Neovascular AMD: Not Only VEGF
Delia N. Sang, MD, FACS
Ophthalmic Consultants of Boston
Schepens Eye Research Institute
Harvard Medical School, Boston, MA
The histopathogenesis of neovascular age-related macular degeneration (AMD) is not completely defined, but appears to be multifactorial, involving increased permeability, angiogenesis, anomalous matrix alteration, and remodeling with an inflammatory component.1, 2 The overwhelming success of the use of anti-VEGF blockade with ranibizumab in the treatment of neovascular AMD3 has revolutionized the outcomes associated with treatment of exudative AMD, with both compelling stabilization of choroidal neovascularization and high rates of improvement or stabilization of vision. However, clearly not all patients benefit uniformly or to the same degree,4, 5, 6 and this is likely related to stages of the neovascularization where factors other than VEGF may play a dominant role in the progression of the disease. In addition, there is both ocular7, 8 and systemic evidence to suggest that inflammatory mechanisms play a role in AMD.9, 10
Angiogenesis
The term "vasculogenesis" refers to early stages of endothelial differentiation from endothelial precursor cells into primitive vascular structures. This is in contrast to "angiogenesis", which refers to neovascularization originating from pre-existent formed vessels.11 The stages of angiogenesis are associated with degeneration of basement membranes, vascular permeability, degradation of extracellular matrix, endothelial cell proliferation and migration, and endothelial cell organization with formation of tubules and lumens, forming differentiated vessels.11 The causative factors controlling the different stages of angiogenesis are likely to be responsive to different approaches to treatment, and provide justification for combination or sequential therapy in AMD. Also, evidence for both cellular and immunologic participation in the development of drusen has been established.12
Role of Complement
There is increasing evidence to suggest that the complement system plays a key role in the development of drusen,7 and in the conversion from the dry to the exudative form of the condition, particularly in stages prior to the emergence of the neovascular membrane.
It has been hypothesized that initiation of the complement cascade with persistent activation is associated with development of AMD.13 Complement components C3a and C5a upregulate VEGF expression,14 and have been documented both in specimens from humans with AMD and in experimental models of laser-induced choroidal neovascularization.15, 16 Both complement and associated immune complexes may be implicated in damage to the retinal pigment epithelium (RPE).17
The complement factor H (CFH) gene has been identified as associated with development of AMD,18 supporting the hypothesis that inflammatory events are involved in the histopathogenesis of AMD. Up to 43 to 50% of AMD cases may be associated with the expression of a single nucleotide substitution in CFH,19, 20 and CFH has been reported to be associated with a significantly increased risk for AMD with an odds ratio of 2.45 to 5.57.19, 20 A CFH polymorphism, with a tyrosine to histidine replacement at position 402 of the CFH gene, appears to be a major risk factor for exudative macular degeneration,21 and this has been shown to be true for both dry geographic atrophy and choroidal neovascularization.22 This risk variant has also been associated with the finding of complement both in choroidal capillaries and other vessels.23 Importantly, this involves a region of the CFH gene that binds heparin and C-reactive protein,24 and certain C-reactive protein haplotypes influence the expression of the CFH gene.25 In addition, CFH genotypes may be associated with size of choroidal neovascular lesions and response to treatment.26
Role of Macrophages
Hageman et al12 and Penfold et al27 published compelling evidence to suggest that drusen formation is associated with an inflammatory response, and that macrophages are particularly implicated. Grossnicklaus and others28 have demonstrated macrophages as the predominant type of leukocyte involved in choroidal neovascular membranes in histological studies. Cousins et al identified activation of monocytes as a possible biomarker for exudative AMD29 and macrophages in choroidal neovascularization were found to be derived from bone marrow in experimental choroidal neovascularization.30
Earlier reports documented that lymphocytes, fibroblasts, and myofibroblasts may also play a role.31 Anderson et al hypothesized that degenerated RPE cells participate in drusen formation and act as a nidus for a chronic inflammatory response.7 In a laser-induced animal model, a rise in VEGF precedes macrophage infiltration,32 although both macrophages and RPE cells are associated with a second, later rise in VEGF. Kamei et al have demonstrated the presence of oxidized lipoproteins in choroidal neovascular membranes, and that macrophages express cell surface receptors for those lipoproteins.33
Role of PDGF
PDGF (platelet-derived growth factor) is a growth factor which may also play a role in ocular neovascularization, and has been documented to be a mitogen for pericytes, as well as smooth muscles cells, fibroblasts, and other mesenchymal cell types. Pericytes appear critical for the maintenance of established blood vessels and may be important in the maturation process in choroidal neovascularization.
It has been hypothesized that there is a period of VEGF dependency of endothelial cells34 that overlaps with a period of responsiveness to PDGF withdrawal in newly formed neovascular vessels.35 Established mature vessels may be associated with resistance to anti-VEGF blockade36 or rapid recurrence of choroidal neovascularization in AMD.
PDGF has been demonstrated to stimulate angiogenesis and pericyte recruitment.37, 38 Loss of pericytes in retinal vessels is thought to be associated with abnormalities of the vasculature and instability, including the formation of microaneurysms and vascular permeability. Loss of pericytes with PDGF-blockade may be associated with regression of maturing neovascularization.35 Pericytes produce VEGF-A under selected conditions, possibly thereby protecting endothelial cells when generalized suppression of VEGF-A is present.39
Jo et al demonstrated increased suppression of neovascularization in experimental models with specific blockade of both VEGF-A and PDGF-B as compared with blockade with VEGF-A alone.36 In addition, neovascularization became more resistant to VEGF-A blockade with progression. Suppression of VEGF-A and PDGF-B together caused regression of neovascularization not responsive to anti-VEGF-A alone.36
Role of Integrins
Integrins are transmembrane proteins composed of heterodimers with a single α and β chain, which are cell surface receptors, several of which have been demonstrated to stimulate endothelial cell migration and macrophage recruitment under specific conditions.40
Evidence for participation of β1 integrins in angiogenesis is accumulating, in particular in terms of maturation and organization of neovascularization. Activation of integrin adhesion receptors appears to play a key role in both embryologic and pathologic angiogenesis, and progression of β1 integrin expression from angiogenesis to maturation of vessels in the central nervous system has been demonstrated.41 Blockade of integrins has been shown to inhibit endothelial cell proliferation and induce apoptosis.42 In addition, several documented inhibitors of angiogenesis, including endostatin, have been demonstrated to bind to integrins during activation of anti-angiogenic function.43 VEGF-mediated cell adhesion and endothelial cell migration has been shown to be suppressed by anti-α5β1 antibodies.44
Recent increasing interest in the role of integrin receptors in tumor angiogenesis11 has raised questions of integrin involvement in neovascular AMD. The integrin α5β1 is the most important of the fibronectin receptors and blockade of α5β1 can lead to (1) suppression of tube formation stimulated by VEGF, and (2) apoptosis of proliferation but not quiescent endothelial cells.42 Importantly, in an experimental model, anti-α5β1 has been demonstrated to inhibit angiogenesis both stimulated by VEGF and independent of VEGF, while anti-VEGF treatment inhibited only VEGF-stimulated angiogenesis.42
Conclusion
VEGF is a dominant factor in pathological neovascularization in AMD and its inhibition is important in the suppression of choroidal neovascularization and VEGF-related permeability.45 However, the complex and sequential nature of progressive AMD is recognized and anti-VEGF monotherapy alone may not be as effective as combination therapy in controlling established neovascularization.36
Increasing awareness of the structural and inflammatory aspects of angiogenesis, as well as the recognition of factors in addition to VEGF which appear to be important in the different stages of maturation of neovascularization, may hopefully lead to optimization of pharmaceutical approaches to exudative AMD.
Several complement, integrin, and PDGF-targeted therapeutic agents are currently in clinical trials being initiated for treatment of AMD. The sequential participation of these factors in angiogenesis suggests a rationale for combination therapy in the treatment of AMD.
REFERENCES