downregulates inßammatory cytokines. It can be administered subcutaneously and is in phase 2 and 3 studies in patients with drusen (NCT00466076). A small, randomized controlled study demonstrated efÞcacy after 12 weeks of subcutaneous injections [186]. It is not certain that drusen disappearance, the end point of this study, is an appropriate surrogate end point for long-term visual acuity preservation in AMD eyes. The CAPT trial demonstrated no long-term visual beneÞt to laser photocoagulation-induced drusen resorption [187]. The two treatment modalities, however, have different mechanisms of action. Laser treatment induces inßammation, and glatiramer acetate is anti-inßammatory.

Amyloid b oligomers are toxic to cells (soluble monomers are not). Amyloid diseases typically exhibit abundant Þbrils of various lengths that are an end product of stepwise protein/peptide misfolding. These Þbrils accumulate as extracellular deposits. Drusen vesicles probably contain Þbrillar amyloid composed in part of amyloid b, which may damage RPE cells and/or incite inßammation that contributes to AMD progression [188Ð190]. Anti-amyloid-b antibody blocks ocular pathology in a CFH-deÞcient mouse. After 3 months of prophylactic treatment with GSK1532968 (6F6), there was signiÞcant lowering of amyloid-b deposition at all dose levels and a signiÞcant lowering of activated complement C3 deposition at 300 and 600 mg doses [191]. RN6G (PF-4382923, PÞzer) is a humanized monoclonal antibody that targets the C-termini of amyloid b-40 and amyloid b-42. These peptides have been implicated in neurodegenerative diseases. Treatment with intravenous RN6G is intended to prevent the accumulation of amyloid b-40 and amyloid b-42 and to prevent their cytotoxic effects. A phase 1 clinical trial has been completed successfully (NCT00877032), and a phase 2 trial is underway for treatment of subjects with advanced nonexudative AMD. GSK93377 (GlaxoSmithKline) is a humanized monoclonal antibody directed against amyloid-b. It is administered intravenously, and a phase 2, multicenter, randomized, double-masked, placebocontrolled, parallel-group study in adult patients with GA due to AMD is in progress (NCT01342926). Patients will be treated monthly with placebo, 3 or 6 mg/kg GSK93377. The primary end point is the rate of change in GA area from baseline.

1.3.4Neurotrophic Agents

Neuroprotectants can rescue photoreceptors in preclinical models of retinal degeneration including light damage, glaucoma, and RP. The mechanism(s) by which neurotrophic factors promote retinal survival in these models is not established fully and may vary depending on the disease setting [192, 193].

Basic Þbroblast growth factor (bFGF), ciliary neurotrophic factor (CNTF), brainderived neurotrophic factor (BDNF), pigment epithelium-derived factor (PEDF), and interleukin-1 (IL-1) seem to provide the broadest degree of protection against photoreceptor degeneration [194Ð198]. Some of these molecules can be produced by RPE and/ or retinal cells [199Ð202]. The RPE and retina, for example, seem to produce FGF, and the retina produces CNTF [203]. Constitutive production of bFGF is probably important for photoreceptor survival normally [204, 205]. In addition, rod cells produce cone

survival factors [206, 207]. Heat shock proteins may also promote retinal survival in some paradigms of retinal degeneration, including light damage, ischemia, and RP [208Ð211]. Alpha-2 adrenergic agonists protect photoreceptors against light damage, probably due to speciÞc induction of bFGF expression in photoreceptors [212]. Sustained delivery of CNTF can slow photoreceptor degeneration in animal models of RP although it can be associated with side effects such as decreased electroretinogram amplitude [205, 213]. In animal models of RP, neurotrophins do not prevent photoreceptor death; they merely delay it. In addition, the duration and amplitude of pathway activation by a given neurotrophin receptor modulate the biological response [192].

Many of the neurotrophic factors that seem to provide the broadest degree of protection are ligands for two major families of membrane bound receptor tyrosine kinases: FGF receptors and Trk neurotrophin receptors [194, 195]. Ligand binding to FGF and CNTF receptors activates various enzymes, e.g., phosphatidyl inositol 3-kinase (PI3-K), mitogen-activated protein kinases (MEK, ERK), and Akt, which in turn can inhibit apoptosis (see Chaum [192] for references). In addition, phosphorylation of cAMP response element binding protein 1 (CREB1) and activating transcription factor 1 (ATF1) is an intrinsic response to photoreceptor injury arising from photoreceptor gene mutations and is found in AMD eyes [214]. CNTF induces CREB1/ATF1 phosphorylation in normal retinas and induces increased phosphorylated CREB1/ATF1 in canine retina with the rcd 1 mutation [214]. At least in some cases, the photoreceptor rescue effect of neurotrophic factors (e.g., BDNF, CNTF, and bFGF) may be mediated via Muller cells [199Ð202, 215Ð217] although a direct effect on photoreceptors is possible [218]. Nerve growth factor (NGF) and BDNF may inhibit cell death through multiple pathways, e.g., PI3-K activation and c-jun protein inhibition (see Chaum [192] for references). Gene therapy to modulate expression of components of the signaling pathways stimulated by neurotrophins (e.g., Akt, antiapoptotic genes, or heat shock protein) might be superior to treatment with neurotrophic factors themselves [192].

The pathophysiology of photoreceptor death associated with light damage, mechanical injury, and inherited retinal degeneration share similarities but also differ in important ways [148, 192, 193]. The relevance of light damage models and animal models of RP to AMD is not clear although mutations in ABCA4 have been associated with AMD in some, but not all, studies [219Ð221]. Furthermore, strategies that are protective in a given retinal degeneration model, e.g., Bcl-2 overexpression or oxidative stress reduction, may not be effective in another [222Ð225]. Finally, the therapeutic effect of an intervention may not only depend on the disease but also on the way the therapy is delivered. For example, the route of delivery (e.g., intravitreal injection vs. viral vector-mediated transfection vs. cell-based delivery system) and the steady state level seem to be important in determining the effectiveness of bFGF-mediated photoreceptor rescue in RP models [226Ð229]. Thus, one should contemplate strategies for neuroprotection in AMD based on the results of light damage experiments and animal models of RP with caution.

Currently, a brimonidine sustained release implant (brimonidine (alpha-2 adrenergic receptor agonist) formulated in the Allergan Novadur sustained release delivery system) and topical tandospirone (AL-8309B, Alcon, serotonin 1A receptor agonist) are in clinical trials (NCT00658619, NCT00890097, respectively) for GA based on their effectiveness in preventing retinal degeneration in preclinical light

Fig. 1.7 Effect of high-dose CNTF on the increase of retinal thickness of the right eye of an 85-year-old GA patient. Individual OCT images of the macula at baseline (top) and month 12 (bottom). The dark areas indicated by arrows show the outer nuclear layer. The images were both obtained as 7-mm long scans offset 5¡ from the horizontal, beginning at the midpoint of the temporal aspect of the optic nerve. Because these custom scans are oriented with respect to the midpoint of the temporal aspect of the optic nerve, rather than with the presumed foveal center, registration and foveal centration difÞculties inherent with time domain Stratus OCT are minimized. Accordingly, these images were obtained from similar, if not identical retinal locations. The variation in choroidal shadowing is not related to differing retinal scan locations; rather, it represents differences in thickness of the overlying layers after treatment with NT-501 and, possibly, to differences in overall image saturation. The qualitative widening of the outer layer complex was observed consistently among subjects (adapted with permission from Zhang et al. [232])

damage models. Serotonin 1A agonists are neuroprotective in animal models of excitotoxic neuronal damage [230]. Neuroprotection may arise from their hyperpolarizing effects on cells, mediated via G protein-coupled K+ channels, and/or stimulation of NGF release by neurons [231].

A randomized, double masked, sham control phase 2 study comparing high dose (20 ng/day, n = 27), low dose (5 ng/day, n = 12), and sham (n = 12) treatment with CNTF intravitreal implants (in one eye only) has been completed in patients with GA [232]. The CNTF Study (NCT00447954) utilized intravitreal implants of genetically modiÞed RPE that overexpress CNTF and are contained within a semipermeable capsule that has small pores, which permit CNTF to escape into the vitreous cavity and protect the allogeneic RPE cells from immune rejection. The primary end point was the change in best-corrected visual acuity at month 12. CNTF treatment resulted in a dose-dependent, statistically signiÞcant increase in retinal thickness (as measured with OCT) by month 4 (P < 0.001) (Fig. 1.7). The high-dose cohort had signiÞcantly greater retinal thickness than the low-dose cohort (P < 0.05). CNTF-induced increased retinal thickness has been observed in laboratory animals with RP-like conditions [213, 233]. In mice, this thickness change reßects, in part, increased photoreceptor nuclear size and increased amounts of euchromatin, and in rcd-1 dogs it reßects increased photoreceptor nuclear size as well as swelling of photoreceptors and/or Muller cell processes with expansion of the outer limiting membrane towards the RPE (Fig. 1.8).

The change in retinal thickness was followed by stabilization of vision (loss of less than 15 letters) in the high-dose cohort (96.3%) compared with low-dose