Molecular Targets on Blood Vessels for Cancer Therapies in Clinical Trials
Molecular Targets on Blood Vessels for Cancer Therapies in Clinical Trials
This review covers progress to date in the identification of molecular targets on blood vessels in cancers, as well as agents that act on those targets, with emphasis on those currently in clinical trials. Current vascular-targeting therapies comprise two general types—antiangiogenic therapy and antivascular therapy. Advances in antiangiogenic therapies, particularly inhibitors of vascular endothelial growth factors and their receptors, have clarified the capacity of these inhibitors to change tumor-associated vessel structure to a more normal state, thereby improving the ability of chemotherapeutics to access the tumors. The responses of other antiangiogenesis target molecules in humans are more complicated; for example, αvβ3 integrins are known to stimulate as well as inhibit angiogenesis, and cleavage of various extracellular proteins/proteoglycans by matrix metalloproteinases produces potent regulators of the angiogenic process. Antivascular therapies disrupt established blood vessels in solid tumors and often involve the use of ligand-based or small-molecule agents. Ligand-based agents, irrespective of the antiangiogenic capacity of the ligand, target antivascular effectors to molecules expressed specifically on blood vessels, such as aminopeptidase N, fibronectin extra-domain B, and prostate-specific membrane antigen. Small-molecule antivascular agents, which are not targeted to molecules on blood vessels, rely on physical differences between the vasculatures in tumors and those in normal tissues.
Targeting therapeutic agents specifically to disease lesions, as opposed to systemic delivery, can reduce toxicity and enhance therapeutic efficacy. Despite these advantages, such specificity remains an elusive goal in molecular medicine. The tumor vasculature is an attractive target in anticancer therapy for several reasons: (1) it is easily accessed by a variety of therapeutic molecules, (2) it represents a vital route for delivery of oxygen and other nutrients essential for tumor growth, (3) it is one of the main routes of metastatic spread, (4) the genetic stability of vascular endothelial cells (in comparison with tumor cells) makes them attractive targets for long-term therapy, and (5) tumor vasculature–specific selection markers should make this form of treatment applicable to a range of solid tumors. Thus, extensive effort has been devoted to the discovery of novel therapeutic agents that can target the tumor vasculature.
Current vascular-targeting therapies are of two general types: antiangiogenic agents, which block endothelial cell survival and prevent the supply of blood to tumors, and antivascular agents, which disrupt established blood vessels in solid tumors. Two types of antivascular agents are currently being studied: ligand-directed agents and other small molecules. Success in animal models targeting a variety of molecules expressed on blood vessels within tumors has led to testing of several vascular-targeting agents in clinical trials. Here we review progress in the identification of molecular targets on vasculature in cancers, as well as agents that act on those targets, with particular emphasis on those currently in clinical trials.
Target Molecules for Antiangiogenesis Therapy
Most of the molecules identified as targets on tumor blood vessels are associated with angiogenesis and stromal reorganization. Moreover, many are expressed by endothelial cells, and some are produced by the tumor cells themselves, by pericytes, or by components of the extracellular matrix. A variety of vascular-targeting agents that affect cancer through the inhibition of angiogenesis are currently in clinical trials (Table 1) and are reviewed in the following paragraphs.
Vascular Endothelial Growth Factors and Their Receptors
The main targets of many antiangiogenic drugs are the vascular endothelial growth factors (VEGFs) and their receptors (VEGFRs). A single administration of the humanized monoclonal antibody against VEGF, bevacizumab (Avastin), decreased tumor perfusion, vascular volume, microvascular density, and interstitial pressure within tumors of rectal carcinoma patients. Batchelor et al recently reported that daily doses of cediranib (AZD2171), a tyrosine kinase inhibitor of VEGFRs, led to decreases of vessel size and blood flow in tumors, and alleviation of edema in patients with recurrent glioblastoma. Given in combination with standard chemotherapy, bevacizumab has been shown to prolong survival in patients with metastatic colon, breast, and lung cancer.[4,5]
These findings support the concept that inhibition of VEGF transiently changes atypical vascular structure and permeability to a more normal state, effects that reduce interstitial pressure within tumors and thereby improve the access of oxygen or chemotherapeutics to the tumors. Therefore, it is indicated that anti–VEGF-signaling agents should be combined with cytotoxic chemotherapy to induce a tumor response sufficient for improved survival of patients. Alternatively, in addition to the blocking of VEGF signaling, inhibition of other growth factor signaling pathways by small-molecule multikinase inhibitors (Table 1) would be required. Indeed, treatment with sorafenib (BAY 43-9006, Nexavar), prolongs progression-free survival in patients with advanced renal cell carcinoma.[7,8]
Integrins mediate cellular attach ment to the extracellular matrix and subsequent signal transduction. Many types of integrins—eg, αvβ3, αvβ5, α5β1, and α6β4—are enhanced on growing endothelial cells and/or tumor cells, and on cancer-associated but not mature endothelia. Such observations have indicated their potential in the targeting of both endothelial and tumor cells. Systemic administration of anti-αvβ3 antibody or peptide inhibitors of αvβ3 integrin has been shown to block tumor angiogenesis and to reduce the growth and invasiveness of human tumors in animals.[10,11] However, findings from mouse models of genetic ablation were less clear—eg, angiogenesis was not blocked in mice lacking β3, β5, or β6 integrin. Moreover, angiogenesis was enhanced in mice lacking αv integrin, results indicating that αvβ3 integrin could have both positive and negative regulatory roles in various phases of angiogenesis. Phase II trials are currently underway to test a humanized anti-αvβ3 antibody (Abegrin, formerly known as Vitaxin) for stage IV metastatic melanoma and androgen-independent prostate cancer that metastasizes to bone.
The αvβ3 integrin serves as a receptor for a variety of extracellular matrix proteins, including vitronectin, fibronectin, fibrinogen, thrombospondin 1, laminin, and several collagens, via their exposed arginine-glycine-aspartic acid (RGD) tripeptide sequence.[13,14] Cyclic RGD-containing peptides are the most commonly studied integrin inhibitors. One such peptide, cilengitide (EMD 121974, an N-methylated form of cRGDfV), selectively inhibits the binding of vitronectin to the αvβ3 receptor. In preclinical mouse models, daily administration of cilengitide led to shrinkage of orthotopically-implanted human brain tumor cells in mice and prolonged the ir survival. Phase II studies of cilengitide for patients with glioblastoma or prostate cancer are ongoing.
Our group has developed a novel vehicle for gene therapy that consists of a chimerized adeno-associated virus and single-stranded bacteriophage (AAVP). The prototype was confirmed to target αv integrin exposed on the epithelial cell surface and to inhibit tumor growth in mice.
Previously, matrix metalloproteinases (MMPs) were thought to act primarily by their promotion of the migration or invasion of tumor cells, stromal cells, or epithelial cells via degradation or reorganization of tumor-associated stroma and basement membrane. These characteristics make MMPs attractive targets for treatment of late-stage (metastatic) disease. However, phase III trials of broad-spectrum MMP inhibitors for the most part have not resulted in extended survival, because the effects of MMPs are more complex than were anticipated.
Proteolysis by MMPs regulates aspects of cell signaling that control the homeostasis of the extracellular environment. For example, cleavage of various substrates by MMPs can give rise to angiostatin and endostatin, both potent inhibitors of angiogenesis. Therefore, selecting particular MMPs as targets in the treatment of cancer will require the knowledge that a particular MMP acts specifically on the target (and not on related molecules that could be considered "antitargets," ie, molecules that could have deleterious effects if blocked or disrupted).
Our work in this regard focuses on MMP-2 and MMP-9, which are present on the tumor epithelium and in the stroma of several types of cancer, including that of breast, colon, skin, and lung. We showed that a cyclic peptide containing the sequence HWGF, isolated from in vivo phage display screening, selectively inhibits MMP-2 and MMP-9 but not several other MMPs. Moreover, our prototype synthetic peptide CTTHWGFTLC has been shown to prevent tumor growth and invasion in animal models, to improve survival of mice bearing human tumors, and to block migration of endothelial cells.
Aminopeptidase N (APN, also known as CD13) is distributed widely, yet it plays tissue-specific roles that depend on differential cleavage of various biologically active peptides such as enkephalins, endorphins, and angiotensins. We found APN to be expressed exclusively on endothelial cells in mouse and human tissues undergoing angiogenesis, but not on those in the absence of angiogenesis.[21,22] Others have shown that 61% of 62 clinical samples of squamous cell carcinomas of the lung showed expression of APN in interstitial cells, including fibroblasts and endothelial cells, but not in the tumor cells themselves. Mice deficient in the APN gene grow normally but exhibit an impaired angiogenic response under pathologic conditions.
APN antagonists such as anti-APN antibodies and bestatin have shown antiangiogenic effects and have led to reduced tumor volume in breast cancer models. APN is therefore an attractive target, not only for antiangiogenic therapy but also attractive for ligand-based therapy. In comparison to αvβ3 integrin, we have shown that APN is an alternative and more specific binding target than for peptides containing the NGR motif. Fusing the anticancer agent tumor necrosis factor (TNF) with the peptide CNGRC, an APN ligand efficiently targets activated blood vessels and shows antitumor effects in murine models. This ligand-directed drug is currently in phase I trials (Table 1).