EPA
United States
Environmental Protection
Agency
Office of
Toxic Substances
Washington DC 20460
EPA-560/11-79-001
Toxic Substances
Support Document
Test Data Development
Standards:
Chronic Health Effects
Toxic Substances
Control Act
Section 4
-------
EPA 560/11-79-001
SUPPORT DOCUMENT
TEST DATA DEVELOPMENT STANDARDS:
CHRONIC HEALTH EFFECTS
TOXIC SUBSTANCES CONTROL ACT
SECTION 4
MAY 1979
HEALTH REVIEW DIVISION
OFFICE OF TOXIC SUBSTANCES
ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
-------
-11-
PREFACE
The Proposed Chronic Toxicity Test Standards
(F.R. ), written under Section 4(a) and 4(b) of
the Toxic Substances Control Act (TSCA), are designed to
ensure the development of reliable and adequate test
data for assessing the chronic health effects of natural
and synthetic chemicals. Moreover, the testing
requirements are not to create unnecessary economic
barriers to technological innovation in the chemical
industry. To fulfill these responsibilities, scientists
and staff of the Environmental Protection Agency (EPA)
have reviewed the literature and discussed the signi-
ficant scientific and economic issues both in Agency
Workgroup meetings and with consultants and reviewers
from outside the Agency. The results of this effort are
reflected in detail in this Support Document and the
Preamble to the Chronic Health Effects Standards.
*
The Support Document is not intended to be a
comprehensive scientific treatise which reviews all the
literature pertinent to chronic toxicity testing.
However, the review and dialogue with scientists were
thorough, wide-ranging, and thoughtfully performed in
order that reasonable and effective standards could be
developed. Along with the Preamble, the Support
Document identifies and discusses in detail the most
significant issues that pertain to the proposed
standards and records the Agency's reasoning behind
their development.
The EPA encourages all interested parties to review
the scientific and economic reasoning expressed in the
Support Document and provide comment to the Agency.
Such input can significantly benefit the development of
Final Test Standards. .All comments will be carefully
reviewed by EPA, and all major points will be addressed
in the final Support Document.
Written comments should bear the document control
number EPA 560/11-79-001 and should be submitted to the
Document Control Officer (TS-793), Office of Toxic Sub-
stances, U.S. Environmental Protection Agency, 401 M
Street, S.W., Washington, D.C. 20460.
-------
-Ill-
ACKNOWLEDGEMENT
The * Office of Toxic Substances gratefully acknowledges
the efforts of the authors, Diane Beal, William Pegram,
and H.G. Williams and the scientific and editorial
contributions of the TSCA Section 4 Workgroup. Special
appreciation goes to the secretarial staff who typed
the manuscript, particularly Carolyn Blake,
Barbara Featherstone, Juanita Herman, Eleanor Jones,
and Bobbe Ward.
-------
-iv-
CONTENTS *'
Preface ii
Acknowledgment iii
Introduction 1
I. Need for Chronic Toxicity Studies I- 1
A. Environmental Factors and Chronic
Health Effects I- 1
B. Environmental Chemicals and Chronic Health
Effects: The Need for Chronic Toxicity
Testing I- 2
C. Impact of Chronic Disease on Health in
the United States I- 5
D. Economic Costs of Chronic Disease. I- 6
E. Environmental Factors Associated with
Oncogenicity in Man I- 9
F. Initiation, Promotion, and Other Chemical
Interactions Which Enhance Oncogenic Poten-
tial of Low-Level Chemical Exposure 1-11
G. Identification of Chemicals with Chronic
Health Effects: The Use of Epidemiology,
Short-Term Tests for Predicting Oncogeni-
city, and Long-Term Chronic Toxicity Tests 1-16
Definitions: Appendix A 1-24
References 1-27
Tables and Figures 1-37
II. Scientific Aspects of the Proposed Chronic
Health Effects Test Standards II- 1
A. Overview of Proposed Standards II- 1
B. Study Design Issues II- 6
1. Test Species, Strain and Sex II- 6
2. Age at Start of Test 11-15
3. Number of Animals/Test Group.' 11-17
-------
-v-
4. Number of Dose Levels and Dose
Selection 11-20
5. Controls 11-25
6. Route(s) of Exposure 11-28
7. Period of Exposure and Observation 11-29
C. Study Conduct Issues 11-33
1. Clinical Procedures 11-32
;
2. Pathological Procedures. 11-39
D. Data Collection and Reporting Issues 11-42
E. Good Laboratory Practice Issues 11-44
F. References 11-51
III. Economic Aspects of the Proposed Chronic
Health Effects Test Standards Ill- 1
A. Summary Ill- 1
B. Methodology and Assumptions Ill- 4
1. Methodelogy Ill- 4
2. Assumptions concerning Study Design Ill- 4
3. Costing Assumptions Ill- 8
4. Items Excluded from Estimates III-ll
5. Variations in Costs. 111-12
6. Use of Ranges 111-13
C. Fixed Costs 111-14
1. Summary 111-14
2. Protocol Design and Study Submission 111-14
3. Project Management and Preparation of
Final Report 111-15
4. Statistical Analysis 111-15
D. Oncogenic Effects 111-17
-------
-vi-
1. Summary 111-17
2. Variable Costs. 111-18
3. Prechronic Testing Costs 111-26
E. Chronic Effects 111-28
1. Summary 111-28
2. Variable Costs 111-29
3. Prechronic Testing Costs .'. r... 111-38
F. Combined Chronic Effects...! 111-41
1. Summary:'. 111-41
2. Variable Costs Ill-41
* -
3. Prechronic Testing Costs.. 111-42
4. Cost Savings Due to Combined Test 111-42
:'_")..•
G. Other Data on Testing Costs 111-44
H. Cost of Alternative'Standards. 111-46
IV. Confidentiality Issues f IV- 1
V. Differences Between TSCA Section 4(b) Test
Standards and FIFRA Guidelines V- 1
-------
Support Document
Test Data Development Standards
Chronic Health Effects
Toxic Substances Control Act
Section 4
Introduction
Under Sections 4 (a),(b) of the Toxic Substances
Control Act (TSCA), the Administrator o,ffthe
Environmental Protection Agency (EPA) is proposing
standards for testing chemicals for oncogenic and non-
oncogenic chronic effects induced in laboratory animals
after long-term, repeated or continuous exposure.
Health effects which are irreversible, of long duration,
or occur only after long-term exposure are classified as
chronic'. The objectives of Chronic tests are to detect
such toxic effects in all affected target organs and
systems and define toxic and no-observed-effect levels
(NOEL) for Non-oncogenic chronic effects in order to
develop data for assessing human risk.
Many chemical agents in the environment are capable
of interacting with biological systems and, thus, of
inducing chronic health effects (NAS, 1975). The
irreversible or long-continuing nature of these health
effects and the insiduous way they may develop—perhaps
years after a single exposure or the initiation of long-
term, low-level exposure to a toxic chemical substance—
make testing of chemicals for chronic toxicity a
necessary regulatory measure for disease prevention
(Hayes, 1975; NAS, 1975; NAS-NRC, 1977).
So that the public might understand the scientific
and economic rationale behind the development of the
Chronic Health Effects Test Standards, the Administrator
of EPA is making available the Support Document for
these Standards. The document addresses issues relevant
to the design and conduct of chronic toxicity tests,
discusses the costs of these tests and, in addition,
provides other background information pertinent to
understanding the need for and the implementation of
For a discussion of EPA's use of the term oncogenicity and other
cancer-related terms, consult "Definitions," Appendix A to
Section I of this document.
-------
Chronic Health Effects Test Standards. The Support
Document is composed of the following sections:
I. Need for Chronic Toxicity Studies
II. Scientific Aspects of the Proposed
Standards
III. Economic Issues
IV. Confidentiality Issues
V. Differences Between TSCA Test Standards
and FIFRA Guidelines
-------
I. Need for Chronic Toxicity studies
A. Environmental Factors and Chronic Effects
B. Environmental Chemicals and Chronic Health
Effects: The Need for Chronic Toxicity
Testing
C. Impact of Chronic Disease on Health in the
United States
D. Economic Costs of Chronic Disease
E. Environmental Factors Associated with
Oncogenicity in Man
1. Physical Factors
2. Biological Factors
3. Chemical Factors
F. Initiation, Promotion/ and Other Chemical
Interactions Which Enhance Oncogenic
Potential of Low-Level Chemical Exposure
1. Cocarcinogenesis
a. Initiation and Promotion
b. Additive and Synergistic Inter-
actions
2. Metabolic Activation of Chemicals and
Carcinogenicity
G. Identification of Chemicals with Chronic
Health Effects: The Use of Epidemiology,
Short-Term Tests for Predicting Oncogeni-
city, and Long-Term Chronic Toxicity Tests
1. Epidemiology
2. Short-Term Tests for Predicting Onco-
genicity
3, Long-Term Chronic Toxicity Tests
a. Oncogenicity Tests
b. Non-oncogenicity Chronic
Toxicity Tests
Definitions: Appendix A
References
Tables and Figures
-------
I. Need for Chronic Toxicity Studies
A. Environmental Factors and Chronic Health
Effects
In its report Human Health and the
Environment; Some Research Needs (1977), the Second
Task Force for Research Planning in Environmental Health
Science characterized the relationship between
environmental factors and chronic health effects in the
following way:
There is virtually no currently-recognized chronic
disease to which environmental factors are not
contributing, either (a) by direct or indirect
action on the organ itself, or (b) through an
influence on modulating mechanisms of
neuroendocrine or immunologic nature. For everyone
of the multitude of cardiovascular, pulmonary,
renal, hepatic, pancreatic, gastrointestinal,
cutaneous, and neurologic diseases as well as
disorders of the hematopoietic system, examples
exist of toxic manifestations brought about by
direct actions of environmental stresses.
Environmental factors, such as chemicals, radiation,
infectious agents, drugs, smoking, nutritional
deficiency and other factors related to diet, have been
associated with chronic health effects (T.F., 1978;
S.T.F. » 1977). The effects may range, for example, from
a subclinical, sustained impairment of the respiratory
tract evidenced only under stress, to chronic diseases
such as emphysema, arteriosclerosis, diabetes mellitus,
cirrhosis of the liver, nephrosis of the kidney, and
benign or malignant tumors, to premature death as a
delayed response to a toxic substance gradually
accumulated in the body (Albert, 1975; Blodgett et al.
1975). However, the data base for evaluating the
relationships between most environmental factors and
chronic health effects is presently inadequate or not
available to government scientists. Recognizing this
problem, Congress enacted the Toxic Substances Control
Act to ensure the availability or development of
adequate test data for regulatory decision-making with
regard to chemical substances and mixtures.
-------
1-2
B. Environmental Chemicals and Chronic Health
Effects; The Need for Chronic Toxicity
Testing
The large and diverse productivity of the
chemical industry and the wide-spread use of chemicals
in modern industrial and agricultural societies ensure
that humans will continue to undergo long-term exposure
to exogenous chemicals. Chemicals in commerce around
the world may number as high as 100,000 with another
1000 new chemicals coming on the market each year.
About 63,000 chemicals are currently in common use
(Maugh, 1978).
In the United States alone, approximately
45,000 chemicals are in current production, according to
the chemical inventory under development by EPA. In
1973 the production of the top fifty chemical substances
in the United States was about 410 billion pounds (C&EN
May 6, 1974). In 1977, production of organic chemicals
indicated on C&EN's "Top 50 Chemical Products List" was
156.9 billion pounds, an 8 percent increase over 1976.
The production for the 50 largest-volume chemicals in
1977 was 482.7 billion pounds — a production record —
and indicators suggest that for 1978, production may
reach 500 billion pounds (C&EN May 1, 1978). These
figures do not include the production of inorganic or
organic chemicals produced in lesser volumes.
Some of these chemicals which enter the
environment undoubtedly are capable of inducing chronic
health effects in man. These effects may take years to
manifest themselves, however, and establishment of
causal relationships will be very difficult.
Consequently, suspect chemical must be identified and
carefully laboratory tested.
Approximately 6000 chemicals have been
laboratory-tested for toxicity—about 6 percent of the
chemicals in commerce. The quality and uniformity of
the testing, however, varies significantly. EPA's study
Preliminary Assessment of Suspected Carcinogens in
Drinking Water"(1975) demonstrates the need for a
regulatory testing program for chronic toxicity of
environmental chemicals. Of the 253 organic chemicals
identified in drinking water, a large number lack
sufficient data for evaluating whether they produce
tumors, gene mutations, birth defects or other serious
chronic effects. Among those organic chemicals
identified were chloroform, carbon tetrachloride,
-------
1-3
benzene, haloethers, chloro-olefins, and polynuclear
aromatic hydrocarbons—all of which have been associated
with carcinogenesis either in man or laboratory
animals. Preliminary epidemiologic studies of aggregate
populations in Louisiana, Ohio, and New Jersey support
the hypothesis that carcinogens in drinking water are
related to the occurrence of human cancer (Harris, Page
and Reiches, 1977).
Since chemicals in surface and drinking water
are prime contenders for entering the human food chain
at low dose levels and repeated exposure, the lack of
knowledge concerning their chronic toxicity is
particularly disturbing and, moreover, argues for an
effective chronic toxicity testing program for
chemicals.
The Task Force on Environmental Cancer and
Heart and Lung Disease (1978), composed of scientists
from EPA, NCI, NHLBI, NIOSH, and NIEHS, has thus far
focused its attention on chemical pollutants and found
evidence associating them with cancer and heart and lung
disease, but the evidence is "diffuse and in many cases
inconclusive." Known or suspected relationships which
the Task Force identified are in Figure 1.
As Figure 1 indicates, of the three diseases,
most is known about the association between
environmental chemicals and cancer. As of December
1977, the International Agency for Research on Cancer
(IARC) had evaluated 368 chemicals selected according to
2 main criteria:
a. that there is evidence of a human
exposure; \
b. that there is some evidence of
carcinogenicity in experimental animals
and/or some evidence or suspicion of
human risk.
»-
IARC reported that 26 chemicals or industrial processes
were associated epidemiologically with cancer in'humans
(Table 1), and for 221 chemicals (Table 2), some
"evidence of carcinogenicity was found in at least one
species of experimental animals" (Tomatis et al.,
1978). The remaining 121 chemicals lacked sufficient
data for an adequate evaluation; however, most likely
some of these as well as additional synthetic and
-------
1-4
natural chemicals will eventually be associated with
human cancer.
Evidence linking environmental chemicals with
non-oncogenic diseases is not as well-developed as for
cancer; however, there are clear indicators that
chemically-induced non-oncogenic health effects pose a
considerable health problem. The causal role of
asbestos in the chronic human disease asbestosis is well
known (Merewether, 1930). The association between
'soft1 water and the death rate from cardiovascular
disease has been verified by studies in Japan, United
States, United Kingdom, and Sweden although the
chemical causal factors have not yet been clearly
identified (Kobayashi, 1957; Schroeder, 1974). Research
such as Schroeder's study of recondite toxicity of trace
elements in laboratory animals (1973) — a study of
"subtle metabolic changes consistent with reasonable
survival" associated with trace elements — indicates
possible association between industrial elements and
chronic diseases. Signs of recondite toxicity include
hypertension, atherosclerosis, diabetes mellitus,
coronary artery occlusion, benign and'malignant tumors,
as well as alterations in carbohydrate and lipid
metabolism, glycosuria, proteinuria, shortened
longevity, reproductive abnormalities1 and weight loss in
older animals.
Other studies in both humans and laboratory
animals have shown association between environmental
chemicals and the expression of chronic diseases. For
example, a relation has been shown between long-term
occupational exposure to vinyl chloride and the
oncogenic effect, angiosarcoma of the liver, and also
the non-oncogenic effect, chronic liver injury—a rather
nonspecific peri'portal fibrosis which in advanced stages
resembles primary biliary cirrhosis (Thomas et al.,
1975). There have been associations between long-term
exposure to fluorocarbons and cardiac palpitation in the
human (Speizer et al., 1975), and long-term exposure to
hydrocarbons and rapidly progressing glomerulonephritis
in the human kidney (Beirne and Brennan, 1972; Harman,
1971; Zimmerman et al., 1975). Moreover, experimental
toxicology has demonstrated that B-cell destruction or
dysfunction in the pancreas can be caused' by chemicals,
suggesting that some diabetes might result from exposure
to toxic chemicals in the environment (Longnecker,
1977).
-------
1-5
These findings attest to the diversity of
adverse chronic health effects associated with chemical
exposure. But typically, the chronic effects are subtle
and difficult to study directly. Exposures in the real
world are usually small; their effects may often be
attributed to other agents to which the subject is
exposed voluntarily as well as involuntarily. However,
the biological implications of long-term, low-level
exposure "may be quite serious--especially for chronic
degenerative disease" (Burger, 1976). Therefore, the
need for testing selected chemicals for these effects is
critical.
C. Impact of Chronic Disease on Health in the
United States
' '*' '
-------
1-6
Of the 2 million people who died in 1975,
1,028,415 succumbed to diseases of the circulatory
system, 365,538 to neoplasms, 109,276 to disease of the
respiratory system, 73,189 to disease of the digestive
system, and 28,029 to disease of the genitourinary
system (Paringer and Berk, 1977).
Prevalence of non-oncogenic chronic conditions
as estimated by the National Center for Health
Statistics totals approximately 174.1 million cases:
chronic circulatory conditions 36.4 million; (Wilder,
1974) chronic skin and musculoskeletal conditions 25.2
and 25.4 million respectively (Wilder, 1974); chronic
conditions of the genitourinary, nervous, endocrine,
metabolism, blood and blood-forming systems 23.2 million
(Scott, 1977); selected chronic respiratory conditions
46.8 million (Wilder, 1973); and selected chronic
digestive conditions, 17.0 million (Wilson, 1973).
With regard to cancer, of the present
population in the United States, nearly 55 million
people (one out of four) will eventually develop the
disease; approximately 700,000 new cancer cases are
expected in 1978, an increase from 675,000 in 1976 and
690,000 in 1977 (Cancer Facts and Figures, 1978).
Furthermore, since cancer usually results in death if it
is not treated successfully, less than half of all
cancer patients live five years after first diagnosis
(Cancer Facts and Figures. 1978; Cairns, 1975). The
American Cancer Society estimates that 390,000 people
will die from cancer in 1978.
Furthermore, the ability of a large number of
people to pursue their lives actively is significantly
affected by chronic disease. According to the National
Center for Health Statistics, "In 1974, almost 7 million
people or 3.3 percent of the noninstitutionalized
population were unable to perform what they considered
their major activity, 7.3 percent were limited in the
kind or amount of major activity, and 3.5 percent were
limited in other activities as a direct result of
chronic diseases." In total, about 30 million persons
had some degree of limitation of activity as a result of
chronic diseases (Wilder, 1977).
D. Economic Cost of Chronic Disease
The cost of health care for people who suffer
from chronic disease is very large. For example, (using
a 4 percent discount for lost future earnings) in fiscal
-------
1-7
year 1975, the total economic cost for diseases of the
circulatory systems was approximately $50.4 billion, for
diseases of the respiratory system $19.7 billion, for
diseases of the digestive system $22.8 billion and
disease of the genitourinary system $8.2 billion (Berk,
Paringer and Mushkin, 1978).
The economic burden which cancer alone inflicts
on society is extraordinarily large. The General
Accounting Office of the United States (GAO) estimated
in 1976 the cost of cancer to be $15 billion per year,
3.5 billion of which goes to care and treatment and the
remainder to loss of earning power and productivity
(GAO, 1976).
Rice and Hodgson (1978) of the National Center
for Health Statistics recently reported, however, the
total economic cost of neoplasms in 1975 was in the
range of $19 to $22 billion which breaks down into
direct cost of $5.3 billion and indirect costs ranging
from $13.6 to $17.0 billion, depending on the discount
rate (Cf. Paringer and Berk, 1977). Direct costs are
outlays for prevention, detection, and treatment for the
illness and indirect costs are the loss in the economy's
output because of disability, morbidity and premature
death of workers (NCHS, 1977).
If "social costs" are computed into the
estimate, the yearly cost of cancer soars even higher.
Abt (1975) attempted to quantify social costs of
cancer--costs of psychosocial deteriorations that are
brought about by disease but are not reflected in the
direct and indirect economic costs—and estimated the
minimum annual social cost of cancer "to be about $2.5
billion, excluding extinction costs and about $138
billion if costs of extinction are included."
(Extinction costs are the collective social costs of
cancer mortality.) Consequently, the annual financial
burden of cancer on people and society in the United
States may be as high as $150 billion.
Moreover, between 1950 and 1976, total health
expenditures rose at an average rate of 9.9 percent, and
in 1976 expenditures were 14 percent greater than in the
previous year. Monetary costs for health care will
continue to rise, suggesting that the economic burden of
chronic disease upon society is ever increasing.
-------
1-8
Furthermore, the average age of the population
has grown older, a trend, it is thought, that will
continue (NCHS, 1977):
Assuming that women average 2.1 births and
that recent death rates prevail, the total
population will be about 262.5
million in the year 2000. The number of
children under age 20 will increase by only
6.3 percent (from 74.6 million to 79.3), while
the number of elderly people will increase by
36.6 percent (from 22.4 million to 30.6
million).
Given this aging of the population, the incidence of
chronic diseases and, therefore, treatment-related costs
to society are expected to increase since these diseases
are found most commonly in the middle and elderly age
groups.
If the incidence of chronic disease could be
reduced by prevention—by testing chemical substances
for chronic health effects—even if the reduction is
only a small fraction, economic and health gains to
society would be significant. The "Forward" to the
study of environmental health issues, Human Health and
the Environment; Some Research Needs (1977), states:
This report is produced against a background of
our nation's increasing recognition that:
1. The costs of medical diagnosis and
treatment are astronomical and growing; a
further increase in investments in these
areas will not yield significant improve-
ments in average longevity, productivity,
or quality of life;
2. Prevention of illness, disability, and
premature death will yield the greatest
benefits to society; and
3. Identification, evaluation, and subsequent
modification of the role of environmental
factors in causing illness and premature
death promise early and major payoffs in
the prevention and control of disease.
The TSCA reflects this view; testing of
selected chemicals is an effective way of modifying the
-------
1-9
role of environmental factors in causing illness and
premature death.
E. Environmental Factors Associated with
Oncogenicity in Man
The percent of cancers among the human
population^in the United States caused by environmental
factors has been variously estimated from 70 to 90
percent (Higginson and Muir, 1973). These percentages
are essentially based on epidemiological studies which
have shown geographic variations in'the incidence of
cancer in various organs'(Dunham and Bailar, 1968;
Higginson, 1969) and the risks of cancer among migrants
conforming within one. or two generations to those of the
adopted country (Haen^zel e_£ al., 1972). As of yet,
there is no consensus of scientific opinion regarding
the exact percentage of cancer due to these factors, but
the important fact remains that they, i.e., physical.
biological, and chemical factors in the environment,
contribute substantially to cancer incidence in the
human population.
1. Physical Factors
*,"
Exposures to ultraviolet light and ionizing
radiation have been shown to cause cancer (Jablon, 1975;
Upton, 1975). Ultraviolet light from the sun is
strongly associated with squamous cell carcinoma of the
skin and, to lesser degrees, with basal cell epithelioma
and malignant melanoma (Emmett, 1973). Ultraviolet light
apparently interacts directly with nucleic acids of the
skin cells (Upton, 1975). Ionizing radiation —
exposure to X-ray, radionucleides, or radiopharma-
ceuticals — can have leukemogenic or other carcinogenic
effects on the human (Jablon, 1975). Acute lympho-
blastic and chronic myeloid leukemia may be caused by
ionizing radiation; an increased incidence of tumors of
the thyroid, respiratory tract, breast, gastrointestinal
'system (except stomach) and' lymphosarcoma has been
diagnosed in the population of Hiroshima and Nagasaki
who received gamma and neutron radiation from the atomic
explosions. In the case of the radium dial painters,
radium isotope which localized in the bone eventually
induced osterosarcomas and leukemia. But if one
excludes skin cancer caused by prolonged exposure to the
sun, a malady usually treatable, radiation presently
plays a relatively small part, "a minute portion," in
overall cancer incidence (Doll, 1977a; Upton, 1975).
-------
1-10
2. Biological Factors
Other environmental agents that have been
associated with cancer are microbes and parasites
(Gross, 1978; Heath et al., 1975). DMA viruses —
particularly the Epstein-Barr virus (EBV) and herpes
simplex virus type 2 (HSV-2)— are suspect human
oncogens. They have been associated with cancer in
humans and have shown tumor formation in lower
animals. However, proof of oncogenicity in humans has
not been demonstrated for either virus. Suggestive
evidence links the metabolic activity of intestinal
bacterial flora to colon cancer and possibly breast
cancer, uriaary tract infection to gastric cancer, and
infestation by the parasite Shistosoma haematobium with
increased risk of malignancy. However, it is for these
biological agents as it was for the viruses, proof of
oncogenicity in the human is not yet conclusive (Heath
et al., 1975). Consequently, whether biological agents
cause cancer in the human is still an open question
(Doll, 1977a).
3. Chemical Factors
The exogenous agents most often associated
with the etiology of cancer are natural and synthetic
chemicals. Boyland (1969) estimates that 90 percent of
human cancers are evoked by chemical agents. In a
recent study, cancers attributable to occupational
factors — carticularly synthetic chemicals — were
estimated between 20 and 40 percent, a finding
considerably higher than previous estimates (Bridbord et
al., 1978). Epstein (197*0 has suggested four broad
categories of environmental chemicals:
(a) Natural chemicals which are normal
dietary components, such as nitrates
and nitrites;
(b) natural fungal or plant toxins in
crops, such as aflatoxins and
cycasins;
(c) complex organic and inorganic
mixtures, for example the community
air, water, and occupational
pollutants, such as coke tar pitch
volatiles;
-------
1-11
(d) synthetic chemicals-agricultural
chemicals, such as pesticides,
fertilizers, food additives, fuel
additives, household chemicals and
industrial chemicals.
There is some controversy about which one
of these four is the most significant category in the
etiology of chemically-induced cancer. (Berg, 1977;
Doll, 1977b; Weisburger, Cohen, and Wynder, 1977; Wynder
and Gori, 1977; Devesa and Silverman, 1978; and
Schneiderman, 1978). Nevertheless, it is with the
synthetic and natural chemical substances, developed and
produced by industry, which may present "an unreasonable
risk of injury to health or the environment" (with the
exception of pesticides and food additives covered by
FIFRA and FFDCA respectively) that TSCA is primarily
concerned. Knowledge of the carcinogenic activity of
some of these chemicals, the interaction among
environmental chemicals that can enhance tumorgenicity
at low dose, the role of a variety of host factors in
chemically-induced carcinogenesis, and the awareness of
the potential for ever increasing exposure (often
involuntary) of modern agricultural and industrial
societies have caused the Congress and the EPA to focus
on these chemicals.
F. Initiation^ Promotion, and Other Chemical
Interactions Which Enhance Oncogenic
Potential of Low-Level Chemical Exposure
Relatively little is known about the response
of a total population to lifelong oncogenic exposures
which result from low-levels of both natural and
synthetic oncogens in the environment (Saffiotti, 1977);
the latent period associated with carcinogenesis can
delay discovery of carcinogenic action for as much as
forty years. Moreover, tumor type and incidence of
chemically-induced cancers may be indistinguishable from
cancers associated with other causes. However, exposure
to even low-levels of carcinogens may pose serious
risks, particularly under long-term exposure conditions,
because of the chemical interactions that can occur in
the human body. Cocarcinogenic interactions can
significantly enhance the oncogenic effects of chemical
substances administered at low doses to both human and
laboratory animals.
-------
1-12
1. Cocarcinogenesis
Cocarcinogenic action refers to the
interaction of two or more factors which augment tumor
induction when administered either together with or
subsequent to a sub-optimal dose of a carcinogen
(Berenblum, 1974). Berenblum classifies these different
kinds of cocarcinogenic actions as follows:
1. Additive action, when the cocarcinogenic
agent itself possesses definite
carcinogenic activity.
2. Synergistic action, when the combined
effect exceeds the summation of their
separate actions.
3. 'Incomplete1 carcinogenic action,
responsible for only one phase of
carcinogenesis, i.e. operating as
initiator only or as promoter only.
4. Preparative action, by rendering the target
organ or tissue more responsive to
carcinogenic action.
5. Permissive influences of carcinogenic '
action, e.g. through solvent effect, or
by influencing the rate of absorption of
the carcinogen into the cell, or by
affecting the metabolism of the carcinogen
prior to its action, or by influencing its
rate of detoxification and excretion/ etc.
6. Influence on viral action, by favoring
virus release from its hidden site:
depressing the immune response of the
animal, activating an incomplete virus,
or by rendering the target organ responsive
to the virus.
7. Conditional influence on the induced tumor
during its previsible stage, by encouraging
its continual growth, in the case of a
hormone-dependent tumor, or by countering
the immune resistance of the host to the
tumor.
Of these various cocarcinogenic actions, incomplete
carcinogenic action (initiation and promotion),
-------
1-13
synergistic action, and additive action have received
considerable attention and will be at the focus of the
following discussion.
Initiation and Promotion
Carcinogenesis is generally viewed as
complex biochemical and cellular process which occurs in
two stages: initiation followed by promotion.
Initiation.is the rapid induction of imperceptible and
essentially irreversible changes in cells exposed to a
carcinogen, (such as benzo(a)pyrene) which result in the
production of transformed but dormant cells. Promotion
is the stimulation of these transformed cells with an
agent (such as phorbol esters) which ultimately
causes them to grow and proliferate into progressively-
growing tumors (Boutwell, 1974; Van Duuren,
1966). It has been found that promoters administered
subsequent to a tracer dose of a carcinogenic polycylic
aromatic hydrocarbon lead to extensive tumor formation
on the skin of mice (Haddow, 1959, Hecker, 1971).
Evidence suggests, moreover, that the
induction of lung cancer by exposure to cigarette smoke
is due to relatively small amounts of the initiating
carcinogen and larger amounts of promoting agents
(Wynder and Hoffman, 1967, 1972; Van Duuren et al.,
1971; Bock, 1972; Wynder and Mabushi, 1972). It has
been found that cigarette smoke significantly enhances
the development of lung cancer among uranium miners
(Archer et al., 1976) and asbestos workers (Nicholson,
.1976) and contains ingredients which are powerful
promoters of polyclic aromatic carcinogens (Van Duuren
et al., 1966). These findings show that three classes
of.carcinogens—pplycyclic aromatic hydrocarbons,
ionizing radiation, and asbestos—accompanied by
promoter agents have a markedly enhanced tumor response
in humans (Albert and Burns, 1977).
Because of this initiation-promotion
interaction, low-level exposures to oncogens in an
environment "loaded" with many other chemical
substances, some of which may serve as promoters, may
show a much larger tumor incidence than the actual dose
would indicate and/or a shortened latent period for
tumor development. In either case, the impact on human
health could be serious.
-------
1-14
b . Additives and Synergistic
Intera c t i. onsi
Additive and synergistic (often
called syncarcinogenic) interactions between chemicals
and oncogens enhance the oncogen's potency to induce
tumorgenesis. It has been shown that some exogenous
chemicals (even with different chemical structures)
which are not oncogenic alone at low doses may, when
administered simultaneously or consecutively at the same
low dose, react with the target organ and result in the
development of tumors. Examples are the hepatotrophic
oncogens: dimethyl- and diethylnitrosamine, nitrb-
somorpholine and dimethylaminobenzene (Schma'hl, 1977).
When these four chemicals are administered singly in
subdivided doses, no heptomas develop even though the
total dose is the same^for the combination study. But
if the doses of the four chemicals are administered
simultaneously, liver tumors develop, evidencing an
additive effect. Montesano et al. (1974) reported a 31
percent incidence of squamous cell carcinomas of the
tracheo-bronchial tract in hamsters with benz(a)pyrene
followed by diethylnitrosamine under exposure conditions
where neither carcinogen induced tumors alone. With
regard to spontaneously occurring tumors, it would seem
likely that exposure to environmental chemicals could
synergistically or additively enhance these oncogenic
processes (Crump et al., 1976).
2. Metabolic Activation of Chemicals and
Carcinogenicity ""L~"" "^"
Chemical oncogens may be divided into two
broad classes: direct acting ultimate carcinogens and
procarcinogens. Direct acting ultimate carcinogens,
such as nitrogen or sulfur mustard, methyl
methanesulfonate, bis-(chloromethyl) ether, and others,
have a chemical structure which makes them inherently
reactive, thus able to react directly with cellular and
molecular receptors in the target cells. Most of the
known environmentally important oncogens, such as
polycyclic aromatic hydrocarbons, aromatic amines, amino
azo dyes, aflatoxins, and others are procarcinogens;
they require metabolic activation to reactive
derivatives which can chemically interact with receptors
in target cell(s) (Weisburger, 1978). For example, the
liver's mixed function oxidase system is thought to
activate the oncogens 3,4 benzypyrene, 2-acetylamino
fluorene, and aflatoxin B (Felton and Nebert, 1975).
-------
1-15
Examples of chemicals requiring and not
requiring metabolic activation are found in the simple
alkyl group (Fig. 2). Dimethylnitrosamine is
metabolically dealkylated by mixed-function oxidases in
the endoplasmic reticulum, forming the monoalkyl
derivative which spontaneously decomposes to the
reactive monoalkyldiazonium ions. ' On the other hand,
N-methyl-N-nitrosourea does not require enzymatic
activation; the reaction with water and other cellular
nucleophiles results in formation of the monoalkyl
derivative which likewise spontaneously decomposes to the
monoalkyldiazonium ions (Miller, 1978).
In-vivo, the reactive chemical oncogen or
metabolite has relatively electron-deficient atoms
(electrophiles) that react with nucleophilic sites in
the cell, i.e., atoms in the cell that easily share
electrons. The macro-molecules, DNA, RNA, and protein
in the cell in the target tissue have numerous
nucleophilic sites to which the chemical derivatives
bind, a process which correlates with the induction of
transformed cells and the subsequent formation of
neoplasia.
The crucial parameter in determining
whether a given carcinogen is active under certain
conditions, the degree and extent of its activity, the
site it affects, and under some conditions the time
required to elicit the effect, is the ratio of the
enzyme-mediated activation-detoxification reactions
(Weisburger and Williams, 1975). The introduction of
exogenous chemicals into the human body, even at low
levels, can modify the level of metabolic enzyme
activity and as a consequence the activation-
detoxification ratio. An increase in metabolic
activation may enhance the oncogenic effects of low-dose
chemicals. Given the myriad chemicals which occur in
our environment, it is possible that the metabolic
activation of some procarcinogens will be enhanced.
Chemicals which have been shown to cause enzyme
induction and thus enhance activation or detoxification
reactions are polycyclic aromatic hydrogens and their
quinones, phenobarbital, certain chlorocarbon pesticides,
antioxidants, some hormones, and some dietary
ingredients.
-------
1-16
G. Identification of Chemicals with Chronic
Health Effects: The Use of Epidemiology,
Short-Term Tests for Predicting Oncogenicity,
and Long-Term Chronic Toxicity Test's^~~7~^
t
Epidemiological studies may provide the most
satisfactory data for investigating human health effects
resulting from exposure to environmental chemicals. .
Such studies avoid the uncertainties of extrapolating
experimental animal or in vitro system, data to humans
and the high doses of the bioassay to the probable low-
dose of human exposure. However, epidemiology has
inherent limitations. A program designed to prevent
adverse health effects by identifying and regulating
toxic chemicals must rely heavily on the direct testing
of chemical substances for health effects. The.
following discussion reviews epidemiology, short-term
tests, and long-term tests and explains why scientists
at EPA presently view data developed by adequately
designed and properly conducted long-term tests as the
most reliable and timely means for evaluating chemicals
for chronic health Effects in human.
" 1. Epidemiology
Epidemiology has been defined as the study
of the origin, nature, pathology, and prevention of
diseases temporarily prevalent in a community or
throughout a large area. Epidemiologists analyze human
incidence and mortality rates, identifying significant
changes in incidence or mortality for each disease; they
compare inter- and intra-country rates and compute t
baseline or spontaneous rates for each disease and .
excessive rates which may be associated with
environmental factors; and they attempt to establish a
quantitative relationship between the dose of the
environmental factors to which a population is exposed
and the incidence of disease within the population.
Human epidemiology provides post facto
information about the effects of chemicals on the human
since it gathers health data from a population already
exposed to a chemical. It is the most reliable' data 'for
human risk assessment and a very important method for
detecting adverse health effects, for example, the
increased'incidence of bladder cancer among, workers in
benzidene (IARC, 1972) and B-napthylamine (I.ARC,.,"
1974). But it is limited by a number of factors, some
of which are the following:
-------
1-17
a. Because of the long latency period
for the development of most cancers and other chronic
diseases in humans, epidemiological studies.do not
detect disease until after a population has been exposed
and a significant number of its members contract the
disease. Disease prevention, therefore, is effective
only for later populations for whom exposure is limited
or prevented by the findings of the earlier study. For
cancer, where the latency period is 5-40 years, many
years may pass before an effect is observed, and there
is little that can be done to protect people previously
exposed who have induced but as yet clinically
unrecognizable tumors. In addition, since
epidemiological studies are retrospective, they cannot
be used to assess the risk of new chemical substances or
substances which have only recently entered the environ-
ment. ,
b. People are exposed to chemicals in
great varieties and concentrations in their job,, diets,
and surrounding environment. Singling out one factor as
a cause of disease is difficult, especially in the case
of cancer because of the possibility of synergistic
effects and long latency periods of oncogens.
c. The population under study must be
well-defined with regard to exposure; the inclusion of
unexposed people in the study dilute it, causing an
underestimate of risk.
d. When .investigating chronic low-level
environmental exposure, it is difficult to establish a
control population, i.e., a major fraction of the
population with little or no exposure.
e. The high degree of mobility in the
U.S. society complicates the exposure pattern. People
change doctors and hospitals often, making data
collection difficult.
f. Individuals are resistent to
participation in medical studies. Investigators are
often unable to obtain medical histories and records.
g. Epidemiological studies are
questionable when it comes to detecting toxic activity
of environmental chemicals at low-levels in the
atmosphere. Chronic toxic effects of chemicals may be
too subtle to identify; they may be masked by other
-------
1-18
effects, attributed to other causes, or be one of a
number of participatory causes for a disease process.
Occupational exposure to chemicals such as 2-
napthylamine, vinyl chloride, and asbestos gave rise,
respectively, to rare cancers of the bladder, angio-
sarcomas of the liver, and mesotheliomas of the lung
cavity. Such specific and rather rare tumors may be
detected by epidemiological studies. But for environ-
mental exposures to chemicals which induce a variety of
common tumors, epidemiology probably will be unable to
associate the malady with the caused factor (Bridbord
et al. 1978; Berg, 1977).
To prevent chemically-induced chronic
diseases, timely and reliable data on chronic health
effects of chemical substances are essential. While
epidemiological studies are extremely useful, they
cannot be relied upon as the primary means of
developing data EPA needs for regulatory actions
concerning toxic substances. Testing of selected
chemicals for oncogencity is essential.
-------
1-19
iShort-term Tests for Predicting Oncogenicity
Toxicologists generally agree that a common goal
is the development of a battery of reliable short-term
tests for predicting oncogenicity. Given the thousands
of chemicals to which humans are exposed in the environ-
ment, most of which have not been tested adequately if
at all for oncogenic effects, the need for quick,
inexpensive short-term tests with high predictive value
is obvious. These tests will be useful in identifying
possible oncogenic chemicals and establishing their
priority for testing in the long-term bioassay.
Moreover, the data developed in the short-term tests
will assist in the interpretation of the long-term
effects data.
Although a variety of short-term in vitro tests
are under development (tests based on unscheduled DNA
synthesis in human skin fibroblasts; transformation of
epithelial cells or fibroblasts in vitro, and others
(Bridges, 1976; Miller, 1978; Purchase je_t al., 1976),
the Salmonella/microsome mutation test developed by Ames
and McCann appears to be the most popular (Ames, 1971;
Ames it*!-, 1975; McCann and Ames, 1977); however, the
in vitro cell transformation test has been shown to be
an equally good indicator of oncogenicity (Purchase et
al.., 1976).
The Salmonella/microsome test and the in vitro
cell transformation tests are relatively inexpensive,
give quick results, and show sensitivity to very small
chemical doses. Moreover, correlations with the results
of long-term animal bioassay have been good. Validated
against 175 chemicals whose oncogenic effects in animals
are well-known, the Salmonella/microsome test detected
157 positives (90 percent); of 108 chemicals identified
as non-oncogenic in animal tests, theSalmonella/micro-
some test detected 94 (87 percent) (McCann and Ames,
1977). In an evaluation performed by Purchase and
colleagues (1976), again using compounds found oncogenic
in animal tests, the Salmonella/microsome test was found
91 percent accurate in detecting oncogens and 93 percent
accurate in detecting non-oncogens. And in this same
evaluation, the cell transformation test, using neonatal
Syrian hamster kidney fibroblasts, detected 91 percent
of the carcinogens and 97 percent of the noncarcinogens.
A 90 percent reliability for the test chemicals
still means, however, that a significant number of other
chemicals could be misevaluated, given the extremely
-------
1-20
large and varied universe of chemicals either in the
environment, underdevelopment, or pending distribution
(Purchase _et ^., 1976). And too, until there is a
rigorous method of validating the short-term test
systems, perhaps along the lines suggested by Saffiotti
(1978), the reliability of the short-term tests will
remain in doubt.
EPA does not think that these short-term tests are
presently capable of replacing the long-term bioassay
for several important reasons. The Salmonella/microsome
test detects the mutagenicity of a chemical, not its
capacity to produce tumors, the "endpoint" of the long
term bioassay. Insofar as chemically-induced carcino-
genicity and mutagenicity correlate, the test may be a
satisfactory indicator, but the extent of this correla-
tion is not conclusively known; not all mutagens are
demonstrable carcinogens and vice versa. Compounds such
as metals, hormones, chrysene, urethan, and thioaceta-
mide, all of whom have carcinogenic activity in intact
animals but are not bacterial mutagens, would not be
detected by the most sensitive and accurate Salmonella/
microsome test (Tardiff, 1978).
For all short-term tests which use an in vitro
metabolic activating system, there is the possibility
the in vitro system does not mimic the whole animal or
human system. Nor do the short-term tests mimic
pharmacologic distribution of the chemical, macro-
molecular repair capabilities, immunological mechanisms
or other factors of the mammalian system. Consequently,
extrapolation of short-term in vitro test data to man would
be precarious. In fact, for the performance of human risk
assessments there is presently no mechanism to extrapolate
in vitro data to man.
Two other potential weaknesses of the short-term tests
concern mammalian aging processes and sex-linked
differences. Since one characteristic of oncogenicity
is the latent period, the mammalian aging process may be
a significant parameter in oncogenicity testing. In his
review of toxic drug reactions in old age, Zbinden
(1973) indicates that there is a higher susceptibility
of old people for toxic drug reaction because of the
aging of tissue, decrease in enzyme activity, and
general deterioration. Such conditions could possibly
Alter response to low-level exposure to environmental
chemicals,and short-term tests are not sensitive to them.
And sex-linked differences in response to chemical
oncogens have also been shown in human and experimental
animals; however, short-term tests presently are not
capable of mimicing these differences.
-------
1-21
Thus, the state-of-the-art and our understanding
of short-term tests are not sufficiently advanced to
allow such tests to be used in place of long-term tests.
VLong-term Chronic Toxicity Tests
a. Oncogenicity Tests
A properly conducted, long-term oncogenicity
test is presently the definitive test model for
estimating the oncogenic risk of chemicals for humans
(Miller, 1978; NAS/NRC, 1977; Page, 1977; Saffiotti,
1978). Having mammalian tumor-induction as its end-
point, the oncogenicity bioassay is the only source of
direct evidence (other than in the human) of chemically-
induced tumors in the mammalian species. Moreover, of
all test systems, it comes closest to mimicing human
routes of exposure and metabolic/ pharmacologic
processes which activate and distribute chemicals.
In keeping with the great importance of the
oncogenioity bioassay, national and international
advisory groups have reviewed the procedures for con- ,
ducting long-term oncogenicity tests several times:
NAS/NRC, 1969; WHO, 1969; Berenblum, 1969; FDA, 1971; Ad
Hoc Committee on the Evaluation of Low Levels on
Environmental Chemical Carcinogens, 1971. The National
Cancer Institute has improved its Bioassay Program
during the past few years
and published its methods in the form of guidelines
(Sontag, .et_.al..f 1976).
Recently, an IARC/WHO ad hoc work group revised
its assessment criteria for evaluating chemically-
induced oncogenicity in humans and/or experimental
animals and concluded:
In the presence of appropriate experimental
carcinogenicity data and in,,the absence of
adequate human data, it is reasonable to
regard chemicals for which there is 'strong
evidence1 of carcinogenicity (i.e., unques-
tionable production of malignant tumors in
animals) as if they were carcinogenic to
humans (Tomatis et^ al. , 1978).
Most types of human cancers can now be
chemically-induced in animals in the oncogenicity
bioassay. Animal models for assessing carcinogenicity
include: bronchogenic lung cancer; carcinoma of the
-------
1-22
larynx; large bowel cancer; carcinoma of the pancreas;
kidney carcinoma; urinary bladder carcinoma, and mammary
cancer (Saffiotti, 1978). And virtually all oncogens
known to be active in humans have also been shown
positive in animal tests.
The long-term oncogenicity bioassay is not
always an optimal test system. For example, some
critics emphasize that substances used in animals test
systems are usually specific agents, whereas, humans are
exposed to multiple environmental stresses. This same
criticism applies, of course, to short-term tests.
Other critics assert that the high doses administered to
test animals are not comparable to the low-levels of
exposure experienced in the environment. However high
dosing is necessary to assure the sensitivity of the
test, moreover, it has been shown that high dosing does not
necessarily cause false positive results (Innes et al.,
1969). (Cf. discussion of dose in Section II.B.Sa).
Some commentators contend that extrapolation of
laboratory data from test animals to man can be very
difficult because interspecies variations in the
metabolism of chemicals, in the enzyme systems, in life-
span are variables whose consequences are largely
unknown for testing. In this regard, the National
Academy of Sciences/National Research Council's report
on Pest Control (1975) concluded:
As a working hypothesis, in the absence of
countervailing evidence for the specific
agent in question.... it appears reasonable
to assume that the lifetime cancer incidence
induced by chronic exposure in man can be
approximated by the lifetime incidence in-
duced by similar exposure in laboratory
animals at the same total dose per body
weight.
EPA agrees that evidence sufficiently indicates that
results shown in the long-term bioassay can predict the
human risk and should be used for the identification and
removal of cancer causing chemicals (Rail, 1978;
Tomatis, 1974; Tomatis, 1977).
For further discussion, consult Section II of this
document.
b. Non-Oncogenic Chronic Toxicity Tests
Given the large volume of chemicals in commerce
and under development which require testing, the large
-------
1-23
investment of tine and money in the traditional, long-
term bioassay, the possibly serious impact that toxic
chemicals in the environment may have on human health,
and the awareness that epidemiological studies cannot
provide primary data in most cases, scientists have
attempted to design an optimal chronic toxicity test
system which will provide adequate toxicologic
information while reducing the investments of time and
money for each chemical test.
Short-term (3-4 months) chronic tests have been
proposed by Weil and McCollister (1963) and McNamara
(1976). Although such studies are useful in
establishing no-effect dose levels for environmental
chemicals, scientists at EPA do not think such tests
provide adequate data for Section 4 test requirements.
Since humans are exposed to synthetic chemicals
in the environment for probably a lifetime, chronic
toxicity tests for assessing their health effect should
be for a comparable period of time—lifetime or near-
lifetime for studies in laboratory animals (Loomis,
1968; NAS, 1975).
Testing for this period of time in animals
(24-30 months) allows observation of toxic effects
related to the aging process; age related factors such
as altered tissue sensitivity, changing metabolic and
physiological capability, and spontaneous disease, which
can influence the degree and nature of toxic responses,
can be assessed (WHO, 1978). Moreover, with long-term
testing, chemicals may produce different toxic responses
with repeated dosing; different metabolic pathways may
beCcome involved; bioaccumulation of the chemical in the
tissues of the host may occur. Only with long-term
dosing can these responses as well as effects on the
longevity of the animal's life be monitored.
In short, the non-oncogenic chronic toxicity
test is the most reliable means to establish a no-
observed-adverse- effects level which can then be used
to define the lifetime "acceptable daily intake" (ADI)
of a chemical for humans.
-------
Appendix A to Section I
Definitions
In this document, terminology specific to cancer is
consistent with definitions and usage found in the
cancer literature. To assure a common understanding of
key terms, however, the following definitions and
references which characterize the disease are provided.
Cancer — an extremely complex, multistaged disease
process or group of disease processes which gives rise
to malignant tumors (Anderson and Scotti, 1976;
Berenblum, 1974; Weinstein et al., 1975).
Tumor — a neoplasm, an abnormal mass of tissue
developed by progressive growth and proliferation of
transformed cells at a rate uncoordinated with
contiguous normal cells. Growth continues after
cessation of the triggering stimulus. Although composed
of cells and intercellular substances apparently similar
to those found in embryonic and mature tissues, the
tissue mass is characterized by growth activity rather
than function; function may, however, not be lacking.
More than 270 types of human neoplasms, including
cancer, have been recognized and defined histologically,
each of which may be distinct in its behavior. Tumors
are generally classified as benign or malignant,
depending on their biological characteristics (Anderson
and Scotti, 1976; Berenblum, 1974; Willis, 1967.
Benign Tumor — a tissue mass that is usually
slow-growing and expansive but which after reaching a
certain size may cease to expand. A benign tumor may
press against or push aside normal tissues, but its
cells do not invade adjoining tissue or metastasize.
The tumor appears as a circumscribed, well-demarcated
growth, usually surrounded by a fibrous capsule.
Histologically, benign tumors are composed of well-
differentiated, mature tissue closely imitating the
normal tissue of their origin (Anderson and Scotti,
1976). Although some benign tumors do not endanger the
life of the host, if they are situated so as to interfer
with some vital organ, cause hemorrhage, or promote
unregulated hormone production, they can be life-
threatening (Berenblum, 1974). Moreover, some benign
tumors may become malignant and thus of serious concern
(Cf. discussion under "Oncogenesis).
Malignant Tumor — a cancer, a tissue mass that
usually grows by expansion and invasion of surrounding
-------
1-25
tissue. Generally unencapsulated and poorly demarcated,
nalignant tunors often spread by local invasion of sur-
roundinq tissue or metastasis, i.e., transformed cells
break off fron the tumor mass and migrate through blood
or the lymphatic systen and establish colonies in
distant organs. Histologically, the cells and struc-
tural organization of malignant tumors generally exhibit
an inadequate maturation and, to varying degrees, a lack
of differentiation so pronounced that often the tissue
of origin is difficult to identify. Unless t-he
malignant tunor is removed or its cells killed, it
usually causes the death of the host. Approximately TOO
different types of cancer have been classified (Andorson
and Scotti,1976; Berenblum, 1974).
Carcinogenesis -- the production of a cancer, or
malignant tumor. Although most cancers originate in
epithelial tissues, — particularly lining ^nithelial of
the skin, hollow organs, and respiratory, digestive and
genitourinary systems — malignant neoplasms oth<^r than
carcinoma are included in the definition of rarrino-
genesis, i.e., neoplasms which arise in the connective
tissue (sarcomas) and in the blood-forming tissues
(leukemias). Moreover, since the distinction between
malignant neoplasms and benign neoplasms is in some
cases very difficult to judge (see discussion below),
the inclusion of benign neoplasms in the definition of
carcinogenesis, not withstanding the fact that a cancer
is not benign, has been suggested: "Carcinogpnesis is
the generation of benign and malignant neonlasia in 1-he
broadest possible sense including generation of sarco-
mata and leukemia" (Hecker, 1976).
Oncogenesis — (from the Greek onkos, meaning bulk
or mass and the Latin genere, meaning to make or create)
the production of tumors in the broadest possible sense,
both benign and malignant.
Since EPA's employment of the term oncoqenesis may
be controversial, a brief description of the ^nency's
rationale follows.
In the past, a tumorigen, when defined as an agent
which induced only benign tumors, was considered less
dangerous than a carcinogen, a malignant tumor-inducing
agent. This classification is unsatisfactory, however.
The conversion of benign papillomas of the skin into
squamous carcinomas which invade and metastasize has
been shown (Berenblum, 1974); moreover, benign growths
such as leukoplakea of the oral cavity, ostertis defor-
mans (a bone disease), and polyps in the digestive tract
(particularly villous papillomas) appear to become
-------
1-26
malignant (Levin et al., 1974). The International
Agency for Research on Cancer (IARC) has concluded:
Many chemicals induce both benign and
malignant tumors; few instances are
recorded in which only benign neoplasms
are induced by chemicals that have been
studied extensively. Be.n.ign tumors may
represent a.state in the evolution of a
malignant neoplasm or they may be 'end-
points' which do not readily undergo
transition to malignant neoplasm. If a
substance is found to induce only benign
neoplasms in experimental animals, the
chemical should be suspected of being a
carcinogen and requires investigation
(Tomatis et al., 1978).
The IARC position generally reflects scientific
opinion on the issue (Mrc-k Commission Report,
1969).. In agreement, EPA has taken the position that
for the purposes of performing risk assessment for human
health-effects, all tumorigens mu-st be considered as
potential carcinogens; "any evidence of tumorigenic
activity in animals is a signal that the agent is a
potential human carcinogen11 (Albert* et al., 1977; EPA,
1976).
Consequently, the term oncogenesis and its
derivatives seem most appropriate for describing the
disease process which includes benign and malignant
neoplasms, including the generation of sarcomata and
leukemia. An oncogen then is a chemical, physical, or
biological agent which induces either benign or
malignant neoplasms, or both.
Irreversibility of the Oncogenic Effect — refers
to the irreversible change of a normal cell to a neo-
plastic cell effected by an oncogen. The neoplastic
cells are self-replicating, i.e., clone neoplastic
daughter cells which in turn replicate neoplastic cells,
and this process continues, forming the tumor. Once
transformation of the target cell has occurred, the
cells are capable of irreversible autonomous growth
regardless of whether exposure to the oncogen continues.
• . • - *
••"• Latent Period - a period of time between exposure
to an oncogen and manifestation of a tumor — estimated
to be 5 to 40 years in the human.
-------
1-27
REFERENCES
Abt, C.C. 1975. The Social Cost of Cancer.
Social Indicators Research. 2:175-190.
»
Ad Hoc Committee on the Evaluation of Low Levels of
Environmental Chemical Carcinogens: Evaluation of
Envrionmental Carcinogens. Report to the Surgeon
General. 1971. (NIH-77762), Washington.
Albert, R.E. 1975. The Costs and Effects of
Chronic Exposure to Low-Level Pollutants in the
Environment—Testimony Before the Subcommittee of the
Environment and the Atmosphere/House of Representa-
tives. (Microform) Washington.
Albert, R.E. and Burns, F.J. 1977. Carcinogenic
Atmospheric Pollutants and the Nature of Low-Level
Risks. In: Hiatt, H.H., Watson, J.D. and Winsten,
J.A. Origins of Human Cancer-Book A. Cold Spring
Harbor Laboratory, pp.289-292.
Albert, R.E., Train, R.E. and Anderson, E. 1977.
Rationale developed by the Environmental Protection
Agency for the Assessment of Carcinogenic Risks. J.
Natl. Cancer Inst. 58:1537-1539.
j
Ames, B.N. 1971. The detection of Chemical
Mutagens With Enteric Bacteria. In: Hollaender, A.'
Chemical Mutagens: Principles and Methods for Their
Detections, Vol-I. Plenum, New York.
Ames, B.N., McCann, J. and Yamosaki, E.r 1975.
Methods for Detecting Carcinogens and Mutagens With the
Salmonella/Mammalian Microsome Mutagenicity Test.
Mutat. Res. 31:347.
Anderson, E.V. 1974. Growth Slows in Top 50
Chemicals Outputs. C&EN. May 6, 1974.
Anderson, W.A.D. and Scotti, T.M. 1976. Synoposis
of Pathology. 9th ed. C.V. Mosley Co., St. Louis
Archer, V.E., Gillam, J.D., Wagoner, J.K. 1976.
Respiratory Disease Mortality among Uranium Miners.
Ann. N.Y. Acad. Sci. 271:280.
Ashford, N. 1976. Crisis in the Workplace: Occupa-
tional Disease and Injury. A Report to the Ford
Foundation. MIT Press. Cambridge, Mass.
-------
1-28
Beirne, G.J. and Brennan, J.T. 1972. Glomeru-
lonephritis Associated with Hydrocarbon Solvents. Arch.
Environ. Health 24:365.
Berenblum, I. 1969. Carcinogenicity Testing.
UICC Tech. Rep. Ser., Vol. 2 Geneva, Switzerland.
Berenblum, I. 1974. Carcinogenesis as a
Biological Problem. North-Holland Publishing Co.,
Amsterdam.
Berg, J.W. 1977. World-Wide Variations in Cancer
Incidence as Clues to Cancer Origins. In: Hiatt,
H.H., Watson, J.D., Winsten, J.A., ed. Origins of Human
Cancer, Book A. Cold Spring Harbor Laboratory, pp.15-
19.
Berk, A., Paringer, L. and Mushkin, S.J. 1978.
The Economic Cost of Illness Fiscal 1975. Medical
Care. 16:785-790.
Blodgett, J.E., McCullough, J.M., Biniek, J.P.,
Costello, G.A., Justus, J.R. 1975. Effects of Chronic
Exposure to Low-Level Pollutants in the Environment.
U.S. Government Printing Office, Washington.
Bock, F.G. 1972. Detection of Weak Carcinogenic
Stimuli. In: Nakahara, W., Takayama, S., Sugimura, T.,
Odashima, S. Toxicss in Chemical Carcinogenesis. Uni-
versity of Tokyo, pp.335-70.
Boyland, E. 1969. The Correlation of Experimental
Carcinogenesis and Cancer in Man. Progress in
Experimental Tumor Research. 11:222-234.
Boutwell, R.K., 1974. The Function and Mechanism
of Promoters of Carcinogenesis, CRC Crit. Rev.
Toxicol. 2:419-443.
Bridbord, ,K., Decoufle, P. Fraumeni, J.E., Hoel,
D.G., Hoover, R.N., Rail, D.P., Saffiotti, U.,
Schneiderman, M.A., Upton, A.C. 1978. Estimates of the
Fraction of Cancer in the United States Related to
Occupational Factors. Prepared by NCI, NIEHS, and
NIOSH.
Bridges, B.A. 1976. Short-Term Screening Tests
for Carcinogens. Nature. 261:195-200.
Burger, E.J. 1976. Protecting the Nation's
Health. D.C. Heath: Lexington, Mass.
Cairns, J. 1975. The Cancer Problem. Scientific
American. 233(5):64-78.
-------
1-29
Cancer Facts and Figures: 1978. American Cancer
Society, New York.
Chiazze, L., Levin, D.L. and Silverman, D.T.
1977. Recent Changes in Estimated Cancer Morality.
In: Hiatt, H.H., Watson, J.D., Winsten, J.A., ed.
Origins of Human Cancer. Book A. Cold Spring Harbor
Laboratory, pp. 33-44.
Committee for the Revision of NAS Publication
1138. 1977, Principles and Procedures for Evaluat-
ing the Toxicity of Household Substances. National
Academy of Sciences, Washington.
Crump, K.S., Hoel, D.G., Langley, H. and Peto, R.
1976. Fundamental Carcinogenic Processes and Their
Implication for Low-Dose Risk Assessment. Cancer Res.
36:2973.
Devesa, S.S. and Silverman, D.T. 1978. Cancer
Incidence and Morality Trends in the United States:
1935-74. J. Natl. Cancer Inst. 60(3):545-571.
Doll, R. 1977a. Strategy for Detection of Cancer
Hazards to Man. Nature. 265:589-96.
Doll, R. 1977b. Introduction. In: Hiatt, H.H.,
Watson, J.D., Winsten, J.A. Origins of Human Cancer.
Book A. Cold Spring Harbor Laboratory.
Dunham, L.J. and Bailar, J.C. 1968. World Maps of
Cancer Mortality: Rates and Frequency Ratios. J. Natl.
Cancer Inst. 41:155.
Emmett, E.A. 1973. Ultraviolet Radiation as a
Cause of Skin Tumors. CRC Crit. Rev. Toxicol. 2:211-
255.
Environmental Protection Agency. 1975.
Preliminary Assessment of Suspected Carcinogens in
Drinking Water—Report to Congress. EPA, Washington.
Environmental Protection Agency. 1977. Monitoring
to Detect Previously Unrecognized Pollutants in Surface
Water. EPA, Washington.
Epstein, S.S. 1974. Environmental Determinants of
Human Cancer. Cancer Research. 34:2425-2435.
Felton, T.S. and Nebert, D.W. 1975. Mutagenesis
of Certain Activated Carcinogens Associated With
Genetically Mediated Increases in Mono-Oxygenase
Activity and Cytochrome P-450. J. Biol. Chem. 250:6769.
-------
1-30
Food and Drug Administration Advisory Committee on
Protocols for Safety Evaluation: Panel on Carcinogene-
sis Report on Cancer Testing in the Safety of Food Addi-
tives and Pesticides. 1971. Toxicol. Appl.
Pharmacol. 20:419-438.
General Accounting Office. 1976. Federal Efforts
to Protect the Public From Cancer Causing Chemicals.
June 16, 1976, p. 1. <
Grdss, L. 1978. Viral Etiology of Cancer and
Leukemia: A Look Into the Past, Present and Future—
G.H.A. Clowes Memorial Lecture. Cancer Research.
38:485-493.
Haddow, A. 1959. The Chemical and Genetic Mecha-
nisms of Carcinogenesis. I. Nature and Mode of Action,
and II. Biologic Alkylating Agents. In: Homburger, F.
(ed.). The Physiopathology of Cancer. Hoeber and
Harper, New York. pp.565-85.
Haenszel, W., Kirihara, M., Segi, M., Lee, R.K.C.
1972. Stomach Cancer Among Japanese in Hawaii. JNCI.
49: 969-988.
Harman, J.W. 1971. Chronic Glomerulonephritis and
the Nephrotic Syndrome Induced in Rats With N,N'-
diacetyl benzidine. J. Pathol. 104:119.
i
Harris, R.H., Page, T. and Reiches, N.A. 1977.
Carcinogenic Hazard of Organic Chemicals in Drinking
Water. In: Hiatt, H.H., Watson, J.D. and Winsten, J.A.
1977. Origins of Human Cancer. Book A. Cold Spring
Harbor Laboratory -
i
Hayes, W.J. 1975. Toxicology of Pesticides.
Williams and Wilkins Co., Baltimore.
Heath, C.W., Caldwell, G.G., Fearino, P.C. 1975.
Viruses and Other Microbes. In: Fraumeni, J.F. Persons
High Risk of Cancer. Academic Press, New York.
Hecker, E. 1971. Isolution and Characterization of
Cocarcinogenic Principles from Croton Oil. In: Busch,
H. ed. Methods in Cancer Research. Academic Press, New
York. pp.439-84.
Hecker, E. 1976. Definitions and Terminology in
Cancer (Tumor) Etiology. 2. Krebsforsch. 86:219-230.
Higginson, J. 1969. Present Trends in Cancer
Epidemiology. Proc. Canad. Cancer Congr. 8:40-75.
-------
1-31
Higginson, J. and Muir, C. 1973. Epidemiology.
In Holland, J.F. and Frei, E. eds. Cancer Medicine. Lea
and Febiger, Philadelphia.
IARC. 1972. IARC Monographs on the Evaluation of
the Carcinogenic Risk of Chemicals to Man, 1, Some
Inorganic Substances, Chlorinated Hydrocarbons, Aromatic
Amines, N-Nitroso Compounds and Natural Products, Lyon.
IARC. 1974. IARC Monographs on the Evaluation of
the Carcinogenic Risk of Chemicals to Man, 4, Some
Aromatic Amines, Hydrazine and Related Substances,
N-Nitroso Compounds and Miscellaneous Alkylating Agents,
Lyon.
Innes, J.R.M., Ulland, B.M., Valerico, M.G.,
Petrucelli, L., Fishbein, L., Hart, E.R., Pallotta,
A.J., Bates, A.J., Falk, R.R., Gart, H.L., Klein, J.J.,
Mitchell, M., and Peters, J. 1969. Bioassay of Pesti-
cides and Industrial Chemicals for Tumorigenicity in
Mice: A Preliminary Note. J. Natl. Cancer Inst.
42:1101-1114.
Jablon, S. 1975. Radiation. In: Fraumeni, J.F.
ed. Persons at High Risk of Cancer. Academic Press,
New York.
Kay, K. 1977. Chemical Agents From Occupational
Sources: Reactions to Environmental Agents. In: Lee,
D.H.K. ed. Handbook of Physiology, Section 9. American
Physiological Society, Bethesda.
Kobayashi, J. 1957. Geographic Relationship
Between the Chemical Nature of River Water and Death-
Rate from Apoplexy. Ber. d. Ohara Inst. f. Laudiverts
Beologue. 11:12-21.
Levin, D.L., Devesa, S.S., Godwin, J.D. and
Silvermann, D.T. 1974. Cancer Rates and Risks. 2nd.
ed. U.S. Government Printing Office, Washington.
Longnecker, D.S. 1977. Environmental Factors and
Diseases of the Pancreas. EHP. 20:105-112.
Loomis, T.A. 1968. Essentials of Toxicology. Lea
and Febiger, Philadelphia.
i
Maltoni, C. and Lefemine, G. 1974.
Carcinogenicity to Bioassays of Vinyl Chloride. I.
Research Plan and Early Results.* Environ. Res. 7:387-
405.
-------
1-32
Maugh, T.H. 1978. Chemicals: How many are there?
Science. 199:162.
McCann, J. and Ames, B.N. 1977. The Salmonella/
Micro-some Mutagenicity Test. Predictive Value for
Animal Carcinogenicity. In: Hiatt, H.H., Watson, J.D.
and Winstein, J.A. Origins of Human Cancer, Book C.
Cold Spring Harbor Laboratory.
McNamara, B.P. 1976. Concepts in Health
Evaluation of Commercial and Industrial Chemicals.
In: Mehlman, M.A., Shapiro, R.E., Blumenthal, H. eds.
New Concepts in Safety Evaluation. Advances in Modern
Toxicology, Vol. 1, Part 1. John Wiley and Sons, New
York.
Merewether, E.R.A. 1930. The Occurrence of
Pulmonary Fibrosis and Other Pulmonary Affections in
Asbestos workers. J. Ind. Hyg. 12:198-222, 239-257.
Miller, E.G. 1978. Some Current Perspective on
Chemical Carcinogenesis in Humans and Experimental
Animals: Presidential Address. Cancer Research.
38:1479-1496.
Montesano, R., Saffiotti, U., Ferrero, A., Kaufman,
D.G. 1974. Brief Communication: Synergistic Effects of
Benzo(a)pyrene and Diethylnitrosamine on Respiratory
Carcinogenesis in Namatus. J. Natl. Cancer Inst.
53:1395.
Mrak, E.M., Chairman. 1969. Report of the
Secretary's Comnission on Pesticides and Their Rela-
tionship to Environmental Health. U.S. Government
Printing Office, Washington.
National Academy of Sciences/National Research
Council. 1969. Guidelines for Estimating Toxicologi-
cally Insignificant Levels of Chemicals in Food.
Washington.
National Academy of Sciences/Ndtional Research
Council Environmental Studies Board. 1975.
Contemporary Pest Control Practices -ind Prospects: The
Report of the Executive Committee 1. Pest Control: An
Assessment of Present and Alternative Technologies.
Washington.
-------
1-33
National Academy of Sciences/National Research
Council. 1975. Principles for Evaluating Chemicals in
the Environment. Report Prepared for the Environmental
Protection Agency by the Environmental Studies Board and
the Committee on Toxicology. Washington.
National Academy of Sciences/National Research
Council. Committee for the Revision of NAS Pabl. 1138.
1977. Principles and Procedures for Evaluating the
Toxicity of Household Substances. Washington.
National Center for Health Statistics: Vital
Statistics of the United States. 1975. Health Resources
•Administration, DHEW, Rockville, MD.
National Center for Health Statistics. 1977.
Statistics Needed for Determining the Effects of the
Environment on Health. U.S. Government Printing Office,
Washington.
National Center for Health Statistics. 1977.
Health: United States—1976-1977. U.S. Government
Printing Office, Washington.
Nicholson, W.J. 1976. Asbestos - The TLV
Approach. Ann. N.Y. Acad. Sci. 271:759.
Page, N.P. 1977. Concepts of a Bioassay Program
in Environmental Carcinogenesis. In: Kraybill, H.F.
and Mehlman, M.A. eds. Environmental Cancer, vol. 3 in
Advances in Modern Toxiocology. Wiley and Sons, New
York.
Paringer, L.C. and Berk, A. 1977. Costs of
Illness and Disease Fiscal Year 1975: Report #Bl.
Public Services Laboratory of Georgetown University,
Washington.
Purchase, I.F.H., Longstaff, E., Ashby, j., Styles,
J.A., Anderson, D., Lefevre, P.A., and Westwood, F.R.
1976. Evaluation of Six Short-Term Tests for Detecting
Organic Chemical Carcinogens and Recommendations for
Their Use. Nature. 264:624-627.
Rail, D.P. 1978. Validity of Extrapolation of
Results of Animal Studies to Man. Paper Presented at
the New York Academy of Sciences Week, Washington, June
28, 1978.
-------
1-34
Rice, D.P. and Hodgson, T.A. 1978. Social and
Economic Implications of Cancer in the United States.
Paper Presented to the Expert Committee on Cancer
Statistics of the WHO and IARC at Madrid, Spain. June
20 to 26, 1978.
Saffiotti, U. and Autrup, H. eds. 1978. In Vitro
Carcinogenesis. Guide to the Literature, Recent
Advances and Laboratory Procedures. DHEW Pub. No. (NIH)
78-844.
Schmahl, D. 1977. Multiple Environmental
Carcinogens: Their Significance. Food. Cosmet.
Toxicol. 15:475-485.
Schneiderman, M.A. 1978. The Links Between the
Environment and Health. In: Brown, G.E. and Rogers,
P.G. Sponsors. Proceedings of the National Conference
on the Environment and Health Care Costs—U.S. House of
Representatives, August 15, 1978. U.S. Government
Printing Office, Washington.
Schroeder, H.A. 1973. Recondite Toxicity of Trace
Elements. In: Hayes, W.I., ed. Essays in Toxicology.
Vol. 4. Academic Press, New York, pp. 108-199.
Schroeder, H.A. 1974. Role of Trace Elements in
Cardiovascular Diseases. Mod. Clinics of N. Amer.
58:381-396.
Scott, G. 1977. Prevalence of Chronic Conditions
of the Genitourinary, Nervous, Endocrine, Metabolic, and
Blood and Blood-Forming Systems and of Other Selected
Chronic Conditions, United States—1973. NCHS Publica-
tion, Rockville, MD.
Storck, W.J. 1978. Production Rises for Most
Major Chemicals. C&EN. May 1, 1978.
Second Task Force for Research Planning in
Environmental Health Science. 1977. Human Health and
the Environment: Some Research Needs. U.S. Government
Printing Office, Washington.
Selikoff, I.J., Hammond, B.C. and Churg, J.
1968. Asbestos Exposure, Smoking and Neoplasia. J.
Amel. Med. Assoc. 204:106-112.
Sontag, J.A., Page, N.P., and Saffiotti, U. eds.
1976. Guidelines for Carcinogen Bioassay in Small
Rodents. National Cancer Institute Carcinogenesis Tech-
nical Report Series No. 1. DHEW Publ. No. (NIH) 76-
801.
-------
1-35
Speizer, F.E., Wegman, D.H. and Ramirez, A.
1975. Palpitation Rates Associated With Flourocarbon
Exposure in a Hospital Setting. New Engl. J. Med.
292:624-626.
Tardiff, R.G. 1978. In Vitro Methods of Toxicity
Evaluation. Ann. Rev. Pharmacol. Toxicol. 18:357-69.
Task Force on Environmental Cancer and Heart and
Lung Disease. 1978. Environmental Pollution and Cancer
and Heart and Lung Disease. U.S. Environmental Protec-
tion Agency, Washington.
Thomas, L.B. 1975. Vinyl Chloride Induced Liver
Disease From Idiopathic Portal Hypertension (Banti's
Syndrome) to Angiosar comas. New Engl. J. Med. 292:17.
Tomatis, L. 1974. Inception and Duration of
Tests. In: Carcinogenesis Testing of Chemicals. L.
Golberg (ed.). CRC Press, Inc. Cleveland, Ohio.
pp. 23-27.
Tomatis, L. 1977. Validity and Limitations of
Long-Term Experimentation in Cancer Research. IARC Pub.
16:299-307.
Tomatis, L. , Agthe, C., Bartsch, H. , Huff, J.,
Montesano, R., Saracci, R., Walker, E., Wilbourn, J.
1978. Evaluation of the Carcinogenicity of Chemicals:
A Review of the Monograph Program of the International
Agency for Research on Cancer (1971 to 1977). Cancer
Research. 38:877-885.
Upton, A.C. 1975. Physical Carcinogenesis:
Radiation, History and Sources. In: Becker, F.I., ed.
Cancer: A Comprehensive Treatise, Vol. I. Plenum Press,
New York.
Van Duuren, B.L., Sivak, A., Segal, A., Orris, L.
and Langseth, L. 1966. The Tumor Promoting Agents of
Tobacco Leaf and Tobacco Smoke Condensate. J. Natl.
Cancer Inst. 37:519.
Van Duuren, B.L., Sivak, A., Katz, C. and
Melchionne, S. 1971. Cigarette Smoke Carcinogenesis:
Importance of Tumor Promoters. J. Natl. Cancer Inst.
47:235-40.
Weil, C.S. and McCollister, D.D. 1963. Safety
Evaluation of Chemicals. Relationship Between Short-and
Long-Term Feeding Studies in Designing an Effective
Toxicity Test. Agric. Food Chem. 11:486-491.
-------
1-36
Weinstein, I.B., lamuguchi, N., Gebert, R. and
Kaighn, M.E. 1975. Use of Epithelial Cell Cultures for
Studies on the Mechanism of Transformation by Chemical
Carcinogens. In Vitro. 11:130-141.
Weisburger, E.K. 1978. Mechanisms of Chemical
Carcinogenesis. Ann. Rev. Pharmacol. Toxicol. 18:395-
415.
Weisburger, J.H. 1975. Chemical Carcinogenesis.
In: Casarett, L.J. and Daull, J. Toxicology: The Basic
Science of Poisons. Macmillan, New York. pp.333-378.
Weisburger, J.H. and Williams, G.M. 1975.
Metabolism of Chemical Carcinogens. In: Becker, P.P.
Cancer: A Comprehensive Treatise. Vol. 1. Plenum
Press, New York. pp. 185-234.
Weisburger, J.H., Cohen, L.A. and Wynder, E.L.
1977. On the Etiology and Metabolic Epidemiology of the
Main Human Cancers. In: Hiatt, H.H., Watson, J. and
Winsten, J.A. 1977. Origins of Human Cancer. Book A.
Cold Spring Harbor Laboratory, pp.567-601.
Wilder, C.S. 1973. Prevalence of Selected Chronic
Respiratory Conditions, United States—1970. NCHS
Publication, Rockville, MD.
Wilder, C.S. 1974. Prevalence of Chronic
Circulatory Conditions, United States—1972. NCHS
Publication, Rockville, MD.
Wilder, C.S. 1977. Limitation of Activity Due to
Chronic Conditions, United States—1974. NCHS
Publication, Rockville, MD.
Wilder, C.S. 1974. Prevalence of Chronic Skin and
Musculoskeletal Condition, United States—1969. NCHS
Publication, Rockville, MD.
Willis, R.A. 1967. Pathology of Tumors, 4th ed.
Butterworth Co., London.
Wilson, R.W. 1973. Prevalence of Selected Chronic
Digestive Conditions, United States—July-December 1968.
NCHS Publication, Rockville, MD.
World Health Organization. 1969. Principles for
the Testing and Evaluation of Drugs for Carcinogenicity.
WHO. Techn. Rep. Ser. 426. Geneva.
-------
1-37
World Health Organization. 1978. Acute, Subacute
and Chronic Toxicity Tests. In: Principles and Methods
for Evaluating the Toxicity of Chemicals. World Health
Organization. Geneva, pp.95-115.
Wynder,- E.L. and Gori, G.B. 1977. Contribution of
the Environment to Cancer Incidence: An Epidemiologic
Exercise. J. Natl. Cancer Inst. 58(4):825-832.
Wynder,E.L.. and Hoffman, D. 1967. Tobacco and
Tobacco Smoke. Academic Press, New York.
Wynder, E.L. and Hoffman, D. 1972. Less Harmful
Ways of Smoking. J. Natl. Cancer Inst. 48:1748-58.
Wynder, E.L. and Mabushi, K. 1972. Etiological and
Preventive Aspects of Human Cancer. Prev. Med. 1:300-
34. =
Zbinden, G. 1973. Progress in Toxicology. Vol. 1.
Springer-Verlag, New York.
Zimmerman, S.W., Groehler, K. and Beirne, G.J.
1975. Hydrocarbon Exposure and Chronic Glomerulonephri-
tis. Lancet. 2:199.
-------
TABLES AND FIGURES
SECTION I
-------
Table 1
Chemicals or industrial processes associated with cancer induction in humans: comparison of target organs and main routes of
exposure in animals and humans
Chtmical or indus-
trial process
1. Aflatoxins
2. 4-Aminobi-
phenyl
3. Arsenic com-
pounds
4. Asbestos
5. Aununine
(manufac-
ture of)
6. Benzene
7. Benzidine
6. Bisfchloro-
methyl) ether
9. Cadmium-us-
ing Indus-
tries (possi-
bly cadmium
oxide)
10. Chlorampheni-
col
11. Chtoromethyl
methyl ether
(possibly as-
sociated with
bb(chloro-
methyl) ether
12. Chromium
(chromate-
producing
industries)
Main type of ex-
posure"
Environmen-
tal, occupa-
tionaK
Occupational
Occupa-
tional, me-
dicinal, and
environ-
mental
Occupational
Occupational
Occupational
Occupational
Occupational
Occupational
Medicinal
Occupational
Occupational
Humans
Target organ
Liver
Bladder
Skin, lung, liver*
Lung, pleura! cav-
ity, gastrointes-
tinal tract
Bladder
Hemopoietlc sys-
tem
Bladder
Lung
Prostate. lungr
Hemopoietic sys-
tem
Lung
Lung, nasal cavi-
ties'
Mam route of
exposure*
P.O.. inhala-
tion0
Inhalation,
skin. p.o.
Inhalation.
P.O., skin
Inhalation.
P.O.
Inhalation.
akin. p.o.
Inhalation.
skin
Inhalation.
skin. p.o.
Inhalation
Inhalation.
P.O.
P.O.. injec-
tion
Inhalation
Inhalation
Animal
Rat
Fish. duck, mar-
moset, tree
shrew, monkey
Rat
Mouse, rat
Mouse
Mouse, rabbit, dog
Newborn mouse
Rat
Mouse, rat. dog
Mouse
Mouse, rat. ham-
ster, rabbit
Rat, hamster
Rat
Mouse, rat
Rabbit, dog
Rat
Mouse
Mouse
Rat
Hamster
Dog
Mouse, rat
Mouse
Rat
Rat
Animals
Target organ
Liver, stomach, co-
lon, kidney
Liver
Liver, trachea
Liver
Local
Lung
Bladder
Liver
Mammary gland, in-
testine
Inadequate, negative
Inadequate, negative
Lung, pleura
Local
Local
Various sites'
Liver
Negative
Local, liver, intestine
Inadequate
Liver
Liver
Zymbal gland, liver.
colon
Liver
Bladder
Lung, nasal cavity
Skin
Local, lung
Local
Local, testis
Route of exposure
p.o.
p.o
i.t.
i.p.
s.c. injection
i.D.
P.O..
s.c. injection
s.c. injection
P.O.
Topical, i.v.
Inhalation or i.t.
Intrapleural
i.p.. s.c. injection
p.o.
p.o.
P.O.
s.c. injection
Topical, s.c. in-
jection
s.c. injection
p.o.
s.c. injection
p.o.
P.O.
Inhalation
Topical
s.c. injection
s.c. injection
s.c. or i.m. injec-
tion
No adequate tests
Mouse
Rat
Mouse, rat
Rat
Initiator
Lung'
Local, lung'
Local'
Local
Lung
Skin
Inhalation
s.c. injection
s.c. injection
s.c.. i.m. injection
Intrabronchial im-
plantation
-------
Table i -Continued
Humans
Animals
Chemical or indus-
trial process
13. Cyclophospha-
mide
Mam type of ex-
posure" Target organ
Medicinal Bladder
Main route of
exposure6. Animal
P.O.. injec- Mouse
tion
Rat
Target organ
Hemopoietic system.
lung
Various sites
Bladder'
Mammary gland
Various sites
Route of
i.p.. s.c.
P.O.
i.p.
i.p.
I.V.
exposure
injection
14. Diethylstilbes- Medicinal Uterus, vagina
trot
P-o.
Mouse
Mouse
Rat
Hamster
Squirrel monkey
Mammary
Mammary, lymphore-
ticular, testis
vagina
Mammary, hypophy-
sis' bladder
Kidney
Uterine serosa
15. Hematite min- Occupational Lung
ing (? radon)
Inhalation
16. Isopropyl oils Occupational Nasal cavity, lar- Inhalation
ynx
17. Melphalan Medicinal Hemopoietic sys- P.O.. injec-
tem tion
18. Mustard gas Occupational Lung, larynx Inhalation
19. 2-Naphthylam- Occupational Bladder Inhalation.
ine skin. p.o.
20. Nickel (nickel Occupational Nasal cavity, lung Inhalation
refining)
Mouse, hamster. Negative
guinea pig
Rat Negative
No adequate tests
Mouse
Rat
Mouse
Hamster,
monkey
Mouse
Rat. rabbit
Initiator
Lung, lymphosarco-
mas
Local
Lung
Local, mammary
dog. Bladder
Liver, lung
Inadequate
21. N,W-Bis(2- Medicinal Bladder
chloroethyl)-
2-naphthy-
lamine
22. Oxymetholone Medicinal Liver
23. Phenacetin Medicinal Kidney
24. Phenytoin Medicinal
p.o.
p.o.
p.o.
Rat Lung
Mouse, rat, ham- Local
ster
Mouse, rat Local
Mouse Lung
Rat Local
No adequate tests
No adequate tests*
p.o.
s.c. injection, s.c.
implantation
Local
s.c. implantation
s.c. injection, s.c.
implantation
s.c. implantation
Inhalation, i.t.
s.c. injection
Skin
i.p.
i.p.
inhalation, i.v.
s.c. injection
p.o.
s.c. injection
p.o.
Inhalation
s.c., i.m injection
i.m. implantation
i.p.
s.c. injection
Lymphoreticular p.o.. injec-
tissues tion
25. Soot, tars, and Occupa- Lung, skin (sere- Inhalation.
oils tional. en- turn) skin
vironmen- -
tal
26. Vinyl chloride Occupational Liver. brain,c Inhalation,
lungc skin
Mouse
Mouse, rabbit
Mouse, rat
Lymphoreticular tis- p.o., i.p.
sues
Skin
Topical
Lung, liver, blood Inhalation
vessels, mammary.
Zymbal gland, kid-
ney
" The main types of exposures mentioned are those by which the association has been demonstrated; exposures other than those
mentioned may also occur.
* The main routes of exposure given may not be the only ones by which such effects could occur.
e Indicative evidence.
* The induction of tumors of the nasal cavities in rats given phenacetin has been reported recently (S. Odashima, personal
communication, 1977).
Tomatie ct al. , 1976. Po.rmi.acl.oxv to xo-p-rodMc-c. ToA»!c. 1. granted by ptibl.'i.oVic.-r Cu.rK.t-i. HC
-------
Table 3.
List of chemicals for which there is some evidence of carcinogenic^ m excenmental animals only or tor which the data were
inadequate for evaluation of the presence or absence of carcinogenicity ((ARC monographs. Volumes 1 to 16
For the 26 compounds evaluated as carcinogenic to humans, see Table 1.
1. Acetamide"1
2. Acridine orange
3. Acriflavinium chlo-
ride
4. Actinomycins*
5. Adriamycin
6. Aldrin
7. Amaranth
8. 5-Aminoaco-
naphthene
9. p-Aminoazoben-
zene*
10. o-Aminoazo-
toluene*
11. p-Aminobenzoic
acid
12. 2-Amino-5-(5-nitro-
2-furyl)-i.3.4-
thiadiazole*
13. 4-Aminc-2-nitro-
phenol
14. Amltrole*
15. Aniline
16. Anthranilic acid
17. Apholate
18. Aramite*
19. Arsenic trioxide
20. Aurothioglucose*
21. Azaserine*
22. Aziridine'
23. 2-(1-Aziridinyl)-
ethanol*
24. Aziridyl benzoqui-
none*
25. Azobenzene*
26. Barium chromate
27. Benz(a)acridine*
28. Benz(c)acridine*
29. Benzo(b)fluoran-
thene*
30. Benzo(/)fluoran-
thene*
31. Benzo(a)pyrene*
32. Benzo(e)pyrene*
33. Benzyl chloride*
34. Benzyl violet 4B*
35. Beryllium*
36. Beryllium oxide*
37. Beryllium phos-
phate*
38. Beryllium sulfate*
39. Beryl ore*
40. BHC (technical
grades)*
41. Bis(t-aziridinyl)-
morpholinophos-
phine sulfide'
42. Bis(chloroethyl)
ether*
43. 1,2-Bis(chloro-
methoxy)ethane*
44. 1.4-Bis(chloro-
methoxymethyl)-
benzene*
45. Blue VRS*
46. Brilliant blue FCF*
47. 1,4-Butanediol di-
methane-sulfo-
nate (Myleran)*
48. 0-Butyrolactone*
49. y-Butyrolactone
50. Cadmium acetate
51. Cadmium chloride*
52. Cadmium powder*
53. Cadmium sulfate'
54. Cadmium sulfide*
55. Calcium arsenate
56. Calcium chromate'
57. Cantharidin*
58. Carbaryl
59. Carbon tetrachlc-
ride*
60. Carmoisine
61. Catechol
62. Chlorambucil*
63. Chlorinated diben-
zodioxins
64. Chlormadinone
acetate*
65. Chlorobenzilate*
66. Chloroform
67. Chloropropham
68. Chloroquine
69. p-Chloro-o-tolui-
dine (hydrochlo-
ride)
70. Cholesterol
71. Chromic chromate*
72. Chromium acetate
73. Chrysene*
74. Chrysoidine*
75. C.I. Disperse Yel-
low 3
76. Cinnamyl anthrani-
late
77. Citrus red No. 2*
78. Copper 8-hy-
droxyquinoline
79. Coumarin*
80. Cycasin*
81. Cyclochlorotine*
82. 2,4-D and esters
83. Daunomycin*
84. D & C Red No. 9
85. Oichlorodiphenyldi-
chloroethane
(ODD)
86. 1.1-Dichloro-2,2-
bis(p-chloro-
phenyl)ethylene
(DDE)
87. DDT*
88. Diacetylaminoazc-
toluene
89. /V.A/-Diacetylbenzi-•
dine*
90. Dial late*
91. 2,4-Diaminoanisole
(sulfate)
92. 4,4'-Diaminodi-
phenyl ether*
93. 1.2-Diamino-4-ni-
trobenzene
94. 1,4-Diamino-2-ni-
trobenzene
95. 2.6-Diamino-3-
(phenylazohpyri-
dine (hydrochlo-
ride)
96. 2,4-Diamino-
toluene'
97. 2.5-Dtammotoluene
(sulfate)
98. Diazepam
99. Diazomethane*
100. Dibenz(a.ftiacn-
dtne*
101. Dibenz(a./)acridine*
102. Dibenz(a,ft)anthra-
cene*
103. Dibenzo(c,g)carba-
zoie*
104. Dibenzo(/>.rsf)pent*
aphene*
105. Dibenzo(a,e)pyrene*
106. Dibenzo(a./»)
pyrene*
107. Dibenzo(a.f)pyrene*
108. Dibenzo(a./)pyrene*
109. 1£-Dibromo-3-
chloropropane*
110. Dibutylnitrosamine"
111. o-Dichlorobenzene
112. p-Dichlorobenze
113. 3,3'-Oichlorobenzi-
dine*
114. rrans-Dichlorobu-
tene
115. 3,3'-Dichloro-4.4'-
diamino-diphenyl
ether*
116. Dieldrin*
117. Diepoxybutane*
118. 1.2-Diethylhydra-
zine*
119. Diethylnitrosamine*
120. Diethyl sulfate*
121. Diglycidyl resorci-
nol ether
122. Dihydrosafrole*
123. Dimethisterone
124. Dimethoxane*
125. 3.3'-Dimethoxyben-
zidine*
126. p-Dimethylamino-
azobenzene*
127. p-Dimetnylaminc-
benzenediazo-so-
dium sulfonate
128. fra/is-2-[(Dimethyla-
mino)methylami-
no]-5-[2-(5-nitro-
2-furyl)vinyl]-
1.3.4-oxadiazole*
129. 3.3'-Dimethylbenzi-
dine*
130. Dimethylcarbamoyl
chloride*
131. 1.1-Dimethylhydra-
zine*
132. 1,2-Dimethylhydra-
zine*
133. Dimethylnitrosa-
mine*
134. Dimethyl sulfate*
135. Dinitrosopentame-
thylenetetramme
136. 1.4-Dioxane*
137. 2.4'-Diphenyldi-
amine
138. Disulfiram
139. Dithranol*
140. Dulcin
141. Endrin
142. Eosin (disodium
salt)
143. Epichlorohydrin*
144. 1-Epoxyethyl-3.4-
epoxycyclohex-
ane*
145. 3,4-Epoxy-6-meth-
ylcyclohexylme-
thyl-3.4-epoxy-6-
methyl carboxyl-
ata*
146. c/s-9,10-Epoxy-
stearic acid
147. Estradiol mustard*
148. Ethinylestradiol*
149. Ethionamide*
150. Ethylene dibrom-
ide*
151. Ethylene oxide
152. Ethylene sulfide*
153. Ethylenethiourea*
154. Ethyl methane-
sulfonate*
155. Ethyl Selenac
156. Ethyl Tellurac
157. Ethynodiol diace-
tate*
158. Evans blue*
159. Fast green FCF*
160. Ferbam
161. 2-(2-Formylhydra-.
zmo)-4-(5-nitro-2-
furyl)thiazole*
162. Fusarenon-X
163. Glycidaldehyde*
164. Glycidyl oleate
165. Glycidyl stearate
166. Griseofulvin*
167. Guinea green B*
168. Heptachlor
169. Hexamethylphos-
phoramide*
170. Hycanthone (mesy-
late)*
171. Hydrazine*
172. Hydroquinone
173. 4-Hydroxyazoben-
zene
174. 8-Hydroxyquinoline
175. Hydroxysenkirkine
176. Indenod .2.3-
cd)pyrene*
177. Iron dextran*
178. Iron dextrin*
179. Iron oxide
180. Iron-sorbijol-citric
acid complex
181. Isatidine*
182. Isonicotmic acid
hydrazide*
183. Isopropyl alcohol
184. Isosafrole*
185. Jacobine
186. Lasiocarpine*
187. Lead acetate*
188. Lead arsenate
189. Lead carbonate
190. Lead chromate
191. Lead phosphate*
192. Lead subacetate*
193. Ledate
194. Light green SF*
195. Lindane*
196. Luteoskyrin*
197. Magenta*
198. Maleic hydrazide*
199. Maneb
200. Mannomustine (di-
hydrochloride)*
201. Medphalan
202. Medroxyprogester-
one acetate*
203. Merphalan*
204. Mestranol*
205. Methoxychlor
206. 2-Methylaziridine*
207. Methylazoxymetha-
nol acetate*
208. Methyl carbamate
209. A/-Methyl-A/,4-dini-
trosoaniline*
210. 4.4'-Methylenebis(2-
chloroaniline)*
211. 4.4'-Methylenebis(2-
methylaniline)*
212. 4.4'-Methylenedi-
aniline
213. Methyl iodide*
214. Methyl methane-
sulfonate*
215. A/-Methyl-W'-nitro-
/V-nitrosoguani-
dine*
216. Methyl red
217. Methyl Selenac
218. Methylthiouracil*
219. Metronidazole*
220. Mirex*
221. Mitomycin C*
222. Monocrotaline*
223. Monuron*
224. 5-(Morpholino-
methyl)-3-[(5-
rtitrofurfuryli-
dene)-amino]-2-
oxazolidinone*
225. 1-Naphthylamine*
226. Native carrageen-
ans*
227. Nickel carbonyl*
228. Nickelocene*
229. Nickel oxide*
230. Nickel powder*
231. Nickel subsulfide*
232. Nirtdazole*
233. 5-Nitroace-
naphthene*
234. 4-Nitrobiphenyl*
235. Nitrofuraldehyde
semicarbazone
236. 1-[(5-Nitrofurfuryli-
dene)amino]-2-
imidazolidinone*
237. W-[4-(5-Nitro-2-
furyl)-2-thiazolyl]-
acetamide*
-------
Table 7 -Continued
238.
239.
240.
241.
242.
243.
244.
245.
246.
247.
248.
249.
250.
251.
252.
253.
254.
255.
256.
257.
258.
259.
260.
Nitrogen mustard
(hydrochloride)*
Nitrogen mustard
N-oxide (hydro-
chlorideV
Nitrosoethylurea*
Nitrosomethylurea*
N-Nitroso-A/-methyl-
urethan*
Norethisterone*
Norethisterone ace-
tate*
Norethynodrel*
Norgestrel
Ochratoxin A
17/3-Oestradiol*
Oestriot
Oestrone*
Oil orange SS*
Orange 1*
Orange G
Oxazepam*
Oxyphenbutazone
Parasorbic acid*
Patulin*
Penicillic acid*
Phenicarbazide*
Phenobarbital so-
dium*
261.
262.
263.
264.
265.
266.
267.
268.
269.
270.
271.
272.
273.
274.
275.
276.
277.
278.
Phenoxybenzamine*
Phenylbutazone
m-Phenylenedi-
amine (hydro-
chloride)
p-Phenylenedi-
amine (hydro-
chloride)
N-Phenyl-2-
naphthylamine*
Polychlorinated bi-
phenyls*
Ponceau MX*
Ponceau 3R*
Ponceau SX
Potassium arsenite
Potassium bis(2-hy-
droxyethyl)-di-
thiocarbamate*
Progesterone*
Pronetalol hydro-
chloride*
1 ,3-Propanesul-
tone*
Propham
/3-Propiolactone*
n-Propyl carba-
mate*
Propylene oxide*
279.
280.
281.
282.
283.
284.
285.
286.
287.
288.
289.
290.
291.
292.
293.
294.
295.
296.
297.
298.
299.
300.
301.
302.
303.
Propylthiouracil'
Pyrimethamine*
p-Quinone
Oumtozene'
Reserpine
Resorcmol
Retrorsine*
Rhodamme B"
Rhodamme 6G*
Riddelhine
Saccharated iron*
Safrole*
Scarlet red
Selenium com-
pounds
Semicarbazide (hy-
drochloride)*
Seneciphylline
Senkirkine
Sodium arsenate
Sodium arsenite
Sodium bichromate
Sodium diethyldi-
thiocarbamate
Sterigmatocystin*
Streptozotocin*
Strontium chrc-
mate*
Styrene oxide
304.
305.
306.
307.
308.
309.
310.
311.
312.
313.
314.
315.
316.
317.
318.
319.
320.
321.
322.
323.
324.
325.
Succinic anhy-
dride*
Sudan 1*
Sudan II*
Sudan III
Sudan brown RR
Sudan red 78
Sunset yellow FCF
2.4.5-T and esters
Tannic acid*
Terpene polychlori-
nates*
Testosterone*
Tetraethyl & tetra-
methyl lead
Thioacetamide*
4.4'-Thioaniline*
Thiouracil*
Thiourea*
Thiram
o-Toluidine (hydro-
chloride)
Trichloroethylene*
Trichlorotriethy-
lamine hydro-
chloride
Triethylene glycol
diglycidyl ether*
Tris(aziridinyl)-p-
326.
327.
328.
329.
330.
331.
332.
333.
334.
335.
336.
337.
338.
339.
340.
341.
342.
benzoquinone*
Tris(l-aziridinyl)-
phosphine oxide
Tris(l-aziridinyl)-
phosphine sul-
fkJe*
2.4.6-Tris(1-aziridi-
nyl)-s-triazine*
1 ,2,3-Tris(chloro-
methoxy)-pro-
pane*
Tris(2-methyl-1-
azirWinyl)phos-
phine oxide
Trypan blue*
Uracil mustard*
Urethan*
Vinyl cyclohexane
2.4-Xylidine (hydro-
chloride)
2.5-Xylidine (hydro-
chloride)
Yellow AB
Yellow OB*
Zectran
Zinc chromate hy-
droxide*
Zineb
Ziram
• Asterisk, chemicals for which there is some evidence of carcinogenicity in experimental animals only.
Tomatis et al., 1978. Permission to reproduce Table 2 granted by publisher Cancejr Research
-------
Pollutants
Disease
Aldehydes
Aldrin/Dieldrin
Arsenic
Asbestos
Benzene
Benzidine
Benzo-a-Pyrene
Beryllium
Cadmium
Calcium Chromate
Carbon Monoxide
DDT
Lead
Nickel Carbonyl
Oxidants
Oxides of Nitrogen
Particulates
Silica
Sulfates
Sulfur Oxides
Vinyl Chloride
Cancer
5r» Heart
• Lung
Figure 1. Known or Suspected Links Between Selected Pollutants
and Disease (Task Force on Environmental Cancer and Heart and
Lung Disease, 1978). Permission to use figure granted by EPA.
-------
DIMETHYLNITROSAMINE
N-METHYL-N-NITROSOUREA
CH-
CH.
N-NO
E.R.
NADPH
CH.
CH,
j.
H2N
~K
i n o
";"]
non
DNA, RNA, PROTEIN
CH3-DNA, CH3-RNA WITH O -CH3-G, 7-CH3-G, 3-CH3-A, ETC.
CH3-PROTEIN WITH 1- and 3-CH3-Histidine, S-CH3-Cysteine, ETC
Figure 2.: The in vivo conversion of dimethylnitrosamine
and of N-methyl-N-nitrosourea to a reactive electrophile
and its reaction with cellular macromolecules.
Miller, 1978. Permission to use Figure 2 granted by publisher of Cancer Research.
-------
II. Scientific Aspects of the Proposed Chronic Health
Effects Test Standards
A. Overview of Proposed Standards
1. Objectives
2. Need for Standards
3. Summary of Standards
B. Study Design Issues
1. Introduction
2. Test Species, Strain, and Sex
Objective
a. Test Species
i. Oncogenicity Studies
ii. Non-oncogenic Chronic
Toxicity Studies
iii. Combined Chronic Toxicity
Studies
iv. Alternative Species
b. Test Strain
c. Sex
3. Age at Start of Test
Objective
a. Weanlings
b. In Utero
c. Neonatals
4. Number of Animal/Test Group
Objective
a. Number
b. Randomization
5. Number of Dose Levels and Dose
Selection
Objective
a. Oncogenicity Studies
b. Non-oncogenic Chronic Toxicity
Studies t.
c. Combined Toxicity Studies
-------
6. Controls
Objective
a. Matched Controls
b. Positive Controls
c. Historical Controls
7. Route(s) of Exposure
Objective
Discussion
8. Period of Exposure and Observation
Objective
a. Period of Exposure
i.Oncogenicity Test Standards
ii. Non-oncogenic Chronic
Toxicity Test Standards
A. Rodents
B. Nonrodents
b. Period of Observation
C. Study Conduct Issues
1. Introduction
2. Clinical Procedures
Objective
a. Clinical Observations
b. Clinical Chemistry
3. Pathology Procedures
Objectives
a. General
b. Gross Necropsy
c. Microscopic Examination
D. Data Collection and Reporting Issues
1. Final Report
2. Interim Reports
E. Good Laboratory Practice Issue
1. Introduction
2. Personnel
3. Animal Care and Facility
4. Dietary Requirements
-------
Objective
a. Diet
b. Standardization of Diet
5. Contaminant Analysis Requirements
Objective
Discussion
6. Safety and Health Standards
F. References
-------
II. Scientific Aspects of the Proposed Chronic Health
Effects Test Standards
A. OVERVIEW OF TEST STANDARDS
1. Objectives
Method of choice in a chronic toxicity study
depends on objective(s) of the study and intended use of
results (Weisburger and Weisburger, 1967; Arcos, Argus,
and Wolf, 1968; NAS, 1975; Page 1977a; WHO, 1978a).
Based on this, three general types of chronic toxicity
studies have evolved: (1) studies to determine whether
one or a combination of chemicals possesses toxic
activity; (2) studies to determine structure-activity
relationships; and (3) studies to define the
mechanism(s) of action (Arcos, Argus, and Wolf, 1968).
The objective of chronic toxicity test standards to be
promulgated under section 4(b) of TSCA is to test a
variety of high concern, individual environmental
chemicals for chronic toxicity . The ultimate use of
test data will be to help evaluate possible risks to
humans due to these chemicals.
Part 772, Subpart D, Chapter I of Title 40 of the
Code of Federal Regulations will prescribe three chronic
health effects test standards: Section 772.113-2 will
prescribe oncogenic effects test standards; Section
772.113-3 non-oncogenic chronic effects test standards;
and Section 772.113-4 combined chronic effects test
standards. The main objective of studies carried out
under Section 772.113-2 will be to determine the
oncogenic potential of chemicals; that of Section
772.113-3 will be to determine chronic toxicity
potential other than oncogenicity, and that of Section
772.113-4 will be to determine any chronic toxicity
potential including oncogenicity.
The Agency is proposing combined chronic effects
standards because it anticipates that certain chemicals
will be required to undergo both oncogenicity and non-
oncogenic chronic toxicity testing. In order to maximize
efficient use of test animals, laboratory facilities and
personnel resources, the combined test may be used
I/ The proposed test standards are not designed to study
synergistic effects such as promotion and cocarcinogenesis,
structure-activity relationships or mechanism(s) of action.
However, because data on such effects and relationships are
relevant to the objectives of chronic toxicity testing, if such
data are developed, they must be submitted to EPA. Such data may
be required for certain chemicals as specified in specific
Section 4(a) test rules.
-------
II-2
instead of separate studies for oncogenic and non-
oncogenic chronic effects. The proposed combined test
meets the standards necessary for assessing both onco-
genicitv/non-oncogenic chronic toxicity (see parts B and
C of this section) while at the same time decreasing the
cost and resource needs of such testing. As discussed
in Section III of this document, an oncogenic effects
test is estimated to cost approximately $40/0,0001'; a
non-oncogenic effects test about $500,OQO1'; and a
combined chronic effects test $800,OOO1'. Thus there
will be a savings of around $150,000 if both types of
effects must be determined and the combined chronic
effects test is used.
2. Need for Standards
Under Section 4 of TSCA, it is the
responsibility of the Agency to assure that sufficient
data are developed and in a manner and quality so that
hazard identification and risk assessment can be
performed on chemicals identified in Section 4(a) test
rules. In order to assure this, TSCA Section 4(b)
specifies that each test rule must include standards for
development of test data. This requirement itself is
controversial in some quarters and is predicated on the
fact that all chemicals cannot be completely evaluated
by the same procedure (Hayes, 1975"; Loomis, 1974).
After reviewing "state-of-the-art" in chronic
toxicity testing, EPA agrees that total standardization
of test procedures, even for studies with the same
objective, is neither possible nor desirable. However,
EPA believes that certain minimum standards can and must
be established for certain factors of study design and
conduct common to all chronic toxicity studies with the
same objective. Based on historical experience, such
test standards would define the minimum chronic toxicity
testing requirements needed for hazard evaluations and
would assure that adequate and reliable data are
developed for TSCA proposes. Therefore, the standards
presently proposed under Section 4(b) of TSCA would
serve to standardize only those aspects of design and
conduct of chronic toxicity studies which EPA finds to
be essential to their scientific acceptability. Their
intent is not to structure toxicological investigation
so as to stifle original research or to prevent
'serendipitous1 discoveries.
I/ These estimated costs do not include the costs for prechronic
toxicity studies.
-------
II-3
Other advantages of such standards would be a
better ability to compare studies and results on the
same chemical or group of chemicals, a help in
facilitating assessment and comparison of all studies
including those from foreign countries, and a help in
facilitating cost estimates (Berenblum, 1969; Page,
1976; Page 1977a).
3. Summary of Standards
An ideal test system would present a sensitive,
reliable, and specific tool for detection of all
possible long-term toxic effects of chemicals (J. H.
Weisburger, 1976). It would also be economical, fast,
foolproof, and simulate the human situation. Although
presently available long-term animal studies do not
fully meet these ideals (Tomatis, 1977; Shimkin, 1977),
there is general agreement that such studies are the
most reliable indicators of a chemical's chronic
toxicity potential (D'Aguanno, 1974; Page, 1977a-b;
Rail, 1977). For example, no shortened tests for
oncogenicity can currently be substituted for lifetime
studies when it comes to establishing 'the absence of
oncogenic risk (WHO, 1969; WHO, 1978b).
The chronic toxicity test standards proposed for
use in Section 4(a) test rules represent minimum
requirements necessary to conduct an adequate study for
hazard evaluation and are generally consistent with
those recommended by other Federal (NAS, 1977; FIFRA,
1978; Sontag, Page, and Saffiotti, 1976; FDA, 1971; Ad
Hoc Committee on the Evaluation of Low Levels of
Environmental Chemical Carcinogens, 1970) and interna-
tional (Canada, MHW, 1975; WHO, 1978b) agencies. They
are designed to assure that TSCA test data will be
developed in conformity with the following long-
established criteria for chronic toxicity testing: (1)
use of sensitive and reliable animal test systems; (2)
optimal exposure conditions to reveal chronic toxicity;
(3) elimination of extraneous factors that might
influence conduct of the test and interpretation of
results; (4) in-depth pathology examination to detect
minute as well as more obvious adverse changes; and (5)
complete documentation of all data to allow those
responsible for interpretation of human relevance to
make the best judgments possible (Page, 1977b).
-------
II-4
Briefly, the main aspects of the proposed TSCA
test standards are as follows:
(1) The tester must use both sexes of two
mammalian species, usually rat and mouse, for the
oncogenicity studies; two mammalian species, usually rat
and dog, for non-oncogenic chronic toxicity studies; and
three mammalian species, usually rat, mouse, and dog,
for combined chronic toxicity studies;
(2) The tester must begin to dose animals as
soon as possible after weaning and environmental
acclimatization but no later than six (6) weeks of age
for rodents and ten weeks of age for dogs;
(3) Each rodent group must contain at least 50
animals for oncogenicity studies and 58 animals for non-
oncogenic and combined chronic toxicity studies; each
non-rodent group must contain at least 6 animals;
(4) For oncogenicity studies, the tester must
use at least three (3) dose levels (in addition to
controls). The high dose level (HDL) is the maximum
dose level that can be administered for the duration of
the test period, with demonstrable but only slight
toxicity in test aninmals, and no substantial reduction
in longevity due to effects other than tumors; the
second dose level is a specified level (1/4 to 1/2) of
the HDL; and the third dose level is to be no more
than 1/2 of the second dose level and no less than 1/10
of the HDL.
For non-oncogenic chronic toxicity studies, the
tester must use at least three (3) dose levels (in
addition to controls). The high dose level (HDL) must
induce chronic effects including mortality; the low dose
must not induce any observable evidence of adverse
effects (NOEL); and the middle dose must be appro-
priately spaced to demonstrate a "dose-response"
relationship.
For combined chronic toxicity studies the tester
must use at least three (3) dose levels (in addition to
controls) for the mouse and dcg and at least four (4) or
five (5) dose levels (in addition to. controls) for the
rat. The three dose levels for the mouse are those
specified in the oncogenicity effects test standards and
for the dog those specified in the non-oncogenic chronic
effects test standards. For the rat, the high dose
level is the specified HDL for the non-oncogenic chronic
effects test standards; the next three dose levels are
those specified for the oncogenic effects test
-------
II-5
standards; and the fifth dose level must not induce any
observable evidence of adverse effects (NOEL). If one
of the dose levels being specified from the oncogenic
effects test standard is predicted to induce no adverse
effects other than tumors, i.e., be a NOEL, a fifth dose
is not required;
(5) The tester must use a matched control group
identical in every respect to the exposed groups except
for exposure to test substance;
(6) Route(s) of administration is (are),
whenever possible, to be comparable to expected or known
•route(s) of human exposure;
(7) The tester must administer the test
substance to mice for a minimum of 24 months but no
longer than 30 months and to dogs for a minimum of 24
months. The tester must administer the test substance
to rats for a minimum of 24 months but no longer than 30
months for oncogenicity studies and for a minimum of 30
months for non-oncogenic and combined chronic toxicity
studies;
(8) The tester must feed test animals specified
standardized diets and analyze feed and vehicle, if any,
for certain specified contaminants;
(9) Appropriately trained employees must observe
all animals at least every 12 hours throughout the test
period;
(10) Technical employees must weigh and
clinically examine each animal at least once each week
during the first 13 weeks of the study and every 2 weeks
thereafter. Certain quantitative clinical chemistry
determinations including hematology, blood chemistry,
urinalysis, function tests, and residue analysis are to
be made on a minimum of eight (8) predesignated rodents
in each test group and on all non-rodents. The only
determinations to be made in oncogenicity studies are
hematology determinations at one year and at study
termination. For non-oncogenic and combined chronic
toxicity studies, all of the clinical chemistry tests
are to be performed at least at 3, 6, 12, 18, 24 months,
and at study termination.
(11) The tester must conduct a detailed necropsy
and histopathology examination of all animals with
approximately thirty (30) to forty (40) tissues
routinely examined microscopically;
-------
Il-fi
(12) The sponsor must submit to EPA a full and
detailed report of test conditions, of all observations
made on test animals, and of any data analysis
conducted;
(13) The tester must conduct all studies
according to specified good laboratory practice
standards.
Exact and complete details of these test standards
are to be found in the proposed test standards
themselves.
In the discussions to follow, the Agency outlines
the scientific or other basis for each of the standards
proposed and discusses the major issues associated with
the development and use of these standards.
B. DESIGN ISSUES
1. Introduction
A study to determine toxic effects of a
chemical consists of three elements—study design, study
conduct, and data analysis. Each plays an essential
role in the final outcome and usefulness of long-term
animal studies. A properly designed study ensures that
an adequate data base is generated during the study to
meet its objectives; a properly conducted study ensures
the quality of the data base generated, and a properly
analyzed study enables a decision to be made as to
chronic toxicity potential of the test substance under
the given test conditions. The following is a discus-
sion of aspects involved in proper design of a chronic
toxicity study.
2. Test Species, Strain, and Sex
Objective: The standards for selection of
test species, strain and sex are set forth to ensure
that sensitive and reliable animal test systems are
selected.
a. Test Species
i. Oncogenicity Studies. Testing is required
in at least two mammalian species, usually the rat and
the mouse. Present knowledge indicates that development
of chemically induced neoplasia is the result of a
complex series of biological interactions which are
subject to and controlled by numerous endogenous and
exogenous modifying factors (Hueper and Conway, 1964; J.
-------
II-7
H. Weisburger, 1973). Because of this complex patho-
genesis, no one species of animal can be predicted to
give biological responses similar to those of humans.
As discussed in detail below, certain scientific,
technical, and economical considerations, thus, become
the deciding factors in selecting appropriate test
animals. Included among these factors are the life span
of the species, susceptibility to tumor induction by
chemical oncogens, reliability as a model for the human,
quality of the animals in terms of health and stamina,
physical size, genetic stability and reproducibility,
availability in sufficient number, purchase cost,
maintenance costs, knowledge of spontaneous tumor
incidence, and availability of historical information
'(FDA, 1959; NAS, 1961; Clayson, 1962; Weisburger and
Weisburger, 1967; Canada, MHW, 1975; Page, 1977a; WHO,
1978b; FSC, 1978).
One of the most important considerations in
selecting animals species for evaluation of potential
chemical oncogens is the normal life span of the
animals. Because tumors usually develop only after a
long latent period, (e.g., humans 5 to 40 years), it is
usually necessary to extend the period of testing over
the greater part of or the entire life span of the
animals (Zwickey and Davis, 1959; NAS, 1961; Magee,
1970; NAS, 1975; Canada, MHW, 1975; Page, 1977b; WHO,
1978b; FSC, 1978). This latency period depends on the
life span of the animal and the potency of the
chemical. Even with potent chemical oncogens, it takes
3 minimum of 1/8 to 1/4 of the life span of a given
species for chemically induced tumors to develop
(Weisburger and Weisburger, 1967; D'Aguanno, 1974). For
this reason, species with long life spans such as the
rabbit (7 years), dog (> 10 years) and monkey (> 10
years) are generally not used for routine oncogenicity
testing (Arcos, Argus, and Wolf;, 1968). Comparing the
latent period for induction of neoplasms with
benzo(a)pyrene in a range of animal species, it was
noted that while the rat, mouse, guinea pig, and rabbit
all responded within two years, it required up to ten
years to develop in monkeys (Hartwell, 1951; E.K.
Weisburger, 1971; Canada, MHW, 1975). Although the dog
is relatively unique in that it shows the same organ
specificity to aromatic amine carcinogenesis as the
human, it is not uncommon for the latent period to be
five to ten years in these animals (Canada, MHW, 1975;
J.H. Weisburger, 1975; Stula, Barnes, ^t a±., 1978a-b).
Susceptibility of species to particular chemical
oncogens, classes of chemical oncogens, or routes of
exposure must also be considered in selection of
-------
II-8
appropriate test species (Hueper and Conway, 1964;
Shimkin, 1974; NAS, 1975; J. H. Weisburger, 1976;
Sontag, 1977). Species specificity reflects the
definite and sometimes decisive role which the host
organism plays in controlling the response to oncogenic
exposure (Hueper and Conway, 1964; J. H. Weisburger,
1976). For example, guinea pigs show an apparent
resistance to a wide variety of chemical oncogens (Page,
1977b). Most strains of rats are resistant to induction
of tumors by 2-naphthylamine, a potent oncogen in other
species including humans (J. H. Weisburger, 1973). Rat
skin is relatively resistant to hydrocarbon oncogens
while mouse skin is susceptible; therefore, the mouse is
the species of choice when testing this class of
chemicals (Clayson, 1962).
Not only must a species be sensitive to tumor
induction by chemical oncogens, but it must also be a
reliable model for the human. If extrapolation of data
to humans is to be meaningful, a species must have a low
potential for generating both false positive and false
negative results. Of the 26 chemicals known to be
oncogenic in humans, all but three, arsenic, benzene,
and phenacetin-containing analgesics, have induced
tumors in some animal model; the majority showing
effects in rat and mouse (Clayson, 1978b; Tomatis,
Agthe, et_ al_. , 1978). Of 56 other chemicals suspected
of being oncogenic in humans, all have tested positive
in animal studies (Maugh, 1978). While it is possible
to estimate the false negative rate (3/26) for long-term
animal studies, it is not possible to estimate a false
positive rate because of the complexity of determining
which chemical oncogens are active in humans (see
discussion in Section I).
Other factors also influence the choice of species
to be used. Since test animals must be maintained for
the greater part of or for their entire life span, it is
important that the animals be of high quality in regard
to their health, vigor, and stamina. Ideally,
spontaneous disease rates should not exceed 5 percent
(FSC, 1978). Because proper evaluation of data requires
that large numbers of animals be used (see Section
B.4.), appropriate species should be small in size for
maintenance and cost purposes, be available in
sufficient numbers, be genetically stable and readily
reproduce. Proper evaluation of data also dictates
that the species have a low spontaneous tumor rate with
these tumors appearing as late in life as possible.
Information regarding the susceptibility to chemically
induced tumors, the type and incidence of spontaneous
diseases, and the metabolism and pharmacokinetics of the
-------
II-9
test substance also aids in the selection of appropriate
species.
At this time, it is generally agreed that the only
mammalian species that adequately meet these selection
criteria are rodents, especially the rat and mouse
(Barnes and Denz, 1954; FDA, 1959; WHO, 1961; Berenblum,
1969; Magee, 1970; Peck, 1974; NAS, 1975; Canada, MHW,
1975; Sontag, Page and Saffiotti, 1976; J.H. Weisburger,
1976; NAS, 1977; Sontag, 1977; WHO, 1978b)- These two
species have a relatively short life span, are suscep-
tible to chemically induced oncogenesis and are small in
size, readily available, relatively cheap, fertile,
hardy, and well-standardized and studied.
Since it is not presently possible to predict
which species will give oncogenic responses similar to
those of humans and since there are many examples of .
species variation to oncogens on record among rodents,
it is considered necessary to test a chemical in more
than one species (Shubik and Sice, 1956; FDA, 1959;
Delia Porta, 1963; Berenblum, 1969; FDA, 1971; NAS,
1975; Page, 1977b, Sontag, 1977; WHO, 1978b; FSC,
1978). The more species a chemical is tested in, the
less chance there is of missing an oncogenic response
due to metabolic or other variables (Sontag, 1977). Due
to economic, space, and time considerations, however,
more than two species are rarely used1/. Testing in one
rodent species costs approximately $200,000 and requires
4-5 years to complete.
Testing in more than one species -not only
increases the chance of detecting an oncogenic response
but also increases the confidence in extrapolating to
predict carcinogenic potential t;o humans. Although a
positive oncogenic effect in one species is considered
adequate evidence of a potential hazard to humans,
positive results in more than one species verify the
potential hazard and give an indication of the relative
potency of the chemical in different species. Although
no unqualified negative answer is ever possible, a lack
of evidence of tumor induction in at least two species
is generally regarded as the most reliable criterion of
non-oncogenicity (WHO, 1969; WHO, 1978b). This is very
important since epidemiological data from humans is
never adequate to define a negative oncogen (Clayson,
1978a).
I/ Testing in only two species, of course, is not foolproof as
demonstrated by the inability to induce tumors in animals with
arsenic, known to be oncogenic to humans (Bencko, 1977).
-------
11-10
ii. Non-oncogenic Chronic Toxicity
Studies. Testing is required in at least two mammalian
species, one the rat and the second a non-rodent,
usually the dog.
(A) Need for Two Species
The requirement for two species in non-
oncogenic chronic toxicity testing is due to the uncer-
tainties of extrapolating to humans from a single
surrogate for such effects (Barnes and Denz, 1954;
Benitz, 1970; Loomis, 1974; Hayes, 1975). As in the
case of oncogenicity, significant differences in
susceptibility and mode of action for non-oncogenic
chronic toxicity exist between species and have been
identified as due to differences in distribution,
metabolism, and cytology (Albert, 1973). McConnel,
Moore, et_ jl^. (1978) reported a positive correlation
between liver burden of TCDD in mice, rats, guinea pigs,
and monkeys and the degree of toxicity induced in these
species. Williams (1978) describes numerous species
variations in detoxification mechanisms. Hayes (1975)
cites species differences in toxicity to norbormide of
230 fold and to thalidomide of 1,000 fold. Hodge,
Smith, et al. (1963) reports a 250 fold difference in
sensitivity to fluoroacetate among mammals and if the
toad is included, a 25,000 fold difference. In
addition, cause of death from fluoroacetate differs
between species; the dog displays central nervous system
effects, the rabbit shows cardiac effects, and man and
monkey show a mixed response. Casida and Baron (1976)
cite marked species differences in susceptibility and
clinical signs of neurotoxicity due to organophosphorus
compounds. In this case, humans and mature hens appear
to respond similarly; the dog, cat, pig, cow, sheep, and
horse, though susceptible, show different patterns of
toxicity; and the rat appears to be refractory. These
differences in species susceptibility demonstrate the
weaknesses inherent to relying on only one animal model
for extrapolation to humans for non-oncogenic chronic
toxicity effects.
(B) Need for a Rodent and a Non-Rodent
Among laboratory rodents, mice and rats remain
the species of choice under most general circumstances
for chronic toxicity studies (FSC, 1978). As discussed
in the section on selection of species for oncogenicity
testing, the life span, small size, ready availability,
cost, fertility and hardiness of rats and mice and the
fact that they are well-standardized and studied make
-------
11-11
them suitable for chronic toxicity testing. Of these
two species, the rat was chosen as the species of
general choice for the non-oncogenic chronic toxicity
studies because of the enhanced capability to conduct
clinical evaluations with larger-sized animals and
because of the more extensive work done with this
species. It is felt that blood collection in the mouse
is too stressful and that blood volume is too limited.
Nephroses, chronic pulmonary disease, hypertension
and deposition of lipids in the aorta, susceptibility to
infections, senility as reflected by morphological
changes in the CNS associated with aging, growth, and
survival in terms of the median life span have been
described in laboratory rats. Although differences in
long-term pathogenesis between man and rat have been
identified (e.g., arteriosclerosis) the qualitative
similarities are striking.
While the use of a rodent in chronic toxicity
testing is non-controversial, the use of a non-rodent is
highly controversial. Those challenging the need for a
non-rodent contend that nearly all chronic effects will
be observed in the rodent (Aviado, 1978). Examination
of the available data, however, shows important differ-
ences. For example, the report by Vettorazzi (1975)
shows differences in response between the rat and dog in
two year feeding studies with binaparcyl, amitrole,
dichlofluanid, thiabendazol, and dodine. Hodge, Downs,
et_ _al^ (1968) found similarities in the toxicities of
monuron and diuron while linuron showed species
differences between rat and dog. This latter example
shows that even where similarities exist between species
for one chemical differences may exist for related
substances.
EPA has decided to require use of a non-rodent in
chronic toxicity studies for three reasons; First, EPA
believes it is necessary for Section 4(a) purposes to
utilize as efficient a test system as practical to
detect all possible chronic effects. Since nononcogenic
chronic effects are sometimes detectable in non-rodents
but not in rodents (or vice versa), it is felt that both
types of species must be used. Second, use of the non-
rodent will enhance the capability to conduct more
precise clinical evaluations. Third, since the Agency
will likely propose exposure limits for certain
chemicals tested under Section 4 where positive test
results are found, dose-response data with two species
will provide a firmer scientific basis for risk
estimations. Generally, dose-related effects obtained
with larger animals more nearly approximate those
-------
11-12
expected in humans. For example, Litchfield (1961) in
an analysis of 39 signs of toxicity of six drugs in the
rat, dog and human found the dog to be a better
predictor of human effects than the rat.
Dogs have been extensively used as the usual non-
rodent species of intermediate life span between humans
and the common laboratory rodents. They are of a
convenient size and disposition for clinical examina-
tions, and can be obtained as purebreds if genetic
similarity is desired. The cost of using dogs
($226,400) instead of another rodent species ($257,300)
is insignificant. ' Because of the important
differences in information obtained from using both a
rat and dog arid because cost of such studies are
equivalent, EPA believes that a two year dog study is
feasible and necessary to adequately determine chronic
toxicity.
Non-human primates have also been promoted as
surrogates for humans based primarily upon their
phylogenetic relationship and similarity in physio-
logical functions. However, they are generally
difficult and expensive to obtain in a defined (disease-
free) condition with most animal sources from foreign
countries having restricted exportation quotas. For
routine testing, the Advantages of non-human primates
over dogs do not offset the procurement, cost, and
management problems involved in their use.
iii. Combined Chronic Toxicity Studies. Testing
is required in at least three mammalian species, two
rodents, usually the rat and mouse, and a non-rodent,
usually the dog.
As discussed in detail in the previous two
sections, at least two species, usually the mouse and
rat, must be used to adequately test a chemical for
potential oncogenic effects and two species, usually the
rat and dog, must be used to adequately test a chemical
for potential non-oncogenic chronic effects. Based on
these assessments, EPA believes that to adequately
assess both oncogenic and non-oncogenic chronic effects
in one test all three species must be used. The mouse
is used to detect oncogenic effects, the dog non-
oncogenic chronic effects, and the rat both types of
effects. Use of the dog to detect oncogenic effects is
not appropriate because the duration of the study is too
short for tumor development in a long-lived species such
/For cost estimates see section III.
-------
11-13
as the dog except for the most potent of oncogens.
(iv) Alternative Species. The Agency recognizes
that the mouse, rat and dog are not always the most
appropriate species in which to test certain chemicals.
Because of this, EPA allows use of alternative species.
The main objective of the test standards is not only to
show toxic effects but to obtain the most meaningful
data for extrapolation to the human situation. Any
procedure which allows this to be accomplished will be
accepted by EPA if an appropriate rationale can be shown
to support it . Acceptable rationale for use of another
species would be to show that is a better model for a
given chemical because of its known sensitivity to a
given class of chemical toxicants or because of its
known metabolic or pharmacokinetic handling of the
chemical.
b. Test Strains
No concensus exists in the scientific community
as to the most appropriate type of strain to use,
outbred, inbred or hybrid (NCTR/NCI, 1979). Because of
this, EPA has decided not to require use of either type
but leaves the decision to the tester.
Outbred stocks are animals that are maintained by
a breeding system that tends to minimize inbreeding and,
thus, maximize genetic variation (ILAR, 1976). Inbred
strains are animals that have been brother-sister mated
for 20 generations or its equivalent (Clayson, 1962;
NAS, 1974). Genetic variation within an inbred strain
is minimal. Hybrids are animals resulting from a cross
between two inbred strains (ILAR, 1976). Use of any one
of the three types of strains has its advantages and
disadvantages.
-------
11-14
Arguments over use of inbred or hybrid strains
versus outbred stock characterize the two possible
approaches to long-term testing, namely to simulate the
human situation or to maximize sensitivity of the test
model (Tomatis, 1977). Certain researchers and scien-
tific committees recommend use of outbred stock because,
like humans, they are genetically heterogenous (Delia
Porta, 1963; Arcos, Argus, and Wolf, 1968? PDA, 1971;
NAS, 1975; Canada, MHW, 1975). Because of their
heterogeneity, it is believed that at least a few
animals will respond if a chemical has activity and that
extrapolation of the data to humans is more meaningful.
In reality, compared to humans, outbred strains have
narrow genetic bases because each colony has a
restricted and self-limiting gene pool (Sontag, 1977).
Compared to inbred strains, the outbred stocks have a
wider genetic base and generally longer life spans, and
are more resistant to infections (Sontag, 1977; FSC,
1978). The major disadvantage with using outbred stocks
is their fluctuating incidence of spontaneous diseases
including neoplastic lesions and variable end-response.
Because of this, larger numbers of animals than would be
the case for inbreds are needed to secure statistically
significant results (Weisburger and Weisburger, 1967).
Other researchers and scientific committees
recommend use of inbred or hybrid strains because of
their genetic homogeneity (Zwickey and Davis, 1959;
Weisburger and Weisburger, 1967; Berenblum, 1969;
Testing, 1975; Sontag, Page and Saffiotti, 1976; FSC,
1978). This genetic stability of inbred and hybrid
strains allows for greater uniformity of response,
highly predictive spontaneous disease and tumor
incidences and better reproducibility of test results.
However, their homogeneity may also be their major
disadvantage in that they are more likely to be totally
resistant to certain potentially toxic chemicals than
are outbred animals (Clayson, 1962; Sontag, 1977).
Hybrids have the additional advantage over inbred
strains in that they are more vigorous, less disease
prone and have a longer life expectancy. Their
disadvantage is that they are more difficult to obtain
and, therefore, tend to be more expensive.
c. Sex
Because of known sex differences in response
and possible target organs, the tester must use both
sexes to show the full range of activity of a test
substance. Sex differences in response to chemical
toxicants are well documented in both humans (American
Cancer Society, 1978) and test animals (Bock, 1964; J.
-------
1 I-J.O
H. Weisburger, 1975; Canada, MHW, 1975). Sex-linked
responses may involve differences in incidence of
effects, organ specificity, and latent period. For
example, N-2-fluorenylacetamide induces liver cancer
primarily in male rats (Miller, 1970) while 7, 12-
dimethylbenz(a)anthracene induces a higher incidence and
multiplicity of skin tumors in male mice than in female
mice (E. K. Weisburger, 1971). Dichlofluanid damages
the testes in dogs (Vettorazzi, 1975). Hodge, Downs, et
al. (1968) found an abnormal blood pigment in female
dogs at 25 ppm while at 625 ppm the abnormal blood
pigment was present in both sexes.
3. Age at Start of Study
Objective: The standards for age at
start of a study are designed to maximize the sensi-
tivity of the animal test system and to allow sufficient
time for toxic effects to develop.
a. Weanlings. According to the proposed test
standards, the tester must.begin to dose animals as soon
as possible after their weaning and environmental
acclimatization but no later than six (6) weeks of age
for rodents and ten (10) weeks of age for dogs.
There is general agreement that animals roust
be started on treatment at a young age in order to
maximize the sensitivity of the animal bioassay system
(NAS, 1961; Weisburger and Weisburger, 1967; Arcos,
Argus, and Wolf, 1968; WHO, 1969; Berenblum, 1969;
Magee, 1970; Tomatis, 1974; NAS, 1975, Canada, MHW,
1975; Sontag, Page, and Saffiotti, 1976; Weisburger,
1976; WHO, 1978b; FSC, 1978). However, a great deal of
discussion and some controversy has arisen over the past
several years over the use of prenatal and neonatal
animals versus use of weanling animals. (FIFRA, 1978;
Rice, 1976; Toth, 1968). EPA requires the use of
weanling animals in general oncogenicity testing. This
is because weanling animals actively undergoing protein
synthesis and cellular proliferation/maturation, and
subject to alterations in physiology occurring during
sexual maturation usually are more responsive to
chemical toxicants including chemical oncogens than
adult animals (Weisburger and Weisburger, 1976). Use of
weanling animals also allows the test substance to be
administered for the major portion of the animals' life
span, allowing for extensive exposure and maximum time
for development of toxic effects.
b. In utero. Theoretically, in order to
detect all age-related effects, animals should be
exposed during all phases of their lives, including
during gestation (Canada, MHW, 1975; Tomatis, 1974).
-------
11-16
However, after reviewing the available data, EPA
concluded that the scientific basis coupled with certain
technical problems is such that prenatal and neonatal
animals should not be required for general oncogenicity
testing instead of weanlings.
In utero testing for potential chemical
oncogens took on new importance when diethylstilbestrol
(DES) was shown to be oncogenic to humans following
prenatal exposure (Herbst, Ulfelder, and Poskanzer,
1971a-b). Tumors in offspring of exposed mothers have
now been shown to be induced by at least 30 chemicals
(Tomatis, 1974; Rice, 1976). However, all of these
chemicals were already known to produce tumors in adult
and/or newborn animals when they were tested (Tomatis,
1974; Rice, 1976). Because too few chemicals not
oncogenic in adults have been tested via in ut.erp
exposure, the extent of enhanced sensitivity of this
method is difficult to quantify. Fetal tissue may be
more susceptible or resistant to certain chemical
oncogens simply due to their lack of metabolic
competence or to changes in sensitivity of tissues at
different stages of development (Rice, 1976). EPA
believes more work needs to be done to validate this
test method before it can be adopted as a testing re-
quirement.
EPA is also hesitant to require in utero testing,
because of various technical difficulties. Restricting
administration of the test substance to the period of
gestation may result in inadequate exposure and thereby
negate use of any negative results. Use of this method
does not decrease the length of time of the study,
especially in regard to negative results. The amount of
chemical administered is very difficult to predict
because of possible fetal toxicity including terato-
genicity. To achieve use of maximum tolerated doses, it
is usually necessary to adjust levels for various ages.
thus making dose-response relationships difficult to .
evaluate.
c. Neonatal: Following the pioneer work of
Pietra, Spencer and Shubik (1959), it was thought that
using neonatal animals might allow for reducing the
number of test animals and length of exposure for
oncogenicity and chronic toxicity studies (Delia Porta,
1963). Studies of drugs and pesticides also indicated
that newborn animals were generally more sensitive than
adults with a mean ratio of response to a given dose
(newborn/adult) of 2.9 and a range of 0.6-10.0 (Hayes,
1975). Careful review of available data indicates that
neonatal treatment alone cannot be recommended in some
instances since certain known oncogens have gone
undetected in this method (Delia Porta and Terracini,
-------
11-17
1969). EPA believes that more work needs to be done
also to validate this test method. Use of neonatal
animals also entails many of the same technical problems
as use of prenatal animals, i.e., inadequate exposure
and differences in metabolic competence and tissue
sensitivity.
4. Number of Animals/Test Group
Objective. The standards for selection of
numbers of animals per test group are set forth to
provide for sufficient number of test animals per group
so that reliable statistical analysis can be used to
evaluate the validity of test results.
a. Number. Each group of rodents must contain
at least 50 animals for oncogenicity studies and 58
animals for non-oncogenic and combined chronic toxicity
studies. The eight additional animals required for the
non-oncogenic and combined chronic toxicity studies are
the eight predesignated animals need for clinical
chemistry studies. Each group of dogs must contain at
least 6 animals.
Ideally, in arriving at these numbers, EPA would
consider the desired sensitivity of the test, expected
incidences of comparable spontaneous diseases or tumors
in control animals, and percentage of animals that are
expected to survive to an adequate age to show chronic
effects (referred to as the effective number of animals)
(Page, 1977a). However, as discussed below, the final
numbers of animals required reflect a compromise between
the optimal sensitivity and practical factors such, as
costs and availability of animals (in case of dogs or
primates).
The number of animals in each test group is an
integral part of the sensitivity (i.e., the ability of
the study to demonstrate the effect(s) studied and the
reliability of the observed effect(s)) required for
statistical analysis of test data (Barnes and Denz,
1954; Shubik and Sice, 1956; Vos, 1959; MAS, 1961;
Arcos, Argus, and Wolf, 1968; Berenblum, 1969; Magee,
1970; NAS, 1977; Page, 1977b; FSC, 1978). Sensitivity
of a test depends on the smallest difference in disease
or tumor incidences between exposed and control groups
the study is designed to detect and the degree of
confidence with which this difference should be detected
(Clayson, 1962; Arcos, Argus, and Wolf, 1968). Type of
data is also a consideration; data may be qualitative,
i.e., presence or absence of effects, or quantitative,
i.e., latent period or time to effect.
-------
11-18
In statistical significance testing the minimal
detectable difference, (delta), between exposed and
control groups is dependent upon the standard error of
the mean. The standard error is defined as (&-/YZ? )
where (G~ ) refers to the standard deviation of the
distribution and n refers to sample size. Because the
standard error term is used as the denominator in the
minimal detectable difference term, the smaller the
difference to be detected, the larger the sample size
requirements in the protocol.
In chronic toxicity testing, sensitivity also
involves minimizing the probability of false positives
and false negatives. A false positive occurs when an
effect is judged to be the result of exposure to the
chemical (e.g. oncogenicity based upon the test results)
when the results are actually spurious or not the result
of exposure to the test substance. A false negative is
defined as failure to correctly classify or detect the
toxic effect. The risk or probability which a tester
allows for false positives is referred to as Type I
error (alpha); 1 minus the probability of alpha is
denoted as the level of statistical confidence (i.e.,
the ability to detect a true positive). Probability of
not detecting a toxic effect is referred to as Type II
error (beta); 1 minus the probability of beta is denoted
as statistical power (i.e., the ability to detect a true
negative) (Fears, Tarone, and Chu, 1977; Fleiss,
1973). In determining the level of confidence and power
in study design, a statistical distribution type for
occurrence of Type I and II errors is assumed, such as
the lognormal, Poisson or binomial.
Because an unlimited number of animals can not be
tested each time, a limitation on sensitivity must be
set for general testing of unknown chemicals for
toxicity potential. Results of biological investiga-
tions generally have been considered to be significant
if the probability (p) that the difference in incidences
of diseases or tumors between control and exposed groups
due to chance is relatively low, i.e., p=0.05 (NAS,
1961). This "p" value is referred to as the signifi-
cance level, or probability of false positives. As
illustrated in Table 1 for p=0.05, the decision as to
whether results produced in an exposed group by the test
substance is different from the result in the control
group is dependent on the incidence of comparable
spontaneous diseases or tumors in control animals and on
the effective number of exposed animals (Page, 1977b).
-------
11-19
Table 1
Incidence of Tumors in Exposed Groups
Required for Significance
(p=0.05) Depending on Test Group Size and
Incidence of Tumors in Controls*
Incidence of Tumors No. of Animals Per Group**
in Controls (%) 10 25 50 75 100
0
10
20
30
40
5056
70
80
90
100
20*
40
52
64
72
12*
28
40
52
62
8*
24
36
47
58
656
21
34
45
55
•Calculations based upon tabulations of Mainland and
Murray (1952) and presented by Page (1977a).
**Exposed and control groups of same size.
In practice, one usually cannot predict in advance
how many test animals will survive to an adequate age to
show effects or whether effects that may be induced will
represent a generalized elevation of the type which
arise spontaneously in controls or will represent a type
practically nonexistent in controls. For these reasons,
one cannot specify error rates (e.g., 95% probability of
detecting a 10$ increase in tumor*incidence between
exposed and control groups) to determine sample size for
each specific study. Because one cannot specify error
rates, historically, a minimum of 50 rodents or 4-8 non-
rodents per test group has been selected (Goldenthal and
D'Aguanno, 1959; Benitz, 1970; Newberne, 1975; NAS,
1975; Sontag, Page and Saffiotti, 1976; Page, 1976;
Page, 1977a-b; FIFRA, 1978). This represents a
compromise between the desired sensitivity of the test
system and practical considerations, such as cost and
quantity of work. As discussed previously, for
biological investigations the significance level is
-------
11-20
usually set at p=0.05 (i.e., there is no greater chance
of being wrong than once in 20 such studies). The
economic factor is easily seen by the cost of the non-
oncogenic chronic health effects test which requires 50
rats per group and 6 dogs per group and is estimated to
cost $550,000 (see Section III). Another factor that
must also be considered is quanitity of work. EPA
agrees with Benitz's basic concept that more useful
information can be obtained in thorough studies carried
out in a relatively small number of animals than in
incomplete studies using an excessive number of animals
(Benitz, 1970). „ , ..
'•- • ' <
EPA also agrees with Benitz that clinical chemistry
testing cannot be done on all 50 rodents and that 8
animals per group will give sufficient information for
Section U needs (Benitz, 1970). These 8 animals are in
addition to the normal 50 rodents per group because the
manipulation of the animals during the clinical chemis-
try tests may alter the effect of the test substance.
- i '
b. Randomization: Another important element in
statistical analysis of test data is proper allocation
of animals to exposed and control grdups (Arcos, Argus,
and Wolf, 1968; Berenblum, 1969; Roe and Tucker, 1971*;
Peck, 1974; Sontag, Page, and Saffiotti, 1976; NAS,
1977; Page, 1977b; FSC, 1978). Randomization of animals
is necessary to ensure that unintentional selection
biases are not introduced into the study. The test
standards do not specify use >of a given randomization
method. NCI recommends that all animals be from the .-
same supply source, ideally within 2-3 days of the same*
age, and optimally assigned to test .groups and cages by
the use of a table of random number (Sontag, Page, ami
Saffiotti, 1976). Other methods of randomization are
available such as Weiner's complete"block design
(Weiner, 1972).
• v
5. Number of DoseiLevels and Dose Selection
Objective: The standards for selection
of dose levels are designed to optimize exposure condi-
tions to reveal any toxic response, to-ensure-that :
adequate data will be available at the end of the,study,
and to obtain information, if possible, on dose-response
relationships.
a. Oncogenicity Test Standards. The range of
dose levels selected for oncogenicity testing must
include one which'is the maximum that can be admini-
stered for the duration of the test period, with
induction of.demonstrable but only slight toxicity, and
no substantial reduction in longevity due to effects
other than tumors (HDL); one which is a specified level
-------
11-21
to 1/2) of the HDL, and one which is less than 1/2
of second dose but no less than 1/10 of the HDL.
i. High Dose Level — 'Need: The main controversy in
selecting dose levels for oncogenicity tests involves
the use of an upper dose level that is in the toxic
range (Page, 1977b; Munro, 1977). Use of such a high
dose level is intended to achieve maximum sensitivity of
the test system without altering its accuracy (NAS,
1961; Weisburger and Weisburger, 1967; WHO, 1969;
Friedman, 1974; Sontag, Page, and Saffiotti, 1976; WHO,
1978b; FSC, 1978). As discussed in the previous
section, the number of animals (50 per group) tested is,
of necessity, extremely small compared to the size of
the human population potentially at risk. A chemical
which induces tumors in one percent of the U.S. popu-
lation would result in over two million new cancer
cases. Yet an exposed group of 50 animals must have at
least a twelve percent incidence of tumors (if controls
have a zero percent incidence) for results to be
considered statistically significant at the p=0.05 level
(Table 1). It is believed that a chemical oncogen,
which at environmental or occupational levels is a
threat to humans, will cause a statistically significant
detectable increase in tumor incidence when administered
at much higher dose levels to a small population of test
animals. Within the limitation of the animal's ability
to tolerate the chemical for long-term exposure without
death or substantial life-shortening effects, usually
the higher the dose is the higher the tumor incidence
and the shorter the latency period (J.H. Weisburger,
1976). Based on this data,
-------
11-22
animals, and no substantial reduction in longevity due
to effects other than tumors. EPA has defined its
required upper dose level slightly different than NCI
because it wants to make sure that this level is
slightly toxic but without an effect on longevity.
Weight reduction is only one acceptable demonstrable
parameter of slight toxicity.
Major criticisms of using a toxic dose, whether
defined as a MTD or HDL, are: (1) at high levels,
normal metabolic pathways may become saturated leading
to aberrations in metabolic pathways which are pre-
disposing to tumor development; (2) this dose level may
be incompatible with normal physiological function; and
(3) this dose level may be unrealistic compared to human
exposure (Munro, 1977). A ten percent decrement in body
weight is not considered normal physiology. EPA
recognizes such problems but agrees with the present
general scientific opinion that if the high dose group
shows no sign of chronic toxicity, decrease in mean
weight, the oncogenic potential to the chemical
substance may be underestimated, i.e., the study will
have a higher probability of being a false negative
(J.H. Weisburger, 1976; Sontag, Page and Saffiotti,
1976; Page, 1977b; WHO, 1978; FSC, 1978). Oncogenicity
is a dose dependent effect with some chemical oncogens
having fairly steep dose-response curves (Weisburger and
Weisburger, 1967). For these chemicals, the tumor yield
falls off rather rapidly with relatively minor altera-
tions in dose level. By not employing highest possible
doses, the effect might go undetected.
•* i
The term "maximally tolerated dose" (MTD) has
almost as many different connotations as there are in-
dividuals who use it and is, therefore, subject to
misinterpretation. In practice, predicted MTD's have
often not been achieved such that toxic effects
considered "not tolerated" occurred. Most importantly
because EPA is requiring use of a slightly toxic dose,
and to avoid confusion in use of MTD, EPA prefers to use
the term HDL. It is believed that this will allow the
emphasis to be placed on the important biological ques-
tions that need to be addressed.
iii. High Dose-Level--Determination: To predict
the HDL most accurately involves undertaking of expen-
sive, time-consuming pharmacokinetic and metabolic
studies. With these types of information, it is pos-
sible to predict accumulation in the body and dose-
dependent changes in the metabolic profile. In prac-
tice, however, to select the HDL, it is necessary to use
an empirical approach based on results of a subchronic
toxicity study (Sontag, Page, and Saffotti, 1976).
Review of data from the NCI bioassay program (Page,
-------
11-23
1977b) shows that considerable variability exists in
being able to assess the length of the subchronic
toxicity study needed to accurately predict long-term
toxicity. Clearly, the use of 4-6 weeks subchronic
studies is not satisfactory (Burchfield, Storrs and
Kraybill, 1974). Indeed, for some chemical substances,
a toxicity study of 20-30 weeks or longer may be needed
to provide an accurate estimate of two-year toxicity.
For most however, reasonable estimates can be made on
the basis of a 90 day study. After reviewing available
data and cost figures, EPA believes that routine
subchronic toxicity studies of at least ninety-days are
necessary to properly predict long-term toxicity in
oncogenicity studies.
iv. Second and Third Doses: In theory, only one
dose level is needed to detect oncogenic potential of a
chemical. But, because of problems in being able to
accurately predict this HDL, and because under TSCA a
risk assessment must be conducted once a chemical
oncogen has been identified, EPA is requiring the use of
at least three dose levels. Use of three dose levels is
in agreement with recommendations of other major sci-
entific bodies (Canada, MHW, 1975; MAS, 1977; FIFRA,
1978; WHO, 1978; FSC, 1978).
A second dose level acts as a safety factor if the
HDL is overestimated and survival rate or toxicity level
is such that results from this group cannot be used. A
second dose level is also needed because the HDL may not
produce maximum tumor incidence because of competing
cytotoxic and oncogenic activities (Arcos, Argus and
Wolf; 1968).
If results from all three dose levels can be used,
use of the third dose level provides for better
knowledge of dose-response relationships, and,
therefore, for better precision at estimating human
health risks (NAS, 1961; Page, 1977b). Use of at least
three dose levels allows for concurrent data collection
for both determining oncogenic potential and doing
curvilinear risk assessment. This is both cost and time
effective in that a second study will not have to be
done to obtain this information. The third dose level
also acts as a safety factor in case results from the
HDL cannot be used or in the extreme case where results
from both HDL and the second dose level cannot be used.
Ranges are set for the second (1/4-1/2 HDL) and
third (1/2 second dose - 1/10 HDL) dose levels for two
reasons: (1) to achieve the purpose(s) for using each
of the dose levels as discussed in the previous two
paragraphs; and (2) to allow for variations in dose-
response. The range for the second dose level is in
-------
11-24
agreement with that recommended by NCI (Sontag, Page and
Saffiotti, 1976) and FIFRA (1978). The third dose level
is the same as that recommended by the FIFRA guidelines
except for setting a lower limit of 1/10 HDL. This
lower limit was set because of the limited sensitivity
of the study. With the use of only 50 animals per
group, it is anticipated that the sensitivity of the
study will be too low to detect activity below this
level for most oncogens. This dose level is not meant
to be a threshold or NOEL value for oncogens.
b. Non-oncogenic Chronic Toxicity Test Standards.
A minimum of three dose levels are required for
non-oncogenic chronic toxicity studies in order to meet
the objectives of such tests. The objectives are to
detect chronic effects and to establish dose-response
patterns and "no-observed-adverse-effect levels. There-
fore, the range of dose levels selected for chronic
toxicity must include one eliciting a clearly toxic
response, one showing no-observable effects (NOEL), and
at least one intermediate level showing a certain level
of toxicity. The intermediate dose level(s) is required
to be selected so as to maximize information on dose-
response relationships and to allow application of risk-
analysis models.
As in oncogenicity studies, to predict the high and
low dose levels most accurately involves the undertaking
of expensive, time-consuming pharmacokinetic and meta-
bolic studies. With these types of information, it is
possible to predict accumulation in the body and dose-
dependent changes in effects. In practice, however, to
select the appropriate high and low dose levels, it is
necessary to use an empirical approach based on results
of subchronic toxicity study. As stated previously,
review of data from the NCI bioassay program (Page,
1977b) shows that considerable variability exists in
being able to predict long-term toxicity. After
reviewing available data, EPA believes that routine
subchronic toxicity studies of ninety-days duration are
necessary to properly predict long-term toxicity.
c. Combined Toxicity Test Standards.
For combined toxicity studies the tester must use
at least three (3) dose levels (in addition to controls)
for the mouse and dog. The three dose levels for the
mouse are those specified in the oncogenicity effects
test standards and for the dog those specified in the
non-oncogenic chronic effects test standards. The
reason for this is that the data from the mice studies
will be used for determining oncogenicity potential
while the data from the dog studies will be used only
-------
11-25
for determining non-oncogenic chronic toxicity
potential..
The requirements for rats in the combined effects
studies are different from the other two species because
data from such studies will be used to determine both
oncogenic and non-oncogenic chronic toxicity potential.
In order to accomplish this purpose, enough dose levels
must be used to insure that proper and sufficient
information is collected on both effects. As discussed
in the previous section, one of the dose levels selected
for non-oncogenic chronic toxicity testing must elicit a
clearly toxic response, one must show no-observable
effects (NOEL) and one is to be an intermediate level
showing a certain level of toxicity. For the oncogenic
toxicity testing, one dose level is to elicit slight
toxicity with the other two being some specified
fraction of the higher slightly toxic dose.
EPA believes these two tests can be combined if
four or five dose levels are used to provide adequate
information to study both effects. One dose must
clearly elicit a toxic response to fulfill the objec-
tives of the upper dose for the non-oncogenic chronic
toxioity studies; a second dose must elicit only slight
toxicity as defined in the oncogenicity effects stand-
ards to fulfill the objectives of the high dose level
for the oncogenicity studies; the third and fourth doses
must be the lower two doses defined in the oncogenic
studies in order to adequately assess for oncogenicity
potential; the fifth dose must be a no-observable effect
level to fulfill the second objective of the non-
oncogenic chronic effects test standards. The require-
ments for an intermediate dose level to show dose
response relationship(s) for non-oncogenic chronic
effects can be fulfilled by the second dose level
described above. The requirements of the fifth dose
level can be fulfilled if either the third or fourth
dose levels are predicted to induce no adverse effects
other than tumors. Use of these four or five dose
levels in the rat should enable one to determine both
the oncogenicity and chronic toxicity potential of a
chemical and a NOEL for effects other than oncogenicity.
6. Controls
Objectives. The standards for use of control
groups are designed to assess possible contribution(s)
made by any factor(s) other than the test substance
itself and to determine normal survival rates and
spontaneous disease and tumor incidences, factors which
are necessary for proper interpretation of test results.
-------
11-26
a. Matched Controls. The tester must use a
matched control group identical in every respect to the
exposed groups except for exposure to the test
substance. This will be a vehicle control group if a
vehicle is used to administer the test substance or a
negative (untreated) control group if no vehicle is
used.
Study of matched controls is necessary for two
reasons. First, as in most toxicity studies, it is
necessary to assess the possible contribution made by
any factor(s) other than the test substance itself.
Second, and extremely crucial for interpretation of
results from chronic toxicity studies, is the need to
determine the normal life span and extent of natural
diseases including spontaneous tumors in comparable
control animals (Arcos, Argus and Wolf, 1968; WHO, 1969;
Roe and Tucker, 1974; Shimkin, 1974; Bickerton, 1974;
NAS, 1975; NAS, 1977; WHO, 1978a-b; FSC, 1978).
Information on the normal life span of test animals
is needed to determine certain correction factors which
must be taken into account when expressing the final
results. Determination of the concurrent spontaneous
tumor rate is important because it determines the
statistical significance of the disease or tumor rate in
exposed groups. As shown in Table 1, the higher the
spontaneous tumor rate in controls, the higher incidence
of tumors exposed groups must have for them to be
statistically significant. Therefore, studies designed
to detect weakly active chemical toxicants must be
carefully controlled.
If the required matched control group is a vehicle
control group or if the toxic effects of the vehicle are
unknown, the tester may want to also use a negative or
untreated control group to determine whether any effects
observed in the vehicle-exposed control group or chemi-
cally exposed groups are due to ancillary materials used
in the study. Use of these two types of controls should
help safeguard against attributing a chronic effect to
the test substance when in fact it may be due to the
vehicle instead.
b. Positive Controls: A third type of control is
the positive control. For purposes of these test
standards, this type of control is not required for
every chemical but may be required by EPA when it is
necessary to ascertain whether the test animals are
sensitive to or respond in a predictable manner to known
chemical toxicants, to assess the relative potency of
certain chemical toxicants and to test the reliability
of a laboratory conducting chronic health effects
studies (NAS, 1961; WHO, 1969; J.H. Weisburger, 1974;
-------
11-37
Sontag, Page and Saffiotti, 1976; NAS, 1977; Sontag,
1977; Page, 1977b; WHO, 1978b). Establishing that test
animals are capable of responding in a predictable
manner to a known chemical toxicants provides a degree
of confidence that tests with unknown agents can be
accepted as valid.
As stated above, inclusion of positive control
groups will not be required for each unknown chemical
required to be tested. However, it may be required or
recommended that they be used in relationship to testing
of a series of compounds (Sontag, Page, and Saffiotti,
1976; Page, 1977b; WHO, 1978b). Selection of a positive
control substance ideally should be on the basis of
chemical similarity to the agents under test (Arcos,
Argus, and Wolf, 1968; Peck, 1974; NAS, 1975; Sontag,
Page, and Saffiotti, 1976). However, in a large testing
program, where chemicals having diverse structures are
being tested, proper considerations of structure may not
be possible. For large-scale testing programs of orally
administered agents, the most common oncogens used
appear to be N-2-fluorenylacetamide (2-AAF), diethyl-
nitrosamine (DEN), safrole, and 3-aminotriazole. Target
organs for 3-aminotriazole are primarily the thyroid and
liver, whereas the other three are primarily liver
oncogens. DEN and 2-AAF are potent carcinogens, whereas
safrole and 3-aminotriazole are less potent. Other com-
pounds used include uracil mustard, nitrogen mustard,
urethane, 7, 12-dimethylbenz(a)anthracene, N-dimethyl-
stilbenzmine, and N-methyl-4-diethylaminoazobenzene, N-
methyl-N'-nitro-n-nitrosoguanidine and N-(4-(5-nitro-2-
furyl)-2-thiazolyl)- formamide.
It should be recognized that there are species/
strain differences in response to some positive control
substances. Thus, sensitivity of animals in use should
be considered in selecting positive control substances.
While the NCI guidelines recommend only a single
dose level for positive controls (Sontag, Page, and
Saffiotti, 1976), EPA may require that several dose
levels be used in order to equate response(s) to proce-
dural design aspects, such as concentration, method of
administration and length of exposure. While a low dose
of a positive control substance might mimic the weak re-
sponse of the chemical under study, a higher dose could
serve to quickly verify consistency and sensitivity of
response of test animals. Use of multiple dose levels
of a positive control substance might also allow for
semi-quantitative interpretation of results with the
test substance.
For some strong carcinogens such as 2-AAF and DEN
(Sontag, Page, and Saffiotti, 1976), number of animals
-------
11-28
required, especially at high dose levels, may only need
to be 20-25 per group.
Regardless of the positive control substance used,
precautions should be taken to minimize exposure of
personnel or other animals in the laboratory. Use and
handling of positive control chemicals should be
carefully monitored and conform to proper safety precau-
tions.
c. Historical Controls: The fourth type of
control group is the historical or colony controls which
should include all control animals for a specific strain
or breed studied within the most recent five-year period
(ILAR, 1976). These controls are a valuable source of
information on the normal life span and spontaneous
disease and tumor incidence of each individual strain.
They cannot, however, replace matched controls because
of the accuracy needed in determining the spontaneous
disease and tumor incidence (NAS, 1977). Data on
historical controls is required to be submitted as a
part of the study plan.
7. Route(s) of Exposure
Objective: The standards for selection of
route(s) of exposure are set forth to provide the most
reliable test system.
In agreement with other standards (Shubik and Sice,
1956; FDA, 1959; Berenblum, 1969; Peck, 1974; NAS, 1975;
Canada, MHW, 1975; NAS, 1977; WHO, 1978a; FSC, 1978),
EPA is generally requiring that test substances be
administered by the route that duplicates or most
closely simulates the major known or expected route by
which human exposure occurs. This is the accepted
method because results are generally directly amenable
to evaluation in terms of potential human health
hazards.
However, if humans are exposed via several routes,
the major route of exposure may not be the most impor-
tant parameter. In this case, EPA may consider the most
important parameter to be the route which is anticipated
to be the most sensitive in terms of chronic toxicity.
For example, even though greatest exposure to a chemical
may be to the skin, the more hazardous exposure may be
through inhalation or ingestion. This is due to the fact
that the mode of administration influences toxicity at a
given site since it dictates concentration at the site
(WHO, 1969; J. H. Weisburger, 1975). In deciding on
route of exposure, EPA will consider not only human use
or exposure but also specific properties of the chemical
including its absorption, distribution and metabolism,
-------
11-29
organs affected, dose, and length of treatment. As will
be discussed later, potential safety hazard to labora-
tory personnel must also be considered.
Each route of exposure has advantages and
disadvantages. Weisburger and Weisburger (1967), Arcos,
Argus and Wolf (1968), Magee (1970), Canada, MHW, (1975)
and Page (1977b) describe in detail various con-
siderations involved in using each specific route.
8. Period of Exposure and Observation
Objective: The standards for selection of
exposure and observation periods are designed to make
exposure conditions optimal for revealing any long-term
toxic effects.
a. Period of Exposure:
i. Oncogenicity Test Standards. The tester
must administer the test substance to rats and mice for
a minimum of 24 months but no longer than 30 months.
A few powerful chemical oncogens can induce statis-
tically significant incidences of tumor in a short time
period and after only one or a few doses (Clayson, 1962;
J.H. Weisburger, 1976). However, most chemical oncogens
need to be administered continuously for long periods of
time before statistically valid conclusions can be
obtained (Zwickey and Davis, 1959; NAS, 1961; Magee,
1970; NAS, 1975; Canada, MHW, 1975; Page, 1977b; WHO,
1978b; FSC, 1978). As discussed previously, to maximize
sensitivity of the test system and, thus, to enhance
confidence in a "negative" result, a test substance must
be administered continuously for the greater part of an
animal's life span (Tomatis, 1974; Page, 1977b).
Present controversy is over testing for a long but
finite period or for the full life span of the test
animals.
EPA, in agreement with most federal agencies and
scientific groups, is requiring use of a long (24-30
months) but finite period of exposure (Berenblum, 1969;
Magee, 1970; J. H. Weisburger, 1973; Peck, 1974; Sontag,
Page, and Saffiotti, 1976; WHO, 1978b; FSC, 1978). Use of
a finite period of exposure has several advantages.
Although induced-tumor incidences increase with time,
incidences of spontaneous tumors in controls also
increase especially during the latter part of an
animal's life span. Over the life span of the animal,
spontaneous tumor incidences may increase so much as to
conceal a true positive (Berenblum, 1969). Sacrificing
at a given time enables better evaluation of data since
with this method time-at-risk and age are constant for
-------
11-30
exposed and control groups. Quality of tissue samples
ar= anticipated to be higher in animals killed at a
finite time. Tissue quality generally decreases as age
increases (Berenblum, 1969). Termination of a study
after a given period is more cost effective because it
helps facilitate both the pathology examination and
evaluation of test data. There will be no need to
continue a study simpl because one or two animals have
survived for an • unusually long time (Berenblum,
1969). There will also be no need for an above average
quality and quantity of staff to take care of aged
animals.
The Agency is requiring a minimum duration of 24
months for both the rat and mouse. While this duration
for rats may not be controversial, its use for mice may
be controversial. However, EPA believes that under the
hygenic and dietary conditions prescribed by the
proposed test standards most strains of mice will have a
high survival rate at 24 months. The NCI has found that
approximately 75 percent of its control B6C3P1 mice
survive to 25-1/2 months of age (Page, 1977b). Such an
increase in exposure time will help to maximize the
sensitivity of the study. The one assumption being made
is that 24 months is adequate time to detect virtually
all chemicals which might be potential oncogens.
Experimental evidence seems to indicate that very few
tumors will be observed for the first time in a study
only after 24 months (J. H. Weisburger, 1976).
ii. Non-oncogenic Chronic Toxi-
city Test Standards.
A. Rodent. Duration of dosing in non-
oncogenic chronic toxicity tests for rodents, like for
the oncogenicity tests, has long been the subject of
controversy. Life span studies in the rodent are
recommended by the NAS (1975) for evaluating chemicals
in the environment. NAS (1977) cites life span studies
in rodent with sacrifice when mortality reaches 80
percent. Schroeder (1973) described the effects of 30
trace elements on life span- and foufld the median life
span to be a sensitive indication of effect. Gray
(1977) showed that mortality due to chronic progressive
nephrosis had a dose-response relationship in a long-
term toxicity test.
Toxicological assessment of chemical effect by life
time exposure with determination of effects on growth,
mortality, patterns of disease, reserve function
capacity of organ systems, sensitivity to infection, and
median life span provide necessary information for
hazard evaluation to man and his environment. EPA is
requiring studies of at least 30 months for two reasons:
-------
II-3J
(1) to determine effects on median life span; and (2) to
determine changes in sensitivity with age. EPA believes
that use of this long but finite period will allow
sufficient information to be collected on both of these
effects while allowing certain advantages. These
advantages, as discussed in the previous section,
include better evaluation of data, better quality tissue
samples, and facilitation of the pathology examination
and data evaluation.
B. Non-rodent. Duration of non-rodent studies
have likewise been a controversial subject. Fitzhugh
(1959) cites in the Appraisal of the Safety of Chemicals
in Foods, Drugs, and Cosmetics, "Although previous
outlines have recommended only one year in dogs, newer
information confirms our opinion that insidious changes
require 2 years or more to develop in this species."
NAS (1977) recommends a 1 to 2 year study in dogs.
Loonis (1974) recommended a duration of 2-7 years in
non-rodents.
A two year study in the dog represents
approximately 10 percent of this species life span.
Therefore, in terms of life span, two year dog studies
are equivalent to subchronic toxicity testing in
rodents. Utility of the 10 percent life span study for
subchronic toxicity effects has been aptly demonstrated
in the rodent (Hayes, 1975; Boyd, 1972). Historical
utility of the two-year study in dogs is aptly demon-
strated by their use in setting acceptable daily intake
(ADI) for thirty-three pesticides (Vettorazzi, 1975).
Several investigators have found additional effects
when dogs were exposed for longer than six months.
Braun, Sung, et^ al^. (1977) found changes in some
clinical chemistry parameters at 2 years but not at 18
months while investigating long-term toxicity of
tetrachlorobenzene. Weil, Woodside, ^t_ al_. (1971)
observed hematologic changes at two years which were not
seen at 1 year in dogs exposed to propylene glycol.
Histopathologic changes were observed in kidney and
liver of dogs exposed to methomyl for 2 years but not at
one year (Kaplan and Sherman, 1977). Case, Smith and
Nelson (1976) noted central nervous system stimulation
in dogs at one year but not at six months after feeding
netopam. Herrman, Wiegleb and Leuschner (1977) found
side effects in the last weeks of a 12-month experiment
which were attributed to exhaustion of fluid and
electrolyte reserves in the dogs following exposure to
etozolin.
After reviewing this data, EPA believes that a two
year study in the dog is necessary to adequately assess
the chronic toxicity potential of chemicals for TSCA
-------
11-32
Section 4(a) purposes.
b. Period of Observation: Another issue related
to the duration of study is whether there should be an
observation period between termination of exposure and
time of kill. The major reason for recommending such an
observation period is to allow time for any reversible
lesions to regress and for any irreversible lesions to
progress (Sontag, Page, and Saffiotti, 1976; Page,
1977b). This is said to enable the pathologist to
better assess any chemically induced irreversible toxic
manifestation such as precancerous lesions. Most
advocates of this issue recommend a 3-6 month
observation period but this is empirical (FSC, 1978).
Such a period should be determined by the pharmacoki-
netic properties of the test substance and the lesions
induced. A major difficulty with such an additional
period of time is the need to decrease time of exposure
to the chemical, thereby, decreasing sensitivity of the
test or to increase the duration of the study. EPA is
not requiring an observation period but if a testing
laboratory wants to include such an observation period,
it should include additional groups of animals at the
start of a study which can be taken off of treatment
early and then observed.
10. Interim Kill
Objective. The standards for interim kill are
set forth to allow testers to obtain at their discretion
additional information which may help in final
interpretation of test results.
EPA is not requiring interim kill of test animals
because the proposed test standards are primarily
designed to detect differences in type and incidence of
toxic effects and not differences in latent periods or
pathogenesis of the conditions, because of the increased
costs due to the need for more animals and more patho-
logy, and because of the uncertain use of the resultant
data in risk assessments. Admittedly, however*,informa-
tion from interim kills may be useful for evaluating the
pathogenesis and latent period of induced lesions and
for assessing reversible changes (NAS, 1977; FSC,
1978). If the tester wants to include an interim kill
in a study, the Agency would welcome such data; however,
the number of animals at the start of the study must be
increased by the number scheduled to be killed. This is
to assure that sufficient numbers of animals are at risk
for the long-term exposure to determine if there is a
statistically significant difference in number of
exposed animals with toxic manifestations compared to
control animals.
-------
11-33
C. Study Conduct Issues
1. Introduction
As shown in part III of this document, studies
designed to detect oncogenicity and chronic toxicity
involve extensive investment in animals, materials,
personnel and time. Because of this, each test animal
becomes a very valuable asset, especially the further
the study progresses. Loss of any test animal due to
disease or loss of any tissues due to autolysis or
cannibalism must, therefore, be avoided. At least two
aspects of the conduct of any well-designed chronic
toxicity study readily determine if the study will be
executed successfully. These are proper and timely
observation of each test animal during the test period
and proper conduct of the pathology examination.
2. Clinical Procedures
Objective. The standards for clinical
observation and tests are designed to detect and assess
toxic effects at the clinical level, to help eliminate
certain extraneous factors that might adversely
influence conduct of the test and/or interpretation of
results, and to help ensure the quality of animals and
their tissues, thereby, providing reliable animal test
systems.
a. Clinical Observations. Appropriately trained
employees must observe all test animals at least every
12 hours throughout the test period. Each animal must
also be weighed and clinically examined at least once
each week during the first 13 weeks of the study and
every 2 weeks thereafter.
General clinical observation and examination of
test animals is a neglected area in all toxicological
assessments (Canada, MHW, 1975; FSC, 1978). Fox (1977)
and Arnold, Charbonneau, et^ ^1^. (1977) provide detailed
discussions on clinical monitoring and assessment of
animals. In chronic toxicity studies, proper clinical
appraisal of each test animal must be directed towards
keeping the animal alive as long as reasonably possible
while still ensuring that its tissue specimens will pro-
vide relevant and useful data (Weisburger and Weisbur-
ger, 1967; Canada, MHW, 1975; FSC, 1978). The Agency
believes that proper observation of each test animals at
least every 12 hours throughout the test period will
alert the investigator to early onset of an infectious
disease or degeneration of health due to the test
substance, will decrease loss of tissue samples due to
cannibalism or autolysis, and will help assure that not
more than five percent of the animals in each group are
-------
11-34
lost during the study. Twice daily observation also
allows for observing animals during light and dark
periods and therefore for detecting photo-effects.
EPA may not accept studies with losses greater than
five percent per group because it believes such losses
can generally be prevented by proper conduct of the
study and because losses greater than five percent
decrease the sensitivity of the study. Such a decrease
in sensitivity may interfere with the detection of the
effect and the interpretation of data especially for
weak oncogens. In addition, the losses due to
cannibalism or autolysis are often those animals
experiencing the effect, e.g., oncogenicity, and thus,
the most valuable data may be lost. Arnold, Charbonneau
et_ al^. (1977), reported that proper observation
procedures in their laboratory reduced tissue losses due
to autolysis to one percent. EPA believes that the cost
of careful and frequent clinical observation and
examination is a warranted expenditure in order to
provide maximum data for evaluating the test and results
obtained.
In addition to daily observations noted above, each
animal must be palpated and carefully examined at least
once a week for the initial 13 weeks while the animal is
growing rapidly and then once every two weeks thereafter
(Canada, MHW, 1975; Fox, 1977). Palpation of the animal
allows not only for detection of certain tumors but also
enables detection of any enlargements of internal
organs. These detailed physical examination provide
indications as to the general health of the animal and
its tumor burden.
Body weights and feed consumption need to be
monitored because they are indicators of general health
and because this information is needed to evaluate the
test (Arcos, Argus and Wolf, 1968; NAS, 1977; Page,
1977b; WHO, 1978b). Changes in body weight of the HDL
group in oncogenicity studies are used to determine if
this dose is adequately toxic. If the test substance is
being administered in the diet, feed consumption data
are needed to determine how much is being consumed and
therefore, determine the dosage of test substance admin-
istered. Body weight may be directly related to
decreased food consumption, especially for test sub-
stances that are somewhat unpalatable to the exposed
animals.
b. Clinical Chemistry. Certain quantitative
clinical chemistry determinations including hematology
(the only determination to"be made in oncogenicity
studies), blood chemistry, urinalysis, function tests
and residue analysis are to be made on a minimum of
-------
11-35
eight predesignated rodents in each test group and on
all non-rodents. For oncogenicity studies, the tester
must perform the hematology determinations at one year
and at study termination. For non-oncogenic and com-
bined chronic toxicity studies, the tester must perform
the specified studies at least at 3,6,12,18,and 24
months and/or at study termination.
i. General.
The objectives of clinical biochemical tests
in toxicity studies are to monitor for disease states in
exposed as well as control animals (Street, 1970;
Cornish, 1971). These studies provide early indicators
of toxicity which lead to identification of target
organ(s), as well as evidence of reversibility of toxic
effects. The purpose of clinical biochemical testing in
chronic toxicity studies is different from the usual
diagnostic use in clinical medicine in that in the test
situation there is a rather homogenous population of
test animals, randomly partitioned into several groups,
one of which is a matched control. Variability which is
routinely seen in clinical situations can be reduced by
controlling test sampling time and conditions (diet,
temperature, humidity, lighting).
This study design, coupled with methods which have
increased sensitivity and precision, permit statistical
comparison between control and exposed groups. These
statistical analyses can identify very small group
differences which may be within the normal range but are
biologically significant (Street, 1970).
ii. Blood Chemistry.
Of the forty plus enzymes and other blood
components which have been identified, only a few are
functional components of the blood (e.g., thrombin,
plasmin, cholinesterase, etc.) (Todd and Sanford,
1976). The majority of enzymes are from tissues and
organs which the blood perfuses. The amount of enzyme
in the blood at any time is a function of its rate of
entry into and loss from the blood and is indicative of
the normal or disease state of the organ or tissue being
perfused (Todd and Sanford, 1976). The mechanisms
proposed for entry into the blood includes organ release
due to tissue necrosis, increased membrane permeability
and increased production. Removal from the blood is
controlled by chemical degradation and excretion.
Therefore, elevation of enzymes can be due to increased
release from organs and tissues or a decrease in removal
from blood.
-------
11-36
In a similar manner, other components of blood,
i.e., electrolytes, urea, glucose, are regulated by
rates of entry into and loss from the blood (Cantarow
and Trumper, 1975). Homer Smith is cited (Cantarow and
Trumper, 1975) as stating that "the composition of the
blood plasma is determined by what the kidneys keep
rather than by what the mouth ingests." Thus, EPA
believes that it is necessary to examine the readily
accessible blood, not only because of explicit
information but also because of its translation to the
clinical situation.
Necessity for using an operational set or panel of
laboratory tests for organ and system evaluation is due
to the following functions (Hyde, Mellor and Raphael,
1976). Rate of organ damage (indicated by enzyme
release) is related to dose of toxicant and susceptibi-
lity of the organ or system to the toxicant. Residence
time of the enzyme in blood is determined by rate of
loss and can be measured as enzyme half-life. Differ-
ences in rates may well account in part for difficulties
encountered in clinical chemistry determinations and
most certainly suggest that rigid quality control
utilizing sensitive and precise methods are necessary
for evaluation of these parameters and that multiple
clinical biochemical parameters are necessary to fully
evaluate organ and system status.
Representative panels are shown in Table 2. Devel-
opment of automated analysis has reduced the cost for
most if not all the determinations. Sample size
requirements have been reduced and generally require
less than 1 ml of serum, thereby eliminating the need to
sacrifice the test animal.
-------
11-37
TABLE 2
Representative'clinical Chemistry Panels
Renal Panel
Blood Urea nitrogen
Creatinine
Serum protein and electrophoresis
Osmolarity
Calcium
Magnesium
Phosphorus
Uric acid
Electrolytes
i
Liver Panel
Bilirubin, total and conjugated
Urine for bile pigments
Aspartate aminotransferase (GOT)
Alkaline phosphatase ,
Serum protein and electrophoresis
, Gamma-glutamyl transpeptidase
Leucine aminopeptidase
Alanine aminotransferase
Ornithine carbamyl transferase (
Muscle Heart Panel
Creatine kinase
Aspartate aminotransferase (GOT)
Lactate dehydrogenase
Hypertension Panel
Renal panel plus:
Serum electrolytes
Acid-base balance: blood pH, P^ bicarbonate
Lactate dehydrogenase
Triglycerides
Cholesterol
Pancreatic Panel
Serum amylase
Urine amylase
Serum lipase
-------
11-38
EPA is of the opinion that clinical biochemical
parameters performed with adequate sensitivity and pre-
cision and evaluated by appropriate statistical tests
are necessary for complete assessment of chronic
toxicity.
iii. Urinalysis.
EPA is also requiring routine urinalysis.
Such routine urinalysis may be the most used and abused
procedure in toxicological evaluation. One frequently
finds statements as to the doubtful usefulness of
routine urinalysis (NAS, 1977). It should be emphasized
that routine semiquantitative urinalysis is a screening
procedure and only that. As with any screening test, a
follow-up with a quantitative determination is required
when a positive is found.
As pointed out earlier, the analysis of blood
components is an useful indicator of renal function.
Urinalysis is required to detect early kidney damage
since enzymes of renal origin do not appear to any
extent in blood (Todd and Sanford, 1976). Urinary
protein correlates well with renal pathology (G'.u>y,
1977). Urinary glucose is an indicator of pancreas
function and of proximal tubular function of the kidneys
(Cantarow and Trumper, 1975). Urinary excretion is also
a major pathway for elimination of xenobiotics (Cantarow
and Trumper, 1975).
After reviewing the available data, EPA agrees with
Berndt (1976) that routine urinalysis is not only very
beneficial but necessary for complete toxicological
evaluation.
iv. Function Tests.
Function tests like blood chemistry tests are
necessary to detect and characterize toxic effects as
expressed by abnormal biochemical and physiological
functions (Benitz, 1970). These tests allow detection
of all types of changes including non-morphological
changes (NAS, 1975). EPA realizes that it is impossible
and not practical to give an exhaustive list of all
function tests or to make specific recommendations for
certain classes of chemicals. It also agrees with
Benitz (1970) that it is more judicious to select the
most meaningful tests depending upon the toxicity
results in subchronic toxicity studies or in the early
phases of the chronic toxicity study. EPA believes,
however, that certain organ systems (i.e., liver,
kidney, pulmonary and cardiovascular systems) must be
-------
11-39
routinely monitored via some function tests because of
their key positions in the metabolism, distribution
and/or excretion of toxic substances. Such key
positions increase the chance that these systems will be
affected by the test substance.
v. Residue Analysis.
The non-oncogenic and combined chronic
toxicity studies require measurement of levels of test
substance and certain metabolites in plasma, urine,
feces and target organs. With such data, comparisons
can be made between the amount of concentration of a
test substance or its metabolite(s) in a target tissue
or body fluid and the, corresponding level of toxicity,
it can be used to determine whether a steady-state con-
centration is achieved, and it will give some general
understanding of the absorption, disposition and elimi-
nation of the test substance and how these parameters
vary with concentration and between species. These
types of information in turn allow more complete and
accurate risk assessments to be made. They also form a
bridge between animal studies and human exposure
(Burchfield, Storrs and Green, 1977). EPA believes that
the importance of the data supplied by residue analysis
studies outweigh its cost.
3. Pathology Procedures.
Objective. The standards for pathology
procedures are set forth to ensure that an in-depth
pathology examination is done to detect minute as well
as more obvious toxic changes.
a. General. The tester must conduct detailed
necropsy and histopatholpgy examination of all animals
with approximately thirty (30) to forty (40) tissues
routinely examined microscopically.
It is essential in chronic toxicity studies to
include adequate consideration of pathology (Shubik and
Sice, 1956; Magee, 1970; Roe and Tucker, 1974; FSC,
1978). Final success of an entire study depends on the
current diagnosis and interpretation of the significance
of any lesion(s) noted both of which are part of the
pathology evaluation. To plan and conduct an animal
bioassay meticulously and then to permit an inadequate
pathology evaluation would be ludicrous. On the other
hand, no amount of subsequent pathology interpretation
will extract useful information from poorly planned or
executed studies. Two essential criteria for success in
the pathology examination itself are properly prepared
-------
11-40
tissue specimens and qualified personnel (Barnes and
Denz, 1954; PDA, 1959; Weisburger and Weisburger, 1967;
Page; 1977b; FIFRA, 1978; WHO, 1978b; FSC, 1978). A
lack of either will predispose a study to failure.
A complete pathology examination consists of sever-
al steps: macroscopic examination (necropsy), histo-
logic preparation of tissues, microscopic diagnosis,
recording and tabulation of lesions, and evaluation of
the test (Page, 1977b). These procedures represent a
major aspect of a chronic toxicity study both in terms
of time and expense. In the NCI bioassay program,
approximately 40 percent of the overall cost of a study
is for the pathology examination. One-fourth to one-
third person-years of a pathologist's time" is needed
just to examine the 15,000 to 20,000 tissues from one
study (Page, 1977b). The requirements are even greater
for the proposed TSCA test standards. Because of
economic and time considerations and because of the
shortage of qualified staff (GAO, 1978), controversy has
developed over how to cut back on the pathology
examination without drastically reducing the degree of
confidence in the final interpretation of the data
(Sontag, 1977). As discussed below, EPA does not
believe that for TSCA Section 4 needs the amount of
pathology can be reduced without compromising test
results. '
b. Gross Necropsy. There seems to be complete
accord that a detailed necropsy must be performed on
each test animal and that all of its major organs and
tissues must be preserved (FDA, 1959; Weisburger and
Weisburger, 1967; WHO, 1969; FDA, 1971; Peck, 1974;
Sontag, Page, and Saffiotti, 1976, NAS, 1977; WHO,
1978b; FSC, 1978). Extent and accuracy of the pathology
findings and, thus, those of the study are dependent on
extent and accuracy of the gross necropsy. It repre-
sents the last and most important chance to obtain
biological evidence.
i
To obtain maximum usable information from a study,
animals should be necropsied and tissue samples taken
immediately after death or kill. Because this is not
practical in all cases, the Agency has reached a compro-
mise that the necropsies be done within 16 hours of
'death. EPA believes that if the time of necropsy is
increased further certain important tissues will be lost
to autolysis and also that the fine details of dther
tissues will be lost. Thus, the sensitivity and the
reliability of the study data will be decreased.
-------
11-41
c. Microscopic Examination. While there seems to
be little controversy regarding the need for a careful
and thorough gross necropsy, considerable disagreement
exists as to number of animals and number of tissues
from each animal that should be examined
microscopically. Many guidelines use such terms as
"complete" or "adequate" examination, leaving selection
of animals and tissues up to the responsible patho-
logist1 s judgment (PDA, 1971; Magee, 1970; WHO, 1961).
Most recently, a series of committees have recommended
that as a minimum routine procedure, microscopic examin-
ation be done on all major tissues and gross lesions for
only the high-dose and control groups supplemented by
gross lesions and target organs from other dose groups
(Peck, 1974; NAS, 1977; FIFRA, 1978, FSC, 1978). This
type of recommendation places primary reliance upon the
gross necropsy examination for detection of lesions and
is based on the assumption that induction of lesions is
dose dependent. Peck (1974) recommends complete micro-
scopic examination on a representative number of animals
from the high-dose and control animals supplemented by
target organs from other dose groups. NCI has stringent
requirements in that all test animals must undergo
extensive microscopic examination (Sontag, Page, and
Saffiotti, 1976).
Controversy over number of tissues to be examined
is also considerable. Abrams, Zbinden and Bagdon (1965)
felt that 18 different tissues should be routinely ex-
amined while the WHO (1961) recommended an initial exam-
ination of only five organs (lungs, liver, spleen, kid-
neys and urinary bladder), plus organs showing gross
lesions. The NCI bioassay program was the first to
introduce an expanded protocol that required microscopic
examination of approximately 30 tissues sections plus
blood smear, tissue masses, and gross lesions from each
test animal (Page, 1977b). Zbinden (1976) believing
that a compromise was needed between an all-encompassing
evaluation of all possible tissues and a superficial
examination, assigned tissues into priority classes
according to the frequency with which morphological
changes are likely to occur. Fears and Douglas (1978a-
b) have also proposed for comment suggested procedures
for reducing the pathology workload.
EPA is in agreement with NCI that all test animals
must be thoroughly examined microscopically. In the
Agency's opinion, all test animals must be examined
thoroughly since there can be changes in organ specifi-
city as the dose level is varied. Throughout the devel-
opment of the chronic toxicity standards, it has been
the policy of EPA to try and maximize the sensitivity of
-------
11-42
the test system, and, thereby, enhance confidence in its
ability to detect chemical toxicants. One of the most
crucial ways of accomplishing this is by thorough micro-
scopic examination of each test animal. This enables
detection of not only gross lesions but also detection
of early small lesions, those in small organs and those
that have metastasized and are in a minute form.
D. Data Collection and Reporting Issues
1. Final Report
Under the proposed test standards, the Agency
is requiring submission of an extensive final report.
The sponsor must submit a full and detailed report,
generally in tabular or graph form, of all test
conditions, all observations made on the animals, and
the complete data analysis. Summaries only of the
results are not acceptable. Examples of the data that
must be collected are information pertaining to changes
in body weight, food consumption, dose administered,
diseases and treatments, time of death, method of kill
or cause of death, presence of any pathological lesions
and time of first observation, if known, all signs of
toxicity and their time of onset, regression or
progression of any lesion, irreversible lesion incidence
by tissue and type, multiplicity of specific lesion
types, and pathology diagnosis.
There are several purposes for such detailed data
requirements. Foremost is EPA's need to be able to
check the accuracy of the sponsor's analysis and to make
its own independent evaluation of the study. Without
such information, EPA scientists simply could not accom-
plish these tasks. It is commonly recognized that
inadequate documentation is one of the most difficult
problems confronting scientists in evaluating chronic
toxicity results (Berenblum, 1969; Sontag, Page, and
Saffiotti, 1976; Page, 1976; Page, 1977a-b; FSC, 1978).
Neither summaries of studies nor the current journal
style of publication provide the full report of test i
conditions and animal observations that is necessary if
another scientist is to evaluate the empirical and
analytical bases for the conclusions reached by the
sponsor. The ability to engage in such reviews is
particularly important in light of recent discoveries
that call into question the validity of many toxicity
studies submitted to EPA and the Food and Drug Adminis-
tration in the past. (For further discussion of this -
point refer to the preamble to the proposed "Good
Laboratory Practice Standards for Health Effects.")
-------
11-43
Detailed reports serve other functions as well.
They will facilitate the peer review process, thereby
helping to assure that reliance will be placed upon the
studies which merit it. Moreover, as additional
scientists review the data, the fresh perspectives they
bring may lead to new insights into the data and a
better understanding of the study's significance. And,
the knowledge that the data will be subject to scrutiny
will provide an incentive for the sponsor to produce the
best data. Finally, EPA will be assured that the
necessary information will be available for all
chemicals that are tested, thus facilitating efficient
and reliable comparison and evaluation of different
studies.
Review of the data will focus on three elements:
study design, study conduct, and the conclusions drawn
from the data. Examination of the first two elements
will provide the first major indication of the study's
worth; if a study is not properly designed and con-
ducted, the resulting data will be inadequate, unreli-
able, or insufficient to interpret. Without proceeding
any further, the reviewer will already know he must be
suspicious of the data base and the validity of the
study. Roe and Tucker (1974) have listed some of the
common faults found in chronic toxicity tests that can
be found by analysis of study design and conduct:
(1) inadequate randomization; (2) unintentional
variations such as differences in room temperature and
humidity; (3) high loss of animals without post-mortem
examination; (4) poor records of necropsy findings; (5)
use of non-standard post-mortem techniques; and (6)
failure to match microscopic findings with macroscopic
ones.
A reviewer would then look at the various data in
order to assess the validity of conclusions reached by
the sponsor, i.e., whether the test clearly indicates a
toxic response(s), whether the results are suspicious or
inconclusive; or whether there is no evidence of
toxicity (Page, 1977b). To do this, the totality of the
biochemical, clinical pathology, and statistical
evidence gathered on the test substance during the study
must be available (D'Aguanno, 1974; FSC, 1978).
In summary, extensive documentation is necessary to
assess the validity and reliability of each study and to
allow independent scientific analysis of the results.
For the majority of chemicals to be tested under the
proposed chronic toxicity standards, the decision as to
their chronic toxicity potential will probably be based
solely on evidence obtained from one such study, making
-------
11-44
it especially imperative that each study meet certain
scientific criteria. Hence, the need to incorporate
data collection considerations into the original study
design and to require detailed reports.
2. Interim Report.
Besides requiring submission of a detailed
final report, EPA is also requiring submission of
quarterly interim reports. The purpose of these reports
is three-fold: (1) to monitor the progress of a study;
(2) to learn of significant findings of such parameters
as survival, weight changes, clinical test results and
accumulative incidence of tumors and toxicity; and (3)
to learn of any catastrophic events which might affect
the quantity or quality of data and thereby, affect the
interpretation of the data or completely invalidate a
study. Such monitoring of studies will allow for
communication between EPA, the sponsor and the tester
during the study. More importantly, such monitoring
will make EPA aware of any possible imminently hazardous
chemicals or potentially hazardous chemicals which
should be considered for regulation before the final
report is submitted. EPA believes that the best way to
accomplish this is by requiring submission of summary
reports on a quarterly basis.
E. Good Laboratory Practice Issues
1. Introduction
Along with the test standards, the Agency has
proposed Good Laboratory Practice (GLP) standards to
assure the quality and integrity of data submitted from
health effects testing, including chronic toxicity
studies (Chapter 40, Part 772, Subpart B, section
772.110-1 of the Code of Federal Regulations). The GLP
standards set forth criteria for such matters as test
substance characterization, general personnel require-
ments, administration of testing programs, facilities
and equipment, facilities operation, and general study
design and conduct. These are only general criteria,
however, that apply to all health effects laboratory
testing under TSCA. Specific, more detailed, require-
ments for good laboratory practice are to be developed
in test standards for particular effects testing, as
appropriate. In the chronic toxicity test standards
several of these more detailed criteria have been
proposed.
-------
11-45
2. Personnel
The Agency considers adequate training and
experience a key factor in assuring the quality of data
generated. Therefore, the Agency has adopted stringent
requirements for specific key personnel involved in the
design, conduct and analysis of chronic toxicity
studies. The Good Laboratory Practices Standards for
Health Effects (Section 772.110-1) proposes general
educational and experience requirements for the study
director, and other personnel while the chronic toxicity
test standards (Section 772.113-1) set forth specific
qualification requirements for the pathologists,
veterinarians and certain technical employees.
As stated previously, two essential ingredients to
success in the pathology phase of a study are quality
tissue specimens and qualified personnel. Qualified
personnel includes everyone involved in the study. Key
to a successful animal care operation is a well-super-'
vised, well-trained and motivated animal care-taker
staff interested and concerned about the health of
individual animals and their role in quality research
(Page, 1977a-b; NAS, 1977). Obtaining quality tissue
specimens is dependent in part on the humane care given
to animals and astute observations of the veterinarians
and animal care-taker staff. Well qualified histology
technicians are essential to obtaining professionally
prepared tissue slides. Careful microscopic evaluation
cannot make up for lost animals or for inadequate gross
necropsy and histologic preparation. In simple terms,
the pathologist reviewing tissues slides can report only
lesions placed before him. Need for highly qualified
pathologists is equally important and cannot be over
emphasized. Final analysis of a study is based primari-
ly on pathology findings, and the extent and accuracy of
these findings depend on the qualifications of the
pathologist.
3. Animal Care and Facility
Quality of test animals, animal facilities and
animal husbandry practices contribute in no small manner
to successful outcome of a study (Who, 1978, FSC,
1978). Again, long-term animal studies are too
important and too costly to begin a study with animals
of inferior quality (Sontag, Page and Saffiotti,
1976). An appropriate animal facility with adequate
environmental controls is an essential element in
assuring that animals do not succumb from causes extra-
neous to the study. The facility must be designed and
maintained to meet the high standards of animal care and
-------
11-46
chemical and biological hazard controls required for
reliable and safe chronic toxicity studies. Operation
of the facility must strive to prevent entrance of
extraneous factors at all levels of containment, from
facility to animal room to individual cage (Page,
1977a). Several publications are available that provide
detailed information and recommendations on animal
facilities and their operations (Arcos, Argus and Wolf,
1968; Canada, MHW, 1975, NAS, 1976; Sontag, Page, and
Saffiotti, 1976; NAS, 1977). i
Successful outcome of a study is also dependent
upon the quality and extent of animal care and husbandry
(Page, 1977b). Primary goals of animal husbandry
practices are to promote health and humane care of
animals and to control environmental variables among
individual animals and test groups. Page (1977b) has
listed some of the most important husbandry factors that
can influence oncogenicity tests (Table 3). EPA believes
that these factors can be controlled or prevented by
routine practice of strict hygiene and disease preven-
tion measures and close clinical observation.
Table 3*
Husbandry Factors That Can Influence Chronic
Toxicity Tests
1. Infectious Diseases
a. Microbiologic
b. Parasitic
2. Chemical Pollutants in
a. Feed
b. Water
c. Bedding
d. Air
3. Operations Management
a. Cannibalism
b. Autolysis
c. Vermin Infestation
d. Diet
e. Prevention of Cross-Contamination
'Page, 1977b
-------
11-47
4. Dietary Requirements
Objectives. The standards for selection of
diets are set forth to help ensure the quality of test
animals throughout a study and to eliminate certain
extraneous factors that might influence conduct of a
study and interpretation of results.
a. Diet: The tester must feed test animals
standardized diets. A standarized diet for rodents is
specified in the test standards.The diet is a very
significant environmental source of variance in animal
research studies including long-term toxicity studies
(Barnes and Denz, 1954; FDA, 1959; NAS, 1961; Roe and
Tucker, 1974; NAS, 1977). It helps determine the
general health and longevity of test animals; it affects
sensitivity of the test system; and it affects
reproducibility of the study. The specific diet fed to
test animals must meet all of their nutritional require-
ments, be palatable and be free of toxic or infectious
microorganisms in order to promote good health and
longevity.
A negative relationship between obesity or a low
level of nutrition and longevity has been shown in
humans (Armstrong, Dublin, et^ al_., 1951) as well as in
various laboratory animal species (Berg and Simms, 1960;
Lane and Dickie, 1958; Silberberg and Silberberg, 1955;
McCay, Crowell, et_ al_., 1935; McCay, Maynard, et al.,
1939; Nolen, 1972; Ross, 1961; Ross and Bras, 1975;
Ross, 1977). Onset of age-associated changes in kidney
function as indicated by decreased protein use and
increased PAH transport was delayed in animals fed a
restrictive diet (Tucker, Mason and Beauche, 1976).
Prevalence and severity of kidney lesions were reduced
when caloric intake was decreased (Ross, 1976).
Incidence of myocardial and prostatic diseases could be
altered by diet composition (Ross, 1976) and a similar
relationship was observed between diet and
glomerulonephropathy.
Diets are known to affect the sensitivity of the
oncogenicity test in two ways. Experimental evidence
has shown that both caloric value and individual dietary
components and contaminants alter markedly the induction
rate and progression of chemically induced and spon-
taneous tumors in the rodent (Romburger, 1974; Newberne
and Rogers, 1976). A series of papers have been
published by Ross and his associates showing inter-
actions between dietary nutrient concentrations and
incidence of neoplasms (Ross, Bras and Ragbeer, 1970;
Ross and Bras, 1973; Ross and Bras, 1975; Ross and Bras,
-------
11-48
1976; Ross, 1977; ). Dietary effects on induction rate
and tumor progression may be due to changes in meta-
bolism, changes in intestinal flora or changes in the
host's immunological processes (Newberne, 1974). Test
results suggest dietary factors may have as much influ-
ence on the incidence of lesions in animals involved in
long-term studies as some test substances being
evaluated. Accordingly, it is essential to include the
diet as an environmental factor to be controlled in
long-term studies. This is particularly critical if
test results are to be replicated or if data generated
at different laboratories are to be compared.
b. Standardization of Diet: The Agency, along
with certain scientific groups, is concerned not only
with regard to effects that the diet has on test
sensitivity but also in being able to standardize
testing conditions (Newberne, 1974; NAS, 1975; Canada,
MHW, 1975; NAS, 1978; PSC, 1978). Standardization of
the diet will enable more reliable comparisons to be
made between test substances and between test systems.
Three types of diets are available for use in long-
term rodent studies. Closed formula rations are made
from natural ingredients and are the normal commercially
available rations. Information on ingredient composi-
tion of closed formula rations are not readily available
since they generally are trade secrets. Since they are
made from natural ingredients, there can be considerable
variation in nutrient concentration as sources of
ingredients are changed. These rations are the most
readily available, the most widely used and generally
thought to be the most economical.
Open formula rations contain both natural ingredi-
ents and ingredients of varying degrees of refinement.
They differ from closed formula rations in that both
quantitative and qualitative ingredient composition is
readily available and can be adjusted to meet specific
needs of individual testers. Uniformity of these
rations can be assured from lot to lot.
Synthetic diets are formulated only with specified
chemicals of known refinement. The major problem with
formulating such diets is lack of knowledge as to the
complete nutrient requirements of any test animal.
They are also expensive and generally not very
palatable. Their major advantages are absence of
naturally occurring contaminants and complete stan-
darization of ingredients (ILAR, 1976).
-------
11-49
Open formula rations most closely fill the need to
maximize testing conditions and to standardize the diet
and hence are proposed by EPA. Requirement of an open
formula ration is not anticipated to significantly
increase the cost of long-term animal studies (see
Section II). Experience of NIH has shown that their
cost for rations has decreased since they began using
open formula rations. No shortage of diet is anticipated
when Section 4 rules are promulgated since several
companies are selling the ration and more are expected
to begin selling it.
5. Contaminant Analysis Requirements
Objective. The standards for contaminant
analysis are designed to eliminate certain extraneous
factors that night influence conduct of a test and
interpretation of results.
Contaminants. The tester must analyze feed and
vehicles, if any, for certain specified contaminants.
Contaminants can be introduced with either one or
more ingredients in a diet or vehicle or through diet
and vehicle mixing apparatus (NAS, 1978). These may be
industrial contaminants such as the polychlorinated
biphenyls and insecticides or naturally occurring
contaminants such as aflatoxins. The objective for
analyzing for these contaminants is to establish a
profile of the kinds and amount of each contaminant,
test animals are exposed to during a study. If
unacceptable concentrations are detected, a change in
ration or source may be indicated. These data will also
be utilized when test results are evaluated to ascertain
any effect they may have on interpretation of results.
The list of contaminants provided in Appendix A
(Section 772.113-1 of the test standards) are known to
interfere with long-term chronic toxicity studies (ILAR,
1976). The list originally resulted from a four year
surveillance study conducted by the National Center for
Toxicological Research. All of these contaminants were
detected in analytically significant quantities in the
diets used. The presence of such agents in animal feed
or vehicle has serious implications. Indeed complex
interactions between critical levels of given contami-
nants might singely or in concert with other test
material produce results which are at best unreliable or
at worst misinterpreted. EPA believes that the maximum
permissible levels given in Appendix A are realistic and
based on scientific evidence.
-------
11-50
6. Safety and Health Standards
EPA recognizes that long-term toxicity studies
especially oncogenicity studies have inherent risks
associated with them. These include possible con-
tamination of laboratory facilities, laboratory person-
nel and the general environment by potential oncogens
and toxicants (Sansone, Losikoff and Pendleton, 1977a-bj
Page, 1977b; IARC, 1978). Once such events are detected
it is too late to undo any exposure that may have
occurred. Therefore, proper precautions and codes of
practice must be followed throughout a study. The
Agency recommends that guidelines developed by the DHEW
Toxicology Subcommittee for Carcinogen Standards
(August, 1978) and those developed by the International
Agency for Research on Cancer (IARC, 1978) be followed.
Other publications dealing with this subject have
been prepared and include the following: The National
Cancer Institute Safety Standards for Research Involving
Chemical Carcinogens (1977); the National Center for
Toxicological Research Carcinogen Standards (1978) and
the Code of Practice for the Safe Handling of Chemical
Carcinogens in Research Establishments (UK, 1978). hese
sources may be used for additional guidance.
-------
11-51
REFERENCES
Abrams, W.B., Zbinden, G. , and Bagdon, R. 1965.
Investigative Methods in Clinical Toxicology. J. New
Drugs, 5:199-207.
Ad Hoc Committee on the Evaluation of Low Levels of
Environmental Chemical Carcinogens. 1970. Evaluation
of Environmenta.1 Carcinogens. Report to the Surgeon
General, USPHS (NIH-77762), Department of Health,
Education, and Welfare. Washington, D.C. 20pp.
Albert, A. 1973. Selective Toxicity (5th ed.).
Chapman and Hall, London.
Albert, R.A., Anderson, E., and Train, R.E. 1976.
Interim Procedures and Guidelines for Health Risk and
Economic Impact Assessments of Suspected Carcinogens. J.
Nat. Cancer Inst., 58:1537-1541.
American Cancer Society. 1978. 1978 Cancer Facts
and Figures. American Cancer Society, New York.
Arcos, J.C., Argus, M.F., and Wolf, G. 1968.
General Characteristics of Tumors and the Testing of
Carcinogens. In: Chemical Induction of Cancer: Struc-
tural Basis and Biological Mechanisms, Volume 1, Aca-
demic Press, New York, pp. 303-463.
Armstrong, D.B., Dublin, L.I., Wheatley, G.M., and
Marks, H.H. 1951. Obesity and Its Relation to Health
and Disease. J. Am. Med. Assoc., 147:1007-1014.
Arnold, D.L., Charbonneau, S.M. , Zawidzke, Z.Z.,
and Grice, B.C. 1977. Monitoring Animal Health During
Chronic Toxicity Studies. J. Environ. Path. Toxicol.,
1:227-239.
.,
Association of Food and Drug Officials of the U.S
1959 (3rd Printing, 1975). Appraisal of the Safety of
Chemicals in Foods, Drugs and Co.smetics. Austin. 107
PP.
Aviado, D.M. 1978. Overwhelming Simularities and
Minimal Differences in Toxic Responses of Rats and Dogs
(Part I: 1974-1978 and Part II: 1966-1974). Personal
Communication to EPA.
Barnes, J.M., and Denz, F.A. 1954. Experimental
Methods Used in Determining Chronic Toxicity: A Critical
Review. Pharmacol. Rev. 6:191-242.
-------
11-52
Bencko, V. 1977. Carcinogenic, Teratogenic, and
Mutagenic Effects of Arsenic. Environ. Health Perspect.,
19:179-182.
Benitz, H.F. 1970. Measurement of Chronic Toxi-
city. In: Methods in Toxicology. G.E. Paget (ed.).
F.A. Davis Company, Philadelphia, pp.82-131.
Berenblum, I. (ed.). 1969. Carcinogenicity
Testing. UICC Technical Report Series, 2:158-196.
Geneva.
•
Berg, B.N., and Simms, H.S. 1960. Nutrition and
Longevity in th Rat. II. Longevity and Onset of Disease
with Different Levels of Food Intake. J. Nutr. 71:255-
263.
W.O. Berndt. 1976. Renal Function Tests: What Do
They Mean? A Review of Renal Anatomy, Biochemistry, and
Physiology. Environ. Health Perspect. 55-71
Bickerton, R.K. 1974. Spontaneous Tumors and
Related Pathologic Changes. In: Carcinogenesis Testing
of Chemicals. L. Golberg (ed.). CRC Press, Inc.,
Cleveland, Ohio, pp. 55-70.
Bock, F.G. 1964. Early Effects of Hydrocarbons on
Mammalian Skin, Prog. Exp. Tumor Res., 4:126-168.
Boyd, E.M. 1972. Predictive Toxicometrics.
Williams and Wilkins, Baltimore.
Braun, W.H., Sung, L.Y., Keyes, D.G., and Kociba,
R.J. 1977. Pharmacokinetic and Toxicological Evaluation
of Dogs Fed 1,2,4,5-Tetrachlorobenzene in the Diet for
Two Years. Toxicol. Appl. Pharmacol. 41:218-219.
Burchfield, H.P., Storrs, E.E., and Green, E.E.
1977. Role of Analytical Chemistry in Carcinogenesis
Studies. Adv. Modern Toxicol., 3:173-207.
Burchfield, H.P., Storrs, E.E., and Kraybill, H.F.
1974. In: Pesticides Environmental Quality and Safety,
Suppl. Vol. 3 F. Coulston and F. Korte (eds.). George
Thieme, Stuttgart.
Canada, Minister of Health and Welfare. 1975. The
Testing of Chemicals for Carcinogenicity, Mutagenicity
and Teratogenicity. Ottawa. 183pp.
Cantarow and Trumper. 1975. Clinical Biochemistry
(7th Ed.). A.L. Latner (ed.) W.B. Saunders Co.
-------
11-53
Case, M.T., Smith, J.K., and Nelson, R.A. 1976.
Chronic Oral Toxicity Studies of Netopam Hydrochloride
in Rats and Dogs. Toxicol. Appl. Pharmacol., 36:301-
305.
Casida and Baron. 1976. Pesticides Induced Delayed
Neurotoxicity, Proceedings of a Conference. EPA 600/1-
76-025. Washington, D.C.
Clayson, D.B. 1962. Chemical Carcinogenesis,
Little, Brown and Co., Boston, Massachusetts. 467pp.
Clayson, D.B. 1978a. Relationship Between Labora-
tory and Human Studies. J. Environ. Path. Toxicol.,
1:31-40.
Clayson, D.B. 1978b. Overview, Fact, Myth and
Speculation. J. Environ. Path. Toxicol., 2:1-8.
Cornish, H.H. 1971. Problems .Pos.ed by Observation
of Serum Enzyme Changes in Toxicology. CRC Critical
Rev. Toxicol., 1:1-32.
D'Aguanno, W. 1974. Interpretation of Test Results
in Terms of Carcinogenic Potential to the Test Animal:
The Regulatory Point of View. In: Carcinogenesis
Testing of Chemicals. L. Golberg (ed.). CRC Press,
Inc-., Cleveland, Ohio. pp.41-44.
Delia Porta, G. 1963. The Study of Chemical
Substances for Possible Carcinogenic Action. In: Pro-
ceedings of the European Society for the Study of Drug
Toxicity. Vol. II-Viewpoints on the Study of Drug
Toxicity. Excepta Medica Foundation. New York, pp.29-
40.
Delia Porta, G., and Terracini, B. 1969. Chemical
Carcinogenesis in Infant Animals. In: Progress in
Experimental Tumor Research, Vol. I.I. F. Homburger
(ed.). S. Karger, Basel, pp.334-363.
Fears, T.R. and Douglas, J.F. 1978a. Suggested
Procedures For Reducing the Pathology Workload in a
Carcinogen Bioassay Program, Part I. J. Environ. Path.
Toxicol., 1:125-137.
Fears, T.R. and Douglas, J.F. 1978b. Suggested
Procedures For Reducing the Pathology Workload in a
Carcinogen Bioassay Program, Part II: Incorporating
Blind Pathology Techniques and Analysis For Animals with
Tumors. J. Environ. Path. Toxicol., 1:211-222.
-------
11-54
Fears, T.R., Tarone, R.E., and Chu, K.C. 1977.
False Positive and False Negative Rates for Carcino-
genicity Screens. Cancer Res. 37:1941-1945.
Festing, M.F.W. 1975. A Case for Using Inbred
Strains of Laboratory Animals in Evaluating The Safety
of Drugs. Fd. Cosmet. Toxicol., 13:369-375.
FIFRA. 1978. Proposed Guidelines for Registering
Pesticides in the U.S.; Hazard Evaluation: Humans and
Domestic Animals. Fed. Reg., 43:37336-37403.
Fitzhugh, O.G. 1959. Chronic Oral Toxicity. In:
Appraisal of the Safety of Chemicals in Foods, Drugs,
and Cosmetics. Association of Food and Drug Officials
of the U.S. Austin, pp.36-45.
Fleiss, J.L. 1973. Statistical Methods for Rates
and Proportions, Wiley and Sons, New York.
Food and Drug Administration. 1976. Nonclinical
Laboratory Studies. Proposed Good Laboratory Practice
Regulations. Fed. Reg., 41:51206.
Food and Drug Administration. 1978. Nonclinical
Laboratory Studies. Good Laboratory Practice Regula-
tions. Fed. Reg., 43:59986-60025.
FDA Advisory Committee on Protocols for Safety
Evaluations. 1971. Panel on Carcinogenesis Report on
Cancer Testing in the Safety Evaluation of Food Addi-
tives and Pesticides. Toxicol. and Applied Pharmacoli,
20:419-438.
Food Safety Council. 1978. Chronic Toxicity
Testing. In: Proposed System for Food Safety Assess-
ment. Report of the Scientific Committee of the Food
Safety Council, pp.97-108.
Fox, J.G. 1977. Clinical Assessment of Laboratory
Rodents on Long Term Bioassay Studies. J. Enviro. Path.
Toxicol., 1:199-226.
Friedman, L. 1974. Dose Selection. In: Carcino-
genesis Testing of Chemicals. L. Golberg (ed.). CRC
Press, Inc. Cleveland, Ohio. pp.21-22.
General Accounting Office (GAO). 1978. Personnel
Shortages Affecting the Implementation of Chemical
Control Acts. Prepared by the U.S. General Accounting
Office for the House Subcommittee of Consumer Protection
and Finance (December 28, 1978).
-------
11-55
Golberg, L. (Ed.). 1974. Carcinogenesis Testing of
Chemicals. CRC Press, Inc. Cleveland, Ohio, pp.144.
Goldenthal, E.I. and D'Aguanno, W. 1959. Evalua-
tion of Drugs. In: Appraisal of the Safety of Foods,
Drugs and Cosmetics. Assoc. of Food and Drug Officials
of the U.S. pp.60-67.
Gary, J.E. 1977. Chronic Progressive Nephrosis in
the Albino Rat. CRC Critical Rev. Toxicol., 5:115-144.
Hartwell, J.L. 1951. Survey of Compounds Which
Have Been Tested for Carcinogenic Activity, 2nd ed.
Public Health Service Publ. No. 149., U.S. Government
Printing Office, Washington.
Hayes, W.J., Jr. 1975. Toxicology of Pesticides.
Williams and Wilkins Company, Baltimore.
Herbst, A.L., Ulfelder, H., and Poskanzer, D.C.
1971a. Adenocarcinoma of the Vagina: Association of
Material Stilbestrol Therapy with Tumor Appearance in
Young Women. New Eng., J. Med. 284:878-881.
Herbst, A.L., Ulfelder, H., and Poskanzer, D.C.
1971b. Registry of Clear-Cell Carcinoma of Genital
Trace in Young Women. New Eng. J. Med., 285:407.
Herrman, J.M., Wiegleb, J., and Leuschner, F.
1977. Toxikologische Untersuchurgen Mit Etozolin.
Arzneim-Forsch., 27:1758-1767.
Hodge, B.C., Downs, W.L., Smith, D.W. and Maynard,
E.A. 1968. Oral Toxicity of Linuron (3-(3,4-Dichloro-
phenyl)-l-methoxy-l-methylurea) in Rats and Dogs. Fd.
Cosmet. Toxicol., 6:171-183.
Hodge, H.C., Smith, D.W. and Chen, P.P. 1963.
Biological Effects of Organic Flourides. In: Flourine
Chemistry, Vol. 3. J.H. Simons (ed.). Academic Press,
New York.
Hoffmann, D., and Wynder, E.L. 1974. Positive
Controls in Environmental Respiratory Carcinogenesis.
In: Carcinogenesis Testing of Chemcials. L. Golberg
(ed.). CRC Press, Inc. Cleveland, Ohio. pp.35-39.
Homburger, F. 1974. Modifiers of Carcinogenesis.
In: The Physiopathology of Cancer, Vol. 1. F.
Homburger (ed.). Karger, Basel.
Hueper, W.C., and Conway, W.D. 1964. Chemical
Carcinogenesis and Cancer. Thomas. Springfield,
Illinois. 744pp.
-------
11-56
Hyde, T.A., Mellor, L.D., and Raphael, S.S. 1976.
Lynch1s Medical Laboratory Technology. Senior Author,
S.S. Raphael, Vol. 1, Section 1, 3rd ed. W.B. Saunders.
IARC. 1978. General-Quality Assurance Standards,
Appendix B. A Manual on the Safety of Handling
Carcinogens in the Laboratory (Draft).
Institute of Laboratory Animal Resources (ILAR).
1976. Long-Term Holding of Laboratory Rodents. ILAR
News, 19(4):1-25.
Institute of Laboratory Animal Resources (ILAR).
1977. Guide for the Care and Use of Laboratory Animals.
DHEW Publication No. (NIH) 77-23. Public Health
Service, National Institute of Health, Washington,
D.C. 55pp.
Kaplan, A.M., and Sherman, H. 1977. Toxicity
Studies with Methyl-N-[(Methylamine-carbonyl) oxy]-
ethanimidothioate. Toxicol. Appl. Pharmacol., 40:1-17.
Laboratory Chemical Carcinogen Safety Standards
Subcommittee of the DHEW Committee to Coordinate
Toxicology and Related Programs. Guidelines for the
Laboratory Use of Chemical Substances Posing a Potential
Occupational Carcinogenic Risk. Draft. August, 1978.
Lane, P.W., and Dickie, M.M. 1958. The Effect of
Restricted Food Intake on the Life Span of Genetically
Obese Mice. J. Nutr., 64:549-554.
Litchfield, J.T. 1961. Forecasting Drug Effects in
Man from Studies in Laboratory Animals. J.A.M.A.,
177:34-38.
Loomis, T.A. 1974. Essential of Toxicology (2nd
ed.); Lea and Febiger Publishers, Philadelphia.
Mainland, D. and Murray, I.M. 1952. Tables for Use
in Fourfold Contingency Tests. Science, 116:591-594.
Maugh, T.H. 1978. Chemical Carcinogens: The
Scientific Basis for Regulation. Science, 201:1200-
1205.
McCay, C.M., Cromwell, M.E., and Maynard, L.A.
1935. The Effect of Retarded Growth Upon the Length of
Life Span and Upon the Ultimate Body Size. J. Nutr.,
10:63.
-------
11-57
McCay, C.M., Maynard, L.A., Sperling, G. and
Barnes, L.I. 1939. Retarded Growth, Life Span, Ultimate
Body Size and Age Changes in the Albino Rat After
Feeding Diets Restricted in Calories. J. Nutr., 18:1.
McConnel, E.E., Moore, J.A., Baseman, J-.K., and
Harris, M.W. 1978. The Comparative Toxicity of
Chlorinated Dibenzo-p-dioxins in Mice .and Guinea Pigs.
Toxicol. Appl. Pharmacol. 44:335-356.
Magee, P.N. 1970. Tests for Carcinogenic Poten-
tial. Methods in Toxicology, 1:158-196.
McNamara, B.P. 1974. Toxicological Test Methods.
Assoc. Pood and Drug Officials, U.S. Quarterly Bull.,
38:33-50.
Miller, J.A. 1970. Carcinogenesis by Chemicals: An
Overview. Cancer Res., 30:559-576.
Munro, I.e. 1977. Considerations in Chronic
Toxicity Testing: The Chemical, The Dose, The Design.
J. Environ. Path. Toxicol., 1:183-197.
National Academy of Sciences/National Research
Council Subcommittee on Carcinogenesis, Food Protection
Committee, Food and Nutrition Board. 1961. Problems in
the Evaluation of Carcinogenic Hazard from Use of Food
Additives, Cancer Res. 21:429-456.
National Academy of Sciences/National Research
Council. 1975. Principles for Evaluating Chemicals in
the Environment. National Academy of Sciences. Washing-
ton, D.C. pp.115-133.
National Academy of Science/National Research
Council. Committee for the Revision of NAS Publ. 1138.
1977. Principles and Procedures for Evaluating the
Toxicity of Household Substances. NAS, Washington, D.C.
pp.130.
National Academy of Science/National Research
Council. Committee on Laboratory Animal Diets, Institute
of Laboratory Animal Resources. 1978. Control of Diets
in Laboratory Animal Experimentation. NAS, NRC,
Washington, D.C.
NCTR/NCI Symposium on the Use of Inbred and Outbred
Animals in Toxicological Testing. 1979. J. Toxicol.
Environ. Health, 5:1-168.
Newberne, P.M. 1974. Diets. In: Carcinogenesis
9 (ed') CRC Press'
-------
11-58
Newberne, P.M. 1975. Pathology: Studies of Chronic
Toxicity and Carcinogenicity. pp.650-656.
Newberne, P.M., and Rogers, A.E. 1976. Nutritional
Modulation of Carcinogenesis. In: Fundamentals in
Cancer Prevention. P.' Magee (ed.). University Park
Press, Baltimore, Md.
Nolen, G.A. 1972. Effect of Various Restricted
Dietary Regimens on the Growth, Health and Longevity of
Albino Rats. J. Nutrition, 102:1477-1494.
Page, N.P. (Chairman). 1976. Report of the Subtask
Group on Carcinogen Testing to the Interagency Collabo-
rative Group on Environmental Carcinogenesis. (2/10/76).
pp.24.
Page, N.P. 1977a. Chronic Toxicity and Carcino-
genicity Guidelines. J. Enviro. Path, and Toxicol.,
1:161-182.
Page, N.P. 1977b. Concepts of a Bioassay Program
in Environmental Cancer. Adv. Modern Toxicol., 3:87-
171.
•
Peck, H.M. 1974. Design of Experiments to Detect
Carcinogenic Effects of Drugs. In: Carcinogenesis
Testing of Chemicals. L. Golberg (ed.), CRC Press,
Inc., Cleveland, Ohio. pp.1-14.
Pietra, G., Spencer, K., and Shubik, P- 1959.
Response of Newly Born Mice to a Chemical Carcinogen.
Nature (Lond.), 183:1689.-
Rail, D.P. 1977. Species Differences in Carcino-
genesis Testing. In: Origins of Human Cancer, Book C,
Human Risk Assessment. H.H. Hiatt, J.D. Watson and J.A.
Winsten (eds.). Cold Spring Harbor Laboratory,
pp.1383-1390.
Rice, J.M. 1976. Carcinogenesis: A Late Effect of
Irreversible Toxic Damage during Development. Environ.
Health Perspect. 18:133-139.
Roe, F.J.C., and Tucker, M.J. 1974. Recent Devel-
opments in the Design of Carcinogenicity Tests on Labor-
atory Animals. Proc. European Soc. for the Study of
Drug Toxicity, 15:171-174.
Ross, M.H. 1961. Length of Life and Nutrition in
the Rat. J. Nutr., 75:177-210.
-------
11-59
Ross, M.H. 1976. Nutrition and Longevity in
Experimental Animals. In: Nutrition and Aging. M.
Winick (ed.). John Wiley and Sons, New York. pp.43-57.
Ross, M.H. 1977. Dietary Behavior and Longevity.
Nutr. Rev., 35:257-265.
Ross, M.H., and Bras, G. 1973. Influence of
Protein Under-and Over-Nutrition on Spontaneous Tumor
Prevalence in the Rat. J. Nutr., 103:944.
Ross, M.H., and Bras, G. 1975. Pood Preference and
Length of Life. Science, 190:165.
Ross, M.H., and Bras, G. 1976. Tumor Incidence
Patterns and Nutrition in the Rat. J. Nutr., 36:245-
259.
Ross, M.H., and Bras, G. and Ragbeer, M.S. 1970.
Influence of Protein and Caloric Intake upon Spontaneous
Tumor Incidence of the Anterior Pituitary Gland of the
Rat. J. Nutr., 100:177-189.
Sansone, E.B., Losikoff, A.M., and Pendleton, R.A.
1977a. Potential Hazards from Feeding Test Chemcials in
Carcinogen Bioassay Research. Toxicol. Appl. Pharmacol.
39:435-450.
Sansone, E.B., Losikoff, A.M., and Pendleton, R.A.
1977b. Sources and Dissemination of Contamination in
Material Handling Operations. Am. Ind. Hyg. Assoc.,
38:433-422.
Schroeder, H.A. 1973. Recondite Toxicity of Trace
Elements. In: Essays in Toxicology, Vol. 4. W.J.
Hayes, Jr. (ed.). Academic Press, New York. pp!07-199.
Shimkin, M.B. 1974. Species and Strain Selection.
In: Carcinogenesis Testing on Chemicals. L. Golberg
(ed.). CRC Press, Inc., Cleveland, Ohio. p.15.
Shimkin, M.B. 1977. Reflections on the Art of
Bioassay for Carcinogenesis. Adv. in Modern Toxicol.,
3:373-381.
Shubik, P. 1974. Interpretation of Test Results in
Terms of Significance to Man. In: Carcinogenesis
Testing of Chemicals. L. Golberg (ed.). CRC Press,
Inc., Cleveland, Ohio. pp.45-60.
Shubik, P., and Sice, J. 1956. Chemical Carcino-
genesis as a Chronic Toxicity Test: A Review. Cancer
Res. 16:728-742.
-------
11-60
Silberberg, M., and Silberberg, R. 1955. Diet and
Lifespan. Physiol. Rev. 35:347-362.
Sontag, J.M. 1977. Aspects in Carcinogen
Bioassay. In: Origins of Human Cancer. Book C-Human
Risk Assessment. H.H. Hiatt, J.D. Watson and J.A.
Winsten (eds.). Cold Spring Harbor Laboratory.
pp.1327-1338.
Sontag, J.M., Page, N.P., and Saffiotti, U. 1976.
Guidelines for Carcinogenesis Bioassay in Small Rodents.
DHEW Publ. No. (NIH) 76-801. pp.65.
Street, A.E. 1970. Biochemical Tests in
Toxicology. In: Methods in Toxicology. G.E. Paget,
(ed.). F.A. Davis Company, Philadelphia. pp.313-337.
Stula, E.F., Barnes, J.R., Sherman, H., Reinhardt,
C.F., and Zapp, J.A., Jr. 1978a. Liver and Urinary
Bladder Tumors in Dogs from 3f,3-Dichlorobenzidine. J.
Environ. Path. Tox., 1:475-490.
Stula, E.F., Barnes, J.R., Sherman, H., Reinhardt,
C.F., and Zapp, J.A., Jr. 1978b. Liver and "Lung Tumors
in Dogs from 4,4'-Methylene-Bis (2-Methylaniline). J.
Environ. Path. Tox., 1:339-356.
Todd, J.C., and Sanford, A.H. 1976. Clinical
Diagnosis by Laboratory Methods (15th ed.). I. Davidson
and J.B. Henry (eds.). Saunders, Philadelphia.
Tomatis, L. 1974. Inception and Duration of
Tests. In: Carcinogenesis Testing of Chemicals. L.
Golberg (ed.). CRC Press, Inc., Cleveland, Ohio, pp.23-
27. ' -
Tomatis, L. 1977. The Value of Long-Term Testing
for the. Implementation of Primary Prevention. In: •'
Hiatt, H.H., Watson, J.D., and Winsten, J.A., eds.
Origins of Human Cancer, Book C, Human Risk Assessment.
Cold Spring Harbor Laboratory. pp.1339-1357.
Tomatis, L., Agthe, C., Bartsch, H., Huff, J.,
Montesano, R., Saracci, R., Walker, E., and Wilbourn, J.
1978. Evaluation of the Carcinogenicity of Chemicals: A
Review of the Monograph Program of the International
Agency for Research "on Cancer (1971 - 1977). Cancer
Res. 38:877-885.
Tucker, S.M., Mason, R.L. and Beauchene, R.E. 1976.
Influence of Diet and Feed Restriction on Kidney
Function in Aging Male Rats. J. Gerontol., 31:264-270.
-------
11-61
Vettorazzi, G. 1975. Toxicological Decisions and
Recommendations Resulting from the Safety Assessment of
Pesticide Residues in Food. CRC Critical Reviews in
Toxicology. 4:125-182.
Vos, B.J. 1959. Statistics. In: Appraisal of the
Safety of Chemicals in Foods', Drugs, and. Cosmetics,
Assoc. of Food and Drug Officials of the U.S. pp.93-96.
Weil, C.S., Woodside, M.D., Smyth, H.F., Jr., and
Carpenter, C.Pl 1971. Results of Feeding Propylene
Glycol in the Diet to Dogs for Two Years. Fd. Cosmet.
Toxicol. 9:479-490.
Weisburger, E.K. 1971. Testing of New Compounds
for Long-Term Toxicity. J. Soc. Cosmet. Chem. 22:825-
838. . •..,.•
•' I. • , . !
Weisburger, J.H. 1973. Chemical Carcinogenesis.
In: Cancer Medicine. J. Holland and E.-Frei (eds.).
Lea and Febiger. Philadelphia, pp.45-89.
Weisburger, J.H. 1974. Inclusion of Positive
Control Compounds. In: Carcinogenesis Testing of
Chemicals. L. Golberg (ed.). CRC Press, Inc., Cleve-
land, Ohio. pp.29-34.
Weisburger, J.H. 1975. Chemical Carcinogenesis.
In: Toxicology: The Science of Poisons. L. Casarett and
J. Doull (eds.) McMillan. New York. pp.333-378.
Weisburger, J.H. 1976. Bioassays and Tests for
Chemical Carcinogens. In: Chemical Carcinogens. C.E.
Searle (ed.). ACS Monograph 173. pp.1-23.
Weisburger, J.H., and Weisburger, E.K. 1967. Test
for Chemical Carcinogens. Methods Cancer Res. 1:307-
398. .
Williams, R.T. 1978. Species Variation in 4the
Pathways of Drug Metabolism. Environ. Health Perspect.,
22:133-138. N
World Health Organization. 1961. Evaluation of
Carcinogenic Hazards in Food Additives. Wld. Health
Org. Tech. Rep. Ser. No. 222. Genea. pp.3-23.
World Health Organization. 1967. Procedures for
Investigating Intentional and Unintentional Food
Additives. Wld. Health Org. Tech. Rep. Ser. No. 348.
Geneva, pp.
-------
11-6?,
World Health Organization. 1969. Principles for
the Testing and Evaluation of Drugs for Carcinogenicity.
Wld. Health Org. Tech. Rep. Ser. No. 426. Geneva, pp.
World Health Organization. 1978a. Acute, Subacute
and Chronic Toxicity Tests. In: Principles and Methods
for Evaluating the Toxicity of Chemicals. Geneva.
pp.95-115.
World Health Organization. 1978b. Carcinogenicity
and Mutagenicity. In: Principles and Methods for
Evaluating the Toxicity of Chemicals. Geneva, pp.236-
291.
Zbinden, G. 1976. Progress in Toxicology, Special
Topics, Vol.2. Springer-Verlag, New York. pp.8-18.
Zwickey, R.E., and Davis, K.J. 1959. Carcinogen-
icity Screening. In: Appraisal of the Safety of
Chemicals in Poods, Drugs and Cosmetics. The Associa-
tion of Food and Drugs Officials of the United States.
pp.79-82.
-------
III. Economic Aspects of the Proposed Chronic Health Effects
Test Standards
A. Summary
B. Methodology and Assumptions
1. Methodology
2. Assumptions Concerning Study Design
a. Species
b. Group Size
c. Control Groups
d. Route(s) of Administration
e. Duration of Treatment
f. Number of Dose Levels
g. Number of Animals Purchased
3. Costing Assumptions
a. Overhead Rate
b. Pathologist Requirements
c. Veterinarian
d. Technical Employee (Necropsy)
e. Technical Employee (Animal Care)
f. Salary Rates
g. Variable Costs
h. Caging
i. Animal Space Requirements
4. Items Excluded From Estimates
a. Transitional Costs
b. Regulatory Liaison
c. Chemical Characterization
d. Ancillary Studies
e. Stability Studies, Contaminant Analysis,
Other Studies
5. Variations In Costs
6. Use Of Ranges
C. Fixed Costs
1. Summary
2. Protocol Design and Study Submission
3. Project Management and Preparation of Final Report
4. Statistical Analysis
-------
D. Oncogenic Effects
1. Summary
2. Variable Costs
a. Animal Procurement
b. Rations
c. Animal Care
i. Clinical Examination
ii. Observation
iii. Feeding
iv. Cage Cleaning
v- Summary Chart
d. Clinical Laboratory Tests
e. Microscopic Examinations
f. Necropsy and Histological Preparation
3. Prechronic Testing Costs
E. Chronic Effects
1. Summary
2. Variable Costs
a. Animal Procurement
b. Rations (rat only)
c. Animal Care (rat only)
i. Clinical Examination
ii. Observation
iii. Feeding
iv. Cage Cleaning
v. Summary Chart
d. Rations and Animal Care (dog)
e. Clinical Laboratory Tests
f. Microscopic Examinations
g. Necropsy and Histological Preparation
3. Prechronic Testing Costs
F. Combined Chronic Effects
1. Summary
2. Variable Costs
3. Prechronic Testing Costs
4. Cost Savings Due To Combined Test
G. Other Data on Testing Costs
H. Cost of Alternative Standards
-------
III. ECONOMIC ASPECTS OF THE PROPOSED CHRONIC HEALTH EFFECTS
TEST STANDARDS
A. SUMMARY
The estimated costs of testing a chemical according to the
proposed standards are provided in Table 1 below. I/
TABLE 1
Summary of Costs
Oncogenic Chronic Combined
Tests Tests Tests
Fixed Costs; $ 45,000 $ 45,000 $ 45,000
Variable Costs;
Animal Procurement $ 2,500 $ 10,300 $ 12,000
Rations 1,300 1,500 2,400
Animal Care 96,300 270,800 339,500
Clinical Laboratory 2,600 21,100 24,700
Tests
Microscopic 90,000 75,000 152,100
Examination
Necropsy and 150,000 105,000 212,600
Histological
Preparation
Total Variable
Cost $342,700 $483,700 $743,300
Total Cost $387,700 $528,700 $788,300
Prechronic
Testing Costs; 50,900 100,500 125,900
Total Costs
including
Prechronic $438,600 $629,200 $914,200
I/ The following terms will be used throughout this chapter;
Term Refers to
Oncogenic Oncogenic effects test
standard
Chronic Non-oncogenic chronic
toxicity standard
Combined Combined chronic effects
test standard
-------
Ill - 2
Most estimates of testing costs provide only a total
cost for a test, and give little or no detail on costs for
subparts of the test. EPA has adopted a more detailed ap-
proach in order to provide a basis for:
1) discussion and improvement of the cost
estimates, and
2) calculation of the costs of alternatives
to the proposed test standards.
The following points, however, regarding the cost estimates
should be emphasized:
. The estimated costs are those of meeting the minimum
requirements of the test standards. Actual testing
costs will be greater to the extent that optional
testing is performed.
. Actual costs of long-term testing will vary substan-
tially, depending on a number of factors, including,
for example, the chemical tested, laboratory perform-
ing the test, number of dose levels, species, route(s)
of exposure, extent of pathology conducted, duration
of tests, inclusion of ancillary studies, and other
factors.
. As indicated by the detailed calculations in the sec-
tions which follow, EPA's cost estimates are calculated
based on a number of assumptions concerning labor produc-
tivity and wage and overhead rates. These assumptions
represent a consensus judgment among EPA oncologists,
toxicologists, and pathologists derived from personal
experience in conducting chronic health effects testing.
In cases where published numbers or other documented
numbers were judged reliable by EPA, these numbers were
used and are referenced accordingly.
-------
Ill - 3
. Costs of Good Laboratory Practice (GLP) items in the
test standards do not receive explicit treatment
in most cases in the estimates, and thus are included
as part of overhead.
. Many cost items are included as part of overhead in
the estimates and therefore do not receive explicit
treatment. EPA solicits comment as to areas which
merit explicit treatment.
. Section 4 test rules will supply additional informa-
tion (e.g. route(s) of exposure) and may modify the
test standards to suit particular chemicals. EPA
will develop cost estimates to account for these
additions and modifications at the time test rules
are proposed and promulgated.
The above costs are on a per chemical tested basis. The
total annual cost of TSCA Section 4 test rules requiring these
tests will depend on the per chemical costs and the identity
and number of chemicals tested for particular effects. The
latter items are undecided and thus it is difficult to esti-
mate the testing costs that will result. If one assumes that
approximately 15 to 30 chemicals a year will be tested under
Section 4 test rules using these particular test standards,
and test costs will range from approximately $400,000 (for an
oncogenic study where prechronic testing has already been per-
formed) to $900,000 (for a combined study where prechronic
testing is required), the total annual cost is estimated to
range from $6 to $27 million.
-------
Ill - 4
B. METHODOLOGY AND ASSUMPTIONS
1. Methodology
Costs were determined by separating each of the standards
into the following subparts and estimating the cost of each:
. Fixed Costs (Section C)
Protocol Design and Study Plan Submission
Project Management and Preparation of Final
Report
Statistical Analysis
. Variable Costs (Sections D-2, E-2, F-2)
Animal Procurement
Rations
Animal Care
Clinical Laboratory Tests
Microscopic Examinations
Necropsy and Histological
Preparation
. Prechronic Testing Costs (Sections D-3, E-3, F-3)
Calculation of costs in each of these subparts consisted
of three elements:
. Determination of the requirements of the rule;
. Determination of assumptions concerning labor
productivity, wage rates, etc.; and
. Calculation of cost.
Costs were then summed to give the total for each of the test
standards.
2. Assumptions Concerning Study Design
The cost estimates are designed to estimate the cost of
satisfying the minimum requirements of the standards. EPA
recognizes that some sponsors may choose to perform testing
in addition to what is assumed for cost estimation. This
testing is regarded in these estimates as optional and is
not included. EPA solicits comment as to whether any of
this additional testing is really not optional in the sense
that sponsors would be induced by EPA regulatory require-
ments, particularly performance requirements, to perform
this testing.
A number of assumptions about key variables were made in
estimating the costs of the test standards. The key variables
-------
Ill - 5
and the assumption made for each test standard are discussed
below. Each section first sets forth the requirements of the
standard and then explains how the item is treated in the
cost estimates. Where EPA expects that sponsors will perform
the testing in a given manner, this is also noted.
a. Species
The tester must use the following species for
each test:
Oncogenic: at least two rodent species, the
laboratory mouse and rat
Chronic: at least two mammalian species, a
laboratory rat and a non-rodent
Combined: at least three mammalian species:
two rodent species, the laboratory
mouse and rat, and a non-rodent.
The dog is recommended as the non-rodent. Alternate
species may be utilized if justified to EPA. The cost
estimates are based on the minimum required number of
species for each test and have assumed use of the mouse,
rat, and dog. EPA believes that the large majority of
testing will be conducted with these speciesi because of
the requirements and recommendations of the standards,
and because of the scientific and economic advantages
of these species.
b. Group Size
The minimum group size requirements of the test stan-
dards are as follows:
oncogenic: 50 rodents per group
chronic: 58 rats per group, 6 dogs per group
combined: 50 mice, 58 rats, and 6 dogs per group.
The standards require that when interim kill is planned, the
number of animals at the start be increased by the number
of animals scheduled to be killed before completion of the
study.
The cost estimates are based on the minimum allowable
group size and assume no interim, kill. In assuming 58
rats per group, EPA assumed that the minimum of 8 rats
per group will survive the clinical laboratory tests and
need not be sacrificed (see Section II for further dis-
cussion) .
-------
Ill - 6
c. Control Groups
The standards require the use of matched control
groups. I/ If the test substance is administered to the
exposed groups by the use of a vehicle whose toxic pro-
perties are not known, the tester may, at his discretion,
use a negative (untreated) group. 2/ The EPA may require
a positive control group 3/ for particular chemicals when
the sensitivity of the test animal cannot be documented
for the chemical class to which the test substance belongs.
The cost estimates assume that only a matched control
group will be used. The rationale for exclusion of the
cost of negative (untreated) and positive control groups is
as follows:
. The addition of a negative control group to a
study would increase the cost by a considerable
amount. For this and scientific reasons, EPA
expects that if a vehicle is used, the vehicle
chosen will be well-characterized and therefore
a negative control group will not be used in most
cases.
. A positive control group may be required by EPA.
Where this requirement is specified in a parti-
cular test rule, the cost of this group will
be included in the cost estimates accompanying
the particular test rule. Both because of the
considerable increase in cost due to positive controls
and because of the safety problems associated with
using known toxic agents in the laboratory, EPA
expects that the use of positive control groups,
where not required by EPA, will be relatively in-
frequent and therefore has not included their cost
in these generalized cost estimates.
I/ A matched control group is identical in every respect
to an exposed group except that it is not exposed to
the test substance. If a vehicle is used in administering
the test substance, the matched control group would
receive the vehicle but not the test substance.
2/ A negative (untreated) control group is a control group
which receives neither the test substance nor a vehicle.
3/ A positive control group is a control group which does not
receive the test substance but instead receives a known
toxic agent. When used, the positive control group
serves as an internal quality control group to ascertain
whether the test animals are sensitive to known toxic
agents and whether the test strain or species reacts
similarly to another strain or species when exposed
to the same known standard toxicant..
-------
Ill - 7
d. Route(s) of Administration
The routes of administration will be specified in the
individual test rules and EPA's cost estimates accompanying
those test rules will reflect the route(s) of administration.
Since most studies are expected to be feeding studies, the
cost estimates presented here have been based on the costs
of a feeding study.
e. Duration of Treatment
The test standards require:
oncogenic: 24-30 months for mouse and rat
(24 month minimum)
chronic: minimum of 24 months for dog,
30 months for rat,
combined: minimum of 24 months for mouse
and dog, 30 months for rat.
The cost estimates are based on the minimum allowable test
period. EPA anticipates most testing will conclude after
this minimum period.
f. Number of Dose Levels
The standards require:
oncogenic and chronic: at least 3 dose levels
(in addition to controls)
combined: at least 3 dose levels (in addition to
controls) for mouse and non-rodent; at
least 4 or 5 dose levels (in addition to
controls) for rat.
The cost estimates use 3 dose levels for all tests and
species (plus controls) except for the combined test for the
rat because this is the minimum required. In the combined
study for the rat, the estimates assume that 4 dose levels
(plus controls) will be used as often as 5 dose levels (plus
controls). An average of these two levels is therefore used for
these cost estimates. This approach is employed because EPA
is unable to predict at this time how often 4 dose levels
will be required versus how often 5 dose levels will be re-
quired.
g. Number of Animals Purchased
Although the standards do not address the issue of the
number or percentage of animals in control groups that
must survive to termination of the study, good survival
is clearly important. ' One of the ways to accomplish this
is through starting the study with healthy animals. EPA
-------
Ill - 8
believes many laboratories will order more animals than
will be used in the study for prescreening purposes. These
cost estimates assume test laboratories will purchase 20%
more animals than actually started with the test to assure
starting the study with healthy animals.
3. Costing Assumptions
A number of specific cost assumptions were made in esti-
mating the costs of each test standard. The cost assumptions
which apply to all three test standards are explained below.
a. Overhead Rate
An overhead rate of 100% is applied to all labor costs
in the cost estimates. This includes costs that testing
laboratories may separate into overhead, general and admin-
istrative, and fee or profit. There are three areas in which
this assumption may be limited:
1) Overhead is used in the estimates to capture all
costs not explicitly estimated. In some cases, it
may be desirable to develop a specific cost estimate
for items now lumped into overhead.
2) Overhead rates specific to particular parts of the
tests could be used, as opposed to a uniform rate.
If significant differences in overhead exist for
different parts of the test, use of specific rates
could improve the cost estimates for particular parts
of the standards and for variations in the rule require-
ments.
i
EPA is uncertain as to whether significant overhead
differences can be identified for different parts of a
test. Some factors suggests rough uniformity of over-
head - e.g., highly-paid pathologists work with expen-
sive microscopes. Other factors suggest that use of a
uniform rate may be inappropriate - e.g., social
security expenses and the cost of running a payroll
system vary more with the number of employees, than
with the wage bill, and thus do not increase pro-
portionally with wage rates. :Given other causes of
variations in the cost estimates, it is doubtful that
use of specific rates will significantly improve the
estimates.
3) Assuming a single overhead rate is .appropriate, then
100% may still not be the best estimate. However, the
fact that the cost estimates here are within 15% of
those of the National Cancer Institute when adjusted
-------
Ill - 9
for differences in the standards (see "Cost of
Alternative Standards" section) lends support
to the belief that the use of 100% (in conjunc-
tion with the manpower estimates and wage rates
used here) is probably reasonable.
b. Pathologist Requirements
The standards provide that a Board-Certified or Board-
Eligible pathologist with a minimum of three years of ex-
perience in pathology of the species of laboratory animals
to be used, may directly supervise other doctorate patholo-
gists for conducting certain procedures. The cost estimates
do not specify separate categories of pathologists with
different salary rates, but merely use one category and rate.
The salary rate used is intended as an average of the differ-
ent rates that exist for these categories. Comment is soli-
cited as to whether EPA should attempt to use a weighted
average.
c. Veterinarian
The standards require that a veterinarian who is Board-
Certified or eligible for certification by the American College
of Laboratory Animal Medicine, and who has a minimum of two
years of experience in laboratory animal science, be responsible
for the health status and care of all test animals. The cost
of the veterinarian is included in the overhead estimate, and
not costed explicitly.
d. Technical Employee (Necropsy)
Necropsies must be performed by a pathologist or by a
technical employee under the personal supervision of a
pathologist. Such technical employee must be certified by
the American Society of Clinical Pathology (HTASCP), or
have equivalent training and experience. The cost estimates
assume that most, if not all, of :the necropsies are performed
by such technical employees. The time required for super-
vision by pathologists is included in the "Microscopic Ex-
amination" section; the assumption concerning the number of
sections per year that a pathologist can examine microscopically
is based on 4 hours per day at the microscope with a portion
of the remaining time to be spent in necropsy supervision.
e. Technical Employee (Animal Care)
Technical employees responsible for the daily observa-
tion and care of test animals must be certified or eligible
for certification by the American Association of Laboratory
-------
Ill - 10
Animal Science (LTAALAS), or have equivalent training and
experience. The cost estimates assume that these technical
employees shall conduct the twice daily observations and the
biweekly clinical examinations, but that the feeding and
cage cleaning will be conducted by personnel with less train-
ing (termed "technicians" in the estimates).
f. Salary Rates
The cost estimates assume the following average annual
salary rates for each type of employee:
Pathologist $50,000
Senior Statistician $40,000
Intermediate Statistician $24,000
Junior Analyst $15,000
Technical employee (HTASCP certified) $15,000
for necropsy
Animal weight, clinical exam, $12,000
and daily observation
technical employee (LTAALAS
• certified, or eligible)
Technician to clean cages and feed $ 8,000
animals
g. Variable Costs
All variable costs are assumed to be strictly proportion-
al to the size of the experiment; e.g., doubling the number of
animals doubles the variable costs. While this assumption is
probably not valid over a large range, for a small range
(± 25%) it may be more reasonable. This range is sufficient
to capture the variation in cost for many of the possible
modifications to the standards (See "Cost of Alternate Stan-
dards" Section H, Table 5).
h. Caging
These estimates assume 5 mice per cage and 2 rats per
cage. Two rats per cage is an average, with actual caging
practices varying from 1 to 3 rats per cage. Since many of
the animal care operations are costed on a per cage basis,
the effect of this assumption is to make animal care for rats
much more costly than for mice and possibly more costly than
will be the case in practice.
i. Animal Space Requirements
Animal space requirements are included as part of overhead
This was done for two reasons:
-------
Ill - 11
1) EPA believes this to be customary industry practice,
although it recognizes some laboratories will explicitly
cost animal space.
2) EPA estimates that the cost for animal space will be
small relative to the total costs of the standards.
If one assumes about 100 square feet for 400 mice,
250 square feet for 400-500 rats, and 700 square feet
for 48 dogs, and a cost of $6/square foot/year, the
costs are $1,200, $3,000 and $8,400 respectively for
a two year test.
4. Items Excluded From Estimates
The cost of a number of items have been excluded from these
cost estimates. The items and the rationale for exclusion are
explained below.
a. Transitional Costs
The cost estimates represent only the marginal cost of an
additional chronic effects test in a laboratory, and not the
transitional costs that will be incurred the first few times a
laboratory conducts tests in accordance with these standards.
These transitional costs, e.g., design of a protocol, establish-
ment of a quality assurance unit, and development of computer
programs for statistical analysis, will be incurred in varying
degrees from laboratory to laboratory depending on current
practices and compliance with FDA Good Laboratory Practice stan-
dards.
b• Regulatory Liaison
No estimate is provided for the costs of regulatory liaison
with EPA. The extent of these costs will depend on the degree of
variance from the EPA standard of the submitted protocol. The in-
formal nature of the mechanisms for handling submitted study plans
should minimize these costs.
c. Chemical Characterization
The required chemical characterization for chronic effects test-
ing would in almost every case have been performed prior to short
term testing of the chemical and would not be done again prior to
chronic health effects testing. Therefore, no cost is included in
these estimates for chemical characterization.
d« Ancillary Studies
The cost of pharmacokinetic and metabolic studies are not in-
cluded in the estimates because they are optional studies undertaken
at the discretion of the sponsor.
-------
Ill - 12
e. Stability Studies, Contaminant Analysis, Other Studies
Section 772.113 requires that certain chemical tests be per-
formed to determine:
1) the stability of the test substance in the
test mixture;
2) methods to assure homogeneous mixing of the
test substance in the test mixture (feed or
other carrier material);
3) concentration of test substance in each test
mixture;
4) contaminants in the feed or vehicle; and
5) nutrient content of feed.
The costs of these tests are expected to vary considerably
from chemical to chemical and may be quite expensive. These
variations are due to the analytical methodologies required
for a particular chemical and the extent to which the analy-
tical methodology needs to be developed. Because of this
variability per chemical, EPA may address this item in greater
detail when chemicals are selected for testing in Section 4
test rules.
5. Variations In Costs
The actual cost of testing that will be conducted in re-
sponse to Section 4 testing requirements is likely to vary
considerably from the estimates provided here for oncogenic,
chronic, and combined testing. There are three reasons for
these variations:
1) The cost of conducting exactly the same test
will vary from laboratory to laboratory due to
differences in labor productivity, salary and
overhead rates, equipment, etc.
2) The cost of testing chemicals in the same
laboratory will vary depending on a number of
factors. These factors include: the number of
dose levels, species, route(s) of exposure, ex-
tent of pathology conducted, duration of tests,
and inclusion of ancillary studies.
3) The cost estimates probably make some estimation
errors and oversimplify some factors.
The cost estimates are those to meet the minimum requirements
of the standard. The actual cost of testing will be greater to the
extent that optional testing is conducted and transitional costs
are incurred.
-------
Ill - 13
6. Use Of Ranges
While ranges have not been used in the estimates, every
number used in the calculations below could undoubtedly be more
accurately reflected by a range. While this might initially
seem desirable, there are two practical problems in implement-
ing the concept. First, the information requirements and num-
ber of calculations of the costing methodology used here are
already substantial; use of ranges would increase these. Second,
the appropriate range estimate for subparts of the standard or
for the entire standard would be difficult to determine. This
is true because:
1) Range estimates almost always have an implicit or
explicit confidence level associated with them —
e.g., 95% of the time the number will fall within
the range.
2) When range estimates with confidence levels are
added together, the range for the sum with the
same confidence level (e.g. 95%) is more narrow
than that obtained by adding all the low ends of
the ranges for one estimate and the high ends for
the other estimate. The range estimate produced
by all the low ends and high ends would be too
broad to be meaningful.
For these reasons, the appropriate range estimates are very
difficult to determine.
-------
Ill - 14
C. FIXED COSTS
1. Summary
The estimates assume that the following testing costs
do not vary with the size of the experiment or the type
of test (oncogenic, chronic, or combined):
Protocol Design and Study Plan
Submission $ 8,300
Project Management and Preparation of
Final Report $20,000
Statistical Analysis $18,200
$46,500,
or approximately
$45,000
Clearly, this is somewhat of an oversimplication —
these costs will be greater for a combined study than for
either an oncogenic or chronic study, and may differ somewhat
with the size of the experiment; however, the differences are
not expected to be significant.
The rule requirements, assumptions, and calculations
for these items are explained below.
2. Protocol Design and Study Submission
Rule Requirements; The rule requirements for study plan
submission are stated in detail in Good Laboratory Practices
Subpart B, Section 772.110-l(g)(1) and in the General Chronic
Health Effects Standards, Subpart D, Section 772.113-l(f).
Assumptions; The following assumptions were used in calculating
the cost of protocol design and study submission:
Most of the effort in protocol design is estimated
to be a one-time transitional cost, i.e., a cost
incurred the first time a laboratory conducts a
test in accordance with these standards. The cost
to modify this previously developed protocol to
apply to the particular chemical at issue should
be relatively minor, approximately one person-month
of senior personnel effort. It is the latter cost
which is estimated here.
Salary of senior personnel required is approximately
$50,000. This represents an approximate average
of those personnel required (study director, senior
statistician, pathologist, etc.).
-------
Ill - 15
Cost;
T7T2 person-year x ($50,000 x 100% overhead) = $8,333 or
approximately
$8,300
3. Project Management and Preparation of Final Report
Rule Requirements; The explicit requirements in the
standards for project management are contained in the Good
Laboratory Practice Standards, Subpart B, Section 772.110-1
(c)(2), (3), and (4). This includes requirements for testing
facilities management, a study director, and a quality assurance
unit. The reporting requirements for the final report are
specified in the general requirements for chronic health
effects studies contained in Subpart D, Section 772.113-1(j).
The Agency plans to propose specific standards for data
formatting later this year.
Assumptions; The following assumptions were used in
calculating the cost of project management and preparation of
final report;
Transitional costs (e.g., establishment of a
quality assurance unit or adoption of new
reporting requirements) are not included in the
estimates.
A precise estimate of report costs can not be made
at this time. The analysis of data formatting pro-
cedures to be conducted later this year should provide
a basis for estimating these reporting costs.
Cost; A preliminary estimate is that the costs of project
management and reporting will be approximately $20,000. EPA
solicits comment on how this cost estimate might be refined.
*• Statistical Analysis
Rule Requirements; Information specified in Subpart D,
Section 772.113-1(j)(2)(ii) must be reported for each test
group. In many cases, this will require the computation
of medians, means, standard deviations, and other statistical
measures.
Assumptions; The following assumptions were used in calcu-
lating the cost of statistical analysis;
-------
Ill - 16
Computerized statistical programs will be the least
costly way of performing this analysis. Computer
costs are expected to be minimal, once the statistical
programs are operational on the computer. The costs
to develop these programs are transitional costs
and are not included in the estimates.
The approximate staff time requirements are as follows:
senior statistician
intermediate statistician
junior analysts
The salary rates are as follows
senior statistician
intermediate statistician
junior analysts
Cost:
1 person-month
1 person-month
3 person-months
$40,000
$24,000
$15,000
1/12 person-year x ($40,000/yr x 100% overhead)
1/12 person-year x ($24,000/yr x 100% overhead)
3/12 person-year x ($15,000/yr x 100% overhead)
Total
$ 6,666
4,000
7,500
$18,166 or
approximately
$18,200
-------
Ill - 17
D. ONCOGENIC EFFECTS
1. Summary
The estimated costs of the oncogenic effects test standards
are shown below in Table 2.
Table 2
Costs of Oncogenic Effects Test Standards
Mouse Rat Total
Fixed Costs; - - $ 45fOOO I/
Variable Costs;
Animal Procurement $ 1,100 $ 1,400 $ 2,500
Rations 300 1,000 1,300
Animal Care 32,800 63,500 96,300
Clinical Laboratory Tests 1,300 1,300 2,600
Microscopic Examinations 45,000 45,000 90,000
Necropsy and Histological 75,000 75,000 150,000
Preparation
Total Variable Costs $155,500 $187,200 $342,700
Total Oncogenic Costs - - $387,700
Prechronic Testing Costs; - - 50,900
Total Costs Including $438,600
Prechronic
These estimates assume testing in 400 rats and 400 mice.
This assumes testing at 3 different dose levels plus a control
group, 50 animals in each group and testing of both sexes —
(3 dose levels + 1 control group) x 50 animals x 2 sexes =
4 x 50 x 2 = 400 animals of each species.
These estimates assume that the duration of the test is 2
years. In the cost calculations below, the duration factor will
be expressed as 2 years, 104 weeks, or 730 days, depending on how
other assumptions are expressed.
The details of how the variable costs of oncogenic testing
and associated prechronic testing were calculated are provided
below.
T/ See Section C, "Fixed Costs" for details.
-------
Ill - 18
2. Variable Costs
Variable costs of oncogenic testing include the costs of ani-
mal procurement, rations, animal care, clinical laboratory tests,
microscopic examinations, and necropsy and histological prepara-
tion. Each of these costs are calculated separately, and separ-
ate estimates are presented for mice and rats. These calculations
are discussed below.
a. Animal Procurement
Rule Requirement; EPA does not stipulate that a specific
strain or stock of rats or mice must be used, nor does EPA
express a preference among inbred, outbred, and hybrid
strains. However, the rule does require that test animals be
from established strains and/or stocks. As part of the study
plan submission, the sponsor must provide historical data on
the lifespan and types and incidences of disease for the se-
lected strains.
Assumptions; The following assumptions were used in cal-
culating the cost of animal procurement;
. Mice cost $2.35 each;
. Rats cost $2.85 each;
. 480 rats and 480 mice will be purchased.
It is assumed that 20% more rats and mice will be purchased
than started on the test so that testing can begin on healthy
animals. (See Section B, page III-8).
Cost; Animal procurement is estimated to cost $2,500, cal-
culated as follows;
480 mice x $2.35/mouse = $1,130
480 rats x $2.85/rat = $1/370
$2,500
b. Rations
Rule Requirement; A standardized rodent diet contain-
ing specified nutrient levels and produced from certain feed
stocks or ingredients (see Appendix A, Subpart D, Section
772.113-1) is required unless information is provided by the
sponsor to justify deviation. Diets for species other than
the rodent must be approved by EPA.
-------
Ill - 19
Assumptions; The following assumptions were used in cal-
culating the cost of rations:
. Mice consume 4.5 gm of food per day;
. Rats consume 13.5 gm of food per day;
. Rations cost $235/ton. I/
Cost: Rations are estimated to cost $1,360, calculated as
follows:
400 rats x 13.5 gm/day x 730 days = 3,942 kg.
3,942 kg. x .0011 tons/kg, x $235/ton = $1,020
400 mice x 4.5 gm/day x 730 days = 1,314 kg.
1,314 kg. x .0011 tons/kg, x $235/ton = $ 340
$1,360 or
approximately
$1,300
c. Animal Care
Animal care consists of clinical examination, observation,
feeding, and cage cleaning.
i. Clinical Examination
Rule Requirements; Clinical examination, including the
weighing of each animal, must be conducted at least once a
week during the first 13 weeks, and every two weeks there-
after. Clinical examination must include observation re-
lating to food and water consumption, morbidity, mortality
and causes thereof, pharmacologic effects, and behavioral
changes.
Assumptions; The following assumptions were used in cal-
culating the cost of clinical examinations.
A technical employee can weigh and examine 25
animals per hour;
The salary of the technical employee is $12,000
per year.
Cost; Clinical examination is estimated to cost $21,600,
calculated as follows:
I/ Low bid to lupply NIH-31 rodent diet containing 18% crude
~" protein to NIH for 12 months was $235.20 per ton. (Letter
from Joseph Knapka, NIH, to Carl Morris, EPA dated
January 11, 1979).
-------
Ill - 20
Mice;
400 animals ± 25 animals/hour = 16 hours/week
Weekly exams for 13 weeks:
13 weeks x 16 hours/week 208 hours
Biweekly exams for 91 weeks
(91 weeks * 2) x 16 hours/week 728 hours
936 hours*
936 hours
2,080 hours/year = .45 person-years
. 45 person-years x ($12,000/yr. x 100%
overhead) = $10,800
i
Rats;
The above calculation applies to rats
as well - i.e., the cost is; $10,800
Total $21,600
ii. Observation
Rule Requirement; A technical employee must observe the
test animals every 12 hours throughout the test period.
Assumptions; The following assumptions were made in cal-
culating the cost of observation;
The time requirements for observation are assumed to
be as follows:
- For the first 18 months of the study, 1 hour per day
for 400 mice and 3 hours per day for 400 rats;
- For the last 6 months of the study where the health
status of more animals will be problematical, 3 hours
per day for 400 mice and 5 hours per day for 400 rats.
. The salary of the technical employee is $12,000/yr.
»
Cost; The estimated cost of observation is $42,200, cal-
culated as follows:
Mice;
First 18 months:
(365 days x 1.5 years) x 1 hour/day = 547.5 hours
Last 6 months:
(365 days x .5 years) x 3 hours/day = 547.5 hours
1095 hours
-------
Ill - 21
1,095 hours
2,080 hours/year = .53 person-years
.53 person-years x ($12,000/yr. x 100%
overhead) = $12,720
or approximately
$12,700.
Rats;
First 18 months:
(365 days x 1.5 years) x 3 hours/day = 1,642.5 hours
Last 6 months:
(365 days x .5 years) x 5 hours/day = 912.5 hours
2,555 hours
2,555 hours
2,080 hours/year =1.23 person-years
1.23 person-years x ($12,000/yr. x 100%
overhead) = $29,520 or
approximately
$29,500.
Total $42,200
iii. Feeding
Rule Requirement; For a feeding study, the test sub-
stance must be administered ad libitum (continuously avail-
able).
Assumptions; The following assumptions were made in cal-
culating the cost of feeding:
i
1 cage will contain 5 mice and
1 cage will contain 2 rats;
A technician can process 30 cages/hour for feeding
the test animals;
The technician's salary is $8,000/year;
The feed containers for mipe and rats will be
refilled once every third day. ' ;
Cost; The estimated cost of feeding is $17,500, calcu-
lated as follows:
Mice;
400 mice * 5 mice/cage = 80 cages
80 cages * 30 cages/hour = 2.67 hours/feeding.
-------
Ill - 22
2.67 hours/feeding x (730 days x 1/3)
650 hours
2,080 hours/year = .31 person-years
.31 person-years x ($8,000/yr. x 100%
overhead) =
Rats:
650 hours
$4,998 or
approximately
$5,000.
400 rats f 2 rats/cage = 200 cages
200 cages * 30 cages/hour = 6.67 hours/feeding
6.67 hours/feeding x (730 days x 1/3) = 1,623 hours
1,623 hours
2,080 hours/year = .78 person-years
.78 person-years x ($8,000 x 100%
overhead) =
Total
$12,480 or
approximately
$12,500
$17,500
iv. Cage Cleaning
Rule Requirement; The recommendations of HEW Pub-
lication No. (NIH) 74-23, "Guide for the Care and Use of
Laboratory Animals" apply except where standards are spe-
cified in the Animal Welfare Act of 1970 (9CFR Part 3).
HEW Publication No. 74-23 states that for routine mainte-
nance of small rodents, one to three changes per week
of litter or bedding should suffice.
Assumptions; The following assumptions were made in
calculating the cost of cage cleaning:
1 cage will contain 5 mice;
1 cage will contain 2 rats;
Cage cleaning will occur twice a week (average of the
one to three changes per week cited above);
A technician can clean 30 cages per hour;
The technician's salary is $8,000/year.
Cost; The estimated cost of cage cleaning is $15,000,
calculated as follows:
-------
Ill - 23
Mice;
400 mice * 5 mice/cage = 80 cages
80 cages * 30 cages/hour = 2.67 hours/cage cleaning
2.67 hours/cage cleaning x (104 weeks x 2 times) = 555 hours
555 hours
2,080 hours/year = (.27 person-years)
.27 person-years x ($8,000/yr. = $4,320 or
100% overhead) approximate-
ly $4,300
Rats;
400 rats ± 2 rats/cage = 200 cages
200 cages * 30 cages/hour = 6.67 hours/cage cleaning
6.67 hours/cage cleaning x (104 weeks x 2) = 1,387 hours
1,387 hours
2,080 hours/year = .67 person-years
.67 person-years x ($8,000 x 100% overhead) = $10,720 or
approximate-
ly $10,700
Total $15,000
v. Summary Chart
Summarizing the above costs, the estimated costs of animal
care are as follows:
Mouse Rat Total
Clinical Examination $10,800
Observation 12,700
Feeding 5,000
Cage Cleaning 4,300
Total Animal Care Costs $32,800
d. Clinical Laboratory Tests
Rule Requirements: The following qu
determinations are required twice durin
$10,800
29,500
12,500
10,700
$63,500
$21,600
42,200
17,500
15,000
$96,300
antitative hematologic
g the study on a mini-
mum of 8 predesignated animals per group: hematocrit, hemo-
globin, erythrocyte count, total and differential leukocyte
-------
Ill - 24
counts, platelet count, and prothrombin and clotting time. If
hematologic evidence of anemia is present at one year, reti-
culocyte counts must be performed within one week of the de-
termination.
Assumptions; The following assumptions were made in cal-
culating the cost of clinical laboratory tests:
The cost of a hematologic determination is $10/animal;
8 animals will be tested per group (this is the mini-
mum requirement of the standard).
Cost; The estimated cost of clinical laboratory examina-
tion is $2,600, calculated as follows:
Mice;
' 8 animals/group x 8 groups = 64 animals
64 animals x 2 times = 128 tests
128 tests x $10/test = $1,280
Rats;
The above calculation applies to rats as
well — i.e., the cost is: $1,280
Total $2,560 or
approximately
$2,600
e. Microscopic Examinations
Rule Requirements; Microscopic examination must be per-
formed on all tissues described in Subsections (b)(2)(ii),
(b)(2)(v), and (b)(2)(vi) of Section 772.113-2. Subsection
(b)(2)(v) requires that when there is clinical evidence of
specific toxicologic or pharmacologic effects related to spe-
cific target organs, the necropsy and microscopic examinations
of these target organs must be conducted in greater detail.
Subsection (b)(2)(vi) requires that sections of tissues
from a minimum of ten animals from each test group and in all
animals in which clinical or grossly observable evidence of
disease is present be examined microscopically. If .micro-
scopic examination reveals evidence of disease in any of the
tissues, then these target tissues must be examined micro-
scopically in all animals.
Assumptions; The following assumptions were used in cal-
culating the cost of microscopic examinations:
The routine microscopic examination of Subsection (b)
(2)(ii) represents approximately 50 sections per ani-
mal .
-------
Ill - 25
The additional examinations required by Subsection
(b)(2)(v) represents approximately a 10% increase
over the routine requirements, or 5 additional
sections,
The special examinations of Subsection (b)(2)(vi)
represent only 1-2 sections. These 1-2 sections will
be examined in anywhere from 20-100% of the animals.
This requirement should represent, on average, no
more than 1 section per test animal,
Total sections required per animal are therefore as
follows:
routine: 50
additional: 5
special: 1
56 sections.
A pathologist can read slides from 50,000 sections
per year. This figure is from an NCI survey which
assumed four hours a day at the microscope. This
figure has been used in the literature to calculate
pathology workload I/
The salary of the pathologist is $50,000 per year.
Cost; Microscopic examination is estimated to cost
$90,000, calculated as follows:
Mice;
400 animals x 56 sections/animal = 22,400 sections
22,400 sections
50,000 sections/year = .45 person-years
.45 person-years x ($50,000/year x 100%
overhead) = $45,000
Rats;
The above calculation applies to rats as well—
i.e., the cost is: $45,000
$90,000
f. Necropsy and Histological Preparation
Rule Requirements; The rule requirements are stated
in detail in the standards. The requirements provide for:
I/ Page, Norbert "Chronic Toxicity and Carcinogenicity
Guidelines" Journal of Environmental Pathology and Toxicology
Vol. 1, pp. 161-182, 1977-
-------
Ill - 26
an initial examination of the external surfaces
and all orifices followed by an internal examination
of tissues and organs in situ;
inflation of urinary bladder and lungs;
trimming specifications;
multiple sections (step cuts) must be made on each
tissue or organ that contains gross evidence of a
neoplasm or lesion and on each tissue or organ
in which a metastasis may be anticipated.
Assumptions; The following assumptions were used
in calculating the cost of necropsy and histological
preparation:
13 hours/animal of technical time are required
for gross necropsy and examination, preparation
of tissues, and staining of slides;
The salary of the necropsy technical employee
is $15,000 per year.
Cost; Necropsy and histological preparation is
estimated to cost $150,000, calculated as follows;
Mice;
400 animals x 13 hours/animal = 5,200 hours
5,200 hours
2,080 hours =2.5 person-years
2.5 person-years x ($15,000/year x 100%
overhead) = $ 75,000
Rats;
The above calculation applies to rats as
well — i.e., the cost is $ 75,000
Total $150,000
3. Prechronic Testing Costs
Rule Requirement: A preliminary toxicology study
of at least 90 days must be utilized to predict dose
levels. If such a study has been completed previous-
ly, it may be submitted for this purpose.
Assumptions; The following assumptions are used in
calculating the cost of prechronic testing;
The prechronic testing procedures of the National
Cancer Institute (NCI) will be used. These procedures
are essentially those described in "Guidelines for
-------
Ill - 27
Carcinogen Bioassay in Small Rodents", National
Cancer Institute Carcinogens Technical Report Series,
No. 1, February 1976, pp. 11-15. Briefly, they
consist of the following studies:
~ acute toxicity test; 5 dose levels plus controls,
5 animals/ sex/dose level, 2 species, 14-day
observation period following single dosing,
and a 16-day quarantine period.
~ repeated-dose test; 5 dose levels plus controls,
5 animals/ sex/dose level, 2 species, and a
14-day exposure period. No additional observation
following exposure period, and a 16-day quarantine
period.
- subchronic test; 5 dose levels plus controls,
10 animals/sex/dose level, 2 species, 90-day
exposure period, and a 16-day quarantine
period. No additional observation following
exposure period. Pathological examination
of 32 tissues in control group and high dose
group. Target tissues thus identified are
then examined in next-to-highest dose group
and proceeding through lower dose groups until
no effects are found in the particular tissue.
The cost of these prechronic procedures will be the
same as currently estimated by NCI. NCI estimates
that their feeding studies cost approximately
$268,000 and that prechronic test costs constitute
approximately 19% of this total. I/
NOTE; The fact that prechronic studies are assumed for costing
purposes does not indicate that EPA is thereby providing
guidance as to appropriate prechronic testing procedures
Cost: Prechronic studies are estimated to cost $50,900,
calculated as follows:
19% of total cost of $268,000 = $50,920, or
approximately
$50,900.
I/ Personal communication with Donald Minnick, Tracer Jitco,
April 20, 1979. Tracer Jitco is the prime contractor for
NCI's Carcinogen Bioassay Program.
-------
Ill - 28
E. CHRONIC EFFECTS
1. Summary
The estimated costs of the chronic effects standards are
shown below in Table 3.
Table 3
Cost of Chronic Effects Test Standards
Rat Dog Total
Fixed Costs; - - $45,000 I/
Variable Costs;
Animal Procurement $ 1,600 $ 8,700 $ 10,300
Rations 1,500 - 2/ 1,500
Animal Care 95,600 175,200 2/ 270,800
Clinical Laboratory Tests 9,600 3/ 11,500 21,100
Microscopic Examination 62,000 13,000 75,000
Necropsy and Histological 87,000 18,000 105,000
Preparation
Total Variable Costs $257,300 $226,400 $483,700
Total Chronic Costs - - $528,700
Prechronic Costs; $100,500
Total Costs Including Prechronic $629,200
The estimates assume testing in 464 rats and 48 dogs.
This assumes testing at 3 different dose levels plus a control
group, and testing in both sexes — (3 dose levels + 1 control
group) x 2 sexes =4x2=8 groups in each species. Group
size in the rat is 58 — this consists of 50 per group as in
the oncogenic study plus an additional 8 animals required
for clinical laboratory tests. Group size for the dog is 6 —
the clinical laboratory tests are performed on all dogs and
no additional animals are required for this purpose.
Rat; 8 groups x 58 rats/group = 464 rats
Dog; 8 groups x 6 dogs/group = 48 dogs.
The details of how the variable costs of chronic testing
and associated prechronic testing were calculated are provided
_ _ __
I/ See Section C, "Fixed Costs" for detailed discussion.
2/ Cost of rations for dog is included in animal care.
3/ This represents the costs of the tests themselves.
The addition of 8 rats per group costs an additional $34,200
which appears in this table under animal procurement, rations,
etc. The true cost of these tests in the rat is therefore
approximately $43,800.
-------
Ill - 29
2. Variable Costs
Variable costs of chronic testing include the costs
of animal procurement, rations, animal care, clinical laboratory
tests, microscopic examinations, and necropsy and histological
preparation. Each of these costs were calculated separately
for the rat; for the dog, the costs of rations and animal
care were developed as a single estimate. These calculations
are discussed below.
a. Animal Procurement
Rule Requirement; EPA does not stipulate that
a specific strain or stock must be used, nor does
EPA express a preference among inbred, outbred, and
hybred strains. However, test animals must be from
established strains and/or stocks. As part of the
study plan submission, the sponsor must provide historical
data on the lifespan and types and incidence of disease
for the strains.
Assumptions: The following assumptions were used
in calculating the cost of animal procurement:
. Rats cost $2.85 each;
. Dogs cost $150 each; I/
. 560 rats and 58 dogs will be purchased. 2/
Cost: Animal procurement is estimated to cost
$10,300, calculated as follows:
560 rats x $2.85/rat = $ 1,600
58 dogs x $150/dog = $ 8,700
$10,300
b. Rations (rat only)
Rule Requirement; A standardized rodent diet
containing specified nutrient levels and produced
from certain feed stocks or ingredients (see Appendix
A, Subpart D, Section 772.113-1) is required unless
I/ Two estimates were obtained here: $300/beagle from a non-
profit testing laboratory and $150/beagle from a commercial
testing laboratory. The latter estimate was for a 7-8
month old dog which would be more expensive than the dogs
used for the test standards, where dosing begins by 10 weeks
and purchase (to allow for acclimatization period) would
occur even earlier. EPA believes $150/dog is a conservative
and reasonable figure.
2/ It is assumed that 20% more rats and dogs will be purchased
than started on the test so that testing can begin on
healthy animals. (See Section B, page III - 8).
-------
Ill - 30
information is provided by the sponsor to justify de-
viation. Diets for species other than the rodent must
be approved by EPA.
Assumptions; The following assumptions were used
in calculating the cost of rations:
Rats consume 13.5 gm of food per day;
Rations cost $235/ton; I/
Duration of the test is 2 1/2 years or 913 days
(see Section B, page III - 7).
Cost; Rations are estimated to cost $1,500, cal-
culated as follows:
464 rats x 13.5 gm/day x 913 days = 5,719 kg.
5,719 kg. x .0011 tons/kg, x $235/ton =
$1,478 or
approximately $1,500
c. Animal Care (rat only)
Animal care consists of clinical examination,
observation, feeding, and cage cleaning.
i. Clinical Examination
Rule Requirements; Clinical examination, including
the weighing of each animal, must be conducted at
least once a week during the first 13 weeks, and
every two weeks thereafter. Clinical examination
must include observation relating to food and water
consumption, morbidity, mortality and causes thereof,
pharmacologic effects, and behavioral changes.
Assumptions; The following assumptions were used
in calculating the cost of clinical examinations:
A technical employee can weigh and examine 25
animals per hour;
The salary of the technical employee is $12,000
per year.
Duration of test is 2 1/2 years or 130 weeks.
(See Section B, page III - 7).
T/Low bid to supply NIH-31 rodent diet containing 18% crude
protein to NIH for 12 months was $235.20 per ton. (Letter
from Joseph Knapka, NIH, to Carl Morris, EPA dated January 11,
1979).
-------
Ill - 31
Cost; Clinical examination is estimated to cost
$15,400, calculated as follows:
464 animals * 25 animals/hour =18.6 hours/week
Weekly exams for 13 weeks:
13 weeks x 18.6 hours/week 242 hours
Biweekly exams for 117 weeks:
(117 weeks * 2) x 18.6 hours/week 1088 hours
1330 hours
1330 hours
2080 hours/year = .64 person-years
.64 person-years x ($12,000/yr. x 100%
overhead) = $15,360 or
approximately
$15,400
ii. Observation
Rule Requirement; A technical employee must observe
the test animals every 12 hours throughout the test period.
Assumptions; The following assumptions were made in
calculating the cost of observation:
The time requirements for the two daily observations
are assumed to be as follows:
- For the first 18 months of the study, 3.5 hours
per day for 464 rats;
- For the last 12 months of the study where the
health status of more animals will be problema-
tical, 5.8 hours per day for 464 rats.
The salary of the technical employee is $12,000/yr.
Duration of the test is 2 1/2 years.
Cost; The estimated cost of observation is $46,600,
calculated as follows:
First 18 months:
(365 days x 1.5 years) x 3.5 hours/day = 1,916 hours
Last 12 months:
(365 days x 1.0 years) x 5.8 hours/day = 2,117 hours
4,033 hours
4,033 hours
2,080 hours/year =1.94 person-years
-------
Ill - 32
1.94 person-years x ($12,000 x
100% overhead) = $46,560 or
approximately
$46,600
iii. Feeding
Rule Requirement; For a feeding study, the test substance
must be administered ad libitum (continuously available).
Assumptions; The following assumptions were made in
calculating the cost of feeding;
1 cage will contain 2 rats;
A technician can process 30 cages/hour for
feeding the test animals;
The technician's salary is $8,000/year;
The feed containers for rats will be refilled
once every third day;
Duration of test is 2.5 years or 913 days.
(See Section B, page III - 7).
Cost; The estimated cost of feeding is $18,100, calculated
as follows:
464 rats f 2 rats/cage = 232 cages
232 cages f 30 cages/hour = 7.73 hours/feeding
7-73 hours/feeding x (913 days x 1/3) = 2,352 hours
2,352 hours
2,080 hours/year = 1.13 person-years
1.13 person-years x ($8,000/year x 100%
overhead) = $18,080 or
approximately
$18,100
iv. Cage Cleaning
Rule Requirement; The recommendations of HEW (NIH)
Publication No. 74-23, "Guide for the Care and Use of Laboratory
Animals" apply except where standards are specified in the
Animal Welfare Act of 1970 (9CFR Part 3). HEW Publication
No. 74-23 states that for routine maintenance of small rodents,
one to three changes per week of litter or bedding should
suffice.
Assumptions; The following assumptions were made
in calculating the cost of cage cleaning;
-------
Ill - 33
1 cage will contain 2 rats;
Cage cleaning will occur twice a week
(average of the one to three changes
per week cited above);
A technician can clean 30 cages per hour;
. The technician's salary is $8,000/year;
. Duration of test is 2.5 years or 130 weeks
(see Section B, page III - 7).
Cost; The estimated cost of cage cleaning is $15,500,
calculated as follows:
464 rats* 2 rats/cage = 232 cages
232 cages * 30 cages/hour = 7.73 hours/cage cleaning
7.73 hours/cage cleaning x (130 weeks x 2) = 2,010 hours
2,010 hours
2,080 hours/year = .97 person-years
.97 person-years x ($8,000/yr. x 100%
overhead) = $15,520 or
approximately
$15,500
v- Summary Chart
Summarizing the above costs, the estimated costs of animal
care for rats afe as follows:
Rat
Clinical Examination 15,400
Observation 46,600
Feeding 18,100
Cage Cleaning 15,500
$95,600
d. Rations and Animal Care (dog)
Rule Requirements; The requirements for dogs are as
specified in the rations and animal care sections above.
Assumptions; The following assumptions were made
in calculating the cost of rations and animal care
for the dog.
Duration of the test is 2 years or 730 days
(see Section B, page III - 7).
Dogs will receive very good animal care because
each dog represents a sizable investment of
resources.
-------
Ill - 34
Rations and animal care for dogs will cost
$5/dog/day. I/
Cost; Rations and animal care for the dog is estimated
to cost $175,200, calculated as follows:
$5/dog/day x 48 dogs x 730 days = $175,200
e. Clinical Laboratory Tests
Rule Requirements; A number of determinations are required
on a minimum of eight rodents in each group and on all non-rodents
These determinations are specified in detail in Subpart D,
Section 772.113-3(b)(1)(ii). Briefly summarizing, hematology,
blood chemistry, urinalysis, function tests, and residue analysis
are required at least at 3,6,12,18, and 24 months and at study
termination. Function tests are also required at the beginning
for non-rodents.
Assumptions; The following assumptions were made in
calculating the cost of clinical laboratory tests:
These determinations will be performed each
time in 8 rats per group and 6 non-rodents
per group. 8 rats per group represents the
minimum requirement of the standard whereas
6 non-rodents per group results from using
6 non-rodents per group for chronic testing,
which is the minimum requirement for the*entire
chronic test.
The cost of these determinations is $25/rat
and $40/dog each time the set of determinations
is performed.
Hematology, blood chemistry, urinalysis, and
residue analysis will be performed 5 times,
function tests will be performed 6 times in
the non-rodents. All analyses will be performed
6 times for the rodent. The non-rodent require-
ments differ from the rodent requirements
because the minimum duration periods differ
between the two species. In the non-rodent,
all tests except the function test will only
be performed 5 times in the 2 year test period.
The function tests in the non-rodent are required
at the beginning of the study and thus must be
performed a minimum of six times.
I/ EPA assumption based on limited industry data.
-------
Ill - 35
Costs estimates were derived based on all
tests being performed 6 times since the error
will be small relative to other estimation
errors.
Cost; Clinical laboratory tests are estimated to cost
$21,100, calculated as follows:
Rat;
6 times x 8 rodents/group x 8 groups x
$25/rat/time = $ 9,600
Dog;
6 times x 6 dogs/group x 8 groups x
$40/time = $11,520
Total $21,120
or approximately
$21,100.
f. Microscopic Examinations
Rule Requirements; Microscopic examination must be per-
formed on all tissues described in Subsections (b)(2)(ii),
(b)(2)(v), and (b)(2)(vi) of Section 772.113-3. Subsection
(b)(2)(v) requires that when there is clinical evidence
of specific toxicologic or pharmacologic effects related
to specific target organs, -the necropsy and microscopic
examinations of the suspected target organs must be conducted
in greater detail. Subsection (b)(2)(vi) requires that
sections from certain tissues be microscopically examined
from a minimum of ten animals from each test group and
from all animals with clinical or grossly observable evidence
of disease. If microscopic examination shows evidence
of disease in any of these tissues, then these target
tissues must be examined microscopically for all animals.
Assumptions; The following assumptions were used
in calculating the cost of microscopic examinations:
The routine microscopic examination of Subsection
(b)(2)(ii) represents approximately 60 sections
per animal.
The additional examinations of Subsection (b)(2)(v)
represent approximately a 10% increase over the
routine requirements, or 6 sections per animal.
The special examinations of Subsection (b)(2)(vi)
represent only 1-2 sections. These 1-2 sections
will be examined in anywhere from 20-100% of the
animals. This requirement should represent, on
average, no more than 1 section per test animal.
-------
Ill - 36
Total sections required per animal are therefore
as follows:
routine: 60
additional: 6
special: _1
67 sections.
A pathologist can read slides from 50,000 sections
per year for the rat. This figure is from an NCI
survey which assumed four hours a day at the micro-
scope. This figure has been used elsewhere
to calculate pathology workload. I/
A pathologist can read slides from 25,000 sections
per year for the dog. 2/
The salary of the pathologist is $50,000 per year.
Cost; Microscopic examination is estimated to cost $75,000,
calculated as follows:
Rats;
464 rats x 67 sections/animal = 31,088 sections
31,088 sections
50,000 sections/year = .62 person-years
.62 person-years x ($50,000/year x 100%
overhead) = $62,000
Dogs;
48 dogs x 67 sections/animal = 3,216 sections
3,216 sections = .13 person-years
25,000 sections/year
.13 person-year x ($50,000/year x 100%
overhead) * $13,000
Total $75,000
V Page, Norbert "Chronic Toxicity and Carcinogenicity Guidelines"
Journal of Environmental Pathology and Toxicology, Vol. 1, pp.
161-182, 1977.
2/ EPA assumption, based on an extrapolation from the 50,000
figure for the rat.
-------
Ill - 37
g. Necropsy and Histological Preparation
Rule Requirements; The rule requirements are stated
in detail in the standards (Section 772.113-3(b)(2)).
The requirements provide for:
an initial examination of the external surfaces
and all orifices followed by an internal examination
of tissues and organs in situ;
inflation of urinary bladder and lungs;
trimming specifications; and
multiple sections (step cuts) on each tissue
or organ that contains gross evidence of a
neoplasm or lesion and on each tissue or organ
in which a metastasis may be anticipated.
Assumptions; The following assumptions were used
in calculating the cost of necropsy and histological
preparation.
13 hours/rat and 26 hours/dog I/ of technical
employee time are required for gross necropsy
and examination, preparation of tissues, and
staining of slides.
The salary of the necropsy technical employee
is $15,000 per year.
Cost; Necropsy and histological preparation
is estimated to cost $105,000, calculated as follows;
Rats;
464 animals x 13 hours/animal = 6,032 hours
6,032 hours
2,080 hours/year = 2.9 person-years
2.9 person-years x ($15,000/year x 100%
overhead) = $87,000
I/ 26 hours/dog is an extrapolation from the 13 hours/rat
~ figure.
-------
Ill - 38
48 animals x 26 hours/animal = 1,248 hours
1,248 hours
2,080 hours/year = .6 person-years
.6 person-years x ($15,000/year x 100%
overhead) = $18,000
Total 105,066
3. Prechronic Costs
Rule Requirement; A preliminary toxicology study of
at least 90 days must be utilized to predict dose levels.
If such a study has been completed previously, it may be
submitted for this purpose.
Assumptions; The following assumptions are used
in calculating the cost of prechronic testing.
The following testing will be assumed;
- rats; National Cancer Institute (NCI)
prechronic testing procedures will be
used. These procedures are essentially
those described in "Guidelines for Carcinogen
Bioassay in Small Rodents", National Cancer
Institute Carcinogens Technical Report Series,
No. 1, February 1976, pp. 11-15. Briefly, they
consist of the following studies;
- acute toxicity; 5 dose levels plus controls,
5 animals/sex/dose level, 2 species, 14-day
observation period following single dosing,
and a 16-day quarantine period.
repeated dose; 5 dose levels plus controls, 5
animals/sex/dose level, 2 species, and a 14-day
exposure period. No additional observation
following the exposure period and a 16-day
quarantine period.
subchronic; 5 dose levels plus controls,
10 animals/ sex/dose level, 2 species, 90-day
exposure period, and a 16-day quarantine
period. No additional observation following
the exposure period. Pathological examin-
ation of 32 tissues in the control group and
high dose group. Target tissues thus identified
are examined in next-to-highest dose group
and proceeding through lower dose groups
until no effects are found in the particular
tissue.
-------
Ill - 39
- Dogs; The same number of dose levels and
dogs per group are used as in the long-term
study; therefore, 48 dogs will be used.
- subchronic; 3 dose levels plus controls,
6 dogs/sex/dose level, a 90-day exposure
period, and no additional observation
following the exposure period. Hematology,
blood chemistry, and urinalysis on all
dogs, three times during the study..
Pathological examination of same tissues
as in the chronic study.
Cost of the NCI prechronic study for the rat is
only one half of the cost of the study for
both rats and mice.
Cost of the prechronic dog study can be calculated
using the same assumptions as in the long-term
study. The fixed costs for a prechronic dog study
are assumed to be $10,000.
Cost; Prechronic testing is estimated to cost
$100,500, calculated as follows:
Rat; $50,900 I/ x 1/2 = $25,450, or
"" approximately
$25,500.
Dog;
Animal Procurement;
48 dogs x $150/dog 2/ = $ 7,200
Rations and Animal Care;
48 dogs x $5/dog/day 3/ x 90
days = $21,600
I/ Cost of NCI prechronic testing for rats and mice is 19% of
cost of long term NCI test ($268,000), or $50,900 (see
Section D-3, page III - 27).
2/ See Section E - 2, pg. Ill - 29.
3/ See Section E - 2, pg. Ill - 34.
-------
Ill - 40
Clinical Chemistry;
48 dogs x 3 times x $40/set I/ of
observations =
$ 5,760
Microscopic Examination;
48 dogs x 60 sections/dog 2/ = 2,880
sections
2,880
25,000 sections/yr.
.12 person-years
.12 person-years x ($50,000/yr. x
100% overhead) =
$12,000
Necropsy and Histological Preparation;
48 dogs x 26 hours/dog 3/ = 1,248 hours
1,248 hours
2,080 hours/year = .6 person-years
.6 person-years x ($15,000/yr. x
100% overhead) =
$18,000
Fixed Costs;
Total Cost
$74,560
$10,000
$74,560 or
approximately
$75,000
Total Prechronic Costs for Rat and Dog
$100,500
I/ See Section E-2, page 111-34.
^/ See Section E-2, page 111-36.
3/ See Section E-2, page 111-37.
-------
Ill - 41
P. COMBINED CHRONIC EFFECTS TEST
1. Summary
The estimated cost of the combined chronic effects
standards are shown below in Table 4.
Table 4
Cost of Combined Chronic Effects Test Standards
Mouse Rat Dog Total
Fixed Costs; $ - - $ 45,000 I/
Variable Costs;
Animal Procurement $ 1,100 $ 2,200 $ 8,700 $ 12,000
Rations 300 2,100 - 2,400
Animal Care 32,800 131,500 175,200 339,500
Clinical Laboratory
Tests - 13,200 11,500 24,700
Microscopic Examination 53,800 85,300 13,000 152,100
Necropsy and Histological
Preparation 75,000 119,600 18,000 212,600
Total Variable Cost $163,000 $353,900 $226,400 $743,300
Total Combined Chronic $788,300
Cost
Prechronic Costs; 125,900
Total Costs Including Prechronic $914,200
2. * ' Variable Costs
The figures for the dog in Table 4 are exactly the same as those
presented in Table 3, Section E, because the tests are identical. The
figures for the mouse are the same as in Table 2, Section D, except
for 2 iteiirs;
1) Hematology is not required in the combined study for
the mouse and thus no cost is shown.
2) Microscopic examination requirements differ slightly.
Since the pathology requirements are more extensive in
the chronic study than in the oncogenic, the costs are
greater in the combined study for the mouse (which uses
the chronic requirements) than in the oncogenic. This
difference is estimated to amount to: 67 sections
(chronic) - 56 (oncogenic) = 11 sections/animal.
I/ See Section C for detailed discussion.
-------
Ill - 42
11 sections/mouse x 400 mice = 4,400 sections
4,400 sections
50,000 sections/year = .088 person-years
.088 person-years x ($50,000/year x 100% overhead) = $ 8,800
Thus the cost for microscopic examination of the mouse in the
combined study is $8,800 greater than in the oncogenic study
where the cost is $45,000, or $53,800.
The cost estimates for the rat reflect the fact that 4 or 5
dose levels will be required in the combined chronic tests where
only 3 are required in the chronic test. Since EPA is unable to
predict what percentage of testing will be requiring at 4 vs. 5
dose levels, equal amounts of testing at both dose levels has been
assumed for the purposes of cost estimation. The figures in Table 4
for the rat are 1.375 times the corresponding figures in Table 3.
This figure has been calculated as shown below:
5 dose levels (4 dose levels 4- control) T 4 dose levels (3 dose
levels plus control) =1.25
6 dose levels (5 dose levels + control) T 4 dose levels (3 dose
levels plus control) =1.5
Average = (1.25 + 1.5)/2 = 1.375.
3. Prechronic Testing Costs
The prechronic testing done in the combined test is assumed to
be identical to that done in the oncogenic study for the mouse and
rat and that done in the chronic study for the dog. Costs are there-
fore estimated to be approximately $125,900 calculated as follows:
Oncogenic Study - mouse, rat: $50,900
Chronic Study - dog: 75,000
Total Prechronic Study Costs $125,900
4. Cost Savings Due To Combined Test
The cost savings due to combining the oncogenic and chronic test
standards in the combined test are approximately $125,000 - $150,000.
The cost saving from the combined test standard (without prechronic
testing) would be approximately $125,000, calculated as follows:
Oncogenic Costs Without Prechronic (Table 2) $387,700
Chronic Costs Without Prechronic (Table 3) 528,700
Total Cost for Both Tests $916,400
Combined Costs Without Prechronic (Table 4) -790,900
Cost Savings $125,500
-------
Ill - 43
The above calculation assumes that no prechronic testing need be
performed. If prechronic testing is required for all tests, then
there would be an additional cost savings in the combined test of ap-
proximately $25,000, calculated as follows:
Oncogenic Prechronic Testing Costs $ 50,900
Chronic Prechronic Testing Costs 100,500
Total for Both Tests $151,400
Combined Prechronic Testing Costs -125,900
Cost Savings $ 25,500
The cost savings would arise primarily from the fact that when
both the oncogenic and chronic tests are performed separately, test-
ing for the rat in 3 dose levels plus controls has to be performed
twice. When the combined test is performed, testing for the rat
would be at either 4 or 5 dose levels plus controls, and thus there
would be a saving of 1-2 dose levels plus a control group over that
required when separate tests are performed for oncogenic and chronic
effects. Additional savings would result from the fact that the
fixed costs for a combined study are lower than the sum of the fixed
costs for the oncogenic and chronic studies when conducted separate-
ly. Other less significant savings would occur in clinical labora-
tory tests (approximately $1300), however, these savings are offset
slightly by increased microscopic examination requirements ($8800).
-------
Ill - 44
G. OTHER DATA ON TESTING COSTS
The National Cancer Institute (NCI) contracts with testing
laboratories to perform oncogenicity testing. Cost data from these
contracts, when properly adjusted for differences in testing proced-
ures between NCI and EPA, provides a rough check on the cost estimates
developed in Sections D, E, and F.
The cost of testing a chemical according to the NCI protocol,
as indicated by current cost data from contracts, is as follows: I/
$268,000 for a feeding study;
$382,000 for gavage; and
$402,000 for an inhalation study.
The NCI cost of a feeding study including prechronic testing ($268,000)
can be compared to the EPA estimate ($439,000) by making the following
adjustments and then comparing the adjusted total ($382,900) with EPA's
estimate ($439,000). The adjustments to NCI's cost are shown below:
Modified NCI protocol adjusting
for differences in number of
NCI protocol dose levels, pathology, and
(2 dose levels) 2/ other costs
Fixed Costs $ 81,700 $ 81,700
Variable Costs
- Pathology 96,500 160,800 3_/
- Other 89,800 119,700 4/
Other Costs
Animal Pro-
curement - 2,500 5/
Statistical - 18,200 \/
Analysis
Total Cost $268,000 $382,900
I/ Personal communication wTth Donald Minnick, Tracer jTtco.
Tracer Jitco is the prime contractor for the NCI
Carcinogenesis Bioassay Program.
2/ NCI's estimated fixed costs are 30.5% of total cost (includes
prechronic, project management, and technical report)
and pathology is 36% of total cost (Personal communication with
Donald Minnick). Therefore, fixed costs are .305 x $268,000
= $81,740, or approximately $81,700; pathology costs are
.36 x $268,000 = $96,480, or approximately $96,500; and
other costs are simply the remainder of the $268,000.
3/ EPA estimates that the pathology requirements per animal
"~ are approximately 25% greater than those of NCI. EPA also
requires 3 dose levels plus a control group whereas NCI requires
2 dose levels. The EPA pathology cost would therefore be 1.25 x
4/3 x $96,500 = $160,833, or approximately $160,800.
-------
Ill - 45
No adjustment is made for differences in observation re-
quirements. While EPA requires that a technical employee ob-
serve the animals every 12 hours (twice daily), and NCI guide-
lines require observation only every 24 hours, actual practice
of NCI contractors is to observe the animals at the beginning
and end of the 9-hour duty-day period (twice daily). Thus,
the only cost difference would be that EPA requires either
longer shifts, an extra shift, or some overlap, all of which
might lead to some increase in cost.
The EPA estimate of $439,000 for a feeding study is thus
approximately 15% higher than the adjusted NCI cost of $382,900.
The fact that the EPA estimate is within 15% of the adjusted NCI
costs is an indication that the costing procedures and assump-
tions used here are probably reasonable, at least in the aggre-
gate.
_4/ Other variable costs are increased proportionately,
4/3 x $89,800 = $119,733, or approximately $119,700.
This is due to the difference in the required number
of dose levels.
5/ NCI supplies animals for testing so its costs do
not include the cost of animal procurement. EPA esti-
mate of $2,500 is used.
6/ NCI does not require statistical analysis. EPA's
~~ estimate of $18,200 is used.
-------
Ill - 46
H. COST OF ALTERNATIVE STANDARDS
Estimates of the cost of alternative standards are use-
ful in making decisions as to what standards EPA should re-
quire and also as an indication of the additional cost that
testers will incur for adding optional items, i.e., dose
levels, duration, tests, number of animals, etc.
Innumerable variations in the standards could be present-
ed here. Instead, the attempt here has been to focus on items
which may be controversial and which appear, based on the esti-
mates, to have a large impact on the cost of the standards. The
costs presented in Table 5 on the next page are the marginal cost
of varying that item assuming all other parts of the standard are
as proposed.
However, the reader is cautioned that the cost for various
parts of the protocol are interrelated and cannot always be summed.
For example, the cost of adding a dose level is clearly dependent on
the extent of the pathology required. Increasing both requirements
at the same time would have a greater impact on the cost than the sum
of the individual costs. Conversely, decreasing both requirements
would have less of a cost saving than the sum of the individual costs,
-------
TABLE 5
Proposed
Item Test
Dose Levels Oncogenic
Oncogenic
Chronic
Chronic
Combined
Combined
Combined
Duration Oncogenic
of Test
Oncogenic
Chronic
Chronic
Combined
Combined
Combined
Species
Mouse
Rat
Rat
Dog
Mouse
Dog
Rat
Mouse
Rat
Rat
Dog
Mouse
Rat
Dog
Requirement Variation
3 dose levels +1 dose level
plus control
3 dose levels +1 dose level
plus control
3 dose levels ±1 dose level
plus control
3 dose levels ±1 dose level
plus control
3 dose levels ±1 dose level
plus control
3 dose levels ±1 dose level
plus control
4 or 5 dose levels±l dose level
plus control
24-30 months ±6 months
24-30 months ±6 months
30 month ±6 months
minimum
24 month minimum ±6 months
24-30 months ±6 months
30 month +6 months
minimum ""
24 month minimum ±6 months
Effect on Cost
±$38,900 I/
±$46,800
±$64,300
±$56,600
±$40,800
±$56,600
±$64,300
M
M
1^
±$ 8,300 2/ ,
*-
±$16,100
±$19,400
±$43,800
±$ 8,300
±$26,700
±$43,800
I/ EPA assumes tHat varying tHe number of dose levels will affect all variable costs
~~ proportionately. The cost variation was obtained by dividing the variable costs
for each species specified in Tables 2, 3, and 4 by the number of dose levels plus
controls (4 for all tests except in the combined test in the rat where the
average factor of 4.5 + 1 = 5.5 was used).
2/ EPA assumes that varying the duration of exposure will affect only the cost of
rations and animal care. Any indirect effects on pathology due to different age
of animals, number of lesions, etc., were assumed to be insignificant. EPA
assumes that the number of times clinical laboratory tests are performed is not
changed.
-------
TABLE 5 (continued)
Microscopic
Examinations
Chronic
Species
Mouse
Rat
Proposed
Requirement
Approximately 56
sections
Complete microscopic
examination in all
animals (approximately
67 sections/animal)
Variation
NCI requirements
(approx. 20% less)
Effect on Cost
- $ 9,000
- $ 16,000 3/
Essentially FIPRA
proposed guidelines
(1) Complete micro-
scopic examination
of all animals in high
dose level and in
control group, and in 15
rats/group in inter-
mediate and low dose groups.
(2) For remainder of animals,
examination of a small num-
ber of tissues plus any
target tissues identified
from (1) above.
3_/ The cost savings of §16,000 was calculated as follows:
(1) 67 sections/animal x 232 animals =
(High dose plus control groups —
58 rats/group, 4 groups)
67 sections/animal x 60 animals =
(15 rats/group in intermediate
and low dose levels, 4 groups)
15,544 sections
4,020 sections
00
EPA assumes that the tissues required to be examined plus the target tissues
identified from (1) will lead to microscopic examination of approximately 20
sections/animal in the remaining animals, thus
(2) 20 sections/animal x 172 animals =
(Remainder of animals or
58 - 15 - 43 rats/group, 4 groups)
3,440 sections
23,004 sections
Base case = 67 sections x 464 animals = 31,088 sections
Difference = 8,084 fewer sections
.16 x ($50,000/yr. x 100% overhead)
8,084 sections
50,000 sections/yr. = .16 person year;
$16,000
-------
Table 5 (continued)
Item
Test
Species
Clinical
Laboratory
Chronic
Rat
Proposed
Requirement
8 additional rats
per group in addi-
tion to the 50.
Variation
Essentially
FIFRA proposed
guidelines:
No additional
animals (8 are
part of 50)
Effect on Cost
-$34r200 4/
V A variable cost savings of $34,200 may be achieved by using 50 rather than 58 rats since
fewer animals must undergo other parts of the protocol. This savings was calculated as
follows:
Variable Costs (Rat)
Less Clinical Laboratory Tests
257,300 (from Table 3)
9,600 (from Table 3)
247,700
8
x $247,700 = $34,165 cost savings due to use of 50 rats for all cost elements.
-------
IV. Confidentiality Issues
A. Substantiation Policy
As indicated in the Section of the Preamble
entitled "Confidentiality and Public Access to Infor-
mation," EPA is proposing to require persons asserting
confidentiality claims for information contained in
health and safety studies to provide substantiation for
those claims in the form of written answers to specific
questions.
EPA's substantiation proposal for health and safety
data submitted under Section 4 is based on both policy
and legal considerations. First, for each of these
types of information there is strong evidence that
Congress intended public disclosure of the data to the
fullest extent possible. Second, because EPA expects a
high volume of requests for disclosure, EPA must have
the substantiation readily available in order to reduce
the administrative burden of responding to Freedom of
Information Act (FOIA) requests. Third, EPA must be
prepared to respond to these requests speedily. By
including in TSCA such provisions as the Section 4(d)
(3) requirement to publish a Federal Register notice of
receipt of data, Congress evinced a decision that the
public should be involved in the Section 4 data review
process to the extent possible. EPA will strive to
assure that non-confidential information be available to
the public as soon as possible.
Finally, the substantiation burden is further
justified by the need to discourage ill-founded
claims. Substantiation aids the submitter in under-
standing the findings which must be made to support a
confidentiality claim. Experience with the inventory
reporting regulations indicates that detailed substan-
tiation requirements significantly reduced the number of
claims. This should result in a higher level of public
information, and a reduced burden on EPA to evaluate and
protect information erroneously claimed to be confiden-
tial.
EPA considered requiring substantiation of a claim
of confidentiality for information contained in health
and safety studies to be submitted only after receipt of
an FOIA request for the information.
-------
IV-2
However, as indicated above, EPA believes that this
policy would not accomplish a number of purposes EPA
believes necessary to an effective implementation of
TSCA:
- speedy response to FOIA requests
- lessened administrative burden on the
Agency
- discouragement of overly broad confiden-
tiality claims
EPA's proposed Section 4 policy concerning the
substantiation of confidentiality claims for information
contained in health and safety studies is similar to
those proposed under the Section 5 Premanufacture
notification rulemaking (44 FR 2242, January 10, 1979).
B. Substantiation Questions
EPA's proposed procedures for Section 4 would
require persons claiming confidentiality or health and
safety data to address the following questions:
1. Will disclosure of the information
claimed as confidential in your health and safety study
disclosure a process used in manufacturing or processing
a chemical substance or mixture? If so, describe how
the information will disclose the process.
2. Will disclosure of the information
claimed as confidential in your health and safety study
disclose proportions of a mixture comprising any of the
chemical substances in the mixture? If so, describe how
the information will disclose the proportions.
3. Will disclosure of the information
claimed as confidential in your health and safety study
disclose information which is not related in any way to
the effects of a substance on health or the environment?
If so, describe how the information will disclose this
information.
4. Has any of the information claimed as
confidential been disclosed on a patent? If so, why
should it be treated as confidential?
5. Is this information available to the
public or your competitors without restriction?
6. How do you protect this information
from undesired disclosure?
-------
IV-3
7. Has this information been disclosed to
others?, If so, , what precautions have you taken in
regard to these disclosures? Has the information been
disclosed to the public or competitors?
8. In the case of information concerning
proportions of a mixture comprising one or more chemical
substances, does the mixture leave the control of your
company and move in commerce? If so, can the mixture be
analyzed to determine the proportion of the chemical
substance in it?
9. Has EPA, another Federal Agency, or any
Federal court made any pertinent confidentiality deter-
minations regarding this information? If so, please
attach copies of such determinations.
10. How long should confidential treatment
be given? Until a specific date, the occurrence of a
specific event, or permanently? Why?
11. What harmful effects to your competi-
tive position, if any, do you think would result from
disclosure of the process? The proportions? Infor-
mation not related in any way to the effects of a sub-
stance on the health or the environment? How would a
competitor use this information? Would the harmful
effects be substantial? What is the causal relationship
between disclosure and the harmful effects?
-------
V. Differences Between TSCA Section 4(b) Test
Standards and FIFRA Guidelines
A. Introduction
EPA's Office of Pesticide Programs has pro-
posed testing guidelines (43 FR 37336, Aug. 22, 1978)
under the authority of Sections 3(c)(2) and 25(a) of the
Federal Insecticide, Fungicide, and Rodenticide Act as
amended (FIFRA) (86 stat. 973; 89 stat. 751; 7 U.S.C.
136 et seg.), which contain data requirements comparable
to those presently being proposed under TSCA Section
4(b). The Agency's policy is to reduce the burden on
the regulated public which might arise from conflicting
requirements under these different sets of regulations
Therefore, although the proposed TSCA test standards and
FIFRA guidelines differ in some significant ways, the
final TSCA test standards and FIFRA guidelines will be
consistent to the extent permitted by the different
laws. The major differences between the two proposed
regulations are identified below to help facilitate the
comparison of the two proposals by the public.
B. Section 772.113-1 General
1. Subsection(d): The FIFRA guidelines
(Section 163.80-3(b) (5-7,12) do not contain a separate
section on good laboratory practice standards as do the
TSCA test standards. Good laboratory practice standards
appear throughout the FIFRA guidelines and are less
detailed in some respects.
2. Subsection(e): The FIFRA guidelines
(Section 163.80-3(b)(1)) do not contain as detailed
personnel requirements as the TSCA test standards do.
For example, the TSCA test standards have proposed two
types of qualified pathologists with specified training
and experience and do not presently allow for substitu-
tion of persons with equivalent training and experience
as the FIFRA guidelines do.
3. Subsection(f): The TSCA test standards
require the submission of a study plan at least 90 days
before a study is initiated. The FIFRA guidelines do
not require submission of a study plan prior to study
initiation but do require submission of a test protocol
with the final report (Section 163.83-1). A study plan
submission requires other information in addition to the
test protocol.
4. Subsection(g): The TSCA test standards
specify that the test substance or mixture administered
must contain no less than 90 percent of the specified
test sub&tance concentration during the time of
-------
V-2
administration; FIFRA guidelines specify only that no
mixture of test or control substance be maintained or
used during the period exceeding the known stability of
the test or control substance. Also, only the TSCA
standards specified that the initial mean concentration
of the test substance must not vary more than 5% from
the concentration designated in the test protocol.
5. Subsection(h-i): Only the TSCA test
standards require that all feed be used within 90 days
of its manufacture, that all rodents be fed a specified
standardized diet, and that feed and vehicles be
analyzed for specified contaminants.
6. Subsection(j): The TSCA test standards
require the' submission of "Interim Quarterly Summary Re-
ports." FIFRA guidelines do not have such a require-
ment.
-------
V-3
C. Section 772.113-2 Oncogenic Effects
Test Standards
1. Subsection(a)(3): The TSCA test
standards require that animals be weaned and environ-
mentally acclimatized before dosing; the FIPRA
guidelines (Section 163.83-2(c)(3)), under certain
specified circumstances, permit dosing in utero.
2. Subsection(a)(5): The FIFRA
guidelines (Section 163.83-2(c)(4)) require both an
untreated and a vehicle control group, if the toxic
properties of the vehicle are unknown. The TSCA test
standards leave the decision to the discretion of the
tester. The TSCA test standards indicate positive
control groups may be required for particular chemicals.
3. Subsection(a)(7): The TSCA test
standards specify the frequency of exposure for the
different routes of exposure; the FIFRA guidelines do
not.
4. Subsection(a)(8): The TSCA test
standards require the tester to administer the test sub-
stance to both rats and mice for a minimum of 24 months
but no longer than 30 months. The FIFRA guidelines
(Section 163.83-2(c)(6)) require the test substance to
be administered to mice for a minimum of 18 months and
not ordinarily longer than 24 months; to rats 24 months
and 30 months, respectively.
5. Subsection(a)(9): Both the TSCA
test standards and FIFRA guidelines (Section 163.83-
2(c)(7)) require at least three dose levels in addition
to controls. However, they define the highest dose
level and lowest dose level slightly differently. The
TSCA test standards require a preliminary toxicology
study of at least 90 days to select the dose levels and
require the sponsor to submit the rationale for dose
selection as part of the study plan submission. FIFRA
guidelines state that dose levels are generally
predicted from subchronic data.
6. Subsection(b)(1): The TSCA test
standards require that qualified veterinarian(s) be
responsible for the health status and care of all test
animals. The FIFRA guidelines do not specify the
qualifications of the veterinarian.
7. Subsection(b)(1): The TSCA
standards provide that the animals must be observed
every 12 hours and that losses greater than 5% in any
group due to cannibalism, autolysis of tissues,
misplacement of animals, and similar management problems
-------
may not be acceptable. The FIFRA guidelines (Section
163.83-2(c)(10)) require each test animal be observed at
least daily, with losses not to exceed 10 percent.
8. Subsection(b)(2): The TSCA test
standards specify the responsibilities of the two types
of qualified pathologists.
9. Subsection(b)(2): The tissues to
be examined microscopically are essentially the same for
both regulations. However, only the TSCA test standards
require oral mucous membranes and aorta be examined and
only the FIFRA guidelines require routine examination of
the sciatic nerve.
D. Section 772.113-3 Non-oncogenic
Chronic Effects Test Standards
1. Subsection (a)(l): The TSCA test
standards require in addition to a rodent, generally the
rat, the use of a non-rodent, generally the dog.
2. Subsections 772.113-2(a)(3,5,7):
The differences in these subsections have been discussed
above in part C, paragraphs 1-3.
3. Subsection 772.113-3(a)(8): The
TSCA test standards require the tester to administer the
test substance to the rat for at least 30 months. The
FIFRA guidelines (Section 163.83-1(c)(6)) require
administration for at least 24 months but no longer than
30 months. In addition, the TSCA test standards require
the test substance be administered to dogs for at least
2 years.
4. Subsection 772-113-3(a)(9): FIFRA
guidelines (Section 163.83-l(c)(7)) require that the
highest dose be higher than that expected for human
exposure. Under the TSCA test standards the tester must
conduct a preliminary toxicology study of at least 90
days to select the dose levels and must submit the
rationale for dose selection as a part of the study plan
submission.
5. Subsection 772-113-3(b)(1)(i):
The differences in this subsection have been discussed
above in part C, paragraphs 6-8.
6. Subsection(b)(l)(i): The TSCA
test standards require that urinalysis and certain
specified function test be performed while the FIFRA
guidelines leave these decisions to the discretion of
the tester.
-------
V-5
7. Subsection 772-113-3(b)(2)(ii):
The TSCA test standards require that all of the tissues
listed in this subsection be examined from all test ani-
mals. The FIFRA guidelines (Section 163.83-l(c)(16))
require a limited number of tissues from all test
animals be examined and other specified tissues from
test animals in certain test groups be examined.
E. Section 772.113-4 Combined Chronic
Effects Test Standards
FIFRA guidelines (Section 163.80-5)
allow combined testing to be conducted if the data
requirements of each individual test are satisfied. They
do not, however, provide any specified test methods.
TSCA Section 4 test standards (Section 772.113-4)
propose a test method for studying both oncogenic and
chronic toxicity effects simultaneously.
------- |