Regulating human safety: How dose selection in toxicity studies impacts human health hazard assessment and subsequent risk management options.

In the EU, one of the key determinants in the regulation and management of substances to ensure adequate protection of human health is the outcome of toxicity studies. These studies should therefore be performed in a way that the data generated are adequate to fulfil all regulatory requirements. However, in recent years, an increasing number of toxicity studies use dose levels that induce only slight, or even no toxicity, while the top dose lies well below the limit dose of 1000 mg/kg bw/d. The results of these studies have limited value for the hazard and subsequent risk assessment and risk management of substances. This paper shows why conducting toxicity studies with too low doses has severe consequences for among others classification and labelling, identification of endocrine disruptors, health impact assessment, and incident management. With this paper we aim to raise awareness on this issue and want to stress the importance of the use of sufficiently high dosing in toxicity studies. Given their central role in toxicity testing, it is therefore key to adapt where necessary the descriptions in OECD test guidelines and guidance documents on requirements for dose level setting, to make sure they are as explicit and unambiguous as possible.

Use of the kinetically-derived maximum dose concept in selection of top doses for toxicity studies hampers proper hazard assessment and risk management.

The KMD (kinetically-derived maximum dose) is an increasingly advocated concept that uses toxicokinetic data in the top dose selection for toxicity testing. Application of this concept may have serious regulatory implications though, especially in the European Union. The basic assumption is that the relationship between internal and external dose (IED) shows an inflection point where linearity transits into non-linearity due to saturation of underlying processes; top doses in toxicity tests should not be above the inflection point, provided human exposures are well below this point. A critical analysis of the KMD concept and its underlying assumptions shows, however, that the IED relationship is non-linear over the whole dose range, without any point of inflection. The KMD concept thus aims to estimate a non-existing point, rendering it invalid for use in toxicity testing. Moreover, the concept ignores the key question in toxicology: What kind of toxic effects occur at which doses? These and several other reservations against the KMD concept are discussed and illustrated with three existing applications of the KMD approach. Hence, we recommend to abolish the KMD concept for selecting top doses in toxicity testing. This requires the updating of regulations, guidance documents and OECD test guidelines.

Why Do Countries Regulate Environmental Health Risks Differently? A Theoretical Perspective.

Why do countries regulate, or prefer to regulate, environmental health risks such as radiofrequency electromagnetic fields and endocrine disruptors differently? A wide variety of theories, models, and frameworks can be used to help answer this question, though the resulting answer will strongly depend on the theoretical perspective that is applied. In this theoretical review, we will explore eight conceptual frameworks, from different areas of science, which will offer eight different potential explanations as to why international differences occur in environmental health risk management. We are particularly interested in frameworks that could shed light on the role of scientific expertise within risk management processes. The frameworks included in this review are the Risk Assessment Paradigm, research into the roles of experts as policy advisors, the Psychometric Paradigm, the Cultural Theory of Risk, participatory approaches to risk assessment and risk management, the Advocacy Coalition Framework, the Social Amplification of Risk Framework, and Hofstede's Model of National Cultures. We drew from our knowledge and experiences regarding a diverse set of academic disciplines to pragmatically assemble a multidisciplinary set of frameworks. From the ideas and concepts offered by the eight frameworks, we derive pertinent questions to be used in further empirical work and we present an overarching framework to depict the various links that could be drawn between the frameworks.

An integrative test strategy for cancer hazard identification.

Assessment of genotoxic and carcinogenic potential is considered one of the basic requirements when evaluating possible human health risks associated with exposure to chemicals. Test strategies currently in place focus primarily on identifying genotoxic potential due to the strong association between the accumulation of genetic damage and cancer. Using genotoxicity assays to predict carcinogenic potential has the significant drawback that risks from non-genotoxic carcinogens remain largely undetected unless carcinogenicity studies are performed. Furthermore, test systems already developed to reduce animal use are not easily accepted and implemented by either industries or regulators. This manuscript reviews the test methods for cancer hazard identification that have been adopted by the regulatory authorities, and discusses the most promising alternative methods that have been developed to date. Based on these findings, a generally applicable tiered test strategy is proposed that can be considered capable of detecting both genotoxic as well as non-genotoxic carcinogens and will improve understanding of the underlying mode of action. Finally, strengths and weaknesses of this new integrative test strategy for cancer hazard identification are presented.

International regulatory needs for development of an IATA for non-genotoxic carcinogenic chemical substances.

Although regulatory requirements for carcinogenicity testing of chemicals vary according to product sector and regulatory jurisdiction, the standard approach starts with a battery of genotoxicity tests. If any of the in vivo genotoxicity tests are positive, a lifetime rodent cancer bioassay may be requested, which allows the detection of non-genotoxic carcinogens (NGTxC). However, under most chemical regulations the cancer bioassay is rarely requested, specific requests to obtain information on non-genotoxic mechanisms of carcinogenicity are few, and there are no OECD approved screening methods. When the in vitro genotoxicity battery is negative, usually no further carcinogenicity testing is requested. Consequently NGTxC might remain unidentified and therefore the risks they may pose to human health will not be managed. In contrast to genotoxic carcinogens NGTxCact through a large variety of specific mechanisms, and a panel of tests covering multiple biological traits will be needed. The development of an Integrated Approach to Testing and Assessment (IATA) of NGTxC could assist regulatory decision makers. We examine what NGTxC are and discuss chemical regulatory requirements and limitations. With a strong drive to reduce animal testing and costs in mind, it is essential that proper and robust alternatives for animal testing (3Rs) methods for addressing non-genotoxic modes of action are developed and used. Therefore relevant in vitro mechanisms and assays are described and tentatively organized in levels of information, indicating both a possible structure of the future IATA for NGTxC and associated OECD Test Guideline development priorities.

Evaluation of the performance of the reduced local lymph node assay for skin sensitization testing.

The local lymph node assay (LLNA) is the preferred method for classification of sensitizers within REACH. To reduce the number of mice for the identification of sensitizers the reduced LLNA was proposed, which uses only the high dose group of the LLNA. To evaluate the performance of this method for classification, LLNA data from REACH registrations were used and classification based on all dose groups was compared to classification based on the high dose group. We confirmed previous examinations of the reduced LLNA showing that this method is less sensitive compared to the LLNA. The reduced LLNA misclassified 3.3% of the sensitizers identified in the LLNA and misclassification occurred in all potency classes and that there was no clear association with irritant properties. It is therefore not possible to predict beforehand which substances might be misclassified. Another limitation of the reduced LLNA is that skin sensitizing potency cannot be assessed. For these reasons, it is not recommended to use the reduced LLNA as a stand-alone assay for skin sensitization testing within REACH. In the future, the reduced LLNA might be of added value in a weight of evidence approach to confirm negative results obtained with non-animal approaches.

On the impact of second generation mating and offspring in multi-generation reproductive toxicity studies on classification and labelling of substances in Europe.

The possible impact on classification and labelling decisions of effects observed in second generation parental (P1) and offspring (F2) parameters in multi-generation studies was investigated. This was done for 50 substances classified as reproductive toxicants in Europe, for which a multi-generation study was available. The P1 and F2 effects were compared to parental (P0) and first generation offspring (F1) effects with regard to type of effect as well as incidence, magnitude and severity (IMS), at any dose level. For every study with unique P1/F2 effects, or differences in IMS, the influence of the P1/F2 findings on the classification decision was investigated. Unique P1/F2 generation findings did not play a crucial role in the classification decision of any of the 50 classified substances, except for fenarimol. This substance however provided abundant alerts on the basis of its endocrine activity and developmental neurotoxicity and would therefore also be expected to be identified as a developmental neurotoxicant in an Extended One Generation Reproductive Toxicity Study (EOGRTS). These findings, in addition to the increased number of parameters analysed, increased statistical power and reduced animal use, provide strong further support for replacement of the classical two-generation reproductive toxicity study by the EOGRTS in regulatory reproductive toxicity assessment.

Finding maximal transcriptome differences between reprotoxic and non-reprotoxic phthalate responses in rat testis.

The chemical legislation of the EU, Registration, Evaluation, and Authorization of Chemicals (REACH), stipulates that about 30 000 chemical substances are to be assessed on their possible risks. Toxicological evaluation of these compounds will at least partly be based on animal testing. In particular, the assessment of reproductive toxicity is a very complicated, time-consuming and animal-demanding process. Introducing microarray-based technologies can potentially refine in vivo toxicity testing. If compounds of a distinct chemical class induce reproducible gene-expression responses with a recognizable overlap, these gene-expression signatures may indicate intrinsic features of certain compounds, including specific toxicity. In the present study, we have set out the first steps towards this approach for the reproductive toxicity of phthalates. Male rats were treated with a single dose of either reprotoxic or non-reprotoxic phthalates, and were analyzed 24 h afterwards. Subsequently, histopathological and gene-expression profiling analyses were performed. Despite ambiguous histopathological observations, we were able to identify genes with differential expression profiles between the reprotoxic phthalates and the non-reprotoxic counterparts. This shows that differences in gene-expression profiles, indicative of the type of exposure, may be detected earlier, or at lower doses, than classical pathological endpoints. These findings are promising for 'early warning' biomarker analyses and for using toxicogenomics in a category approach. Ultimately, this could lead to a more cost-effective approach for prioritizing the toxicity testing of large numbers of chemicals in a short period of time in hazard assessment of chemicals, which is one of the objectives of the REACH chemical legislation.

Retrospective analysis of relative parameter sensitivity in multi-generation reproductive toxicity studies.

Current suggestions towards amending the OECD two-generation protocol include omission of the second generation and inclusion of additional parameters. This study analysed the relative parameter sensitivity in 18 individually published multi-generation studies with substances toxic to fertility. Among parameters that most often determined the reproductive LOAEL were weight of testis, dam and pup as well as litter size. Several other parameters were found to be unaffected in all studies evaluated. Some substances affected a specific set of parameters, indicating that rarely affected parameters may prove crucial in individual situations. This argues for the inclusion of a wide spectrum of parameters to cover all possible effects. Less sensitive parameters, mechanistically related to more sensitive ones, may be omitted as they will unlikely contribute to the overall LOAEL. This study gives first insights and needs follow-up by more extensive analyses before firm conclusions on the design of the two-generation study protocol can be drawn.

A retrospective analysis of developmental toxicity studies in rat and rabbit: what is the added value of the rabbit as an additional test species?

In contrast to most toxicological tests, developmental studies are usually required in both a rodent and a non-rodent species. This study retrospectively assessed the added value of the rabbit developmental test when a rat developmental test is available. In contrast with previous reviews, we looked at developmental toxicity instead of teratogenicity, and took into account maternal toxicity in the evaluation of developmental toxicity. We analyzed data for 54 substances classified for developmental toxicity and 73 substances considered to be teratogenic in the rabbit and not in the rat in two previous reviews. On average, the rat and the rabbit developmental toxicity studies were similarly sensitive: the average ratio of the NOAELs between the two species was about one, and for most compounds there were no differences between rat and rabbit studies in terms of classification for developmental toxicity. For certain substances the developmental study in either one of the two species appeared to be more sensitive than in the other species. However, these differences are partly due to differences between studies other than the test species used. Overall, our analysis does not clearly indicate that the evaluation of developmental toxicity, as opposed to other types of toxicity, would specifically require the rabbit as an additional test species. The discrimination between direct and indirect (i.e., as a consequence of maternal toxicity) developmental effects was often doubtful, and is one of the factors that could explain the apparent differences between the two species. A more accurate assessment of maternal toxicity might improve the reliability of the results from a single developmental toxicity study. More knowledge about the interaction between maternal and developmental effects is required before decisions on omitting the requirement for the developmental toxicity testing in a second species can be considered.

Quantitative extrapolation of in vitro whole embryo culture embryotoxicity data to developmental toxicity in vivo using the benchmark dose approach.

If in vitro data are to be used as a basis for hazard characterization, a translation of an in vitro concentration toward an in vivo dose must be made. In this study we examined the correlation between dose descriptors from the in vitro Whole Embryo Culture (WEC) test and in vivo developmental toxicity tests. We applied the Benchmark Dose (BMD) approach to estimate equipotent in vitro concentrations (Benchmark Concentrations [BMCs]) and equipotent in vivo doses (BMDs). Using the data generated in an European Center for the Validation of Alternative Methods validation study we found that the BMCs were highly reproducible among laboratories. The three endpoints analyzed (head length, crown-rump length, and total morphological score) were strongly correlated. A clear in vitro-in vivo correlation was found between BMCs and BMDs. However, a considerable uncertainty would remain if the BMDs were estimated from the BMC using this correlation: the confidence interval of such an in vivo dose estimate would span various orders of magnitude. Differences in toxicokinetic properties among the compounds explained at least part of the scatter of the in vitro-in vivo correlation. But also heterogeneity in the design of the available in vivo studies underlies much of the scatter, and this puts a limit on validating in vitro data as predictors of in vivo data. Further analysis of the in vitro-in vivo correlation would therefore require high-quality in vivo data, generated by appropriate (and similar) study designs.

A retrospective analysis of the added value of the rat two-generation reproductive toxicity study versus the rat subchronic toxicity study.

This study aims to evaluate the added value of the two-generation reproductive toxicity study when a subchronic study (90-day repeated dose toxicity study) is available. The analysis includes a total of 47 reproductive toxic and 75 non-reproductive toxic substances, for which a two-generation study was available. For each of these compounds the outcomes of both study types were compared, in view of the question what the impact would have been both for the derived NOAEL and for classification regarding toxicity to fertility. On average, only a small difference (less than twofold) in overall NOAELs was found between the rat two-generation study and the rat subchronic study. For individual compounds the differences could be larger (up to around a factor of 10), but differences of this magnitude equally occur between NOAELs of subchronic studies (testing the same substance). The two generation study did have an impact on classification for toxicity to fertility: about one-third of the substances shown to be toxic to fertility in the two-generation study did not show any sign of that in the 90-day study. If the subchronic study did show toxicity to reproductive organs this often occurred at (much) higher doses than other toxic effects in the same study. Therefore, apart from including more fertility endpoints, a larger dose spacing (or more dose groups) in the subchronic study might increase its detection rate of fertility toxic substances. The consequences that these findings may have for risk assessment and risk management are discussed, especially in the context of REACH.

A retrospective analysis of the two-generation study: what is the added value of the second generation?

Increasing pressure is exerted by some stakeholders to reduce the two-generation study to a one-generation study, a measure that would considerably reduce the number of animals and other costs involved in these lengthy studies. The present study retrospectively evaluates 176 multi-generation studies to assess potential differences between the first and the second generation, both in terms of the types of effects observed and in terms of the effective doses. All substances classified as reproductive toxicants by the Directive 92/32/EEC or considered as toxic to fertility by the California EPA for which we found a multi-generation study were included (n=58 studies). The second generation in the two-generation studies considered affected neither the overall NOAEL nor the critical effect. Therefore, it had no impact on the ensuing risk assessment, nor on classification and labeling. However, several substances did show an increased sensitivity of the F(1) adults in comparison to the P(0). These results support the proposal of replacing the current two-generation study by a one-generation study with a more extensive assessment of parameters at F(1) adulthood.

Comparative in vitro-in vivo percutaneous penetration of the fungicide ortho-phenylphenol.

The validity of in vitro and in vivo methods for the prediction of percutaneous penetration in humans was assessed using the fungicide ortho-phenylphenol (OPP) (log Po/w 3.28, MW 170.8, solubility in water 0.7 g/L). In vivo studies were performed in rats and human volunteers, applying the test compound to the dorsal skin and the volar aspect of the forearm, respectively. In vitro studies were performed using static diffusion cells with viable full-thickness skin membranes (rat and human), nonviable epidermal membranes (rat and human), and a perfused pig ear model. For the purpose of conducting in vitro/in vivo comparisons, standardized experimental conditions were used with respect to dose (120 microg OPP/cm(2)), vehicle (60% aqueous ethanol), and exposure duration (4 h). In human volunteers, the potentially absorbed dose (amount applied minus dislodged) was 105 microg/cm(2), while approximately 27% of the applied dose was excreted with urine within 48 h. In rats these values were 67 microg/cm(2) and 40%, respectively. In vitro methods accurately predicted human in vivo percutaneous absorption of OPP on the basis of the potential absorbed dose. With respect to the other parameters studied (amount systemically available, maximal flux), considerable differences were observed between the various in vitro models. In viable full-thickness skin membranes, the amount systemically available and the potentially absorbed dose correlated reasonably well with the human in vivo situation. In contrast the K(p)/maximal flux considerably underestimated the human in vivo situation. Although epidermal membranes overestimated human in vivo data, the species differences observed in vivo were reflected correctly in this model. The data generated in the perfused pig ear model were generally intermediate between viable skin membranes and epidermal membranes.