I. Introduction
1. Moving away from animal-based toxicity testing
Currently, around 100,000 substancesFootnote 1 are produced globally in volumes of up to 400 million tonnes, making the chemicals industry one of the largest industries in the world.Footnote 2 Global chemicals production and use are expected to double by 2030 compared to 2017.Footnote 3 Chemical substances exist in various forms, such as basic chemicals (eg, petrochemicals), speciality chemicals derived from basic chemicals (eg, plastics, electronic chemicals, catalysts, adhesives etc.) or as components in products like medicines, pesticides, biocides, fertilisers and household products (eg, detergents, soaps, dyes).Footnote 4
Chemicals can be detrimental to human health (eg, negative impact on the reproductive system, carcinogenic or endocrine disrupting properties), and the environment (eg, contamination of soil and water, resulting in reproductive, immunological, and neurological damage or death to wildlife, biodiversity loss and ozone layer damage).Footnote 5 Therefore, the production, use and marketing of chemical substances and consumer products are subject to quality standardsFootnote 6 and safety requirementsFootnote 7 outlined by various national and international laws, regulations, and guidelines. These regulatory requirements typically involve conducting experimental studies using animals,Footnote 8 non-animal methods, or a combination of both. Animal experiments remain integral to safety testing, contributing significantly to the ongoing prevalence of animal testing in Europe: According to current statistics, out of nearly 8 million animals used for scientific purposesFootnote 9 in the European Union and Norway in 2020,Footnote 10 approximately 1.4 million were used for regulatory purposes.Footnote 11
This situation contrasts with the clear intent of the European Parliament to transition away from animal testingFootnote 12 and with the 3Rs principle of Replacement, Reduction and RefinementFootnote 13 enshrined in the European legislation on animals used for scientific purposes.Footnote 14 This also applies to testing for regulatory purposes.Footnote 15 However, there is a persistent reliance on animal testing in the current regulatory testing systems, largely reflecting how (industrialised) societies aim to minimise potential risks to human health and the environment by adhering to the well-established practice of animal testing. This results in risk-averse legislation and application of the law, particularly concerning non-animal methods and new technologies.Footnote 16 Animal models became the established standard for regulatory safety testing following significant health crises in the twentieth century, such as the sulfanilamide elixir poisonings in the 1930s, the discovery of thalidomide causing widespread birth defects in the 1960s, and the health risks associated with diethylstilbestrol use until 1971.Footnote 17 In response, governments around the world enacted laws mandating comprehensive animal testing before new substances could be marketed.Footnote 18
Scientists argue that today’s regulatory systems represent an emotional rather than scientifically grounded response to past health crises.Footnote 19 These systems are founded on the belief that substantial advancements in medical knowledge throughout the twentieth century were made possible primarily through the use of animal models.Footnote 20 Consequently, there is a prevailing view that ensuring public safety requires extensive animal testing of substances.Footnote 21 Decades-old testing procedures to evaluate substance toxicity were standardised and incorporated into international guidelines to facilitate production and international trade. As a result, animal models became the gold standard, without ever undergoing comprehensive quality control or validation,Footnote 22 making it very difficult to move away from animal testing and integrate new scientific advancements in toxicity testing.Footnote 23
The continued adherence to an unsustainable animal-based testing paradigm despite the availability of scientifically sound and potentially better non-animal test methods is increasingly criticised in the scientific community.Footnote 24 Not only does this practice contradict the requirements and objectives formulated by legislators, but it also ignores the issue of poor repeatabilityFootnote 25 and translatabilityFootnote 26 of animal test data: Often, toxicity testing outcomes cannot be reproduced between animal species or sexes, or compared to control animals used in previous toxicity testing for the same substance.Footnote 27 An analysis of animal testing to predict the toxicity of pharmaceuticals in humans, involving 2,366 different drugs, concluded that the predictive value of animal tests in terms of determining whether a substance will be toxic in humans was “barely greater than that which would result merely by chance.”Footnote 28 The referenced study noted that “the fact that a compound shows no toxic effects in animals provides essentially no insight into whether the compound will also show no toxic effects in humans’.Footnote 29 In a subsequent study, the same authors found that even preclinical tests in monkeys failed to predict toxicity in humans reliably and concluded that the data “suggest strongly that a lack of toxicity in any species cannot be reliably used to imply a probable lack of toxicity in any other species.”Footnote 30 Toxicologist Thomas Hartung poignantly sums up the issues with translatability between species due to the differences in physiology, life span, and exposure to environmental factors: “We are not 70kg rats.”Footnote 31
Consequently, animal testing cannot reliably predict whether a substance will be harmful to humans. In fact, some substances that have been proven safe for animals have resulted in toxic outcomes and sometimes death for humans.Footnote 32 At the same time, it is assumed that many substances that had toxic effects in animals and were therefore discontinued before reaching the clinical testing phase in humans would have been effective and beneficial for humans.Footnote 33 It has been suggested in the literature that certain widely used drugs, such as paracetamol, aspirin and penicillin, known to be toxic or lethal to various animal species, would not have been developed and made available for human use if they had undergone animal testing.Footnote 34 The process of extrapolating results from animals to humans is also highly influenced by policy considerations and it can have varying outcomes depending on the practice of the responsible agency and specific legislative requirements (eg, requirement to balance risk against benefit or to factor in economic considerations), which can lead to diverse safety assessments.Footnote 35
Despite these demonstrable problems with animal-based toxicity testing, the most important legislation for assessing and managing chemicals in Europe,Footnote 36 the Regulation on Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH),Footnote 37 along with the Regulation on Classification, Labelling and Packaging of Substances and Mixtures (CLP Regulation),Footnote 38 is largely based on an animal testing paradigm. REACH, in force since 2007, aims to ensure a high level of protection for human health and the environment from risks posed by chemicals, and to promote the use of alternative methods for assessing substance hazards to reduce animal testing. REACH applies to all chemical substances, from those in our daily lives to industrial chemicals, requiring companies to gather information on the properties and uses of substances they manufacture or import in quantities of one tonne or more per year and to assess and manage associated risks. For this purpose, REACH contains standard information requirements, which are largely based on standard (animal) tests to assess substances in terms of specific human health and environmental effects.Footnote 39, Footnote 40
While REACH is not the sole EU legislation regulating chemicals,Footnote 41 there appears to be more stagnation within the REACH framework in terms of transitioning away from animal testing and adopting advanced non-animal safety assessment methods compared to other chemical safety legislations, such as those governing food safety or cosmetics.Footnote 42
Because of the thousands of chemicals on the market with either lacking or incomplete toxicity data, toxicology experts see a particular challenge in applying a traditional animal-centric approach to meet REACH requirements. They worry that applying this approach to the existing chemicals and those expected to enter the market in the future will take too much time and resources.Footnote 43 In addition, six of the most common animal tests provided for in the OECD guidelines for chemical safety assessments have been shown to produce different outcomes in repeat experiments.Footnote 44 Numerous analyses of traditional animal models used to assess systemic toxicityFootnote 45 have indicated that animal testing in this area is largely unreliable.Footnote 46 There is also growing concern about the suitability of animal testing for assessing the safety of new substances, such as nanomaterials,Footnote 47 the potentially harmful effects of combined exposure to multiple substances (mixture risk assessment, MRA),Footnote 48 and providing solutions to emerging health questions, eg, the effect of chemicals on neurodevelopment.Footnote 49
For these reasons, scientists are increasingly calling for a new system based on so-called New Approach Methodologies (NAMs) that allow for more human-relevant assessments of the adverse effects of chemicals on human health and the environment, using fewer or no animals.Footnote 50, Footnote 51 Despite this, REACH remains highly rigid and unadapted to new (non-animal) methodologies, with regulatory authorities still reluctant to accept safety assessments that have not been conducted using animals,Footnote 52 maintaining animal testing as the gold standard. Non-animal test methods are still measured against this standard,Footnote 53 thereby limiting the potential for the development and validation of new human-relevant non-animal models and methods.Footnote 54
Keeping this situation in mind, the present study provides an overview of the regulation of chemical safety within the scope of REACH as an important example of the challenges to implementing the 3Rs in regulatory safety testing, shedding light on the REACH-specific challenges to the increased implementation of NAMs. The significance of REACH cannot be overstated as it governs chemical safety across Europe. Additionally, because it regulates the import of substances and products from non-EU countries into the EU, it holds crucial importance for countries like Switzerland, which primarily export to the EU and are thus directly affected by EU legislation.Footnote 55
2. Current situation: numbers on the rise for compliance with REACH despite clear intent to reduce and replace
The European Commission initially estimated that 2.6 million animals would be needed to comply with REACH after its introduction.Footnote 56 However, the number of animals used so far has already doubled,Footnote 57 and this figure is expected to keep increasing. For one, this is due to ongoing compliance checks of already registered substances by the European Chemicals Agency (ECHA), which is responsible for overseeing compliance with REACH. Theses checks can result in requests by the Agency to conduct animal testing for registrations deemed non-compliant.Footnote 58 Furthermore, it is expected that recent amendments to REACH, as well as the CLP Regulation, will lead to more animal testing: An amendment to Annexes VI to X of REACH,Footnote 59 which came into force in 2022, requires several additional animal tests for substances that have either been tested in vitro and the results have come out positive or that are otherwise of a specific concern.Footnote 60 It is expected that this amendment will cost millions of animal lives for substances that have already been registered.Footnote 61 A 2023 amendment to the CLP regulationFootnote 62 included several new hazard classes relating to endocrine disruption in humans and the environment, with animal testing being the main method of testing for these endpoints, which is also expected to cost additional animal lives.Footnote 63 It also expected that the EU Commission plan to revise REACH as part of its Chemicals Strategy for Sustainability with the aim of tackling information gaps to keep humans and the environment safe from potential hazards will also entail more animal testing if carried out as planned.Footnote 64
At the same time, the purpose of REACH is not just to guarantee a high level of protection for human health and the environment, but also to promote alternative methods for assessing the hazards of substances. Testing on animalsFootnote 65 to meet the REACH requirements is only permitted as a last resort.Footnote 66 These principles and requirements are also in line with the objectives of Directive 2010/63/EU on the protection of animals used for scientific purposes.Footnote 67
The significant disparity between these ambitions and legal goals to move away from animal testing and the actual increase in animal testing is especially pertinent considering that non-animal methods have advanced to the extent that defaulting to animal testing is no longer deemed necessary. A new, more reliable, and efficient regulatory system based on the routine use of non-animal methods is seen as feasible.Footnote 68 Hence, the current regulatory system is inadequate, which appears partly due to issues in the formulation and design of the relevant legal bases as well as other factors, warranting further investigation.Footnote 69
The ECHA has faced criticism for showing a significant bias towards animal testing.Footnote 70 This perspective is supported by official sources: In 2021, the European Parliament adopted a resolution urging the development of an EU-wide action plan to “accelerate the shift to innovative methods that do not involve animal use in research, regulatory testing, and education.”Footnote 71 The resolution expresses concern that the number of animals used in experiments has not substantially decreased since Directive 2010/63/EU came into effect. It also highlights bureaucratic hurdles in accepting non-animal testing methods, inadequate enforcement of their use, and insufficient funding for their development.Footnote 72 Furthermore, it calls on the European Commission to set reduction goals in view of Article 13 of the REACH regulation, which mandates the revision of test methods when non-animal alternatives become availableFootnote 73 and specifically urges the European Chemicals Agency (ECHA) to formulate a replacement and reduction strategy.Footnote 74
Following a European Citizen’s Initiative (ECI) to strengthen the ban on animal testing for cosmetics in January 2023,Footnote 75 the European Commission committed to launching a roadmap for phasing out animal testing in chemical safety assessments.Footnote 76 So far, there have been two workshops on the roadmap involving EU member states and stakeholders from research, industries, and animal welfare organisationsFootnote 77 as well as a public consultation.Footnote 78 Furthermore, in its strategic plan for 2023–2026, the ECHA included a commitment to “enhance cooperation with the European Commission, other institutional partners, the scientific community and stakeholders to support the development of a roadmap towards full replacement of animal testing”.Footnote 79 The effectiveness of these measures in reducing animal testing to meet regulatory requirements, specifically for REACH compliance, remains to be seen.
II. The link between natural sciences and law with regard to chemical safety: regulatory toxicology and the implementation of the 3Rs
1. What is regulatory toxicology?
Regulatory toxicologyFootnote 80 describes the legally mandated collection and use of information on potentially hazardous properties of substances and safe levels of human and environmental exposure to these hazards. This information is used to make decisions about risk and exposure management, hazard communication through classification and labelling, and other regulatory actions, such as imposing restrictions or bans on dangerous substances. The aim is to assess the risks and set management standards to prevent human and environmental health hazards from exposure to all kinds of substances, such as industrial chemicals, pharmaceuticals, foods, cosmetics, biocides, plant protection products, etc.Footnote 81
Regulatory toxicology links private industries, the natural sciences, and the law. Regulatory agencies are tasked with managing or carrying out the technical, scientific, and administrative aspects of the relevant regulations.Footnote 82 They base their decisions on toxicological evaluations of substances by private industries, scientific institutions, and interdisciplinary committees with toxicological expertise. Both the relevant legislation and the decisions and actions taken by the regulatory agencies are informed by scientific as well as non-scientific criteria. Scientific criteria include everything to do with the physicochemical, toxicological, and ecotoxicological properties of the substances, while non-scientific aspects include considerations regarding costs, compatibility of actions with intersecting laws, liability issues, risk vs. benefits of substances with the potential to harm humans or the environment, public opinion etc.Footnote 83
Due to globalisation, substances are traded across national and international jurisdictions. This necessitates standardisation of chemical safety evaluation and harmonisation of legal requirements to promote international acceptance of national or regional regulatory decisions, and ultimately to guarantee safe exposure to chemicals no matter where they are used. For this purpose, numerous international programs, collaborations, and organisations have been established and many international agreements have been adopted.Footnote 84
Depending on the regulated area and substances in question, different laws can apply and intersect, potentially resulting in repeated testing for the same substance due to varying protection objectives.Footnote 85, Footnote 86 In its 2020 Chemical Strategy for Sustainability, the European Commission vowed to address the inconsistencies and lack of transparency and harmonisation of chemical safety assessments in the current regulatory landscape, explicitly mentioning the goal of avoiding the duplication of work.Footnote 87 It is yet to be seen if the corresponding legislative proposals put forward by the European CommissionFootnote 88 – and recently adopted by the European Council for negotiation with the European ParliamentFootnote 89 – will help reduce animal testing and avoid repeat testing.
Depending on the specific safety needs of the regulatory area, the relevant legislation contains provisions on how to meet these needs.Footnote 90 For instance, REACH, the regulation at the centre of this study, mandates the collection of data on the physical, toxicological, and ecotoxicological properties of substances as a basis for hazard and risk assessment to establish how to safely use the substance in question. The necessary information for this purpose is generally acquired by studying existing data and performing tests with animal models, non-animal methods, or a combination of both.Footnote 91
The testing required to assess substance hazards and the risks from exposure to hazards is largely conducted on animals, the core belief being that animal models serve as surrogates for the human organism, allowing an examination of the reactions of the organism to substances or how substances affect specific organs.Footnote 92 However, as explained above, this animal testing paradigm is being increasingly called into question for various reasons,Footnote 93 one fundamental one being that scientific developments have brought forward a wide array of new methodologies that can replace current animal models or that produce more reliable human-relevant information without even having to mimic the responses of animal test subjects on a method-by-method basis.Footnote 94
In its latest report on global chemicals, the United Nations Environment Programme highlights the 3Rs principle – replacing, reducing, and refining animal testing, first introduced by scientists William Russell and Rex Burch in 1959Footnote 95 – as an international priority in the assessment of chemical hazards.Footnote 96 It traces the beginning of this prioritisation of the 3Rs in toxicology across countries to a landmark 2007 report by the United States National Academy of Sciences,Footnote 97 which provides potential ways forward for more efficient toxicity testing based on modern non-animal methods.
Despite these developments and explicit legal requirements to test with means other than animals wherever possible and to use animals as a last resort only,Footnote 98 many scientifically sound and available new approach methodologies (NAMs) face challenges to their implementation. Various reasons have been identified for this impasse, including inflexible legislation and a test method validation system – the process of verifying that methods perform as intendedFootnote 99 – which has proven inadequate to accommodate new non-animal methods in a timely manner for regulatory approval and acceptance.Footnote 100
2. Chemical safety assessment as an integral part of regulatory toxicology
Chemical safety assessmentsFootnote 101 are integral to regulatory toxicology, their purpose being to establish if exposure to a chemical through its use or presence in the environment can have adverse effectsFootnote 102 on human health or other organisms, to determine safe levels of exposure for humans and the environment,Footnote 103 and/or to establish risk management measures and operational conditions.Footnote 104 To this end, substances are generally assessed according to a multi-stage process consisting of a hazard identification,Footnote 105 a dose-response assessment and hazard characterisation,Footnote 106 an exposure assessment,Footnote 107 and a risk characterisation.Footnote 108, Footnote 109, Footnote 110
Note that a hazard stands for the innate properties of a substance and its potential to cause harm, while a risk represents a percentage expressing the likelihood of harm through exposure to the hazard.Footnote 111 It is also worth noting that “safe” is not a scientifically defined term. The safety of a substance for human health is still generally determined by extrapolating test results to real life, which depends on a number of extrapolations (ie, from animal to human; from high test dose to low dose; from average to individual human; from short term to long term exposure)Footnote 112 and is influenced by policy considerations. This can lead to varying outcomes depending on the assessment practice of the agency responsible and the legislative requirements,Footnote 113 resulting in diverse safety assessments even where repeat experiments produce the same results.Footnote 114
Within the scope of the chemical safety assessment, specific studies are performed to identify the hazards of substances depending on the nature and latencyFootnote 115 of their toxic effects.Footnote 116 The most common studies include acute toxicity testing, repeated dose toxicity testing, genotoxicity studies, and reproductive and developmental toxicity tests. Acute toxicity studies, for which there are now a range of internationally accepted non-animal methods, generally provide information on health hazards from short-term exposure and single-dose tests (eg, acute oral and dermal effects, or eye irritation and corrosion). Testing of a substance can also be performed over a longer period to assess the various risks associated with repeated use. This is called repeated dose toxicity testing, which is done by administering the test substance daily over a specified period. Genotoxicity testing is aimed at acquiring information on the potential of the test substance to interfere with DNA and cause cancer. Reproductive and developmental toxicity tests are performed to establish the effects of a substance on reproductive functions and the health of offspring.Footnote 117
Further studies aimed at addressing specific concerns include developmental neurotoxicity testing and studies on endocrine disruption. The former specifically address neurological effects of repeated exposure of offspring to substances in utero and after birth, while the latter are performed to assess hormonal disruption through exposure to the test substance.Footnote 118
3. New approach methodologies (NAMs) and the 3Rs in regulatory toxicology
(a) Definition and purpose
Within the scope of toxicity testing, new approach methodologies (NAMs) are generally understood as alternatives to animal-based testing or other methods, technologies, and approaches that can be used to obtain information about substance hazards and risks without using animals or in combination with animal-based tests.Footnote 119, Footnote 120 NAMs include method-by-method replicas of existing animal tests as well as methods that can deliver human-relevant information without requiring an entire organism.Footnote 121 Non-animal strategies can be experimentalFootnote 122 or non-experimentalFootnote 123 in nature and can consist of combinations of various non-animal and/or non-testing methods, strategies, and technologies such as integrated testing strategies (ITS) or integrated approaches to testing and assessment (IATA).Footnote 124
NAMs are generally divided into three categories: (1) one-on-one replacement, which describes a non-animal test that entirely replaces an animal test; (2) combining datasets, where a combination of methods and information sources are applied in the absence of a one-on-one replacement (eg, combining in vitro cell cultures with organs-on-a-chip and computational models); and (3) reformulated research questions, which are not geared towards an animal model but are instead broken down into separate mechanistic steps (eg, through the application of so-called adverse outcome pathwaysFootnote 125).Footnote 126 The various non-animal methods include in vitro methods (eg, using human cell cultures, tissues, organoids etc.), in chemico approaches (ie, chemical reactivity methods), in silico methods (ie, computational tools) and new technologies such as omics.Footnote 127, Footnote 128 Another way to acquire information, not technically a NAM but often used alongside NAMS to replace or reduce animal testing, is to use existing data. This data can originate from animal tests, human data, or NAMs.Footnote 129
The purpose of NAMs is to provide information on substance hazards and exposure in a way that complies with the principle of the 3Rs to replace, reduce, and refine animal testing, and to increase the efficiency of the current chemical safety testing system.Footnote 130 Article 13(1) REACH states that, in particular for human toxicity, information must be generated whenever possible by means other than vertebrate animal tests, thereby prioritising replacement. It also constitutes a legal basis for the use of NAMs as a standard for toxicity testing because the technology needed to obtain reliable information to assess the safety of substances is already available, and new methods and strategies are continuously being developed.Footnote 131
Despite this clear legal requirement, NAMs-based information still has a hard time being accepted within the scope of REACH. This has a lot to do with the fact that many NAMs don’t fit into the mould of the current animal-based paradigm of chemical safety assessments.Footnote 132 This is particularly true for NAMs for complex endpoints such as chronic toxicity, systemic toxicity, reproductive toxicity, and carcinogenicity: Unlike NAMs for less complex endpoints such as skin and eye irritation, they do not result in an apical outcome – ie, in an observable outcome in a whole organism – and thus don’t function as one-to-one replacements for animal models.Footnote 133 For example, it has been shown that the ECHA often rejects data generated without animal testing, unless NAMS are used that are in international guidelines, and requests the performance of (new) animal tests.Footnote 134 However, only a small number of NAMs for simple topical endpoints, such as eye and skin irritation, have been incorporated in the OECD TG and the EU Test Method Regulation so far.Footnote 135
According to Article 13(3) REACH, where testing is required to comply with the requirements of REACH, it must be conducted according to the test methods laid down in a Commission Regulation, ie, the EU Test Method Regulation,Footnote 136 or in accordance with other international test methods recognised by the Commission or the ECHA.
This raises the following questions: How do NAMs end up in international guidelines and why haven’t more NAMS been accepted in these guidelines?
(b) Validation of NAMs for regulatory acceptance
The validation of a non-animal test method according to internationally accepted principles and protocols generally facilitates the inclusion of a test method or approach in international test guidelines and regulations such as the OECD Guidelines for the Testing of Chemicals (OECD TG)Footnote 137 or the EU Test Method Regulation.Footnote 138, Footnote 139 Validation is a condition for the inclusion of NAMs in the OECD system of Mutual Acceptance of Data in the Assessment of Chemicals (MAD system), which is aimed at harmonising the multitude of national and regional regulatory requirements to promote acceptance of data across countries, and thus avoid duplication of chemical risk assessments.Footnote 140
Validation is generally understood as a process in which a test methodFootnote 141 is independently and transparently evaluated as to whether it is fit for purpose, in other words, if the test results are reliable – ie, reproducible in different laboratories according to a standardised protocol – and if the test method is relevant – ie, capable of measuring or predicting the sought-after biological effects.Footnote 142 A common feature of the different existing approaches to validation is a multi-step process that takes place after the development or update of a test method: first, a laboratory that is independent from the test developer assesses the test method in terms of its transferability, reproducibility, and relevance (prevalidation), and either proposes further optimisation of the test or, if the test method performance proves to be acceptable, refers it to full inter-laboratory reliability and relevance assessment (validation via ring-trial study).Footnote 143 The validation process as set out in the OECD Guidance Document 34 is based on the validation principles of internationally accepted validation bodies, such as the EU Reference Laboratory for alternatives to animal testing (EURL ECVAM)Footnote 144 and the US Interagency Coordinating Committee on the Validation of Alternative Methods (ICCVAM),Footnote 145 among others.Footnote 146
Even though this formal validation process has strongly contributed to the global harmonisation of test method acceptance,Footnote 147 it has proven to be a barrier to the timely validation of NAMs: The validation and uptake of NAMs in international test guidelines can take up to 20 years from the development of a NAM and has thus shown a strong need for a paradigm change in validation concepts.Footnote 148 The focus on single-method validation has been inadequate for assessing the validity of NAMs, particularly for complex endpoints such as reproductive toxicity, as it does not allow for new and more flexible concepts of validating NAMs with a performance-based approach.Footnote 149, Footnote 150
In that sense, during the most recent meeting of the International Cooperation on Alternative Test Methods (ICATM)Footnote 151 in 2023, representatives of the member validation organisations concluded that the OECD Guidance Document 34 is no longer suitable to accommodate current scientific developments, and is in need of revision. In particular, concern was expressed about the adequacy of multi-laboratory ring trials in assessing the transferability and performance of NAMs. Proposals were made to improve the validation procedure, such as a move away from the “gold standard” of the animal model towards human-relevant reference criteriaFootnote 152 and better coordination of peer reviews to avoid duplicate validation of methods.Footnote 153 At the same time, the OECD is currently calling for an “urgent mobilisation of national and regional resources to support the acceleration of new method validation for the safety testing of chemicals.”Footnote 154
III. The legal framework for chemical safety in the EU under REACH and the challenges to the implementation of NAMs for compliance with REACH
The acronym REACH stands for Registration, Evaluation, Authorisation, and Restriction of Chemicals. REACH was adopted in 2006 and came into force in 2007 in response to inadequate safety regulations for chemicals in the European Union. Prior to its inception, the regulation of chemicals was deemed inadequate due to its fragmentation, and the view prevailed that there was a general lack of information about the safety of existing substances.Footnote 155 Furthermore, the burden of proof to guarantee the safety of chemicals was on the regulators, ie, public authorities were tasked with gathering and evaluating information on substances from chemicals industries.Footnote 156 It was therefore decided that a new chemical safety legislation was needed and, in 1999, the European Commission was engaged to draft a document for a new chemicals strategy that would ensure that manufacturing companies and importers of chemicals into the European market would register important data on the substances with a centralized agency so that the data could be used to monitor and restrict substances of very high concern.Footnote 157 In 2000, the European Commission published a White Paper with a concrete REACH proposal. Following stakeholder debates and a public consultation process, the European Commission published the final proposal in 2003, which was subsequently debated in the European Parliament,Footnote 158 before the Regulation entered into force on 1 June 2007.
Simply put, REACH requires that substances are evaluated for specific human health and environmental effects, which are called endpoints (eg, skin corrosion, reproductive toxicity, short-term toxicity in fish, etc). To obtain information on these endpoints, REACH stipulates, among other things, the generation of information via testing on animals. The information requirements depend on how much of a substance is produced or imported annually into the European Economic Area (EEA).Footnote 159, Footnote 160
In the context of REACH, the Regulation (EC) 1272/2008 on the classification, labelling, and packaging of substances and mixtures (CLP Regulation) plays an important role. The CLP Regulation is based on the United Nations’s Globally Harmonised Systems (GHS) and requires manufacturers, importers, and downstream users of substances or mixtures to classify, label, and package their hazardous chemicals appropriately before placing them on the market.Footnote 161, Footnote 162 The CLP Regulation outlines the criteria for the hazard classification of substances.Footnote 163 When registering a substance under REACH, one requirement is to provide information regarding the hazard classification of that substance.Footnote 164
1. Aim and scope of REACH: protecting human and environmental health and promoting alternatives to animal testing
The purpose of REACH is to ensure a high level of protection of human health and the environment, including the promotion of alternative methods for the assessment of hazards of substances,Footnote 165 as well as the free circulation of substances on the internal market of the European Economic Area (EEA), while enhancing competitiveness and innovation (Article 1(1) REACH).
By incorporating the promotion of alternative methods in its purpose Article the Regulation makes clear that the 3Rs are an integral component of REACH. The fact that this is not to be understood as just an empty phrase, is emphasized in Article 13 REACH, which states that all information on intrinsic properties of substances must be generated by other means than testing on vertebrate animals wherever possible (1) and that the methods must be regularly reviewed and improved with a view to reducing testing on vertebrate animals and the number of animals involved (2). Article 25 REACH stipulates that testing on vertebrates for the purpose of the Regulation can only be undertaken as a last resort. Annexes VII to X REACH, which contain the standard information requirements for substances depending on the tonnage levels of production or import, also explicitly state in their introductory sections that all existing data should be consulted first before new tests may be carried out, and in their respective column 2 provide for possible adaptations to the standard information requirements, ie, scenarios in which a specific test listed in column 1 may be omitted.
REACH applies to manufacturers, importers, and downstream users who produce in the EEA, import into the EEA, place on the EEA market, and use substances on their own, in mixtures or articlesFootnote 166 in the EEAFootnote 167 depending on the intended production or import volumeFootnote 168 and obliges them to do so in a way that does not harm human health and the environment.Footnote 169, Footnote 170
2. Basic legal structure of REACH: substance registration, evaluation, authorization, and restriction
(a) Registration: “No Data, No Market”
Registration is mandatory for the manufactureFootnote 171 and placement on the EEA market of a substance on its own, in a mixture, or articles in quantities of one tonne or more per year.Footnote 172, Footnote 173 In other words: Without prior registration, chemicals can neither be produced nor imported (no data, no market). Substance registration is meant to ensure documentation for substances on the market to guarantee their safety for human health and the environment throughout the entire supply chain.Footnote 174 To register a substance, the manufacturer or importer into the European Union is required to apply to the European Chemicals Agency (ECHA)Footnote 175 by submitting a technical dossier containing specified information.Footnote 176, Footnote 177, Footnote 178 If the production or import volume surpasses 1 tonne per year, registrants are required to submit additional (eco)toxicological information on the substance and data on its physicochemical properties.Footnote 179 With increasing tonnage levels, additional information is required.Footnote 180 In addition to the technical dossier, registrants must perform a chemical safety assessment for substances in quantities of 10 tonnes or more per year.Footnote 181 To avoid duplication of studies, REACH provides for data-sharing and dispute settlement mechanisms for cases in which previous registrants fail or refuse to provide information on their testing results for the substance in question.Footnote 182
(b) Evaluation: Dossier and substance evaluation
During the evaluation phase, the ECHA performs a mandatory examination of testing proposals for the highest tonnage levels.Footnote 183 If an animal test is planned for a substance in the highest tonnage levels (Annexes IX and X REACH), registrants must submit the (animal) testing proposal before carrying it out. The ECHA must publish the proposal on its website and invite public commenting, ie, call on third parties to submit existing scientifically valid information on the substance in question.Footnote 184 Based on the information received, the ECHA must then draft a decision and give the registrant 30 days to comment on it.Footnote 185 The ECHA – or the Member State authority in cases of substance evaluation (Article 46 REACH) – must then consider the registrant’s comments and may amend the draft decision accordingly.Footnote 186, Footnote 187 This mechanism is intended to ensure that existing information is used and to avoid duplicate tests on vertebrate animals. However, evidence suggests that comments made by registrants in support of non-animal approaches are often not considered by the ECHA.Footnote 188
In contrast to its obligation to perform a compliance check for testing proposals for substances in the highest tonnage levels, the ECHA has discretion regarding compliance checks for all other registrations under REACH.Footnote 189 If the ECHA establishes insufficient compliance, it may require registrants to submit further information, eg, an (additional) animal test (Article 41 para 3 of REACH) after granting registrants their right to comment within 30 days of receipt (Article 50 para 1 of REACH). The ECHA must examine any additional information submitted by registrants (Article 42(1) REACH). As established by the European Court of Justice, the obligation to examine new information extends to information acquired using NAMs, even when the ECHA has explicitly demanded additional animal tests from the registrant to complete the registration dossier.Footnote 190 However, REACH does not provide a deadline after which the ECHA may no longer perform compliance checks. This means the ECHA can revisit a dossier at any time after registration and request new testing.Footnote 191
Under the coordination of the ECHA, the competent authorities of the Member States are responsible for evaluating the substances (not the registration dossiers) to determine if and which substances pose a risk to human health and the environment.Footnote 192, Footnote 193 If there is reason to suspect that the substance poses a threat, the registrants can be obligated to provide additional information, eg, based on animal testing for complex endpoints, such as endocrine disruption properties,Footnote 194 but also information not included in the standard information requirements in Annexes VII to X REACH.Footnote 195
In its report on the Integrated Regulatory Strategy to identify substances of concern for 2022, the ECHA stated that it had detected numerous substances that it suspected of having hazardous properties and that had been registered with insufficient data. The Agency concluded that more data was needed to assess compliance with REACH and the need of risk management action. Despite stating that it intended to assess all available data and group chemicals to avoid unnecessary animal testing, it is vague in terms of how many additional animal tests are required.Footnote 196 This reflects the concerns expressed by experts in the field of NAMs-based safety testing that the ECHA is requesting new animal tests for substances that are already widely used and that it is thereby not made sufficiently clear that animal testing will effectively only be required as a last resort.Footnote 197
(c) Authorisation and restriction
In the case of substances of very high concern listed under Annex XIV REACH,Footnote 198 registrants must apply for authorisation of production and/or import.Footnote 199 Annex XVII REACH contains a list of restricted substances, which may not be produced, imported, or used unless the conditions of the restriction in column 2 of Annex XVII REACH specify otherwise. Restrictions can also apply to substances in quantities of less than one tonne per year.Footnote 200
3. Standard information requirements and adaptations: animal testing and NAMs
(a) Standard information requirements according to Annexes VII–X
According to the general information requirements set out in steps 1–4 of Annex VI REACH, which are mandatory for all substances manufactured or imported in quantities of one tonne or more per year, the registrant must consider information needs for compliance with the relevant Annexes VI–X REACH and first gather and share all existing information on the substance. If information gaps are identified, the registrant must generate new data, ie, directly perform new tests according to Article 13 REACH, or in case the production or import volume exceeds 100 tonnes per year and the registrant plans to generate information on the substance with animal tests,Footnote 201 propose a testing strategy first before performing a test.Footnote 202
Annexes VII–X REACH each contain a column 1 with standard information requirements on the physicochemical substance propertiesFootnote 203 and toxicological and ecotoxicological information.Footnote 204 Column 1 lays down specific endpoints and test methods, which must be conducted according to the Test Method Regulation or in accordance with other international test methods recognised by the European Commission or the ECHA.Footnote 205 Column 2 specifies rules for adaptations to the testing requirements, ie, for omission or modification of testing or the need for additional testing.
(b) Chemical safety assessment
If a substance is produced or imported in quantities of 10 tonnes or more per year, the registrant must provide, in addition to the standard information required for the registration dossier set out in Annexes VI–X REACH, a chemical safety report containing a chemical safety assessment as set out in Annex I of REACH.Footnote 206 The chemical safety assessment consists of gathering existing information on the intrinsic substance properties, a human health hazard assessment, a physicochemical hazard assessment, an environmental hazard assessment,Footnote 207 and an assessment as to the existence of (very) persistent, (very) bioaccumulative and toxic properties (PBT or vPvB).Footnote 208 If the outcome of the assessment concludes that the substance fulfils certain hazard class criteria according to the CLP Regulation or is assessed to be have (very) persistent, (very) bioaccumulative and toxic properties, the chemical safety assessment must additionally include an exposure assessment and a risk characterization as described in Annex I, Sections 5 and 6 REACH.Footnote 209, Footnote 210
Note that a comprehensive human health hazard assessment required as part of the chemical safety assessment includes determining human exposure levels that should not be exceeded (Derived No-Effect Level, DNEL).Footnote 211 The standard human toxicity tests required in Annex VII, like skin sensitization, eye/skin irritation, and acute toxicity, cannot provide the necessary information to establish a Devired No-Effect Level. Currently, no validated non-animal test methods to determine a DNEL exist; therefore, for substances produced or imported in quantities over 10 tonnes per year, live animal testing remains a standard requirement.Footnote 212
(c) Adaptations to the standard information requirements according to Annex XI
In addition to the adaptations provided for in column 2 of Annexes VII–X, Annex XI REACH gives substance registrants the possibility to adapt their standard testing regime, ie, to waive animal testing. Testing according to the standard information requirements of Annexes VII–X may be omitted if it does not appear scientifically necessary. This can be the case when there is either existing data from previous testing or historical human data, sufficient weight of evidence from several independent sources, adequate results from qualitative or quantitative structure-activity relationship models ((Q)SAR), acceptable results from validated in vitro methods, or sufficient information from grouping or read-acrossFootnote 213 approaches.Footnote 214 Testing can also be omitted if it is technically impossible to testFootnote 215 or the registrant can demonstrate that humans or the environment will not be exposed to substance hazards (substance-tailored exposure-driven testingFootnote 216).
The ECHA can review dossiers to verify if the adaptations of the standard information requirements, along with the related justifications, comply with the rules for adaptations under Annexes VII to X and the general rules outlined in Annex XI.Footnote 217 This is emphasized by the information on the ECHA support page for registrants: “Importantly, omitting testing on animals must not compromise the safe use of substances. Therefore, every adaptation you use instead of submitting the standard information needs a valid and documented justification.”Footnote 218 The requirement to justify the submission of NAMs-based data highlights the animal-centric nature of the REACH Annexes and contradicts the principle that animal testing should be avoided whenever possible and used only as a last resort.Footnote 219
However, in a recent judgment by the European Court of JusticeFootnote 220 on the use of adaptations to fulfill standard information requirements after an explicit decision by the ECHA asking registrants to perform an additional animal test, the Court held that “a registrant has, generally and therefore especially where ECHA issues it with a decision asking it to complete its registration dossier with a study involving animal testing, not simply the possibility but the obligation Footnote 221 to generate information obtained by means other than animal testing ….” According to the Court, this conclusion follows from recital 47 of REACH, according to which “it is necessary to replace, reduce or refine testing on vertebrate animals,” the related general provisions of REACH, which require that information be generated whenever possible by means other than animal testing and that animal tests should be performed as a last resort only,Footnote 222 and the aim of REACH to promote alternative methods for hazard assessments.Footnote 223 This decision provides an important clarification as to what should be considered the standard information source for chemical safety assessments: Registrants have an obligation to use NAMs and may only resort to animal testing when there is truly no other way to obtain the necessary information.
In light of this judgment, there are calls to halt any new animal testing requested by the ECHA following compliance checks to enable a transparent scientific assessment of submitted testing proposals.Footnote 224 Others are calling into question the current system under REACH with its “rigidly hazard-focused, tick-box approach” in its entirety.Footnote 225
(d) Legal inconsistencies impeding the use of non-animal methods to adapt the standard information requirements
Article 13(3) REACH stipulates that test methods used for compliance with REACH must be carried out according to internationally accepted test method guidelines or otherwise be internationally accepted. It also states that information on the intrinsic properties of substances may be generated in accordance with other test methods Footnote 226 as long as the requirements set out in Annex XI REACH are met. For example, the Annex XI weight-of-evidence approach allows for “the use of newly developed test methods, not yet included in the test methods referred to in Article 13(3)” if they provide sufficient “weight of evidence,” which may lead to the “conclusion that a substance has or has not a particular property”. If this is the case, animal testing must be omitted.Footnote 227, Footnote 228 This wording expressly permits the use of NAMs that have not been validated (yet).
This is where the legal requirements become difficult to follow: The information acquired according to Annex XI of REACH must be adequate for the purpose of classification and labelling under the CLP Regulation. However, the criteria for classification as specified in Annex I of the CLP Regulation require either human or animal data or information based on tests according to test guidelines or other internationally accepted test methods.Footnote 229, Footnote 230 In turn, these criteria impede the use of non-guideline and/or unvalidated NAMs.Footnote 231 This merits further investigation to better understand what legislative amendments are needed to accommodate the uptake of new non-animal methodologies, particularly test methods that have not yet been officially validated but produce sufficient data to satisfy regulatory needs.
Some scientists have proposed alternative interpretations of Annex XI to allow for an increased use of NAMs for REACH compliance.Footnote 232 However, the underlying paradigm of animal-based testing remains an obstacle to the acceleration of NAMs acceptance: In practice, the ECHA generally only accepts NAMs-based data when the test method functions as a direct replacement of the standard animal test and has been fully validated.Footnote 233
Therefore, while the urgency of accelerating validation procedures is evident and the benefits of validating test methods are indisputable – such as promoting data acceptance across countries, ensuring regulatory consistency, building public confidence, and facilitating effective risk management, which ultimately reduces duplication in chemical risk assessments – there is also an equally pressing need to amend the relevant laws. The amendments must be designed to allow decision-makers greater flexibility in accepting information derived from innovative, not yet fully validated test methods, thus aligning legislation with the legal obligation to adhere to the 3Rs principle and, specifically, to use non-animal methods whenever possible and conduct animal testing as a last resort only.
4. Planned revision of REACH with negative implications for animals and the 3Rs
In its 2020 Chemicals Strategy for Sustainability, the European Commission announced a revision of REACH and the CLP Regulation, in which it stated its intention to amend EU chemicals legislation to ensure a safer use of chemicals and to simplify the legal framework.Footnote 234 In its Inception Impact Assessment of 2021 on the planned revision of REACH and the CLP Regulation, the European Commission held that the revision aims, among other things, to tackle substance information gaps.Footnote 235 In its presentation of policy plans,Footnote 236 the Commission listed increased information requirements for low tonnages and the identification of the most harmful substances as the main goal of its Chemicals Strategy for Sustainability. While some animal tests are set to be deleted (eg, acute oral toxicity in rats) or replaced (eg, replacing short-term toxicity testing in fish with in vitro testing or fish embryo toxicity), the Commission states that it intends to close the human health information gaps for 5,800 low tonnage substances by subjecting them to information requirements for higher tonnage bands such as repeated dose toxicity and reproductive and developmental toxicity. Furthermore, more animal testing is also planned for endocrine-disrupting chemicals. There is, however, no further illustration of whether and how the last resort requirement of Article 25 REACH will be promoted and ensured as part of the planned amendment.
Scientists estimate that millions of animal lives could be lost if the REACH revision proceeds without transitioning from the animal-based testing standard to a next-generation risk assessment (NGRA) using non-animal methods to meet the increased information requirements for identifying endocrine disruption.Footnote 237 The fact that more animal testing is being actively planned is incomprehensible given the global trend towards the use of NAMsFootnote 238 and the European Commission’s own words that “safety testing and chemical risk assessment need to innovate in order to reduce dependency on animal testing but also to improve the quality, efficiency and speed of chemical hazard and risk assessments”.Footnote 239
This is all the more untenable in light of the questionable reliability and relevance of animal testing.Footnote 240 It is also seen as a missed opportunity to facilitate the transition to a more modern and efficient chemical safety regulation that would truly be in line with the objectives of REACH to promote innovation and NAMs.Footnote 241 The proposed amendments to REACH and the CLP Regulation also contrast with other European regulatory fields, such as food safety, where the competent regulatory bodies, namely the European Food Safety Authority (EFSA), are moving towards less animal testing and more new technologies.Footnote 242
The final REACH revision proposal was initially planned for 2022 and was then postponed to the end of 2023 by the European Commission, with implementation of the revision initially expected to enter into force between 2025–2027.Footnote 243 However, as of now, the proposal has not yet been introduced, and it remains uncertain whether the revision will proceed as planned or when it will be implemented.Footnote 244 The delay has raised public and environmental health concerns due to a lack of identification and effective management of hazardous substances.Footnote 245 The European commission has been criticised for stalling the revision following lobbying by chemicals industry groups opposed to anticipated restrictions and bans on highly toxic chemicals.Footnote 246, Footnote 247 The Commission has remained vague about the reasons for the delay, emphasizing that quality takes precedence over speed, the need to ensure compatibility with other regulations, and to design a regulation that simplifies the administrative burden for registrants, all of which takes time.Footnote 248
5. Animal testing for cosmetics-only substances continues for compliance with REACH
There has been growing concern and frustration among the public,Footnote 249 the scientific community,Footnote 250 and manufacturers of chemical productsFootnote 251 over ongoing animal testing for ingredients in cosmetic products to comply with REACH requirements, even though the Cosmetics Regulation prohibits animal testing for marketing cosmetic products – both final products and ingredients used exclusively in cosmetics (cosmetics-only substancesFootnote 252).Footnote 253 The European Parliament, in its resolution on a global ban to end animal testing for cosmetics, has also clearly stressed the importance of upholding the EU citizen’s will to end animal testing for cosmetics.Footnote 254
There has been an ongoing conflict between the Cosmetics Regulation and REACH since 2014,Footnote 255 with the ECHA and ECHA Board of Appeal having thus far interpreted the relevant provisions in favour of animal testing for compliance with REACH: They hold that the Cosmetics Regulation is solely intended to protect end users of cosmetic products – ie, consumers and professionals who work with cosmetic products. Therefore, they consider new animal tests justified if conducted to determine the compliance of a cosmetic ingredient with REACH, specifically to assess the risks associated with worker exposure and environmental health.Footnote 256, Footnote 257
This interpretation by the ECHA and the ECHA Board of Appeal has now been upheld by a 2023 ruling by the General Court of the European Court of Justice (GC).Footnote 258 The Court determined that the relevant provisions do not establish the primacy of one regulation over the other.Footnote 259 It also concluded that the scope of the animal testing ban under the Cosmetics Regulation is limited to studies performed to assess the safety of end users only, leading to the conclusion that workers involved in the manufacturing of a cosmetic product or its ingredients are excluded from the scope of the Cosmetics Regulation, as they are exposed to the product or ingredient throughout its production cycle and under conditions that differ from those associated with the use of the final product.Footnote 260
As a result of the limited scope of the animal testing ban under the Cosmetics Regulation, over 100 studies on animals have been performed for sixty-three cosmetics-only substances for human-health endpoints after the deadlines for animal testing set by the Cosmetics Regulation. Most of these animal tests are assumed to have been conducted to fulfil REACH requirements.Footnote 261
These developments underscore the urgent need to address the protection goals of the different laws involved to bring the practice of testing substances in line with the political decision to ban animal testing for cosmetic products. For example, manufacturers of cosmetic products have demonstrated how they implement specific risk management measures to ensure worker safety during the production phase, thus rendering animal testing for this purpose unnecessary.Footnote 262 Such proposals must be integrated into the legislation to uphold the fundamental commitment to ending animal testing for cosmetics.
6. Summary of current challenges to the implementation of NAMs under REACH
Despite the significant potential of NAMs to replace animals in chemical safety assessments, various obstacles hinder their broader adoption within the current regulatory framework under REACH and impede the shift away from the animal testing paradigm. These obstacles, some of which have been mentioned throughout this study, will be briefly outlined and summarised below.
(a) Inadequate and contradictory legislation
Despite the legal requirement to use animal testing as a last resort only, REACH largely requires guideline animal tests for compliance with the standard information requirements in its Annexes, especially for higher tonnage production or import volumes. Test method guidelines are updated following lengthy validation procedures, thus offering only a very small number of validated NAMs to choose from. At the same time, the option to adapt the standard information requirements by using new, not yet validated non-animal test methods is hampered by contradicting legal requirements that essentially force registrants to perform animal tests. Therefore, suggestions have been made to replace the current system of fixed testing requirements towards a more open and flexible system of assessing chemical safety that does not follow a tick-box approach of mandated tests and leaves room to fulfil the legal requirement of putting safe chemicals on the market in a result-oriented way.Footnote 263
Furthermore, scientists are criticising the widely unquestioned paradigm of testing on animals with doses that are multiple times higher than that to which humans and the environment are exposed to in reality, with the aim of coming at conservative answers to maximise safety, all the while ignoring the fact that this approach is scientifically questionable and can produce distorted results.Footnote 264
Some stakeholders advocate for the requirement of testing proposals for all substances, not just those produced or imported in volumes of 100 tonnes or more per year.Footnote 265 The ECHA must publish the testing proposal on its website and allow third parties to submit existing information on the substance.Footnote 266 This process aims to promote transparency and collaboration in the safety assessment, ultimately helping to avoid unnecessary or duplicate animal tests. By extending the duty of registrants to submit testing proposals and the ECHA’s obligation to publish these proposals for all planned animal testing, regardless of production or import volume, proponents of this amendment argue that it would encourage the use of NAMs instead of animal studies.Footnote 267 Another related demand is to amend the last resort requirement under REACH to specify that animal testing can only be performed if the NAMs-based results are insufficient for risk assessment and risk management.Footnote 268
There is also a conflict between REACH and the Cosmetics Regulation as animal testing is still required to assess the safety of substances intended solely for cosmetics under REACH. A revision of both regulations is needed to ensure that the intentions of the EU Parliament and EU citizens to ban animal testing for cosmetics are respected. It has been proposed that only existing information or data generated using NAMs should be permitted to meet the standard information requirements under REACH for cosmetics-only substances.
(b) Restrictive validation system
To achieve quicker validation and broader acceptance of NAMs the current validation system is in need of an overhaul. So far, only NAMs that function as direct replacements for animal test methods have been validated, while other scientifically sound approaches that ensure human-relevant results are still being met with scepticism. Furthermore, method-by-method replacements have only proven successful for simple endpoints such as eye or skin irritation, while battery approaches to recreate whole organisms for more complex endpoints have not yet been validated and included in regulations.
It currently takes up to 20 years from the development of test methods to their validation, which is untenable and shows that the legal side of toxicity testing is lagging far behind scientific developments. The (largely successful) ban on animal testing for cosmetics was possible not least because of increased technological developments and investments in non-animal test methods.Footnote 269 Therefore, the fact that the paradigm shift in chemical safety testing for compliance with REACH is being hampered by a rigid and outdated validation system despite the availability of the necessary technology is becoming increasingly unjustifiable.
Another issue lies in how NAMs are obviously being held to stricter standards than animal models with respect to the variability of test results. As mentioned in the introduction, there is ample evidence of poor transferability of test results from animals to humans and between species, as well as for the uncertainty and variability of animal test results.Footnote 270 Therefore, scientists are calling for the same standard of variability to apply to NAMs as animal methods, and also for the uptake of new approaches to validation that consider the relevance of NAMs for humans (performance-based validation) instead of their ability to replicate animal models.Footnote 271
(c) Low acceptance by the ECHA
The ECHA has pointed to an inadequate use of non-testing approaches by registrants such as read-across and Qualitative or Quantitative structure-activity relationship models ((Q)SAR)Footnote 272 and that it has therefore required additional data generation via standard (animal) testing.Footnote 273 This has been attributed to an unreasonably high burden of proof required to justify the use of adaptations to the standard information requirements.Footnote 274 Given the last resort requirement, it is questionable why the ECHA does not – or is not obligated – to accept NAMs-based information by default if the regulatory needs are adequately addressed, and, where more information is necessary, to require additional NAMs-based data first, before resorting to animal testing as a genuine last resort.
NAMs are also having a hard time because there is a lack of multidisciplinary expertise amongst regulators who have been traditionally trained to interpret results from animal tests and therefore expect information that is directly comparable or equivalent to animal models. Therefore, experts in the field are calling for more collaboration between academia, industry, and regulatory agencies, as well as enhanced knowledge generation in academia to foster knowledge transfer and a better understanding of NAMs and how NAMs-based chemical safety assessments work.Footnote 275
Another concern is that the ECHA often requires testing of substances used solely in cosmetics to comply with REACH, aiming to protect workers and the environment. This occurs despite industries suggesting practical solutions to mitigate worker exposure to these substances without resorting to animal testing. These solutions include utilising consumer-related exposure data, considering various exposure scenarios for workers, and implementing adequate occupational safety measures to safeguard workers from potential exposure.Footnote 276
(d) Recent and planned amendments to REACH and CLP
Some recent amendments to REACH have included further animal testing requirements. A 2022 amendment to Annexes VI to X of REACH requires several additional animal tests for substances that have either been tested in vitro and the results have come out positive, or that are otherwise of a specific concern. It is expected that this amendment will cost millions of animal lives for substances that have already been registered. An amendment to the CLP Regulation of 2023 included several new hazard classes relating to endocrine disruption in humans and the environment, with animal testing being the main method of testing for these endpoints, which is also expected to cost additional animal lives.
Furthermore, based on what has been communicated by the European Commission so far, it is also expected that the pending revision of REACH and the CLP Regulation will result in increased animal testing. Consequently, both the recent and the forthcoming amendments, should they be carried out as planned, are incompatible with the requirement to adhere to the 3Rs, specifically, to exhaust all non-animal methods first and use animal testing as a last resort only, and the final goal of moving away from animal testing according to EU Law. This is also in obvious contrast to other regulatory areas, such as food safety, where the European Food Safety Authority (EFSA) is readily adopting new technologies.
(e) Registrant hesitation
For all the stated reasons above, it has been observed that registrants often use animal models pre-emptively as a path of least resistance, and also because of a fear of legal liabilities in case of wrong substance classification.Footnote 277
(f) Summary
The implementation of NAMs under REACH faces significant challenges, including contradictory and inadequate legislation, a slow and restrictive validation system, low acceptance by the ECHA, and legislative amendments that have increased animal testing requirements, with an upcoming revision expected to add even more animal tests to the standard information requirements. Consequently, registrants often resort to animal testing preemptively. To overcome these obstacles, legislative and regulatory reforms are necessary to create a more flexible and adaptive framework. This includes updating validation processes, ensuring equal standards for NAMs and animal tests, enhancing expertise on NAMs among regulators, and aligning contradicting legislations to avoid duplicating animal tests for the same substance or carrying out animal tests for cosmetics-only substances.
IV. Conclusion and outlook
The analysis presented in this paper highlights significant challenges and opportunities associated with the implementation of New Approach Methodologies (NAMs) in regulatory toxicity testing under REACH. Despite the explicitly stipulated final goal of moving away from animal testing in EU law, the legal requirement to replace, reduce, and refine animal testing and to use animals as a last resort only, the current regulatory framework under REACH still heavily relies on animal-based methods due to several entrenched practices and systemic barriers. These include contradictory and inadequate legislation, a slow and restrictive validation system for NAMs, and low acceptance by the European Chemicals Agency (ECHA). Additionally, recent and forthcoming amendments to REACH and the Classification, Labelling and Packaging (CLP) Regulation are poised to increase the requirements for animal testing, further complicating the transition towards non-animal methods.
The persistence of animal testing within REACH can be attributed to historical precedence, risk-averse regulatory practices, and the deeply ingrained belief in the reliability of animal models for ensuring human safety. However, the scientific community is increasingly critical of the translatability and predictability of animal tests. Evidence suggests that many substances deemed safe in animal studies have proven harmful to humans, and vice versa, raising concerns about the efficacy of animal-based testing. Moreover, the financial and ethical costs associated with animal testing cannot be overlooked, further necessitating the shift towards more humane and scientifically advanced methodologies.
The current system also faces significant practical challenges, such as the vast number of chemicals requiring safety assessments and the limitations of animal testing in addressing complex and emerging health issues, such as the effects of nanomaterials and chemical mixtures. The inability of animal models to adequately mimic human physiology and predict human-specific responses underscores the urgent need for alternative approaches.
Looking ahead, it is imperative to address the aforementioned challenges to foster the broader adoption of NAMs and advance the 3Rs principles. Legislative and regulatory reforms are necessary to create a more flexible and adaptive framework that prioritizes NAMs over animal testing. This includes updating the validation processes to ensure they are more inclusive of NAMs that do not function as 1:1 replacements of animal models, enhancing the expertise on NAMs among regulators, and aligning conflicting legislations to prevent redundant animal testing for the same substance.
Additionally, fostering collaboration between academia, industries, and regulatory agencies will be crucial. Such collaboration can enhance the generation and transfer of knowledge regarding NAMs, thereby improving their acceptance and implementation in regulatory contexts. Moreover, increasing investment in the development and validation of NAMs can provide more reliable and human-relevant data, ultimately leading to safer and more ethical chemical safety assessments.
Promising developments include the European Parliament’s resolution calling for an EU-wide action plan to accelerate the shift to non-animal testing methods, the EU Commission roadmap for phasing out animal testing in chemical safety assessments under way, and the strategic plans by the ECHA to enhance cooperation with various stakeholders with regard to NAMS. These initiatives, if effectively implemented, could significantly reduce the reliance on animal testing and promote the use of NAMs.
To conclude, the road to fully integrating NAMs in regulatory toxicity testing under REACH is fraught with obstacles, but urgently needed to benefit human health, environmental safety, and animal welfare. By embracing scientific advancements and regulatory flexibility, a more humane and scientifically robust framework for chemical safety assessment is possible. In this regard, it is also imperative that legal scholars bring their expertise into the field and begin working alongside scientists to assist in overcoming the legislative and regulatory challenges and to facilitate the regulatory acceptance of NAMs.
Acknowledgments
This work was carried out as part of the National Research Program (NRP)79 “Advancing 3R”, project no. 206392, funded by the Swiss National Science Foundation (SNSF).
Competing interests
The author has no conflicting interests to declare.
Open access
