Ординатура / Офтальмология / Английские материалы / Retinal and Choroidal Angiogenesis_Penn_2008

123.E. Yamada et al., TIMP-1 promotes VEGF-induced neovascularization in the retina,

Histol. Histopathol. 16, 87-97 (2001).

124.K. Ohno-Matsui et al., Inducible expression of vascular endothelial growth factor in adult mice causes severe proliferative retinopathy and retinal detachment, Am. J. Pathol. 160, 711-719 (2002).

125.S. A. Vinores et al., Experimental models of growth factor-mediated angiogenesis and blood-retinal barrier breakdown, Gen. Pharmacol. 35, 233-239 (2000).

126.M. J. Tolentino et al., Pathologic features of vascular endothelial growth factor-induced retinopathy in the nonhuman primate, Am. J. Ophthalmol. 133, 373-385. (2002).

127.A. Madan and J. S. Penn, Animal models of oxygen-induced retinopathy, Front. Biosci. 8, d1030-d1043 (2003).

128.R. P. Danis and I. H. Wallow, Microvascular changes in experimental branch retinal vein occlusion, Ophthalmology 94, 1213-1221 (1987).

129.C. J. Pournaras, M. Tsacopoulos, K. Strommer, N. Gilodi, and P. M. Leuenberger, Experimental retinal branch vein occlusion in miniature pigs induces local tissue hypoxia and vasoproliferative microangiopathy, Ophthalmology 97, 1321-1328 (1990).

130.C. A. Wilson and D. L. Hatchell, Photodynamic retinal vascular thrombosis. Rate and duration of vascular occlusion, Invest. Ophthalmol. Vis. Sci. 32, 2357-2365 (1991).

131.R. P. Danis, Y. Yang, S. J. Massicotte, and H. C. Boldt, Preretinal and optic nerve head neovascularization induced by photodynamic venous thrombosis in domestic pigs, Arch. Ophthalmol. 111, 539-543 (1993).

132.M. Minamikawa, K. Yamamoto, and H. Okuma, H. [Experimental retinal branch vein occlusion. 4. Pathological changes in the middle and late stage]. Nippon Ganka Gakkai Zasshi 97, 920-927 (1993).

133.C. J. Pournaras, Retinal oxygen distribution: Its role in the physiopathology of vasoproliferative microangiopathies, Retina 15, 332-347 (1995).

134.R. P. Danis, D. P. Bingaman, Y. Yang, and B. Ladd, Inhibition of preretinal and optic nerve head neovascularization in pigs by intravitreal triamcinolone acetonide, Ophthalmology 103, 2099-2104 (1996).

135.R. Danis et al., Intravitreous anti-raf-1 kinase antisense oligonucleotide as an angioinhibitory agent in porcine preretinal neovascularization, Curr. Eye Res. 26, 45-54 (2003).

136.H. Akiyama et al., Inhibition of ocular angiogenesis by an adenovirus carrying the human von Hippel-Lindau tumor-suppressor gene in vivo, Invest. Ophthalmol. Vis. Sci. 45, 1289-1296 (2004).

137.R. J. Koletsky and P. Ernsberger, Obese SHR (Koletsky rat): a model for the interactions between hypertension and obesity, Genet. Hyperten. 218, 373-375 (1992).

138.S. S. Huang, S. A. Khosrof, R. J. Koletsky, B. A. Benetz, and P. Ernsberger, Characterization of retinal vascular abnormalities in lean and obese spontaneously hypertensive rats, Clin. Exp. Pharmacol. Physiol. 22(Suppl. 1), S129-S131 (1995).

139.M. Mizutani, T. S. Kern, and M. Lorenzi, Accelerated death of retinal microvascular cells in human and experimental diabetic retinopathy, J. Clin. Invest. 97, 2883-2890 (1996).

140.F. Podesta et al., Bax is increased in the retina of diabetic subjects and is associated with pericyte apoptosis in vivo and in vitro. Am. J. Pathol. 156, 1025-1032 (2000).

141.L. Zheng, S. Szabó, and T. Kern, Poly(ADP-ribose) polymerase is involved in the development of diabetic retinopathy via regulation of NF-6B, Diabetes 53, 2960-2967 (2004).

142.T. W. Gardner, D. A. Antonetti, A. J. Barber, K. F. LaNoue, and M. Nakamura, New insights into the pathophysiology of diabetic retinopathy: potential cell-specific therapeutic targets, Diabetes Technol. Ther. 2, 601-608 (2000).

143.A. M. Joussen et al., Leukocyte-mediated endothelial cell injury and death in the diabetic retina, Am. J. Pathol. 158, 147-152 (2001).

144.M. Lorenzi and C. Gerhardinger, Early cellular and molecular changes induced by diabetes in the retina, Diabetologia 44, 791-804 (2001).

145.A. P. Adamis, Is diabetic retinopathy an inflammatory disease? Br. J. Ophthalmol. 86, 363-365 (2002).

146.G. Romeo, W. H. Liu, V. Asnaghi, T. S. Kern, and M. Lorenzi, Activation of nuclear factor-kappaB induced by diabetes and high glucose regulates a proapoptotic program in retinal pericytes, Diabetes 51, 2241-2248. (2002).

147.Y. Du, M. A. Smith, C. M. Miller, and T. S. Kern, Diabetes-induced nitrative stress in the retina, and correction by aminoguanidine, J. Neurochem. 80, 771-779 (2002).

148.A. M. Joussen et al., Nonsteroidal anti-inflammatory drugs prevent early diabetic retinopathy via TNF-alpha suppression, Faseb J. 16, 438-440 (2002).

149.T. Abiko et al., Characterization of retinal leukostasis and hemodynamics in insulin resistance and diabetes: role of oxidants and protein kinase-C activation, Diabetes 52, 829-837 (2003).

150.K. Yamashiro et al., Platelets accumulate in the diabetic retinal vasculature following endothelial death and suppress blood-retinal barrier breakdown, Am. J. Pathol. 163, 253-259 (2003).

151.A. M. Joussen et al., Suppression of Fas-FasL-induced endothelial cell apoptosis prevents diabetic blood-retinal barrier breakdown in a model of streptozotocin-induced diabetes, Faseb J. 17, 76-78 (2003).

152.S. Mohr, Potential new strategies to prevent the development of diabetic retinopathy,

Expert Opin. Investig. Drugs 13, 189-198 (2004).

153.Y. Du, V. Sarthy, and T. Kern, Interaction between NO and COX pathways in retinal cells exposed to elevated glucose and retina of diabetic rats, Am. J. Physiol. 287, R735-R741 (2004).

Chapter 5

NEOVASCULARIZATION IN MODELS OF BRANCH RETINAL VEIN OCCLUSION

Ronald P. Danis, MD,1 and David P. Bingaman, PhD, DVM2

1Director of the Fundus Photograph Reading Center, Professor of Ophthalmology, Department of Ophthalmology and Visual Sciences, University of Wisconsin-Madison, and 2Assistant Director, Ocular Angiogenesis & Diabetic Retinopathy Programs, Retina Discovery Unit, Alcon Research, Ltd., Fort Worth, Texas

1.INTRODUCTION

Retinal ischemia is the primary cause of preretinal, optic nerve head, and iris neovascularization (NV) in human ocular disease. Causes of retinal vascular occlusion include diabetes mellitus, radiation, emboli, thrombosis, and inflammation. Occlusions, which can subsequently induce neovascularization, may affect either large or small caliber retinal vessels. Diabetic retinopathy and radiation retinopathy typically produce ischemia by affecting the microcirculation. Large vessel occlusions are identified by the primary site of obstruction: branch or central retinal vein occlusion, and branch or central retinal artery occlusion.

NV, as a complication of retinal ischemia, is highly prevalent in human retinal disease. Diabetic retinopathy (DR) is one of the most common causes of acquired blindness in developed nations, causing about 12% of cases of new blindness in the U.S. annually. Diabetes mellitus afflicts nearly

103

J.S. Penn (ed.), Retinal and Choroidal Angiogenesis, 103–117.

© Springer Science+Business Media B.V. 2008

14 million Americans.1 Approximately 5% of all diabetic patients develop ocular neovascularization. Branch retinal vein occlusion (BVO) is the second most common retinal vascular disease;2 about 50% of large BVO cases have significant ischemia, and of these about 40% will develop neovascularization.3 Central retinal vein occlusions are also common clinical problems; approximately 30% of these cases have severe ischemia, and of these 40 to 60% may also suffer neovascular complications.4

Retinal ischemia can stimulate pathological angiogenesis in multiple ocular tissues, such as the posterior segment (at the optic nerve head or growing out of the retina) or the anterior segment on the anterior surface of the iris. Posterior segment NV requires vitreous to serve as a collagen scaffold for neovascular growth. Importantly, eyes with the vitreous removed by surgical vitrectomy do not develop retinal NV, except where there is persistent vitreous. Posterior segment NV can penetrate the internal limiting membrane to develop along the posterior vitreous face or emanate into the vitreous gel. Posterior segment NV is also associated with a highly variable fibrous component. Typically, the early clinical appearance is that of “naked” vessels growing out of the retina or optic nerve into the vitreous. While these vessels may appear as simple vascular proliferation to the clinician, histopathology invariably demonstrates a fibrous component. With continued proliferation, the fibrous component tends to become more evident as whitish tissue accompanying the vessels. In most cases, untreated preretinal and optic nerve head NV evolves with a time course of months to years. The vessels gradually accumulate fibrous extravascular tissue until the vascular component eventually begins to atrophy and the lesion involutes. During involution, the fibrous component predominates, and the vessels may become grossly unapparent on clinical examination, although histopathology often shows some perfused vasculature.

Blinding complications occur as a consequence of the fibrous component of posterior segment NV. Optic nerve head NV may lead to bleeding into the vitreous cavity. Preretinal NV may also result in vitreous hemorrhage, but in addition, it may lead to a potentially more grave complication: tractional retinal detachment. Vitreous hemorrhage and retinal traction occur when the cellular component of the fibrovascular tissue contracts, causing rupture of the fragile new vessels and detachment of the retina. Vitreous hemorrhage, when mild, may clear spontaneously and cause only mild or transient visual impairment. Severe hemorrhage or tractional retinal detachment involving or threatening the macula can be blinding if left untreated and often requires surgical intervention.

Iris NV, also known as rubeosis iridis, most often develops first as a lacy configuration of vessels around the pupil, on the iris surface, or as small tufts at the pupillary sphincter. In more severe cases, the NV grows across the

entire iris surface and across the iridocorneal angle at the base of the iris. As in posterior segment NV, a fibrous tissue component of the NV proliferates with the vessels, and this eventually leads to contractile changes within the membrane. Contraction on the iris surface may cause abnormal enlargement of the pupil (anisocoria) and eversion of some of the posterior iris tissue through the pupil (ectropion uvea). The iridocorneal angle includes the trabecular meshwork, which is the major pathway for egress of aqueous humor from the globe. Contraction of a fibrovascular membrane may cause obstruction of the trabecular meshwork and closure of the angle, which frequently results in a very dire complication: neovascular glaucoma. Neovascular glaucoma is characterized by a very high intraocular pressure that may not be responsive to medication. Patients with neovascular glaucoma may experience severe pain, severe optic nerve damage, and retinal infarction. Surgical intervention may salvage the eye in some cases. Another mechanism whereby iris NV may cause pressure elevation and vision loss is through bleeding into the anterior segment, causing hyphema. The blockade of the iridocorneal angle by red blood cells can produce ocular hypertension.

Interestingly, the threshold for ischemia-induced injury to ocular tissues appears to vary with the underlying etiology. Iris NV is much more prevalent among cases of central retinal vein occlusion,4 whereas DR and BVO are more commonly associated with posterior segment NV. Both forms of intraocular NV can occur simultaneously in the same eye. Overall, iris NV is not observed as often as posterior segment NV in retinal clinical practice, because DR and BVO are more common diseases. Treatment of preretinal and optic nerve head NV with laser also inhibits iris NV, which limits its presentation in ischemic eyes that otherwise are at risk.

The mechanisms by which ischemia leads to ocular NV are detailed elsewhere in this text. Briefly, NV is caused by a complex interplay of factors, including hypoxia-regulated soluble growth factors (VEGF, HGF, IGF-1, PEDF, etc.), extracellular matrix components found in the vitreous and fibrovascular tissue, and the influence of immune cells. Macrophages invariably occur in specimens of pathological NV of all types. Their importance is generally acknowledged, since macrophages are a prominent source of angiogenic growth factors as well as chemo-attractants.

The potential for pharmacotherapeutic modulation of angiogenesis is the principal stimulus for the development of relevant animal models of ocular NV.

2.ANIMAL MODELS OF OCULAR NEOVASCULARIZATION FROM BVO

The study of BVO in animal models has been attractive to researchers for decades. It is technically possible to occlude retinal veins by laser photocoagulation, a relatively non-invasive methodology. Animals with eye sizes comparable to man, such as pigs and primates, can therefore be studied clinically and histopathologically to monitor the evolution of retinal ischemia and neovascular responses. The relatively large eyes of primates and pigs allow surgical implants or other drug delivery devices not accommodated by smaller rodent eyes. The study of NV can be approached advantageously through ocular models approximating the human condition.

Determining appropriate therapeutic targets is one of the challenges of studying posterior segment NV in animals. Another is the relative rarity with which this is created in laboratory animals. BVO models have been reported in mice, rats, rabbits, cats, dogs, various primate species, and pigs.5-12 Pigs provide the only BVO model where preretinal and optic nerve head (ONH) NV are consistently produced.13,14 The underlying causes of resistance of other species to the development of ischemia-induced retinal NV are unknown. However, inter-species (and even inter-strain) variations may affect the threshold for manifestation of ocular NV from the same stimulus.15

Another problematic feature of models of ocular NV due to BVO is the difficulty in quantifying the extent of NV. During efficacy studies, it is mandatory to have reproducible and standardized outcome measures. In BVO models, ordinal semi-quantitative grading systems have been established and applied in primate iris NV and pig retinal NV. However, these methods are dependent on subjective evaluation of photographic images or gross pathology with histopathological confirmation. More recently, histopathological assessment techniques have been employed to provide more reliable quantification.16

As noted, species differences in the neovascular proliferation following BVO are quite large. In monkeys, preretinal and optic nerve head NV are quite rare, whereas iris NV tends to be robust and quantifiable. Quantification of iris NV in the monkey model has relied exclusively on angiographic imaging.17-19 Iris angiography can be employed to determine the extent of the neovascular response in a masked manner. Since fundus features are not seen in the angiographic images, there is less potential for observer bias with this method.

3.PORCINE MODELS OF BVO

BVOs in pigs were produced by intense argon laser photocoagulation by Kohner and colleagues more than 30 years ago.6 These researchers demonstrated many of the ultrastructural and clinical findings seen in human patients with BVO; however, they did not observe pathological NV despite long term follow-up. The studies of experimental BVO in pigs from Moorfield’s group in the 1970’s did not describe NV despite careful histopathological evaluation.6 It is not clear why a group in a different laboratory was later able to produce a 100% incidence of preretinal and optic nerve head NV from a similar stimulus.14

Pournaras and colleagues first noted neovascular changes following argon laser-induced BVO in miniature pigs.13 They reported a high rate of vitreous hemorrhage with their laser technique. NV was demonstrated histopathologically in about 50% of their animals. Although Pournaras’ group has not employed the model to assay NV per se, they have used their model of BVO in pigs to define the degree of hypoxia achieved using intraocular oxygen sensors. They have studied hypoxic reactions of the tissues, soluble growth factor regulation, and laser effects in ischemic retinas.13,20

Danis and colleagues standardized this model to reliably produce quantifiable NV in domestic pigs.14 An innovation was the use of intravenous Rose Bengal dye as a photodynamic agent. Rose Bengal is a phthalocyanine dye with structural similarities to fluorescein; however, it fluoresces and produces oxygen free radicals from hydrolysis at 555 nm. This absorption peak is close to the 514-nm light produced by the green argon laser. Consequently, use of this dye allows retinal branch veins to be reliably closed with one treatment session with minimal vitreous hemorrhage (Figure 1). When 50% or more of the retinal venous territory was occluded in this manner, a 100% incidence of preretinal and optic nerve head NV was observed (Figure 2). The procedure was effective for producing ocular NV in both domestic and miniature pigs.21

Vascular occlusion in this pig model is produced by combined thermal necrosis and photodynamic thrombosis. Intense thermal burns are produced despite low laser power due to the long duration required to block vessel perfusion. The thermal necrosis appears necessary to produce permanent occlusions. When branch points of the retinal veins are targeted, a focal constriction of the vasculature can be exploited to advantage in producing the occlusion. Immediately after occlusion, retinal edema and intraretinal hemorrhage are produced as clinical indications of increased intravascular pressure. When multiple branch veins are occluded, retinal detachment is commonly produced. A vitritis often develops that may include fibrin

Figure 5-1. Immediate post-treatment fundus photograph after photodynamic thrombosis of a superior branch retinal vein of a pig eye.

Figure 5-2. (A) Fundus photograph of a porcine optic nerve head prior to BRVO. (B) Fundus photograph of the same porcine optic nerve head with neovascular tissue in the center of the optic nerve head.

membranes. After 3 to 4 weeks, retinal traction, schisis cavities, retinal venous collaterals, and retinal atrophy become apparent. NV invariably develops at the optic nerve head and is usually visible clinically with indirect ophthalmoscopy by 6 to 8 weeks. Clinical observations in miniature pigs up to 6 months after laser suggest that the neovascular development plateaus around 3 months after BVO. Spontaneous involution is not apparent, perhaps in part because NV in pigs often includes a heavy fibrous matrix which resembles, from the beginning, involutional NV in human diabetics.21

NV in the porcine BVO model may appear as preretinal fibrovascular tufts (as noted above), but also as “naked” vessels extending into the vitreous cavity with very little perivascular tissue (Figure 3). The endothelial cells of even the naked vessels appear to have relatively few fenestrations, which correlates with the lesser degree of fluorescein leakage compared to human preretinal NV. Additional differences between porcine BVO and human BVO histopathology include a pronounced inflammatory infiltrate into the vitreous in pigs. Macrophages are routinely found accompanying the neovascular tissue in humans as well, but the inflammatory infiltration around the NV and in the vitreous is more dramatic in the pig BVO model. The pig also often features full thickness retinal necrosis of the ischemic area, with inner retinal damage probably mediated by infarction from vascular occlusion and outer retinal damage perhaps due to the exudative retinal detachment that accompanies BVO in pigs. Because the porcine retina lacks a macula lutea (it possesses an area of increased cone photoreceptors toward the “area centralis”), the study of one of the most prevalent complications of BVO in humans, macular edema, is problematic in this model.

Quantifying the NV response in the porcine BVO model led to the development of a 5-step ordinal grading scheme. This scheme initially employed clinical grading of the fundus using stereoscopic color fundus photographs combined with histopathological confirmation of the NV.22 Later projects that involved intravitreal injections, which caused vitreous or lens opacity, underscored the weakness of relying on clinical grading and photography, and masked grading was then performed at the time of gross histology under a dissecting microscope.16 The confirmation of the clinical or gross histopathological grade with light microscopic study was necessary because avascular membranes are commonly produced in this model and are difficult to distinguish from neovascular membranes. In addition, the inner layer of schisis cavities may resemble neovascular vitreous membranes. Fluorescein angiography was thought to be unreliable in the detection of NV, because unlike in humans, preretinal NV in the pig does not always leak profusely.

Figure 5-3. Light photomicrograph demonstrating neovascularization of the retina (hematoxylin and PAS, original mag 40X.)

In the case of pig BVO, the gross histopathological grade is confirmed by

microscopy, since the clinical grade is sometimes considered unreliable.14,16,22,23 Unlike the situation in oxygen-induced retinopathy

models where histopathological features of preretinal NV are easily quantified from microscopic sections or whole mounts, pig BVO-induced retinal NV grows as a tuft into the vitreous. At present, this presentation of NV has not been adequately quantified from whole mounts (which distort and compress the NV) or from cross sections (which can confirm the presence of NV, but are not suitable for quantitation of an irregular endophytic mass). Since the fibrous proliferation that accompanies the intravitreal NV can mimic fibrinous nonvascular vitreous membranes and schisis cavities from vitreous traction, histopathology has been necessary to