or 50 mg/kg daily by oral gavage) or vehicle control began on day 5 (early stage), day 10 (intermediate stage), or day 15 (late stage) after tumor cell injection. As shown in Fig. 5A, the oral administration of ZD6474 (25 or 50 mg/kg) produced near-complete suppression of tumor growth (lung tumor weight and volume) and impaired pleural invasion and effusion formation compared with the control even for well-established lung tumors. Treatment with ZD6474 did not produce any obvious toxicity, including diarrhea or body weight loss. The mean vessel density in the lung tumors as indicated by the number of CD31-positive cells was decreased from 42 ± 11 in control mice to 10 ± 8 and 5 ± 4 in mice given ZD6474 at 25 or 50 mg/kg, respectively. Therapy with ZD6474 (50 mg/kg) significantly (P < 0.001) prolonged survival by at least 3-fold relative to control (Fig. 5B) even when initiation of therapy was delayed until 15 days after tumor injection when mice had locally advanced disease. These data show that ZD6474 targets lung cancer angiogenesis and can both prevent and reverse lung cancer progression and locoregional metastasis and suggest that further study of ZD6474 for lung cancer prevention and treatment is warranted.
ZD6474 Treatment Blocks Activation of VEGFR2 and EGFR for Tumor and Tumor-Associated Endothelial Cells and Inhibits Secretion of Proangiogenic Factors in Orthotopic Lung Adenocarcinomas
As shown in Table 1 and Fig. 6A, treatment with ZD6474 reduced the activation of VEGFR2 and EGFR without significant effect upon total VEGFR2 and EGFR expression. Double staining for CD31/pVEGFR2 and CD31/pEGFR (Fig. 6B) further revealed that the activation of VEGFR2 and EGFR in both tumor and tumor-associated endothelial cells in tumor tissues were blocked by ZD6474. One potential means of tumor resistance to therapy with agents that target VEGF and its receptors is the increased expression of proangiogenic factors. To characterize the in vivo effects of ZD6474 treatment upon growth factor expression, we did immunohistochemical analyses of H441 tumor tissues. The expression of several proangiogenic factors, including TGF-α, basic fibroblast growth factor, and interleukin-8 (Table 1), were decreased in tumors treated with ZD6474 compared with controls.
|Group .||pVEGFR2* .||pEGFR* .||TGF-α* .||bFGF* .||IL-8* .||CD31† .||PCNA‡ .||Caspase-3‡ .|
|Vehicle||169 ± 14||173 ± 12||233 ± 17||146 ± 6||232 ± 11||42 ± 11||110 ± 24||0.3 ± 0.7|
|ZD6474 25 mg/kg||106 ± 8§||109 ± 9§||178 ± 10∥||126 ± 4∥||148 ± 12§||10 ± 8§||31 ± 7§||31 ± 8§|
|ZD6474 50 mg/kg||97 ± 8§||104 ± 5§||181 ± 5∥||106 ± 11∥||142 ± 7§||5 ± 4§||6 ± 4§||35 ± 7§|
|Group .||pVEGFR2* .||pEGFR* .||TGF-α* .||bFGF* .||IL-8* .||CD31† .||PCNA‡ .||Caspase-3‡ .|
|Vehicle||169 ± 14||173 ± 12||233 ± 17||146 ± 6||232 ± 11||42 ± 11||110 ± 24||0.3 ± 0.7|
|ZD6474 25 mg/kg||106 ± 8§||109 ± 9§||178 ± 10∥||126 ± 4∥||148 ± 12§||10 ± 8§||31 ± 7§||31 ± 8§|
|ZD6474 50 mg/kg||97 ± 8§||104 ± 5§||181 ± 5∥||106 ± 11∥||142 ± 7§||5 ± 4§||6 ± 4§||35 ± 7§|
NOTE: Logarithmically growing H441 cells (5 × 105; single-cell suspensions) were injected into the left lungs of mice. Treatment (vehicle or ZD6474 at 25 or 50 mg/kg) began on day 5 following tumor injection, and mice were killed after 4 wks of treatment. Primary lung tumors were harvested, fixed, and stained with specific antibodies.
Abbreviations: bFGF, basic fibroblast growth factor; IL-8, interleukin-8; PCNA, proliferating cell nuclear antigen.
Mean ± SD intensity of positive cells from tumor tissues was measured in 10 random 0.039-mm2 fields at ×200 magnification using the Optimas Image Analysis software and was expressed as percentage of tumor to adjacent tumor free tissue expression.
Mean ± SD CD31-positive cells in tumors was quantified in 10 random 0.159-mm2 fields at ×100 magnification for each tumor using Scion software based on the NIH Image program for Macintosh.
Mean ± SD PCNA- or cleaved caspase-3–positive cells was quantified in 10 random 0.159-mm2 fields at ×100 magnification.
P < 0.001, compared with vehicle.
P < 0.01, compared with vehicle.
Blocking VEGFR2 and EGFR Signaling by ZD6474 Inhibits Cell Proliferation and Induces Cell Apoptosis in Orthotopic Lung H441 Adenocarcinomas
As shown in Table 1, cell proliferation within lung tumors, as indicated by proliferating cell nuclear antigen staining, was markedly reduced, and apoptosis, as determined by TUNEL and caspase-3 cleavage, was significantly enhanced after ZD6474 treatment. To differentiate between tumor and tumor-associated endothelial cell proliferation and apoptosis, double immunofluorescence studies of CD31/bromodeoxyuridine, CD31/caspase-3, and CD31/TUNEL were completed for the lung cancer specimens. ZD6474 treatment resulted in decreased proliferation (CD31/bromodeoxyuridine) and increased apoptosis (CD31/TUNEL and CD31/caspase-3) for both tumor and tumor-associated endothelial cells in H441 tumors (Fig. 7A and B). The ability of ZD6474 to enhance endothelial cell apoptosis and impair endothelial cell proliferation strongly suggest that it can induce both antivascular and antiangiogenic effects.
To further define the downstream signaling pathways responsible for the effects of ZD6474, we employed CD31/pAkt and CD31/pMAPK fluorescent double-labeling techniques. Treatment with ZD6474 was associated with an inhibition of phosphorylation of Akt and MAPK in both tumor (green) and tumor-associated endothelial (Fig. 6B, yellow) cells. We next carried out a time course experiment to analyze the kinetics of the effects of ZD6474 upon proliferation and apoptosis in H441 lung adenocarcinoma tumors. Treatment with ZD6474 began on day 15 after tumor injection, and lung tumors were harvested after 1, 3, 5, 7, or 9 days of treatment. As shown in Fig. 7C, ZD6474 exerted an antiproliferative effect upon tumor endothelial cells as early as day 3 after therapy that was maximal on day 5. The effect of ZD6474 upon endothelial cells preceded the antiproliferative effect upon tumor cells by 2 days. ZD6474 treatment exerted an apoptotic effect for both endothelial and tumor cells within 24 h of treatment initiation. Taken together, these data show that blocking the activation of VEGFR2 and EGFR targets both tumor and tumor-associated endothelial cell proliferation and survival.
Angiogenesis, the growth of new capillary blood vessels from preexisting vessels, is a fundamental process that is required for a wide variety of physiologic and pathophysiologic processes (33). In addition to providing nutrients and oxygen and removing catabolites, proliferating endothelial cells found in and around tumors produce multiple growth factors that can promote tumor cell growth, invasion, and survival (34, 35). Angiogenesis, therefore, provides both a perfusion effect and a paracrine effect to a growing tumor (33), and tumor and endothelial cells can drive each other and perpetuate and amplify the malignant phenotype. These observations led Folkman to propose a two-compartment model of malignancy consisting of endothelial cell and tumor cell populations (36). Both of these compartments offer potential targets for anticancer strategies, and it will be prudent to target both the cancer cells and the angiogenic endothelium, especially given the heterogeneity of the vasculature and microenvironment (37, 38) within human tumors, and that malignant tumors can produce multiple angiogenic factors (39). In the current study, we found that both the VEGF/VEGFR2 and EGF/EGFR axes can directly influence the tumor and endothelial compartments in a human lung cancer model via autocrine and paracrine mechanisms, and that targeting these pathways can prevent and reverse lung cancer progression.
The effects of VEGFR, EGFR, and other tyrosine kinase inhibitors alone and in combination have typically been studied in s.c. xenograft models of tumor growth. However, xenograft models do not adequately address the interaction between the specific organ environment and tumor cells (40). Furthermore, gene expression for tumors is markedly altered when they are implanted s.c. compared with orthotopically (41). Indeed, we recently reported that treatment with paclitaxel was effective against human lung cancer cells growing as s.c. in mice but was virtually ineffective when the same cells were growing in the lung (15). In the current study, double immunofluorescence staining revealed overexpression and activation of VEGFR2 and EGFR for tumor and tumor-associated endothelial cells in clinical specimens of human lung adenocarcinoma. VEGF and VEGFR2 and EGF/TGF-α and EGFR in the tumor tissues may contribute to an autocrine and paracrine loop resulting in the co-expression of activated VEGFR2 and EGFR in both tumor and tumor-associated endothelial cells. Thus, we studied the effects of combined blockade of VEGFR2 and EGFR upon human lung adenocarcinoma growing in the lungs of the nude mice. Although our model is limited by the use of immunocompromised mice to allow for the growth of the human cancer cells, our data show that that ZD6474 therapy is highly effective in the lung microenvironment. We are currently expanding our studies of ZD6474 as part of the management of lung cancer to include metastatic disease in models of lung cancer brain and bone metastasis.
Recent studies show that response to monotherapy with gefitinib and other small-molecule inhibitors of the EGFR tyrosine kinase was associated with EGFR mutations of tumor cells (42, 43), and that further mutations of EGFR can induce resistance to these agents (44). Evidence also suggests that EGFR gene copy number may contribute to resistance to EGFR antagonism (45). In preclinical models of renal cancer, tumor response to treatment with an EGFR inhibitor was dependent upon and correlated with tumor cell expression of EGF and/or TGF-α (11). Expression of the ligands EGF and/or TGF-α induced expression of EGFR on tumor-associated endothelial cells, thereby enhancing tumor sensitivity to the treatment. In our studies, H441 human lung adenocarcinomas, which express wild-type EGFR, growing in the lungs of mice responded to treatment with ZD6474 primarily targeting VEGFR2 and EGFR with significant inhibition of tumor progression and improved survival of mice without significant toxicity. Double immunofluorescence staining revealed that ZD6474 therapy blocked activation of VEGFR2 and EGFR for tumor and tumor-associated endothelial cells of human lung adenocarcinomas growing in mice. Both EGFR and VEGFR2 expression by tumor-associated endothelial cells, therefore, provided two major targets for therapy.
Although the effects of VEGF via its receptors were initially thought to be specific for the vasculature, recent data show that VEGF can also play a role in multiple other processes, including hematopoiesis (46), immune function (47), hepatic integrity (48), recruitment and mobilization of endothelial cell progenitors (49), and neurologic function (50). Our data show that human lung adenocarcinoma cells can, in some cases, express VEGFR2 and suggest that VEGF may play a direct role in tumor cell survival. Tumor cell expression of VEGFRs suggests that anti-VEGF strategies could have both direct and indirect effects upon both a tumor and its vasculature and provides an effective strategy for human lung cancer therapy. It will therefore be critical to screen for the expression of both the ligand(s) and receptor(s) by tumor and tumor-associated endothelial cells when designing therapeutic strategies for biologically targeted agents.
Lung cancer is a particularly heterogeneous disease, and its successful treatment will likely require a multiple targeted approach. The simultaneous blockade of VEGFR and EGFR pathway exerts significant antitumor effects in human lung adenocarcinoma with effects on both tumor and tumor-associated endothelial cells supporting a clinical strategy for the use of targeted therapies against the vasculature for lung cancer. In our studies of H441 lung adenocarcinomas, activated VEGFR2 and EGFR were detected only in tumor tissues and not in the vasculature adjacent to the tumor and endothelial cell proliferation, and apoptosis was observed only in tumor-associated endothelial cells and not in the vasculature of the adjacent lung. These data suggest that differences between the dividing tumor-associated endothelial cells and quiescent endothelial cells provide an opportunity for selective therapeutic intervention. However, in clinical trials with agents that target VEGF, hemorrhage and thrombosis have been observed as complications that may have been due in part to disruption of the normal vascular integrity (8). Furthermore, although our data with ZD6474 suggest that the quiescent vasculature of the lung may not be targeted, the interface between normal lung and tumor vasculature may be targeted. Thus, therapy with these agents will have to be monitored closely and optimized for each patient with careful attention paid to the interaction between the endothelial and tumor cells within the tumor microenvironment.
Grant support: The University of Texas M. D. Anderson Cancer Center Physician Scientist Program Awards (M.S. O'Reilly and R.S. Herbst), Department of Defense grant DAMD 17-02-1-0706, and AstraZeneca.
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.