Integrated Probabilistic and Deterministic Modeling Techniques in Estimating Exposure to Water-Borne Contaminants: Part 2: Pharmacokinetic Modeling

INTEGRATED PROBABILISTIC AND DETERMINISTIC MODELING
TECHNIQUES IN ESTIMATING EXPOSURE TO WATER-BORNE
CONTAMINANTS: PART 2: PHARMACOKINETIC MODELING
IN Blancato1*, FW Power2 , CR Wilkes3, AM Tsang2 , SC Hem1, SS Olin4
]US Environmental Protection Agency, Human Exposure and Atmospheric Sciences Division
2Anteon Corporation, Las Vegas, NV USA 89119
3Wilkes Technologies, Inc., Bethesda, MD USA 20814
4Intemational Life Sciences Institute, Risk Science Institute, Washington, DC USA 20036
ABSTRACT
The Total Exposure Model (TEM) uses deterministic and stochastic methods to estimate the
exposure of a person performing daily activities of eating, drinking, showering, and bathing
(see part 1), There were 250 time histories generated, by subject with activities, for the three
exposure routes, oral, dermal, and inhalation, and these were input to the physiologically
based pharmacokinetic (PBPK) model, via ERDEM (Exposure Related Dose Estimating
model). The chemicals modeled were trichloroethylene (TCE), trichloroacetic acid (TCA),
and dichloroacetic acid (DCA). Time histories of concentrations and Areas Under the Curve
(AUC) were determined for the liver, kidney, and venous blood. They were combined to
determine the distribution at each time step and hence define the 5th, 50th and the 95th
percentiles. The important pathways and the basis for their predominance are shown. Thus
highly variable exposures can be related to actual dose to various organs of the human body.
DISCLAIMER
The United States Environmental Protection Agency through its Office of Research and
Development collaborated in the research described here and the Office of Water partially
funded the work under Cooperative Agreement CX-822663-01. This work is being submitted
for peer review and has not yet been approved for final publication. Mention of trade names
or commercial products does not constitute endorsement or recommendation for use.
INDEX TERMS
Human Activities, VOCs and SVOCs, Exposure Assessment, Pharmacokinetic Modeling,
Modeling Indoor Pollutants
INTRODUCTION
Scientists have been working to develop techniques for analyzing and representing residential
water use and the subsequent exposure to waterborne contaminants. In this work the main
chemical concerns are water borne contaminants. The Total Exposure Model (TEM) was
used to estimate the exposure pattern of these contaminants. Then the Exposure Related
Dose Estimating Model (ERDEM) is run to determine the concentrations, area under the
curve (AUC), and amount in urine for the chemicals and their metabolites in the liver, kidney,
and venous blood. There were 250 exposure patterns determined from TEM for an adult male
and they have been run through ERDEM. Parameter values are taken from values suggested
by Abbas and Fisher, 1997, Fisher, et al, 1998, and Clewell, et al, 2000. Some values were
then adjusted using ERDEM in order to fit data for the first subject reported by Fisher, et al,
1998.
* Contact author email: Blancato.jerry@epa.gov

-------
WATER USE EXPOSURE MODEL
A water use exposure model, TEM (Total Exposure Model, discussed in Part 1) was
developed by Dr. Charles Wilkes. TEM models the exposure of subjects to chemicals in
water through ingestion, inhalation and dermal contact. Activities considered in this
modeling approach include showering, bathing, hand washing, toilet use, clothes washing,
dish washing, and direct and indirect consumption of water and beverages. The pattern of use
is determined by activity pattern studies such as NHAPS, (Kliepeis, et al 1996, Tsang, et al
1996).
EXPOSURE RELATED DOSE ESTIMATING MODEL
The ERDEM model is a Physiologically Based Pharmacokinetic (PBPK) Model (developed
using the ACSL1 engine) that takes input from up to eight exposure routes with multiple
chemicals in each scenario of an exposure route. Inputs may also be time histories of the
exposures. There may be multiple chemicals and multiple metabolites of each. All chemicals
and metabolites are treated as circulating. The static lung models the exchange between the
blood and air. A two compartment gastro-intestinal tract is modeled. Outputs can be time
histories of any of the variables (specified in a special history file). Plots or even ASCII time
histories can be generated for post processing and analysis. The outputs from the 250 model
runs for this study were generated as ASCII time histories and run through a special SAS
program for analysis and curve generation
PBPK MODEL PARAMETERS FOR TCE, TCA, AND DCA AND METABOLITES
The parameter values were chosen from the work of others but some elimination and
metabolism parameters were adjusted with ERDEM model runs using data reported by Fisher,
et al, 1998. The volumes and blood flows are given in Table 1 for the compartments used in
the simulation. The remaining 9% is the blood residing in the blood vessels (not shown in
Table 1). The alveolar ventilation rate was input from the TEM model as a time history of up
to seven activities and then converted to an approximate cardiac output by multiplying by the
factor 0.854 (based on the relative values used in earlier male trichloroethylene modeling,
Fisher, et al, 1998).
Tablel: Volume and Blood Flow
Compartment Name
Tissue Volume (% Body Wt)
Blood Flow( % Card.Output)
Dermis
9a
4.8d
Fat
l?b
4.8b
Liver
2.6b
24.0b
Static Lung
1.4b

Kidney
0.4^
19.7b
Rapidly Perfused Tissue
4.6b
27.5°
Slowly Perfused Tissue
56c
19.2b
a. McDougal, et al, 1990,value reduced to 9%.
b. Fisher, et al, 1998.
c. Value reduced to account for the Dermis.
d. Estimated from Corley, et al, 1990.
The partition coefficients used in the ERDEM model runs are shown in Table 2.
1 Advanced Continuous Simulation Language, owned by Aegis Technologies.

-------
Table 2: Partition Coefficients for TCE and It's Metabolites in the Human Male
Compartment to Blood or Air
TCEa
TCAa
TCOHa
DCA,
Mouse0
TCOG,
Mouse0
Arterial Blood to Air
11.15

Dermis to Venous Blood
1.38b

Fat to Venous Blood
52.34

Kidney to Venous Blood
1.08
0.66
2.15
0.8
1.4
Liver to Venous Blood
4.85
0.66
0.59
0.8
0.6
Rapidly Perfused to Venous Blood
4.85

Static Lung to Arterial Blood
0.39
0.47
0.66
0.16
1.1
Slowly Perfused to Venous Blood
1.38
0.52
0.91
0.43
1.1
a. Fisher, et al, 1998
b. Value chosen the same as the Slowly Perfused Tissue.
c. Abbas and Fisher, 1997
METABOLISM, ELIMINATION, GI, AND SKIN PERMEATION PARAMETERS
There are five metabolisms in the Liver that are modeled for TCE and TCOH (Table 3).
Elimination is modeled for TCA and DCA (Table 4). Urine flow is modeled for TCA, DCA,
and TCOG (also in Table 4). The skin permeation coefficient and the gastro-intestinal tract
parameters (stomach to portal blood, stomach to intestine, and intestine to portal blood) are all
given in Table 4. The results of ERDEM model runs for inhalation exposures with these
values provides a good fit with experimental values (A in the figures from data of Abbas and
Fisher) of concentration in the blood and urine measurements for TCE, TCOH, TCA, DCA,
and TCOG. Metabolism and urine parameters were fit for subject 1 (Abbas and Fisher) and
then scaled by body weight for the other subjects. Figures 1 and 2 show subject 2 results for
TCE and TCA.
Table 3: Parameters for Metaboli
tes of TCE
Parent Chemical
Metabolite
Saturable Metabolism
Linear Metabolism

Vmax,mg/h/kgBWa
Km,mg/La
Linear rate const.
TCE
TCA(O.l)
0.6
10.8

TCE
TCOH(0.9)
5.4
10.8

TCOH
TCA

7.0a
TCOH
DCA
O.lb
10b

TCOH
TCOG
30.0
160.0

a. Values c
etermined from fitting to experimental data from Fisher, et al, 1998.
b. Clewell, et al, 2000.
Table 4: Elimination rates, Urine Flow, Skin and GI Parameters
Chem
Liver Lin
Urine Rate
Skin Perm.
Stomach to
Stom-Portal
Intest-Portal

Elim Rate
Const(l/hr)
Coef.
Intest. Rate
Blood Rate
Blood Rate

Const(l/hr)

(cm/hr)
eonst.(l/h)e
const (l/hr)e
const (l/hr)e
TCE
N/A
N/A
0.0157d
2.18
13.65
0.044
TCA
0.2a
0.519b
3.58E-6d
2.18
13.65
0.044
DCA
7.0873°
0.00795°
1.84E-6d
2.18
13.65
0.044
TCOG
N/A
40.0a
N/A
N/A
N/A
N/A
a. Determined from fit to experimental data from Fisher, et al, 1998.
b. Estimated from the urine data for subject 1 of Fisher, et al, 1998
c. Clewell, et al, 2000

-------
d.
e.
IN Mcdougal, personal communication.
Abbas and Fisher, 1997 for the mouse, corn oil gavage. GI parameters modified
based on Staats, et al, 1990 for water..
10.0

AftV

¦o.l
0 24 48
TIME,HOURS
Figurel: Concentration of TCE in
Venous Blood for Subject 2 Versus
Measured Data
10.
0
c

1.
0
m

(

0.
1

k-=====r-

A
1

' "T -

0 48
TIME,HOURS
Figure 2; Concentration of TCA in Venous
Blood for Subject 2 Versus Measured Data
96
EXPOSURE INPUTS AND PARAMETER SETTINGS
There were two exposure scenarios, one with 0.1 mg of TCE in the water and the other with
the addition of 0.03 mg of TCA and DCA. The subjects were exposed to these chemicals by
inhalation of the volatilized TCE from the water, by ingestion of fluids, and by dermal
contact. The TCE in inhaled air was modeled by using the TEM model to determine 250
inhalation patterns over a 24-hour period. The 250 dermal and ingestion exposures were also
modeled by TEM for TCE, TCA and DCA. The concentration of the TCE given by TEM for
ingestion is reduced due to volatilization by 22% for drinks taken directly from the faucet
(direct consumption) and by 75% for drinks taken after some processing of water from the
faucet (indirect consumption). There is no loss due to volatilization for TCA and DCA.
RESULTS FROM ERDEM MODEL RUNS
The three routes, dermal, ingestion, and inhalation, were modeled separately and all together
for 0.1 milligrams/Liter of trichloroethylene (TCE) in the water. The 250 output time
histories from ERDEM were input to a SAS® program to determine the percentiles at each
time step. The 5th, 50th, and 95th percentiles were plotted at each time step. There were time
histories of concentration and area under the concentration-time curve (AUC) generated for
the kidney, liver, and venous blood for TCE and the metabolites, DCA, TCA, TCOH, and
glucuronidated TCOH - referred to as TCOG. Time histories were generated also for
exhaled air and for total amount in the urine for TCA, DCA, and TCOG. The chemicals TCA
and DCA at concentrations of 0.03 milligrams/Liter were added to the water and separate time
histories were generated from the TEM model for ingestion and dermal. There is no
inhalation exposure to TCA and DCA due to low volatilization. The presence of DCA due to
metabolism of TCE is minute and is only of consequence when it is already present in the
water. TCA is a substantial metabolite of TCE. Thus the presence of TCA in the water may

-------
not cause a significant increase in the AUC of TCA. The dominance of the inhalation route is
dependent on the volatility of the chemicals in the water.
E-2
A
U
C
E-4
mg-
h/L
E-6







""""""95th pe
xerHrT'le"
..		—-

s*'
¦^50th Per
cent-He

/ /
/ S-
-'"5th Per
:enti le

/ /
1 / /




0

!0
L
0
TIME,HOURS
Figure 3: AUC of TCA in Venous Blood for
Dermal Exposure
cerrtTIe
5th PgL \
10th Pe
Sth Perc
errtn le
0	20
TIME,HOURS
Figure 4: AUC of TCA in Venous Blood for
Ingestion Exposure
E-l
A
U
C
E-B
mg-
h/L
E-5







		9Tth f
-—loth p
er.csi*erW
ercerrtile

/ X

--'yth Pe
rcentile

/ /
// /




///




6
2
0
i
0
E-l
A
U
C
E-3
mg-
h/L
E-5







"9Sth P§p
¦"Tfoth Per
cuttle*

/ /

~"5th Perc
inti le

/ /
///




///




0
40
20
TIME,HOURS
Figure 6: AUC of TCA in Venous Blood for All
Exposures
TIME,HOURS
Figure 5: AUC of TCA in Venous Blood for
Inhalation Exposure
RANGE OF 5th AND 95th PERCENTILES
The range of the 5th and the 95th percentiles for the AUC of TCA in Venous Blood for
exposure to 0.1 mg/L of TCE are:
1.	Dermal - 0.0004 to 0.01 mg-h/L - a factor of 25, Figure 3.
2.	Ingestion - 0.0043 to 0.053 mg-h/L - a factor of 12.3, Figure 4.
3.	Inhalation - 0.014 to 0.038 mg-h/L - a factor of 27.1, Figure 5,
4.	All three routes - 0.04 to 0.5 mg-h/L - a factor of 12.5, Figure 6.
The inhalation and dermal routes are highly variable depending on the activities of the
subject, while the drinks of water taken via the ingestion route are more likely to occur many
times throughout the day. Thus the AUC for all three routes is less variable than the dermal
and inhalation routes due to less variability in the AUC for the ingestion route.
FACTORS AFFECTING THE DOSE FOR EACH PATHWAY
The absorption of chemical by the inhalation route is affected by the activity pattern of an
individual, volatility of the chemical(s) in the water, and the blood to air partition coefficient.
The rate that the chemical enters the blood is dependent on the lung to blood partition
coefficient. Similarly, the absorption of chemical by the ingestion route is affected by the
ingestion pattern of the individual as well as the gastro-intestinal tract parameters for that
chemical. The absorption of chemical by the dermal route in ERDEM is dependent on the
permeation coefficient for the chemical. The rate of release into the blood is dependent on the

-------
partition coefficient from skin to blood.
CONCLUSIONS
This work demonstrates the utility of coupling an exposure model with a pharmacokinetic
model to help determine the importance of different exposure factors and patterns on the
toxicologically relevant dose. Due to inherent non-linearities in the pharmacokinetics there is
often not a direct linear relationship between exposure and dose. Exposure models, such as
TEM, are able to give estimates of exposure that take into account the impact of human
activities and the natural trans-media relationships of chemicals such as TCE. Coupling these
results with a well-formulated pharmacokinetic model, as we have done here, enables the risk
assessor to begin to make the connections between source, exposure, and relevant dose.
REFERENCES
Abbas R, and Fisher JW. 1997. A physiologically based pharmacokinetic model for
trichloroethylene and its metabolites, chloral hydrate, trichloroacetate,
dichloroacetate, trichloroethanol, and trichloroethanol glucuronide in B6C3F1 mice.
Toxicology and Applied Pharmacology. Vol 147, pp 15-30
Clew ell HJ, Gentry PR, and Allen BC, Covington TR, et al. 2000. Development of a
physiologically based pharmacokinetic model of trichloroethylene and its
metabolites for use in risk assessment. Environmental Health Perspectives, Vol 108
(Supplement 2), pp 283-305.
Corley RA, Mendrala AL, Smith FA, et al. 1990. Development of a physiologically
based pharmacokinetic model for chloroform. Toxicology and Applied
Pharmacology, Vol 103, pp 512-527.
Fisher JW, Mahle D, and ABBAS R. 1998. A human physiologically based pharmacokinetic
model for trichloroethylene and its metabolites, trichloroacetic acid and free
trichloroethanol. Toxicology and Applied Pharmacology. Vol.152, pp 339-359.
Kliepeis NE, Nelson WC, Ott WR, et al 2001. The National Human Activity Pattern Survey
(NHAPS): a resource for assessing exposure to environmental pollutants. Journal of
Exposure Analysis and Environmental Epidemiology. Vol 11, pp 231 -252
McDougal JN, and Jepson GW. 1990. Dermal absorption of organic chemical vapors in rats
and humans. Fundamental and Applied Toxicology. Vol 14, pp 299-308.
McDougal JN, Jepson GW, Clewell III HJ, et al. 1986. A physiological pharmacokinetic
model for dermal absorption of vapors in the rat. Toxicology and Applied
Pharmacology. Vol 85, pp 286-294
Staats.DA, Fisher JW, and Connolly RB. 1990. Gastrointestinal absorption of xenobiotics in
physiologically based pharmacokinetic models, a two-compartment description. The
American Society for Pharmacology and Experimental Therapeutics. Vol 19, No. 1,
pp144-148
Tsang AM, Kliepeis NE 1996. Descriptive statistics tables from a detailed analysis of the
National Human Activity Pattern Survey (NHAPS) data. EPA 600-R-96-148,
Washington, D.C., U.S. Environmental Protection Agency.

-------
nerl-rtp—HEASD-02-069 TECHNICAL REPORT DATA
Date
1, Report No. EPA/600/A-02/076 2-
3.
4. Title and Subtitle
INTEGRATED PROBABILISTIC AND DETERMINISTIC MODELING
TECHNIQUES IN ESTIMATING EXPOSURE TO WATER-BORNE
CONTAMINANTS: PART 2: PHARMACOKINETIC MODELING
5. Report]
6. Performing Organization Code
NERL/HEASD/HERB/IO
7. Author(s) JN Blancato*. FW Power , CR Wilkes, AM Tsang , SC Hem, SS
Olin
8. Performing Organization
Report No.
9.Per forming Organization Name and Address
National Exposure Research Laboratory
Human Exposure & Atmospheric Sciences Division
Human Exposure Research Branch
Las Vegas, NV 702-798-2456 Blancato.Jerry@epa.gov
10. Program Element No.
11. Contract/Grant No.
12.Sponsoring Agency Name and Address
National Exposure Research Laboratory
Human Exposure & Atmospheric Sciences Division
Human Exposure Research Branch
Las Vegas, NV
13. Type of Report and Period
Covered Conference
Proceedings
14.Sponsoring Agency Code
15. Supplementary Notes
16. Abstract ABSTRACT
The Total Exposure Model (TEM) uses deterministic and stochastic methods to estimate the exposure of a person
performing daily activities of eating, drinking, showering, and bathing (see part 1). There were 250 time histories
generated, by subject with activities, for the three exposure routes, oral, dermal, and inhalation, and these were input to
the physiologically based pharmacokinetic (PBPK) model, via ERDEM (Exposure Related Dose Estimating model). The
chemicals modeled were trichloroethylene (TCE), trichloroacetic acid (TCA), and dichloroacetic acid (DCA). Time
histories of concentrations and Areas Under the Curve (AUG) were determined for the liver, kidney, and venous blood.
They were combined to determine the distribution at each time step and hence define the 5th, 50th and the 95th percentiles.
The important pathways and the basis for their predominance are shown. Thus highly variable exposures can be related to
actual dose to various organs of the human body.
17. KEY WORDS AND DOCUMENT ANALYSIS
A. Descriptors
B. Identifiers / Open Ended
Terms
C. COSATI

18. Distribution Statement
19. Security Class (This
Report)
21. No. of Pages
6
20. Security Class (This
Page)
22. Price
Form Available: Network Neighborhood\Rnight\Groups\HEASD\Forms\Tech-Form-2220-l
* Contact author email: Blancato.jerry@epa.gov

-------