Cumulative Risk Assessment of Drinking Water Disinfection ByProducts

1
Cumulative Risk Assessment of Drinking Water
Disinfection By-Products Linda K. Teuschler1,
Glenn E. Rice1, John C. Lipscomb1, Richard C.
Hertzberg1, Fred W. Power2, Jerry N. Blancato2,
Charles Wilkes3, Jane Ellen Simmons4 U.S.
Environmental Protection Agency, Office of
Research and Development, 1NCEA 2NERL, 4NHEERL,
3Wilkes Technologies, Inc.
Linking External Exposure and PBPK Models
Multiple Route Internal Dose Estimates
Strategy of The CRPF Approach
Results/Conclusions
Science Question
Step 1 Group Chemicals into Common Toxic Mode
of Action (MOA) Subclasses Step 2 Conduct
Dose Response Modeling of Toxicology Data for
Each Individual Chemical Estimate Index
Chemical Equivalent Doses (ICED) for DBPs Within
Each Subclass Develop Relative Potency Factors
(RPF) Within Each Subclass (e.g., using ratios of
Effective Doses) Step 3 Estimate Each
Subclass Risk Using RPF Method (assumes same
toxic MOA within each subclass) Step 4
Estimate Total Mixture Risk Using Response
Addition (assumes statistical and toxicological
independence between subclasses)
Various dose metrics can be used for DBP mixtures
depending on physicochemical characteristics and
available data. Internal dose estimates can be
made beginning with environmental concentrations,
and linking external exposure modeling outputs
with pharmacokinetic models.
What are peoples real world aggregate exposures
to drinking water disinfection by-product (DBP)
mixtures? How do we estimate cumulative risk
from multiple-route exposures to DBP mixtures?
Humans are exposed daily to hundreds of DBPs
via oral, dermal, and inhalation routes. Some
positive epidemiological studies suggest
reproductive and developmental effects and cancer
are associated with consumption of chlorinated
drinking water. However, human health risk
estimates made using animal data based only on
oral exposures do not reflect the same magnitude
of risks found in positive epidemiological
studies. Thus, it is hypothesized that this
difference may be accounted for by evaluating
simultaneous exposures to multiple DBPs via all
three exposure routes.
Realistic multiple route exposure estimates can
be developed for specific populations that
account for human activity patterns and water use
patterns, and physico-chemical properties of
DBPs. The scientific basis for evaluating DBP
mixture health risks is improved beyond methods
based on individual DBP concentrations and single
route risk analyses. The CRPF method provides a
way to assess mixtures whose components cause a
common health effect through different toxic
MOAs.
Dose-Response Modeling Using Single Chemical,
Oral Toxicology Data for Each Individual DBP
Assumes No Portal of Entry Effects
Impact and Outcomes

Human Exposures in the Home via Oral, Dermal
and Inhalation Routes

This research Provides data and tools supporting
the evaluation of alternative drinking water
treatment technologies developed to meet new
promulgated standards. Aids EPAs Office of Water
with interpreting epidemiological studies on DBP
health risks. Provides information and tools to
key interested parties (American Water Works
Association, the American Chemistry Council,
stakeholder groups) who implement programs to
protect human health. Provides methods and data
that enrich the available library of cumulative
risk assessment methods beyond what is currently
published in EPA guidance
Build Human Dose Response Model
Adjust to Internal Animal Dose (PBPK Modeling
or Administered Dose Times Bioavailable)
Research Goals
Estimate ED10s, RPFs
The Safe Drinking Water Act Amendments of 1996
require EPA to begin evaluating mixtures of
contaminants in drinking water. The goal of this
research was to determine the feasibility of
conducting a Cumulative Risk Assessment (CRA) for
DBP mixtures. Three significant findings
resulted. 1) External exposure modeling was
conducted and linked with Physiologically Based
Pharmacokinetic (PBPK) modeling to produce
internal dose measures from dermal, oral and
inhalation exposures for use in risk assessment.
This new approach provided realistic human
exposure distributions for integrating health
risks for multi-route DBP exposures. 2) These
internal doses were estimated for an adult female
and an adult male, each of reproductive age, and
for a child (age 6), accounting for human
activity patterns that affect contact time with
drinking water. Exposure distributions were
estimated for 13 major DBPs, including most of
the DBPs currently regulated by EPA. 3) A new
risk assessment method, the Cumulative Relative
Potency Factors (CRPF) approach, integrates the
principles of dose addition and response
addition, additivity concepts presented in Agency
guidance, to produce multiple-route, chemical
mixture risk estimates using internal doses. The
CRPF approach evaluates human health risks using
total internal doses and single chemical, oral
toxicology dose-response data based on knowledge
or assumptions regarding toxic modes of action.
The CRPF approach is a new method that can be
generalized for the evaluation of other
environmental mixtures whose components act by
more than one toxic mode of action.
Response
In this research, the Total Exposure Model (TEM)
provided 24 hour time history simulations using
data on chemical properties, human activity
patterns, human drinking water intake, and
building characteristics. These time histories
were joined with pharmacokinetic data and became
input to the PBPK model, ERDEM (Exposure Related
Dose Estimating Model). Multi-route exposure,
internal doses were then estimated.
Internal HED
Future Directions

• Physico-Chemical Properties, e.g.,
• - Mass transfer coefficients
• - Henrys Law constants
• - Octanol/water partition coefficients
• - Chemical concentrations
• Human Behavioral Parameters, e.g.,
• - Frequencies/duration of
• showering, bathing, faucet use
• - Appliance factors, flow rates,
• water temperatures, cycles
• Human Intake Characteristics, e.g.,
• - Tap water consumption
• - Skin permeability coefficients
• Breathing rates
• Building Characteristics, e.g.,
• - Household air volume
• - Whole house air exchange rate

Example of ERDEM 48 Hour Absorbed
Doses Bromodichloromethane for Adult Female
This research establishes the feasibility of
analyzing multi-route exposures to DBP mixtures
using a CRA approach. Important data gaps
include Careful treatment of MOA and common MOA
subclasses for the major DBPs for developmental
and reproductive effects and cancer. Development
of physico-chemical properties and external
exposure factors for several major DBPs. PBPK
modeling of the pregnant uterus and DBP doses to
the fetus. Improved uncertainty and sensitivity
analyses.
Note EDx is the Effective Dose causing a
response in x of the population of interest.
Schematic of CRPF to Illustrate DBP Mixture
Cancer Risk
TEM
Estimate Subclass Risks
Group DBPs into Common MOA Subclasses
RPFs TADs (Total Absorbed Doses)
Sum to get Subclass ICEDs
Sum to get Mixture Risk
EDx(BDCM) DBCM EDx(DBCM) TAD
DBCM ICED CHBr3 ICED BDCM Dose ICED of
BDCM (mg/Kg day)
Genotoxic DBPs BDCM (Index Chemical), DBCM,
CHBr3
References
EDx(BDCM) CHBr3 EDx(CHBr3) TAD
Teuschler, L.K., G.E. Rice, C.R. Wilkes, J.C.
Lipscomb, F.W. Power. 2004. A Feasibility Study
of Cumulative Risk Assessment Methods for
Drinking Water Disinfection By-Product Mixtures.
Journal of Toxicology and Environmental Health
Part A, 2004, 67755-777. U.S. EPA. 2003. The
Feasibility of Performing Cumulative Risk
Assessments for Mixtures of Disinfection
By-Products in Drinking Water. EPA/600/R-03/051.
ORD/NCEA Cincinnati, OH.
Non-Genotoxic DBPs DCA (Index Chemical) TCA
TCA ICED DCA Dose ICED of DCA (mg/Kg day)
EDx(DCA) TCA EDx(TCA) TAD
Note AUC Area Under the Curve