Organ culture complicates that assumption with an inconvenient fact: the tissue continues to lose endothelial cells while it is being stored, and the loss is not merely a laboratory curiosity. It can determine whether a graft remains viable after surgery.
In standard eye-bank organ culture, donor corneas are maintained at warm temperatures, typically between 31°C and 37°C, for as long as four to five weeks. During that interval, endothelial cell density declines. The paradox is that a preservation method designed to keep tissue usable also creates a controlled period of biological stress. That stress exposes compromised donor corneas before they reach the operating room.
This is why corneal endothelial cell density loss in organ culture should not be interpreted only as deterioration. It is also a form of quality control. The more uncomfortable question is whether current thresholds and storage practices distinguish adequately between tissue that is merely losing cells and tissue that has already lost the capacity to recover.
Warm storage is not passive storage
Corneal organ culture is often discussed as though the tissue is placed into a medium and temporarily removed from biological time. It is not. Warm storage preserves the cornea under conditions that allow continued cellular activity, but that activity comes with metabolic demands and progressive endothelial attrition.
The corneal endothelium is particularly unforgiving because human endothelial cells have limited proliferative capacity. Once cells are lost, neighboring cells must spread and enlarge to cover the remaining posterior surface. The cornea may remain clear for a period, but the tissue is no longer biologically equivalent to the original donor state. Cell density, cell morphology and reserve capacity all matter.
The relevant point is not that every decline predicts immediate graft failure. It does not. The point is that organ culture makes the decline visible before transplantation, when it can still influence tissue selection.
A German eye-bank dataset illustrates the trajectory. In donor corneas stored under standard organ-culture conditions without dextran, central endothelial cell density fell from an average of 2629 cells/mm² at 28 days to:
- 2526 cells/mm² at 35 days;
- 2364 cells/mm² at 42 days;
- 2045 cells/mm² at 77 days.
The numbers do not describe a sudden collapse. They show a persistent downward movement. That distinction matters. A graft can cross a clinically important threshold gradually, while still appearing acceptable if the decision is based on a single early measurement or on overall clarity alone.
Organ culture does not stop endothelial loss. It turns that loss into a measurable test of whether the donor tissue still has a credible margin of safety.
The storage temperature itself is not the only variable. Medium composition, donor biology, time from procurement, dissection quality and the condition of the tissue at entry into culture all influence the observed result. The temperature range creates the environment; it does not explain every individual outcome.
The rate of loss is modest—until it is not
A mean decline of 0.61% per day may sound reassuring when presented without context. It is less reassuring when interpreted as a cumulative process extending across several weeks.
An analysis of 1031 organ-cultured cornea-scleral discs between 2012 and 2016 found a mean endothelial cell-density loss of 280 cells/mm² over culture, corresponding to a mean relative daily loss of 0.61% per day. That average is useful for describing the population. It is not a personal forecast for every donor cornea.
This is where the language of averages can obscure the practical problem. A mean value does not reveal the distribution of losses. It does not identify the donors that decline rapidly, nor does it explain why two corneas placed under apparently comparable conditions may emerge with different endothelial profiles. The exact molecular mechanisms behind this individual variation remain unresolved.
The endothelial response is also not captured fully by cell counts. Density is a numerical summary of a cellular population, not a complete assessment of its condition. Morphological changes—such as variation in cell size and shape—can indicate that the remaining cells are compensating for loss. Yet morphology is not interchangeable with viability, and neither metric should be treated as a complete substitute for the other.
This is the first major limitation of a density-centred paradigm: a graft may sit above a threshold while carrying a depleted functional reserve. Conversely, a lower density does not automatically mean that the tissue is unusable in every surgical context. The clinical meaning depends on the procedure, the recipient, the method of assessment and the expected endothelial demand after transplantation.
What the available figures actually show
| Observation | What it indicates | What it does not prove |
|---|---|---|
| Mean relative loss of 0.61% per day | Endothelial decline continues during organ culture | That every donor loses cells at the same rate |
| 9.9% to 13.3% loss after up to four weeks | Medium and culture conditions may influence the measured decline | That one medium is universally superior in every setting |
| ECD below 2000 cells/mm² is linked to higher graft-failure risk | Low density can represent an important warning sign | That every graft below the threshold must fail |
| Up to 30% of cultured donor corneas are discarded | Organ culture removes a substantial proportion of unsuitable tissue before surgery | That all discarded tissue was biologically worthless for every possible research use |
| Decline from 2629 to 2045 cells/mm² over the reported storage interval | Prolonged storage can materially reduce the endothelial reserve | That storage duration alone explains the entire decline |
The table is deliberately less dramatic than many clinical summaries. It separates what the data support from what they do not. That distinction is particularly important in donor tissue research, where a clean numerical threshold can be tempting to overinterpret.
Preservation media are not interchangeable background fluids
Corneal preservation media comparison is often reduced to a question of which formulation produces the lower percentage of cell loss. That is a useful starting point, but not a sufficient endpoint.
In one evaluation of organ-cultured human corneas maintained for up to four weeks at 32°C, the mean overall endothelial cell loss was 13.3% in Dulbecco Modified Eagle Medium and 9.9% in CorneaMax Medium. The difference suggests that medium composition can influence the trajectory of endothelial preservation. It does not justify the broader claim that a single formulation resolves the problem of endothelial degradation.
The media are part of a system. Their effects are entangled with storage temperature, culture duration, donor characteristics and the way endothelial density is assessed. A percentage loss measured under one protocol cannot be transferred casually to another protocol, particularly when the counting method, sampling region or preparation of the cornea differs.
The more useful question is therefore not simply which medium wins. It is what the medium allows the laboratory to observe reliably.
A preservation medium has at least two roles:
1. It supports the tissue sufficiently for the cornea to remain assessable during storage.
2. It exposes tissue that cannot maintain adequate endothelial integrity under the conditions required for transplantation.
Those roles can pull in opposite directions. A highly permissive environment might preserve marginal tissue longer without resolving its underlying weakness. A more demanding environment may reveal endothelial vulnerability earlier, but the resulting discard rate could rise. Without a standardized global agreement on maximum organ-culture duration and universal assessment criteria, comparisons remain conditional rather than absolute.
That is not a minor technical footnote. If two eye banks use different media, temperatures and counting practices, their reported endothelial cell-density losses may describe different biological and operational realities. Calling the results equivalent because both are expressed in cells/mm² is a category error.
Organ culture as a stress test for graft quality
The strongest case for organ culture is not that it preserves every donor cornea. It is that it filters out some corneas before transplantation.
Up to 30% of donor corneas maintained in organ culture are discarded before surgery, primarily because endothelial cell counts become insufficient. From a logistics perspective, this looks like loss. From a transplant-safety perspective, it is also the purpose of the screening process.
The donor cornea is not judged only on the day it enters the eye bank. It is observed across time. If endothelial density declines to a level associated with increased failure risk, the tissue can be removed from the clinical supply rather than transferred into a recipient with a compromised cellular reserve.
Clinical studies indicate that endothelial grafts with ECD below 2000 cells/mm² are significantly more likely to fail. A separate suitability threshold of 2200 cells/mm² has been used for penetrating keratoplasty in an NHSBT study. These values are not interchangeable rules, but together they demonstrate the threshold dilemma: the number considered acceptable depends on the procedure, the evidence base and the risk tolerance of the tissue bank.
A graft intended for endothelial keratoplasty is not evaluated in exactly the same conceptual frame as a graft intended for penetrating keratoplasty. The surgical manipulation differs. The amount and distribution of transplanted endothelium differ. The recipient environment differs. Yet the central biological constraint remains: the graft must contain enough functional endothelial cells to maintain corneal deturgescence after implantation.
The phrase “cell count” can make this sound more mechanical than it is. A density measurement is an estimate derived from a selected region and a particular assessment method. It does not guarantee uniformity across the entire graft. Nor does it show whether the remaining cells are metabolically robust, structurally stable and capable of compensating for additional post-transplant loss.
This is why endothelial cell viability assessment should be treated as a layered process rather than a single pass-fail number. The assessment may involve:
- central and peripheral endothelial density rather than one undifferentiated count;
- cell morphology and the degree of cellular enlargement or shape variation;
- evidence of progressive decline during culture;
- the intended surgical application;
- the timing of the last assessment before release;
- whether the tissue has remained within the defined storage conditions.
The practical value of organ culture lies in the trend. A donor cornea that remains relatively stable over the observation period presents a different risk profile from one that loses density rapidly, even if both happen to cross the same threshold on the final day.
The meaningful question is not whether a cornea passes one count. It is whether its endothelium demonstrates enough stability to survive the next biological demand.
The threshold dilemma: 2000, 2200, or something more useful?
Thresholds are attractive because they create order. They allow eye banks to standardize decisions, communicate with surgeons and manage limited tissue inventories. But the number itself can become a surrogate for judgment.
The 2000 cells/mm² threshold is clinically important because grafts below it show a significantly greater likelihood of failure. The 2200 cells/mm² suitability threshold reported in the NHSBT context adds another margin. Yet neither figure should be treated as a universal biological border separating safe tissue from unsafe tissue.
The endothelium does not become nonfunctional when the count moves from 2001 to 1999 cells/mm². Biology rarely respects administrative cutoffs with that kind of precision. Thresholds are operational tools built around risk, not natural laws.
The more defensible interpretation is that lower density narrows the margin for additional injury and future attrition. A cornea may remain clear at a lower density under one set of circumstances and fail under another. The recipient’s ocular environment, the surgical technique, the graft design and the postoperative course all influence the final result.
This also explains why storage duration cannot be considered independently from graft integrity. Organ culture commonly permits storage for four to five weeks, but longer availability is not synonymous with unchanged quality. In the German data, the central ECD decline from 2629 cells/mm² at 28 days to 2364 cells/mm² at 42 days is not a trivial extension of storage. By 77 days, the reported average had fallen to 2045 cells/mm², close to the range at which risk becomes clinically consequential.
The issue is not whether every cornea should be released as early as possible. Early release may reduce the opportunity to observe delayed endothelial deterioration. The issue is whether storage duration is being treated as a neutral logistical convenience when it is, in fact, part of the biological selection process.
An eye bank that stores tissue longer gains scheduling flexibility and may improve inventory management. It also accepts more time for endothelial attrition to reveal itself. That trade-off is not a flaw in organ culture. It is the central bargain.
Why cell morphology and viability still complicate the count
Endothelial cell morphology in organ culture remains important because density alone cannot describe how the surviving monolayer is coping.
As cells are lost, neighboring cells enlarge and alter their shape to cover the posterior corneal surface. Morphological irregularity may therefore reflect compensation rather than immediate failure. But compensation has limits. A monolayer can preserve transparency while operating with reduced reserve, and the apparent stability of the cornea can conceal a more fragile cellular state.
This is where conventional paradigms become too comfortable. The field often treats density as the primary currency of graft quality because it is quantifiable and clinically familiar. The result is a paradigm deficit: the easiest measurement becomes the dominant measurement.
A stronger assessment framework would ask several linked questions:
- How many cells are present?
- How uniformly are they distributed?
- How stable is the density over the culture interval?
- What morphological changes accompany the decline?
- Does the tissue remain viable under the specific conditions used by the eye bank?
- Is the remaining endothelial reserve appropriate for the intended transplant procedure?
The answers will not necessarily produce one perfect score. They may instead produce a more honest profile of the graft.
This matters for research as much as for transplantation. Human donor tissue used to study corneal physiology, Fuchs dystrophy, Descemet membrane mechanics or endothelial recovery must be characterized by its storage history. A cornea that has spent several weeks under warm organ culture is not simply “human cornea.” It is human cornea after a defined period of post-mortem stress, medium exposure and endothelial attrition.
That history can affect experimental interpretation. If researchers compare donor endothelial cells, stromal biomechanics or anterior-segment tissue responses without accounting for preservation conditions, they may attribute storage-related variation to disease biology or donor phenotype. The result is a familiar scientific problem: a controllable pre-analytical variable mistaken for a biological discovery.
Post-mortem degradation is not a single process
The phrase post-mortem corneal endothelial degradation can suggest a linear decline beginning at death and continuing at a predictable rate. The available evidence is more complicated.
Endothelial loss during organ culture is measurable, but the individual trajectory depends on factors that are not fully explained by storage duration. Donor-to-donor variation remains substantial enough that identical culture conditions do not guarantee identical outcomes. The molecular mechanisms responsible for that variation are not completely defined.
This uncertainty should not be used to dismiss organ culture. It should make the interpretation more disciplined.
The tissue enters the eye bank with a history. Time to procurement, donor age and medical condition, handling, tissue preparation and baseline endothelial status may all influence the subsequent course. Once in culture, temperature and medium create additional pressures. The final cell count is therefore the result of a sequence, not a single storage event.
The distinction is especially important when comparing organ culture with hypothermic storage. The two approaches should not be assumed to produce identical cell-counting methodologies or identical long-term post-transplant endothelial loss. Culture media, assessment protocols and storage conditions differ. A result derived under one preservation system cannot be imported into another without examining the underlying method.
In practical terms, the relevant record for a donor cornea should include more than a final number. A meaningful tissue history includes the preservation mode, culture duration, temperature range, medium and serial endothelial assessments where available. Without that context, the count becomes detached from the process that produced it.
The limits of current preservation protocols
Organ culture is often positioned as a solution to the logistical problem of donor cornea storage. It is better understood as a controlled compromise.
It permits storage at warm temperatures for several weeks, creates time for endothelial assessment and helps remove tissue with inadequate viability before transplantation. Those are substantial advantages. But the same period also exposes the endothelial population to progressive loss. The protocol preserves usability by accepting attrition.
The central unresolved issue is not whether organ culture works. It clearly provides a valuable quality-control function. The harder issue is whether current protocols measure the right forms of quality early enough and with enough precision.
A future assessment model may need to combine:
1. Serial endothelial density measurements, so that the trajectory is visible rather than reduced to a final snapshot.
2. Morphological analysis, identifying whether the remaining cells show signs of compensatory stress.
3. Procedure-specific thresholds, rather than treating all keratoplasty grafts as biologically identical.
4. Standardized reporting of culture conditions, including medium, temperature and duration.
5. Better characterization of donor variability, especially the reasons some tissues decline faster under apparently similar conditions.
6. Research-use annotation, distinguishing tissue unsuitable for transplantation from tissue still valuable for laboratory studies.
The last point deserves more attention. A cornea rejected for clinical transplantation because its endothelial density falls below a surgical threshold may remain useful for research on stromal structure, Descemet membrane biology, trabecular meshwork relationships or anterior chamber histology. Clinical unsuitability is not the same as scientific worthlessness.
That distinction also improves the ethics of tissue procurement. A screening system that discards tissue from the clinical pathway should still preserve accurate information about why the tissue was rejected and what biological features remain available for investigation. Otherwise, the research community receives samples stripped of the very context needed to interpret them.
A better question for the field
The conventional question is: does the donor cornea meet the endothelial cell-density threshold?
The more useful question is: what does the culture trajectory reveal about the graft’s remaining reserve, and how does that reserve match the intended use?
That shift sounds modest, but it changes the logic of quality control. It moves the field away from a static number and toward a time-dependent assessment of tissue behavior. It also makes room for the inconvenient cases: corneas that remain above threshold but decline rapidly, corneas with lower counts but stable morphology, and tissues whose clinical value differs from their research value.
Despite the consensus around density thresholds, the data do not support pretending that a single number captures endothelial competence. A mean loss of 0.61% per day, a 9.9% to 13.3% decline across media, and the rejection of up to 30% of cultured corneas all point in the same direction: the endothelium is not a passive passenger in storage.
Organ culture matters because it exposes that fact before transplantation. It is not a perfect preservation method, and it does not prevent endothelial cell loss. Its real contribution is more demanding: it makes the tissue demonstrate whether it can withstand time.
The next advance in corneal preservation will therefore not come from repeating that organ culture is useful. That is already established. The challenge is to stop treating endothelial density as the entire verdict and start reading the graft’s behavior as evidence. Until then, the field risks confusing a convenient threshold with a complete account of viability.