1. Mutagenic MOA Carcinogens: How High is the Burden of Proof ?
RASS Telecom
09/10/08
Rita Schoeny, Ph.D.
Senior Science Advisor
Office of Water, U.S. EPA

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

3. Risk Assessment is constantly evolving
Science and Judgment
Describe all defaults
Ensure they are health protective, documented, departures are warranted
Cancer Guidelines 2005
Use data before defaults
Rather than determine how much data needed to depart from default
Including default procedures such as linear low dose risk

4. Mode of Action and Cancer Assessment
MOA is the keystone to all aspects of the assessment process
True for other endpoints
and is the major factor in
harmonization among risk
assessments

5. 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
Examples of relevance – alpha2 u –globulin, atrazine hormonal interaction.
Keystone of the revised Cancer Guidelines
Determine relevance of data for humans
Determine conditions of risk
Choose low dose extrapolation process
Expect to see more use in harmonizing “cancer vs. non-cancer” assessment
Affects the bottom line choices for risk management

6. Mode of Action Exposure 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”
Key event I
Key event
Toxicity

7. Mode of Action Frameworks
U.S. EPA
IPCS
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
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
EPA’s framework is consistent with others, e.g., the International Programme on Chemical Safety and by the International Life Sciences Institute

8. MOA/Human Relevancy ILSI/IPCSNO
Is the weight of evidence sufficient to establish a mode of action (MOA) in animals?
Proceed with
risk assessment
YES
Can human relevancy of the MOA be reasonably excluded on the basis of fundamental, qualitative differences in key events between
animals and humans?
YES
MOA not
Relevant NO
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?
YES NO
MOA not
Relevant
Proceed with
Risk assessment 8

9. 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
Necessary (but not sufficient)

10. Postulated Mode Of Action
Chloroform
CYP2E1
Metabolism
Oxidative
Phosgene
Sustained Toxicity
Lots of data for phosgene production, cell death and proliferation
In rodent bioassay studies, Chlorofom has been shown to result in liver and kidney tumors. The postulated MOA to explain these tumor responses involves bioactivation through CYP2E1 oxidative metabolism as the rate limiting step in chloroform’s mode of action. Metabolism by this pathway produces cytotoxic metabolite within the target organ, in particular phosgene that injures and kills cells, cytotoxicity is followed by regenerative cell proliferation, and it cytotoxicity/regenerative proliferation is sustained, eventually tumor development. So,these are the key sequence of events that will be considered with respect to chloroform;s included tumorigenesis in the rodent kidney and liver.
Regenerative Cell Proliferation
Key Events
Tumor Development

11. MOA and Kids
Supplemental Guidance for Assessing Susceptibility from Early-Life Exposure to Carcinogens
Effects observed in childhood
Early life exposures that contribute to later life effects
MOA determines whether quantitative adjustment is made
If applied with exposure adjustments comes out to about 1.6 x.

12. 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
If applied with exposure adjustments comes out to about 1.6 x.
“When MOA cannot be established, the policy choice would be to use linear extrapolation to lower doses such that risk estimates are based on a lifetime average daily exposure without further adjustment.
EPA expects to issue additional guidance to address moa other than mutagenicity when sufficient data are available and analyzed.

13. Framework for Determining a Mutagenic Mode of
Public Comment
completed 12/07
Framework for Determining
a Mutagenic Mode of
Action for Carcinogenicity
Using EPA’s 2005 Cancer
Guidelines and Supplemental
Guidance for Assessing
Susceptibility from Early-Life
Exposure to Carcinogens
External peer review
completed 05/08
In the process of revising the document. Aiming for end of October. Then reviewed by the RAF, SPC and OMB.
www. epa.gov/
osa/mmoaframework/
pdfs/MMOA-ERD-FINAL
pdf

14. 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.”
MOA determinations follow Cancer Guidelines
Framework
If insufficient data to support MOA, use low dose
linear extrapolation and no ADAF
Determination of mutagenic MOA is as
scientifically rigorous as any other MOA

15. What is a mutagen?
A chemical that induces biologically relevant mutations in any one of a number of validated mutation assays
Mutation assays detect the induction of mutants
Mutants are cells with genetic alterations that can be passed to viable daughter and granddaughter cells -- heritable

16. What is “Mutagenic”?
EPA does not have a standard definition of “mutagenic”
Operationally for use in “mutagenic MOA for cancer”
“. . . capacity of either the carcinogen or its metabolite to react with or bind to DNA in a manner that causes mutations. In this context, mutagens usually (though not always) produce positive effects in multiple test systems for different genetic endpoints, particularly gene mutations and structural chromosome aberrations, both in vitro and in
vivo.”
The peer reviewers hated it
This definition is for cancer MOA only. New definition focusses on distinction between geno toxic and mutagenic. Focus on mutation as a permanent change in DNA, one that is heritable.
Germ cell mutagenicity Guidelines mostly lists test that are relevant. Doesn’t really define mut.
Panel chose not to try to parse out differences between genotoxic and mutagenic.

17. Framework: Multi-step Process
Risk assessment
is an iterative
process
Visualize the Framework as series of linear steps
Step 1 is assemble relevant data
Genetic toxicity testing, tumor data, pk, SAR, etc.
Framework describes test batteries

18. Step 2: Evaluate Data Quality
Look at primary papers
Judge against current
acceptability criteria
e.g. were tests done at
cytotoxic levels
Cites publications for evaluating quality (e.g. Cimino 2006, OECD, ICH, IWGT, DHHS 2006)
Keep, but weigh
International Conference on Harmonization. Note that older studies in particular may have used conditions of testing no longer considered acceptable. Important when faced with contradictory results.

19. Gene- tox Tests Measure Different Events
Cancer Hazard ID
Gene- tox Tests Measure Different Events
Genotoxicity Assays
Type of Damage
Mouse Lymphoma
Chromosome
Aberrations CHO cells
Ames Bacterial Mutagenicity
Point mutation
Yes No
Oligonucleotide insertion or deletion
Allele Loss
Small Chromosome alteration ?
Large Chromosome alteration
Aneuploidy
And a lot of DNA interactions are not on this particular chart. One point is that DNA damage is not mutation.
TERA’s Dose-Response Assessment Boot Camp
Adapted from M. Moore (2004)
TERA's Dose-Response Assessment Boot Camp 19

20. Step 3: WOE for Mutagenic Activity -- 1
Evaluation requires someone expert in gene-tox (all tests don’t measure same things)
Categorize data – suggest use of our table in Appendix A.
Put in all data with notes on quality
Use consistent terms for assay types or endpoints: positive, negative, inconclusive, contradictory
Present summary of
database
Categorize data – suggest use of our table in Appendix A.
Put in all data with notes on quality
Conclusion for each assay type (e.g. Salmonella)
Conclusion for each type of effect (e.g. point mutation or clastogenicity)
Inconclusive refers to the data set of test results that cannot be
definitively termed sufficient or insufficient due to any of the following:
borderline responses;
insufficient number of test strains, or poor performance of the test organisms;
inadequate testing of exogenous metabolic activation for in vitro assays;
inadequate doses or concentrations (either too high or too low);
inadequate dose spacing;
inadequate sampling time(s);
positive responses observed only at unacceptable levels of cytotoxicity; or
statistical significance in the absence of biological significance.
OECD 1997

21. WOE for Mutagenic Activity -- 2
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 mutation
Hierarchy of data utility
DNA interaction ≠DNA damage ≠mutation
e.g. most useful are mutations in relevant genes in humans
WOE for mutagenic activity: negative, data are inadequate, data are of questionable quality, data are equivocal, data are positive
No Checklist
No Minimum Data Set
Cell transformation doesn’t measure a genetox effect. Aneuploidy is often a late stage observation in tumors.
Positive results in one or more in vivo test(s) in mammalian systems and/or positive results in multiple species in vitro

22. How to Weigh the Evidence as to Whether a Chemical Causes Specific Tumors by a Mutagenic Mode of Action (Mutation is THE Key Event)
(Listed in decreasing order of relevance/importance)
Cancer relevant oncogene/tumor suppressor gene mutations can be detected in the target tissue following chemical exposure
Surrogate gene mutations can be detected in the target tissue following chemical exposure
DNA adducts (known to be mutagenic adducts) can be detected in the target tissue following chemical exposure
Primary DNA damage can be detected in the target tissue following chemical exposure
Gene mutations and/or DNA adducts or other measures of primary DNA damage can be detected in vivo.
Evidence that the chemical can induce mutations, cytogenetic damage, DNA adducts and/or primary DNA damage in vitro.

23. Not Finished yet Mutagenicity + carcinogenicity ≠ Mutagenic MOA Step 4
Apply MOA Framework
Hypothesized MOA
Experimental support:
Dose-response concordance
Strength, consistency, specificity of association
Temporal relationship
Biological plausibility and coherence
Consideration of the possibility of other MOAs
Relevance to humans
Step 4

24. Key Events DNA changes resulting in mutation
“ For a chemical to act by a mutagenic MOA, either the chemical or its direct metabolite is the agent inducing the mutations that initiate cancer.”
“This is contrasted with a MOA wherein mutagenicity occurs as an indirect effect of another key event in carcinogenesis.”
Properties for mutagenicity as the key event
Long list in Guidelines: early tumor response, initiator, target tissue is exposed to DNA-reactive chemical, mutation is early event, mutation in oncogenes, etc
Indication of types of data supporting WOE
Consistency across assays
Induction of ≥ type of effect
Effects in vivo
Mutation in absence of cytotoxicity
Belongs to a class of compounds with established mutagenic MOA
Including the “Kid Guidance 12”

25. Tumor Induction: Time-related Accumulation of Events
Mutagenic Carcinogen
Multiple events
Initiating
Mutation
Tumor
Nonmutagenic Carcinogen
Toxicity
Altered Gene Expression
Cell Proliferation
Initiating
Mutation
Multiple events
Tumor

26. Applying the MOA Framework
Types of data supporting WOE
Consistency across assays
Induction of ≥ 1 type of
effect
Effects in vivo
Mutation in absence of cytotoxicity
Belongs to a class of compounds with established mutagenic MOA
Including the “Supplemental Guidance 12”
Positive in at least one in vivo mutagenicity assay
Observation of multiple genetic effects (point mutations, chromosome aberrations)
Positive in vitro results at noncytotoxic concentrations in somatic cells from a diverse range of phylogenetically distinct species (bacteria, yeast, fungi, plants, insects mammals or humans).
DNA reactivity or ability to bind to DNA (e.g. DNA adduct formation, UDS, DNA strand breaks)

27. Cyclophosphamide Cytotoxic, alkylating Alkylating Cytotoxic
Pharmaceutical – antineoplastic, (Hodgkin’s, non-Hodgkin's, leukemia, multiple myeloma, neuroblastomas, ovarian adenocarcinomas, some lung). Used as immunosuppressant in arthritis treatment, scleroderma, glomerulonephritis, chronic hepatitis, multiple sclerosis, organ transplants.
Given orally or i.v..
Associated with bladder cancer and leukemia.
Animals – rat: bladder, hematopoietic, neurological sarcomas, reticulum cell sarcomas, hemangioendotheliomas, osteosarcoma, neoplasms of lung, liver, testis, mammary.
Mice: mammary, pulmonary sarcomas and adenomas, liver tumors, skin. Carcinogenic transplacentally.
Carcinogenic in animals
transitional-cell carcinomas of the urinary bladder in ♂ rats & malignant tumors at numerous sites: bladder, lung, liver, testis, ovaries in rats & mice
Carcinogenic in humans, causing bladder cancer (IARC, 1998)
Alkylating
Cytotoxic

28. Postulated Mode Of ActionCP
Metabolism Cyt p 450s
Phosphoramide mustard, PAM Acrolein
DNA damage
Metabolism to phosphoramide mustard (PAM)
DNA damage (e.g.,DNA adduct formation)
Induction of multiple adverse genetic events (mutation and/or chromosomal aberrations) and/or cytotoxicity
Regenerative proliferation ( cell proliferation, organ weight, hyperplasia)
Bladder tumors
Tumor
Development
Mutations

29. Cyclophosphamide GAP
GAP developed by Mike Walters horizontal line divides + from - results; looking from left to right, assays are categorized according to species. As you can see, CP is positive in multiple systems at multiple endpoints across phylogeneticallt distinct species.
Starting at the very left, we see that CP is + for gene mutations in bacteria (S.typhimurium& E.coli),yeast (S. cerevisiae), insects(somatic cell mutations in Drosophilia melanogaster) & mammalian cells (mouse lymphoma),+ for chromosome aberrations in
Pink lines represent germinal cell assays+ sex-linked recessive lethals, heritable translocations, dominant lethal mutations in Drosophilia & dominant lethal mutations, heritable translocations in rats &/or mice
Positive results in multiple systems at multiple endpoints across many species both in vitro & in vivo. In vitro, need S9

30. CP In Vivo Tests: Animals
Gene Mutation Assays
Positive Mouse Spot Test ( mg/kg)
Positive Muta Mouse (lacZ) 100 mg/kg x 5 days in bone marrow
B6C3F1 mouse (lacI) 100 mg/kg MF increased in lungs and urinary bladder
No transgenic studies in rats
The assay showing increased MFs in the urinary bladder will be important for site concordance when we get to the framework analysis.

31. CP In Vivo Tests: Humans
Micronuclei peripheral blood lymphocytes (PBL) & buccal epithelials 26/26 nurses handling CP
Structural chromosome aberrations & SCE, gene mutations or DNA damage (Comet assay) in PBL or bone marrow, patients
Structural chromosome aberrations in children
Mutation of p53 in bladder tumors (cumulative doses of mg/kg)
6-Thioguanine-resistant T lymphocytes from multiple sclerosis patients (750 mg/m2)

32. So CP Is Mutagenic
And it’s carcinogenic
Apply MOA Framework

33. Dose Resp Concordance Mutation is key event
Expect mutations and / or DNA interaction at lower dose than tumors
Mutation is not the key event
Expect increased mutation at doses higher than those required for tumor induction
(the increase in mutations likely results as a secondary effect of cytotoxicity or cell proliferation)

34. CP Dose / Resp Concordance
Rodents
Lowest effective dose [induction of SCE in rat bone marrow (0.62 mg/kg)]
Consistent with data showing significant tumor formation in the urinary bladder of male rats at 1.25 and 2.5 mg/kg/day (488 mg total)
Humans
Chromosome aberrations & SCEs 2 hrs after dosing mg/kg
p53 mutations at a cumulative dose of 6 g
Cohort of 6171survivors of non-Hodgkin's lymphoma; 48 developed cancer of the urinary tract – those receiving a total dose of 20g had a 2.4-fold  risk of bladder cancer; 20-49g, a 6-fold  risk

35. Temporality: Evaluate time-to-mutation
Mutagenic carcinogens would be expected to show a positive mutation response after relatively short treatment periods
Mutant Frequency
Time in Weeks
Nonmutagenic carcinogens would be expected to be negative after long chronic treatment, or show a positive response only after long chronic treatment

36. CP TEMPORAL ASSOCIATIONS
SCE bone marrow of Fischer 344 rats dosed with 20 mg/kg (ip) CP after 30 min. (1 hr after 5 or 10 mg/kg)
Chromosomal aberration & micronuclei in human bone marrow 24 hrs post therapeutic dose of 40 mg/kg (iv)
Cytotoxicity & regenerative proliferation in the rat also occur early:
Bladder damage (ulceration of mucosa, necrosis of bladder epithelium)—1 day
Regeneration of bladder epithelia – 36 hrs
Hyperplasia of bladder epithelia – 48 hrs
Malignant bladder tumors — weeks
(Dearfield et al., 1985) SCE increase beginning at 30 min. Evaluated up to 6 hours post treatment.
(Bochkov et al., 1986) Cab and MN

37. CP Database Plausibility & Coherence
Qualitative & quantitative data for key events leading to tumors
Concordance of most key events in animal models & humans
No stop/recovery studies found, but there is evidence suggesting that CP-associated cancers may occur up to several years after drug treatment has ceased.
Gaps in human data (e.g., DNA adducts & cell proliferation) do not compromise the analysis

38. MOA Relevance Rats Humans PAM generation Yes Yes
DNA adducts Yes Plausible
Mutagenicity Yes Yes
Bladder
cytotoxicity Yes Yes
Epithelial regeneration Yes Plausible
Hyperplasia Yes Yes
Bladder tumors Yes Yes
Therapeutic dose range from mg/kg, iv; 3X weekly
orally: 1-5 mg/kg/day; iv doses of mg/kg have been given over 4 days for bone-marrow transplants

39. Postulated Mode Of Action
Chloroform
CYP2E1
Metabolism
Oxidative
Phosgene
Sustained Toxicity
Lots of data for phosgene production, cell death and proliferation
In rodent bioassay studies, Chlorofom has been shown to result in liver and kidney tumors. The postulated MOA to explain these tumor responses involves bioactivation through CYP2E1 oxidative metabolism as the rate limiting step in chloroform’s mode of action. Metabolism by this pathway produces cytotoxic metabolite within the target organ, in particular phosgene that injures and kills cells, cytotoxicity is followed by regenerative cell proliferation, and it cytotoxicity/regenerative proliferation is sustained, eventually tumor development. So,these are the key sequence of events that will be considered with respect to chloroform;s included tumorigenesis in the rodent kidney and liver.
Regenerative Cell Proliferation
Key Events
Tumor Development

40. CCl3 Genetic Activity Profile
Here is a genetic profile that displays results of all the assays, each line is a study and it is presented by phylogenetic order from left to right bacteria to whole mammal assays. Top are positive and bottom are negative
Predominantly negative responses i n particular for assays that are most indicative of DNA reactivity and mutagenesis, gene mutation tests. Although a few positive reported, they must be interpreted carefully, some are inconclusive due to study design and conduct issues, or they may be in conflict with negative tests conducted in the same organism/assay, some are from assays that are not necessarily indicative of DNA reactivity and mutageneis, yeast recombination SCEs, and tend to yield a high frequency of false positives,

41. Mutagenicity: Lines of Evidence
Negative in vitro
Conflicting evidence in vivo
Initiation-Promotion Studies
CCl3 is not an initiator
Molecular Based Approaches
Negative for tumors in p53 +/- transgenic mouse cancer bioassay
Negative for mutations in LacI transgenic B6C3F1 mice
Negative for mutation in rat GST transfected bacteria
Additional key lines of evidence that reinforces the conclusions for ruling out mutagenicity as a mode of action :
Chloroform lack initiating activity in initiation-promotion tests
Using molecular based models
Negative for any tumors in the p53 mouse transgenic cancer assay, which most effectively responds to mutagenic carcinogens at doses (up to 240mg/kg) and conditions (co gavage) similar to those that tumors are found out in the traditional bioassay. (IS P53 DERVIVE FROM B6 STRAIN?)
And also negative in a mutation mouse transgenic model where one can measure mutation induction in the relevant target tissue.
Lastly adding support that other metabolic pathways do not lead to mutagenicity, was negative in a Ames assay where the bacteria were genetically engineered to have multiple copies of rat Glutathione s transferase and thus GST catalyze conjugation would be favored.

42. Mutagenicity CCl3: Conclusions
Weight of Evidence
Mutagenicity is not a component of chloroform induced neoplasia
Highlight the genetox and related metabolism data – suffice it say that there are sufficient data linking cell tox and regenerative hyperplasia to the kidney and liver tumors.
Given that cancer is a genetic disease, and thus there must be a genetic alteration in genes controlling cell proliferation or differentiation, it is important to determine in a mode of action analysis whether the agent can produce mutations through its direct DNA reactivity.
Chloroform has been subject to a large number of genetic toxicology assays and the WOE indicates mutagenicity is not….
Although reductive metabolism might lead to mutagenic metabolites, the reductive metabolism of chloroforms not major pathway.

43. Metabolism: Conclusions
Predominate pathway
P450 (CYP2E1)-mediated oxidative pathway
Phosgene key reactive metabolite
The following play little, if any role in chloroform induced tumors--
Reductive P450 metabolism & free radical production
GST catalyzed conjugation
Three possible metabolic pathways, two play little if any role
Predominant is oxidative via CY2E1
Phosgene is produced by this pathway as the only quantatitively significant metabolite, which would react with cellular macromolecule to injure or kill cells.

44. MOA Conclusions for Chloroform
Hypothesized MOA Well Supported
Other MOAs NOT Well Supported
Human Relevance Presumed (also epidemiological data on chlorinated water)
Applies to Children (but not more susceptible)
Consistent with Nonlinear Dose Response
Risk Approach Based on Protection Against Sustained Toxicity/Proliferation

45. Consider What data are available? What data are optimal?
Screening genetox data, batteries of test designed for hazard identification
What data are optimal?
Real, live MOA data (e.g. time course studies in relevant human genes)
What data are practical?
Something less than what was available for cyclophosphamide
Requires some strategic thought in test design.