Skip to Content

Fuchs' Endothelial Corneal Dystrophy: Why Your Cornea Clouds Overnight (2026 Guide)

Cornea guttata, genetics, DMEK vs DSEK evidence, the triple procedure, and honest risk data
Fuchs' Endothelial Corneal Dystrophy 2026: Cornea Guttata, DMEK vs DSEK Guide | Agaaz Ophthalmics

Cornea & Endothelium · Surgeon Series

The cornea has one
irreplaceable layer.
In Fuchs' dystrophy, it's failing.

Fuchs' endothelial corneal dystrophy doesn't touch the parts of the eye that get the attention — the lens, the retina, the optic nerve. It targets a single monolayer of cells on the inside of the cornea that never divides and never regenerates. When enough of it is gone, the cornea can no longer pump itself clear, and vision clouds — often worse first thing in the morning. Here is what that cell layer does, why it fails, and what to do when it does.

~15µmDMEK graft
thickness
2–3×more common
in women
0µmhost stroma
removed in DMEK
12 minreading time

Section 01 — One Cell Layer, No Backup

Every other corneal
layer can heal.
This one cannot.

The cornea is built from five stacked layers, and four of them can repair themselves after injury to some meaningful degree — the epithelium regrows within days, the stroma remodels over months. The innermost layer, the endothelium, cannot. Adult human corneal endothelial cells essentially stop dividing after early life. When one dies, its neighbours stretch and spread to cover the gap rather than being replaced by new cells. Density only ever goes down.

That single layer does one job: it actively pumps fluid out of the stroma above it, using energy-dependent Na+/K+-ATPase transport, keeping the stroma in a state of relative dehydration that is exactly what makes it optically clear. A healthy adult starts adult life with a substantial endothelial cell reserve and loses a small percentage per decade simply from normal aging — a loss the remaining cells comfortably absorb by spreading further. Fuchs' endothelial corneal dystrophy (FECD) accelerates that loss well beyond the normal aging curve, and adds a second problem on top of it: abnormal, wart-like excrescences of collagen called guttae, which form on the endothelium's basement membrane (Descemet's membrane) and further disrupt the pump's function even where cells remain.

The core answer, in about 120 words

Fuchs' dystrophy is a primary, usually bilateral disease of the cornea's single innermost cell layer. That layer cannot regenerate, so cell loss is permanent by definition — the only question is whether the remaining cells can still pump enough fluid out of the stroma to keep it clear. In Fuchs' dystrophy, cells die faster than normal aging accounts for, and abnormal deposits called guttae accumulate on Descemet's membrane and further impair function even in cells that survive. Early disease is often asymptomatic, found incidentally on a slit-lamp exam. Once pump capacity is exceeded, the stroma swells with fluid — worse overnight, better by midday — and vision clouds. Advanced disease that fails conservative measures is treated with endothelial keratoplasty, not a full corneal transplant.

The five layers
Epithelium, Bowman's layer, stroma, Descemet's membrane, endothelium. Only the innermost, the endothelium, is a single non-regenerating cell layer.
One cell, no mitosis
Adult human endothelial cells essentially do not divide. Lost cells are compensated by neighbouring cells enlarging and spreading, not by new cell production.
Guttae
Wart-like excrescences of abnormal collagen that accumulate on Descemet's membrane — the hallmark slit-lamp finding, visible before symptoms often appear.
The pump
Active, energy-dependent fluid transport out of the stroma keeps it relatively dehydrated and transparent. A failing pump means fluid accumulates — and the stroma swells.

Section 02 — Interactive

The number that decides
everything: cell density.

Endothelial cells are lost with ordinary aging and, faster, with Fuchs' dystrophy — but a lost cell is never replaced. What matters clinically is not any single cell, but how many are left to share the pumping workload between them. Drag the slider below to see how falling cell density, and the guttae that accumulate alongside it, change corneal clarity.

Interactive: endothelial cell density vs corneal clarity
Illustrative model built around widely used clinical teaching thresholds for endothelial reserve — not a diagnostic tool, and not a substitute for specular microscopy.
2000 cells/mm²
drag to rotate
Cornea / stroma Guttae on Descemet's membrane Endothelial monolayer

A newborn cornea has a very high endothelial cell density that declines steadily through life even with no disease at all — ordinary aging attrition, absorbed comfortably because there is so much reserve to begin with. Fuchs' dystrophy steepens that decline. Widely used clinical teaching places meaningful functional risk once density falls into roughly the 500 to 1,000 cells/mm² range, with clear decompensation risk becoming substantial below roughly 500 cells/mm² — thresholds drawn from clinical experience and specular microscopy literature rather than a single precise cut-off that applies identically to every eye.

Density alone is not the whole story. Two corneas with the same raw cell count can behave differently depending on cell shape uniformity. As cells die and survivors spread to cover the gap, they lose their normal even hexagonal shape (polymegathism describes size variability, pleomorphism describes shape variability). A cornea with more shape variability per remaining cell tends to be functioning closer to its limit than the raw density number alone would suggest — which is why a specular or confocal microscopy report includes both figures, not density in isolation.

Section 03 — Not the Same as Surgical Endothelial Loss

A genetic disease
and a surgical side effect
are not the same thing.

It is easy to conflate Fuchs' dystrophy with the endothelial cell loss that any cataract operation causes, because both involve the same cell layer and the same eventual risk — corneal edema. They are, however, different mechanisms with different timelines, and the distinction changes how each is managed.

FactorFuchs' endothelial dystrophyEndothelial loss after cataract surgery
OriginPrimary, largely genetic — a TCF4 repeat expansion in most late-onset casesMechanical and ultrasonic stress of phacoemulsification on a previously normal endothelium
LateralityUsually bilateral, often asymmetric between eyesConfined to the operated eye
OnsetSlow, over years to decades; guttae often precede symptoms by yearsOccurs during the surgical period itself
Hallmark signGuttae on Descemet's membrane, a "beaten-metal" slit-lamp appearanceReduced cell count with increased size/shape variability, without guttae
First-line managementObservation, hypertonic saline, monitoringNone needed in most eyes — normal reserve absorbs typical surgical loss
Surgical fix if neededEndothelial keratoplasty (DMEK/DSEK)Rare; more often prevented by gentler phaco technique and a protective OVD in eyes known to have reduced reserve

The clinically important overlap is this: a patient who already has Fuchs' dystrophy going into cataract surgery starts with less endothelial reserve than an average eye, and phacoemulsification always costs some cells regardless of how gently it is performed — the topic covered in full in our guide to corneal endothelial cell loss after cataract surgery. Stack a genetic head start toward decompensation on top of an unavoidable surgical cost, and the margin for error narrows considerably. That is exactly why preoperative specular microscopy and a protective viscoelastic technique both get more attention in a Fuchs' eye than in a routine cataract case, a point this guide returns to later.

Section 04 — Genetics & Guttae

Why it runs
in families.

Fuchs' dystrophy is not a single mutation with one clean story — it is a genetically heterogeneous disease with a well-characterised common variant and several rarer ones, layered on top of environmental and simply age-related contributions that are still being worked out.

Inheritance
Most familial late-onset cases follow an autosomal dominant pattern, though sporadic cases with no known family history also occur.
TCF4 / CTG18.1
The most common genetic association identified in late-onset Fuchs' dystrophy — an expanded trinucleotide repeat in an intron of the TCF4 gene.
COL8A2
A much rarer, earlier-onset form of the disease has been linked to mutations in the COL8A2 gene, affecting a structural collagen the endothelium produces.
Krachmer grading
A slit-lamp grading scale, roughly 0 to 6, scoring guttae confluence and central corneal involvement — used to track progression visit to visit.

Not everyone carrying the TCF4 repeat expansion goes on to develop clinically significant disease, and the repeat length itself varies in how strongly it predicts eventual severity — which is part of why a genetic marker alone is not a diagnosis. What actually establishes Fuchs' dystrophy clinically is the combination of a slit-lamp exam showing guttae, a Krachmer grade tracked over time, and specular or confocal microscopy quantifying cell density and morphology. Family history raises suspicion; the exam and the imaging confirm it.

Section 05 — The Evidence: DMEK vs DSEK

Two ways to replace
one failing layer —
they are not equivalent.

When conservative measures — hypertonic saline, monitoring, sometimes a hairdryer held at arm's length to encourage morning evaporation — are no longer enough, the endothelium itself is replaced. Endothelial keratoplasty has replaced full-thickness penetrating keratoplasty as the standard approach for endothelial disease, and within endothelial keratoplasty, DMEK and DSEK are not interchangeable choices.

Eyes reaching high-quality best-corrected visionDMEK favoured
Speed of visual recoveryDMEK favoured
Rebubbling (graft reattachment) rateDMEK higher
Rejection rate vs full-thickness transplant (PK)Both far lower than PK

Illustrative visualisation of the direction and rough magnitude reported across comparative DMEK/DSAEK series — see References for source literature. Not pooled meta-analytic effect sizes.

DSEK (Descemet stripping endothelial keratoplasty) transplants donor endothelium and Descemet's membrane on a thin carrier of posterior stroma, commonly cited at roughly 100 microns thick — tissue that is comparatively forgiving to handle in the operating theatre. DMEK (Descemet membrane endothelial keratoplasty) strips that stromal carrier away entirely, transplanting only the donor's Descemet's membrane and endothelium, a graft on the order of 10 to 20 microns. That thinness is exactly why DMEK tends to recover vision faster and reach a higher visual ceiling — there is no stromal-to-stromal interface left to scatter light — and exactly why it is technically harder: unfolding tissue that thin inside the anterior chamber, without tearing or flipping it, is a genuinely different skill than positioning a DSEK carrier. Comparative literature from groups publishing extensively on both techniques (see References) consistently reports this trade: better visual outcomes with DMEK, offset by a steeper learning curve and a higher early rebubbling rate while a surgeon builds DMEK-specific experience.

0approx. DMEK
graft thickness
0typical DSEK
graft thickness
0host stroma
removed in DMEK

Section 06 — Risk, Honestly Stated

DMEK is not simply
"the better one."
It is a different trade.

A guide that presents DMEK as a strictly superior upgrade over DSEK is not being honest about the tissue-handling risk that comes with it. Both are real, well-established procedures with a documented complication profile, and the right choice depends on the eye and the surgeon's experience with each.

RiskDMEKDSEK
Graft detachment / rebubblingDocumented, materially higher than DSEK, most common early in a surgeon's DMEK experienceLower than DMEK, though not zero
Primary graft failureUncommon in experienced hands; thin tissue can be damaged before or during insertionUncommon; thicker, more forgiving tissue to handle
RejectionLower than DSEK and substantially lower than full-thickness PKLow, somewhat higher than DMEK, much lower than PK
Visual quality ceilingGenerally the highest achievable of the endothelial keratoplasty techniquesSlightly lower on average, attributed to the stromal-to-stromal graft interface
Surgeon learning curveSteep — unfolding a 10–20µm graft inside the eye is technically demandingMore forgiving; widely adopted as the first endothelial keratoplasty technique learned
If the graft failsCan be repeated, or converted to DSEK or PKCan be repeated, or converted to PK
The honest framing. DMEK's advantage in final vision comes bundled with a technically harder, less forgiving procedure. DSEK trades a slightly lower visual ceiling for a gentler learning curve and fewer early reattachment problems. Neither is a strictly dominant choice — it depends on surgeon experience and the specific eye in front of them, the same way it does anywhere else in ophthalmic surgery.

Section 07 — When Cataract and Fuchs' Coexist

One eye,
two problems —
one operation, or two?

Fuchs' dystrophy and cataracts share an age profile, so it is common for a patient to need both a cataract removed and endothelial disease addressed around the same time. When that happens, cataract removal, IOL implantation and endothelial keratoplasty can be combined into a single operation — a "triple procedure" — or staged as two separate surgeries.

Illustrative Scenario — Composite, Not an Individual Patient
The cataract that isn't the only problem

A patient in their late sixties presents with gradually blurring vision and is found to have both a visually significant cataract and confluent central guttae with early morning haze that clears somewhat by afternoon — the classic diurnal pattern. Cataract surgery alone would remove the lens opacity but do nothing for the failing endothelium underneath, and the added stress of phacoemulsification on an already-compromised cell layer raises real concern about postoperative corneal decompensation if the endothelial disease is left untreated.

A triple procedure — phacoemulsification, IOL implantation, and DMEK in one sitting — addresses both problems in a single recovery period. The tradeoff is IOL power calculation: endothelial keratoplasty tends to produce a small, fairly predictable hyperopic shift, which surgeons account for by adjusting the target lens power at the time of combined surgery rather than measuring a cornea that has not yet stabilised.

Illustrative composite based on published triple-procedure decision criteria and IOL power adjustment literature for combined phaco-DMEK surgery — not a specific patient record.

The alternative, staging the two procedures, lets the surgeon measure the eye for IOL power after the corneal component has stabilised, at the cost of a second operation and a longer total recovery timeline for the patient. Neither approach is universally correct; the decision generally weighs guttae severity and central corneal thickness against the patient's tolerance for two procedures versus one longer recovery.

Section 08 — The Agaaz Range

Protecting a layer
that was already fragile
before the surgery began.

Agaaz Ophthalmics does not manufacture corneal donor tissue or endothelial keratoplasty instrumentation, but the intraocular case built around the cataract portion of surgery on a Fuchs' eye — whether standalone or combined into a triple procedure — is exactly the territory Agaaz supplies for cataract and other intraocular surgery.

OP-VISC & PURE-VISC — HPMC viscoelastics
Dispersive OVD
In eyes with reduced endothelial reserve — guttae on slit-lamp exam, a low cell count on specular microscopy, or a known Fuchs' diagnosis — a dispersive viscoelastic is the coating agent of choice during phacoemulsification: it adheres to and protects the endothelial surface through the phaco energy delivered near it, rather than being swept out of the chamber like a purely cohesive OVD under high aspiration flow.
PURE-HYAL — sodium hyaluronate 1.4%
Cohesive OVD
Paired with a dispersive coat in the soft-shell technique, a cohesive sodium hyaluronate OVD maintains anterior chamber space and pressurises the working area during lens removal. The two roles are complementary, not interchangeable — which is why the soft-shell approach layers both rather than using one OVD at a higher volume.
MOXGUARD — intracameral moxifloxacin
Endophthalmitis prophylaxis
Delivered into the anterior chamber once the OVD is aspirated at the close of the cataract portion of any combined or standalone procedure, on the same infection-prophylaxis evidence base used in routine cataract surgery.

The rheology behind this pairing is covered in full in the cohesive vs dispersive OVD guide, applied here to the specific population it matters most for. View the complete portfolio →

Section 09 — FAQ

Frequently asked questions
about Fuchs' dystrophy.

Fuchs' endothelial corneal dystrophy (FECD) is a primary, usually bilateral, slowly progressive disorder of the cornea's innermost cell layer, the endothelium. Endothelial cells normally pump fluid out of the corneal stroma to keep it clear; in FECD these cells die off faster than expected for age and form wart-like deposits called guttae on the underlying Descemet's membrane. As functioning cell density falls, the pump can no longer keep pace with fluid entering the stroma, and the cornea swells and clouds.

During sleep, the closed eyelid stops the normal evaporation of fluid from the front of the cornea, so fluid that a weakened endothelial pump cannot clear accumulates overnight and the stroma swells further. After waking, several hours of open-eye evaporation partially offsets the extra fluid, and vision typically sharpens through the morning and into the afternoon. This diurnal pattern — worse on waking, better by midday — is one of the more distinctive early symptoms of the disease.

Both are forms of endothelial keratoplasty that replace only the diseased inner cell layer rather than the whole cornea. DSEK (Descemet stripping endothelial keratoplasty) transplants donor endothelium and Descemet's membrane together with a thin carrier layer of posterior stroma, typically on the order of 100 microns thick. DMEK (Descemet membrane endothelial keratoplasty) transplants only the donor's endothelium and Descemet's membrane, without any stromal carrier — a graft roughly 10 to 20 microns thick. DMEK generally shows faster visual recovery and a higher proportion of eyes reaching very good corrected vision, but the tissue is more delicate to prepare and unfold, and rates of postoperative graft detachment requiring an air or gas "re-bubbling" are typically higher than with DSEK, especially earlier in a surgeon's DMEK experience.

No, and the distinction matters clinically. Fuchs' dystrophy is a primary, largely genetically driven disease of the endothelium that a patient has independent of any surgery. Endothelial cell loss after cataract surgery is a separate, mechanical phenomenon — a normal cornea loses a measurable share of endothelial cells from the physical and ultrasonic stress of phacoemulsification. The two can compound each other: a patient who already has reduced endothelial reserve from Fuchs' dystrophy has less margin to absorb the additional surgical cell loss that any cataract operation causes, which is exactly why preoperative endothelial assessment and protective OVD technique both get more attention in these eyes.

Yes — this is commonly called a "triple procedure": cataract removal, intraocular lens implantation, and endothelial keratoplasty (typically DMEK) performed in a single operation rather than as two staged surgeries. It is generally considered for patients who have both a visually significant cataract and endothelial disease advanced enough to need transplantation. The tradeoff is IOL power calculation, since endothelial keratoplasty can produce a small, fairly consistent hyperopic shift that surgeons account for in the lens power formula; a staged approach lets the surgeon measure the eye after the corneal component has stabilised, at the cost of a second operation and a longer total recovery timeline.

Often, yes. Most cases follow an autosomal dominant inheritance pattern, and the most common genetic association identified in late-onset Fuchs' dystrophy is an expanded trinucleotide repeat (CTG18.1) in the TCF4 gene. A much rarer, earlier-onset form has been linked to mutations in the COL8A2 gene. Not everyone with the genetic marker develops clinically significant disease, and family history alone is not a diagnosis — a slit-lamp exam showing guttae, ideally supported by specular or confocal microscopy of endothelial cell density and morphology, is what actually establishes it.

References & Evidence Base

Peer-reviewed
citations.

Krachmer JH, Purcell JJ Jr, Young CW, Bucher KD. "Corneal endothelial dystrophy: a study of 64 families." Arch Ophthalmol. 1978. Landmark original clinical description of the familial pattern of Fuchs' dystrophy. PubMed search
Wieben ED, Aleff RA, Tosakulwong N, et al. "A common trinucleotide repeat expansion within the TCF4 gene is highly associated with Fuchs' corneal dystrophy." PLoS One. 2012. The TCF4 CTG18.1 repeat association study. PubMed search
Biswas S, Munier FL, Yardley J, et al. and related literature. COL8A2 mutations in early-onset Fuchs' endothelial corneal dystrophy. Hum Mol Genet. and subsequent replication studies. PubMed search
Eghrari AO, Gottsch JD. "Fuchs' corneal dystrophy" — genetics and pathophysiology review. Expert Rev Ophthalmol. PubMed search
Price MO, Price FW Jr, and related series. Comparative outcome literature on Descemet membrane endothelial keratoplasty (DMEK). Ophthalmology / Cornea / JCRS. PubMed search
Comparative DMEK vs DSAEK series (Melles group and others). Visual acuity, recovery time and rebubbling-rate literature comparing the two endothelial keratoplasty techniques. PubMed search
Arshinoff SA. "Dispersive-cohesive viscoelastic soft shell technique." J Cataract Refract Surg. 1999. The foundational description of combining dispersive and cohesive OVDs for endothelial protection. PubMed search
American Academy of Ophthalmology. "Fuchs Endothelial Dystrophy" — EyeWiki clinical reference on diagnosis, grading and management. eyewiki.aao.org
StatPearls / NCBI Bookshelf. "Fuchs Endothelial Dystrophy" and endothelial keratoplasty clinical reference chapters. NCBI Bookshelf
Eye Bank Association of America. Annual Statistical Report — tracks the shift in US corneal transplant volume from penetrating keratoplasty toward endothelial keratoplasty techniques over the past decade. restoresight.org

Continue Reading

Related guides
from Beyond Vision.

Start writing here...