International Regulation of Toxicological Test Systems

Abstract

Toxicity testing for regulatory purposes began after the thalidomide (Contergan®) tragedy in Germany around 1960, when unexpectedly the hypnotic drug caused severe limb malformations in newborns whose mothers had taken the drug during pregnancy. This accident initiated in the first place regulatory testing for drugs, while safety testing of other chemicals, e.g., pesticides, industrial chemicals, cosmetics, and food ingredients, has become mandatory at the international level about 20 years later. Meanwhile regulatory safety testing of chemicals is mandatory in all OECD (Organization for Economic Cooperation and Development, http://www.oecd.org/) member states, which are the major industrial countries except Brazil, China, and India. Initially test requirements as well as individual tests were developed at the national level, and therefore, they differed quite significantly due to experience at the national level of both in industry and regulatory agencies. As a consequence, international industry had to conduct different sets of safety tests to meet the legal requirements in each of these countries. In essence, differences in test requirements created a considerable financial burden both for the industry and also for patients and consumers, who in the end have to pay for testing. Thus, the differences in testing requirements, which were not based on scientific grounds, resulted in barriers to international trade.

Finally, the major industrial nations agreed to correct the situation by harmonizing the safety testing requirements. At the OECD, industry and regulatory agencies took the lead in this activity, since regulatory testing has a higher priority for them than for research-driven academic institutions. It was the goal of the harmonization activity to achieve mutual acceptance of data (MAD) by OECD member countries when the data were produced according to standardized testing requirements. Today harmonization of test guidelines has been achieved for regulatory testing in all areas of toxicology. In addition, the criteria for development and validation of new test methods have been harmonized, since test guidelines have to be updated continuously according to scientific progress and also to meet the needs of testing for newly emerging endpoints.

In this chapter the concept of the mutual acceptance of data (MAD) produced in standardized and harmonized toxicity tests is described and also the criteria for developing, validating, and achieving international acceptance of new tests methods.

International Harmonization of Guidelines for Toxicity Testing

For historical and economic reasons, standardization and harmonization of test guidelines were developed independently for the major areas of toxicology, e.g., safety testing of drugs, cosmetics, industrial chemicals, and pesticides/biocides. The major reason for this are differences in regulatory requirements, which take into account differences in use and exposure of humans and the environment. The most important step was in 1982 the adoption of the OECD Guidelines for the Testing of Chemicals (OECD TGs) http://www1.oecd.org/ehs/testguid which are mandatory today for testing of all chemicals except human and veterinary drugs. It is important to note that OECD TGs have been accepted not only for toxicity testing but also for physicochemical properties and for environmental safety including accumulation and degradation of biotic and abiotic systems.

The most important consequence of the adoption of harmonized OECD TGs by OECD member states is the concept of mutual acceptance of data (MAD) that are generated by testing according to OECD TGs by regulatory agencies of all OECD member countries. Another important requirement for MAD is that testing must be conducted according to GLP (good laboratory practice), which is an accepted measure to enforce quality assurance. If a toxicity test has been conducted according to an OECD TG and to GLP, the test must only be performed once and not any more for each national regulatory agency. The harmonization of TGs by the OECD has not only financial advantages for industry and consumers, but it also improves ethical standards from the animal welfare perspective. The latter aspect is a major driving force in OECD member countries. Finally, from the scientific perspective, harmonization of TGs is most welcomed since data for specific endpoints of toxicity are now produced according to the same TG, which provides for a better comparison of the results of studies conducted for the same endpoint.

It has been criticized that OECD TGs are quite rigid, and it is time consuming and laborious to update them once they have been accepted, since at the OECD decisions are not made by majority vote, but all OECD member states have to agree. Although progress may be delayed by unanimous agreement, this has not really happened during the past 20 years, since the procedure for submitting new TGs or for updating existing TGs has continuously been improved.

In Germany and all EU member states, testing of chemicals has in the first place to be conducted according EU TGs, which are almost identical to OECD TGs. When the EU legislation was changed in 2007 from the EU Directive 76/769/EEC on the regulation of dangerous substances and preparations to the EU Directive 1907/2006 on the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH, http://echa.europa.eu/web/guest/regulations/reach/), only testing according to OECD TGs is accepted in Europe.

Since the concept of using standardized OECD TGs for toxicity testing has proven successful and also the mutual acceptance of data (MAD), OECD TGs are now used for safety testing of all chemical substances and products except drugs, e.g., biocides, pesticides, cosmetics, as well as food and feed additives.

Taking into account the successful implementation of the OECD TGs into regulatory practice, in 1990 the national and international agencies that are responsible for the regulation of drug safety agreed on an international harmonization of the test guidelines for human and veterinary drugs. They were harmonized by the ICH (International Conference on Harmonisation, http://www.ich.org/), which is formed by regulatory agencies and the drug industry of the major economic regions Europe, Japan, and the USA. ICH test guidelines (TGs) are used not only for toxicity testing but also for all other areas of preclinical drug testing, e.g., efficacy testing and pharmaceutical quality control. As described for OECD TGs, results of tests that were conducted according to ICH guidelines will only be accepted internationally according to MAD, if testing has been conducted according to “GLP,” and produced according to “good manufacturing practice (GMP).” Again, the harmonization of TGs has led to significant reduction of testing in animals, since regulatory agencies around the world are now accepting the results of a test that was conducted according to ICH guidelines.

Table 1 summarizes the most important areas, which require safety testing in animals and in which the test guidelines have been harmonized at the international level. Table 1 shows that in addition to drugs, industrial chemicals, and pesticides, international TGs have also been harmonized for hormones and biologicals by the pharmacopoeias and for vaccines by the WHO. So far, the harmonization of international TGs for toxicity and safety testing has been the most successful approach to reduce animal testing for regulatory purposes.

Table 1 International harmonization of methods for safety testing

Validation and Acceptance of New OECD Test Methods

Regulators will only accept new nonanimal tests, also termed “alternative tests” (e.g., in vitro or in silico tests), if the new tests allow to classify and label chemicals in the same way as the current animal tests. The OECD has, therefore, decided that in vitro toxicity tests can be accepted for regulatory purposes only after a successful experimental validation study has been conducted. To approach this problem scientifically, European and American scientists agreed in 1990 in Amden, Switzerland, on a definition of experimental validation and the essential steps in this process. At this workshop, validation was defined as the process by which reproducibility and relevance of a toxicity testing procedure are established for a particular purpose (Balls et al. 1990), regardless of whether the method is an in vitro or in vivo test. The essential steps of the experimental validation process were defined in the following manner:

• 1. Test development in one or several laboratories

• 2. Experimental validation under blind conditions in several laboratories in a ring trial

• 3. Independent assessment of the results of the validation trial

• 4. Regulatory acceptance

Steps 2 and 3 are the essential part of a formal validation study conducted for regulatory purposes. The report of this workshop (Balls et al. 1990) encouraged scientists to initiate several international validation studies. Since the Draize eye test has been the most widely criticized toxicity test, a worldwide validation study on nine nonanimal alternatives to the Draize eye test was coordinated by the EU Commission’s Centre for the Validation of Alternative Methods (ECVAM, http://ecvam.jrc.it/) and the British Home Office. However, this and other extensive international validation attempts failed (Balls et al. 1995b).

Therefore, the leading scientists involved met for a second validation workshop in Amden in 1994 to improve the concept of the validation procedure. The second Amden validation workshop recommended the inclusion of new elements into the validation process (Balls et al. 1995a), which had not sufficiently been identified in the first Amden validation workshop. The following three essential elements were added:

• 1. The definition of a biostatistically based prediction model

• 2. The inclusion of a prevalidation stage between test development and formal validation under blind conditions

• 3. A well-defined management structure

As to in vitro tests, a prediction model should allow the prediction of in vivo endpoints in animals or humans from the endpoints determined. The prediction model must be defined mathematically in the standard operation procedure (SOP) of the test that will undergo experimental validation under blind conditions with coded chemicals (Balls et al. 1995a). In order to assess the limitations of a new test before it will be evaluated in a validation study, the test should be standardized in a prevalidation study with a few test chemicals in a few laboratories (Curren et al. 1995). This will ensure that the in vitro test method, including the prediction model, is robust and that the formal validation study with coded chemicals is likely to be successful. Finally, the goal of a validation study has to be defined clearly, and the management structure has to ensure that within the study the scientists who are responsible for essential tasks can conduct their duties independently from the sponsors and the managers of the study, e.g., biostatistical analysis, and the selection, coding, and shipment of the test chemicals.

The improved concept of experimental validation for regulatory purposes defined in the second Amden workshop was accepted by ECVAM, in 1995, and in 1996 by US regulatory agencies and also by the OECD (OECD 1996). After this agreement at the international level, scientists have tried to follow the ECVAM/US/OECD principles for validation in new validation trials. The improved validation concept was immediately introduced into ongoing validation studies, e.g., the ECVAM/COLIPA validation study on in vitro phototoxicity tests and the ECVAM validation study of in vitro skin corrosivity test. Taking into account the experience from successful validation studies, in 2005 the OECD has published a “guidance document on the validation and international acceptance of new or updated test methods for hazard assessment” (OECD 2005).

Example of the Successful Validation and Regulator Acceptance of a New Test Method

Since no standard guideline for the testing of photoirritation potential, either in vivo or in vitro, for regulatory purposes existed at the international level in 1991, the OECD, the European Commission (EC), and the European Cosmetics, Toiletry, and Perfumery Association (COLIPA) established a joint program to develop and validate in vitro photoirritation tests. In the first phase of the study, which was funded by GD Research and Technology of the EU Commission and coordinated by ZEBET (Zentralstelle zur Erfassung und Bewertung von Ersatz- und Ergänzungsmethoden zum Tierversuch, the national German validation center), in vitro phototoxicity tests established in laboratories of the cosmetics industry were evaluated and also a new assay, the 3T3 NRU PT test, which is a photocytotoxicity test using the mouse fibroblast cell line 3T3 and neutral red uptake (NRU) as the endpoint for cytotoxicity.

In the prevalidation study conducted with 20 test chemicals, the 3T3 NRU PT in vitro phototoxicity test was the only in vitro test in which all of the 20 test chemicals were correctly identified as phototoxic or non-phototoxic. Quite independently, a laboratory in Japan subsequently obtained the same correct results in the 3T3 NRU PT, when testing the same set of 20 test chemicals. In the second phase of the study, which was funded by ECVAM and coordinated by ZEBET, the 3T3 NRU PT test was validated with 30 carefully selected test chemicals in 11 laboratories in a blind trial. A representative set of test chemicals covering all major classes of phototoxins was selected according to results from standardized photopatch testing in humans. The results obtained in this in vitro test under blind conditions were reproducible, and the correlation between in vitro and in vivo data was almost perfect (Spielmann et al. 1998a).

Therefore, the ECVAM Scientific Advisory Committee (ESAC) concluded in 1998 that the 3T3 NRU PT is a scientifically validated test which is ready to be considered for regulatory acceptance (ESAC 1998). However, the EU expert committee on the safety of cosmetics, the Scientific Committee on Cosmetology and Non-Food Products (SCCNFP), criticized that an insufficient number of UV filter chemicals (widely used as sun blockers) were tested in the formal validation study. In a subsequent blind trial on UV filter chemicals, which was again funded by ECVAM and coordinated by ZEBET, the phototoxic potential of all test chemicals was predicted correctly in the 3T3 NRU PT in vitro phototoxicity test (Spielmann et al. 1998b). Therefore, in 1998, the EU, having accepted the 3T3 NRU PT test as the first experimentally validated in vitro toxicity test for regulatory purposes, officially applied to the OECD for worldwide acceptance of this in vitro toxicity test. Early in 2000, the European Commission has officially accepted and published the 3T3 NRU PT phototoxicity test in Annex V of Directive 67/548 EEC on the Classification, Packaging and Labelling of Dangerous Substances (EU Commission 1983). Thus, this in vitro test is the first formally validated in vitro toxicity test that has been accepted into Annex V, and it is the only phototoxicity test that is accepted for regulatory purposes in Europe. In 2002 the OECD has accepted the 3T3 NRU PT phototoxicity test at the worldwide level as the first in vitro toxicity test into the OECD guidelines for the testing of chemicals (OECD 2004).

New Developments During the Last Decade

Recent EU legislations, e.g., the seventh Amendment of the EU Cosmetics Directive (EU Commission 2003) and the EU chemical regulation REACH (EU Commission 2006), are enforcing the use of nonanimal methods to replace the use of toxicity testing in animals. In fact, in 2013 the EU Cosmetics Directive stipulates a marketing ban of cosmetic ingredients that were tested in animals. Due to public and private research funding activities in Europe, this deadline may be met as far as testing for local side effects on eye and skin is concerned. As a consequence, during the past decade, several nonanimal test methods for skin and eye irritation have successfully been developed and validated, and some of them have been accepted for regulatory purposes at the international level by the OECD. Testing of new chemicals for local toxicity can now be conducted according to the following OECD in vitro toxicity test methods (http://www.oecd-ilibrary.org/environment/oecd-guidelines-for-the-testing-of-chemicals-section-4-health-effects_20745788) shown in Table 2:

Table 2 In vitro toxicity test for local eye and skin toxicity accepted by the OECD for regulatory purposes (classification and labelling)

In addition for eye irritation, several “human cornea construct” in vitro models are currently undergoing prevalidation and also a tiered in vitro testing scheme and prediction model for the detection of skin sensitizers based on in vitro assays, addressing three different steps in the development of skin sensitization, an in silico method, a protein reactivity assay, and a dendritic cell activation assay. A prevalidation study with 54 chemicals showed a high predictivity for sensitization potential in humans, which was better than that of the in vivo local lymph node assay (LLNA) (Bauch et al. 2012).

In 2012 the OECD has also accepted two in vitro hormone receptor assays for assessing the estrogenic and antiestrogenic potential of chemicals (http://www.oecd-ilibrary.org/environment/oecd-guidelines-for-the-testing-of-chemicals-section-4-health-effects_20745788):

• 1. OECD TG 455 the “Performance-based test guideline for stably transfected transactivation in vitro assays to detect estrogen receptor agonists”

• 2. OECD TG 457 the “BG1Luc estrogen receptor transactivation test method for identifying estrogen receptor agonists and antagonists”

The past two decades have seen unprecedented scientific and technological advances, including the birth of functional genomics, the fast-paced growth of computing power and computational biology/bioinformatics, the establishment of robotic platforms for high-throughput screening of chemicals, and the sequencing of the human genome. Together, these advances have triggered a revolution in molecular biology and have made available a wide range of new tools for studying the effects of chemicals on cells, tissues, and organisms in a rapid and cost-efficient manner.

This convergence of factors, coupled with increased recognition of the limitations of conventional in vivo tests and the need to evaluate the safety of an increasingly large number of chemical substances and mixtures, has led authorities such as the US National Research Council and other to call for a shift in toxicity testing towards the elucidation of toxicity pathways at the cellular level – an approach commonly referred to as “Toxicity testing in the 21st century” (US National Research Council 2007). According to this concept an adverse outcome pathway (AOP) delineates the documented, plausible, and testable processes by which a chemical induces molecular perturbations and the associated biological responses which describe how the molecular perturbations cause effects at the subcellular, cellular, tissue, organ, whole animal, and population levels of observation. The AOP can then be used to form categories by integrating knowledge of how chemicals interact with biological systems (i.e., the molecular initiating events) with knowledge of the biological responses. Taking into account the new concept, an OECD expert group is working on developing an in vitro sensitization test based on the adverse outcome pathway (AOP) approach.

Since the new concept of toxicity pathways has been welcomed by scientists in academia, industry, and regulatory agencies, the OECD is proposing that all new test methods should take into account the AOP concept and it should also be considered when existing tests are updated.