Laser treatments
Thermal laser photocoagulation
Ophthalmologists have traditionally used thermal laser destruction of CNV as the primary treatment of exudative ARMD based on the results of the MPS. This study, which was initiated in the 1980s and supported by the National Institutes of Health, demonstrated that laser photocoagulation of extrafoveal, juxtafoveal, and subfoveal choroidal neovascularization limited the risk of large reductions in visual acuity compared with observation alone.
Patients were eligible for laser photocoagulation if they had classic choroidal neovascularization, as determined on RSFA. However, only 13-26% of patients with exudative ARMD presented with this pattern. Therefore, whether laser photocoagulation was beneficial most patients with other leakage patterns on RSFA is unclear because they were not eligible for laser photocoagulation in the MPS (MPS Group, 1982, 1986, 1991, and 1993; Freund, 1993; Moisseiev, 1995). Moreover, at least one half of the enrolled subjects have persistent or recurrent choroidal neovascularization within 2 years of treatment (MPS Group Arch Ophthalmol, 1986, pp 694-701 and 503-12,; MPS Group 1991).
Although data from the arm of the MPS exploring treatment of choroidal neovascularization under the fovea suggested that laser photocoagulation was better than observation, many clinicians have not treated subfoveal choroidal neovascularization with thermal photocoagulation because of the induction of an iatrogenic, immediate central scotoma (MPS group, 1991 and 1993). Therefore, researchers have searched for an alternative means of treating subfoveal CNV by using a variety of laser derivatives.
Feeder-vessel photocoagulation
ICG angiography allows the precise identification of feeder vessels to the subfoveal choroidal neovascularization in a subset of patients with ARMD. Therefore, several investigators have attempted to apply laser photocoagulation to the extrafoveal vessels to eliminate the source of the choroidal neovascularization while preserving much of the overlying foveal tissue iatrogenically destroyed with traditional thermal laser photocoagulation.
In 1998, Shiraga et al reported that 70% of patients treated in this manner had complete resolution of choroidal neovascularization, and 68% had stable or improved visual acuity. However, other investigators have reported poor closure rates and outcomes (Bloom, 1998; Freund, 1998; Staurenghi, 1998). In 2002, Piermarocchi and associates found that the number of eyes with detectable feeder vessels detectable by using ICG increased from 22.4% to 84.2% after PDT; therefore, they suggested the use of PDT as an adjunct to photocoagulation of the feeder vessels.
Transpupillary thermotherapy
With TTT, the subfoveal CNV complex is slowly heated with infrared (810 nm) diode laser energy to occlude the CNV complex with treatment of a single large spot. The infrared wavelength is thought to traverse the retina and RPE to maximally affect the CNV membranes while minimizing thermal injury to the overlying neurosensory retina.
Although the precise mechanism of CNV destruction is unclear, findings from 1 study of color Doppler imaging suggested TTT alters choroidal blood flow (Ciulla, 2001). Proponents of TTT suggest that it may play a role in managing occult subfoveal choroidal neovascularization because current therapies are limited. In an uncontrolled retrospective series of 16 eyes in 15 patients undergoing TTT for occult subfoveal choroidal neovascularization, 94% eyes had decreased exudation, as observed on fluorescein angiography, and no eyes showed any deleterious effects (Reichel, 1999).
The Verteporfin in Photodynamic Therapy (VIP) trial, an uncontrolled phase I-II safety and efficacy study of 113 patients showed that results in patients with occult choroidal neovascularization who received TTT were similar to those of patients treated with verteporfin at 6 and 12 months (Algvere, 2003). Another uncontrolled trial with 69 patients demonstrated that TTT compared favorably to the natural history of occult choroidal neovascularization (Thatch, 2003).
The Transpupillary Thermotherapy of Occult Subfoveal Choroidal Neovascular Membranes in Patients with Age-Related Macular Degeneration Trial (TTT4CNV), a randomized, prospective, double-blind, placebo-controlled study, enrolled 303 patients in 22 centers. Eyes with subfoveal occult CNV membranes and visual acuity between 20/50 and 20/200 were randomly assigned to TTT or sham treatment (Optimed, 2002; Nader, 2004) Evaluation of all participants did not reveal a statistically significant difference between the groups; however, subgroup analysis of the 116 patients with a visual acuity 20/100 or worse showed a statistically significant benefit to visual acuity in the TTT group at 18 months (Schultz, 2005).
Photodynamic therapy
With PDT, laser energy and intravascular dyes (eg, photosensitizers) are used to achieve a therapeutic effect. After intravenous injection and sufficient time to concentrate the photosensitizer in neovascular tissue, the CNV membrane is stimulated with a specific wavelength of light to activate the photosensitizer, which reacts with water to create oxygen and hydroxyl free radicals (Aveline, 1994). These free radicals, in turn, induce occlusion of the pathologic vasculature by means of massive platelet activation and thrombosis while preserving the normal choroidal vasculature and nonvascular tissue (Allison, 1991; Hunt, 1999).
The intensity of the exciting wavelength is ideally low enough to spare the non-neovascular irradiated tissues from thermal damage. Important variables in this reaction include the concentration of the intravascular dye, the photochemical behavior of the dye, and the interval between the injection and the onset of irradiation, the intensity and specificity of the exciting irradiation, and the duration of irradiation (Hope-Ross, 1994; Moriarty, 1994; Reichel, 1994).
Verteporfin therapy
In April 2000, the FDA approved verteporfin (Visudyne; QLT Therapeutics, Inc, Vancouver, British Columbia, Canada, and Novartis Ophthalmics, Bulach, Switzerland) for use in patients with predominantly classic, subfoveal CNV caused by ARMD. Marketing approval was granted in Europe in July 2000, and the drug is currently commercially available in more than 70 countries for the treatment of predominantly classic CNV (Novartis, 2004).
Verteporfin is a modified porphyrin with an absorption peak near 689 nm that is delivered intravenously for 10 minutes. After a 5-minute delay, the CNV complex is irradiated through the pupil with a large-spot diode laser at 689 nm for 83 seconds. The laser energy activates the intravascular photosensitizer and stimulates the photodynamic action within the CNV. Verteporfin is cleared rapidly from the body, resulting in minimal skin sensitivity after 5 days.
In 1999 and 2001, the 1- and 2-year results of the Treatment of AMD with PDT (TAP) study were published. TAP consisted of 2 randomized, prospective, double-blind, placebo-controlled phase III trials with 609 subjects. First-year data showed that the proportion of eyes with < visual-acuity loss of <15 letters on a standardized eye chart was 67% in the treated group versus 39% in control group (P <0.001) when the CNV was predominantly classic. However, no significant differences in visual acuity were demonstrated when the area of classic CNV was <50% of the entire complex. Also, researchers noted that 90% of the subjects required retreatment at 3 months and an average of more than 3 repeat treatments over the first year (TAP Study Group, 1999).
Second-year data showed that 59% of treated eyes had a favorable visual outcome vs 31% in the control group when the lesion was predominantly classic (Bressler, 2001). The TAP trial was unmasked after 2 years of follow-up, and investigators continued with an open-label extension (to 36 mo) in 124 of the 159 original TAP participants with predominantly classic CNV. The data revealed that visual acuity remained nearly constant and the number of required repeat treatments decreased (Blumenkranz, 2002). Because of the success of this trial, the VIP trial, another randomized, prospective, double-blind, placebo-controlled clinical trial, was developed to examine many of the patients who were excluded from the TAP study.
The VIP study was designed to evaluate the efficacy of PDT in 339 subjects with total occult subfoveal choroidal neovascularization, predominantly classic choroidal neovascularization with visual acuity better than 20/40, or choroidal neovascularization secondary to pathological myopia. One-year results of the occult-ARMD arm showed no significant difference between unfavorable visual-acuity outcomes in patients with exudative ARMD treated with verteporfin (51%) and those receiving placebo (54%); however, at 2 years, rates were 55% versus 68%, respectively (P = 0.023). On average, verteporfin-treated patients received 5 treatments over 24 months of follow-up. On the basis of this data, the study group recommended verteporfin for the treatment of purely occult subfoveal CNV with recent disease progression in all patients except those with large lesions with good visual acuity (VIP Study Group, 2001; Bressler, 2002).
Because the FDA desired additional data before approving verteporfin for occult CNV, the Visudyne in Occult (VIO) trial was developed as a 24-month study to examine patients with only occult CNV. Enrollment of 364 subjects was completed in September 2003, and the trial is currently in its second year of follow-up (QLT, 2005).
Several groups have evaluated the efficacy of verteporfin in a variety of clinical situations previously lacking sufficient data. Retrospective data from the TAP and VIP studies suggested some treatment benefit for small, minimally classic lesions. The Visudyne in Minimally Classic Trial (VIM) was a randomized, prospective, double-blind, placebo-controlled study of the use of verteporfin in patients with minimally classic choroidal neovascularization. Phase II data in 117 patients suggested that small, recently progressive, minimally classic choroidal neovascularization might benefit from verteporfin therapy (Bressler, 2003; Gonzalez, 2003). Two-year follow-up data revealed fewer verteporfin-treated eyes lost 3 or more lines of vision on a standard visual-acuity chart or converted to a predominantly classic lesion versus placebo (P = 0.01) (Rosenfeld, 2004).
A phase III study, the Visudyne Minimally Classic (VMC) Trial, was consequently initiated in late 2003 to further evaluate verteporfin for the treatment of minimally classic choroidal neovascularization (QLT, 2005). On April 1, 2004, the United States Centers for Medicare and Medicaid Services (CMS) agreed to reimburse physicians for PDT of occult and minimally classic subfoveal choroidal neovascularization due to ARMD provided that the lesion is 4 disc areas or smaller at least 3 months before initial treatment with evidence of progression (ie, loss of 5 or more letters on standard visual-acuity charts, increase of at least 1 disc diameter, or appearance of blood) within 3 months of treatment (CMS, 2004). In light of that decision, the VMC trial was halted in mid-2004 (QLT, 2005).
Because 80% of vision loss in verteporfin-treated patients occurs within 6 months of choroidal neovascularization, the Verteporfin Early Retreatment (VER) trial was conducted as a phase III study of 323 patients to compare the benefit of retreatment at 6-week intervals versus the standard 12 weeks (Riddle, 2003; LuEster T. Mertz Retinal Research Center). Twelve-month interim results of the 2-year trial did not show improved outcomes with the 6-week versus standard treatment (Stur, 2004).
The Verteporfin with Altered (Delayed) Light in Occult (VALIO) study was developed to evaluate whether delaying the light application to 30 minutes after the initiation of verteporfin infusion (vs standard 15 min) improves outcomes in occult CNV. Phase II data at 6-month follow-up showed that the group treated at 30 minutes after infusion lost 1.3 lines of vision, whereas the group treated at 15 minutes lost 2-3 lines; the difference was not statistically significant (Riddle, 2003; Slakter, 2003). One-year data substantiated the 6-month findings (Singerman, 2004). Because verteporfin is the only agent currently approved for PDT, additional photosensitizing products are under development.
Rostaporfin therapy
Rostaporfin (Photrex; formerly SnET2; Miravant Medical Technologies, Santa Barbara, CA) is a purpurin with a structure similar to that of chlorophyll and maximal absorption at 664 nm (Peyman, 1997; Moshfeghi, 1998). Like verteporfin, the preconstituted solution of rostaporfin is intravenously infused over 10-20 minutes (Regillo, 2000). In December 2001, enrollment for a phase III placebo-controlled, double-masked clinical trial of 920 patients was completed. Two-year follow-up data showed that 58% of patients receiving a 0.5-mg/kg dose of SnET2 lost <15 letters compared with 42% of patients receiving placebo (P = 0.0045). Rostaporfin was well tolerated and had an acceptable safety profile (Thomas, 2004). On September 30, 2004, the FDA requested an additional confirmatory clinical trial before approving final marketing; this trial is scheduled to begin in mid-2005 (Miravant Medical Technologies, 2004).
Other PDT agents
Motexafin lutetium (Optrin; Pharmacyclics Inc, Sunnyvale, CA) is activated by 732-nm light and can be used as both an imaging and a photosensitizing agent. It had shown promise in phase II trails involving 75 patients with ARMD; however, 77% of subjects receiving therapeutic doses developed peripheral-extremity paresthesias (Blumenkranz, 2000). In October 2001, Pharmacyclics gained the worldwide rights to develop and market the product from Alcon (Pharmacyclics, 2001); however, further development is currently stalled because of adverse effects observed in the phase II trials (Riddle, 2003).
Talaporfin sodium (Light Sciences Corporation, Snoqualmie, WA) is currently in an early clinical trial in Europe (Light Sciences Corporation, 2004).
In addition, preclinical studies of ATX-S10 (Na) (Allergan Inc, Irvine, CA, and Photochemical Co, Ltd, Okayana, Japan) have demonstrated the ability to occlude choroidal vessels in nonhuman primates (Obana, 2000).
Receptor-targeted PDT
Researchers are in the preclinical stages of developing receptor-targeted PDT. Instead of a nonspecific vaso-occlusion based on a high, generalized, intravascular concentration of photosensitizer, conjugated photosensitizer is concentrated in neovascular tissue by binding to receptors expressed preferentially in CNV membranes. By conjugating verteporfin to a VEGF receptor-2 (VEGFR2) antagonist and then performing PDT, investigators achieved 100% angiographic closure in 17 rat laser-injury models of CNV. Histologic examination revealed minimal collateral damage to the surrounding retina structures in treated subjects compared with verteporfin-treated controls (Renno, 2004).
Combination PDT
In 2003, Schmidt-Erforth and associates found that expression of VEGF is enhanced after PDT with verteporfin. Therefore, investigators have reasoned that PDT combined with intravitreal triamcinolone acetonide (a corticosteroid with antiangiogenic properties) might enhance the effect of PDT by limiting VEGF expression. A noncomparative pilot study of 26 eyes with CNV secondary to ARMD showed that subjects treated with verteporfin PDT followed by intravitreal injection of triamcinolone acetonide 4 mg showed significant visual improvement over baseline (P = .01) in previously untreated eyes, with few repeat treatments over 12 months of follow-up (Spaide, 2005).
Four additional studies are planned by using combination verteporfin and triamcinolone acetonide, including 1 sponsored by the National Eye Institute, which is expected to enroll around 300 subjects (QLT Inc, 2005). In addition, a recent report showed that the combination of intravitreal ranibizumab, a VEGF inhibitor, and verteporfin caused less angiographic leakage than either modality alone in monkey eyes (Husain, 2005).
Although standard PDT with verteporfin has shown promise in treating some forms of CNV, numerous repeat treatments are often required, it is expensive, and it typically slows vision loss rather than improves it. In addition, PDT can damage adjacent normal tissue containing the photosensitizer (Nishikawa, 2002). Findings in immunohistopathologic specimens suggest that PDT with verteporfin caused only short-term damage to the CNV membranes, which often returned to baseline in weeks (Grisanti, 2004). Therefore, other pharmacologic interventions to treat subfoveal CNV membranes are in all stages of development.
Antiangiogenic agents
Although the exact stimulus, or more likely stimuli, that precipitates CNV formation remains speculative, recent evidence indicates that a combination of inflammatory cytokines are involved in angiogenesis. Circulating endothelial progenitor cells, monocytes, circulating and resident macrophages, endothelial cells, and even astrocytes have been implicated as potential cellular sources for cytokine release during CNV formation. One theory proposes that macrophages secrete angiogenic growth factors as an initial response to injury to the Bruch membrane (Ishibashi, 1985).
Whatever initiates CNV formation, angiogenic growth factors are ultimately involved, and an imbalance of angiogenic promoters and inhibitors occurs. Surgically excised and post mortem CNV tissue, as well RPE cells, are immunoreactive for various proangiogenic growth factors, including VEGF, transforming growth factor-beta (TGF-b), platelet-derived growth factor (PDGF), and basic fibroblast growth factor (bFGF or FGF-2) (Amin, 1994; Kvanta, 1995; Reddy, 1995; Lopez, 1996).
VEGF inhibitors
Findings from animal and clinical studies have established VEGF as a key mediator in ocular angiogenesis. Investigators have reported up-regulation of VEGF expression in experimentally induced CNV in rats (Yi, 1997). Another group developed a model of retinal and subretinal neovascularization by using a transgenic model driven by overexpression of VEGF and showed that excessive VEGF is sufficient for intraretinal and subretinal neovascularization (Okamoto, 1997).
In human clinical trials, particular attention has focused on the development of pharmaceutical agents to block VEGF expression or neutralize it after it is expressed. Investigators have inhibited preretinal neovascularization in experimental models with antibodies against VEGF (Adamis, 1996). Others have shown similar effects using VEGF-neutralizing chimeric proteins, which were constructed by joining the extracellular domain of high-affinity VEGF receptors with immunoglobulin G (IgG) (Aiello, 1995).
Pegaptanib sodium
Pegaptanib sodium (Macugen; Eyetech Pharmaceuticals, Inc, New York, NY and Pfizer, Inc, New York, NY) is an anti-VEGF pegylated aptamer (a DNA or RNA molecule selected from random pools on the basis of its ability to bind other molecules). Pegaptanib sodium demonstrated both safety and efficacy in clinical trials. This intravitreally administered polyethylene glycol (PEG)–conjugated oligonucleotide was specifically designed to bind and neutralize VEGF165, hypothesized to be the predominant VEGF isomer associated with CNV in humans.
A phase I trial of 15 subjects receiving a single injection of pegaptanib sodium, demonstrated 80% with stable or improved vision at 3 months. More impressive was that 27% of patients had significantly improved vision, a finding not observed with many standard ARMD treatment modalities (Eyetech Study Group, 2002). Although small, a phase II trial of 21 patients provided data ion support the phase I data. When injections of pegaptanib sodium were combined with verteporfin PDT, 6 (60%) of 10 patients had significantly improved vision versus 2.2% treated with PDT alone (Eyetech Study Group, 2003).
The VEGF Inhibition Study in Ocular Neovascularization (VISION) Study comprised 2 phase II-III multicenter, randomized, placebo-controlled trials. Enrollment of 1186 subjects was completed in July 2002. The 12-month data for all types of choroidal neovascularization showed that 70% of subjects receiving a 0.3-mg intravitreous injection every 6 weeks lost <3 lines of vision versus 55% of control subjects receiving sham injection (P <0.001) (Gragoudas, 2004). Furthermore, follow-up data revealed less vision loss for subjects receiving maintenance therapy with pegaptanib sodium for 2 years than with those receiving therapy for 1 year (P <0.05) (Schwartz, 2004; VISION Study Group, 2004).
The treatment group did not have an increased rate of permanent ocular or systemic complications. After 7545 injections, complications included 12 subjects (0.16% per injection) with endophthalmitis (which was significantly reduced after the injection technique was changed), 5 retinal detachments (3 rhegmatogenous, 2 exudative; 0.07% per injection), and 5 traumatic cataracts (0.07% per injection) (D'Amico, 2004). The 2-year data revealed no new safety concerns (Roach, 2004; Eyetech Pharmaceuticals, 2004). The FDA accepted a new drug application (NDA) for wet ARMD in August 2004, as did the European Medicines Agency (EMEA) in September 2004 (Eyetech Pharmaceuticals, Inc, and Pfizer Inc, 2004). On December 17, 2004, the FDA approved the drug, which became available for consumer use in the United State in January 2005 (FDA, 2005).
Ranibizumab
Ranibizumab (Lucentis, formerly rhuFab V2; Genentech Inc, South San Francisco, CA, and Novartis Ophthalmics, Basel, Switzerland), an intravitreally injected, recombinant, humanized, monoclonal antibody Fab fragment designed to actively bind and inhibit all isoforms of VEGF, has shown promise in early human trials. A phase Ib-II randomized, single-agent study showed that 94% of the 50 patients receiving ranibizumab had stable vision and that 44% had significantly improved vision at 6 months (Genentech, 2002; Heier, 2003 and 2004).
The Minimally Classic/Occult Trial of the Anti-VEGF Antibody Ranibizumab (formerly, RhuFab) In the Treatment of Neovascular AMD (MARINA) was a phase III randomized, prospective, double-blind, placebo-controlled comparison of ranibizumab against sham controls. Investigators enrolled 716 patients. At 12-month follow-up 95% of those treated with monthly ranibizumab injections has improved or stable vision versus 62% of control subjects receiving sham treatment (P <0.0001) (Genentech, May 23, 2005).
In addition, a 2-year phase I/II study, (RhuFab V2 Ocular Treatment Combining the Use of Visudyne to Evaluate Safety (FOCUS), was conducted to evaluate the efficacy of concurrent ranibizumab and verteporfin PDT in 162 subjects with predominantly classic CNV. About 90% of subjects receiving ranibizumab and verteporfin in combination maintained or improved visual acuity versus 68% of those receiving verteporfin alone at 12 months (P = 0.0003) (Genentech, May 31, 2005).
Another prospective, randomized, multicenter, double-blind phase III trial, Anti-VEGF Antibody for the Treatment of Predominantly Classic Choroidal Neovascularization in AMD (ANCHOR), was performed to compare 2 dosages of ranibizumab with verteporfin alone in 423 subjects with predominantly classic exudative ARMD. Results are expected in late 2005 (Genentech, May 23, 2005).
Investigators in an ongoing study of alternative dosing (PIER study) started enrolling approximately 180 patients in September 2004 to evaluate 3-month intravitreal dosing intervals versus the standard 1-month intervals (Heier, 2004). Results are expected in mid-2006 (Genentech, May 23, 2005).
Bevacizumab
Bevacizumab (Avastin, Genentech Inc, South San Francisco, CA) is a full-length humanized monoclonal antibody against human VEGF (whereas ranibizumab is a fragmented humanized monoclonal antibody against human VEGF) that the FDA approved for the treatment of metastatic colorectal cancer (Genentech, September 2005; Reddy, 2005). A preclinical study in 132 monkeys demonstrated drug leakage from laser-induced choroidal neovascularization after intravenous administration. Therefore, researchers initiated the Systemic Avastin for Neovascular ARMD (SANA) Study, an open-label uncontrolled pilot study of 9 subjects with subfoveal choroidal neovascularization to evaluate the efficacy of intravenous bevacizumab. Patients were infused with 5 mg/kg bevacizumab every 2 weeks for 2-3 treatments. Follow-up through 12 weeks revealed significant improvements in mean visual acuity (P = 0.008) and central retinal thickness (P = 0.001) over baseline with a marked reduction in leakage on RSFA (Michels, 2005).
VEGF trap
The VEGF trap (Regeneron Pharmaceuticals, Tarrytown, NY, and Aventis, Strasbourg, France) is a high-affinity recombinant fusion protein consisting of the immunoglobulin domain 2 of the VEGF-R1 receptor and domain 3 of the VEGF-R2 receptor fused to the crystallizable fragment of human IgG. This antigen selectively binds and neutralizes all exogenous VEGF-A molecular isoforms as well as placental growth factor. Administration can be either local or intravenous.
In preclinical evaluations, the VEGF trap was evaluated as a possible antiangiogenic agent in tumor therapy (Holash, 2002). Using murine choroidal and retinal neovascularization models, Campochiaro and associates determined that this agent inhibited choroidal neovascularization, preretinal neovascularization, and retinal vascular leakage. It also reduced breakdown of the blood-retinal barrier (Saishin, 2003).
In a randomized, double-masked, ascending dose, placebo-controlled phase I trial, 25 subjects with advanced wet ARMD received either placebo or 1 or 3 intravenous doses of the VEGF trap. Results revealed a dose-dependent decrease in retinal thickness; however, a dose-dependent increase in blood pressure was also observed. Therefore, a phase I study is planned to begin in mid-2005 to evaluate the VEGF trap administered intravitreally (Regeneron Pharmaceuticals, 2005).
Small interfering RNA therapy
RNA interference (RNAi) is a method of posttranscriptional gene silencing in which double-stranded RNA is used to target a specific messenger RNA (mRNA) transcript. Small interfering RNA (siRNA) destroys targeted mRNAs, thereby silencing the expression of the target gene.
The siRNA molecule is 21-nucleotide double-stranded RNA that mediates RNAi, specifically targets the pathologic mRNA, such as VEGF. One siRNA molecule can destroy hundreds of mRNA, resulting in the suppression of thousands of VEGF proteins. Instead of antagonizing the VEGF after it is produced, siRNA can stop the production of VEGF altogether.
In preclinical trials, siRNA directed against VEGF effectively silenced VEGF expression in murine models, and it inhibited laser-induced choroidal neovascularization in murine and nonhuman primate models with no signs of toxicity (Tolentino, 2004; Reich, 2003).
Cand-5 therapy
In August 2004, Acuity Pharmaceuticals filed an investigational new drug application (IND) with the FDA to begin phase I trials for Cand-5 (Acuity Pharmaceuticals, Philadelphia, PA), an siRNA directed against VEGF (Acuity Pharmaceuticals, August 2004.). These trials were initiated for the treatment of exudative ARMD in September 2004 (Acuity Pharmaceuticals, October 2004).
Sirna-027 therapy
Sirna-027 (Sirna Therapeutics, San Francisco, CA), is a modified siRNA that specifically targets VEGF receptor I, a component of the angiogenic pathway found on endothelial cells. In September 2004, an IND was filed with the FDA for the treatment of exudative ARMD (Sirna Therapeutics Inc, September 2004). A phase I open-label, dose-escalation study enrolling up to 30 subjects to receive intravitreal Sirna-027 was initiated in November 2004, with preliminary results expected in mid-to-late 2005 (Sirna Therapeutics Inc, November 2004).
Pigment epithelium-derived factor inducer
Researchers have attempted to stimulate intravitreal production of native pigment epithelium–derived factor (PEDF), a naturally occurring potent antiangiogenic protein deficient in eyes with choroidal neovascularization (Holekamp, 2002) by using gene therapy (Takita, 2003). PEDF inhibits angiogenesis by inducing apoptotic death of endothelial cells stimulated to form new vessels (Stellmach, 2001). In a laser-induced murine model, choroidal neovascularization was reduced after intravitreal PEDF was produced from an adenoviral vector (Mori, 2001). One study demonstrated that increased intravitreal PEDF has shown up to 85% inhibition of neovascularization in models of laser-induced choroidal neovascularization, transgenic VEGF, and retinopathy of prematurity models (Rasmussen, 2001).
GenVec, Inc (Gaithersburg, MD) developed a PEDF-producing adenovirus vector called pigment epithelium-derived factor on an adenovirus vector (AdPEDF). A phase I dose-escalation study in 28 subjects with severe ARMD demonstrated safety and suggested efficacy of intravitreal AdPEDF. A new phase I study is planned for 20 subjects with ARMD less severe than that of the previous patients to compare the safety and efficacy of 2 doses (GenVec, 2005).
Squalamine
Squalamine lactate (Evison; Genaera Corporation, Plymouth Meeting, PA) is an antiangiogenic aminosterol that was originally found in the body tissues of the cancer-resistant dogfish shark. Squalamine lactate inhibits signaling of growth factors, including that related to VEGF, integrin expression, and cytoskeletal formation. Systemic intravenous administration has inhibited iris neovascularization in primate models (Genaidy, 2002), oxygen-induced retinopathy in murine models (Higgins, 2000), and laser-induced choroidal neovascularization in a rat model (Ciulla, 2003).
A phase I-II trial of 40 subjects in Mexico revealed that once-weekly intravenous injections of squalamine for 4 consecutive weeks preserved vision in 100% and improved visual acuity by 3 lines or more in 26% at 4 months (Genaera Corporation, 2003).
Three phase II clinical trials are currently underway. The largest, MSI-1256F-209, is a 100 patient prospective, randomized, controlled trial of the effects of 20 or 40 mg given intravenously every week for 4 weeks followed by maintenance every 4 weeks for 48 weeks and 12 months of observation for exudative ARMD. Investigators completed enrollment for this trial in June 2005.
The second trial, MSI-1256F-208, is a 45-patient prospective controlled trial of intravenous squalamine 10, 20, or 40 mg given initially in combination with verteporfin PDT and then alone for an additional 6 months followed by 12 months of observation for exudative ARMD. Enrollment is closed.
The final trial, MSI-1256F-207, is an 18-patient, open-label, parallel-group trial comparing 3 doses of intravenous squalamine given weekly for 4 weeks followed by 4 months of follow-up on exudative ARMD. Enrollment is also closed (Genaera Corporation, June 8, 2005).
Two phase III trials are scheduled to begin in June 2005 and run concurrently with the phase II trials. Investigators will evaluate the effects intravenous squalamine of 20 and 40 mg versus placebo dosed weekly for 4 weeks followed by a maintenance dose every 4 weeks for a total of 104 weeks in all forms of exudative ARMD (Genaera Corporation, June 27, 2005).
Microstructure modulators
Combretastatin A4-phosphate prodrug
Combretastatin A4-phosphate Prodrug (CA4P; Oxigene Inc, Watertown, MA) is an analog of colchicine that binds tubulin, an intracellular structural protein necessary for cell division. It was originally derived from the root bark of the South African willow tree Combretum caffrum.
Murine models of VEGF overexpression and laser-induced CNV demonstrated that CA4P was effective in preventing and treating CNV (Nambu, 2003). A phase I-II trial of 20 patients to study the safety and efficacy of intravenous CA4P administered once a week for 4 weeks with 6-month follow-up recently concluded. Oxigene has since announced a shift in their efforts to develop either a topical or periocular injection as a means of CA4P delivery (Oxigene, 2005).
Steroid compounds
Researchers report that corticosteroid compounds possess angiostatic properties by altering degradation of the extracellular matrix (Folkman, 1987) and by inhibiting inflammatory cells, which invariably participate in neovascular responses (Ohkuma, 1983). Researchers now favor intravitreal administration of corticosteroids because the blood-ocular barrier is bypassed, improved consistency of therapeutic steroid levels is achieved, and systemic adverse effects are minimized. These injections demonstrated efficacy in subretinal (Ishibashi, 1985; Ciulla, 2001) and preretinal (Danis, 1996) neovascularization in animal models.
Triamcinolone acetonide
Uncontrolled pilot studies of choroidal neovascularization in ARMD have involved the off-label use of intravitreally administered triamcinolone acetonide (Kenalog; Bristol-Myers Squibb, New York, NY) because of the drug's long half-life and corticosteroid properties. One study of 30 eyes receiving a single triamcinolone acetonide injection showed that 11 experienced improved or stabilized vision within 1-3 months of treatment, with regression of the choroidal neovascularization to inactive fibrosis. Outcomes were similar in 15 eyes except for slow extension and exudation from recurrent choroidal neovascularization, whereas 4 had no obvious treatment benefit (Penfold, 1995).
In later, studies, data showed a favorable effect on the course of the disease over follow up of 6 (Danis, 2000), 12 (Ranson, 2002), and 18 (Challa, 1998) months. However, the lack of control subjects complicated the researchers' ability to assess treatment efficacy versus the natural course of the disease. The authors speculated that intravitreal triamcinolone had a beneficial effect on ARMD-related CNV by inhibiting leukocytes, including macrophages, which normally release angiogenic factors (Penfold, 1995; Danis, 2000; Challa, 1998).
A randomized, double-masked, placebo-controlled clinical trial of 151 eyes receiving a single 4-mg injection of intravitreal triamcinolone showed significant antiangiogenic effects at 3 months after treatment. However, no improvement visual acuity was seen at 1 year. The authors speculated that triamcinolone might be effective at a higher or more sustained dose than that studied or in combination with other modalities (Gillies, 2003).
Other investigators have consequently evaluated high doses and combination therapy. A prospective, comparative, nonrandomized study of 187 subjects showed that the difference in visual acuity between subjects receiving intravitreous triamcinolone acetonide 25 mg and no treatment was significant (P) = 0.001) at 1 and 3 months (Jonas, 2004). In addition, a study of 26 patients receiving combination intravitreous triamcinolone and verteporfin therapy demonstrated a significant improvement in visual acuity over baseline at 6 months for newly treated choroidal neovascularization (PP = 0.007) (Spaide, 2003).
Visagen (Regenera Limited, Nedlands, Australia) has announced the development of a triamcinolone acetonide formulation used strictly for intraocular applications. The company anticipates sponsoring clinical trials in to formally gain approval for several ophthalmic indications in the near future (Trudinger, 2004). Other researchers are developing a preservative-free formulation that theoretically decreases the 0.8% sterile endophthalmitis rate observed with traditional intravitreal triamcinolone acetonide injections (Moshfeghi, 2003; Heriot, 2004).
Although many systemic adverse effects are avoided by using intravitreous corticosteroid injections, potential vision-threatening side effects, including an increased risk of infectious and sterile endophthalmitis, ocular hypertension, and progression of cataracts, can occur. However, intraocular pressure is often responsive to topical medication and usually needs no medication 3-6 months after the injection (Jonas and Hugger, 2003; Jonas and Degenring, 2003). In addition, each pars-plana injection increases the risk of retinal detachment, vitreous hemorrhage and endophthalmitis. To minimize these risks, researchers have developed new steroid compounds to obtain the antiangiogenic effect without the adverse effects and complications of repeated intravitreal injections.
Anecortave acetate
In 1985, a class of steroid with minimal glucocorticoid and mineralocorticoid activity was developed (Crum, 1995) and is now undergoing evaluation in human trials as anecortave acetate (Retaane; Alcon Laboratories, Inc, Fort Worth, TX). The lack of corticosteroid activity (Clark, 1997; McNatt, 1999) minimizes common elevations in intraocular pressure and accelerated cataract formation.
In addition, anecortave acetate was formulated for injection into the sub-Tenon space with a specially designed cannula (Anecortave Acetate Clinical Study Group, 2003). Findings from early animal models suggested that administration of this compound inhibited growth of a highly vascularized intraocular tumor in a murine model (Clark, 1999) and retinal neovascularization in a retinopathy of prematurity rat model (Penn, 2001).
A study of 128 patients in a phase II-III randomized, prospective, placebo-controlled trial, designed to evaluate the clinical safety and efficacy of juxtascleral injection of anecortave acetate versus placebo for the treatment of subfoveal choroidal neovascularization showed that baseline vision (P = 0.01), stabilization of vision (P = 0.03), and prevention of severe vision loss (P = 0.02) was statistically superior to baseline at 12 months. However, the dropout rate was nearly 50% (Anecortave Acetate Clinical Study Group, 2003). Investigators announced 12-month data from a 530-subject, phase III, 2-year, randomized, prospective clinical study in which they directly compared anecortave acetate with verteporfin PDT. No statistical difference was found between the 2 modalities in subjects with less than 3 lines of vision loss (Regillo, 2004).
Alcon received an approvable letter from the FDA in May 2005 and is currently awaiting further direction from the FDA before seeking final approval (Alcon, 2005).
Implantable corticosteroids
Because intraocular corticosteroids have shown antiangiogenic effects with repeated intravitreal administration, the development of sustained release intraocular implants to achieve near-constant intraocular steroid concentrations without repeated injections was pursued. In 1 study in rats, triamcinolone acetate microimplants inhibited laser-induced CNV (Ciulla, 2003) Furthermore, researchers at Bausch & Lomb (Rochester, NY) and Control Delivery Systems (Watertown, MA) have developed Retisert, a nonbiodegradable intravitreal implant that releases fluocinolone acetonide for up to 3 years to treat posterior uveitis.
Early phase III studies involving diabetic macular edema showed that 58.5% of subjects receiving the 0.5-mg implant developed serious adverse effects, such increased intraocular pressure, vitreal hemorrhage, and cataracts at 1 year compared with 10.7% of the standard-care group (Control Delivery Systems, 2003). Another study of 14 patients receiving high-dose fluocinolone acetonide implants for non–age-related subfoveal choroidal neovascularization revealed a variety of complications, including elevated intraocular pressure (14 of patients), cataract progression (14/ patients), and nonischemic central retinal vein occlusion (4 patients) (Holekamp, 2005). Therefore an ARMD indication has not been actively pursued.
A similar biodegradable dexamethasone implant (Posurdex; Allergan Inc, Irvine, CA, and Sanwa Kagaku Kenkyusho Co, Ltd, Nagoya, Japan) was safety and efficacy in phase II trials for the treatment of macular edema due to diabetes mellitus, occlusion of a branch or central retinal vein, uveitis, or surgery (Oculex). However, no trials are currently under way to evaluate the dexamethasone implant in ARMD.
Radiation therapy
Because CNV membranes are composed of rapidly proliferating pathologic endothelial cells, the membranes may be sensitive to means of inhibiting rapid cell division, such as radiation therapy. Given the apparent dose-response effect, some groups have delivered ionizing radiation to the macula by using modalities that may limit the exposure of ionizing radiation to normal radiosensitive structures of the eye, such as the optic nerve or lens. These methods have included stereotactic external photon beam irradiation of the posterior pole; brachytherapy, in which radioactive plaques are sutured to the posterior pole of the eye and explanted several days later (Finger, 1996 and 1999); and proton-bean irradiation, which deposits almost all of its energy at the desired depth in the eye at a point called the Bragg peak and undergoes little scattering (Yonemoto, 1996; Ciulla, 2002).
Although data from early pilot studies suggested a possible benefit, later reports were conflicting regarding the efficacy of radiation therapy for exudative ARMD. Data from more recent studies, which included nonrandomized or historic controls, suggested no beneficial effect of low-dose radiation therapy for exudative ARMD (Spaide, 1997 and 1998; Prettenhoefer, 1998; Rosen, 1998). However, findings from other small randomized (Anders, 1998; Char, 1999) and non-nonrandomized (Sasai, 1997; Krott, 1998; Mauget-Faysse, 1998; Star, 1999; Subasi, 1999) trials suggested that radiotherapy was beneficial, particularly with classic CNV (Tholen, 1998). In addition, data from a small, uncontrolled prospective trial suggested that irradiation can stabilize occult subfoveal CNV (Donati, 1999), whereas another uncontrolled trial suggested no benefit (Weinberger, 1999).
Controlled studies have demonstrated mixed results. Two prospective controlled studies involving a relatively many fractions of low ionizing radiation failed to demonstrate a treatment benefit for external-bean radiation (Radiation Therapy for Age-related Macular Degeneration Study Group, 1999; Marcus, 2001). However, 2 smaller prospective controlled studies of fewer high-radiation fractions demonstrated statistically significant vision improvement in patients compared with control subjects (Bergink, 1998; Char, 1999). Because of the positive outcomes, a prospective, controlled, pilot study to evaluate external-beam radiation on CNV in a small number of high-energy fractions was sponsored by the National Eye Institute. An interim analysis of findings, known as the ARMD Radiotherapy Trial (ARMDRT), showed that 43% of radiated eyes and 50% of nonradiated eyes had moderate visual loss at 12 months' follow-up (P = 0.60) (Marcus, 2003).
Theragenics Corporation (Buford, GA) is currently pursuing radiation therapy by using the palladium-103 isotope in their Therasight Ocular Brachytherapy System. A pilot study, which started in October 2004, was designed to evaluate the safety and efficacy of brachytherapy in an estimated 30 subjects with exudative ARMD (Theragenics Corporation, 2004).
Surgical Care: Vitreoretinal surgeons have attempted to remove CNV membranes with direct surgical excision of the CNV complex. However, results were disappointing for exudative ARMD. Researchers speculate that the CNV membranes in ARMD grow both anterior and posterior to the RPE. The damaged RPE that remains after removal of the CNV membranes causes atrophy of the underlying choriocapillaris, leading to neural retinal disorganization (Lambert, 1992; Mandelcorn, 1993; Heimann, 1994; Ormerod, 1994; Hudson, 1995; Del Priore, 1996).
In 1998, the National Eye Institute of the National Institutes of Health awarded funding to the Submacular Surgery Trial (SST). This study was designed as a randomized, multicenter, prospective clinical comparison of surgery versus observation to specifically evaluate patients with large or poorly demarcated and new subfoveal CNV, submacular hemorrhage from CNV associated with exudative ARMD, or subfoveal CNV due to presumed ocular histoplasmosis (POHS) or idiopathic causes. Patients were followed up for 2 years and assessed for stabilization or deterioration of their visual acuity, a change in contrast sensitivity, cataract development, surgical complications, and quality of life. The trials did not o show any benefit of submacular surgery over observation (Bressler, 2004; Hawkins, 2004).
In addition, other surgeons have suggested several experimental techniques to treat CNV; however, these approaches have not been tested in large randomized controlled trials. A novel approach in the treatment of subfoveal CNV includes macular translocation, in which the retina is shifted away from the underlying subfoveal CNV membrane. This procedure, performed only in small pilot studies, involves substantial risks and requires surgical refinement before it is adopted in large studies (Machemer, 1993).
Macular translocation led to the development of limited translocation. In this procedure, pars plana vitrectomy is followed by detachment of the temporal retina through 1 or more retinotomies and then reattachment of the retina after the sclera is surgically shortened (de Juan, 1998; Lewis, 1999).
In addition, several groups are investigating the possibility of transplanting RPE after surgical excision of the CNV membranes in ARMD because RPE transplantation might theoretically facilitate the repair of RPE defects that occur after CNV excision. However, whether functional repair occurs is unclear because only a few people have undergone this procedure in pilot studies (Algvere, 1994; Gouras, 1996; Little, 1996; Algvere and Algvere and Berglin, 1997).
