United States
Environmental Protection
Agency
TOXICOLOGY HANDBOOK
Government Institutes, Inc.
-------
United States
Environmental Protection
Agency
EPA TOXICOLOGY HANDBOOK
Government Institutes, Inc
-------
Government Institutes, Inc., Rockville, Maryland 20850
Copyright © 1986 by Government Institutes. All rights reserved.
Published September 1986. Seventh printing January 1994.
No part of this work, except for material in the public domain, may be reproduced
or transmitted in any form or by any means, electronic or mechanical, including
photocopying, recording, or any information storage and retrieval system, without
permission in writing from the publisher. All requests for permission to reproduce
material from this work should be directed to Government Institutes, Inc., 4
Research Place, Suite 200, Rockville, Maryland 20850.
The publisher makes no representation or warranty, express or implied, as to the
completeness, correctness, or utility of the information in this publication. In
addition, the publisher assumes no liability of any kind whatsoever resulting from
the use of or reliance upon the contents of this book.
EPA Disclaimer
This document has not completed final review within EPA, and is for internal
Agency use/distribution only. There has been an EPA workgroup review of the
development of this handbook prior to this draft.
ISBN: 0-86587-142-6
Printed in the United States of America
-------
TABLE OF CONTENTS
PAGE
LIST OF FIGURES viii
LISf OF TABLES ix
LIST OF ACRONYMS . x
GLOSSARY . . xii
EXECUTIVE SUMMARY . ES-1
1.0 INTRODUCTION 1-1
1.1 Purpose and Scope of Che Toxicology Handbook 1-1
1.2 Handbook Organization 1-1
2.0 FUNDAMENTAL CONCEPTS IN TOXICOLOGY i 2-1
2.1 Mechanisms of Chemical Toxicity 2-1
2.2 Essential Toxicological Information 2-3
2.3 Limitations- to Toxicological Knowledge- 2-5
2.4 Exposure to Toxic Chemicals 2-6
3.0 DOSE-RESPONSE RELATIONSHIPS .... 3-1
3.1 Identification of No-Effect Levels From Dose-Response
Curves 3-1
3.1.1 Chemicals with Thresholds 3-1
3.1.2 Determination of Threshold Values '. 3-1
3.1.3 Chemicals Without Thresholds 3-5
3.2 Other Uses of Dose-Response Curves . 3-9
4.0 IMPORTANT PARAMETERS IN TOXICITY ASSESSMENTS 4-1
4.1 Route of Exposure 4-1
4.1.1 Ingestion . 4-1
4.1.2 Inhalation 4-3
4.1.3 Dermal Absorption . 4-3
4.1.4 Exposure by Injection 4-4
4.2 Duration/Frequency of Exposure 4-4
4.3 Species of Exposed Organism 4-4
contiviueQ-
iii
-------
Table of Concents - continued
PAGE
4.3.1 Absorption Differences Among Species 4-5
4.3.2 Differences In Chemical Metabolism Among
Species 4-5
4.4 Individual Characteristics of Exposed Organisms 4-5
4.4.1 Genetic Traits 4-5
4.4.2 Sex and Hormonal Status 4-6
4.4.3 Nutritional Status and Dietary Factors 4-6
4.4.4 Age and Maturity 4-6
4.5 Toxicological Endpoints 4-7
4.5.1 Criteria for Selection of Endpoints 4-7
4.5.2 Common Toxicological Endpoints and Measuring
Techniques 4-8
4.6 Key Guidance and Implementation Documents 4-13
4.7 Background References . . 4-13
5.0 TYPICAL PROTOCOLS USED IN TOXICOLOGICAL STUDIES 5-1
5.1 Studies in Laboratory Animals 5-1
5.1.1 Acute Studies 5-1
5.1.2 Subchronic Studies 5-4
5.1.3 Chronic Studies 5-4
5.2 Clinical Studies in Humans 5-*
5.3 Epidemiological Studies 5-5
5.3.1 Types of Epidemiological Studies 5-5
5.3.2 Uncertainties and Limitations Associated with
Epidemiological Data .* . . . 5-8
5.4 Ecotoxlcological Studies 5-10
5.4.1 Bloassays . 5-10
5.4.2 An Example of an Acute Toxicity Study on Fish . . . 5-11
5.4.3 Limitations of Ecotoxlcological Studies 5-14
5.5 Key Guidance and Implementation Documents 5-14
5.6 Background References 5-14
6.0 EXTRAPOLATION OF TOXICOLOGICAL DATA FROM ANIMALS TO HUMANS ... 6-1
contlnued-
iv
-------
Table of Contents - continued
PAGE
6.1 Selection of Appropriate Studies 6-1
6.2 Conversion of Dose Levels 6-2
6.3 Correction for Toxicokinetic Differences 6-6
6.A High-Dose to Low-Dose Extrapolation 6-8
6.4.1 Dose-Response Models 6-10
6.A.2 Threshold Versus Nonthreshold Models 6-12
6.4.3 Toxicokinetic Considerations in High-to-Low Dose
Extrapolation 6-12
6.5 Sources of Uncertainty 6-13
6.5.1 Sources of Uncertainty in Extrapolation frcni
Animal Studies 6-13
6.5.2 Sources of Uncertainty in Epidemiological
Studies 6-14
6.6 Approaches to Dealing with Uncertainty 6-15
6.7 Key Guidance and Implementation Documents 6-17
6.8 Background References ." 6-17
7.0 EXPOSURE ASSESSMENT 7-1
7.1 Collection of Occurrence Data .7-1
7.1.1 Site History 7-1
7.1.2 Sampling and Analysis of Environmental Data .... 7-3
7.1.3 Legal Requirements 7-10
7.2 Identification and Analysis of Exposed Populations .... 7-12
7.2.1 Population Identification and Enumeration 7-13
7.2.2 Population Characterization 7-13
7.2.3 Activity Analysis : 7-14
7.2.4 Development of Exposure Coefficients . 7-14
7.2.5 Nonhuman Population Analyses ... 7-14
7.3 Calculation of Exposure . 7-16
7.4 Estimation of Past Exposure 7-17
7.5 Prediction of Future Exposure 7-18
7.5.1 Environmental Stability 7-18
7.5.2 Environmental Mobility 7-22
continued-
v
-------
Table of Contents - continued
PACE
7.5.3 Environmental Partitioning 7-25
7.5.4. Population Predictions 7-26
7.6 Key Guidance and Implementation Documents 7-26
7.7 Background References 7-28
8.0 RISK ASSESSMENTS 8-1
8.1 Risk Assessment Process 8-1
8.1.1 Toxicological Evaluation 8-2
8.1.2 Dose-Response Evaluation 8-2
8.1.3 Exposure Assessment 8-2
8.1.4 Risk Characterization 8-4
8.2 Risk Characterization at Hazardous Waste Sites 8-4
8.2.1 Characterization of Noncarclnogenic Risks 8-4
8.2.2 Characterization of Carcinogenic Risks 8-5
8.3 Application of Risk Assessment at Hazardous Waste Sites . . 8-6
8.4 Key Guidance and Implementation Documents 8-6
8.5 Background References . 8-7
9.0 SUMMARY OF TOXICOLOGICAL INFORMATION ON DIGXINS *u
-------
Table of Contents - continued
PAGE
10.2.1 Noncarclnogenlc Studies : . . . . 10-2
10.2.2 Mutagenic and Carcinogenic Studies 10-3
10.3 Quantitative Indices of Toxicity 10-7
10.3.1 Noncarclnogenlc Effects 10-7
10.3.2 Carcinogenic Effects 10-9
10.4 Special Concerns 10-9
10.5 References 10-9
11.0 SUMMARY OF T0X1C0L0GICAL INFORMATION ON LEAD 11-1
11.1 Chemical Properties and Environmental Stability 11-1
11.2 Summary of Health Effects Data 11-1
11.2.1 Noncarclnogenlc Studies 11-2
11.2.2 Mutagenic and Carcinogenic Studies
11.3 Quantitative Indices of Toxicity 11-6
11.4 Special Concerns 11-6
11.5 References 11-7
T t X-L
vii
-------
LIST OF FIGURES
FIGURE PAGE
ES-1 Overview of Che Risk Assessment Process Applied Co a
Hazardous Waste Site ES-2
ES-2 Hypothetical Dose-Response Curve wich a Threshold " ES-4
ES-3 Hypothetical Dose-Response Curve with no Threshold ; . . . . ES-5
2-1 Overview of the Risk Assessment Process Applied Co a
Hazardous Waste Site 2-2
2-2 Examples of Mechanisms by Which Two Chemicals Alter a Cell
Ular Funcclon . . 2-4
3-1 Hypothetical Dose-Response Curve 3-2
3-2 Cellular Defense Mechanisms as a Basis for the Threshold
Phenomenon 3-3
3-3 Tissue Reserve as a Basis for the Threshold Phenomenon . . . 3-4
3-4 Hypothetical Dose-Response Curve (Logarithmic Scale) for
Nitrite-Induced Hemoglobin Oxidation . 3-7
3-5 Hypothetical Relationship Between Risk and Dose for Chemicals
with no Threshold 3-8
3-6 Hypothetical Dose-Response Curve of Two Chemicals, A
and B . . . . 3-11
3-7 Hypothetical Dose-Response Curves for Chemicals with
Beneficial and Adverse Effects .. 3-13
4-1 Factors Considered During Toxicity Assessments ~ • 4-2
5-1 Design of Typical Retrospective (Case-Control) Study .... 5-6
5-2 Design of Typical Prospective (Cohort) Study 5-7
5-3 ' Design of Typical Prevalence (Cros3-Sectional) Study .... 5-9
6-1 Illustration of Different Possible Shapes of the Dose-
Response Curve in the Low Dose Range 6-9
6-2 Dose-Response Curve Extrapolations by Six Mathematical
Models 6-11
8-1 Risk Assessment Process at Hazardous Wastes Sites 8-3
viii
-------
LIST OF TABLES
TABLE PAGE
3-1 Hypothetical Results of an Experiment to Define a Dose-
Response Curve for Nitrite-Induced Hemoglobin Oxidation . . . 3-6
3-2 Quantitative Indices Derived from Dose-Response Curves ... 3-10
4-1 Measurement of Common Toxicological Endpoints 4-9
5-1 Typical Testing Protocols in Animals 5-2
5-2 Ecotoxicological Tests for Which EPA has Developed
Guidelines 5-12
6-1 Statistical Terms and Their Use 6-3
6-2 Statistical Analysis of Hypothetical Data 6-4
6-3 Comparison of Dose Conversions Using Surface Area and Weight
Equivalence 6-5
6-4 Dose Conversion Factors 6-7
6-5 Guidelines for Selection of Uncertainty Factors 6-16
7-1 Checklist of Important Site History Information 7-2
7-2 Chemicals Frequently Occurring at Hazardous \Jaste Sites for
Which Sampling Data May be Required 7-4
7-3 Standard Intake Assumptions for Humans 7-15
7-4 Physical and Chemical Properties That May Influence
Environmental Fate . • . . . 7-19
7-5 Site Characteristics That May Influence Environmental
Fate 7-20
7-6 Environmental Persistence of Some Compounds That May Occur at
Hazardous Waste Sites 7-23
9-1 Acute Toxicities of Various Chlorodioxins by Oral
Exposure 9-3
9-2 Acute Lethality of TCDD 9-4
9-3 Effects of TCDD in Animals Following Acute Exposure 9-5
9-4 Effects of TCDF in Animals Following Acute Exposure 9-7
9-5 NOAEL and LOAEL Values Obtained from Subchronic and Chronic
Oral Toxicity Studies of 2,3,7,8-TCDD 9-8
9-6 Summary of Carcinogenic Effects of TCDD 9-12
10-1 Carcinogenicity of TCE 10-4
10-2 Risk Estimates for TCE 10-8
ix
-------
LIST OF ACRONYMS
ACGIH
American Conference of Governmental
Industrial Hygienists
ADI
Acceptable Daily Intake
AL
Acceptable Level
ALA
Delta-Aminolevulinic Acid
ALA-D
Delta-Aminolevulinic Acid Dehydratase
CAS
Chemical Abstract Service
CDC
Centers for Disease Control
CERCLA
Comprehensive Environmental Response,
Compensation and Liability Act of 1980
CHO
Chinese Hamster Ovary
CNS
Central Nervous System
DNA
Deoxyribonucleic Acid
EC
Effective Concentration
ED
Effective Dose
EEG
Electroencephalogram
EL
Exposure Level
EPA
United States Environmental Protection
Agency
FDA
U.S. Food and Drug Administration
FEL
Frank Effect Level
FR
Federal Register
IARC
International Agency for Research on
Cancer
ICAIR
Interdisciplinary Planning and Information
Research
HI
Hazard Index
IQ
Intelligence Quotient
IRCP
International Commission on Radiological
Protection
LC
Lethal Concentration
LD
Lethal Dose
LOAEL
Lowest Observed Adverse Effect Level
LOEL
Lowest Observed Effect Level
MATC
Maximum Allowable Toxicant Concentration
MCL
Maximum Contaminant Level
'MTD
Maximum Tolerated Dose
NAS
National Academy of Science
NCI
National Cancer Institute
NCV
Nerve Conduction Velocity
NHANES
National Health and Nutrition Examination
Survey
NIOSH
National Institute of Occupational Safety
and Health
NOAEL
No Observed Adverse Effect Level
NOEL
No Observed Effect Level
-.crs tirvued
-------
List of Acronyms - continued
NTP National Toxicology Program
OERR Office of Emergency and Remedial Response
OSU Office of Solid Waste
OWPE Office of Waste Programs Enforcement
PCDF Polychlorlnated-dlbenzofuran
PND Post Natal Day
QA Quality Assurance
QC Quality Control
RCRA Resource Conservation and Recovery Act of
1976
SD Standard Deviation
SEM Standard Error of the Means
SNARL Suggested No Adverse Response Level
STEL Short-Term Exposure Limit
SW Slow Wave
TI Therapeutic Index
TC Toxic Concentration
TCDD Tetrachlorodibenzo-p-dioxin
TCDF Tetrachlorodibenzo-fcran
TD Toxic Dose
TLV Threshold Limit Value
VSD Virtually Safe Dose
WHO World Health Organization
xi
-------
GLOSSARY
Abiotic - Nonliving, especially the nonliving elements in ecological systems.
Absorbed dose - The amount of a chemical that enters the body of an exposed
organism.
Absorption - The uptake of water or dissolved chemicals by a cell or an
organism.
Absorption factor - The fraction of a chemical making contact with an organism
that is absorbed by the organism.
Active transport - An energy-expending mechansim by which a cell moves a
chemical across the cell membrane from a point of lower concentration to
a point of higher concentration, against the diffusion gradient.
Acute - Occurring over a short period of time; used to describe brief
exposures and effects which appear promptly after exposure.
Additive Effect - Combined effect of two or more chemicals equal to the sum
of their individual effects.
Adsorption - The process by which chemicals are held on the surface of a
mineral or soil particle. Compare with absorption.
Ambient - Environmental or surrounding conditions.
Animal studies - Investigations using animals as surrogates for humans, on cne
expectation that results in animals are pertinent to humans.
Antagonism - Interference or inhibition of the effect of one chemical by the
action of another chemical.
Assay - A test for a particular chemical or effect.
Background level - Normal, ambient environmental concentration of a chemical.
Bias - An inadequacy in experimental design that leads to results or
conclusions not representative of the population under study.
Bioaccumulation - The retention and concentration of a substance by an
organism.
Bioassav - Test which determines the effect of a chemical on a living
organism.
Bioconcentration - The accumulation of a chemical in tissues of an organism
(such as fish) to levels that are greater than the level in the medium
(such as water) in which the organism resides (see bioaccumulation).
xii
-------
Biodegradation - Decomposition of a substance into more elementary compounds
by the action of microorganisms such as bacteria.
Biomagnification - The serial accumulation of a chemical by organisms in the
food chain, vith higher concentrations of the substance in each
succeeding trophic level.
.Biotransformation - Conversion of a substance into other compounds by
organisms; includes biodegradation.
Cancer - A -disease characterized by the rapid and uncontrolled growth of
aberrent cells into malignant tumors.
Carcinogen - A chemical which causes or induces cancer.
CAS registration number - A number assigned by the Chemical Abstracts Service
to identify a chemical.
Central nervous system - Portion of the nervous system which consists of the
brain and spinal cord; CNS.
Chromosome - Rodlike structure in the nucleus of a cell that forms during
' mitosis; composed of DNA and protein; chromosomes contain the genes
responsible for heredity.
Chronic - Occurring over a long period of time, either continuously or
intermittently; used to describe ongoing exposures and effects that
develop only after a long exposure.
Chronic exposure - Long-term, low level exposure to a toxic chemical.
Clinical studies - Studies of humans suffering from symptoms induced by
chemical exposure.
Confounding factors - Variables other than chemical exposure level which can
affect the incidence or degree of a parameter being measured.
Cost/benefit analysis - A quantitative evaluation of the costs which would be
incurred versus the overall benefits to society of a proposed action such
as the establishment of an acceptable dose of a toxic chemical.
Cumulative exposure - The summation of exposures of an organism to a chemical
over a period of time.
Degradation - Chemical or biological breakdown of a complex compound into
simpler compounds.
Demography - The study of the characteristics of human populations such as
size, growth, density, distribution and vital statistics.
Dermal - Of the skin: through or by the skin.
xiii
-------
Dermal exposure - Contact between a chemical and the skin.
Diffusion - The. movement of suspended or dissolved particles from a more
concentrated to a less concentrated region as a result of the random
movement of individual particles; the process tends to distribute them
uniformly throughout the available volume. -
Dose - The quantity of a chemical to which an organism is exposed. (See
absorbed dose.)
Dose-response - A quantitative relationship between the dose of a chemical and
an effect caused by the chemical.
Dose-response curve - A graphical presentation of the relationship between
degree of exposure to a chemical (dose) and observed biological effect or
response.
Ecology - The study of the interrelationships between living organisms and
their environment, both physical and biological.
Ecosystem - The interacting system of a biological community and its nonliving
environment.
Ecotoxicological studies - Measurement of effects of environmental toxicants
on indigenous populations of organisms.
Endangerment assessment - A site-specific risk assessment of the actual or
potential danger to human health or welfare and the environment from the
release of hazardous substances or waste. The endangerment assessment
document is prepared in support of enforcement actions unaer CERCLA or RCRA.
Endpoint - A biological effect used as an index of the effect of a -chemical on
an organism.
Environmental fate - The destiny of a chemical alter release Co che
environment; involves considerations such as transport through air, soil
and water, bioconcentration, degradation, etc.
Enzyme - A protein, synthesized by a cell, that acts as a catalyst in a
specific chemical reaction.
Epidemiological studies - Investigation of elements contributing to disease or
toxic effects in human populations.
Exposure - Contact with a chemical or physical agent.
Exposure assessment - The determination or estimation (qualitative or quanti-
tative) of the magnitude, frequency, duration, route, and extent (number
of people) of exposure to a chemical.
Exposure coefficient - Term which combines information on the frequency, mode,
and magnitude or contact with contaminated sedius tz ;iili i quantitative
value of the amount of contaminated medium contacted per day.
xiv
-------
Exposure level, chemical - The amount (concentration) of a-chemical at the
absorptive surfaces of an organism.
Exposure scenario - A set of conditions or assumptions about sources, exposure
pathways, concentrations of toxic chemicals and populations (numbers,
characteristics and habits) which aid the investigator in evaluating and
quantifying exposure in a given situation.
Extrapolation - Estimation of unknown values by extending or projecting from
known values.
Half-life The length of time required for the mass, concentration, or
activity of a chemical or physical agent to be reduced by one-half.
Hematopoiesis - The production of blood and blood cells; hemopoiesis.
Hepatic - Pertaining to the liver.
Hepatoma - A malignant tumor occurring in the liver.
Histology - The study of the structure of cells and tissues; usually
involves microscopic examination of tissue slices.
Homeostasis - Maintenance of a constant internal environment in an organism.
Hormone - A chemical substance secreted in one part of an organism and
transported to another part of that organism where is has a specific
effect.
Human equivalent dose - A dose which, when administered to humans, produces an
effect equal to that produced by a dose in animals.
Hydrology - The study of the properties, distribution, behavior and effects
of water on the earth's surface, in the soil and underlying rocks ar.d in
the atmosphere.
Hydrolysis - The breakdown of a chemical into two parts concomitant with
addition of the elements of water (H- and 0H-) to the products.
Hypoxia - A deficiency of oxygen.
Intake - Amount of material inhaled, ingested, or absorbed dermally during a
specified period of time.
Integrated exposure assessment - A summation over time, in all media, of the
magnitude of exposure to a toxic chemical.
In vitro studies - Studies of chemical effects conducted in tissues, cells or
subcellular extracts from an organism (i.e., not in the living organism).
xv
-------
In vivo 9tudies - Studies of chemical effects conducted in intact living
organisms.
Irreversible effect - Effect characterized by the inability of the body to
partially or fully repair injury caused by a toxic agent.
Latency - Time from the first exposure to a chemical until the appearance of a
toxic effect.
LC^q - The concentration of a chemical in air or water which is
expected to cause death in 50 percent of test animals living in that air
or water.
LD^q - The dose of a chemical taken by mouth or absorbed by the skin
which is expected to cause death in 50 percent of the test animals so
treated.
LOAEL - Lowest-Observed-Adverse-Effect Level; the lowest dose in an experiment
which produced an observable adverse effect.
Materials balance - An accounting, of the mass flow of a substance from sources
of production, through distribution and use, to disposal or distribution,
and including any releases to the environment.
Metabolism - The sum of the chemical reactions occurring within a cell or a
whole organism; includs the energy-releasing breakdown of molecules
(catabolism) and the synthesis of new molecules (anabolism).
Metabolite - Any product of metabolism, especially a transformed chemical.
Modeling - Use of mathematical equations to simulate and predict real events
and processes.
Monitoring - Measuring concentrations of substances in environmental media or
in human or ocner biological tissues.
Mutagen - An agent that causes a permanent genetic change in a cell other than
that which occurs during normal genetic recombination.
Mutagenicity - The capacity of a chemical or physical agent to cause
permanent alteration of the genetic material within living cells.
Necrosis - Death of cells or tissue.
NOAEL - No-Observed-Adverse-Effect Level; the highest dose in an experiment
which did not produce an observable adverse effect.
Oncology - Study of cancer.
Oral - Of the mouth; through or by the mouth.
xvi
-------
Pathogen - Any disease-causing agent, usually applied to living agents.
Pathogenic - Causing or capable of causing disease.
Pathology - The study of disease.
Permissible dose - The dose of a chemical that may be received by an
individual without the expectation of significantly harmful result.
Pharmacokinetics - The dynamic behavior of chemicals inside biological
systems; It includes the processes of uptake, distribution, metabolism,
and excretion;
Population at risk - A population subgroup that is more likely to be exposed
to a chemical, or is more sensitive to a chemical, than is the general
population.
Potentiation - The effect of one chemical to increase the effect of another
chemical.
Prevalence study - An epidemiological study which examines the relationships
between diseases and exposures as they exist in a defined population at a
particular point in time.
Prospective study - An epidemiological study which examines the development of
disease in a group of persons determined to be presently free of the
disease.
Qualitative - Descriptive of kind, type or direction, as opposed to size,
magnitude or degree.
Quantitative - Descriptive of size, magnitude or degree.
Receptor - (1) In biochemistry: a specialized molecule in a ceil that binds a
specific chemical with high specificity and high affinity; (2) In
exposure assessment: an organism that receives, may receive, or has
received environmental exposure to a chemical.
Renal - Pertaining to the kidney.
Reservoir - A tissue in an organism or a place in the environment where a
chemical accumulates, from which it may be released at a later time.
Retrospective study - An epidemiological study which compares diseased persons
with non-diseased persons and works back in time to determine exposures.
Reversible effect - An effect which is not permanent, especially adverse
effects which diminish when exposure to a toxic chemical is ceased.
Risk - The potential for realization of unwanted negative consequences or
events.
xvii
-------
Risk assessment - A qualitative or quantitative evaluation of the environ-
mental and/or health risk resulting from exposure to a chemical or phys-
ical agent (pollutant); combines exposure assessment results with toxi-
city assessment results to estimate risk.
Risk estimate - A description of the probability that organisms exposed to a
specified dose of chemical will develop an adverse response (e.g.,
cancer).
Risk factor - Characteristic (e.g., race, sex, age, obesity) or variable
(e.g., smoking, occupational exposure level) associated with increased
probability of a toxic effect.
Risk specific dose - The dose associated with a specified risk level.
Route of exposure - The avenue by which a chemical comes into contact with an
organism (e.g., inhalation, ingestion, dermal contact, injection).
Sink - A place in the environment where a compound or material collects (see
reservoir).
Sorption - A surface phenomenon which may be either absorption or adsorption,
or a combination of the two; often used when the specific mechanism is
not known.
Stochastic - Based on the assumption that the actions of a chemical substance
result from probabilistic events.
Stratification - (1) The division of a population into subpopulations for
sampling purposes; (2) Che separation of environmental media into layers,
as in lakes.
Subchronic - Of intermediate duration, usually used to describe studies or
levels of exposure between five and 90 days.
Synergism - An interaction of two or more chemicals that results in an effect
that is greater than the sum of their effects taken independently.
Systemic - Relating to whole body, rather than its individual parts.
Teratogenesis - The induction of structural or functional development
abnormalities by exogenous factors acting during gestation; interference
with normal embryonic development.
Teratogenicity - The capacity of a physical or chemical agent to cause non-
hereditary congenital malformations (birth defects) in offspring.
Therapeutic Index - The ratio of the dose required to produce toxic or lethal
effect to dose required to produce non-adverse or therapeutic response.
xviii
-------
Threshold - The lowest dose of a chemical at which a specified measurable
effect is observed and below which it is not observed.
Time-Weighted Average - The average value of a parameter (e.g., concentration
of a chemical in air) that varies over time.
Tissue - A group of similar cells.
Toxicant - A harmful substance or agent that may Injure an exposed organism.
Toxicity - The quality or degree of being poisonous or harmful to plant,
animal or human life.
Toxicity Assessment - Characterization of the toxlcologlcal properties and
effects of a chemical, including all aspects of its absorption,
metabolism, excretion and mechanism of action* with tpecial emphasis on
establishment of dose-response characteristics.
Transformation - Acquisition by a cell of the property of uncontrolled growth.
Uncertainty Factor - A number (equal to or greater than one) used to divide
NOAEL or LOAEL values derived from measurements in animals or small
groups of humans, in order to estimate a NOAEL value for the whole human
population..
xix
-------
EXECUTIVE SUMMARY
1.0 INTRODUCTION
Risk assessment is the process of characterizing the risk either to human
health or to the environment as the result of chemical releases into the
environment from some specific site (such as a factory or waste storage
facility). The process of collecting and interpreting the Information needed
to perform a risk assessment consists of two main branches: toxicity assess-
ment (characterization of the inherent toxicity of a chemical) and exposure
assessment (determination of how much of the chemical is coming into contact
with humans or other species). Figure ES-1 is a generalized scheme that
provides an overview of the risk assessment process and illustrates some of
the major elements involved.
Examination of Figure ES-1 makes clear that risk assessment requires the
interaction and cooperation of scientists from a variety of disciplines,
including geology, hydrology, meteorology, ecology, biochemistry, chemistry
and toxicology. This handbook is intended to provide a description of the
toxicity assessment process in a way that is useful to non-toxicologists.
2.0 FUNDAMENTAL CONCEPTS IN TOXICOLOGY
Living organisms are composed of cells, and all cells must carry out a large
number of chemical reactions in order to maintain themselves and perform their
functions. Introduction of a foreign chemical into a cell may interfere with
one or more of these cellular reactions, leading to impaired cell function or
viability.
Toxicology is the study of how specific chemicals cause injury to living cells
and whole organisms. Studies are performed to determine how easily the
chemical enters the organism, how it behaves inside the organism, how rapidly
it is removed from the organism, what cells are affected by the chemical ar.d
what cell functions are impaired. With respect to the risk assessment process,
the ultimate goal is usually to derive a reliable estimate of the amount of
chemical exposure which is considered acceptable for humans or other
organisms. It is important to recognize that, for many chemicals, current
toxicological knowledge is insufficient to answer this question with
assurance.
3.0 DOSE-RESPONSE RELATIONSHIPS
The relationship between degree of exposure to a chemical (dose) and the
magnitude of chemical-induced effects (response) is described by a dose-
response curve. In general, dose-response curves fall into two groups: those
in which no response is observed until some minimum (threshold) dose is
reached, and those in which no threshold is apparent.
A hypothetical dose-response curve with a threshold is shown in Figure ES-2.
The most important part of this curve is the dose at which significant effects
first begin to occur. The highest dose which does not produce an observable
adverse effect is the No-Observed-Adverse-Effect-Level (NOAEL), and the lowest
ES-1
-------
TOXICITY ASSESSMENT
EXPOSURE ASSESSMENT
Geology, Hydrology,
Meterology, Types and
Amounts or Cnemicals
Site Characteristics
Dose
Duration
Route
Species
Effects
Quality
Affected Populations
Human, Plant, Animal
Intake Assumptions
Sampling/Analysis
Hazardous Wwte
Disposal Problem
Quantitative Indices
NOAEL
LOAEL
Animal Epidemiological
Clinical Ecotoxicotogical
Uncertainty/
Margin of Safety
RISK
ASSESSMENT
FIGURE ES-1 OVERVIEW OF THE RISK ASSESSMENT PROCESS
APPLIED TO A HAZARDOUS WASTE SITE
ES-2
-------
dose vhich produces an observable adverse effect is the Lowest-Observed-
Adverse-Effect-Level (LOAEL). With respect to the toxicity assessment
process, identification of these points is a primary objective of the
toxicologist. Other useful points are the TD-q (the dose which produces 50%
of the maximum toxic effect or produces the toxic effect in 50% of the exposed
organisms) and the TD.0Q (the lowest dose which produces the maximum toxic
effect or produces the toxic effect in all exposed organisms).
An example of a dose-response curve that does not have a threshold is shown in
Figure ES-3. For chemicals of this sort, there is no dose that is free of
risk, but as the dose decreases to low levels, so does the probability that
the effect will occur. Cancer is an adverse response that is believed to have
no observable threshold. For these types of chemicals, selection of a "safe"
exposure limit is made on the basis of what risk level is acceptable to
society.
4.0 IMPORTANT PARAMETERS IN TOXICITY ASSESSMENT
Even under controlled laboratory conditions, it is not always simple to obtain
reliable and useful dose-response data. There are a number of important
variables which determine the characteristics of dose-response curves and must
be considered in performing toxicity tests and interpreting toxicity data. The
most important of these variables are discussed below.
4.1 Route of Exposure
The toxicity of some chemicals depends on whether the route of exposure is by
ingestion, inhalation or dermal contact. In addition, there may be local
responses at the absorption site (gastrointestinal tract, lungs, skin), since
the concentration of the chemical is highest at chat xocation. Ihe rouce or
exposure employed in experimental animal studies is normally selected based on
the anticipated route of exposure of humans to the specific chemical.
4.2 Duration/Frequency of Exposure
The toxicity of many chemicals depends not only on dose (the amount of
chemical contacted or absorbed each day) but also on the length of exposure
(number of days, weeks or years). This is especially true for chemicals which
produce irreversible injuries to cells or tissues. Thus, brief exposure to a
low dose of such a chemical may produce so little damage that no significant
injury occ\irs, but continued exposure will result in an accumulation of damage
that eventually becomes apparent as a significant injury. For this reason, a
full toxicological evaluation of any chemical must include consideration of
the time-dependence of any adverse effects chat occur. Typically, studies
focus on acute (one-day), subchronic (roughly 5 to 90 days) and chronic
(lifetime) exposures.
4. 3 Test Species
For obvious reasons, laboratory investigations of chemical toxicity employ
animals as test species. Unfortunately, not all animal species are equally
ES-3
-------
T 1 1 ——r
LOAEL
NOAEL
Dose, Arbitrary Units (Logarithmic Scale)
FIGURE ES-2 HYPOTHETICAL DOSE-RESPONSE CURVE WITH A THRESHOLD
ES-4
-------
—i\j , 1 1 1 r
10"1 -
Dose, Arbitrary Units (Logarithmic Scale)
FIGURE ES-3 HYPOTHETICAL DOSE-RESPONSE CURVE WITH NO THRESHOLD
ES-5
-------
sensitive to the toxic effects of chemicals, and information which is obtained
in some animals may not be directly relevant to humans. This is often a
result of differences among species vith respect to absorption, excretion or
®etabolism of a chemical, although other factors (e.g., genetic
susceptibility) may also be involved. It is for this reason that thorough
toxicity assessments involve studies with several species of animals.
^ Individual Characteristics
Individual members of a population (especially humans) are not identical and
usually do not respond identically to equal exposures to a chemical. This
variation has two elements: variation among subgroups of the population as a
function of age, race or sex, and variation between individuals within a
subgroup (e.g., white males aged 20 to 30 years). Whether the reasons for
these variations are understood'or not, it is important to identify any
subgroups that may have greater sensitivity to a chemical than the general
Population.
Toxicological Endpoints
^en a chemical is given to a test animal, it is often possible to detect and
Measure a number of changes in the animal. Those changes which are used by
the toxicologist as an index of the chemical's toxicity are called "end-
Points." Some commonly measured endpoints are carcinogenicity, hepato-
toxicity, mutagenicity, neurotoxicity, renal toxicity, reproductive and fetal
toxicity and teratogenicity. One of the most important parts of any toxicity
study is selection of the best endpoint to monitor. Usually the endpoint
which is most sensitive (i.e., the parameter in which a measurable change
'•irst occurs as dose levels increase) is judged to be most appropriate as an
lridex or toxicity. For example, a scuay aeasuring a sensitive oi-jchacical
•••ndicator of early liver injury is more valuable than a study measuring only
gross injury (necrosis or cirrhosis) to the liver.
5>0 TYPICAL protocols used in toxicological studies
A typical experiment designed to determine the dose-response curve for a
chemical might involve giving six groups of animals a series of increasing
doses of the chemical (e.g., 0, 1, 3, 10, 30 or 100 mg/kg) and measuring the
magnitude or frequency of the response at each dose level.
A well-performed experiment of this sort may be sufficient to describe the
shape of the dose-response curve for a particular test species. However, as
discussed above, results in one species are not always directly applicable to
humans or other species. Thus, any data which help to define a dose-response
curve directly in humans are especially valuable. One source of human data is
clinical experience with patients seeking medical attention for health
Problems resulting from exposure to chemicals. While this is useful in
identifying toxic effects in humans, quantitative information on the amount of
chemical causing the effects is usually not known. Moreover, even when the
exposure, is known, it is generally high on the dose-response curve and is not
°f direct help in defining the N0AEL in humans.
ES-6
-------
A second source of data in humans Is epidemiological studies. An epidemio-
logical study usually seeks to determine whether a particular health effect is
associated with exposure to a certain chemical. Typically, incidence of the
health effect is compared between two groups of humans, one group known to be
exposed to the chemical at a higher level than the other. These studies.are
often difficult to analyze due to interference by confounding variables such
as age, weight, diet,.smoking, etc., which may also be related to the health
effect under study. Even when meaningful associations are established,
epidemiological studies rarely supply dose-response information. It is
precisely because of the difficulty in obtaining useful dose-response data in
humans that experiments in animals are so important in toxicity assessments.
Contamination of the environment with a toxic chemical may threaten not only
humans but wildlife and plants as well. Assessment of the toxic effects of
chemicals on fish, terrestrial animals and plants is performed in much the
same way as for laboratory animals. Representative species are exposed to
various doses or exposure levels of the chemical for varying lengths of tir.e,
and signs of adverse effects are noted. From data of this sort"it is possible
to correlate concentrations of the chemical in the environment to specific
levels of risk to indigenous flora and fauna.
6.0 EXTRAPOLATION OF TOXICOLOGICAL DATA FROM ANIMALS TO HUMANS
Since most toxicological data on a chemical are obtained from studies in
animals, it is frequently necessary to extrapolate from results in animals to
predicted results in humans in order to derive appropriate guidelines or
standards regarding acceptable human exposure limits.
' ' f «•« cp 1 ^
The initial step in this process is selection of the most appropriate anixal
studies which have been performed. Criteria for selection of studies include
the use of appropriate endpoints, optimal analytical methods, proper experi-
mental design and correct data analysis. Studies which provide a clear
picture of dose-response relationships are especially valuable in estimating
no-effect dose levels.
6.2 Conversions and Corrections
Once suitable studies have been identified, it is necessary to convert the
doses administered to the animals to equivalent doses in humans. Since doses
are normally described in terms of amount per unit weight (e.g., ag/kg), this
would be simple if all animals were equally sensitive on this basis. However,
it appears that a better correlation among different species exists when dose
is expressed as amount per unit surface area (mg/m ), and so it is necessary
to make this correction. Another correction which is desirable (if sufficient
data are available) is for differences between animals and humans in
absorption, metabolism or excretion rates of chemicals.
ES-7
-------
6.3
High-Dose to Low-Dose Extrapolation
One of the most difficult problems In extrapolating data from animal studies
to expected results in humans concerns dose levels. Most laboratory
experiments are performed at dose levels of the chemicals that produce clear,
e&sily measured responses in the test animals. However, exposure to chemicals
in the environment often occurs at dose levels low enough that adverse effects
are not immediate or obvious. Thus, it is necessary to extrapolate results
obtained at high doses to results expected at low doses. This is especially
difficult in the case of a carcinogen. In this case, extrapolation must span
a very large change in dose levels (four to six orders of magnitude).
Extrapolations are done mathematically by application of equations (models)
thought to describe the dose-response curve, but unfortunately, the shape of
the dose-response curve at low doses is not known with certainty. The result
Is that different models yield predictions about risk at low doses that may
differ by several orders of magnitude.
6.4 Dealing with Uncertainty
There are many sources of uncertainty inherent in biological research. With
respect to the toxicity assessment process, the problems associated with
high-dose to low-dose extrapolation, Inter-species extrapolation and variation
In the sensitivity of sub-groups and individuals have been mentioned above.
Because of the uncertainty created by these and other problems, it is general-
ly necessary and appropriate to include a "margin of safety" when establishing
a toxicity level. For chemicals that appear to exhibit thresholds in the
dose-response curve (i.e., noncarcinogens), this is accomplished by dividing
Che dose for which no adverse effect has been observed (the NOAEL) by an
appropriace uncertainty factor. The basic philosophy behind the selection of
uncertainty factors is that when information is lacking and estimates must be
wade, it is better to err on the low side (i.e., be conservative) than to run
the risk of setting too high a value. A similar philosophy underlies the
»eans of dealing with uncertainty in estimating risk from chemicals considered
to be without thresholds (e.g., carcinogens). A conservative model is used to
Wake dose extrapolations, and then the upper 95Z confidence limit of the
extrapolation is employed. Taken together, this ensures that these risk
Estimates will be highly conservative (i.e., they may be very low, but will
°ot be too high).
7-0 EXPOSURE ASSESSMENT
The second portion of the risk assessment process is the collection and
evaluation of data on the actual and potential exposure of humans and other
8Pecles to chemicals on or near the site. As opposed to toxicity assessment,
whlch is performed primarily in the laboratory, exposure assessment involves a
^tailed study of the specific site. Specifically, the toxicologist needs
to know what chemicals are being released from the site into the environment,
what the pattern of chemical distribution is in and around the site and the
expected pattern of chemical distribution in the future.
I Collection Occurrence Data
The first step in answering these questions is collection of data regarding
the identities and concentrations of chemicals which occur in and around
ES-8
-------
Che site. Consideration of the history of the site (Identity, amount and
means of storage of chemicals) often helps to focus concern on one or two
specific chemicals known to be present. However, since 1« 'prmation of this
sort is sometimes incomplete or unreliable, it may be neo «sary to test for a
variety of pollutants. In order to provide an adequate description of- the
distribution of chemicals in the environment, samples of air, soil, sediment,
surface water, groundwater and biota (e.g.* fish, shellfish) should be
collected at the site and at a series of distances and directions from the
site. In some cases, mathematical models of chemical transport in air, water
or soil are used to calculate environmental concentrations of chemicals in
locations where they have not been or cannot be measured directly.
Since the results of the chemical analysis of environmental samples are
usually an essential part of any legal proceedings regarding the site, it is
very important to follow strict scientific and legal procedures during
collection and analysis of the samples.
?•2 Identification of Exposed Organisms
No matter"hov toxic a chemical may be or how concentrated it may be in the
environment, no Injury can occur to an organism if the chemical does not come
into contact with the organism. For this reason, an essential part of the
exposure assessment process is identification of organisms (humans and other
speciesj that live in or periodically enter the contaminated area.
Additional important information includes a description of how the organisms
come in contact with the chemical (inhalation, ingestion in water, dermal
contact, etc.) and how long they are exposed.
7.3 Calculation of Exposure
Given quantitative data on the concentration and distribution cf chemicals in
and around the site, it is necessary to calculate the degree of exposure of
humans and of environmental organisms that come into contact with the
chemical. For human exposure to 3 chemical in water, for example, this
calculation involves finding the product of the concentration in the water at
the location where the exposure occurs times the intake of water per day by
the exposed humans at that location. If the absorption factor Is known, the
absorbed dose can be calculated. Similar calculations may be made for other
routes of exposure, and the total dose from all routes of exposure is found by
addition.
7.4 Estimation of Past and Future Exposure
A thorough risk assessment may require knowledge not only of present exposure,
but of past and future exposure as well. Past exposure to some chemicals can
be estimated from measurements of the levels of the chemical in the body.
This is useful only for chemicals which tend to accumulate in the body.
Estimates of past or future exposure levels must usually be calculated, based
on the expected movement and stability of the chemical in the environment.
Movement of a chemical is determined by the properties of the chemical and the
3logic md characteristic of :hs :ita. sample, 'f •>.
ES-9
-------
chemical is readily soluble and is found to occur primarily in groundwater,
the rate and direction of groundvater flow will be the principal controlling
force in chemical movement. Conversely, if a chemical is strongly adsorbed to
surface soil, weather conditions controlling soil erosion (such as annual
rainfall and prevailing wind patterns) would be of chief concern.
Stability of a chemical in the environment is determined by the rate at which
it undergoes reactions. Such reactions include microbial degradation,
photolysis, hydroloysis and reduction or oxidation. Usually these reactions
decrease the amount and the toxicity of the chemicals, but some reactions -
actually increase toxicity.
8.0 RISK ASSESSMENT
The final step in the risk assessment process is integration of the results of
the toxicity assessment process and the exposure assessment process. In
simplest terms, what Is required is a comparison of the exposure levels with
the levels believed to pose health risks. For example, if the groundwater
being tapped by wells used for human consumption contains SO Ug/L of
trichloroethylene, and a concentration of 5 Ug/L of trichloroethylene is
believed to pose a risk of cancer (1/10 ), then the level of risk to humans
is 1/10 . "Conversely, if air levels of chloroform are I ppm, and levels of 10
ppm in air are considered to be safe, then no present risk exists.
In real -life, risk assessment may not.be quite so simple. For example, there
may be significant (and legitimate) differences of opinion between scientists
regarding interpretation of key data on toxicity,' predictions of chemical fate
Sr.transport, etc. In addition, when data is lacking and judgments must be
made, it is not uncommon that dissimilar conclusions may be reached by
different scientists, depending on Che assumptions which are csplsyed.
Thus, risk assessment involves a blend of scientific fact, judgment and common
sense. Clearly, however, the keystone to a sound risk assessment is
collection of -a sound, reliable and complete set of exposure data along with a
thorough toxicity assessment*
9.0 TOXIC0L0GICAL INFORMATION SUMMARIES
The last three chapters of the handbook present summaries of the current
understanding of selected chemicals which occur at many hazardous waste sites:
dloxlns and furans, trichloroethylene and lead. These summaries contain a
greater level of technical detail, similar to that which a practicing
toxlcologlst would prepare. For the nontoxicologist, they present both
specific information on frequently occurring chemicals and an opportunity to
aPply the understanding obtained from the earlier chapters of the handbook.
ES-1Q
-------
FOREWORD
This handbook was prepared by ICAIR, Life Systems, Inc., under U.S.
Environmental Protection Agency (EPA) Contract 68-01-7037 during the period
February, 1985 to June, 1985. The program was directed by Mr. Timothy E.
Tyburski. The handbook was compiled by Mr. Jeffrey Beaton with technical
support from Dr. William Brattin, Dr. Carol Maczka, Mr. Kevin Gleason,
Ms. Yvonne Hales, Mr. Steve Lavenhar, Ms. Betty Neustadter, Ms. Lee Ann Smith
and Ms. Jo Ann Duchene. Advice about the topics and the appropriate level of
detail was obtained from four toxicologists experienced in giving expert
witness testimony in hazardous waste litigation actions: Dr. Herbert Cornish,
Dr. Curtis Klaassen, Dr. Andrew Reeves and Dr. James Selkirk.
Mr. R. Charles Morgan and Ms. Kathleen Plourd, Health Sciences Section, were
the lead Technical Contacts for the Office of Waste Programs Enforcement
during the preparation of this handbook. ICAIR would also like to acknowledge
the contribution of the members of the EPA Workgroup to this program:
Dr. Arthur Chiu, ORD Carcinogen Assessment Group; Dr. Leroy Folmer and
Mr. Abraham Mittelman, OWPE; Dr. Charles Nauman, ORD Exposure Assessment
Group; Dr. Larry Valcovic, ORD Reproductive Effects Assessment Group; Dr. Judy
Bellin, OSW; and Dr. Steve Ostrodka, EPA Region V.
-------
1.0 INTRODUCTION
The U.S. Environmental Protection Agency (EPA) is responsible for investigat-
ing hazardous wast? disposal sites under the Comprehensive Environmental
Response, Compensation and Liability Act of 1980 (CERCLA) and the Resource
Conservation and Recovery Act of 1976 (RCRA). The activities under CERCLA are
directed by the EPA Office of Emergency and Remedial Response (OERR). The
activities under RCRA are directed by the EPA Office of Solid Waste (OSW).
The EPA Office of Waste Programs Enforcement (OWPE) provides program manage-
ment for EPA solid waste and emergency/remedial response enforcement activi-
ties. This handbook was prepared under the direction of OWPE. All three
offices (OWPE, OERR and OSW) are directed by the EPA Assistant Administrator
for Solid Waste and Emergency Response.
1.1 Purpose and Scope of the Toxicology Handbook
Enforcement actions at hazardous waste sites require regional EPA personnel to
understand general toxicological principles and to be able to effectively
interact with toxicologists. The purpose of this handbook is to explain to
non-toxicology-trained personnel those principles of toxicology relevant to
hazardous waste site investigations. This handbook is not highly technical
and is intended to provide a working familiarity with relevant toxicological
principles. It is not intended to teach a nontoxlcologist how to perform
hazard, exposure or risk assessments. Sources of more detailed information
are provided in .the handbook. By understanding the principles of toxicity,
exposure, risk and endan'germent assessment processes and the relevant toxi-
cology of the chemicals -of high concern presented in this handbook, regional
personnel will be better able to conduct the hazardous waste site investiga-
tions. Better informed hazardous waste site investigations will result in
rewer problems in suosequenc Federal activities such as litigacion and
remedial actions.
1.2 Handbook Organization
The Toxicology Handbook is divided into eleven chapters in addition to an
executive summary and glossary in the front and an index at the back. A
description of the contents of each chapter appears in section 2.0. At the
conclusion of chapters four through eight, key guidance and implementation
documents are provided to direct the user to additional technical- information.
Background references used in preparing chapters are listed at the end of each
chapter.
1-i
-------
2.0 FUNDAMENTAL CONCEPTS IN TOXICOLOGY
Toxicology is the study of how chemical substances, either natural or
man-made, cause undesirable effects in living organisms. Knowledge derived
from toxicological studies has many applications, including the provision of
recommendations to public officials charged with the protection of human
health and the environment from hazardous substances.
Risk to human health and the environment from toxic chemicals is a matter of
grave concern to modern society. Large amounts of hazardous substances have
been inadequately or improperly disposed of ", and reports of toxic, chemicals
entering air, water and food occur with alarming regularity. However, simple
detection of a hazardous chemical in the environment is not necessarily cause
for alarm. Indeed, modern chemical detection techniques are so sensitive that
it is possible to detect contaminants at very low levels. Additionally,
nearly any chemical (including such common items as coffee, alcohol, salt,
even water) can produce adverse effects when consumed in high enough amounts.
Clearly, determination of the problems posed by a chemical detected in the
environment must involve an evaluation of how much of a chemical is present,
judged in terms of how toxic that chemical is.
The process of characterizing the inherent toxicity of a chemical is termed
toxicity assessment. This process involves determining what adverse effects
the chemical causes in exposed organisms, and how much of the chemical is
required to produce thes^ effects. The important concepts involved in
toxicity assessment are introduced in sections 2.1 through 2.3 of this chapter
and are explained in more detail in chapters 3.0 through 6.0 of this handbook.
The process of determining how much of a chemical is in the environment and
may come into contact with humans or other organisms is termec exposure
assessment. The important concepts in this process are introduced in
section 2.4 of this chapter and described in more detail in chapter 7.0.
The process of integrating available Information on the ir.herar.t toxicity of a
chemical with information on how much of a chemical cay sake contact vith
exposed organisms is termed risk assessment. A description of the main
elements in this process, especially as they are performed at hazardous waste
sites, is contained in chapter 8.0. Figure 2-1 provides an overview of the
risk assessment process.
Chapters 9.0 through 11.0 provide detailed summaries of toxicological
information on chemicals of common concern at hazardous waste sites: dioxin
(and furan), trichloroethylene and lead.
2.1 Mechanisms of Chemical Toxicity
All living organisms are composed of cells. In humans and other higher
organisms, Nearly all cells are specialized to perform some specific function
to benefit the entire organism. For example, muscle cells are specialized to
perform motion, retinal cells are specialized to detect light and red blood
cells are specialized to carry oxygen. In addition to these specialized
2-1
-------
TOXICITY ASSESSMENT
Research
Animal Epidemiological
Clinical Ecotoxicoiogicai
Data
Dose
Duration
Route
Species
Effects
Quality
Quantitative Indices
NOAEL
LOAEL
TDjo
TDioo
Uncertainty/
Margin of Safety
roxic Level
EXPOSURE ASSESSMENT
Site Characteristics
igy. Hydrology,
iiogy, Typtfs and
iicai».
Means of Storage
Sampling/Analysis
Chemicals
Concentrations
Distributions
Human, Plant, Animal
Intake Assumptions
Affected Populations
Hazardous Waste
Disposal Problem
*
RISK
ASSESSMENT
FIGURE 2-1 OVERVIEW OF THE RISK ASSESSMENT PROCESS
APPLIED TO A HAZARDOUS WASTE SITE
2-2
-------
functions, all cells must perform certain basic functions, such as generating
the energy they need and maintaining and repairing themselves. All of these
specialized and basic functions are the direct result of chemical reactions
which occur in the cells. It is essential to the proper functioning of a cell
that all of the chemical reactions occurring within it proceed at the proper
rate; any significant increase or decrease in even one key reaction may lead
to failure of the cell.
In view of the large number of chemical reactions that are occurring in all
cells at all times, and the importance that these reactions not be disturbed,
it is perhaps not surprising that introduction of a foreign chemical into a
cell may result in adverse effects. Different types of chemicals may disturb
cellular chemical reactions by different means. Some chemicals may react with
key cellular molecules, thereby changing their properties, damaging them or
rendering them ineffective. Other chemicals may substitute for normal body
chemicals, leading to formation of unusual new products, which may be tnor-e or
less toxic than the original chemicals, and preventing formation of normal
products.
Figure 2-2 summarizes two simple examples of these sorts of toxic mechanisms.
Red blood cells contain hemoglobin (Hb) , a chemical specialized for the
binding and release of oxygen molecules (0^). Carbon monoxide (CO) is a
chemical that is sufficiently similar to oxygen that it may substitute for
oxygen and bind with hemoglobin. Hemoglobin molecules which bind a molecule
of carbon monoxide are thus rendered unable to.carry their normal product
(oxygen), and death may result from oxygen starvation of the cells. Nitrite
(NO ) is a chemical that has a similar result, but involves a different
mecnanism. Hemoglobin contains one atoij^of iron, and this is in a particular
form, (tensed ferrous icn, written as Fe ). Nitrite reacts with this ferrous
ion, changing it to ferric ion (re,"">). This change completely destroys the
ability of hemoglobin to carry oxygen ana, as with carbon monoxide, death lay
ensue from oxygen starvation at the cellular level.
Ultimately, the toxic effects of any chemical must be due to its interfersr.ce
with some important cellular reaction, but cha details will vary with each
individual chemical.
2.2 Essential Toxicological Information
A full understanding of the toxicological effects of a chemical requires
detailed investigation of a number of aspects of how a chemical behaves in an
organism, including:
• Absorption - How readily does the chemical enter the organism through
the skin, stomach or lungs?
• Distribution - What cells of the organism does the chemical enter? Do
some cells take up more than others?
• Excretion - How rapidly does the organism get rid of the chemical?
Does it tend to accumulate in the organism over long periods?
2-3
-------
Normal Cell Reaction
Lungs
02 + Hb i * Htr02
Tissues
Carbon Monoxide Poisoning
O,
Lung C
CO + Hb Hb-CO U Hb-02
Nitrite Poisoning
02
N02-+ Hb(Fe2+) N03- + Hb(Fe3+) ^—jj—* Hb-02
Hb = Hemoglobin
02 = Oxygen
CO = Carbon Monoxide
N05 = Nitrite Ion
NO5 = Nitrate Ion
Fe2+ = Ferrous Ion
Fe3+ = Ferric Ion
FIGURE 2-2 EXAMPLES OF MECHANISMS BY WHICH TWO
CHEMICALS ALTER A CELLULAR FUNCTION
2-4
-------
• Toxic Effects - What are the toxic effects that occur in an organism
exposed to the chemical? What cells are injured? Are the effects
permanent?
• Sensitive Populations - Does the toxicity of the chemical depend on
age, sex or race? If so, why? What is the most sensitive sub-group?
• Mechanism - What chemical reactions are altered by the chemical? How
does the chemical do this?
While all of these questions must ultimately be answered in order to fully
describe the toxicity of a chemical, it is possible (and often necessary) to
make regulatory decisions regarding a chemical with a more limited amount of
knowledge. Specifically, the key information that is required is an estimate
of the highest amount (dose) of a chemical that does not produce any
undesirable or unacceptable effects in the exposed organism (plant, animal,
human).
For some chemicals, there may be no dose that is without risk of causing
adverse effects. This is suspected to be true for chemicals which cause
cancer (carcinogens). In this case, even a single molecule of the chemical
might react within a cell causing a change in the system controlling cell
growth that cannot be reversed, leading (in the worst case) to a fatal tumor.
For other chemicals, there may be a certain minimum level of exposure (a
threshold value) below which no significant or observable effects occur. This
is due to one or both of two reasons: (1) there is considerable "reserve
capacity" in many tissues, such that limited damage (e.g., a 5Z loss of
hemoglobin) does not cause any decrease in function; and (2) most cells have
at iaast a limited capacity Co repair or compensate for cellular damage. 2:
course, both the reserve and the repair capacity of a tissue can be
overwhelmed by too much of a harmful chemical and it is at that point that the
organism as a whole begins to suffer adverse effects.
Thus, the essential toxicological questions that must be answered in order to
make informed regulatory decisions regarding a chemical are:
• Does the chemical cause effects that do not have a dose-response
threshold (e.g., cancer)?
• For effects that do have a dose-response threshold, what is that
threshold value?
A generalized description of the means of answering these questions through
toxicological research is presented in chapters 3.0, 4.0, 5.0 and 6.0, of this
handbook.
2.3 Limitations to Toxicological Knowledge
Existing toxicological knowledge cannot provide answers to all the questions
of concern to regulatory agencies. One reason is simply the large number of
-------
chemicals of potential or known concern that are in use today. Toxicological
investigation and characterization of a chemical is a slow and expensive
process and evaluation of all existing chemicals would greatly exceed the
capacity of the toxicological community. Additionally, new chemicals are
being developed and proposed for use at a rate that further overloads the
capacity to perform additional toxicological studies.
When it becomes necessary to make a decision on a chemical in the absence of
full and detailed toxicological knowledge, the only solution is to make
reasonable assumptions. Often there is partial information on the chemical or
knowledge of a similar chemical. In these cases, assumptions may be made by
extension or extrapolation of existing knowledge to achieve answers to
questions that have not been studied directly.
Such extensions and extrapolations of information are never certain, and it is
possible for knowledgeable and reasonable scientists to differ considerably in
the extrapolations which they feel are appropriate.
The optimal solution to this problem is to choose defensible and well reasoned
assumptions to yield a range of potential risk including that which is most
conservative (i.e., provides for the greatest margin of safety). When the
requirements for human health concern are addressed in terms of cost and
feasibility, it is important to remember that present toxicological knowledge
is insufficient in many cases to provide definitive answers about a chemical
and it is, therefore, important to characterize this uncertainty.
2.4 Exposure to Toxic Chemicals
No matter how toxic a chemical may be, it cannot cause an effect on a living
organism unless it comes into contact wicr. that organism. This contact
between chemical and organism is termed exposure. There are two key aspects
of exposure that are important determinants of the effect on the organism.
First is the site of contact between the chemical and the organism (skin,
eyes, lungs, gastrointestinal tract). This is important since the amount of
chemical actually entering the body (absorbed) depends on the sase with which
it can cross these body surfaces. The second aspect is the amount of chemical
making contact with the organism. This amount is usually described as a
concentration in each of the relevant environmental media (air, water, soil,
food) . From these environmental concentrations It is possible to calculate
the amount of chemical contacting the organism, based on the rate of contact
between the organism and the media (e.g., liters of air breathed per day,
liters of water consumed per day, etc.).
From these considerations, it is clear that an evaluation of the exposure to
humans or other organisms by a chemical in the environment requires answers to
the following questions:
• In what environmental media (air, water, soil, food) is the chemical
present ?
• How much of the chemical is or is likely to be present in these media?
2-6
-------
• How many and which organisms are or will likely become exposed to the
chemical?
• By what routes may the chemical make contact with exposed organisms?
• What is the frequency and duration of contact?
Collection of this information constitutes the exposure assessment process. A
more detailed description of the issues and requirements of this process is
presented in chapter 7.0.
2-7
-------
3.0 DOSE-RESPONSE RELATIONSHIPS
As discussed in chapter 2.0, complete assessment of the toxicity of a chemical
requires detailed study of a number of areas relating to how the chemical
behaves in the body. However, as input to regulatory decision making, the
major objective of a toxicity assessment is the determination of the maximum
dose which produces no significant adverse effect (response) in exposed
organisms. This information is obtained by deriving the dose-response curve
for the chemical.
3.1 Identification of No-Effect Levels From Dose-Response Curves
The dose-response relationship is the most fundamental concept in toxicology.
The dose-response curve describes the relationship that exists between degree
of exposure to a chemical (dose) and the magnitude of the effect (response) in
the exposed organism. By definition, no response is seen in the absence of
chemical. As the amount of chemical exposure increases, the response becomes
apparent and increases. Depending on the mechanism by which the chemical
acts, the curve may rise with or without a threshold (Figure 3-1). In both
cases, the response normally reaches a maximum after which the dose-response
curve becomes flat.
3.1.1 Chemicals with Thresholds
Many Chemicals produce responses that display a threshold value (an exposure
below which no responses can be detected). One example of-this is shown in
Figure 3-2. Nitrite (N02 ) is a chemical which produces injury to ^d blood
cells by oxidizing the iron atom in hemoglobin from the ferrous (Fe ) to the
ferric ''?eJ"r^ form. Within the red blood cell is a system designed to protect
the cell against this kind of damage. It involves providing a ''sacriiiciai;i
chemical, glutathione (GSH), that is oxidized by the nitrite instead of the
iron. Only when the amount of nitrite exceeds the ability of the cell to
supply glutathione is the iron of hemoglobin oxidized.
A second reason that a threshold may occur in a dose-response curve can also
be understood by consideration of nitrite toxicity. There is a considerable
excess of oxygen-carrying capacity in normal blood, and even when some
hemoglobin becomes oxidized (and hence non-functional), the amount of oxygen
supplied to tissues is still in excess of the tissues' needs. Only when a
large amount of hemoglobin has been oxidized will there be an adverse effect
in tissues consuming oxygen (Figure 3-3).
3.1.2 Determination of Threshold Values
Regardless of the precise reason for the existence of a threshold, it is
important to determine the threshold value with some accuracy. This value is
often called the No-Observed-Adverse-Effect Level (NOAEL). The lowest value
where a significant adverse effect is first seen is the Lowest-Observed-Adverse-
Effect Level (LOAEL). A typical experiment designed to define the shape of
the dose-response curve and provide an estimate of the NOAEL and the_ LOAEL
involves exposing groups of experimental subjects (e.g., mice or rats) to a
3-1
-------
No Threshold
Value
/-Threshold Value
60 80
Dose, Arbitrary Units
100
120
FIGURE 3-1 HYPOTHETICAL DOSE-RESPONSE CURVES
The dose-response curve on the left illustrates a no threshold
effect; there is a response at all doses greater than zero.
The dose-response curve on the right illustrates an effect
with a threshold; no response occurs until some minimum
(threshold) dose Is exceeded.
3-2
-------
Cellular
GSH
Oxidized
Hemoglobin
Threshold
Value
FIGURE 3-2 CELLULAR DEFENSE MECHANISMS AS A BASIS FOR THE
THRESHOLD PHENOMENON
This figure illustrates the red blood cell's ability to
counteract nitrite's harmful oxidation of hemoglobin. The
cell produces a chemical (GSH) which preferentially reacts
with NO. , protecting hemoglobin. Only when the dose of
nitrite exceeds the cell's ability to produce GSH is
hemoglobin oxidized. The curves shown are hypothetical.
3-3
-------
Oxygen Transport Capacity
No Effect
on Tissues
Tissue Oxygen
Requirement
Oxidized
Hemoglobin
Oxygen
Deficiency
in Tissues
Threshola
Dose of N03, mg/kg
FIGURE 3-3 TISSUE RESERVE AS A BASIS FOR THE THRESHOLD PHENOMENON
This figure illustrates that oxidation of hemoglobin by
nitrite can reduce the oxygen transport capacity of blood
significantly prior to the occurrence of oxygen starvation
in tissues. The curves shovm are hypothetical.
-------
series of doses of the chemical. An example of the results of such an experi-
ment is shown in Table 3-1. There are several useful ways in which these data
may be displayed graphically. In prior examples (Figures 3-1, 3-2 and 3-3),
the dose was plotted on a linear scale. It is usually more convenient to plot
the doses on a logarithmic scale. This type of plot expands the low-dose
region of greatest interest, and compresses the high-dose region. The data
shown in Table 3-1 are plotted in Figure 3-4 using a logarithmic dose, scale.
Based on the smooth curve drawn through the data points in Figure 3-4, the
NOAEL value is about 0.3 mg/kg, and the LOAEL value is between 1 and 3 mg/kg.
Usually NOAEL and LOAEL-values are not derived by extrapolation between doses,
but are selected from the doses actually administered. In this case, the
NOAEL (the highest dose which did not produce a statistically significant
effect) is 1.0 mg/kg, and the LOAEL (the lowest dose which did produce a
statistically significant effect) is 3.0 mg/kg. (See Section 6.1 for
discussion of statistical significance.)
When a dose-response experiment is well-designed, it will span a series of
doses cn both sides of the NOAEL and-the LOAEL, permitting accurate
identification of each. Poorly designed experiments may only test doses that
are too high or too low (not spanning the NOAEL/LOAEL range). It is difficult
or impossible to derive reasonable estimates of the NOAEL or the LOAEL from
this sort of data.
A very important point which must always be remembered is that NOAEL and LOAEL.
values depend on the effect (endpoint) being measured. For "example, selecting
lethality, hemoglobin oxidation or reduction of red blood cell glutathione
levels as endpoints, NOAEL values might be 30, 3 or 0.3 mg/kg, respectively.
Choosing the most appropriate endpoint is not always simple and is discussed
in more detail in chapter 4.0.
3.1.3 Chemicals Without Thresholds
Some chemicals produce adverse effects that are characterized by a
dose-response curve with no threshold (see Figure 3-1). This is usually
because the cells that are affected have little or no "defense" against the
chemical and have little or no ability to repair or compensate for damage that
is done. For example, recent research suggests that there may be no threshold
for the effects of lead on the nervous system in infants and children.
Chemicals that produce cancer (carcinogens) are also considered to belong to
this group.
For chemicals with no threshold, any exposure is associated with some degree
of risk. A hypothetical relationship between dose and risk of cancer is_ghown
in Figure 3-5. The risk of cancer is gxpressed as a frequency, where 10 for
example, means that if one million (10 ) pgople wgre exposed for their
lifetime, on average one cancer case (1/10 - 10 ) would occur. Risk
decreases as dose decreases, but never reaches zero for chemicals without
thresholds. Figure 3-5 assumes a linear relationship exists between risk and
dose at low dose levels, but this is not known with certainty. This matter is
discussed more fully in section 6.4.
-------
TABLE 3-1 HYPOTHETICAL RESULTS OF AN EXPERIMENT TO DEFINE A DOSE-
RESPONSE CURVE FOR NITRITE-INDUCED HEMOGLOBIN OXIDATION
Number
of rats
Dose of
Nitrite, mg/kg
Oxidized Hemoglobin, %
(mean ± SD)
5
6
7
8
3
1
10
10
10
10
10
10
10
10
1
3
10
30
60
0
0.1
0. 3
80 ± 12(a)
3 ± 3
4 ± 2
4 ± 4
This table shows the results of a hypothetical experiment in which eighty rats
were divided in eight groups of ten and administered varying doses ot nitrite
by stomach tube. After one hour, the level of oxidized hemoglobin in the
blood of each rat was measured. Results presented are the mean and standard
deviation (SD) for each dose group.
(a) Statistically different (P<0.05) from control (group 1).
3-6
-------
LOAEL
NOAEL
30.0 100.0
Dose of NO2", mg/Kg
FIGURE 3-4 DOSE-RESPONSE CURVE (LOGARITHMIC SCALE)
FOR NITRITE-INDUCED HEMOGLOBIN OXIDATION
This figure is a plot of the hypothetical data shown in
Table 3-1. The logarithmic scale magnifies the low-dose
region of the curve, making it easier to identify NOAEL
and LOAEL values. The mean data values are shown by the
dots: ir.d the standard -i-viations are rfhown bv the
vertical bars.
3-7
-------
10"
1
10"'
— 4
Dose, mg/kg/day
FIGURE 3-5 HYPOTHETICAL RELATIONSHIP BETWEEN RISK AND DOSE
FOR CHEMICALS WITH NO THRESHOLD
This figure illustrates that for chemicals with no threshold,
as dose decreases so does risk (frequency of the effect),
but the frequency never reaches z^ro unless the dose is
zero. For example, a dose of 10 mg/kg/da^(l yg/kg/day)
would cause an effect in about 1/10,000 (10 ) exposed
organisms.
3-3
-------
While it would be ideal from the health effects perspective to set zero
exposure (and zero effect) as a goal, it is usually not technically or
economically feasible. Therefore, a judgement must be made as to what risk
level (greater than zero) is "acceptable." This choice is basically a cost-
benefit judgement made by public officials and by society, assessing the
magnitude of the risk from the chemical in comparison to risks from other
sources (accident, crime, flood, tornado, etc.), taking the cost and feasi-
bility of achieving the selected acceptable level into consideration.
Once a risk level is established (e.g., 10 ^), an acceptable exposure limit
can simply be determined from a curve similar to the one shown in Figure 3-5.
Unfortunately, it is not possible to make direct experimental measurements of
dose-response relationships at such low response rates (this would require
testing many millions of animals at each dose). Therefore, curves such as
shown in Figure 3-5 must be extrapolated from measurements made at high doses
(involving hundreds of animals at each dose). Performance of such
extrapolations is discussed in chapter 6.0.
3.2 Other Uses of Dose-Response Curves
In addition to providing a convenient means of determining no-effect levels
for chemicals, dose-response curves also help characterize the toxic properties
of a chemical and are useful in comparing the toxicity of several chemicals.
In characterizing or comparing dose-response curves, a number of quantitative
indices are commonly employed to aid in the description of the shape and
location of a dose-response curve. Table 3-2 lists some of these parameters
and describes their meaning and usefulness.
It is common to select the mid-point on the dose-response curve for comparison
of che toxicity of two chemicals. This is simply oeeause i; ii usually :Eci5r
to determine the mid-point accurately than to make an accurate estiniaca the
NOAEL. In general terms, the mid-point is referred to as the (the dose
which produces 50% of the effect). When a toxic effect is being measured, the
term is used. When the effect being measure-d is lethality, the terra
is used. Secause lethality is unambiguous and simple to neasure, an
value is often the first characterization or a chemical's toxicity to be
derived. However, lethality is usually too crude an index of adverse effects
to be useful in assessing toxicity of a chemical in the environment, and a
TD^q or based on a more sensitive endpoint should be used whenever
possible in evaluating or comparing chemical toxicity.
The slope of a dose-response curve is another important variable to be con-
sidered in assessing the toxicity of a chemical. Figure 3-6 shows two dose-
response curves, where chemical B is more toxic than chemical A at low doses,
even though the for chemical B is higher (less toxic) than for chemical
A. The reasons why the slopes of dose-response curves differ for different
chemicals involve the particular cellular mechanisms and metabolic functions
which they affect. While the details of these mechanisms and functions are of
concern to the toxicologist in understanding chemical toxicity, understanding
is not necessary for evaluating the degree of chemical toxicity. The important
concept to grasp is that chemicals with steep dose-response curves need to be
treated with caution since there may be only a small difference between a dose
producing no effect and a dose producing serious outcome.
3-9
-------
TABLE 3-2 QUANTITATIVE INDICES DERIVED FROM DOSE-RESPONSE CURVES
Index Definition/Use
ED Effective dose; may be used to mean any effect, but usually is
reserved to describe non-toxic or beneficial effects.
TD Toxic dose
LD Lethal dose
ED^q Dose which produces 50% of an effect (either 50% of the maximum
change or a Significant change in 50% of an exposed population).
TD50 Dose which produces 50% of a toxic effect.
LDj.q Dose which causes death in 50% of an exposed population.
EC,TC, Analogous to ED, TD, LD, but refers to the concentration of chemical
LC in water, and is used in studies of aquatic or marine organisms.
TI Therapeutic Index, a ratio describing the margin of safety between
beneficial and adverse effects.
3 —10
-------
100
E
3
°
o
o>
-------
A special category of chemicals is one with beneficial as well as adverse
effects. Fluoride ion is an example. Low levels of fluoride provide
protection to teeth from decay, but higher levels cause tooth disfigurement
and bone damage. Figure 3-7 shows hypothetical dose-response curves for two
chemicals (A and B) with both beneficial and adverse effects. When the curves
do not overlap significantly (Figure 3-7, Chemical A), it is relatively easy
to designate a dose that yields beneficial effects without adverse effects.
When the curves are close together (Figure 3-7, Chemical B), the maximum
beneficial level cannot be reached without producing adverse effects. The
closeness of the beneficial and the adverse dose response curves is described
by the Therapeutic Index (TI), which is the ratio of the mid-points of the two
curves:
TI - (TD50)/(ED50)
When this index has a value of 10 or higher the two curves are well separated
and doses for beneficial purposes are reasonably safe. Lower values of TI
indicate that the curves are close to each other and that doses for beneficial
purposes may result in adverse effects as well.
3-12
-------
Chemical
100
Beneficial
Effects
Curve
Toxic
Effects
Curve
o
0)
LLI
ED,
TD.
100
1.000
Dose of Chemical A, Arbitrary Units
Chemical
100
Beneficial
Effects
Curve
S
3
§
x
«
2
Toxic
Effects
Curve
ED,
TD,
o
-------
4.0
IMPORTANT PARAMETERS IN TOXICITY ASSESSMENTS
As discussed in section 2.2, Che essential question to answer in assessing the
toxicity of a chemical is: what is the maximum exposure to the chemical that
is safe? The answer to this question often depends on a number of factors,
including route of exposure, length of exposure, species and individual
characteristics of the exposed organism and the nature or endpoint of the
toxic effect being measured (Figure 4-1). This chapter describes why the
toxicity of a chemical usually depends on these factors and how these factors
are investigated and characterized in the toxicity assessment process.
*.1 Route of Exposure
For a chemical to exert a toxic effect on an organism, it must first gain
access to the cells and tissues of that organism. In humans, the major routes
by which toxic chemicals enter the body are through ingestion, inhalation and'
dermal absorption. The absorptive surfaces of the tissues involved in these
three routes of exposure (gastrointestinal tract, lungs, skin) differ from
each other with respect to the rate with which chemicals move across them. A
summary of the factors which influence absorption of chemicals through these
three routes of exposure is presented below.
4.1. 1 Ingestion
Ingestion brings chemicals into contact with the tissues of the gastro-
intestinal tract. The normal function of the gastrointestinal tract is
absorption of foods and fluids that are ingested. The gastrointestinal tract
is also effective in absorbing toxic chemicals that are contained in the food
-r v.t-s** The if ibsomr.ion generally depends on whether the chemical
is hydroTT'hilic (easily -soluble in water) on lipophilic (easily soluble in
organic solvents or fats). Lipophilic compounds (e.g., organic solvents) are
usually well absorbed, since the chemical can easily diffuse across the
nembrar.es of the cells which line the gastrointestinal tract. Hydrophilic
compounds ("e.g., metal ions) cannot cross the cell lining in this way, and
must be ''carried'' across by transport systems in the cells. The extent or* che
transport depends on how efficient the transport system is and how closely the
chemical resembles the normal compound for which the transport system is
intended.
If a chemical is a weak organic acid or base, it will tend to be absorbed by
diffusion in the part of the gastrointestinal tract in which it exists in its
most lipid-soluble (least ionized or polar) form. Since gastric juice in the
stomach is acidic and the intestinal contents are nearly neutral, the polarity
of a chemical can differ markedly in these two areas of the gastrointestinal
tract. A weak organic acid is in its least polar form while in the stomach
and, therefore, tends to be absorbed through the stomach. A weak organic base
is in it least polar form while in the intestine and, therefore, tends to be
absorbed through the intestine.
Another important determinant of absorption from the gastrointestinal tract is
the interaction of the chemical with gastric or intestinal contents. Many
chemicals tend to bind to food, ana so a cnemicai ingesceu xn Jooa oJcen
not absorbed as efficiently as when it is ingested in water. Additionally,
4-1
-------
TOXICITY ASSESSMENT
Route of Exposure
• Ingestion
• Inhalation
• Dermal Absorption
• Injection
Species Differences
* Absorption Rates
* Metabolic Activity
* Excretion Rates
Selection of Endpoints
• Dose Sensitivity
• Severity of Response
• Reversible/Irreversible
Effects
Duration of Exposure
• Accumulation
• Repair Processes
• Delayed Effects
individual Differences
• Genetic Traits
• Sex and Hormones
• Nutrition and Diet
» Age and Maturity
Common Endpoints
• Carcinogenicity
• Hepatotoxicity
• Mutagenicity
• Neurotoxicity
• Renal Toxicity
• Reproductive Toxicity
• Teratogenicity
FIGURE 4-1 FACTORS CONSIDERED DURINC TOXICITY ASSESSMENTS
-------
some chemicals may not be stable in the strongly acidic environment of the
stomach and others may be altered by digestive enzymes or intestinal flora
(bacteria which reside in the intestines) to yield different chemicals with
altered toxicological properties. For example, intestinal flora can reduce
aromatic nitro groups to aromatic amines, which may be carcinogenic.
4.1.2 Inhalation
Inhalation brings chemicals into contact with the lung. Most inhaled
chemicals are gases (e.g., carbon monoxide) or vapors of volatile liquids
(e.g., trichloroethylene). Absorption in the lung is usually high because the
surface area is large and blood vessels are in close proximity to the exposed
surface area. Gases cross the lung via simple diffusion, with the rate of
absorption depending on the solubility of the toxic agent in blood. If the
gas has a low solubility (e.g., ethylene), the rate of absorption is limited
by the rate of blood flow through the lung, whereas the absorption of readily
soluble gases (e.g., chloroform) is limited only by the rate and depth of
respiration.
Chemicals may also be inhaled in solid or liquid form as dusts or aerosols.
Liquid aerosols, if lipid-soluble, will readily cross the cell membranes by
passive diffusion. The absorption of solid particulate matter is highly
dependent upon the size and chemical nature of the particles. The rate
of absorption of particulates from the alveoli is determined by the compound's
solubility in lung fluids, with poorly .soluble compounds being absorbed at a
slower rate than readily soluble compounds. Certain small Insoluble particles
may remain in the alveoli indefinitely. Larger particles (2 to 5 microns) are
deposited in the tracheobronchiolar regions of the lungs where they are
cleared by coughing and sneezing or they are swallowed and deposited in the
gastrointestinal tract. Particles of five microns or larger are uouall/
deposited in the nasopharyngeal region where they are subsequently either
expelled or swallowed.
4.1.3 Dermal Absorption
Absorption of toxicants through the epidermal layer of the skin is hindered by
the densely packed layer of horny, keratinized epidermal cells. Absorption of
chemicals occurs much more readily through scratched or broken skin. There
are significant differences in skin structure from one region of the body to
another (palms of hands versus facial skin), and these differences further
influence dermal absorption.
Absorption of chemicals by the skin Is roughly proportional to their lipid
solubility, and can be enhanced by application of the chemical in an oily
vehicle and rubbing the resulting preparation into the skin. Some
lipid-soluble compounds can be absorbed by the skin in quantities sufficient
to produce systemic effects. For example, carbon tetrachloride can be
absorbed by the skin in amounts large enough to produce liver injury.
-------
4.1.4 Exposure by Injection
In toxicity assessment, the route of exposure employed in experimental animal
studies is normally chosen to be the same as the anticipated route of exposure
of humans to the specific chemical. However, in studying chemicals toxicolo-
gists frequently administer chemicals to laboratory animals by injection
(parenterally), the most common routes being subcutaneous (s.c.), intraperi-
toneal (i.p.), intramuscular (i.m.) and intravenous (i.v.). These routes are
employed because they are often more convenient and yield more reproducible
results than oral, dermal or inhalation routes of exposure. However, results
from studies of this sort must be interpreted with caution, since parenteral
administration bypasses the normal absorptive processes, and a parenteral dose
may be more toxic than the same dose given by ingestion, inhalation or dermal
application.
4.2 Duration/Frequency of Exposure
The toxicity of many chemicals depends on the length of time over whicn
exposure has occurred. There are several reasons for this dependency. First,
some chemicals are not readily eliminated from the body, so that continued
exposure to low doses (each too small to produce an effect) may lead to
accumulation of the chemical in the body at levels which are high enough to
produce adverse results. For example, cadmium is strongly retained in the
body and tends to accumulate in the kidney. When levels become high enough
(usually after many years), kidney dysfunction begins to occur.
A second reason why toxic effects may depend on duration of exposure is
related to the ability of cells to repair themselves. When an injury to a
cell cannot be quickly reversed by repair processes, there is a candency for
;ha injury zo accumulate in th= c = ll ^2 i :f i~crza3i~s
duration. Thus, a dose of a chemical that causes a small, but Irreversible,
injury may have no immediately apparent effect, but a clear adverse response
may develop with continued exposure.
Final!", sons adverse effects sinroLy rs-nin? «n extended period of tine to
develop, even though they might be the result of exposure months or years
earlier. Lead exposure, for example, may impair the development of the
nervous system in young children, but this effect requires an extended period
of exposure, and does not become apparent for several years. Similarly, the
development of tumors following exposure to a carcinogen may take months or
years to occur.
4.3 Species of Exposed Organism
It is generally true that if a chemical is found to be toxic in one species of
organism (e.g., rat) it will also be toxic in similar organisms (e.g., other
mammals, including humans). However, there are often significant differences
in the sensitivity of different species to a chemical, and sometimes there are
qualitative differences in the types of effects that occur. The reason for
these differences among species is usually related to differences in the
absorption or metabolism of the chemical or to differences in anatomic
f ijrj n 1. OH
-------
4.3.1 Absorption Differences Among Species
The rate of absorption of chemicals across the skin, lungs or gastrointestinal
tract is determined primarily by the properties of the cells at the surfaces
of these tissues, and there are some significant differences in these cells
among species. For example, the skin of the rat and rabbit are more
permeable, the skin of the cat is less permeable, and the skin of the pig,
guinea pig, and monkey are similar in permeability characteristics to those
observed in humans. Additionally, physical and chemical conditions which
influence gastrointestinal absorption may also differ among species. For
example, gastrointestinal transit time, surface area to volume ratio and pH in
various parts of the gastrointestinal tract often differ among species.
Finally, the bacterial populations in the gastrointestinal tract vary among
species. Some bacteria may convert one chemical into another one that is more
or less absorbable and thus alter the apparent toxicity of the chemical, or
they may convert a nontoxic chemical into a toxic one.
4.3.2 Differences in Chemical Metabolism Among Species
Metabolism is the name applied to any chemical reaction which a chemical may
undergo while in the body. The liver and kidney are especially active in
these reactions, but metabolism of a chemical may occur in any tissue. Nearly
all chemicals are modified by one or more reactions, but the nature and extent
of these reactions may vary widely among different organisms. The rate of
metabolism of chemicals is often the limiting" step in detoxification and/or
excretion of chemicals, so differences in metabolic activity can markedly
influence how long a toxic chemical endures in the body. In addition, meta-
bolism of a chemical may sometimes generate a more toxic chemical. For
— - . ¦ - o 2 ^ 1 n ** r* *4 *r\ c 31 s , "ZSTbi — S* ins 3. 3.nd
hamstars, while i: is poorly methylated in nice, rabbits, rats and humans.
Since methylation (addition of a methyl group, CH ) may increase the toxicity
of pyridine, the effects produced by equal doses In these two groups of
animals mav be more adverse in the animals which methylate pyridine efficiently.
4.4 Individual Characteristics of Exposed Organisms
Just as there are significant differences among species with respect to toxicity
of some chemicals, so there may also be significant differences among subgroups
of a population (as a function of sex, race or age) and among individuals in a
population. The principal factors which underlie these variations in sensitivity
are outlined below.
4.4.1 Genetic Traits
The genetic makeup of an individual is expressed by the presence or absence of
key enzymes in cells, and differences in these enzymes underlie much of the
variation in susceptibility of individual members of a particular strain or
population. For example, the variation in the susceptibility of some rabbits
to the toxic effects of atropine is explained by the presence of an enzyme,
atropine esterase, in the blood of the resistant animal.
4-5
-------
It should be recognized that laboratory animals used in most experimental
studies of chemical toxicity are highly inbred in order to achieve a very
uniform genetic composition, so variations in chemical sensitivity between
individual animals of the same species are usually small. Conversely, humans
are genetically highly heterogeneous and variations between individuals, even
of the same age, sex and race, can be significant. For example, two
sub-groups have been identified in the human population with the respect to
the ability to acetylate certain chemicals; slow and rapid acetylators.
Acetylation (addition of the acetyl group, CH^CO-) is significant because it
can lead to detoxification of some chemicals. Slow acetylators have less
N-acetyltransferase in their livers than rapid acetylators, and this is the
enzyme that catalyzes the acetylation process. Therefore, slow acetylators
are more likely to develop toxic effects to certain chemicals.
U.U.2 Sex and Hormonal Status
Differences in toxicity between sexes have been demonstrated in studies on the
effects of chloroform, benzene and some organophosphate insecticides. For
example, female mice show little response to chloroform exposures that are
lethal for males. The difference has been shown to be under direct endocrine
(hormonal) control. As another example, female rats and rabbits are more
susceptible to the toxic effects of parathion and benzene, respectively, than
are male rats and rabbits. These sex-related effects become reversed after
castration and administration of hormones. Pregnancy, with its increased
hormonal activity, has been shown to markedly increase the susceptibility of
mice to some types of pesticides, and similar-effects have been reported for a
lactating animal exposed to heavy metals. Hyperthyroidism (excessive
secretion of thyroid hormone) and h.yperinsulinism (excessive secretion of
insulin ) Tna'1- also alter the suscertibilitv of animals, including humans. to
toxic zhimicals.
U.U.I Nutritional Status and Dietary Factors
Humans are able to achieve large adjustments in the absorption and metabolism
of foods and minerals to compensate for fluctuations in dietary intake levels.
These metabolic adaptations frequently influence the absorption and/or
metabolism of toxic chemicals as well. For example, long-term ingestion of a
diet low in essential minerals (iron, calcium, zinc) leads the body to
increase absorption and retention of these minerals. However, along with this
adaptation to retain essential minerals, the absorption of toxic metals
(cadmium, barium) also increases. Generally, low calorie or protein-deficient
diets result in increased sensitivity to a number of toxic chemicals.
U.U.k Age and Maturity
Some chemicals are more toxic to one age group than another (usually being
more toxic to infants and children than adults). In some cases this is only
because infants and children drink and eat proportionately larger amounts than
do adults, and thereby ingest proportionately larger doses. However, infants
and children may be inherently more sensitive to chemicals for reasons related
to the development process. For example, lead ingestion has much more severe
effects on the nervous system or infants and cnnaren tnen it does on acui;s.
4-6
-------
Additionally, the ability of the young to metabolize and detoxify chemicals is
usually less than for adults. Elderly humans may be more sensitive to some
chemicals, since the detoxifying capacity of liver and the excretory capacity
of kidney tend to decrease in old age.
4.5 Toxicological Endpoints
Exposure of an organism to a chemical often results in multiple effects. For
example, long-term exposures to dioxin results in hepatotoxicity (liver
toxicity), genotoxicity (chromosomal damage), teratogenicity (structural/func-
tional abnormality), fetotoxicity (injury to developing fetuses) and carcino-
genicity (growth of malignant tumors). Effects which are measured by the
toxicologist as an index of a chemical's toxicity are called "endpoints". The
criteria for identifying the endpoint most appropriate for use in toxicity
assessment include dose sensitivity, the severity of the effect, and whether
the effect is reversible or irreversible.
4.5.1 Criteria for Selection of Endpoints
4.5.1.1 Dose Sensitivity
The most appropriate endpoint for use in the toxicity assessment process is
usually the one in which a measurable change can first be detected in response
to increasing doses. For example, pyridine is toxic to the central nervous
system (CNS), the liver and the kidney. However, CNS toxicity can be
demonstrated at much lower doses than adverse kidney and liver effects. In
studying pyridine, then, CNS effects are appropriate as the most sensitive
endpoint.
^.5. 1.2 Severity of Response
The selection of a toxicological endpoint is sometimes based on the extent of
damage co a particular organ following exposure. A toxic chemical may produce
harmful effects in a number of organs, but the severity of the response r.av be
quite different. For example, carbon tetrachloride exposure -nay result in
mild damage to the kidney, but severe damage and loss of function in the
liver. In studying carbon tetrachloride, then, effects on the liver are the
most appropriate endpoint.
Sometimes a low dose of a chemical may produce an effect that is not in itself
clearly adverse. For example, a low dose of acrylamide may cause slowed
axonal transport in nerve cells without measurably affecting the ability of
the cells to carry nerve impulses.
To distinguish between detectable effects which are adverse and those which
are not, the term LOEL (Lowest-Observed-Effect Level) is used, as distinct
from LOAEL (Lowest-Observed-Adverse-Effect Level). Similarly, NOEL (No-
Observed-Effect Level) implies no detectable effect of any sort, while NOAEL
(No-Observed-Adverse-Effect Level) may include some effect which is judged not
to be adverse.
4-7
-------
The decision as to whether an effect is adverse or not must be judged on the
basis of whether the change is an early indication of a more serious conse-
quence or whether the change is not of significant concern.
4.5.1.3 Reversible Versus Irreversible Effects
Some toxic effects are reversible and others are not. In a tissue that has a
strong ability to regenerate (e.g., the liver), most injuries are reversible,
whereas injury to the CNS is largely irreversible, since specialized cells of
the CNS cannot divide or be replaced. Carcinogenic effects of chemicals are
also irreversible. Irreversible effects are often chosen as toxicological
endpoints since these effects are likely to produce serious consequences
following chronic (long-term, low level) exposure to a chemical.
4.5.2 Common Toxicological Endpoints and Measuring Techniques
Table 4-1.lists endpoints that are commonly used to assess the toxic effects
of a chemical, along with the experimental means of measuring such effects.
More detailed descriptions of these endpoints are given below.
4.5.2.1 Carcinogenicity
Cancer is a complex group of diseases whose causes are not yet fully under-
stood, but there is ample evidence that some chemicals can cause or promote
certain types of tumors in animals or humans. The carcinogenic potential of a
chemical can be measured with lifetime animal bioassays, short-term carcino-
genicity tests (with bacterial or cultured mammalian cells), or limited in
vivo bioassays. Each of these methods is associated with certain advantages
and disadvantages, as discussed below.
Standard lifetime animal bioassays are long-term experiments conducted to
measure the effect of a chemical on frequency of tumor occurrence.
Typically, large groups of animals (at least 50 per sex per dose) are exposed
:o the ;hemicai for cneir lifetime, and the number and types of tumors occur-
ring in exposed animals are compared to control animals.
These studies are considered to be the most predictive of carcinogenicity
screening tests. However, substantial controversy exists over certain standard
practices used in the bioassays. For example, to compensate for the relative
insensitivity of these studies, the maximum tolerated dose (MTD) is frequently
used to maximize the likelihood of detecting carcinogenicity. The use of MTD
is controversial because high doses of a chemical may produce physiological
conditions that affect the induction and development of tumors. Normal
detoxification and repair mechanisms may be overwhelmed by the use of the MTD,
or different absorption, distribution, metabolism or excretion may result from
the use of the MTD. These events might result in a response at the MTD that
may not be indicative of effects at lower exposure levels.
A similar controversy exists over the use of strains of test animals that are
very susceptible to carcinogens. The purpose of using these animals is to
4-8
-------
TABLE 4-1 MEASUREMENT OF COMMON TOXICOLOGICAL ENDPOINTS
Toxicological Endpoint Parameters Measured
Behavioral Toxicity
Carcinogenicity
Hematologic Toxicity
Hepatoxicity
Inhalation Toxicity
Mutagenicity
Neurotoxicity
Renal Toxicity
Reproductive Toxicity
Teratogenicity
Motor function (motor activity,
coordination strength), sensory
function (vision, audition),
integrative systems (learning and
memory).
Tumor frequency in tissues, detected
by gross observation or histological
examination.
Hematocrit, hemoglobin levels,
changes in cellular components
(erythrocytes, leucocytes,
platelets), plasma components, and
foreign substances.
Gross and microscopic examination,
organ weight, liver function (bile
formation, lipid metabolism, protein
metabolism, carbohydrate metabolism,
metabolism of foreign compounds,
serum enzyme activities).
Gross anatomy, microscopic and
uxtrastructural anatomy, changes in
function.
Chromosome alterations, bacterial
mutations, DNA damage.
Gross observation, clinical
evaluation, neurological exams,
behavioral tests, neurohisto-
pathological tests, neurochemical
tests.
Urinalysis, function tests
(clearance, glomerular filtration
rate), gross and microscopic
examinations, organ weight.
Fertility, litter size and survival,
gestation survival, postnatal body
weight.
Gross abnormalities, skeletal and
abnormalities, functional/behavioral
deviations.
4-9
-------
increase the ability to detect carcinogenic potential in chemicals. However,
these sensitive strains may have very high spontaneous tumor frequencies, and
the meaning and validity of a positive test result in such a sensitive strain
is not entirely clear. Because of these uncertainties, it is very desirable
to perform lifetime carcinogenicity bioassays using two or more different
animal species.
There are two major types of short-term carcinogenicity tests used to indicate
carcinogenic potential: mutagenicity tests and transformation tests using
cultured mammalian cells. Mutagenicity experiments are often used to evaluate
the potential for inducing tumors because of basic similarities in the
postulated molecular mechanisms of chemical carcinogenesis and mutagenesis.
Some mutagenicity tests, especially the Ames test, have been extensively
validated and shown to correlate very well with known carcinogens, but there
is still a significant frequency of false-positive or false-negative results
(approximately 10% each for the Ames test). Positive results in mutagenicity
tests support other experimental findings of carcinogenic potential and are
generally considered tc provide suggestive evidence of carcinogenic.hazard.
They do not constitute definite proof of a chemical's carcinogenicity in
humans, nor do negative results rule out the possibility of carcinogenic
potential.
A major disadvantage of mutagenicity tests using bacterial test systems is the
basic biological differences between bacterial cells and human cells, making
extrapolations to human health effects somewhat tenuous. Testing in mammalian
cells provides a stronger basis for extrapolating to human health effects, but
the test methods are not as well developed or validated as those using bacteria.
The primary short-term test of this sort is based on mammalian cell transforma-
tion. Transformation occurs when cultured cells develop uncontrolled growth,
an event analogous co the formation of a cumor an organism. A number zz
transformation tests using mammalian cells nave been developed in recent years
and are in widespread use. The cells are treated with the chemical in question
and the transformation frequency is measured. Cell transformation is usually
detected by observing changes in the cultured cells, and is confirmed by
injecting the transformed calls into animals where they becnme nalignant
tumors. A major disadvantage is that the carcinogenic potential or a chemical
may depend on its metabolism in the living organism, with one or more metabolic
products being more carcinogenic than the original chemical. In a case such
as this, a transformation test might yield negative results, while positive
results would be obtained in a lifetime animal bioassay.
Limited in vivo bioassays can provide evidence of the tumorogenic potential of
a chemical without the great time investment and expense required for a
lifetime bioassay. These tests generally yield results in 30 weeks or less
and use mice or rats as test animals. Examples of limited in vivo bioassays
include skin tumor formation in mice, breast cancer induction in rats, or
altered liver foci (an early step in liver tumor formation) in mice or rats.
While limited in vivo bioassays are not an adequate substitute for lifetime
bioassays, they are more useful as predictors of carcinogenicity than
short-term tests, and positive results in well-designed and executed limited
in vivo bioassays are additional supportive evidence of carcinogenic hazard.
4-10
-------
It is not uncommon that a chemical will yield different results in different
tests of carcinogenic potential (positive in some tests, negative in others).
In such cases, the probability that the chemical is a human carcinogen is
determined by a "weight-of-evidence" approach (USEPA 1984). In general,
positive results in animal systems or positive finding in epidemiological
studies are required before a chemical is classified as a probable human
carcinogen. This may be supported by positive results in bacterial or cell
system tests, but positive results in these systems are insufficient alone.
4.5.2.2 Hepatotoxicity
Liver damage is a frequent response to exposures to toxic chemicals. Since
the liver is such a vital organ, a variety of procedures have been developed
over the years to assess extent of liver damage. Because the liver has
considerable reserve capacity, tests that measure its ability to perform its
functions may not reveal an effect until the liver is already extensively
damaged. A more sensitive test involves measurement of liver enzymes in blood
serum. This test is based on the observation that when liver cells are
damaged, some of the active enzymes within the cells escape into the blood.
This increase in liver enzymes can be measured simply by collecting a sample
of blood and measuring enzyme activity. A disadvantage of this test is that
liver enzymes do not endure very long in the blood, and so only an on-going
injury can be detected. Finally, evidence of liver damage may be detected
both during and well after a chemical-induced injury by microscopic
examination of the liver for signs of abnormality. Therefore, microscopic
examination for histological changes is another excellent endpoint.
4.5.2.3 Mutagenicity
Mutagenasi,: is the induction of caanges ip generic T.a serial cr.ar --"i
transmitted during cell division. If mutations are present in the ^er.eci-
material of eggs or sperm, the fertilized ovum may not be viable. A mutation
may also result in congenital abnormalities or death of a fetus at a later
developmental period. There are a number of powerful tests for mutagenic
potential of chemicals, no:: ir.vclvir.g bacteria or other cells in culture.
For example, the Ames test measures the frequency of a certain type of
mutational event in the bacterial species Salmonella typhimurium. Other
valuable tests examine Chinese hamster ovary (CHO) cells for alterations in
chromosome structure, or determine whether unscheduled DNA synthesis (a strong
indicator of damage to the DNA) is occurring in other cultured mammal cells.
4.5.2.4 Neurotoxicity
The nervous system is of special toxicological concern, since chemical-induced
injury to nerve cells is often irreversible and may lead to adverse health
effects. There are many means of measuring nervous system functions, includ-
ing tests of reflexes, coordination, conditioned responses in animals and
intelligence (IQ) tests in humans. In addition, there are sophisticated means
of analyzing the status of individual nerves by measuring the rate at which
they transmit nerve impulses or the rate at which they synthesize and transport
cellular materials.
-------
The EPA has published guidelines that focus on delayed neurotoxicity as an
endpoint. Delayed neurotoxicity is a syndrome in which damage to the peri-
pheral nervous system and some portions of the CNS may result in paralysis.
The domestic hen is typically the species chosen for evaluation of delayed
neurotoxicity. In the acute delayed neurotoxicity test, a single dose of the
test material is administered orally to groups of adult domestic hens. The
hens are observed daily for at least 21 days for behavioral abnormalities,
ataxia (inability to coordinate muscles) and paralysis. Selected sections of
nervous tissues are also examined histopathologically. In tests for sub-
chronic delayed neurotoxicity, groups of hens are administered the test
substance orally for 90 days, followed by an observation period of seven days.
As in the acute studies, the hens are observed daily for behavioral abnor-
malities, ataxia and paralysis, and selected sections of nerve tissue are
examined histopathologically.
4.5.2.5 Renal Toxicity
Damage to the kidneys is another common and serious consequence of exposure to
a toxic chemical. As with the liver, techniques to assess kidney injury may
include functional tests in the intact animal, along with direct histological
examination of kidney tissue. Urinalysis offers another convenient and
sensitive means of detecting kidney damage. For example, detection of
substances not normally present in urine (proteins, cells or cell fragments,
glucose) is strong evidence that the kidney has .been injured.
4.5.2.6 Reproductive Toxicity
Fertility and reproductive toxicity studies are usually performed in rats or
mice at dose levels that produce no overt toxicity in the exposed adults. In
a epical study, the 22I2 parart is exposed to a chemical for *0 to 80 days
and the female for 14 days prior to mating. The percentage of females that
become pregnant is determined. The number of stillborn and live offspring,
and their weight, growth, survival and general condition during the first
three weeks of life are also recorded.
The perinatal (during late pregnancy) and lactational (during nursing)
toxicities of chemicals may be measured in a similar fashion. Pregnant female
rats are exposed to the chemical from the fifteenth day of gestation to the
time of weaning. Parameters measured may include all of those above, as well
as analysis of milk for presence of the chemical.
4.5.2.7 Teratogenicity
Teratology is defined as the study of functional or physical defects induced
during development of an animal from the time of conception to birth.
Teratogenic studies are usually performed in rats and/or rabbits with doses of
the test chemical that produce no maternal toxicity. Teratogens are most
effective when administered during the period of organogenesis, so pregnant
females are usually exposed on days 6 to 15 of gestation. Prior to delivery,
some females are sacrificed and examined for the number of fetal implantations
in the uterus. Dead and living fetuses are counted, weighed and examined for
gross malformations. These fetuses are examined microscopically for more
4-12
-------
subtle effects, and some are cleared of soft tissue and examined for skeletal
abnormalities. Since teratogens can produce functional as well as morphologic
changes, offspring of other females are sometimes monitored after delivery for
changes in behavior or development.
4.6 Key Guidance and Implementation Documents
USEPA. 1984. U.S. Environmental Protection Agency. Proposed guidelines for
carcinogenic risk assessment. Fed. Regist. 49-46294.
USEPA. 1984. U.S. Environmental Protection Agency. Proposed guidelines for
mutagenicity risk assessment. Fed. Regist. 49-46314.
USEPA. 1982. U.S. Environmental Protection Agency, Office of Toxic
Substances. Analyses of limited bioassays as potential carcinogenicity
screening tests. Washington, DC: U.S. Environmental Protection Agency
68-01-6196.
USEPA. 1981. U.S. Environmental Protection Agency, Office of Pesticides and
Toxic Substances. Health effects test guidelines, Washington, DC: U.S.
Environmental Protection Agency 560/6-82-001.
USEPA. 1979. U.S. Environmental Protection Agency. Proposed health effects
test standards for Toxic Substances Control Act test rules and proposed good
laboratory practice standards for health effects. Fed. Regist. July 26, 1979,
: 4405-1-44093.
USEPA. 1978. U.S. Environmental Protection Agency, Office of Toxic
Substances. Short-term tests for health and ecological effects. Washington,
DC: U.S. Environmental Protection Agency: oC0/9-78-037.
4.7 Background References
Doull J. 1980. Factors influencing toxicology. In: Doull J., Xlaasser. CD,
Air.aur >10, eda. Cisarecc and Doull'3 toxicology. The basic science of poi-
sons , 2nd ed. New York: Macmillan Publishing Co., Inc., pp 70-83.
D'Souza J, Caldwell J, Smith RL. 1980. Species variations in the
N-methylation of pyridine. Xenobiotica 10:151.
Fisher PB, Weinstein BI. 1981. In vitro screening tests for potential
carcinogens. In:. Sontag JM, ed. Carcinogens in industry and the environ-
ment. New York: Marcel Dekker, Inc., pp 113-166.
Freeman AE. 1978. In vitro testing of chemical carcinogens: an overview.
In: Berky J, Sherrod PC, eds. Short term in vitro testing for carcinogene-
sis, mutagenesis and toxicity. Philadelphia: Franklin Institute Press, pp
8-22.
Gilman AG, Mayer SE, Melmon KL. 1980. Pharmacodynamics: mechanisms of drug
action and the relationship between drug concentration and effect. In:
Jij-iTiati, no, oooomar. -o, Ji.ja.i • - ...a.—."5-"
cal basis of therapeutics, 6th ed. New York: Macmillan Publishing Co., Inc.,
pp 28-39.
4-13
-------
Klaassen CD. 1980. Absorption, distribution and excretion of toxicants. In:
Doull J. Klaassen CD, Amdur MO, eds. Casarett and Doull's toxicology. The
basic science of poisons, 2nd ed. New York: Macmillan Publishing Co., Inc.,
pp 28-55.
Klaassen CD. 1980. Heavy metals and heavy metal antagonists. In: Gilman
AG, Goodman LS, Gilman A, eds. The pharmacological basis of therapeutics 6th,
ed. New York: Macmillan Publishing Co., Inc., pp 1615-1637.
Klaassen CD. 1980. Nonmetallic environmental toxicants: air pollutants,
solvents and vapors and pesticides. In: Gilman AG, Goodman LS, Gilman A,
eds. The pharmacological basis of therapeutics, 6th ed. New York: Macmillan
Publishing Co., Inc., pp 1638-1659.
Klaassen CD. 1980. Principles of toxicology. In: Gilman AG, Goodman LS,
Gilman A, eds. The pharmacological basis of therapeutics, 6th ed. New York:
Macmillan Publishing Co., Inc., pp 1602-1614.
Klaassen CD, Doull J. 1980. Evaluation of safety: toxicologic evaluation.
In: Doull J, Klaassen CD, Amdur MO, eds. Casarett and Doull's toxicology.
The basic science of poisons, 2nd ed. New York: Macmillan Publishing Co.,
Inc., pp 11-27.
Loomis TA. 1978. Essentials of Toxicology. Philadelphia: Lea and Febiger.
Mayer SE, Melmon KL," Gilman AG. 1980. Introduction; the dynamics of drug
absorption, distribution ana elimination. In: Gilman AG, Goodman LS, Gilman
A, eds. The pharmacological basis of therapeutics, 6th ed. New York:
Macmillan Publishing Co., Inc., pp 28-39.
•S
Meal RA. 1980. Metabolism of toxic substances. In: Doull J, Klaassen CD,
Amdur MO, eds. Casarett and Doull's toxicology. The basic science of poi-
sons, 2nd ed. New York: Macmillan Publishing Co., Inc., pp 56-69.
0'Flaherty EJ. 1981. Toxicants and drugs: kinetics and dynamics. New York:
John VJiley and Sons, Inc.
Weisburger JH, Williams GM. 1980. Chemical Carcinogens. In: Doull J,
Klaassen CD, Amdur MO, eds. Casarett and Doull's toxicology. The basic
science of poisons, 2nd ed. New York: Macmillan Publishing Co., Inc., pp
11-27.
A-i4
-------
5.0
TYPICAL PROTOCOLS USED IN TOXICOLOGICAL STUDIES
As discussed in chapters 3.0 and 4.0, assessment of the toxicity of a chemical
involves identification of the adverse effects which the chemical causes and
systematic study of how these effects depend upon dose, route and duration of
exposure and test organism. This information is derived from studies which
may be divided into four general categories:
• Studies in laboratory animals evaluate the toxicity of a chemical with
special reference to predicting the toxicity in humans.
• Clinical studies are case-by-case investigations of the symptoms and
dis'eases in humans who are exposed to a toxic chemical at doses high
enough to require medical attention.
• Epidemiological studies seek ,to determine whether a correlation exists
between chemical exposure and frequency of disease or health problems
in large groups of human populations.
• Ecotoxicological studies assess the toxic effects of chemicals on
indigenous aquatic and terrestrial plants and animals.
This chapter describes the usual experimental designs (protocols) used in
these studies.
5. 1 Studies in Laboratory Animals
Table 5-1 summari2es protocols that might be used in testing the toxicity of a
chemical in laboratory animals. In order to determine how the effects of a
chemical depend on exposure levels, all studies involve administration of a
series of doses. To investigate how the effects of a chemical depend on
duration of exposure, chemicals are administered for one day or in one dose
(acute), for 5 to 90 days (subchronic) and for long periods (2 years to
lifetime). To determine how effects may depend on the characteristics of the
test organism, che chemical is administered to boch sexes of two or more
species. To identify the cells and tissues that are affected by the chemical,
a broad range of endpoints are evaluated for chemical induced changes.
5.1.1 Acute Studies
In an acute study, animals are given a brief exposure to the chemical (a
single oral dose, a 4-hour inhalation exposure or a 24-hour dermal exposure)
and are observed for subsequent effects. The chemical is usually tested in
two animal species, most often rats, mice, dogs or guinea pigs.
Often the first'endpoint to be measured is lethality. Determination of oral
LD n values requires the testing of four to six dose levels with five to ten
animals/sex/dose level. The test chemical is administered once to each test
animal and the number of animals that die, the time of death, and toxic signs
observable directly and at necropsy (animal autopsy) are recorded. Acute
dermal and inhalation studies are similar, except that emphasis is placed on
signs oi injury co che Sitin or lungs.
5-1
-------
TABLE 5-1 TYPICAL TESTING PROTOCOLS IN ANIMALS
Type of Study
ACUTE
Oral
Dermal
i
to
Test
Species
Rats, mice,
guinea pigs
Dogs
Rabbits
Inhalation Rats
Number
of Dos
Levels
?a)
4-6
3-4
4-5
Approximate No.
of Animals of
Each Sex Per
Dosing. Heglmen Do^e Level
Single dose
Single ilose
Single ..ppli-
catiou for
24 hour.-.
4-Hour
exposure
5-10
2-3
Observation
Period Typical Observations
14 days
14 days
14 days
(evaluated
24 hours ,
7 days and
14 days)
14 days
Survivors, body weight
changes (day 14), gross
hlstopathology and
toxicities, clinical
chemistry (dog only).
Survivors, body weight
changes (day 14), gross
toxicity and hlstopath-
ology, especially of
skin.
Survivors, body weight
changes (day 14), gross
toxicity and hlstopath-
ology, especially of
lungs.
cont inued-
(a) 11.eludes a dose level of zero (control).
-------
Table 5-1
- continued
Number
Test of Dos^
Type of Study Species Levels
SUBC1IU0NIC
Oral Rats 3 - A
Dogs 3-4
l'ennal Rabbits 3-4
inhalation Rats 3-4
CHRON.C
oral Rats 3-4
Dog6 3-4
inhalation Rats 3-4
(b) Minimum of 5 days per week.
Approximate No.
of Animals of
Each Sex Per Observation
Dose Level Period Typical Observations
20 90 days
6 90 days
10 90 days
10 90 days
Survivors, body weight
changes, diet consump-
tion, urinalysis,
hematology, clinical
chemistry, gross and
microscopic examina-
tion of major tissues
and organs.
50 2 years
6 2 years
50 2 years
Survivors, body weight
changes, diet consump-
tion, urinalysis,
hematology, clinical
chemistry, gross and
microscopic examina-
tion of major tissues
and organs.
-------
Acute toxicity studies are useful in (1) providing a quantitative measure of
acute toxicity (LD^) for comparison with other chemicals, (2) identifying the
functions or organs most severely affected and (3) defining the appropriate
doses to be used in longer-term studies (subchronic, chronic).
5.1.2 Subchronic Studies
Following acute toxicity testing, the chemical is next tested for toxicity
following subchronic exposure, usually between five and 90 days. Doses are
administered daily to both sexes of at least two species (e.g., rats and dogs)
by the expected route(s) of human exposure. At least three dose levels of the
test chemical are used. These doses are selected to span the full
dose-response range and to define the NOAEL or LOAEL if possible. A variety
of parameters are monitored as described in Table 5-1. If one or more unique
endpoints are recognized as being especially characteristic of the chemical's
effects, more detailed attention is focused on then.
5.1.3 Chronic Studies
Chronic studies are performed similarly to subchronic studies, except that
emphasis is placed on searching for evidence of slowly emerging adverse
effects (e.g., cancer). Doses are generally selected to be low enough that
most animals survive the full exposure period. Larger numbers of animals are
employed to obtain statistically significant results in endpoints that
naturally vary among individuals (e.g., longevity, tumor frequency).
In summary, testing protocols in animals are designed to identify the
principal adverse effects of a chemical as a function of dose, route of
exposure, species and sex of test animals and duration of exposure. When
carefully performed, these studies will yiela w0aŁL or LOAZL values for
aost sensitive noncarcinogenic enapoint for each route and lengtti of exposure,
and, if carcinogenic, a series of exposure levels corresponding to excess
cancer risks of 10 , 10 and 10
5.2 Clinical Studies in Huaans
The medical' community often reports detailed descriptions of human diseases
and other health problems resulting from exposure to toxic chemicals.
Exposures may be accidental (e.g., a farmer applying pesticide without proper
protection) or intentional (e.g., suicide or homicide). In view of the
difficulties and uncertainties in extrapolating toxicological information from
animals to humans, this sort of direct toxicological observation is especially
valuable in characterizing toxic responses of clinical significance in humans.
Unfortunately, clinical studies are rarely sufficient to provide a complete
description of a chemical's toxicity. This is because clinical observations
are usually available on only a small number of individuals, and quantitative
information on the exposure levels causing the effect are rarely known. This
absence of quantitative dose information diminishes the usefulness of clinical
studies in estimating a NOAEL or LOAEL in humans. Additionally, even when
exposure levels are known, these levels are usually high on the dose-response
curve and so are not of direct use in defining the NOAEL in humans. Finally,
clinical scutiies, Jo ,wc awwCur.c for Iszzzzc sush is 2gs, smoking or previous
exposure to other chemicals.
-------
5.3 Epidemiological Studies
Epidemiological studies seek to determine whether or not correlations exist
between the frequency or prevalence of a disease or health condition in human
populations and some specific factor such as concentration of a toxic chemical
in the environment. The major advantages of epidemiological studies are that
they are based on large numbers of humans and exposure levels are usually
sub-clinical. Thus the data are directly relevant, with no need to
extrapolate from data in animals or to make projections from a small number of
humans exposed to a high dose of the chemical.
5.3.1 Types of Epidemiological Studies
This section provides a discussion of the three basic types of epidemiological
studies and the considerations associated with data derived from such studies.
5.3.1.1 Retrospective Studies
In most instances the most feasible approach in terms of cost, time and
statistical power is through a retrospective (case-control) study, which
compares diseased persons (cases) with non-diseased persons (controls) and
works back in time to determine exposure. The validity of a retrospective
study depends upon careful selection of the control group. The control group
should be similar to the case group in all respects except exposure for the
risk factor under investigation. The distribution for age, sex, race,
socioeconomic status, education, emotional status and other potentially
confounding factors should be the same for each group.
The design of a typical retrospective study is shown in Figure 5-1.
Weaknesses in retrospective studies include confounding factors and biases.
Errors in detecting a cause and effect relationship can stem from failure to
account and adjust for all confounding factors related to the disease and risk
factor under consideration. This task is complicated in retrospective studies
by the lack of accurate historical data. Due to the dependence on recall
data, retrospective studies are especially subject to biases. For example,
diseased patients may be more likely to recall exposure than non-diseased
patients.
5.3.1.2 Prospective Studies
Prospective (cohort) studies examine the development of a disease or condition
in a group (cohort) of persons who have been determined to be presently free
of the disease or condition. The cohort consists of subgroups who have and
have not been exposed to a toxic chemical. The subgroups of the cohort are
then followed for several years to observe differences in the rate at which
disease develops in relation to exposure to the toxic chemical. The design of
a typical prospective study is shown in Figure 5-2.
Compared to retrospective studies, prospective studies are advantageous in
that exposure amounts are observable and likely to be more reliable than
raccllad ercpcsura which my have occurred years before. Often elaborate
5-5
-------
Sample Population
Cases (Have Disease) Controls (No Disease)
Trace Exposure
To Risk Factor
Back In Time
Exposed To
Unexposed To
Exposed To
Unexposed To
Risk Factor Risk Factor Risk Factor Risk Factor
FIGURE 5-1 DESIGN OF TYPICAL RETROSPECTIVE (CASE-CONTROL) STUDY
A group of persons (cases) having an injury, condition or
disease (e.g., cancer) is selected and their past history
with respect to a risk factor (e.g., exposure to a carcino-
gen) is compared to a group of persons (controls) who do
not have the injury condition or disease. Careful statis-
tical analysis of the data is performed to determine
whether there is an association between the specific risk
factor and the condition.
5-6
-------
Sample Population
Disease-Free
Individuals
Individuals
With Disease
Exposed To
Risk Factor
With
Disease
Without
Disease
Unexposed To
Risk Factor
•With
Disease
Without
Disease
FIGURE 5-2 DESIGN OF TYPICAL PROSPECTIVE (COHORT) STUDY
First, a disease-free group is selected from the general
population. This group is then divided into subgroups
according to the presence or absence of a risk factor
(e.g., exposure to a carcinogen). A toxicological endpoint
(e.g., cancer incidence) is measured in the two groups to
determine whether a relationship exists between presence of
the risk factor and development of a disease (e.g., cancer).
-------
physical examinations and environmental monitoring are part of the prospective
study protocol and subjects are followed for many years. This permits
observation of and adjustment for confounding factors. Also, moving forward
in time allows latency (the time from initial exposure to disease diagnosis)
to be more accurately measured. Unfortunately, failure to observe some study
variable that later is found to be important can negate the value of a study;
thus prospective study protocols specify that many characteristics be
observed. Not surprisingly, prospective studies are more expensive and time
consuming than retrospective studies.
The most difficult challenge in a prospective study is selecting the two
subgroups. For example, it would be difficult to select a comparison subgroup
(i.e., not exposed to risk factor) for a subgroup of industrial workers
exposed to toxic chemical. Other individuals exposed to similar stresses and
living similar lifestyles, but having no exposure to the chemical, may be
impossible to identify.
5.3.1.3 Prevalence Studies
Prevalence (cross-sectional) studies examine the relationships between
diseases and exposure as they exist in a defined population at one particular
time. Analysis of data collected in a prevalence study focuses on the
correlation between the incidence of a disease and selected risk factors
(e.g., exposure to a carcinogen). Sometimes it is possible to obtain
dose-response curves that relate the frequency and/or severity of some
biological effect to the intensity of the exposure. The design of a typical
prevalence study is summarized in Figure 5-3.
The key limitation to prevalence studies stems from the fact that they
represent a "snapshot" in time. They may point out a relationship, but do not
describe how such a relationship may have developed. More importantly,
since prevalence studies eliminate the time relationship between exposure to
an environmental hazard and development of a biological effect, some
cause-and-effect relationships may not be detected. One important value cf 1
prevalence study is that it can identify the best source of controls for a
retrospective study and is essentially the first step in conducting a
prospective study.
5.3.2 Uncertainties and Limitations Associated with Epidemiological Data
The conclusions derived from epidemiological studies can be strengthened when
the investigators are aware of and deal with the most common limitations
associated with such studies. These include the following:
• Confounding Factors. Confounding factors are variables which the
epidemiologist cannot control, but which may influence the parameter
being measured. For example, smoking is a confounding variable in
measurements of cancer frequency, and age and weight are confounding
factors in measuring blood pressure. When confounding factors are
recognized, it is sometimes possible to correct for them. However,
this is often not possible, and in some cases the nature of all
confounding variafaj.es are not even known.
5-8
-------
Sample Population
With
Disease
And
Exposed
To
Risk Factor
With
Disease
And
Unexposed
To
Risk Factor
Without
Disease
And
Exposed
To
Risk Factor
Without
Disease
And
Unexposed
To
Risk Factor
FIGURE 5-3 DESIGN OF TYPICAL PREVALENCE (CROSS-SECTIONAL) STUDY
A group of in individuals is selected and each is carefully
investigated to determine (1) the present degree of exposure
to a toxic chemical (i.e., at the time of the study) and
(2) the present health status of the individual (presence
or absence of specific diseases or adverse effects).
5-9
-------
• Bias. Bias is the collection of data that is not truly representative
of the whole population, but is characteristic of some subgroup. Bias
may be introduced, for example, by failing to adjust for age, socio-
economic status, smoking, weight or other variables, or by failing to
follow all members of the exposed and control groups during prospective
studies. This is especially true when effects being measured have a
long latency, as is true for cancer.
• Data Analysis. The primary question in any epidemiological study is
whether an association or correlation between some risk factor and a
biological effect is indicative of a causal relationship or is simply
random. A number of statistical techniques, such as multivariate
analysis, determine whether a significant association should be
accepted as causal, and proper use and interpretation of these
statistical methods is essential. In addition, it is important that
the data make sense biologically. For example, given the long latency
period for cancer, a large increase in cancer incidence in one month
does not make sense biologically.
5.4 Ecotoxicological Studies
The presence of toxic chemicals in the environment may adversely affect the
abundance, species composition and diversity, stability, productivity and
physiological condition of indigenous fish and wildlife populations. Methods
used for assessing the effects of toxic chemicals on these populations include
the collection, identification and counting of organisms, biomass measure-
ments, measurements of metabolic rates, and measurements of the toxicity,
bioaccumulation and biomagnification of toxic chemicals. Testing for these
effects ?.ay be conducted with representative species of plankton, neriphvton,
macrop'nytes, macrcir.vertebrates, fish, amphibians, reptiles, birds on mammals.
These single-species studies are usually very similar in design and objectives
to studies using inbred laboratory animals, and many of the key considerations
are the same (dose-response relationships, effect of duration, differences in
sensitivity between species, etc.). The main difference is that species
employed in ecotoxicological tests are selected to be representative of
indigenous fish and wildlife, while laboratory animals are intended to serve
as models for humans.
This section discusses the types of laboratory experiments (short-term,
intermediate and long-term bioassays) which are performed in order to assess
the toxicity of chemicals on representative test species, provides an example
of a typical single-species experiment and discusses some of the limitations
of ecotoxicological data relative to long-term exposure potential.
5.4.1 Bioassays
Bioassays measure the responses of test organisms to a particular chemical.
Bioassays are useful in determining the suitability of certain environmental
conditions for life, the effects of environmental factors (e.g., pH,
temperature) on the toxicity of a chemical and the comparative sensitivity of
organisms to a chemical. Bioassavs are typically classified according to
duration of exposure (short-term, intermediate or long-term).
5-10
-------
5.4.1.1 Short-Term Bioassays
Short-term (acute) bioassays are generally used to determine the level of a
toxic chemical that produces lethality in a specified percentage of test
organisms over a specified period of time. Acute toxicity tests may be
categorized as range-finding or definitive tests. Range-finding tests a.re
usually 24-hour tests conducted to determine the concentration to be used in
the definitive tests. The test organisms are exposed to a wide range of
concentrations to determine the highest concentration that killed no (or few)
organisms and the lowest concentration that killed all (or most) organisms.
Definitive tests then employ a series of concentrations between those highest
and lowest lethal concentrations.
Experimentally, a 507. effect is the most reproducible adverse effect. The
most frequently used measure of acute toxicity is the median lethal concentra-
tion or dose (1C ^ or LD ). Acute definitive tests provide an indication of
concentrations tnat should be used in conducting partial or complete life-cycle
tests.
5.4.1.2 Intermediate-Term Bioassays
No sharp division exists between short- and intermediate-term bioassays or
between intermediate- and long-term bioassays. Generally, tests lasting eight
days or less are considered to be short-term; tests lasting 8 to 90 days are
considered intermediate-term. Duration of exposure is the only difference in
these tests.
5.4.1.3 Long-Term Bioassays
Long-tarm bioassays ,-easurs .sublethal sffiztz that occur :hr~ugr. chronic
exposure to concentrations lower than those causing acute effects. Vher.
long-term bioassays are conducted, exposure continues over as much of the life
cycle as possible. In life-cycle and partial life-cycle bioassays, the
objective is to determine the maximum allowable toxicant concentration (MATC),
("the concentration of a toxic chemical that cay be present without causin2
significant harm). Parameters measured by this type of bioassay include
growth, reproduction, maturation, spawning, hatching, survival, behavior and
b ioaccumulation.
The EPA has developed guidelines for conducting a variety of ecotoxicological
tests. A list of these guidelines appears in Table 5-2. These guidelines
present methodologies for each test. The EPA has also developed support
documents that provide the scientific rationale used in the development of the
test guidelines.
5.4.2 An Example of an Acute Toxicity Study on Fish
An example of a representative single-species study is the acute toxicity test
for fish. A range-finding test Is initially conducted to determine the range
of concentrations to be used in the definitive test. For the definitive test,
a minimum of 20 fish are exposed to each of five or more concentrations in two
~r 3C-? r®r-licate test chambers. Controls are exposed to the same experi-
mental conditions as the test fish, but are not exposed to the test chemical.
-------
TABLE 5-2 ECOTOXICOLOGICAL TESTS FOR WHICH EPA
HAS DEVELOPED GUIDELINES
Daphnld Acute Toxicity Test
Daphnid Chronic Toxicity Test
Mysid Shrimp Acute Toxicity Test
Mysid Shrimp Chronic Toxicity Test
Oyster Acute Toxicity TEst
Oyster Bioconcentration Test
Penaeid Shrimp Acute Toxicity Test
Algal Acute Toxicity Test
Fish Acute Toxicity Test
Fish Bioconcentration Test
Fish Early Life Stage Toxicity Test
Seed Germination/Root Elongation Toxicity Test
Early Seedling Growth Toxicity Test
Plant Uptake and Translocation Test
Avian Dietary Test
Bobvhite Peproducticn Test
Mallard Reproduction Test
Lemna Acute Toxicity Test
-------
Recommended species for this test include the rainbow trout (Salmo gairdneri) ,
bluegill sunfish (Lepomis macrochirus) and fathead minnow (Pimephales
promeleas). These species were selected based on the following factors:
• A large toxicity data base exists for each species
• All species are readily available and easy to maintain
• All species are widely distributed in the aquatic environment
• Economically important species are represented
In conducting this test, juvenile fish.of the same age and appearance are
used. After a specified period of acclimation, the fish are exposed to the
test chemical under either flow-through or static conditions.
Data that should be collected, recorded or derived include:
• Detailed information about the test fish, including: the scientific
name, average weight (wet weight), standard length, age, source,
history, observed diseases, treatments, mortalities, acclimation
procedures and food used.
• Detailed information about the test system, including: number of
replicates used, number of organisms per replicate, loading rate, flow
rate for flow-through tests, levels of dissolved oxygen, pH and the
temperature and lighting regime.
• Information about the test conditions, including: solvent used, the
test chemical concentration in the stock solution, the highest solvent
concentration in the test solution, and a description of the
solubility of the test chemical in water, other solvents used and the
concentration of the test chemical in each test chamber just before
the start of the test and at all subsequent sampling periods.
• Methods and data records of all chemical analyses of water quality
parameters and test substance concentrations, including method
validations and reagent blanks.
• The number and percentage of test organisms that died, and the number
that showed any abnormal effects in the control and in each test
chamber at each observation period.
• The 24-, 48-, 72- and 96-hour values and confidence limits and
the methods used to calculate the LC,- values and their confidence
limits.
• When available, the no-observed-effect level (the highest
concentration tested at which there were no mortalities or abnormal
behavioral or physiological effects).
• The concentration-response curve at each observation period for which
LC50 values were calculated.
5-13
-------
5.4.3 Limitations of Ecotoxicological Studies
The short-term nature of most studies on representative species from the
ecosystem limits their usefulness in assessing the affects of long-term
exposure. Specific limitations include the following:
• Bioaccumulation. The concentration of a chemical within an organism
may be increased by bioaccumulation. The properties of a chemical
which contribute to high bioaccumulation include a high partition
coefficient and resistance to degradation. A short-term study may not
allow sufficient time for bioaccumulation to play a significant role.
• Effects of latency. For some ecotoxicological endpoints (e.g.,
carcinogenicity), the observed effect may be delayed from the time of
initial exposure. Short-term studies may, therefore, fail to detect
the majority of late-occurring effects.
• Interactions and synergisms. It is likely, especially at a hazardous
waste site, that many different chemicals escape from the site into the
environment. This greatly complicates the task of assessing the hazard
involved, since interactions between chemicals can increase or decrease
the toxicity of specific chemicals.
• Fluctuations in susceptibility. Susceptibility may vary consi'derably
over the lifetime of the organism (e.g., during rapid growth periods)
and, therefore, may not be addressed in a short-term study.
• Sporadic or uneven exposure. Long-term exposure of indigenous species
in the environment may include periods of uneven or sporadic exposure.
This could be caused, for example, by variations in chemical or
biological degradation rates of the chemical, seasonal animal
migration patterns and changes in river water, level and flow. Test
animals, on the other hand, tend to receive constant exposure for a
specified period. The effect of such different treatments, even when
the totai dose is the same, is still unknown.
5.5 Key Guidance and Implementation Documents
DeBell G, ed. 1970. Environmental Handbook. New York: Ballantine Books.
NAS. 1981. National Academy of Sciences, National Research Council. Testing
for effects of chemicals on ecosystems. Washington, DC: National Academy
Press.
USEPA. 1978. Short-term tests for health and ecological effects.
Washington, DC: Office of Toxic Substances, EPA 600/9-78-037.
5.6 Background References
Blair A, Spirtas R. 1981. Use of occupational cohort studies in risk
assessment. In: Richmond CR, Walsh PJ, Copenhauer ED, eds. Health risk
analysis-proceedings of the chird life sciences symposium. Philadelphia,
Pennslyvania: Franklin Institute Press, pp. 97-108.
5-14
-------
Cantor KP. 1981. Human case-control studies in risk assessment. In:
Richmond CR, Walsh PJ, Copenhauer ED, eds. Health risk analysis-proceedings
of the third life sciences symposium. Philadelphia, Pennslyvania: Franklin
Institute Press, pp. 109-120.
Chiazze L, Lundin FE, Watkins D, eds. 1983. Methods and issues in
occupational and environmental epidemiology. Ann Arbor, Michigan: Ann Arbor
Science Publishers.
Lave, LB. 1982. Methods of risk assessment. In Lave LB, eds. Quantitative
risk assessment in regulation. Washington, DC: Brookings Institution,
pp. 23-54.
Lauwerys RR. 1980. Occupational toxicology. In: Doull J, Klaasen DC, Amdur
MO, eds. Casarett and Doull's toxicology. The basic science of poisons, 2nd
ed. New York: Macmillan Publishing Co., Inc., pp. 699-709.
Mclntyre AD, Mills CF, eds. 1976. Ecological toxicology Research. New York:
Plenum Press .
5-15
-------
6.0
EXTRAPOLATION OF TOXICOLOGICAL DATA FROM ANIMALS TO HUMANS
If it were possible to obtain detailed toxicological information on a chemical
directly from studies in humans, it would not be too difficult to calculate
appropriate exposure limits or risk estimates that define the levels at or
below which no significant adverse effects would occur in an exposed human
population. However, good quantitative toxicity data in humans for a specific
toxic chemical are often limited or absent, and derivation of exposure limits
that are applicable to humans nearly always requires extrapolation of results
obtained in animals. In view of the many potential differences among species
with respect to sensitivity to toxic chemicals (see section 4.3), it is not
surprising that extrapolations of this sort are rather complicated and
sometimes involve considerable uncertainty. This chapter describes the major
problems and uncertainties involved in deriving human exposure standards from
studies in animals, and the means that are currently used to circumvent these
problems and uncertainties.
6.1 Selection of Appropriate Studies
The first step in derivation of an exposure limit or risk estimate is review of
existing toxicological data and selection of the most applicable study or
studies. The factors which are important in making chis selection are:
• Route of Exposure. Because absorption and toxicity of a chemical
often depend on the route of exposure (see section 4.1), it is
important that an inhalation standard be based on inhalation data, that
a water standard be based on ingestion data and so on. Exceptions to
this rule should be viewed with caution.
• Duration of Exposure. 3ecause the coxic effects of some chemicals cend
to accumulate with time (see section A.2), it is important to select a
study involving long-term exposure for derivation of a standard
intended to provide protection from chronic exposure.
• Species cz Animal. Since there are sometimes considerable differences
in sensitivity to a chemical among species (see section A.3), it is
nearly always desirable to select studies involving humans when
sufficient data are available. When human data are sparse, care
should be taken not to select a study in an animal that is known to be
significantly different from humans in sensitivity or response to the
chemical.
• Endpoint. Many chemicals produce multiple effects (see section 4.4).
It is important to identify the most appropriate endpoint of a
chemical's toxicity and select a study which has determined a NOAEL or
LOAEL value on the basis of that endpoint.
• Statistical Significance. Fluctuation in measurements and variations
among animals is an inherent aspect of toxicological investigations.
It is not always simple to determine whether a chemical has caused an
effect or not. Statistical analysis of the data is the most objective
-neans of dnsverir.g the question: "How certain is it that the chemical
-------
did (or did not) cause an effect?" Table 6-1 summarizes some common
statistical terms, and Table 6-2 illustrates the statistical analysis
of some sample data. In general, an effect is not considered to be
statistically significant unless there is 95% confidence that a change
did occur.
• Study Quality. While statistical analysis is very useful in judging
the significance (or lack thereof) of experimental results, more
subjective analysis of the quality of the study must also be
performed. This analysis should consider, for example, possible flaws
in the measuring techniques, failure to consider all important
variables, biased experimental design and so on. This sort of
critiquing of a study requires intimate familiarity with all aspects
of the study and can only be done by an experienced toxicologist.
6.2 Conversion of Dose Levels
When the study cr studies selected as most appropriate involve animals, the
doses administered to the animals must be converted to an equivalent dose in
humans. When an animal is exposed to a chemical in a laboratory study
involving oral exposure, it is customary to describe the amount of chemical
the animal ingests in units of mg chemical per kg body weight (mg/kg). If
humans and animals were equally sensitive to the chemical on this basis
(weight equivalence), then a dose without effect in the animal would also be
without effect in humans. However, a considerable body of laboratory data in
animals and clinical data in humans indicates that doses expressed' in these
units are not toxicologically equivalent in animals of different sizes, and
that a dose producing no effect in mice might indeed produce an effect in
humans.
Studies of effect levels of chemicals indicate that a better correlation among
different species exists between toxicity and dojje when doses are expressed in
units of mg chemical per unit surface area (mg/m ). The theoretical basis for
this correlation is not obvious, but a large number of physiological
parameters, including surface area, are approximately proportional to the
two-thirds power of body weight. These parameters include metabolic rates,
oxygen consumption, blood volume, kidney function, thyroid function, brain
weight, liver weight, cardiac output, blood pressure and extracellular water
volume. Since many of these parameters are related to the absorption,
distribution, excretion, metabolism and mechanisms of toxicity of chemicals,
surface area is used as a convenient physical parameter which is proportional
to the physiological rates and functions that are directly affected.
The use of surface area equivalence in dose conversions yields calculated
doses that are lower than if weight equivalence is used. Comparative data are
presented in Table 6-3, which shows the doses in several species calculated to
be equivalent to a dose of 2 mg in a 20 g mouse (100 mg/kg). Using weight
equivalence, the same dose in a 70-kg human is 7,000 mg (100 mg/kg), but if
surface area equivalence is used, the dose is only 776 mg (11 mg/kg).
Conversion of doses (expressed in units of mg/kg) in one species to
surface-area equivalent doses (also expressed as mg/kg) in another species may
be accomplished by simply multiplying by a conversion factor, as follows:
6-2
-------
TABLE 6-1 STATISTICAL TERMS AND THEIR USE
Sample Size (N) - The sample size must be sufficient to establish statistical
significance of the results. Statistical significance increases as function
of the number of animals used in the experiment. The acceptable number needed
for an experiment depends on the type of effect studied (acute, chronic,
etc.), the length of the experiment, and the degree of control that can be
placed over the experiment.
Mean - In samples, as well as in populations, there are generally a preponder-
ance of values somewhere around the middle of the range of observed values.
The mean is the most widely used measure of this central tendency. The mean
is calculated as the sum of all the sample values divided by the sample size.
X - 2i
N
Measures of Dispersion and Variability - Whereas the mean provides a measure
of central tendency, other measures provide an indication of how tightly the
data are distributed around a mean or, conversely, how variable the measure-
ments are. Many such distributions are found to conform to a normal "bell
shaped" distribution. Generally, statistical significance increases as
variability decreases. The most common measure of variability is variance,
the sum of the squares of the deviations from the mean divided by N-l.
„ 4 2 (X, -X)2
Variance * —^i^
N— 1
The positive square root of the variance is called the standard deviation
(SD). If random samples of size N are drawn from a normal population, the
standard error of the means (SEM) is calculated by dividing the SD by the
square root of the sample size.
SEM SD
,0.5
Statistical Significance - Many investigations address the hypothesis that a
given chemical produces a significant effect in a treated group as compared to
an untreated control group (i.e., the sample data are derived from two
statistically different populations). The P-value is the probability that the
hypothesis is false based on the data collected. The smaller the P-value the
greater the confidence in the truth of the hypothesis. Generally, the greater
the observed differences between the treated and untreated groups the smaller
the P-value. The significance level a is the criteria by which the hypothesis
is rejected or accepted. Many studies use a 5Z significance level, i.e., if
P>.05, the hypothesis is rejected. However, if a test result fails to meet
the criteria of significance for proving an effect has occurred, that does not
prove the effect did not occur. The strength of a negative conclusion is
evaluated by a power test. A power test provides an estimate of the
probability that an experiment would detect an effect if it were present. For
small effects, tests with high power (large numbers of animals, precise
measuring techniques) are required.
6-3
-------
TABLE 6-2 STATISTICAL ANALYSIS OF HYPOTHETICAL DATA
Blood Pressure, mm Hg
Control
(Not Exposed)
Exposed
to Chemical
100
110
93
107
85
108
94
86
105
78
107
92
115
116
112
96
101
115
111
99
X
SD
96.9
11.1
106.A
8.8
P<0.05
Assume 20 animals are divided into two groups of ten. The first group
(control) is not exposed to the chemical, while the second group is exposed to
a dose of some chemical suspected of causing Increased blood pressure. After
exposure, the mean blood pressure of each animal is measured, with the results
shown above. Exposure to the chemical did produce an increase in mean blood
pressure (X), but is this really caused by the chemical or is this change
random? To answer this, first the standard deviations are derived, and then a
P value is calculated. Since P is less than 0.05, there is a 95% probability
that the change observed did not occur by random, but is a real effect caused
by the chemical.
6-4
-------
TABLE 6-3 COMPARISON OF DOSE CONVERSIONS USING
SURFACE AREA AND WEIGHT EQUIVALENCE
Species
Mouse
Rat
Guinea Pig
Rabbit
Cat
Monkey
Dog
Human
Surface.
Weight, g Area, cm
Calculated Dose, mg
(a)
20
200
400
1,500
2,000
4,000
12,000
70,000
46
325
564
1,272
1,381
2,975
5,766
18,000
Weight
Equivalence
2
20
40
150
200
400
1,200
7,000
Surface Area
Equivalence
2
14
24
55
59
128
248
776
(a) Based on a dose of 2 mg in a 20 g mouse (100 mg/kg).
Adapted from Klaassen and Doull (1980).
6-5
-------
DH ¦
-------
TABLE 6-4 DOSE CONVERSION FACTORS
Species
Weight, kg
w
Mouse
0.02
0.066
Rat
0.2
0.142
Guinea Pig
0.4
0.179
Rabbit
1.5
0.278
Cat
2.0
0.306
Monkey
4.0
0.385
Dog
12.0
0.555
Human
70.0
1.000
6-7
-------
Similar corrections may be employed to adjust for differences in absorption
across skin or lung and differences in rates of metabolism. Unfortunately,
reliable quantitative toxicokinetic data in animals and/or humans are
frequently lacking and objective corrections for differences in absorption or
metabolism are not possible in all cases.
6.4 High-Dose to Low-Dose Extrapolation
The design of experimental dose-response studies is limited by two important
practical considerations. First, the researcher is restricted to the study of
animal populations of manageable size, usually 100 to 1,000 animals. Second,
the researcher is limited to the use of exposures that will produce a
measurable response in a test population of the size studied. The task of
the toxicologist is then to extrapolate results obtained at high doses where
effects can be detected to expected results at low levels (more characteristic
of human exposure from the environment) where effects cannot be measured
directly.
Despite wide gaps in our knowledge of the metabolism and ultimate fate of
chemicals in man, properly conducted animal experiments have yielded results
that are predictive of the effects in humans. Using appropriate statistical
treatment of the results of experiments on animals, a plausible estimate of
risk can be calculated that approximates the true risk in populations exposed
to known concentrations of a toxip chemical. Such estimates are designed to
be highly conservative (i.e., the true risk is almost.certainly lower than the
estimate).
Animal experiments, using relatively small numbers of subjects, must be per-
formed at doses high enough to provide measurable toxic effects in a
relatively short time period. Since adverse effects at very low expcsuras are
often not apparent, estimates of risk from low exposures are based on the
downward extrapolation of the dose-response curve from relatively high dose
levels. Therefore, to estimate the probability of effects (response) at dose
levels outside the experimental range, it is necessary to make an assumption
concerning the shape of the dose-response curve at the low dose range.
Many mathematical models have been developed which make this downward
extrapolation. The dose-response curve, depending on the mathematical model
used, may be convex, linear, or concave at low doses (see Figure 6-1). A
curve that is concave at low doses will lie above one that is linear and a
curve that is convex will lie below one that is linear. Consequently, if the
curve is approximated at low doses by a straight line, the approximation will
overestimate risk if the true response curve is convex, and underestimate risk
if the true response curve is concave.
Six models for extrapolating to low dose are routinely used: the probit model,
logit model, Weilbull model, linearized multistage model, one-hit model, and
the gamma multi-hit model. (These models are discussed in detail in Section
6.4.1) The models are used to estimate a virtually safe dose (VSD), which
6-8
-------
(a) = concave
(b) = linear
(c) = convex
Dose, Arbitrary Units
FIGURE 6-1 ILLUSTRATION OF DIFFERENT POSSIBLE SHAPES OF
THE DOSE-RESPONSE CURVE IN THE LOW DOSE RANGE
5-9
-------
is the dose level corresponding to a desired low level of response. Although
each of the six models are frequently used in the assessment of risk, there is
substantial scientific disagreement about their relative merits. The Proposed
Guidelines for Carcinogen Risk Assessment (USEPA 1984) recommends that the
linearized multistage model be utilized for high-to-low dose extrapolation
unless there is mechanistic or other biological evidence that indicates the
greater suitability of an alternative extrapolation model, or there is
statistical or biological evidence that excludes the use of the linearized
multistage model.
Figure 6-2 is an example of how the six models are applied to extrapolating
low dose-response from incidence of liver tumors in mice exposed to high
levels of DDT. While each of the models was found to fit the experimental
data nearly equally well, they lead to very large differences when extrap-
olated to low doses, differing in this case by three to five orders of
magnitude (1,000- to 100,000-fold different). The one hit and the linearized
multi-stage models are most conservative (i.e., give the highest risk
estimate) .
6.4.1 Dose-Response Models
The following section describes the six models currently being used for
high-to-low dose extrapolation, their assumptions, similarities and limita-
tions .
There are two basic classes of dose-response models:
• Tolerance Distribution Models
• Stochastic or Mechanistic Models
These categories are not always distinct and some models may belong to both
categories.
Tolerance distribution models are based or. the concept that each individual in
the population has its own tolerance to the test chemical. If a dose does not
exceed the tolerance of an individual, then there will be no response by that
individual. If the dose exceeds the tolerance, then a response will be
observed. It is assumed that the distribution of tolerances within the
population follows a normal (bell-shaped) distribution. The probit and logit
models fall into this class.
Stochastic or mechanistic models including the one-hit, gatmna multi-hit,
linearized multistage and Weibull models, are derived from assumptions
regarding the mechanism of action of the toxic chemical upon its target site.
The "hit" theory for interaction between radiation particles and susceptible
biological targets has generated this general class of models. This "hit"
theory is also applicable to the action of toxic chemicals upon their target
sites. The assumptions forming the basis of this theory include:
6-i0
-------
-J I I 1 L.
10-6 10~* 10"2 1 102
Dose, ppm DDT in Daily Diet
FIGURE 6-2 DOSE-RESPONSE CURVE EXTRAPOLATIONS BY SIX MATHEMATICAL MODELS
This figure Illustrates the large differences between six
mathematical models for extrapolation to the low-dose
. range from a set of cancer frequency data at high doses of
DDT.
Adapted from Brown (1974).
-------
1. The organism has some number of critical targets (susceptible cells
or molecules);
2. The organism responds if one or more of these critical targets are
injured;
3. A critical target is injured if it is "hit" by one or more toxic
chemicals; and
4. The probability of a "hit" in the low-dose region is proportional to
the dose level of the toxic chemical.
It should be noted that stochastic/mechanistic models of carcinogenesis
generally become linear at low dose levels. This is because they assume that
all population members have identical susceptibility to the carcinogen. If
this was not the case, the population dose-response relationship at very low
dose rates would be concave even though no absolute threshold for the
carcinogen exists. Many researchers believe that the shape of the true
dose—response curve at low exposure levels is convex, i.e., may have some
degree of upward curvature (see Figure 6-1). Therefore, linearity provides
conservative or overestimated extrapolated risk estimates at low doses.
6.4.2 Threshold Versus Nonthreshold Models
All the models described previously are nonthreshold models; i.e., they assume
some positive.probability of observing a response no matter how low the dose.
There has been much discussion over the existence of thresholds. Some
scientists argue that responses to carcinogens are less likely to show a
threshold effect than responses to other toxic substances because cance"r may
be produced by an event in a single cell. Even if a threshold level does
exist, this level would probably vary among members of the population at risk
and the resulting dose-response curves would be indistinguishable from those
described by nonthreshold models. For this reason, it appears that
nonthreshold models are appropriate to use in extrapolating the risk from
exposure to toxic chemicals.
6.4.3 Toxicokinetic Considerations in High-To-Low Dose Extrapolation
Toxicokinetics is the study of the time course of an administered chemical and
its metabolites in the body. The six models currently available for high-to-
low dose extrapolation described above assume that the biological fate of the
administered chemical is directly proportional to the administered dose.
However, this circumstance is not always the case. It is now known that many
chemicals are only carcinogenic after they have been metabolized and that the
metabolic processes involved may not be proportional to the administered dose.
The high doses typically used in carcinogenesis bloassays often saturate
normal metabolic (detoxification) processes, resulting in nonproportional
relationships between nominal and effective doses. If detoxification pathways
are saturated, then the effective dose will increase more rapidly than a
proportional relationship would predict. The overall consequence of a
nonproportional relationship is that the carcinogenic response may increase in
6-12
-------
a nonlinear manner with increasing dose. Therefore the mathematical models
used for extrapolation may overestimate the risks associated with low dose
levels of toxic chemicals.
6.5 Sources of Uncertainty
In the assessment of risks associated with toxic chemical exposure, there are
a number of sources of significant uncertainty. Many of the sources of
uncertainty have been described in preceding sections, and are only
summarized here.
6.5.1 Sources of Uncertainty in Extrapolation from Animal Studies
6.5.1.1 High-to-Low Dose Extrapolations
Toxicological studies are often conducted at doses of chemicals much higher
than those to which human populations are exposed in the environment. For
responses thought to have no threshold (e.g., cancer), prediction of the
toxicological response at low doses of a chemical must be done mathematically.
Frequently, there is considerable uncertainty in this process, most often
because the data are too limited to define the dose-response curve precisely.
In addition, the best mathematical equation to describe the dose-response
curve may not be know.
6.5.1.2 Scalirg Factors
Rodents are smaller than humans, they live approximately l/35th as long, and
their rates of metabolism and cell division are much faster. These
differences influence chemical toxicity. There is some uncertainty whether
the proper way to scale or convert doses from animals to humans is on the
basis of relative body weights, relative surface areas, relative life spans,
or on some other basis.
6.5.1.3 Differences in Species' Sensitivity
There is a large data base of evidence that different species, even different
strains within species, have markedly different sensitivities to toxic
chemicals. Animal species can exhibit differences in sensitivity to a
chemical of as much as 100-fold. These differences are not accounted for in
scaling corrections and assumptions.
6.5.1.4 Individual Variation in Human Sensitivity
Even beyond differences among species, there is greater individual variation
in sensitivity among humans Chan among test animals. This variation is a
result of the fact that humans are genetically heterogeneous, while test
animals are bred for genetic homogeneity. The range of human sensitivities
creates an important source of uncertainty of unquantified magnitude in
assessing human risks.
6-13
-------
6.5.1.5 Interactions and Synergisms
Animals are usually exposed to only one test chemical in carefully controlled
settings. Humans, however, are exposed to a wide variety of other chemicals
and environmental conditions. The risk from multiple exposures may be greater
than the sum of the risks from exposure to individual toxic chemicals. Yet,
most synergisms have not been identified, let alone quantified.
6.5.2 -Sources of Uncertainty in Epidemiological Studies
6.5.2.1 Confounding Factors
Epidemiological studies are performed in an uncontrolled setting, introducing
many unknown factors that can obscure true relationships between cause and
disease. While it is usually possible to identify some confounding factors
and take them into account, it is very difficult to eliminate this problem
entirely. Eyen the best, epidemiological studies cannot usually detect effects
occurring at less than 50% above the normal rate.
6.5.2.2 Effects of Latency
In the case of carcinogens, latency periods (the time from the first exposure
to diagnosis of disease) of 20 years or more are common. Studies which do not
follow exposed subjects for their full lifetime can fail to detect a
carcinogenic effect entirely or can miss the majority of later occurring
cases.
6.5.2.3 Failure to Follow All Members of the Exposed Group
A common failure of occupational studies is the failure to follow employees
who change jobs. Since ill employees are the most likely to leave employment,
ignoring this factor can be significant. Their state of health, or even
whether they are still alive, is often not determined. This tends to
understate the true risk. Follow-up problems are even greater for many
nonoccupational studies.
6.5.2.4 Ignorance of True Exposure Levels and Duration
Dose information is often highly speculative, especially for exposure
incidences taking place two or more decades in the past. This problem is
especially severe in epidemiological studies concerning toxic chemicals
released in the environment.
6.5.2.5 Errors in Describing the Study and Control or Failures to Adjust for
Age, Socioeconomic Status, and Other Variables
Errors in -describing the group exposed and the comparison control group can
introduce serious errors into quantitative estimates. Similar errors can
occur from failure to adjust for age, income, race and other variables.
6-14
-------
6.6 Approaches to Dealing with Uncertainty
When there are insufficient data to permit clear decisions in the toxicity
assessment process, two strategies are available. The first is to assume
worst-case values; this strategy almost certainly ensures that calculated
values will not be too high, but may yield values that are lower than really
necessary. The EPA's calculation of cancer risk estimates is a prominent
example of the application of this strategy.
A second approach to dealing with uncertainty is to employ uncertainty
factors. These factors are intended to provide a sufficient "margin of error"
to account for uncertainties arising from all of the possible sources
described in sections 6.5.1 and 6.5.2. It is a standard practice to employ
uncertainty factors in t-he derivation of noncarcinogenic guidelines and
standards.
Table 6-5 .provides EPA guidelines for selection of uncertainty factors for
evaluating acceptable daily intake (ADI) of noncarcinogenic chemicals based on
a NOAEL. A minimum uncertainty factor of 10 is generally employed to account
for variations between individuals and to provide a basic margin of safety.
Additional uncertainty is added (usually by factors of 10) to account for use
of data from animals, or for use of limited or poor quality data, or data from
short-term studies.
The following is a hypothetical example of the use of this approach in
estimating an ADI for humans.
Assume a hypothetical compound is widely used in the United States, and that
it is toxic to humans and animals, causing injury primarily to lung.tissue. A
two-year feeding study ir. rats indicates chat doses of up to 170 ppm in the
diet (8.5 mg/kg/day) do not cause significant injury to lung or other tissues.
This is identified as the NOAEL. Since no useful long-term or acute human
data exist, an uncertainty factor of 100 is appropriate, and the ADI is
calculated as follows:
ADI - 8.5^mg/kg/day - 0.085 mg/kg/day
Assuming 70 kg as the average weight of an adult human, and assuming
consumption of 2 L/day of water, if exposure were entirely through water, the
maximum acceptable water concentration would be:
(0.085 mR/kg/day)(70 kg) ,
(2 L/day) mg/L
If exposure were also occurring by other routes, the permissible level in
water would be reduced accordingly. Assuming that 20% of the ADI could come
from water, the maximum permissable level from water would be:
(0.20)(3.0 mg/L) - 0.6 mg/L
6-15
-------
TABLE 6-5 GUIDELINES FOR SELECTION OF UNCERTAINTY FACTORS
Uncertainty Factor
1. Use a 10-fold factor when extrapolating from valid experimental results
of studies on prolonged ingestion by humans. This 10-fold factor
protects the sensitive members of the human population estimated from
data gathered on average healthy individuals.
2. Use a 100-fold factor when extrapolating from valid results of long-term
feeding studies on experimental animals when results of studies of human
ingestion are not available or scanty (e.g., acute exposure only). This
represents an additional 10-fold uncertainty factor in extrapolating data
from the average animal to the average human.
3. Use a 1,000-fold factor when extrapolating from short-term study results
from experimental animals when no useful long-term or acute human data
are available. This represents an additional 10-fold uncertainty factor
in extrapolating from short-term to chronic exposures.
4. Use an additional uncertainty factor of between 1 and 10, depending on
the sensitivity of the adverse effect, when deriving an ADI from a LOAEL.
This uncertainty factor drops the LOAEL into the range of a NOAEL.
Adapted from Dourson and Stara (1983).
6-16
-------
Application of appropriate uncertainty factors in calculations of this sort
ensures that the resulting values will be sufficiently conservative that no
adverse effects will occur in any exposed human population.
It is important to realize that methods for calculating ADI's are constantly
being reviewed by EPA for consistency with the latest toxicological knowledge.
The method illustrated in this example, therefore, may be revised.
6. 7 Key Guidance and Implementation Documents
Food Safety Council. 1980. Proposed system for food safety assessment.
Washington, DC: Food Safety Council.
USEPA. 1984. U.S. Environmental Protection Agency. Proposed guidelines for
carcinogenic risk assessment. Fed. Regist. 49-46294.
USEPA. 1984. U.S. Environmental Protection Agency. Proposed guidelines for
mutagenicity risk assessment. Fed. Regist. 49-46314.
USEPA. 1983. U.S. Environmental Protection Agency, Environmental Criteria
and Assessment Office. Guidance and methods for the use of acceptable daily
intakes (ADIs) in health risk assessment. Cincinnati, OH: U.S. Environmental
Protection Agency. ECAO-CIN-401.
USEPA. 1981. U.S. Environmental Protection Agency, Office of Pesticides and
Toxic Substances. Health effects test guidelines. Washington, DC: U.S.
Environmental Protection Agency 560/6-82-001.
USEPA. 1979. U.S. Environmental Protection Agency. Proposed health effects
test standards for Toxic Substances Control Act test rules and proposed zood
laboratory practice standards for health effects. Fed. Regist. July 26, 1979,
44:44054-44093.
USEPA. 1977. U.S. Environmental Protection Agency. Interim procedures and
guidelines for health risk economic impact assessments of suspected
carcinogens. Fed. Regist. 41:21402-21405.
6.8 Background References
Albert RE, Train RE, Anderson E. 1977. Rationale developed by the Environ-
mental Protection Agency for the assessment of carcinogenic risks. J. Natl.
Cancer Inst. 58:1537-1541.
Anderson MW, Hoel DG, Kaplan NL. 1980. A general scheme for the incorpora-
tion of pharmacokinetics in low-dose risk estimation for chemical carcinogene-
sis: example—vinyl chloride. Tox. Appl. Pharm. 55:154-161.
Brown CC. 1984. High-to low-dose extrapolation in animals. In: Rodricks
JV, Tardiff RG, eds. Assessment and management of chemical risks. Washington
DC: American Chemical Society, pp. 57-79.
6-17
-------
Calabrese EJ. 1984. Principles of animal extrapolation. New York, NY: John
Wiley and Sons.
Chevron Chemical Company. 1975.. Paraquat poisoning; a physician's guide for
emergency treatment and medical management. San Francisco: Chevron Environ-
mental Health Center, pp. 66.
Crouch E, Wilson R. 1979. Interspecies comparison of carcinogenic potency.
J. Toxicol. Enviro. Health 5:1095-1118.
Crump KS, Hoel DE, Langley CH, Peto R. 1976. Fundamental carcinogenic pro-
cesses and their implication for low dose risk assessment. Cancer Res.
36:2973-2979.
Crump KS, Howe R. 1980. Approaches to carcinogenic, mutagenic and terato-
genic risk assessment. Summary report. Washington, DC: U.S. Environmental
Protection Agency. Contract no. 68-01-5975.
Dourson ML, Stara JF. 1983. Regulatory history and experimental support of
uncertainty (safety) factors. Reg. Toxicol. Pharmacol. 3:224-238.
Freireich EJ, Gehan EP, Rail DP, Schmidt LH, Skipper HE. 1966. Quantitative
comparison of toxicity of anti-cancer agents in mouse, rat, hamster, dog,
monkey, and man. Cancer Chemotherapy Rep. 50:219-244.
Gart J, Chu K, Tarone R. 1979. Statistical issues in interpretation of
chronic bioassay tes-ts for carcinogencity. J. Nat. Can. Inst. 62:957-978.
Gehring PJ, Watanabe PG, Blau GE. 1979. Risk assessment of environmental
carcinogens utilizing pharmacokinetic data. Ann. NY Acad. Sci. 329:137-152.
Hoel DG, Kaplan NL, Anderson MW. 1983. Implications of nonlinear kinetics
risk estimation in carcinogenesis. Science 219:1032-1037.
Klaassen CD, Doull J. 1980. Evaluation of safety: toxicologic evaluation.
In: Doull J, Klaassen CD, Amdur MO, eds. Casarett and Doull's Toxicology.
New York: Macmillan Publishing Co., p. 21.
Klippel CH. 1979. Surface area versus skin area. New Eng. J. Med. 301:730.
Lundin FE, Wagoner JK, Archer VE. 1971. Radon daughter exposure and respira-
tory cancer: quantitative and temporal aspects. NI0SH/NIEHS Joint Monograph
1. U.S. Department of Health, Education, and Welfare.
Mantel N, Byran WR. 1961. "Safety" testing of carcinogenic agents. J. Nat.
Cancer Ins. 27:455.
Menzel DB, Smolko ED. 1984. Interspecies Extrapolation. In: Rodricks JV,
Tardiff RG, eds. Assessment and management of chemical risks. Washington,
DC: American Chemical Society, pp 23-25.
6-18
-------
Rothraan KJ, Keller AZ. 1972. The effect of joint exposure at alcohol and
tobacco on risk of cancer of the mouth and pharyns. J. Chron. Dis.
25:711-716.
Selikoff IJ, Hammond EC, Churg J. 1968. Asbestos exposure, smoking and neo-
plasia. J. Am. Med. Assoc. 204:106-112.
Weisburger JH, Williams GM 1981. Carcinogen testing: current problems and
new approaches. Science 214:401-407.
6-19
-------
7.0 EXPOSURE ASSESSMENTS
Exposure assessment is the process of collecting all the necessary information
to answer the following questions for a specific hazardous waste site:
• What chemicals are present?
• In what media (air, soil, water) do these chemicals occur?
• What amount or concentration of each chemical is present?
• What living organisms (humans, wildlife, plants) are exposed?
• By what routes are these organisms exposed?
• What pattern and degree of exposure is expected in the future?
When this information is qombined with toxicological data derived from the
toxicity assessment process (chapters 3.0 through 6.0), it is possible to
derive reasonable estimates of the risks posed to the exposed organisms.
The purpose of this chapter is to explain various aspects of exposure
assessments as related to actions at hazardous waste sites. T~»is chapter does
not describe the technical means of collecting and analyzing Aposure data or
the methodology of conducting exposure assessments. Rather, it explains the
types of data which are required to permit the toxicologist to perform a risk
assessment. To accomplish this purpose, this chapter describes the steps in
assessing present, past, and future exposures.
7.1 Collection of Occurrence Data
Adverse effects in living organisms are not produced unless a toxic chemical
contacts the organism at a sufficient concentration and for a sufficient
duration to initiate a toxic effect. Therefore, whether or not a toxic effect
occurs depends not only on the properties of the chemical and the character-
istics of the organism in question, but also on a number of exposure-related
factors. The major exposure-related factors that Influence toxicity are
route, duration and frequency of exposure. Therefore, one type of
information required in conducting an exposure assessment is knowledge of the
identity and concentrations of toxic chemicals in each possible exposure
medium (i.e., air, groundwater, surface water, soil and biota). Two important
aspects of collecting this type of information include an evaluation of the
history of the hazardous waste site and the collection and analysis of
sampling data. An Important component of the collection/analysis of data
involves the consideration of certain legal requirements (e.g., chain-of-
custody requirements, quality assurance/quality control (QA/QC) procedures) to
ensure the integrity of the data.
7.1.1 Site History
In the assessment of exposure resulting from the presence of toxic chemicals
at a hazardous waste site, an important initial step is an evaluation of the
history ol the site. The types of information typically required in order to
adequately characterize the history of a hazardous waste site are presented in
Table 7-1. Especially important to the performance of an exposure assessment
is information on the identity of chemicals present at the site, their concen-
trations, and the manner in which they were originally disposed or stored.
7-1
-------
TABLE 7-1 CHECKLIST OF IMPORTANT SITE HISTORY INFORMATION
1. ~ Facility Ownership
2. ~ Facility Type
3. ~ Facility Location, Size Configuration, Physical Description
A. ~ Time-Frame of Waste-Related Activities
5. ~ Types of Activities/Operations at the Site
6. ~ Waste Disposal/Storage Methods
7. ~ Identity/Quantity of Disposed Waste
8. ~ Chemical Composition of Waste
9. ~ Site Incidents (e.g., fires, explosions)
10. ~ Records of Previous Site Investigations
U. ~ Records of Previous Sampling Activities
12. D Records of Previous Response Actions
N/A • Not applicable.
D/A - Documented and Attached.
7-2
-------
Once this information is obtained, predictions can be made regarding which
chemicals warrant concern, what media are likely to be contaminated and which
exposure routes will be of concern. Based on this information, a sampling
plan can be devis.ed. Therefore, the results of the site history analysis
serve to focus and direct subsequent exposure assessment activities.
7.1.2 Sampling and Analysis of Environmental Data
A second important aspect of th-e collection of occurrence data at hazardous
waste sites involves the sampling and analysis of environmental media for
chemicals of concern. To a large extent, this activity receives its direction
from the results of the site history analysis. Since site history analyses
typically provide information on those chemicals expected to be of concern and
in what environmental media those chemicals may occur, sampling efforts can be
directed accordingly. However, while site history analysis is an important
initial step in performing an exposure assessment, the available site history
information is often incomplete and/or unreliable. In such cases, it may be
necessary to collect sampling data on a variety of chemicals that may be
present at the site. Table 7-2 presents a list of chemicals for which
sampling may be required when site history data are'insufficient.
The results of sampling efforts should yield data on the amount of toxic
chemicals released from each on-site source into each environmental medium.
Estimates of chemical release to each environmental medium may be qualitative
or quantitative. Qualitative information is useful for discriminating
between major and minor sources of releases and for estimating the nature and'
relative magnitude of releases. However, quantitative release data are re-
quired since it is these data that will be used in subsequent steps of the
exposure assessment process to calculate doses of toxic chemicals incurred by
exposed populations (receptors). Quantitative data may be obtained through
either modeling or monitoring (sampling) activities or a combination of model-
ing and monitoring. The media which must be analyzed for toxic chemicals
include the atmosphere, surface water, groundwater, soil, and the tissues of
organisms (biota) which are consumed by others (especially species consumed by
humans).
7.1.2.1 Atmospheric Contamination
Emissions of contaminated fugitive dusts (airborne wastes and contaminated
soil particles) and volatilization of toxic chemicals are the most likely
sources of atmospheric contamination at hazardous waste sites. Fugitive dust
emissions can result from wind erosion of waste and contaminated soil,
vehicular traffic over contaminated roads, heavy equipment activity at the
site or incineration of wastes during remediation. Volatilization of
contaminants at hazardous waste sites can occur from lagoons, inadequately
covered landfills or from spills or leaks.
Mathematical equations for making quantitative estimates of atmospheric
contamination patterns are termed atmospheric dispersion models. These models
are employed to determine airborne chemical concentrations as a function of
space and time and they typically require information on characteristics of
che source, physical -r.d ;hisi:al prDpertias, ar.d data t. Iccal rseteorolocy.
-------
TABLE 7-2 CHEMICALS FREQUENTLY OCCURRING AT HAZARDOUS WASTE SITES
FOR WHICH SAMPLING DATA MAY BE REQUIRED
Chemical Abstract
Number Volatiles Service (CAS) Number
1.
Chlorcme thane
74-87-3
2.
Bromomethane
74-83-9
3.
Vinyl chloride
75-01-4
4.
Chloroethane
75-00-3
5.
Methylene chloride
75-09-2
6.
Acetone
67-64-1
7.
Carbon disulfide
75-15-0
8.
1,1-Dichloroethene
75-35-4
9:
1,1-Dichloroethane
75-34-3
10.
trans-1,2-Dichloroethene
156-60-5
11.
Chloroform
67-66-3
12.
1,2-Dichloroethane
107-06-2
13.
2-Butanone
78-93-3
14.
1,1,1-Trichloroethane
71-55-6
15.
Carbon tetrachloride
56-23-5
16.
Vinyl acetate
108-05-4
17.
Bromodichloromethane
75-27-4
18.
1,1,2,2-Tetrachloroethane
79-34-5
19.
1,2-Dichloropropane
78-87-5
20.
trans-1,3-Dichloropropene
10061-02-6
21.
Trichloroethene
79-01-6
22.
Dibromochlorome thane
124-48-1
23.
1,1,2-Trichloroethane
79-00-5
24.
Benzene
71-43-2
25.
cis-1,3-Dichloropropene
10061-01-5
26.
2-Chloroethyl vinyl ether
110-75-8
27.
Bromofonn
75-25-2
28.
2-Hexanone
591-78-6
29.
4-Me thyl-2-pent anone
108-10-1
30.
Tetrachloroethene
127-18-4
31.
Toluene
108-88-3
32.
Chlorobenzene
108-90-7
33.
Ethylbenzene
100-41-4
34.
Styrene
100-42-5
15.
Xylene
1330-20-7
cjntinued-
7-4
-------
Table 7-2 - continued
Number
Semi-Volatiles
36.
N-Nitrosodimethylamine
37.
Phenol
38.
Aniline
39.
bis(2-Chloroethyl) ether
40.
2-Chlorophenol
41.
1,3-Dichlorobenzene
42.
1,4-Dichlorobenzene
43.
Benzyl alcohol
44.
1,2-Dichlorobenzene
45.
2-Methylphenol
46. -
bis(2-Chloroisopropyl) ether
47.
4-Methylphenol
48.
N-Nitroso-N-dipropylamine
49.
Hexachloroethane
50.
Nitrobenzene
51.
Isophorone
52.
2-Nitrophenol
53.
2,4-Dimethylphenol
54.
Benzoic acid
55.
bis(2-Chloroethyoxy) methane
56.
2,4-Dichlorophenol
57.
1,2,4-Trichlorobenzene
58.
Naphthalene
59.
4-Chloroaniline
60.
Hexachlorobutadiane
61.
4-Chloro-3-methylphenol
(p-Chloro-m-cresol)
62.
2-Methylnaphthalene
63.
Hexachlorocyclopentadiene
64.
2,4,6-Trichlorophenol
65.
2,4,5-Trichlorophenol
66.
2-Chloronaphthalene
67.
2-Nitroaniline
68.
Dimethyl phthalate
69.
Acenaphthylene
70.
3-Nitroaniline
Chemical Abstract
Service (CAS) Number
62-75-9
108-95-2
62-53-3
111-44-4
95-57-8
541-73-1
106-46-7
100-51-6
95-50-1
95-48-7
39638-32-9
106-44-5
621-64-7
67-72-1
98-95-3
78-59-1
88-75-5
105-67-9
65-85-0
111-91-1
120-83-2
120-82-1
91-20-3
106-47-8
87-68-3
59-50-7
91-57-6
77-47-4
88-06-2
95-95-4
91-58-7
88-74-4
131-11-3
208-96-8
99-09-2
continued-
7-5
-------
Table 7-2 - continued
Chemical Abstract
Number Seml-Volatiles Service (CAS) Number
71. Acenaphthene 83-32-9
72. 2,4-Dinitrophenol 51-28-5
73. 4-Nitrophenol 100-02-7
74. Dibenzofuran 132-64-9
75. 2,4-Dinitrotoluene 121-14-2
76. 2,6-Dinitrotoluene 606-20-2.
77. Diethyl phthalate 84-66-2
78. 4-Chlorophenyl phenyl ether 7005-72-3
79. Fluorene 86-73-7
80. 4-Nitroaniline 100-01-6
81. 4,6-Dinitro-2-methylphenol 534-52-1
82. N-Nitrosodiphenylamine 86-30-6
83. 4-Bromophenyl phenyl ether 101-55-3
84. Hexachlorobenzene 118-74-1
85. Pentachlorophenol 87-86-5
86. Phenanthrene 85-01-8
87. Anthracene 120-12-7
88. Di-n-butyl phthalate 84-74-2
89. Fluoranthene 206-44-0
90. Benzidine 92-87-5
91. Pyrene 129-00-0
92. Butyl benzyl phthalate 85-68-7
93. 3,3'-Dichlorobenzidine 91-94-1
94. 3enzo(a)anthracene 56-55-3
95. bis(2-Ethylhexyl) phthalate 117-81-7
96. Chrysene 218-01-9
97. Di-n-octyl phthalate 117-84-0
98. Benzo(b)fluoranthene 205-99-2
99. Benzo(k)fluoranthene 207-08-9
100. Benzo(a)pyrene 50-32-8
101. Indeno(l,2,3-cd)pyrene 193-39-5
102. Dibenz(a,h)anthracene 53-70-3
103. Benzo(ghi)perylene 191-24-2
continued-
7-6
-------
Table 7-2 - continued
Chemical Abstract
Number Pesticides Service (CAS) Number
104. alpha-BHC 319-84-6
105. beta-BHC 319-85-7
106. delta-BHC 319-86-8
107. gamma-BHC (Lindane) 58-89-9
108. Heptachlor 76-44-8
109. Aldrin 309-00-2
110. Heptachlor epoxide 1024-57-3
111. Endosulfan I 959-98-8
112. ' Dieldrin 60-57-1
113. 4,4'-DDE 72-55-9
114. Endrin 72-20-8
115. Endosulfan II 33213-65-9
116. 4,4'-DDD 72-54-8
117. Endrin aldehyde 7421-93-4
118. Endosulfan sulfate 1031-07-8
119. 4,4'-DDT 50-29-3
120. Endrin ketone 53494-70-5
121. Methoxychior 72-43-5
122. Chlordane 57-74-9
123. Toxaphene 8001-35-2
124. Aroclor-1016 12674-11-2
125. Aroclor-1221 11104-28-2
126. Aroclor-1232 11141-16-5
127. Aroclor-1242 53469-21-9
128. Aroclor-1248 12672-29-6
129. Aroclor-1254 11097-69-1
130. Aroclor-1260 11096-82-5
7-7
-------
When sampling is employed to obtain quantitative data on atmospheric
contamination it must be recognized that the distribution of a wind-bcrne
chemical varies over time with down-wind distance from the source. Therefore,
sampling may have to be conducted over extended time periods and distances.
Air quality data should provide adequate temporal as well as spatial resolu-
tion of air pollutant concentrations. -An air sampling survey designed to
provide good temporal resolution should include samples having a short enough
averaging time to measure the effects resulting from a variety of possible
combinations of source strength and meteorological phenomena. Obtaining
samples averaged over a short interval requires either continuous sampling or
frequent grab samples and results in a large quantity of data. Monitoring
long enough to sample all variations of source strength and meteorology may
take as long as one year. Continuous monitoring for one year usually assures
that samples are taken during each season, on all days, and during all hours.
Such sampling provides the best data describing the temporal distribution of
toxic chemical concentrations. The cost, however, may be unreasonably high
for maintaining and operating monitoring stations. There is also a long lag
time between initiation of the monitoring program and receipt of the final
results. A compromise is often made between the two situations, using the
best available methodology and ensuring that the sampling schedule includes
the time period expected to yield the highest toxic chemical concentrations.
7.1.2.2 Surface Water/Groundwater Contamination
Runoff and overland flow of toxic chemicals (from leaks, spills, etc..) are the
most likely sources of surface water contamination at hazardous waste sites.
Leaching of toxic chemicals from contaminated soils and the vertical migration
of coxic chemicals from lagoons ara the most likely sources of groundwater
contamination at hazardous waste sites.
Quantitative estimations of the degree of contamination of surface waters are
often derived by the application of mathematical models designed to provide
information on runoff releases and overland flow of toxic chemicals. These
models require data on the sorption partition coefficients for contaminants of
concern (sorption partion coefficients may be derived from octanol-water
partition coefficients). A model also exists for the rapid estimation of the
extent of groundwater contamination. This model requires data on various
site-specific and chemical-specific characteristics and provides order-of-
magnitude estimates of groundwater contamination.
Surface and groundwater contamination data may also be obtained via sampling.
As with air quality data, water sampling data should provide adequate temporal
and spatial description of water-borne chemical concentrations. The variation
of chemical concentrations in groundwater samples will generally be less than
chemical concentrations in surface water. Of particular concern when dealing
with water samples is the stability (integrity) of the sample. Complete
preservation of a sample is nearly impossible since biological, chemical or
physical changes in the sample will usually occur. However, to the extent
possible, adequate Integrity of samples must be maintained.
7-8
-------
7.1.2.3 Soil Contamination
Sources of surface soil contamination at hazardous waste sites include spills,
lagoon failure, contaminated runoff or intentional placement of waste on or in
the ground. Toxic chemicals can also be leached from surface soils to
subsurface layers. Generally, the substances of concern at uncontrolled
hazardous waste sites are non-polar and will bind (adsorb) strongly to organic
soil particles as a result of their hydrophobic properties.
No modeling methods are currently available to estimate toxic chemical concen-
trations in surface soils and direct sampling must be conducted to obtain data
on surface soil contamination. Analyses of toxic chemical concentrations in
subsurface soils may be performed by means of modeling, sampling, or a combi-
nation of both.
7.1.2.4 Contamination of Biota
In addition to assessing concentrations of toxic chemicals in air, water and
soils, performance of an exposure assessment requires the collection of data
on the occurrence of toxic chemicals in the tissues of plants and animals that
may be consumed as food by other organisms. Of special concern are organisms
(such as fish or shellfish) which are consumed by humans. The quality of
results in assessing toxic chemical concentrations in food depends entirely on
the method of sampling and sample preparation. Since even processed foods are
not homogeneous in quality, the design, sampling and interpretation of the
results requires extra care due to possible interactions between chemical,
food and nutrients. Important factors that must be considered in the sampling
and analysis of biota include both the characteristics of the chemical of
concern and the food of concern. Significant characteristics of chemicals
include the physical/chemical properties and sources of release. The food
characteristics that are important include the potential for contamination,
consumption data (regional patterns) and dynamics of food consumption.
7.1.2.5 Data Validation and Interpretation
In the collection of occurrence data for use in an exposure assessment, the
first task is to obtain relevant data that are complete, accurate and repre-
sentative of the actual situation and conditions. The measurements should be
comparable and maintained in consistent units. After collection of data, the
next task is to verify and validate the data values. This involves removing
erroneous and irrelevant measurements, testing for outliers (extremely high or
low values) and testing data for accuracy, precision and representativeness.
A detailed explanation should be given to justify the removal or inclusion of
questionable data values in the final data set. A decision must be made on
how to handle values below the detection limit or data reported with negative
values. The final task is to perform statistical tests which will allow for
interpretation of the data. Since the statistical methods employed are
usually based on the assumption of the normality of the data values, appro-
priate tests should be performed and transformations (such as logarithmic)
should be used if needed. The accuracy and precision of the estimates should
be reported in the form of standard deviation, coefficient of variation and
confidence limits.
7-9
-------
7.1.2.6 Collection of Occurrence Data—A Hypothetical Example
In order to illustrate the nature of the data required in evaluating the
occurrence of a toxic chemical in the environment, the following hypothetical
example is presented.
Prior to initiating an enforcement action at a hazardous waste site pursuant
to Section 106 of CERCLA, an endangerment assessment is required. One of the
first steps performed in preparing the endangerment assessment is an evalua-
tion of the history of the site. In evaluating the history of a particular
site it is determined that the waste present in the largest quantity is
1,1,2,2-tetrachloroethylene (perchloroethylene).
Results of the site history analysis indicate that the perchloroethylene was
disposed of in steel drums buried in deep, unlined pits. The site is located
several thousand feet from a reservoir which is a popular fishing location and
which also serves as the drinking water supply for several nearby towns.
Available toxicological information on perchloroethylene indicates that this
compound is mutagenic, carcinogenic and hepatotoxic. There is, therefore, a
high level of concern regarding the presence of this chemical at the site. A
review of data on the physical and chemical characteristics of perchloro-
ethylene reveals that the chemical is expected to be mobile in groundwater
and, therefore, could be expected to migrate to the reservoir if released from
the buried drums. A sampling plan, is subsequently devised to obtain data on
concentrations of perchloroethylene in the groundwater between.the site and
the reservoir, and in the reservoir itself. Sampling is also scheduled for
game fish species present in the reservoir because of the potential for human
exposure through ingestion of contaminated fish. Air sampling is not included
in the initial sampling plan since the depth of burial of the waste-filled
drums reduces concern for volatilized emissions. Similarly, soil sampling is
not scheduled because of the depth of waste location and because no human
exposure through contact with potentially contaminated soils is expected due
to the nature of the site (on relatively inaccessible private property and
fenced in).
When sampling is concluded and valid occurrence data for perchloroethylene are
available, the data can be used to calculate predicted human exposure levels
for the two expected exposure routes (drinking water and ingestion of
contaminated biota).
7.1.3 Legal Requirements
It is not always possible to know in advance whether or not monitoring data in
general, and exposure monitoring data in particular, may be used in enforce-
ment actions. Therefore, it is good practice to make the assumption, when
collecting monitoring data, that the data may be used in enforcement actions.
Accordingly, there are certain policies and procedures which should be
followed in order to ensure that the collected monitoring data will be accept-
able as evidence in a court of law. The purpose of this section is to present
a brief summary of some of the more important factors which must be taken into
consideration so that monitoring data may be acceptable for use in enforcement
cases. For more detailed information on this topic, the reader is referred co
the list of publications at the end of this chapter.
7-10
-------
Prior to the collection and analysis of samples, a draft sampling/analysis
plan should be prepared and circulated to appropriate technical and legal
personnel for review and comment. The final sampling/analysis plan should
then formally incorporate or rebut all comments received. The final plan
should also detail the sampling project's objectives, survey methods,
personnel and equipment requirements, safety program and equipment, custody
procedures, quality assurance procedures and report schedules.
Once samples have been collected they should be appropriately marked with tags
and maintained under Chain-of-Custody procedures. Possession of all samples
must be traceable from the time they are collected until they are introduced
as evidence in legal proceedings. Samples are accompanied by a Chain-of-
Custody Record, and all transfers betveen individuals must be documented by
signatures and records of the dates and times of transfer. Samples are
shipped (if by mail, registered with return receipt requested) to an appro-
priate laboratory for analysis, with a separate custody record identifying the
sample contents accompanying each shipment.
At the receiving laboratory, a designated sample custodian accepts custody of
the samples and verifies that the information on the sample tags matches that
on the Chain-of-Custody Record. After assigning a unique laboratory
identification number to each sample tag, the sample custodian enters the
sample tag data into a bound logbook arranged by project code and station
number. The sample custodian then distributes samples to appropriate analysts
who also maintain a record to show the Chain-of-Custody for each sample or
sample aliquot. All laboratory observations and calculations relevant to -a
sample are recorded by the analysts in serialized logbooks. The logbooks must
contain information sufficient to allow the analysts to recall and describe
succinctly each step of the analyses performed. Moreoyer, sufficient detail
is necessary to enable others to reconstruct the procedures followed should
the original analyst be unavailable for testimony. Any irregularities
observed during the analytical process need to be noted.
After an organization has completed its work for a particular project, all
documents generated should be assembled in the organization's files. The file
then becomes accountable. Any records taken from the file must be signed out.
Using proper Chain-of-Custody procedures, data obtained from the analyses of
samples are returned to the Sampling/Analysis Project Leader for use in the
preparation of a draft project report. In preparing reports, the ability to
substantiate and defend the contents is of foremost Importance. All draft
reports should be dated and numbered and be accountable. All draft copies of
the report must be returned to the Sampling/Analysis Project Leader after
review. Once comments have been incorporated and the final report is avail-
able, all draft copies should be disposed of properly.
The following is a list of documents which provide more detailed information
on the procedures which should be followed to ensure that sampling data will
be acceptable as evidence In a court of law.
7-11
-------
1. The EPA's "Handbook for Analytical Quality Control in Water and
Wastewater Laboratories" (Booth 1979) addresses in detail the
following areas: laboratory facilities, instruments, glassware
requirements, and reagents. It further deals with analytical
performance, data handling, and data reporting. Separate chapters
are devoted to the special requirements for trace organic analysis,
water and wastewater sampling, microbiology, aquatic biology, and
safety.
2. The "Quality Assurance Guidelines for Biological Testing" (Stanley
1978) discusses the following elements of quality assurance: quality
assurance policy and objectives, design and analysis of experiments,
sampling, precision and assurance of tests, physical environment of
research, chemicals and reagents, control of performance, and data
handling and reporting.
3. . The "Handbook for Sampling and Preservation of Water and Wastewater"
(3erg 1982) is not an official EPA manual, but is a reference to be
used as an input to EPA manuals and guidelines. The report presents
general considerations for sampling; sample preservation and hand-
ling methods, sampling methods for wastewaters, surface waters, and
bottom sediments; methods for collection of microbial samples, and
statistical analysis methods.
4. The "Quality Assurance Guidelines for IERL-CI Project Officers"
(Stratton and Bonds 1979) is designed for use, as the-title implies,
by EPA project officers-. The report provides quality assurance
guidelines for procurement of projects requiring sampling and
analysis, guidelines for monitoring such projects and auditing
procedures. The document should assist in providing an
understanding of what EPA project officers expect in terms of
quality assurance.
5. Another source of EPA quality assurance procedures is the "Manual of
Analytical Quality Control for Pesticides and Related Compounds in
Human and Environmental Samples" (Sherma 1977). A general
description of pesticide residue analytical procedures is provided
following a discussion on inter- and intra-laboratory quality
control. Also covered are procedures for analysis of samples
Including extraction, isolation, and gas chromatography. Although
the manual deals primarily with pesticides, many of the procedures
and recommendations apply to the analysis of any organic chemical.
7.2 Identification and Analysis of Exposed Populations
The Identification and analysis of exposed human populations generally
Involves four distinct steps, each of which requires unique information. In
the first step, population data are compared with occurrence data in order to
quantify the population that will actually or potentially come into contact
with contaminated air, water, soil or food. The second step, population
characterization, involves determining whether any groups within the exposed
oopulation may experience a greater risk than the average population as a
7-12
-------
result of exposure. High risk groups might include women of childbearing age,
the chronically ill, infants, children or the elderly. The third step,
activity analysis, involves an examination of the activities in which a
potentially or actually exposed population engages. The final step,
development of exposure coefficients, yields information critical to the
calculation of exposure levels. Each of these steps is discussed in further
detail below. The analysis of nonhuman populations is also discussed.
7.2.1 Population Identification and Enumeration
The first step in identifying and analyzing exposed populations is to
determine the number of individuals exposed and the routes of exposure. For
humans, the Census of Population (most recently conducted in 1980) can be
accessed to determine the size, distribution and demographic characteristics
of a geographically defined population. Census data are especially useful in
quantifying populations exposed as a result of their presence in a specific
locale (e.g., those exposed to toxic chemicals in ambient air or soil). For
example, identification/enumeration of populations exposed to airborne"
toxic chemicals may be accomplished by overlaying an isopleth map of
chemical concentrations around a source on census maps.
In the analysis of populations that may be exposed to chemicals present in
surface or groundwater, all persons in the service area of a water supply
system that draw water from a contaminated -source must be considered as
potentially exposed through ingestion and dermal exposure while bathing.
Information concerning local drinking water sources and populations served can
usually be obtained from the local Department of Public Works, Planning
Department or Health Department. Information on public departments or private
drinking water treatment companies that use groundwater as their raw water
supply, as well as the number of households drawing water from private veils,
will also generally be available from these sources. Swimmers in concaminaced
waters may also comprise a portion of the exposed population.
Dermal exposure to contaminated soils could constitute an exposure route for
individuals who work outdoors or for children playing outdoors. In addition,
children sometimes ingest toxic chemicals contained in soil through pica
behavior (eating soil). Direct soil exposure is, in most cases, minor in
magnitude when compared to other routes, and it is often difficult to quantify
the actual level of transmission of soil-absorbed toxic chemicals across skin.
Exposure to airborne contaminated soil particulates and substances
volatilizing from soil may, however, be quite significant, and must be
considered as an integral component of the overall site air contamination/
exposure analysis.
Human exposure to contaminated food will usually be associated with fruit and
vegetables grown in home gardens, or with fish or game residing in
contaminated areas.
7.2.2 Population Characterization
After exposed populations have been identified and enumerated, they should be
characterized by -ge and se:: f^ctcrs -±t.zi the physiological parameters that
determine dose (e.g., breathing rate, skin surface area, food and water
7-13
-------
ingestion) are often age- or sex-specific. Also, from a toxicity standpoint,
subpopulations defined by age or sex may be especially susceptible to toxic
chemicals. Thus, characterization of exposed populations permits the
determination of exposure distributions within the population at large, and
the delineation of specific high-risk subpopulations.
7.2.3 Activity Analysis
Activities engaged in by members of a given population or subpopulation can
dramatically affect the level of exposure to environmental chemicals. For
example, persons whose lifestyle or employment involves frequent strenuous
activity will ventilate larger volumes of air per unit time than will those
living a less strenuous life and will, therefore, experience a greater level
of exposure to airborne chemicals. Another example would be if activities
take persons away from the source of the exposure such as weekends in the
country. Also seasonal factors can affect exposure levels, e.g., dermal
contact with contaminated soil is considerably reduced during freezing
weather.
Key activity-related exposure determinants to be quantified in the activity
analysis phase for each exposure mechanism are:
• Ingestion: amount of contaminated food or water ingested
(per unit time).
« Inhalation: length of time (frequency and duration) spent in each
related activity; nature of the activity in terms of
light, medium, heavy, or maximum exertion (per unit time).
• Dermal exposure: length of time (frequency and duration) spent in
each related activity (per unit time).
7.2.4 Development of Exposure Coefficients
The final component of the identification and analysis of exposed human
populations is the development of exposure coefficients. Ar. exposure coeffi-
cient is a term which combines information on the frequency, mode and magni-
tude of contact with each contaminated medium to yield a quantitative value of
the amount of contaminated medium contacted per day. Exposure coefficients
are developed for each exposure route and are used in calculating exposure
levels. Developing an exposure coefficient requires that certain assumptions
be made regarding intake of air, food and water. Some standard intake
assumptions are provided in Table 7-3.
7.2.5 Nonhuman Population Analyses
Nonhuman populations are generally more difficult to quantify and characterize
than are human populations because less information is available and the data
that are available are often measured in units other than numbers of
organisms. Populations may, for example, be expressed as herds or biomass.
The adequacy and sources of nonhuman population data vary considerably,
depending on the organism. As might be expected, most research effort has
been invested in plant and animals species that have economic significance.
7-14
-------
TABLE 7-3 STANDARD INTAKE ASSUMPTIONS FOR HUMANS
,v Adult Adult Child
Respiratory Rate Male Female (10 yrs)
Resting (L/min) 7.5 7.0 4.8
Light Activity (L/min) 20.0 19.0 13.0
Fluid Intake(b)
Milk (L/day) 0.30 0.20 0.45
Tap Water (L/day) 0.15 0.10 0.20
Other (L/day). . 1.50 1.10 0.75
Total (L/day) 1.95 1.40 1.40
(a}
Food Consumption 1,500 g/day — —
(a) Adapted from Callahan et al. (1983)
(b) Adapted from International Commission on Radiological Protection (IRCP)
(1975).
(c) EPA usually uses 2 L/day as total fluid consumption for adults and 1 L/day
for children.
7-15
-------
In most cases, it is necessary to gather the data from diverse sources to
enumerate nonhuman populations. The major sources of data include journal
articles and other publications, Federal and state agencies, and selected
private agencies. When data from these sources are not available, field
surveys may be required.
7.3 Calculation of Exposure
Calculation of doses of contaminants incurred by exposed populations is
accomplished by integrating the results of the collection of chemical
occurrence data (described in section 7.1) with the results of exposed
population analyses (described in section 7.2).
To calculate the dose incurred, the concentration of the chemical in an
environmental medium is first multiplied by the appropriate exposure
coefficient. This calculation provides an estimate of the total amount of
each chemical to which the population is exposed on a daily basis. However,
this value must be adjusted to account for the extent.to which each chemical
is transferred across the membranes of the exposed organism (i.e., the extent
of absorption). This adjustment is accomplished by multiplying total daily
exposure values by an absorption factor. Usually, absorption factors cited in
the toxicological literature are employed in this calculation. When empiri-
cally derived absorption factors are not available, an absorption factor of
unity is applied, thereby generating a conservative (worst-case) estimate of
the dose incurred. Finally, this whole-body dose estimate (mg/day) is
converted to terms of mg of contaminant/kg of body mass/day by dividing it by
the body mass representative of the receptor population.
Sample Calculation
The following hypothetical example illustrates the procedure for calculation
of dose incurred:
A factory is located near a toxic dump site which has improperly disposed of
large quantities of carbon tetrachloride (CC1,) . The factory draws water for
drinking and industrial uses from a well whicR is found to ^ontain 1.6 mg/L of
CC1,, and average air concentrations are 0.1 ppm (0.63 mg/m ). The factory
employs adult males who work 8-hour shifts doing light physical labor.
Calculation of Exposure Coefficient:
Since the adult males spend about one-half of their waking hours in the
factory, the amount of water they consume may be estimated to be one-half of
average total water consumption for adults (2 L/day/2 ¦ 1 L/day). The
inhalation exposure coefficient is calculated as follows:
(20 L/min) (60 min/hr) (8 hr/day) • 9,600 L/day
Converting to more convenient units, this becomes:
(9,600 L/day) (10~3 m3/L) - 9.6 o3/day
7-16
-------
Absorption Coefficients:
Inspection of published toxicokinetic information on CC1, reveals that
gastrointestinal absorption is about 90% and absorption In the lung is about
302, so the absorption coefficients are 0.90 and 0.30, respectively.
Dose Calculation:
A. Daily Dose via Water ¦ (Exposure Coefficient)(Water Concentration)
» (1 L/day)(1.6 mg/L)
¦ 1.6 mg/day
Daily Dose via Air « (Exposure Coefficient)ŁAir Concentration)
- (9.6 m /day)(0.63 mg/m )
* 6.0 mg/day
B. Absorbed Dose ¦ (Daily Dose)(Absorption Coefficient)
For Water: (1.6 mg/day)(0.90) « 1.4 mg/day
For Air: (6.0 mg/day)(0.30) » 1.8 mg/day
Total Absorbed Dose: 1.4 mg/day + 1.8 mg/day ¦ 3.2 mg/day
C. Conversion to mg/kg basis for adult (70 kg):
(3.2 ms;/day) „ ,, ,,
—(7Q g j 7 - 0.05 mg/kg/day
It is important in all caculations of exposure levels that assumptions
regarding body weight, water and air intake, food consumption, etc., be
clearly identified as such, co distinguish assumed average values from actual
measured values.
7.4 Estimation of Past Exposure
In some instances, it may be desirable to obtain estimates of past exposure
levels to certain chemicals. This can sometimes be accomplished by measuring
the concentration of the chemical in body fluids or tissues of exposed
organisms.
Following the absorption of a chemical by a particular organism, the
chemical distributes throughout the organism and may accumulate in various
parts of the organism. The site of accumulation may or may not be the major
site of toxic action. When a chemical accumulates at a site other than the
site at which it produces toxic action, the site of accumulation may be
referred to as a "storage site." The accumulation in this storage site may
serve as a protective mechanism, whereby the concentration of the chemical in
the storage site keeps the concentration of the chemical in a, vulnerable organ
at a low level. However, the concentration of the chemical in the storage
site is in equilibrium with the chemical in blood, and therefore it is slowly
released into the circulation at a rate equal to the rate at which the
chemical is excreted or removed from the blood. This tends to maintain the
chemical in the blood for long periods of time. Some of the major storage
areas fcr chemicals include plasma proteins, liver, kidney, fat ar.d bcr.e.
7-17
-------
Therefore, measurements of toxic chemical concentrations in these storage
areas are the most useful in terms of estimating past exposure levels.
It should be noted that this means of estimating past exposure levels has a
major limitation. Not all chemicals accumulate to the same extent. There-
fore, estimation of past exposure levels by measuring accumulated chemical
concentrations is useful only for those chemicals that tend to accumulate.
This method provides no useful information on past exposure to chemicals that
are efficiently cleared from the body.
7.5 Prediction of Future Exposure
In order to predict future exposure scenarios, it is necessary to generate
estimates of the direction of travel and future ambient concentrations of
toxic chemicals within all environmental media. Some important physical/chemi-
cal properties of chemicals that may influence their environmental fate are
presented in Table 7-4. Some important site characteristics that may influ-
ence environmental fate are presented in Table 7-5. Whenever mathematical
models are employed to calculate expected future exposures, it is important to
identify calculated values as such (to avoid confusion with measured values)
and to cite the model employed. Population predictions, such as future popu-
lation characteristics and distribution patterns, are also necessary so that
they can be compared with future chemical distribution/concentration data in
order to predict future exposure scenarios.
7.5.1 Environmental Stability
In analyses of the environmental fate of chemicals, all processes that are
likely to lead to transformation of chemicals in various environmental media
must be identified. In conducting environmental fate analyses, each
transformation process is assessed and, when possible, rates of all processes
are quantified. Usually, information regarding the significance and rates of
transformation processes in various media is available in the technical
literature. Some transformation processes likely to be important in assessing
the fate of chemicals are described below.
Photolysis is the degradation of a chemical caused by exposure to light. This
process may be important to chemicals in the atmosphere, in surface water of
sufficient clarity for penetration of light, and on the surface of the soil.
Photolysis typically results in the degradation of chemicals through rupture
of covalent bonds. Factors determining the rate of photolysis include: flux
of photons to the substrate media, chemical-specific light adsorption
coefficients, and reactor yield constants (i.e., the efficiency of the
degradation process with available photon energy).
Examples of photolysis include the splitting of hydrogen iodide by the
reaction
2HI ——> H2 + I2
and the splitting of ketene (CH^CO) into carbon monoxide (CO) and carbene
(methylene) (CH^).
7-18
-------
TABLE 7-4 PHYSICAL AND CHEMICAL PROPERTIES THAT
MAY INFLUENCE ENVIRONMENTAL FATE
Adsorption Coefficient
Blodegradatlon Rate
Boiling Point
Henry's Law Constant
Hydrolysis Rate
Melting Point
Molecular Weight
Octanol/Water Partition Coefficient
Photolysis Rate
Vapor Pressure
Volatilization Rate
Water Solubility
7-19
-------
TABLE 7-5 SITE CHARACTERISTICS THAT MAY INFLUENCE
ENVIRONMENTAL FATE
Ambient Moisture
Ambient Humidity Levels
Ambient Temperatures
Ambient Wind Velocity and Direction
Geologic Characteristics
Hydrologic Characteristics
Soil Characteristics
Topographic Features of Site and Surrounding Area
Vegetative Cover of Site and Surrounding Area
Watershed/Land Use Characteristics
7-20
-------
Indirect photolysis is an important transformation process for many chemicals.
This process involves intermediate compounds present in the environmental
medium which undergo direct photolytic transformation and thereby become
reactive with a toxic chemical (usually as a strong oxidizer). These
intermediates subsequently react with a chemical, effecting its trans-
formation. Indirect photolysis rates are governed by the concentration and
light adsorption coefficient of the intermediate compound and the rate of its
subsequent reaction with the toxic chemical.
Oxidation is the process of removing electrons from a chemical. These are
accepted by another chemical (an oxidizing agent). Chlorine in drinking water
and ozone in the atmosphere are two Important oxidizing agents. Oxidation by
oxygen is also important, and occurs in air, surface water, groundwater and
soil. The rate of oxidation is determined by the concentration of the
oxidizing agent, and its reaction rate constant with respect to the chemical.
For example, cupric ion is the oxidizing agent in the following reaction:
Fe + Cu2+ Fe2+ + Cu
Here, two electrons are transferred from the Fe atom to the Cu atom. Thus,
the Fe atom is oxidized (becomes positively charged by the loss of two
electrons) while the Cu atom receives the two electrons and becomes neutral
(is reduced).
Hydrolysis is the splitting of a chemical bond by insertion of watet.
Hydrolysis may occur in the air, surface water, groundwater or soil.
Hydrolysis is most important for organic chemicals which contain as part of
their structure one .or more easily polarized, electronegative groups
covalently bonded co carbon atoms. Covalent inorganic chemical ccr.pcurda zloc
undergo hydrolytic transformation, usually at rapid rates. Hydrolysis is
highly pH-dependent, and can be acid- or base-catalyzed.
The hydrolysis of an organic cyanide (nitrile) is shown below as an example of
this type of reaction:
2H 0
RC=N =—> RC0,H + NH,
t Z 4
H
Chemical Speclation. Metallic chemical compounds undergo a wide range of
transformation processes, forming and breaking down complexes with inorganic
or organic ligands present in the substrate medium. These processes, referred
to collectively as speciation, can occur in all environmental media. The
final speciation of a metal in a given environment (i.e., its degree of
association with various complexing agents) affects its bioavailability,
solubility, volatility and sorption properti^. For example, lead chloride
(PbCl,) is a fairly soluble form of lead (Pb ), and could easily dissolve in
groundwater and escape from a toxic dump site. However, if the soil were rich
in carbonates (CO^ ), the lead would form lead carbonate (PbCO^), which is
much less soluble:
PbCl2 + C032" - PbC03 + 2C1~
7-21
-------
Bioconcentration and biotransformation are processes mediated by organisms.
Bioconcentration is the uptake and accumulation of a chemical vithin the
tissue of exposed organisms (fish, plants, birds, etc.). often within specific
tissues or organs. Bioconcentration rates are a function of the uptake and
excretion rates of the exposed organisms and the lipophilic properties of the
chemical. Biotransformation is a collective tern for the various reactions
that occur as a result of metabolism by organisms such as* soil bacteria.
Biotransformation commonly causes the degradation and detoxification of toxic,
organic chemicals. Both bioconcentration and biotransformation are dependent
on the metabolic rates of the organisms which, in turn, are functions of
various environmental parameters, including the availability of nutrients,
sunlight, moisture, ambient temperature, and pHx. These transformation
processes occur In the water, soil, groundwater and, to a limited extent, in
the air.
An example of a biodegradation pathway is illustrated by the microbial
degradation of the insecticide parathion. In this pathway, Pseudoacnas
stutzerl initially converts parathion to p-nitrophenol which is subsequently
utilized by Pseudomonas aeruginosa.
—• stutzeri P. aeruginosa
Parathion » p-Nitrophenol > Further Degradation Products
Table 7-6 presents some information on the environmental stability
(persistence) of selected chemicals from Table 7-2.
7.5.2 Environmental Mobility
The movement of chemicals in the environment is determined by "he physical and
chemical properties of the chemicals themselves and by the characteristics or
the environment. Chemicals may move both within media (intrameclxa transport)
and from one environmental medium to another (intermedia transport).
7.5.2.1 Intramedia Transport
Atmospheric transport of chemicals is governed by diffusion and by
meteorological variables such as air currents (wind speed and direction)i
humidity, particulate concentrations and air temperature differentials. The
size of the particle to which a toxic chemical is adsorbed is an important
determinant of the extent to which it will be transported through the
atmosphere.
In water, diffusion through the water and currents within the water are the
principal routes of transport. Aquatic transport processes are, therefore,
dependent on whether the water body is a lake, slow river, swift stream,
estuary or ocean. For example, tidal movement is the overriding transport
mechanism within an estuary, while thermal stratification may control movement
within a deep lake or reservoir.
Chemicals are transported within soil principally by water flow. Aside from
the rate of water flow, the key parameter is the interaction (sorption and
7-22
-------
TABLE 7-6 ENVIRONMENTAL PERSISTENCE OF SOME COMPOUNDS
THAT MAY OCCUR AT HAZARDOUS WASTE SITES
Chemical
Volatiles;
Vinyl chloride
Chloroform
1,1,1-Trichloroethane
Carbon tetrachloride
Dibromochlororae thane
Bromoform
Chlorobenzene
Semi-Volatiles:
bis(2-Chloroethyl) ether
2-Chlorophenol
2,4-D ime t hyIpheno1
2,i-Qichlorophenol
Hexachlorocylcopentadiene
2,4,6-T richlorophenol
2-Chloronaphthalene
2,4-Dinitrophenol
2,6-Dinitrotoluene
Benzidine
Environmental Persistence
Half life: 25 minutes
Half life in bluegill: 1 day
Half life in water: 15 months
Hydrolysis half life:
freshwater: 5-9 months
marine : 39 months
Half life in bluegill: 1 day
Half life projection: 274 years
Hydrolysis half life: 686 years
Half life in water> 5.8 hours
Half life in bluegill: 4-7 days
Half life in bluegill: 1 day
Complete biodegradation: 2 months
Half life: less than 1 day
Half life in aerated water: 8-9 days
Half life: 11 days
Half life in rat blood: 20 hours
Biomagnified in water
No significant microbial degradation
in 64 days
Greater than 64 days
Half life in dogs and rats: 68-88 hours
continued-
7-23
-------
Table 7-6 - continued
Chemical
Semi-Volatiles (continued):
3,3'-Dichlorobenz idine
Pesticides;
Dieldrin
Chlordane
Adapted fron McNeils et al. (1984)
Environmental Persistence
Long life In soil
Half life in water: 723 days
Very persistent
7-24
-------
desorption) of Che chemical with soil particles. Sorption is controlled by
the characteristics of the chemical and by the type (i.e., organic content) of
the soil. Siterspeciflc Information on soil composition should be used
whenever possible to characterize soil transport when disposal sites or other
areas of soil contamination are identified by location.
7.5.2.2 Intermedia Transport
The physical and chemical properties of a particular chemical typically have
the greatest impact on intermedia transport. Transport from air to water is
dominated by two process: (1) scavenging via liquid deposition (rain, snow),
and (2) gravitational settling (dry deposition). Transport from water to air
is mediated by air-stripping, aerosol formation and volatilization. Air-
stripping and aerosol formation are most important under agitated conditions.
The importance of volatilization for a chemical can be evaluated by assessing
the chemical's solubility in water, vapor pressure, Henry's Law Constant,
molecular weight and activity coefficient.
Transport from ground to air may also occur through volatilization from soil.
Volatilization of organic chemicals from soil is strongly influenced by
sorption phenomena and local frequency of precipitation. Volatilization from
soil may also be influenced by seasonal differences in air and ground tem-
peratures; volatility from soli is enhanced when the land is warmer than sur-
rounding air. The other major ground-to-air transport process, resuspension
of dusts, also is affected by precipitation as well as by wind speed.
Chemicals are transported from air to ground by dry and wet deposition.
Transport between water and land is, to a large extent, controlled by the
solubility and leachability of a chemical and the affinity of the chemical for
the organic content of soil and sediment. Soil particles upon vhich
chemicals are sorbed are washed Into water bodies by surface runoff; once the
chemical is in the water, equilibrium partitioning between water and suspended
solids is established.
7.5.3 Environmental Partitioning
Partitioning refers to the relative distribution of a chemical among environ-
mental compartments. Partitioning analysis is essentially a qualitative
effort that is undertaken when the time or financial resources do not allow
for a detailed analysis of the environmental transport and transformation of a
chemical. A partitioning analysis may be based on examination of a
chemical's physical and chemical properties, analogy with other chemicals
whose fate is fairly well documented or mathematical modeling.
Examination of a chemical's physical and chemical properties can often permit
an estimate of its environmental partitioning. The following are the most
useful parameters for estimating the relative partitioning of a chemical:
• Vapor pressure is indicative of a chemical's ability to volatilize to
the atmosphere.
7-25
-------
• Water solubility of a chemical is indicative of its ability to be dis-
tributed by the hydrologic cycle. Water solubility also affects other
degradation pathways (e.g., photolysis, hydrolysis, oxidation) and
specialized transport pathways (e.g., volatilization from solution and
washout from the atmosphere by rain).
• Octanol/water partition coefficient (K ) is indicative of a chemical's
tendency to partition itself between an organic phase (e.g., fish,
soil) and an aqueous phase. It is particularly important for esti-
mating bioconcentration factors for aquatic life.
• Boiling point, in addition to serving as an indicator of the physical
state of a chemical, provides an indication of its volatility.
• Henry's Law Constant is indicative of an organic chemical's tendency
to volatilize from water to air.
• Adsorption coefficient (K ) is indicative of the extent to which a
chemical partitions between solid and solution phases (e.g., water-
saturated or unsaturated soil, runoff water, and sediment).
Adsorption potential to soil/sediment is indicative of a chemical's
environmental mobility.
Qualitative analysis of the fate of a chemical can also be made by analogy
with other chemicals whose fate is well documented. If the chemical under
study is structurally similar tp a previously studied chemical, some parallel
can be drawn to the environmental fate of the analogue.
A more detailed mathematical modeling analysis of partitioning aay also be
undertaken when resources permit and where environmental loadings (i.e.,
emission) data are available.
7.5.4. Population Predictions
As noted in section 7.2, performing an exposure assessment requires a
characterization of exposed populations (number, age, sex, race, activities,
etc.). Generally, predictions regarding future exposure scenarios require the
same types of information on exposed populations as assessments of present
exposures. That is, potentially exposed populations must be identified, enum-
erated and characterized, an activities analysis must be performed and
exposure coefficients must be estimated. (Section 7.2 of this handbook,
provides descriptions of these aspects of exposed population analyses.) Based
on this information, predictions regarding future exposure levels can be made.
7.6 Key Guidance and Implementation Documents
In 1984, the EPA proposed guidelines for use in conducting exposure
assessments (49 FR 46304). These guidelines, which will be revised by EPA as
necessary to reflect the benefit of experience, provide a general approach and
framework for carrying out exposure assessments for specified pollutants. In
some cases these guidelines may be useful only as a rough template; in other
cases they may serve as a model which can be closely followed.
7-26
-------
According to the guidelines, the five major topics to be addressed in most
exposure assessments include:
• sources
• exposure pathways
• monitored or esCimated concentration levels and duration
• exposed population(s)
• integrated exposure analysis
These five topics are appropriate for exposure assessments, whether the
assessments are of a global, national, regional, local, site-specific or
workplace-related nature. The topics are also appropriate for exposure
assessments on new or existing chemicals and radionuclides. They are also
applicable to both single media and multimedia assessments.
In general, the guidelines should be followed, to the extent possible, when an
exposure assessment is conducted as a required element in a regulatory process
or when an exposure assessment is conducted on a discretionary basis to
support regulatory or programmatic decisions. Application of these guidelines
to the conduct of exposure assessments results in a number of advantages
including the following:
1. While there will usually be gaps in the data required to perform an
exposure assessment, the use of the guidelines should help avoid
inadvertent errors of omission.
2. Use of the guidelines should promote consistency among various
exposure assessment activities. For example, use of the guidelines
should result in consistency with respect to common physical,
chemical and biological parameters and with respect to assumptions
about typical exposure situations. Consistent presentation of the
possible range estimates will enhance the comparability of results
and allow for improvement of exposure assessment procedures through
the sharing of common data and experiences.
3. The primary objective of most exposure assessments is to provide
reliable data and/or estimates for subsequent use in conducting a
risk assessment. The guidelines provide a common approach to
format which simplifies the process of reading and evaluating
exposure assessments, thereby facilitating the integration of the
exposure and hazard assessments.
In addition to EFA's proposed exposure assessment guidelines, a number of
other documents provide valuable information regarding the performance of
exposure assessments. These documents are listed below.
1. Superfund Exposure Assessment Manual (Versar 1984)—This document
presents an integrated methodology designed to guide the execution
of the major components of qualitative and quantitative exposure
assessments for hazardous waste sites.
7-27
-------
2. Exposure Assessment Methodologies for Hazardous Waste Sites (McNeils
et al. 1984)—This document presents a uniform approach to the
estimation of exposures to Important receptors from defined
hazardous waste constituents via all important exposure pathways.
The efficient use of available or historical information and the
need for exploratory programs is emphasized. Applicable theoretical
and empirical models which are accessible are described in
sufficient detail to facilitate decisions regarding their use in
designing and conducting an exposure assessment program.
3. Remedial Investigations Guidance Document (JRB 1984)—This document
provides a detailed structure for field studies involving data
collection for remediation decisions pertaining to hazardous waste
sites.
7. 7 Background References
Anderson E, Browne N, Duletsky S, Warn T. 1984. GCA Corporation.
Development of statistical distributions or ranges of standard factors used in
exposure assessments. Revised draft final report. Washington, DC: U.S.
Environmental Protection Agency, Office of Health and Environmental
Assessment, Exposure Assessment Group. Contract No. 68-02-3510.
Berg EL. 1982. Handbook for sampling and sample preservation of water and
wastewater. Cincinnati, OH: U.S. Environmental Protection Agency.
EPA-6C0/4-82-029.
Booth RL. 1979. Handbook for analytical quality control in water and waste-
water laboratories, Cincinnati, OH: U.S. Environmental Protection Agency.
EPA-600/4-79-019.
IRCP. 1975. International Commission on Radiological Protection. Report of
the task group on reference man. No. 23. Oxford, England: Pergamon Press.
JRB. 1984. Remedial investigations guidance document. Washington, DC: U.S.
Environmental Protection Agency. Contract No.: 68-03-3113.
McNelis DN, LBarth DS, Khare M, LaPoint TW, Yfantis EA. University of Nevada,
Las Vegas. 1984. Exposure assessment methodologies for hazardous waste
sites. Las Vegas, Nevada: Environmental Monitoring Systems Laboratory,
Office of Research and Development. Cooperative Agreement No. CR 810550-01.
Sherma J. 1977. Manual of analytical quality control for pesticides and
related compounds in human and environmental samples. Research Triangle Park,
NC: U.S. Environmental Protection Agency. EPA-600/1-79-008.
Stanley RE. 1978. Quality assurance guidelines for biological testing. Las
Vegas, Nevada: U.S. Environmental Protection Agency. EPA-600/4-78-043.
Stratton CL, Bonds JD. 1979. Quality assurance guidelines for IERL-CI
project officers. Cincinnati, OH: U.S. Environmental Protection Agency.
EPa-600 / 9- 7 1.
7-28
-------
USEPA. 1984. U.S. Environmental Protection Agency. Office of Health and
Environmental Assessment. Proposed guidelines for exposure assessments;
request for comments. Fed. Reglst., Nov. 23, 1984, 49 46304.
Versar. 1984. Superfund exposure assessment manual. Washington, DC: U.S
Environmental Protection Agency. Contract Nos,: 68-01-6271, 68-03-3149.
Whitmore RW. 1984. Research Triangle Institute. Methodology for
characterization of uncertainty in exposure assessments. Washington, DC:
U.S. Environmental Protection Agency, Office of Health and Environmental
Assessment, Exposure Assessment Group. Contract No. 68-01-6626.
7-29
-------
8.0 RISK ASSESSMENTS
Risk assessments are invaluable scientific tools for hazardous waste site
investigations. Management decisions regarding the regulation of chemicals
present at hazardous waste sites require a thorough assessment of the
potential risk to human health and the environment resulting .from exposure to
the chemicals present. Hazardous waste site risk assessments may be utilized
for many purposes including:
• Screening and determining response priorities.
• Ranking the relative toxicities of a large number of chemicals at a
site.
¦ Estimating risk associated with various clean-up alternatives.
• Supporting the assertion that an enforcement action is warranted under
CERCLA.
This chapter will discuss the structure of the overall risk assessment process
and explain how information developed during the toxicity assessment and
exposure assessment processes is combined to determine the magnitude of the
problem (i.e. probability of risk) at a site. This chapter will also
introduce some specific applications of risk assessment- at hazardous waste
sites, such as the Hazard Ranking System, Endangerment Assessments and
Feasibility Studies.
8.1 Risk Assessment Process
In the simplest sense, population risks from toxic chemicals are a function of
cwo measurable factors: toxicity and exposure. In order to cause a risk to
humans, a chemical must be (1) toxic, (2) present in the environment at some
significant level and (3) accessible for human exposure. Risk assessment is
an evaluation and interpretation of the available scientific evidence on these
points, providing a judgment and, if appropriate, an estimate of the
probability that risk exists.
The risk assessment process consists of one or more of the following four
steps:
• Toxicological Evaluation
0 Dose-Response Evaluation
• Exposure Assessment
• Risk Characterization
The toxicological evaluation and dose-response evaluation collectively
comprise the toxicity assessment process at a hazardous waste site. The risk
assessment process is usually initiated by performing a toxicological
evaluation and a preliminary exposure assessment on the chemlcal(s) of
8-1
-------
concern. If the results of either are negative (i.e., the chemical is not
toxic or is not present at some significant level), then it is not necessary
to proceed with the risk assessment.
Figure 8-1 provides an overview of the risk assessment process.
8.1.1 Toxicologlcal Evaluation
The toxicologlcal evaluation should answer the question "Does the chemical
have an adverse effect?" It is a qualitative evaluation of the scientific
data to determine the nature and severity of actual or potential health
hazards associated with exposure to a chemical substance. This step involves
a critical evaluation and interpretation of toxicity data from
epidemiological, clinical, animal and in vitro studies. Factors that should
be considered during the toxicologlcal evaluation for each contaminant include
routes of exposure, types of effects, reliability of data, dose, mixture
effects and the strength of evidence supporting the conclusions of the
toxicological evaluation. The toxicologlcal evaluation should also identify
any known quantitative indices of toxicity (e.g., NOAEL, LOAEL, carcinogenic
risk factors, etc.). The elements of the toxicological evaluation are
discussed in detail in Chapters 3.0, 4.0 and 5.0 of this handbook.
8.1.2 Dose-Response Evaluation
Once the toxicological evaluation indicates that a chemical is likely to cause
a particular adverse effect, the next step is to determine the potency of the
chemical. The product of the dose-response evaluation is an estimate of the
relationship between the dose of a chemical and the incidence of the adverse
effect in the human population.
The dose-response evaluation for non-carcinogenic chemicals provides an
estimation of the NOAEL or LOAEL and the margin of safety associated with the
conditions of exposure of the human populations at potential risk.
The dose-response evaluation for carcinogenic chemicals provides an estimation
of the probability or range of probabilities that a carcinogenic effect will
occur under the conditions of exposure of the human populations at risk.•
These estimates of probability are derived using mathematical models of the
dose-response relationship.
The dose-response relationship is introduced in chapter 3.0 of this handbook
and discussed in detail in chapter 6.0.
8.1.3 Exposure Assessment
The exposure assessment describes the likely degree of human exposure to a
chemical of concern at a hazardous waste site. The objectives of a
site—specific exposure assessment are to Identify actual or potential routes
of exposure, characterize the population exposed and determine the extent of
the exposure at a site. The product of the exposure assessment process, as
detailed in chapter 7.0, is an estimation of exposure levels or doses incurred
for chemicals of concern at the site.
8-2
-------
Toxicity Assessment
Does the chemical cause an
adverse effect?
(lexicological Evaluation)
What is the relationship
between the dose and the
adverse effect in humans?
(Dose-Response Evaluation)
Exposure Assessment
What are the current and
projected exposure levels
at the site?
Risk Characterization
What is the estimated
incidence of the adverse
effect in a given population?
FIGURE 8-1 RISK ASSESSMENT PROCESS AT HAZARDOUS WASTES SITES
The risk assessment process at hazardous waste sites Is an
evaluation and Interpretation of the available scientific
evidence on the toxicity and exposure potential for a
chemical of concern. The results of the Toxicity and
Exposure Assessments are integrated to yield a complete
characterization of risk, at the site.
8-3
-------
8.1.A Risk Characterization
The final step in risk assessment, risk characterization, is the process of
estimating the incidence of an adverse health effect under the conditions of
exposure described in the exposure assessment. It is performed by integrating
the information developed during the toxicity assessment (toxicological
evaluation and dose-response evaluation) and the exposure assessment to yield
a complete characterization of risk for a given hazardous waste site.
The final risk assessment should Include a summary of the risks associated
with a site and such factors as the weight of evidence associated with each
step of the process, the estimated uncertainty of the component parts, the
distribution of risk across various sectors of the population and the
assumptions contained within the estimates.
8.2 Risk Characterization at Hazardous Waste Sites
This section discusses the characterization of risk at hazardous waste sites
in greater detail. Two discrete steps are required to fully characterize
potential risks from exposure to toxic chemicals at hazardous waste sites:
1. Characterize noncarclnogenic risks.
2. Characterize carcinogenic risks.
Typically, human populations are exposed to a mixture of chemicals at a
hazardous waste site rather than a single chemical. This phenomenon occurs
when.a series of unrelated chemicals are placed in the same area for disposa
or storage, eventually come in contact with each other, and are released as &
mixture to the environment. The EPA has published "Proposed Guidelines ior
the Health Risk Assessment of Chemical Mixtures" (USEPA 1985) which provices
guidelines for assessing the effects of multiple toxicant or multiple
carcinogen exposures. The following discussion of risk characterization at
hazardous waste sites is consistent with these proposed guidelines.
8.2.1 Characterization of Noncarcinogenic Risks
Risk assessment for single noncarcinogenic compounds at a site generally
results in the derivation of an exposure level that is considered acceptable
or is not expected to cause adverse effects. This exposure level may be
expressed in a variety of ways such as the Acceptable Daily Intake (ADI)» a"t
Ambient Air Standard, or Water Quality Criteria. The term "acceptable level
(AL) will be used here to Indicate any derived criteria, health standards or
advisories.
Characterizing risks from noncarcinogenic compounds involves comparing the
expected exposure level (E) to the AL. The resultant ratio (referred to as
the hazard index, HI ¦ E/AL) is a numerical indicator of the transition
between acceptable and unacceptable exposure levels. When making this
comparison it is important to ensure that the units for the exposure level an
acceptable level are the same. It may be necessary to apply a scaling factor
or exposure coefficient to the estimated exposure level, as appropriate, to
8-4
-------
standardize the units in the ratio. When HI > 1 there is a potential health
risk to the human populations associated with exposure to that chemical.
Therefore, the HI of a mixture nay be defined as:
HI - E1/AL1 + E2/AL2 + ... + E1/AL1
where:
E. - exposure level to the i**1 toxic che^Lcal
ALi - maximum acceptable level for the i toxic chemical
The EPA guidelines assume dose additivlty for exposure to multiple
noncarcinogens. This assumption of additivity is most properly applied to
chemicals that induce the same effect by the same mechanism. Thus, it may be
desirable to group the chemicals by type of adverse effect and derive a
separate HI for each chemical group to avoid overestimating risk. As with
single chemicals, when HI approaches unity, the potential for risks increases.
8.2.2 Characterization of Carcinogenic Risks
Carcinogenic risks are estimated as the probability or range of probabilities
that a specific adverse effect will occur under the conditions of exposure of
the human population at risk. One way of expressing this numerical estimate
of risk is as a carcinogenic potency factor or "unit cancer risk" which is
defined as the excess risk due to a continuous lifetime exposure to one unit
of carcinogen concentration. This carcinogenic potency factor can be used to
convert estimated intake directly to incremental risk. This relationship may
be defined as:
? - dB
where:
? ¦ incremental risk
d « exposure level (mg/kg/day)
B » carcinogenic potency factor (mg/kg/day )
For multiple compounds, the above equation may be generalized co:
Total incremental risk at the site is then equal to + ... + Pj
This equation is based on the following assumptions:
• Individual intakes are small.
• Actions by several carcinogens in mixture are independent (i.e., not
synergistic or antagonistic).
• Route-specific cancer risks are additive.
• Total risk for the site is additive.
8-5
-------
ERRATA (PAGES 8-4, 8-5)
Page 8-4, Section 8.2.1, Paragraph 2
Change the first sentence to read: "One method that toxicologists use to
characterize noncarcinogenic risks involves comparing the expected exposure
levels (E) to the "acceptable level" (AL) (USEPA 1985)."
Page 8-5, Paragraph 2
Add this paragraph immediately before Section 8.2.2: "This method of character-
izing noncarcinogenic risks has not been adopted as official EPA policy. This
is an oversimplification of the procedures used to estimate noncarcinogenic
risks associated vith exposure to multiple chemicals. There are many assump-
tions made in developing these calculation procedures which the user should be
aware of. These assumptions are detailed in the proposed mixture guidelines
(USEPA 1985)."
-------
Therefore, characterizing the potential risks at a site from carcinogenic
chemicals involves determining the incremental risk associated with exposure
to each potential carcinogen at the site and summing the incremental risks to
determine the total risk at the site.
For a detailed discussion of the uncertainties, assumptions and limitations
associated with risk assessments of chemical mixtures refer to the proposed
EPA guidelines (USEPA 1985).
8.3 Application of Risk Assessment at Hazardous Waste Sites
Risk assessments at hazardous waste sites may be performed for many reasons.
Examples of some different applications of risk assessment are briefly
described below. The Hazard Ranking System, which is described in Appendix A
of the National Contingency Plan, is a risk assessment approach to screen
hazardous waste 3ites for inclusion on the National Priorities List.
The Public Health Assessment component of the Remedial Investigation and
Feasibility Studies at Superfund sites prescribes an initial risk assessment
methodology for selecting Indicator chemicals at a site. Chemicals are
screened and assigned an "indicator score" based on concentrations present at
the site and their toxicity. Chemicals with the top scores are designated as
"indicator chemicals" for all analyses at the site. This allows subsequent
analyses at the site to focus on five to ten "indicator chemicals".rather than
potentially thousands oŁ individual chemicals that may be present at a site.
The goal of the Feasibility Study at a Superfund site is to develop and
evaluate alternative remedial actions. This evaluation involves an analysis
of the effectiveness (i.e., risk assessment)-©f each proposed alternative at
reducing potential public health risks at the site.
Finally, any time EPA initiates an enforcement action against a responsible
party at a hazardous waste site, it must support and justify its actions with
an endangerment assessment. The endangerment assessment, which is a type of
risk assessment, evaluates the magnitude and probability of actual or
potential harm to public health and welfare or the environment from the actual
or threatened release of hazardous substances from a site. A more detailed
discussion of the endangerment assessment process can be found in the
Endangerment Assessment Handbook (ICA1R 1985).
8.A Key Guidance and Implementation Documents
ICAIR, Life Systems, Inc. 1985. The endangerment assessment handbook.
Draft. Washington, DC: U.S. Environmental Protection Agency, Office of Waste
Programs Enforcement. Contract No. 68-01-7037.
ICF Incorporated. 1985. Superfund public health evaluation process
procedures manual. Draft. Washington, DC: U.S. Environmental Protection
Agency, Office of Emergency and Remedial Response. Contract No. 68-01-6872.
8-6
-------
Schultz HL, Palmer WA, Dixon GH et al. Versar Inc. 1984. Superfund exposure
assessment manual. Final Draft. Washington, DC: U.S. Environmental
Protection Agency, Office of Toxic Substances, Office of Solid Waste and
Emergency Response. Contract Nos. 68-01-6271 and 68-03-3149.
USEPA. 1985. U.S. Environmental Protection Agency. Office of Waste Programs
Enforcement. Draft endangerment assessment guidance. Memorandum from Jack W.
HcGraw. Washington, DC: U.S. Environmental Protection Agency. January 16,
1985.
USEPA. 1984. U.S. Environmental Protection Agency. Office of Pesticides and
Toxic Substances. Chemical activities status report, fourth edition, volume
2. Washington, DC: U.S. Environmental. Protection Agency. EPA
560/TIIS-84-0016.
USEPA. 1982. U.S. Environmental Protection Agency. Office of Pesticides.and
Toxic Substances. Graphical exposure modeling system (GEMS) user's guide.
Draft. Washington, DC: U.S. Environmental Protection Agency.
Whitmore RW. Research Triangle Institute. Methodology for characterization
of uncertainty in exposure assessments. Washington, DC: U.S. Environmental
Protection Agency, Office of Health and Environmental Assessment, Exposure
Assessment Group. Contract No. 68-01-6826.
8.5 Background References
ICF Incorporated. 1983. Scientific support document: The scientific basis
for the risk evaluation process. Draft. Washington, DC: U.S. Environmental
Protection Agency, Office of Emergency and Remedial Response.
ICF Incorporated. 1983. Superfund feasibility study guidance, chapter 4,
risk evaluation. Washington, DC: U.S. Environmental Protection Agency,
Office of Emergency and Remedial Response.
Morgan RC, Clemens R, Davis 3D, et al. (No date). Endangerment assessments
for superfund enforcement actions. Washington, DC: U.S. Environmental
Protection Agency, Office of Waste Programs Enforcement.
Schaum J. 1984. Short course on integration of exposure and risk assessment.
Part 3. Exposure assessment methods. Paper presented at the Annual Meeting
of Society for Environmental Toxicology and Chemistry, Arlington, VA.
USEPA. 1985. U.S. Environmental Protection Agency. Environmental Criteria
and Assessment Office. Proposed guidelines for the health risk assessment of
chemical mixtures. Fed. Regist., Jan. 9, 1985, 50 1170.
USEPA. 1984. U.S. Environmental Protection Agency. Risk assessment and
management: framework for decision making. Washington, DC: U.S.
Environmental Protection Agency.
USEPA. 1984. U.S. Environmental Protection Agency. Office of Health and
Environmental Assessment. Proposed guidelines for carcinogen risk assessment.
Fed. Regist., Nov. 23, 1984, 49 46294.
8-7
-------
USEPA. 1984. U.S. Environmental Protection Agency. Office of Health and
Environmental Assessment. Proposed guidelines for exposure assessment. Fed.
Regist., Nov. 23, 1984, 49 46304.
USEPA. 1984. U.S. Environmental Protection Agency. Office of Health and
Environmental Assessment. Proposed guidelines for mutagenicity risk
assessment. Fed. Regist., Nov. 23, 1984, 49 46314.
USEPA. 1984. U.S. Environmental Protection Agency. Office of Health and
Environmental Assessment. Proposed guidelines for the health assessment of
suspect developmental toxicants. Fed. Regist., Nov. 23, 1984, 49 46324.
8-8
-------
9.0
SUMMARY OF TOXICOLOGICAL INFORMATION ON DIOXINS AND FURANS
This chapter provides a summary of current toxicological Information on
dloxlns and furans. This class of compounds was chosen for consideration here
because of their high inherent toxicity and because they are often found as
contaminants at hazardous waste sites. This chapter is similar to the sort of
summary which might be prepared by a toxicologlst for a hazard assessment, and
therefore it is somewhat more technical and detailed than chapters 1 to 8 of
this handbook. A nontoxicologist would not be expected to prepare a summary
of this sort, but should, on the basis of the information presented in
chapters 1 to 8 of the handbook, be able to understand and apply this
information.
9.1 Chemical Properties and Environmental Stability
Dioxins are a class of compounds characterized by a structure comprising two
benzene rings interconnected to each other through a pair of oxygen atoms
(dibenzo-p-dioxin). Individual dloxlns differ from each other with-respect to
substituents present on the benzene rings. Various chlorine-substituted
dioxins have been found to be highly toxic, the degree of toxicity depending
on the substitution pattern. Especially toxic is 2,3,7,8-tetrachloro-
dibenzo-p-dioxin (TCDD).
Dioxins are stable and highly persistent in the environment. Under laboratory
conditions, soil samples incubated with 1, 10 or 100 ppm of dioxin contained
75% to 85% of the original material up to 160 days after treatment (Kearney et
al. 1972). In another study (Young 1974), soil samples from areas sprayed
10 years earlier contained dioxin levels of 10, 11, 30 and 710 ppt (parts per
trillion).
Some studies indicate that dioxins are subject to p'notodecomposition or
biodegradation. Crosby and Wong (1977) applied herbicide formulations
containing chlorodioxins to glass petri dishes, rubber plant leaves or co loam
soil. About 60% to 70% of the compound vas lost in approximately six hours.
Esposito et al. (1980) reviewed studies on the bioaegradabiiity of dioxins ir.
aqueous and soil environments. However, due to problems of extracting dloxlns
from test soils, it was concluded that such transformation has not been
conclusively established. Chlorodioxins (e.g., TCDD, 2,7-DCDD and DCDD)
undergo photodecomposition in nonpolar solvents, but photodecomposltion is
very slow in aqueous solutions or in wet or dry soils (Crosby et al. 19-71).
Dioxins are produced as unwanted contaminants in the industrial manufacture
of chlorophenols and their derivatives. The primary sources of dioxins in the
environment are the manufacturing and disposal sites of these industries.
Important exposure routes for dioxins include dermal contact or ingestion of
contaminated water or soil. Dioxins are lipophilic and tend to bioaccumulate
in fatty tissue, so ingestion of fish (especially bottom-feeders) from
contaminated waters is a significant human exposure route.
Furans are class of polyhalogenated aromatic compounds which are similar to
dioxins. For example, 2,3,7,8-tetrachlorodibenzo—furan (TCDF) is similar in
struecurs and toxicity to the chlorinated dioxin (TCDD) isomer. Very few
9-1
-------
pertinent studies on polychlorinated dibenzofurans (PCDFs), including TCDF,
have been located. Summaries of the available studies are included in the
appropriate sections of this chapter, but several sections (i.e., Mutagenic
and Carcinogenic Studies, Quantitative Indices of Toxicity) have no specific
references to PCDFs.
9.2 Summary of Health Effects Data
Exposure to TCDD causes thymic atrophy, decreased body weight, liver damage,
skin lesions (chloracne), renal function impairment, hematologic effects,
adrenal atrophy, reproductive system damage, immunosuppression, fetotoxic,
teratogenic and mutagenic effects and cancer. The chemical TCDD is an
extremely toxic substance, with an LD^q value reported as low as 0.6 vg/kg
following oral administration.
The effects of greatest concern associated with exposure to TCDD are liver
damage, thymic atrophy, fetotoxic and teratogenic effects and carcinogenic
potential.
Reported manifestations of TCDF toxicity are similar, including general edema
(accumulation of fluid), liver effects, skin lesions (chloracne), thymic
atrophy, immunosuppression and lethality.
9.2.1 Noncarcinogenlc Studies
9.2.1.1 Acute Effects
Various chlorodioxins have produced lethality following exposures for short
durations via oral administration in the rat, mouse, guinea pig» rabbit,
monkey and dog (Schwetz et al. 1973, KcConnell et al. 1978a, 1978b).
Estimates of LD_n values for guinea pigs and mice exposed to various
chlorodioxins range from 283.7 vg/kg to 5,000 vg/kg. These values are
summarized in Table 9-1. The most toxic chlorodioxin is TCDD, which has been
shown Co Induce lethality in a. variety of animals (rat, guinea pi?, mouse,
monkey, rabbit) at dose levels between 0.6 vg/kg and 190 ug/kg following oral
administration. The dermal value was 270 vg/kg. Estimates of LD^Q
values for this compound through various routes of exposure are summarized in
Table 9-2.
A summary of studies providing data on the sub—lethal effects of acute
exposure to TCDD is presented in Table 9-3. The effects were reported to
occur following single exposures ranging from 0.1 to 300 vg/kg in four animal
species (rat, guinea pig, chicken, mouse). Liver damage is the most
consistently reported effect in most species. Rats receiving a single dose of
100 pg/kg of TCDD showed severe liver damage, thymic atrophy and jaundice
(Gupta et al. 1973). In the same study, thymic and liver damage of lesser
severity occurred at lower dose levels (25 and 50 Vg/kg). In another study
(Greig et al. 1973), rats exposed to TCDD (300 Vg/kg) exhibited jaundice,
multinucleated parenchymal cells of the liver and gastric hemorrhage.
Histopathologic liver changes were observed five weeks after single oral doses
9-2
-------
TABLE 9-1 ACUTE TOXICITIES OF VARIOUS CHLORODIOXINS BY ORAL EXPOSURE
Chlorine —50' yB/k8
Positions Guinea Pig House
2,8 300,000
2,3,7 29,400 3,000
2,3,7,8 0.6-2.0 284
1,2,3,7,8 3.1 337
1,2,4,7,8 1,125 5.000
1.2.3.4.7.8 72.5 825
1,2,3,6,7,"8 70-100 1,250
1.2.3.7.8.9 60-100 1,440
Adapted from McConnell et al. (1978b).
9-3
-------
TABLE 9-2 ACUTE
LETHALITY OF TCDD
Route
Species
Sex
of Exposure
LD.q, yg/kg
Reference
Rat
M
Oral
22
Schwetz et al. (1-973)
Rat
M,F
Oral
100
Harris et al. (1973)
Rac
F
Oral
190
Greig et al. (1973)
Guinea Pig
M
Oral
0.6
Schwetz et al. (1973)
Rabbit
M.F
Oral
115
Schwetz et al. (1973)
M.F
Dermal
272
M.F
Intra-
252
peritoneal
Mouse
M
Oral
114
Vos et al. (1974)
Mouse
M
Oral
283.7
McConnell et al. (1978b)
Monkey
F
Oral
<70
McConnell et al. (1978a)
Adapted from Esposito et al. (1980) and NT? (1982a,b).
9-4
-------
TABLE 9-3* EFFECTS 01 TCDD IN ANIMALS FOLLOWING ACUTE EXPOSURE
Species
Rac
Guinea Pig
a
Rat
Chicken
Rat
Rat
Rat
Mouse
Rat
Rat
Dose (Ug/kg)
25, 50 or 100
100
3.0
300
25 - 50
10
0.1
50
50
10
10, 25, or 50
Route
Effects
Reference
Oral
Intra-
peritoneal
Oral
Oral
I.lver damage, thymic atrophy
Jaundice, 43Z mortality
Hemorrhage, adrenal atrophy,
cellular depletion of lymphoid
organs, 90% mortality
Weight loss, gastric hemorrhage,
liver damage (cellular changes),
jaundice
Pericardial edema
Hematologic effects
li. creased liver weights
(1972)
Liver damage
Liver damage
Decreased renal function
Decreased renal function
Gupta et al. (1973)
Gupta et al. (1973)
Greig et al. (1973)
Greig et al. (1973)
Welssburg and Zlnkl (1973)
Cunningham and Williams
Harris et al. (1973)
Harris et al. (1973)
Anaizi and Cohen (1978)
Hook et al. (1978)
A.iapted from NAS (1977), NTP (1982) and Esposico et al. (1980).
-------
of TCDD as low as 50 ug/kg were administered to male and female CD rats,
and one week after a single dose of 50 Ug/kg was administered to female
CD-I mice (Harris et al. 1973). Increased liver weights were found in male
Wistar rats seven days after single intraperitoneal doses of 0.1 lig/kg
(Cunningham and Williams 1972).
Acute exposure to TCDF appears to cause effects similar to, but less severe
than, the effects caused by TCDD. A summary of the effects of acute exposure
to TCDF is presented in Table 9-4.
Acute toxicity studies show that TCDF is toxic in chickens and guinea pigs.
McKinney et al. (1976) reported that Leghorn chickens given 5 Ug/kg TCDF
showed fluid accumulation, enlarged hearts and reduced size of thymuses and
spleens. The compound also caused lethality.
Harltev guinea pigs administered TCDF showed oral LD^q values of 5 to
10 ug/kg,. with a mean time to death of 11 days. As a comparison, TCDD was
lethal to 50% of the test guinea pigs following administration of 2.0 ug/kg.
Death occurred within 20.6 days (Moore et al. 1976).
Mice and rats were found to be resistant to TCDF (Moore et al. 1976). Rats
administered TCDF at a dose of 1,000 Ug/kg failed to show any toxic effects.
Male mice administered a single oral dose of TCDF up to 6,000 ug/kg did not
exhibit adverse effects, except for mild effects on the liver. Mice
administered subcutaneous injections of TCDF (6,000 ug/kg) showed an increase
in liver weights as well as liver-to-body-weight ratio. In contrast, TCDD
caused adverse effects at levels of 0.2 ug/kg in mice (Kociba et al. 1976) and
0.1 ug/kg in rats (Vos et al. 1974).
9.2.1.2 Subchronic and Chronic Effects
Longer exposures to TCDD reportedly cause effects similar to those following
acute exposure including thymic atrophy, liver damage, renal function
impairment, hematological effects, hormonal alterations, isnunssuppression,
nervousness and irritability. A summary of major studies providing
dose-response effects is presented in Table 9-5. Those studies providing
dose-response data indicating the greatest sensitivity to TCDD are described
below.
Doses as low as 0.1 ug/kg/day caused a slight degree of liver degeneration in
rats in a subchronic 13-week (5 doses per week) study (Kociba et al. 1976).
Dose levels of 1.0 ug/kg/day increased levels of serum bilirubin and alkaline
phosphatase and caused pathologic changes in the livers of rats. A NOAEL of
0.01 ug/kg TCDD was reported for rats.
(a) A series of letters, numbers or a combination of both indicates a specific
strain (i.e., a race or stock) of animals that all have common hereditary
characteristics.
9-6
-------
TABLE 9-4 EFFECTS OF TCOF IN ANIMALS FOLLOWING ACUTE EXPOSURE
Species Dose (Mg/kg) Route
Chicken 5, 25 Oral
Effects
Reference
Guinea Pig
Rat
Moase
Moiise
5 to 10
1,000
6,000
6,000
Oral
Oral
Oral
Subcutaneous
Enlarged heart, reduced size of
thymus, spleen
I.D50
No toxic effects
Mild liver toxicity
Increased liver weight and liver-
to-body-weight-ratlo
McKlnney et a 1. (1976)
Moore et al. (1976)
Moore et al. (1976)
Moore et al. (1976)
Moore et al. (1976)
Adapted from Esposito et al.. (1980), NTP (1982).
-------
TABLE 9-5 NOAEL AN1) LOAEL VALUES OBTAINED FROM SUBCMRONIC AND CHRONIC ORAL
TOXICITY STUDIES OF 2,3.7,8-TCDD
Species
Rat
Rac
Rac
RaC
Mouse
Monkey
RaC
Rut
Mcuse
Duration
of Exposure
13 weeks
13 weeks
16 weeks
28 weeks
13 weeks
36 weeks
104 weeks
104 weeks
104 weeks
Endpolnts
Decreased body weight,
liver pathology
Toxic hepatitis
Elevated porphyrin
levels
Fatty changes in the
liver, decreased body
weight
Toxic hepaticLs
Pancytopenia
Degenerative and necrotic
changes in the liver
Toxic hepatitis
Dermatitis and amyloidosis
' NOAEL LOAEL
Mg/kg/day pg/kg/day
Reference
0.01
ND
(a)
ND
ND
0.001
0.0014
ND
0.1
Kociba et al. (1976)
0.07 0.14 NTP (1980a)
0.0014 0.014 Goldstein et al. (1982)
0.014 King and Roesler (1974)
0.014 NTP (1980a)
2 Allen et al. (1977)
0.01 Kociba et al. (1978, 1979)
0.007 NTP (1980a)
0.001 NTP (1980a)
(ti) Not determined.
-------
Increased mortality was observed in female Sprague-Dawley rats maintained for
2 years on a diet that provided a TCDD dose of 0.1 ug/kg/day, while no
increased mortality vas observed in male rats at this dose or in animals
receiving doses of 0.01 or 0.001 ug/kg/day (Kociba et al. 1978, 1979). At
termination of the study, gross and histologic examination indicated that the
liver vas the most severely affected organ, with degenerative, necrotic and
inflammatory changes observed. Increases in urinary excretion rates of the
metabolites, coproporphyrin and uroporphyrin, in the high and middle dose
females vere consistent vith the observed liver damage. Primary liver injury
was dose-related with the lowest dose representing a NOAEL.
When TCDD was administered by gavage (by stomach tube) in com oil-acetone
(9:1) at dose levels of 0.0, 0.5, 0.05 or 0.01 ug/kg/week, toxic hepatitis was
observed in male Osborne-Mendel rats at incidences of none out of 74 tested
(0/74), 14/50, 0/50 and 1/50 and in female rats at incidences of 0/75, 32/49,
1/50 and 0/50 (NTP 1980a"). No other non-neoplastic lesions were observed,
even though extensive histologic examinations were performed. The two
preceding studies support a NOAEL for rats of =0.001 yg/kg/day, with a LOAEL
of 0.05 yg/kg/day, and a frank effect level (FEL) for liver injury and
possibly decreased survival of 0.5 Ug/kg/day.
Non-neoplastic effects of chronic exposure to TCDD in mice were described in
studies investigating the carcinogenic potential of TCDD. In a National
Toxicology Program (N^) (1980a) bioassay, histologic examinations were
performed on B6C3F1 mice treated biweekly with TCDD by gavage in corn
oil-acetone (9:1) for 104 weeks followed by an additional 3-week observation
period. The doses for male animals were 0.0, 0.01, 0.05 and 0.5 yg/kg/week,
and for female animals, 0.0, 0.04, 0.2 and 2.0 ug/kg/week. The only
non-neoplastic adverse effect was toxic hepatitis, which occurred in males at
incidences of 0/~3, 3/49, 3/49 and 44/50, and in females at incidences of
0/73, 1/5G, 2/46 and 34/47, respectively, in the control, low, medium and high
dose groups. In another study, weekly dosing by stomach tube of TCDD at doses
of 0.0, 0.007, 0.7 or 7.0 ug/kg/week for one year resulted in amyloidosis
(deposition of amyloid) of the kidney, spleen and liver, and dermatitis at the
•time of death in sale Swiss sice (Tcth et al. 1978, 1979). The incidences of
these effects in the control, low, medium and high dose groups, respectively,
was 0/38, 5/44, 10/44 and 17/43. In the high dose group, the amyloidosis was
extensive and considered to be the cause of early mortality. Severe toxic
effects were observed at doses of 1 ug/kg/day (early mortality) and 0.28 to
0.07 ug/kg/day (toxic hepatitis), while a LOAEL for dermatitis and amyloidosis
of 0.001 ug/kg/day was reported.
Several epidemiologic studies and case reports involving dioxln exposure In
human subjects have been reported (Esposito et al. 1980). Effects observed
include skin lesions (chloracne, prophyria cutanea tarda), liver function
impairment and neurological disorders (polyneuropathy, peripheral nerve
damage). An International Agency for Research on Cancer (IARC 1982)
evaluation of human exposure data concluded that these studies are inadequate
since they involve multiple chemical exposures.
9-9
-------
The chemical TCDD has been reported to be fetotoxic and teratogenic when
administered alone or in combination with other chemicals. Various studies
have been identified in the available literature based on TCDD exposure alone.
Effects observed were kidney anomalies, intestinal hemorrhage, general edema,
cleft palate and fetal death. Adverse effects on reproduction were also
reported.
Intestinal hemorrhage, general edema and a reduction in fetal weights were
reported in rats following the administration of 0.125 yg/kg/day in studies by
Sparschu et al. (1971). In the same studies, the number of fetuses was
reduced and fetal death increased at 0.5 Mg/kg/day. No structural
malformations were reported at 0.03 ug/kg/day. Courtney and Moore (1971)
reported cleft palate and kidney abnormalities in offspring from mice
administered TCDD at doses of 1.0 ug/kg or 3.0 Ug/kg. Similarly, kidney
malformations were reported by the same authors in offspring from rats which
received subcutaneous injections of 0.5 yg/kg/day on day 9, 10, or 13 and 1*»
of gestation.
Murray et al. (1979) completed a three-generation reproduction study using
Sprague-Davley-rats fed TCDD continuously in the diet (at levels of 0, 0.001,
0.01, and 0.1 yg/kg/day). Significant decreases were observed in fertility,
litter size, gestation survival, postnatal survival, and postnatal Body weight
for the 0.01 and 0.1 yg/kg groups. No apparent adverse effect on reproduction
was seen at the 0.001 Ug/kg dose level.
Nisbet and Paxton (1982) reevaluated the'data of Murray et al. (19.79) using
different statistical methods. From this reevaluation it was concluded chat
TCDD significantly reduced the gastational index, decreased fetal weight, and
increased liver-to-body weight ratios and the incidence of dilated renal
pelvis i.i both lower dose groups. Nisbet and Paxton (1982) concluded that the
dose of j.CCI iJg/kg/iay was not a NCAEL but a L0AEL in this stucy.
Luster et al. (1980) examined bone marrow, immunologic parameters, and host
susceptibility in B6C3FlTa} mice following pre- and postnatal exposure to
TCDD. •Doses cf 0, 1.0, 5.0 or 15.0 yg/kg TCDD/body weight were given to dams
on day 14 of gestation and to offspring on days 1, 7 and 14 following birth.
Neonatal body, liver, spleen, and thymus weights were decreased and bone
marrow toxicity occurred in the 5.0 and 15.0 ug/kg groups. Red blood cell
counts, hematocrits, and hemoglobin were decreased at the highest dose tested.
Oishi et al. (1978) studied the subchronic toxicity of polychlorinated
dibenzofurans (PCDFs) in rats. Test animals were fed diets -containing 1 or
10 ppm PCDF for four weeks. The PCDF markedly depressed normal body weight
gain. In rats fed diets containing 10 ppm of PCDF, significantly decreased
thymus, ventral prostate, and s«minal vesicle weights were found, and the
animals developed chloracne-like lesions on the ears within three weeks.
Hemoglobin and hematocrit values (percentage of the volume of a blood sample
occupied by cells) were decreased in rats fed either diet.
9.2.2 Mutagenic and Carcinogenic Studies
®r.'jdies the mutagenicity of TCDD have produced conflicting results. The
chemicals TCDD reportedly produces mutagenic effects in various bacterial
9-10
-------
systems. However, results were negative in tests employing other Indicator
test systems, including cytogenetic (chromosome analysis) tests and dominant
lethal assays. Hussain et al. (1972) reported that TCDD (2 ug/mL) increased
the Incidence of reverse mutations in Escherichia coli. Similarly, TCDD (dose
not specified) was reported to be mutagenic.without metabolic activation in
Salmonella typhimurium test strain TA 1532. a Green et al. (1977) gave 0.25,
0.5, 1.0, 2.0, or 6.0 ug/kg TCDD (dissolved in 1 part acetone: 9 parts corn
oil) by gavage to male and female Osborne-Mendel rats twice weekly for 13
weeks and observed an increased incidence of chromosomal breaks in female rats
dosed with 4 ug/kg and in males dosed with 2 Ug/kg or 4 ug/kg.
Mutagenic effects (with or without metabolic activation) were not detected
when C.eiger and Neal (1981) examined the mutagenicity of TCDD (up to
20 ug/plate) using the S. typhimurium test strains TA1535, TA100, TA1538,
TA98, and TA1537.
The carcinogenic potential of TCDD has been studied extensively. A summary of
the results of selected comprehensive studies is presented in Table 9-6. The
results of these studies show that TCDD-exposed animals exhibited malignant
lesions involving multiple organ systems including accessory digestive organs
(liver), endocrine (thyroid, adrenal), renal, reproductive (testes), and nasal
structures. Representative studies are described below.
Croups of ten male Sprague-Dawley rats were fed a diet containing TCDD for 78
weeks at dose levels ranging from 1 ppt to 500 ppt or 1 ppb to 1,000 ppb (Van
Miller et al. (1977). These dose levels represent approximate weekly dose
levels of 0.0003 to 0.1 Ug/kg or 0.4 to 500 ug/kg. Animals exposed at 5 ppt,
50 ppt, 500 ppt or 5 ppb showed an overall incidence of neoplasm of 38Z
(23/60). No neoplasms were reported or observed following exposure to I ppt
TCDD. In the 5 ppt group, 5/10 animals had six neoplasms (earduct carcinoma,
lymphocytic leukemia, adenocarcinoma, malignant histiocytoma iwitn meta-
stases) , angiosarcoma and Leydig-cell adenoma). Neoplasm were also observed
in the following groups: at 50 ppt, three in 3/10; at 500 ppt, four in 4/10;
at I ppb, five in 4/10; at 5 ppb, ten in 7/10. Neoplasms were not observed in
the controls. Rats administered TCDD at 50, 500 or 1,000 ppb exhibited 100%
mortality by the fourth week.
In another stujiy (Kociba et al. 1978), groups of 100 Sprague-Dawley rats (50
males and 50 females) received diets containing TCDD at 0, 22, 210, or
2,220 ppt (equivalent to 0.0, 0.001, 0.01 and 0.1 ug TCDD/kg/day) for two
years . Administration of 0.01 ug/kg/day increased the incidence of
hepatocellular hyperplastic nodules (female: 18/50 versus 8/86 controls) and
focal alveolar hyperplasia in the lungs (P<0.05). Dietary intake of
0.1 ug/kg/day increased the incidence of hepatocellular carcinomas (female:
11/49 versus 1/86) and squamous cell carcinomas of the lung (female: 7/49
versus 0/86), hard palate/nasal turbinates (male: 4/50 versus 0/85; female:
4/49 versus 0/86), and tongue (male: 3/50 versus 0/85) (P<0.05). Also
increased in frequency by the 0.1 ug TCDD/kg/day were adenoma of the adrenal
cortex (male) and hepatocellular hyperplastic nodules (female).
9-11
-------
TABLE 9-6 SUMMARY OF CARCINOCENIC EFFECTS OF TCDD
Species/Sex
(Number)
Ral:/
M (50)
!• (50)
Dose
Rat/M (10) 1 ppt
Rat/M (10) 5-500 ppt
Ral /M (10) 1-5 ppb
0.001 Mg/kg
0.01 pg/kg
0.1 Pg/kg
Duration Route
78 wks
2 years
Diet
Diet
Effects
Reference
No neoplasm.
Van Miller et al,
(1977)
Ear duct carcinoma, benign tumor
of the kidney and testes,
lymphocytic leukemia, skin
carcinomas and benign muscle
tumors.
Cholangiocarclnoma of liver,
squamous cell tumor of lung,
angiosarcoma in skin, glioblas-
toma in brain, malignant histio-
cytomas in peritoneum.
No significant increase in tumors. Kociba et al. (1978)
Liver cancer.
Liver cancer, squamous cell car-
cinoma of the lung, hard palate/
nasal turbinates, or tongue
(P=0.05).
Moiise/F (30) 0.015 pg/kg/wk 99-104 wks Dermal Fibrosarcoma in Integumentary
system (8/27, P=0.007).
Rat/M (50) 0.5 pg/kg/wk 104 wka
Ral/F (50) 0.5 pg/kg/wk 104 wka
Cavage Follicular cell adenomas of
thyroid (10/50, P=0.00l).
Cavaj>e Neoplastic nodules .of the liver
(J 2/49, P=0.006).
NTP (1982b)
NTP (1982a)
continued-
-------
Tal le 9-6 - continued
Sp*.cies/Sex
(Humber) Dose Durat ton Routo
Mo».se/M&F 2.0 yg/kg/wk 104 wksi Gavage
Moose/F 2.0 pg/kg/wk 104 wks Gavage
Adapted from Esposito et al. (1980), NTP (1982a,b).
Effects
Reference
Hepatocellular carcinoma NTP (1982a)
(17/50, P=0.002 in M);
(6/47, P=0.14 in F).
Follicular cell adenomas of the
thyroid (5/46, P=0.009)
-------
The NTP has performed a chronic bioassay in both Osborne-Mendel rats and
B6C3F1 mice to determine the carcinogenicity of a mixture containing 31%
1,2,3,6,7,8-HCDD and 67Z 1,2,3.7,8,9-HCDD (NTP 1980). Other dioxins,
including di-, tri-, tetra-, pentachlorodibenzo-p-dioxin, and
bromopentachlorodibenzo-p-dioxin, were present at less than 0.09Z. The
mixture was administered to the test animals in corn oil-acetone (9:1).by
gavage two times/week. The male and female rats, and the male mice received
doses of 0.0; 1.25, 2.5 or 5 yg/kg/week, and the female mice received doses of
0.0, 2.5, 5.0 or 10 ug/kg/week. Treatment was continued for 104 weeks
followed by a 3-4 week observation period. In both test species, exposure to
HCDD produced a dose-related "toxic hepatitis" and an increased incidence of
hepatocellular nodules or tumors (adenomas and carcinomas). Liver tumor
incidence was statistically significant in both male and female mice and in
female rats.
The NTS P.982a) conducted a study for 104 weeks using Osborne-Mendel rats and
B6C3F1 3 , mice. The rats and male mice were administered TCDD at 0, 0.01,
0.05 or 0.5 Ug/kg/wk by gavage in two divided doses, and the female mice were
given 0, 0.04, 0.2, or 2.0 yg/kg/wk. Incidences of follicular cell thyroid
adenomas in male rats (P<0.001) and of neoplastic nodules in livers of female
rats (P-0.006) increased significantly. TCDD increased the numbers of
hepatocellular carcinomas in male mice (P-0.002) and in females (P-0.014).
The total liver tumors (carcinomas and adenomas) were increased in males
(?<0.001) and females (P-0.002). In addition, female mice had increased
incidence of follicular cell thyroid adenomas. These studies indicate that
TCDD is an animal carcinogen.
Epidemiologic studies on industrial workers and herbicide applicators are
consistent with the conclusion from animal studies that TCDD is a carcinogen.
However, since TCDD is usually a contaminant of phenoxy acids and/or
chloroDhenols, human exposure is always to multiple chemicals. i.nerefore, the
evidence for human carcinogenicity from these studies is only suggestive due to
the difficulty of evaluating the risk of TCDD exposure in the presence of the
confounding effects of the ether chemicals (USEPA 1984).
9.3 Quantitative Indices of Toxicity
9.3.1 Noncarcinogenic Effects
Recommended exposure limits to TCDD to ensure human safety have been
established by many agencies. The National Academy of Sciences (NAS 1977),
before TCDD was considered to be a carcinogen, suggested an ADI of 0.0001 yg/
TCDD/kg/day based on a 13-week feeding study in rats (Kociba et al. 1976).
The reported NOEL in that study (0.01 yg/kg) was divided by an uncertainty
factor of 100 to determine the ADI. The NAS then calculated a suggested-
no-adverse-effect-level in drinking water of 0.0007 yg/L, based on the average
weight of a human adult (70 kg) and an average daily intake of water of 2 L,
with water representing 202 of total Intake.
The USEPA (1984) has calculated an ADI of 10 ^ yg/kg/day, using a LOAEL, based
on toxic effects and reduced fertility, of 0.001 yg/kg/day and employing an
ur.cartair.t" fzzzzz zz 1,00C. "oir.g ; bioac:::aulatior. factor of 5 000, and
9-14
-------
assuming_g daily consumption of 6.5 g of fish, a water concentration of
2.0 x 10 Ug/L was derived. It was noted that this value may not be
sufficiently low to protect against the carcinogenic effects of TCDD (USEPA
1984).
9.3.2 Carcinogenic Effects
Since there is no recognized safe concentration for a human carcinogen, and
TCDD is a suspected human carcinogen, the recommended concentration of TCDD in
water is zero (USEPA 1984). Assuming daily consumption of 2 L of water and
6.5 g of fish and shellfish, the concentration^ of TCD^ in water that
correspond to excess cancer risks of 10 , 10 or 10 were calculated to be
1.3 x 10 , 1.3 x 10 or 1.3 x 10 ug/L, respectively (USEPA 1984). Because
of the tendency for aquatic organisms to bioconcentrate TCDD, most of this
risk is due to consumption of the fish or shellfish. If no contaminated fish
or shellfish were eaten, the water-concentrations corresponding to.excess
canger risks of 10 , 10 and 10 ' would be 2.2 x 10 , 2.2 x 10 and 2.2 x
10 ug/L, respectively. These criteria are below the limit of detection of
TCDD in water (approximately 3 x 10 Ug/L) by current analytical methods.
The Food and Drug Administration (FDA) issued a health advisory stating that
fish with residues of TCDD >_50 ppt should not be consumed, but fish with
residues of < 25 ppt pose no serious health concern (USEPA 1984). The Centers
for Disease Control . (CDC) has established 1 ppb TCDD as a level of concern in
residential soils.
9.4 Special Concerns
The special concerns related to TCDD are its extreme toxicity and persistence
in the environment. It is possibly, on a molecular basis, the aost poisonous
synthetic chemical known (Esposico et al. i960). It is resistant to
biodegradation and has a high affinity to soil, especially soil with a
significant organic content.
The data from animal studies suggest :hat fetuses and newborns may be at
greater risk from TCDD exposure than the general population.
9.5 References
Allen JR, Barsotti DA, Van Miller JP, Abrahamson LJ, Lalich JJ. 1977.
Morphological changes in monkeys consuming a diet containing low levels of
TCDD. Food Cosmet. Toxicol. 15:401.
Anaizi NH, Cohen J. 1978. The effects of TCDD on the renal tubular secretion
of phenolsulfonphthalein. J. Pharmacol. Exp. Ther. 207(3):748-755.
Courtney KD, Moore JA. 1971. Teratology studies with 2,4,5-T and
2,3,7,8-TCDD. Toxicol. Appl. Pharmacol. 20:396-403.
Crosby DG, Wong AS, Plimmer JR, Woolson EA. 1971. Photodecomposition of
chlorinated dibenzo-p-dioxins. Science 173:748-749.
Crosby DG, Wong AS. 1977. Environmental degradation of
2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Science 195:1337-1338.
9-15
-------
Cunningham HM, Williams DT. 1972. Effect of 2,3,7,8-tetrachloro-
dibenzo-p-dioxin on growth rate and the synthesis of lipids and proteins in
rats. Bull. Environ. Contam. Toxicol. 7(1):45—51-
Esposito MP, Tieman TO, Dryden FE. 1980. Dioxins. Industrial Environmental
Research Laboratory, Office of Research and Development. EPA 600/2-80-197,
187-306.
Geiger LE, Neal RA. 1981. Mutagenicity testing of 2,3-7,8-tetrachloro-
dibenzo-p-dioxin in histidine auxotrophs of Salmonella typhimurium. Toxicol.
Appl. Pharmacol. 59(1):125-129.
Goldstein JA, Linko P, Bergman H. 1982. Induction of porphyria in the rat by
chronic versus acute exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin.
Biochem, Pharmacol. 31(8):1607-1613.
Green S, Moreland F, Shen C. 1977. Cytogenetic effects of 2,3,7,8-TCDD on
rat bone marrow cells. Food and Drug Administration Bylines 6:292-29*+.
Greig JB, Jones G, Butler WH, Barnes JM. 1973. Toxic effects of
2,3,7,8-TCDD. Food Cosmet. Toxicol. 11:585-595.
Gupta BN, Vos JG, Moore JA, Zinkl JG, Bullock BC. 1973. Pathologic effects
of 2,3,7 ,.8-tetrachlorodibenzo-p-dioxin in laboratory animals. Environ. Health
Perspect. 5:125-140.
Harris MW, Moore JA, Vos JG, Gupta-BN. 1973. General biological effects of
TCDD in laboratory animals. Environ. Health Perspect. 5:101-109.
Hcok JB, et al. 1978. Renal effects of 2,3,7,8-TCDD. Environ. Sci. Res.
12:381-388.
Hussain SL, Ehrenberg L, Lofroth G, Gejvall T. 1972. Mutagenic effects of
TC3D on bacterial systems. Asibio. 1:32-33.
IARC. 1982. International Agency for Research on Cancer. IARC monographs on
the evaluation of carcinogenic risk of chemicals to humans. Suppl 4. IARC
Lyon. France.
Kearney PC, Woolson EA, Ellington CP. 1972. Persistence and metabolism of
chlorodioxins in soils. Environmental Sci. Tech. 6(12):1017-1019.
King ME, Roesler AR. 1974. Subacute intubation study on rats with the
compound 2,3,7,8-tetrachlorodioxin. United States Environmental Protection
Agency. NTIS PB-257-677, p 27.
Kociba RJ, Keeler PA, Park CN, Gehring PJ. 1976. 2,3,7,8-tetrachloro-
dibenzo-p-dioxin (TCDD)-Results of a 13-week oral toxicity study in rats.
Toxicol. Appl. Pharmacol. 35:553-574.
Kociba RJ, Keyes DG, Beyer JE, et al. 1978. Results of a two-year chronic
;c::ici;y and oncogenicity study of 2,3,7,S-tstrachlorcdibenzo-p-dioxir. in
rats. Toxicol. Appl. Pharmacol. 46(2):279-303.
9-16
-------
Kociba RJ, Keyes DG, Beyer JE, Carreon RM, Gehring PJ. 1979. Long-term
toxicologic studies of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in
laboratory animals. Ann. NY Acad. Science. 320:397-404.
Luster MI, Boorman GA, Dean JH, et al. 1980. Examination of bone marrow,
immunologic parameters and host susceptibility following pre- and postnatal
exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Int. J.
Immunop'uarmacol. 2(4): 301-310,
McConnell EE, Moore JA, Dalgard DW. 1978a. Toxicity of 2,3,7,8-tetrachloro-
dibenzo-p-dioxin in rhesus monkeys (Macacas mulatta) following a single oral
dose. Toxicol. Appl. Pharmacol. 43:175-187.
McConnell EE, Moore JA, Haseman JK, Harris MW. 1978b. The comparative
toxicity of chlorinated dibenzo-p-dioxins in mice and guinea pigs. Toxicol.
Appl. Pharmacol. *4:335-356.
McKinney JD, Chae K, Gupta GN, Moore JA, Goldstein JA. 1976. Toxicological
assessment of hexachlorobiphenyl isomers and 2,3,7,8-tetrachlorobenzofuran in
chicks. I. Relationships of chemical parameters. Toxicol. Appl. Pharmacol.
36:65-80.
Moore JA, Gupta BN, Vos JG. 1976. Toxicity of 2,3,7,8-tetrachlorodibenzo-
furan - preliminary results. Proc. -Natl. Conf. PCB's, November, 77-80.
Murray FJ, et al. 1978. Three generation reproduction study of rats
ingesting TCDD. Toxicol. Appl. Pharmacol. 41:200-201.
Murray FJ, Smith FA, Nitschke KL, Humiston CG, Kociba RJ, Schvets BA. 1979.
Ihree-generacion reproduction study of rats given 2,3,7,3-tetrachlcro-
dibenzo-p-aioxin (TCDD) in the diet. Toxicol. Appl. Pharmacol. 50:241-251.
NAS. 1977. National Academy of Sciences. Drinking water and health,
vashir.gtor., DC: National Academy
-------
Oishi S, Morita M, Fukuda H. 1978. Comparative toxicity of polychlorinated
biphenyls and dibenzofurans in rats. Toxicol. Appl. Pharmacol. 43:13-22.
Schwetz BAf Norris JM, Sparschu GL, et al. 1973. Toxicology of chlorinated
dibenzo-p-dioxins. Environ. Health Perspect. 5:87-99.
Sparschu GL, Dunn FL, Howe VK. 1971. Study of the teratogenicity of
2,3,7,8-tetrachlorodibenzo-p-dioxin in the rat. Food Cosmet. Toxicol.
9:405-412.
Toth K, Somfai-Relle S, Sugar J, Bence J. 1979. Carcinogenicity testing of
herbicide 2,4,5-trichlorophenoxyethanol containing dioxin and of pure dioxin
in Swiss mice. Nature (Lond). 278(5704):548-549.
Toth K, Sugar J, Somfai-Relle S, Bence J. 1978. Carcinogenic bioassay of the
herbicide, 2,4,5-trichlorophenoxyethanol (TCPE) with different
2,3,7,8-cetrachlorodibenzo-p-dioxin (dioxin) content in Swiss mice. Prog.
Biocheo. Pharmacol. 14:82-93.
USEPA. 1984. United States Environmental Protection Agency. Ambient water
quality criteria for 2,3,7,8-tetrachlorodibenzo-p-dioxin. Washington, DC:
Office of Water Regulations and Standards. U.S. Environmental Protection
Agency. EPA 440/5-84-007.
Van Miller JP, Lalich JJ, Allen JR. 1977. Increased incidence of neoplasms
in rats exposed to low levels of 2,3,7,8-tetrachlorodiberizo-p-dioxin.
Chemosphere 6:537-544.
Vos JG, Mocre JA. Zinkl JG. 1974. Toxicity of 2,3,7,8-TCDD in C5781/6 mice.
Toxicol. Appl. "Pharmacol. 29:229-2^1.
Weissburg JB, Zinkl JG. 1973. Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin
upon hemostasis and hematologic function in the rat. c-nviron. Health
rerspect. 5:119-123.
Young AL. 1974. Ecological studies on a herbicide-equipment test area
(TA-C-52A). Elgin Air force Armament Laboratory. Technical Report
AF ATL-TR-74-12.
9-18
-------
10.0
SUMMARY OF TOXICOLOGICAL INFORMATION ON TRICHLOROETHYLENE
This chapter provides a summary of current toxlcologlcal Information on
trichloroethylene. This compound was chosen for consideration here because of
its high inherent toxicity and because it is often found as a contaminant at
hazardous waste sites. This chapter is similar to the sort of summary -which
might be prepared by a toxicologist for a hazard assessment, and therefore it
is somewhat more technical and detailed than chapters 1 to 8 of this handbook.
A nontoxicologist would not be expected to prepare a summary of this sort, but
should, on the basis of the information presented in chapters 1 to 8 of the
handbook, be able to understand and apply this information.
10.1 Chemical Properties and Environmental Stability
Trichloroethylene (TCE) is a nonflammable, colorless liquid used primarily as
•a solvent, in metal degreasing, in drycleaning operations and in refrigerants
and fumlgants. In the past, TCE has also been used as an anesthetic and as an
extractant in food processing. Trichloroethylene has a sweet odor, resembling
that of chloroform. Trichloroethylene is practically insoluble in water, but
highly soluble in lipids.
Although TCE can be formed during the water chlorlnation process, Che major
source of environmental levels of this chemical is volatilization of TCE
during manufacturing. Trichloroethylene is widely distributed in the environ-
ment; however, there is no indication that it is persistent or that it bioac-
cumulates in the food chain. Trichloroethylene has been detected in air,
food, human tissues and in groundwater. Concentrations in groundwaters range
from 18 to 22,000 ppb. An EPA finished water survey of ten cities reported
TCE in half the supplies tested (levels ranged from 0.1 to 0.5 ug/L (USZPa
1975).
10.2 Summary of Health Effects Data
Expcsur® to TCE can produce CNS depression (e.g., unconsciousness, numbness,
incoordination), rir.rr liver iansge, kidr.ey necrosis (death of xidnay cells;.,
painful breathing and cardiac arrhythmia. The most critical endpoint for
toxicological evaluation, however, is the carcinogenic potential of TCE.
The most important routes of exposure to TCE are inhalation and ingestion.
Trichloroethylene is readily absorbed both orally and in the lungs. In mice
and rats 95Z and 98Z, respectively, of an orally administered dose of TCE was
absorbed from the gastrointestinal tract within 72 hours (Dekart et al. 1984).
Human data concerning TCE absorption has been obtained using the inhalation
route of exposure. Stewart et al. (1962) reported TCE concentrations of
4.5 to 7 mg/L in blood within two hours of exposure to a time-weighted average
concentration of 1,420 mg/m . Estimates of retention in the lung have ranged
from 28Z to 24Z. Dermal absorption is also rapid, but In most cases the
opportunity for this route of exposure is insignificant.
The metabolism of TCE is an important factor in its toxicity., Trichloro-
ethylene is metabolized to active intermediates (e.g., 2,2,3-trichloroxirane,
dichloroacetic acid) that may responsible for some of the long-term effects
of TCE. Additionally, there are important differences in the metabolism among
10-1
-------
different species, making interspecies comparisons difficult. Trichloro-
ethylene does not appear to be teratogenic. Available evidence concerning
mutagenicity is mixed. Trichloroethylene was mutagenic in two microbial
mutagenicity assays (an Ames test and a host-mediated assay in mice).
However, exposure of mouse sperm cells to TCE had no significant effect on
litter size (dominant lethal assay).
10.2.1 Noncarclnogenic Studies
10.2.1.1 Animal Studies
Trichloroethylene, like most other chlorinated aliphatic hydrocarbons,
produces depression of the CNS and inhibits cardiac function in animals
following acute exposure. At higher doses liver damage occurs. Evidence
indicates that repeated acute exposures are no more harmful to laboratory
animals ;hat a single acute exposure (i.e., there is no accumulation).
Grandjean (1960) exposed male rats to 200 and 800 ppm TCE vapor for 4 to
11 weeks. After a single three-hour exposure the animals were tested in a
previously trained rope-climbing experiment. The number of spontaneous climbs
to receive a reward was significantly increased in comparison to controls.
The observed effect, attributed to TCE-induced excitability, was, however, not
dose dependent. In a subsequent study Grandjean (1963) exposed rats to Ł00 or
800 ppm TCE for six hours and measured swimming performance and motor
activity. At 400 ppm performance was retarded slightly; at 800 ppm
performance was significantly adversely affected. One hour after termination
of exposure no significant changes were observed. •
Adams et al. (1951) exposed several species to various levels or iCE vapor
7 hours/day, 5 days/week for 6 months. The authors determined maximum
no-effect levels of 400 ppm in monkeys, 200 ppm in rabbits and rats and
100 ppm in guinea pigs. Prendergast et al. (1967) subjected rats, rabbits,
guinea pigs, monkeys and dogs to either 730 ppm TCE (5 days/week, 8 hours/day
fcr six weeks) or 35 ppm (continuously for 90 days) . In both these
experiments, ine only effects of TCE exposure ocserved were occasional, siignt
body weight loss or below normal body weight gain.
Kiannerle and Eben (1973) exposed male rats to 55 ppm TCE vapor for 15 weeks
(8 hours/day, 5 days/week). The only observable effect seemed to be increased
liver weights. Clinical hematological values were within normal ranges.
Tucker et al. (1982) exposed male and female CD-I mice to TCE in drinking
water at concentrations of 0, 100, 1,000, 2,500 or 5,000 mg/L for six months.
The 100-mg/L dose produced no observed effects in the mice. The 1,000-mg/L
dose produced an increase in the ratio of liver weight to total body weight in
males only. At the 2,500-mg/L dose males exhibited increased ketone (an
indication of diabetes) and protein levels in the urine and increased relative
liver weights. These effects were not observed in the females. The highest
dose produced decreased body weights, increased liver and kidney weights and
increased urinary ketone and protein levels.
10-2
-------
Stott et al. 1982 administered various doses of TCE to mice and rats by gavage
(5 days/week) for three weeks. At 250 mg/kg minimal hepatotoxic effects were
observed in the mice. Doses of 500 mg/kg/day or greater produced
centrilobular hypertrophy (swelling of the middle of the lobe) in the mouse
livers. In contrast, a dose of 1,000 mg/kg/day resulted in no liver damage in
the rats.
10.2.1.2 Human Studies
In humans, acute exposure to high levels of TCE results in CNS depression such
as incoordination and unconsciousness. Although there have been incidences of
acute exposure, clinical analysis of these incidents rarely establishes
clear-cut exposure limited to only TCE. Chronic exposure of humans to TCE in
occupational settings has resulted in few reports of liver damage.
Single oral doses of TCE ranging from 7.6 to 35 g have been reported to produce
adverse symptoms and eventual recovery. Feldman (1970) reported on a person
exposed to TCE vapors from an overheated, degreasing unit (exposure level not
reported). Symptoms included nausea, vomiting, blurred vision and facial
numbness. An 18-month recovery period resulted in restoration of facial
sensation and motor function. In fatal cases of acute TCE exposure, no tissue
abnormalities were observed at autopsy (Kleinfeld and Tabershaw 1954).
Epidemiological evidence has been generally reported on workers exposed
occupationally. These studies rarely include an unexposed control group.
Demographic characteristics of. the exposed group are not always provided.
Another serious flaw includes lack of information about possible exposures to
other chemicals in the workplace.
Sordodej and Vyskocil (1956) examined 75 persons occupationally exposea to TCE
in drycleaning or degreasing establishments. The TCE concentration varied
from 50 to 630 ppm; workers had been exposed from six months to 25 years.
Symptoms including alcohol intolerance, shivers, giddiness, anxiety and
cardiac abnormalities and sleep disturbances were found to be significantly
correlated -«rith the duration of exposure.
Takamatsa (1962) studied 50 males and female factory workers exposed to TCE
for approximately 2.5 years. Airborne TCE concentrations ranged from 25 to
250 ppm. Workers exposed to less than 50 ppm TCE showed no apparent ill
effects. Six workers in a degreasing room (150 to 250 ppm TCE) reported
headaches, dizziness, giddiness, drunken feeling, flushing of the face,
burning throat and fatigue.
10.2.2 Mutagenic and Carcinogenic Studies
10.2.2.1 Animal Studies
A number of carcinogenic studies have been performed in rats and mice that
indicate the carcinogenic potential of TCE. Experimental details for these
studies are provided in Table 10-1. The National Cancer Institute (NCI)
(1976) conducted a 78-week carcinogenic study in rats and mice. Five-week-old
10-3
-------
TABLE 10-1 CARCINOGENICITY OF TCE
Route
Dose
Duration of
Treatment
Oral
0
1
*-
Oral
Oral
fnhala-
Lion
1,169 mg/kg/day 5 days/week
2,339 mg/kg/day for 78 weeks
869 mg/kg/day 5 days/week
1,739 ag/kg/day for 78 weeks
length of Species
Lxperlment (strain)
549 mg/kg/day 5 days/week
1,097 mg/kg/day for 78 weeks
500 mg/kg 5 days/week
1,000 mg/kg for 103 weeks
1,000 mg/kg 5 days/week
for 103 weeks
100 ppm 6 hours/day,
500 ppm 5 days/week
for 18 months
100 ppm
500 ppm
90 weeks
mouse
(B6C3FI)
90 weeks
110 weeks rat
(Osborne-
Mendel)
103 weeks rat
(Fisher
344)
103 weeks mouse
(B6C3F1)
30 months mouse
(NMRl)
Sex
M
M
M
Comments
F
F
F
F/M
F/M
F/M
F/M
F/M
F
F
F
M
M
M
Hepatocellular tumors
occurred in 26/50 (low-dose)
and 31/48 (high-dose) of the
treated males as compared
with 1/20 control male mice
(P<0.05 for both dose levels).
Hepatocellular tumors occurred
in 4/50 (low-dose) and 11/47
(high-dose) of the treated,
females as compared with 0/20
control female mice (P<0.05
for the high-dose level only).
No effects were seen in rats.
High-dose males showed
significant increase in
kidney adenocarcinomas.
Increased incidence of hepa-
tocellular carcinoma observed
in treated maleB and females.
llistopathological examination
were made on all animals. No
carcinogenic effect was
observed in either sex of rats
or hamsters, or in male mice.
In female mice, the incidence <
of lymphomas was higher in the
low-dose (17/30) and the high-
Reference
NCI (1976)
NTP (1982)
NTP (1982)
Henschler
et al.
(1980)
continued-
-------
Table 10-1 - continued
Duration of
Route
Dose
Treatment
100 ppm
6 hours/day.
500 ppm
5 days/week
for 18 months
100 ppm
6 hours/day,
500 ppm
5 days/week
for 18 Diunths
Inhala-
50 ppm
7 hours/day,
tion
150 ppm
5 days/week
450 ppm
for 104 weeks
l.t-iigth ol
Experiment
36 months
.10 months
107 weeks
50 ppm
150 ppm
450 ppm
7 hours/day,
5 days/week
for 104 weeks
107 weeks
Species,
(strain)
> Sex
rat
F/M
(Wistar)
F/M
F/M
hamster
F/M
(Syrian)
F/M
F/M
mouse
F
(IRC)
F
V
F
rat
F
(Sprague-
F
Dawley)
F
Comments Reference
dose (18/28) groups of animals
than in the control (9/29)
group.
Mice exposed to 150 and Fukuda et
450 ppm TCE had three times al. (1983)
the number of lung tumors
observed in the low-dose
animals and the controls. A
statistically significant
increase was seen when the
number of lung adenocarcinomas
in mice exposed to 150 and
450 ppm TCE was pompared with
the number of lung adenocar-
cinomas in the low-dose and
control animals.
-------
mice and seven-week-old rats were gavaged 5 days/week with TCE with a high or
low dose. No treatment-related effects were observed in rats. There was a
significant increase in hepatocellular carcinomas in the male mice at both
doses and in the females of the high-dose group. Due to suspected contamina-
tion of the original test material with epichlorohydrin (a known carcinogen),
the bioassay was repeated. In the repeat bioassay, rats (F344/N) and-mice
(B6C3F1) a of both sexes were administered TCE by gavage for 103 weeks. Rats
received doses of 500 or 1,000 mg/kg; mice were administered 1,000 mg/kg.
Trichloroethylene was not found to be carcinogenic in female F344/N rats.
The experiment with male rats was considered to be inadequate, since these
rats received dose levels of TCE that exceeded the maximum tolerated dose.
Trichloroethylene was demonstrated to be carcinogenic in both sexes of
B6C3F1 mice, producing hepatocellular carcinomas.
The NTP (1982) completed a bioassay using Fisher 344^a^ rats and B6C3Fl^a^
mice. Trichloroethylene was administered by gavage at doses of 500 or
1,000 mg/kg in rats and 1,000 mg/kg in mice for 103 weeks. High-dose male
rats exhibited*a significant increase in kidney tubular adenocarcinomas. A
dose-related reduction in survival was noted in male rats. Toxic nephrosis
(degeneration of the cells in the kidney tubules) was found in treated rats
dying during the course of the study. In the treated mice, body weights and
survival were reduced in males. In both sexes there was a significant
increase in the incidence of hepatocellular carcinomas.
Henschler et al. (1980) .studied the effect of chronic inhalafion of TCE on the
tumor incidence in NMRI a mice, Wistar rats arid Syrian hamsters. Groups of
male and female animals of each species were exposed by inhalation to 100 or
500 ppm TCE for IS months. At 30 months (for mice and hamsters) and 36 months
(for rats) , no statistically significant increased tumor incidences vers
observed in any group except for malignant lymphomas in female sice in both
treated groups.
Fukuda et al. (1983) examined the effect of chronic inhalation of reagent
grade TCE (99.8" pura; or. the tumor incidence in female ICR" ' -nice and
Sprague-Dawiey racs. Groups of h9 to 51 animals of each species were exposed
to 0, 50, 150 or 450 ppm TCE for 104 weeks. In mice, the incidences of
pulmonary adenocarcinomas in the 150 ppm group and the 450 ppm group were
significantly higher than that of controls. The average number of lung tumors
per mouse in groups exposed to 150 ppm TCE and 450 ppm TCE was more than three
times that of the controls. No increased tumor incidences were observed in
rats.
10.2.2.2 Human Studies
Axelson et al. (1978) conducted an epidemiological study of cancer deaths
among a group of 518 men occupationally exposed to TCE. Exposure was
(a) A series of letters, numbers or a combination of both indicates a specific
strain (i.e., a race or stock) of animals that all have common hereditary
characteristics.
10-6
-------
estimated by measuring a TCE metabolite, trichloroacetic acid (TCA), in the
urine. A TCA level of >100 mg/L was considered to be high exposure,
corresponding to more than 30 ppm in air. Close agreement was found between
observed and expected numbers of cancer deaths based on national Swedish death
rates. Since the sample was small, the authors hesitated to rule out an
increased cancer risk from TCE exposure, especially from rare tumors.
10.3 Quantitative Indices of Toxicity
A number of estimates of noncarcinogenic indices of toxicity and estimates of
carcinogenic risk have been calculated for TCE. These are summarized in
Table 10-2.
10.3.1 Noncarcinogenic Effects
There have been several risk estimates for TCE calculated with regard to
noncarcinogenic endpoints. The National Academy of Sciences (NAS 1980)
calculated a Suggested No-Adverse-Response Level (SNARL) of 105 mg/L of
drinking water based on observation that the lowest oral dose of TCE reported
to produce inebriation was approximately 300 mg/kg. A 100-fold safety factor
was used in the calculation, recognizing that the minimum effect level for
inhibition of reflexes was undoubtedly lower than 300 mg/kg. It was assumed
that the sole source of TCE was drinking water and that a 70-kg human consumes
2 L/day. A seven-day SNARL of 15 mg/L was also calculated by dividing the
one-day value by 7. A chronic minimum effect level was not established due to
lack of appropriate data.
The EPA Office of Drinking Water (USEPA 1979) computed a one-day SNARL value
of 2 mg/L. This value was based on a study in which human volunteers wera
exposed via inhalation to 110 ppic of TCE for an eight-hcur period. The
calculation considered the most sensitive subpopulacion (childrar.) and uoed ar.
uncertainty factor of 100. A ten-day SNARL was conservatively estimated by
dividing the one-SN'ARL by 10 (200 Ug/L) . A chronic SNARL of 74 yg/L was
derived using the sininun effect level of 55 ppm observed in the Kimmerle and
7 s "..J-
The EPA Office of Water Regulations and Standards (USEPA 1980) established an
ADI of 38 mg/day based on the noncarcinogenic endpoints noted in the MCI
(1976) bioassay (dose-related decreased survival, chronic kidney disease).
The lowest dose (548 mg/kg/day, time-weighted average) was considered a
L0EL. Assuming daily consumption of 2 L of water and 6.5 grams of
contaminated fish, th . level that protects against the toxic effects of TCE
was calculated to be 18.3 mg/L.
The American Conference of Governmental Industrial Hygienists (ACGIH 1980)
has recommended a Threshhold Limit Value (TLV) of 50 ppm and a Short-Term
Exposure Limit (STEL) of 150 ppm. These levels are recommended to provide
workers with adequate protection against the toxic effects of TCE. An
8 hour TLV of 100 ppm has been proposed by the National Institute for
Occupational Safety and Health (NIOSH) (1973) for an eight-hour, workday.
10-7
-------
TABLE 10-2 RISK ESTIMATES FOR TCE
Organization
NAS (1977)
NAS (1980)
USEFA (1979)
(Office of Drinking Water)
USEPA (1980)
(Office Water Regulations
and Standards)
ACGIH (1980)
NIOSH (1973)
Noncarcinogenic 10~6 Excess Cancer Risk
6,3 Vg/L
105 mg/L (one-day SNARL) 3.0 yg/L
15 mg/L (seven-day SNARL)
2 mg/L (one-day SNARL)
200 yg/L (ten-day SNARL)
74 yg/L (chronic SNARL)
38 mg/day (ADI) 2.7 yg/L
18.3 mg/L (Ambient Water
Standard)
50 ppm (TLV)
150 ppm (STEL)
100 ppm (TLV, 8-hour)
10-8
-------
10.3.2 Carcinogenic Effects
The IARC has performed an assessment of the degree of evidence for the
carcinogenicity of TCE to humans and experimental animals which was published
by the World Health Organization (WHO) (WHO 1982). This assessment concluded
that TCE is a Group 3 chemical (inadequate evidence of carcinogenicity in
animals or humans). Nonetheless, several carcinogenic risk assessments have
been performed.
A statistical assessment of human cancer risk associated with TCE in drinking
water was performed by NAS (1977) based on the NCI (1976) study. For TCE at a
concentration of 1 ug/L, the estimated risk was calculated to be 0.36 to
1.1 x 10 . The concentration corresponding to a cancer frequency of 10~ is
9.1^to 27.7 ug/L. The upper bound 95% confidence limit estimate was 1.6 x
10 for 1 ug/L. This corresponds^to a concentration of 6.3 Ug/L for an
excess lifetime cancer risk of 10 . Using the data from NTP (1982), another
estimate was made using the multistage model (USEPA 1983). Averaging both the
male and female data sets, the estimated upper 95% confidence estimate cf
lifgtime risk per yg/L was 3.3 x 10 . The concentration corresponding tc a
10 risk level is 3.0 ug/L. The EPA Office of Water Regulations and
Standards (USEPA 1980), using the data from NCI (1976), published a
recommended criterion of 2.7 ug/L for the 10 risk level.
10.4 Special Concerns
Since the toxicity of TCE is highly dependent on its metabolism, the
possibility exists that there will be age and sex differences and additional
risk in extrapolating results from one species to another. Additionally,
there is the possibilitv of synergistic effects with alcohol.
10.5 References
ACGIH. 1980. American Conference of Governmental Industrial Hygienists.
Threshold limit values :or chemical substances and physical agents in '.he
workroom environment with intended changes for 195G. Cincinnati, Ohio:
American Conference of Governmental Industrial Hygienists, p. 406.
Adams EH, Spencer NC, Rcwe KU, McCollister DO, Irish DD. 1951. Vapor
toxicity of trichloroethylene determined by experiments on laboratory animals.
AMA Arch. Ind. Hyg. Occup. Med. 4:469.
Axelson 0, Andersson K, Hogstedt C, Holmberg B, Molina G, de Verdier A. 1978.
A cohort study on trichloroethylene exposure and cancer mortality. J. Occup.
Med. 20:194.
Bardodej Z, Vyskocil J. 1956. The problem of trichloroethylene in
occupational medicine. AMA Arch. Ind. Health 13:581.
Dekant W, Metzler M, Henschler D. 1984. Novel metabolites of
trichloroethylene through dechlorination reactions in rats, mice and humans.
Biochem. Pharmacol. 33:2021-2027.
10-9
-------
Feldman RG, Mayer RM» Taub A. 1970# Evidence for a peripheral neurotoxic
effect of trichloroethylene. Neurology 20:599-606.
Fukuda K, Takemoto K, Tsuruta H. 1983. Inhalation carcinogenicity of
trichloroethylene in mice and rats. Industrial Health 21:243-254.
Grandjean E. 1963. The effects of short exposures to trichloroethylene on
swimming performances and motor activity of rats. Am. Ind. Hyg. Assoc. J.
24:376.
Grandjean E. 1960. Trichloroethylene effects on animal behavior. Arch.
Environ. Health 1:106.
Henschler D, Romen W, Elsasser HM, Reichert D, Ecler E, Radwan Z. 1980.
Caicinogenic'ty study of trichloroethylene by long-term inhalation in three
animals species. Arch. Toxicol. 43:237-248.
Kimmerle G, Eben A. 1973. Metabolism, excretion and toxicology of
trichloroethylene after inhalation. I. Experimental exposure on rats. Arch.
Toxicol. 30:115.
Kleinfeld M, Tabershaw IR. 1954. Trichloroethylene toxicity. AMA Arch. Ind
Hvg. Occup. Med. 10:134.
NAS. 1980. National Academy of Sciences. Drinking water and health. Vol.
III. Washington, DC: National Academy Press.
NAS. 1977. National Academy of Sciences. Drinking water and health. Vol I
Washington.. DC: National Academy Press.
NCI. 197b. National Cancer Institute. Carcinogenesis bioassay of
trichloroethylene. U.S. Department of Health Education and Welfare, Public
Health Service.
NIOSH. 1973. National Institute for Occupational Safety and Heaitti.
Criteria for recommended standard occupational exposure to trichloroethylene.
HFM 73-11025. Washington, DC: U.S. Government Printing Office.
NTP. 1982. National Toxicology Program. Carcinogenesis bioassay
experimental and status report, March 1982. National Toxicology Program:
Research Triangle Park, NC.
Prendergast JA, Jones RA, Jenkins LJ, Siegel J. 1967. Effects on
experimental animals of long-term inhalation of trichloroethylene, carbon
tetrachloride, 1,1,1-trichloroethane, dichlorodifluoromethane, and
1,1-dichloroethylene. Toxicol. Appl. Pharm. 10:270.
Stewart RD, et al. 1962. Observations on the concentrations of
trichloroethylene in blood and expired air following exposure to humans. Am.
Ind. Hyg. Assoc. J. 23:167.
10-10
-------
Stott WT, Quast JF, Watanabe PG. 1982. The pharmacokinetics and
macromolecular interactions of trichloroethylene in mice and rats. Toxicol.
Appl. Pharmocol. 62:137-151.
Takamatsa M. 1962. Health hazards in workers exposed to trichloroethylene
vapor. II. Exposure to trichloroethylene during degreasing operation in a
communicating machine factory. Kumamoto Med. J. 15:43.
Tucker An, Sanders VM, Barnes DW, Bradshav TJ, White KL, Sain LE, GBorzelleca
JF, Munson AE. 1982. Toxicology of trichloroethylene in the mouse. Toxicol.
Appl. Pharmacol. 62:351-357.
USEPA. 1983. United States Environmental Protection Agency. Health
assessment document for trichloroethylene. ECAO. Research Triangle Park, NC.
December 1983 draft.
USEPA. 1980. United States Environmental Protection Agency. Ambient water
quality for trichloroethylene. Washington, DC: U.S. Environmental Protection
Agency. EPA 440/5-80-077.
USEPA. 1979. United States Environmental Protection Agency. SNARL;for
trichloroethylene. Health Effects Branch, Criteria and Standards Division,
Office of Drinking Water, Washington, DC.
USEPA. 1975. Uniced States Environmental Protection Agency. Identification
of organic compounds in effluents from industrial sources.
WHO. 1982. World Health Organization. IARC monographs on the evaluation of
the carcinogenic risk of chemicals to humans. Chemicals, industrial processes
¦ind industries associated with cancer ir. humans. International Agency for
Research on Cancer Monographs. Vol 1 to 29, Supplement 4. Geneva: World
Health Organization.
10-11
-------
11.0
SUMMARY OF TOXICOLOGICAL INFORMATION ON LEAD
This chapter provides a summary of current toxicological information on
lead. This compound was chosen for consideration here because of its high
inherent toxicity and because it is often found as a contaminant at hazardous
waste sites. This chapter is similar to the sort of summary which might be
prepared by a toxlcologist for a hazard assessment, and therefore it Is
somewhat more technical and detailed than chapters 1 to 8 of this handbook. A
nontoxicologist would not be expected to prepare a summary of this sort, but
should, on the basis of the information presented in chapters 1 to 8 of the
handbook, be able to understand and apply this information.
11.1 Chemical Properties and Environmental Stability
Lead is a metallic element designated by the symbol Pb. Contamination of the
environment with lead has increased dramatically since the industrial
revolution, due primarily to lead emissions into the air. These emissions may
be inhaled directly, or may settle on soil or water and be ingested. The
chemical form of lead emissions and its form in soil, water or the food chain
varies considerably, but is nearly always some oxide, salt or complex of the
lead ion (Pb ). Lead oxides, salts and complexes may be altered by
speciation reactions, but the lead ion itself is stable under normal
environmental conditions.
11.2 Summary of Health Effects Data
Lead exposure produces adverse effects on matiy systems of the body, including
the hematopoietic, cardiovascular, nervous, endocrine, renal, reproductive and
digestive systems. Acute lead intoxication in humans is characterized byr
encephalopathy (disease of the brain), abdominal pain, hemolysis (destruction
of red blood cells), liver damage, renal tubular necrosis, seizures, coma ana
cardiorespiratory arrest. Severe poisonings of this sort are rare, and most
concern regarding lead toxicity focuses on insideous injury to the hema-
topoietic system, the nervous system and the cardiovascular system relieving
chronic exposure to lew levels oi lead. Of particular concerr. sre iats zhat
suggest that some of these effects may not have a threshold value.
Lead-induced effects have been extensively investigated both in animals and in
humans. Studies in animals are useful since- dose levels and the chemical form
of administered lead are known, but they are limited by differences in lead
absorption and metabolism between animals and humans. Studies in humans are,
therefore, more directly useful, but accurate knowledge of exposure levels,
exposure routes or chemical form are rarely known.
One means of solving this problem is to assess human exposure to lead by
measuring the concentration of lead in the blood (PbB). This value reflects
the magnitude of current or recent-past exposure to lead, and it is possible
to calculate the exposure that produced the observed PbB value. While this
approach permits the use of human data in identifying NOAEL and LOAEL values
for lead, it is limited by the fact that the calculation of lead exposure from
measured PbB values may not be highly accurate. Moreover, many of the toxic
11-1
-------
effects of lead are irreversible and current or recent-past exposure levels
may not reflect exposure levels at the time the injury actually occurred.
The following sections provide brief descriptions of some representative
studies of the key adverse health effects of lead in animals and humans.
11.2.1 MoncaTcinogenic Studies
11.2.1.1 Effects of Lead on Hematopoiesis
Lead inhibits several key enzymes involved in heme (a component of hemoglobin)
biosynthesis. The activity of delta-aminolevulinic acid dehydratase (ALA-D)
appears to be very sensitive to lead, and inhibition had been reported at
quite low PbB values. Hernberg and Nikkanen (1970) found that ALA-D activity
was inversely correlated with PbB values in a group of subjects with 50Z
inhibition at a PbB level of 16 ug/dL. Other reports have confirmed these
observations across age groups and exposure categories.
The inhibition of ALA-D is reflected by increased levels of its substrate,
delta-aminolevulinic acid (ALA), both in urine and in whole blood or plasma.
The toxicological significance of an increase In cellular or plasma ALA levels
is uncertain, but it appears that ALA can inhibit release of neurotransmitter
from nerve cells, even at ALA levels as low as 1.0 yM (Brennan and Cantrill
1979).
Another enzyme of heme biosynthesis that is inhibite^+by lead is
ferrochelatase. This enzyme inserts ferrous ion (Fe ) into protoporphyrin to
form heme. In lead exposure, the porphyrin acquires a zinc ion in lieu of
ferrous ion, forming zinc protoporphyrin (ZPP). A correlation-rbetween ?b3 and
erythrocyte ZPP has been extensively documented, with a threshold value of
about 15 to 30 yg/dL (Roels et al. 1975, Piomelli et al. 1982).
One of the most characteristic effects of chronic lead intoxication is anemia.
The mechanism of lead-associated anemia appears to be a combination of reduced
hemoglobin production dncl shortened erythrocyte survival, r.saucaa
hemoglobin production is a consequence of inhibition of heme synthesis (as
described above), coupled with a decreased production of globin. It is also
clear that there is a hemolytic component to lead-induced anemia in humans,
owing to shortened erythrocyte survival (Hernberg et al. 1967, Leikin and Eng
1963).
Exposure to lead results in inhibition of heme biosynthesis not only in
erythrocytes, but in other tissues as well. This results in decreased
activity of a number of heme-containing enzymes, including cytochrome P-450
(which is important in metabolism of many drugs, chemicals and natural
compounds) and the renal enzyme 1-hydroxyglase (which mediates the final step
in the synthesis of the active form of vitamin D). Inhibition of the latter
enzyme impairs biosynthesis of the active form of vitamin D, and this may
produce a series of adverse effects related to calcium absorption and cell
development.
11-2
-------
11.2.1.2 Effects of Lead on the Nervous System
Studies in animals indicate that the perinatal period of ontogeny is a time of
particular sensitivity of the nervous system to lead exposure. In nursing
female rats, ingestion of water containing levels as lov as 20 mg Pb /L
(2 mg/kg/day) causes effects on neurotransmitter metabolism in the offspring,
and higher exposure levels result in a number of morphologic, biochemical and
electrophysical changes (Govoni et al. 1980).
Grant et al. (1980) exposed rats indirectly to lead in utero and during
lactation through the mother's milk and, after weaning, directly through
drinking water. There were delays in the development of surface righting+and
air righting reflexes in subjects given water containing-50 or 250 mg Pb /L
(5 or 25 mg/kg/day). Generally similar results were obtained when suckling or
young rats were exposed to lead directly (as opposed to indirect exposure via
the mother).. Cory-Slechta and Thompson (1979) supplied weanling
rats with drinking water solutions containing 25, 150 or 500 mg Pb /L (2.5,
15 or 50 mg/kg/day). Animals exposed to-50 ppm or 300 ppm lead solutions
showed significantly higher response rates than matched controls during
conditioned response training. Jason and Kellogg (1981) reported a
developmental lag in activity around post-natal day (PND) 15 to 18, as
measured in an automated activity chamber. Rat pups were dosed on PND 2 to 14
with lead at 25 or 75 mg/kg. A delay in the characteristic decrease in
activity that normally .occurs in pups at that age was observed, indicating
that lead-exposed pups were significantly more active than control subjects at
PND 18. These alterations in behavior are indicative of altered, neural
functioning in the CNS.
Bushnell and Bowman (1979) investigated the effect of lead on learning in
rnesus monkeys. Lead acetate was fed to the animals for 12 months, resulting
in PbB levels of 30 to 50 ug/aL (versus 5 yg/dL in controls). Exposed
monkeys were significantly retarded in their ability to learn a visual
discrimination task.
lr. conclusion, it appears thaz aiiaraticr.s in behavior and neural deveirrrsent
in rats and monkeys occur as a consequence of oral exposure to lead. These
alterations are presumably indicative of altered neural functioning,
especially in the CNS, but whether such alterations represent biologically
significant impairment in overall functioning of the lead-exposed subjects is
not yet clear.
There are many reports of lead-induced nervous system injury in humans. Morgan
and Repko (1974) reported deficits in hand-eye coordination and reaction time
in an extensive study of behavioral functions in 190 lead-exposed workers.
Similar studies (Arnvig et al. 1980, Haenninen et al. 1978, Valciukas et al.
1978) have found disturbances in visual motor performance, IQ test
performance, hand dexterity, mood, nervousness and coping in lead workers with
PbB values of about 50 to 80 yg/dL.
Seppalainen et al. (1975) measured nerve conduction velocity .(NCV) in 26 lead
workers whose PbB levels had been monitored regularly for several years. Most
11-3
-------
PbB values ranged between 35 and 60 Ug/dL with occasional values as low as 20
or as high as 70 Ug/dL. There was a clear decrease in mean NCV in nerves of
the arm in exposed workers compared to controls (P<0.01). The authors
emphasized that the data showed evidence of neurophysiological effects In
workers whose PbB values never exceeded 70 pg/dL.
Melgaard et al. (1976) observed a clear association between lead exposure and
peripheral nerve dysfunction in 20 automobile mechanics exposed to tetraethyl
lead and other lead compounds in lubricating and high-pressure oils. Half of
the workers had elevated PbB levels (60 to 120 yg/dL) and showed definite
electromyographic deficits (abnormal electrical activity in skeletal muscles).
The mean blood lead level for the control group was 18.6 ug/dL.
Thus, considerable evidence exists that peripheral nerve dysfunction occurs in
adults at PbB levels as low as 30 to 50 ug/dL. The question as to whether
these reflect mild, reversible effects or are true early warning signals of
progressively more serious peripheral neuropathies is still a matter of seme
dispute.
As in experimental animals, the developing child appears to be especially
sensitive to lead-induced nervous system injury. De la Burde and Choate
(1972, 1975) observed neurological dysfunctions including fine motor
dysfunction, impaired concept formation and altered profiles in 70 preschool
children exhibiting pica behavior (the tendency to eat dirt). These children
displayed elevated PbB levels (30 to 100 yg/dL, mean - 58 ug/dL) in comparison
with 70 matched control subjects not engaging in pica. Continuing CNS
impairment in the lead-exposed group, as assessed by a variety of
psychological and neurological tests, was observed when the children were
seven to eight years old, despite the observation that many of their PbB
levels had by then decreased significantly from the initial study.
The relationship between low-lead exposure, psychometric function and
electrophysiological response in children aged 13 to 78 months was explored in
studies by Milar et al. (1980, 1981; , Otto et al. (1981) and Senignus ec al.
(198i). Psychometric evaluation revealed lower scores for children vitn Pb3
levels of 30 Ug/dL or higher compared to children with PbB levels under
30 ug/dL, but the observed IQ deficits were confounded by poor home caregiver
environment scores in children with elevated PbB levels. Electrophysiological
assessments, including analyses of slow potentials during sensory conditioning
and electroencephalogram (EEG) spectra, did provide evidence of CNS effects of
lead in the same children. A significant linear relationship between PbB
(ranging from 7 to 59 ug/dL) and slow wave (SW) voltage was observed (Otto et
al. 1981). Analyses of quadratic and cubic trends in SW voltage did not
reveal any evidence of a threshold for this effect.
Beattie et al. (1975) identified 77 retarded children and 77 normal children
matched for age, sex and geography. Of 64 matched pairs, 11 of the retarded
children came from homes with high concentration of lead in the water. By
contrast, none of the control children came from such homes. In a follow-up
study, PbB values of the mental retardates, measured during the second week of
life, were found to be significantly higher than those of control subjects
(25.5 t 8.9 versus 20.9 = 7.9 yg/dL) t,rloore ec al. 1977). «h«sn compared with
11-4
-------
studies of children suffering neurobehavioral deficits produced by direct
exposure to lead, these studies suggest that the brain of the fetus is
considerably more sensitive to the toxic effects of lead than the brain of the
young child.
11.2.1.3 Effects of Lead on Blood Pressure
Several epidemiological studies have Indicated that chronic lead exposure may
be associated with Increased blood pressure in humans (Dingvall-Fordyce and
Love 1963, Beevers et al. 1980), although other studies have not detected such
an association (Cramer and Dahlberg 1966).
More recently, Harlan et al. (1985) examined the relationship between FbB and
blood pressure -by statistical analysis of the data base obtained during the
second National Health and Nutrition Examination Survey (NHANES-II). This
survey collected information on a representative cross-section of over 20,000
members of the U.S. population. Using simple regression analysis, a direct,
nearly linear relationship was found between PbB and blood pressure in both
men and women aged 12 to 74 years. Since blood pressure is known to be
related to factors such as age and body mass, multiple regression analyses
were performed to separate confounding factors. After accounting for these
and other variables, PbB was found to retain a statistically significant
relationship to blood pressure in males (P<0.05) but not in females. The
authors cautioned that causal inferences about effects of PbB on blood
pressure should not be drawn from this cross-sectional survey, although the
results obtained were consistent with a direct effect.
Pirkle et al. (1985) employed the NHANES-II data base to perform a detailed
statistical analysis sf the relation between PbB and blood, pressure in white
~al2S aged '40 Zz 59 "?ars. This sub-population was selected because, in this
age range, the effaczs cf age on blood pressure are small, and confining the
analyses to white males obviated the confounding effects of sex and race.
Regression analyses correcting for age and body mass indicated that PbB values
correlated to both svstolic and diastolic blood pressure. Segmented
regrassior. analyses indicated there was no zhrashoic beiow which bleed
pressure was not related to ?ba.
11.2.2 Mutagenic and Carcinogenic Studies
Studies of cytogenetic (chromosome appearance) abnormalities in persons
exposed to lead have yielded mixed results. For example, 0'Riordan and Evans
(1974) reported no significant chromosomal damage in male workers with PbB
values from 40 to 120 ug/dL. However, Forni et al. (1976) found the incidence
of abnormal metaphases doubled (PS0.05) in workers exposed for one month to
air lead levels of about 0.8 mg/m .
A number of ingestion studies on the carcinogenic potential of various lead
salts in animals have been reported. The most common observation was
increased frequency of renal tumors, although evidence of tumors in other
tissues has been noted. The doses of lead producing these effects were quite
high, generally 0.1Z to 1Z in the diet (equivalent to about 50 to
500 mc Pb/kg/dav). For example, Azar et al. (1973) reported dose-decendent
11-5
-------
increases in renal tumor frequency in male rats fed 500 to 2,000 ppm Pb in the
diet. It should be noted that these doses are associated with moderate to
severe non-carcinogenic effects in rats.
A number of epidemiological studies of industrial workers with elevated lead
exposure have been conducted to evaluate the role of lead in the induction of
human neoplasia (Cooper 1976, Dingwall-Fordyce and Lane 1963, Lane 1964,
McMichael and Johnson 1982). In general, these studies made no attempt to
consider the types of lead compounds to which workers were exposed, or to
determine probable routes of exposure. While a number of these studies found
an association between lead exposure and the frequency of various cancer
types, no study was sufficiently free of confounding factors to permit a clear
conclusion.
The IARC has performed an assessment of the degree of evidence for the
carcinogenicity of lead and lead compounds in humans and experimental animals
(WHO 1982), This assessment concluded that lead and lead compounds are Group
3 compounds (sufficient evidence for carcinogenicity of some lead salts in
animals, but inadequate evidence for carcinogenicity in humans).
11.3 Quantitative Indices of Toxicity
Many studies in animals indicate that adverse effects occur in pups born to
dams exposed to doses of lead from 5 to 150 mg/kg/day. Similarly, studies in
young animals exposed to lead directly indicate that doses of 5 to
500 mg/kg/day cause behavioral or neurological effects.
Studies in humans suggest that PbB values of around 25 ug/dL or higher are
associated with adverse effects in adults, and even lower values nay be
associated with adverse effects in children. Some effects may occur without i
threshold value. Equations have been developed which describe She
relationship between PbB values and lead exposure via inhaled air and
ingested food, water or soil. While there is some variability between
different studies, application of these equations makes clear that daily
ingestion or inhalation of several hundred oiercgra:?.s or less cf ?b could
yield these PbB values.
The current MCL for lead in drinking water is 50 yg/L (USEPA 1976). This
corresponds to a total AD I of 100 yg/'day. The present TLV for inorganic lead
compounds in inhaled air is 0.15 mg/m , and the STEL is 0.45 mg/m (ACGIH
1980). There is growing awareness that present exposure limits may not
provide an adequate margin of safety (e.g., see NAS 1977), and several
agencies are evaluating recent data on lead toxicity to determine if revisions
in existing guidelines and standards are appropriate.
11.4 Special Concerns
There are a number of special concerns associated with lead toxicity. First,
most adverse effects develop only slowly, but are basically irreversible.
Thus, by the time injury is recognized, permanent harm may already have been
done. Second, the fetus in utero and the young child are especially
sensitive. Third, as analytical techniques and testing protocols become Tiore
11-6
-------
powerful and sophisticated, adverse effects of lead are being detected at
levels previously thought to be safe. Indeed, some researchers feel there may
be no threshold for some lead-Induced effects.
11.5 References
ACGIH. 1980. American Conference of Governmental Industrial Hygienists.
Threshold limit values for chemical substances and physical agents in the
workroom environment with intended changes for 1980. Cincinnati, Ohio:
American Conference of Governmental Industrial Hygienists, p. 243.
Arnvig E, Grandjean P, Beckmann J. 1980. Neurotoxic effects of heavy lead
exposure determined with psychological tests. Toxicol. Lett. 5:399-404.
Beattie AD, Moore MR, Goldberg A. 1975. Role of chronic low-level lead
exposure in the aetiology of mental retardation. Lance 1 (7907) .*589-592.
Beevers DG, Cruickshank JK, Yeoman WB, Carter GF, Goldberg A, Moore MR. 1980.
Blood-lead and cadmium in human hypertension. J. Environ. Pathol. Toxicol.
4:251-260.
Benignus VA, Otto DA, Muller KE, Seiple KJ. 1981. Effects of age and body
lead burden on CNS function in young children. II: EEG spectra.
Electroencephalogr. Clin. NeuroRfiysiol. 52:240-248."
Breennan MW, Cantrill RC. 1979. Delta-aminolaevulinic acid is a potent-
agonist for GABA autoreceptors. Nature (London) 280:514-515.
Bushnell ?J, Bowman RE. 1979. Reversal learning deficits in young monkeys
exposed to lead. Pharmacol. Biochem. Behav. 10:733-742.
Cooper WC. 1976. Cancer mortality patterns in the lead industry. Ann. N.Y.
Acad. Sci. 271:250-259.
Cory-Sieehca DA, Thompson T. 1979. Behavioral toxicity of cr.ronic
postweaning lead exposure in the rate. Toxicol. Appl. Pharmacol. 47:151-159.
Cramer K, Dahlberg 1. 1966. Incidence of hypertension among lead workers: a
follow-up study based on regular control over 20 years. Br. J. Ind. Med.
23:101-104.
De la Burde B, Choate MS, Jr. 1972. Does asymptomatic lead exposure in
children have latent sequelae? J. Pediatr. 81:1088-1091.
De la Burde B, Choate MS, Jr. 1975. Early asymptomatic lead exposure and
development at school age. J. Pediatr. 87:638-642.
Dingwall-Fordyce I, Lane RE. 1963. A follow-up study of lead workers. Br.
J. Ind. Med. 20:313-315.
11-7
-------
Govoni S, Memo M, Lucchi L, Spano PF, Trabucchi M. 1980. Brain
neurotransmitter systems and chronic lead intoxication. Pharmacol. Res.
Commun. 12:447-460.
Granick JL, Sassa S, Granick S, Levere RD, Kappas A. 1973. Studies in -lead
poisoning. II: Correlation between teh ratio of activated and inactivated
delta-aminolevulinic acid dehydratase of whole blood and the blood lead level.
Biochem. Med. 8:149-159.
Grant LD, Kimmel CA, West GL, Martinez-Vargas CM, Howard JL. 1980. Chronic
low-level lead toxicity in the rat. II. Effects on postnatal physical and
behavioral development. Toxicol. App. Pharmacol. 56:42-58.
Haenninen H, Hernberg S, Mantere P, Vesanto R, Jalkanen M. 1978.
Psychological performance of subjects with low exposure to lead. J. Occup.
Med. 20:683-689.
Harlan WR, Landis R, Schmounder RL, Goldstein NG, Harlan LC. . 1985. Blood
lead and blood pressure. Relationship in the adolescent and adult U.S.
population. J. Am. Med. Assoc. 253:530-534.
Hernberg S, Nikkanen J. 1970. Enzyme inhibition by lead under normal urban
conditions. Lancet 1 (7637):63-64.
Hernberg S, Nurminen M, Hasan J. 1967. Nonrandom shortening of red cell
survival times in men exposed to lead. Environ. Res. 1:247-261.
IARC 1980. International Agency for Research on Cancer. Lead and lead
compounds. In: IARC monographs on the evaluation of the carcinogenic risk of
chemicals to humans: some metals and metallic compounds; October, 1979; Lyon,
France. Geneva, Switzerland: World Health Organization/IARC; pp. 325-416.
Jason MK, Kellogg CK. 1981. Neonatal lead exposure: effects on development
cf behavior and striatal dopamine neurons. Pharmacol. Biochem. Behav.
15:i-6«9.
Lane RE. 1964. Health control in inorganic lead industries: a follow-up of
exposed workers. Arch. Environ. Health 8:243-250.
Leiken S, Eng G. 1963. Erythrokinetic studies of the anemia of lead
poisoning. Pediatrics 31:996-1002.
Litman DA, Correia MA. 1983. L-tryptophan: A common denominator of
biochemical and neurological events of acute hepatic porphyrias? Science
222:1031-1033.
McMichael AJ, Johnson HM. 1982. Long-term morality profile of
heavily-exposed lead smelter workers. J. Occup. Med. 24:375-378.
Melgaard B, Clausen J, Rastogi SC. 1976. Electromyographic changes in
automechanics with increased heavy metal levels. Acta. Neurol. Scand.
54:227-240.
11-8
-------
Milar CR, Schroeder SR, Hushak P, Boone L. 1981. Failure to find
hyperactivity in preschool children with moderately elevated lead burden. J.
Pediatr. Psychol. 6:85-95.
Milar CR, Schroeder SR, Mushak P, Dolcourt JL, Grant LD. 1980. Contributions
of the caregiving environment to increased lead burden of children. Am. J.
Ment. Defic. 84:339-344.
Morgan BB, Jr, Repko JD. 1974. Evaluation of behavioral functions in workers
exposed to lead. In: Xintaras C, Johnson BL, DeGroot I, eds. Behavioral
toxicology: early detection of occupational hazards. Washington, DC:
Department of Health, Education and Welfare; pp^. 248-266.
NAS. 1977. National Academy of Sciences. Drinking water and health.
Washington, DC: National Academy of Sciences.
Otto DA, Benignus VA, Muller KE; Barton CN. 1981. Effects of age and body
lead burden on CNS function in young children. I: Slow cortical potentials.
Electroencephalogr. Clin. Neurophysiol. 52:229-239.
Pirkle JL, Schwartz J, Landis JR, Harlan WR. 1985. The relationship between
blood lead levels and blood pressure and its cardiovascular risk implications.
Am. J. Epidem. 121:246-258.
PiomelliS, Seaman C, Zullow D, Curran A, Davidow B. 1982. Threshold for
lead damage to heme synthesis in urban children. Proc. Natl. Acad. Sci.
U.S.A. 79:3335-3339.
Reels HA, Lauwerys PR, Buchet JP, Urelust MT. 1975. Response of free
erythrocyte protoporphyrin and urinary-d-aminolevulinic acid in men and vomen
moderately exposed to lead. Int. Arch. Arbeitsmed. 34:97-108.
Seppalainen AM, Tola S, Hember S, Kock B. 1975. Subclinical neuropathy at
"safe" lavels of lead exposure. Arch. Environ. Health 30:180-i33.
USEPA. 1984. U.S. Environmental Protection Agency. Air quality criteria for
lead. Review Draft. Research Triangle Park, N.C., EPA report No.
EPA-600/8-33-0283.
USEPA. 1976. U.S. Environmental Protection Agency. National interim primary
drinking water regulations. Washington, DC: U.S. Gov. Printing Office, pp.
69-75.
Valciukas JA, Lilis R, Eisinger J, Blumberg WE, Fischbein A, Selikoff IJ.
1978. Behavioral indicators of lead neurotoxicity: results of a clinical
field survey. Int. ARch. Occup. Environ. Health 41:217-236.
WHO. 1982. World Health Organization. IARC monographs on the evaluation of
the carcinogenic risk of chemicals to humans. Chemicals, industrial processes
and industries associated with cancer in humans. International Agency for
Research on Cancer Monographs. Vol. 1 to 29, Supplement. Geneva: World
Health Organization.
11-9
-------
Zielhuis RL. 1975. Dose-response relationships for inorganic lead. I.
Biochemical and haematological responses. Inc. Arch. Occup. Environ. Health
35:1-18.
11-10
-------
SUBJECT INDEX
Subject Page
acceptable dally intake, example
calculation 6-15
acute studies 5-1
animal studies 5-1
bioassays 5-10
carcinogen 2-5,
4-8
cellular defense mechanisms 3-1
chain of custody 7-11
chronic studies- 5-4
clinical studies 5-4
detection limits 2-1
dioxin-furans 9-1
dose 2-5
dose, conversions of 6-2
dose conversion, example 6-6
dose conversion factors 6-7
dose-response curves 3-1
dose-response models 6-10
cotoxicol-ogical guidelines 5-12
ecotoxicologicai studies 5-10
ecotoxicological studies, limitations.5-14
ecotoxicologicai study, example 5-11
ED-, • • 3-9
effacsive concentration (EC) 3-10
effective dose (ED) 3-10
endangerment assessment 8-6
endpoints 4-7
endpoints, -assuring techniques.......—3,
4-9
environmental fate/stability 7-18
environmental persistence 7-23
epidemiological studies 5-5
epidemiological studies, limitations..5-8
exposure 2-6
exposure assessments .....7-1
exposure calculation, present 7-16
exposure coefficients 7-14
exposure, duration of 4-4
exposure estimation, past 7-17
exposure prediction, future 7-18
exposure, routes of 4-1
extrapolation, animals to humans 6-1
extrapolation, high to low dose 6-8
extrapolation, mathematical models of.6-8
hepatoxicity 4-11
incremental risk 8-5
inj ection 4-4
Subject Page
Intake assumptions 7-15
laboratory animal studies 5-1
LD 3-10
lela : 11-1
legal requirements 7-10
lethal concentration (LC) 3-10
lethal dose (LD) 3-10
limitations, toxicological 2-5,
3-5
L0AEL 3-1
LOEL 4-7
mathematical models for extrapolation.6-8
mechanisms, chemical toxicity 2-1
metabolism 4-5
mutagenicity 4-11
neurotoxicity 4-11
N0AEL 3-1
NOEL 4-7.
occurrence data, collection of 7-1
population analysis, exposure..' 7-13
protocols 5-2
renal toxicity 4-12
repair capacity 2-5
reproductive toxicity.- 4-12
reserve capacity 2-5
risk assessment 8-1
risk characterization 8-4
risk levels, non-threshold chemicals..3-5
spiicies differences 4-5
site niscory information checklist.... 7-2
statistical analysis, example 6-4
statistical terms 6-3
subchronic studies 5-4
surface area equivalence ..6-2
teratogenicity 4-12
testing protocols 5-2
therapeutic index (TI) 3-10,
3-12
threshold 3-1
toxic concentration (TC) 3-10
toxic dose (TD) 3-10
toxicoklnetlc differences 6-12
trichloroethylene (TCE) 10-1
uncertainty factors.... 6-15
uncertainty factors, selection of 6-16
uncertainty, sources of 6-13
weight equivalence 6-2
1-1
-------
About Government Institutes
Government Institutes, Inc. was founded in 1973 to provide continuing education and
practical information for your professional development. Specializing in environmental,
health and safety concerns, we recognize that you face unique challenges presented by
the ever-increasing number of new laws and regulations and the rapid evolution of new
technologies, methods and markets.
Our information and continuing education efforts include a Videotape Distribution Ser-
vice, over 140 courses held nation-wide throughout the year, and over 150 publications,
making us the world's largest publisher in these areas.
Government Institutes, Inc.
4 Research Place, Suite 200
Rockville, MD 20850
(301)921-2300
Other related books published by Government Institutes:
TSCA Handbook, 2nd Edition - The law firm of McKenna, Conner &
Cuneo provides a comprehensive look ar. your requiremer.es under the
Toxic Substances Control Act (TSCA). Includes a copy of the TSCA
law, charts, tables, figures and multiple indexes. Softcover/
490 pages/Nov '89/$95 ISBN: 0-86587-791-2
TSCA Inspection Guidance - This manual was developed by EPA to
support its field inspectors in conducting compliance monitoring
inspections of the manufacturing, processing, distribution, use
and disposal of chemicals. Contains detailed inspection guidance
for ensuring compliance with DMN, Recordkeeping & Reporting,
Testing, Imports, Exports, and a review of the current
biotechnology program under TSCA. Softcover/400 pages/Nov '93/$85
ISBN: 0-86587-358-5
TSCA Confidential Business Information Security Manual - Contains
the stringent procedures to be followed by EPA personnel, other
Federal agencies, and contractors ir. handing information,
submitted or collected under TSCA, that is claimed by companies as
confidential business information. Softcover/154 pages/Oct '93/
$69 ISBN: 0-86587-363-1
Good Laboratory Practice Compliance Inspection Manual - Make sure
your laboratory procedures comply with TSCA and FIFRA. The
questions EPA will ask when conducting laboratory inspections of
facilities, techniques, recordkeeping, safety and quality
assurance programs are detailed. Softcover/116 pages/Aug '85/$48
ISBN: 0-86587-584-8
Pesticides Inspection Manual - Written for EPA inspectors
conducting Pesticide Inspections, this manual provides all of the
necessary guidance to carry out the standard field procedures for
a pesticide inspection. Softcover/231 Pages/Sep ' 89/$73
ISBN: 0-86587-784-X
Call the above number for our current
book/video catalog and course schedule.
/
------- |