Ординатура / Офтальмология / Английские материалы / Diabetic Retinopathy_Lang_2007

55Lamberts SWJ, Van Der Lely AJ, De Herder WW, Hofland LJ: Drug therapy: octreotide. N Engl J Med 1996;334:246–254.

56Hyer SL, Sharp PS, Brooks RA, Burrin JM, Kohner EM: Continuous subcutaneous octreotide infusion markedly suppresses IGF-I levels whilst only partially suppressing GH secretion in diabetics with retinopathy. Acta Endocrinol (Copenh) 1989;120:187–194.

57Mallet B, Vialettes B, Haroche S, Escoffier P, Gastaut P, Taubert JP, Vague P: Stabilization of severe proliferative diabetic retinopathy by long-term treatment with SMS 201-995. Diabetes Metab 1992;18:438–444.

58Lee HK, Suh KI, Koh CS, Min HK, Lee JH, Chung H: Effect of SMS 201-995 in rapidly progressive diabetic retinopathy. Diabetes Care 1988;11:441–443.

59Kirkegaard C, Norgaard K, Snorgaard O, Bek T, Larsen M, Lund-Andersen H: Effect of one year continuous subcutaneous infusion of a somatostatin analogue, octreotide, on early retinopathy, metabolic control and thyroid function in type I (insulin-dependent) diabetes mellitus. Acta Endocrinol (Copenh) 1990;122:766–772.

60Shumak SL, Grossman LD, Chew E, Kozousek V, George SR, Singer W, Harris AG, Zinman B: Growth hormone suppression and nonproliferative diabetic retinopathy: a preliminary feasibility study. Clin Invest Med 1990;13:287–292.

61McCombe M, Lightman S, Eckland DJ, Hamilton AM, Lightman SL: Effect of a long-acting somatostatin analogue (BIM23014) on proliferative diabetic retinopathy: a pilot study. Eye 1991;5: 569–575.

62Kuijpers RW, Baarsma S, van Hagen PM: Treatment of cystoid macular edema with octreotide. N Engl J Med 1998;338:624–626.

63Grant MB, Mames RN, Fitzgerald C, Hazariwala KM, Cooper-DeHoff R, Caballero S, Estes KS: The efficacy of octreotide in the therapy of severe nonproliferative and early proliferative diabetic retinopathy: a randomized controlled study. Diabetes Care 2000;23:504–509.

64Boehm BO, Lang GK, Jehle PM, Feldmann B, Lang GE: Octreotide reduces risk for vitreous hemorrhages and loss of visual acuity in patients with high risk proliferative diabetic retinopathy. Horm Metab Res 2001;33:300–306.

65Growth Hormone Antagonist for Proliferative Diabetic Retinopathy Study Group. The effect of a growth hormone receptor antagonist drug on proliferative diabetic retinopathy. Ophthalmology 2001;108:2266–2272.

66Davies RR, Turner SJ, Alberti KG, Johnston DG: Somatostatin analogues in diabetes mellitus. Diabet Med 1989;6:103–111.

67Colao A, Merola B, Ferone D, Marzullo P, Cerbone G, Longobardi S, Di Somma C, Lombardi G: Acute and chronic effects of octreotide on thyroid axis in growth hormone-secreting and clinically non-functioning pituitary adenomas. Eur J Endocrinol 1995;133:189–194.

68Aiello LP: Clinical implications of vascular growth factors in proliferative retinopathies. Curr Opin Ophthalmol 1997;8:19–31.

69Duh E, Aiello LP: Vascular endothelial growth factor and diabetes: the agonist versus antagonist paradox. Diabetes 1999;48:1899–1906.

70Ferrara N, Gerber HP, LeCouter J: The biology of VEGF and its receptors. Nat Med 2003;9: 669–676.

71Porta M, Allione A: Current approaches and perspectives in the medical treatment of diabetic retinopathy. Pharmacol Ther 2004;103:167–177.

72Croxen R, Baarsma GS, Kuijpers RW, van Hagen PM: Somatostatin in diabetic retinopathy. Pediatr Endocrinol Rev 2004;1(suppl 3):518–524.

Bernhard O. Boehm, MD

Division of Endocrinology and Diabetes, Ulm University Robert-Koch-Strasse 8

DE–89081 Ulm (Germany)

Tel. 49 731 500 44504, Fax 49 731 500 44506, E-Mail bernhard.boehm@uniklinik-ulm.de

prevalence of diabetic retinopathy is expected to rise as the number of people with diabetes increases due to the demographic effects of population growth, aging and urbanization and the growing prevalence of obesity and physical inactivity [3]. This will further add to the human and economic burden that diabetic retinopathy and its ensuing vision loss are already imposing on our society [2].

Severe visual loss in patients with diabetes occurs primarily as a consequence of retinal neovascularization and complications resulting from intraocular angiogenesis; moderate visual loss results primarily from diabetic macular edema (DME) related to altered permeability of the retinal vasculature. Proliferative diabetic retinopathy is more commonly reported in patients with type 1 diabetes, whereas DME is more commonly associated with type 2 diabetes [4]. While the pathogenesis of diabetic retinopathy is incompletely understood, evidence suggests that it is one of several ocular diseases characterized by neovascularization and increased vascular leakage ultimately driven by the effects of vascular endothelial growth factor (VEGF) [5–7].

Laser photocoagulation is the current standard of care for the treatment of sight-threatening diabetic retinopathy. Focal photocoagulation, primarily used for treating DME, applies small-sized burns to leaking microaneurysms, while scatter (panretinal) photocoagulation is employed for proliferative retinopathy and indirectly treats neovascularization by placing burns throughout the fundus [8]. While the use of laser photocoagulation has greatly reduced the risk of developing severe vision loss, this is accomplished by attendant destruction of retinal tissue that can lead to side effects, such as loss of peripheral vision, alterations in color perception, and perceptions of night blindness. Pars plana vitrectomy is another option for treating complications of severe proliferative retinopathy and/or hemorrhage [5, 8]. Although instrumentation and surgical techniques have improved during the past decade [9], pars plana vitrectomy is still associated with several complications [5, 9–11], and for this reason, visual acuity (VA) outcomes are still poor [9].

The mechanism by which scatter laser photocoagulation reduces proliferative retinopathy is not known. It has been proposed that light energy absorbed by melanin in the retinal pigment epithelium destroys highly metabolically active outer retinal cells, reducing retinal oxygen consumption and facilitating improved oxygen diffusion from the choriocapillaris through the laser scars [12]. Increased oxygen tension may lead to vasoconstriction, further reducing the edema. Therefore, laser photocoagulation is directed at reducing the retinal neovascularization or macular edema rather than reversing the underlying biological process of diabetic retinopathy. The risk of VA loss is reduced, and substantial recovery of reduced VA is relatively unusual [4, 5, 13].

Because of the limitations, potential side effects and complications of currently available treatments for diabetic retinopathy, research continues to be directed toward the development of novel, more effective and nondestructive therapeutic modalities. Over the past decade, a large body of work has established VEGF as a major regulator of both physiological and pathological vessel growth and vascular permeability, and that it plays a key role in ocular neovascular diseases, such as age-related macular degeneration and diabetic retinopathy [6]. Importantly, studies in animal models have demonstrated that the single isoform VEGF164 (the rodent counterpart of human VEGF165) is particularly important in the pathogenesis of diabetic retinopathy [14].

Macugen® (pegaptanib sodium), a pegylated synthetic ribonucleic acid (RNA) oligonucleotide, specifically inhibits the actions of VEGF165 [15]. The oligonucleotide portion of the molecule, called an aptamer, was designed to bind selectively to extracellular VEGF165 as compared with antisense oligonucleotides, which have an intracellular site of action. To increase the in vivo residence time of pegaptanib, a 40-kDa branched polyethylene glycol molecule has been conjugated to the oligonucleotide. Based on the efficacy demonstrated in two large, multicenter, randomized clinical trials (the VEGF Inhibition Study in Ocular Neovascularization, or VISION, trials) [16], Macugen has been approved for the treatment of neovascular age-related macular degeneration in the United States, Canada and Brazil and has received a recommendation for market authorization by the Committee for Human Medicinal Products of the European Union. The selective pharmacologic blockade of the 165 isoform of VEGF with pegaptanib also has potential applicability in the treatment of other diseases characterized by retinal revascularization and increased retinal vascular permeability. Indeed, a phase II clinical trial exploring its safety and efficacy in patients with DME has reported encouraging early results [17].

This chapter will first review the role of VEGF165 in the pathogenesis of ocular neovascular diseases, including diabetic retinopathy. A description of the development of pegaptanib sodium as an anti-VEGF agent will follow, together with a review and discussion of the recent phase II clinical trial evaluating its use in patients with DME [17]. The findings from this trial not only validate the hypothesis that VEGF165 plays an important role in the pathogenesis of diabetic retinopathy, but also offer the promise of a new and less destructive treatment option for DME.

VEGF in Ocular Neovascular Disease

VEGF Is a Pluripotent Growth Factor

VEGF (also known as VEGF-A) is a member of the VEGF-platelet- derived growth factor family, which also includes VEGF-B, VEGF-C, VEGF-D

Table 1. Proand antiangiogenic factors [21]

and VEGF-E [for a general review of VEGF, see ref. 18]. It was isolated independently by 2 groups, first as a vascular permeability factor [19] and second as a potent endothelial cell mitogen [20]. Although investigations during the 1980s suggested numerous proangiogenic and antiangiogenic factors, in a list that has since continued to grow (table 1) [21], only VEGF convincingly showed all the characteristics of a necessary and sufficient inducer of angiogenesis [22]. Alternative splicing of the VEGF gene yields at least 6 distinct biologically active human isoforms, each comprised of a differing number of amino acids (e.g., 121, 145, 165, 183, 189 and 206). VEGF165, the predominant

isoform, is a 45-kDa homodimeric glycoprotein existing both free in the cytoplasm and bound through a heparin-binding domain to the cell surface and extracellular matrix and is the isoform principally responsible for mediating the pathological effects of VEGF in ocular neovascular diseases [23, 24]. VEGF189 and VEGF208 also contain this heparin-binding domain and are strongly basic, and for the most part, they are sequestered in the extracellular matrix while VEGF121 is acidic, lacks the heparin-binding domain and is secreted [25]. VEGF is a ligand for 2 receptor tyrosine kinases, VEGFR-1 and VEGFR-2, mediating their activation of downstream signal transduction cascades [18]. While VEGFR-2 is believed to be the principal receptor for VEGF signaling in angiogenesis [18], VEGFR-1 also plays a key role in pathological ocular neovascularization through mediating monocyte chemotaxis to VEGF [24, 26].

Much research effort has been applied toward understanding the function of VEGF with the goal of inhibiting, if not reversing, pathological angiogenesis. However, it is becoming increasingly evident that VEGF is a pluripotent growth factor that is active not only in angiogenesis but also in a variety of physiological contexts. For example, there is recent evidence that VEGF serves as a neurotrophic role, lending hope that the administration of VEGF may have benefits in the treatment of neurodegenerative diseases and optic neuropathies [27, 28]. In the eye, VEGF121 appears to be sufficient to exert this neuroprotective action, which may serve to counteract the effects of retinal ischemia [29]. In addition, VEGF secretion by the retinal pigment epithelium has been implicated in trophic maintenance of the choriocapillaris [30], much like the trophic role it plays in other vascular beds [31].

VEGF also has been implicated in a variety of other vital and required physiological processes, including bone growth [32, 33], wound healing [34, 35], female reproductive cycling [32, 36], vasorelaxation [37], skeletal muscle regeneration [38], glomerulogenesis [39] and protection of hepatic cells [40]. Given this wide range of actions, antiangiogenic therapies that target VEGF need vigorous monitoring of safety, an issue of particular relevance for systemic administration [41]. In this context, there is already evidence that intravenous administration of a monoclonal antibody that binds all VEGF isoforms is associated with an increased incidence of hypertension, thromboembolism and hemorrhage [42– 47].

VEGF in Physiological and Pathological Angiogenesis

VEGF has a variety of properties in physiological and pathological neovascularization (table 2) [14, 18, 19, 23, 24, 26, 48–55]. In addition to its role as a potent endothelial cell mitogen, VEGF serves as an endothelial cell survival factor [56] and as a chemoattractant for bone marrow-derived endothelial cells [48, 57, 58]. It also induces the synthesis of several enzymes whose actions affect angiogenesis, including the matrix metalloproteinases and plasminogen activator; together, these promote degradation of the extracellular matrix,

Table 2. Key properties of VEGF in physiological and pathological angiogenesis [14, 18, 19, 23, 24, 26, 48–55]

Endothelial cell mitogen

Endothelial cell survival factor

Chemoattractant for bone marrow-derived endothelial cells

Potent enhancer of vascular permeability

Expression induced by hypoxia

Chemoattractant for monocyte lineage cells

Proinflammatory cytokine, promoting leukocyte adhesion

Inducer of synthesis of key enzymes

Matrix metalloproteinases

Plasminogen activator

Endothelial nitric oxide synthase

permitting blood vessel extravasation [49–51]. VEGF also induces endothelial nitric oxide synthase, leading to the upregulation of nitric oxide, a stimulator of angiogenesis [52, 53]. In addition, VEGF acts as a chemoattractant for monocyte lineage cells, which are believed to contribute to pathological ocular neovascularization [23, 24], and to promote local adhesion of leukocytes [14, 54].

Historically, much of the impetus for the isolation of angiogenic factors stems from the hypothesis that antiangiogenic approaches could serve to starve malignant tumors [59]. VEGF was evaluated in this context in the early 1990s when it was found that tumor vascularization and growth could indeed be inhibited by injections of a monoclonal antibody to VEGF [60]. Subsequently, the role of VEGF in supporting tumor growth was intensively examined, with the first anti-VEGF agent developed for clinical use (bevacizumab, an anti-VEGF monoclonal antibody) as an anticancer therapeutic [61]. VEGF has been implicated in several other classes of disorders involving dysregulation of angiogenesis, including hematological malignancies, inflammation, brain edema and several pathological conditions of the female reproductive tract [18].

In the course of evaluating the properties of VEGF, 2 were recognized that are of particular relevance to the pathogenesis of diabetic retinopathy. First, synthesis of VEGF is upregulated by hypoxia, which provides a mechanistic basis for VEGF-mediated ocular neovascularization in response to ischemia [18]. Secondly, VEGF is the most potent known inducer of vascular permeability, 50,000 times more potent than histamine [55]. Several additional mechanisms contributing to VEGF-mediated increases in vascular permeability have been elucidated, including induction of fenestrations in the endothelium [62], dissolution of tight junctions [63] and induction of leukostasis and subsequent injury to the endothelium [54, 64]. These properties of VEGF support a direct

involvement of VEGF in the macular edema that often accompanies diabetic retinopathy.

VEGF in Ocular Neovascularization

An extensive series of clinical and preclinical investigations has confirmed that VEGF plays a central role in promoting ocular neovascularization [7, 65–73]. Clinical studies have demonstrated elevated ocular levels of VEGF in patients with anterior segment neovascularization [7], retinal vein occlusion [7], neovascular glaucoma [74], retinopathy of prematurity [75], and DME [76 –78]. In other studies, increased expression of VEGF was detected within the macula of patients with age-related macular degeneration when compared with controls [79] and in choroidal neovascular membranes from patients with either age-related macular degeneration or diabetic retinopathy [80–82].

Preclinical studies examining VEGF in a variety of animal models of ocular neovascularization demonstrated that increased intraocular levels of VEGF can induce ocular neovascularization and that inactivation of VEGF in the eye can prevent the occurrence of ocular neovascularization [67–73, 83–90].

In one of the first of these preclinical studies, Miller et al. [83] reported that experimentally induced retinal vein occlusion in monkeys resulted in iris neovascularization and an associated increase in ocular VEGF levels. The severity of iris neovascularization was proportional to the concentration of VEGF [83]. In other studies, injection of VEGF into the vitreous of monkeys produced many of the features characteristic of diabetic retinopathy, including intraretinal and preretinal neovascularization, microaneurysm formation, intraretinal hemorrhage and edema, and areas of capillary nonperfusion with endothelial cell hyperplasia [72, 84]. Qualitatively similar data were obtained using molecular biological techniques. Injection of recombinant adenovirus vectors expressing VEGF into rodent eyes increased VEGF production in the retinal pigment epithelium, with resulting choroidal neovascularization [85, 86]. The severity and extent of vascular proliferation correlated with the amount of virus delivered [86]. Similarly, ocular neovascularization occurred in transgenic mice engineered to overexpress VEGF in the retinal pigment epithelium [73] or in photoreceptors [87]. In the latter study, neovascularization was sufficient in some instances to cause retinal detachment [87].

A variety of models have also been employed to demonstrate that blockade of VEGF and its receptors can inhibit the development of ocular neovascularization. Injection of anti-VEGF antibodies was shown to prevent the neovascularization of the monkey iris that normally followed laser occlusion of the retinal vein [68], and antibodies or their Fab fragments were also effective in preventing choroidal neovascularization in a photocoagulation-induced model in monkeys [70] and in a rat corneal wound model [67]. Other approaches for

inactivating VEGF that have prevented ocular neovascularization included administration of soluble VEGFR chimeric proteins by injection [69, 88] or expression from an adenovirus vector [89], injection of pegaptanib [14], and injection of an anti-VEGF antisense oligonucleotide [90].

VEGF in Diabetic Retinopathy

While the pathophysiology of diabetic retinopathy involves a complex interaction between many factors, current evidence supports a pivotal role of VEGF. Progression of diabetic retinopathy begins with alterations in the retinal vasculature characterized by the degeneration of retinal capillary pericytes, thickening of the basement membrane, and adhesion of leukocytes to the endothelium. These changes are accompanied by blockages of retinal capillaries, loss of endothelial cells, and the formation of acellular vessels, resulting in areas of local nonperfusion [4, 91]. The resultant hypoxia leads to local upregulation of factors such as VEGF [92]. Many retinal cell types express VEGF, including all classes of neurons, glia, endothelial cells, pericytes, and retinal pigment epithelium cells [30, 93, 94]. Hypoxia leads to dramatic increases in VEGF expression from these cells [30, 93, 95].

Several biochemical pathways are believed to be important in linking hyperglycemia to vascular injury in the retina, including the accumulation of polyols, advanced glycation end products and reactive oxygen intermediates; these compounds can produce vascular injury by affecting cellular metabolites and by induction of growth factors [4, 96]. Both advanced glycation end products [97] and reactive oxygen intermediates [98] can directly induce VEGF expression. While causative mechanisms remain to be fully elucidated, increased VEGF levels have been consistently observed in eyes with diabetic retinopathy [7, 65, 66, 99–101]. Early studies demonstrated that VEGF levels were higher in eyes with proliferative diabetic retinopathy than those with nonproliferative diabetic retinopathy; this finding has since been corroborated by other investigators [99–101].

Several other factors were subsequently shown to be elevated in conjunction with VEGF in diabetic retinopathy, including interleukin (IL)-6 [100], stromal-derived factor 1 [101] and angiopoietin II [102]. Similarly, elevated levels of VEGF, together with angiotensin II [76], IL-6 [77], stromal-derived factor 1 [101], and intercellular adhesion molecule (ICAM)-1 [78], have also been demonstrated in association with DME. In recent work, both VEGF and erythropoietin levels were found to be independent predictors of proliferative diabetic retinopathy [103]. To what extent these various factors act independently of VEGF production is unclear; both angiotensin II [104, 105] and stromal-derived factor 1 [106] induce VEGF production while ICAM-1 is upregulated in response to VEGF [24, 107].

Fig. 1. Retinal VEGF mRNA levels are increased in early diabetes. Adapted from Qaum et al. [108].

More recent evidence suggests that the proinflammatory effects of VEGF are important in the pathogenesis of ocular neovascular diseases such as diabetic retinopathy. Specifically, VEGF-mediated upregulation of ICAM-1, an adhesion receptor for leukocytes, may provide a link connecting some of the vascular changes seen in diabetic retinopathy to elevated VEGF levels. This conclusion is based on 2 lines of evidence. First, studies demonstrated that expression of both retinal VEGF mRNA [108] and ICAM-1 [109] was upregulated in rodent models of diabetic retinopathy. Second, studies in nondiabetic rats found that retinal ICAM-1 was upregulated in response to VEGF [24, 54]. Subsequently to its upregulation, ICAM-1 may then contribute to the vascular damage characteristic of diabetic retinopathy by promoting leukocyte entrapment (leukostasis) in capillaries, with accompanying local nonperfusion, vascular leakage and endothelial cell damage.

Importantly, VEGF165 has been established as the predominant pathological isoform responsible for inflammation and vascular injury characteristic of diabetic retinopathy as well as the ocular neovascularization that follows ischemia [14, 23, 108]. In rats made diabetic by injection of streptozotocin, retinal VEGF levels were increased 3.2-fold after 1 week (fig. 1) [108]. The effects of this elevation may have been compounded by the enhanced pathogenicity of VEGF164; VEGF164 has been shown to be approximately twice as potent as VEGF120 (the rodent counterpart of human VEGF121) in mediating the upregulation of