Inhibitors of Tumor Angiogenesis
Kari B. Wisinski
William J. Gradishar
Neoplasms occur when a normal cell acquires molecular changes that allow the cell to divide and metastasize in an unsupervised manner. This process requires a supply of nutrients, which is provided by the vasculature. In the early 1970s, Judah Folkman first proposed the theory of tumor angiogenesis, which he defined as the process of recruitment of new vessels to a tumor as it grows.1 His hypothesis was that tumors are unable to grow beyond a certain size in the absence of a new vascular supply. Subsequent data have supported this hypothesis, indicating that angiogenesis is a critical step in tumor growth and metastases.2,3 A concentrated effort is now underway to target angiogenesis as a component of cancer therapy.
The growth and maturation of new vasculature is a highly complex process involving interactions of multiple cellular pathways and communication between cells and the extracellular matrix. Many of the key molecules have been identified and have become potential therapeutic targets. As one of the first steps in this process, tumors produce signaling molecules that activate angiogenesis or down-regulate the expression of inhibitors of angiogenesis. This step alters the microenvironment to favor an “angiogenic switch.”4,5 This environment supports the proliferation of endothelial cells. In the normal microenvironment, endothelial cells are usually quiescent. In contrast, once the “angiogenic switch” occurs, proliferating endothelial cells form new blood vessels. These blood vessels are structurally abnormal with a leaky basement membrane, which allows tumor cells to become integrated into the vessel wall. This neo-vasculature is then responsible for the supply of nutrients to the tumor. Angiogenesis has been considered an ideal target for anti-neoplastic therapy, since it is a unique process that occurs only at tumor sites and not in normal human organs, a fact that may limit the toxicities related to these agents.
Three Food and Drug Association (FDA)-approved drugs target angiogenesis. This chapter reviews the process of angiogenesis and the rationale for the development of these angiogenesis inhibitors as antineoplastic therapies. The chemistry, pharmacology, and indications for each of the current FDA-approved angiogenesis inhibitors will be covered. In addition, an extensive discussion of the clinical studies leading to the FDA’s approval will be presented. Finally, the common toxicities seen with these agents and their limitations in the treatment of malignancies will be discussed.
Tumor Angiogenesis
Since Judah Folkman initially proposed the concept of tumor angiogenesis in 1971, much work has been done to elucidate the key players in this critical event. Initiation of tumor angiogenesis leads to a cascade of events that must occur in concert in order for successful tumor angiogenesis to occur. Without angiogenesis, tumors cannot exceed a few cubic millimeters in size.2
Angiogenic Switch
The normal microenvironment has an anti-angiogenic disposition. As tumors grow they are initially dependent on the normal microenvironment for oxygen and other nutrients. However, since the diffusion limit of oxygen is around 100 μm, as tumors expand, those cells that are no longer within the oxygen diffusion perimeter become hypoxic.6 In the setting of hypoxia, multiple intracellular molecules are activated. One well-characterized result of hypoxia is the activation of the hypoxia inducible factor-1 (HIF-1) transcriptional complex.7 Activation of this complex leads to up-regulation of pro-angiogenic factors, such as vascular endothelial growth factor (VEGF), platelet-derived growth factor (PDGF), and nitric oxide synthase (NOS), which are released into the microenvironment. Consistent with this finding, HIF-1 is up-regulated in many cancers.7 However, like most pathways in oncogenesis, redundancy is present and the HIF system is also influenced by many oncogenic pathways, including the insulin-like growth factor-1, epidermal growth factor (EGF), mutant Ras, and Src kinase pathways as well as tumor suppressor mutations, including PTEN, p53, p14ARF, and pVHL (von Hippel-Lindau).7 The activation of HIF-1 is a critical step in initiation of the “angiogenic switch” in which the normal microenvironment changes from anti-angiogenic to pro-angiogenic and new blood vessel formation is started. Drugs targeting HIF-1 have been developed for clinical use and are currently being tested in early phase trials.
The Vascular Endothelial Growth Factor (VEGF) and VEGF Receptor Family
Activation of the VEGF pathway is critical in both physiologic and pathologic angiogenesis. The complexities of this system are beyond the scope of this chapter; therefore, only a brief overview is provided. Detailed reviews are available elsewhere.8, 9, 10 The VEGF receptor family is made up of three tyrosine kinase receptors. Of these, only VEGFR-1 and VEGFR-2 bind the main ligand important in tumor angiogenesis, VEGF-A.8,10 VEGF-A, usually referred to as VEGF, belongs to a family of genes including placenta growth factor (PLGF), VEGF-B, VEGF-C, and VEGF-D. The biology of VEGF and its primary receptors, VEGFR-1 and VEGFR-2, will be discussed in this chapter (Table 26-1).
TABLE 26.1 The VEGF signaling family | ||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ||||||||||||||||||||||||||||||
The two circulating isoforms of VEGF, VEGF121 and VEGF165, are the major mediators of tumor angiogenesis.9 One of the main inducers of VEGF expression is tumor hypoxia.11 This is mediated in part by HIF-1.12 Interestingly, loss of the von Hippel-Lindau (vHL) tumor suppressor gene in renal cell carcinomas results in constitutive overexpression of HIF-1 leading to elevated VEGF expression in these tumors.13 Several other major growth factors including EGF, transforming growth factors alpha (TGF-α) and beta (TGF-β), keratinocyte growth factor, insulin-like growth factor-1, fibroblast growth factor (FGF), and PDGF also have been shown to up-regulate VEGF mRNA expression.8 In addition, oncogenic mutations or amplifications of the Ras oncogene and inflammatory cytokines (IL-1α and IL-6) are associated with VEGF gene induction.14,15 VEGF production can also originate from tumor-associated stromal cells9 suggesting that paracrine or autocrine release of these factors cooperate with hypoxia to induce tumor angiogenesis via VEGF upregulation. Vascular permeability is also mediated by VEGF, which forms fenestrations in blood vessels.10 Therefore, agents targeting VEGF are also thought to exert antitumorigenic effects by allowing improved delivery of other agents, such as chemotherapy, to tumors.16
The VEGF-1 and VEGF-2 receptors are expressed normally on vascular endothelial cells engaged in angiogenesis as well as on bone marrow-derived cells. In addition, many solid and hematologic tumors also express the VEGF receptors, primarily VEGFR-1.9 These receptors have a similar protein structure with seven immunoglobulin-like domains in the extracellular region, a single transmembrane region, and a consensus tyrosine kinase sequence that is interrupted by a kinase-insert domain.8 Interestingly, the VEGF receptors can also be expressed inside the cell, where they can promote cell survival by an “intracrine” mechanism.9
VEGFR-1 (also known as Flt-1) was the first identified VEGF receptor; however, its role in angiogenesis remains controversial (Table 26-1). The data suggest that the function of VEGFR-1 differs depending on the developmental stage of the animal, the cell type on which it is being expressed, and the ligand to which it is binding (VEGF-A, VEGF-B, or PLGF).8,9 Studies linking VEGFR-1 to angiogenesis have found that expression of VEGFR-1 is up-regulated by hypoxia. Similar to the VEGF, this is mediated by HIF-1. However, VEGFR-1 has also been identified as a negative regulator of VEGF, especially in embryonic development. In addition, its signal transduction properties are very weak. In general, VEGFR-1 is thought to have little role in tumor angiogenesis.
VEGFR-2 (also known as KDR or Flk-1) is the main regulator of endothelial cell proliferation and survival, as well as vascular permeability (Table 26-1).10 Upon VEGF ligand binding, the receptor undergoes dimerization and tyrosine phosphorylation. In endothelial cells, this VEGFR-2 activation leads to initiation of several signaling cascades.9 This includes activation of the mitogenactivating protein kinase (MAPK) pathway, which involves phospholipase C-γ (PLC γ), protein kinase C (PKC), Raf kinase, and MEK. Activation of this pathway results in DNA synthesis and cell growth. Other pathways initiated by VEGFR-2 activation are the phosphatidylinositol 3′-kinase (PI3K)/Akt pathway that leads to increased endothelial cell survival and the Src family pathway that results in cell migration. Many of these pathways are abnormally expressed in tumors and have been targeted with specific inhibitors. In addition, the neuropilin (NRP) receptors also bind to VEGF and can act as co-receptors with VEGFR-2 to regulate angiogenesis; therefore, these receptors may be targeted by novel agents in the future.9
Other Molecular Pathways Important in Tumor Angiogenesis
Matrix metalloproteinases (MMPs) are a family of structurally related zinc-containing endopeptidases that are involved in the degradation of extracellular matrix components (ECM).17 In angiogenesis, MMPs mediate remodeling and invasion of the ECM by new vessels by regulating endothelial cell attachment, proliferation, and migration. These observations suggest that matrix metalloproteinase inhibitors (MMPIs) could inhibit tumor progression at both the primary tumor site and sites of metastases. Several MMPIs have been developed for clinical use, but results in phase III clinical trials of patients with advanced malignancy were disappointing.18,19
Interaction between endothelial cells and the extracellular matrix is critical to tumor angiogenesis. The integrins are transmembrane receptors that bind to extracellular matrix proteins and play a significant role in this interaction. Studies have implicated a number of endothelial cell integrins in the regulation of endothelial cell growth, survival, and migration during angiogenesis.20 In addition, integrin signaling is dysfunctional in cancer cells and their expression may correlate with prognosis.21 Several inhibitors of these endothelial cell integrins have been developed and tested in early phase clinical trials.20 Clinical studies indicate signs of potential therapeutic benefit from these agents; however, further testing is still needed.
Another receptor tyrosine kinase pathway involved in tumor angiogenesis involves the tie-2 receptor.9 This receptor is expressed principally on the vascular endothelium. Its major ligands are angiopoietin-1 (ang-1) and angiopoietin-2 (ang-2). In concert with VEGF, this pathway stabilizes and matures new capillaries. Ang-1, ang-2, and tie-2 expression have been correlated with prognosis in several cancers, including early stage bladder cancers and breast
cancer.22,23 Peptibodies against ang-2 have been developed and are in early clinical development.
cancer.22,23 Peptibodies against ang-2 have been developed and are in early clinical development.
Another pathway that also appears to play an important role in tumor angiogenesis is the notch receptor pathway. The notch receptors are cell surface receptors implicated in cell fate, differentiation, and proliferation.9,24 The ligands for these receptors are the transmembrane proteins, jagged and delta-like ligand (Dll), which are expressed on adjacent cells. Vascular endothelial cells express notch 1 and 4, as well as jagged 1, Dll-1, and Dll4. Notch-Dll4 signaling is essential for vascular development in the embryo and Dll4 is up-regulated in tumor vasculature. This is thought to be in part VEGF-mediated. Novel agents targeting this pathway are also currently in early phase clinical development.
Preclinical Development of VEGF Inhibitors
VEGF and its receptors are overexpressed in a number of tumor types, including colorectal, lung, breast, renal cell, and endometrial carcinomas, as well as hematologic malignancies, such as acute myelogenous leukemia.8,9 The majority of retrospective studies correlating VEGF or VEGFR expression to prognosis suggest that high levels are associated with a worse prognosis.25, 26, 27 Efforts to determine the prognostic significance of another marker of angiogenesis, microvascular density have had conflicting results, in part secondary to lack of standardized laboratory techniques. However, it appears that high microvascular density is a poor prognostic factor in several malignancies.28,29
In vitro and in vivo studies indicate the importance of VEGF signaling in enhancing proliferation and inhibiting apoptosis of endothelial cells. In addition to tumor angiogenesis, VEGF signaling is critical in developmental endothelial cell growth and in endothelial cell growth after normal tissue injury. However, established blood vessels are not VEGF-dependent. Therefore, agents targeting this pathway should have effects only in the peritumoral vasculature and not on the established vasculature critical for normal human physiology. Based on the data indicating the importance of VEGF signaling in tumor angiogenesis, animal studies were done and demonstrated that inhibition of VEGF signaling interrupts tumor growth and invasion.30, 31, 32, 33 The abundance of preclinical data supporting the antineoplastic potential of these agents led to the development of multiple drugs targeting the VEGF pathway.
Antibodies to Vascular Endothelial Growth Factor
Bevacizumab (Avastin)
Chemistry
Bevacizumab is a recombinant humanized monoclonal immunoglobulin G (IgG) antibody generated by engineering VEGF binding residues of a murine neutralizing antibody into the framework of a normal human IgG.30 It consists of approximately 93% human and 7% murine protein sequences.34 Bevacizumab has a molecular weight of approximately 149 kD (kilodaltons). It comes in a clear to slightly opalescent, colorless to pale brown, sterile, pH 6.2 solution for intravenous infusion.
Mechanism of Action
Bevacizumab binds and neutralizes the biologically active forms of VEGF by recognizing the binding sites for the VEGF receptors. This prevents the interaction of VEGF with its receptors on the surface of endothelial cells. Bevacizumab neutralizes all isoforms of human VEGF with a dissociation constant (Kd) of 1.1.34 Inhibition of VEGF-induced proliferation of endothelial cells as well as tumor angiogenesis in vitro have been noted with bevacizumab. Preclinical studies showed that the administration of bevacizumab to mice with various tumor xenografts caused growth inhibition and decreased metastatic progression. Given that VEGF is produced and acts locally in a tumor site, the capacity of an antibody to interfere with its effects is somewhat surprising. Unlike antibodies targeting receptors, ligands can be difficult to sop up.
Absorption, Distribution, and Elimination
Bevacizumab is administered as an intravenous formulation. In a population pharmacokinetic analysis of 491 patients who received 1 to 20 mg/kg of bevacizumab weekly, every 2 weeks, or every 3 weeks, the estimated half-life of bevacizumab was approximately 20 days (range 11 to 50 days).35 The predicted time to reach steady state was 100 days. The clearance of bevacizumab varied with body weight, sex, and tumor burden. Bevacizumab has been tested at doses from 3 to 20 mg/kg, but a clear dose-response relationship has not been identified. In colorectal cancer, a dose of 5 mg/kg was more effective than 10 mg/kg.36 On the other hand, in renal cell cancer and non-small cell lung cancer, higher doses up to 15 mg/kg are more effective than lower doses.37,38 Demographic data suggest that no dose adjustments are necessary for age or sex. No studies have been conducted to examine the pharmacokinetics of bevacizumab in patients with renal or hepatic impairment.
Drug Interactions
Bevacizumab has no known drug interactions. However, in a phase I study of bevacizumab and sunitinib in advanced renal cell carcinoma, two patients developed severe microangiopathic hemolytic anemia associated with hypertension, thrombocytopenia, renal insufficiency, and hemolysis. Evidence of less severe microangiopathic hemolytic anemia was also noted in three other patients on the study.39 This combination should not be used outside of a clinical trial.
Pregnancy and Breast Feeding
Bevacizumab is a pregnancy category C agent. There are no studies of bevacizumab in pregnant women. Human IgG is known to cross the placental barrier; therefore, bevacizumab has the potential to cause fetal harm. Reproduction studies in rabbits treated with approximately 1 to 12 times the recommended human dose of bevacizumab resulted in teratogenicity, including an increased incidence of fetal skeletal alterations. It is not known whether bevacizumab is secreted in human milk. Although human IgG is excreted in human milk, some data suggest that breast milk antibodies do not enter the neonatal and infant circulation in substantial amounts. Recommendations for bevacizumab therapy during nursing must take into account the long half-life of the drug.
Special Considerations for Pediatric and Geriatric Populations
The safety, efficacy, and pharmacokinetic profile of bevacizumab in pediatric patients have not been established. Bevacizumab has been
administered in the elderly population and no dose modifications are recommended in this group. The section on special populations has further details regarding the data on bevacizumab use in these groups of patients.
administered in the elderly population and no dose modifications are recommended in this group. The section on special populations has further details regarding the data on bevacizumab use in these groups of patients.
Therapeutic Uses
Renal Cell Carcinoma
Renal cell carcinoma was felt to be a potentially promising target disease for bevacizumab because most renal clear cell carcinomas have a mutation in the vHL tumor suppressor gene which leads to HIF-1-mediated VEGF production.13 A randomized phase II trial of single-agent bevacizumab versus placebo was conducted in patients with metastatic renal clear cell carcinoma who had progressed after, or could not receive interleukin-2. A total of 116 patients were randomly assigned to placebo, bevacizumab 3 mg/kg every 2 weeks, or bevacizumab 10 mg/kg every 2 weeks. No responses were noted in the low-dose group, while a partial response was seen in 10% of patients in the higher dose group. Toxicities included proteinuria, hypertension, and epistaxis. There was a difference between PFS in the high-dose group compared with the placebo group (4.8 versus 2.5 months; P < 0.001), but no improvement in overall survival occurred in any of the cohorts.40 Subsequent studies have been undertaken with various combination therapies.
TABLE 26.2 Food and drug administration (FDA)-approved drugs targeting the VEGF signaling pathway | |||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| |||||||||||||||||||||||||||
The most promising combination in renal cell carcinoma has been with interferon. A multicenter, randomized, double-blind, placebo-controlled phase III study of 649 patients with previously untreated metastatic renal cell carcinoma was performed. Patients were randomly assigned to either interferon alpha-2a (9 million international units subcutaneously three times weekly) and bevacizumab (10 mg/kg every 2 weeks) or interferon alone.37 Progression-free survival (PFS) was significantly longer in the combination group (10.2 versus 5.4 months; P = 0.0001). The combination group had a higher response rate as well (31% versus 13%; P = 0.0001). Hypertension (26% versus 9%), headaches (23% versus 16%), venous thromboembolism (3% versus <1%), and bleeding (33% versus 9%) of any grade were more frequent in the bevacizumab combination arm than with interferon alone. Adverse events included grade 3 or 4 gastrointestinal perforations or thromboembolic events in 1% and 3% of bevacizumab-treated patients, respectively (Table 26-2).
Other combinations with bevacizumab have also been evaluated in renal cell carcinoma. Thalidomide plus bevacizumab was not found to be effective.41 A phase II trial combined bevacizumab with erlotinib, an epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor.42 All 63 patients were treated with bevacizumab 10 mg/kg every 2 weeks and erlotinib 150 mg orally daily. Of 59 evaluable patients, a partial response was seen in 15 (25%), and
61% had stable disease after 8 weeks of treatment. Median survival was not reached at 15 months. Grade 3 or 4 toxicities included hypertension, diarrhea, and rash. Although this was a well-tolerated regimen, a randomized phase II follow-up did not show any additional benefit in PFS with the addition of erlotinib to bevacizumab.43 Thus, the best data for bevacizumab treatment in advanced renal cell carcinoma remains in combination with interferon alpha.
61% had stable disease after 8 weeks of treatment. Median survival was not reached at 15 months. Grade 3 or 4 toxicities included hypertension, diarrhea, and rash. Although this was a well-tolerated regimen, a randomized phase II follow-up did not show any additional benefit in PFS with the addition of erlotinib to bevacizumab.43 Thus, the best data for bevacizumab treatment in advanced renal cell carcinoma remains in combination with interferon alpha.
Colorectal Cancer
Bevacizumab was approved by the FDA in 2004 for use with 5-fluorouracil (5-FU)-based chemotherapy as first-line therapy for patients with metastatic colorectal cancer. This approval was based on two clinical trials in patients with previously untreated metastatic disease.36,44 In the initial randomized phase II trial, 104 patients were assigned to one of the three treatment regimens: 5-FU/leucovorin (LV), 5-FU/LV plus bevacizumab 5 mg/kg every 2 weeks, or 5FU/LV plus bevacizumab 10 mg/kg every 2 weeks.36 The 5-FU/LV in all groups was given weekly for the first 6 weeks of an 8-week cycle. Although a difference in overall median survival was detected, it was not statistically significant. The time to progression was improved in the 5 mg/kg bevacizumab group when compared with the control group (9 versus 5.2 months; P = 0.005), as was the overall response rate (40% versus 17%; P = 0.029). However, the time to progression and overall response in the cohort that received the higher dose bevacizumab were no better than for the control group.
The phase III trial randomly assigned 411 patients to IFL (irinotecan, bolus 5-FU, and LV) and 402 patients to IFL plus bevacizumab 5 mg/kg every 2 weeks.44 A third arm with 5-FU/LV plus bevacizumab was discontinued after a preplanned safety analysis indicated that the IFL + bevacizumab regimen was safe. The group that received bevacizumab had a better overall response rate (44.8 versus 34.8; P = 0.004) and improved PFS. The study met its primary endpoint with in improvement in median survival from 15.6 months in the IFL cohort to 20.3 months in the IFL plus bevacizumab group (P < 0.001). Grade 3 or 4 adverse events were significantly increased with bevacizumab when compared to placebo including hypertension (11% versus 2.3%), leucopenia (37% versus 31.1%), and diarrhea (32.4% versus 24.7%) (Table 26-3). This study led to FDA approval of bevacizumab with 5-FU-based chemotherapy as first-line treatment for metastatic colorectal cancer (Table 26-2).
TABLE 26.3 Rates of grade 3 or higher adverse events reported in large phase III studies with bevacizumab | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
