Ординатура / Офтальмология / Английские материалы / Retinal and Choroidal Angiogenesis_Penn_2008

overall and can be associated with the later stages of NPDR or PDR. Today, no pharmacological therapy is approved for the treatment of PDR or DME. Panretinal or grid laser photocoagulation, respectively, and surgical interventions, such as vitrectomy and removal of preretinal membranes, are the only options available. Although panretinal photocoagulation (PRP) is relatively effective in preventing further deterioration, its mechanism of action is unknown. Similar to the exudative AMD treatements, laser photocoagulation in diabetic patients is a cytodestructive procedure, and the visual field of the treated eye is irreversibly compromised.

Emerging Pharmacologic Therapies for Exudative AMD

Figure 23-2. Emerging Pharmacologic Therapies for Exudative AMD.

An effective pharmacological therapy for pathological ocular angiogenesis and retinal edema would provide substantial benefit to an everincreasing segment of the population. Safe and effective treatments would have numerous benefits for the patient, such as avoiding invasive surgical or damaging laser procedures, improving quality of life, and prolonging work productivity. Moreover, societal costs associated with providing assistance and healthcare to the visually impaired could be dramatically reduced. Figure 2 provides a non-exhaustive list of the various clinical trials that are

currently being conducted in the area of ocular angiogenesis. The list and the following text describing these novel agents are simply a snapshot in time, more so than the other chapters in this book. The information related to the clinical outcomes changes very rapidly. Therefore, the following information is provided as a summary towards understanding the types of treatments that may one day ameliorate, or possibly even cure, these blinding diseases.

2.LASER-INDUCED INHIBITION OF PSNV

2.1Laser Photocoagulation for Exudative AMD and PDR

2.1.1Laser Photocoagulation in Exudative AMD

The presence of CNV under or near the fovea has visually devastating consequences and is the most common form of late stage AMD, where the prevalence of exudative AMD is nearly 2 times that of geographic atrophy.2 Moreover, because AMD is a bilateral disease, up to 26% of patients diagnosed with unilateral exudative AMD will develop CNV in the fellow eye within 5 years of follow-up.3 Although only 10-20% of all AMD patients will progress to the exudative late stage of the disease, patients with CNV represent 80% of all AMD patients with severe loss of visual acuity (20/200 or worse).4 Until the last 5-6 years, laser photocoagulation was the only clinically validated treatment for pathological CNV in patients with exudative AMD.

Laser photocoagulation therapy involves the selective heating of ocular tissues through the absorption of a specific wavelength of light by ocular pigments. Temperature increases between 10 and 20 o C provide enough thermal damage to denature proteins and other large molecules.5 The most common wavelength for laser photocoagulation of CNV lesions is within the green spectrum (488-514 nm).

Over roughly two decades, the Macular Photocoagulation Study Group (MPS) has conducted numerous prospective clinical trials demonstrating the utility of laser photocoagulation in reducing the risk of severe vision loss in patients with small, well-defined extrafoveal, juxtafoveal, and subfoveal CNV associated with exudative AMD.6-11 Unfortunately, 80-85% of CNV lesions evaluated by fluorescein angiography are not small enough and/or sufficiently demarcated to be eligible for treatment under MPS guidelines. Additionally, during the course of these trials a variety of issues were

identified, such as recurrent CNV development post-treatment and immediate, permanent central vision loss in patients with subfoveal CNV.12,8 Laser photocoagulation, nonetheless, has been a major advancement in the treatment of exudative AMD versus observation alone.

2.1.2Photodynamic Therapy (PDT)

Photodynamic therapy (PDT) is a relatively selective and localized treatment for CNV that is based on the oxidation of biological tissues by a photodynamic reaction.13 An intravenous photosensitizing dye is first delivered to the target tissue via the systemic circulation and then locally excited using a specific wavelength of light delivered via an ophthalmic laser. The photosensitizer alone generally does not cause cellular damage; however, when activated by laser light, electrons are released through a photochemical reaction and reactive oxidative species are generated within the target vasculature.14 The reactive oxidative species subsequently react with the cell membranes of the endothelium and blood constituents, producing platelet activation and thrombosis. If maintained, this microvascular thromobosis can lead to reduced vascular permeability and eventual involution of the lesion.15 As the first step beyond standard laser photocoagulation, PDT for exudative AMD laid the foundation for the advent of pharmacological inhibition of ocular angiogenesis.

Verteporfin (Visudyne®, Novartis AG) was the first ocular photodynamic therapy to be approved in the United States. In 2000, it was released for the treatment of patients with predominantly classic subfoveal CNV secondary to exudative AMD.16 Verteporfin is a modified benzoporphyrin dye that can be activated intravascularly by low-intensity, non-thermal diode laser light (689 nm). The marketed product is a lipid emulsion delivered by intravenous infusion over 10 minutes at 6mg/mm2 of body surface area. It achieves peak plasma levels by the end of the infusion and has been shown to rapidly and selectively accumulate on low-density lipoprotein receptors that are highly expressed on proliferating choroidal endothelial cells.17,18 Verteporfin is activated at 15 minutes following completion of the infusion. In an important step forward in AMD therapy at the time, the Treatment of Age-Related Macular Degeneration with Photodynamic Therapy (TAP) study demonstrated that PDT with verteporfin can safely and effectively occlude leaking choroidal vessels, retard CNV lesion growth, and slow the progression of vision loss in patients with predominantly classic (welldefined margins during fluorescein angiography) subfoveal CNV lesions at 12 and 24 months. However, PDT with verteporfin did not restore lost vision and required repeat treatment at approximately 3-month intervals in most patients.19,20 In addition, this therapy has provided a treatment benefit in

patients with subfoveal occult CNV lesions (poorly demarcated lesions) with no classic subfoveal component after 24 months of therapy, but not at 12 months, according to data from the AMD Verteporfin in Photodynamic therapy (VIP-AMD) trial.21 In the VIP trial, PDT was superior to placebo at preserving vision and reducing the risk of severe vision loss (>6 lines). This led to a follow-up study assessing Visudyne® therapy for patients with occult CNV and no classic component, i.e., the Visudyne® In Occult AMD (VIO) trial. However, an initial report from the company related to 2-year data indicates that the primary outcome measure was still undetermined and that subgroup analyses were being conducted. Nonetheless, a subsequent retrospective observational cohort study comparing Visudyne® therapy in patients with predominantly classic versus occult lesions found no clinically relevant difference in mean visual acuity between the 2 treatment groups.22

A variety of other clinical trials have been conducted with verteporfin. Two-year follow-up data from the Visudyne in Minimally Classic Trial (VIM) revealed that fewer verteporfin-treated eyes, as compared to placebo, had severe vision loss (>3 lines on a standard Snellen visual acuity chart) or

converted to a predominantly classic lesion in the reduced fluence PDT group.23,24 Other studies such as VER (the Verteporfin Early Retreatment

trial)25 and VALIO (the Verteporfin with Altered (delayed) Light I Occult trial)26 have been conducted. To date, verteporfin PDT has been approved in over 50 countries for extended indications, including occult CNV, pathological myopia, and presumed ocular histoplasmosis syndrome.27 Guidelines for using verteporfin PDT to treat patients with CNV related to various etiologies have been published.28

Verteporfin therapy has been generally safe and well tolerated.29 Systemic adverse events appear to be transient and mild to moderate, including injection site reactions, photosensitivity reactions, and infusionrelated back pain.30 In addition, PDT can damage adjacent normal tissue containing the photosensitizer.31 Notably, acute, severe visual acuity loss occurred in 0.7% and 4.9% of treated patients in the TAP and VIP trials, respectively. Although this type of acute visual loss is uncommon with PDT, additional treatment may be necessary to stabilize visual acuity.32

Although PDT with verteporfin initially showed promise as a monotherapy in treating several types of CNV, several factors have led to its falling out of favor as a first-line or sole treatment for exudative AMD. Perhaps most importantly, PDT typically only slows vision loss rather than achieving an improvement. Final visual results are, on average, in the 20/200 range and the treatment generally needs to be repeated 2-4 times per year33-37 because verteporfin often mediates only transient damage to the CNV.38 In fact, studies have demonstrated that immediately following PDT, vascular permeability is actually enhanced and VEGF expression is elevated.39,40

Therefore, most retina specialists who continue to use verteporfin PDT employ it along with other pharmacological agents, such as intravitreal or subtenon triamcinolone acetonide (Kenalog®, Bristol-Myers Squibb), intravitreal ranibizumab (Lucentis®, Genentech), or bevacizumab (Avastin®, Genentech).41 Published results when using PDT alone versus combination therapy with these agents appear to indicate that combination therapy is a superior treatment paradigm.41-46

2.1.3Feeder Vessel Photocoagulation

Feeder vessel photocoagulation (FVP) treatment is another potential laser therapy for CNV that is based on the hypothesis that a small number of intrachoroidal “feeder vessels” supply the entire CNV complex. If true, closure of these feeder vessels should result in changes in flow characteristics within the CNV complex and subsequent starvation of lesion growth and maintenance. The technique requires high-speed indocyanine green (ICG) angiography to detect the feeder vessels located at a distance from the subfoveal CNV and then delivery of laser energy to achieve vessel

closure.47 Several independent reports suggested initial successful clinical outcomes in pilot studies.48,49 The main advantage touted for FVP is that the

photocoagulated area is small and remote from the CNV, thereby completely avoiding the fovea. However, the major drawbacks are that (1) feeder vessels can be correctly identified in only 22–42% of patients with macular CNV,50

(2) feeder vessel visualization is indirect, therefore the ability to accurately aim the laser beam is limited, and (3) repeat treatment may be required because of reperfusion. There have been no large, prospective, controlled, randomized clinical trials to fully evaluate this novel technique. However, combined treatment of FVP with PDT has shown that pretreatment with PDT can increase the rate of feeder vessel identification,48 which could be used as an alternative approach to persistent or recurrent CNV.

2.2Panretinal Photocoagulation for PDR

Panretinal photocoagulation (PRP) involves placing multifocal laser burns in the peripheral retina, sparing the macula, and it has been shown in

randomized clinical trials to reduce the risk of vision loss in the majority of patients with PDR.7,51-53 PRP is indicated in patients with PDR that exhibit

well-established preretinal NV or NV at the disc, where the NV may be associated with hemorrhage and generally some visual acuity may already have been compromised.54,55 Complications of the procedure, although infrequent, include loss of peripheral, night, or color vision.7 Along with PRP, vitrectomy is often performed in patients with severe PDR involving

nonclearing vitreous hemorrhage and/or tractional retinal detachment. Vitrectomy has been shown to prevent severe vision loss, be cost-effective, and improve quality of life.56-58

The mechanism through which PRP inhibits further development of retinal angiogenesis is still incompletely understood. PRP also has been termed “retinal ablation,” indicating that the laser is directly destroying tissue responsible for producing the pro-angiogenic stimuli. Other hypotheses suggest that PRP reduces the sum metabolic demand of the diabetic retina and thereby diminishes hypoxic ischemia, or that laser damage actually increases the production of endogenous anti-angiogenic molecules, such as TGFβ or PEDF.59-64 Consistent with these concepts, PRP has been shown to reduce VEGF levels in the vitreous, aqueous (also reported a decrease in HGF levels), and plasma.65-67 Regardless of the mechanism of action, PRP treatment necessitates permanent tissue destruction to elicit an efficacious response.

3.PHARMACOLOGICAL INHIBITION OF PSNV

3.1Inhibitors of VEGF-mediated Ocular Angiogenesis

The VEGFs (VEGF-A, -B, -C, -D, -E and placental growth factor [PlGF]),

are a family of homodimeric glycoproteins that bind with varying affinities to VEGFR1-3.68,69 VEGF (or VEGF-A) and its requisite tyrosine kinase

receptors (or RTKs), VEGFR1 and VEGFR2, contribute to vascular morphogenesis and neovascular pathology predominantly through two mechanisms: (1) new vessel growth (vasculogenesis and/or angiogenesis) and (2) vascular permeability.70-74 In the eye, VEGF is a critical factor during retinal vascular development.75 Moreover, ocular tissues respond to a variety of stimuli, such as hypoxia or inflammation, by the induction of VEGF resulting in blood-retina barrier breakdown (i.e., enhanced vascular permeability) and PSNV.76,77 Animal models have been used to demonstrate the critical role of VEGF signaling in ocular disease. Early work demonstrated that intravitreal injection of soluble VEGFR chimeric proteins suppressed retinal NV in the mouse model of oxygen-induced retinopathy (OIR).78 When using a non-human primate model of retinal ischemia induced by branch vein occlusion, VEGF was shown to be spatially and temporally correlated with the ocular NV.79,80 Moreover, intravitreal injection of VEGF produced retinal ischemia and micro-angiopathy in the same non-human primate species.80 Similar to the earlier results using soluble VEGFRs, a neutralizing anti-VEGF monoclonal antibody injected

intravitreally inhibited the NV displayed in a primate model and supported the premise that anti-VEGF therapies may have promise for human ocular disease.81

As discussed in previous chapters, OIR models produce preretinal NV similar to that found in human ocular diseases such as retinopathy of prematurity (ROP) and PDR and are widely used as screening assays for pharmacological efficacy studies. In rodent OIR models, retinal VEGF levels are correlated with the incidence and severity of pathology, and intravitreal

injection of VEGFR inhibitors significantly inhibits preretinal NV formation.82,83 The rodent and primate models of laser-induced CNV are

commonly used experimental surrogates for exudative AMD and have been shown to be VEGF-dependent.84-86

VEGF, VEGFR1, and VEGFR2 have been localized in ocular fluids and neovascular membranes obtained from patients with DR and exudative AMD, and are associated with increased severity of disease.65,87-90 Successful results from clinical trials using various molecules to block aspects of VEGF signaling have validated the concept first identified in the animal models.91-94 The strategies used will be discussed in detail below; however, several important questions regarding VEGF inhibition still remain unanswered:

(1)Which VEGF isoform(s) is predominantly responsible for pathological ocular angiogenesis in humans? Recent preclinical evidence

suggests that the VEGF165 isoform may be a primary mediator of ocular disease.95,96 However, published clinical trial results from intravitreal injection of an anti-VEGF antibody fragment (Lucentis®) that binds all

soluble human isoforms suggest that enhanced efficacy is achieved when addressing multiple VEGF isoforms.97

(2)Will chronic blockade of ligand-receptor interaction, as occurs with molecules that act as VEGF “sponges,” induce overexpression of VEGF receptors, leading to the potential for a rebound effect? Early evidence from clinical trials and preclinical studies using VEGF ligand antagonists seem to

suggest that this phenomenon could be a reality in some patients following cessation of therapy.98,99

(3)Does VEGF signaling play a role in homeostasis of the neural retina, and could chronic VEGF inhibition lead to retinal degeneration? Preliminary

evidence in animal models suggests a role for VEGF in both neurogenesis and neuroprotection.100,101 To date, this potential side effect has not been

observed in the early clinical results from the anti-VEGF trials; however, the treatment durations reported may be too short to assess this aspect.

3.1.1Macugen®

Pegaptanib sodium 0.3 mg (EYE001, Macugen®; Eyetech/Pfizer) was approved by the FDA in late 2004 for the treatment of patients with all forms of subfoveal neovascular (wet) AMD, becoming the first pharmacological agent to be approved for ocular angiogenesis.102 Pegaptanib is a 28-base ribonucleic acid aptamer identified via a chemical screening process

(systematic evolution of ligands by exponential enrichment, SELEX), and is the first aptamer to be approved for human use.103,91 Through its unique three-

dimensional structure, the aptamer selectively binds to the major soluble human VEGF isoform, VEGF165, essentially acting as a high-tech sponge for extracellular VEGF.92 The molecule is pegylated, i.e., covalently attached to two 20-kDa polyethylene glycol (PEG) moieties, to increase residency time within the eye following intravitreal injection. Moreover, it has a modified sugar backbone that resists degradation by endoand exonucleases.92,91

Phase III results from 2 randomized, double-blind, multicenter, doseranging, controlled clinical trials, representing a total of 1186 patients with various forms of wet AMD, demonstrated that after one year of intravitreal injections every 6 weeks, patients treated with pegaptanib exhibited a statistically significant reduction in risk of visual acuity loss from baseline up to 54 weeks.91 More specifically, intravitreal injection of pegaptanib provided a significant reduction in “moderate” and “severe” vision loss (loss of 15 letters or more and loss of 30 letters or more of visual acuity, respectively).91 It has been noted that the level of risk reduction and the percentage of patients (nearly 10%) that demonstrate an improvement in vision are relatively similar to that observed with photodynamic therapy.104 Although patients were randomized to receive control (sham injection), 0.3 mg, 1 mg, or 3 mg pegaptanib, dose levels above the approved 0.3 mg did not demonstrate additional efficacy.91,105 Pegaptanib therapy was less effective during the second year of treatment, and efficacy and/or safety beyond 2 years is unknown.105 Because of the use of fluorescein angiography for detecting the CNV lesion growth, the mechanism of action remains unclear as to true anti-angiogenic effects versus inhibition of retinal vascular permeability.91 The most significant adverse events were endophthalmitis, traumatic lens injury, and retinal detachment, and the company reports that these events most likely were attributed to the injection procedure and not to the study drug.91 The risk of serious endophthalmitis and subsequent vision loss following repeated intravitreal injections is a concern, although the risk per injection was shown to be very small in the phase III study (0.16%), once the injection procedure was modified related to local antimicrobial therapy. Intravitreal pegaptanib also is being evaluated in phase III clinical trials for the treatment of patients with DME.

3.1.2Lucentis®

Intravitreal Lucentis® (0.5 mg ranibizumab, rhuFab2; Genentech/Novartis

Ophthalmics) was approved in June 2006 and has revolutionized the treatment of exudative AMD.106,107 A high level of excitement has been

associated with the release of ranibizumab and its demonstrated ability to improve visual acuity in roughly 1/3 of treated patients.108 Ranibizumab is a 48-kDa humanized monoclonal antibody fragment that binds to all isoforms

of VEGF and is delivered by intravitreal injection in patients with exudative AMD and DME.46,93,106,109-113 This affinity-matured Fab (MB1.6 variant)

functions similar to the 139-kDa, humanized full-length anti-VEGF

monoclonal antibody, Avastin® (Fab-12 variant), that was the first pharmacological anti-angiogenic therapy approved for human use.114,115,107 In

1993, Ferrara et al. at Genentech were the first to demonstrate that inhibition of soluble VEGF produced by tumor cells could suppress tumor growth in mice.116 Humanization of the mouse anti-VEGF monoclonal antibody then provided the ability to test this treatment strategy in man.117 Although antiVEGF therapy was an obvious target for multiple ocular diseases, early pharmacokinetic work in monkeys demonstrated that a full monoclonal antibody, trastuzumab (148 kDa), may not provide adequate tissue distribution to the retina and choroid following a single intravitreal injection.118 The smaller ranibizumab, in contrast, was shown to penetrate the posterior segment following intravitreal administration. Further preclinical studies have shown that ranibizumab has a 3-day terminal halflife in monkey eyes following a single intravitreal injection (500 or 2000 μg/eye).109 Intravitreal administration of ranibizumab prevented the development of laser-induced CNV in monkeys and decreased fluorescein leakage from already-formed CNV.86 During this preclinical study, all treated eyes exhibited acute, self-limiting anterior chamber inflammation that seemed to diminish with repeated injections.

Genentech (www.gene.com) has completed multiple clinical trials using Lucentis® as a sole therapy or in combination with Visudyne® treatment in patients with wet AMD, with several additional trials remaining open. An open-label, dose-ranging study, using 150–2000 μg ranibizumab administered via a single intravitreal injection in 27 patients, demonstrated that 500 μg was the maximum tolerated dose.30 At higher doses, eyes injected with ranibizumab exhibited significant intraocular inflammation that was self-limiting and without infectious endophthalmitis. A phase I/II multidose study also demonstrated that repeated intravitreal injections of 0.3 or 0.5 mg ranibizumab over 6 months provided an acceptable safety profile and improved visual acuity with decreased fluorescein leakage from CNV in patients with exudative AMD.119 Moreover, a multiple escalating dose study

in 32 patients with primary or recurrent subfoveal CNV showed that multiple intravitreal injections of ranibizumab (0.3–2.0 mg) were well-tolerated, and

median and mean visual acuity improved in all study groups up to 20 weeks.112

Clinical results released from two phase III Lucentis® trials (MARINA and FOCUS) demonstrate that the anti-VEGF therapy has the potential to improve, not just stabilize, vision in wet AMD patients. In the phase III MARINA (Minimally classic/occult trial of the Anti-VEGF antibody Ranibizumab In the treatment of Neovascular AMD) study, 716 patients with minimally classic or occult wet AMD were randomized 2:1 to receive intravitreal injections of ranibizumab (0.3 or 0.5 mg) or sham injections every 28 days for two years. At 12 months, when addressing the primary endpoint of maintaining visual acuity, 95% of the patients injected with 0.3

or 0.5 mg ranibizumab lost <15 letters (ETDRS eye chart) compared to baseline versus 62% of the sham-treated patients.120,106 Patients treated with

0.3 or 0.5 mg ranibizumab gained an average of 6.5 and 7.2 letters of visual acuity, respectively, compared to baseline, whereas the sham-treated group lost an average of 10.4 letters. Importantly, the benefit in visual acuity in the ranibizumab-treated patients was maintained at 24 months. Serious ocular adverse events that ocurred more frequently in ranibizumab-injected eyes were uveitis (1.3%) and presumed endophthalmitis (1%).

In the phase III ANCHOR (Anti-VEGF Antibody for the treatment of Predominantly Classic Choroidal Neovascularization in AMD) study, 423 patients with predominantly classic subfoveal wet AMD were randomized 1:1:1 to receive PDT followed by a sham injection or an intravitreal injection of either 0.3 or 0.5 mg ranibizumab followed by a sham PDT treatment for 24 months.111 Data at 12 months show that the primary endpoint of maintaining visual acuity was met and that 40% of patients treated with 0.5 mg ranibizumab, versus 5.6% of controls, had improved vision by 15 letters or more as compared to baseline. Serious ocular adverse events reported in this study also were uveitis (0.7%) and presumed endophthalmitis (1.4%).

Numerous other studies are ongoing in wet AMD, e.g., HORIZON, a phase III open-label extension study allowing patients exiting the above trials to continue to receive the investigational therapy; PrONTO (Prospective Optical Coherence Tomography Imaging of Patients with Neovascular AMD Treated with Intra-Ocular Lucentis®), a prospective, open-label, uncontrolled study designed to evaluate the effectiveness of a reduced number of treatments; and PIER (A Phase IIIb, multicenter, randomized, double-masked, sham injection-controlled study of the efficacy and safety of ranibizumab in subjects with subfoveal choroidal NV with or without classic CNV secondary to AMD), involving a fixed treatment regime of 3 initial monthly injections followed by quarterly injection