Do Alternative Treatments for CNV Have a Place in the Era of VEGF Inhibition?

KARTIK KUMAR, MD · R.C. ANDREW SYMONS, MB BS, PhD, FRANZCO

Visual loss due to choroidal neovascularization (CNV) in age-related macular degeneration (AMD) has long been one of the most feared conditions of old age. A variety of therapeutic modalities have been investigated over the years. Most treatments have either been ineffective or resulted in a degree of retinal destruction, leading to irreversible visual loss.

Recently, inhibitors of vascular endothelial growth factor A (VEGF) have revolutionized the treatment of CNV. They represent the first treatment modality that can frequently induce reversal of visual loss. While retinal specialists have certainly welcomed this new modality, its intro duction has raised questions about what role, if any, older therapies for AMD play in this new era. This article will review the landmark trials that led to approval of the VEGF inhibitors, as well as more recent data that have enhanced our knowledge of their efficacy and risk profiles, followed by a discussion of older AMD treatments and the potential for combination regimens.

VEGF ANTAGONISTS

Pegaptanib sodium (Macugen, OSI/Eyetech), bevacizumab (Avastin, Genentech), and ranibizumab (Lucentis, Genentech) are the 3 commercially available VEGF inhibitors available for use in ophthalmology. Pegaptanib was the first of the 3 to have its efficacy proved in a randomized controlled trial and the first to receive Food and Drug Administration (FDA) approval. It is an oligonucleotide aptamer that binds specifically to the 165 amino acid isoform of VEGF, which animal studies suggest is somewhat selective for involvement in pathological neovascularization in contrast to physiological angiogenesis.¹ Pegaptanib was first studied in the VISION-1 trial, a combination of 2 randomized controlled studies comparing pegaptanib with sham injections.² The study looked at patients with exudative AMD regardless of lesion type. Patients received injections every 6 weeks for a total of 48 weeks. At 12 months, patients receiving pegaptanib were found to be significantly protected against visual loss.

Ranibizumab is the Fab fragment of a murine monoclonal antibody that binds all active isoforms of VEGF-A at a single antigen-binding site. Ranibizumab was originally studied in the MARINA³ and ANCHOR⁴ trials. MARINA investigated the treatment of occult or minimally classic CNV, and ANCHOR studied the treatment of predominantly classic CNV. Both were 2-year randomized, controlled studies. In addition, patients were randomized between 2 doses of ranibizumab. Patients randomized to the ranibizumab treatment arms received monthly injections throughout the trial. MARINA demonstrated that 95% of patients receiving ranibizumab lost less than 15 letters of visual acuity (VA) at 12 months as compared to 62% of controls. Additionally, it also demonstrated 25% of the low-dose group and 34% of the high-dose group gained more than 15 letters at 12 months, as compared to 5% of the control group. The results were maintained at 24 months.

In the ANCHOR trial, the control group was treated with photodynamic therapy (PDT) every 3 months as needed plus sham injections. The findings were similar to those of the MARINA trial, with 95% of patients receiving ranibizumab losing >15 letters at 12 months as compared with 64% of patients in the PDT group losing >15 letters. Almost 36% of patients in the low-dose ranibizumab group gained 15 letters or more, and 40% of the high-dose group gained at least 15 letters, as compared with about 5% in the PDT control group. These landmark studies paved the way for ranibizumab to be incorporated into everyday practice. These studies and later experience showed that ranibizumab was far more effective than pegaptanib; now little reason remains to use pegaptanib.

Bevacizumab is a humanized monoclonal antibody that binds all active isoforms of VEGF. The original mouse monoclonal antibody that is the basis for bevacizumab was also the starting point for ranibizumab, although an affinity maturation process was performed as part of the design process of ranibizumab. Bevacizumab was first used for metastatic colon cancer, during the window period after it had become obvious that ranibizumab was highly effective in treatment of CNV but before it was commercially available. Bevacizumab was first infused intravenously for the treatment of CNV and was later injected into the vitreous. A few studies have been performed to demonstrate the efficacy of bevacizumab. The findings, in terms of VA and macular thickness, were consistent with the results of ranibizumab studies. However, no randomized, controlled studies have been performed to date to allow for FDA approval of the drug. Nonetheless, bevacizumab is significantly cheaper than ranibizumab and the off-label use of bevacizumab for the treatment of CNV is prevalent.

Intravitreal injections with VEGF inhibitors may be complicated by retinal detachment, endophthalmitis, vitreous hemorrhage, lenticular damage, and uveitis. In addition, ranibizumab and bevacizumab have possible class effects of hypertension, thrombosis, and bleeding. In the SAILOR trial, there was a trend toward a higher stroke risk in the 0.5 mg ranibizumab group than in the 0.3 mg group. Because of the side effects and burden of frequent intraocular injections, alternative treatments are being investigated.

WHICH LESIONS SHOULD BE TREATED?

Ranibizumab and pegaptanib were tested for the treatment of CNV lesions involving the central fovea. Therefore, there are no trial data supporting the use of these medications for juxta- or extra-foveal lesions. In practice, almost all physicians are using the medications for juxtafoveal lesions, and many physicians are using them for extrafoveal lesions. Although not supported by the trial data, these are reasonable approaches, particularly for lesions where a scotoma is likely to be symptomatic and possibly disabling.

There is some controversy concerning whether occult lesions should be treated in the absence of documented disease progression. This controversy arose because some such lesions can remain stable for long periods and because of the reaction to the sudden loss of vision sometimes observed in treating occult lesions with PDT. The MARINA study's inclusion criteria included evidence of disease progression as defined on the basis of observable blood, recent vision loss, or a recent increase of at least 10% of the lesion's greatest linear diameter. It is probably reasonable to add further indications for the treatment of occult lesions with ranibizumab or bevacizumab in practice.

Many physicians would treat in a situation where they had previously documented absence of a lesion and later observed an occult lesion. Likewise, in a situation where a patient presents with an occult lesion without substantial fibrosis and a chronic symptomatic reduction in vision that is consistent with the level of retinal thickening on optical coherence tomography (OCT), it may be reasonable to perform a therapeutic trial of a VEGF inhibitor.

The ANCHOR trial excluded patients with permanent structural damage to the central fovea, and the MARINA trial excluded patients with subfoveal fibrosis or atrophy. CNV can be defined histologically as a form of fibrotic or wound-healing process. Clinicians readily recognize the occurrence of sub-retinal "fibrosis" as part of the natural history of CNV, but there is no formal definition to separate this phenomenon from the remainder of CNV lesions.

It may be possible to incorporate clinical and OCT observations to create a useful definition of clinical fibrosis. Anecdotal evidence from the clinic suggests that cystoid degeneration overlying regions of clinical fibrosis may show a response to VEGF inhibition in terms of a diminution of retinal thickness by OCT. However, improvement in visual acuity is unlikely. We would generally prefer not to treat such patients except in the tragic situation where the eye to be treated is the eye with better vision.

HOW SHOULD I DECIDE WHEN TO RETREAT?

The ANCHOR and MARINA trials proved the efficacy of monthly injections of ranibizumab. The question that immediately arose was whether so many injections were really necessary. The PIER study demonstrated that when patients changed to quarterly dosing after 3 initial monthly injections with ranibizumab some of their initial gains in VA were lost.⁵ Therefore the PIER protocol is an unsatisfactory alternative to monthly injections.

Most clinicians appear to be tailoring their treatment regimens to a patient's response to treatment. The PrONTO study explored the effects of using OCT and clinical examination to guide retreatment with ranibizumab.⁶ After an initial course of 3 injections at monthly injections, subsequent treatments were performed when retreatment criteria were met. These criteria were a worsening of visual acuity by 5 letters, the presence of macular fluid on OCT, an increase in central retinal thickness of at least 100μ on OCT, new classic CNV and new macular hemorrhage. The visual acuity gains made at the 3 month time point were largely sustained at the end of 12 months. However, there was variability of the mean visual acuity and OCT thickness, suggesting that the protocol allowed rebounds that were then followed by catch-up treatment.

Criticisms can be made of the PrONTO protocol. Greater gains might have been obtained if mandatory monthly injections had been given for a longer period. Also, VEGF has a mitogenic effect on vessels and a permeability effect. Since many of the retreatments occurred after evidence of a recurrence of permeability of the neovascular membranes, we might presume that unnecessary exposure of the membranes to VEGF had occurred and therefore there was unnecessary progression of membranes.

These criticisms argue for an even more conservative protocol than that of the PrONTO study. At baseline, fluorescein angiography (FA) and OCT should be performed. In most cases, at least 3 initial monthly injections should be performed, as per the trials. Treatment should be continued as long as there is improvement in VA or in OCT indications of disease activity, and it should be continued for at least 1 extra injection after stabilization appears to have occurred to ensure that there is no extra gain to be made. When VA and OCT appear to have stabilized, performance of FA is probably of value to detect leakage, indicating lesion activity that was not evident on OCT scanning. When the decision is made not to continue to treat, monthly examinations and OCT scans should be obtained for several months to ensure that there is no return of lesion activity. Since the risk of adverse effects is low, it is better to err on the side of treating whenever there is a doubt about whether a lesion is still active.

IS THERE STILL A ROLE FOR LASER PHOTOCOAGULATION?

Laser photocoagulation with the aim of ablating neovascular tissue is the oldest of the therapeutically useful techniques for treating CNV. The Macular Photocoagulation Study (MPS) demonstrated its efficacy. It is still the only treatment with formal proof of its efficacy in extrafoveal lesions. The utility of laser photocoagulation is limited by the absolute scotomas that occur immediately after treatment and by the risk of recurrence. Laser photocoagulation still has a place in the armamentarium of many clinicians for extrafoveal lesions, where a complete treatment is possible with less resultant visual disability. Given the reasonably benign risk profile of intravitreal injections of ranibizumab, it is reasonable to use a VEGF inhibitor whenever there is the likelihood of significant visual disability resulting from laser treatment.

DO INTRAVITREAL STEROIDS HELP CNV?

Intravitreal triamcinolone acetonide has been used in the treatment of CNV in AMD. Corticosteroids may modify the pathophysiology of CNV in several ways: they inhibit the inflammatory response involved in neovascularization; in vitro they inhibit the response or choroidal endothelial cells to basic fibroblast growth factor expression, which is a potent proangiogenic factor; and they decrease vascular permeability, which may decrease influx of proangiogenic factors and reduce retinal thickening.

Several small studies describe the effects of triamcinolone monotherapy for CNV in AMD with mixed results. The studies were performed on a variety of lesions, with a variety of combinations of classic and occult components, and with subfoveal, juxtafoveal, and extrafoveal locations. However, no consistent results were obtained. In the largest of the studies,⁷ 151 eyes with any component of classic AMD were injected with a single 4-mg dose of triamcinolone. In this study, triamcinolone was not protective against loss of vision; however, at 3 months there was significant protection against increase in membrane size.

Several smaller studies with 25-mg injections did demonstrate an increase in VA over controls; however, further randomized, controlled studies, which are no longer warranted in the age of VEGF inhibition, would be required to provide convincing evidence of this effect. Finally, serious adverse effects of corticosteroids include elevated intraocular pressures, endophthalmitis, and an increased rate of cataract formation. There is no longer a role for use of intravitreal steroids in isolation.

PHOTODYNAMIC THERAPY

A major breakthrough in CNV treatment came with PDT with liposomal verteporfin (Visudyne, QLT/Novartis). Photoactivated verteporfin creates reactive oxygen species that oxidize lipids in cellular membranes, thus disrupting cellular structures, causing thrombosis and occlusion of active CNV. The efficacy of PDT with verteporfin in preventing visual loss from predominantly classic CNV in AMD over 24 months was demonstrated in the TAP studies⁸ and its efficacy for preventing visual loss in patients with occult lesions with no classic component was demonstrated in the VIP studies.⁹

Treatment with PDT with verteporfin resulted in little in the way of visual improvement. Therefore, its use has largely disappeared due to its lack of efficacy compared with VEGF inhibition. Moreover, PDT is not without its concerns. Animal histological studies shows choriocapillaris occlusions and retinal pigment epithelium (RPE) atrophy secondary to PDT. Human data echoed the concern regarding choriocaipllaris occlusions, but found no correlation with RPE atrophy. The study also suggested there to be a direct correlation between the number of PDT treatments and likelihood of choriocapillaris occlusion.¹⁰ There is evidence that using lower laser power than was used in the VIP and TAP studies, in an approach known as "reduced fluence," may have equal efficacy and cause less choroidal hypoperfusion.¹¹

OTHER THERAPIES

Several other forms of treatment have been and are still undergoing evaluation for a possible role in the treatment of AMD, including transpupillary thermotherapy, in which a long pulse diode laser is used to occlude CNV, brachytherapy, and submacular surgery. None of these has demonstrated routine success in randomized, controlled studies.

COMBINATION THERAPIES

■ Intravitreal Triamcinolone and PDT. The addition of triamcinolone to PDT is thought to improve results due to the antiangiogenic and anti-inflammatory effects of triamcinolone. This may reduce reperfusion of the choroidal neovascular membrane and decrease the likelihood of recurrence. A number of studies have been performed on the entire spectrum of CNV lesions and have demonstrated beneficial effects, including a mild increase in VA. However, most studies regarding this combination have been nonrandomized. Furthermore, 2 studies noted a decrease in VA post-treatment. In the age of VEGF inhibition, this approach has a very small role.

■ VEGF Inhibition and Intravitreal Triamcinolone. One small retrospective study investigated the potential benefits of a combination of bevacizumab and triamcinolone as compared to bevacizumab alone. Based on OCT readings, the study found that patients receiving combination therapy were more likely to maintain a strong biological response to the injections than patients receiving bevacizumab alone. The major flaw of this study was that bevacizumab was only injected every 3 months, allowing time for lesion progression between injections. Additionally, the study was performed retrospectively and involved only 43 eyes.

■ VEGF Inhibition and PDT. Several studies have been performed comparing the combination of PDT and intravitreal VEGF inhibitors vs. PDT monotherapy. The studies have generally shown that combination therapy is more effective than PDT alone. But few if any studies have compared the potential benefits of combination therapy to intravitreal VEGF inhibitors alone. Studies are currently recruiting to determine whether photodynamic therapy with verteporfin provides supplemental benefit when used in combination with VEGF inhibition.

SUMMARY

In this exciting new era of anti-VEGF agents, VEGF inhibition alone is an appropriate treatment strategy for most CNV lesions in age-related macular degeneration. A role may exist for combinations of VEGF inhibition with triamcinolone or PDT in patients exhibiting poor response to VEGF inhibition or who have poor access to retinal care facilities. In addition, there are new therapeutic agents on the horizon and these drugs will be the subject of later articles in this series. RP

REFERENCES

1. Ishida S, Usui T, Yamashiro K, et al. VEGF164-mediated inflammation is required for pathological, but not physiological, ischemia-induced retinal neovascularization. J Exp Med. 2003;198:483-489.
2. Gragoudas ES, Adamis AP, Cunningham ET Jr, et al. Pegaptanib for neovascular age-related macular degeneration. N Engl J Med. 2004;351:2805-2816.
3. Rosenfeld PJ, Brown DM, Heier JS, et al; MARINA Study Group. Ranibizumab for neovascular age-related macular degeneration. N Engl J Med. 2006;355:1419-1431.
4. Brown DM, Kaiser PK, Michels M, et al. ANCHOR Study Group. Ranibizumab versus verteporfin for neovascular age-related macular degeneration. N Engl J Med. 2006;355:1432-1444.
5. Regillo CD, Brown DM, Abraham P, et al. Randomized, double-masked, sham-controlled trial of ranibizumab for neovascular age-related macular degeneration: PIER Study year 1. Am J Ophthalmol. 2008;145:239-248.
6. Fung AE, Lalwani GA, Rosenfeld PJ, et al. An optical coherence tomography-guided, variable dosing regimen with intravitreal ranibizumab (Lucentis) for neovascular age-related macular degeneration. Am J Ophthalmol. 2007;143:566-583.
7. Gillies MC, Simpson JM, Luo W, et al. A randomized clinical trial of a single dose of intravitreal triamcinolone acetonide for neovascular age-related macular degeneration: one-year results. Arch Ophthalmol. 2003;121:667-673.
8. Bressler NM; TAp Study Group. Photodynamic therapy of subfoveal choroidal neovascularization in age-related macular degeneration with verteporfin: two-year results of 2 randomized clinical trials-tap report 2. Arch Ophthalmol. 2001;119:198-207.
9. Verteporfin In Photodynamic Therapy Study Group. Verteporfin therapy of subfoveal choroidal neovascularization in age-related macular degeneration: two-year results of a randomized clinical trial including lesions with occult with no classic choroidal neovascularization--verteporfin in photodynamic therapy report 2. Am J Ophthalmol. 2001;131:541-560.
10. Dewi NA, Yuzawa M, Tochigi K, Kawamura A, Mori R. Effects of photodynamic therapy on the choriocapillaris and retinal pigment epithelium in the irradiated area. Jpn J Ophthalmol. 2008;52:277-281.
11. Michels S, Hansmann F, Geitzenauer W, Schmidt-Erfurth U. Influence of treatment parameters on selectivity of verteporfin therapy. Invest Ophthalmol Vis Sci. 2006;47:371-376.