Currently there is a huge unmet medical need for effective therapies to treat neurological disorders, including Alzheimer’s disease, Parkinson’s disease, multiple sclerosis, spinal cord injury, psychiatric disease, amyotrophic lateral sclerosis, stroke, and age-related macular degeneration, said Sally Temple, a principal investigator and the scientific director of the Neural Stem Cell Institute. All of these diseases have a common mechanism of cell loss, which provides a rationale for contemplating the replacement of these cells with stem cell–based products, she said. The neurological system is extraordinarily complex, containing multiple cell types and involving complex interactions within and between each neural region; therefore, it is challenging to produce the specific neural cell types that are authentic and appropriate for each specific region. Despite the complexity, there has been progress made in several areas, including retinal pigment epithelium transplantation, oligodendrocyte replacement, and the transplantation of human neural stem cells. During this session, speakers discussed opportunities in these areas, along with challenges encountered and lessons learned.
Adult Retinal Pigment Epithelial Stem Cells
Retinal pigment epithelium (RPE) transplantation is under investigation as a potential method for treating age-related macular degeneration (AMD). AMD, the major cause of vision loss in adults over 65 is caused by the degeneration of RPE cells (AOA, 1994), which are highly specialized cells that support the health and integrity of photoreceptors and the choriocapillaris, the network of capillaries that supplies nutrients to the retina (Sonoda et al., 2009). Patients with AMD experience vision distortion that leads to central vision loss, which can significantly impair their ability to read or recognize faces. Temple and her colleagues are attempting to replace the damaged RPE with new cells from donor eyes. Their protocol involves the harvesting of adult retinal
pigment epithelial stem cells (RPESCs) from donor eyes; these cells are normally dormant but can be encouraged to self-renew in vitro. Because the starting material is specific to the eye and is the natural precursor to RPE, donor RPESCs are especially well suited to make RPE compared to other types of stem cells, Temple said. Once the donor cells are purified and expanded, 50,000 to 150,000 cells are transplanted into the patient’s eye beneath the retina. In the future, Temple said, it would be ideal if RPESCs could be activated in vivo, so that the process of cell removal, replication, and transplantation could be avoided.
Several studies are under way to investigate different sources of RPE for transplantation, and each has its own benefits and drawbacks, Temple said. For example, using embryonic stem cells or induced pluripotent stem (iPS) cells has the benefit of unlimited expansion, but it also has the potential drawback of tumorigenicity. In contrast, RPESCs will likely not increase the risk of tumors in the recipient, but the cells cannot be perpetually replicated.
The effects of advancing studies that use poorly defined cell types or cells that are not normally present in the nervous system are a major concern, Temple said. Unproven therapies springing up in clinics are deeply concerning because they are of unproven safety and efficacy and risk patient health as well as patient trust in the medical system.
Human Embryonic Stem Cell–Derived RPE
An alternative stem cell–based approach to treat AMD was described by Peter Coffey, a professor at the Neuroscience Research Institute of the University of California, Santa Barbara. He told participants about a United Kingdom–based partnership between the University College London Institute of Ophthalmology, the Moorfields Eye Hospital, and Pfizer Neusentis which investigated the possibility of using human embryonic stem cells (hESCs) for the treatment of AMD. As discussed above, RPE cells are critical for retinal function and maintenance of vision. Coffey and his team aimed to develop and deliver a small artificial membrane, or “patch” of RPE (derived from hESCs), to replace damaged or lost RPE in patients’ eyes. The success of the project relied on four main steps according to Coffey: proper cell characterization, high-quality manufacturing, the development and validation of quality control assays, and a demonstration of the safety of hESC-derived RPE.
The characterization of the cells was fairly straightforward, Coffey said. The team performed a series of analyses of phenotype, cellular
ultra-structure, immune activity, and in vitro and in vivo function to demonstrate that the hESC-derived RPE was indistinguishable from native fetal or adult RPE. The group went on to perform whole genome transcript analysis on approximately 40 cell lines of RPE to demonstrate that the hESC-derived RPE was identical to other sources of RPE.
Manufacturing the RPE patches in the UK required submission of an investigational medicinal product dossier (IMPD), which is a document that details the steps that will be taken to manufacture the product and how safety and quality will be ensured.1 Preparing the IMPD took Coffey’s team 9 months and required the development of an extremely detailed manufacturing process. During the process of therapeutic development, researchers often acquire new information and may wish to make modifications; however, the regulatory pathway requires that the process be “locked down” at some point so that the regulators can perform their evaluation. Any non-minor changes to the process are challenging, time-consuming, and costly, Coffey said.
Quality control assays were developed in order to test various aspects of the hESC-derived RPE cells, including their sterility, viability, identity, purity, potency, sensitivity, accuracy, and robustness. Tests for the presence of contaminating hESCs were critical because of concerns about tumorigenicity and teratoma formation. Stringent regulatory requirements stated there must be zero pluripotent cells on the RPE patch, Coffey said. Therefore the team employed specific RPE culture conditions that do not support hESC survival, and they confirmed total hESC depletion using image analysis and flow cytometry, he said. Regulators also required that the hESC-derived cells be transplanted into an animal model to evaluate for tumorigenicity. A 6-month-long tumorigenicity study in immunodeficient mice demonstrated that the cells did not form teratomas, Coffey said.
After a lengthy and complex regulatory process that involved seven different regulatory bodies and cost approximately £10 million, the group received regulatory approval to conduct the first round of clinical trials. In August 2015 the first human patient received the small patch of hESC-derived RPE, which was placed behind the retina, and Coffey’s team expects the 12-month clinical outcomes data in late 2016. The team is now working with a chip manufacturer to build a process for scaling up
production of the patches, with the hope of eventually getting the patch licensed as a therapeutic.
Despite the fact that the product has taken 8 years to develop to this point, the process was streamlined and accelerated because the project’s interdisciplinary team, which included engineers, regulators, good manufacturing practice facilities, and clinicians, had all the necessary skill sets to take the therapy from development to clinic, Coffey said. Rather than approaching the process serially, beginning with laboratory development and moving onto the other steps as appropriate, the team tried to tackle the project in a more holistic manner.
Ann Tsukamoto, the former executive vice president of Scientific and Strategic Alliances at StemCells, Inc., shared her company’s experience with developing human neural stem cells (HuCNS-SCs®). StemCells, Inc., worked with HuCNS-SCs for more than 18 years before shutting down operations in May 2016. With a fluorescence-activated cell sorting protocol, HuCNS-SCs are isolated and purified from human fetal brain tissue, then expanded, banked, and cryopreserved under conditions suitable for clinical applications, Tsukamoto said. The banked cells are tested for safety and biological properties including genetic modifications, normal karyotype, and potency. HuCNS-SCs do not require pre-differentiation before transplant and do not form tumors in vivo, Tsukamoto said.
In animal studies using immunosuppressed or immunodeficient mice, HuCNS-SC cells were shown to migrate throughout the central nervous system (CNS) and differentiate in a site-specific manner depending on where the cells took up residence. For example, HuCNS-SCs differentiated into myelin-producing oligodendrocytes in white matter areas of the CNS, Tsukamoto said. The HuCNS-SCs self-renew in vivo, which is an important feature for cells that are being used to correct a lifelong disorder, she said. Furthermore, HuCNS-SCs exhibit several neuroprotective properties that could be beneficial in trying to regenerate disease target areas within the CNS, such as the production of neurotropic factors, phagocytic activity, and an anti-inflammatory effect, she said.
HuCNS-SCs have been tested in humans for a variety of disorders of the brain, retina, and spinal cord. In total, HuCNS-SCs have been
transplanted into 55 patients, and there have been no safety concerns observed, Tsukamoto said. Evidence from a mouse model of infantile neuronal ceroid lipofuscinosis (also known as Batten disease, which is a fatal, inherited disorder of the nervous system) demonstrated that, upon transplantation, HuCNS-SCs engrafted and produced therapeutic benefit through the protection of endogenous neurons (Tamaki et al., 2009). Subsequent human clinical studies were performed in patients with Batten disease, and the transplanted HuCNS-SCs migrated, self-renewed, and engrafted, and in post-mortem exams researchers found that the donor cells had survived long term. One of the patients with Batten disease who received HuCNS-SCs is still alive, 8 years post transplant, Tsukamoto said.
HuCNS-SCs were also tested in patients with Pelizaeus Merzbacher disease, a leukodystrophy characterized by the inability to form myelin, the insulating sheath that is wrapped around nerve axons to facilitate the conduction of electrical impulses. In a mouse model of the disease, injection with HuCNS-SCs resulted in mature, compact myelin formation (Uchida et al., 2012). Following cell transplant, diffusion tensor imaging of the children’s brains showed de novo myelination, Tsukamoto said. Unfortunately, Phase II trials for Batten disease and Pelizaeus Merzbacher disease were not carried out because of a lack of eligible patients and a poor understanding of the natural history of the disease.
HuCNS-SCs have also been investigated as a potential treatment for the dry form of AMD, with the cells being injected into the sub-retinal space. Testing in the Royal College of Surgeons (RCS) rat, a model of inherited retinal degeneration, demonstrated that the photoreceptor layer was protected long-term by an injection of neural stem cells. The protective effect was the result of multiple mechanisms, including phagocytosis of the outer segments, stabilization of the synapses of the cells, neuroprotection through secretion of neurotrophic factors, and proliferation of endogenous RPE layers, Tsukamoto said. In a Phase I clinical study of 15 human patients with AMD, there was a decelerated progression of the rate of geographic atrophy and improved visual acuity.
Researchers went on to test HuCNS-SCs as a possible treatment for thoracic and cervical spinal cord injuries. In animal models of thoracic spinal cord injury, injected cells migrated, engrafted, and resulted in restored motor function (Cummings et al., 2005). These findings led to a clinical study in which 12 patients received injections of 20 million cells each at four injection sites. There was overall sensory improvement in 7
of the 12 patients, with sustained effects beginning at 3 months post transplant, Tsukamoto said. In a clinical trial conducted in patients with cervical spinal cord injury, and most patients experienced a significant increase in upper extremity motor strength. Immunosuppression was a key factor in getting the HuCNS-SCs to survive, Tsukamoto said. Clinical trials for thoracic and cervical spinal cord injuries showed that most patients experienced restoration of sensory function, but the restoration was not permanent for all patients. Thoracic patients received immunosuppression for 9 months, and there was no loss of function at 12 months. Cervical patients, on the other hand, received only 6 months of immunosuppression and experienced a decline in restored motor function by 12 months. The mixed results of these studies unfortunately resulted in the company closing its doors, Tsukamoto said.
Looking Ahead and Learning from Past Challenges
One major hurdle for researchers in the AMD field, Temple said, is to determine how to replace other damaged cell types such as photoreceptors, the cells that lie adjacent to the RPE and rely on it for normal function and survival. Another goal for the field will be to produce a functional retina of full thickness from three-dimensional organoids. Finally, it will be important to overcome the challenge of getting regenerated or transplanted retinal ganglion cells to project inside the optic nerve in order to connect back to the brain, Temple said.
At the conclusion of their presentations, Temple, Coffey, and Tsukamoto discussed specific challenges and lessons learned during the process of developing a cell replacement strategy, and their individual ideas are listed in Box 4-1 below.
Cell Dosing and Source
In the discussion following the presentations, a workshop participant inquired about the rationale underlying the selection of the cell dose for these therapies. Temple, whose approach to AMD therapy uses 50,000 to 150,000 cells per injection, said that their decision was based on the relatively small size of the diseased area and the number of cells that needed to be replaced, in addition to data from a previous study. The selection of number of cells for the patch for AMD was limited by the size of the patch (3 × 6 millimeters) and by how many cells could actually grow on it, Coffey said. In using HuCNS-SCs to treat AMD,
Tsukamoto said, she and her colleagues chose to dose with 20 million cells because that type of cell migrates and they wanted to cover a larger area. In the case of using HuCNS-SCs to treat spinal cord injuries, she said, researchers at StemCells, Inc., used allometric scaling based on the volume of the patient’s spinal cord compared to the cord size used in animal model tests.
Brian Fiske asked about the rationale for choosing the source of cells for the therapies described by the speakers. Tsukamoto said that research on fetal-derived neural stem cells began well before important discoveries were made using other cells such as hESCs or iPS cells. The easiest cell population to identify, purify, and expand was the brain stem cells, so her team pursued that, she said, and ultimately the rationale behind the source of the cells they chose was the timing of the science. Previous work that demonstrated that RPE cells from organisms such as salamanders could regenerate the RPE and the neural retina in about 4 weeks was the inspiration for her team, Temple said. This evidence spurred Temple’s team to look for adult cells that were multipotent and could self-renew, to see if a similar regenerative process could be activated in human retinas, she said.
Tolerance and Immunosuppression
The challenge of achieving permanent tolerance in a setting with ongoing inflammation was an issue raised by Cynthia Dunbar, a workshop co-chair and the president of the American Society of Gene and Cell Therapy, and she asked the speakers to comment on why and when they decided to stop immunosuppression regimens after transplantation. There are currently no good models with which to examine graft-versus-host disease and tolerance induction, Tsukamoto said, and for their experiments they looked to data that had been published on Parkinson’s disease patients who received fetal tissue grafts. Although they cannot rule out differences in the patient population (thoracic versus cervical spinal cord injury) that may have impacted the clinical outcome of the trial, shortening the length of the immunosuppression may also have contributed to the loss of the transplanted cells and subsequent decrease in motor function, she said.
The speakers in this session were all involved in developing cell-based therapies for AMD, and a workshop participant asked if they had any mechanism for sharing data between them. “I think that would be wonderful,” Temple said, adding that it is critical to share the failures and negative results from trials so that mistakes are not repeated. However, she said, there need to be incentives to create consortia. There has been a good amount of information sharing among those involved in the RPE trials, Coffey said, noting that knowledge sharing is critically important for the entire cell-based therapy community, not just those working on RPE.
Panelists were asked if they believed that the regulatory burden imposed on the therapies is appropriate for the seriousness of the disease, particularly from a patient perspective. The impact of disease on patients’ lives may be something that regulatory agencies do not fully understand, the workshop participant noted, and perhaps risk–benefit analysis may be disease specific. It is important for charities, patient advocates, and hospitals to inform the regulators about how large the unmet clinical need is, and to share the perspective of the patients, Coffey said. Though they had an overall positive experience with FDA, it would have been helpful to have more interaction and a two-way conversation between the researchers and the regulators, Temple said. In the United Kingdom, the relationship between researchers and regulators is one of continual back and forth, Coffey said, which is markedly different than the relationship in the United States. At The Michael J. Fox Foundation, Fiske said, researchers have found ways to engage with FDA and have more open conversations by hosting workshops on specific topics.