Hazard & Dose Response Assessment: Roadmaps & Methods for Using 21st Century Data. Michael L. Dourson

Transcription

1 Hazard & Dose Response Assessment: Roadmaps & Methods for Using 21st Century Data Michael L. Dourson Toxicology Excellence for Risk Assessment Center 1 Seminar Overview What is harmonization in risk assessment? The risk paradigm with a quiz The Future Epigenetics Toxicology 21 Collaboration 2 1

2 What is Harmonization in Risk Assessment? Understanding each others judgments and methods, with a future move to resolve differences Sharing information among groups, with contemplation of developing a unified source where all can contribute Creating problem formulation frameworks to sort through the confusing landscape of different safe doses, safe concentrations, or adverse effect probabilities 3 Most Extensive Data Requirements (human epidemiology studies) > quality, > certainty Moderate Data Requirements (in vitro and animal studies and anecdotal reports of human health effects) > quality, > certainty Quantitative Health Based OELs Health Based OELs Regulatory, Authoritative Traditional (TLVs, MAKs, WEELs, PELs, MACs, RELs) Working Provisional OELs (internal company, trade association, vendor limits) Hierarchy of OELs More toxicological and epidemiological data allow one to move up the hierarchy of OELs, but one needs to work on the problem formulated by the risk manager. Prescriptive Process Based OELs (REACH DNELs/DMELs) Least Data Requirements (in vitro and animal studies) Hazard Banding Strategies Pharmaceutical banding Occupational exposure bands Control Banding = Hazard Bands + Exposure Risk Assessment + Exposure Management 4 Task force: M. Guillemin, D. Heidel, M. Jayjock, C. Laszcz Davis, P. Logan, A. Maier, J. Mulhausen, K. Niven, D. O Malley, J. Perkins, S. Ripple 2

3 The Risk Paradigm (in part) Step 1. Evaluate Toxicology data to derive a safe dose Dose Response Measure (NOAEL) Safe Dose (OEL) = Uncertainty Factors (U A x U H x U D ) Step 2. Characterize risk: Hazard Quotient (HQ) = Exposure (UCL) Safe Dose (LCL) 5 Extrapolation Observed x Response 5/10% 0% x Human Exposure OEL HQ UF BMDL NOAEL x Confidence Limit LOAEL BMD x Dose x x No Observed Adverse Effect Level (NOAEL) Lowest Observed Adverse Effect Level (LOAEL) Benchmark dose (BMD) Benchmark dose lower limit (BMDL) UF = Uncertainty Factor 6 3

4 Risk Paradigm Quiz? If we had a array of effects that were all linked into one syndrome of toxicity, what would be the point of departure for a dose response assessment? If we had only 8 fingers, what would be the uncertainty factor covering experimental animal to human and within-human variation? 7 Risk Paradigm: Critical Effect Risk assessment is preventive medicine. Thus, toxicologists, epidemiologists, and clinicians are needed in judgment of critical effect conduct hazard identifications collaboratively Focus on effects of medical significance Critical effect is the first adverse effect, or its known precursor, that occurs as dose rate increases (EPA, 2013). 8 4

5 NAS, Health implications of perchlorate ingestion 9 Risk Paradigm: Uncertainty Factors Uncertainty factors for within human variability, experimental animal to human extrapolation, LOAEL to NOAEL, subchronic to chronic, and lack of certain data. Misconceptions: Studies with small n are not useful. The variability of the human population is large; an uncertainty factor of 10-fold with human data is often not enough. 10 5

6 Uncertainties to Consider in Noncancer Dose Response Assessment 1 Population Cumulative Response 0.1 UF H H human variability A animal to human L LOAEL to NOAEL S subchronic to chronic D data gap Human PBPK UF A UF L Animal LOAELs UF S UF D Sub-chronic Reproductive NOAELs or BMDs OEL Concentration (mg/m 3 ) 11 Factor of 10 Enough? Dourson, M.L., G. Charnley and R. Scheuplein,

7 Factor of 10 Enough? Human NOAEL or BMD a Animal NOAEL or BMD 13 Factor of 10 Enough? 5b. 14 7

8 Susceptibility Factors Internal Dose Factors Age Sex Genetic factors Preexisting disease External Exposure Factors Exposure variability Individual concurrent exposures 15 The Future: Addressing Human Variability: Chemical Specific Adjustment Factor (CSAF) International Programme on Chemical Safety (IPCS, 2005) 16 8

9 Intent of the CSAF HKAF Using the CSAF MOA known Data in range of exposures Data in representative population Statistical test (CV) 17 Impact of CYP2C9 Polymorphism on Warfarin Dose QC CV QF CF/PF QR CR/PR QS CS/PS QL CL/PL Cinh QP Gas Exchange Fat Rapidly Perfused Slowly Perfused Liver VMax KM KF Metabolism Cexh QC QF CA QR CA QS CA QL CA Gentry, P.R., C.E. Hack, L. Haber, A. Maier and H.J. Clewell, 3rd Toxicol Sci. 70(1):

10 The Future: Epigenetics Epigenetics heritable change in gene expression NOT caused by DNA code Epigenetics >> Genetics in variability? 19 Frameworks for Consideration 20 10

11 Kramer et al. Criteria Kramer et al. (2006) developed the following criteria for considering genetic (and epigenetic) information in risk assessment and development of OELs: The gene product must be relevant to the pathophysiology of a clearly defined and consistent phenotype (again toxic MOA needed). Gene function must be associated with exposure to a regulatedpollutant or, at the very least, to a disease-progression process known to be associated with exposure to the chosen regulated pollutant. The mutation (or expression change) must be functionally relevant. The magnitude or frequency of occurrence in the population must be measured, and variation across populations (e.g., geography, race) must be considered. There must be a high magnitude of association (i.e., preferably a relative risk >1.5) between the phenotype of interest and an adverse health effect. 21 Toxicology 21: Systems Biology-based Toxicology Testing? >84,000 chemicals on the Toxic Substances Control Act inventory >100,000 chemicals registered in REACH ~1000 new industrial chemicals and pesticides introduced to the market annually Only a fraction of new chemicals are evaluated more than superficially for human risk because current testing paradigm is slow, expensive, requires large numbers of animals, and involves considerable scientific understanding to develop credible extrapolations

12 The Future: Toxicology 21---Systems Biology-based Toxicology Testing? Driving impetus (US) Toxicity Testing 21st Century: A Vision and a Strategy (NAS, 2007) The Vision Cheaper High throughput Predictive analyses Minimize animal testing Focus on relevant dose levels More informative and efficient Characterizehuman variability Improvescientific basis of risk assessment Human cells minimal interspecies extrapolation 23 Tox 21: Outcome Pathways of NAS Exposure Uptake-Delivery to Target Tissues Perturbation Cellular response pathway Biologic inputs Normal Biological Function Adaptive Responses Early cellular changes Cell injury, Inability to regulate Adverse Outcomes (e.g., mortality, Reproductive Impairment) Molecular initiating event Toxicity Pathway Perturbed cellular response pathway Adverse Outcome Pathway Adverse outcome relevant to risk assessment NAS,

13 Some Risk Assessment Uses of Systems Biology Hazard characterization: Hypothesis generation for AOPs/MOAs (maturing) Hypothesis testing of AOPs/MOAs (developing) Endpoint identification (immature/developing) Dose-response assessment: Characterize dose-response on biomarker data (developing) Decreased need for low dose extrapolation (developing) Reduced extrapolation across species (developing/immature) Exposure assessment Use biomarkers of effect to combine exposures (immature) High-throughput exposure assessments (EPA s ExpoCast program); RAIDAR and USETOX models immature 25 Stepwise Process for Estimating Genomic BMD Values Fit Each Gene with Statistical Model Calculate Dose at which response significantly deviates from control (i.e., BMD) Group Genes by Cellular Function Proliferation Apoptosis Estimate Dose at Which Cellular Function is Perturbed Pathway Transcriptional Response BMR NOAEL LOAEL -3 SD 34.1% 34.1% 13.6% 13.6% -2 SD -1 SD Mean +1 SD +2 SD +3 SD BMDL BMD Dose Thomas et al. Tox. Sci. 98:240, 2007 Yang et al. BMC Genomics 8:387,

14 Thomas et al. (2012 Mutation Research 746: ) found a strong correlation between transcriptional BMDs for specific pathways and traditional BMDs 27 Risk Assessment Needs Phenotypic anchoring for clinical, critical, or adverse effects Markers for common critical effects, such as decreased body weight & models for specific icities e.g., neurotoxicity Address communication among tissues; endocrine effects, in vitro to in vivo extrapolation Incorporate metabolism, differences in individuals and durations Test volatile chemicals 28 14

15 Susceptibility Factors? Internal Dose Factors Age Sex Genetic factors Preexisting disease External Exposure Factors Exposure variability Individual concurrent exposures Employment, diet, medications, psychosocial 29 Exposure Variability Often variability in exposure is large and scenariobased: Time Location Source of contamination Equipment (e.g., PPE) Worker and source mobility Environmental conditions Typically estimate: Mean, Median and UCL Representative and reasonable worst case 30 15

16 The Future Individualized Data Historically, sparse exposure data - 13% of epidemiologic studies used quantitative exposure Amount and quality of exposure data has been increasing new technology, regulations, and concepts (e.g. biomarkers and the exposome), may promote putting the E into G x E interaction studies Simple inexpensive direct reading exposure measurement techniques should allow for a broader and more comprehensive exposure assessment References [Rappaport and Kupper 2008; Smith and Rappaport 2009] 31 The EXPOSOME the measure of all the exposures of an individual in a lifetime and how those exposures relate to disease

17 Trend: Cumulative Risk EPA working hard on guidance on multiple exposure routes and stressors edia/webinars/2013/ NIOSH Initiatives focus on total exposure Total Worker HealthTM Exposome cs/exposome/ 33 Meek et al.,

18 Contemporary: PBPK It is now routine to ask folks whether or not a PBPK model is available for the chemical of interest. Numerous PBPK papers; some have been given top awards (RASS of SOT papers of the year): Sweeney, L. et al. (2001). Proposed occupational exposure limits for select glycol ethers using PBPK models and Monte Carlo simulation. Toxicol. Sci. 62(1): Kirman, C.R., et al. (2004). Addressing nonlinearity in the exposure-response relationship for a genotoxic carcinogen: cancer potency estimates for ethylene oxide. Risk Anal. 24: Contemporary: BMD Clear advantages and disadvantages exist in the use of a benchmark dose (BMD) Uses responses near the range of observation. Includes a measure of variability in the response. Determines a consistent measure of response. Applies to fewer, more robust, toxicity data sets. Accounts for more dose response of critical effect Casarett and Doull (Sixth Edition) page

19 Contemporary: Categorical Regression RfD Definition Regression model "without appreciable risk" r < 10 2 "is likely to be" P(*) > 0.95 "deleterious effect" severity = moderate or frank New RfD Definition P ( r < 10 2 at dose<rfd ) > 0.95 where r = P (severity >1) Hertzberg R.C. and M.L. Dourson, Contemporary: Categorical Regression Advantages: provides a consistent basis for calculating risk above the RfD all useful data can be categorized accounts for severity of toxic effect Limitations: animal to human extrapolation is still needed data are transformed into categories which loses information Haber et al Patty s Toxicology, Volume 1 (Fifth Edition) pages

20 Expedite the Future: Open Collaborative R&D Occupational Alliance for Risk Science (OARS) Alliance for Risk Assessment (ARA) Risk Information Exchange (RiskIE) International Toxicity Estimates for Risk (ITER) Beyond Science and Decisions: From Problem Formulation to Dose Response Mixtures and Combined Exposures Peer Review

21 Alliance for Risk Assessment (ARA) ( Stakeholder Process ARA Process States, Fed. Agencies, Public Interests, Industry Initiation of Risk Issue Steering Committee Risk Document Development Training and Certification Document Draft Peer Reviews Risk Information Exchange (RiskIE) Non-profit Collaborators Risk Communication Risk Research And Tools Peer Review & Consult Release to Public 41 Beyond Science and Decisions: From Problem Formulation to Dose Response 37 case studies 42 21

22 A Variety of Possibilities Essentially ( ), Hormesis ( ), Toxicity (---) Nonmontonic (red) Adverse Effect Background Range Beneficial Effect Increasing Dose Wherein lies endocrine disruption? 43 Summary Defend traditional risk methods; practice contemporary methods; harmonize at every opportunity Develop individual & cumulative exposures, measures of susceptibility, & data for genomics and 21st century toxicology so as to improve the biological basis of OELs and lead to more credible management decisions. Expedite the future thru collaboration 44 22

23 EXTRA SLIDES

24 Biologically-Informed D-R Modeling TiO 2 Tumor Progression TiO2 Lung Burden What we used to do! Inflammatory Cell Proliferation (Macrophages, PMN, etc.) Change in Alveolar Air:Blood Barrier / Loss of Integrity (Proteins in BALF) Fibrosis (Incidence) Cell Proliferation (BrdU labeling) Allen et al., in preparation Tumor (Incidence) 47 Key Limitations in TK How to assess the relative contribution of different enzyme systems, Reconciling differences between in vitro and in vivo data Role of other rate-limiting factors Impacts of epigenetics The lack of toxicokinetic data for many allelic variants, and The effect of co-exposures which could lead to either induction or inhibition. Haber et al. (2002) 48 24

25 The Future: Subpopulations Increasing focus on aging workforce Increasing focus on Children s risk EPA FQPA Factor California Children s Risk Guidance Plus a host of other susceptibility factors (background disease, etc.) Reference Results Trichloroethanol Glucuronide Renwick (1998) Child kinetic within factor < 3.16 in 19/22 chemicals, < 5 for 22/22 Theophylline Methotrexate Ganciclovir Naumann and Faria Child kinetic within factor of 3.16 for 3/3 C hem ical Digoxin Ciprofloxacin Rane (1992) Skowronski and Abdel- Rahman (2001) Calabrese (1986) chemicals, newborn poorer at clearance; tk factor of 3.16 covers 71% 5/6 chemical, kinetic factor <3.16 Animal adult; young LD 50 ratios, 86% of chemicals covered by total factor of 10 Chlorpheniramine Ceftibuten Busulfan Aztreonam Alfentanil Ratio of Child to Adult Value (1 Indicates Unity) Dourson, M.L., G. Charnley and R. Scheuplein,

26 (Epi)Genetics for Risk Growing but limited data (e.g. GWAS) Mode of action and adverse outcome pathways Assessing gene-environment interaction is the foundation Design studies to maximize information on both genetic and epigenetic variation Use of genetic and epigenetic data requires attention to ethical, legal, societal, and political implications. Schulte et al Tox 21: The Black Box Revealed Exposure Effect Water Perchlorate Blood Perchlorate Perchlorate uptake in thyroid Altered T3, T4, TSH Thyroid Histopathology Tumors CNS Exposure Internal Dose Effective Dose Early Biological Effect Altered Function: Critical Effect Clinical Disease Susceptibility Adapted from Schulte (1989); Farland et al

27 Tox 2: Biomarkers of Effect Statistically significant The authors think this argues against a hormone MOA, but does it? 53 Flow Charts for Evaluating Chemical Mixtures Whole Mixture Data Available Component Data Available Whole Mixture of Concern Sufficiently Similar Mixture Toxicologically Similar Components Mix of Toxicologically Similar & Independent Components Toxicologically Independent Components Mixture RfD/RfC; Slope Factor Health Evaluations Available Interactions Data Dose Addition Relative Potency Factors Integrated Additivity Methods Response Addition Whole Mixture Exposure Assessment Component Exposure Assessment Hazard Quotient; Risk Estimate Epidemiological Evaluations, Toxicity Profiles Interaction-Based Hazard Index, Interaction Profiles, Weight of Evidence, PBPK Models Hazard Index Index Chemical-Based Risk Estimate; Hazard Quotient Risk Estimate EPA, various references 54 27

28 RiskIE Risk Information Exchange An interactive Database to Communicate In-Progress Risk & Toxicity Assessments Includes over 7800 projects being conducted by more than 30 organizations representing 15 countries Available at the Alliance for Risk Assessment (ARA) website 55 ITERInternational Toxicity Estimates for Risk Provides chronic human health risk values and cancer classifications from organizations around the world for over 650 chemicals, including values from journal publications after quality assurance Includes synopsis on the underlying basis and rationale for each risk value and differences in risk values Links to each organization s website or source document - A forum through which independent parties can share - their peer reviewed risk values after peer review 56 28

29 Advancing Risk Assessment: Integrating Committee Recommendations Problem formulation should be linked to risk management; doing so does not pollute the risk assessment science The safe dose concept has evolved; CSAFs should be used Mode Of Action (MOA) is the assessment s organizing principle, but integrate key events in dose-effect continuum Key dose-dependent transitions are the norm; understanding MOA is essential for dose response Cumulative risk and mixtures assessment is iterative and should focus on the lowest tier needed to understand risk Biomonitoring is now interpretable; communication essential Dourson, Becker, Haber, Pottenger, Bredfeldt, and Fenner Crisp (Crit Rev Toxicol, 2013; 43(6): ) 57 Risk Assessment Peer Review Cornerstone principles Scientific robustness Selection of appropriate panel expertise & chair Transparency Independence Distinguish conflict-ofinterest from bias Avoid COI Balance biases Rule of thirds; ~1/3 of the panel should be Experience risk assessors Chemical or relatedchemical experts Effect experts Balanced affiliations 58 29