United States
Environmental Protection
Agency
Office of Research and
Development
Washington DC 20460
EPA'600/8-91/011A
March 1991
Workshop Review Draft
Interim
*;>
Guidance for
Dermal Exposure
Assessment
Review
Draft
(Do Not
Cite or Quote)
Notice
This document is a preliminary draft. It has not been formally
released by EPA and should not at this stage be construed to
represent Agency policy. It is being circulated for comment on its
technical accuracy and policy implications.
-------
DRAFT EPA/600/8-91/011 A
NOT QUOTE OR CITE March 1991
Workshop Draft
INTERIM GUIDANCE FOR DERMAL EXPOSURE ASSESSMENT
NOTICE
THIS DOCUMENT IS A PRELIMINARY DRAFT. It has not been formally released by the U.S.
Environmental Protection Agency and should not at this stage be construed to represent Agency
policy. It is being circulated for comment on its technical accuracy and policy implications.
Office of Health and Environmental Assessment
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
•DISCLAIMER "
This document is a draft for review purposes only and does not constitute Agency policy.
Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
11
-------
CONTENTS
Page
Disclaimer ii
Contents iii
Tables vi
Figures ix
Foreword x
Preface xi
Authors and Reviewers xii
1.0 INTRODUCTION 1-1
2.0 METHODOLOGIES FOR ESTIMATING DERMAL EXPOSURE 2-1
2.1 Contamination of Environmental Media 2-4
22 Dermal Contact Scenarios 2-4
2.3 Surface Area 2-9
2.3.1 Body Surface Area 2-9
2.3.2 Surface Area of Component Body Parts 2-10
2.3.3 Exposed Surface Area and Clothing 2-11
2.4 Dermal Adherence of Soil 2-11
2.5 Experimental Methods for Estimating Cutaneous Absorption 2-IS
3.0 MECHANISMS OF DERMAL ABSORPTION 3-1
3.1 Structure of the Skin 3-1
32 Fate of Compounds Applied to the Skin 3-4
32.1 Transport Processes Occurring in the Skin 3-5
322 Loss Processes Occurring in the Skin 3-6
3.3 Factors That Influence Percutaneous Absorption 3-10
3.3.1 Skin-Specific Factors 3-10
3.3.2 Compound-Specific Factors 3-21
4.0 TECHNIQUES FOR MEASURING DERMAL ABSORPTION 4-1
4.1 In Vivo Studies 4-3
4.1.1 Quantification of Radioactivity, Parent Compound, or Metabolite Levels in
Excreta 4-4
4.1.2 Quantification of Parent Compound or Metabolite in Blood, Plasma, or
Tissues 4-4
4.1.3 Quantification of the Disappearance of the Compound from the Surface of the
Skin or from the Donor Solution 4-5
4.1.4 Measurement of a Biological Response 4-7
4.1.5. Stripping Method 4-8
42 In Vitro Techniques 4-8
42.1 Diffusion Cells 4-9
A22 Isolated Perfused Tubed-Skin Preparation 4-16
42.3 Stratum Comeum Binding Technique 4-17
4.3 Comparison of In Vitro and In Vivo Percutaneous Absorption Values 4-17
4.4 Interspecies Comparison of Percutaneous Absorption Values 4-20
111
-------
CONTENTS (continued)
5.0 ESTIMATION OF DERMAL ABSORPTION RATE VALUES FOR COMPOUNDS IN
WATER 5-1
5.1 Theoretical Basis of Kp and J Values 5-2
5.2 Dermal (Percutaneous) Absorption Rate Equation Parameters 5-5
53 Evaluation of Kp from Experimental Studies 5-11
53.1 In Vitro Approaches 5-11
53.2 In Vivo Approaches 5-13
5.4 Interconversion Among Kp, J, and Percent Absorbed Values 5-15
5.5 Percutaneous Absorption of Water-Containing Lipophilic Compounds 5-15
6.0 DERMAL ABSORPTION OF COMPOUNDS FROM SOIL 6-1
6.1 Experimentally Derived Values 6-1
62 Factors Affecting the Dermal (Percutaneous) Absorption of Soil-Adhered Components . 6-5
63 Methodologies for Estimating the Dermal Absorption of Soil-Adsorbed Compounds ... 6-10
63.1 Interpreting Existing Data 6-10
63.1.1 TCDD 6-11
63.12 TCB 6-14
63.13 BaP 6-15
63.1.4 DDT 6-17
63.2 Estimating Dermal Uptake When Data Are Lacking 6-18
63.2.1 Use of Structural Analogues 6-19
63.2.2 Use of Values of Percent Applied Dose Absorbed for the Neat
Compound 6-21
63.23 Use of Kp Values for the Compound in Water or Other Fluids 6-22
63.2.4 Theoretical Modeling 6-23
6.23.5 Default Values for Organic Compounds 6-24
63.2.6 Extraction of Contaminants from Soil 6-25
7.0 DERMAL ABSORPTION OF CHEMICAL VAPORS 7-1
7.1 Experimentally Derived Values 7-1
72 Factors Affecting the Dermal Absorption of Vapors 7-6
73 Methodologies For Estimating the Dermal Absorption of Chemical Vapors in the Absence
of Experimental Data 7-7
8.0 METHODS FOR PREDICTING PERMEABILITY COEFFICIENTS OF AQUEOUS
CONTAMINANTS 8-1
8.1 Empirical Correlations 8-2
82 Theoretical Skin Permeation Models 8-11
82.1 Heterogeneous Structural Model 8-11
82.2 Two Parallel Pathway Model 8-12
823 Three Parallel Pathway Model 8-13"
82.4 Albery and Hadgraft Model 8-13
82.5 Kasting, Smith & Cooper Model 8-14
IV
-------
CONTENTS (continued)
9.0 RELATIVE CONTRIBUTION OF DERMAL EXPOSURE TO TOTAL ABSORBED DOSE . . 9-1
9.1 Contribution of Dermal Exposure to the Total Absorbed Dose of Neat Compounds .... 9-1
9.2 Contribution of Dermal Exposure to the Total Absorbed Dose of Compounds in Aqueous
Media 9-2
9.3 Contribution of Dermal Exposure to the Total Absorbed Dose of Chemical
Vapors 9-14
9.4 Contribution of Dermal Exposure to the Total Absorbed Dose of Compounds in the
Soil 9-16
9.5 Summary of Conditions That Enable Dermal Uptake to Become a Significant Route of
Exposure 9-17
10.0 STEPWISE DERMAL EXPOSURE ASSESSMENT PROCESS 10-1
10.1 Contact With Compounds in Aqueous Media 10-2
Step 1. Select Values for Exposure Parameters 10-2
Step 2. Select a Permeability Coefficient for the Compound of Interest 10-3
Step 2a. Experimental Values of Permeability Coefficient 10-3
Step 2b. Consideration of Factors that Affect the Experimental Kp Values .... 10-3
2b.l Experimental Kp from Neat Compounds 10-4
2b.2 Species Differences 10-4
2b.3 Regional Variation 10-5
2b.4 Evaporation and Occlusion 10-5
2b.5 Metabolism 10-5
2b.6 Age of the Skin 10-6
2b.7 Skin Condition 10-6
2b.8 Hydration 10-6
2b.9 Temperature 10-6
2b.lO Study Type 10-6
Step 2c. Estimation of Kp in the Absence of Experimentally Derived Valu es ... 10-7
Step 2d. Model Validation 10-8
Step 3. Integration of the Information to Determine Dermal
Absorption 10-8
10.2 Contact with Compounds in Soil 10-9
Step 1. Select Values for Exposure Parameters 10-9
Step 2. Select a Value for Percent of Applied Dose Absorbed 10-10
Step 3. Calculate the Absorbed Dose of the Soil-Adsorbed Compound 10-12
103 Use of Dermal Absorption Data in Risk Assessment 10-12
11.0 REFERENCES 11-1
Appendix A: RATIONALE FOR SELECTING RECOMMENDED COMPOUND-SPECIFIC KP
VALUES
Appendix B: GLOSSARY
-------
TABLES
Page No.
Table 2-1 Assumptions of Outdoor Soil Exposure Duration .................. 2-6
Table 2-2 Assumptions of Frequency of Exposure to Soil .................... 2-7
Table 2-3 Surface Area by Body Part for Adults (m2) ...................... 2-11
Table 2-4 Total Body Surface Area of Male Children in Square Meters .......... 2-12
Table 2-5 Soil Adherence Values ..................................... 2-15
Table 2-6 Ranges of Possible Values and Recommended Defaults for Dermal
Exposure Factors ......................................... 2-17
Table 3-1 Evidence for the Two-Phase Model of the Stratum Comeum .......... 3-3
Table 3-2 Comparison of Vapor Pressure and Disposition of Radioactivity After
Topical Application of Radiolabeled Control Compounds to Pig Skin
Under Standardized Conditions ............................... 3-7
Table 3-3 Effect of Anatomical Region on In Vivo Percutaneous Absorption of
Pesticides in Humans ...................................... 3-12
Table 3-4 Percutaneous Absorption in Monkeys as Related to Site of Application
and Test Compound ...................................... 3-13
Table 3-5 Effect of Gender and Body Site on the Permeability of Rat Skin ....... 3-14
Table 3-6 Rat Skin Thickness Measurement from Frozen Sections .............. 3-15
Table 3-7 Regional Variation in Stratum Comeum Thickness in Humans ......... 3-15
Table 3-8 In Vitro Percutaneous Absorption of Triclocarban in Human Adult and
Newborn Abdominal and Foreskin Epidermis ..................... 3-16
Table 3-9 Effect of Temperature on Permeability Coefficients for Model
Compounds Permeating Hairless Mouse Skin In Vitro ............... 3-20
Table 3-10 In Vitro Permeability Coefficients and Partition Data for Various Phenol
Compounds ............................................ 3-22
Table 3-11 Permeability of Hairless Mouse Skin to Selected Phenols as Function of
pH .............................. . .................... 3-24
Table 3-12 Permeability of Human Skin (In Vitro) to Alcohols ................. 3-25
Table 3-13 Percutaneous Absorption of Topical Doses of Several Compounds in the
Rhesus Monkey ......................................... 3-27
Table 4-1 Experimental Techniques Used to Obtain Kp or J Values Reported in the
Dermal Permeability Database ............................... 4-2
Table 4-2 Comparison of In Vivo Methods for Determining Mean Bioavailability ... 4-5
Table 4-3 Comparison of the "Direct" and "Indirect" Methods of Dutkiewicz and
Tyras (1967) for Determination of Percutaneous Absorption Rate ....... 4-7
Table 4-4 In Vitro Percutaneous Absorption of Triclocarban in Human Adult
Abdominal Epidermis ..................................... 4-11
Table 4-5 Effect of Receptor Fluid Composition on the Relative Absorption of
Hydrophobic Compounds ................................... 4-12
VI
-------
Table 4-6 Percutaneous Absorption of Radiolabeled Compounds on Pig Skin In
Vitro and In Vivo 4-14
Table 4-7 Influence of the Epidermis on Percutaneous Absorption of Compounds
Through Pig Skin 4-14
Table 4-8 Total Absorption of Various Compounds by Skin in Vivo and in Vitro
(Modified Tests) (Expressed as Percent of Applied Dose) 4-18
Table 4-9 Ranking of the Relative In Vitro Percutaneous Absorption of Different
Species 4-20
Table 4-10 Permeability of Animal Skin Relative to Human Skin 4-21
Table 4-11 Percutaneous Absorption of Nitroaromatic Compounds in Human and
Monkey Skin 4-22
Table 4-12 Relative In Vitro Percutaneous Absorption of Water and Paraquat
Through Human and Animal Skin 4-23
Table 4-13 Summary of Factors that May Affect the Use of Kp Data in Cutaneous
Exposure Assessment 4-24
Table 5-1 Diffusion Coefficients at Infinite Dilution in Water at 25°C 5-8
Table 5-2 Diffusion Coefficients at Infinite Dilution in Nonaqueous Liquids 5-8
Table 5-3 Physiological and Biochemical Parameters Used in This Model 5-16
Table 5-4 Abbreviations Used in Figure 5-2 5-17
Table 6-1 Dermal Absorption of Soil-Adhered Organic Compound 6-6
Table 6-2 Properties of Soil Used in the Skowronski et al. (1988, 1989) Studies .... 6-8
Table 6-3 Dermal Absorption Percentages for Dilute Concentrations of TCDD
in Soil 6-12
Table 6-4 Dermal Absorption Percentages for Dilute Concentrations of TCB
in Soil 6-14
Table 6-5 Dermal Absorption Percentages for Dilute Concentrations of BaP
in Soil (24-Hour Duration of Exposure) 6-16
Table 6-6 Dermal Absorption Percentages for Dilute Concentrations of DDT
in Soil (24-Hour Duration of Exposure) 6-17
Table 7-1 Dermal Vapor Absorption in Rats In Vivo 7-2
Table 7-2 Estimated Human Permeability Constants for Vapor Phase Organic
Compounds 7-3
Table 7-3 Flux Values for Organic Compounds in Permeating Human Skin In Vitro
as a Saturated Vapor and as a Liquid 7-4
Table 7-4 Estimated Permeability Coefficient Values (cm/hr) for Alcohol and
Alkane Saturated Vapors 7-5
Table 7-5 Flux and Permeability Coefficient Values for Permeant Gases in
Humans 7-5
Vll
-------
TABLES (continued)
Table 8-1 Permeability Coefficient for Human Skin (Aqueous Solutions) and
Octanol/Water Partition Coefficients (Neat) of Organic Compounds:
Alphabetical Ordering of Compounds Having Published Permeability
Coefficients 8-4
Table 8-2 Algorithm for Calculating Permeability Coefficients from Octanol/Water
Coefficients 8-8
Table 8-3 Regression Equations Developed by Various Authors 8-10
Table 9-1 Estimation of PEA or EEA Uptake in Man 9-2
Table 9-2 Relative Contribution (%) of Dermal and Oral Exposure to Dose 9-3
Table 9-3 Absorption Constants (Fraction Absorbed) for Various Routes of
Exposure 9-5
Table 9-4 Effect of Drinking Water Concentration on Relative Exposure via All
Routes to a Child's Total Body Burdens in Summer (Rural) 9-6
Table 9-5 Lifetime Equivalent Exposure Factors (Expressed as Percent of Total
Exposure) for Trans- 1,2-Dichloroethylene in Tap Water 9-7
Table 9-6 Relative Contribution of Different Routes of Exposure to the Absorbed
Dose of VOCs in Drinking Water 9-9
Table 9-7 Minimum and Maximum Conditions for Dermal Absorption Defined as
by Brown and Hattis (1989) 9-11
Table 9-8 Classification of Kp Values by Order of Magnitude, Based on the Values
Listed in Tables A-2 and A-3 9-12
Table 9-9 Contribution of Skin Uptake to the Total Absorbed Dose of Chemical
Vapors in the Rat 9-15
Table 10-1 Default Values for Water-Contact Exposure Parameters 10-2
Table 10-2 Default Value for Soil 10-10
Table 10-3 Dermal Absorption Fractions for Dilute Concentrations of Contaminants
in 1 mg Soil/cm2 10-15
vni
-------
FIGURES
Page No.
Figure 3-1 Structure of the Skin 3-2
Figure 3-2 Two-Phase Model of the Stratum Corneum 3-3
Figure 3-3 Transport/Loss Processes Occurring in the Skin 3-4
Figure 3-4 Major Routes of Diffusion Through the Skin 3-5
Figure 3-5 Regional Variation in the Percutaneous Absorption of Hydrocortisone in
Humans 3-12
Figure 4-1 IPPSF Preparation and Perfusion System 4-16
Figure 5-1 Determination of Percutaneous Absorption Lag Time 5-7
Figure 5-2 Example of a Physiologically-Based Pharmacokinetic Model 5-15
IX
-------
FOREWORD
The Exposure Assessment Group (EAG) within OHEA of EPA's Office of Research and
Development has three main functions: (1) to conduct exposure assessments; (2) to review
assessments and related documents; and (3) to develop guidelines for exposure assessments. The
activities under each of these functions are supported by and respond to the needs of the various
program offices. In relation to the third function, EAG sponsors projects aimed at developing
or refining techniques used in exposure assessments.
The purpose of this document is to provide guidance on how to conduct dermal exposure
assessments. These procedures are not official Agency guidelines, rather they represent the
opinions and recommendations of the authors. Dermal exposure is the least well understood and
most uncertain area in the exposure assessment field. We hope that this guidance will improve
the scientific basis for approaching this challenging aspect of risk assessment
Michael A. Callahan
Director
Exposure Assessment Group
-------
PREFACE
The Exposure Assessment Group of the Office of Health and Environmental Assessment
has prepared this guidance document at the request of the Office of Emergency and Remedial
Response (Superfund) of the Office of Solid Waste and Emergency Response.
XI
-------
AUTHORS
The Exposure Assessment Group (EAG) within EPA's Office of Health and
Environmental Assessment was responsible for the preparation of this document. The original
report was prepared by Technical Resources, Inc. under EPA Contract Number 68-W8-0082.
Revisions and additional prerpartion was provided by an expert workgroup with the support of
ILSI Risk Science Institute under EPA Cooperative Agreement Number CR-817457-01-0 and
VERSAR Inc. under EPA Contract Number 68-DO-0101. Kim Hoang of EAG served as EPA
task manager (as well as contributing author) providing overall direction and coordination of the
production effort as well as technical assistance and guidance.
Technical Resources, Inc.
Ronald Brown
Louis Cofone
Isaac Divan
Abrarnah Mittelman
Kathleen Plourd
VERSAR Inc.1
Jeffrey Driver
Robert Fares
Bentley Gregg
Nica Mostaghim
Greg Schweer
Gary Whitmyre
Dermal Work Group
Robert Bronaugh
Jeffrey Driver
Larry Fishbein
Richard Guy
Karen Hammerstrom
Kim Hoang
Frank Marzulli
John Schaum
Janet Springer
FDA, Center for Food Safety and Applied Nutrition
RiskFocus®\VERSAR Inc.
ILSI Risk Science Institute
University of California, San Francisco
EPA, Office of Health and Environmental
Assessment
EPA, Office of Health and Environmental
Assessment
Consultant in Toxicology
EPA, Office of Health and Environmental
Assessment
EPA, Center for Food Safety and Applied Nutrition
1 Special thinks to Sylvia Johnson and Sally Gravely for providing word processing and other editorial support.
xu
-------
REVIEWERS
This document has undergone extensive internal and external review. The final review
will be conducted at the Dermal Workshop to be held on April 2 and 3, 1991, by the following
participants:
Environmental Protection Aeencv Reviewers:
Gerry Ackland Office of Modeling, Monitoring Systems, and
Quality Assurance
Linda Birnbaum Office of Health Research
Jerry Blancato Office of Modeling, Monitoring Systems, and
Quality Assurance
Nancy Chiu Office of Drinking Water
Christina Cinalli Office of Toxic Substances
Ernest Falke Office of Toxic Substances
Larry Hall Office of Health Research
Karen Hammerstrom Office of Health and Environmental Assessment
Kim Hoang Office of Health and Environmental Assessment
Ann Jarabek Office of Health and Environmental Assessment
Leonard Keifer Office of Toxic Substances
Carol Kimmel Office of Health and Environmental Assessment
Jim Konz Office of Emergency and Remedial Response
Curt Lunchick Office of Pesticide Programs
Terry O'Bryan Office of Emergency and Remedial Response
Jean Parker Office of Health and Environmental Assessment
Bruce Peirano Office of Health and Environmental Assessment
John Schaum Office of Health and Environmental Assessment
Robert Zendzian Office of Pesticide Programs
Other Government Aeencv Reviewers:
Robert Bronaugh FDA, Center for Food Safety and Applied Nutrition
Claire Franklin Health & Welfare Canada
Major James McDougal Wright Patterson Air Force Base
Thomas McKone Lawrence Livermore National Laboratory
William Reifenrath Letterman Army Institute of Research
Curtis Travis Oak Ridge National Laboratory
xm
-------
EPA Contractor Reviewers:
Ron Brown
Jeffrey Driver
Larry Fishbein
Carol Henry
Frank Marzulli
Abe Mittleman
Gary Whitmyre
Technical Resources, Inc.
RiskFocus®\VERSAR Inc.
ILSI Risk Science Institute
ILSI Risk Science Institute
Consultant in Toxicology
Technical Resouces, Inc.
RiskFocus®\VERSAR Inc.
Academic Reviewers:
Kenneth Bischoff
Anders Bowman
Annette Bunge
Halina Brown
Yie Chen
Gordon Flynn
Richard Guy
Dale Hattis
Howard Maibach
Ronald Wester
University of Delaware
University of California, San Francisco
Colorado School of Mines
Clark University
Rutgers University
University of Michigan
University of California, San Francisco
Clark University
University of California, San Francisco
University of California, San Francisco
Private Sector Reviewers:
Michael Bird
Emily Currie
Stephen Frantz
Clay Frederick
Cindy Fuller
Bun Hakkinen
J. M. Holland
Bruce Houtman
John Kao
Richard Nolan
Boyd Poulsen
Tim Roy
Tom Spencer
Joe Yang
Exxon Biomedical
Chemical Manufacturers Association
Bushy Run Research Center, Union Carbide Corp.
Rohm & Haas Co.
Woodward-Clyde Consultants
Proctor and Gamble
Upjohn Corp.
Dow Elanco Co.
Smith Kline Beecham Pharmaceuticals
Dow Chemical
Syntex Corp.
Mobil Oil Corp.
Cyngus Labs
Mobil Oil Corp.
xiv
-------
1.0 INTRODUCTION
Guidance is needed to assist exposure/risk assessors in determining when dermal exposure may be a route
of concern, and to assist in making preliminary estimates of dermal exposure and absorption. The goal of this
document is to describe the pertinent input parameters and to discuss their practical use in dermal exposure
assessment Specifically, this document:
• Summarizes the available state of knowledge and evaluates the validity of existing methods and
techniques;
Clarifies the capabilities and limitations of these methods and elaborates upon the uncertainties
associated with them;
Provides guidance in a stepwise approach to dermal exposure assessment;
• Describes the use of dermal permeability data including estimation techniques, the use of data
reported as percent of applied dose absorbed, and the use of default values when experimental or
empirical data are lacking; and
• Provides a database of dermal permeability coefficient values.
The exposure/risk assessor should be aware, however, that a number of significant uncertainties remain such
as assessment of dermal contact with soil and vapors, and the absence of experimentally derived permeability or flux
values for many of the compounds commonly encountered at hazardous waste sites. In the latter case, the previous
lack of guidance on appropriate default values to select in the absence of experimentally derived values places a
major limitation on the ability of Agency scientists to conduct site-specific dermal exposure/risk assessments. This
guidance document is divided into the following sections:
Section 2 Presents methods for estimating dermal exposure (i.e., the amount of contaminant that
contacts skin). This section includes discussion of exposure pathways, contact duration
and frequency, body surface area, and soil adherence.
Section 3 Describes the biological mechanisms of dermal absorption, i.e., amount of contaminant
that crosses the skin and enters the body; includes discussion of skin structure, transport
processes, metabolism, and factors that influence dermal absorption such as body site and
hydration level. This section establishes the theoretical basis for absorption issues
presented in Sections 4.0 through 8.0.
Section 4 Describes laboratory techniques for measuring dermal absorption. It includes a number
of in vivo and in vitro methods and comparisons of these methods.
1-1
-------
Sections
Section 6
Section?
Sections
Section 9
Section 10
Describes how to estimate flux, permeability coefficient, and absorption fraction from
experimental studies. Methods for interconversion among these parameters are also
described.
Describes the available studies assessing dermal uptake from soil and assesses those
factors which will affect dermal absorption from soil including organic matter content and
exposure duration.
Describes the available literature involving dermal absorption of chemical vapors, and the
basis for a possible future predictive approach.
Presents a number of approaches for estimating dermal absorption rates involving water
contact
Offers exposure/risk assessors guidance to determine when dermal exposure will
contribute significantly to total absorbed dose. It presents existing literature comparing
routes of exposure as proportion of total dose for aqueous media, soil, and vapors.
Describes a step-by-step procedure to conduct a dermal exposure assessment. Default
assumptions are included for situations where information is not available.
Section 11 Provides references cited in the previous sections.
In addition to these sections, appendices have been added to provide the user of this document with ready
access to important reference material.
Appendix A Describes the basis for selecting a single recommended Kp value under "standard"
scenario conditions for the compounds in the dermal permeability database.
Appendix B Provides definitions of terms used in this document.
1-2
-------
2.0 METHODOLOGIES FOR ESTIMATING DERMAL EXPOSURE
Estimation of the dose of a compound absorbed across the skin requires knowledge of the amount of
material that comes into contact with the skin and the rate at which this material is absorbed. The purpose of this
section is to present methods for estimating the amount of material that comes into contact with the skin, or
exposure. Subsequent sections will address methodologies for estimating the dermal absorption rate.
Dermal exposure has been addressed in several of the existing multimedia exposure assessment guidance
documents developed by EPA, including:
• Methods for Assessing Exposure to Chemical Substances (EPA, 1983);
Estimating Exposures to 23,7,8-TCDD (EPA, 1988a);
• Methodology for Assessment of Health Risks Associated with Multiple Pathway Exposure to
Municipal Combustor Emissions (EPA, 1986);
• Superfund Exposure Assessment Manual (EPA, 19885);
Risk Assessment Guidance for Superfund (EPA, 1989b); and
Exposure Factors Handbook (EPA, 1989a).
These documents serve as a valuable resource of information to assist in the assessment of dermal exposure.
To avoid repetition of this material, this section will summarize the important concepts presented in earlier EPA
documents, and will refer the reader back to these documents for a more detailed discussion of these topics.
Dermal exposure to environmental contaminants can occur during a variety of activities and may be
associated with a number of different environmental media:
• Water (e.g., bathing, washing, swimming);
• Soil (e.g., outdoor recreation, gardening, construction);
• Sediment (e.g., wading, fishing);
Liquids (e.g., use of commercial products); and
• Vapors (e.g., use of commercial products).
This document focusses primarily on water and soil contact because these are currently the most well studied
and probably the most prevalent The following methodologies have been described in previous EPA documents
2-1
-------
to estimate the dermally absorbed dose of compounds in contaminated water or soil. For contact with contaminated
water, dermally absorbed dose can be estimated by (EPA, 1989b):
CWxSAxKxETxEF xEDxCF
Dermally Absorbed Dose (mg/kg/day) • '
BW x AT
where:
CW = Chemical concentration in water (mg/L);
SA * Skin surface area available for contact (cm2);
Kp = Chemical-specific dermal permeability constant (cm/hr); distance travelled through
skin/unit time (this constant is determined at steady state and infinite dose, and varies with
skin thickness);
ET « Event time (hours/day);
EF = Event frequency (days/year);
ED « Exposure duration (years);
CF » Volumetric conversion factor for water (1171000 cm3);
BW « Body weight (kg); and
AT = Averaging time, a pathway-specific period of exposure for non-carcinogenic effects (i.e.,
ET x EF x ED x 365 days/year), and a 70-year lifetime for carcinogenic effects (i.e., 70
years x 365 days/year).
For contact with contaminated soil, two documents (EPA, 1986 and 1989b) provide a variation of this
equation that replaces the dermal absorption rate constant with a term to account for the percent of the initial dose
present in soil that is absorbed. Dermally absorbed dose is calculated using this approach by:
nnr. ^^ - CSxSAxAF+XBSx EF x ED x CF
' .... • BW X AT . • • •• .:.-.. .":C:;...:. ,;:; : •
where:
CS B Chemical concentration in soil (mg/kg);
2-2
-------
Material belongs to:
Office o; Toxic Substances Library
U.S. Er.v onmental Protection Agency
SA = Skin surface area available for contact (cm2); 401 M Street, S.W. TS-793
Washington, D.C. 20460
AF « Soil-to-skin adherence factor (mg/cm2/event); (202)382-3944
ABS & Absorption factor (unitless);
EF * Event frequency (events/year);
ED « Exposure duration (years);
CF « Conversion factor (10"* kg/mg);
BW • Body weight (kg); and
AT « Averaging time, a pathway-specific period of exposure for non-carcinogenic effects (i.e.,
EF x ED x 365 days/year), and a 70-year lifetime for carcinogenic effects (i.e., 70 years
x 365 days/year).
Although the apparent simplicity of the absorption fraction (% absorbed) approach makes it appealing, it
is not practical to apply it to water contact scenarios, such as swimming, because of the difficulty in estimating the
total material contacted. Attempts have been made to apply dermal absorption models that consider the thickness
of a thin film of chemical and water on the skin (EPA, 1983; Wester et al., 1987; Wester and Maibach, 19895).
However, for exposure scenarios of interest in environmental settings, especially when considering exposure to a
contaminated aqueous medium, this thickness is difficult to establish. There is essentially an infinite thickness of
material available, and the contaminant will be continuously replaced, thereby increasing the amount of available
material by some large, but unknown, amount. In contrast, not all of the soil contaminant in a thick layer of dirt
applied to the skin can be considered to be bioavailable, nor can it be considered to constitute a dose. However, if
the amount of contaminant in the adhered soil can be established, the absorption fraction approach may be practical.
A more thorough discussion of the dose of a compound that can be considered to be bioavailable in an aqueous or
soil matrix can be found in Sections 5.4 and 6, respectively.
Use of an absorption factor (% absorbed) for estimating dermal absorption of contaminants in soil or
aqueous media presents at least one other uncertainty. The percentage of the initial dose absorbed is rarely measured
at steady-state, and is generally given as a proportion of the applied dose that is absorbed (or lost from the
application site) after a contact duration, or at various time steps during such contact. Furthermore, percent absorbed
is dependent upon amount applied per unit area, becoming less with increasing amount applied. The significance
of this manner of reporting is that the percentage of the dose that is absorbed after 1 hour would not necessarily be
the same as the percentage absorbed after 24 hours, nor is it equivalent to 1/24 the percentage absorbed after 24
hours. In contrast, flux (J) and permeability coefficient (Kp) values are generally determined under conditions of
steady-state, or near steady-state. Additionally, when Pick's Law of diffusion is applicable (no chemical-related skin
2-3
-------
damage), the permeability coefficient is constant over the range of contaminant concentrations generally of interest.
This provides a much more consistent basis for comparing the dermal absorption potential of chemicals of concern.
Therefore, the permeability coefficient-based approach is advocated over the absorption fraction approach
for determining the dermally absorbed dose of compounds in an aqueous medium. Because of the lack of data
demonstrating the scientific reliability of using aqueous K,, data for compounds bound to soil and reduced
uncertainty in defining an applied dose, the absorption fraction-based approach is presently recommended for
determining the dermally absorbed dose of soil contaminants.
The following sections provide a background for understanding the dermal exposure variables used in the
dermal exposure equations for soil and water. The range of possible values for each parameter is discussed along
with recommended default values to be used when site specific data are not available.
2.1 Exposed Populations
The components of a dermal contact scenario include identification of the contaminated media (water, soil,
air) and the pathways by which it reaches exposed populations, identification and quantification of exposed
populations, exposure time and frequency, and duration of contact. Detailed guidance for performing a quantitative
population analysis is provided in the Superfund Exposure Assessment Manual (SEAM) (EPA, 1988b).
Populations that experience dermal exposure to groundwater or surface water contaminants can be identified
by recognizing geographically-defined sources of recreational surface waters such as contaminated rivers, lakes, and
ponds. The exposed population would include swimmers in those specific contaminated waters, who can experience
dermal exposure to contaminants over a major proportion of their bodies' surface area. Quantification of the
population exposed in this manner may be possible by contacting local governmental agencies concerned with
recreation. In the absence of site-specific data, SEAM (EPA, 1988b) recommends using a national average value
of 34% of the total population of the area to quantify the number of persons that swim outdoors in natural water
bodies.
Populations potentially exposed to contaminated groundwater or surface water also include persons served
by a water supply system that draws its water supply from a contaminated source. Such a population may be
dermally exposed while showering or bathing. Information regarding local surface water or groundwater sources,
the populations served, and the number of households that draw water from private wells should be available from
local public works or health departments.
2-4
-------
Populations that are exposed by dermal contact with contaminated soil may include neighborhood children
playing at contaminated sites, workers, and gardeners. Children who play outside are expected to be exposed to
contaminated soil more frequently and with greater magnitude than adults. The number of children and other age
groups potentially exposed can be estimated by referring to census data.
22 Contamination of Environmental Media
Since contaminants may be transported in environmental media as particulates or in dissolved form, any
contact with a medium that contains such substances can potentially result in dermal absorption. Generally,
contaminants can be found in air as vapors, dissolved in water, or mixed with soil. In addition, particle-associated
contaminants can be carried in air, water, or soil.
The extent of the dermal exposure and absorption of a specific contaminant from soil, water, sediments, or
vapor can only be evaluated following a determination of the concentration of the contaminant in the medium of
interest. Exposure concentrations may be estimated using monitoring data alone, or as in most exposure assessments,
using a combination of monitoring data and environmental fate and transport models. The selection and use of such
models are explained in the following documents:
Superfund Exposure Assessment Manual (SEAM) (EPA, 1988b);
Exposure Factors Handbook (EPA, 1989a);
• Selection Criteria for Models Used in Exposure Assessments: Groundwater Models (EPA, 1988c);
and
• Selection Criteria for Models Used in Exposure Assessments: Surface Water Models (EPA, 1987).
2.3 Exposure Time, Frequency and Duration
Exposure duration and event time/frequency are three of the variables necessary for application of the dermal
dose equations given in Section 2.0. Exposure duration (expressed in units of years) can be defined as the overall
time period over which dermal contact events occur. Event time (i.e., hour/event) is the time over which a single
contact event occurs. Event frequency (i.e., events/year) refers to how often the contact event occurs. The
following discussion summarizes the values for these parameters for dermal exposure to contaminants in soil and
water.
2-5
-------
Soil Contact Time. Frequency and Duration
Ranges of values for soil exposure time and frequency have been based on research concerning the soil
ingestion pathway. These values are considered applicable to scenarios involving the dermal exposure route because
human activities associated with incidental soil ingestion are likely to present opportunities for dermal exposure.
For example, Hawley (1985) used existing literature and professional judgment to develop scenarios for estimating
exposure of young children, older children, and adults to contaminated soil These exposure time values are
presented in Table 2-1. Hawley's assumptions included consideration of seasonal factors; estimates were divided
into winter months, when soil contact would be limited in some parts of the United States, and summer months,
when more opportunity existed for such contact Hawley (1985) assumed an outdoor soil exposure of five days a
week during a period of six months, for young children (2.5 years of age). The contact time was estimated to be
approximately 12 hours, since children retain soil on their skin after coming indoors. For older children, the average
outdoor playtime, during which dermal exposure to soil would occur, was estimated at five hours per day 6 days per
week from May to September. Adults were expected to be exposed to outdoor soil two days a week for eight hours
during the summer months.
Table 2-1. Assumptions of Outdoor Soil Exposure Time
Value Reference
12 hours/day, 5 days/week, 6 months/year (age 2.5 years) Hawley, 1985
5 hours/day, 6 days/week, 5 months/year (older children) Hawley, 1985
8 hours/day, 2 days/week, 3 months/year (adults) Hawley, 1985
If the contaminated soil is carried into the house from the surrounding area, household dust may be
contaminated at levels approaching those found outdoors (EPA, 1988a). Furthermore, indoor use of some
commercial products, such as pesticides, may result in increased indoor contaminant levels. Therefore, dermal
contact time, frequency and duration with contaminated indoor dust would need to be considered in settings where
this exposure pathway may occur.
No actual data could be found on the residence time of soil residues on skin. It probably corresponds
roughly to the time between washings or about 8 to 24 hours. Soil contact time (as discussed in Section 6) may
influence the absorption fraction value. It may also be useful for evaluating the experimental conditions used to
generate absorption fraction estimates. Since soil residue times are probably 8 to 24 hours, experiments conducted
2-6
-------
over similar times would provide the best basis for absorption fraction estimates.
A range of values for the frequency of soil contact has been developed in past exposure assessments. These
frequencies were derived from considerations of seasonal factors that may influence activities and soil conditions.
These are summarized in Table 2-2.
Table 2-2. Assumptions of Frequency of Exposure to Soil
Range Reference
247-365 days/year Schaum, 1984
180 days/year Paustenbach et al., 1986
130 days/year (<2-5 yr) Hawley, 1985
130 days/year (older children) Hawley, 1985
43 days/year (adults) Hawley. 1985
The upper end of the range presented above is based on the rationale that in warmer climates, people who
actively garden or play outdoors could have contact with soil almost every day. On this basis, a default value of
365 day/yr is recommended. However, in cooler climates or where the contaminated soil is located outside the
residential property, a lower frequency (e.g., 180 days/year) should be assumed.
The exposure duration over which soil contact could occur depends primarily on how long a person lives
near a contaminated site. EPA (1989a, 1989b) has reviewed census data and concluded that the time people spend
at a residence averages about 30 years, with an upper estimate of 30 years. On this basis, the upper estimate of 30
years is recommended as a default.
Water Contact Time. Frequency, and Duration
Approximately 90% of the American population bathes every day, and 5% average more than one bath per
day (Tarshis. 1981). Seventy-five percent of the men and 50% of the women use showers as a primary means of
bathing. On this basis, a default bathing frequency of 365 days/year is recommended.
While no data have been presented to show gender or age differences in shower time, EPA (1989a)
presented a cumulative frequency distribution of shower times for a population as a whole. Based on records of 2500
Australian households, a median shower time of 7 minutes and a 90th percentile of 12 minutes were reported (James
and Knuiman, 1987). Furthermore, EPA (1989a) estimates that shower-flow rates range from 5 to 15 gallons per
minute. Brown and Hattis (1989) assumed a 20-minute bath time to use estimate the dermal absorption from
2-7
-------
hypothetical bathtub scenarios. Assuming an average shower of 7 min plus a few minutes for water residues to dry
a default of 10 min is recommended.
Much less data is available on the time and frequency of swimming. The SEAM (EPA, 1988b) suggests
that national average values based on the Department of the Interior (DOI) Bureau of Outdoor Recreation data (DOI,
1973) be applied. These are:
Exposure Frequency: 7 days per year
Exposure Time: 2.6 hours per day
Unfortunately, a distribution or range of values around these means was not provided. Care must be used
when applying these numbers as they represent the national mean and do not consider geographic region-specific
factors such as proximity or availability of surface waters for recreation, nor do they consider seasonal factors, such
as temperature, that are related to recreational water use. Furthermore, certain subpopulations (e.g., competitive
swimmers) will encounter a greater mean exposure frequency and time. For example, Vandeven and Hemnton
(1989) used an exposure duration of 3 hours and a frequency of 20 events per year in calculations intended to
illustrate the importance of dermal exposure to contaminated water that may leach from hazardous waste sites. For
default purposes, the DOI values of 7 days/yr and 2.6 hr/day are recommended.
The exposure duration for both swimming and bathing will be determined primarily by how long a person
lives in one residence. As stated above, census data suggests that this averages about 30 years and has an upper
estimate of 30 years. The average value of 30 years is recommended as a default
2A Surface Area
This Section describes how to obtain values for total body surface area and the surface area of component
body parts that may be exposed to contaminated media. In addition, a discussion is included concerning the effect
of clothing on determining the effective surface area exposed.
The U.S. EPA Exposure Factors Handbook (EPA. 1989a) and the Standard Factors Used in Exposure
Assessments (EPA, 1985) present reviews of measurement techniques that have been used in the past to estimate
surface area. In addition, these documents present summary data for skin surface areas for different genders and
ages.
Determination of the surface areas of the component body pans has been performed by a number of authors
2-8
-------
as part of their determination of whole body surface areas. The surface areas of anatomical parts have been reported
by gender, age, and ethnic group. Early studies have reported surface areas for such component parts as head, trunk,
upper arms, forearms, hands, thighs, legs, and feet Unfortunately, these early studies were based on very limited
numbers of subjects, and the values reported may not be meaningful when attempting to extrapolate to other
populations. For example, Boyd's (1935) data set consists of measurements on two children measured at various
intervals over less than a year. DuBois and DuBois' (1916) data set is limited to four adult males, and one adult
female. More extensive studies, such as those performed by Fujimoto and Watanabe (1969) are useful because of
the number of subjects measured (201), but their applicability may be limited due to initial biases; this study is
limited to Japanese subjects prescreened to fit a "standard Japanese physique."
EPA (1985) used available direct measurement data, including the Japanese data outlined above, to generate
equations that estimated surface area of body pans as a function of height and weight These equations were then
applied to the U.S. adult population, using height and weight distribution data obtained from the National Health and
Nutrition Examination Survey II (NHANES II) (NCHS, 1983). Insufficient data precluded the development of
similar equations for children.
For adults, EPA (1985) used regression equations that relate height and weight to the surface area of the
head, trunk, upper extremities, and lower extremities. Because of the low regression coefficient (r2) values of the
equations for male and female heads, and for female hands, the equations are considered to be inaccurate predictors
of the surface area for these body parts.
EPA (1985) also estimated percentiles estimated the percentile estimates for the surface areas of body parts
using the regression equations, and the NHANES II height and weight database. For children, the available
measurements of the body part surface areas were summarized as a percentage of the total surface area. Data on
the surface areas of specific body parts for adults and children are presented in the Exposure Factors Handbook
(EPA, 1989a). Note that the percent of total body surface area contributed by the head decreases from childhood
to adult status, whereas that contributed by the legs increases.
One inherent assumption of many exposure scenarios developed in the past is that clothing prevents dermal
contact and subsequently absorption of contaminants. This assumption may in fact be faulty in cases where the
contaminant is carried in a fine dust or liquid suspension, which may be able to penetrate clothing. Studies using
persona] patch monitors placed beneath clothing of pesticide workers show that a significant proportion of the dermal
exposure may occur at anatomical sites that are covered by clothing (Maddy et al., 1983). Nevertheless, hand
exposure is believed to exceed exposure of all other body areas in such workers. Furthermore, studies have
demonstrated that hands cannot be considered to be protected from exposure even if waterproof gloves are worn.
2-9
-------
This may be because of contamination on the interior surface of the gloves, removal of gloves during machine
adjustments, and handling of the outside of the gloves while putting them on or taking them off (Maddy et al., 1983).
Depending on their specific tasks, pesticide workers have been shown to experience 12% to 43% of their total
exposure through their hands, approximately 20 to 23% through their heads and necks, and 36% to 64% through their
torsos and arms, despite the use of protective gloves and clothing (Fenske et al., 1985,1986). These studies were
conducted with fine mists and vapors. Clothing is expected to limit the extent of the exposed surface area in cases
of soil contact EPA (1989a) presents two adult clothing scenarios for outdoor activities:
Typical case: Individual wears long sleeve shirt, pants, and shoes. The exposed skin
surface is limited to the head and hands (2,000 cm2);
Reasonable worst case: Individual wears a short sleeve shirt, shorts, and shoes. The exposed
skin surface is limited to the head, hands, forearms, and lower legs
(5300 cm2).
For swimming and bathing scenarios, past exposure assessments have assumed that 75% to 100% of the
skin surface is exposed (Vandeven and Herrinton, 1989; Wester and Maibach, 1989a). For default purposes, surface
areas of 20,000 cm2 for adults and 10,000 cm2 for children are recommended as conservative values for entire skin
surface area. The range of possible values are shown in Tables 2-3 and 2-4 (EPA, 1989a).
For soil contact scenarios, dermal exposure was expected to occur at the hands, legs, arms, neck, and head
(McKone and Layton, 1986) with approximately 26% and 30% of the total surface area exposed for adults and
children, respectively. Less conservative scenarios have limited exposure to the arms, hands, and feet (e.g.,
Vendeven and Herrinton, 1989). An indoor dust contact scenario considered exposure to be limited to feet and palms
(EPA, 1986). The clothing scenarios presented above, suggest that roughly 10 to 25 % of the skin area may be
exposed to soil. Using the upper end of this range, a default of 5,000 cm2 for adults and 2,500 cm2 for children is
recommended.
2-10
-------
Table 2-3. Surface Area by Body Part for Adults (m2)
Men
Body Pan
Head
Trunk
Upper extremities
Arms
Upper Arms
Forearms
Hands
Mean (s.A)
0.118
0.569
0.319
0.228
0.143
0.114
0.084
Lower extremities 0.636
Legs
Thighs
Lower legs
Feet
TOTAL
0.505
0.198
0.207
0.112
(0.0160)
(0.0140)
(0.0461)
(0.374)
(0.0143)
(0.0127)
(0.0127)
(0.0994)
(0.0885)
(0.1470)
(0.0379)
(0.0177)
1.94 (0.00374)*
Min.
0.090
0.306
0.169
0.109
0.122
0.0945
0.0596
0.283
0.221
0.128
0.093
0.0611
1.66
-Max.
-0.161
-0.893
-0.429
-0.292
-0.156
-0.136
-0.113
-0.868
-0.656
-0.403
-0.2%
-0.156
-2.28b
n
29
29
48
32
6
6
32
48
32
32
32
32
48
Women
Mean (s.d.)
0.110
0.542
0.276
0.210
0.0746
0.626
0.488
0.258
0.194
0.0975
(0.00625)
(0.712)
(0.0241)
(0.0129)
-
-
(0.00510)
(0.0675)
(0.0515)
(0.0333)
(0.0240)
(0.00903)
1.69 (0.00374)*
Min.
0.0953
0.437
0.215
0.193
0.0639
0.492
0.423
0.258
0.165
0.0834
1.45
-Max.
-0.127
-0.867
-0.333
-0.235
-
-
-0.0824
-0.809
-0.585
-0.360
-0.229
-0.115
-2.09b
n
54
54
57
13
-
-
12
57
13
13
13
13
58
* median (standard error)
b percentiles (5* - 95th)
s.d. = standard deviation.
s.e. - standard error for the 5th to 95th percentile of each body part.
n = number of observations
Source: Adapted from EPA (1985).
2-11
-------
Table 2-4. Total Body Surface Area of Male Children in Square Meters
Age (yr)»
2<3
3<4
4<5
5<6
6<7
7<8
8<9
9<10
10 < 11
11<12
12<13
13<14
14<15
15 < 16
16 < 17
17<18
3<6
6<9
9< 12
12<15
15<18
5
0.527
0.585
0.633
0.692
0.757
0.794
0.836
0.932
1.01
1.00
1.11
1.20
1.33
1.45
1.55
1.54
0.616
0.787
0.972
1.19
1.30
10
0.544
0.606
0.658
0.721
0.788
0.832
0.897
0.966
1.04
1.06
1.13
124
1.39
1.49
1.59
1.56
0.636
0.814
1.00
124
1.55
15
0.552
0.620
0.673
0.732
0.809
0.848
0.914
0.988
1.06
1.12
1.20
127
1.45
1.52
1.61
1.62
0.649
0.834
1.02
127
1.59
25
0.569
0.636
0.689
0.746
0.821
0.877
0.932
1.00
1.10
1.16
125
1.30
1.51
1.60
1.66
1.69
0.673
0.866
1.07
132
1.65
Percentile
50
0.603
0.664
0.731
0.793
0.866
0.936
1.00
1.07
1.18
123
1.34
1.47
1.61
1.70
1.76
1.80
0.728
0.931
1.16
1.49
1.75
75
0.629
0.700
0.771
0.840
0.915
0.993
1.06
1.13
128
1.40
1.47
1.62
1.73
1.79
1.87
1.91
0.785
1.01
128
1.64
1.86
85
0.643
0.719
0.7%
0.864
0.957
1.01
1.12
1.16
1.35
1.47
1.52
1.67
1.78
1.84
1.98
1.96
0.817
1.05
136
1.73
1.94
90
0.661
0.729
0.809
0.895
1.01
1.06
1.17
125
1.40
1.53
1.62
1.75
1.84
1.90
2.03
2.03
0.842
1.09
1.42
1.77
2.01
95
0.682
0.764
0.845
0.918
1.06
1.11
124
129
1.48
1.60
1.76
1.81
1.91
2.02
2.16
2.09
0.876
1.14
1.52
1.85
2.11
* Lack of height measurements for children < 2 years in NHANES U precluded calculation of surface areas for this
age group.
Source: EPA I989a.
2-12
-------
2.5 Dermal Adherence of Soil
A number of studies have attempted to determine the values of dermal soil adherence. Lepow et al. (1975)
employed preweighed self-adhesive labels to sample a standard area on the palm of the hands. The preweighed
labels were pressed on a single area, and often pressed several times on the given area to obtain an adequate sample.
In the laboratory, labels were equilibrated in a desiccant cabinet for 24 hours (comparable to the preuse weighing
desiccation), then the total weight was again recorded. The mean weight of hand din for the 22 hand samples was
11 mg; on a 21.5 cm2 preweighed label, this amounts to 0.51 mg/cm2. LePow et al. (1975) stated that this hand din
samples represented only a small fraction (percent not specified) of the total amount of surface din present on the
hands, since much of the din may be trapped in skin folds and creases; moreover, there may have been patchy
distribution of the din on the hands. However, 0.5 mg/cm2 appears to be an adequate estimate for skin adherence
based on data reported by other researchers using different methods.
Roels et al. (1980) assessed lead levels removed from children's hands by rinsing the hands in 500 ml dilute
nitric acid and compared these lead levels to lead levels in soil to determine the amount of soil adhering to the hands.
The lead content of the hand rinse and of representative soil samples were determined, and an estimate of the amount
of soil (g) on the hand was calculated by dividing the hand lead amount (fig) by the soil lead amount (ug/g). The
mean soil amount adhering to the hand was 0.159 g. Sedman (1989) used this estimate and the average surface area
of the hand of an eleven year old (i.e., 307 cm2) to estimate the amount of soil adhering per unit area of skin (0.9
mg/cm2). The Sedman (1989) estimate assumed approximately 60 percent (185 cm2) of the lead on the hand was
recovered by the method employed by Roels et al. (1980).
Que Hee et al. (1985) used soil having panicle sizes ranging from 44 to 833 urn diameters, fractioned into
six size ranges, to estimate the amount of soil adhering to skin. For each range of particle size, the amount of soil
that adhered to the palm of the hand of a small adult was determined by applying approximately 5 g of soil for each
size fraction removing excess soil by shaking the hands, and then measuring the difference in weight before and after
soil application. Several assumptions were made when applying the studies results to other soil types and exposure
scenarios. These assumptions include: (1) soil is composed of panicles of the indicated diameters; (2) all soil types
and particle sizes adhere to the skin to the degree observed in the study; and (3) an equivalent weight of particles
of any diameter adhere to the same surface area of skin. On average, 312 mg of soil adhered to the small adult
palm. The surface area of the palm of a small adult (approximately 14 years old with an average total body surface
area of 16,000 cm2 and a total hand surface area of 400 cm2 is assumed to be approximately 160 cm2. Based on
these assumptions, 0.2 mg of soil adhered to 1 cm2 of skin.
Sedman (1989) used the above estimates from Lepow et al (1975), Roels et al. (1980), and Que Hee et al.
2-13
-------
(1985) to develop a maximum soil load that could occur on the skin given the types of procedures employed in each
study to determine their measurements. A rounded arithmetic mean of 0.5 mg/cm2 was calculated from the three
studies. Sedman (1989) stated that this was near the maximum load of soil and normally it is unlikely that most skin
surfaces would be covered with this great an amount of soil.
Driver et al. (1989) conducted soil adherence experiments which involved the use of various soil types
collected from sites in Virginia. A total of five soil types were collected: Hyde, Chesapeake, Panorama, lackland,
and Montalto. Both top soils and subsoils were collected for each soil type. The soils were also characterized by
cation exchange capacity, organic content, clay mineralogy, and particle size distribution. The soils were dry sieved
to obtain particle sizes of < 250 urn and < 150 urn. For each soil type, the amount (mg) of soil adhering to adult
male hands, using both sieved and unsieved soils, was determined gravimetrically (i.e., measuring the difference in
soil sample weight before and after soil application to the hands). An attempt was made to measure only the
minimal or "monolayer" of soil adhering to the hands. This was done by mixing a pre-weighed amount of soil over
the entire surface area of the hands for a period of approximately 30 seconds, followed by removal of excess soil
by gently nibbing the hands together after contact with the soil. Excess soil which was removed from the hands was
collected and weighed with the original soil sample.
Driver et al. (1989) measured average adherences of 1.40 mg/cm2 for particle sizes less than 150 urn, 0.95
mg/cm2 for particle sizes less than 250 um and 0.58 mg/cm2 for unsieved soils. The analysis of variance statistics
showed the most important factor affecting adherence variability was particle size, with a variance (F) ratio far in
excess of the 0.999 significance value (p < 0.001). The next most important factor of soil type and subtype with
a F ratio also in excess of 0.999 significance ftxO.OOl). The interaction of soil type and particle size was also
significant, but at a lower 0.99 significance level (p < 0.01).
Driver et al. (1989) found statistically significant increases in adherence with decreasing particle size;
whereas, Que Hee et al. (1985) found relatively small changes over particle size. Also, the amount of adherence
found by Driver et al. (1989) was greater than that of Que Hee et al. (1985). Although it appears that soil particle
size may affect adherence, exact quantitative relationships cannot be derived at this time because of insufficient data.
It is suggested that this is an area for further study.
The U.S. EPA Superfund Exposure Assessment Manual (EPA 1988b) reported an upper-bound soil-to-skin
adherence value of 2.77 mg/cm2. This estimate was based on a study by Harger (1979, as cited in EPA 1988b)
conducted for the State of Michigan, Toxic Substances Control Division, which estimated the adherence of kaolin
on human hands using the gravimetric approach (weighed difference between amount applied to hands and amount
shaken off). The soil adherence for potting soil was also estimated with a reported value of 1.45 mg/cm2.
2-14
-------
Table 2-5 summarizes the soil adherence studies and values recommended in SEAM for use in Superfund
exposure assessments. Since the selected adherence value represents the amount of soil on the skin at the time of
measurement, they must be interpreted as representing a single event. Assuming that the amount measured on the
skin represents its accumulation between washings and that people wash at least once per day, then these adherence
values could also interpreted as daily contact rates. However, since the residence time of soils on skin has not been
studied (see above discussion) and since the adherence studies are independent of time, this is not recommended.
Table 2-5. Soil Adherence Values
Reference
Lepow, 1975
Roels et al., 1980
Que Hee et al., 1985*
Driver et al., 1989b
Size Fraction (urn)
-
-
<44
44-149
149-177
177-246
<150
<250
unsieved
Soil
0.5
0.9 - 1.5
0.17
0.17
0.19
0.18
1.40
0.95
038
Adherence (mg/cm2)
human
children
children
children
children
children
human
human
human
Harger, 1979
(as cited in EPA 1988b)
Yang et al.. 1989s
1.45 (potting soil)
2.77 (kaolin)
< 150
rat
1 Assume hand size * 160 cm2.
b Five different soil types and 2-3 soil horizons.
c Rat skin "monolayer" (i.e., minimal amount of soil covering the skin).
The following analysis was used to review the data in Table 2-4 for purposes of recommending a default
value. The soil adherence value from the Yang et al. (1989) study which used rat skin was not included for
consideration because of the uncertainties associated with using this value for human dermal exposure scenarios.
The value from the Harger (1979) study using kaolin was also not considered since it did not involve a common soil
type. Among the remaining studies, the Lepow (1975) and Roels (1980) studies have the advantage that they were
conducted under actual field conditions and disadvantage that they involved collection methods with unknown
efficiencies. The use of collection methods which were less than 100% efficient, suggest that the estimates may be
low. However, only hand samples were collected which suggests that the estimates may be high for other parts of
the body which probably have less soil contact. Finally, only children were surveyed and they may not be
2-15
-------
representative of adults. The Que Hee et al. (1989), Driver et al. (1989) and Harger (1979) studies all used the
gravimetric methods which provide comparable estimates of adherence potential. However, these studies were
conducted under laboratory conditions. They all examined adherence to hands only after intimate contact with soil.
Parts of the body which have less intimate contact with the soil will likely have lower values. In summary, all the
studies have uncertainties which make it very difficult to recommend a default value. A range of values from 0.2
to 1.5 mg/cm2-event appear possible. Until better data is available, a conservative central value of 1.0 mg/cm2-event
is recommended for use in dermal exposure assessments where site-specific soil adherence values have not been
estimated. Since this value is derived from hand measurements only, it may overestimate average adherence for the
entire exposed skin area. However until better data is available, more representative values cannot be confirmed.
2.6 Dermal Absorption of Compounds in Soil: The Soil "Monolayer" Concept
Some investigators (Yang et al. (1989) Roy et al (1989)) have postulated that soil absorption occurs only
from a "monolayer" of soil and that the absorbed dose is independent of the total amount of soil on the skin. This
monolayer has not been well defined but could be interpreted as a single layer of soil particles. Assuming tightly
packed 100 urn particles, approximately 10,000 particles would fit on 1 cm2 and weigh 8 mg/cm2 (assuming particle
density of 1,500 mg/cm3). Since this value is higher than measured in all human skin adherence studies, it suggests
that the adhered particles have an average diameter less than 100 urn or are not tightly packed.
Yang et al. (1989) attempted to measure a soil "monolayer" by covering the in vitro rat skin surface (which
had been shaved) with a minimal amount of soil, then removing the excess soil by gently tapping the inverted
diffusion cell and weighing the excess. This procedure indicated that 9 mg/cm2 adhered to the rat skin. Yang et
al. (1989) found a lower percentage absorption from a thick soil layer (56 mg/cm2) compared to a "monolayer" (9
mg/cm2), with 1.3 and 8.4% of benzo(a)pvrene (BaP) absorbed in % hours, respectively, from each type of soil, even
though both soil doses contained about 1 ppm of (BaP) in the soil. However, nearly identical quantities of
benzo(a)pyrene (1.3 ng) were absorbed from the thick layer and the "monolayer" of soil. The constant amount of
(BaP) absorbed from soil over the 96 hr testing period, regardless of the amount of soil applied, led Yang et al.
(1989) to conclude that the small amount of (BaP) absorbed was entirely derived from the "monolayer" of soil in
direct contact with the skin. Further, Yang et al. (1989) concluded that the degree of soil binding of (BaP) impedes
its movement into the skin to the extent that soil migration of (BaP) would be negligible from layers of soil above
the "monolayer". Yang et al. (1989) also reported the percutaneous absorption of (BaP) was significantly less from
petroleum-crude contaminated soil than from (BaP) from crude, indicating that sorption to soil retarded absorption.
It was also pointed out that percutaneous absorption of lipophilic compounds would be generally greater in the rat
2-16
-------
than in humans. Thus, Yang et al. (1989) suggested that it would be reasonable to expect the absorption rate of
(BaP) from soils would be lower through human skin than through rat skin.
Yang et al. (1989) also conducted exploratory in vitro experiments to assess the minimum amount of soil
that would adhere to the shaved rat skin. Close examination of skin sections after removal of excess soil indicated
that the components remaining on the skin were predominantly silt and clay fractions (<50 urn particle size). This
fine fraction contains the bulk of the soil organic carbon content, which is the dominant sorbent for lipophilic
compounds. Thus, clay and silt are more effective sorbents than sand. Based on these preliminary experiments, it
is interesting to note that the 9 mg/cm2 monolayer of soil estimated by Yang et al. (1989) compares with 1 mg/cm2
reported by Taylor, E.A. (1961) as the amount of liquid above which a pooling effect occurs on human skin. Yang
et al. (1989) carried out the percutaneous absorption experiments exclusively with soil particles of <150 urn, and
approximately 9 mg/cm2 of soil (which was found to be the minimum amount required for a "monolayer" coverage
of rat skin in both in vitro and in vivo tests). This value was acknowledged to be higher than the minimal amounts
of soil (monolayer) tested by Driver et al. (1989) with human skin (Table 2-5). Yang et al. (1989) stated that the
differences between rat and human soil adherence findings may be a result of differences between rat and human
skin texture, the types of soils tested, soil moisture contents, or possibly the methods of measuring soil adhesion.
In summary, the monolayer concept offers a potential means of interpreting data concerning the percutaneous
absorption of compounds from soil. Section 6 presents further discussion of this issue.
2.7 Summary
The range of possible values and recommended default for each of the exposure factors in the absorbed dose
equations for dermal contact with water and soil are summarized below in Table 2-6.
2-17
-------
Table 2.6. Ranges of Possible Values and Recommended
Defaults for Dermal Exposure Factors
Event
frequency
Event time
Exposure
duration
Skin surface
area
Soil-to-skin
adherence
factor
Water Contact
Bathing
Range
5-15 min/event
Adult:
15,000-23,000
cm2
2-18 yr. old:
See Table 2-3
and
2-4
Default
365 day/yr
10
min/event
30 years
20,000 cm2
10 yr. old:
10,000 cm2
Swimming
Range
15,000-23,000
cm2
2-18 yr old:
see Table 2-3
and
2-4
Default
7days/yr
2.6
hr/event
30 years
20,000
cm2
10 yr. old:
10,000
cm2
Soil contact
Range
Default
365
events^r
not applicable
Adults:
3800-5800
cm2
2-18 yrs.
see Table 2-
3
and 2-4
0.2-1.5
mg/cm2-
event
30 years
5,000 cm2
2^00 cm2
1.0
mg/cm2-
event
In summary, the supporting data for the dermal exposure factors are needed in several areas; therefore,
research is strongly recommended in the following areas:
• Soil adherence levels, especially levels found on parts of bodies other than hands during normal
soil contact activities;
• Residence time of soil residues on skin;
• Soil contact frequency, especially as a function of age, location, activity, climate, etc.;
• Swimming event times and frequencies.
2-18
-------
3.0 MECHANISMS OF DERMAL ABSORPTION
As explained in Section 2.0, the permeability coefficient (Kp) is a key parameter in estimating dermal
absorption. The effective use of Kp values in dermal exposure assessments requires understanding of the processes
that affect the transport of compounds across the skin. An understanding of these processes will enable the
exposure/risk assessor to evaluate the appropriateness of using the available Kp values to estimate dermally absorbed
dose in site- or scenario-specific exposure/risk assessments. In this section, the mechanisms by which compounds
are absorbed (or removed) from the skin are explored, and factors that affect this absorption process are considered.
This section will show how the exposure/risk assessor can use this information to make the qualitative judgments
about the appropriateness of Kp values generated in a study with a defined set of conditions that may be different
from those encountered in the exposure scenario of interest.
3.1 Structure of the Skin
The general anatomy and morphology of the skin have been well characterized. Several detailed reviews
of the physical nature of the skin are available (e.g.. Marks et al., 1988; Elias, 1987), and the reader is directed to
these reviews for a thorough discussion of this topic. However, percutaneous absorption is highly influenced by the
structure of the skin. Therefore, a brief review is presented in this section to better allow the exposure/risk assessor
to interpret compound-specific dermal absorption rate data relative to the structure of the skin.
The skin is composed of two layers: the epidermis, a non-vascular layer about 100 urn thick, and the dermis,
a highly vascularized layer about 500 to 3,000 urn thick. The outermost layer of the epidermis, the stratum corneum
is about 10-40 um thick. This layer is thought to provide the major barrier to the absorption into the circulation of
most substances deposited on the skin layer into the circulation. It is composed of dead, partially desiccated and
keratinized epidermal cells. Below this layer lies the viable epidermis, a region about 50-100 um thick, containing
at its base the germinative or basal cell layer whose cells move outward to replace the outer epidermis as it wears
away. This layer generates about one new cell layer per day, which results in the stratum corneum becoming totally
replaced once every two to three weeks. The basal layer epidermis contains enzymes that metabolize certain
penetrant substances. Enzymes may also be active in the stratum corneum (Marzulli et al., 1969) if cofactors are
not required.
Below the epidermis lies the dermis which is a collagenous, hydrous layer. The hair follicles and sweat
ducts originate deep within the dermis and terminate at the external surface of the epidermis. They occupy only
about 1% of the total skin surface, and therefore their role as transport channels for the passage of substances from
3-1
-------
the external environment to the capillary bed is thought to be negligible for most chemicals (Scheuplein and Blank,
1971). The structure of the skin is shown diagrammatically in Figure 3-1.
HAIR SHAFT
OPENING OF
SWEAT GLAND
DERMIS
FATTY _
SUBCUTANEOUS
TISSUE
ADIPOSE
(FAT) TISSUE
NERVE ENDING
SWEAT GLAND
CORNIFIED]
a >^ LAYER
BASAL
LAYER
• EPIDERMIS
SEBACEOUS GLANO
•\HAIR ERECTOR MUSCLE
PAPILLA OF GROWING
HAIR
PAPILLA OF
INACTIVE HAIR
CAPILLARY
Figure 3-1. Structure of the Skin
Source: Ritchie and Carola (1983).
As mentioned above, the stratum comeum is generally considered to be the rate-limiting diffusion barrier
for most compounds. Because of the importance of this layer in determining the rate and extent of dermal
absorption, the following discussion will focus on its structure and function.
Michaels et al. (1975) have described the stratum comeum as a heterogeneous structure containing about
40% protein (primarily keratin), 40% water, and 15% to 20% lipoidal. Lipids in the stratum comeum exist
principally in the form of triglycerides, fatty acids, cholesterol, and phospholipids. Michaels et al. (1975) have
3-2
-------
conceptualized the stratum comeum as being composed of parallel arrays of proteinaceous cells separated by thin
layers of lipoidal material in a "bricks and mortar" arrangement (Figure 3-2).
f
3i iLi— )C
iLi
Interstitial Llpld Phast Prottlnactous Phast
Figure 3-2. Two-Phase Model of the Stratum Comeum
Source: Michaels et al. (1975).
As outlined in Table 3-1, evidence for this two-phase (lipid and protein) model comes from permeation,
freeze-fracture, histochemical, biochemical, and x-ray diffraction studies.
Table 3-1. Evidence for the Two-Phase Model of the Stratum Comeum
Physicochemical evidence for two pathways of transport of lipid and water-soluble molecules;
• Freeze-fracture morphology,
• Histochemistry and fluorescence staining of iipids in frozen sections;
• Dispersion by lipid solvents;
• Co-localization of lipid catabolic enzymes; and
Membrane isolation and characterization, including x-ray diffraction.
Source: Elias et al. (1987)
Evidence for this two-phase system based on the physicochemical properties of permeable compounds was
offered by Raykar et al. (1988). These researchers reported that solutes with octanol/water partition coefficients less
than 1,000 had similar partition coefficients for whole stratum comeum/water and delipidized stratum corneum/water.
3-3
-------
This finding suggests that these compounds were taken up largely or entirely by the protein domain in the stratum
comeum. Conversely, for highly lipophilic compounds (octanol/water partition coefficient greater than 1,000) an
increasing divergence between whole stratum corneum/water and delipidized stratum corneum/water partition
coefficient values was observed with increasing lipophilicity.
Freeze-fracture and x-ray diffraction studies have demonstrated that the lipid in the stratum comeum is
limited to the interstitial areas; no lipid is found in the cytosol of the keratinized cells. Lipids have also been
identified as being localized to the intercellular areas by histochemical staining. Lipid solvents easily disperse the
stratum comeum into a cellular suspension, thereby lending support for a lipid-rich intercellular area.
32 Fate of Compounds Applied to the Skin
Numerous environmental pollutants are known to permeate the skin's diffusional barrier and enter the
systemic circulation via the capillaries in the dermis. However, as shown in Figure 3-3, compounds that come into
contact with the skin can also experience other fates, including:
• Evaporation from the surface of the skin;
• Penetration into the stratum comeum, followed by reversible or irreversibe binding; or
• Penetration into the viable epidermis, followed by metabolism.
Skin lay«rs
vehicle
Drug
Loss procMsas
-•» surface loss
stratum
viable
eomeum
epidermis
dermis
partition
partition
metabolism 4
irreversible binding
-t» metabolism
Figure 3-3. Transport/Loss Processes Occurring in the Skin
Source: Guy and Hadgraft (1989a)
3-4
-------
In many of the studies used to generate the Kp values found in the deimal database, the extent of dermal
absorption was estimated by measuring the loss of compound from the skin surface. However, if loss processes such
as those presented in Figure 3-3 are occurring, an over-estimation of the extent and rate of dermal absorption will
be made. Therefore, this section reviews not only the processes by which compounds are absorbed across the skin,
but other loss processes as well, to enable the exposure/risk assessor to effectively use the Kp values found in the
dermal
3.2.1 Transport Processes Occurring in the Skin
The two-phase structure of the stratum comeum suggested by the studies reviewed in the previous section
has a marked effect on the permeation of compounds through this diffusional barrier. Penetrant molecules can follow
either an intercellular or transcellular route through the stratum comeum, as shown in Figure 3-4, depending on their
relative solubility and partitioning in each phase.
Intcrctllular
Transccllular
Transapp*ndag«al
Figure 3-4. Major Routes of Diffusion Through the Skin
Source: Guy and Hadgraft (1989a)
Evidence for the existence of separate pathways for hydrophilic and hydrophobic molecules was offered by
Michaels et al. (1975). These investigators reported that when a penetrant exists in both nonionic and ionic forms,
the nonionic (and therefore more lipophilic) form is the better skin penetrant. Flynn (1985) made the same argument
after considering the total literature.
An alternative to transport of a compound through transcellular or intercellular pathways in the stratum
comeum shown in Figure 3-3 is penetration via skin appendages, such as hair follicles, sebaceous glands, and sweat
3-5
-------
glands. These appendages could serve as diffusional shunts through rate-limiting barriers, thereby facilitating the
dermal absorption of topically applied chemicals. However, since they occupy less than 1% of the skin surface for
humans, their role as transport channels for the passage of substances from the external environment to the capillary
bed is often negligible. Nevertheless, for slowly diffusing chemicals, transappendageal absorption may be a
contributing or even dominant pathway of dermal permeation, especially during the period immediately following
application of the compound to the skin (Blank and Scheuplein, 1969). For example, the high follicular density of
the scalp may enhance the follicular absorption route during swimming. Using a hydrophobia compound
(benzo[a]pyrene) and a more polar compound (testosterone), Kao et al. (1988) recently examined the extent to which
the transappendageal route contributes to the dermal absorption of these compounds. For the hydrophobic
benzo[a]pyrene, the rate and degree of dermal permeation was significantly greater in the haired strain (Balbc) than
in the hairless strain (SKH) of mouse. Conversely, there was little difference in the in vitro dermal permeation of
topically applied testosterone between the skin of haired and hairless strains of mice. Hairless animals tend to have
a better developed stratum comeum than haired species and strains within a species. Appendageal penetration was
also discussed by Tregear (1966) and by Guy and Hadgraft (1984). However, although the mice were genetically
related, other differences in skin structure cannot be ruled out to account for the observed differences in percutaneous
absorption.
323 Loss Processes Occurring in the Skin
As mentioned above, a compound coming into contact with the skin not only can cross the diffusional
barrier and be taken up by the systemic circulation, but also can evaporate from the surface of the skin, penetrate
into stratum comeum and become bound, or penetrate into the viable epidermis and become metabolized. These
processes are described below.
Evaporation from the Surface of the Skin
Reifenrath and Spencer (1989) have reviewed the processes that play a role in the evaporation of compounds
from the skin. These include wind, humidity, temperature and vapor pressure. Processes that lead to increased
volatilization of the compound on the skin surface (high wind speed and temperature, low humidity) will accelerate
the loss of compound from the skin, and reduce the available dose for absorption. The role of vapor pressure on
the disposition of a dermally applied compound is demonstrated in Table 3-2. Hawkins and Reifenrath (1984)
showed that evaporation accounted for only 4% of the applied radioactive dose of DDT, a relatively nonvolatile
compound applied to pig skin in vitro, but is responsible for the loss of 65% of a volatile compound such as
3-6
-------
diisopropyl fluorophosphonate. Despite their loss by evaporation from the skin surface, volatile compounds tend to
be good skin peneirants.
Table 3-2.
Comparison of Vapor Pressure and Disposition of Radioactivity After Topical Application
Compound
DDT
Parathion
Malathion
Lindane
Benzole acid
N,N-Diethyl-m-toluamide
Diethyl malonate
Diisopropyl fluorophosphonate
Vapor Pressure
(mm Hg at 20°C)
1.5 x Iff7
4.7 x Iff6
5.5 x ID"6
3.3 x 10'5
3.8 x KT4
1.03 x 1(T3
2.49 x 10-'
5.79 x la1
Evaporation Loss; Percent
of Applied Radioactive Dose
4 ± 5
7 ± 0.6
17 ± 6
26 ± 5
5.7 ± 0.3
21 ± 6
40 ±10
65 ± 8
Source: Hawkins and Reifenrath (1984)
Percutaneous absorption studies are often conducted by covering the site of application with an occlusive
wrap to protect the site or prevent loss of the compound. Application of occlusive wrap will limit the evaporation
of volatile compounds from the surface of the skin, and may therefore increase the extent or rate of percutaneous
absorption of these compounds. For example, Bronaugh et al. (1985) demonstrated that the percutaneous absorption
of a single dose of volatile fragrance compounds in monkeys was increased after occlusion of the application site
with plastic wrap. However, occlusion of the site also increases the hydration of the stratum comeum, which may
also be responsible for the increased absorption seen in this study. The effect of occlusion on percutaneous
absorption was discussed by Buck (1988; 1989).
Because of the potential for evaporation from the skin to affect the rate and extent of percutaneous
I absorption, the exposure/risk assessor is advised to take this factor into consideration when using the Kp values found
iin the dermal database in calculations of dermally absorbed dose in specific exposure scenarios. Although this factor
is perhaps most important to consider when assessing the absorption of a neat compound applied to the skin, volatile
compounds in aqueous solution or a soil medium will also evaporate from those vehicles.
3-7
-------
Binding of Compounds in the Skin
Compounds can be retained in the skin, to some degree, by temporarily partitioning into the protein or lipid
phases of the various dermal strata. Compounds can also be retained in the skin by virtue of either reversible or
irreversible binding to skin tissue. As shown in Figure 3-3, irreversible binding and eventual sloughing of the cells
in the stratum corneum may limit the percutaneous absorption of a compound.
Binding of a compound may also occur in the epidermal or dermal layers of the skin. Such binding, and
the establishment of a reservoir of the compound in the skin, could result in the creation of a pharmacokinetic
compartment with a slow turnover rate.
Wester et al. (1987) recently used compound binding to the stratum corneum as a means to evaluate the total
skin absorption of environmental chemical contaminants in ground and surface water. Using the assumption that
any chemicals bound to the skin will be ultimately absorbed into the body, these investigators underscored the
importance of percutaneous absorption as a route of exposure to environmental pollutants during swimming or
bathing.
Metabolism
Metabolism is an important factor in determining both the rate and amount of percutaneous absorption. The
metabolic activity of the epidermis, in turn, depends on the distribution and activity of specific enzyme systems and
on the rate of chemical diffusion.
As shown in Figure 3-3, metabolism may influence the bioavailability of topically applied compounds. The
skin may act as a site of "first pass" metabolism serving, in most cases, to assist in chemical detoxification. For
example, in vitro studies involving topically applied benzo(a)pyrene (BaP) and testosterone to viable human, mouse,
rat, guinea pig, rabbit, and marmoset skin were carried out by Kao et al. (1985) to demonstrate the importance of
metabolic processes. Moreover, it has been shown (Kao et al., 1984) mat enzyme induction, via pretreatment in vivo
with TCDD, affects cutaneous metabolism in vitro following treatment with lipophilic compounds such as BaP. A
recent study by Thohan et al. (1989) demonstrated that pretreatment with Arochlor 1254 administered
intraperitoneally results in a greater degree of 7-ethoxycoumarin deethylase induction in skin microsomes than in
hepatic microsomes in the rat
3-8
-------
In their review of cutaneous xenobiotic metabolism, Noonan and Wester (1989) have strongly argued for
greater emphasis on metabolic activity versus sole reliance on passive diffusion processes. According to their thesis,
the lipid-saturated stratum corneum, the primary diffusional barrier, acts as a sink for lipophilic compounds such as
BaP and steroids. The metabolic activity of the viable epidermis may then be the rate-limiting factor affecting
delivery of the compound to the vasculature in the dermis. While it is often assumed that skin metabolic rates are
significantly lower than hepatic rates, the activity ratio of epidermis (primary cutaneous metabolic site) to liver is
comparable for certain enzymes such as aromatic hydrocarbon hydroxylase (AHH), 7-ethoxycoumarin deethylase,
aniline hydroxylase, and NADP-cytochrome c reductase (Noonan and Wester, 1989). Such activity comparisons may
be misleading, however, unless they take into account heterogeneity in enzyme distribution.
Numerous studies have been carried out demonstrating cutaneous metabolism of PAHs (polycyclic aromatic
hydrocarbons). Incubation of BaP with human epithelial cell culture has yielded metabolites such as 3- and 9-
hydroxy-benzo(a)pyrene, 7,8- and 9,10-dihydrodiol derivatives, and 1,6-, 6,12-, and 3,6-quinone derivatives (Fox et
al., 1975). While Pohl et al. (1976) determined the AHH activity to BaP in whole skin to be 2% of that in the liver,
Noonan and Wester (as mentioned above) have estimated the epidermal AHH activity as 80% relative to the liver,
a far more significant contribution even assuming variation in enzyme distribution. This assumes that all activity
is in epidermis and that epidermis is 2.5% by weight of whole skin.
Skin has also been shown to contain PAH detoxification enzyme systems. Bentley et al. (1976) have
identified the presence of epoxide hydrase in rat skin, an enzyme involved in the detoxification of benzopyrene
epoxides via conversion to their corresponding dihydrodiols. Skin tissues also contain conjugating enzyme systems
capable of enhancing elimination of compounds such as BaP. Glucuronidation of hydroxylated benzo(a)pyrene has
been reported by Harper and Calcutt (1960). While sulfate conjugation of certain steroids has been reported (Berliner
et al., 1968; Faredin et al., 1968), similar sulfate conjugation for PAHs has not yet been shown to occur.
Other detoxification systems such as hydrolytic esterases have been found in epidermal tissue. Chellquist
and Reifenrath (1988) evaluated the in vitro distribution and fate of diethyl malonate using both normal and heat-
treated pig skin. Nearly complete hydrolysis of diethyl malonate occurred in normal tissue. Human, guinea pig, and
rat skin also contain such hydrolytic esterases. Diflucortolone valerate (DFV) was rapidly hydrolyzed in vitro using
guinea pig or rat skin (t1/2 = 30-60 minutes), but it was slowly hydrolyzed using human skin (5-15% metabolism in
7 hrs). Esterases seem to be primarily concentrated in the epidermal layer of the skin. Tauber and Rost (1987)
reported that esterase activity for the hydrolysis of steroids esterified at the 21 position is 20-fold greater per unit
volume in the epidermis than in the dermis. However, due to the greater overall mass of the dermal layer, total ester
hydrolysis is approximately the same in the epidermal and dermal tissues.
3-9
-------
Deamination and dealkylation reactions in skin have also been repotted. Hakanson and Moller (1963)
incubated norepinephrine with rat, rabbit, mouse, and human skin. They demonstrated the availability of monoamine
oxidase by identifying the deaminated metabolite dehydroxymandelic acid. Similarly, Pohl et al. (1976) have
identified the presence of mixed function oxidase dealkylating activity fordeethylation of 7-ethoxycoumarin in mouse
skin.
Cutaneous acetylation reactions have also been demonstrated in hamster skin (Kawakubo et al., 1988). A
high correlation of N-acetylation activity was observed between skin and liver for metabolism of 2-aminofluorene
(2-AF) and p-aminobenzoic acid (PAB A). The importance of such metabolic activity in the skin is underscored by
the role of N-acetylation in activation of carcinogenic arylamines. Acetylating enzymes in the skin have also been
shown to reduce azo bonds during in vitro percutaneous penetration studies (Collier et al., 1989b).
A method for maintaining viability of skin in diffusion cells for studying metabolism in conjunction with
percutaneous absorption was published by Collier et al (1989a). Using this method, the skin absorption/metabolism
of numerous compounds has been studied (Bronaugh et al., 1989; Storm et al., 1990; Nathan et al.. 1990).
Approximately 5% of absorbed butylated hydroxytoluene (BHT) and acetyl ethyl tetramethyltetralin (AETT) were
metabolized while no detectable metabolism of caffeine, DDT, and salicylic acid was seen in hairless guinea pig skin
(Bronaugh et al., 1989). Only small amounts of absorbed benzo(a)pyrene and 7-ethoxycoumarin were found to be
metabolized in rat, fuzzy rat. hairless guinea pig, mouse, and human skin (Storm et al., 1990). However,
p-aminobenzoic acid (PABA) and benzocaine were extensively acetylated on the primary amino group during
percutaneous absorption in the hairless guinea pig and human (Nathan. 1990).
Processes such as pathway-specific transport through the stratum corneum, evaporation from the surface of
the skin, binding in the stratum comeum, and metabolism in the epidermis all affect the extent to which compounds
are absorbed by the skin, as well as the rate of percutaneous absorption. These processes have been discussed in
Section 32.2 as factors that can result in loss of compound from the surface of the skin. However, there are
numerous other factors that the exposure/risk assessor should be aware of that affect the process of percutaneous
absorption. These additional factors are addressed in Section 33.
3.3 Factors That Influence Percutaneous Absorption
The rate and amount of percutaneous absorption of a compound depend highly on the physical-chemical
nature of both the skin and the compound that comes into contact with the skin. This section reviews how dermal
3-10
-------
factors (e.g., skin thickness, hydration, temperature) and compound-specific factors (e.g.. lipophilicity, polarity.
volatility, and solubility) are involved in determining the rate and amount of absorption by the dermal route.
3 J.I Skin-Specific Factors
The use and evaluation of Kp values in the database requires the exposure/risk assessor to know the
conditions under which the results were obtained. (See Section 5.1 for definition of Kp.) As discussed below, factors
(e.g., species, site of application and the age, degree of hydration, temperature, condition of the skin) can have a
marked effect on the extent and rate of percutaneous absorption.
Site of Application or Exposure
A common assumption used in dermal exposure assessment is that Kp values obtained from one site of
application on the body are appropriate for all areas of the skin where percutaneous absorption may occur. However,
as reported by Feldmann and Maibach (1967), and shown in Figure 3-5, the extent of absorption of a compound such
as hydrocortisone in humans is dependent on the anatomical site to which the compound is applied.
(Ventral)
Forearm (DortoD
Fool Arch (Plantar)
Ankle (lateral)
Bock
Scotp
Anlla
Forehead
Jew Angle
Scrotum
HYDROCORTISONE ABSORPTION—totol
•ffoct of onolomn
Figure 3-5. Regional Variation in the Percutaneous Absorption of Hydrocortisone in Humans
Source: Feldmann and Maibach (1967)
3-11
-------
Feldmann and Maibach later extended their investigation to include pesticides (Maibach et al., 1971). As
shown in Table 3-3, a marked variation exists in the dose of parathion and malathion absorbed at different anatomical
sites in humans.
Table 3-3. Effect of Anatomical Region on In Vivo Percutaneous Absorption
of Pesticides in Humans
Anatomical Region
Forearm
Palm
Foot, ball
Abdomen
Hand,dorsum
Forehead
Axilla
Jaw angle
Fossal cubitalis
Scalp
Ear canal
Scrotum
Parathion
Dose Absorbed
(Percent)
8.6
11.8
13.5
18.5
21.0
36.3
64.0
33.9
28.4
311
46.6
101.6
Malathion
Dose Absorbed
(Percent)
6.8
5.8
6.8
9.4
12.5
23.2
28.7
69.9
Source: Maibach et al. (1971)
While the data of Table 3-3 show that human palm and forearm skin are of comparable resistance to skin
penetration by parathion and malathion, in vivo, data with another compound (tri-n-butylphosphate) suggest that
plantar skin is considerably more protective than anterior forearm when tested in vitro with this more hydrophilic
compound (Marzulli, 1962). (Palmar and plantar skin are thought to be alike in physicochemical and structural
makeup). The data of Marzulli (1962) compare the permeability of scrotum, post auricular, scalp, thigh, instep,
anterior forearm, plantar, chest, abdomen, ankle, leg, and nail. Regional variation effects appear to be modified in
accordance with the type of compound involved (hydrophilic or hydrophobia).
3-12
-------
Many of the K,, values reported in the database come from studies using experimental animals. As expected,
the percutaneous absorption of compounds also demonstrates regional variation in experimental animal species.
Franklin et al. (1989) have shown that the absorption of a series of pesticides applied to the foreheads of rhesus
monkeys was approximately twice that observed when the compound was applied to the forearms of these animals,
as shown in Table 3-4.
Table 3-4. Percutaneous Absorption in Monkeys
as Related to Site of Application and Test Compound
Compound
Aminocarb
Azinphosmethyl
Diethyl toluamide
Fenitrothion
Cis-permethrin
Trans-permethrin
Forehead
Percent Absorbed
74
47
33
49
28
21
±
±
±
±
±
±
4
10
11
4
6
3
Forearm
Percent Absorbed
37
32
14
21
9
12
±
±
±
±
±
±
14
9
5
10
3
3
Source: Franklin et al. (1989)
Variations in percutaneous absorption between different body sites have been reported in the rat, a
commonly used species in percutaneous absorption studies. For example, Bronaugh et al. (1983) have shown that
differences in the permeability of rat skin may be related not only to body site, but also to the sex of the animal.
Differences have been observed in the measured permeability constants for water, urea, and cortisone in excised male
and female rat skin taken from the back of the animals, and between permeability constants measured using
abdominal and back skin from male rats (Table 3-5). However, body site differences in skin permeability are not
always observed for some species. For example, Behl et al. (1983b) found relatively little difference in permeability
coefficients of methanol obtained using abdominal or dorsal skin from the hairless mouse.
3-13
-------
Table 3-5. Effect of Gender and Body Site on the Permeability of Rat Skin
Male
Water
Back
Abdomen
Urea
Back
Abdomen
Cortisone
Back
Abdomen
Permeability
Constant
(cm/hr ± SD)
0.00049 ± 0.00004
0.00131 ± 0.00021
0.00016 ± 0.00005
0.00188 ± 0.00055
0.00017 ± 0.00004
0.00122 ± 0.00006
Lag
Time
(hr ± SD)
2.4 ± 0.1
1.7 ± 02
15.0 ± 1.8
16.5 ± 43
33.4 ± 4.4
32.9 ± 2.4
No. of
Determin-
ations
7
4
6
4
8
4
Female
Permeability Lag No. of
Constant Time Determin-
(cm/hr ± SD) (hr ± SD) ations
0.0009310.00011 10 ±0.1 4
0.00048 ± 0.00013 11.1 ±0.6 3
0.00047 ± 0.0001 1 20.0 ± 2.6 3
Source: Bronaugh et al. (1983)
Although gender-related permeability differences have not been measured directly in humans, animal data
that demonstrate gender differences are frequently noted in toxicity studies, and these differences are taken into
account when extrapolating animal toxicity to humans. Any regional permeability differences that are observed may
be due to the gender- and site-related differences in the thickness of the stratum comeum and/or whole skin. For
example, site- and sex-specific differences in the stratum corneum thickness in the rat, as shown in Table 3-6. may
explain the results reported in Table 3-5. A competent stratum comeum is expected to provide better barrier capacity
than a thick, disorganized stratum comeum. Thus, thickness is not the only measure of regional variation in skin
permeability.
3-14
-------
Table 3-6. Rat Skin Thickness Measurement from Frozen Sections
Type of Skin
Male
Back
Abdomen
Female
Back
Abdomen
Stratum
Corneum
(urn)
34.7 ± 2.3
13.8 ± 0.7
18.2 ± 1.0
13.7 ± 0.6
Whole
Epidermis
(urn)
61.1 ± 3.0
30.4 ± 1.5
31.2 ± 1.5
34.8 ± 1.8
Whole
Skin
(mm)
2.80 ±0.08
1.66 ±0.06
2.04 ±0.05
0.93 ±0.02
Source: Bronaugh et al. (1983)
As shown in Table 3-7, similar site-specific differences in skin thickness exist in humans as well.
Table 3-7. Regional Variation in Stratum Comeum Thickness in Humans
Skin Area Stratum Comeum Thickness
um
Abdomen 15.0
Volar forearm 16.0
Back 10.5
Forehead 13.0
Scrotum 5.0
Back of Hand 49.0
Palm 400.0
Sole 600.0
Source: Scheuplein and Blank (1971)
Despite the important implications that stratum comeum thickness may have on body site-related variations
in the rate or extent of percutaneous absorption, Elias et al. (1981) found site-specific variations in permeability to
be directly proportional to the lipid content of the stratum comeum. This factor undoubtedly plays a significant role
in determining the rate and extent of percutaneous absorption for highly lipophilic compounds.
Regional differences in the extent of percutaneous absorption can have a significant impact on calculations
of dermally absorbed dose and on any subsequent risk assessment that uses these values. Guy and Maibach (1984)
have proposed a methodology for incorporating regional permeability differences into the assessment of dermally
absorbed dose.
3-15
-------
Age of the Skin
Infants and children represent a population at high risk for the toxic effects of environmental pollutants
because of, among other reasons, their immature detoxification pathways and rapidly developing nervous systems.
Infants and children are also at increased risk for dermal exposure to toxic compounds because of their greater
surface-to-volume ratio. Reports of toxic effects occurring in infants after the topical application of various drugs
or pharmaceutical agents are not uncommon in the literature. These toxic effects, however, are most likely the result
of the increased surface-to-volume ratio in infants that results in greater total absorption of the compound, and not
to the increased permeability of the skin of infants relative to that of adults. Full-term infants have been shown to
have a completely functional stratum corneum with excellent barrier properties (Atherton and Rook, 1986).
To investigate possible age-related changes in dermal permeation, Wester et al. (1985) compared the in vitro
percutaneous absorption of triclocarban in adult and newborn abdominal and foreskin epidermal preparations
(split-thickness, frozen skin samples). Using a static diffusion cell system at 37°C, these researchers showed that
the total dose of triclocarban absorbed across abdominal skin preparations excised from human adults, a newborn
infant (S-day-old female donor) and an older infant (9-month-old male donor) was similar, as shown in Table 3-8.
Table 3-8. In Vitro Percutaneous Absorption of Triclocarban in Human Adult
and Newborn Abdominal and Foreskin Epidermis
Type Dose Absorbed
(% ± SD)
Static system, 37°C
Adult abdominal 0.23 ± 0.15
Newborn abdominal 0.26 ± 0.28
Infant abdominal 0.29 ± 0.09
Adult foreskin 0.60 ± 0.2S
Newborn foreskin 2.5 ± 1.6
Static system, 23°C
Adult abdominal 0.13 ± 0.05
Continuous flow system, 23°C
Adult abdominal 0.6 ± 2.0
Human in vivo 0.7 ± 2.8
Source: Wester et al. (1985)
3-16
-------
Unlike full-term infants, pre-term infants may demonstrate an increased percutaneous absorption of some
compounds, probably because of the lack of a fully developed stratum corneum. Barker et al. (1987) investigated
the effect of gestational age on the permeability of an infant's skin to sodium salicylate. They showed that
absorption of this compound across excised, full-thickness abdominal skin was several orders of magnitude greater
in infants of 30 weeks of gestation or less than in full-term infants. The epidermis of these pre-term infant skin
samples was thin, with relatively little formation of a keratinized stratum corneum.
Although dermal permeability remains relatively invariant in humans as a function of age, at least one
experimental animal species, the hairless mouse, undergoes a period early in life where the percutaneous absorption
of some compounds is increased. Behl et al. (1984) reported that a three- to five-fold increase in permeability of
hairless mouse skin occurs in animals less than 120 days of age relative to that observed in older animals.
At the other end of the age spectrum, older adults also constitute a high-risk category for the toxic effects
of environmental pollutants. Although the effects of increasing age on the gastrointestinal absorption of toxic
chemicals have been addressed, the effects of aging in humans on the percutaneous absorption of toxic compounds
appear to have been largely ignored. However, Banks et al. (1989) have recently reported that the percutaneous
absorption of TCDD and 2,3,4,7,8-pentachlorodibenzofuran decreases as a function of increasing age in male F-344
rats. These findings suggest that the potential for systemic toxicity occurring in older animals after dermal exposure
to these nalogenated hydrocarbons is reduced. Behl et al. (1983) also observed a slightly reduced permeability of
hairless mouse skin from older animals (441 days) to phenol; however, these investigators concluded that any decline
in permeability was probably the result of animal variability rather than age. A review of the world's literature on
skin permeability as related to age suggested that age-related differences in skin permeability (child to adult) are
generally less than species-related differences (mouse to human). The evidence suggests that old and young skin
are alike in barrier function (Marzulli and Maibach, 1984).
Skin Condition
For most compounds, the rate of percutaneous absorption is limited by diffusion through the stratum
corneum. However, the epidermal barrier may not be intact in diseased or damaged skin. Persons with diseased
or damaged skin may be at special risk for the toxic effects of environmental pollutants as a result of increased
percutaneous absorption. Damage to the skin may occur from mechanical injury (cuts, wounds, abrasions) or other
insults such as sunburn. Any skin condition that compromises the capability of the stratum corneum to serve as a
permeability barrier, including psoriasis, eczema, rashes, or dermatitis, may also result in increased percutaneous
absorption in affected individuals (Brown et al., 1984).
3-17
-------
A number of studies have been conducted to quantify chemical absoiption through abnormal skin in vitro
and in vivo. Increased absorption of hydrocortisone (Solomon and Lowe. 1979) and propylene glycol (Komatsu and
Suzuki, 1982) have been observed in vitro using damaged hairless mouse and rabbit skin, respectively. However,
the in vitro model may be inadequate for studying this process. While increased blood flow may not affect
percutaneous absorption through normal skin, it may indeed affect the absorption rate through skin denuded of
stratum comeum. Also, it is not currently possible to prepare split-thickness epidermal sheets from skin in which
the stratum comeum has been stripped off (Scott and Dugard, 1986). These limitations have largely restricted
investigators in this field to the use of in vivo studies.
Increased in vivo absorption of mannitol and octyl benzoate have been observed in tape-stripped rat skin.
Tape stripping removes the stratum comeum and provides a simple model for a psoriatic or eczematous state in skin.
Using monkey skin affected by eczematous dermatitis, Bronaugh et al. (1986) demonstrated a doubling of the total
absorption of hydrocortisone across diseased skin versus that across normal skin sites. However, this effect is not
seen with triamcinolone acetonide, whose absorption through normal skin is so rapid that there is no increase when
applied to damaged skin. Earlier work in this laboratory (Bronaugh and Stewart, 1985) also demonstrated that the
greatest increases in dermal penetration in damaged skin were observed for compounds that were poorly absorbed.
The barrier layer of the skin cannot only be damaged by disease processes or mechanical injury, but also
by exposure to the chemical penetrant itself, especially at high concentrations. The effect of concentration on the
permeability of the skin to a particular compound is reviewed in Section 3.3.2.
Hydration
As discussed previously, the thickness of the stratum comeum is a major determinant of the dermal
permeation. The thickness of the stratum comeum is inversely proportional to permeability. However, thickness
of the stratum corneum in vivo and in vitro is positively correlated with the relative environmental humidity and
degree of hydration of this layer.
Therefore, one would expect well-hydrated skin to be less permeable than relatively dry skin as a result of
its increased thickness. This, however, is not generally the case. As a rule, hydration increases the permeability of
skin for most compounds. Therefore, there is an increased potential for percutaneous absorption of environmental
pollutants in scenarios such as bathing, swimming or showering where the skin is well hydrated.
3-18
-------
Circulation to the Skin
Prolonged dermal exposure to organic solvents is known to result in vasodilation in areas that come into
contact with these compounds (e.g., Engstrom et al., 1977). If the rate of chemical accumulation in the epidermis
(via diffusion across the stratum corneum) is equal to or greater than the circulatory perfusion rate, then the rate-
limiting step for dermal permeation could become that of capillary transfer. The relationship between the rates of
capillary transfer and diffusion can be described by the following equation (Scheuplein and Blank, 1971):
-D *L - 0 L (Cm - O
dx
(3.1)
where:
C. e Concentration of the diffusing compound in arterial blood;
C = Concentration of the diffusing compound in tissue adjacent to the capillary walls;
L * The thickness (cm) of the layer beneath the stratum comeum;
0 * Peripheral blood flow;
D «= Average membrane diffusion coefficient; and
dc
_ e Change in chemical concentration over the change in unit distance through the layers.
ax
The symbol 0 represents the transfer coefficient into capillary circulation and, in practical terms, is
equivalent to the resistance to capillary transfer. If this resistance is small relative to resistance to diffusion across
the stratum comeum, then the latter would be the rate-limiting step. For all situations except those involving gases
and small, highly lipophilic compounds, the diffusion resistance is likely to be substantially greater than capillary
resistance. Thus, circulatory flow should not be rate limiting in most cases.
Skin Temperature
The assumption in using the values in the dermal database for compounds in aqueous media is that
temperature has little effect on the Kp or flux (J). However, humans are exposed to ambient or drinking water
supplies during activities such as bathing, showering, or swimming that may differ markedly from skin surface
3-19
-------
temperature. Therefore, it is important to consider the potential impact that water temperature may have on the rate
or extent of percutaneous absorption of the compound of interest
Jetzer et al. (1988) recently examined temperature-related changes in K,, using hairless mouse skin in a
diffusion cell apparatus. Their results for three model compounds are presented in Table 3-9.
Table 3-9. Effect of Temperature on Permeability Coefficients for Model
Compounds Permeating Hairless Mouse Skin In Vitro
Temperature (°Q
Pei-meant of Donor1
n-Butanol 10
20
30
37
Phenol 10
20
30
37
p-Nitrophenol 10
20
30
37
K
[cm/hr]
0.00237
0.00470
0.00805
0.01432
0.01602
0.01932
0.02881
0.04375
0.00289
0.00608
0.00109
0.01753
}± s.d.)
(0.00117)
(0.00025)
(0.00180)
(0.00239)
(0.00109)
(0.00270)
(0.00148)
(0.00020)
(0.00033)
(0.00046)
(0.00010)
(0.00237)
* receptor fluid temperature = 37°C
Source: Jetzer etal. (1988)
Keeping the receptor solution at 37°C to mimic the physiological state, but exposing the stratum corneum
to aqueous solutions of the compound at temperatures from 10°C to 37°C allowed these investigators to represent
environmentally relevant exposure scenarios with this in vitro test system. As shown in Table 3-9, for these three
compounds, Kp varies three- to seven-fold as a function of donor solution temperature. Durrheim et al. (1980), as
well, have shown that percutaneous absorption rates can vary over the temperature range of 29°C to 37°C. At the
other end of the temperature spectrum, Loron and Cohen (1984a) have reported a two-fold increase in Kp in vitro
going from a donor solution temperature of 37°C to 50°C. Therefore, the potential exists for percutaneous absorption
to increase in vivo when skin temperatures are elevated during bathing or showering with warm water. Frequently,
in the absence of skin damage, a 10-fold increase in temperature results in a doubling of skin permeability.
3-20
-------
These results suggest that the exposure/risk assessor should take skin temperature into consideration when
using the Kp values in the database to estimate dermally absorbed dose in specific exposure scenarios; however, this
factor probably changes Kp values by less than an order of magnitude.
Miscellaneous Factors
In addition to the variables discussed in this section, there are several other factors that may affect the rate
and degree of dermal penetration, including release of the compound from the vehicle in which it is formulated and
multiple versus single-dose application. These factors are potentially important and should be considered by the
exposureAisk assessor where applicable. For a more detailed examination of these topics, the reader is directed to
reviews that address these topics in greater detail (e.g., Wester and Maibach, 1983).
3.3.2 Compound-Specific Factors
In addition to the skin-specific factors discussed above, the physicochemical nature of the penetrant
compound also plays a role in the rate and extent of absorption of that compound. These factors are reviewed below.
Partition Coefficients
The best penetrants are those that are soluble in both lipids and water, whereas, compounds that are largely
soluble only in either lipids or water, but not both, are not good penetrants. The relative solubility of a compound
in an organic or water phase can be represented by a partition coefficient. Several investigators have attempted to
demonstrate a correlation between percutaneous absorption and partitioning behavior. The use of this approach to
predict Kp from partitioning behavior is explored in Section 8.0; however, the results reported in Table 3-10 by
Roberts et al. (1977) are illustrative of these efforts. As shown in Table 3-10, Kp values tend to increase with
increasing lipid solubility. This relationship also exists for compounds such as steroids (Scheuplein et al., 1969).
The most commonly used measure of partitioning behavior is the octanohwater partition coefficient (KDW)
or its logrithmic form (log K,,w). However, discrepancies have been noted in the relationship between skin
permeability and lipophilicity as expressed by the log K,w for some compounds, notably certain phenols (Jetzer et
al., 1986). This lack of correlation is particularly striking for the nitrophenols. However, when "oiT/water (o/w)
partition coefficients based on either n-hexane, dichloromethane, chloroform, or silicone rubber as the
water-immiscible phase are used, permeability coefficients for the various phenolic compounds follow the expected
dependency on partitioning (Jetzer et al., 1986). On this basis, it appears that the log K^ may not always properly
reflect the lipophilicity of certain classes of chemicals, and, thus, may be an inconsistent predictor of skin
3-21
-------
permeability. Furthermore, these various partition coefficients are in themselves, individual and unique measures
of lipophilicity, and should not be used interchangeably.
Table 3-10. In Vitro Permeability Coefficients and Partition
Data for Various Phenol Compounds
Solute
Resorcinol
Nitrophenol
Nitrophenol
Phenol
Methyl hydroxybenzoate
Cresol
Cresol
Cresol
Naphthol
Chlorophenol
Ethylphenol
3,4-Xylenol
Bromophenol
Chlorophenol
Thymol
Chlorocresol
Chloroxylenol
2,4,6-Trichlorophenol
2,4-Dichlorophenol
(cm/min)
0.000004
0.000093
0.000094
0.000137
0.000152
0.000254
0.000262
0.000292
0.000465
0.000551
0.000581
0.000600
0.000602
0.000605
0.000880
0.000916
0.000984
0.000990
0.001001
LogK.w
0.8
1.96
2.00
1.46
1.96
1.96
1.95
1.95
2.84
2.15
2.40
2.35
2.59
2.39
3.34
3.10
3.39
3.69
3.01
Source: Roberts et al. (1977)
3-22
-------
The capacity of a substance to penetrate the skin is at least partially dependent on the polarity of a
compound; that is, the extent to which the substance, at a molecular level, is associated with a nonsymmetrical
distribution of electron density. Polar compounds are generally poorly absorbed through the skin, whereas nonpolar
compounds are more readily absorbed. The extent of the polarity of a molecule can be expressed quantitatively
through its dipole moment, which is a function of the magnitude of the partial charges on the molecule and the
distance between the charges. The degree of polarity associated with a molecule is a function of spacing and
proportion of electronegative atoms (e.g., nitrogen, oxygen, fluorine), particularly if they are tonizable, versus the
occurrence of nonelectronegative atoms (e.g., hydrogen, carbon). Thus, placing an electronegative functional group
on a non-polar compound will increase its polarity, but in many cases, the molecular size and structure will also
determine a compound's polarity. The greater the polarity of a compound, the less is the lipophilicity; lipophilicity
can most readily be measured based on partition coefficients (see above).
The most polar compounds (i.e., those least able to penetrate the skin) are those which spontaneously
dissociate to form ions in an aqueous environment; such compounds are referred to as electrolytes. Electrolytes can
be inorganic salts, which are readily dissociated, or weak organic acids and bases, whose state and extent of
ionization depend on the pH of the environment Weak organic acids or bases in their non-ionized form are much
more soluble in lipids and are absorbed more readily through the skin than when in their ionized forms. Generally,
the smaller the pK, for an acid and the larger the pK, for a base, the more extensive will be the dissociation in
aqueous environments at normal pH values, and the greater will be the electrolytic nature of the compound. Thus,
the potential for absorption through the skin can be at least qualitatively determined from the ratio of ionized to
unionized compound as defined by the Henderson-Hasselbach equation:
„//-,*„ +log fo"^
[unionized]
Several investigators have shown that electrolytes in dilute solution (and therefore in the ionized form)
penetrate the skin poorly. It is interesting to note that small ions such as sodium, potassium, bromine, and aluminum
penetrate the skin with permeability constants of about 10'3 cm per hour, similar to the rate reported by Scheuplein
(1965) for water. Wahlberg (1968) and Skog and Wahlberg (1964) reported similar results for the chloride salts of
Cobalt, Zinc, Cadmium, and Mercury; sodium chromate; and silver nitrate applied to guinea pig skin in vivo or in
vitro.
3-23
-------
However, larger nonpolar compounds that would be expected to exist as non-electrolytes in aqueous solution
(e.g., urea, thiourea, glucose, glycerol) permeate nude mouse skin in vitro with Kp values on the order of 10~* cm/hr
(Ackermann and Flynn, 1987).
Since the ionization state is a function of the pH of the applied solution, changes in pH can affect the
penetration of an ionizable compound. For example, Wahlberg (1971) snowed marked variations in the absorption
of chromium from unbuffered solutions with a wide range of pH values when applied to guinea pig skin. The
dramatic effect that ionization state can have on permeability was demonstrated by Huq et al. (1986) for a series of
phenolic compounds as shown in Table 3-11.
Table 3-11. Permeability of Hairless Mouse Skin to
Selected Phenols as a Function of pH
Permeant pK,
4-Nitrophenol 7.15
2,4-Dinitrophenol 3.%
2,4,6-Trichlorophenol 6.0
Donor
PH
3.46
6.20
7.56
10.16
2.0
3.5
3.5
4.35
4.65
6.0
7.7
5.0
6.0
7.4
(cm/nr)
0.0012
0.0011
0.0007
0.00005
0.0151
0.0116
0.0105
0.00506
0.00326
0.000315
0.0
0.0174
0.0087
0.00409
Source: Huq et al. (1986)
As shown in Table 3-11, essentially no 2,4-dinitrophenol permeates hairless mouse skin in vitro in aqueous
solution at pH levels comparable to those found in environmentally relevant dermal exposure scenarios. In contrast,
at pH levels at or below the pKa for this compound, mouse skin is fairly permeable to 2,4-dinitrophenol.
In addition, if the pH value of the applied solution results in very acidic or alkaline conditions on the skin,
there is a potential to increase the rate of absorption of a compound because of destruction of the barrier layer (Zatz,
1983).
3-24
-------
Chemical Structure
Changes in chemical structure across a series of homologous compounds have the potential to alter the
permeability characteristics of these compounds. For example. Blank et aL (1967) demonstrated the effect of
increasing chain length on the permeability coefficient of aqueous solutions of normal alcohols (Table 3-12). This
change in K,, is most likely a result of the increase in lipophilicity.
Table 3-11 Permeability of Human Skin (In Vitro) to Alcohols
Compound
(Aqueous Solutions)
Water
Methanol
Ethanol
Propanol
Butanol
Pentanol
Hexanol
Heptanol
Octanol
Nonanol
V
(cm/hr)
0.0005
0.0005
0.0008
0.0014
0.0025
0.0060
0.0130
0.0320
0.0520
0.0600
U.I.
-1.38b
-0.77b
-0.31b
0.30°
0.65b
1.56b
2.03b
2.41b
2.97b
3.47d
* Blank et al. (1967)
b Hansch-Leo Log P Database (online database). Pomona College Medicinal Chemistry Project (data from hard
copy printout).
c Information System for Hazardous Organics in Water (online database). Baltimore, MD: Chemical Information
System.
d CHEMEST or AUTOCHEM Estimation.
Schaefer et al. (1987) have also shown how minor modifications in chemical structure can markedly alter
the percutaneous absorption of a series of closely related androgens. For example, the addition of two hydrogen
molecules to a double bond in the A ring of testosterone to yield dihydrotestosterone, results in a 30-fold decrease
in the relative absorption of the latter compound over the former.
3-25
-------
The marked changes in percutaneous absorption that may result from small structural differences point out
the hazards of using values for structurally similar compounds in dermal exposure assessment, in the absence of
experimentally derived values for the compound of interest
The molecular size and weight of a compound appear to have less of an effect on the rate or extent of
percutaneous absorption than lipophilicity. Large maoomolecules penetrate skin slowly because of a combination
of molecular size and poor lipid solubility. Summarizing the work of several researchers, Grasso and Lansdown
(1972) have noted that macromolecules such as colloidal sulfur, albumin, dextran, and polypeptides penetrate the skin
poorly if applied in an aqueous solvent However, these macromolecules will permeate the skin more readily if
applied in a solvent with high lipid solubility.
As mentioned in Section 3.22, volatilization of a compound from the surface of the skin represents a
process that may result in loss of the applied compound. Since volatilization of the compound will alter the amount
on the skin surface available for absorption, estimates of percutaneous uptake should account for this loss process.
Volatilization can be prevented in experimental studies by the application of occlusive wraps or devices over
the site of compound exposure. However, occlusion generally results in enhanced absorption of the test compound.
The relevance of absorption data obtained in studies where occlusive wraps or devices are used must be questioned
when the data are to be used in exposure/risk assessment because they may result in overly conservative estimates
of percutaneous absorption.
Compound Concentration
A major determinant of the amount of a compound absorbed across the skin is the concentration or the
amount of the compound at the skin surface. Wester and Maibach (1976*) demonstrated that the total amount of a
compound absorbed increases as a function of the applied amount, as shown in Table 3-13.
3-26
-------
Table 3-13. Percutaneous Absorption of Topical Doses of Several Compounds
in the Rhesus Monkey
Time
(hr)
0-24
24-48
48-72
72-96
96-120
Total %
SD
Total ug
# Animals
Testosterone (vz/cm2)
40
2.5"
1.7
1.1
0.8
0.6
6.7
4.2
2.7
6
250
0.5
1.2
0.6
0.4
0.2
2.9
1.4
7.2
3
400
0.5
0.8
0.4
0.3
0.2
2.2
1.0
8.8
3
1600
0.5
1.2
0.5
0.4
0.3
2.9
1.7
46.4
3
4000
0.1
0.5
0.3
0.3
0.2
1.4
0.8
56.0
3
Hydrocortisone
(uz/cm2)
40
0.9
0.7
0.3
0.1
0.1
2.1
0.6
0.84
3
Benzole acid*
(ue/cm2)
40
29.9
3.0
0.4
0.2
0.1
33.6
5.1
134.4
3
2000
13.5
2.8
0.7
0.3
0.2
17.4
1.2
348.0
3
* Values for benzole acid are not corrected for incomplete urinary excretion. All other values (that is, for the other
chemicals) are corrected.
b All data are presented as the percentage of the applied dose that was absorbed.
Source: Wester and Maibach (1976).
Furthermore, when Pick's simple law of diffusion is applicable, skin penetration at steady state is
proportional to the concentration (driving force) of the penetrant (Tregear, 1966). Pick's law does not apply when
the penetrant damages the skin.
Liron and Cohen (1984) reported that the penetration of propionic acid from n-hexane solution through
porcine skin in vitro was relatively greater at higher concentrations. The authors postulate that this effect may be
the result of a breakdown of the permeability barrier by exposure to the acid used in the study.
The permeability barrier of the skin can also be damaged by the delipidizing effect of organic solvents.
Numerous investigators (Scheuplein and Blank, 1973; Roberts et al., 1977; Baranowska-Dutkiewicz, 1982; Behl et
al, 1983b; Huq et al., 1986), have demonstrated increased flux rates for various organic solvents across both human
and animal skin relative to the permeability of more dilute aqueous solutions of the same compounds. Each of these
researchers has attributed this increased permeability to the delipidization and subsequent damage of the stratum
comeum.
3-27
-------
4.0 TECHNIQUES FOR MEASURING DERMAL ABSORPTION
The permeability coefficient (Kp) and flux (J) values presented in the dermal permeability database have
been generated from both in vivo and in vitro studies that utilized a wide range of experimental techniques. Ideally,
the absorption rate values used in a percutaneous exposure assessment should be obtained under conditions that
mimic environmentally relevant exposure scenarios. However, few studies of this nature have been conducted.
Studies of human subjects are very costly, and the conditions of the experiment more difficult to control. Also,
ethical constraints may rule out the testing of toxic compounds in humans. In the absence of human in vivo
percutaneous absorption values, exposure/risk assessors are required to use available data from in vivo animal and
in vitro animal or human studies. The challenge, therefore, is to extrapolate the results obtained in animal studies
to those expected in humans, and to evaluate the capability of in vitro percutaneous absorption rate values to predict
the percutaneous absorption of toxic compounds in vivo. This section will examine these issues and provide a set
of general guidelines for evaluating absorption data for use in a cutaneous exposure assessment.
Because of the wide variation in the experimental techniques used to obtain percutaneous absorption rate
data, exposure/risk assessors are urged to examine the relevance of the available data for their particular exposure
scenario when estimating percutaneously absorbed dose. For example, since rat skin is generally three to five times
more permeable than human skin, the use of Kp or J data from rat studies may result in a conservative (i.e., higher)
estimate of percutaneously absorbed dose in humans. Conversely, variation in the thickness of the stratum corneum
at different body sites may cause the exposure assessor to underpredict the Kp or J for whole-body exposure if data
from whole-hand immersion studies are used. Therefore, it is important to be aware of factors that affect the rate
or extent of absorption when conducting a cutaneous exposure assessment. Many factors, such as body site variation,
skin metabolism, and binding in the skin, have been addressed in Section 3.0. Other factors that may limit the
unqualified use of Kp or J values in cutaneous exposure assessment are inherent in the experimental technique used
to obtain the values or the design of the study. To assist exposure/risk assessors in their selection of the most valid
Kp values, the studies conducted to obtain the values present in the dermal permeability database are categorized by
experimental technique in Table 4-1. This section will review these techniques, and examine various factors that
may affect how data generated using these methods can be used in a cutaneous exposure assessment These factors
are summarized in Table 4-13 at the end of this Section.
4-1
-------
Table 4-1. Experimental Techniques Used to Obtain Kp or J Values Reported
in the Dermal Permeability Database
Technique*
Reference6
IN VIVO
Quantification of Radioactivity,
Parent Compound or Metabolites in
Excreta
Quantification of Radioactivity
or Parent Compound in Blood,
Plasma or Tissues
Quantification of the Disappearance
of the Compound from the Surface of
the Skin or from the Donor Solution
Measurement of a Biological Response
IN VITRO
Diffusion Cell/Quantification of
Radioactivity in Receptor Solution
Baranowska-Dutkiewicz, 1982
Bronaugh and Stewart, 1986
Dutkiewicz and Tyras, 1967,1968
Engstrom et al., 1977
Guest et al., 1984
Engstrom et aL, 1977
Guest et al., 1984
Johanson et al., 1988
Skog and Wahlberg, 1964
Baranowska-Dutkiewicz, 1981,1982
Dutkiewicz and Tyras, 1967
Frederickson, 1961a,b
Knaak et al., 1984a,b
Lopp et al., 1986
Skog and Wahlberg, 1964
Wahlberg, 1971
Frederickson, 1961a
Ackerman and Flynn, 1987
Behl et al., 1983a,b, 1984
Bond and Barry, 1988
Bronaugh and Stewart, 1986
Bronaugh et al., 1986a,b
DelTerzo et al., 1986
Durrheim et aL, 1980
Frederickson, 1961b
Garcia et al.. 1980
Guest et al., 1984
Huq et aL. 1986
Jetzeretal., 1986.1988
Lopp et al.. 1986
Shackelford and Yielding, 1987
Scheuplein and Blank, 1973
Scott et al., 1987
4-2
-------
Table 4-1. Experimental Techniques Used to Obtain K_ or J Values Reported
in the Dermal Permeability Database (Com)
Technique* Reference1"
Diffusion Cell/Quantification Blank and McAuliffe, 198S
of Parent Compound in Blank et al., 1967
Receptor Solution Dugard et al., 1984
Huq et al., 1986
Jetzer et al., 1986,1988
Roberts et al., 1977
Scheuplein and Blank, 1973
Scott et al., 1987
Southwell et al., 1984
Tsuruta, 1977, 1982
* Techniques for studies that provided Kp values for chemical vapors are described in Section 7.0.
b Citations as they appear in the reference list
This section focuses on techniques to obtain Kp or J values for compounds applied to the skin in neat form
or in various liquid vehicles. Sections 6.0 and 7.0 have been included to examine the methods used to quantify the
percutaneous absorption of soil contaminants and vapors, respectively.
4.1 In Vivo Studies
Quantitative percutaneous absorption rate values can be obtained in living animals and humans using a
variety of techniques. The application of the compound of interest to the skin in vivo may be more physiologically
relevant than the use of in vitro methods. However, in vivo techniques allow only indirect measurement of the
absorption of the compound across the skin. Also, the results of in vivo studies are often reported as the percent of
the applied dose that is absorbed, thereby limiting their use for cutaneous exposure assessment, if a Kp is required.
Various authors (Wester and Maibach, 1986,1989a; Scott and Dugard, 1989) have reviewed the advantages
and limitations of many of the techniques available to obtain in vivo percutaneous absorption rate values. These
advantages and limitations are summarized below.
4-3
-------
4.1.1 Quantification of Radioactivity, Parent Compound, or Metabolite Levels in Excreta
Percutaneous absorption is commonly determined by measuring the appearance of radioactivity in the excreta
following the topical application of a labeled compound. Following application of the radiolabeled compound, the
total amount of radioactivity excreted in urine or urine plus feces is determined. The total radioactivity in the excreta
is a mixture of the parent compound and any labeled metabolites that may have resulted from metabolism of the
parent in the skin and the body.
Any radioactive label retained in the body, or excreted by another route, will not be detected in the urine
or feces. Therefore, Feldmann and Maibach (1969, 1970, 1974) used the following expression to correct for any
nonassayed radioactivity:
Percent Absorbed - Tota1 r^ioactivity following topical administration
Total radioactivity following parenteral administration
The percutaneous absorption of a large number of compounds has been quantified using this technique.
However, since absorption is expressed as percent of the applied dose, none of the K,, values in the database result
from studies that employed this approach. As the percent absorbed may vary with different amounts applied, it is
desirable that the actual dosing regimen is reported.
An alternative to measuring the amount of radioactive label in the excreta involves measuring levels of a
urinary metabolite over time. For example, Baranowska-Dutkiewicz (1982) estimated the percutaneous absorption
of aniline based on the amount of p-aminophenol excreted in the urine over a 24-hour period. Similarly, Dutkiewicz
and Tyras (1967, 1968) estimated the percutaneous uptake of ethylbenzene and styrene from the appearance of
mandelic acid in the urine of exposed individuals. Use of this technique to quantify percutaneous absorption requires
that the urinary metabolites of a parent compound be known, and the relationship between administered dose of the
parent compound and amount of the metabolite in the urine be characterized.
4.1.2 Quantification of Parent Compound or Metabolite in Blood, Plasma, or Tissues
In some cases, unlabeled parent compound can be measured in blood, plasma, or tissues after topical
administration. However, because of the difficulty in detecting and quantifying low levels of many compounds in
the plasma, this approach has been used for only a few compounds.
4-4
-------
Wester and Maibach (1983) measured actual levels of nitroglycerin (NTG) in the plasma after topical
administration in addition to following the radiolobel. These investigators estimated the mean percutaneous
absorption of this drug using three measurements: (1) area under the curve (AUC) of the plasma NTG concentration-
time profile; (2) AUC of the total radioactivity in the plasma; and (3) urinary total radioactivity. Their results are
presented in Table 4-2.
Table 4*2. Comparison of In Vivo Methods for Determining Mean Bioavailability
Method Percutaneous Absorption (%)
Plasma nitroglycerin AUC 56.6 ± 2.5
Plasma total radioactivity AUC 112 ± 6.7
Urinary total radioactivity 72.7 ± 5.8
Source: Wester and Maibach (1983)
Wester and Maibach (1983) have speculated that the difference between the percutaneous absorption based
on the AUC of the parent compound in the plasma, and percutaneous absorption based on the measurement of total
radioactivity in either plasma or urine, is due to the first-pass metabolism of this compound occurring in the liver.
Radioactivity can be measured in the tissues, as well as in blood or plasma, after topical administration of
a compound. This approach can be used not only to characterize the tissue distribution of the radiolabel after
cutaneous exposure (e.g., Skowronski et al., 1989), but also to quantify the rate or extent of absorption. For example,
Poiger and Schlatter (1980) determined the extent of percutaneous absorption of TCDD in a soil matrix applied to
the backs of hairless mice by monitoring the appearance of radiolabel in the liver. Shu et al. (1987) conducted a
similar study of percutaneous absorpiton of TCDD applied in soil. However, estimation of the percent of TCDD
absorbed in these studies requires knowledge of the distribution of compound in the body and the percent of the body
burden of the compound that resides in the liver.
4.1.3 Quantification of the Disappearance of the Compound from the Surface of the Skin or from the
Donor Solution
An older technique that was used to measure in vivo percutaneous absorption involves determining the loss
of material from the surface as it penetrates the skin. It is assumed that the difference between applied dose and
residual dose is the amount of penetrant absorbed. The difficulties inherent in skin recovery, volatility of penetrant,
and errors associated with using the difference between amount of the compound applied and amount remaining
4-5
-------
make this an inaccurate method to obtain quantitative percutaneous absorption rate information. As shown by
Frederickson (1961a), this approach can be especially problematic for compounds that permeate the skin slowly.
Percutaneous absorption rate constants for several important environmental pollutants (e.g., ethylbenzene,
toluene, styrene, xylene) were obtained by measurement of the disappearance of the compound from the donor
vehicle (Dutkiewicz and Tyras, 1967). Any difference in the amount of the compound in the donor solution before
and after immersion of the whole hand in the liquid for a prescribed period is assumed to result from uptake across
the skin. Evaporation of the compound is prevented by placing a beaker containing the donor solution in a
polyethylene bag and securing the open end of the bag around the subject's forearm.
The studies conducted by Dutkiewicz and Tyras (1967, 1968) using this technique are of special interest
because they have generated Kp values for important environmental pollutants using human subjects and aqueous
solutions of the compounds. Furthermore, the values generated by this technique have been used by other
investigators to determine the relative contribution of percutaneous absorption to total body burden (Brown et al.,
1984; Shehata, 1985; Brown and Hattis, 1989) or to validate theoretical skin permeability models (Brown et al.,
1990). Several researchers, however, have identified problems inherent in this approach. Sato and Nakajima (1978)
suggested that the rate of absorption measured by Dutkiewicz and Tyras (1967,1968) may be a combination of the
rate at which the compound is absorbed by the systemic circulation and rate at which the compound is taken up by
the stratum comeum. Maibach (1989) has expressed similar concerns over the use of this technique to provide valid
Kp values.
Parallel studies conducted by Dutkiewicz and Tyras (1967) may validate their findings. In their study of
ethylbenzene uptake, Dutkiewicz and Tyras (1967) determined the absorption of this compound not only by
measuring the difference in ethylbenzene concentration before and after immersion of the hand in an aqueous
solution, but also by monitoring the appearance of the main metabolite of ethylbenzene, mandelic acid, in the urine,
over a 24-hour period. The results are summarized in Table 4-3.
4-6
-------
Table 4-3. Comparison of the "Direct" and "Indirect" Methods of Dutkiewicz and Tyras (1967)
for Determination of Percutaneous Absorption Rate
Method"
Direct
Direct
Indirect
Mean Initial
Concentration
Ethylbenzene
(mg//)
112.0
156.2
142
Flux
(ug/cm2/hr)
118
215.7
189.4
(cm/hr)
1.05
1.38
1.33
1 Exposure occurred over a 2-hour period for the indirect and 1-hour for the direct method.
Assuming the relationship between urinary mandelic acid excretion and exposure to ethylbenzene is
adequately characterized, the similarity in the Kp values obtained by the "indirect" and "direct" methods of
Dutkiewicz and Tyras (1967) suggests that these Kp values for ethylbenzene may be valid. However, because the
method has reasonable potential for error, the experiment should be repeated.
The method used by Knaak et al. (1984 a.b) to determine the percutaneous absorption of tridemeton and
parathion requires either surface disappearance or plasma elimination data. This technique is described more fully
in Section 5.32.
4.1.4 Measurement of a Biological Response
Biological or pharmacological responses have been used to estimate percutaneous absorption for a limited
number of compounds. Responses such as vasodilation or vasoconstriction have been monitored as indices of
compound absorption. For example, laser Doppler velocimetry has been used as a noninvasive technique to monitor
the vasodilatory effects of topically applied nicotinate compounds (Guy et al., 1985b; Kohli et al., 1987).
One study in the dermal permeability database uses a biological response to obtain absorption rate values.
Frederickson (1961a) determined the percutaneous flux of paraoxon in cats by monitoring the inhibition of plasma
cholinesterase in these animals after topical administration of the compound. Determination of the rate of absorption
is complicated by the kinetics of this reaction which involves a conversion of the parent compound to parathion for
the anticholinesterase action.
4-7
-------
Biological response measurements are useful, noninvasive means for determining in vivo percutaneous
absorption rates when validated. However, the responses measured by these techniques do not distinguish between
the rate of absorption and the capacity of the compound to produce the effect. Therefore, given its current state of
development, this approach generally provides only a semi-quantitative index of percutaneous absorption.
4.1.5 Stripping Method
In a series of papers, Rougier et al. (1983,1985,1987,1989) recently measured percutaneous absorption
in vivo and examined the relationship between absorption and the concentration of the compound present in the
stratum comeum reservoir after a relatively short exposure period. For example, Rougier et al. (1987) tested the
relationship between percutaneous penetration and the amount present in the stratum comeum in vivo in humans
using four organic compounds (benzole acid sodium salt, benzoic acid, caffeine, and acetyl salicyclic acid). The first
applications of radiolabeled test compound onto the skin allowed the total absorption to be determined by measuring
the amounts of the chemicals excreted in the urine during the first 24 hours. The second applications. 48 hours after
the first, on the contralateral site of the body enabled an assessment of the total amount of the chemical present in
the stratum comeum at the end of 30 minutes by tape-stripping (IS strippings with adhesive tape). Rougier et al.
(1987) determined that the amount of the four compounds penetrating human skin in vivo after four days correlated
well (r = 0.97) with the amount of compound localized in the stratum comeum 30 minutes after application.
Despite the potential usefulness of this approach, it is a relatively recent development and none of the values
in the dermal permeability database were obtained using this method. Although, initial findings (Rougier et al., 1983,
1985. 1987; Dupuis et al., 1986) suggest that this may prove to be a valid approach, it appears that the stripping
method has been evaluated by only one laboratory, with only a few compounds which are not important
environmental pollutants.
4.2 In Vitro Techniques
As indicated in Table 4-1, in vitro studies provided many of the Kp values in the dermal permeability
database. This may be because values in in vivo percutaneous absorption studies are more often reported as percent
of the initial dose that is absorbed. In contrast, until recently results of in vitro percutaneous absorption studies were
reported as Kp or J. Because of the reliance on in vitro studies to provide Kp values, the various in vitro
experimental techniques and factors that affect in vitro percutaneous absorption will be examined in this section.
4-8
-------
Bronaugh and Maibach (1983) reported advantages of using in vitro methods for obtaining percutaneous
absorption rates. For example, these techniques permit:
• Investigation of percutaneous absorption separate from other pharmacokinetic factors that affect
cutaneous uptake;
• Larger numbers of assays;
Sampling directly under the skin; and
• Measurements of the permeability of highly toxic compounds using human tissues.
Furthermore, in vitro techniques are rapid, inexpensive, and easy to perform. However, despite these
methodological advantages, the conditions present in in vitro studies can be quite different from those present in the
in vivo state. Because of these differences, it is important to determine the validity of using in vitro data in a
percutaneous exposure assessment. To do this, the exposure/risk assessor should be aware of the advantages and
limitations of each of the commonly used in vitro techniques. It will then be possible to determine which in vitro
techniques best represent the in vivo physiological state of the skin in order to identify the most appropriate approach
to assessment of percutaneously absorbed dose.
4.2.1 Diffusion Cells
In vitro percutaneous absorption rates are most often measured using diffusion cells (e.g., glass, teflon,
stainless steel). Studies conducted in the 1960s and 1970s commonly employed a two-cell (side-by-side) diffusion
chamber. This technique involves mounting a piece of excised skin in the chamber, putting the radiolabeled
penetrant compound in one cell, and a receptor fluid, usually water or saline, in an adjacent cell. Tregear (1966)
has commented that the validity of using excised skin in an in vitro diffusion study depends on the following three
assumptions:
• Skin surface conditions in vitro are similar to those in vivo;
• The dermis does not affect penetration; and
• No living process affects permeability.
A limitation in the use of the two-chambered diffusion cell, such as the Franz cell, is that skin surface
conditions may be different from those experienced by the living organism (Franz, 1975). When occluded in the
diffusion cell, the skin becomes hydrated, thereby altering the permeability characteristics of the skin (Scott, 1986).
4-9
-------
An alternative technique that minimizes this problem uses a one-chambered unoccluded diffusion cell (Franz, 1975).
The unoccluded one-chambered cell may more closely parallel in vivo skin conditions when the stratum comeum
is exposed to the atmosphere (but not during swimming). Furthermore, this arrangement allows for the evaporation
of volatile compounds. However, in the one-chambered cell, unless there is an excess of material applied, one
measures percent absorption rather than determines a penetration rate with attendant K,,. A flow-through cell can
automate sample collection, improve partitioning of water-insoluble compounds into the receptor fluid, and replenish
skin nutrients to maintain the viability of skin for metabolism studies (Bronaugh and Stewart, 1985). Determination
of a permeability constant from water, however, requires that the skin be covered with an aqueous solution.
Therefore, either type of cell can be used equally well
Factors such as skin surface condition (especially those that affect volatilization of the test compound),
thickness of the barrier layer, and skin viability may affect the degree of percutaneous absorption in vivo and in vitro.
Therefore, these factors should be considered. A variable that may affect the predictive capacity of this in vitro
technique that is unique to the diffusion cell is the solubility of the penetrant molecule in the receptor fluid. This
factor can also be controlled in a diffusion cell. The degree to which these factors affect the rate of in vitro
percutaneous absorption in the diffusion cell apparatus, and the advantages and limitations of this technique, are
examined below.
Volatility of the Test Compound
As discussed in Section 3.0, evaporation can account for a significant percentage of the total dose of a
compound applied to the skin. For example, Reifenrath and Robinson (1982) have shown that evaporation accounts
for as much as 26% of the total dose of lindane applied to pig skin in a covered diffusion cell at 24°C. Therefore,
volatility is a factor that should be considered.
Receptor Fluid Compatibility
While the use of saline solution in the receptor chamber of a diffusion cell may be appropriate for
measurements that determine the percutaneous flux of hydrophilic compounds, it may not be appropriate for water-
insoluble lipophilic compounds. In a living organism, a lipophilic compound is readily taken up by blood (a
relatively lipophilic medium of large capacity) once it enters the cutaneous capillaries. In a static diffusion cell, the
receptor fluid serves the same role as blood does in vivo. However, unlike in the in vivo state, the receptor
compartment volume in a diffusion cell is of a finite size. If the receptor compartment volume is relatively small,
and if the compound is not metabolized in the skin, the concentration gradient across the cutaneous membrane will
4-10
-------
decrease until equilibrium is approached (Riley and Kemppainen, 1985). This limitation can be improved by using
a flowing receptor medium and by changing the receptor fluid to one that can serve as a better solvent for the
penetrating molecule.
Wester and colleagues (1985) reported that markedly different values for the percutaneous absorption of
triclocarban in human abdominal skin were obtained in static and continuous flow diffusion cells (Table 4-4). The
relative insolubility of this compound in the aqueous receptor phase may be responsible for the discrepancy between
the results obtained by the two systems. In the continuous flow system, the extent of triclocarban absorption was
similar to that measured for the in vivo penetration of this compound in humans (7.0 ± 2.8%). Because of the greatly
increased volume of saline, the solubility of triclocarban in the receptor fluid did not limit absorption (Wester et al.,
1985; Wester and Maibach, 1986).
Table 4-4. In Vitro Percutaneous Absorption of Triclocarban in Human Adult Abdominal Epidermis
Type of System Dose Absorbed
(% ± SD)
Static, 37°C 0.23 ± 0.15
Static, 23°C 0.13 ± 0.05
Continuous flow through, 23°C 6.0 ± 2.0
Source: Wester et al. (1985)
Some percutaneous penetration studies have been conducted using lipophilic receptor fluids such as human
and animal serum, aqueous alcohol mixtures, and various surfactant solutions. Bronaugh and Stewart (1984)
proposed the use of the non-ionic surfactant PEG 20 oleyl ether in the receptor fluid in combination with split-
thickness skin (most of dermis removed with dermatome). Only when split-thickness skin was utilized was
absorption of lipophilic compounds enhanced. Increased skin penetration was obtained as compared to values
obtained using serum, albumin, or alcoholic solutions; and no damage to the barrier properties of skin was detectable.
However, the viability of skin is not maintained under these conditions and so metabolism cannot be studied. If
viability is maintained with physiological buffer, partitioning of lipophilic compounds into the receptor will not be
complete. Skin content of the compound at the end of the experiment will need to be included with receptor fluid
values to determine the total absorbed test compound. The difference in the relative absorption of benzo(a) pyrene
and DDT, two relatively hydrophobia compounds, across rat skin in a diffusion cell using either saline or a nonionic
surfactant solution as the receptor fluid is shown in Table 4-5.
4-11
-------
Table 4-5. Effect of Receptor Fluid Composition on the Relative Absorption of
Hydrophobia Compounds
Absorbed % of
Compound
Benzo(a)pvienet
DDT*
Receptor Fluid
Normal Saline
6% PEG-20 oleyl ether in water
Normal Saline
0.5% PEG-20 oleyl ether in water
Applied Dose
3.7
56.0
1.8
60.6
±
±
±
±
0.1
0.9
0.1
2.9
* Applied in an acetone vehicle (15 ul/cm2) to haired rat skin (300 urn)
b Applied in an acetone vehicle (15 ul/cm2) to fuzzy rat skin (200 urn)
Source: Modified from Bronaugh and Stewart (1986)
As shown in Table 4-5, benzo(a)pyrene and DDT were poorly absorbed into the saline receptor fluid;
however, the presence of the nonionic surfactant in the receptor chamber markedly increased the extent to which
these compounds were absorbed. By comparison, Bronaugh and Stewart (1986) determined that the in vivo
percutaneous absorption of benzo(a)pyrcne and DDT was 48.3 ±2.1% (haired rat) and 69.5 ± 1.7% (fuzzy rat),
respectively.
Therefore, the exposure/risk assessor should be aware that the use of in vitro percutaneous absorption data
obtained in studies in which saline was used as the receptor fluid may result in underestimates of the in vivo
percutaneous absorption of lipophilic compounds. The surfactant solution is appropriate only if it does not alter
permeability characteristics of the skin. The rate-limiting step for diffusion through the stratum corneum of
chemicals with limited water solubility may be the transfer to the receptor fluid, especially in a static receptor system.
Kp is then dependent on the stratum comeum/receptor fluid partition coefficient.
Full- vs. Split-Thickness Skin
The second of Tregear's (1966) assumptions regarding the validity of using excised skin in a diffusion cell
study is that the dermis does not affect penetration. Under in vivo conditions, the greatest percentage of a compound
applied to the skin will be taken up by capillaries found in the dermis at a depth of approximately 200 urn.
Therefore, topically applied compounds do not have to penetrate the dermis to be absorbed. However, when full-
thickness skin is used in an in vitro diffusion cell study, the compound must penetrate the stratum comeum,
4-12
-------
epidermis, and dermis before reaching the receptor fluid. Furthermore, the cutaneous microcirculation is destroyed
in the preparation of skin for use in a diffusion cell.
Since dermal absorption is a passive process, the rate of absorption will be inversely proportional to the
thickness of the barrier layer (as shown in Equation 5.3, Section 5.0). Therefore, to ensure that in vitro percutaneous
absorption rates are comparable with those measured in vivo, the excised skin prepared for a diffusion cell should
be of the same thickness as the effective in vivo penetration barrier. The cutaneous layer present in the full-thickness
skin samples commonly used in in vitro studies may present an artificial barrier to percutaneous absorption,
especially for lipophilic compounds. Unlike the stratum corneum, cutaneous tissue is primarily an aqueous medium.
Therefore, this aqueous phase represents a potential barrier to the absorption of lipophilic compounds.
A number of researchers have investigated how using full- or split-thickness skin in vitro affects the
relationship between in vitro and in vivo percutaneous absorption of hydrophilic and lipophilic compounds. Hawkins
and Reifenrath (1986) examined the penetration of compounds with octanol/water partition coefficients spanning
several orders of magnitude using either full- or split-thickness pig skin in vitro and compared these values to those
obtained in vivo, as shown in Table 4-6. Total radioactivity was measured to determine the percent of applied
radioactive dose which penetrated pig skin in vitro and in vivo.
From the results reported in Table 4-6, it appears that the dermis can provide a significant barrier for highly
lipophilic compounds such as lindane and testosterone. With more hydrophilic compounds, the degree of
percutaneous penetration in vitro more closely approximates percutaneous permeation values obtained in vivo. Also,
as might be expected, removal of the epidermal layer of pig skin in vitro enhances the percutaneous absorption of
lipophilic compounds such as DDT and progesterone to a lesser extent than hydrophilic compounds, such as benzoic
acid, because the epidermal layer may not serve as the rate-limiting barrier for the diffusion of lipophilic compounds
across the skin. Absorption into the circulation takes place at the dermal-epidermal boundary in vivo so that the
penetrant reaches the capillaries prior to traversing the dermis, hence the dermis does not act as a significant barrier
to penetration. This is illustrated in Table 4-7.
4-13
-------
Table 4-6. Percutaneous Absorption of Radiolabeled Compounds on Pig Skin
In Vitro and In Vivo
Compound
Caffeine
Benzoic acid
N,N-Diethyl-m-toIuamide
Fluocinolone acetonide
Malathion
Parathion
Testosterone
Lindane
Progesterone
Percent of Applied Radioactive Dose*
Log P* In Vivo* Whole Skind
0.01
1.95
2.29
2.48
2.98
2.98
3.31
3.66
3.78
23
28
9
6
4.4
19
6.0
8
10
i
±
±
±
±
±
±
±
±
9
6
4
1
0.3
2
0.3
1
1
20
20
14
2
16
1
3
1
1
± 2
±13
± 4
± 1
±11
± 1
± 2
± 1
± 1
Split Thickness Skin
Rawd
18
17
19
1.1
21
12
9
6
5
±
±
±
±
±
±
±
±
±
3
6
13
0.9
6
5
4
2
2
Adjusted*
21
21
21
2
24
21
13
9
9
± 4
± 7
±13
± 2
i 7
± 5
± 5
± 4
± 4
* The applied dose of all compounds was 4 ug/cm2
b Log of the octanol/water partition coefficient
c Duration of cutaneous exposure was 48 hrs, followed by 5 days of monitoring excreta for radioactivity prior to
animal sacrifice and tissue analyses.
d Duration of cutaneous exposure was 50 hrs, followed by analysis of radioactivity in the skin (without dermis) and
receptor fluid.
e Sum of the radioactivity from split-thickness skins (with dermis) and radioactivity in the receptor fluid.
Source: Hawkins and Reifenrath (1986)
Table 4-7. Influence of the Epidermis on Percutaneous Absorption
of Compounds Through Pig Skin
Percent Absorbed
Compound
Benzoic acid
Testosterone
Progesterone
DDT
LogP
1.95
3.31
3.78
5.0
Epidermis
Present
15 ±4
4 ±2
1.7 ± 0.6
0.7 ± 0.3
Epidermis
Removed
88 ± 9
15 ± 8
7 ± 5
1.2 ± 0.5
Ratio
5.9
3.8
4.1
1.7
Source: Hawkins and Reifenrath (1986)
4-14
-------
Skin Viability
The third of Tregear's (1966) assumptions regarding the use of excised skin in an in vitro diffusion study
is that no living process affects permeability. However, Section 3.0 outlines the xenobiotic metabolizing capacity
of the skin, and how metabolism may affect both the rate and amount of percutaneous absorption. Nevertheless,
in vitro percutaneous absorption studies do not always consider skin viability and metabolic capacity and, in fact,
may involve skin that has been previously frozen. The influence of metabolic capacity effects on skin penetration
needs further study, but for our purposes, they may be small with regard to cutaneous exposure assessment For
example, the difference between viable and nonviable human skin may often be less than the permeability difference
between mouse and human skin. Exploratory studies (Hawkins and Reifenrath, 1984) suggest that the metabolite
capacity of mouse skin is much greater than that of swine or human skin.
Hawkins and Reifenrath (1984) measured the absorption of N,N-diethyl-m-toluamide (m-DEET) in full-
thickness pig skin used immediately after excision and in pig skin stored for a period of one to six weeks at -80°C.
Absorption of this compound through the frozen skin samples increased as a function of storage time. However,
the authors of the two studies present in the dermal permeability database (Bronaugh et al., 1986 and DelTerzo et
al., 1986) reported that freezing of the skin used in their studies had no effect on the integrity of these preparations,
or subsequent permeation of the topically applied compounds. Nevertheless, freezing the skin may affect the
biotransformation of compounds that can be metabolized by the skin. For example, Holland et al. (1984) have shown
that TCDD-induced BaP metabolism is markedly reduced by freezing mouse skin prior to in vitro use.
Collier et al. (1989) described a method for maintaining the viability of skin in a flow-through diffusion cell
for at least 24 hours using HEPES-buffered Hank's balanced salt solution as the receptor fluid. Viability was
assessed by monitoring glucose utilization, metabolism of test compounds, and by electron microscopy. A complete
tissue culture medium is not required to maintain viability of skin. The simplified balanced salt solution is less
expensive and less likely to interfere with analytical techniques.
Storing skin samples for prolonged periods at cold temperatures that are above freezing apparently has little
effect on tissue viability. Kemppainen et al. (1986) have shown that the permeability of monkey skin to T-2 toxin
(12,13-epoxy-n-chothecene) and cutaneous metabolism of this compound are unaffected by storage of the skin at 4°C
in air-tight plastic bags for up to 10 days.
From these studies, it appears that the in vitro percutaneous absorption and metabolism of several
compounds is closely linked to the viability of the skin preparation. Although skin viability can probably be
4-15
-------
maintained for periods up to 10 days by keeping the tissue cool (4°C) but not frozen (Kemppainen et al., 1986). two
techniques have been developed to better ensure tissue viability by maintaining the excised skin in a short-term organ
culture system.
Kao et al. (1984) developed a "static" system for simultaneously maintaining tissue viability and measuring
in vitro percutaneous absorption. Researchers in this laboratory (Holland et aL, 1984) have also developed a
"dynamic" culture system. This system consists of a waterjacketed multisample skin penetration chamber that is
continuously perfused with oxygenated culture medium. This system not only provides the excised skin with an
oxygen-rich tissue culture medium, but it also serves as a flow-through diffusion chamber for permeability studies.
4.2.2 Isolated Perfused Tubed-Skin Preparation
To overcome the potential limitations posed by in vitro systems. Riviere et al. (198S, 1986) have used
viable, isolated perfused tubed-skin preparations for determining in vitro percutaneous absorption rates. This system,
an isolated perfused porcine skin flap (IPPSF), is viable for at least ten hours. The tubed flap is transferred to an
isolated organ perfusion apparatus (Figure 4-1).
Figure 4-1. IPPSF Preparation and Perfusion System
Source: Riviere et al. (1985,1986)
4-16
-------
This IPPSF preparation may be valuable to quantify the rate and degree of first-pass cutaneous metabolism
of percutaneously absorbed xenobiotics without the confounding effects of biotransformation of the compound in
other metabolically active tissues. It also permits percutaneous absorption to be modeled as a function of applied
dose and cutaneous blood flow. Although the IPPSF preparation shows potential for generating information useful
for human exposure/risk assessment, its use is rather limited now.
4.2.3 Stratum Corneum Binding Technique
Wester et al. (1987) have proposed an in vitro method to estimate the percutaneous absorption of chemical
contaminants in aqueous solution, based on the binding of these compounds to powdered stratum comeum. The
method needs to be validated.
The stratum comeum binding technique was not used to obtain values reported in the dermal permeability
database. However, Wester et al. (1987) used the stratum corneum binding data to estimate the dose of p-nitroaniline
that would be absorbed from water after bathing or swimming for 30 minutes. Caution should be exercised in using
stratum comeum binding data for this purpose.
4.3 Comparison of In Vitro and In Vivo Percutaneous Absorption Values
Parallel studies of in vitro and in vivo percutaneous absorption have been compared by Franz (1975,1978).
The in vitro percutaneous absorption of 12 organic compounds was evaluated using previously frozen, full-thickness
human abdominal skin that was compared with the in vivo percutaneous absorption values obtained by Feldman and
Maibach (1970). Although some in vitro and in vivo values did not agree quantitatively, Franz (1978) conducted
modified tests with four compounds. The technique was able to distinguish compounds of low permeability from
those of high permeability, and the correlation showed fairly good agreement (Franz, 1978). Skin viability and
receptor fluid compatibility were not taken into consideration.
4-17
-------
Table 4-8. Total Absorption of Various Compounds by Skin In Vivo and In Vitro (Modified Tests)
(Expressed as Percent of Applied Dose)
Compound
Nicotinic acid
Hippuric acid
Thiourea
Caffeine
Absorption In Vivo*
0.32
1.0
3.7
22.1
±
±
±
±
0.10
0.4
1.3
15.8
(3)
(6)
(4)
(4)
1*
21
3
21
7
Absorption In
2.3
1.25
4.6
24.1
±
±
±
±
0.9
0.5
2.3
7.8
Vitro'
(4)
(4)
(5)
(4)
' Mean ± standard deviation; the values in brackets represent the number of subjects studied.
b Number of days urine was collected.
Source: Franz (1978)
Reasonably good agreements of in vitro and in vivo percutaneous absorption data have been obtained for
relatively hydrophilic compounds using a standard diffusion cell technique. For example, the percent absorption of
benzoic acid measured in vivo (42.6%) by Feldman and Maibach (1970) was similar to the value obtained with the
static in vitro diffusion cell (44.9%) (Franz, 1975). Bronaugh et al. (1982a) also reported good agreement between
in vitro and in vivo values for another relatively hydrophilic compound, acetylsalicylic acid.
Investigators have had difficulty predicting the in vivo percutaneous absorption of lipophilic compounds by
using in vitro techniques. Although further research is needed, these limitations have been improved by two major
modifications in the standard diffusion cell technique proposed by Bronaugh and Stewart (1984):
• Use of a split-thickness skin preparation in which the dermis is markedly reduced or completely
removed; and
• Use of receptor fluids that provide a greater solubility for the permeant than saline.
The use of a flow-through receptor medium fluid compartment to provide a greater volume of receptor
media has also been shown to improve the predictive capability of in vitro percutaneous absorption studies. The
capability of these modifications to improve the predictive capability of in vitro percutaneous absorption studies has
been documented in previous sections of the report.
For example, the in vivo percutaneous absorption of the relatively lipophilic compounds DDT and
fenvalerate can be closely approximated when split-thickness in vitro skin preparations are used. However, the
results obtained by using full-thickness preparations are quite different from those measured in vivo (Grissom et al.,
4-18
-------
1987a). Recognition of the poor diffusion of lipophilic compounds through the relatively aqueous dermis and of the
insolubility of these compounds in a static aqueous receptor fluid has led to the increased use of split-thickness skin
preparations and compatible receptor media or flow-through diffusion cells to measure the in vitro percutaneous
absorption of lipophilic compounds.
Differences in the degree of in vitro and in vivo percutaneous absorption may be a function of the time at
which these measurements were taken. Yang and colleagues have shown that for anthracene (Yang et al., 1986a)
and benzo(a)pyrene (Yang et al., 1986b) the total amounts of absorption measured in vitro and in vivo are somewhat
different when measured at day 1 or 2 of a 5-day study, but they do coalesce over time. These investigators have
attributed the greater differences observed at earlier time points to the systemic uptake, metabolism, and elimination
of the compound that occurs in vivo. This results in a time delay between absorption of the compound across the
stratum comeum and measurement of the compound or metabolites in exhaled air or excreta. The time that it takes
for these processes to occur accounts for the apparent lag in the in vitro results. Yang and coworkers (1986b) have
speculated that the coalescence of the in vitro and in vivo values over time occurs because "lag time" becomes less
of a factor in the determination of cumulative recovery of absorbed BaP in vivo.
Yang et al. (1986b) used non-viable skin preparations for their in vitro studies of BaP absorption and
obtained a reasonably good approximation of in vivo results after 3 to 4 days of absorption. However, as shown in
Section 3.0, Kao et al. (1985) observed that the in vitro percutaneous absorption of BaP is markedly affected by the
viability of the skin preparation. The differences in the results of these studies may result from differences in applied
dose. Yang et al. (1986b) used a higher dose (9-10 ug/cm2) than Kao et al. (1985) (2.5 ug/cm2). This higher dose
may have saturated the enzymes responsible for metabolizing BaP in the skin and may have resulted in the
development of diffusion-dependent conditions.
In summary, in vitro percutaneous absorption values are generally good predictors of the rate or extent of
percutaneous absorption that occurs in the intact animal. However, the factors described in the previous sections of
this report may affect the predictive accuracy of in vitro percutaneous absorption, especially for compounds that are
neither very hydrophilic nor very lipophilic. In many cases, failure to control for these variables will lead to a poor
correlation between in vitro and in vivo percutaneous absorption values. Therefore, if in vitro K,, data are used to
estimate percutaneously absorbed data, the exposure/risk assessor should examine the conditions under which the
in vitro percutaneous absorption study was conducted, to determine how well the in vitro Kp can be expected to
approximate the results obtained in vivo.
4-19
-------
4.4 Interspecies Comparison of Percutaneous Absorption Values
Percutaneous absorption studies in which the compound is applied to human skin in vivo provide the most
relevant information for human exposure/risk assessment. However, the toxicity of many compounds of interest to
exposure/risk assessors limits their testing in humans in vivo. Alternatively, data from studies using experimental
animals or in vitro techniques provides the Kp data necessary for estimating percutaneously absorbed dose in humans.
Examination of the K- values found in the dermal permeability database reveals that the majority of these
values were obtained in studies in which the compound was applied in vivo or in vitro to the rat skin. The
advantages of using this species to obtain percutaneous absorption rate data have been reviewed by Zendzian (1989).
For example, these animals: (1) are readily available to the research community; (2) have a defined genetic
background thereby minimizing the degree of individual variation in handling xenobiotic compounds; and (3) have
a surface area sufficient for dose application. However, rat skin, as well as skin from the mouse, rabbit and guinea
pig have consistently been shown to be more permeable to topically applied compounds than human skin.
Furthermore, the male rat differs from the female rat in skin permeability of the dorsal skin.
Wester and Maibach (1986) have summarized the results of several investigators that have ranked, from
highest to lowest, the relative in vitro percutaneous absorption of different species, as shown in Table 4-9.
Table 4-9. Ranking of the Relative In Vitro Percutaneous Absorption of Different Species
Tregear (1966)
Rabbit
Rat
Guinea pig
Human
Marzulli et al. (1969)'
Mouse
Guinea Pig
Goat
Rabbit
Horse
Cat
Dog
Monkey
Weanling pig
Man
Chimpanzee
McCreesh
Rabbit
Rat
(1965)
Guinea pig
Cat
Goat
Monkey
Dog
Pig
* Based on studies involving organophosphate compounds.
Source: Wester and Maibach (1986)
4-20
-------
To use percutaneous absorption rates obtained from animal studies in exposure/risk assessment, it would
be useful to understand how rat, mouse, rabbit, guinea pig and human skin compare. Bronaugh et al. (1982b) have
summarized some of these data in Table 4-10.
Table 4-10. Permeability of Animal Skin Relative to Human Skin1*
Reference and Compound
Tregear (1966)
Ethylenebromide
Paraoxon
Thioglycolic acid
Water
LogP
1.96
—
0.09
-1.38
Pig
0.8
1.4
3.3
1.4
Rat
2.3
3.3
3.0
Guinea
Pig Mouse Mouse
1.5
3.0
2.3
1.0
Hairless
Rabbit
3.3
Chowhan and Pritchard (1978)
Naproxin
2.3
3.5
Durrheim et al. (1980)
Butanol
Ethanol
Octanol
Stoughton (1975)
Betamethasone
5-Fluorouracil
Hydrocortisone
Bronaugh et al. (1982b)
Acetylsalkylic acid
Benzoic acid
Urea
0.65
-0.31
2.97
—
-0.95
1.61
1.19
1.87
-2.11
1.8
1.5
0.6
1.3
1.1
1.5
1.2 1.0 4.9
0.2 0.6 2.0
1.5 4.8 0.9 5.8
8.7
2.0
* Values for human skin in all studies were assigned a value of 1.0.
b All values are based on in vitro determinations.
Source: Bronaugh et al. (1982b) for permeability data; Hansch and Leo (1979) for Log P data.
4-21
-------
Other investigators have also demonstrated the similarity in the relative permeabilities of human and pig
skin for compounds with a range of log P values using in vitro studies. Monkey skin has been shown to be a good
model for human skin in a number of in vitro studies. Bronaugh and Maibach (1985) found the in vitro percutaneous
absorption of nitroaromatic compounds to be similar in human and monkey skin, as shown in Table 4-11, but the
excised human skin tended to be somewhat less permeable.
Table 4-11. Percutaneous Absorption of Nitroaromatic Compounds
in Human and Monkey Skin*
AoDlied Dose (%)
Compound
p-Nitroaniline
p-Amino-2-nitrophenoI
2,4-Dinitrochlorobenzene
2-Nitro-p-phenylenediamine
Nitrobenzene
1 Based on in vitro studies.
Source: Bronaugh and Maibach (1985)
LogP
1.39
0.96
1.90
0.53
1.85
Human
48.0 ±11.0
45.1 ± 8.0
32.5 ± 8.7
21.7 ± 2.6
7.8 ± 12
Monkey
62.2 ± 6.1
482 ± 7.8
48.4 ± 3.9
29.6 ± 4.3
62 ± 1.0
Walker et al. (1983) compared the relative in vitro percutaneous absorption rates of water and paraquat
through excised human and experimental animal skin. As shown in Table 4-12, the absorption rates of water in the
excised hairless rat and hairless mouse skin are about 1.5- to 3-fold greater, respectively, than in human skin; Kp
values measured in rabbit and guinea pig skin were about 3- to 5-fold greater, respectively, than human skin.
However, the Kp values for paraquat obtained for all animal models were markedly greater than those of humans,
ranging from 40-fold greater for the rat to 1,500-fold greater for the hairless mouse.
Several trends are evident from the data presented in Tables 4-9 through 4-12. The percutaneous absorption
of many compounds in the pig and monkey is similar to that found in humans. However, although rat, mouse, rabbit,
or guinea pig skin may be useful models for human skin, they may overestimate percutaneous absorption in humans.
Therefore, the use of percutaneous absorption values obtained in experimental animal studies will almost always
result in a conservative (i.e., higher) estimate of percutaneously absorbed dose in humans if animal Kp values are
used in exposure/risk assessment.
4-22
-------
Table 4-12. Relative In Vitro Percutaneous Absorption of Water and Paraquat
Through Human and Animal Skin
Permeability constant (cm/hr x 10"s)
Water Paraquat
Human
Rat
Hairless Rat
Nude Rat
Mouse
Hairless Mouse
Rabbit
Guinea Pig
93
103
130
152
164
254
253
442
0.7
27.2
35.3
35.3
97.2
1065.0
92.9
196.0
Source: Walker et al. (1983)
The variability in estimates of animal skin permeability relative to that of human skin makes it difficult to
suggest a factor to correct for the increased permeability of animal skin when using these values in a human
exposure/risk assessment. Vanderslice and Ohanian (1989) observed from the data reported from Scheuplein and
Blank (1973) and Flynn et al. (1980), that a five-fold difference in the percutaneous absorption rate of a series of
alkanols exists in excised human and mouse skin, respectively. Based on this observation, Vanderslice and Ohanian
(1989) adjusted the Kp values obtained in mice or rats by a factor of 5 to approximate human absorption rates.
In addition, McDougal et al. (1990) observed, on average, a two- to three-fold difference in the Kp of
chemical vapors between rat and human skin. Therefore, from these observations, and the relative permeabilities
summarized in Table 4-10, it may be reasonable to correct the percutaneous absorption rates from mouse and rat
studies by a factor of three to five to obtain more realistic estimates of human Kp values. However, the relatively
small database that is currently available makes it difficult to validate this approach for environmental pollutants.
In summary, the values present in the dermal permeability database have been obtained by a variety of
techniques. Before using these values to estimate the percutaneously absorbed dose of environmental pollutants in
humans, the exposure/risk assessor should be aware of the limitations of the technique used. In addition, if the value
4-23
-------
comes from either an in vitro or experimental animal study, the exposure/risk assessor should explain how this value
might approximate the values expected for human skin in vivo. General guidance on the evaluation of these factors
has been presented in this section, and is summarized in Table 4-13.
Table 4-13. Summary of Factors That May Affect the Use of K, Data in
Cutaneous Exposure Assessment
Factor
Comments
1. Type of Study Used to Obtain Data
In Vivo/Radioactivity
In Vivo/Parent or Metabolite
In Vivo/Biological Response
In V/vo/Stripping
In V/w/Disappearance
In Vitro/Diffusion Cell
In Vitro/IPPSF
In Vitro/Stratum Comeum Binding
May not represent absorption of the parent
compound.
Sensitive assay is often needed to detect parent
compound. One needs to know the pharmacokinetic
behavior of the compound if metabolite data are
used.
Response may indicate absorption and potency of
the compound. Not quantitative.
No data are present in database from this method.
Used to provide K,, data for key compounds in
database, but may measure both absorption and
binding in stratum comeum. Evaporative loss may
confound findings.
Commonly used technique to provide K,, values.
Need to examine the conditions of the study to
determine how they may have affected the results.
Consider species differences and lipophilicity of test
materials.
New technique; shows promise, but little available
data.
Shows promise, but not well validated.
In Vitro to In Vivo Comparison of
Percutaneous Absorption Values
In vitro results are often good predictors of in vivo
data; however, one needs to carefully examine the
conditions under which the in vitro studies were
conducted. Species and lipophilicity affect results.
3. Interspecies Comparison of
Percutaneous Absorption Values
Percutaneous absorption values obtained using
monkey or pig skin often approximate values
obtained in humans. Rat. rabbit, mouse, and guinea
pig skins are generally five to ten times more
permeable than human skin.
4-24
-------
5.0 ESTIMATION OF DERMAL ABSORPTION RATE VALUES
FOR COMPOUNDS IN WATER
The rate at which a compound is absorbed through the skin (percutaneous absorption) is commonly
represented in one of three ways:
• Flux (J): The amount of chemical absorbed across a defined surface area
of the skin per unit time (mg/cm2/hr).
• Permeability coefficient (Kp): A flux value, normalized for concentration, that represents the
rate at which the chemical crosses the skin's rate-limiting
barrier layer (cm/hr).
Percent absorbed The percentage or fraction of the applied dose that is absorbed
across the skin. Duration of exposure or decontamination time
(usually 24 hrs) and observation time (usually 5 days) should
be indicated along with amount applied per unit area and
percent absorbed.
As presented in Chapter 2, the current equation used to estimate the percutaneously absorbed dose of a compound
in aqueous media uses the permeability coefficient as a measure of percutaneous absorption. When Pick's law
prevails under the experimental conditions of percutaneous absorption studies, the permeability coefficient can be
evaluated from measured fluxes across the skin or it can be grossly approximated from the percent of the compound
absorbed through the skin. Since experimental conditions vary greatly among in vitro studies, and between in vitro
and in vivo systems, estimation of a permeability coefficient from measured flux or percent absorbed is not always
possible.
The following section will present the derivation of Pick's law and the assumptions underlying the application
of this equation in the evaluation of the permeability coefficient from experimental data. Under the simple case
where Pick's law holds, the interconversion of Kp, J, and percent absorbed will be explored. This derivation is then
applied to the in vitro experimental system of the diffusion cell, with some experimental conditions identified when
Pick's law is violated. A brief description of in vivo percutaneous absorption measure is then provided, with some
reference to a possible equivalent in vivo "permeability coefficient".
5-1
-------
5.1 Theoretical Basis of Kp and J Values
Fick's first law of diffusion is widely used to relate the flux J of a compound across the skin as a function of
its concentration gradient across the stratum comeum:
The use of this simple flux equation requires that steady-state diffusion occurs one-dimensionally, and that there
is no convection in the same direction as the one-dimensional diffusion. Applying Fick's law to diffusion across
the skin, the steady-state assumption means that, physically, the volumes of the adjacent solutions on the two sides
of the skin must be much greater than the volume of the skin, that these solutions should be well-mixed, and that
the concentration of the compound at the surface of the skin should be constant The lack of convection transport
often implies a dilute solution. The concentration difference is measured at the two sides of the skin.
The skin can be considered to be a membrane, which has chemical properties different from the properties of
the solutions. The permeability coefficient is a function of the pathlength of chemical diffusion (/), the stratum
comeum/vehicle partition coefficient (KJ, and the diffusion coefficient (DJ of the compound within the membrane:
(5-2)
Combining equations 5.1 and 52 yields an expanded solution for percutaneous flux:
Km ZL A C
_!__: (5-3)
The following assumptions, employed by Blank (1964) and Scheuplein (1965), are necessary for this
relationship to be valid:
5-2
-------
• Molecular diffusion across the stratum corneum is the rate-limiting step in absorption;
• The full thickness of the stratum corneum contributes to the diffusion barrier,
• No active transport occurs;
• The stratum corneum is considered a homogeneous medium;
• Penetrant and vehicle molecules diffuse across the stratum corneum as individual entities; i.e., there
is no carrier effect;
• There are no size-limiting pores to affect absorption;
• The stratum corneum is not changed progressively by the vehicle or penetrant; and
• Penetrant concentration changes do not alter stratum comeum or vehicle properties.
Although some of these assumptions are violated (e.g., homogeneous nature and constant thickness of the
stratum corneum), equations 5.2 and 5.3 are nevertheless useful for the estimation of Kp or J, respectively, of
relatively small, polar, nonelectrolytes. The use of these equations also assumes that there is no metabolism in the
skin which could damage the barrier function of the skin. For some water insoluble compounds, the viable epidermis
may serve as the rate-limiting barrier (Bronaugh and Collier, 1990); thus, the assumption that the stratum corneum
is the rate-limiting barrier may not always apply.
A major assumption involved in the use of equations 5.1 through 5.3 is that a steady state diffusion exists
across the stratum corneum. However, time is required after initial contact with the skin for the compound to
accumulate in the stratum comeum and establish a constant concentration gradient. This period, termed the lag time
(I), is represented by Scheuplein and Blank (1971) as:
The time required to reach steady state can have a significant impact on the use of equations 5.1 through
5.3 to estimate Kp or J for compounds in environmentally relevant exposure scenarios. For example, the lag time
for some compounds may be on the order of hours (Dugard, 1986). The actual rate of absorption is less prior to
reaching a steady state (Figure 5-1), so the use of equations 52 and 53 will result in an overestimate of Kp or J for
relatively short exposure periods (e.g., 10-15 minutes) such as those that are common for environmentally relevant
5-3
-------
exposure scenarios (e.g., bathing or showering). Therefore, the use of equations 5.1 through 5.3 will result in a
conservative (higher) estimate of dermally absorbed dose for many environmental pollutants. Under the conditions
where Pick's law prevails, the dermally absorbed dose can be evaluated by Equation 2.1 using the permeability
coefficient in aqueous vehicles. However, the percent absorbed cannot be estimated under these conditions of infinite
exposure.
The parameters affecting the rate of percutaneous absorption include the permeability coefficient K,, and the
concentration gradient across the skin. Kp can be measured directly under in vitro experimental conditions, or can
be estimated theoretically from equation (52). In the next section (5.2), the effects of each parameter on the
percutaneous flux evaluated by equation (5.3) are discussed. The experimental measurement of Kp is discussed in
Section 5.3.
I
OMnrMrM
TUMbnM
Figure 5-1. Determination of Percutaneous Absorption Lag Time
Source: Treagear (1966) and Guy and Hadgraft (1989a)
-------
5J Dermal (Percutaneous) Absorption Rate Equation Parameters
Partition coefficient K_
The partition coefficient K,,, defines the equilibrium ratio of the concentration of the compound in the
membrane (stratum comeum) divided by that in the adjacent solution (vehicle). The use of this partition coefficient
implies that equilibrium exists across the skin surface. The properties of this partition coefficient are such that
diffusion can often occur from a region of low concentration into a region of high concentration. If the compound
is more soluble in the skin than in the surrounding solution, then the concentration of the compound in the skin at
equilibrium will be greater than the concentration in the vehicle. If the compound is less soluble in the skin, the
concentration in the skin will be less than the concentration in the adjacent solution.
The method to determine K,,, values described by Scheuplein (1965) involves:
• Allowing a known quantity of dry stratum comeum to come to equilibrium with a known
concentration of a solution of the compound under study;
• Determining the concentration of the compound in the solution both initially and after equilibrium
is reached; and
• Taking the difference between these concentrations to determine the amount of the compound taken
up by the stratum comeum.
The partition coefficient is then determined by:
~ m Concentration in the stratum corneum (5-5)
* Concentration in the vehicle,
Several investigators have developed modifications to this approach. For example, Bronaugh et al. (1981)
enclosed dried, weighed pieces of stratum comeum in filter paper (to facilitate removal of the tissue) and exposed
the tissue to various vehicles containing 14C-N-nitrosodiethanolamine (NDELA). Once the stratum comeum was
removed, dried, and reweighed, the content of 14C-NDELA in the membrane was determined by liquid scintillation
counting. The partition coefficient was then estimated according to equation 5.5.
5-5
-------
Blank and McAuliffe (1985) recognized that the technique proposed by Scheuplein (1965) was not adequate
to determine K^, values for organic compounds that partition from an organic solvent into the relatively lipophilic
stratum comeum. They developed an approach based on the following assumptions:
• The internal (receptor) and external (donor) surface layers of the stratum comeum have the same
activity coefficient for the test compound, and that both surface layers are equally hydrated at
equilibrium;
• The concentration of an organic compound in the unsaturated aqueous receptor of a diffusion cell
remains constant after equilibrium is reached;
• The receptor concentration, at equilibrium is identical to the concentration of die compound added
to the water which has been allowed to come to equilibrium with the donor without the presence
of the skin; and
• The concentration of the organic compound within the stratum comeum (i.e., on both sides of the
stratum comeum at the receptor surface boundary and the donor surface boundary) are identical
because, at equilibrium, no net flux occurs.
Therefore, the concentration of the compound in the stratum comeum at the receptor boundary in an in vitro
test system can be estimated by multiplying the concentration of the compound in the aqueous solution by the stratum
comeum/water partition coefficient. This value can be divided by the known concentration of the compound in an
organic the vehicle (according to equation 5.5) to estimate the stratum comeum/organic vehicle K,,,.
From an exposure/risk assessment perspective, this approach may be especially important for determining
the percutaneous absorption of organic compounds from nonaqueous vehicles. For example, benzene is known to
be a component of gasoline. An exposure assessment for benzene should account for percutaneous absorption of
benzene from gasoline coining in contact with the skin. The approach proposed by Blank and McAuliffe (1985) for
measuring the stratum corneum/gasoline K,,, value for benzene can be used in this hypothetical situation, and the
resultant K,,, value can be incorporated into equation 5.3 to determine flux.
Despite the usefulness of these approaches, stratum comeum/vehicle partition coefficient values are rarely
reported in the literature. In the absence of experimentally derived K,,, values, one can approximate K,,, values for
nonelectrolytes in aqueous solution from octanol/water partition coefficient (K^J values using the empirical
relationship proposed by Roberts et al. (1975):
5-6
-------
Octanol/water partition coefficient values for a large number of compounds have been compiled by Hansch
and Leo (1979) and are also reported in the EPA documents such as Superfund Exposure Assessment Manual (EPA,
19885).
Alternatively, one can use the approach proposed by Scheuplein (1965) to estimate K,,, that has as its basis
the measurement of the difference between the initial vehicle concentration and the concentration of the compound
in the solution after exposure to a known mass of skin. This relationship has been represented by Durrheim et al.
(1980) as:
cv
*•«' •
where C0 and Ce are the initial and equilibrium concentrations of the compound in the vehicle, respectively. WT is
the weight of the tissue and V^ is the solution phase volume, which was 2 ml in their system.
Diffusion Coefficient
The other parameter that must be determined when estimating permeability constant and/or flux is the
diffusion coefficient (Dm). Most diffusion coefficients in liquids are in the order of 10'5 cm2/sec to 10"* cm2/sec
(Cussler, 1984). This is true for most solutes, at infinite dilution in water at 25°C (Table 5-1) or in nonaqueous
liquids (Table 5-2). Exceptions occur for high-molecular-weight solutes like albumin and polystyrene, where
diffusion can be one hundred times slower.
The diffusion coefficients in liquids are about ten thousand times slower than those in dilute gases. This
means that diffusion often limits the overall rate of processes occurring in liquids. The apparent diffusion
coefficients in the stratum comeum are much lower, ranging from KT4 to 10"7 times (lower than) the values for water
(Kasting et al. 1987).
The most common basis for estimating diffusion coefficients in liquids is the Stokes-Einstein equation. This
equation holds when the ratio of solute to solvent radius exceeds 5. When the solute size is less than 5 times that
of the solvent, this equation tends to break down. Adaptations of the Stokes-Einstein equation for small solutes and
macromolecules have been developed. For solute and solvent of similar size, several empirical correlations have
been proposed (Cussler, 1984).
5-7
-------
Table 5-1. Diffusion Coefficients at Infinite Dilution in Water at 25°C.
Solute
Argon
Air
Bromine
Carbon dioxide
Carbon monoxide
Chlorine
Ethane
Ethylene
Helium
Hydrogen
Methane
Nitric oxide
Nitrogen
Oxygen
Propane
Ammonia
Benzene
Hydrogen sulfide
Sulfuric acid
Nitric acid
Acetylene
Methanol
Ethanol
1-Propanol
2-Propanol
n-ButanoI
Benzyl alcohol
Formic acid
Acetic acid
Propionic acid
Benzoic acid
Glycine
Valine
Acetone
Urea
Sucrose
Ovalbumin
Hemoglobin
Unease
Fibrinogen
D (x 10'5 cmVsec)
2.00
2.00
1.18
1.92
2.03
125
120
1.87
628
4.50
1.49
2.60
1.88
2.10
0.97
1.64
1.02
1.41
1.73
2.60
0.88
0.84
0.84
0.87
0.87
0.77
0.821
1.50
121
1.06
1.00
1.06
0.83
1.16
(1380 - 0.0782c, + 0.00464 c,2)1
(0.5228 - 0265c,)«
0.078
0.069
0.035
0.020
* Known to very high accuracy, and so often used for calibration; c, is in moles per liter.
Source: Data from Cussler (1976) and Sherwood et al. (1975) (as reported in Cussler, 1984).
5-8
-------
Table 5-2. Diffusion Coefficients at Infinite Dilution in Nonaqueous Liquids.
Solute* Solvent D (x 10"5 cm2/sec)
Acetone Chloroform 2.35
Benzene 2.89
n-Butyl acetate 1.71
Ethyl alcohol (15°) 2.20
Ethyl ether 2.14
Ethyl acetate 2.02
Methyl ethyl ketone 2.13
Acetic acid Benzene 2.09
Aniline 1.96
Benzoic acid 1.38
Cyclohexane 2.09
Ethyl alcohol (15°) 2.25
/i-Heptane 2.10
Methyl ethyl ketone (30°) 2.09
Oxygen (29.6°) 2.89
Toluene 1.85
Acetic acid Acetone 3.31
Benzoic acid 2.62
Nitrobenzene (20°) 2.94
Water 4.56
Carbon tetrachloride n-Hexane 3.70
Dodecane 2.73
n-Hexane 4.21
Methyl ethyl ketone (30°) 3.74
Propane 4.87
Toluene 4.21
Benzene Ethyl alcohol 1.81
Camphor (20°) 0.70
Iodine 1.32
lodobenzene (20°) 1.00
Oxygen (29.6°) 2.64
Water 1.24
Carbon tetrachloride 1.50
Benzene n-Butyl alcohol 0.988
Biphenyl 0.627
p-Dichlorobenzene 0.627
Propane 1.57
Water 0.56
Acetone (20°) Ethyl acetate 3.18
Methyl ethyl ketone (30°) 2.93
Nitrobenzene (20°) 2.25
Water 3.20
Benzene /i-Heptane 3.40
* Temperature 25°C except as indicated.
Source: Data from Reid et al. (1977) as reported in Cussler (1984).
5-9
-------
For structurally analogous compounds, Guy et al. (198Sa) have proposed the following relationship to
approximate Dm:
where Dm is the diffusion coefficient for the compound of interest across the stratum comeum, DA is the diffusion
coefficient of a structurally analogous compound, and where MWA and MWm are the molecular weights of
components A and m, respectively. Brown and Hattis (1989) specifically used this approach to estimate Dm values
for compounds in their pharmacokinetic model (described in Section 9.0).
Pathleneth of Chemical Diffusion
The pathlength of chemical diffusion is determined by the nature of the chemicals as they penetrate the
stratum comeum (SC). The SC behaves as a solution phase and the penetrant molecule must dissolve in the SC as
it penetrates the skin. The two-phase (lipid and protein) model of the SC is generally accepted, and penetrant
molecules can follow the intercellular and/or transcellular route depending on their relative solubility and partitioning
in each phase. Though obviously not homogeneous, the SC is treated as uniform in character when an apparent
pathlength of chemical diffusion is determined. This pathlength is clearly (longer and) not the thickness of the SC,
and can not be readily measured experimentally. The thickness of the SC will therefore somewhat overestimate the
theoretical Kp, when the Kp is determined by Equation 5.2.
As shown previously in Table 3-7, the thickness of the SC in humans ranges from 10 to 50 urn for most
areas of the body; however, this layer can be as thick as 400 to 600 urn on the palm and sole, respectively
(Scheuplein and Blank, 1971). Because of this wide range in stratum comeum thickness, one should ideally know
the average thickness of the stratum comeum that may come into contact with aqueous solutions of environmental
pollutants in various scenarios when estimating percutaneously absorbed dose.
Concentration Gradient
The removal of most compounds by capillaries in the dermal layer is assumed to be efficient. When there are
no flow-limitations to the removal of penetrant compounds, the concentration of the compound of interest at the
5-10
-------
interior border of the diffusional barrier is assumed to be essentially zero. Therefore, the concentration difference
of solute across the membrane (AC) can be represented simply by the concentration of the solute in the vehicle. This
value is obtained experimentally and is specific for each exposure assessment
5.3 Evaluation of K,, from Experimental Studies
5.3.1 In Vitro Approaches
Permeability coefficient values can be evaluated from Pick's law (Equation 5.1) under specific in vitro
diffusion cell studies. Both the donor and receptor (receiver) cell compartments of the diffusion cell contain
homogeneous dilute solutions of one solute, with the donor cell at higher concentration. The donor and receptor cells
are separated by a sample of skin tissue. The solute diffuses from the fixed higher concentration in the donor cell
into the less concentrated solution in the receptor cell. Since the skin is chemically different from these solutions,
the partition coefficient defines the equilibrium ratio of the concentration of the solute in the membrane divided by
that in the adjacent donor solution. To determine the permeability coefficient of a solute, the concentration in both
the donor and receptor cells are measured as a function of time.
The main assumption of this diffusion cell is that the flux across the skin quickly reaches its steady-state
values, even though the concentrations in the donor and receptor cells are still changing with time. In this pseudo
steady-state, the flux across the skin can be approximated by Pick's law of diffusion (Equation 5.1).
*,
The mass balance for the donor and receiver cells give:
- -SAJ (5-10)
(5-11)
where SA is the surface area of the skin.
5-11
-------
If V,^, B Vnei^v = V, and the concentrations C,^, and C,^^ can be measured as a function of time,
the permeability coefficient Kp can be evaluated as follows:
(5-12)
' (SA)(Q
where the slope can be obtained by plotting C,^^ or £&„„ as a function of time and measuring the steady-state
slope.
Alternatively, the system of equations (5.10 and 5.11) can be solved using the initial condition:
at time t = 0, C^ - C,^ = C0^ - C°mtiva
and K can be evaluated as:
K. - ._L_ In [ C « - C ' w ] / [ (V- C_w ] (5-13)
V QAy /f. ln t C*»», - t'rmivtr J / I *"**«•-
where C(daoor) and C(reeeiVCT) are the measured concentrations at time L
This approach is used by Flynn and colleagues (e.g., Ackerman and Flynn, 1987; Behl et al., 1983a,b;
Durrheim et al.. 1980; Jetzer et al., 1986,1988) to estimate K,,.
Alternatively, one can plot J as a function of time, obtain a steady-state flux (!„), and estimate Kp by:
Several of the recommended compound-specific Kp values in Tables 10-1 and 11-2 were estimated from
experimentally derived J values using this relationship and the assumption that a steady-state rate of flux exists.
5-12
-------
The above analysis assumes that skin behaves as a homogeneous membrane with averaged properties in the
diffusion coefficient, partition coefficient and pathlength of chemical diffusion, and where no metabolism occurs.
The measured permeability coefficient is therefore an empirical value, and it encompasses all these underlying
assumptions, as well as those inherent in the use of Pick's law. Skin is actually a very complex organ, and
percutaneous absorption in vivo entails the consideration of physiological conditions often ignored in the in vitro
experimental systems. The nature of the chemicals exposed to the skin and their interactions with the various
components of the skin will determine in large part how well the above assumptions hold. Modifications of the in
vitro experimental conditions wiU in pan contribute to the validity of using Pick's first law (e.g., using volatile
vehicles or flow-through cells). An in-depth mathematical analysis of these in vitro experimental systems is currently
being undertaken by Bunge (through a cooperative agreement with U.S. EPA/ORD/OHEA). In most cases, under
steady-state diffusion, Pick's first law usually provides a first estimate of the permeability coefficients. To apply
these values in actual dermal exposure assessment, care has to be taken to compare the exposure conditions in actual
scenarios to the experimental conditions to ensure that the measured Kp provides an adequate estimate of
percutaneous absorption.
5JJ In Vivo Approaches
The current in vivo approaches usually report a percent absorbed as the measure of percutaneous absorption.
This percent absorbed is often the quantification of radioactivity of the parent compound and metabolites from
exhaled air, excreta, blood, plasma, or tissues. Several investigators have also reported the disappearance of the
compound from the surface of the skin or from the donor solution as percent absorbed. In most studies, the chemical
was exposed to the skin at a finite dose for a certain period of time, and the percent absorbed determined as
described above. Under these experimental conditions, transdermal flux and Kp can not be easily determined from
the measured percent absorbed.
Guy (1989) has proposed an equivalent measure to steady-state flux from percent absorbed data. When
small finite amounts of a compound are applied to the skin in vivo, the compound becomes depleted during the time
course of the study. As a result, the flux of the permeant first increases to a maximum, then decreases as the source
becomes depleted, making it impossible to estimate a suitable Kp from a steady-state flux. Because of this limitation,
Guy (1989) characterized the measured maximum percentage dose absorbed per hour per cm2 as being the closest
empirical value to the theoretical steady-state flux per unit area of skin. This approach requires the assumption that
the maximum rate of percutaneous uptake is the limiting step for absorption. This condition seems to have been
applicable to data of Roberts et al. (1977) that were evaluated by Guy. However, a Kp developed from maximum
rate of percutaneous uptake from a finite source, as suggested by Guy (1989), poses serious concerns as revealed
5-13
-------
by a review of percent absorbed data of Wester and Maibach (1976). These data, involving application of
testosterone on monkey skin, demonstrate that the maximum total penetration during a 24-hour period ranges from
1 ug at 40 fig/cm2 dosage to 20 ug at 4000 ug/cm2 dosage, indicating a 5-fold range in Kp values by this approach.
To determine Kp directly from in vivo studies, pharmacokinetic data which describes the absorption,
distribution, metabolism, and elimination of the compound is usually required. A pharmacokinetic model can be
constructed to include the transdermal flux as part of the mass balance across the skin. The flux can be assumed
to approach a pseudo steady-state condition; this can be accomplished experimentally. The body can be represented
by several physiological compartments to describe the distribution of the compound in various organs, or can be
lumped into a few major groups of tissues which characterize the kinetic data. Figure 5-2 gives an example of such
a physiologically-based pharmacokinetic model. Table 5-3 contains the physiological and biochemical parameters
used in the model, and Table 5-4 lists the abbreviations used in Figure 5-2 and Table 5-3. The value of Kp can thus
be fitted to the kinetic data, usually obtained as concentration over time of the compound in the body. This approach
has been used by Knaak et al. (1984b) for some pesticides, and by McDougal et al. (1990) to determine permeability
coefficients of several organic vapors.
5.4 Interconversion Among Kp, J, and Percent Absorbed Values
As shown in Section 53, from in vitro studies, Kp and J can be measured using diffusion cells under
steady-state conditions. If the concentration of the solute in the donor cell were to be considered essentially infinite
(i.e., the quantity of solute in the donor cell is so large that no decrease in the solute concentration can be observed
over time), a percent absorbed can not be obtained. However, if the donor cell concentration is finite, a pseudo
steady-state condition can be established across the skin to apply Pick's law, and, the percent absorbed can be
estimated. Results of in vitro testing are usually reported in terms of Kp and J. The percent of applied dose
absorbed can also be calculated at any time by measuring the concentration in the receptor fluid, using the following
equation:
The above equation can be used when the concentration or the total amount in the donor solution is so large
that decreases in donor concentration cannot be discerned (i.e,. pseudo infinite dose) as long as concentrations in the
receptor fluid can be measured.
5-14
-------
Inhalation
C,
ven
vp
vl
,
Alveolar Space
Lung Blood
Fat Tlasue Group
Richly Perfused
Tissue Group
Poorly Perfused
Tissue Group
Ingestlon ^
D K
Uver (Metabolizing)
Tissue Group
Metabolites
Skin Tissue Group
Cair
Dermal Absorption
Figure 5-2. Example of a Physiologically-Based Pharmacokinetic Model
5-15
-------
Table 5-3. Physiological and Biochemical Parameters Used in the Model
Parameters
Rats
Mice
Male
Female
Human
Body weight (kg)
Alveolar ventilation rate (1/min)
Q.
Blood flow rate (Vmin)
Q,
Qr
Q,
Qp
Q,
Q.
Tissue volume (1)
v,
v.
VP
vt
v.
Partition coefficient
N
Pt/i
P,
P,
P
P.
P.
Metabolic constant
Vm (mg/min)
K^ (mg/1 blood)
Absorption coefficient
K_ (cm/hr)
k (nun'1)
Surface Area exposed (m2)
.035
0.083
.104
.0092
.0434
.0074
.0389
.0052
.0315
.015
.220
.0140
.035
18.9
__
108.994
3.179
1.058
3.719
—
.00586
2.9378
.668
.0035
.035
.023
.002
.0012
.0035
.0058
.00115
.0038
.0021
.0273
.0017
.00035
.0039
1.472
.0025
.0028
.019
.0017
.0097
.00195
.0048
.00095
.0027
.0015
.0195
.0012
.00025
16.9
__
121.893
4.159
1.183
4.159
—
.003
1.472
11
70
7.5
6J2
.31
2.76
1.26
1.55
.31
14
3.5
36.4
1.72
7
10.3
505.4
108.994
3.719
3.72
3.72
505.4
.703
32.043
.17
.292
Appropriate unit conversions are provided in the actual program set up for the PB-PK model simulation
5-16
-------
Table 5-4. Abbreviations Used in Figure 5-2
Q. Alveolar ventilation rate (liters/min)
Q, Cardiac blood output (liters/min)
Q, Blood flow rate to fat tissue group (liters/min)
Q, Blood flow rate to richly perfused tissue goup (liters/min)
Q, Blood flow rate to poorly perfused tissue goup (liters/min)
Q, Blood flow rate to liver tissue group (liters/min)
Q, Blood flow rate to skin tissue group (liters/min)
V,, Vrt Vp, V,, V. Volumes of tissue groups (liters) corresponding, respectively, to fat,
richly perfused, poorly perfused, liver and skin tissue groups
N Blood/air partition coefficient
P* Skin/air partition coefficient
P,, P,, P,, P,, P. Tissue/blood partition coefficient corresponding, respectively, to fat, richly
perfused, poorly perfused, liver and skin tissue goups
V. Maximum velocity of metabolism (mg/min)
K. Michaelis constant (mg/liter)
A. Amount metabolized (rag)
CM Concentration in arterial blood (mg/liter)
C_ Concentration in venous blood (mg/liter)
C^, Cw, Cv. C«, C, Venous concentrations in tissue goups, corresponding, respectively,
to fat, richly perfused, poorly perfused, liver and skin tissue groups
C.J, Concentration in inhaled air (mg/liter)
C. Concentration in alveolar air (mg/liter)
K, Skin Permeability coefficient (cm/h)
A Surface area of dermal exposure cm2)
D Gavage dose (mg)
k Gut absorption time constant (min')
Appropriate unit conversions are provided in the actual program set up for the PB-PK model simulation
5-17
-------
The evaluation of Kp or J from measured in vivo percent absorbed is usually not as simple, as was discussed
in the last section. If a pseudo steady-state flux can be assumed, then a gross measure of Kp can be approximated
using Pick's law with the caveat expressed under 5.3.2. A pharmacokinetic model would allow an evaluation of Kp,
though the kinetic data necessary to develop such a model are often very extensive.
5.5 Percutaneous Absorption of Lipophilic Compounds in Water
There is limited data relating to penetration of highly lipophilic compounds dissolved in water. This
scenario has special interest to the EPA because water sources that may contact skin during bathing or swimming
are often contaminated with lipophilic compounds.
Blank and McAuliffe (1985) studied the penetration of benzene and aqueous solutions of benzene through
human abdominal skin in vitro. The fluxes were reported to be 2.11 ± 1.08 ul/cm2/h for pure benzene and 0.22 ±
0.05 ul/cm2/hr for benzene in water. The log K,,w is 2.13 and Blank and McAuliffe (1985) reported the Kp for
benzene in water to be 0.1 cm/hr. These data show only a 10-fold higher penetration rate for neat benzene, despite
a 1000-fold higher benzene concentration in the neat material than in the aqueous solution. (The maximum solubility
of benzene in water had been independently determined.) The enhanced flux of benzene in water is largely explained
on the basis of partition capacity and skin hydration.
A more recent paper by Dal Pozzo et al. (1991) involves a study of permeation rates through isolated human
abdominal and breast skin epidermis. Twelve nicotinic acid derivatives with a wide range of (isopropyl myristate-
water) partition coefficients (log P = -1.93 to +3.32) were tested as pure liquids and as aqueous solutions. The
results show a positive correlation between flux and partition coefficient for both neat and water-dissolved
compounds up to tog P = 1. Above these P values, Kp declined for neat liquids and increased (followed by a
plateau) for aqueous solutions. Again, we observe the enhancing effect of water on the flux of lipophilic compounds.
Dal Pozzo et al. (1991) warn, however, that in vitro permeation of the more water-insoluble compounds might be
artificially limited by the model used where water is the receptor solution. (Isopropyl myristate-water partition
coefficients were reported by Dal Pozzo et al. (1991) to more closely approximate stratum comeum-water partition.)
Scheuplein and Blank (1973) might argue against the use of partition coefficients in the manner used by Dal Pozzo
et al. (1991).
5-18
-------
6.0 DERMAL ABSORPTION OF COMPOUNDS FROM SOIL
Dermal contact with contaminated soil represents a potentially significant route of exposure to toxic
compounds. For example, workers at industrial facilities or hazardous waste sites can be exposed to soil
contaminated directly during the manufacture, transport, storage, or disposal of such compounds. The general
population can also be exposed to compounds applied directly to soil, as in the case of pesticides applied to lawns
and gardens. The general population may also be exposed to soil that is contaminated indirectly as a result of wind
erosion, surface water runoff, or fallout from municipal incinerators. Therefore, soil can become contaminated from
various sources, and activities such as playing in the din or gardening can result in exposure of different segments
of the general population to contaminated soil.
A number of methodologies have been developed to estimate the exposure of an individual to toxic
compounds in a soil matrix. While some approaches are designed to estimate dermal exposure (e.g., EPA, 1986),
others include an absorption factor to permit the calculation of absorbed dose (e.g., Schaum, 1984; Kimbrough et
al., 1984; Paustenbach et al., 1986; Eschenroeder et al., 1986; EPA, 1989b). For most chemicals, the parameters
in these approaches have not been well characterized. Estimates of the amount of soil adhering to human skin are
reported in Section 2.0. However, major uncertainties exist in the extent to which a chemical is percutaneously
absorbed and in the extent to which a chemical will partition from soil to skin. In addition to being chemical-
dependent, percutaneous absorption of a chemical in a soil matrix may depend on characteristics of the soil, such
as particle size and organic carbon content, which affect partitioning of the chemical between soil and skin. If
percutaneous absorption is presented as the fraction of applied dose that is absorbed, the soil loading rate will affect
the fraction absorbed. However, if percutaneous absorption is represented by a flux (concentration times permeability
coefficient) across the skin, the flux should not be affected by changing the soil loading.
To assist the exposure/risk assessor in determining the extent to which soil-adhered organic compounds are
absorbed percutaneously, the available experimental data and the factors that affect the percutaneous absorption of
compounds from soil will be examined in this section. Insufficient data exist to develop specific guidance that can
be used to estimate the dermal uptake of specific soil-adsorbed compounds. Guidance will be provided, however,
to estimate upper boundaries on percutaneous absorption for chemicals where experimental data are not available.
6.1 Experimentally Derived Values
Poiger and Schlatter (1980) investigated the percutaneous absorption of TCDD from contaminated soil.
These researchers applied a paste of 67% soil and 33% water and various levels of radiolabeled TCDD to the backs
of hairless rats, and monitored the appearance of TCDD in the liver to determine the percent of the applied dose that
6-1
-------
was absorbed. Seventy-five mg of the soil/water paste were spread over an area of 3 to 4 cm2. Three dose levels
were used — 26,350, and 1300 total ng of TCDD. Soil loading rates were 19 to 25 mg/cm2. TCDD concentrations
in the soil/water paste were 0.3 ppm, 5 ppm, and 17 ppm. Poiger and Schlatter (1980) stated that reproducible
quantities of TCDD in the liver were measurable only after administration of 50 uCi or more. Since 26 ng
corresponds to about 5 uCi, results for this dosage are not included in this report even though they are reported by
Poiger and Schlatter (1980). If the percentage of total body burden is known, the percentage of applied dose
absorbed can be calculated. Poiger and Schlatter noted that Fries and Marrow (1975) found 70% of the total body
burden of TCDD located in the livers of female rats. Using 70% as a correction factor, the percentage of the total
applied dose absorbed can be estimated to be approximately 3% in 24 hours.
Shu et al. (1988) also studied percutaneous absorption of TCDD in soil. Shu et al. (1988) used protocols
that approximate exposure conditions more closely resembling those that have been experienced by humans,
including exposures to TCDD in the low concentration (i.e., part per billion) range, exposure durations less than 24
hours, and exposure to TCDD in soil contaminated with waste oil. Shu et al. applied soil containing concentrations
of 10 to 123 ppb TCDD and from 0 to 2% oil to rats and measured the percent of applied dose appearing in the liver
after 24 hours of exposure. They also measured the percent of applied dose in the liver after a 4-hour exposure
period compared to a 24-hour exposure period. Finally, they measured the percent applied dose in the liver for
environmentally contaminated soil from Times Beach containing 123 ppb TCDD. There was little difference in the
percutaneous absorption of TCDD under these various conditions. The percentages of applied dose in the liver
ranged from a low of 0.54% (10 ppb TCDD, 0.5% oil, 24-hour exposure) to a high of 1.01% (concentrations of
TCDD and oil not specified, 24-hour exposure). Shu et al. (1988) also estimated that if 100% of an orally
administered dose of TCDD were absorbed, 50% of that dose would appear in the liver 24 hours after administration.
Therefore, the percentages of applied dose measured in the liver after dermal application were divided by 0.5 to
arrive at a total percent absorbed ranging from 1.08% to 2.02% of the applied dose.
Roy et al. (1989,1990) applied TCDD, both neat and in soil, to rat skin in vivo and in vitro and to human
skin in vitro. These investigators found that approximately 77% of a topically applied dose of 70 ng of neat TCDD
was absorbed across rat skin after 96 hours. The traction absorbed was similar whether the TCDD was studied in
vivo or in vitro. Application to rat skin of the same applied dose used in the study of neat TCDD, but adsorbed to
low organic carbon soil at a concentration of 1 ppm, resulted in absorption of 16.3% of the applied dose in vivo and
7.7% of the applied dose in vitro. Application of TCDD in a soil matrix reduced the percentage of TCDD absorbed
by a factor of 5 in the in vivo test and by a factor of 10 in the in vitro test, compared to absorption of the neat
compound. The percentage of applied dose absorbed in vitro using human skin and low organic carbon soil was
2.4% after 96 hours or one-third that observed when rat skin was tested in vitro. TCDD was also applied to rat skin
6-2
-------
in vitro at a concentration of 1 ppm in a high organic carbon content soil. The percentage of applied dose absorbed
after 96 hours was 1.0%. or about one-eighth of the absorption obtained in vitro using the low organic carbon soil.
The in vivo results of Poiger and Schlatter (1980), Shu et al. (1988), and Roy et al. (1989, 1990) are
comparable when one takes into account different vehicles, soil loading rates, and exposure durations. Poiger and
Schlatter (1980), and Shu et al. (1988) used soil loading rates of about 21 mg/cm2, while Roy et al. used loadings
ranging from 6 to 10 mg/cm2. Poiger and Schlatter used TCDD-free soils from Seveso, Italy, which were ground
to a powder of homogeneous particle size and applied in a soil-water paste. Organic carbon content was unspecified.
Shu et al. (1988) used TCDD-free and contaminated soils obtained from Verona and Times Beach, Missouri, with
unspecified organic carbon content In some experiments, Shu et al. (1988) added 0.5% or 2% crankcase oil to the
soils. Roy et al. (1989, 1990) used a low organic carbon content (0.77%) and a high organic carbon content
(19.35%) soil. Exposure durations in the Roy et al. (1989,1990) experiments were 96 hours compared to 24 hours
for the other studies. Approximate concentrations of TCDD in soil were 0.3 ppm, 5 ppm, and 17 ppm in Poiger and
Schlatter (1980), 0.01 and 0.1 ppm in Shu et al. (1988), and 1 ppm in Roy et al. (1989,1990). Excluding the Poiger
and Schlatter (1980) results for the 0.3 ppm concentration, which the authors stated was not reproducible because
radioactivity was less than 50 uCi, and using the authors' factors for converting percent dose in liver to percent
applied dose absorbed (70% and 50% of total absorbed dose assumed to be in the liver by Poiger and Schlatter
(1980) and Shu et al. (1988), respectively), percents of applied dose absorbed ranged from 1.1% (Shu et al. (1988),
24 hour exposure, 10 ppb TCDD, 0.5% crankcase oil) to 16.3% (Roy et al. (1989,1990), % hour exposure, 1 ppm,
low organic carbon content soil). Given the variation in vehicles and exposure durations, these in vivo results seem
consistent.
Roy et al. (1989,1990) also tested 33',4,4'-tetrachlorobiphenyl (TCB) at a concentration of 1,000 ppm in
low organic carbon content soil in vivo in rats and in vitro in rat and human skin. After 96 hours, the percentages
of applied dose absorbed from low organic carbon content soil applied to rat skin were 49.8% in vivo and 31.9%
in vitro, while 7.4% of the dose applied in vitro to human skin was absorbed. Roy et al. (1989) also applied TCB
in high organic content carbon soil to rat skin in vitro, 9.6% of the applied dose was absorbed over 96 hours.
The in vitro and in vivo percutaneous absorption of another highly lipophilic compound, BaP, was
determined by Yang et al. (1989). Soil contaminated with crude petroleum oil at a concentration of 1% and with
BaP at a concentration of approximately 1 ppm was applied to split thickness rat skin in a diffusion chamber, at 9
mg soil/cm2 (described by the authors as a "monolayer") or at 56 mg soil/cm2. The percentages of BaP absorbed
after 96 hours of exposure were 8.4% and 1.3% of the initial applied dose at exposure levels of 9 and 56 mg
soil/cm2, respectively. Similar results were obtained when BaP at a concentration of 1 ppm in a soil containing 1%
6-3
-------
crude petroleum oil was applied in vivo to the backs of female rats in a monolayer (9 mg/cm2) of contaminated soil.
Approximately 9.2% of the dose of BaP applied in vivo was absorbed after 96 hours.
In the Yang et al. in vitro study of the effect of the soil loading on the amount absorbed, the quantity of
BaP absorbed, 1.3 ng/cm2, was the same regardless of the amount of soil and hence of BaP applied to the skin.
The authors propose that the migration of BaP present in layers of soil above the "monolayer" is impeded by the
extensive binding of BaP to soil particles.
Wester et al. (1990) studied the percutaneous absorption of DDT and BaP from soil in vitro using human
skin and in vivo in rhesus monkeys. Soil composed of 26% sand, 26% clay, and 48% silt, containing 10 ppm 14C-
labeled DDT or BaP was applied to human skin samples with surface areas of 1 cm2. The soil loading rate was
40 mg/cm2. After 24 hours, the surface of the sample was washed with soap and water. The investigators found
that the DDT tended to bind to the skin rather than enter the human plasma receptor phase. After 25 hours, an
average of 95.6% of the of the topically applied radioactivity was found in the surface wash, 1% was in the skin
sample, and 0.04% was in the receptor fluid. Results were similar for BaP. After 24 hours, 91.2% of the BaP
applied in soil was found in the surface wash, 1.4% was found in the skin, and 0.01% was found in the plasma
receptor fluid.
Wester et al. (1990) also measured in vivo percutaneous absorption in rhesus monkeys. The percentage of
topically applied DDT absorbed over a 24-hour exposure period averaged 3.3%. Percentages of applied dose of DDT
absorbed for three subjects were 2.7%, 3.4%, and 3.7%. The percentage of topically applied BaP absorbed over 24
hours averaged 13.2%. Percentages of applied dose of BaP absorbed for four subjects were 13.1%, 10.8%, 18.0%,
and 11.0%. The results of Poiger and Schlatter (1980), Shu et al. (1988), Roy et al. (1990), Yang et al. (1989), and
Wester et al. (1990) are presented in Table 6-1.
The only published data on percutaneous absorption of volatile organic chemicals (VOCs) in soil are the
studies of Skowronski et al. (1988,1989,1990), wherein, percutaneous absorption of benzene, toluene, and xylene
were measured. The Skowronski et al. data are not presented in Table 6-1 because the conditions used in these
experiments were not sufficiently similar to conditions of environmental exposure to allow use of the data in these
types of exposure assessments. The measured percutaneous absorption rates may not be appropriate for use in
exposure assessments for several reasons. First, the concentrations of the VOCs in soil were about 21% of the soil-
VOC mixture. While such a mixture might be encountered in the environment, in general, much lower
concentrations would be expected. Second, the skin was occluded during the experiment, preventing losses as a
result of evaporation. In an environmental exposure to VOCs in soil, the skin would not be expected to be occluded.
-------
Third, the soil was first placed on the rat and the VOC was subsequently added via syringe (Skowronski et al., 1990).
Under these conditions, and given the high concentrations of VOCs in the mixture, the VOCs might not be mixed
with and bound to the soil to the extent they would in the environment The absorption from sandy and clay soils
may be more representative of absorption of the neat compounds than of soil-bound compounds.
62 Factors Affecting the Dermal (Percutaneous) Absorption of Soil-Adhered Components
The factors outlined in Section 4.0 are known to affect the absorption of neat compounds and compounds
in aqueous or organic solvents. Presumably, these factors have the potential to influence the absorption of soil-bound
compounds as well For example, skin temperature, pH, moisture content, and surface area exposed have been
proposed by Skowronski et al. (1989) as factors that can affect the desorption of organic compounds from soil and
subsequent uptake across the skin. Although the capacity of these factors to affect the rate or extent of percutaneous
absorption of soil-absorbed compounds has not been demonstrated experimentally, factors unique to the soil matrix,
such as the physical and chemical characteristics of the soil, and the chemical's propensity for binding to soil
particles, have been shown to affect the dermal bioavailability of a soil-bound compound.
6-5
-------
Table 6-1. Dermal Absorption of Soil-Adhered Organic Compounds
Com-
pound
TCDD
SHe of
Appl-
catfon
Back
Back
Back
Back
Back
Back
Back
Back
Back
Back
Back
Back
Back
Back
Back
Back
Breast
Back
Back
Soil
loading
(mp/cn?)
19-25*
19-251
21
21
21
21
21
21
21
21
21
21
21
9-10
9-10
9-10
5-6
9
56
Surface
area
exposed
(cm2)
3-4
3-4
12
12
12
12
12
12
12
12
12
12
12
7
1.77
1.77
1.77
1.77
1.77
Total
applied
dose
(ng)
350
1300
2.7*
2.7'
26.9*
26.9f
26.9'
26.9f
26.9'
26.91
26.9
26.9'
26.9'
70
17.7
17.7
17.7
15.5
100.0
Concen-
tration
(ppm)
5
17
0.01
0.01
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.123
1
1
1
1
1
1
Percent
appled
dose
absorbed
2.4*
3.1'
1.08b*
1.22"
1.3b
1.58b'e
1.44M
1.26b
2.02b
1.24b
1.98b
1.46b'q
1.58b'
16.3"
7.7°
1.0°
2.4°
8.4°
1.3°
Total
dose
absorbed
(ng)
8.4
40.3
0.029
0.033
0.35
0.43
0.39
0.34
0.54
0.33
0.53
0.39
0.43
11.4
1.4
0.18
0.42
1.3
1.3
Exposure
period
(hrs)
24
24
24
24
24
24
24
4
24
4
24
24
24
96
96
96
96
96
96
Soil
organic
content
(%)
NRh
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
0.77
0.77
19.4
0.77
1.64
1.64
Animal
Species
Rats
Rats
Rats
Rats
Rats
Rats
Rats
Hairless
rats
Hairless
rats
Haired rats
Haired rats
Rats
Rats
Rats
Rats
Rats
Humans
Rats
Rats
Study
Method
In vivo
In vivo
In vivo
In vivo
In vivo
In vivo
In vivo
In vivo
In vivo
in vivo
In vivo
In vivo
In vivo
In vivo
In vitro
In vitro
In vHrv
In vitro
in vitro
Ref.
1
1
i
J
j
I
i
1
i
1
i
1
i
k
k
k
k
1
1
-------
Com-
pound
BaP
DDT
TCB
Site of
Appl-
cation
Back
Back
Back
Back
Back
NR
Back
NR
Back
Back
Breast
Back
Soil
loading
(mg/cnr)
9
9
9
9
40
40
40
40
9-10
9-10
5-6
9-10
Surface
area
exposed
(cm2)
7
7
7
7
12
1
12
1
L 7
1.77
1.77
1.77
Total
applied
dose
(ng)
63
63
63
63
4800
400
48009
400
70,000
17.700
10.600
17.700
Concen-
tration
(ppm)
1
1
1
1
10
10
10
10
1.000
1,000
1,000
1,000
Percent
appied
dose
absorbed
1.1
3.7
5.8
9.2
13.2°
1.4"
3.3°
1.4"
49.8°
31.9°
7.4°
9.6°
Total
dose
absorbed
(ng)
0.69
2.3
3.6
5.8
633.6
5.6
158.4
5.6
34.860
5.646
786
1.700
Exposure
period
(hrs)
24
48
72
96
24
24
24
24
96
96
86
96
Soil
organic
content
(%)
1.64
1.64
1.64
1.64
NR
NR
NR
NR
0.77
0.77
0.77
19.4
Animal
Species
Rats
Rats
Rats
Rats
Rhesus
monkeys
Humans
Rhesus
monkeys
Humans
Rats
Rats
Humans
Rats
Study
Method
in vivo
In vivo
in vivo
in vivo
In vivo
in vitro
in vivo
in vitro
In vivo
In vitro
In vitro
In vitro
Ref.
1
1
1
1
m
m
m
m
m
k
k
k
Values were estimited by dividing the percentages of the applied dose on the liver after dermal application by 0.7 (fraction of the total body burden of TCDO estimated lo be in the liver).
Values were estimated by dividing the percentages of the applied dose in the liver after dermal application by 0.5 (fraction of the total body burden of TCDD measured in the liver).
Presence of 0.57% crankcase oil as a co-contaminant
Presence of 2.8% crankcase oil as a co-contaminant
12.5 ng/kg x assumed mean weight of rats (0.215 kg - range 180-250 g).
125 ngTkg x assumed mean weight of rats (0.215 kg - range 180-250 g).
Percutaneous absorption was determined by percent absorbed « ("C unna
Not reported.
Poiger and Schlatter (1980).
Shu et al. (1988).
Roy et al. (1989,. 1990) and Driver et al. (1989).
'. urinary excretion following topical applicationXC14 excretion following i.v. administration) x 100.
Yangetal. ... ^
Wester el al. (19*). . .
Percentage of applied dose occurring in animal body mass ad determined by a complete tissue analysis.
Percentage of applied dose in the receptor fluid (using PEG-20 qleyl ether) following topical application.
Percentage of applied dose in the skin following topical application.
Laboratory contaminated.
Environmentally contaminated.
Applied as a soil/water paste (approximately V4 water by weight).
-------
Roy et al. (1989,1990) repotted an eight-fold reduction in the percent of TCDD absorbed through rat skin
in vitro when the compound was applied in soil with high organic carbon content (19.35%) as compared to
application in soil of low organic carbon content (0.77%) (see Table 6-2). By contrast, Poigcr and Schlatter (1980)
reported that 0.05% or less of a dose of TCDD applied to the skin of rats in an activated carbon/water paste was
found in the liver after 24 hours of exposure compared to about 2% of TCDD applied in a soil/water paste. These
results suggest that the TCDD is even more strongly adsorbed to activated carbon than to soil. Although TCDD in
the soil was bioavailable, essentially none of the TCDD was bioavailable when applied in activated carbon.
Table 6-2. Properties of Soil Used in the Roy et al. (1989,1990) Studies
Soil Type
Sand (%)
Silt (%)
Clay(%)
Organic Matter (%)
Particle Size, mm (%)
0.05 - 0.1
0.1 -025
0.25-0.5
03 - 1.0
1.0 -2.0
Hyde
193
70.1
10.4
19.4
593
193
15.9
4.1
1.0
Chapanoke
15.1
682
16.7
0.77
673
183
113
2.0
0.1
Poiger and Schlatter (1980) have shown that aging of TCDD in soil before oral administration of a TCDD/soil
paste to rats affected the percentage of the applied dose that appeared in the liver. Ingestion of soil aged with TCDD
for 10 to 15 hours before oral administration resulted in 24.1% of the dose appearing in the liver, only 16% of the
dose appeared in the liver after ingestion of soil aged with TCDD for 10 days before oral administration. The
authors postulate that the decreased oral bioavailability that occurs as a function of soil contact time with TCDD is
probably the result of a "strengthening" of the binding of TCDD to soil particles. From these two time points, it is
not possible to extrapolate the uptake of TCDD that would occur at even longer contact times with soil, nor is it
possible to determine how these results might be applicable for dermal uptake of TCDD. However, it is significant
to note that Shu et aL (1988) reported little difference in the percent of dose appearing in the liver of rats dermally
exposed to soil contaminated with TCDD in the laboratory (contact time unreported, but presumably on the order
of hours to days) and soil from Times Beach, Missouri that contained an equivalent amount of TCDD and was
presumably contaminated many years before the study was conducted.
6-8
-------
Poiger and Schlatter (1980) and Shu et al. (1988) have shown that the concentration of TCDD in the soil
matrix, within the range of concentrations studied, has little effect on the percentage of the applied dose that is
subsequently absorbed and appears in the liver (Table 6-1), provided the amount of soil applied to the skin is
constant The amount of TCDD in the liver, however, increased in direct proportion to the concentration. Yang et
al. (1989) demonstrated the effect of changing the amount of soil applied to the skin on the percentage and amount
of applied dose absorbed in their study of BaP. Although the percent of the initial applied dose absorbed was
different after application of 1 ppm BaP in 9 or 56 mg soil/cm2, the amount (i.e., mass) of the compound that was
absorbed from either dose was the same (approximately 1.3 ng). Although this factor has not been investigated for
other soil-adsorbed compounds, these limited findings suggest that the bioavailability of tightly-bound, soil-adsorbed
compounds may be limited to the amount of compound in the "monolayer" of soil in direct contact with the surface
area of skin.
The capacity of one component of a mixture of soil-adsorbed compounds to impede or accelerate the
absorption of another compound has been studied to a limited degree. Shu et al. (1988) demonstrated that co-
contamination of soil with TCDD and various concentrations of waste crankcase oil had little effect on the dermal
bioavailability of soil-adhered TCDD. Given the potential for numerous contaminants to co-exist in soil, this factor
requires further study to understand its impact on percutaneous absorption.
In summary, soil characteristics, environmental conditions, physical-chemical properties, and exposure time
are some of the important factors that affect the percutaneous absorption of soil-adsorbed compounds. Further,
factors that affect the percutaneous absorption of neat compounds or compounds in aqueous or organic vehicles (see
Section 3.0) may also influence the percutaneous absorption of soil-adsorbed compounds. However, studies are
lacking to support the presence of such effects.
63 Methodologies for Estimating the Dermal Absorption of Soil-Adsorbed Compounds
Section 63.1 provides guidance in interpreting the data summarized in Section 6.1. Section 6.3.2 provides
guidance for estimating percentages absorbed when there are no data on percutaneous absorption of the chemical of
interest applied in soil.
6.3.1 Interpreting Existing Data
The data summarized in Section 6.1 and Table 6-1 on dermal uptake of TCDD, TCB, BaP, and DDT in soil
must be used with extreme caution in exposure assessments. Percutaneous absorption rates presented in terms of
percentage of applied dose absorbed over a given exposure period are not universally applicable to exposure
6-9
-------
assessments. As discussed in Sections 6.1 and 6.2, these percentages are dependent on the amount of soil applied
to the skin and to the time the soil remains on the skin, as well as to other factors related to the composition of the
soil and the physicochemical properties of the penetrant. Rates of percutaneous absorption also differ between
various animal surrogates and humans and between in vitro and in vivo testing systems. The following section
provides guidance for evaluating the existing studies and selecting or estimating percutaneous absorption rates which
are consistent with existing data and protective of human health and the environment
6.3.1.1 TCDD
All of the data related to TCDD in soil are summarized in Table 6-1. Most of the data from the three
TCDD studies were collected in vivo using rats. For the reproducible data reported in Poiger and Schlatter (1980),
one can estimate that about 3% of the two dose levels of TCDD (5 ppm and 17 ppm) applied in a soil/water paste
at a rate of about 20 mg soil/cm2 was absorbed in vivo in the rat over 24 hours. From the Shu et al. data, one can
estimate that about 1% to 2% of TCDD applied at concentrations of 10 ppb, 100 ppb, and 123 ppb and a soil loading
of about 20 mg/cm2 was absorbed in vivo in the rat over 24 hours. The Roy et al. (1989, 1990) in vivo rat data
indicate that 163% of TCDD at 1 ppm in low organic carbon content soil applied at 10 mg/cm2 was absorbed over
96 hours. Given that the exposure duration used by Roy et al. (1989,1990) was four times that of the other two
studies, that the soil loading was half that used in the other two studies, and that the organic carbon content of the
soil was unusually low, the results for the three studies — 2% to 3% absorbed over 24 hours from a soil loading
of 20 mg/cm2 and an unknown organic carbon content; 16.3% absorbed over 96 hours from a soil loading of about
10 mg/cm2 and an organic carbon content of 0.77%) — appear consistent
Roy et al. (1989,1990) also performed in vitro studies of TCDD using rat and human skin samples. They
observed that 7.7% of TCDD applied to rat skin in vitro at a concentration of 1 ppm in low organic carbon content
soil was absorbed over % hours. The comparable percentage absorbed using human skin samples was 2.4%. When
TCDD was applied to rat skin in vitro in a high organic carbon content soil, 1.0% of the applied dose was absorbed
over 96 hours.
Roy et al. (1989,1990) data indicate that absorption of TCDD is greater when measured in vivo in the rat
(16.3%) than when measured in vitro in the rat (7.7%). Assuming that the differences between in vivo and in vitro
testing results are similar for rats and humans, a correction factor of 2 (16.3/7.7) can be calculated for converting
in vitro human data to in vivo human absorption fractions. Using this correction factor the estimated percent
absorbed for humans in vivo is 5% (2.4% x 16.3/7.7) over 96 hours for a loading of low organic carbon content soil
6-10
-------
of approximately 10 mg/cm2. Assuming the same relationships for high organic carbon content soil, one can estimate
a percentage of applied dose absorbed in rats in vivo of 2.0% over 96 hours (1.0% x 16.3/7.7).
The data from Roy et al. (1989,1990) also indicate that absorption of TCDD measured in vitro in the rat
(7.7%) is greater than absorption in human skin measured in vitro (2.4%). Assuming that the ratio of percent
absorbed in human skin to percent absorbed in rat skin is the same whether measured in vivo or in vitro, a factor
of 0.31 (2.4/7.7) for converting rat in vivo data to human in vivo absorption fractions can also be calculated from
the Roy et al. data. Thus, one can estimate an amount absorbed for humans in vivo over % hours of 0.62% (2.0%
x 2.4/7.7) for a soil loading of 10 mg/cm2 of high organic carbon content soil Applying this correction factor to
the Poiger and Schlatter (1980) and Shu et al. (1988) provides an estimation of the fraction of TCDD absorbed over
24 hours in vivo in the rat for a loading rate of 20 mg soil/cm2 (at most 2 to 3%). Thus, one can estimate fractions
absorbed under these conditions of about 1% in humans.
The differences between percentages absorbed for low and high organic carbon content soils are quite large
and represent a range that might be expected based on differences in soil composition. The high organic carbon
content soil used by Roy et al. (1989,1990) contains 19.35% organic carbon. This level is higher than would be
expected in most environmental soils. If organic carbon content is unknown or less than 19.35%, use of the
percentages absorbed for the low organic carbon content soil are recommended.
One critical factor that has not yet been considered in this estimation process is the soil loading. Typical
soil loadings in exposure scenarios are on the order of 0.5 to 1 mg/cm2 rather than the 10 to 20 mg/cm2 used in the
three TCDD studies. Section 2 of this document recommends 1 mg/cm2 as a default value. As the soil loading
decreases, the amount absorbed should remain the same for a given concentration while the percentage absorbed
increases. If 1% of TCDD in 20 mg soil/cm2 is absorbed over 24 hours, one would expect that 20% of TCDD at
the same concentration in a 1 mg soil/cm2 loading and 40% of a TCDD at the same concentration in a 0.5 mg/cm2
soil loading would be absorbed over a 24 hour exposure period. Applying the same assumptions to the 5%
absorption percent for a 10 mg/cm2 soil loading and a 96 hour exposure GOW organic soil) estimated from the Roy
et al. data, one can calculate 50% absorbed for a 1 mg/cm2 soil loading and 100% absorbed for a 0.5 mg/cm2 soil
loading.
Roy et al. (1989,1990) also studied percutaneous absorption of a total applied dose of 70 ng neat TCDD,
the same total applied dose of TCDD used in the soil experiments, and percutaneous absorption of an "infinite" dose
(246 mg TCDD/cm2). The percentages of the low (70 ng) applied dose absorbed, measured in vivo in the rat, were
333% after 8 hours and 77.4% measured after % hours. After 96 hours, 76.6% of the low dose was absorbed in
6-11
-------
vitro in rat skin. The fraction absorbed of the "infinite" dose (246 mg) was measured in vitro in both rat and human
skin. After 72 hours, 4.1% of the "infinite" dose was absorbed by the rat skin and 1.1% was absorbed by the human
skin. These data indicate that absorption of neat TCDD at low doses can approach 100%. However, at high doses,
the fraction absorbed is likely to be much less than 100%.
Table 6-3 presents some recommended values for fraction of TCDD absorbed for a range of exposure
scenarios. These values are calculated from the Roy et al. (1990), Shu et al. (1988), and Poiger and Schlatter (1980)
results, using the assumptions discussed above. These values apply only for soil concentrations of TCDD in the
range of approximately 1 ppb to 20 ppm. No data are available for concentrations of TCDD greater than 17 ppm
in soil.
6-12
-------
Table 6-3. Dermal Absorption Percentages for Dilute Concentrations of TCDD in Soil.
Soil
Loading
(mg/cm2)
0.5
0.5
1
1
10
10
20
20
1 Assumes a TCDD
Exposure
Duration
(hr)
24
24
24
24
96
concentration of 1 ppm.
,m Oh.i At al /1Ofiff\ nnrt Di
Percent
Absorbed
40*
12-100*
20d
6.2-50*
2"
0.62-5°
lb
0.31-2.5*
->!nor an/1 Cxhlaftor AlOfiHA
Amount
Absorbed
(ng/cm2)*
0.2f
0.06-OJf
0.2
0.06-0.5
0.2
0.06-0.5
0.2
0.06-0.5
absorbed in rats of 2% to 3% over 24 hours with a soil loading of 20/mg cm2 was estimated. Conversion from
rat to human percutaneous absorption is accomplished using conversion factors derived from Roy et al. (1990)
{(2% to 3% in rat in vivo over 24 hours) x (2.4% in human skin in vitro over 96 hours/7.7% in rat skin in vitro
over 96 hours) = 0.6% to 0.9% in humans, or approximately 1% over 24 hours).
Estimated from Roy et al. (1990) study and assumptions described in text. The low end of the range represents
estimated percent absorbed from high organic carbon content soil. The high end of the range represents estimated
percent absorbed from low organic carbon content soil.
Calculated from estimated percentage absorbed for a soil loading of 20 mg/cm2 and a 24 hour exposure duration
by assuming that the quantity of TCDD absorbed is constant (e.g., 1%/100% x 20 mg soil/cm2 x 1 ng TCDD/mg
soil = 0.2 ng TCDD/cm2 absorbed over 24 hours; 0.2 ng TCDD/cm2 absorbed/D.5 ng TCDD/cm2 applied x 100%
= 40%, which is the estimated percentage absorbed over 24 hours for a concentration of 1 ppm TCDD applied
in soil at 0.5 mg/cm2).
Calculated from estimated percentage absorbed for a soil loading of 10 mg/cm2 and a % hours exposure duration
by assuming that the quantity of TCDD absorbed is constant (e.g., 0.62%/100% x 10 mg soil/cm2 x 1 ng TCDD/1
mg soil = 0.062 ng TCDD/cm2 absorbed over 96 hours; 0.062 ng TCDD absorbed/0.5 ng TCDD applied x 100%=
12%, which is the percentage absorbed over % hours for 1 ppm TCDD in high organic carbon content soil applied
in 0.5 mg soil/cm2.)
Below a soil loading of 0.2 mg/cm2 for the 24 hour data, below 0.06 mg soil/cm2 for the 96 hour loading of high
organic soil, and below 0.5 mg/cm2 for the % hour loading of low organic soil, the amount absorbed will begin
to decrease with the soil loading and the percentage of applied dose absorbed will be 100% in all cases. At a soil
loading of 0.1 mg/cm2 and a concentration of 1 ppm TCDD, the estimated absorption at 24 hours would be 0.1
ng/cm2 or 100% of the total applied dose. The estimated exposure at 96 hours would be 0.06 to 0.1 ng/cm2.
6-13
-------
6.3.1.2 TCB
Roy et al. (1989) studied percutaneous absorption of TCB applied in soil containing a concentration of 1000
ppm of TCB. They studied percutaneous absorption in vivo and in vitro in rat skin and in vitro in human skin. In
low organic carbon content soil, 49.8% of the applied dose was absorbed in % hours in rat skin in vivo and 31.9%
was absorbed in rat skin in vitro. The ratio of in vivo to in vitro absorption is 1.6. In human skin, 7.4% of the TCB
applied in low organic carbon content soil was absorbed over % hours. Assuming that absorption in human skin
in vivo would be higher than absorption in human skin in vitro at the same ratio that absorption in rat skin in vivo
is higher than absorption in rat skin in vitro, one can predict that the percentage of applied dose absorbed by humans
over 96 hours would be about 12% (i.e., 7.4% x 49.8/31.9), for a soil loading of 10 mg/cm2.
Another way to calculate absorption in humans from these data is to assume that differences between
absorption in vitro in rat skin and human skin are analogous to differences between absorption in vivo. The ratio
of absorption in human skin in vitro to absorption in rat skin in vitro is 0.23 (7.4/31.9). Applying this ratio to the
percentage of applied dose absorbed for rat skin in vivo, one can estimate a percentage absorbed of TCB over 96
hours in humans of 12% (49.8% x 7.4/31.9) for a soil loading of 10 mg/cm2.
Roy et al. (1989) also studied percutaneous absorption of TCB in vitro in rat skin, applying the TCB in high
organic carbon content soil, with all other parameters kept constant Over % hours, 9.6% of the TCB was absorbed.
Using the relationships for the experimental results using low organic carbon content soil, one can estimate a
percentage of applied dose absorbed in rats in vivo of 15% over % hours (9.6% x 49.8/31.9). Then, one can
estimate an amount absorbed in humans over 96 hours of 3.5% (15% x 7.4/31.9) for a soil loading of 10 mg/cm2
of high organic carbon content soil.
Thus, the estimate for human absorption of TCB applied in soil at a rate of 10 mg soil/cm2 over 96 hours
of exposure ranges from 3.5% to 12% depending on the carbon content of the soil. Estimates of percentages of
applied dose absorbed for other soil loadings are presented in Table 6-4. Like the estimates for TCDD, these
estimates apply only to relatively dilute solutions of TCB in soil. The data are all generated using a concentration
of 1,000 ppm TCB in soil.
6-14
-------
Table 6-4. Dermal Absorption Percentages for Dilute Concentrations of TCB in Soil.
Soil
Loading
(mg/cm2)
0.5
1
10
Exposure
Duration
(hr)
%
%
96
Percent
Absorbed
70-100°
35-100°
3.5-12"
Amount
Absorbed
(ug/cm2)'
0.35-0.5d
03S'l.tf
0.35-1.2
* Assumes a TCB concentration of 1,000 ppm.
b Estimated from Roy et al. (1990) data and assumptions described in text. The low end of the range represents
percentage absorbed from high organic carbon content soil. The high end of the range represents estimated
percentage absorbed from low organic carbon content soil.
c Calculated from estimated percentage absorbed for a soil loading of 10 mg/cm2 and a % hours exposure duration
by assuming that the quantity of TCB absorbed is constant (e.g., 3.5%/100% x 10 mg soil/cm2 x 1 pg TCB/1 mg
soil« 0.35 ug TCB/cm2 absorbed over 96 hours; 0.35 ug TCB absorbed/0.5 ug TCB applied x 100%= 70%, which
is the percentage absorbed over 96 hours for 1000 ppm TCB in high organic carbon content soil applied in 0.5
mg soil/cm2.)
d Below a soil loading of 0.35 for high organic carbon content soil and 1.2 mg/cm2 for low organic carbon content
soil, the amount absorbed begins to decrease, while the percentage of applied dose absorbed will be 100%. At
a low organic carbon content soil loading of 1 mg/cm2 and a concentration of 1000 ppm TCB, the estimated
absorption at % hours would be 1.0 ug/cm2 or 100% of the total applied dose.
63.13
BaP
Yang et al. (1989) studied percutaneous absorption of BaP applied in soil containing 1 ppm BaP and 1%
crude oil in vivo and in vitro in rats. Yang et al. (1989) used soil with an organic carbon content of 1.64%. In one
series of in vivo experiments using a soil loading of 9 mg/cm2 and a dose of 63 ng, they reported percentages of
applied dose absorbed of 1.1% (0.7 ng) at 24 hours, 3.7% (2.3 ng) at 48 hours, 5.8% (3.6 ng) at 72 hours, and 9.2%
(5.4 ng) at 96 hours. In an in vitro experiment comparing percentages absorbed for two different soil loadings, they
reported 8.4% (13 ng) of applied dose absorbed for a loading of 9 mg/cm2 and a dose of 15.5 ng and 1.3% (1.3 ng)
for a loading of 56 mg/cm2 and a dose of 100 ng.
Wester et al. (1990) studied percutaneous absorption of BaP in vitro in human skin and in vivo in rhesus
monkeys. They used a soil loading of 40 mg/cm2- a concentration of 10 ppm BaP in the soil, and an exposure
duration of 24 hours, at which time excess soil and BaP were removed with soap and water. The soil was composed
of 26% sand, 26% clay, and 48% silt.
The Wester et al. (1990) in vitro data for human skin are consistent with the Yang et al. (1989) in vivo data
for rat skin collected at 24 hrs. In Wester et al. the percentage of applied dose in skin and fluid after 24 hours
6-15
-------
averaged 1.41%. The amount absorbed was 5.6 ng/cm2 (40 mg soil/cm2 x 10 ng BaP/mg soil x .0141). Assuming
that the amount absorbed over a given time period is proportional to the concentration (Pick's Law), the amount
absorbed for a concentration of 1 ppm, as used in Yang et al. (1989), would be 0.56 ng/cm2 over 24 hours. This
is about six times higher than the 0.1 'ng/cm2 flux observed by Yang et al. (1989) over 24 hours in rats in vivo.
Given the difference in the experimental systems [0.1 ng/cm2 measured in vivo in the rat for a concentration of 1
ppm BaP and 5.6 ng/cm2 (comparable to 0.56 ng/cm2 for a concentration of 1 ppm) measured in vitro in human skin
samples for a concentration of 10 ppm], a six-fold difference in estimated fluxes for the 1 ppm concentration seems
reasonable.
In rhesus monkeys, however, the average flux measured by Wester et al. (1990) was about an order of
magnitude higher than would be predicted from the Yang et al. (1989) in vivo and in vitro rat testing and the Wester
et al. in vitro testing of human skin. The average percentage of BaP absorbed from a soil loading of 40 mg/cm2 with
a concentration of 10 ppm BaP was 13.2% over 24 hours in rhesus monkeys. The amount absorbed was 53 ng/cm2
(40 mg soil/cm2 x 10 ng/mg x 0.132), about 9 times greater that the 5.6 ng/cm2 flux measured by the Wester et al.
(1990) in vitro testing in human skin (40 mg soil/cm2 x 10 ng/mg x 0.0141). The flux measured by Wester et al.
(1990) of 53 ng/cm2 in rhesus monkeys is equivalent to a flux of 5.3 ng/cm2 for a concentration of 10 ppm. A flux
of 5.3 ng/cm2 is 63 times larger than the amount absorbed in rats over the same time period measured by Yang et
al. (1989).
The absorption percent of 13.2% for a 40 mg/cm2 soil loading is recommended for use in exposure
assessment because it is much larger than percents absorbed in other species and in in vitro studies and the
differences in the results are not readily explainable. Table 6-5 shows recommended percents absorbed and amounts
absorbed for a concentration of 1 ppm, based on the rhesus monkey data.
1 Yang et al. in vivo 24 hours - 9 mg/soil / cm2 x 0.011 = 0.1 ng / cm2.
6-16
-------
Table 6-5. Dermal Absorption Percentages for Dilute Concentrations of BaP in Soil
(24 Hour Duration of Exposure)
Soil
Loading
(mg/cm2)
0.5
1
10
20
40
Percent
Absorbed
lOOf
loo*
53e
26C
13b
Amount
Absorbed
(ng/cm2)'
5
10
53
53
53
* Assumes a BaP concentration of 10 ppm.
b Average percent applied dose absorbed measured over 24 hours in rhesus monkeys by Wester et al. (1990) data.
c Calculated from estimated percentage absorbed for a soil loading of 40 mg/cm2 and a 24 hours exposure duration
by assuming that the quantity of BaP absorbed is constant (e.g., 13.2%/100% x 40 mg soil/cm2 x 10 ng BaP/1
mg soil = 53 ng BaP/cm2 absorbed over 24 hours; 53 ng BaP absorbed/00 mg soil/cm2 x 10 ng BaP/mg soil) x
100% = 53%, which is the estimated percentage absorbed over 24 hours for 10 ppm BaP in soil applied at 10 mg
soil/cm2).
d Below a soil loading of 5.3 mg/cm2, the amount absorbed begins to decrease, while the percentage of applied dose
absorbed will be 100%. At a soil loading of 0.5 mg/cm2 and a concentration of 10 ppm BaP, the estimated
absorption at 24 hours would be 5 ng/cm2 or 100% of the total applied dose.
6.3.1.4 DDT
Wester et al. (1990) studied percutaneous absorption of DDT in vitro in human skin and in vivo in rhesus
monkeys. As in the study of BaP, they used a soil loading of 40 mg/cm2, a concentration of 10 ppm DDT in the
soil, and an exposure duration of 24 hours, at which time excess soil and DDT were removed with soap and water.
The soil used in the BaP experiment was also used in the DDT experiment. In the in vivo experiment using human
skin, 1 % of the applied dose of DDT was found in the skin after 24 hours and 0.04% was found in the receptor fluid,
for a total of 1.04% of the applied dose absorbed over 24 hours.
In rhesus monkeys, the average percentage of DDT absorbed from a soil loading of 40 mg/cm2 with a
concentration of 10 ppm DDT was 3.3%. The amount absorbed was 13 ng/cm2 (40 mg soil/cm2 x 10 ng/mg x
0.033).
The absorption percent of 3.3% for a 40 mg/cm2 soil loading is recommended for use in exposure
assessment because it is larger than percents absorbed in other species. Table 6-6 shows recommended percents
absorbed and amounts absorbed for a concentration of 10 ppm, based on the rhesus monkey data.
6-17
-------
Table 6-6. Dermal Absorption Percentages for Dilute Concentrations of DDT in Soil
(24 Hour Duration of Exposure)
Soil
Loading
(mg/cm2)
0.5
1
10
20
40
Percent
Absorbed
100"
loo*
13C
6.6=
3.3b
Amount
Absorbed
(ng/cm2)'
5
10
13
13
13
1 Assumes a DDT concentration of 10 ppm.
b Average percutaneous absorption measured over 24 hours in rhesus monkeys by Wester et al. (1990) data.
c Calculated from estimated percentage absorbed for a soil loading of 40 mg/cm2 and a 24 hours exposure duration
by assuming that the quantity of DDT absorbed is constant (e.g., 3.3%/100% x 40 mg soil/cm2 x 10 ng DDT/1
mg soil * 13 ng DDT/cm2 absorbed over 24 hours; 13 ng DDT absorbed/00 mg soil/cm2 x 10 ng BaP/mg soil)
x 100% e 13%, which is the estimated percentage absorbed over 24 hours for 10 ppm DDT in soil applied at 10
mg soil/cm2).
d Below a soil loading of 1.3 mg/cm2, the amount absorbed begins to decrease, while the percentage of applied dose
absorbed will be 100%. At a soil loading of 0.5 mg/cm2 and a concentration of 10 ppm DDT, the estimated
absorption at 24 hours would be 5 ug/cm2 or 100% of the total applied dose.
6 J3. Estimating Dermal Uptake When Data Are Lacking
This report has identified only seven chemicals whose percutaneous absorption from a soil matrix has been
studied — TCDD, TCB, BaP, DDT, benzene, toluene, and xylene. Only the first four chemicals have been studied
under conditions which resemble environmental exposures at Superfund sites. The three VOC studies used occlusion
to prevent evaporation. Therefore, the VOC studies do not represent exposures likely to occur under field conditions.
The studies of the four semivolatile compounds attempted to duplicate environmental exposures to contaminants in
soil. However, even in interpreting these studies, differences in soil loading, soil organic carbon content, and time
exposed to the soil between the experiments and the exposure scenarios must be resolved.
The principal goal of this report is to provide methodologies for assessing exposure to chemicals in water,
and the intent of Section 6 is merely to summarize existing studies of percutaneous absorption of chemicals in soil.
In-depth analysis of modeling approaches and other methods of estimating percutaneous absorption are deferred to
Phase II of the Interim Guidance for Dermal Exposure Assessment.
6-18
-------
Nevertheless, the Agency is presently assessing dermal exposure to chemicals in soil, and this section will
provide what guidance is available for chemicals with no data. Several methods of making these estimates will be
considered: (1) use of data on TCDD, TCB, BaP, and DDT to estimate uptake of structurally similar chemicals;
(2) use of data on uptake of the neat compound as a surrogate for uptake of the compound in soil; (3) use of Kj,
values generated for the compound in water (not presently recommended); (4) use of theoretical models to estimate
uptake; (5) use of default values based on physicochemical properties; and (6) use of data on the amount of the
compound in the soil that can be extracted under conditions expected to extract at least as much compound as would
partition to the skin during the exposure period.
When the assessor has no data, he or she should consider such things as uptake of the neat compound,
modeling results, and default values to arrive at a best estimate of dermal uptake from soil. Where uncertainty
remains high, the assessor should use a default of 100% absorption over 8 to 12 hours for a soil loading of 0.5 to
1 mg/cm2.
EPA's Exposure Assessment Group (BAG) is in the process of collecting and analyzing the literature on
percutaneous absorption of neat compounds and on methods of estimating dermal exposure to chemicals in soil.
Metals are the first chemicals which will be studied. EPA assessors are invited to call EAG to obtain the latest
information from these assessments.
6.3.2.1 Use of Structural Analogues
Several chemicals are similar in structure to TCDD, TCB, or BaP. Where chemicals are similar in structure,
the uptake measured for these chemicals may be used for the analogue chemicals, provided there is no reason to
suspect that the uptake of the analogue chemicals could be significantly higher.
TCDD — Analogues of TCDD include the other polychlorinated dibenzo-p-dioxins (PCDDs), which differ
from TCDD only in the number and location of chlorine substituents, and polychlorinated dibenzofurans (PCDFs),
which differ from TCDD in having one oxygen bridge instead of two between the aromatic rings and in the number
and location of the chlorine substituents. Brewster et al. (1989) studied absorption of neat TCDD applied in acetone
to rats in vivo at six dosages ranging from 11 ng to 72.5 ug and three PCDFs — 2,3.7.8-tetrachlorodibenzofuran
(TCDF); 1,2,3,7,8-pentachlorodibenzofuran (1-PeCDF); and 2,3,4,7,8-pentachlrodibenzofuran (4-PeCDF)—at three
dosages ranging from 7.2 pg to 72 pg. The lowest dosage of PCDFs used in Brewster et al. (1989) is 100 times
higher than the dose used in Roy et al. (1990), 300 to 3,000 times the dosages in Shu et al. (1988), and 5 to 20 times
the dosages used in Poiger and Schlatter (1980). At the lowest dosage used for all four chemicals (7.2 ug), the
6-19
-------
percents of applied dose absorbed after 72 hours of exposure in the rat in vivo were 17.8% of TCDD, 48.8% of
TCDF, 253% of 1-PeCDF, and 34.2% of 4-PeCDR Brcwster et al. (1989) report that the percents absorbed for
TCDF and 4-PeCDF were significantly greater than the percent absorbed for TCDD. However, at the next highest
dosage. 36 ug, uptakes were 19% for TCDD, 17.9% for TCDF, 8.3% for 1-PeCDF, and 24.5% for 4-PeCDF, with
only the 4-PeCDF uptake exceeding the percent absorbed for TCDD. At the highest dosage, 72 ug, the uptake of
TCDD was about the same as that of 4-PeCDF and higher than that of the other two TCDFs. These results are
consistent with the Roy et al. results for in vivo rat studies. At 96 hours, Roy et al. (1990) observed that 77% of
the TCDD applied at a dose of 70 ng was absorbed. At 72 hours, Brcwster, et al. (1989) observed that 40.3% of
the TCDD applied at a dose of 72 ng was absorbed. Both studies observed high absorption of the neat compound.
Roy et al. observed mat 16.3% of the dose applied in a low organic carbon content soil was absorbed over 96 hours.
The percents absorbed of TCDD in soil, estimated for humans in vivo in Section 6.3.1 range from 0.62 to 5% for
a soil loading of 10 mg/m2, from 62 to 50% for a soil loading of 1 mg/cm2, and from 12 to 100% for a soil loading
of 0.5 mg/cm2. Since the percutaneous absorption of neat PCDFs is similar to the absorption of neat TCDD and
recommendations for percents absorbed from soil for humans are large, use of the percents absorbed for TCDD in
soil are recommended for use in assessments of other polychlorinated dibenzo-p-dioxins and polychlorinated
dibenzofurans.
PCBs — Analogues of TCB include the other polychlorinated bipnenyls (PCBs), which differ only in the
number and location of the chlorine substituents. Samples from contaminated waste sites are often analyzed for
Araclors, commercial mixtures of PCB isomers, which were widely used as lubricants, hydraulic fluids, heat transfer
fluids, and dielectric fluids in applications where fire retardant fluids were required. Araclors are mixtures of PCBs
containing 1 to 10 chlorines. The greatest percentage of these mixtures consists of the most persistent Araclors —the
tetra-, penta-, and hexachlorinated biphenyls.
Data exist on absorption of neat hexachlorobiphenyl and neat Araclor mixtures containing 42% and 54%
chlorine by weight. Shah et al. (1987) studied absorption of neat 2,4,5,2',4',5'-hexachlrobiphenyl in vivo in adult
and young rats. At a dose of 16 ug/cm2,33.5% of the dose was absorbed in the adult rat and 40.7% of the dose was
absorbed in the young rat after 72 hours. Wester et al. (1983) studied percutaneous absorption of PCB mixtures
containing 42% and 54% chlorine by weight in vivo in rhesus monkeys and guinea pigs. Doses were removed from
the skin after 72 hours. Collection of urine and feces continued for many days. For a dose of 4.1 ug/cm2 of 42%
PCBs, total PCBs excreted were 17.3% and 33.9% of the applied dose for two monkeys. For a dose of 19.3 ug/cm2
of 42% PCBs, the total PCBs excreted were 15.2% and 19.4% of the applied dose for two monkeys. In guinea pigs,
33.2% of a dose of 4.6 pg/cm2 of 42% PCBs was excreted, and 55.6% of a dose of 52 jig/cm2 of 54% PCBs was
excreted. In the same study, percents dose excreted after parenteral administration were 65.8% for 42% PCB
6-20
-------
administered to Rhesus monkeys, 52.3% for 42% PCB administered to guinea pigs, and 61.5% for 4% PCB
administered to guinea pigs. According to Wester et al. (1983), the fraction of the dose absorbed topically is equal
to the total 14C urinary excretion following topical administration divided by total 14C urinary excretion following
parenteral administration. However, it does not appear that the authors made this correction, since they refer to the
data obtained from topical application interchangeably as the percent dose excreted and the dermal absorption. Since
the parenteral and topical doses of radioactivity were equal, it seems reasonable to divide the percent excreted in both
urine and feces after topical application by the percent excreted in urine and feces after parenteral application to
obtain the percent of applied dose absorbed. With these adjustments, the average percents of applied dose absorbed
for a 24-hour exposure are 39% of 42% PCBs in Rhesus monkeys at a dose of 4.1 ug/cm2, 26% of 42% PCBs in
Rhesus monkeys at a dose of 19.3 ug/cm2,63% of 42% PCBs administered to guinea pigs at a dose of 4.6 ug/cm2,
and 90% of 54% PCBs administered to guinea pigs at a dose of 52 ug/cm2.
In Roy et al. (1990), 48% of a dose of 70 ug of TCB applied in low organic carbon content soil was
percutaneously absorbed after 96 hours in vivo in the rat. This absorption rate is somewhat larger but comparable
to the rate measured in Shah et al. in rats over 72 hours for neat hexachlorobiphenyl. From Roy et al. (1980) data.
it has been estimated in Section 6.1 that the comparable percutaneous absorption in humans would be 12% of applied
dose over 96 hours. The uptake of neat PCBs in Rhesus monkeys is similar to the uptake of the neat compound
from soil in the rat (26% to 39% for a 24-hour exposure for monkeys and 33% to 41% for a 72-hour exposure for
rats). Uptake of TCB appears to equal or exceed uptake of similar PCBs. Therefore, it is appropriate to apply the
recommendations in Section 6.1 to all PCBs and Araclors.
BaP and DDT — BaP is one of a class of polyaromatic hydrocarbons (PAHs). This is a broad and not well-
defined class. Physicochemical properties and percutaneous absorption vary over a wide range, and it would not be
appropriate to use the recommendations in Section 6.2 for BaP in assessing other PAHs.
No environmentally significant analogues to DDT have been identified.
6.3.2.2 Use of Values of Percent Applied Dose Absorbed for the Neat Compound.
In many cases, percutaneous absorption of the neat compound will have been studied when data on
absorption of the chemical applied in soil is not available. As mentioned in the introduction to this subsection,
collection and analysis of these data are outside the scope of this report, and an ongoing EPA/EAG project is
addressing this issue.
6-21
-------
This approach must be used with extreme caution. One should determine whether the applied dose in the
experiment with the neat compound is similar to the applied dose in the exposure scenario. If the experimental dose
is much larger, which may often be the case, it may be more appropriate to use the experimental results for total
amount of compound absorbed rather than the experimental results for percent of applied dose absorbed in the
exposure assessment For example, if the neat applied dose was 4 ug/cm2, a common dosage used in in vitro
experiments, and the percent of applied dose absorbed was 5% over 24 hours, the experimentally determined
absorbed dose would be 200 ng/cm2. If the applied dose in the exposure scenario is 1 ug/cm2, multiplying the
applied dose by the 5% applied dose absorbed produces an uptake for the exposure scenario of only SO ng/cm2,
compared to 200 ng/cm2 for the neat compound. In this case, it would be more appropriate to assume an absorbed
dose of 200 ng/cm2 for the exposure scenario based on uptake of the neat compound, comparable to absorption of
20% of the applied dose, rather than 5% of the applied dose. However, a materials balance should be done to ensure
that the total applied dose in the exposure scenario is greater than or equal to this assumed absorbed dose.
A second aspect of this approach is the problem of applying measurements using animal or in vitro models
to humans. In general, rat skin is more permeable than human skin, and data obtained from the rat may be
considered an upper-bound on absorption in humans in the absence of contrary data. Rhesus monkey skin is believed
to behave more like human skin. Guinea pig skin, on the other hand, is more similar to rat skin than human skin
with regard to permeability.
When data collected in vitro are used, the data should be compared to in vivo results if possible, and the
assessor should feel confident that the in vitro experimental protocol does not prevent absorption, particularly when
percutaneous absorption measured in vitro is low. For example, sometimes receptor fluids are used which prevent
penetration in vitro because they do not adequately solubilize the penetrant.
6-3JU Use of Kp Values for the Compound in Water or Other Fluids.
Since there are many Kp values for organic compounds in water, it is tempting to assume an infinite soil
loading, convert the concentration of penetrant in the soil to mg penetrant/cm3 of soil by multiplying by the soil
density, and use this concentration and the Kp for the penetrant in water to calculate an upper-bound on the flux.
This approach is not recommended. EPA/EAG has begun studying this approach, and so far it appears to have little
predictive value for absorption of chemicals in soil. This issue will be addressed further in the second phase of the
Interim Guidance.
6-22
-------
6 J J.4 Theoretical Modeling
A method for estimating the dermal absorption of compounds from a soil matrix has been proposed by
McKone (1989). Central to this approach is a fugacity model that uses the physical-chemical properties of the
compound and the soil to estimate transport across the combined soil and skin layer, taking evaporation into account.
Although this model has yet to be validated with large numbers of compounds, it appears to provide a potentially
useful approach for estimating compound-specific percutaneous absorption values (i.e., percent of applied dose
absorbed during a specified exposure duration) for organic soil pollutants.
Using the McKone model, Burmaster and Maxwell (1990) estimated the uptake of benzene, naphthalene,
phenanthrene, fluoranthene, benzo[a]pyrene, and indeno(l,2,3-cd)pvrene for a 12-hour exposure duration as
a function of soil loading.
For BaP, the one compound for which data are available, the McKone (1989) model estimated 99% uptake
for a loading of 0.1 mg soil/cm2, about 55% uptake for 1 mg soil/cm2, about 2% for 10 mg soil/cm2, and less than
1% for 32 mg soil/cm2 over 12 hours. Yang et al. (1989) measured a 1.1% uptake of BaP in the rat after 24 hours
for a 9 mg soil/cm2 loading and a 1.3% uptake at 96 hours for a 56 mg soil/cm2 loading. In human skin, Wester
et al. (1990) measured a 1.41% uptake of BaP after 24 hours for a 40 mg/cm2 soil loading. The results from the
McKone (1989) model seem consistent with experimental data in the rat. However, Wester et al. measured an
average uptake of 132% after 24-hour exposure in Rhesus monkeys, a much higher absorption than that predicted
by the model or by the Yang et al. (1989) results in the rat and the Wester et al. results using human skin in vitro.
The models predictions at lower soil loadings (1 mg/cm2 and 0.3 mg/cm2) are consistent with the recommendations
in Section 6.2 to use a 100% uptake for BaP for low soil loadings.
For indeno(l,2,3-cd)pyrene, Burmaster and Maxwell (1990) estimated about 97% uptake over 12 hours for
a soil loading of 0.1 mg/cm2, about 25% uptake for a loading of 1 mg/cm2, and 1% uptake for a loading of 10
mg/cm2. For fluoranthene, they estimated about 90% uptake for loadings between 0.1 and 3 mg/cm2, a 40% uptake
for a loading of 10 mg/cm2, about 8% uptake for a loading of 32 mg/cm2, and less than 1% at 100 mg/cm2. For
naphthalene, they predicted about 10% uptake for soil loadings between 0.1 and 3 mg/cm2, increasing to about 30%
uptake at a loading of 100 mg/cm2. The model predicts that uptake of benzene is about 1 to 2% for loadings of 0.1
to 10 mg/cm2 and increases to about 8% for a loading of 100 mg/cm2.
6-23
-------
Burmaster and Maxwell concluded that uptake fraction is a strong, nonlinear function of the number of rings
and recognized the octanol-water partition coefficient (K,,w) and the dimensionless Henry's Law constant (K,,) as
physicochemical properties which can be used to predict the uptake fraction.
6.3.2.5 Default Values for Organic Compounds
McKone (1989) observed that there are few data on dermal uptake of chemicals from a soil matrix with
which to test his model. However, he drew several generalizations from a theoretical analysis of dermal uptake of
chemicals from soil. He observed that dermal uptake is influenced strongly by the soil loading on the surface of the
skin and the K^w and Kj, of the penetrant K,,w is a unitless partition coefficient and represents the ratio of the
concentration of the penetrant in octanol to the concentration of the penetrant in water under equilibrium conditions.
Kj, is the dimensionless form of the Henry's Law Constant, H. H is equal to the vapor pressure divided by the water
solubility of the penetrant.
Kk - " (6-1)
* (R x T)
where,
H = Vapor pressure (atm) / water solubility (mol/m3);
R = The ideal gas constant (8.205 x 10"5 m3-atm/mol-°K);
T = Ambient temperature (degrees Kelvin).
McKone (1989) drew some generalizations, which are applicable to the default soil loading of 1 mg/cm2:
For KH < 0.001 and K,,w £ 106, assume 100% uptake over 12 hours;
• For K,, ^ 0.1, assume no more than 3% uptake over 12 hours; and
• For K,, between 0.01 and 0.1, assume no more than 40% uptake in 12 hours, with uptake likely
to be much lower when K,,w is greater than 10.
For compounds not listed in the above three defaults (i.e., those with K,, between 0.001 and 0.01, and those
with both KJ, less than 0.001 and Kow greater than 106), assume 100% uptake if no other data are available.
6-24
-------
The McKone model is only applicable for organic compounds; however, metals constitute an important class
of soil pollutants. Nevertheless, no studies were identified in which the dermal uptake of metals from soil was
reported.
633.6 Extraction of Contaminants from Soil
In some cases, it may be possible to determine the maximum amount of contaminant that is likely to be
released from soil to the skin. If soil collected from the site is subjected to extraction procedures which are expected
to be at least as efficient in extracting the chemical from soil as the dermal contact described in the exposure
scenario, the extraction efficiency (percent extracted) may be used as an upper-bound estimate of the percent of
applied dose absorbed.
In Phase II of this Guidance, EPA/EAG will investigate extraction procedures and identify chemicals that
can be assessed using this methodology.
6-25
-------
7.0 DERMAL ABSORPTION OF CHEMICAL VAPORS
Volatile compounds in the vapor phase are assumed to be absorbed primarily by the respiratory tract.
However, the capacity of compounds in the vapor phase to cross the skin has been demonstrated by a number of
researchers (Hanke et al., 1961; Dutkiewicz and Piotrowski, 1961; Riihimaki and Pfaffli, 1978, Wieczorek, 1985;
McDougal et al., 1985,1987,1990). Furthermore, McDougal et al. (1987) have pointed out that because of the large
surface area of the skin, compounds with even small flux values may be extensively absorbed across the skin.
Therefore, it is important to examine the conditions under which dermal exposure to chemical vapors may contribute
significantly to absorbed dose.
7.1 Experimentally Derived Values
Both in vivo and in vitro techniques have been used to obtain dermal absorption rate measurements for
chemical vapors. In vivo exposures have been conducted by placing rats (McDougal et al., 1985, 1987, 1990) or
human volunteers (Hanke et al., 1961; Dutkiewicz and Piotrowski, 1961; Riihimaki and Pfaffli, 1978; Wieczorek,
1985), wearing facemasks to provide respiratory protection, in whole-body exposure chambers. The amount of the
compound absorbed in the human exposure studies is determined either by monitoring the appearance of metabolites
in blood or postexposure urine samples or by measuring the concentration of parent compound in expired air. In
their rat studies, McDougal et al. (1985, 1987, 1990) measured blood concentrations of the compounds over the 4-
hour exposure period. These researchers then used a physiologically based pharmacokinetic (PB-PK) model to
optimize the K,, values that gave a best fit to the blood concentration-time profile data. The permeability constant
(Kp) and flux (J) values obtained by McDougal et al. (1985,1987,1990) using this approach for a series of volatile
organic compounds can be found in the dermal permeability database, which is described in Appendix B. In
addition, these values are summarized in Table 7-1.
7-1
-------
Table 7-1.
Dermal Vapor Absorption in Rats In Vivo
Compound
Styrene
m-Xylene
Toluene
Perchloroethylene
Benzene
Halothane
Hexane
Isoflurane
Concentration
(ppm)
3,000
5,000
8,000
12,500
40,000
50,000
60,000
50,000
Flux
(mg/cm2/hr)
0.0211
0.0151
0.0206
0.0541
0.0191
0.0180
0.0065
0.00%
Permeability Constant*
(cm/hr)
1.753 ± 0.105
0.723 ± 0.003
0.721 ± 0.007
0.668 ± 0.080
0.152 ± 0.006
0.045 ± 0.005
0.031 ± 0.004
0.025 ± 0.004
* Permeabilities are expressed as means ± SD, which are obtained by an estimation program. These numbers are
not a measure of the variability in the rat population, but reflect the confidence that could be placed on the mean
value for permeability.
Source: McDougal et al. (1990)
McDougal et al. (1990) have reviewed the existing literature on the dermal absorption of vapors in humans.
Absorbed doses in human vapor exposure studies were converted to Kp (McDougal et al. 1990) using the following
relationship derived from Pick's Law:
X ••• ABS
" * SA x C x t
(7-1)
where,
ABS
SA
C
t
The total absorbed dose (mg);
Surface area (cm2);
Exposure concentration (mg/cm3); and
Time (hrs).
These estimated values are presented in Table 7-2 for a number of studies.
7-2
-------
Table 7-2.
Estimated Human Permeability Constants for
Vapor Phase Organic Compounds*
Compound
Styreneb
Styrenec
Styrcne'
m-Xyleneb
m-Xyleneb
Tolueneb
Perchloroethyleneb
Benzene11
Aniline'
Aniline'
Aniline*
Methoxychloi*
C
(mg/cm3)
2.55 x 1(T3
3.25 xlO"3
1.37 x 1(T3
1.30 x 1(T3
2.61 x 1(T3
226 x 1(T3
4.07 x Iff3
1.00 x ia3
5.00 x ia3
LOO x ia2
2.00 x ia2
3.27 x Iff3
t
(hr)
3.5
2.0
2.0
3.5
3.5
3.5
3.5
7.0
6.0
6.0
6.0
3.5
ABS
(mg)
60.1
175.0
45.0
20.8
44.5
26.4
47.1
10.0
2.1
Flux
(mg/cm/hr)
1.90 x KT4
2.50 x 104
4.00 x KT4
KP
(cm/hr)
0.35
1.42
0.87
0.24
0.26
0.18
0.17
0.08
0.04
0.03
0.02
0.01
* See text for equation; K_ estimated assuming a skin surface area of 19,000 cm2.
b Riihimalti and Pfaffli, 1978
c Wieczorek, 1985
d Hanke et al., 1961
e Dutkiewicz and Piotrowski,
Source: McDougal et al. (1990)
In addition to the in vivo approach for obtaining dermal absorption rate values for chemical vapors, a
modification of the two-chamber diffusion cell has been used to obtain in vitro measurements of dermal vapor
penetrations.
Using a similar experimental design, Barry et al. (1984) obtained flux values for a number of organic
compounds permeating human skin in vitro as either a vapor or a neat liquid (Table 7-3).
7-3
-------
Table 7-3.
Flux Values for Organic Compounds Permeating Human Skin In Vitro
as a Saturated Vapor and as a Liquid
Compound
Anisole
Benzaldehyde
Aniline
Benzyl alcohol
2-Phenylethanol
420
410
260
52
27
Flux* (ug/cm2/hr)
Vapor
(100)
(70)
(50)
(12)
(8)
990
1,970
1,870
540
650
Liquid
(300)
(720)
(U60)
(240)
(60)
' Flux (± sd)
Source: Bany et al. (1984)
An alternative approach to estimating the permeability of chemical vapors through the skin was used by
Scheuplein and Blank (1971). From measured solubilities of the compounds in the skin, data on the "sorption" of
these compounds to dry stratum comeum tissue, and an assumed thickness of the stratum corneum, these
investigators estimated gas-phase Kp values for a series of homologous alcohols and alkanes. These values are
summarized in Table 7-4.
7-4
-------
Table 7-4. Estimated Permeability Coefficient Values (cm/hr)
for Alcohol and Alkane Saturated Vapors
Compound
Carbon Number
c,
C^
C3
C4
Cj
C6
Cj
Cg
Cj
c,.
cu
CM
Alcohols
0.05
0.02
0.01
0.02
0.04
0.12
0.11
0.04
0.05
0.05
Alkanes
0.0016
0.0062
0.0104
0.022
0.058
0.103
0.124
0.254
0.92
Source: Scheuplein and Blank (1971)
In addition to organic vapors, other gases are also known to permeate the skin. Scheuplein and Blank
(1971) summarized the results of investigators who obtained K,, values for permeant gases (Table 7-5).
Table 7-5. Flux and Permeability Coefficient Values for Permeant Gases in Humans
Chemical
Helium
Argon
Nitrogen
Carbon Dioxide
Oxygen
Water
Flux
(umole/cmVhr)
0.18
0.20
0.11
0.49
0.49
27.8
(cm/hr)
0.67
0.21
0.25
0.24
0.46
0.0007
Source: Scheuplein and Blank (1971)
7-5
-------
The values presented in Table 7-5 have not been validated by recent investigators and, except for oxygen, they were
obtained for the efflux1 of the gas across human skin in vivo.
Hursh et al. (1989) recently conducted a study to investigate the dermal uptake of mercury vapor, a
lexicologically significant inorganic compound. By measuring the difference between the accumulated radioactivity
on exposed and unexposed forearms of human volunteers, these researchers were able to quantify the dermal uptake
of mercury. The mean uptake rate for five subjects exposed to various mercury vapor concentrations was 0.024 ng
mercury per cm2 skin per minute per ng mercury per cm3 air (units expressed by the authors). This value, however,
probably represents both uptake into the systemic circulation, and uptake and retention by the stratum comeum. The
rate of uptake by skin was estimated as about 2.6% of the rate of uptake by the lung. Desquamation of the stratum
comeum causes loss of about half of the skin uptake (which is released by the outward movement of mercury-
containing cells below the stratum comeum).
7.2 Factors Affecting the Dermal Absorption of Vapors
The factors that affect the dermal absorption of neat compounds or compounds in aqueous solution would
be expected to affect the absorption of chemical vapors in a similar fashion. One factor known to affect dermal
absorption rate is the concentration of the compound in the donor solution (Section 4.0). The primary difference
between a pure liquid and the vapor phase of the compound, relative to dermal permeation, is the concentration
difference between these states. The pure liquid form is far more concentrated than the vapor form. As a result,
the flux of the pure liquid form is expected to be greater than that of the compound in the vapor-phase. The results
of Barry et al. (1984) support this hypothesis.
McDougaJ et al. (1990) point out that the physical form of the chemical, i.e., liquid or vapor, will affect
the concentration of the compound at the skin surface, but should not affect the solubility of the compound in the
skin. This statement is true unless damage occurs to the stratum comeum from the pure liquid compound.
Similar to compounds in the neat form or in an aqueous vehicle, the rate of dermal permeation of organic
vapors is highly correlated with their lipid solubility. McDougal et al. (1990) demonstrated that a general trend
between fat/air partition coefficients and Kp values exists in the rat. For example, styrene vapor, with a fat/air
partition coefficient approximately seven times greater than that of benzene, penetrates rat skin about 11.5-fold
1 Efflux is here defined as the unidirectional outward flow (of an inhaled gas) from the blood
stream to the skin surface.
7-6
-------
greater than benzene. As shown below, this relationship should prove useful for making rough predictions about the
dermal absorption of organic chemical vapors in the absence of experimental data.
13 Methodologies For Estimating the Dermal Absorption of Chemical Vapors in the Absence of
Experimental Data
Techniques for estimating compound-specific Kp values for chemical vapors in the absence of experimental
data have not been reported. Based on the flux of neat, liquid compounds applied to the skin and compounds in the
vapor phase (Table 7-3), the neat, liquid flux value, if available, could provide a very rough, order-of-magnitude
estimation of the flux of the compound in the vapor phase. This approach is only valid, however, if the neat, liquid
compound does not damage the stratum comeum. However, since Kp = J/AC at steady-state, the Kp of the liquid
form of the compound would be less than that for the vapor, because of the increased concentration of the liquid
form. (The greater number of molecules of liquid in contact with skin are in part counterbalanced by more rapid
movement of vapor phase molecules.)
The relationship between lipid solubility and permeability for a number of organic compounds in the vapor
phase suggests that a linear regression equation using these variables may serve as a useful tool to provide rough,
order-of-magnitude predictions for similar compounds. Using the K_ and fat/air partition coefficient data provided
by McDougal et al. (1990), the following regression equation can be used to estimate Kp values of organic vapors,
if fat/air partition coefficient data are known:
KP** " (KFA ' 0.00049) - 0.0385 (r2 * 0.956) <7'2)
where Kpe, is the expected Kp value and KFA is the fat/air partition coefficient. Fat/air partition coefficient values
for the rat are reported by Gargas et al. (1989) for 55 compounds. However, a lack of data from other researchers
suggests that further validation of this relationship is required. Also, insufficient data were available to develop a
similar relationship to estimate the dermal absorption of organic vapors in humans. Although this approach must
be used cautiously, it nevertheless may be useful for providing a rough approximation of the dermal permeation rates
of organic vapors in humans if values predicted for the rat are divided by a factor of two to four.
7-7
-------
8.0 METHODS FOR PREDICTING PERMEABILITY COEFFICIENTS
OF AQUEOUS CONTAMINANTS
As evidenced by the relatively small number of K,, or J values present in the dermal database, permeability
coefficients are available for only a few environmental pollutants. This lack of pertinent data seriously limits the
ability of Regional or Program offices to conduct dermal exposure assessments. In the absence of experimentally
derived values, the assumption of 100% percutaneous absorption has been used by some Program Offices; however,
this approach has been criticized as being overly conservative. For contaminants in water, this assumption translates
into using the exposed dose as the absorbed dose, without applying Equation 2.1 in the calculation. Alternately, the
Superfund Risk Assessment Guidelines has recommended the use of the permeability coefficient of water in
Equation 2.1 in cases where absorption data are lacking.
From the current literature, two general types of structure-activity models (empirical and theoretical) have
been recommended to estimate permeability coefficients for chemicals in aqueous solution:
• Empirical models were based on actual experimental permeability coefficients for one or several
classes of chemicals. In general, the permeability coefficients of several chemicals of the same
class were measured, and an empirical correlation was determined, usually as a function of some
physical or chemical properties, such as partition coefficient and molecular weight, although other
physicochemical properties have also been used. This relationship was then employed to predict
the permeability coefficient of compounds in the same chemical class. When the correlation was
obtained from a wide range of chemicals, it can be used more widely to provide a first estimate
of the permeability coefficient of any compound, based on a few known physicochemical
properties.
• Theoretical skin permeability models were developed based on an assessment of the contribution
of physicochemical properties of the skin and their interaction with the chemicals. Some models
might describe the percutaneous absorption of certain classes of chemicals better than others,
depending on the assumptions of the skin structure and composition as they affect the percutaneous
absorption processes.
The models are described in this section, together with the assumptions and limitations of their validity and
accuracy in predicting permeability coefficients and fluxes of water contaminants. The conditions under which the
current Superfund approach would provide a conservative estimate of percutaneous absorption will also be identified.
8-1
-------
8.1 Empirical Correlations
The general characteristics of the skin, as described in Chapters 3 and 4, determine to a large extent the
limits of percutaneous absorption of chemicals from contaminated water. As a vehicle, water hydrates the skin and
may enhance absorption. The relative partitioning of the contaminants between water and skin, and between the skin
and the systemic circulation, govern the overall absorption of the chemicals into the body. Therefore, water solubility
of the pollutants sets the upper concentration limit for the absorbed dose. Other physicochemical properties of the
pollutants define their interactions with the various components of the skin and determine its diffusion through the
skin.
Flynn (1990) proposed the following working model of the skin to assess the permeation of chemicals by
their physicochemical determinants. The two main layers of the skin are the epidermis and the dermis. The stratum
comeum constitutes the outer dead layer of the epidermis and provides the main barrier to absorption of many
chemicals into the skin. The stratum comeum is composed of sheets of flattened keratinous material interspaced with
intercellular lipids. Flynn (1990) estimted that keratin occupies 65% of this mass and lipids about 25%. There is
evidence that penetrant chemicals are transferred through the stratum comeum via an aqueous pore pathway and a
non-polar pathway. Data from many sources indicate that the stratum corneum behaves as a hydrophobia membrane
to organic nonelectrolytes. Without the stratum corneum, the living epidermis would appear to provide little
resistance to diffusion, except for extremely polar molecules which are limited in their capacity to penetrate the
viable tissues. However, there is only limited information about the permeability of the living epidermis per se, and
in the absence of the statum corneum, the living epidermis provides little resistance to absorption of polar
compounds.
Flynn (1972,1990) examined in vitro absorption data of aqueous solutions of organic compounds, both large
and small, polar and nonpolar (including homologous alkanols, hydrocortisone and its esters, and vidarabine), through
human and hairless mouse skin as a function of the ether/water partition coefficient From these data, several zones
of permeability behavior can be defined based on the degree of polarity of the permeants:
• For extremely nonpolar molecules such as long chain alcohols, hexanoate, and heptanoate esters
of hydrocortisone, permeation is controlled by an aqueous diffusion barrier
log K,,, > 3, then log K, « -1
8-2
-------
• For middle chain-length alkanols and the rest of the hydrocoitisones, permeation is controlled by
a lipid medium. In this region, log Kp is proportional to log K,,w:
0.5 < log K,, < 3, then -5 < log Kp < -2
• For extremely polar molecules like water, methanol, ethanol, vidarabine, and esters (both small and
large molecules), the permeation is not a direct function of K,,,.
log Kow < 4.5, then log Kp = -5
• For nonelectrolytes (both polar and nonpolar), a size effect was observed and the small molecules
permeate better than the large ones.
• No data were available for small polar nonelectrolytes (other than water).
• The weak electrolytes (salts and acids) are pH dependent, with Kp varying within at least one order
of magnitude for a pH range from 2 to 9, and with those at lower pH being the better skin
penetrants. Unionized forms are thought to be better skin penetrants than ionized forms, and
compounds may be at their peak penetration capacity at their pK (dissociation constant) values.
In general, porous membranes allow the diffusion of free (non-ionized) and ionic species of weak
electrolytes with equal facility, while a lipoidal membrane supports the permeation of undissociated ionic molecules
regardless of polarity, and ionic species in dissociated forms permeate very slowly, even in ion pairs like salt. The
flux of these dissociated ions are several log orders less than the flux of free species at the same concentration.
By plotting log Kp values obtained for a series of homologous alkanols, hydrocortisone-21 esters, and other
miscellaneous compounds against their log Kow values, Flynn and Stewart (1988) were able to ascribe upper and
lower bounds on the Kp values for compounds based on their relative polarity. Three patterns emerged from this
evaluation. For highly polar compounds (log K,w < -2.301), a lower bound Kp value of 10~* cm/hr was assigned.
Compounds with log KgW values from -2.301 to 2.000 had Kp values that increased somewhat linearly from 10~* to
10"2 cm/hr. The following relationship was developed to empirically determine Kp values in this Kow range:
8-3
-------
log Kf « log ATW - 3.698
(8-1)
The upper limit for the most lipophilic compounds in the study (log K^ > 2.000) was a K,, value of about 10'2
cm/hr.
In a later publication, Flvnn (1990) applied the same model to human skin. Flynn (1990) compiled
published permeability coefficients from the existing literature and looked at their dependence on KDW and molecular
weight. Permeability data obtained in aqueous solutions were tabulated for about 100 compounds, many of which
were drugs (Table 8-1). Two groupings of the Kp's were observed as a function of molecular weight. The
algorithms in Table 8-2 were proposed by Flynn (1990) for estimating the permeability coefficients in human skin
as a function of KDW and molecular weight
Table 8-1.
Permeability Coefficients for Human Skin (Aqueous Solutions) and Octanol/Water Partition Coefficients (Neat)
of Organic Compounds:
Alphabetical Ordering of Compounds Having Published Permeability Coefficients
Compound
aldosterone
amobarbital
atropine
barbital
benzyl alcohol
4-bromophenol
23-butanediol
butanoic acid (butyric acid)
n-butanol
2-butanone
butobarbital
4-chlorocresol
2-chlorophenol
4-chlorophenol
Molecular
Weight
360.44
22627
289.38
184.19
108.13
173.01
90.12
88.10
74.12
72.10
21224
142.58
128.56
128.56
Kp
(cm/hr)
3.0x10-*
2.3xlO'3
8.5x10-*
l.lxlO-4
6.0xlO'3
3.6xl(T2
4.0xlO'5
1.0xl(T3
2JxlO"3
4.5xlO'3
1.9xl
-------
Compound
chloroxylenol
chlorpheniramine
codeine
coitexolone ( 1 1-desoxy- 17-hydroxycorticosterone)
cortexone (deoxycorticosterone)
corticosterone
cortisone
0-cresol
m-cresol
p-CKSOl
n-decanol
2,4-dichlorophenol
diethylcarbamazine
digitoxin
ephedrine
B-estradiol
B-estradiol (2)
eslriol
estrone
ethanol
2-elhoxy ethanol (Cellosolve)
ethyl benzene
ethyl ether
4-ethylphenol
etorphine
fentanyl
fentanyl (2)
fluocinonide
Molecular
Weight
156.61
274.80
299.30
346.45
330.45
346.45
360.46
108.14
108.14
108.14
158.28
127.55
19929
764.92
165.23
272.37
272.37
28837
270.36
46.07
90.12
10620
74.12
122.17
411.50
336.50
336.50
494.55
Kp
(cm/hr)
52xl(T2
22xlO'3
4.9xl(T5
7.4xia5
4.5X10"4
6.0xlO-5
LOxlO'5
1.6xlO-2
1.5xlO'2
1.8xlO'2
7.9xl(T2
6.0xlO'2
UxlO-4
1.3xlO'5
6.0xl(T3
3.0X10"4
52x10°
4.0X10'5
3.6x10°
7.9XKT4
2.5x10"*
12
1.6xl(T2
3.5xlO'2
3.6xlO'3
5.6xlO'3
1.0xl(T2
1.7xlO'3
logKp
-1.28
-2.66
-4.31
-4.13
-3.35
-422
-5.00
-1.80
-1.82
-1.75
-1.10
-122
-3.89
-4.89
-2.22
-3.52
-228
-4.40
-2.44
-3.10
-3.60
0.08
-1.80
-1.46
-2.44
-225
-2.00
-2.77
log
Kow
3.39
7
0.89
2.52
2.88
1.94
1.42
1.95
1.96
1.95
4.00
3.08
?
1.86
1.03
2.69
2.69
2.47
2.76
-0.31
-0.54
3.15
0.83
2.40
1.86
4.37
4.37
3.19
8-5
-------
Compound
heptanoic acid (enanthic acid)
n-heptanol
hexanoic acid (caproic acid)
n-hexanol
hydrocortisone
hydrocortisone (2)
[hydrocortisone-21-yl]-N,N dimethyl succinamate
[hydrocortisone-21-yl]-hemipimelate
[hydrocoilisone-2 1 -hemisuccinate
[hydrocortisone-2 1 -yl]-hexanoate
[hydrocortisone-21-yl]-6-hydroxy hexanoate
[hydrocortisone-2 1 -yl]-octanoate
[hydrocortisone-2 1 -yl]-pimelamate
[hydrocortisone-2 1 -yl]-proprionate
[hydrocortisone-2 1 -yl]-succinamate
hydromorphone
hydroxypregnenolone
17a-hydroxyprogesterone
isoquinoline
meperidine
methanol
methyl-[hydrocortisone-21-yl]-succinate
methyl-[hydrocortisone-2 1 -yl]-pimelate
methyl-4-hydroxy benzoate
morphine
2-naphthol
naproxen
nicotine
nitroglycerine
Molecular
Weight
130.18
11620
116.16
102.18
362.47
362.47
489.60
504.60
462.50
460.60
476.60
488.70
503.60
41830
461.60
285.30
330.45
330.45
129.15
247.00
32.04
476.60
518.60
152.14
285.30
144.16
23026
16223
227.09
Kp
(cm/hr)
2.0xlO-2
32xlO-2
1.4xlO'2
Uxlfr2
3.0xl04
UxlO"4
6.8xl(T5
1.8xlO'3
6.3xKT*
1.8xlO'2
9.1X10"4
62xlO'2
8.9XKT4
3.4xlO'3
2.6xlO's
UxlO'5
6-OxlO-4
6-OxlO-4
UxlO-2
3.7xlO'3
5.0X10-4
llxlO4
5.4xl(T3
9.1xlO'3
9.3xlO-«
2.8xl(T2
4.0XKT4
1.9xl(T2
l.lxlO-2
logKp
-1.70
-1.50
-1.85
-1.89
-5.52
-3.93
-4.17
-2.75
-3.20
-1.75
-3.04
-1.21
-3.05
-2.47
-4.59
-4.82
-322
-322
-1.78
-2.43
-3.30
-3.68
-227
-2.04
-5.03
-1.55
-3.40
-1.71
-1.96
log
Kow
2.50
2.72
1.90
2.03
1.53
1.53
2.03
326
2.11
4.48
2.79
5.49
2.31
3.00
1.43
1.25
3.00
2.74
2.03
2.72
-0.77
2.58
3.70
1.96
0.62
2.84
3.18
1.17
2.00
8-6
-------
Compound
3-nitrophenoI
4-nitrophenol
tf-nitrosodiethanolamine
n-nonanol
octanoic acid (caprylic acid)
n-octanol
ouabain
pentanoic acid (valeric acid)
n-pentanol
phenobarbital
phenol
pregnenolone
progesterone
n-propanol
resorcinol
salcylic acid
scopolamine
styrene
sucrose
sufentanyl
testosterone
thymol
toluene
2,4,6-trichlorophenol
water
3,4-xylenol
Molecular
Weight
139.11
139.11
134.13
14426
144.21
130.23
584.64
102.13
88.15
232.23
94.11
316.47
314.45
60.10
110.11
138.12
303.35
104.10
342.30
387.50
288.41
15021
92.10
162.00
18.01
122.17
Kp
(cm/hr)
5.6xlO'3
5.6xlO'3
6.0x10-*
6.0xlO'2
2.5xlO-2
52xlO'2
7.8xlO'7
2.0xlO'3
6.0xlO'3
4.6x10^
S.lxlO"3
1.5xlO'3
1.5xlO'3
1.4xlO"3
2.4x10-*
6.3xlO'3
5.0xlO'5
6.5x10-'
5.2X10-6
UxlO'2
4.0x10-*
5.2xlO"2
1.0
5.9xlO'2
5-OxlO-4
3.6xlO'2
logKp
-2.25
-225
-5.22
-U2
-1.60
-1.28
-6.11
-2.70
-2.22
-3.34
-2.09
-2.82
-2.82
-2.85
-3.62
-2.20
-4.30
-0.19
-5.28
-1.92
-3.40
-128
0.00
-1.23
-3.30
-1.44
log
Kow
2.00
1.96
?
3.62
3.00
2.97
?
1.30
1.56
1.47
1.46
3.13
3.77
0.25
0.80
2.26
124
2.95
-2.25
4.59
3.31
334
2.75
3.69
-1.38
2.35
8-7
-------
Table 8-2.
Algorithms for Calculating Permeability Coefficients from
Octanol/Water Coefficients*
log K,, < 0.5
O.SilogK^O.0
OJ3.0
logKow>3.5
Low Molecular Weight
Compounds
log Kp « -3
log KP = -3.5 •»• log Kw
log KP = -0.5
High Molecular Weight
Compounds
log KP « -5
log Kp* -5.5-1- log Kow
log KP = -1.5
'Where Kp = Permeability Coefficient
A similar predictive algorithm was proposed by Vanderslice and Ohanian (1989), based on their evaluation
of Kp values for potential drinking water contaminants in aqueous media. This lists of chemicals used to derive the
following correlations are shown in Table 8-3.
For compounds with:
< 1, then Kf <, 0.001 cm/hr
(8-2)
1 < K^< 500, then Kf «
cm/hr
(8-3)
^
1000
Kaw> 500, then K<, 0.1 cm/hr
(8-4)
Several other investigators have developed regression equations for the prediction of Kp from partition
coefficients and/or molecular weight values for different classes of compounds in aqueous media. Lien and Tong
(1973) looked at the percutaneous absorption experimental data for nonelectroytes through rabbit whole skin and
dermis in vitro (data from Treherne, 1956), aliphatic alcohols through human epidermis in vitro as a function (data
from Scheuplein, 1966), steroids through human epidermis in vitro (data from Scheuplein et al., 1969), nicotinic acid
8-8
-------
derivatives in situ (data from Stoughton et al., 1960) and developed several empirical correlations for the evaluation
of the permeability coefficient as a function of various physicochemical properties. Guy and Hadgraft (19895)
summarized the experimental data and structure-activity empirical correlations for the following chemical classes:
alkanoic acids (data from Liron and Cohen, 1984a,b), alkanols (data from Behl et al., 1980; Scheuplein and Blank,
1971), nicotinic acid esters (data from Houk and Guy, 1988; Stoughton et al., 1960), nonsteroidal anti-inflammatory
drugs (data from Yano et al., 1986), phenols (data from Houk and Guy, 1988; Roberts et aL, 1977), phenylboronic
acids (data from Clendenning and Stoughton, 1962), polynuclear aromatics (data from Roy et al., 1987), and steroids
(data from Idson and Behl, 1987; Scheuplein et al., 1969). Tsuruta (1975,1982) reported a good correlation for the
solubility of the compounds in water and the absorption rate for a series of aliphatic or aromatic hydrocarbons.
Table 8-3 summarized the classes of chemicals and the physicochemical properties which have been correlated to
permeability coefficients.
8-9
-------
TABLE 8-3. Regression Equation! Developed by Various Authors
Chemical class
Aliphatic alcohols:
Water (lie)
Methanol
Elhanol
n-Propanol
n-Bulanol
n-Pentanoi
n-Hexanol
n-HepUnol
n-Octanol
Phenols:
Resorcinol
p-Nitrophenol
m-Nitrophenol
Phenol
Methyl hydfoxybenzoate
m-Ciesol
o-Cresol
p-Cresol
P-Naphthol
o-Chlorophenol
p-Ethylphenol
3.4-Xytenol
/>-Bromophenol
p-Chlorophenol
Thymol
Chlorocresol
Chloroxylenol
2.4,6-Trichlorophenol
2.4-Dichlorophenol
Phenols and esters of nicotinic
acid:
reiorcinol
catechol
p-melhoxyphenol
phenol
p-cresoJ
p-bromophenol
p-iodophenol
4-chloro-m-cresol
n-butylphenot
fl-pentylphenol
methyl m'cotinate
ethyl nicotinate
n-butyl nicotinate
n-pentyl nicotinale
n-hexyl nicotinate
Phenvlboronic (P) acids:
m-carbamido-P
m-carboxy-P
p-carbonxy-P
m-amino-P
p-methoxy-P
P
p-chloro-P
P-methyl-P
Experimental
system
Human epideimii
in vitro
Full thickness
hairless mouse in
vitro; same
chemicals as
Scheuplein (1966)
Human epidermis
in vitro
Ispropyl myristate
membrane (1PM)
Tetradecane
membrane (TD)
Human skin
in vitro
Percutaneous absorption function
log K. (cm/hr) = 0.420 log K. -2 354
log K. (cm/hr) = 0.544 log KL -2.884
log K! (cm/hr) = 0.934 log £„ -2.891
K,: olive oil/water partition coefficient
K^,: ocunolAvater partition coefficient
K^.^: stratum comeum/water partition coefficient
logRpsO-SOlogK..^^
log K, « -0.36 Gog K..)5 +2 J9 Oog K,,) -5 3.
log 1C = -0.48 flog 1C,)3 +2 .32 Oog 1C,) -2.2
log 1C, = -0.40 Oog 1C,)7 +2 J5 flog 1C,) -4.0
log C = 0.573 Oog 1C>3.749
log C = 0.212 Oog 1C,)2 +1.133 flog IC,)-3.999
log C = 0.417 Oog K*.)-2.463
K^,: octanol/water partition coefficient
K,^: benzeneAvater partition coefficient
C: molar concentration to cause a standard biological
response (i.e., boron penetration into the dermis).
Reference
Data
Scheuplein
(1966)
Behl et al.
(1980)
Roberts et al.
(1977)
Houk & Guy
(1988)
Houk & Guy
(1988)
,„
Clendeiming
&. Stoughton
(1962)
Equation
Lien & Tong
(1973)
Guy &
Hadgraft
(1989b)
Guy 4
Hadgraft
(1989b)
Guy &
Hadgraft
(1989b)
Guy &
Hadgraft
(1989b)
Lien &. Tong
(1973)
8-10
-------
Chemical class
Nonelectrovhes:
. ethyl iodide
methanol
ethanol
' thiourea
glycerol
urea
glucose
Steroids:
Progesterone 0*o)
Pregnenolone (Pe)
Hydroxy-Pe
Hychoxy-Po
Cortex one
Testosterone
Conexolone
Corticosterone
Cortisone
Hydrocortisone
Aldosterone
Estrone
Estradiol
Estriol
Hydrocortisone and 21 -esters
(acetate through hepunoate)
Nonsterioidtl anli -inflammatory
drues (NSAIDSs):
Alclofenac
Aspirin
Bufexamac
Flufenamic acid
Flurbiprofen
Iboprofen
Indomelhacin
Naproxen
Salicylic acid
Methylsalicylate
EthyUallcylaie
n-propylsalicylate
n-butylsalicylate
Ethylene glycol monosalicylate
Salicylamide
Salicyluric acid
( Nicotonic Acid esters:
Nicotonic acid (NA)
NA-HQ
Methyl nicotinate
Ethyl nicotinate
Butyl nicotinate
Hexyl nicotinate
Octyl nicotinate
Tetrahydrofurfuryl
nicotinate
Experimental
sys
Rabbit whole skin
in vitro
Rabbit dermis
in vitro
Human epidermis
in vifro
Hairless mouse
skin in vitro
Human in vivo
Human skin in
situ
Percutaneous absorption function
log K. (cm/hr) = -1.006 tog MW -1J71
log 1C (cm/hr) = -1.836 tog MR, -0.982
log 1C (cm/hr) = 0.392 tog K_ -2.761
log 1C (cm/hr) = -0.060 (tog K.W)2+OJ09 Oog 1C,) -2.591
log K. (cm/hr) = 0.360 (tog K_) -0.964 log (MRd)-U99
log K, (cm/hr) = 0.385 (tog ICJ -0.856 log MW-1.51
log K. (cm/hr) * 0.100 (tog K,.) -0.970
log 1C (cm/hr) = -0.622 tog MR, -0.395
log K, (cm/hr) = -0.575 tog MW -0.098
1C = permeability constant
MW = molecular weight
MRd = molar refnctivitiet
log K. (cm/hr) = 0.818 tog 1C,.,. - 3.556
log 1C (cm/hr) = 1.262 tog K_,. • 5.21 1
log 1C (cm/hr) = 2.626 log 1C,,. - 7.537
log K. (cm/hr) = 0.891 log 1C. - 5.175
log 1C (cm/hr) = -0.207 log (K./.)2-!-4'4 log (K,M)-5.425
Kin/.: hexadecane/water partition coefficient
K^.,.: amyl caproate/water partition coefficient
K,^.: stratum comeum/water partition coefficient
K,M: ether/water partition coefficient
log K, = 0.56 log K.,. -3.39
log [% dose absorbed] = -0.23 Oog JC^,)2 + 1.14 Oog K,.) +
0.42
log (1/C) = 1.008 tog K^. «• 1.230 tog S + 6.604
C: threshold molar concentration to induce visible
erythema (skin reddening)
K.,.: ether/water partition coefficient
S: molar solubility (mole/liter H,O)
Reference
Data
Treheme
(1956)
Scheuplein et
al. (1969)
Idson & Behl
(1987)
Yano et al.
(1986)
Stoughton et
aL(1960)
Equation
Lien & Tong
(1973)
Lien &. Tong
(1973)
Guy it
Hadgraft
(1989b)
Guy &
Hadgraft
(1989b)
Lien A Tong
(1973)
8-11
-------
Chemical clau
Coiticoneriods:
Predn'uofone
9o>Fluorahydroccftitcne
Methylpicdniiolone
Hydroconiiane
Hydnooititane acetate
Prednisokme acute
Dexaincthasone
9a-Fluorohydrocortitane
acetate
Triamctnolone acetonide
Fluocinolone acetonide
Flunnditnoione acetonide
Miscellaneoui:
oarbitone
phenobaibitone
butobaibitone
amytobaibitone
nicotine
salicylic acid
iioquinoline
Polynuclear aromatics (PNAs):
^-ring:
acenaphthykne
acenaphthene
dibenzofuran
fluorcne
9,10-dihydropheiunthrene
1 -methyl -
fluocenenzothiophene
dibenzoihiophene
phenanthrene
anthmcene
cufaazok
2-methylanthrcene
9,10^1imethyUnthimcene
1 -Methylphenanthrene
9-Methylanthncene
3,6-DifflelhylphenanthreM
2-Ethylanlhrmcene
4- and 5 -ring:
fliwranthene
pyrene
23-benzofluoKnc
benz(a)anthracene
chiyiene
benzoOcXIuoranthene
benzo(c)pyiciie
Denzo(a)pyi6!ie
peiylene
1 J3,4-dibenzanthracene
1 J^^dibenzanthncene
Pare itraltht-CRatn Alkanolc
acMsCj-ji:
accSTacM
Jwtyricacid
DCttlMM>iC ftCkl
iKUMlcadd
hepUMicadd
octaMkacid
Experimental
lyitem
Human Skin
in vitro
IPM }
TO )
Rat ilcin in vitro
Porcine ikin in
vitro
Percutaneoui abKiiption function
log 1/C = 1553 log K,,. + 1.139 log S + 6.101
C: molar concentration to induce vatoconstriction
log K, = 0.66 log K,,. - 2.02
Ky.: tetndecane/water partition coefficient
log K, = 0.7 llog K« -0.03
(IPM: Itopropyl myrittate membrane;
TO: tetradecane membrane)
log (% applied do»e) = f (log Kw)
Kp invenely related to melting points.
Reference
Data
Katz&
Shaikh (196S)
Hadgraft &.
Ridout (1987.
1988)
Roy et al.
(1987)
Lironi
Cohen
(1984a,b)
Equation
Lien tt long
(1973)
Guy &
Hadgraft
(19896)
Guy &
Hadgrath
(1989.)
8-12
-------
Chemical class
Aliphatic and »rom*tic
> hydrocarbons:
benzene
toluene
, stymie
ethylbenzene
* o-xylene
n-pentane
2-methylpenUne
n-hexane
ft -heptane
n -octane
Aliphatic hydrocarbons:
i^sJichlaroethane
larachloroethylene
l.U^-tetn-
chloreethane
trichloromelhane
1.1.2-trichloroethane
dichloromethane
1,1,1-trichJofethane
trichloroelhylene
Experimental
system
Rat skin in vitro
Mice in vitro
Percutaneous absorption function
log J = 1.41 log S -0.297
Percutaneous absorption rate (nM/min/cm* of skin) as a
function of solubility in water
S< 16 (mM at 25°C)
J~ 30.8 + (2.13)S rsO.87
S< 16
J"^ -52.8 + (6.59)8 r=1.00
Reference
Data
Tsunila
(1982i)
Tsuruta
(1975)
Equation
Tsuruta
(1982)
Tsuruta
(1975)
Since these equations were derived empirically using data from compounds in specific structural classes,
their use cannot be advocated for structurally dissimilar compounds. Therefore, the use of this approach is somewhat
limited. Guy and Hadgraft (1989b) explored the validity of using the regression equation derived empirically for
one class of compounds under a set of defined experimental conditions to predict the Kp values for a compound in
a structurally dissimilar class. Guy and Hadgraft (1989b) found that when the Kp values for phenols are compared
to the predictions obtained by the equation developed by Lien and Tong (1973) for alkanols, there is a tendency to
overpredict the absorption values for the phenols, with the most marked deviation in predictions of Kp values for
phenols with log K,w < 2.0.
8J
Theoretical Skin Permeation Models
An alternative to the previously discussed approaches for obtaining default Kp values is the use of theoretical
skin permeation models, such as those proposed by Michaels et al. (1975), Bemer and Cooper (1987), and Albery
and Hadgraft (1979). These models have the capability to predict the dermal permeation rate of a compound from
a saturated aqueous solution based on knowledge of the compound's lipid/aqueous phase partition coefficient, water
solubility, and molecular weight Another model has been developed based purely on Fickian diffusion mechanisms
of percutaneous absorption (Kasting et al., 1987). Although these predictive models have been previously evaluated
for their ability to predict Kp values for a series of drugs (Michaels et al, 1975; Osbome, 1986), the use of this
approach for predicting 1C, values for environmental pollutants has only recently been examined (Brown et al., 1990).
The theoretical basis of these models is presented below.
8-13
-------
8.2.1 Heterogeneous Structural Model
Michaels et al. (1975) developed a mathematical model to predict the rate of percutaneous absorption that
is consistent with the proposed two-phase structure of the stratum corneum (Figure 3-2). This model was the first
to provide a rationale for quantifying the rate of absorption of compounds through the skin based on knowledge of
the physicochemical properties of the compounds. In this model, molecules are assumed to migrate through the
stratum corneum by dissolution in the aqueous or lipid phase and by Fickian diffusion. If one makes further
assumptions about diffusional path length through the stratum corneum, the spatial geometry of this layer, the average
diffusivity of solutes in aqueous gels, and a lower-bound value for the thickness of the stratum corneum (15 um),
the Kp of solutes in an aqueous medium can be predicted from the following relationship:
Dp 0.16 + (P DJDJ
where P is a lipid phase/protein phase partition coefficient in the stratum corneum, and Dt/Dp is the ratio of the
diffusivities of the compound in the lipid and protein phases.
Although Michaels et al. (1975) used mineral oil/water partition coefficient values to approximate the
partitioning behavior of the compound in the lipid and protein phases of the stratum corneum, Brown et al. (1990)
found that use of readily available octanol/water partition coefficients does not significantly affect the predictive
capability of this model. Michaels et al. (1975) have shown that when Kp values for a series of drugs are plotted
against their mineral oil/water partition coefficient, a value of 2 x Iff3 for DL/DP gives the best theoretical fit to the
data. Alternatively, DL and Dp values can be estimated using equations 8-7 and 8-8, as shown below.
Therefore, with data concerning a compound's KDW and DL and Dp values, or the Di/Dp ratio as provided
by Michaels et al. (1975), this model can be easily used to estimate compound-specific Kp values.
8.2.2 Two Parallel Pathway Model
The two parallel pathway model proposed by Bemer and Cooper (1987) is conceptually similar to that
proposed by Michaels et al. (1975) in that it assumes that permeation of a molecule through the stratum corneum
can occur via a polar (aqueous) pathway or a nonpolar (lipid) pathway. Assuming that the flux through the polar
and nonpolar pathways are additive, Kp can be estimated in this model by the following relationship:
8-14
-------
)L (8-6)
where Ap and AL are the area fractions of the polar and lipophilic pathways, respectively, Dp and I\ are diffusion
coefficients for these pathways, respectively, and L is the thickness of the stratum corneum. Bemer and Cooper
(1987) have proposed values of 0.1 and 0.9 for Ap and AL respectively; however, these values have no physiological
basis. Similar to Michaels et al. (1975), Bemer and Cooper (1987) used mineral oil/water partition coefficients (P);
however, the use of readily available K,,w values does not affect the predictive capability of the model (Brown et al.,
1990). The following equations, based on Bueche's free-volume theory for rigid molecules, can be used to estimate
Dp and DL:
where Dp(0) and DL(O) are constants with approximate values of 3.8 x 10"5 and 1.7 x 10"5 cm2/hr, respectively,
Bp = BL = 0.016, and M is the molecular weight of the compound. It is important to recognize that Dp(o>, DL(o), Bp,
and BL are empirically derived constants obtained from studies with only a few compounds.
8.2.3 Three Parallel Pathway Model
Bemer and Cooper (1987) have proposed an expansion of the two parallel pathway model that accounts for
permeation through a heterogeneous oil-water multilaminate pathway, as well as the distinct polar and nonpolar
routes. Because this heterogeneous pathway model is difficult to solve, Bemer and Cooper (1987) developed
equations to predict upper- and lower-bound Kp values using this model. The equation to obtain upper-bound Kp
values is shown below:
(8.9)
0 -•*,
The values for all parameters in this model are the same as those used for the two parallel pathway model, except
AL is now equal to 0.5.
8-15
-------
8.2.4 Albery and Hadgraft Model
Another model that can account for penetration of a molecule via transcellular or intercellular pathways was
developed by Albery and Hadgraft (1979). The Kp for either pathway can be calculated by:
K, - JL + L (8-10)
' ak_i tr(PDA
where a and D are the area fraction and diffusion coefficient, respectively for each pathway, 7 is an empirically
determined parameter, k., is the interfacial transfer constant (3.6 cm/hr), L is the thickness of the stratum corneum,
and P is the water partition coefficient. Albery and Hadgraft (1979) assigned values of unity and 7 x lO* for the
area fraction of the transcellular and intracellular routes, respectively, values that are much more physiologically
based than those by Bemer and Cooper (1987) in their models. Values for transcellular and intercellular diffusion
coefficients are 1.9 x Iff6 cm2/hr and 9.7 x 10"4 cm2/hr, respectively, and y has values of 1 for the transcellular route
and 5 for the intercellular route. The total K,, for both routes can be determined by:
K, (total) - !__ (8-11)
' l/Kf (trans) «• MKf (inter)
8.2.5 Kasting, Smith & Cooper Model
This model is based on a purely Fickian diffusion mechanism for percutaneous absorption in which the
stratum corneum is treated as a homogeneous liquid membrane and provides the limiting barrier function (Kasting
et al., 1987). Pick's first law for a steady-state flux (JJ through a homogeneous membrane of thickness h is:
/ « (£) • K C - (£).C (8-12)
*tt *"?"' "•«• **» TT *
where ^ is the membrane-vehicle partition coefficient and Cm is the permeant concentration at the boundary of
the membrane. To compare the penetration of different permeants, it is easier to compare maximum penetration rates
from saturated solutions than to try to ensure that Cm is the same for each compound. Since thermodynamics ensures
that saturation can be obtained at the outer boundary of the membrane by bringing it into contact with a saturated
solvent, the maximum flux is:
8-16
-------
Jm''S (8-13)
"
where S is the membrane solubility of the permeant. By assuming that the stratum comeum behaves like a lipophilic
membrane, one can get good estimates for relative solubilities of lipophilic compounds in a model lipid or by
estimating them from ideal solution theory.
The diffusion coefficient is of the form:
D - D0 exp (-Pv) <8'14)
where v is the van der Waals volume of the permeant and D0 and 0 are properties of the skin. This model of the
diffusion coefficient assumes that the stratum comeum lipids form a polymeric membrane with defined thermal
transitions, and that the diffusion is governed by the size of these permeants.
Substituting the expression for D into the maximum flux equation and rearranging, the following function
is obtained:
V
-------
log Kp - -3.15 + log tfw - ( 6.95 x 10'3 ) • MW (*•'!«)
This equation therefore presents a theoretical model with empirical parameters.
Bronaugh (personal communication, 1991) used the same Flynn (1990) data set, and fitted a linear regression of log
Kp as a function of molecular weight and log Kow to yield the following equation:
tog Kf - -2.61 + 0.67 log *_ - ( 6.1 x 10'1) MW (*•»)
8-18
-------
9.0 RELATIVE CONTRIBUTION OF DERMAL EXPOSURE
TO TOTAL ABSORBED DOSE
A central issue for exposure/risk assessors is determining when dermal exposure is likely to be a major route
of concern compared with the other routes of exposure. This section will review the results of several studies
designed to address this question, and from the results of these studies, will identify the conditions that would enable
dermal uptake to become a significant route of exposure relative to other routes.
9.1 Contribution of Dermal Exposure to the Total Absorbed Dose of Neat Compounds
Engstrom et al. (1977) determined that immersion of both hands in liquid xylene for IS minutes would result
in the systemic absorption of 35 mg of this compound. Based on previous data, these investigators estimated that
inhalation of xylene at 100 ppm for the same period (IS minutes) would result in the absorption of about 40 mg of
xylene (60% retention of 10 L/min ventilation). Although it is unlikely that continuous dermal exposure to liquid
xylene would occur over a 15-minute period in the workplace, it is not unreasonable to expect that intermittent
topical application of this solvent, or exposure to aqueous solution of xylene, would occur over the same period.
Guest et al. (1984) compared pulmonary and percutaneous absorption of 2-propoxyethyl acetate (PEA) or
2-ethoxyethyl acetate (EEA) in beagle dogs. The dogs were exposed to SO ppm of either compound (299 mg/m3
of PEA and 270 mg/m3 of EEA) for 300 minutes with breath samples taken periodically both during and after
exposure. From these data, Guest et al. (1984) estimated a pulmonary absorption rate of 74% of inhaled PEA and
68% of inhaled EEA. Percutaneous absorption rates (fluxes) were measured both in vivo and in vitro in beagle dogs
by applying an excess amount of neat compound. In vivo absorption rates for PEA, calculated for exposure periods
of 30 and 60 minutes, were 1.2 mg/cm2-hr and 0.8 mg/cm2-hr, respectively. In vivo absorption rates for EEA,
calculated for exposure periods of 30 and 60 minutes, were 1.7 mg/cm2-hr and 0.8 mg/cm2-hr, respectively. In vitro
absorption rates, calculated under steady-state conditions, were 2.3 mg/cm2-hr for EEA and 1.5 mg/cm2-hr for PEA.
Guest et al. (1984) used the absorption data obtained in vivo in the dog to estimate absorbed dose in humans exposed
via the pulmonary route to SO pm of PEA or EEA for 1 hour and via the dermal route to the neat compounds in
contact with both hands for 0.5 or 1 hour. Using these assumptions, the dose absorbed via the skin was estimated
to be three to four times greater than the dose absorbed via the respiratory tract, as shown in Table 9-1.
9-1
-------
Table 9-1. Estimation of PEA and EEA Uptake in Man
Estimated Uptake (me)
Route of Exposure
Inhalation (SO ppm)*
Skin (hands)*
Exposure Time (hr)
1.0
0.5 to 1.0
PEA
270
580 to 760
EEA
220
640 to 645
• 1 *>< m3 air inhalaH nor h/Mfr- atwuwntuin* 1ACL, /T>CA'\ nr fSlOi, fCK A\
b 175 cm tall human, surface area 1.85 m2; hands comprise 4% surface area, human dermal uptake comparable to
uptakes measured in vivo in beagles; dermal uptake of PEA of 1,2 mg/cm2-hour for exposures of OJ hour and
0.8 mg/cm2-hour for exposure of 1 hour, dermal uptake of EEA of 1.7 mg/cm2-hour for exposure of 0.5 hour and
0.8 mg/cm2-hour for exposure of 1 hour.
Source: Guest et al. (1984)
The studies conducted by Engstrom et al. (1977) and Guest et al. (1984) are two examples of how dermal
exposure serves as a significant route of exposure to neat organic compounds.
92 Contribution of Dermal Exposure to the Total Absorbed Dose of Compounds in Aqueous Media
Brown et al. (1984) demonstrated the importance of dermal exposure, relative to oral absorption for the
uptake of VOCs from drinking water. Using the fluxes from aqueous solutions reported by Dutkiewicz and Tyras
(1967,1968), Brown et al. (1984) used Pick's Law to calculate average Kp values for toluene, ethylbenzene. and
styrene. These Kp values were then used in the following equation1 to calculate absorbed dose via the dermal route:
Dermal Dose * Permeability constant (—__.)x
cm2xhr
duration of exposure (hr) x total body surface area (cm2) :x'" • '
amount rf body surface area exposed (%)x concentration (!OL) > body weight
** • ...
Oral doses of these compounds from the consumption of drinking water were obtained by:
1 This equation is taken directly from Brown et al. (1984). According to the context of Brown et al. (1984), the hours in the
units of the permeability constant are in the denominator, if the equation wen: written in the units used in this Interim Guidance
report, Kp would be in units of cm/hour, arrived at by multiplying the Brown et al. (1984) permeability constant in L/(cm2 x hr)
by a conversion factor of 1.000 cm3/L. Also, the body surface area exposed should be a decimal traction, rather than a percent
9-2
-------
O al Do « concentration (mg/L) x amount consumed (LJday)
body weight (kg)
Brown et al. (1984) then proposed three scenarios in which the estimated oral and dermal doses were
compared:
Scenario 1: 70 kg adult bathing 15 minutes, 80% immersed (skin absorption), 2 liters water consumed
per day (ingestion), 18,000 cm2 body surface area.
Scenario 2: 10.5 kg infant bathed 15 minutes, 75% immersed (skin absorption), 1 liter water
consumed per day (ingestion), 4,000 cm2 body surface area.
Scenario 3: 21.9 kg child swimming 1 hour, 90% immersed (skin absorption), 1 liter water consumed
per day (ingestion), 8,800 cm2 body surface area.
The estimated contribution of each route of exposure for each scenario is presented in Table 9-2. In these scenarios,
dermal exposure contributes significantly to the estimated dose of these compounds from drinking water, even when
the gastrointestinal absorption efficiency was assumed to be 100% (note that the respiratory exposure is ignored in
these computations and scenarios).
Table 9-2. Relative Contribution (%) of Dermal and Oral Exposure to Dose
Compound*
Toluene
Ethylbenzene
Styrene
Concentration
(mg/L)
0.005
0.10
0.5
0.005
0.10
0.5
0.005
0.10
0.5
Scenario lb
Dermal
67 (67)
63 (64)
59 (65)
75 (63)
63 (63)
68 (63)
67 (54)
50 (54)
59 (54)
Oral
33 (35)
37 (36)
41 (35)
25 (37)
37 (37)
32 (37)
33 (46)
50 (46)
41 (46)
Scenario 2
Dermal
44 (43)
46 (43)
45 (43)
44 (42)
46 (42)
45 (42)
29 (33)
35 (34)
29 (33)
Oral
56 (57)
54 (57)
55 (57)
56 (58)
54 (58)
55 (58)
71 (67)
65 (66)
71 (67)
Scenario 3
Dermal
91 (89)
89 (89)
89 (89)
91 (88)
89 (88)
89 (88)
83 (84)
84 (84)
83 (84)
Oral
9 (11)
11 (ID
11 (ID
9 (12)
11 (12)
11 (12)
17 (16)
16 (16)
17 (16)
* Permeability coefficients calculated by Brown et al. (1984) as follows:
toluene = 0.001 L/cm2-hr (1 cm/hr);
ethylbenzene = 0.00095 L/cm2 (0.95 cm/hr); and
styrene « 0.00065 L/cm2-hr (0.65 cm/hr).
b Numbers in parentheses are recalculated percentages, because Brown et al. (1984) had rounded their estimated
doses prior to calculating the relative contribution percentages.
Source: Brown et al. (1984)
9-3
-------
By examining Equations 9-1 and 9-2, one can evaluate the factors that lead to increased dermal absorption
relative to ingestion. Since the concentration of the chemical in water and the body weight of the exposed individual
appear in both Equations 9-1 and 9-2, the relative quantities absorbed via each route should remain the same when
either or both of these factors are changed. Increasing the duration of dermal exposure will increase the contribution
of dermal exposure relative to ingestion as components of the total dose via the dermal and ingestion routes of
exposure. As shown in Scenario 3 of Table 9-2, a child swimming for one hour in water containing 5 ppb to 0.5
ppm concentrations of three compounds will dermally absorb 6 to 10 times the amount of the compound relative to
the amount that would be absorbed orally after ingestion of one liter of drinking water with the same concentrations
of chemical contaminant (even assuming 100% oral uptake). Scenario 3 has a dermal contact time of 1 hour, while
Scenarios 1 and 2 represent only IS minutes dermal contact, and this increase in exposure duration contributes to
the increases in relative contribution of dermal exposure as shown in Table 9-2.
Table 9-2 is essentially verbatim from Brown et al. (1984) who also presented the estimated dose data in
mg/kg. The percentages for each chemical within each scenario should be exactly the same (within one percent or
so), but Brown et al. (1984) had computed the percentages from their calculated dermal and oral doses. Their
estimated dermal and oral doses had been rounded to two or only one significant figure; hence, the percentages
within scenarios for each of the chemicals are artificially different Table 9-2 shows the Brown et al. (1984) values,
as well as unrounded, recalculated values in parentheses.
In Table 9-2, note that the concentration of pollutant in the water varies over a 100-fold range, but has little
impact on the contribution of dermal vs. oral exposure relative to total dose. The larger the Kp, the larger the
contribution of dermal exposure to total absorbed dose; for example, toluene with a Kp of 0.001 L/cm2 (1.0 cm/hr)
showed slightly higher dermal exposures than ethylbenzene with a Kp of 0.00095 IVcm2-hr (0.9S cm/hr) which was
even larger still than styrene with a Kp of 0.00065 L/cm2-hr (0.6S cm/hr). Increasing the surface area of skin
exposed will increase the contribution of dermal exposure to total exposure; while increasing the ingestion rate of
water will increase the contribution of ingestion. The effect of changing the surface area or the ingestion rate is not
as clearly illustrated by the Brown et al. (1984) scenarios as the effects of changing exposure duration and Kp.
Surface area exposed and ingestion rate would be more important factors if they varied over wider ranges than those
of the Brown et al. (1984) scenario.
Several investigators (Cothern et al., 1984; Shehata, 1985; McKone, 1987) have shown that the respiratory
route can contribute significantly to the total body burden of VOCs that volatilize from drinking water. Using an
indoor air quality model developed by Wadden and Scheff (1983), Shehata (1985) estimated the amount of benzene,
toluene, and xylene introduced into indoor air from drinking water sources (i.e., baths, showers, toilets, dishwashers,
washing machines, and boiling water and other cooking). This information was used to estimate the dose of each
9-4
-------
compound absorbed via the respiratory tract by occupants exposed to these pollutants in a "typical" house. The
relative contribution of each route of exposure was then estimated using the assumptions listed in Table 9-3.
Table 9-3. Absorption Constants (Fraction Absorbed) for Various Routes of Exposure
Chemical
Benzene
Toluene
Xylene
Inhalation (%)
50.0
93.0
64.0
Oral(%)
100.0
100.0
100.0
Dermal (L/cm2 x
2.0 x 1(T5
9J x 10-*
3.8 x Iff4
hr)
Source: Shehata (1985)
The inhalation dose was calculated by:
Inhalation dose (mg) - Indoor air concentration (mg/m3) (9-3)
.x daily respiratory rate (m*/day) x fraction absorbed
The method used to calculate the oral dose was similar to that used by Brown et al. (1984), with an
additional term to account for ingestion of the pollutant through food consumption. It was stated in Shehata (1985)
that the dermal Kp values for benzene and toluene were taken from the Brown et al. (1984) study; however, benzene
was not one of the chemicals reported on by Brown et al. (1984). Also, the xylene Kp was "extrapolated" by Shehata
(1985), but it is unclear how this value was determined. The relative contribution of each route of exposure to the
total daily dose of each compound for a child in one exposure scenario (that is, one living in a rural area in the
summer) is presented in Table 9-4.
9-5
-------
Table 9-4. Effect of Drinking Water Concentration on Relative Exposure Via All
Routes to a Child's Total Body Burdens in Summer (Rural)*
Chemical
Benzene
Toluene
Xylene
Drinking Water
Concentration
(mg/L)
0.0
0.005
0.05
0.5
5.0
50.0
0.0
0.005 - 50.0**
0.0
0.005 - 50.0"
Percent Relative Contribution
Inhalation
0.0
1.0
5.0
15.0
20.0
20.0
0.0
22.0
0.0
21.0
Oral
100.0
99.0
95.0
84.0
79.0
78.0
0.0
45.0
0.0
62.0
Dermal
0.0
0.0
0.0
1.0
1.0
1.0
0.0
32.0
0.0
18.0
* The assessment includes all identified pathways of exposure (not just drinking water), and includes contributions
of the chemical from food. That is why the oral route contributes 100% to the exposure for benzene when there
is 0.0 mg/L benzene in the drinking water.
b The same percent relative contribution was reported at 0.005, 0.05,0.5, 5.0, and 50.0 mg/L.
Source: Shehata (1985)
The concentration of the compound in drinking water has no effect on the relative contribution of dermal
exposure. Of the three compounds presented in Table 9-4, dermal absorbed dose of toluene (the compound with the
highest estimated Kp) contributes more to total absorbed dose than dermal absorbed dose of benzene or xylene. The
findings from Shehata (1985) suggest dermal exposure is less important than those estimates from other studies, but
considering the uncertainties in the approach, these findings cannot be fully evaluated.
Hall et al, (1989) used the concept of pathway-exposure factors, expressed in units of (mg/kg-day)/(mg/L),
to compare the relative contribution of oral, respiratory, and dermal pathways to the percent of the lifetime equivalent
exposure to trans- 1,2-dichloroethyIene attributable to each of the three pathways (Table 9-5).
9-6
-------
Table 9-5. Lifetime Equivalent Exposure Factors (expressed as Percent of Total
Exposure) for Trans-1.2-Dichloroethylene in Tap Water
Fluid Ingestion Indoor Inhalation Dermal Absorption
Best Estimate 31 41 28
Upper Bound 21 65 14
Source: Hall et al. (1989)
Hall et al. (1989) developed their model such that the pathway exposure factor, F, is used to translate the
water-supply concentration, Cw in mg/L, into the total equivalent average lifetime exposure, e in mg/kg-day, as
follows:
e (mg/kg-day) - F (L/kg-day) x Cw (mg/L) (9-4)
For each of the three pathways (i.e., water ingestion, inhalation, dermal absorption), the equivalent lifetime
exposure within a population is composed of three age categories. The overall exposure factors, F, are calculated
as the weighted sum of the pathway-exposure factors, f (age group), for each of the three age categories:
F'^f (infant) + ^Lf (child) + *Lf (adult) (9-5)
where 2/70, 14/70, and 54/70 reflect the fraction of time (in years) that each population cohort spends in each age
category. For the water ingestion pathway, the intake values per unit body weight were 0.11 and 0.044 L/kg-day
for the infant and child, respectively. For adults, the fluid intakes were bracketed using the adult average fluid intake
of 1.4 L/day for a 60 kg female (0.023 L/kg-day) and 2.0 L/day for a 73 kg male (0.027 L/kg-day) for a "best-
estimate" average of 0.025 L/kg-day. For the "upper bound" average, the moderately active adult average intake was
taken to be 3.7 L/day for 65.5 kg average adult, 0.056 L/kg-day. Thus, the overall weighted sum of the water
ingestion pathway was 0.031 L/kg-day for the best estimate and 0.055 L/kg-day for the upper bound.
Indoor inhalation values are calculated according to the indoor air model for VOCs developed by McKone
(1987). The assumptions listed in Hall et al. (1989) include that the typical household has four occupants and uses
900 L/day of water contaminated with 1 mg/L of trans-1,2-dichloroethylene. The time-dependent concentration
profile in shower stall, bathroom, and household air, and the resulting effective lifetime doses were derived using
the McKone (1987) model and Hall et al. (1989) used an almost identically verbatim set of assumptions as appears
9-7
-------
in McKone (1987). The approach resulted in estimated inhalation pathway-exposure factors of 0.041 L/kg-day for
typical households (best estimate) and 0.17 L/kg-day for the upper-bound estimate.
For the dermal route. Hall et al. (1989) assumed dermal exposure occurred only during bathing and
showering, and made the following simple assumptions:
• Resistance to diffusive flux through layers other than the stratum corneum is negligible;
* Steady-state diffusive flux is proportional to the concentration difference between the skin surface
and internal body water,
• An adult spends from 10 to 20 minutes in a bath or shower each day;
• During bathing, roughly 80% of the skin is in contact with water, and during showers, roughly
40% of the skin is in contact with water; and
• Children and infants spend approximately 1 hour/week in bathing or swimming.
The exposure, e, from dermal absorption is given by the expression:
e *J,fftSA
where:
J, = Steady-state flux across the stratum corneum (mg/cm2-h);
T = Duration in the shower or bath (h);
f, * . Fraction of the skin surface in contact with water (unitless); and
SA = Surface area of the skin (cm2).
This exposure equation and the Pick's law equation are utilized to obtain lifetime equivalent exposure factors for
dermal absorption using a weighted sum approach comparable to that used for ingestion for the three age groups.
The best estimate approach, assuming a 10-minute bath for adults, yielded a pathway exposure factor of 0.028 L/kg-
day, and the higher (upper-bound) estimate assumed a 20-minute bath for adults and yielded a pathway exposure
factor of 0.037 L/kg-day.
In the very typical, "best estimate" scenario, each exposure pathway contributes about equal amounts of 1,2-
dichloroethylene to the total dose, with range of 28 to 41% of the total. In a reasonable approximation of more
9-8
-------
extensive exposure, the relative contribution of dermal exposure to the total dose of 1,2-dichloroethylene decreases,
but the dermal contribution remains similar to the dose received via ingestion of drinking water, with 14 and 21%
of the total, respectively.
Each of the previous calculations of absorbed dose were based on the assumption that a steady-state rate
of dermal uptake was occurring. However, Brown and Hattis (1989) pointed out that their previous use of this
assumption (Brown et al., 1984) may be invalid for the relatively brief exposure periods of 15 to 20 minutes bathing
as encountered in typical exposure scenarios. Brown and Hattis (1989) re-examined the data from their earlier study
(Brown et al., 1984) relative to this issue for ethylbenzene, toluene, and styrene, and reported estimated doses for
60 minute exposures, which were greater than the doses in Brown et al. (1984). Also included in Brown and Hattis
(1989) were estimates shown in Table 9-6 using a pharmacokinetic model to estimate the dermal absorption of VOCs
(ethylbenzene, tetrachloroethylene, and trichloroethylene) from aqueous solutions. Using this approach. Brown and
Hattis (1989) estimated a "minimum" and "maximum" adult daily dose of three VOCs in drinking water received
via dermal absorption, and compared these doses to those expected from oral and respiratory uptake. The oral dose
values were obtained by assuming an intake of 2 liters of drinking water/day. The inhalation dose was based on the
model of McKone (1987). The relative contribution of each route to daily dose is presented in Table 9-6. Only one
compound, ethylbenzene, had a dermal uptake as large as uptake by inhalation and ingestion, when conditions were
set to allow for maximal dermal absorption; for the maximum concentration exposure scenarios, the dermal percents
of total daily dose were less for both tetrachloroethylene and trichloroethylene, but were about one-third and about
one-half the contributions of respiratory and oral exposures, respectively, as percents of the total daily dose.
Table 9-6. Relative Contribution of Different Routes of Exposure to the
Absorbed Dose of VOCs in Drinking Water
Ethylbenzene
Tetrachloroethylene
Trichloroethylene
Oral
37.9
26.0
41.7
343
40.6
343
Percent of Total Dose
Respiratory
54.5
373
56.9
46.8
58.2
49.2
Dermal
7.6
36.7
1.5
18.9
1.2
16.5
Min1
Max
Min
Max
Min
Max
1 Hypothetical conditions that result in a "minimum" or "maximum" amount of the compound being absorbed (see
text).
Source: Adapted from Brown and Hattis (1989).
9-9
-------
The parameters changed by Brown and Hauls (1989) to reflect these "minimum" and "maximum" conditions
of dermal uptake include physiological and anatomical factors, such as percentage of skin fat, and percentage of
blood fat (Table 9-7). These two parameters may affect the skin/blood partitioning of the compounds. Curiously,
the skin thickness value was also increased by Brown and Hattis (1989) under "maximum" uptake conditions. The
parameter that was varied to the greatest extent by Brown and Hattis (1989) was an unitless empirical constant. This
constant (CONST) was incorporated into the dermal absorption equation to account for the various uncertainties in
estimating the other parameters of the dermal absorption rate equation, including the stratum comeum/water partition
coefficient (K,,,), diffusion coefficient (D), surface area of exposed skin, and thickness of the stratum comeum, as
well as to account for deviations from Pick's law. Therefore, it is probably accurate to represent the different doses
as occurring at "maximum" or "minimum" levels of uncertainty. Whatever its meaning, increasing the value of this
parameter (CONST) effectively increases the rate of dermal absorption. This comparison of the Brown and Hattis
(1989) estimates for route-specific absorbed doses is similar to that of Shehata (1985); for those compounds with
greater permeability coefficients, the dermal absorption rate is a major factor in determining the contribution of a
compound's dermal exposure to total absorbed dose relative to its exposure via other routes.
Jo et al. (1990a) performed studies to determine the chloroform doses to individuals during a 10-minute
shower based on analyses of exhaled breath. Breath samples were collected, commencing at exactly 5 minutes after
termination of showering, by using non-rebreathing two-way valves until a Tedlar sampling bag was filled, with
purified, humidified air being supplied through the valve from an inhalation bag. The pre-shower breath samples
from all subjects were less than the detection limit of 0.83 ug/m3. In one experiment, subjects showered normally
with municipal water containing chloroform at 53 to 36 ug/L. and the measured exhaled breath concentrations ranged
from 6.0 to 21.0 ug/m3 (Least square mean (LSM) =13 ug/m3). In a companion experiment, subjects showered in
the same shower stalls, but wore rubber clothes and boots to preclude dermal contact but to permit inhalation
exposure. The water concentrations of chloroform ranged from 10 to 37 ug/L, and the measured exhaled breath
concentrations ranged from 2.4 to 10 ug/m3 (LSM = 6.8 ug/m,). [Jo et al. (1990a) reported that the LSM of the
breath concentrations after normal showering (13 ug/m3) represents the sum of inhalation and dermal exposures to
chloroform in showering water, while the LSM of the breath concentrations on those subjects wearing rubber suits
(6.8 ug/m3) represents inhalation exposure only. The increased exhaled air concentration after exposure to a normal
shower (inhalation plus dermal) was statistically significant when compared to exposure from inhalation only, and
Jo et al. (1990a) concluded that the difference, 6.2 ug/m3. represents the dermal exposure. Thus, the dermal exposure
as a fraction of inhalation exposure, was 6.2 * 6.8 = 0.91176 (= 92%). [Jo et al. (1990a) reported the factor was
0.95, and Jo et al. (1990b) reported 0.93, but it would seem 0.92 is the correct factor.] Jo et al. (1990a,b) concluded
that the mean internal dose to chloroform during showering was approximately equal for the dermal and the
inhalation exposure routes.
9-10
-------
Jo et al. (1990b) compared the chloroform dose as estimated from showering with the water ingestion dose.
For one shower per day, the inhalation exposure was 0.24 pg/kg-day and the dermal exposure was 022 pg/kg-day,
for a total of 0.46 pg/kg-day. The chloroform ingestion doses were estimated based on daily water ingestion rates
of either 0.15 L or 2 L, and the "chloroform water concentration used was the value proportional to mean shower
air concentration used for the inhalation dose calculation" (although the specific concentration used was not reported).
Jo et al. (19905) reported that the chloroform dose estimated for a daily 0.15 L water ingestion was 0.05 pg/kg-day
and for a daily 2-L water ingestion was 0.7 pg/kg-day. Thus, the dermal and inhalation doses from one 10-minute
shower per day were similar (022 and 0.24 pg/kg-day, respectively), and the doses from these routes are greater than
the ingestion dose estimated from a 0.15-L water intake rate (O.OS pg/kg-day).
Table 9-7. Assumed Minimum and Maximum Conditions for Dermal Absorption as
Defined by Hattis and Brown (1989)
Parameter
Constant
Skin Thickness (cm)
Skin Fat (%)
Blood Fat (%)
Minimum
1
0.02
2.5
2.7
Maximum
20
0.1
2.0
0.9
The importance of dermal contact with water can be evaluated by comparing the possible exposure occurring
during normal water contact relative to that occurring as a result of ingestion. Water contact occurs most frequently
via showering, daily bathing and washing. Assuming the average size adult takes a 10 min bath or shower, the
absorbed dose in mg/day can be estimated as follows:
Dermal dose « C Kf ET A
C Kp (10 minAfoy) (1 hr/60 min) (20,000 cm2) (9-7)
C Kf (3.3 x 103 cm2-hr/day)
where,
C = Contaminant concentration in water (mg/cm3);
9-11
-------
Kp * Permeability coefficient (cm/hr);
ET = Exposure time (min/day);
A = Exposed skin area (cm2).
Assuming an average adult ingestion rate (IR) of 2L/day and absorption fraction (AF) of 1 (where AF is the fraction
of the administered dose which becomes absorbed), the ingestion dose in mg/day is estimated as:
Ingestion dose • C IR AF
« C 2L/day) (1,000 cm3/L)
- C (2,000 cm^/day)
Now the two routes can be compared as follows:
Dermal dose * Ingestion dose
C Kp (3.3 x 103 cm2-hrlday) - C (2,000 cm3/day) (9-9)
Kp * 0.6 cm/hr
This set of assumptions and arithmetic relationships would suggest that when Kp exceeds 0.6 cm/hr, the daily
exposure due to dermal contact while bathing would exceed exposure by direct ingestion. Of the 66 environmental
contaminants where Kp values have been measured (Tables A-2 and A-3), only 3 (ethyl benzene, styrene and toluene)
have a K,, greater than this value. As shown in Table 9-8, a Kp value of 0.6 cm/hr ranks near the top (the highest
is ethylbenzene with Kp = 1.37 cm/hr) of all measured Kp values and most are much less. This table provides the
number of compounds corresponding to Kp values classified by order of magnitude.
Table 9-8. Classification of Kp Values by Order of Magnitude Based
on the Values Listed in Tables A-2 and A-3
Kp (cm/hr) No. of Compounds
10-° 2
Iff1 6
Iff2 23
Iff3 18
Iff4 9
Iff5 6
Iff6 2
9-12
-------
This analysis suggests that dermal exposure while showering or bathing is not important to consider for most
contaminants, but may be similar to oral exposure for the small percentage of compounds which penetrate fastest.
However, this conclusion must be considered speculative at this point for several reasons. First, this list of chemicals
is clearly not representative of the dermal characteristics of inorganics, as only three are included. Also, the degree
to which it is representative of organic environmental contaminants is unknown. Secondly, multiple uncertainties
are introduced in the measurement and application of the Kp data. Kp values are typically derived from experiments
conducted over 8 to 96 hours and may not be representative of what happens during the first 10 minutes of contact.
Thirdly, most dermal experiments are conducted at ambient temperatures and the elevated temperatures during
showering and bathing would likely increase volatilization, reducing the amount available for dermal absorption.
Finally, this conclusion could be affected to some extent by the assumption for exposure duration. If time spent
showering or bathing differs from the assumed 10 minutes/day, thus would affect the exposure proportionally.
Similar conclusions can be drawn for swimming. Assuming that an individual swims an average of
70 minutes/week, the exposure duration and resulting exposure levels will be the same as in the bathing scenario
described above. The discussion in Chapter 2 suggests an upper estimate of 20 days/year at 3 hr/day. This totals
to 60 hours/year which is about the same as the cumulative total of daily 10 min baths. Thus, in situations where
the same water source is used for swimming, bathing and drinking, the absorbed dose from dermal water contact
may be twice as much as would occur from bathing alone, but would still be much less than direct ingestion for most
contaminants.
Some experimental support for these conclusions can be found in the recent work by Jo et al. (1990a.b).
Chloroform levels in breath after 10-minute showers were measured where the subjects first wore no clothing and
then wore protective rubber suits. The breath levels dropped by about half when wearing the rubber suits, leading
Jo et al. (1990a) to conclude that the chloroform dose from inhalation and dermal contact were about equal during
normal showering. Jo et al. (1990b) also evaluated direct consumption of the water and utilized ingestion rates of
0.1S and 2 L/day. The dose from ingestion of 2 L/day would be about 3 times greater than the dermal (or inhalation)
dose alone, but the other assumed ingestion rate, 0.15 L/day, yielded estimates for ingestion dose that were about
3 times less than the dermal dose alone. Although, Kp values for chloroform have not been measured, they have
been estimated to be in the range of 10"1 cm/hr. As suggested in the conclusion above, for compounds with Kp
values in this upper range, the dermal dose during bathing and ingestion dose should be similar.
9-13
-------
93 Contribution of Dermal Exposure to the Total Absorbed Dose of Chemical Vapors
In the studies outlined in Section 9.2, organic compounds that volatilized from drinking water were assumed
to be absorbed by the respiratory tract However, organic compounds in the vapor phase can also be absorbed across
the skin, as discussed in Section 7.1.
Several investigators have estimated the contribution that dermal exposure to vapors would have to the total
uptake of volatile compounds. Blank and McAuliffe (1985) estimated that an adult with a skin surface area of 2 m2
working in ambient air containing 10 ppm benzene and with 100 cm2 of skin in contact with gasoline containing 5%
benzene would absorb 7.5 pi of benzene from inhalation, 7.1 pi of benzene from dermal contact with liquid
gasoline, and 1.5 ul from body exposure to ambient air in one hour. Blank and McAuliffe (1985) estimated the
respiratory uptakes using a fraction absorbed for benzene in the respiratory tract of 0.46. This value was taken from
Rusch et al. (1977), who cited an earlier study by Teisinger et al. who reported that the average retention of inhaled
benzene vapor was 46.3% (as cited in Rusch et al., 1977). Dermal uptakes were estimated using a flux of
0.072 pl/cm2-hr, which was measured in vitro using human skin. Therefore, based on these assumptions, of the total
dose of benzene absorbed by this worker, approximately 9% will be contributed by dermal absorption of vapor phase
benzene.
Riihimaki and Pfaffli (1978) reported that sedentary volunteers exposed to 20 ppm xylene by inhalation for
3.5 hours absorbed a mean 3572 umole of compound. Over a similar exposure period, subjects wearing respiratory
protection, and exposed to 300 or 600 ppm xylene, absorbed 196.7 or 419.1 umoles of xylene, respectively, via the
dermal route. Although the exposure concentrations are different, the results of this study suggest that a worker
exposed to xylene at high concentrations, and wearing only respiratory protection, could absorb a dose of xylene
through the skin comparable to that absorbed via the respiratory tract at lower concentrations.
As discussed in Section 7.0, McDougal et al. (1987.1990) obtained K,, values for chemical vapors in the
rat Using the following relationship, these researchers estimated the skin uptake of chemical vapors that could occur
in rats if no respiratory protection was used:
SUn uptake ratio ~
Qr
9-14
-------
where:
SA
Alveolar ventilation rate; and
Surface area.
Using this equation, the contribution of skin uptake in a mixed respiratory/dermal exposure scenario for a 210 g rat
with a total SA of 267 cm2 and alveolar ventilation rate of 4.84 liters/hr was calculated for a series of VOCs. These
results are presented in Table 9-9.
Table 9-9. Contribution of Skin Uptake to the Total Absorbed Dose
of Chemical Vapors in the Rat
Chemical
Styrene
m-Xylene
Toluene
Perchloroethylene
Benzene
Halothane
Hexane
Isoflurane
Concentration
(ppm)
3,000
5,000
8,000
12,500
40,000
50,000
60,000
50,000
Flux
(mg/cm2/hr)
0.0211
0.0151
0.0206
0.0541
0.0191
0.0180
0.0065
0.0096
Permeability
Constant
(cm/hr)
1.753 ± 0.105
0.723 ± 0.003
0.721 ± 0.007
0.668 ± 0.080
0.152 ± 0.006
0.045 ± 0.005
0.031 ± 0.004
0.025 ± 0.004
Skin Uptake in a
Mixed Exposure
(%)
9.4
3.9
3.7
3.5
0.8
0.2
0.1
0.1
Source: McDougal et al. (1990)
Under these conditions, skin uptake of chemical vapors is expected to contribute about 9% of the total dose
of styrene, and 3 to 4% of the total dose of m-xylene, toluene, and perchloroethylene. The findings of McDougal
et al. (1990) that rat skin is generally 2 to 4 times more permeable to chemical vapors than human skin suggests that
the contribution of chemical vapors to skin uptake in humans would be 2 to 4 fold less.
The studies of Blank and McAuliffe (1985), Riihimaki and Pfaffli (1978), and McDougal et al. (1987,1990)
suggest that dermal exposure to chemical vapors would contribute less than 10% of the total body burden of a VOC.
In addition, Hursh et al. (1989) has described a scenario in which the dose of a volatile inorganic compound,
mercury, taken up by the skin is about 2.6% of the dose that would be retained by the lung under the same
9-15
-------
conditions. Therefore, dermal exposure probably accounts for relatively little of the uptake of vapors of volatile
organic or inorganic compounds.
Although the contribution of dermal exposure to the total absorbed dose of vapors may be minimal relative
to other routes of exposure, the dose upon which risk estimations of volatile compounds are based may be
underestimated by a factor of up to 10% for individuals with a substantial area of skin exposed to the contaminant.
The results of these studies also suggest that workers wearing respiratory protection without chemical protective
clothing may be at risk for absorbing a significant amount of the compound into the body, depending on the air
concentration of the compound, and the rate at which the vapors of ithe compound are absorbed through the skin.
9.4 Contribution of Dermal Exposure to the Total Absorbed Dose of Compounds in the Soil
A comparison of the relative importance of different exposure pathways has also been performed for
exposure to solvents in the soil. Using the approach developed by McKone (1989), Howd et al. (1990) have
estimated the dermal uptake of a number of VOCs from soil. From these values, and estimates of the ingestion and
inhalation of soil-bound VOCs, Howd et al. (1990) compared the relative uptake of several soil-adhered compounds
from each route of exposure in three different exposure scenarios. It is important to recognize, however, that this
exercise only evaluated the relative contribution of each route of exposure to the uptake of compounds that are still
adsorbed to soil particles. Volatilization of the compound from soil and ambient atmospheric concentrations of the
compound will increase the relative contribution of the inhalation route to the total absorbed dose from all sources.
Similarly, ingestion of drinking water and food will increase the dose from the oral route.
An analysis approach can be used to evaluate the relative importance of soil ingestion vs dermal contact
with soil. Using typical parameter values for children, the exposures can be estimated as follows:
Ingestion exposure * 200 mg/day
Dermal exposure ~ (contact rate) (total skin area) (fraction skin exposed)
- (1 mg/cm2-day) (10.000 cm2) (0.25) <9'12>
- 2,500 mg/day
9-16
-------
These two scenarios suggest that a child may eternally contact 12 times more soil than he or she ingests.
Accordingly, the absorbed dose would be greater by the dermal route if the gastrointestinal tract absorption fraction
is less than 12 times greater than the dermal absorption fraction. Assuming 50% absorption in the gut, if dermal
absorption exceeds about 4%, then dermal exposure would be greater than the ingestion exposure. As explained in
Chapter 6, very little data are available on dermal absorption fractions from soil. But, these limited data suggest that
absorption fractions greater than 4% are possible. Key assumptions in this analysis are that the contact rate is a valid
average for the entire exposed skin area and that the number of days of exposure by ingestion equal the number for
exposure by contact Also, the fraction of exposed skin used above (025) assumes the child is wearing shorts and
short-sleeved shirt In cold climates, this is probably not a valid assumption for much of the year.
9.5 Summary of Conditions That Enable Dermal Uptake to Become a Significant Route of Exposure
From the results of the studies presented in this section, we can begin to identify the compound- and skin-
specific factors that would result in dermal absorption becoming a major contributor to the total body burden of a
compound.
Any factor that increases the rate or extent of dermal absorption of a compound, and has no effect on the
compound's pulmonary or oral uptake, would be expected to increase the relative contribution of the dermal pathway
to total dose. Two conditions that would be expected to increase dermal absorption relative to oral or respiratory
uptake are damage to the stratum comeum and exposure of a large surface area of the exposed skin.
Prolonged exposure time might be expected to increase dermal and respiratory uptake equally. However,
this factor would also be expected to increase the contribution of either of these routes to total absorbed dose, relative
to the dose received by oral exposure. As shown by Brown et al. (1984) for a series of VOCs in drinking water,
exposure time is a major factor in determining contribution of dermal uptake relative to ingestion.
As discussed in Section 3.0, a number of other factors affect the rate of dermal absorption. For example,
increased temperature of the aqueous media may result in increased dermal absorption, but it may also increase the
volatilization rate of VOCs from tap water, thereby making these compounds more available for respiratory uptake.
Chemical vapors are not expected to be significantly adsorbed by the skin; however, when adsorption and
subsequent absorption does occur, it appears that the process is most affected by the lipid solubility of the compound.
For soil-adsorbed compounds, the volatility of the compound and its capacity to bind to soil particles may be the
major determinants of the relative contribution of dermal exposure to total absorbed dose.
9-17
-------
From these factors, some general guidelines can be proposed to indicate when the dose received from dermal
absorption might equal or exceed that from other routes of exposure. For organic compounds in aqueous media,
dermal exposure may predominate if:
• The compound is known to have a high Kp (e.g., ethylbenzene, styrene, toluene); and
• Conditions are present that would lead to increased dermal exposure (e.g., prolonged exposure
times, damaged stratum corneum, large areas of exposed skin).
Based on our current limited understanding of dermal absorption, the following interim conclusions can be
drawn concerning when dermal contact with water, soils and vapors; may be a concern:
1. For most contaminants, dermal contact with water during bathing or swimming will generally pose
less threat than direct consumption of the water. The fastest penetrating contaminants (Kp >
0.6 cm/hr) may pose hazards similar to direct consumption. Although these chemicals may not
increase the total risk substantially, they may significantly impact the cost of remedial action. This
would occur in a situation where the water was considered unsafe to drink and the remedial action
plan called for replacement of drinking water only, which could be accomplished via use of bottled
water. Since it now appears that these chemicals would pose an equal risk via contact during
bathing, it would be equally important to replace the water used for bathing and showering. For
practical purposes, this suggests that replacing the entire household water supply would be
necessary. It has not been well established how many of the environmental contaminants may have
Kp values in this upper range, but it appears to be a minority. Those which have been identified
include ethyl benzene, styrene, and toluene.
2. It appears that much more soil is dermally contacted than is ingested during normal exposure
scenarios. Unless the absorption fraction via ingestion exceeds the absorption fraction via dermal
contact by a factor of 12 or more, then dermal contact poses a greater threat than soil ingestion.
On this basis, dermal absorption appears to be more significant for those chemicals where
experimental evidence is available: dioxin, PCB and DDT. It is very difficult to speculate which
other soil contaminants may be a concern, because the predictive models have not been validated.
However, those compounds which are moderately volatile, and are neither very hydrophilic nor
lipophilic, would appear to permeate readily.
3. Current studies suggest that dermal exposure may be expected to contribute no more than 10% to
the total body burden of those compounds present in the vapor phase. However, those compounds
9-18
-------
adsorbed to soil that have moderate volatility and are neither very hydrophilic nor lipophilic (e.g.,
phenol) can be expected to permeate the skin readily. Dermal absorption may contribute
significantly to the total absorbed dose for these compounds.
9-19
-------
10.0 STEPWISE DERMAL EXPOSURE ASSESSMENT PROCESS
This section describes the steps involved in quantifying the dermal absorption of contaminants from
environmental media. A prior review of earlier sections of this document will assist the exposure/risk assessor with
the decision-making process outlined in this step-by-step approach to estimate dermally-absorbed dose. Therefore,
the exposure/risk assessor is encouraged to use this stepwise guidance in conjunction with the earlier supporting
chapters.
The steps to necessary to consider in determining the values of the exposure and absorption parameters are:
1. Identification of contaminated environmental media;
2. Characterization and quantification of chemical contaminants;
3. Identification and quantification of dermally exposed individuals;
4. Identification of activities resulting in dermal exposure, and quantification of exposure duration and
frequency; and
5. Selection/estimation and evaluation of dermal absorption values.
Default values are provided for each of the exposure factors included in the procedures presented in this
interim guidance document. These default values are to be used in situations where site specific information is not
available. No default value is provided for the medium concentration term in the dose equation since this is a purely
site specific value. The default values for event time/frequency, exposed surface area and absorption factor are based
on conservative assumptions, or measurements, judged to represent upper percentile values. The defaults for body
weight and lifetime are based on average values. Since the default value for many of the individual exposure factors
is an upper estimate, their combination will create an upper percentile scenario. It is difficult to accurately
characterize the severity of this scenario for several reasons. First, it depends on the concentration term, for which
defaults cannot be provided and may significantly impact the severity of the scenario. Secondly, the distributions
of most of the individual exposure factors have not been well defined. Finally, the Agency has not yet agreed on
how to define scenario severity. A number of efforts are currently underway for this purpose and it is hoped that
standard scenarios will be defined soon that can be incorporated into future versions of this document.
Separate guidance is provided for aqueous and soil media because different approaches are proposed to
estimate the dermally absorbed dose of compounds from these media (the estimation of the absorption of compounds
in water follows a K_-based approach, while procedure for estimating the dermal absorption of soil-bound compounds
utilizes absorption fraction data). Therefore, identification of contaminated media is an important first step in this
10-1
-------
process. Although proposed approaches to estimate the dermal absorption of chemical vapors were discussed in
Section 7.0 of this document, no specific guidance can be offered at this time to conduct such a dermal exposure
assessment Similarly, no specific guidance can be offered on estimating the dermal absorption of compounds in
sediment. However, future versions of this document will provide guidance in both of these areas.
10.1 Contact With Compounds in Aqueous Media
The steps required to identify appropriate values for the expiosure/absorption parameters, and to use these
values to estimate the dermally absorbed dose of a compound in aqueous solution, are presented in this section.
Step 1. Select Values for Exposure Parameters
Site-specific measurement or modeling is required to identify values for the concentration of the
contaminant(s) of interest in water. However, lacking site-specific data to the contrary, the default values presented
in Table 10-1 are recommended for the remaining parameters.
Table 10-1. Default Values for Water-Contact Exposure Parameters
Parameter
Skin Surface Area Available for Contact (cm2)'
Children (1-6 years)
Adults
Exposure Time (hours/day)
Exposure Frequency (days/year)
Exposure Duration (years)
Body Weight (kg)b
Children (1-6 years)
Adults
Averaging Time (years)
Bathing
10,000
20,000
10 min/event
365 events/yr
30
40
70
70
Swimming
10,000
20,000
2.6 hr/day
7days/yr
30
40
70
70
1 Refer to percentile estimates of body surface area presented in the Exposure Factors Handbook (EPA, 1989a),
or use the approximations provided in this table.
b Refer to the age-specific estimates of body weight presented in the Exposure Factors Handbook (EPA, 1989a),
or use the approximations provided in this table.
10-2
-------
The supporting rationale for each of these defaults along with discussions of their possible ranges of values
is presented in Section 2. The rationales for each default is briefly summarized here. The exposed skin area is based
on the assumption that the people are entirely immersed during bathing or swimming; the corresponding body areas
were presented in Section 2.0. The bathing frequency of 365/days per year is based on information that most people
bathe once per day. The bathing duration is based on the upper-end of the range (to be representative of baths as
well as showers and considering that some water residue remains on skin for a brief period after bathing) given in
the Exposure Factors Handbook (EPA, 1989a). The swimming frequency and duration is based on average values
from Risk Assessment Guidance for Superfund (EPA I989b). The overall exposure period of 30 years represents
the likely time that a person spends in one residence, and the body weight and lifetime values are population
averages (EPA, 1989a).
Step 2. Select a Permeability Coefficient for the Compound of Interest
This step outlines the procedures for identifying an appropriate experimentally derived Kp value for
compounds in aqueous solution, if such a value exists. In the absence of an experimentally derived Kp value for the
compound of interest, the procedures for estimating this value using readily obtainable physicochemical property data
are presented.
Step 2a. Experimental Values of Permeability Coefficient
A list of all measured Kp values that have been reported in the open, scientific literature for environmental
pollutants was reviewed and recommendations made as to which compound-specific measurements were most
appropriate for human exposure assessment for use with this interim guidance document for aqueous-phase and neat
compounds, respectively. These recommended compound-specific Kp values, presented in Tables A-2 and A-3 of
Appendix A. These measured Kp values are assumed to be more reliable than modelled predictions. Thus, if the
compound of interest is listed in these tables, the corresponding Kp value should be used. The rationale for selecting
the specific Kp values presented in these two tables are described in Appendix A.
Step 2b. Consideration of Factors that Affect the Experimental Kp values
As discussed in Section 3.0 of this interim guidance document, a number of skin- and compound-specific
factors are known to affect the extent and rate of dermal absorption. The exposure/risk assessor should consider all
of the following factors in determining how the experimentally derived Kp values obtained from the database are
applicable to a particular exposure scenario. Although it is impossible to determine quantitatively how these factors
will affect the calculation of dermally absorbed dose, it is important for the exposure/risk assessor to determine if
10-3
-------
this is a conservative, reasonable, or unconservative estimated dose, biased on the conditions in the exposure scenario
and those under which the experimental data were obtained.
Once a Kp value is selected, the exposure/risk assessor should try to assess the exposure conditions against
the experimental conditions. For the experimental Kp listed in Tables A-2 and A-3, the criteria for their selection
is discussed in Appendix A.
2b.l Experimental Kp from Neat Compounds
Application of the neat compound to the skin or the compound in a nonaqueous vehicle may result in Kp
values that are greater or less than those that might be expected if the compound were in aqueous solution (see
Section 3.0). Blank and McAuliffe (1985) measured the flux of pure benzene and benzene in aqueous solutions
across human skin in vitro to be 2.1 and 0.22 ul/cm2-hr, respectively. Therefore, for volatile organic chemicals with
physicochemical properties similar to those of benzene, it could be expected that the Kp from aqueous solution is
about one order of magnitude lower than that in neat solution. Specific guidance cannot be offered for other
chemical classes on the use of Kp values for the neat compound in an environmental scenario where exposure to
contaminated water occurs. However, depending on the integrity of the stratum corneum after exposure to the neat
compound, these Kp values probably represent an approximation within an order of magnitude of the Kp that would
be expected if the compound were applied in an aqueous solution. The exposure/risk assessor is encouraged to
compare these values with Kp data obtained for structurally analogous compounds applied in aqueous solution to
human skin.
2b3. Species Differences
Numerous studies have shown experimental animal skin to be more permeable to environmental pollutants
and other compounds than human skin (see Section 3.0). Studies using rat, mouse, or rabbit skin in vivo or in vitro
tend to overpredict human Kp values by three- to five-fold; however, the monkey and pig have been shown to have
skin permeation rates comparable to those found for humans. For a first approximation, experimental Kp values
obtained from studies using mouse, rat or rabbit skin do provide a conservative estimated for human exposure.
2b J Regional Variation
Guy and Maibach (1984) have developed a technique to account for variations in dermal absorption which
depend on the body area where the compound is applied. Although a lack of data limits the full validation of their
10-4
-------
approach, it nevertheless represents a means to quantitatively assess the effect of regional differences in dermal
absorption on total absorbed dose. However, in the absence of a validated quantitative approach to account for
regional variation in dermal absorption, the exposure/risk assessor should determine the site of application in the
experimental study, and consider how absorption at the experimental site relates to the particular application site in
the environmental exposure scenario under study. For example, data obtained from application of the compound to
the forehead may overpredict the dermal absorption of a compound in situations such as wading, where only the legs
or feet are exposed.
2b.4 Evaporation and Occlusion
If conditions in the exposure scenario may lead to rapid evaporation of the compound from the skin surface,
the absorbed dose would be overpredicted if these conditions were not reproduced experimentally. Occlusion of the
skin site of application in experimental studies may increase the rate and extent of absorption. Therefore, the use
of data from experimental studies in which the skin site was occluded may result in an overestimate of absorbed
dose.
2b.5 Metabolism
Several priority pollutants induce a dermal toxic effect as parent compounds or as metabolites. These
compounds are considered to be reacted or metabolized at the skin, and absorption into the systemic circulation is
usually of secondary importance, unless a systemic toxic effect is also observed. In general, first-pass metabolism
of many compounds in the skin refers to the clearance process which results in a reduced amount of parent
compound entering the systemic circulation. When first-pass metabolism produces a systemically toxic metabolite,
this process should be accounted for in K,, values obtained from in vivo studies or in vitro studies using viable skin
preparations. However, the use of values from most in vitro studies will overestimate the extent of dermal absorption
of a parent compound in situations where extensive first-pass skin metabolism is occurring.
2b.6 Age of the Skin
Pre-term infants may represent a high-risk group for the dermal absorption of environmental pollutants as
a result of an immature stratum corneum; however, the stratum comeum of the full-term infant is intact. The ability
of the stratum corneum to serve as an effective barrier in older adults is unknown, but animal studies suggest that
dermal absorption may decrease slightly as a function of age. These factors should be considered, whenever
necessary, by the exposure/risk assessor.
10-5
-------
2b.7 Skin Condition
Numerous studies have shown that any condition that compromises the ability of the stratum corneum to
serve as a permeability barrier (i.e., skin disease or injury) will potentially result in increased dermal absorption in
affected individuals.
2b.8 Hydration
Increased skin hydration will occur in common environmental exposure scenarios (i.e., bathing, showering,
swimming). Since hydration of the skin tends to increase the absorption of topically applied compounds, the use
of Kp values obtained from studies in which the skin is not fully hydrated will result in an underestimate of dermally
absorbed dose in the scenarios mentioned above.
2b.9 Temperature
In vitro studies have shown that raising the donor solution temperature over 37°C, such as might occur in
a bathing or showering scenario, increases K^; lowering the donor temperature below 37°C, such as might occur
during swimming scenarios, reduces Kp. Therefore, the temperature of the donor solution in the experimental study,
if known, should be compared to that in the exposure scenario, to determine how Kp might be affected.
2b.lO Study Type
The techniques used to obtain dermal absorption rates can markedly affect the resultant Kp values. The
study design of both in vivo and in vitro studies should be examined, relative to the discussions in Section 3.0, for
their ability to provide estimates of Kp appropriate to the calculation of dermally absorbed dose in specific exposure
scenarios.
Step 2c. Estimation of Kp in the Absence of Experimentally Derived Values
A list of over 200 priority pollutants were compiled across the EPA Program Offices. For these priority
pollutants, estimates of Kp are provided in Table A-4 of Appendix A. From the structure-activity relationships
presented in Chapter 8 (see Table 8-2), the following models are selected to provide an estimate of the Kp values:
10-6
-------
For the alcohols and phenols, the specific chemical class linear correlations between log Kp and
log KDW are selected. These correlations were developed for these specific classes of chemicals,
and would work best for similar compounds in the same class. They are:
Alcohols: log 1C (cm/hr) = 0.54 x log K,w - 2.88
(Lien & Tong, 1973)
Phenols: log Kp (cm/hr) = -0.36 x (log ICJ2 + 239 x log IC,W - 5.2
(Guy & Hadgraft, 1989)
For all other compounds, Flynn's empirical correlation and Kasting et al. 's theoretical model are
selected to provide estimates of Kp. Flynn's (1990) correlation was developed based on
experimental observations of over 100 compounds with molecular weight ranging from 18 to 584,
with a log 1C,,, ranging from -2 to over 4. Kasting's (1897) model took into account the
dependence of the permeability coefficient on both molecular weight and 1C,,,. Both Guy's (1991)
equation and Bronaugh's (1991) equation were selected for this model since they represent the
most general interpretation of the Kasting's (1897) model.
Algorithms for Calculating Permeability Coefficients from OctanoI/Water Coefficients'
log Kow < 0.5
OJ 3.0
log IC,W > 3 .5
Low Molecular Weight
Compounds
logK,, = -3
logIC. = -3.5 -HoglC.,,
log Kp = -0.5
High Molecular Weight
Compounds
logKp = -5
log Kp = -5.5 + log IC,W
log Kp = -1.5
'Where Kp « Permeability Coefficient.
Source: Flynn (1990)
log Kf - -3.15 + log *„ - (6.95 x 10'3) xMW
(Guy, private communication, 1991)
(10-1)
10-7
-------
log Kf - -2.61 •+ 0.67 x log Km - (6.1 x 10'3) x MW (10-2)
(Bronaugh, private communication, 1991)
Step 2d. Model Validation
From Table A-4, it is clear that for metals, which do not have a K^ and for priority pollutants with very
high KQW Gog KO» > 4). the current Structure-Activity correlations, which are all based on a linear function of
molecular weight and K,,w, can not and do not provide a reasonable estimate. This was due to the fact that most
correlations were developed with the assumption that the principal barrier to percutaneous absorption is the stratum
comeum, and that the stratum comeum is composed of lipophilic and hydrophilic areas. A compound therefore
needs to have some solubility in both lipids and water for its Kp to be well predicted by these models. For very
hydrophilic or lipophilic molecules, this assumption breaks down. Flynn's correlations are derived from experimental
observations, and do provide some limit to the values of the Kp. For TCDD with a log K,,w larger than 46, this
model predicts a Kp on the order of 0.01 cm/hr, while the Kasting's model predicts a Kp on the order of 100 cm/hr.
This discrepancy is largely due to the fact that the Kasting's model predicts a linear dependence of log Kp on log
K,,, at all ranges of log Kow. Clearly, the models need to be modified for the extremities of the log K,,w ranges.
Work is underway to develop specific recommendations for the metals in inorganic, ionic and organic
metallic forms (by Guy et al. at UCSF). Similarly, the models are being examined and modified for the high Kow
range (by Guy, Bronaugh and Hoang). These recommendations and modifications will be presented at a Work Shop
sponsored by EPA in April 1991 in Washington, D.C. for the reviewers' comments.
Step 3. Integration of the Information to Determine Dermal Absorption
For each population segment and for each chemical, integrate the information identified above into the
dermal absorption formula to generate dermal absorption values, as shown below. Use the following formula from
the RAGS (EPA, 1989b) to calculate dermal exposure and uptake from water
CWxSAxlLETxEDxCF /1A,,
DermaBy Absorbed Dose (mg/fc- <10'3)
where:
CW « Chemical concentration in water (mg/L);
10-8
-------
SA = Skin surface area available fa- contact (cm2);
Kp = Chemical specific dermal permeability constant (cm/hr);
ET = Exposure time (hours/day);
EF « Exposure frequency (days/year);
ED = Exposure duration (years);
CF «= Volumetric conversion factor for water (1 L/1000 cm3);
BW = Body weight (kg); and
AT = Averaging time, a pathway specific period of exposure for non-carcinogenic effects (i.e.,
ET x EF x ED x 365 days/year), and a 70 year lifetime for carcinogenic effects (i.e., 70
years x 365 days/year),
10.2 Contact with Compounds in Soil
The steps required to identify appropriate values for the exposure/absorption parameters, and the use of these
values to estimate the dermally absorbed dose of an organic compound adsorbed to soil, are presented in this section.
Step 1. Select Values for Exposure Parameters
Site-specific measurement or modeling is required to identify values for the concentration of the
contaminant(s) in soil. However, lacking site specific data to the contrary, the default values presented in Table 10-4
are recommended for the remaining parameters.
10-9
-------
Table 10-2. Default Values for Soil
Parameter Default Value
Skin Surface Area Available for Contact* (cm2)
Children 2.500
Adults 5,000
Adherence Factor (mg/cm2-event) 1.0
Exposure Frequency (events/year) 365
Exposure Duration (years) 30
Body Weight6 (kg)
Children (6-12 years) 30
Adults 70
Lifetime (years) 70
* Refer to age-specific percentile estimates of body surface area presented in the Exposure Factors Handbook (EPA,
1989a), or use the approximations provided in this table.
b Refer to the age-specific percentile estimates of body weight presented in the Exposure Factors Handbook (EPA,
1989a), or use the approximations provided in this table.
The contact rate is a conservative central estimate of the measured adherence values presented in
Section 2.0. The exposed skin area is based on the assumption that the exposed people are wearing shorts and short
sleeved shirts; the corresponding body areas were presented in Section 2.0. The exposure frequency of 365 days
per year is based on the judgement that in a warm climate, a serious gardener or child who likes to play outdoors
could have daily soil contact. The overall exposure of 30 years represents an upper-estimate of time that a person
spends in one residence (EPA, 1989a). The bodyweight and lifetime values are population averages (EPA, 1989a).
See Section 2 for further details on how these values were derived.
Step 2. Select a Value for Percent of Applied Dose Absorbed
As discussed in Section 6.0, data are available for the percent of applied dose absorbed from soil for only
a few compounds. Thus, specific guidance can only be offered for estimating the dermal absorption of the
compounds listed in Table 6-1 (TCDD, TCB, BaP, and DDT). Section 6 provides a detailed analysis on the use of
existing data on this issue. Table 6-1 summarizes the studies of dermal absorption of chemicals from soil. Tables
6-3 through 6-6 provide recommended values for percent of applied dose absorbed for TCDD, TCB, BaP, and DDT
at soil loadings ranging from 0.5 to 20 to 40 mg/cm2 and concentrations of the contaminant ranging from 1 to 1,000
10-10
-------
ppm, depending on the chemical. Table 10-5 is a summary of recommended percents of applied dose absorbed for
the default soil loading of 1 mg/cm2. Section 6.3.1 contains information on how these values were selected.
Table 10-3. Dermal Absorption Fractions for Dilute Concentrations of Contaminants in 1 mg soil/cm2.
Chemical
Tetrachlordibenzo-p-dioxin (TCDD)
Tetrachlordibenzo-p-dioxin (TCDD)
33'-4,4'tetrachlorobiphenyl (TCB)
Benzo-fl-pyrene (BaP)
Dichlorodiphenyl-trichloroethane,l,l,l-
trichloro-2,2-bis[/>-chlorophenyl]ethane (DDT)
Concentration
(ppm)
1
1
1000
10
10
Exposure
Duration
(hr)
24
96
96
24
24
Percent
Absorbed
20*
62-S&
35-100b
lOO6
100°
* Based on data from Shu et al. (1988) and Poiger and Schlatter (1980) and assumptions in Section 6.3.1.1. See
Table 6-3 for more values.
b Estimated from Roy et al. (1990) study and assumptions in Sections 6.3.1.1 and 6.3.1.2. The low end of the
range represents estimated percent absorbed from high organic carbon content soil; the high end of the range
represents estimated percent absorbed from low organic carbon content soil. See Table 6-3 for more values for
TCDD and Table 6-4 for more values for TCB.
0 Estimated from Wester et al. (1990) data and assumptions in Section 6.3.1.3 and 6.3.1.4. See Table 6-5 for more
values for BaP and Table 6-6 for more values for DDT.
Since experimental data on dermal bioavailability are available for only a few soil contaminants, the
exposure/risk assessor is likely to be without information necessary to calculate the dermally absorbed dose of a soil-
adsorbed compound. In the absence of experimentally derived values, an assumption of 100% uptake of the
compound from soil can be made. However, this assumption will result in overly conservative estimates of dermally
absorbed doses for compounds that permeate the skin poorly or compounds that are tightly bound to soil particles.
Section 63.2 discusses five approaches to assigning values to chemicals that do not have data. These
methods include using values measured for structural analogues, using values of the percent of applied dose absorbed
of the neat compound, using Kp values for the chemical in water (an approach that is not recommended), theoretical
modeling, use of default values based on the chemical's KgW and dimensionless Henry's Law Constant, and
determination of the total amount of the chemical which is extractable from the soil as an upper bound on the amount
that could be absorbed. These approaches are discussed in detail in Section 6.3.2.
10-11
-------
Step 3. Calculate the Absorbed Dose of the Soil-Adsorbed Compound
The formula to use for calculating dermal uptake of a compound from soil (EPA, 1989b) is:
DemoJty Absorbed Dose (mglkg-day) - ^*MP*P«Qr (10-4)
where:
CS B Chemical concentration in soil (mg/kg);
SA B Skin surface area available for contact (crn2);
AF = Soil-to-skin adherence factor (mg/cm2 - day);
ABS = Absorption factor (unitless);
EF ft Exposure frequency (events/year);
ED B Exposure duration (years);
CF B Conversion factor (10"* kg/mg);
BW = Body weight (kg); and
AT B Averaging time, a pathway specific period of exposure for non-carcinogenic effects (i.e.,
EF x ED), and a 70-year lifetime for carcinogenic effects (i.e., 70 years x 365 days/year).
103 Use of Dermal Absorption Data in Risk Assessment
The procedures outlined in this document allow the exposure/risk assessor to derive an estimate of dermally
absorbed dose. This absorbed dose information can be used to quantify the risks from exposure to toxic compounds
when used in conjunction with toxicity data in the risk characterisation paradigm outlined in the RAGS (EPA,
1989b), and other documents. However, as stated in the RAGS (EPA, 1989b):
No RfDs [References Doses] or [Carcinogenic potency] slope factors are
available for the dermal route of exposure. In some cases, however,
noncarcinogenic or carcinogenic risks associated with dermal exposure can be
evaluated using an oral RfD or oral slope factor, respectively.
Since the use of a K,, or percent absorbed value in Equations 10.1 and 102 provides an estimate of absorbed dose,
the exposure/risk assessor should only use oral RfDs or slope factors that have been adjusted to express the toxicity
10-12
-------
expected from an absorbed dose, when conducting a risk assessment. Appendix A in the RAGS (EPA, 1989b) gives
a detailed description of how to convert oral RfDs and slope factors derived using exposure (administered) dose to
values based on absorbed dose.
The lack of dermal RfDs and slope factors requires that oral (or inhalation) toxicity values be used to
calculate risk from dermal exposure to toxic compounds. However, the exposure/risk assessor should be aware of
the limitations and uncertainties associated with this approach. The most obvious limitation is that the risk associated
with point-of-entry (skin) effects for locally acting toxic agents cannot be estimated from oral toxicity data.
Furthermore, unlike orally administered compounds, dermally applied chemicals would not be subjected to first-pass
hepatic metabolism before reaching the systemic circulation. Therefore, a toxic effect attributable to an active
metabolite might be more pronounced if the compound was administered orally. Conversely, the dermal absorption
of a toxic parent compound that undergoes little or no first-pass metabolism may result in a greater dose of the toxic
moiety entering the systemic circulation than if the compound was absorbed orally.
These are only two uncertainties associated with the use of oral toxicity data to characterize the human
health risks that may result from dermal exposure to toxic compounds. Several working groups in the Agency are
currently addressing these and other related issues regarding the route-to-route extrapolation of toxicity data.
10-13
-------
11.0 REFERENCES
Ackerman, C., and Flynn, G. 1987. Ether-water partitioning and permeability through nude mouse skin in vitro.
I. Urea, Thiourea, Glycerol and Glucose. Intl. J. Pharmacol. 36: 61-66.
Albery, WJ., and Hadgraft, J. 1979. Percutaneous absorption: Theoretical description. J. Pharm. Pharmacol. 31:
140-147.
Albery, WJ., Guy, R.H., and Hadgraft, J. 1983. Percutaneous absorption: Transport in the dermis. Intl. J.
Pharmacol. 15: 125-148.
Atherton, DJ., and Rook, A. 1986. The newborn. P. 229 in Rook, A., Wilkinson, D.S., Ebling, FJ., Champion,
R.H., and Burton, J.L., Eds. Textbook of Dermatology, 4th ed. Blackwell Scientific, New York.
Banks, Y.B., Brewster, D.W., and Birnbaum, L.S. 1989. Age-related changes in dermal absorption of TCDD and
2,3,4,7,8-pentachlorodibenzofuran(4PeCDF). Toxicologist 9: 118.
Baranowska-Dutkiewicz, B. 1981. Absorption of hexavalent chromium by skin in man. Arch. Toxicol. 47: 47-50.
Baranowska-Dutkiewicz, B. 1982. Skin absorption of aniline from aqueous solutions in man. Toxicol. Lett. 10:
367-372.
Barker, N., Hadgraft, J., and Rutter, N. 1987. Skin permeability in the newborn. J. Invest. Dermatol. 88(4): 409-
411.
Barry, B.W., Harrison, S.M., and Dugard, P.H. 1984. Vapor and liquid diffusion of model penetrants through
human skin: Correlation with thermodynamic activity. J. Pharm. Pharmacol. 37: 226-236.
Bartnik, F.G., Reddy, A.K., Klecak, G., Zimmerman, V., Hostynek. JJ., and Kunstler, K. 1987. Percutaneous
absorption, metabolism, and hemolytic activity of /i-butoxyethenol. Fund. Appl. Toxicol. 8: 59-70.
Beech, J.A. 1980. Estimated worst case trihalomethane body burden of a child using a swimming pool. Med.
Hypotheses 6: 303-307.
Behl, C.R., Flynn, G.L., Kurihara, T., Harper, N., Smith, W., Higuchi, W.I., Ho, N.F.H., and Pierson CJ-. 1980.
Hydration and percutaneous absorption: I. Influence of hydration on alkanol permeation through hairless mouse skin.
J. Invest. Derm. 75: 346-352.
Behl et al., 1983. TO BE ADDED LATER
Behl, C.R., Linn, E£., Flynn, G.L., Pierson, Cl., Higuchi, W.I., and Ho, N.F.H. 1983b. Permeation of skin and
eschar by antiseptics. I. Baseline studies with phenol. J. Pharm. Sci. 72(4): 391-3%.
Behl, C.R., El-Sayed, A.A., and Flynn, G.L. 1983a. Influence of hydration on n-alkanol permeation through rat
skin; Comparison with hairless and Swiss mice. J. Pharm. Sci. 72: 79-82.
Behl, CR., Flynn, GL., Kurihara, T., Smith, W.M., Bellantone, N.H., Gatmaitan, O., Higuchi, W.I., and Ho, N.F.H.
1984. Age and anatomical site influences on alkanol permeation of skin of the male hairless mouse. J. Soc. Cosmet.
Chem. 35: 237-252.
11-1
-------
Behl, CJR., Bellantone, N.H., and Flynn, GJ-. 198S. Influence of age on percutaneous absorption of dmg
substances. Pp. 183-212 in R.L. Bronaugh and H.I. Maibach, Eds., Percutaneous Absorption. Marcel Dekker, New
York.
Bentley, P., Schuassmann, H., Sims, P., and Oesch, F. 1976. Epoxides derived from various polycyclic
hydrocarbons as substrates of homogenous and microsome bound epoxide hydratase. Eur. J. Biochem. 69: 97-103.
Berliner, D., Pasqualini, J., and Gallegos, A. 1968. The formation of water soluble steroids by human skin. J.
Invest. Dermatol. 50: 220-224.
Bemer, B., and Cooper, E.R. 1987. Models of skin permeability. Pp. 41-56. In B. Bemer and A.S. Keygondus.
Eds., Transdermal Deli very of Drugs. Vol. H. CRC Press, Boca Raton, FL, VOL. II. 41-56.
Blank, I.H. 1964. Penetration of low molecular weight alcohols into skin. I. The effect of concentration of alcohol
and type of vehicle. J. Invest. Dermatol. 43: 415.
Blank and Scheuplein, 1969. TO BE ADDED LATER
Blank, I.H. 1985. The effect of hydration on the permeability of the skin. Pp97-106 in RJ-. Bronaugh and H.I.
Maibach, Eds., Percutaneous Absorption. Marcel Dekker, New York.
Blank, I.H., and McAuliffe, D J. 1985. Penetration of benzene through human skin. J. Invest. Dermatol. 85: 522-
526.
Blank, I.H., Scheuplein, RJ., and MacFarlane, DJ. 1967. Mechanism of percutaneous absorption. III. The effect
of temperature on the transport of non-electrolytes across the skin. J. Invest. Dermatol. 49(6): 582-589.
Bond, J., and Barry, B. 1988. Limitations of hairless mouse for in vitro permeation studies through human skin:
Hydration damage. J. Invest. Dermatol. 90: 486-489.
Boyd.E. 1935. The Growth of the Surface Area of the Human Body, University of Minnesota Press, Minneapolis.
Brewster et al. 1989. TO BE ADDED LATER.
Bronaugh, RJ-., Congdon, E.R., and Scheuplein, RJ. 1981. The effect of cosmetic vehicles on the penetration of
N-nitrosodiethanolamine through excised human skin. J. Invest Dermatol. 76(2): 94-96.
Bronaugh. RJL., Stewart, R P., Congdon, E.R.. and Giles, AJL 1982a. Methods for in vitro percutaneous absorption
studies. I. Comparison with in vivo results. Toxicol. Appl. Pharmacol. 62: 474-480.
Bronaugh, RJ-., Stewart, RJ7., and Congdon, EJ*. 1982b. Methods for in vitro percutaneous absorption studies.
II. Animal models for human skin. Toxicol. Appl. Pharmacol. 62:481-488.
Bronaugh, RJ-., and Maibach, H.I. 1983. In vitro percutaneous absorption. Ppll7-129 in F.N. Marzulli and H.I.
Maibach, Eds., Dermatotoxicology. Hemisphere, Washington, D.C.
Bronaugh, RJL., Stewart, RJF., and Congdon, EJ*. 1983. Differences in permeability of rat skin related to sex and
body site. J. Soc. Cosmet. Chem. 34: 127-135.
Bronaugh, RJL., and Stewart, R.F. 1984. Methods for in vitro percutaneous absorption studies. III. Hydrophobia
compounds. J. Pharm. Sci. 73: 1255.
11-2
-------
Bronaugh, RX. 1985. Determination of percutaneous absorption by in vitro techniques. Pp. 267-280 in R.L.
Bronaugh and H.I. Maibach, Eds., Percutaneous Absorption. Marcel Dekker, Inc., New York.
Bronaugh, R.L., and Maibach, H.I. 1985. Percutaneous penetration of nitroaromatic compounds. J. Invest. Derm.
84: 180-183.
Bronaugh, and Stewart 1985: TO BE ADDED LATER
Bronaugh, RX., Stewart, R.F., Wester, R.C., Bucks, D., Maibach, H.I., and Anderson, J. 1985. Comparison of
percutaneous absorption of fragrances by humans and monkeys. Food. Chem. Toxicol. 23(1): 111-114.
Bronaugh, RX., and Stewart, R.F. 1986. Methods for in vitro percutaneous absorption studies. VI. Preparation of
the barrier layer. Am. Pharm. Assoc. 75(5): 487-491.
Bronaugh, RX., Stewart, R.F., and Simon, M. 1986a. Methods for in vitro percutaneous absorption studies. VI.
Preparation of barrier layer. J. Pharm. Sci. 75: 487-491.
Bronaugh, RX., Stewart, R.F., and Simon, M. 1986b. Methods for in vitro percutaneous absorption studies. VII.
Use of excised human skin. J. Pharm. Sci. 75: 1094-1097.
Bronaugh, RX., Weingarten, D.P., and Lowe, NJ. 1986. Differential rates of percutaneous absorption through the
eczematous and normal skin of a monkey. J. Invest. Dermatol. 87(4): 451-453.
Bronaugh, RX., Stewart, R.F., and Storm, JE. 1989. Extent of cutaneous metabolism during percutaneous
absorption of xenobiotics. Toxicol. Appl. Pharmacol. 99: 534-543.
Bronaugh, RX., and Collier S.W. 1990. In vitro methods for measuring skin penetration. Cosmetics and Toiletries
105: 86-93.
Brown, H.S., Bishop, D.R., and Rowan, C.A. 1984. The role of skin absorption as a route of exposure for volatile
organic compounds (VOC's) in drinking water. Am. J. Public Health 74(5): 479-484.
Brown, H.S., and Hattis, D. 1989. The role of skin absorption as a route of exposure to volatile organic
compounds in household tap water A simulated kinetic approach. J. Am. Coll. Toxicol. 8: 839-851.
Brown, R.. Cofone, L., Diwan, I., Mittelman, A., Hoang, K., and Schaum, J. 1990. Ability of Theoretical Skin
Permeability Models to Predict Permeability Coefficient or Flux Values for Environmental Pollutants. Paper
presented at the Principles of Route-to-Route Extrapolation Meeting, March 1990, Hilton Head, S.C. Sponsor
Buck, 1988 and 1989. TO BE ADDED LATER
Bucks, D.A.W., Wester, R.C., and Maibach, H.I. 1985. Skin binding during percutaneous penetration. Pp. 43-56
in RX. Bronaugh and H.I. Maibach, Eds., Percutaneous Absorption. Marcel Dekker, New York.
Burmaster and Maxwell. 1990. TO BE ADDED LATER
Chellquist, E.M., and Reifenrath, W.G. 1988. Distribution and fate of diethyl malonate and diisopropyl
fluorophosphate on pig skin in vitro. J. Pharm. Sci. 77(10): 850-854.
Chowhan, Z.T., and Pritchard, R. 1978. Naproxin I. Comparison of rabbit, rat and human skin. J. Pharm. Sci. 67:
1272-1274.
11-3
-------
Clenddening, W.E. and Stoughton, R.R. 1962. Importance of the aqueous/lipid partition coefficient for percutaneous
absorption of weak electrolytes. J. Invest. Dermatol. 39: 47.
Collier, S.W.. Sheikh, N.M., Sakr, A.. Lichtin. J J... Stewart, RP., and Bronaugh, R.L. 1989a. Maintenance of skin
viability during in vitro percutaneous absorption/metabolism studies. Toxicol. Appl. Pharmacol. 99: 522-533.
Collier, S.W., Storm, J£., and Bronaugh, Rl. 1989b. Structure-related differences in the metabolic rate of aniline
during azo color metabolism in sencoar mouse skin. lexicologist 9: 164.
Cothem, C.R., Coniglio, W.A., and Marcus, WJ. 1984. Techniques for the Assessment of the Carcinogenic Risks
to the U.S. Population due to Exposure from Selected Volatile Organic Compounds from Drinking Water via
Ingestion, Inhalation and Dermal Routes. U.S. Environmental Protection Agency, Office of Drinking Water,
Washington, D.C. Available from National Technical Information Service, Springfield, Va., as PB 84-213941.
Cussler, E.L. 1976. Multicomponent Diffusion. Amsterdam: Elsevier. (as cited in Cussler 1984).
Cussler, E.L. 1984. Diffusion - Mass Transfer in Fluid Systems. Cambridge University Press.
Dal Pozzo, A., Donzelli, G., Liggeri, £., Rodriguez, L. 1991. Percutaneous absorption of nicotinic acid derivatives
in vitro. J. Pharm. Sci 80(1): 54-57.
DelTerzo, S., Behl, C.R., Nash, R.A., Bellantone, N.H.. and Malick, A.W. 1986. Evaluation of the nude rat as a
model: Effects of short-term freezing and alkyl chain length on the permeabilities of n-alkanols and water. J. Soc.
Cosmet. Chem. 37: 297-307.
DOI (U.S. Department of Interior). 1973. Outdoor Recreation: A Legacy for America. U.S. Department of Interior,
Washington, D.C.
Driver, J.H., Whitmyre, O.K., Konz, JJ., Roy, T.A., Yang, JJ., and Kreuger. AJ. 1990. Dermal exposure
assessments for toxicants in soil: Soil adherence to skin and dermal bioavailability. Pp. 597-618 in Total Exposure
Assessment Methodology. A New Horizon. Proceedings of the EPA/A and WMA Specialty Conference, November
1989, Las Vegas, Nevada.
Driver, J.H., Konz, JJ., and Whitmyre, G.K. 1989. Soil adherence to human skin. Bull. Environ. Contam. Toxicol.
43: 814-820.
DuBois, D., and DuBois, G.F. 1916. A formula to estimate the approximate surface area if height and weight be
known. Arch. Int. Med. 17: 863-871.
Dugard, P.H., Walker, M., Mawdsley. S J., and Scon, R.C. 1984. Absorption of some glycol ethers through human
skin in vitro. Environ. Health Perspect 57: 193-197.
Dugard, P.H. 1986. Absorption through the skin: Theory in vitro techniques and their applications. Food Chem.
Toxicol. 24(6/7): 749-753.
Dupuis, D., Rougier, A., Roguet, R., and Lotte, C. 1986. The measurement of the stratum comeum reservoir A
simple method to predict the influence of vehicles on in vivo percutaneous absorption. Br. J. Dermatol. 115: 233-
238.
Durrheim, H., Flynn, G.L., Higuchi, W.I., and Behl, CJR. 1980. Permeation of hairless mouse skin. I. Experimental
methods and comparison with human epidermal permeation by alkanols. J. Pharm. Sci. 69: 781-786.
11-4
-------
Dutkiewicz, T., and Piotrowski, J. 1961. Experimental investigations on the quantitative estimation of aniline
absorption in man. Pure Appl. Chem. 3: 319-323.
Dutkiewicz, T., and Tyras, H. 1967. A study of the skin absorption of ethylbenzene in man. Br. J. Med. 24: 330-
332.
Dutkiewicz, T., and Tyras, H. 1968. Skin absorption of toluene, styrene, and xylene by man. Br. J. Ind. Med. 25:
243-246.
Elias, P.M., Cooper, E.R., Korc, A., and Brown, B.E. 1981. Percutaneous transport in relation to stratum corneum
structure and lipid composition. J. Invest. Dermatol. 76: 297-301.
Elias, P.M., Feingold, K.R., Menon, G.K., Grayson, S., Williams, ML., and Grubauer, G. 1987. The stratum
comeum two-compartment model and its functional implications. Pp. 1-9 in B. Shroot and H. Schaefer, Eds.,
Pharmacology and the Skin. Vol. 1. Skin Pharmacokinetics. Karger, Basel.
Engstrom, K., Husman, K., and Riihimaki, V. 1977. Percutaneous absorption of m-xylene in man. Int. Arch.
Occup. Environ. Health 39: 181-189.
EPA. 1983. Methods for Assessing Exposure to Chemical Substances. Volume 3. Methods for Assessing Exposure
Criteria from Disposal of Chemical Substances. Report No. EPA 560/5-83-016. U.S. Environmental Protection
Agency .and Office of Toxic Substances, Washington, D.C.
EPA. 1985. Development of Statistical Distributions or Ranges of Standard Factors Used in Exposure Assessments.
Report No. EPA/600/8-85-010. U.S. Environmental Protection Agency, Office of Health and Environmental
Assessment, Washington, D.C. Available from National Technical Information Service, Springfield, Va., as PB 85-
242667/AS.
EPA. 1986. Methodology for the Assessment of Health Risks Associated with Multiple Pathway Exposure to
Municipal Waste Combustor Emissions. Staff paper submitted for review to the Science Advisory Board, October,
1986. U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards, Environmental
Assessment Office.
EPA. 1987. Selection Criteria for Mathematical Models Used in Exposure Assessments: Surface Water Models.
EPA/600/8-87/042. U.S. Environmental Protection Agency, Exposure Assessment Group, Office of Health and
Environmental Assessment.
EPA. 1988a. Estimating Exposure to 23.7,8-TCDD. Report No. EPA/600/6-88/OOSA. U.S. Environmental
Protection Agency, Office of Health and Environmental Assessment, Washington, D.C.
EPA. 1988b. Superfund Exposure Assessment Manual. Report No. EPA/540/1-88/001. U.S. Environmental
Protection Agency, Office of Remedial Responses, Washington, D.C.
EPA. 1988c. Selection Criteria for Mathematical Models Used in Exposure Assessments: Groundwater Models.
EPA/600/8-88/075. U.S. Environmental Protection Agency, Exposure Assessment Group, Office of Health and
Environmental Assessment.
EPA. 1989a. Exposure Factors Handbook. Report No. EPA/600/8-89/043. U.S. Environmental Protection Agency,
Office of Health and Environmental Assessment, Exposure Assessment Group, Washington, D.C.
11-5
-------
EPA. 1989b. Risk Assessment Guidance for Superfund. Human Health Evaluation Manual Part A. Interim Final.
U.S. Environmental Protection Agency, Office of Solid Waste and Emergency Response, Washington, D.C.
Eschenroeder, A., Jaeger, RJ., Ospital, JJ., and Doule. CJP. 1986. Health risk analysis of human exposures to soil
amended with sewage sludge contaminated with poiychlorinated dibenzodioxins and dibenzofurans. Vet. Hum.
Toxicol. 28(5): 435-442.
Faredin, I., Toth, I., Fazekas, A., Kokai, K., and Julesz, M. 1968. Conjugation in vitro of (4-14C)
dehydroepiandrosterone to (4-14C) dehydroepiandrosterone sulfate by normal human female skin slices. J.
Endocrinol. 41: 295-296.
Feldmann, RJ., and Maibach, H.I. 1970. Absorption of some organic compounds through the skin in man. J.
Invest. Dermatol. 54: 399-404.
Feldmann, RJ., and Maibach, H.I. 1974. Percutaneous penetration of some pesticides and herbicides in man.
Toxicol. Appl. Pharmacol. 28: 126-132.
Feldmann, RJ., and Maibach, H.I. 1967. Regional variation in percutaneous penetration of 14C cortisol in man.
J. Invest. Derm. 48(2): 181-183.
Feldmann, RJ., and Maibach, H.I. 1969. Percutaneous penetration of steroids in man. J. Invest. Dermatol., 52:
89-94.
Fenske, R.A., Leffingwell, J.T., and Spear, R.C. 1985. Evaluation of fluorescent tracer methodology for dermal
exposure assessment. Pp. 377-393 in R.C. Honeycutt, G. Zweig, and N. Ragsdale, Eds., Dermal Exposure Related
to Pesticide Use. ACS Symposium Series No. 273. American Chemical Society, Washington, D.C.
Fenske, R.A., Leffingwell, J.T.. and Spear, R.C. 1986. A video imaging technique for assessing dermal exposure.
II: Fluorescent tracer testing. Am. Ind. Hyg. Assoc. J. 47: 771-775.
Fisher, L.B. 1985. In vitro studies on the permeability of infant skin. Pp. 213-222 in R.L. Bronaugh and H.I.
Maibach, Eds., Percutaneous Absorption. Marcel Dekker. New York.
Flynn. 1972. TO BE ADDED LATER
Flynn, G.L., Durrheim, H., and Higuchi, WJ. 1980. Permeability of hairless mouse skin. II. Membrane sectioning
techniques and influence on alkanol permeabilities. J. Pharmaceut. Sci. 70(1): 52-56.
Flynn, GJ_. 1985. Mechanism of percutaneous absorption from physicochemical evidence. Pp. 17-42 in R.L.
Bronaugh and H.I. Maibach, Eds., Percutaneous Absorption. Marcel Dekker, New York.
Flynn, GJL, and Stewart, B. 1988. Percutaneous drug penetration: Choosing candidates for transdermal
development Drug Dev. Res. 13: 169-185.
Flynn, G.L. 1990. Physiochemical determinants of skin absorption. Pp. 93-127. T.R. Gerity and C J. Henry. Eds.,
Principles of Route-to-Route Extrapolation for Risk Assessment Elsevier Science Publishing Co., Inc.
Foreman, M.I. 1988. A critical review of percutaneous penetration studies. Pp. 81-88 in R.M. Marks, SP. Barton,
and C. Edwards, Eds., The Physical Nature of the Skin. MTP Press, Ltd., Lancaster.
11-6
-------
Fox, C., Selkirk, J., Price, R, Cray, R., Sanford, K., and Fox, M. 1975. Metabolism of benzo(a) pyrene by human
epithelial cells in vivo. Cancer Res. 35: 3551-3557.
Franklin, CA., Somers, D.A., and Chu, I. 1989. Use of percutaneous absorption data in risk assessment. J. Amer.
Coll. Toxicol. 8(5): 815-827.
Franz, TJ. 1975. Percutaneous absorption: On the relevance of in vitro data. J. Invest. Dermatol. 64(3): 190-195.
Franz, TJ. 1978. The finite dose technique as a valid in vitro model for the study of percutaneous absorption in
man. Curr. Probl. Dermatol. 7: 58-68.
Fredriksson, T. 196 la. Studies of the percutaneous absorption of parathion and paraoxon. V. Rate of absorption
of paraoxon. J. Invest. Dermatol. 8: 233-236.
Fredriksson, T. 1961b. Studies of the percutaneous absorption of parathion and paraoxon. m. Rate of absorption
of parathion and paraoxon. Acta Derm.-Venereol. 41: 353-362.
Fries, G.F., and Marrow, G.S. 1975. Retention and excretion of 2,3,7,8-tetrachlorodibenzo-p-dioxin by rats. J.
Agric. Fd. Chem. (23): 265.
Fujimoto, S., and Watanabe, T. 1969. Studies on the body surface area of Japanese. Acta Medica Nagasaki, Japan,
13: 1-13.
Garcia, B., Marty, J., and Wepierre. J. 1980. Etude des facteurs conditionnant 1'absorption percutanee des alcools
incorpores dans des melanges eau-monooleate de sorbetanne polyoxyethylene-myristate d-isopropyle. Int. J.
Pharmaceut. 4: 205-217.
Gargas, M.L., Burgess, RJ., Voisard, D.E., Cason, G.H., and Andersen, ME. 1989. Partition coefficients of low-
molecular-weight volatile chemicals in various liquids and tissues. Toxicol. Appl. Pharmcol. 98: 87-89.
Gehan, E.A., and George, S.L. 1970. Estimation of human body surface from height and weight. Cancer
Chemother. Reports, Part I, 54: 225-235.
Graham, MJ., Williams, F.M., Rettie, A.E., and Rawlins, M.D. 1987. Aldrin metabolism in the skin. Pp. 252-255
in B. Shroot and H. Schaefer, Eds., Pharmacology and the Skin. Vol. 1. Skin Pharmacokinetics. Karger, Basel.
Grasso, P., and Lansdown, A.B.G. 1972. Methods of measuring, and factors affecting, percutaneous absorption.
J. Soc. Cosmet Chem. 23: 481-521.
Grissom, RJ5., Brownie, C. and Guthrie, FJ£. 1987a. In vivo and in vitro dermal penetration of lipophilic and
hydrophilic pesticides in mice. Bull. Environ. Contam. Toxicol. 38: 917-924.
Grissom, R.E., Monteiro-Riviere, N.A., and Guthrie, F£. 1987b. A method of preparing mouse skin for assessing
in vitro dermal penetration of xenobiotics. Toxicol. Lett. 36(3): 251-258.
Guest, D., Hamilton, MJ-., Deisinger, PJ., and Divincenzo, G.D. 1984. Pulmonary and percutaneous absorption
of 2-propoxyethyl acetate and 2-ethoxyethyl acetate in beagle dogs. Environ. Health Perspect. 57: 177-183.
Guy, R.H., Hadgraft, J., and Maibach, HJ. 1982. A pharmacokinetic model for percutaneous absorption. Intl. J.
Pharmacol. 11: 119-129.
11-7
-------
Guy, R.H., and Hadgraft, J. 1984. Prediction of drug disposition kinetics in skin and plasma following topical
application. J. Pharm. Sci. 73: 883-887.
Guy, R.H., and Maibach, H.I. 1984. Correction factors for determining body exposure from forearm percutaneous
absorption data. J. App. Toxicol. 4(1): 26-28.
Guy, R.H., and Hadgraft, J. 1985a. Mathematical models of percutaneous absorption. Pp. 3-16 in R.L. Bronaugh
and H.I. Maibach, Eds., Percutaneous Absorption. Marcel Dekker, New York.
Guy, R.H., and Hadgraft, J. 1985b. Skin metabolism: Theoretical. Pp. 57-64 in R.L. Bronaugh and H.I. Maibach,
Eds., Percutaneous Absorption. Marcel Dekker, New York.
Guy, R.H., Hadgraft, J., and Maibach, H.I. 198Sa. Percutaneous absorption in man: A kinetic approach. Toxicol.
Appl. Pharmacol. 79: 123-139.
Guy, R.H., Tur, E., and Maibach, H.I. 1985b. Optical techniques for monitoring cutaneous microcirculation. Int.
J. Dermatol. 24: 88-94.
Guy, R.H. 1989. Developing Predictive Models of Percutaneous Absorption. I. Transport of Phenols. Submitted
to the US EPA under Cooperative Agreement CR-812474. University of California, San Francisco, School of
Pharmacy.
Guy, R.H.. and Hadgraft, J. 1989a. Mathematical models of percutaneous absorption. Pp 13-26 in R.L. Bronaugh
and H.I. Maibach, Eds., Percutaneous Absorption. Marcel Dekker, New York.
Guy, R.H., and Hadgraft, J. 1989b. Structure-activity correlations in percutaneous absorption. Pp95-109 in R.L.
Bronaugh and H.I. Maibach, Eds., Percutaneous Absorption. Marcel Dekker, New York.
Guy, R.H., and Hadgraft, J. 1990. TO BE ADDED LATER.
Hakanson, R.. and Moller, H. 1963. On metabolism of noradrenaline in the skin: Activity of catechol-0-
methyltransferase and monoamine oxidase. Acta Derm. Venerol. (Stockholm) 43: 552-555.
Hall, L.C., Bogen, K.T., McKone, IE., Mallon, B., and Hall, C.H. 1989. Health Risk Assessment of trans-\,2-
Dichloroethylene. Report No. UCRL-21062. Lawrence Livermore Laboratory.
Hanke, J., Dutkiewicz, T., and Piotrowski, J. 1%1. The absorption of benzene through human skin. Med. Pr. 12:
413-426. (in Polish; English summary)
Hansch, C, and Leo, A. 1979. Substituent Constants for Correlations in Chemistry and Biology. Wiley-
Interscience, New York.
Harger, 3RE. 1979. A Model for the Determination of an Action Level for Removal of Curine Contaminated Soil.
Memorandum to P.S. Cole, Executive Director, Toxic Substances Control Division. Lansing, Michigan. (As cited
in EPA 1988b)
Harper, K., and Calcutt, G. 1960. Conjugation of 3:4-benzpyrenols in mouse skin. Nature 186: 80-81.
Hawkins, G.S., and Reifenrath, W.G.. 1984. Development of an in vitro model for determining the fate of
chemicals applied to skin. Fund. Appl. Toxicol. 4: S133-S144.
11-8
-------
Hawkins, G.S., and Reifemath, W.G. 1986. Influence of skin source, penetration cell fluid, and partition coefficient
on in vitro skin penetration. J. Pharm. Sci. 75(4): 378-381.
Hawley, J.K. 1985. Assessment of health risk from exposure to contaminated soil. Risk Anal. 5: 289-302.
Hill, M.S. 1985. Patterns of time use. Pp. 133-166 in Juster F.T. and F.P. Stafford, Eds., Time. Goods, and Well-
Being. University of Michigan, Institute for Social Research, Survey Research Center, Ann Arbor, Mich.
Holland, J.M., Kao, J.Y., and Whitaker, MJ. 1984. A multisample apparatus for kinetic evaluation of skin
penetration in vitro. The influence of viability and metabolic status of the skin. Toxicol. Appl. Pharmacol. 72: 272-
280.
Houk, J., Guy, R.H. 1988. Membrane models for skin penetration studies. Chemical Reviews. 88(3): 455-471.
Howd, R.A., Schum, G.M., McKone, T.E., and Wong, J J. 1990. Risk Estimation for Solvents in Soil. Poster
presented at Society of Toxicology Meeting, Miami Beach, Florida.
Howes, D., and Black, J. 1983. Percutaneous absorption of 2-nitro-p-phenylenediamine. Int. J. Cosm. Sci. 5: 215-
226.
Huq, A.S., Ho, N.F.H., Huisari, N., Flynn, G.L., Jetzer, W.E., and Condie, L., Jr. 1986. Permeation of water
contaminative phenols through hairless mouse skin. Arch. Environ. Contain. Toxicol. 155: 557-566.
Hursh, J.B., Clarkson, T.W., Miles, E.F., and Goldsmith, L.A. 1989. Percutaneous absorption of mercury vapor
by man. Arch. Environ. Health 44(2): 120-127.
Idson, B. and Behl, CR. 1987. In transdermal delivery of drugs. Pp. 85. Vol. 3. A.F. Kydonieus and B. Berner,
Eds. CRC Press, Boca Raton, FL.
James, I.R., and Knuiman, M.W. 1987. An application of Bayes methodology to the analysis of diary records from
a water use study. J. Am. Stat. Assoc. 82(399): 705-711.
Jetzer, W.E., Huq, A.S., Ho, N.F., Flynn, G.L., Duraiswamy, N., and Condie, L., Jr. 1986. Permeation of mouse
skin and silicone rubber membranes by phenols: Relationship to in vitro partitioning. J. Pharm. Sci. 7511: 1098-
1103.
Jetzer, W.E., Hou, S.Y.E., Huq, A.S., Duraiswamy, N., Ho, N.H., and Flynn, G.L. 1988. Temperature dependency
of skin permeation of waterbome organic compounds. Pharm. Acta Helv. 63: 197-201.
Joe, W.K., Weisel, CD., and Lioy, PJ. 1990a. Chloroform exposure and the health risk associated with multiple
uses of chlorinated tap water. Risk Anal. 10(4): 581-585.
Joe, W.K., Weisel, C.P., and Lioy, PJ. 1990b. Routes of chloroform exposure and body burden from showering
with chlorinated tap water. Risk Anal. 10(4): 575-580.
Johanson, G., and Fernstrom, P. 1986. Percutaneous uptake rate of 2-butoxyethanol in guinea pig. Scand. J. Work
Environ. Health 12: 499-503.
Johanson, G., and Fernstrom, P. 1988. Influence of water on the percutaneous absorption of 2-butoxyethanol in
guinea pigs. Scand. J. Work Environ. Health 14: 95-100.
11-9
-------
Johanson, G., Boman, A., and Dynesius, B. 1988. Percutaneous absorption of 2-butoxyethanol in man. Scand. J.
Work Environ. Health 14: 101-109.
Kao, J., Hall, J., Shugart, LJR., and Holland, J.M. 1984. An in vitro approach to studying cutaneous metabolism
and disposition of topically applied xenobiotics. Toxicol. Appl. Pharmacol. 75: 289-298.
Kao, J., Patterson, F.K., and Hall, J. 1985. Skin penetration and metabolism of topically applied chemicals in six
mammalian species, including man: An in vitro study with benzo[a]pyrene and testosterone. Toxicol. Appl.
Pharmacol. 81: 502-516.
Kao, J., and Hall, J. 1987. Skin absorption and cutaneous first pass metabolism of topical steroids: In vitro studies
with mouse skin in organ culture. J. Pharm. Exp. Ther. 241: 482-487.
Kao, J., Hall, J., and Helman, G. 1988. In vitro percutaneous absorption in mouse skin: Influence of skin
appendages. Toxicol. Appl. Pharmacol. 94: 93-103.
Kasting, GJ3., Smith, R.L., and Cooper, E.R. 1987. Effect of lipid solubility and molecular size on percutaneous
absorption. Pharmacol. Skin. 1: 138-153.
Katz, M., and Shaikh, Z.I. 1965. Percutaneous corticosteroid absorption correlated to partition co-efficient. J.
Pharm. Sci., 54: 591.
Kawakubo, Y., Manabe, S., Yamazoe, Y., Nishikawa, T., and Kato, R. 1988. Properties of cutaneous acetyl-
transferase catalyzing N- and O-acetylation of carcinogenic arylamines and N-hydroxyarylamine. Biochem.
Pharmacol. 37(2): 265-270.
Kemppainen, B.W., Riley, R.T., Page, J.G., Hoerr, FJ., and Joyave, J. 1986. Evaluation of monkey skin as a model
for in vitro percutaneous penetration and metabolism of [3H]T-2 toxin in human skin. Fund. Appl. Toxicol. 7: 367-
375.
Kemppainen, B.W., Pace, J.G., and Riley, R.T. 1987. Comparison of in vivo and in vitro percutaneous absorption
of T-2' toxin in guinea pigs. Toxicon 25(11): 1153-1162.
Kimbrough, R J)., Falk, R, Stehr, P., and Fries, G. 1984. Health implications of 23,7,8-tetrachlorodibenzo dioxin
(TCDD) contamination of residential soil. J. Toxicol. Environ. Health 14:47-93.
Knaak, JJJ., Yee, K., Ackerman, C.R., Zweig, G., and Wilson, B.W. 1984a. Percutaneous absorption of triadimefon
in the adult and young male and female rat. Toxicol. Appl. Pharmacol. 72: 406-416.
Knaak, J.B., Yee, K., Ackerman, C.R., Zweig, G., Fry, D.M., and Wilson, B.W. 1984b. Percutaneous absorption
and dermal dose - Cholinesterase response studies with parathion and carbaryl in the rat Toxicol. Appl. Pharmacol.
76: 252-263.
Kohli, R., Archer, W.I., and LiWan P, A. 1987. Laser velocimetry for the non-invasive assessment of the
percutaneous absorption of nicotinates. Inc. J. Pharm. 36: 91-98.
Komatsu, R, and Suzuki, M. 1982. Studies on the regeneration of the skin barrier and the changes in 32P
incorporation into the epidermis after stopping. Br. J. Dermatol. 106: 551-560.
Lepow, ML., Bruckman, L., Gillette, M., Markowitz, S., Rubino, R., and Kapish, J. 1975. Investigations into
sources of lead in the environment of urban children. Environ. Res. 10: 415-426.
11-10
-------
Levin, A., Maibach, H.I., and Wester, R.C. 1984. Assessment of Dermal Absorption of Contaminants in Drinking
Water (Draft Final Report). Report submitted to U.S. Environmental Protection Agency, Office of Drinking Water
Lien, EJ., and Tong, GJ-. 1973. Physicochemical properties and percutaneous absorption of drugs. J. Soc. Cosmet.
Chem. 24: 371-384.
Liron, Z., and Cohen, S. 1984a. Percutaneous absorption of alkanoic acids. I. A study of operational conditions.
J. Pharm. Sci. 73(4): 534-537.
Liron, Z., and Cohen, S. 1984b. Percutaneous absorption of alkanoic acids. II. Application of regular solution
theory. J. Pharm. Sci. 73(4): 538-542.
Maddy, K.T., Wang. R.G., and Winter, C.K. 1983. Dermal Exposure Monitoring of Mixers. Loaders and
Applicators of Pesticides in California, pp. 115-1089. State of California, Worker Health and Safety Unit.
Lopp, A., Shevchuk, I., Kirso, U. 1986. Fluorescence methods for the measurements of penetration and metabolism
of carcinogens in mouse skin. Cancer Biochem. Biophys. 8:185-191.
Maibach, H.I., Feldmann, RJ., Milby, T.H. and Serat, W.F. 1971. Regional variation in percutaneous penetration
in man. Arch. Environ. Health 23: 208-211.
Maibach, H.I. 1989. Comments on presentations by speakers at symposium I. J. Am. Coll. Toxicol. 8(5): 861-862.
Marks, R.M., Barton, S.P., and Edwards, C., Eds. 1988. The Physical Nature of the Skin. MTP Press. Ltd..
Lancaster.
Marzulli, F.N. 1962. Barriers to skin penetration. J. Invest. Dermatol. 39: 397-393.
Marzulli, F.N., Braun, D.W.C., Maibach, H.I. 1969. Techniques for studying skin penetration. Tox. & Appl.
Pharm. Suppl. 3: 76-83
Marzulli, F.N., and Maibach, H.I. 1984. Permeability and reactivity of skin as related to age. J. Soc. Cosmet.
Chem. 35: 95-102
McDougal, J.R, Jepson, G.W., Clewell, HJ., Ill, and Andersen, M£. 1985. Dermal absorption of dihalomethane
vapors. Toxicol. Appl. Pharmacol. 79: 150-158.
McDougal, J.N., Jepson, G.W., Clewell, HJ., MacNaughton, M.G., and Andersen, M£. 1986. A physiological
pharmacokinetic model for dermal absorption of vapors in the rat Toxicol. Appl. Pharmacol. 85(2): 286-294.
McDougal, J.N., Jepson, G.W., Clewell, HJ., Ill, and Andersen, M£. 1987. Pharmacokinetics of organic vapor
absorption. Pharmacol. Skin. 1: 245-251.
McDougal, J.N., Jepson, G.W., Clewell, HJ., IE, Gargas, ML, and Andersen, M.E. 1990. Dermal absorption of
organic chemical vapors in rats and humans. Fund. Appl. Toxicol. 14: 299-308.
McKone, TJE. and Layton, D.W. 1986. Screening the potential risks of toxic substances using a multimedia
compartment model: Estimation of human exposure. Reg. Toxicol. Pharmacol. 6: 359-380.
McKone, T.E. 1987. Human exposure to volatile organic compounds in households tap water. The indoor inhalation
pathway. Environ. Sci. Technol. 21: 1194-1201.
11-11
-------
McKone, TJE. 1989. Dermal uptake of chemicals from a soil matrix. Risk anal. 10:407-419
McLaughlin, T. 19S4. Review of Dermal Absorption. U.S. Environmental Protection Agency, Office of Research
and Development, OHEA/EAG, Washington, D.C.
Mcreesh, 1965. TO BE ADDED LATER
Menczel, E. 198S. Skin binding percutaneous penetration. Pp43-56 in RJ-. Bronaugh and H.I. Maibach, Eds.,
Percutaneous Absorption. Marcel Dekker, New York.
Michaels, A.S., Chandrasekaran, S.K., and Shaw, J.E. 197S. Drug permeation through human skin: Theory and in
vitro experimental measurement. AICE J. 21(5): 985-996.
Nathan, D., Sakr, A., Litchtin, IX., and Bronaugh, R.L. 1990. In vitro skin absorption and metabolism of benzole
acid, p-aminobenzoic acid and benzocaine in the hairless guinea pig. Pharmaceut. Res. 7: 1147-1151.
NCHS. 1983. National Center for Health Statistics, National Health and Nutrition Examination Survey 1976 -1988
(cited in EPA, 1985).
Noonan, P.K., and Wester, R.C. 1989. Cutaneous metabolism of xenobiotics. Pp.53-75 in R.L. Bronaugh and H.I.
Maibach, Eds., Percutaneous Absorption. Marcel Dekker, New York.
Osbome. D.W. 1986. Computational methods for predicting skin permeability. Pharmaceut. Manuf. 3: 41-48.
Paustenbach, DP., Shu, H.P., and Murray, FJ. 1986. A critical examination of assumptions used in risk
assessments of dioxin contaminated soil. Reg. Toxicol. Pharmacol. 6: 284-307.
Piotrowski, J. 1957. Quantitative estimation of aniline absorption through the skin in man. J. Hyg. Epid. Microb.
Immunol. 23: 1.
Pohl, R., PhilpoU R., and Fouls. J. 1976. Cytochrome P-450 content and mixed function oxidase activity in
microsomes isolated from mouse skin. Drug Metab. Dispos. 4: 442-450.
Poiger, H., and Schlatter, C.H. 1980. Influence of solvents and adsorbents on dermal and intestinal absorption of
TCDD. Food Cosmet. Toxicol. 18: 477-481.
Poulsen, B J., and Flynn, G.L. 1985. In vitro methods used to study dermal delivery and percutaneous absorption.
Pp. 431-460 in RJ-. Bronaugh and H.I. Maibach. Eds., Percutaneous Absorption. Marcel Dekker, New York.
Que Hee, S.S., Peace, B., Clark, C.S., Boyle, J.R., Bomschein, K.L., and Hammond, P.B. 1989. Evolution of
efficient methods to sample lead sources, such as house dust and tend dust, in the homes of children. Environ. Res.
38: 77-95.
Raykar, P.V., Fung, M.C., and Anderson, B.D. 1988. The role of protein and lipid domains in the uptake of solutes
by human stratum corneum. Pharm. Res. 5(3): 140-150.
Reid, R.C., Sherwood, T.K., Prausnitz. J.M. 1977. Properties of gases and liquids. 3rd ed. McGraw-Hill.
New York.
Reifenrath, W.G., and Robinson, P.B. 1982. In vitro skin evaporation and penetration characteristics of mosquito
repellents. J. Pharm. Sci. 71: 1014-1018.
11-12
-------
Reifenrath, W.G., and Spencer, T.S. 1989. Evaporation and penetration from skin. Pp313-334 in R.L. Bronaugh
and H.I. Maibach, Eds., Percutaneous Absorption. Marcel Dekker, New York.
Riihimaki. V., and Pfaffli, P. 1978. Percutaneous absorption of solvent vapors in man. Scand. J. Work Environ.
Health 4: 73-85.
Riley, R.T., and Kemppainen, B.W. 1985. Receptor fluid penetrant interactions and the in vitro cutaneous
penetration of chemicals. Pp. 387-392 in R.L. Bronaugh and H.I. Maibach, Eds., Percutaneous Absorption. Marcel
Dekker, New York.
Ritchie, D.D., and Carola, R. 1983. Biology, 2nd ed. Reading, Mass.: Addison-Wesley Publishing Company, p.
469.
Riviere, JE,, Bowman, K.F., and Monteiro-Riviere, N.A. 1985. Development of In Vitro Perfused Porcine Skin
Raps for Study of Percutaneous Absorption of Xenobiotics. Report No. AD-A183-508. Annual report submitted
to U.S. Army Medical Research and Development Command, Fort Detrick, Maryland (released 1987).
Riviere, JE., Bowman, K.F., Monteiro-Riviere, N.A., Dix, Z.P., and Carver, M.P.. 1986. The isolated perfused
porcine skin flap. I. A novel in vitro model for percutaneous absorption and cutaneous toxicology studies. Fund.
Appl. Toxicol. 7(3): 444-453.
Riviere, JE. and Williams, PL. 1989. Quantitative prediction of in vivo isolated absorption profiles using in vitro
isolated perfused porcine skin. Toxicologist 9: 60.
Roberts et al., 1975. TO BE ADDED LATER
Roberts, M.S., Anderson, R.A., and Swarbrick, J. 1977. Permeability of human epidermis to phenolic compounds.
J. Pharmacol. Pharmac. 29: 677-683.
Robl, M.G., Collier, S.W., and Bronaugh, RJ-. 1989. Evaluation of hairless Hartley guinea pig skin preparations
used for in vitro studies. Toxicologist 9: 61.
Roels, H.A., Buchet, J.P., Lauwenys, R.R., Branx, P., Claeys-Thoreau, F., Lafontaine, A., and Verduyn, G. 1980.
Exposure to lead by oral and pulmonary routes of children living in the vicinity of a primary lead smelter. Environ.
Res. 22: 81-94.
Rougier, A., Dupuis, D., Lotte, C, Roguet, R., and Schaefer, H. 1983. In vivo correlation between stratum comeum
reservoir function and percutaneous absorption. J. Invest. Dermatol. 81: 275-278.
Rougier, A., Dupuis, D., Lotte, C., and Roguet, R. 1985. The measurement of the stratum comeum reservoir. A
predictive method for in vivo percutaneous absorption studies: Influence of application time. J. Invest. Dermatol.
84:66-68.
Rougier, A., Lotte, C., and Maibach, H.I. 1987. In vivo percutaneous penetration of some organic compounds
related to anatomic site in man: Predictive assessment by the stripping method. J. Pharm. Sci. 76(6): 451-454
Rougier, A., DuPuis, D., Lotte, C., and Maibach, H.I. 1989. Stripping method for measuring percutaneous
absorption in vivo. Pp. 415-434 in R.I. Bronaugh and H.I. Maibach, Eds., Percutaneous Absorption. Marcel Dekker,
New York.
11-13
-------
Roy, T.A., Yang, J J., Czerwinski, M.H. 1987. Evaluating the percutaneous absorption of polynuclear aromatics
using in vivo and in vitro techniques and structure activity relationship. Pp. 471. In vitro toxicology Approaches to
Validation. A.M. Golberg, Ed. Mary Ann Liebert, Inc., New York.
Roy, T.A., Yang, JJ., Krueger, AJ., and Mackerer, CR. 1990. Percutaneous absoiption of neat 2,3,7,8-tetra-
chlorodibenzo-p-dioxin (TCDD) and TCDD sotted on soils. Toxicologist 10(1): 308.
Roy, T.A., Yang, J J., Krueger, AJ., Driver, J.H. and Konz, J J. 1989. Dermal Absorption of Dioxins and PCBs
from Soil. Draft report prepared by Versar, Inc., for USEPA, Office of Toxic Substances, Exposure Evaluation
Division, Exposure Assessment Branch. EPA Contract No. 68-02-4254, task no. 240.
Rusch, G.M., Leong B.K., and Laskin, S. 1977. Benzene metabolism. J. Toxicol. Env. Health 2 (Suppl.): 23-36.
Sato, A., and Nakajima, T. 1978. Differences following skin or inhalation exposure in the absorption and excretion
kinetics of trichloroethylene and toluene. Br. J. Ind. Med. 35: 43-49.
Schaefer, H., Lambrey, B., Caron, D., Illel, B., and Renucci, F. 1987. Methods in skin pharmacokinetics:
Introduction. Pp. 50-56 in B. Shroot and H. Schaefer. Eds., Pharmacology and the Skin. Vol. 1. Skin
Pharmacokinetics. Karger, Basel.
Schaum. J. 1984. Risk Analysis of TCDD Contaminated Soil. Report No. EPA/600/8-84/031. U.S. Environmental
Protection Agency, Washington, D.C.
Scheuplein, RJ. 1965. Mechanism of percutaneous absorption. I. Routes of penetration and the influence of
solubility. J. Invest. Dermatol. 45: 334.
Scheuplein, RJ. 1966. Mechanism of percutaneous absorption. I Routes of penetration and the influence of
solubility. J. Invest. Dermatol. 45: 334.
Scheuplein, RJ., et al. 1969. Percutaneous absorption of steroids. J. Invest. Dermatol. 52: 63
Scheuplein, RJ., and Blank, I.H. 1971. Permeability of the skin. Physiol. Reviews 51(4): 702-747.
Scheuplein, RJ., and Blank, I.H. 1973. Mechanism of percutaneous absorption. IV. Penetration of nonelectrolytes
(alcohols) from aqueous solutions and from pure liquids. J. Invest Dermatol. 60: 286-326.
Scheuplein, RJ. 1975. Site variations in diffusion and permeability. Pp. 1731-1752 in A. Jarrett, Ed., The
Physiology and Pathophysiology of the Skin. Academic Press, New York, (cited in Flynn, 1985)
Scott, R.C. 1986. Chamber designs for in vitro percutaneous penetration studies. Principles and practices of in vitro
percutaneous penetration studies as related to bioavailability and bioequivalence. Am. Assoc. Pharm. Sci. First
National Meeting (cited in Elias et al., 1987).
Scott, R.C., and Dugard, P.H. 1986. A model for quantifying absorption through abnormal skin. J. Invest.
Dermatol. 86: 208-213.
Scon, R.C. 1987. Percutaneous absorption. Pp. 103-110 in B. Shroot and H. Schaefer. Eds., Pharmacology and
the Skin, Vol I. Skin Pharmacokinetics. Karger. Basel.
Scott, R.C., and Ramsey, J.D. 1987. Comparison of the in vivo and in vitro percutaneous absorption of a lipophilic
molecule (cypermethrin, a pyrethroid insecticide). J. Invest. Dermatol. 89: 142-146.
11-14
-------
Scott, R., Dugard, P., Ramsey, J., and Rhodes, C. 1987. In vitro absorption of some o-phthalate diesters through
human and rat skin. Environ. Health Perspect. 74: 223-227.
Scott, R.C., and Dugard, P.H. 1989. The properties of skin as a diffusion barrier and route for absorption. Pp. 93-
114 in M.W. Greaves and S. Shuster, Eds., Pharmacology of the Skin, Vol. II. Springer-Verlag, New York.
Sedman, R.M. 1989. The development of applied action levels for soil contact: A scenario for the exposure of
humans to soil in a residential setting. Environ. Health Perspect. 79: 291-313.
Shah, P.V., Sunler, MR., lannou, Y.M., Fisher, HI,., and Hall, LL. 1985. Dermal absorption and disposition of
13-diphenylguanidine in rats. J. Toxicol. Environ. Health 15: 623-633.
Shahetal. 1987. TO BE ADDED LATER.
Shackleford, J.M., Yielding, Kl. 1987. Application of the fiber-optic perfusion fluorometer to absorption and
exsorption studies in hairless mouse skin. Anatomical Record 219: 102-107.
Shehata, T.A. 1985. A multi-route exposure assessment of chemically contaminated drinking water. Toxicol. Ind.
Health 1(4): 277-298.
Shehata-Karam, H., Monteiro-Riviere, N.A., and Guthrie, F.E. 1988. In vitro penetration of pesticides through
human newborn foreskin. Toxicol. Lett. 40: 233-239.
Sherwood, T.K., Pigford, R.L., Wilke, C.R. 1975. Mass transfer. New York: McGraw-Hill (As cited in Cussler
1984).
Scheuplein et al. 1969: TO BE ADDED LATER.
Shu, H., Teitelbaum, P., Webb, A.S., Marple, L., Brunck, B., DeiRossi, D., Murray, FJ., and Paustenbach, D. 1988.
Bioavailability of soil-bound TCDD: Dermal bioavailability in the rat. Fund. Appl. Toxicol. 10: 335-343.
Skog, E. and Wahlberg, J.E. 1964. A comparative investigation of the percutaneous absorption of metal compounds
in the guinea pig by means of the radioactive isotopes: 51Cr, 58Co, 65Zn, HOmAg, 115mCd, 203Hg. J. Invest.
Dermatol. 43: 187-192.
Skowronski, G.A., Turkall, R.M., and Abdel-Rahman, M.S. 1988. Soil adsorption alters bioavailability of benzene
in dermally exposed male rats. Am. Ind. Hyg. Assoc. J. 49(10): 506-511.
Skowronski, G.A., Turkall, R.M., and Abdel-Rahman, M.S. 1989. Effects of soil on percutaneous absorption of
toluene in male rats. J. Toxicol. Environ. Health 26: 373-384.
Skowronski. G.A., Turkall, R.M., Kadry, A.R.M., and Abdel-Rahman, M.S. 1990. Effects of soil on the dermal
bioavailability of m-xylene in male rats. Environ. Res. 51: 182-193.
Solomon, A., and Lowe, NJ. 1979. Percutaneous absorption in experimental dermal disease. Br. J. Dermatol. 100:
717-722.
Southwell, D., Barry, B., and Woodford, R. 1984. Variations in permeability of human skin within and between
specimens. Int. J. PharmaceuL 18: 299-309.
11-15
-------
Storm, J.E., Collier, S.W., Stewart, R.F., and Bronaugh, RJL 1990. Metabolism of xenobiotics during percutaneous
penetration: Role of absorption rate and cutaneous enzyme activity. Fund. Appl. Toxicol. IS: 132-141.
Stoughton, R.B., and Clendenning, W.E., Lruse, D. 1960. Percutaneous absorption of nicotinic acid and derivatives.
J. Invest. Dermatol. 35: 337.
Stoughton, R.B. 1975. Animal models for in vivo percutaneous absorption. Pp. 121-132 in H.I. Maibach, Ed.,
Animal Models in Dermatology. Churchill Livingstone, Edinburgh.
Suiter, C, Wilhelm, K.P., Bucks, D.A.W., Maibach, H.I., and Guy, R.H. 1989. Partitioning of chemicals into
stratum comeum (SC). lexicologist 9: 166.
Tarshis, B. 1981. The "Average American" Book. New American Library, New York. 191 pp.
Tauber, U., and Rost, K.L. 1987. Esterase activity of the skin including species variations. Pp. 170-183 in B.
Shroot and H. Schaefer, Eds., Pharmacology and the Skin, Vol.1. Skin Pharmacokinetics. Karger, Basel.
Thohan, S., Barr, J., and Sipes, I.G. 1989. Metabolism of 7-ethoxycoumarin (7-EC) by adult F-344 rat skin.
Toxicologist 9: 240.
Timmer, S.G., Eccles, J., and Obrien, K., 1985. How children use time. Pp. 335-380 in Juster, F.T., and Stafford.
F.P., Eds. Time, Goods, and Well-Being. University of Michigan, Institute for Social Research, Survey Research
Center, Ann Arbor, Mich.
Tregear, R.T. 1966. Molecular movement: The permeability of skin. Pp. 1-52 in Physical Functions of Skin.
Academic, New York.
Treherne, J.E. 1956. The permeability of skin to some nonelectrolytes. J. Physiol. 133: 171-180.
TRI, 1990. Pharmacokinetics of Chemical Mixtures. Report submitted to U.S. Environmental Protection Agency,
Office of Research and Development, OHEA/EAG, July, 1990.
Tsuruta, H. 1975. Percutaneous absorption of organic solvents. 1. Comparative study of the in vivo percutaneous
absorption of chlorinated solvents in mice. Ind. Health 13: 227-236.
Tsuruta, H. 1975. Percutaneous absorption of trichloroethylene in mice. Ind. Health 16: 145-148.
Tsuruta, H. 1977. Percutaneous absorption of organic solvents. 2. A method for measuring the penetration rate of
chlorinated solvents through excised rat skin. Ind. Health 15: 131-139.
Tsuruta, H. 1978. Percutaneous absorption of trichloroethylene in mice. Ind. Health 16: 145-148.
Tsurata, H. 1982. Percutaneous absorption of organic solvents. III. On the penetration rates of hydrophobia solvents
through the excised rat skin. Ind. Health 20: 335-345.
Tsuruta, H. 1986. Percutaneous absorption of chemical substances and toxicity-especially about organic solvents.
Eisei Kagaku 32: 229-241.
Vanderslice, R.R., and Ohanian, E. V. 1989. Dermal Absorption of Drinking Water Contaminants. Poster presented
at the Society of Toxicology Meeting, Atlanta, Georgia.
11-16
-------
Vandeven, J., and Herrinton, L. 1989. Assessing the Importance and Calculation of Dermal Exposures at Hazardous
Waste Sites. Poster Presented at Society for Environmental Toxicology and Chemistry Meeting, Washington, D.C.
Wadden, P.A., and Scheff, P.A. 1983. Indoor Air Pollution: Characterization, Prediction, and Control. John Wiley
& Sons, New York.
Wahlberg, J.E. 1968. Percutaneous absorption from chromium ("Cr) solutions of different Ph. 1.4-12.8. An
experimental study in the guinea pig. Dermatologica 137: 17-25.
Wahlberg, J.E. 1971. Vehicle role of petrolatum. Absorption studies with metallic test compounds in guinea pigs.
Acta Derm-Venerol. 51: 129-134.
Walker, M., Dugard P.H., and Scott, R.C. 1983. In vitro percutaneous absorption studies: a comparison of human
and laboratory species. Hum. Toxicol. 2: 561.
Wester, R.C., and Maibach, H.I. 1976. Relationship of topical dose and percutaneous absorption in Rhesus monkey
and man. J. Invest. Derm. 67: 518-520.
Wester etal. 1983. TO BE ADDED LATER.
Wester, R.C., and Maibach, H.I. 1983. Cutaneous pharmacokinetics: 10 steps to percutaneous adsorption. Drug
Metab. Rev. 14(2): 169-205.
Wester, R.C., and Maibach, H.I. 1985a. Influence of hydration on percutaneous absorption. Pp231-244 in R.L.
Bronaugh and H.I. Maibach, Eds., Percutaneous Absorption. Marcel Dekker, New York.
Wester, R.C., and Maibach, H.I. 1985b. In vivo animal models for percutaneous absorption. Pp251-266 in R.L.
Bronaugh and H.I. Maibach, Eds., Percutaneous Absorption. Marcel Dekker, New York.
Wester, R.C., and Maibach, H.I. 1985c. Structure-activity correlations in percutaneous absorption. Ppl07-124 in
RL. Bronaugh and H.I. Maibach, Eds., Percutaneous Absorption. Marcel Dekker, New York.
Wester, R.C., Maibach, H.I., Surinchak, J., and Bucks, D.A.W. 1985. Predictability of in vitro diffusion systems.
Effect of skin types and ages on percutaneous absorption of triclocarban. Pp. 223-226 in R.L. Bronaugh and H.I.
Maibach, Eds., Percutaneous Absorption. Marcel Dekker, New York.
Wester, R.C., and Maibach, H.I. 1986. Dermatopharmacokinetics: A dead membrane or a complex multifunctional
viable process. Prog. Drug Metab. 9: 95-109.
Wester, R.C., Mobayen, M., and Maibach, H.I. 1987. In vivo and in vitro absorption and binding to powdered
stratum corneum as methods to evaluate skin absorption of environmental chemical contaminants from ground and
surface water. J. Toxicol. Environ. Health 21: 367-374.
Wester, R.C., and Maibach, H.I. 1989a. Skin contamination and absorption of chemicals from drinking water while
bathing or swimming. Pp. 191-196 in R.L. Bronaugh and H.I. Maibach, Eds., Percutaneous Absorption:
Mechanisms, Methodology, Drug Delivery, 2nd ed. Marcel Dekker, New York.
Wester, R.C., and Maibach, H.I. 1989b. Human skin binding and absorption of contaminants from ground and
surface water during swimming and bathing. J. Am. Coll. Toxicol. 8(5): 853-859.
11-17
-------
Wester, R.C., Maibach, HI, Bucks, D.A.W., Sedik, L., Melendres, J., Liao, C, and DiZio, S. 1990. Percutaneous
absorption of [14C]DDT and [14C]Benzo[a]pyrene from soil. Fund. Appl. Toxicol. 15: 510-516.
Wieczorek, H. 1985. Evaluation of low exposure to styrene. II. Dcnnal absorption of styrene vapors in humans
under experimental conditions. Int. Arch. Occup. Environ. Health 57: 71-75.
Williams, P.L., and Riviere, JJE. 1988. A Pharmacokinetic Model Describing Xenobiotic Percutaneous Absorption
in the Isolated Perfused Porcine Skin Flap. Paper presented at the 2nd Annual Symposium - Frontiers of
Pharmacokinetics and Pharmacodynamics, Little Rock, Arkansas. Sponsor
Yang, J J., Roy, T.A., and Mackerer, C.R. 1986a. Percutaneous absorption of benzo[a]pyrene in the rat: Comparison
of in vivo and in vitro results. Toxicol. Ind. Health 2(4): 409-415.
Yang, J J., Roy, T.A., and Mackerer, C.R. 1986b. Percutaneous absorption of anthracene in the rat: Comparison
of in vivo and in vitro results. Toxicol. Ind. Health 2(1): 79-84.
Yang, JJ., Roy, T.A., Krueger, AJ., Neil, W. and Mackerer, CR. 1989. In vitro and in vivo percutaneous
absorption of benzo[a]pyrene from petroleum crude-fortified soil in the rat. Bull. Environ. Contain. Toxicol. 43:207-
214..' - • ••
Yano, T., Nakagawa, A., Tsuji, M., Noda, K. 1986. Skin permeability of various non-steroidal anti-inflammatory
drugs in man. Life Sci. 39: 1043-1050.
Zatz, JX. 1983. Fundamentals of transdermal controlled drug administration: Physicochemica! considerations.
Drug. Dev. Ind. Pharmacy 9(4): 561-577.
Zendzian, RP. 1989. Skin penetration method suggested for Environmental Protection Agency requirements. J.
Am. Coll. Toxicol. 8(5): 829-835.
11-18
-------
APPENDIX A
RECOMMENDED COMPOUND-SPECIFIC KP VALUES
-------
APPENDIX A: RECOMMENDED COMPOUND-SPECIFIC K,, VALUES
The dermal permeability database contains Kp values derived from studies conducted under a wide range
of experimental conditions. In estimating a dermal absorbed dose, the exposure/risk assessor would ideally select
Kp values from experimental studies that mimic as closely as possible the conditions found in the environmental
exposure scenario. For example, if the exposure/risk assessor was interested in determining the dose of a compound
absorbed through the skin during a shower, the most appropriate Kp value to use would be one generated in a study
in which the compound was applied to the skin in an aqueous vehicle. Therefore, the exposure/risk assessor is
encouraged to access the dermal permeability database and examine not only the Kp value, but also the conditions
under which it was generated, to select the most appropriate Kp value for a particular exposure scenario.
The Agency recognizes that some users of this document may not have access to dBase IV, the software
necessary to run and format the dermal permeability database. Therefore, the Kp values in the database were
evaluated and specific values for each compound were selected as being the most appropriate to use in calculating
a dermal absorption dose for bathing/showering in contaminated water. "Standard" exposure conditions have been
used to develop this scenario, as explained below. The rationale used to select these compound-specific Kp values
is provided in Section 3 of this appendix.
A.1 Exposure Scenario: Bathing/Showering with Contaminated Water
Exposure/risk assessors are often required to estimate the dose of a compound absorbed dermally from
bathing or showering with contaminated water. For the purposes of this exercise, the Agency assumes that "standard"
exposure conditions involve an adult with undamaged skin bathing or showering for 10 minutes a day in water with
a temperature of approximately 110° F (43°C). The concentration of contaminants in the water is assumed to be
in the part per million (ppm) range. It is further assumed that essentially 100% of the body will come into contact
with the water during bathing or showering.
A3. Rationale for Selecting the Most Appropriate Kp Values for Each Scenario
A set of criteria was developed to evaluate the suitability of the individual Kp values reported in the database
for use in calculating the dermally absorbed dose of a contaminant received by the individuals in the standard
exposure scenario described above. Although these criteria have been developed specifically for this exercise, they
can be used, in conjunction with the evaluation factors outlined briefly in Section 7.0, and in detail in Section 3.0,
as general guidance for evaluating the appropriateness of the Kp value for a particular exposure scenario. These
criteria are described below, and are summarized in Table A-l.
A-l
-------
Table A-l. Selection Criteria for Recommended K Values
Parameter
Higher Priority
Lower Priority
Vehicle
Species
Method
Number of Animals
Gender/Sex
Age
Chemical concentration
Citation Reference
Number of Replicates
Skin Condition
Shaved
Occluded
Temperature
Duration of Exposure
PH
Aqueous solution*
Human*
In viw?b
More0
Femaleb
Middle range*
Lower concentrations*
Recent (<15 years)0
More0
Undamaged*
Skin is undamaged*
Yes*
Similar to scenario*
Longer durations'*
Tested close to
neutrality0
Non-aqueous solution; neat
Non-human
In vivo
Less
Male
Premature or aged
Higher concentrations
Older (>15 years)
Less
Damaged
Skin is damaged
No
Different from scenario
Shorter durations
Not tested close to
neutrality
* Corresponds more closely to exposure scenario.
b Results in more conservative estimate of dose.
0 More scientifically/statistically valid results.
Vehicle
The vehicle in which the compound of interest is dissolved can have a dramatic effect on the rate at which
the compound is absorbed (see Section 4.0). For the exposure scenario, Kp values from studies in which the
compound was applied to the skin in an aqueous solution were selected in preference to those values generated from
studies in which other vehicles were used.
A-2
-------
Species
Marked variability in dermal permeation rates exists in different species (see Sections 3.0 and 4.0).
Therefore, equal weight cannot be given to all studies. Whenever possible, Kp values from studies in which human
skin was used were given the greatest priority. Since monkey and pig skin have been shown to be good models for
human skin, studies using these species were given the next priority. All other studies received lower priority.
Method
In vivo studies more closely represent the exposure conditions in the scenario; however, in vitro studies
allow Kp values to be measured directly and were given higher priority for selection.
Number of Animals
A study which used a greater number of animals/treatment group was given a higher priority over one with
fewer animals, for reasons of statistical significance.
Gender/Sex
As shown in Section 3.0, skin from female animals has been shown to be generally more permeable to
applied compounds than skin from male animals. Therefore, to allow for a more conservative estimate of absorbed
dose to be made, Kp values generated from studies using female skin were selected over those using male skin, if
all other parameters were equal.
Age
A trend toward increased permeability of the skin has been shown in studies using premature and slightly
decreased permeability has been demonstrated in aged animals (see Section 3.0). Although Kp values for individuals
in these older age categories would be the most conservative, they would not be representative of the majority of
the population. Therefore, studies in which these age groups were used received less priority for selection than
dermal absorption studies of animals or humans in the middle age range, if all other parameters were equal.
A-3
-------
Chemical Concentration
Since Kp is defined as flux normalized for concentration, Kp values should remain constant over a range
of concentration values. However, situations occur in the database where different Kp values are reported at different
concentrations of compound applied to the skin. This effect may be present because high concentrations of organic
solvents can extract lipids from the stratum corneum, thereby altering the diffusional barrier properties of this layer.
Therefore, Kp values generated in studies in which relatively dilute solutions of the compound of interest were
applied to the skin were selected over those that used more concentrated solutions, in the absence of other factors
that affect the selection of appropriate permeability coefficients.
Citation Reference
Investigators have increasingly become aware of the factors; that affect dermal absorption rates and have
taken these factors into account in the design of more recent studies. Although the primary criteria to evaluate a
study are the method, vehicle, etc., more confidence has been placed in relatively recent data (<15 years) over old
data, because of the increased recognition of the need to account for factors that affect dermal absorption.
Number of Replicates
Studies that used more replicates/dose were selected over those that used fewer replicates/dose because of
the increased scientific validity and statistical significance of these results, assuming all other factors are equal.
Skin Condition
Diseased or injured skin is generally more permeable to chemical compounds than healthy, intact skin (see
Section 4.0). Kp values from studies with damaged skin were not identified as the most appropriate value for any
compound in the database.
Shaved
Shaving the skin of experimental animals may be necessary to allow in vivo diffusion chambers to adhere
to the skin. However, this process may also damage the skin and result in artificially high estimates of Kp. Any
indication that the skin was damaged after shaving or was not given a 24-hour period after shaving to recover was
given a lower priority than if the skin was undamaged.
A-4
-------
Occluded
Occlusion of the site of application on the skin with plastic wrap in vivo, or a closed diffusion chamber in
vitro, results in hydration of the stratum corneum and subsequent increased permeability of this layer relative to a
nonoccluded state (see Section 4.0). Since this increased degree of hydration after occlusion corresponds to the
degree of hydration that is most likely found during bathing or showering, studies which used occluded conditions
were selected over studies which used nonoccluded skin sites, if all other factors were equal.
Temperature
The temperature of the donor solution is known to have a dramatic effect on the resultant rate of absorption
of some compounds (see Section 4.0). As a result, attempts were made to select a Kp for each compound that was
obtained in a study in which the temperature of the donor solution approximated that of the aqueous environment
for each scenario (45°C-50°C for bathing/showering). This field is not included in the database, because many
studies do not report this information. However, when temperature is reported, these data can be found in the
"comments" field.
Duration of Exposure
The scenario assumes a 10-minute bathing or showering duration. If studies reported 15, 30, and/or 60
minutes duration (or longer), the longest duration was used, all other factors being equal, as the most conservative
estimate of dose.
PH
The pH of in vitro and in vivo testing should have been conducted near pH (neutrality), since this would
most closely simulate the actual conditions of the scenario. At high or low pH, there may be dissociation of the
parent compound.
AJ Compound-Specific Default Kp Values
The lists of the experimental K_ values selected as outlined in the above criteria are included in Table A-2
for studies done in aqueous vehicle and Table A-3 for studies done with neat chemicals. The rationale for the value
chosen in the tables are discussed in the following two tables.
A-5
-------
Table A-2. Recommended Default Permeability Coefficient Values for Compounds in Aqueous Media
Chemical
Aniline
Benzene
p-Bromophenol
23-Butanediol
Butanol
2-Butanone
2-Butoxyethanol
Carbon disulfide
Chlorocresol
4-Chloro-3-cresol
2-Chlorophenol
p-Chlorophenol
Chloroxylenol
Kj, Human Single
(cm/hr) Study
4.1 x Iff2
1.11 x Iff1
3.61 x Iff2
<5.0 xlff5
2.5 xlff3
5.0 xlff3
\2 xlff2
5.3 x Iff1
5.5 xlff2
1.19 x Iff1
3.31 x Iff2
3.63 xlff2
5.9 x Iff2
Reference
Baranowska-Dutkiewicz, 1982
Blank and McAuliffe, 1985
Roberts et al., 1977
Blank et al.. 1967
Scheuplein and Blank. 1973
Blank et al.. 1967
Johanson et al.. 1988a
Baranowska-Dutkiewicz, 1982
Roberts et al., 1977
Huq et al., 1986
Roberts et al., 1977
Roberts et al., 1977
Roberts et al., 1977
Chromium-hexavalent
(sodium chromate)
Cobalt (Cobaltous chloride)
m-Cresol
o-Cresol
p-Cresol
Decanol
2,4-Dichlorophenol
2,4-Dimethylphenol
2,4-Dinitrophenol
Ethanol
2-Ethoxyethanol
Ethylbenzene
Ethyl ether
p-Ethylphenol
Glucose
Glycerol
Heptanol
Hexanol
Mercuric chloride
Methanol
Methyl ethyl ketone
Methyl hydroxybenzoate
Methyl mercury-dicyandiamide
B-Naphthol
2-Nitrophenol
4-Nitrophenol
3-Nitrophenol
Nonanol
Octanol
Pentanol
Phenol
Potassium mercuric iodide
2.1 xlff3
5.5 xlff4
1.52 x Iff2
1.57 x Iff2
1.75 x Iff2
8.0 xlff2
6.0 x Iff2
1.10 x Iff1
<3.15 x
8.0
3.0
1.37
1.7 x
3.49 x
9.5 x
1.4 x
Iff3
x Iff4
x Iff4
xlff2
Iff2
Iff5
Iff4
3.76 x Iff2
2.77 x Iff2
1.55 x Iff3
1.6 x Iff3
5.0 x Iff3
9.12 x Iff3
3.34 x Iff3
2.79 x Iff2
1.01 x Iff1
5.58 x Iff3
5.64 x Iff3
6.0 x Iff2
6.1 x Iff2
6.0 x Iff3
8.22 x Iff3
1.05 x Iff2
Baranowska-Dutkiewicz, 1981
Wahlberg, 1971
Roberts et al., 1977
Roberts et al., 1977
Roberts et al., 1977
Scheuplein and Blank, 1973
Roberts et al., 1977
Huq et al., 1986
Huq et al., 1986
Scheuplein and Blank, 1973
Blank et al., 1967
Dutkiewicz and Tyras, 1967
Blank et al., 1967
Roberts et al., 1977
Ackermann and Flynn, 1987
Ackermann and Flynn, 1987
Blank et al., 1967
Bond and Barry, 1988
Wahlberg, 1971
Southwell et al., 1984
Blank et al., 1967
Roberts et al., 1977
Skog and Wahlberg, 1964
Roberts et al., 1977
Huq et al., 1986
Roberts et al., 1977
Roberts et al.. 1977
Scheuplein and Blank, 1973
Southwell et al.. 1984
Scheuplein and Blank. 1973
Roberts et al., 1984
Skog and Wahlberg, 1964
A-6
-------
Table A-2. Recommended Default Permeability Coefficient Values for Compounds in Aqueous Media (com.)
Chemical
(cm/hr)
Human Single
Study
Reference
Propanol
Resorcinol
Sodium chromate
Styrene
Thiourea
Thymol
Toluene
2,4,6-Trichlorophenol
Urea
Water
3,4-Xylenol
1.7 x Iff3
2.4 x Iff4
3.4 x Iff3
6.7 x Iff1
9.6 x Iff5
5.28 x Iff2
1.01
5.9 x Iff2
1.2 x Iff4
1.5 x Iff3
3.6 x Iff2
Blank et al., 1967
Roberts et al., 1977
Baranowska-Dutltiewicz, 1981
Dutkiewicz and Tyras, 1968
> Ackermann and Flynn, 1987
Roberts et al., 1977
Kutkiewicz and Tyras, 1968
Roberts et al., 1977
> Ackermann and Flynn, 1987
Bronaugh et al., 1986
Roberts et al., 1977
Table A-3. Recommended Default Permeability Coefficient Values for Neat Compounds
Chemical
2-(2-Butoxyethoxy)-ethanol
Dibutyl phthalate
Di-(2-ethylhexyl) phthalate
Diethyl phthalate
Dimethyl phthalate
2-(2-Ethoxyethoxy)-ethanol
2-Ethoxyethyl acetate
2-Methoxyethanol
2-(2-methoxyethoxy)-ethanol
1 -Methoxypropan-2-ol
Parathion
Kp Human Single Reference
(cm/hr) Study
3.57 x Iff5
2.3 xlff6
5.7 x Iff6
1.14 x ia5
3.32 x Iff5
1.32 x Iff4
8.07 x Iff4
2.89 x Iff3
2.06 x Iff4
1.25 x Iff3
1.0 x Iff2
Dugard et al., 1984
Scott et al., 1987
Scott et al., 1987
Scott et al., 1987
Scott et al., 1987
Dugard et al., 1984
Dugard et al., 1984
Dugard et al., 1984
Dugard et al., 1984
Dugard et al., 1984
Knaak et al., 1984b
Aniline
No studies reported Kp data for this chemical; however, two studies repotted the flux of aniline from
aqueous media. Baranowska-Dutkiewicz (1982) reported the average absorption rate was 0.82 mg/cm2-hour for
aniline across human skin exposed in vivo (immersion of hands) to a 2% aqueous solution of aniline for 60 minutes.
A Kp of 0.041 cm/hr can be calculated by dividing the flux value by the concentration of aniline in aqueous solution,
20 mg/cm3. This study provides a very useful Kp for our scenarios because the data were obtained in humans in
vivo for aniline in dilute aqueous solutions. The absorption rate at 60 minutes was selected over that measured at
30 minutes to ensure that a steady state rate of absorption was occurring. The absorption rate from the 2% solution
was chosen because it is higher than the 1% solution also tested.
A-7
-------
In addition to the flux values obtained for aniline in aqueous solution, Baranowska-Dutkiewicz (1982) also
obtained flux values for liquid aniline and reported flux values for liquid aniline from other researchers. Although
these values for neat exposures are similar to those obtained for aqueous solutions of aniline, they are not as relevant
for the described exposure scenario. The flux value in the database from Tsuruta (1986) is the one also cited by
Baranowska-Dutkiewicz (1982) and Piotrowski (1957) for liquid aniline. The flux obtained by Baranowska-
Dutkiewicz (1982) may underpredict the flux expected in the scenario, because the temperature of the aqueous
solutions was about 20°C. Also, the body pan exposed to the solution was only the hand, which has a stratum
comeum that is significantly thicker than other tissues.
Benzene
Four studies reported Kp data for this chemical. The study by Blank and McAuliffe (1985) was selected
as the most appropriate Kp value to estimate absorbed dose in the scenarios. These investigators reported a Kp of
0.111 cm/hr for the penetration of an aqueous solution of benzene through human abdominal skin in vitro. One other
study, that of Dutkiewicz and Tyras (1968, as cited by Baranowska-Dutkiewicz, 1982), also reported an absorption
rate constant for benzene in aqueous solution, but no information was reported on the experimental conditions.
Other values from the Blank and McAuliffe (1985) paper wen: reported based on the penetration of benzene
through skin from various organic solvents, and are therefore inappropriate for our scenarios. The other values
reported in the database are either for liquid benzene (Tsuruta, 1982:1986) or benzene in the vapor phase (McDougal
et al., 1987; 1990).
p-Bromophenol
A single Kp value has been reported for this compound. Roberts et al. (1977) used human abdominal
epidermal membrane and a number of aqueous concentrations of p-bromophenol in vitro to determine a Kp of 6.02
x 10*4 cm/min (3.61 x 10*2 cm/hr). The stirred receptor cell contained distilled water. The temperature of the system
was maintained at 25°C. A number of different concentrations of 2,4-dichlorophenol in aqueous solution were used
in a series of such experiments to determine the Kp value.
23-Butanediol
Only one Kp was available in the literature for this compound. Blank et al. (1967) reported the Kp value
as less than 0.05 x 10"3 cm/hr for 23-butanediol in aqueous solution applied to human abdominal skin in vitro. The
A-8
-------
sensitivity of the assay presumably prevented a more quantitative characterization of the Kp. Therefore, 5 x 10"5
cm/hr was selected as an upper bound estimate of the Kp for this compound.
Butanol
Kp data for butanol can be found in seven references (Del Terzo et aL, 1986; Behl et al., 1983a; 1983b,
1984; Garcia, 1980; Scheuplein and Blank, 1971,1973). Values reported by Scheuplein and Blank (1971) and by
Del Terzo (1986) were from other studies. Blank et al. (1967) and Scheuplein and Blank (1973) presented the only
Kp for butanol derived using human skin. Unfortunately, Blank et al. (1967) provided only a range of values for
the permeability coefficient Therefore, the recommended Kp is from Scheuplein and Blank (1973). A Kp of 2.5
x 10"3 cm/hr for abdominal epidermis was generated in vitro at 2S°C using an aqueous butanol solution (0.1 M) and
a distilled water receptor.
2-Butanone (Methyl ethyl ketone)
Only one study has been identified that reported a Kp value for 2-butanone (Blank et al., 1967). This study
was conducted using human abdominal skin in vitro and an aqueous vehicle; therefore, the values are especially
applicable to the scenario. Blank et al. (1967) reported a Kp of 4 to 5 x 10~3 cm/hr for 2-butanone as an aqueous
solution to the human abdominal skin in vitro. The upper-bound Kp value of 5 x 10~3 cm/hr is recommended for
this compound.
2-Butoxyethanol
There are flux values in the database for neat 2-butoxyethanol applied in vivo to guinea pig skin (Jchanson
and Femstrom, 1986,1988) and human skin (Johanson et al., 1988), and in vitro to human skin (Dugard et al., 1984).
In addition, Johanson et al. (1988) reported flux values for five aqueous concentrations of this compound applied
to guinea pig skin in vivo. The Kp calculated by Johanson et al. (1988a) for the 5% (v/v) solution is 12 x 10"2 cm/hr
is the only value available for the aqueous solution of this compound, and is therefore the recommended value for
this compound.
2>(2-Butoxyethoxy)ethanol
The Kp for this compound was reported in one reference only (Dugard et al., 1984). The value of this
coefficient, 3.57 x 10"5 cm/hr, was generated using neat exposures to whole human abdominal skin, in vitro.
A-9
-------
Carbon disulfide
The permeability of this compound has been tested in only one study. Baranowska-Dutlriewicz (1982) cited
a previous study in Polish (Baranowska, 1968) that reported dermal absorption rate values of 023 to 0.79 mg/cm2-
hour for carbon disulfide (CSJ in aqueous solutions applied to human skin at concentrations from 0.42 to 1.49 g/1.
A Kp of approximately 0.54 cm/hr results from either the upper bound or lower bound flux estimate. Although this
Kp was obtained using human skin and an aqueous solution of the compound, it should be used cautiously, because
little information is known about the conditions under which the study was conducted. A flux value of 9.7
mg/cm2/hr for liquid CS2 applied to human skin was also reported (Baranowska, 1968, as cited in Baranowska-
Dutkiewicz, 1982), but not selected because of the absence of an aqueous vehicle.
4-ChIoro-3-cresol (4-Chloro-3-methylphenol)
Huq et al. (1986) reported a Kp value for this compound of 1.19 x 10"' cm/hr. The experiment used whole
hairless mouse skin in an in vitro stirred cell system maintained at a temperature of 37°C. The receptor
compartment contained a saline solution (pH = 62) while the donor compartment was loaded with an aqueous
solution of 4-chloro-3-cresol (0.5 mg/ml, pH = 6.18). The estimated stratum corneum permeability was calculated
to be 235 x 10"3 cm/hr, based on the whole skin experimental value. In addition, Roberts et al. (1977), using human
abdominal epidermal membranes in vitro, determined the Kp of an unspecified chlorocrcsol isomer to be 9.16 x 10~'°
cm/min (5.5 x 10"2 cm/hr). A series of different concentrations of chlorocresol in aqueous solution were used. The
stirred receptor cell contained distilled water. The temperature of the system was maintained at 25°C. Because
human tissues were used, the Kp of 5.5 x 10*2 cm/hr reported by Roberts et al. (1977) is the recommended value for
chlorocresol. [See p-Bromophenol for procedural details of Roberts et al. (1977)].
2-ChlorophenoI (o-chlorophenol)
Two papers reported Kp data for 2-chlorophenol, Huq et al. (1986) for hairless mouse skin and Roberts et
at (1977) for human skin. The recommended Kp value is therefore based on the data generated by Roberts et al.
(1977). A series of in vitro experiments were performed using human abdominal epidermal membranes and a
number of aqueous concentrations of o-chlorophenol to determine a Kp of 5.51 x 10"4 cm/min (331 x 10"2 cm/hr).
[See p-BromopnenoI for details of Roberts et al. (1977)].
A-10
-------
p-Chlorophenol
A single Kp value has been reported for this compound. Roberts et al. (1977) used human abdominal
epidermal membranes and a number of aqueous concentrations of p-chlorophenol in vitro to determine a Kp of 6.05
x Iff4 cm/min (3.63 x 10'2 cm/hr). [See p-Bromophenol for details of Roberts et al. (1977)].
Chloroxylenol (unspecified isomer)
Only a single Kp value has been identified for this compound. Roberts et al. (1977) performed a number
of in vitro permeability experiments on human abdominal epidermal membranes with a series of different
concentrations of chloroxylenol in aqueous solution. The Kp was determined to be 9.84 x Iff4 cm/min (5.90 x 10~2
cm/hr). [See p-Bromophenol for details of Roberts et al. (1977)].
m-Cresol
Only single Kp value has been identified for this compound. Roberts et al. (1977) used human abdominal
epidermal membranes and a number of aqueous concentrations of m-cresol, in vitro, to determine a Kp of 2.54 x Iff4
cm/min (1.52 x Iff2 cm/hr). [See p-Bromophenol for details of Roberts et al. (1977)].
o-Cresol
A single Kp value has been identified for this compound. Roberts et al. (1977) used human abdominal
epidermal membranes and a number of aqueous concentrations of 0-cresol, in vitro, to determine a Kp of 2.62 x 10*4
cm/min (1.57 x 10"2 cm/hr). [See p-Bromophenol for details of Roberts et al. (1977)].
p-Cresol
A single Kp value has been identified for this compound. Roberts et al. (1977) used human abdominal
epidermal membranes and a number of aqueous concentrations of p-cresol, in vitro, to determine a Kp of 2.92 x Iff4
cm/min (1.75 x 10"2 cm/hr). [See p-Bromophenol for details of Roberts et al. (1977)].
A-ll
-------
Cobaltous chloride
Only one study (Wahlberg, 1971) provided data to calculate a Kp for any of the cobalt salts. Flux values
were obtained after application of a 2.00% and a 5.69% aqueous CoClj solution amended with radiolabeled cobalt
to unspecified sites on the skin of guinea pigs in vivo. Flux values were repotted in ug Co cm*2 hr'1; therefore, the
proportion of cobalt in CoCl2, by weight, was used to estimate the aqueous concentration of the cobalt in solution,
which was subsequently used to calculate Kp.
The resultant Kp values for each concentration of CoClj applied to the skin were similar, 5.55 x 1CT4 and
4.96 x 10*4 cm/hr, for the 2.00% and 5.69% CoCl2, respectively. The value of 5.5 x 10*4 cm/hr was selected as being
the recommended value for calculating the dermally absorbed dose in these scenarios, because it was obtained using
the more dilute solution.
Wahlberg (1971) used a "disappearance technique" to obtain the flux values reported in this paper. The
uncertainties surrounding the use of data obtained by this technique are discussed in Section 4.1.3. This technique
involves following the disappearance of a radiolabeled compound from the skin, rather than the direct penetration
of the compound across the skin. Unfortunately, cobalt Kp values generated in other studies are unavailable. These
uncertainties should be recognized when using these data to calculate the dermally absorbed dose of cobalt.
Decanol
Scheuplein and Blank (1973) reported the only available value for the Kp of decanol. In vitro experiments
were run using human abdominal epidermis and an aqueous solution of decanol (3 x 10"4 M) at 25°C to provide a
Kp of 8.0 x 1(T2 cm/hr.
Dibutyl phthalate
Only one study was identified that reported Kp values for dibutyl phthalate (Scott et al., 1987). These
investigators obtained Kp values for neat dibutyl phthalate applied to rat and human abdominal epidermal membranes
in vitro. The K value obtained using human skin was 23 x 10"* cm/hr.
A-12
-------
2,4-Dichlorophenol
Two studies reported the Kp for this compound. The Kp for 2,4-dichlorophenol has been reported by Huq
et al. (1986) for hairless mouse skin, and by Roberts et al. (1977) for human skin. The recommended Kp value was
reported by Roberts et al. (1977) from in vitro permeability experiments on human abdominal epidermal membranes
to determine a Kp of 10.01 x 1(T* cm/min (6.01 x 10"2 cm/hr). (See p-Bromophenol for details of Roberts et al.,
1977.)
Di.(2-ethylhexyl) phtalate
Only one study was identified that reported Kp values for this compound (Scott et al., 1987). These
investigators obtained Kp values for neat di-(2-ethylhexyl)phthalate applied to rat and human abdominal epidermal
membranes in vitro. The value obtained using human skin was 5.7 x 10"6 cm/hr.
Diethyl phthalate
Only one study was identified that reported Kp values for diethyl phthalate (Scott et al., 1987). These
investigators obtained Kp values for neat diethyl phthalate applied to rat and human abdominal and rat epidermal
membranes in vitro. The Kp value obtained using human skin was 1.14 x Iff5 cm/hr.
2,4-Dimethylphenol
The Kp for this compound has only been reported by Huq et al. (1986) for hairless mouse skin. Experiments
were run in vitro using stirred cells maintained at 37°C. The receptor contained saline solution (pH = 6.2). The
donor was an aqueous solution of 2,4-dimethylphenol (0.5 mg/ml, pH = 6.31) and a Kp of 1.10 x 10"' cm/hr was
reported.
Dimethyl phthalate
Two studies reported the dermal permeability of this compound. Both Scott et al. (1987) and Dugard et
al. (1984) obtained dermal absorption rate constants for dimethyl phthalate. However, only Scott et al. (1987)
reported their results as Kp. The Kp for neat dimethyl phthalate applied to human abdominal epidermal membranes
in vitro was 332 x 10*s cm/hr. Scott et al. (1987) also obtained a K value for this compound applied to rat skin.
A-13
-------
2,4-Dinitrophenol
Two studies reported the dermal permeability of this compound. letter et al. (1986) and Huq et al. (1986)
performed a number of in vitro permeability experiments on hairless mouse abdominal skin sections. These two
studies originated from the same research team and used the same techniques. The experiments were performed over
a range of receptor and donor cell pH values. The receptor contained a saline solution, and the system temperature
was maintained at 37°C. A Kp value of 3.15 x Iff3 cm/hr (Huq et al., 1986) was obtained at pH 6.0. However, at
experimental conditions that most closely matched the exposure scenario conditions (donor and the receptor pH were
7.1 and 7.6, respectively), no absorption of this compound across the skin was observed. Therefore, the Kp value,
obtained at pH 6.0 (3.15 x 10"3 cm/hr) represents a reporting limit, not an absolute value. Huq et al. (1986) also
reported a test conducted with the donor cell pH at 2.0 with an experimental permeability coefficient for whole skin
of 1.51 x 10"' cm/hr and an estimated permeability coefficient of stratum corneum of 2.28 x 10*1 cm/hr.
Ethanol
The dermal permeation coefficient for ethanol has been determined in four papers (Del Terzo et al., 1986;
Behl et al., 1984; Garcia et al., 1980; Scheuplein and Blank, 1973). Scheuplein and Blank (1973) are the only
authors that presented their own experimentally derived K_ for ethanol through human epidermis. In vitro
experiments conducted at 25°C, using an aqueous solution of ethanol (0.1 M), resulted in a reported Kp of 0.8 x 10*3
cm/hr.
2-(2-Ethoxyethoxy) etbanol
The Kp for this compound was reported only by Dugard et al. (1984). The coefficient was determined using
whole human abdominal skin in vitro. Using a diffusion cell system maintained at 30°C. these investigators obtained
a Kp of 132 x Iff4 cm/hr.
2-Ethoxyethanol
Two studies reported permeability data for this compound. Kp values for 2-ethoxyethanol applied to human
skin in vitro have been reported by both Blank et al. (1967) and Dugard et al. (1984). However, only the Blank et
al. (1967) study involved application of this compound to the skin in an aqueous vehicle. From the range of Kp
values reported by Blank et al. (1967) for 2-ethoxyethanol (2 to 3 x 10~* cm/hr), the upper bound of this range, 3
x 10"* cm/hr, was selected as the recommended Kp value.
A-14
-------
2-Ethoxyethyl acetate
Two studies reported permeability data for this compound. Dugaid et al. (1984) obtained a Kp of 8.07 x
10"4 cm/hr for 2-ethoxyethyl acetate permeating through human abdominal epidermal membranes in vitro. One other
study reporting a dermal absorption rate for this compound was identified in the literature (Guest et al., 1984), but
only reported a flux for the neat compound.
Ethylbenzene
Four papers reported dermal permeability data for the compound. Dutkiewicz and Tyras (1967) determined
the flux of ethylbenzene in aqueous solution across human skin in vivo. Using a "direct" method, one that involves
measuring the disappearance of compound from the donor solution, these investigators report a mean flux of 118
ug/cm2/hr and 215.7 ug/cm2/hr for ethylbenzene penetrating through the skin from aqueous solutions of 112.0 and
156.2 mg/1. From these flux rates and aqueous concentrations, it is possible to calculate Kp values of 1.05 cm/hr
and 1.38 cm/hr under these conditions. Although the validity of the direct method developed by these researchers
has recently been questioned, the Kp value of 1.38 cm/hr is supported by a Kp of 1.33 cm/hr calculated by using a
flux value obtained by determining the amount of ethylbenzene absorbed from the measurement of a urinary
metabolite, mandelic acid. Therefore, the 1.38 cm/hr Kp value from this study appears valid, and is selected as the
recommended Kp to use in calculating the dermally absorbed dose in these scenarios.
The other dermal absorption rate values in the database are either a summary of the Dutkiewicz and Tyras
(1967) results reported by Baranowska-Dutkiewicz (1982), or results obtained after applying neat ethylbenzene to
rat skin in vitro (Tsuruta, 1982; Tsuruta, 19865).
Ethyl ether
Only one study has been identified that reports a Kp value for ethyl ether (Blank et al., 1967), and the values
are especially applicable to the scenario. Blank et al. (1967) reported a Kp of 1.5 to 1.7 x 10"2 cm/hr for ethyl ether
applied as an aqueous solution to the human abdominal skin in vitro. The upper bound Kp of 1.7 x 10*2 cm/hr is
the recommended value for this compound.
A-15
-------
p-Ethylphenol
A single Kp value has been reported for this compound. Roberts et al (1977) used human abdominal
epidermal membrane and a number of aqueous concentrations of p-ethylphenol in vitro to determine a Kp of S.81
x Iff4 cm/min (3.49 x 10'2 cm/hr). (See p-Bromophenol for details of Roberts et al., 1977).
Glucose
Only one paper reported a Kp for glucose. This compound is not an environmental pollutant, but is included
in the database as an example of a highly polar organic compound. Ackermann and Flynn (1987) reported a Kp of
9.5 x 10"5 cm/hr for glucose in saline solution permeating through full thickness nude mouse abdominal skin in vitro.
Since most compounds enter the circulation via the capillaries present just below the epidermal layer, the
use of full-thickness skin may result in artificially high estimates of Kp, especially for highly lipophilic compounds
whose penetration through the dermis would be impeded by the aqueous nature of this layer. However, since the
Kp for glucose permeating only through the dermis (0.29 cm/hr) is about three to four orders of magnitude greater
than the value for full-thickness skin, it is clear that the epidermis (and probably specifically the stratum corneum)
represents the diffusional barrier for this compound, and that the full-thickness skin Kp value is valid.
Glycerol
Only one paper reported a Kp for glycerol. This compound is not an environmental pollutant, but is
included in the database as an example of a highly polar organic compound. Ackermann and Flynn (1987) reported
a Kp of 1.4 x 10~* cm/hr for glycerol in saline solution permeating through full thickness nude mouse skin in vitro.
Since most compounds enter the circulation via the capillaries present just below the epidermal layer, the
use of full-thickness skin may result in artificially high estimates of Kp, especially for highly lipophilic compounds
whose penetration through the dermis would be impeded by the aqueous nature of this layer. However, since the
Kp for glycerol permeating only through the dermis (0.41 cm/hr) is about three to four orders of magnitude greater
than the value for full-thickness skin, it is clear that the epidermis (and probably specifically the stratum comeum)
represents the diffusional barrier for this compound, and that the full- thickness skin Kp value is valid.
A-16
-------
Heptanol
Four studies reported Kp values for heptanol, but only two of these presented values using human epidermal
tissue. Scheuplein and Blank (1973) used an aqueous heptanol solution and performed experiments at 25°C. Blank
et al (1967) used similar experimental procedures, but performed the experiments at 30°C. Since the latter
temperature is closer to human body temperature and the Kp was produced with in vitro human abdominal epidermal
membrane, the K of 3.76 x 10"2 cm/hr reported by Blank et al. (1967) is recommended as the default value.
Hexanol
The dermal permeability coefficient for hexanol has been reported in several papers. However,
experimentally derived values for human skin were reported only by Bond and Barry (1988) and by Scheuplein and
Blank (1973). The Bond and Barry in vitro abdominal tissue experiments were performed at 31°C, and are therefore
more representative of human body temperature that the 25°C used by Scheuplein and Blank (1973), and the Kp of
27.7 x 10" 3 cm/hr is the recommended default value.
Mercuric chloride
Two papers reported permeability data for this compound. Wahlberg (1971) investigated the absorption of
aqueous solutions of mercuric chloride (HgCl2) applied to guinea pig skin in vivo. Since flux was reported as pg
Hg cm'2 hr"1, the Kp was calculated on a percentage weight basis for Hg in the HgC^ solution. Kp values of 1.55
x 10'3, 2.09 x 10"*, and 8.13 x KT4 cm/hr were calculated from the flux data provided by Wahlberg (1971) for Hg
in 0.14%, 2.17%, and 6.49% aqueous HgCl2 solutions, respectively. The Kp obtained at the lowest concentration
(1.55 x 10~3 cm/hr) was selected as the recommended value for calculating a dermally absorbed dose in the scenario.
Similar values were obtained for Hg in HgCI2 solutions by these researchers in an earlier study (Skog and Wahlberg,
1964). Both Wahlberg (1971) and Skog and Wahlberg (1964) used the disappearance technique to obtain flux values
for Hg. See Section 4.1.3 for a discussion of the limitations of this technique.
Methanol
Five papers presented Kp values for this compound. The Kp for methanol has been determined for rat skin
by Del Terzo et al. (1986) and Behl et al. (1983a), and by Behl et al. (1984) for mouse skin. Scheuplein and Blank
(1973) presented data for experiments conducted using human abdominal epidermal tissue in an in vitro diffusion
apparatus, and loaded the donor compartment with an aqueous methanol solution (0.1 M) to determine a Kp of 0.5
A-17
-------
x Iff3 cm/hr. Southwell et al. (1984) used water conditioned stratum comeum of human abdominal skin in vitro and
reported a Kp for methanol of 1.6 x 10'3 cm/hr.
2-Methoxyethanol
For 2-Methoxyethanol, the only Kp was identified in the study by Dugard et al. (1984). A value of 2.89
x Iff3 cm/hr was obtained for neat 2-methoxyethanol permeating through human abdominal epidermal membranes
in vitro tested at 30°C.
2-(2-Methoxyethoxy)ethanol
The only Kp identified for this compound was from the study by Dugard et al. (1984). The coefficient was
determined using neat material with whole human abdominal skin, in vitro. The diffusion cell system was
maintained at 30°C, and the Kp was determined to be 2.06 x 10*4 cm/hr.
l-Methoxypropan-2-ol
Only one Kp was identified from the literature for this compound. Dugard et al. (1984) obtained a value
of 125 x Iff3 cm/hr for neat l-methoxypropan-2-ol applied to human skin in vitro tested at 30°C.
Methyl hydroxybenzoate (unspecified isomer)
A single study was identified that provided a Kp value for mis compound. Roberts et al. (1977) obtained
a Kp of 1.52 x Iff4 cm/min for methyl hydroxybenzoate permeating me human abdominal epidermal layer in vitro
from an aqueous solution. Conversion of this value to a Kp in cm/hr yields 9.12 x 10~3 cm/hr. (See /vBromophenol
for details of Roberts et al., 1977.)
Methyl mercury-dicyandiamide
Only one study was identified that provided permeability data for this compound. Skog and Wahlberg
(1964) investigated the absorption of aqueous solutions of methyl mercury-dicyandiamide applied to guinea pig skin
in vivo. Since flux was reported as pg Hg cm"2 hr"1, the Kp was calculated on a percentage weight basis for Hg in
the methyl mercury-dicyandiamide solution. Values of 3.34 x Iff3 and 4.39 x 10'3 cm/hr were calculated from the
flux data provided by Skog and Wahlberg (1964) for mercury in aqueous solutions of 0.04 M and 0.08 M,
A-18
-------
respectively, of this compound. The Kp obtained at the lower concentration (3.34 x 10'3 cm/hr) was selected as the
recommended value for calculating a dermally absorbed dose in these scenarios.
In this study, Skog and Wahlberg (1964) used the disappearance technique to estimate the flux of mercury
across guinea pig skin. See Section 4.1.3 for a discussion of the limitations of this technique.
B-Naphthol
A single Kp value has been identified for this compound. Roberts et al. (1977) used human epidermal
abdominal membranes and a number of aqueous concentrations of 6-naphthol in vitro to determine a Kp of 4.65 x
Iff4 cm/min (2.79 x 10"2 cm/hr). (See p-Bromophenol for details of Roberts et al., 1977.)
2-NitrophenoI
Two studies provided Kp values for this compound. The Kp for this compound, 1.01 x 10*' cm/hour, was
reported by both Jetzer et al. (1986) and Huq et al. (1986) for hairless mouse skin. However, the experimental value
for the coefficient originated in the Huq et al. (1986) paper. The experiments used whole abdominal mouse skin,
in an in vitro stirred cell system maintained at 37°C. The receptor contained saline solution (pH = 6.2). The donor
was loaded with an aqueous solution of 2-nitrophenol (0.5 mg/ml, pH - 3.46).
4-Nitrophenol
Four papers provided Kp data for this compound. Jetzer et al. (1986), Jetzer (1988), and Huq et al. (1986)
performed a number of in vitro experiments using hairless mouse abdominal skin. The recommended Kp, 5.58 x
10"3 cm/hr, based on a value of 0.93 x KT* cm/min, comes from work by Roberts et al. (1977) who used human
epidermal tissues. (See p-Bromophenol for details of Roberts et al., 1977.)
3-Nitrophenol
A single Kp value has been identified for this compound. Roberts et al. (1977) used human abdominal
epidermal membrane and a number of aqueous concentrations of 3-nitrophenol in vitro to determine a Kp of 0.94
x Iff4 cm/min (5.64 x 10*3 cm/hr). (See p-Bromophenol for details of Roberts et al., 1977.)
A-19
-------
Nonanol
The only Kp value identified for nonanol was provided by Scheuplein and Blank (1973). The in vitro
experiments used human abdominal epidermal tissues and an aqueous solution of nonanol (1.4 x 10~3 m/1) to
determine a K value of 6.0 x 10*2 cm/hr.
Octanol
Five papers presented permeability data for this compound. The Kp for octanol has been determined by
three researchers (Del Terzo et al., 1986; Behl et al., 1984; Garcia et al., 1980) using rat or mouse skin. Southwell
et al. (1984) and Scheuplein and Blank (1973) reported the only experimentally derived octanol Kp values using
human skin. Similar equipment and experimental conditions were used by both groups. However, Southwell et al.
(1984) used well-hydrated skin. The K,, values reported in both papers are close (6.1 x 10~2 cm/hr in Southwell et
al., 1984 versus 52 x 10~2 cm/hr in Scheuplein and Blank, 1973), but the value presented by Southwell et al. (1984)
is recommended for use as a default because of the skin hydration conditions and because the value represents a
newer data set.
Parathion
Three studies reported permeability data for this compound. Although parathion is one of the most widely
studied dermal penetrants, only Knaak et al. (1984b) reported K. values for this compound, and no studies involved
the application of aqueous solutions of parathion. Knaak et al. (1984) calculated Kp values for parathion applied to
rat skin in vivo for adult male and female animals, based either on the t1/2 for plasma elimination of the compound
or the ti/2 for loss of parathion from the skin (see Section 5.3.2 for a description of this technique). The Kp value
obtained using plasma elimination data from female rats. 1.0 x 10"2 cm/hr, was selected as being the most appropriate
of the four Kp values available, because female rat skin is more permeable to parathion and the Kp from plasma
elimination is probably more accurate than the Kp from surface loss for a moderately volatile compound such as
parathion. However, since water tends to accelerate the uptake of compounds across the skin (see Section 3.0), this
Kp value for neat parathion probably underestimates the Kp that would be expected for an aqueous solution of
parathion.
Flux values from the studies by Fredericksson (1961a, b) are also included in the database. However, since
the absorption rate was determined indirectly as a function of the rate of acetylcholinesterase inhibition, it is unclear
A-20
-------
whether these values represent the rate of parathion absorption, or the absorption of parathion and its subsequent
conversion to its active metabolite, naraoxon.
conversion to its active metabolite, paraoxon
At least 12 studies have been conducted that reported percent absorbed values for parathion. Although flux
values can theoretically be derived from these data, none of the studies used aqueous solutions of parathion, and
therefore, they each have limited use for our scenarios.
Pentanol
Scheuplein and Blank (1973) presented the only experimentally derived data for Kp of pentanol. The in
vitro experiments used human abdominal epidermal tissue and were performed at 25°C using an aqueous solution
of pentanol (0.1 M). The K was determined to be 6.0 x 10° cm/hr.
Phenol
Six papers presented permeability data for this compound. Experimentally derived Kp values for phenol
were presented by a number of researchers using mouse skin (Huq et al., 1986; Jetzer et al., 1986, 1988; Behl et al.,
1983b). Experiments using human skin were performed by Southwell et al. (1984) and Roberts et al. (1977).
Southwell et al. (1984) used a 1% (w/v) aqueous phenol solution penetrating in vitro through human abdominal
stratum corneum at 22°C, and calculated a mean steady state flux = 1.49 ug/cm2/hr (Kp calculated as 1.49 x 10"4
cm/hr). Roberts et al. (1977) used a number of aqueous phenol concentrations in vitro for human abdominal
epidermal membranes to determine a Kp of 1.37 x 10~s cm/min (8.22 x 10~3 cm/hr). (See p-Bromophenol for details
of Roberts et al, 1977.) Because this is a higher value than the Kp obtained in the Southwell et al. (1984) study, and
was obtained with more dilute solutions and higher temperatures more closely approximating human body
temperatures, the Kp obtained in the Roberts et al. (1977) study is recommended when calculating the dermally
absorbed dose of phenol.
Potassium mercuric iodide
Only one paper presented permeability data for this compound. Skog and Wahlberg (1964) investigated
the absorption of aqueous solutions of potassium mercuric iodine (KjHgl^ applied to guinea pig skin in vivo. Since
flux was reported as muM Hg cm"2 hr"1, the Kp was calculated on a percentage weight basis for Hg in the K2HgI4
solution. Values ranging from 4.48 x 10"3 to 1.05 x 10~2 cm/hr were calculated from the flux data provided by Skog
and Wahlberg (1964) for Hg in a range of aqueous K2HgI4 solutions. The Kp values in the middle range of
A-21
-------
concentrations were all approximately 1 x 10"2 cm/hr. Therefore, tiie value of 1.05 x 10"2 cm/hr was selected as
being representative for this compound.
In this study, Skog and Wahlberg (1964) used the disappearance technique to estimate the flux of mercury
across guinea pig skin. See Section 4.13 for a discussion of the limitations of this technique.
Propanol
Kp values for propanol penetrating human abdominal epidermal membranes in vitro have been reported in
two papers (Blank et al., 1967; Scheuplein and Blank, 1973). Experiments were performed at 30°C by Blank et al.
(1967) and are, therefore, more representative of contact scenarios than the 25°C used by Scheuplein and Blank
(1973). The Kp reported by Blank et al. (1967) was 1.7 x 10'3 cm/hr.
Resorcinol
A single Kp value has been identified for this compound. Roberts et al. (1977) used human abdominal
epidermal membranes and a number of aqueous concentrations of resorcinol in vitro to determine a Kp of 0.04 x 10"4
cm/min (2.4 x Iff4 cm/hr). (See p-Bromophenol for details of Roberts et al., 1977.)
Sodium chromate (Hexavalent chromium)
Two studies quantified the dermal absorption of nexavalent chromium from aqueous sodium chromate
solutions. Baranowska-Dutkiewicz (1981) obtained flux values of 1.1,6.5, and 10.0 ug/cm2/hr for Cr** applied to
the skin of humans in vivo in 0.01,0.1, and 0.2 M solutions of sodium chromate, respectively. From these data, Kp
values of 3.4 x 10*3, 2.03 x 10'3, and 1.6 x 10~3 cm/hr could be calculated based on the weight percent of Cr** in
sodium chromate. Similar values were reported by Wahlberg (1971) for sodium chromate solutions applied to guinea
pig skin. The Kp of 3.4 x 10"3 cm/hr was selected as the recommended value to calculate the dermally absorbed dose
in these scenarios for the following reasons: it was calculated from a human in vivo study, an aqueous solution was
applied to the skin, and it is the Kp obtained at the lowest concentration of sodium chromate.
Both Baranowska-Dutkiewicz (1981) and Wahlberg (1971) use a disappearance technique to estimate flux
values for this compound. See Section 4.13 for a discussion of the Limitations of this technique.
A-22
-------
Styrene
Eight papers presented permeability data for this compound. Several of the studies provided flux values
for neat styrene permeating as a liquid (Tsuruta, 1982; Tsuruta, 1986) or as a vapor (Riihimaki and Pfaffli, 1978;
McDougal et ah, 1987, 1990) across the skin; however, only the study by Dutkiewicz and Tyras (1968) reported
values for the absorption of styrene from an aqueous vehicle. The researchers report a flux value of 40 to 180
ug/cm2/hr for mean aqueous concentrations for 66.5 to 269 mg/1 of styrene. This information allows us to calculate
Kp values of 0.60 (lower bound) to 0.67 (upper bound) cm/hr for styrene. Although the lower value may be more
appropriate for these scenarios, because the concentration used in the study more closely approximates the
concentration of styrene found in drinking or surface water supplies, the higher value was selected to give a slightly
more conservative estimate of the absorbed dose.
Little information was reported by Dutkiewicz and Tyras (1968) in this brief communication regarding the
conditions under which the study was conducted, other than to state that the method was developed previously by
these researchers (Dutkiewicz and Tyras, 1967). Since the amount of styrene absorbed was apparently quantified
by monitoring the amount of mandelic acid excreted in the urine, and not by measuring the loss of the compound
from the donor solution, more confidence can be placed in the validity of the flux values as having been calculated
from the actual absorbed dose.
Thiourea
Only one study reported a Kp for this compound. This compound is not an environmental pollutant, but
is included in the database as an example of a highly polar compound. Ackermann and Flynn (1987) reported a Kp
of 9.6 x 10~5 cm/hr for thiourea in saline solution permeating through full thickness nude mouse skin in vitro.
Since most compounds enter the circulation via the capillaries present just below the epidermal layer, the
use of full-thickness skin may result in artificially high estimates of Kp, especially for highly lipophilic compounds
whose penetration through the dermis would be impeded by the aqueous nature of this layer. However, since the
Kp for thiourea permeating only through the dermis (0.62 cm/hr) is about three to four orders of magnitude higher
than the value for full-thickness skin, it is clear that th epidermis (and probably especially the stratum comeum)
represents the diffusional barrier for this compound, and that the full-thickness skin Kp value is valid.
A-23
-------
Thymol
Only a single Kp value has been identified for this compound. Roberts et al. (1977) used human abdominal
epidermal membranes and a number of aqueous concentrations of thymol in vitro to determine a Kp of 8.8 x 10"*
cm/min (5.28 x Iff2 cm/hr). (See p-Bromophenol for details of Roberts et al., 1977.)
Toluene
Kp data for toluene were presented in seven papers. However, only Dutkiewicz and Tyrss (1968) and
Baranowska-DuUtiewicz (1982) used human skin and an aqueous solution. A Kp value of 1.01 cm/hr can be both
read from the graph plotting the rate of absorption against concentration (Dutkiewiez and Tyras, 1968) and calculated
from the flux data of Baranowska-Dutkiewicz (1982).
2,4,6-Trichlorophenol
Two studies presented Kp values for this compound. The permeability of 2,4,6-trichlorophenol was reported
for mouse skin by Huq et al. (1986) and for human skin by Roberts et al. (1977). The recommended default Kp
value is therefore based on the human permeability data. Experiments by Roberts et al. (1977) were performed in
vitro, at 2S°C, using abdominal epidermal tissue and a stirred distilled water receptor, and the donor compartments
were loaded with a number of different concentrations of an aqueous solution of 2,4,6-trichlorophenol to determine
a K value of 9.9 x Iff4 cm/min (5.9 x 1(T2 cm/hr).
Urea
Only one study presented Kp values for urea. This compound is not an environmental pollutant, but is
included in the database as an example of a highly polar compound, Ackermann and Flynn (1987) reported a Kp
of 1.2 x 10"4 cm/hr for urea in saline solution permeating through full thickness nude mouse skin in vitro.
Since most compounds enter the circulation via the capillaries present just below the epidermal layer, the
use of full-thickness skin may result in artificially high estimates of Kp, especially for highly lipophilic compounds
whose penetration through the dermis would be impeded by the aqueous nature of this layer. However, since the
Kp for urea permeating only through the dermis (0.68 cm/hr) is about three to four orders of magnitude higher than
the value for full-thickness skin, it is clear that the epidermis (and probably specifically the stratum corneum)
represents the diffusions] barrier for this compound, and that the full-thickness skin Kp value is valid.
A-24
-------
Water
Seven studies presented Kp values for this compound. Permeability coefficients for water have been
obtained for several species including humans. In fact, the range of "normal" values for the Kp of water through
human skin has been so well characterized that Kp values >2J x 10*3 cm/hr are suggestive of skin damage.
Bronaugh et al. (1986b) pointed out that most Kp values for human skin in vitro have ranged historically from 0.5
x 10~3 to 2.5 x JO"3 cm/hr. From their extensive study, using skin from individuals of both sexes that span a broad
range of ages, Bronaugh et aL (1986b) reported an average Kp value of 1.55 x 10~3 cm/hi. This value is
recommended as the default Kp for water.
Human Kp values for water reported by Bond and Barry (1988) and Bronaugh and Stewart (1986) are
essentially identical to this average value. Values for the Kp of water across animal skin in vitro range from being
very similar to human (e.g., 0.6 to 22 x 10'3 cm/hr for mouse skin as reported by Behl et al., 1984) to being about
3 to 8 times greater (e.g.. Del Terzo et al., 1986).
3,4-Xylenol
Only one identified paper has reported a Kp value for 3.4-xylenol. Roberts et al. (1977) performed a series
of experiments using a number of concentrations of aqueous solutions of 3.4-xylenol, and human abdominal
epidermal tissue in an in vitro system consisting of stirred distilled water receptor at 25°C and the Kp value was
reported to be 6.0 x 10"4 cm/min (3.6 x 10'2 cm/hr). [See p-Bromophenol for details of Roberts et al., 1977).]
A.4 Estimated Kp Values from Structure-Activity Relationships
Table A-4 presents estimated Kp values derived from the models of Lien and Tong (1973), Guy and
Hadgraft (1989b), Flynn (1990). and Kastings et al. (1987). The Kastings et al. (1987) model was used based on
Flynn's (1990) data by both Guy (personal communication, 1991) and Bronaugh (personal communication, 1991).
A-25
-------
CHEMICM MNE
womtm*
VlHTl CH.0UK
CHOWttWHE
MUHYUW CM.OMDE
IJtONDMElmNE
OICHUWOtTHTllHt. 1.2-
OtatOROEtmNE. 1.2-
DICHlOmtHTUK. l.l-
DtOHFOM
OlCHUHOCtHME, l.l-
iiicHLOMtTKn.tiK
niatomimut, 1.1.1-
lllCHOtOttHMt. 1.1.2-
uiatororiuwoMEiHMC
CARBON HIMCHOBIOC
DlCHUmnDPAK, 1.2-
o*oNOMctt.o«a«lwNt
HIMCHOKimitM
tEnumoncTmiK. 1.1.2.2-
EnniEKOIOMMIDE
CMUMOlMNKimK
HEMOO.ORKTHME
OHOHDTOI04
DICHUMOinUDmNEtHME
HEXACHUMEWtADlEK
iisicH.im4Enin.jtna
llsU-CHUMtnNl|E1HM
tanUUH
NMONiA
CTMIOC
OKOMIUN
»ICMI
COWM
2IK
MSEMIC
CASK* DISW.F1DE
SEIE01W
SILtER
CADMIUM
COBAUOW CHUMOE
CHHMIUPI (SODIUM CHROWIE)
MERCIRT
THMltUM
LEAD
MCKWiC OtniDE
niASSIUH HEOCWIC IODIDE
MEHniMEICURf DICYANOIMNIDE
MIU
MCIHN«.
EIHTUNt OXIDE
ewNicM
cuss
mi. Mim.
mi. Mim.
mi. Mim.
mi. Mim.
mi. Mim.
mi. Mim.
mi. Mim.
mi. Mim.
mi. Mim.
mi. Aitm.
mi. Mim.
mi. Mim.
mi. Mim.
mi. Mim.
mi. Mim.
mi. Mim.
mi. Mim.
mi. Mim.
mi. Mim.
HM. Mim.
mi. Mim.
HM. Mim.
mi. Mim.
mi. Him.
HM. AllPH.
mi. EMI
mi. ETHU
HEt./IKM.
Kt./IMK.
MEt./IKM.
MCI./IMM6.
Ml. /IMS.
MEI. /INDUS.
MT./INM6.
HCT /inQico.
HEt. /INDUS.
MET./INMS.
NCI./IMM.
MEI./IMJI6.
MI./INMS.
MEI./IKM.
Wt. /INDUS.
NEI./IKM.
Mtl./INDBS.
MEI./IKM.
MU./IMM.
MISC.
MISC.
MISC.
MISC.
CODE
M.RHX.
CASH)
74-07-1
1501 4
15 00 3
75-01-2
74-03-1
7S-3S-4
75-34-3
101-08-2
UK 3
7S-7I-0
mi i
11 SSI
71-OO-S
75-11-4
SI-23-S
120-01-2
7S-27-4
127-10-4
71-34-S
101-13-4
124-40-1
07-72-1
7S-2S-2
II 14 1
11-68-1
S42-M-1
111-44-4
7440-41-7
7IM-4I-7
S7-12-S
7440-47-3
7440-02-0
7440-SO-O
7440-U-O
7MS-3S-2
7S-IS-0
77(2-41-2
7440-22-4
7440-43-0
1641 If 1
7I7S-II-3
7431-17-1
7440-28-0
7431-12-1
7401-14-7
7101-31-7
S02-M-8
7112-ll-S
01-58-1
75-21-8
MOIEC
VT
SO 5
MS
04. S
04.1
W.I
n
n
111.4
120.1
III. 4
131.4
133.4
137.4
153.0
163
103.0
IIS.O
187.1
181
200.3
231.7
252.0
2S3.I
260.1
IIS
14)
I.Oit
17
20
SO.7
035
154
ii 1
IOM
112.4
121.1
164
200.1
204.4
207.2
271 .S
780.4
10
32
44.1
n llkjnols otenoli
IDS (tM*") Ihuwnl
HM Km lltntTant.Ill) 6uy4H«d»r«ft.lWO
I* Rp n lot K» to
(oi/hr) (oVhr
O.I2030S 0.11
22.MM7 1 M
20.11534
17.70211
15.48010
134 .OK2
01.15150
30.1H5I
11.32541
144. $431
201.0207
30I.02IS
lit .2011
331.0441
070.0821
031.7037
121.0260
2511.880
245.4700
11.20100
101.0243
0511.380
.43
.25
.11
.13
.71
.41
.17
.11
.42
.41
.05
.SI
.01
.12
.0!
.40
11
S
.1)
234.4220 2.17
3215.130 LSI
802SS.H 4.70
11.4W44 1.21
1 0416*4 -1.38
0 169824 -0.77 -3.30 5.06C-04
O.SOIIB1 -0.10
.51 1
14
.07
Tl
*«•
!37
.71
.02
.S3
.34
.00
.01
45
.17
.07
.51
.41
.10
.11
54
.27
.57
.13
.n
.72
.21
.SO
.50
.50
.50
.50
.SO
.50
.50
.SO
5Q
.50
4 00
4 27
3.80
!.51t-01
.24E-01
.SIC -01
.I2E-01
.m-oi
.27E-02
.ISE-02
.S5E-01
.ISE-02
.57E-02
.12E-02
.77E-02
.S5E-02
.01E-OI
.I4E-01
.03E-01
.OK-04
.14C-01
161-04
.OK-04
llt-04
.I9C-02 -I.SO 1.IM-W
.4IE-04
.02E-02 -I.SO 3.I6C-02
llt-01 -I.SO 1.I8E-02
.OKMM
.llt-01
.IK-04
.IK-04
.IK-04
.IK-04
.IK-04
.IK-04
.IK-04
.IK-04
i6£-54
.IK-04
.IK-04
.IK-04
.IK-D4
I.32C-05 -300 I.ODE-03
i.31C-05 -1.00 I.OOE-03
.SK-04 -1.00 1. OOC-OJ
51 1
.22 !
.17
.41
12
M
.0$
.30
11
M
.51
.01
.S7
.n
.30
.20
-M
.11
.$0
.37
.17
54
.40
.11
.IS
.OS
.21
.27
.3)
.SI
.50
.51
.00
.07
M
.70
.N
.11
.OS
.21
.54
51
.51
.04
.12
IS
M
.14
.10
SK-fll
lit M
m M
.24E-01
401 01
.03E-02
.15E-01
.J8C-03
.7K-01
.481-02
2H-02
.5K-02
J9C 01
.OK-02
.OK-02
34C-02
.13E-01
JSt-ll
IK-02
.IK-01
.2SE-01
.3K-OI
KE-01
.ME-02
.511-01
.I2E-04
.4K-01
IK-04
IK-04
.C7E-04
.OK-04
llt-04
.SK-04
.4X-04
.I4E-04
.5K-54
.OK-04
.2K-04
.lit 04
llt-05
111 OS
MC 05
.IK-OS
.SIC -05
.IK-01
.42C-OI
.OK-04
.211-05
.»« -os
.15C-04
2.11
t.M
I OS
2.21
.11
.77
.01
.22
.02
.N
.71
.71
.05
.75
.15
.05
.21
.34
.01
.44
.31
.42
.51
.00
.00
.31
.12
.M
.71
77
.13
17
.00
.01
.01
.«?
.01
.27
.10
.40
.01
.01
.24
.07
.21
.41
.01
M
.12
.08
1« 03
.32E-03
Oil 03
.12E-01
.DK-01
CK-02
.01E-01
.iiE-oi
.SK-01
.2K-02
.02E-02
.7K-02
SIC M
.711-02
.2K-02
.2SE-02
.IK-01
.541 02
211 03
601-03
lit 03
.OK-02
.73E-03
.SK-02
001 01
.OK-04
.41E-03
.IK-03
.13E-03
10E-01
.IK-01
08C-01
.OIE-01
.OK -04
SIC -04
431-04
.IK-04
.IK-04
.OK-04
.SK-04
.4SE-04
.41E-04
19C-04
341-04
.421-05
.121-00
4* 03
.211-04
.llt-04
.12t -04
Table A-4: Pcedicled Kp Values for Priority Pollulanls
-------
n-«lli«nolt
CHEHICAL NAME
ETHAKH
MOMKOl
MIA
BUI wait. 2-
BOTum.. N-
ETKTL ETHER
IHIOUIEA
NEIHOKVETHAMOL 1-
PEN1AMN.
EIHO»EIHMOL. t- '
BUTANEOIOl. 1.3-
6LTCEROL
EIHTLENETHIOMEA
PENIANONE. 4-HETHYl-t-
HEXANOl
BUTmUTHAm.. 1-
ETHMOL. t-(t-*ETHO»YETHKY)-
OCTAMX
EIHOMETHVL ACETATE. 2-
EIHMOL. l-(t-EtHOmTHO«T|-
NONXm.
DECANOL
ETHMDL. t-tt-BUTOWTHBM)-
> GLUCOSE
i DIOXANE. 1.4-
N> DIHETHTimm.. 1.4-
^ BENZENE
TOLUENE
ANILINE
OICHLOROPHENOI. 1 4-
STTREM
ETHYLBENZENE
HIENE
WENT. N-
CRESOL, 0-
CRESOl. H-
CRESa, P-
RESORCim
CHLOROHRZEIN:
ETHfincm.. r-
OIRITROmENOL. 2,4-
OIHETHnnCHOL. 3.4-
CHioRomm. 4-
CHLOROFWNOL. 2-
NimopHEm., t-
NliROPHENOL. 4-
NIIROfHENOL. 3-
CHENICAl CODE
CUBS
HISC. ALCOHOL
HISC. ALCOHOL
MISC.
HISC.
MISC. ALCOHOL
HISC.
HISC.
HISC. ALCOHOL
HISC. AlCOHOl
HISC. ALCOHOL
HISC. ALCOHOL
HISC. ALCOHOL
HISC. ALCOHOL
HISC.
HISC.
HISC. ALCOHOL
HISC. ALCOHOL
HISC. ALCOHX
HISC. AlCOHOl
HISC. ALCOHOL
HISC.
HISC. ALCOHOL
HISC. ALCOHOL
HISC. ALCOHOL
HISC. ALCOHOL
HISC.
MISC.
MONO. NKM PHENOL
HMO. AMH
MONO. AMH
HMO. AMH
HMO. AMH
MONO. AMH
HMO. AMM
HOMO. AMH
MONO. AMH
HMO. AMH
HMO. AMH
HMO. AMH
HMO. AMH
HMO. AMH
HMO. AMH
HMO. AMH
HMO. AMN
MONO. AMH
HMO. AMH
HMO. AMH
HMO. AMH
HMO. AMH
HMO. AMM
mm
mm
mm
mm
mm
mm
mm.
mm
mm
mm
mm
mm
mm
mm.
CASNO
64-17-
71-23
S7-I3-
78-91-
71 36
62-56-6
109-86-4
M-41-0
110-80 5
SI3-8S-I
5*81-5
96-45-7
1 06-16-1
lii-t;-3
111-70 6
III-7I-2
III-I7-S
111-159
III-N-I
14306 1
lll-N-l
112-34-S
Sfl-99-/
606-20-2
II-7S-I
71-43-t
1 08-88-3
62-53-3
IN-IS-t
156-59-2
100-42-5
100-41-4
1330-20-7
108-38-3
95-46-7
108-39-4
106-44-5
106-46-3
106-90-7
123-67-9
95-65-8
105-67-6
1*6-48-9
95-57-1
N-75-S
100-02-7
554-64-7
HDLEC
WT
41
n
to
71
74.1
H.I
71
71
N
N
N
N.I
92 1
N
IN
102
III
111
131
131
134
144
151.3
162
IN.I
161.1
H.I
71.1
92 1
93.1
94
N.I
164.1
IN.2
IN.t
IN.t
IN.I
IN.I
IN.I
lll.l
llt.l
120
Itt.t
Itt.t
1268
1211
131.1
139.1
139.1
to.
.489771
195212
'. 007761
.949644
466835
.762471
.112201
.169624
6.30780
.660693
.794328
.120221
0017378
0 }|877|
15.48811
107 1519
157.039$
8.760829
1.380189
133.1543
4466835
I.6317S3
1937.64*
ltttt.41
1.120226
52.48074
1.281153
134.6962
537.6317
;. 943211
2164031
72.44359
891.2509
1412.537
1SB4.B93
69 12509
91.20108
17.03635
1.309573
691.6309
161.9700
1696243
199.5262
2454706
144.5439
II. 65951
81.28305
100
LOS |Kum>
to. Hcnlfonf.
973
lag Up Kp
(on/hrl
-0.31
-Ml
1.19
I.IS
189
-195
-1.77
LSI
-l.ll
•010
-1.91
-1.71
-I.N
l.ll
1.03
141
1.13
-l.4t
M;
.05
.72
.53
.30
.04
.M
.93
.31
.63
.71
.51
.43
.11
.28
I.IS
-I.N -t.lt
3.4; -toi
4.11 -IN
-1.91 -3.38
.97E-04
.9IE-03
.96E-03
06! -04
.I7E-03
051-03
16! 03
201-04
.48E-04
65E-02
64! 02
.70E-03
.B2E-04
.29E-0!
.19E-03
.63E-R
.16E-II
.20! -04
1 72
-I.S7
1.13
1.73
I.N
1.41
1.N
2.95
3.IS
3 20
I.H
I.N
l.M
I.H
t.M
t.II
1.13
1.31
2 39
t.II
1.79
l.ll
1.00
phenol)
IhuMl)
Flym't tpprtxch (1990)
1990 By cquttlm Mjutted «ilu«
KMtlK90Cooptr(IN7| KMtln«ICao|xr(1987)
FlynTt rtlttlby Guy) Flynt't tfctilby BroMuoti)
(cm/hr) (virnr) (cn/nr) (cM/nr) (a*/hr|
-8.N t
-2 41 3
-2.00 1
l.ll
I.N
1 92
3 52
1.64
I.N
l.ll
I.S4
1.71
t.N
I.IS
-I.N
09E-07
33E-I3
OOE-Ot
23E-62
KE-Ot
tlE-Ot
03E-04
30E-Ot
l9E-Ot
47E-02
B6E-OI
1ZE-02
4IE-03
UE-tt
38E-N
-3.81
-320
-S.6I
-3 21
-2 65
-2 II
-4 45
-4.1;
-1.94
-3.N
-3 M
-44?
-S.2I
-4.11
-231
-1.47
-1 09
-2 17
-3.lt
-I.S3
-2. 65
-3.56
-1.03
-1.31
-142
-SSO
-3 71
-4.0;
-1.3;
YN
-1.04
-1.64
-I.S5
-I.3S
-3.50
-1.30
-I.SS
-I.S4
-LSI
-t 70
-I.H
-1 24
-i.t;
-1.20
-l.ll
-1.34
-l.M
-LSI
-1 SO
I.S5E-04
I.3IE-04
1.45E-N
I.I7E-04
I.4IE-01
1.45E-03
3.5SE-OS
S.37E-OS
I.ISE-Ot
1.09E-04
t.SIE-04
3.60E-OS
S.SOE-N
I.S2E-OS
4.90E-03
3.39E-K
I.I3E-R
2.141-03
1 201-04
I.ISE-OI
1.4IE-03
t.UE-04
I.I9E-OI
4071-62
3.80E-07
3.16E-N
1.66E-04
I.SIE-OS
4.I7E-R
t 51! 03
I.I2E-03
t 29E-02
1.62E-OI
4.47E-OI
3.16E-04
501E-OI
t.«E-tt
t.NE-K
I.7SE-K
t.OOE-03
1.I9E-OI
S.7SE-U
S37E-02
I.3IE-K
;.76E-tt
4.S7E-W
1.95E-tt
t 57! 02
3.1M-02
-3.00
-3.00
-3.00
-3.00
-3.N
-3.00
-3 00
-3.00
-3.00
-3.00
-3.00
-3.N
-3.00
-050
-1 50
-5.00
.OOE-03
.OOE-03
.OOE-03
.OOE-03
.OOE-03
.OOE-03
.OOE-03
.OOE-03
.OOE-03
.OOE-03
.OOE-03
.OOE-03
.NE-03
IM 01
.I6E-M
.OOE-OS
-3.00 1. OOE-03
-050 3.I6E-OI
-1.50 3.I6E-OI
-3.71
-3.17
-5 M
-3.31
-3.01
-t.77
-4.63
-4.45
-t.to
-3.N
-3.N
-4.70
-S.5S
-4.41
-t.N
-1 83
-I.SS
-3.14
-4.41
-I.N
-3 42
-4.11
-I.N
-1.14
-5 20
-4.41
-2/0
-4.13
-LSI
-I.N
-t.N
-t 34
-I.N
-in
-I.H
-I.N
-I.N
-1 95
-l.M
-I.N
-3.11
-1.01
-i.;t
-i.;;
-1 71
-I.IS
-I.N
-t.ss
-t.ti
-t.it
1 66E-04 -3.11
S.4IE-04 -2 76
t.lOE-N -4.39
4.36E-04 -215
966E-04 -263
I.68E-03 -t.4;
t.35E-OS -3.M
3.S6E-OS -3.51
12*03 -t.10
1.1IE-04 -3.21
133104 3.13
I.IIE-OS
t.ltE-N
1.13E-OS
t 21! 03
I.48E-K
t 64E-02
M4E-04
3 94E-05
1 251-02
1.62E-04
1901-05
t.OBE-II
I.14E-I1
I.37E-N
LITE-OS
I.02E-03
S.KE-OS
I.HE-K
I.ME-tt
I.IK-n
4.S4E-U
I.OK-tt
I.I9E-I1
3.71
4.3S
364
2 42
LI;
I 71
t 77
3 62
1.41
t.N
3.41
l.ll
I.It
4.11
3.M
t.s;
3.44
I.N
L34
t.s;
t.tl
I.IS
i.t;
1 631-01 -I.IS
I.29E 04 -321
t.OSE-M -l.ll
I.I2E-02 -I.N
I.Ht-tt -I.N
I.09E-K -Li;
f.lTE-64 -2.75
I08E-62 -1.31
I.NE-tt -1.13
I.70E-02 -I.H
t.OK-tt -l.ll
t.tn-ot -i 79
I.3IE-K -I.IS
4.7IE-D3 -t.tl
I.1IE-03 -l.ll
7.64E-03 -t.lt
7.971-04
I.68E-03
4 08t -OS
I.40E-03
1.36E-03
3.42E-03
I.I5E-04
2 57! 04
7 91! 03
S 25! 04
S.94E-04
I.6/E-04
4.46E-OS
t.30E-04
3.78E-03
l.34E-tt
I.NE-R
I.66E-03
1.38E-04
3.NE-R
L05E-03
1.30E-M
I.I4E-02
I.5IE-I1
I.10E-OS
I.96E-04
2 70E-03
3.60E-04
2 19! 02
4 54! 02
I 66E 03
1 231-03
1 lit 02
S.39E-R
7.I2E-M
S.S2E-04
7.69E-R
I.OK-tt
1. UE-tt
I.I7E-R
LNE-03
4.04E-R
I.49E-R
I.38E-R
I.S3E-R
I.6IE-R
I.13E-R
S.SIE-03
I.63E-03
7.6IE-03
Table A-4: Predicted K,, Values for Priority Pollutants
-------
ctcmcM. MHE
OCMICM.
cuss
CIESa. CHUM- Inn-ipKlflelKM
CKSOL. 4-CHURO-l- KM
MfWHOt, I- KM
OiaUROKUERt. 1.3- KM
OlaUMKRZEHE. 1.4- KM
iwwt KM
HEIHIt HftmntEttMTE KM
CtUKRTLEm. (m-«»cmc) KM
BROMDMm. t- KM
TiiotoraERZEflE. i.t.4- KM
BiRiiRoiouKRE. i.i- KM
DllinaiOlUEIK. 7,4- KM
IRIOtOROPKMI,, t.4,t- KM
Mm. 4.t-oiNiTn-i-HnMn.- KM
KU01C «CIO
Kiuaumncm.
ticHunKiim. i.t-
KIINHOMMTIKIKnEK
•anaimiu
oinorruHiK, HIITOO-
BICHlOmKniDIK. 3.1
IWI01«
oirwnuMtRE. n-Kiiiwo-
OlMinunDWIK, l.t-
Kt (nan-iMfilflel
ra-oi-owSifMtiin. 4-
Tan
MM
KM
KM
KM NKM
KM WON
iinosmiK
IIIUSMIK
am
mm.
MM.
MCMH
MKH
asm
MXEC
in
US
bo Rat
n-tlk*nol«
Jluwil
Tan(.»73
Bhcnolt
(hiMni
Cu|IHMlgr*tl.tHl
Mm.
arac
KIOUIR
oiuwm
isonmK
HtMO»aW8CWt«
rtMlHIM
IUWK.
DK
BOO
DOT
M.MIR
HtfUCHLO*
EKMI*
CW.ORDWE tiuirsl
CHLORDME ICISI
TOMMRE
DI-7-tlHTU€Hl 7WWLME
OIMCTHTI miWlATE
oiourn nnwuiE
DIETHa nnHMAlE
RIIROSMIK
IITROSINIK
IIIMSMIK
KB
KB
KB
KB
KSIICIK
KSTICIK
KSIICIK
KSIICIK
KSIICIK
KSIICIK
KSTICIK
KStlCiOt
KSTICIK
KSIICIK
KSIICIK
KSIICIK
KSIICIK
KSTICIK
KSTICIK
KSTICIK
mm. ESUR
mm. ESUR
mm. tstti
mrmi. ESU*
mi. «MM.
wit. HUH.
, _,
(OK/H-)
51-50-7
I1S-U-3
IS-M-I
S4I-71-I
M-M-I
M-71-3
IM-41-t
Itl-lt-l
ltl-14-t
Sl-tl-S
M-W-t
S34-U-I
SO-M-I
II-M-S
14-74-t
R-M-I
IW-I3-I
17J-1I-I
IM-41-7
M-87-S
M-10-1
nsi-H-i
1741-114
IMOI-M-I
I41-M-I
117-R-l
71-W-S
71-51-1
111-74-1
M-M-t
M-M-I
;j sjs
M-71-3
301-00-t
71-44-1
77-10-1
SIOl-M-t
5103-71-1
8001-35-7
ltl-M-7
M-M-t
M-S7-I
ts-os-o
11-tO-l
|4t.l
I4I.I
144.1
147
IM.t
ISM
IM.S
111 S
IR.I
114.1
117. S
1M.I
tst
IM.4
171
1*3.3
S3.I
M.I
147
IM. t
IM.t
IM.t
IM
3tt
Ml
M.I
IIS
IS4J
ni
3!!
3SS
385
373.5
311
411
411
414
130.1
in
381
m
Itl.t
ItSO.RS
111.1301
tlM.Blt
1MI.17I
l»S.tH
11.17117
44I.M1S
IS41.1Z5
IS.4MtS
14.17361
4M7.7M
I3I.RM
I311S4.3
714435.1
I14M.R
4NSI.M
I.I7R7I
0.537031
1454.708
11.87711
I7I.M3S
31W.J77
IMlt.R
HltSlf.
SI4M74.
1IIM.77
I.7S43M
IM
M.IIWt
194U1.7
I7M.RI
SJ41.I74
4»77I.I
130*57.3
ttlOMI.
IMIO.W
1I117.M
nsm.i
KOOS.14
tt.»OM7
tis.tto*
1*11301.
18307 .M
74.I3IR
IM5.7.8I
3.11
t.M
3.M
3.H
3.M
l.W
t.is
I.M
I.M
I.M
l.W
t.lt
S.W
S.M
4.11
4.M
i.ts
-It7
3.M
1.34
t.M
3.M
4.M
I.IS
l.7t
4.M
-111
t.M
1.71
S.31
3.13
1 7t
S.H
S.N
I.M
3.11
4 17
4.M
S.47
S.47
4.K
1.31
t.47
I.H
4.M
1.17
I.M
-5 71 I.3IE-M
-I.IS S.ltE-R
-1.3* 4.13E-R
-l.tl S.I4E-R
-1.11 4.0K-R
-l.tl S.ttE-R
-I.7S 1.77E-R
-S.M t. 771-14
-1.41
-I.M
-I.It
I.It
-t 71
-S.M
-t.is
-LSI
-s.st
-3.H
-1.11
-3.M
1.47
1.31
-tM
-3.15
-3.77
-111
-4.11
-t.M
-t.M
-S.M
-I.M
115
l.tt
-I.M
-S M
-I.M
-I.M
-i.ii
-l.W
-1.71
I.II
IM
I.M
-7.4*
-l.tl
-1.14
-0.03
-0.01
-O.M
-I'.OJ
1.34
-0.14
-1 M
-O.tl
3 IM 04
l.ME-OI
I.IK-OI
1.
I.
3.
3'.
3
1
4
t.
S9E-01
1K»M
3IE-03
I3E-04
IR-M
411-03
Ort-OJ
OtE-M
1K-M
SSE-R
I7E-04
1SEMH
t.tX«M
«
I
j
1
t
3
t
1
1
3
1
I
t
1
1
7
1
S
1
1
t
7
1
t
1
t
t
I7E-R
3K-II
RE-M
7K-M
7R-II
IK-IS
7SE-03
Mt-R
IR-M
SIE-01
47E«M
KE
-------
CNCMICAL MMC
OlMtTKTUKin. N-NUKKO-
FLUOtAHIHfllt
CWTStKt
MMOIBIFUnUIITWIE
BtBZO-»-PTPItP«
nncia(i.t.3-co)rmK
CHKItW
CUSS
mr. MOM.
WIT MON.
mi. NHM.
POtT. MOM.
mi. «MM.
mt. «MH.
PM.T. MOM.
CODE
n-«1k«noll phenol*
CASW HOICC LOS jhuwil (huninl flyrwi'i .ppnMch (1990) bitlnoJCooperllM;) KlstlngKaoptrllMr)
VT to. fa* llenllong.1973 GuytH>d]r«ft.l990 By equation* Mjmted viluej Flym't dit>|lnr Guy) Ftym'i dit>((nr Broniugh)
loo to to 109 Kp to log Kp to Log ^ to log to to loa Kp to
(cn/hr) (oWhr) |cWt>r) (tn/hr) (cx/hr) (oVhr)
IS-DI-I in.t ansj.sj 4 s; -o»
UI-ll-3 1M 38. JO/80 t.St
Z06-M-0 IK.
Z18-OI-9 t2t
S6-55-1 »l.
2DS-99-I ISI.
!9)-3t-S 171.
niis.oi 495
481317. S S.M
411317.5 S.66
I3304M. t.ll
3S370/I. t.S8
.94
55
II
.11
.(I
08
.171-01 -1 50 3.1SC-OI
.151-04
ezt oi
4M«00
.46C>N
zif-oo
?lt-OI
so
so
!so
.so
so
.16C-R
.IW-OJ
.ist-n
.11
.94
.39
.93
93
.itc-n t.it
.ISC-K LSI
5«'00 -I.M I.3n-OI
.1SC-03 -I./S I.7K-03
48t«00 -O.S3 I.97C-OI
4«.00 -O.II I.IOE-OI
.4K«00 -O.tl I.IK-OI
.Mt«OI -O.OS 9.00C-OI
.IH«OI O.It 1 Jlt'OO
Table A-4: Predicled Kp Values for Priority Pollutants
-------
APPENDIX B
GLOSSARY
-------
APPENDIX B - GLOSSARY
•y: An empirically deteimined parameter used in the Albery and Hadgraft model, equal to 1 for the transcellular route
and 5 for the intercellular route.
AC: Concentration difference of a compound across the dermal barrier.
4>: Peripheral blood flow, used in the equation to describe the relationship between the rates of capillary transfer
and diffusion.
t: Time duration of the shower or bath, in hours, in the Hall et al. (1989) model.
a: The area fraction used in the Albery and Hadgraft model, is equal to 1 for the transcellular route and 7xlO"6 for
the intercellular route.
Ability: An animate object's quality or competence to perform some function or activity, (see Capacity)
&bodr: I" determining the percutaneous absorption of a substance based on the mass-balance of the substance in the
body, the amount of the substance present in the body at any given time, (see ael)
ABS: Total absorbed dose (mg); used in conjunction with surface area, exposure concentration, and time to calculate
dermal absorption from vapors in humans. In the equation to calculate dermal uptake of a compound from soil, it
represents a unitless absorption factor.
Absorbed Dose: The amount of a substance penetrating across an absorption barrier (the exchange boundaries) of
an organism, via either physical or biological processes, after contact (exposure). Sometimes called internal dose.
Absorption Factor: The relative amount of a substance on the skin that penetrates through the epidermis into the
body; reported as the unitless fraction of the applied dose or as the percent absorbed.
Absorption Fraction: Percent absorbed; percent of applied amount that penetrates per unit time based on a specific
contact time.
Administered Dose: The amount of a substance given to a human or test animal in determining dose-response
relationships, especially through skin absorption, ingestion or inhalation, (see Applied Dose)
Adsorption: Adherence of a solid or liquid to a surface without penetrating through the surface layer. Also, the
adherence of ions or organic chemicals onto the surface layer of other materials without being incorporated into or
absorbed by the surface of the material.
Atl: In determining the percutaneous absorption of a substance based on the mass-balance of the substance in the
body, the amount of the substance eliminated from the body with time, (see a^)
AF: Soil-to-skin adherence factor (mg/cm2-day).
AL: Area fraction of the lipophilic pathway. In the Two-Parallel-Pathway model, this value is equal to 0.9, and in
the Three-Parallel-Pathway model, it is equal to 0.5. (see A,,)
Ap: Area fraction of the polar pathway, equal to 0.1 in both the Two-Parallel and Three-Parallel-Pathway models.
(seeAJ
B-l
-------
Applied Dose: The amount of a substance given to a human or test animal in determining dose-response
relationships, especially through skin contact, (see Administered Dose)
Aqueous: Relating to water or substances dissolved or suspended in water, not to be confused with other liquid
solutions or suspensions not containing water.
AT: Averaging time: a pathway-specific period of exposure for non-carcinogenic effects (i.e., ET x EF x ED x 365
days/year), and a 70-year lifetime for carcinogenic effects (i.e., 70 years x 365 days/year).
AUC: Area under the curve; one of the methods used to calculate the mean bioavailability of a substance. AUC
can be representative of the plasma concentration-time profile or the total radioactivity in the plasma.
p: In the Kasting et al. (1987) model, a property of the skin which is temperature dependent and, with the van der
Waals volume, is used to define the volume dependence of the diffusion coefficient of the stratum corneum.
B: A constant used to calculate the diffusivity of a substance in the lipid phase (D^ and protein phase (Dp) in the
Two-Parallel and Three-Parallel-Pathway models. In both models, B = BL = Bp = 0.016.
Bioavailability: The state of being capable of being absorbed and available to interact with the metabolic processes
of an organism. Bioavailability is typically a function of chemical properties, physical state of the material to which
an organism is exposed, and the ability of the individual organism to physiologically take up a chemical.
Body Burden: The amount of a particular chemical stored in the body at a particular time, especially a potentially
toxic chemical in the body as a result of exposure. Body burdens can be the result of long term or short term
storage; for example, the amount of a metal in bone, the amount of a lipophilic substance such as PCB in adipose
tissue, or the amount of carbon monoxide (as carboxyhemoglobin) in the blood.
BW: Body weight (kg).
C: Exposure concentration (mg/cm3); also concentration of a compound of interest (mg/mL). When used to
determine the relationship between the rates of capillary transfer and diffusion, the concentration of the diffusing
compound in tissue adjacent to the capillary walls.
C«: The initial aqueous concentration of the compound, (see Cc)
Ct: In in vitro diffusion cell studies, the concentration of a substance in the donor cell, (see C2)
C2: In in vitro diffusion cell studies, the concentration of a substance in the receptor side, (see C,)
C.: The concentration of a diffusing compound in arterial blood. This is one of the factors used to determine the
relationship between the rates of capillary transfer and diffusion.
Calculate: Determine by mathematical process.
Calculation: A numerical value determined by a mathematical process.
Capacity: An inanimate object's suitability or potential to be able to undergo some activity, function or reaction.
(see Ability)
Cfa: Concentration of a compound in the blood.
B-2
-------
Ct: Equilibrium concentration of a compound in solution after exposure to a known mass of skin, (see CJ
CF: In the equation used to calculate dermal uptake of a compound from soil, this represents a weight conversion
(10"* kg/mg); or the volumetric conversion for water (1 L/1000 cm3); or any conversion factor.
CLb: Clearance of a compound from the blood.
Contaminant: Something that corrupts, infects, makes impure, unsafe or unfit for use.
CS: Chemical concentration in soil (mg/kg).
C^: Steady state concentration, for example in the donor solution.
Cgs^: Steady-state concentration of a compound in the blood.
Cutaneous: Of, relating to, or affecting the skin.
CT: In the Kasting et al. (1987) model, the vehicle concentration, and the common practice is to measure flux as
a function of this parameter and to report the ratio as the permeability coefficient.
Cw: Water solubility of a substance (mg/mL); the amount of the chemical per unit volume present in a fully
saturated aqueous solution. This parameter is used in the Heterogenous Structural. Two-Parallel-Pathways, Three-
Parallel-Pathways, and Albery and Hadgraft models to calculate flux (J).
CW: Chemical concentration in water (mg/L); the aqueous concentration tested.
D: When used to determine the relationship between the rates of capillary transfer and diffusion, the average
membrane diffusion coefficient.
D(: Diffusion coefficient for a compound of interest. This coefficient may be used in conjunction with Db in the
absence of permeability data to approximate Dm.
DA: In the Albery and Hadgraft model, the diffusion coefficient for the stratum comeum, and this value is equal
to 1.9xlff* cm2/hr for the transcellular route and 9.7x10^ cm2/hr for the intercellular route. Also in the Guy et al.
(1985a) model, the diffusion coefficient through the stratum comeum of A, a structurally analogous compound.
Db: The diffusion coefficient of a compound structurally analogous to a compound of interest, (see D.)
Dermal Absorption: The process by which a substance is transported across the skin permeability surface barrier
and taken up into the living tissue of the body; generally synonymous with percutaneous absorption and with dermal
uptake.
Dermal Adherence Capacity: The maximum amount of a specified matrix that can be contained on the skin.
Dermal Adsorption: The process by which materials come in contact with the skin surface, but are then retained
and adhere to the permeability barrier without being taken into the body; generally synonymous with skin adherence.
Dermally Absorbed Dose: The amount of the applied material (the dose) which becomes absorbed into the body.
Dermal Exposure: Contact with the skin by any medium containing chemicals, quantified as the amount on the
skin and available for adsorption and possible absorption.
B-3
-------
Dennis: The highly vascularized inner mesodermic layer of the skin, about 500 to 3,000 urn thick, which is a
collagenous, hydrous layer and contains the outermost nerve endings of the skin.
Diffusion Cell: For in vitro skin penetration testing, any system of chambers for holding test materials, between
the two compartments of which a section of skin or its membrane components is stretched to assess the transport
of a chemical or chemicals from the donor side to the receptor side to measure the flux of the chemical(s) across
the skin or its membrane components.
DL/Dp: The ratio of the diffusivities of the penetrant in the lipid and protein phases, empirically given the value of
2.0x10~3 in the Heterogeneous Structural Model. Alternatively, DL and D_ values can be estimated using equations,
DL = DL(O) x e'BLM and Dp = Dp(0) x e'81*1. (see B and M)
DL: Diffusivity of a substance in the lipid phase, (see Dp)
DL<0): A constant used in the Two-Parallel-Pathway and Three-Parallel-Pathway models, and having a value of
1.7xl(Ts cm2/hr.
Dm: Diffusion coefficient for molecules within the membrane. This coefficient is needed to calculate the
permeability constant (K,,) and/or flux (J).
Dose: The amount of a substance available for interaction with metabolic processes of an organism following
exposure and absorption into the organism. The amount of a substance crossing the exchange boundaries of skin,
lungs, or digestive tract is termed absorbed dose, while the amount available for interaction by any particular organ
or cell is termed the delivered dose for that organ or cell. Theoretically, the sum of the delivered doses plus the
metabolic transformations should equal absorbed dose. (The terms administered dose and applied dose refer to
amounts of a substance made available for absorption, and therefore are measures of exposure rather than uptake.
As such, these terms, sometimes found in the literature, are somewhat confusing and may need to be defined when
used.)
Dp: Diffusivity of a substance in the protein phase, (see D^
Donor: Something which gives up some of itself or one of its components; in a diffusion cell, that material (neat
or test compound in a vehicle) placed on the skin or its membrane components, from which the loss of the test
compound is measured over time to develop an estimated flux value.
Dp<0): A constant used in the Two-Parallel-Pathway and Three-Parallel-Pathway models, and having a value of
3.8xl(T5 cm2/hr.
e: The base of the natural (Naperian) system of logarithms, an irrational number whose approximate value is
2.71828.
ED: Exposure duration (years).
EF: Exposure frequency (days or events/year).
Electrolyte: A substance that will provide ionic conductivity when dissolved in water or when in contact with it:
such compounds may be either solid or liquid. A substance that wilt ionize in solution and increase the conductivity
of the aqueous medium.
Environmental Pollutant: Any entity which contaminates any ambient media, including surface water, groundwater,
soil, or air.
B-4
-------
Epidermis: The outer mesodermic layer of the skin; a non-vascular layer about 100 um thick, with the outermost
layer, the stratum comeum of about 10 to 40 urn thickness composed of dead, partially desiccated and keratinized
epidermal cells; below the stratum comeum lies the stratum germinativum, or viable epidermis, a layer about SO to
100 urn thick composed of rapidly proliferating nucleated cells, generating about one new cell layer per day, resulting
in the stratum comeum becoming totally replaced once every 2 to 3 weeks.
Estimate: To determine a numerical value that is an approximate value of the size or extent, either arrived at by
rough calculation or by best professional judgement using available scientific information. Also a noun, (see
Estimation)
Estimation: A numerical value that is a rough determination of size, extent.
ET: Exposure time (hours/day).
Exposure: Contact of a chemical, physical, or biological agent with the outer boundary of an organism. Quantified
as the concentration of the agent in the medium in contact with the organism and available for absorption integrated
over the time duration of that contact.
Fat/Air Partition Coefficient: The relationship between lipid solubility (and thus permeability into skin) and vapor
phase transport (and thus relative volatilization rate and vapor pressure) for organic compounds. A compound with
a high fat/air partition coefficient would have a high lipid solubility and/or low potential to enter into the vapor
phase, while a compound with a low fat/air partition coefficient would have a low lipid solubility and/or a high
potential to volatilize and remain in the vapor phase. Compare with octanol/water partition coefficient.
Pick's Law of Diffusion: In theory, the transdermal flux of a compound in proportion to the concentration
difference of the compound (AC) across the dermal barrier, represented by J = Kp x AC.
Flux (J): Amount of chemical absorbed across a defined surface area of the skin per unit time (mg/cm2/hr). This
is equal to the dermal permeability coefficient multiplied by the concentration of the chemical. Flux and
concentration are interdependent.
Flux: Amount penetrating per unit area per unit time (mg/cm2/hr) = dermal permeability coefficient times
concentration (flux and concentration are interdependent) = J = Kp x AC.
h: The thickness of the membrane being tested.
Hydrophilic: Literally "water loving"; the property of a chemical to have a strong tendency to bind or absorb water.
Hydrophobic: Literally "water hating"; the property of a chemical to be antagonistic to water or incapable of
dissolving in water (and for many organic chemicals, to be soluble in fats and oils, or non-polar solvents).
In Vitro: Literally "in glass"; a situation in which an experiment is carried out in a vessel (glass container, test tube,
beaker, or diffusion cell) using excised tissues.
In Vivo: Literally "in life"; a situation in which an experiment is carried out using living, intact organisms.
IPPSF: Isolated perfused porcine skin flap; an anatomically intact, viable, isolated perfused tubed-skin flap
preparation for determining in vitro dermal absorption rates. This system was developed to overcome the potential
limitations posed by relatively artificial in vitro systems.
Jm: In the Kasting et al. (1987) model, the maximum flux which is easily obtained by measuring the transport from
B-5
-------
saturated solutions.
Ju: Steady-state flux, i.e., J plotted as a function of time and showing little change over time.
k.j: The interfacial transfer constant used in the Albery and Hadgraft model; equal to 3.6 cm/hr.
Kg,: The stratum corneum/vehicle partition coefficient This constant is determined by dividing the solubility of
the chemical in the stratum comeum by its solubility in the vehicle.
Kn,,: In the Kasting et al. (1987) model, the membrane-vehicle partition coefficient.
KM: Organic carbon/water partition coefficient. The higher the K^., the more likely a chemical is to bind to the
organic carbon in the soil or sediment than to remain in solution.
K.^ Octanol/water partition coefficient, (see P)
Kp: A flux value, normalized for concentration, that represents the rate at which a chemical crosses the stratum
comeum (cm/hr). This constant is generally determined at steady state or near steady state and infinite dose, and
varies with thickness of the skin, specifically the stratum corneum.
Kp*: Chemical-specific dermal permeability constant (cm/hr) - distance travelled through skin (constant is
determined at steady state and infinite dose and varies with thickness).
1: The thickness of the stratum comeum (pin). The range of human values is 10 urn on the back to 400 urn on the
soles of the feet, and 15 to 40 pm on the abdomen. When used to determine the relationship between the rates of
capillary transfer and diffusion, it represents the thickness (cm) of the layer beneath the stratum comeum.
L: Liter. Also, the thickness of a layer, such as the layer of the epidermis below the stratum corneum.
Lipophilic: Literally "lipid loving"; the property of a chemical to have a strong affinity for lipids, fats, or oils; or
being highly soluble in nonpolar organic solvents.
Lipophobic: Literally "lipid hating"; the property of a chemical to be antagonistic to lipids or incapable of
dissolving in or dispersing uniformly in fats, oils, or nonpolar organic solvents.
M: Molecular weight
Matrix: The material or medium in which something is enclosed, embedded, dispersed or dissolved.
Medium (pi. media): Any one of the basic categories of material surrounding or contacting an organism (e.g.,
outdoor air, indoor air, water, soil, sediments) through which chemicals or pollutants can move and reach the
organism.
MWA: The molecular weight of a structurally analogous compound.
MWM: The molecular weight of the compound of interest.
Neat: Pure material, undiluted, free from admixture.
P: The lipid/protein partition coefficient of a substance. This barrier function is treated as a dispersion of
hydrophilic protein in a continuous non-polar matrix through which the substance migrates by dissolution and Fickian
diffusion. Although most correlation studies have used octanol/water partition coefficients (K^), it is not clear that
B-6
-------
this is the ideal solvent system for modeling all interactions of organic compounds with biologic systems.
Partition Coefficient: The vehicle-specific relationship of a chemical and its relative presence within two competing
media (such as between water and n-octanol in the octanol/water partition coefficient); expressed as the ratio of the
amounts in each medium at steady state.
PB: Physiologically-based models to describe the dermal uptake of chemical vapors, (see PK)
Percutaneous: Performed or effected through the skin.
Permeability Coefficient: (see Kp)
Permeable: Penetrable; capable of permitting materials (liquids, gases, dissolved chemicals) to pass through (a
permeable membrane).
Phannacokinetics: The study of the time course of absorption, distribution, metabolism, and excretion of a foreign
substance (e.g., a drug or pollutant) in an organism.
Physiologically-Based Pharmacokinetic Modeling (PBPK): For dermal exposure testing. Use of models to
estimate the dermal permeability constant and amounts absorbed by the best-fit method based on blood concentration-
time profile data or by monitoring the appearance and amounts of metabolites in post-exposure urine samples or by
measuring the concentration of parent compound in the expired air.
PK: Pharmacokinetic models to describe the dermal uptake of compounds in aqueous media, (see PB)
pKt: Negative logarithm of the equilibrium coefficient of an acid; also, the hydrolysis dissociation constant for a
chemical substance.
Polarity: The quality or condition inherent in something (a body, electric cell, membrane, chemical, etc.) that
exhibits opposing properties or powers in opposite parts or directions; a polar molecule has positive and negative
electrical charges which are permanently separated, and polar molecules ionize in solution and impart electrical
conductivity; polar membranes have opposite (or differing) charges on the inside and outside of the membranes.
Qj,: Alveolar ventilation rate, used to calculate the skin uptake ratio of chemical vapors when no respiratory
protection is used.
Receptor Fluid: The liquid which receives some material; in a diffusion cell, the liquid in the compartment opposite
the donor cell, and which started with none or very little of the test material and receives the material transporting
across the whole skin or epidermis, to permit measurements of the flux of the material, (see Donor)
RfD: Reference Dose; an estimate (uncertainty spanning perhaps an order of magnitude) of a daily exposure (mg/kg-
day) to the general human population (including sensitive subgroups) that is likely to be without an appreciable risk
of deleterious effects during a lifetime exposure (chronic RfD) or exposure during a limited time interval (subchronic
RfD).
Route: The pathway a chemical or pollutant enters an organism after contact, e.g., by ingestion, inhalation, or
dermal absorption.
Routes or exposure: The pathways of chemical contact to organisms.
S: Solubility of a compound (nM) at 25°C. This value is used in the equation, log J = 1.41 x log S - 0297, to
calculate flux.
B-7
-------
SA: Surface area (abbreviation); the skin surface area available for contact (cm2), (see Surface Area)
Scenario: A set of facts, assumptions, and inferences about how a process takes place that aids the assessor in
evaluating, estimating, or quantifying the process. The scenario method is a method used in predictive assessments
to account for variability in parameters by setting up one or more alternative sets of assumptions (scenarios),
evaluating each of the resulting scenarios, and then using the resulting estimate(s) as being illustrative of the actual
conditions being evaluated. In the case of a single scenario, a sensitivity analysis is usually performed on the
parameters to evaluate the variability.
Skin Adherence: The property of a material which causes it to be retained on the surface of the epidermis (adheres
to the skin).
Steady State: The status or condition of a system or process that has reached an equilibrium over time and then
does not change, or a condition that changes only negligibly over a specified time.
Stratum Corneum: The outermost layer of the skin (see Epidermis) composed of dead, partially desiccated and
keratinized epidermal cells; thought to provide the major resistance to the absorption into the circulation of substances
that are deposited on the skin.
Surface Area: The number of unit squares equal in measure to the exterior, upper, or outer boundary of an object
or body.
t: The time required after initial contact with the skin for a compound to accumulate in the stratum comeum and
reach a maximum equilibrium concentration. This lag time is represented by the equation L2/6Dm, where L is the
thickness of the stratum comeum and Dm is a diffusion coefficient.
T: The duration of exposure (hours).
tflj: Half-life; in in vitro studies which evaluate the distribution and fate of a chemical, t^ refers to the period of
time in which 50 percent of the test substance is hydrolyzed from the media. In in vivo studies, this term represents
either the plasma elimination or surface loss half life.
V: Molecular volume; van der Waals volume of the permeant.
V,: In in vitro diffusion cell studies, this represents the volume of the donor cell.
V2: In in vitro diffusion cell studies, this represents the volume of the receptor cell.
V,q: Solution phase volume, one of the parameters necessary to estimate K,,, by measuring the difference between
the initial aqueous concentration of a compound and the concentration in the solution after exposure to a known mass
of skin. (seeW,)
YD: Volume of distribution; one of the parameters necessary to calculate the body burden of a compound.
VDU: Steady-state volume of distribution.
Wt: Weight of skin tissue. This is also one of the parameters necessary to estimate K^ by measuring the difference
between the initial aqueous concentration of a compound and the concentration in the solution after exposure to a
known mass of skin, (see V(q)
x: Distance, as in dc/dx which is the change in concentration over the unit distance.
B-8
------- |