Alternatives à l'expérimentation animale pour l'évaluation de la sécurité des produits chimiques en Europe


    Testing the safety of chemicals sometimes requires the use of tests on animals. However, there is a strong push towards the use of alternative testing methods both by regulatory bodies and the public at large. These highlights explain what are these alternative methods, discuss how effective they are, and look into the requirements in the frame of the Regulation on the Registration Evaluation Authorisation and restriction of Chemicals (REACH, Commission regulation 1907/2006) made by the European Chemicals Agency (ECHA)

    How is the safety of chemicals regulated in the European Union?

      In Europe, since 2007, the use of chemicals is regulated under the “REACH” Regulation1. One of the main reasons for developing REACH was that a large number of substances were already in use on the European market for many years, and there was inadequate information on their intrinsic properties and the risks for humans and the environment that their use may pose.

      This Regulation defines the conditions required2 for the use of chemicals on the European market, whether produced or imported, and used as raw materials or intermediaries in industrial processes, or as substances in research laboratories. Companies have thus to register the chemicals that they import, use and/or sell to the European Chemicals Agency (ECHA), and to produce the information needed to determine their safety for human health or the environment.

      In order to assess this safety, there are about thirty standard effects or “endpoints” that have to be evaluated, from acute toxicity including skin irritation, to carcinogenicity and reproductive toxicity, and as regards ecological systems, persistence, bioaccumulation and impact.

      2 REACH prescribes the minimum information requirements for physicochemical, toxicological and ecotoxicological properties of a substance based on the tonnage of the chemical substances manufactured or imported. A key feature REACH is based on the principle that manufacturers, importers and downstream users are responsible for ensuring and showing that they manufacture, place on the market, or use substances that do not adversely affect human health or the environment

      What are the main requirements to register a chemical substance to the ECHA?

        To document a registration file in the context of the REACH regulation, for each of the potential toxic effects of a substance for humans and the environment, there are a number of ways that information can be gathered:

        • a review of the scientific literature that already exists on the properties (hazard and risks) of the substance;
        • an inference based on the information available for substances that are known to be more or less similar; 
        • an evaluation of the potential effects of the substance by use of computer models that compare substances whose structures are comparable (“Quantitative Structure-Activity Relationships” or QSAR models); 

        Presently, the main source of information originates from experimental in vivo studies (67 % on average), with a high percentage of them having been carried out before the REACH Regulation entered in force in 2007 (on average 44 %).

        Under Article 117(3) of the REACH Regulation, one of the objectives is to promote non-animal testing methods. Accordingly, ECHA proposes a Guidance on information requirements and chemical safety assessment, which has been updated to reflect the latest scientific developments in the field and ECHA’s current best practice3.

        3  and 

        Why animal testing has to be done?

          In order to know what the effect on human health or the environment of a substance could be, it is necessary to test it to ensure that the conditions in which the people will handle and use the substance will be safe.

          In the case information on potential toxic effects of a substance is necessary and new tests are needed, there are two possible options:

          • do tests on the chemical properties and on their effects “ in vitro ”, which means on isolated cells in culture, bacteria (in particular for genotoxicity tests), or on artificial tissues; or 
          • do tests on animals for a particular ‘endpoint’ that cannot be evaluated by in vitro methods.

          Some toxic effects are still difficult to anticipate and evaluate on “ in vitro ” models due to the variety and complexity of different physiological processes that could be involved, in particular for reproductive, immunologic or carcinogenic effects. Due to these limitations (e.g. applicability domain of the method or adequacy for classification and labelling purposes), some in vivo testing may still be necessary.

          When these further animal tests are necessary besides the “basic tests” imposed by the regulations, they are conducted under close control to ensure that the smallest possible number of animals is used to get statistically meaningful results, and the company that is registering the chemical must ask for approval for their performance before conducting them.

          Meanwhile, some in vivo tests may be available for other reasons, for example, if registrants under REACH obtain access to such studies as they were conducted to fulfill regulatory requirements outside the EU.

          Did the REACH Regulation in Europe result in a reduction in animal use in laboratory testing?

            For the purposes of registration under REACH, registrants must not undertake any new studies involving vertebrate animals before submitting a testing proposal to ECHA and receiving ECHA’s approbation. When they submit their proposal, the registrants must show in their dossier that they have considered alternatives. During the evaluation of the dossier, testing proposals can be removed, considered as inadmissible and therefore not further processed, put on hold, replaced by adaptations (which means the use of non-animal methods for addressing or omitting information requirements), or appear under the scrutiny of an appeal procedure.

            Overall, only 11 % of REACH information requirements for substances analysed in this report (so from 2009 onwards) have been covered by new experimental studies performed on vertebrate animals, and the instruments provided by REACH to avoid unnecessary animal testing seem to work well.

            The main contributing factor to the reduction of the use of animals in laboratory testing is the registrants’ obligation to share data and existing information on the substance and act jointly when they wish to register the same substance. This ensures that for each substance, the test data are collected, generated and brought together in one joint registration dossier. It appears that 98 % of the substances are registered jointly. Registrants largely make use of existing in vivo studies only (>50 %), and to a lesser degree, of new in vivo studies only (up to about 4 %) to fulfill the requirements. To this end, ECHA provides a Guidance on data-sharing4.

            Thus, when possible, registrants have to make an extensive use of the various alternative “ in vitro ” testing possibilities instead of conducting new studies or proposing further animal tests, particularly on vertebrates. But given that not only the use, but rather the proper use of alternatives to testing on animals is essential to achieve a key objective of REACH, it is also important to reflect upon the quality of alternative adaptations.

            However, with the increasing number of new registration dossiers submitted to the Agency, the absolute number of new experimental studies has increased for all endpoints. Nevertheless, alternative options for addressing information requirements have been used extensively by registrants particularly for the more detailed “high tier” endpoints5, where the numbers for new experimental studies or testing proposals are not as high as could be expected from the number of submitted registration dossier. For example, for several larger groups of substances, tests would be done only for some substances within the group.

            If registrants cannot provide the information required for the “high tier” endpoints with existing experimental data or alternative methods, or by omitting the study on justified grounds, they need to submit a testing proposal to ECHA and await its decision on the proposal. Registrants must nowadays also provide their considerations for alternative methods to justify the need for testing proposals on vertebrate animals.

            So far, the vast majority of examined testing proposals were considered necessary by ECHA and the Member States and resulted in an adopted decision authorising the testing.

            This ensures that for each substance the test data are collected, generated and brought together in one joint registration dossier. Globally, for 6290 registered substances that were analysed in the present report for the endpoints concerning the use of vertebrate animals in the testing procedures:

            • 89 % contain at least one endpoint in the dossiers where an adaptation or other argument was provided instead of an animal study result;
            • 63 % contain at least one case with toxicity data from a similar substance (read –across adaptation);
            • 43 % contain at least one use of a combination of information from several independent sources on previous studies on the same substance to give sufficient evidence to fulfil an information requirement (the so-called “weight-of-evidence” method”);
            • 34 % contain at least one QSAR (Quantitative Structure-Activity Relationship) prediction, which is when computer modeling is used to predict the effect of a substance. ECHA has provided tools and practical guides on how to properly use newly developed methods and to build good quality QSAR predictions. 

            5 High tier human health endpoints include repeated dose toxicity by all exposure routes and of all durations, genetic toxicity in vivo, developmental toxicity, toxicity to reproduction and carcinogenicity.

            For which toxicity end-points are alternative methods most applicable?

              One of the priority areas for ECHA’s regulatory science activities are non-animal alternative methods and new approaches to hazard assessment, in particular rational integration of different lines of evidence and other means of reduction or refinement when non-animal approaches are not yet available. There is a wide range of methods available to perform hazard assessments of chemicals, including “traditional” toxicological studies, in vitro tests, ”read-across”, QSARs, and ”high throughput screening”6 approaches, but further research is needed to combine these approaches, e.g. into integrated testing strategies or similar.

              Registrants already use existing alternative in vitro testing methods and approaches for skin corrosion/irritation and serious eye damage/eye irritation and skin sensitisation. Around 34 % of substances are covered by alternatives and over 56 % of the new studies performed for these endpoints were using these in vitro tests. In some cases (about 20 %), the information requirement was fulfilled by using in vitro test data either alone or together with a so-called read-across approach or RAAF which involves “the use of relevant information from analogous substance(s) to predict properties for the ‘target’ substance(s) under consideration7. ECHA has provided tools and practical guides on how to properly use newly developed methods and to build good quality QSAR predictions.

              Due to limitations in the above approaches (e.g. applicability domain of the method, or adequacy for classification and labelling purposes), some in vivo testing may still be necessary. Nevertheless, alternative options for addressing information requirements have been used extensively by registrants. This is particularly true for endpoints covering acute and local effects, for which registrants mainly use experimental studies (many of them were carried out before REACH) and where the numbers for new experimental studies or testing proposals are not as high as could be expected from the number of submitted registration dossiers.

              When ECHA suspects that registrants have not complied with their legal obligations to use alternative methods, the Agency refers the case to Member State authorities to consider any enforcement action.

              However, especially for high tier endpoints, it is apparent that neither in vitro methods nor QSARs will be able to replace animal testing in a simple one-to-to manner for complex hazard endpoints. If read-across is a promising methodology, ECHA’s evaluation experience was that this approach was often not substantiated with thorough argumentation, and supporting evidence was frequently lacking.

              New approach methodologies (NAMs) are also addressed in a broad context to include in silico (computational) approaches, in chemico and in vitro assays, as well as including information from the exposure of chemicals in the context of hazard assessment. NAMs also include a variety of new testing tools, such as the “high-throughput screening” and “high-content methods” which include genomics, proteomics, metabolomics; as well as some “conventional” methods aiming to improve the understanding of toxic effects through improving toxicokinetic or toxicodynamic knowledge8.


              What are the limits of the alternative methods?

                One of the great challenges is currently to compensate the complexity of a higher organism (metabolism, toxicokinetics, etc.) when extrapolating from data obtained for instance on cell cultures.

                ECHA’s evaluation experience shows that many ‘adaptations’ had quality deficiencies. These include, especially with respect to commonly used read-across adaptations, which use relevant information from analogous substance(s):

                • poor documentation;
                • insufficient substance identification;
                • significant deficiencies in the quality of the source studies;
                • lack of or low quality of supporting data;
                • lack of qualitative and quantitative data to support predictions based on metabolism and elimination (toxicokinetics) of the substance;
                • shortcomings in the toxicological hypothesis.

                The deficiencies related to the supporting evidence are particularly relevant for human health endpoints such as long-term toxicity or genetic toxicity, and for environmental endpoints such as toxicity to birds and fish.

                New approach methodologies (e.g. high throughput in vitro screening) have the potential to further substantiate the hypotheses of read-across approaches as they often use starting points which are directly relevant for humans (e.g. human liver cells). However, in order to increase the robustness and regulatory acceptance of these adaptations for human health and environmental endpoints, additional data is needed, particularly related to mechanisms of toxicity and absorption, distribution, metabolism and excretion (ADME) properties.

                New approach methodologies (e.g. high throughput in vitro screening) have this potential to further substantiate the hypotheses of read-across approaches as these approaches often use starting points which are directly relevant for humans (e.g. human liver cells).

                What does the European chemicals agency ECHA do to promote alternatives to animal testing?

                  The development of new approach methodologies will bring high throughput assessment methods, which can support current alternative approaches and might potentially provide more human relevant information. A challenge for ECHA is to bring them into regulatory use.

                  For some information requirements for end points of low complexity, such as skin irritation, ECHA’s focus is promoting the possibilities of a range of appropriate alternative in vitro methods, which are already available. In addition, the amount of available experimental data on other substances and the generally low complex toxicology of these effects increase the possibility to successfully apply alternatives like read-across and QSARs.

                  ECHA will also continue to contribute to developing and promoting alternative methods in an international context, most specifically through the Organisation for Economic Co-operation and Development (OECD).

                  For the more complex endpoints, the focus of ECHA is on making the shortcomings that are observed more explicit. The publication of the recently updated (2017) Read-Across Assessment Framework to cover environmental endpoints already allows registrants to improve their read-across predictions for these endpoints.

                  ECHA also supports the development by the OECD of the software QSAR Toolbox9, which will be further developed and improved before the third REACH deadline in 2018 for the last class of chemicals not yet registered. It is already an option used after the opportunity given by REACH to waive studies followed by the read-across and the weight of evidence approaches.

                  FURTHER INFORMATION
                  “The use of alternatives to testing on animals for the REACH Regulation” report 

                  How to avoid unnecessary testing on animals?

                  Information on animal testing

                  Practical guide: How to use alternatives to animal testing


                  Publications A-Z

                  Get involved!

                  This summary is free and ad-free, as is all of our content. You can help us remain free and independant as well as to develop new ways to communicate science by becoming a Patron!

                  PatreonBECOME A PATRON!