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.

A Method for Comparing the Impact on Carcinogenicity of Tobacco Products: A Case Study on Heated Tobacco Versus Cigarettes.

Comparing the harmful health effects related to two different tobacco products by applying common risk assessment methods to each individual compound is problematic. We developed a method that circumvents some of these problems by focusing on the change in cumulative exposure (CCE) of the compounds emitted by the two products considered. The method consists of six steps. The first three steps encompass dose-response analysis of cancer data, resulting in relative potency factors with confidence intervals. The fourth step evaluates emission data, resulting in confidence intervals for the expected emission of each compound. The fifth step calculates the change in CCE, probabilistically, resulting in an uncertainty range for the CCE. The sixth step estimates the associated health impact by combining the CCE with relevant dose-response information. As an illustrative case study, we applied the method to eight carcinogens occurring both in the emissions of heated tobacco products (HTPs), a novel class of tobacco products, and tobacco smoke. The CCE was estimated to be 10- to 25-fold lower when using HTPs instead of cigarettes. Such a change indicates a substantially smaller reduction in expected life span, based on available dose-response information in smokers. However, this is a preliminary conclusion, as only eight carcinogens were considered so far. Furthermore, an unfavorable health impact related to HTPs remains as compared to complete abstinence. Our method results in useful information that may help policy makers in better understanding the potential health impact of new tobacco and related products. A similar approach can be used to compare the carcinogenicity of other mixtures.

Testing developmental toxicity in a second species: are the differences due to species or replication error?

Developmental toxicity studies for chemical and pharmaceutical safety are primarily performed in rats. Regulatory frameworks may require testing in a second, non-rodent species, for which the rabbit is usually chosen. This study shows that differences in NOAELs or LOAELs (N(L)OAELs) observed between rat and rabbit developmental toxicity studies performed according to OECD guidelines could just as well be caused by study replication errors, and not necessarily by differences in species sensitivity. This conclusion follows from an analysis of a database with rat and rabbit developmental toxicity studies for over 1000 industrial chemicals, pesticides, veterinary drugs and human pharmaceuticals, which included 143 compounds with multiple oral rat studies and 124 compounds with multiple oral rabbit studies. Our analysis confirms earlier findings that, on average over all compounds, rat and rabbit do not differ in sensitivity to developmental effects. There is substantial scatter in the correlation plots comparing rat and rabbit developmental N(L)OAELs, which is easily interpreted as species differences for individual compounds. However, for compounds tested twice in the same species, these N(L)OAELs may differ up to a factor of 25. Thus, potential interspecies differences in developmental N(L)OAEL will be overwhelmed by the reproducibility error, rendering the added value of a second species study questionable. As N(L)OAELs serve as point of departure (POD) for setting health-based guidance values in risk assessment, the large reproducibility error of N(L)OAELs should be taken into account by the introduction of an additional uncertainty factor. It is recommended to aim for reducing the reproducibility error by applying dose-response (BMD) analysis, optimize study designs and harmonize study protocols.

Is current risk assessment of non-genotoxic carcinogens protective?

Non-genotoxic carcinogens (NGTXCs) do not cause direct DNA damage but induce cancer via other mechanisms. In risk assessment of chemicals and pharmaceuticals, carcinogenic risks are determined using carcinogenicity studies in rodents. With the aim to reduce animal testing, REACH legislation states that carcinogenicity studies are only allowed when specific concerns are present; risk assessment of compounds that are potentially carcinogenic by a non-genotoxic mode of action is usually based on subchronic toxicity studies. Health-based guidance values (HBGVs) of NGTXCs may therefore be based on data from carcinogenicity or subchronic toxicity studies depending on the legal framework that applies. HBGVs are usually derived from No-Observed-Adverse-Effect-Levels (NOAELs). Here, we investigate whether current risk assessment of NGTXCs based on NOAELs is protective against cancer. To answer this question, we estimated Benchmark doses (BMDs) for carcinogenicity data of 44 known NGTXCs. These BMDs were compared to the NOAELs derived from the same carcinogenicity studies, as well as to the NOAELs derived from the associated subchronic studies. The results lead to two main conclusions. First, a NOAEL derived from a subchronic study is similar to a NOAEL based on cancer effects from a carcinogenicity study, supporting the current practice in REACH. Second, both the subchronic and cancer NOAELs are, on average, associated with a cancer risk of around 1% in rodents. This implies that for those chemicals that are potentially carcinogenic in humans, current risk assessment of NGTXCs may not be completely protective against cancer. Our results call for a broader discussion within the scientific community, followed by discussions among risk assessors, policy makers, and other stakeholders as to whether or not the potential cancer risk levels that appear to be associated with currently derived HBGVs of NGXTCs are acceptable.

Head skeleton malformations in zebrafish (Danio rerio) to assess adverse effects of mixtures of compounds.

The EU-EuroMix project adopted the strategy of the European Food Safety Authority (EFSA) for cumulative risk assessment, which limits the number of chemicals to consider in a mixture to those that induce a specific toxicological phenotype. These so-called cumulative assessment groups (CAGs) are refined at several levels, including the target organ and specific phenotype. Here, we explore the zebrafish embryo as a test model for quantitative evaluation in one such CAG, skeletal malformations, through exposure to test compounds 0-120 hpf and alcian blue cartilage staining at 120 hpf, focusing on the head skeleton. Reference compounds cyproconazole, flusilazole, metam, and thiram induced distinctive phenotypes in the head skeleton between the triazoles and dithiocarbamates. Of many evaluated parameters, the Meckel's-palatoquadrate (M-PQ) angle was selected for further assessment, based on the best combination of a small confidence interval, an intermediate maximal effect size and a gentle slope of the dose-response curve with cyproconazole and metam. Additional test compounds included in the CAG skeletal malformations database were tested for M-PQ effects, and this set was supplemented with compounds associated with craniofacial malformations or cleft palate to accommodate otherwise organized databases. This additional set included hexaconazole, all-trans-retinoic acid, AM580, CD3254, maneb, pyrimethanil, imidacloprid, pirimiphos-methyl, 2,4-dinitrophenol, 5-fluorouracil, 17alpha-ethynylestradiol (EE2), ethanol, 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), PCB 126, methylmercury, boric acid, and MEHP. Most of these compounds produced a dose-response for M-PQ effects. Application of the assay in mixture testing was provided by combined exposure to cyproconazole and TCDD through the isobole method, supporting that in this case the combined effect can be modeled through concentration addition.

A general theory of effect size, and its consequences for defining the benchmark response (BMR) for continuous endpoints.

A general theory on effect size for continuous data predicts a relationship between maximum response and within-group variation of biological parameters, which is empirically confirmed by results from dose-response analyses of 27 different biological parameters. The theory shows how effect sizes observed in distinct biological parameters can be compared and provides a basis for a generic definition of small, intermediate and large effects. While the theory is useful for experimental science in general, it has specific consequences for risk assessment: it solves the current debate on the appropriate metric for the Benchmark response in continuous data. The theory shows that scaling the BMR expressed as a percent change in means to the maximum response (in the way specified) automatically takes "natural variability" into account. Thus, the theory supports the underlying rationale of the BMR 1 SD. For various reasons, it is, however, recommended to use a BMR in terms of a percent change that is scaled to maximum response and/or within group variation (averaged over studies), as a single harmonized approach.

Estimating the carcinogenic potency of chemicals from the in vivo micronucleus test.

In this study, we investigated the applicability of using in vivo mouse micronucleus (MN) data to derive cancer potency information. We also present a new statistical methodology for correlating estimated potencies between in vivo MN tests and cancer studies, which could similarly be used for other systems (e.g. in vitro vs. in vivo genotoxicity tests). The dose-response modelling program PROAST was used to calculate benchmark doses (BMDs) for estimating the genotoxic and carcinogenic potency for 48 compounds in mice; most of the data were retrieved from the National Toxicology Program (NTP) database, while some additional data were retrieved from the Carcinogenic Potency Database and published studies. BMD05s (doses with 5% increase in MN frequency) were derived from MN data, and BMD10s (doses with 10% extra cancer risk) were derived from carcinogenicity data, along with their respective lower (BMDL) and upper (BMDU) confidence bounds. A clear correlation between the in vivo MN BMD05s and the cancer BMD10s was observed when the lowest BMD05 from the in vivo MN was plotted against the lowest BMD10 from the carcinogenicity data for each individual compound. By making a further selection of BMDs related to more or less equally severe cancer lesions, the correlation was considerably improved. Getting a general scientific consensus on how we can quantitatively compare different tumour lesion types and investigating the impact of MN study duration are needed to refine this correlation analysis. Nevertheless, our results suggest that a BMD derived from genotoxicity data might provide a prediction of the tumour potency (BMD10) with an uncertainty range spanning roughly a factor of 100.

Comparison of in vitro and in vivo clastogenic potency based on benchmark dose analysis of flow cytometric micronucleus data.

The application of flow cytometry as a scoring platform for both in vivo and in vitro micronucleus (MN) studies has enabled the efficient generation of high quality datasets suitable for comprehensive assessment of dose-response. Using this information, it is possible to obtain precise estimates of the clastogenic potency of chemicals. We illustrate this by estimating the in vivo and the in vitro potencies of seven model clastogenic agents (melphalan, chlorambucil, thiotepa, 1,3-propane sultone, hydroxyurea, azathioprine and methyl methanesulfonate) by deriving BMDs using freely available BMD software (PROAST). After exposing male rats for 3 days with up to nine dose levels of each individual chemical, peripheral blood samples were collected on Day 4. These chemicals were also evaluated for in vitro MN induction by treating TK6 cells with up to 20 concentrations in quadruplicate. In vitro MN frequencies were determined via flow cytometry using a 96-well plate autosampler. The estimated in vitro and in vivo BMDs were found to correlate to each other. The correlation showed considerable scatter, as may be expected given the complexity of the whole animal model versus the simplicity of the cell culture system. Even so, the existence of the correlation suggests that information on the clastogenic potency of a compound can be derived from either whole animal studies or cell culture-based models of chromosomal damage. We also show that the choice of the benchmark response, i.e. the effect size associated with the BMD, is not essential in establishing the correlation between both systems. Our results support the concept that datasets derived from comprehensive genotoxicity studies can provide quantitative dose-response metrics. Such investigational studies, when supported by additional data, might then contribute directly to product safety investigations, regulatory decision-making and human risk assessment.

Empirical analysis of BMD metrics in genetic toxicology part I: in vitro analyses to provide robust potency rankings and support MOA determinations.

Genetic toxicity testing has traditionally been used for hazard identification, with dichotomous classification of test results serving to identify genotoxic agents. However, the utility of genotoxicity data can be augmented by employing dose-response analysis and point of departure determination. Via interpolation from a fitted dose-response model, the benchmark dose (BMD) approach estimates the dose that elicits a specified (small) effect size. BMD metrics and their confidence intervals can be used for compound potency ranking within an endpoint, as well as potency comparisons across other factors such as cell line or exposure duration. A recently developed computational method, the BMD covariate approach, permits combined analysis of multiple dose-response data sets that are differentiated by covariates such as compound, cell type or exposure regime. The approach provides increased BMD precision for effective potency rankings across compounds and other covariates that pertain to a hypothesised mode of action (MOA). To illustrate these applications, the covariate approach was applied to the analysis of published in vitro micronucleus frequency dose-response data for ionising radiations, a set of aneugens, two mutagenic azo compounds and a topoisomerase II inhibitor. The ionising radiation results show that the precision of BMD estimates can be improved by employing the covariate method. The aneugen analysis provided potency groupings based on the BMD confidence intervals, and analyses of azo compound data from cells lines with differing metabolic capacity confirmed the influence of endogenous metabolism on genotoxic potency. This work, which is the first of a two-part series, shows that BMD-derived potency rankings can be employed to support MOA evaluations as well as facilitate read across to expedite chemical evaluations and regulatory decision-making. The follow-up (Part II) employs the combined covariate approach to analyse in vivo genetic toxicity dose-response data focussing on how improvements in BMD precision can impact the reduction and refinement of animal use in toxicological research.

Empirical analysis of BMD metrics in genetic toxicology part II: in vivo potency comparisons to promote reductions in the use of experimental animals for genetic toxicity assessment.

Genotoxicity tests have traditionally been used only for hazard identification, with qualitative dichotomous groupings being used to identify compounds that have the capacity to induce mutations and/or cytogenetic alterations. However, there is an increasing interest in employing quantitative analysis of in vivo dose-response data to derive point of departure (PoD) metrics that can be used to establish human exposure limits or margins of exposure (MOEs), thereby supporting human health risk assessments and regulatory decisions. This work is an extension of our companion article on in vitro dose-response analyses and outlines how the combined benchmark dose (BMD) approach across included covariates can be used to improve the analyses and interpretation of in vivo genetic toxicity dose-response data. Using the BMD-covariate approach, we show that empirical comparisons of micronucleus frequency dose-response data across multiple studies justifies dataset merging, with subsequent analyses improving the precision of BMD estimates and permitting attendant potency ranking of seven clastogens. Similarly, empirical comparisons of Pig-a mutant phenotype frequency data collected in males and females justified dataset merging across sex. This permitted more effective scrutiny regarding the effect of post-exposure sampling time on the mutagenicity of N-ethyl-N-nitrosourea observed in reticulocytes and erythrocytes in the Pig-a assay. The BMD-covariate approach revealed tissue-specific differences in the induction of lacZ transgene mutations in Muta™Mouse specimens exposed to benzo[a]pyrene (BaP), with the results permitting the formulation of mechanistic hypotheses regarding the observed potency ranking. Lastly, we illustrate how historical dose-response data for assessments that examined numerous doses (i.e. induced lacZ mutant frequency (MF) across 10 doses of BaP) can be used to improve the precision of BMDs derived from datasets with far fewer doses (i.e. lacZ MF for 3 doses of dibenz[a,h]anthracene). Collectively, the presented examples illustrate how innovative use of the BMD approach can permit refinement of the use of in vivo data; improving the efficacy of experimental animal use in genetic toxicology without sacrificing PoD precision.

Benchmark dose and the three Rs. Part II. Consequences for study design and animal use.

OECD test guidelines for standard toxicity studies prescribe (minimal) numbers of animals, but these are not substantiated by a quantitative analysis of the relationship between number of animals and the required performance of the associated study design. This paper provides a general approach of how this relationship may be established and discusses the approach in more detail by focusing on the three typical repeated-dose studies (subacute, subchronic, and chronic). Quantitative results derived from simulation studies, including some new results, are summarized and their consequences for study guidelines are discussed. The currently prescribed study designs for repeated-dose studies do not appear to be sufficient when the NOAEL is used for evaluating the data--the probability of not detecting toxicologically significant effects is high. The ensuing need for increasing the number of animals may be avoided by replacing the NOAEL approach by the BMD approach as it increases the probability of detecting the same effects without increasing the number of animals. Hence, applying the BMD approach will result in a virtual reduction in the number of animals. Further, the BMD approach allows for a real reduction in the number of animals on various grounds. It allows for analyzing combined similar datasets, resulting in an increase in precision, which can be translated in animal reduction while keeping the same precision. In addition, applying the BMD approach may be expected to result in animal reduction in the long run, as it allows for distributing the same number of animals over more doses without loss of precision. The latter will reduce the need to repeat studies due to unfortunate dose location.

Benchmark dose and the three Rs. Part I. Getting more information from the same number of animals.

Evaluating dose-response data using the Benchmark dose (BMD) approach rather than by the no observed adverse effect (NOAEL) approach implies a considerable step forward from the perspective of the Reduction, Replacement, and Refinement, three Rs, in particular the R of reduction: more information is obtained from the same number of animals, or, vice versa, similar information may be obtained from fewer animals. The first part of this twin paper focusses on the former, the second on the latter aspect. Regarding the former, the BMD approach provides more information from any given dose-response dataset in various ways. First, the BMDL (= BMD lower confidence bound) provides more information by its more explicit definition. Further, as compared to the NOAEL approach the BMD approach results in more statistical precision in the value of the point of departure (PoD), for deriving exposure limits. While part of the animals in the study do not directly contribute to the numerical value of a NOAEL, all animals are effectively used and do contribute to a BMDL. In addition, the BMD approach allows for combining similar datasets for the same chemical (e.g., both sexes) in a single analysis, which further increases precision. By combining a dose-response dataset with similar historical data for other chemicals, the precision can even be substantially increased. Further, the BMD approach results in more precise estimates for relative potency factors (RPFs, or TEFs). And finally, the BMD approach is not only more precise, it also allows for quantification of the precision in the BMD estimate, which is not possible in the NOAEL approach.

Shape and steepness of toxicological dose-response relationships of continuous endpoints.

A re-analysis of a large number of historical dose-response data for continuous endpoints indicates that an exponential or a Hill model with four parameters adequately describes toxicological dose-responses. No exceptions were found for the datasets considered, which related to a wide variety of endpoints and to both in vivo and in vitro studies of various types. For a given endpoint/study type dose-response shapes were found to be homogenous among chemicals in the in vitro studies considered, while a mild among-chemical variation in the steepness parameter seemed to be present in the in vivo studies. Our findings have various practical consequences. For continuous endpoints, model selection in the BMD approach is not a crucial issue. The often applied approach of using constraints on the model parameters to prevent "infinite" slopes at dose zero in fitting a model is not in line with our findings, and appears to be unjustified. Instead, more realistic ranges of parameter values could be derived from re-analyses of large numbers of historical dose-response datasets in the same endpoint and study type, which could be used as parameter constraints in future individual datasets. This approach will be particularly useful for weak datasets (e.g. few doses, much scatter). In addition, this approach may open the way to use fewer animals in future studies. In the discussion, we argue that distinctions between linear, sub/supralinear or thresholded dose-response shapes, based on visual inspection of plots, are not biologically meaningful nor useful for risk assessment.

Exploring the uncertainties in cancer risk assessment using the integrated probabilistic risk assessment (IPRA) approach.

Current methods for cancer risk assessment result in single values, without any quantitative information on the uncertainties in these values. Therefore, single risk values could easily be overinterpreted. In this study, we discuss a full probabilistic cancer risk assessment approach in which all the generally recognized uncertainties in both exposure and hazard assessment are quantitatively characterized and probabilistically evaluated, resulting in a confidence interval for the final risk estimate. The methodology is applied to three example chemicals (aflatoxin, N-nitrosodimethylamine, and methyleugenol). These examples illustrate that the uncertainty in a cancer risk estimate may be huge, making single value estimates of cancer risk meaningless. Further, a risk based on linear extrapolation tends to be lower than the upper 95% confidence limit of a probabilistic risk estimate, and in that sense it is not conservative. Our conceptual analysis showed that there are two possible basic approaches for cancer risk assessment, depending on the interpretation of the dose-incidence data measured in animals. However, it remains unclear which of the two interpretations is the more adequate one, adding an additional uncertainty to the already huge confidence intervals for cancer risk estimates.

An improved model to predict physiologically based model parameters and their inter-individual variability from anthropometry.

We developed a population physiology model, physB, which provides a statistical description of the physiological characteristics in the human population, in terms of the physiological parameters that are needed in physiologically based pharmacokinetic modelling. The model predicts individual organ weights, blood flows and some respiratory parameters from anthropometric properties (body height and weight, age and gender). It draws on two existing models, PK-Pop and P(3)M, but various changes and improvements were made. The conceptual differences among the three models are discussed and they are quantitatively compared by running all three models for various specific combinations of anthropometric properties.

An exposure-response curve for copper excess and deficiency.

There is a need to define exposure-response curves for both Cu excess and deficiency to assist in determining the acceptable range of oral intake. A comprehensive database has been developed where different health outcomes from elevated and deficient Cu intakes were assigned ordinal severity scores to create common measures of response. A generalized linear model for ordinal data was used to estimate the probability of response associated with dose, duration and severity. The model can account for differences in animal species, the exposure medium (drinking water and feed), age, sex, and solubility. Using this model, an optimal intake level of 2.6 mg Cu/d was determined. This value is higher than the current U.S. recommended dietary intake (RDI; 0.9 mg/d) that protects against toxicity from Cu deficiency. It is also lower than the current tolerable upper intake level (UL; 10 mg/d) that protects against toxicity from Cu excess. Compared to traditional risk assessment approaches, categorical regression can provide risk managers with more information, including a range of intake levels associated with different levels of severity and probability of response. To weigh the relative harms of deficiency and excess, it is important that the results be interpreted along with the available information on the nature of the responses that were assigned to each severity score.

The Relationship Between Internal and External Dose: Some General Results Based on a Generic Compartmental Model.

Statements on how the internal-to-external-dose (IED) relationship looks like are often based on qualitative toxicokinetic arguments. For example, the recently proposed kinetically derived maximum dose (KMD) states that the IED relationship must have an inflection point, due to saturation of underlying processes like metabolism or absorption. However, such statements lack a solid quantitative foundation. Therefore, we derived expressions for the IED relationship for a number of scenarios based on a generic compartmental model involving saturation. The scenarios included repeated or single dose, and saturable metabolism or saturable absorption. For some of these scenarios, an explicit expression for the IED relationship can be derived, for others only implicit expressions can be established, which need to be evaluated numerically. The results show that saturable processes will lead to an IED relationship that is nonlinear over the whole dose range, ie, it can be approximated by a linear relationship at the lower end, whereas the approximation will become gradually poorer with increasing doses. The finding that saturation does not lead to an inflection point in the IED relationship, as assumed in the KMD, implies that the KMD is not a valid approach for selecting the top dose in toxicological studies. An additional use of our results is that the derived explicit expressions of the IED relationship can be fitted to IED data, and, possibly, for extrapolation outside the observed dose range.

Defining embryonic developmental effects of chemical mixtures using the embryonic stem cell test.

The embryonic stem cell test (EST) was applied to evaluate dose addition in combined exposures of teratogenic compounds in the EFSA-defined cumulative assessment group "craniofacial malformations", which was one of the selected cases in the EU-H2020 project "EuroMix". Test compounds were selected through reported effects in rodents, and represented a wide variety of chemical families and modes of action (MOA), including triazoles to inhibit CYP26; (synthetic) retinoids, to activate RAR/RXR; valproic acid, to inhibit histone deacetylase; dithiocarbamates, to disrupt extracellular matrix formation; dioxin (-like) compounds, to activate the aryl hydrocarbon receptor; 17alpha-ethynylestradiol, to activate the estrogen receptor; 5-fluorouracil, to disrupt DNA-synthesis; MEHP and PFOS, to activate peroxisome proliferation activated receptors; and methyl mercury, to induce oxidative stress and inhibit protein function. The EST appeared particularly useful to evaluate differentiation-inhibiting effects of compounds targeting early processes in craniofacial development, possibly related to the early fate of neural crest cells. Mixtures, designed as equipotent concentrations of two compounds with similar or dissimilar MOA, and single compounds showed overlapping dose-responses. This observation is consistent with dose addition in the EST in all studied binary mixtures, irrespective of MOA, and thereby supports the application of dose-addition as a default in cumulative risk assessment.