NEW PARADIGMS FOR ASSESSING AND

MANAGING CHEMICAL CARCINOGENS

Brazilian Benzene Seminar

Brasilia, Brazil

December 5, 2012

Rita Schoeny, Ph.D.Senior Science Advisor, Office of Science Policy, Office of Research and Development U.S. EPA

1

Disclaimer

The views expressed in this presentation are those do the author and do not represent the policy of the U.S. EPA.

2

Some of this is EPA policy

3

So What Is EPA Policy?

Science Policy Defaults, methods, Guidelines Used when there are data or methodology gapsPeer reviewedLots of documentation, which is publicly available

Policy based on scienceMay be set by EPA Executive LevelGenerally involves regulations or other risk management

choices; science is peer reviewed, action involves public comment; May be subject to Federal Advisory Committee Act

Lots of documentation; may be docket; publicly available

4

Examples Science policy

Cancer Guidelines 2005Set a reference dose for effects which are likely to

have a thresholdQuantitative adjustment to cancer risk for early life

exposureAnimal data are relevant to humans unless

demonstrated otherwise

Examples Policy set on science

Drinking Water Regulations are set as close as feasible to a Maximum Contaminant Level Goal

Consideration of residual risk after setting and air regulation requiring Maximum Achievable Control Technology○ What we are currently doing for electrical power

plantsCost / benefit choices

5

6

SDWA ‘96

Will regulation of the contaminant present a meaningful opportunity for health risk reduction?

Is the contaminant known or likely to occur in PWSs with a frequency and at levels posing a threat to public health?

Does the contaminant adversely affect public health?

Regulate with NPDWR

These are questions, demonstrations of risk

7

Risk Assessment Risk Management

’83 Risk Assessment Paradigm ’12?

Hazard Identification

Dose ResponseAssessment

Exposure Assessment

Risk Characterization

Risk Management Options

Statutory, legalconsiderations

Politics

Social Factors

Economics

Available Technology

Mode of Action

A lot has changed since ‘83

8

IPCS FRAMEWORK FOR ANALYSING THE RELEVANCE OF A CANCER MODE OF ACTION

FOR HUMANS

Exposure Science in the 21st Century: A Vision and A Strategy

NRC Silver Book Recommendation

NRC Silver Book recommendation (Chapter 8 “Improving Utility of Risk Assessment”)To make risk assessments most useful for risk

management decisions, the committee recommends that EPA adopt a framework for risk-based decision-making . . . that embeds the Red Book risk assessment paradigm into a process with initial problem formulation and scoping, upfront identification of risk-management options and use of risk assessment to discriminate among these options.

9

Con

firm

atio

n of

Util

ity

Planning & Scopingand Problem Formulation

Informing Decisions

ConceptualModel

AnalysisPlan

ExposureAssessment

Public/ Community/ Stakeholder Involvement

Effects AssessmentHazard Identification

Dose Response

Risk Characterization

Risk Assessment

Draft HHRA FrameworkSilver Book on Utility“Risk assessments should not be conducted unless it is clear that they are designed to answer specific questions, and that the level of technical detail and uncertainty and variability analysis is appropriate to the decision context” (NRC 2009, p. 247).

10

04/12/2023 11

12

NRC Scheme for Biomarkers (Schulte, 1989)

13

A Generalized Conceptual Model (adapted from USEPA, 2002; 2003)

S ources Exposure Pathways/RoutesS tressors EndpointsReceptors Risk Metrics

Activities that generate/release

Stressors or types of stressor releases

Chemical, physical or biological agents that cause an effect

Physical processes or interactions by which a stressor is brought into to contact with

receptor Populationsand/or lifestagesexposed tothe stressor

Measures of stressor effects or biological systems affected

Metrics by which risk is quantified(e.g., disease cases, hazard quotients, magnitude of

effect)

Drinking water disinfection

Variable mixturenitrosamines; dependent on treatment & source water. Ingestion of

nitrosamine mixture in drinking water Consumers of

drinking water; includes sensitive populations & life stages

Cancer, any site or type

Combined risk of cancer from subsetnitrosamines in mixture

Sources Stressors Exposure Receptors Endpoints Risk Pathways/ Metrics Routes

Conceptual Model Nitrosamines in Drinking Water

04/12/2023 16

Cancer Guidelines: What’s Different from 1986? Analyze data before invoking default options. Mode of action is key in decisions Weight-of-evidence narrative replaces the

previous “A-B-C-D-E” classification scheme. Two step dose response assessment

Model in observed range Extrapolate from point of departure

Consider linear and non-linear extrapolation Address differential risks to children

Risk Assessment Science

17

Analyze the available data

Is there too much uncertainty or is critical information lacking?

Invoke a default option

N

Y

Use Data Before Invoking Defaults

Conduct risk assessment

04/12/2023 18

Cancer Guidelines: What’s Different from 1986? Analyze data before invoking default options. Mode of action is key in decisions Weight-of-evidence narrative replaces the

previous “A-B-C-D-E” classification scheme. Two step dose response assessment

Model in observed range Extrapolate from point of departure

Consider linear and non-linear extrapolation Address differential risks to children

Evidence Maps to Cancer ClassificationsHuman Animal Indirect,

OtherIARC US EPA NTP

Sufficient -- --Carcinogenic to

humans (Group 1)

Carcinogenic to humans

Known to Be Human Carcinogen

Limited Sufficient

Strong human mechanistic data

-- Probably carcinogenic to

humans (Group 2A)

Likely to be carcinogenic to

humansReasonably

Anticipated to Be Human Carcinogen

Inadequate

Sufficient Strong

Limited Strong

Possibly carcinogenic to

humans (Group 2B)

Sufficient --

Limited Limited --

Inadequate Inadequate

Strong & same class as other carcinogens

Strong/ convincing

Inadequate Information to Assess

Inadequate Limited -- Not classifiable Suggestive Not classified 19

Zeise EEA Copenhagen Sept 3, 2010

04/12/2023 20

Cancer Guidelines: What’s Different from 1986? Analyze data before invoking default options. Mode of action is key in decisions Weight-of-evidence narrative replaces the

previous “A-B-C-D-E” classification scheme. Two step dose response assessment

Model in observed range Extrapolate from point of departure

Consider linear and non-linear extrapolation Address differential risks to children

21

Mode of Action and Cancer Assessment MOA is the keystone to all aspects of

the assessment process

True for other endpointsand is the major factor in harmonization among riskassessments

22

Why Do You Care about MOA ? MOA is key in Hazard Identification

Helps describe circumstances under which agent is carcinogenic (High dose? Route?)

Relevance of data for humans MOA determines choice of Low Dose

Extrapolation Life stage risk

MOE

Dose

Resp

on

se

(T

um

or

or

No

ntu

mo

r D

ata

)

0%

10%

EnvironmentalExposure Levels

of Interest

LED10 ED10

Nonlinear Default

EmpiricalRange of Observation

Range ofExtrapolation

x

x

xx

x

x NOAEL

LOAEL

x

23

Cancer

Breaking Down the Dichotomy

Non-ThresholdIrreversibleRisk value

Slope Factor Unit Risk Risk-Specific Dose

Non-Cancer

ThresholdReversibleSafety Value

RfD/RfC ADI/TDI MRL

24

Exposure

Toxicity

Key event

Key event

Key event

“. . . a sequence of key events and processes, starting with interaction of an agent with a cell, proceeding through operational and anatomical changes, and resulting in cancer formation. . . Mode of action is contrasted with “mechanism of action,” which implies a more detailed understanding and description of events, often at the molecular level, than is meant by mode of action”

Mode of Action

25

Mode of Action Frameworks

Hypothesized MOA: summary description and identification of key events

Experimental support: Strength, consistency,

specificity of association Dose-response

concordance Temporal relationship Biological plausibility and

coherence Consideration of the

possibility of other MOAs Relevance to humans

Postulated mode of action (theory of the case) Key events Concordance of dose-response relationships Temporal association Strength, consistency and specificity of association of tumour response with key events Biological plausibility and coherence Other modes of action Uncertainties, Inconsistencies, and Data Gaps Assessment of postulated mode of action

U.S. EPA IPCS

26

MOA/Human Relevancy

ILSI/IPCS

YES

NO

NO

YESMOA notRelevant

YESMOA notRelevant

NOProceed withRisk assessment

Proceed withrisk assessment

Is the weight of evidence sufficient to establish a mode of action (MOA) in animals?

Can human relevancy of the MOA be reasonably excluded on the basis of fundamental, qualitative differences in key events between animals and humans?

Can human relevancy of the MOA be reasonably excluded on the basis of quantitative differences in either kinetic or dynamic factors between animals and humans?

27

Key Event A “key event” is an empirically observable

precursor step that is itself a necessary element of the mode of action or is a biologically based marker for such an element.

Key event is necessary, but not sufficient

If a key event doesn’t occur, there is no cancer

If one key event occurs, there may or may not be cancer

28

ChloroformCYP2E1

Phosgene

Regenerative Cell Proliferation

Postulated Mode Of Action

MetabolismOxidative

Sustained Toxicity

Tumor Development

Key Events

29

Postulated Mode Of Action CPMetabolism

Phosphoramide mustard, PAM

Acrolein

DNA damage

Mutations

Tumor Development

Cyt p 450s

04/12/2023 30

Mode of Action Framework

Hypothesized MOA: summary description and identification of key events

Experimental support:Strength, consistency, specificity of associationDose-response concordanceTemporal relationshipBiological plausibility and coherence

Consideration of the possibility of other MOAs

Relevance to humans

04/12/2023 31

Based on “Hill Criteria” for Causality

Experimental support:Strength, consistency,

specificity of association

Dose-response concordance

Temporal relationshipBiological plausibility

and coherence

“None of my nine viewpoints can bring indisputable evidence for or against the cause-and-effect hypothesis and none can be required as a sine qua non. What they can do, with greater or less strength, is to help us to make up our minds on the fundamental question — is there any other way of explaining the set of facts before us, is there any other answer equally, or more, likely than cause and effect?” Hill (1965)

04/12/2023 32

Experimental Support for MOA Strength, consistency, specificity of association

What is the level of statistical and biological significance for each event and for cancer?

Do independent studies and different experimental hypothesis-testing approaches produce the same associations?

Does the agent produce effects other than those hypothesized?

Is the key event associated with precursor lesions?

Section2.4.3.2

04/12/2023 33

Experimental Support for MOA Dose Response Concordance

○ Is precursor induced at lower dose than tumors?

○ If greater incidence of the precursor occurs, does incidence of tumor increase?

○ If the precursor event is more severe is there an increase in tumor incidence?

04/12/2023 34

Experimental Support for MOA Temporal Relationship

○Do the precursor events occur before tumors are observed?

○ Is this observed in independent studies?

Mode of Action: Bladder Tumors, Key EventsCytotoxicity and Regenerative Hyperplasia

DMAIII

Metabolite

Hyperplasia

UrothelialToxicity

RegenerativeProliferation

Tumor

Sustained BrdU Labeling

Measurable Key Events in Target Tissue

BrdULabeling

SEM

04/12/2023 36

Dos

e R

espo

nse

Con

cord

ance

Temporal

Dose (mg/kg bw/day)

Metabolism DMAVDMAIII

Urothelial Toxicity

Regenerative Proliferation

Urothelial Hyperplasia

Transitional Cell

Carcinoma

0.2(2 ppm)

+(wk 3-0.03 ± 0.01 uM)

+(wk 10-6/10, grade 3 or 4)

- - -

1(10 ppm)

+(wk 3-0.12 ± 0.02 uM)

+(wk 3-2/7, grade 3) (wk- 10; 8/10, grade 3 or 4)

slight(wk 10-1.5X inc) - -

4(40 ppm)

+(wk 3-0.28 ± 0.09 uM)

+(wk 3-7/7, grade 3) (wk 10-5/10, grade 3 or 4)

+(wk 10-4.3X inc)

+(wk 10- 4/10) -

9.4(100 ppm)

+(wk 3-0.55 ± 0.15 uM)

+(6 hrs-6/7, grade 3) (24 hrs-4/7, grade 3 or 4)(wk 2 6/10, grade 5)(wk 10-0/10, grade 4 or 5)

+(wk 1- 2.2X inc) (wk 2-3.9X inc) (wk 10-4.2X inc)

+ (wk 8-7/10)(wk 10-9/10)

+ (papilloma first obs at wk 107; carcinoma first obs at wk 87)

Association of Key Precursor Events & Bladder Tumors in F344 Rats

04/12/2023 37

Experimental Support for MOA Biological plausibility and coherence

○ Does the MOA make sense given what is known about carcinogenesis in general, and for the case specifically?

○ Are carcinogenic effects and events consistent across structural analogues?

○ Is the database on the agent internally consistent in supporting the MOA, including relevant non-cancer toxicities?

04/12/2023 38

Possible Key Events

Strength ConsistencyReproducibility

Specificity Temporal Biological GradientDose-Response

Biological Plausibility

Coherence Key Event(Causal)Associated(Marker?)Modulatory?Neither

CAR activation High, Required for tumors1,2

Bear in mind that both ref #1 & #2 involve initiation with den and a rather short period of treatment with pb, 32 wk in ref #1 & 30 wk in ref #2, we do not know what would happen in car ko mice if they were treated with pb for 2 years(JGoodman)

High2 In vivo studies1,2

In vitro nuclear translocation3,4

HighKnock out studies1,2

Earliest event, in vivo and in vitro14

Perhaps PB-induced activation of the phosphatase that appears to facilitate the translocation of CAR from the cytoplasm to the nucleus is the very earliest event?

Not determined(any data?) No data that I am aware of.Only in vitro (93, 58)

Yes HighFits in a logical sequence

Causal(DWolf)(RPeffer) (RBars)(RSchoeny)(CElcombe)Causal, under the experimental conditions outlined in Ref #1 & #2(JGoodman)

Nuclear Receptor Workshop

39

MOA/Human Relevancy

ILSI/IPCS

YES

NO

NO

YESMOA notRelevant

YESMOA notRelevant

NOProceed withRisk assessment

Proceed withrisk assessment

Is the weight of evidence sufficient to establish a mode of action (MOA) in animals?

Can human relevancy of the MOA be reasonably excluded on the basis of fundamental, qualitative differences in key events between animals and humans?

Can human relevancy of the MOA be reasonably excluded on the basis of quantitative differences in either kinetic or dynamic factors between animals and humans?

Concordance Analysis of Key Events: Cytotoxic Mode of Action

Key Event Rodents Humans

Presence of metabolite Yes Yes

Persistent cytototoxicity

Yes Possible

Persistent regenerative proliferation

Yes Possible

Tumors Yes Possible

Concordance analysis of key events is for the MOA and not necessarily chemical specific

Chemical specific and generic information relevant to adverse outcome is useful

α2μ Globulin and Male Rat Kidney Tumors

a2u

-Globulin

Renal Reabsorption

Nucleus

Heterolysosome Amino Acids

Circulation

Chemical-a2u -Globulin "Complex"

Renal Reabsorption

Nucleus

Circulation

Heterolysosome

04/12/2023 42

α2u-Globulin The main story

Protein produced by male ratsMOA– functional changes in epithelial cells of

proximal tubules

• Hyaline droplets accumulate • Tubule cell degeneration• Regenerative cell

proliferation• Expansion of initiated renal tubule cells

Human Relevance of α2u-globulin nephropathy – U.S. EPA

Humans do not possess a protein that is similar to α2u in abundance or binding characteristics; thus, humans would not be at risk of developing a chemically induced protein-mediated nephrotoxic response. Borghoff SJ and Lagarde WH, Toxicol Appl Pharmacol. 1993 Apr;119(2):228-35.

If male rat kidney tumors developIn the absence of female rat kidney tumorsMale rats have increased hyaline dropletsHyaline droplets contain a2u-GlobulinCharacteristic nephrotoxicity

The male rat kidney tumors are not relevant for human health risk and would not be included in a risk assessment.

Is the MoA for phenobarbital plausible in humans?

Key event in MoA Plausible in humans?

Activation of CAR

Induction of CYP2B Increase via CAR can occur

Hypertrophy Yes

Cell proliferation Not likely, based on in vitro and in vivo data

Inhibition of apoptosis Possible but not likely, based on limited in vitro data

Selective clonal expansion Possible but not likely, none reported

Occurrence of liver tumors No, based on epidemiological data

04/12/2023 45

Cancer Guidelines: What’s Different from 1986? Analyze data before invoking default options. Mode of action is key in decisions Weight-of-evidence narrative replaces the

previous “A-B-C-D-E” classification scheme. Two step dose response assessment

Model in observed range Extrapolate from point of departure

Consider linear and non-linear extrapolation Address differential risks to children

46

MOA and Kids

Supplemental Guidance for Assessing Susceptibility from Early-Life Exposure to CarcinogensEffects observed in childhoodEarly life exposures that contribute to later

life effectsMOA determines whether quantitative

adjustment is made

47

Supplemental Guidance Use age-specific values for exposure and potency When data permit, develop separate potency estimates for childhood exposure In risk characterization, mutagenic MOA risk

is increased by age-dependent adjustment factor (used with exposure info for age group)

○ <2 yrs old, 10 fold○ 2 to < 16yrs, 3 fold

No MOA, linear extrapolation without ADAF; non-linear MOA, no ADAF

48

Framework for Determining

a Mutagenic Mode of

Action for CarcinogenicityUsing EPA’s 2005 Cancer

Guidelines and Supplemental

Guidance for Assessing

Susceptibility from Early-Life

Exposure to Carcinogens

This may actually get published at some point in time

49

Framework on Default MOA “ It should also be noted that there is no

‘default MOA.’ The Cancer Guidelines offer some default procedures to use when no MOA can be determined.”

• No MOA is not the same as a Mutagenic MOA

Determination of mutagenic MOA is as scientifically rigorous as any other MOA

I found nothing in IPCS on a default MOA

04/12/20235050

To boldly go where no regulatory toxicologist has gone before . . .

Mutagenicity is the induction of permanent, transmissible changes in the amount, chemical properties, or structure of the genetic material. These changes may involve a single gene or gene segment, a block of genes, parts of chromosomes, or whole chromosomes. Effects on whole chromosomes may be structural and/or numerical (e.g., aberrations and/or aneuploidy).

04/12/202351

Boldly part 2Genotoxicity is the induction of alterations to genetic material. It is a broader term than mutagenicity in that genotoxicity refers to potentially harmful effects on genetic material, which are not necessarily persistent and transmissible. Genotoxicity may be mediated directly or indirectly by chemical or physical agents, and may or may not be associated with mutagenicity. Tests for genotoxicity include tests for mutagenicity, as well as other tests which provide an indication of induced damage to DNA. For example, such tests may include unscheduled DNA synthesis (UDS), sister chromatid exchange (SCE), mitotic recombination, DNA adduct formation, or DNA strand breaks.

52

What does this definition do? Distinguishes between genotoxic and

mutagenic (U.S. EPA Guidelines refer to “mutagenic MOA”; Europe deals with genotoxic).

Includes numeric changes in chromosome as mutagenicBut not likely to be linear at low dose

Includes agents that generate reactive oxygen species (ROS)But need to consider low dose response

53

Framework: Multi-step Process Risk assessment

is an iterative process

Visualize the Framework as series of linear stepsAssemble dataAssess, weigh data qualityWOE for mutagenicityWOE for MOA

54

Step 2: Evaluate Data Quality

Look at primary papers Judge against current

acceptability criteria Cites publications for evaluating quality (e.g.

Cimino 2006, OECD, ICH, IWGT, DHHS 2006)

Keep, but weigh

55

Step 3 Gene- tox Tests Measure Different EventsStep 3 Gene- tox Tests Measure Different Events

Genotoxicity Assays

Type of Damage

Mouse Lymphoma

ChromosomeAberrations CHO cells

Ames Bacterial Mutagenicity

Point mutation Yes No YesOligonucleotide insertion or deletion

Yes No Yes

Allele Loss Yes No NoSmall Chromosome alteration

Yes ? No

Large Chromosome alteration

Yes Yes No

Aneuploidy ? Yes No

Adapted from M. Moore (2004)

Cancer Hazard ID 55TERA’s Dose-Response Assessment Boot Camp

56

Step 3: WOE for Mutagenic Activity Categorize data – suggest use of our table in

Appendix A.Put in all data with notes on qualityUse consistent terms for assay types or

endpoints: positive, negative, inconclusive, contradictory

Present summary of

database

57

In vitro AssaysTest System

Concen-trations

Cytotoxicity observed

Duration of Exposure

Results With metabolic activation (+ S9)

Results Without metabolic activation (- S9)

Conform to relevant guideline

Ref

Gene MutationBacterial Salmonella, reverse mutation

E. coli, reverse mutation

Mammalian CHO gene mutation, hprt locus

Mouse L5178Y, tk locus

Chromosome Mutation

Micronucleus assayChromosomal aberrations

DNA EffectsMammalian

Unscheduled DNA synthesis

Sister chromatid exchanges

Comet assay

DNA adduct analysis

Lower Eukaryote

Saccharomyces cerevisiae, gene conversion

58

WOE for Mutagenic Activity

Conclusions across endpoints: some endpoints carry more weight than others○ e.g. Sperm head morphology may be caused

by modification of protein structure○ Morphologic cell transformation does not

measure mutationWOE for mutagenic activity: negative, data

are inadequate, data are of questionable quality, data are equivocal, data are positive

Apply MOA Framework

59

Mutagenicity + carcinogenicity ≠ Mutagenic MOA

Initiating

Mutation

Multiple events

Toxicity

Altered Gene Expression

Cell Proliferation

InitiatingMutation

Multiple events

Tumor

Mutagenic Carcinogen

Nonmutagenic Carcinogen

Tumor

04/12/202360

Data Preference: WOE Mutagenic MOA

Cancer-relevant oncogene or tumor suppressor gene mutations detected in target tissue after chemical exposure.

Surrogate gene mutations detected in target tissue after chemical exposure.

Chemical-specific DNA adducts (known to be mutagenic adducts) in target tissue after chemical exposure.

Primary DNA damage in target tissue after chemical exposure.

Gene mutations or chromosome aberrations in surrogate tissues after in vivo exposure.

DNA adducts or other measures of DNA damage and/or repair or in surrogate tissues after in vivo exposure.

Mutations, cytogentic damage, DNA adducts and/or primary DNA damage in vitro.

61

What Has a Mutagenic MOA? Cyclophosphamide

Alkylating

Cytotoxic

Cytotoxic, alkylating

62

Mode of Action

Adverse Outcome Pathway

Source to Outcome Pathway

Toxicity Pathway

Source

EnvironmentalContaminant

Exposure

Molecular Initiating Event Cellular Effects Individual Population Community

Application to Levels of Organization Based on

Source to Outcome

63

Application to Levels of Organization Based on Source

to Outcome Source

EnvironmentalContaminant

Exposure

Cellular Effects

Individual

Population

Community

Mode of Action

Adverse Outcome Pathway

Source to Outcome Pathway

Toxicity Pathway

Molecular Initiating Event

04/12/2023 64

Cancer Guidelines: What’s Different from 1986? Analyze data before invoking default options. Mode of action is key in decisions Weight-of-evidence narrative replaces the

previous “A-B-C-D-E” classification scheme. Two step dose response assessment

Model in observed range Extrapolate from point of departure

Consider linear and non-linear extrapolation Address differential risks to children

65

High dose data – what do they tell us?

Res

pons

e

Dose

Take home message

In science, data before defaults Scientifically based low dose

extrapolation before defaults Science may inform policy; policy should

never affect science

66

67

PossibilitiesR

espo

nse

Dose

Interspecies

68

Two Step Approach

Model data in the observed range – to a point of departure

Extrapolate below the POD

UF

Dose

Res

pons

e (T

umor

or

Non

tum

or D

ata)

0%

10%

EnvironmentalExposure Levels

of Interest

LED10 ED10

Nonlinear Default

EmpiricalRange of Observation

Range ofExtrapolation

x

x

xx

x

x NOAEL

LOAEL

x

69

Extend the Observed Range Using Precursor Data Objective of choosing POD is to set it as

close to environmental levels as Supported by dataAppropriate to model

Cancer Guidelines say precursor data are useful for this

Must have MOA

Section3.2.2

70

Endpoint Duration

Feeding

Duration

Drinking water

10% 1% 10% 1%

BMD (mg/kg/d)

BMDL(mg/kg/d)

BMD(mg/kg/d)

BMDL(mg/kg/d)

BMD(mg/kg/d)

BMDL(mg/kg/d)

BMD(mg/kg/d)

BMDL(mg/kg/d)

Tumor 104 weeks 7.74 5.96 6.80 2.22 104

weeks 1.92 1.21 0.88 0.14

Hyperplasia

10weeks 1.36 1.04 0.42 0.32

104 weeks 1.63 1.04 0.74 0.14

104 weeks 1.97 1.61 0.93 0.66

BrdU labeling

10 weeks 0.65 0.29 0.54 0.07 Not determined. Available data not suitable for modeling.

Cytotoxicity

3 weeks 0.68 0.18 0.31 0.02

No reliable dose-response data available

10 weeks 0.02 0.008 0.002 0.0007

Cacodylic Acid: BMDs and BMDLs

Linear or Non-linear?

71

UF

Dose

Re

sp

on

se

(T

um

or

or

No

ntu

mo

r D

ata

)

0%

10%

EnvironmentalExposure Levels

of Interest

LED10 ED10

Nonlinear Default

EmpiricalRange of Observation

Range ofExtrapolation

x

x

xx

x

x NOAEL

LOAEL

x

Two Step Dose Response Process

Another Question First

Is There Something Better?

72

Analyze all data before using defaults

Is there too much uncertainty or is critical information lacking? Invoke a

default option*

N

Y

Analyze the available data

Conduct risk assessment

73

Specify the sequence(s) of key events

dosimetry Key event B1

Key event B2

Key event A1

Key event A2

Key event A3

Assessment endpoint

Mode of Action

Exposure

Tissue dose

Mode of action

Response

Source

PBPK

BBDR

74

Biologically Based Dose Response Model

75

(PBPK)

AppliedDose of Phenobarbital

Multiple dose-responses

and time-courses

76

Reality check (I) There are always data gaps

ArsenicFormaldehydeTCDDphenobarbital

A BBDR model is a description of biological structure with embedded empirical linkages that cover the parts of the overall exposure-dose-response linkage for which data are missing.

77

Reality check (II) As research improves our understanding of

the overall exposure-dose-response linkage, the sophistication of the description of the mode of action increases.

Corresponding iteration of the BBDR model leads to more accurate predictions of dose-response and time-course behaviors.

Will always be some degree of residual uncertainty.

But is the default more uncertain?

And if no BBDR?

78

UF

Dose

Re

sp

on

se

(T

um

or

or

No

ntu

mo

r D

ata

)

0%

10%

EnvironmentalExposure Levels

of Interest

LED10 ED10

Nonlinear Default

EmpiricalRange of Observation

Range ofExtrapolation

x

x

xx

x

x NOAEL

LOAEL

x

Two Step Dose Response Process

Linear or Non-linear

Mutagenesis Paradigm

DNA Damaged DNA

Mutagens/Spontaneous

Damage Sensing

Cellular Response

DNA Repair

Repaired DNA

IncorrectRepair/Replication

Mutant DNA

No repair

Dead Cell79

Demarini 70

Threshold? Demonstrated

By inspection of the dose response curve

Fitting models and checking goodness of fit

Statistical tests for one model or another

Based on MOAMutagenic MOA has

been linearBut should consider

biology of mutation

Does mutagenic MOA mean low dose linear?BBDR should be first choice 80

MMS

MNU

2011 EMS Annual Meeting Pottenger 81

In vitro Mutation Dose-Response: MMS & MNUDoak et al., 2007

MMS

MNU

HPRT MF

NOEL = 1 mg/ml

No NOEL

Johnson et al., 2009

ENU threshold dose-response (Lutz & Lutz model)

In vitro Mutation Dose-Response: ENU

HPRT MF

Slide from Pottenger 82

Take Home Message

MOA informs dose response assessment

DNA damage is not mutation Mutation is not cancer Some genotoxicity endpoints may be

reasonable biomarkersMay be useful for extending the lower end of

dose response curveUseful in MOA

83

Take home message 2

In science, data before defaults Scientifically based low dose

extrapolation before defaults Science may inform policy; policy should

never affect science

84

85

86

NRC 2009 Silver Book 1 Framing questions

and design step. Risk Assessment is

not an end in itself. Characterize

uncertainty and variability

Default before data?

These are strictly my own opinions

87

NRC 2009 Silver Book 2 Dose response

Additivity to background is a major theme○ How differentiate between exogenous and

endogenous damage?○ DNA adducts biomarkers, could have major role○ Does this mean linear all the time?

EPA has expressed preference for BBDR○ Low dose data for adduct formation○ Low dose data for mutation○ Low dose data for other markers

Again my own opinions

88

Cancer

Breaking Down the Dichotomy

Non-ThresholdIrreversibleRisk value

Slope Factor Unit Risk Risk-Specific Dose

Non-Cancer

ThresholdReversibleSafety Value

RfD/RfC ADI/TDI MRL