United States
Environmental Protection
^ir KmI B »Agency
EPA/600/R-09/072
STATUS REPORT:
Advances in Inhalation Dosimetry
of Gases and Vapors with
Portal of Entry Effects
in the Upper Respiratory Tract
September 2009
U.S. Environmental Protection Agency
Washington, DC
-------
DISCLAIMER
This report has been reviewed in accordance with U.S. Environmental Protection Agency
Policy and approved for publication. Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
September 2009
11
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
AUTHORS
John J. Stanek—National Center for Environmental Assessment, U.S. Environmental Protection
Agency, Research Triangle Park, NC
Gary L. Foureman—National Center for Environmental Assessment, U.S. Environmental
Protection Agency, Research Triangle Park, NC
REVIEWERS
This report has been reviewed by scientists internal to the EPA.
RFC WORKGROUP CONTRIBUTORS AND REVIEWERS
Robert Benson—Region 8, U.S. Environmental Protection Agency, U.S. Environmental
Protection Agency, Denver, CO
James S. Brown—National Center for Environmental Assessment, U.S. Environmental
Protection Agency, Research Triangle Park, NC
Ernest Falke—Office of Pollution Prevention and Toxics, U.S. Environmental Protection
Agency, Washington, DC
Anna Lowit—Office of Pesticide Programs, U.S. Environmental Protection Agency,
Washington, DC
Eva McLanahan—National Center for Environmental Assessment, U.S. Environmental
Protection Agency, Research Triangle Park, NC
Deirdre Murphy—Office of Air Quality Planning and Standards, U.S. Environmental Protection
Agency, Research Triangle Park, NC
Babasaheb Sonawane—National Center for Environmental Assessment, U.S. Environmental
Protection Agency, Washington, DC
Paul Schlosser—National Center for Environmental Assessment, U.S. Environmental Protection
Agency, Washington, DC
Marc Stifelman—Region 10, U.S. Environmental Protection Agency, Seattle, WA
September 2009
in
-------
Ravi Subramanian—National Center for Environmental Assessment, U.S. Environmental
Protection Agency, Washington, DC
John Whalan—National Center for Environmental Assessment, U.S. Environmental Protection
Agency, Washington, DC
Paul White—National Center for Environmental Assessment, U.S. Environmental Protection
Agency, Washington, DC
INTERNAL EPA REVIEWERS
Lynn Flowers—National Center for Environmental Assessment, U.S. Environmental Protection
Agency, Washington, DC
Connie Meacham—National Center for Environmental Assessment, U.S. Environmental
Protection Agency, Research Triangle Park, NC
Molly Rosett—Student Services Contractor, National Center for Environmental Assessment,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Reeder Sams—National Center for Environmental Assessment, U.S. Environmental Protection
Agency, Research Triangle Park, NC
John Vandenberg—National Center for Environmental Assessment, U.S. Environmental
Protection Agency, Research Triangle Park, NC
Deborah Wales—National Center for Environmental Assessment, U.S. Environmental Protection
Agency, Research Triangle Park, NC
Debra Walsh - National Center for Environmental Assessment, U.S. Environmental Protection
Agency, Research Triangle Park, NC
EXTERNAL REVIEWERS
Harvey Clewell—The Hamner Institute, Research Triangle Park, NC
Jeffry Schroeter—The Hamner Institute, Research Triangle Park, NC
James Ultman—Pennsylvania State University, University Park, PA
September 2009
iv
-------
TABLE OF CONTENTS
AUTHORS, CONTRIBUTORS. AND REVIEWERS Ill
TABLE OF CONTENTS V
LIST OF TABLES VII
GLOSSARY II
ACRONYMS AND ABBREVIATIONS V
EXECUTIVE SUMMARY VII
1. INTRODUCTION AND PURPOSE 1-1
1.1. PURPOSE 01 THIS REPORT 1-1
1.2. A PERSPECTIVE ON INHALATION DOSIMETRY 1-1
1.2.1. Historical Inhalation Dosimetry 1-2
2. REVIEW OF THE 1994 RFC METHODS FOR PORTAL-OF-ENTRY GAS D0SIMETRY2-1
2.1. GAS CATEGORIZATIONS - GENERAL 2-1
2.1.1. The RfC Methods Gas Categorization Scheme 2-1
2.2. CONCEPTUAL AND HISTORICAL BASIS FOR COMPARATIVE DOSIMETRY
OF INSPIRED GASES IN RFC METHODS - MINUTE
VENTILATION/SURFACE AREA OF RESPIRATORY TRACT (VE/SART) . 2-5
2.2.1. Factors Controlling Comparative Inhaled Dose 2-5
2.3. NORMALIZATION OF INHALED CONCENTRATION TO SURFACE AREA OF
RESPIRATORY TRACT REGIONS 2-12
2.3.1. Dose-Response in Respiratory Tract Tissues is Based on External Exposure
Concentration 2-12
2.3.2. Normalization of External Exposure Concentration to Surface Areas.... 2-13
2.4. APPLICATION OF VE/SA IN CALCULATION OF THE HUMAN EQUIVALENT
CONCENTRATION, HEC: THE DEFAULT APPROACH FOR INSPIRED
GASES 2-14
2.4.1. The Dosimetric Adjustment Factor, DAF 2-15
2.4.2. The DAF for Gases; the Regional Gas Dose Ratio, RGDRr 2-15
2.4.3. Some Example HEC Calculations with Commentary 2-16
2.5. ASSUMPTIONS AND LIMITATIONS IN VE/SA 2-18
2.5.1. The Ve/SA Assumptions 2-19
3. SCIENCE DEVELOPMENTS IN PORTAL-OF-ENTRY GAS DOSIMETRY 3-1
3.1. INTRODUCTION 3-1
3.2. A MODIFIED GAS SCHEME: DESCRIPTORS VERSUS CATEGORIES 3-2
3.3. FLOW-DYE CAST MODELS 3-4
September 2009
v
-------
3.4. COMPUTER MODELING OF FLUID FLOW - INTRODUCTION TO CFD 3-6
3.4.1. CFD Air Flow Models of the Rat ET Region- Correlation with Dye-
Streamlines 3-8
3.4.2. CFD Air Flow Models of the Human ET Region 3-10
3.4.3. CFD Air Flow Models - Predictions of Reactive Gas Distribution in the ET
Region 3-11
3.4.4. Interspecies CFD Air Flow Models Predictions of Gas Distribution in the
ET Region 3-13
3.4.5. Range and Distribution of Flux in ET Regions for Various Species 3-16
3.4.6. Correlation of High Flux with Lesions in the ET Region 3-16
3.5. EVALUATION AND USE OF MODELS IN INTERSPECIES INHALATION
DOSIMETRY 3-21
3.5.1. Overview of CFD-PBPK Hybrid Modeling - Combination of Gas
Transport in the Air Phase into the Liquid/Tissue Phase 3-22
3.5.2. CFD-PBPK Hybrid Modeling and the Overall Mass Transport Coefficient -
Kg 3-25
3.5.3. Results and Analysis of Interspecies Inhalation Dosimetry Modeling.... 3-28
3.5.4. Summary and Conclusions - Modeling Developments and Results 3-42
4. REFERENCES 4-1
September 2009
vi
-------
LIST OF TABLES
Table 2-1. Respiratory tract regions 2-7
Table 2-2. Default surface area values for respiratory effects 2-11
Table 2-3. Hierarchy of model structures for dosimetry and interspecies extrapolation .... 2-14
Table 2-4. RGDRs and minute ventilation/surface area (VE/SA) ratios of the ET, TB, and PU
regions of respiratory tract for various species and humans 2-16
Table 3-1. Nasal flow stream allocation in the rat and summary of CFD simulated flow
apportionment (as a % of total volumetric flow) to various cross-sectional areas
(as mm2) of the ET region 3-9
Table 3-2. Summary of CFD simulated flow apportionment (as a % of total at 15 L/min) on
the cross-sectional area in the middle turbinate (as mm2) of the ET region in
selected human models as analyzed by Wen et al. (2008) 3-11
Table 3-3. Summary of airway flow studies using CFD techniques 3-14
Table 3-4. Estimates of formaldehyde flux to ET surface walls for various species 3-16
Table 3-5. Primary toxicological endpoint(s), uptake, properties, and physicochemical
descriptor for representative gases 3-30
Table 3-6. Animal (rat) to human ratios of modeled or calculated gas Kg and kg values for
each representative gas 3-36
Table 3-7. Comparison of approaches for calculating the DAF for representative gases in
determining the HEC - portal of entry ET or nasal effects 3-37
September 2009
vii
-------
LIST OF FIGURES
Figure 2-1. Gas categorization scheme based on water solubility and reactivity as major
determinants of gas uptake 2-2
Figure 2-2. Diagrammatic representation of the three respiratory tract regions designated in
humans 2-8
Figure 2-3. A representative model of symmetric (A) and asymmetric (B) airway branching
patterns occurring in humans and murine laboratory animals species, respectively.
2-9
Figure 2-4. Calculation and application of VE/SAET in the RjCMethods default approach for
interspecies dosimetric extrapolation through derivation of the regional gas dose
ratio for the extrathoracic region, RGDRET 2-19
Figure 3-1. Representation of the assumptions of uniformity following from VE /SA as applied to
comparative gas dosimetry 3-1
Figure 3-2. A schematic representation of the physicochemical properties of reactivity and water
solubility overlaid with descriptors of their practical limits 3-4
Figure 3-3. Inspiratory airflow patterns in upper respiratory tract of F344 rat and Rhesus
monkey 3-6
Figure 3-4. Major inspiratory flow streams simulated in the F344 rat nose. The external nostril is
to the right 3-9
Figure 3-5. Representations of CFD-simulations of flow streamlines in human nasal cavities 3-11
Figure 3-6. Patterns of simulated wall mass flux of formaldehyde across the nasal septum of the
rat 3-13
Figure 3-7. Nasal wall flux spectra of inhaled formaldehyde simulated in rats, monkey and
humans at normal inspiratory flow rates 3-15
Figure 3-8. Graphs showing (A) the incidence of formaldehyde-induced squamous metaplasias
and (B) modeled formaldehyde flux values along regions assigned to the
perimeter of a transected nasal airway of rats 3-18
Figure 3-9. Schematic diagram of the transverse nasal section 3-19
Figure 3-10. Representation of application of current information to the processes, assumptions,
influences and outcomes for RfC Methods basic default procedures for
comparative gas dosimetry 3-21
Figure 3-11. Schematic of the CFD-PBPK hybrid model of the rat nasal cavity 3-24
Figure 3-12. Plot of Kg (overall MTC) for respiratory epithelium (RE) in rats and humans versus
partition coefficient for the gases shown in Tables 3-5, 3-6, and 3-7 3-40
Figure 3-13. Plot of Kg (overall MTC) for olfactory epithelium (OE) in rats and humans versus
partition coefficient for the gases shown in Tables 3-5, 3-6, and 3-7 3-41
Figure 3-14. A revised schematic representation of the outcomes for interspecies inhalation
dosimetry of gases for the ET region following from the advances presented.. 3-43
-------
GLOSSARY
Aerosol - A suspension of liquid or solid particles in air.
Critical Effect - The first adverse effect, or its known precursor, that occurs as the dose rate
increases. Designation is based on evaluation of overall data base.
Chronic Exposure - Multiple exposures occurring over an extended period of time, or a
significant fraction of the animal's or the individual's lifetime.
Computational fluid dynamics (CFD) - A branch of fluid mechanics that uses numerical
methods and algorithms to solve and analyze problems of fluid flows. Flows may apply to liquid
and gases, including inspired and expired air, and are thus applicable to solving flows within the
respiratory tract. The fundamental bases of any CFD problem are the Navier-Stokes equations,
which define any single-phase fluid flow. These equations can be simplified by removing terms
describing viscosity to yield the Euler equations.
Diffusivity (gas) - The transport of matter from one point to another by random molecular
motions to become equalized with respect to concentration. For gases, rates of diffusion increase
with the temperature and are inversely proportional to the pressure. The interdiffusion
coefficients of gas mixtures are almost independent of the composition. Kinetic theory shows
that diffusion of a pure gas is inversely proportional to both the square root of the molecular
weight and the square of the molecular diameter.
Dorsal - On or near the upper surface (of the nasal tract).
Dosimetric Adjustment Factor (DAF) - A multiplicative factor used to adjust observed
experimental or epidemiological data to human equivalent concentration (HEC) for assumed
ambient scenario. See also regional gas dose ratio (RGDR).
Flux - The rate of flow of energy, gas or particles across a given surface.
Gas - Term referring to a compressible fluid phase of a substance. Fixed gases are gases for
which no liquid or solid can form at the temperature of the gas, such as air at ambient
temperatures.
Inhalation Reference Concentration (RfC) - An estimate (with uncertainty spanning perhaps
an order of magnitude) of a continuous inhalation exposure to the human population (including
sensitive subgroups) that is likely to be without an appreciable risk of deleterious noncancer
health effects during a lifetime. The inhalation reference concentration is for continuous
inhalation exposures and is appropriately expressed in units of mg/m3.
September 2009
11
-------
Henry's Law Constant - The law can be expressed in several equivalent forms, a convenient
form being: Cg = HC1 where Cg and CI are the gas-(g) and liquid-(l) phase concentrations. The
constant (H) is the ratio at equilibrium of the gas phase concentration to the liquid-phase
concentration of the gas (i.e., moles per liter in air/moles per liter in solution).
Kg - The overall mass transfer coefficient describing movement of gas from the air phase into
the liquid phase of the respiratory tract (see also MTC).
kg - The gas phase mass transfer coefficient describing movement of gas from the gas phase to
liquid/tissue boundary (see also MTC).
Lowest-Observed-Adverse-Effect Level (LOAEL) - The lowest exposure level at which there
are statistically and/or biologically significant increases in frequency or severity of adverse
effects between the exposed population and its appropriate control group.
Mass Transfer Coefficient (MTC) - A diffusion rate constant that relates the mass transfer rate,
mass transfer area, and concentration gradient as driving force between and through phases.
These coefficients may also be viewed in terms of resistance to flow and movement. For
purposes of this report (with phases of gas and solid) MTC requires units of mass, time, distance,
and concentration: mol/(s-m2), mol/m3, or m/s. Examples of MTCs used in this report relate to
movement of gases in the respiratory tract. They include the MTC designated for the gas phase
only, kg, and an overall MTC inclusive of both the gas and liquid phases, Kg.
Meatus - A body opening or passage, such as the openings in the nose or ear.
No-Observed-Adverse-Effect Level (NOAEL) - An exposure level at which there are no
statistically and/or biologically significant increases in the frequency or severity of adverse
effects between the exposed population and its appropriate control. Some effects may be
produced at this level, but they are not considered as adverse, nor immediate precursors to
specific adverse effects. In an experiment with several NOAELs, the assessment focus is
primarily on the highest one for a given critical effect, leading to the common usage of the term
NOAEL as the highest exposure without adverse effect.
Physiologically-Based Pharmacokinetic (PBPK) Modeling - A mathematical modeling
technique for predicting the absorption, distribution, metabolism and excretion of a compound in
humans and other animal species. PBPK models strive to be mechanistic by mathematically
transcribing anatomical, physiological, physical, and chemical descriptions of the phenomena
involved in complex pharmacokinetic processes. These models have an extended domain of
applicability compared to that of classical, empirical function based, compartmental
pharmacokinetic models.
Portal-of-Entry Effect - A local effect produced at the tissue or organ of first contact between
the biological system and the toxicant.
September 2009
111
-------
Regional Gas Dose (RGDr) - The gas dose (mg/cm2) of respiratory tract surface area (per
minute) calculated for the respiratory tract region of interest (r) as related to the observed toxicity
(e.g., calculated for the tracheobronchial region for an adverse effect in the conducting airways).
Regions of interest may be the extrathoracic (ET), tracheobronchial (TB), or pulmonary (PU).
Regional Gas Dose Ratio (RGDRr) - The ratio of the deposited gas dose in a respiratory tract
region (r) for the laboratory animal species of interest to that of humans. This ratio is used to
adjust the observed gas exposure level for interspecies dosimetric differences.
Uncertainty Factor (UF) - One of several, generally 3- to 10-fold factors, used in operationally
deriving the inhalation reference concentration (RfC) from experimental data. UFs are intended
to account for (1) the variation in sensitivity among the members of the human population, (2)
the uncertainty in extrapolating laboratory animal data to humans, (3) the uncertainty in
extrapolating from data obtained in a study that is of less-than-lifetime exposure, (4) the
uncertainty in using LOAEL data rather than NOAEL data, and (5) the inability of any single
study to adequately address all possible adverse outcomes in humans. The RfC Methods use 3
for the UF for interspecies extrapolation due to the incorporation of default dosimetric
adjustments.
Vapor - a term referring to a gas phase at a temperature below the critical temperature of the
substance where the same substance can also exist in the liquid or solid state. If the gas is in
contact with the liquid or solid phase, the two phases will be in a state of equilibrium. This report
is intended to consider those agents present as gaseous vapors at ambient temperatures.
Ventral - On or near the lower surface (of the nasal tract).
September 2009
iv
-------
ACRONYMS AND ABBREVIATIONS
AA
Acrylic Acid
Aald
Acetaldehyde
Acr
Acrolein
AEL
Adjusted exposure level
A/H
Animal/Human
BMD
Benchmark dose
BW
Body Weight
C
Gas concentration, variously subscripted (mg/cm3)
r
^air
Concentration of toxicant in inspired air
CFD
Computational Fluid Dynamics
Ci
Inhaled concentration (mg/cm3)
Cex
Exhaled concentration (mg/cm3)
D
Liquid diffusivity (cm2/min)
DA
Diacetyl
DAF
Dosimetric Adjustment Factor
DL
Dorsal Lateral
DM
Dorsal Medial
DMS
Dimethyl sulfate
EA
Ethyl aery late
ET
Extrathoracic respiratory region
f
Breathing frequency
FA
formaldehyde
h2s
Hydrogen sulfide
Hb/g
Blood:gas (air) partition coefficient (unitless)
Ht/b
Tissue:blood partition coefficient (unitless)
Ht/g
Surface-liquid/tissue:gas (air) partition coefficient (unitless)
HEC
Human equivalent concentration
IRIS
Integrated Risk Information System
Kg
Overall MTC, mass transport coefficient (cm/min)
Kgc
Overall mass transport coefficient from gas phase
Kgm
Permeability coefficient
kg
MTC, mass transport coefficient in the gas phase (cm/min)
k i
Mass transport coefficient in the surface-liquid/tissue phase (cm/min)
kmc
Mass transport coefficient in the mucus phase (cm/min)
I
Thickness of the liquid-tissue layer
tmuc
Thickness of the mucus phase diffusion layer
LOAEL
Lowest-Ob served-Adverse-Effect Level
MM
Middle Medial
MTC
Mass Transfer Coefficient
N
Overall transport or flux (mg/cm2-min)
Ng
Flux through the air phase (mg/cm2-min)
Ni
Flux through the surface liquid-tissue phase (mg/cm2-min)
NP
Nasophaynx
OE
Olfactory Epithelium
P
Surface liquid/tissue:air partition coefficient (unitless)
September 2009
v
-------
PBPK
Physiologically Based Pharmacokinetic (model)
PO
Propylene oxide
POD
Point Of Departure
POE
Point of Entry
PU
Pulmonary respiratory tract region
PV
Posterior nasal vestibule
NOAEL
No-Observed-Adverse-Effect Level
RDDR
Regional deposited dose ratio used in derivation of an HEC for particles
RE
Respiratory Epithelium
RGD
Regional gas dose (mg/cm2-min)
RGDRet
Regional gas dose ratio (animal to human) for the extrathoracic region
RGDRpu
Regional gas dose ratio (animal to human) for the pulmonary region
RGDRtb
Regional gas dose ratio (animal to human) for the tracheobronchial region
RfC
Reference concentration
RT
Respiratory tract
SA
Surface area of unspecified respiratory region (cm2)
SAet
Surface area of the extrathoracic region (cm2)
SATb
Surface area of the tracheobronchial region (cm2)
SApu
Surface area of the pulmonary region (cm2)
TB
Tracheobronchial respiratory tract region
TH
Thoracic respiratory tract region
UF
Uncertainty factor(s)
URT
Upper Respiratory Tract
VA
Vinyl acetate
VL
Ventral Lateral
VM
Ventral Medial
Ve
Minute ventilation (L/min)
September 2009
vi
-------
EXECUTIVE SUMMARY
The purpose of this report is to briefly review advances in the state-of-the-science for
inhalation gas dosimetry related to the upper respiratory tract (URT) or nasal tract region. The
primary results from this review include empirical information related to the assumptions
underlying the current methodology used by EPA, as well as how the state of the science and
advanced dosimetry modeling techniques inform these assumptions. A principal conclusion from
this review that may impact EPA's future inhalation dosimetry methods for the URT is that
internal dose equivalency for rats and humans is achieved through similar external exposure
concentrations. This evaluation of internal dose equivalency relates to EPA's methods for
interspecies extrapolation and not to potential differences in dosimetry among individuals in the
human population.
This report reviews advances in science on gas dosimetry in the URT and will serve as a
source document to explore revisions to these aspects of Agency's 1994 RfCMethods (Methods
for Derivation of Inhalation Reference Concentrations and Application of Inhalation Dosimetry;
EPA/600/8-90/066F, October 1994). Future reports will review advances on gas dosimetry for
other regions of the respiratory tract. This report is intended neither to displace the current
procedures for gas dosimetry in RfC Methods nor to reflect on use of these procedures in
assessments currently in IRIS.
Comparative (animal to human) dosimetry is critical to all inhalation assessment
activities that relate effects observed in animals to humans. The basic principle involved in
comparative dosimetry is the determination of the internal dose to the target-tissue. This
principle, in turn, is founded on the fundamentals of risk, as is stated by the NRC in its 1994
publication "Science and Judgment in Risk Assessment" and discussed further in its 2009
publication "Science and Decisions: Advancing Risk Assessment":
"... the target-site dose is the ultimate determinant of risk... " .
The goal of comparative inhalation dosimetry is to characterize the steps leading from
(1) estimation of the internal target-site dose in an animal resulting from a given external air
concentration followed by (2) estimation of the external air concentration to which humans
would be exposed to attain that same internal target-site dose. The external concentration of a
human exposure scenario producing the equivalent internal target tissue dose is termed a human
equivalent concentration (HEC) in the RfC Methods.
For gases producing effects in the URT, the approximation of the internal target-site dose
from the external exposure concentration presented in the 1994 RfC Methods uses species-
September 2009
vii
-------
specific overall minute ventilation (VE) and the overall surface areas (SA) of theURT epithelium
or extrathoracic (ET) region. In the default approach to gas dosimetry, the RfC Methods uses
ratios of these measures as a dosimetric adjustment factor (DAF) that is then applied to the
animal external exposure concentration to arrive at an estimate of a HEC. The main assumptions
underlying procedures on gas dosimetry currently in use by the Agency follow from the
application of overall Ve/SA relationships assuming uniformity of airflow and uniformity of
deposition on surfaces. Chapter 3 of this report is a compendium of theory, observations,
simulations, and results that indicate extensive nonuniformity associated with these measures.
Information presented and discussed regarding these assumptions includes (1) empirical results
of airflow in the URT using casts of animals and humans, (2) observations on the highly
divergent geometry and anatomy of the URT between animals and humans, (3) advanced airflow
simulation methods employed within and between species, (4) advanced simulations of tissue
kinetics in URT tissues, and (5) correlation analysis of lesions and airflow simulations. An
extensive section is presented on state of the science estimations of target-tissue dose based on
the quantitatively linked airflow and tissue kinetic models. Overall, these advances both inform
the original assumptions on airflow and gas deposition of the RfC Methods and present solutions
to accommodate nonuniformity. A primary outcome is that for gas deposition in the URT the
internal target tissue dose equivalency between humans and rats is achieved through equivalency
at the level of the externally applied concentration, i.e., for both rats and humans, the same
external air concentration leads to the similar internal target-site dose to the URT.
In 2005, EPA consulted with a panel of inhalation toxicology experts external to the
Agency to review a range of topics that included the state-of-the-science for gas dosimetry and
default procedures advocated in the RfC Methods including the gas categorization scheme. The
results and recommendations from the 2005 consultation form the basis for the contents of this
report. Chapter 1 provides an introduction to this report and a brief history on the practice of
inhalation dosimetry. Chapter 2 provides a brief, focused review of the default comparative
inhalation dosimetry procedures as they currently exist in the RfC Methods. Chapter 3 of this
report summarizes the scientific advances in gas dosimetry for the URT region of both laboratory
animal species and humans. Chapter 4 provides a listing of source references identified by the
panel of experts as part of their consult and review.
Also featured in this report is a possible scheme for characterizing gases that differs
fundamentally from the current RfC Methods. The RfC Methods current gas scheme relates
physicochemical properties of gases to a category; this category is then related to the observed
toxicity, including that of the target tissue. The scheme featured in this report provides a direct
and simplified descriptor approach for characterizing gases that relates the properties of the gas
directly to the site of the observed toxicity without an intervening category.
September 2009
viii
-------
1. INTRODUCTION AND PURPOSE
1.1. PURPOSE OF THIS REPORT
Methods for Derivation of Inhalation Reference Concentrations and Application of
Inhalation Dosimetry (U.S. EPA, 1994), referred to in this report as the RfCMethods, was made
publically available in 1994.
The RfC Methods was drafted to advance dosimetry in general and for animal-to-human
extrapolation. RfC Methods characterized the exposure-dose-response continuum leading to a
target-tissue dose and provided a complete spectrum of interspecies dosimetry procedures for
gases and particles based on target-tissue dose. Furthermore, it addressed broad areas of
dosimetry. In cases where empirical observations had not yet sufficiently characterized theory,
the RfC Methods provided alternatives based on reasonable scientific theory.
Since 1994, significant advancements have occurred throughout risk assessment sciences;
in particular, interspecies comparative dosimetry of gases eliciting effects in the upper
respiratory tract (URT). Since many of these advancements impact core components of the RfC
Methods a need was recognized to assess the state-of-the-science in this area. Based upon that
need, an expert panel was assembled in 2005 and tasked with reviewing the state-of-the-science
of inhalation dosimetry.
The purpose of this report is to evaluate new science developments in the area of gas
dosimetry and evaluate how these advancements may support future revisions to the RfC
Methods used by EPA in developing RfCs. Specifically, this report summarizes the findings
and recommendations of an expert panel (2005) on interspecies gas dosimetry for portal of entry
effects in the upper respiratory tract. Future reports will review advances in gas dosimetry for
other regions of the respiratory tract. This report relates to methods for interspecies
extrapolation and not to differences in dosimetry among individuals in the human population
(intraspecies variability).
This report is intended as a source document for potential revisions to gas dosimetry
procedures in the RfC Methods. However, it is not intended to displace the current procedures
for gas dosimetry in RfC Methods or to reflect on the use of these procedures in assessments
currently in IRIS.
1.2. A PERSPECTIVE ON INHALATION DOSIMETRY
The 1994 RfC Methods document considers nearly every aspect of dose-response
assessment for inhaled toxicants. This state-of-the-science report is more limited in scope and
focuses on procedures and issues surrounding interspecies comparative dosimetry of gases,
September 2009
1-1
-------
particularly those exhibiting portal-of-entry effects in the URT. Also, it expands on the methods
for estimating the dose delivered to target tissues; the basis of the interspecies comparative
dosimetry established in the RfCMethods.
1.2.1. Historical Inhalation Dosimetry
The 1994 RfC Methods represented a fundamental shift in the approach for inhalation
dosimetry. Therefore, the historical practices to which the RfC Methods offered an alternative
need to be discussed briefly for the reader to understand the basis for the default procedures that
were put forth in the RfC Methods for interspecies extrapolation. These default procedures are a
primary focus of this report.
1.2.1.1. Total Dose from Concentration Inhaled
Perhaps the most simplistic approach for consideration of delivered dose from inhaling a
toxicant is to calculate the mass of the agent in the inspired air and consider that exposure as total
dose,
where VE is minute ventilation expressed here as total volume of air inspired in a day, either for
laboratory test animals or for humans, and Cair is the concentration of the toxicant in the inspired
air. This approach requires multiple assumptions, the most critical being those related to
toxicokinetics including amount absorbed, distribution, metabolism, and excretion
characteristics. A brief description of these assumptions and limitations follows in this report;
however, a complete discussion of the limitations of this approach is not within the scope of this
report.
1.2.1.2. Normalization of Dose
Yet another major issue with this simplistic approach relates to how the total dose could
be normalized for purposes of dose-response and interspecies extrapolation. Body weight is
often used for this purpose, such that the total dose is expressed in units of body weight. This
normalization procedure results in units typically encountered in oral dosing studies despite the
fact that the exposures occurred via inhalation, i.e.,
Nevertheless, this normalization procedure results in units of dose that are the same between
species and that are scaled to the ventilatory volumes of the different species, providing a
common starting point for interspecies extrapolation.
Dose (mg/day) = Ve (m3/day) x Cair (mg/m3)
Equation 1
Normalized Dose (mg/kg-day) = Dose (mg/day)/BW (kg)
Equation 2
September 2009
1-2
-------
Normalization of an inhaled dose using body weight has been used for consideration of
effects that occurred anywhere in the breathing individual, both systemic and portal-of-entry, i.e.
the respiratory tract. For systemic effects from inhaled dose, (i.e., effects that occur elsewhere in
the body following systemic transport via the blood) this normalization procedure is based on
many toxicokinetic assumptions, and it is conceptually reasonable. Accordingly, inhaled
toxicants absorbed into the blood via the respiratory tract and circulated to systemic target tissues
of a given volume or weight resulting in a "dose" per mL of blood or mg of tissue, can be
conceptualized for any test species and for humans. Part of this conceptualization relies on the
consistent interspecies relationship of blood volume, organs and body weights.
On the other hand, exposures involving contact of inhaled materials with portal-of-entry
tissues (i.e. those within the respiratory tract) differ fundamentally from contact of materials with
tissues remote from the portal-of-entry. Both blood and organs are usually evaluated and
functionally described in terms of volume, either directly or indirectly (as weight). Respiratory
tract tissues, however, are most often evaluated and functionally described in terms of epithelial
surface area or tissue weight. Inhaled materials entrained in breathed air pass over and around
thereby exposing these respiratory tract tissue surfaces. Unlike the consistent defined
interspecies relationships for weights and volumes, the interspecies relationships for surface
areas are known to be divergent and not well defined. Thus, normalizing inhaled doses for
effects and exposures based on surface area is not as readily conceptualized for effects and
exposures based on weights and volumes. The considerable information that has accumulated on
the interspecies differences in breathing physiology and airway anatomy makes normalization
procedures for respiratory tract tissues increasingly complex.
It is from this historical perspective of interspecies inhalation dosimetry, where inhalation
exposures were first transformed from external air concentrations to units of oral dosing and then
normalized to body weight, that the advent of RfCMethods is perhaps most appropriately
understood and valued.
September 2009
1-3
-------
2. REVIEW OF THE 1994 RFC METHODS FOR PORTAL-OF-ENTRY GAS
DOSIMETRY
2.1. GAS CATEGORIZATIONS - GENERAL
Numerous model structures have been used for describing aspects of toxicant uptake,
including gases and particles, in the respiratory tract.
Common uptake modeling schemes are often founded on the physicochemical
characteristics of the gases to which they are applied. For the respiratory tract uptake of ozone, a
highly reactive and moderately water soluble gas, Miller et al. (1985) proposed a detailed,
distributed parameter model. Key elements incorporated into this convective-diffusion-chemical
reaction model include (1) anatomic dimensions of the airspace and tissue thickness, (2)
dispersion in the airspace, (3) reactivity in the liquid lining (mucus or surfactant) covering the
cells of the lower respiratory tract, and (4) lateral mass transport resistance from the airspace to
the blood (Overton et al., 1987). To be inclusive of highly reactive and highly soluble gases
(e.g., formaldehyde, hydrogen fluoride), modeling schemes have emphasized the requirement to
account for scrubbing of the gas from the airstream by the upper respiratory tract (Casanova
et al., 1991; Hanna et al., 1989; Morris and Smith, 1982; Morgan and Frank, 1977; Aharonson
et al., 1974). For nonreactive gases, modeling schemes often use the counterpart of these
descriptions and characteristics to provide for a category whereby little toxicant is absorbed in
the upper respiratory tract. Dahl (1990) categorized gases as stable, reactive, or metabolizable
based on their thermodynamic and kinetic properties. Various concepts of "dose" can be related
to these properties and the mechanism of action (e.g., macromolecular bound fraction as dose for
reactive gases versus inhaled dose for stable asphyxiants).
It should be noted, however, that these uptake schemes are frequently based on the
chemical-specific physicochemical characteristics (e.g., solubility and reactivity) of the subject
gases, and described in terms of a qualitative continuum (e.g., low, moderate, or high).
Therefore, any model scheme comprised of discrete categories has limited application to the
broad range of gases that exist and that the RfC methodology must evaluate.
2.1.1. The RfC Methods Gas Categorization Scheme
The three category gas scheme currently in RfC Methods was constructed based on
physicochemical characteristics as determinants of gas uptake as shown in Figure 2-1. A similar
scheme has been developed by the International Commission on Radiological Protection (1993).
The primary determinants used to construct such a conceptual framework and that direct
September 2009
2-1
-------
development of the default dosimetry scheme, i.e., reactivity and water solubility, are elaborated
upon in the following sections. The numerical gas categories are placed on this scheme relative
to their character of these determinants; Category 1 in the upper right hand corner corresponding
to high reactivity and high water solubility; Category 3 in the lower left hand corner
corresponding to low reactivity and low water solubility; and Category 2 occupying the area
intermediate to the other two categories. The legend of this figure also indicates the regions of
the respiratory tract where uptake is likely to occur for these gas categories. Category 1 gases
are indicated to be absorbed in the "extrathoracic" or ET region which corresponds generally to
the nasal cavity. Category 3 gases are indicated to be absorbed in the deeper pulmonary region,
distal to the "ET" region, whereas Category 2 gases are indicated to be absorbed throughout the
entire respiratory tract. Some example gases for each category are provided in Sections 2.1.1.1 -
2.1.1.3. Detail describing the regions of the respiratory tract is included in Section 2.2.1.1.1 of
this report.
n
D
O
if)
w
O
•*->
A3
5
Reactivity ~
Gas Category Scheme Location
Category 1: Do not penetrate to blood ¦ Extrathoracic absorption
(e.g., highly water soluble/ ® Entire tract absorption
rapidly reactive) ~ Predominantly pulmonary
Category 2: Water soluble/Blood absorption
accumulation
Category 3: Water insoluble/
Perfusion limited
Source: U.S. EPA (1994).
Figure 2-1. Gas categorization scheme based on water solubility and reactivity as
major determinants of gas uptake. Reactivity is defined to include both the propensity for
dissociation as well as the ability to serve as a substrate for metabolism in the respiratory
tract. Definitive characteristics of each category and anticipated location (region) for
respiratory tract uptake are shown.
September 2009 2-2
-------
The two categories of gases with the greatest potential for respiratory effects are
Category 1 gases that are highly water soluble and/or rapidly irreversibly reactive, and Category
2 water soluble gases that may also be rapidly reversibly reactive or moderately to slowly
irreversibly metabolized in respiratory tract tissue. The definition of reactivity includes both the
propensity for dissociation as well as the ability to serve as a substrate for metabolism in the
respiratory tract. The scheme does not apply to inert gases that exert their effect by reversible
"physical" interactions of gas molecules with biomolecules.
The objective of the default RfC modeling approach is to describe the effective dose to
the major regions of the respiratory tract by addressing the absorption or "scrubbing" of a
relatively water soluble and/or reactive gas from the inspired airstream as it travels down and
through the respiratory tract from the nasal cavity to the pulmonary region. That is, the dose to
the peripheral regions is influenced by the dose to the region from which the airstream has just
exited.
The appropriateness of assessing proximal to distal (i.e., nasal cavity to the pulmonary
region) dose representative of the scrubbing (uptake) pattern is supported by the proximal to
distal progression pattern of respiratory tract toxicity with increasing concentration that is
observed with many chemicals. Exposure to low concentrations of highly water soluble and/or
irreversibly reactive gases primarily produces effects in the most anterior region of the nasal
tract. At higher concentrations, more severe effects occur in the ET region and toxicity is also
observed to progress to the peripheral regions. The severity of toxicity also progresses distally
with increasing exposure concentrations. Although the default gas models of the RfC Methods
do not describe detailed respiratory tract uptake to the level of local airflow distribution that
actually exists in the tissues of the respiratory tract (e.g., respiratory versus olfactory epithelium),
they do provide a description of the scrubbing of the chemical from the inhaled airstream on a
more general regional scale.
2.1.1.1. "Category 1" Gases
A defining characteristic for what the RfC Methods describes as Category 1 gases is that
they are so highly water soluble and/or rapidly irreversibly reactive in the surface-liquid/tissue of
the respiratory tract that they do not develop a significant backpressure (i.e., reversal in the
concentration gradient at the gas-liquid interface) from the surface-liquid tissue phase during
exhalation. Also, Category 1 gases do not significantly accumulate in the blood which would
reduce the concentration differential between tissue and blood and, hence, reduce the absorption
rate into the tissue. The default model structure is based on these characteristics. Examples of
gases classified as Category 1 are hydrogen fluoride, chlorine, formaldehyde, and volatile
organic acids and esters.
September 2009
2-3
-------
2.1.1.2. "Category 2" Gases
Gases in Category 2 are moderately water soluble and rapidly reversibly reactive or
moderately to slowly irreversibly metabolized in respiratory tract tissue. Ozone, sulfur dioxide,
xylene, propanol, and isoamyl alcohol are examples of Category 2 gases. As has been already
stated, these characteristics are continuous in nature such that the boundaries between categories
cannot be definitive. Thus, some compounds may appear to be defined by either Category 1 or
Category 2. Although sulfur dioxide is reversibly reactive, which would categorize it as a
Category 2 gas, it is also highly soluble suggesting it could be a Category 1 gas. Similarly,
ozone is highly reactive yet only moderately water soluble. More explicit delineation of
categories will be made upon review of the empirical data and the predictability of the model
gases that may appear to fit more than one category.
Because Category 2 gases are not as reactive in the respiratory tract tissue as Category 1
gases, they have the potential for significant accumulation in the blood and thus have a higher
potential for both respiratory and remote toxicity. Thus, the model structure used to describe
uptake of these gases is a hybrid of that for Category 1 and Category 3. The process of
estimating the extent and pattern of uptake of such gases through the respiratory tract and into
the systemic circulation is highly complex. PBPK models, with their capacity to use flows and
partitioning to allow estimation of both absorption on inhalation and the extent of desorption
during exhalation, are necessary to evaluate the steady-state blood concentration of such gases.
The derivation of such model structures, along with their reduction to forms with a minimal
number of parameters, are described in detail in Appendix I of the RfC Methods but are not a
principal part of this report.
2.1.1.3. "Category 3" Gases
Category 3 gases occupy the opposite pole of Category 1 gases as they are relatively
water insoluble and unreactive in surface liquid and tissue. Although exposure would occur
throughout the entire respiratory tract with Category 3 gases, uptake is predominantly in the
pulmonary region with relatively little uptake (or dose) occurring at the more anterior regions of
the respiratory tract. Styrene is an example of a Category 3 gas. The site of toxicity of these
gases is generally remote to the respiratory tract and a compartmental approach can be used to
describe distribution to various systemic tissues. Thus, the default model for Category 3 gases is
similar in structure to the PBPK model used by Ramsey and Andersen (1984) to describe styrene
distribution. The model structure and the derivation of the default dosimetric adjustment based
on this model are described in detail in the RfC Methods but are not a principal part of this report.
September 2009
2-4
-------
2.2. CONCEPTUAL AND HISTORICAL BASIS FOR COMPARATIVE DOSIMETRY
OF INSPIRED GASES IN RfCMETHODS- MINUTE VENTILATION/SURFACE
AREA OF RESPIRATORY TRACT (VE/SART)
The 1994 RfC Methods document is broad in its coverage and consideration of nearly
every aspect of dose-response assessment for inhaled toxicants. This status report is focused on
procedures and issues surrounding interspecies comparative dosimetry of gases, particularly
those exhibiting portal-of-entry effects. The basis of the interspecies comparative dosimetry,
posited by the RfC Methods and expanded upon in this report, is the estimate of delivered dose to
target tissues.
The RfC Methods presents an in depth consideration of what was known about
interspecies dosimetry for inspired gases at the time of publication with expansive commentary
on the fundamental underlying determinants of dosimetry. These included species differences in
anatomical and physiological characteristics of the respiratory tracts, the wide range of
physicochemical properties associated with inhaled chemicals, the diversity of cell types that
may be affected throughout the respiratory tract, as well as the many mechanistic and metabolic
differences, all combining to make characterization of dosimetry particularly complex.
This section briefly reviews and summarizes knowledge of several of the most critical
determinants of inhalation dosimetry that define and control the inhaled gas dose as well as
presenting the underlying basis for the RfC Methods interspecies normalization of the gas dose.
This review provides critical insight into how the RfC Methods default approach for dosimetry of
gases was developed. It is this insight that will serve the reader as the informational backdrop
against which scientific and technical advances described in Chapter 3 have progressed.
It is from this perspective of interspecies inhalation dosimetry established by the RfC
Methods, in conjunction with the historical perspective given in Chapter 1, that the advent of the
RfC Methods is most appropriately understood and valued.
2.2.1. Factors Controlling Comparative Inhaled Dose
RfC Methods states clearly that the various species used in inhalation toxicology studies
do not receive identical doses in comparable respiratory tract regions when exposed to the same
external gas concentration. The health endpoint or response from a gas, therefore, may be more
directly related to the quantitative pattern of target tissue dose which is influenced by such
factors as volumetric apportionment of air within the respiratory tract than to the external
exposure concentration. Such regional deposition and uptake patterns would likely determine
not only the initial doses to the respiratory tract tissues but also the specific pathways and rates
by which the inhaled agents are cleared and redistributed.
September 2009
2-5
-------
The two major factors controlling the deposition and uptake pattern are: (1) respiratory
anatomy and physiology and (2) the physicochemical characteristics of the inhaled toxicant.
Disposition is the more general term that encompasses the processes of deposition, absorption,
distribution, metabolism, and elimination. For gases, initial deposition is used in reference to
contact with the respiratory tract surface that precedes actual absorption. Mechanisms important
for gas deposition include convection, diffusion, chemical reaction (including metabolism),
dissolution, and perfusion. Detailed consideration of these mechanisms is beyond the scope of
this discussion.
Isolation of individual factors is probably necessary for initial understanding, but further
dissection of factors that control inhaled dose into discrete topic discussions may mask the
dynamic nature of the intact respiratory system. For example, although deposition in a particular
respiratory region will be discussed separately from the clearance mechanisms for that region,
retention (the actual amount of inhaled agent found in the respiratory tract at any time) is
determined by the relative rates of deposition and clearance. Retention and the toxicological
properties of the inhaled agent are related to the magnitude of the pharmacologic, physiologic, or
pathologic response. Therefore, although the deposition, clearance mechanisms, and
physicochemical properties of the agent are described in distinct sections, assessment of the
overall toxicity requires integration of the various factors.
Thus, any description of the continuum defined by exposure, dose, and response requires
integration of quantitative knowledge of determinants of chemical disposition, toxicant-target
interactions, and tissue responses into an overall model of pathogenesis. Among other things,
this process would involve determining the dose delivered to the target organ of various species
as well as determining the sensitivity of the target organ to that dose. Once such aspects of
dosimetry have been established and species sensitivity has been accounted for, the effective
chemical concentration in laboratory animals can be quantitatively related to dose responses in
humans. Models employed to perform this interspecies extrapolation would incorporate
parameters such as species-specific anatomical and ventilatory differences, metabolic processes,
as well as the physicochemical properties of the pollutant and should be based upon the
physiological factors that govern transport and removal of the pollutant.
2.2.1.1. Comparative Respiratory Anatomy and Physiology
As introduced earlier, the respiratory systems of humans and various experimental
animals, especially rodents which are the most frequently studied experimental species, differ
markedly in numerous quantitative and qualitative aspects of anatomy and physiology. These
differences affect critical aspects such as air flow patterns in the respiratory tract and thus
deposition of an inhaled agent, as well as retention of that agent in the system. These differences
September 2009
2-6
-------
in anatomy and physiology will be discussed according to respiratory regions and branching
patterns, clearance mechanisms, and cell types. Clearance mechanisms used here include
processes such as the mucociliary escalator, solubilization in various compartments, uptake, and
metabolism.
2.2.1.1.1. Regions of the Respiratory Tract Common among Species
The respiratory tract in both humans and experimental animals including the commonly
used murine species (i.e., rats and mice) can be divided into three similar regions on the basis of
structure and function: the extrathoracic region (ET) that extends from just posterior to the
external nares to just anterior to the trachea, the tracheobronchial region (TB) defined as the
trachea to the terminal bronchioles where proximal mucociliary transport begins, and the
pulmonary region (PU) including the terminal bronchioles and alveolar sacs. The thoracic (TH)
region or lower airways is defined as the TB and PU regions combined. The anatomic structures
included in each of these respiratory tract regions are listed in Table 2-1 and Figure 2-2 provides
a diagrammatic representation of these regions in humans.
Table 2-1. Respiratory tract regions
Region
Anatomic Structures
Other Terminology
Extrathoracic (ET)
Nose
Mouth
Nasopharynx
Oropharynx
Laryngopharynx
Larynx
Head airways region
Nasopharynx (NP)
Upper respiratory tract (URT)
Upper airways
Thoracic (TH)
Tracheobronchial
(TB)
Trachea
Bronchi
Bronchioles (to terminal
bronchioles)
Conducting airways
Pulmonary (PU)
Respiratory bronchioles
(not found in rodents)
Alveolar ducts and sacs
Alveoli
Gas exchange region
Alveolar region
Parenchyma
Source: Adapted from Phalen et al. (1988).
September 2009
2-7
-------
Extrathoracic
Region
Tracheobronchial
Region
Pulmonary
Region
Source: U.S. EPA (1994).
Figure 2-2. Diagrammatic representation of the three respiratory tract regions designated
in humans.
These interspecies similarities occur only at this very general level of organization, as
analysis at any more refined level begins to reveal marked differences. For example, within the
ET region, murine laboratory species have much more extensive and convoluted nasal turbinate
systems than do humans. Functionally, murine species are nasal breathers, whereas humans
breathe oronasally. Also, leading into the TB region, the length of the nasopharynx in relation to
the entire length of the nasal passage also differs between species. In addition, humans have
highly symmetric bifurcations whereas murine species have a highly asymmetric pattern
featuring secondary subdominant or monopodial branching (Figure 2-3). In the PU regions, the
September 2009
2-8
-------
branching patterns of airway ducts between humans and murine species are also markedly
different.
B
Source: Adapted from Lippman and Schlesinger (1984)
Figure 2-3. A representative model of symmetric (A) and asymmetric (B) airway
branching patterns occurring in humans and murine laboratory animals species,
respectively. The numbers represent major airway daughters.
September 2009
2-9
-------
These differences, especially the greater degree of nasal passage complexity in rodents
and the comparatively high ventilatory rates, are likely to result in greater deposition in the ET
region of rodents compared to humans; however, data to support this assertion were limited at
the time the RfC Methods was developed. It is important to note that the extent of upper
respiratory tract removal affects the amount of gas passing to the distal respiratory tract.
Airway size (length and diameter) and branching pattern affect the aerodynamics of the
respiratory system in the following ways:
• The airway diameter affects the aerodynamics and the distance from the agent
molecule or particle to the airway surface.
• The cross-sectional area of the airway determines the velocity for a given
volumetric flow.
• Airway length, airway diameter, and branching pattern variations affect the
mixing between tidal and reserve air.
Differences in airway sizes and branching among species therefore may result in
significantly different patterns of deposition for gases. For example, differences in airway size
and branching pattern are confirmed to be a major source of interspecies differences in inhaled
dose for the TB region. As mentioned above, relative to humans, experimental animals tend to
have tracheas that are much longer in relation to their diameter. A more complete listing and
analysis of comparative airway anatomy differences between humans and murine species is
available in the RfC Methods.
2.2.1.1.2. Surface Areas (SA) of Common Respiratory Regions
The existence of these general functional regions, the ET, TB and the PU, within the
respiratory tracts of humans and murine laboratory species has been thoroughly documented. It
is through, within, around and over these regions that inspiratory and expiratory air flows. When
recast in terms of an agent that may be entrained or present in the breathed air, the implications
become clear and evident that exposure to an agent may occur in this manner to tissues
comprising these respiratory tract regions.
Perhaps in partial recognition of this relationship, considerable effort in the scientific
community has been expended on estimating the surface areas (SA) for these respiratory tract
regions for both humans and a number of the common laboratory test species. Some accepted
values for the SA of these various regions are given in the RfC Methods and reproduced below,
complete with sources listed in the RfC Methods in Table 2-2.
September 2009
2-10
-------
Table 2-2. Default surface area values for respiratory effects
ETa (cm2)
Source
TBa (cm2)
Source
PUa (cm2)
Source
Human
200.0
Guilmette et al.
(1989)
3,200.0
Mercer et al.
(1994a)
540,000
Mercer et al.
(1994b)
Mouse
3.0
Gross et al.
(1982)
3.5
Mercer et al.
(1994a)
500
Geelhaar and Weibel
(1971), Mercer et al.
(1994b)
Hamster
14.0b
20.0
Yu and Xu
(1987)
3,000
Lechner (1978)
Rat
15.0°
Gross et al.
(1982)
22.5
Mercer et al.
(1994a)
3,400
Mercer et al.
(1994b)
Guinea
Pig
30.0
Schreider and
Hutchens
(1980)
200.0
Schreider
and
Hutchens
(1980)
9,000
Tenney and
Remmers (1963)
Rabbit
30.0
Kliment (1973)
300.0
Kliment
(1973)
59,000
Gehretal. (1981)
Source: U.S. EPA (1994).
aET = Extrathoracic; TB = Tracheobronchial; PU = Pulmonary.
bNo measurements of hamster ET surface area were found in the literature. This value is estimated based on
similarity of the other regional surface areas to the rat.
Additional unpublished measurements of the surface area beyond the ethmoid turbinates are included.
Aspects important to this report should be noted from the values in Table 2-2. The first
being that the PU region has the greatest percentage of SA over the respiratory tract, being
hundreds of times larger than ET and TB combined. At low concentrations of those gases
considered highly soluble and reactive and which would undergo significant clearance in both
the ET and TB regions, the residual gas reaching the PU would be dispersed over this large SA
of the PU region. A second point is the converse of the first in that for highly soluble and
reactive gases, the SA of both the ET and TB regions receives a disproportionate amount of gas
relative to the PU region.
Another aspect to be noted from the table is that a single SA figure is given for each
region. It is important to note that SA varies with age, and the SA of the PU region is also
affected by disease. Although it is established that numerous cell types and tissues exist in these
areas, only a single SA is given. For example, perhaps the most problematic region in this
respect is the ET region, where no fewer than four different types of epithelial and submucosal
tissues exist. Some tissues, such as the olfactory epithelium, possess multiple cell types, some of
which are specific targets for certain toxicants. Thus the SAs listed in this table, although
appearing as a single uniform number, describe areas that have extreme cellular and functional
diversity. The review of Harkema et al. (2006) provides an in-depth discussion of the various
September 2009
2-11
-------
epithelial cell types as well as their toxicological and physiological significance in a variety of
species including humans.
2.2.1.1.3. Comparative Respiratory Minute Ventilations (VE)
The means by which exposures to any SA would occur from agents in the air is most
likely and logically via the agent concentration present within the inspired and expired air, the
minute ventilation, VE.
The RfC Methods provides procedures and parameters for calculating typical minute
ventilations, VE, expressed as total volume of air inspired in a day, both for laboratory test
animals and for humans. The default ventilation values for minute ventilation [VE = tidal
volume (VT) x breathing frequency (f)] are based on accepted body weight allometric scaling
equations provided in the literature.
The basis of interspecies allometric scaling is to account for disproportionalities between
species. It should be kept in mind that these disproportionalities manifest because smaller species
have proportionally greater VE per unit body weight than larger species. VE also varies with age,
activity, and disease.
For purposes of interspecies scaling, the RfC Methods specifies a default body weight for
the human of 70 kg, and a corresponding VE of 13.8 L/min or 20 m3/day.
2.3. NORMALIZATION OF INHALED CONCENTRATION TO SURFACE AREA OF
RESPIRATORY TRACT REGIONS
2.3.1. Dose-Response in Respiratory Tract Tissues is Based on External Exposure
Concentration
A central consistent observation from inhalation exposure has been the capacity of the
external exposure concentration to establish and explain the response in the respiratory tract.
Pathology in laboratory animals from inhalation exposure to a variety of agents has characterized
responses in the ET, TB and PU regions of the respiratory tract. Response characterization based
on external exposure concentration has been thorough, extending from the basis of structure and
cell biology, from various mechanisms of toxic action, as well as response to injury. Some
studies and responses to inhalation exposures have been characterized by inclusion of other
variables, e.g. that the response may follow a concentration x duration (Cx T) relationship. Even
with advanced dose-response modeling approaches for inhalation exposures, the inception of any
analysis is the external exposure concentration.
Transformation of the external exposure concentration in air from typical units of ppm or
mass/unit air to other units, such as total mg/day, does not provide a metric with a direct
September 2009
2-12
-------
relationship to response, especially when the response being characterized is a portal-of-entry
response involving one of the surfaces in the respiratory tract. For extremely reactive gases,
deposition may only occur in the first few centimeters of the ET, thus invalidating even those
metrics based on the overall surface area of the ET.
2.3.2. Normalization of External Exposure Concentration to Surface Areas
In their investigation designed to resolve marked differences in tumor response to high
concentrations of inhaled formaldehyde observed in rats (high response) versus mice (low
response), Chang and coworkers (1983) explored several approaches to evaluate whether this
difference in response was due to a smaller dose in mice. The authors clearly demonstrated that
mice were better able than rats to reduce their VE via decreases in respiratory rate, thus logically
resulting in a comparatively smaller dose to mice than rats. When dose was expressed in
conventional terms, |ig/m in/kg body weight, mice were predicted to receive nearly twice the
dose of rats. However, when dose was expressed in terms of target tissue (i.e., nasal cavity)
surface area, mice were predicted to receive approximately one half the dose of rats. This
approach, "normalization" of dose per surface area, offered a satisfactory explanation to account
for the observed differences in tumor response. A conclusion of this work was that normalizing
the dosimetry to nasal surface area could lead to better understanding of species differences in
nasal toxicity. The basis of such an approach was theoretical as the assumptions being made
were that formaldehyde was completely absorbed and uniformly distributed over the entire
surface area. These same theoretical assumptions will be examined in the following sections.
These conceptual considerations for inhaled dose coupled with observations of Chang
and coworkers (1983) that target tissue surface area was explanatory of responses logically led to
the construct of the basic dosimetry scheme put forth by the RfC Methods. In this scheme,
inhaled dose to the respiratory tract is based on species -specific relationships of minute
ventilation and surface areas associated with the target regions of the respiratory tract:
Ve/SAet?tb or pu Equation 3
It is the consideration and exploration of the assumptions underlying this normalization
that will provide the impetus for the next steps and advancements for the RfC Methods. As will
be apparent in Chapter 3 of this report, it is the clarification of these assumptions following from
Ve /SA that will focus and direct advancement of inhalation dosimetry.
September 2009
2-13
-------
2.4. APPLICATION OF VE/SA IN CALCULATION OF I II I HUMAN EQUIVALENT
CONCENTRATION, HEC: THE DEFAULT APPROACH FOR INSPIRED GASES
Overall, this chapter has attempted to present in a concise manner the conceptual basis
for choice of the components and procedures used by the RfC Methods in performing
interspecies inhalation extrapolation for respiratory tract responses. Rather than rely on the
historical and problematic procedure involving transformation of inhaled concentration to total
dose, the RfC Methods relied on empirical evidence to validate the use of the external air
concentration. Instead of imposing the typical dose measures for systemic effects (e.g., mg/kg-
day) on effects occurring in portal-of-entry respiratory tract tissues, extensive information, not
presented here, was synthesized to support a more relevant and explanatory dose measure based
on normalization of target surface areas of the respiratory tract.
The overall goal of the RfC procedures is to estimate toxicokinetically equivalent doses
to target tissues in laboratory animals to those of humans. RfC Methods give application
procedures for these various components to produce an estimate of a human equivalent
concentration (HEC). RfC Methods elaborates upon and outlines practices and data requirements
for other procedures that are arranged in a hierarchy of approaches for estimations of human
equivalent concentrations for gases (see Table 2-3). The approach being discussed and analyzed
in this report is that of the "default" approach only, the most generalized and data limited
situation where information is limited to that discussed thus far in this report, principally species
VE and SA for the various regions of the respiratory tract As indicated in the hierarchy
scheme, the default approach would be bypassed when more sophisticated or chemical-specific
models are available, several of which will be explained further in this report (e.g. PBPK, CFD,
and CFD-PBPK hybrid).
Table 2-3. Hierarchy of model structures for dosimetry and interspecies extrapolation
Optimal model structure
Structure describes all significant mechanistic determinants of chemical disposition, toxicant-target
interaction, and tissue response
Uses chemical-specific and species-specific parameters
Dose metric described at level of detail commensurate to toxicity data
Default model structure
Limited or default description of mechanistic determinants of chemical disposition, toxicant-target
interaction, and tissue response
Uses categorical or default values for chemical and species parameters
Dose metric at generic level of detail
Source: U.S. EPA (1994)
September 2009
2-14
-------
2.4.1. The Dosimetric Adjustment Factor, DAF
The following equation may be generally applied to estimate an HEC. It demonstrates
the application of a correction factor to an external exposure air concentration encountered by the
laboratory animal:
(1) POD(adj) is the air concentration used as the "Point Of Departure" corresponding to
an effect level obtained from, for example, analysis of a laboratory animal inhalation
study. Examples of PODs include the NOAEL (no-observed-adverse-effect-level) or a
BMD (benchmark dose). The subscript "ADJ" refers to adjustments made to the POD for
considerations other than dosimetry (e.g., duration) and is not a focus of this report.
(2) DAFr is a dosimetric adjustment factor for respiratory tract region, r (ET, TB, PU, or
TOTAL). It represents a multiplicative factor used to adjust an observed exposure
concentration in a particular laboratory species to an exposure concentration for humans
that would be associated with the same delivered dose also termed the "regional gas
dose" or RGDr. For gases this factor is the "regional gas dose ratio" or RGDRr.
(3) The POD(hec) is the human equivalent external air concentration of the animal POD
that has been dosimetrically adjusted to a human equivalent concentration (HEC) through
application of the appropriate DAFr, in this case for gases, the RGDRr.
2.4.2. The DAF for Gases; the Regional Gas Dose Ratio, RGDRr
The equations to derive the regional gas dose ration (RGDRr) for different gas categories
and for the various regions of the respiratory tract (versus remote sites) are provided and
described further in RfCMethods. The principal focus of this report is the approach and
procedures used for gases having effects in the extrathoracic (ET) region of the respiratory tract.
The basic default equation given in RfC Methods used to calculate the RGDREt, i.e., the
DAFet, for gases incorporates basic determinants of inhaled dose — species-specific minute
ventilation and surface areas:
POD(hec) (mg/m3) = POD(ADj) (mg/m3) x DAFr
Equation 4
where:
RGDRet =
(VE / SAet)a
(VE / SAet)h
Equation 5
where:
VE = minute ventilation (L/min),
SAet = surface area of the extrathoracic region (cm2), and
September 2009
2-15
-------
A, H = subscripts denoting laboratory animal and human, respectively.
Basically, the RGDRr is used as the DAFEt in Equation 5 to dosimetrically adjust the
experimental POD (duration adjusted or not) to estimate an HEC POD (PODrec) as follows:
POD(hec) (mg/m3) = POD(Adj) (mg/m3) x RGDRr Equation 6
RGDRr can be seen to be equal to the ratio of the RGD in laboratory animal species to that of
humans (RGDr)A/(RGDr)H.
RGDRs have been calculated for several species for all regions of the respiratory tract,
including the ET, and are shown in Table 2-4.
Table 2-4. RGDRs and minute ventilation/surface area (Ve/SA) ratios of the ET, TB,
and PU regions of respiratory tract for various species and humans
Parameter
Hamster
Guinea Pig
Mouse
Rat
Monkey
Rabbit
Human
Vl:(nr7day):i
0.093
0.41
0.064
0.30
2.55
1.475
20
ET Region
SAETb(cm2)
14
30
3
15
63°
30
200
Ve/SAet
0.007
0.014
0.021
0.02
0.04
0.049
0.1
RGDRETd
0.07
0.14
0.21
0.20
0.4
0.5
-
TB Region
SATBb(cm2)
20
200
3.5
22.5
-
300
3200
\e/SAtb
0.005
0.002
0.018
0.019
-
0.005
0.006
RGDRTBd
0.8
0.33
3
3.2
-
0.8
-
PU Region
SAPUb (m2)
0.300
0.900
0.050
0.340
-
5.9
54
Qalv^SApu
0.22
0.32
0.9
0.91
-
0.18
0.26
RGDRPUd
0.8
1.2
3.5
3.5
-
0.67
-
a(male + female)/2 "mature" values from U.S. EPA (1988a, 1988b).
''U.S. Environmental Protection Agency (1994)
°Kepler et al. (1995)
dRGDR = Regional Gas Dose Ratio for the ET, TB or PU regions = (VE /SA)animal/(YE /SA)human
"Alveolar ventilation rate (Qalv) calculated as 0.7 x minute ventilation
2.4.3. Some Example HEC Calculations with Commentary
To make application of this dosimetric procedure clearer, a few examples using the
entries in Table 2-4 are given, each accompanied by some commentary.
September 2009
2-16
-------
Example 1 - In hamsters, effects in the tracheobronchial (TB) region (cellular
degeneration) are noted after exposure to a gas at a POD of 10 mg/m3 with no duration
adjustments to this concentration. Using Table 2-4 and Equation 7, the HEC would be calculated
as:
POD(hec) (mg/m3) = POD(Adj) (mg/m3) x RGDRTB Equation 7
8 mg/m3 = 10 (mg/m3) x 0.8
Commentary: Here the default HEC estimated for this extrapolation from the TB region
of the hamster is quite close in value to the animal POD, due to the appropriate RGDRtb being
close to unity at 0.8. This indicates that at typical ventilatory parameters, hamster and human
lung tissues are predicted to be about equally exposed at a common external concentration.
Example 2 - In mice, effects in the pulmonary region (chronic infiltrate) are noted after a
10 mg/m3 exposure to a reactive gas, duration adjusted to achieve a final POD concentration of
5 mg/m3. Using Table 2-4 and Equation 7 the HEC would be calculated as:
POD(hec) (mg/m3) = POD(Adj) (mg/m3) x RGDRPU Equation 8
17.5 mg/m3 = 5 (mg/m3) x 3.5
Commentary: The default HEC estimated for this extrapolation from the PU region of
the mouse is quite divergent from the adjusted animal POD, due to the appropriate RGDRPU
being considerably greater than unity at 3.5. This indicates that at typical ventilatory parameters
and breathing the same gas concentration, human lung tissues would be receiving considerably
less dose (3.5-fold less) than mouse lung tissues. Another view of the extrapolation is that
human lung tissues would be exposed to the same internal dose at 17.5 mg/m3 as mouse lung
tissues would be at 5 mg/m3.
Example 3 - In rats effects are noted in the nasal cavity, the ET region, (necrosis of the
respiratory epithelium) at an external gas concentration of 100 mg/m3. Since no exposure
duration or other adjustments are needed, 100 mg/m3 is the POD. Using Table 2-4 and Equation
7 the HEC would be calculated as:
POD(hec) (mg/m3) = POD(Adj) (mg/m3) x RGDRet Equation 9
20 mg/m3 = 100 (mg/m3) x 0.20
Commentary: The default HEC estimated for this extrapolation from the ET region of
the rat is also quite divergent from the adjusted animal POD, due to the appropriate RGDRet
September 2009
2-17
-------
being considerably smaller than unity at 0.20, which is opposite of the PU example above. This
indicates that at typical ventilatory parameters and breathing the same gas concentration, human
ET tissues would be receiving more dose (1/0.20 = 5-fold more) than rat ET tissues. Another
view of this extrapolation is that humans ET tissues would be exposed to the same internal dose
at 20 mg/m3 as rat ET tissues would be at 100 mg/m3.
2.5. ASSUMPTIONS AND LIMITATIONS IN W./SA
Although inhalation dosimetry based on the measure of "dose" estimated through
VE /S A|;T_ti>_ pu or Totai has been demonstrated to be more explanatory of inhaled dose and
responses in the respiratory tract and portal-of-entry effects than alternatives discussed in
Sections 1.2.1.1 and 1.2.1.2, such an approach is put forth on assumptions, either explicit or
implicit, whose existence, limitations and caveats need to be considered.
Perhaps the most debated aspect of this dosimetric approach concerns the outcome of the
example above using the ET region. As demonstrated, the default dosimetric adjustment for gas
exposures in this region estimates that human tissues receive a three- to fivefold higher exposure
than do rat tissues. One primary reason the VE/SAr relationship is debated is its marked
divergence from the closely related relationship of Ve /BW. Parallel construction of an RGDR
based on body weight instead of SAET would yield a value of 3 ([(rat VE/0.3 kg)/(human VE/70
kg)] = 0.6/0.2 = 3) indicating that human tissues would receive threefold less dose than rat
tissues.
With Ve/SA regarded as dose normalized to surface area, this increased ratio in humans
indicates that human ET tissues will receive a fivefold higher dose than those of, for example,
rats. This circumstance regarding interspecies differences is represented in Figure 2-4 which
illustrates the actual values used in application of VE/SAETto perform interspecies dosimetry
through derivation of an RGDR. The right hand side of this figure shows the overall outcome
when these differences are applied to interspecies extrapolation; the human ET tissues are
projected to always receive more of dose, up to fivefold more, than are rat ET tissues.
September 2009
2-18
-------
Species
/ ^VeT
animal
(rat)
#,1S L/min / 15 cm2 = 0 012
human !
13,8 L/min ' 200 cm2 0.069
(Ix dose)
*( Vg/ SAlT)a _
{ VE/ SAjj)^
x close)
= RGDKet= 0.2
Figure 2-4. Calculation and application of Ve/SAEt in the RfC Methods default approach
for interspecies dosimetric extrapolation through derivation of the regional gas dose ratio
for the extrathoracic region, RGDREt. The RGDR value of 0.2 may be seen to relate back
to the presumption that humans ET tissues receive the obverse of the RGDR in dose, i.e.
fivefold (1/0.2 =5). Also see Equation 5, above.
This divergent interspecies relationship is reflected in the application of VE/SA in actual
calculation of an HEC where the DAF (the RGDREt) applied to the animal point of departure
approximates the inverse of this ratio, i.e., 0.2-0.3 (i.e., 1/3-5). Thus, human equivalent human
exposures based on responses in the ET from rat results are adjusted downward by this fraction.
This adjustment lowers the overall estimate for an effect concentration in humans by a factor of
3 to 5 even before consideration of uncertainty factors.
2.5.1. The Ve/SA Assumptions
This specific species divergence in dosimetric adjustment has resulted in considerable
scientific debate. This debate has led to the identification and clarification of assumptions made,
either explicitly or implicitly, underlying the default dosimetric procedures for interspecies
extrapolation. It is these assumptions, the most critical and significant of which are listed below,
that provide a basis for further revisions and refinement of the dosimetric procedures used by the
Agency.
Assumption 1 - The flow of the gas at Ve through the respiratory tract region of interest is
uniform. As the units of Ve are L/min there exists no basis to consider nonuniformity of flow of
the gas present or entrained in the breathed air. At the time of the RfC Methods development,
nonuniformity was suspected although there was no substantial basis to quantitatively evaluate
its extent.
September 2009
2-19
-------
Assumption 2 - The SAs of the respiratory tract regions subtending the entrained gas
have to be treated as uniform and equivalent, i.e., the cell types, relative amount, and
distributions are equivalent in human and animal species. This assumption actually follows from
and is a corollary to assumption 1. Although considered valid for the general regions of the
respiratory tract, the available SAs incorporate no further refinement regarding tissue- or cell-
types within any region. It is essentially an averaged surface area that contains widely divergent
nonuniform cell types. Under the RfC Methods the most refinement that can be achieved with
Ve/SA is essentially limited to the SA term in the denominator. As discussed elsewhere
previously in this report and in the RfC Methods, such an assumption may be most problematic
for the ET, a region that may be considered the most anatomically complex, divergent, and
varied in tissue type of all regions in the respiratory tract.
Assumption 3 - The gas inhaled is uniformly and evenly distributed over the entire
surface of the respiratory tract region in question. A further assumption for gases with portal-of-
entry effects is that the deposition/uptake is complete or 100% in the region in question and is
the same in animals and humans as deposition/uptake information for humans is frequently
lacking. These assumptions are somewhat implicit but necessarily follow from the units of
Ve/SA and from the accompanying assumptions of uniform flow and uniform surface area.
Also, inspection of the Ve/SA relationship reveals that modulation of either Ve or SA would
directly influence the "intensity" or flux of the gas deposited to the SA. For example, increasing
Ve and decreasing the SA would increase the flux at the SA; decreasing Ve and increasing the SA
would instead decrease the flux at the SA. Perhaps the most obvious inconsistency of this
assumption is the empirically demonstrable proximal to distal, high to low concentration
response gradient known to occur for gases that produce respiratory tract lesions. Also, for gases
extensively scrubbed from the upper airways (the ET region) such as formaldehyde, the
assumption of 100% uptake/absorption may be valid. However, this assumption cannot be valid
for a great many other gases. These issues as well as many others critical to the practice of
inhalation interspecies dosimetry as given in the historical RfC Methods are evaluated in the
following chapter.
September 2009
2-20
-------
3. SCIENCE DEVELOPMENTS IN PORTAL-OF-ENTRY GAS DOSIMETRY
3.1. INTRODUCTION
The preceding chapter examined and summarized the origins, underlying principles and
concepts, and demonstrated the application of the default procedure in the RfC Methods for
performing inhalation dosimetry of gases that exhibit portal-of-entry effects. This examination
focused on important assumptions made in the application of the basic element of default
dosimetry, i.e., VE/SA. These three basic assumptions are briefly restated and illustrated in
Figure 3-1.
lx dose
'/SA,
Human V./SA,
5x dose
1 - assumption of uniform even flow of gas present in VE through the respiratory tract region
2 - assumption of uniform even SAs of tissues in the respiratory tract region subtending VE
3 - assumption of uniform even deposition of gas present in VE over the entire surface of the respiratory tract
Figure 3-1. Representation of the assumptions of uniformity following from VE /SA
as applied to comparative gas dosimetry. Gas is entrained in the VEand impinges upon the
subtending epithelial surface area (SA) of the various regions (r) of respiratory tract. The
right side of this figure shows the arithmetic outcome of applying VE /SA to the ET regions
of rats and humans.
This figure illustrates a major outcome following from the interspecies extrapolation for
the ET region with these attendant Ve /SAEt assumptions. With the accompanying assumption
that all of the gas is absorbed in the region defined by SA, the approximate fivefold higher value
for Ve/SAet in humans compared to rats indicates that the surface of the human ET receives a
fivefold higher dose than the rat ET.
September 2009
3-1
-------
The intervening period from issuance of the RJC Methods from 1994 to date has seen
substantive advancements for dosimetry of gases in general and interspecies comparative gas
dosimetry in particular. Many of these advancements and investigations directly inform the
above three assumptions.
This chapter will present and discuss information that is representative of what has
become available concerning these assumptions since the 1994 RfC Methods was completed.
3.2. A MODIFIED GAS SCHEME: DESCRIPTORS VERSUS CATEGORIES
On the practical level a gas/vapor dosimetry scheme should be useful as an aid in
predicting the site of toxicity. This is especially applicable in describing the fate of inhaled
gases. Knowledge of two physicochemical properties, water solubility and reactivity, is useful in
predicting the site of uptake of gases/vapors in the respiratory tract tissue and/or absorption into
blood and the potential toxic actions in both inhalation portal-of-entry sites and in sites remote
from the inhalation portal-of-entry. These properties have been repeatedly used as predictors of
the site of gas uptake and effects within the respiratory tract.
Although acknowledged as a continuum, the reactivity-solubility relationship is
nevertheless grouped by the RjC Methods into a numerical category: Category 1 for a highly
water soluble and reactive gas, Category 2 for a gas designated moderately water soluble and
moderately reactive, and Category 3 for a relatively insoluble and nonreactive gas. Use of this
scheme, conceptually represented and described above, yields a numerical gas category which in
turn leads to general principles and statements concerning likely sites and conditions of toxicity.
Medinsky and Bond (2001) put forth a descriptor scheme based on these same two
physicochemical properties whose application to gases would be fundamentally different, direct
and more useful. In this scheme, water soluble gases are defined as gases that readily dissolve in
the mucus lining of the upper respiratory tract followed by diffusion into the underlying
epithelial cells and, potentially, into the blood for systemic distribution. Generally, water-
insoluble gases penetrate the mucus lining more slowly and are transported to the lower
respiratory tract where they may be absorbed into the blood. For gases such as these, deposition
and uptake in the URT would be dependent upon other processes controlling uptake (e.g.
metabolism). The other principal determinant, reactivity, defined in this scheme as the tendency
of a gas to undergo chemical reaction, is simple to understand. However, reactivity at the level
of organization relating to tissue dosimetry is complex. In the distance from the airway to the
blood, reactive gases may undergo interactions with components in the air, mucus, and tissue.
These reactions lead to rapid and substantive decreases in the concentration of the reactive gas
across this distance. In fact, some gases have such high reactivity that very little of the gas itself
ever reaches the blood. This is in contrast to inert gases that do not react with respiratory tract
September 2009
3-2
-------
tissues or fluids such that the gas can ultimately diffuse through the various respiratory layers
with appreciable concentrations eventually reaching the blood for systemic distribution. Thus,
chemical reactivity of the gas controls its molecular interactions with respiratory tissues and
influences penetration of the vapors to the blood.
Although not explicitly discussed in this scheme, metabolism (i.e., catalyzing actions of
enzymes in the respiratory tract) is also considered a component of reactivity in characterizing
gas transport in the tissue and predicting site of gas uptake and effect. However, application of
the descriptor outlined by this scheme excludes metabolism in designating the parent gas as
"reactive". For example, dibasic ester and butadiene are both designated as "non-reactive" gases
because their toxicity is dependent upon their metabolite(s) (Medinsky and Bond, 2001). These
types of gases often produce site-specific toxicity based on the localization of their metabolic
enzymes.
Rather than assigning specific numerical categories to gases, these descriptors are placed
on a chart that represents reactivity and water solubility as continuous variables. This scheme,
along with the descriptors for the boundary conditions of the variables, is shown in Figure 3-2. It
is upon this scheme that information and data can be placed to gain insight regarding sites and
conditions of toxicity. Figure 3-2 also illustrates this application with the placement of
formaldehyde and its predominant toxicity, effects in the ET portal-of-entry region, in the upper
right portion of the chart. This sector of the chart corresponds to high reactivity and high water
solubility, characteristics that would predict gas disposition with resultant effects in the anterior
most portal-of-entry region, i.e., the nasal cavity or "ET" (for extrathoracic) region, as is noted
for formaldehyde. Butadiene, which produces effects in lung tissue, has low water solubility and
is non-reactive, and thus, is the obverse entry to formaldehyde in Figure 3-2. It is important to
note that this scheme provides examples of gases that fit these discrete descriptors, but that the
majority of gases may not fit one particular descriptor. However, examination of these examples
at the extremes should facilitate understanding of the behavior of other gases.
Utilization of such a scheme yields direct information about the nature and site of toxicity
that is based on the determinative variables of water solubility and reactivity. Direct information
on the nature and site of toxicity is crucial to formulating an approach to inhalation dosimetry.
This approach and scheme contrasts with the current RfCMethods scheme where the outcome is
a numerical category whose implications to dosimetry are proscribed, often irrespective of the
sites and conditions of the observed toxicity. The current RfC Methods scheme has resulted in
some confusion in that it is often applied as such and influences dosimetry by placing an
unintended emphasis on the numerical gas category as opposed to the site of toxicant effect.
September 2009
3-3
-------
WUter-lneoluble,
reactiv*
Ozone:
Nose end lungs
Water-aoluble,
reeeUvt
Formaldehyde:
Nose
Water-lnaolubia,
nonrtecUvt
Bubadene, benzene:
Ling
Water-aoluble,
nonreactive
Olbsdceeters
Nose
~
Water Solubility
Source: Adapted from Medinsky and Bond (2001).
Figure 3-2. A schematic representation of the physicochemical properties of reactivity and
water solubility overlaid with descriptors of their practical limits. Examples of specific
chemicals with their primary site of toxicity are also presented.
Such a scheme would have best applicability in situations where gas exposure levels are
relatively low and the effects observed are in the most sensitive target tissues. A higher
concentration and a broader range of responses would result in less applicability of this scheme.
3.3. FLOW-DYE CAST MODELS
Morgan and co-workers (1991) laid the foundation for studying airflow distribution
patterns in the ET region, the nasal tract. In an early work, these investigators prepared clear
acrylic casts from molds of the nasal airways (meatuses), including the larynx and proximal
trachea, of both F344 rats and Rhesus monkeys. Water flow was introduced through the casts in
an anterior to posterior direction using a unidirectional water-dye siphon system. Water-dyes
were introduced to visualize and study the dynamics of flow through this system at
physiologically relevant rates. Although consistent from cast to cast, the character of the
observed flows was not uniform. Rather, the dye-streamlines observed with this technique were
consistent in revealing complex flow patterns in both species with major inspiratory streamlines
September 2009
3-4
-------
observed consistently from cast to cast (Figure 3-3). In rats, major inspiratory streams were
present medially with a superior and inferior ventral stream, a middle stream and a dorsal stream
that bifurcated mid-way through the upper meatuses. Major lateral inspiratory streams were also
present and apportioned in this manner into ventral, middle and dorsal streams. The flow
patterns in monkeys differed significantly and were far more complex and varied than those in
rats. Inspired air flow in monkeys was divided into a number of major streams within the most
anterior portion of the nares, the vestibule. These patterns included a relatively disorganized
medial fan-shaped jet, ventral and dorsal vortices in addition to major streamlines in both ventral
and dorsal regions of the nasal tract. The flow streamlines also allowed for a qualitative
assessment of flow velocity within each species. In rats, the flows appeared similar for all major
streams except for dorsal streams through regions containing olfactory epithelium (OE), which
were slower. In contrast, there were clear velocity differences for monkeys. The fastest
velocities occurred in the anterior fan-jet, with slower flows in the principal passageways and a
markedly slower flow velocity in the dorsal medial meatus.
Consequences of these (and other similar) observations are significant on several levels.
First, the complex but generally consistent and orderly streamlines revealed by the cast method
show a sensitive dependence of nasal airflow patterns on nostril geometry throughout the ET
region. Second, all observations indicate overwhelmingly that flow into the nasal area, either
liquid or air, is in no way uniform but has discernable patterns that could only result in highly
nonuniform deposition onto surfaces (i.e., nasal epithelial surfaces). The consistent patterns of
both streamlines and estimates of widely varying flow velocity also give support to the proposal
that these elements both play a role in the distinct distribution of lesions known to be induced by
agents such as formaldehyde (discussed later).
September 2009
3-5
-------
Male F344 Rat
Dorsal Mcd-al
•;i • < ¦
Male Rhesus Monkey
Superior flo'SOl
Stream —.
manor uorfsm
Midde S ream
Medial AnlHriur
Jet
Ventral Anlenof
Vortex
Middle Lateral
Dorsal Lateral
Stream
SiT.-in
Dorsal Anterior
Vorte*
Superior Ventral Latera
Stream
Inferior Ventral Lateral
Stream
Ventral Lateral
Vortex
" Ventral Anterior
Vortex
Source: Morgan et al. (1991).
Figure 3-3. Inspiratory airflow patterns in upper respiratory tract of F344 rat and Rhesus
monkey. A = major medial streams; B = major lateral streams. Black and white arrows
depict high and low velocity airstreams, respectively. The external nostril is to the left.
On the other hand, use of this approach for dosimetric comparisons is limited. These
limitations include the accuracy, representativeness and resolution of the casting process itself
and the inability to quantitatively evaluate any of the flows and flow rates observed. These
limitations remained to be addressed with quantitative mathematical airflow models which are
presented and discussed in the following sections.
3.4. COMPUTER MODELING OF FLUID FLOW - INTRODUCTION TO CFD
Computer modeling of fluid flow is an established technology that has been applied
extensively in engineering. In general, this modeling approach, known as computational fluid
dynamics (CFD), allows for quantitative prediction of all variables of fluid flow (e.g., pressure
and velocity) based on the mathematical laws governing fluid behavior and, with proper
software, offers a three dimensional visualization of the predicted flow. Fluids are defined to be
inclusive of all substances that can flow and thus the process is applicable to both liquids and
September 2009
3-6
-------
gases. CFD and its predictions may therefore be applied to quantitatively evaluate airflow in the
airways of the respiratory tract.
Regardless of the fluid flow problem at hand, the same basic procedure to develop a CFD
model is followed (see Bailie et al. [2006] for a review). First, the geometry (physical bounds)
and volume of the space is defined. For the airways, this step may be accomplished with high
resolution CAT (Computer Assisted Tomography) or MRI (Magnetic Resonance Imaging) scans
or, for laboratory animals, step-sectioning. This step is followed by dividing the volume
occupied by the fluid into a mesh with numerous small but discrete cells, each cell being either
uniform or non-uniform in shape. The physical modeling (e.g., motion equations) and boundary
conditions are then defined and the model is solved at the level of each individual cell in the
mesh. For the airways, the mesh edges (actually cells) represent the external boundaries of flow
and will consist of an inlet (nares), an outlet (nasopharynx) and some form of restraining walls
(lining tissues/mucus of the ET region). Modeling may be specified such that flow cannot cross
these boundaries with specific values given to the inlet and outlet flow. Typically, the boundary
condition is of a fixed concentration (e.g., C = 0 implies that the gas is so highly reactive that is
disappears at the surface) or of a mixed type (i.e. diffusion flux in the gas is proportional to the
surface concentration, implying that there is a linear, steady-state, diffusion-reaction process in
the underlying tissue). If the modeling results do not meet the boundary specifications, greater or
lesser inlet and outlet flows are chosen and the calculations begin all over again until the
boundary specification are met or approached as closely as possible. The resulting solutions are
then analyzed and presented, thereby defining the solution for the flow field, i.e., the air in the
respiratory tract, with each solution having coordinates of directions and units of velocity
(typically m/s).
Once the flow field is defined, a second set of equations is solved to describe the
transport of the gas present within the air flow field. These transport solutions consider both
convection (i.e., the gas being carried along with the air) and diffusion (i.e., gas transport due to
concentration gradients). The predictions for the gas in this fluid flow presented in CFD
solutions are in terms of flow or "flux"1 usually to a surface. Flux is essentially a surface
bombardment rate. Depending on the nature of the substance, each type of flow/flux has its own
distinct units of measurement along with distinct physical constants. For movement of gas- (or
particle-) containing air within spaces such as the regions of the respiratory tract, the units are
typically for mass flux with a rate of mass flow across a unit area, e.g., kg-m"2-s_1.
The goal of inhalation dosimetry is to estimate dose or concentration in target tissues.
Models of air flow, either dye-flow or CFD, visualize or estimate the movement of materials to
1 Flux comes from the Latin fluxus meaning "flow".
September 2009
3-7
-------
surfaces. For the regions of the respiratory tract, flow models give estimates of the flux of
materials present or entrained in the inhaled air to discrete areas. The rate of transfer across the
boundary/surface (i.e. level of flux) may be regarded as an exposure to air with the agent or
toxicant of concern having units of mass flow to a unit area. Thus, these model outputs may
yield a quantitative estimate of materials flowing to the boundaries of their meshes (i.e., to the
tissue surface) but do not afford an estimate of an actual dose to the tissue, whose units are
typically mass to a volume or weight.
3.4.1. CFD Air Flow Models of the Rat ET Region- Correlation with Dye-Streamlines
The observations on fluid flow in nasal casts clearly illustrate a basis for nonuniform flow
occurring throughout the ET region of the rat. Kimbell and co-workers (1997a) created a three-
dimensional mesh grid reconstructed from serial step-sections of the nasal passages of a male
F344 rat. The model extended from the external nares posteriorly to the nasopharyngeal meatus,
nearly the same range as the acrylic casts discussed above. Computational simulations were then
run to solve equations for the direction and speed of the airflow and pressure at each of the more
than 143,000 nodes of the mesh. Visualization of these simulations (Figure 3-4) illustrates
inspiratory air flow in the anterior nose to consist of five major streams flowing anterior to
posterior as was observed in studies using clear casts of rat nasal airways: (1) ventral lateral
(VL); (2) ventral medial (VM); (3) middle medial (MM); (4) dorsal lateral (DL); and (5) dorsal
medial (DM). Four of these five simulated streams flowed together and then exited via the
nasopharynx. These CFD simulations provided further resolution of the DM stream which
flows into and out of ethmoid recesses and is itself subdivided into two components. The flow
patterns of the DM stream are resolved sufficiently in the simulations to show clear structure
with localized areas of high and low flows to the subtending tissues in this area, the olfactory
epithelium. Simulations also revealed a portion of flow in and through the recesses in the
ethmoid areas flowing in reverse, i.e., posterior to anterior. Furthermore, simulations revealed
marked differences in flow rates. Flow in the ethmoid substructure was more than an order of
magnitude slower than flow in interior and ventral portions of the nasal airway.
September 2009
3-8
-------
B
Source: Adapted from Kimbell et al. (1997a)
Figure 3-4. Major inspiratory flow streams simulated in the F344 rat nose. The
external nostril is to the right. Simulated streamlines of inspiratory airflow emerging from
points near the medial septal wall are shown in (A) and from points near the lateral wall in
(B). Dorsal streams are to the top as shown in Figure 3-3. The simulated flow rate was
288 mL/min.
Localized volumetric flows and their apportionment were quantified by integrating the
velocity component for designated cross-sections of the model. For the cross-section at a level
approximately midway in the airway model, the % apportionments at a flow rate of 100 mL/min
are noted in Table 3-1.
Yang et al. (2007) made similar observations using a male Sprague-Dawley rat over a
range of tidal volumes.
Table 3-1. Nasal flow stream allocation in the rat and summary of CFD simulated flow
apportionment (as a % of total volumetric flow) to various cross-sectional areas (as mm2)
of the ET region.
Flow stream region
Cross-sectional stream areaa(mm2)
Flowb (%)
Dorsal medial
0.65
15.5
Dorsal lateral
1.15
35.6
Middle medial
0.83
29.9
Ventral lateral
0.15
0.9
Ventral medial
0.64
11.5
Total
3.42
93.4
a from Kimbell et al. (1993). SA refers to estimates from a cross-section of the nasal cavity, approximately
midway between the anterior nares and the nasopharynx.
b from Kimbell et al. (1997a) [%DM = %DMS (dorsal medial stream - septum) + %DMN (dorsal medial stream -
nasoturbinate)]
September 2009
3-9
-------
3.4.2. CFD Air Flow Models of the Human ET Region
The observations reported above clearly illustrate the applicability of CFD modeling
techniques for resolution of flow occurring in the ET region of the rat. Two independent
approaches, dye-visualized and computer-simulated, consistently revealed a high degree of
complexity and nonuniformity of airflow patterns in the ET region.
The applicability of this noninvasive approach to investigate nasal airflow characteristics
in humans was soon realized. Several investigators proceeded to develop intricate mesh models
based on highly refined human MRI and CAT scans and to simulate and characterize flow in,
through and around the ET region. Figure 3-5 presents a visual summary of three of these
studies. Results analogous to those observed in rat analyses are obvious with medial, ventral,
lateral and dorsal airflow streams being observed from the simulations. These simulations of
flow in human ET regions also predict low flow apportionments to the dorsal regions with
accompanying vortices. Increased airflow resulted in increased complexity, especially in the
dorsal regions where larger and multiple vortices were simulated.
Localized volumetric flows and their apportionments were also determined in a number
of these studies. The basis of these analyses was by assigned regions of the nasal tract, from
dorsal to ventral, typically a total of six. A partial summary of these results is presented in Table
3-2. These results indicate a wide range of flow values to these subregions of the ET. Also,
simplistic apportionment of percent flow per mm2 surface area of these various regions can be
seen to result in a range of values; e.g., from 0.12 (1.2%/9.7 mm2) in region B as reported by
Wen (2008) to 1.03 (28.7%/27.9 mm2) in region E as reported by Keyhani et al. (1995). These
composite simulated results clearly indicate a high degree of nonuniformity of flow within the
human ET region and variability between models.
September 2009
3-10
-------
(b)
Source: Wen et al. (2008)
Figure 3-5. Representations of CFD-simulations of flow streamlines in human nasal
cavitiesfrom (a) Keyhani et al. (1995), (b) Schreck et al. (1993), and Subramaniam et al.
(1998). Airflow from left to right.
Table 3-2. Summary of CFD simulated flow apportionment (as a % of total at 15
L/min) on the cross-sectional area in the middle turbinate (as mm2) of the ET region in
selected human models as analyzed by Wen et al. (2008).
Regions: Dorsal (A)
to Ventral (E,F)
Wen et al. 2008 (left)
Subramaniam et al 1998
Keyhani et al 1995
Cross-sectional
area
Flow
Cross-sectional
area
Flow
Cross-sectional
area
Flow
A
13.7
11.6
7.9
1.9
15.6
11.4
B
9.7
1.2
15.4
1.9
6.0
3.0
C
23.2
21.4
20.8
11.3
35.5
27.3
D
21.6
20.3
54.8
46.7
27.9
18.3
E
50.3
43.7
20.5
24.4
27.9
28.7
F
42.8
1.8
28.9
13.9
26.5
11.3
Total
161.3
100
148.3
100
139.4
100
3.4.3. CFD Air Flow Models - Predictions of Reactive Gas Distribution in the ET Region
The preceding section briefly described basic aspects of CFD methodology applied to
inspiratory airflow in both laboratory animals and humans. These results indicate that inspired
airflow to the various areas of the ET region is highly nonuniform. Some reasons for the
nonuniformity of flow have their basis in the extensive departures from unimpeded flow that
airway anatomy and geometry in different species impose on the incoming airstream. This
report, however, deals not directly with just flow through and out of the ET region, but more
with the flow of gases as they are directed and distributed to the surfaces within the ET,
especially in humans. This section is concerned with the application of CFD to address the fate
of inspired gases within the upper respiratory tract.
September 2009
3-11
-------
In early efforts, Kimbell and co-workers (1993, 1997b, 2001a, 2001b; Kimbell and
Subramaniam, 2001) used CFD modeling of airflow in the ET regions of laboratory animals and
humans as a basis to describe disposition of inhaled gases using formaldehyde, a highly water
soluble and reactive gas, as an example. These authors, and others following, began this work
based on the assumption that low concentrations of a gas present in inspired air do not modify
the characteristics of the air (e.g., viscosity, density) and thus do not influence airflow itself.
Although gas present in the airstream would not alter the airflow characteristics of the
airstream, the airflow would have an influence on the disposition of the reactive gas through both
the airflow apportionment and the molecular diffusivity of the gas. It is the molecular diffusivity
of the gas that results in a net transport of mass to the surface. And it is the mass flow rates
across a unit area, or mass flux, that are the predictions of CFD solutions. Thus, these authors
carried out gas transport simulations by incorporating the molecular diffusivity of formaldehyde
with units of cm2/sec in air. These authors also made assumptions that limited the goal of the
study to predicting the transport of the inhaled gas to the surface walls of the ET (or rather the
mesh cells at the wall representing the surface wall of the ET).1 The model simulated mass flux
to the walls1 but not into the tissues at the boundary of the mesh wall.
Results from the gas transport simulations were visualized as a contour plot of simulated
formaldehyde mass flux across the nasal septum (Figure 3-6). This visualization is conceptually
consistent for a highly soluble and reactive gas entering the nasal cavity along major airstreams
and passing to the walls in an anterior to posterior gradient. High mass flux regions or "hot
spots", were predicted to include the ventral and lateral margins of the nasoturbinate as well as
the lateral wall. Conversely, low flux regions were predicted in regions more posterior to the
nares and out of the major flowstreams.
The results of the study by Kimbell et al. (1993) were among the first to demonstrate the
application of CFD to regional dosimetry of inhaled gases in predicting quantitative mass flux
patterns to surface (mesh) walls, which acted as a sink. Consistent with other advances related to
predictive dosimetry for the ET region, these results also give indications that considerable levels
of nonuniformity exists across the surfaces of the ET region, in this instance for mass flux. For
example, results discussed in this study indicate that maximal mass flux occurred on the lateral
aspect of the nasoturbinate and the adjacent lateral wall where flux was predicted to be higher, in
a few instances approaching two orders of magnitude higher, than flux on the medial surface of
1 The walls of the mesh were modeled to act as "sinks" such that the reactive gas concentration at the walls was set
to zero effectively allowing maximal uptake consequently eliminating "back pressure" or desorption to limit wall
uptake.
September 2009
3-12
-------
the turbinate. These areas of high flux and their relationships to lesions will be discussed in
Section 3.4.6 regarding target tissue dosimetry.
Source: Adapted from Kimbell et al. (1993).
Figure 3-6. Patterns of simulated wall mass flux of formaldehyde across the nasal
septum of the rat. Red denotes regions of high flux; PV, posterior nasal vestibule; DM,
dorsal medial meatus: MM, middle medial meatus: VM, ventral medial meatus. Airflow
from right to left.
3.4.4. Interspecies CFD Air Flow Models Predictions of Gas Distribution in the ET Region
Subsequent to the initial studies of Kimbell et al. (1993), a number of investigators
developed and published similar sophisticated models for various species and gases of different
solublilites and reactivities. A listing of these studies is provided in Table 3-3.
Inclusion of studies in Table 3-3 was contingent upon a preliminary general evaluative
screening of the studies and their models. All studies listed provide sufficient detail of the CFD
technique to establish the accuracy of the results. These elements include the source of the
geometrical representation used, as well as a description and analysis of the mesh including the
basic structure and character of the individual cells of the mesh. Most also include some
evaluation as to the dependency of modeling results on mesh size (cell number). Statements
regarding the assumptions and simplifications made to the basic Navier-Stokes equations for
calculating results were also provided.
September 2009
3-13
-------
Table 3-3. Summary of airway flow studies using CFD techniques.
References
Species
Region
Source for
Mesh
Construction
Purpose
Elad et al.
(1993)
Human
ET
geometric model
Air flow and flux patterns, particularly to olfactory
region
Tarabichi and
Fanous(1993)
Human
ET
CAT scans
Airflow characteristics
Keyhani et al.
(1995)
Human
ET
CAT Scans
Airflow characteristics at resting breathing rates
Bush et al.
(1998)
Rat
ET
(olfactory)
CAT scans
Airflow data as input to PBPK model simulating
uptake in various compartments
Frederick et al.
(1998)
Rat/Hum
an
ET
(olfactory)
CAT scans
Combine airflow flux with deposition patterns to
determine in human olfactory tissues receives more
or less than rat
Kepler et al.
(1998)
Monkey
s
ET
(olfactory)
video image
Resting airflow rates used to model uptake of
formaldehyde
Shome et al.
(1998)
Humans
pharynx
anatomical
model
Modeling of laminar and turbulent flow in the
pharynx and its effects of morphological changes
Chometon
et al. (2000)
Humans
ET
anatomical
model
Determination of the separation of airflow between
the meatus and olfactory regions
Sarangapani
et al. (2000)
Humans
Lung
Weibel 4-lobe
model
Modeling the influence of aerosol dispersion on
particle deposition in lung
Bush et al.
(2001)
Humans
ET
CAT scans
Modeling diffusion and uptake of ozone and
chlorine in the respiratory tract using mouth and
nasal breathing
Martonen et al.
(2001)
Humans
ET and TB
Casting, then
sliced, scanned
and digitized
Development of upper airway representation for
simulations of airflow and particle transport
Kurtz et al.
(2004)
Humans
ET,nasoph
arynx
3-D model of
human nose
Comparison of experimental odorant nasal
absorption data in humans with simulations
Zhao et al.
(2004)
Humans
ET
(olfactory)
CT scans to
models
Developed a method to convert CT scans of a
human nose into a anatomically-correct 3-D model
to predict airflow and odorant transport
Zhao et al.
(2006)
Humans
ET
(olfactory)
CT>3-D models
Used CT scans and CFD to determine improvement
in nasal airflow and odorant delivery rates in a
patient after surgery
Minard et al.
(2006)
Rat
ET
MR imaging
MR imaging used to validate CFD predictions
De Rochefort
et al. (2007)
Humans
lung
CT
CFD/CT used to model airflow in upper airways
Wen et al.
(2008)
Humans
ET
CT
Comparisons of laminar flow patterns between the
left and right nasal cavities
Corley et al.
(2009)
Rabbits
ET
MR
Airflow in rabbit respiratory and olfactory
epithelium
Furthermore, Kimbell et al. (2001b) constructed anatomically accurate, 3-dimensional
computational fluid dynamics models of nasal passages of F344 rat, Rhesus monkey, and
humans for the purposes of modeling inhaled formaldehyde. The rat and monkey CFD models
were constructed from tracings of airway outlines of the right nasal passages from the nostril to
the nasopharynx from serial-step sections of embedded tissue specimens (Kimbell et al. 1993;
September 2009
3-14
-------
Kepler et al., 1995). The human model was constructed from airway tracings of both sides of the
nasal passages from the nostril through the nasopharynx from magnetic resonance image scans
as described by Subramaniam et al. (1998).
Key
pmol/(mm:z-hr-ppm)
2000
1500
1000
500
0
F344 Hat
Rhesus Monkey
Human
Source: Kimbell et al. (2001b).
Figure 3-7. Nasal wall flux spectra of inhaled formaldehyde simulated in rats, monkey
and humans at normal inspiratory flow rates. Nostrils are to the right.
Simulations configured for uptake of formaldehyde were run for all three of the ET
models, the results of which are shown in Figure 3-7. Despite the absolute difference in size,
with the rat ET being 40-fold smaller than the human ET, comparative aspects regarding flux are
apparent. Visual inspection of Figure 3-7 shows clearly, for example, that relative proportions of
area for the highest formaldehyde flux in the ET region ("red" in color) is highest in rat, with the
rank order following rat > monkey > human. The authors estimated both maximum and average
areas of formaldehyde flux over the whole ET in each species in which the maximum:average
flux ranged from 3- to 10-fold among these three species (Table 3-4).
September 2009
3-15
-------
Table 3-4. Estimates of formaldehyde flux to ET surface walls for various species.
Formaldehyde flux estimate(pmol/[mm2-hr-ppm])A'B
Area
Rat
Monkey
Human
Whole nose: maximum
3210
4492
2082
Whole nose: average
336
508
568
Maximum/Average
10
9
4
Nonsquamous: maximum
2620
4492
2082
Nonsquamous: average
284
535
611
Maximum/Average
9
8
3
Source: Adapted from Kimbell et al. (2001b).
Simulations conducted at inhaled concentration of 1 ppm formaldehyde.
Simulations conducted at flow rates of twice the minute ventilation for rat (576 mL/min), monkey (4.8 L/min),
and human (15.0 L/min).
3.4.5. Range and Distribution of Flux in ET Regions for Various Species
The general ranges of rat and human flux in the ET region estimated from visual
inspection in Figure 3-7 and from the average and maximum ranges in Table 3-4 were further
analyzed by Kimbell et al. (2001b). The ET regions for each species were first partitioned into
20 evenly spaced flux levels between zero and the maximum predicted flux value for each
species; 2620 pmol/(mm2 -hr-ppm) at a flow rate of 576 mL/min in the rat and 2082 pmol/(mm2
-hr-ppm) at a flow rate of 15 L/min in the human. Surface areas of the ET found to be within
these flux levels were then assigned or "binned" accordingly. This strategy allowed for
estimating the distribution of flux levels over the surface area of the ET, each species being
"normalized" to its respective range of flux. The binning revealed that flux values higher than
half the maximum flux value (flux median) were predicted for nearly 20% of human ET surfaces
whereas only 5% of rat ET surfaces were associated with fluxes higher than flux medians. This
relationship was maintained for flux levels higher than 75% of the maximum flux value with
approximately 1.8% of human but only approximately 0.6% of rat ET surfaces were exposed to
this higher level of flux. Distribution within the ET region of what may be considered "high"
flux will be examined in the next section in relation to actual occurrence of lesions in the ET
region.
3.4.6. Correlation of High Flux with Lesions in the ET Region
The information discussed and presented thus far in this section on uniform vs.
nonuniform airflow in the ET region has been for simulations, either ex vivo with water-soluble
dyes through clear acrylic casts of ET regions or in silico using elaborate CFD models based on
September 2009
3-16
-------
tracings or scans of the ET regions. These modeling approaches have given very similar and
internally consistent results concerning patterns and distributions of airflow and gases in the ET
region of various species, including humans. Supporting empirical observations would,
however, provide a more robust basis for these modeling results.
One logical strategy that could provide support and reinforcement for the modeling
results would be to examine the extent of correlation between flux and lesions. More
specifically, such an activity would involve localization of areas in the ET region modeled to
experience high flux from an inspired gas and then examine the extent of correlation of these
areas with actual lesions, both in incidence and severity, observed in response to in vivo
exposures to the same gas. With flux regarded as a surface bombardment rate, flux of dilute gas
in air to ET surfaces could be reasonably anticipated to exhibit some degree of a dose-response
relationship. There are currently several examples of such a correlation analysis in the current
literature. Results from two of these studies, formaldehyde (Kimbell et al., 1997b) and hydrogen
sulfide (Moulin et al., 2002), are summarized here.
Kimbell and colleagues (1997b) investigated the relationship between squamous
metaplasia and areas of high formaldehyde flux in the nasal tissues of adult rats. They first
performed a pathology analysis on the incidence of squamous metaplasia in specific areas of the
nasal tracts of rats that had been exposed via inhalation to formaldehyde at 0, 0.7, 2, 6, 10 or 15
ppm for 6 hr/d, 5 d/wk for 6 months. These specific areas, primarily the lateral meatus and mid-
septum of the anterior portion of the nose, known to have extensive distributions of target tissue,
were observed for analysis by making a transectional cut at the midline. The perimeter of the
airway opening revealed by this cut was then divided into 20 regions based on anatomical
landmarks and the location of major airflow streams. The sectioned head of a male rat was then
used to generate a computational mesh for CFD analysis with normal squamous epithelium
mapped onto the walls of the mesh such that this tissue was a defined area on the airway walls of
the CFD model. Transport of formaldehyde through the air and into the nasal epithelium was
assumed by the model to occur by convective forces and molecular diffusion. Incidence of
squamous metaplasia was then calculated for each region and flux values modeled for each
region, both ranked high to low and statistically analyzed for correlation. The results, shown in
Figure 3-8, provide clear evidence that, at high flux levels (and high exposures) of formaldehyde,
the distribution of squamous metaplasia is related closely to the location of regions of high
formaldehyde flux into airway walls.
September 2009
3-17
-------
A.
100n
15 ppm
to ppm
A 6 ppm
80-
a 70~
1
E 40-
20
10-
120
scuam absorbing
20
8 10 12 14 16 18 20
6
0
2
4
Region number
Source: Kimbell et al. (1997b).
Figure 3-8. Graphs showing (A) the incidence of formaldehyde-induced squamous
metaplasias and (B) modeled formaldehyde flux values along regions assigned to the
perimeter of a transected nasal airway of rats.
Moulin et al. (2002) used a similar approach to investigate the relationship between
lesions in olfactory epithelia and areas of high hydrogen sulfide flux in the nasal cavity of adult
rats. They also performed a pathology analysis on the incidence of lesions (olfactory neuronal
loss and basal cell hyperplasia) in two transverse nasal sections, one directly through the ethmoid
turbinates, obtained from adult male rats (n = 12/concentration) that had been exposed to
hydrogen sulfide at either 0, 10, 30 or 80 ppm for 6 hr/d for 70 days. The CFD model used
consisted of a 3-dimensional nasal airway reconstruction of a mesh from serial cross sections of
an adult rat which provided a simulation of flow and mass transport through the entire airway.
Flux predictions were made at the same level of the transverse nasal section (through the
ethmoid turbinates) that had been demarcated around the perimeter into 39 regions. Transport of
hydrogen sulfide through the air and to the nasal epithelium was assumed by the model to occur
September 2009
3-18
-------
by convective forces and molecular diffusion. The incidence of olfactory lesions and flux levels
for 80 ppm hydrogen sulfide were then calculated for each of the 39 regions. Rank correlations
between lesion incidence and flux were than carried out. Distinct hot spots of regional flux
occurred in the ethmoid turbinate section at those regions corresponding to high airstream flow.
These results are presented in Figure 3-9. These regions of high flux were closely associated
with hydrogen sulfide-induced nasal lesions if that region was lined by olfactory epithelium. An
additional observation made regarding high hydrogen sulfide flux was that pathology was not
observed in those regions that were lined with respiratory rather than olfactory epithelium. This
lack of correlation between high flux and non-olfactory epithelium lesions is apparently due to
resistance of this tissue to hydrogen-sulfide toxicity.
Source: Moulin et al. (2002).
Figure 3-9. Schematic diagram of the transverse nasal section through the ethmoid
turbinates (top left, Section 2 of the nasal cavity) with plot of lesion incidence at 30 and
80 ppm (top right). Note coding of surface areas in schematic to the x-axis of plot. Plot of
predicted hydrogen sulfide flux under different assumptions of uptake: low - 20%, medium
- 40%, and high - 80% (bottom right). Diagram of same section (bottom left) under
intermediate uptake conditions at 80 ppm where red corresponds to 320 pg/(mm2-s) and
blue corresponds to 0 pg/(mm2-s). Plot of predicted flux at 80 ppm (bottom right) on
designated regions of Section 2.
Nasal Region Number
m- Medium Uptake
Low Uptake
hftflh Uptake
10 20 30 40
Nasal Region Number
September 2009
3-19
-------
Similar correlations between lesion incidence and tissue dose were also observed for
acrolein (Schroeter et al., 2008) and diacetyl (Morris and Hubbs, 2009). For acrolein, predicted
air:tissue flux from the rat nasal CFD model compared well with the distribution of nasal lesions
observed in a subchronic inhalation study. In the case of diacetyl, a strong correlation of injury
location and pathology severity scores with CFD-PBPK hybrid model predicted tissue
concentrations was observed in nasal tissues.
These examples provide strong support for a direct relationship between flux and
responses in tissues in four independent cases. The hydrogen sulfide case offers resolution
sufficient to demonstrate expected target tissue specificity as responses were only seen in
olfactory epithelium despite equivalent flux levels in more proximal respiratory epithelium.
Also, in the case of formaldehyde, the relationship between flux intensity and response appears
to be one in which high flux is more predictive of lesions than are low flux levels (i.e., either
appreciably above or below the median flux level). A similar relationship between flux intensity
and response was observed with acrolein. These additional observations indicate the flux-
response relationship to have a high degree of resolution as well as providing predictability of a
dose-response relationship. The predicted nasal tissue concentrations of diacetyl corresponded
with lesion location and severity. Thus, results from these four examples may be regarded further
as providing compelling support for a close correlation between flux and tissue-specific
responses.
In summary, the intent of this portion of the report is to bring forth and discuss recent
findings that inform short comings with the basic assumptions following from the use of Ve/SA,-,
and in particular VE/SAET, as the basis for default interspecies dosimetry of inspired gases.
Chapter 2 presented the reasoning behind the use of Ve as a surrogate for gas dose and use of SAr
for normalization of dose along with acknowledgment of the general advancement this concept
made to inhalation dosimetry. Figure 3-10 provides an illustrative summary and perspective of
what this chapter has presented regarding the default procedure. This figure illustrates the
processes and assumptions of uniformity underlying VE/SA originally shown in Figure 3-1 The
composite evidence and results from this information indicate nonuniform character of flow, of
surface areas, and of deposition for gases in the ET region.
September 2009
3-20
-------
lx dose
Human
1 - nonuniform uneven flow of gas present in VE through the respiratory tract region
2 - nonuniform uneven SAs of tissues in the respiratory tract region subtending VE
3 - nonuniform uneven deposition of gas present in VE over the entire surface of the respiratory tract
Geometry &
Anatomy
Dye-flow
Lesions &
Flux-lesion
correlation
CFD ET models
Figure 3-10. Representation of application of current information to the processes,
assumptions, influences and outcomes for RfCMethods basic default procedures for
comparative gas dosimetry. The text-containing arrows below the figure represent the
information presented in this chapter that has addressed the assumptions of uniformity.
The arrows and labels extending to the right demonstrate the outcome of applying this
information to the overall process introduced in Figure 3-1.
Thus, more extensive evaluative tools will need to be used to integrate newer information
to the end of making more informed estimates of interspecies dosimetry. These tools and the
results from their application are the principal subject of Section 3.5 of this report.
3.5. EVALUATION AND USE OF MODELS IN INTERSPECIES INHALATION
DOSIMETRY
As discussed previously, the default RfC method for interspecies dose-extrapolation for
the ET region is determined by the ratio VE/SAET of animals to humans. When applied for an
interspecies comparison with rats, for example, the calculation results in a RGDR of
approximately 0.2, indicating that the inhalation delivered dose to humans is at least fivefold
greater than to the rat.
September 2009
3-21
-------
Evidence indicates that flow and distribution of VE in the respiratory tract is not uniform,
and that CFD modeling offers refined and characterized disposition of airflow and of the gases
present in VE. Also, CFD offers resolution only to the surface of the tissues in contact with the
flow. It does not and cannot offer resolution of interspecies tissue dosimetry, that is, the
concentration of the gas in the target tissues of the respiratory tract.
The purpose of this section is to present new information that relates these basic issues of
dosimetry in the airways, i.e., flux of gas to the tissue surfaces and the concentration of gas in the
airway tissues. The focus of this section is on the results and insights gained from the
combination of modeling approaches of CFD on airflow and gas disposition in the airways with
the state-of-the-science developments from physiologically-based pharmacokinetic or PBPK
models describing pharmacokinetics in respiratory tract tissues.
Various models have been developed and utilized to examine and quantitatively estimate
dose to target tissues via the inhalation route of exposure in both animals and humans. As
discussed above, 3-D, anatomically accurate CFD models were developed to model inspiratory
airflow and estimate regional uptake and amount of inhaled gas reaching sites in the nose, with
formaldehyde serving as the vanguard example for this application (Kimbell et al., 2001a, 2001b,
1997a, 1997b, 1993; Kimbell and Subramaniam, 2001). One output of these simulations is an
estimate of flux which is the rate of mass transport in the direction perpendicular to the nasal
wall, typically with units of pmol/(mm2-h-ppm). In this regard, the CFD model estimates a
"dose" of gas delivered to the tissue boundary but not into the tissue itself. CFD modeling
simulations can also be utilized to estimate the surface area and volume of specific anatomical
features, the allocation of inspired air to specific flow streams, and gas phase mass transfer
coefficients (Kimbell and Subramaniam, 2001). While representing a major step forward in
describing and refining interspecies inhalation dosimetry compared to the current default RfC
methods, the nature of CFD model output (in terms of flux) does not provide a definitive
measure of target tissue dose. Developments in tissue physiology and kinetic models, however,
have provided this information and allow for more refined and accurate dosimetry in and
between species. Combination of these two modeling approaches led to the development of
CFD-PBPK hybrid models allowing for the most highly refined and accurate estimates of target
tissue dose currently available.
3.5.1. Overview of CFD-PBPK Hybrid Modeling - Combination of Gas Transport in the
Air Phase into the Liquid/Tissue Phase
CFD-PBPK hybrid modeling represents the state-of-the-art science for examining
inhalation dosimetry. As discussed in Bush et al. (1998), combined CFD and PBPK models were
September 2009
3-22
-------
developed to help address how factors related to airway anatomy might be a reason that other
models assuming uniformity, such as the ventilation-perfusion model, failed to fully explain the
effects of gas flow on total vapor uptake in different animal species. As discussed above, the
CFD models developed by Kimbell and colleagues (1997a, 1997b, 1993) helped to address
toxicant transport in the gas phase. The PBPK model developed by Morris et al. (1993) was an
attempt to account for toxicant transport in airways and tissue by dividing nasal flow into
multiple streams and the nasal epithelial compartments into tissue stacks, but its modeling of the
gas phase was less exact than that of CFD. Consequently, Bush and colleagues (1998) developed
a hybrid model based on combining these two aspects, the CFD model for consideration of gas
disposition in the air phase and the PBPK models for consideration of gas disposition into the
liquid and tissue phases within the rat nose.
Bush et al. (1998) proceeded to utilize the CFD model of flux distributions to determine
the distribution of air flow and gas phase mass transfer coefficients along the four major flow
paths (i.e. the DM, DL, MM, and VM) in the rat nose to predict gas transport to the surface
(actually a modeling "boundary"; see below) of the underlying nasal tissues. The VL was
combined with the adjacent DM because the CFD simulations revealed that the VL only carried
about 1% of the inhaled flow. Using these solutions to guide the PBPK model development,
these authors subdivided the four flow paths into four distinct meatuses such that each contained
a major air flow stream predicted from CFD simulations. The underlying tissues in these
meatuses were then further divided into four serial regions attached to a tissue stack containing
mucus, epithelial, and subepithelial compartments. The steady state diffusion equations
governing these 16 airway regions and associated tissue stack compartments constituted the
PBPK portion of the model. The PBPK and CFD models were coupled at the gas-tissue phase
interface by a permeability coefficient - termed Kgm - that incorporated the gas phase mass
transfer coefficient (kg) with a mucus phase diffusion parameter. A graphical representation of
this approach, including areas of airflow and composition of tissues stacks modeled is presented
in Figure 3-11. Similarly structured models to predict interspecies gas dosimetry have been
described for both the rat and human nasal regions for other gases including vinyl acetate (see
Bogdanffy et al. 1999).
September 2009
3-23
-------
DM
DL
R
R
R
R
MM
»-
Mucus
Epithelial
Epithelial
Epithelial
Epithelial
Subepithelial
Mucus
Epithelial
Epithelial
Subepithelial
Source: Bush et al. (1998).
Figure 3-11. Schematic of the CFD-PBPK hybrid model of the rat nasal cavity. This model
divides the tissue into four serial levels along the DM, DL, MM, and VM flow paths. Each
of the resulting 16 tissue stacks is classified as either respiratory (R) or olfactory (O). The
exploded views of R and O tissue stacks indicate their internal compartmental structure.
The model and its predictions were then validated by using overall uptake data from rat
inhalation studies for three nonreactive vapors that were either completely inert (acetone),
reversibly ionized in aqueous media (acrylic acid), or prevented from being metabolized by an
enzyme inhibitor (isoamyl alcohol). This CFD-PBPK hybrid model was thus parameterized and
validated with empirical observational data to model actual uptake into tissues such that actual
tissue concentration of the test case vapors could be predicted.
The results of this modeling work showed variation of surface area, cross-sectional area,
and values of kg along the differing flow paths among the regions, reflecting the convoluted
nature and complexity of the rat nasal geometry. The overall uptake predicted by the CFD-PBPK
model for the toxicants was not sensitive to the exact value of kg estimated by the CFD analysis
even when reduced by 95% or made essentially infinite. Moreover, as one component of Kgm
(the coefficient governing the mass flux of gas from the flowstreams into the underlying mucus
compartments), increasing kg had little effect on the concentration or uptake distribution of gas
into the first epithelial layer. In fact, when kg was simulated in the model to be sufficiently large
to remove the mass transfer resistance of the gas phase, the uptake of both acetone and acrylic
September 2009
3-24
-------
acid was limited by the rate of diffusion in the mucus and tissue despite a 30-fold difference in
their solubilities. These results suggest that, in the rat, diffusion through the gas phase is so rapid
relative to diffusion through the mucus and tissue layers that it does not influence the rate of gas
absorption. This observation allowed for a simplifying assumption to be made that a
compartmental kg value can be determined by averaging the local values obtained by CFD
analysis for those airway sections that constitute the compartment (e.g., respiratory and
olfactory). As further discussed in Frederick et al. (1998), a "compartmental mass transport
coefficient" (i.e., one kg value per tissue compartment) used in this context should be derived and
defined as an "average mass transfer coefficient" based on a difference in "average
compartmental concentrations." This derivation and definition is required when the
compartments of the PBPK model are commonly described as "well-mixed", and the results of
the CFD simulations need to be appropriately expressed so that the "average air compartment
concentration" derived from CFD is defined as the same way as the "average tissue compartment
concentration" in the PBPK model.
3.5.2. CFD-PBPK Hybrid Modeling and the Overall Mass Transport Coefficient - Kg
In the hybrid modeling approach described by Bush et al. (1998), the PBPK and CFD
models were coupled at the gas-tissue phase interface by Kgm. The aim of this approach was to
determine the regional dose within the respiratory tract by characterizing the transport of gases
between the air phase, the intervening surface liquid and tissue, and the blood. Kgm is also
referred to as the overall mass transport coefficient or Kg.
As described in Appendix I of the RfC Methods, the concept of Kg (the overall mass
transport coefficient, MTC) is used to describe transport through several different phases
including air and liquid. The most definitive evaluation of transport is to describe absorption by a
simultaneous solution of the conservation of momentum and mass in the complex 3-D airway
and tissue structure. For purposes here, however, the concept of Kg as a two phase MTC based
on the physicochemical properties of the gas may be used to direct development of model
structures adequate for estimation of regional gas dose. The structure of the two-phase mass
transport model may be further simplified to a minimal number of parameters that may be scaled
to gases differing in their physicochemical properties. The basic structure of this approach,
which relies on Kg, was used in the CFD-PBPK hybrid models developed by Bush et al. (1998)
and subsequently by Frederick et al. (1998), both of which incorporate the output from CFD
simulations to describe anatomy, airflow, kg, and flux of inhaled gas in the portal-of-entry (air
phase) linked to a PBPK model describing the systemic compartments (tissue phase).
September 2009
3-25
-------
The basis of the CFD-PBPK hybrid model is very similar to the approach described in
Appendix I of RfCMethods. There, the respiratory tract is simplified into a two-phase resistance
model (because it is applied as an inverse function, Kg may be viewed as a measure of resistance)
to illustrate the overall mass transport approach in which it is assumed that the blood
concentration is constant. For highly soluble and very reactive gases, such as formaldehyde, the
blood concentration has been assumed to be zero. This type of "boundary condition" is used, for
example, in the CFD simulations for formaldehyde (Kimbell et al., 2001b) and acrylic acid
(Frederick et al., 1998) in which the concentration of gas in the nasal airway lining is zero, so
that transport of the gas occurs primarily through the air phase to the surface-liquid/tissue
interface but not into the tissue phase. Thus, for this approach, it is assumed that the surface-
liquid and tissue phases are a single phase because of the limited data to identify differing
transport parameters for these two phases.
Bush et al. (1998) and Frederick et al. (1998) provide an updated and modified version of
the approach presented in Appendix I to describe gas phase mass transport in which the estimate
of the overall transport or flux, N, across and ainliquid interface is expressed by
N = Kg (Cg — Ct/P) Equation 10
where Kg (cm/min) is the overall mass transfer coefficient, Cg (|imol/cm3) is the air phase gas
concentration, and Q (|imol/cm3) is the concentration in the liquid/tissue phase, and P (unitless)
is the surface liquid/tissue:air partition coefficient.
In general, the overall mass transport coefficient, Kg, from the air phase into the liquid
phase may be determined from the transport coefficients of each individual phase, such that
1/ Kg = 1/ kg + l/(P*kt) Equation 11
where kg (cm/min) is the gas phase mass transfer coefficient as defined above, and kt (cm/min) is
the liquid phase mass transfer coefficient. Contextually, Kg may be considered analogous to a
tissue clearance term used in compartmental pharmacokinetic studies as similar principles apply
(Frederick et al., 1998). In the case where the surface liquid and tissue cannot be assumed to be
a single compartment, a separate partition coefficient and transport coefficient would need to be
incorporated to account for additional compartments. For example, in cases where gas diffuses
through the tissue into the blood and contributes to overall absorption, additional mass transport
resistances must be considered to describe this additional compartment. This is important
because significant accumulation and recirculation of gas in the bloodstream may reduce the
concentration driving force (and thereby reduce the absorption rate) and contribute to the
September 2009
3-26
-------
development of a "back pressure", which may result in desorption during exhalation due to the
reversal in the concentration gradient between the air and tissue.
Following in this approach, kg can be estimated by CFD simulations at several
representative inhalation flow rates with CFD models and species-specific meshes for the
differing nasal geometries of species using defined boundary conditions for the nasal cavity
(Frederick et al., 1998). For example, by conducting CFD simulations assuming that (1) P is very
large to establish a "Ci = 0" boundary condition such that kg ~ Kg; (2) that the compartment is
well-mixed; and (3) that there is negligible change in airflow across the compartment, a
compartmental kg, designated kgc, can be calculated. Thus, kgc describes the compartmental
clearance of gas from the gas phase. In addition, data from CFD simulations at various flow rates
for each species can be used to calculate kgc, since the regional flow, surface area, and fractional
penetration can be evaluated for any defined compartment of the nasal mesh.
A Kgc, overall mass transfer coefficient from the gas phase, can now be estimated that
combines a compartmental mucus phase transport coefficient kmc (analogous to kt in the
definition of Kg - Equation 11) with the gas phase mass transport coefficient kgc calculated from
CFD simulations such that:
1/ Kgc = 1/ kgc + 1/ (P*kmc) Equation 12
The mucus phase mass transport coefficient, kmc, (and thus, by analogy, 1<() can be
calculated as described by Bush et al. (1998) to describe transport to the midpoint of the mucus
layer:
kmc Dmuc/(^muc^2) Equation 13
where Dmuc (cm2/min) is diffusivity of the compound in the mucus (liquid) phase, and tmuc (cm)
is the thickness of the mucus phase diffusion layer.
The definition of the overall mass transport coefficient provided in Equations 11 and 12
may be used to evaluate the conditions in which a single phase, either gas phase or surface-
liquid/tissue phase, determines the overall mass transport coefficient. Gas phase controlled
absorption (i.e., kg « P*kt in Equation 11) is typical for a gas considered reactive and/or
relatively soluble in both the tissue and blood in which the transport resistance through the gas
phase, l/kg, may be greater than the transport resistance in the other phases such that
1/Kg = l/kg Equation 14
The gas phase term, kg, is dependent on flow rate, flow geometry, and the gas phase
diffusivity. In interspecies comparisons, the flow geometry differences of the species are likely
September 2009
3-27
-------
to predominately determine kg as demonstrated in Frederick et al. (1998). Using values of Kg or
Kgc obtained in a single species, Kg can be scaled within the species for a different gas by
decomposing Kg to the individual transport resistances (i.e., the gas phase and surface-
liquid/tissue phase mass transport coefficient).
Liquid phase controlled absorption (i.e., Pkt < kg) is typically identified by a gas of
moderate to low water solubility and low reactivity. In the case of a nonreactive gas, Pkt may be
approximated by the surface-liquid/tissue:gas partition coefficient, the liquid diffusivity (D,
cm2/min), and the thickness of the liquid-tissue layer (t, cm) such that
Pkt = P * |D/(t/2)| Equation 15
An initial difficulty identified in the RfCMethods in using such approaches, i.e., to
determine or decompose an empirically founded Kg, was lack of kg values in airways of
laboratory animals (and humans), and the lack of a data base in which Pk< could be determined.
However, much of this difficulty has been overcome and resolved by the advancement and
validation of CFD models to obtain estimates of kg as the gas phase term is dependent on flow
rate, flow geometry, and the gas phase diffusivity. In cross-species comparisons, the flow
geometry differences of the species are likely to predominately determine kg. For both CFD and
PBPK models, the increase in the amount of data available from various sources defining the gas
and water diffusivities, and partition coefficients for many compounds, as well as parameters
such as tissue surface area, thickness, volume, air flow, etc., used to describe the anatomical and
physiological "compartments" in both animals and humans has aided in advancing these
approaches. Several of the models represented below use kg values determined specifically for
that study or the compartmental (or regional) kg values determined by the CFD simulations
conducted by Frederick et al. (1998) for both animals and humans for the purpose of describing
interspecies inhalation dosimetry.
3.5.3. Results and Analysis of Interspecies Inhalation Dosimetry Modeling
The purpose of this section is to provide an example-based overview of concepts related
to the various aspects of inhalation gas dosimetry discussed in the preceding sections. Table 3-5
provides information for various gases including primary toxicological endpoint(s), measured
nasal uptake in the rat (if available), as well as water solubilities and partition coefficients. This
information was used to provide a physicochemical descriptor for each gas based on the scheme
outlined by Medinsky and Bond (2001). Table 3-6 provides some data on the comparative
calculated or modeled Kg and kg ratios between animals and humans for each of these gases.
Finally, Table 3-7 compares the various methods for interspecies dose-extrapolation (i.e.,
September 2009
3-28
-------
determination of the HEC) between animals and humans for the ET region based on CFD, CFD-
PBPK hybrid, or PBPK modeling for these gases to the default RfC method based on VE /SA. In
addition, the most informative results from the CFD-PBPK hybrid model developed for acrylic
acid are discussed.
In comparing these physicochemical properties, reactivity, and measured uptake for the
representative gases shown in Table 3-5, a general pattern emerges. Those gases with high
uptake (>90% - formaldehyde, acrylic acid) in the ET region are reactive, have higher
liquid/tissue:air partition coefficients (P), and water solubility. Reactive gases with moderate
uptake (acrolein, acetaldehyde, and diacetyl) in the ET region have moderate to low
liquid/tissue:air P values and water solubility. Gases with low uptake (<25% - hydrogen sulfide,
ethyl acrylate and propylene oxide) have among the lowest liquid/tissue:air P values in this group
of gases. However, these gases also exhibit a range of water solubilities and reactivities. In
general, high ET uptake gases tend to be more reactive and scrubbed more efficiently in nasal
tissues with little penetration to the lower respiratory tract and less potential for systemic
distribution. Likewise, low ET uptake gases have the potential to reach the lower respiratory tract
and produce an effect and/or be more readily distributed systemically via the gas exchange area
of the lung. An important consideration in examining the relationship among these properties and
uptake is that they are substantiated by experimental observations of uptake in the rat. In
addition, these generalities are to some extent illustrated when comparing the CFD-predicted
flux values for formaldehyde, acrolein, and hydrogen sulfide. As shown in Tables 3-5 and 3-7,
formaldehyde, the gas with the highest uptake, water solubility and reactivity, has the highest
modeled flux values compared to acrolein and hydrogen sulfide. Taken together, such
information may help one predict in which region of the respiratory tract a given chemical may
preferentially produce an effect or if the chemical might instead or also produce systemic effects.
September 2009
3-29
-------
Table 3-5. Primary toxicological endpoint(s), uptake, properties, and physicochemical
descriptor for representative gases
Endpoint/
Effect
AUptake
in Rat
Liquid/tissue:
air P values
Water
Solubility
Physico-
chemical
Descriptor
References
Formaldehyde
REcTumors and
squamous
metaplasia
> 90%
72,000
(calculated)
400 g/L
soluble-
reactive
Kimbell et al.
(2001b)
Acrylic Acid
OE degeneration
> 90%
6,100
1000 g/L
soluble-
reactive
Frederick
et al. (1998);
Andersen
et al. (2000)
Acrolein
OE degeneration
and atrophy
80- 60%
88 (or 200)
212 g/L
soluble-
reactive
Schroeter
et al. (2008);
Morris (1998)
Acet aldehyde
REandOE
degeneration
80- 40%
140
1000 g/L
soluble-
reactive
Teegaurden
et al. (2008);
Dorman
et al. (2008);
Morris et al.
(1997)
Diacetyl
Nasal, tracheal,
bronchial toxicity
76- 36%
550
200 g/L
soluble-
reactive
Morris et al.
(2009)
Vinyl Acetate
OE degeneration
93- 40%
29
20 g/L
nonsoluble-
nonreactive
Bogdanffy
et al. (1999)
Hydrogen
Sulfide
OE degeneration
and necrosis
26- 18%
2.8
4-5 g/L
nonsoluble-
nonreactive
Schroeter
et al. (2006)
Ethyl Acrylate
OE toxicity
18-25%
86
15 g/L
nonsoluble-
nonreactive
Sweeney
et al. (2004);
Frederick
et al. (2002)
Dimethyl
Sulfate
Nasal tissue
tumors
NAd
100
28 g/L
moderately
soluble-
reactive
Sarangapani
et al. (2004)
Propylene
Oxide
RE hyperplasia
and OE
degeneration
11-23%
68
590 g/L
moderately
soluble-
reactive
Csanady
et al. (2007);
Morris (2004)
AUptake - measured percent of inspired vapor that is retained or deposited in the URT of the rat
BPhysicochemical Descriptor from Medinsky and Bond (2001) (see text for details)
CRE = respiratory epithelium; OE = olfactory epithelium
dNA = not available
However, these generalizations do not always hold. For example, propylene oxide has
less nasal uptake than might be predicted based solely on its water solubility and/or reactivity
and comparable liquid/tissue:air P to that of acrolein, thus highlighting its hazard categorization.
Conversely, the nasal uptake of vinyl acetate is greater than might be predicted based on those
same two properties as well as its relatively low liquid/tissue:air P. In the case of vinyl acetate,
nasal metabolism via carboxylesterase greatly enhances its uptake into nasal tissues and also
September 2009
3-30
-------
plays a role in its toxicity. Yet, the primary toxicity induced by both of these gases is damage to
the nasal epithelium. In addition, it is critical to note that some high uptake gases can be
systemically distributed via the circulation after absorption in nasal tissues just as low uptake
gases absorbed primarily in the lungs can be delivered to the nose. These observations highlight
the complexity of interspecies dose extrapolation for inhalation as well as the limitations in the
application of a strict categorization or descriptor scheme. Therefore, it is important that
dosimetry extrapolations/calculations should be based on the effect and the target tissue, and not
based solely on physicochemical properties.
Using the physicochemical descriptors scheme proposed by Medinsky and Bond (2001),
an attempt was made to characterize or classify the selected gases based on their criteria as
discussed in Section 3.2. For the majority of the chemicals shown in Table 3-5, application of
these descriptors appears straightforward. For example, formaldehyde and acrylic acid are highly
soluble and/or reactive with tissue components with little potential for systemic distribution,
have high liquid/tissue:air P values, and high uptake in the nasal cavity. Acrolein, acetaldehyde,
and diacetyl are also best described as soluble and/or reactive with some potential for systemic
distribution because of their water solubility, moderate to high liquid/tissue:air P values, and
moderate to high uptake in the nasal cavity. Hydrogen sulfide is best described as non-soluble
based on its low water solubility, low liquid/tissue:air P and low uptake, and nonreactive based
on its hypothesized mode of action - inhibition of cyotchrome oxidase due to competitive
binding. On the other hand, ethyl acrylate is best described as non-soluble based on its low water
solubility, moderate liquid/tissue:air P, and low uptake, and non-reactive as its toxicity is
mediated via metabolism to acrylic acid. However, similar to the 1994 RfCMethods, a set of
qualitative descriptors cannot capture the impact of multiple, interacting quantitative properties,
thus highlighting the limitations in applying a strict categorization or descriptor scheme. For
example, in Table 3-5, propylene oxide and dimethyl sulfate could best be characterized as
moderately-soluble and reactive based on their moderate to high water solubility, moderate
liquid/tissue:air P values and ability to react directly with tissue macromolecules such as
glutathione and form DNA adducts, although their nasal uptake is either unknown or low. In the
case of propylene oxide, nasal uptake is lower than might be expected for a water soluble and
reactive chemical because it is not particularly reactive compared to a chemical such as acrolein
which has a similar solubility and liquid/tissue:air P. As discussed above, vinyl acetate also
stands out for a number of reasons. It is relatively non-soluble as a result of its low solubility and
partition coefficient and is non-reactive. Both its toxicity mediated by being a reactive aldehyde
and its acid metabolites and higher than expected uptake is due to the presence of
carboxylesterase in nasal tissues. Further adding to the complexity and limitations of applying
such a scheme in a strict manner is the consideration of metabolism. As noted in Section 3.2,
September 2009
3-31
-------
metabolism is also considered a component of reactivity in characterizing gas transport in the
tissue and predicting site of gas uptake and effect, but based on the examples provided by
Medinsky and Bond (2001), metabolism appears to be excluded in designating the parent gas as
"reactive".
As discussed in the previous section, two parameters integral to these models shown are
Kg and one of its components kg. Kg, the overall mass transfer coefficient, describes the
movement of gas from the air phase into the liquid or tissue phase by combining kg, the gas
phase mass transfer coefficient, with a liquid or mucus phase transfer coefficient, kg or kmc. The
CFD-PBPK model developed by Bush et al. (1998) and expanded upon by Frederick et al.
(1998) describes and highlights the importance of these parameters in the basic model structure
for estimating target tissue dose for a wide range of inhaled gases in regions of the nasal cavity
for different exposure scenarios. The model was initially used to evaluate the rodent nasal
deposition of several poorly metabolized gases, but was further validated using the
physicochemical and toxicity properties information for acrylic acid. Because this model
provides the basis for the approach used for several of the gases presented here, the hybrid
modeling results for acrylic acid (Frederick et al., 1998) are presented below for illustrative
purposes.
The modeling results for acrylic acid demonstrated several findings regarding
interspecies inhalation target tissue dosimetry. First, the CFD simulations provided estimates of
the volume of the airflow through the various regions of the rat and human nasal cavities at
various flow rates. These data confirmed the results observed in other studies (and discussed
above) showing that a relatively small fraction of inspired air ventilates the human olfactory
region compared to the rat. Second, the CFD simulations also showed that where the data can be
compared, the regional kg values for the rat are higher (up to one to two orders of magnitude in
the respiratory epithelium) than those for the human (Table 3-6) for all of the gases listed. Other
data from these studies indicate that this difference is maintained over a range of ventilation
rates. This difference in kg values indicates that rat nasal cavity is much more efficient in
scrubbing gas from inspired air than the human. Since the kg values for the rat are so much
greater, the rat nasal uptake of gases may be described as limited by tissue phase resistance (i.e.
the rate of liquid phase diffusion dominates the extent of uptake in this species as shown in Bush
et al. 1998). Also as a result, these higher kg values for the rat lead to higher Kg values compared
to the human. The remarkably high animal to human overall Kg ratios for the highly water
soluble and reactive (and therefore highly scrubbed) gases formaldehyde and acrylic acid also
reflects this difference in scrubbing efficiency. Third, the available data demonstrate that the
anterior half of the rat nasal cavity is also more efficient in scrubbing organic gases than the
entire human nasal cavity. On a regional basis in the nasal cavity, this interspecies difference in
September 2009
3-32
-------
the delivery of inspired gases in the overall nasal cavity is significant due to differences in air
flow patterns and distribution of target epithelium. Fourth, unlike the rat model that was found to
be sensitive to variation in the liquid/tissue:air P and liquid phase diffusivity, the human model
was sensitive to variation in air phase parameters. Therefore, air phase resistance to uptake of
inhaled vapors is a significant factor that limits uptake in the human nasal cavity. Overall, the
output from these CFD-PBPK hybrid simulations indicates that the target tissue (olfactory
epithelium) in the rat nasal cavity is exposed to two- to threefold greater concentrations of
acrylic acid than human under similar exposure conditions (i.e., equal external exposure
concentration and equivalent species specific ventilatory flow rates). Further model refinement
for acrylic acid by Andersen et al. (2000) indicate that a DAF of approximately 1.3 be applied to
determine the HEC based on target tissue dose at the NOAEL for rats of 25 ppm (see Table 3-7).
The data in Table 3-6 is complementary to Tables 3-5 and 3-7. Table 3-6 compiles the kg
and a calculated Kg as described below for each of the gases listed. The manner in which the
individual kg values were generated is also documented in this table. For formaldehyde, the kg
for mucus-coated epithelium was calculated based on an empirically derived kg for nonmucus-
coated epithelium using human epidermal tissue; these values were used in all subsequent model
simulations for the rat and human (Kimbell et al., 2001b). For acrolein, however, an
approximation of a simplified mass transfer rate was derived by scaling the kg derived for
nonmucus-coated epithelium for formaldehyde based on differences in diffusivity and partition
coefficient; this value was used for both the rat and human nasal CFD model (Schroeter et al.,
2008). For acrylic acid, kg values for both respiratory epithelial (RE) and olfactory epithelial
(OE) region surface were derived directly from rat and human CFD simulations (Section 3.5.2).
Several other examples in this table also used this same approach, employing the CFD derived kg
values from Frederick et al. (1998).
For illustrative and comparative purposes, a Kg value for each gas for both the RE and
OE was calculated by applying the kg values derived from Frederick et al. (1998) and the
chemical specific parameters outlined in Equations 11 and 12. The animal (rat):human Kg for
each gas and for each nasal tissue region resulting from these calculations are presented in Table
3-6. For example, the rat and human OE Kg values for acetaldehyde were calculated using the
data from Frederick et al. (1998) where: 1/ kg in the rat = 0.00012 cm/min, and 1/ kg in the
human = 0.00014 cm/min. For the OE, a tmuc value of 0.001 cm for the rat, and a value of 0.002
cm for the human were used. A partition coefficient (P) of 224 and a liquid diffusivity (D) value
of 0.000846 cm2/min for acetaldehyde were also used. Use of these values resulted in an animal
(rat):human Kg for the OE of 362:185 and thus, an approximately twofold higher rat Kg. It should
be noted that comparisons based on calculated Kg values alone only provide an indication of
September 2009
3-33
-------
qualitative differences in dose estimates and not actual quantitative estimates of target tissue
dose.
Several aspects of Table 3-7 warrant further comment. One basic observation is that the
default DAF for the ET region is approximately 0.2 for each of the gases whereas DAF values
based on modeling are different for nearly every gas, over a range of sevenfold in this group.
These differences are indications of the models' capacity to employ and integrate numerous
critical gas- and species-specific parameters and variables in characterizing gas transport through
the air and tissue phases for the ET region. For example, the models for ethyl acrylate and
dimethyl sulfate indicate that a DAF of 3 or 7, respectively, be applied to the rat POD to
determine the HEC. The model predicted DAF for each gas is based on more detailed dose
metrics: for ethyl acrylate, internal metabolite concentration; and for dimethyl sulfate, specific
DNA adduct concentration. The fact that quantitative differences exist in the DAF values
estimated for the ET region is an indication that the comparative dosimetry is sensitive to some
combination of these parameters and variables. On the other hand, the default method is de facto
restricted in its use of the relationship between just two general parameters, Ve and SAEt, to
characterize gas transport through the air and tissues phases. Another observation from Table 3-
7 is that the DAF values from modeling are all one or greater despite the wide range of gas
descriptions and characteristics shown in Table 3-5, including uptake (11 - >90%), water
solubility (5 - 1,000 g/L), and tissue:air partition coefficient values (<3 to >6000). Additionally,
these modeled outputs of DAF > 1 were achieved through a similarly wide range of dose metrics
including those based on maximum target tissue flux and/or maximum target tissue flux at the rat
NOAEL, target tissue concentration, target tissue metabolite concentration, adduct concentration,
and even changes in intracellular pH. Despite these wide ranges of sensitive parameters and
variables, and gas descriptors ranging from "soluble- reactive", to "nonsoluble-reactive" to
"nonsoluble-nonreactive", the gases in Table 3-7 all achieved the same internal target tissue dose
in both rats and humans at either similar (DAF ~ 1) or greater for humans (DAF >1) external
concentration.
As is the case with all "state-of-the-science" techniques and approaches, limitations and
restrictions need be considered, including those of the CFD-PBPK hybrid models presented and
described here. These hybrid models may be considered somewhat limited in their refinement of
typical PBPK models relative to the linked CFD model. The surface area characterized by the
PBPK tissue "stack" ( e.g., see Figure 3-11) is much less refined and defined than the flux values
to that same area characterized by the linked CFD model. The gas flux portion of the hybrid
model for this area is represented by a localized kg into which some of the flux has been
collapsed and incorporated. Therefore, small localized areas of very high flux may be diluted or
not sufficiently characterized especially for gases exhibiting high flux, i.e., gases that are highly
September 2009
3-34
-------
reactive and that have high uptake. Conversely, this refinement limitation would be less
applicable to gases that are not highly reactive and that have lower uptake. In addition, due to
the common practice of having and utilizing a single constant value for partition coefficients,
these models are most appropriately used under exposure conditions that do not approach the
limits of solubility and the concomitant establishment of biphasic conditions. In general, these
conditions would be those within the linear range of solubility that also allow for maximization
of all clearance processes, i.e. low level, long term exposures. Thus, although the currently
available CFD-PBPK models are considered to provide better estimates of target tissue dose
compared to conventional default methods, they may also be characterized as providing more
certainty for relatively nonreactive gases versus highly reactive gases and for lower rather than
higher concentrations of reactive gases. Improvements and advances in refinement of CFD-
PBPK hybrid models remain in the future.
September 2009
3-35
-------
Table 3-6. Animal (rat) to human ratios of modeled or calculated gas Kg and kg values
for each representative gas
AGas phase MTC - kg
BCalculated Overall
MTC - K, A/H
References
Formaldehyde
Same values for Animal and Human
used: k - nonmucous = 0.41 cm/s;
k - mucous = 4.7 cm/s determined by
uptake simulations
RE: 46
OE: 1.2
Kimbell et al. (2001b)
Acrylic Acid
Regional kg values estimated using
CFD: Rat - RE 333 cm/s, OE 134 cm/s;
Human - RE 6.4 cm/s, OE 117 cm/s
RE: 16
OE: 1-1.6
Frederick et al. (1998);
Andersen et al. (2000)
Acrolein
Same values used for Rat and Human:
mass transfer rate0 for nonmucus-coated
epithelium scaled from Kimbell et al.
2001b by a factor of 362
RE: 1.3-2.3
OE: 1-2
Schroeter et al. (2008);
Morris et al. (1998)
Acet aldehyde
Based on Frederick 1998 values:
Rat - RE 333 cm/s, OE 134 cm/s;
Human-RE 6.4 cm/s, OE 117 cm/s
RE: 2-3
OE: 1-2
Teegaurden et al.
(2008); Dorman et al.
(2008); Morris et al.
(1997)
Diacetyl
Used Frederick 1998 values:
Rat - RE 333 cm/s, OE 134 cm/s;
Human-RE 6.4 cm/s, OE 117 cm/s
RE: 1.7-2.7
OE: 1-2
Morris et al. (2009)
Vinyl Acetate
Based on Frederick 1998 values:
Rat - RE 333 cm/s, OE 134 cm/s;
Human-RE 6.4 cm/s, OE 117 cm/s
RE: 1.1-2
OE: 1-2
Bogdanffy et al. (1999)
Hydrogen
Sulfide
Calculated only for the rat and combined
for the entire nasal cavity: 2.13 cm/s at a
flow rate of 150 ml/min
RE: 1-2
OE: 1 - 2
Schroeter et al. (2006)
Ethyl Acrylate
Used Frederick 1998 values:
Rat - RE 333 cm/s, OE 134 cm/s;
Human-RE 6.4 cm/s, OE 117 cm/s
RE: 1.2-2.2
OE: 1-2
Sweeney et al. (2004);
Frederick et al. (2002)
Propylene Oxide
Not reported
RE: 1.2-2.2
OE: 1-2
Csanady et al. (2007);
Morris (2004)
Dimethyl Sulfate
Based on Frederick 1998 values:
Rat - RE 4.3 cm/h, OE 4.8 cm/h;
Human - RE 1.8 cm/h, OE 1.2 cm/h
RE: 1.2-2.2
OE: 1-2
Sarangapani et al.
(2004)
AGas phase mass transport coefficient - MTC (kg) as reported or the method referenced for model input
BCalculated Overall MTC (Kg) animal (rat)/human ratio (A/H) comparisons for respiratory epithelium (RE) and
olfactory epithelium (OE) shown - based on Frederick et al. (1998) (Equations 11 and 12; see text for details)
cSimplified scaled mass transfer rate used; not a true kg.
September 2009
3-36
-------
Table 3-7. Comparison of approaches for calculating the DAF for representative gases
in determining the HEC - portal of entry ET or nasal effects
aV/SAet
bcfd
cCFD-PBPK
hybrid
CPBPK
References
Formaldehyde
HEC = 0.2 *
AEL
DAF = 1.26 (based
on the target tissue
max flux of 2620
in Rat and 2082 in
Human at 1 ppm)
Kimbell et al.
(2001b)
Acrylic Acid
HEC = 0.2 *
AEL
DAF = 1.36
(based on target
tissue dose at
the Rat NO AEL
of 25 ppm)
Frederick et al.
(1998);
Andersen et al.
(2000)
Acrolein
HEC = 0.2 *
AEL
DAF = 1.4
(based on Rat OE
NOAEL=
0.6 ppm: highest
flux of 682 in Rat
and 476 in Human)
Schroeter et al.
(2008);
Morris et al.
(1998)
Acet aldehyde
HEC = 0.2 *
AEL
DAF = 1.4
(based on
steady-state
tissue
concentrations
at the Rat
NO AEL = 50
ppm
Teegaurden
et al. (2008);
Dorman et al.
(2008); Morris
et al. (1997)
Diacetyl
HEC = 0.2 *
AEL
DAF = 1 (based
on nasal and
tracheal
target tissue
concentration)
Morris et al.
(2009)
Vinyl Acetate
HEC = 0.2 *
AEL
DAF = 1.14
(based on
equivalent
change in OE
intracellular pH
at the Rat
NOAEL)
Bogdanffy
et al. (1999)
Hydrogen
Sulfide
HEC = 0.2 *
AEL
DAF = 2.6 (based
on Rat OE
NOAEL=
10 ppm: highest
flux of 34 in Rat
and 13 In Human)
Schroeter et al.
(2006)
EthylAcrylate
HEC = 0.2 *
AEL
DAF = 3 (based
on target
internal
metabolite
concentration
Sweeney et al.
(2004);
Frederick et al.
(2002)
September 2009
3-37
-------
aV/SAet
bcfd
cCFD-PBPK
hybrid
CPBPK
References
Propylene
Oxide
HEC = 0.2 *
AEL
DAF = 1 (based
on equivalent
concentrations in
RE and venous
blood at <
50ppm)
Csanady et al.
(2007); Morris
(2004)
Dimethyl
Sulfate
HEC = 0.2 *
AEL
RE: DAF = 7;
OE: DAF = 2
(based on tissue
N7mG adduct
concentration)
Sarangapanni
et al. (2004)
A Calculated based on procedures in U.S. EPA (1994) RfC Methodology where:
HEC = DAF (RGDR) x Adjusted Exposure Level (AEL - based on NOAEL, LOAEL, or BMCLx)
DAF or RGDR = /SAET.animalA''e /SAET_human
DAFet = (0.18 L/min/15 cm2)/(13.8 L/min/200 cm2) = 0.18 or 0.2
(SA|.T_amm(l = 15 cm2; VE = 0.18 L/minfor a 250 g rat; SAET_human = 200 cm2; VE = 13.8 L/minfor a 70
kg human)
B Results from CFD simulation modeling - DAF based on comparative animal (rat): human flux values as
indicated
c Results from CFD-PBPK hybrid or PBPK modeling - DAF based on modeled target tissue dose or dose metric
as indicated
September 2009
3-38
-------
The relationship of the air phase and tissue resistance terms in determining Kg (see
Equations 11 and 12) is further evaluated in Figures 3-12 and 3-13. Presented in these two
figures are plots of the calculated Kg, for both the respiratory epithelium (RE) and olfactory
epithelium (OE) in animals and humans, versus the partition coefficients for the gases presented
in Tables 3-5 and 3-6. These plots demonstrate the influence of the partition coefficient on Kg;
the differences in Kg in each tissue and between animals and humans; and thus, the differences in
kg in each tissue and between animals and humans. In both nasal tissues, gases with higher
partition coefficients had higher calculated Kg values. For each gas, the calculated Kg in the rat is
greater than in the human mainly because the regional kg values determined from the CFD
simulations are greater in the rat. The greatest differences between the rat and human were seen
in the RE and for gases with the highest liquid/tissue:air P values (i.e. formaldehyde and acrylic
acid). For gases such as these, uptake in the nasal cavity tends to be higher because the gas phase
mass transfer coefficient (kg) predominates the determination of Kg. Gases with lower partition
coefficients which tend to have lower nasal uptake, such as hydrogen sulfide, since the tissue
phase resistance predominates the determination of Kg. The output from the dosimetry models
for the gases presented in Table 3-7 support the relationship of these parameters and of Kg in that
the calculated rat/human target tissue dose (i.e. DAF) is > or ~ 1.
September 2009
3-39
-------
10000-
1000:
• Rat
V Human
s
o
5
100-
DMS
PO
7
• n
Aald
Vv Acr
EA
V
DA
V
AA
V
FA
10:
W
H2s
VA
-1 1 1 1—I—I I I I |
10
-1—i—i—i i i i | 1 1 1—i—i—i i i i |
100 1000
Partition Coefficient
-1 1—i—i—i i i i | 1 1 1—i—i—i i i i |
10000 100000
Figure 3-12. Plot of Kg (overall MTC) for respiratory epithelium (RE) in rats and humans
versus partition coefficient for the gases shown in Tables 3-5, 3-6, and 3-7. MTC calculated
as described by Frederick et al. (1998) (Equations 11 and 12; see text for details). (H2S -
Hydrogen Sulfide, VA - Vinyl Acetate, PO - Propylene Oxide, E A - Ethyl Acrylate, DMS -
Dimethyl Sulfate, Acr - Acrolein, Aald - Acetaldehyde, DA - Diacetyl, AA - Acrylic Acid,
FA - Formaldehyde).
September 2009
3-40
-------
10000:
1000-
• Rat
V Human
V
AA
V
FA
100-
10-
DMS
V
VA
V
Aald
PO
^ Acr
V?
EA
7
DA
V
H2s
-1 1 1—I—I—I I I I I 1 1 1—I—I—I I I I I 1 1 1—I—I—I I I I I 1 1 1—I—I—I I I I I 1 1 1—I—I—I I I I I
10
100 1000
Partition Coefficient
10000
100000
Figure 3-13. Plot of Kg (overall MTC) for olfactory epithelium (OE) in rats and humans
versus partition coefficient for the gases shown in Tables 3-5, 3-6, and 3-7. MTC calculated
as described by Frederick et al. (1998) (Equations 11 and 12; see text for details). (H2S -
Hydrogen Sulfide, VA - Vinyl Acetate, PO - Propylene Oxide, E A - Ethyl Acrylate, DMS -
Dimethyl Sulfate, Acr - Acrolein, Aald - Acetaldehyde, DA - Diacetyl, AA - Acrylic Acid,
FA - Formaldehyde).
The interspecies dosimetry modeling results for the selected gases presented in Table 3-7
indicate that for the ET region, the dose (i.e., HEC) to animals is either greater (up to sevenfold)
compared to humans or close to unity. Comparison of these approaches demonstrates this point
whether the dosimeter is quantitative (e.g., based on target tissue flux, target tissue concentration
September 2009
3-41
-------
of parent or metabolite) or qualitative (e.g. based on overall or regional mass transfer coefficients
- Kg and kg) over a range of gases having differing solubility, reactivity, uptake, and partition
coefficients.
3.5.4. Summary and Conclusions - Modeling Developments and Results
The overall goal of this section is to summarize and put into context critical modeling
developments that have occurred in the fundamental areas of target tissue dosimetry of gases.
These developments should inform users of the utility and inherent limitations in the existing
default RfC Methods which employ dose estimates from VE and SA. The general aspects to
consider are air phase transport (related to Ve in the default approach) and disposition of gases
into the liquid/tissue phase (related to SA).
The capacity of CFD techniques to solve and describe air phase behavior in complex
geometries represented by different species has been clearly demonstrated. Further, this
technique has been repeatedly shown to be capable of indicating potential exposure (through
flux) of target tissues of the ET region at a highly refined level with considerable accuracy, much
more so than the use of VE and SA ratios as a basic default procedure.
The development of PBPK models to describe the toxicokinetic behavior of gas flow and
disposition into the tissue phase relating to SAEt has been illustrated and demonstrated to be
realistic. Model construction of tissue stacks with underlying blood flow for simulating gas flow
into and through the tissues subtending the areas exposed to CFD-estimated flux has been shown
to correspond with observations.
The capabilities of CFD and PBPK models to describe these individual phases have been
integrated in a quantitative manner with CFD- PBPK hybrid modeling approaches. The
development of this approach through use of a "permeability" coefficient (i.e. overall mass
transport coefficient, Kg) to combine elements of flux from CFD modeling and of permeability
from PBPK modeling in defining the diffusion of the gas through the air and tissue phases has
been demonstrated. For target tissue dose, CFD- PBPK hybrid models represent a best available
model for air and tissue phase elements of target tissue dosimetry that the Ve/SAEt surrogate
attempts to approximate. Finally, this section demonstrates the application of the best available
model of tissue dosimetry, the CFD-PBPK hybrid models. These published applications
consistently indicate that interspecies target tissue doses (human:animal) relative to external
exposure are close to or greater than 1:1.
This outcome of the analysis for the ET region is graphically summarized in Figure 3-14.
This figure may be viewed as a follow-on and complementary representation of Figures 3-1 and
September 2009
3-42
-------
3-10 which provided a backdrop for the information presented in this Chapter. The assumptions
of uniformity for the ET region listed in Figures 3-1 and 3-10 have been altered in Figure 3-15 to
better represent the information presented in this section - namely, that air flow and gas
deposition to surface areas in the ET region are nonuniform. Also, CFD-PBPK hybrid models
have been entered in that area concerned with outcomes to the assumptions of uniformity of the
ET region shown in Figures 3-1 and 3-10. This entry reflects the capability of these modeling
approaches to integrate and apply conditions of nonuniform gas behavior. The far right of Figure
3-14 now indicates the outcome predicted by these models as shown by the examples presented
in this section. The examples presented here generally indicate that the ET regions of animals
and humans receive nearly equivalent doses at equivalent external exposure concentrations. For
some gases, however, the dose to the ET regions of animals is estimated to be three- to sevenfold
greater than of humans.
ET
Geometry
& Anatomy
ET Lesions
&
Flux-lesion correlation
Dye-flow
ET Models
I CFD I
I ET Models I
nonuniform uneven flow of gas in VE
nonuniform uneven distribution of
tissue types
nonuniform uneven denosition over SA
4
CFD PBPK
Hybrid Models
Rat Jf
}o<
™ Human^
lx dose
nr < 1 x rinse
Figure 3-14. A revised schematic representation of the outcomes for interspecies inhalation
dosimetry of gases for the ET region following from the advances presented.
The analysis, summary, and conclusions presented in this report on the state-of-the-
science advances in inhalation dosimetry are intended to serve as a foundation for potential
revisions to gas dosimetry procedures in the RfCMethods. Future reports will review advances
in gas dosimetry for other regions of the respiratory tract. This report relates to methods for
interspecies extrapolation and not to differences in dosimetry among individuals in the human
population (intraspecies variability). This report is intended neither to displace the current
September 2009
3-43
-------
procedures for gas dosimetry in RfC Methods nor to reflect on use of these procedures in
assessments currently in IRIS.
September 2009
3-44
-------
4. REFERENCES
List of specific references cited in the text:
Aharonson EF; Menkes H; Gurtner G; Swift DL; and Proctor DF. (1974). Effect of respiratory
airflow rate on removal of soluble vapors by the nose. J Appl Physiol, 37(5), 654-657.
Andersen M; Sarangapani R; Gentry R; Clewell H; Covington T; and Frederick CB. (2000).
Application of a hybrid CFD-PBPK nasal dosimetry model in an inhalation risk
assessment: an example with acrylic acid. Toxicol Sci, 57(2), 312-325.
Bailie N; Hanna B; Watterson J; and Gallagher G. (2006). An overview of numerical modelling
of nasal airflow. Rhinology, 44(1), 53-57.
Bogdanffy MS; Sarangapani R; Plowchalk DR; Jarabek A; and Andersen ME. (1999). A
biologically based risk assessment for vinyl acetate-induced cancer and noncancer
inhalation toxicity. Toxicol Sci, 51(1), 19-35.
Bush ML; Frederick CB; Kimbell JS; and Ultman JS. (1998). A CFD-PBPK hybrid model for
simulating gas and vapor uptake in the rat nose. Toxicol Appl Pharmacol, 150(1),
133-145.
Bush ML; Zhang W; Ben-Jebria A; and Ultman JS. (2001). Longitudinal distribution of ozone
and chlorine in the human respiratory tract: simulation of nasal and oral breathing with
the single-path diffusion model. Toxicol Appl Pharmacol, 173(3), 137-145.
Casanova M; Morgan KT; Steinhagen WH; Everitt JI; Popp JA; and Heck HD. (1991). Covalent
binding of inhaled formaldehyde to DNA in the respiratory tract of rhesus monkeys:
pharmacokinetics, rat-to-monkey interspecies scaling, and extrapolation to man. Fundam
Appl Toxicol, 17(2), 409-428.
Chang JC; Gross EA; Swenberg JA; and Barrow CS. (1983). Nasal cavity deposition,
histopathology, and cell proliferation after single or repeated formaldehyde exposures in
B6C3F1 mice and F-344 rats. Toxicol Appl Pharmacol, 68(2), 161-176.
Chometon F; Gillieron P; Laurent J; Ebbo D; Kos-man F; Lecomte F; and Sorrel-Dejerine N.
(2000). Aerodynamics of nasal airways with application to obstruction. Ann Otolaryngol
Chir Cervicofac 117:98-104
Corley RA. (2009). Magnetic Resonance Imaging and Computational Fluid Dynamics (CFD)
Simulations of Rabbit Nasal Airflows for the Development of Hybrid CFD/PBPK
Models Inhalation Toxicol, 1-7.
Csanady GA; and Filser JG. (2007). A physiological toxicokinetic model for inhaled propylene
oxide in rat and human with special emphasis on the nose. Toxicol Sci, 95(1), 37-62.
Dahl AR. (1990). Dose concepts for inhaled vapors and gases. Toxicol Appl Pharmacol, 103(2),
185-197.
de Rochefort L; Vial L; Fodil R; Maitre X; Louis B; Isabey D; Caillibotte G;Thiriet M; Bittoun
J; Durand E; and Sbirlea-Apiou G. (2007). In vitro validation of computational fluid
dynamic simulation in human proximal airways with hyperpolarized 3He magnetic
resonance phase-contrast velocimetry. J Appl Physiol, 102(5), 2012-2023.
September 2009
4-1
-------
Dorman DC; Struve MF; Wong BA; Gross EA; Parkinson C; Willson GA; Tan, YM; Campbell,
JL;Teeguarden, JG;Clewell, H J; and Andersen, ME. (2008). Derivation of an inhalation
reference concentration based upon olfactory neuronal loss in male rats following
subchronic acetaldehyde inhalation. Inhal Toxicol, 20(3), 245-256.
Elad DL R; Wenig BL; Einav S. (1993). Analysis of Air Flow Patterns in the Human Nose.
Medical and Biological engineering and Computing, 31, 585-592.
Frederick CB; Bush ML; Lomax LG; Black KA; Finch L; Kimbell JS; Morgan, KT;
Subramaniam, RP; Morris, JB; Ultman, JS. (1998). Application of a hybrid
computational fluid dynamics and physiologically based inhalation model for interspecies
dosimetry extrapolation of acidic vapors in the upper airways. Toxicol Appl Pharmacol,
152(1), 211-231.
Frederick CB; Lomax LG; Black KA; Finch L; Scribner HE; Kimbell JS; Morgan, KT;
Subramaniam, RP; and Morris, JB.(2002). Use of a hybrid computational fluid dynamics
and physiologically based inhalation model for interspecies dosimetry comparisons of
ester vapors. Toxicol Appl Pharmacol, 183(1), 23-40.
Geelhaar A; and Weibel ER. (1971). Morphometric estimation of pulmonary diffusion capacity.
3. The effect of increased oxygen consumption in Japanese Waltzing mice. Respir
Physiol, 11(3), 354-366.
Gehr P; Mwangi DK; Ammann A; Maloiy GM; Taylor CR; and Weibel ER. (1981). Design of
the mammalian respiratory system. V. Scaling morphometric pulmonary diffusing
capacity to body mass: wild and domestic mammals. Respir Physiol, 44(1), 61-86.
Gross EA; Swenberg JA; Fields S; and Popp JA. (1982). Comparative morphometry of the nasal
cavity in rats and mice. J Anat, 135(Pt 1), 83-88.
Guilmette RA; W JD; Wolff RK. (1989). Morphometry of human nasal airways in vivo using
magnetic resonance imaging. J Aerosol Med, 2, 365-377.
Hanna LM; Frank R; and Scherer PW. (1989). Absorption of soluble gases and vapors in the
respiratory system. In H. K. a. P. Chang, M. (Ed.), Respiratory physiology: an analytical
approach, (pp. 277-316). New York, NY: Marcel Dekker, Inc.
ICRP. (1993). Gases and vapours. In I. C. o. R. Protection (Ed.), Human respiratory tract model
for radiological protection: a report of Committee 2 of the International Commission on
Radiological Protection, (pp. 57-66). Sutton, Surrey, United Kingdom: International
Commission on Radiological Protection.
Kepler GM; Joyner DR; Fleishman A; Richardson RB; Gross EA; Morgan KT; Godo MN; and
Kimbell JS. (1995). Method for obtaining accurate geometrical coordinates of nasal
airways for computer dosimetry modeling and lesion mapping. Inhal. Toxicol.
7:1207-1224.
Kepler GM; Richardson RB; Morgan KT; and Kimbell JS. (1998). Computer simulation of
inspiratory nasal airflow and inhaled gas uptake in a rhesus monkey. Toxicol Appl
Pharmacol, 150(1), 1-11.
Keyhani KS; and Mozell MM. (1995). Numerical Simulation of Airflow in the Human Nasal
Cavity. J Biomemechanical Engineering, 117, 429-441.
September 2009
4-2
-------
Kimbell JS; Godo MN; Gross EA; Joyner DR; Richardson RB; and Morgan KT. (1997a).
Computer simulation of inspiratory airflow in all regions of the F344 rat nasal passages.
Toxicol Appl Pharmacol, 145(2), 388-398.
Kimbell JS; Gross EA; Joyner DR; Godo MN; and Morgan KT. (1993). Application of
computational fluid dynamics to regional dosimetry of inhaled chemicals in the upper
respiratory tract of the rat. Toxicol Appl Pharmacol, 121(2), 253-263.
Kimbell JS; Gross EA; Richardson RB; Conolly RB; and Morgan KT. (1997b). Correlation of
regional formaldehyde flux predictions with the distribution of formaldehyde-induced
squamous metaplasia in F344 rat nasal passages. Mutat Res, 380(1-2), 143-154.
Kimbell JS; Overton JH; Subramaniam RP; Schlosser PM; Morgan KT; Conolly RB; and Miller
FJ. (2001a). Dosimetry modeling of inhaled formaldehyde: binning nasal flux predictions
for quantitative risk assessment. Toxicol Sci, 64(1), 111-121.
Kimbell JS; and Subramaniam RP. (2001). Use of computational fluid dynamics models for
dosimetry of inhaled gases in the nasal passages. Inhal Toxicol, 13(5), 325-334.
Kimbell JS; Subramaniam RP; Gross EA; Schlosser PM; and Morgan KT. (2001b). Dosimetry
modeling of inhaled formaldehyde: comparisons of local flux predictions in the rat,
monkey, and human nasal passages. Toxicol Sci, 64(1), 100-110.
Kliment V. (1973). Similarity and dimensional analysis, evaluation of aerosol deposition in the
lungs of laboratory animals and man. Folia Morphol. (Prague) 21, 59-64.
Kurtz DB; Zhao K; Hornung D. E; and Scherer P. (2004). Experimental and numerical
determination of odorant solubility in nasal and olfactory mucosa. Chem Senses, 29(9),
763-773.
Lechner AJ. (1978). The scaling of maximal oxygen consumption and pulmonary dimensions in
small mammals. Respir Physiol, 34(1), 29-44.
Martonen TB; Zhang Z; Yu G; and Musante CJ. (2001). Three-dimensional computer modeling
of the human upper respiratory tract. Cell Biochem Biophys, 35(3), 255-261.
Medinsky MA; and Bond JA. (2001). Sites and mechanisms for uptake of gases and vapors in
the respiratory tract. Toxicology, 160(1-3), 165-172.
Mercer RR; Russell ML; and Crapo JD. (1994a). Alveolar septal structure in different species.
J Appl Physiol, 77(3), 1060-1066.
Mercer RR; Russell ML; Roggli VL; and Crapo JD. (1994b). Cell number and distribution in
human and rat airways. Am J Respir Cell Mol Biol, 10(6), 613-624.
Miller FJ; Overton JH, J; Jaskot RH; and Menzel DB. (1985). A model of the regional uptake of
gaseous pollutants in the lung. I. The sensitivity of the uptake of ozone in the human lung
to lower respiratory tract secretions and exercise. Toxicol Appl Pharmacol, 79(1), 11-27.
Minard KR; Einstein DR; Jacob RE; Kabilan S; Kuprat AP; Timchalk CA; Trease LL; and
Corley RA. (2006). Application of magnetic resonance (MR) imaging for the
development and validation of computational fluid dynamic (CFD) models of the rat
respiratory system. Inhal Toxicol, 18(10), 787-794.
September 2009
4-3
-------
Morgan KT; Kimbell JS; Monticello TM; Patra AL; and Fleishman A. (1991). Studies of
inspiratory airflow patterns in the nasal passages of the F344 rat and rhesus monkey using
nasal molds: relevance to formaldehyde toxicity. Toxicol Appl Pharmacol, 110(2),
223-240.
Morgan MS; Frank R. (1977). Uptake of pollutant gases by the respiratory system. In J. D. P.
Brain, D. F.; Reid, L. M; (Ed.), Respiratory defense mechanisms: part I. (pp. 157-189).
New York, NY: Marcel Dekker, Inc.
Morris JB. (1997). Uptake of acetaldehyde vapor and aldehyde dehydrogenase levels in the
upper respiratory tracts of the mouse, rat, hamster, and guinea pig. Fundam Appl Toxicol,
35(1), 91-100.
Morris JB. (1998). Effect of acrolein vapor on upper respiratory tract uptake of acetaldehyde.
Inhal Toxicol, 10, 843-856.
Morris JB; Banton MI; and Pottenger LH. (2004). Uptake of inspired propylene oxide in the
upper respiratory tract of the f344 rat. Toxicol Sci, 81(1), 216-224.
Morris JB; Hassett DN; and Blanchard KT. (1993). A physiologically based pharmacokinetic
model for nasal uptake and metabolism of nonreactive vapors. Toxicol Appl Pharmacol,
123(1), 120-129.
Morris JB; and Hubbs AF. (2009). Inhalation dosimetry of diacetyl and butyric acid, two
components of butter flavoring vapors. Toxicol Sci, 108(1), 173-183.
Morris JB; and Smith FA. (1982). Regional deposition and absorption of inhaled hydrogen
fluoride in the rat. Toxicol Appl Pharmacol, 62(1), 81-89.
Moulin FJ; Brenneman KA; Kimbell JS; and Dorman DC. (2002). Predicted regional flux of
hydrogen sulfide correlates with distribution of nasal olfactory lesions in rats. Toxicol
Sci, 66(1), 7-15.
NRC (1994). Science and Judgment in Risk Assessment. Committee on Risk Assessment of
Hazardous Air Pollutants Board on Environmental Studies and Toxicology
Commission on Life Sciences National Research Council, Washington, D.C; National
Academy Press,
NRC (2009). Science and Decisions: Advancing Risk Assessment. Committee on Improving
Risk Analysis Approaches Used by the U.S. EPA, National Research Council;
Washington, D.C; National Academy Press,
Overton JH; Graham RC; and Miller FJ. (1987). A model of the regional uptake of gaseous
pollutants in the lung. II. The sensitivity of ozone uptake in laboratory animal lungs to
anatomical and ventilatory parameters. Toxicol Appl Pharmacol, 88(3), 418-432.
Phalen, RF; Stuart, BO; and Lioy, PJ. (1988) Rationale for and implications of particle size-
selective sampling. In: Advances in air sampling: [papers from the ACGIH symposium];
February 1987; Pacific Grove, CA. Chelsea, MI: Lewis Publishers, Inc.; p. 6. (Industrial
hygiene science series).
Ramsey JC; and Andersen ME. (1984) a physiologically based description of the inhalation
pharmacokinetics of styrene in rats and humans. Toxicol. Appl. Pharmacol 73:159-1 75
September 2009
4-4
-------
Sarangapani R; and Wexler AS. (2000). The Role of Dispersion in Particle Deposition in Human
Airways. Tox Sci 54, 229-236
Sarangapani R; Teeguarden JG; Gentry PR; Clewell HJ, 3rd Barton HA; and Bogdanffy MS.
(2004) Interspecies dose extrapolation for inhaled dimethyl sulfate: a PBPK model-based
analysis using nasal cavity N7-methylguanine adducts. Inhal Toxicol 16: 593-605.
Schreck S; Sullivan KJ; Ho CM; and Chang HK. (1993). Correlations between flow resistance
and geometry in a model of the human nose, J. Appl. Physiol. 75 (4) (1993), pp. 1767-
1775.
Schreider JP; and Hutchens JO. (1980). Morphology of the guinea pig respiratory tract. The
Anatomical Record 196(3):313-321
Schroeter JD; Kimbell JS; Andersen ME; and Dorman DC. (2006). Use of pharmacokinetic-
driven computational fluid dynamics model to predict nasal extraction of hydrogen
sulfide in rats and humans. Toxicol. Sci. 94:359-367
Schroeter JD; Kimbell JS; Gross EA; Willson GA; Dorman DC; Tan Y-M; and Clewell HJ.
(2008). Application of physiological computational fluid dynamics models to precit
interspecies nasal dosimetry of inhaled acrolein. Inhal Toxicol 20:227-243
Shome B; Wang LP; Santare MN; Prasad AK; Szeri AZ; and Roberts D (1998) Modeling of the
Airflow in the Pharynx with Application to Sleep Apnea. Amer. Soc. Mechanical
Engineers 120: 416-422.
Subramanian RP; Richardson RB; Morgan KT; Kimbell JS; and Guilmette RA. (1998).
Computational fluid dynamics simulations of inspiratory airflow in the human nose and
nasopharynx. Inhalation Toxicology, 10, 91-120.
Sweeney LM; Andersen ME; and Gargas ML. (2004). Ethyl acrylate risk assessment with a
hybrid computational fluid dynamics and physiologically based nasal dosimetry model.
Toxicol Sci, 79(2), 394-403.
Tarabichi M; and Fanous N. (1993). Finite element analysis of airflow in the nasal valve. Arch
Otolaryngol Head Neck Surg, 119(6), 638-642.
Teeguarden JG; Bogdanffy MS; Covington TR; Tan C; and Jarabek AM. (2008). A PBPK model
for evaluating the impact of aldehyde dehydrogenase polymorphisms on comparative rat
and human nasal tissue acetaldehyde dosimetry. Inhal Toxicol, 20(4), 375-390.
Tenney SM; and Remmers JE. (1963). Comparative quantitative morphology of the mammalian
lung: diffusing area. Nature, 197, 54-56.
U.S. EPA (1988a). Reference physiological parameters in pharmacokinetic modeling. Arms AD;
Travis CC. U.S. Environmental Protection Agency, Office of Research and Development,
Office of Health and Environmental Assessment, Washington, DC; EPA/600/S6-88/004
U.S. EPA. (1988b). Recommendations for and documentation of biological values for use in risk
assessment. U.S. Environmental Protection Agency, Washington, DC; EPA/600/6-
87/008.
September 2009
4-5
-------
U.S. EPA (1994). Methods for derivation of inhalation reference concentrations (RfCs) and
application of inhalation dosimetry. U.S. Environmental Protection Agency, Office of
Research and Development, Office of Health and Environmental Assessment,
Washington, DC; EPA/600/8-90/066F.
U.S. EPA. (2006) Science policy council handbook. 3rd edition. Office of Science Policy
Council, Office of Research and Development, Washington, DC; EPA/100/B-06/002.
Wen JI; Tu J; and Wang S. (2008). Numerical Simulations for Detailed Airflow Dynamics in a
Human Nasal cavity. Resp Physiol neurobiology, 161(2), 125-135.
Yang GC; Scherer PW; and Mozell MM. (2007). Modeling inspiratory and expiratory steady-
state velocity fields in the Sprague-Dawley rat nasal cavity. Chem Senses, 32(3), 215-
223.
Yu CP; and Xu GB. (1987). Predictive models for deposition of inhaled diesel exhaust particles
in humans and laboratory species. Res Rep Health Eff Inst(10), 3-22.
Zhao K; Dalton P; Yang GC; and Scherer PW. (2006). Numerical modeling of turbulent and
laminar airflow and odorant transport during sniffing in the human and rat nose. Chem
Senses, 31(2), 107-118.
Zhao K; and Dalton P. (2007). The Way the Wind Blows: Implications of Modeling Nasal
Airflow. Current Allergy and Asthma reports, 7, 117-125.
Zhao K; Scherer PW; Hajiloo S; and Dalton P. (2004). Effect of Anatomy on Human Nasal Air
Flow and Odorant Transport patterns: Implications for Olfaction. Chem Senses, 29(5),
365-379.
Other source references related to inhalation gas dosimetry:
Aitken, M. L., Franklin, J. L., Pierson, D. J., and Schoene, R. (1986). Influence of body size and
gender on control of ventilation. J. Appl. Physiol. 60:1894-1899.
Andersen ME; Clewell HJ, III; and Frederick CB. (1995). Applying simulation modeling to
problems in toxicology and risk assessment—a short perspective. Toxicol Appl
Pharmacol, 133(2), 181-187.
Andersen ME; Green T; Frederick CB; and Bogdanffy MS. (2002). Physiologically based
pharmacokinetic (PBPK) models for nasal tissue dosimetry of organic esters: assessing
the state-of-knowledge and risk assessment applications with methyl methacrylate and
vinyl acetate. Regul Toxicol Pharmacol, 36(3), 234-245.
Andersen, ME; and Sarangapani, R. (1999). Clearance concepts applied to the metabolism of
inhaled vapors in tissues lining the nasal cavity. Inhal Toxicol, 11(10), 873-897.
Andersen ME; Sarangapani R; Frederick CB; and Kimbell JS. (1999). Dosimetric adjustment
factors for methyl methacrylate derived from a steady-state analysis of a physiologically
based clearance-extraction model. Inhal Toxicol, 11(10), 899-926.
September 2009
4-6
-------
Arens R; McDonough JM; Corbin AM; Hernandez ME; Maislin G; Schwab RJ; and Pack AI.
(2002). Linear dimensions of the upper airway structure during development: assessment
by magnetic resonance imaging. Am J Respir Crit Care Med, 165(1), 117-122.
Arens R; McDonough JM; Corbin AM; Rubin NK; Carroll ME; Pack AI; Liu J; and Udupa JK.
(2003). Upper airway size analysis by magnetic resonance imaging of children with
obstructive sleep apnea syndrome. Am J Respir Crit Care Med, 167(1), 65-70.
Basak SC; Mills D; Hawkins DM; and El-Masri HA. (2003). Prediction of human blood: air
partition coefficient: a comparison of structure-based and property-based methods. Risk
Anal, 23(6), 1173-1184.
Beals, J. A., Funk, L. M., Fountain, R., and Sedman, R. (1996). Quantifying the distribution of
inhalation exposure in human populations: distribution of minute volumes in adults and
children. Environ. Health Perspect. 104:974-979.
Bennett WD; Zeman KL; and Jarabek AM. (2003). Nasal contribution to breathing with
exercise: effect of race and gender. J Appl Physiol, 95(2), 497-503.
Bogdanffy MS. (2002). Vinyl acetate-induced intracellular acidification: implications for risk
assessment. Toxicol Sci, 66(2), 320-326.
Bogdanffy MS; and Jarabek AM. (1995). Understanding mechanisms of inhaled toxicants:
implications for replacing default factors with chemical-specific data. Toxicol Lett,
82-83, 919-932.
Bogdanffy MS; and Sarangapani R. (2003). Physiologically-based kinetic modeling of vapours
toxic to the respiratory tract. Toxicol Lett, 138(1-2), 103-117.
Bogdanffy MS; Sarangapani R; Kimbell JS; Frame SR; and PlowchalkDR. (1998). Analysis of
vinyl acetate metabolism in rat and human nasal tissues by an in vitro gas uptake
technique. Toxicol Sci, 46(2), 235-246.
Bogdanffy MS; and Valentine R. (2003). Differentiating between local cytotoxicity, mitogenesis,
and genotoxicity in carcinogen risk assessments: the case of vinyl acetate. Toxicol Lett,
140-141, 83-98.
Burton RT; Isaacs KK; Fleming JS; and Martonen TB. (2004). Computer reconstruction of a
human lung boundary model from magnetic resonance images. Respir Care, 49(2),
180-185.
Clewell HJ; Andersen ME; and Barton HA. (2002). A consistent approach for the application of
pharmacokinetic modeling in cancer and noncancer risk assessment. Environ Health
Perspect, 110(1), 85-93.
Clewell HJ; Teeguarden J; McDonald T; Sarangapani R; Lawrence G; Covington T; Genry R;
and Shipp A. (2002). Review and evaluation of the potential impact of age- and gender-
specific pharmacokinetic differences on tissue dosimetry. Crit Rev Toxicol, 32(5), 329-
389.
Condorelli P; and George SC. (1999). Theoretical Gas Phase Mass Transfer Coefficients for
Endogenous Gases in the lungs. Annals of Biomedical Engineering, 27, 326-339.
September 2009
4-7
-------
Conolly RB; Kimbell JS; Janszen, D; Schlosser PM; Kalisak D; Preston J; and Miller FJ. (2003).
Biologically motivated computational modeling of formaldehyde carcinogenicity in the
F344 rat. Toxicol Sci, 75(2), 432-447.
Conolly RB; Kimbell JS; Janszen D; Schlosser PM; Kalisak D; Preston J; and Miller FJ. (2004).
Human respiratory tract cancer risks of inhaled formaldehyde: dose-response predictions
derived from biologically-motivated computational modeling of a combined rodent and
human dataset. Toxicol Sci, 82(1), 279-296.
da Costa GC; Churchwell MI; Hamilton LP; von Tungeln LS; Beland FA; Marques MM; and
Doerge DR. (2003). DNA Adduct Formation from Acrylamide via Conversion To
Glycidamide in Adult and Neonatal Mice. Chem Res Toxicol, 16, 1328-1337.
Dahl AR; and Gerde P. (1994). Uptake and metabolism of toxicants in the respiratory tract.
Environ Health Perspect, 102 Suppl 11, 67-70.
Daigle CC; Chalupa DC; Gibb FR; Morrow PE; Oberdorster G; Utell MJ; and Frampton MW.
(2003). Ultrafine particle deposition in humans during rest and exercise. Inhal Toxicol,
15(6), 539-552.
de Zwart LL; Haenen HE; Versantvoort CH; Wolterink G; van Engelen JG; and Sips AJ. (2004).
Role of biokinetics in risk assessment of drugs and chemicals in children. Regul Toxicol
Pharmacol, 39(3), 282-309.
Deloar HM; Watabe H; Nakamura T; Narita Y; Yamadera A; Fujiwara T; and Itoh M. (1997).
Internal dose estimation including the nasal cavity and major airway for continuous
inhalation of CI 502, 1502 and CI 50 using the thermoluminescent dosimeter method. J
Nucl Med, 38(10), 1603-1613.
Dorman DC; Brenneman KA; McElveen AM; Lynch SE; Roberts KC; and Wong BA. (2002).
Olfactory transport: a direct route of delivery of inhaled manganese phosphate to the rat
brain. J Toxicol Environ Health A, 65(20), 1493-1511.
Faller TH; Csanady GA; Kreuzer PE; Baur CM; and Filser JG. (2001). Kinetics of propylene
oxide metabolism in microsomes and cytosol of different organs from mouse, rat, and
humans. Toxicol Appl Pharmacol, 172(1), 62-74.
Gargas ML; Tyler TR; Sweeney LM; Corley RA; Weitz KK; Mast TJ; Paustenbach DJ; and
Hays SM. (2000). A toxicokinetic study of inhaled ethylene glycol monomethyl ether (2-
ME) and validation of a physiologically based pharmacokinetic model for the pregnant
rat and human. Toxicol Appl Pharmacol, 165(1), 53-62.
Gentry PR; Covington TR; Andersen ME; and Clewell HJ, 3rd. (2002). Application of a
physiologically based pharmacokinetic model for isopropanol in the derivation of a
reference dose and reference concentration. Regul Toxicol Pharmacol, 36(1), 51-68.
Gentry PR; Covington TR; and Clewell HJ, 3rd. (2003). Evaluation of the potential impact of
pharmacokinetic differences on tissue dosimetry in offspring during pregnancy and
lactation. Regul Toxicol Pharmacol, 38(1), 1-16.
Gentry PR; Covington TR; Clewell HJ, 3rd; and Anderson ME. (2003). Application of a
physiologically based pharmacokinetic model for reference dose and reference
concentration estimation for acetone. J Toxicol Environ Health A, 66(23), 2209-2225.
September 2009
4-8
-------
Georgieva AV; Kimbell JS; and Schlosser PM. (2003). A distributed-parameter model for
formaldehyde uptake and disposition in the rat nasal lining. Inhal Toxicol, 15(14),
1435-1463.
Gerde P; and Dahl AR. (1991). A model for the uptake of inhaled vapors in the nose of the dog
during cyclic breathing. Toxicol Appl Pharmacol, 109(2), 276-288.
Gerde P; Muggenburg BA; Lundborg M; and Dahl AR. (2001). The rapid alveolar absorption of
diesel soot-adsorbed benzo[a]pyrene: bioavailability, metabolism and dosimetry of an
inhaled particle-borne carcinogen. Carcinogenesis, 22(5), 741-749.
Gerde P; and Scott BR. (2001). A model for absorption of low-volatile toxicants by the airway
mucosa. Inhal Toxicol, 13(10), 903-929.
Gerrity TR. (1995). Regional deposition of gases and particles in the lung: implications for
mixtures. Toxicology, 105(2-3), 327-334.
Ginsberg GL; Asgharian B; Kimbell JS; Ultman JS; and Jarabek AM. (2008). Modeling
approaches for estimating the dosimetry of inhaled toxicants in children. Journal of
Toxicology and Environmental Health, Part A, 71: 166-195.
Ginsberg G; Hattis D; Miller R; and Sonawane B. (2004). Pediatric pharmacokinetic data:
implications for environmental risk assessment for children. Pediatrics, 113(4 Suppl),
973-983.
Ginsberg G; Hattis D; Russ A; and Sonawane B. (2004). Physiologically based pharmacokinetic
(PBPK) modeling of caffeine and theophylline in neonates and adults: implications for
assessing children's risks from environmental agents. J Toxicol Environ Health A, 67(4),
297-329.
Ginsberg G; Hattis D; and Sonawane B. (2004). Incorporating pharmacokinetic differences
between children and adults in assessing children's risks to environmental toxicants.
Toxicol Appl Pharmacol, 198(2), 164-183.
Green T; Toghill A; and Foster JR. (2001). The role of cytochromes P-450 in styrene induced
pulmonary toxicity and carcinogenicity. Toxicology, 169(2), 107-117.
Guilmette RA; C YS; and Griffith WC. (1997). Characterising the Variability in Adult Human
airway Dimensions. Ann Occup Hyg, 41 (Supplement 1), 491-496
Haddad S; Beliveau M; Tardif R; and Krishnan K. (2001). A PBPK modeling-based approach to
account for interactions in the health risk assessment of chemical mixtures. Toxicol Sci,
63(1), 125-131.
Hanna LM; Lou SR; Su S; and Jarabek AM. (2001). Mass transport analysis: inhalation rfc
methods framework for interspecies dosimetric adjustment. Inhal Toxicol, 13(5),
437-463.
Harkema JR; Carey SA; and Wagner JG. (2006). The nose revisited: a brief review of the
comparative structure, function, and toxicologic pathology of the nasal epithelium.
Toxicol Pathol, 34(3), 252-269.
September 2009
4-9
-------
Hinderliter PM; Thrall KD; Corley RA; Bloemen LJ; and Bogdanffy MS. (2005). Validation of
human physiologically based pharmacokinetic model for vinyl acetate against human
nasal dosimetry data. Toxicol Sci, 85(1), 460-467.
Hofmann W; Asgharian B; Bergmann R; Anjilvel S; and Miller FJ. (2000). The effect of
heterogeneity of lung structure on particle deposition in the rat lung. Toxicol Sci, 53(2),
430-437.
Hubal EA; Fedkiw PS; and Kimbell JS. (1996). Mass-transport models to predict toxicity of
inhaled gases in the upper respiratory tract. J Appl Physiol, 80(4), 1415-1427.
Hubal EA; SchlosserPM; Conolly RB; and Kimbell JS. (1997). Comparison of inhaled
formaldehyde dosimetry predictions with DNA-protein cross-link measurements in the
rat nasal passages. Toxicol Appl Pharmacol, 143(1), 47-55.
International Commission on Radiological Protection (ICRP). 1994. Publication 66. Human
Respiratory Tract Model for Radiological Protection. Ann. ICRP 24. Oxford: Pergamon
Press
Jarabek AM. (1995). Interspecies Extrapolations Based on Mechanistic determinants of
Chemical Disposition. Human and Ecological Risk Assessment, 1(5), 641-662.
Kelly JT; Prasad AK; and Wexler AS. (2000). Detailed flow patterns in the nasal cavity. J Appl
Physiol, 89(1), 323-337.
Kimbell JS; Segal RA; Asgharian B; Wong BA; Schroeter JD; Southall JP; Dickens CJ; Brace
G; and Miller FJ. (2007). Characterization of deposition from nasal spray devices using a
computational fluid dynamics model of the human nasal passages. J Aerosol Med.
20(l):59-74.
Kohn MC. (1997). The importance of anatomical realism for validation of physiological models
of disposition of inhaled toxicants. Toxicol Appl Pharmacol, 147(2), 448-458.
Layton, D. W. (1993). Metabolically consistent breathing rates for use in dose assessments.
Health Phys. 64:23-36.
Lippmann M; and Schlesinger RB. (1984). Interspecies comparisons of particle deposition and
mucociliary clearance in tracheobronchial airways. J Toxicol Environ Health, 13(2-3),
441-469.
Litman RS; Weissend EE; Shrier DA; and Ward DS. (2002). Morphologic changes in the upper
airway of children during awakening from propofol administration. Anesthesiology,
96(3), 607-611.
Martonen TB; Schroeter JD; and Fleming JS. (2007). 3D in silico modeling of the human
respiratory system for inhaled drug delivery and imaging analysis. J Pharm Sci.
96(3):603-17.
Morris JB. (1999). A method for measuring upper respiratory tract vapor uptake and its
applicability to quantitative inhalation risk assessment. Inhal Toxicol, 11(10), 943-965.
Morris JB. (2000). Uptake of styrene in the upper respiratory tract of the CD mouse and
Sprague-Dawley rat. Toxicol Sci, 54(1), 222-228.
September 2009
4-10
-------
Morris JB. (2001). Overview of upper respiratory tract vapor uptake studies. Inhal Toxicol,
13(5), 335-345.
Morris JB. (2002). Sensory nerve-mediated nasal vasodilatory response to inspired ethyl
acrylate. Inhal Toxicol, 14(6), 585-597.
Nodelman V; and Ultman JS. (1999). Longitudinal distribution of chlorine absorption in human
airways: a comparison to ozone absorption. J Appl Physiol, 87(6), 2073-2080.
Oberdorster G. (1992). Lung Dosimetry: Extrapolation Modeling from Animals to man. Nickel
and Human health: Current perspectives, 32, 421-436.
Overton JH; Kimbell JS; and Miller FJ. (2001). Dosimetry modeling of inhaled formaldehyde:
the human respiratory tract. Toxicol Sci, 64(1), 122-134.
Pelekis M; GephartLA; and Lerma, SE. (2001). Physiological-model-based derivation of the
adult and child pharmacokinetic intraspecies uncertainty factors for volatile organic
compounds. Regul Toxicol Pharmacol, 33(1), 12-20.
Plowchalk DR; Andersen ME; and Bogdanffy MS. (1997). Physiologically based modeling of
vinyl acetate uptake, metabolism, and intracellular pH changes in the rat nasal cavity.
Toxicol Appl Pharmacol, 142(2), 386-400.
Rigas ML; Catlin SN; Ben-Jebria A; and Ultman JS. (2000) Ozone uptake in the intact human
respiratory tract: relationship between inhaled dose and actual dose. J Appl Physiol 88:
2015-2022.
Santiago LY; Hann MC; Ben-Jebria A; and Ultman JS. (2001) Ozone absorption in the human
nose during unidirectional airflow. J Appl Physiol 91: 725-732.
Sarangapani R; Teeguarden JG; Cruzan G; Clewell HJ; Andersen ME. (2002) Physiologically
based pharmacokinetic modeling of styrene and styrene oxide respiratory-tract dosimetry
in rodents and humans. Inhal Toxicol 14: 789-834.
Sarangapani R; Clewell HJ; Cruzan G; and Andersen ME. (2002) Comparing respiratory-tract
and hepatic exposure-dose relationships for metabolized inhaled vapors: a
pharmacokinetic analysis. Inhal Toxicol 14: 835-854.
Sarangapani R; Gentry PR; Covington TR; Teeguarden JG; and Clewell HJ. (2003) Evaluation
of the potential impact of age- and gender-specific lung morphology and ventilation rate
on the dosimetry of vapors. Inhal Toxicol 15: 987-1016.
Schlesinger RB. (1995) Interaction of gaseous and particulate pollutants in the respiratory tract:
mechanisms and modulators. Toxicology 105: 315-325.
Schroeter JD. (2009). The use of nasal dosimetry models in the risk assessment of inhaled gases.
Toxicol Sci. 108(1): 1-3.
Tardif R; Charest-Tardif, G; Brodeur J; and Krishnan K. (1997). Physiologically based
pharmacokinetic modeling of a ternary mixture of alkyl benzenes in rats and humans.
Toxicol Appl Pharmacol, 144(1), 120-134.
September 2009
4-11
-------
Teeguarden JG; Deisinger PJ; Poet TS; English C; Faber WD; Barton HA; Corley RA; and
Clewell HJ. (2005). Derivation of a human equivalent concentration for n-butanol using a
physiologically based pharmacokinetic model for n-butyl acetate and metabolites n-
butanol and n-butyric acid. Toxicol Sci, 85(1), 429-446.
September 2009
4-12
-------
APPENDIX A. INDEPENDENT EXTERNAL PEER REVIEW - SUMMARY AND
DISPOSITION OF COMMENTS
The "Status Report: Advances in Inhalation Dosimetry of Gases with Portal of Entry
Effects in the Upper Respiratory Tract" has undergone a formal external peer letter review
performed by scientists in accordance with EPA guidance on peer review (EPA, 2006b). The
reviewers were tasked with providing written answers to charge questions on both general and
specific scientific aspects of the report. A summary of significant comments made by the
external reviewers to these charge questions and EPA's responses to these comments arranged
by charge question follow. Editorial comments were considered and incorporated directly into
the document as appropriate.
EXTERNAL PEER REVIEWER COMMENTS
Comments and Response to Charge:
Charge Question 1: The primary focus of this report relates to the pharmacokinetic component
of interspecies gas dosimetry for portal of entry effects in the upper respiratory tract. Issues
related to pharmacodynamics, including variability in response, are specifically excluded from
this report. Is the scope and primary focus of this report clear?
Comments:
All three reviewers agreed that the scope and focus of the report are clear. Two of the reviewers
elaborated that the report was well organized and maintains a clear focus on the issue of
interspecies dosimetry for gases and vapors in the upper respiratory tract.
Response:
No response necessary.
Charge Question 2: Have the principal studies examining interspecies gas dosimetry for portal
of entry effects in the upper respiratory tract that have appeared since the issuance of the 1994
RfCMethods been identified in this report? Please identify and provide a rationale for any other
key studies that should be considered for inclusion.
Comments:
Two of the three reviewers thought the report includes and provides a detailed description of all
of the key studies examining interspecies gas dosimetry for portal of entry effects in the upper
respiratory tract. One of these reviewers further elaborated that the Medinsky and Bond (2001)
paper put the topic of describing gases on terms of their physicochemical properties into
perspective. This reviewer also commented that the majority of the important works describing
airflow were included and described sufficiently to support the conclusions reached. However,
this reviewer listed two more recent references that could be included in Section 3.4.6 to support
September 2009
A-l
-------
the correlation of flux and lesions incidence. These two reviewers provided additional source
references specifically addressing dosimetry in children and ventilation rates among human
populations that could be added to the reference list, but did not specifically state that they
needed to be described as they would not change the conclusions of the report.
The third reviewer noted that not all the source references were cited in the text.
Response:
The reference list is divided into those publications that are cited in the report, and other source
references related to inhalation dosimetry. The suggested references which address dosimetry in
children and ventilation rates among human populations have been added to the reference
sections as other sources (Ginsberg GL et al., 2008; Schroeter, 2009; ICRP, 1994; Aitken et al.,
1986; Layton, 1993; and Beals et al., 1996). Some text describing two other suggested references
(Schroeter et al., 2008; and Morris et al., 2009) for inclusion into Section 3.4.6 has also been
added. This structure delineates the nature of reference materials and addresses the reviewer
comments.
Charge Question 3: The state-of-the-science pertaining to the focus of this report is primarily
presented in Chapter 3. Is the description of those studies in this report concerned with gas
dosimetry appropriate and accurate? Are the analyses and evaluations of the scientific evidence
supported by the studies cited?
Comments:
One of the three reviewers stated he was impressed by the consistency of the cited and described
studies. This reviewer further stated that the analysis in Chapter 3 of the report is accurate and
the conclusions are appropriate.
Another of the three reviewers commented that the major conclusions from the CFD models with
respect to nasal airflow patterns are well-described. The major conclusions related to gas
dosimetry from these studies were also described appropriately and in-depth in this report. In
particular, this reviewer stated that Table 3-7 provides compelling evidence that the evaluations
in this report are supported by the literature cited, noting that the DAF is equal or greater than 1
for each compound listed in this table.
One of the three reviewers thought that the focus of the report was too narrow with respect to
discussion and analyses of dosimetry experiments specifically in areas related to the
development and application of mass transport principles to all airways. This reviewer also cited
the lack of inclusion of information the criteria gases ozone, sulfur dioxide, and nitrogen dioxide,
and that Table 3-5 should be expanded to include these gases. This reviewer did state, however,
that the information presented does support the use of Equation 11 of the report.
Response:
September 2009
A-2
-------
The intent of this report was to focus on presenting the advances in inhalation interspecies
dosimetry for the URT noting that similar evaluations and analyses will be conducted for other
regions of the respiratory tract as well as for extrarespiratory/systemic regions. The intent of this
report was not to provide an exhaustive comparative list of all the chemicals/compounds that
could potentially be included in a report such as this. Specifically, the criteria gases - while
providing large databases to examine for potential comparisons across species - do not rely on
the 1994 RfC Methods to estimate dose. These pollutants rely on actual human dosimetry and
concentration-response relationships in assessing risk and health, not on extrapolations from
animal studies. The animal and human dosimetry information for the criteria pollutants primarily
deal with regions other than the URT. Thus, a chemical such as ozone should provide
information for evaluating interspecies extrapolation for other regions of the respiratory tract.
Charge Question 4: The primary focus of this report relates to the pharmacokinetic component
of interspecies gas dosimetry for portal of entry effects in the upper respiratory tract. Is this
report successful in presenting and evaluating the state-of-the-science relating to this focus?
Comments:
Two of the three reviewers commented that the report was successful in presenting and
evaluating the state-of-the-science related to interspecies upper respiratory tract inhalation
dosimetry, specifically the pharmacokinetic component. The presentation and evaluation of the
key studies and findings was considered accurate and objective, and the quality of the synthesis
of the new information and the analysis of its implications for cross-species dosimetry of the
upper respiratory tract was stated as impressive. One of these reviewers further commented that
this report presents an excellent summary of the numerous models that have been developed
since the mid-1990s, including CFD and CFD-PBPK models, and their application in
interspecies inhalation dosimetry.
One of the three reviewers commented that the theoretical basis of the RfC Methodology
presented in Chapter 2 does not provide an adequate perspective for the material in Chapter 3.
The reason being that the ET default presented in the 1994 RfC document was preceded in the
RfC document by a one-dimensional model in which Kg appeared as a parameter - equivalent to
the Aharonson model. This reviewer thought it would be helpful to present the Aharonson
equation and its theoretical limitations in Chapter 2.
Response:
The purpose of Chapter 2 was to provide a brief review and summary of the major applications
of the 1994 RfC Methods. In practice, the RfC derivation is made from applying the default
dosimeter of Ve/SA without use of the kg term as it often not known as discussed in the
methodology. Although the work of Aharonson is cited in the RfC Methods, it is not described in
detail. Thus, for the purposes of this report, addition of the Aharonson approach would be
historically interesting but such inclusion is not necessary as a foundation for Chapter 3.
However, much of the theory supporting the development of the hybrid modeling that was
initially presented in Appendix I of the 1994 RfC Methods is outlined in Chapter 3.
September 2009
A-3
-------
Charge Question 5: The latter part of the report deals extensively with advanced state-of-the-
science dosimetry models including CFD and the CFD-PBPK hybrid models. The strengths and
limitations of these models to estimate target-tissue are highlighted in Chapter 3. However, as
with any state-of-the-science approach, limitations exist with respect to application and outcome
of these models - and some of which have been recognized and discussed (see in particular
Section 3.5.3). Have the strengths and limitations of these advanced models been sufficiently
characterized? Is there any information that would further support as well as limit the overall
conclusions drawn from the results of these models and, more broadly, the new science
published since creation of RfC Methods?
Comments:
One of the three reviewers stated that the report sufficiently characterizes the strengths and
limitations of the CFD and CFD-PBPK models relied upon for the analysis. In addition, this
reviewer commented that the power and general acceptance of these modeling approaches is
exemplified by their use in the development of inhalation pharmaceuticals and provided
references to be added to the source reference list.
Another of the three reviewers commented that the report clearly states the strengths of these
models and their usefulness in estimating DAFs for inhaled gases in the URT. Additionally, as
related to kg values in Table 3-6, is that different methods were used to compute kg values. For
example, the kg for formaldehyde was calculated using empirical results while the kg values used
in CFD-PBPK models (typically from Frederick et al., 1998) were derived from CFD models.
This reviewer also provided specific commentary in regards to the dosimetry models in two
sections of the report as well as the overall conclusion. In Section 3.4.6: It is stated that a
limitation of the CFD models used in flux versus lesion correlations is that they only include the
URT - this reviewer commented that this isn't really a limitation of the models per se, but is a
convenient use of URT models when pathology is observed in the URT. In Section 3.5, a
comprehensive summary of interspecies nasal dosimetry results using CFD or CFD-PBPK
models for a multitude of gases is presented with the primary conclusion that, in all cases, a DAF
equal to or greater than 1 is achieved. This reviewer commented that this is in sharp contrast to
results from the current RfC Methods, which using default approaches, arrives at a DAF of 0.2
with the implication that since the chemicals considered here cover a wide range of gas
descriptions that these results can be extrapolated to all gases. Thus, one limitation is that since
these models are chemical-specific, one can only infer that similar results would be expected for
other compounds.
One of the three reviewers commented that the impact of different physical-chemical properties
of a gas on mathematically modeling and understanding its dosimetry needs further clarification
and that the classification scheme in Figure 3-2 helps in this regard. This reviewer further
elaborated that it is not clear how the two different solubility parameters given in table 3-5,
namely, liquid/air partition coefficients and water solubility or some combination of the two
should be specifically used in the classification scheme in Figure 3-2. In addition, this review
does not understand why the definition of reactive gas in the report appears to exclude metabolic
reaction.
September 2009
A-4
-------
Response:
The suggested references (Kimbell et al., 2007; and Martonen et al., 2007) regarding the
modeling approaches as applied in pharmaceuticals were added to the source reference list.
The statement made in Section 3.4.6 on the limitation of CFD models, flux, and lesion
correlation was removed.
The intent of discussing the Medinsky and Bond (2001) physicochemical descriptor scheme was
not to imply that it directly impacts dosimetry modeling. The intent was to show how this
particular descriptor scheme may provide an improvement in examining the relationship between
these physicochemical parameters and sites of toxicity versus a strict "numerical categorization"
that may potentially influence default dosimetry application. Thus, the aspect of Figure 3-2 that
improves the understanding of dosimetry is the de-emphasis of the overly simplified and
proscriptive categorization scheme described in the 1994 RjCMethods. Figure 3-2 instead
allows for the realization of the continuous nature of the named variables (reactivity and
solubility) and the uncertainty and shortcomings inherent in this scheme. These points are made
in the text of the Report.
The point regarding the two different solubility parameters given in Table 3-5 is valid, especially
regarding implementation of such models as it is likely that partition coefficients would either
change or become increasingly unstable under exposure conditions approaching the limits of
solubility. Text concerning this point has been added to Section 3.5.3.
The definition of reactivity does not exclude metabolism as discussed in Section 3.2. However,
in the examples provided in Medinsky and Bond (2001), a gas that is metabolized to a toxic
moiety is described as "nonreactive" highlighting a limitation of this scheme. Text discussing
this issue of metabolism and reactivity has been added to Section 3.2 and 3.5.3.
Other Reviewer Comments:
One reviewer was uncomfortable with the attempt to use the categorization suggested by
Medinsky and Bond (2001) as a modified gas/vapor dosimetry scheme (p. 3-2 to 3-4 and 3-28 to
3-31). The reviewer commented that the Medinsky and Bond (2001) categorization suffers from
the same limitation as the 1994 RfC Methods: water solubility and chemical reactivity alone are
simply not adequate to classify gas/vapor uptake/toxicity. Further, the descriptors should at least
include: water solubility, tissue/air partition coefficient, chemical reactivity, pK, and
susceptibility to metabolism in the URT.
Another of the three reviewers commented on the statement (p 3-28) that "Kg may also be scaled
from one gas to other gases by scaling to the water diffusivity and partition coefficient" is not
entirely correct. This was done in Schroeter et al. (2008) to scale a simplified mass transfer rate
on squamous epithelium from formaldehyde to acrolein. The mass transfer term was not a "Kg"
as described earlier in this section and was also not stated correctly on page 3-33. Also, this was
a specific case for comparing two gases that share similar molecular structures.
September 2009
A-5
-------
Response:
The intent of presenting and discussing the Medinsky and Bond (2001) scheme was to show how
this descriptor scheme may provide an improvement in examining the relationship between these
physicochemical parameters and sites of toxicity compared to a strict "numerical categorization"
that may potentially influence default dosimetry application. Thus, the aspect of this descriptor
scheme that improves the understanding of dosimetry is the de-emphasis of the overly simplified
categorization scheme described in the 1994 RfC Methods. Figure 3-2 allows for the realization
of the continuous nature of the named variables (reactivity and solubility) and the uncertainty
and shortcomings inherent in this scheme. Text highlighting the limitations of this scheme has
been added to section 3.5.3.
In considering the reviewer's comment concerning the statement made on p 3-28, the text was
removed. Similarly, corrections to the text and Tables regarding the "mass transfer term" relating
to acrolein not being a true kg were made.
Other specific text suggestions and corrections were considered and incorporated as appropriate
in revising the draft report.
September 2009
A-6
------- |