Medical Device Eye Irritation Testing: Standards, Methods, and Regulatory Compliance

Eye irritation testing is a core requirement of the biological evaluation of medical devices, confirming that device materials and their chemical extracts do not cause inflammation or damage to ocular tissue such as the cornea, conjunctiva, and iris. Performed in accordance with ISO 10993-10, ISO 10993-23, OECD test guidelines, and the GB/T 16886 series, this testing is essential for devices that contact the eye and for regulatory submissions in major global markets. This article explains the technical principles, sample preparation, in vivo and in vitro test methods, and practical considerations that manufacturers need to understand when planning an eye irritation study.

Technical Overview

Eye irritation testing determines whether a medical device, a device material, or a chemical extract of a device can produce irritation or corrosion on the surface of the eye. The procedure evaluates the three principal target tissues of the anterior eye: the cornea, the iris, and the conjunctiva. The test forms part of the overall biological evaluation program required by ISO 10993-1 and its Chinese counterpart GB/T 16886.1, and it is conducted using the specific methodology described in ISO 10993-10 and GB/T 16886.10, with in vitro alternatives increasingly supported by OECD test guidelines and ISO 10993-23.

The need for eye irritation data is driven by the nature, duration, and frequency of body contact defined in the device use profile. Devices that contact the mucous membranes of the eye, either directly or through extracts and run-off during clinical use, generally warrant evaluation. A scientifically justified evaluation strategy should always begin with physical and chemical characterization of the device materials (ISO 10993-18) and a toxicological risk assessment (ISO 10993-17). In some cases, a comprehensive chemical characterization combined with published toxicological literature can justify waiving animal testing, in line with the 3Rs principle of replacement, reduction, and refinement promoted by ISO 10993-2.

• ISO 10993-1 / GB/T 16886.1 – Evaluation and testing within a risk management process
• ISO 10993-2 – Animal welfare requirements applicable to in vivo testing
• ISO 10993-10 / GB/T 16886.10 – Tests for irritation and skin sensitization, including the in vivo eye irritation procedure
• ISO 10993-23 / GB/T 16886.23 – Irritation testing using in vitro methods
• ISO 10993-12 / GB/T 16886.12 – Sample preparation and reference materials
• OECD Test Guidelines 437, 438, 491, and 492 – Recognized in vitro eye irritation methods

Testing Samples

Samples submitted for eye irritation testing should be representative of the finished device in its final commercial form, including all processing, cleaning, and sterilization steps. Where direct application of the intact device is impractical, tests are performed on polar and non-polar extracts prepared according to ISO 10993-12.

Ophthalmic and Ocular-Contact Devices

This category includes contact lenses and their accessories, ophthalmic viscosurgical devices, ocular bandages and dressings, eye shields and patches, and diagnostic instruments that touch the conjunctival surface during use. Because these products contact ocular tissue directly, eye irritation data is almost always expected by reviewers.

Mucous-Membrane Contacting Devices

Devices applied to mucous membranes in the head and face region, or devices whose clinical use could result in contact of extracts or fluids with the eye, are also candidates for evaluation. Examples include certain topical drug-delivery components, wound gels and irrigating solutions used near the face, and supportive materials used during ophthalmic surgery.

Test Extracts and Material Coupons

Extracts are typically prepared using a polar vehicle such as water for injection, 0.9% sodium chloride, or cell culture medium, and a non-polar vehicle such as sesame oil or cottonseed oil. Standard extraction ratios include 6 cm²/mL for thin films below 0.5 mm thickness, 3 cm²/mL for materials between 0.5 mm and 1 mm, and 0.2 g/mL for irregularly shaped solids. Common extraction conditions are 37 °C for 24 hours, or accelerated conditions such as 50 °C for 72 hours, 70 °C for 24 hours, or 121 °C for 1 hour, selected to simulate or exaggerate clinical exposure.

Controls and Reference Materials

Each study includes the extraction vehicle alone as a negative control and a known irritant, such as benzalkonium chloride or sodium dodecyl sulfate, as a positive control for method sensitivity. The pH of each extract is recorded, since strongly acidic or alkaline extracts may be presumed corrosive and excluded from in vivo testing on humane grounds.

Testing Items

Eye irritation studies for medical devices typically cover the following endpoints, whether performed in vivo or in vitro:

• Acute ocular irritation following single exposure to the device extract
• Ocular corrosion, defined as irreversible tissue damage to the eye surface
• Reversibility assessment, with observation extended until any observed effects resolve or are confirmed as persistent
• Corneal effects, including opacity and the surface area of the cornea involved
• Iris effects, such as congestion, swelling, and reaction to light
• Conjunctival effects, including redness and chemosis (swelling)
• In vitro endpoints such as corneal opacity, permeability to fluorescein, and epithelial cell viability
• Optional histopathological examination of treated corneal tissue to strengthen mechanistic interpretation

For devices intended for repeated or prolonged ocular contact, a repeated-exposure irritation protocol may be considered, subject to ethical review and regulatory expectations.

Testing Methods

In Vivo Rabbit Eye Irritation Test

The classical method, described in ISO 10993-10, uses a small number of rabbits, typically three, treated humanely in accordance with ISO 10993-2. A fixed volume of test extract, typically 0.1 mL, is instilled into the conjunctival sac of one eye, while the contralateral eye receives the corresponding control vehicle. The eyes are examined at approximately 1, 24, 48, and 72 hours after exposure, and observations continue up to 21 days if effects persist. Tissue reactions are graded using the standardized scoring scale: corneal opacity and affected area are scored 0 to 4, iris reaction 0 to 2, conjunctival redness 0 to 3, and chemosis 0 to 4. The mean scores across animals and time points, together with the persistence of any lesion, determine the classification of the sample as a non-irritant, mild, moderate, or severe ocular irritant. Materials known or suspected to be corrosive, or that produce unambiguous positive results in other irritation tests, should not be tested in the rabbit eye.

Bovine Corneal Opacity and Permeability (BCOP) Test

The BCOP test, standardized in OECD Test Guideline 437, uses isolated bovine corneas mounted in holders. Corneas are exposed to the device extract, after which opacity is measured quantitatively with an opacitometer and permeability is assessed by the passage of sodium fluorescein dye. The results are combined into an In Vitro Irritancy Score (IVIS). An IVIS of 3 or below indicates the sample is not classified as an eye irritant, while an IVIS of 55 or above indicates serious eye damage; intermediate scores generally require follow-up testing. Optional histopathology of the corneal epithelium, stroma, and endothelium can provide additional confidence in the result.

Isolated Chicken Eye (ICE) Test

Described in OECD Test Guideline 438, the ICE test evaluates corneal swelling, opacity, and fluorescein retention in enucleated chicken eyes exposed to the test extract. The three endpoints are scored and combined to categorize the irritancy potential of the sample, offering an organotypic alternative that avoids the use of live animals.

Reconstructed Human Cornea-Like Epithelium (RhCE) Test

RhCE models, covered by OECD Test Guideline 492, use three-dimensional human corneal epithelial tissues maintained at an air-liquid interface. After topical exposure to the extract, tissue viability is measured using MTT reduction. Viability above 60% relative to the negative control supports a conclusion of no eye irritation, while reduced viability indicates irritant potential. RhCE methods are increasingly applied to medical device extracts and are particularly attractive for screening program during development.

Short Time Exposure (STE) Test

The STE test, standardized in OECD Test Guideline 491, uses rabbit corneal epithelial cells in monolayer culture exposed to the test sample for a short duration, typically five minutes. Cell viability after exposure is used to predict acute ocular irritation, providing a rapid and economical option compatible with high-throughput material screening.

Test Report

A compliant study report issued by accredited third-party laboratories should include full sample and extraction documentation, the identification and source of test systems, individual animal or tissue observations and scores with negative and positive control results, the calculation scheme applied, and a clearly stated conclusion regarding the irritancy classification of the device or extract.

Applications

Eye irritation test data supports a wide range of regulatory and engineering objectives across the product lifecycle:

• Registration dossiers submitted to regulatory authorities in China under the GB/T 16886 framework, including submissions for ophthalmic devices and mucous-membrane contacting products
• Technical documentation for market access in the European Union under the Medical Device Regulation, where ISO 10993 biological evaluation is a harmonized expectation
• Pre-market submissions in the United States and other major jurisdictions where device biocompatibility evidence is required
• Design-phase screening of candidate polymers, coatings, and sterilization residuals before committing to expensive in vivo programs
• Change control, when modifications to materials, suppliers, manufacturing processes, or sterilization methods require re-evaluation of biological safety
• Integrated chemical and toxicological risk assessments that combine extractables and leachables data from ISO 10993-18 with biological test results to justify a waiving rationale

For ophthalmic products in particular, reviewers frequently scrutinize eye irritation results alongside cytotoxicity, sensitization, and acute systemic toxicity data, making a complete and internally consistent evaluation dossier essential for timely approval.

FAQ

When is eye irritation testing required for a medical device?

Testing is generally expected when a device or its extracts contact the surface of the eye or the surrounding conjunctival mucous membranes, as defined by the contact categorization in ISO 10993-1. The required extent of testing depends on contact nature and duration, and a documented risk assessment may justify reducing or waiving testing when chemical characterization demonstrates an adequate margin of safety.

Can in vitro methods replace the rabbit eye test?

In vitro methods such as BCOP, ICE, RhCE, and STE are formally recognized under OECD test guidelines and are increasingly accepted for medical device extracts, particularly as part of a tiered testing strategy. Acceptance varies by jurisdiction and product type, so manufacturers should confirm the applicable expectations with accredited third-party laboratories and the target regulatory authority before finalizing the study plan.

What extraction vehicles and conditions should be used?

Both a polar vehicle, such as saline or culture medium, and a non-polar vehicle, such as sesame oil, are typically extracted in parallel to capture both water-soluble and lipid-soluble leachables. Extraction ratios and temperatures follow ISO 10993-12, with 37 °C for 24 hours being the most common condition for devices that can be tested close to their as-used state.

How are results interpreted for regulatory purposes?

In vivo results are classified by combining mean Draize scores for cornea, iris, and conjunctiva with the persistence and reversibility of lesions. In vitro results are compared with method-specific decision thresholds, such as the IVIS cut-offs in BCOP or the viability limit in RhCE models. Samples showing no irritant response in both polar and non-polar extracts, with valid controls, support a conclusion of no eye irritation.

How long does an eye irritation study take?

In vitro studies, including extraction and tissue treatment, are typically completed within two to three weeks. In vivo rabbit eye studies require approximately three to five weeks, allowing for animal acclimatization, dosing, the observation period of up to 21 days for reversible effects, and reporting. Additional time should be allocated for method validation and protocol review.

Do sterilization residues affect eye irritation outcomes?

Yes. Residuals from ethylene oxide sterilization or migrated process chemicals can be leached into the test extracts and influence results. Samples should therefore be tested in their final sterilized condition, and sterilization residue limits should be verified separately so that any positive eye irritation finding can be correctly attributed to its chemical cause.