Developing Risk Assessment Beyond Science and Decisions

1. Developing Risk AssessmentBeyond Science and Decisions M.E. (Bette) Meek McLaughlin Centre University of Ottawa bmeek@uottawa.ca

2. Outline • The NAS Report • Science & Decisions: Advancing Risk Assessment • Coordinating & Extending the Recommendations • Potential Contribution of Other Initiatives • National & International • Dose Response tailored to Need • Problem Formulation/Dose Response Analysis • Focus of the Workshop

3. The Committee’s Vision Science and Decisions:Advancing Risk Assessment Final Report Released, 2008

4. NAS Committee:Advancing Risk Assessment - Background • “Chemical Risk assessment at a crossroads” • Facing substantial challenges, e.g., • long delays in completing complex risk assessments, some of which take decades • lack of data • the need to address the many unevaluated chemicals in the marketplace • Recommendations for practical improvements to the U.S. Environmental Protection Agency (EPA) • Shorter (2-5 y) and • longer (10-20 y) term

5. NAS Committee:Advancing Risk Assessment - Focus • Improving the technical analysis that supports risk assessment • Chapters 4 to 7 • Improving the utility of risk assessment • Chapters 3 and 8 Efficiency/Tiering?

6. NAS Committee:Advancing Risk Assessment • Design of risk assessment • Uncertainty and variability • Selection and use of defaults • A unified default approach to dose-response • Combined Exposures risk assessment • Improving the utility of risk assessment • Problem formulation/Issue Identification • Stakeholder involvement • Capacity building

8. Phase I Problem Formulation • Begins with a “signal” of potential harm • Positive bioassay or epidemiological study; industrial contamination • What options are there to reduce the hazards or exposures? • How can risk assessment be used to evaluate the merits of the various options?

9. Phase II Planning, Risk Assessment & Confirmation of Utility • Level & complexity consistent with the goals of decision-making • Including uncertainty & variability analysis • Assessment • Meet the need, discriminate among options, adequate process? Hazard Dose Response Exposure Risk

10. The NAS 4-Step Paradigm The Need to Move On Hazard Identification Dose Response Assessment & Characterization Exposure Assessment & Characterization Risk Assessment & Characterization

11. Revised NAS 4-Step Paradigm Hazard Characterization Dose Response Assessment & Characterization Exposure Assessment & Characterization Risk Assessment & Characterization Hazard Characterization (including mode of action) permits us to be more predictive

12. Implications of Regulatory Developments to Consider All Chemicals • Need for increased efficiency in risk assessment • Processing much larger numbers of substances • More/better predictive tools & more efficient toxicity testing • Drawing on new technologies such as the genomics • Requires a fundamental paradigm shift

13. The Committee’s Vision Toxicity Testing in the 21st Century:A Vision and A Strategy Final Report Released June 12, 2007

14. Phase III Risk Management • Based on consideration of a broader range of options and array of impacts, beyond individual effects to include individual health status and ecosystem protection The entire process to protect against political interference, engage stakeholders and meet time constraints (Details to follow)

15. NAS Committee:Advancing Risk Assessment • Design of risk assessment • Uncertainty and variability • Selection and use of defaults • A unified default approach to dose-response • Combined Exposures risk assessment • Improving the utility of risk assessment • Problem formulation/Issue Identification • Stakeholder involvement • Capacity building

16. Selection and Use of Defaults • “EPA should develop clear, general standards for the level of evidence needed to justify the use of agent-specific data and not resort to default” • This is helpful to increase transparency as a basis to separate science judgment from science policy • However: • It sets up “default” as representing something other than: • what we use when we don’t have predictive data about how chemicals induce their effects • Recognizing that the scientific basis of defaults is nebulous, at best,

17. Default Methodology in Risk Assessment x • Curve fitting from high dose studies for adverse late (apical) endpoints (hazard identification) • Linear extrapolation to estimate risk, or: • “Safe Dose” N/LO(A)EL or BMC/D Uncertainty Factors • Interspecies differences/ • human variability (x10) x Response x x NOAEL/BMC Dose

18. Default as the “Gold Standard” Adherence to default is likely: • One of, and arguably, the single most important determinant of lack of progress in regulatory risk assessment to address: • Susceptible subpopulations • Harmonization across endpoints • Combined exposures, • Accuracy, & • Predictive capacity

19. Changing the “Default Mindset” • Rather than challenging the risk assessment community to justify moving from default, emphasis should necessarily be on data first • Reverse the onus • We need also to increase understanding of the erroneous premise that default is always protective • Depends on mode of action (kinetic/dynamic data)

20. NAS Committee:Advancing Risk Assessment • Design of risk assessment • Uncertainty and variability • Selection and use of defaults • A unified default approach to dose-response • Combined Exposures risk assessment • Improving the utility of risk assessment • Problem formulation/Issue Identification • Stakeholder involvement • Capacity building

21. Unified Approach to Default Dose Response Assessment • “A consistent approach to risk assessment for cancer and non-cancer effects is scientifically feasible and needs to be implemented” • “Because the RfD and RfC do not quantify risks for different magnitudes of exposure…their use in risk-risk and risk-benefit comparisons and risk management decision-making is limited” • This seemed to prevail over discussions related to modes of action, background exposures & susceptibility

22. The “Recommendation” • EPA implement a phased-in approach to consider chemicals under a unified dose-response assessment framework that includes a systematic evaluation of background exposures and disease processes, possible vulnerable populations, and modes of action that may affect human dose-response relationships = mode of action

23. Figure 5.8 New unified process for selecting approach and methods for dose-response assessment for cancer and noncancer .

24. Curve fitting at high dose for point of departure for late (apical) endpoints Linear extrapolation or N/LO(A)EL or BMC/D UF Interspecies differences/human variability (x10) Earlier endpoints in the most relevant species, considering kinetic and dynamic data, to address extrapolations Biologically(MoA) Based Default x x Response x x NOAEL/BMC Dose

25. Characterize Uncertainty Mode of Action (MoA) information is core to evolution of the risk assessment paradigm • Understand • Human Relevance • Susceptibility • Dose Extrapolation • Mode of Action • Toxicity Pathways • Identify Key Events • Promote • Harmonized approach for all endpoints • Inform • Multiple Chemical Risks

26. Exposure Tissue Dose Biologically Effective Dose Early Responses Late Responses Pathology Physiologically Based Pharmacokinetic Models Tissue Dose Metric Mode of Action Toxicokinetics Toxicodynamics Exposure-Response Continuum(Source to Outcome Pathway) Mode of Action involves identification of several key events between exposure and effect

27. IPCS/ILSI MOA/HR (WOE) Framework Q1. Is the weight of evidence sufficient to establish the MoA in animals? Q2. Fundamental qualitative differences in key events? Q3. Fundamental quantitative differences in key events? “Key Events” established based on “Hill Criteria” Implications of Kinetic & Dynamic Data for Dose– Response Comparison of “Key Events” & relevant biology between animals & humans

28. Implications for Dose-Response Analysis Key Event 1 Key Event 2 What is the shape of the dose-response curve in the range of both observation and inference for the rate limiting key events, based on an understanding of MOA? Key Event Tumours

29. Default to Biologically Based Default Biologically Based Key Events x x Response x x NOAEL/BMC Dose

30. NAS Committee:Advancing Risk Assessment • Design of risk assessment • Uncertainty and variability • Selection and use of defaults • A unified default approach to dose-response • Combined Exposures risk assessment • Improving the utility of risk assessment • Problem formulation/Issue Identification • Stakeholder involvement • Capacity building Efficiency/Tiering?

31. Efficiency/Tiering? Figure 5.8 New unified process for selecting approach and methods for dose-response assessment for cancer and noncancer .

32. The Need for Efficiency:Evolving Legislative Mandates • Canada • “Categorization” (i.e., systematic priority setting) for 23, 000 chemicals by Sept., 2006 under the Canadian Environmental Protection Act (CEPA) • Europe • Registration, Evaluation and Authorization of Chemicals (REACH) (2007) • Volume trigger and hazard based • Consistency between Existing and New Chemicals • Industry Responsibility • U.S. • Voluntary Testing Initiatives • Renewal of TOSCA

33. CEPA 1999 Existing Substances Program CATEGORIZATION of the Domestic Substances List (DSL) (First Phase) (n=23,000) INCREASING REFINEMENT OF PRIORITIES + COMPLEXITY OF ASSESSMENT Decisions of Other Jurisdictions Public Nominations Greatest Potential for Human Exposure Substances that are Persistent or Bioaccumulative DECREASING NUMBERS OF SUBSTANCES “Inherently Toxic” to Humans “Inherently Toxic” to non-Human Organisms SCREENING ASSESSMENT (Second Phase) No further action under this program Risk Management CEPA-Toxic IN-DEPTH ASSESSMENT - Priority Substances List (Third Phase) DECREASING UNCERTAINTY No further action under this program Risk Management CEPA-Toxic

34. Simple and Complex Priority Setting Tools EXPOSURE Simple Exposure Tool (SimET) - Relative ranking of all DSL substances based on submitters (S),quantity (Q) and expert ranked use (ERU) Complex Exposure Tool (ComET) - Quantitative plausible maximum age-specific estimates of environmental and consumer exposure for individuals based on use scenario (sentinel products), phys/chem properties & bioavailability Potential for exposure influential in setting priorities Included simple use profiling for all 23, 000 chemicals, more complex use profiling for priorities HAZARD Simple Hazard Tool (SimHaz) - Identification of high or low hazard compounds by various agencies based on weight of evidence and expert opinion/consensus Complex Hazard Tool (ComHaz) - Hierarchical approach for multiple endpoints & data sources (e.g., (Q)SAR) including preliminary weight of evidence framework

35. ▲Efficiency - Screening Health Assessments • Includes early (formal) Issue Identification to focus resources • Draws on work completed in other jurisdictions • Considers weight of evidence of hazard and “margins of exposure” (comparing upper bounding estimates of exposure to lowest effect levels) • High hazard, proposed “toxic” • Margin wide, “set aside” • Margin small, additional assessment (Priority Substance) • Decision making on the basis of adequacy of the margins, based on consideration and clear delineation of confidence/uncertainties

36. Post CategorizationWhat’s Changed? Priorities not necessarily what we expected The value of a “risk-based” framework to consider all chemicals Consumer vs. environmental exposure/profile Volume ≠ exposure Persistence/bioaccumulation ≠ exposure Integrated approach across compounds (less chemical-specific; more contextually relevant) Drawing to much greater extent on profiles of substances with similar uses, properties and hazards

37. Post CategorizationWhat’s Changed? (cont’d) Focus Strategically targetting testing and “right sizing” risk management Only as much research/survey/assessment as is required to set a substance aside as a non-priority or to inform risk management increasing efficiency Full assessments where required Even for these few, early focus

38. Other Relevant Initiatives • International initiatives on mode of action (MOA)/extension to dose response (IPCS/ILSI) (supported by HC; US EPA) • MOA and Key Events Frameworks; CSAF • PCS Coordinating Predictive Use of MOA • In collaboration with OECD/others • ILSI/Health Canada initiative on Problem Formulation/Issue Identification(2007) • IPCS Combined Exposures Framework • In collaboration with OECD/others

39. Other Relevant Initiatives • IPCS Tiered Uncertainty Analysis (2007) • IPCS Characterizing and Applying PBPK Models in Chemical Risk Assessment http://www.who.int/ipcs/methods/harmonization/index.html http://www.hc-sc.gc.ca/exsd-dse http://www.ilsi.org/ResearchFoundation

40. The Role of Formal Issue IdentificationMore than a Statement of the Issue; A Process • Early Consideration of All Relevant (assimilated) information/expertise • Relying as much as possible on existing assessments, “peers” • Determining need for risk assessment based on consideration of factors such as nature and feasibility of risk management • Determining focus and scope of risk assessment, based on potential options for management • Ensuring that any assessment meets the considered need • Communication and formal engagement • Stakeholders/risk managers/public

41. Problem Formulation - Process Dialogue/Engagement with Risk Assessors/Risk Managers/Advisors/Peers First Questions • What is the problem? • What factors need to be considered? • What is the role of risk assessment in decision making? Establish the Focus and Scope of the Risk Assessment Is the preliminary focus acceptable? No Role Risk Assessment to be considered in decision making No Yes Develop Risk Assessment Procedures Proceed with Risk Assessment Make decision another way Seek stakeholder and peer input Perform Risk Assessment Risk Management Peer Review Has the problem been addressed? Develop Risk Characterization Approach and Tools No Yes Iterate Implement Risk Management Decisions

42. Establish the Focus and Scope of the Risk Assessment • Management Constraints • Regulatory considerations • Objectives • Screening or full • Broad or narrow • Time and resources • Socio-economic considerations • Political considerations • Characterization of Assessment • Assimilate universe of known data • Consider relevant exposure scenarios • Determine degree of acceptable uncertainty • Plan for addressing uncertainty • Develop communications plan • Development of preliminary hypotheses Seek stakeholder and peer input Seek stakeholder and peer input No Develop Approach Is the preliminary approach acceptable? Seek stakeholder and peer input • Forms of Assessment • Examples • Semi-Quantitative/Quantitative • Deterministic/Probabilistic • PB/PK • MOA Seek stakeholder and peer input Yes Proceed with Risk Assessment

43. PCS “Combined Exposures” Framework Sample Tiered Exposure and Hazard Considerations Yes, no further action required Tiered Exposure Tiered Hazard Assessments Assessments Tier 0 Tier 0 Simple semi-quantitative estimates of exposure Dose addition for all components Input from exposure or hazard assessments (iterative process) Is the margin of exposure adequate ? Tier 1 Tier 1 Refined potency based on individual POD, refinement of POD Generic exposure scenarios using conservative point estimates Increasing refinement of hazard models (MOA) Increasing refinement of exposure models Tier 2 Tier 2 Refined exposure assessment, increased use of actual measured data More refined potency (RPF) and grouping based on MOA Tier 3 Tier 3 PBPK or BBDR; probabilistic estimates of risk No, continue Probabilistic Exposure Estimates Framework includes also a decision tree for when to consider combined effects of multiple chemicals

44. Some Outstanding Questions on the NAS Analysis • Were all of the relevant options along the continuum of straight “default” to more mode of action based approaches to dose-response analysis adequately considered? • Were their attendant relative uncertainties characterized? • Was the analysis considered in the context of tiered, efficient assessment strategies? Problem Formulation

45. Evolution of Risk Assessment 21st Century 1980s 2000s 1990s RA/RM Paradigm Guidelines/Methods Dosimetry/PbPK Mode of Action Susceptible Populations Mixtures Toxicity Pathways Integrated Approaches CompTox 1983 2007 1994 2009

46. How Quickly does Risk Assessment Evolve? • In 1984, the “benchmark dose” was introduced • Now receiving widespread acceptance 25 yrs. later • Much guidance, many recommendations on analysis of uncertainty since the 1980’s • Several NAS & EPA reports • Incorporation of PBPK modelling • Since the late 1980’s • Formal consideration of mode of action; chemical specific adjustment factors (National & International) • Since the late 1990’s • Tiered assessment (‘94 NAS report) • Problem formulation • EPA Guidance on Risk Characterization (2000)

47. Barriers to Change • A function (in part) of: • The traditional (default) approach is “simple”, easy to explain to stakeholders & somewhat consistent • Can be “codified” “institutionalized” (requires limited expert interpretation) • Lack of long term planning for regulatory science • Meeting short term deadlines for numbers of assessments vs. understanding efficiencies of longer term investment in methodology • Sometimes predicated on science policy basis, without transparency • Lack of coordination of regulatory risk assessment/research

48. About Advisory Groups on Risk Assessment Methodology • They bring extraordinary amounts of multidisciplinary expertise to the table • They work extraordinarily hard, and • They offer insights that might not have otherwise been realized

49. About Advisory Groups • However: • Much expertise resides in Governments who commission the input • They (& others) cannot contribute directly • Time is often limited • Sometimes leading to partial analysis, contradictions • It’s often easier to make recommendations than to implement them • Constraints may not be well recognized • There isn’t continuity • Who explains and develops?

50. About Advisory Groups • So: • Follow-up is essential • Engaging the broadest range of expertise possible , or • All of the good work is for nought!