1. Emerging Methods for Controlling Chemicals E. Donald Elliott Professor (adj) of Law, Yale Law School e.donald.elliott@yale.edu Gail Charnley Elliott PhD HealthRisk Strategies charnley@healthriskstrategies.com

2. Current Regulatory Practice NAS, Decision Making for Regulating Chemicals in the Environment (1975) Whole body animal tests High doses Uncertainty factors

3. Problems #1 High Cost – ~$500,000/bioassay Low Throughput Information Lacking on Safety of Most Chemicals

4. Too Many ChemicalsToo Little Data (%) Judson, et al EHP (2009) 9912

5. Is New Chemical Regulation Technology-Forcing Toxicology? “Ambitious new testing programs such as REACH are literally impossible without new methods for screening and prioritizing.” E. Donald Elliott, Needed: A Strategy for Implementing the Vision, 29 Risk Analysis 482 (Feb., 2009). Conventional Implementation of REACH  68,000 substances (vs. 29,000 predicted)  54 million animals (ECHA: “only” 9 million)  $13.5 billion  “regulatory toxicology needs to move into the 21 st century”  “many core methods have remained unchanged for 40 years.” Thomas Hartung & Costanza Rodiva, 460 Nature 1080 (Aug 27, 2009).

6. Problems #2 Inter-species extrapolations High dose low dose extrapolations ?

7. Problems #3 Cruelty to Test Animals

8. The Solution? Automated Micro-Arrays Pathway- Based Risk Assessment Computational Toxicology

9. Toxicity Testing in the 21 st Century National Academy of Sciences (2007)

10. “envisions a not-so-distant future in which virtually all routine toxicity testing would be conducted in human cells or cell lines in vitro by evaluating cellular responses in a suite of toxicity pathway assays using high throughput tests, implemented with robotic assistance.”* * M.Andersen & D.Krewski, Toxicol Sci. 2009 Feb;107(2):324-30

11. “ toxicity pathway assays ” ProteinRNA Normal pathway

12. “ toxicity pathway assays ” ProteinRNA ? Perturbation

13. An (Imperfect) Analogy http://en.wikipedia.org/wiki/Rabbit_test

14. ISCHEMIA REPERFUSION EVIDENCE

15. Source: http://www.epa.gov/NCCT/practice_community/Andersen_EPA_CPCP_27mar2 008.pdf

16. Toxicity Testing in the 21 st Century National Academy of Sciences (2007)

17. Overview How we got where we are now Why our current approach doesn’t work well Proposal for a new approaches to testing for the purpose of risk assessment/regulatory decision- making Reaction so far

18. Qualitative statutory requirements for public health protection reasonable certainty that no harm will result provide an ample margin of safety to protect public health no employee will suffer material impairment of health safe significant risk of harm fulfill those requirements without being arbitrary & capricious

19. Regulatory Risk Analysis Quantitative risk assessment: developed in reaction to laws and judicial decisions calling for limits on chemical exposures that protect public health & to create a judicially reviewable record Hazard identification Dose-response assessment Exposure assessment Risk characterization Economics Feasibility Social factors Politics Protect public health

20. Regulatory Risk Analysis Quantitative risk assessment: developed in reaction to laws and judicial decisions calling for limits on chemical exposures that protect public health & to create a judicially reviewable record Hazard identification Dose-response assessment Exposure assessment Risk characterization Economics Feasibility Social factors Politics Protect public health

21. Types of toxicity tests Acute, single dose Repeated dose (14 days, 90 days, 2 years) Reproductive toxicity tests Developmental toxicity tests Neurological toxicity tests Cancer bioassays Mode-of-action studies

22. Traditional toxicity paradigm: Exposure-response continuum Focus = adverse outcome Whole-animal based

23. The Problem (1) Tests are dated, expensive, time- consuming, and animal-intensive –Standard test methods date to 1930s –$1 million+ per substance –Low throughput: cancer bioassay takes 2 years to do, another year+ to evaluate results –Cancer bioassay uses 800 animals Low sensitivity (false negatives) Poor specificity (false positives) Ethics

24. The Problem (2) Uncertain relevance of laboratory animal tests to humans –Interspecies extrapolation required –High-to-low dose extrapolation required

25. High-to-low-dose extrapolation DATA ARE HERE WE ARE HERE

26. The Problem (2) Uncertain relevance of laboratory animal tests to humans –Interspecies extrapolation required –High-to-low dose extrapolation required Poor information on potential special sensitivities –Life-stage, genetic, nutritional, etc. Need to regulate with incomplete information –Default assumptions –Uncertainty factors

27. Setting chemical exposure limits Highest non-toxic dose in rats 10x 100x Safe dose People might be more sensitive than rats Some people might be more sensitive than others

28. The Problem (3) Test large numbers of existing chemicals, many of which lack basic toxicity data Test the large number of new chemicals and novel materials, such as nanomaterials, introduced each year

29. Toxicity Testing in the 21 st Century A Vision and a Strategy The National Academies Press www.nap.edu

30. Goals Minimize animal use Increase throughput Reduce time & expense Replace assumptions with data Produce toxicity in a way that provides information useful for risk assessment –identify conditions of use that are unlikely to produce toxicity = “safety”

31. Vision Transform routine toxicity testing and human health risk assessment & management –Computational toxicology –High & medium-throughput screening assays in vitro human cell lines toxicogenomics & other -omics –Physiologically based pharmacokinetic models –Pathway-based, not adverse-outcome-based Identify critical pathways that when perturbed can lead to adverse health outcomes

32. Pathway-based toxicity testing: Intersection of exposure and biologic function

33. Example: Perchlorate & thyroid function I-I- I-I- thyroid hormones thyroid hormones fetal/childhood growth & development metabolism & organ function thyroid cell

34. Example: Perchlorate & thyroid function I-I- I-I- thyroid hormones thyroid hormones metabolism & organ function fetal/childhood growth & development compensate thyroid cell

35. Example: Perchlorate & thyroid function I-I- I-I- thyroid hormones thyroid hormones lower metabolism & impaired organ function abnormal fetal/childhood growth & development thyroid cell

36. Pathway-based toxicity testing: Intersection of exposure and biologic function Exposure leads to perturbation of normal biologic pathways Biologic function is compromised if host is unable to adapt because: –Perturbation is sufficiently large –Underlying nutritional, genetic, life-stage, or disease status Compromised biologic function leads to toxicity & disease = mode of action

37. Source: NAS/NRC (2006) Health Implications of Perchlorate Ingestion

38. ISCHEMIA REPERFUSION EVIDENCESource: William Pennie, Pfizer

39. Challenges Identifying normal biologic pathways Validation: determining whether perturbations, such as a particular pattern of changes in gene expression, are adverse –Changes may not indicate adverse effect –Multiple/complex etiologies of most diseases –Support tort liability arguments without rigor Regulatory guidance development

40. Bimodal reactions so far Chemical industry We’ll have the means to identify safer substances & to reduce risks more effectively We’ll get sued every time a chemical perturbs something Environment NGOs We’ll be able to identify low-dose & cumulative effects, synergies, & people with special sensitivities Uncertainty factors might be replaced by data, leading to less stringent exposure limits

41. The classic case in American law is International Harvester Co. v. Ruckelshaus, 478 F.2d 615 (D.C. Cir. 1973). Judge Leventhal held the extent of proof required should be a function of the consequences of an error in one direction or another. Balancing the Risks of False Positives and False Negatives

42. “Extraordinary conclusions require extraordinary proof” - - Carl Sagan Aaron Wildavsky, Searching for Safety (New Brunswick, N.J.: Transaction Books, 1988)  improve resilience vs. improve anticipation

44. Uses of Pathway-Based in vitro Assays Private use to eliminate candidates Screening/setting agency priorities Requiring additional testing Public information Considered in weight of evidence in standard-setting/rulemaking Material basis for approval Admissible in damage cases Material basis for regulation Validation + _