Toxicity Testing and Safety Assessment in the 21st Century: Creating a Research Program to Accelerate Change

Home / New Perspectives / Overarching Challenges / Toxicity Testing and Safety Assessment in the 21st Century: Creating a Research Program to Accelerate Change

Overarching Challenges

Toxicity Testing and Safety Assessment in the 21st Century: Creating a Research Program to Accelerate Change

Melvin Andersen, The Hamner Institutes for Health Sciences

Published: February 11, 2011

About the Author(s)
Melvin E. Andersen, PhD, is Director, Institute for Chemical Safety Sciences, The Hamner Institutes for Health Research, Research Triangle Park, NC. Over a 40-year toxicology career, he has worked in the federal government (US Navy, Department of Defense and EPA), private industry (ICF Kaiser) and academia (Colorado State University). He has over 400 published papers and book chapters in a research career focused on computational approaches for dose response modeling and human health risk assessment.  Increasingly, his programs at The Hamner are becoming more closely aligned with accelerating acceptance of toxicity testing approaches outlined by the 2007 NAS report, “Toxicity Testing in the 21st Century: A Vision and A Strategy”. He holds a PhD in Biochemistry and Molecular Biology from Cornell University.

Melvin E. Andersen
Director Institute for Chemical Safety Sciences
The Hamner Institutes for Health Sciences
6 Davis Drive
P.O. Box 12137
Research Triangle Park, NC 27709-2137

With few exceptions, toxicology has relied on high dose testing in animals with default methods for extrapolating results to low level exposures in human populations. This approach remains as the operating paradigm for chemicals management into the 21st century. There is widespread disaffection with this approach from both regulatory agencies and the regulated communities. These traditional methods are expensive, exorbitant in use of animals, and have low throughput. Over the past 30 years, there have been attempts to incorporate pharmacokinetics and mode of action into the risk assessment process for high-value compounds. These research efforts are costly and have had limited success in changing the existing risk framework. In addition, most of these efforts focus heavily on explaining high dose rodent effects rather than on understanding the biological basis for dose response relationships expected in humans exposed to chemicals at relevant environmental levels.

With progress in molecular, cellular and computational biology, new tools are available for studying the responses of cells, tissues, and organisms to chemical stressors. As a result, the US National Academy of Sciences, at the request of the US Environmental Protection Agency (EPA), issued its 2007 landmark report (“Toxicity Testing in the 21st Century: A Vision and A Strategy“), which envisioned a transformative change in toxicity testing that would greatly increase human relevance, significantly reduce the cost and time required to conduct chemical safety assessments, and markedly reduce and potentially eliminate high-dose animal testing (NAS, 2007, reprinted as Krewski et al., 2010). These tools would enable chemical manufacturers and government regulators to predict regions of exposure that are expected to be without adverse consequences rather than making predictions of the incidence of specific adverse responses in human populations. This transformation in toxicity testing of chemical substances would move risk assessment away from reliance on high dose studies to an assessment of the likelihood of effects at relevant environmental exposures of human populations. This toxicity testing transformation is now in progress and will take clear shape over the next 5 years. Whether it continues to be a risk assessment based process is not entirely clear.

Regulatory Drivers

The US EPA has recognized that new approaches to toxicity testing are needed to increase throughput and provide more relevant data to inform risk assessment for pesticides and for many other chemical compounds. In 2007, EPA launched a high throughput screening program (ToxCast™) to develop more cost-effective approaches for prioritizing the toxicity testing of thousands of chemicals. Additionally, the US EPA pesticide program has outlined specific short and long-term objectives to accelerate implementation of the NRC vision within a more risk based context.

In the European Union (EU), the Cosmetics Directive (2003) banned the sale of finished cosmetic products that have undergone animal testing after March 2009, as well as animal testing of chemical ingredients intended for cosmetics products, except for reproductive toxicity, repeated dose studies, and toxicokinetics, which are technically banned effective 2013. New approaches will be needed to assess likely chronic responses from results obtained with in vitro assays. Other regulatory initiatives in Europe (with REACH [Registration, Evaluation, Authorisation, and Restriction of Chemicals]) and in the USA (with reauthorization of TSCA [Toxic Substances Control Act]) are elevating the questionable value of even more extensive requirements for in-life animal testing for human health safety assessments.

Clearly, new technologies are urgently needed to create expeditious approaches for testing and data interpretation for a variety of industrial products at relevant exposures. Such tools will be useful for prioritization of lead chemistries, formal toxicity testing, risk/safety assessments and prioritization of compounds for more detailed evaluations. These new test methods are likely to guide development of modern safety approaches on a global basis including broader influences on developing economies in Asia and South America.

The Opportunity: New 21st Century Chemical Risk Assessment

A modern suite of in vitro technologies, when properly applied, promises better assessment of low dose risks, places biomonitoring results in a risk assessment rather than hazard context, and establishes new models for human risk assessment that do not rely on default methodologies applied to animal studies. These technologies also support replacement of animals in toxicity testing as is consistent with directions in Europe and general practices of humane science.

The development and validation of new risk assessment tools can be accomplished in a number of ways. The Hamner Institute for Chemical Safety Sciences (ICSS) leadership believes that the most expedient approach is to develop a complete suite of tools – in vitro assays, computational systems modeling of pathway function, and in vitroin vivo extrapolation models – around strategically chosen prototype pathways. Once in place, these methodologies can be more quickly and cost effectively expanded to include any number of additional toxicity pathways for risk assessment of broad sets of chemical classes. Other potential strategies, already moving forward, are based on prioritization of more limited in vivo testing from in vitro assay results (Judson et al., 2010) or direct use of in vitro results for hazard identification (Crump et al., 2010). These do not have as their primary focus a risk context, which is the unifying component of the NRC vision.

To develop and optimize the tools and approaches necessary to evaluate dose response relationships and low dose risks in biologically relevant in vitro assays, resources will need to be brought to bear in pathway assay design, computational systems biology pathway models, and in vitro to in vivo extrapolation. This initiative will also require the coordination and management of a larger research program with interactions between industry, academia, NGOs and regulatory bodies. The Hamner ICSS, with key support from several sponsors, has already initiated research programs to implement and validate approaches along the lines envisioned by the NRC 2007 report. The Hamner has also engaged with the Human Toxicology Program Consortium, organized by The Humane Society of the United States, to accelerate the application of new in vitro toxicity test methods. The Hamner ICSS is pursuing a broad research plan for designing new in vitro assays and in using prototype compounds representing major toxicity pathways and their application to risk assessment.

Implementation: 21st Century Chemical Risk Assessment

Application of 21st Century Toxicology in Risk Assessment: The 2007 NRC report includes a figure that maps the new tools and approaches onto the prevailing framework for risk assessment. This figure has been updated (Krewski et al., 2011). While the new version (Figure 1) shows the components involved in risk assessment, it does not capture the manner in which individual activities come together to allow quantitative human health risk (safety) assessments. More recently it was suggested (Boekelheide & Andersen, 2010) that the DNA-damage toxicity pathway be modeled by panels of in vitro assays, similar to an integrated testing strategy, to ascertain both adversity and the process by which the new technologies – toxicity pathways, computational biology, and pharmacokinetic modeling – will be used for risk/safety assessment (Figure 2).

Pathway Prototypes: The overall goal is to develop appropriate panels of short-term studies that will in time completely supplant in-life studies for risk assessment approaches for any compound. The pathway targeted research effort, in turn, will focus on developing/refining assays at several levels of biological organization – i.e., molecular, cellular and organ level. Results from these assays will be used together with the computational systems biology pathway modeling to understand the structure and signaling circuitry in a fashion similar to work with p53-DNA damage networks (Batchelor et al., 2009; Loewer et al., 2010). As pointed out in the NRC report, these computational models support biologically-oriented dose response assessments with the capacity to account for transitions between sub-threshold levels, homeostatic regions, adaptation and overtly adverse regions of concentration.

Two other aspects will deserve consideration for extrapolations: (1) linkage of in vitro results in cell systems with in vivo responses and (2) in vitroin vivo extrapolations to predict exposures that could lead to human risks. The pragmatic need to link the new in vitro methods with older risk assessments based on in-life animal studies argues for using well-studied compounds as part of the calibration of new approaches. Whenever possible, existing animal datasets should be used for comparisons rather than conducting new in vivo studies.

Cellular response studies are the central element of the research program, requiring evaluation of the dose response for specific molecular and cell level responses and the mapping and modeling of pathways and pathway perturbations. A series of in vitro and in vivo genomic time-course, dose-response studies on the prototype compounds would provide the linkage to cell response behaviors in vitro and permit clear development of the mode-of-action based approach for risk assessment. Linkage of in vivo pathway responses, such as the rat uterotrophic assays (Kwekel et al., 2005) with transcriptomics responses studies of uterine cells in vitro (Naciff et al., 2009) would be an example of calibrating pathway and dose response information between in vitro and in vivo systems.

Physiologically-based pharmacokinetic (PBPK) modeling will also have an important role in establishing the exposure condition capable of providing an in vivo tissue concentration equivalent to those causing adverse responses in vitro (Figure 2). Reverse dosimetry with PBPK modeling (Clewell et al., 2008) will be used to assess the margin of safety (MOS). The MOS would be determined by predicting the in vivo human concentrations expected from realistic exposure situations and comparing these concentrations with those causing adverse responses in vitro.

The overall step-wise process for using in vitro studies in risk assessment (Boekelheide & Andersen, 2010) includes:

  • Collecting in vitro data on a panel of pathway assays with increasing degrees of cellular perturbations
  • Analyzing results from the panel of assays to propose both the target toxicity pathway and the ‘adverse’ response concentration for some designated percentage response
  • Developing computational systems biology pathway models to support dose-response extrapolation
  • Applying PK modeling to assess the exposure levels of compound in the environment that would be required to lead to an adverse response in humans
  • Carrying out MOS calculation to assess the ratio of exposure required to give a tissue response compared to current exposure.

This final step in the process represents the ultimate goal of the Hamner project in terms of demonstrating how the results of this research should be applied in a risk or safety assessment context. Initial application with well-studied prototypes – for instance, p53–mdm2 DNA damage stress, Nrf2-Keap1 oxidative stress and nuclear receptor signaling pathways such as estrogen, androgen, thyroid or CAR – could serve to accelerate change to a new risk assessment methodology highlighting both opportunities and challenges in moving expeditiously to a new toxicity testing platform.

Figure 1: Risk/Safety Assessment following Toxicity Testing in the 21st Century. A 1983 National Research Council (NRC) publication “Risk Assessment in the Federal Government” outlined four steps in the risk assessment process — hazard identification, dose-response assessment, exposure assessment, and the integrated process of risk characterization. Each of these steps is captured in this figure designed to show their correspondence with the new testing components described in the 2007 NRC report Toxicity Testing in the 21st Century: A Vision and A Strategy. Dose-response assessment covers the activities circumscribed by rectangle outlined by thick blue lines. The other steps are shown by the dotted lines along the bottom of the figure. The figure is reproduced from a recent publication (Krewski et al., 2011).


Figure 2: A schematic of the use of the data collected on toxicity pathways for risk assessment based on in vitro panels of assays for specific toxicity pathways. The panel of assays for a specific pathway provides a point of departure for the risk assessment as an in vitro concentration. Computational systems biology modeling of pathway circuitry and dynamics indicates the shape of the dose response at lower doses, leading to an acceptable concentration proposed for a human population. The acceptable concentration is then converted to an exposure level through techniques of reverse dosimetry implemented by pharmacokinetic modeling. Adapted with modification (Boekelheide & Andersen, 2010).

©2011 Melvin Anderson

Batchelor, E., Loewer, A. & Lahav, G. (2009). The Ups and Downs of p53: Understanding Protein Dynamics in Single Cells. Nat. Rev. Cancer. 9, 371-377.

Boekelheide, K. & Andersen, M.E. (2010). A Mechanistic Re-definition of Adverse Effects – A Key Step in the Toxicity Testing Paradigm Shift. ALTEX. 27, 243-252.

Clewell, H.J., Tan, Y.M., Campbell, J.L. & Andersen, M.E. (2008). Quantitative Interpretation of Human Biomonitoring Data. Toxicol. Appl. Pharmacol. 231, 122-133.

Crump, K.S., Chen, C. & Louis, T.A. (2010). The Future Use of in Vitro Data in Risk Assessment to Set Human Exposure Standards: Challenging Problems and Familiar Solutions. Environ. Health Perspect. 118, 1350-1354.

Judson, R.S., Houck, K.A., Kavlock, R.J., Knudsen, T.B., Martin, M.T., Mortensen, H.M., et al. (2010). In Vitro Screening of Environmental Chemicals for Targeted Testing Prioritization: The ToxCast Project. Environ. Health Perspect. 118, 485-492.

Krewski, D., Acosta, D., Andersen, M., Anderson, H., Bailar, J.C., Boekelheide, K., et al. (2010). Toxicity Testing in the 21st Century: A Vision and a Strategy. J. Toxicol. Environ. Health B Crit. Rev. 13, 51-138.

Krewski, D., Westphal, M., Al-Zoughool, M., Croteau, M.C. & Andersen, M.E. (2011). New Directions in Toxicity Testing. Annu. Rev. Public Health. 32, in press.

Kwekel, J.C., Burgoon, L.D., Burt, J.W., Harkema, J.R. & Zacharewski, T.R. (2005). A Cross-Species Analysis of the Rodent Uterotrophic Program: Elucidation of Conserved Responses and Targets of Estrogen Signaling. Physiol. Genomics. 23, 327-342.

Loewer, A., Batchelor, E., Gaglia, G. & Lahav, G. (2010). Basal Dynamics of p53 Reveal Transcriptionally Attenuated Pulses in Cycling Cells. Cell. 142, 89-100.

Naciff, J.M., Khambatta, Z.S., Thomason, R.G., Carr, G.J., Tiesman, J.P., Singleton, D.W., et al. (2009). The Genomic Response of a Human Uterine Endometrial Adenocarcinoma Cell Line to 17Alpha-Ethynyl Estradiol. Toxicol. Sci. 107, 40-55.

NAS. (2007). Toxicity Testing in the 21st Century: A Vision and a Strategy. National Academies Press, Washington, DC.