Mode of Action and Dosimetry Considerations in Interspecies Extrapolation

1
Mode of Action and Dosimetry Considerations in
Interspecies Extrapolation

• Melvin E. Andersen, Ph.D.
• CIIT Centers for Health Research
• Board of Scientific Counselors (BOSC)
• Risk Assessment Workshop
• February 2-3, 2005
• Washington, DC

2

• Outline
• Interspecies extrapolation defaults
• Dosimetry models in risk assessment and what
  weve learned.
• Mode of Action models and dose-response where
  we are going
• Some Recommendations, especially about the idea
  of defining defaults

3

• Defaults
• The process we follow in the absence of
  information
• Change default process as information becomes
  available
• Its a bit backwards.
• What kind of information do we want to have to do
  a risk assessment and what do we do when it is
  lacking?

4

• Default Interspecies Scaling (Model independent)
• FDA 1950s Two safety factors, 10-each rules
  of fingers animal-to-human, among human
  populations
• Safety factors
• Cancer Risk Assessment (1970s) linear risk
  models
• Surface area adjustment (bw1/bw2)0.33
• Reference Dose (1980s) non-linear risk models
• Uncertainty factors
• 10-interspecies
• 10-intraindividual
• others data base, NOAEL/LOAEL/Duration
• Proliferation of independent, uncertainty
  factors

5

• What are we accounting for?
• Dosimetry for a given exposure in test animals
  and humans how do tissue doses vary between
  species
• Tissue Response for a specified tissue dose,
  how will the response vary between test animals
  and humans.

• In the context of some presumption about the
  dose-response relationship?

6
What tools help us evaluate these
relationships? Pharmacokinetics calculate the
tissue dose of active forms of the toxic chemical
for various doses, dose-routes, and animal
species Pharmacodynamics calculate the degree
of response for any level of tissue dose in
different species
7
Physiologically Based Pharmacokinetic (PBPK)
Modeling
You can be wrong!
Air
Metabolic Constants Tissue Solubility Tissue
Volumes Blood and Air Flows Experimental System
Lung
Body
Tissue Concentration
X
Fat
X
X
X
X
X
X
Liver
X
Model Equations
Time
Define Realistic Model
Make Predictions
Collect Needed Data
Refine Model Structure
8
Using PBPK models - 1985

• Identify toxic effects in animals and people
• Evaluate available data on mode(s) of action,
  metabolism, chemistry of compound, metabolites
  and related chemicals
• Describe potential mode(s) of action
• Propose relation between response and tissue dose
• Develop a PBPK model to calculate tissue doses
• Estimate tissue dose during toxic exposures with
  model
• Estimate risk in humans assuming similar tissue
  response for equivalent target tissue dose

9
Extrapolations supported by these models

• High doses to low doses
• Dose route inhalation, oral, dermal
• Between species
• Across classes of chemicals
• in vitro to in vivo
• Dosing scenarios

10
What weve learned with these dosimetry models

• The process how to do it.
• Broad utility, i.e., 1000 papers by 2002
• Mode of action specific extrapolation defaults
• Analysis tools assess variability, evaluate
  model sensitivity, improve experimental design
• Assumptions in process made more explicit
• Identify, quantify and reduce uncertainty

11
RfD and RfC Calculations Distinguish PK and PD
components of Interspecies Uncertainty
Factor. Default the PK and PD portions are
regarded as having equal value (10)1/2 May be
changed with information about a specific
compound, leading to data-derived uncertainty
factors (Renwick and colleagues)
12
Dose-Response Models

• Linear models some risk probability at all
  doses
• Threshold models use independent, uncertainty
  factors. Why??

13
BBDR models for tissue level responses, MVK
cancer model and carcinogenesis
14
Exposures produce toxicity and DNA-protein
cross-links. With toxicity, there is increased
cell proliferation, tumor promotion and cancer.
15
Cellular responses
Tissue Phase Reactions Cl2 HOCl
HCl CH2O HCOOH
Ventilation
Dosimetry Inhaled Stressors
Adaptive State
Stressed State
Molecular Biochemical Histologic Physiological
16

• In general,
• Lack organized information about basic responses
  of cell signaling pathways to toxic compounds
• High coverage/signal transduction studies, in a
  functional genomics context, are changing the
  situation

17
Common Signaling Themes. Science Nov. 2004.
18
MAP-Kinase Families
And others. (Johnson and Lapadat, Science,
2002).
19
MAPKinase Cascades
Input
phospahatases
MAPKKK MAPKKK-PO4 MAPKKK
PO4
MKKK
MAPKK
MAPKK PO4
MAPKK-(PO4)2
MKK
MAPK MAPK-(PO4)2 MAPK
PO4
MAPK
Target Protein Transcriptional Factor
Function
20
Platelet Derived Growth Factor (PDGF)

• Positive feedback loop through cPLA2-AA-PKC
• Negative feedback loop through MKP and PP2A

21
The fading paradigm
UNCERTAINTY
Exposure
Tissue Dose
Biologically Effective Dose
Early Responses
Late Responses
Pathology
Physiologically Based Pharmacokinetic Models
Tissue Dose Metric
Mode of Action
22
Perturbations of biological function
Exposure Tissue Dose Biological
Interaction(s) Perturbation
Affected Biological Function
Systems Inputs
Molecular or Pathway Target
23
(No Transcript)
24
Dosimetry Recommendations

• Use mode of action based defaults
• Develop parameter data bases for human PBPK
  models
• Encourage dosimetry based approaches for
  cumulative aggregate risk assessments
• Expand suite of validated human PBPK models
• Improve understanding of parameters important for
  understanding PK of lipophilic compounds

25
More Fundamental Recommendations

• Develop clear articulation of the underlying
  models linear and non-linear - used for risk
  assessment and the implicit structure/assumptions
• Explain how data when available are to be used in
  these assessments
• Discuss rationale for defaults used when data are
  unavailable
• Develop mechanistic dose-response models for
  obligatory precursor cellular responses to toxic
  stressors, e.g., chloroform or chlorine, etc., or
  to altered signaling pathways from EACs

26
What do we want to do with dose response models

• The process what do we think we are doing and
  how should we do it.
• Find out if independent use of uncertainty
  factors is actually a rational approach
• Analysis tools assess variability, evaluate
  sensitivity, improve experimental design
• Assumptions in extrapolation process made more
  explicit
• Identify, quantify and reduce uncertainty