ISO 10993-3:2026 Biological evaluation of medical devices — Part 3: Evaluation of genotoxicity, carcinogenicity, reproductive toxicity and developmental toxicity

Foreword

ISO (the International Organization for Standardization) is a worldwide federation of national standards bodies (ISO member bodies). The work of preparing International Standards is normally carried out through ISO technical committees. Each member body interested in a subject for which a technical committee has been established has the right to be represented on that committee. International organizations, governmental and non-governmental, in liaison with ISO, also take part in the work. ISO collaborates closely with the International Electrotechnical Commission (IEC) on all matters of electrotechnical standardization.

The procedures used to develop this document and those intended for its further maintenance are described in the ISO/IEC Directives, Part 1. In particular, the different approval criteria needed for the different types of ISO document should be noted. This document was drafted in accordance with the editorial rules of the ISO/IEC Directives, Part 2 (see www.iso.org/directives).

ISO draws attention to the possibility that the implementation of this document may involve the use of (a) patent(s). ISO takes no position concerning the evidence, validity or applicability of any claimed patent rights in respect thereof. As of the date of publication of this document, ISO had not received notice of (a) patent(s) which may be required to implement this document. However, implementers are cautioned that this may not represent the latest information, which may be obtained from the patent database available at www.iso.org/patents. ISO shall not be held responsible for identifying any or all such patent rights.

Any trade name used in this document is information given for the convenience of users and does not constitute an endorsement.

For an explanation of the voluntary nature of standards, the meaning of ISO specific terms and expressions related to conformity assessment, as well as information about ISO's adherence to the World Trade Organization (WTO) principles in the Technical Barriers to Trade (TBT), see www.iso.org/iso/foreword.html.

This document was prepared by Technical Committee ISO/TC 194, Biological and clinical evaluation of medical devices, in collaboration with the European Committee for Standardization (CEN) Technical Committee CEN/TC 206, Biological and clinical evaluation of medical devices, in accordance with the Agreement on technical cooperation between ISO and CEN (Vienna Agreement).

This fourth edition of ISO 10993-3 cancels and replaces ISO 10993-3:2014 and ISO/TR 10993-33:2015.

The main changes are as follows:

A list of all parts in the ISO 10993 series can be found on the ISO website.

Any feedback or questions on this document should be directed to the user’s national standards body. A complete listing of these bodies can be found at www.iso.org/members.html.

Introduction

The basis for biological evaluation of medical devices is often empirical and driven by relevant concerns for human safety. The risk of serious and irreversible effects, such as cancer or second-generation abnormalities, is of particular public concern. It is inherent in the provision of safe medical devices that such risks be minimized to the greatest extent feasible. The assessment of mutagenic, carcinogenic, reproductive toxicity and developmental toxicity hazards is an essential component of the control of these risks. Not all test methods for the assessment of genotoxicity, carcinogenicity, reproductive toxicity and developmental toxicity are equally well developed, nor is their validity well established for the testing of medical devices.

Significant issues with test sample size and preparation, scientific understanding of disease processes and test validation can be cited as limitations of available methods. Since the previous edition of this document, many genotoxicity test methods have been updated with revised OECD guidelines. However, these generally provide clearer recommendations but little alteration in test methods. Scientifically sound alternatives to the proposed testing can be acceptable insofar as they address relevant matters of safety assessment of the medical device.

In the selection of tests needed to evaluate a particular medical device, a careful assessment of expected human uses and potential interactions with biological systems is important, particularly in such areas as reproductive toxicity and developmental toxicity.

This document presents test methods and evaluation strategies for the identification of specific biological harms. Testing is not always necessary or appropriate in evaluating biological risks associated with exposure to medical device materials but, where it is appropriate, it is important that maximum test sensitivity is achieved.

In view of the multitude of possible outcomes and the importance of factors such as extent of exposure, species differences and mechanical or physical considerations, risk assessment is typically performed on a case-by-case basis. Suggestions for risk consideration and integration of ISO 10993-17 and ISO 10993-18^[7^] are provided.

1 Scope

This document specifies strategies for risk estimation and evaluation of biological harms with respect to:

This document is applicable when the need to evaluate a medical device for potential genotoxicity, carcinogenicity, reproductive toxicity and developmental toxicity has been established.

This document is not applicable to active pharmaceutical ingredients of device-drug combination products or biological components of device-biologic combination products which are covered by regulations.

NOTE

Guidance on selecting relevant biological effects for medical devices is covered in ISO 10993-1.

2 Normative references

The following documents are referred to in the text in such a way that some or all of their content constitutes requirements of this document. For dated references, only the edition cited applies. For undated references, the latest edition of the referenced document (including any amendments) applies.

• ISO 10993-1:2025 Biological evaluation of medical devices — Part 1: Requirements and general principles for the evaluation of biological safety within a risk management process
• ISO 10993-2 Biological evaluation of medical devices — Part 2: Animal welfare requirements
• ISO 10993-12:2021 Biological evaluation of medical devices — Part 12: Sample preparation and reference materials
• ISO 10993-17 Biological evaluation of medical devices — Part 17: Toxicological risk assessment of medical device constituents
• ISO 14971:2019 Medical devices — Application of risk management to medical devices
• OECD 471:2020 OECD Guideline for the Testing of Chemicals — Bacterial Reverse Mutation Test
• OECD 473:2016 OECD Guideline for the Testing of Chemicals — In vitro Mammalian Chromosome Aberration Test
• OECD 487:2016 OECD Guideline for the Testing of Chemicals — In vitro Mammalian Cell Micronucleus Test
• OECD 490:2016 OECD Guideline for the Testing of Chemicals — In vitro Mammalian Cell Gene Mutation Tests Using the Thymidine Kinase Gene

3 Terms and definitions

For the purposes of this document, the terms and definitions given in ISO 10993-1, ISO 10993-12 and the following apply.

ISO and IEC maintain terminology databases for use in standardization at the following addresses:

3.1 carcinogenicity test

3.2 developmental toxicity evaluation

3.3 genotoxicity test

3.4 reproductive toxicity evaluation

3.5 sham-incubated vehicle

3.6 threshold of toxicological concern

4.1 General

ISO 10993-1 indicates circumstances where the potential for genotoxicity, carcinogenicity, reproductive toxicity and developmental toxicity are relevant biological harms for consideration in an overall biological safety evaluation. Evaluation of the biological harms shall consider the following factors:

The decision to waive in vitro genotoxicity, carcinogenicity, reproductive toxicity and developmental toxicity testing shall be justified. The decision to conduct in vivo testing shall be justified.

If exposure to constituents identified in accordance with ISO 10993-18^[7^] is determined to be within acceptable levels based on a toxicological risk assessment conducted in accordance with ISO 10993-17, no additional testing needs to be performed to address the relevant risks.

4.2 Other considerations

Toxicity evaluation can be warranted for additional states of the device, such as wear debris generated from the device, absorbable materials, or materials that cure in situ (e.g. cements, adhesives, pre-polymer mixtures) unless toxicological risk assessment based on material information determines no cause for concern. For guidance on test sample preparation for in situ curing devices, see ISO 10993-12.

NOTE

Material information can include chemical composition, degradation products, debris size and morphology, pre-polymerization, post-polymerization products, etc.

5.1 General

Assessment of genotoxic risk should be made:

1. a) by conducting toxicological risk assessment based on chemical characterization according to ISO 10993-17 and ISO 10993-18^[7^]; or
2. b) by performing genotoxicity testing in the standard test battery according to 5.2.2.

In certain cases (e.g. a material which has not been used in a legally marketed device for its intended use), both toxicological risk assessment of medical device constituents based on chemical characterization and genotoxicity testing of a medical device can be appropriate.

NOTE 1

Genotoxicity testing can also be appropriate when chemical characterization data is insufficient for a toxicological risk assessment.

If the device is made with and releases nanomaterials, additional evaluation shall be considered.

NOTE 2

See ISO/TR 10993-22^[9^] and Annex D for more information.

If chemical characterization of the device identifies genotoxicants which are not easily detected using non-targeted analytical approaches described in ISO 10993-18^[7^] (e.g. glutaraldehyde, formaldehyde), targeted analysis of genotoxic residuals shall be conducted. Where compositional profiling or analytical chemistry reveals the presence of chemical constituents with inadequate genotoxicity data and where quantitative structure activity relationship (QSAR) information is inconclusive (see ISO 10993-17), testing of individual chemicals or device components containing the chemical should be considered. In either case, the test sample shall be justified and documented.

5.2.1 General

No single test is capable of detecting all types of genotoxic agents. Therefore, the usual approach is to conduct a battery of in vitro and, under certain circumstances, also in vivo tests.

Genotoxicity tests are designed to detect the two major classes of genetic damage:

NOTE

Bacterial reverse mutation assays have been shown to detect relevant small-scale mutational changes produced by the majority of genotoxic carcinogens detected by rodent bioassays.

5.2.2 In vitro genotoxic test battery

When genotoxicity testing is performed, it shall be conducted in accordance with Annex B and the test battery shall include:

1. a) a test for gene mutations in bacteria (see OECD 471), modified for use with medical devices, see Clause B.2, and
2. b) one of the following three validated in vitro mammalian cell tests:
   • 1) an in vitro test with cytogenetic evaluation of chromosomal damage with mammalian cells (see OECD 473), modified for medical devices, in accordance with Clause B.3 and B.4.2, or
   • 2) an in vitro mouse lymphoma tk^+/- assay using L5178Y (tk^+/-) cells (see OECD 490), modified for medical devices, including detection of small (slow growing) and large colonies, see Clause B.3 and B.4.1, or
   • 3) an in vitro mammalian cell micronucleus test for chromosomal damage and aneugenicity (see OECD 487), modified for medical devices, see Clause B.3 and B.4.3.

NOTE 1

If the medical device is known to contain nanomaterials testing according to OECD 471 is not recommended, see Annex D.

All tests (bacterial and in vitro mammalian tests) shall be performed in the presence and absence of metabolic activation. Metabolic activation is achieved by addition of a co-factor-supplemented post-mitochondrial fraction (S9).

NOTE 2

An example for a S9 recipe is given in Annex E.

In vitro mammalian tests shall include both a short-term exposure with and without S9 and a long-term exposure without S9.

NOTE 3

See B.3.2.3 for descriptions of short-and long-term exposure.

Results from both chromosomal aberration assays and the in vitro mouse lymphoma tk^+/- assay (MLA) have a relatively high level of congruence for compounds that are regarded as genotoxic but yield negative results in the bacterial reverse mutation assay. The MLA detects the broadest set of genetic damage, including both small-scale and large-scale genetic damages. The micronucleus assay detects chromosomal damage in the form of whole chromosome loss or chromosomal fragments and uniquely detects aneuploidy. The three tests are equally acceptable as an in vitro mammalian genotoxicity test (see OECD 473, OECD 487 and OECD 490).

The hypoxanthine guanine phosphoribosyl transferase gene mutation assay (HGPRT/HPRT) method (see OECD 476^[84^]) is not included in the recommended test battery for device testing. This test can be appropriate for medical devices containing nanomaterials (see Annex D).

5.2.3 In vivo genotoxicity testing

In vivo genotoxicity tests are not part of the recommended genotoxicity test battery because they are relatively insensitive and not likely to detect genotoxicity at the concentrations of substances generally found in medical devices or in medical device extracts^[73^][74^]. These concentrations can be insufficient to meet the limit dose required by the test methods. However, under certain circumstances, an in vivo test can be useful.

These circumstances can include:

1. a) assessment of nanomaterial or other particulates suspected of sequestration in particular organs or tissues (see Annex D, Reference [30] and ISO/TR 10993-22^[9^]); or
2. b) when additional relevant factors [e.g. toxicokinetics (see ISO 10993-16^[6^]), organ-specific toxicity] need to be considered.

In vivo tests are valid only with data demonstrating the test sample reached the target cells, organs, or tissues. Mutational analysis in target organs of rats or mice can be determined using one of several rodent transgenic mutation assays (see OECD 488^[27^]); chromosomal damage can be assessed in the erythrocyte micronucleus assay (see OECD 474^[24^]), the in vivo comet assay (see OECD 489^[28^]), or the in vivo chromosomal aberration assay (see OECD 475^[25^]). See Clause B.5 for additional considerations for in vivo assays.

All animal studies shall be performed in a facility meeting the requirements of ISO 10993-2.

5.2.4 Follow-up evaluation

If in vitro genotoxicity tests are performed in accordance with 5.2.2 and if the tests are negative, the test sample can be considered non-genotoxic and further testing is not necessary.

Equivocal results (i.e. elevated results that do not meet criteria for a positive result), can require repeating tests with a modified protocol such as adjusting dilution of the test sample or extract.

When any single in vitro test is positive or equivocal, then the following options shall be considered:

1. a) Identification of genotoxic constituents by chemical characterization (information gathering or analytical chemistry testing) or other testing, with appropriate steps taken to eliminate or manage any identified risks.
2. b) If confounding factors (e.g. osmolality, pH, cytotoxicity) are suspected to have caused the positive response or equivocal in an initial genotoxicity test, the dosing or test conditions may be modified with a justification to address confounding factors.
3. c) Presume that a genotoxicity risk exists and manage the risk in accordance with ISO 14971.

5.3 Test sample preparation

Test samples should be extracted in vehicles compatible with the test system (see Annex A). Use of extraction vehicles that are incompatible with the test system requires additional manipulation (e.g. solvent removal) that can alter the composition of the extract and, therefore, affect the results of the test. Use of extraction vehicles that are incompatible with the test system shall be justified.

Extraction shall be performed in accordance with ISO 10993-12 unless otherwise justified.

Tests can be performed on solutions (e.g. soluble or liquid devices), suspensions (e.g. Method A in Annex A), extracts (e.g. Method B in Annex A) of the finished product, medical device components or individual chemicals from the medical device. If individual chemicals are tested, then the appropriate OECD guidelines shall be followed.

NOTE

Some authorities having jurisdiction prefer Method C in Annex A while others do not accept Method C.

6.1 General

The objective of carcinogenicity studies is to determine tumorigenic potential of medical devices or medical device constituents in animals and to evaluate the relevant risk to patients. Mechanisms of carcinogenesis can be genotoxic or non-genotoxic. Medical devices shall be evaluated for both types of carcinogenicity, preferably by alternative approaches, such as chemical characterization and toxicological risk assessment, that do not use in vivo testing.

Animal testing for carcinogenicity is technically challenging. It requires implantation of many medical devices, representative samples of the medical device, or the application of concentrated medical device extracts to achieve the elevated doses necessary for a valid test in a reasonable number of animals. For these reasons, animal testing is generally not justified when the risks can be adequately assessed by other means.

NOTE

For additional information on animal carcinogenicity testing, see Annex C.

6.2.1 General

Carcinogenicity should be evaluated initially using chemical characterization according to ISO 10993-18^[7^] and toxicological risk assessment according to ISO 10993-17. When medical devices or materials are not readily amenable to analytical chemistry testing and adequate chemical characterization information is available from information gathering, it is always preferable to use the chemical information and toxicological risk assessment including assessment of available genotoxicity data to evaluate carcinogenicity rather than performing an animal carcinogenicity study.

NOTE 1

Some authorities having jurisdiction can require consideration of carcinogenicity testing for novel materials (e.g. when data are not available to provide an adequate assessment).

NOTE 2

Evaluation of solid-state carcinogenesis due to release of particulates or fibres from a medical device can be partially addressed by toxicological risk assessment according to ISO 10993-17. This does not address all the risks. See Annex C for more information.

If the results of toxicological risk assessment indicate the risk of carcinogenicity is acceptable, no further assessment is required.

Carcinogenicity testing shall not be performed when risks can be adequately assessed or managed without generating new carcinogenicity test data or when the benefit-risk analysis obviates the need for a carcinogenicity assessment. If the toxicological risk of carcinogenicity is unacceptable, the risk shall be managed accordingly (see also ISO 14971).

If the medical device contains a chemical with inadequate toxicological data, biological testing to address carcinogenicity shall be justified and documented based on the need for additional information. The assessment should address both genotoxic (see 6.2.2) and non-genotoxic carcinogenicity (see 6.2.3) on the individual chemical identified in the toxicological risk assessment.

NOTE 3

The history of human clinical use can be considered in the risk assessment. The clinical evidence for carcinogenicity risk is generally not monitored over an adequate timeframe and this aspect of medical device safety can be unknown.

6.2.2 Genotoxic carcinogens

The carcinogenicity risk from genotoxic carcinogens can be established by genotoxicity testing (if unknown) or by toxicological risk assessment (of identified genotoxins). Genotoxicity testing requires a battery of tests (see 5.2.2).

Where genotoxicity of a medical device is established, either based on genotoxicity testing or prior material knowledge of the medical device, additional evaluation of carcinogenic risk shall be performed in accordance with ISO 10993-17. It can be appropriate to identify the genotoxic constituent and focus the evaluation on only that constituent.

Genotoxic chemicals that are DNA reactive but with adverse effects that can be mitigated by rapid detoxification before coming in contact with DNA or by a demonstrated effective repair of induced damage can be evaluated by the use of a No observed effect level and uncertainty factors^[75^].

NOTE

While genotoxic carcinogenicity can include non-threshold mechanisms, there are exposure levels below which impact on lifetime cancer risk is considered acceptable (usually less than 1 in 100 000 anticipated additional lifetime incidence).

6.2.3 Non-genotoxic carcinogens

Non-genotoxic carcinogens do not cause direct DNA damage but induce cancer via other mechanisms (see References [54], [55] and [60]). These carcinogens can induce their effects (leading to cancer) upon repeated exposure leading to prolonged interference with homeostatic processes that lead to irreversible and uncontrolled cellular proliferation. The estimation of carcinogenic risk for non-genotoxic chemicals is complicated by:

There are many practical challenges to conducting a rodent carcinogenicity test (see 6.1). Therefore, a toxicological risk assessment according to ISO 10993-17 should be used.

If a medical device constituent is not genotoxic and carcinogenicity data are not available, then a tolerable intake (TI) based on subchronic or chronic systemic toxicity data can be used to evaluate carcinogenicity risk. If there are insufficient data to set a TI, then the TTC may be used to estimate the risk in accordance with ISO 10993-17. Recent publications indicate that the TTCs based on Cramer Class are protective for non-genotoxic carcinogens (see Reference [62]).

NOTE 1

Non-genotoxic carcinogens have thresholds, below which the risk of cancer is unlikely (see Reference [61]).

NOTE 2

Approaches for prediction of non-genotoxic carcinogenicity are under active investigation (e.g. References [58] and [60]).

NOTE 3

Examples of non-genotoxic carcinogens can be found in the supplementary material in Reference [62].

7.1 General

Reproductive toxicity evaluation shall be considered for medical devices intended for use in individuals of reproductive age under any of the following conditions:

Developmental toxicity evaluation shall be considered for:

7.2 Evaluation methods

Assessment of reproductive and developmental risk should be made:

1. a) by toxicological risk assessment based on chemical characterization according to ISO 10993-17 and ISO 10993-18^[7^]; or
2. b) by performing developmental and reproductive testing as described in Annex G.

Current biological test methods for the evaluation of reproductive toxicity and developmental toxicity are for single substances and are not validated for mixtures (e.g. medical device extracts). Evaluation for medical devices shall first consider a toxicological risk assessment focusing on reproductive toxicity and developmental toxicity in accordance with ISO 10993-17. Exposure estimates can be determined through chemical characterization outlined in ISO 10993-18^[7^].

Exposure estimates can also be improved with

performed according to ISO 10993-18^[7^].

If it is determined that the amount of a reproductive and developmental toxicant in the medical device is above acceptable limits, the presence of the toxicant shall be carefully considered for risk control where possible and in the benefit-risk analysis in the (overall) risk management file in accordance with ISO 14971.

The following factors should be considered in the risk assessment:

NOTE

For more information on reproductive toxicity and developmental toxicity, see Annex G.

When the remaining risk is unacceptable due to a lack of robust experimental data, individual chemical substances can be tested with in vitro or in vivo biological test methods (e.g. see Reference [63], Clause G.2 and OECD 43^[14^]). For example, testing an individual chemical in a developmental toxicity study can yield an effect level or no effect level to help refine the risk assessment.

8 Test report

The test report shall include at least the following details:

1. a) the International Standard used (including its year of publication);
2. b) a description of the test sample (e.g. identifier, material composition, processing, size and weight) including its intended use;
3. c) a description, rationale and justification of the test methods used (e.g. OECD 471) and the test details (including at least conditions, materials, dose and test sample preparation);
4. d) a description of analytical methods including quantification limits;
5. e) a statement of conformity of the laboratory to appropriate current best laboratory or quality practices, for example, Good Laboratory Practices (GLP) or ISO/IEC 17025^[8^], where applicable;
6. f) a description of S9 mix (including manufacturers as applicable) and final % of S9 fraction used in testing;
7. g) historical control data;
8. h) test results including summary;
9. i) statistical methods, if applicable;
10. j) any deviations from the procedure or any unusual features observed;
11. k) interpretation and discussion of results;
12. l) further details regarding follow-up testing as specified in Annex B, Annex C or Annex G, if applicable;
13. m) date of the test;
14. n) name and signature of the responsible person.

A.1 General

This annex gives guidance for selecting an appropriate test sample preparation method for genotoxicity testing of medical devices. The selection of the appropriate test sample preparation method including vehicles, is critical to obtain relevant results from genotoxicity tests, because inappropriate test sample preparation can result in an underestimation of genotoxicity risk.

A.2.1 General

The ideal test sample preparation method for genotoxic tests is exhaustive extraction.

ISO 10993-12 may be used to conduct exaggerated extraction for test sample preparation, as this is considered more practical and achievable. For some medical devices (e.g. liquids), other test sample preparation methods can be more appropriate.

NOTE 1

The selection of a test sample preparation method for any material or medical device intended for use in humans requires a structured approach that takes the chemical composition and physicochemical properties of the material or medical device into consideration.

The test sample preparation should follow one of the two proposed methods – Method A or Method B – described in Figure A.1.

Figure A.1

When the test sample requires extraction in accordance with ISO 10993-12, Method B (see A.2.3) shall be used. Otherwise, Method A (see A.2.2) should be used.

The method used should be justified.

NOTE 2

Some authorities having jurisdiction prefer Method C while others do not accept Method C.

The solvents or extraction vehicles shall be compatible with the test sample and the test system. Some vehicles can react with or alter the test material (e.g. cause degradation, swelling). Care should be taken to avoid altering the extractable profile resulting in clinically irrelevant compounds.

A.2.2 Method A

The test sample is either dissolved or suspended (or partially dissolved) in the solvent. The final volume of test sample preparation added to the test system varies with the compatibility of the solvent in the test system.

Solutions should be used immediately after preparation to prevent sorption on to containers or to prevent other changes in composition. If samples are not used immediately, justification should be provided (e.g. stability of the solution). Solution pH shall not be adjusted unless a rationale is provided. The solution shall not routinely be processed by filtration, centrifugation or other methods to remove suspended particulates. However, if such processing is necessary, the process, outcome and rationale shall be documented.

If cytotoxicity of the solution is expected, preliminary dose range assays can be useful to achieve the proper dose selection. However, dosing at a single, maximum level is acceptable if cytotoxicity is not expected. Refer to the genotoxicity methods in Annex B for maximum doses and solvent compatibility with the bacterial and mammalian cell tests.

NOTE

For detailed information on Method A, see ISO 10993-12.

A.2.3 Method B

The test sample shall be extracted in a vehicle compatible with the test system. The recommended polar vehicles are saline (for the bacterial assays) and cell culture medium without serum (for the mammalian in vitro assays). The recommended non-polar vehicle is DMSO. Other non-polar vehicles such as undiluted PEG 400 and ethanol may be used. The use of alternative extraction vehicles shall be justified and documented.

Extracts should be used within 24 h after preparation to prevent sorption on to the container or other changes in composition. Extract pH shall not be adjusted unless a rationale is provided. The extract shall not routinely be processed by filtration, centrifugation or other methods to remove suspended particulates. However, if such processing is necessary, the process, outcome and rationale shall be documented.

NOTE

For detailed information on Method B, see ISO 10993-12.

A.3.1 General

There is an optional test sample preparation method that can be applied to medical device extraction.

NOTE

Some authorities having jurisdiction require or prefer this extraction method, historically referred to as "Method B" in ISO 10993-3:2014 for the test sample preparation for genotoxicity testing. The former "Method B" is revised as "Method C" in this document.

A preliminary step is required to measure the amount of extractable material in the test sample. The test sample preparation (see A.3.2.1) and method (see A.3.2.2) shall be carried out in order to select either Method C option 1 or Method C option 2.

A.3.2.1 General

The test sample is immersed in a solvent that extracts chemical substances from the test sample but does not dissolve the test sample. If test sample appearance or weight confirms partial degradation, use Method B (see A.2.3). Two or more solvents are tested to determine which solvent extracts the higher percentage of extractables from the test sample. Common solvents are methanol and acetone. The extracted, evaporated test sample residue is dissolved or suspended in a vehicle compatible with the genotoxicity test system.

For the in vivo test, the extracted and evaporated test sample residue is dissolved or suspended in vehicles compatible with the test system. The selection of the highest dose and administration route are the same as those given in Annex B.

A.3.2.2 Method (pre-experiment)

The method is as follows:

A.3.2.3 Expression of results

For each solvent, calculate the mass of the extracted residuals, W_R, in the flask by determining the change in mass of the flask using Formula (A.1):

(A.1)

where

m₁

is the mass of the empty flask, in g;

m₂

is the mass of the flask after evaporation of the extract, in g.

Calculate the percentage of extractables using Formula (A.2):

(A.2)

where

e%

is the percentage of extractables;

m₃

is the mass of the test sample prior to extraction, in g.

Record and report the percentage of extractables for each vehicle.

The study report should include the rationale for selection of the solvent and the percentage of residue for each solvent tested.

The appropriate option of Method C is selected according to the percentage of extractables obtained in the pre-test are shown in Table A.1.

Method C option 1 (exaggerated) is described in detail in A.3.2.4 and Method C option 2 (enhanced) is documented in A.3.2.5.

A.3.2.4 Method C option 1 (exaggerated)

The methods are as follows:

A.3.2.5.1 General

For Method C option 2, the recommended polar vehicles are saline for the bacterial assays and cell culture medium without serum for the mammalian in vitro assays. Acetone is the recommended non-polar vehicle, as it is volatile and compatible with the test system. Methanol and ethanol are compatible volatile, non-polar vehicles. The selection of vehicle should be justified.

The test sample is immersed in polar and non-polar extraction vehicles that extract leachables from the test sample but do not dissolve the test sample.

A.3.2.5.2 Method with non-polar volatile vehicles (e.g. acetone)

Immerse the test sample in the extraction vehicle in a glass container. If the test sample is the finished medical device, the extraction should be performed using the appropriate extraction ratio according to ISO 10993-12. The vehicle should cover the test sample without cutting, if possible.

NOTE 1

Details on the extraction of absorbent materials are included in ISO 10993-12.

Extracts should be reduced to volume fraction of 10 %. They should not be heated, to avoid chemical changes of residues or loss of volatile compounds.

NOTE 2

The Soxhlet extraction method can be considered as an alternative method as long as the vehicle is extracted to a volume fraction of approximately 10 %, not dryness.

Evaporation of the extract following extraction is not applicable for cases when the suspected residue in the medical device is highly volatile (e.g. ethylene oxide, low molecular weight acrylate monomers).

The volatile acetone, methanol or ethanol extracts can be reduced to a volume fraction of approximately 10 % of the original volume in order to concentrate the extract or reduced to dryness (see Method C option 1). When volatile contaminants are expected, extracts shall not be evaporated to dryness in order to preserve these substances.

A.3.2.5.3 Method with polar vehicles (e.g. saline, cell culture medium without serum)

Immerse the test sample in the extraction vehicle in a glass container. If the test sample is the finished medical device, the extraction should be performed using the appropriate extraction ratio according to ISO 10993-12. The vehicle should cover the test sample without cutting, if possible.

Extract the test sample for (72 ± 2) h at (37 ± 1) °C for cell culture medium and for (72 ± 2) h at (50 ± 2) °C or for (24 ± 2) h at (70 ± 2) °C for saline with constant agitation. For medical devices that contain heat labile or heat sensitive materials, extraction at 37 °C in saline or another vehicle according to ISO 10993-12 is acceptable with justification.

A.4.1 Degradable or absorbable materials

For medical devices made with degradable or absorbable materials, particulates can form during extraction. If during test sample preparation particulates form, the particulates should not be filtered or centrifuged or allowed to settle prior to introducing the sample to the test system. These generated particulates remain part of the test sample. If the particulates are sufficient to interfere with the test, additional genotoxicity testing with removal of particulates may be conducted if adequate scientific evidence can be provided to demonstrate that the particulates interfered in the assay and that the extract manipulation does not otherwise alter the extract composition.

For absorbable materials that require extraction, the extraction conditions described in ISO 10993-12 shall be used. Any adjustment applied to the extracts prior to biocompatibility testing shall be justified.

NOTE

For more information on test sample preparation for testing of absorbable medical devices, see ISO 10993-13^[3^], ISO 10993-14 ^[4^], ISO 10993-15^[5^], ISO 10993-18^[7^] and ISO/TS 37137-1^[12^].

A.4.2 Wear debris from metals, alloys and ceramics

Wear debris can cause adverse biological effects due to various factors such as increased release of metal ions. During clinical use, metal ions released from inorganic materials (e.g. stainless steel, titanium alloy, hydroxyapatite, tricalcium phosphate, alumina, zirconia) due to wear or corrosion can present a genotoxic hazard. For example, in vivo genotoxic effects have been seen in periprosthetic lymphocytes of patients implanted with metal-on-metal hip implants^[36^]. Many metal ions play an important regulatory role in the body and this role is dependent on their chemical characteristics and valence state. Metal ions bind to proteins in the physiological fluids (such as blood, lymphatics and urine) and can partition to different fractions for biodistribution. Metal ions are internalized into the cell where they can bind to nuclear material and alter cell signalling locally or systemically.

Therefore, the metal ions’ genotoxicity profile should be evaluated and assessed as far as possible based on literature data; refer to:

Although the assays in this document are designed to measure the genotoxic potential of final medical device extracts (in solution) and not particulates, the standard assays can still be appropriate to assess the genotoxic potential of wear debris or particles, with minor changes to methods. For example, the bacterial colonies can require counting by hand rather than by automatic colony counter.

NOTE

Some wear debris can be present in nanoscale form. Modified methods and a modified test battery for assessment of nanomaterials have recently been published. See Reference [29] and Annex D.

A.4.3 Low molecular weight compounds (LMWCs)

Many medical devices constructed from polymers can contain LMWCs. When the test sample is likely to be composed of one or more LMWC (e.g. <1 000 Da), it should be evaluated using Method A (see A.2.2).

A.4.4 Medical devices containing nanomaterials (NMs)

General guidance regarding nanomaterials can be found in Annex D.

B.1.1 General

When evaluating medical devices for genotoxic potential, a medical device or material, or its extract, can be applied to test systems. If extracts are evaporated and residues are used, justification should be provided to show that extractables are not lost or degraded (see Annex A, Method C). It is important to maximize exposure of the test system to the substances present in the medical device. When determining the highest amount of test sample to be used, criteria to consider should include cytotoxicity and solubility in the final treatment mixture. It can be useful to determine toxicity and insolubility in a preliminary experiment. The cytotoxicity of a test sample can be altered in the presence of metabolic activation systems. Insolubility should be assessed as precipitation in the final mixture under the actual test conditions and evident to the unaided eye.

If the test sample is cytotoxic (or precipitates) a full study with several different analysable concentrations of the test sample shall be used. If the test sample is non-cytotoxic, dosing at a single, maximum level is appropriate. For insoluble medical devices or materials, a study conducted with only one dose, the neat extract, is acceptable. For non-cytotoxic soluble or partially soluble medical devices or materials, the top dose should be selected based on the test system and used as the single study dose (see dose selection in B.2.2.2).

B.1.2 Test sample preparation

The selection of a test sample preparation method for any medical device shall consider the chemical composition and physicochemical properties of the material(s) used in the medical device. Annex A should be consulted for test sample preparation guidance. Two extracts shall be tested; one polar and one non-polar.

NOTE

Additional information is provided in Annex A.

B.1.3 Stability

Test extracts shall be used within 24 h of preparation. Extracts should, if possible, be used immediately after preparation to prevent adsorption onto the extraction container or changes in composition. If an extract is stored longer than 24 h, then the stability and homogeneity of the extract under the storage conditions shall be verified.

B.2.1 General

Methods evaluating medical devices using the bacterial reverse mutation test shall be performed in accordance with OECD 471.

B.2.2.1 Vehicles added to the test system

The vehicles shall be compatible with the survival of the bacteria and the S9 activity. Saline is the recommended polar vehicle. The recommended non-polar vehicle is DMSO. Other non-polar vehicles such as undiluted PEG 400 (see Annex A, Method B), ethanol and acetone (see Annex A, Method C) may be used. The use of other vehicles shall be justified and documented.

B.2.2.2 Dose selection

For non-cytotoxic liquid or soluble medical devices, the recommended maximum test dose is 100 µl per plate. For soluble non-cytotoxic solid test samples (e.g. powder, see Annex A, Method A), 5 mg per plate should be used as the maximum test concentration. For extracts, material or vehicle extraction ratios are described in Annex A or ISO 10993-12:2021.

If the test sample is cytotoxic (or interfering precipitates), a full study with at least five different analysable concentrations of the test sample shall be used. However, a study conducted with only one dose, the highest dose, or neat extract, is acceptable if the test sample is not cytotoxic or genotoxic. If test samples or extracts are toxic, preliminary experiments, using test concentrations at two-fold intervals (e.g. 100 %, 50 %, 25 %) can be appropriate in choosing a dose range. Smaller intervals can be appropriate when subsequent experiments are performed.

B.2.2.3 Method

Bacterial strains used are Salmonella typhimurium TA98, TA100, TA1535, TA1537 (or TA97, TA97a) and TA 102 (or E. coli WP2 uvrA, E. coli WP2 uvrA pKM101). Genotypes of the strains should be confirmed by tests affirming the genetic markers for DNA repair functions and cell wall permeability.

The test can be conducted using the plate incorporation or the preincubation method. The choice of test method shall be based on medical device material information and the capability of the test method to detect the medical device materials. Chemical classes that include short chain aliphatic nitrosamines, divalent metals, aldehydes, azo-dyes and diazo compounds, pyrrolizidine alkaloids, allyl compounds and nitro compounds are detected more efficiently using the preincubation method (see OECD 471:2020 and Reference [69]).

The test shall be conducted both in the presence and absence of metabolic activation (S9 mix) (see Annex E). The test shall include negative and positive controls. The negative control is a control blank (in the case of Annex A, Method A) or the sham-incubated extraction vehicle, i.e. vehicle incubated as for the preparation of test sample extracts (see Annex A, Method B and Method C). The positive controls (with and without metabolic activation) for each tester strain shall be used based on the recommendations in OECD 471.

One hundred (100) µl of fresh overnight cultures of bacterial cells are added to a tube with 100 µl test sample (solution or extract) or control and 500 µl S9 mix or phosphate buffer control. Two (2) ml warm soft agar (approximately 45 °C) with minimal amounts of histidine (for Salmonella) or tryptophan (for E. coli) are added to the tube and the contents poured onto a minimal medium E agar plate. See Reference [38] for methods and compositions for agar and S9 mix. Triplicate plates shall be used for each test and control. Plates are incubated for 2 d to 3 d at 37 °C. Revertant colonies are counted and compared to the number of spontaneous revertant colonies on negative control plates. The background lawn on each individual plate should be evaluated for evidence of cytotoxicity (see Table B.1).

The test should meet the assay acceptance criteria for the controls as described below:

1. a) The mean number of spontaneous revertants in the control should be within the laboratory's historical negative control data or the range reported in the literature.
2. b) The mean of the number of revertants in each positive control should exhibit significant increase compared with that in the respective solvent control or within the laboratory's historical positive control data.

Statistical analysis may be applied, but generally positive responses are recognized if mutant counts are equal to or greater than two-fold-increase concurrent negative control data for the strains with high levels of spontaneous mutations (TA98, TA100, TA97, TA102 and the E. coli strain) and three-fold-increase concurrent negative control data, for the strains with lower background levels of spontaneous mutations (TA1535 and TA1537).

B.3.1 General

Methods evaluating medical devices using in vitro mammalian cell genetic toxicity tests shall be performed in accordance with:

1. a) OECD 473 for in vitro mammalian chromosomal aberration test;
2. b) OECD 487 for in vitro mammalian cell micronucleus test;
3. c) OECD 490 for in vitro mammalian cell gene mutation tests using thymidine kinase.

The methods given in OECD 473, OECD 487 and OECD 490 may be modified for medical device evaluation as described in B.3.2.

B.3.2.1 Vehicles added to the test system

For mammalian cell tests, a cell culture medium (without serum) should be used as a polar extraction vehicle and DMSO should be used as a non-polar vehicle. Non-polar, organic vehicles such as undiluted PEG 400, ethanol (see Annex A, Method B) and acetone (see Annex A, Method C) can be used, if justified. The use of culture medium with serum instead of a non-polar extraction vehicle, shall be justified and documented (e.g. the test sample is not compatible with any of the non-polar extraction vehicles listed in this document). Generally, organic vehicles should not exceed a volume fraction of 1 % in the cell culture, except when cell culture medium with serum extracts are applied neat to the cells. Depending on the serum concentration in the extraction vehicle (i.e. cell culture medium with or without serum), additional serum may be added prior to applying to cells.

Acetone is recommended as a volatile vehicle that can be evaporated to concentrate the extract (see Annex A, Method C).

B.3.2.2 Exposure concentrations

The criteria to be considered when determining the highest concentration are cytotoxicity, solubility in the test system and changes in pH or osmolality.

For non-cytotoxic insoluble medical devices and materials, neat cell culture medium extract shall be tested without dilution and non-polar test extract (e.g. acetone, DMSO, ethanol, PEG 400 extracts) shall be tested at a volume fraction of maximum 1 % after dilution in the cell culture medium prior to dosing the cells.

For medical devices which are incompatible with cell culture medium, physiological saline or water can be used but should be justified. The extracts based on physiological saline or water should be conducted under exaggerated conditions according to ISO 10993-12 (e.g. 50 °C for 72 h). They should be added to the cell culture medium up to a maximum of 10 %.

If the test sample is cytotoxic (or precipitates), at least three (for the chromosomal aberration and the in vitro micronucleus assays) or four (for the mouse lymphoma assay) informative test concentrations, (including the cytotoxicity limit and several less toxic or precipitating concentrations) shall be evaluated. However, a study conducted with only one dose, the highest dose, or neat extract, is acceptable if the test sample is not cytotoxic or genotoxic.

For preliminary experiments, test concentrations using two-fold intervals (e.g. 100 %, 50 %, 25 %) can be appropriate. Smaller intervals can be appropriate when subsequent experiments are performed to target an appropriate dose-response range. The recommended maximum test concentration depends on the test sample preparation method selected.

1. a) For soluble non-cytotoxic solid test samples (see Annex A, Method A or ISO 10993-12:2021), 2 mg/ml should be used as the maximum test concentration.
2. b) For non-cytotoxic liquid or soluble gel test samples, 10 % (final concentration after dilution in the cell culture medium) should be used as the maximum test concentration.
3. c) For non-cytotoxic solid or liquid test samples that are not soluble at concentrations mentioned in a) or b), one or more concentrations tested shall be insoluble in the final treatment mixture. Test samples that are cytotoxic below this concentration shall be tested up to a cytotoxic concentration including an informative dose-response. However, a study conducted with only one dose, the highest dose, or neat extract, is acceptable if the test sample is not cytotoxic or genotoxic.
4. d) For extracts (see Annex A, Method B or ISO 10993-12), a limit study using 100 % (i.e. neat) test sample extract in cell culture medium (i.e. polar extract) and maximum 1 % (final concentration after dilution in the cell culture medium) non-polar extract is acceptable and a dosing study is unnecessary if cytotoxicity is not observed.For medical devices which are incompatible with cell culture medium, physiological saline or water can be used but should be justified. The extracts based on physiological saline or water should be conducted under exaggerated conditions (e.g. 50 °C for 72 h). They should be added to the cell culture medium up to a maximum of 10 %.For non-cytotoxic test samples that produce precipitates, one or more concentrations tested should not produce precipitates in the final treatment mixture. Test samples that are cytotoxic below 100 % (cell culture medium extract or 1:10 dilution for a saline extract) or 1 % (non-polar extract) shall be tested up to a cytotoxic concentration using at least three (for the chromosomal aberration and the in vitro micronucleus assays) or four (for the mouse lymphoma assay) test concentrations in a dose response.However, a study conducted with only one dose, the highest dose or the neat extract is acceptable if the test sample is not cytotoxic or genotoxic.
5. e) Residues from evaporated samples (see Annex A, Method C) should be re-suspended in cell culture medium (ideally) or another compatible solvent and tested up to the maximum concentration of 5 mg/ml, if feasible. Test samples that are cytotoxic below this dose shall be tested up to a cytotoxic concentration using dilutions informative for a dose-response. However, a study conducted with only one dose, the highest dose or the neat extract, is acceptable if the test sample is not cytotoxic or genotoxic.Any precipitate should not interfere with the scoring.
6. f) For test samples prepared according to Method B in Annex A, the recommended maximum test concentration depends on the nature of the extract (e.g. a cell culture medium extract can be applied undiluted, saline extract can be applied as a 1:10 dilution).

B.3.2.3 Methods

For in vitro mammalian cell tests, the culture medium without serum is used as a polar vehicle for extraction. The extract is tested neat following supplementation with serum before dosing the cells. The non-polar extract, such as a DMSO, polyethylene glycol400 (PEG 400), acetone or ethanol test extract, should be tested at maximum 1 % following dilution with serum supplemented cell culture medium. If the non-polar extract is culture medium with serum, it should be tested neat. For cytotoxic test extracts, acceptable cytotoxicity limits of (55 ± 5) % cytotoxicity for chromosomal aberration (CA) and in vitro micronucleus (MNvit) assays and the relative total growth (RTG) window of 10 % to 20 % for the mouse lymphoma assay (MLA) shall be considered for the selection of test extract doses (see OECD 473, OECD 487 and OECD 490).

Alternative polar vehicle may be used; see B.3.2.2 d).

For a thorough evaluation, all three of the following experimental conditions shall be conducted using a short-term treatment with and without metabolic activation and long-term treatment without metabolic activation (including the addition of serum):

1. a) short-term treatment without metabolic activation for 3 h to 6 h for CA and MNvit or for 3 h to 4 h for MLA;
2. b) short-term treatment with metabolic activation for 3 h to 6 h for CA and MNvit or for 3 h to 4 h for MLA;
3. c) long-term treatment without metabolic activation: for about 1,5 normal cell cycle lengths for CA, 1,5 to 2,0 normal cell cycle lengths for MNvit or for approximately 24 h for MLA. Following treatment, cells are collected, washed and given fresh medium with serum for expression of mutations or chromosomal aberrations.

B.3.2.4 Controls

Appropriate negative and positive controls shall be included in the test.

All negative controls for extractions shall be tested as sham-incubated vehicles.

The following positive controls can be used (see OECD 473, OECD 476^[84^], OECD 487^[26^] and OECD 490):

1. a) without metabolic activation:

   — methyl methane sulphonate (CAS RN^®1

   1 CAS Registry Number^® is a trademark of the American Chemical Society (ACS). This information is given for the convenience of users of this document and does not constitute an endorsement by ISO of the product named. Equivalent products may be used if they can be shown to lead to the same results.

   66-27-3); or

   — mitomycin C (CAS RN 50-07-7);

2. b) with metabolic activation:

   — benzo(a)pyrene (CAS RN 50-32-8); or

   — cyclophosphamide (monohydrate) CAS RN 50-18-0 (CAS RN 6055-19-2).

NOTE

If validated, other positive controls can also be used for individual assays.

B.3.2.5 Metabolic activation

The metabolic activation system should be identified including the source of the S9 fraction, formulation of the co-factor mix and the final concentration administered to the cell cultures. Examples of metabolic activation used in the genetic toxicity tests can be found in References [47], [48], [49], [50], [51], [52] and [53].

NOTE

An example for a S9 recipe is given in Annex E.

B.4.1.1 Principle

The mouse lymphoma assay (MLA) is a forward mutation assay in L5178Y (TK+/-) heterozygotes (one normal and one mutant chromosome) that detects a wide variety of genetic alterations, including both mutations and chromosomal damages, at the thymidine kinase (TK) enzyme locus. The detection of the rare (TK-/-) non-functional mutants depends on the selection agent trifluorothymidine (TFT), an inhibitor which is taken up only by the functional (TK+/-) cells. As the major population (TK+/-) is inhibited, the mutants (TK -/-) can grow out into visible colonies.

In the MLA, two distinct phenotypic classes of TK mutants are generated: the normal growing mutants (large colonies) and slow growing mutants (small colonies). Induction of normal growing mutants is associated with substances primarily acting as point mutagens (small scale damage). The induction of small colony mutants has been associated with substances that induce gross structural changes at the chromosomal level, likely affecting genes vital to growth. Consequently, both small and large colony mutants shall be counted in order to recover all of the mutants and to provide some insight into the type(s) of damage (mutagens versus clastogens) induced by the test agent.

B.4.1.2 MLA-specific methods

L5178Y cells are generally grown in RPMI culture. The MLA can be performed following the soft agar or the microwell method (see OECD 490). In the MLA, the L5178Y suspension cultures are exposed to the polar or non-polar test medical device extracts (or the test sample) with the addition of serum, both with and without an exogenous source of metabolic activation. Samples are incubated for short and long incubations as described in B.3.2.3. At the end of the treatment period, cells are collected and washed by pelleting and then maintained in an active growth state for 2 d to allow phenotypic expression of induced mutations. Cytotoxicity is assessed by calculating the relative suspension growth (RSG) after treatment and the expression period, comparing the suspension growth of treated cultures to that of the vehicle control. Following phenotypic expression, cells are seeded in medium containing the selective agent TFT to detect mutant colonies. A dilution of cells is plated in medium without selective agent to determine the cloning efficiency (viability). Acceptable cytotoxicity targets are determined by the RTG [TG is equal to the RSG multiplied by the relative cloning efficiency (RCE)]. For more details, see OECD 490. After a cloning time of 10 d to 15 d, colonies are counted. Mutant frequency (MF) is calculated based on the number of mutant colonies corrected by the cloning efficiency at the time of mutant selection. For a valid experiment with cytotoxic test samples, an RTG window of 10 % to 20 % shall be attained with one dose, with others closely spaced above this value.

The following criteria should be met before determining the overall results for a specific test chemical:

1. a) testing of three experimental conditions (short treatment with and without metabolic activation, approximately 24 h treatment without metabolic activation) unless one of them results in positive results;
2. b) adequate number of cells and analysable concentrations (see OECD 490:2016, paragraphs 27 and 34 to 35);
3. c) criteria for the selection of top concentration that are consistent with those described in B.3.2.2 (see OECD 490:2016, paragraphs 28 to 30 for additional information).

Depending on the method used (i.e. soft agar or microwell method), the test shall meet the assay acceptance criteria for the negative control as described in Table B.2.

For MLA, every test should also be evaluated as to whether the positive control(s) meets at least one of the following two acceptance criteria developed by the International Workshops on Genotoxicity Testing (IWGT) work group:

1. a) The positive control should demonstrate an absolute increase in total MF, that is, an increase above the spontaneous background MF [an induced MF (IMF)] of at least 300 ×10^-6. At least 40 % of the IMF should be reflected in the small colony MF.
2. b) The positive control has an increase in the small colony MF of at least 150 × 10^-6 above that seen in the concurrent untreated or solvent control (a small colony IMF of 150 × 10^-6).

MLA relies on the use of a predefined induced mutant frequency (i.e. increase in MF above concurrent control) designated as the global evaluation factor (GEF) for defining positive and negative responses. For the soft agar MLA method, the GEF is 90 × 10^-6 and for the microwell MLA method, the GEF is 126 × 10^-6. The test sample is considered positive if the test sample dosed culture has an induced mutant frequency (IMF) that exceeds the assay’s GEF and the increase is concentration related (e.g. using a trend test).

RSG is a comparison of the growth of the test sample with growth of the negative control after treatment and expression (at least three population doublings). RTG shall be at least 10 % for a valid test. RTG incorporates both RSG and relative cloning efficiency (RCE) of the treated samples in comparison with the negative control. Cloning efficiency accounts for the proportion of cells plated in the selective medium that actually grow into colonies. Both RSG and RTG are monitored to ensure that enough cells will survive the treatments and plating to ensure a valid test. Sometimes, the toxicity of the test sample interferes with growth and expression of the mutations or interferes with the growth of mutated cells into countable colonies.

NOTE

For more details on MLA-specific methods, see OECD 490.

B.4.2.1 Principle

The chromosomal aberration assay is an assessment of clastogenicity (large scale genetic damage) in cell lines such as Chinese Hamster Ovary (CHO) cells (see B.4.2.2) or primary cell cultures, including human or other mammalian peripheral blood lymphocytes (see B.4.2.3). The use of other cell lines (e.g. CHL/IU cells) and types should be justified based on their demonstrated performance in the test. Structural alterations in chromosomes visible in a microscope are scored. These alterations are generally dependent on breaks in chromosomes that occur during active logarithmic growth of the cells. Cells are arrested in metaphase using colcemid or colchicine so that alterations are visible in condensed chromosomes.

B.4.2.2 Specific methods for CA in CHO cells

CHO cells are generally grown in Ham’s F-12 or McCoy’s culture medium. CHO cells adherent in cell culture dishes are exposed to the polar (cell culture medium) and non-polar test medical device extracts (or the test sample) with the addition of serum, both with and without an exogenous source of metabolic activation for short and long incubations, as described in B.3.2.3. After the short treatments, cells are washed with PBS and fresh cell culture medium added for the expression period of approximately 1,5 doubling times. Cytotoxicity is determined by comparison of the growth of treated cultures with that of the vehicle control [e.g. relative population doubling (RPD) or relative increase in cell count (RICC)]. If the test sample is toxic, a dose-response including one sample near the toxicity limit of (55 ± 5) % should be included among the samples. 1 h to 3 h prior to harvest, colcemid or colchicine is added to the cells. At harvest, cells are treated with a hypotonic solution [e.g. 0,075 M potassium chloride (KCl)] and recovered by centrifugation, including the monolayer culture and any cells that have been dislodged during the incubation period. Cultures are then washed twice with fixative (3:1 methanol: glacial acetic acid) and the cell suspension dropped onto microscope slides. Slides are stained with 5 % Giemsa for subsequent scoring of chromosomal aberrations. All slides, including those of the positive and negative controls, shall be independently coded before microscopic analysis for chromosomal aberrations. At least 300 well-spread metaphases shall be scored per concentration and control, sampling from the replicates (e.g. 100 cells per replicate in the case of triplicate cultures, if possible).

The percentage of cells with structural chromosomal aberration(s) shall be evaluated. Chromatid and chromosome-type aberrations classified by sub-types (breaks, exchanges) shall 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. Either the percentage of polyploidy or endo-reduplicated cells, or both, is reported when seen.

B.4.2.3 Specific methods for CA in human lymphocytes

Human lymphocytes should be sampled according to human ethical principles in the area of use.

Human peripheral blood lymphocytes should be obtained from young (approximately 18 to 35 years of age; if the background incidence is low, donors of more than 35 years of age can be used), 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. Freshly drawn whole blood samples are used for testing. The culture medium should contain a mitogen, such as phytohemagglutinin (PHA). Cells in cell culture dishes are exposed to the polar (cell culture medium) and non-polar test medical device extracts (or the test sample) with the addition of serum, both with and without an exogenous source of metabolic activation for short and long incubations, as described in B.3.2.3. After the short treatments, cells are washed with fresh cell culture medium and grown for the expression period of approximately 1,5 doubling times. For the long treatment, cells are incubated continuously with extracts in the absence of metabolic activation for approximately 1,5 cell divisions. Cytotoxicity is determined by comparison of the mitotic index (MI) of treated cultures with that of the vehicle control. If the test sample is toxic, a dose-response including one sample near the toxicity limit of (55 ± 5) % should be included among the samples. 1 h to 3 h prior to harvest, colcemid or colchicine is added to the cells. At harvest, cells are treated with a hypotonic solution and recovered by centrifugation. Cultures are then washed with fixative and the cell suspension dropped onto microscope slides. Slides are stained with Giemsa for subsequent scoring of chromosomal aberrations. All slides, including those of the positive and negative controls, shall be independently coded before microscopic analysis for chromosomal aberrations. When single cultures are used per concentration, at least 300 well spread metaphases should be scored in this single culture. The 300 cells should be equally divided among the replicates, when replicate cultures are used. 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 (e.g. 50 cells for the positive controls).

The percentage of cells with structural chromosomal aberration(s) shall be evaluated. Chromatid and chromosome-type aberrations classified by sub-types (breaks, exchanges) shall 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. Either the percentage of polyploidy or endo-reduplicated cells, or both, is reported when seen.

B.4.2.4 Evaluation of the results

The following criteria should be met before determining the overall results for a specific test chemical:

1. a) testing of three experimental conditions (short treatment with and without metabolic activation, long treatment without metabolic activation) unless one resulted in positive results;
2. b) adequate number of cells and analysable concentrations (see OECD 473:2016, paragraphs 21 and 31);
3. c) criteria for the selection of top concentration that are consistent with those described in B.3.2.2 (see OECD 473:2016, paragraphs 22-24 for additional information).

The test shall meet the assay acceptance criteria for the controls as described below: concurrent negative control is considered acceptable for addition to the laboratory historical negative control database; concurrent positive control should induce responses that are compatible with those generated in the historical positive control data base and produce a statistically significant increase compared with the concurrent negative control.

The test sample is considered positive if at least one of the test concentrations exhibits a statistically significant increase compared with the concurrent negative control; any of the results are outside the distribution of the historical negative control data (e.g. Poisson based 95 % control limits; see OECD 473:2016, paragraph 39); and the increase is concentration related (e.g. using a trend test). Generally, for such test, a concentration related increase is defined as a statistically significant increase in two consecutive test concentrations.

NOTE

For more details on evaluation of results, see OECD 473.

B.4.3.1 Principle

The in vitro micronucleus test detects micronuclei (MN) in the cytoplasm of interphase cells. MN may originate from acentric chromosome fragments (i.e. lacking a centromere), or whole chromosomes that are unable to migrate to the poles during the anaphase stage of cell division. The MNvit test provides a basis for investigating chromosome-damaging potential in vitro that detects both aneugens and clastogens. For detection of MN, it is necessary that cells have undergone cell division during or after exposure to the test chemical. MN represents damage that has been transmitted to daughter cells, whereas chromosome aberrations scored in metaphase cells may not be transmitted. OECD 487 allows the use of protocols with and without the actin polymerization inhibitor cytochalasin B (cytoB). The addition of cytoB prior to mitosis results in cells that are binucleate and therefore allows for the identification of cells that have completed cell division. However, OECD 487 also allows for the use of protocols without cytokinesis block, if evidence is provided that the cell population has undergone mitosis.

B.4.3.2 MNvit-specific methods

A variety of common cell lines, such as CHO or primary cell cultures including human or other mammalian peripheral blood lymphocytes, may be used, with cell culture extracts chosen as recommended for the selected cell line. The method is the same as that described for the CA assay (see B.4.2). After exposure, washing and expression, the cultures are harvested (colcemid or colchicine is not added), stained and dropped onto slides. Micronucleus frequencies shall be analysed in at least 2 000 binucleate cells per test concentration and control. Flow cytometric methods allow the assessment of many more cells.

The following criteria should be met before determining the overall results for a specific test chemical:

1. a) testing of three experimental conditions (short treatment with and without metabolic activation, long treatment without metabolic activation) unless one resulted in positive results;
2. b) adequate number of cells and analysable concentrations (see OECD 487:2016, paragraphs 28 and 44 to 46);
3. c) criteria for the selection of top concentration that are consistent with those described in B.3.2.2 (see OECD 487:2016, paragraphs 24 to 31 for additional information).

The test shall meet the assay acceptance criteria for the controls as described below:

The test sample is considered positive if:

Generally, for such test, a concentration related increase is defined as a statistically significant increase in two consecutive test concentrations.

NOTE

For more details on MNvit-specific methods, see OECD 487.

B.5.1 General

In vivo genotoxicity testing is not part of the standard genotoxicity test battery and should be reserved for rare situations (see ISO 10993-1).

NOTE

The in vivo assays describe testing into a cytotoxic range or to a maximum dose for a valid test that can be challenging for medical device assessment.

Methods evaluating medical devices using in vivo mammalian genotoxicity tests should be according to:

1. a) OECD 474^[24^] for mammalian erythrocyte micronucleus test;
2. b) OECD 475^[25^] for mammalian bone marrow chromosomal aberration test;
3. c) OECD 488^[27^] for transgenic rodent somatic and germ cell gene mutation assays;
4. d) OECD 489^[28^] for in vivo mammalian alkaline comet assay.

B.5.2.1 Vehicles

The vehicles shall be selected in accordance with ISO 10993-12 and shall not produce toxic effects at the dose levels used. Well-known vehicles include water, physiological saline, methylcellulose solution, PEG and freshly refined vegetable oil (e.g. cottonseed, sesame oil) of the quality defined in various pharmacopoeias. If other vehicles are used, their inclusion shall be justified and supported with reference data indicating their compatibility.

B.5.2.2 Routes of administration and doses levels

Saline solutions can be injected intravenously and other vehicles or extracts should be injected intraperitoneally at a maximum dose according to ISO 10993-11^[2^]. Other routes of exposure may be acceptable where justified and the extract is shown to be taken up by this route. The maximum volume of liquid that can be administered by gavage or injection at one time depends on the size of the test animal. Except for irritating or corrosive substances which will normally induce exacerbated effects with higher concentrations, variability in test volume should be minimized by adjusting the concentration to ensure a constant volume at all dose levels. The use of volumes different than those mentioned above shall be justified.

C.1 General

Many medical device products have been used in humans for decades without evidence of implant-related tumorigenesis. International Agency for Research on Cancer (IARC) classifies many products as group 3, unclassifiable, due to a lack of evidence, see WHO IARC Monographs on identification of carcinogenic harms to human. However, it is not understood whether the risk from these products can be identified via current chemical carcinogenicity test methods or other carcinogenicity evaluation methods.

Carcinogenicity tests are generally performed using long term (lifetime) bioassays in experimental animals (rodents). Shorter term, transgenic assays, such as mouse rasH2, have also been chosen as more appropriate for medical devices, because the tumour development time occurs prior to the development of tumours caused by solid state carcinogenesis^[39^]. However, the limitations of the carcinogenicity bioassay have been increasingly noted.^[40^],[41^],[42^],[44^],[45^] These include high costs, large numbers of animals and difficulty in interpreting results due to high and variable background incidence of certain tumours in the test species. More importantly, the ability of the bioassay to accurately predict human carcinogens has been called into question.

OECD Guidance Document 116^[13^] and OECD 451^[22^] on carcinogenicity testing require several hundred animals for a study (at least 50 animals per sex per dose level with a concurrent control). In addition, the high dose should be sufficient to cause either mild toxicity or a systemic exposure that is well in excess of clinical exposure (i.e. 100 to 150-fold) (see OECD Guidance Document 116^[13^]).

Given these requirements, carcinogenicity testing of medical devices requires implantation of many medical devices or the application of medical device extracts. Implantation has a limited ability to exaggerate the clinical dose, especially in mice, and can still be unable to provide true clinical exposure (e.g. using subcutaneous implantation for an intravascular medical device). Use of medical device extracts requires an extraction that yields a sufficient amount of the extractables that reflect the spectrum of extractables with respect to polarity. Often, multiple extraction vehicles are required.

In either situation, it is exceedingly difficult to exaggerate the dose of agent to the extent applicable with tests of chemical agents or drugs, where dose exaggeration is necessary for a valid test.

C.2.1 Rodent carcinogenicity test

There are no OECD guidelines for conducting carcinogenicity tests on medical devices. Available guidelines on carcinogenicity testing indicate that a well conducted rodent bioassay requires the administration of at least three exposure levels (see Reference [22]) to a sufficient number of animals so that at least 20 to 30 of each sex at each exposure level survives to the end of the study^[46^]. Implantation of medical devices should follow expected human use, if feasible. Requirements for exaggerated dosing can result in implantation of large numbers of medical devices that can stress the animal, confounding study interpretation. The dosing and route of administration of medical device extracts should be justified. The duration of studies is generally for two years, or the majority of the lifespan of the animal. OECD 451^[22^] provides guidance on duration, termination of the study and survival; further guidance, including consideration of the acceptability of a negative carcinogenicity is found in OECD Guidance Document 116^[13^].

C.2.2 Transgenic animal carcinogenicity tests

Shorter term, transgenic assays (e.g. mouse rasH2) have been recommended as more appropriate for medical devices, because the tumour development time occurs sooner, generally within six months, ostensibly prior to the development of tumours caused by solid state carcinogenesis^[39^]. However, the experiments of Palazzi included negative implant controls, but no positive implant controls. The lack of positive medical device controls makes interpretation of negative results uncertain. However, the technical limitations described in Clause C.1 and C.2.1 (e.g. number of animals, route of administration of the test sample) also apply to the transgenic animal assays.

C.2.3 Combination of chronic toxicity and carcinogenicity assays

It is possible for tests to be designed to examine simultaneously in a single study both chronic toxicity and carcinogenicity, but the same issues apply (see C.2.1). In addition, substantial preliminary dose-response studies shall be performed prior to the test (OECD 453^[23^]) to ensure the dose groups are appropriate. This helps to prevent or minimize premature mortality from chronic or cumulative systemic toxicity compromising the statistical evaluation of data derived from animals surviving to the end of the study period (i.e. normal lifespan). The attempt to combine the assessments of both chronic toxicity and carcinogenicity thus does not always result in any savings of time, test agents or animals over the performance of the two tests separately.

C.2.4 Test sample preparation

When carcinogenicity tests are necessary as part of an evaluation of biological safety, the medical device should be tested in a form representative of the finished medical device state. Use of either materials, defined chemicals or characterized extracts of medical devices may be considered.

Additional testing can be warranted for additional states of the medical device such as, wear debris generated from the medical device or materials that cure in situ (e.g. cements, adhesives, pre-polymer mixtures). For guidance on in situ curing medical devices, see ISO 10993-12. The selection of the test sample should be justified and documented. For implantation studies to evaluate carcinogenicity, the amount of material implanted should be representative of an exaggerated human dose which provides a conservative approach (e.g. 100-fold). The highest dose will be limited by the physical constraints of the animal model.

C.2.5 Test methods

If testing of an extract is considered necessary, the carcinogenicity tests should be performed according to OECD 451^[22^] or OECD 453^[23^], and OECD Guidance Document 116^[13^]. Tissues evaluated should include relevant tissues from the list indicated in OECD 451^[22^] or OECD 453^[23^], as well as the implantation site and adjacent tissues.

C.3 Foreign body carcinogenesis

Tumours induced by implants are well known in experiments using rats and this phenomenon is called “foreign body carcinogenesis” or “solid state carcinogenesis.” The phenomenon is summarized as follows.

Tumours usually develop around or near an implant with a frequency that is dependent on several factors such as:

The same material that produces tumours as a film or sheet will, for the most part, produce fewer or no tumours when implanted as a powder, a thread or a porous material. See Reference [67].

Tumours due to this phenomenon first begin to be detected after 8 months to 9 months following implantation. The incidence continues to increase after this latency period.

On the other hand, many reports indicate a difference in the incidence of tumour formation among different materials of similar shape and size using the same experimental animal protocol.

Mechanistic understandings were summarized in an IARC Monograph (see References [67] and [68]).

C.4 Animal welfare considerations

The conduct of carcinogenicity studies by implantation requires surgically invasive methods on a large number of both test and control (sham operated) animals. Alternatives should be considered according to the three R-principles and ISO 10993-2.

D.1 General

Where the release of nanomaterials (NM) from the medical device is demonstrated, the genotoxicity of the nanomaterials should be evaluated. Information related to the release should be provided including type of media, dispersion method, surfactant use, duration of treatment, test system and dosimetry, etc., before genotoxicity testing (see ISO/TR 10993-22^[9^]).

D.2 Genotoxicity assessment

The presence of serum or proteins in the medium can impact the extent of particles internalized in the cells. Characterization of NMs should occur in the test media (as well as in other media, as appropriate). Consideration should be given to potential artefacts occurring with NMs in the test media (e.g. agglomeration).

Recommended approach for assessing genotoxicity of NMs:

1. a) the use of carefully defined referenced NMs, including characterization of these NMs in the test medium;
2. b) assessment of uptake of NMs and their distribution within cells and in vivo systems (see ISO/TR 10993-22^[9^]);
3. c) dose ranges carefully chosen to avoid artefacts related to system overload;
4. d) a modified test battery^[76^] with two in vitro mammalian mutagenicity assays (replaces bacterial mutation assays):
   • 1) mouse lymphoma [L5178Y (TK+/-) assay (MLA)] (see OECD 490) or HPRT gene mutation assay (HPRT) (see OECD 476^[84^]); and
   • 2) in vitro chromosomal aberration assay (see OECD 473) or in vitro MN assay with assay modification as described earlier (see OECD 487) or in vitro MLA (see OECD 490);

     NOTE 1

     Given that NMs generally do not require metabolic activation, there is no compelling reason to perform an in vitro assay with S9 and there is no reason for default testing in vivo. Thus, NMs are generally tested in the absence of S9, unless composed of organic materials or agents likely to be affected by mammalian metabolism.

     NOTE 2

     As bacterial assays such as S. typhimurium and E. coli tester strains have limited uptake of nanoparticles and also lack the capability for mammalian-specific responses to nanomaterials, they therefore are considered to be inappropriate for evaluation of the risk.

     NOTE 3

     The mouse lymphoma [L5178Y (TK+/-) assay (MLA)] alone is insufficient for a comprehensive genotoxicity assessment of nanomaterials.

5. e) adherence to tests and cell systems described in ISO/TR 10993-22^[9^] and Reference [64]; and
6. f) effort on understanding mechanisms of genotoxicity.

Animal testing should only be considered if there is evidence with supporting literature or test data of specific accumulation of nanoparticles in organs, and then the relevant target organ and tissue should be tested [e.g. in vivo comet assay (see OECD 489^[28^]), transgenic rodent mutation assay (see OECD 488^[27^]), in vivo erythrocyte micronucleus test (see OECD 474^[24^]), in vivo bone marrow chromosomal aberration test (see OECD 475^[25^])]. The choice of in vivo test methods shall be justified and documented.

Annex E Recipe for S9 mix

Traditionally, the recipe was different for bacterial assays and for mammalian cells. There was independent development of bacterial and mammalian cell tests, but both recipes result in NADPH generating systems. S9 mix is now used with bacterial and in vitro mammalian tests. An example of a S9 mix recipe is provided in Table E.1. Other recipes can be used if validated.

The used S9 mix concentration, including manufacturer and quality assessment details, should be described in the test report.

F.1 General

An endocrine disruptor (ED) is an exogenous substance or mixture that alters function(s) of the endocrine system and consequently causes adverse health effects in an intact organism, or its progeny, or (sub)populations (IPCS/WHO, 2002^[65^]). Substances identified as EDs for human health should be evaluated for reproductive toxicity and developmental toxicity (see 7.1).

EDs should be identified using established lists including the ED list from the European Commission2

2 The ED list from the European commission is available at https://edlists.org/the-ed-lists/list-i-substances-identified-as-endocrine-disruptors-by-the-eu.

3 The ECHA list is available at https://echa.europa.eu/ed-assessment.

Evaluation shall be performed in a toxicological risk assessment in accordance with ISO 10993-17 based on endocrine disruption data. When risk from an endocrine disruptor is unacceptable and controlling the risk is not practical, the risk can be addressed by conducting a benefit-risk analysis according to ISO 14971.

F.2 Evaluation of endocrine disruptors

If any substances in or on the medical device are identified as known endocrine disruptors, the toxicological risk assessment should consider reproductive toxicity and developmental toxicity for that substance according to ISO 10993-17.

When screened against the lists above, if any of the extractable compounds are identified as known endocrine disruptors [e.g. dibutyl phthalate (CAS RN 84-74-2)], in addition to the systemic toxicity endpoints, the point of departure (POD) in the toxicological risk assessment should consider the possible developmental or reproductive endpoint of concern, and the most conservative POD should be selected.

When a medical device constituent is not on a list of EDs, the specific constituent should be assessed according to ISO 10993-17.

G.1 General

Developmental and reproductive toxicity (DART) refers to the adverse effects that certain substances can have on the reproductive system and the development of an organism, particularly during pregnancy. These effects can occur in both humans and animals and can involve various biological processes. It is an important area of concern in toxicology, as exposure to certain chemicals or agents can have long-lasting and serious consequences for individuals and populations.

Reproductive toxicity involves the adverse effects of substances on the reproductive system, including male and female reproductive organs and processes. Reproductive toxicity can lead to changes in fertility, impaired reproductive function and adverse outcomes in pregnancy such as miscarriage.

Developmental toxicity refers to the adverse effects on the developing organism, usually during embryonic or foetal development. Developmental toxicity can result in structural abnormalities, functional deficits or other adverse outcomes in the developing foetus. Exposures during critical periods of development can have long-term consequences.

G.2 Biological test methods

If biological test methods are to be used, the study design and logistics should be discussed with a testing laboratory that has experience with DART studies. Table G.1 provides a summary of some common OECD study designs used to address DART endpoints. The summary is included as a reference to aid in understanding of study designs (e.g. developmental toxicity versus reproductive toxicity) and to facilitate discussions with the test laboratory.

NOTE

Standard DART study designs are not validated for mixtures (e.g. medical device extracts), or intraperitoneal and intravenous dosing methodologies. There can be situations where a DART study can be adapted for an implantable medical device (e.g. breast implants, hydrogels to treat intrauterine adhesions, surgical meshes or sutures); however, the dosing methodology would be specific to the intended use of the medical device and the study design would be highly customized to address the endpoints of potential concern.

Although in vivo approaches are not validated for mixtures such as medical device extracts, there are available in vitro approaches, such as the mouse embryo assay (MEA) that are currently applied to lot release of assisted reproduction technology (ART) devices. MEA is used to assess the potential for embryotoxicity of medical devices in contact with gametes or embryos. MEA assesses blastocyst development from either one-cell or two-cell stage embryos, and medical devices can be tested as a liquid without extraction or dilution or extracted. Acceptance criteria are based on the percentage of embryos that develop to expanded blastocyst (see Reference [71]).

G.3 Identification of reproductive and developmental toxicants

The following list can be used to identify whether a specific chemical substance can be classified as a reproductive and developmental toxicant:

Due to the biological complexity of DART endpoints, as well as limited high-quality data to support model development, no single computational approach is currently available to predict reproductive toxicity and developmental toxicity, including endocrine disruption. There are, however, computational tools available that can lend support for these endpoints and provide additional information in a weight-of-evidence argument.

There are numerous test protocols for assessing reproductive toxicity and developmental toxicity. Reference [63] provides an overview of reproductive and developmental test considerations for pharmaceuticals; however, most of the chemicals extracted from medical devices are chemicals that have been tested for reproductive toxicity and developmental toxicity under OECD test guidelines rather than ICH. Therefore, Table G.1 provides an overview of some common OECD test guidelines for the testing of reproductive toxicity and developmental toxicity in mammalian species. Understanding the general test method should be appropriate for identifying relevant PODs for reproductive toxicity and developmental toxicity in the toxicological risk assessment. The selected PODs should be justified and documented.

Bibliography