Considerations for Developing an Autologous Induced Pluripotent Stem Cell (iPSC)-Derived Retinal Pigment Epithelium (RPE) Replacement Therapy

Abstract

Cell-replacement therapies are a new class of treatments, which include induced pluripotent stem cell (iPSC)-derived tissues that aim to replace degenerated cells. iPSCs can potentially be used to generate any cell type of the body, making them a powerful tool for treating degenerative diseases. Cell replacement for retinal degenerative diseases is at the forefront of cell therapies, given the accessibility of the eye for surgical procedures and a huge unmet medical need for retinal degenerative diseases with no current treatment options. Clinical trials are ongoing in different parts of the world using stem cell–derived retinal pigment epithelium (RPE). This review focuses on scientific and regulatory considerations when developing an iPSC-derived RPE cell therapy from the development of a robust and efficient differentiation protocol to critical quality control assays for cell validation, the choice of an appropriate animal model for preclinical testing, and the regulatory aspects that dictate the final approval for proceeding to a first-in-human clinical trial.

The recent use of the induced pluripotent stem cell (iPSC) technology to generate different cell types has revolutionized the field of regenerative medicine (Takahashi et al. 2007; Yu et al. 2007; Hirschi et al. 2014). The problems associated with immune rejection, insufficient availability of replacement tissues, and how well the body can integrate a new allogeneic cell/tissue can be circumvented by using the patient's own iPSCs to generate new and healthy autologous tissues and, potentially, organs (Vonk et al. 2015; Madrid et al. 2021). Stem cell–based therapy is being used to treat retinal degenerative diseases like age-related macular degeneration (AMD) (Schwartz et al. 2012, 2015; Song et al. 2015; Da Cruz et al. 2018; Kashani et al. 2018), retinitis pigmentosa (RP) (Uy et al. 2013; Tuekprakhon et al. 2021), diabetic retinopathy, and Stargardt's retinal degeneration (Schwartz et al. 2015). Retinal pigment epithelium (RPE)-based cell therapy is one of the most advanced treatments for some of these diseases (Bharti et al. 2011; Monsarrat et al. 2016; Deinsberger et al. 2020).

RPE is a polarized monolayer present at the back of the eye, which performs many functions critical for healthy vision; some of these include nourishing the overlying photoreceptors and removing waste products of the visual cycle (Bharti et al. 2006). Healthy RPE cells are therefore responsible for good vision and retinal homeostasis. A patient's fibroblasts or peripheral blood mononuclear cells (PBMCs) can be isolated and reprogrammed into iPSCs. These cells are then differentiated to form RPE by different protocols, including spontaneous or directed differentiation (Osakada et al. 2009b; Bharti et al. 2011; Pennington et al. 2015; Ben M'barek et al. 2017; Sharma et al. 2019). Studying the developmental pathways of the RPE in vivo underlies which signaling pathways and associated growth factors are important for RPE differentiation in vitro. Differentiation of iPSCs toward RPE requires activation and inhibition of different pathways at different phases of the differentiation cycle (Song et al. 2015; Da Cruz et al. 2018; Sharma et al. 2022). Our laboratory has developed a triphasic differentiation protocol to generate RPE from iPSCs, in which media with different growth factors are used in all three phases (Sharma et al. 2022).

As per regulatory guidelines and for patient safety, the quality control (QC) criteria are to be stringent for RPE cells that are planned to be transplanted into a patient. As the differentiation process is very long and there are many factors influencing differentiation, it is necessary to monitor this process at every step. As part of good manufacturing practices (GMPs) and Food and Drug Administration (FDA) guidelines, a variety of tests must be performed at every stage of the differentiation process starting with the source material—the iPSCs—and ending with the final product—the RPE (Rao and Malik 2012; Baghbaderani et al. 2015; Jha et al. 2021). In addition to regularly monitoring sterility of cultures (these tests are not discussed in this review), QC tests include testing for iPSC and RPE purity and maturity markers by flow cytometry, measuring the trans-epithelial resistance of RPE tissue after they mature, and performing microscopic evaluation to further confirm RPE maturity (Schwartz et al. 2015; Mandai et al. 2017; Kashani et al. 2018; Sharma et al. 2019; Nishida et al. 2021). There are many regulatory guidelines that need to be followed when using an iPSC-RPE product as an investigational new drug (IND) (Jha et al. 2021). These are discussed in detail later in this review. Under Title 21 Code of Federal Regulations (CFR) guidelines, the iPSC-RPE has to be tested in animals to determine its safety and efficacy (Jha and Bharti 2015). It is vital to see whether, in preclinical studies, transplanted RPE cells are effective at mitigating the retinal degeneration in an efficacy model, and whether they have the potential to cause a tumor or any other adverse effect in a safety model (Jha et al. 2021).

RPE DEVELOPMENT AND DIFFERENTIATION

RPE has a central role in maintaining the health of the outer blood–retina barrier (oBRB) and its dysfunction and degeneration are associated with retinal degenerations (Bharti et al. 2014). This property makes RPE a key target for cell therapies—RPE replacement could halt or slow down retinal degenerations. To develop cell therapies for retinal diseases, RPE cells derived from iPSCs need to closely resemble the native cells, phenotypically and functionally. To this end, it is essential to know which signaling pathways act during the development of the human RPE to reproduce them in vitro during stem cell differentiation.

The vertebrate eye develops from an evagination of the eye field in the ventral forebrain to form the optic vesicle. The proximal region of the optic vesicle will become the optic stalk, the distal region will become neural retina, while the dorsal region will become the RPE. Once the evagination of the optic vesicle reaches the surface ectoderm, its distal part invaginates forming the optic cup. The tissues surrounding the optic cup influence specification of cell types. Four signaling pathways are particularly important: the transforming growth factor β (TGF-β) superfamily, WNT, the hedgehog (HH) family, and fibroblast growth factors (FGFs) (Amato et al. 2004; Yang 2004; Bharti et al. 2006). The mesenchyme surrounding the eye directly influences RPE development. After removal of the mesenchyme, the chick optic vesicle shows down-regulation of RPE markers and up-regulation of neuroretinal markers. ACTIVIN A, a member of the TGF-β superfamily, activates the TGF-β pathway and was shown to compensate for the missing mesenchyme ex vivo (Fuhrmann et al. 2000). A different branch of the TGF-β superfamily, the bone morphogenetic protein (BMP) pathway, also proved to be essential for RPE development (Müller et al. 2007). Bmp4 and Bmp7 are expressed in the surface ectoderm, the mesenchyme surrounding the optic vesicle and the presumptive RPE. Inhibiting BMP signaling, with the protein NOGGIN, prevents RPE development and induces expression of neuroretinal genes (Steinfeld et al. 2017). On the other hand, stimulating BMP signaling in the optic vesicle induces RPE development in the presumptive optic stalk and neuroretina (Müller et al. 2007). WNT activation is essential for RPE specification, since RPE commitment was blocked upon β-CATENIN genetic inactivation (Hägglund et al. 2013). The HH pathway is also thought to play a role in RPE development since proteins of this pathway have been detected in RPE (Neumann and Nuesslein-Volhard 2000; Stenkamp et al. 2000). Interestingly, inhibiting HH signaling in Xenopus leads to a reduction or loss of RPE markers (Perron et al. 2003). FGFs are negative regulators of early RPE development (Bharti et al. 2012). The surface ectoderm is the main source of FGFs for the eye primordium, specifically FGF1 and FGF2 (De Iongh and McAvoy 1993; Pittack et al. 1997). Removal of the surface ectoderm induces the transition of the presumptive neuroretina into RPE-like cells, while exposure of the presumptive RPE to FGF alters its development to neuroretina (Pittack et al. 1997; Hyer et al. 1998; Nguyen and Arnheiter 2000).

Several groups attempted in vitro RPE differentiation using spontaneous differentiation or developmentally guided protocols (Schwartz et al. 2012, 2015; Sharma et al. 2022). Spontaneous differentiation requires replacing iPSC culture medium with one that does not contain any specific growth factor. This process is inefficient since it requires 20–25 wk of culture for RPE differentiation and the yield of RPE cells is <10%. Developmentally guided or directed differentiation uses the knowledge from work done on embryonic RPE development by supplementing culture media with specific growth factors at the right time point (Osakada et al. 2009a; Sharma et al. 2022). This system improves efficiency, shortens differentiation time from 25 wk to 10 wk, and reduces iPSC line to line variability. The first step for RPE production is the induction of the anterior neuroepithelium and the eye field. The role of insulin-like growth factor (IGF) in directing differentiation toward the anterior neuroepithelium was first described in Xenopus, where ectopic expression of IGF mRNA led to the induction of ectopic eyes and head-like structures containing brain tissue (Pera et al. 2001). Efficient neural induction from stem cells can be achieved with dual inhibition of SMAD signaling (Chambers et al. 2009). In addition, WNT/β-CATENIN and BMP inhibitors were shown to be up-regulated during eye field specification (Meyer et al. 2009). To recreate this process in vitro, we supply IGF-1 in combination with dual SMAD inhibition (TGF-β, SB431452; BMP, LDN193189) and WNT inhibitor, CK1–7 (Fig. 1; Sharma et al. 2022). The second step is the induction of the RPE from the anterior neuroepithelium. We add an FGF inhibitor, PD0325901, to the neuroectoderm-induction media (Fig. 1; Sharma et al. 2022), since FGF removal is sufficient to convert the presumptive neuroretina to RPE (Pittack et al. 1997). The third step is the appearance of RPE from the eye field. At this point, activation of BMP, TGF-β, WNT, and HH pathways appear to be important in determining RPE fate, as described above. We found that supplying a medium containing ACTIVIN A that triggers TGF-β/SMAD pathway activation is sufficient to generate cells with a committed RPE phenotype (Sharma et al. 2022). Other studies enhance RPE derivation by adding a WNT agonist to the ACTIVIN A–containing medium (Nadar et al. 2015). Because the differentiation process alone may not yield a pure population of RPE cells, an enrichment process is often needed. Several laboratories manually select and expand pigmented colonies (Pennington et al. 2015; Ben M'barek et al. 2017; Da Cruz et al. 2018). Instead, our laboratory enriches RPE cultures by negative selection of cells expressing neuronal surface markers with CD24 and CD56 antibodies (Sharma et al. 2019, 2022). Finally, to obtain a fully mature RPE culture ready to be used as a clinical product, WNT inhibition through primary cilium stimulation with PGE2 was shown to enhance iPSC-RPE maturation (May-Simera et al. 2018; Sharma et al. 2022).

Figure 1.

Schematic showing the developmentally guided directed differentiation protocol for retinal pigment epithelium (RPE) generation. (Figure adapted from Sharma et al. 2022, with permission from Elsevier © 2022.)

iPSC-RPE CELL-BASED THERAPY: IN-PROCESS AND CLINICAL PRODUCT VALIDATION

When using tissue differentiated from iPSCs for developing a replacement therapy, stringent QC assays to characterize every step of the manufacturing process and the final product have become imperative. There are FDA guidelines available to check safety when using cells as a source of treatment (Mendicino et al. 2019). However, iPSC-based technologies are still in their infancy as far as the FDA roadmap is concerned. iPSC-based treatments are broadly categorized into autologous or allogeneic approaches. Autologous or patient-specific cell therapies use the patients’ own cells to make iPSC cells and the iPSC-derived product is transplanted back into the patient, whereas allogeneic therapies use iPSCs generated from one specific healthy donor and the derived tissue is used to treat larger populations. For the autologous approach, the process is the product; since every individual gets their iPSC-derived tissue transplanted, validating the process is as critical as the product. But for allogeneic transplants, hypothetically, one master iPSC bank can be used to treat the entire population; hence, the validation steps are focused more on the final product. The therapeutic and interventional clinical trial space is using both the allogeneic and autologous approaches. The iPSC-RPE field is currently using both these strategies; two groups are using the autologous approach to target neovascular wet AMD and dry AMD, while two other groups are exploiting the allogeneic approach for developing therapies for geography atrophy (dry AMD) (Mandai et al. 2017; Sharma et al. 2019) (ClinicalTrials.gov identifiers: NCT04339764, NCT02463344, NCT01344993). Here, we will primarily focus on the autologous-based cell-therapy approach for dry AMD and will discuss the critical steps of the manufacturing pipeline, QC of the product, and validation of the process. We have categorized this section into iPSC validation and iPSC-derived RPE validation.

iPSC Validation

1. Somatic cell donor screening,

2. Quality of iPSC, and

3. Sterility: endotoxin, mycoplasma, bacteria, fungus testing.

Somatic Cell Donor Screening

For making iPSCs, somatic cells such as blood cells and fibroblasts are often used. The Centers for Disease Control and Prevention (CDC) and the American Association of Blood Banks (AABB) have a list of assays to conduct for screening patients for pathogenic diseases (Mendicino et al. 2019; Jha et al. 2021). For autologous cell therapy, the donor screening (www.cdc.gov/transplantsafety/protecting-patient/screening-testing.html) is done mainly to protect GMP operators and there is no strict regulatory requirement to do so. Hence, once donors are screened for disease pathogens (hepatitis B, HIV, Nile virus, etc.), it is not critical to screen the source material before starting the process of iPSC induction (Creasey et al. 2019).

iPSC Quality Assessment

Reprogramming of somatic cells to iPSC is a time-extensive process. The process of generating iPSCs has been refined since its discovery (González et al. 2011). Previously, integrating viral vectors (e.g., retroviral, lentiviral, and inducible lentiviral) were the only options, but now nonintegrating transgene expressing technologies (e.g., Sendai, adenoviral, plasmid DNA transfer, loxp lentivirus, piggyBac, polyarginine-tagged polypeptide, RNA-modified synthetic mRNA) are also available (Yu et al. 2011; Rao and Malik 2012). There is a wide range of somatic cell types that can be used and, depending upon the type, the reprogramming efficiency varies (Rao and Malik 2012). At the National Eye Institute (NEI) phase I/IIa clinical trial (ClinicalTrials.gov identifier: NCT04339764) we are using episomal vectors (Sharma et al. 2019). As far as iPSC-based RPE clinical trials, the choice has been skin fibroblasts or peripheral blood CD34⁺ cells (Mandai et al. 2017; Sharma et al. 2019). The limiting factor and a point for consideration is whether there are GMP protocols for generating iPSCs for a specific cell type; for example, CD34⁺ cells have protocols established for GMP of iPSC lines (Mack et al. 2011). Figure 2 highlights the streamlined autologous GMP of iPSC-RPE for the NEI phase I/IIa clinical trial to treat dry AMD (Sharma et al. 2019).

Figure 2.

Schematic of clinical manufacturing process of the induced pluripotent stem cell-retinal pigment epithelium (iPSC-RPE) patch at the National Eye Institute (NEI) for phase I/IIa autologous cell-based replacement therapy for dry age-related macular degeneration (AMD). (oriP) Minimal replicator of Epstein–Barr virus.

Regardless of the method used for iPSC generation, clinical-grade iPSC quality is validated by checking for the expression of iPSC markers, either by qRT-PCR, flow cytometry or immunofluorescence of surface (SSEA4, TRA-1-81) and nuclear (OCT-4, SOX-2) markers (Rao and Malik 2012; Baghbaderani et al. 2015). Stringent QC thresholds for iPSC quality in terms of percentage-positive iPSC markers are usually preferred. In iPSC-derived RPE cell therapy, the Japanese group led by Dr. Takahashi used immunostaining and teratoma assay to assess the pluripotency of the iPSC lines (Mandai et al. 2017). The iPSCs also need to be characterized for sterility (endotoxin, mycoplasma, fungus, bacteria), donor matching (short-tandem repeat assay or HLA-matching), and zero footprints—absence of the reprograming vectors (plasmid loss determination by copy number) (Rao and Malik 2012; Baghbaderani et al. 2015; Steichen et al. 2019).

To ensure the genomic integrity of iPSCs, different assays can be carried out. For example, conventional G-banding determines the potential karyotypic abnormalities, whole genome or exome sequencing can determine the status of oncogenic and disease-inducing alterations, copy number variations (CNVs), or insertion or deletion mutations (INDELs) (Sharma et al. 2020). The very first autologous iPSC-derived RPE transplantation clinical study in Japan was halted because of potential oncogenic mutations discovered in iPSCs of the second patient enrolled in their trial (Mandai et al. 2017).

RPE Validation during Differentiation and Maturation

1. Molecular validation,

2. Structural validation,

3. Functional validation,

4. Junctional integrity validation, and

5. Tumorigenicity, dose toxicity, and distribution.

   The quality of the final product depends on the differentiation protocol; details of the NEI protocol have been covered in the previous section and published extensively (May-Simera et al. 2018; Sharma et al. 2020, 2021, 2022). Here we focus on the assays that validate RPE purity, maturity, and functionality. Several of the other RPE differentiation protocols have assays that focus on the final product (Ben M'barek et al. 2017; Mandai et al. 2017; Kashani et al. 2018). However, our differentiation protocol has three stages of RPE development: RPE progenitors (day 25), immature RPE (intermediate, day 40), and mature RPE (day 75, final product)—this has allowed us to develop “go/no go” assays for the clinical manufacturing process (Sharma et al. 2019, 2022).

The final RPE product has been transplanted either in cell suspension or as a monolayer patch (Schwartz et al. 2015; Mandai et al. 2017; Kashani et al. 2018). There are various kinds of natural, synthetic materials like parylene or biodegradable materials like poly(lactic-co-glycolic acid) (PLGA), and polyester that have been used to seed RPE cells and transplant the tissue as a sheet (Kashani et al. 2018; Sharma et al. 2019). Poly(glycerol sebacate) (PGS) scaffolds are biodegradable scaffolds used in the transplantation of other retinal cells like photoreceptors (Lee et al. 2021). Biodegradable scaffolds like PLGA or PGS are beneficial because they provide a support to secrete extracellular matrix proteins, an essential factor in cellular interactions and cell signaling (Sharma et al. 2020). Moreover, they enable cells to build native niches like the Bruch's membrane underneath iPSC-RPE (Sharma et al. 2019). This allows generation of a mature functional iPSC-RPE monolayer with basal infoldings, which are otherwise missing on a plastic surface (Sharma et al. 2020). A significant and obvious benefit of using a biodegradable matrix is that, following transplantation and degradation of the scaffold, there will be no hindrance to the exchange of nutrients and metabolites. At the same time, another important point to consider is that the substrate, even though it is biodegradable, should have enough physical strength to be easily manipulated during the transplantation procedure.

RPE sheets have also been grown on a collagen substrate without scaffold support, and later the collagen was dissolved using collagenase enzyme to obtain only the cell sheet that was transplanted in one patient (Mandai et al. 2017). A recent report from Nishida et al. showed transplantation in animals of iPSC-derived RPE sheets grown in a polydimethylsiloxane (PDMS)-based device, which is 19.5 mm in length, 1 mm in width, and 1.6 mm in depth (Nishida et al. 2021). Authors transplanted albino nude rats and rabbits and showed RPE transplant integration in the back of the eye.

In the next section, we highlight the QC steps at the intermediate and the end-stage product. The process of characterization and validation of the differentiating tissue can be designed during various manufacturing stages. It helps in benchmarking the quality of the product.

Molecular Validation

The developing RPE has a specific gene-expression pattern in its early and late stages of maturity. These signature genes, discovered or established by various laboratories, can be used to study the molecular maturity of differentiating RPE by assays such as qRT-PCR, flow cytometry, western blot (WB), or immunofluorescence (IF) staining. For instance, the early signs of RPE commitment fate are determined by MITF and PAX6 coexpression (Bharti et al. 2012). As the cells mature, TYRP1 and CRALBP start to appear, while BEST1, RPE65, and ALDHA3 markers decide the final stage of maturity (Sharma et al. 2019). Our group uses the flow cytometry assay to determine the purity of differentiating cells. Other groups have used qRT-PCR and immunostaining-based assays to confirm the maturity of the end product (Mandai et al. 2017).

Structural Validation

RPE has a unique cellular architecture that can be observed by high-resolution scanning and transmission electron microscopic (SEM, TEM) techniques. The apically located melanosomes, the apical mesh of microvilli, basally located nucleus, and basal infoldings confirm the polarity and structural maturity of the iPSC-RPE monolayer (Strauss 2005; Bharti et al. 2011). These structures can be validated by SEM and TEM (Fig. 3). RPE polarity is preserved by a specific distribution of proteins on its apical and basal sides. For example, EZRIN and Na/K ATPase are expressed on the apical membrane, while COLLAGEN IV is the marker for the basal side of RPE cells (Miyagishima et al. 2016). RPE monolayer secretes pigment epithelium-derived factor (PEDF) predominantly on the apical side for photoreceptor health and vascular endothelial-derived growth factor (VEGF) predominantly on the basal side for choriocapillaris (Strauss 2005; Maminishkis et al. 2006; Miyagishima et al. 2016). ELISA assay is typically used to detect these cytokine levels and can be used to confirm the polarity of RPE tissue in vitro. Several groups use VEGF and PEDF secretion assays to confirm the maturity of RPE tissue (Osakada et al. 2009a; Ben M'barek et al. 2017; Mandai et al. 2017; Da Cruz et al. 2018; Kashani et al. 2018).

Figure 3.

Transmission and scanning electron microscopic (TEM and SEM) images of induced pluripotent stem cell-retinal pigment epithelium (iPSC-RPE) validates maturity at the structural level.

Functional Validation

Inside the eye, RPE remains in direct contact with the outer segments (OS) of photoreceptors to complete the replenishment of visual cycle pigments and uptake shed OS through phagocytosis (Lakkaraju et al. 2020). Established in vitro assays can assess RPE phagocytic ability via WB, IF, or flow cytometry for uptake and digestion of OS and rhodopsin WB—a key OS protein (Lakkaraju et al. 2020). In the laboratory, RPE cells are “fed” with bovine OS, and, after 4 h of “feeding,” cells are tested for the presence of rhodopsin either by WB, IF, or flow cytometry. For example, Mandai et al. checked the phagocytosis capacity by feeding the porcine photoreceptor outer segment (POS) labeled with FITC (Osakada et al. 2009a; Mandai et al. 2017). For flow cytometry-based assays, OS are labeled with pH-sensitive or any other fluorescent dye. The phagocytic ability can be directly correlated with the fluorescent intensity, which can be quantified by flow cytometry-based assay. Our data show that iPSC-RPE are comparable to primary human RPE in their ability to phagocytose OS (Miyagishima et al. 2016).

Morphometric Validation

RPE is a monolayer of hexagonal cells and its hexagonality is thought to correlate with RPE health and maturity (Schaub et al. 2020; Ortolan et al. 2022). Thus, morphometric analysis can indicate this tissue's health. Our group has developed an artificial intelligence (AI)-based software that can determine shape metrics, such as cell hexagonality and number of neighbors of each RPE cell (Schaub et al. 2020; Ortolan et al. 2022). The cells are first stained for border markers, using anti-ZO-1 antibodies or phalloidin (stains for F-ACTIN), and subsequently imaged with a fluorescence microscope (Fig. 4). Images are then segmented with our machine-learning-based software—RESHAPE. This morphometric assay is used for QC of the iPSC-RPE patch (Sharma et al. 2019, 2020). Others have used anti-ZO-1 and phalloidin staining to qualitatively assess RPE shape (Schwartz et al. 2015; Song et al. 2015).

Figure 4.

Example of induced pluripotent stem cell-retinal pigment epithelium (iPSC-RPE) stained with a ZO-1 antibody and its corresponding cell border segmentation performed by RESHAPE, a machine-learning software for morphometric analysis.

Junctional Integrity

Another important RPE maturity feature is related to the tight junctions that the cells express in the monolayer (Strauss 2005; Maminishkis et al. 2006). As RPE cells mature into a monolayer, these junctional complexes get stronger by addition of more proteins onto them, leading to fully sealed paracellular spaces between cells (Rizzolo 2007). The junctional integrity of an RPE monolayer is measured by passing an electric current from the apical toward the basal side of the monolayer. The resistance to the flow of current defines the tightness of RPE junctions. There are several ways to study this—with EVOM or ENDOM or using a modified Ussing chamber (Maminishkis et al. 2006). Our clinical study is the only clinical study to date that uses this functional readout of RPE monolayer maturity as several hundred ohm·cm² of TER.

The in vitro validation of the end product is followed by safety and efficacy testing in animal models. The choice of animal model is based on the target disease and the mode of transplantation.

ANIMAL MODELS FOR PRECLINICAL STUDIES

Preclinical efficacy and toxicity studies are required for regulatory approval before the transplant can be tested in humans (Chader 2002). For a first-in-human trial, toxicity studies are required to be performed as per good laboratory practice (GLP) standards and are discussed in our regulatory section. Here, we focus on efficacy studies that are typically performed in a retinal degeneration animal model for an RPE-based transplant. Immune rejection of a xeno transplant by the animal is a major cause of failure of human transplants in such preclinical studies; hence, a strong immune-suppression regimen is required to allow survival of the human cells over the duration of the study. Important factors that contribute to the success of cellular therapy are the mode of injection, how cells integrate into the eye, and whether the transplanted cells survive (Frey-Vasconcells et al. 2012; Sharma et al. 2020; Arzi et al. 2021). Some of the earliest RPE transplantations were done in monkeys and in Royal College of Surgeon Rats (RCS) (Binder et al. 2007). The RCS rat model is a well-established animal model used for RPE patch transplantation (Binder et al. 2007). In this rat model, a homozygous mutation in the MERTK gene renders the host RPE cells incapable of phagocytosing shed POS, causing retinal degeneration and secondary photoreceptor loss (Binder et al. 2007). Researchers have tested the ability of human RPE cells injected as a suspension or a patch to rescue retinal degeneration seen in this model (Wang et al. 2008). However, rescue effects also observed in some cases following sham surgery have confounded some of the results obtained from this model (Wang et al. 2008). Our group has tested their RPE transplants in pigs. The pig is a useful model to study human RPE transplants because the pig eye contains a cone photoreceptor-rich area, called the visual streak, which is reminiscent of the human macula that degenerates in AMD (Hafezi et al. 2000). Furthermore, the size and anatomy of the pig eye are very similar to the human eye, making it practical to test surgical approaches meant for the human eye (Choi et al. 2021). For these reasons, we tested our iPSC-derived RPE patch in a pig eye with laser-damaged RPE, such that the human RPE patch could rescue the pig retina from degenerating (Sharma et al. 2019). Our work shows the feasibility of delivering a patch of the size intended for a human eye, into a pig eye. Follow-up of the RPE patch with clinically relevant techniques such as fundus imaging, optical coherence tomography, and electroretinograms of the pig retina confirmed RPE-transplant integration and its ability to rescue the pig retina from degenerating (Sharma et al. 2019). Our experience suggests that testing of an RPE patch in an animal model that better recapitulates the human eye and human disease condition helps build confidence in the clinical product to be delivered to patients.

REGULATORY CHALLENGES FOR DEVELOPING AN iPSC-BASED CELL THERAPY

Many regulatory challenges need to be overcome to develop a pluripotent stem cell–based therapy. Here, we focus on an iPSC-based product for clinical testing. These challenges may significantly impact the final product and must be adequately addressed at the outset of product development. Among these challenges are the establishment of a reliable and consistent manufacturing process, the validation of analytical in-process controls—including the specificity, sensitivity, accuracy, and reproducibility of each assay—and the execution of the necessary preclinical studies to validate the product's safety and efficacy prior to use in humans.

The FDA's Center for Biologics Evaluation and Research (CBER) governs cell-based therapies in the United States. Several sections of the CFR offer general guidelines for the development of an iPSC-derived product (Mendicino et al. 2019).

The development process for cell therapy products differs depending on the chosen strategy between autologous and allogeneic. Because each strategy has its own requirements and challenges, it is essential to identify these limitations early to implement strategies that improve the safety and efficacy of the product for a particular clinical intervention. See Jha et al. (2021) and Creasey et al. (2019) for additional information on the latest regulatory considerations for developing autologous cell therapy and the approval pathway for the cell therapy biologics license application (BLA).

Overall, the regulatory consideration for developing a phase I IND application for cell-based therapies requires three major sections: chemistry, manufacturing, and control (CMC), preclinical studies, and clinical studies (Fig. 5; FDA 2008a,b).

Figure 5.

General requirements for an investigational new drug (IND) application for the U.S. Food and Drug Administration (FDA). In the Code of Federal Regulations (CFR), the FDA has issued guidelines for the development of regulated products. Several sections of Title 21 of the CFR provide general guidelines for developing an induced pluripotent stem cell (iPSC)-derived product. (CMC) Chemistry, manufacturing, and control, (POC) proof-of-concept, (GLP) good laboratory practice, (IRB) institutional review board, (iRPE) iPSC-derived retinal pigment epithelium.

CMC information is a detailed description of the clinical product that includes the cell source, the collection method, and any associated handling, culturing, processing, storage, shipping, and testing. It is one of the most essential components of a phase I IND application for cell therapy (FDA 2008a). To ensure product safety, characterization, quality, purity, stability, and strength (including potency) (details in 21 CFR 312.22(a) for the U.S. FDA), FDA requires sufficient CMC information (Fig. 6).

Figure 6.

Chemistry manufacturing and controls requirements for a phase I investigational new drug (IND) application for autologous and allogeneic induced pluripotent stem cell (iPSC) product.

Critical quality attributes (CQAs) are the properties of cell therapy products that provide a mechanistic understanding of product characterization and ensure manufacturing consistency and stability. CQAs are particularly advantageous for the development of autologous products because they assist in identifying variability and its source during the manufacturing process (Jha et al. 2021). The FDA mandates preclinical studies that need to be GLP-compliant (details available in 21 CFR Part 58 of the FDA) to evaluate the safety of iPSC-derived cell therapy products intended for clinical use (Mendicino et al. 2019). These studies evaluate the tumorigenic, toxic, and migratory properties of iPSC-derived cell therapy products. The functionality of a cell therapy product is another critical quality issue that preclinical studies should address. The functionality will likely determine the success of integration of the transplanted cells in the target tissue and the ability to become part of a functional unit. To ensure that transplanted human cells survive long enough in animals to reveal their tumorigenic potential, preclinical studies employ a variety of animal models, including both immunocompetent (with immune suppression) and immunocompromised animals (Harding and Mirochnitchenko 2014; Jha et al. 2021).

In addition to FDA approval of the IND application, an institutional review board (IRB) and a data safety monitoring board help ensure patient safety. Establishing the product's safety in a patient cohort is the foundation for all subsequent trials. Because a phase I study is intended to be a safety trial, the first patient cohort should be selected so that if the product fails to meet its safety profile, it will cause minimal or no harm to patients. The identification and comprehension of a patient's prognosis are crucial to effective risk management. Patients must be made aware of the potential dangers associated with a first-in-human procedure.

CLINICAL DEVELOPMENT OF AUTOLOGOUS iPSC-RPE PATCH TRANSPLANT

Our laboratory at the NEI is conducting a phase I/IIa trial for testing the safety of autologous iPSC-RPE patches for treating dry AMD patients. The protocol plans to enroll up to 16 patients with a primary end point of 1 yr and secondary end point follow-up of 5 yr. The first autologous transplantation of an iPSC-RPE scaffold in the eye of a patient with dry AMD was completed in August 2022 at the NEI (www.nei.nih.gov/about/news-and-events/news/first-us-patient-receives-autologous-stem-cell-therapy-treat-dry-amd).

The estimated completion date of Clinical Trial NCT04339764 is 2026–2028. With the current timeline, if successful, this technology would likely be avaiable for FDA approval 8–10 yr from now.

The time taken at NEI to manufacture an iPSC-RPE patch is 164 d (Fig. 2). Once the iPSC culture is ready, it takes 40 d of differentiation to obtain pure RPE cells. Our RPE cells grow on PLGA scaffolds for an additional 5 wk before the cell therapy product is ready to be transplanted.

Previously, it was suggested that the costs incurred in manufacturing clinical grade iPSCs are close to $800,000 (Huang et al. 2019). Our manufacturing cost (reagents, supplies, and assays) is well below $50,000. Because of the academic nature of this trial, the cost of GMP facility operation and management is not feasible to calculate at this stage. It is anticipated that GMP facility operation costs will drop significantly as this project reaches phase III when products for multiple patients are being manufactured simultaneously.

The success rate of iPSC reprograming is quite high in our manufacturing process; on average, we have been able to pick 20–30 clones from every donor. The major advantage of our autologous approach is that the therapy does not require systemic immunosuppression. Given the typical age of dry AMD patients, they are predisposed to multiple life-threatening diseases. Having a healthy immune system greatly increases the chances of success. The main disadvantages of the autologous approach are the large costs of manufacture and the fact that the technology of generating personalized iPSC-RPE patches is logistically cumbersome. To alleviate this issue, we have set up several freezing points along the manufacturing process that allow us to keep backups of the product intermediates. With ongoing technological innovation in microfluidics and AI-based image analysis, it is anticipated that, in the coming decade, manufacturing of autologous cell therapies will be streamlined and the cost will drop by several orders of magnitude.

CONCLUDING REMARKS

The current decade has witnessed an increasing demand for the use of stem cell–based therapy for the treatment of several diseases, including retinal degenerative diseases (Bharti et al. 2014). Visual health is an important area of research—eye diseases if left untreated can progressively lead to blindness (Flaxman et al. 2021). There are many hurdles for developing iPSC-derived products for clinic care. Herein we have covered the essential steps that led to the development of iPSC-RPE clinical-grade tissue and led to a first-in-human study for the treatment of dry AMD. Our data suggest that it is critical to develop a tissue that faithfully resembles the native tissue by using the knowledge accumulated by studies on tissue development; proper QC assays to assess the quality and functionality of the product; testing its safety and efficacy by transplanting it in relevant animal models; and, finally, following the guidelines set by the FDA to ensure timely reach to the patients.

In vivo, the RPE interacts with other cell types in the retina-like photoreceptors, endothelial cells, pericytes, and fibroblasts (Sharma et al. 2020; Song et al. 2022). So, the next step of cell-replacement therapy would be going from a two-dimensional culture to three-dimensional culture to increase the success of integration of the new tissue and to treat late-stage diseases. Our group has recently developed an in vitro system that recapitulates the oBRB, by embedding endothelial cells, pericytes, and fibroblasts in a bio-ink, which is then bioprinted on the basal side of a biodegradable scaffold on which RPE cells were seeded (Song et al. 2022).

The RPE is primarily responsible for the health of the photoreceptors and can support preservation of vision only when photoreceptors are still present. In cases in which photoreceptors have already been lost, RPE transplantation alone will not suffice. As such, other groups have aimed at the generation of photoreceptors from iPSC as a viable option for cell-replacement therapy (Lee et al. 2021). Transplantation of iPSC photoreceptors is still at the proof-of-concept stage, and safety and efficacy studies of iPSC photoreceptor scaffolds is being tested in animals (Lee et al. 2021). Currently, researchers are looking into ways to develop the technology that would enable transplantation of RPE photoreceptors as a dual patch for cell-replacement therapy. Hopefully, more clinical trials can be developed in the near future by leveraging our experience to cure currently untreatable diseases.