Role of Angiogenesis Inhibitors in Cancer Treatment

July 1, 2001
Lee M. Ellis, MD, FACS

Wenbiao Liu, MD

Niels Reinmuth, MD

Raymond M. Shaheen, MD

Young Do Jung, MD, PhD

Syed Ahmad, MD

Oncology, ONCOLOGY Vol 15 No 7, Volume 15, Issue 7

Angiogenesis is essential for the growth of both primary andmetastatic tumors. This process, more complex than was previously thought,

ABSTRACT: Angiogenesis is essential for the growth of both primary andmetastatic tumors. This process, more complex than was previously thought,requires the coordinated activities of multiple factors and cell types. Fortumors to develop a neovascular blood supply, tumor and host cells must secretepro-angiogenic factors that offset the activities of inhibitory angiogenicfactors. In addition, the newly derived tumor endothelium must respond tosurvive in a relatively caustic microenvironment. Thus, endothelial-cellsurvival factors are essential in the maintenance of this neovasculature.Because redundant factors and pathways regulate angiogenesis, inhibition of anysingle pathway is unlikely to lead to prolonged response in most patients withsolid malignancies. Since anti-angiogenic therapy is unlikely to induce tumorregression, the criteria for efficacy must be evaluated by means other than thestandard criteria used to evaluate cytotoxic chemotherapy regimens.Understanding the basic principles that drive tumor angiogenesis will lead tothe development of therapies that will likely prolong survival without thetoxicity associated with standard chemotherapy. [ONCOLOGY 15(Suppl 8):39-46,2001]


Currently, the field of tumor angiogenesis is undergoing moreexplosive growth than any other field in cancer research. More than 860 paperswere published on various aspects of tumor angiogenesis in 1999, and scientists,clinicians, patients, and the media have closely scrutinized this research.Regrettably, trials of anti-angiogenic agents thus far have produced mostlytoxicity data. Given the complexity of this process, the basic biology ofangiogenesis also must be better understood before more effectiveanti-angiogenic therapy can be developed.

By definition, "angiogenesis" is the establishment ofa neovascular blood supply derived from preexisting blood vessels, whereas"vasculogenesis" is the embryonic establishment of a blood supply frommesodermal precursors such as angioblasts or hemangioblasts. "Tumorangiogenesis" more accurately refers to a combination of angiogenesis andvasculogenesis, in which the main blood supply to a tumor is derived frompreexisting blood vessels, but circulating endothelial cell precursors maycontribute to the growing endothelial cell mass.

Angiogenesis is an essential step in both the growth of primarytumors and metastasis.[1,2] A neovascular blood supply is also essential forincreasing the chance that tumor cells will gain access to the circulation andsubsequently begin the process of growth at a different site. Once a tumorestablishes an invasive phenotype in the organ of metastasis, it must thenestablish its own neovascular blood supply for growth. Numerous investigatorshave established the association between tumor angiogenesis and tumormetastasis.[3]

Balance Between Stimulatory and Inhibitory Angiogenic Factors

Pathologic angiogenesis occurs when the effect of stimulatorymolecules outweighs the effect of inhibitory molecules (Table1). Intensivestudy led to the realization that the angiogenic process involves more thansimple proliferation of endothelial cells; rather, it requires endothelial cellsto divide, invade the basement membrane, migrate, and undergo differentiationand capillary-tube formation (Figure 1). This process is driven by angiogenicmolecules and also by other factors, such as degradative enzymes, which mediatethe above processes. Interestingly, tumor angiogenesis and tumor-cell invasionare very similar processes.

Growth Factors

The best characterized of the stimulatory angiogenic factors isthe vascular endothelial growth factor (VEGF). VEGF has also been associatedwith the aggressive phenotype in numerous solid malignancies.[4-9] VEGF is a 32-to 44-kDa protein secreted by nearly all cells. VEGF is expressed as fourisoforms derived from alternate splicing of the mRNA.[10] The smaller isoforms,VEGF-121 and VEGF-165 (the numbers denote the number of amino acids), aresecreted from cells. The larger isoforms, VEGF-189 and VEGF-205, arecell-associated, and their functions are not well known at this time.

One distinguishing factor of VEGF is its ability to inducevascular permeability. In fact, this factor was originally named the vascularpermeability factor (VPF) and was subsequently found to be homologous to VEGF.[11-13] The extent of vascular permeability induced by VEGF is 50,000 times thatof histamine, the gold standard for induction of permeability. This action byVEGF allows proteins to diffuse into the interstitium and to form the latticenetwork onto which endothelial cells migrate.

Endothelial GrowthFactor-Receptor Family

Receptors for VEGF are expressed almost exclusively onendothelial cells. VEGF receptors have been found on cells of neural origin,Kaposi’s sarcoma cells, hematopoietic precursor cells, and other rare tumorcell types.[14,15] The current nomenclature for the VEGF receptors lists threereceptors: VEGFR-1/Flt-1, VEGFR-2 KDR/Flk-1, and VEGFR-3/Flt-4. Thesetyrosine-kinase receptors require dimerization to induce intracellular signalingupon binding to specific ligands. The receptors for VEGF may mediate distinctfunctions within the endothelial cell; for example, VEGFR-1 may be important inmigration, whereas VEGFR-2 may be important in the induction of permeability andcell proliferation.

Recently, the angiopoietin family of ligands has been found toplay an important role in homeostasis of tumor vasculature. The angiopoietinsare proteins involved in angiogenesis that bind to the endothelial-cell-specifictyrosine kinase receptor Tie-2. Angiopoietin-1 (Ang-1) acts as an agonist and isinvolved in endothelial-cell differentiation and stabilization.[16] In contrast,Ang-2 binds to Tie-2 and blocks the binding of Ang-1 to this receptor.[17,18]This blockade leads to endothelial cell destabilization and vascularregression.[19]

New Theories of Angiogenesis

A recent theory of tumor angiogenesis suggests that this processinvolves the co-option of preexisting blood vessels in addition to vascularregression and subsequent neovascularization.[19] Initially, tumors co-optexisting blood vessels within an organ for their nutrient blood supply. Shortlythereafter, the existing vasculature becomes destabilized, most likely throughthe release of Ang-2 by endothelial cells. This loss of vascular integrity leadsto relative hypoxia within the tumor, which, in turn, leads to upregulation ofVEGF in the tumor cells. These events then lead to a robust angiogenic response.At that stage, the newly developed endothelial cells require stabilization,achieved through release of Ang-1 by endothelial cells and possibly throughcontinued response to VEGF. Thus, the process of angiogenesis depends on thetemporal coordination of factors that regulate pathways for the establishment ofstable conduits that provide a nutrient blood supply to the tumor.

Numerous nonspecific angiogenic molecules and factors affect thegrowth of cell types other than endothelial cells. These include fibroblastgrowth factors (acidic and basic); transforming growth factor-alpha andepidermal growth factor (EGF), both of which bind to the EGF receptor;platelet-derived growth factor (PDGF); platelet-derived endothelial cell growthfactor (PD-ECGF); angiogenin; and the CXC chemokine IL-8, macrophageinflammatory protein (MIP), PF-4, and growth-regulated oncogene (GRO)[20] (Table1).

These factors are known to be angiogenic in in vivo models butare not specific for endothelial cells. However, as noted earlier, angiogenesisis not driven by a single molecule or family of molecules, but rather depends onthe cooperation and integration of various factors that lead to endothelial cellproliferation, migration, invasion, differentiation, and capillary-tubeformation. It has yet to be determined whether inhibiting the activity of asingle angiogenic factor will lead to vascular compromise of significantduration. More likely, the redundancy in the angiogenic process will select forother angiogenic factors when a specific angiogenic factor is targeted.

Upstream Regulation of Angiogenic Factors

In formulating antiangiogenic regimens, it is essential tounderstand the cascade of events that leads to upregulation of angiogenic factorexpression and secretion. Signals that upregulate angiogenic factors includeextracellular signals, intrinsic upregulation of signal transduction activity,and loss of tumor suppressor genes. Examples of these signals are discussedbelow.

External signals that lead to the induction of angiogenic factorexpression include environmental stimuli such as hypoxia or a decrease inpH.[21-23] Hypoxia is the most potent stimulus for inducing angiogenic factors,especially VEGF. Hypoxic induction of VEGF is probably mediated through Srckinase activity, which then leads to downstream induction of signaling cascadesand eventually to an increase in the activity of hypoxia-inducible factor-1(HIF-1) alpha .[24,25] This factor then increases the transcription of the VEGFgene, which in turn leads to the induction of angiogenesis. Other externalfactors that increase the angiogenic response include various cytokines andgrowth factors. Insulin growth factors -I and -II, epidermal growth factor,hepatocyte growth factor, interleukin-1, and platelet-derived growth factor haveall been shown to upregulate VEGF.[26-28] Thus, anti-angiogenic therapy couldinvolve downregulation of upstream targets of the angiogenic factors rather thantargeting the angiogenic factors themselves.[24,29]

Once a growth factor or a cytokine binds to its receptor, acascade of intracellular signaling events is initiated. Two specific signaltransduction pathways are well known to mediate the upregulation of angiogenicfactors: the PI-3 kinase/Akt signal transduction pathway, which eventually leadsto stabilization of HIF-1 alpha,[30,31] and the mitogen-activated protein (MAP)kinase pathway, in which activation of extracellular signal-regulatedkinases-1/2 (Erk-1/2) activates factors that increase transcription of the VEGFgene.[32] Activated Ras and Src also have been demonstrated in in vivo models tobe associated with increased VEGF production and angiogenesis.[33] Again,therapeutic strategies that target the upstream effector molecules inangiogenesis may be a rational means of preventing angiogenesis. Inhibitors ofsignal transduction molecules have been demonstrated to inhibit angiogenesis inin vivo tumor models.[24]

Protein products of tumor suppressor genes such as the vonHippel-Lindau (VHL) or p53 genes also regulate angiogenesis. The wild-type VHLprotein represses transcriptional regulation of the VEGF gene.[34-36] A loss ofheterozygosity with a mutation in the remaining VHL allele leads to loss oftranscriptional control of the VEGF gene and overexpression of VEGF. Mutant p53has also been associated with an increase in angiogenesis.[37] Reinsertion ofthe wild-type p53 gene into cells with mutant p53 can downregulate VEGFexpression and angiogenesis.[29] Thus, the process of angiogenesis is driven byexternal forces (including environmental stimuli), aberrations in internalsignaling, and alterations in tumor suppressor gene function.

Anti-Angiogenic Therapy

Overall Expectations

The knowledge that angiogenesis is essential for tumor growthand formation of metastases has led to one of the largest research efforts everundertaken in an attempt to discover effective anti-angiogenesis compounds.Angiogenesis is not only a pathologic process but also is essential forhomeostasis. Physiologic angiogenesis is important in reproduction, woundhealing, and menses, as well as in coronary artery and peripheral vasculardiseases. Thus, therapeutic efficacy requires a balance in which angiogenesis intumors is limited while the host is protected from toxic effects.

In addition to potential toxicity from anti-angiogenic therapy,therapy duration and criteria for efficacy are other issues to be considered.Because most anti-angiogenic therapy is intended to decrease the development ofnew blood vessels, the traditional end points for treatment success or failuremust be redefined. For example, a desirable response for standard chemotherapyis a 50% decrease in the cross-sectional area of a tumor; in contrast, thedesired end point after anti-angiogenic therapy might be inhibition of furthertumor growth. Thus, criteria for the effectiveness of anti-angiogenic therapy,whether in the clinic or in the laboratory, must be considered from a newperspective relative to those for conventional therapies.

Although some reports exist of tumor regression in experimentalmodels of angiogenesis,[38,39] such findings are rare. The vast majority ofstudies in this field demonstrate that anti-angiogenic therapy leads toinhibition of tumor growth rather than regression of established tumors.[40,41]The ability to interpret experimental studies appropriately is critical toensure that extrapolations to the clinical setting are not fraught withunrealistic expectations. For example, a typical growth curve for asubcutaneously implanted tumor may demonstrate that anti-angiogenic therapysignificantly decreases the growth of a tumor. In this preclinical model, this"positive" result may lead to clinical trials of that same agent. Inthe clinic, however, inhibition of tumor growth can be interpreted as"progressive disease" and the therapy thus considered a failure,particularly if tumor-imaging studies are done at short intervals.Anti-angiogenic therapy may well require a longer period of administration thandoes chemotherapy to obtain a desirable response.

Effective anti-angiogenic therapy will probably need to bedelivered on a chronic basis. Chronic administration will require that the agentbe delivered easily (perhaps by the oral route) and have few cumulativelong-term side effects. As noted in the previous paragraph, the effect ofanti-angiogenic therapy may require longer evaluation intervals. One must alsoconsider that the goal of standard anti-angiogenic therapy is to decrease bloodvessel formation, not cause tumor regression. Therefore, uniform criteria shouldbe developed for determining the effectiveness of anti-angiogenic therapy (eg,time to progression, survival, and quality of life); these criteria willprobably differ from current criteria for tumor response to cytotoxic agents.

Overall Strategies

Despite the simplified view that anti-angiogenic therapy simplyinterferes with the blood supply to a tumor, developmental strategies foranti-angiogenic therapies are quite diverse and distinct. Anti-angiogenicstrategies can be classified into four major categories: (1) Those that decreasethe activity of specific angiogenic factors; (2) those that decrease theactivity of endothelial survival factors; (3) those that increase the activityof naturally occurring anti-angiogenic agents, ie, angiostatin, endostatin,thrombospondin, etc; and (4) those that indirectly downregulate activity ofangiogenic and survival factors. (Matrix metalloproteinase inhibitors are notspecific anti-angiogenic agents and will not be discussed in this article.)

Anti-VEGF Approaches

Because VEGF has been linked to the angiogenesis andaggressiveness of many disease types, this prototype molecule will be used indescribing strategies to decrease the activity of angiogenic factors. Anti-VEGFstrategies include neutralizing antibodies, ribozymes, soluble receptors, orantisense infection. Other anti-VEGF strategies include neutralizing antibodiesto the receptors for VEGF, and the use of tyrosine kinase inhibitors that blockdownstream signaling, even upon binding of the ligand to its receptor.Currently, several of the above-mentioned strategies are being assessed inclinical trials, and all of the strategies have shown promise in the preclinicaltrials.

One of the earliest strategies used to inhibit VEGF activityinvolved the use of neutralizing antibodies to VEGF where the antibody is ahybrid and the variable region recognizes the epitope, whereas the Fc ishumanized and is not recognized as foreign by the host. This region should alsointeract with human Fc receptor-bearing effector cells and/or human complement.This strategy is similar to that used for anti-VEGF receptor antibodies. Theantibodies must be administered by the intravenous route.

The other commonly used strategy for inhibiting VEGF activity isthe use of tyrosine kinase inhibitors.[42] These are small molecules thatprevent kinase activation upon binding of its ligand to its receptor. Althoughthese compounds are claimed to be selective for their specific targets, inreality these tyrosine kinase inhibitors do have some cross-reactivity withother receptors and require a much higher dose to achieve an effect. Thus, forall essential purposes, these tyrosine kinase inhibitors are selective. Theseinhibitors can be administered by the intravenous route, although the newergeneration of tyrosine kinase inhibitors can be given orally. This is of greatadvantage to patients who may need to take anti-angiogenic therapy chronically.

Anti-VEGF and Apoptosis

Studies from our laboratory have examined anti-VEGF receptorantibodies and tyrosine kinase inhibitors in mouse models of colon cancer andliver metastasis.[41,43] Interestingly, the compounds were similarly effectivein decreasing hepatic tumor burden, vessel count, and proliferative index of thetumor cells. An unexpected finding from these studies was the increase in thenumber of tumor cells undergoing apoptosis in the treated metastases. Weinvestigated this phenomenon to determine if endothelial cell apoptosis was thepreemptive cause of tumor cell apoptosis. We established a technique ofdouble-staining to identify endothelial cells and superimpose terminaldeoxynucleotidyl transferase-mediated UTP end labeling (TUNEL)-positive cellsthat would selectively identify those endothelial cells undergoingapoptosis.[41] Results showed an increase in the wave of endothelial cellapoptosis that preceded an increase in the number of tumor cells undergoingapoptosis.[43] These data suggest that endothelial cell apoptosis occurs beforetumor cell apoptosis, again demonstrating the selectivity of VEGF and itsactivity for endothelial cells. This also implicates VEGF as a survival factorfor tumor endothelium.

Because anti-VEGF therapy leads to an increase in tumor andendothelial cell apoptosis, one would surmise that this therapy could lead to adecrease in tumor size. There are reports of studies in subcutaneous xenograftmodels where tyrosine kinase inhibitors to the VEGF receptor and otherangiogenic factor receptors can cause regression of established tumors.[42,44]However, in our model of colon cancer liver metastasis, tumor growth wasinhibited but eventually the continued tumor growth led to the animals’deaths. This probably is because tumors contain redundant mechanisms forangiogenesis, and anti-angiogenic therapy directed at a specific factor may leadto selection of cells whose angiogenesis is driven by a different factor.[2,45]

Endothelial Cell Proliferation

A second anti-angiogenic strategy involves agents that decreasethe activity of endothelial cell survival factors. Typically, angiogenesis issimply thought of as the development of a new vasculature within tumors whereendothelial cells migrate, proliferate, invade the basement membrane, anddifferentiate to form a primitive capillary network. However, the tumormicroenvironment is caustic, with low pH and low oxygen tension. Therefore, tosurvive, these fragile endothelial cells must be exposed to endothelial cellsurvival factors that prevent apoptosis in adverse conditions.

Endothelial cell survival factors include pericytes that maystabilize endothelium and function either by cell-to-cell contact or secretionof endothelial cell survival factors, such as VEGF or Ang-1. In the absence ofpericytes, VEGF and Ang-1 are necessary for endothelial cell survival.[46] Thesefactors can be secreted by endothelial cells, tumor cells, or nonmalignant cellswithin the microenvironment. VEGF inhibits endothelial cell apoptosis byactivation of various survival pathways, including the Akt pathway, IAP, A1, andthe MAP kinase pathway [47]. Angiopoietin-1 binds to the specific endothelialcell receptor, Tie-2, and activates the Akt pathway that mediates survival inmany cell types.[48]

Another very important mechanism for endothelial cell survivalis the binding of integrins on the endothelial cell surface to the extracellularmatrix. At first, integrins were thought to be important only in cell-to-cellcontact and binding to the extracellular matrix, but it is now known thatintegrins may mediate intracellular signaling, either alone or in combinationwith other receptors.[49] The integrins alpha 5, beta 3 and alpha 5, beta 5 havebeen shown to act as survival factors for endothelial cells; disruption of thebinding between integrins and the extracellular matrix may lead to endothelialcell death.[50-52] It is likely that integrins lead to activation of focaladhesion kinase, resulting in downstream signaling, which initiates endothelialcell survival mechanisms.[52]

Several agents are currently being evaluated that inhibitintegrin activation.[50,53] Specifically, small molecules have been developedthat may inhibit integrin activation, and antibodies have been synthesized thatblock its binding to the extracellular matrix. Some of these compounds are inearly clinical trials, and other compounds are being evaluated in preclinicaltrials.

Natural Anti-Angiogenic Agents

Another anti-angiogenic strategy is one that increases theactivity of naturally occurring anti-angiogenic agents, including angiostatin,endostatin, and thrombospondin. A great deal of publicity has surrounded thediscovery of angiostatin and endostatin, as these agents were first discoveredas fragments of larger molecules (angiostatin is a fragment of plasminogen,endostatin is a fragment of collagen XVI-II). [38,54] The exact mechanism bywhich these two compounds lead to a decrease in angiogenesis is not yet clearlyunderstood. In addition, the naturally occurring angiogenic antagonistthrombospondin is being evaluated in preclinical trials.

Another anti-angiogenic compound that has received moreattention for its other activities is the interferon family of proteins.[55-58]One of these cytokines, interferon-alpha, was shown to cause regression oflife-threatening childhood hemangiomas in a study published in the early1990s.[57] Further laboratory studies have demonstrated that interferon-alphaand interferon-beta can downregulate basic fibroblast growth factor (b-FGF)levels, and regulation of basic FGF by the interferons has also beendemonstrated in other tumor systems.[59] More recently, reports havedemonstrated the efficacy of interferon-alpha in regression of other tumors inchildren.[56] The efficacy of interferon therapy may be dependent on chroniclow-dose therapy as opposed to higher-dose therapy, which is often associatedwith intolerable side effects.

Finally, another category of anti-angiogenic therapy indirectlydownregulates the activity of angiogenic or endothelial cell survival factors.VEGF and other angiogenic factors are often upregulated in response to stress,such as hypoxia, low pH, or cytokines. Strategies that down-regulate theupstream signaling pathways to VEGF and other angiogenic factors may indirectlydownregulate VEGF activity and angiogenesis. Studies done in our laboratory aswell as others have demonstrated that several growth factor receptors areinvolved in induction of VEGF upon binding of its ligand to its receptor(epidermal growth factor-receptor [EGF-R], insulin-like growth factor-1-receptor[IGF-R1]).[26,60] Strategies to inhibit the activity of these receptors lead toa decrease in in vivo VEGF production and angiogenesis, which in turn leads todecreased tumor growth.

It is also known that tumor suppressor genes, such as p53 andVHL, repress transcription of VEGF. We have shown that infection of a wild-typep53 gene can lead to downregulation of VEGF in a colon cancer cell line with amutated p53 gene and inhibit angiogenesis in vivo.[29] It is possible thatanti-VEGF therapy may be beneficial in patients with von Hippel-Lindausyndrome, which is almost certainly due to overexpression of VEGF in theformation of multiple vascular tumors.[61]

Metronomic Therapy

Recently, several publications have advocated the use oflow-dose continuous chemotherapy as anti-angiogenic therapy. The premise behindthis strategy is that endothelial cells are susceptible to chemotherapy, much asbone marrow progenitor cells are susceptible. Standard chemotherapy protocolsinclude a rest period between regimens in order that the bone marrow canrecover. However, dividing tumor endothelial cells that may have been injuredduring the chemotherapy are also able to recover and reestablish blood conduitsduring this recovery period. Thus, with continuous, low-dose chemotherapy, thedividing tumor endothelial cells are damaged by the chemotherapy but are notgiven sufficient time to recover and reestablish tumor vessels. The low doseprevents significant toxicity to bone marrow progenitor cells.

In tumor cells resistant to cyclophosphamide in vitro, Browderand colleagues showed that repetitive delivery of low-dose cyclophosphamidedecreased angiogenesis and tumor growth.[62] When this dosing regimen was usedin combination with TNP-470, tumors were eradicated. It was thought that thisstrategy provided sustained apoptosis of endothelial cells.

In a modified approach, Klement and associates delivered dailydoses of vincristine with an antibody toVEG-FR-2 and again found marked tumor growth inhibition.[63] This strategy wasformulated to inhibit the activity of VEGF survival function while damagingtumor endothelium with chemotherapy. While this "metronomic"therapeutic approach seems promising, it must be realized that low-dosecontinuous chemotherapy has been administered for numerous diseases withdisappointing results. In fact, the basis of the development of some of thenewer oral analog of fluorouracil and its derivatives is to deliver chronicchemotherapy by a convenient method (oral) with less toxicity and potentiallybetter outcomes. At present, this dosing regimen appears to be equal, but notsuperior, to conventional chemotherapy delivered intravenously regardingresponse and survival of patients with unresectable disease.


As stated previously, most local tumors can be treated bysurgery or radiation, or both. However, the true challenge in oncology lies intreating metastatic cancers. The host microenvironment plays a major role inmodulating gene expression in tumors growing at different sites, and this holdstrue for angiogenic factor expression as well. In our laboratory, we have foundthat VEGF expression is actually higher in liver metastasis than in primarytumors. It would be naive for oncologists to think that anti-angiogenic activitywould be equally efficacious in different tumors growing at different sites. Inaddition, the endothelium is phenotypically distinct at different sites, andtherefore each tumor may not only express different angiogenic factors, but theendothelium may have different angiogenic factor receptors.[64]

Future Directions

It is foreseeable that in the future we will need to obtainbiopsies of tumors growing at various metastatic sites and analyze expression ofvarious genes within these biopsies (Figure 2). The revolution of microarraytechnology may allow us rapidly to identify angiogenic factors that may bedriving angiogenesis in specific tumors at specific sites. At that point, wethen may be able to direct appropriate anti-angiogenic therapy toward specifictargets. It is also possible that continued antiangiogenic therapy against aspecific target may lead to selection of clones whose angiogenesis is driven bya different tumor. Therefore, it may be important to "restage"patients with repeat biopsy of these tumors to determine adequately theangiogenic profile of tumors within the course of their growth, and ideally,response to anti-angiogenic regimens.


1. Folkman J: What is the evidence that tumors are angiogenesisdependent? J Natl Cancer Inst 82:4-6, 1989.

2. Fidler IJ, Ellis LM: The implications of angiogenesis to thebiology and therapy of cancer metastasis. Cell 79:185-188, 1994.

3. Weidner N: Tumor angiogenesis: Review of current applicationsin tumor prognostication, New Cancer Strategies: Angiogenesis Antagonists,Washington, DC, April 3-4, 1995. Cambridge Healthtech Institute.

4. Takahashi A, Sasaki H, Kim SJ, et al: Markedly increasedamounts of messenger RNAs for vascular endothelial growth factor and placentagrowth factor in renal cell carcinoma associated with angiogenesis. Cancer Res54:4233-4237, 1994.

5. Takahashi Y, Kitadai Y, Bucana CD, et al: Expression ofvascular endothelial growth factor and its receptor, KDR, correlates withvascularity, metastasis, and proliferation of human colon cancer. Cancer Res55:3964-3968, 1995.

6. Takahashi Y, Cleary KR, Mai M, et al: Significance of vesselcount and vascular endothelial growth factor and its receptor (KDR) inintestinal-type gastric cancer. Clin Cancer Res 2:1679-1684, 1996.

7. Brown LF, Berse B, Jackman RW, et al: Expression of vascularpermeability factor (vascular endothelial growth factor) and its receptors inadenocarcinomas of the gastrointestinal tract. Cancer Res 53:4727-4735, 1993.

8. Brown LF, Berse B, Jackman RW, et al: Increased expression ofvascular permeability factor (vascular endothelial growth factor) and itsreceptors in kidney and bladder carcinomas. Am J Pathol 143:1255-1262, 1993.

9. Toi M, Kondo S, Suzuki H, et al: Quantitative analysis ofvascular endothelial growth factor in primary breast cancer. Cancer77:1101-1106, 1996.

10. Tischer E, Mitchell R, Hartman T, et al: The human gene forvascular endothelial growth factor. Multiple protein forms are encoded throughalternate exon splicing. J Biol Chem 266:11947-11954, 1991.

11. Dvorak HF, Dvorak AM, Manseau EJ, et al: Fibrin-gelinvestment associated with line 1 and line 10 solid tumor growth, angiogenesis,and fibroplasia in guinea pigs. Role of cellular immunity, myofibroblasts,microvascular damage, and infarction in line 1 tumor regression. J Natl CancerInst 62:1459-1472, 1979.

12. Senger DR, Galli SJ, Dvorak AM, et al: Tumor cells secrete avascular permeability factor that promotes accumulation of ascites fluid.Science 219:983-985, 1983.

13. Dvorak HF, Nagy JA, Berse B, et al: Vascular permeabilityfactor, fibrin, and the pathogenesis of tumor stroma formation. Ann N Y Acad Sci667:101-111, 1992.

14. Ziegler BL, Valtieri M, Porada GA, et al: KDR receptor: akey marker defining hematopoietic stem cells. Science 285:1553-1558, 1999.

15. Ferrer FA, Miller LJ, Lindquist R, et al: Expression ofvascular endothelial growth factor receptors in human prostate cancer. Urology53:567-572, 1999.

16. Papapetropoulos A, Garcia-Cardena G, Dengler TJ, et al:Direct actions of angiopoietin-1 on human endothelium: Evidence for networkstabilization, cell survival, and interaction with other angiogenic growthfactors. Lab Invest 79:213-223, 1999.

17. Lauren J, Gunji Y, Alitalo K: Is angiopoietin-2 necessaryfor the initiation of tumor angiogenesis? Am J Pathol 153:1333-1339, 1998.

18. Davis S, Yancopoulos GD: The angiopoietins: Yin and Yang inangiogenesis. Curr Top Microbiol Immunol 237:173-185, 1999.

19. Holash J, Maisonpierre PC, Compton D, et al: Vesselcooption, regression, and growth in tumors mediated by angiopoietins and VEGF.Science 284:1994-1998, 1999.

20. Moore BB, Arenberg DA, Addison CL, et al: Tumor angiogenesisis regulated by CXC chemokines. J Lab Clin Med 132:97-103, 1998.

21. Levy AP, Levy NS, Wegner S, et al: Transcriptionalregulation of the rat vascular endothelial growth factor gene by hypoxia. J BiolChem 270:13333-13340, 1995.

22. Shweiki D, Neeman M, Itin A, et al: Induction of vascularendothelial growth factor expression by hypoxia and by glucose deficiency inmulticell spheroids: Implications for tumor angiogenesis. Proc Natl Acad Sci USA92:768-772, 1995.

23. Shweiki D, Itin A, Stoffer D, et al: Vascular endothelialgrowth factor induced by hypoxia may mediate hypoxia-initiated angiogenesis.Nature 359:843-845, 1992.

24. Ellis LM, Staley CA, Liu W, et al: Downregulation ofvascular endothelial growth factor in a human colon carcinoma cell linetransfected with an antisense expression vector specific for c-src. J Biol Chem273:1052-1057, 1998.

25. Mukhopadhyay D, Tsiokas L, Zhou XM, et al: Hypoxic inductionof human vascular endothelial growth factor expression through c-src activation.Nature 375:577-581, 1995.

26. Akagi Y, Liu W, Xie K, et al: Regulation of vascularendothelial growth factor expression in human colon cancer by insulin-likegrowth factor-I. Cancer Res 58:4008-4014, 1998.

27. Akagi Y, Liu W, Xie K, et al: Regulation of vascularendothelial growth factor expression in human colon cancer by interleukin-1. BrJ Cancer 80:1506-1511, 1999.

28. Tsai JC, Goldman CK, Gillespie GY: Vascular endothelialgrowth factor in human glioma cell lines: Induced secretion by EGF, PDGF-BB, andbFGF. J Neurosurg 82:864-873, 1995.

29. Bouvet M, Ellis LM, Nishizaki M, et al:Adenovirally-mediated wild-type p53 gene transfer down-regulates vascularendothelial growth factor expression and inhibits angiogenesis in human coloncancer. Cancer Res 58:2288-2292, 1998.

30. Mazure NM, Chen EY, Laderoute KR, et al: Induction ofvascular endothelial growth factor by hypoxia is modulated by aphosphatidylinositol 3-Kinase/Akt signaling pathway in Ha-ras-transformed cellsthrough a hypoxia inducible factor-1 transcriptional element. Blood90:3322-3331, 1997.

31. Maxwell PH, Dachs GU, Gleadle JM, et al: Hypoxia-induciblefactor-1 modulates gene expression in solid tumors and influences bothangiogenesis and tumor growth. Proc Natl Acad Sci USA 94:8104-8109, 1997.

32. Jung YD, Nakano K, Liu W, et al: Extracellularsignal-regulated kinase activation is required for upregulation of vascularendothelial growth factor by serum starvation in human colon carcinoma cells.Cancer Res 58:4804-4807, 1999.

33. Rak J, Mitsuhashi Y, Bayko L, et al: Mutant ras oncogenesupregulate VEGF/VPF expression: Implications for induction and inhibition oftumor angiogenesis. Cancer Res 55:4575-4580, 1995.

34. Pal S, Claffey KP, Dvorak HF, et al: The von Hippel-Lindaugene product inhibits vascular permeability factor/vascular endothelial growthfactor expression in renal cell carcinoma by blocking protein kinase C pathways.J Biol Chem 272:27509-27512, 1997.

35. Mukhopadhyay D, Knebelmann B, Cohen HT, et al: The vonHippel-Lindau tumor suppressor gene product interacts with Sp1 to repressvascular endothelial growth factor promoter activity. Mol Cell Biol17:5629-5639, 1997.

36. Levy AP, Levy NS, Goldberg MA: Hypoxia-inducible proteinbinding to vascular endothelial growth factor mRNA and its modulation by the vonHippel-Lindau protein. J Biol Chem 271:25492-25497, 1996.

37. Takahashi Y, Bucana CD, Cleary KR, et al: p53, vessel count,and vascular endothelial growth factor expression in human colon cancer. Int JCancer 79:34-38, 1998.

38. O’Reilly MS, Boehm T, Shing Y, et al: Endostatin: anendogenous inhibitor of angiogenesis and tumor growth. Cell 88:277-285, 1997.

39. Lode HN, Moehler T, Xiang R, et al: Synergy between anantiangiogenic integrin alphav antagonist and an antibody-cytokine fusionprotein eradicates spontaneous tumor metastases. Proc Natl Acad Sci USA96:1591-1596, 1999.

40. Warren RS, Yuan H, Matli MR, et al: Regulation by vascularendothelial growth factor of human colon cancer tumorigenesis in a mouse modelof experimental liver metastasis. J Clin Invest 95:1789-1797, 1995.

41. Shaheen RM, Davis DW, Liu W, et al: Antiangiogenic therapytargeting the tyrosine kinase receptor for vascular endothelial growth factorreceptor inhibits the growth of colon cancer liver metastasis and induces tumorand endothelial cell apoptosis. Cancer Res 59:5412-5416, 1999.

42. Cherrington JM, Strawn LM, Shawver LK: New paradigms for thetreatment of cancer: The role of anti-angiogenic agents. in Woude (ed.):Advances in Cancer Research. Vol. 79, pp 1-38. San Diego, California, AcademicPress, 2000.

43. Bruns CJ, Liu W, Shaheen RM, et al: Vascular endothelialgrowth factor is an in vivo survival factor for tumor endothelium in a murinemodel of colorectal liver metastases. Cancer 89:488-499, 2000.

44. Boehm T, Folkman J, Browder T, et al: Antiangiogenic therapyof experimental cancer does not induce acquired drug resistance. Nature390:404-407, 1997.

45. Takahashi Y, Bucana CD, Liu W, et al: Platelet derivedendothelial cell growth factor in human colon cancer angiogenesis: role ofinfiltrating cells. J Natl Cancer Inst 88:1146-1151, 1996.

46. Holash J, Maisonpierre PC, Compton D, et al: Vesselcooption, regression, and growth in tumors mediated by angiopoietins and VEGF.Science 284:1994-1998, 1999.

47. Liu W, Ahmad SA, Reinmuth N, et al: Endothelial cellsurvival and apoptosis in the tumor vasculature. Apoptosis 5: 323-328, 2000.

48. Papapetropoulos A, Fulton D, Mahboubi K, et al:Angiopoietin-1 inhibits endothelial cell apoptosis via the Akt/survivin pathway.J Biol Chem 275:9102-9105, 2000.

49. Eliceiri BP, Cheresh DA: The role of alphav integrins duringangiogenesis: Insights into potential mechanisms of action and clinicaldevelopment. J Clin Invest 103:1227-1230, 1999.

50. Eliceiri BP, Klemke R, Stromblad S, et al: Integrinalphavbeta3 requirement for sustained mitogen-activated protein kinase activityduring angiogenesis. J Cell Biol 140:1255-1263, 1998.

51. Scatena M, Almeida M, Chaisson ML, et al: NF-kappaB mediatesalphavbeta3 integrin-induced endothelial cell survival. J Cell Biol141:1083-1093, 1998.

52. Akiyama SK: Integrins in cell adhesion and signaling. HumCell 9:181-186, 1996.

53. Allman R, Cowburn P, Mason M: In vitro and in vivo effectsof a cyclic peptide with affinity for the alpha(nu)beta3 integrin in humanmelanoma cells. Euro J Cancer 36:410-422, 2000.

54. O’Reilly MS, Holmgren L, Shing Y, et al: Angiostatin: Anovel angiogenesis inhibitor that mediates the suppression of metastases by aLewis lung carcinoma. Cell 79:315-328, 1994.

55. Dong Z, Greene G, Pettaway C, et al: Suppression ofangiogenesis, tumorigenicity, and metastasis by human prostate cancer cellsengineered to produce interferon-beta. Cancer Res 59:872-879, 1999.

56. Kaban LB, Mulliken JB, Ezekowitz RA, et al: Antiangiogenictherapy of a recurrent giant cell tumor of the mandible with interferon alfa-2a.Pediatrics 103:1145-1149, 1999.

57. Ezekowitz RAB, Mulliken JB, Folkman J: Interferon alfa-2atherapy for life-threatening hemangiomas of infancy. N Engl J Med 326:1456-1463,1992.

58. Ellis LM, Fidler IJ: Angiogenesis and metastasis. Euro JCancer 32A:2451-2460, 1996.

59. Singh RK, Gutman M, Bucana CD, et al: Interferons alpha andbeta downregulate the expression of basic fibroblast growth factor in humancarcinoma. Proc Natl Acad Sci 92:4562-4566, 1995.

60. Perrotte P, Matsumoto T, Inoue K, et al: Anti-epidermalgrowth factor receptor antibody C225 inhibits angiogenesis in human transitionalcell carcinoma growing orthotopically in nude mice. Clin Cancer Res 5:257-265,1999.

61. Pal S, Claffey KP, Dvorak HF, et al: The von Hippel-Lindaugene product inhibits vascular permeability factor/vascular endothelial growthfactor expression in renal cell carcinoma by blocking protein kinase C pathways.J Biol Chem 272:27509-27512, 1997.

62. Browder T, Butterfield CE, Kraling BM, et al: Antiangiogenicscheduling of chemotherapy improves efficacy against experimental drug-resistantcancer. Cancer Res 60:1878-1886, 2000.

63. Klement G, Baruchel S, Rak J, et al: Continuous low-dosetherapy with vinblastine and VEGF receptor-2 antibody induces sustained tumorregression without overt toxicity. J Clin Invest 105:R15-R24, 2000.

64. Pasqualini R, Ruoslahti E: Organ targeting in vivo usingphage display peptide libraries. Nature 380:364-366, 1996.

65. Fidler IJ, Kerbel RS, Ellis LM: Biology of Cancer:Angiogenesis, in DeVita VT, Hellman S, Rosenberg SA (eds.): Cancer: Principlesand Practice of Oncology, 6th ed, pp 137-147. Philadelphia, Pennsylvania,Lippincott Williams & Wilkins Publishers, 2001.

66. Jung YD, Ahmad SA, Akagi Y, et al: Role of the tumormicroenvironment in mediating response to anti-angiogenic therapy. CancerMetastasis Rev 19(1-2):147-157, 2000.