Proceedings of the ISEA Bioavailability Symposium
              Durham, North Carolina

Use of In Vitro Bioaccessibility/Relative Bioavailability
     Estimates for Metals in Regulatory Settings:
                  What Is Needed?
                     Prepared for:
           U.S. Environmental Protection Agency
       Technical Workgroup Bioavailability Committee
                     October 2007

-------
Table of Contents
Panel Members	2
Introduction	3
Proceedings	4
Panel Discussion Session 1	9
Panel Discussion Session II	18
Summary	20
Appendix A: Questions for the Panel (submitted in advance)	22
Appendix B: Abstracts	25
Appendix C: Presentations	26
Appendix D: Agenda
Panel Members
       Nick Basta, Ohio State University
       Karen Bradham, U.S. Environmental Protection Agency; (co-chair)
       Mike Beringer, U.S. Environmental Protection Agency
       Stan Casteel, University of Missouri
       Mark Cave, British Geological Survey (co-chair)
       Heather Jamieson, Queen's University
       Brian Laird, University of Saskatchewan
       Yvette Lowney, Integral Consulting
       Agnes Oomen, Rijksinstituut Voor Volksgezondheid En Milieu
       Pat Rasmussen, Health Canada; (co-chair)
       Sohel Saikat, UK Environment Agency
       Kirk Scheckel, U.S. Environmental Protection Agency
       Rosalind Schoof, Integral Consulting, Inc.; (co-chair)
       Dave Thomas, U.S. Environmental Protection Agency
       Joanna Wragg, British Geological Survey
       Aaron Yeow, U.S. Environmental Protection Agency

       Notes recorded and draft prepared by Mark Follansbee, Sycracuse Research Corporation,
       Environmental Science Center

Abbreviations
%           Percent
ABA        Absolute Bioavailability
As           Arsenic
BARGE      Bioaccessibility Research Group of Europe
BGS         British Geological Survey
Cr           Chromium
Cu           Copper
DMA        Dimethylarsenic
DQO        Data Quality Objective
g            gram
g/kg         gram per kilogram
                                    Page 2

-------
GI           Gastrointestinal
HERA       Human and Ecological Risk Assessment
HF          Hydrofluoric acid
HHRS       Human Health Risk Assessment
ICCVAM    Interagency Coordinating Committee for Validation of Alternative Methods
IAEA        International Atomic Energy Agency
ISEA        International Society of Exposure Analysis
IVG         In vitro gastrointestinal
k            thousand
jig           microgram
mg/kg       milligram per kilogram
MMA       methylarsonic acid
NaAS        Sodium arsenate
Ni           Nickel
NIST        National Institute of Standards and Technology
OSWER     Office of Solid Waste and Emergency Response
Pb           Lead
PBET        Physiologically Based Extraction Technique
ppm         parts per million
PRG         Preliminary Remediation Goal
Q&A        Question and Answer
RA          Risk Assessment
RBA         Relative Bioavailability
RIVM       Rijksinstituut Voor Volksgezondheid En Milieu
SBRC       Solubility/Bioavailability Research Consortium
SERDP      Strategic Environmental Research and Development Program
SOP         Standard Operating Procedure
TIM         TNO Gastrointestinal Model
TRV         Toxicity Reference Value
UEF         Urinary Excretion Fraction
UK          United Kingdom
US EPA     United States Environmental Protection Agency
XANES      X-Ray Absorption Near-Edge Spectroscopy
XRD         X-Ray Diffraction
Zn           Zinc
Introduction
Oral ingestion of soil and dust is a key pathway for human exposures to metal and metalloid
contaminants. It is widely recognized that the site-specific bioavailability of metals in soil and
dust may be reduced relative to the metal bioavailability in media such as water and food, and
adjustments for oral relative bioavailability are becoming more accepted. Both animal models
and in vitro bioaccessibility models have been used to estimate relative bioavailability of metals
in soil and dust.  Although animal models are often considered the "gold standard", they may be
costly or otherwise prohibitive at certain sites and may not be sensitive enough to test
                                     Page3

-------
environmentally relevant samples for all contaminants. Routine application of in vitro metal
bioaccessibility models in regulatory settings is being held up by different perceptions of what is
required of these models in terms of validation.

This symposium provided the opportunity for international experts to exchange their views on
methods for assessing relative bioavailability/bioaccessibility for application in risk assessments
at contaminated sites.  The symposium speakers presented recent developments in animal
models, new in vitro models, the role of mineralogical analyses in assessing relative
bioavailability, and the application of physiologically based models as research tools.  In
addition, two panel discussions addressed specific research questions and discussed future
research needs in this area. Recognizing the multi-disciplinary nature of exposure assessment,
this symposium included representation from many disciplines including risk assessment,
toxicology, environmental  geochemistry, geology, soil, and analytical chemistry from the U.S.,
Europe, and Canada.
Proceedings

Introduction and Overview
K. Bradham; U.S. Environmental Protection Agency, Research Triangle Park, NC
M. Beringer; U.S. Environmental Protection Agency, Kansas City, KS.
A. Yeow; U.S. Environmental Protection Agency, OSRTI, Washington, DC
P. E. Rasmussen; Health Canada, Ottawa, ON, CANADA.
R. A. Schoof; Integral Consulting, Inc., Mercer Island, WA.
M. R. Cave; British Geological Survey, Nottingham, UNITED KINGDOM.

       Karen Bradham provided definitions and background for the discussions. Oral
bioavailability of metals is important for Human Health Risk Assessment (HHRA) of metals and
decision-making. Absorption from the gastrointestinal tract (GI) depends on the metal,
chemical, and physical form, as well as biological factors.  Karen acknowledged the many
definitions of bioavailability, but she asked attendees to consider the definition in the Guidance
for Evaluating the Oral Bioavailability of Metals in Soils for Use in Human Health Risk
Assessment (OSWER 9285.7-80): "the fraction of an ingested dose that crosses the
gastrointestinal epithelium and becomes available for distribution to internal target tissues and
organs"  Absolute bioavailability (ABA) and relative bioavailability (RBA) are also important
definitions. ABA: ratio of the amount of metal absorbed compared to the amount ingested.
RBA: the ratio of the bioavailability of a metal in one exposure context (i.e., physical chemical
matrix or physical chemical form of the metal) to that in another exposure context.
Bioaccessibility is a related term, typically referring to a measure of the physiological solubility
of the metal at the portal of entry into the body. Bioavailability information can be used to
improve the accuracy of risk calculations and inform decisions at hazardous waste sites. Karen
provided an overview of the three forms of studies: in vivo, in vitro, and mineralogical/speciation
studies. Highlights from the 2006 ISEA symposium: general agreement that criteria are needed
to assess in vivo and in vitro methods; some countries are  allowing limited site-specific
adjustments; and few are comfortable with allowing existing site-specific methods to be
                                      Page 4

-------
universally applied.  Karen reviewed the panel discussion questions for the audience, which were
disseminated to the panelists prior to the symposium.

Q&A: No time for questions
Evaluating the Unavailability of Soil-Borne Contaminants at Waste Sites
M. Beringer; U.S. Environmental Protection Agency, Kansas City, KS.
A. Yeow; U.S. Environmental Protection Agency, OSRTI, Washington, DC.

       Mike Beringer discussed EPA's new guidance on using bioavailability studies to gather
site-specific information at Superfund sites. EPA developed this guidance because there was a
need for a consistent basis for approaching sites and evaluating new bioavailability methods.
The bioavailability guidance is limited in scope to oral ingestion of metals at hazardous waste
sites for HHRA. The assessment of bioavailability is consistent with existing EPA guidance.
The new guidance provides a decision framework for collecting site-specific bioavailability
information to make quantitative adjustments. The guidance also recommends using a validated
methodology and provides recommended evaluation criteria for the development of new
methods. The recommended evaluation criteria are based on the Interagency Coordinating
Committee for Validation of Alternative Methods (ICCVAM) criteria for method validation  and
regulatory acceptance. The ICCVAM criteria are widely accepted  internationally.  Mike noted
that the evaluation criteria are not all critical to the validation or acceptance of a methodology.
Mike highlighted the importance of establishing a correlation between an in vitro method and an
in vivo method. Mike provided an overview of the methods for evaluating lead bioavailability
and bioaccessibility. The juvenile swine model  and the in vitro bioaccessibility method for lead
have been evaluated and accepted for site-specific HHRA by EPA using the ICCVAM criteria.
These methods are described in a companion document to the guidance, referred to as the Lead
Technical Support Document.  Mike noted that the evaluation  covered a broad range of metal
forms and a range of bioavailability values. The documentation also outlines some limitations of
the accepted methodology.  Mike listed some future activities to include support for
implementation of the guidance and evaluation of arsenic methodologies.

Q&A
          What was the in vitro method?
              o  Mike responded that the in vitro assay published by John Drexler was
                 evaluated (Drexler and Brattin, 2007).
          For the in vivo bioavailability: in vitro bioaccessibility correlation, how did you
          derive the confidence interval?
              o  Mike responded that this was a prediction interval, not a confidence interval.
          Do you recommend using the prediction interval in the RA?
              o  Mike responded that the EPA recommendation suggests that one should use
                 the best-fit line.
          What is the mechanism for using new methods that are similar to the existing
          method? Do we need to show equivalence or go through the entire process?
                                     Page 5

-------
                 Mike responded that the method described in Drexler and Brattin (2007) is
                 recommended because it has satisfied the ICCVAM criteria and suggested that
                 we discuss the process for similar methods later in the symposium.
Method Development and the Application of Oral Unavailability Data in U.S. Risk
Assessments
R. A. Schoof; Integral Consulting, Inc., Mercer Island, WA.

Rosalind Schoof posed several (5) questions related to EPA's new guidance on bioavailability.
       1. How is validation defined?
       2. What are validation requirements?
       3. What are regulatory acceptance requirements?
       4. What is the process for method development and validation?
       5. What methods are already considered validated based on history of use?

       Rosalind expressed concerns about how strictly the validation process will be defined.
Validation requires an understanding of relevance and reliability. Rosalind noted that the
ICCVAM criteria were developed for evaluating the alternative toxicity models, not for
bioavailability.  These may or may not necessarily be  applicable to bioavailability methods.
Rosalind provided an overview of the process leading from research to acceptance and
implementation. Validation occurs when its performance characteristics, advantages, and
limitations have been adequately documented for a specific purpose.  Rosalind identified three
types of tests: definitive, screening, and adjunct. Are bioavailability test methods adjunct?
Rosalind offered that some test methods  might be considered validated based on history of use
(e.g., in vivo pharmacokinetic studies of bioavailability). Rosalind noted that validated studies
still require scientific interpretation and oversight.  GI physiology differs among species, so
animal models are predictive, but not perfect. Rosalind expressed concerns related to her
perception that EPA is going to require validation for  all metals. A number of other metals that
are of concern in soil, are not typically found at high enough concentration in soil to allow for
animal study and may therefore not be evaluated using the EPA process (cost of method
validation is too high).  Rosalind ended with four questions:
       1. How reliable are the in vivo methods?
       2. Are the in vivo methods considered validated?
       3. Should validation of in vitro protocols be required on a metal-specific basis?
       4. Should results of methods that have not been validated be considered in risk
       assessment? If not then they may not be developed using site money. If so how?

Q&A
          Related to in vivo testing of other metals, I agree that concentrations are too low to
          establish a correlation. Can you  comment  on this limitation?
             o  Rosalind agreed that this  would be  difficult for metals other than lead and
                 arsenic.
                                      Page 6

-------
          The concept of saying that concentrations are too low is not necessarily true.  We can
          develop methods to assess low concentrations. The document EPA developed is not
          universal to all metals.  We need to examine relevant samples.
             o   Rosalind agreed and responded that the concern is high background in diet
                 and other sources to distinguish additional exposure from site soil to detect
                 site exposure.
Arsenic Bioaccessibility Testing Using Various Extraction Methods: Results and
Relation to Relative Oral Bioavailability as Measured in the Cynomolgus Monkey
Y. W. Lowney; Exponent, Boulder, CO.
S. Roberts; University of Florida, Gainesville, FL.
S. Saikat; UK Environment Agency, Oxfordshire, UNITED KINGDOM.

       Yvette Lowney discussed the in vivo bioavailability method using the Cynomolgus
monkey model and in vitro methods using a variety of methods.  The Cynomolgus monkey
model was used because the oral bioavailability of the monkeys tends to correlate well with
humans.  Excretion of As into urine is relatively rapid—within 48 hours nearly all of the As has
been excreted.  The Cynomolgus monkey studies used a low arsenic diet: Soil dose <1 g/kg and
As dose <1 mg/kg. Five animals were used to develop an RBA estimate. Urinary and fecal
recoveries were very good. Controls (high and low RBA) were used. Sample RBA ranged from
5% to 31%. No RBA was found to be higher than 32%.  Soluble NaAs was used and found to
have an RBA of nearly 100%.  The study (published by Roberts et al., 2007) used a variety of
soil types. In vitro model development began with the Solubility/Bioavailability Research
Consortium (SBRC) method, with some alterations and consideration of The National Institute
for Public Health and the Environment (RIVM) and Bioavailability Research Group of Europe
(BARGE) methods.  The SBRC method had good correlation for some  samples, but not all (even
when different pH levels were used). It could not be determined why all soils did not have good
correlation. Phosphate additions improved the correlation, but did not work for all soil samples.
Other alterations to the method were made, but even with hydroxylamine additions correlation
was not good enough. When the RIVM and unified BARGE methods were used, none of these
methods worked (correlated) well for all soil samples. There is a need to develop a method that
is predictive for all soils or to understand which soils will work with the method. Available
methods do not provide a 1:1 correlation, but may be predictive. Yvette noted that the soil
samples have been extensively studied: mineralogy, speciation, etc.  Mathematic modeling may
not necessarily work; however, the factors that control arsenic bioavailability can be stated.
Research status: data suggests RBA <30% from in vivo bioavailability testing, need more work
for the in vitro model.

Q&A
          What is going on with the soils that affect the correlation? We need to define the
          parameters for when soil samples will and will not work with the method.
              o   Yvette responded that she agrees; however, soil sample source (e.g., mine
                 tailings or orchard samples) and mineralogy cannot explain correlation.
          Were the monkeys fasted?
              o   Yvette responded that the monkeys were fasted overnight.
                                     Page 7

-------
          Can you give background as to why correlation is the standard as opposed to
          predicting or over predicting in vitrol
              o  Yvette responded that she agrees that an in vitro assay either predicts or over
                 predicts (slope of the line is flatter than it should be). The objective is to have
                 the in vitro assay predictive of bioavailability.
          Have you considered reabsorption of the arsenic? Should you consider a different
          method, perhaps to add some resin beads to absorb the arsenic?
              o  Yvette agrees that reabsorption and precipitation could be a problem. Yvette
                 noted that the Hawaii volcanic soil sample had very low recovery.
Assessing Bioavailability Using the Swine Model
S. Casteel; University of Missouri, Columbia, MO.
G. Pent; University of Missouri, Columbia, MO.
C. Weis; US EPA, Denver, CO.
W. Brattin; Syracuse Research Corp, Denver, CO.

       Stan Casteel gave an overview of the swine model (in vivo bioavailability) used for Pb,
As, Cd, and Cr. Juvenile swine were used as a surrogate for children. Naive animals are used,
so background is not typically a concern. The model allows multiple tissue endpoints to assess
exposure and absorption.  The model is reproducible. Intravenous and oral routes can be used.
Most oral  exposures are fed to the pig in a dough ball (moistened, powdered feed). Twice daily
fasted dosing is typically used for consecutive 12-14 days (subchronic exposure).  Three (3)
levels of soil sample and three levels of reference standard are used (along with negative
controls).  For arsenic, urinary arsenic (48 hours) approximates the oral absorption fraction
(ABA). After 5 days of exposure, arsenic excretion is a linear function of dose (independent of
time).  RBA is the ratio of urinary excretion fraction (UEF) of test material to standard. In
general 80 ppm is lowest level in soil that they will use—because of the amount of soil that
needs to be used to achieve a measurable urinary arsenic concentration.  With increasing
exposure dose, variability increases (heteroscedascity).  Research has found RBAs of 26% to
72% for the test soils.

Q&A
          Is it relevant to do an in vivo study when soil arsenic levels are as high as 4000 ppm?
             o  Stan said that they probably do not, but that is not his decision.
          Can you explain the variability in the control group?
             o  Stan responded that there is some As in the low As feed, but that it is likely
                 due to biological variability.
Assessing Soil Arsenic Bioavailability in the Laboratory Mouse
D. Thomas; US EPA, Research Triangle Park, NC.
M. Hughes; US EPA, Research Triangle Park, NC.
K. Herbin-Davis; US EPA, Research Triangle Park, NC.
P. Seales; US EPA, Research Triangle Park, NC.
                                      PageS

-------
Dave Thomas discussed the results of the pilot study in mice. Dave noted that this is a method
that is still in development.  Goals are to develop a mouse model and to determine whether the
mouse is a good model for As absorption in humans.  Mice were chosen because they have been
well studied, can be manipulated experimentally and a large body of As information is available
for this animal model. Dietary exposure was used. Total arsenic absorption (body burden and
excretion) was used.  Dave noted that pharmacodynamic issues are significant for As
(methylated forms are more toxic); in the GI tract methylation and demethylation reactions do
occur, as well as thiol forms of As.  Similarly in the tissues (post absorption) methylation and
thiolation occur.  The method permits the evaluation of arsenic forms (inorganic As,
methylarsonic acid (MMA), dimethylarsenic (DMA)). Mice are fed purified diet which is non-
detect for As (cellulose in the diet is replaced  with soil). The diet preparation method needs
further development.  Eight (8) to 9 days of exposure are used (steady state). Food consumption
is not impacted by addition of soil to the diet.  One limitation is stress involved with use of the
metabolism cage (mice lose weight). Future work will repeat existing research, examine other
soils, and refine methods of arsenic speciation, examine the relationship between soil source and
patterns of arsenic metabolism, and examine dietary factors that may impact absorption (dietary
fat, iron, and copper levels).

Q&A
          Are you also interested in comparing the mouse model with other in vivo models?
             o  Dave replied that yes, for those soils that have been evaluated by others.
          Why do they lose weight?
             o  Dave responded that this is likely the stress of the open bottom metabolism
                 cage.
          Is a loss of 10% of body mass likely to impact the kinetics of absorption?
             o  Dave responded that we do not know and we would have to design a study to
                 evaluate that issue.
          How many mice in each study?
             o  Dave responded there are three replicates per cage.
Panel Discussion Session I
Question for Stan: at 80 ppm As, how much soil is delivered to the pig?
       Stan responded that is approximately 5-6 grams of soil per dose.  Food impacts
       absorption so this is a concern.
Question for Stan: are there different quantities of soil at different dose levels?
       Yes.
Question for Stan: do the different quantities impact absorption?
       No. However, at higher doses we have an increase in variability, but not a change in
       absorption.
What is the cost per sample for various methods?
       Yvette: monkeys are approximately $90-100k for three soils; in vitro assay for 10 soils is
          approximately $10k.
       Stan: swine are approximately $60k for two soils.
       Dave: mice are probably several thousand dollars per  sample. We have not looked at this
          yet.
                                     Page 9

-------
Is the purpose of in vivo methods to calibrate the in vitro tests or to use data for a site?
       Stan: no. the in vitro assay plays a role, as well as soil sample characterization
          (mineralogy).
       Aaron: EPA wants something cost effective and quick. We are looking to use in vivo
          models for correlation/calibration to have less reliance on animal models.
       Yvette: although in vitro methods will allow us to generate data, there are sites where
          animal studies are warranted.
Which animal model is best? Primate vs. swine vs. mouse vs. human? Which is best for
developing a correlation?
       Aaron: we have accepted the in vitro model for lead based on in vivo bioavailability.  For
          As, we have not yet accepted an in vivo bioavailability model. Until we compare the
          animal models we cannot identify an animal model.
       Mike added that juvenile swine were used for lead to mimic the child for lead. Monkeys
          may be good models for adult exposure and for As because As is a carcinogen.
       Rosalind agreed. Arsenic absorption may be more applicable to adults.
Question for the panel: How can we determine the better model swine or primates?
       Stan replied that the dosing methods are not the same. We are currently evaluating a
          single soil for comparison.  The age issue is pertinent to the discussion. Another
          consideration is using a naive animal vs. using an animal that was exposed over and
          over to the same metal.
       Yvette noted that we have not yet cross-dosed different species with the same test soil.
          That work needs to be done. We need to do this work  and discuss the benefits of the
          various models.
          General agreement that this would be helpful.
          Kirk Scheckel suggested that this would be a huge investment and an opportunity for
             EPA to attack the problem.  European researchers and regulators agree that this
             would be helpful.
Mark Cave: what is the best way forward to work together to achieve this goal?
       Mike agreed that the available data should be pulled together to identify what needs to be
          done to fill in the data gaps.
       Suggestion to plan the next phase: get real soil, characterize the soil, and compare the
          in vivo bioavailability and in vitro results back to the epidemiological data.
          Responses noted that the epidemiological data are confounded by background
             exposures to arsenic exposures in the diet. This complicates the biomarker
             analysis. The UK Environment Agency is working on this issue and the BARGE
             group is also considering this.
Question for the Panel: Can the rat data and swine data be used for arsenic? A lot of research has
been done with  swine. Yvette asked for what chemicals is there a preponderance of data.
       Stan responded that he has dosed between 30-50 arsenic soil samples.
       Rosalind suggested that we should not leave other animal models behind until we have a
          correlation between juvenile swine and the adult human.
       It was also noted that Toxicity Reference Values (TRY) are from rodents—that should be
          a consideration  for model selection.
Marc Stifelman noted that recoverability and mass balance were good for non-human primates,
less so with swine (older Region 10 studies).
                                     Page 10

-------
       Stan replied that the urinary excretion fraction is not affected by total uptake.  His
          laboratory does not collect mass balance data, because they no longer have these
          difficulties.
       Dave Thomas noted that recovery in the mouse model is 80-85% (urinary and fecal, not
          considering tissue retention, which is expected to account for the other 15%).
       Stan noted that this information is necessary for ABA, not for KB A.
       Mark Maddaloni noted that discrepancies in mass balance might be a cause for concern
          because it is these tissue levels that are having the ultimate toxic effect.
       Rosalind agreed that pharmacokinetic considerations need to be evaluated in developing
          an animal model.
       Stan agreed and noted that differences in distribution are easily observed with different
          routes of exposure.
       Yvette agreed and stated we need to be careful in selecting a single animal model because
          our data set is based on a small subset of metals.

Questions for Panel Discussion Session I (evaluation/site application and in vivo research):
    1 - For regulatory applications such as site clean-up decisions, is an in vitro model that is
    predictive (i.e., correlates with an in vivo model) adequate even  if we don't know why it is
    predictive, or is it important to know why a model is predictive?
       From the audience, a suggestion to not be closed minded about model acceptance without
          complete understanding.
       Kirk noted that complete understanding of the model may be the ultimate goal to help
          evaluate long-term stability of contaminants at sites.
       Rosalind would prefer not to need to understand mechanism as long as the model is
          predictive.
       Stan agrees.  We do not need to know the answer—the utility of the in vitro method
          depends upon consistency in correlation for soil types of interest. How much
          confidence we have in adding a new soil type depends upon how well we understand
          the factors controlling the bioavailability.

    2 - Given that many animal models require use of soil concentrations exceeding those of
    public health concern, what efforts are necessary, if any, to show that bioavailability at these
    concentrations reflect oral absorption at lower environmental concentrations? In other
    words, how important is it to evaluate whether bioavailability is concentration dependent?
       Pat asked that if 80 ppm is the lower limit for arsenic studies in swine, what do we do for
          soils between 20 ppm  (Canadian soil level) and 80 ppm? It was agreed that this is a
          concern for monkeys,  too.
       Nick offered that when you are below 100 ppm perhaps you are stuck with in  vitro
          assays. Nick noted that the soil type that doesn't work may be the most informative.
          Why doesn't it work?  That is the important piece of information.
       Kirk suggested that rather than arsenic, can we look at a tracer.
       Mark Maddaloni responded that we have to identify a tracer that we know is very similar
          to arsenic.

    3 - Should a single animal model be specified, or are studies in a variety of different animal
    models more likely to improve our understanding of relative bioavailability in humans.
                                     Page 11

-------
       No. Agreement to retain various animal models for further evaluation.

   4 - Is there a single animal model that will work for all chemicals considering the widely
   varying toxicokinetic profiles?
       Not at this time.
In-vitro Bioaccessibility of Soil-borne Contaminants: An Environment Agency Perspective
S. Saikat; UK Environment Agency, Oxfordshire, UNITED KINGDOM.

       Sohel Saikat gave a brief overview of contamination in the United Kingdom.  The UK
must use in vitro methods because of ethical constraints.  Five in vitro methods that have been
used in the UK were evaluated using an inter-laboratory approach (metals tested: As, Pb, and
Ni).  Labs that used the same methods and same SOP produced consistent results; however,
different methods did not produce comparable results.  They also investigated the utility of in
vitro methods for in vivo bioavailability for As (the Cynomolgus monkey was used).  In general,
the results showed that the five in vitro methods evaluated are not adequate for these  arsenic soils
(i.e., poor correlation between in vivo and in vitro). Sohel also reported that of the three methods
tested, none was adequate for all metals (As, Pb, and Ni).  Sohel noted that bioavailability and
bioaccessibility are considered the same for many UK researchers.  Sohel suggested that more
work is needed to develop the in vitro assay including appropriate validation with an  in vivo
bioavailability model.  Sohel suggested furthering the knowledge of geochemical
characterization (method screening, geochemical matching, geochemical classification, and
biomarkers) to further the in vitro method. Sohel suggested focused discussion on data gaps,
harmonization of approaches among international regulators, and sharing of data among the
research community.

Q&A
       -   What is a reasonable correlation to move forward?
             o  We don't know.  It was also noted that the correlation will change  (as will
                 p-value) when more data become available.
          Mark Maddaloni noted that the important area is when the bioavailability  is low.
          Sohel agreed.
Measurement of Metal Bioaccessibility in Urban Household Dust and Corresponding
Garden Soils
P. E. Rasmussen; Health Canada, Ottawa, ON, CANADA.

Patricia Rasmussen noted high variability of bioaccessible metals in house dust and the many
geochemical differences between outdoor soil and indoor dust.  Patricia noted that sample
preparation (size fraction) and analytical procedures must be consistent for comparisons of
outdoor soil and indoor dust. She observed that tracking-in models do not work in the city of
Ottawa (indoor dust metal concentrations are not predicted accurately using outdoor soil metal
concentrations).  Patricia cautioned that dust may differ from outdoor soil (e.g., some metals are
                                     Page 12

-------
found at higher concentrations and with higher bioaccessibilities; dust also has higher
concentrations of organic carbon).  Patricia reported data for Zn, Ni, Cu, inorganic carbon,
organic carbon and a variety of other metals in dust and corresponding garden soil, and noted
that the higher organic content of indoor dust is associated with greater variability in Ni and Zn.
Patricia reported that for some metals (e.g. copper) bioaccessibility is affected by particle size,
but noted that speciation can override the influence of particle size (likely due to different
species of the metal in the various size fractions). She indicated that although smallest size
fractions yield important information, she finds it necessary to sieve house dust to larger
fractions (e.g. 80 microns or 150 microns) due to the small amount of dust that is typically
obtained when sampling inside homes.  Patricia noted that in her pilot study, the High Volume
Small Surface Sampler (HVS3) vacuum sampler was used (from Rasmussen et  al., 2008 [HERA,
in press]). She provided an overview of the sources of variability in bioaccessibility (variability
in both the numerator and the denominator of the % bioaccessibility equation).  Variability in the
numerator (bioaccessible metal) may result from differences in mass to volume ratio, pH, other
constituents in the sample such as buffers, complexing agents, means of physical mixing and
means of separation (filtration vs. centrifugation).  Variability in the denominator (total metal
concentration) may result from differences in recovery from using alternative methods to
determine total metal concentrations (EPA 3051 vs. HF  total digestion methods). Patricia
suggested qualitative statements (low [<20%], medium [20-59%], and high [>60%]) could be
used to categorize the bioaccessibility of metals for outdoor soil and indoor dust, in light of the
many sources of large variability. Patricia suggested that indoor-outdoor ratios need to be
determined to understand indoor exposures. She offered that representative indoor dust data is
being collected across Canada (Canadian House Dust Study) to develop a national baseline
against which site-related measurements can be compared.
Q&A
       -   What is the German VDI method?
              o  Patricia reported that this is a whole house vacuum method.
Assessment of the Use of Dynamic Human Stomach Models for In-vitro Measurement of
the Bioaccessibility of Arsenics and Chromium in Soils - Can They Replace Animal
Testing?
M. R. Cave; British Geological Survey, Nottingham, UNITED KINGDOM.

       Mark Cave reported on his research to develop dynamic in vitro human stomach models:
TEVI-1 and Model Gut.  These models have been used for pharmacology research.  The TEVI-l
system (TIM) includes a stomach phase and a small intestinal phase.  TIM seeks to mimic human
physiological processes (e.g, temperature, pH changes, peristalsis, secretions, and absorption).
While TIM has been validated for glycemic response, there has been limited testing to date using
soil samples. Results for soil samples showed good correlation for a single As and Pb soil, but
not for Cd. TIM compared well with human subject data for fed/fasted ratio for Pb. The British
Geological Survey (BGS) is developing a standard reference material: 104 ppm As, and 79 ppm
Pb (BGS 102). Data suggest that the arsenic in this sample is of low bioaccessibility.  A
chromium (Cr VI) soil sample (approx 3500-ppm total Cr; 1400 ppm Cr VI) was also evaluated.
Low bioaccessibility was reported for this sample using BARGE and TIM assays.  The dynamic
in vitro models show promising results, however, these methods are not intended to replace batch
                                     Page 13

-------
methods (they may be used as reference methods, to characterize factors underlying differences
in absorption, and provide an alternative to animal models).

Q&A
          How much soil is used for TIM and how much do they cost?
             o   10 g of soil and a few thousand Euros.
       -   Does TIM include a mastication process?
             o  No, but pH changes over time.
          Is redox state controlled?
             o  I am not sure, but I don't think so.
The Use of In Vitro Bioavailability Studies in Human Health Risk Assessment: Scientific
Research and Application by Policy Makers
A. G. Oomen; RIVM, Bilthoven, THE NETHERLANDS.
W. I. Hagens; RIVM,  Bilthoven, THE NETHERLANDS.
J. P. A. Lijzen; RIVM, Bilthoven, THE NETHERLANDS.
E. B. P. Kessels; Actief Bodembeheer de Kempen, Eindhoven, THE NETHERLANDS.
A. J. A. M. Sips; RIVM, Bilthoven, THE NETHERLANDS

      Agnes Oomen noted that for most risk assessments oral bioavailability is equal to the
bioavailability in the studies underlying the reference toxicity study (default RBA is 100%).
Agnes reported that this has changed recently for lead, where RBA for soil lead is 74% (based on
80  percentile value).  The intervention value (similar to Preliminary Remediation Goals) in the
Netherlands is 530 ppm.  She stated the default RBA for lead may be adjusted with reliable site-
specific  data (including in vitro data). Agnes discussed an area of the Netherlands contaminated
by an historical Zn smelter. Contaminated slags and soils from the area were evaluated for
bioavailability of Pb and As. RBA information was not applied because for lead it didn't have a
major consequence for the site; whereas for As the soil did not pose a human health risk, but an
ecological risk.

Q&A
          Might the soils be impacted by lead-based paint, since they are so high?
             o  Agnes responded no.
Assessing Contaminant Bioavailability in Soil when In Vitro Gastrointestinal Methods are
the Only Option
N. T. Basta, Ohio State University, Columbus, OH.
K. G. Scheckel, U. S. Environmental Protection Agency, Cincinnati, OH.
K. D. Bradham; U.S. Environmental Protection Agency, Research Triangle Park, NC.

       Nick Basta discussed the utility of in vitro methods when they are the only option.  Nick
reported that the majority of samples to date have been from highly contaminated soils (often
>2000 ppm Pb; >1500 ppm As). Nick noted that at highly contaminated soils adjustment of a
cleanup level is likely to be influenced by RBA. Development of methods that can be used for
                                    Page 14

-------
moderately contaminated soils (near 400 ppm) may have greater utility for more sites. Nick
noted that most in vivo studies require highly contaminated soils (limit of detection issues). In
vitro methods can be used at moderately contaminates sites, even down to near background
levels. He questioned whether we are confident of in vitro methods at low levels. Nick
suggested that he has greater confidence if it has been validated at high levels for the same
contaminant. If the method will be used for a different contaminant, he is hesitant. Also, the
source of the arsenic (or metal of interest) and factors that control the bioavailability of the
sample (e.g., weathering) need to be better understood to use the method for other samples and
concentration ranges.  Nick believes that different extraction methods can be used, but first these
factors must be understood. Nick provided an overview of some historical data on arsenic and
collaboration with other researchers, including data on scorodite, iron oxide, and phosphate.
When only in vitro methods can be used, As speciation can be used to identify the form of
arsenic in the sample and to determine whether the method has been calibrated for those species.
Nick suggested that more work is needed in speciation and in vivo bioavailability to in vitro
bioaccessibility correlation and round robin studies. Nick added that collaboration maximizes
resources and saves time. Thus, there is a need to share characterized soils and share data.

Q&A
       -   Which of the speciation methods is essential?
              o  Nick responded that advanced spectroscopy will be discussed later. You need
                 to understand the soil type and weathering of the contaminant to select a
                 method.
The Bioaccessibility of Nickel in Contaminated Soils, Can It Be Explained Using Solid
Phase Distribution Data?
J. Wragg, British Geological Survey, Nottingham, UNITED KINGDOM.
M. Cave, British Geological Survey, Nottingham, UNITED KINGDOM.
C. Ollson, Jacques Whitford Ltd., Ottawa, ON, CANADA.
K. J. Reimer, Royal Military College of Canada, Kingston, ON, CANADA.

       Joanna Wragg discussed Ni bioaccessibility and factors that are known to impact Ni
bioaccessibility. Joanna reported on data showing differences in geogenically-influenced soils
and anthropogenically influenced soils.  Geogenic soils tend to have less organic carbon and
lower levels of Ni. Joanna compared Physiologically Based Extraction Technique (Ruby) and In
Vitro Gastrointestinal (Basta) assays for soil samples that were characterized by chemical
speciation.  For anthropogenically influenced soils, the two methods compared well: no
statistically significant difference in Ni bioaccessibility.  Speciation identified 9 distinct soil
components.  PBET and IVG likely measured a mixture of several of these 9 contaminants. For
geogenic-influenced soils, Joanna did not see a significant difference between stomach and
intestinal phase of the PBET and IVG assays.  She used cluster analysis to understand the
speciation data for the anthropogenic soils. In general, anthropogenic soils were more
bioaccessible. Joanna reported that aging of the geogenic sources impacted the bioaccessibility
of iron complexes. Joanna believes that geochemical data can support the use of bioaccessibility
data for risk assessment.
                                     Page 15

-------
Q&A
       -   What pH was used for anthropogenic and glycine extractions?
             o  Joanna responded glycine was used and pH of 1.8.
Importance of Metal Speciation in Understanding Bioavailability
Kirk G. Scheckel; U.S. Environmental Protection Agency, Cincinnati, OH.

       Kirk Scheckel noted that speciation can support remediation and bioavailability. Kirk
differentiated between the issues: what do we want to know and what do we need to know.  Both
chemical and physiology factors influence bioavailability.  Kirk noted that research time at the
synchrotron facility is free and the only cost is travel associated with getting to the facility.  The
synchrotron allows for characterization on the atomic scale. For example x-ray absorption near-
edge spectroscopy (XANES) can measure down to 10 ppm. Kirk discussed a case study using
phosphate amended soil speciation at Joplin, MO. Changes in speciation were observed with
different soil treatments.  Kirk reported on data showing alteration in bioavailability as a function
of time after phosphate treatment where there was a reduction in bioavailability as the time since
treatment increased.  He noted that requirements are an appropriate measure (methodology and
samples) as well as knowledge of the reason for the observed measurement  (outliers).

Q&A
          Concerning the figure with the in vivo comparison, was what was shown the KB A or
          the ABA?
             o   In vitro was RBA others were ABA.
          pH has ranged from 1.5 to 2.5. What should be used?
             o   2.5 was best correlation between PBET and in vivo (closer to 1:1 slope).
          Is 250-micron size fraction only correct size fraction?
             o   Kirk replied that we need to harmonize our size fraction to allow
                 comparability among assays.
Direct Identification of Metal Compounds in Contaminated Soil Mine Tailings and House
Dust Using Synchrotron-based Methods
Heather E. Jamieson, Queen's University, Kingston, ON, CANADA.
S. R. Walker, Queen's University, Kingston, ON, CANADA.
S. E. Fawcett, Queen's University, Kingston, ON, CANADA.
A. Lanzirotti, University of Chicago, Chicago, IL.
P.E. Rasmussen, Health Canada, Ottawa, ON, CANADA.
S. Beauchemin, Natural Resources Canada, Ottawa, ON, CANADA.
M. Parsons, Geological  Survey of Canada, Halifax, NS, CANADA.

      Heather Jamieson noted that bioavailability is a function of mineral, grain size, and
encapsulation. She suggests researchers spend several hours characterizing soils under the
microscope for every hour of synchrotron X-ray beam time. Heather discussed two cases for As.
The first case was from the Giant mine (an abandoned gold mine in Yellowknife near Slave Lake
in Canada), while the second case was a series of sites such as Goldenville and Montague
                                    Page 16

-------
(abandoned gold mines in Nova Scotia).  The Nova Scotia sites are used recreationally today.
They are working on paired speciation and bioaccessibility assays for the samples.  Heather
noted that many As-bearing secondary minerals are nanocrystalline (tens of nanometers), but
porosity, cementation, and disaggregation can impact grain size. Heather also discussed the
Ottawa house dust samples that Pat Rasmussen discussed earlier. Heather believes that
synchrotron methods combined with classical mineralogical studies can provide valuable
characterization information on the metal form in the sample.

Q&A
       -   No time for questions.
Bioaccessibility of Arsenic Adsorbed onto or Incorporated within Freshly Synthesized Iron
Oxide Minerals Using the Simulator of the Human Intestinal Microbial Ecosystem
(SHIME)
Brian D. Laird, University of Saskatchewan, Saskatoon, SK, CANADA.
T. Van De Wiele,  University of Ghent, Ghent, BELGIUM.
D. Peak, University of Saskatchewan, Saskatoon, SK, CANADA.
W. Verstraete, University of Ghent, Ghent, BELGIUM.
S. D.  Siciliano, University of Saskatchewan, Saskatoon, SK, CANADA.

       Brian Laird suggested that understanding the inverse relationship between
bioaccessibility and concentration is a data gap. Does it result from a thermodynamic limitation
(liquid:solid ratio) or a kinetic limitation (residence time)? Complex mineralogy in mine tailings
makes developing a relationship based on mineralogy challenging. Another hypothesis
considered was the impact of colon microbes on As bioavailability. Does microbial activity in
the colon also impact other metals? Brian found that for scorodite, stomach phase
bioaccessibility was determined by the liquid:solid ratio. By contrast, in the small intestine
scorodite bioaccessibility was determined by residence time. Isotherm analysis requires that the
sample mineralogy is identical. Other research has shown that scorodite bioaccessibility was
reduced in the sterile colon as compared to the microbially active colon. The  opposite was seen
with ferrihydrite-As(V). This effect was greater at higher concentrations of ferrihydrite-As(V).
The effect of GI microbial activity may pose a challenge for validation of in vitro models with
in vivo results. Brian acknowledged that the toxicological implications of the microbial activity
are unknown.

Q&A
       -  Does physical mixing affect bioaccessibility?
              o  Brian: Yes.  These are kinetic constraints. Contact with solution is important.
                 These models are operationally defined.
          Have you captured the dynamics of the GI tract?
              o  Brian: No. We have not. The kinetics of the absorption  (pulling arsenic out
                 of the solution) is important, but how do you establish what the rate of
                 absorption is?
          Do we know whether arsenic is  absorbed in the colon?
                                     Page 17

-------
              o  Brian replied yes. Some data are available concerning rate of arsenic
                 absorption in various stages of the GI tract. For example, Brian found only
                 one study of absorption throughout the GI tract (a study from 50 years ago in
                 Japanese).
          Do the animal models we use have microbial flora in the upper GI?
              o  Brian responded: yes. Some species have greater and other species have
                 lesser gut microflora.  Brian noted that cecum of the mouse is relatively large,
                 so the microbial alteration may be more important to understand in this
                 species.
Panel Discussion Session II

1 - Is it sufficient for in vitro models to have correlation with animal results or is it necessary to
make the models accurate physiologic mimics of human gut dissolution?
       Mark Maddaloni stated that good correlation across multiple soil types is the objective.
   Research to develop a physiological model is not necessary (assuming we have validation);
   however, we would like to have detailed information on physiological mechanisms.  Nick
   Basta agrees.

2 - Is it acceptable to use the terms "bioaccessible", "soluble", "migrateable", and "extractable"
interchangeably in the numerator of the % bioaccessibility equation?
   In general the terminology has been consistent for the presentations today. Some differences
       in sample preparation (e.g., pH of digestion and sieving) have come out today. We
       should work to make these consistent.
   Agnes noted that separation method (dialysis vs. centrifugation) also has a large influence on
       results.  This, too, should be standardized.
   Nick noted that the methods with more colloidal material may be influenced by separation
       method to a greater degree than other simpler methods.
   Brian noted that we need to consider the role of concentration on kinetics of dissolution.
   Kirk noted that we are not trying to establish equilibrium in the in vitro tests.
   Yvette noted that for other metals (like barium) we may be reaching the solubility limitation.
   Marc Stifelman noted that for risk assessment purposes we can  probably establish a
       reasonable upper bound for liquid:solid ratio. Others agreed that would be a start.

3 - Should bioaccessibility values be used on their own without supporting information? If not,
what additional information should be included for each soil sample analyzed (e.g.,
geochemistry)?
   Rosalind believes that supporting information is important for better understanding the site.
       Patricia  asked for clarification of the question "be used for what?"
   Are we using the bioaccessibility information to make site decisions or for comparing
       sources? Bioaccessibility is operationally defined.
   Beverly Hale agreed. We should collect as much information as possible.
   Rosalind followed up that for many of the  studies we need to understand the limitations and
       how we  can augment the information.
                                     Page 18

-------
   Kirk noted that electron microprobe studies tell you about chemical composition, it does not
       provide speciation. The results can be misleading for As. XRD and microscopy can give
       you some very good information also.
   Mark noted that the bioaccessibility and extraction methods are best used in combination.
   Nick added that weathering is an important factor. The speciation of soil samples is
       important and these methods should be used.

What can be done where we have limitations on methods for low concentration and in vivo
bioavailability methods?
   Marc Stifelman noted that he agrees with testing unknowns in as many species as available.
       We can probably feed monkeys low levels.
   Yvette agrees—we should study soils where the metal concentration is of environmental
       relevance (e.g., 50-100  ppm for As). Moving from there to the risk range is not a scary
       extrapolation.  Mark Maddaloni  agrees.
   Marc Stifelman noted that with the exception of Florida, most Superfund soil cleanup goals
       range from 20-245 ppm As.
   Nick suggested that if we can characterize the species that is controlling As solubility, then
       the effect of solubility is likely to be the same at 30 ppm as it is at 3000 ppm.
   Kirk and others raised concerns with this related to pH and other factors (as demonstrated
       with scorodite  research).

Particle size distribution issue:  Rosalind noted that for oral studies many have sieved to 250-|im
particle size (based on adherence of soils to skin and incidental soil ingestion pathway). Recent
research on dust is looking at smaller particle size fraction (<150|im).  There may or may not be
enrichment in the finer fraction. What is the appropriate particle size range cutoff?
   Kirk offered that for consistency we should stick with one size fraction.
   Rosalind agrees that harmonization is an important issue.
   Yvette added that when they examined soils from SERDP, the majority of lead mass was in
       the <75 |im size fraction.
   Patricia noted that house dust has a different particle size distribution. A key issue is to use
       the same procedure for indoor and outdoor dust.  Patricia noted that it is physically
       challenging to  get down below 50 jam. Mark Richardson's paper (Health Canada)
       encouraged the further characterization of size fractions (look at different cuts).
   Agnes added that in the Netherlands they agree that smaller particles tend to stick to the
       hand, but if even a few  large particles adhere to the hand those can dramatically impact
       the results.  RIVM uses the <2 mm  size fraction as a result.
   Rosalind noted that the particle size  fraction to keep in mind is the size fraction commonly
       used for site characterization, <250 jim.
   Patricia noted that the cost of sieving is relatively cheap (a few dollars for full
       characterization of a sample).  A more complete characterization may result in greater
       consistency across sites. Patricia and Yvette agreed.
   Yvette noted that the 3050 method found >100% bioaccessibility for barium (this is a known
       limitation of the 3050 method).  Agreement that consistency in digestion method is
       important. Heather agrees that results from total digestion may be misleading.
                                     Page 19

-------
Summary
       The EPA guidance also recommends using a validated methodology and provides
       recommended evaluation criteria for the development of new methods.

       Establishing a correlation between an in vitro method and an in vivo method was
       generally considered an important step.  A 1:1 correlation is not necessary, only that the
       model is predictive for soil types and contaminant concentrations of interest.

       Agreement that mimicking physiology is not necessary, nor is complete characterization
       and understanding of processes, for an in vitro method to be accepted.  What is important
       is that the in vitro method is correlated with an in vivo model and that it is predictive for
       soil types and contaminant concentrations of interest.

       While not strictly required, insight into the factors underlying the absorption processes
       for a given metal and complete characterization of the metal in the media of interest will
       inform the limits of a test method for a given metal species (e.g., concentration range).
       This information will also determine confidence in using the method for other forms of
       the metal or soil phases.

       Any candidate method for evaluation must be presented with bounds of valid use (e.g.,
       metal forms) and methodological constraints (e.g., concentration range, pH, and
       liquid:solid ratio).

       Agreement that the next logical step is to develop an arsenic assay (in vivo and in vitro).

       Cost is a factor for an in vitro assay. EPA is seeking in vitro methods that are cost
       effective and relatively quick. The in vivo model against which it is correlated is not
       necessarily similarly constrained.

       At this time no decision has been made concerning which animal model is preferred for
       metals other than soil borne lead from mining, milling, and smelting sites (for which EPA
       has  an SOP).

       For selection of an in vivo model for As, comparisons among the available animal models
       using the same soils are required to make this determination.

       A single animal model may not work for all metals because of differences in
       pharmacokinetics.

       A more complete characterization of concentration and bioaccessibility for various
       particle size fractions may result in greater consistency across sites.

       There is a need for harmonization for digestion methods and sieving across sites.


                                     Page 20

-------
•      A combination of laboratory, microscopic, and synchrotron methods may provide useful
       information for bioaccessibility, bioavailability, and risk assessment.

•      Agreement was made to identify, characterize, and share standard reference materials for
       in vivo and in vitro assays.

•      Agreement to share information and data, as well as soil samples, to identify data gaps
       and research needs.

•      Agreement to collect soil  samples from a wide array of relevant sites, preferring
       weathered soils over spiked soils (because they may differ).

•      Agreement to meet again  before the 2008 ISEA meeting.  This will foster collaboration
       and further discussion.
Although this work was reviewed by EPA and approved for publication, it may not necessarily

reflect official Agency policy.
                                     Page 21

-------
General References:

Basta, N., Foster, J., Dayton, E., Rodriguez, R., Casteel, S., The effect of dosing vehicle on
arsenic bioaccessibility in smelter-contaminated soils. Journal of Environmental Science and
Health, Part A 2007, 42, (9), 1275-1281.

Casteel, S., Evans, T.,  Turk, J., Basta, N., Weis, C., Henningsen, G., Hoffman, E., Refining the
risk assessment of metal-contaminated soils. InternationalJournal of Hygiene and
Environmental Health 2001, 203, (5-6), 473-474.

Cave, M., Taylor, H., Wragg, J., Estimation of the bioaccessible arsenic fraction in soils using
near infrared spectroscopy.  Journal of Environmental Science and Health, Part A 2007,  42, (9),
1293- 1301.

Drexler, J., Brattin, W., An in vitro procedure for estimation of lead relative bioavailability: with
validation.  Human and Ecological Risk Assessment 2007, 13, (2), 383 - 401.

EPA.  2007. Guidance for Evaluating the Oral Bioavailability of Metals in Soils for Use in
Human Health Risk Assessment. OSWER Report 9285.7-80.

EPA.  2007. Estimation of Relative Bioavailability of Lead in Soil and Soil-like Materials Using
in Vivo and in Vitro Methods.  OSWER 9285.7-77.

ICCVAM (Interagency Coordinating Committee on the Validation of Alternative Methods).
1997. Validation and Regulatory Acceptance of Toxicological Test Methods:  A Report of the Ad
Hoc Coordinating Committee on the Validation of Alternative Methods. NIH Publication
973981. National Institute of Environmental Health Sciences, Research Triangle Park, N.C.

Kelly, M.,  Brauning, S., Schoof, R.,  Ruby, M., Assessing Oral Bioavailability of Metals in
Soil, A 1 &A2, 2002  pp. 75 -88, Battelle Press Columbus, OH.

Laird, B., Van De Wiele, T., Corriveau, M., Jamieson, H., Siciliano, S., Evaluation of the
bioaccessibility of metals and metalloids in Eastern Canadian mine tailings using an in vitro
gastrointestinal model, the simulator of the human intestinal microbial ecology (SHIME).
Epidemiology 2006, 17, (6), S40-S40.

Oomen, A., Hack, A.,  Minekus, M., Zeijdner, E^Cornelis, C., Schoeters, G., Verstraete,  W.,
Van de Wiele, T., Wragg, J., Rompelberg, C., Sips, A., Van Wijnen, J.,  Comparison of five in
vitro digestion models to study the bioaccessibility of soil contaminants. Environ. Sci. Technol.
2002, 36, (15), 3326-3334.

Rodriguez, R., Basta, N., Casteel, S., Pace, L., An in vitro gastrointestinal method to estimate
bioavailable arsenic in contaminated soils and solid media. Environ. Sci. Technol. 1999, 33, (4),
642 -649.
                                     Page 22

-------
Roberts S., Munson J., Lowney Y., Ruby M. Relative Oral Bioavailability of Arsenic from
Contaminated Soils Measured in the Cynomolgus Monkey. Toxicological Sciences 2007
95(l):281-288

Ruby, M., Davis, A., Schoof, R., Eberle, S., Sellstone, C., Estimation of lead and arsenic
bioavailability using a physiologically based extraction test. Environ. Sci.  Technol. 1996,30,
(2), 422-430.

Saikat, S., Barnes, B., Westwood, D., A review of laboratory results for bioaccessibility values
of arsenic, lead and nickel in contaminated UK soils. Journal of Environmental Science and
Health, Part A 2007, 42, (9), 1213-1221.

Scheckel, K., Ryan, J., Allen, D., Lescano, N., Determining speciation of Pb in phosphate-
amended soils: Method limitations. Sci. Total Environ. 2005, 350, (1-3), 261-272.

Wragg, J., Cave, M., Nathanail, P., A Study of the relationship between arsenic bioaccessibility
and its solid-phase distribution in soils from Wellingborough, UK.  Journal of Environmental
Science and Health Part A-Toxic/Hazardous Substances & Environmental Engineering 2007, 42,
(9), 1303-1315.
                                     Page 23

-------
Appendix A: Questions for the Panel (submitted in advance)

Questions for Panel Discussion Session I (evaluation/site application and in vivo research):
    1 - For regulatory applications such as site clean up decisions, is an in vitro model that is
    predictive (i.e., correlates with an in vivo model) adequate even if we don't know why it is
    predictive, or is it important to know why a model is predictive?
    2 - Given that many animal models require use of soil concentrations exceeding those of
    public health concern, what efforts  are necessary, if any, to show that bioavailability at these
    concentrations reflect oral absorption at lower environmental concentrations? In other
    words, how important is it to evaluate whether bioavailability is concentration dependent?
    3 - Should a single animal model be specified, or are studies in a variety of different animal
    models more likely to improve our understanding of relative bioavailability in humans.
    4 - Is there a single animal  model that will work for all chemicals considering the widely
    varying toxicokinetic profiles?
    5 - Is relative oral bioavailability of chemicals in soil so routine that it is "ready" for
    guidance/regulation that will inherently limit science-based research?

Questions for Panel Discussion Session II (in vitro and speciation/mineralogy research):
    1 - Is it sufficient for in vitro models to have correlation with animal results or is it necessary
    to make the models accurate physiologic mimics of human gut dissolution?
    2 - Is it acceptable to use the terms  "bioaccessible", "soluble", "migrateable", and
    "extractable" interchangeably in the numerator of the % bioaccessibility equation?
    3 - Should bioaccessibility  values be used on their own without supporting information?  If
    not,  what additional information should be included for each soil sample analyzed (e.g.,
    geochemistry)?
    4 - How can we leverage resources  to answer specific research questions to advance the
    understanding of bioavailability/bioaccessibility?
       - Develop Standard Reference Materials or Certified Reference Materials?
       - Round robin testing?
    5 - In lieu of comparing in vitro model  results to an in vivo model, what criteria  should be
    considered when evaluating whether a particular in vitro method is appropriate for providing
    screening level data versus data to derive quantitative site-specific bioavailability
    adjustments?
    6 - What other metals, or metal species, are of interest for developing additional in vivo and
    in vitro bioavailability assays? What criteria are used by regulatory agencies for ranking or
    prioritizing contaminants?
    7 - Is it feasible and cost effective to develop a single standardized in vitro method for each
    metal for simulating oral bioaccessibility? If so, what steps would be needed for that to
    happen? What criteria would be used to determine the difference between the methods?
    - What is available (and in  process) and how available/quantities (related to above) how
    prioritized types of materials needed (ex: soils, tailings, etc.)
    - Discussions of desired specifications by type (uses will drive Data Quality Objectives)
    - Funding (ex: to National Institute  of Standards and Technology for Standard Reference
      Materials) vs. volunteer analysis via round robins (to develop consensus values)  and how
       determine "reliable" methods?  International Atomic Energy Agency could be a model.
                                      Page 24

-------
Appendix B: Abstracts

-------
Abstract 580
Introduction and Overview
K. Bradham; U.S. Environmental Protection Agency, RTF, NC

In human health risk assessments, soil and dust ingestion can be a major route of
exposure to many soil contaminants, including metals and metalloids. Site-specific soil
physical and chemical characteristics, as well as internal biological factors, determine the
oral bioavailability of soil contaminants. Within a single sample, this contamination may
be from multiple sources of metals and may exist as different forms and species. Both
animal models and in vitro bioaccessibility models have been used to estimate relative
bioavailability of metals in soil and dust. The bioavailability estimates for soil have a
direct impact on current human health risk assessment and risk management practices.
This introduction and overview to the symposium will include definitions and
specification information necessary for setting the stage for the presentations on recent
developments in animal models, new in vitro models, the role of mineralogical analyses
in assessing relative bioavailability, and the application of physiologically-based models
as research tools. Information will also be presented regarding the panel discussions and
specific research questions provided to the panelists and presenters for discussion during
this symposium.

Although this work was reviewed by EPA and approved for publication, it may not
necessarily reflect official Agency policy.

-------
Abstract 520
Evaluating the Unavailability of Soil-Borne Contaminants at Waste Sites
M. Beringer1, A. Yeow2; *U.S. Environmental Protection Agency, Kansas City, KS,
2OSRTI, Washington, DC

Site-specific bioavailability is an important consideration in determining potential threats
to human health that are posed by metals-contaminated soils at waste sites. It is important
to consider bioavailability because metals may be absorbed to a lesser or greater extent
following ingestion of contaminated soils as compared to the fraction absorbed in the
studies used to establish toxicity values, such as a reference dose or a cancer slope factor.
U.S. EPA's Office of Superfund Remediation and Technology Innovation (OSRTI) has
led an effort to develop guidance on evaluating and incorporating bioavailability
adjustments into human health risk assessments. The guidance outlines a decision
framework for deciding when to collect and incorporate site-specific bioavailability
information; recommends a process for documenting the data collection, analysis, and
site-specific implementation of a validated method; as well as provides recommended
method validation and regulatory acceptance criteria for evaluating alternative
methodologies. U.S. EPA has used these criteria to evaluate two separate methodologies
for predicting the relative bioavailability of lead in soil and soil-like materials. The
Agency has determined that both an in vivo swine bioavailability bioassay and an in vitro
bioaccessibility assay have sufficiently satisfied these criteria. Thus, they are considered
regulatory methodologies appropriate for determining the relative bioavailability  of lead
for quantitative use in site-specific risk assessments. This presentation will summarize the
bioavailability guidance document and the basis for the Agency's decision regarding the
two methodologies for predicting lead relative bioavailability.

-------
Abstract 514
Method Development and the Application of Oral Unavailability Data in U.S. Risk
Assessments
R. A. Schoof; Integral Consulting Inc., Mercer Island, WA

USEPA guidance includes provisions for site-specific adjustments in exposure estimates
to account for differing relative bioavailability of chemicals in soil and in the exposure
media of toxicity studies. The relative oral bioavailability of lead and arsenic in soil has
been assessed in a series of studies that have included both animal models and in vitro
test systems. USEPA recommends a default assumption that children will absorb only 60
percent as much lead from soil compared with absorption from water or diet. For arsenic,
no such default assumption has been generally accepted. In vivo bioavailability studies
should be designed to account for variations in metabolism and excretion of chemicals.
The absorption and disposition of lead and arsenic differ substantially. Consequently,
bioavailability studies to assess these two chemicals have been designed to reflect  these
differences. In vivo methods used to assess the bioavailability  of soil-borne chemicals are
typically modified versions of methods widely used in biomedical research. These
methods have been modified to address constraints associated with use of doses relevant
to environmental concentrations, the need to reflect weathering behavior in soils over
time, and the need to generate data applicable to human health risk assessments.
Regulatory acceptance of bioavailability data in a site-specific risk assessment is
generally dependent on the use of a validated test method or a  careful scientific review of
the test method employed. In vivo bioavailability data generated by traditional study
designs is usually considered to be acceptable, but in vitro studies face a greater burden in
obtaining acceptance. In the US a process developed by the Interagency Coordinating
Committee on the Validation of Alternative Methods for validating newly developed
alternative toxicity methods provides relevant guidance for assessing in vitro methods.
Efforts to obtain regulatory acceptance of in vitro studies of relative bioavailability will
be reviewed.

-------
Abstract 461
Arsenic Bioaccessibility Testing Using Various Extraction Methods: Results and
Relation to Relative Oral Unavailability as Measured in the Cynomolgus Monkey
Y. W. Lowney1, S. Roberts2, S. Saikat3; Exponent, Boulder, CO, 2University of Florida,
Gainesville, FL, 3UK Environment Agency, Wallingford, UNITED KINGDOM

Recent research has established that the absorption of arsenic from soils following
ingestion exposures is lower than absorption of soluble arsenic from water. Because
regulatory toxicity values for arsenic are based on studies of human exposures to arsenic
in water, understanding the relative oral bioavailability (RBA) of arsenic from soils is
important for accurate assessment of exposure to arsenic associated with soil ingestion.
Because of the site-specific nature of the controls on arsenic bioavailability, site-specific
information is important in assessing potential risk associated with arsenic in soils. To
conduct an animal study of bioavailability for every site affected by arsenic would be
time- and cost-prohibitive, and may counter policies regarding the use of animals in
research. Therefore, recent efforts have focused on developing practical and economical
bench-top (in vitro) procedures to measure the fraction of contaminants in soils  that,
following ingestion, would be available for absorption into systemic circulation.
This presentation will discuss recent in vivo testing of arsenic bioavailability in  the
cynomolgus monkey, and the results of in vitro extraction tests that have been designed
to predict the in vivo bioavailability results. The in vitro methods include an extraction
protocol that has been validated as predictive of the RBA of lead from soil, other
extraction methods that have been reported in the literature, methodologies developed by
the Dutch RIVM, and the RIVM method as modified by BARGE. This presentation
provides the results of extraction testing of splits of several soils that were evaluated in
vivo, and discusses the differences in results between the extraction methods, and the
correlation between the results from extraction testing  and RBA as measured in
cynomolgus monkeys. Results indicate that additional  work remains to identify  an
individual in vitro method that is able to predict the in  vivo bioavailability  satisfactorily
for all soils.

-------
Abstract 519
Assessing Unavailability Using the Swine Model
S. Casteel1, G. Pent1, C. Weis2, W. Brattin3; University of Missouri, Columbia, MO,
n                                                        o
 Environmental Protection Agency, Region VIII, Denver, CO, Syracuse Research Corp,
Denver, CO

Bioavailability of site-specific environmental contaminants is critical to exposure
assessment. Determining the bioavailability of contaminants in a diverse range of soils,
allows scientifically derived data to dictate site-specific remedies to reduce the risk for
sensitive human populations. Based on a series of dosing trials in a juvenile swine model,
site-specific estimates of relative bioavailability of metals and organic compounds, is
highly variable and is matrix and chemical species dependent. Results for lead- arsenic-
and cadmium-contaminated soils support the view that soil metals are  not always as well
absorbed as soluble forms; therefore use of default assumptions for assessing human
health risk may overestimate the hazard.

Since the selection of appropriate animal models enhances the science and reduces
uncertainty in human risk assessment it is critical to use the best available model with
reasonable constraints. Numerous rodent model studies are in the literature.  Studies in
non-human primate (NHP) models have been fewer due to higher purchase and per diem
costs, housing availability, zoonotic concerns, and animal rights attention.
Criteria useful in selecting the appropriate animal model include  behavior, age, size, ease
of bleeding, anatomical considerations and gastrointestinal physiology. Research
management factors such as historical database, costs, model availability, and animal
rights group interest in the model were also involved in selection.

The swine model has the versatility to assess the bioavailability of a wide variety of
materials,  including metals, organic compounds, and biodistribution of gold-, palladium-
and silver-nanoparticles. When assessing site-specific contaminants, pigs are dosed 2
hours before each feeding twice daily for 14-to-15 consecutive days, at constant dosing
times. Multiple doses are provided as a more likely real-world reflection of exposure.
This approach has been successfully applied at numerous sites for estimation of relative
bioavailability for lead, arsenic, cadmium, vanadium, and chromium.

-------
Abstract 521
Assessing Soil Arsenic Unavailability in the Laboratory Mouse
D. Thomas, M. Hughes, K. Herbin-Davis, P. Seales; U.S. Environmental Protection
Agency, RTF, NC

Variation among soils in the bioavailability of arsenic can be a critical determinant of the
risk posed by exposure to these soils. Although in vitro techniques can provide vital data
on aspects of bioavailability of metals and metalloids from soils, these results must be
validated in an animal model. A useful animal model provides a measure of
bioavailability and allows comparison of bioavailability for different soil matrices. Inbred
strains of laboratory mice are potentially good models for development of a
bioavailability assay. Laboratory mice are well characterized physiologically and can be
manipulated experimentally (e.g., altered dietary components, altered genotype). There is
also a large body of data on the absorption, metabolism, disposition, and excretion of
inorganic and methylated arsenicals in the mouse which is germane to evaluating the
differences and similarities between mouse and human. Initial studies are comparing
arsenic bioavailabilities in soils with known arsenic contents with the bioavailability of
sodium arsenate. Here, soils (e.g. NIST  SRM 2710) or sodium arsenate are added to a
standard powdered mouse chow (AIN-93G purified rodent diet) at the one percent
(weight/weight) level. Adult female C57BL/6 mice have had free access to this amended
chow and tap water for nine days. Urine and feces are collected on a daily basis and food
intake is monitored throughout this period. At the end of the nine-day exposure period,
mice are euthanized and tissues collected. Data on food consumption and arsenic contents
of excreta and selected tissues are used to calculate the bioavailability of arsenic in each
soil matrix. Development and refinement of this animal model should provide a
convenient and rapid means to assess the absolute and relative bioavailability of arsenic
in soils. These data may be of great value in risk assessment. (This abstract does not
reflect US EPA policy.)

-------
Abstract 518
In-vitro Bioaccessibility of Soil-borne Contaminants: An Environment Agency
Perspective
S. Saikat; Environment Agency, Oxfordshire, UNITED KINGDOM

In the UK, interest in the use ofin-vitro bioaccessibility data in risk assessment, has been
stemmed mainly from the problem associated with elevated arsenic in mineralised and
mining effected areas. There is an expectation that in-vitro bioaccessibility data can be a
'quick fit solution' to dealing with land contamination where contaminant level exceeds
corresponding generic assessment criteria (e.g. Soil Guideline Values).

Studies undertaken by the Environment Agency, however, indicate that a number of in-
vitro methods (e.g. physiologically based, semi-physiologically based and simple
chemical leaching) are currently available to measure in-vitro bioaccessibility but no
information to indicate that they are being validated with in-vivo data for UK soils. The
study also indicated that reproducibility of different in-vitro methods, operating
procedures and reporting of results could contribute to a large variation in in-vitro
bioaccessibility data. Laboratories use same in-vitro method irrespective of chemicals,
concentrations, mineralogy and soil types. Moreover, no reference material containing in-
vivo data is available to measure the accuracy ofin-vitro methods.

In order to appreciate and make the best of research progress achieved, a review of
outstanding issues is required to consolidate efforts and develop appropriate partnerships.
Scientists, risk assessors and regulators need to balance their expectations of the in-vitro
approach in terms of its capabilities and weaknesses in order to make it more useful in
risk assessment.

-------
Abstract 524
Measurement of Metal Bioaccessibility in Urban Household Dust and
Corresponding Garden Soils
P. E. Rasmussen; Health Canada, Ottawa, ON, CANADA

Large uncertainties are associated with the measurement of gastric bioaccessibility of
metals in household dust, caused in part by the heterogeneous nature of settled dust
samples, and in part by variations in analytical parameters. A modified version of
European Standard EN 71-3 Toy Safety Protocol was used as a rapid screening method
for estimating gastric bioaccessibility of metals in urban geochemical surveys of
household dust and corresponding garden soil samples in Ottawa, Canada. In this study,
gastric bioaccessibility is defined as the concentration of metal leached from the test
sample into 0.07 M HC1 (2 h at 37°C; pH 1.5), expressed as a percentage of the total
metal concentration. To improve measurements of the total metal concentration (the
denominator in the bioaccessibility equation) several modifications were made to the US-
EPA 3051 microwave digestion protocol. Increasing the microwave digestion time to 30
min ramp followed by 30 min hold (compared to 5.5 min total digestion time specified by
EPA3051) increased total metal recoveries by 15-20%. Increasing the acid volume to
sample mass ratio to 1000 (compared to ratios of 20 to 100 specified by EPA3051)
increased total metal recoveries by 30-60%. Similarly, for the simulated gastric extraction
(the numerator), increasing the acid volume to sample mass ratio to 2000 (compared to
ratios of 50 to 500 specified by  the Toy Safety protocol) typically increased the extraction
efficiency by 20 to 50%. Analytical reproducibility is improved using smaller sieve
fractions (<60 micron is best); however, settled dust samples collected in this study were
typically very small (1-2 g) necessitating the use of a larger size fraction (<150 micron).
In light of the inherent variability associated with settled dust measurements, estimates of
gastric bioaccessibility are grouped into simple categories: low (19% and less), medium
(20 to 59%) and high (over 60%).

-------
Abstract 391
Assessment of the Use of Dynamic Human Stomach Models for In-vitro
Measurement of the Bioaccessibility of Arsenics and Chromium in Soils - Can They
Replace Animal Testing?
Mark Cave1, Helen Taylor1, Joanna Wragg1, Andrew Broadway21 British Geological
Survey, Key worth, Nottingham; University of Edinburgh, School of GeoSciences,
Edinburgh, UK

The development of methods for estimating the oral bioavailability of soil contaminants
may reduce costs of site remediation and soil cleaning, while still maintaining the
required protection level. Currently, simple batch in vitro extraction methods, which
broadly mimic the physico-chemical conditions in the human gastro intestinal tract, have
been developed as screening methods for bioaccessibility measurement. Regulatory
authorities, however, require that the in-vitro methods should produce data that is
demonstrated to be comparable to the in-vivo situation. It has been shown, however, that
the GI tract of young pigs is similar to humans and that they can be used to validate the
results of in-vitro tests. Animal studies are, however, time consuming, costly, have
ethical considerations and there are concerns regarding their relevance to the human. The
food and drug industry has worked to produce dynamic in-vitro systems specifically
designed to mimic the human gastrointestinal system (Wickham, 2007, Minekus, 1995 ).
Such systems may be as relevant as animal models for soil bioavailability studies and
have a part to play in either estimating bioaccessibility or in validating the simpler batch
tests. This paper will discuss the results obtained for the bioaccessibility of arsenic and
chromium from soils with both in-vivo bioavailability and batch in-vitro bioaccessibility
data.

Wickham, M. (2007): The Model Gut, 2007 (27April),
http://www.ifr.ac.uk/science/platform/MG/default.html.
Minekus, M, Marteau, P, Havenaar, R and Huisintveld, JHJ. (1995): "A
Multicompartmental Dynamic Computer-Controlled Model Simulating the Stomach and
Small-Intestine", Atla-Alternatives to Laboratory Animals, 23(2), 197-209.

-------
Abstract 517
The Use of In vitro Unavailability Studies in Human Health Risk Assessment:
Scientific Research and Application by Policy Makers
A. G. Oomen1, W. I. Hagens1, J. P. A. Lijzen1, E. B. P. Kessels2, A. J. A. M. Sips1;
National Institute for Public Health and the Environment, Bilthoven, THE
NETHERLANDS, 2Actief Bodembeheer de Kempen, Eindhoven, THE
NETHERLANDS

Today, a relative bioavailability factor of "1" is used for human health risk assessment of
contaminated soils. This implicates the assumption that there is no difference in the
bioavailability of a contaminant from soil compared to the bioavailability from the matrix
used in the studies underlying the Intervention Value for remediation, which is typically a
food or water matrix. However, there is ample evidence demonstrating that the
bioavailability of a contaminant from soil can be considerably lower than from food or
water. Integrating oral bioavailability of contaminants from soil in human health risk
assessment will increase the realistic outcome of risk assessment through soil ingestion.
The research in  this presentation focuses on the contaminant lead, since lead is frequently
encountered at human lexicologically high concentrations in soil in the Netherlands.
Furthermore, soil ingestion is an important pathway of exposure for lead, especially for
children, leading to potential adverse effects. Therefore, the need for a realistic but still
protective risk assessment for human health is high.

The RIVM has developed a simple experimental tool, an in vitro digestion model, to
supply information on the bioavailability of a contaminant in the human body after
ingestion of contaminated soil. This model has been used to estimate the bioaccessibility
of lead in specific soils. With CSOIL, site specific risk assessment of human health can
be modeled to answer specific policy issues.

In this presentation, the experimental setup and outcome of such a project is given.
Furthermore, the implementation of the results and the scientific advice towards policy
makers is addressed. Actions taken by policy makers following the recommendations are
discussed.

Taken together, this presentation gives  an overview on the involvement of the RIVM in
vitro digestion model in site specific risk assessment in the Netherlands.

-------
Abstract 516
Assessing Contaminant Unavailability in Soil when In Vitro Gastrointestinal
Methods are the Only Option
N. T. Basta1, K. G. Scheckel2, K. D. Bradham3; ^hio State University, Columbus, OH,
0                                                    1
 U. S. Environmental Protection Agency, Cincinnati, OH,  U.S. Environmental
Protection Agency, Research Triangle Park, NC

Incidental soil ingestion is an important exposure pathway for assessing public health
risks associated with contaminated soils. The bioavailability of Pb, As, and possibly other
contaminants in soils can be determined by conducting dosing trials using acceptable
surrogate animal models. To overcome the difficulty and expense associated with in vivo
trials, in vitro gastrointestinal (IVG) methods, that simulate human gastrointestinal
conditions, have been developed. Bioaccessible Pb and As determined by several IVG
methods has been shown to be correlated with in vivo bioavailability data.
Soils must have a very high contaminant concentration, often > 500 or 1000 mg/kg, to
accurately measure bioavailability from animal  dosing trials. Most contaminated soils are
not highly contaminated. These moderately contaminated soils require risk assessment
but are below the "detection limits" of animal models. IVG methods will be the only
methods that can be used for exposure assessment of moderately contaminated soils.
Soil chemistry, mineralogy, and other geomedia properties are likely to have more
influence on contaminant bioavailability in moderately contaminated soils than highly
contaminated soils. Can we rely on  IVG methods to assess contaminant bioavailability in
moderately contaminated soil without method validation based on in vivo bioavailability
data? Soil and contaminant chemistry requirements necessary  for accurate application of
IVG methods to access contaminant (bio)availability will be presented.

-------
Abstract 406
The Bioaccessibility of Nickel in Contaminated Soils, Can It Be Explained Using
Solid Phase Distribution Data?
J. Wragg1, M. Cave1, C. Ollson2, K. J. Reimer3; British Geological Survey, Nottingham,
UNITED KINGDOM, 2Jacques Whitford Ltd., Ottawa, ON, CANADA, 3Royal Military
College of Canada, Kingston, ON, CANADA

In recent years there has been increased use of bioaccessibility testing to determine the
fraction of potentially harmful elements (PHEs) available for uptake in the human
gastrointestinal tract, and which therefore may pose a risk to human health. The data
produced by such tests may be incorporated into human health risk assessments to
determine the risk posed to the critical receptor, by a given land use, for the ingestion
pathway.

In tandem, research has focussed on identifying the physico-chemical sources of
bioaccessible PHEs (and in some cases non-bioaccessible PHEs) in soils. Identification of
the  physico-chemical hosts of PHEs can be achieved by the use sequential extractions.
Application of these techniques in conjunction with bioaccessibility methods can  aid the
understanding of soil-contaminant relationships and how contaminant bioaccessibility
and mobility may impact on human health risk assessment and future land use. To date,
most interest has centred on arsenic but more recently focus has shifted to nickel (Ni).
The physico-chemical sources of Ni in soils collected from Sudbury,  Ontario will be
described, after indentification by the use of the CISED (Chemometric Identification of
Substrates and Element Distribution) extraction technique with confirmatory X-ray
diffraction information. The information surrounding the solid phase  distribution  of Ni in
the  soils will then be used to provide an understanding of measured bioaccessibility data.

-------
Abstract 407
Importance of Metal Speciation in Understanding Bioavailability
K. G. Scheckel; U.S. Environmental Protection Agency, Cincinnati, OH

The speciation or chemical form of metals governs their fate, toxicity, mobility, and
bioavailability in contaminated soils, sediments and water. To assess these chemical
properties and to accurately gauge their impact on human health and the environment we
need to characterize metals at the atomic level. One can employ an array of techniques to
address speciation including XRD, DRS, TEM, TGA, and XPS. In addition to these tools,
researchers have used synchrotron radiation methods to elucidate metal speciation.
The complexity of metal contaminated sites has and continues to be simplified to a
measure of the total metal content. While total metal content is a critical measure in
assessing risk of a contaminated site, total metal content alone does not provide
predictive insights on the bioavailability, mobility, and fate of the metal contaminants.
Our ability to determine metal speciation in soils enhances efforts to understand the
mobility, bioavailability, and fate of contaminant metals in environmental systems, to
assess health risks posed by them, and to develop methods to remediate metal
contaminated sites. To attain in situ atomic level information on the speciation of metals
we utilize high-energy synchrotron  X-rays to probe chemical structure. At the Advanced
Photon Source (APS) of Argonne National Laboratory (Argonne, IL), we incorporate X-
ray absorption (XAS), X-ray fluorescence (XRF), and micro-tomography spectroscopies
to analyze environmental samples to determine the true,  in situ speciation of metal
contaminants. These innovative research tools are expanding our ability to directly
identify the role of metal speciation on many dynamic processes that influence risk.

-------
Abstract 587
Direct Identification of Metal Compounds in Contaminated Soil, Mine Tailings and
House Dust Using Synchrotron-based Methods
H. E. Jamieson1, S. R. Walker1, S. E. Fawcett1, A. Lanzirotti2, P. Rasmussen3, S.
Beauchemin4, M. Parsons5; Queen's University, Kingston, ON, CANADA, 2University
of Chicago, Chicago, IL, 3Health Canada, Ottawa, ON, CANADA, 4Natural Resources
Canada, Ottawa, ON, CANADA, 5Geological Survey of Canada, Halifax, NS, CANADA

Contaminated soils can be expected to contain multiple hosts of the metal or metalloid of
concern, especially in the case of mine-impacted soils or tailings, where the
concentrations are orders of magnitude above soil quality guidelines. For example, we
have determined that a single sample of arsenic-rich gold mine tailings contains, in
addition to the primary arsenopyrite (FeAsS), five secondary oxidation products namely
scorodite (FeAsO4 2H2O) , amorphous Fe arsenate,  kankite (FeAsO4 3.5H2O), yukonite
(Ca-Fe arsenate), and arsenic bound to iron oxyhydroxides. At sites where ore roasting
was used, tailings and soils contain AsIII-bearing roaster-generated iron oxides, as well
as AsV-bearing iron oxyhydroxides generated by sulfide weathering. We have also
determined that antimony is present in multiple mineral forms and oxidation states in
mine waste impacted sediments. The detailed and direct identification of these As- and
Sb-bearing phases was achieved using a combination of synchrotron-based techniques
microanalytical including microXRF (X-ray fluorescence), microXANES (X-ray near
edge spectroscopy) and microXRD (X-ray diffraction) on target grains in polished thin
sections with a <10 micron spatial resolution.

Synchrotron-based techniques have also been applied household dust and shown that for
a sample from a background urban environment Cu and Zn are associated with distinct
matrices. Copper is dominantly hosted in an organic phase while Zn is associated with
inorganic minerals.

Each solid host of a metal or metalloid may exhibit  different response to bioaccessibility
tests, as these phases are known to vary in  solubility. The multiplicity of mineral hosts
has significant implications for the design of sampling programs that aim to obtain
representative ingestable material. Where applicable, synchrotron-based microanalysis
provides a tool to unambiguously characterize contaminants  in complex samples.

-------
Abstract 525
Bioaccessibility of Arsenic Adsorbed onto or Incorporated within Freshly
Synthesized Iron Oxide Minerals Using the Simulator of the Human Intestinal
Microbial Ecosystem (SHIME)
B. D. Laird1, T. Van De Wiele2, D. Peak1, W. Verstraete2, S. D. Siciliano1; University of
Saskatchewan, Saskatoon, SK, CANADA, 2University of Ghent, Ghent, BELGIUM

The bioaccessibility of arsenic adsorbed to amorphous ferrihydrite or incorporated within
amorphous scorodite was measured in the stomach, small intestine, and colon stages of
the Simulator of the Human Intestinal Microbial Ecosystem (SHIME), an in vitro
gastrointestinal model that incorporates the microbial community found in the human
colon. Arsenic concentrations adsorbed to ferrihydrite ranged between 500 and 9500 ppm
while the arsenic concentrations of amorphous scorodite mixed with freeze-dried iron
oxide ranged between 4500  and 450,000 ppm. The SHIME digests of these arsenic-
bearing minerals were used to construct arsenic dissolution isotherms for the stomach,
small intestine, and colon SHIME. Subsequently, the Kd of arsenic in gastrointestinal
fluids and the mechanism of concentration-dependent constraints on arsenic
bioaccessibility was evaluated. Additionally, the colon digest was repeated with sterilized
colon SHIME suspension to investigate the role gastrointestinal microorganisms on the
bioaccessibility of arsenic adsorbed onto or incorporated into iron oxide minerals. These
experiments investigated the mechanisms by which concentration and gastrointestinal
microbes affect arsenic bioaccessibility.

-------
Appendix C: Presentations

-------
 Introduction and Overview for Symposium "Use of In
 Vitro Bioaccessibihty/Relative Bioavailabihty Estimates in
 Regulatory Settings: What is Needed?
 Karen Bradham, Mike Beringer,
 Aaron Yeow (EPA)
 Pat Rasmussen (Health Canada)
 Rosalind Schoof (Integral Consulting, Inc.)
 Mark Cave (British Geological Survey)
 Symposium Roadmap

 • Welcome

 • Setting the stage

 • Definitions of bioavailability

 • ISEA 2006

 • ISEA 2007 overview and logistics

 - Panel discussion questions

 • Closing remarks
AEPA
>EFA
    Oral bioavailability of metals
    • Site-specific human health risk
     assessments

    • Risk assessments used to
     determine whether a contaminated
     site poses a current or future threat
     to human health that warrants
     remedial action

    • Oral ingestion of soil and dust -
     "risk driver" for human exposure to
     metal contaminants
   Exposed to contaminated soil - oral ingestion

   • Toxicity of an ingested chemical depends,
    on the degree to which it is absorbed from
    the gastrointestinal tract into the body

   • Metals can exist in a variety of chemical
    and physical forms

   • Not all forms of a given metal are
    absorbed to the same extent
     • Physical, chemical, biological
     • Matrix: metal from a contaminated soil
      absorbed vs. ingestion from dietary
      exposure
                                                                                                                                          iii »i,v,i!,t> Simpson et al.
                                                                                                                                          The Gut: Inside out.
                                                                                                                                          Physiology and biology
SB*


      Definitions of bioavailability and related terms


      • Many different meanings across various disciplines

      • Guidance for Evaluating the Oral Bioavailability of Metals in Soils
       for Use in Human Health Risk Assessment (OSWER 9285.7-80)

      • Bioavailability: "The fraction of an ingested dose that crosses the
       gastrointestinal epithelium and becomes available for distribution
       to internal target tissues and organs"



       http://www.epa.gov/superfund/bioavailability/guidance.htm
     Definitions of bioavailability and related terms continued


    • Bioavailability of metals in a particular matrix may be expressed as
      absolute bioavailability or relative bioavailability

    • "Bioavailability expressed as a fraction (or percentage) of a dose is
      commonly referred to as absolute bioavailability (ABA)"

    • ABA - "ratio of the amount of metal absorbed compared to the amount
      ingested"

                              Absolute
                            Bioavailability
                                                                                             http:/MMW.epa.gov/superfund/bioavailability/guidance.htm

-------
   Definitions of bioavailability terms continued
  • Relative bioavailability (RBA): "The ratio of the bioavailability of a metal in one
    exposure context (i.e., physical chemical matrix or physical chemical form of
    the metal) to that in another exposure context'

  • RBA - ratio of the absolute bioavailability of metal present in a test material
    compared to the absolute bioavailability of metal in some appropriate
    reference material

  • RBA is usually the most important for risk assessment
     •  extent to which the absolute bioavailability of a metal increases or
       decreases in context with the exposure matrix (e.g., water vs. soil)

  • "A related term, pertaining to bioavailability assessment, is bioaccessibility.
    This usually refers to a measure of the physiological solubility of the metal at
    the portal of entry into the body (IVBA)

    http://mvw.epa.gov/superfund/bioavailability/guidance.htm
 Use and application - oral bioavailability

* If the oral dose used for risk assessment is based
  on studies using metal administered in food or
  water, then the risk from ingesting metal
  contaminated soil might be over or
  underestimated
• Risk assessments that adjust for metal
  bioavailability may reduce the burden of
  unnecessary and costly remedial action
  • Small adjustment in oral bioavailability-
    significant impacts on estimated risks  and
    cleanup goals
  • Bioavailability data used to improve the
    accuracy of exposure and  risk calculations at a
    site
AEPA
 >EFA
   Methods for Assessing Bioavailability in Soil

     - In vivo methodologies
       • Quantification of metal present in various tissues and excrement
       • Used to develop quantitative bioavailability adjustments

     •  In vitro methodologies
       • Physiologically-based extraction tests
       • Measures bioaccessibility (e.g. solubility)
       • Used for screening purposes and reducing uncertainty

     - Mineralogical/speciation studies
       • Importance of solid phase distribution,  speciation and particle size
       • Provides supporting bioavailability/bioaccessibility information
    2006 Symposium

   "Childhood exposures to bioavailable and bioaccessible metals in soil and
     household dust in residential environments"
   Joint ISEE/ISEA Conference on Environmental Epidemiology and
     Exposure, September 2-6, 2006, in Paris

   Organized by Health Canada (Pat Rasmussen) and British Geological
     Survey (Joanna Wragg and Mark Cave)

   Abstracts published in Epidemiology 17 (6) pp S39-S42 (2006)
SERA
    Some observations from the 2006 symposium

    •  Participants found the International Society of Exposure Assessment
      (ISEA) to be perfect "home" for the cross-disciplinary nature of
      bioaccessibility/bioavailability research
    •  Many countries were represented, and in some areas similar views were
      expressed:
      - Everyone agreed that a set of criteria is needed, to assess both in vivo
        and in vitro methods
      - In addition to Canada, some countries (UK, USA, and EU) were
        allowing some limited site-specific adjustments
      -At that time, few would be comfortable allowing existing site-specific
        methods to be universally applied
      - Everyone expressed support for the annual ISEA BA session in 2007
 2007 ISEA Symposium "Use of In Vitro Bioaccessibility/Relative
 Bioavailability Estimates in Regulatory Settings: What is Needed?"

 • Morning session 9:00 am - 12:30 pm:
    - Site evaluation and application
    - In vivo research
    - Panel discussion session I

 • Afternoon session 1:30 pm - 6:00 pm:
    - In vitro research
    - Mineralogy and speciation
    - Panel discussion session II
    - Closing remarks

-------
&EFA


    Panel discussion and post ISEA

    • Two panel discussion sessions

    • Questions provided to all presenters/panelists prior to symposium

    • EPA plans to develop symposium proceedings to capture current state of
     the science and research needs
      - Proceedings will be posted on EPA website
        http://epa.gov/superfund/bioavailability/links.htm
      - Presentations included in symposium proceedings

    • Plan to publish a series of papers from symposium as a special issue
      - Some recommendations have been made for several journals
      -Symposium organizers will contact presenters following ISEA
        conference
                                                                                         vvEPA
                Questions for Panel Discussion Session I:
                Evaluation/site application and in vivo research
                                                                                           • For regulatory applications such as site clean up decisions, is an in vitro model
                                                                                           that is predictive (i.e., correlates with an in vivo model) adequate even if we
                                                                                           don't know why it is predictive, or is it important to know why a model is
                                                                                           predictive?

                                                                                           • Given that many animal models require use of soil concentrations exceeding
                                                                                           those of public health concern, what efforts are necessary, if any, to show that
                                                                                           bioavai I ability at these concentrations reflect oral absorption at lower
                                                                                           environmental concentrations?  In other words, how important is it to evaluate
                                                                                           whether bioavai I ability is concentration dependent?

                                                                                           • Should a single  animal model be specified, or are studies in a variety of
                                                                                           different animal models more likely to improve our understanding of relative
                                                                                           bioavai I ability in humans?

                                                                                           • Is there a single animal model that will work for all chemicals considering the
                                                                                           widely varying toxicokinetic profiles?

                                                                                           • Is relative oral bioavai I ability of chemicals in soil so routine that it is "ready" for
                                                                                           guidance/regulation?
AEPA
                Questions for Panel Discussion Session I
                and speciation/mineralogy research
AB*
                                                                                                          Questions for Panel Discussion Session II: In vitro
                                                                                                          and speciation/mineralogy research continued
• Is it sufficient for in vitro models to have correlation with animal results or is it
 necessary to make the models accurate physiologic mimics of human gut
 dissolution?

• Is it acceptable to use the terms "bioaccessible", "soluble", "migratable", and
 "extractable" interchangeably in the numerator of the % bioaccessibility
 equation?

• Should bioaccessibility values be used on their own without supporting
 information? If not, what additional information should be included for each soil
 sample analyzed (e.g., geochemistry)?

• How can we leverage resources to answer specific research questions to
 advance the understanding of bioavailability/bioaccessibility?
  - Develop Standard Reference Materials or Certified Reference Materials?
  - Round robin testing?
                                                                                            • In lieu of comparing in vitro model results to an in vivo model, what criteria
                                                                                             should be considered when evaluating whether a particular in vitro method is
                                                                                             appropriate for providing screening level data versus data to derive quantitative
                                                                                             site-specific bioavai I ability adjustments?

                                                                                            • What other metals, or metal species, are of interest for developing additional in
                                                                                             vivo and in vitro bioavai I ability assays? What criteria are used by regulatory
                                                                                             agencies for ranking or prioritizing contaminants?

                                                                                            • Is it feasible and cost effective to develop a single standardized in vitro method
                                                                                             for each metal for simulating oral bioaccessibility?  If so, what steps would be
                                                                                             needed for that to happen? What criteria would be used to determine the
                                                                                             difference between the methods?
For additional information, visit the following
websites:
 http://epa.gov/superfund/bioavailability/links.htm

 http://www.epa.gov/superfund/bioavailability/guidance.htm

 http://www.epa.gov/heasd/mdab/mdab.htm

 www.bgs.ac.uk/barge

 To obtain Proceedings from Health Canada Sponsored Workshops
 on Bioaccessibility/Bioavailability in Contaminated Site Assessment
 (2005, 2006, and 2007) go to: http://www.cntc.ca

-------

                                                                                  11
               Evaluating the Bioavailability
                of Soil-Borne Contaminants
                          at Waste Sites
                          ISEA Symposium
                           Durham, N.C.
                          October 15,2007
                   Mike Beringer & Aaron Yeow
               U.S. Environmental Protection Agency
                                                                                          Need for Additional Bioavailability Guidance
                                                                                         • Default assumption likely overestimates health risks
                                                                                            • Bioavailability is equal in soil, diet and water
                                                                                            • Relative bioa vail ability or RBA is 1.0
                                                                                            • Lead is the exception where default RBA is 0.60
        • Existing guidance supports bioavailability adjustments
           • Does not address when data collection should be pursued
           • Does not address how to evaluate site-specific bioavailability
        • Limited use of site-specific bioavailability information
           • Absence of rapid and inexpensive tools
           • Lack of criteria for evaluating alternative test methods

 Guidance for Evaluating the Oral Bioavailability
 of Metals in Soils for Use in
 Human Health Risk Assessment
^^^•^••^^^H

 • Limited in scope

 • Outlines a decision framework - series of questions
    • Is a validated method available?
    • Does the added value exceed the costs?
        • Addresses site-specific documentation
           • Basis for relying on the selected method
           • Data translation
           • Sample collection

        • Recommends criteria for evaluating alternative methods
                                                                                  I   I
                                                                                          Validation of Bioavailability Test Methods
         • Relying on ICCVAM criteria (Interagency Coordinating
          Committee for Validation of Alternative Methods)
            • http://iccvam.niehs.nih.gov/

         • Method validation criteria
            • Demonstrate method is reliable and relevant for its proposed use

         • Regulatory acceptance criteria
            • Method fulfills a specific regulatory need

         • Regulatory methodologies
            • Must satisfy both sets of criteria
            • Appropriate for making quantitative site-specific adjustments
N
II
        Method Validation Criteria
        (ICCVAM, 1997)
         • Scientific and Regulatory Rationale

         • Relationship Between Test Method Endpoint and Biological Effect

         • Detailed Protocol and Known Limitations

         • Within-Test Variability and Reproducibility Among Labs

         • Test Method Performance with Representative Agents

         • Comparison to Existing Test Method

         • Data in Accordance with Good Laboratory Practices (GLP)

         • Validity Assessment Data Available for Review

         • Independent Scientific Review
        Regulatory Acceptance Criteria
        (ICCVAM, 1997)
         • Independent Scientific Peer Review

         • Detailed Protocol with SOPs

         • Adequately Predicts Bioavailability and Demonstrates a Linkage

         • Representative Chemicals Tested

         • Generates Data Useful for Risk Assessment Purposes

         • Documentation of Strengths and Limitations

         • Robust and Transferable

         • Time and Cost Effective

         • Can Be Harmonized

         • Suitable for International Use

         • Reduction of Animal Use

-------
II
        Estimation of Relative Bioavailability of Lead
        in Soil and Soil-Like Materials Using
        In Vivo and In Vitro Methods (Lead TSD)
        • Describes in vivo and in vitro methodologies
           • Juvenile swine model
           • Simplified bioaccessibility method


        • Characterizes 19 soil and soil-like test materials
           • Mineral phase
           • Particle size distribution
           • Matrix association
           • Clear differences in RBA between materials
           • Data not sufficient for predictions based on mineral content alone


        • Evaluates the correlation between both methods

        Lead TSD Transmittal  Memo
        • Evaluated both methods using ICCVAM criteria
           • Broad range of relative bioavailability
           • Variety of mineralogical forms
           • Pairwise comparison shows a good fit (r2=0.92)

        • Both methods considered regulatory methodologies
           • Weight-of-evidence determination
           • Method validation and regulatory acceptance criteria achieved
           • Appropriate for use in site-specific risk assessment

        • Outlines limitations and considerations for use
           • Quality assurance
           • Sample lead concentration limits
           • Particle size and soil mineralogy
           • Extrapolation to adults
           • Valid for soil samples from mining and milling sites
11
        Lead - Correlation Between In Vivo RBA
        and In  Vitro Bioaccessibility (IVBA)
                                                                                   I   I
                                                                                            Future Activities
        • Formation of Bioavailability Committee
           • Information archive
           • Provide technical support to the USEPA Regions
           • Develop additional guidance
           • Review new bioavai I ability methods

        • Evaluation of other metals
           • Formal consideration of arsenic bioavailability data
           • Possible derivation of default values for other metals

-------
     Method Development and the
  Application of Oral Bioavailability
       Data in Risk Assessments
               Rosalind A. Schoof

               Integral Consulting, Inc.
                Mercer Island, WA
              rschoof@integral-corp.com
New EPA bioavailability guidance
calls for use of validated  methods

• How is validation defined?
• What are validation requirements?
• What are regulatory acceptance requirements?
« What is the process for method development
  and validation?
• What methods are  already considered validated
  based on history of use?
                  ISEA 2007
  Validation - Definitions

  Scientific process designed to characterize
  operational characteristics, advantages, and
  limitations of a test method

  The process by which the reliability and
  relevance of a test method are evaluated for
  the purpose of supporting a specific use
  Validation requirements

  Relevance - Linked to mechanism of
  toxic effect and proposed uses

  Reliability - Objective measure of
  method's intra- and interlaboratory
  reproducibility
Who oversees validation efforts?

• Interagency Coordinating Committee for
  Validation of Alternative Methods (ICCVAM)
•   http://iccvam.niehs.nih.gov
• Organization for Economic Cooperation and
  Development (OECD)
• American Society for Testing and Materials
  (ASTM)
ICCVAM validation criteria
http://iccvam.niehs.nih.gov/docs/guidelines/validate.pdf
•Scientific and regulatory
rationale
• Endpoint relation to
effect of interest
• Detailed protocol
•Within test variability
and reproducibility among
labs
• Performance demo with
reference chemicals
• Data comparison with
standard test
• Limitations described
•GLP data
• Data available for review

-------
ICCVAM regulatory acceptance criteria
http://iccvam.niehs.nih.gov/docs/guidelines/validate.pdf
• Peer review
• Protocol with SOPs
•Measures endpoint of
interest, and linkage with
existing test
• Representative chemicals
tested
• Useful for risk
assessment
• Strengths and limitations
identified
• Robust and transferable
•Time and cost effective
• Can be harmonized
• International acceptance
possible
•Minimizes animal use
What is the process and when is a
method validated and implemented?

•  Process: Research > development >
  pre-validation > validation > review >
  agency consideration > implementation
•  Validated when its performance
  characteristics, advantages, and
  limitations have been adequately
  documented for a specific purpose.
 Types and uses of test methods

 • Definitive tests - Used to measure toxic
  effects
 • Screening methods - Support
  preliminary hazard decisions
 • Adjunct tests - Used to increase the
  information base and/or aid in the
  interpretation of results from definitive
  methods
                                Evolution of test methods

                                • Development of study design
                                • Refinement of test protocol
                                • Assurance of transferability
                                • Determination of performance
                                  characteristics
Currently accepted methods

• Considered validated based on history
  of use
• Applies to many approaches to
  measuring oral bioavailability of
  chemicals
                                Currently accepted in vivo methods
                                of measuring bioavailability
                                • Blood concentration over time
                                 (area under the curve, or AUC

                                • Absorbed fraction in urine
                                 and/or tissues

                                • Comparison of tissue concentrations

                                • Unabsorbed fraction in feces

-------
 Variations in gastric anatomy
Validation for toxicity tests vs.
bioavailability methods
• Are validation criteria for development of
  alternative toxicological methods appropriate
  for methods of testing relative bioavailability
  of chemicals in soil?
Validation/acceptance issues for oral
bioavailability studies of metals in soil

• Endpoint of interest is relative
  bioavailability for oral exposures
• For some metals there is no standard
  method available for comparison
• SBRC in vitro method meets validation
  and regulatory acceptance criteria
• Should animal studies continue to be
  used?
Does the in vitro method meet
validation criteria?
 > Rationale and
  relationship of endpoint
  to effect of interest are
  documented
 • Detailed protocol
  available
 • Reproducibility
  established in
  interlaboratory study
 • Performance
  demonstrated for
  representative chemicals
• Comparison of
 performance with
 existing test established
• Limitations described
> Data quality documented
• Data reviewed both in
 peer-reviewed
 publications and in
 independent peer review
 process
Bioavailability method development
questions for metals

• How reliable are oral in vivo study methods?
• Should validation of in vitro protocols be
  required on a metal-specific basis?
• Should results of unvalidated methods be
  considered in risk assessments?
  If yes, how?

-------
                            Exponent
 _ pone
 Relative Oral Unavailability of
 Arsenic From Soils:
 in vivo and beyond
              Yvette Lowney
              Presentation to ISEA
              October 2007
                                          Presentation Topics
in vivo research
- cynomolgus monkeys
in vitro extraction testing
  SBRC method
  Other chemical extractions
  RIVM/Unified Barge
Geochemistry and soil characterization
Future efforts
Research Funding
   SBRC
   SERDP
   Industry sponsors
   UK Environment
                                          Urinary Arsenic Excretion:
                                          Timeline Following Exposure
Integrated Assessment of Beneficial Uses and Development of Candidate Cleanup Levels
                                    1

-------
                                Exponent
Study Design
 cynomolgus monkey
 Low arsenic diet prior to dosing
 Dosed with slurry of soil in water
   Soil dose 
-------
                            Exponent
in vitro Extraction Method

 SBRC, phosphate, hydroxlamine
 buffered "gastric" solution
 <250 um particle size
 1 g soil: 100 mL fluid
 End-over-end rotation at 37°C
 1 hr
 Filter (0.45 um)
 ICP-MS for As
SBRC Method at pH 1.5 and 2.5
           ARSENIC BIOACCESSIBILITY (°
Integrated Assessment of Beneficial Uses and Development of Candidate Cleanup Levels
                                     3

-------
                       Exponent
Integrated Assessment of Beneficial Uses and Development of Candidate Cleanup Levels
                              4

-------
                               Exponent
Status of iw r/'tro Method

 Indications that in vitro methods may
 be predictive of in vivo results for RBA
 of arsenic
 Method validated for lead correlates
 well with arsenic for most soils
 Need method that is predictive for all
 soils across a diversity of soil types
 RIVM and UBM correlate with each
 other, poorer relation with in vivo data
 Avaijable approaches don't provide 1:1
 relation between in vitro and in vivo,
 but could find good correlation
Soil Characterization Data


 Conventional parameters
   Arsenic concentration
   TOO
 - Metals
 -SoilpH

 Arsenic source

 Particle size distribution

 Arsenic mineralogy (18 phases)

 Extractable iron oxide
Integrated Assessment of Beneficial Uses and Development of Candidate Cleanup Levels

                                        5

-------
                                   Exponent
Soil Arsenic Mineralogy Data
                                                     Research Status
                                                     ' 117 VIVO
                                                      - Robust database suggests RBA <30%
                                                        More data?
                                                     • in vitro
                                                      - Method validated for lead correlates well with arsenic
                                                        for most soils
                                                        Need method that is predictive for all soil types
                                                        Available approach  doesn't provide 1:1 relation
                                                        between in vitro and in vivo, but could find good
                                                        correlation
                                                        Progress is potentially rapid
                                                     1 Predictive models
                                                        Currently no model that is robust across all soil types
                                                        Theoretically possible
                                                        Likely to be 'informed' by in vitro method
                                                        development
Integrated Assessment of Beneficial Uses and Development of Candidate Cleanup Levels

                                             6

-------
                       Exponent
Integrated Assessment of Beneficial Uses and Development of Candidate Cleanup Levels
                              7

-------
                                                                                      UilJjiy
                                                             VirDiiiiJiiy—iiDDiDD j/jii-iJ:; !A=, rjd, rjr,
                                                             Pb, '••]], o/:J-i/jJfj fjo;;j{jou/jd:; (dJo/J/j:;^
                                                             DD'fj iiud Au-, Pd- iiud AcHjiiijo^iiriidE:
                                                             JuviuiJE: DWiui DurrogiiiE: for 'jhUd/i/j,
                                                             jJil'j'VE: JLJViJjijE: {jicJD UDid i/J  -ill 5PA il/jd
                                                             jJrjJ ={jojj=orsd ^UjdiiD,
                                                             OrnJ i/iJO^u/E: for 'J2-'J4 eoijDEJcU'fJys;
                                                             diiy=  si! "J do^E: Jeveii;,
                                                            Statistical Power of the Study > 90%.
                                                            Doses selected to reflect low-dose
                                                            human exposure (25-160 ug/kg BW).
                                                            Multiple responses to assess RBA—
                                                            blood, urine, liver, kidney, and bone.
Absorption of Soil Lead
         by
   Immature Swine
                                                           PbMn(M)O
                                                           average RBAs (25% - 75%) are associated with PbO,
                                                           PbFe(M)O, PbPO4, and Pb-Slags
                                                           lower RBAs (<25%) are associated with PbS, PbSO4,
                                                           Pb(M)O, PbFe(M)SO4, and metallic Pb


-------
                                            Conceptual Model

General Subchronic As Study Design

-------
   coal fly ash
   agricultural/orchard pesticides
   wood preservation
                                                             Absorbed As primarily excreted in urine
                                                             Urinary Excretion Fraction fUFr) is an
                                                             approximation of trie oral AF or ABA,
                                                             USr does not account for As excreted in bile or
                                                             -\s distributed to tissue compartments.
                                                             "•'BA of 2 orally dosed materials ftest an
of UEF(As.test) / UEF(A   fx. Keep in mind this is
really a ratio of slopes of As excreted as a
function of As dosed.
       'kriamiJs 21 mJ
7-10 Qmups o/4-j
[jigs dosed for-12-14
corjsecuiive diiyj
Abaor-bsd As
=s'iirrjai=d by J-.s
e/.creied irj urirje (24 or
43 br;—US:^—urirjyry
excretion Traction
Urinary As excretion—a
linear function of dose
and independent of
time after day 5

-------
                                                            Rssulis Coal Combusion By-
                                                                        Prod LldS

                                                         Using sodium
                                                         arsenate as si relative
                                                         frame of rsfsrerics,
                                                         trie arsenic R3A
                                                         sstirnatsa 5ire
                                                         for Test Material 1
                                                         and 50% for Test
                                                         Material 2.
                                                                                            1.jnoT.-.T«!Mawii:
Eidpoo:
D«i«1
;^?:.
Dri i: U
AC 5m
InnuttK R3A i'?;*> "cdii»i;« £Ur-*l>
7t.: .^as:: .
.TTCd
0*4^«-0.!J)
. ;i •.!:-; •'.
M cr-:.':.
;~:(ca-;"*i
T«tywR»;:
7?:-".
35IJ3-«-eJf]
r: .+; •. ••
0«CH5-C-!3J
}»[94C-cn}
                                                             Jiiiiwsi Tow;j3hj|j rJsri 3Jsi33
Usincj sodium
arssnais as a relsiii /3
frame of rsfsrsnoe,
She arssriio P.3A
sstirnaiss sirs
for Test Material 1
and 26% for Test
Material 2.

-------
Absorption
                                                                                                          Gastrointestinal microbial metabolism
                                                                                                            (methylation, demethylation, thiolation)
                                                                                                               Postabsorptive metabolism
                                                                                                                  (methylation, thiolation)

-------
                                                                                 Water bottle  Food hopper

                                                                                                 Feces collector
                                                                                   Urine collector
                                                                                                        Soils
                                                                                                                      Diet
         110700 AIM 93G Pljl
-------

                                                      S 2
                                                      8
                                                           \\
                                                           \\
\ \
                                               NaAsV 170/2 170/5 170/4

-------
                                           Sift*
                                         I  «MA»
                                           SDMAs
M*At(V)    170H     170/5     170/4

-------
                                     Environment
                                     taencj
 In-vitro Bioaccessibihty of Soil-
 borne Contaminants: An
 Environment Aaencv Perspective
                                                                                                  Environment
                                                                                                  Agency
                                                                             Outline
                                                                                                         £01
                                                             Risk assessments and land contamination
                                                             Review of currently available in-vitro methods
                                                             Conclusion to date
                                                             Forward Look: what is required?
                                                                 Risk assessment approach
                                                        Preliminary Screening
                                                                           2RA (e.g. SG\
                                                                                DQRA (site specific bio-availability
                                     Environment
                                     Agency
              Definitions
Absolute bioava liability: Is
fraction of an administered
 •  Needs in-vivo studies
Relative bioavailability: Cor
different forms of a chemic;
containing the chemical
 •  Needs in-vivo studies
                    measure of the uptake or the
                    ose absorbed by the body.
                        ve bioavailabilities of
                        - different exposure medic
Oral bioaccessibility: Is a measure of dissolution, or the
fraction of contaminant released into solution from the sc
during digestion
 • Attempt to measure in in-vitro model
                                                                                                       nmeni
                                                                                                  Agency
                                                                      hy in-vitro?
                                                             **'*
Technical expertise
Reproducibility
Ethical reservation


Absorption is likely to be
dependent on solubility

-------
Environment Aaencvs work Droaramme
                                                                                                        N Environment
                                                                                                        FAgency
Questions
                                                                                                                cm
                                                                1) What are the different in-vitro methods in the UK?

                                                                2) Can they produce comparable results?

                                                                3) Can they adequately predict bioavailability?

                                                                4) Can one method be suitable for different chemicals and
                                                                  different soils?

                                                                5) Do we have sufficient awareness of these issues and
                                                                  their importance?
          Q 1: Methods available
                                                                   Q 2: Inter-lab data comparability
                                                                  Study undertaken with UK labs producing bioaccess
                                                                  data - available at
                                                                  ileven labs participated (two from overseas)
                                                                  Three prepared samples from UK supplied in triplicate (
                                                                  one human tested lead contaminated sample from
                                                                  Maddaloni etal. 1998)
                                                                  Labs were asked to analyse for As, Pb and Ni using the
                                                                  normal protocol.
                                                                  'CONTEST soil proficiency-testing scheme' followed

                                                                  in-vitro data evaluated as 'consistent' or 'inconsistent'

-------
                          /5?X Environment
                          V^iV'Agency
Data comparability (summary)
      (excludes overseas and one UK l
   Test soils   Arsenic
            R: 13-88     R: 1462-8219  R: 1-5
            Med: 18     Med: 1911    Med: 2
            R: 121-7011  R: 2920-84979 R: 5-25
            Med: 194    Med: 10480   Med: 9
            R:5-9      R: 348-542    R: 1.35-2
            Med: 5      Med: 477     Med: 2
                                                                                                Environment
                                                                                                Agency
                                                                                                      em
                                                  Can UK labs produce comparable results?


                                                  No, largely due to variability in the types
                                                  of in-vitro methods used

                                                  (But labs using the same method and same
                                                  operating procedure produced comparable results)
   Q 3: Predicting bioavailability
 Study undertaken with Exponent USA (report in
 prep.)

  • Aim: Evaluate selected in-vitro methods for their
   ability to predict bioavailability
  • Used 13 in-vivo (Cynomolgus monkey) tested US
   soils with arsenic bioavailability data obtained from
   a previous study (Roberts etal. 2007)
  • In-vitro data produced was studied against in-vivo
   bioavailability data
                                                   Q 3: Predicting bioavailability for arsenic
                                                                              7 (0.37 excl. outlier)
                                                                Intestine phase
                                                                              5 (0.32 excl. outlier)
                                                        PBET (Ruby <* ol. 1996)"  0.18

                                                       SBRC/SBET (Drexler 1998)" 0.27
                                     Environment
                                     Agency
Can in-vitro data adequately predict
bioavailability?

Not adequately for the soils tested. For
UK soils it is unknown as none have
gone through in-vivo studies
                                                                                     (3& Environment
                                                                                     *9jf Agency
                                                      Q 4: Applicability of one method to
                                                               various chemicals
                                                          In-vitro methods
                                                    (Drexler 1998)

                                                     SBRC/SBET
                                                    (Drexler 1998)
                                                                                  R2 (range 0-1)
                                                                 0.17a (0.37 <

-------
                                        Environment
                                        Agency
                                        \ Environment
                                     Xi  Agency
Q 5: Do we  have sufficient awareness
                 in the UK?

 Can one method be suitable for different
 chemicals and different soils?
 Questionable based on evidence
 currently available
   Perception that bioaccessibility and bioavailability are
   the same thing

   There are reports of extrapolation of bioaccessibility
   data from literature or different sites
   Inconsistency in the practice (e.g. lab procedure) and
   use of data in risk assessment
                                       S Environment
           Conclusions to date
Ability to predict bioavailability by m-vitro methods used in the
UK is uncertain for UK soils

Considerable inter-laboratory variability of in-vitro data
Laboratories use same method irrespective of
 • chemical
 • chemical form
 • matrix
Contaminants of concern differ from country to country
In-vitro bioaccessibility testing is an ongoing research area
    Forward look: what is required?
   More needs to be done to develop in-vitro methods
   including validation with appropriate in vivo data

   What can be done to increase confidence in in-vitro
   data?
    • Multiple lines of evidence to compliment in-vitro
     bioaccessibility methods/data
                                                                                                            nmeni
                                                                                                       Agency
                                                                                           required?
                                                                 Focussed discussion to identify and define what are
                                                                 the essential data gaps

                                                                 International harmonisation through a framework of
                                                                 partnership (repeat and working in part would delay
                                                                 achieving the common shared objectives)

                                                                 More linkage between research scientist and
                                                                 regulatory scientist (and with Policy makers)

-------
HC Science Forum  Oct 31 2006
                             Pat E. Rasmussen
      Outline of Presentation
         Consumer Safety
                Branch
              Measurement of Metal
             Bioaccessibility in Urban
                Household Oust and
            Corresponding Garden Soil
Differences between house dust and soil

High variability of bioaccessible metals in dust

Solution - simple categories of bioaccessibility:
  low, medium and high

What is needed? significance for risk  assessment
                                         CanadS
           The "tracking-in" models to
           estimate indoor  exposures
      Coarse fraction of soil
                                                                              leaves, twigs


                                                                                 stones
                                                                               agglomerates
                                                                               natural and/or
                                                                               manufactured
          Coarse Fraction of House Dust
                                         textiles

                                      "I am Canadian"
                                        beer cap
                                         plastics

                                        metals


                                       organics
   Sample preparation technique
          used for soils  ..

-------
HC Science Forum  Oct 31  2006
              ....are applied to house dust
                                  To compare indoor dust with
                              outdoor soil,  analytical approaches
                                       must  be consistent


                                                                                             Same size fraction
                                                                                                To calculate indoor/outdoor
                                                                                                metal ratios (dust/soil)


                                                                                             Same analytical approach
                                                                                                Aggressive digestion for
                                                                                                "total metal" determination,
                                                                                                to ensure equally efficient
                                                                                                recoveries in different
                                                                                                media
                                                                                                Weak extraction to
                                                                                                estimate "bioaccessible
                                                                                                metal" fraction
               Metals in Dust -  References
                               Detailed Ottawa studies: the unique
                                 geochemical profile of  house dust

                                                                                        After drying and sieving to fine
                                                                                          fraction, settled dust may look
                                                                                          the same as soil.

                                                                                        But...
                                                                                        •  Key metals have higher total
                                                                                          concentrations.
                                                                                        •  Bioaccessibility of key metals
                                                                                          higher in dust.
                                                                                        •  Organic content higher

                                                                                        Dust is very heterogeneous.
                                                                                        •  Particle size
                                                                                        •  Speciation / mineralogy
           Ottawa:  an example of  "urban  background"
                  dust
                  410
                     soil   dust
               10     18    16
                            soil    dust
                            2.2    2.7
dust
 28 soil
  5.1   dust
       1.1
                                                soil
                                                0.18
                   Zn (ppm)    Ni (ppm)    Cu (ppm)   Org C (%)   Inorg C (%)
                             High  variability of bioaccessible
                              metal  in dust compared to soil





-------
HC  Science Forum Oct 31 2006
                 Influence  of particle size
Bioaccessibility as a function of mineralogy
                                                                  Zinc species


                                                                  Zn hydroxyl carbonate
                                                                  Zn/Fe-oxides
                                                                  ZnS
                  <36ftm    80-150 fim

                   % of total zinc in house dust

                      52          65

                      22          16

                      26          19
            Influence of Speciation (XAS)
 Influence of Speciation (XAS)
                Summary of histograms
         Sources of variability
 - Bioaccessible metals in house dust-
            Particle size is an important control on metal
            concentration.
            Metal concentration commonly increases as particle
            size decreases.
            Metal bioaccessibility commonly increases as
            particle size decreases.
            However, particle size is not the only control on
            concentration & bioaccessibility: the opposite
            trends may occur depending on metal Speciation.
            Analytical reproducibility is improved using smaller
            size fractions.

-------
HC  Science Forum Oct 31 2006
            Selection of Sampling Method -
              Depends on Purpose of Study

Adaptation  of Toy Safety Protocol

            for House  Dust

  European Standard EN 71-3: for the migration of
   certain elements from toys

  • Used in Product Safety lab at HC
  • Children as target population
  • Extraction uses only dilute HCI  (pH 1.5) to
   simulate stomach acid
  • No added ingredients (complexing agents)
  • Omits mouthing/mastication - assumes toy is
   small enough to be swallowed
  • Omits passage through intestine
         Modified Toy Safety Protocol Yields Similar
             Results to Other Gastric  Simulations



                                                                 100


                                                                  80


                                                                  60


                                                                  40


                                                                  20


                                                                  0
                                                                           Copper in NIST 2583 Indoor Dust
                                                                             error bar = 1 sd; n = 6 replicates
                            RSD=18 %
                            at ratio of
                            5000
                                                                         1000    2000    3000    4000    SOW   6000

-------
HC Science Forum Oct 31 2006
                                                                    Modifications to US-EPA 3051
                                                               microwave digestion method for total metals
                                                                        in dust and soil samples

                                                               Increase microwave digestion time to 30
                                                               minutes ramp time & 30 min hold time
                                                               (EPA3051 specifies 5.5 min total)
                                                               - recovery improved by 15-20%

                                                               Increase acid volume to sample mass
                                                               ratio to at least 1000 (EPA 3051 specifies
                                                               ratios from 20 to 100)
                                                               - recovery improved by 30-60%
              Sample mass to acid volume ratio
                            HIST 1648
                            Urban PM
                         Ni    Cu    Zn


                        • 10 mg • 30 mg • 100 mg

-------
HC Science Forum Oct 31 2006

          LOW
          19% or less
LOW
19% or less
          HIGH
          60% and higher
HIGH
60% and higher
Zn
                                  Indoor: Outdoor Ratio
    Arguments for measuring metal
     bioaccessibility in indoor dust
          for risk assessments
             Bioaccessible extraction      13   7   23

-------
HC Science Forum Oct 31 2006
                 What is needed?
                                                          Relevance to Risk Assessment
           Thanks to our great lab team
                                                                Acknowledgments & Funding


               Michelle Nugent, Christine Levesque, Marc Chenier,
                   Jianjun Niu,    Monique Lanouette
                                                                     http://www.cntc.ca
                                                                                                     7

-------
                The use of in vitro bioavailability
                in human health risk assessment

                Scientific research and application by policy makers
                                                                           Human risk assessment of

                                                                           contaminated soils in the Netherlands

                                                                           • Intervention Value for substance in soil derived from MPR
                                                                           • MPR: Maximum Permissible Risk (mg/kg/day)
                                                                             - Exposure level above which there is a potential health risk
                                                                             - Based on toxicological and epidemiological studies
                                                                           • For most substances it is implicitly assumed that:
                                                                             - Oral bioavailability in all exposure matrices = oral
                                                                              bioavailability in the studies underlying the MPR
                                                                             - Relative bioavailability =1
Human risk assessment of lead-
contaminated soils in the Netherlands
• Most substances: relative bioavailability = 1
•Lead
  - Recently changed
  - Intervention value for lead in soil 530 mg/kg
    for scenario "living with garden" in the
    Netherlands
  - Relative oral bioavailability = 0.74 for soil
    ingestion
     • P80 for relative bioavailability of lead in tested
      soils (soils low in organic matter)
     • P80 for difference between fasted and average
      state in gastrointestinal tract
Status application relative oral
bioavailability in risk assessment

• In the assessment of the risk of contaminated soils a
 default value for the relative bioavailability of a
 contaminant from soil is applied (0.74)
• The default value can be changed if reliable site-specific
 information is available
• Recommendation by RIVM to government to accept the
 use of in vitro determined bioaccessibility for estimation
 of relative bioavailability factor (2006)
• Government will probably seek advise from other
 institutes (Health Council, Technical Soil Committee)
Dutch Kempen area
• Zinc smelter for over a century
• Area contaminated with cadmium, lead, arsenic, etc
Experimental set-up


• Intention: relative bioavailability as input for
 an area specific policy


• What is bioavailability of lead and arsenic in
 soils and slags?
• 20 zinc slags, 16 soils
• In vitro determination of bioaccessibility
• Estimation of relative oral bioavailability

-------
In vitro determination of bioaccessibility
             astric      duodenal juic
                                       Estimation relative oral bioavailability
        5 min     2 h      2 h
       pH ±6.5  pH ±1.2   pH±5.5
chyme   Pellet
     (destruction)
                                                                         small intestine  F Fraction of F absorbed by small intestine
                                                                          portal vein

                                                                                    F = Fraction reach

                                                                        Internal exposure F = Fb xFaxFh
                                                                                    Fh = Fraction of Fapassing liver without being metabolised

                                                                                                    nic circulation = bioavailable fractio
 Estimation relative oral bioavailability of
 lead
 Described in Oomen et al.
 „  ,  •   r   Fsoll     FbsoilxFasoilxFhsoi
 Relative F=   sml  -    '     '      '
 Relative F=
             0.5
                                       Research on oral bioavailability lead
                                       and arsenic from zinc slags
                                       Relative bioavailability lead and arsenic from soils and
                                         slags from the Dutch Kempen area
                                                                             Soils

                                                                             Slags
                                                                                         Lead
                                                                                                    Arsenic
                                                    0.83 ±0.11   0.15 ±0.13
                                                    (n=13)       (n=12)
                                                    0.35 ± 0.15   0.02 ± 0.01
                                                    (n=l7)       (n=11)
 Application in risk assessment
 • Intention: relative bioavailability as input for an area specific
  policy
 • In this case not applied

 • Lead
   -  Relative bioavailability of lead from Kempen soils (0.83) and slags
     (0.35) was high
   -  Minor consequences for area specific policy
   -  Soil concentrations in the Dutch Kempen may lead to ecological
     risks, but in almost all cases not to human health risks
                                       Team

                                       •RIVM
                                          - Marco Blokland
                                          - Esther Brandon
                                          - Menno Duits
                                          - Werner Hagens
                                          - Johannes Lijzen
                                          - Agnes Oomen
                                          - Adrienne Sips
                                          - Frank Swartjes
                                       •ABdK
                                          - Eric Kessels
                                                                                                      V-  ^

-------
 Assessment of the use of dynamic human stomach models
  for in-vitro measurement of the bioaccessibility of Arsenic
  and Chromium in soils - Can they replace animal testing?
         Mark Cave1, Helen Taylor1, Joanna Wragg1, Andrew Broadway2

           1 British Geological Survey, Keyworth. Nottingham, NG12 SGG UK
2 University of Edinburgh, School of GeoSciences, Crew Building, West Mains Road, Edinburgh, EH9 3JN,
                         Scotland, UK.
Kingsley Dunham Centre
Keyworth
Nottingham NG12 SGG
Tel 0115 936 3100
                                                                       Investigating human digestion
Human clinical studies - EXPENSIVE and
time consuming (not to mention gross
intervention effects and ethics)

Animal studies - may provide an alternative •
ethical considerations and concerns on
relevance to the human

In vitro models - a requirement for accurate
in-vitro screening tools - currently
oversimplified

          Dynamic  in Vitro models


     •  TIM-1 and Tiny TIM -TNO Quality of Life, Zeist, The
       Netherlands


     •  Model Gut - Institute of Food Research, Norwich, UK
                     TIM-1 system
                stomach and small intestine
                  body temperature
                  gastrointestinal pH curves
                                                                                                   seccetion ol gastric add and salivary
                                                                                                   and gas t FK enzymes

                                                                                                   secretion of We. pancreatic juice

                                                                                                   absorption of digested products and
                                                                                                   mi*

                                                                                                   continuous proces* control and data


                                                                                                   modular set-up, htgh flexibility
                                                                                                        _^k  mf± it.  fK
                  TIM Validation
                                                                                                               Batch in-vitro

                                                                                                               TIM

                                                                                                               Minipig in vivo

-------
                           Maddaloni soil
                                                             FASTED
                                                             FED
                                                                                                BGS reference  Material
                                                                                        .-
                                                                                                 Total As104±3mgkg-1

                                                                                                 Total Pb 79 ±3 mg kg-1
                                                                             Chromium In Glasgow:

                                                                             • Between 1830 and 1968 the Rutherglen area
                                                                             of Glasgow was home to the world's largest
                                                                             producer of chromium chemicals.

                                                                             • Chromite ore processing residue (COPR) and
                                                                             waste from the chemical works was used as
                                                                             landfill in south-east Glasgow during the 19th
                                                                             and 20th Centuries.

                                                                             •An estimated 2.5 million  tons dry weight of
                                                                             COPR  waste was sent  to landfill  around
                                                                             Southern Glasgow and South Lanarkshire.

                                                                             •In Glasgow, 15 contaminated sites have been
                                                                             identified. The  average Cr(VI)  concentration
                                                                             found was  700 mg/kg, with the highest being
                                                                             15,600mg/kg.
                         sibilityin COPRContar
 4000


 3500


 3000


 2500


? 2000


 1500


 1000


  500
                                                                                         Institute for Food Research (UK)
• First model to combine emerging
knowledge of the physical/
mechanical aspects of digestion
with the biochemistry  in a single
predictive system

• It is the only simulation available
that can handle real food and
pharmaceutical preparations

• Also the only model that allows
access at any stage of 'digestion'
permitting sample collection and
analysis at anytime point

-------
          Echo-Planar
 t" ;l        Magnetic
'<^*-'' Resonance Imaging
               (EPI)
                     Meal in fundus
  *   «•;    t."    *l   0  I
                                     Meal in an
  §(•   s.i    s^    ci    K    ?
                            Meal in antrum
                                                   Dilution Maps and Gastric Mixing
                                                                36 min     48 min     72 min
                                                          Marciani etal. Journal of Nutrition,130. 2000
                    Analysed samples are composed of EQUAL vi
                                                                      Water Blank and Soil Run w ilh Water
                                                                      Volume of Acid and Enzyme will deper

                                                                         British Geological Survey
                                                                             C Samples
                                                                  10.00  15.00  20.00  25.00  30.00  35.00  40.00

                                                                          Time (min)
Soil 1 Digest in Water
                      IFR Model Gut
                                                                           Conclusions

                                                                Dynamic in-vivo models show promising results.
                                                                As yet we do not have a definitive study for soils.
                                                                At present they probably would not replace batch in-
                                                                vitro methods but they could be used as a 'reference
                                                                methods'.
                                                                Commercial units are likely to be available in the near
                                                                future.
                                                                Could provide a more cost effective, ethical, and
                                                                scientifically valid alternative to animal testing.

-------
                                       .ML.
               Soil 1,11300 mg kg'1 As (Calciner soil)
                                                                                                .ML.
                                                                         Soil 8,1180 mg kg'1 As (Iron slag soil)
 3500

 3000
f
„ 2500

; 2000
!
1
J 1500

 1000
                                        CISED extractiot
                                                          | 300
                                                                                                         CISED extractio
       S- Ca  Ca- S- As   Fe- As    Fe    Fe- S   Fe- As
                                                                      Pb-Ca   Ca    Ca   Si-Fe  Ca-S  Si-Pb  Fe-Ca  Fe-Ca
                                  IFR Model Gut
                                                                                           'olume against Tim
                                                                                                Soil Run with Mlk
                                                                                             ^e of acid and enzyme will depend
                                                                                              on how the pH responds.
                                                                                              British Geological Survey
                                                                                                  6 Samples
                                       iDigea in Full Fat Milk
                                                                                    20.00  30.00  40.00  50.00  60.00  70.00

                                                                                             Time (min)
i 35

1 30

I 25

I 20
LU
  15
y = 0.7337x +3.8731
  R2 = 0.7753
         5     10    15    20    25     30
                       Relative Bioav ail ability %
                                                           TIM dynamic
                                                           model
                                                                                                 .1 t hi III I I' < Ul

-------
            Tiny - TIM
Small intestine compartment
                                  Hollow fibre MW cut-off filte
Simulated Peristalsis in the TIM model
                Bioaccessibility of Chem icals
                       Not New
Jacob Helm (1797-1802) Investigations of the human digestive
system and the physiologic manifestation of digestion using
gastric and enterocutaneous fistula [Kisch B (1954) Jacob Helm's
Observations and Experiments on Human Digestion: A One
Hundred and Fiftieth Anniversary, J Hist Med Allied Sci, IX, 311-
328]
Carson and Woelfel (1913) The solubility of white lead  in human
gastric juice and its bearing on the hygiene of the lead. Am J Public
Health. 3, 755-769.
Reiman C.K & MinotA.S. (1920) Absorption and Elimination of
manganese ingested as oxides and silicates. Journal of Biological
Chemistry, 45,133-143

-------

-------
   Assessing Contaminant (Bio)availability in Soil
       when In vitro Gastrointestinal Methods
                are the Only Option

                       Nick Basta
      Professor of Soil and Environmental Chemistry
      School of Environment and Natural Resources
                  Ohio State University

                   Dr. Kirk Scheckel
     National Risk Management Research Laboratory
                U.S. EPA, Cincinnati, OH

                   Dr. Karen Bradham
         National Exposure Research Laboratory
          U.S. EPA, Research Triangle Park, NC
                                                                 U.S. EPA
                                            Guidance for Evaluating the Oral Bioavailability of
                                         Metals in Soils for Use in Human Health Risk Assessment
                                                         OSWER 9285.7-80, May 2007

                                            Recommended Criteria for Validation of Test Methods
                                                           adapted from ICCVAM

                                        "Data generated adequately measure or predict the toxic endpoint of
                                        interest and demonstrate a linkage between either the new test
                                        and effects in the target species."

                                                  In vitro gastrointestinal (IVG) method must
                                                 be correlated with an acceptable in vivo model
                                                           IVG must be predictive
     Correlation of IVG method with an in vivo model
       ii
       U) ra
         § m
         at.
                                                                             IVG Method Correlation Studies
                                                                             most on highly contaminated soils
                RBA As = 0.942 IVG + 7.11 r = 0.91*
                     IVG Gastric As
                   % Bioaccessible As
                                     Basta et al. 2003.
                                     Grant R825410 Final Report.
                                     submitted to U.S. EPA ORD
                                                                    Most correlation studies conducted
                                                                    on highly contaminated wastes

                                                                    often > 2,000 mg/kg contaminant of
                                                                    concern
                                         Estimating RBA of Pb in Soil and Soil-like materials
                                         (OSWER 9285.7-77, May 2007)
                                         Most of 19 solid waste materials from smelter origin
                                         Pb content: 1,590 to 14,200 mg/kg, median 7,225 mg/kg

                                         Estimating RBA of Arsenic in Contaminated Soils and Solid Media
                                         (Rodriguez et al., 1999)
                                         As content: 233 to 17,500 mg/kg, median 1,460 mg/kg
   Contaminant Concentration in Soil / Solid Waste
     when will bioavailability adjustments be made?
                                                        Assessing Bioavailability of
                                                       Moderately Contaminated Soil
 Highly Contaminated
                       Risk= [Soil]
                                     (EF)(ED)(IR)(BIO)
                 (BW) (AT)

example: 100 mg/kg As target
The greatest utility of IVG or in vivo methods may be
to assess risk for soils with mod. level contamination
 Pb paint, pesticides, coal ash, CCA, cattle dips, etc.
    Moderately
   Contaminated
reasonable adjustment

    Background
High level: 7,000 mg/kg total As
bioavailability < 2% to be < target
unreasonable adjustment

Moderate level: 300 mg/kg As
bioavailability < 33% to be < target
reasonable adjustment
            Moderately contaminated
         urban and/or old industrial sites
                                                      Page  1

-------
      Bioavailable (in vivo) vs. Bioaccessible (in vitro)
     Method Detection Limits and Contaminant Levels
most in vivo dosing studies require highly contaminated soil
> 500 to 5,000 mg/kg contaminant

Moderately contaminated soil levels
could be < 1000 mg/kg Pb; < 100 mg/kg As
Below in vivo detection limits

Below in vivo working range but
easily measured by IVG methods

A Strong Advantage of IVG methods
the ability to estimate (bio)availability
at moderate contaminant levels
                                       Highly Contaminated


                                           Moderately
                                          Contaminated
                                           only in vitro


                                           Background
    Are we confident to use IVG methods to Estimate
           Contaminant (Bio)availability in Soil
           for Moderately Contaminated Soils?

More confident to use IVG methods validated for highly
contaminated soil if the contaminant source is the same

IVG method validated for scorodite (As contaminant)
in highly contaminated smelter soil;
OK to use IVG method on soil with moderate levels of scorodite

Confident to use IVG method on soil with moderate levels of CCA
contamination? other sources of As contamination (pesticides?)
     What do we need to establish confidence in using
the IVG method for CCA / other arsenic contaminated soils?
1. Evidence the IVG method is correlated with in vivo
   for the CCA (other) arsenic sources
   study conducted with highly contaminated CCA soil(s)

2. Contaminant (i.e., arsenic) speciation in soil
   identify the source term for arsenic (CCA or other?)
                     How do we do it?
            Contaminant Source and Speciation
 contaminant source
 mineral source: mineral speciation method
 i.e., galena in smelter-contaminated soil
 Soil-sequestered contaminant                 Fe  OH
 i.e., As sequestered by reactive soil Fe oxides  o
                                                                  Soil chemical speciation
                                                                  methods should be used
                                                                  when soil is the sink
                             Sir
                                                                                                                o

                                                                                                              O As OH

                                                                                                                OH

                                                                                                             soil sequestered
                                                                                                                arsenic
         Contaminant Speciation in Soil / Solid Waste
         using Advanced Spectroscopic Investigation
         more from Kirk Scheckel and H. Jamieson at 4 p.m.
       Contaminant Speciation in Soil using Extraction
              Extract Different Contaminant Pools
       70
               EXAFS for As
                   in soil
               eV
                                   Advanced Photon Source
                                  Argonne National Laboratory
« 50

- 40

| 30

•520

5? 10

   0
                                                                             Soil A
                                                                                       Soil B
                                 water
                                 NaOAc
                                 0.1 M PCM
                                 NH2OH(HCI|
                                 NH40X
                                                                                                          easily dissolved
                                                                                                          weakly adsorbed
                                                                                                          strongly adsorbed

                                                                                                          part of Fe oxide fraction
 Based on "selective" extraction of soil contaminants
 must be very carefully applied and interpreted
 must have knowledge of extractant and soil / solid waste system
                                                         Page 2

-------
             Contaminant Speciation is Essential

     Knowledge of chemical form of the soil / solid waste
           contaminant controlling bioaccessibility

contaminant species in old orchard soil same as
contaminant species in smelter soil (in vivo correlation study)?

Yes: then we are more confident to use the IVG (in vitro) method for
the orchard soil
               ^^^LL^/          Pesticides
                                  in old orchards     % "
                                                                                       A Case Study
                                                                            OSU In Vitro Gastrointestinal Method
                                                                                            Simulated Gl extraction at 37°C
                                                                                             Gastric bioaccessibility and
                                                                                             Intestinal bioaccessibility
                                                                                 Development of Chemical Methods to Assess
                                                                                 the Availability of Arsenic in Contaminated
                                                                                 Media, R825410

                                                                                 U.S. EPA, Office of Research and Development
                                                                                 National Center for Environmental Research

                                                                                 N.T. Basta, R.R. Rodriguez, and S.W. Casteel
                                                                                 $431,677; Nov 1996 to October 2000.
        Development and Evaluation of IVG Methods
          Correlation Studies with Immature Swine
          In vitro bioaccessible vs. in vivo bioavailable
                                                                       Approach to Development of an IVG Method
                                                                                      Occam's Razor
                       15 As-Contaminated Soils
      Swine Dosing Trial
     Measure Urinary As
      Bioavailable
         Arsenic
       Stan Casteel
     Univ. of Missouri
                                      In Vitro
                               Gastrointestinal Methods
                  Compare       Bioaccessible
                                      Arsenic
                            Nick Basta, Robin Rodriguez
                                Oklahoma State Univ.
              single       batch    dynamic
            extraction    in vitro    in vitro
        Simple                               Complex

                    RBA Method complexity

"Pluralitas non est ponenda sine neccesitate" - William of Occam (1280-1349)

Plurality should not be assumed without necessity

If you have two equally likely solutions to a problem - pick the simplest

"Everything should be made as simple as possible,  but not simpler"
 -A. Einstein
        Bioaccessible vs. Bioavailable
    OSU IVG correlated with immature swine

                           OSU IVG correlation with in vivo
  eo                        As with dosing vehicle
SS   RBA AS = o 942 IVG + 7 11 r=09r*    Rodriguez et al. 1999.
«f so                         ES&T 33:642-649
.£ 40                        As without dosing vehicle
2                          Basta et al., 2007. J. Environ.
5 so                        Health Sci. Part A 42:1275-1181.
g                          Special Publication (BARGE):
io 20                        Bioaccessibility of Soil Contaminants
»                          C. Gr0n and J. Wragg (eds.)

|                          Pb with/out dosing vehicle
K  "„    .„    ,„    ,„    „    Schroder etal., 2004
                            J. Environ, dual., 33:513-521.

                           Cd with/out dosing vehicle
                           Schroder et al., 2003.
                            ES&T37:1365-1370.
                                                                                   Research on OSU IVG USEPA Project
                                                                                         still continuing after 10 yr
                % Bioaccessible As
     Basta et al. 2003.
 Grant R825410 Final Report.
 submitted to U.S. EPA ORD
                                                                                                            the soil isn't contaminated
                                                                     1997
                                                                                        2007
                                                              PageS

-------
           OSU IVG USEPA Project Productivity

   Publications (Basta): 39
     10 refereed publications
     27 proceedings /abstracts (11 at international conferences)
     2 (book chapter, technical report)
   5 Conference Symposia (3 international)
   4 Graduate Student Ph.D. dissertation and M.S. Theses

   Collaborative research
   Soil samples, reports, data (including bioavailability) sent to
    13 research groups
   many joint publications / proceedings / symposium
   U.S. EPA ORD (NERL, NRMRL)

   Round robin validation studies
   Bioavailability Research Group of Europe (BARGE)
     Arsenic Solid Phase Speciation of OSU IVG Soils


            Four solid phase species identified by EXAFS

   Scorodite (FeAsO4- H2O)   43 to 76% As;  mean 61 % As

               SorbedAs  7.3 to 28% As; mean 17% As

           "Elemental" As  6.3 to 43%;  mean 16%

                 Lollingite  Oto8%; mean 4.6%
Bioavailable Arsenic and Solid Phase Speciation (EXAFS)
          Arsenic identified as Scorodite inversely related
                 to Relative Bioavailable Arsenic

   Koch et al. (2005): Soils containing scorodite had lowest bioaccessible As
             of 6 soils from a military base in gastric solution
        Bioavailable As and Soil Reactive Fe Oxide
                Soil Chemical Extraction Methods
                                                                                                     Intercept= 46.1
                                                                                                       Slope= -4.91
                                                                                                        /• = -0.91**
                                                                                            % As in Feox
                                                                                                                 Bioavailable As
                                                                                                                 inversely related
                                                                                                              to As in Soil Feox pool
                  Arsenic Fractionation
                Phosphate extractable arsenic
                                     Phosphate will increase
                                     bioavailable arsenic
                                     and bioaccessible arsenic

                                     Basta et al. 2007. JEHS Part A,
                                     42:1275-1281
           % Arsenic Extracted by Phosphate
      Assessing Contaminant Bioaccessibility in Soil
          when In vitro Gastrointestinal Methods
                    are the Only Option
We could extrapolate the OSU IVG methods for highly contaminated
smelter waste soils to soils/solid waste where scorodite or As
sequestered by Fe oxide was the source term to moderately
contaminated soils

As-sorbed to Fe oxide: A likely source term for many As-
contaminated soils and solid wastes

Arsenic Speciation by chemical extraction / EXAFS could be
performed to verify that the form of As is sorbed to Fe-oxides
                                                           Page 4

-------
                         U.S. EPA
     Guidance for Evaluating the Oral Bioavailability of
 Metals in Soils for Use in Human Health Risk Assessment
                 OSWER 9285.7-80, May 2007

"A detailed protocol for the test method	and a description of
the known limitations of the test including a description of the
classes of materials that the test can and cannot accurately assess."

   > Specify the contaminant chemical speciation and

   > whether the IVG method has been correlated with in vivo for the
     contaminant species in the test material
                                                                                       may be down the track?
                                                                                      Future Research Needs
             Collaborative Efforts are Essential
            Maximize Use of Resources ($$), Time
                 Research is expensive
                 Especially Bioavailability Data

                 >  Share soils, bioavailability data
                 >  Cooperative round robin studies
                 >  etc
Thank you for your attention
More information? Please contact:
          Nick Basta
     School of Environment
     and Natural Resources
       basta.4@osu.edu
       www.snr.osu.edu
                                                         Page 5

-------

-------
 The bioaccessibility of soil Nickel,  can  it
       be explained using solid phase
                 distribution data?
                       Joanna Wragg
Kingsley Dunham Centre
Keyworth
Nottingham NG12 5GG
Tel 0115 936 3100
                                                                                     Health  Effects
                                             ACUTE
                                       Death
                                        • 90 min exposure to c.400
                                          mg rrr3 metallic Ni
                                        • Ingestion (by a child) of
                                          570 mg kg-1 bwNiSO4
                                       Headache, nausea, vomiting
                                       etc
                                        • Inhalation of Ni carbonyl
                                       Gl distress, muscular pain,
                                       exacerbation of dermatitis
                                        • After ingestion of Ni
                                          compounds
                                       CHRONIC
                                 Respiratory effects
                                 (bronchitis, emphysema,
                                 asthma etc)
                                  • Repeated inhalation
                                   toxicity
                                 Effects on organ weight and
                                 nervous system,
                                 reproductive toxicity etc
                                 (animal studies)
                                  • Repeated oral toxicity
                                 Skin sensitivity from Ni and
                                 its compounds
                                  • Dermal studies
                                   /hypersensitivity tests

                   Nickel in Soil
    Geogenic and Anthropogenic
    Sources           "
     •  Mined predominantly as Ni-Fe
       sulphide

     •  Dominates in ultramafic and
       igneous rocks
       Many industrial uses
        • Ni-Steel alloys,
         electroplating, jewellery,
         catalyst etc...
Soil chemistry is based on the most
stable, divalent ion
Ni (II) is stable over a range of Eh and
PH conditions
Solubility increases with decreasing
l'n
Associations with numerous soil
components/minerals

   Mn °xides
   Sulphides
   Carbonates
   Organic matter
   Etc	
                                             Soils  Under Investigation
                   Organic
                     c
  Anthropogenically Influenced

s   |e   Nj %    H
   •            •
 Soi| 1    c1   c 6 25  c 28%

 Soil 2    c. 1.4   c. 6.4   c. 6.5%

        SPeciati°"
Soil 1

Soil 2
        c. 90%    c. 10% Fe
       Oxidic Ni*  oxide & trace Ni
       c. 99.6%   c. 0.4% Fe
       Oxidic Ni*  oxide & trace Ni
                                                                  *Oxidic Ni includes all fornis of Ni oxide/hydroxide,
                                                                   Ni-Fe oxide/hydroxide and complex multi-metal

                                                                           oxide/hydroxides

                                                                   ©NERC All rights rescued
                                                                                                      Geogenically Influenced
Sample

  1
  2

  3

  4

  5

  6
  Ni
mg kg-1


 19.4


 65.9

 45.9

 78.9
PH

7.2

6.5

7.2

6.0

7.2

5.8
Organic
  C

 3.52

 1.69

 4.70

 3.02

 2.15

 2.62
          Bioaccessibility Methods

  In vitro gastric and small intestinal simulations
   • <250um size fraction
   • 37°C
   • PBET modified from Ruby et al.,
        intestinal pH
      • Agitation via shaking, not Ar bubbling
     IVG
      • Fluids and residence times as per
      Basta et al., 2007
                                     Differences in  Bioaccessibility Method
                                         Parameters for anthropogenic Ni
Parameter
Soil:Solution Ratio
Stomach Fluid
Composition
Intestine pH
Intestine Fluid
Composition
PBET
1:100
1.25 g M pepsin
7.0
pH adjusted with
Na2C03
IVG
1:150
10 g M pepsin, 0.15M
NaCI, no citrate, malate
or acetic acid
5.5
pH adjusted with
NaHCO3
                                                            4 hours
                                                                                 1 hour
 Intestine Residence
      Time
       Are the results likely to differ significantly?

-------
                                    .ML.
       Solid  Phase Distribution of Nickel

                    CISED Test
hemometric dentification of  ubstratesand  lement  istributions

•Separate aliquots of aqua regia of increasing
concentration            "          "    Centrifugation

•Passed through the sample under centrifugal
force

•Determination by ICP-AES

•Chemometric data processing                 045u

•Identification of physico-chemical hosts and     membrane
the metal distributions within the sample        L'"""" "* ' leachate
under test
                                                                  Anthropogenic Contamination
  Differences

 Between Method
  Not significant

   (Stomach or
 Intestine, for both   •*
     soils)       E :

  Within method    |
   Significant      g
differences between   |
   stomach and     s
   Intestine data

   Between Soils
    Significant
 difference between
stomach or intestine    <
      data        «
                                                              Although there are methodological differences
                                                                                SoihSolution ratio
                                                                              Fluid Chemistry etc...
                                                                 There is no appreciable difference in the
                                                                       measured Ni bioaccessibility
£JL*. Britu. . 4 •&& BHii*h
<£f> GeotajfcJi Sumy ^PVIkT 'VI^T ^tfl 1 V*7 Ge«lo(fc»l Sum*
CISED Components - Soil 2
™ 4™ ™° Component
am lm 133°
JOCD icoo Organic
Carbonate (1)
Ca/AI mixed
^ !™ ™ ™ " assemblage
| im ™ ™ AI/Fe oxyhydroxide
1 ™ ™ o Fe oxide (1)
™o 5 10 15 ™o 5 10 15 ™o 5 10 15 Ex changeable
™ ™ Fe oxide (2)
Z ™ Fe oxide (3)
» o " Carbonate (2)
i'" »•
Soil 2 - Ni distribution
Component Composition
S
Ca
Ca-AI-Ni
AI-Fe-Ni
Fe-Ca
Ca-Mg-Fe
Fe-P
Fe-AI
Ca-Mg
fek. 4p~

Ni mg kg'1
10.7
208
1070
865
0
45.6
220
553
0

-------
 3000

 2500

 2000
Soil 2
                             .ML.
' 1000

  500
                                             Natural Contamination
                                   Not
                                   significantly
                                   different
                                   P>0.05
Cluster
1
2
3
4
5
5
7
8
9
10
11
12
Name
High Carbonate
Carbonate
High Carton ate
(2)
Carbonate (2)
Organic
Fertilizer
Exchangeable
Fe/AI
Phosphate
Fe/AI
oxyhydroxide
Fe/AI silicate
Mn oxide
Fe oxide
Ni
1.67
0-2.8
0.6-0.9
0
0
0.03-0.1
0-0.1
3.72
0.5-1.3
0.04-1 .4
0.06-2.6
0.01-1.9
                                                         Briifeh
                                                         Gesbfka! Sum,
                                   Bioaccessible
                                   Ni associated
                                      with
                                   components
                                   extracted over
                                    the 1st 7-8
                                    extraction
                                     phases

                                    RELATED
                                      TO
                                     Clusters

                                    1-8 and 11
2  3  4 5  6  7 8  9 10 11  12 13 14
        Extract Number
                                                                        Summary
                                       Higher total and bioaccessibility data observed for the
                                       anthropogenic soils
                                       No appreciable differences between the PBET and the IVG
                                       data
                                       Soil components acting as sources of Ni bioaccessibility were
                                       similar for both contaminant types
                                       • Non Fe dominated components
                                          • Organic, Fertilizer/Exchangeable/Carbonates/Mn oxides
                                       • Naturally contaminated soils
                                          • Additional contribution from  Fe oxyhydroxide
                                            components
                                          • Thought to be an effect of the different ageing of the Ni
                                            within the soil matrix

-------
            Summary (2)
Extraction Efficiencies of the CISED
 • Similar for the anthropogenically contaminated
  soils, c. 32%
 • > extraction efficiency for the soil with a higher
  total Ni content
 • 5-15% for the geogenically influenced soils
   • Higher extraction rates are not observed in
     soils that have the highest total Ni
     concentrations or absolute Ni bioaccessibilities
        Final Thoughts .....
Can solid phase distribution help explain
measured bioaccessibility data?
 • YES, Inclusion of geochemical information and
  testing provides a wealth of additional data
   • Help explain the bioaccessibility data
   • Support the Risk Assessment process
 • Previously adopted method for As and Pb studies
  in the UK
 • Being applied in to Cr issues in the UK and
  Canada
                                                                         Summary (3)
Anthropogenic soil
bioaccessibility data
 • Significant decrease in
  Ni concentration in the
  intestine phase
 • Not observed for
  geogenic Ni soils
 • Function of Ni solubility
  with pH???
                                         t  «  Id I) 12
                                                                    Acknowledgements
    Mark Cave, Chemical & Biological Hazards
    Programme, British Geological Survey, UK
    Ken Reimer, Kim House, Ashley Campbell, Megan
    Lord-Hoyle, Environmental Sciences Group, Royal
    Military College, Canada
    Bruce Conard, BRConard Consulting, Canada
    Chris Ollson, Jacques Whitford Ltd, Canada

-------
 Importance of Metal  Speciation
in Understanding Bioavailability
                                                                         ..to protect human health & the natural environment.
                                                                                        SPECIATION
                  Kirk G. Scheckel
              US EPA, Cincinnati, OH
                                        Remediation
                                   Indicative of metal fate and transport
                                   Helps decide remediation strategy, i.e.
                                   dig-and-haul, in-situ amendment,
                                   monitored natural attenuation
                                   Evaluate effectiveness of remediation
                                                                                                  Understand the variability of biologically
                                                                                                  available metal uptake
                                                                                                  Ability to manipulate system to reduce
                                                                                                  bioavailability
                                                                                                  Develop comprehensive predictive
                                                                                                  models based on speciation
    Systematic Characterization of Exposure-Dose-
    Response Continuum and the Evolution of
    Protective to Predictive Dose-Response Estimates
   Chemical Exposure

      Protective

          Exposure
 Toxicological Response

          Qualitative
           Solution Specie


      Predictive
Site of
biological
response
What
do we
 need
 and
 what
do we
 want
  to
know?
                                                                                 EFFECT OF SOURCE OF Pb ON BIOAVAILABILITV TO RATS
Q- 300

§200
m      SOIL: 3,900 ppm Pb

      SOIL: 810 ppm Pb
  °0     10   20     30
        DOSE (mg Pb/Kg BW)

 EFFECTOF SOURCEOFPh ON BIOAVAILABILITY TO HUMAN
                                                                                    Stsse

                                                                                     SOURCE
                                                                                                      BLOOD Pb (Jlg/dL)
                                                                                                     SOIL Pb (lOOOmg/kg)
                                                                                     SMELTER/URBAN      1.1 - 7.6
                                                                                     MINING             0 - 4.8
         Components of Bioavailability
                                                                               Levels of Detail in Research
  PHYSIOLOGICAL
     •  Pharmacokinetics - Nutritional needs (Ca, P,
       Fe, etc)


  CHEMICAL
     •  Form of metal
     •  Geochemical matrix
     •  Particle size
                                  Atomic     Molecular  Microscopic Macroscopic    Field
XRF
XPS
XAS
Requires
synchrotron
radiation.
•XRD
•TGA
• FTIR
• DRS


• Enhanced
Visual
Analysis:
1. SEM
2. TEM
3. AFM
• Field Plots
• Equilibrium
Studies
• Kinetic
Studies
• Extractions
• Visual/
Intuitive
Insight
• Field Plots



-------
               Advanced Photon Source
             (Argonne National Laboratory, Argonne, IL)
                                                                        Principal Synchrotron Techniques
                                                                          Used in Environmental Science
                                                                                  • X-ray Fluorescence (XRF): chemical
                                                                                  composition (quantification, mapping)

                                                                                  • X-ray Absorption Fine Structure
                                                                                  (XAFS) Spectroscopy: chemical
                                                                                  :>tx-cii!."io;i * ex; dun - vi state, coordination,
                                                                                  nearest neighbors)

                                                                                  • Surface Scattering and Diffraction:
                                                                                  surface structure, sorption processes

                                                                                  • Microtomography: 3D imaging of
                                                                                  internal microstructure (porosity, fluid
                                                                                  flow, composition)
       X-ray Absorption Spectroscopy
X-ray Absorption Spectroscopy: Measure
energy-dependence of the x-ray absorption
coefficient fi(E) [either Iog(l0/I) or (If/lo)]
of a core-level of a selected element
Element Specific: Ele
    XANES = X-ray Absorption Near-Edge Sp

    EXAFS = Extended X-ray Absorption Fine-Stnlc
                                                                             Pb Immobilization & Bioavailability

                                                                           • Smelter contaminated site in Joplin, MO
                                                                           • Various P amendments
                                                                           • In-vitro and in-vivo bioavailability
                                                                           • Synchrotron speciation
 X-ray Absorption Spectroscopy &

                  PCA/LCF

ion in a Smelter Contaminated Soil: Results of in-situ Remediation with Phospho
                                   Lead (Pb) Speciation (%)
                                                                                   Effect of Time and 1% P Treatment on  Soil

                                                                                   Lead Bioavailability  Joplin Swine Ryan et ai  2004


                                                                                             Control Soil (all),y = 6.44 + 198(1 -e"'0021 x), R2 = .95

                                                                                      250     1 % P Treated Soil (3mo), y = 3.42 + 180(1 - a'-0022*), R2 = .95
                                                                                             1% P Treated Soil (18 mo), y=10.02 + 150(1 -e"°°19*,, R2 = .92
                                                                                   \        1 % P Treated Soil (32 mo), y = 6.09 + S2(1 - e"0""1*,, R2 = .94
                                                                                   5J1 200     PbOAc(all),y=6.49 + 200(1 - e''0023x), R2 = .S4
                            10% Biosolids   63
                                                                                   <  100
                                                                                                200     400     600     800

                                                                                                 Lead Dose (jig Pb (pg BW day)'1)

-------
           Soil Lead Bioavailability
             Joplin 18 mo Sample
    Conclusion: Joplin  Field  Experiment
 1 Bioavailability of soil lead is not a simple function of
                                                                    total soil lead.
                 Rat   Swine    In vitro  Human
   Control
   Treated
                  7.2    21.6    21pH2.5   13.1
                               39pH2.0
                               51 pH1.5
• Soil lead bioavailability can be measured by
    • Swine
    • Rat
    • Human
    • I n vitro

• Soil lead bioavailability can be changed by addition of
  materials to soil.

• The addition of materials to the soil altered the
 geochemistry of soil lead.
               What's next?
 Do we have enough information from expensive in-
 vivo animal bioavailability studies tojustify
 acceptance of affordable in-vitro extraction
 bioaccessibility?

 What role does spectroscopic speciation play in
 support of in-vivo and in-vitro research?
 Requirements for  using  bioavailability in
         risk management decisions

 1) An appropriate measure (methodology and
    samples)
 2) Knowledge of the reason for the observed
    measurement
 3) Knowledge of the long-term stability of the
    measurement
             Lessons  learned

Total metal content is not a good indicator of exposure or risk

Soil chemistry important in determination of
bioavailability/phytoavai lability
   Form is important
 •  Particle size is important
   Adsorption is important
    • Fe/Mn are important adsorptive surfaces
     Organic matter is important adsorptive surface

Cannot always assume an increase concentration in the foodchain
equates to increase transfer through the foodchain (plant uptake)

Predicting the potential transfer of soil metals requires a holistic
evaluation of soil, plant, animal, and human processes which may
increase or reduce the transfer (bioavailability)

-------
      Direct Identification of Metal Compounds
                             in
        Contaminated Soil, Mine Tailings, and
                       House Dust
         Using Synchrotron-based Methods

    H. E. Jamieson, S.R. Walker, S.E. Fawcett, Queen's University
                A. Lanzirotti, University of Chicago
                 P. Rasmussen, Health Canada
            S. Beauchemin, Natural Resources Canada
            M.B. Parsons, Geological Survey of Canada
                                                                             Mineral Form
        FeAsS
  Grain Size
                                    Modified from Ruby et al, 1999


                                            AS20,
                        Fe-As oxides
  Encapsulation or Rimming of Grains


     qtz   As


   decreasing
                                                                                                BIOAVAI LABILITY
                Analytical Techniques
              Analytical Techniques
Synchrotron-based
•  Micro X-ray Fluorescence
  (jjXRF): Element mapping
•  Micro X-ray Diffraction (jiXRD):
  Identify microcrystalline
  phases
•  Micro X-ray Absorption Near
  Edge Structure (jiXANES):
  oxidation state of As & other
  elements
•  Macro X-ray Absorption
  (XANES): local molecular
  environment
                               Classical
                                 Electron Microprobe (EMPA):
                                 quantitative chemical analysis

                                 Petrography: visual
                                 characterization
Synchrotron-based
•  Micro X-ray Fluorescence
  (jjXRF): Element mapping
•  Micro X-ray Diffraction (jiXRD):
  Identify microcrystalline
  phases  lOmfofOfl
•  Micro X-ray Absorption Near
  Edge Structure (jiXANES):
  oxidation state of As & other
  elements   10 micron
•  Macro X-ray Absorption
  (XANES): local  molecular
  environment  10mm
Classical

•  Electron Microprobe (EMPA):
  quantitative chemical analysis
          10 mteron
•  Petrography: visual
  characterization
          1-2 micron
  Results in
  grain-scale
  characterization
Giant Mine: Large abandoned site near Yellowknife, NWT
 Arsenopyrite-bearing
 ore was roasted to
 liberate Au
 Total arsenic in tailings, soils, se'jy nts 10>
                          ,
                                                                           Roaster-generated Fe oxides:
                                                                           nanocrystalline maghemite yFe2O3and hematite aFe2O3

-------
                                           Nanocrystalline
                                           Maghemite or
                                           Maghemite
                                           & Hematite
                                            Mixture of As+3
                                            and As+6
                                                                      Walker etal (2005) Canadian Mineralogist 43, 1205-1224
Abandoned gold mines (more than 65
sites) in Nova Scotia


Montague
            Arsenic concentrations
Total As content in <150 um fraction of tailings, soil, and mill residue samples (ppm)
Site
Caribou
Goldenville
Montague
N Brookfield
Max
313,000
210,000
62,100
9,170
Min
15,200
7,200
318 (soil)
195 (soil)
Median
72,600
38,900
10,600
1,590
                                                                      At North Brookfield
                                                                      sulfide ore was roasted
                                                                      to liberate gold
                                                                                                         Roaster   Chlorination
                                                                                                    Ml"   Stacks     Plant
   Fe-arsenate rimmed Fe-oxyhydroxide
      Three different Fe-arsenates
                                                                     Almost amorphous
                                                                     - two broad weak rings
                                  Mixture of lepidocrocite and goethite
                                  (polymorphs of FeOOH)
                                                                    As2O5 = 27.9%
                                                                    SO4=2.4%
 -All concentrations by EPMA

-------
      Rimmed Grain -  Scorodite
                                                                                 Yellowish  grain - Kankite

                                                                                              £_
                                                                                                                 Kankite
                                                                                                                 FeAsO4-3.5H2O
                      Yukonite
                           Yukonite
                           -Ca3Fe7(As04)e(OH)9.18H20
         M^^^E'r^^MHM
  Transmitted and reflected light   Calks (molar) = 0.3
                                                                              Tailings from gold-ore roasting
                                                                                          North Brookfield
                                                                       Abundant roaster derived As-
                                                                       bearing Fe-oxide (hematite
                                                                       aFe2O3 or maghemite yFe2O3)
                                                                       in some samples
                                                                                                   68%  Asi+
                                                                                                   32%  As!+
  Arsenic-bearing minerals identified in IMS gold mine tailings (to date)
                                                                                Information on Grain Size from jaXRD
   Kankite
FeAs04-3.5H20
                            As-Ca bearing Fe
                            oxyhydroxide = Yukonite
                                              As-bearing Fe
                                              oxyhydroxide
Amorphous Fe arsenate
                                                                                                           As-bearing Hematite
Many As-bearing secondary minerals are nanocrystalline, but
porosity, cementation and disaggregation affect grain size

-------
     Metal Speciation of House Dust: |aXRF mapping
                                                                    Zn macro XANES spectra of known species
  Example of fitting results forZn
                              In contrast, Cu in
                              house dust is
                              largely bound
                              to organic species
 General Conclusions

A combination of synchrotron and classical
mineralogical analysis can provide direct
identification of metal-bearing grains
Information on grain size and intergrowth
textures may also be provided

Most samples contain at least three mineral
forms of a particular metal
Many particles are aggregates or mixtures of
more than one mineral
   Acknowledgements
Funding for this project provided by
  MITHE (Metals in the Human Environment) Research
     Network
  NSERC
  Indian and Northern Affairs Canada
  Geological Survey of Canada
  Health Canada
Correspondence: jamieson@geol.queensu.ca

-------
        Bioaccessibility of arsenic
     adsorbed onto or incorporated
     within freshly synthesized iron
         oxide minerals using  the
                     SHIME
         B.D. Laird1-2, K. Dekker1-2, T.R. Van De Wiele3, D.
              Peak2, W. Verstraete3, S.D. Siciliano2

            ^interdisciplinary Graduate Program of Toxicology
            2University of Saskatchewan, Saskatoon, Canada
                  3University of Ghent, Belgium
Bioaccessibility and Risk Assesment
 D Total contaminant
   concentration in a mineral
   may overestimate exposure

 D Bioaccessibility measured
   using In vitro GI models as a
   surrogate for bioava liability

 D Not yet widely incorporated
   into risk assessment
   •  How do we go about
      validating in vitro models?
    Bioaccessibility vs Concentration:
    Inverse Relationships

Two Possible Mechanisms:    A Cause for concern?
     Thermodynamic
     limitation of As
     dissolution
     D  Bioaccessibility a
        function of LS Ratio

   2. Kinetic limitation of
     arsenic dissolution
     D  Bioaccessibility a
        function of residence
        time
 Bioaccessibility vs Concentration

 D How does mineralogy impact the
   relationship between bioaccessibility and
   concentration?
   • Amorphous Scorodite
     D Arsenic incorporated into iron oxide mineral
     D Readily soluble at low pH

   • As(V) - Ferrihydrite
     D Arsenic adsorbed onto surface
     D Ferrihydrite sparingly soluble at low pH
    Isotherms in  Soil  Chemistry
 GI Microbes and As Toxicokinetics
     Shape of isotherm is a function of the chemical
        processes controlling dissolution
 D Gastrointestinal microbes increase arsenic
   bioaccessibility of mine tailings in colon stage of
   SHIME

 D Gastrointestinal microbial community may contain
   both arsenate reducing bacteria and iron reducing
   bacteria

 D Small intestine is the primary site of absorption
   •  Duodenum & Jejunum: Lower microbial densities
   •  Ileum: Higher microbial densities
                                                              D  Arsenic is potentially absorbed in the proximal
                                                                 colon

-------
Hypotheses
1.  Arsenic dissolution in the SHIME is at
   equilibrium
   D HA: Kinetics limit dissolution of arsenic in the
     SHIME


2.  Gastrointestinal microbes impact the
   bioaccessibility of arsenic absorbed to
   ferrihydrite and incorporated within scorodite
                                                      Scorodite Bioaccessibility in the
                                                      Stomach
                                                             Scorodite - Stomach
                                                                                          000  §
                                                                  Arsenic Concentrationmjneral ™g kg"1
Scorodite Bioaccessibility in the
Small Intestine
                                                      Scorodite Isotherm:  Stomach
      Scorodite Small Intestine
                                   2 Oe+5  .O
                                                            E
                                                            §
                                                                 Scorodite Stomach
         Arsenic Concentrationmineral mg
                                                                       Concentrationso|utjon ng L"1
Scorodite Isotherm: Small Intestine
                                                      Conclusions: Arsenic Bioaccessibility
                                                      versus Concentration
         Scorodite Small Intestine
I  26«
|

§  1645

O
            0.0  5.06+4   1.0645   1.5645  2.0645  2.5645

            Arsenic Concentrationso|utjon fj.g L"1
                                                         When an  inverse relationship between
                                                           arsenic concentration and
                                                           bioaccessibility is observed:
                                                           • As bioaccessibility from scorodite in the
                                                             stomach likely determined by LS Ratio

                                                           • As bioaccessibility from scorodite in the
                                                             small intestine may be determined by
                                                             residence time

-------
Conclusions: Use of Isotherms in
Bioaccessibility Research
GI Microbe Impact on Arsenic
Bioaccessibility
   Use of isotherms may not be applicable for use
   with real-world samples
   • Mineralogy MUST be constant between samples
     for interpretation of isotherm results
D  Should test impact of LS ratio and residence
   time when inverse relationship observed
   • Current recommendations focus on LS ratio
   • May result in underestimation of bioaccessibility
     and exposure in risk assessments
 § 100
 o
GI Microbe Impact on Arsenic
Bioaccessibility
Conclusions: Impact of GI Microbes
on arsenic  bioaccessibility
      3.50
      3.00
          Ferrihydrite-As(V)
                                  Active Colon
                                  Sterile Colon
                                                              GI microbes capable of increasing or decreasing
                                                              arsenic bioaccessibility from arsenic-bearing
                                                              minerals
                                                              • GI microbial activity associated with iron reduction
                                                              • Effect of GI microbes likely mineralogy-dependent

                                                              Unknown toxicological implications of this
                                                              microbial impact on bioaccessibility
                                                              • Mechanisms potentially also active in ileum
                                                              • May pose challenging for the validation of in vitro
                                                                models using in vivo animal dosing experiments
Acknowledgements

D  International Society of Exposure Analysis
D  NSERC MITHE-RN network
D  NSERC
D  University of Saskatchewan Department of Soil
   Science
D  University of Saskatchewan Department of
   Toxicology

-------