A stem cell journey in ophthalmology: From the bench to the clinic

Abstract Debilitating diseases of the eye represent a large unmet medical need potentially addressable with stem cell‐based approaches. Over the past decade, the California Institute for Regenerative Medicine (CIRM) has funded and supported the translation, from early research concepts to human trials, of therapeutic stem cell approaches for dry age‐related macular degeneration, retinitis pigmentosa, and limbal stem cell deficiency. This article chronicles CIRM's journey in the ophthalmology field and discusses some key challenges and questions that were addressed along the way as well as questions that remain.


| INTRODUCTION
Limbal stem cell deficiency (LSCD) is a rare, progressive corneal disease that is found throughout the world, has wide-ranging etiology, and can culminate in blindness. Surgical tissue transplantation is the main treatment option, but is often unsuccessful.
Developing treatments for debilitating and blinding eye diseases was a relatively early focus of regenerative medicine research. The accessibility and small size of the eye are an advantage for cell therapy manufacture and delivery. These distinguishing features, combined with the many sophisticated ophthalmic imaging tools for assessing and monitoring clinical outcomes, make the eye an attractive target for evaluating stem cell-based therapies. Beginning in 2008, the California Institute for Regenerative Medicine (CIRM), together with its grantees, pioneered the development of cell-based approaches for severe eye diseases including diseases of the retina such as dry AMD and RP as well as the corneal disease, LSCD. This article briefly describes the progress made from the bench to early clinical trials, beginning in the trail blazing years of the now burgeoning field of regenerative medicine. We highlight initial questions and what has been learned along the way as well as remaining questions and potential challenges as these programs progress in clinical development.  [1][2][3] This early research relied on the spontaneous differentiation of hESC into RPE when cultured, a process that is slow and yields a low percentage of RPE. 4 The observation that some hESC lines exhibited more robust RPE differentiation capacity than others, producing higher frequencies of pigmented RPE cells, highlighted the importance of selecting an optimal, well characterized starting cell line before beginning translational activities.

| AGE-RELATED MACULAR DEGENERATION
A key question was how to deliver stem cell-derived RPE to the retina. What devices or tools could be used? Would the cells survive and persist? For RPE to perform their multiple essential functions, which include interacting with and supporting the PR and participating in the visual cycle, 5 investigators reasoned that the transplanted RPE cells would need to mimic endogenous RPE and form a polarized monolayer in contact with both the underlying Bruch's membrane on which RPE cells depend for survival and the overlying PR that they support. In advanced AMD, known as geographic atrophy (GA), the Bruch's membrane is damaged or dysfunctional, raising concerns about the attachment and survival of transplanted RPE cells and their ability to function in the retina. 6 Mark Humayun, an early CIRM awardee and cofounder of the California Project to Cure Blindness (CPCB), addressed this concern by designing an implant composed of a polarized monolayer of hESCderived RPE (hESC-RPE) on an ultrathin, synthetic parylene substrate designed to both provide a surface for RPE adhesion and to mimic the diffusion barrier characteristics of Bruch's membrane. In a comparative study in rats, the survival of hESC-RPE after subretinal implantation was significantly improved when RPE cells were delivered as a polarized monolayer on parylene compared to as a cell suspension. 7 With CIRM funding, the CPCB developed good manufacturing practice (GMP) procedures to produce transplantable hESC-derived RPE and manufacture the composite implant. They also developed and validated tools and procedures to surgically insert it into human retinas, patients improved their ability to visually fixate, suggesting that the implanted hESC-RPE monolayer supports visual function in the overlying, previously nonfunctional retina. 9 The investigators speculate that these improvements may be due to revival of dormant PR in the area of GA, resulting from direct integration with the hESC-RPE.
Importantly, there were no safety concerns. Patients in the study received immunosuppressive therapy for approximately 2 months postimplantation. Analysis of OCT images indicated continued presence of the implanted hESC-RPE and integration with the recipient retinal tissue. In addition, the observed functional improvements were maintained for at least 120 days at the time of reporting, suggesting persistence of the implanted RPE after immunosuppression was stopped.

| Alternative approaches for dry AMD
A number of alternative cell-therapy approaches are being investigated for dry AMD using different cell types or different starting sources of cells to derive RPE and alternative/no scaffold to deliver the cells ( Table 1). All of these approaches are in early stages of clinical development. All but one use allogeneic cells while one approach employs autologous iPSC-derived RPE. It remains to be seen whether an autologous approach can adequately support a disease with a large patient population such as AMD and whether the potential advantages of an autologous approach (reduced/no immunogenicity) will outweigh the challenges and cost of manufacturing.

| RETINITIS PIGMENTOSA
CIRM began funding development of a stem cell therapy for RP in 2008. Briefly, RP is an inherited degenerative disease of the PRs, typically with earlier onset than AMD. The literature documents that RP can be caused by more than one hundred different mutations in multiple genes, many of them rod-specific, making it challenging to address with a gene therapy approach. RP is characterized initially by loss of the rod PRs that are located in the peripheral retina and are responsible for low-light vision. Clinically, this anatomical loss of rod PR manifests as loss of peripheral vision, diminished ability to see in low light conditions and difficulty seeing at night. Subsequently, because the more important cone PR that enable high-resolution color vision rely upon cone survival factors released by the rods, loss of rods leads to loss of cones, eventually culminating in total blindness. Cellreplacement of lost PR would require re-establishing synapses and neural connections, events that are likely hard to accomplish. More feasible treatment strategies, at least in the near term, tend to focus on preserving cone PRs, either directly or by preserving the rods. Such strategies could in principle arrest disease progression, which would have a major impact on patients, particularly in earlier stages of disease. Beginning in 2011, CIRM has supported the translation and development of two novel therapeutic approaches for RP.     16 and the retina has been hypothesized to be an immuneprivileged site capable of retaining an allogeneic cell graft without eliciting rejection. However, the RP disease process appears to compromise the BRB, [17][18][19] as evidenced by the frequent formation of macular edema 20,21 ; and in addition, subretinal injection breaches the barrier, albeit temporarily. Immunosuppression will therefore be employed in this phase 1 trial.

| Alternative approaches for RP
The landscape of regenerative medicine approaches for RP remains limited (

| Alternative approaches for LSCD
The main treatment option for LSCD is a surgical approach using either allogeneic or autologous tissue for the graft. Graft failure due to rejection is a common complication in the case of allografts.
Expanded autologous LSC (Holoclar) is approved for use in Europe but not in the United States (as indicated above). A number of alternative autologous approaches are in early stages of clinical development as well as an approach using cadaveric LSC (Table 3). Several regulatory challenges specific to cell-based therapies needed to be overcome. These included demonstrating that the final cell product does not contain residual undifferentiated, potentially tumorigenic cells and does not induce tumors in preclinical animal models. Additional challenges included developing a potency assay and demonstrating manufactured lot consistency when the cellular product's mechanism of action is complex and not well understood. In some cases, a specifically designed delivery tool needed to be developed and tested in large animal preclinical models.

| DISCUSSION
The optimal route of delivery of a cell therapy for a retinal disease will likely depend on the mechanism of action. For a therapy that exerts a neuroprotective effect via a paracrine mechanism, intravitreal delivery appears to be the preferred route, allowing for diffusible factors to broadly access the retina. Intravitreal delivery has several additional advantages in that it is minimally invasive, allows for repeat injection and is routinely used in outpatient settings, including in the optometrist's office. In contrast, retinal cell replacement strategies may necessitate subretinal delivery, requiring specific tools and surgical expertise that may best be deployed at centers of excellence. A CIRM supported study has demonstrated that subretinal delivery is feasible and that durable graft retention can be achieved. The potential for long-term clinical benefit thus could outweigh the risks.

The need for immunosuppression remains an open question.
Whether the eye is an immune privileged site remains to be deter-

| Future directions
The development of stem cell therapies for debilitating eye diseases progressed rapidly and measurably on several fronts during the past 5 years. Advancements include (a) the development of innovative tools and technologies that incorporate patients' QOL measures; (b) evolution of a regulatory paradigm that previously required a change of 13 letters on an eye chart in order to approve a therapy to acceptance of a mobility endpoint when appropriate; and (c) the advancement of gene therapy approaches for some rare eye diseases.
New therapeutic modalities and genomics will expand the field further.
As with all cell therapies, the development of a robust and scalable manufacturing process is critical to the final success of the therapy. The increasing adoption of expedited regulatory pathways, including regenerative medicine advanced therapy designation, has escalated the importance of defining critical quality attributes for each cell and gene product as early in the process as possible and of rapidly establishing a mature manufacturing process suitable for commercialization.