Many advocates of cell therapies are of the opinion that pluripotent cells—ES and iPS cells—represent a significant breakthrough in the progress of regenerative medicine. Alan Trounson and Natalie DeWitt, writing in the “Science and Society” section of the journal Nature, describe their potential as “unique and extraordinary.”1 Not that there isn’t also some skepticism. The claims for pluripotent cells, according to Theodore Friedman, a former chairman of the National Institute of Health’s Recombinant DNA Advisory Committee “have been characterized by the kinds of exaggerations and elevated expectations that were seen in the field of gene therapy just a few years earlier.”2 Claims, he might have added, that are now broadly seen as the epitome of hype in the history of biomedicine.3
There’s no question that deciphering the biological basis of pluripotency is an exquisite piece of experimental science, but so was the discovery of Pluto. Why should we conclude that new stem cell therapies will result? Trounson and DeWitt cite two parameters to support their enthusiasm: the scalability of pluripotent cells—the ability to expand them almost indefinitely—plus their pluripotency. Do these stand up to scrutiny?
The first of these is significant, though certainly not unique. We have met several examples already of expansive cell populations: ReNeuron’s “CTX0E03” cells and SanBio’s “SB623” cells, to name just two. Conversely, we have encountered instances where the inability to scale the production of an appropriate stem cell population severely limited its applicability, the most prominent being the fetal cells used for the Parkinson’s trials. A major problem in that instance was the variability between the cells derived from aborted fetuses. One scalable source that could be applied to a series of patients would certainly represent success. We have also identified variability and lack of scalability as a problem with MSC therapies. So perhaps the point about pluripotent cells is not that they are uniquely scalable, but that they might allow us to scale up the cells we actually want, rather than forcing us to work with the cells that happen to be intrinsically scalable.
The ability to generate any cell type we want, at least in principle, is clearly enticing. Many of the cells that practioners would like to get their hands on are not readily available. We hear all the time of clinical practice restricted by the lack of availability of human tissue: only 10 percent of the demand for organ transplants is currently being met.4 There are, for example, far more candidates for liver transplants than there are cadaveric livers available. What if we could just make liver cells starting from pluripotent cells? And, beyond that, there are therapies we can’t even conceive of because the cells are simply not available. No one has ever been able to realistically contemplate a cell replacement therapy for motor neuron disease that actually started with human motor neurons because where would you start to source such cells? Not that such a therapy is available now, but step one has been completed: these cells have now been generated from pluripotent cells, giving the adventurous among the regenerative medicine community a substrate to work on.
There is, however, a third feature of pluripotent cells that distinguishes them fundamentally from what has gone before: their capacity for histogenesis. When we discussed multipotential neural stem cells in chapter 2, we highlighted their capacity to generate all the major cell types of the central nervous system (neurons, oligodendrocytes, and astrocytes). If we are now saying that the potential of pluripotent cells includes the generation of neural stem cells, which in turn have the potential to generate all the major cell types of the nervous system, you might ask: what has actually been achieved? We seem to have taken a circuitous route and arrived at the same place. There are two differences. First, starting with pluripotent cells gives us more control over precisely the type of neurons and glia we end up with. We’ll see exactly how important this is when we reconsider Parkinson’s disease. But, second, the neural stem cells derived from the pluripotent cells have a property that earlier generations of cultured neural cells never achieved: the capacity for histogenesis. Pluripotent cells can actually build tissue from scratch.
If you cultured the ReNeuron CTX0E03 cells appropriately, predictably they would give rise to neurons and glia. But they would be jumbled neurons and glia, with none of the structure that you would encounter if you looked at proper brain tissue. So while these cells were derived from human cerebral cortex, they have no capacity to organize themselves into something that resembles cortex. Compare this with the fetal progenitor cells from which they are derived. The cells of the fetal ventricular zone (which we met in chapter 3) build the adult cerebral cortex essentially single-handedly. Certainly, they need a blood supply to provide nutrients and an endocrine system to provide the right hormonal milieu, and they will eventually need to talk to other brain regions to ensure the right networks are created. It is also true that this region of the fetal brain incorporates cells from neighboring regions as development proceeds. But essentially, this population of ventricular cells builds the cortical structure unaided. To go back to the epigenetic concept that emerged in chapter 9, these cells have the correct genes primed to direct a cascade of behaviors that will ultimately lead their progeny to create human cerebral cortex in all its complexity. This is the wonder of development.
Remarkably, human pluripotent cells have this same capacity. If human ES or iPS cells are cultured appropriately, they make cerebral cortical precursor cells, which will attempt to build a cerebral cortex in the tissue culture dish. How far they get will depend on how they are handled. If they are grown in the conventional fashion as a thin layer coating the surface of a plastic dish, then they won’t get very far. They will try to form a tubular structure equivalent to the neural tube they would form early in brain development, but on the flat surface of the dish this turns into individual flower shaped rosettes, as if the tube had been salami-sliced into a series of thin sections and laid side by side. Modest though this is, this constitutes histogenesis far beyond anything conventionally sourced neural stem cells ever achieve.
But to see the pluripotent derivatives at their best, you need a more permissive culture. If you allow them to float free, then they form true three-dimensional structures, and start to organize themselves into a rudimentary brain, populated with neuroepithelial cells. Then these neuroepithelial cells begin to generate neurons. Just as in vivo, they make the deep cortical neurons before they make the superficial neurons, so they build the cortex starting from the inside and growing outward, just as happens in normal development. Eventually, they build something that looks remarkably like the brain of a six-week-old fetus.5
But then they hit a snag. Without a blood supply, they can only go so far. No cell in the mature brain is more than a few cell diameters away from a blood vessel, and without the nutrients the blood provides, the cells can’t survive. The little cortical organoids have no blood supply and they soon outgrow the capacity of the culture environment to meet their needs, bringing progress to a halt.
Overcoming this limitation is a major area of research in the regenerative medicine community. How far researchers get will play a large role in setting ambitions in this area for the next decade. If it turns out that pieces of cortex (or any tissue) can be grown to a reasonable degree of size and maturity, then the possibility arises (at least in theory) that such tissue pieces could be used to replace the areas of brain lost to stroke, for example. This would, however, need to be accomplished with a degree of consistency and accuracy considerably beyond what has currently been achieved. Quite apart from the technical challenge, the ethical and logistical problems facing such an approach are enormous. In truth, were this strategy to prove feasible, it would surely impact other somatic tissues well in advance of brain, and in fact progress across a broad range of tissues is progressing apace.6 One could conceive of the transplantation of liver or pancreas organoids, for example. An exciting prospect indeed.
So where will these advantages of pluripotent cells lead in the arena of brain repair? There are three indications involving neural tissue where pluripotent cell-derived products are threatening to make an early impact: spinal cord injury, Parkinson’s disease, and disorders of the retina. This is not to say that other indications are not also in play. The original drive to replace cells following stroke has switched to pluripotent cells, and encouraging preclinical findings have emerged there.7 Similarly, therapies for Huntington’s disease have also moved on to employ pluripotent cell derivatives.8 Other CNS disorders seem likely to follow a similar path.
The most advanced, somewhat surprisingly, are therapies for spinal cord injury; “surprisingly” because the challenge in spinal cord injury does not appear to align with the strengths of pluripotent cells. We saw in chapter 6 that the primary problem in spinal cord injury is the rupture of spinal axons, a problem of axonal regrowth rather than replacement. A transection (partial or complete) of the spinal column breaks the nerve pathways running between the brain and the spinal cord, meaning the brain can no longer control motor activity below the break. Voluntary movement is lost.
In 2009, the FDA approved a clinical trial sponsored by Geron Corporation for a differentiated cell product (GNROPC1) derived from human ES cells for the treatment of spinal cord injury. The uneven progress of this trial presents a cautionary tale. The first patient was treated in 2010, only to have Geron terminate the study a year later with just four of the eight planned patients having been treated. This withdrawal was not a consequence of an anticipated failure of the trial, the Company maintained. True, the clinical results showed little sign of efficacy (preliminary though they were), but GNROPC1 had appeared safe, which is as much as can ever be reliably concluded from a phase 1 clinical trial. Rather, Geron put its decision down to financial concerns. By killing the study, the company claimed, they would save $25 million, enough to fund half a dozen phase 2 clinical trials of their two cancer products.
There was, however, a suspicion that this was not the whole story. The scientific rationale for the Geron approach always looked somewhat tenuous. The product they developed (GNROPC1) was a preparation of oligodendrocyte progenitor cells (OPCs), differentiated from a human ES cell line procured from James Thomson, the original developer of the human ES cell technology. It was never clear why this particular population of neural progenitor cells should have a therapeutic effect in spinal cord injury. The cells’ primary function, as their name suggests, is to produce oligodendrocytes. It was never clear how these myelinating cells of the nervous system should have any positive impact on lesioned spinal neurons. This is particularly true since the mature oligodendrocytes are known to inhibit axon regrowth and are thought to be one of the negative factors preventing spinal cord repair. Animal studies, however, suggested that there was functional recovery following engraftment of these progenitor cells, and this had opened the possibility of a clinical trial.
It is also the case that Geron had had an earlier scare when some of their treated rats developed spinal cysts. These were neither malignant nor a cause for concern, argued Geron, but the FDA put a hold on the study until more data could be provided. When these new data suggested that the cysts were the result of an epithelial cell contaminant, which was excluded from subsequent cell preparations, the trial was restarted. Not without some raised eyebrows, however, particularly since this was the first-in-human trial of a decidedly novel therapy, where more caution might have been expected. Audrey Chapman and Courtney Scala have argued that Geron’s original decision to go ahead was flawed. In their view the preclinical data were inadequate, the regulators’ decision-making process opaque, and the design of the study prevented a meaningful consent by patients.9 No doubt Geron would point to the 22,000-page dossier it had had to submit to the FDA in order to obtain clinical trial approval as evidence that they’d walked the extra mile, and then some.10 “Don’t be the first one out the door” was the conclusion of Michael West, Geron’s chief scientific officer (CSO), “The first one out the door gets all the arrows in his back.”11
Either way, the Geron withdrawal looked like a serious early setback for ES-based therapies and the end of the road for Geron’s particular application, but neither prediction has proven correct. In 2014, GNROPC1 (now renamed “AST-OPC1”) reemerged in the hands of the California-based Asterias Biotherapeutics, Inc., which, on the back of a $14.3-million grant from the California Institute of Regenerative Medicine (CIRM), had acquired the Geron technology.
The Asterias trial has increased considerably in complexity. There are now five cohorts of patients either in progress or planned. In each case, the patients have suffered cervical level lesions, a variation from the earlier thoracic lesion studies. These clinical studies are ongoing and clinical data are still sparse. The first cohort of three patients showed no clinical improvement, though the second cohort of five patients (who received a larger dose of cells) has been reported to show some clinical progress. These are small numbers, naturally, and we will have to wait until the end of the study to know the true outcome. In the meantime, there are strong testimonials from the small number of individual patients who have seen dramatic improvement.12
Asterias has also completed a number of animal studies, building on the earlier preclinical data,13 reporting both behavioral and pathology assessments. They have adopted a novel way of integrating the behavioral data into one overall score of the animal’s gait. This has the advantage of lessening the uncertainty about which might be the most proper measure of locomotor performance, and of removing the suspicion that the investigators make lots of readings and simply report the most favorable. Using that combined parameter, rats with cervical spinal cord lesions showed improved overall scores following engraftment of AST-OPC1 cells after four months, compared with ungrafted animals.
Of the pathological outcomes, they saw reduced cavitation and increased myelination following engraftment of the cells. Somewhat like the cyst formation following stroke, cavitation following spinal cord injury is a consequence of the loss of neural tissue. Again, as in stroke, scar tissue forms and a space remains where previously there was neural tissue. A reduction is cavitation therefore represents success in terms of reducing the advancing pathology. The increased myelination follows more directly from the cell therapeutic of choice: oligodendrocyte progenitor cells would be expected to generate myelinating cells. Consequently, this program certainly looks more robust than it once did, but as ever, we await the pivotal clinical trial results to see the real outcome.
What difference has it made taking pluripotent cells as the starting material, compared to the earlier studies we considered? Have the attributes of pluripotency, scalability, and capacity for histogenesis made a demonstrable difference? Histogenesis is not much in evidence here. The ES cells have been employed primarily for their pluripotency, in this case their capacity to generate OPCs. The scalability of the ES cells, and indeed of the OPCs, has allowed the expansion of specific cell types into clinical production and is very much in evidence. Human OPCs are not a trivial cell type to acquire from primary brain tissue, and consequently little direct research has been done on this cell population. But the extensive research into rodent OPCs that has been conducted over the years has proved applicable to their human counterparts, facilitating this approach enormously.
What’s still missing here, however, is a satisfactory manufacturing process. A major shortcoming of the fetal cell approach for Parkinson’s disease was the uncontrolled mix of cell types that emerged from the fetal brain: progenitor cells, young neurons of multiple types, and various contaminating cells—vascular endothelial cells, fibroblasts, and others. While the AST-OPC1 cells are nowhere near as disparate, neither are they a wholly characterized population.
In the original study describing the preparation of the OPC1 cells, more than 95 percent of the differentiated cells were OPCs.14 In their more recent publications, Asterias says that the proportion of OPCs in the clinical preparations is between 30 and 70 percent, with the identity of the remaining cells uncertain.15 They label cells with a marker called “Nestin,” a protein expressed in a wide variety of neural progenitor cells, but also found in a variety of other cell types during development. The cells do not label with markers that would identify neurons or recognized glial cell types. So, we can conclude that these interloper cells are probably neural, but of unknown phenotype.
This relatively uncharacterized and variable mix of cells is not really satisfactory. The sponsors of this trial have done the best they can, and as far as possible, they have shown the cells to be safe, at least in the preclinical studies. Nonetheless, you would hardly be reassured if you bought a prescription drug at the pharmacist to be told that the active ingredient is somewhere between 30 and 70 percent of the total material in the tablet, and the makers were somewhat uncertain as to what else was in there.
Achieving more consistency and reproducibility remains a major goal for the field, particularly as this Asterias case is actually one of the more satisfactory examples of current practice: at least they know the identity of half the their cells. In many cases, it is simply impossible from the published literature to discern the true make up of the cell populations that are going into clinical trials.
Consider this intriguing comparison: if an academic research group wished to publish their data describing the biological activity of a population of cells, they would submit their manuscript to an academic journal, who would send it for assessment to a number of peer reviewers, experts in the field. The first question these reviewers would address would be: what are the cells in question, and how do the research team know that these cells are what they think they are? If the team had not demonstrated to the satisfaction of the reviewers that they had exhaustively characterized and identified their cells, it is highly unlike that the paper would be accepted for publication, certainly not in any reputable journal. Remarkable then, that regulators are relatively relaxed about exactly this question when the population of cells in question is about to be injected into someone’s head. Surely this cannot be allowed to go unchallenged for too much longer. We need to know what’s in the syringe.
There is a genuine conundrum here. A regenerative medicine project usually begins with an experimentally defined cell preparation, typically improvised by academic researchers, who will characterize the cells, at least in part, and show that they have some activity in an animal or cellular model. If successful, this project then gets adopted by a commercial entity, usually a small company. They optimize the protocols and scale-up, aiming for a clinical trial. Before the cells progress far into the clinical program, the protocols will often be handed over again, sometimes to a newly created development arm of the original company, and sometimes to an external contract research organization (CRO) that specializes in commercial production. Finally, a cell preparation emerges that finds its way into increasing numbers of patients. But the conundrum is: how can the anyone be certain that the cells emerging from this end process are equivalent to the cells the academic researchers started with. Unless they can, how do they know that the cells going into patients have equivalent biological activity to those tested in the preclinical models?
One answer of course is that the late development cells can be tested in the same animal models in which the earlier cells had been shown to work, and the best research groups do indeed do this. The problem is that such models are invariably slow and expensive. What the company really needs are fast and easy release assays, so that each batch of cells can be shown to be comparable. Partly because of cost, partly because of the difficulty of determining the mode of action of cell therapies, these release assays are slow to emerge.
The second area where pluripotent cells spell progress is Parkinson’s disease. Here the approach seems much more methodologically secure. When we left this area in chapter 5, it was marooned without a suitable supply of cells. There was proof of concept that the replacement of dopaminergic neurons could bring about clinical improvement, but the source of the neurons—human fetal cells—was inconsistent and unreliable. Perfect, it would seem, for a defined cell product derived from pluripotent cells. At least three groups worldwide have declared their intention to begin trials with such cells in the near future. A consortium comprising Malin Parmar from Lund and Agnete Kirkeby from Copenhagen recently announced support from Novo Nordisk in a project aimed at generating new dopaminergic cell therapies.16 In the United States, a parallel program is under way sponsored by BlueRock Therapeutics and driven by stem cell biologists Gordon Keller and Lorenz Studer with the support of Bayer.17 And in Japan a group led by Jun Takahashi in Kyoto are taking a similar approach.18
This progress rests firmly on the earlier pioneering work coupled with insights that have emerged with the advances in stem cell science. Significantly, it also hangs on some important findings from basic developmental biology. It turns out that, from the outset, biologists had misunderstood the origin of the midbrain dopaminergic neurons. Chapter 3 described how neurons are derived from the neuroepithelial cells of the neural tube. Broadly that is true, but it now transpires that the dopamine progenitors come from a distinct structure—the floor plate—adjacent to the region originally thought to produce these cells. This probably explains in part the variability that was seen with the early fetal grafts, and the insight has provided a more accurate process of dopamine cell production and concomitant improvement in the markers by which that process can be monitored. Coupled with the scalability inherent in the pluripotent cell starting material, these developments have all led to a much more robust and efficient manufacturing process being devised for the production of the dopaminergic neuron progenitor cells required for Parkinson’s therapy.19
Cell therapy for Parkinson’s disease has also taken another important step. Several times in this book we’ve had cause to bemoan scientists’ poor understanding of the mode of action of cell therapeutics. As we’ve seen, we do not know how a number of cell therapies actually work. This has never really been the case for the dopaminergic therapy, where the evidence clearly supported replacement of neurons as the significant factor. Nonetheless, this hypothesis still needed proof.
Lorenz Studer and his colleagues provided this proof in mice using an elegant new technology, optogenetics. They took mice that had been lesioned using the same model we met in chapter 5: dopamine neurons on one side of the brain are destroyed, and the animal rotates, chasing its tail. This parkinsonism is then “cured” by the engraftment of dopaminergic cells, in this case human dopaminergic neurons derived from pluripotent cells, the sort of cells that are now about to enter clinical trials. This is all as we’ve come to expect: the lesion gives the mice a parkinsonian pattern of behavior, and the cell transplant restores the behavior to normal. The question is: can we be sure that the dopaminergic activity of the cells is responsible for bringing about the change in behavior?
This is the same predicament that Brian Cummings and his team faced in 2005 with the spinal cord grafts that we considered in chapter 6; they overcame the problem with the judicious use of tetanus toxin. But in the intervening years, technology had moved on, and Studer’s team were able to employ a much more sophisticated approach.
The trick was that the cells they used were not just ordinary ES cells. They had been engineered to express a strange light-sensitive protein called halorhodopsin. As its name suggests, this protein is related to the rhodopsin, which in the eye normally mediates the capture of light. But this chimeric protein is part light receptor and part membrane chloride pump, and has a particular property. If the chloride pump is activated in dopaminergic neurons, it inhibits their firing, and this activation can be achieved (courtesy of the halorhodopsin attachment) by simply shining light on it. So, if you shine light on the dopaminergic cells, they cease activity.
This allowed the experimenters to perform an elegant experiment. First, they could show that hemiparkinsonian mice, engrafted with the cells, did indeed recover. The dopaminergic cells had worked, reducing the parkinsonian behavior. Then, using a fiber optic light source fed directly into the mouse brain, the chloride pump was activated in the transplanted cells, turning off their activity. If the behavioral recovery were truly dependent on the dopaminergic activity of the grafted cells, this should lead to a loss of the recovery and the reemergence of the parkinsonian behavior, and, sure enough, this is precisely what the experimenters observed. This experiment proved conclusively that it was truly the dopaminergic activity of the transplanted neurons that was responsible for the functional recovery.
Though an enormously influential experiment in the context of cell therapies, it is by no means the only application of optogenetics to therapies more generally. There are many scenarios where the ability to turn neurons on or off can deliver novel therapeutic approaches. In the treatment of pain, for example, optogenetics are being investigated as a means to lessen the ability of pain neurons to fire. The use of gene therapy to deliver this type of treatment is very much in its infancy but will surely gain prominence in coming years.20
The very strong impression now is that in the age of pluripotent human stem cells and given the substantial improvements in understanding and manufacturing, the huge investment in cell therapies for Parkinson’s disease might be about to pay off.
The treatment of retinal disorders—particularly age-related macular degeneration (ARMD)—is probably the most dynamic area of neural transplantation in the whole regenerative medicine arena. Several reports have emerged recently of well-structured clinical trials with different pluripotent cell–derived preparations. To read these reports is to be struck not only by the advantages conferred by the pluripotent cells, but equally by the advancements and sophistication that have emerged since the earlier studies.
Recall that the core problem with ARMD is the degeneration of the pigment epithelium—the dark-colored layer of cells that underlies the retina (see figure 6.2). Since this layer is required to maintain the viability of the photoreceptors of the retina, loss of epithelium results in degeneration of photoreceptors with a corresponding loss of vision. We saw in chapter 5 that efforts to replace the pigment epithelium with cells taken either from the periphery of the patient’s own retina or from cadaveric retina produced inconclusive or problematic outcomes. Both sources of material are clearly limited and are themselves “old” pigment epithelium.
So the scalability and pluripotency of pluripotent stem cells are an enormous advantage here. Several protocols have been devised that produce pigment epithelium from human ES or iPS cells. Indeed, a recognizable manufacturing process for the production of a true advanced therapeutic medicinal product is now emerging. The study led by Lyndon da Cruz and Peter Coffey at the London Project to Cure Blindness incorporates a clear linear process to generate a defined product, which (subject to release criteria) is provided to a surgeon, who then employs a specifically designed delivery tool to engraft the tissue into patients. No doubt the sponsorship of Pfizer, while signaling a growing interest of Big Pharma in regenerative medicine, has also helped develop manufacturing capability. Whatever other skills pharmaceutical companies might bring, they certainly understand the manufacture of medicines.
Histogenesis is also starting to emerge as a feature. In the da Cruz and Coffey study and also in the California study reported by Amir Kashani and colleagues,21 the medicinal product is not merely cells dissociated from a culture dish—as has been the case in most studies reported in this book—but rather a sheet of pigment epithelial cells, which da Cruz and Coffey call a “patch,” prepared on a specifically engineered membrane. The Kashani product is prepared on the synthetic polymer “parlene,” which has the advantage of being already established for use in a medical device. In both cases, the cells have tissue integrity as a result of having been grown on a synthetic structure. They have already formed a polarized epithelium sitting on a membrane equivalent to Bruch’s membrane, the structure on which they would sit in the undamaged eye. This is important because the breakdown of Bruch’s membrane is an integral part of ARMD pathology in the first place. So by combining the capacity of the ES cells to generate pigment epithelium together with an artificial membrane, these researchers have built a structure capable of truly replacing lost retinal tissue.
Not all have embraced this approach, and it does have its own problems. Steven Schwartz and his colleagues at the Geffen School of Medicine in Los Angeles have preferred a dissociated cell approach, pointing out that the surgery required to deliver a cell suspension to the virtual space behind the retina is considerably simpler than that required to deliver the structured implant. Another problem with the retinal patch is that the size of the delivery tool and the patch itself means that they can’t be effectively tested in rodents, simply because they’re too big. This led the London group to run a preclinical trial in pigs, whose eyes are closer to humans’ in size and structure than those of rodents. They were able to demonstrate successful delivery of the patch in twenty animals in this fashion. Which approach is the safer and more efficacious will presumably emerge from the clinical trials. Furthermore, clearly progress across multiple fronts will be required to make these advanced therapies a success: in stem cell science, certainly, but also in surgical devices, materials design, and manufacturing process.
All three of the initiatives mentioned above are now in clinical trials. Schwartz and colleagues have reported the most advanced data from the largest group of patients. They’ve treated eighteen patients across four centers, separated into two patient cohorts. One cohort comprises older patients with the dry form of age-related macular degeneration, and the other, younger patients with Stargardt disease, a genetic juvenile form of macular degeneration. The protocol these researchers have employed (and which has been broadly adopted) is to treat the worse-affected eye, and to use the better eye as a control. The emerging safety data are good, though some patients have suffered complications from the surgery and others from the immunosuppression (about which more later). But by and large, there were no adverse events associated with the cellular therapy, and while numbers are still small, the efficacy data are very encouraging. An increase in pigmentation around the damaged area was observed in thirteen of the eighteen patients, suggesting the grafts had taken. More significantly, visual acuity demonstrably improved in nine ARMD patients and stabilized in a further three. The visual acuity outcome in the Stargardt patients was more modest, improving in three patients and stabilizing in a further three. Visual acuity was measured using a letter chart similar to one many of us have encountered at our local opticians. Each patient was asked to name letters of decreasing size. The smaller the letters a patient could distinguish, the better the patient’s visual acuity.
With their more technically ambitious approaches, da Cruz/Coffey and the Kashani studies have fewer treated patients to report currently. Both studies showed that the patch had engrafted successfully, with evidence for functional pigment cells now underlying the retina. The London study reported two patients, and while the numbers are again small, the improvement is marked, with both patients able to read an increased number of letters on the chart—from 10 to 39 in one case, and 8 to 29 in the other—compared to before treatment. Kashani reports five patients. The improvement here appears more modest. Four of five patients showed no significant improvement, while the fifth could read 17 letters more than before the treatment. In both studies again, the safety data were good.
In all the cases just cited, the pluripotent cells of choice are human ES cells. What about the newer iPS cells? In Japan, Michiko Mandai and colleagues commenced a study with two separate trials in 2017, both using a sheet of pigment epithelium similar to those above but derived from iPS cells.22 One study was autologous: the retinal cells came from iPS cells generated from skin fibroblasts donated by the patient himself. The other was allogeneic: the iPS cells came from an unrelated donor and were thus more directly comparable to the ES cell studies. Both of these trials, however, are currently on hold. There were to have been two patients in the autologous study, and five in the allogeneic study, but in each case only a single patient has been treated. While there is no reason to believe that treatment won’t recommence in the future, the reasons for the halt are informative, and address issues around safety we have ignored up to now: genetic mutation and tissue rejection.
1. Trounson, A., and DeWitt, N. D., “Pluripotent Stem Cells Progressing to the Clinic,” Nature Reviews Molecular Biology 17, no. 3 (2016): 194–200; quotation on 194. http://doi.org/10.1038/nrm.2016.10.
2. Friedmann, T., “Lessons for the Stem Cell Discourse from the Gene Therapy Experience,” Perspectives in Biology and Medicine 48, no. 4 (2005): 585–591; quotation on 585. http://doi.org/10.1353/pbm.2005.0089.
3. See, for example, Addison, C., “Spliced: Boundary-Work and the Establishment of Human Gene Therapy,” BioSocieties 12, no. 2 (2016): 257–281. http://doi.org/10.1057/biosoc.2016.9.
4. Giwa, S., Lewis, J. K., Alvarez, L., Langer, R., Roth, A. E., Church, G. M., et al., “The Promise of Organ and Tissue Preservation to Transform Medicine,” Nature Biotechnology 35, no. 6 (2017): 530–542. http://doi.org/10.1038/nbt.3889.
5. See Lancaster, M. A., Renner, M., Martin, C.-A., Wenzel, D., Bicknell, L. S., Hurles, M. E., et al., “Cerebral Organoids Model Human Brain Development and Microcephaly,” Nature 501, no. 7467 (2013): 373–379. http://doi.org/10.1038/nature12517.
6. Clevers, H. “Modeling Development and Disease with Organoids,” Cell 165, no. 7 (2016): 1586–1597. http://doi.org/10.1016/j.cell.2016.05.082.
7. Tornero, D., Wattananit, S., Gronning Madsen, M., Koch, P., Wood, J., Tatarishvili, J., et al., “Human Induced Pluripotent Stem Cell–Derived Cortical Neurons Integrate in Stroke-Injured Cortex and Improve Functional Recovery,” Brain 136, no. 12 (2013): 3561–3577 http://doi.org/10.1093/brain/awt278.
8. Ma, L., Hu, B., Liu, Y., Vermilyea, S. C., Liu, H., Gao, L., et al., “Human Embryonic Stem Cell–Derived GABA Neurons Correct Locomotion Deficits in Quinolinic Acid–Lesioned Mice,” Stem Cell 10, no. 4 (2012): 455–464. http://doi.org/10.1016/j.stem.2012.01.021.
9. Chapman, A. R., and Scala, C. C., “Evaluating the First-in-Human Clinical Trial of a Human Embryonic Stem Cell–Based Therapy,” Kennedy Institute of Ethics Journal 22, no. 3 (2012): 243–261. http://doi.org/10.1353/ken.2012.0013.
10. Alper, J., “Geron Gets Green Light for Human Trial of ES Cell–Derived Product,” Nature Biotechnology 27, no. 3 (2009): 213–214. https://www.nature.com/articles/nbt0309-213a.
11. Michael West, as quoted in Baker, M., “Stem-Cell Pioneer Bows Out: Geron Halts First-of-Its-Kind Clinical Trial for Spinal Therapy,” Nature 479, no. 7374 (2011): 459. http://www.nature.com/news/stem-cell-pioneer-bows-out-1.9407.
12. See, for example, Aldrich, M., “Experimental Stem Cell Therapy Helps Paralyzed Man Regain Use of Arms and Hands,” USC News, September 9, 2016. https://news.usc.edu/107047/experimental-stem-cell-therapy-helps-paralyzed-man-regains-use-of-arms-and-hands/.
13. See, for example, Manley, N. C., Priest, C. A., Denham, J., Wirth, E. D., III, and Lebkowski, J. S., “Human Embryonic Stem Cell–Derived Oligodendrocyte Progenitor Cells: Preclinical Efficacy and Safety in Cervical Spinal Cord Injury,” Stem Cells Translational Medicine 6, no. 10 (2017): 1917–1929. http://doi.org/10.1002/sctm.17-0065.
14. Keirstead, H. S., Nistor, G., Bernal, G., Totoiu, M., Cloutier, F., Sharp, K., and Steward, O., “Human Embryonic Stem Cell–Derived Oligodendrocyte Progenitor Cell Transplants Remyelinate and Restore Locomotion after Spinal Cord Injury,” Journal of Neuroscience 25, no. 19 (2005): 4694–4705. https://www.jneurosci.org/content/jneuro/25/19/4694.full.pdf.
15. Priest, C. A., Manley, N. C., Denham, J., Wirth, E. D., III, and Lebkowsi, J. S., “Preclinical Safety of Human Embryonic Stem Cell–Derived Oligodendrocyte Progenitors Supporting Clinical Trials in Spinal Cord Injury,” Future Medicine 10, no. 8 (2015): 939–958. https://www.futuremedicine.com/doi/pdfplus/10.2217/rme.15.57.
16. “Collaboration between Lund University Researchers and Novo Nordisk Paves the Way for Large-Scale Cell Therapy against Parkinson’s Disease,” Lund University, Faculty of Medicine, May 16, 2018. https://www.med.lu.se/english/news_archive/180516_parkinson.
17. “Bayer and Versant Establish iPSC Therapeutics Company BlueRock with $225M,” GEN, December 12, 2016. http://www.genengnews.com/gen-news-highlights/bayer-and-versant-establish-ipsc-therapeutics-company-bluerock-with-225m/81253536.
18. “Kyoto University Reprograms Stem Cells to Fight Parkinson’s in Monkeys: A Breakthrough for Therapy,” Japan Times (Reuters), August 31, 2007. https://www.japantimes.co.jp/news/2017/08/31/national/science-health/kyoto-university-team-reprograms-stem-cells-fight-parkinsons-disease-monkeys/.
19. Parmar, M.. “Towards stem cell based therapies for Parkinson’s disease,” Development 145, no. 1 (2018): dev156117–4. http://doi.org/10.1242/dev.156117.
20. Williams, S., “Optogenetic Therapies Move Closer to Clinical Use: With a Clinical Trial Underway to Restore Vision Optogenetically, Researchers Also See Promise in Using the Technique to Treat Deafness, Pain, and Other Conditions,” Scientist, November 16, 2017. https://www.the-scientist.com/?articles.view/articleNo/50980/title/Optogenetic-Therapies-Move-Closer-to-Clinical-Use/.
21. Kashani, A. H., Lebkowski, J. S., Rahhal, F. M., Avery, R. L., Salehi-Had, H., Dang, W., et al., “A Bioengineered Retinal Pigment Epithelial Monolayer for Advanced, Dry Age-Related Macular Degeneration,” Science Translational Medicine 10, no. 435 (2018): eaao4097. https://
22. Mandai, M., Watanabe, A., Kurimoto, Y., Hirami, Y., Morinaga, C., Daimon, T., et al., “Autologous Induced Stem-Cell–Derived Retinal Cells for Macular Degeneration,” New England Journal of Medicine 376, no. 11 (2017): 1038–1046. http://doi.org/10.1056/NEJMoa1608368 http://doi.org/10.2217/rme-2017-0130.