Presented to the BOSC Risk Assessment Workshop

1
EPA Risk Assessment Practice Extrapolation
Between Species
Kerry L. Dearfield, Ph.D. Senior Scientist for
Science Policy Office of the Science
Advisor United States Environmental Protection
Agency
Presented to the BOSC Risk Assessment
Workshop February 2-3, 2005
2
Staff Paper

• EPA Publication EPA/100/B-04/001
• Found on Office of the Science Advisor web site
  www.epa.gov/osa
• Published March 2004

3
Document as Staff Paper

• Document is an EPA Staff Paper
• Presents the perspectives of EPA risk assessors
  on how they understand risk assessment is
  conducted at the Agency
• Also presents staff recommendations for EPA and
  interested stakeholders to consider for how EPA
  can move forward to strengthen and improve its
  risk assessment practices
• Staff paper as it stands does not represent EPA
  policy

4
Use Data Before Invoke Defaults
Analyze the available data

Is there too much uncertainty or is critical
information lacking?
Invoke a default option
Y
N
The primary goal of EPA actions is public
health protection, accordingly, as an agency
policy, the defaults used in the absence of
scientific data to the contrary should be health
protective (SAB 1999).
5
Outline of Presentation

• Relevance of animal data to humans and issues
• Interspecies extrapolation and issues
• Target organ concordance
• Route to route extrapolation

6
Relevance of Animal Data to Humans

• Positive adverse effects in animal studies
  indicate the agent under study can have
  toxicological potential in humans unless there
  are data to indicate otherwise
• Based on assumption there are important
  similarities between animal models and humans
  (e.g., most known human carcinogens are positive
  in animal models)
• Prefer information from animals that are as
  similar to people as possible

7
Relevance of Animal Data to Humans

• The most sensitive responding species (given
  several data sets to choose from) are selected
• Mode of action information is becoming more
  useful to help determine relevance

8
Relevance of Animal Data to Humans Issues

• Contrast adverse with beneficial or adaptive
  changes that may be observed also, if and when
  does adaptive become adverse
• Severity of effect how relate qualitatively and
  quantitatively to the toxicity
• Reversibility of effect
• Paucity of comparative data on metabolism for
  specific chemicals and other interspecies
  differences that can affect toxicity

9
Interspecies Extrapolation

• Use of PBPK models can enhance the calculation of
  internal dose for systemic toxicants and help
  refine the interspecies extrapolation
• For RfD derivation, the UF of 10 is divided into
  PK and PD components
• For RfC derivation, PK methods help derive the PK
  component of the UF the PD component is a
  default of 3 (unless data are available)
• Most physiological endpoints scale by body weight
  to the ¾ power

10
Interspecies Extrapolation Issues

• Do all PBPK parameters scale to the same ¾ power?
• Uncertainty around the extent of interspecies
  variability
• Species specific sensitivity
• Does the most sensitive animal estimate an
  average human, or most sensitive human?

11
Target Organ Concordance

• No evidence that a mechanism(s) in one specie is
  necessarily target organ concordant in another
• Site concordance is not assumed a priori
• However, if mode of action is established, there
  is an expectation for site concordance when
  making the mode of action case this is a case
  by case circumstance

12
Route to Route Extrapolation

• If an agent causes an internal effect by one
  route of exposure, it will cause the effect by a
  different route if it is absorbed by the other
  route to give an internal dose unless data are
  available to indicate otherwise
• Assumes that the internal dose to the tissue of
  interest is the ultimate determinant of toxicity

13
3 Examples

• Alpha-2u-globulin specific to male rat kidney
  tumors
• Thyroid tumors animal model demonstrates mode of
  action to indicate non-linear extrapolation for
  humans
• Concordance of endpoint may not be the best
  predictor of developmental effects in people

14
The End

• Thank you very much

15
Extra Slides

• Extra slides

16
Risk Assessment at EPA

• EPA conducts risk assessment in order to provide
  the best possible scientific characterization of
  the risk in question, based on a scientifically
  sound, rigorous analysis of available information
  and knowledge.
• Risk assessment informs decision makers about the
  science implications of the risk in question.

17
EPA Risk Assessment Approach

• Confidence in our risk assessments is critical.
• Approach is to use to fullest extent site- and
  chemical-specific data relevant to the decision
  needed.
• Without such information, we use defaults to
  ensure we cover the uncertainty of the remaining
  data or lack of data.

18
EPA Risk Assessment Approach (cont.)

• The data and information we use in developing
  risk assessments has inherent uncertainty and
  variability.
• Due to the general uncertainty and variability of
  the data, information, and methodologies EPA
  assesses, we tend to take a more health and
  environmentally protective stance to ensure we do
  not underestimate risk.

19
General Nature of Comments

• EPA is being criticized for its risk assessment
  practices
• Generally, the nature of these criticisms are
• EPA must not intermingle policy judgments within
  the scientific assessment of risk
• Risk assessments should not rely on conservative
  (worst case) assumptions that distort outcomes
  and yield estimates that grossly overstate risk
• Risk assessments should acknowledge the presence
  of considerable uncertainty