How to Treat Corneal Edema: From Drops to Surgery

Corneal edema treatment depends on the cause and severity, but the options range from simple salt-based eye drops that draw fluid out of the cornea to full-thickness transplants and, increasingly, injections of lab-grown cells that may one day replace donor tissue altogether. For mild or temporary swelling, such as the edema that follows routine cataract surgery, hypertonic saline drops are often enough. When the inner cell layer of the cornea is permanently damaged, as in Fuchs’ dystrophy or pseudophakic bullous keratopathy, the conversation shifts toward surgery. The past two decades have reshaped that surgical landscape, moving from full-thickness corneal transplants toward selective replacement of just the diseased layer, and now toward techniques that avoid donor tissue entirely.

Why the Cornea Swells in the First Place

The cornea stays clear because a single layer of cells on its inner surface, the endothelium, continuously pumps fluid out of the corneal tissue. This works through what researchers call a “pump-leak” mechanism: the stroma naturally absorbs water (the “leak”), while the endothelium actively transports ions to pull that water back out (the “pump”).¹ When enough endothelial cells die or stop working, the pump fails and the cornea waterloggs. That extra water scatters light, producing the hazy, cloudy vision that defines corneal edema.

Human endothelial cells barely divide in adults, so any significant loss tends to be permanent. The two most common diseases that destroy them are Fuchs’ endothelial dystrophy, a genetic condition in which cells gradually die off, and pseudophakic bullous keratopathy, where surgical trauma from cataract surgery or a poorly positioned intraocular lens damages the endothelium.² Other triggers include angle-closure glaucoma with high intraocular pressure, chemical injuries, herpes infections, and even implanted lenses that sit at the wrong angle.³ The underlying water-channel proteins involved differ between diseases. In bullous keratopathy, several aquaporin channels become abnormally distributed compared with Fuchs’ dystrophy, suggesting the two conditions accumulate fluid through different molecular pathways even though the result looks similar at the slit lamp.⁴

Morning Blur and Other Symptoms Worth Knowing About

If you have early Fuchs’ dystrophy, you may have noticed that your vision is worst first thing in the morning and slowly improves over the day. That pattern is real and measurable. A study tracking corneal thickness and visual quality throughout the day in people with mild Fuchs’ found that the cornea was thickest and vision was poorest in the morning, with both improving significantly by afternoon.⁵ During sleep, your eyes are closed and fluid accumulates because the endothelial pump can’t keep up without the evaporation that normally happens when your eyes are open. A hair dryer held at arm’s length pointed at the eyes each morning is a folk remedy that some ophthalmologists genuinely recommend for early-stage disease, and it works on the same principle: evaporation helps the cornea thin temporarily.

Beyond blurry vision, advanced corneal edema can be painful. When fluid pushes all the way through to the surface epithelium, it creates fluid-filled blisters (bullae) that can rupture, exposing corneal nerves and causing sharp, stinging pain. That condition, bullous keratopathy, is a major reason treatment becomes urgent even when a patient’s remaining vision is limited.

How Doctors Track the Swelling

Corneal thickness is the main objective measure of edema. A normal central cornea is roughly 530 to 550 micrometers thick; an edematous one can easily exceed 700 or 800. Three tools are commonly used to measure it: ultrasound pachymetry, specular microscopy, and anterior segment optical coherence tomography (AS-OCT). All three correlate well with each other, but AS-OCT has a practical advantage: it can measure thickness even in severely swollen corneas where the other two instruments sometimes fail. In one study comparing the three devices after cataract surgery, AS-OCT could take readings in all eyes on postoperative day one, while ultrasound pachymetry and specular microscopy managed measurements in only about 12 to 16 percent of eyes that same day.⁶

Specular microscopy also serves a different purpose: it photographs the endothelial cells themselves, letting doctors count cell density and spot abnormal shapes. In a chemical eye injury case, AS-OCT showed acute corneal thickening of about 940 micrometers during the initial injury, yet a year later, even though the cornea looked clinically clear, specular microscopy revealed persistent severe endothelial cell loss, with density down to just 648 cells per square millimeter (healthy adults typically have 2,000 to 3,000).⁷ That kind of subclinical damage is invisible without imaging and can catch patients off guard years later when the remaining cells finally give out.

Hypertonic Saline Drops and Other Topical Treatments

The simplest treatment for corneal edema is a hypertonic saline solution, typically a 5% sodium chloride drop or ointment. The high salt concentration creates an osmotic gradient that draws water out of the swollen cornea. A randomized, double-blind trial comparing 5% hypertonic saline to placebo after cataract surgery found that the saline group had faster corneal thinning (about 11% reduction versus 7% in the placebo group by day four) and better visual acuity. About 92% of patients in the treatment group reported clear vision within a week, compared with 57% in the placebo group.⁸ No serious side effects were recorded, though the drops can sting briefly on instillation.

Hypertonic saline works best when the edema is expected to be temporary, as after uncomplicated cataract surgery where the endothelium was stressed but not destroyed. For progressive diseases like Fuchs’ dystrophy, the drops offer temporary symptom relief but don’t address the underlying cell loss. Some patients use the ointment at bedtime to reduce overnight fluid accumulation and wake up with better morning vision, which can buy time before surgery becomes necessary.

ROCK Inhibitors as Eye Drops

A class of drugs called Rho kinase (ROCK) inhibitors has generated excitement because these drugs appear to coax corneal endothelial cells into migrating and dividing, something adult human endothelial cells normally refuse to do. In laboratory and animal studies, ROCK inhibitor eye drops improved corneal clarity and promoted wound healing in damaged endothelium.⁹ Ripasudil, a ROCK inhibitor already approved in Japan for glaucoma, has been repurposed for corneal endothelial therapy in clinical studies, though it is not yet widely available for this use outside of trials and select centers.

Where ROCK inhibitor drops have gained the most clinical traction is in combination with a minimally invasive procedure called Descemet stripping only, discussed in the next section. As standalone drops for corneal edema, the evidence remains early-stage, and many ophthalmologists consider them experimental for this indication.

Descemet Stripping Only, a Transplant-Free Approach

For patients with Fuchs’ dystrophy specifically, a newer approach skips donor tissue altogether. In Descemet stripping only (DSO), the surgeon removes just the diseased central portion of the Descemet’s membrane, the thin basement membrane covered in guttae (the wart-like excrescences that characterize Fuchs’). The idea is that healthy peripheral endothelial cells will then migrate inward to repopulate the stripped area, especially if helped along by ROCK inhibitor eye drops.

A study of 23 eyes at Sydney Eye Hospital found that 22 of 23 achieved corneal clearance at an average of about four weeks after DSO with topical ripasudil. Vision improved in all patients who cleared, and satisfaction was high. One patient failed to clear and went on to a conventional transplant at week 12.¹⁰ Patients who used ripasudil alongside DSO recovered vision faster than those who had DSO alone (roughly four and a half weeks versus six and a half weeks) and had higher endothelial cell counts at follow-up.¹¹

The catch is that about 39% of patients in the Sydney study experienced a relapse of edema when they stopped using ripasudil, and the cornea cleared again when they restarted the drops.¹⁰ That dependency is a significant limitation. Side effects were generally mild, with eye irritation being the most common local complaint and gastrointestinal upset occurring in about a quarter of patients. DSO is still considered a relatively niche procedure, best suited for carefully selected Fuchs’ patients with adequate peripheral endothelial reserves. It is not applicable to bullous keratopathy or other causes of edema where the entire endothelium is depleted.

Bandage Lenses and Anterior Stromal Puncture for Pain Relief

When corneal edema is advanced and causes painful epithelial blistering, but the patient isn’t a good candidate for transplant surgery (due to other eye conditions, poor vision potential, or general health), the goal shifts from restoring vision to controlling pain. Bandage contact lenses physically shield exposed nerves from eyelid friction. Anterior stromal puncture (ASP) is a simple procedure in which a needle creates small scars in the superficial cornea, helping the epithelium adhere more firmly and reducing blister formation. A study combining ASP with a bandage contact lens in bullous keratopathy patients found that irritation symptom scores dropped from about 8 out of 10 before surgery to about 5 afterward, and epithelial blisters disappeared on average within about 16 days.¹² Neither of these interventions improves the underlying edema, but they make a real difference in comfort for patients who are otherwise suffering.

Endothelial Keratoplasty, the Current Gold Standard Surgery

When the endothelium is irreversibly damaged, the definitive treatment is replacing the diseased tissue. Older full-thickness penetrating keratoplasty (PK), where the entire cornea is swapped out, has largely been supplanted by endothelial keratoplasty, in which only the thin inner layers are transplanted. The two main variants are DSAEK (Descemet’s stripping automated endothelial keratoplasty), which transplants a thin slice of stroma along with the endothelium, and DMEK (Descemet membrane endothelial keratoplasty), which transplants only the Descemet’s membrane and endothelium without any stromal tissue at all.

DMEK delivers the best visual outcomes. A meta-analysis comparing ultrathin DSAEK to DMEK found that DMEK eyes had significantly better visual acuity at three months, six months, and one year.¹³ A large ten-year follow-up study reported that the median time to achieve good visual acuity (roughly 20/25 equivalent) was about 8 months after DMEK, 12 months after DSAEK, and nearly 38 months after PK. By ten years, 99% of DMEK eyes had reached that level of vision.¹⁴ A separate systematic review confirmed this hierarchy, with five-year visual acuity best in DMEK, intermediate in DSAEK, and worst in PK.¹⁵

DMEK does come with tradeoffs. It has a higher rate of needing a “rebubbling” procedure, where an air or gas bubble is reinjected to re-attach the transplanted membrane if it detaches in the early postoperative period. Multiple systematic reviews have consistently found higher rebubbling rates with DMEK compared to DSAEK, though overall graft rejection remains rare for both.¹⁶ A Cochrane review of head-to-head trials recorded no primary graft failures and only one graft rejection event across 144 eyes.¹⁷ DSAEK remains the better choice in certain complex eyes, such as those with prior glaucoma shunts, anterior chamber abnormalities, or limited surgical view that makes DMEK technically difficult. PK still has a role when there is full-thickness corneal scarring or disease affecting layers beyond the endothelium.

Managing Graft Rejection

Even with modern techniques, transplant rejection can happen. Endothelial rejection presents with sudden onset of blurry vision, eye redness, and pain, and it is a true emergency that requires prompt treatment. The mainstay is aggressive topical corticosteroid drops, sometimes supplemented with systemic steroids. A randomized trial comparing local steroids alone to local plus systemic steroids for endothelial graft rejection found that the overall failure-to-reverse rate was 8%, with all three failures occurring in the local-steroid-only group, though the study was not large enough to show a statistically significant difference between the two strategies.¹⁸ The takeaway for patients: if you’ve had a corneal transplant and notice a sudden drop in vision, see your ophthalmologist the same day. Early, aggressive anti-inflammatory treatment gives you the best chance of saving the graft.

When Transplants Keep Failing, the Keratoprosthesis Option

Some patients cycle through multiple failed corneal transplants. For them, a keratoprosthesis, an artificial corneal device, may be the last resort. The Boston Type I keratoprosthesis (KPro) is the most widely used model. It consists of a clear optical cylinder held in place by a donor cornea carrier that is sutured into the eye. A retrospective comparison found that patients with prior graft failures who received a KPro had a lower two-year rate of device failure compared to those who underwent repeat PK.¹⁹ A more recent small case series found that all four KPro devices were retained at 12 months, with three out of four eyes reaching 20/100 or better by the end of the first year, though half of the eyes developed complications requiring management.²⁰

KPro eyes require lifelong monitoring, daily antibiotic drops to prevent infection, and sometimes glaucoma management. Complications like retroprosthetic membranes, stromal melting, and infection remain concerns. For that reason, keratoprostheses are reserved for patients who have no reasonable chance of success with another donor transplant.

Cell Therapy and the Future Beyond Donor Tissue

The most transformative work in corneal edema treatment is happening in cell therapy. Instead of transplanting donor tissue, researchers are culturing endothelial cells in the lab and injecting them directly into the eye, typically with a ROCK inhibitor to help the cells attach and survive. A landmark trial published in the New England Journal of Medicine injected cultured human endothelial cells along with a ROCK inhibitor into 11 eyes with bullous keratopathy. At 24 weeks, all 11 eyes had endothelial cell densities above 500 cells per square millimeter (a meaningful recovery), and 10 of 11 had densities above 1,000. Corneal thickness normalized in 10 of 11 eyes, and 9 of 11 gained at least two lines of visual acuity.²¹

The same research group recently published long-term follow-up stretching to ten years. Corneal transparency was maintained in about 93% of treated eyes at five years and roughly 84% at ten years. About 80% still had cell densities above 1,000 cells per square millimeter at five years, and no serious adverse reactions occurred over the entire follow-up period.²² Those ten-year results are genuinely remarkable for a cell-based therapy and have fueled regulatory-track trials. A prospective randomized trial (ESCALÓN) is already underway: at 12 months, mean corneal thickness dropped from about 697 micrometers to 571, and about 89% of subjects had gained meaningful visual improvement.²³

Another experimental angle involves loading cultured endothelial cells with magnetic nanoparticles and using an external magnet to guide them onto the back of the cornea after injection. In an animal model, the magnetically delivered cells formed a functional monolayer on the cornea, expressed the right tight-junction markers, and didn’t migrate to other structures in the eye. Intraocular pressure remained stable throughout the study.²⁴ Magnetic delivery is still preclinical, but it addresses one of the practical headaches of cell injection: keeping patients face-down for hours after the procedure so that gravity settles the cells onto the cornea.

AI-Assisted Diagnosis of Endothelial Disease

One of the quieter shifts in the field is the use of artificial intelligence to analyze specular microscopy images. Counting endothelial cells and detecting guttae (the hallmark deposits of Fuchs’ dystrophy) has traditionally been time-consuming and operator-dependent. Deep learning models can now segment individual cells and guttae in microscopy images, calculate cell density, and even estimate the ratio of guttae area to healthy cell area.²⁵ In one study, that guttae area ratio was the single strongest predictor of Fuchs’ dystrophy severity grading, outperforming cell density alone. A separate pilot using a deep learning classifier to detect Fuchs’ from widefield specular images achieved an area-under-the-curve of 0.96 in its development set.²⁶ Performance dropped on an external validation set, as is typical for AI models applied to images from different clinics, but the trajectory is clear: automated screening could eventually catch early endothelial disease before symptoms start, allowing earlier intervention with emerging therapies like ROCK inhibitors or cell injection while there are still healthy cells to work with.