Therapy for Neovascular Age-Related Macular Degeneration
STEPHAN MICHELS, MD, PHILIP J. ROSENFELD, MD, PhD
VASCULAR ENDOTHELIAL GROWTH FACTOR AND AGE-RELATED MACULAR DEGENERATION
Vascular endothelial growth factor (VEGF) induces vascular endothelial cell proliferation, maintains new blood vessels by promoting vascular endothelial cell survival, increases vascular permeability, and acts as a macrophage chemotactic factor.1-5 This multifunctional protein has been implicated as a major growth factor in ocular neovascularization and macular edema due to its angiogenic and vascular permeability properties. Evidence supporting the role of VEGF in neovascular age-related macular degeneration (AMD) includes the upregulation of VEGF in the retinal pigment epithelium (RPE) and the retinal outer nuclear layer in association with choroidal neovascularization (CNV).6 Experimental evidence in support of VEGF includes the over-expression of VEGF in different mouse models, which results in deep retinal neovascularization or choroidal neovascularization.7-10 However, damage to Bruch's membrane is necessary to induce the growth of CNV from the choriocapillaris into the subretinal space.9 These studies suggest that the VEGF is associated with CNV in situ, and overexpression of VEGF could be sufficient for the induction of neovascularization in both the retina (retinal angiomatous proliferation) and choroid in AMD.
VEGF is a dimeric 36-46 kd glycosylated protein with an N-terminal signal sequence and a heparin-binding domain. In humans, 5 different VEGF isoforms are known with varying numbers of amino acids due to alternative splicing of the VEGF mRNA (eg, VEGF206, VEGF189, VEGF165, VEGF145, and VEGF121).11, 12 VEGF165 is the predominant isoform of VEGF, with the other isoforms present in smaller amounts. Both VEGF206 and VEGF189 display the heparin-binding characteristic attributed to VEGF, and these isoforms are believed to be the nondiffusable or matrix-bound forms of VEGF. In contrast, VEGF145 and VEGF121 do not bind heparin and may represent the diffusable form of VEGF. VEGF165 has intermediate properties and is thought to be both matrix-bound and diffusable. An additional diffusable form of VEGF, known as VEGF110, has been identified as a proteolytic breakdown product of VEGF165 and has been shown to be functionally active. Plasmin, a ubiquitous protease, is capable of hydrolyzing VEGF165 to VEGF110,13, 14 and VEGF110 may have an important role in the growth of CNV.
VEGF INHIBITION AS A TREATMENT STRATEGY FOR AMD
One possible approach to prevent or inhibit the growth of CNV would be to block the action of VEGF. The first successful report of VEGF inhibition resulting in a biologically meaningful response was in 1993 when a murine anti-VEGF monoclonal antibody was used to inhibit VEGF-mediated tumor growth in vivo.15 Subsequently, a humanized monoclonal anti-VEGF antibody known as bevacizumab (Avastin) was developed by Genentech, Inc.16 This full-length antibody binds all forms of VEGF with high affinity (kd ~8x10_� M), including the biologically active proteolytic fragment known as VEGF 110. Bevacizumab used in conjunction with chemotherapy has been approved by the US Food and Drug Administration (FDA) for the systemic treatment of metastatic colorectal carcinoma. Although using systemic bevacizumab for the treatment of CNV could theoretically succeed, Genentech proceeded with an alternative strategy, an intravitreal injection as the preferred mode of drug delivery.
Local drug delivery has certain theoretical advantages, as opposed to systemic delivery, the most obvious being that a high concentration of drug can be delivered to the diseased tissue with less risk of systemic side effects. For the successful treatment of CNV using a pars plana injection, the drug must be capable of penetrating the internal limiting membrane, the full thickness of the retina, and entering the subretinal and even the sub-RPE space to bind VEGF. The size of a full-length antibody (148 kd) and its inability to penetrate the inner retina precluded the use of bevacizumab for the local treatment of CNV,17 which led to the development of ranibizumab, previously known as rhuFab V2, (Lucentis) for the treatment of CNV secondary to AMD.
Molecular and Pharmacologic Properties
Ranibizumab is a recombinant humanized antibody fragment (Fab) derived from one of the antigen-binding arms of the full-length anti-VEGF monoclonal antibody bevacizumab. Compared with the molecular weight of bevacizumab, ranibizumab has been reduced from 148 kd to 48 kd. Six mutations were introduced and then selected by affinity maturation, resulting in a variant known as Y0317 that can bind all the VEGF forms with a 120- to 140-fold higher affinity compared with the original binding fragment.18 Because of its smaller size, ranibizumab is better suited to intravitreal drug administration. In rhesus monkeys, an intravitreal Fab is capable of penetrating the full thickness of the retina, resulting in high retinal drug concentrations and reaching the retinal pigment epithelium within 1 hour and remaining there for up to 7 days. Intravitreal drug delivery was found to result in nondetectable serum levels of ranibizumab, thus supporting the theoretical advantage of local vs systemic drug delivery.17 In addition, animal studies have demonstrated a rapid clearance of intravenous ranibizumab with a serum half-life time of 3.05 hours.19 Following intravitreal delivery, the half-life of ranibizumab was found to be 2.9 to 3.2 days in rabbits with detectable levels observed out to 11 days based on fluorophotometric and vitreous sampling studies.17, 20-22
The safety of intravitreal ranibizumab was evaluated in cynomolgus monkey eyes at doses of 250 �g, 750 �g, and 2000 �g. Retreatment was performed every 2 weeks for 3 months. Ranibizumab induced a dose-related inflammatory reaction within the anterior chamber that was most severe after the first injection and became attenuated with subsequent injections. Most eyes showed no to mild anterior chamber reaction 7 days after treatment. Retinal perivascular infiltrates composed of lymphocytes, macrophages, and plasma cells were seen in 7 out of 28 treated eyes. These lesions diminished during the recovery, and no alterations in electroretinography testing or visually evoked potentials were detected, and no vascular leakage was detected by fluorescein angiography (FA). Fifteen animals developed antibodies to ranibizumab, and these antibodies were directed against the humanized backbone of the antibody.23
The safety and efficacy of 500 �g intravitreal ranibizumab given at 2-week intervals were evaluated in a study using the laser-induced CNV monkey model.24 As in the previous safety study, all eyes developed an anterior chamber reaction within 24 hours of the first intravitreal injection. Inflammation resolved within 7 days, and subsequent injections produced less inflammation. Repeated injections every 2 weeks starting 3 weeks before laser injury prevented the formation of clinically significant CNV. The study also suggested a beneficial effect in treating established CNV. These promising experimental results with regard to safety and efficacy led to clinical trials in patients with neovascular AMD.
Figure 1. Treatment protocol for phase I/II study FVF2128g: 300 �g or 500 �g ranibizumab vs usual care.
The first clinical study, designated FVF1770g, was a phase I single-injection, dose-ranging investigation to identify the maximum tolerated dose of ranibizumab In this study, 27 patients received a single intravitreal injection of ranibizumab ranging from 50 �g to 1000 �g. Dose-limiting toxicity was predetermined to be 2 or more patients developing a 2+ or greater inflammatory response. This dose-limiting toxic response was observed at the 1000 �g dose level, and all inflammation was self limited with no associated sequelae. As a result, the 500 �g dose was identified as the maximum tolerated dose. Doses of 300 �g and 500 �g were then used for the next investigation exploring multiple intravitreal injections of ranibizumab.25
The phase I/II clinical trial, designated FVF 2128g, explored the safety and tolerability of 4 to 8 intravitreal injections of ranibizumab (Figure I).26, 27 In this open-label, randomized, controlled trial, 64 patients were enrolled in 2 treatment groups (300 �g or 500 �g every 28 days) or to a usual care group. Usual care was defined as verteporfin photodynamic therapy (PDT) or observation. All usual care subjects were given the option of crossing over to receive ranibizumab at day 98 of the study. Inclusion criteria included eyes with classic-containing CNV as defined by FA, and prior verteporfin therapy was permitted. Twenty-five subjects received the 300 �g dose, and once that dose was found to be safe through day 98, an additional 28 subjects received the 500 �g dose. Initially, 11 subjects received usual care.
Overall, the CNV was classified as minimally classic in 39% of the subjects, predominantly classic in 33%, and classic CNV after verteporfin therapy in 28%. The most common adverse event was a transient, reversible inflammation of grade 2+ or greater in 26% of treated subjects during the first 3 months of the study. Of the 64 patients receiving intravitreal injections in this series, 3 patients had a serious ocular adverse event. There was one case each of endophthalmitis, recurrent uveitis, and a central retinal vein occlusion. All 3 events resolved with recovery of visual acuity to preevent levels or better. At follow-up day 98, both treatment groups showed a mean gain of visual acuity with 12.6 letters in the 500 �g group and 7.35 letters in the 300 �g group. In contrast, subjects enrolled in the usual care group lost 5.1 letters at day 98. The positive trend continued up through day 210 with a mean gain of visual acuity compared with a baseline of 15 letters in the 500 �g group and 12.8 letters in the 300 �g group. Overall 97.5% of all treated patients were improved or stable (�15 letters) at day 210, with 45% of patients gaining 3 or more lines in ETDRS visual acuity, 52.5% gaining at least 2 lines, 75% gaining at least 1 line, and 85% gaining any visual acuity. 28 Only ~2.5% of patients lost at least 3 lines of visual acuity by day 210. Short-term results on central retinal thickness in optical coherence tomography were reported for a subgroup of patients in this trial receiving either 300 �g or 500 �g of intravitreal ranibizumab. The average central retinal thickness at baseline was 300 microns �92 (SD). After multiple injections, central retinal thickness decreased to 262 microns �164 by day 28 and 201 microns �33 by day 84.29
Figure 2. Treatment protocol for phase I/II
study FVF 2425g: ranibizumab dose escalation strategy.
In another phase I/II study, designated FVF2425g, 3 different dose-escalating regimens were explored to determine if a dose higher than 500 �g could be safely administered to 30 patients (Figure 2). The goal was to determine whether these higher doses, if safe, had more apparent efficacy compared with the lower doses. The doses started at 300 �g and ranged up to 2000 �g. Ranibizumab was given intravitreally as frequently as ever 2 weeks and was well tolerated. All patients received 16 weeks of treatment and were followed through 20 weeks. Only 3 subjects experienced a grade 2+ or higher intraocular inflammation after the first or second injection, and only one patient showed a grade 2+ intraocular inflammation at a dose higher than 500 �g. None of the patients demonstrated inflammatory retinal infiltrates as seen in the animal experiments. By day 140, mean visual acuity improved 13.6 letters in 9 patients from Group 1, 11.9 letters in 9 patients in Group 2, and 5.2 letters in 9 patients of Group 3. Overall, 44% of patients had an improvement of at least 15 letters, 48% had stable visual acuity, and 7% had at least a 15-letter decrease in visual acuity compared with baseline. No serious ocular adverse events, in particular endophthalmitis, were identified. Results appeared similar for patients retreated every 2 or 4 weeks. The study concluded that frequent and higher doses of ranibizumab are well tolerated and suggested a beneficial effect on visual acuity.30, 31
Figure 3. Treatment protocol for FOCUS study.
NEW CLINICAL TRIALS
The phase I/II studies established that ranibizumab therapy appeared to be safe, well tolerated, and beneficial. As a result of these positive outcomes, 3 additional registration studies are underway. A phase II study, called FOCUS, is designed to evaluate AMD patients with predominantly classic subfoveal CNV (Figure 3). All patients will receive verteporfin therapy in addition to 13 monthly injections of ranibizumab (500 �g) or 13 monthly sham injections. Verteporfin therapy will be given 1 week prior to the intravitreal injection. Patients are randomized 2:1 to receive active treatment or sham. Long-term followup, at 6 and 12 months after the final visit (week 52), is planned.
In a phase III trial known as MARINA, AMD patients with subfoveal minimally classic or occult-only CNV are randomized (1:1:1) to receive 24 monthly treatments of either 300 �g or 500 �g ranibizumab or a sham injection (Figure 4). In another phase III trial called ANCHOR, AMD patients with predominantly classic CNV are being randomized 1:1:1 to receive 24-monthly intravitreal injections of 300 �g or 500 �g of ranibizumab or verteporfin therapy (Figure 4). Depending on their randomization, patients will receive either a ranibizumab injection or a sham injection every month and will then be evaluated by FA every 3 months to determine if there is leakage from CNV. They will then receive either a verteporfin infusion or a placebo infusion with standard PDT laser light exposure. The primary endpoint for all 3 studies, as determined by the FDA, is the proportion of patients losing at least 15 letters of visual acuity.
Figure 4. Study design for phase III studies
ANCHOR and MARINA.
From a retina specialist's perspective, a major disadvantage of all 3 registration studies is that patients are injected with ranibizumab every month for 2 years regardless whether leakage from CNV is detected. Once ranibizumab is approved, it seems unlikely that we will treat our patients using this approach. Most likely, we will inject our patients monthly until no leakage from CNV is detected, and then we will withhold additional injections until we detect recurrent leakage from CNV. At the Bascom Palmer Eye Institute we have limited experience, treating 23 patients using this approach after they completed their injection portion of the phase I/II studies and who were then enrolled in an extension study. To date, we have followed these patients for >18 months, and by using this intermittent injection strategy, we have been able to maintain their improved visual acuity and prevent lesion growth. To investigate this treatment strategy further, we have obtained approval from Genentech and the FDA to conduct a 2-year, investigator-sponsored trial called the PrONTO Study (Prospective Optical Coherence Tomography (OCT) Imaging of Patients with Neovascular Age-Related Macular Degeneration (AMD) Treated with Intra-Ocular Lucentis). This open-label, nonrandomized clinical study is currently underway at the Bascom Palmer Eye Institute, and we are enrolling AMD patients with all major lesion types of subfoveal CNV.
Anti-VEGF therapy using ranibizumab is a promising new treatment that may provide neovascular AMD patients with an opportunity for visual acuity improvement. In our experience, intravitreal drug delivery through the pars plana can be a safe, low risk procedure with endophthalmitis as a rare complication, provided simple precautions are implemented such as the use of topical 5% betadine and a sterile lid speculum. Fortunately, the widespread use of intravitreal triamcinolone has prepared the retina specialist for intravitreal injections in the outpatient setting and the coming age of intravitreal anti-VEGF therapy. Unless unexpected long-term complications arise from chronic ranibizumab therapy, the ongoing clinical trials should show that ranibizumab therapy can stabilize and improve vision in the majority of patients with neovascular AMD.
From the Bascom Palmer Eye Institute, University of Miami School of Medicine, Miami, Fla. Dr. Rosenfeld has received research support from Genentech, Inc. to perform clinical trials using ranibizumab and support to attend scientific meetings to present clinical trial results. He has also served on Genentech advisory boards. Additional support has been provided by an unrestricted grant from Research to Prevent Blindness, Inc. and NIH center grant P30 EY14801. Dr. Michels is a recipient of a research grant from the German Research Foundation (DFG). As of September 2004, Dr. Michels will be at the University Eye Hospital Vienna, Austria, W�hringer G�rtel 18-20, 1090 Vienna, Austria.
Address correspondence to: Philip J. Rosenfeld, MD, PhD, Bascom Palmer Eye Institute, 900 N.W. 17th St., Miami, FL 33136. Telephone: (305) 326-6148. Fax: (305) 326-6417. E-mail: email@example.com.
1. Feletou M, Staczek J, Duhault J. Vascular endothelial growth factor and the in vivo increase in plasma extravasation in the hamster cheek pouch. Br J Pharmacol. 2001; 132(6):1342-8.
2. Tilton RG, Chang KC, LeJeune WS, et al. Role for nitric oxide in the hyperpermeability and hemodynamic changes induced by intravenous VEGF. Invest. 1999; 40(3):689-96.
3. Thakker GD, Hajjar DP, Muller WA, et al. The role of phosphatidylinositol 3-kinase in vascular endothelial growth factor signaling. J Biol Chem. 1999;274(15):10002-7.
4. Cursiefen C, Chen L, Borges LP. Et al. VEGF-A stimulates lymphangiogenesis and hemangiogenesis in inflammatory neovascularization via macrophage recruitment. J Clin Invest. 2004;113(7):1040-50.
5. Bernatchez PN, Rollin S, Soker S, et al. Relative effects of VEGF-A and VEGF-C on endothelial cell proliferation, migration and PAF synthesis: Role of neuropilin-1. J. Cell Biochem. 2002; 85(3):629-39.
6. Kliffen M, Sharma HS, Mooy CM. et al. Increased expression of angiogenic growth factors in age-related maculopathy. Br J Ophthalmol. 1997;81(2):154-62.
7. Campochiaro PA. Retinal and choroidal neovascularization. J Cell Physiol. 2000; 184(3):301-10.
8. Okamoto N, Tobe T, Hackett SF, et al. Transgenic mice with increased expression of vascular endothelial growth factor in the retina: a new model of intraretinal and subretinal neovascularization. Am J Pathol. 1997; 151(1):281-91.
9. Schwesinger C, Yee C, Rohan RM, et al. Intrachoroidal neovascularization in transgenic mice overexpressing vascular endothelial growth factor in the retinal pigment epithelium. Am J Pathol. 2001;158(3):1161-72.
10. Spilsbury K, Garrett KL, Shen WY, et al. Overexpression of vascular endothelial growth factor (VEGF) in the retinal pigment epithelium leads to the development of choroidal neovascularization. Am J Pathol. 2000; 157(1):135-44.
11. Plate KH ,Warnke PC. Vascular endothelial growth factor. J Neurooncol, 1997; 35(3):365-72.
12. Ferrara N, Gerber HP, LeCouter J. The biology of VEGF and its receptors. Nat Med. 2003; 9(6):669-76.
13. Fairbrother WJ, Champe MA, Christinger HW, et al. Solution structure of the heparin-binding domain of vascular endothelial growth factor. Structure. 1998;6(5):637-48.
14. Keyt BA, Berleau LT, Nguyen HV, et al. The carboxyl-terminal domain (111-165) of vascular endothelial growth factor is critical for its mitogenic potency. J Biol Chem. 199; 271(13):7788-95.
15. Kim KJ, Li B, Winer J, et al. Inhibition of vascular endothelial growth factor-induced angiogenesis suppresses tumour growth in vivo. Nature. 1993; 362(6423):841-4.
16. Presta LG, Chen H, O'Connor SJ, et al. Humanization of an anti-vascular endothelial growth factor monoclonal antibody for the therapy of solid tumors and other disorders. Cancer Res. 1997; 57(20):4593-9.
17. Mordenti J, Cuthbertson RA, Ferrara N, et al. Comparisons of the intraocular tissue distribution, pharmacokinetics, and safety of 125I-labeled full-length and Fab antibodies in rhesus monkeys following intravitreal administration. Toxicol Pathol. 1999; 27(5):536-44.
18. Chen Y, Wiesmann C, Fuh G, et al. Selection and analysis of an optimized anti-VEGF antibody: crystal structure of an affinity-matured Fab in complex with antigen. J Mol Biol. 1999; 293(4):865-81.
19. Gaudreault J, Damico LA, Haughney PC, et al. A pharmacokinetic model describing ocular and systemic disposition of ranibizumab in rabbits. IOVS. 2004; 45:ARVO E-Abstract 1854.
20. Gaudreault J, Escandon E, Maruoka M, et al. Vitreal pharmacokinetics of rhuFab V2 in rabbits using a non-invasive method. IOVS. 2002; 43:ARVO E-Abstract 2801.
21. Gaudreault J, Reich M, Arata A, et al. Ocular pharmacokinetics and antipermeability effect of rhuFab V2 in animals. IOVS. 2003; 44:ARVO E-Abstract 3942.
22. Damico LA, Gaudreault J, Bender BC, et al. Pharmacokinetic study of ranibizumab (Lucentis�) following subconjunctival, intracameral and intravitreal administration in rabbits. IOVS. 2004; 45:ARVO E-Abstract 3952.
23. O'Neill CA, Christian B, Murphy CJ, et al. Safety evaluation of intravitreal administration of rhuFab VEGF in cynomoglus monkeys for 3 months. IOVS. 2000; 41:ARVO E-Abstract 732.
24. Krzystolik MG, Afshari MA, Adamis AP, et al. Prevention of experimental choroidal neovascularization with intravitreal anti-vascular endothelial growth factor antibody fragment. Arch Ophthalmol. 2002; 120(3):338-46.
25. Rosenfeld PJ, Schwartz SD, Blumenkranz MS, et al. Safety of rhuFabV2, an anti-VEGF antibody fragment, as a single intravitreal injection in patients with neovascular AMD. Vitreous Society. 2001;121-122.
26. Heier JS, Sy JR, McCluskey ER, and the rs Study Group. RhuFab V2 in wet AMD - 6 month continued improvement following multiple intravitreal injections. IOVS. 2003; 45:ARVO E-Abstract 972.
27. Heier JS. Review of Lucentis (ranibizumab, rhuFab V2) phase I/II trial results: 6 month treatment of exudative AMD. IOVS. 2004; 45:ARVO E-Abstract 1109.
28. Heier JS. The VEGF antibody approach: the rhuFab collaborative trial. RhuFab V2 for the treatment of wet AMD. American Academy of Ophthalmology Subspecialty Day 2003, Anaheim, CA.
29. Antoszyk AN, Sy JP, McCluskey ER, and the rs Study Group. RhuFab V2 in wet AMD: Changes in OCT measures of the ocular edema. American Society of Retina Specialists, 2003.
30. Puliafito CA, Rosenfeld PJ, McCluskey ER. RhuFab V2 dose escalating trial: safety and tolerability of 3 escalating dosing regimens in subjects with age-related macular degeneration (AMD). [Abstract] Retina Society. 2003.
31. Rosenfeld PJ, Villate N, Feuer WJ, et al. RhuFab V2 (anti-VEGF antibody fragment) in neovascular AMD: safety, tolerability and efficacy of multiple, escalating dose intravitreal injections. IOVS. 2003; 44:ARVO E-Abstract 970.