Depicting brighter possibilities for treating blindness

See allHide authors and affiliations

Science Translational Medicine  29 May 2019:
Vol. 11, Issue 494, eaax2324
DOI: 10.1126/scitranslmed.aax2324


Advances in preclinical research are now being translated into innovative clinical solutions for blindness.

The World Health Organization (WHO) estimates that approximately 1.3 billion people live with some form of vision impairment. Degenerative diseases of the retina, such as age-related macular degeneration (AMD) and inherited retinal dystrophies (IRDs), are major causes of untreatable blindness, with glaucoma, pathological myopia, and corneal blindness also having a major health impact (1). Recent clinical trials have led to approval in 2017 of the first treatment, gene therapy, for treating blinding retinal degeneration. This follows decades of research on visual processing, genetics, animal models, mechanisms of vision loss, vector design, imaging, and microsurgery that have provided invaluable preclinical data and clinical proof-of-concept results. Substantial investment by charities, public agencies, and businesses has moved innovative therapeutic strategies, such as pharmacotherapy, gene therapy, stem cell therapy, and optogenetics, into clinical development. In this Focus article, celebrating Science Translational Medicine’s 10th anniversary, we examine some of the exciting advances—from gene and cell therapy to prosthetics and neuroprotection—made in treating retinal degeneration over the past decade.

Progress in ophthalmology is intrinsically linked to increased understanding of the morphology and function of the visual system. Vision is a complex process that begins in the retina, the specialized neurosensory organ that is established in the eye during embryonic development (2). Retinal photoreceptor cells (rods and cones) convert light into neuronal signals that are processed by other retinal cell types including retinal ganglion cells. The resulting visual information travels via the optic nerve to higher centers in the brain, where it is processed and decoded into visual perception. Specific characteristics make the eye particularly suited for diagnostic and therapeutic exploration: easy access, small volume, high internal compartmentalization, and stable cell populations. The optical transparency of the eye allows direct visualization with high-resolution imaging and precise evaluation of disease stage and response to therapy. Moreover, the relative immune privilege of the eye, especially the subretinal space, reduces adverse responses to injected vectors and gene products. However, the impact of this immune privilege on cell-based therapies is less clear.

AMD is a leading cause of irreversible blindness and central vision loss in the elderly (Fig. 1). It is a multifactorial disease in which cumulative damage over a lifetime leads to progressive deterioration of key retinal structures, including the retinal pigment epithelium (RPE), Bruch’s membrane, the choroid, and photoreceptor cells. Although there has been substantial progress in treating the neovascular form of AMD (characterized by growth of abnormal, leaky blood vessels), there is no effective or approved treatment for the atrophic “non-neovascular” form (associated with gradual loss of photoreceptors and RPE cells). Oxidative stress, inflammation, heredity, environmental factors, and demographic factors are implicated in AMD etiology but not fully understood and are potential therapeutic targets. Despite continued gains in understanding AMD pathophysiology, relevant animal models and prospective controlled clinical studies are lacking. Furthermore, early markers of the disease that could be targets of future preventive treatment have yet to be identified or validated.

Fig. 1 Relevance of therapies for treating retinal degeneration and stage of disease.

Gene replacement therapy is appropriate during the early stages of retinal degeneration when the photoreceptor cells (rods and cones) are still intact (stage I). Early intervention with gene replacement or gene editing could potentially reverse vision loss and lead to close to normal visual outcomes (stage II). Neuroprotective strategies, particularly those for preserving cones, are the best approach for treating disease where there is ongoing photoreceptor cell degeneration (stage III). Neuroprotection of cones can stave off loss of high acuity vision by protecting the fovea (stage III). Stem cell therapy, optogenetic therapy, and retinal prostheses are used to restore vision during the later stages of retinal degeneration, after the loss of cone outer segments (stage III). These approaches can be applied independently of the causal mutation and are expected to restore a low degree of vision in blind patients (stage IV). Pierre-Auguste Renoir, “Madame Henriot,” National Gallery of Art, Washington, DC.


IRDs may be inherited as Mendelian traits or through mitochondrial DNA, involve the entire retina or just the macula, affect either rod or cone photoreceptor cells predominantly, and may occur as single, syndromic, or systemic forms, with peripheral and central loss of vision (Fig. 1). The numerous challenges associated with the development of therapies should be considered alongside the extensive genetic and phenotypic heterogeneity of IRDs. In addition, genotype-phenotype correlations are difficult: Different mutations in the same gene can cause different diseases, or the same mutation can result in different phenotypes.


There have been some notable successes over the past decade using gene replacement strategies for treating blinding retinal diseases. Leber congenital amaurosis (LCA) is the most severe IRD that usually develops in early childhood. In the autosomal-recessive form of LCA caused by RPE65 mutations, a biochemical deficit impairs the ability of photoreceptor cells to respond to light. Delivery of wild-type RPE65 complementary DNA (cDNA) to RPE target cells in animal models and in humans led to substantial improvements in light sensitivity, visual fields, and functional vision as demonstrated on multi-luminance mobility testing. These studies led to regulatory approval in 2017 for gene therapy to treat LCA in children and adult patients with confirmed biallelic RPE65 mutation–associated retinal dystrophy (3). The successes with RPE65 gene therapy have paved the way for more than 30 gene replacement trials worldwide. Phase 3 trials of gene replacement therapy for choroideremia, achromatopsia, and Leber hereditary optic neuropathy are ongoing. A recent promising strategy delivered an antisense oligonucleotide to restore correct splicing of a common LCA-causing variant, CEP290, that results in a splicing defect (4).

Gene-independent strategies could overcome the complexities of the mechanisms underlying photoreceptor cell degeneration in AMD and the genetic heterogeneity of IRDs and their often dominant transmission. Of these strategies, neuroprotection aims to prevent or slow the progressive degeneration of photoreceptor cells. The retina-specific neurotrophic factor rod-derived cone viability factor rescued cone structure and function independently of genetic mutations and the mechanisms and extent of rod degeneration. This is a potential therapeutic strategy for a broad spectrum of retinal dystrophies (5).

Stem cells are at the center of another mutation-independent approach for vision restoration that replaces degenerated cells. The proposed cell therapies are based on human pluripotent stem cells (PSCs), which expand indefinitely in culture and are a potentially unlimited source of retinal cells (RPE cells, photoreceptor cells, and retinal ganglion cells) for cell replacement. Methods to differentiate human PSCs toward retinal lineages have improved in the past decade, particularly the development of three-dimensional culture systems for generating human retinal organoids. Based on preclinical data, several stem cell–based therapies for RPE replacement for AMD and IRDs are currently under development or clinical evaluation (6). Stem cell therapy could be used to restore vision in a wide range of retinal degenerative conditions, provided that functional integration into the host tissue occurs and that immune responses to transplanted cells can be avoided or limited.


Retinal prostheses are able to reactivate remaining retinal circuits at the level of bipolar or ganglion cells after photoreceptor cell loss. Both epiretinal and subretinal implants are able to stimulate a light-insensitive degenerated retina and to restore partial vision in blind people (7, 8). One implant has received market approval in Europe and in the United States; a photovoltaic wireless subretinal prosthesis is currently under clinical evaluation. The wireless device is characterized by photovoltaic electrodes with their own local return circuit and independent function (potentially giving higher resolution) and a simpler surgical procedure than for wired implants.

Another technology called optogenetics makes cells light sensitive through expression of an optogene encoding a light-activated channel or pump in the remaining inner retinal cells. Optogenes can be targeted to specific cell types using adenovirus-associated viral vectors equipped with cell type–specific promoters (9). Optogenetic therapy could be used to resensitize a degenerated retina to visible light independent of the mutation causing photoreceptor cell loss. Optogenes targeted to cones, bipolar cells, or ganglion cells in animal models of retinitis pigmentosa have been shown to restore visual function and behavioral responses to visual cues. Furthermore, efficient expression of optogenes in cones and ganglion cells has been demonstrated in the retina of nonhuman primates in vivo and in postmortem human retinas in vitro. The choice of the target cell type depends on the state of retinal degeneration. Cone targeting is expected to produce optimal results, followed by targeting of bipolar cells and, lastly, ganglion cells. All of these approaches require the patients to wear goggles that stimulate the optogenes with appropriate light intensity. Ganglion cell–based optogenetic stimulation is in phase 1 clinical trials for treating advanced retinitis pigmentosa ( NCT03326336).


Recent therapeutic strategies for treating blindness are very encouraging, and the field is poised to address the next set of challenges. Efficiently timed delivery of genes or small molecules to the appropriate cells is critical for success. For developmental IRDs, maximum efficacy would likely require intervention during gestation, which is fraught with safety and ethical issues. The rescue of photoreceptor cell function also depends on whether there are any viable cells left in the retina; there may be nothing left to treat if the intervention is too late. Remodeling of the retina late in disease may also limit the ability of therapies to restore retinal function. Retinal remodeling may also affect the feasibility of gene transfer through subretinal injection. Subretinal gene transfer requires general anesthesia and carries risks, making delivery through the routine office procedure of intravitreal injection appealing. However, there are very few gene delivery agents that can reach the appropriate cellular targets in an organ the size of the human eye after intravitreal delivery. Furthermore, some test compounds are unstable (antisense oligonucleotides), necessitating repeated injections. Delivery of antisense oligonucleotides via a slow-release compound may sustain the therapeutic effect and avoid the risk of repeated injections.

Several common IRD genes have relatively large cDNAs that are incompatible with the packaging constraints of the current set of viral vectors. Effective delivery will require, for example, developing vectors with a larger cargo hold, trimming the transgene cassettes, halving the cargo and delivering multiple complementary drugs, and developing methods to safely and specifically edit the endogenous DNA or RNA. So far, clinical trials have not targeted autosomal-dominant mutations causing retinal degeneration. This may require delivery of two different reagents to inactivate the native faulty gene and deliver the normal gene. A gene-editing approach for autosomal-dominant IRDs, for example, would also require delivery of two different components, CRISPR and Cas9 (10). Thus, from dosage, safety, and regulatory perspectives, gene transfer of large genes or genes for autosomal-dominant retinal degenerative diseases is far more complicated than “simple” gene replacement.

The validation of a robust response to a therapeutic intervention can also be a challenge for diseases causing profoundly abnormal baseline vision. Traditionally, drugs for ophthalmic indications have been approved based on one criterion: the ability to read lines of letters on an eye chart. This standard reflects the function of the foveal cone cells that occupy 1/1000 of the area of the retina. Additional criteria are required to assess potential benefits for aspects of vision carried out by the extrafoveal retina. National regulatory agencies now request evaluation of the impact of disease and therapies on functional vision (patient-reported outcomes, performance-based tests, or daily activities).

IRDs and other diseases that affect ganglion cells can lead to atrophy of the optic nerve. Therapy cannot improve vision after optic nerve loss. However, brain-machine interface technologies using electrode arrays or optogenetics can stimulate the visual pathway downstream of the retina. Electrical stimulation of the primary visual cortex is one possible scenario that is currently in clinical trials. An early feasibility study is evaluating the safety of the visual cortical prosthesis and surgery, as well as the reliability of the system and the usefulness of any restored vision ( NCT03344848).

The expanding armamentarium of gene therapy and gene-editing agents will allow testing of interventions for a variety of IRDs. As safety data accumulate on new vectors and routes of administration, regulatory bodies may relax the regulatory burden. This, in turn, will reduce the cost of clinical trials. Improved properties of therapeutics will allow many of them to be delivered safely during routine office procedures. As interventions specific to single genetic targets continue to develop, interventions that could treat IRDs regardless of genetic cause will also emerge. The latter include neuroprotective agents, those that enhance metabolic and nutritional pathways and those that can activate more distal neurons in the visual pathway. The next major challenges will be to understand the effects of such therapies on brain plasticity and to demonstrate the impact of vision preservation or restoration in real life.


Acknowledgments: We thank K. Marazova and D. Dalkara for valuable comments. Funding: This work was supported by LabEx LIFESENSES (ANR-10-LABX-65), ERC Synergy “HELMHOLTZ”, Banque publique d’Investissement, Foundation Fighting Blindness. Competing interests: J.-A.S. has relationships with Pixium Vision, GenSight Biologics, and SparingVision and holds multiple patents related to treating blindness. J.B. is a co-inventor on patent US8147823B2 that covers “Method of treating or retarding development of blindness,” but waived any potential financial gain in 2002. B.R. declares no competing interests.

Stay Connected to Science Translational Medicine

Navigate This Article