Certain types of retinal damage can be repaired surgically, while others remain beyond the reach of current medicine. The retina, unlike skin or bone, has almost no built-in ability to regenerate lost neurons in humans. That biological limitation shapes everything about retinal treatment: surgeons can reattach a detached retina and restore good vision, lasers can stabilize diabetic eye disease, and injections can halt abnormal blood vessel growth, but once the light-sensing photoreceptor cells die, no approved therapy can regrow them. The landscape is shifting, though, with gene therapy already on the market for one inherited condition and stem cell transplants, electronic implants, and optogenetic approaches in various stages of testing.
Why the Retina Cannot Heal Itself
Fish like zebrafish can regenerate an injured retina almost completely. They do this through cells called Müller glia, which reprogram themselves into progenitor cells capable of producing new neurons. Mammalian Müller glia share some of that stem-cell-like potential, but their native regenerative capacity is extremely limited.1PubMed Central. Insights on the Regeneration Potential of Müller Glia in the Mammalian Retina In mice and humans, Müller glia respond to injury with scarring and inflammation rather than by producing replacement neurons.2PubMed. Age-dependent Müller glia neurogenic competence in the mouse retina Researchers have tried coaxing mammalian Müller glia to behave more like their fish counterparts under controlled lab conditions, but the efficiency remains far too low for clinical use.3PubMed Central. Advances in the study of Müller glia reprogramming in mammals
This regeneration gap means that every treatment strategy for retinal disease falls into one of two categories: stop the damage before cells die, or find a way to replace or work around the cells that have already been lost. The first category is where current medicine excels. The second is where most of the experimental work lives.
How Retinal Cells Are Lost
Photoreceptors, the rods and cones that convert light into electrical signals, are the cells most commonly lost in retinal disease. When the retina detaches from its supporting layer, photoreceptor death begins quickly. In animal models, dying cells appear within a day of detachment, peak around day three, and then drop off sharply after a week.4PubMed. Preventing retinal detachment-associated photoreceptor cell loss in Bax-deficient mice If the detachment persists, losses mount rapidly: one experimental study found that roughly half the photoreceptor layer was gone by two weeks, and only about a tenth remained after a month.5PubMed. Photoreceptor decay over time and apoptosis in experimental retinal detachment
The death pathways are complex. Photoreceptor loss was long attributed solely to apoptosis, but research now shows that other forms of programmed cell death, including autophagy and a regulated form of necrosis, also play significant roles. Blocking any single pathway alone does not fully protect photoreceptors, which is one reason neuroprotective drugs have been difficult to develop.6PubMed Central. Photoreceptor cell death and rescue in retinal detachment and degenerations
Photoreceptor loss can also trigger downstream problems. Some research indicates that as photoreceptors degenerate, changes cascade through the retina, eventually affecting retinal ganglion cells, the neurons that carry visual information to the brain. These changes can include vessel displacement, tissue remodeling, and ganglion cell death over time.7PubMed Central. Retinal Ganglion Cell Death as a Late Remodeling Effect of Photoreceptor Degeneration That said, other animal studies have found certain ganglion cell types surviving well over a year after complete photoreceptor loss, suggesting the cascade is not universal and that the inner retina may remain a viable target for intervention even in advanced disease.8PLOS ONE. Retinal Ganglion Cells are Resistant to Photoreceptor Loss in Retinal Degeneration
Reattaching a Detached Retina
Retinal detachment is the condition where the word “repair” most clearly applies. Surgery physically repositions the retina against its nutrient-supplying layer, and two main techniques dominate. Scleral buckling involves stitching a silicone band to the outside of the eye, which pushes the wall inward to meet the detached retina. Pars plana vitrectomy takes the opposite approach: the surgeon goes inside the eye, removes the vitreous gel, uses a laser to seal the retinal tear, and fills the cavity with a gas bubble or silicone oil to press the retina flat.9PubMed Central. Pars plana vitrectomy versus scleral buckling for repairing simple rhegmatogenous retinal detachments Both approaches have high anatomical success rates for uncomplicated detachments, and the choice between them depends on the location and character of the tear, the patient’s lens status, and the surgeon’s judgment.
Visual recovery after detachment repair depends heavily on whether the macula, the central high-resolution part of the retina, was involved. When the macula stays attached, outcomes tend to be excellent. When the macula detaches (“macula-off”), about three in five patients still recover to 20/50 or better, roughly a third end up between 20/60 and 20/200, and only about one in twenty are left with vision worse than 20/200.10PubMed. Visual recovery in macula-off rhegmatogenous retinal detachments Interestingly, within the first week of macular detachment, the exact day of surgery does not appear to change the outcome significantly. That finding reassures both patients and surgeons that a day or two of scheduling delay in that window is unlikely to cost vision.
Two factors that do predict recovery are the duration and height of the macular detachment. Shorter detachment times and a flatter separation between the retina and its supporting tissue correlate with better final acuity.11PubMed Central. Postoperative recovery of visual function after macula-off rhegmatogenous retinal detachment This makes intuitive sense: less separation means less oxygen deprivation for the photoreceptors, and less time detached means fewer cells lost before surgery.
Laser Photocoagulation for Diabetic Eye Disease
Diabetic retinopathy does not detach the retina in the same way a tear does, but it threatens vision through leaking and abnormal blood vessels. Laser photocoagulation has been a workhorse treatment for decades. The mechanism is counterintuitive: the laser deliberately destroys patches of peripheral retinal tissue. Those patches were consuming oxygen, and eliminating them allows more oxygen from the blood supply to reach the inner retina. As oxygen tension rises, the signals that drive abnormal vessel growth, particularly a molecule called VEGF, decrease, and the destructive cycle of new vessel formation and leaking slows or stops.12PubMed Central. Laser Treatment for Diabetic Retinopathy: History, Mechanism, and Novel Technologies
Laser treatment does not restore lost vision. It stabilizes the disease, trading peripheral retinal tissue you can afford to lose for the survival of central vision you cannot. Patients sometimes notice a modest reduction in peripheral vision or night vision after extensive laser treatment, but the tradeoff is overwhelmingly favorable compared to untreated proliferative diabetic retinopathy, which can lead to blindness.
Anti-VEGF Injections
For conditions driven by abnormal blood vessel growth, including wet age-related macular degeneration and diabetic macular edema, injections of anti-VEGF drugs directly into the eye have become the standard first-line therapy. These drugs block the molecular signal that promotes leaky, destructive new blood vessels. In patients with choroidal neovascularization (abnormal vessels growing beneath the retina), anti-VEGF treatment over twelve months led to a meaningful reduction in retinal thickness, and about 44% of treated eyes gained fifteen or more letters of visual acuity on a standard eye chart.13PubMed Central. Results of Intravitreal Anti-VEGF Injection in Choroidal Neovascularization Caused by Pathologies Other Than Age-Related Macular Degeneration
The catch is that anti-VEGF therapy requires repeated injections, sometimes monthly or bimonthly, for years. Newer formulations aim to extend the interval between injections, but for now, the treatment burden remains one of the major challenges patients face. The drugs also cannot reverse structural damage that has already occurred. They are most effective when started early, before extensive scarring or cell loss sets in.
Gene Therapy for Inherited Retinal Disease
The first gene therapy approved for any inherited retinal disease is voretigene neparvovec, marketed as Luxturna. It treats patients who carry mutations in both copies of the RPE65 gene, a condition that causes Leber congenital amaurosis and other forms of severe early-onset vision loss.14PubMed. Post-approval outcomes of voretigene neparvovec (Luxturna®) retinal gene therapy: A systematic review and meta-analysis The treatment delivers a working copy of RPE65 into the retinal pigment epithelium using an engineered virus, restoring a key step in the visual cycle that these patients’ cells cannot perform on their own.15PubMed Central. Gene Therapy Using Recombinant Adeno-Associated Virus for Leber Congenital Amaurosis Induced by RPE65 Mutation
Luxturna works best in patients who still have surviving photoreceptors for the corrected RPE cells to support. It does not regrow photoreceptors that have already died, so the window for treatment matters. This underscores a theme across retinal repair: interventions are most powerful when they happen before irreversible cell loss. The list of genetic targets for retinal gene therapy is growing rapidly, with dozens of clinical trials in progress for conditions like retinitis pigmentosa, Stargardt disease, and achromatopsia.
Stem Cell Transplantation
Where gene therapy corrects a malfunctioning cell, stem cell transplantation aims to replace cells that are gone. The most clinically advanced approach involves transplanting retinal pigment epithelium (RPE) cells, the support layer beneath the photoreceptors that degenerates in macular degeneration. In a first-in-human trial for dry age-related macular degeneration, researchers transplanted RPE progenitor cells derived from adult RPE stem cells into the subretinal space. The treatment was safe, with no tumors or serious inflammation, and the average visual acuity gain was about 21 letters at one month and held steady through twelve months.16Cell Stem Cell. First-in-human clinical trial of adult retinal pigment epithelium stem cell-derived progenitor cell transplantation for dry age-related macular degeneration
Safety is a central concern with any stem cell therapy. Pluripotent stem cells have the theoretical potential to form tumors, which is why extensive safety testing precedes human trials. In one study examining the tumorigenic potential of iPSC-derived RPE cells, no tumors formed after subcutaneous transplantation over fifteen months of monitoring, and no tumors were found after subretinal transplantation in animal models over six to twelve months.17PubMed Central. Tumorigenicity studies of induced pluripotent stem cell (iPSC)-derived retinal pigment epithelium (RPE) for the treatment of age-related macular degeneration These results are encouraging, but the field is still early. Transplanting RPE cells to support surviving photoreceptors is one thing; replacing photoreceptors themselves is far harder.
Photoreceptor transplantation has shown promise in animal models. Researchers have transplanted rod precursor cells into mice that lack functional rods, and the transplanted cells formed working synaptic connections, responded to light, and drove visually guided behavior.18PubMed Central. Restoration of vision after transplantation of photoreceptors Translating this to humans is a major engineering challenge, however. Transplanted cells need to survive, integrate into the correct retinal layer, wire up to the right downstream neurons, and function over the long term. One of the biggest hurdles is the immune environment: microglia, the retina’s resident immune cells, play a pivotal role in determining whether transplanted cells survive or are rejected.19PubMed Central. Microglia Activation and Immunomodulatory Therapies for Retinal Degenerations
Scaffolds and Tissue Engineering
Delivering individual cells into the subretinal space and hoping they land in the right orientation is inefficient. One approach to improve survival and integration is to grow cells on thin scaffolds that can be surgically implanted as a sheet. Researchers have developed 3D-printed scaffolds with pores sized to hold individual RPE cells, and testing in pig and rabbit eyes showed the scaffolds were stable, biocompatible, and caused no adverse effects on retinal structure or function over periods up to fourteen months.20PubMed. Single-cell-pore-sized 3D printed scaffolds for retinal pigment epithelial cell therapy Scaffold-based delivery could help transplanted cells maintain their polarity and barrier function, both of which are critical for RPE to do its job properly. The same concept is being explored for photoreceptor progenitor delivery, where spatial organization matters even more.21PubMed Central. Scaffolds and stem cells: delivery of cell transplants for retinal degenerations
Electronic Retinal Implants
For patients who have already lost most of their photoreceptors, electronic retinal implants bypass the dead cells entirely. These devices use electrode arrays placed on or beneath the retina to electrically stimulate the surviving inner retinal neurons. Devices like the Argus II and Alpha AMS, designed for retinitis pigmentosa patients, have demonstrated the ability to restore basic visual function: users can perceive light, detect motion, and recognize large objects.22PubMed Central. Can bionic eyes restore vision? Breakthroughs, challenges, and future frontiers in ophthalmology – A comprehensive review
A newer device, the PRIMA subretinal microchip, takes a different approach for patients with atrophic (dry) macular degeneration. Rather than an external camera, PRIMA uses photovoltaic pixels that convert projected infrared light into electrical stimulation. At four years of follow-up, patients could reliably recognize letters and sequences of letters, and with a zoom feature, prosthetic acuity improved by up to eight lines on a standard eye chart.23PubMed Central. Prosthetic Visual Acuity with the PRIMA Subretinal Microchip in Patients with Atrophic Age-Related Macular Degeneration at 4 Years Follow-up That is a meaningful gain, though the resulting vision is still far from normal. These devices offer a functional floor rather than a cure, giving patients enough visual input to navigate their environment and perform some daily tasks.
Optogenetics and Light-Sensitive Proteins
Optogenetic therapy tries to create artificial photoreceptors by genetically modifying surviving retinal cells to produce light-sensitive proteins. The idea is that even after photoreceptors die, cells deeper in the retina, such as bipolar cells or ganglion cells, can be made responsive to light if they express the right proteins. In blind mouse models, researchers found that expressing a channelrhodopsin protein in retinal ganglion cells restored light sensitivity and visual acuity more effectively than targeting bipolar cells, at least with current gene delivery methods.24PubMed Central. Comparison of AAV-Mediated Optogenetic Vision Restoration between Retinal Ganglion Cell Expression and ON Bipolar Cell Targeting Early human trials of optogenetic therapies are underway, though the vision restored so far is rudimentary, and the approach typically requires special goggles that amplify light to the wavelengths the engineered proteins respond to.
Neuroprotective Drugs
Rather than replacing dead cells, neuroprotective strategies aim to keep threatened cells alive longer. Several classes of compounds are under investigation, including bile acids like TUDCA, steroid hormones like progesterone, therapies targeting the retinal dopamine system, and various neurotrophic factors, which are natural proteins that promote neuronal survival.25PubMed Central. Neuroprotective strategies for retinal disease None of these approaches has reached mainstream clinical use yet. Part of the difficulty, as noted earlier, is that photoreceptor death involves multiple overlapping pathways, so blocking any single one provides incomplete protection.6PubMed Central. Photoreceptor cell death and rescue in retinal detachment and degenerations The most effective strategy may eventually involve cocktails targeting several death pathways simultaneously, but that adds complexity and safety concerns.
Tracking Recovery After Surgery
Optical coherence tomography (OCT), a non-invasive imaging technique that produces cross-sectional pictures of the retina, has become essential for monitoring how well the retina heals after surgery. After macular hole repair, for instance, OCT can show whether specific retinal layers have regenerated. In one series, the external limiting membrane restored in all eyes after successful closure, but the ellipsoid zone, a marker of photoreceptor integrity, recovered in about 69%, and the cone outer segment tips line in only about 17%.26PubMed Central. Inner Retinal Layer and Outer Retinal Layer Findings after Macular Hole Surgery Assessed by means of Optical Coherence Tomography Recovery of these deeper layers progresses slowly, sometimes over a year or more, spreading from the edges of the closed hole toward its center. Persistent defects on OCT at these layers predict worse final acuity, giving surgeons a way to counsel patients about realistic expectations.
Factors associated with better ellipsoid zone recovery include shorter eye length, longer follow-up time, smaller initial hole size, and sufficient inner retinal thickness at the repair site.27Scientific Reports. Changes in each retinal layer and ellipsoid zone recovery after full-thickness macular hole surgery The practical takeaway is that recovery from retinal surgery is not a single event. It unfolds over months, and the fine structure of the retina can keep improving well after the initial wound has healed.
When the Brain Has to Relearn How to See
Even when a retinal intervention succeeds in sending visual signals to the brain, the brain’s ability to interpret those signals is not guaranteed. This is especially relevant for people born with severe visual impairment. Studies of individuals whose congenital cataracts were surgically removed found that their ventral visual cortex showed reduced selectivity for faces compared to normally sighted people, and the way their brains mapped visual categories differed substantially.28PubMed Central. Visual and Auditory Object Representations in Ventral Visual Cortex After Restoring Sight in Humans The brain regions that would normally process visual information had been partially repurposed for other senses during the period of blindness. Some of that cross-modal wiring may actually help rather than hinder recovery, but the broader point is that restoring retinal function and restoring useful vision are not the same thing. The brain’s plasticity creates both opportunities and limitations, and the age at which vision is lost or restored matters enormously.
Retinal Organoids and the Search for New Drugs
One reason retinal therapies have been slow to develop is the difficulty of testing them. Animal models do not perfectly replicate human retinal disease, and you cannot biopsy a living person’s retina to study it. Retinal organoids, three-dimensional miniature retinas grown from stem cells in a dish, are changing that. Researchers have created organoids from patient-derived cells that recapitulate key features of diseases like retinitis pigmentosa and age-related macular degeneration, including the patterns of photoreceptor death seen in actual patients.29PubMed. Generation of Human Patient iPSC-derived Retinal Organoids to Model Retinitis Pigmentosa These organoids can then be used to screen drug candidates in a system that more closely resembles the human eye than a mouse model would.30PubMed Central. Pluripotent stem cell-derived retinal organoids for disease modeling and development of therapies Organoids that model AMD-associated photoreceptor degeneration are now being developed specifically as drug screening platforms, which could accelerate the pipeline for neuroprotective and regenerative treatments.31Cell Death & Disease. Human retinal organoid model of disease-relevant photoreceptor cell death amenable to drug screening
Organoids are not perfect stand-ins for a full retina. They lack blood vessels, immune cells, and the full layered architecture of the real thing. But they offer something no animal model can: human genetics in a human cellular context. For diseases driven by specific mutations, that difference is substantial.