• vascular endothelial growth factor;
  • signaling;
  • resistance;
  • cancer


  1. Top of page
  2. Abstract

Angiogenesis is essential for the growth of primary tumors and for their metastasis. This process is induced by factors, such as vascular endothelial growth factors (VEGFs), that bind to transmembrane VEGF receptors (VEGFRs). VEGF-A is the primary factor involved with angiogenesis; it binds to both VEGFR-1 and VEGFR-2. The inhibition of angiogenesis by obstructing VEGF-A signaling has been investigated as a method to treat solid tumors, but the development of resistance to this blockade has complicated treatment. The major mechanisms of this resistance to VEGF-A blockade include signaling by redundant receptors, such as the fibroblast growth factors, angiopoietin-1, ephrins, and other forms of VEGF. Other major mechanisms of resistance are increased metastasis of hypoxia-resistant tumor cells, recruitment of cell types capable of promoting VEGF-independent angiogenesis, and increased circulation of nontumor proangiogenic factors. Additional mechanisms of resistance to VEGF-A blockade include heterogeneity of responsiveness among tumor cells, use of anti-VEGF-A agents at insufficient doses or for insufficient duration, altered sensitivity to anti-VEGF-A agents by mutations in endothelial cells or vascular remodeling, maintenance of vascular sleeves that allow for easy regrowth of tumor vasculature upon discontinuation of therapy, vascular cooption, and intussusceptive angiogenesis. An understanding of these mechanisms may lead to the development of targeted therapies that overcome this resistance. Some of these approaches include the combined inhibition of redundant angiogenic pathways, proper patient selection for various therapies based on gene expression profiles, blockade of cellular migration by inhibition of colony-stimulating factor, or the use of agents to disrupt vascular architecture. Cancer 2012;3455–3467. © 2011 American Cancer Society.


  1. Top of page
  2. Abstract

Vasculogenesis and angiogenesis are the essential processes that form the extensive human vasculature, providing oxygen, essential nutrients, and transport of cells and cellular products.1 Vasculogenesis2 occurs in the embryo when endothelial cells arising from the mesoderm form vascular channels. The vascular architecture of the mature organism then forms by the process defined as angiogenesis through extensive branching, sprouting,3 and intussusceptive microvascular growth,4 followed by association with the vascular smooth muscle.5 Angiogenesis occurs throughout the lifetime of the individual and takes part in tissue repair, hemostasis, the menstrual cycle, implantation of the embryo, and overall growth and maturation.6 The pathogenesis of many diseases occurs by dysfunction in the formation of vessels (eg, pre-eclampsia),7 inappropriate vessel overgrowth (eg, diabetic retinopathy),8 insufficiency or occlusion of vessels (coronary and cerebrovascular disease), and neovascularization (tumorigenesis).9, 10

Angiogenesis is controlled by various growth, inhibitory, and feedback mechanisms that are tightly regulated.11 Stasis in angiogenesis12 can be achieved either by removing or interrupting positive growth factors13 or by administration of negative factors.14 Because the growth and progression of solid tumors depends on angiogenesis, inhibition of this process has been investigated as a therapeutic strategy for the treatment of cancer.15-18 Here, we discuss how the use of inhibitors of angiogenesis have compound effects along with conventional chemotherapy in malignancy and metastasis.15-17 Several antiangiogenic agents have been developed for the treatment of various cancers (Table 1); however, tumor growth still occurs over time, and this problem of resistance is the focus of our review.

Table 1. Antiangiogenic Agents Targeting Vascular Endothelial Growth Factor Signaling Approved in the United States for the Treatment of Cancer
Agent (Supplier)ReferenceAgent TypeTarget(s)Cancer Type(s)Line of Therapy
  • Abbreviations: c-fms, transmembrane glycoprotein receptor tyrosine kinase; c-KIT, stem cell factor receptor; CRC, colorectal cancer; CSF-1R, colony-stimulating factor 1 receptor; FGFR, fibroblast growth factor receptor; FLT-3, fms-like tyrosine kinase 3; GIST, gastrointestinal stromal tumor; HCC, hepato-cellular carcinoma; HER2, human epidermal growth factor receptor 2; Itk, interleukin-2 receptor inducible T-cell kinase; Lck, leukocyte-specific protein tyrosine kinase; NSCLC, nonsmall cell lung cancer; PDGFR, platelet-derived growth factor receptor; Raf, v-raf 1 murine leukemia viral oncogene homolog 1; RCC, renal cell carcinoma; RET, rearranged during transfection; TKI, tyrosine kinase inhibitor; VEGF, vascular endothelial growth factor; VEGFR, vascular endothelial growth factor receptor.

  • a

    Pending results of proceedings of the US Food and Drug Administration.

Bevacizumab (Genentech, South San Francisco, Calif)Genentech 201119Monoclonal antibodyVEGFMetastatic HER2-negative breastaFirst-line therapy in combination with paclitaxel in chemotherapy-naive patients
    Metastatic CRCFirst-line or second-line therapy with 5-fluorouracil–based chemotherapy
    Unresectable, locally advanced, recurrent, or metastatic nonsquamous NSCLCFirst-line therapy with carboplatin and paclitaxel
    Metastatic RCCFirst-line therapy in combination with interferon alpha or after nephrectomy
    GlioblastomaSecond-line therapy
Sunitinib (Pfizer, New London, Conn)Pfizer 201120TKIVEGFR-2, PDGFR-β, RET, c-KIT, CSF-1R, and FLT-3Advanced RCCFirst-line therapy
    GISTSecond-line therapy
    Unresectable or locally advanced pancreatic neuroendocrine tumorsFirst-line therapy
Sorafenib (Bayer, Leverkusen, Germany)Bayer 201021TKIVEGFR-2 and VEGFR-3, PDGFR-β, c-KIT, Raf, and FLT-3Unresectable HCCFirst-line therapy
    Advanced RCCSecond-line therapy
Pazopanib (GlaxoSmithKline, London, United Kingdom)GlaxoSmithKline 201022TKIVEGFR-1, VEGFR-2, and VEGFR-3; PDGFR-α and PDGFR-ß; FGFR-1 and FGFR-3; c-KIT; Itk; Lck; and c-fmsAdvanced RCCFirst-line therapy or after prior cytokine therapy

Factors of Angiogenesis

Angiogenesis occurs under the influence of various biochemical interactions and also possibly by mechanical stimulation like shear stress of exercise.23 Once the need for a new vessel is established, proangiogenic factors result in proliferation and infiltration of endothelial cells through degradation of the extracellular matrix by myeloid24 and various circulating inflammatory cells. Finally, smooth muscle migration from pericytes leads to vascular wall maturation and the formation of a new vessel.3

Endothelial cell proliferation is induced by soluble factors like vascular endothelial growth factors (VEGFs),25, 26 fibroblast growth factors (FGFs),27, 28 angiopoietin-1,29 and insulin-like growth factors (IGFs),30 as well as the transcription factor hypoxia-inducible factor (HIF).31 These factors interact with cognate cell surface receptors and various coreceptors like syndecans32 to promote proliferation. In addition, enzymes like matrix metalloproteinase digest extracellular matrix, and endothelial cells interact with various integrins to cause migration and apoptosis,33 making way for the new vessel. Finally, antiangiogenic factors (eg, endostatin) secreted by tumor cells also affect the development of tumor vasculature and may be therapeutically induced as a potential strategy to inhibit angiogenesis.9, 34 All these processes occur simultaneously with each process playing its essential and crucial part to ensure functional vessel formation and stabilization.

Vascular Endothelial Growth Factor Receptors

The VEGFs are a family of proteins that play a dominant role in angiogenesis. Five of these factors are produced in humans and were named in the order of their discovery: placental growth factor 1 (PlGF-1) and PlGF-2, VEGF-A, VEGF-B, VEGF-C, VEGF-D.35 Orf virus (Poxviridae)-encoded VEGF-E, and, more recently, VEGF-F identified in snake venom36 complete the total of 7 proteins discovered thus far with ability to interact with VEGF receptors (VEGFRs).

VEGF-A is a major factor involved in angiogenesis; after coupling with its appropriate receptor, it causes migration and multiplication of endothelial cells. VEGF-A also increases expression of proangiogenic integrins (αvβ3), increases vessel permeability, causes vasodilation, and is chemotactic to macrophages and granulocytes. In addition, VEGF-A increases expression of enzymes responsible for extracellular matrix degradation, resulting in creation of the blood vessel lumen. The combination of all the above processes leads to infiltration and formation of the vessel into the supporting tissue.37, 38

Conflicting evidence has been reported from preclinical models regarding the role of VEGF-B in angiogenesis and cardiovascular development.37-39 VEGF-C and VEGF-D are mostly involved with lymphangiogenesis,37 whereas PlGF, similar to VEGF-A, plays an important role in normal and tumor-associated angiogenesis.37 Provided there is no interference in the signaling of these proangiogenic factors, all these processes may occur normally throughout life.

These proteins are products of an 8-exon VEGF gene through which alternate splicing can alter expression, biologic activity, and heparin-binding affinity. Such differences in splicing can also make them behave as either proangiogenic or antiangiogenic.40 The presence of these various isoforms also may predict response and toxicity to various antiangiogenic agents in development.41

The VEGFRs,42 namely VEGFR-1 (fms-like tyrosine kinase [FLT]-1), VEGFR-2 (fetal liver kinase [FLK] 1/kinase insert domain receptor [KDR]), and VEGFR-3 (FLT-4), consist of a split intercellular tyrosine kinase domain, a single transmembrane component, and 7 immunoglobulin-like domains that form the extracellular component. Different binding affinities exist among VEGFRs and VEGF family members; specifically, VEGF binds to both VEGFR-1 and VEGFR-2; whereas VEGF-B, PlGF-1, and PlGF-2 primarily activate VEGFR-1. VEGF-C and VEGF-D have highest affinity for VEGFR-3, VEGF-E primarily binds VEGFR-2, and VEGF-F may interact with both VEGFR-1 and VEGFR-2.35, 38 Differences also exist in the level of involvement of each receptor in specific processes. For example, VEGFR-1 and VEGFR-2 are involved primarily in vasculogenesis and angiogenesis, whereas VEGFR-3 mostly takes part in lymphangiogenesis.35 VEGFR-2 mediates most of the known responses of VEGF; however, VEGFR-1 may modulate the function of VEGFR-2. The neuropilin-1 and neuropilin-2 receptors form complexes with VEGFR-1 and VEGFR-2, enhance VEGFR-mediated VEGF signaling, and may transmit VEGF signaling even in the absence of VEGFRs through interactions with other types of tyrosine kinases (Fig. 1).43 Tumor cells are known to over express VEGF and at times may express multiple VEGF ligands simultaneously, whereas endothelial cells represent the most significant source of VEGFR expression. Specifically, VEGFR-2 is globally expressed by endothelial cells, whereas VEGFR-1 and VEGFR-3 expression is restricted to endothelial cells in distinct locations.35

thumbnail image

Figure 1. Vascular endothelial growth factor receptors (VEGFRs) and neuropilins mediate proangiogenic signaling. The receptors VEGFR-1 and VEGFR-2 bind VEGF family ligands and are the key receptors involved in angiogenesis. VEGFR-3 binds VEGF-C and VEGF-D and is involved primarily in lymphangiogenesis. The neuropilin-1 and neuropilin-2 receptors that bind a range of signaling molecules (including variants of VEGF-A and the semaphorins) form complexes with VEGFRs and potentiate VEGF signaling. PlGF indicates placental growth factor; FLT, fms-related tyrosine kinase (VEGF/vascular permeability factor receptor); FLK-1/KDR, fetal liver kinase 1/kinase insert domain receptor; FLT, fms-related tyrosine kinase.

Download figure to PowerPoint

Inhibition of Angiogenesis

Angiogenesis is an essential process for the growth of primary tumors and their metastasis. The critical process of blood vessel recruitment18 may occur even at a premalignant stage to promote proliferation.44 Inhibition of angiogenesis at such an early stage can be valuable to prevent the initial growth of the primary tumor and the occurrence and growth of metastases, potentially making conventional chemotherapy agents more effective.45-47

Stasis in angiogenesis can be achieved either by blocking or inhibiting proangiogenic factors or by administering compounds that, by themselves, have endogenous antiangiogenic properties. Endogenous compounds with antiangiogenic properties include vasostatin, angiostatin, endostatin, and platelet factor-4, among others. Many drugs, such as anti-VEGF antibodies (eg, bevacizumab [Genentech, South San Francisco, Calif]) and tyrosine kinase inhibitors (TKIs; eg, sunitinib [Pfizer; New London, Conn], sorafenib [Bayer; Leverkusen, Germany], pazopanib [GlaxoSmithKline; London, United Kingdom]), have been developed to date for the treatment of various cancers (Table 1). Some of these agents have been incorporated into various existing chemotherapy regimens, whereas others are used as single agents.

Mechanisms of Resistance to Vascular Endothelial Growth Factor Receptor Blockade

VEGF-targeted monoclonal antibodies (eg, bevacizumab) and receptor-targeted monoclonal antibodies inhibit receptor-ligand interaction and the subsequent activation of receptors through binding ligand or binding the extracellular domain of the receptor, respectively; whereas TKIs bind the intracellular tyrosine kinase domain of the receptor to block activation of downstream signaling.48 In addition to currently approved indications, sorafenib, sunitinib, and pazopanib (all of which are multitargeted antiangiogenic TKIs) are in later stages of clinical development for the treatment of other cancer types. Other investigational TKIs (cediranib [AstraZeneca; Wilmington, Del], BIBF 1120 [Boehringer Ingelheim; Ingelheim, Germany], brivanib [Bristol-Myers Squibb; New York, NY], motesanib [Amgen; Thousand Oaks, Calif], ABT-869 [Abbott; Abbott Park, Il], and axitinib [Pfizer]) also are being evaluated in phase 3 clinical trials for the treatment of a number of malignancies (Table 2).

Table 2. Multitargeted Antiangiogenic Agents Targeting Vascular Endothelial Growth Factor Signaling in Phase 3 of Clinical Development for the Treatment of Cancer
Agent (Supplier)ReferenceAgent TypeTargetsCancer Type(s)aLine of Therapy
  • Abbreviations: c-fms, transmembrane glycoprotein receptor tyrosine kinase; c-KIT, stem cell factor receptor; CRC, colorectal cancer; CSF-1R, colony-stimulating factor 1 receptor; FGFR, fibroblast growth factor receptor; FLT-3, fms-like tyrosine kinase 3; HCC, hepatocellular carcinoma; Itk, interleukin-2 receptor inducible T-cell kinase; Lck, leukocyte-specific protein tyrosine kinase; NSCLC, nonsmall cell lung cancer; PDGFR, platelet-derived growth factor receptor; Raf, v-raf 1 murine leukemia viral oncogene homolog 1; RCC, renal cell carcinoma; RET, rearranged during transfection; src, v-src sarcoma viral oncogene homolog; TKI, tyrosine kinase inhibitor; VEGFR, vascular endothelial growth factor receptor.

  • a

    Based on phase 3 trials listed in (accessed on March 10, 2011).

Sunitinib (Pfizer, New London, Conn)Pfizer 201120TKIVEGFR-2, PDGFR-β, RET, c-KIT, CSF-1R, and FLT-3Advanced/metastatic NSCLCSecond-line or maintenance therapy
    Advanced/metastatic RCCAdjuvant therapy after surgery
    Advanced breastFirst-line or second-line in combination with chemotherapy
Sorafenib (Bayer, Leverkusen, Germany)Bayer 201021TKIVEGFR-2 and VEGFR-3; PDGFR-β; c-KIT; Raf; and FLT-3Advanced NSCLCFirst-line in combination with chemotherapy and third-line or fourth-line therapy
    Advanced/metastatic HCCFirst-line therapy alone or in combination with erlotinib or in combination with chemotherapy
    Advanced/metastatic HER2-negative breastSecond-line in combination with chemotherapy
    Advanced RCCSecond-line therapy; adjuvant therapy after nephrectomy
    Advanced/metastatic differentiated thyroidSecond-line therapy
    Unresectable stage III/IV melanomaSecond-line therapy in combination with chemotherapy
    Advanced/metastatic pancreaticSecond-line therapy in combination with chemotherapy
Cediranib (AstraZeneca; Wilmington, Del)Wedge 200549TKIVEGFR-1, VEGFR-2, and VEGFR-3; PDGFR-α and PDGFR-ß; FGFR-1; and c-KITAdvanced/metastatic NSCLCFirst-line and second-line in combination with chemotherapy
    Metastatic CRCFirst-line in combination with chemotherapy
    Recurrent glioblastomaSecond-line with and without chemotherapy
    Relapsed gynecologicSecond-line in combination with chemotherapy and maintenance therapy
BIBF 1120 (Boehringer Ingelheim, Ingelheim, Germany)Hilberg 200850TKIVEGFR-1, VEGFR-2, and VEGFR-3; PDGFR-α and PDGFR-ß, FGFR-1, FGFR-2, and FGFR-3; FLT-3; and srcAdvanced NSCLCSecond-line in combination with chemotherapy
    Advanced ovarianFirst-line in combination with chemotherapy
Pazopanib (GlaxoSmithKline, London, United Kingdom)GlaxoSmithKline 201022TKIVEGFR-1, VEGFR-2, and VEGFR-3; PDGFR-α and PDGFR-ß; FGFR-1 and FGFR-3; c-KIT; Itk; Lck; and c-fmsInflammatory HER2-positive breastSecond-line therapy in combination with lapatinib
    Advanced gynecologicSecond-line therapy
    Advanced/metastatic RCCFirst-line and second-line therapy and adjuvant therapy after nephrectomy
    Metastatic soft tissue sarcomaSecond-line therapy
Brivanib (Bristol-Myers Squibb, New York, NY)Huynh 200851TKIVEGFR-1, VEGFR-2, and VEGFR-3; and FGFR-1, FGFR-2, and FGFR-3Advanced HCCFirst-line or second-line therapy
    Metastatic CRCSecond-line therapy in combination with cetuximab
Motesanib (Amgen, Thousand Oaks, Calif)Fujisaka 2010,52 Polverino 200653TKIVEGFR-1, VEGFR-2, and VEGFR-3; PDGFR-β; c-KIT; and RETAdvanced nonsquamous NSCLCSecond-line in combination with chemotherapy
ABT-869 (Abbott, Abbott Park, Ill)Albert 2006,54 Shankar 200755TKIVEGFR-1, VEGFR-2, and VEGFR-3; PDGFR-β; c-KIT; CSF-1R; and FLT-3Advanced/metastatic HCCFirst-line or second-line therapy
Axitinib (Pfizer, New London, Conn)Rugo 2005,56 Hu-Lowe 200857TKIVEGFR-1, VEGFR-2, and VEGFR-3; PDGFR-β; and c-KITMetastatic RCCFirst-line or second-line therapy
    Advanced pancreaticFirst-line in combination with chemotherapy

Inhibition of the VEGF pathway as an approach to antiangiogenic therapy is well established; however, not all patients respond to anti-VEGF therapy, and patients that do respond eventually may experience progressive disease.58 Because proangiogenic VEGF signaling seems highly integrated with other pathways, such as platelet-derived growth factor (PDGF)59, 60 and FGF signaling,61, 62 it is believed that targeting additional angiogenic and/or proliferative compensatory pathways could prevent or overcome resistance. Synergistic angiogenic effects have been described among these signaling pathways and VEGF, and both PDGF and FGF signaling have been associated with resistance to VEGF-targeted therapy.63-67

Clinically, resistance to VEGF-targeted therapy may be defined as the evidence of disease progression according to Response Evaluation Criteria in Solid Tumors despite therapy.68 However, it is recognized that resistance to VEGF-targeted and other targeted therapies may fall within a span of outcomes and is not identical68, 69; these outcomes may include intrinsic nonresponsiveness (failure to achieve any tumor shrinkage) with more rapid progression, minor antitumor effects followed by progression, or prolonged tumor shrinkage that may be followed by a slower progressive disease. Patients who experience the latter outcome may present with progressive disease with a lower total tumor burden than before therapy.69 Malignancies can develop various patterns to circumvent VEGFR blockade,70, 71 as described below, and overcoming such resistance poses challenges for future cancer therapy.

Major Mechanisms

Redundant signaling of receptors

Human cells possess various alternate pathways to overcome aberrations in normal signaling. These same pathways may offer resistance to currently available anti-VEGF drugs. It has been demonstrated that treatment with anti-VEGF drugs in pancreatic tumors up-regulates FGF-1 and FGF-2, angiopoietin-1, ephrin-A1, and ephrin-A2.63 Increased levels of FGF-2, stromal cell-derived factor 1 (SDF-1), and circulating endothelial cells (CECs) are observed upon disease progression during VEGFR blockade in patients with glioblastoma multiforme.72 Overlapping VEGFR specificity, such as the overlap of VEGF-C and VEGF-D, can take over signaling in the event of VEGF-A blockade.73 Up-regulation of PDGF-C in tumor-associated fibroblasts can be observed in tumors that are refractory to anti-VEGF therapy and, hence, represent another alternate pathway by which tumors can resist therapy.74 Such pathways can result in an escape mechanism from proangiogenic blockade and result in alternate signaling toward the direction of least resistance from currently available agents (Fig. 2).

thumbnail image

Figure 2. Alternate pathways contribute to resistance to antivascular endothelial growth factor (anti-VEGF) therapy. With anti-VEGF therapy, alternate factors (fibroblast growth factor [FGF], angiopoietin, ephrin, stromal cell-derived factor 1 [SDF-1], and other VEGFs) can bind to VEGF receptor 2 (VEGFR-2) and other receptors on the surface of the endothelial cell. These compensatory pathways allow proangiogenic signaling to continue and facilitate resistance to anti-VEGF therapy. PDGF-C indicates platelet-derived growth factor C.

Download figure to PowerPoint

Selection of hypoxia-resistant malignant clones

Under anti-VEGF therapy, particular clones can be selected that have the ability to survive antiangiogenic therapy-induced hypoxia and potentially metastasize (Fig. 3). Increased invasiveness and metastasis have been observed in pancreatic islet cell tumors and glioblastomas in mice upon VEGFR and PDGFR blockade.75 Similarly, in breast cancer and malignant melanoma, antiangiogenic therapy inhibits primary tumor growth but facilitates metastasis.76 Although hypoxia is induced by anti-VEGF therapy, the selection of hypoxia-resistant clones (through loss of p53) also has been observed with VEGFR-2 inhibition.77 These clones require fewer proangiogenic factors to promote their growth and proliferation and, thus, are indirectly resistant to current therapy. Because therapy with selective anti-VEGF agents only blocks 1 factor, other proangiogenic factors still can override the blockade and use alternate pathways to favor angiogenesis and extracellular matrix degradation for possible increased tumor invasiveness and metastasis. Thus, hypoxia, which leads to increased VEGF signaling/angiogenesis, may continue to be a driving force for angiogenesis (after transient hypoxia is induced by initial anti-VEGF therapy) through the selection of hypoxia-resistant clones.

thumbnail image

Figure 3. Increased metastasis/invasiveness during vascular endothelial growth factor (VEGF) blockade is illustrated. Anti-VEGF therapy can lead to the inhibition of primary tumor growth, but an increase in metastasis may occur because of the selection of resistant tumor cells. These resistant clones migrate away from the primary tumor into surrounding tissue and organs during anti-VEGF therapy.

Download figure to PowerPoint

Myeloid and circulating cell recruitment for vascular endothelial growth factor-independent angiogenesis

Bone marrow-derived circulating endothelial precursors (CEPs), CECs, pericyte progenitors, tumor-associated macrophages, endothelial TEK tyrosine kinase 2 (Tie 2)-expressing monocytes, fibroblasts, and neutrophils can be recruited to intratumor areas of hypoxia to override anti-VEGF therapy.72, 78-80 Such circulating cells, like the tumor-associated protein gamma response 1-positive and cluster of differentiation molecule 11B (integrin alpha M) (CD11b+/Gr1+) myeloid cells, can result in resistance to anti-VEGF therapy by contributing to angiogenesis aided by endogenous granulocyte-colony–stimulating factor (G-CSF)81 and possibly by exogenous administration of G-CSF after various chemotherapy regimens or other various causes of neutropenia (Fig. 4).

thumbnail image

Figure 4. Circulating and supporting cells continue angiogenesis during vascular endothelial growth factor (VEGF) blockade. When the VEGF pathway is blocked, other cell types facilitate resistance and continued angiogenesis. These include bone marrow-derived circulating endothelial precursors (CEPs), circulating endothelial cells (CECs), pericyte progenitors, tumor-associated macrophages, endothelial TEK tyrosine kinase 2 (Tie 2)-expressing monocytes, fibroblasts, and neutrophils. CD11b+Gr1+ indicates protein gamma response 1-positive and cluster of differentiation molecule 11B (integrin alpha M)-positive; G-CSF, granulocyte-colony–stimulating factor.

Download figure to PowerPoint

Increase in circulating nontumor proangiogenic factors

Dose-dependent increases in circulating proangiogenic nontumor secreted factor levels results in tumor angiogenesis by a resultant increase in alternate signaling.82 Systemic increases in nontumor proangiogenic factors like VEGF, G-CSF, SDF-1, osteopontin, and stem cell factor have been observed after administration of systemic VEGFR/PDGFR tyrosine kinase inhibitors because of inhibition of normal pathways. An increase in expression of these factors possibly may fuel alternate signaling pathways for angiogenesis, overwhelm anti-VEGF therapy targets, and cause resistance or may require dynamic adjustment in dosage of therapy. PlGF, FGF-B, or basic fibroblast growth factor (bFGF) and hepatocyte growth factor (HGF) observed in colon cancer recurrence83 can signal a systemic compensatory mechanism, which would warrant using inhibitors of VEGFR tyrosine kinase, bFGF, or mesenchymal-epithelial transition factor (MET). Hypothetically, these similar mechanisms also could fuel progression of secondary metastatic disease (Fig. 5).

thumbnail image

Figure 5. Systemic increases in circulating proangiogenic factors contribute to resistance to antivascular endothelial growth factor (anti-VEGF) therapy. During anti-VEGF therapy, levels of nontumor systemic factors (other VEGFs, granulocyte-colony–stimulating factor [G-CSF], stromal cell-derived factor 1 [SDF-1] and SDF-2, osteopontin, stem cell factor, placental growth factor [PlGF], and fibroblast growth factor [FGF]) may increase. The up-regulation of these factors may overwhelm VEGF receptor blockade and allow signaling through alternate angiogenic pathways.

Download figure to PowerPoint

Other postulated mechanisms
Heterogeneous angiogenic factor requirement of tumor cells.

The heterogeneity of various tumor types also can make them inherently less or more responsive to various antiangiogenic factors. Anti-VEGF therapy in neuroblastoma appears to be less effective than in Wilms tumor xenografts in preclinical studies.84 Similar heterogeneity has been reported within renal cell carcinoma, in which some tumors do not respond to anti-VEGF therapy, whereas other subsets will have an impressive response of stable disease for up to 3 to 5 years.85-87

Inadequate duration or dosage of therapy.

Some metastases are so small before the initiation of VEGF blockade that they can rely on diffusion from surrounding tissue for their nutritional requirements. Such micrometastases, which can survive without neovascularization, can persist beyond the currently recommended duration of anti-VEGF therapy and progress after tumor dormancy.88 For patients with advanced renal cell carcinoma, a higher dose of anti-VEGF therapy was required in clinical trials than that previously calculated as optimal from preclinical studies. On the basis of these results, calculating the adequate dose of such agents to provide optimal inhibition of VEGF signaling may pose a problem.85

Endothelial cell mutations.

On initial development of antiangiogenic drugs, it was believed that the endothelial cells forming the tumor vasculature were not prone to the same increased mutation rate as that of the malignancy itself, but it has been demonstrated recently that tumor endothelial cells are cytogenetically abnormal and can have tumor-related mutations.89-91 Such mutations potentially can lead to conformational changes in receptors and the proangiogenic factor requirement of tumor-associated vessels. Changes also may affect the expression profile and the resultant sensitivity to available antiangiogenic agents.

Organization of vessel architecture.

Detailed vascular architecture analysis of melanoma metastasis after anti-VEGF therapy has indicated that such tumor vessels have enhanced vessel diameter, mature pericytes, immunoreactivity for desmin, PDGFR-β, and late-stage maturity marker α smooth muscle actin indicating increased vessel maturity, all of which may be because of anti-VEGF therapy.92 Vascular remodeling by PDGF-B and ephrin-B2 up-regulation causes increased pericyte coverage, hence decreasing the amount of VEGF required for malignant angiogenesis.93, 94 Such enhanced vessel architecture forms a complicated barrier to current therapy, which only reverses certain features of the vessel structure and not all of the various structural elements of a mature vessel.

Vascular sleeves.

Although endothelial cells may die from anti-VEGF therapy, empty vascular sleeves persist within the tumor that can serve as channels for endothelial cell proliferation when anti-VEGF therapy is halted.95 These channels are formed by the persistence of pericytes and vascular basement membrane forming a scaffold for endothelial cell regrowth. Endothelial cell proliferation then rapidly occurs through this scaffold, suggesting either that maintenance antiangiogenic therapy or drugs consistent with the destruction of these supporting structures may be necessary to overcome resistance and inhibit growth.

Vascular cooption.

Vascular cooption is the process by which malignant cells grow around pre-existing normal vessels and draw their oxygen and other essential nutrients without the need for new vasculogenesis.96 This process has been reported previously in lung cancer cells97 and glioblastomas98, 99 under anti-VEGF therapy.

Intussusceptive angiogenesis.

In the absence of VEGF stimulation, blood vessels can split into new vessels without the need for endothelial proliferation.100 Intussusceptive angiogenesis is a relatively fast process in which existing endothelial cells migrate and are remodeled by increasing in volume and becoming thinner. Because existing endothelial cells are assimilated, the requirement for VEGF-mediated endothelial proliferation can be overcome by the tumor. It is believed that the factors that influence this process are blood flow dynamics and shear stress and that such dynamics result in endothelial-endothelial and endothelial-pericyte interactions, cytoskeletal rearrangements, and adaptation of gap junction complexes. It is also hypothesized that inhibition of sprouting angiogenesis may even stimulate intussusceptive angiogenesis.

Future Prospects

Although many pathways of resistance to anti-VEGF therapy have been identified to date (Fig. 6), all of this information adds to our understanding of this critical process of angiogenesis in tumor growth and metastasis. By the combined inhibition of various redundant proangiogenic factors and their pathways, these mechanisms of resistance potentially may be overcome. Such a response has been noticed in concurrent inhibition of HIF-1 and VEGF,101 antibody-directed blockade of PlGF,102 antibody-directed blockade of integrin αvβ3,103 and anti-interleukin-6 therapy in gliomas104 and other examples. In addition, several multitargeted antiangiogenic TKIs currently are approved (Table 1) and/or in phase 3 of clinical development (Table 2) for the treatment of various cancer types.

thumbnail image

Figure 6. This is an overview of the mechanisms of resistance to vascular endothelial growth factor (VEGF) blockade. Resistance to VEGF blockade may occur through a variety of mechanisms, including alternative signaling pathways, increased metastasis and invasion caused by the selection of resistant tumor cells, proangiogenic circulating and supporting cells, and increased circulating proangiogenic factors.

Download figure to PowerPoint

Because the VEGF gene is highly polymorphic, discovery and further elaboration of isoforms of VEGF-A or other proangiogenic factors may result in the ability to select subgroups that would benefit from existing anti-VEGF therapy. Depending on their gene expression profile, such subgroups may have relatively worse or better inherent disease outcomes. The identification of certain subgroups also may help us to identify the candidates for antiangiogenic therapy who are most likely to experience benefits in overall survival or to select for patients who are at higher risk of drug-induced toxicity41 for which tailored, pre-emptive strategies can be undertaken.

Blocking the migration of myeloid cells, CECs, and inflammatory cells by the inhibition of colony-stimulating factors105, 106 also appears to be a new promising front for antiangiogenic therapy. This hypothesis can be applied in combination with chemotherapy and also should be taken into consideration with liberal administration of parenteral G-CSF to avoid any of its possible proangiogenic effects.

Vascular architecture disruption also can be influential in tumor antiangiogenesis and prevention of metastasis. Endothelial-specific pathways like delta-like 4 (Dll-4)-mediated notch signaling attenuation, can cause excessive branching and sprouting. Such excessive disruption can result in functionally defective, chaotic vessel architecture, leading to growth inhibition.107

By using information thus gained from studies of the above-mentioned mechanisms of resistance, targeted therapy to potentially overcome resistance and possibly assist conventional chemotherapy agents can be developed. Such agents can be used as single agents, in combination with currently available chemotherapy agents, or in new metronomic regimens that use each drug at its most critical step, thus inhibiting both tumor growth and metastasis.


  1. Top of page
  2. Abstract

Financial support for medical and editorial assistance was provided by Boehringer Ingelheim Pharmaceuticals, Inc.


Dr. Roman Perez-Soler is a consultant and a speaker for Genentech and Roche.


  1. Top of page
  2. Abstract