Ocular tissue procurement: a case of hidden microbial breach

The variation does not read like ordinary statistical drift. It follows changes inside the procurement pipeline, most notably the transition from whole-globe enucleation to in situ corneal excision. When a large eye bank changed its retrieval method, contamination rose from 12.5% to 19.4% before downstream protocols were recalibrated.

That transition is one of the clearest published signals that ocular tissue procurement contamination is not simply a random event. It is a systems variable, shaped by technique, decontamination sequence, cold-chain latency, storage conditions, and the microbial ecology of the donor surface at the moment of recovery. For biobank operations, the implication is direct: tissue quality is not determined at the end of the process. It is largely designed upstream.

The shifting landscape of procurement: enucleation versus in situ excision

The distinction between whole-globe enucleation and in situ excision defines the first boundary of the contamination problem. Enucleation removes the intact globe for ex vivo corneoscleral dissection under laboratory conditions. Historically, that approach has been associated with lower contamination rates in part because the donor surface can be decontaminated again and the dissection can take place in a controlled environment.

In situ excision changes the exposure profile. The corneal button is removed in the field, often at the bedside or in a mortuary. The limbus remains exposed to ambient air and donor flora for a longer interval before the tissue is sealed into storage medium. The procedure may shorten the interval between recovery and preservation, but it also moves a larger part of the dissection process into an environment with less control over airflow, surface contact, instrument handling, and sequence discipline.

The retrospective dataset covering 4,546 corneas shows what can happen when a procurement program crosses that boundary. The reported contamination rate rose from 12.5% during an enucleation-dominant workflow to 19.4% in the period immediately after the technique transition. The increase was not necessarily permanent. Rates generally stabilized as staff retrained and the decontamination sequence was adjusted. But the transition period itself became a defined risk window, one that would not be visible in routine donor eligibility metrics.

Several variables move together during such a change:

• The retrieval environment changes. A controlled laboratory dissection is replaced by a procedure performed closer to the donor and the surrounding environment.
• The timing of preservation changes. In situ retrieval can reduce donor-to-preservation time, which may support endothelial viability and later graft performance.
• The contamination pathway changes. The exposed ocular surface, limbus, instruments, gloves, and work area become part of the same operational chain.
• Staff behavior becomes a major variable. A protocol can be technically sound on paper and still produce different results while a team is learning a new sequence.
• The data may lag behind the intervention. Contamination can become visible in storage or distribution records only after the procurement change has already been implemented.

This is why a change in retrieval method should not be managed as a simple equipment or staffing update. It is a process change with its own quality-control signal. An eye bank that monitors only graft outcomes may discover the problem late, after contaminated tissue has already entered the storage and allocation workflow. Q-sample results, culture patterns, and batch-level clustering provide earlier warning.

The trade-off is not one-sided. In situ retrieval may offer a practical advantage by reducing the time before preservation, while simultaneously increasing the period during which the ocular surface is exposed to contamination before sealing. The relevant question is therefore not whether one method is universally safe and the other universally unsafe. It is whether the local workflow has been calibrated to the method being used.

Technique is not a neutral variable in ocular tissue procurement. A documented shift from enucleation to in situ excision pushed contamination rates from 12.5% to 19.4% before protocol recalibration.

A useful procurement record should make that transition visible. It should capture the retrieval method, operator or team, timing of recovery and preservation, decontamination sequence, storage pathway, and later microbiological result. Without those fields, a spike in contamination appears as an isolated laboratory event rather than a process signal.

Microbial profiles and the challenge of resident ocular flora

The organisms that survive procurement decontamination form a recognizable profile, but the profile depends on where and when the sample is taken. In limbal swab cultures from 297 donors, representing 594 corneas recovered in situ, post-decontamination isolates were dominated by skin and environmental flora. The reported organisms included Staphylococcus epidermidis at 5.5%, Pseudomonas at 3.4%, Escherichia coli at 1.7%, and Klebsiella at 1.4%.

These figures describe residual flora after standard povidone-iodine decontamination supplemented by adjunctive topical antibiotics. They do not represent the initial microbial burden on the donor surface. That distinction matters because a negative or low-yield surface sample is not proof that the tissue entered storage in a sterile state. Sampling is limited by the site, the timing, the amount of material collected, and the distribution of organisms across the ocular surface.

The storage-phase pattern looks different. In the retrospective review of 4,546 corneas, fungi accounted for 61.9% of contamination events in organ-cultured storage. Candida species alone represented 45% of all contamination events. Bacterial species accounted for 34.4%.

The difference between the surface-swab profile and the storage profile is operationally important. The organism detected during procurement is not necessarily the organism that later proliferates in storage. The storage medium creates a different selective environment: antibiotic supplements may suppress some bacterial growth while allowing yeast species to persist or become more apparent over time.

That does not mean the storage medium creates contamination from nothing. It means that the organisms introduced during recovery do not all behave in the same way after preservation. A low bacterial yield at procurement cannot be read as a complete forecast of storage safety, especially when the medium and incubation conditions favor the detection of organisms that were not prominent in the original surface sample.

For this reason, storage-medium sampling — the Q-sample — should be treated as a primary control point rather than a ceremonial confirmation of an earlier swab. A procurement program that relies only on donor-eye microbial screening may detect the wrong part of the risk. Surface screening can describe the donor environment, but storage surveillance shows what remains viable and detectable after the tissue has entered the biobank workflow.

Why decontamination is a reduction step, not sterilization

Povidone-iodine reduces ocular surface microflora, but it does not convert the recovered tissue into a sterile object. The residual isolates documented in the 297-donor swab study were found after povidone-iodine application and adjunctive topical antibiotic use. That result is not evidence of protocol failure by itself. It is evidence that the intervention has limits.

The distinction between reduction and sterilization should shape both training and audit language. If a team describes decontamination as a sterilizing event, it may underweight:

• the condition of the donor surface before recovery;
• the interval between death, preparation, excision, and preservation;
• contact between instruments and non-sterile external surfaces;
• the possibility of uneven chemical coverage;
• the time required for the decontamination step to work;
• the microbial behavior of organisms that survive into storage.

Aseptic recovery of ocular tissue is therefore not a single action. It is a sequence in which every handoff can change the result. The quality system should record the sequence well enough to distinguish a contaminated donor surface from a breach introduced during recovery or handling.

Debunking the sepsis myth: systemic infection and corneal safety

Donor bacteremia and systemic sepsis are routinely flagged in tissue-procurement workflows as exclusion criteria or heightened-risk markers. The available evidence does not support treating systemic infection and corneal contamination as equivalent findings.

A prospective evaluation of 264 corneas from 136 donors compared microbial growth in septicemic donors — those with a positive blood culture at or near procurement — with growth in non-septicemic controls. Corneal contamination occurred in 7.14% of the septicemic group and 9.01% of the non-septicemic group. The difference was not statistically significant, with a reported p-value of 0.30. Across the full cohort, only one corneal tissue grew the same bacterial strain identified in the donor’s blood culture.

The underlying anatomy explains why the two risks should not be collapsed into one. The cornea is avascular. Systemic bacteremia does not seed the corneal stroma through vascular perfusion in the same way that bloodborne organisms reach vascularized parenchymal organs. The blood–ocular barriers also separate internal ocular compartments from ordinary systemic circulation.

That does not make the cornea immune to contamination. It changes the route that deserves attention. Organisms can reach the ocular surface through external colonization, contact, ambient exposure, handling, instruments, and imperfect decontamination. These variables are not interchangeable with bloodstream status. The relevant risk factors are surface flora density, exposure interval, retrieval technique, decontamination efficacy, and the conditions under which the tissue is sealed into storage.

The operational consequence is uncomfortable but useful. Excluding septicemic donors on the assumption that sepsis automatically increases corneal contamination may discard viable tissue without producing a corresponding reduction in corneal microbial risk. That does not mean sepsis is irrelevant to donor assessment. It remains important clinical information for other tissues and for the overall medical record. It means that corneal eligibility should not be governed by a categorical assumption unsupported by the comparative contamination data.

A stronger documentation model would separate three questions:

1. Was systemic infection present or suspected?

2. What does that information imply for the specific tissue being recovered?

3. What independent evidence exists about the ocular surface, procurement conditions, and storage microbiology?

That separation prevents a broad clinical label from replacing tissue-specific evidence. It also makes later audits more useful, because the eye bank can determine whether a decision was based on an established corneal risk, a general donor-safety rule, or a precaution carried over from another tissue system.

Systemic sepsis does not predict corneal contamination. The avascular cornea is not exposed to bloodborne flora through ordinary vascular perfusion, and septicemic donors did not show higher contamination rates than controls in the comparative evaluation.

Quantifying the risk: audit data on graft failure and contamination rates

Downstream clinical outcomes provide an important calibration point, but they must be linked to the correct denominator. A contamination event detected before distribution is not the same as an undetected event that reaches surgical use, and neither is identical to a post-keratoplasty infection from an unspecified source.

A five-year risk assessment from the Lions Eye Donation Service in Melbourne covered 3,122 recovered corneas. Twenty-four corneas, or 0.8%, were excluded before surgical distribution after microbiological testing of the storage medium — the Q-sample — returned positive growth. Of the 2,516 corneas that proceeded to distribution, three ultimately failed because of undetected microbial contamination. That corresponds to an undetected-contamination rate of 0.12% in the distributed group.

The two rates are often placed beside each other, but they describe different points in the pipeline. The pre-distribution exclusion rate of 0.8% is approximately 6.7 times the 0.12% rate observed for undetected contamination leading to graft failure in the distributed group. It is not a difference of two orders of magnitude, and these figures do not by themselves establish a formal tolerance threshold for an eye bank.

That qualification is more than a mathematical correction. The first rate reflects contamination detected early enough to prevent distribution. The second reflects failures that escaped the available screening and were identified only after tissue had proceeded to clinical use. They therefore measure different kinds of performance: the yield of the detection system and the residual risk after detection has operated.

The clinical literature also reports post-keratoplasty infection in two broad bands: endophthalmitis at 0.2% to 0.77% and microbial keratitis at 6.5% to 10.5%. The endophthalmitis range is broadly comparable to the Melbourne estimate for undetected contamination resulting in graft failure, but the comparison should not be treated as a direct equivalence. Postoperative infections can arise from several sources, including perioperative conditions, the recipient environment, and other routes that are not necessarily graft-borne. Microbial keratitis is particularly broad as an endpoint and should not be used as a proxy for procurement contamination without additional attribution.

The practical value of the Melbourne data lies in the separation of events. A useful audit should distinguish:

• contamination detected in storage before distribution;
• tissue withdrawn for a positive or suspicious microbiological result;
• tissue distributed despite a negative or non-diagnostic screen;
• later graft failure associated with microbial contamination;
• postoperative infection where the origin remains uncertain;
• organism identity and whether it matches a donor, storage, or clinical isolate.

Without that separation, a single “contamination rate” can conceal whether the program is detecting more events early or allowing more events to pass through undetected. Both changes matter, but they require different interventions.

Longitudinal Q-sample data are particularly valuable because they establish the local pattern of a specific workflow. A shift in positivity may reflect a change in retrieval method, a new decontamination product, altered storage conditions, staff turnover, delayed transport, or a change in the laboratory’s sampling and reporting practice. The number becomes meaningful when it is connected to those process variables.

The Melbourne figures should therefore be used as a reference point for questions, not as a universal pass–fail boundary. An eye bank can ask whether its own pre-distribution exclusion rate is changing, whether undetected events are clustering by operator or retrieval method, and whether the organism profile has shifted. It cannot infer from two published percentages alone that its workflow is operating within a formally defined universal tolerance.

Refining decontamination protocols in the era of modern biobanking

Ocular tissue procurement now sits at the intersection of three pressures: the potential viability benefit of faster in situ retrieval, the contamination profile of the donor surface at excision, and the limited predictive power of any single pre-transplant microbiological screen. Resolving that tension requires instrumentation rather than intuition.

The evidence supports several operational refinements.

First, the transition from enucleation to in situ excision should be treated as a defined risk period. A new method changes more than the location of the procedure. It changes timing, exposure, instrument handling, staff choreography, and the point at which the tissue enters storage. Q-sample results should be reviewed specifically around the transition, rather than averaged into a long-term rate that hides the early signal.

Second, decontamination protocols should be written and audited as reduction steps. The goal is to lower the microbial burden and limit survival into storage, not to claim that the ocular surface has become sterile. Training should make the residual-risk model explicit. Staff need to understand why contact time, coverage, sequence, and instrument discipline matter even when the procedure appears routine.

Third, storage-medium surveillance should include fungal risk, not only bacterial risk. The organ-culture data show a predominance of fungal contamination events, with Candida accounting for a substantial share of the total. A bacterial-only mental model can misread a storage system in which antibiotics alter the relative visibility and persistence of different organisms.

Fourth, donor-eye microbial screening should be interpreted as one layer of evidence. A surface swab can be useful, but it cannot substitute for storage surveillance, process records, and outcome monitoring. The sample captures a limited point in time and may not predict which organisms will remain viable under the conditions of organ culture or hypothermic storage.

Fifth, sepsis status should remain in the clinical assessment without being treated automatically as a corneal exclusion trigger. The prospective comparison did not show a higher corneal contamination rate in septicemic donors. A tissue-specific decision should instead incorporate the ocular findings, retrieval conditions, microbiological results, and the intended use of the tissue.

Hypothermic storage and an unresolved standardization question

The public literature does not establish a universal global standard for routine pre-transplant microbiological testing of hypothermic storage media, including media such as Optisol-GS, across international eye banks. That is an unresolved standardization question, not a confirmed absence of practice. Requirements, testing cadence, sample handling, and release decisions may vary between jurisdictions and institutions.

The distinction matters because organ-culture findings cannot simply be transferred to hypothermic workflows without qualification. Hypothermic storage suppresses microbial proliferation but does not eliminate the possibility of contamination. A Q-sample collected under one storage regime may also behave differently from a sample collected under another, depending on timing, medium composition, incubation, and laboratory practice.

Programs using hypothermic protocols should therefore establish their own baseline rather than infer equivalent performance from organ-culture benchmarks. That baseline should connect:

• the retrieval method and procurement setting;
• the interval from recovery to preservation;
• the decontamination sequence;
• the storage medium and temperature;
• the timing and method of microbiological sampling;
• the organisms recovered;
• the number of tissues withheld or discarded;
• the clinical outcomes associated with distributed tissue.

This approach does not require pretending that every eye bank uses the same system. It acknowledges that a meaningful comparison depends on comparable definitions and denominators. Until broader standardization is resolved, local data are not a substitute for external evidence, but they are the only reliable way to understand how a particular workflow behaves.

Modern biobanking rewards that kind of specificity. A program can have a low overall contamination rate and still carry a concentrated risk during a method transition, a particular recovery shift, or a period of delayed preservation. Conversely, a higher detection rate may reflect a more sensitive surveillance system rather than a worsening clinical risk, especially if contaminated tissue is consistently removed before distribution.

The central point is straightforward: contamination risk in ocular tissue procurement is not a fixed property of the donor. It is a measurable output of technique choice, decontamination design, storage conditions, and post-recovery surveillance. Donor eligibility remains important, but it cannot carry the entire quality system. The programs most able to protect tissue quality are those that make each stage visible, compare the right denominators, and treat microbial findings as process information rather than isolated bad luck.