Soil Dioxin Relative Unavailability Assay Evaluation Framework
United States OSWER 9200.2-136
Environmental
Protection Agency
Soil Dioxin Relative Bioavailability Assay
Evaluation Framework
February 2014
-------
Soil Dioxin Relative Unavailability Assay Evaluation Framework
Table of Contents
1. Introduction 1
2. KB A Assay Requirements 2
2.1. Application of KB A to Risk Assessment 2
2.2. Calculating RBATEQ of PCDD/F in Soil 3
2.3. RBAxEQ in Soil for Noncancer Risk Assessment 4
2.4. RBAxEQ in Soil for Cancer Risk Assessment 6
2.5. Selection of Animal Model for Predicting RB A in Humans 6
2.6. Dosing Regimen 7
2.7. Measurement of Internal TEQ Dose 8
2.8. Confidence in RB A Estimates 9
2.9. Soil Characterization 10
3. Summary and Conclusions 10
4. References 10
-------
Soil Dioxin Relative Unavailability Assay Evaluation Framework
Soil Dioxin RBA Assay Evaluation Framework
1. Introduction
The Risk Assessment Guidance for Superfund (RAGS) Part A (USEPA, 1989) discusses making
adjustments to Superfund site-specific risk assessments when the medium of exposure in an
exposure assessment differs from the medium of exposure assumed by the toxicity value (cancer
slope factor, reference dose value, etc.) based upon site-specific bioavailability data. An
important consideration in assessing risks from exposures to dioxin in soil is whether an
adjustment is needed in the application of the oral cancer slope factor (CSF) and/or oral chronic
reference dose (RfD) for 2,3,7,8-tetrachloro-dibenzo-p-dioxin (TCDD). This adjustment would
account for differences in the bioavailability of TCDD (and lexicologically related
polychlorinated dibenzo-p-dioxins [PCDD] and polychlorinated dibenzofuran congeners
[PCDF]) in soil and in the test medium used in the critical study(s) on which the CSF and/or RfD
were based (e.g., dietary exposure vs. exposure to soil). An adjustment would be considered
appropriate if evidence were sufficient to indicate that the relative bioavailability (RBA) of the
PCDD/F mixture in soil was less than 100%.
EPA recently compiled and summarized studies conducted to estimate relative bioavailability
(RBA) of TCDD and PCDD/F in soils (USEPA, 2010). Nine studies were identified that
collected data on soil RBA based on bioassays conducted in guinea pigs (McConnell et al., 1984;
Umbreit et al., 1986; Wendling et al., 1989), rabbits (Bonaccorsi et al., 1984); rats (Budinsky et
al., 2008; Finley et al., 2009; Lucier et al., 1986; Shu et al., 1988) or swine (Budinsky et al.,
2008; Wittsiepe et al., 2007). These studies used various experimental designs for dosing
animals, metrics for estimating bioavailability, and data reduction methods for calculating soil
absolute bioavailability (ABA) or RBA (Table 1). The extent to which variations in
experimental design affects RBA estimates has not been rigorously evaluated. Only one study
has compared RBA estimates for the same test materials in more than one assay; the outcome
was dissimilar estimates of RBA for 2 soils based on a single dose rat bioassay and a multiple
dose swine assay (Budinsky et al., 2008).
The current status of methods for estimating RBA of PCDD/F in soil can be considered as being
in the early development phase. Although various methods have been explored, no single
methodology has been determined to be optimal; furthermore, advancements and refinements of
methodologies is expected to continue to progress towards the establishment of standard
procedures. The evolution of varying methodologies into generally accepted and validated
methods for use in risk assessment occurred in the history of the development of the juvenile
swine assay for soil lead RBA (USEPA, 2007a,b).
-------
Soil Dioxin Relative Unavailability Assay Evaluation Framework
Until standard procedures for estimating KB A of PCDD/F in soil are established, there is a need
for a consistent approach to evaluate the strengths and weaknesses of assays designs that are
proposed or implemented to support in risk assessments. This report offers a framework for
making such evaluations. Specific design parameters that should be subject to evaluation are
identified and relevant scientific literature is cited where more in depth discussion can be found.
Whenever possible, minimal requirements for study designs are proposed. This report also
identifies issues that have yet to be resolved regarding how RBA assays should be designed and
which could be objectives of further research to develop RBA assays for soil PCDD/F and
applications to risk assessment.
2. RBA Assay Requirements
This report is organized into subsections that discuss important experimental design features that
should be considered in evaluating the potential utility of a given RBA assay design to support
risk assessment. Minimal requirements are identified at the start of each subsection and are
followed with discussions of the rationale for the requirements.
2.1. Application of RBA to Risk Assessment
Requirement 1: PCDD/F risk assessment requires estimates of the RBA for soil
Exposures from soil are almost always to mixtures of PCDD/F congeners that have varying toxic
potency and, very likely, different RBA (USEPA, 2010). Variations in toxic potency of the
congeners are accounted for in risk assessment by assigning Toxicity Equivalence Factors (TEF)
to concentrations of PCDD/F in soil, with TEF reflecting the relative toxic potency of each
congener, relative to 2,3,7,8-TCDD (TCDD, Equation 1).
CTEQ= I,Ci-TEFi Eq. (1)
where CTEQ is the concentration of 2,3,7,8-TCDD Toxic Equivalents, Ct is the concentration of
congener /', and TEFt is the TEF of congener /'. The CTEQ value is used in the appropriate
equation for average daily intake (ADITEQ), which is then used in the appropriate risk equation
(e.g., Equations 2 - 4):
ADITEQ = CTEQ • IRS Eq. (2)
-------
Soil Dioxin Relative Unavailability Assay Evaluation Framework
HQ = -- Eq. (3)
V H V '
CR = CSFTCDD • ADITEQ Eq. (4)
where IR$ is the soil ingestion rate, HQ is the hazard quotient, RfD is the reference dose, CR is
the cancer risk, and CSF is the cancer slope factor. The corresponding adjustments for RBA
would be (Equations 5 and 6):
HQ = Eq.
V RfDTCDD 4
CR = CSFTCDD • ADITEQ • RBATEQ Eq. (6)
where RBATEQ is the RBA for total TEQ in the soil.
2.2. Calculating RBATEQ of PCDD/F in Soil
Requirement 2: Calculation of RBAjEQ requires quantification of the total TEQ external dose
and total TEQ internal dose, as well as the excretion fraction for TEQ (or experimental designs
that ensure that the excretion fractions for TEQ are the same when administered in the soil or
reference material).
The general form of the calculations used to estimate RBA for PCDD/F is given in Equations 7
and 8:
RBA = Eq. (7)
ABARM M v '
ABA = AF = — - —1— Eq. (8)
ED (1-£F) M V '
where ABAxM and ABARM are absolute bioavailability for PCDD/F in the test material (e.g., soil)
and reference material (e.g., PCDD/F in a suitable vehicle), respectively; AF is the absorbed
-------
Soil Dioxin Relative Unavailability Assay Evaluation Framework
fraction of the dose; ID and ED are the internal dose (e.g., body burden) and external dose,
respectively, of the test or reference material; and EF is the fraction of the absorbed dose
eliminated by metabolism and excretion. Although the elimination fraction (EF) appears in the
expression for absolute bioavailability (ABA in Equation 8), it does not need to be considered in
the calculation of KB A (Equation76), as long as elimination kinetics are similar for the PCDD/F
absorbed from the test material and reference materials (i.e., EFTM=EFRM). However, ifEFxM
were to exceed EFjM, the ID/ED ratio will overestimate KB A. The validity of the assumption of
equal elimination kinetics of the test and reference materials is an important issue in the
estimation of KB A for PCDD/F congeners, because the metabolic elimination of PCDD/F s is
dose-dependent. Dose-dependency derives from the induction of cytochrome P450 (CYP450),
which is the primary mechanism for metabolic elimination of PCDD/F. This issue is addressed
further in the data analysis sections of this report.
The units of ID and ED in Equation 8 can be either congener mass (i.e., g or moles congener) or
TEQ (i.e., g or moles TCDD equivalents). When expressed in units of TEQ, the RBA outcome
is RBAxEQ, which is the parameter needed to estimate RBA-adjusted risk in Equations 4 and 5.
Equation 7, expressed in units of TEQ, is applicable to a single congener (e.g., TCDD) in soil or
to a mixture of congeners. However, when applied to the mixture of congeners, the parameters
ID and ED become sums of the TEQs for individual congeners that make up the ID or ED
(Equation 9):
--1— Eq.(9)
-
The data requirements for Equation 9 are quantification of the total TEQ external dose and total
TEQ internal dose, as well as the excretion fraction for TEQ (or experimental designs that ensure
2.3. RBAjEQ in Soil for Noncancer Risk Assessment
Requirement 3: For noncancer risk assessment two RBA estimates are needed: (1) RBA for
TEQ in corn oil (ABATEQ,corn oii/ABATcDD!Corn oil); and (2) RBA for TEQ in soil (ABATEQ,soii
/AB ATEQ,corn oil)-
According to USEPA 2011, EPA is considering a chronic oral RfD based on epidemiology of
Seveso, Italy cohort(s). These cohorts experienced relatively high acute multi-pathway
exposures (inhalation, dermal, soil ingestion, ingestion of contaminated produce) shortly after an
industrial accident (explosion) dispersed TCDD into the Seveso community. The dose metric in
the dose-response modeling that supports the RfD is blood TCDD. The Point of Departure
(POD) was translated into an average daily intake by use of a PBPK model which was calibrated
-------
Soil Dioxin Relative Unavailability Assay Evaluation Framework
to achieve an oral bioavailability of 87% (based on an ingestion balance study conducted in a
single individual who ingested a single dose [3H]TCDD dissolved in corn oil; Poiger and
Schlatter, 1986).
Based on the above considerations, the proposed RfD assumes 87% absolute bioavailability of
TCDD from corn oil (AB Acorn Oii=87%). Therefore, the appropriate RBA for TCDD in soil
would be (Equation 10):
PDA _ ABATCDD:SOU
KtlATCDDsoii - — - Eq. (10)
ABATCDDcornoa
and the appropriate application of the RBA to the TCDD Hazard Quotient (HQ) would be
(Equation 11)
HQ = - ™L - = ADlTCDD,soiVRBATCDD,soil
V1LUU RfDTCDD/RBATCDDtSoil RfDTCDD M ^ '
However, EPA assesses risks for total TCDD TEQ in soil, not just for TCDD alone. Therefore,
the RBAsou in Equation 1 1 must represent the RBAsoii for TEQ (RBAxEQ) and not just the RBA for
TCDD.
This raises several problems. The 87% ABA assumption used in the basis for the RfD represents
the bioavailability of TCDD and would not necessarily apply to the bioavailability of TEQ for a
mixture of dioxin congeners because bioavailability appears to be dependent on chlorination
(USEPA, 2010). Therefore, the appropriate RBA adjustment for TEQ in soil would be (Equation
12):
RBATEOsoil = •
ABATEQ corn oii ABATCDDcOrnoil
and the appropriate application of the RBA to the TEQ Hazard Quotient (HQ) would be
(Equation 13):
HO = 'so Ea
VTEQ RfDTCDD/RBATEQjSoil RfDTCDD 4'
Operationally, this translates into two requirements for a soil RBA bioassay for TEQ to be used
in noncancer risk assessment: (1) estimate of RBA for TEQ in corn oil (ABATEQ,com
oii/AB ATcDD,com oil); and (2) estimate of the RBA for TEQ in soil (ABATEQ,soii /AB ATEQ,com oil).
-------
Soil Dioxin Relative Unavailability Assay Evaluation Framework
2.4. RBAjEQ in Soil for Cancer Risk Assessment
Requirement 4: For cancer risk assessment two RBA estimates are needed: (1) estimate of RBA
for TEQ in food (ABAxEQ,food/ ABAxcDD,food); and (2) estimate of the RBA for TEQ in soil
soil /ABAxEQ,food)-
According to USEPA 201 1, EPA is considering a cancer oral slope factor (OSF) based on
epidemiology of occupational cohort(s). These cohorts experienced multi-pathway exposures
(e.g., inhalation, dermal, dust ingestion). The dose metric in the dose-response modeling that
supports the OSF is blood TCDD. The Point of Departure (POD) was translated into an average
daily intake by use of a pharmacokinetics model which assumed 80% bioavailability of TCDD in
the diet (a source for this value was not found in a cursory review of USEPA 201 1 or its
precursor, USEPA 2003).
Based on logic similar to that described above for the noncancer risk assessment, the appropriate
RBA adjustment for soil TEQ for use in cancer risk assessment would be (Equations 14 and 15):
RBATEO soil = - oo
TEQ.SOU ABATEQfood ABATCDDJood 4
- ADITEQ-OSFTCDD „
TEQ - — — - Eq.
Operationally, this translates into 2 requirements for a soil RBA bioassay for TEQ for use in
cancer risk assessment: (1) estimate of RBA for TEQ in food (ABATEQ,food/ ABATcDD,food); and
(2) estimate of the RBA for TEQ in soil (ABATEQ,soii /ABATEQ,food)-
2.5. Selection of Animal Model for Predicting RBA in Humans
Requirement 5: There is no general consensus on the preferred animal model for estimating
RBA for PCDD/F. RBA assays for congener mixtures in soil have been conducted in rats and
swine, and these two assay yield different estimates
Differences are evident between RBA estimates for test soils assayed in swine and rats (USEPA,
2010). This included large differences in the average RBA values for the same test material
assayed in swine and rats (Budinsky et al., 2008), as well as regression coefficients for the effect
-------
Soil Dioxin Relative Unavailability Assay Evaluation Framework
of congener chlorine content on KB A that are in opposite directions. KB A varies with congener
chlorination. The direction of the relationship (i.e., positive or negative slope) is not the same
when estimated based on data from swine or rat assays. Data from swine assays indicates an
increase in RBA with increasing chlorine content (Budinsky et al., 2008; Wittsiepe et al., 2007),
whereas, data from rat assays indicates a decrease in RBA with increasing chlorination
(Budinsky et al., 2008; Finley et al., 2009). These differences suggest substantially different
RBA estimates may be obtained depending on the animal model used. The dependence of RBA
on congener chlorination suggests that soil RBA will depend on the congener composition of the
soil (as well as the bioassay used to estimate RBA).
2.6. Dosing Regimen
Requirement 6a: External doses of TEQ should not exert overt systemic toxicity that alters
PCDD/F distribution or impairs elimination (metabolism or excretion). External dose should be
well below the LDso and preferably, well below to LDoi.
Requirement 6b: Multiple dose levels of TEQ should be administered to allow and evaluation
of the dependence of RBA on dose.
Requirement 6c: External doses of TEQ delivered in the test (e.g., soil) and reference material
(e.g., corn oil) must result is similar (or overlapping ranges) of internal doses of TEQ. This is
needed to prevent different levels of induction of CYP450 and different elimination fractions of
TEQ for the test and reference material.
Requirement 6d: There is no general consensus as to whether single doses or repeated doses
should be administered. Regardless of the dosing schedule, a sufficient cumulative (and non-
toxic) external dose must be delivered to allow quantification of the internal dose of the
administered congeners that comprise >95% of the administered TEQ.
Requirement 6e: For assay of RBA of PCDD/F in soils, the administered soil should be the
<250 |im fraction
As noted previously in reference to Equations 7 and 8, measurement of the elimination fraction
(EF) is not needed in the calculation of RBA as long as the elimination fraction is not different
following administration of the PCDD/F dose in test or reference materials. However, because
the internal TEQ dose (e.g., liver dose) can induce CYP450 (which increases elimination rate),
the elimination fraction may vary with internal TEQ dose. Therefore, dosing regimens for the
test and reference materials should be matched to achieve similar internal TEQ doses (Finley et
al., 2009; USEPA, 2010). Establishing internal dose equivalents for TEQ requires forehand
knowledge of TEQ RBA for the test material of interest, which, of course, will not be known (if
-------
Soil Dioxin Relative Unavailability Assay Evaluation Framework
it were, there would be no need to assay the test material). Therefore, administering multiple
dose levels of TEQ is recommended to achieve overlap of the corresponding internal TEQ doses.
Use of multiple dose levels will allow evaluation of the external dose-internal dose relationship
and detection of any nonlinearities that might suggest dose-dependence of elimination kinetics.
The RBA can be calculated from the regression relationships for the reference and test materials
(USEPA, 2007b).
Calculation of RBA for total TEQ in the test material requires that the internal doses of TEQ
contributed from each administered congener be quantified and summed (Equation 9). To
achieve this, the administered dose of each congener must be sufficient to achieve a
corresponding internal dose that is above the detection limit. Those congeners that are below the
detection limit must be assigned values that will introduce uncertainty into the RBA estimate
(e.g., one half detection limit). There is no general consensus as to whether single doses or
repeated doses should be administered. Given the relatively slow elimination kinetics, it is
unlikely that steady state conditions are feasible. However, repeated dosing will allow the
accumulation of the more rapidly eliminated congeners and congeners having low RBA, and
may improve detection and quantification of these congeners in the internal dose. Whether or
not single or repeated dosing is feasible will depend, in part, on the animal model selected.
Detection and quantification of all congeners in the internal dose may not always be possible for
congeners have very low RBA. Minimum objectives for quantification of the internal dose
should be established in the study design and results evaluated against these objectives. As a
general default, the administered doses should ensure detection of >95% of the administered
TEQ.
In risk assessment applications, the grain size fraction that is most likely to adhere to human skin
is typically of primary importance. It is generally accepted that for moisture contents found in
typical surface soils, this is the <250 jim fraction (Berstrom et al., 2011; Kissel et al., 1996;
Siciliano et al., 2009; Yamamoto et al., 2006). Therefore, unless a strong argument can be made
an alternative, the assay should estimate the RBA for the <250 jim fraction.
2.7. Measurement of Internal TEQ Dose
Requirement 7: Tissues selected for assay of PCDD/F congeners should provide reliable
predictions of the TEQ body burden. There is no general consensus regarding which tissue
would satisfy this requirement, and it is likely to vary across animal species. Ideally, if whole
body (gastrointestinal tract excluded) is not analyzed for TEQ, selected tissues should include
those that collectively contribute >50% of total body burden. At a minimum, this should include
liver and adipose.
-------
Soil Dioxin Relative Unavailability Assay Evaluation Framework
Calculation of KB A for TEQ requires quantification of the relationship between the administered
(external) dose and internal dose (Equations 7 and 8). Absorbed PCDD/F is widely distributed
and partitions into tissue lipid. Therefore, the internal dose is the total TEQ body burden
(excluding unabsorbed TEQ in the gastrointestinal tract). Ideally this could be achieved by
quantifying the entire body burden of administered PCDD/F congeners, however, this may not
be feasible for most animal models. In most mammalian species in which the whole body
distribution of PCDD/F (e.g., TCDD) has been studied, that largest fractions of the body burden
>50% reside in liver and adipose (USEPA, 2003). RBA assays of congener mixtures have
measured internal dose as PCDD/F concentrations or burdens of liver (Finley et al., 2009), liver
plus adipose combined (Budinsky et al., 2008) or combined adipose, blood, brain, liver, and
muscle (Wittsiepe et al., 2007). Given the relatively large contribution of adipose and liver to
body burden, at a minimum, these two tissues should be assayed. Estimation of the PCDD/F
burdens in tissue requires measurement of the PCDD/F concentrations and the total mass of each
tissue. This is easily accomplished for the liver but is more difficult for adipose. The
experimental design should address how the adipose mass (or volume) is to be estimated and, if
not actually measured, what assumptions are to be made about it mass or volume.
2.8. Confidence in RBA Estimates
Requirement 8: The study design must provide: (1) statistical confidence limits on the estimate
(e.g., 95% confidence limits) of the RBA and; (2) an evaluation of reproducibility of RBA
estimates when the same test materials are assayed.
The RBA calculation shown in Equation 7 is typically a ratio of mean ABA values obtained from
a sample of measurements of ABA from a group or groups of animals that received doses of the
test or reference material. The resulting RBA from Equation 7 represents an estimate of the
mean RBA. Estimating confidence limits on the mean RBA requires estimating the confidence
limit on a ratio of mean values for ABA, where each mean has an associated uncertainty that
must be estimated from the sample distributions. Several different computational strategies for
calculating confidence limits on the RBA from single or multiple dose level assays of PCDD/F
have been described (USEPA, 2007a; 2010). These include application of Fieller's theorem and
bootstrap methods. The statistical design for estimating confidence on the RBA should be
articulated in the study design.
In addition to confidence limits on each RBA estimate, reproducibility of RBA estimates should
be evaluated. The only way to accomplish this is to assay the same test material several times
and compare outcomes. Where this is not feasible (e.g., budget limitations) the study design
must address how uncertainty in the reproducibility of the assay would be addressed in any
application of the RBA estimate to risk assessment.
-------
Soil Dioxin Relative Unavailability Assay Evaluation Framework
2.9. Soil Characterization
Requirement 9: Study designs intended to estimate KB A of PCDD/F in soils should include a
characterization of the soil, including a complete analysis of PCDD/F congeners, as well as soil
characteristics. Minimum soil characteristics should include total solids, pH, total organic
carbon, and grain size distribution.
The expectation is that adherence of PCDD/F to constituents of soil (e.g., organic carbon) is an
important determinant of KB A. The soil characteristics that most greatly influence PCDD/F
KB A have not been identified. However, an important objective will be to utilize data obtained
from soil KBA studies, data on soil characteristics, and in vitro extraction methods to establish
methods to predict KBA that circumvent the need for expensive animal bioassays. Therefore,
collection of data on the characteristics of soils (composition, mineralogy) that are assayed is
highly desirable. At a minimum, soil should be evaluated for PCDD/F congener composition,
total solids, pH, total organic carbon, and grain size distribution.
3. Summary and Conclusions
This report provides the basis for minimum requirements of assays intended to estimate KBA of
PCDD/F in soils for applications to risk assessment. Given that the methodology for assaying
PCDD/F KBA in soils is evolving, greater experience with various experimental designs is likely
to prompt modifications to the requirements identified in this report. The minimal requirements
identified in this report are summarized in Table 2.
4. References
Bergstrom, C., Shirai, J., Kissel, J., 2011. Particle size distributions, size concentration
relationships, and adherence to hands of selected geologic media derived from mining, smelting,
and quarrying activities. Sci Total Environ 409:4247-4256.
Bonaccorsi A, diDomenico A, Fanelli R, Merlin F, Motta R, Vanzati R, Zapponi GA. 1984. The
influence of soil particle adsorption on 2,3,7,8-tetrachlorodibenzo-p-dioxin biological uptake in
the rabbit. Disease, metabolism, and reproduction in the toxic response to drugs and other
chemicals. Arch Toxicol Suppl 7:431-434.
10
-------
Soil Dioxin Relative Unavailability Assay Evaluation Framework
Budinsky RA, Rowlands JC, Casteel S, Pent G, Gushing CA, Newsted J, Giesy JP, Ruby MV,
Aylward LL. 2008. A pilot study of oral bioavailability of dioxins and furans from contaminated
soils: Impact of differential hepatic enzyme activity and species differences. Chemosphere
70(10): 1774-1786.
Finley B, Fehling K, Warmerdam J, Morinello EJ. 2009. Oral bioavailability of polychlorinated
dibenzo-p-dioxins/dibenzofurans in industrial soils. Hum Ecol Risk Assess 15:1146-1167.
Kissel, J.C., Richter, K.Y., Fenske, R.A., 1996. Factors affecting soil adherence to skin in hand-
press trials. Bull Environ Contam Toxicol 56:722-728.
Lucier GW, Rumbaugh RC, McCoy Z, Hass R, Harvan D, Albro P. 1986. Ingestion of soil
contaminated with 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) alters hepatic enzyme activities
in rats. Fundam Appl Toxicol 6:364-371
McConnell EE, Lucier GW, Rumbaugh RC, Albro PW, Harvan DJ, Hass JR, Harris MW. 1984.
Dioxin in soil: Bioavailability after ingestion by rats and guinea pigs. Science 223:1077-1079.
Poiger, H, Schlatter, C (1986) Pharmacokinetics of 2,3,7,8-TCDD in man. Chemosphere 15:
1489-1494.
Shu H, Paustenbach D, Murray FJ, Marple L, Brunck B, Rossi DD, Teitelbaum P. 1988.
Bioavailability of soil-bound TCDD: Oral bioavailability in the rat. Fund Appl Toxicol 10:648-
654.
Siciliano, S.D., James, K., Zhang, G., Schafer, A.N., Peak, J.D., 2009. Adhesion and enrichment
of metals on human hands from contaminated soil at an Arctic urban brownfield. Environ Sci
Technol 43:6385-6390.
Yamamoto, N., Takahashi, Y., Yoshinaga, J., Tanaka, A., Shibata, Y., 2006. Size distributions of
soil particles adhered to children's hands. Arch Environ Contam Toxicol 51:157-163.
Umbreit TH, Hesse EJ, Gallo MA. 1986. Bioavailability of dioxin in soil from a 2,4,5-T
manufacturing site. Science 232:497-499.
U.S. EPA (U.S. Environmental Protection Agency). 1989. Risk Assessment Guidance for
Superfund: Volume III - Part A, Process for Conducting Probabilistic Risk Assessment. Office of
Emergency and Remedial Response, U.S. Environmental Protection Agency, Washington, DC
(Available online at: http ://www. epa. gov/oswer/ri skassessment/rags3 adt/).
U.S. EPA (U.S. Environmental Protection Agency). 2003. Exposure and Human Health
Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds. NAS
Review Draft, December 2003. National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Washington, DC.
Available online at: http://www.epa.gov/ncea/pdfs/dioxin/nas-review/
11
-------
Soil Dioxin Relative Unavailability Assay Evaluation Framework
U.S. EPA (U.S. Environmental Protection Agency). (2007a) Estimation of Relative
Bioavailability of Lead in Soil and Soil-Like Materials by In Vivo and In Vitro Methods. U.S.
Environmental Protection Agency, Office of Solid Waste and Emergency Response.
Washington, DC 20460. OSWER 9285.7-77. Available online at:
http://www.epa.gov/superfund/heal th/contaminants/bioavailability/lead_tsd_main.pdf.
U.S. EPA (U.S. Environmental Protection Agency). (2007c) Guidance for Evaluating the Oral
Bioavailability of Metals in Soils for Use in Human Health Risk Assessment. U.S.
Environmental Protection Agency, Office of Solid Waste and Emergency Response.
Washington, DC 20460. OSWER 9285.7-80. Available online at:
http://www.epa.gov/superfund/heal th/contaminants/bioavailability/bio_guidance.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2011) Final Report
Bioavailability of Dioxins and Dioxin-Like Compounds in Soil. U.S. Environmental Protection
Agency, Office of Superfund Remediation and Technology Innovation, Environmental Response
Team - West Las Vegas, NV 89119. Available online at:
http://epa.gov/superfund/health/contaminants/dioxin/pdfs/Final dioxin RBAReport 12 20 10.
p_df
U.S. EPA (U.S. Environmental Protection Agency). (2011) EPA's Reanalysis of Key Issues
Related to Dioxin Toxicity and Response to NAS Comments (External Review Draft). National
Center for Environmental Assessment, Office of Research and Development, U.S. Environmental
Protection Agency, Washington, DC. EPA/600/R-10/038A. Available online at:
http://cfpub.epa.gov/ncea/iris_drafts/recordisplay.cfm?deid=222203
Van den Berg M, Birnbaum LS, Denison M, De Vito M, Farland W, Feeley M, Fiedler H,
Hakansson H, Hanberg A, Haws L, Rose M, Safe S, Schrenk D, Tohyama C, Tritscher A,
Tuomisto J, Tysklind M, Walker N, Peterson RE. 2006. The 2005 World Health Organization re-
evaluation of human and mammalian toxic equivalency factors for dioxins and dioxin-like
compounds. Toxicol Sci 93(2):223-241.
Wendling T, Hileman F, Orth R, Umbreit T, Hesse, Gallo M. 1989. An analytical assessment of
the bioavailability of dioxin contaminated soils to animals. Chemosphere 18:925-932.
Wittsiepe J, Erlenkamper B, Welge P, Hack A, Wilhelm M. 2007. Bioavailability of PCDD/F
from contaminated soil in young Goettingen mini-pigs. Chemosphere 67(9):S355-S356.
Yamamoto, N., Takahashi, Y., Yoshinaga, J., Tanaka, A., Shibata, Y., 2006. Size distributions of
soil particles adhered to children's hands. Arch Environ Contam Toxicol 51:157-163.
12
-------
Soil Dioxin Relative Unavailability Assay Evaluation Framework
Table 1. Reported Variations in Experimental Designs for TCDD or PCDD/F RBA Assays
Experimental Design Parameter
Animal models
Soil test materials
Dosages
Dosing regimens
Dose vehicles for soil
Dose vehicles for reference
Measured bioavailability metrics
Interval between dosing and tissue
collection
Data reduction methods
Implemented Design
• Guinea pig
• Rabbit
• Rat
• Swine
• In situ contaminated soil
• Laboratory spiked soil
• Subtoxic
• Systemically toxic
• Similar tissue levels of PCDD/F achieved
received soil and reference
• Similar tissue levels of PCDD/F achieved
received soil and reference
in animals that
in animals that
• Single dose
• Repeated dose
• Single dose level
• Multiple dose levels
• Aqueous suspension
• Food mix (e.g., dough ball)
• Acetone/corn oil
• Acetone/hexane
• Corn oil
• Gum acacia
• Liver TCDD
• Liver PCDD/F
• Adipose, liver PCDD/F
• Adipose, blood, brain, liver, muscle PCDD/F
• 1 day
• 6 days
• 7 days
• 30 days
• 60 days
• SoiLreference tissue concentration ratio
• SoiLreference slope ratio for dose-tissue PCDD/F
• Absolute bioavailability based on intravenous reference
dosing
Based on Bonaccorsi et al., 1984; Budinsky et al., 2008; Finley et al., 2009; Lucier et al., 1986; McConnell et
al., 1984; Shu et al., 1988; Umbreit et al., 1986; Wendling et al., 1989;Wittsiepe et al., 2007)
13
-------
Soil Dioxin Relative Unavailability Assay Evaluation Framework
Table 2. Minimum Experimental Design Requirements for PCDD/F RBA Assays
Design Parameter
RBA for TEQ
Calculating RBA TEQ in soil
RBATEQ in soil for noncancer
risk assessment
RBATEQ in soil for cancer
risk assessment
Animal model
Dosing regimen
Internal TEQ dose metrics
Confidence in RBA
Estimates
Soil characterization
#
1
2
3
4
5
6a
6b
6c
6d
6e
7
8
9
Requirement
PCDD/F risk assessment requires estimates of the RBA for soil TEQ
(RBATEO)
Calculation of RBATEQ requires quantification of the total TEQ external
dose and total TEQ internal dose, as well as the excretion fraction for
TEQ (or experimental designs that ensure that the excretion fractions for
TEQ are the same when administered in the soil or reference material).
For noncancer risk assessment two RBA estimates are needed: (1) RBA
for TEQ in corn oil (ABATEQ!Comoii/ABATCDD,comoii); and (2) RBA for
TEQ in soil (ABATEO,SOii /AB ATEO,com oil).
For cancer risk assessment two RBA estimates are needed: (1) estimate
of RBA for TEQ in food (ABATEQ food/ ABATCDD food); and (2) estimate of
the RBA for TEQ in soil (ABATEO,soll /ABATEafood).
There is no general consensus on the preferred animal model for
estimating RBA for PCDD/F. RBA assays for congener mixtures in soil
have been conducted in rats and swine, and these two assay yield
different estimates of RBATE0.
External doses of TEQ should not exert overt systemic toxicity that alters
PCDD/F distribution or impairs elimination (metabolism or excretion).
External dose should be well below the LD50 and preferably, well below
toLDoi.
Multiple dose levels of TEQ should be administered to allow and
evaluation of the dependence of RBA on dose.
External doses of TEQ delivered in the test (e.g., soil) and reference
material (e.g., corn oil) must result in similar (or overlapping ranges) of
internal doses of TEQ. This is needed to prevent different levels of
induction of CYP450 and different elimination fractions of TEQ for the
test and reference material.
There is no general consensus as to whether single doses or repeated
doses should be administered. Regardless of the dosing schedule, a
sufficient cumulative (and non-toxic) external dose must be delivered to
allow quantification of the internal dose of the administered congeners
that comprise >95% of the administered TEQ.
For assay of RBA of PCDD/F in soils, the administered soil should be the
<250 um fraction
Tissues selected for assay of PCDD/F congeners should provide reliable
predictions of the TEQ body burden. There is no general consensus
regarding which tissue would satisfy this requirement, and it is likely to
vary across animal species. Ideally, if whole body (gastrointestinal tract
excluded) is not analyzed for TEQ, selected tissues should include those
that collectively contribute >50% of total body burden. At a minimum,
this should include liver and adipose.
The study design must provide: (1) statistical confidence limits on the
estimate (e.g., 95% confidence limits) of the RBA and; (2) an evaluation
of reproducibility of RBA estimates when the same test materials are
assayed.
Study designs intended to estimate RBA of PCDD/F in soils should
include a characterization of the soil that includes a complete analysis of
PCDD/F congeners, as well as soil characteristics to including, at a
minimum: total solids, pH, total organic carbon, and grain size
distribution.
14
------- |