BIO AVAILABILITY OF ARSENIC AND LEAD IN
ENVIRONMENTAL SUBSTRATES
Roseanne M. Lorenzana
Bruce Duncan
U.S.EPA Region 10
Risk Evaluation Branch
Seattle, WA 98101
Mike Ketterer
Joe Lowry
John Simon
U.S.EPA National Enforcement Investigation Center
Denver, CO
Michael Dawson
University of Technology
Sydney, Australia
Robert Poppenga
Michigan State University
East Lansing, MI
February 1996
EPA910/R-96-002
-------
ACKNOWLEDGMENTS
This report has been compiled by Dr. Roseanne M. Lorenzana (U.S.EPA Region 10, Office of
Environmental Assessment, Seattle, Washington) and is based on technical work performed by the
following people:
Roseanne M. Lorenzana, DVM, PhD
P. Bruce Duncan, PhD
John Yearsley, PhD
Don Matheny
U.S.EPA Region 10, Office of Environmental Assessment, Seattle, Washington
Joe Lowry, PhD
Mike Ketterer, PhD
John Simon
U.S.EPA National Enforcement Investigation Center, Denver, Colorado
Michael Dawson, PhD
University of Technology, Sydney, Australia
Robert Poppenga, DVM, PhD
Brad Thacker, DVM, PhD
Michigan State University, East Lansing, Michigan
David Maughan, Science Applications International Corporation
ESA Laboratory, Inc., Bedford, Massachusetts
The early phases of planning and study design were conducted by Drs. Raleigh Farlow, Dana
Davoli and Michael Watson (U.S.EPA Region 10).
Review comments were provided by Drs. Ron Landy (U.S.EPA/ORD/ORSI), David J. Thomas
(U.S.EPA/ORD/FffiRL), Michael Watson (U.S.EPA Region 10) and David Frank (U.S.EPA
Region 10).
Special appreciation is made to the U.S.EPA Region 10 hazardous waste site managers, Mary
Kay Voytilla and Chris Field, for authorizing the financial and technical support for this study.
-------
EXECUTIVE SUMMARY
A study using immature swine as test animals was performed to determine if arsenic and lead were
absorbed from the gastrointestinal tract into the bloodstream following oral dosing of soil or slag
from the Ruston/North Tacoma Superfund site located in Tacoma, Washington, or following oral
dosing of tailings or dust from the Triumph Mine Tailings site located in Triumph, Idaho.
A data evaluation methodology was developed to estimate the extent of arsenic absorption using
the results from multiple dose groups. The methodology provided an estimate of the 95%
confidence limits for the calculated mean absolute and relative bioavailiability of arsenic in the soil
and slag samples from the Ruston/North Tacoma Superfund site. Relative bioavailability was
calculated using the oral control group data and absolute bioavailability was calculated using the
intravenous control group data. These values are shown below:
Bioavailability (arsenic)
Relative (mean)
95% confidence limit
Absolute (mean)
95% confidence limit
Test Material
Soil
78%
56-111%
52%
44-61%
Slag
42%
27-63%
28%
20-37%
Because the toxicity criteria for arsenic were developed from oral ingestion studies, the relative
bioavailability estimates would be appropriate for use, if desired, in adjusting arsenic exposure
estimates.
Significant increases in blood lead concentrations were observed following oral dosing of the soil
and slag samples from the Ruston/North Tacoma Superfund site, and following oral dosing of the
tailings or dust from the Triumph Mine Tailings site. However, this experiment did not provide
reliable bioavailability estimates for lead from any of these test materials.
-------
TABLE OF CONTENTS
ACKNOWLEDGMENTS i
EXECUTIVE SUMMARY ii
INTRODUCTION 1
Background 1
Objectives 2
MATERIALS 3
Environmental Test Materials 3
Technical Grade Test Materials 3
Test Facilities and Animals 4
Analytical Reagents and Standards 4
METHODS 6
Study Protocol 6
Preliminary Study 6
Final Study 6
Dose Preparation 8
Biological Sample Collection and Handling Protocol 10
Quality Assurance 10
Characterization of Arsenic and Lead 11
Environmental Samples 11
Arsenic and Lead Aqueous Dosing Solutions 12
Water and Animal Feed 12
Biological Samples 13
Data Evaluation Methodology 15
RESULTS 17
Range Finding Experiment 17
Quality Assurance Review of Final Study 17
Grain and Water Analyses 17
Dose Analyses 17
Clinical Observations 21
Urine Samples 21
Arsenic Concentrations in Blood 21
Arsenic and Lead Concentrations in Urine 22
Lead Concentrations in Blood 27
Percent Recovery of Orally Administered Arsenic and Lead 32
Estimates of Bioavailability 32
DISCUSSION 42
REFERENCES 46
in
-------
INTRODUCTION
Background
This study was initiated in late 1990 by U.S.EPA Region 10 to provide site-specific empirical data
to improve the certainty of exposure estimates for the assessment of risks due to arsenic and lead
contaminated residential yard and driveway materials (soil and slag) at a hazardous waste site in
Tacoma, WA.
A smelter operated at the site from 1890 to 1986, first as a lead smelter, and after 1912 as a copper
smelter that specialized in the processing of ores with high arsenic concentrations. Air emissions
resulted in contaminated soils in adjacent residential areas, and smelter slag had been used for
landscaping and top grade on gravel driveways and roads. The Ruston-Vashon Island Arsenic
Exposure Pathways Study indicated an association between human residents' proximity to the
hazardous waste site and urinary arsenic (Polissar, 1987; Kalman et al., 1990; Polissar et al., 1990).
As a result of a screening evaluation, two smelter-related contaminants were identified for detailed
evaluation in the risk assessment: arsenic and lead (U.S.EPA, 1992). In early 1992 when site cleanup
options were developed, there were few literature reports of arsenic and lead bioavailability from solid
matrices such as food or soil. Studies in which animals were orally dosed with arsenic contaminated
soils were difficult to interpret due to inter-animal or inter-group variation, or questionable
representativeness for soils at the Tacoma study area (Freeman et al., 1993; Griffin & Turck, 1991;
Boyajian, 1987). However, information from these studies supported a reasonable assumption that
bioavailability of arsenic in soils was reduced as compared to drinking water. Based on best
professional judgement and public health protectiveness, a relative bioavailability factor of 0.8 was
assumed for soil and 0.4 for slag. Due to the widespread nature of contamination in the surrounding
residential area and the potential impacts of cleanup activities on the community, confirmation of the
literature based bioavailability estimates was needed prior to finalizing cleanup decisions.
Both arsenic and lead were present in elevated concentrations in the environmental matrices (soil and
slag). Because these contaminants could not be separated, the uptake of both was evaluated. No
other reports were identified in which the bioavailability of co-contaminants were simultaneously
studied. Since this approach was developed in late 1990, other investigators have continued
development of the immature swine model and further demonstrated its value in bioavailability studies
(LaVelle et al., 1991; Weis et al., 1994; DuPont, 1993).
Chemical and physical characteristics of the soil and slag as well as the chemical form of the
contaminants can have significant effects on the bioavailability of arsenic and lead (Chaney et al.,
1989). Whereas bulk analyses provide total concentrations of arsenic or lead, physical/chemical
characterization provides information about the three dimensional arrangement of elements in the soil,
slag, tailing or dust matrices. Following ingestion, this composition influences the ability of arsenic
and lead to move from the matrix to gastrointestinal fluids and then to body tissues.
EPA 910/R-96-002 Page 1 of 48
-------
Objectives
Objectives of the study were to identify physical and/or chemical indicators of the environmental
matrices' (soil, slag, tailings) potential to release biologically available forms of arsenic and lead, and
to obtain tissue data indicating whether arsenic and lead are absorbed into the body following oral
exposure.
The first objective was to examine physical and chemical characteristics and identify which of these
may be important determinants of the materials' potential to release arsenic and lead when orally
ingested by human beings. The results of these physical and chemical analyses will be included in a
subsequent report.
The second objective was to utilize the immature swine model to examine urine and blood
concentrations of arsenic and lead, respectively, as evidence of their gastrointestinal absorption, and,
if possible, to determine estimates of the extent of absorption. Samples of soil or slag from highly
contaminated lots in the vicinity of the smelter in Tacoma were evaluated in replicate at multiple dose
levels. Single samples of soil or tailings from a former mining site in Idaho provided preliminary
information useful for designing a bioavailability study, if needed, for that site. No smelter was in the
area of the former mining site. The results of the animal studies are described in this report.
The physical and chemical characteristics of the smelter site and mining site samples will be provided
in a subsequent report.
EPA 910/R-96-002 Page 2 of 48
-------
MATERIALS
Environmental Test Materials
A composite of soils was collected in the residential area surrounding the former Asarco smelter in
Tacoma, Washington. Decontaminated stainless steel hand auger, mixing bowl, spoon and sample
containers were used to collect fifteen surface (zero - 3 inch depth) soil samples.
Sampling and decontamination procedures were carried out according to a standard operating
protocol developed to ensure the use of safe methods as well as careful technique as part of the
quality assurance plan. Briefly, equipment was decontaminated by washing with detergent and water,
rinsing with distilled and reagent-grade water, rinsed with pesticide-grade methanol and allowed to
air dry.
Samples were collected from two residential properties (vacant lots) within one to two blocks of the
smelter stack where previous investigation indicated elevated arsenic concentrations and access was
granted. Five random samples were collected from one site and ten from the other.
A composite of slag was collected in the residential area surrounding the former Asarco smelter in
Tacoma, Washington. Decontaminated equipment was used to collect five surface (zero - 3 inch
depth) samples from residential driveways known to contain smelter slag and where access was
granted.
A composite of soil was collected from a residential location in Triumph, Idaho which provided
access to EPA. Decontaminated equipment was used to collect twelve randomly chosen surface
(zero to 3 inch depth) soil samples from a residence within 200 feet adjacent to and downwind of the
mine tailings. The individual samples were thoroughly mixed together then placed in a sample
container.
A sample of surface (zero to 3 inch depth) tailings from the upper eastward tailings pile in Triumph,
Idaho was collected using decontaminated equipment.
A sample of subsurface (ten - eleven feet below surface) tailings from the same borehole as the
surface tailings piles in Triumph, Idaho was collected using decontaminated equipment.
A composite sample of vacuum cleaner dust was collected from residences in Triumph, Idaho.
Decontaminated equipment was used to collect and mix together three individual dust samples
obtained from vacuum cleaner bags at three separate residences. Chosen residences were those who
agreed to EPA's access. Vacuum dust contents were a result of residents' routine cleaning practices.
This composite was included in the physical-chemical analyses but was not used in the animal dosing.
Technical Grade Test Materials
Sodium arsenate heptahydrate (Na2HAsO4-7H2O, Sigma Chemical No. A756). Lead acetate
trihydrate (Pb(C2H3O2)2-3H2O, Aldrich Chemical No. 31651-2). Radiogenic lead (92% 206Pb,
National Institute of Standards and Technology no. 983).
Standard reference materials used as quality control specimens were obtained from the Center for
Disease Control and Kaulson Laboratory (blood lead and blood arsenic, Kaulson #0141), from the
EPA910/R-96-002 Page 3 of 48
-------
National Institute of Standards and Technology and the Centre de Toxicologie du Quebec (urinary
arsenic NIST2670, Centre#S-9206, #S-200, #S-189, urinary lead, NIST2670 and vegetable matter
arsenic and lead, NIST1547).
Test Facilities and Animals
The swine study was conducted from April through September 1992 at the Michigan State University
Pesticide Research Center, East Lansing, MI under the direction of Drs. Robert Poppenga and Brad
Thacker. Cages, containment areas and study equipment were isolated from other equipment and
animals both prior and during the course of this study.
Thirty-nine female crossbred swine approximately 15 kilograms (approximately 40-50 days old) were
obtained from a commercial producer. The sires were a Hampshire hybrid and the dams were
crossbred Landrace/Large White/Duroc. Following preconditioning, animals were acclimated for at
least seven days prior to the study. Swine were randomly assigned to experimental groups with
stratification for litter of origin (to control for age, genetic background and environmental factors
prior to weaning) and body weight.
Animals, individually identified by plastic ear tags, were fed a standard swine corn and soybean ration
diet equal to 2% of body weight twice a day for the duration of the study except for fasting prior to
dosing. This quantity was sufficient for normal growth of swine this age and size. Water was
provided ad libitum in stainless containers. Multi-element analyses were conducted on both feed and
water samples, and analyzed using plasma mass spectrometry as described below. The animals'
drinking water and water samples for analyses were obtained following flushing the system for several
minutes.
Immature swine were preferred as the test animal for this study because of characteristics comparable
to young children (the age group at greatest risk of ingesting soil or other material containing
contaminants). These included similar body size, weight, bone-to-body weight ratio and
gastrointestinal anatomy and physiology. In addition, unlike other species such as rats or rabbits, the
rate of growth and maturation is slower (a smaller portion of the prepubertal period will occur during
the experiment), the cecum (a diverticulum of the large intestine where prolonged exposure to
digestive enzymes and fluids occurs) is small, and coprophagia (reingestion of feces) is not required
to maintain normal nutritional status. Like humans, swine are monogastric omnivores (stomach and
intestinal fluid and bacterial composition are different than herbivores or carnivores), are adaptable
to a periodic feeding schedule and have a gall bladder which excretes bile into the small intestine when
food is present (some contaminants, such as lead are excreted in bile). Unlike the rat, metabolism and
excretion of arsenic in swine is similar to humans. The results of pharmacokinetic studies of lead in
immature swine and humans are similar (Weis et al., 1994).
Analytical Reagents and Standards
Trace metal grade concentrated nitric acid and hydrochloric acid (Baker Instra-Analyzed) were used
without further purification. The sodium borohydride was 98% pure (J.T. Baker Inc.) All other
common chemicals were of reagent grade purity. Fresh distilled, deionized water was used as the
solvent for all solutions.
Element stock solutions were obtained from either PlasmaChem Associates or Spex Industries.
Calibration working standards were prepared daily from these standards and were matrix matched
EPA910/R-96-002 Page 4 of 48
-------
to the digestion or fusion solutions. Calibration verification standards were prepared from a second
source (Environmental Protection Agency, Environmental Resources Associates, PlasmaChem or
Spex Industries).
A stock solution of lead (1000 mg/L) was prepared by dissolving a portion of National Institute of
Standards and Technology (NIST) Standard Reference Material (SRM) 981 Common Lead Isotopic
Standard wire in 1 N nitric acid. This stock was used to prepare the primary calibration working
standards in a 10% v/v nitric acid aqueous matrix for lead isotope measurements. Calibration
verification standards (0.002 mg/L) were prepared from a PlasmaChem Associates stock solution (10
mg/L).
NIST SRM 2670 Toxic Metals in Freeze-Dried Urine which is provided at normal and elevated levels
was reconstituted before use. The other reference materials were used as purchased including NIST
SRM 955 Lead in Bovine Serum; NIST SRM 1645, River Sediment from Indiana Harbor near Gary,
Indiana; NIST SRM 2704, River Sediment from Buffalo, New York; PACS-1 a marine sediment
obtained from the National Research Council of Canada, Marine Analytical Chemistry Standards
Program, Ottawa, Ontario, Canada; and United States Geological Survey Geochemical Exploration
Reference Samples GXR-1 (Jasperiod, Drum Mountains, Utah), GXR-2 (Soil, Park City, Utah),
GXR-3 (Hot Springs Deposit, Humboldt County, Nevada), and GXR-4 (Porphyry Copper Mill
Heads, Utah).
EPA 910/R-96-002 Page 5 of 48
-------
METHODS
Study Protocol
Preliminary Study. A preliminary range finding experiment utilizing three swine was performed to
assess the proposed animal handling methodology, the analytical methodology and to determine
appropriate doses.
Dose
Animal No. kg mg soil/kg mg Pb/kg mg As/kg ug 206Pb/kg
34 8.2 18.3 0.025 0.029 4.46
35 10 20.0 0.027 0.032 4.46
36 8.6 41.9 0.056 0.067 5.65
Final Study Protocol. The final study included thirty-six swine in treatment groups consisting of
positive control groups, a negative control group and groups receiving environmental media. The
positive control group (8 animals) received a single intravenous or gavage administration of sodium
arsenate and lead acetate. The negative control group (4 animals) received only the aqueous vehicle
with no arsenic or lead included. Data for baseline blood arsenic from two animals in the positive
control group that received intravenous lead acetate but not sodium arsenate were included in the
negative control group. Therefore, there were four animals in the negative or untreated group for
lead and six animals in the untreated group for arsenic. The group receiving environmental media (21
animals) received a single oral administration of one of four quantities of soil at 25, 60, 100 or 150
milligram (mg) soil per kilogram (kg) of body weight (BW) (0.04, 0.10, 0.16 or 0.24 mg As per kg
BW and 0.03, 0.08, 0.14 or 0.20 mg Pb per kg BW), or a single oral administration of one of three
quantities of slag at 60, 100 or 150 mg slag per kg BW (0.61, 0.10, 1.01 or 1.52 mg As per kg BW
and 0.23, 0.38 or 0.57 mg Pb per kg BW). Three swine were also included to provide preliminary
data from mine waste contaminated samples. Each animal received either 100 mg/kg of residential
soil (1 animal), or surface tailings (1 animal), or subsurface tailings (1 animal). This was equivalent
to 0.15, 1.11 or 0.18 mg As per kg BW and 0.22, 0.43 or 0.43 mg Pb per kg BW, respectively. The
final study design is shown in Table 1.
Doses of soil administered to the swine were in the high end of the range expected for normal children
(i.e. 0.8 grams per day) up to the range depicting a child with pica (i.e. 10,000 grams per day) for soil
(U.S.EPA, 1989).
Animals in all treatment groups except the untreated controls and the intravenous controls received
a single intravenous dose of 206Pb in an aqueous solution immediately prior (within minutes) to
receiving test material described above. This intravenous administration of the 206Pb-enriched solution
and gavage administration of test materials enabled the comparison of intravenous and oral dose
responses within the same individual animal. This approach has the advantage of identical clearance
and other physiological factors influencing the elimination of lead. The dose of 206Pb was based on
results of the preliminary study and intended to add an insignificant amount (<10%) to the total dose
of lead.
Intravenous or gavage doses of sodium arsenate equivalent to the highest doses of arsenic in the
environmental material (0.61-1.52 mg As/kg BW, slag) were not administered. Acute toxicity has
been reported in swine and in humans exposed to highly bioavailable forms of arsenic in this dose
range (Osweiler et al., 1985; ATSDR, 1993). Therefore, intravenous and oral sodium arsenate doses
greater than 1 mg/kg were not administered.
EPA910/R-96-002 Page 6 of 48
-------
Twenty-four hours prior to dosing, the animals' weights were obtained and pre-dosing blood samples
were collected. Animals were preanesthetized and while under general anesthesia a self-retaining
catheter was placed in the urinary bladder. The catheter balloon was filled with a saline solution and
the catheter was connected to an empty, sterile urine collection bag. The bags were attached to the
back of the animal with adhesive tape. Animals fully recovered in the 24 hours prior to the
experiment. Following a 12 hour overnight fast, a second predose blood sample was collected. All
urine and feces were collected during the 24 hours prior to dosing.
At the beginning of the experiment following a 12 hour overnight fast (time 0), each animal was given
a single administration of the appropriate test material. Solutions of sodium arsenate and lead acetate
were administered separately and not mixed together prior to administration. Intravenous doses were
administered over a 2-3 minute period into the jugular vein using an 18 gauge butterfly infusion set
attached to a disposable syringe. Frequent withdrawals of blood into the infusion tubing confirmed
needle placement in the vein. Gavage arsenic solutions, lead solutions or environmental media were
mixed in a total of 40-ml of sterile double distilled water and administered directly into the stomach
via a lubricated tube passed through a mouth gag. An additional 20-ml of water was used to flush
the tube after initial dosing to assure that all material had reached the stomach. The animals were
cradled in the handler's arms while another person placed the gastric tube for dose administration.
The animals' behavior indicated that this handling technique minimized stress.
Clinical observations were recorded. Animals were observed at short (-15 minute) intervals
following dosing for signs of toxicity or emesis (vomiting). And, at 3-4 times during each of the
subsequent days. Food and water were provided four hours after dosing. At the end of the study,
urinary catheters were removed and the animals were returned to a swine confinement facility prior
to being sold or used for other research.
A maximum of eighteen animals could be efficiently handled for dosing and sample collection. The
experiment was conducted in two portions as indicated below. Control group animals (intravenous,
oral and untreated) were included in each portion. The dates were as follows: Range finding - dose
administered on 2/9/92; soil treatment group - dose administered on 7/24/92; slag treatment group -
dose administered on 9/11/92.
EPA 910/R-96-002 Page 7 of 48
-------
Table 1. Study Design
Treatment
Group
Untreated
Control - Oral
Gavage
Control - Oral
Gavage
Control - Oral
Gavage
Control - Oral
Gavage
Control -
Intravenous
Control -
Intravenous
Control -
Intravenous
Control -
Intravenous
Oral Test Soil
Oral Test Soil
Oral Test Soil
Oral Test Soil
Oral Test Slag
Oral Test Slag
Oral Test Slag
Oral Test Mine
Soil
Oral Test Mine
Tailings
Oral Test
Subsurface
Tailings
Test.
Material
Deionized Water
Aqueous Lead Acetate
Aqueous Sodium Arsenate;
Aqueous Lead Acetate
Aqueous Sodium Arsenate;
Aqueous Lead Acetate
Aqueous Sodium Arsenate;
Aqueous Lead Acetate
Aqueous Lead Acetate
Aqueous Sodium Arsenate;
Aqueous Lead Acetate
Aqueous Sodium Arsenate;
Aqueous Lead Acetate
Aqueous Sodium Arsenate;
Aqueous Lead Acetate
Aqueous Suspension
Aqueous Suspension
Aqueous Suspension
Aqueous Suspension
Aqueous Suspension
Aqueous Suspension
Aqueous Suspension
Aqueous Suspension
Aqueous Suspension
Aqueous Suspension
Dosing
Frequen
cy
Single
Single
Single
Single
Single
Single
Single
Single
Single
Single
Single
Single
Single
Single
Single
Single
Single
Single
Single
Dose Rate
mg soil/kg
BW
Water,
only
ND
ND
ND
ND
ND
ND
ND
ND
25
60
100
150
60
100
150
100
100
100
mg As/kg
BW
Water,
only
ND
0.01
0.11
0.31
ND
0.01
0.11
0.31
0.04
0.10
0.16
0.24
0.61
1.01
1.52
0.15
1.11
0.18
ug Pb/kg
BW
Water,
only
656
18
214
263
656
18
214
263
34
81
135
202
227
378
567
217
425
2464
ug 206Pb/kg
BW
Water,
only
1.68
1.68
1.68
1.68
ND
ND
ND
ND
1.68
1.68
1.68
1.68
1.68
1.68
1.68
1.68
1.68
1.68
No. of
Animals
4
1
1
1
1
1
1
1
1
3
3
3
3
3
3
3
1
1
1
ND = not dosed with this material
Dose Preparation
Intravenous and oral technical grade dose materials were prepared the day prior to use. Sodium
arsenate heptahydrate or lead acetate trihydrate were weighed on a Mettler Haining balance, and
mixed with deionized, distilled water in a volumetric flask. Arsenic or lead solutions were
administered separately to avoid chemical interactions. Solutions were submitted to the laboratory
for confirmation of arsenic and lead concentrations.
Approximately 0.25 grams of NIST SRM 983 Radiogenic Lead Isotopic Standard lead wire was
digested in a polytetrafluoroethylene centrifuge tube with 5 milliliters (ml) of glacial acetic acid and
2 ml of 30% hydrogen peroxide. The tube was heated, the digest diluted with deionized water and
filtered using a 45 micron membrane filter. The solution was further diluted in sterile saline solution
to give the final concentration. The isotopic composition of NIST SRM 983 Radiogenic Lead
solution was reported by NIST as: 92.1497% 206Pb, 6.5611% 207Pb, 1.2550% 208Pb, and 0.0342%
204Pb.
EPA910/R-96-002
Page 8 of 48
-------
Five samples of driveway slag collected from the vicinity of the Asarco/Ruston site were dried in a
circulating air oven at 80 degrees centigrade for 48 hours. Each of the five materials were pre-
screened using 1 millimeter (mm) plastic sieve, and the oversize material (greater than 1 mm) was set
aside. The less than 1 mm material was sieved using a 100 mesh stainless steel sieve (Tyler). The
less than 100 mesh material from each sample was analyzed and used in blending.
Preliminary analyses of each <100 mesh slag sample was conducted using microwave aqua regia
digestion. Approximately 0.5 grams of samples were accurately weighed into a fluorinated ethylene
propylene microwave digestion vessel: 2.5 ml of nitric acid and 7.5 ml of hydrochloric acid were
added, and the mixtures were heated for about 15 minutes while maintaining a stable pressure of
about 160 psig. The contents were diluted with distilled, deionized water, filtered through a 0.45
micron cellulose nitrate filter, and the volume of each solution was adjusted to 100 ml using distilled,
deionized water. Arsenic and lead were determined in the individual digests by inductively coupled
plasma atomic emission spectroscopy and inductively coupled plasma mass spectrometry; the latter
techniques employed rhodium and iridium as internal standards. The slag blend was produced by
combining a 400 gram subsample of each slag sample in a 2 liter plastic jar, which was then tumbled
for about 2 hours on a Norton ball mill tumbler.
The blended slag material was tested for homogeneity by spreading the entire lot of material onto a
large piece of paper and removing ten individual increments of about 0.05 grams. Each individual
increment was digested using 5 ml of nitric acid in a polytetrafluoroethylene test tube at 95 degrees
centigrade for about 16 hours, followed by filtration (0.45 micron cellulose nitrate) and dilution to
100 ml with distilled, deionized water. Arsenic and lead were determined in the digest using plasma
mass spectrometry with germanium and platinum as internal standards.
Sixteen individual soil samples were dried in a circulating air oven at 80 degrees centigrade for about
48 hours. Each dried sample was sieved using a 100 mesh stainless steel sieve and less than 100 mesh
fractions were analyzed and used in blending. Individual samples were analyzed following digestion
of a 0.25 gram subsample using 5 ml of nitric acid in a polytetrafluoroethylene test tube at 85 degrees
centigrade for about 2 hours, followed by filtration (0.45 micron cellulose nitrate filter) and dilution
to 100 ml with distilled, deionized water. Arsenic and lead were determined in the individual digests
using plasma mass spectrometry with rhodium and iridium as internals standards. Based upon lead
results exceeding 500 milligrams per kilogram, a blend was produced from nine of the sixteen
samples. Blending was conducted as described above for the slag blend. The blend was tested for
homogeneity at the 0.05 gram level as described above for the slag material.
Mining site materials were adequately homogeneous at the 0.25 gram level. Vacuum cleaner dust
was dried at 80 degrees centigrade for about 36 hours, then sieved with a 100 mesh stainless steel
sieve. The less than 100 mesh fraction was analyzed. The surface soil composite, subsurface tailings
composite and surface tailings composite were individually dried, sieved and analyzed as described
above for dust.
EPA 910/R-96-002 Page 9 of 48
-------
Biological Sample Collection and Handling Protocol
Whole blood was collected from the jugular vein in sterile EDTA-treated containers using a new 20
gauge needle for each animal. For blood collection, animals were restrained by leaning the animal's
back against the handler's legs in a nose down position. This handling technique enabled animals to
rest quietly with minimum of restraint during blood collection. Blood was collected from all animals
prior to dosing and at the intervals shown:
Prior to dosing: 24 and 12 hours
After dosing: 15, 30, 60 minutes, 1.5, 3, 6, 12, 24, 48, 72, 96, 144 hours
Urine volume was recorded and acidified urine samples were collected from the urine collection bags.
Samples were grouped in the intervals shown.
Composite Intervals:
Before dosing: 24 hrs - 12 hrs, and 12 hrs through dosing (time 0)
After dosing: time 0-12 hrs, 12-24 hrs, 24-48 hrs, 48-72 hrs, 72-96 hrs, 96-144 hrs
All feces produced during the experiment were collected and frozen. Samples were grouped
in the intervals are shown below:
Before dosing: 24 hrs - time 0
After dosing: time 0 through 3 days, 3 days through 7 days
All urine and blood samples were stored at 0-5 °C in secured facilities prior to shipment to the
analytical lab. Standard chain-of-custody procedures were followed. Holding times for blood, urine
and feces have not been officially established by EPA; however, for water samples, the standard
holding time is 6 months. In this study, blood and urine samples were analyzed within 30 days of
collection. Some blood arsenic samples were reanalyzed within 120 days. Feces samples were
analyzed within 90 days of collection.
Samples were prepared as follows. Feces were mixed together using an electric drill and auger
adding only enough distilled, deionized water to form a thick paste. The homogenized composite for
each time interval was weighed, a portion was removed and dried at 50 degrees centigrade. Percent
dry weight was determined by the weight difference before and after drying. Following collection,
blood and urine samples were divided at the animal testing facility. Blood and urine samples were
maintained at 4 degrees centigrade at all times during preparation and analysis. A portion of each
blood, urine and fecal sample was prepared for each of the testing laboratories and one archive
portion was retained.
Quality Assurance
Quality Assurance (Q A) Proj ect Plans were developed by the analytical laboratory, ES A Laboratories
(43 Wiggins Avenue, Bedford, MA 01730) and the animal facility, Michigan State University (MSU,
E. Lansing, MI) and submitted to Region 10 EPA. QA Coordinators were designated at each
laboratory and were responsible to assure procedures described in the Plans were followed. Reports
were submitted to Region 10 EPA for quality assurance verification and audit of the raw data.
EPA910/R-96-002 Page 10 of 48
-------
Split samples of a minimum of 10% of the blood and urine samples were independently analyzed by
the U. S.EPA National Enforcement and Investigations Center (Denver, CO) utilizing different sample
preparation and analytical techniques. Independently acquired quality control blood, urine and fecal
samples were included in batches of samples sent to both labs at the rate of 7-13% The samples'
appearance was similar to the biological specimens and labeling was not distinguishable from other
samples except for a unique sample number.
Characterization of Arsenic and Lead
Environmental Samples. Characterization of the environmental dose materials included multi-element
determinations, homogeneity testing, particle size analysis, moisture content, organic matter content,
pH, lead isotopic analysis and mineralogical evaluation.
To analyze the soil, slag, tailings and dust and verify concentrations of lead and arsenic in the positive
control dosing solutions, a Jarrell-Ash model 61 inductively coupled argon plasma atomic emission
spectrometer was used with a fixed cross-flow nebulizer, mass flow meters for all gas streams, and
a peristaltic sample delivery pump. Spectral background and inter-element interference corrections
were applied. The Jarrell-Ash model 61 inductively coupled argon plasma atomic emission
spectrometer was also used with a hydride generation system and mass flow meters for all gas
streams. Spectral background and inter-element interference corrections were not necessary.
To determine the lead isotope ratios, the lead and arsenic in animal food and water, and for the
analyses of lead in urine and blood, a Sciex Elan model 250 inductively coupled argon plasma mass
spectrometer, equipped with mass flow meters for all gas streams and a peristaltic sample delivery
pump was used. A refrigerated circulating bath was used to maintain the nebulizer spray chamber
at a temperature of 10 °C. Meinhard TR-C concentric glass nebulizers (I.E. Meinhard Associates,
Santa Ana, CA) were used. The ion optics of the spectrometer are the updated version: voltage
adjustments consist of a barrel lens setting of 11, the plate lens at 2, the Einzel at 95, and the photon
stop at 37. The instrument was operated in the multichannel (peak hopping) mode, with single
measurements being taken at the nominal mass value of each peak. The low resolution mode was
used, producing peak widths of 1.0 - 1.1 m/z at 10% height. The program "Spectrum Display" was
used to collect data, which were directed to a personal computer for storage and manipulation using
Statgraphics software.
The following digestion and extraction sample preparation procedures were used. Samples of soil,
slag, tailings and dust for multi-element analyses were prepared by a potassium hydroxide fusion
method. The fusion consisted of mixing 0.25 grams of sample with 2.0 grams of potassium hydroxide
in a 10 mL pyrolytic graphite crucible. The sample-potassium hydroxide mixture was placed in an
electrically heated muffle furnace and heated for one hour at each of the following temperatures: 150,
300, and 450 °C. Following the last heating step, the fused mixtures were removed from the furnace,
allowed to cool, and carefully immersed in a plastic beaker containing 15 mL of distilled, deionized
water, 5 mL hydrochloric acid, and 5 mL nitric acid. After an hour of agitation on a rotary shaker,
0.5 mL of 30% hydrogen peroxide was added. The beakers were capped (a small slit was made in
the plastic cap so gases could vent) and then agitated overnight on an oscillating shaker, followed by
filtering through 0.45 micron cellulose nitrate membrane disposable filter units. The fusates were
diluted to 100 mL with deionized water and then transferred to high-density polyethylene bottles for
storage. Each fusate contained 5 mL of concentrated hydrochloric acid and 5 mL of concentrated
nitric acid per 100 mL total volume.
EPA 910/R-96-002 Page 11 of 48
-------
Samples of soil, slag, tailings, animal food and water for lead isotopic analyses were prepared by a
nitric acid digestion method. The digestions were performed in capped 30 mL
polytetrafluoroethylene centrifuge tubes which were heated to approximately 90 °C for 18 hours in
an air convection oven. The digestates consisted of a 0.25 gram portion of a sample in 5 mL
concentrated nitric acid. Following heating, the digestates were allowed to cool, and 20 mL of
distilled, deionized water added. The digestates were filtered through 0.45 micron cellulose nitrate
membrane disposable filter units, diluted to 100 mL with distilled, deionized water, and then
transferred to high density polyethylene bottles for storage.
Lead isotopic analyses were conducted. Lead has four stable isotopes 204Pb, 206Pb, 207Pb, and 208Pb
of which the three higher mass isotopes are the daughter nuclides of radioactive decay from uranium
and thorium. The isotope, 204Pb, has no known radioactive parent of significance. As a result the
absolute abundance of 204Pb remains constant, while the absolute abundance of the other isotopes
increases systematically with time. Consequently the isotopic composition of a given sample will be
dependent on its age and the relative proportions of thorium and uranium in the parent strata. Isotope
abundances of lead are commonly expressed in a relative fashion. For example, the abundance of
208Pb is typically about 50% and the 206Pb is approximately 25% therefore the 208Pb:206Pb ratio is 2.
Each lead ore body has its own specific set of isotope ratios (i.e. its "isotopic fingerprint").
Lead isotopic ratios were determined by plasma mass spectrometry, using pneumatic nebulization,
in the nitric acid digestates. Thallium was added to the digestates and served as an internal standard
for mass discrimination correction. Long counting times of 100 seconds were used and the mass
spectrometer was operated in a peak hopping mode with a dwell time of 50 milliseconds and the low
resolution mode. Single measurements were made for each preparation (three preparations per
sample). NBS SRM 981 (Common Lead) was used as the control sample from which bias is inferred.
Arsenic and Lead Aqueous Dosing Solutions. Arsenic and lead aqueous dosing solutions were
analyzed as follows. Three levels of dilution of each solution of sodium arsenate or lead acetate were
analyzed by plasma emission spectroscopy. Each dilution was analyzed in triplicate and one dilution
for each of the solutions was spiked appropriately with arsenic or lead. Each dilution was matrix
matched to the calibration standards. In addition, three second source reference standards were
analyzed to verify the calibration standards. The filtered radiogenic lead isotopic standard solution
was analyzed seven times by plasma mass spectrometry. The isotopes 206Pb, 207Pb, and 208Pb were
measured with 203T1 and 205T1 as internal standards.
Water and Animal Feed. Water and animal feed were analyzed as follows. A composite of three
drinking water samples was collected from the animal watering system, approximately 3 days prior
to the first animal dosing, day 3 of the first animal treatment group and at the first dosing of the
second animal treatment group. A composite feed sample was prepared from five grab samples from
the feed batch consumed by the swine during the study. Multi-element analyses were conducted on
both feed and water samples. About 0.5 grams of pulverized feed was digested with 5 mL nitric acid
in polytetrafluoroethylene 30 mL screw cap centrifuge tubes. Each vessel was heated with
microwave energy in a CEM Model 205 digestion system at 400 watts. Ten mL purified water was
added to each cooled sample. Solutions were filtered through 0.45 micron filters and then diluted
to 100 mL. Water was analyzed after adding 5 mL nitric acid per 95 mL. Solutions, blanks and
spiked samples were analyzed using plasma mass spectrometry. Rhodium and iridium were used as
internal standards.
EPA 910/R-96-002 Page 12 of 48
-------
Biological Samples. Biological samples were analyzed as follows. Blood arsenic concentrations
were analyzed by atomic absorption spectrophotometry (Hitachi Z-6100) and gaseous hydride
generation (HFS-2 Hydride Generation). Duplicate 1.0 ml aliquots of sample are pipetted into acid
washed glass tubes. A standard acid mixture of 30:10:1 (HNO3:HC1O4:H2SO4) is added to each
sample then heated at 128 degrees centigrade for 10-15 minutes, increased to 150 degrees for 15
minutes, increased to 220-250 degrees and digested to white HC1O4 fume for 5-10 minutes. Then
cooled. Aliquots of the following were added sequentially with mixing: 2N HC1, 20% urea, 20% KI
and 10% ascorbic acid.
Blood lead concentrations were determined as follows. Plasma source mass spectrometry
measurements were made at 203T1,205T1,206Pb, 207Pb, and 208Pb. For all scans, an equal measurement
time of thirty seconds was used for each isotope. Each digestion batch represented an instrumental
analysis run along with calibration standards of 0.0005, 0.002, 0.005, and 0.010 mg L"1 lead,
calibration blanks, and continuing calibration standards all containing 0.100 mg L"1 thallium as an
internal standard. For a number of batches the calibration curve range was extended by analyzing
standards to bracket the highest specimen concentration. For total lead, the signals for the three lead
isotopes were summed and ratioed to the sum of the signals for the two thallium isotopes . For the
individual lead isotope calibration the signal for each isotope was ratioed to the sum of the signals for
the two thallium isotopes. The calibration blank average lead/thallium signal ratio was subtracted and
calibration curves were fitted by regression (y=mx). For the digests, the average batch digestion
blank value was subtracted.
Blood digestion procedures prior to lead analyses involved the following. Polytetrafluoroethylene
30 mL screw-cap centrifuge tubes were used as the digestion vessels. Each vessel was rinsed twice
with water and 3.0 mL nitric acid was added and the tubes were filled with water. The filled tube was
placed in a preheated 95 °C convection oven for at least four hours and usually overnight. The nitric
acid leach was then repeated with fresh solution. For each tube prior to use, 1.0 mL of concentrated
nitric acid was added to an empty tube and the tube heated for one hour at 130 °C in a forty place
block digestor (Techne Model DG-1). The tube was cooled, 9.0 mL water was added and the
contents were shaken. This solution was analyzed by plasma source mass spectrometry for lead. If
the analysis indicated a count rate of less than 50 counts per second (about 50,000 counts per second
for 0.100 mg L"1 lead and less than 20 counts per second for a calibration blank at 208Pb), then the
tube was deemed clean. If the count rate was greater than 50 counts per second, the tube was
rejected and taken through another cleaning and analysis sequence. Tubes were subjected to the
cleaning and analysis acceptance procedure between use. Generally for the early batches the tubes
had a repeat cleaning rate of about 30%. For the later batches, few of the tubes required repeated
cleaning.
Each specimen vial was shaken and 0.5 gram of a specimen was weighed into a clean digestion vessel.
A new disposable pipet tip was used to transfer each aliquot. The internal standard of 0.010 mL of
100 mg L"1 thallium was added followed by 1.0 mL of nitric acid. The vessel was capped and placed
in the preheated 130 °C block digestor. After five to seven minutes each tube was vented. Each tube
was vented two or three more times over the one hour digestion period. After cooling, 8.5 mL of
water was added to each vessel. The specimens for two swine were digested as a batch along with
six blanks and two different levels of NIST SRM 955 lead in bovine blood. One specimen for each
swine was also digested in duplicate for each batch. A total of eighteen digestion batches were
processed.
EPA 910/R-96-002 Page 13 of 48
-------
Urinary arsenic concentrations were determined using a Jarrell-Ash Model 61 inductively coupled
argon plasma atomic emission spectrometer was used with a hydride generation system and mass flow
meters for all gas streams. The incident power was 1.25 kilowatts and the reflected power was less
than five watts. Argon flows of 20 and 0.625 L min"1 for the torch and sample gas, respectively, were
used. The observation height was set at 13 mm. Spectral background and inter-element interference
corrections were not necessary. A glass hydride gas stripping cell provided mixing of the
continuously introduced digested solution and reducing agents, separation of the hydrides and
hydrogen from the spent liquid, and mixing of the gaseous products with the argon carrier gas for
transmission to the injector of the plasma torch. The digestates were introduced into the hydride
generator at a flow rate of 3.0 mL min"1. The reducing agents were a potassium iodide solution (8%
w/v) pumped at a flow rate of 1.2 mL min"1 followed by an alkaline tetrahydroborate solution (2.4%
w/v sodium tetrahydroborate in 0.1 N sodium hydroxide) pumped at a flowrate of 1.2 mL min"1.
Each digestion batch represented an instrumental analysis run along with calibration standards,
calibration blanks, and continuing calibration standards. The known addition for each specimen was
used to correct for calibration slope rotational error. For the digests, the average batch digestion
blank value was subtracted.
Urine digestion procedures prior to arsenic analyses involved the following. Each specimen tube was
shaken and 2.0 mL of a specimen was transferred into a clean Corex 30 mL centrifuge tube. Two
small glass beads and 0.5 mL of concentrated sulfuric acid was added. The vessels were heated for
approximately one hour at 160 °C in a forty place block digestor (Techne Model DG-1). Completion
of this step was indicated by viscous black appearance of the digestates. The temperature of the
block digestor was then adjusted to 300 °C. Fuming occurs when digestion is complete.
Approximately 30 minutes was required to produce white sulfur trioxide fumes. After 10 minutes
of fuming, the tubes were removed from the block digestor, and then 0.2 mL of concentrated nitric
acid was added dropwise to each vessel. The tubes were then returned to the block digestor for
approximately two minutes, whereupon most digestates were colorless or slightly yellow. For those
vessels containing a dark colored solution, the heated nitric acid treatment was repeated until the
solution cleared. After the vessels cooled, 7.8 mL distilled, deionized water and 2 mL concentrated
hydrochloric acid was added to each vessel. Each vessel was mixed on a vortex mixer. Two separate
5 mL aliquots of the digestate were transferred to two 10 mL polypropylene screw cap test tubes.
Ten to 100 microliters of a 5 mg L"1 arsenic as monosodium methylarsonate standard (Diamond
Shamrock, Houston, Texas) was added to one of the tubes to formulate a known addition. The
remainder of each digestate was transferred to another test tube for storage. The specimens for four
swine were digested as a batch along with six blanks and the two levels of NIST SRM 2607. Two
specimens for each batch were also digested in duplicate and one was spiked with known amount of
arsenic. A total of nine digestion batches were processed.
Urinary lead concentrations were determined using plasma source mass spectrometry measurements
made at 203T1,205T1,206Pb, 207Pb, and 208Pb. For all scans, an equal measurement time of thirty seconds
was used for each isotope. Each digestion batch represented an instrumental analysis run along with
calibration standards of 0.0005, 0.002, 0.005, and 0.010 mg L"1 lead, calibration blanks, and
continuing calibration standards all containing 0.100 mg L"1 thallium as an internal standard. For a
number of batches the calibration curve range was extended by analyzing standards to bracket the
highest specimen concentration. For total lead, the signals for the three lead isotopes were summed
and ratioed to the sum of the signals for the two thallium isotopes. The calibration blank average
lead/thallium signal ratio was subtracted and calibration curves were fitted by regression (y=mx). For
the digests the average batch digestion blank value was subtracted.
EPA 910/R-96-002 Page 14 of 48
-------
Urine digestion procedures prior to lead analyses involved the following. Each specimen tube was
shaken and 5.0 mL of each was transferred into a clean digestion vessel (see section 4.b.ii, above, for
the cleaning verification procedure). A new disposable pipet tip was used to transfer each aliquot.
The internal standard of 0.01 mL of 100 mg L"1 thallium was added followed by 1.0 mL of nitric acid.
The vessel was capped and placed in the preheated 13 0 °C block digestor. After five to seven minutes
each tube was vented. Each tube was vented once over the one hour digestion period. After cooling,
4.0 mL of water was added to each vessel. The specimens for four swine were digested as a batch
along with six blanks and two different levels of NIST SRM 2670 reconstituted freeze dried urine.
Two specimen per batch were digested in duplicate and one specimen was spiked with a known
amount of lead. A total of nine digestion batches were processed.
Data Evaluation Methodology
Methodology was developed to estimate the extent of arsenic and lead absorption following oral
exposure to environmental materials. Obj ectives were (a) To address the presence of ubiquitous pre-
existing background concentrations in biological samples, (b) To enable comparisons of experimental
groups of swine receiving high intakes of environmental toxicants with groups receiving low, non-
toxic doses of equivalent technical grade chemicals, and (c) To provide an estimate of variability of
the calculated biological availability of the metal/metalloid.
Area-under-the-curve (AUC) for blood concentration versus time was determined from zero to 144
hours for individual test animals. The time series of observations for each individual was corrected
for background by subtracting each animal's average pre-experiment blood arsenic or lead
concentration from the concentrations observed after dosing.
The Student's t-test was used to evaluate the difference between the mean pre-experiment blood
arsenic concentration of the twenty-seven treated animals and the mean blood arsenic concentration
of the six untreated animals. In each of the treated groups, a regression model was determined which
described the relationship (with subtraction of endogenous background) between AUC and the dose.
Dose was expressed as micrograms (jig) of arsenic (As) per kilogram (kg) of body weight (BW) and
the regression model passed through the origin.
Widely differing predose values and postdose variances in blood lead data negated standard analyses
of variance methods. In Table 10, total blood lead, the percent change between the two pre-dose
baseline concentrations in an individual animal ranged between minus 34% to plus 24%. Within a
treatment group, the coefficient of variation in the pre-dose baseline blood lead concentrations ranged
from 17 to 65%. These widely differing pre-dose baseline values limit the sensitivity of the
experiment. A treatment group must demonstrate a uniformly large change in blood lead
concentrations in order to conclude statistical significance by standard analysis of variance methods.
EPA 910/R-96-002 Page 15 of 48
-------
However, within a treatment group the coefficient of variation in the calculated area-under-the-curve
(see Table 20) ranged from 50-133%. This degree of variance within a group precluded detection
of significant differences.
Inspection of the blood lead concentration results indicated increases in concentrations. Therefore,
as an alternative approach, each animal in the untreated and treated groups was tested for significant
upward or downward trend in blood lead concentration. The trends were assessed for significance
by both Pearson's r and Kendall's tau correlation analyses, with the time sequence of successive
observations serving as the independent variate. In this approach, all animals were individually tested
against the hypothesis that treatment had no nonrandom effects upon blood lead concentration. Inter-
group differences were judged by comparing the intra-group tests. In this way, for example, soil-
dosed animals all showing a significant increase are clearly responding differently than a population
of untreated animals all showing either no change or the opposite change in the same parameter. A
value of P less than 0.05 was considered statistically significant. This statistical approach is not
standard but has been previously applied to the evaluation of blood parameters in immature swine
studies when unexplained pattern of variances was observed (Lorenzana et al., 1985a; Lorenzana et
al. 1985b).
In each of the treated groups, a regression model was determined which described the relationship
(with subtraction of pre-existing background) between AUC and the dose. Dose was expressed as
micrograms (jig) of lead (Pb) per kilogram (kg) of body weight (BW). Only the regression model
for the intravenously dosed animals passed through the origin. The intravenous lead infusion was
introduced directly into the bloodstream; therefore, concentration responses in the blood are expected
to be linear. Linear intravenous responses have been previously reported (Freeman et al., 1994; Weis
etal., 1993; Aungstetal., 1981).
Non-linear responses at low oral exposures have been previously reported (Freeman et al., 1992;
Weis et al., 1993; Aungst et al., 1981). Because a non-linear range was not identified in this study,
regression models for slag and soil-exposed groups were not forced through the origin. The
regression model describing the relationship between area-under-the-curve (AUC) and dose therefore
is not adequate to describe the relationship between lead intake and blood concentrations at intake
doses below the experimental range.
For each regression, the 95% confidence intervals for the regression coefficient (slope) were
determined. These confidence limits were used in Monte Carlo analyses to determine confidence
limits for bioavailability.
EPA 910/R-96-002 Page 16 of 48
-------
RESULTS
Range Finding Experiment
Results indicated that larger doses of the soil (higher blood lead concentrations) were needed in order
to obtain greater precision in the lead isotope ratio data, and to determine concentration responses
versus time. It was also determined that it was less stressful for the pigs to have urinary collection
bags than to obtain urine directly from the catheter. Water consumption, appetite and clinical
appearance were normal, and urinary output was adequate. No gross lesions, including bladder
lesions, were observed at necropsy. Information from the range finding experiment was used to refine
the final design of the experiment which is reported in this study.
Quality Assurance Review of Final Study
The quality assurance review is summarized in Appendix A. Data utilized for bioavailability estimates
met quality assurance criteria developed for this study and were verified by two different laboratories.
Grain and Water Analyses
Metal residue results are shown in Table 2, and indicate that nominal concentrations of arsenic and
lead were present in the feed. No arsenic was detected in drinking water. Approximately two parts
per billion lead were detected in drinking water.
Table 2. Grain and Tap Water Results
. Grain 1
Element . „ .
(mg/kg)
Antimony 0.127
Arsenic 0.462
Cadmium 0.188
Cobalt 0.191
Copper 249.
Iron 169.
Lead 1.30
Manganese 43.
Molybdenum 2.48
Nickel 2.74
Thallium < 0.05
Selenium 1.93
Zinc 63.5
Grain 2
(mg/kg)
0.096
0.382
0.216
0.188
167.
136.
1.09
42.
2.58
2.67
<0.05
0.927
51.2
Grain3
(mg/kg)
0.135
0.592
0.217
0.210
189.
203.
1.29
44.4
2.48
2.64
<0.05
2.04
77.2
Tap Water
(mg/L)
< 0.004
< 0.0005
0.120
0.180
0.0015
0.035
0.002
< 0.0002
< 0.010
0.380
Dose Analyses
The arsenic and lead concentrations determined by chemical analyses were used for bioavailability
calculations. The concentrations of arsenic in the control gavage and intravenous solutions were,
respectively, 143 (sd=3.0) or 837 (sd=14.5) milligrams arsenic per liter (sodium arsenate solutions).
The concentrations of lead in the control gavage and intravenous solutions were, respectively, 177
(sd=6.8) or 1034 (sd=26) milligrams lead per liter. The concentration of 206Pb in the dosing solution
was 2.97 milligrams lead per liter with a standard deviation of 0.013 mg/L.
The elemental composition of soil, slag and mining samples are shown in Table 3. These
environmental substrates contained arsenic concentrations ranging from approximately 1500
milligrams arsenic per kilogram of soil (mg/kg) to approximately 11,000 mg/kg. Lead concentrations
EPA910/R-96-002
Page 17 of 48
-------
ranged from approximately 1,300 mg Pb/kg to 25,000 mg Pb/kg. Other metal compounds were also
present in the substrates.
The isotopic composition of soil, slag and mining samples are present in Table 4. No significant bias
was indicated by the NBS 981 determinations and imprecisions were within acceptable limits. The
similarities of the ratios for the mining site samples strongly suggest the same lead parent strata for
these environmental substrates. The same can be concluded for the two smelter site samples. The
source discrimination power of the isotope technique is illustrated by the differences detected between
the mining site samples, the smelter site samples and the standard reference materials.
Table 3. Elemental Concentrations in Environmental Substrates
Element
Aluminum
Antimony
Arsenic
Barium
Beryllium
Boron
Cadmium
Calcium
Chromium
Cobalt
Copper
Iron
Lead
Lithium
Magnesium
Manganese
Molybdenum
Nickel
Phosphorus
Selenium
Silicon
Silver
Sodium
Strontium
Sulfur
Thallium
Tin
Titanium
Vanadium
Yttrium
Zinc
Smelter Site Soil Composite Sample
Average
mg/Kg
65000
145
1600
526
ND
26
7
16500
91
17
2670
34800
1350
17
8630
768
15
59
935
ND
274000
15
16500
240
510
ND
53
4850
99
15.8
332
Std Dev
mg/Kg
1160
4.6
31
7.0
2.2
0.7
1.1
273
0.4
0.4
50
660
31
1.2
114
14
1.2
3.3
19
19
12800
0.2
560
3.6
23
6.5
5.2
91
2.3
0.1
8.5
LCL
mg/Kg
62100
134
1530
509
0
24
4
15800
89
16
2540
33200
1270
14
8350
733
12
50
887
0
242000
14
15100
231
453
0
40
4630
93
15.4
311
UCL
mg/Kg
67900
156
1680
544
8
28
10
17100
92
18
2790
36500
1430
20
8920
803
18
67
983
52
306000
15
17900
249
568
14
66
5080
105
16.1
353
Smelter Site Slag Composite Sample
Average
mg/Kg
18200
3350
10100
274
6.9
18
16
41700
401
269
5220
224000
3780
8.0
6910
850
1770
93
527.2
ND
180000
18.2
5000
133
3570
ND
367
1590
58
11
11400
Std Dev
mg/Kg
520
309
407
8.1
0.3
1.8
2.9
1180
23
9.1
169
4310
118
0.9
163
22
67
2.3
30.2
15
11100
1.4
320
4.7
128
26
14.8
62
2.6
0.3
458
LCL
mg/Kg
16900
2580
9100
254
6.0
14
8.4
38700
343
246
4800
214000
3480
5.7
6500
796
1600
88
452.2
0
152000
14.8
4210
121
3260
0
331
1440
52
10.7
10200
UCL
mg/Kg
19500
4120
11100
294
7.7
23
23
44600
459
291
5640
235000
4070
10.3
7310
904
1930
99
602.2
24
207000
21.6
5790
144
3890
75
404
1750
65
12.1
12500
EPA910/R-96-002
Page 18 of 48
-------
Table 3. Elemental Concentrations in Environmental Substrates
Element
Aluminum
Antimony
Arsenic
Barium
Beryllium
Boron
Cadmium
Calcium
Chromium
Cobalt
Copper
Iron
Lead
Lithium
Magnesium
Manganese
Molybdenum
Nickel
Phosphorus
Selenium
Silicon
Silver
Sodium
Strontium
Sulfur
Thallium
Tin
Titanium
Vanadium
Yttrium
Zinc
Mine Site Surface Tailings Sample
Average
mg/Kg
20400
646
11100
571
6.2
144
76
38800
62
3.0
270
52100
4250
11
14137.0
5840
36
80
1980
ND
172000
30
953
60
34500
ND
74
1510
526
26.1
11000
Std Dev
mg/Kg
822
21
379
24
0.4
6.8
3.4
1030
3.6
0.9
6.3
1863
177
1.2
430.6
176
2.4
7.5
87
16
17600
1.5
139
1.8
1360
15
5.1
65
21
0.3
410
LCL
mg/Kg
18400
595
10200
512
5.3
127
68
36200
53
0.8
254
47500
3810
8
13067.2
5410
30
61
1760
0
129000
26
607
56
31100
0
62
1350
472
25.3
9980
UCL
mg/Kg
22500
697
12000
629
7.1
161
85
41300
71
5.2
285
56700
4690
14
15206.9
6280
42
98
2200
54
216000
34
1299
65
37900
48
87
1670
579
26.9
12000
Mine Site Subsurface Tailings Sample
Average
mg/Kg
20200
1720
1810
2160
1.7
116
6.1
36100
32.5
10.5
504
51600
24600
13.1
11200
1360
16.7
61
885
ND
223000
303
1110
57
43900
ND
ND
1520
144
11.7
5410
Std Dev
mg/Kg
433
10
34
64
0.1
3.6
1.8
378
0.1
0.7
11.6
809
511
0.3
265
34
0.8
2.5
32.4
8.5
22700
5.3
79
1.2
291
15
4.2
9.8
2.4
0.2
86
LCL
mg/Kg
19100
1700
1730
2000
1.4
107
56
35100
32.2
8.8
475
49600
23400
12.4
10600
1270
14.8
54
805
3.2
166000
290
918
54
43200
0
0
1500
138
11.1
5200
UCL
mg/Kg
21200
1740
1900
2310
2.0
125
65
37000
32.8
12.2
532
53600
25900
13.8
11900
1450
18.6
67
966
45.5
279000
316
1310
60
44600
58
13
1540
149
12.2
5630
EPA910/R-96-002
Page 19 of 48
-------
Table 3. Elemental Concentrations in Environmental Substrates
Element
Aluminum
Antimony
Arsenic
Barium
Beryllium
Boron
Cadmium
Calcium
Chromium
Cobalt
Copper
Iron
Lead
Lithium
Magnesium
Manganese
Molybdenum
Nickel
Phosphorus
Selenium
Silicon
Silver
Sodium
Strontium
Sulfur
Thallium
Tin
Titanium
Vanadium
Yttrium
Zinc
Mine Site House Dust Sample
Average
me/Kg
14500
54
713
477
ND
95
14
19900
44
5
99
12500
665
8
7300
841
7
39
1060
ND
104000
6
127000
79
6400
ND
19
1360
62
7.6
1510
Std Dev
me/Kg
770
10
43
24
1.2
10
0.9
1150
1.6
1.0
5.9
606
32
0.7
375
44
1.9
3.0
48
14
4530
1.8
4040
4.9
356
15
7.5
70
2.5
0.4
78
LCL
me/Kg
12600
29
606
416
0
70
11
17000
40
2
84
11000
587
6.5
6370
731
2.2
31
943
0
93100
2
117000
67
5520
0.0
0.9
1190
56
6.5
1310
UCL
mg/Kg
16400
80
819
538
3
121
16
22700
47
7
114
14000
744
10.1
8230
951
11.7
46
1180
45
116000
11
137000
91
7290
42
38
1530
68
8.7
1700
Mine Site Surface Soil Sample
Average
mg/Kg
38800
183
1540
1180
3.8
129
32.4
31600
82
7
130
31300
2170
23.5
19700
2290
27
70
1620
ND
309000
17
4070
130
6070
22.8
30
2590
276
25
4500
Std Dev
mg/Kg
1620
6.0
23
33
0.2
3.9
0.5
490
0.5
0.2
5.0
595
14
0.9
523
27
3.3
3.7
51
14
587
1.5
63
2.3
44
0.8
5.2
67
3.4
.4
55
LCL
mg/Kg
34800
168
1490
1100
3.3
119
31.3
30400
80
6
117
29800
2140
21.2
18400
2220
18
60
1490
0
307000
13
3910
125
5960
20.8
17
2420
267
21
4360
UCL
mg/Kg
42900
198
1600
1270
4.2
139
33.6
32800
83
7
142
32800
2210
25.8
2100
2350
35
79
1750
50
310000
21
4220
136
6180
24.8
43
2750
284
29
4630
Std Dev = Standard Deviation
ND = Confidence interval included zero
LCL = Lower confidence level at 0.05 significance level
UCL = Upper confidence level at 0.05 significance level
Table 4. Lead Isotope Ratios
Mine Site Surface Soil
Mine Site Surface Tailings
Mine Site Subsurface Tailings
Mine Site House Dust
Smelter Site Soil
Smpltpr Sitp Slao
NBS SRM 981
Certified Value
Deviation
% Deviation
Pb-204 to Pb-206
Average
0.05063
0.04996
0.05053
0.05062
0.05566
n ns4Qi
0.05905
0.05904
•0.00001
-0.013
Std Dev
0.00051
0.00041
0.00062
0.00082
0.00135
n nni9i
0.00088
%RSD
1.02
0.82
1.23
1.62
2.43
9 9fi
1.50
Pb-207 to Pb-206
Average
0.79935
0.79167
0.79862
0.79480
0.86104
n x^m
0.91105
0.91464
0.00359
0.393
Std Dev
0.00269
0.00282
0.00257
0.00284
0.00378
n nrpsx
0.00774
%RSD
0.34
0.36
0.32
0.36
0.44
n -in
0.85
Pb-208 to Pb-206
Average
1.98702
1.96605
1.95291
1.99505
2.10150
9 nQ9sn
2.15507
2.16810
0.01303
0.601
Std Dev
0.01537
0.01825
0.01141
0.00772
0.01145
n rn/ns
0.03490
%RSD
0.77
0.93
0.58
0.39
0.54
n fix
1.62
n = 3 for all samples
n = 27 for the SRM
Clinical Observations
EPA910/R-96-002
Page 20 of 48
-------
All animals appeared clinically normal during the course of the study. One animal in the untreated
group was found dead five days into the study. Necropsy did not reveal the cause of death. Animals
were not sacrificed at the end of the study.
Urine Samples
Individual animal urine volumes are provided in Table 5. Several urine samples were lost due to
disconnection of the catheter with the urine collection bag, some urine leaked around the catheters
and several animals lost their catheters during the study. Consequently, analytical results do not
represent all of the arsenic or lead which may have been excreted in urine
Table 5. Individual Animal Urine Volumes (milliliters)
Animal# (group #)
126(1)
137(1)
195(1)
190(1)
133(2)
199(2)
140(3)
188(3)
141(8)
182(8)
138 (9)
196 (9)
130 (4)
131 (4)
136 (4)
139(5)
145 (5)
148(5)
129(6)
144(6)
150(6)
142 (7)
147(7)
149 (7)
192(10)
194(10)
189(10)
180(11)
197(11)
179(11)
186(12)
191 (12)
184(12)
177(13)
181 (14)
187(15)
Hour Intervals After Dosing
pre-24
370
550
770
2290
270
290
370
310
840
330
890
330
580
560
125
530
430
210
300
440
540
490
2210
1130
310
280
250
870
1090
1070
590
1440
2790
430
0-12
310
510
630
410
190
310
140
90
190
580
630
350
170
610
440
530
570
210
190
220
370
180
470
430
210
410
650
230
650
570
310
390
910
170
12-24
610
650
750
870
310
90
140
130
390
580
710
350
290
450
570
610
490
190
280
390
230
890
830
350
750
730
170
790
750
80
550
1150
190
24-48
1070
530
1250
1070
410
90
430
110
650
980
1330
1150
570
790
950
1350
1070
990
750
350
650
1310
630
410
690
870
1370
2130
1230
1270
110
1490
1510
2530
190
48-72
790
810
950
2050
900
270
810
170
530
1670
1230
1210
580
1710
930
1170
610
900
870
650
470
900
570
390
1110
1690
2110
1490
510
2050
350
2310
2990
2890
210
72-96
1050
1190
2110
1290
610
690
1070
90
650
920
880
1070
800
770
950
1040
1300
1330
540
580
215
690
780
590
730
970
1550
282
450
2630
190
1390
770
2690
80
96-144
160
3340
4310
3750
1570
730
1570
570
1440
1080
915
2710
390
1490
3340
2875
1575
2810
870
2940
875
1670
450
690
2390
3270
3630
4290
470
4750
1090
4190
4710
5290
990
Empty cell indicates no sample was obtained.
Arsenic Concentrations in Blood
Arsenic concentrations in blood are shown in Table 6. The detection limit was 1 ug/L and a value
of one-half the detection limit was utilized when no arsenic was detected. No significant difference
(p<0.5) was found between the mean pre-experiment blood arsenic concentration of the twenty-seven
treated animals and the mean blood arsenic concentration of the six untreated animals. Subtraction
of the pre-experiment blood arsenic concentrations from each experimental observation eliminates
the effect of background concentrations on the determination of bioavailability. Elevated blood
EPA910/R-96-002
Page 21 of 48
-------
arsenic concentrations were not prolonged indicating that gastric emptying was not delayed. Blood
arsenic concentrations returned to pre-dosing levels by the last sampling at 144 hours.
Area-under-the-curve results of time versus blood arsenic concentrations are shown in Table 7.
Arsenic and Lead Concentrations in Urine
Arsenic and lead concentrations in urine are shown in Table 8 and Table 9. The detection limits were
2 ug/L for arsenic and 0.8 ug/L for lead.
EPA 910/R-96-002 Page 22 of 48
-------
Table 6. Arsenic Concentrations in Blood (micrograms per liter; ug/L)
pig
ID#
#126
#137
#195
#190
#133
#199
#140
#188
# 141
#182
# 138
# 196
#130
#131
# 136
# 139
# 145
#148
# 129
# 144
#150
#142
# 147
# 149
#192
#194
# 189
# 180
#197
#179
# 186
# 191
#184
#177
# 181
# 187
Treat-
ment
Group
1
1
1
1
2
2
3
3
8
8
9
9
4
4
4
5
5
5
6
6
6
7
7
7
10
10
10
11
11
11
12
12
12
13
14
15
Pre-Dosing
pre-1
0.5
1
4
1
1
134
1
1
0.5
1
2
1
0.5
0.5
0.5
2
0.5
1
1
1
1
1
0.5
2
5
2
*
1
1
2
1
0.5
*
2
1
2
pre-2
0.5
1
3
0.5
1
0.5
2
0.5
1
0.5
2
0.5
0.5
0.5
0.5
1
1
1
1
0.5
3
2
1
3
2
1
1
0.5
1
2
0.5
1
1
2
2
2
Hour Intervals After Dosing
0.25
0.5
2
3
0.5
0.5
25
0.5
0.5
14
496
111
2
0.5
0.5
0.5
1
0.5
2
0.5
2
3
1
2
7
2
2
3
3
3
3
7
*
2
2
2
3
0.5
0.5
2
2
0.5
5
86
32
1
12
369
89
125
0.5
2
0.5
5
1
5
5
5
3
1
4
10
0.5
9
10
20
13
7
35
23
8
7
2
*
1.0
0.5
1
3
0.5
6
89
45
1
*
318
97
0.5
1
5
0.5
9
3
9
22
9
8
11
13
43
3
15
24
48
48
28
77
37
20
32
2
4
1.5
0.5
1
3
1
7
90
47
0.5
7
254
94
0.5
2
7
1
21
16
16
32
14
36
19
19
51
17
28
53
100
60
67
118
125
44
49
6
5
3.0
1
0.5
3
0.5
8
99
62
1
6
183
76
0.5
9
15
4
52
39
37
57
40
80
50
55
56
78
83
87
231
180
11
177
224
58
83
10
10
6.0
1
3
4
2
5
137
55
1
4
88
50
2
12
14
16
34
37
37
58
50
80
93
84
83
136
125
49
290
211
249
243
423
196
82
10
11
12
1
2
4
2
3
44
27
2
2
42
28
1
4
6
4
12
*
13
15
12
24
46
37
36
52
53
22
109
111
49
90
142
78
26
8
9
24
1
2
2
2
2
9
6
2
1
12
9
0.5
1
2
0.5
3
5
4
3
4
5
8
7
10
10
11
8
16
20
14
20
22
14
5
4
3
48
1
2
2
2
1
3
2
1
1
6
4
1
0.5
0.5
0.5
2
1
2
0.5
1
4
3
2
5
4
3
3
4
5
6
4
6
4
2
2
2
72
1
2
2
2
0.5
2
3
0.5
1
3
2
0.5
0.5
1
0.5
2
0.5
3
0.5
1
3
0.5
1
3
3
1
3
4
4
2
3
3
4
2
2
2
96
0.5
1
1
1
1
1
1
2
1
3
1
1
0.5
0.5
0.5
2
0.5
1
0.5
1
3
1
1
2
3
2
4
2
3
12
2
3
4
2
2
4
144
*
1
2
0.5
0.5
1
3
1
0.5
3
2
*
0.5
1
0.5
1
0.5
2
0.5
1
5
0.5
1
3
2
1
1
2
2
3
2
1
2
0.5
1
1
SRMK
3
0.5
0.5
0.5
0.5
121
0.5
1
0.5
4
107, 0.5
148
147
0.5
10
120, 108
144
0.5
148
0.5
120, 1
0.5
0.5
0.5
0.5
148
147
0.5
0.5
0.5
0.5
160
2
0.5
0.5
140
Group #s: 1 = Neg Control; 2 = Low oral sol salt; 3 = High oral sol salt; 4 = Low oral soil; 5 = Med/Low oral soil; 6 = Med/hi oral soil; 7 = High oral soil; 8 = Pos control, low IV;
9 = Pos control, high IV; 10 = Low oral slag; 11= Med oral slag; 12 = High oral slag; 13 = Surface tailings; 14 = Subsurface tailings; 15 = Residential soil; * = Missing data;
H =Standard Reference Material (SRM) samples which were blinded and placed among the test blood samples
EPA910/R-96-002
Page 23 of 48
-------
Table 7. Arsenic Blood Area-Under-The-Curve
Results
Pig No.
141
182
138
133
199
140
130
131
136
139
145
148
129
144
150
142
147
149
192
194
189
180
197
179
186
191
184
177
181
187
Trmt Grp
8
9
2
3
4
5
6
7
10
11
12
13
14
15
Treatment
IV/Control
IV/Control
PO/Control
PO/Control
Soil
25
mg/kg
Soil
60
mg/kg
Soil
100
mg/kg
Soil
150
mg/kg
Slag
60
mg/kg
Slag
100
mg/kg
Slag
150
mg/kg
SurfTail
SubTail
Soil
Pig Wt (kg)
18
14
15
15.5
11
15.5
17
18.5
16
17
15.5
16.5
16
15.5
18.5
16
17
19.5
13.5
13
14
15
11
14
12.5
16.5
14.5
11
14.5
14.5
Total Dose
(ug)
163
4302
1657
129
3381
1707
681
741
641
1634
1490
1586
2563
2483
2964
3845
4085
4686
8192
7889
8496
15171
11125
14160
18964
25032
21998
12212
2630
2239
Total Dose
(mg/kg)
0.01
0.31
0.11
0.01
0.31
0.11
0.04
0.04
0.04
0.10
0.10
0.10
0.16
0.16
0.16
0.24
0.24
0.24
0.61
0.61
0.61
1.01
1.01
1.01
1.52
1.52
1.52
1.11
0.18
0.15
AUC
ug/ml/hr
18
1971
784
59
1310
625
85
146
82
338
291
400
440
385
709
846
773
837
1017
1117
144
2656
2365
2348
2332
3421
1752
655
36
144
EPA910/R-96-002
Page 24 of 48
-------
Table 8. Total Arsenic Eliminated in Urine (microsrams)
Animal# (Group #)
126(1)
137(1)
195(1)
190(1)
133(2)
199(2)
140(3)
188(3)
141 (8)
182(8)
138 (9)
196 (9)
130 (4)
131 (4)
136 (4)
139(5)
145 (5)
148(5)
129(6)
144(6)
150(6)
142 (7)
147(7)
149 (7)
192(10)
194(10)
189(10)
180(11)
197(11)
179(11)
186(12)
191(12)
184(12)
177(13)
181(14)
187(15)
Hour Intervals After Dosing
pre-24
29.6
26.4
67.8
57.3
40.5
29.0
67.7
36.9
84.0
25.7
72.1
87.5
52.2
43.1
10.5
32.3
37.4
21.4
40.8
26.8
33.5
38.7
75.1
80.2
7.1
44.8
30.0
83.5
73.0
89.9
3.5
60.5
53.0
63.2
0-12
31.0
32.6
39.1
31.6
90.6
1221.4
905.8
19.2
119.1
3306.0
882.0
36.1
340.0
242.8
849.2
588.3
723.9
1455.3
1305.3
1386.0
1517.0
2304.0
2453.4
2279.0
1142.4
4510.0
3152.5
3335.0
5050.5
4902.0
2759.0
1216.8
180.2
164.6
12-24
44.5
46.8
42.0
57.4
64.8
186.3
425.6
14.2
63.2
1090.4
323.1
45.2
121.2
224.6
223.4
122.6
442.0
418.0
462.0
585.0
680.8
717.3
515.4
451.5
1500.0
1423.5
1042.1
1706.4
2049.8
632.8
313.0
89.7
149.2
24-48
92.0
89.0
80.0
77.0
68.9
59.3
226.2
22.4
89.1
480.2
244.7
64.4
124.3
122.5
128.3
108.0
159.4
182.2
108.8
195.3
110.5
200.4
192.8
283.7
172.5
244.5
413.7
383.4
365.3
508.0
163.9
892.5
146.5
101.2
61.8
48-72
102.7
75.3
53.2
67.7
44.1
83.4
115.8
27.0
84.3
202.1
98.4
65.3
96.9
155.6
122.8
124.0
95.8
114.3
101.8
120.9
65.8
108.9
127.7
153.7
46.6
113.2
158.3
117.7
101.0
135.3
92.1
184.8
80.7
54.9
46.0
72-96
73.5
89.3
69.6
60.6
68.9
70.4
117.7
11.1
75.4
163.8
90.6
77.0
81.6
101.6
95.0
75.9
80.6
103.7
39.4
133.4
31.4
114.5
148.2
113.3
32.9
95.1
125.6
10.2
72.5
84.2
46.7
77.8
87.8
40.4
15.0
96-144
9.1
187.0
94.8
142.5
94.2
48.2
171.1
49.6
122.4
125.3
40.3
140.9
25.7
174.3
163.7
123.6
45.7
188.3
107.9
129.4
81.4
160.3
85.5
75.9
112.3
160.2
199.7
124.4
64.4
171.0
130.8
209.5
155.4
100.5
78.2
Total (ug)
352.9
520.1
378.7
436.8
431.5
1669.0
1962.2
143.5
553.4
5367.7
1679.1
428.9
789.7
1021.4
1582.3
1019.9
1228.0
2485.8
2081.1
579.0
2137.1
2686.2
3539.0
3797.3
3158.7
2206.9
6907.2
5211.7
4980.2
7655.4
7385.2
4756.5
2000.2
566.9
514.6
Empty cell indicates no sample was obtained.
EPA910/R-96-002
Page 25 of 48
-------
Table 9. Total Lead Eliminated in Urine (micrograms)
Animal# (Group #)
126(1)
137(1)
195(1)
190(1)
133 (2)
199 (2)
140 (3)
188(3)
141 (8)
182(8)
138 (9)
196 (9)
130 (4)
131 (4)
136 (4)
139(5)
145 (5)
148(5)
129(6)
144(6)
150(6)
142(7)
147(7)
149 (7)
192(10)
194(10)
189(10)
180(11)
197(11)
179(11)
186(12)
191(12)
184(12)
177(13)
181 (14)
187(15)
Hour Intervals After Dosing
pre-24
2.48
3.30
1.62
2.52
0.54
2.48
1.65
0.22
1.61
3.19
2.28
2.31
0.73
3.07
2.86
0.63
1.01
1.46
0.78
1.20
3.26
1.57
3.48
2.21
3.16
1.21
1.12
1.83
1.91
2.40
2.03
0.53
3.60
1.40
1.03
0-12
2.42
2.40
1.07
1.27
0.15
26.51
6.33
0.06
12.33
85.72
57.52
95.59
1.94
1.89
14.61
3.13
3.02
4.98
18.62
11.80
7.83
3.77
7.52
3.53
7.40
14.07
8.53
4.62
7.38
8.65
6.27
6.98
12.87
22.75
2.77
12-24
1.65
1.76
0.90
1.57
1.61
4.86
4.42
0.60
8.11
29.47
56.11
1.51
3.24
5.99
2.16
0.49
3.68
13.09
8.38
5.54
2.89
5.61
1.42
2.49
6.44
5.03
3.36
2.89
4.27
4.73
2.10
5.67
8.40
2.20
24-48
5.24
3.60
2.13
1.39
2.21
4.23
5.46
0.15
11.05
34.30
44.29
64.40
2.51
4.11
7.22
2.57
2.35
3.56
9.30
3.89
4.23
2.88
6.05
1.76
3.73
7.05
4.25
2.34
3.32
4.06
0.65
8.34
6.80
13.92
1.14
48-72
2.84
2.43
1.90
1.44
1.26
4.27
3.56
0.92
5.57
41.42
21.77
49.25
2.32
4.79
4.56
3.28
1.77
1.53
8.61
2.28
2.82
1.89
3.31
1.76
1.67
1.71
4.43
3.13
3.62
2.05
14.35
4.39
3.59
8.67
0.71
72-96
1.79
2.86
3.59
2.58
1.77
6.49
1.18
0.35
2.93
28.06
23.67
49.22
1.84
2.39
1.81
1.56
1.04
1.86
3.02
2.73
0.47
1.45
1.87
1.83
1.53
4.46
4.34
0.68
2.43
2.10
0.86
3.20
4.24
4.30
0.16
96-144
0.26
6.68
4.31
0.75
1.88
8.03
2.04
1.82
2.02
33.26
12.35
64.23
0.51
3.13
5.01
4.89
1.58
3.65
4.61
3.53
1.93
2.34
1.26
1.38
4.06
6.87
3.04
4.29
1.79
4.28
3.05
7.54
6.59
7.41
3.47
Total
ug in urine
14.19
19.72
13.89
8.99
8.89
54.38
23.00
3.91
42.00
189.07
378.78
10.62
19.54
39.19
17.57
10.25
19.26
57.26
32.60
22.82
15.22
25.62
11.68
20.87
40.60
29.61
18.41
21.43
25.40
29.90
32.55
39.75
65.44
10.45
Empty cell indicates no sample was obtained.
EPA910/R-96-002
Page 26 of 48
-------
Lead Concentrations in Blood
Lead concentrations in blood are shown in Table 1 0. In animals orally exposed to smelter soil or slag,
maximum blood lead concentrations were detected at 6 or at 12 hours (Tmax mean=8.5; sd=3.4;
n=20). In animals orally exposed to mine site substrates Tmax was at 6 hours except the animal which
received mine site soil. In this animal, Tmax occurred at 24 hours. The detection limits were 0.19
ug/dL, 0.04 ug/dL, 0.05 ug/dL and 0.09 ug/dL for total lead, 206Pb, 207Pb and 208Pb, respectively.
Elevated blood lead concentrations were not prolonged indicating that gastric emptying was not
delayed.
Analytically measured total blood lead concentrations are the sum of intravenously administered 206Pb,
background lead and lead absorbed from the environmental substrates. Because animals were
randomly assigned to treatment groups and dose substrates were administered on the basis of body
weight, the concentration of blood lead can be described by:
where
Cpb,t = total lead analytically measured, ug/dL
CPbbk = background lead concentration; average of pre-dosing concentrations, ug/dL
CPb 206 = intravenously administered 206Pb, ug/dL
CPbpo = perorally absorbed lead, ug/dL
The average abundance of 208Pb in the environmental substrates was 51.9% of all lead isotopes with
a standard deviation of 0.37%. The 208Pb abundance in blood lead concentrations of the four
untreated experimental animals averaged 50.7% of all lead isotopes with a standard deviation of
1 .67%. Variability in the 208Pb concentrations was due to the low concentration of lead in these blood
samples. The 208Pb abundance in the background blood lead concentrations approximated that of the
environmental substrates. Intravenous 206Pb was administered with the peroral materials at one tenth
or less of the estimated total lead dose . Negligible quantities of 208Pb were contributed by the 206Pb-
enriched dose because of the low abundance of 208Pb in the 206Pb solution and the low intravenous
dose.
Therefore, analytically measured blood 208Pb concentrations were attributable to absorption of 208Pb
from background sources and orally administered lead-containing substrates. Total lead contributed
by background sources and the perorally administered substrates can be calculated as follows:
+ P =1 Q97 * P
^Pb,po i.y^- 1 ^Pb208
is the analytically measured 208Pb concentrations and the constant of 1 .927 is the inverse
of the abundance of 208Pb in the environmental substrates (5 1 .9%). Even though the 208/206 isotope
ratios differ for the environmental substrates (see Table 4), the use of an average percent abundance
of 208Pb for all substrates does not significantly influence the outcome of the calculation because of
the relatively low standard deviation of 0.37% as compared to the standard deviations and coefficients
of variation in the blood lead concentrations. The intravenous 206Pb concentration is calculated as
follows:
EPA910/R-96-002 Page 27 of 48
-------
Cpb,206
The calculated total blood lead concentration (CPb calc) which includes lead from background sources
and lead from perorally administered substrates but not lead from intravenously administered 206Pb
is determined by subtraction:
r = r r
^Pb.calc M>b,t " ^Pb,206
Calculated total blood lead concentrations (CPbcalc) minus background are shown in Table 11.
Calculated intravenously administered 206Pb (CPb206) concentrations are shown in Table 12. The
coefficient of variation in area-under-the-curve of intravenously administered 206Pb was approximately
34% (mean = 25.6; sd=7.9; n=28) as shown in Table 21.
The relative contributions to variation in blood lead concentrations were evaluated. Analytical
measurement precision was evaluated by thirty-seven pairs of duplicate analyses. The average
standard deviation was 0.185 ug/dL, 0.057 ug/dL, 0.050 ug/dL and 0. 101 ug/dL for total lead, 206Pb,
207Pb and 208Pb, respectively. The coefficients of variation were 3.6%, 3.2% 5.3% and 5.0%,
respectively. Replicate analyses of calibration standards prepared from NIST SRM 981 Common
Lead with certified isotopic compositions were also performed. In accordance with counting
statistics, the isotope ratios for the lower concentration standard were more variable than those for
the higher concentrations. For analyses of the low standard (0.0005 mg/L; n=71), the coefficient of
variation was 6.24%.
Background blood lead concentration may contribute to variation. The variability of background lead
concentrations can be estimated from the blood lead results of the untreated animals. Table 13
presents a summary of these results. The average coefficient of variation of the background lead is
approximately 1 1.4%. For example, for a background lead of 2 ug/dL the standard deviation would
be 0.23 ug/dL.
Additional discussion of variation is contained in Appendix B, Evaluation of Sources of Variation.
EPA 910/R-96-002 Page 28 of 48
-------
Table 10. Blood Lead Concentrations (total analytically measured;
Hg/dl)
Pig ID#
126
137
195
190
133
199
140
188
141
182
138
196
130
131
136
139
145
148
129
144
150
142
147
149
192
194
189
180
197
179
186
191
184
177
181
187
Grp#
1
2
3
8
9
4
5
6
7
10
11
12
13
14
15
Trmt
Neg
Neg
Neg
Neg
POLo
POLo
POHi
POHi
IV Lo
IV Lo
IV Hi
IV Hi
Soil
25
mg/kg
Soil
60
mg/kg
Soil
100
mg/kg
Soil
150
mg/kg
Slag
60
mg/kg
Slag
100
mg/kg
Slag
150
mg/kg
SurfTail
SubTail
Soil
Hour Intervals After Dosing
Pre#l
1.37
1.87
1.48
1.02
1.46
1.41
0.50
1.32
1.69
1.47
1.23
0.97
2.23
1.92
1.71
1.87
1.22
2.31
0.75
1.55
1.57
1.35
2.13
1.76
2.05
1.26
2.09
1.19
0.96
0.71
2.14
1.15
0.92
Pre#2
1.49
2.20
1.35
0.90
1.62
1.00
1.31
0.55
1.34
1.57
1.49
1.22
0.97
1.67
2.11
2.34
1.87
1.38
1.86
0.90
1.60
1.73
1.53
1.97
1.38
1.58
1.13
1.16
2.02
1.43
0.82
0.73
1.04
1.51
1.11
0.88
0.25
1.43
2.09
1.28
0.97
2.35
2.14
2.37
1.08
12.32
103.50
87.95
292.26
1.49
2.76
2.50
2.78
2.22
2.29
2.74
1.42
2.31
2.17
2.13
2.80
1.88
2.04
1.65
1.47
2.39
2.00
1.46
1.90
2.19
1.67
1.13
0.5
1.35
2.10
1.65
1.04
2.22
6.62
2.61
1.18
12.48
80.88
68.09
2.51
2.49
2.47
2.47
2.57
3.35
1.71
3.14
2.60
2.15
2.76
1.86
2.09
2.21
1.42
2.47
2.54
1.51
1.31
2.27
2.35
1.78
1
1.36
2.52
1.30
1.01
2.61
12.95
2.87
1.34
12.46
63.43
55.07
154.77
1.69
2.75
2.35
2.37
2.32
2.28
3.48
1.81
2.23
2.55
2.67
1.74
2.10
3.03
1.53
2.53
2.32
1.72
1.37
3.09
2.35
1.76
1.17
1.5
1.17
2.41
1.40
0.93
2.20
15.47
2.78
1.61
11.90
51.05
41.55
117.72
1.72
2.60
2.58
2.51
2.45
2.24
3.27
2.16
2.33
2.69
1.87
3.09
2.05
2.09
5.45
3.05
2.61
2.18
1.75
1.63
4.24
2.75
2.92
1.25
3
1.31
1.95
1.40
1.01
2.29
19.12
3.06
4.06
10.10
37.38
46.45
74.46
1.73
2.79
2.45
3.89
2.42
2.40
3.47
3.94
2.62
2.76
2.39
3.53
1.86
2.65
7.88
2.26
3.02
2.63
2.74
1.88
6.10
7.62
8.99
1.40
6
1.13
2.03
1.59
0.86
2.11
22.50
3.51
8.77
10.00
26.59
22.92
45.18
1.86
2.85
2.43
7.15
2.61
2.88
5.19
8.59
3.67
4.25
2.65
4.00
2.50
3.99
8.80
3.08
3.94
4.11
3.78
3.33
7.41
8.51
16.04
1.84
12
1.18
2.10
1.21
0.96
2.24
14.26
3.95
5.67
7.80
19.99
14.67
30.18
1.96
3.10
2.30
6.42
2.67
2.66
5.63
6.49
3.92
3.61
2.81
5.34
2.86
3.92
6.60
2.54
3.03
3.70
3.02
2.37
6.22
6.62
12.56
1.96
24
1.20
1.93
1.05
0.97
1.80
10.67
3.01
3.54
5.23
15.80
10.33
23.79
1.67
2.65
2.16
4.90
2.17
2.44
4.26
4.35
2.78
2.85
2.44
3.82
2.12
3.41
4.51
2.13
2.60
3.42
2.50
2.20
4.97
5.30
10.21
1.88
48
1.09
1.88
0.92
0.86
1.44
6.71
2.11
2.05
3.41
12.00
7.11
15.28
1.40
2.32
1.66
3.47
1.83
1.69
3.13
2.65
2.09
2.22
1.79
2.71
1.63
2.39
2.63
1.31
1.89
2.82
1.51
1.41
2.34
2.68
5.27
1.52
72
1.05
2.32
1.13
0.94
1.19
4.27
1.71
1.52
2.44
9.06
6.60
11.34
1.48
2.07
1.60
3.00
1.39
1.73
2.51
1.91
1.69
1.76
1.38
2.16
1.21
2.21
1.78
1.24
1.51
2.53
1.20
1.22
2.06
2.18
4.33
1.30
96
1.10
1.70
1.26
0.84
1.02
3.25
1.47
1.23
2.03
7.97
4.75
9.81
1.34
1.71
1.29
2.49
1.31
1.04
1.97
1.54
1.47
1.49
1.77
1.91
1.42
1.97
1.48
0.90
1.64
2.07
1.09
1.05
1.63
2.01
3.19
1.28
144
1.68
1.46
0.93
1.04
3.25
1.24
1.08
1.70
7.88
4.04
1.18
1.60
1.79
2.30
1.71
1.00
2.05
1.27
1.52
1.52
1.27
1.56
1.52
1.79
1.62
1.10
1.49
1.99
0.99
1.08
1.94
1.90
2.49
1.49
Empty cell indicates no sample.
EPA910/R-96-002
Page 29 of 48
-------
Table 11. Blood Lead Concentrations (Pb,calc) jig/dl
PIG ID
126
137
195
190
133
199
140
188
141
182
138
196
130
131
136
139
145
148
129
144
150
142
147
149
192
194
189
180
197
179
186
191
184
177
181
187
Grp
1
1
1
1
2
2
3
3
8
8
9
9
4
4
4
5
5
5
6
6
6
7
7
7
10
10
10
11
11
11
12
12
12
13
14
15
Trmt
Neg
Neg
Neg
Neg
POLo
POLo
POHi
POHi
IV Lo
IV Lo
IV Hi
IV Hi
Soil
25
mg/kg
Soil
60
mg/kg
Soil
100
mg/kg
Soil
150
mg/kg
Slag
60
mg/kg
Slag
100
mg/kg
Slag
150
mg/kg
SurfTail
SubTail
Soil
Hour Intervals After Dosing
Pre 1
1.37
1.87
1.48
1.02
1.39
1.26
0.43
1.32
1.69
1.47
1.23
0.93
2.23
1.84
1.61
1.86
1.14
2.02
0.75
1.49
1.50
1.35
2.03
1.68
1.98
1.18
2.09
1.17
0.88
0.68
2.11
1.06
0.82
Pre 2
1.49
2.20
1.35
0.90
1.50
0.96
1.17
0.53
1.34
1.57
1.49
1.22
0.97
1.64
2.08
2.24
1.86
1.29
1.70
0.90
1.60
1.70
1.51
1.94
1.26
1.51
1.10
1.08
1.94
1.36
0.79
0.68
0.98
1.40
1.10
0.80
0.25
1.43
2.09
1.28
0.97
1.37
1.39
1.32
0.50
12.32
103.50
87.95
292.26
0.76
1.84
1.84
1.90
1.63
1.31
1.80
1.02
1.64
1.34
1.18
2.04
1.31
1.41
1.05
0.79
1.85
1.29
0.82
1.27
1.35
1.02
0.90
0.5
1.35
2.10
1.65
1.04
1.07
5.75
1.51
0.53
12.48
80.88
68.09
1.75
1.77
1.85
1.68
1.49
2.19
1.16
2.38
1.64
1.20
1.96
1.30
1.46
1.64
0.82
1.95
1.78
0.83
0.72
1.69
1.50
1.07
1
1.36
2.52
1.30
1.01
1.43
12.33
1.62
0.76
12.46
63.43
55.07
154.77
1.02
1.98
1.65
1.83
1.70
1.26
2.30
1.31
1.46
1.39
1.92
1.15
1.41
2.48
0.86
1.93
1.55
0.98
0.85
2.41
1.45
1.06
0.83
1.5
1.17
2.41
1.40
0.93
1.09
14.91
1.75
1.03
11.90
51.05
41.55
117.72
1.05
1.92
1.72
1.74
1.78
1.21
1.99
1.73
1.53
1.78
1.02
2.14
1.39
1.42
4.81
2.33
2.00
1.45
1.08
1.02
3.55
1.94
2.19
0.83
3
1.31
1.95
1.40
1.01
1.18
18.81
1.95
3.54
10.10
37.38
46.45
74.46
1.04
1.94
1.60
3.15
1.70
1.25
2.48
3.52
1.80
1.97
1.37
2.78
1.30
1.92
7.19
1.60
2.47
1.80
2.04
1.31
5.40
6.87
8.34
0.90
6
1.13
2.03
1.59
0.86
1.15
22.08
2.45
8.45
10.00
26.59
22.92
45.18
1.18
2.19
1.64
6.56
2.09
1.88
4.31
8.21
2.98
3.69
1.79
3.29
1.93
3.36
8.16
2.50
3.34
3.49
3.22
2.78
6.80
7.88
15.35
1.32
12
1.18
2.10
1.21
0.96
1.36
14.26
3.21
5.32
7.80
19.99
14.67
30.18
1.33
2.38
1.55
5.78
2.05
1.96
5.00
6.24
3.37
3.11
2.13
4.60
2.32
3.46
6.24
2.13
2.67
3.22
2.66
2.02
5.81
6.19
12.15
1.49
24
1.20
1.93
1.05
0.97
1.09
10.53
2.53
3.29
5.23
15.80
10.33
23.79
1.24
2.09
1.60
4.27
1.74
1.84
3.70
4.00
2.44
2.75
1.94
3.42
1.73
3.03
4.22
1.81
2.28
3.09
2.20
1.91
4.63
4.95
9.85
1.52
48
1.09
1.88
0.92
0.86
1.03
5.61
1.78
1.96
3.41
12.00
7.11
15.28
1.13
2.04
1.29
3.19
1.62
1.34
2.79
2.46
1.89
2.17
1.49
2.41
1.35
2.14
2.47
1.05
1.76
2.55
1.33
1.24
2.18
2.53
5.13
1.24
72
1.05
2.32
1.13
0.94
0.93
4.16
1.46
1.37
2.44
9.06
6.60
11.34
1.27
1.83
1.27
2.71
1.24
1.49
2.24
1.80
1.52
1.61
1.20
1.99
1.01
1.98
1.66
1.08
1.43
2.35
1.07
1.09
1.88
2.18
4.14
1.11
96
1.10
1.70
1.26
0.84
0.85
3.25
1.31
1.10
2.03
7.97
4.75
9.81
1.16
1.55
1.05
2.30
1.11
0.92
1.80
1.45
1.38
1.37
1.62
1.71
1.28
1.78
1.35
0.79
1.62
1.88
0.96
0.95
1.47
1.87
3.07
1.11
144
1.68
1.46
0.93
0.88
3.25
1.13
1.03
1.70
7.88
4.04
1.09
1.53
1.62
2.14
1.49
0.88
1.79
1.25
1.46
1.40
1.17
1.42
1.39
1.62
1.53
0.98
1.39
1.85
0.91
0.94
1.79
1.85
2.41
1.37
Empty cell indicates no sample.
EPA910/R-96-002
Page 30 of 48
-------
Table 12. Blood Lead-206 Concentration (206Pb) jig/dl
PIG ID
126
137
195
190
133
199
140
188
141
182
138
196
130
131
136
139
145
148
129
144
150
142
147
149
192
194
189
180
197
179
186
191
184
177
181
187
Grp
1
1
1
1
2
2
3
3
8
8
9
9
4
4
4
5
5
5
6
6
6
7
7
7
10
10
10
11
11
11
12
12
12
13
14
15
Trmt
Neg
Neg
Neg
Neg
POLo
POLo
POHi
POHi
IV Lo
IV Lo
IV Hi
IV Hi
Soil
25
mg/kg
Soil
60
mg/kg
Soil
100
mg/kg
Soil
150
mg/kg
Slag
60
mg/kg
Slag
100
mg/kg
Slag
150
mg/kg
SurfTail
SubTail
Soil
Hour Intervals After Dosing
Pre 1
0.1
0.15
0.1
0
0
0.1
0.1
0
0.1
0.29
0
0.1
0.1
0
0.1
0.1
0.1
0.1
0
0
0.1
0
0
0.1
0.1
Pre 2
0.12
0.04
0.14
0.02
0
0.03
0.03
0.1
0.01
0.09
0.16
0
0
0.03
0.02
0.03
0.12
0.07
0.03
0.08
0.08
0.07
0.03
0.05
0.06
0.11
0.01
0.08
0.25
0.98
0.75
1.05
0.58
0.73
0.92
0.66
0.88
0.59
0.98
0.94
0.4
0.67
0.53
0.95
0.76
0.57
0.63
0.6
0.68
0.54
0.71
0.64
0.63
0.84
0.65
0.23
0.5
1.15
0.87
1.1
0.65
0.76
0.72
0.62
0.79
1.08
1.16
0.55
0.76
0.83
0.95
0.8
0.56
0.63
0.57
0.6
0.52
0.76
0.68
0.59
0.58
0.85
0.71
1
1.18
0.62
1.25
0.58
0.67
0.77
0.7
0.54
0.62
1.02
1.18
0.5
0.77
0.96
0.75
0.59
0.69
0.55
0.67
0.6
0.77
0.74
0.52
0.68
0.9
0.7
0.34
1.5
1.11
0.56
1.03
0.58
0.67
0.68
0.86
0.77
0.67
1.03
1.28
0.43
0.8
1.16
0.85
0.95
0.66
0.67
0.64
0.72
0.61
0.73
0.67
0.61
0.69
0.81
0.73
0.42
3
1.11
0.31
1.11
0.52
0.69
0.85
0.85
0.74
0.72
1.15
0.99
0.42
0.82
0.91
1.02
0.75
0.56
0.73
0.69
0.66
0.55
0.83
0.7
0.57
0.7
0.75
0.65
0.5
6
0.96
0.42
1.06
0.32
0.68
0.66
0.79
0.59
0.52
1
0.88
0.38
0.69
0.79
0.86
0.71
0.57
0.63
0.64
0.58
0.6
0.62
0.56
0.55
0.61
0.63
0.69
0.52
12
0.88
0
0.74
0.35
0.63
0.72
0.75
0.64
0.62
0.7
0.63
0.25
0.55
0.56
0.68
0.74
0.54
0.46
0.36
0.41
0.36
0.48
0.36
0.35
0.41
0.43
0.41
0.47
24
0.71
0.14
0.48
0.25
0.43
0.56
0.56
0.63
0.43
0.6
0.56
0.35
0.34
0.5
0.5
0.4
0.39
0.38
0.29
0.32
0.32
0.33
0.3
0.29
0.34
0.35
0.36
0.36
48
0.41
1.1
0.33
0.09
0.27
0.28
0.37
0.28
0.21
0.35
0.34
0.19
0.2
0.1
0.3
0.3
0.28
0.25
0.16
0.26
0.13
0.27
0.18
0.17
0.16
0.15
0.14
0.28
72
0.26
0.11
0.25
0.15
0.21
0.24
0.33
0.29
0.15
0.24
0.27
0.11
0.17
0.05
0.18
0.17
0.2
0.23
0.12
0.16
0.08
0.18
0.13
0.13
0.18
0
0.19
0.19
96
0.17
0
0.16
0.13
0.18
0.16
0.24
0.19
0.2
0.12
0.17
0.09
0.09
0.15
0.15
0.2
0.14
0.19
0.13
0.11
0.02
0.19
0.13
0.1
0.16
0.14
0.12
0.17
144
0.16
0
0.11
0.05
0.09
0.07
0.17
0.16
0.22
0.12
0.26
0.02
0.06
0.12
0.1
0.14
0.13
0.17
0.09
0.12
0.1
0.14
0.08
0.14
0.15
0.05
0.08
0.12
Empty cell indicates no sample.
Table 13. Summary of Background Lead
Concentrations
in the Four Untreated Animals
Number of
Sample
Analyses
13
14
14
14
^Pb,bk
ug/dL
1.25
2.05
0.95
1.32
^Pb,bk
ug/dL
0.144
0.249
0.064
0.202
Coefficient of
Variation - %
11.5
12.1
6.73
Percent Recovery of Orally Administered Arsenic and Lead
EPA910/R-96-002
Page 31 of 48
-------
Mass balance estimates could not be accurately calculated due to the loss of significant volumes of
urine, and due to the failure of fecal analytical results to meet the quality assurance criteria for bias
(see Appendix A, Quality Assurance Audit Report).
Estimates of Unavailability
Bioavailability of arsenic could be estimated using a linear regression model passing through the
origin described the relationship between AUC and the dose (mg As/kg BW). Figures 1 A, IB, 1C
and ID illustrate the relationships, and Table 14 shows the results of the regression analyses.
The relationships are expressed as
AUCC = mc * Dosec (I)
and
AUC = m * Dose (2)
s s s
where m is the regression coefficient (slope), c the control value to which others are compared, and
s the soil or other environmental media for which bioavailability (F) is being estimated. Rearranging
equations (1) and (2) provides
mc = AUCc -H Dosec (3)
and
ms = AUCs -H Doses
The conventional bioavailability expression (Gibaldi and Perrier, 1982) is
Dose * AUC
F =
Dose * AUC
s c
Rearranging (5) produces
AUC + Dose
F = ° = (6)
AUC ^ Dose
c c
By substituting (3) and (4) in equation (6), F can be expressed as the ratio of slopes
m
F = _i
m
EPA 910/R-96-002 Page 32 of 48
-------
For each regression line, the 95% confidence limits for the slopes were determined as
m±t( .05 n-l)Sm ( 8 )
Where m is the slope, t(05 ^.^ is the two-tailed critical value for n-1 degrees of freedom, and sm is the
standard error of m. The confidence intervals are shown in Table 15.
The confidence limits for each group were used in Monte Carlo analyses of equation (7) to determine
confidence limits for F. The results of the Monte Carlo analyses are shown in Table 16.
Because multiple dose levels and replicates were not evaluated for the mining site environmental
substrates, bioavailability estimates would be highly uncertain. Area-under-the-curve results for the
individual animals are reported in Table 7.
Table 14. Blood Arsenic Regression Analyses Results
Treatment Group
Positive Control (i.v.)
Positive Control (p.o.)
Oral Soil
Oral Slag
in (slope)
6489
4424
3351
1826
r2
0.999
0.984
0.976
0.907
df
2
2
11
8
p-level
.00074
.00521
<.00001
.000021
Table 15. Slope and Confidence
Intervals of the Slope for Blood Arsenic
AUC vs. dose
Treatment Group
Positive Control (i.v.)
Positive Control (p.o.)
Oral Soil
Oral Slag
m (slope)
6489
4424
3351
1825
± 95%
762
1380
349
475
Table 16. Bioavailability Estimates for Arsenic
F = ms/mc
Control (p.o.yControl (i.v.)
Soil/Control (p.o.)
Soil/Control (i.v.)
Slag/Control (p.o.)
Slag/Control (i.v.)
Mean
0.68
0.78
0.52
0.42
0.28
95% Limits
Lower Upper
0.47
0.56
0.44
0.27
0.20
0.92
1.11
0.61
0.63
0.37
Median
0.68
0.76
0.53
0.41
0.28
EPA910/R-96-002
Page 33 of 48
-------
Figure 1. Dose response relationship between area-under-the-curve and arsenic
dose (jig As/kg BW). Regression line and 95% confidence intervals
of the regression are also shown.
A. Intravenous sodium arsenate (i.v. control)
B. Oral sodium arsenate (p.o. control)
C. Oral smelter site soil (soil)
D. Oral smelter site slag (slag)
i.v. control
0 0-05 0,1 0,15 0,2 0,25 0.3 035
2COO
1800
160C-
1JTO
E.I200-
J
' 1003
j
| 8M
600
150-
200-
p.o. control
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
Arsent Dow (nwj / IKI &fi}
- MO
soil
CK> 0.1 015 0? 025
Arsenic CJose (ng / kg BVtf)
slag
0 0.2 OJ 0.6 OS 1 1.2 1.4 1.6
Awnx now (mg^ltq BW)
EPA910/R-96-002
Page 34 of 48
-------
This experiment did not provide reliable bioavailability estimates for lead However, all animals
receiving an oral dose of an environmental substrate (n = 24) had a positive correlation coefficient
whereas the untreated animals (n = 4) had negative but nonsignificant coefficients at p<0.05. The
positive but nonsignificant correlation coefficients in a few treated animals may have resulted from
elevated background blood lead concentrations relative to the increase in blood lead due to
experimental exposures. Except for the instances noted above, oral exposures to the lead-containing
environmental substrates resulted in significant increases in blood lead concentrations.
Increases due to intravenous exposures to soluble lead acetate were significant. With subtraction of
background, a linear regression model passing through the origin described the relationship between
AUC and intravenously administered lead acetate (ug Pb/kg BW). Figure 2A illustrates the
relationship and Table 17 shows the results of the regression analysis. The use of data
transformations or alternate regression models did not result in improved correlation indices.
Table 17. Blood Lead Regression Analyses Results
Treatment Group
Positive Control (i.v.)
Oral Soil
Oral Slag
m (slope)
3.329
0.2152
0.1092
r2
0.916
0.05
0.04
df
3
11
8
p-level
.005
0.4
0.5
Figures 2B and 2C show the relationships between blood lead AUC and orally administered soil or
slag, respectively. Three of the four animals receiving oral doses of aqueous lead acetate had
significant increases in blood lead concentrations. However, Figure 2D shows that no systematic
relationship between dose and blood lead concentrations could be identified.
The confidence limits shown in Table 18 for each environmental substrate group were used in Monte
Carlo analyses of equation (7) to determine confidence limits for F of lead. The results of the Monte
Carlo analyses are shown in Table 19. Bioavailability estimates shown in Table 19 do not include
values of m less than or equal to zero. These estimates are included to demonstrate the analytical
method; however, the unreliability of these results is indicated by the inclusion of both zero and one
hundred percent in the 95% confidence interval.
Table 18. Slope and Confidence
Intervals of the Slope for Blood Lead
AUC vs. dose
Treatment Group
Positive Control (i.v.)
Oral Soil
Oral Slag
m (slope)
3.329
0.2152
0.1092
± 95%
1.995
0.8608
0.5468
Table 19. B
F = ms/mc
Soil/Control (i.v.)
Slag/Control (i.v.)
oavailabi
Mean
o.io§
0.04§
itv Estimates for Lea
95% Limits
Lower Upper
0
0
1.25
0.82
d
Median
o.n§
0.04§
§ Unreliable estimates which include both zero & 100% in the confidence interval.
EPA910/R-96-002
Page 35 of 48
-------
Figure 2. Dose response relationship between area-under-the-curve and lead
dose (jig Pb/kg BW). Regression line and 95% confidence intervals
of the regression line are also shown.
A. Intravenous lead acetate (upper left)
B. Oral smelter site soil (upper right)
C. Oral smelter site slag (lower left)
D. Oral lead acetate (lower right)
3500-
"3T
.1 3000-
S
| 2500-
O.
a 2°0
f 1500-
1000-
500-
0-,
250 n
0 100 200 300 400 500 800 700
Lead Intake Dose, (micrograms/Jiilogramj
0 50 100 150 200 £50
Lead Intake Dose (microgramaMogram)
*
regression
95% CL
95% CL
data
~ 250-,
I
t
,200-
a.
100-
0 100 200 300 400 500 600
Lead Intake Dose (microsrams/kilogram)
800
700-
"a 600-
£ 500-
a.
» 400-
300-
200
100-
s.
100 200 300 400 500 600 700
Lead Inlake Dose (miorograms/kilogfam)
EPA910/R-96-002
Page 36 of 48
-------
Because multiple dose levels and replicates were not evaluated for the mining site environmental
substrates, bioavailability estimates for lead and arsenic would be highly uncertain and were therefore
not determined. Area-under-the-curve results for the individual animals are reported in Table 20.
Table 20. Area-Under-the-Curve Results
for Calculated Blood Lead (Pb, calc) and Lead-206 (Pb, 206) Concentrations
PIG ID#
130
131
136
139
145
148
129
144
150
142
147
149
192
194
189
180
197
179
186
191
184
177
181
187
133
140
199
188
Grp#
4
4
4
5
5
5
6
6
6
7
7
7
10
10
10
11
11
11
12
12
12
13
14
15
2
3
2
3
Trmt
Soil
25
mg/kg
Soil
60
mg/kg
Soil
100
mg/kg
Soil
150
mg/kg
Slag
60
mg/kg
Slag
100
mg/kg
Slag
150
mg/kg
SurfTail
SubTail
TSoil
PO PbAc
PO PbAc
PO PbAc
PO PbAc
Dose
ug/kg
34
34
34
81
81
81
135
135
135
202
202
202
227
227
227
378
378
378
567
567
567
425
2464
217
17.47
214
263
656
AUC
(Pb,calc)
31.2
9.32
0
151.75
2.14
26.77
95.67
215.98
41.08
52.48
18.12
57.42
10.73
53.69
194.88
21.7
13.39
150.9
68.45
77.6
215.83
145.3
553.52
66.11
0
64
704
207
Tmax
(hr)
(Pb,calc)
12
12
6
6
12
12
6
12
6
12
12
12
12
6
6
6
6
6
6
6
6
6
24
12
6
6
Cmax
(ug/dL)
(Pb,calc)
0.38
0.44
4.63
0.24
0.74
3.14
7.35
1.82
1.87
0.7
2.62
0.85
1.71
7.06
1.37
1.32
2.23
2.38
2.09
5.82
6.13
14.27
0.71
2
21.13
7.97
T'/2
(hr)
(Pb,calc)
28
21
30
14
21
49
21
34
26
16
35
56
35
28
33
AUC
(Pb,206)
32.75
36.07
41.63
29.15
35.98
32.14
18.63
20.56
23.79
21.28
37.1
29.5
20.4
26.97
20.62
19.22
16.03
28.48
17.56
20.71
21.85
14.47
19.74
20.12
39
9
26
14
Tmax (hr)
(Pb,206)
0.25
0.25
1.5
0.25
0.5
1.5
1.5
0.5
3
1.5
3
1.5
1.5
3
3
1.5
1.5
3
1
1.5
3
1
1.5
3
0.5
0.5
1
0.5
Cmax
(ug/dL)
(Pb,206)
0.71
0.91
0.81
0.78
0.79
1.06
1.05
0.58
0.79
1.11
1.01
0.88
0.56
0.66
0.66
0.64
0.58
0.78
0.69
0.57
0.64
0.83
0.68
0.43
1.06
1.1
0.83
0.6
T '/2 (hr)
(Pb,206)
54
0.41
54
32
18
16
43
33
39
25
56
24
19
63
36
39
25
15
Cl
(dL/hr)
(Pb,206)
0.97
0.89
0.72
1.03
0.9
0.88
1.4
1.17
1.28
1.22
0.86
1.26
1.1
0.96
0.96
1.37
1.12
0.96
1.39
1.61
1.25
1.12
1.04
1.25
0.69
1.04
2.11
1.86
Empty cell indicates no value could be determined.
Semi-simultaneous intravenous administration of 206Pb-enriched solution and gavage administration
of test substrates also provides a means to estimate lead bioavailability as well as kinetic parameters.
Each animal orally exposed to smelter site or mining site environmental substrates also received a
simultaneous intravenous dose of 206Pb (1.68 micrograms of lead per kilogram of body weight).
Area-under-the-curve, the time of maximum concentration (Tmax) and maximum concentration (Cmax)
obtained by inspection of the data for blood 206Pb concentrations and are shown in Table 20. For
each animal, the pharmacokinetic parameter describing the rate of elimination, Kel, was obtained
EPA910/R-96-002
Page 37 of 48
-------
from the slope of the blood concentration (expressed as the natural logarithm, In) versus time
elimination curve. Half-life, T1/2, was obtained by application of equation (9).
T1/2 = In2/Kel
Clearance was obtained by application of equation (10)
(9)
Cl = Dose/AUC
(10)
Representative blood concentration (Pbcalc) versus time and blood concentration (206Pb) curves are
shown in Figure 3. The average AUC(0-96) was 24.3 ug.h/dL, the average Cmax was 0.8 ug/dL and
the average clearance was 1.1 dL/hr (Table 21). Average half-life of 206Pb was estimated to be 33
hours and the average 206Pb Tmax was 1.5 hr (Table 21). The delay in Tmax was observed consistently
in all animals.
Table 21. Kinetic Parameters for Blood Lead (Pb, calc)
and Lead-206 (Pb, 206)
Mean
SD
%CV
n
Oral
Tmax (hr)
(Pb,calc)
9
4
47%
26
T 1/2 (hr)
(Pb,calc)
30
11
38%
15
Intravenous
AUC
(Pb,206)
25
8
34%
28
Tmax (hr)
(Pb,206)
1.51
0.98
65%
28
Cmax (ug/dL)
(Pb,206)
0.78
0.19
24%
28
T 1/2 (hr)
(Pb,206)
33
17
51%
18
Cl (dL/hr)
(Pb,206)
1.16
0.32
27%
28
The mean clearance of intravenous lead acetate was 3.5 dL/hr (sd=0.5; n=4) and the mean Tmax was
0.3 hr (sd=0.1; n=4) (Table 22). Half-life of intravenous lead acetate was estimated using the
terminal slopes from 24-96 hours, 48-144 hours and 72-144 hours (Table 23). Figures 4A, 4B, 4C
and 4D show the lead elimination curves for individual animals receiving intravenous lead acetate.
Estimates of lead bioavailability from smelter site and mining site substrates shown in Table 24 were
obtained by application of equation (11)
(11)
Bioavailability (F) = AUC (oral} . Dose (i
AUC (iv). Dose (po)
Table 22: Kinetic Parameters for Blood Lead
(intravenous lead acetate)
PIG ID#
141
182
138
196
Mean
SD
%CV
n
Grp#
8
8
9
9
Trmt
IV PbAc
IV PbAc
IV PbAc
IV PbAc
Dose
ug/kg
18
263
214
656
AUC
(Pb,t)
282
1473
932
1870
Tmax (hr)
(Pb,t)
0.50
0.25
0.25
0.25
0.31
0.13
40%
4
Cmax (ug/dL)
(Pb,t)
12
104
88
293
Cl (dL/hr)
(Pb,t)
3.21
2.94
3.61
4.15
3.48
0.53
15%
4
EPA910/R-96-002
Page 38 of 48
-------
Table 23. Half-Life Estimates for
Intravenous Lead Acetate
PIG ID
141
182
138
196
Mean
SD
%CV
n
Grp
8
8
9
9
Trmt
IV PbAc
IV PbAc
IV PbAc
IV PbAc
T'/2(hr)
Calculated for the time period
shown
24-96 Hr
29
63
53
53
50
15
29%
4
48-144Hr
41
138
77
144
100
50
50%
4
72-144Hr
46
347
77
99
142
138
97%
4
Table 24: Unavailability Estimates
of Lead Based on
Stable Isotope Method
PIG ID
130
131
136
139
145
148
129
144
150
142
147
149
192
194
189
180
197
179
186
191
184
Grp
4
4
4
5
5
5
6
6
6
7
7
7
10
10
10
11
11
11
12
12
12
Trmt
Soil
25
mg/kg
Soil
60
mg/kg
Soil
100
mg/kg
Soil
150
mg/kg
Slag
60
mg/kg
Slag
100
mg/kg
Slag
150
mg/kg
Dose
ug/kg
34
34
34
81
81
81
135
135
135
202
202
202
227
227
227
378
378
378
567
567
567
Absolute
Bioavailability
4
1
0
12
0
2
7
12
3
2
1
2
1
2
6
1
1
2
1
1
4
EPA910/R-96-002
Page 39 of 48
-------
Figures. Blood Lead (Pbcalc) versus time and blood 206Pb versus time
relationship for an animal simultaneously administered an oral dose of
100 mg soil/kg BW smelter site soil and intravenous 206Pb (animal
#129).
3,5
!3
•n
1 1
•a
a
I0-5
Time, hours
Figure 4.
Blood lead elimination curves for
individual animals receiving
intravenous lead acetate. Half-life of
lead acetate was estimated using the
terminal slopes from 24-96 hour, 48-
144 hour and 72-144 hour time
periods.
A. Animal #141 (17.5 |ig Pb/kg BW) C.
B. Animal #182 (263 |ig Pb/kg BW) D.
Animal #138 (214 |ig Pb/kg BW)
Animal #196 (657 |ig Pb/kg BW)
EPA910/R-96-002
Page 40 of 48
-------
Half Lite D«teTtntaatlon
4 cases: r=-,985: p=,015
P1G_141 = 1.4M - .01 7 • TIME
Es!. t t/2 = 40.8 tit
(48- H 4 Sours)
IBS
TIME
= 2.5
2
$
1 b_
JO
Half-Life Dete rmtnaliofi
4ca8e8:r=-.S39Lp=,)6l
PIG 18? = 1.4W • .005 • TIME
Esi. I1»2= 138SOUIS
(49 -144 hours)
IBB
TIME
Half-Life Determination
4 caM9; rs-.96£: ps,03Q
PIG 136= 2.159- JX» = TIME
Est.t W2 = TThou
-------
DISCUSSION
This study provided site specific information useful for evaluating human exposures to arsenic and
lead contaminated soil, slag and tailings. Although this study demonstrated the challenges posed in
a simultaneous evaluation of two contaminants, it is not uncommon for these contaminants to occur
together in the environment. Arsenic and lead were absorbed into the blood following oral exposure
to the environmental substrates. The methods and results of the physical and chemical studies will
be discussed in a forthcoming report.
Dose rates were selected that would attain detectable blood lead concentrations. However, this
resulted in doses of arsenic greater than 1 mg/kg when slag and surface tailings were administered.
Due to potential toxicity, equivalent reference oral or intravenous arsenic dose could not be
administered. Therefore, a data evaluation methodology was developed to address these dose
differences.
The conventional bioavailability calculation method assumes a linear dose response that passes
through the origin. The data evaluation methodology developed for this study utilized the linear dose
response relationship observed between arsenic intake and blood concentrations. When multiple dose
levels are included in the study design, bioavailability may be estimated by this method when either
the dose (milligrams arsenic per kilogram body weight) or the response (area-under-the-curve) of the
group receiving the environmental media are within the experimentally observed range of the control
group. Although arsenic bioavailability has been previously studied, no studies have demonstrated
a linear relationship prior to applying the conventional calculation methodology.
Arsenic metabolism studies have commonly measured urinary concentrations due to the ease of
specimen collection, availability of analytical methods and observation that urine is the predominant
excretion route (for review see ATSDR 1993). Absolute bioavailability estimates for oral sodium
arsenate (mean=68%, CI=47-92%) derived from this study are comparable with estimates from
studies utilizing urinary data in humans and rabbits (Buchet et al. 1981a,b; Freeman et al., 1993).
It is readily observed that a linear dose response relationship passing through the origin assumes that
blood concentrations are nonexistent when substrate intake is zero. Therefore, the experimental
design and data evaluation methodology addressed the presence of endogenous background
concentrations in the blood samples. Bioavailability is overestimated if background concentrations
are not considered. If the origin is omitted from the regression model, the influence of background
concentrations are mistakenly double-counted, the regression coefficient is reduced and estimates of
bioavailability are overstated.
The experimental design and data evaluation methodology provided information to characterize
natural variability and uncertainty in the bioavailability estimates. In previous studies of arsenic
bioavailability from environmental substrates, confidence intervals were not evaluated (Freeman et
al., 1993).
Lead was bioavailable from all substrates studied. There was a higher degree of variability and
uncertainty in the bioavailability estimates for lead as compared to arsenic. The variability and
uncertainty of F is influenced by how well the regression model describes the dose-response
relationship, whether background concentrations had been adequately characterized, and if there were
an adequate number of experimental observations.
EPA 910/R-96-002 Page 42 of 48
-------
In this study, differing predose blood lead concentrations and postdose variances in animals orally
exposed to environmental substrates resulted in low correlation indices. Although analytical
measurement variability is incorporated in the coefficient of variation for background lead and 206Pb,
the data suggest that differences between individual animals in blood clearance of lead (as indicated
by 206Pb) and in background concentrations had more influence on variation in blood lead
concentration than did the analytical measurements.
When high variance is anticipated, a greater number of experimental animals per treatment group are
required. In this study, preliminary results did not indicate the magnitude of the variance eventually
observed in the lead results of the final study. Therefore, a significant limitation in this study for
estimating bioavailability of lead was the number of animals per treatment group in both the reference
and environmental substrate (soil and slag) groups. The results from the mine site soil or tailings
exposures suggested the need for a sensitive study design and protocol in the event of a future
bioavailability study.
It's interesting to note that arsenic concentrations measured in the same animals did not show the
same degree of variance. The variance in blood lead concentrations in this study could not be
explained with the available data.
To date, no other complete lead bioavailability study utilizing the immature swine model has been
published in the open literature. However, a high inter-animal variability was documented in a project
report. Although consistent increasing concentrations in blood lead were identified in animals
following dosing with lead-contaminated soil, investigators were unable to demonstrate statistical
significance using standard analysis of variance methods (Dupont, 1993). Like the present study, the
variance was high and the groups not large enough. Establishing treatment groups with minimal
baseline blood lead concentrations and small variance, increasing group sizes, and switching to a sub-
chronic dosing protocol have been approaches taken by one group currently active in this area of
research (Chris Weis, U.S.EPA Region 8, personal communication, 1995).
Confidence in the blood lead bioavailability estimates were low and were not recommended for use
in regulatory decision-making. The dose interval for lead ranged from approximately 600 to 4,000
micrograms (total dose) for soil, and from approximately 3,000 to 9,000 micrograms for slag. A non-
linear relationship between lead intake and blood concentrations has been described for children's
intake of less than approximately 1,000 micrograms per day (U.S.EPA, 1994; Sherlock and Quinn,
1986). The dose range and increments in this study were not sufficient to characterize a similar non-
linear relationship at low doses.
Bioavailability of lead was shown to be greater than zero percent by the significant increases in
individual animals' blood lead concentrations. However, the confidence intervals of the regression
coefficients (slope) included zero slope. A zero slope indicates that the same amount (total mass) of
lead was absorbed at each intake dose and that the absorbed amount was not dependent on intake
dose. That is, the percent of the total mass of lead absorbed from the dosed materials increases with
decreasing intake dose. Therefore, a slope (m) less than or equal to zero is considered an artifact of
the evaluation methodology and the variability observed in this study.
Bioavailability of lead is not a constant; therefore, steady-state blood concentration would not be
expected to be directly proportional to the amount of lead intake (Aungst et al., 1981). Even with
the modifications utilized in this study, the conventional calculation is inadequate to describe
bioavailability that varies with dose. Further refinement of the bioavailability algorithm is needed.
EPA910/R-96-002 Page 43 of 48
-------
The wide degree of natural intersubject variability is a well recognized phenomenon in human studies
of pharmaceutics' bioavailability. Crossover study designs are preferable to concurrent controls to
address this issue (Tse et al., 1991). However, variation in kinetic parameters can occur over time,
even within the same subject, primarily due to changes in clearance. The stable isotope technique
where isotopically characterized materials are simultaneously administered by different routes almost
completely eliminates the influence of intraindividual variability (Wolen, 1986).
Simultaneous administration of the intravenous stable isotope, 206Pb, and oral substrate permits study
of absorption characteristics isolated from elimination and distribution kinetics when the latter are
assumed to be identical for the stable isotope and the oral substrate. The statistical power of using
stable isotopes for evaluation of differences in absorption is superior to conventional crossover or
concurrent control designs. With only 4-6 subj ects in a treatment group, it's possible to detect a 20%
difference with a probability of 0.8, whereas other designs would require 8 or more animals. (Wolan,
1986). The number of animals per treatment group limited the power of the present study. Estimates
of F based on the stable isotope method ranged from 0 to 12% for soil and from 0.5% to 6% for slag.
These estimates are comparable to the mean estimates derived from the modified bioavailability
calculations.
In animal studies, half-life in blood of 8-25 days has been reported. In those studies, it was
demonstrated that bioavailability and kinetic parameters can be reliably determined from truncated
blood-time data (Aungst et al., 1981; Castellino & Aloj, 1964; Weis et al., 1993). Half-life estimates
developed from this study demonstrate the sensitivity of the estimate to the frequency of observations
during the terminal phase. Half-life estimates based on 72-144 hour data are comparable to previous
reports.
Findings of this study may challenge the assumption that the elimination and distribution of trace
doses of lead are identical to larger doses. The Cmax of intravenous lead acetate (dose range
approximately 300 to 8000 micrograms) occurred at the first sampling. In contrast, Cmax of
intravenous 206Pb (dose range approximately 20 to 30 micrograms) occurred at 1.5 hours. On a per
microgram basis, the apparent AUC for low doses (approximately 30 micrograms) is greater than for
higher doses (>300 micrograms). Delayed intravascular Cmax has been reported following intravenous
administration of tracer doses of radiogenic lead to humans (Chamberlain et al. 1978). And, a
secondary peak in plasma concentrations 2-4 hours following intravenous lead administration to
human volunteers has been observed (De Silva, 1981). Differential tissue affinities for lead have been
described where a greater percentage of the dose is present in extravascular tissues than in plasma
minutes (<1 hr) following intravenous administration (Bornemann and Colburn, 1985).
From the perspective of low-dose risk assessments for lead, similarities in tissue lead binding would
strengthen the extrapolation of laboratory animal/human dose-response relationships. An hypothesis
for the delayed Cmax could be high affinity saturable binding of lead in an extravascular tissue. Since
the intravenous dose was administered via the jugular vein, sites in the lungs may provide such
binding. Research on lead disposition and kinetics has not suggested a special affinity for lead in
pulmonary tissue. However, only the work cited above was identified as reporting blood
concentration profiles of trace doses. Tissue disposition studies have typically been conducted several
hours up to days or weeks following exposure. Saturable binding effects have been observed in
pharmaceutics studies utilizing the stable isotope method (Schmid et al., 1980). In pharmaceutics,
the presence of saturable binding results in decreased drug concentration at the therapeutic target
tissue. Lead binding proteins have been isolated from target tissues and may play a role in lead
EPA 910/R-96-002 Page 44 of 48
-------
toxicity (Fowler et al., 1993). The toxicological significance of saturable binding of lead in
extravascular tissue remains to be investigated.
The relative insensitivity of swine to clinical lead intoxication may suggest dissimilarities between
swine and other laboratory animals in the affinity of lead to blood components and the critical tissues
(Lassen & Buck, 1979; Osweiler et al., 1985). However, the time course of blood 206Pb
concentrations in this study may suggest similarities of extravascular tissue affinities for tracer-level
lead doses in immature swine and humans.
Blood clearance of intravenously administered lead acetate observed in this study was comparable
with previously published animal studies (Aungst et al., 1981). It is noteworthy that clearance of
lead from blood following intravenous administration increased as the dose increased. Further study
is required to determine if single dose kinetics are predictive of steady state and to verify dose
dependent clearance of low doses of lead. The disposition and clearance of lead at low intravenous
doses and comparison with disposition and clearance of low oral doses warrants further investigation.
Enhancing the scientific understanding of the behavior of low doses of lead in the body gains
importance as the blood lead concentration of concern becomes lower and the efficacy of abatement
of different exposure sources is debated.
EPA 910/R-96-002 Page 45 of 48
-------
REFERENCES
Agency for Toxic Substance and Disease Registry. (1993) Toxicological Profile for Arsenic. April
1993, U.S. Dept. of Health & Human Services, Atlanta, Georgia.
Aungst, B.J., Dolce, J.A. and Fung H-L. (1981) The Effect of Dose on the Disposition of Lead in
Rats after Intravenous and Oral Administration. Toxicol Appl Pharmacol 61:48-57.
Bornemann, L.D. and Colburn, W.A. (1985) Pharmacokinetics model to Describe the Disposition of
Lead in the Rat. J Toxicol Environ Health 16:631-639.
Boyajian, S. (1987) Bioavailability of Arsenic in a Refractory Matrix. Masters Thesis. School of
Public Health and Community Medicine, University of Washington, Seattle, WA.
Buchet, J.P., Lauwerys, R., Roels, H. (1981) Comparison of the Urinary Excretion of Arsenic
Metabolites After a Single Oral Dose of Sodium Arsenite, Monomethylarsonate, or Dimethylarsinate
in Man. Int. Arch. Occup. Environ. Health 48, 71-79.
Buchet, J.P., Lauwerys, R., Roels, H. (1981): Urinary Excretion of Inorganic Arsenic and Its
Metabolites After Repeated Ingestion of Sodium Metaarsenite by Volunteers. Int. Arch. Occup.
Environ. Health 48,111-118.
Castellino, N. and Aloj, S. (1964) Kinetics of the Distribution and Excretion of Lead in the Rat. Brit
JIndustrMed2\:3Q8.
Chamberlain, A.C., Heard, M.J., Little, P., Newton, D., Well, A.C. and Wiffen, R.D. (1978)
Investigations into lead from motor vehicles. Report of work at Environmental and Medical Sciences
Division, AERA, Harwell. HL78/4122 (C.10).
Chaney, R.L., Mielke, H.W. and Sterrett, S.B. (1989) Speciation, mobility, and bioavailability of soil
lead. Environ Geochem Health 11 (Suppl), 109-125.
DeSilva, P.E. (1981) Lead in Plasma: Its Analysis and Biological Significance. Thesis, Master of
Public Health, Commonwealth Institute of Health, University of Sydney, Australia.
DuPont Speciality Chemicals. (1993) Absolute bioavailability of soil lead in rats and microswine.
Haskell Laboratory Report no. 671-93. E.I. duPont de Nemours and Co., Haskell Laboratory for
Toxicology and Industrial Medicine, Newark, DE.
Freeman, G.B., Johnson, J.D., Liao, S.C., Feder P.I., Davis, A.O., Ruby, M.V., Schoof, R.A.,
Chaney, R.L. and Bergstrom, P.D. (1994) Absolute Bioavailability of Lead Acetate and Mining
Waste Lead in Rats. Toxicology 91(1994): 151 -163.
Freeman, G.B., Johnson, J.D., Killinger, J.M., Liao, S.C., Feder, P.I., Davis, A.O., Ruby, M.V.,
Chaney, R.L., Lovre, S.C., Bergstrom, P.D. (1992) Relative Bioavailability of Lead from Mining
Waste Soil in Rats. Fund Appl Toxicol 19:388-398.
EPA 910/R-96-002 Page 46 of 48
-------
Freeman, G.B., Johnson, J.D., Killinger, J.M., Liao, S.C., Davis, A.O., Ruby, M.V., Chaney, R.L.,
Lovre, S.C., Bergstrom, P.D. (1993) Bioavailability of Arsenic in Soil Impacted by Smelter Activities
Following Oral Administration in Rabbits. FundAppl Toxicol 21:83-88.
Fowler, B. A., Kahng, M.W., Smith, D.R., Conner, E. A., Laughlin, N.K. (1993) Implications of Lead
Binding Proteins for Risk Assessment of Lead Exposure. JExpo Analysis Environ Epidemiol 3(4),
441-448.
Gibaldi, M. and Perrier, D. Pharmacokinetics. Marcel Dekker, Inc., New York, 1982.
Griffin, S. and Turck, P. (1991) Bioavailability in Rabbits of Sodium Arsenite Adsorbed to Soils.
Toxicologist 11, Abstract 118.
Kalman, D.A., Hughes, J., van Belle, G., Burbacher, T., Bolgiano, D., Coble, K., Karle Mottet, N.,
and Polissar, L. (1990) The Effect of Variable Environmental Arsenic Contaminating on Urinary
Concentrations of Arsenic Species. Environ Health Perspectives 89:145-151.
Lassen, E.D. and Buck, W.B. (1979) Experimental Lead Toxicosis in Swine. Amer J Vet Res
40(10):1359-1364.
LaVelle, J.M., Poppenga, R.H., Thacker, B.J., Giesy, J.P., Weil, C., Othoudt, R., Vandervoort, C.
(1991) Bioavailability of lead in mining wastes: an oral intubation study in young swine. Chemical
Speciation Bioavail 3(3/4): 105-111.
Lorenzana, R.M., Beasley, V.R., Buck, W.B., Ghent, A.W., Lundeen, G.R. and Poppenga, R.H.
(1985a) Experimental T-2 Toxicosis in Swine I. Changes in Cardiac Output, Aortic Mean Pressure,
Catecholamines, 6-keto-PGFla, Thromboxane B2 and Acid-Base Parameters. Fund Appl Toxicol
5:879-892.
Lorenzana, R.M., Beasley, V.R., Buck, W.B. and Ghent, A.W. (1985b) Experimental T-2 Toxicosis
in Swine II. Effect of Intravascular T-2 Toxin on Serum Enzymes and Biochemistry, Blood
Coagulation and Hematology. FundAppl Toxicol 5:893-901.
Lovering, E.G., McGilveray, I.J., McMillan, I, and Tostowaryk, W. (1975) Comparative
bioavailabilities from truncated blood level curves. JPharm Sci 64:1521-1524.
Osweiler, G.D., Carson T.L., Buck, W.B. and Van Gelder, G.A. (1985) Clinical and Diagnostic
Veterinary Toxicology. Kendall/Hunt, Dubuque, Iowa.
Polissar, L., Ruston/Vashon Island Arsenic Exposure Pathways Study. (1987) University of
Washington School of Public Health and Community Medicine.
Polissar, L., Lowry-Coble, K., Dalman, D. A., Hughes, J. P., van Belle, G., Covert, D.S., Burbacher,
T.M., Bolgiano, D., Mottet, N.K. (1990) Pathways of human exposure to arsenic in a community
surrounding a copper smelter. Environ Research 53:29-47.
Schmid, J., Prox, A., Zipp, H. and Koss, F.W. (1980) The Use of Stable Isotopes to Prove the
Saturable First-pass Effect of Methoxsalen. BiomedicalMass Spectro 7(11-12):560-564.
EPA 910/R-96-002 Page 47 of 48
-------
Sherlock, J.C. and Quinn, MJ. (1986) Relationship between blood lead concentrations and dietary
lead intake in infants: The Glasgow Duplicate Diet Study 1979-1980. Food Additives and
Contaminants 3(2):167-176.
Tse, F.L.S., Robinson, W.T. and Choc, M.G. Study Design for the Assessment of Bioavailability and
Bioequivalence in Pharmaceutical Bioequivalence. Marcel Dekker, New York, 1991.
United States Environmental Protection Agency (1989) Exposure Factors Handbook. EPA/600/8-
89/043 Office of Health and Environmental Assessment, Washington, DC
United States Environmental Protection Agency (1992) Baseline Risk Assessment Ruston/No.
Tacoma Operable Unit Commencement Bay Nearshore/Tideflats Superfund Site, Tacoma,
Washington.
United States Environmental Protection Agency (1994) Guidance Manual for the Integrated
Exposure Uptake Biokinetic Model for Lead in Children. EPA/540/R-93/081 Office of Emergency
and Remedial Response, Washington, DC
Weis, C., Poppenga, R., Thacker, B. and Henningsen, G. (1993) Pharmacokinetics of Lead in Blood
of Immature Swine Following Acute Oral and Intravenous Exposures. Toxicologist 13(1):613.
Weis, C.P., Poppenga, R.L., Thacker, B.J., Henningsen, G.M., Curtis, A., Jolly, R., Harpstead, T.
(1994) Pharmacokinetics of soil lead absorption into immature swine following subchronic oral and
intravenous exposure. Toxicologist 14(1): 119.
Wolen, R.L. (1986) The Application of Stable Isotopes to Studies of Drug Bioavailability and
Bioequivalence. JClin Pharmacol 26:419-424.
EPA 910/R-96-002 Page 48 of 48
-------
APPENDIX A.
QUALITY ASSURANCE AUDIT REPORT
Chain of Custody records were adequately maintained.
1. Results for Standard Reference Materials (SRM). SRM samples indistinguishable from
experimental samples were included at an overall frequency of approximately 5%.
a. Blood Arsenic
i. Corrective action (repeated digestion and analysis) was initiated due to low
recoveries for arsenic in blood.
ii. Reanalyses indicated low to acceptable (within SRM control limits or 75-125%
where no limits exist) bias.
b. Blood Lead: Acceptable (within SRM control limits or 75-125% where no limits
exist) recoveries were observed.
c. Urinary Arsenic: Potential high bias for arsenic was indicated.
d. Urinary Lead: Acceptable (within SRM control limits or 75-125% where no limits
exist) recoveries were observed. A 0% recovery was reported for a single SRM with
a true value that was equal to the detection limit. This recovery was therefore not
evaluated. The frequency of SRM analyses was significantly low. As a result, a
representative analysis of bias could not be performed based upon SRM performance.
e. Fecal Analyses: Potential high bias for both arsenic and lead was indicated by the
finding of concentrations higher than the verified SRM concentrations.
2. RESULTS FOR INSTRUMENT CALIBRATION VERIFICATION STANDARDS.
a. All Arsenic Data: Inadequate frequency of standard analysis (a single standard was
analyzed at the beginning of each analytical run without subsequent standard analysis
during or at the end of each run). Acceptable frequency: The beginning and end of
the analytical sequence and 1 per 10 samples in between.
b. Blood Lead: Recovery acceptable (90-110%) only for mid-range (24-25 ug/dl)
standards. Greater recovery ranges were observed for lead standards at
concentrations of < 5 ug/dl.
c. Urinary Lead: Biased low in some animals and high in others.
d. Fecal Lead Analyses: Not Analyzed
EPA910/R-96-002 Appendices
-------
3. Results for Blank Analyses. In all matrices, for those blank samples which were analyzed, no
significant indications of lead or arsenic contamination were observed (i.e., sample
concentrations were > 5 times the amount observed in any blank).
a. The frequency of blank analyses made assessment of false positive or negative errors
near the detection limits difficult. Therefore, certainty near the detection limits is
unknown in many instances.
4. Results for Duplicate Sample Analyses. Duplicate sample analyses were conducted at a
frequency of 6-10%.
a. Blood Arsenic: Acceptable (+ 20% RPD) duplicate results were obtained for
concentrations at > 5-10 times the reported detection limit. Quantitative certainty
near the detection limit is indeterminate due to the inherent increase in analytical
variability. Below 5-10 times the detection limit, the %RPD criterion was not
evaluated.
b. Blood Lead: Acceptable (+ 20% RPD) duplicate results were obtained for
concentrations at > 5-10 times the reported detection limit. Quantitative certainty
near the detection limit is indeterminate due to the inherent increase in analytical
variability. Below 5-10 times the detection limit, the %RPD criterion was not
evaluated.
c. Urinary Arsenic: Acceptable (within + 20% RPD).
d. Urinary Lead: Overall acceptable (within + 20% RPD). RPDs of 21 % and 27% were
not believed to be significantly exceed the criterion for this matrix. As a result, data
was not qualified based upon these RPDs.
e. Fecal Analyses: Acceptable (within + 20% RPD) for all arsenic analysis. Overall
acceptability was observed for lead analysis with three observed exceedances (21%,
23%, 56% RPD). Because only one exceedance was significantly above the criterion
and all matrix spike recoveries for these samples were within the acceptance range,
data qualification was not required.
5. RESULTS FOR MATRIX SPIKE ANALYSES. Matrix spike sample analyses were
conducted at a frequency of 6-10%.
a. Blood Arsenic: Acceptable (within 75-125%). A matrix spike was performed twice
on one sample resulting in recoveries of 136% and 93%. As a result, data
qualification was not recommended.
b. Blood Lead: Not Analyzed.
c. Urinary Arsenic: Unacceptable recoveries occurred when indigenous urinary arsenic
concentrations were greater than five times the amount spiked. As a result, recoveries
for these samples were not applicable to this analysis.
d. Urinary Lead: Potential high bias was indicated.
EPA910/R-96-002 Appendices
-------
e. Fecal Analyses: Acceptable (within 75-125%).
6. Summary.
Potential bias may exist in the approximately 30% of the urinary lead data. Variable
instrument verification standard recoveries were observed for blood lead results at or below
5 ug/dl. Sample quantitation below this concentration may be uncertain. Instrument
performance with regard to fecal lead analysis is unknown. A relatively low frequency of
instrument verification standard and blank analyses represent deficiencies in the information
needed to support the useability of all arsenic data. The overall useability of ESA laboratory
data is recommended for preliminary decision making purposes only until confirmatory
(verification) analyses become available.
True Values for Blind Standard Reference Materials
Standard Concentration Standard Type
4.6 ug/dl
10.7 ug/dl
18 + 6 (33%) ug/dl
30 + 4 (8%) ug/dl
200 + 50 (25%) ug/1
150 + 30(20%)ug/l
109 + 4 (4%) ug/1
10 ug/1
37.5 ug/1
60 ug/1
82.4 ug/1
60 ug/1
480+ 100 (21%) ug/1
0.87 + 0.03(3%)ug/g
0.060 + 0.018 (30%) ug/g
Blood-Pb
Blood-Pb
Blood-Pb
Blood-Pb
Blood-As
Blood-As
Urine-Pb
Urine-Pb
Urine- As
Urine- As
Urine- As
Urine- As
Urine- As
Feces-Pb
Feces-As
EPA910/R-96-002
Appendices
-------
APPENDIX B.
EVALUATION OF SOURCES OF VARIATION
The error associated with the isotope ratios at low concentrations is important because the
intravenously administered lead and the background lead are at these levels. The calculation of the
intravenously administered lead concentration relies upon the isotope ratio. The average relative
standard deviation of 3.6% total lead and 5.3% for 208Pb when propagated as the sum of squares
would indicate the average relative standard for the isotope ratio would be 6.2% which is consistent
with that observed for the lower concentration standards of 6.3%. This agreement indicates that
these uncertainties have been adequately defined. The observed 6.3% relative standard deviation for
the ratio must be propagated with those for total lead and 208Pb to estimate the uncertainty associated
with the calculated intravenously administered lead values. This propagated uncertainty than must
be propagated with the total lead uncertainty to derive that of the sum of the background and the
perorally absorbed lead. Finally the uncertainty of the background lead must be propagated with this
later uncertainty to calculate the uncertainty of the perorally absorbed lead.
The variability of the background lead for the duration of an experiment can be estimated from the
negative control experiments. The negative control experiments did not, however, capture all the
variability of the background lead. A number of substrate dose experiments indicate systematic
sloping of the background lead. This phenomenon is indicated by the difference in the "pre 1" and
"pre 2" values compared to the specimen values near the end of an experiment. Experiment 197 is
an example of this phenomenon where the "pre 1" and "pre 2" values were 2.11 and 1.93 ug/dL,
respectively, while the 144 hour specimen was 1.39 ug/dL. Subtraction of the average of the "pre"
values of 2.02 ug/dL could result in a systematic error of as much 0.63 ug/dL. In summary, the
subtraction of the background could contribute from about 0.10 to as much as 0.25 ug/dL random
error (standard deviation) and on a number of experiments perhaps as much as an additional 0.63
ug/dL systematic error. Interpolation of the sloping background might reduce the systematic error
component.
Now that the uncertainties have been estimated, propagation of these errors for the maxima values
for the high oral slag experiments will be used to examine the analytical contribution to the total error.
It should be noted that this example did not have sloping backgrounds as the uncertainty or perhaps
bias associated with this phenomenon is not really an analytical uncertainty. Table 3-7 contains data
extracted from Table 3-4 for the high oral slag experiments maximums.
Table 3-7 High Oral Slag Lead in Blood Maxima Data
Specimen Specimen Specimen
184-B8 186-B8 191-B8
Parameter ug/dL ug/dL ug/dL
Measured Pb 7.41 3.78 3.33
Measured Pb208 3.53 1.67 1.44
Calculated Pblv 0.61 0.56 0.55
Calculated Pbpo+bk 6.80 3.22 2.77
Measured Pbbk 0.98 0.835 0.68
Calculated Pbnn 5.82 2.38 2.09
EPA910/R-96-002 Appendices
-------
The total uncertainty of concern is calculated from the three perorally absorbed values. The
average value for the three specimens was 3.43 ug/dL with a standard deviation of 2.07
ug/dL. The variance is 4.3. The variance associated with each measured and calculated
parameter for each specimen is given in the following table.
Table 3-8 Analysis of Variance for Each High Oral Slag
Lead in Blood Maximum
Specimen Specimen Specimen
Parameter 184-B8 186-B8 191-B8
Parameter (ug/dL)2 (ug/dL)2 (ug/dL)2
Measured Pb 0.0712 0.0185 0.0144
Measured Pb208 0.0350 0.0078 0.0058
Calculated Pblv 0.3848 0.0887 0.0666
Calculated Pbpo+bk 0.4559 0.1072 0.0809
Measured Pbbk 0.0125 0.0091 0.0060
Calculated Pbno 0.4684 0.1163 0.0869
Summing the variances for each parameter for the three specimens provides the variance for
the average. Table 3-9 presents these summed variances for each parameter and gives the
percentage of the total variance represented by each parameter.
Table 3-9 Analysis of Variance for the Average High Oral
Slag Lead in Blood Maxima
Variance Percent of
Parameter (ug/dL)2 Observed Variance
Measured Pb 0.1040 2.42
Measured Pb208 0.0487
Calculated Pblv 0.5401 12.5
Calculated Pbpo+bk 0.6441
Measured Pbbk 0.0276 0.64
Calculated Pbpo 0.6717 15.6
Observed Pbno 4.305
This analysis indicates that the analytical error might represent about 16% of the total
uncertainty. The analysis of variance indicates a large percentage of the analytical variance
is associated with the derivation of the intravenous tracer concentration. However, the
standard deviation observed between the three swine for the tracer maxima was only 0.032
ug/dL which gives a variance of only 0.0010. This observed variance is much less than that
calculated from the propagation of errors given above. The propagation of error technique
assumes the measurements are independent while the measured total lead and measured Pb208
are not independent and were determined at the same time. Using the observed variance, the
EPA910/R-96-002 Appendices
-------
analytical variance for the perorally adsorbed lead changes from 0.6717 to 0.1327 which
represents just 3.1% of the overall variance observed. Furthermore, if the uncertainty
associated with the isotope ratios is that of the substrates of 1.67% rather than the 6.3%
relative standard deviation used, analytical uncertainty would be an even smaller component
of the total experimental uncertainty.
It should be noted that a few groups of other substrate dose experiments showed
considerable more variability for the tracer. For example, the medium high soil experiments
had tracer maxima of 1.28, 0.55, and 0.82 ug/dL. The variance was 0.6075 which is similar
to that calculated by the propagation of error. However, the lead/time curve for the tracer
for each experiment takes on a shape of a peak, indicating that the observed variance of the
maxima is not due to analytical variance. If the variance was analytical in nature than one
would expect a single point spike as opposed to a multiple point peak.
Finally, for the example selected, the total uncertainly for the perorally absorbed lead
expressed as a standard deviation was about equal to the average. For many of the other
substrate dose experiments the standard deviation was greater than the average. Although
it is evident, both intuitively and statistically, that lead was perorally absorbed from the
substrates, this type of variability limits the ability to distinguish differences in the amount of
lead perorally absorbed at different dose levels. Distinction of these differences would be
aided by a greater number of swine per substrate dose level. Moreover, if swine could be
selected that had lower and less variable background lead, similar to negative control swine
190, then fewer number of swine per dose level would be needed to discern a difference.
EPA910/R-96-002 Appendices
------- |