The mission to develop therapies for non-hematopoietic indications through the use of somatic tissue derived stem and progenitor cells was conceived in 1994 by pioneers in the fields of stem cell biology and neurobiology, Drs. Irving Weissman (Stanford), Fred (Rusty) Gage (Salk Institute & UCSD), and David Anderson (Caltech). Their vision to identify and purify tissue stem cells preceeded the discovery of the human embryonic stem cell (1998) and induced pluripotent stem cells (2007). The Company’s founders developed the concept of identifying other non-hematopoietic tissue stem cells for use in regenerative medicine based upon two key areas of scientific discovery. First, the identification and purification of the mouse and human hematopoietic stem cells, which at the single cell level could self-renew, produce all blood cell lineages, and engraft long-term in vivo. The second scientific discovery was the existence of neural stem cells in rodents and man. A convergence of these ideals led to the concepts that somatic ‘adult’ stem cells, such as the human neural stem cell, could be purified, expanded in culture, and could provide the cells for treating, and possibly curing a spectrum of human diseases.
The concept of stem cells for regenerative medicine has its foundation in the pioneering works by E. Donnall Thomas (Ferrebee & Thomas, 1960) using bone marrow transplantation for hematopoietic rescue. Over the next fifty years, the co-evolution of scientific discovery with computers and fluorescence based cell sorting (Loken et al. 1988) laid the ground work for the identification, isolation, and characterization of the mouse (Spangrude et al, 1988) and human (Baum et al., 1992) hematopoietic stem cells and in 1998 the human embryonic stem cells (ESC) (Thomson et al., 1998) and more recently, reprogrammed induced pluripotent stem cells (iPS) (Takahashi et al., 2009).
Retinal diseases are viewed as an optimal target for cell transplantation approaches because of ease of access, out-patient surgical procedure, the size of the eye, and the availability of non-invasive tests for visual function assessment following cell transplantation.

Age-related macular degeneration (AMD) is the major cause of visual impairment in Western countries and is the leading form of blindness in Americans over the age of 50 years. Prevalence of AMD increases with age, and it is estimated that up to one-third of individuals age 75 and older have some form of AMD (Zarbin, 2004). In the United States (US) alone, more than 8 million people carry the diagnosis (Jager, 2008) and the projected prevalence is expected to increase 50% by the year 2020 (Friedman, 2004). In China the prevalence of AMD ranges from 5.65 to 15.6% in different provinces and a ratio of 5.3:1 is observed for dry/wet AMD (Penfold, 2005).   To date, no curative therapy exists for AMD, and the condition represents a significant, unmet medical need.

AMD may be classified into early, intermediate, and advanced stages (Age-Related Eye Disease Study Report Number 3, 2000). Advanced AMD can be further classified as non-neovascular (dry, atrophic, non-exudative, non-neovascular) or neovascular (wet or exudative).

The primary pathology in AMD involves progressive deterioration of the retinal pigmented epithelium (RPE), Bruch’s membrane, and the choriocapillaris–choroid complex, which ultimately leads to loss of photoreceptor cell support (Leibowitz, 1980; Ferris, 1984; Bressler, 1988; Vingerling 1995). The RPE performs a range of critical functions important to photoreceptor maintenance including phagocytosis of outer segments, delivery of nutrients, and support of the visual pigment cycling (Zeiss, 2010).

In later stages of the disease, patchy degeneration of the photoreceptors, RPE, and underlying choriocapillaris is seen in the macula, the region of sharpest visual acuity. This degeneration eventually develops into clinically recognizable retinal areas referred to as GA. As the pathology of AMD evolves, patients experience loss of central vision that slowly progresses to blindness. The onset of GA is strongly correlated with predictable loss of visual acuity over the ensuing one to two years (Sunness, 2007).

Early clinical changes in AMD may be limited to mild visual loss or the incidental finding of drusen on ocular examination. However, as the disease progresses, patients may experience blurred vision, visual scotomas, decreased contrast sensitivity, abnormal dark adaptation, and difficulty with small print. Further insidious visual loss in patients with non-neovascular AMD occurs over months to years and can manifest as central or pericentral visual scotomas.

GA is associated with thinning of the RPE, and overlying atrophic changes in the macula result in the loss of photoreceptors (Suness, 1999). Natural history studies of GA have characterized the expected rate of visual acuity loss as a function of baseline visual acuity, and reported that a higher rate of acuity loss was associated with best-corrected visual acuity (BCVA) of > 20/50; 41% of subjects losing greater than three lines of vision at two years and 70% at four years (Sarks, 1988; Swaroop, 2007). Researchers have acknowledged that clinical trials including patient populations with this range of visual acuity will maximize the likelihood of detecting progression of disease and treatment effects (Swaroop, 2007).

Although significant progress has been made in developing novel treatments for patients with neovascular AMD, no proven treatment exists for the advanced non-neovascular or GA form of AMD. Lifestyle and dietary modifications are recommended once early AMD is recognized; including cessation of tobacco use, increased intake of antioxidants, control of hypertension, and reduction in body-mass index, but these adjustments typically have minimal impact on the disease course.

Retinal disorders are considered an appropriate target for cell therapy, and the combination of unique anatomic and clinical features render the eye particularly favorable for clinical testing.

Photoreceptor death is the final irreversible event in many blinding diseases, and even a modest reduction in the rate of photoreceptor loss by a neuroprotective approach may lead to significant prolongation of useful vision. This strategy may be particularly relevant in the setting of the rate of changes observed in the natural history of AMD (Luther, 1982). Transplants to slow the loss of existing neurons (i.e., photoreceptors) in the degenerating retina may be the first useful method for maintaining some degree of vision (Vugler, 2007). In many ways, the eye is an ideal organ in which to assess transplant success because objective techniques are available to monitor visual function and anatomy.

The unique attributes of the eye particularly facilitative of testing cell-based neuroprotective interventions are as follows:

• Retinal degeneration is characterized by discreet and small anatomic regions of disease eminently scalable for rescue by cell transplantation.
• The subretinal space is relatively immune privileged and consequently is potentially permissive to long-term allograft survival.
• Subretinal surgery is a minimally invasive vitreo-retinal technique and can be performed as an out-patient procedure.
• Measures of visual acuity, direct fundus examination, retinal electrophysiology, and in vivo retinal imaging offer objective, noninvasive methods of retinal function and outcomes.
• Outcomes from the treatment eye can be compared with outcomes from the untreated fellow eye. 
The advantages of cell-based neuroprotective approaches for GA AMD are:
• GA AMD has no curative therapy and represents an area of significant unmet need
• The final common pathway of vision deterioration in GA AMD is photoreceptor loss
• The rate of progression in GA AMD affords the prospect of early intervention and subsequent rescue of degenerating photoreceptors and possible slowing or stabilization of the rate of vision loss
• Cell-based approaches targeting neuroprotection of photoreceptors do not require the complex phenotype recapitulation, integration, and attachment expected by photoreceptor and RPE replacement strategies
• Neurotrophic factors potentially delivered by cell transplantation circumvent the obstacle posed by the blood-retinal barrier, which impedes access of intermittent systemically administered large molecules to the neural retina
• Proof of concept for neuroprotection of host photoreceptors by subretinal transplantation using various tissue and cells has been established in non-clinical studies
• Previous clinical trials have demonstrated the feasibility and procedural safety for allogeneic subretinal transplantation in subjects with retinal degeneration

Based on the results of non-clinical studies in the dystrophic RCS rat model (Section 4.3), subretinal transplantation of hNSC represents a candidate neuroprotection strategy to prevent or slow photoreceptor loss secondary to GA AMD

The goal of this clinical research is to develop a safe and effective therapy for GA AMD. A successful therapy will either ultimately stabilize or slow the progression of visual loss, thereby reducing the physical, emotional, and social burdens related to this major blinding disorder.  An approach that protects or rescues retinal cells represents an inherently more feasible cell transplantation strategy and arguably an equally worthy regenerative goal to that of replacement. Indeed, the theme of retinal protection can be found in the rationale for administration of growth/trophic factors, dietary supplementation with vitamin A, protection from damaging effects of light, and historical retinal transplant trials (Petrukhin, 2007; Jager, 2008). The underlying strategy of this latter approach is not to specifically replace the photoreceptor cells, but to protect and stabilize the threatened rods and cones (Milam, 1993).

A treatment strategy based on protecting photoreceptor cells by providing trophic and other physiological pathways of support is attractive for several reasons. A retinal-protective strategy does not have the complexity of approaches that require replacement, integration, and survival of a specific photoreceptor cell. It has also been shown that downstream retinal neural circuitry remains intact despite loss of photoreceptor function, so the underlying pathways for signal transmission from the retina are in place if marginal photoreceptors can be stabilized or enhanced (Bergmann, 2004; Suter, 2007). It is conceivable that the loss of the supportive function for the neural retina in GA AMD can be offset by the biological properties of an exogenous cell, such as hNSC.

hNSC cells have been shown to be effective in photoreceptor and vision rescue in a rodent model of retinal degeneration, the RCS rat.  With the exception of nonhuman primates, animals do not have a structure equivalent to the macula, thus no animal model can fully recapitulate all the pathological hallmarks of AMD. However, the RCS rat does demonstrate the consequences of dysfunction in the phagocytosis pathway of the RPE cells and the resulting secondary loss of photoreceptors, which are both common pathological elements in many blinding diseases including AMD. Therefore, insights into the potential therapeutic value of an intervention can be derived from rodent models that replicate the main disease elements. The progressive photoreceptor loss in the RCS rat permits both histological and behavioral assessments of experimental neuroprotective treatments that may also affect the disease course in AMD and other blinding diseases, all of which share photoreceptor death as the final common pathway.

Upon transplantation into the subretinal space at postnatal day 21 (P21) in the RCS rat model, hNSC cells protect host photoreceptors and preserve visual function. Cone photoreceptor density was found to remain constant over time, consistent with the sustained visual acuity and luminance sensitivity functional outcomes. hNSC cell behavior and fate have also illustrated the following: 1) radial migration from the injection site in the subretinal space, 2) survival throughout the seven-month experiment with maintenance of an immature phenotype at the latest time point, and 3) modest cell division, with no evidence of uncontrolled growth or tumor-like formation.

A neuroprotective strategy to prevent progression of vision loss relies upon the preservation of existing photoreceptors at a time prior to irreversible damage. In the RCS rat model, P21 correlates to a time point at which maximal protection of photoreceptor and preservation of vision can be assessed. Hence, P21 transplantation in the RCS rat was chosen to assess efficacy based on previous studies showing that even though rod number and morphology in the P21 rats are virtually indistinguishable from age-matched normal rats, rod function by physiological recording (electroretinogram [ERG]) is already severely compromised. Time points beyond P21 in this model are associated with rapid and progressive photoreceptor loss to the degree that the outcome of protective strategies cannot be as reliably measured. Thus, hNSC transplantation at the P21 stage of photoreceptor degeneration will support clinical evaluation across a broad spectrum of disease severity in humans including early stages of vision pathology.

The collective preclinical data strongly suggest that subretinal hNSC transplantation may result in the preservation of host photoreceptors by adopting one or more of the underlying RPE functions critical for maintenance of vision including phagocytosis of outer segments and delivery of soluble neurotrophic factors, both of which are deficient in deteriorating RPE. Given these findings, hNSC cells appear to be well suited for cell-based therapy in GA AMD.

Since RPE loss is neither global nor complete in GA AMD, particularly in earlier stages, sufficient residual RPE may exist for support of the visual cycling pathway (conversion of all trans-retinol to 11-cis-retinal). A cone-specific visual cycle involving the Müller cell has also been recently identified and could serve as an important pathway for cone chromophore cycling independent of RPE status (Wang, 2011). Thus, it seems reasonable that donor cells need not express the entire spectrum of RPE activity to exert a clinical benefit on vision and it is also possible that donor cells could have a direct or an indirect beneficial effect on the host RPE that also contributes to protection of photoreceptor integrity (i.e., antioxidant or immunomodulatory activity).

The natural history of GA AMD has been studied and the range of visual decline based on BCVA is generally known (Suness, 1999; Suness, 2007). The onset of GA in the setting of AMD represents a disease stage in which the subsequent clinical progression of visual deterioration can be predicted, but sufficient photoreceptors, visual function, and RPE remain amenable to rescue (Suness, 2007). Transplantation of hNSC cells into the subretinal space could lead to neuroprotection of vulnerable photoreceptors and produce a therapeutic effect over a measurable portion of the macula and fovea.