This Article Abstract Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Pepper, M. S. Search for Related Content PubMed PubMed Citation Articles by Pepper, M. S.

(Arteriosclerosis, Thrombosis, and Vascular Biology. 1997;17:605-619.)
© 1997 American Heart Association, Inc.

Articles

Manipulating Angiogenesis

From Basic Science to the Bedside

Michael S. Pepper

From the Department of Morphology, University Medical Center, Geneva, Switzerland.

Correspondence to Michael S. Pepper, MD, PhD, Département de Morphologie, Centre Médical Universitaire, 1 rue Michel-Servet, 1211 Genève 4, Switzerland. E-mail michael.pepper{at}medecine.unige.ch

	Abstract

Top Abstract Introduction Mechanisms of Angiogenesis Clinical Applications of... Perspectives for the Future References

 
Abstract Considerable progress has been made recently in understanding the molecular mechanisms of angiogenesis, which like most other biological processes is the result of subtle and often complex interactions between molecules that have regulatory (eg, cytokines and their receptors) and effector (eg, extracellular matrix, integrins, and proteases) functions. The title of this review was chosen to reflect a recent trend in which knowledge acquired through a molecular/cell biological approach is being rapidly transferred to the clinical setting. As a result, by manipulating angiogenesis either positively or negatively, considerable therapeutic benefit can now be envisaged in physiological and pathological settings in which neovascularization is a prominent component.

Key Words: blood vessels • antiangiogenesis • protease • ischemia • therapeutic angiogenesis • tumor • cytokine

	Introduction

Top Abstract Introduction Mechanisms of Angiogenesis Clinical Applications of... Perspectives for the Future References

 
The establishment and maintenance of a vascular supply is an absolute requirement for the growth of normal and neoplastic tissues, and as might be predicted, the cardiovascular system is the first organ system to develop and become functional during embryogenesis. Two processes are responsible for the formation of new blood vessels, both of which result in the formation of simple, endothelium-lined, capillary-like tubes. The first is vasculogenesis, which is the primary in situ differentiation of endothelial cells from mesodermal precursors, and their subsequent organization into a primary capillary plexus. The second is angiogenesis, which is defined as the formation of new blood vessels by a process of sprouting from preexisting vessels.^{1 2}

On the basis of observations made in the first half of this century, from a developmental perspective it has been suggested that endothelial cells arise either from angioblasts (which differentiate exclusively into endothelial cells) or from hemangioblasts (which have the dual capacity to differentiate into either endothelial or hematopoietic cells).^{3 4} However, although the existence of the angioblast has been well established, definitive proof for the existence of the hemangioblast is still awaited. The formation of capillary-like tubes is implicit in the current definitions of both vasculogenesis and angiogenesis, and it is currently assumed that primary endothelial cell differentiation is strictly limited to vasculogenesis. Furthermore, while angiogenesis occurs both during development and in postnatal life, it is believed that vasculogenesis is limited to early embryogenesis. However, that endothelial "stem cells" may persist into adult life, where they contribute to the formation of new blood vessels through the formation of circulating endothelial precursors (possibly angioblasts), is a novel and exciting possibility that is currently being explored. If this hypothesis does indeed prove to be correct, our current definition of angiogenesis will have to be modified.

In addition to its role during development, angiogenesis is required for the maintenance of functional and structural integrity of the organism during postnatal life.^{2 5} Thus, it occurs during wound healing, in inflammation, in situations of ischemia, and in female reproductive organs (in the ovary before ovulation and during corpus luteum formation; in the placenta and mammary gland during pregnancy). Neovascularization in these situations is tightly regulated and limited by the metabolic demands of the tissues concerned. Angiogenesis also occurs in pathological situations, such as proliferative retinopathy, and is an important constituent of the inflammatory pannus that destroys articular cartilage in rheumatoid arthritis. Uncontrolled angiogenesis is the central mechanism underlying the formation of juvenile hemangiomas, in which the density of newly formed blood vessels appears to greatly exceed the metabolic needs of the tissue concerned.

Much of our interest in angiogenesis comes from the notion that for tumors to grow beyond a critical size, they must recruit endothelial cells from the surrounding stroma to form their own endogenous microcirculation and that this process is driven by the metabolic requirements of the rapidly growing tumor itself.⁶ Thus, during tumor progression, two phases can be recognized: a prevascular phase and a vascular phase. The transition from the prevascular to the vascular phase is referred to as the "angiogenic switch."⁷ The prevascular phase is characterized by an initial increase in tumor growth followed by a plateau, during which the rate of tumor cell proliferation is balanced by the rate of cell death (apoptosis). This phase may persist for many years and can be recognized clinically as carcinoma in situ, which is characterized by few or no metastases. During the vascular phase, which is characterized by exponential growth, tissue invasion, and the hematogenous spread of tumor cells, the rapid increase in tumor growth is largely due to a decrease in the rate of tumor cell apoptosis.^{8 9} An inverse relationship thus exists between tumor dormancy/tumor cell apoptosis and tumor angiogenesis. In a sense, tumor angiogenesis might almost be considered "appropriate," in that newly formed vessels serve to meet the metabolic demands of the rapidly growing tumor. Although angiogenesis may be beneficial to the tumor itself, it is clearly detrimental to the organism, since it is permissive for continued tumor growth and also allows for the dissemination of tumor cells and the formation of metastasis.

In summary, virtually every subspecialty in medicine in one way or another deals with angiogenesis-associated physiological or pathological processes, and without exception, every organ system in the body has many diseases in which angiogenesis is an important component (Table 1). This in itself makes the study of angiogenesis mandatory in both basic science and clinical settings. Yet the study of angiogenesis does not require this justification. As will be detailed in the following section, angiogenesis as biological process is extraordinarily rich and touches on virtually every aspect of modern molecular cell biology.

View this table: [in this window] [in a new window]   Table 1. Angiogenesis Is Relevant to Virtually All Medical and Surgical Disciplines

	Mechanisms of Angiogenesis

Top Abstract Introduction Mechanisms of Angiogenesis Clinical Applications of... Perspectives for the Future References

 
The classic light-microscopic observations of Clark and Clark¹⁰ in transparent tails of living amphibian larvae were among the first to reveal the precise sequence of events leading to the formation of new capillary blood vessels. These and subsequent observations in nondevelopmental settings^{11 12 13 14} have provided a detailed histological account of new blood vessel formation, which can be summarized as follows. In response to a local angiogenic stimulus, endothelial cells of preexisting capillaries or postcapillary venules become "activated." Although the precise molecular consequences of this activation process remain to be clearly defined, it is followed by local vasodilatation, increased vascular permeability, and the accumulation of extravascular fibrin, as well as proteolytic degradation of the basement membrane of the parent vessel. Thin, cytoplasmic processes are then extended from the activated endothelial cells, and directed migration occurs into the surrounding matrix toward the angiogenic stimulus. Migrating endothelial cells elongate and align with one another to form a capillary sprout, and endothelial cell division, which occurs proximal to the migrating tip, further increases the length of the sprout. The solid sprout gradually develops a lumen proximal to the region of proliferation. Contiguous tubular sprouts anastomose at their tips to form a functional capillary loop in which blood flow is soon established. Vessel maturation is accomplished by reconstitution of the basement membrane. Together, these cellular functions contribute to the process of capillary morphogenesis, ie, the formation of patent, endothelium-lined, tubelike structures.

A Balance Between Positive and Negative Regulators
It is usually stated that with the exception of angiogenesis, which occurs in response to tissue injury or in the female reproductive organs, endothelial cell turnover in the healthy adult organism is very low. The maintenance of endothelial quiescence is thought to be due to the presence of endogenous negative regulators, since positive regulators are frequently detected in adult tissues in which there is apparently no angiogenesis. The converse is also true, namely, that positive and negative regulators often coexist in tissues in which endothelial cell turnover is increased. This has led to the notion that endothelial activation status is determined by a balance between positive and negative regulators: in activated (angiogenic) endothelium, positive regulators predominate, whereas endothelial quiescence is achieved and maintained by the dominance of negative regulators (Figure).^{7 15} Used initially in the context of tumor progression to describe the passage from the prevascular to the vascular phase (see above), the notion of the angiogenic switch is also applicable to developmental and physiological as well as pathological angiogenesis. Although this still remains to be definitively demonstrated in vivo, the current working hypothesis is that the switch involves either the induction of a positive regulator and/or the loss of a negative regulator.

View larger version (0K): [in this window] [in a new window]   Figure 1. Positive and negative regulators of angiogenesis. A large number of factors, listed here alphabetically, have been shown to regulate angiogenesis in the experimental setting. However, with the exception of VEGF/VPF, for almost all of these factors definitive studies are still required to demonstrate their role in the endogenous regulation of angiogenesis. It is nonetheless generally assumed that the "switch" to the angiogenic state may involve either the loss of a negative regulator, the induction of a positive regulator, or both, although definitive proof for this notion is also still awaited. (Reproduced from Reference 55 with copyright permission from Edward Arnold Publishers, London, UK.)

The multiple cell functions that occur during angiogenesis can be classified into a phase of activation (which encompasses initiation and progression) and a phase of resolution (which encompasses termination and vessel maturation). While a great deal is known about the factors that induce the activation phase, very little is known about the factors involved in the phase of resolution, in which it is assumed that the dominant activity of negative regulators is called into play. Furthermore, it is at present unclear as to whether the resolution phase is an active phase or whether it is the consequence of exhaustion of the positive regulators that predominated during the phase of activation. If the latter hypothesis is correct, it would be necessary to assume that endothelial cells have the inherent capacity to synthesize their own basement membrane and organize into capillary-like tubes and that this is mediated in part by the autocrine activity of endogenous regulators. With respect to activated endothelium, an important distinction must be made between physiological and pathological settings: although many of the same positive and negative regulators are operative in both, endothelial cell proliferation in the former is tightly controlled, whereas in the latter, uncontrolled angiogenesis implies continuous dominance of positive regulators, which results in unchecked endothelial cell growth.

Among the factors that affect endothelial cell activation status, either positively or negatively, are cytokines. On the basis of the observation that a given tissue can profoundly influence the way in which its cellular components respond to a given cytokine, it has been suggested that cytokines should be seen as "specialized symbols in a language of intercellular communication, whose meaning is controlled by context."¹⁶ Context is determined by (at least) three parameters: (1) the presence and concentration of other cytokines in the pericellular environment of the responding cell; (2) interactions between cells, cytokines, and the extracellular matrix; and (3) the geometric configuration of the cells (and thus their cytoskeleton). With respect to angiogenesis, the notions of both the angiogenic switch as well as context are proving to be central to our understanding of the molecular mechanisms that govern this process.

The cytokines that have been the most extensively studied in the context of angiogenesis are VEGF, aFGF, and bFGF. The finding that in vitro VEGF and FGF positively regulate many endothelial cell functions, including proliferation, migration, extracellular proteolytic activity, and tube formation, has led to the notion that these factors are direct-acting positive regulators.^{17 18 19 20} However, although a role for VEGF in developmental and tumor angiogenesis has been clearly established, much controversy still exists as to whether the FGFs are relevant to the endogenous control of neovascularization in vivo. Furthermore, although a large number of factors have been demonstrated to be active in the experimental setting, it does not necessarily follow that they are relevant to the endogenous regulation of new blood vessel formation in the intact organism. In the case of molecules that are active during the phase of activation, only one, namely VEGF, meets most of the criteria required for the definition of a vasculogenic or angiogenic factor.^{20 21}

The ultimate target for both positive and negative regulators is the endothelial cell. This has led to the notion that angiogenesis regulators may act either directly on endothelial cells or indirectly by inducing the production of direct-acting regulators by inflammatory and other nonendothelial cells. Thus, in contrast to VEGF and FGF (which are direct endothelial cell mitogens), TGF-ß and TNF- inhibit endothelial cell growth in vitro and have therefore been considered as direct-acting negative regulators. However, both TGF-ß and TNF- are angiogenic in vivo, and it has been demonstrated that these cytokines induce angiogenesis indirectly by stimulating the production of direct-acting positive regulators from stromal and chemoattracted inflammatory cells. In this context, then, TGF-ß and TNF- are considered to be indirect positive regulators.^{20 22 23} In view of its capacity to directly inhibit endothelial cell proliferation and migration, reduce extracellular proteolysis, and promote matrix deposition in vitro, TGF-ß has also been proposed to be a potential mediator of the phase of resolution.²³ In vitro, TGF-ß also promotes the organization of single endothelial cells embedded in three-dimensional collagen gels into tubelike structures,^{24 25} a phenomenon that is likely to be representative of the phase of resolution.

Other cytokines that have been reported to regulate angiogenesis in vivo include HGF, EGF/TGF-, PDGF-BB, interleukins (IL-1, IL-6, and IL-12), interferons, GM-CSF, PlGF, proliferin, and proliferin-related protein. Chemokines that regulate angiogenesis in vivo have to date only been identified in the -C-X-C- family and include IL-8, platelet factor IV, and gro-ß. Angiogenesis can also be regulated by a variety of noncytokine or nonchemokine factors, including enzymes (angiogenin and PD-ECGF/TP), inhibitors of matrix-degrading proteolytic enzymes (TIMPs) and of PAs (PAIs), extracellular matrix components/coagulation factors or fragments thereof (thrombospondin, angiostatin, hyaluronan, and its oligosaccharides), soluble cytokine receptors, prostaglandins, adipocyte lipids, and copper ions (Figure).^{2 20 22 26}

Cell–Extracellular Matrix Interactions: Integrins and Proteases
Alterations in at least four endothelial cell functions occur during angiogenesis: (1) an increase in proliferation, which provides new cells for the growing and elongating vessel, with a subsequent return to the quiescent state once the new vessel is formed; (2) an initial increase and subsequent decrease in locomotion (migration), which allows the cells to translocate toward the angiogenic stimulus and to stop once they reach their destination; (3) endothelial cell-to-cell interactions²⁷ ; and (4) interactions with the extracellular matrix.²⁸ The extracellular matrix is an intricate and complex network of proteinaceous fibers and other macromolecules that profoundly influences cellular function and tissue architecture. Among the molecules that are relevant to cell–extracellular matrix interactions are integral membrane proteins, including integrins, which provide a link between the extracellular matrix and the cytoskeleton, and extracellular proteases and their inhibitors, which mediate focal degradation of the extracellular matrix during cellular invasion.

Integrins are heterodimeric cell-surface receptors composed of two noncovalently associated transmembrane glycoproteins ( and ß) that connect adhesive proteins in the extracellular matrix to the cytoskeleton. Integrins not only mediate attachment of cells to their substratum but are also involved in intracellular signal transduction. At present, 15 different and 8 different ß subunits have been identified, which associate to form more than 20 receptors recognizing one or more extracellular ligands.²⁹ Endothelial cells express a number of different integrins, one of which, namely _vß₃, has been shown to be particularly important during angiogenesis. _vß₃ is a receptor for a number of proteins with an exposed Arg-Gly-Asp (RGD) tripeptide moiety, including vitronectin, fibronectin, fibrinogen, laminin, thrombospondin, osteopontin, and von Willebrand factor. In vivo, this receptor is not widely expressed. It appears to be most prominent on cytokine-activated endothelial cells during angiogenesis in a wide variety of settings and is also expressed by smooth muscle cells in postangioplasty restenosis, atherosclerotic plaques, and healing arterial wounds.³⁰ A significant body of experimental evidence has demonstrated that _vß₃ antagonists (antibodies and cyclic RGD peptides) inhibit angiogenesis during development, wound healing, retinal neovascularization and in growing tumors (in which they induce tumor regression).^{31 32 33 34 35 36} A number of angiogenic cytokines, including bFGF, VEGF, and TGF-ß1, have been shown to increase expression of the _v and ß₃ subunits in endothelial cells,^{37 38 39 40} and _vß₃ antagonists have been shown to markedly inhibit angiogenesis induced by bFGF and TNF- in the chicken chorioallantoic membrane and rabbit corneal micropocket assays.^{31 41} The relevance of _vß₃ to angiogenesis and its potential as an important therapeutic target have therefore been clearly established.

Basement membrane degradation, extracellular matrix invasion, and capillary lumen formation are essential components of the angiogenic process, all of which are dependent on a cohort of proteases and protease inhibitors produced by endothelial and nonendothelial cells. Extracellular proteolysis has also been implicated in the regulation of cytokine activity, and one of the consequences of matrix degradation is the generation of a variety of degradation products, many of which themselves have biological activity.^{28 42 43} Although a number of enzymatic systems have been implicated in extracellular proteolytic events, many of the relevant enzymes belong to one of two families: the serine proteases, in particular the PA/plasmin system, and the MMPs.^{44 45} Although many mechanisms, including transcriptional and translational controls and secretion and activation of proenzymes, are involved in the regulation of extracellular proteolysis, one mechanism that appears to be particularly relevant to cell migration and morphogenesis is spatial localization. Spatial localization, which appears to have evolved to concentrate proteolysis near the cell surface as well as to restrict its activity to the immediate pericellular environment, can be achieved by two mechanisms: first, by binding to cell-surface receptors and matrix-binding sites, and second, by the coproduction of protease inhibitors.²⁸ By preserving matrix integrity and thereby ensuring normal tissue architecture, protease inhibitors play an important permissive role during angiogenesis. These findings have led to the notion that a precise protease-antiprotease equilibrium allows for localized pericellular matrix degradation during cell migration, while at the same time protecting the extracellular matrix against inappropriate destruction.⁴² However, since the net balance of proteolysis required for invasion is always likely to be positive, it has been suggested that antiproteolysis could be effective in inhibiting angiogenesis. In this respect, the requirement for MMP^{46 47 48 49} and PA^{50 51} activity during experimentally induced angiogenesis in vivo has been clearly demonstrated. Taken together, these studies provide evidence for a causal role for the MMP and PA/plasmin systems during angiogenesis in vivo (without, however, identifying the target cell) and point to the potential therapeutic benefits that can be gained from inhibiting angiogenesis by interfering with extracellular proteolytic cascades.

Accumulation of extravascular fibrin is one of the hallmarks of angiogenesis.¹⁸ Fibrin accumulates in the extracellular milieu after injury and also as a consequence of vascular hyperpermeability, which is seen during inflammation and in tumors. In contrast to matrices composed of collagen or other macromolecules, fibrin constitutes a provisional matrix that is progressively removed and replaced by other matrix components, including collagen. Although the mechanisms by which fibrin induces mature matrix formation are poorly understood, fibrin itself is chemotactic for inflammatory cells and has been shown to regulate endothelial cell and fibroblast migration.⁵² Recent observations in mice, whose genes for various components of the PA/plasmin system have been inactivated, have raised the possibility that this system has evolved primarily because of its requirement for fibrinolysis. In essence, mice deficient in plasminogen or doubly deficient in uPA and tPA have no obvious phenotypic abnormalities at birth. However, progressive multiple-organ failure (including reduced fertility) occurs in young adult mice and appears to be the result of generalized microvascular thrombosis as well as widespread extravascular fibrin accumulation.^{53 54 55} In addition, a recent study on wound healing in plasminogen-deficient mice has clearly defined a role for the PA/plasmin system during keratinocyte migration in vivo.⁵⁶ These observations raise the possibility that the PA/plasmin system is important in settings in which fibrin is a significant component of the extracellular matrix. Definitive proof for this hypothesis has come from observations on the phenotype of mice deficient in both plasminogen and fibrinogen: removal of fibrinogen was found to alleviate the diverse spontaneous pathologies associated with plasminogen deficiency and restore wound healing time to normal.⁵⁷ These observations strongly suggest that the fundamental and possibly only physiological role of the PA/plasmin system is to mediate fibrinolysis. With respect to angiogenesis, it had been abundantly demonstrated that uPA, uPA receptor, and PAI-1 are expressed by endothelial cells during angiogenesis in vivo and that in vitro, and all of these components are induced by VEGF and bFGF, two well characterized angiogenic factors.²⁸ (It should be noted that from a quantitative point of view, alterations in expression of the PA/plasmin system are usually far more dramatic than those seen with the MMP system.) However, developmental and physiological angiogenesis appears to occur normally in PA-, uPA receptor–, PAI-1–, and plasminogen-deficient mice.²⁸ Possible explanations for these apparent discrepancies include redundancy (in which different proteins share the same function) and compensation (in which removal of one protein results in upregulation of another protein with a related function). However, it has also been suggested that the lack of phenotype in knockout mice may be an indication that in certain settings, uPA, tPA, PAI-1, and plasmin(ogen) may have no functional role. This may be true, for example, during development, in which fibrin is not a major component of the extracellular matrix. Why, then, are all of these components expressed during cellular invasion and tissue remodeling in a precisely controlled temporospatial manner? One explanation might be that the mechanisms that regulate angiogenesis are unable to distinguish situations that require fibrinolysis from those that do not. Thus, irrespective of the context, a consistent pattern of changes in expression of a cohort of genes will occur in endothelial cells in response to VEGF and bFGF. This might include proteases and protease inhibitors as well as alterations in synthesis of matrix components and integrins. This hypothesis would predict that although uPA and uPA receptor are expressed by endothelial cells during angiogenesis, deletion of these genes would have no consequence in settings in which fibrin is not a major component of the extracellular matrix, and this in turn might account for the lack of a developmental phenotype in the knockout mice. This hypothesis also predicts that it should be possible to inhibit angiogenesis by interfering with the PA/plasmin system in settings such as wound healing, inflammation, and tumor growth, in which fibrin is a major component of the extracellular matrix. Since it is likely that coexpression of proteases and protease inhibitors has evolved in part to prevent excess protease activity, the existence of common regulatory mechanisms that coinduce uPA and PAI-1, irrespective of the context in which angiogenesis is occurring, might mean that like uPA, PAI-1 production in situations not requiring fibrinolysis is also an epiphenomenon. Many questions have been raised by these seminal observations in gene-inactivated mice. For example, is the role of the PA/plasmin system in angiogenesis strictly limited to situations in which migrating endothelial cells encounter and degrade fibrin? If so, is PA expression an epiphenomenon of cell migration (ie, expression where function is not required) in situations in which fibrin is not a major component of the extracellular matrix? The answers to these and other questions should be forthcoming in the near future from further studies in knockout mice.

In summary, angiogenesis is dependent on precisely controlled sequential alterations in a number of endothelial cell functions, including proliferation, migration, and cell–extracellular matrix interactions, all of which are potential targets for antiangiogenic strategies. As will be described below, considerable therapeutic benefit can now be obtained through positive or negative manipulation of the angiogenic process, and this is due in large part to the rapid transfer to the clinical setting of knowledge acquired through a cell biological approach.

	Clinical Applications of Angiogenesis Research

Top Abstract Introduction Mechanisms of Angiogenesis Clinical Applications of... Perspectives for the Future References

 
Based on a large body of experimental evidence in which molecular components of the aforementioned and other processes have been targeted, it has now been clearly demonstrated that considerable therapeutic benefit can be derived from manipulating angiogenesis, either positively or negatively. Table 2 provides a nonexhaustive list of those situations likely to benefit from stimulation of angiogenesis (sometimes referred to as "therapeutic angiogenesis") or inhibition of angiogenesis (antiangiogenesis).

View this table: [in this window] [in a new window]   Table 2. Manipulation of Angiogenesis in the Clinic

Therapeutic Angiogenesis
It goes without saying that stimulation of angiogenesis may be of benefit in wound healing and fracture repair. However, the notion that therapeutic angiogenesis may be beneficial in the treatment of ischemia^{58 59 60 61 62} has recently been substantiated by a large amount of experimental data. Two settings that may benefit from this form of therapy (which is sometimes referred to as "molecular bypass grafting") are coronary artery disease and peripheral arterial occlusive disease, in which the objective is to reduce tissue hypoxia in areas of hypovascularization by local stimulation of angiogenesis. It should be pointed out that at this stage, when referring to therapeutic angiogenesis, the creation of new vascular channels may be the result of either angiogenesis itself or the recruitment (through vasodilatation) of existing vascular channels.

With respect to the heart, current therapy for myocardial ischemia relies on drugs that reduce myocardial oxygen demand, mechanical endovascular revascularization procedures (angioplasty), or bypass surgery.⁶⁰ However, compensatory neovascularization is an important physiological process that occurs in chronic myocardial ischemia.⁶² In this respect, it has recently been demonstrated in experimental models of myocardial ischemia/infarction in the pig and rat that VEGF and VEGF receptors 1 and 2 are increased in chronically ischemic myocardium as well as in regions of ischemia surrounding an area of infarction, the ligand being increased in myocytes and its receptors in endothelial cells.^{63 64 65} Further studies have revealed that hypoxia is a potent inducer of VEGF in cultured cardiac myocytes.^{63 66 67} Finally, bFGF activity has been shown to be increased in myocardium after coronary artery ligation, in parallel with an increase in collateral blood flow,⁶⁸ and elevated levels of bFGF (but not VEGF) have been detected in the pericardial fluid of patients with unstable angina.⁶⁹ These and other observations on the molecular mechanisms of physiological angiogenesis in ischemic myocardium have contributed to the notion that supraphysiological or pharmacological stimulation of angiogenesis may provide a useful alternative/adjunct to more conventional forms of therapy. As will be described below, this notion has recently received considerable experimental support in animal models.

Application of exogenous growth factors in the treatment of myocardial ischemia has thus far been convincingly demonstrated in dogs, pigs, and rabbits. In these models, ischemia is induced in the following ways: application of an ameroid constrictor to a coronary artery (ameroid swells as it slowly absorbs liquid, progressively reducing coronary diameter); injection of microbeads into the coronary microcirculation; and injection of an artificial thrombus into a segment of the coronary artery rendered stenotic by laser ablation. Angiogenic factors that have been assessed in these models include VEGF, aFGF, and bFGF (all with or without heparin). The cytokines were administered in a number of ways: single or multiple intracoronary bolus doses via an indwelling catheter; injection of cytokine-impregnated microcapsules into the coronary arteries, which served to induce ischemia and to provide a slow-release source of the factor; extraluminal (periadventitial) administration via slow-release beads implanted in close proximity to the occlusion; the use of osmotic minipumps placed inside the chest wall, with the microcatheter tip placed in the myocardium just distal to the occlusion; and intrapericardial administration, either transatrially or by means of osmotic minipumps. In all of these studies, a significant beneficial effect was documented, including reduction in infarct size, improved systolic function, increased coronary blood flow, and an increase in vessel density in infarcted (and noninfarcted) regions, as demonstrated by histology.^{70 71 72 73 74 75 76 77 78 79 80 81 82} In one study in which 4 hours of balloon catheter–induced ischemia was followed by reperfusion, although a reduction in infarct size was observed after intracoronary administration of bFGF, this occurred without an increase in vascular density or a significant vasodilator effect.⁸³ These authors have suggested that bFGF may have an early cytoprotective effect on ischemic myocytes and that in the setting of ischemia followed by reperfusion, mechanisms that, in the face of prolonged ischemia would normally activate angiogenesis, become attenuated. It has been demonstrated that the vasomotor responses of physiologically induced collateral vessels, as well as arterioles in collateral vessel–perfused myocardium, are altered when compared with the responses of vessels in normally perfused myocardium. However, normal vasomotor reactivity (regulated by ß-adrenergic and endothelium-dependent mechanisms) could be restored if collateral vessels were induced by sustained periadventitial administration of VEGF, aFGF, and bFGF.^{75 76 77} In all of the studies described above, collateral vessels were induced by application of the angiogenic factor itself. However, one recent study has demonstrated the efficacy of a gene therapy approach in a porcine model of ameroid constrictor–induced myocardial ischemia. In this study, intracoronary application of recombinant adenovirus expressing human FGF-5 (which, unlike bFGF, contains a signal peptide and is therefore likely to enter the classic secretory pathway) resulted in marked improvement in myocardial function and blood flow as well as an increase in capillary density.⁸⁴ Finally, the use of angiogenic factors to increase blood flow to ischemic myocardium is a concept that is currently undergoing preliminary testing in several clinical trials.⁶⁰

With respect to the peripheral vasculature, considerable morbidity and significant perioperative mortality are associated with chronic leg ischemia.⁸⁵ Current forms of therapy include exercise (which stimulates collateral vessel formation), surgical revascularization, and endovascular interventional therapy. Less commonly employed forms of therapy include sympathectomy and the use of vasodilator drugs. The beneficial effect of therapeutic angiogenesis in the induction of collateral vessel formation has been clearly demonstrated in animal models of hindlimb ischemia in the rabbit and rat, in which ligation and excision of the femoral artery (together with branches of the common iliac artery) are followed by administration of an angiogenic factor. bFGF, VEGF, and PDGF-BB have been used successfully in these models and have been administered either intramuscularly or intra-arterially proximal to the site of ligation as a single bolus dose or by continuous infusion via an osmotic minipump. Improvement in a variety of parameters has been reported, including clinical/functional parameters (movement and muscle atrophy); time to peak flow after suprasystolic occlusion; transcutaneous PO₂; calf blood pressure ratio (ie, ischemic versus healthy limb by Doppler measurement); ^99mTc radioisotopic perfusion scan; quantitative serial angiography; blood flow as measured with radiolabeled microspheres; and capillary density, capillary-to–muscle fiber ratio, and muscle necrosis (assessed by histochemical techniques).^{86 87 88 89 90 91 92 93 94} A recent study has demonstrated that combined administration of bFGF and VEGF in a rabbit model of hindlimb ischemia induces an increase in collateral vessel formation, as determined by calf blood pressure ratio, luminal diameter of the stem collateral artery, and the number of opacified collateral vessels (serial angiography), as well as capillary density, all of which were superior to the effect seen with either cytokine alone.⁹⁵ These in vivo studies confirm earlier in vitro findings on synergism between bFGF and VEGF⁹⁶ and point to the potentially important therapeutic applications of this observation. Finally, the feasibility of a site-specific gene transfer approach using VEGF has recently been demonstrated in a rabbit ischemic hindlimb model, in which plasmid was applied to the polymer coating of an angioplasty balloon and delivered percutaneously to the iliac artery.⁹⁷ A similar gene transfer approach is currently undergoing preliminary testing in clinical trials,⁸⁵ and the results in a single nondiabetic patient with severe peripheral ischemia have recently been reported.⁹⁸ Four weeks after percutaneous gene transfer, an increase in collateral vessels at the knee, midtibial, and ankle levels was observed, and this improvement persisted at 12 weeks. Increased blood flow was documented by intra-arterial Doppler flow studies. The patient developed three spider angiomas on the ipsilateral foot/ankle; one was excised and revealed proliferative endothelium, while the other two regressed by 8 weeks. Transient peripheral edema was observed after 7 days but resolved with treatment by 4 weeks. However, despite the development of collateral vessels and improvement in blood flow, limb gangrene could not be reversed and below-knee amputation was required 5 months after gene therapy.

Other settings that may benefit from therapeutic angiogenesis include improvement of survival of free or pedicled skin flaps^{99 100 101} and transplantation of islets of Langerhans.¹⁰² It has also been reported that microvessel density is increased in infarcted brain tissues associated with thromboembolic events and that higher vessel counts correlate with longer survival.¹⁰³ These results suggest that, as for other ischemic tissue, the creation of new vascular channels may be therapeutically beneficial in the central nervous system. In accord with this notion, it has been shown that intraventricular or intracortical application of bFGF promotes angiogenesis (increase in vessel density) and prevents neuronal degeneration in models of chronic cerebral ischemia in rats and gerbils.^{104 105 106 107 108} However, whether the ability of bFGF to prevent neuronal degeneration was due to neovascularization or to a direct trophic effect on neuronal or nonneuronal elements was not established.

What of the future? In addition to optimizing treatment schedules (route of administration, dose, and secondary effects), one important issue will be to determine whether it is possible to limit therapeutic angiogenesis to the region of ischemia only and to subsequently remove the angiogenic stimulus once new vessels have been formed. With respect to ischemia in general, it is possible that if neovascularization depends on upregulation of endothelial tyrosine kinase receptors and if such receptors are upregulated directly or indirectly by hypoxia,^{109 110} then therapeutic angiogenesis should be a localized phenomenon (in the area of ischemia) and should also be self-limiting. Understanding the nature of the limiting factors in physiologically induced collateral vessel formation will almost certainly allow for further improvements in the pharmacological induction of neovascularization (including recruitment by vasodilatation). Finally, it will be important to determine whether therapeutic angiogenesis would best be achieved by administering angiogenic polypeptides or by a gene transfer approach (see below).

Inhibition of Angiogenesis
Foremost among the settings in which the notion of antiangiogenesis has been promoted is the inhibition of solid tumor growth and the spread of metastasis. In his classic article on tumor angiogenesis in 1974, Folkman⁶ highlighted a number of points in the angiogenic process that might be targets for antiangiogenesis therapy. These included "(1) interruption of tumor-angiogenesis factor (TAF) synthesis; (2) blockade of transmission of TAF through tissues; (3) prevention of endothelial cell mitosis; (4) prevention of vessels from penetrating tumor." These and other processes have indeed proven to be appropriate targets for antiangiogenic agents in the context of both tumor growth as well as other settings in which angiogenesis is prominent.¹¹¹

A large amount of descriptive data has clearly pointed to the importance of VEGF and its cognate receptors in tumor angiogenesis.¹⁸ These observations have led to the development of strategies aimed at inhibiting tumor growth by interfering with cytokine-receptor interactions, including anti-VEGF antibodies, soluble VEGF receptors, antisense VEGF, a VEGF-toxin conjugate, and a dominant negative approach using a truncated form of VEGF receptor-2.^{112 113 114 115 116 117 118 119 120 121} Since VEGF is an endothelium-specific mitogen, these findings unambiguously demonstrate the essential requirement for angiogenesis in tumor growth. They also point to the importance of establishing an angiogenic profile in patients with cancer and other chronic angiogenesis-associated diseases, for by knowing which angiogenic cytokine is involved, it will be possible in the near future to specifically tailor antiangiogenic therapy to the individual needs of the patient. Another concept that has recently received convincing experimental support is that of the inverse relationship between tumor dormancy/tumor cell apoptosis and tumor angiogenesis (see above). This follows from Folkman's earlier concept of "`antiangiogenesis' as a means of causing tumors to remain avascular and dormant."⁶ Thus, it has been demonstrated that administration of the potent angiogenesis inhibitor angiostatin, an internal-cleavage product of plasminogen that inhibits the growth of a number of tumors in immunoincompetent mice, achieves its effect by inducing a high rate of tumor cell apoptosis. When angiostatin administration was terminated, angiogenesis proceeded unabated, the rate of tumor cell apoptosis was dramatically decreased (in the face of an unaltered rate of tumor cell proliferation), and tumors entered an exponential phase of growth.^{8 9 122} These findings demonstrate that one of the consequences of antiangiogenesis therapy is the induction and/or maintenance of tumor dormancy; this phenomenon may be explained by either the lack of nutrient supply or the removal of endothelial cell–derived tumor cell survival and growth factors. These observations highlight the importance of reciprocal trophic interactions between tumor cells and endothelial cells. They also serve to illustrate that long-term (possibly even lifelong) antiangiogenesis therapy will be required if one is to achieve a beneficial therapeutic effect.

In addition to VEGF/VEGF receptor antagonists and angiostatin, a large number of polypeptide and nonpolypeptide inhibitors have been successfully used for the inhibition of tumor angiogenesis. Some of these are potential endogenous inhibitors (ie, they may be involved in the physiological regulation of angiogenesis) and include inhibitors of matrix-degrading proteases (TIMPs and PAIs), interferons, thrombospondin, platelet factor 4, a 16-kD fragment of prolactin, C-X-C chemokines, IL-12, and certain steroids and their metabolites. Other inhibitors include AGM-1470/TNP-470 (a fumagillin derivative) and other angiostatic antibiotics, synthetic metalloproteinase inhibitors, angiostatic polysaccharides, suramin analogues, integrin (particularly _vß₃) antibodies and other antagonists, genistein (a tyrosine kinase inhibitor), and thalidomide (Table 3).^{2 111}

View this table: [in this window] [in a new window]   Table 3. Inhibitors of Angiogenesis

Other situations in which antiangiogenesis therapy has been successful include juvenile hemangioma, experimental inflammatory arthritis, and proliferative retinopathy (in which IFN- and thalidomide are currently being tested in clinical trials).² The proven efficacy of IFN-2a in inducing early resolution of juvenile hemangiomas is likely to be related to its capacity to inhibit angiogenesis in vivo,^{123 124} which in turn may be the result of either IFN-2a's direct inhibitory effect on endothelial cells¹²⁵ or an indirect consequence of the reduction of angiogenic factor production by stromal cells.¹²⁶ Although inhibitors of angiogenesis have not been studied in clinical trials in patients with rheumatoid arthritis, it is likely that TIMPs (possibly via a gene therapy approach) as well as other MMP inhibitors will prove to be efficacious due to their combined antiangiogenic and chondroprotective effects.

Finally, since angiogenesis is prominent in the ovary and uterus during specific phases of the female hormonal cycle and is critical for early postimplantation development and formation of the placenta, the concept of antiangiogenesis in the context of birth control certainly merits investigation. Furthermore, it is possible that certain forms of infertility may be associated with aberrant angiogenesis.

Therapeutic Modulation of Angiogenesis: Direct Administration of Polypeptides or Gene Transfer?
Essentially, the therapeutic modulation of angiogenesis can be achieved by two means: either through the use of single or multiple doses of a recombinant protein (or a nonpolypeptide angiogenesis regulatory factor) or by a gene transfer approach. A central issue in future clinically applied angiogenesis research will be to determine which of these two treatment modalities will be the most effective, which will have the least number of side effects, and which will be the least costly.

Factors that favor the use of polypeptide and nonpolypeptide regulators include the ability to regulate their dose and thus be able to define a therapeutic window between efficacy and toxicity, and the ability to clearly define a toxicity profile that will allow withdrawal of treatment if and when necessary. Factors against the use of this approach are (1) the considerable cost involved in producing sufficient quantities of pyrogen-free material; (2) the requirement for repeated or prolonged administration of protein (eg, via an indwelling catheter); and (3) the appearance of secondary effects similar to those seen during prolonged administration of bFGF, which is associated with a decrease in arterial pressure, moderate thrombocytopenia, and moderate reversible anemia.⁸¹

Gene therapy involves the introduction of genetic material into somatic cells of an organism with the aim of achieving high levels of sustained gene expression without provoking adverse host reactions. With respect to angiogenesis, gene therapy can be considered from two perspectives: stimulation of angiogenesis to overcome local tissue hypoxia and antiangiogenesis for the inhibition of neovascularization in chronic diseases including cancer, inflammatory arthritis, and proliferative retinopathy. At the present time, gene transfer has only been used successfully for therapeutic angiogenesis in animal models of limb⁹⁷ and myocardial⁸⁴ ischemia, and the use of this approach in the treatment of critical limb ischemia in humans has recently entered clinical trials.⁸⁵ Successful use of this approach for the inhibition of angiogenesis has not been reported. Factors that favor this approach would include the ability to induce or inhibit angiogenesis at the site of ischemia, if local integration of the vector could be achieved without promoting extralesional angiogenesis at undesirable sites. Although in one animal study in the heart <2% of delivered virus escaped the coronary circulation,⁸⁴ evidence has recently been provided for "spillover" of a VEGF-containing plasmid into the distal circulation after arterial gene transfer into a patient with severe peripheral ischemia.⁹⁸ Although these lesions were self-limiting, they do represent unwanted angiogenesis and clearly indicate that in the future, patients who receive angiogenic gene therapy will require careful monitoring. Factors against the use of gene therapy include the general problems encountered in current forms of gene transfer (ie, retroviruses require dividing cells, adenoviruses frequently induce nonspecific inflammation, use of liposomes and naked DNA is generally inefficient, etc) as well as limitations in regulating the amount and duration of growth factor administration. It is important to recall that adenovirus-mediated gene transfer does not result in stable incorporation of the transgene into the host genome; transgene expression is therefore transient. This would certainly be desirable for therapeutic angiogenesis, since a relatively brief period of transgene expression is likely to be all that is required to induce collateral vessels that, once formed, are likely to be maintained by local hemodynamic factors. Other limitations include the presence of predictable as well as unpredictable secondary effects, which may be related to either the vector or the gene product that it encodes. In the case of VEGF, which is also a potent vascular permeability–inducing factor, unilateral peripheral edema has recently been observed in a patient with severe peripheral ischemia after arterial gene transfer.⁹⁸ Careful monitoring for the presence of local or distal edema will be important if potentially serious complications are to be avoided in an organ or limb in which the circulation is already compromised.

Finally, many gene therapy approaches to date have concentrated on transferring genes into patients with single-gene inherited disorders, ie, transfer of a functionally normal copy into a cell with a defective or abnormal gene without actually replacing the defective gene itself. These are usually single-gene recessive disorders but may also include dominant disorders in which a wild-type gene will override the abnormal dominant phenotype. Other approaches have been used to stimulate the immune system, treat chronic diseases, and provide molecular markers. Human trials are presently under way for genetic diseases (eg, adenosine deaminase deficiency, cystic fibrosis, and familial hypercholesterolemia), AIDS, cancer, rheumatoid arthritis, and cardiovascular diseases (ischemia and restenosis).^{127 128} Although traditionally favored target tissues include lymphocytes (adenosine deaminase deficiency), airway epithelium (cystic fibrosis), skeletal myocytes, hepatocytes, and in the case of cancer, tumor cells themselves, endothelial cells may provide a useful alternative because of their intimate contact with circulating blood. With this approach, it is possible to deliver recombinant genes directly into the vasculature at specific sites by local application of the virus, surgical insertion of vascular grafts seeded with endothelial cells engineered to express the gene of interest, or induction of newly formed vessels containing engineered endothelial cells.^{129 130} Despite the fact that the duration of gene expression after transfer appears at present to be a limiting factor, as a drug delivery system for the treatment of local or systemic vascular and nonvascular diseases, gene therapy employing endothelial cells may prove to be an important alternative.

In summary, stimulation or inhibition of angiogenesis has provided very encouraging results in animal models of a variety of diseases and appears to be meeting with a certain degree of success in the early phases of clinical trials in humans as an adjunct to established/conventional forms of therapy. However, the short-term prospects for therapeutic angiogenesis look slightly better than those for antiangiogenesis. Thus, as opposed to antiangiogenesis therapy, which is likely to require long-term (possibly lifelong) administration of inhibitory agents, therapeutic angiogenesis is likely to necessitate only brief periods of treatment. Both nonspecific and specific side effects of antiangiogenesis therapy, including those related to inhibition of physiological angiogenesis, are therefore likely to pose more of a problem than side effects seen with stimulation of angiogenesis. It is interesting to note that high circulating levels of angiogenic factors such as VEGF and bFGF as seen in patients with a variety of cancers (see below) do not appear to promote a generalized or systemic angiogenic state. Similarly, overt side effects associated with uncontrolled nonspecific angiogenesis have not been reported in animal models of ischemia requiring long-term administration of angiogenic factors (for example, see Reference 131¹³¹ ) or in transgenic mice, which systemically overexpress human bFGF in a constitutive manner.¹³² On the other hand, long-term use of IFN-2a in children with complicated hemangiomas may result in the development of spastic diplegia² and other unpredictable neurological complications, and it is likely that many other antiangiogenic agents will have both predictable as well as unpredictable toxicity profiles. There is no doubt, however, that in the long term when used in conjunction with other treatment modalities, antiangiogenesis is likely to have as much of an impact on patients with chronic diseases including cancer as therapeutic angiogenesis will have on patients with ischemia.

Quantitative Parameters for Diagnosis, Prognosis, and Monitoring of Therapy
Major advances have recently been made in the quantitative assessment of measurable parameters directly associated with angiogenesis. These include histological quantitation of intratumoral vessel density^{133 134} and the measurement of angiogenic factors in body fluids and tissue extracts.¹³⁵ Assessment of these parameters has potentially important clinical applications, including use as diagnostic markers, as indicators of prognosis and disease recurrence, and for monitoring the response to therapy.

The first demonstration that quantitation of microvessel density might be a reliable predictor of metastasis came from studies on tumor biopsy specimens of patients with lymph node–negative breast cancer.¹³⁶ Subsequently, a large number of studies in breast carcinoma and a variety of other tumor types, including lung, prostate, head and neck, rectal, testicular and bladder carcinoma, malignant melanoma, soft tissue tumors, and multiple myeloma, has provided a positive association between tumor angiogenesis and the risk of metastasis, tumor recurrence, or death.^{133 134} However, the inability to demonstrate this association in a number of other studies has led some authors to question the usefulness of a static measure of microvessel density as a prognostic tool in cancer.¹³⁷ One of the requirements for this type of assessment is that immunostained sections must first be scanned at low magnification to select those areas with the greatest numbers of distinctly highlighted microvessels, so-called neovascular "hot spots," which almost certainly introduces an element of subjectivity into the measurement. This and other issues related to the measurement process (including the appropriate antibody or lectin markers employed for highlighting endothelial cells) as well as the complexity of the nature of tumor cell growth, invasion, and metastasis formation may explain some of the variability between reported studies.

By measuring levels of positive and negative regulators of angiogenesis in body fluids and tissue extracts, it will be possible in the future to establish an angiogenic profile for patients with cancer and other angiogenesis-associated diseases.¹³⁵ To date, virtually all reported studies have been concerned with measuring a single, positive regulator, and most have focused on bFGF, which from a historical point of view antedated VEGF as an important positive regulator of angiogenesis. Thus, although bFGF is detectable in trace amounts in the sera and plasma of normal adults (Reference 138¹³⁸ and M.S. Pepper et al, unpublished observations, 1996), bFGF is increased in maternal and cord (fetal) sera at term¹³⁹ and is further elevated in pregnancies complicated by diabetes mellitus.¹⁴⁰ Elevated levels of bFGF have also been detected in the sera of patients with prostatic¹⁴¹ and ovarian¹⁴² carcinoma. bFGF is present in the urine of normal individuals and is slightly higher in females than males.¹⁴³ Elevated levels of bFGF have been detected in the urine of patients with a wide spectrum of tumors, including urogenital (bladder and kidney), breast, and lung, as well as hematological malignancies including lymphoma.^{143 144 145 146} High levels of bFGF have also been detected in the cerebrospinal fluid of children with brain tumors.¹⁴⁷ With regard to VEGF, which is detectable in the serum^{148 149} but not the plasma (M.S. Pepper et al, unpublished data, 1996) of apparently healthy individuals, elevated levels have been detected in the sera of patients with pulmonary, ovarian, and uterine tumors¹⁴⁸ ; in malignant ascites¹⁵⁰ ; in cyst fluid and tissue extracts of glioblastomas^{149 151} ; in the ocular fluid of patients with diabetic retinopathy and other retinal disorders¹⁵² ; in the sera of patients with inflammatory bowel disease¹⁵³ ; and in the urine of women undergoing gonadotrophin treatment.¹⁵⁴ Other angiogenic cytokines that have been detected at elevated levels in cancer patients include the following: (1) aFGF in the urine of patients with bladder cancer¹⁵⁵ ; (2) TGF-ß in the plasma of patients with prostatic¹⁵⁶ and hepatocellular^{157 158} carcinoma and in the sera of patients with ovarian carcinoma¹⁴² ; (3) HGF in urine and tissue extracts of individuals with bladder carcinoma¹⁵⁹ and in the sera of patients with hepatocellular carcinoma¹⁶⁰ ; (4) EGF/TGF- in the urine of patients with glial tumors¹⁶¹ and metastatic breast cancer¹⁶² ; and (5) angiogenin, TNF-, GM-CSF, and interleukins-2, -6, -7, -8, and -10 in patients with ovarian carcinoma.¹⁴² In one study, plasma TGF-ß1 levels were found to correlate positively with the extent of tumor vascularity but not with tumor size or underlying liver disease in hepatocellular carcinoma.¹⁶³ Although studies published to date have measured only positive regulators of angiogenesis, it will be important in the future to also consider negative regulators, since their loss may be permissive for allowing endothelial cells to enter the activation phase of angiogenesis. Finally, measurement of positive and negative regulators and the establishment of an angiogenic profile may also help to determine their importance in settings such as juvenile hemangioma, rheumatoid arthritis, and possibly infertility.

	Perspectives for the Future

Top Abstract Introduction Mechanisms of Angiogenesis Clinical Applications of... Perspectives for the Future References

 
As part of a service to medical/surgical disciplines, it is envisaged that it will ultimately be possible to provide an angiogenesis profile to clinicians who treat angiogenesis-related diseases. A routine service of quantitative histological assessment in diseases where angiogenesis (or, more precisely, blood vessel density) is known to be a significant prognostic factor (eg, in breast cancer^{133 134} ) could also be provided. The objectives of establishing these parameters can be summarized as follows: (1) to assist in our understanding of the biology of an angiogenesis-associated disease. For example, studies on the angiogenic profile may help us to identify which angiogenic cytokines are relevant to the progression of a particular tumor. (2) to provide diagnostic and prognostic parameters at initial presentation, to assist in judging the response to therapy, and as indicators of recurrence in asymptomatic patients after therapy. For example, urinary bFGF levels may serve as a prognostic indicator for certain solid tumors, and bFGF levels in the urine of cancer patients are reduced following surgical removal of the tumor.¹⁴³ Similarly, serum HGF levels have been directly correlated with patient mortality in hepatocellular carcinoma.¹⁶⁰ (3) to help us design more specific antiangiogenic therapies, ie, to tailor the therapy to the needs of the individual on the basis of the profile of positive and negative regulators. Although an enormous amount of progress has been made in identifying positive and negative regulators of angiogenesis and the mechanisms by which they regulate this process, many important fundamental questions remain. A number of issues that merit further investigation are discussed below.

First, although it is currently assumed that vasculogenesis is limited to early development, it will be important in the future to determine whether endothelial cell precursors (in the form of angioblasts or hemangioblasts) persist into adult life. For example, it might be possible by applying techniques currently used in hematopoiesis research to determine whether circulating precursors contribute to neovascularization in the adult. From a therapeutic point of view, this may have important implications both for stimulation as well as inhibition of angiogenesis.

Second, we are clearly entering an era in which a genetic approach to understanding the pathogenesis of vascular disorders,¹⁶⁴ with particular emphasis on angiogenesis, will require identification of mutations in endothelial cell receptor tyrosine kinases (VEGF receptors-1, -2, and -3; Tie-1; and Tie-2). Thus, receptor tyrosine kinase mutations would be expected to be important in the pathogenesis of developmental vascular malformations and might play a role in the development of hemangiomas as well as chronic vasoproliferative disorders (cancer, arthritis, and retinopathy), which are likely to be multigenic. For example, it would be important to define whether increased susceptibility/predisposition to some of these chronic disorders is linked to a genetically based proangiogenic state that may result from increased activity of positive regulators (angiogenic factors and receptors) or decreased activity of inhibitors. With respect to vascular malformations, candidate loci for an autosomal dominant form of venous malformation¹⁶⁵ and for the Klippel-Trénaunay-Weber syndrome¹⁶⁶ have been identified. One could envisage starting, for example, with polymerase chain reaction–based exon sequencing (or other simpler "screening" techniques) of endothelial receptor tyrosine kinases on DNA from individuals with vascular malformations. In addition to peripheral blood leukocytes, the source of DNA could be biopsy specimens of the lesions themselves, since the frequent unilateral and localized nature of these lesions raises the possibility that they might be chimeric. It should also be noted that sufficient DNA is present in a single archival pathological slide to undertake polymerase chain reaction–based screening of paraffin-embedded material. Knowing this may facilitate large-scale retrospective studies on material that is otherwise difficult to obtain.

Third, novel pharmacological and gene therapy approaches need to be developed for the stimulation and inhibition of angiogenesis. In addition, extensive clinical evaluation of current therapeutic strategies will almost certainly require testing in multicenter trials. It will also be important to develop animal models that are relevant to angiogenesis-associated diseases and that could be exploited in the search for novel therapeutic strategies. For example, establishment of transgenic mouse strains by site-directed overexpression of positive regulators (eg, VEGF in the skin as a model for hemangiomas) could be combined with assessment of the therapeutic potential of angiogenesis inhibitors using a gene therapy approach.

Finally, although the future of applied research in angiogenesis is clearly very bright, it will be important to maintain and even accelerate the current rate at which knowledge obtained through a molecular/cell biological approach is transferred to the clinical setting. One way of achieving this objective would be through creation of an "Angiogenesis Unit." Central to the concept of a "Unit" is the idea that it should continue to serve endeavors in basic science research while providing a service to virtually all existing classic medical/surgical disciplines. In this way, data obtained from clinical studies could be fed back into basic research and vice versa; a constant interchange between the two areas would mean that they inevitably force each other to focus on relevant issues and that the boundaries between basic and applied research would gradually become less distinct. To function with maximal efficiency, a "Unit" should have a certain degree of autonomy to transcend the boundaries imposed by classic departmental structure. One solution might be that an "Angiogenesis Unit" form part of a "Vascular Biology Unit," in which problems relevant to vascular biology as a whole could be addressed. This would have the added advantage of providing a platform for vascular biology, which like angiogenesis, transcends the boundaries of classic medical/surgical disciplines. An alternative solution would be that the "Unit" begin in a classic large basic science or clinical department but that the spirit of integration be maintained to avoid the limitations (intellectual and otherwise) that would inevitably be imposed through association with a single department or discipline. But why go so far as creating an "Angiogenesis Unit," which is not entirely without risk, if all of the facets of angiogenesis research, both basic and applied, could be effectively carried out on an independent basis in the various disciplines concerned? Possibly the single most important reason would be to attain a critical mass of highly competent individuals, both clinicians and scientists, who through their physical proximity and competitive nature would feed and motivate each other and who, under the right leadership, would ensure that a consistently high degree of excellence, creativity, and rigorous quality control be maintained. Since angiogenesis knows no boundaries in modern medicine and the boundaries between basic and clinical science are rapidly merging, a highly focused and integrated approach must surely be the key to success in the future.

	Selected Abbreviations and Acronyms

 

a = acidic b = basic G-CSF = granulocyte colony stimulating factor FGF = fibroblast growth factor HGF = hepatocyte growth factor IFN = interferon IL = interleukin MMP(s) = matrix metalloproteinase(s) PA(I)(s) = plasminogen activator (inhibitor)(s) PD = platelet derived PlGF = placental growth factor t = tissue type TGF = transforming growth factor TIMP(s) = tissue inhibitor(s) of metalloproteinases TNF = tumor necrosis factor TP = thymidine phosphorylase u = urokinase type (V)E(C)GF = (vascular) endothelial (cell) growth factor

	Acknowledgments

 
Work from the author's laboratory was supported by grants from the Swiss National Science Foundation and grants-in-aid from the Juvenile Diabetes Foundation (International) and the Sir Jules Thorn Charitable Overseas Trust. The author would like to thank Drs Roberto Montesano, Lelio Orci, and Jean-Dominique Vassalli for their enthusiastic support; Dr Michael Morris for helpful discussions; Dr Giulio Gabbiani for his interest in our work; and Stefano Mandriota for reading the manuscript.

Received December 16, 1996; accepted January 14, 1997.

	References

Top Abstract Introduction Mechanisms of Angiogenesis Clinical Applications of... Perspectives for the Future References

 

1. Risau W, Flamme I. Vasculogenesis. Annu Rev Cell Dev Biol. 1995;11:73-91.[Medline] [Order article via Infotrieve]
2. Folkman J. Clinical applications of research on angiogenesis. N Engl J Med. 1995;333:1757-1763.[Free Full Text]
3. Wagner RC. Endothelial cell embryology and growth. Adv Microcirc. 1980;9:45-75.
4. Dieterlen-Lièvre F. La genèse du système sanguin chez l'embryon. La Recherche. 1993;254:528-534.
5. Pepper MS. Positive and negative regulation of angiogenesis: from cell biology to the clinic. Vasc Med. 1996;1:259-266.[Medline] [Order article via Infotrieve]
6. Folkman J. Tumor angiogenesis. Adv Cancer Res. 1974;19:331-358.[Medline] [Order article via Infotrieve]
7. Hanahan D, Folkman J. Patterns and emerging mechanisms of the angiogenic switch during tumorigenesis. Cell. 1996;86:353-364.[Medline] [Order article via Infotrieve]
8. Holmgren L, O'Reilly MS, Folkman J. Dormancy of micrometastases: balanced proliferation and apoptosis in the presence of angiogenesis suppression. Nat Med. 1995;1:149-153.[Medline] [Order article via Infotrieve]
9. O'Reilly MS, Holmgren L, Chen C, Folkman J. Angiostatin induces and sustains dormancy of human primary tumors in mice. Nat Med. 1996;2:689-692.[Medline] [Order article via Infotrieve]
10. Clark ER, Clark EL. Microscopic observations on the growth of blood capillaries in the living mammal. Am J Anat. 1939;64:251-301.
11. Cliff WJ. Observations on healing tissue: a combined light and electron microscopic investigation. Philos Trans R Soc Lond B. 1963;246:305-325.
12. Schoefl GI. Studies on inflammation, III: growing capillaries—their structure and permeability. Virchows Arch Pathol Anat. 1963;337:97-141.
13. Ausprunk DH, Folkman J. Migration and proliferation of endothelial cells in preformed and newly formed blood vessels during tumor angiogenesis. Microvasc Res. 1977;14:53-65.[Medline] [Order article via Infotrieve]
14. Paku S, Paweletz N. First steps in tumor-related angiogenesis. Lab Invest. 1991;65:334-346.[Medline] [Order article via Infotrieve]
15. Bouck N. Tumor angiogenesis: the role of oncogenes and tumor suppressor genes. Cancer Cells. 1990;2:179-185.[Medline] [Order article via Infotrieve]
16. Nathan C, Sporn M. Cytokines in context. J Cell Biol. 1991;113:981-986.[Free Full Text]
17. Ferrara N, Houck K, Jakeman L, Leung DW. Molecular and biological properties of the vascular endothelial growth factor family of proteins. Endocr Rev. 1992;13:18-32.[Medline] [Order article via Infotrieve]
18. Dvorak HF, Brown LF, Detmar M, Dvorak AM. Vascular permeability factor/vascular endothelial growth factor, microvascular hyperpermeability, and angiogenesis. Am J Pathol. 1995;146:1029-1039.[Abstract]
19. Christofori G. The role of fibroblast growth factors in tumour progression and angiogenesis. In: Bicknell R, Lewis CE, Ferrara N, eds. Tumour Angiogenesis. Oxford, England: Oxford University Press. In press.
20. Pepper MS, Mandriota S, Vassalli J-D, Orci L, Montesano R. Angiogenesis regulating cytokines: activities and interactions. Curr Top Microbiol Immunol. 1996;213:31-67.