Emerging therapies for inherited retinal degeneration

See allHide authors and affiliations

Science Translational Medicine  07 Dec 2016:
Vol. 8, Issue 368, pp. 368rv6
DOI: 10.1126/scitranslmed.aaf2838


Inherited retinal degenerative diseases, a genetically and phenotypically heterogeneous group of disorders, affect the function of photoreceptor cells and are among the leading causes of blindness. Recent advances in molecular genetics and cell biology are elucidating the pathophysiological mechanisms underlying these disorders and are helping to identify new therapeutic approaches, such as gene therapy, stem cell therapy, and optogenetics. Several of these approaches have entered the clinical phase of development. Artificial replacement of dying photoreceptor cells using retinal prostheses has received regulatory approval. Precise retinal imaging and testing of visual function are facilitating more efficient clinical trial design. In individual patients, disease stage will determine whether the therapeutic strategy should comprise photoreceptor cell rescue to delay or arrest vision loss or retinal replacement for vision restoration.


Research into retinal degenerative diseases has garnered interest beyond the field of ophthalmology and vision research because of the unique accessibility of the human retina (which is part of the central nervous system) and progress in identifying gene mutations leading to retinal degeneration. Diseases of inherited retinal degeneration typically lead to severe visual impairment or blindness early in life and are the leading cause of legal blindness among working-age adults in England and Wales, having replaced diabetic retinopathy (1). Therapeutic options are currently limited, as is the case for many neurodegenerative diseases, but new therapeutic approaches including gene therapy, stem cell therapy, optogenetics, and retinal prostheses are on the horizon. Many of these approaches have entered the clinical phase of development, and retinal prostheses have received market approval in Europe and the United States. This Review summarizes these new therapeutic strategies and evaluates their capabilities for halting vision loss or even for restoring vision.


Multiple biochemical pathways and mechanisms are affected in the photoreceptor cells of patients with retinal degenerative diseases. These include phototransduction, the structure of the outer and inner segments of photoreceptor cells and the connecting cilium, vesicle trafficking, chaperone function, lipid metabolism, transcription and RNA splicing, retinal development, and synaptic function. In the retinal pigment epithelium (RPE), which underlies and nourishes the photoreceptor cells, affected mechanisms include the visual cycle that regenerates light-sensitive rhodopsin, phagocytosis, membrane trafficking, and ion transport (2, 3). This pathophysiological complexity poses challenges for developing new treatments.

Recent advances in molecular genetics have enabled the identification of genes encoding enzymes and retinoid-binding proteins of the visual cycle that are mutated in Leber congenital amaurosis (LCA), Stargardt disease (STGD), cone dystrophy (COD), cone-rod dystrophy (CRD), and retinitis pigmentosa (RP) (Fig. 1) (4). Two key enzymes of the visual cycle are RPE-specific 65-kDa protein (RPE65) and lecithin-retinol acyltransferase (LRAT) (5). Mice lacking RPE65 (Rpe65−/−) or LRAT (Lrat−/−) display a block in the visual cycle that leads to an absence of 11-cis-retinal and rhodopsin, resulting in severe impairment of rod photoreceptor function and eventual degeneration of the retina (Fig. 1) (6). Mutations in RPE65 (7, 8) and LRAT (9) cause both RP and LCA in humans. Initial studies in Rpe65−/− mice aimed to bypass the biochemical defect caused by the loss of these enzymes through oral delivery of 9-cis-retinal (Fig. 1) (10, 11). A recently completed single-center, open-label, proof-of-concept clinical trial in which 14 patients with RPE65- and LRAT-related LCA were treated with 9-cis-retinyl acetate (QLT091001) reported that this treatment was well tolerated and resulted in improved vision (12). Most recently, an international multicenter, open-label, proof-of-concept study in 18 patients with RPE65- or LRAT-related RP confirmed safety and efficacy of a once-daily oral dose of QLT091001 (40 mg/m2 per day) for seven consecutive days (13). Long-term data are not available for QLT091001 but are available for the same class of compound (that is, retinoids such as adapalene, isotretinoin, retinol, tazarotene, and tretinoin). In retinal degenerative diseases caused by mutations in ABCA4, such as STGD, COD, and CRD, a genetically defective visual cycle exacerbates the nonenzymatic dimerization of vitamin A, leading to the aggregation of toxic bisretinoids and enhanced accumulation of lipofuscin (4). Possible therapeutic approaches include sequestration of toxic all-trans-retinal (14) to reduce vitamin A availability in photoreceptor cells using RPE65 inhibitors or retinol-binding protein 4 (RBP4) antagonists or by reducing vitamin A dimerization through oral delivery of C20-deuterated vitamin A (15, 16).

Fig. 1. Overview of the visual cycle.

Vision is triggered by light activation of visual pigments, such as rhodopsin, in the photoreceptor cells of the retina. In healthy subjects, the visual cycle follows a sequence of reactions resulting in the conversion of all-trans-retinal to all-trans-retinol. All-trans-retinol diffuses into the cells of the RPE through the interphotoreceptor matrix mediated by the interphotoreceptor binding protein (IRBP). In the RPE, all-trans-retinol is esterfied in a reaction catalyzed by LRAT to all-trans-retinyl esters. These retinyl esters are converted to 11-cis-retinol by RPE65, and 11-cis-retinol is then oxidized to 11-cis-retinal. This molecule diffuses back into the outer segments of photoreceptor cells where the cycle is completed by recombination of 11-cis-retinal and opsins to form the visual pigment rhodopsin. There are many points for therapeutic intervention in a dysfunctional visual cycle caused by mutations of inherited retinal degenerative diseases. In STGD, genetic mutations in the ABCA4 gene lead to increased formation of N-retinylidene-N-retinylethanolamine (A2E). Treatment with deuterated vitamin A or aldehyde traps results in decreased accumulation of A2E and other cytotoxic molecules in animal models. Uptake of retinol from serum by RPE cells, which can occur in age-related macular degeneration (AMD) or STGD, can be blocked by RBP4 antagonists. In certain forms of LCA, mutations in LRAT or RPE65 lead to a deficiency in 11-cis-retinol; gene augmentation therapy enables recovery of RPE65 activity. Alternatively, these forms of LCA can be treated by oral delivery of the synthetic retinoid QLT091001. This drug is converted to 9-cis-retinoid that is then oxidized to 9-cis-retinal, which diffuses into the photoreceptor cells and combines with opsin to form isorhodopsin that then participates in the phototransduction cascade. Other retinal dystrophies may require slowing down of the visual cycle by inhibiting LRAT or RPE65 to reduce the amount of toxic molecules formed in the RPE, such as the bisretinoid A2E.

Credit: H. McDonald/Science Translational Medicine


Neurodegeneration is a hallmark of inherited diseases of retinal degeneration. It may be due to both primary induced cell death triggered by causative mutations and secondary effects such as oxidative stress (17). Neuroprotective approaches aim to preserve neuronal viability by either blocking cell death or strengthening endogenous prosurvival mechanisms. In contrast to gene- or mutation-specific tailored therapies, the advantages of neuroprotective strategies are that they do not depend on any specific mutation and may provide a longer time window for treatment (18). Neuroprotective therapies also may act synergistically with gene replacement therapy, which shows promise but may not completely arrest the degenerative process, despite the visual improvement observed in trials for RPE65-related retinal diseases (19). Neuroprotective approaches will also be of value in treating RP, the most common inherited retinal degenerative disease, in which the respective mutations remain unknown in up to 50% of cases (20). RP is characterized by the loss of rod photoreceptor cells, leading to night blindness. The cone photoreceptor cells are affected later in the disease process, although they may not necessarily express the causative gene. Secondary cone photoreceptor cell degeneration may be caused by either increased exposure to light and oxygen or the loss of endogenous protective mechanisms (21).

Several neuroprotective neurotrophic factors, such as brain-derived neurotrophic factor, basic fibroblast growth factor, pigment epithelium–derived factor, and glial cell–derived growth factor, have been examined in preclinical studies. Neuroprotection has been assessed in terms of preventing the death of photoreceptor cells or increased endogenous release of other neuroprotective molecules (18). Over the past decade, rod-derived cone viability factor (RdCVF) has been identified in preclinical studies as a promising treatment candidate for RP and other rod-cone dystrophies. RdCVF has a paracrine trophic effect on cone photoreceptor cells and also protects them from photo-oxidative damage (2123). This trophic factor offers a mutation-independent treatment option for a variety of retinal degenerative diseases. Methods of delivering RdCVF to the retina using gene therapy viral vectors, such as adeno-associated viral vector (AAV), are currently being evaluated (24).

Another well-studied neuroprotective candidate is ciliary neurotrophic factor (CNTF). The neuroprotective effects of CNTF and promotion of rod photoreceptor cell survival have been demonstrated in animal models of retinal degenerative diseases, with positive effects on rod and cone survival (25). CNTF has also been tested in clinical trials using encapsulated cell technology in the eye; the implanted capsule contains human RPE cells engineered to secrete CNTF, ensuring controlled and sustained delivery of CNTF to the vitreous cavity of the eye (25). Although a two-line or three-line improvement on a standard visual acuity chart was detected in three of seven patients with RP in a phase 1 safety trial (26), several prospective randomized multicenter trials failed to detect a significant improvement in their respective endpoints in patients with early-stage RP ( number NCT00447980) or late-stage RP (number NCT00447993) (27). Higher-dose regimens could potentially be investigated, but there are concerns about higher doses of CNTF suppressing retinal function despite apparent preservation of the retina’s structure (28).

Gene replacement therapy

Gene replacement therapy (also referred to as gene addition or gene augmentation therapy) is the most straightforward option for treating inherited retinal degenerative diseases caused by a single recessive gene defect. In this type of treatment, the addition of a normal copy of the gene into affected cells compensates for the loss of gene function due to mutation. The best example is gene therapy for RPE65-associated LCA. In 2008, three groups reported success using an AAV2 vector to deliver a healthy RPE65 gene to the retina of three LCA patients (2931). In these trials, patients recovered some visual function without side effects, sparking renewed interest worldwide in using gene therapy for treating retinal diseases. These findings recently have been confirmed in 12 patients with disease-causing RPE65 mutations, who were followed for up to 2 years after a single subretinal injection of the AAV2 vector carrying the therapeutic gene. There were no serious adverse events, and improvements were reported in one or more measures of visual function in 9 of the 12 patients (32).

This first set of gene therapy trials also tackled some of the key questions concerning the benefits of early intervention (33), optimal dosing (34), and surgical technique (35). For example, one clinical study suggested that subretinal injections were associated with procedural risks in the foveal region of the retina (35). Disease progression affects the retinal architecture, rendering it prone to damage and detachment during surgery. The risk of compromising residual central vision in late-stage patients may represent a roadblock for subretinal surgical manipulation, especially in diseases like X-linked retinoschisis where the retinal architecture is particularly fragile. New developments in viral vector delivery may overcome these hurdles; for example, the newest AAV vectors administered intravitreally can deliver therapeutic genes deep into the retina (36, 37). One such AAV vector, AAV.7m8, mediated widespread gene delivery to the outer retina and rescued the disease phenotype in a mouse model of X-linked retinoschisis (Rs1h−/− mouse; Fig. 2) and RPE65-associated LCA type 2 (rd12 mouse) (37).

Fig. 2. Gene replacement therapy for retinal degeneration.

The illustration shows gene replacement therapy using the AAV vector AAV.7m8 carrying the therapeutic Rs1 gene in a mouse model (Rs1h−/−) of X-linked retinoschisis. Intravitreal injection of AAV.7m8 resulted in delivery of a healthy copy of the Rs1 gene to rod photoreceptor cells of the Rs1h−/− mouse. This resulted in the secretion of retinoschisin, a protein that helps to maintain retinal integrity, resulting in structural and functional improvements in the Rs1h−/− mouse retina (4).

Credit: H. McDonald/Science Translational Medicine

Another important result of the LCA clinical trials was the finding by one group that the photoreceptor cells may continue to degenerate in treated patients (19) and that the initial functional improvements in vision may deteriorate in the midterm (38, 39). However, data from another group showed that both visual function and retinal thickness remained stable for at least 3 years after gene therapy administration apart from one patient who developed a macular hole that may have been related to the surgical procedure rather than the gene therapy vector per se (40). These findings suggest either an unsolved dosage problem or the need for a combination therapy that restores vision while also protecting the photoreceptor cells. This point has strong implications for gene replacement therapy, which is effective only if sufficient target cells remain. Therefore, identifying patients with sufficient retinal tissue using imaging, such as spectral-domain optical coherence tomography (SD-OCT), and functional tests will be key to successful gene therapy and combinatorial gene therapy that includes codelivery of the therapeutic gene together with a gene encoding a trophic factor.

Other positive outcomes for gene replacement therapy have emerged from a clinical trial for choroideremia, a rare X-linked recessive inherited disorder with degeneration of the choriocapillaris and the RPE that eventually leads to blindness (41). In this clinical trial launched in late 2011, six patients with choroideremia due to a mutated REP1 gene received a modified AAV vector containing the therapeutic gene delivered subretinally and targeted to photoreceptor cells in the fovea. Other clinical trials using AAV vectors are aiming to treat MERTK-associated retinal dystrophy (NCT01482195), Leber hereditary optic neuropathy (NCT02161380, NCT01267422, and NCT02064569), and X-linked retinoschisis (NCT02416622 and NCT02317887). AAV vectors may also become a potential tool for delivery of genome-editing molecular machineries, such as clustered regularly interspaced short palindromic repeats (CRISPR)/Cas9 (CRISPR-associated protein) (42). Genome editing with CRISPR/Cas9 may prove especially useful for the treatment of inherited retinal degenerative diseases caused by autosomal dominant mutations (43). Another approach for treating dominantly inherited retinal degenerative diseases, especially those involving a dominant-negative molecular mechanism, would be a suppress-and-replace strategy (44). Further challenges for gene therapy include different transfection rates in the retina and RPE, as well as the size limitations imposed by the viral vectors. Twenty patients with STGD are currently enrolled in a phase 1/2a clinical trial to assess the safety and efficacy of gene replacement therapy using lentiviral vectors to deliver a replacement ABCA4 gene. Lentiviral vectors also are being used for gene therapy to treat RP due to Usher syndrome type 1B caused by mutations in the MYO7A gene. The first gene therapy trials for patients with achromatopsia are also on the horizon (NCT02341807).

As research advances continue in gene sequencing, diagnostics, patient selection, delivery vectors, and surgical techniques, and clinical trial data continue to accumulate, there is a growing conviction that gene therapy holds promise not only for treating inherited retinal degenerative diseases but also for protecting the retina from further deterioration.

Optogenetic therapy

Optogenetic therapy is a way of restoring vision by targeting specific cell types in the retina with “optogenes” that enable the cells to become light-sensitive (45, 46). The best application of optogenetic therapy may be for blindness caused by degeneration of photoreceptor cells. With a single injection into the eye, this therapy has the potential to restore some long-term visual function. A requirement is the presence of at least some retinal cell types in the degenerating retina, such as bipolar and ganglion cells, which can be targeted by optogenes. Optogenetic therapy does not rely on specific genetic mutations and therefore could be used to treat a wide variety of inherited retinal degenerative diseases. In a study in both a mouse model of RP and in ex vivo human retinal tissue, expression of an archaebacterial halorhodopsin optogene in light-insensitive cone photoreceptor cells was shown to substitute for the native phototransduction cascade and to restore light sensitivity to the cells (47). In the RP mouse, resensitized photoreceptor cells exhibited a full visual cycle, activated cortical circuits, and mediated visually guided behaviors (48).

Important factors for vision restoration by optogenetic therapy are the speed at which the target cells can react to and detect motion, the light sensitivity achieved, and the absorption spectrum of the target cells, which is optimally red-shifted to avoid light damage (47). There are a wide variety of optogenes available (47, 4953), and the choice depends on careful ranking of the advantages and disadvantages of each. The cell types targeted can be the remaining light-insensitive cones in the fovea (48), bipolar cells (54), or ganglion cells (55). Careful assessment of the status of these remaining retinal cells using high-resolution noninvasive imaging will inform the decision about which retinal cell type to select. Targeting of optogenes can be achieved with AAV vectors. The design of the vectors and their preclinical evaluation in nonhuman primates will likely be a major determinant of success. Given that only one optogene will be used initially, color vision will not be restored, but in the future this may be possible by combining different optogenes (47). The degree of restored vision will depend on the targeted cell type and the state of degeneration of the patient’s retina. Optimum resolution could be restored if central light-insensitive cones in the fovea are still functional and can be targeted. A major limitation of optogenetic therapy is the lack of adaptation of the engineered light-sensitive cells. Therefore, external goggles that record the visual scene and project it onto the retina in a spectrum and intensity that matches that of the optogene-expressing target cells will be required (46). The outcome of the therapy will depend on many factors, such as the degenerative state of the retina, the ability of the patient to learn the new retinal language driven by the optogene, and the rehabilitation program provided. Once optogenetic therapy reaches clinical trials, the treatment will likely be improved by learning from the visual experience of patients and then redesigning the optogenes, the targeting vectors, and the encoding of the projected visual stimuli.

Retinal prostheses

Retinal prostheses are intended to restore some visual function in patients suffering from photoreceptor cell degeneration and vision loss associated with various retinal diseases from inherited retinal degenerative diseases to age-related macular degeneration (AMD). The principle is to introduce visual information into the degenerated retina through electrical stimulation of neurons in the inner retinal layer. Although inner retinal neurons also degenerate after photoreceptor cell loss, different studies have shown that a substantial number of these neurons persist even at very late stages of disease (56, 57); however, a profound remodeling of the neurons and their synaptic contacts does occur (58). Despite these major changes in the inner retina, the first clinical trials demonstrated that the residual inner retinal neurons can still transfer useful visual information to the brain.

Different strategies to activate inner retinal neurons with electrode arrays have been developed. An electrode array can be inserted in between the sclera and choroid (59, 60), in the subretinal space (61), in the vitreous body on the surface of the retina (epiretinal implant) (62, 63), or around the optic nerve (optic cuff) (64). All of these strategies have demonstrated efficacy in generating phosphenes (visual sensations typically induced by mechanical, electrical, or magnetic stimulation) in patients implanted with these prostheses. The greater visual resolution of subretinal and epiretinal implants has enabled the reading of single short words by patients receiving these implants (61, 62). Both epiretinal and subretinal implants have received regulatory approval.

The newest devices aim to restore useful visual information for face and object recognition, as well as text reading and orientation in unknown environments. Psychophysical studies have shown that devices generating visual information through 625 independent pixels projected on a foveal visual field of 1.7° (or on a 10° × 7° viewing window) stabilized at 15° eccentricity should enable patients to achieve these goals (6567). Different strategies for electrode distribution and geometry have been proposed to increase the resolution of individual electrodes in the retinal implant. These strategies include the use of bipolar (68) and quasi-monopolar electrical stimulation (69) and the introduction of a ground grid surrounding all stimulating electrodes (68). The greatest resolution would be expected with a very conductive ground grid, which has motivated research into innovative semiconductor materials such as graphene or diamond (70, 71). The ground grid configuration could explain the high resolution achieved with photovoltaic subretinal implants recently demonstrated in blind rats (72). Alternative strategies are investigating flexible photosensitive polymers that can adapt to the curvature of the eye (73, 74). Clinical studies are needed to confirm the potential of these new devices to produce a high visual acuity in blind patients.

Stem cell therapy

In clinical situations characterized by the absence of cell targets for gene therapy, such as in the advanced stages of retinal degeneration where photoreceptor cells have already degenerated, retinal cell replacement could be a means to preserve or restore vision. Stem cells, defined by their capacity for proliferation and pluripotency, could provide an unlimited source of appropriately differentiated replacement cells. Human embryonic stem cells (hESCs) are derived from the human blastocyst, whereas human induced pluripotent stem cells (hiPSCs) can be generated from human fibroblasts or other adult cell types. The cell types needed to be replaced to ensure regeneration of the outer retina are photoreceptor and RPE cells. RPE cells are not photosensitive, but they support photoreceptor cell survival and function.

Landmark clinical trials of hESC-derived RPE cells transplanted into the eyes of human patients are in progress. Building on preclinical data (75), Schwartz et al. recently reported the results of two phase 1/2 clinical trials (NCT01345006 and NCT01344993) (76). The trials enrolled patients with STGD and atrophic AMD. The cell transplants were targeted to the transition zone between degenerating and normal retina with the goal of rescuing photoreceptor cells in this zone by replenishing the RPE. The authors reported increased subretinal pigmentation consistent with engraftment of transplanted hESC-derived RPE cells in 13 of 18 patients enrolled; there was no evidence of hESC-related safety issues due to adverse proliferation of the cell transplant or immune rejection. This study involved perioperative immunosuppression of patients. The favorable safety and tolerability profile of hESC-derived RPE cell transplantation was echoed in an Asian cohort of four recipients also with AMD or STGD (77). An ongoing trial at University College London, sponsored by Pfizer, is treating patients with exudative AMD using a monolayer of hESC-derived RPE cells immobilized on a polyester membrane (NCT01691261). These pioneering clinical trials should reveal the relative merits of hESCs and hiPSCs for clinical applications not only in terms of safety and efficacy but also in terms of RPE cell generation efficiency and consistency. Differences in surgical methods and engraftment success with cell suspensions versus transplantation of RPE monolayers will also be important to evaluate.

RPE cells lack light sensitivity and so cannot restore vision once the photoreceptor cell layer has completely degenerated. Photoreceptor cell transplantation poses specific challenges, including the requirement for synaptic integration and the elaboration of a polarized and compartmentalized neuronal cell structure. Regenerating the lost photoreceptor cell layer by transplanting donor-derived rod photoreceptor precursor cells (78) has been shown to restore visual function in rd1 mice with retinal atrophy (79), but this has not yet been tested clinically. It has been suggested (80) that a coexisting strain-specific mutation in Gpr179 (81, 82) causing inner retinal dysfunction could have influenced most or all studies that have used the rd1 mouse model since 1948. However, a recent survey showed that Gpr179 mutations are not as ubiquitous in rd1 mice as once thought (83). Notably, photoreceptor cell replacement has also been investigated in animals with partial or early retinal degeneration (8487). Photoreceptor-like cells derived from hESCs (88) and hiPSCs (87, 89) have been successfully transplanted into animal models, suggesting that renewable stem cell sources can support human photoreceptor cell replacement. Because RPE cells also may be compromised along with photoreceptor cells, a multilayered outer retinal construct of photoreceptor cell and RPE cell layers is a future goal (90). To this end, exciting advances in three-dimensional microorganogenesis (91, 92) enabling the growth of stem cell–derived primitive optic cups in vitro may allow the generation of RPE and photoreceptor cell layers in tandem. Recently, a study from Japan investigated the potential use of hESC-derived neuroretinal sheet grafts transplanted into nonhuman primates (93). These grafts were observed differentiating into a range of retinal cell types including rod and cone photoreceptor cells, and there also seemed to be synaptic connections between the transplant and the host retina.

Most of the current data on the optimization of donor photoreceptor cell transplantation relate to rod photoreceptor cells (86, 94, 95), which normally contribute to peripheral and dim-light vision. In an important recent development, daylight vision improvements were reported in a preclinical study of cone-like cell transplantation into the Cpfl1 mouse that lacks cone photoreceptors (96). Cell replacement approaches are attractive as they are broadly applicable to many inherited retinal degenerative diseases because they act independently of the specific genetic defect.

New tools and outcome measures to accelerate clinical trials

Clinical trials for inherited retinal degeneration are likely to require special considerations with respect to outcome measures. In these retinal diseases, both peripheral and central vision may be affected either in a centrifugal (initially affecting the macula and the periphery later) or in a centripetal (initially affecting the periphery) manner. Therefore, outcome measures must be chosen according to the specifics of the trial design and the aim of the intervention (that is, restoring vision or slowing down retinal degeneration) (97). The outcome measures should have high reliability and repeatability and permit standardization across study sites using established normative data. They should derive from tests that are ideally noninvasive, safe, easy, and quick to perform to minimize the burden placed on both patients and staff (98). Visual acuity has been the primary outcome measure for efficacy evaluation in most treatment trials for retinal diseases, but only a change of ≥15 letter scores (equivalent to ≥3 lines on a standardized visual acuity chart) is considered significant (99). In inherited retinal degenerative diseases with diffuse generalized photoreceptor cell loss such as in RP, visual acuity typically declines slowly and may be only affected in late stages when the damage begins to involve cone photoreceptor cells in the very center of the retina.

In a recent study, using adaptive optics scanning laser ophthalmoscopy (AOSLO), good visual acuities (≥80 letters corresponding to 20/25 or better) were still maintained in patients with retinal dystrophies, whereas cone densities in vivo were 52% below normal (100). AOSLO imaging capabilities have improved sufficiently to allow visualization of the retina at cellular resolution (101); however, further refinements to enhance resolution, reliability, and repeatability are needed before AOSLO can be used in clinical trials. SD-OCT delivers tomographic images of retinal layers with a resolution similar to that of histological tissue sections (Fig. 3) (102). This imaging modality, in combination with segmentation analysis, also provides quantitative measures of retinal structure that facilitate the precise tracking of disease progression (103, 104). Further refinements have led to the development of OCT angiography, which is a pioneering noninvasive technique based on a highly efficient split-spectrum amplitude-decorrelation angiography algorithm (105). This technique provides information about both retinal and choroidal vascular blood flow and is particularly useful in evaluating vascular dysfunction (105). Such in vivo high-resolution imaging modalities promise to enhance early diagnosis and may provide surrogate markers to monitor the efficacy of therapies. For example, in three patients with RP treated over a 2-year period with CNTF, AOSLO imaging was applied to evaluate the structure and distribution of cone photoreceptor cells and revealed reduced rates of cone loss in the CNTF-treated eye of patients compared to the contralateral sham-treated eye. In contrast, there were no significant changes in visual acuity, visual sensitivity, and electroretinography responses (106). The U.S. Food and Drug Administration recommends visual function changes as a primary endpoint in measuring the efficacy of any new compound for treating eye diseases but may consider anatomical markers such as the area of RPE atrophy measured by autofluorescence of the fundus (107). To serve as a surrogate endpoint, a marker must be “reasonably likely, based on epidemiological, therapeutic, pathophysiological, or other evidence, to predict clinical benefit” (107), and the best surrogate endpoint is a marker that changes along with clinical endpoints (107). In vivo imaging of the retina at a histological-level resolution could be combined with functional high-resolution mapping, for example, using microperimetry, to establish structure-function correlations that would enable the establishment of structural endpoints and the development of new outcome measures (108). New software tools like visual field and modeling analysis can help to create topographical models of the hill of vision derived from microperimetry and static visual fields to aid the visualization of field defects and quantify the magnitude and extent of the loss of retinal sensitivity (109). In a clinical trial, volumetric static analyses showed greater power for detecting changes to the retina compared to kinetic perimetry (32). The approach of structure-function correlations has been implemented in one of the largest natural history studies so far in the field of inherited retinal degenerative diseases: The Natural History of the Progression of Stargardt Disease: Retrospective and Prospective Studies (ProgStar, NCT01977846) (110).

Fig. 3. Advanced noninvasive imaging of the retina.

(A) The retina can be imaged using SD-OCT (black box) and AOSLO. (B) Shown is an SD-OCT tomographic section of the retina through the foveal area in a healthy subject. The retinal lamina can be visualized and quantified, leading to surrogate marker endpoints. Photoreceptor cells (white boxes) can be visualized at cellular resolution with AOSLO. (C to E) Shown are AOSLO images of rod and cone photoreceptor cells from the same subject at two different locations (scale bar, 75 μm). (C) Single cone and rod photoreceptor cells are visible. On the basis of different reflectivity, they may appear brighter or darker, as shown in (D). (E) At the fovea of the retina, cone photoreceptor cells are densely packed. Images were acquired using a custom-built AOSLO device at the Moorfields Eye Hospital, London, U.K. and are courtesy of M. Michaelides and A. Dubis.

Credit: H. McDonald/Science Translational Medicine


Emerging therapies, either alone or in combination, will further expand the treatment options for patients with inherited retinal degenerative diseases across the severity spectrum (Fig. 4). Currently, only a small segment of patients—those with severe RP, who had vision in the past but now barely have light perception—are eligible for retinal prosthesis surgery approved by regulatory agencies. It remains to be seen whether, with further preclinical development, photoreceptor cell replacement or optogenetic therapy will provide superior visual gains in patients with profound vision loss from late-stage inherited retinal degenerative diseases. For those with better anatomical and functional reserve at baseline, pharmacotherapy, gene therapy, neuroprotection, or RPE transplantation may be the best strategies for decelerating disease progression, preserving residual vision, or augmenting visual function. With gene therapy in phase 3 trials (for example, NCT00999609 for RPE65-LCA) and with stem cell–derived RPE transplantation in phase 2 trials (for example, NCT01345006 for STGD), it is conceivable that gene therapy will be the next new treatment to become available for specific patients with retinal degenerative diseases.

Fig. 4. Different strategies for treating retinal degeneration.

Gene replacement therapy is appropriate during the early stages of disease when retinal photoreceptor cells are still intact. Pharmacotherapy and neuroprotective strategies are the best approaches to treat the disease with ongoing photoreceptor cell degeneration. Stem cell therapy, optogenetic therapy, and retinal prostheses are needed to restore vision during the later stages of retinal degeneration. STGD is an example of a retinal degenerative disease where gene therapy, pharmacotherapy, and stem cell therapy are being tested in early clinical trials. [Figure modified from (19).]

Credit: H. McDonald/Science Translational Medicine

The cost-effectiveness of these emerging therapies for rare retinal diseases is a significant issue to consider. According to one analysis (111), the Argus II retinal prosthetic implant was found to be a cost-effective intervention compared to the usual cost of care for RP patients, taking into account the cost of professional assistance for daily activities and social benefits received for visual disabilities. Currently, orphan drugs represent some of the highest per-patient costs in medicine, and there is an inverse relationship between the annual cost of an orphan drug and the prevalence of the target disease (112). Thus, interventions, such as gene and stem cell therapies, could be made more cost-effective by broadening their applicability to all inherited retinal degenerative diseases instead of targeting specific diseases. One example of a disease-independent treatment strategy is stem cell transplantation; however, the pursuit of personalized stem cell medicine using patient-derived hiPSCs could drive up costs because of the need for batch-specific safety and quality assays for hiPSCs derived from each individual patient. However, in certain populations with a high degree of genetic homogeneity, one cost-saving strategy would be to establish a limited bank of specified donor cells that could serve a large proportion of recipients in this population. For example, it has been estimated that donor cells derived from 150 selected homozygous human leukocyte antigen (HLA)–matched individuals could match 93% of the U.K. population (113). The challenge remains to develop cost-effective treatments that are tailored to deliver improved vision to large groups of patients with inherited retinal degenerative diseases at early, intermediate, and late stages of disease, ideally, without being restricted to those with specific genetic defects. A broad phenotype-based treatment strategy exploiting common disease pathways will expand the number of patients who can be treated compared to strictly genotype-based approaches.

In summary, there is a wide range of emerging therapeutic approaches for treating inherited retinal degenerative diseases that may be applicable to a variety of incurable conditions with a broad spectrum of severity.


  1. Acknowledgments: We thank M. Michaelides and A. Dubis for the AOSLO images and M. Kasilian for assistance with the figures. Funding: J.-A.S. is supported by grants from Labex LifeSenses (ANR-10-LABX-65), Banque publique d’investissement, the Foundation Fighting Blindness (FFB), ERC Synergy grant “HELMHOLTZ,” Pixium Vision, GenSight Biologics, Sanofi-Fovea, and Gene Signal. H.P.N.S. was supported by the FFB Clinical Research Institute (CRI); a grant to FFB CRI by the U.S. Department of Defense USAMRMC TATRC (W81-XWH-07-1-0720 and W81XWH-09-2-0189), the Shulsky Foundation, and the Ocular Albinism Research Fund (Clark Enterprises Inc.); and an unrestricted grant to the Wilmer Eye Institute from Research to Prevent Blindness, Baylor–Johns Hopkins Center for Mendelian Genetics (National Human Genome Research Institute/NIH). R.W.S. was supported by the Austrian Science Fund (FWF #J3383-B23) and the Foundation Fighting Blindness Clinical Research Institute. M.S.S. was supported by the Shulsky Foundation. B.R. was supported by the Gebert RüF Foundation, the SNSF, Synergia, the NCCR Molecular Systems Engineering, the European Research Council, and the European Union projects SeeBetter, TREATRUSH, OptoNeuro, and 3×3D Imaging. Competing interests: J.-A.S. is a founder of Fovea Pharmaceuticals, Ophtimalia, Streetlab, Pixium Vision, GenSight Biologics, Chronocam, and Chronolife. J.-A.S. is a co-inventor on pending patent U.S. 8394756 B2 (Methods of increasing RdCVF1 or RdCVF2 polypeptides in retinal cells), pending patent U.S. 8957043 B2 (Methods of treating retinitis pigmentosa using nucleic acids encoding RdCVF1 or RdCVF2), and pending patent U.S. 8071745 B2 [Polynucleotides encoding rod-derived cone viability factor (rdcvf) and methods of using the same]. H.P.N.S. is a paid consultant of Boehringer Ingelheim Pharma GmbH & Co. KG, Daiichi Sankyo Inc., Gerson Lehrman Group, Guidepoint, and Shire. H.P.N.S. is a member of the Scientific Advisory Board of the Astellas Institute for Regenerative Medicine, GenSight Biologics, Intellia Therapeutics Inc., and Vision Medicines Inc. H.P.N.S. is member of the Data and Safety Monitoring Board or Committee of the following entities (not including the NIH): Genentech Inc./F. Hoffmann–La Roche Ltd, Genzyme Corp./Sanofi, and ReNeuron Group Plc/Ora Inc. D.D. has received grants and personal fees from GenSight Biologics and is co-inventor on patent #9193956 (Adeno-associated virus virions with variant capsid and methods of use thereof), with royalties paid to Avalanche Biotechnologies. B.R. is the chair of the Scientific Advisory Board of GenSight Biologics.
View Abstract

Stay Connected to Science Translational Medicine

Navigate This Article