Richard Billington

Acute toxicity testing of chemicals—Opportunities to avoid redundant testing and use alternative approaches

Assessment of the acute systemic oral, dermal, and inhalation toxicities, skin and eye irritancy, and skin sensitisation potential of chemicals is required under regulatory schemes worldwide. In vivo studies conducted to assess these endpoints can sometimes be associated with substantial adverse effects in the test animals, and their use should always be scientifically justified. It has been argued that while information obtained from such acute tests provides data needed to meet classification and labelling regulations, it is of limited value for hazard and risk assessments. Inconsistent application of in vitro replacements, protocol requirements across regions, and bridging principles also contribute to unnecessary and redundant animal testing. Assessment of data from acute oral and dermal toxicity testing demonstrates that acute dermal testing rarely provides value for hazard assessment purposes when an acute oral study has been conducted. Options to waive requirements for acute oral and inhalation toxicity testing should be employed to avoid unnecessary in vivo studies. In vitro irritation models should receive wider adoption and be used to meet regulatory needs. Global requirements for sensitisation testing need continued harmonisation for both substance and mixture assessments. This paper highlights where alternative approaches or elimination of tests can reduce and refine animal use for acute toxicity requirements.

Recognition of Adverse and Nonadverse Effects in Toxicity Studies

One of the most important quantitative outputs from toxicity studies is identification of the highest exposure level (dose or concentration) that does not cause treatment related effects that could be considered relevant to human health risk assessment. A review of regulatory and other scientifi c literature and of current practices has revealed a lack of consistency in definition and application of frequently used terms such as No Observed Effect Level (NOEL), No Observed Adverse Effect Level (NOAEL), adverse effect, biologically significant effect, or toxicologically significant effect. Moreover, no coherent criteria were found that could be used to guide consistent interpretation of toxicity studies, including the recognition and differentiation between adverse and nonadverse effects. This presentation will address these issues identified first by proposing a standard set of definitions for key terms such as NOEL and NOAEL that are frequently used to describe the overall outcome of a toxicity study. Second, a coherent framework is outlined that can assist the toxicologist in arriving at consistent study interpretation. This structured process involves two main steps. In the first, the toxicologist must decide whether differences from control values are treatment related or if they are chance deviations. In the second step, only those differences judged to be effects are further evaluated in order to discriminate between those that are adverse and those that are not. For each step, criteria are described that can be used to make consistent judgments. In differentiating an effect from a chance finding, consideration is given inter alia to dose response, spurious measurements in individual parameters, the precision of the measurement under evaluation, ranges of natural variation and the overall biological plausibility of the observation. In discriminating between the adverse and the non-adverse effect consideration is given to: whether the effect is an adaptive response, whether it is transient, the magnitude of the effect, its association with effects in other related endpoints, whether it is a precursor to a more signifi cant effect, whether it has an effect on the overall function of the organism, whether it is a specific effect on an organ or organ system or secondary to general toxicity or whether the effect is a predictable consequence of the experimental model. In interpreting complex studies it is recognised that a weight of the evidence approach, combining the criteria outlined here to reach an overall judgment, is the optimal way of applying the process. It is believed that the use of such a scheme will help to improve the consistency of study interpretation that is the foundation of hazard and risk assessment.

Application of toxicokinetics to improve chemical risk assessment: implications for the use of animals.

While toxicokinetics has become an integral part of pharmaceutical safety assessment over the last two decades, its use in the chemical industry is relatively new. However, it is recognised as a potentially important tool in human health risk assessment and recent initiatives have advocated greater application of toxicokinetics as part of an improved assessment strategy for crop protection chemicals that could offer greater efficiency, use fewer animals and provide better data for risk assessment purposes. To explore the potential scientific and animal welfare benefits of increased use of toxicokinetic data across the chemical industry, an international workshop was held in 2008. Experts from a wide range of chemical industry sectors, including industrial chemicals, agrochemicals and consumer products, participated in the meeting as well as representatives from relevant regulatory authorities. Pharmaceutical industry experts were also invited, in order to share experiences from the extensive use of toxicokinetics in drug development. Given that increased generation of toxicokinetic data could potentially result in an increased number of animals undergoing testing, technologies and strategies to reduce and refine animal use for this purpose were also considered. This paper outlines and expands upon the key themes that emerged from the workshop.

Assessment of diurnal systemic dose of agrochemicals in regulatory toxicity testing--an integrated approach without additional animal use.

Integrated toxicokinetics (TK) data provide information on the rate, extent and duration of systemic exposure across doses, species, strains, gender, and life stages within a toxicology program. While routine for pharmaceuticals, TK assessments of non-pharmaceuticals are still relatively rare, and have never before been included in a full range of guideline studies for a new agrochemical. In order to better understand the relationship between diurnal systemic dose (AUC(24h)) and toxicity of agrochemicals, TK analyses in the study animals is now included in all short- (excluding acute), medium- and long-term guideline mammalian toxicity studies including reproduction/developmental tests. This paper describes a detailed procedure for the implementation of TK in short-, medium- and long-term regulatory toxicity studies, without the use of satellite animals, conducted on three agrochemicals (X11422208, 2,4-D and X574175). In these studies, kinetically-derived maximum doses (KMD) from short-term studies instead of, or along with, maximum tolerated doses (MTD) were used for the selection of the high dose in subsequent longer-term studies. In addition to leveraging TK data to guide dose level selection, the integrated program was also used to select the most appropriate method of oral administration (i.e., gavage versus dietary) of test materials for rat and rabbit developmental toxicity studies. The integrated TK data obtained across toxicity studies (without the use of additional/satellite animals) provided data critical to understanding differences in response across doses, species, strains, sexes, and life stages. Such data should also be useful in mode of action studies and to improve human risk assessments.

Use of toxicokinetics to support chemical evaluation: Informing high dose selection and study interpretation

Toxicokinetic (TK) information can substantially enhance the value of the data generated from toxicity testing, and is an integral part of pharmaceutical safety assessment. It is less widely used in the chemical, agrochemical and consumer products industries, but recognition of its value is growing, as reflected by increased reference to the use of TK information in new and draft OECD test guidelines. To help promote increased consideration of the important role TK can play in chemical risk assessment, we have gathered practical examples from the peer-reviewed literature, as well as in-house industry data, that highlight opportunities for the use of TK in the selection of dose levels. Use of TK can help to ensure studies are designed to be of most relevance to assessing potential risk in humans, and avoid the use of excessively high doses that could result in unnecessary suffering in experimental animals. Greater emphasis on the potential contribution of TK in guiding study design and interpretation should be incorporated in regulatory data requirements and associated guidance.

A 1-year toxicity study in dogs is no longer a scientifically justifiable core data requirement for the safety assessment of pesticides.

A review of publications on pesticides assessing the need for 1-year toxicity studies in dogs was performed. Four key peer-reviewed papers with different approaches investigated the value of a 1-year dog study in addition to a 3-month study. Despite different databases and approaches, each concluded with the recommendation to limit the testing of pesticides in dogs to a duration of 3 months. The combined weight of evidence presented in this review reinforces these independent conclusions. Therefore, the routine inclusion of a 1-year dog study as a mandated regulatory requirement for the safety assessment of pesticides is no longer justifiable and a globally harmonized approach should be taken to match the latest legislation of the European Union and the US EPA.

The mouse carcinogenicity study is no longer a scientifically justifiable core data requirement for the safety assessment of pesticides

Regulatory tests investigating pesticide carcinogenicity potential routinely comprise a battery of in vitro and in vivo genotoxicity studies and two cancer bioassays, one in rats and one in mice. The genotoxicity testing strategy essentially ensures that genotoxic compounds are eliminated, and any carcinogens identified in subsequent lifetime studies are probably nongenotoxic in character. Assessment of 202 pesticide evaluations from the European Union review programme under Directive 91/414/EEC indicated that the mouse carcinogenicity study contributed little or nothing to either derivation of an acceptable daily intake (ADI) for assessment of chronic risk to humans, or hazard classification for labelling purposes. From a pesticide approval perspective, the mouse study did not influence a single outcome. From a risk assessment perspective, the ADI for just one pesticide was based on tumours in mice and this would have barely changed if the mouse data had not been available. In total, only 10 (5%) pesticide ADIs were based solely on the mouse carcinogenicity study and even in these few cases, a similar value would have been identified from other studies if the mouse study had not been available. For pesticides with treatment-related tumours only in mice, just three, or 1.5%, were classified as carcinogens and all were in the lowest category, Category 3 (R40). For pesticides with treatment-related tumours in mice and rats, the mouse data were probably the main, if not the only, cause for another three cases of R40 classification. Absence of the mouse studies would not have influenced assignment of the higher, Category 2 (R45), cancer classification for any substance with treatment-related tumours in both species as all decisions for these substances were limited to Category 3 or ‘unclassified’ outcomes. Over 100,000 mice were used to test these pesticides. This review shows that the mouse carcinogenicity studies did not provide significant information over and above that provided by the rat studies, and underpins the opportunity, from both a scientific and an animal welfare perspective, to remove the mouse carcinogenicity study from regulatory data requirements for the testing of pesticides.

Feasibility of the extended one-generation reproductive toxicity study (OECD 443).

The extended one-generation reproduction toxicity study (OECD 443, adopted 28-July-2011) produces more information with fewer animals than the two-generation study (OECD 416), by including F1 neurotoxicity and immunotoxicity assessments, and omitting an F2 generation if there are no relevant F1 findings. This saves >1000 animals per compound. Feasibility studies based on draft OECD443 were conducted in industrial GLP laboratories in Europe and USA, using vinclozolin, methimazole and lead acetate. A fourth study was conducted with 2,4-dichlorophenoxyacetic acid (2,4-D) in response to a regulatory request for reproduction and developmental neurotoxicity data. The studies effectively profiled vinclozolin as an anti-androgenic developmental toxicant, methimazole as a developmental anti-thyroid agent, and lead acetate as a systemic and developmental toxicant. The 2,4-D study demonstrated the value of toxicokinetic data in dose setting and data interpretation. These results illustrate the variety of reproductive and developmental endpoints which can be captured in this complex but manageable study design. Time constraints for triggering further (F2) testing are summarized.

An integrated approach for prospectively investigating a mode-of-action for rodent liver effects

Registration of new plant protection products (e.g., herbicide, insecticide, or fungicide) requires comprehensive mammalian toxicity evaluation including carcinogenicity studies in two species. The outcome of the carcinogenicity testing has a significant bearing on the overall human health risk assessment of the substance and, consequently, approved uses for different crops across geographies. In order to understand the relevance of a specific tumor finding to human health, a systematic, transparent, and hypothesis-driven mode of action (MoA) investigation is, appropriately, an expectation by the regulatory agencies. Here, we describe a novel approach of prospectively generating the MoA data by implementing additional end points to the standard guideline toxicity studies with sulfoxaflor, a molecule in development. This proactive MoA approach results in a more robust integration of molecular with apical end points while minimizing animal use. Sulfoxaflor, a molecule targeting sap-feeding insects, induced liver effects (increased liver weight due to hepatocellular hypertrophy) in an initial palatability probe study for selecting doses for subsequent repeat-dose dietary studies. This finding triggered the inclusion of dose-response investigations of the potential key events for rodent liver carcinogenesis, concurrent with the hazard assessment studies. As predicted, sulfoxaflor induced liver tumors in rats and mice in the bioassays. The MoA data available by the time of the carcinogenicity finding supported the conclusion that the carcinogenic potential of sulfoxaflor was due to CAR/PXR nuclear receptor activation with subsequent hepatocellular proliferation. This MoA was not considered to be relevant to humans as sulfoxaflor is unlikely to induce hepatocellular proliferation in humans and therefore would not be a human liver carcinogen.

Human relevance framework for rodent liver tumors induced by the insecticide sulfoxaflor

Abstract Sulfoxaflor, a novel active substance that targets sap-feeding insects, induced rodent hepatotoxicity when administered at high dietary doses. Specifically, hepatocellular adenomas and carcinomas increased after 18 months in male and female CD-1 mice at 750 and 1250 ppm, respectively, and hepatocellular adenomas increased after 2 years in male F344 rats at 500 ppm. Studies to determine the mode of action (MoA) for these liver tumors were performed in an integrated and prospective manner as part of the standard battery of toxicology studies such that the MoA data were available prior to, or by the time of, the completion of the carcinogenicity studies. Sulfoxaflor is not genotoxic and the MoA data support the following key events in the etiology of the rodent liver tumors: (1) CAR nuclear receptor activation and (2) hepatocellular proliferation. The MoA data were evaluated in a weight of evidence approach using the Bradford Hill criteria for causation and were found to align with dose and temporal concordance, biological plausibility, coherence, strength, consistency, and specificity for a CAR-mediated MoA while excluding other alternate MoAs. The available data include: activation of CAR, Cyp2b induction, hepatocellular hypertrophy and hyperplasia, absence of liver effects in KO mice, absence of proliferation in humanized mice, and exclusion of other possible mechanisms (e.g., genotoxicity, cytotoxicity, AhR, or PPAR activation), and indicate that the identified rodent liver tumor MoA for sulfoxaflor would not occur in humans. In this case, sulfoxaflor is considered not to be a potential human liver carcinogen.