(1)

In Part A, the following Chapter is added:

‘A.25   DISSOCIATION CONSTANTS IN WATER (TITRATION METHOD — SPECTROPHOTOMETRIC METHOD — CONDUCTOMETRIC METHOD)

INTRODUCTION

This test method is equivalent to OECD test guideline 112 (1981)

Prerequisites

Guidance information

Qualifying statements

Standard documents

This test method is based on methods given in the references listed in the section “Literature” and on the Preliminary Draft Guidance for Premanufacture Notification EPA, August 18, 1978.

METHOD — INTRODUCTION, PURPOSE, SCOPE, RELEVANCE, APPLICATION AND LIMITS OF TEST

The dissociation of a substance in water is of importance in assessing its impact upon the environment. It governs the form of the substance which in turn determines its behaviour and transport. It may affect the adsorption of the chemical on soils and sediments and absorption into biological cells.

Definitions and units

Dissociation is the reversible splitting into two or more chemical species which may be ionic. The process is indicated generally by

RX ⇌ R ++ X –

and the concentration equilibrium constant governing the reaction is

For example, in the particular case where R is hydrogen (the substance is an acid), the constant is

or

Reference substances

The following reference substances need not be employed in all cases when investigating a new substance. They are provided primarily so that calibration of the method may be performed from time to time and to offer the chance to compare the results when another method is applied.

It would be useful to have a substance with several pKs as indicated in Principle of the method, below. Such a substance could be:

Principle of the test method

The chemical process described is generally only slightly temperature dependent in the environmentally relevant temperature range. The determination of the dissociation constant requires a measure of the concentrations of the dissociated and undissociated forms of the chemical substance. From the knowledge of the stoichiometry of the dissociation reaction indicated in Definitions and units, above, the appropriate constant can be determined. In the particular case described in this test method the substance is behaving as an acid or a base, and the determination is most conveniently done by determining the relative concentrations of the ionised and unionised forms of the substance and the pH of the solution. The relationship between these terms is given in the equation for pKa in Definitions and units, above. Some substances exhibit more than one dissociation constant and similar equations can be developed. Some of the methods described herein are also suitable for non-acid/base dissociation.

Quality criteria

Repeatability

The dissociation constant should be replicated (a minimum of three determinations) to within ± 0,1 log units.

DESCRIPTION OF THE TEST PROCEDURES

There are two basic approaches to the determination of pKa. One involves titrating a known amount of substance with standard acid or base, as appropriate; the other involves determining the relative concentration of the ionised and unionised forms and its pH dependence.

Preparations

Methods based on those principles may be classified as titration, spectrophotometric and conductometric procedures.

Test solutions

For the titration method and conductometric method the chemical substance should be dissolved in distilled water. For spectrophotometric and other methods buffer solutions are used. The concentration of the test substance should not exceed the lesser of 0,01 M or half the saturation concentration, and the purest available form of the substance should be employed in making up the solutions. If the substance is only sparingly soluble, it may be dissolved in a small amount of a water-miscible solvent prior to adding to the concentrations indicated above.

Solutions should be checked for the presence of emulsions using a Tyndall beam, especially if a co-solvent has been used to enhance solubility. Where buffer solutions are used, the buffer concentration should not exceed 0,05 M.

Test conditions

Temperature

The temperature should be controlled to at least ± 1 °C. The determination should preferably be carried out at 20 °C.

If a significant temperature dependence is suspected, the determination should be carried out at least at two other temperatures. The temperature intervals should be 10 °C in this case and the temperature control ± 0,1 °C.

Analyses

The method will be determined by the nature of the substance being tested. It must be sufficiently sensitive to allow the determination of the different species at each test solution concentration.

Performance of the test

Titration method

The test solution is determined by titration with the standard base or acid solution as appropriate, measuring the pH after each addition of titrant. At least 10 incremental additions should be made before the equivalence point. If equilibrium is reached sufficiently rapidly, a recording potentiometer may be used. For this method both the total quantity of substance and its concentration need to be accurately known. Precautions must be taken to exclude carbon dioxide. Details of procedure, precautions, and calculation are given in standard tests, e.g. references (1), (2), (3), (4).

Spectrophotometric method

A wavelength is found where the ionised and unionised forms of the substance have appreciably different extinction coefficients. The UV/VIS absorption spectrum is obtained from solutions of constant concentration under a pH condition where the substance is essentially unionised and fully ionised and at several intermediate pHs. This may be done, either by adding increments of concentrated acid (base) to a relatively large volume of a solution of the substance in a multicomponent buffer, initially at high (low) pH (ref. 5), or by adding equal volumes of a stock solution of the substance in e.g. water, methanol, to constant volumes of various buffer solutions covering the desired pH range. From the pH and absorbance values at the chosen wavelength, a sufficient number of values for the pKa is calculated using data from at least 5 pHs where the substance is at least 10 per cent and less than 90 per cent ionised. Further experimental details and method of calculation are given in reference (1).

Conductometric method

Using a cell of small, known cell constant, the conductivity of an approximately 0,1 M solution of the substance in conductivity water is measured. The conductivities of a number of accurately-made dilutions of this solution are also measured. The concentration is halved each time, and the series should cover at least an order of magnitude in concentration. The limiting conductivity at infinite dilution is found by carrying out a similar experiment with the Na salt and extrapolating. The degree of dissociation may then be calculated from the conductivity of each solution using the Onsager equation, and hence using the Ostwald Dilution Law the dissociation constant may be calculated as K = α2C/(1 – α) where C is the concentration in moles per litre and α is the fraction dissociated. Precautions must be taken to exclude CO2. Further experimental details and method of calculation are given in standard texts and references (1), (6) and (7).

DATA AND REPORTING

Treatment of results

Titration method

The pKa is calculated for 10 measured points on the titration curve. The mean and standard deviation of such pKa values are calculated. A plot of pH versus volume of standard base or acid should be included along with a tabular presentation.

Spectrophotometric methods

The absorbance and pH are tabulated from each spectrum. At least five values for the pKa are calculated from the intermediate spectra data points, and the mean and standard deviation of these results are also calculated.

Conductometric method

The equivalent conductivity Λ is calculated for each acid concentration and for each concentration of a mixture of one equivalent of acid, plus 0,98 equivalent of carbonate-free sodium hydroxide. The acid is in excess to prevent an excess of OH– due to hydrolysis. 1/Λ is plotted against √C and Λo of the salt can be found by extrapolation to zero concentration.

Λo of the acid can be calculated using literature values for H+ and Na+. The pKa can be calculated from α = Λi/Λo and Ka = α2C/(1 – α) for each concentration. Better values for Ka can be obtained by making corrections for mobility and activity. The mean and standard deviations of the pKa values should be calculated.

Test report

All raw data and calculated pKa values should be submitted together with the method of calculation (preferably in a tabulated format, such as suggested in ref. 1) as should the statistical parameters described above. For titration methods, details of the standardisation of titrants should be given.

For the spectrophotometric method, all spectra should be submitted. For the conductometric method, details of the cell constant determination should be reported. Information on technique used, analytical methods and the nature of any buffers used should be given.

The test temperature(s) should be reported.

LITERATURE:

(2)

In Part B, Chapter B.5 is replaced by the following:

‘B.5   ACUTE EYE IRRITATION/CORROSION

INTRODUCTION

This test method is equivalent to OECD test guideline (TG) 405 (2012). OECD test guidelines for Testing of Chemicals are periodically reviewed to ensure that they reflect the best available science. In previous reviews of this test guideline, special attention was given to possible improvements through the evaluation of all existing information on the test chemical in order to avoid unnecessary testing in laboratory animals and thereby address animal welfare concerns. TG 405 (adopted in 1981 and updated in 1987, 2002, and 2012) includes the recommendation that prior to undertaking the described in vivo test for acute eye irritation/corrosion, a weight-of-the-evidence analysis should be performed (1) on the existing relevant data. Where insufficient data are available, it is recommended that they should be developed through application of sequential testing (2) (3). The testing strategy includes the performance of validated and accepted in vitro tests and is provided as a supplement to this test method. For the purpose of Regulation (EC) No 1907/2006 concerning the registration, evaluation, authorization and restriction of chemicals (REACH) (2), an integrated testing strategy is also included in the relevant ECHA Guidance (21). Testing in animals should only be conducted if determined to be necessary after consideration of available alternative methods, and use of those determined to be appropriate. At the time of drafting of this updated test method, there are instances where using this test method is still necessary or required under some regulatory frameworks.

The latest update mainly focused on the use of analgesics and anesthetics without impacting the basic concept and structure of the test guideline. ICCVAM (3) and an independent international scientific peer review panel reviewed the usefulness and limitations of routinely using topical anesthetics, systemic analgesics, and humane endpoints during in vivo ocular irritation safety testing (12). The review concluded that the use of topical anesthetics and systemic analgesics could avoid most or all pain and distress without affecting the outcome of the test, and recommended that these substances should always be used. This test method takes this review into account. Topical anesthetics, systemic analgesics, and humane endpoints should be routinely used during acute eye irritation and corrosion in vivo testing. Exceptions to their use should be justified. The refinements described in this method will substantially reduce or avoid animal pain and distress in most testing situations where in vivo ocular safety testing is still necessary.

Balanced preemptive pain management should include (i) routine pretreatment with a topical anesthetic (e.g. proparacaine or tetracaine) and a systemic analgesic (e.g. buprenorphine), (ii) routine post-treatment schedule of systemic analgesia (e.g. buprenorphine and meloxicam), (iii) scheduled observation, monitoring, and recording of animals for clinical signs of pain and/or distress, and (iv) scheduled observation, monitoring, and recording of the nature, severity, and progression of all eye injuries. Further detail is provided in the updated procedures described below. Following test chemical administration, no additional topical anesthetics or analgesics should be applied in order to avoid interference with the study. Analgesics with anti-inflammatory activity (e.g. meloxicam) should not be applied topically, and doses used systemically should not interfere with ocular effects.

Definitions are set out in the Appendix to the test method.

INITIAL CONSIDERATIONS

In the interest of both sound science and animal welfare, in vivo testing should not be considered until all available data relevant to the potential eye corrosivity/irritation of the chemical have been evaluated in a weight-of-the-evidence analysis. Such data include evidence from existing studies in humans and/or laboratory animals, evidence of eye corrosivity/irritation of one or more structurally related substances or mixtures of such substances, data demonstrating high acidity or alkalinity of the chemical (4) (5), and results from validated and accepted in vitro or ex vivo tests for skin corrosion and eye corrosion/irritation (6) (13) (14) (15) (16) (17). The studies may have been conducted prior to, or as a result of, a weight-of-the-evidence analysis.

For certain chemical, such an analysis may indicate the need for in vivo studies of the ocular corrosion/irritation potential of the chemical. In all such cases, before considering the use of the in vivo eye test, preferably a study of the in vitro and/or in vivo skin corrosion effects of the chemical should be conducted first and evaluated in accordance with the sequential testing strategy in test method B.4 (7) or the integrated testing strategy described in ECHA Guidance (21).

A sequential testing strategy, which includes the performance of validated in vitro or ex vivo eye corrosion/irritation tests, is included as a Supplement to this test method, and, for the purpose of REACH, in ECHA Guidance (21). It is recommended that such a testing strategy be followed prior to undertaking in vivo testing. For new chemicals, a stepwise testing approach is recommended for developing scientifically sound data on the corrosivity/irritation of the chemical. For existing chemicals with insufficient data on skin and eye corrosion/irritation, the strategy can be used to fill missing data gaps. The use of a different testing strategy or procedure or the decision not to use a stepwise testing approach, should be justified.

PRINCIPLE OF THE IN VIVO TEST

Following pretreatment with a systemic analgesic and induction of appropriate topical anesthesia, the chemical to be tested is applied in a single dose to one of the eyes of the experimental animal; the untreated eye serves as the control. The degree of eye irritation/corrosion is evaluated by scoring lesions of conjunctiva, cornea, and iris, at specific intervals. Other effects in the eye and adverse systemic effects are also described to provide a complete evaluation of the effects. The duration of the study should be sufficient to evaluate the reversibility or irreversibility of the effects.

Animals showing signs of severe distress and/or pain at any stage of the test or lesions consistent with the humane endpoints described in this test method (see Paragraph 26) should be humanely killed, and the chemical assessed accordingly. Criteria for making the decision to humanely kill moribund and severely suffering animals are the subject of an OECD Guidance document (8).

PREPARATIONS FOR THE IN VIVO TEST

Selection of species

The albino rabbit is the preferable laboratory animal and healthy young adult animals are used. A rationale for using other strains or species should be provided.

Preparation of animals

Both eyes of each experimental animal provisionally selected for testing should be examined within 24 hours before testing starts. Animals showing eye irritation, ocular defects, or pre-existing corneal injury should not be used.

Housing and feeding conditions

Animals should be individually housed. The temperature of the experimental animal room should be 20 °C (± 3 °C) for rabbits. Although the relative humidity should be at least 30 % and preferably not exceed 70 %, other than during room cleaning, the aim should be 50-60 %. Lighting should be artificial, the sequence being 12 hours light, 12 hours dark. Excessive light intensity should be avoided. For feeding, conventional laboratory diets may be used with an unrestricted supply of drinking water.

TEST PROCEDURE

Use of topical anesthetics and systemic analgesics

The following procedures are recommended to avoid or minimize pain and distress in ocular safety testing procedures. Alternate procedures that have been determined to provide as good or better avoidance or relief of pain and distress may be substituted.

Application of the test chemical

The test chemical should be placed in the conjunctival sac of one eye of each animal after gently pulling the lower lid away from the eyeball. The lids are then gently held together for about one second in order to prevent loss of the material. The other eye, which remains untreated, serves as a control.

Irrigation

The eyes of the test animals should not be washed for at least 24 hours following instillation of the test chemical, except for solids (see paragraph 18), and in case of immediate corrosive or irritating effects. At 24 hours a washout may be used if considered appropriate.

Use of a satellite group of animals to investigate the influence of washing is not recommended unless it is scientifically justified. If a satellite group is needed, two rabbits should be used. Conditions of washing should be carefully documented, e.g. time of washing; composition and temperature of wash solution; duration, volume, and velocity of application.

Dose level

(1)   Testing of liquids

For testing liquids, a dose of 0,1 ml is used. Pump sprays should not be used for instilling the chemical directly into the eye. The liquid spray should be expelled and collected in a container prior to instilling 0,1 mL into the eye.

(2)   Testing of solids

When testing solids, pastes, and particulate chemicals, the amount used should have a volume of 0,1 ml or a weight of not more than 100 mg. The test chemical should be ground to a fine dust. The volume of solid material should be measured after gently compacting it, e.g. by tapping the measuring container. If the solid test chemical has not been removed from the eye of the test animal by physiological mechanisms at the first observation time point of 1 hour after treatment, the eye may be rinsed with saline or distilled water.

(3)   Testing of aerosols

It is recommended that all pump sprays and aerosols be collected prior to instillation into the eye. The one exception is for chemicals in pressurised aerosol containers, which cannot be collected due to vaporisation. In such cases, the eye should be held open, and the test chemical administered to the eye in a simple burst of about one second, from a distance of 10 cm directly in front of the eye. This distance may vary depending on the pressure of the spray and its contents. Care should be taken not to damage the eye from the pressure of the spray. In appropriate cases, there may be a need to evaluate the potential for “mechanical” damage to the eye from the force of the spray.

An estimate of the dose from an aerosol can be made by simulating the test as follows: the chemical is sprayed on to weighing paper through an opening the size of a rabbit eye placed directly before the paper. The weight increase of the paper is used to approximate the amount sprayed into the eye. For volatile chemicals, the dose may be estimated by weighing a receiving container before and after removal of the test chemical.

Initial test (in vivo eye irritation/corrosion test using one animal)

It is strongly recommended that the in vivo test be performed initially using one animal (see Supplement to this test method: A Sequential Testing Strategy for Eye Irritation and Corrosion). Observations should allow for determination of severity and reversibility before proceeding to a confirmatory test in a second animal.

If the results of this test indicate the chemical to be corrosive or a severe irritant to the eye using the procedure described, further testing for ocular irritancy should not be performed.

Confirmatory test (in vivo eye irritation test with additional animals)

If a corrosive or severe irritant effect is not observed in the initial test, the irritant or negative response should be confirmed using up to two additional animals. If an irritant effect is observed in the initial test, it is recommended that the confirmatory test be conducted in a sequential manner in one animal at a time, rather than exposing the two additional animals simultaneously. If the second animal reveals corrosive or severe irritant effects, the test is not continued. If results from the second animal are sufficient to allow for a hazard classification determination, then no further testing should be conducted.

Observation period

The duration of the observation period should be sufficient to evaluate fully the magnitude and reversibility of the effects observed. However, the experiment should be terminated at any time that the animal shows signs of severe pain or distress (8). To determine reversibility of effects, the animals should be observed normally for 21 days post administration of the test chemical. If reversibility is seen before 21 days, the experiment should be terminated at that time.

Clinical observations and grading of eye reactions

The eyes should be comprehensively evaluated for the presence or absence of ocular lesions one hour post-TCA, followed by at least daily evaluations. Animals should be evaluated several times daily for the first 3 days to ensure that termination decisions are made in a timely manner. Test animals should be routinely evaluated for the entire duration of the study for clinical signs of pain and/or distress (e.g. repeated pawing or rubbing of the eye, excessive blinking, excessive tearing) (9) (10) (11) at least twice daily, with a minimum of 6 hours between observations, or more often if necessary. This is necessary to (i) adequately assess animals for evidence of pain and distress in order to make informed decisions on the need to increase the dosage of analgesics and (ii) assess animals for evidence of established humane endpoints in order to make informed decisions on whether it is appropriate to humanely euthanize animals, and to ensure that such decisions are made in a timely manner. Fluorescein staining should be routinely used and a slit lamp biomicroscope used when considered appropriate (e.g. assessing depth of injury when corneal ulceration is present) as an aid in the detection and measurement of ocular damage, and to evaluate if established endpoint criteria for humane euthanasia have been met. Digital photographs of observed lesions may be collected for reference and to provide a permanent record of the extent of ocular damage. Animals should be kept on test no longer than necessary once definitive information has been obtained. Animals showing severe pain or distress should be humanely killed without delay, and the chemical assessed accordingly.

Animals with the following eye lesions post-instillation should be humanely killed (refer to Table 1 for a description of lesion grades): corneal perforation or significant corneal ulceration including staphyloma; blood in the anterior chamber of the eye; grade 4 corneal opacity; absence of a light reflex (iridial response grade 2) which persists for 72 hours; ulceration of the conjunctival membrane; necrosis of the conjunctivae or nictitating membrane; or sloughing. This is because such lesions generally are not reversible. Furthermore, it is recommended that the following ocular lesions be used as humane endpoints to terminate studies before the end of the scheduled 21-day observation period. These lesions are considered predictive of severe irritant or corrosive injuries and injuries that are not expected to fully reverse by the end of the 21-day observation period: severe depth of injury (e.g. corneal ulceration extending beyond the superficial layers of the stroma), limbus destruction > 50 % (as evidenced by blanching of the conjunctival tissue), and severe eye infection (purulent discharge). A combination of: vascularisation of the cornea surface (i.e., pannus); area of fluorescein staining not diminishing over time based on daily assessment; and/or lack of re-epithelialisation 5 days after test chemical application could also be considered as potentially useful criteria to influence the clinical decision on early study termination. However, these findings individually are insufficient to justify early study termination. Once severe ocular effects have been identified, an attending or qualified laboratory animal veterinarian or personnel trained to identify the clinical lesions should be consulted for a clinical examination to determine if the combination of these effects warrants early study termination. The grades of ocular reaction (conjunctivae, cornea and iris) should be obtained and recorded at 1, 24, 48, and 72 hours following test chemical application (Table 1). Animals that do not develop ocular lesions may be terminated not earlier than 3 days post instillation. Animals with ocular lesions that are not severe should be observed until the lesions clear, or for 21 days, at which time the study is terminated. Observations should be performed and recorded at a minimum of 1 hour, 24 hours, 48 hours, 72 hours, 7 days, 14 days, and 21 days in order to determine the status of the lesions, and their reversibility or irreversibility. More frequent observations should be performed if necessary in order to determine whether the test animal should be euthanized out of humane considerations or removed from the study due to negative results

The grades of ocular lesions (Table 1) should be recorded at each examination. Any other lesions in the eye (e.g. pannus, staining, anterior chamber changes) or adverse systemic effects should also be reported.

Examination of reactions can be facilitated by use of a binocular loupe, hand slit-lamp, biomicroscope, or other suitable device. After recording the observations at 24 hours, the eyes may be further examined with the aid of fluorescein.

The grading of ocular responses is necessarily subjective. To promote harmonisation of grading of ocular response and to assist testing laboratories and those involved in making and interpreting the observations, the personnel performing the observations need to be adequately trained in the scoring system used.

DATA AND REPORTING

Evaluation of results

The ocular irritation scores should be evaluated in conjunction with the nature and severity of lesions, and their reversibility or lack of reversibility. The individual scores do not represent an absolute standard for the irritant properties of a chemical, as other effects of the test chemical are also evaluated. Instead, individual scores should be viewed as reference values and are only meaningful when supported by a full description and evaluation of all observations.

Test report

The test report should include the following information:

Interpretation of the results

Extrapolation of the results of eye irritation studies in laboratory animals to humans is valid only to a limited degree. In many cases the albino rabbit is more sensitive than humans to ocular irritants or corrosives.

Care should be taken in the interpretation of data to exclude irritation resulting from secondary infection.

LITERATURE:

Table 1

Grading of ocular lesions

SUPPLEMENT TO TEST METHOD B.5  (4)

A SEQUENTIAL TESTING STRATEGY FOR EYE IRRITATION AND CORROSION

General considerations

In the interests of sound science and animal welfare, it is important to avoid the unnecessary use of animals, and to minimise testing that is likely to produce severe responses in animals. All information on a chemical relevant to its potential ocular irritation/corrosivity should be evaluated prior to considering in vivo testing. Sufficient evidence may already exist to classify a test chemical as to its eye irritation or corrosion potential without the need to conduct testing in laboratory animals. Therefore, utilizing a weight-of-the-evidence analysis and sequential testing strategy will minimise the need for in vivo testing, especially if the chemical is likely to produce severe reactions.

It is recommended that a weight-of-the-evidence analysis be used to evaluate existing information pertaining to eye irritation and corrosion of chemicals and to determine whether additional studies, other than in vivo eye studies, should be performed to help characterise such potential. Where further studies are needed, it is recommended that the sequential testing strategy be utilised to develop the relevant experimental data. For substances which have no testing history, the sequential testing strategy should be utilised to develop the data needed to evaluate its eye corrosion/irritation. The initial testing strategy described in this Supplement was developed at an OECD workshop (1). It was subsequently affirmed and expanded in the Harmonised Integrated Hazard Classification System for Human Health and Environmental Effects of Chemical Substances, as endorsed by the 28th Joint Meeting of the Chemicals Committee and the Working Party on Chemicals, in November 1998 (2), and updated by an OECD expert group in 2011.

Although this testing strategy is not an integrated part of test method B.5, it expresses the recommended approach for the determination of eye irritation/corrosion properties. This approach represents both best practice and an ethical benchmark for in vivo testing for eye irritation/corrosion. The test method provides guidance for the conduct of the in vivo test and summarises the factors that should be addressed before considering such a test. The sequential testing strategy provides a weight-of-the-evidence approach for the evaluation of existing data on the eye irritation/corrosion properties of chemicals and a tiered approach for the generation of relevant data on chemicals for which additional studies are needed or for which no studies have been performed. The strategy includes the performance first of validated and accepted in vitro or ex vivo tests and then of TM B.4 studies under specific circumstances (3) (4).

Description of the stepwise testing strategy

Prior to undertaking tests as part of the sequential testing strategy (Figure), all available information should be evaluated to determine the need for in vivo eye testing. Although significant information might be gained from the evaluation of single parameters (e.g. extreme pH), the totality of existing information should be assessed. All relevant data on the effects of the chemical in question, and its structural analogues, should be evaluated in making a weight-of-the-evidence decision, and a rationale for the decision should be presented. Primary emphasis should be placed upon existing human and animal data on the chemical, followed by the outcome of in vitro or ex vivo testing. In vivo studies of corrosive chemicals should be avoided whenever possible. The factors considered in the testing strategy include:

TESTING AND EVALUATION STRATEGY FOR EYE IRRITATION/CORROSION

LITERATURE:

(3)

In Part B, Chapter B.10 is replaced by the following:

‘B.10    In Vitro Mammalian Chromosomal Aberration Test

INTRODUCTION

This test method is equivalent to OECD test guideline 473 (2016). It is part of a series of test methods on genetic toxicology. An OECD document that provides succinct information on genetic toxicology testing and an overview of the recent changes that were made to these Test Guidelines has been developed (1).

The purpose of the in vitro chromosomal aberration test is to identify chemicals that cause structural chromosomal aberrations in cultured mammalian cells (2) (3) (4). Structural aberrations may be of two types, chromosome or chromatid. Polyploidy (including endoreduplication) could arise in chromosome aberration assays in vitro. While aneugens can induce polyploidy, polyploidy alone does not indicate aneugenic potential and can simply indicate cell cycle perturbation or cytotoxicity (5). This test is not designed to measure aneuploidy. An in vitro micronucleus test (6) would be recommended for the detection of aneuploidy.

The in vitro chromosomal aberration test may employ cultures of established cell lines or primary cell cultures of human or rodent origin. The cells used should be selected on the basis of growth ability in culture, stability of the karyotype (including chromosome number) and spontaneous frequency of chromosomal aberrations (7). At the present time, the available data do not allow firm recommendations to be made but suggest it is important, when evaluating chemical hazards to consider the p53 status, genetic (karyotype) stability, DNA repair capacity and origin (rodent versus human) of the cells chosen for testing. The users of this test method are thus encouraged to consider the influence of these and other cell characteristics on the performance of a cell line in detecting the induction of chromosomal aberrations, as knowledge evolves in this area.

Definitions used are provided in Appendix 1.

INITIAL CONSIDERATIONS AND LIMITATIONS

Tests conducted in vitro generally require the use of an exogenous source of metabolic activation unless the cells are metabolically competent with respect to the test chemicals. The exogenous metabolic activation system does not entirely mimic in vivo conditions. Care should be taken to avoid conditions that could lead to artifactual positive results, i.e. chromosome damage not caused by direct interaction between the test chemicals and chromosomes; such conditions include changes in pH or osmolality (8) (9) (10), interaction with the medium components (11) (12) or excessive levels of cytotoxicity (13) (14) (15) (16).

This test is used to detect chromosomal aberrations that may result from clastogenic events. The analysis of chromosomal aberration induction should be done using cells in metaphase. It is thus essential that cells should reach mitosis both in treated and in untreated cultures. For manufactured nanomaterials, specific adaptations of this test method may be needed but are not described in this test method.

Before use of the test method on a mixture for generating data for an intended regulatory purpose, it should be considered whether, and if so why, it may provide adequate results for that purpose. Such considerations are not needed, when there is a regulatory requirement for testing of the mixture.

PRINCIPLE OF THE TEST

Cell cultures of human or other mammalian origin are exposed to the test chemical both with and without an exogenous source of metabolic activation unless cells with an adequate metabolizing capability are used (see paragraph 13). At appropriate predetermined intervals after the start of exposure of cell cultures to the test chemical, they are treated with a metaphase-arresting chemical (e.g. colcemid or colchicine), harvested, stained and metaphase cells are analysed microscopically for the presence of chromatid-type and chromosome-type aberrations.

DESCRIPTION OF THE METHOD

Preparations

Cells

A variety of cell lines (e.g. Chinese Hamster Ovary (CHO), Chinese Hamster lung V79, Chinese Hamster Lung (CHL)/IU, TK6) or primary cell cultures, including human or other mammalian peripheral blood lymphocytes, can be used (7). The choice of the cell lines used should be scientifically justified. When primary cells are used, for animal welfare reasons, the use of primary cells from human origin should be considered where feasible and sampled in accordance with the human ethical principles and regulations. Human peripheral blood lymphocytes should be obtained from young (approximately 18-35 years of age), non-smoking individuals with no known illness or recent exposures to genotoxic agents (e.g. chemicals, ionizing radiations) at levels that would increase the background incidence of chromosomal aberrations. This would ensure the background incidence of chromosomal aberrations to be low and consistent. The baseline incidence of chromosomal aberrations increases with age and this trend is more marked in females than in males (17) (18). If cells from more than one donor are pooled for use, the number of donors should be specified. It is necessary to demonstrate that the cells have divided from the beginning of treatment with the test chemical to cell sampling. Cell cultures are maintained in an exponential cell growth phase (cell lines) or stimulated to divide (primary cultures of lymphocytes), to expose the cells at different stages of the cell cycle, since the sensitivity of cell stages to the test chemicals may not be known. The primary cells that need to be stimulated with mitogenic agents in order to divide are generally no longer synchronized during exposure to the test chemical (e.g. human lymphocytes after a 48-hour mitogenic stimulation). The use of synchronized cells during treatment is not recommended, but can be acceptable if justified.

Media and culture conditions

Appropriate culture medium and incubation conditions (culture vessels, humidified atmosphere of 5 % CO2 if appropriate, incubation temperature of 37 °C) should be used for maintaining cultures. Cell lines should be checked routinely for the stability of the modal chromosome number and the absence of Mycoplasma contamination (7) (19), and cells should not be used if contaminated or if the modal chromosome number has changed. The normal cell cycle time of cell lines or primary cultures used in the testing laboratory should be established and should be consistent with the published cell characteristics (20).

Preparation of cultures

Cell lines: cells are propagated from stock cultures, seeded in culture medium at a density such that the cells in suspensions or in monolayers will continue to grow exponentially until harvest time (e.g. confluence should be avoided for cells growing in monolayers).

Lymphocytes: whole blood treated with an anti-coagulant (e.g. heparin) or separated lymphocytes are cultured (e.g. for 48 hours for human lymphocytes) in the presence of a mitogen [e.g. phytohaemagglutinin (PHA) for human lymphocytes] in order to induce cell division prior to exposure to the test chemical.

Metabolic activation

Exogenous metabolising systems should be used when employing cells which have inadequate endogenous metabolic capacity. The most commonly used system that is recommended by default, unless otherwise justified, is a co-factor-supplemented post-mitochondrial fraction (S9) prepared from the livers of rodents (generally rats) treated with enzyme-inducing agents such as Aroclor 1254 (21) (22) (23) or a combination of phenobarbital and β-naphthoflavone (24) (25) (26) (27) (28) (29). The latter combination does not conflict with the Stockholm Convention on Persistent Organic Pollutants (30) and has been shown to be as effective as Aroclor 1254 for inducing mixed-function oxidases (24) (25) (26) (28). The S9 fraction typically is used at concentrations ranging from 1 to 2 % (v/v) but may be increased to 10 % (v/v) in the final test medium. The use of products that reduce the mitotic index, especially calcium complexing products (31) should be avoided during treatment. The choice of type and concentration of exogenous metabolic activation system or metabolic inducer employed may be influenced by the class of chemicals being tested.

Test chemical preparation

Solid test chemicals should be prepared in appropriate solvents and diluted, if appropriate, prior to treatment of the cells (see paragraph 23). Liquid test chemicals may be added directly to the test system and/or diluted prior to treatment of the test system. Gaseous or volatile test chemicals should be tested by appropriate modifications to the standard protocols, such as treatment in sealed culture vessels (32) (33) (34). Preparations of the test chemical should be made just prior to treatment unless stability data demonstrate the acceptability of storage.

Test conditions

Solvents

The solvent should be chosen to optimize the solubility of the test chemicals without adversely impacting the conduct of the assay, e.g. changing cell growth, affecting the integrity of the test chemical, reacting with culture vessels, impairing the metabolic activation system. It is recommended that, wherever possible, the use of an aqueous solvent (or culture medium) should be considered first. Well established solvents are for example water or dimethyl sulfoxide. Generally organic solvents should not exceed 1 % (v/v) and aqueous solvents (saline or water) should not exceed 10 % (v/v) in the final treatment medium. If not well-established solvents are used (e.g. ethanol or acetone), their use should be supported by data indicating their compatibility with the test chemicals, the test system and their lack of genetic toxicity at the concentration used. In the absence of that supporting data, it is important to include untreated controls (see Appendix 1) to demonstrate that no deleterious or clastogenic effects are induced by the chosen solvent.

Measuring cell proliferation and cytotoxicity and choosing treatment concentrations

When determining the highest test chemical concentration, concentrations that have the capability of producing artifactual positive responses, such as those producing excessive cytotoxicity (see paragraph 22), precipitation in the culture medium (see paragraph 23), or marked changes in pH or osmolality (see paragraph 5), should be avoided. If the test chemical causes a marked change in the pH of the medium at the time of addition, the pH might be adjusted by buffering the final treatment medium so as to avoid artifactual positive results and to maintain appropriate culture conditions.

Measurements of cell proliferation are made to assure that a sufficient number of treated cells have reached mitosis during the test and that the treatments are conducted at appropriate levels of cytotoxicity (see paragraphs 18 and 22). Cytotoxicity should be determined with and without metabolic activation in the main experiment using an appropriate indication of cell death and growth. While the evaluation of cytotoxicity in an initial test may be useful to better define the concentrations to be used in the main experiment, an initial test is not mandatory. If performed, it should not replace the measurement of cytotoxicity in the main experiment.

Relative Population Doubling (RPD) or Relative Increase in Cell Count (RICC) are appropriate methods for the assessment of cytotoxicity in cytogenetic tests (13) (15) (35) (36) (55) (see Appendix 2 for formulas). In case of long-term treatment and sampling times after the beginning of treatment longer than 1,5 normal cell cycle lengths (i.e. longer than 3 cell cycle lengths in total), RPD might underestimate cytotoxicity (37). Under these circumstances RICC might be a better measure or the evaluation of cytotoxicity after 1,5 normal cell cycle lengths would be a helpful estimate using RPD.

For lymphocytes in primary cultures, while the mitotic index (MI) is a measure of cytotoxic/cytostatic effects, it is influenced by the time after treatment it is measured, the mitogen used and possible cell cycle disruption. However, the MI is acceptable because other cytotoxicity measurements may be cumbersome and impractical and may not apply to the target population of lymphocytes growing in response to PHA stimulation.

While RICC and RPD for cell lines and MI for primary culture of lymphocytes are the recommended cytotoxicity parameters, other indicators (e.g. cell integrity, apoptosis, necrosis, cell cycle) could provide useful additional information.

At least three test concentrations (not including the solvent and positive controls) that meet the acceptability criteria (appropriate cytotoxicity, number of cells, etc) should be evaluated. Whatever the types of cells (cell lines or primary cultures of lymphocytes), either replicate or single treated cultures may be used at each concentration tested. While the use of duplicate cultures is advisable, single cultures are also acceptable provided that the same total number of cells are scored for either single or duplicate cultures. The use of single cultures is particularly relevant when more than 3 concentrations are assessed (see paragraph 31). The results obtained in the independent replicate cultures at a given concentration can be pooled for the data analysis (38). For test chemicals demonstrating little or no cytotoxicity, concentration intervals of approximately 2 to 3 fold will usually be appropriate. Where cytotoxicity occurs, the test concentrations selected should cover a range from that producing cytotoxicity as described in paragraph 22 and including concentrations at which there is moderate and little or no cytotoxicity. Many test chemicals exhibit steep concentration response curves and in order to obtain data at low and moderate cytotoxicity or to study the dose response relationship in detail, it will be necessary to use more closely spaced concentrations and/or more than three concentrations (single cultures or replicates), in particular in situations where a repeat experiment is required (see paragraph 47).

If the maximum concentration is based on cytotoxicity, the highest concentration should aim to achieve 55 ± 5 % cytotoxicity using the recommended cytotoxicity parameters (i.e. reduction in RICC and RPD for cell lines and reduction in MI for primary cultures of lymphocytes to 45 ± 5 % of the concurrent negative control). Care should be taken in interpreting positive results only to be found in the higher end of this 55 ± 5 % cytotoxicity range (13).

For poorly soluble test chemicals that are not cytotoxic at concentrations lower than the lowest insoluble concentration, the highest concentration analysed should produce turbidity or a precipitate visible by eye or with the aid of an inverted microscope at the end of the treatment with the test chemical. Even if cytotoxicity occurs above the lowest insoluble concentration, it is advisable to test at only one concentration producing turbidity or with a visible precipitate because artifactual effects may result from the precipitate. At the concentration producing a precipitate, care should be taken to assure that the precipitate does not interfere with the conduct of the test (e.g. staining or scoring). The determination of solubility in the culture medium prior to the experiment may be useful.

If no precipitate or limiting cytotoxicity is observed, the highest test concentration should correspond to 10 mM, 2 mg/ml or 2 μl/ml, whichever is the lowest (39) (40) (41). When the test chemical is not of defined composition, e.g. a substance of unknown or variable composition, complex reaction products or biological material (UVCB) (42), environmental extract etc., the top concentration may need to be higher (e.g. 5 mg/ml), in the absence of sufficient cytotoxicity, to increase the concentration of each of the components. It should be noted however that these requirements may differ for human pharmaceuticals (43).

Controls

Concurrent negative controls (see paragraph 15), consisting of solvent alone in the treatment medium and treated in the same way as the treatment cultures, should be included for every harvest time.

Concurrent positive controls are needed to demonstrate the ability of the laboratory to identify clastogens under the conditions of the test protocol used and the effectiveness of the exogenous metabolic activation system, when applicable. Examples of positive controls are given in the table 1 below. Alternative positive control chemicals can be used, if justified. Because in vitro mammalian cell tests for genetic toxicity are sufficiently standardized, the use of positive controls may be confined to a clastogen requiring metabolic activation. Provided it is done concurrently with the non-activated test using the same treatment duration, this single positive control response will demonstrate both the activity of the metabolic activation system and the responsiveness of the test system. Long term treatment (without S9) should however have its own positive control as the treatment duration will differ from the test using metabolic activation. Each positive control should be used at one or more concentrations expected to give reproducible and detectable increases over background in order to demonstrate the sensitivity of the test system (i.e. the effects are clear but do not immediately reveal the identity of the coded slides to the reader), and the response should not be compromised by cytotoxicity exceeding the limits specified in the test method.

Table 1

Reference chemicals recommended for assessing laboratory proficiency and for selection of positive controls

PROCEDURE

Treatment with test chemical

Proliferating cells are treated with the test chemical in the presence and absence of a metabolic activation system.

Culture harvest time

For thorough evaluation, which would be needed to conclude a negative outcome, all three of the following experimental conditions should be conducted using a short term treatment with and without metabolic activation and long term treatment without metabolic activation (see paragraphs 43, 44 and 45):

In the event that any of the above experimental conditions lead to a positive response, it may not be necessary to investigate any of the other treatment regimens.

Chromosome preparation

Cell cultures are treated with colcemid or colchicine usually for one to three hours prior to harvesting. Each cell culture is harvested and processed separately for the preparation of chromosomes. Chromosome preparation involves hypotonic treatment of the cells, fixation and staining. In monolayers, mitotic cells (identifiable as being round and detaching from the surface) may be present at the end of the 3-6 hour treatment. Because these mitotic cells are easily detached, they can be lost when the medium containing the test chemical is removed. If there is evidence for a substantial increase in the number of mitotic cells compared with controls, indicating likely mitotic arrest, then the cells should be collected by centrifugation and added back to cultures, to avoid losing cells that are in mitosis, and at risk for chromosome aberration, at the time of harvest.

Analysis

All slides, including those of the positive and negative controls, should be independently coded before microscopic analysis for chromosomal aberrations. Since fixation procedures often result in a proportion of metaphase cells which have lost chromosomes, the cells scored should, therefore, contain a number of centromeres equal to the modal number +/- 2.

At least 300 well-spread metaphases should be scored per concentration and control to conclude a test chemical as clearly negative (see paragraph 45). The 300 cells should be equally divided among the replicates, when replicate cultures are used. When single cultures are used per concentration (see paragraph 21), at least 300 well spread metaphases should be scored in this single culture. Scoring 300 cells has the advantage of increasing the statistical power of the test and in addition, zero values will be rarely observed (expected to be only 5 %) (44). The number of metaphases scored can be reduced when high numbers of cells with chromosome aberrations are observed and the test chemical considered as clearly positive.

Cells with structural chromosomal aberration(s) including and excluding gaps should be scored. Breaks and gaps are defined in Appendix 1 according to (45) (46). Chromatid- and chromosome-type aberrations should be recorded separately and classified by sub-types (breaks, exchanges). Procedures in use in the laboratory should ensure that analysis of chromosomal aberrations is performed by well-trained scorers and peer-reviewed if appropriate.

Although the purpose of the test is to detect structural chromosomal aberrations, it is important to record polyploidy and endoreduplication frequencies when these events are seen. (See paragraph 2).

Proficiency of the laboratory

In order to establish sufficient experience with the test prior to using it for routine testing, the laboratory should have performed a series of experiments with reference positive chemicals acting via different mechanisms and various negative controls (using various solvents/vehicle). These positive and negative control responses should be consistent with the literature. This is not applicable to laboratories that have experience, i.e. that have an historical data base available as defined in paragraph 37.

A selection of positive control chemicals (see Table 1 in paragraph 26) should be investigated with short and long treatments in the absence of metabolic activation, and also with short treatment in the presence of metabolic activation, in order to demonstrate proficiency to detect clastogenic chemicals and determine the effectiveness of the metabolic activation system. A range of concentrations of the selected chemicals should be chosen so as to give reproducible and concentration-related increases above the background in order to demonstrate the sensitivity and dynamic range of the test system.

Historical control data

The laboratory should establish:

When first acquiring data for an historical negative control distribution, concurrent negative controls should be consistent with published control data, where they exist. As more experimental data are added to the control distribution, concurrent negative controls should ideally be within the 95 % control limits of that distribution (44) (47). The laboratory's historical negative control database should initially be built with a minimum of 10 experiments but would preferably consist of at least 20 experiments conducted under comparable experimental conditions. Laboratories should use quality control methods, such as control charts (e.g. C-charts or X-bar charts (48)), to identify how variable their positive and negative control data are, and to show that the methodology is ‘under control’ in their laboratory (44). Further recommendations on how to build and use the historical data (i.e. criteria for inclusion and exclusion of data in historical data and the acceptability criteria for a given experiment) can be found in the literature (47).

Any changes to the experimental protocol should be considered in terms of their consistency with the laboratory's existing historical control databases. Any major inconsistencies should result in the establishment of a new historical control database.

Negative control data should consist of the incidence of cells with chromosome aberrations from a single culture or the sum of replicate cultures as described in paragraph 21. Concurrent negative controls should ideally be within the 95 % control limits of the distribution of the laboratory's historical negative control database (44) (47). Where concurrent negative control data fall outside the 95 % control limits they may be acceptable for inclusion in the historical control distribution as long as these data are not extreme outliers and there is evidence that the test system is ‘under control’ (see paragraph 37) and evidence of absence of technical or human failure.

DATA AND REPORTING

Presentation of the results

The percentage of cells with structural chromosomal aberration(s) should be evaluated. Chromatid- and chromosome-type aberrations classified by sub-types (breaks, exchanges) should be listed separately with their numbers and frequencies for experimental and control cultures. Gaps are recorded and reported separately but not included in the total aberration frequency. Percentage of polyploidy and/or endoreduplicated cells are reported when seen.

Concurrent measures of cytotoxicity for all treated, negative and positive control cultures in the main aberration experiment(s) should be recorded.

Individual culture data should be provided. Additionally, all data should be summarised in tabular form.

Acceptability Criteria

Acceptance of a test is based on the following criteria:

Evaluation and interpretation of results

Providing that all acceptability criteria are fulfilled, a test chemical is considered to be clearly positive if, in any of the experimental conditions examined (see paragraph 28):

When all of these criteria are met, the test chemical is then considered able to induce chromosomal aberrations in cultured mammalian cells in this test system. Recommendations for the most appropriate statistical methods can be found in the literature (49) (50) (51).

Providing that all acceptability criteria are fulfilled, a test chemical is considered clearly negative if, in all experimental conditions examined (see paragraph 28):

The test chemical is then considered unable to induce chromosomal aberrations in cultured mammalian cells in this test system.

There is no requirement for verification of a clearly positive or negative response.

In case the response is neither clearly negative nor clearly positive as described above or in order to assist in establishing the biological relevance of a result, the data should be evaluated by expert judgement and/or further investigations. Scoring additional cells (where appropriate) or performing a repeat experiment possibly using modified experimental conditions (e.g. concentration spacing, other metabolic activation conditions (i.e. S9 concentration or S9 origin)) could be useful.

In rare cases, even after further investigations, the data set will preclude making a conclusion of positive or negative results, and therefore the test chemical response will be concluded to be equivocal.

An increase in the number of polyploid cells may indicate that the test chemicals have the potential to inhibit mitotic processes and to induce numerical chromosomal aberrations (52). An increase in the number of cells with endoreduplicated chromosomes may indicate that the test chemicals have the potential to inhibit cell cycle progress (53) (54) (see paragraph 2). Therefore, incidence of polyploid cells and cells with endoreduplicated chromosomes should be recorded separately.

Test report

The test report should include the following information:

LITERATURE:

(4)

In Part B, Chapter B.11 is replaced by the following:

‘B.11   Mammalian Bone Marrow Chromosomal Aberration Test

INTRODUCTION

This test method is equivalent to OECD test guideline 475 (2016). It is part of a series of test methods on genetic toxicology. An OECD document that provides succinct information on genetic toxicology testing and an overview of the recent changes that were made to these Test Guidelines has been developed (1).

The mammalian in vivo bone marrow chromosomal aberration test is especially relevant for assessing genotoxicity because, although they may vary among species, factors of in vivo metabolism, pharmacokinetics and DNA-repair processes are active and contribute to the responses. An in vivo assay is also useful for further investigation of genotoxicity detected by an in vitro system.

The mammalian in vivo chromosomal aberration test is used for the detection of structural chromosome aberrations induced by test chemicals in bone marrow cells of animals, usually rodents (2) (3) (4) (5). Structural chromosomal aberrations may be of two types, chromosome or chromatid. While the majority of genotoxic chemical-induced aberrations are of the chromatid-type, chromosome-type aberrations also occur. Chromosomal damage and related events are the cause of many human genetic diseases and there is substantial evidence that, when these lesions and related events cause alterations in oncogenes and tumour suppressor genes, they are involved in cancer in humans and experimental systems. Polyploidy (including endoreduplication) could arise in chromosome aberration assays in vivo. However, an increase in polyploidy per se does not indicate aneugenic potential and can simply indicate cell cycle perturbation or cytotoxicity. This test is not designed to measure aneuploidy. An in vivo mammalian erythrocyte micronucleus test (Chapter B.12 of this Annex) or the in vitro mammalian cell micronucleus test (Chapter B.49 of this Annex) would be the in vivo and in vitro tests, respectively, recommended for the detection of aneuploidy.

Definitions of terminology used are set out in Appendix 1.

INITIAL CONSIDERATIONS

Rodents are routinely used in this test, but other species may in some cases be appropriate if scientifically justified. Bone marrow is the target tissue in this test since it is a highly vascularised tissue and it contains a population of rapidly cycling cells that can be readily isolated and processed. The scientific justification for using species other than rats and mice should be provided in the report. If species other than rodents are used, it is recommended that the measurement of bone marrow chromosomal aberration be integrated into another appropriate toxicity test.

If there is evidence that the test chemical(s), or its metabolite(s), will not reach the target tissue, it may not be appropriate to use this test.

Before use of the test method on a mixture for generating data for an intended regulatory purpose, it should be considered whether, and if so why, it may provide adequate results for that purpose. Such considerations are not needed, when there is a regulatory requirement for testing of the mixture.

PRINCIPLE OF THE TEST METHOD

Animals are exposed to the test chemical by an appropriate route of exposure and are humanely euthanised at an appropriate time after treatment. Prior to euthanasia, animals are treated with a metaphase-arresting agent (e.g. colchicine or colcemid). Chromosome preparations are then made from the bone marrow cells and stained, and metaphase cells are analysed for chromosomal aberrations.

VERIFICATION OF LABORATORY PROFICIENCY

Proficiency Investigations

In order to establish sufficient experience with the conduct of the assay prior to using it for routine testing, the laboratory should have demonstrated the ability to reproduce expected results from published data (e.g. (6)) for chromosomal aberration frequencies with a minimum of two positive control chemicals (including weak responses induced by low doses of positive controls), such as those listed in Table 1 and with compatible vehicle/solvent controls (see paragraph 22). These experiments should use doses that give reproducible and dose related increases and demonstrate the sensitivity and dynamic range of the test system in the tissue of interest (bone marrow) and using the scoring method to be employed within the laboratory. This requirement is not applicable to laboratories that have experience, i.e. that have a historical database available as defined in paragraphs 10-14.

Historical Control Data

During the course of the proficiency investigations, the laboratory should establish:

When first acquiring data for a historical negative control distribution, concurrent negative controls should be consistent with published control data, where they exist. As more experimental data are added to the historical control distribution, concurrent negative controls should ideally be within the 95 % control limits of that distribution. The laboratory's historical negative control database should be statistically robust to ensure the ability of the laboratory to assess the distribution of their negative control data. The literature suggests that a minimum of 10 experiments may be necessary but would preferably consist of at least 20 experiments conducted under comparable experimental conditions. Laboratories should use quality control methods, such as control charts (e.g. C-charts or X-bar charts (7)), to identify how variable their data are, and to show that the methodology is ‘under control’ in their laboratory. Further recommendations on how to build and use the historical data (i.e. criteria for inclusion and exclusion of data in historical data and the acceptability criteria for a given experiment) can be found in the literature (8).

Where the laboratory does not complete a sufficient number of experiments to establish a statistically robust negative control distribution (see paragraph 11) during the proficiency investigations (described in paragraph 9), it is acceptable that the distribution can be built during the first routine tests. This approach should follow the recommendations set out in the literature (8) and the negative control results obtained in these experiments should remain consistent with published negative control data.

Any changes to the experimental protocol should be considered in terms of their impact on the resulting data remaining consistent with the laboratory's existing historical control database. Only major inconsistencies should result in the establishment of a new historical control database, where expert judgement determines that it differs from the previous distribution (see paragraph 11). During the re-establishment, a full negative control database may not be needed to permit the conduct of an actual test, provided that the laboratory can demonstrate that their concurrent negative control values remain either consistent with their previous database or with the corresponding published data.

Negative control data should consist of the incidence of structural chromosomal aberration (excluding gaps) in each animal. Concurrent negative controls should ideally be within the 95 % control limits of the distribution of the laboratory's historical negative control database. Where concurrent negative control data fall outside the 95 % control limits, they may be acceptable for inclusion in the historical control distribution as long as these data are not extreme outliers and there is evidence that the test system is ‘under control” (see paragraph 11) and no evidence of technical or human failure.

DESCRIPTION OF THE METHOD

Preparations

Selection of animal species

Commonly used laboratory strains of healthy young adult animals should be employed. Rats are commonly used, although mice may also be appropriate. Any other appropriate mammalian species may be used, if scientific justification is provided in the report.

Animal housing and feeding conditions

For rodents, the temperature in the animal room should be 22 °C (± 3 °C). Although the relative humidity ideally should be 50-60 %, it should be at least 40 % and preferably not exceed 70 % other than during room cleaning. Lighting should be artificial, the sequence being 12 hours light, 12 hours dark. For feeding, conventional laboratory diets may be used with an unlimited supply of drinking water. The choice of diet may be influenced by the need to ensure a suitable admixture of a test chemical when administered by this route. Rodents should be housed in small groups (no more than five per cage) of the same sex and treatment group if no aggressive behaviour is expected, preferably in solid floor cages with appropriate environmental enrichment. Animals may be housed individually only if scientifically justified.

Preparation of the animals

Healthy young adult animals (for rodents, ideally 6-10 weeks old at start of treatment, though slightly older animals are also acceptable) are normally used, and are randomly assigned to the control and treatment groups. The individual animals are identified uniquely using a humane, minimally invasive method (e.g. by ringing, tagging, micro-chipping or biometric identification, but not ear or toe clipping) and acclimated to the laboratory conditions for at least five days. Cages should be arranged in such a way that possible effects due to cage placement are minimised. Cross contamination by the positive control and the test chemical should be avoided. At the commencement of the study, the weight variation of animals should be minimal and not exceed ± 20 % of the mean weight of each sex.

Preparation of doses

Solid test chemicals should be dissolved or suspended in appropriate solvents or vehicles or admixed in diet or drinking water prior to dosing the animals. Liquid test chemicals may be dosed directly or diluted prior to dosing. For inhalation exposures, test chemicals can be administered as a gas, vapour, or a solid/liquid aerosol, depending on their physicochemical properties. Fresh preparations of the test chemical should be employed unless stability data demonstrate the acceptability of storage and define the appropriate storage conditions.

Solvent/vehicle

The solvent/vehicle should not produce toxic effects at the dose levels used, and should not be suspected of chemical reaction with the test chemicals. If other than well-known solvents/vehicles are used, their inclusion should be supported with reference data indicating their compatibility. It is recommended that wherever possible, the use of an aqueous solvent/vehicle should be considered first. Examples of commonly used compatible solvents/vehicles include water, physiological saline, methylcellulose solution, carboxymethyl cellulose sodium salt solution, olive oil and corn oil. In the absence of historical or published control data showing that no structural aberrations or other deleterious effects are induced by a chosen atypical solvent/vehicle, an initial study should be conducted in order to establish the acceptability of the solvent/vehicle control.

Controls

Positive controls

A group of animals treated with a positive control chemical should normally be included with each test. This may be waived when the testing laboratory has demonstrated proficiency in the conduct of the test and has established a historical positive control range. When a concurrent positive control group is not included, scoring controls (fixed and unstained slides) should be included in each experiment. These can be obtained by including within the scoring of the study appropriate reference samples that have been obtained and stored from a separate positive control experiment conducted periodically (e.g. every 6-18 months) in the laboratory where the test is performed; for example, during proficiency testing and on a regular basis thereafter, where necessary.

Positive control chemicals should reliably produce a detectable increase in the frequency of cells with structural chromosomal aberrations over the spontaneous level. Positive control doses should be chosen so that the effects are clear but do not immediately reveal the identity of the coded samples to the scorer. It is acceptable that the positive control be administered by a route different from the test chemical, using a different treatment schedule, and for sampling to occur only at a single time point. In addition, the use of chemical class-related positive control chemicals may be considered, when appropriate. Examples of positive control chemicals are included in Table 1.

Table 1

Examples of positive control chemicals

Negative controls

Negative control group animals should be included at every sampling time and otherwise handled in the same way as the treatment groups, except for not receiving treatment with the test chemical. If a solvent/vehicle is used in administering the test chemical, the control group should receive this solvent/vehicle. However, if consistent inter-animal variability and frequencies of cells with structural aberrations are demonstrated by historical negative control data at each sampling time for the testing laboratory, only a single sampling for the negative control may be necessary. Where a single sampling is used for negative controls, it should be the first sampling time used in the study.

PROCEDURE

Number and sex of animals

In general, the micronucleus response is similar between male and female animals (9) and it is expected that this will be true also for structural chromosomal aberrations; therefore, most studies could be performed in either sex. Data demonstrating relevant differences between males and females (e.g. differences in systemic toxicity, metabolism, bioavailability, bone marrow toxicity, etc. including e.g. a range-finding study) would encourage the use of both sexes. In this case, it may be appropriate to perform a study in both sexes, e.g. as part of a repeated dose toxicity study. It might be appropriate to use the factorial design in case both sexes are used. Details on how to analyse the data using this design are given in Appendix 2.

Group sizes at study initiation should be established with the aim of providing a minimum of 5 analysable animals of one sex, or of each sex if both are used, per group. Where human exposure to chemicals may be sex-specific, as for example with some pharmaceuticals, the test should be performed with the appropriate sex. As a guide to maximum typical animal requirements, a study in bone marrow at two sampling times with three dose groups and a concurrent negative control group, plus a positive control group (each group composed of five animals of a single sex), would require 45 animals.

Dose levels

If a preliminary range-finding study is performed because there are no suitable data already available to aid in dose selection, it should be performed in the same laboratory, using the same species, strain, sex, and treatment regimen to be used in the main study (10). The study should aim to identify the maximum tolerated dose (MTD), defined as the highest dose that will be tolerated without evidence of study-limiting toxicity, relative to the duration of the study period (for example, by inducing body weight depression or hematopoietic system cytotoxicity), but not death or evidence of pain, suffering or distress necessitating humane euthanasia (11).

The highest dose may also be defined as a dose that produces some indication of toxicity to the bone marrow.

Chemicals that exhibit saturation of toxicokinetic properties, or induce detoxification processes that may lead to a decrease in exposure after long-term treatment may be exceptions to the dose-setting criteria and should be evaluated on a case-by-case basis.

In order to obtain dose response information, a complete study should include a negative control group and a minimum of three dose levels generally separated by a factor of 2, but not greater than 4. If the test chemical does not produce toxicity in a range-finding study or based on existing data, the highest dose for a single administration should be 2 000 mg/kg body weight. However, if the test chemical does cause toxicity, the MTD should be the highest dose administered and the dose levels used should preferably cover a range from the maximum to a dose producing little or no toxicity. When target tissue (bone marrow) toxicity is observed at all dose levels tested, further study at non-toxic doses is advisable. Studies intending to more fully characterise the quantitative dose-response information may require additional dose groups. For certain types of test chemicals (e.g. human pharmaceuticals) covered by specific requirements, these limits may vary.

Limit test

If dose range-finding experiments, or existing data from related animal strains, indicate that a treatment regime of at least the limit dose (described below) produces no observable toxic effects, (including no depression of bone marrow proliferation or other evidence of target tissue cytotoxicity), and if genotoxicity would not be expected based upon in vitro genotoxicity studies or data from structurally related chemicals, then a full study using three dose levels may not be considered necessary, provided it has been demonstrated that the test chemical(s) reach(es) the target tissue (bone marrow). In such cases, a single dose level, at the limit dose, may be sufficient. For an administration period of > 14 days, the limit dose is 1 000 mg/kg body weight/day. For administration periods of 14 days or less, the limit dose is 2 000 mg/kg/body weight/day.

Administration of doses

The anticipated route of human exposure should be considered when designing an assay. Therefore, routes of exposure such as dietary, drinking water, topical, subcutaneous, intravenous, oral (by gavage), inhalation, intratracheal, or implantation may be chosen as justified. In any case, the route should be chosen to ensure adequate exposure of the target tissue(s). Intraperitoneal injection is generally not recommended since it is not an intended route of human exposure, and should only be used with specific scientific justification. If the test chemical is admixed in diet or drinking water, especially in case of single dosing, care should be taken that the delay between food and water consumption and sampling should be sufficient to allow detection of the effects (see paragraphs 33-34). The maximum volume of liquid that can be administered by gavage or injection at one time depends on the size of the test animal. The volume should not normally exceed 1 ml/100 g body weight except in the case of aqueous solutions where a maximum of 2 ml/100 g may be used. The use of volumes greater than this should be justified. Except for irritating or corrosive test chemicals, which will normally produce exacerbated effects at higher concentrations, variability in test volume should be minimised by adjusting the concentration to ensure administration of a constant volume in relation to body weight at all dose levels.

Treatment schedule

Test chemicals are normally administered as a single treatment, but may be administered as a split dose (i.e. two or more treatments on the same day separated by no more than 2-3 hours) to facilitate administering a large volume. Under these circumstances, or when administering the test chemical by inhalation, the sampling time should be scheduled based on the time of the last dosing or the end of exposure.

There are little data available on the suitability of a repeated-dose protocol for this test. However, in circumstances where it is desirable to integrate this test with a repeated-dose toxicity test, care should be taken to avoid loss of chromosomally damaged mitotic cells as may occur with toxic doses. Such integration is acceptable when the highest dose is greater or equal to the limit dose (see paragraph 29) and a dose group is administered the limit dose for the duration of the treatment period. The micronucleus test (test method B.12) should be viewed as the in vivo test of choice for chromosomal aberrations when integration with other studies is desired.

Bone marrow samples should be taken at two separate times following single treatments. For rodents, the first sampling interval should be the time necessary to complete 1,5 normal cell cycle lengths (the latter being normally 12-18 hours following the treatment period). Since the time required for uptake and metabolism of the test chemical(s) as well as its effect on cell cycle kinetics can affect the optimum time for chromosomal aberration detection, a later sample collection 24 hours after the first sampling time is recommended. At the first sampling time, all dose groups should be treated and samples collected for analysis; however, at the later sampling time(s), only the highest dose needs to be administered. If dose regimens of more than one day are used based on scientific justification, one sampling time at up to approximately 1,5 normal cell cycle lengths after the final treatment should generally be used.

Following treatment and prior to sample collection, animals are injected intraperitoneally with an appropriate dose of a metaphase-arresting agent (e.g. colcemid or colchicine), and samples are collected at an appropriate interval thereafter. For mice this interval is approximately 3-5 hours prior to collection and for rats it is 2-5 hours. Cells are harvested from the bone marrow, swollen, fixed and stained, and analysed for chromosomal aberrations (12).

Observations

General clinical observations of the test animals should be made and clinical signs recorded at least once a day, preferably at the same time(s) each day and considering the peak period of anticipated effects after dosing. At least twice daily during the dosing period, all animals should be observed for morbidity and mortality. All animals should be weighed at study initiation, at least once a week during repeated-dose studies, and at euthanasia. In studies of at least one-week duration, measurements of food consumption should be made at least weekly. If the test chemical is administered via the drinking water, water consumption should be measured at each change of water and at least weekly. Animals exhibiting non-lethal indicators of excessive toxicity should be humanely euthanised prior to completion of the test period (11).

Target tissue exposure

A blood sample should be taken at appropriate time(s) in order to permit investigation of the plasma levels of the test chemicals for the purposes of demonstrating that exposure of the bone marrow occurred, where warranted and where other exposure data do not exist (see paragraph 44).

Bone marrow and chromosome preparations

Immediately after humane euthanasia, bone marrow cells are obtained from the femurs or tibias of the animals, exposed to hypotonic solution and fixed. The metaphase cells are then spread on slides and stained using established methods (see (3) (12)).

Analysis

All slides, including those of positive and negative controls, should be independently coded before analysis and should be randomised so the scorer is unaware of the treatment condition.

The mitotic index should be determined as a measure of cytotoxicity in at least 1 000 cells per animal for all treated animals (including positive controls), untreated or vehicle/solvent negative control animals.

At least 200 metaphases should be analysed for each animal for structural chromosomal aberrations including and excluding gaps (6). However, if the historical negative control database indicates the mean background structural chromosomal aberration frequency is < 1 % in the testing laboratory, consideration should be given to scoring additional cells. Chromatid and chromosome-type aberrations should be recorded separately and classified by sub-types (breaks, exchanges). Procedures in use in the laboratory should ensure that analysis of chromosomal aberrations is performed by well-trained scorers and peer-reviewed if appropriate. Recognising that slide preparation procedures often result in the breakage of a proportion of metaphases with a resulting loss of chromosomes, the cells scored should, therefore, contain a number of centromeres not less than 2n ± 2, where n is the haploid number of chromosomes for that species.

DATA AND REPORTING

Treatment of Results

Individual animal data should be presented in tabular form. The mitotic index, the number of metaphase cells scored, the number of aberrations per metaphase cell and the percentage of cells with structural chromosomal aberration(s) should be evaluated for each animal. Different types of structural chromosomal aberrations should be listed with their numbers and frequencies for treated and control groups. Gaps, as well as polyploid cells and cells with endoreduplicated chromosomes are recorded separately. The frequency of gaps is reported but generally not included in the analysis of the total structural aberration frequency. If there is no evidence for a difference in response between the sexes, the data may be combined for statistical analysis. Data on animal toxicity and clinical signs should also be reported.

Acceptability Criteria

The following criteria determine the acceptability of the test:

Evaluation and Interpretation of Results

Providing that all acceptability criteria are fulfilled, a test chemical is considered clearly positive if:

If only the highest dose is examined at a particular sampling time, a test chemical is considered clearly positive if there is a statistically significant increase compared with the concurrent negative control and the results are outside the distribution of the historical negative control data (e.g. Poisson-based 95 % control limits). Recommendations for appropriate statistical methods can be found in the literature (13). When conducting a dose-response analysis, at least three treated dose groups should be analysed. Statistical tests should use the animal as the experimental unit. Positive results in the chromosomal aberration test indicate that a test chemical induces structural chromosomal aberrations in the bone marrow of the species tested.

Providing that all acceptability criteria are fulfilled, a test chemical is considered clearly negative if in all experimental conditions examined:

Recommendations for the most appropriate statistical methods can be found in the literature (13). Evidence of exposure of the bone marrow to a test chemical may include a depression of the mitotic index or measurement of the plasma or blood levels of the test chemical(s). In the case of intravenous administration, evidence of exposure is not needed. Alternatively, ADME data, obtained in an independent study using the same route and same species can be used to demonstrate bone marrow exposure. Negative results indicate that, under the test conditions, the test chemical does not induce structural chromosomal aberrations in the bone marrow of the species tested.

There is no requirement for verification of a clear positive or clear negative response.

In cases where the response is not clearly negative or positive and in order to assist in establishing the biological relevance of a result (e.g. a weak or borderline increase), the data should be evaluated by expert judgement and/or further investigations of the existing experiments completed. In some cases, analysing more cells or performing a repeat experiment using modified experimental conditions could be useful.

In rare cases, even after further investigations, the data will preclude making a conclusion that the test chemical produces either positive or negative results, and the study will therefore be concluded as equivocal.

The frequencies of polyploid and endoreduplicated metaphases among total metaphases should be recorded separately. An increase in the number of polyploid/endoreduplicated cells may indicate that the test chemical has the potential to inhibit mitotic processes or cell cycle progression (see paragraph 3).

Test Report

The test report should include the following information:

Summary

LITERATURE:

(5)

In Part B, Chapter B.12 is replaced by the following:

‘B.12   Mammalian Erythrocyte Micronucleus Test

INTRODUCTION

This test method is equivalent to OECD test guideline 474 (2016). It is part of a series of test methods on genetic toxicology. An OECD document that provides succinct information on genetic toxicology testing and an overview of the recent changes that were made to these Test Guidelines has been developed (1).

The mammalian in vivo micronucleus test is especially relevant for assessing genotoxicity because, although they may vary among species, factors of in vivo metabolism, pharmacokinetics and DNA repair processes are active and contribute to the responses. An in vivo assay is also useful for further investigation of genotoxicity detected by an in vitro system.

The mammalian in vivo micronucleus test is used for the detection of damage induced by the test chemical to the chromosomes or the mitotic apparatus of erythroblasts. The test evaluates micronucleus formation in erythrocytes sampled either in the bone marrow or peripheral blood cells of animals, usually rodents.

The purpose of the micronucleus test is to identify chemicals that cause cytogenetic damage which results in the formation of micronuclei containing either lagging chromosome fragments or whole chromosomes.

When a bone marrow erythroblast develops into an immature erythrocyte (sometimes also referred to as a polychromatic erythrocyte or reticulocyte), the main nucleus is extruded; any micronucleus that has been formed may remain behind in the cytoplasm. Visualisation or detection of micronuclei is facilitated in these cells because they lack a main nucleus. An increase in the frequency of micronucleated immature erythrocytes in treated animals is an indication of induced structural or numerical chromosomal aberrations.

Newly formed micronucleated erythrocytes are identified and quantitated by staining followed by either visual scoring using a microscope, or by automated analysis. Counting sufficient immature erythrocytes in the peripheral blood or bone marrow of adult animals is greatly facilitated by using an automated scoring platform. Such platforms are acceptable alternatives to manual evaluation (2). Comparative studies have shown that such methods, using appropriate calibration standards, can provide better inter- and intra-laboratory reproducibility and sensitivity than manual microscopic scoring (3) (4). Automated systems that can measure micronucleated erythrocyte frequencies include, but are not limited to, flow cytometers (5), image analysis platforms (6) (7), and laser scanning cytometers (8).

Although not normally done as part of the test, chromosome fragments can be distinguished from whole chromosomes by a number of criteria. These include identification of the presence or absence of a kinetochore or centromeric DNA, both of which are characteristic of intact chromosomes. The absence of kinetochore or centromeric DNA indicates that the micronucleus contains only fragments of chromosomes, while the presence is indicative of chromosome loss.

Definitions of terminology used are set out in Appendix 1.

INITIAL CONSIDERATIONS

The bone marrow of young adult rodents is the target tissue for genetic damage in this test since erythrocytes are produced in this tissue. The measurement of micronuclei in immature erythrocytes in peripheral blood is acceptable in other mammalian species for which adequate sensitivity to detect chemicals that cause structural or numerical chromosomal aberrations in these cells has been demonstrated (by induction of micronuclei in immature erythrocytes) and scientific justification is provided. The frequency of micronucleated immature erythrocytes is the principal endpoint. The frequency of mature erythrocytes that contain micronuclei in the peripheral blood also can be used as an endpoint in species without strong splenic selection against micronucleated cells and when animals are treated continuously for a period that exceeds the lifespan of the erythrocyte in the species used (e.g. 4 weeks or more in the mouse).

If there is evidence that the test chemical(s), or its metabolite(s), will not reach the target tissue, it may not be appropriate to use this test.

Before use of the test method on a mixture for generating data for an intended regulatory purpose, it should be considered whether, and if so why, it may provide adequate results for that purpose. Such considerations are not needed, when there is a regulatory requirement for testing of the mixture.

PRINCIPLE OF THE TEST METHOD

Animals are exposed to the test chemical by an appropriate route. If bone marrow is used, the animals are humanely euthanised at an appropriate time(s) after treatment, the bone marrow is extracted, and preparations are made and stained (9) (10) (11) (12) (13) (14) (15). When peripheral blood is used, the blood is collected at an appropriate time(s) after treatment and preparations are made and stained (12) (16) (17) (18). When treatment is administered acutely, it is important to select bone marrow or blood harvest times at which the treatment-related induction of micronucleated immature erythrocytes can be detected. In the case of peripheral blood sampling, enough time must also have elapsed for these events to appear in circulating blood. Preparations are analysed for the presence of micronuclei, either by visualisation using a microscope, image analysis, flow cytometry, or laser scanning cytometry.

VERIFICATION OF LABORATORY PROFICIENCY

Proficiency Investigations

In order to establish sufficient experience with the conduct of the assay prior to using it for routine testing, the laboratory should have demonstrated the ability to reproduce expected results from published data (17) (19) (20) (21) (22) for micronucleus frequencies with a minimum of two positive control chemicals (including weak responses induced by low doses of positive controls), such as those listed in Table 1 and with compatible vehicle/solvent controls (see paragraph 26). These experiments should use doses that give reproducible and dose-related increases and demonstrate the sensitivity and dynamic range of the test system in the tissue of interest (bone marrow or peripheral blood) and using the scoring method to be employed within the laboratory. This requirement is not applicable to laboratories that have experience, i.e. that have a historical database available as defined in paragraphs 14-18.

Historical Control Data

During the course of the proficiency investigations, the laboratory should establish:

When first acquiring data for a historical negative control distribution, concurrent negative controls should be consistent with published control data, where they exist. As more experimental data are added to the historical control distribution, concurrent negative controls should ideally be within the 95 % control limits of that distribution. The laboratory's historical negative control database should be statistically robust to ensure the ability of the laboratory to assess the distribution of their negative control data. The literature suggests that a minimum of 10 experiments may be necessary but would preferably consist of at least 20 experiments conducted under comparable experimental conditions. Laboratories should use quality control methods, such as control charts (e.g. C-charts or X-bar charts (23)), to identify how variable their data are, and to show that the methodology is ‘under control’ in their laboratory. Further recommendations on how to build and use the historical data (i.e. criteria for inclusion and exclusion of data in historical data and the acceptability criteria for a given experiment) can be found in the literature (24).

Where the laboratory does not complete a sufficient number of experiments to establish a statistically robust negative control distribution (see paragraph 15) during the proficiency investigations (described in paragraph 13), it is acceptable that the distribution can be built during the first routine tests. This approach should follow the recommendations set out in the literature (24) and the negative control results obtained in these experiments should remain consistent with published negative control data.

Any changes to the experimental protocol should be considered in terms of their impact on the resulting data remaining consistent with the laboratory's existing historical control database. Only major inconsistencies should result in the establishment of a new historical control database where expert judgement determines that it differs from the previous distribution (see paragraph 15). During the re-establishment, a full negative control database may not be needed to permit the conduct of an actual test, provided that the laboratory can demonstrate that their concurrent negative control values remain either consistent with their previous database or with the corresponding published data.

Negative control data should consist of the incidence of micronucleated immature erythrocytes in each animal. Concurrent negative controls should ideally be within the 95 % control limits of the distribution of the laboratory's historical negative control database. Where concurrent negative control data fall outside the 95 % control limits, they may be acceptable for inclusion in the historical control distribution as long as these data are not extreme outliers and there is evidence that the test system is ‘under control’ (see paragraph 15) and no evidence of technical or human failure.

DESCRIPTION OF THE METHOD

Preparations

Selection of animal species

Commonly used laboratory strains of healthy young adult animals should be employed. Mice, rats, or another appropriate mammalian species may be used. When peripheral blood is used, it must be established that splenic removal of micronucleated cells from the circulation does not compromise the detection of induced micronuclei in the species selected. This has been clearly demonstrated for mouse and rat peripheral blood (2). The scientific justification for using species other than rats and mice should be provided in the report. If species other than rodents are used, it is recommended that the measurement of induced micronuclei be integrated into another appropriate toxicity test.

Animal housing and feeding conditions

For rodents, the temperature in the animal room should be 22 °C (± 3 °C). Although the relative humidity ideally should be 50-60 %, it should be at least 40 % and preferably not exceed 70 % other than during room cleaning. Lighting should be artificial, the sequence being 12 hours light, 12 hours dark. For feeding, conventional laboratory diets may be used with an unlimited supply of drinking water. The choice of diet may be influenced by the need to ensure a suitable admixture of a test chemical when administered by this route. Rodents should be housed in small groups (no more than five per cage) of the same sex and treatment group if no aggressive behaviour is expected, preferably in solid floor cages with appropriate environmental enrichment. Animals may be housed individually only if scientifically justified.

Preparation of the animals

Healthy young adult animals (for rodents, ideally 6-10 weeks old at start of treatment, though slightly older animals are also acceptable) are normally used, and are randomly assigned to the control and treatment groups. The individual animals are identified uniquely using a humane, minimally invasive method (e.g. by ringing, tagging, micro-chipping or biometric identification, but not ear or toe clipping) and acclimated to the laboratory conditions for at least five days. Cages should be arranged in such a way that possible effects due to cage placement are minimised. Cross contamination by the positive control and the test chemical should be avoided. At the commencement of the study, the weight variation of animals should be minimal and not exceed ± 20 % of the mean weight of each sex.

Preparation of doses

Solid test chemicals should be dissolved or suspended in appropriate solvents or vehicles or admixed in diet or drinking water prior to dosing the animals. Liquid test chemicals may be dosed directly or diluted prior to dosing. For inhalation exposures, test chemicals can be administered as a gas, vapour, or a solid/liquid aerosol, depending on their physicochemical properties. Fresh preparations of the test chemical should be employed unless stability data demonstrate the acceptability of storage and define the appropriate storage conditions.

Test Conditions

Solvent/vehicle

The solvent/vehicle should not produce toxic effects at the dose levels used, and should not be capable of chemical reaction with the test chemicals. If other than well-known solvents/vehicles are used, their inclusion should be supported with reference data indicating their compatibility. It is recommended that wherever possible, the use of an aqueous solvent/vehicle should be considered first. Examples of commonly used compatible solvents/vehicles include water, physiological saline, methylcellulose solution, carboxymethyl cellulose sodium salt solution, olive oil and corn oil. In the absence of historical or published control data showing that no micronuclei and other deleterious effects are induced by a chosen atypical solvent/vehicle, an initial study should be conducted in order to establish the acceptability of the solvent/vehicle control.

Controls

Positive controls

A group of animals treated with a positive control chemical should normally be included with each test. This may be waived when the testing laboratory has demonstrated proficiency in the conduct of the test and has established a historical positive control range. When a concurrent positive control group is not included, scoring controls (fixed and unstained slides or cell suspension samples, as appropriate for the method of scoring) should be included in each experiment. These can be obtained by including within the scoring of the study appropriate reference samples that have been obtained and stored from a separate positive control experiment conducted periodically (e.g. every 6-18 months); for example, during proficiency testing and on a regular basis thereafter, where necessary.

Positive control chemicals should reliably produce a detectable increase in micronucleus frequency over the spontaneous level. When employing manual scoring by microscopy, positive control doses should be chosen so that the effects are clear but do not immediately reveal the identity of the coded samples to the scorer. It is acceptable that the positive control be administered by a route different from the test chemical, using a different treatment schedule, and for sampling to occur only at a single time point. In addition, the use of chemical class-related positive control chemicals may be considered, when appropriate. Examples of positive control chemicals are included in Table 1.

Table 1

Examples of positive control chemicals

Chemicals and CASRN

Ethyl methanesulphonate [CASRN 62-50-0]

Methyl methanesulphonate [CASRN 66-27-3]

Ethyl nitrosourea [CASRN 759-73-9]

Mitomycin C [CASRN 50-07-7]

Cyclophosphamide (monohydrate) [CASRN 50-18-0 (CASRN 6055-19-2)]

Triethylenemelamine [CASRN 51-18-3]

Colchicine [CASRN 64-86-8] or Vinblastine [CASRN 865-21-4] — as aneugens

Negative controls

Negative control group animals should be included at every sampling time and otherwise handled in the same way as the treatment groups, except for not receiving treatment with the test chemical. If a solvent/vehicle is used in administering the test chemical, the control group should receive this solvent/vehicle. However, if consistent inter-animal variability and frequencies of cells with micronuclei are demonstrated by historical negative control data at each sampling time for the testing laboratory, only a single sampling for the negative control may be necessary. Where a single sampling is used for negative controls, it should be the first sampling time used in the study.

If peripheral blood is used, a pre-treatment sample is acceptable instead of a concurrent negative control for short-term studies when the resulting data are consistent with the historical control database for the testing laboratory. It has been shown for rats that pre-treatment sampling of small volumes (e.g. below 100 μl/day) has minimal impact on micronucleus background frequency (25).

PROCEDURE

Number and sex of animals

In general, the micronucleus response is similar between male and female animals and, therefore, most studies could be performed in either sex (26). Data demonstrating relevant differences between males and females (e.g. differences in systemic toxicity, metabolism, bioavailability, bone marrow toxicity, etc. including e.g. in a range-finding study) would encourage the use of both sexes. In this case, it may be appropriate to perform a study in both sexes, e.g. as part of a repeated dose toxicity study. It might be appropriate to use the factorial design in case both sexes are used. Details on how to analyse the data using this design are given in Appendix 2.

Group sizes at study initiation should be established with the aim of providing a minimum of 5 analysable animals of one sex, or of each sex if both are used, per group. Where human exposure to chemicals may be sex-specific, as for example with some pharmaceuticals, the test should be performed with the appropriate sex. As a guide to maximum typical animal requirements, a study in bone marrow conducted according to the parameters established in paragraph 37 with three dose groups and concurrent negative and positive controls (each group composed of five animals of a single sex) would require between 25 and 35 animals.

Dose levels

If a preliminary range-finding study is performed because there are no suitable data already available to aid in dose selection, it should be performed in the same laboratory, using the same species, strain, sex, and treatment regimen to be used in the main study (27). The study should aim to identify the maximum tolerated dose (MTD), defined as the highest dose that will be tolerated without evidence of study-limiting toxicity, relative to the duration of the study period (for example, by inducing body weight depression or hematopoietic system cytotoxicity, but not death or evidence of pain, suffering or distress necessitating humane euthanasia (28)).

The highest dose may also be defined as a dose that produces toxicity in the bone marrow (e.g. a reduction in the proportion of immature erythrocytes among total erythrocytes in the bone marrow or peripheral blood of more than 50 %, but to not less than 20 % of the control value). However, when analysing CD71-positive cells in peripheral blood circulation (i.e., by flow cytometry), this very young fraction of immature erythrocytes responds to toxic challenges more quickly than the larger RNA-positive cohort of immature erythrocytes. Therefore, higher apparent toxicity may be evident with acute exposure designs examining the CD71-positive immature erythrocyte fraction as compared to those that identify immature erythrocytes based on RNA content. For this reason, when experiments utilise five or fewer days of treatment, the highest dose level for test chemicals causing toxicity may be defined as the dose that causes a statistically significant reduction in the proportion of CD71-positive immature erythrocytes among total erythrocytes but not to less than 5 % of the control value (29).

Chemicals that exhibit saturation of toxicokinetic properties, or induce detoxification processes that may lead to a decrease in exposure after long-term administration may be exceptions to the dose-setting criteria and should be evaluated on a case-by-case basis.

In order to obtain dose response information, a complete study should include a negative control group and a minimum of three dose levels generally separated by a factor of 2, but not greater than 4. If the test chemical does not produce toxicity in a range-finding study or based on existing data, the highest dose for an administration period of 14 days or more should be 1 000 mg/kg body weight/day, or for administration periods of less than 14 days, 2 000 mg/kg/body weight/day. However, if the test chemical does cause toxicity, the MTD should be the highest dose administered and the dose levels used should preferably cover a range from the maximum to a dose producing little or no toxicity. When target tissue (bone marrow) toxicity is observed at all dose levels tested, further study at non-toxic doses is advisable. Studies intending to more fully characterise the quantitative dose-response information may require additional dose groups. For certain types of test chemicals (e.g. human pharmaceuticals) covered by specific requirements, these limits may vary.

Limit test

If dose range-finding experiments, or existing data from related animal strains, indicate that a treatment regime of at least the limit dose (described below) produces no observable toxic effects, (including no depression of bone marrow proliferation or other evidence of target tissue cytotoxicity), and if genotoxicity would not be expected based upon in vitro genotoxicity studies or data from structurally related chemicals, then a full study using three dose levels may not be considered necessary, provided it has been demonstrated that the test chemical(s) reach(es) the target tissue (bone marrow). In such cases, a single dose level, at the limit dose, may be sufficient. When administration occurs for 14 days or more, the limit dose is 1 000 mg/kg body weight/day. For administration periods of less than 14 days, the limit dose is 2 000 mg/kg/body weight/day.

Administration of doses

The anticipated route of human exposure should be considered when designing an assay. Therefore, routes of exposure such as dietary, drinking water, topical subcutaneous, intravenous, oral (by gavage), inhalation, intratracheal, or implantation may be chosen as justified. In any case, the route should be chosen to ensure adequate exposure of the target tissue(s). Intraperitoneal injection is generally not recommended since it is not an intended route of human exposure, and should only be used with specific scientific justification. If the test chemical is admixed in diet or drinking water, especially in case of single dosing, care should be taken that the delay between food and water consumption and sampling should be sufficient to allow detection of the effects (see paragraph 37). The maximum volume of liquid that can be administered by gavage or injection at one time depends on the size of the test animal. The volume should not normally exceed 1 ml/100 g body weight except in the case of aqueous solutions where a maximum of 2 ml/100 g may be used. The use of volumes greater than this should be justified. Except for irritating or corrosive test chemicals, which will normally produce exacerbated effects at higher concentrations, variability in test volume should be minimised by adjusting the concentration to ensure administration of a constant volume in relation to body weight at all dose levels.

Treatment schedule

Preferably, 2 or more treatments are performed, administered at 24-hour intervals, especially when integrating this test into other toxicity studies. In the alternative, single treatments can be administered, if scientifically justified (e.g. test chemicals known to block cell cycle). Test chemicals also may be administered as a split dose, i.e., two or more treatments on the same day separated by no more than 2-3 hours, to facilitate administering a large volume. Under these circumstances, or when administering the test chemical by inhalation, the sampling time should be scheduled based on the time of the last dosing or the end of exposure.

The test may be performed in mice or rats in one of three ways:

Other dosing or sampling regimens may be used when relevant and scientifically justified, and to facilitate integration with other toxicity tests.

Observations

General clinical observations of the test animals should be made and clinical signs recorded at least once a day, preferably at the same time(s) each day and considering the peak period of anticipated effects after dosing. At least twice daily during the dosing period, all animals should be observed for morbidity and mortality. All animals should be weighed at study initiation, at least once a week during repeated dose studies, and at euthanasia. In studies of at least one-week duration, measurements of food consumption should be made at least weekly. If the test chemical is administered via the drinking water, water consumption should be measured at each change of water and at least weekly. Animals exhibiting non-lethal indicators of excessive toxicity should be humanely euthanised prior to completion of the test period (28). Under certain circumstances, animal body temperature could be monitored, since treatment-induced hyper- and hypothermia have been implicated in producing spurious results (32) (33) (34).

Target tissue exposure

A blood sample should be taken at appropriate time(s) in order to permit investigation of the plasma levels of the test chemicals for the purposes of demonstrating that exposure of the bone marrow occurred, where warranted and where other exposure data do not exist (see paragraph 48).

Bone marrow / blood preparation

Bone marrow cells are usually obtained from the femurs or tibias of the animals immediately following humane euthanasia. Commonly, cells are removed, prepared and stained using established methods. Small volumes of peripheral blood can be obtained, according to adequate animal welfare standards, either using a method that permits survival of the test animal, such as bleeding from the tail vein or other appropriate blood vessel, or by cardiac puncture or sampling from a large vessel at animal euthanasia. For both bone marrow or peripheral blood-derived erythrocytes, depending on the method of analysis, cells may be immediately stained supravitally (16) (17) (18), smear preparations are made and then stained for microscopy, or fixed and stained appropriately for flow cytometric analysis. The use of a DNA specific stain [e.g. acridine orange (35) or Hoechst 33258 plus pyronin-Y (36)] can eliminate some of the artifacts associated with using a non-DNA specific stain. This advantage does not preclude the use of conventional stains (e.g. Giemsa for microscopic analysis). Additional systems [e.g. cellulose columns to remove nucleated cells (37) (38)] also can be used provided that these systems have been demonstrated to be compatible with sample preparation in the laboratory.

Where these methods are applicable, anti-kinetochore antibodies (39), FISH with pancentromeric DNA probes (40), or primed in situ labelling with pancentromere-specific primers, together with appropriate DNA counterstaining (41), can be used to identify the nature of the micronuclei (chromosome/chromosomal fragment) in order to determine whether the mechanism of micronucleus induction is due to clastogenic and/or aneugenic activity. Other methods for differentiation between clastogens and aneugens may be used if they have been shown to be effective.

Analysis (manual and automated)

All slides or samples for analysis, including those of positive and negative controls, should be independently coded before any type of analysis and should be randomised so the manual scorer is unaware of the treatment condition; such coding is not necessary when using automated scoring systems which do not rely on visual inspection and cannot be affected by operator bias. The proportion of immature among total (immature + mature) erythrocytes is determined for each animal by counting a total of at least 500 erythrocytes for bone marrow and 2 000 erythrocytes for peripheral blood (42). At least 4 000 immature erythrocytes per animal should be scored for the incidence of micronucleated immature erythrocytes (43). If the historical negative control database indicates the mean background micronucleated immature erythrocyte frequency is < 0,1 % in the testing laboratory, consideration should be given to scoring additional cells. When analysing samples, the proportion of immature erythrocytes to total erythrocytes in treated animals should not be less than 20 % of the vehicle/solvent control proportion when scoring by microscopy and not less than approximately 5 % of the vehicle/solvent control proportion when scoring CD71+ immature erythrocytes by cytometric methods (see paragraph 31) (29). For example, for a bone marrow assay scored by microscopy, if the control proportion of immature erythrocytes in the bone marrow is 50 %, the upper limit of toxicity would be 10 % immature erythrocytes.

Because the rat spleen sequesters and destroys micronucleated erythrocytes, to maintain high assay sensitivity when analysing rat peripheral blood, it is preferable to restrict the analysis of micronucleated immature erythrocytes to the youngest fraction. When using automated analysis methods, these most immature erythrocytes can be identified based on their high RNA content, or the high level of transferrin receptors (CD71+) expressed on their surface (31). However, direct comparison of different staining methods has shown that satisfactory results can be obtained with various methods, including conventional acridine orange staining (3) (4).

DATA AND REPORTING

Treatment of Results

Individual animal data should be presented in tabular form. The number of immature erythrocytes scored, the number of micronucleated immature erythrocytes, and the proportion of immature among total erythrocytes should be listed separately for each animal analysed. When mice are treated continuously for 4 weeks or more, the data on the number and proportion of micronucleated mature erythrocytes also should be given if collected. Data on animal toxicity and clinical signs should also be reported.

Acceptability Criteria

The following criteria determine the acceptability of the test:

Evaluation and Interpretation of Results

Providing that all acceptability criteria are fulfilled, a test chemical is considered clearly positive if:

If only the highest dose is examined at a particular sampling time, a test chemical is considered clearly positive if there is a statistically significant increase compared with the concurrent negative control and the results are outside the distribution of the historical negative control data (e.g. Poisson-based 95 % control limits). Recommendations for the most appropriate statistical methods can be found in the literature (44) (45) (46) (47). When conducting a dose-response analysis, at least three treated dose groups should be analysed. Statistical tests should use the animal as the experimental unit. Positive results in the micronucleus test indicate that a test chemical induces micronuclei, which are the result of chromosomal damage or damage to the mitotic apparatus in the erythroblasts of the test species. In the case where a test was performed to detect centromeres within micronuclei, a test chemical that produces centromere-containing micronuclei (centromeric DNA or kinetochore, indicative of whole chromosome loss) is evidence that the test chemical is an aneugen.

Providing that all acceptability criteria are fulfilled, a test chemical is considered clearly negative if, in all experimental conditions examined:

Recommendations for the most appropriate statistical methods can be found in the literature (44) (45) (46) (47). Evidence of exposure of the bone marrow to a test chemical may include a depression of the immature to mature erythrocyte ratio or measurement of the plasma or blood levels of the test chemical. In case of intravenous administration, evidence of exposure is not needed. Alternatively, ADME data, obtained in an independent study using the same route and same species can be used to demonstrate bone marrow exposure. Negative results indicate that, under the test conditions, the test chemical does not produce micronuclei in the immature erythrocytes of the test species.

There is no requirement for verification of a clear positive or clear negative response.

In cases where the response is not clearly negative or positive and in order to assist in establishing the biological relevance of a result (e.g. a weak or borderline increase), the data should be evaluated by expert judgement and/or further investigations of the existing experiments completed. In some cases, analysing more cells or performing a repeat experiment using modified experimental conditions could be useful.

In rare cases, even after further investigations, the data will preclude making a conclusion that the test chemical produces either positive or negative results, and the study will therefore be concluded as equivocal.

Test Report

The test report should include the following information:

LITERATURE:

(6)

In Part B, Chapter B.15. is deleted.

(7)

In Part B, Chapter B.16. is deleted.

(8)

In Part B, Chapter B.18. is deleted.

(9)

In Part B, Chapter B.19. is deleted.

(10)

In Part B, Chapter B.20. is deleted.

(11)

In Part B, Chapter B.24. is deleted.

(12)

In Part B, Chapter B.47. is replaced by the following:

‘B.47   Bovine Corneal Opacity and Permeability Test Method for Identifying (i) Chemicals Inducing Serious Eye Damage and (ii) Chemicals Not Requiring Classification for Eye Irritation or Serious Eye Damage

INTRODUCTION

This test method is equivalent to OECD test guideline (TG) 437 (2013). The Bovine Corneal Opacity and Permeability (BCOP) test method was evaluated by the Interagency Coordinating Committee on the Validation of Alternative Methods (ICCVAM), in conjunction with the European Centre for the Validation of Alternative Methods (ECVAM) and the Japanese Center for the Validation of Alternative Methods (JaCVAM), in 2006 and 2010 (1)(2). In the first evaluation, the BCOP test method was evaluated for its usefulness to identify chemicals (substances and mixtures) inducing serious eye damage (1). In the second evaluation, the BCOP test method was evaluated for its usefulness to identify chemicals (substances and mixtures) not classified for eye irritation or serious eye damage (2). The BCOP validation database contained 113 substances and 100 mixtures in total (2)(3). From these evaluations and their peer review it was concluded that the test method can correctly identify chemicals (both substances and mixtures) inducing serious eye damage (Category 1) as well as those not requiring classification for eye irritation or serious eye damage, as defined by the United Nations (UN) Globally Harmonized System of Classification and Labelling of Chemicals (GHS) (4) and Regulation (EC) No 1272/2008 on Classification, Labelling and Packaging of Substances and Mixtures (CLP) (7) and it was therefore endorsed as scientifically valid for both purposes. Serious eye damage is the production of tissue damage in the eye, or serious physical decay of vision, following application of a test chemical to the anterior surface of the eye, which is not fully reversible within 21 days of application. Test chemicals inducing serious eye damage are classified as UN GHS Category 1. Chemicals not classified for eye irritation or serious eye damage are defined as those that do not meet the requirements for classification as UN GHS Category 1 or 2 (2A or 2B), i.e. they are referred to as UN GHS No Category. This test method includes the recommended use and limitations of the BCOP test method based on its evaluations. The main differences between the original 2009 version and the updated 2013 version of the OECD test guideline concern, but are not limited to: the use of the BCOP test method to identify chemicals not requiring classification according to UN GHS (paragraphs 2 and 7); clarifications on the applicability of the BCOP test method to the testing of alcohols, ketones and solids (paragraphs 6 and 7) and of substances and mixtures (paragraph 8); clarifications on how surfactant substances and surfactant-containing mixtures should be tested (paragraph 28); updates and clarifications regarding the positive controls (paragraphs 39 and 40); an update of the BCOP test method decision criteria (paragraph 47); an update of the study acceptance criteria (paragraph 48); an update to the test report elements (paragraph 49); an update of Appendix 1 on definitions; the addition of Appendix 2 for the predictive capacity of the BCOP test method under various classification systems; an update of Appendix 3 on the list of proficiency chemicals; and an update of Appendix 4 on the BCOP corneal holder (paragraph 1) and on the opacitometer (paragraphs 2 and 3).

It is currently generally accepted that, in the foreseeable future, no single in vitro eye irritation test will be able to replace the in vivo Draize eye test to predict across the full range of irritation for different chemical classes. However, strategic combinations of several alternative test methods within a (tiered) testing strategy may be able to replace the Draize eye test (5). The Top-Down approach (5) is designed to be used when, based on existing information, a chemical is expected to have high irritancy potential, while the Bottom-Up approach (5) is designed to be used when, based on existing information, a chemical is expected not to cause sufficient eye irritation to require a classification. The BCOP test method is an in vitro test method that can be used under certain circumstances and with specific limitations for eye hazard classification and labeling of chemicals. While it is not considered valid as a stand-alone replacement for the in vivo rabbit eye test, the BCOP test method is recommended as an initial step within a testing strategy such as the Top-Down approach suggested by Scott et al. (5) to identify chemicals inducing serious eye damage, i.e. chemicals to be classified as UN GHS Category 1, without further testing (4). The BCOP test method is also recommended to identify chemicals that do not require classification for eye irritation or serious eye damage, as defined by the UN GHS (UN GHS No Category) (4) within a testing strategy such as the Bottom-up approach (5). However, a chemical that is not predicted as causing serious eye damage or as not classified for eye irritation/serious eye damage with the BCOP test method would require additional testing (in vitro and/or in vivo) to establish a definitive classification.

The purpose of this test method is to describe the procedures used to evaluate the eye hazard potential of a test chemical as measured by its ability to induce opacity and increased permeability in an isolated bovine cornea. Toxic effects to the cornea are measured by: (i) decreased light transmission (opacity), and (ii) increased passage of sodium fluorescein dye (permeability). The opacity and permeability assessments of the cornea following exposure to a test chemical are combined to derive an In Vitro Irritancy Score (IVIS), which is used to classify the irritancy level of the test chemical.

Definitions are provided in Appendix 1.

INITIAL CONSIDERATIONS AND LIMITATIONS

This test method is based on the ICCVAM BCOP test method protocol (6)(7), which was originally developed from information obtained from the Institute for in vitro Sciences (IIVS) protocol and INVITTOX Protocol 124 (8). The latter represents the protocol used for the European Community-sponsored prevalidation study conducted in 1997-1998. Both of these protocols were based on the BCOP test method first reported by Gautheron et al. (9).

The BCOP test method can be used to identify chemicals inducing serious eye damage as defined by UN GHS, i.e. chemicals to be classified as UN GHS Category 1 (4). When used for this purpose, the BCOP test method has an overall accuracy of 79 % (150/191), a false positive rate of 25 % (32/126), and a false negative rate of 14 % (9/65), when compared to in vivo rabbit eye test method data classified according to the UN GHS classification system (3) (see Appendix 2, Table 1). When test chemicals within certain chemical (i.e., alcohols, ketones) or physical (i.e., solids) classes are excluded from the database, the BCOP test method has an overall accuracy of 85 % (111/131), a false positive rate of 20 % (16/81), and a false negative rate of 8 % (4/50) for the UN GHS classification system (3). The potential shortcomings of the BCOP test method when used to identify chemicals inducing serious eye damage (UN GHS Category 1) are based on the high false positive rates for alcohols and ketones and the high false negative rate for solids observed in the validation database (1)(2)(3). However, since not all alcohols and ketones are over-predicted by the BCOP test method and some are correctly predicted as UN GHS Category 1, these two organic functional groups are not considered to be out of the applicability domain of the test method. It is up to the user of this test method to decide if a possible over-prediction of an alcohol or ketone can be accepted or if further testing should be performed in a weight-of-evidence approach. Regarding the false negative rates for solids, it should be noted that solids may lead to variable and extreme exposure conditions in the in vivo Draize eye irritation test, which may result in irrelevant predictions of their true irritation potential (10). It should also be noted that none of the false negatives identified in the ICCVAM validation database (2)(3), in the context of identifying chemicals inducing serious eye damage (UN GHS Category 1), resulted in IVIS ≤ 3, which is the criterion used to identify a test chemical as a UN GHS No Category. Moreover, BCOP false negatives in this context are not critical since all test chemicals that produce an 3 < IVIS ≤ 55 would be subsequently tested with other adequately validated in vitro tests, or as a last option in rabbits, depending on regulatory requirements, using a sequential testing strategy in a weight-of-evidence approach. Given the fact that some solid chemicals are correctly predicted by the BCOP test method as UN GHS Category 1, this physical state is also not considered to be out of the applicability domain of the test method. Investigators could consider using this test method for all types of chemicals, whereby an IVIS > 55 should be accepted as indicative of a response inducing serious eye damage that should be classified as UN GHS Category 1 without further testing. However, as already mentioned, positive results obtained with alcohols or ketones should be interpreted cautiously due to potential over-prediction.

The BCOP test method can also be used to identify chemicals that do not require classification for eye irritation or serious eye damage under the UN GHS classification system (4). When used for this purpose, the BCOP test method has an overall accuracy of 69 % (135/196), a false positive rate of 69 % (61/89), and a false negative rate of 0 % (0/107), when compared to in vivo rabbit eye test method data classified according to the UN GHS classification system (3) (see Appendix 2, Table 2). The false positive rate obtained (in vivo UN GHS No Category chemicals producing an IVIS > 3, see paragraph 47) is considerably high, but not critical in this context since all test chemicals that produce an 3 < IVIS ≤ 55 would be subsequently tested with other adequately validated in vitro tests, or as a last option in rabbits, depending on regulatory requirements, using a sequential testing strategy in a weight-of-evidence approach. The BCOP test method shows no specific shortcomings for the testing of alcohols, ketones and solids when the purpose is to identify chemicals that do not require classification for eye irritation or serious eye damage (UN GHS No Category) (3). Investigators could consider using this test method for all types of chemicals, whereby a negative result (IVIS ≤ 3) should be accepted as indicative that no classification is required (UN GHS No Category). Since the BCOP test method can only identify correctly 31 % of the chemicals that do not require classification for eye irritation or serious eye damage, this test method should not be the first choice to initiate a Bottom-Up approach (5), if other validated and accepted in vitro methods with similar high sensitivity but higher specificity are available.

The BCOP validation database contained 113 substances and 100 mixtures in total (2)(3). The BCOP test method is therefore considered applicable to the testing of both substances and mixtures.

The BCOP test method is not recommended for the identification of test chemicals that should be classified as irritating to eyes (UN GHS Category 2 or Category 2A) or test chemicals that should be classified as mildly irritating to eyes (UN GHS Category 2B) due to the considerable number of UN GHS Category 1 chemicals underclassified as UN GHS Category 2, 2A or 2B and UN GHS No Category chemicals overclassifed as UN GHS Category 2, 2A or 2B (2)(3). For this purpose, further testing with another suitable method may be required.

All procedures with bovine eyes and bovine corneas should follow the testing facility's applicable regulations and procedures for handling animal-derived materials, which include, but are not limited to, tissues and tissue fluids. Universal laboratory precautions are recommended (11).

Whilst the BCOP test method does not consider conjunctival and iridal injuries, it addresses corneal effects, which are the major driver of classification in vivo when considering the UN GHS classification. The reversibility of corneal lesions cannot be evaluated per se in the BCOP test method. It has been proposed, based on rabbit eye studies, that an assessment of the initial depth of corneal injury may be used to identify some types of irreversible effects (12). However, further scientific knowledge is required to understand how irreversible effects not linked with initial high level injury occur. Finally, the BCOP test method does not allow for an assessment of the potential for systemic toxicity associated with ocular exposure.

This test method will be updated periodically as new information and data are considered. For example, histopathology may be potentially useful when a more complete characterisation of corneal damage is needed. As outlined in OECD Guidance Document No. 160 (13), users are encouraged to preserve corneas and prepare histopathology specimens that can be used to develop a database and decision criteria that may further improve the accuracy of this test method.

For any laboratory initially establishing this test method, the proficiency chemicals provided in Appendix 3 should be used. A laboratory can use these chemicals to demonstrate their technical competence in performing the BCOP test method prior to submitting BCOP test method data for regulatory hazard classification purposes.

PRINCIPLE OF THE TEST

The BCOP test method is an organotypic model that provides short-term maintenance of normal physiological and biochemical function of the bovine cornea in vitro. In this test method, damage by the test chemical is assessed by quantitative measurements of changes in corneal opacity and permeability with an opacitometer and a visible light spectrophotometer, respectively. Both measurements are used to calculate an IVIS, which is used to assign an in vitro irritancy hazard classification category for prediction of the in vivo ocular irritation potential of a test chemical (see Decision Criteria in paragraph 48).

The BCOP test method uses isolated corneas from the eyes of freshly slaughtered cattle. Corneal opacity is measured quantitatively as the amount of light transmission through the cornea. Permeability is measured quantitatively as the amount of sodium fluorescein dye that passes across the full thickness of the cornea, as detected in the medium in the posterior chamber. Test chemicals are applied to the epithelial surface of the cornea by addition to the anterior chamber of the corneal holder. Appendix 4 provides a description and a diagram of a corneal holder used in the BCOP test method. Corneal holders can be obtained commercially from different sources or can be constructed.

Source and Age of Bovine Eyes and Selection of Animal Species

Cattle sent to slaughterhouses are typically killed either for human consumption or for other commercial uses. Only healthy animals considered suitable for entry into the human food chain are used as a source of corneas for use in the BCOP test method. Because cattle have a wide range of weights, depending on breed, age, and sex, there is no recommended weight for the animal at the time of slaughter.

Variations in corneal dimensions can result when using eyes from animals of different ages. Corneas with a horizontal diameter > 30,5 mm and central corneal thickness (CCT) values ≥ 1 100 μm are generally obtained from cattle older than eight years, while those with a horizontal diameter < 28,5 mm and CCT < 900 μm are generally obtained from cattle less than five years old (14). For this reason, eyes from cattle greater than 60 months old are not typically used. Eyes from cattle less than 12 months of age have not traditionally been used since the eyes are still developing and the corneal thickness and corneal diameter are considerably smaller than that reported for eyes from adult cattle. However, the use of corneas from young animals (i.e., 6 to 12 months old) is permissible since there are some advantages, such as increased availability, a narrow age range, and decreased hazards related to potential worker exposure to Bovine Spongiform Encephalopathy (15). As further evaluation of the effect of corneal size or thickness on responsiveness to corrosive and irritant chemicals would be useful, users are encouraged to report the estimated age and/or weight of the animals providing the corneas used in a study.

Collection and Transport of Eyes to the Laboratory

Eyes are collected by slaughterhouse employees. To minimise mechanical and other types of damage to the eyes, the eyes should be enucleated as soon as possible after death and cooled immediately after enucleation and during transport. To prevent exposure of the eyes to potentially irritant chemicals, the slaughterhouse employees should not use detergent when rinsing the head of the animal.

Eyes should be immersed completely in cooled Hanks' Balanced Salt Solution (HBSS) in a suitably sized container, and transported to the laboratory in such a manner as to minimise deterioration and/or bacterial contamination. Because the eyes are collected during the slaughter process, they might be exposed to blood and other biological materials, including bacteria and other microorganisms. Therefore, it is important to ensure that the risk of contamination is minimised (e.g., by keeping the container containing the eyes on wet ice during collection and transportation and by adding antibiotics to the HBSS used to store the eyes during transport [e.g. penicillin at 100 IU/ml and streptomycin at 100 μg/ml]).

The time interval between collection of the eyes and use of corneas in the BCOP test method should be minimised (typically collected and used on the same day) and should be demonstrated to not compromise the assay results. These results are based on the selection criteria for the eyes, as well as the positive and negative control responses. All eyes used in the assay should be from the same group of eyes collected on a specific day.

Selection Criteria for Eyes Used in the BCOP Test Method

The eyes, once they arrive at the laboratory, are carefully examined for defects including increased opacity, scratches, and neovascularisation. Only corneas from eyes free of such defects are to be used.

The quality of each cornea is also evaluated at later steps in the assay. Corneas that have opacity greater than seven opacity units or equivalent for the opacitometer and cornea holders used after an initial one hour equilibration period are to be discarded (NOTE: the opacitometer should be calibrated with opacity standards that are used to establish the opacity units, see Appendix 4).

Each treatment group (test chemical, concurrent negative and positive controls) consists of a minimum of three eyes. Three corneas should be used for the negative control corneas in the BCOP test method. Since all corneas are excised from the whole globe, and mounted in the corneal chambers, there is potential for artifacts from handling upon individual corneal opacity and permeability values (including negative control). Furthermore, the opacity and permeability values from the negative control corneas are used to correct the test chemical-treated and positive control-treated corneal opacity and permeability values in the IVIS calculations.

PROCEDURE

Preparation of the Eyes

Corneas, free of defects, are dissected with a 2 to 3 mm rim of sclera remaining to assist in subsequent handling, with care taken to avoid damage to the corneal epithelium and endothelium. Isolated corneas are mounted in specially designed corneal holders that consist of anterior and posterior compartments, which interface with the epithelial and endothelial sides of the cornea, respectively. Both chambers are filled to excess with pre-warmed phenol red free Eagle's Minimum Essential Medium (EMEM) (posterior chamber first), ensuring that no bubbles are formed. The device is then equilibrated at 32 ± 1 °C for at least one hour to allow the corneas to equilibrate with the medium and to achieve normal metabolic activity, to the extent possible (the approximate temperature of the corneal surface in vivo is 32 °C).

Following the equilibration period, fresh pre-warmed phenol red free EMEM is added to both chambers and baseline opacity readings are taken for each cornea. Any corneas that show macroscopic tissue damage (e.g, scratches, pigmentation, neovascularisation) or an opacity greater than seven opacity units or equivalent for the opacitometer and cornea holders used are discarded. A minimum of three corneas are selected as negative (or solvent) control corneas. The remaining corneas are then distributed into treatment and positive control groups.

Because the heat capacity of water is higher than that of air, water provides more stable temperature conditions for incubation. Therefore, the use a water bath for maintaining the corneal holder and its contents at 32 ± 1 °C is recommended. However, air incubators might also be used, assuming precaution to maintain temperature stability (e.g. by pre-warming of holders and media).

Application of the Test Chemical

Two different treatment protocols are used, one for liquids and surfactants (solids or liquids), and one for non-surfactant solids.

Liquids are tested undiluted. Semi-solids, creams, and waxes are typically tested as liquids. Neat surfactant substances are tested at a concentration of 10 % w/v in a 0,9 % sodium chloride solution, distilled water, or other solvent that has been demonstrated to have no adverse effects on the test system. Appropriate justification should be provided for alternative dilution concentrations. Mixtures containing surfactants may be tested undiluted or diluted to an appropriate concentration depending on the relevant exposure scenario in vivo. Appropriate justification should be provided for the concentration tested. Corneas are exposed to liquids and surfactants for 10 minutes. Use of other exposure times should be accompanied by adequate scientific rationale. Please see Appendix 1 for a definition of surfactant and surfactant-containing mixture.

Non-surfactant solids are typically tested as solutions or suspensions at 20 % w/v concentration in a 0,9 % sodium chloride solution, distilled water, or other solvent that has been demonstrated to have no adverse effects on the test system. In certain circumstances and with proper scientific justification, solids may also be tested neat by direct application onto the corneal surface using the open chamber method (see paragraph 32). Corneas are exposed to solids for four hours, but as with liquids and surfactants, alternative exposure times may be used with appropriate scientific rationale.

Different treatment methods can be used, depending on the physical nature and chemical characteristics (e.g. solids, liquids, viscous vs. non-viscous liquids) of the test chemical. The critical factor is ensuring that the test chemical adequately covers the epithelial surface and that it is adequately removed during the rinsing steps. A closed-chamber method is typically used for non-viscous to slightly viscous liquid test chemicals, while an open-chamber method is typically used for semi-viscous and viscous liquid test chemicals and for neat solids.

In the closed-chamber method, sufficient test chemical (750 μl) to cover the epithelial side of the cornea is introduced into the anterior chamber through the dosing holes on the top surface of the chamber, and the holes are subsequently sealed with the chamber plugs during the exposure. It is important to ensure that each cornea is exposed to a test chemical for the appropriate time interval.

In the open-chamber method, the window-locking ring and glass window from the anterior chamber are removed prior to treatment. The control or test chemical (750 μl, or enough test chemical to completely cover the cornea) is applied directly to the epithelial surface of the cornea using a micro-pipet. If a test chemical is difficult to pipet, the test chemical can be pressure-loaded into a positive displacement pipet to aid in dosing. The pipet tip of the positive displacement pipet is inserted into the dispensing tip of the syringe so that the material can be loaded into the displacement tip under pressure. Simultaneously, the syringe plunger is depressed as the pipet piston is drawn upwards. If air bubbles appear in the pipet tip, the test chemical is removed (expelled) and the process repeated until the tip is filled without air bubbles. If necessary, a normal syringe (without a needle) can be used since it permits measuring an accurate volume of test chemical and an easier application to the epithelial surface of the cornea. After dosing, the glass window is replaced on the anterior chamber to recreate a closed system.

Post-Exposure Incubation

After the exposure period, the test chemical, the negative control, or the positive control chemical is removed from the anterior chamber and the epithelium washed at least three times (or until no visual evidence of test chemical can be observed) with EMEM (containing phenol red). Phenol red- containing medium is used for rinsing since a colour change in the phenol red may be monitored to determine the effectiveness of rinsing acidic or alkaline test chemicals. The corneas are washed more than three times if the phenol red is still discoloured (yellow or purple), or the test chemical is still visible. Once the medium is free of test chemical, the corneas are given a final rinse with EMEM (without phenol red). The EMEM (without phenol red) is used as a final rinse to ensure removal of the phenol red from the anterior chamber prior to the opacity measurement. The anterior chamber is then refilled with fresh EMEM without phenol red.

For liquids or surfactants, after rinsing, the corneas are incubated for an additional two hours at 32 ± 1 °C. Longer post-exposure time may be useful in certain circumstances and could be considered on a case-by-case basis. Corneas treated with solids are rinsed thoroughly at the end of the four-hour exposure period, but do not require further incubation.

At the end of the post-exposure incubation period for liquids and surfactants and at the end of the four-hour exposure period for non-surfactant solids, the opacity and permeability of each cornea are recorded. Also, each cornea is observed visually and pertinent observations recorded (e.g., tissue peeling, residual test chemical, non-uniform opacity patterns). These observations could be important as they may be reflected by variations in the opacitometer readings.

Control Chemicals

Concurrent negative or solvent/vehicle controls and positive controls are included in each experiment.

When testing a liquid substance at 100 %, a concurrent negative control (e.g. 0,9 % sodium chloride solution or distilled water) is included in the BCOP test method so that nonspecific changes in the test system can be detected and to provide a baseline for the assay endpoints. It also ensures that the assay conditions do not inappropriately result in an irritant response.

When testing a diluted liquid, surfactant, or solid, a concurrent solvent/vehicle control group is included in the BCOP test method so that nonspecific changes in the test system can be detected and to provide a baseline for the assay endpoints. Only a solvent/vehicle that has been demonstrated to have no adverse effects on the test system can be used.

A chemical known to induce a positive response is included as a concurrent positive control in each experiment to verify the integrity of the test system and its correct conduct. However, to ensure that variability in the positive control response across time can be assessed, the magnitude of irritant response should not be excessive.

Examples of positive controls for liquid test chemicals are 100 % ethanol or 100 % dimethylformamide. An example of a positive control for solid test chemicals is 20 % w/v imidazole in 0,9 % sodium chloride solution.

Benchmark chemicals are useful for evaluating the ocular irritancy potential of unknown chemicals of a specific chemical or product class, or for evaluating the relative irritancy potential of an ocular irritant within a specific range of irritant responses.

Endpoints Measured

Opacity is determined by the amount of light transmission through the cornea. Corneal opacity is measured quantitatively with the aid of an opacitometer, resulting in opacity values measured on a continuous scale.

Permeability is determined by the amount of sodium fluorescein dye that penetrates all corneal cell layers (i.e., the epithelium on the outer cornea surface through the endothelium on the inner cornea surface). One ml sodium fluorescein solution (4 or 5 mg/ml when testing liquids and surfactants or non- surfactant solids, respectively) is added to the anterior chamber of the corneal holder, which interfaces with the epithelial side of the cornea, while the posterior chamber, which interfaces with the endothelial side of the cornea, is filled with fresh EMEM. The holder is then incubated in a horizontal position for 90 ± 5 min at 32 ± 1 °C. The amount of sodium fluorescein that crosses into the posterior chamber is quantitatively measured with the aid of UV/VIS spectrophotometry. Spectrophotometric measurements evaluated at 490 nm are recorded as optical density (OD490) or absorbance values, which are measured on a continuous scale. The fluorescein permeability values are determined using OD490 values based upon a visible light spectrophotometer using a standard 1 cm path length.

Alternatively, a 96-well microtiter plate reader may be used provided that; (i) the linear range of the plate reader for determining fluorescein OD490 values can be established; and (ii), the correct volume of fluorescein samples are used in the 96-well plate to result in OD490 values equivalent to the standard 1 cm path length (this could require a completely full well [usually 360 μl]).

DATA AND REPORTING

Data Evaluation

Once the opacity and mean permeability (OD490) values have been corrected for background opacity and the negative control permeability OD490 values, the mean opacity and permeability OD490 values for each treatment group should be combined in an empirically-derived formula to calculate an in vitro irritancy score (IVIS) for each treatment group as follows:

IVIS = mean opacity value + (15 × mean permeability OD490 value)

Sina et al. (16) reported that this formula was derived during in-house and inter-laboratory studies. The data generated for a series of 36 compounds in a multi-laboratory study were subjected to a multivariate analysis to determine the equation of best fit between in vivo and in vitro data. Scientists at two separate companies performed this analysis and derived nearly identical equations.

The opacity and permeability values should also be evaluated independently to determine whether a test chemical induced corrosivity or severe irritation through only one of the two endpoints (see Decision Criteria).

Decision Criteria

The IVIS cut-off values for identifying test chemicals as inducing serious eye damage (UN GHS Category 1) and test chemicals not requiring classification for eye irritation or serious eye damage (UN GHS No Category) are given hereafter:

Study Acceptance Criteria

A test is considered acceptable if the positive control gives an IVIS that falls within two standard deviations of the current historical mean, which is to be updated at least every three months, or each time an acceptable test is conducted in laboratories where tests are conducted infrequently (i.e., less than once a month). The negative or solvent/vehicle control responses should result in opacity and permeability values that are less than the established upper limits for background opacity and permeability values for bovine corneas treated with the respective negative or solvent/vehicle control. A single testing run composed of at least three corneas should be sufficient for a test chemical when the resulting classification is unequivocal. However, in cases of borderline results in the first testing run, a second testing run should be considered (but not necessarily required), as well as a third one in case of discordant mean IVIS results between the first two testing runs. In this context, a result in the first testing run is considered borderline if the predictions from the 3 corneas were non-concordant, such that:

Test Report

The test report should include the following information, if relevant to the conduct of the study:

LITERATURE:

Appendix 4

THE BCOP CORNEAL HOLDER

The BCOP corneal holders are made of an inert material (e.g. polypropylene). The holders are comprised of two halves (an anterior and posterior chamber), and have two similar cylindrical internal chambers. Each chamber is designed to hold a volume of about 5 ml and terminates in a glass window, through which opacity measurements are recorded. Each of the inner chambers is 1,7 cm in diameter and 2,2 cm in depth (11). An o-ring located on the posterior chamber is used to prevent leaks. The corneas are placed endothelial side down on the o-ring of the posterior chambers and the anterior chambers are placed on the epithelial side of the corneas. The chambers are maintained in place by three stainless steel screws located on the outer edges of the chamber. The end of each chamber houses a glass window, which can be removed for easy access to the cornea. An o-ring is also located between the glass window and the chamber to prevent leaks. Two holes on the top of each chamber permit introduction and removal of medium and test chemicals. They are closed with rubber caps during the treatment and incubation periods. The light transmission through corneal holders can potentially change as the effects of wear and tear or accumulation of specific chemical residues on the internal chamber bores or on the glass windows may affect light scatter or reflectance. The consequence could be increases or decreases in baseline light transmission (and conversely the baseline opacity readings) through the corneal holders, and may be evident as notable changes in the expected baseline initial corneal opacity measurements in individual chambers (i.e., the initial corneal opacity values in specific individual corneal holders may routinely differ by more than 2 or 3 opacity units from the expected baseline values). Each laboratory should consider establishing a program for evaluating for changes in the light transmission through the corneal holders, depending upon the nature of the chemistries tested and the frequency of use of the chambers. To establish baseline values, corneal holders may be checked before routine use by measuring the baseline opacity values (or light transmission) of chambers filled with complete medium, without corneas. The corneal holders are then periodically checked for changes in light transmission during periods of use. Each laboratory can establish the frequency for checking the corneal holders, based upon the chemicals tested, the frequency of use, and observations of changes in the baseline corneal opacity values. If notable changes in the light transmission through the corneal holders are observed, appropriate cleaning and/or polishing procedures of the interior surface of the cornea holders or replacement have to be considered.

Corneal holder: exploded diagramme

(13)

In Part B, Chapter B.48 is replaced by the following:

‘B.48   Isolated Chicken Eye Test Method for Identifying i) Chemicals Inducing Serious Eye Damage and ii) Chemicals Not Requiring Classification for Eye Irritation or Serious Eye Damage

INTRODUCTION

This test method is equivalent to OECD test guideline (TG) 438 (2013). The Isolated Chicken Eye (ICE) test method was evaluated by the Interagency Coordinating Committee on the Validation of Alternative Methods (ICCVAM), in conjunction with the European Centre for the Validation of Alternative Methods (ECVAM) and the Japanese Centre for the Validation of Alternative Methods (JaCVAM), in 2006 and 2010 (1) (2) (3). In the first evaluation, the ICE was endorsed as a scientifically valid test method for use as a screening test to identify chemicals (substances and mixtures) inducing serious eye damage (Category 1) as defined by the United Nations (UN) Globally Harmonized System of Classification and Labelling of Chemicals (GHS) (1) (2) (4) and Regulation (EC) No 1272/2008 on Classification, Labelling and Packaging of Substances and Mixtures (CLP) (12). In the second evaluation, the ICE test method was evaluated for use as a screening test to identify chemicals not classified for eye irritation or serious eye damage as defined by UN GHS (3) (4). The results from the validation study and the peer review panel recommendations maintained the original recommendation for using the ICE for classification of chemicals inducing serious eye damage (UN GHS Category 1), as the available database remained unchanged since the original ICCVAM validation. At that stage, no further recommendations for an expansion of the ICE applicability domain to also include other categories were suggested. A re-evaluation of the in vitro and in vivo dataset used in the validation study was made with the focus of evaluating the usefulness of the ICE to identify chemicals not requiring classification for eye irritation or serious eye damage (5). This re-evaluation concluded that the ICE test method can also be used to identify chemicals not requiring classification for eye irritation and serious eye damage as defined by the UN GHS (4) (5). This test method includes the recommended uses and limitations of the ICE test method based on these evaluations. The main differences between the original 2009 version and the updated 2013 version of the OECD test guideline include, but are not limited to, the use of the ICE test method to identify chemicals not requiring classification according to the UN GHS Classification System, an update to the test report elements, an update of Appendix 1 on definitions, and an update to Appendix 2 on the proficiency chemicals.

It is currently generally accepted that, in the foreseeable future, no single in vitro eye irritation test will be able to replace the in vivo Draize eye test to predict across the full range of irritation for different chemical classes. However, strategic combinations of several alternative test methods within a (tiered) testing strategy may be able to replace the Draize eye test (6). The Top-Down approach (7) is designed to be used when, based on existing information, a chemical is expected to have high irritancy potential, while the Bottom-Up approach (7) is designed to be used when, based on existing information, a chemical is expected not to cause sufficient eye irritation to require a classification. The ICE test method is an in vitro test method that can be used, under certain circumstances and with specific limitations as described in paragraphs 8 to 10 for eye hazard classification and labelling of chemicals. While it is not considered valid as a stand-alone replacement for the in vivo rabbit eye test, the ICE test method is recommended as an initial step within a testing strategy such as the Top-Down approach suggested by Scott et al. (7) to identify chemicals inducing serious eye damage, i.e., chemicals to be classified as UN GHS Category 1 without further testing (4). The ICE test method is also recommended to identify chemicals that do not require classification for eye irritation or serious eye damage as defined by the UN GHS (No Category, NC) (4), and may therefore be used as an initial step within a Bottom-Up testing strategy approach (7). However, a chemical that is not predicted as causing serious eye damage or as not classified for eye irritation/serious eye damage with the ICE test method would require additional testing (in vitro and/or in vivo) to establish a definitive classification. Furthermore, the appropriate regulatory authorities should be consulted before using the ICE in a bottom up approach under other classification schemes than the UN GHS.

The purpose of this test method is to describe the procedures used to evaluate the eye hazard potential of a test chemical as measured by its ability to induce or not toxicity in an enucleated chicken eye. Toxic effects to the cornea are measured by (i) a qualitative assessment of opacity, (ii) a qualitative assessment of damage to epithelium based on application of fluorescein to the eye (fluorescein retention), (iii) a quantitative measurement of increased thickness (swelling), and (iv) a qualitative evaluation of macroscopic morphological damage to the surface. The corneal opacity, swelling, and damage assessments following exposure to a test chemical are assessed individually and then combined to derive an Eye Irritancy Classification.

Definitions are provided in Appendix 1.

INITIAL CONSIDERATIONS AND LIMITATIONS

This test method is based on the protocol suggested in the OECD Guidance Document 160 (8), which was developed following the ICCVAM international validation study (1) (3) (9), with contributions from the European Centre for the Validation of Alternative Methods, the Japanese Center for the Validation of Alternative Methods, and TNO Quality of Life Department of Toxicology and Applied Pharmacology (Netherlands). The protocol is based on information obtained from published protocols, as well as the current protocol used by TNO (10) (11) (12) (13) (14).

A wide range of chemicals has been tested in the validation underlying this test method and the empirical database of the validation study amounted to 152 chemicals including 72 substances and 80 mixtures (5). The test method is applicable to solids, liquids, emulsions and gels. The liquids may be aqueous or non-aqueous; solids may be soluble or insoluble in water. Gases and aerosols have not been assessed yet in a validation study.

The ICE test method can be used to identify chemicals inducing serious eye damage, i.e., chemicals to be classified as UN GHS Category 1 (4). When used for this purpose, the identified limitations for the ICE test method are based on the high false positive rates for alcohols and the high false negative rates for solids and surfactants (1) (3) (9). However, false negative rates in this context (UN GHS Category 1 identified as not being UN GHS Category 1) are not critical since all test chemicals that come out negative would be subsequently tested with other adequately validated in vitro test(s), or as a last option in rabbits, depending on regulatory requirements, using a sequential testing strategy in a weight-of-evidence approach. It should be noted that solids may lead to variable and extreme exposure conditions in the in vivo Draize eye irritation test, which may result in irrelevant predictions of their true irritation potential (15). Investigators could consider using this test method for all types of chemicals, whereby a positive result should be accepted as indicative of serious eye damage, i.e., UN GHS Category 1 classification without further testing. However, positive results obtained with alcohols should be interpreted cautiously due to risk of over-prediction.

When used to identify chemicals inducing serious eye damage (UN GHS Category 1), the ICE test method has an overall accuracy of 86 % (120/140), a false positive rate of 6 % (7/113) and a false negative rate of 48 % (13/27) when compared to in vivo rabbit eye test method data classified according to the UN GHS classification system (4) (5).

The ICE test method can also be used to identify chemicals that do not require classification for eye irritation or serious eye damage under the UN GHS classification system (4). The appropriate regulatory authorities should be consulted before using the ICE in a bottom up approach under other classification schemes. This test method can be used for all types of chemicals, whereby a negative result could be accepted for not classifying a chemical for eye irritation and serious eye damage. However, on the basis of one result from the validation database, anti-fouling organic solvent-containing paints may be under-predicted (5).

When used to identify chemicals that do not require classification for eye irritation and serious eye damage, the ICE test method has an overall accuracy of 82 % (125/152), a false positive rate of 33 % (26/79), and a false negative rate of 1 % (1/73), when compared to in vivo rabbit eye test method data classified according to the UN GHS (4) (5). When test chemicals within certain classes (i.e., anti-fouling organic solvent containing paints) are excluded from the database, the accuracy of the ICE test method is 83 % (123/149), the false positive rate 33 % (26/78), and the false negative rate of 0 % (0/71) for the UN GHS classification system (4) (5).

The ICE test method is not recommended for the identification of test chemicals that should be classified as irritating to eyes (i.e., UN GHS Category 2 or Category 2A) or test chemicals that should be classified as mildly irritating to eyes (UN GHS Category 2B) due to the considerable number of UN GHS Category 1 chemicals underclassified as UN GHS Category 2, 2A or 2B and UN GHS No Category chemicals overclassifed as UN GHS Category 2, 2A or 2B. For this purpose, further testing with another suitable method may be required.

All procedures with chicken eyes should follow the test facility's applicable regulations and procedures for handling of human or animal-derived materials, which include, but are not limited to, tissues and tissue fluids. Universal laboratory precautions are recommended (16).

Whilst the ICE test method does not consider conjunctival and iridal injuries as evaluated in the rabbit ocular irritancy test method, it addresses corneal effects which are the major driver of classification in vivo when considering the UN GHS Classification. Also, although the reversibility of corneal lesions cannot be evaluated per se in the ICE test method, it has been proposed, based on rabbit eye studies, that an assessment of the initial depth of corneal injury may be used to identify some types of irreversible effects (17). In particular, further scientific knowledge is required to understand how irreversible effects not linked with initial high level injury occur. Finally, the ICE test method does not allow for an assessment of the potential for systemic toxicity associated with ocular exposure.

This test method will be updated periodically as new information and data are considered. For example, histopathology may be potentially useful when a more complete characterisation of corneal damage is needed. To evaluate this possibility, users are encouraged to preserve eyes and prepare histopathology specimens that can be used to develop a database and decision criteria that may further improve the accuracy of this test method. The OECD has developed a Guidance Document on the use of in vitro ocular toxicity test methods, which includes detailed procedures on the collection of histopathology specimens and information on where to submit specimens and/or histopathology data (8).

For any laboratory initially establishing this assay, the proficiency chemicals provided in Appendix 2 should be used. A laboratory can use these chemicals to demonstrate their technical competence in performing the ICE test method prior to submitting ICE data for regulatory hazard classification purposes.

PRINCIPLE OF THE TEST

The ICE test method is an organotypic model that provides short-term maintenance of the chicken eye in vitro. In this test method, damage by the test chemical is assessed by determination of corneal swelling, opacity, and fluorescein retention. While the latter two parameters involve a qualitative assessment, analysis of corneal swelling provides for a quantitative assessment. Each measurement is either converted into a quantitative score used to calculate an overall Irritation Index, or assigned a qualitative categorisation that is used to assign an in vitro ocular hazard classification, either as UN GHS Category 1 or as UN GHS non-classified. Either of these outcomes can then be used to predict the potential in vivo serious eye damage or no requirement for eye hazard classification of a test chemical (see Decision Criteria). However, no classification can be given for chemicals not predicted as causing serious eye damage or as not classified with the ICE test method (see paragraph 11).

Source and Age of Chicken Eyes

Historically, eyes collected from chickens obtained from a slaughterhouse where they are killed for human consumption have been used for this assay, eliminating the need for laboratory animals. Only the eyes of healthy animals considered suitable for entry into the human food chain are used.

Although a controlled study to evaluate the optimum chicken age has not been conducted, the age and weight of the chickens used historically in this test method are that of spring chickens traditionally processed by a poultry slaughterhouse (i.e., approximately 7 weeks old, 1,5 - 2,5 kg).

Collection and Transport of Eyes to the Laboratory

Heads should be removed immediately after sedation of the chickens, usually by electric shock, and incision of the neck for bleeding. A local source of chickens close to the laboratory should be located so that their heads can be transferred from the slaughterhouse to the laboratory quickly enough to minimise deterioration and/or bacterial contamination. The time interval between collection of the chicken heads and placing the eyes in the superfusion chamber following enucleation should be minimised (typically within two hours) to assure meeting assay acceptance criteria. All eyes used in the assay should be from the same group of eyes collected on a specific day.

Because eyes are dissected in the laboratory, the intact heads are transported from the slaughterhouse at ambient temperature (typically between 18 °C and 25 °C) in plastic boxes humidified with tissues moistened with isotonic saline.

Selection Criteria and Number of Eyes Used in the ICE

Eyes that have high baseline fluorescein staining (i.e., > 0,5) or corneal opacity score (i.e., > 0,5) after they are enucleated are rejected.

Each treatment group and concurrent positive control consists of at least three eyes. The negative control group or the solvent control (if using a solvent other than saline) consists of at least one eye.

In the case of solid materials leading to a GHS NC outcome, a second run of three eyes is recommended to confirm or discard the negative outcome.

PROCEDURE

Preparation of the Eyes

The eyelids are carefully excised, taking care not to damage the cornea. Corneal integrity is quickly assessed with a drop of 2 % (w/v) sodium fluorescein applied to the corneal surface for a few seconds, and then rinsed with isotonic saline. Fluorescein-treated eyes are then examined with a slit-lamp microscope to ensure that the cornea is undamaged (i.e., fluorescein retention and corneal opacity scores ≤ 0,5).

If undamaged, the eye is further dissected from the skull, taking care not to damage the cornea. The eyeball is pulled from the orbit by holding the nictitating membrane firmly with surgical forceps, and the eye muscles are cut with a bent, blunt-tipped scissor. It is important to avoid causing corneal damage due to excessive pressure (i.e., compression artifacts).

When the eye is removed from the orbit, a visible portion of the optic nerve should be left attached. Once removed from the orbit, the eye is placed on an absorbent pad and the nictitating membrane and other connective tissue are cut away.

The enucleated eye is mounted in a stainless steel clamp with the cornea positioned vertically. The clamp is then transferred to a chamber of the superfusion apparatus (18). The clamps should be positioned in the superfusion apparatus such that the entire cornea is supplied with the isotonic saline drip (3-4 drops per minute or 0,1 to 0,15 ml/min). The chambers of the superfusion apparatus should be temperature controlled at 32 ± 1,5 °C. Appendix 3 provides a diagram of a typical superfusion apparatus and the eye clamps, which can be obtained commercially or constructed. The apparatus can be modified to meet the needs of an individual laboratory (e.g. to accommodate a different number of eyes).

After being placed in the superfusion apparatus, the eyes are again examined with a slit-lamp microscope to ensure that they have not been damaged during the dissection procedure. Corneal thickness should also be measured at this time at the corneal apex using the depth measuring device on the slit-lamp microscope. Eyes with; (i), a fluorescein retention score of > 0,5; (ii) corneal opacity > 0,5; or, (iii), any additional signs of damage should be replaced. For eyes that are not rejected based on any of these criteria, individual eyes with a corneal thickness deviating more than 10 % from the mean value for all eyes are to be rejected. Users should be aware that slit-lamp microscopes could yield different corneal thickness measurements if the slit-width setting is different. The slit-width should be set at 0,095 mm.

Once all eyes have been examined and approved, the eyes are incubated for approximately 45 to 60 minutes to equilibrate them to the test system prior to dosing. Following the equilibration period, a zero reference measurement is recorded for corneal thickness and opacity to serve as a baseline (i.e., time = 0). The fluorescein score determined at dissection is used as the baseline measurement for that endpoint.

Application of the Test Chemical

Immediately following the zero reference measurements, the eye (in its holder) is removed from the superfusion apparatus, placed in a horizontal position, and the test chemical is applied to the cornea.

Liquid test chemicals are typically tested undiluted, but may be diluted if deemed necessary (e.g. as part of the study design). The preferred solvent for diluted test chemicals is physiological saline. However, alternative solvents may also be used under controlled conditions, but the appropriateness of solvents other than physiological saline should be demonstrated.

Liquid test chemicals are applied to the cornea such that the entire surface of the cornea is evenly covered with the test chemical; the standard volume is 0,03 ml.

If possible, solid test chemicals should be ground as finely as possible in a mortar and pestle, or comparable grinding tool. The powder is applied to the cornea such that the surface is uniformly covered with the test chemical; the standard amount is 0,03 g.

The test chemical (liquid or solid) is applied for 10 seconds and then rinsed from the eye with isotonic saline (approximately 20 ml) at ambient temperature. The eye (in its holder) is subsequently returned to the superfusion apparatus in the original upright position. In case of need, additional rinsing may be used after the 10-sec application and at subsequent time points (e.g. upon discovery of residues of test chemical on the cornea). In general the amount of saline additionally used for rinsing is not critical, but the observation of adherence of chemical to the cornea is important.

Control Chemicals

Concurrent negative or solvent/vehicle controls and positive controls should be included in each experiment.

When testing liquids at 100 % or solids, physiological saline is used as the concurrent negative control in the ICE test method to detect non-specific changes in the test system, and to ensure that the assay conditions do not inappropriately result in an irritant response.

When testing diluted liquids, a concurrent solvent/vehicle control group is included in the test method to detect non-specific changes in the test system, and to ensure that the assay conditions do not inappropriately result in an irritant response. As stated in paragraph 31, only a solvent/vehicle that has been demonstrated to have no adverse effects on the test system can be used.

A known ocular irritant is included as a concurrent positive control in each experiment to verify that an appropriate response is induced. As the ICE assay is being used in this test method to identify corrosive or severe irritants, the positive control should be a reference chemical that induces a severe response in this test method. However, to ensure that variability in the positive control response across time can be assessed, the magnitude of the severe response should not be excessive. Sufficient in vitro data for the positive control should be generated such that a statistically defined acceptable range for the positive control can be calculated. If adequate historical ICE test method data are not available for a particular positive control, studies may need to be conducted to provide this information.

Examples of positive controls for liquid test chemicals are 10 % acetic acid or 5 % benzalkonium chloride, while examples of positive controls for solid test chemicals are sodium hydroxide or imidazole.

Benchmark chemicals are useful for evaluating the ocular irritancy potential of unknown chemicals of a specific chemical or product class, or for evaluating the relative irritancy potential of an ocular irritant within a specific range of irritant responses.

Endpoints Measured

Treated corneas are evaluated prior to treatment and at 30, 75, 120, 180, and 240 minutes (± 5 minutes) after the post-treatment rinse. These time points provide an adequate number of measurements over the four-hour treatment period, while leaving sufficient time between measurements for the requisite observations to be made for all eyes.

The endpoints evaluated are corneal opacity, swelling, fluorescein retention, and morphological effects (e.g. pitting or loosening of the epithelium). All of the endpoints, with the exception of fluorescein retention (which is determined only prior to treatment and 30 minutes after test chemical exposure) are determined at each of the above time points.

Photographs are advisable to document corneal opacity, fluorescein retention, morphological effects and, if conducted, histopathology.

After the final examination at four hours, users are encouraged to preserve eyes in an appropriate fixative (e.g. neutral buffered formalin) for possible histopathological examination (see paragraph 14 and reference (8) for details).

Corneal swelling is determined from corneal thickness measurements made with an optical pachymeter on a slit-lamp microscope. It is expressed as a percentage and is calculated from corneal thickness measurements according to the following formula:

The mean percentage of corneal swelling for all test eyes is calculated for all observation time points. Based on the highest mean score for corneal swelling, as observed at any time point, an overall category score is then given for each test chemical (see paragraph 51).

Corneal opacity is evaluated by using the area of the cornea that is most densely opacified for scoring as shown in Table 1. The mean corneal opacity value for all test eyes is calculated for all observation time points. Based on the highest mean score for corneal opacity, as observed at any time point, an overall category score is then given for each test chemical (see paragraph 51).

Table 1

Corneal opacity scores

Fluorescein retention is evaluated at the 30 minute observation time point only as shown in Table 2. The mean fluorescein retention value of all test eyes is then calculated for the 30-minute observation time point, and used for the overall category score given for each test chemical (see paragraph 51).

Table 2

Fluorescein retention scores

Morphological effects include “pitting” of corneal epithelial cells, “loosening” of epithelium, “roughening” of the corneal surface and “sticking” of the test chemical to the cornea. These findings can vary in severity and may occur simultaneously. The classification of these findings is subjective according to the interpretation of the investigator.

DATA AND REPORTING

Data Evaluation

Results from corneal opacity, swelling, and fluorescein retention should be evaluated separately to generate an ICE class for each endpoint. The ICE classes for each endpoint are then combined to generate an Irritancy Classification for each test chemical.

Decision Criteria

Once each endpoint has been evaluated, ICE classes can be assigned based on a predetermined range. Interpretation of corneal swelling (Table 3), opacity (Table 4), and fluorescein retention (Table 5) using four ICE classes is done according to the scales shown below. It is important to note that the corneal swelling scores shown in Table 3 are only applicable if thickness is measured with a slit-lamp microscope (for example Haag-Streit BP900) with depth-measuring device no. 1 and slit-width setting at 9, equalling 0,095 mm. Users should be aware that slit-lamp microscopes could yield different corneal thickness measurements if the slit-width setting is different.

Table 3

ICE classification criteria for corneal swelling

Table 4

ICE classification criteria for opacity

Table 5

ICE classification criteria for mean fluorescein retention

The in vitro classification for a test chemical is assessed by reading the GHS classification that corresponds to the combination of categories obtained for corneal swelling, corneal opacity, and fluorescein retention as described in Table 6.

Table 6

Overall in vitro classifications