&EPA
United States
Environmental Protection
Agency
Office of the Administrator
Science Advisory Board
Washington DC 20460
SAB-EC-88-040D
September 1088
Final Report
Appendix D:
Strategies for
Health Effects Research
Report of the Subcommittee
on Health Effects
Research Strategies Committee
-------
NOTICE
This report has b«?en written as a part of the activities
of tfieScience Advisory Board, a public advisory group providing
extramural scientific information and advice to the Administrator
and other officials of the Environmental Protection Agency.
The Board is structured to provide a balanced, expert assessment
of scientific matters related to problems facing the Agency.
This report has not been reviewed for approval by the Agency;
hence, the contents of this report do not necessarily
represent the views and policies of the Environmental Protection
Agency or of other Federal agencies. Any mention of trade
names or commercial products do not constitute endorsement or
recommendation for use.
-------
U.S. Environmental Protection Agency
Science Advisory Board
Research Strategies Committee
Health Effects Group
Chair
Dr. David Rail
Director
National Institute of Environmental Health Sciences
111 Alexander Drive, Bldg. 101
Research Triangle Park, NC 27709
Members
Dr. Eula Bingham
Department of Environmental Health
University of Cincinnati Medical College
Kettering Laboratory
3223 Eden Avenue
Cincinnati, Ohio 45267
Dr. Bernard Goldstein
Chairman, Department of Environmental and Community Medicine
UMDNJ-Robert Wood Johnson Medical
675 Hoes Lane
P1scataway, New Jersey 08854-5635
Dr. David Hoel
Director, Division of Biometry and Risk Assessment
National Institute of Environmental Health Sciences
Research Triangle Park, North Carolina 27709
Dr. Jerry Hook
Vice President, Preclinical R&D
Smith, Kline and French Laboratory
709 Swedland Road
King of Prussia, PA 19406
Dr. Philip Landrigan
Director, Division of Environmental and Occupational Medicine
Mt. Sinai School of Medicine
1 Gustave Levy Place
New York, New York 10029
Dr. Donald Mattison
Director, Division of Human Risk Assessment
National Center for Toxicological Research
Jefferson, Arkansas 72079
-------
Dr. Frederica Perera
School of Public Health
Division of Environmental Sciences
Columbia University
60 Haven Avenue
New York, New York 10032
Dr. Ellen Silbergeld
Chief, Toxics Program
Environmental Defense Fund
1616 P Street, N. W.
Room 150
Washington, D. C. 20036
Dr. Arthur Upton
Director, Institute of Environmental Medicine
New York University Medical Center
550 First Avenue
New York, New York 10016
Support Staff
Dr. C. Richard Cothern
Executive Secretary
Science Advisory Board
U.S. Environmental Protection Agency
Ms. Mary Winston
Staff Secretary
Science Advisory Board
U.S. Environmental Protection Agency
Ms. Renee Butler
Staff Secretary
Science Advisory Board
U.S. Environmental Protection Agency
-------
ABSTRACT
This document attempts co delineate the long-tern health effects
research needs (both basic and applied) considered most supportive of EPA
programs. Chapter 1 provides a historical perspective , describes the
nature and sources of environmental determinants of health and disease,
touches on the underlying mechanisms of toxicity with implications tor risk
assessment and disease prevention, and indicates some of the areas where
research support is clearly inadequate.
Chapter 2 draws a distinction between the basic and applied long-term
health effects research needs of EPA programs by providing specific
examples that illustrate the need for research addressing "generic" issues
as well as various research activities that have application to specific
problems and specific settings but which must be carried on over a period
of several years. An attempt has been made- to explain how EPA uses/depends
on basic research of the type conducted by other Federal Agencies,
particularly as it relates to the regulatory mission of the Agency.
In Chapter 3 the toxic metal lead is used as the paradigm to illustrate
the place of and necessity for long-sustained, basic research activity in
the development of a foundation for constructive action in important
problems in environmental health. Continued long-range and basic research
investigations on lead toxicity are at one and the same time perhaps among
the more justifiable and yet less supportable of such activities in the
entire field of environmental health sciences.
A number of leading-edge/long-term basic research activities with
potential application to environmental health problems are described in
Chapter 4. It attempts to highlight those activities which perhaps
have the greatest promise in this area, llany of these include various
aspects of the "new molecular biology" research field, such as the study of
oncogenes and proto-oncogenes, the development and use of biomarkers to
determine internal dose and exposure and for relating exposure to disease.
Other newer developments in neurotoxicology, immunotoxicology and
reproductive toxicology are described. An important area of basic research
includes methods development and validation, tiagnetic resonance imaging is
discussed as a very promising new technique that should find many us etui
applications in studies of the internal structures, states, and
compositions of various biological systems.
Finally, in Chapter 5 the problem of estimation of population risks is
addressed, particularly as it relates to the role of animal data in the
quantification of possible human health risks. Factors considered here
include choice of mathematical model or extrapolation procedures, primary
versus secondary or indirect modes of action and threshold mechanism,
problems in species extrapolation and determination of biologically
effective dose. Some specific problems in human epidemiologic studies and
population risks analysis are also described. Factors affecting the
balance of basic research on cancer and non-cancer endpoints within any
Federal organization are also discussed. Long-term, basic research into
both cancer and non-cancer endpoints is recognized as being essential if
the EPA is to formulate a broad regulatory policy in the most accurate
manner possible.
-------
Chapter 1
ENVIRONMENTAL FACTORS AND HU11AN HEALTH
Arthur Upton and Philip Landrigan
HISTORICAL PERSPECTIVE
The past century has seen the conquest of those diseases which
have caused the greatest morbidity and mortality, in previous
generations. In the developed countries of the world, the average
life expectancy has doubled, now surpassing the biblical ideal of
"three score and ten" years (Figures 1 and 2). This transformation,
which would have seemed miraculous to our great grandfathers, has
resulted from advances in our understanding of the relationship
between health and Che environment, broadly speaking. These advances,
and the resulting improvements in agriculture, nutrition, sanitation,
public health, and medicine, have all hut eliminated infectious and
parasitic diseases as major causes of death in the industrialized
world. Replacing such afflictions as major causes of death in the
industrialized world are abnormalities in early growth and
development, chronic degenerative diseases, and cancer (Figure 3).
These diseases, viewed until recently largely as hereditary or
inevitable accompaniments of aging, are now attributed increasingly to
environmental causes. Our challenge is to identify the causes and to
control them (4).
-------
-2-
NATURE AND SCJURCES OF QJVIRONMEN1AL DETERMINANTS
OF HEALTH AN DISEASE
The "environment", defined broadly, encompasses all external
factors that may act on the human mind and body. Many of the factors
are produced or altered by man himself. They include chemical and
physical agents in air, food, water, drugs, cosmetics, consumer
products, the home, and the workplace. The "environment" is thus
complex and constantly changing. Inevitably, moreover, it contains a
myriad of agents in varying combinations and from multiple sources.
Furthermore, because the effects of different agents interact in
various ways, the ultimate impacts of any given environmental agent
may depend on the effect of other agents and the conditions of
exposure (4).
Air
Acute episodes of atmospheric pollution, sucn as those listed in
Table 1, have been observed to cause transitory increases in morbidity
and mortality. The effects of chronic exposure, however, are at an early
stage of documentation but this varies, depending on the pollutants
in question and their concentrations in the atmosphere (4).
On chronic exposure at relatively high concentrations in the
workplace, a variety of pollutants are known to cause toxic effects.
Examples include various gases (e.g., carbon monoxide, vinyl chloride,
coke oven emissions, radon), metals (e.g., lead, mercury, arsenic,
nickel) and dusts (e.g., asbestos, silica, cotton fibers, coal) (5).
Also well documented are the effects of chronic exposure to
cigarette smoke. The incidence of lung cancer has risen
-------
-3-
precipitously, in parallel with the antecedent increase in cigarette
consumption (Figure 4). This increase has also occured in parallel with
a massive use and environmental dispersal of asdestos. In smokers,
furthermore, there is a systematic relationship between the amount of
smoke inhaled and mortality from lung cancer (Figure 5), other cancers,
heart disease, and respiratory diseases. Lesser effects have been
tentatively attributed to passive inhalation of cigarette smoke in
chronically exposed nonsmokers.
The ultimate effects of chronic low-level exposure to other widely
prevalent combustion products and their derivatives (such as sulfur
dioxide, ozone, nitrogen dioxide, benzo(a)pyrene, and various
suspended particulates) are less well understood.
Although the air pollution produced as a result of coal combustion
is a direct cause of respiratory fatalities, there is no exact measure
of their number; however, several estimates have been made of the
number of fatalities attributable to the combustion of coal in
generating electricity (where about 70% of coal combustion occurs).
Inhaber (8), for example, estimated that between 5 and 500 fatalities
result per 1000 Hwe of electric power produced each year from
pollution generated by coal fired plants. A more recent survey by
experts in this area puts the estimate between zero and 1000
fatalities per year per 1000 Mwe of electric power produced (9,10).
On the basis of a value of 7 x 105 llwe of electric power produced in
the U. S. by the consumption of coal, the estimates imply that up to
700,000 fatalities per year may result from combustion of coal in the
U. S. Within the uncertainties of this estimate, it agrees well with
-------
-4-
a recent inference by Vdlson that "50,01)0 among the 2 million persons
who die each year in Che United States may have their lives shortened
by air pollution" (11). One may question, therefore, the extent to
which current ambient air standards provide adequate protection
against the potential long-term health effects of coal combustion
products, which cannot be specified with certainty on the basis of
existing knowledge (12).
It is noteworthy in the above context that indoor pollution with
combustion products nay lead to health effects in the chronically
exposed, especially children. Of increasing concern is the extent to
which elevation of the radon concentrations within houses and
buildings, by weather-stripping and other heat-saving measures, may
enhance the risk of lung cancer in their occupants (13-15).
Other air-borne pollutants with potential health effects include
allergens of various kinds. Although susceptibility to such agents
differs widely among individuals, sizable populations are at risk (4;.
Ihe full significance of air-borne agents as causes of disease is far
from established and strongly merits continued study (<0.
Water
In the third world microbial contamination of drinking water still
constitutes a major cause of death. Although this type of pollution
no longer exists on a significant scale in developed countries, the
chemical composition of drinking water has been implicated tentatively
in the two leading causes of death in the U. S.: cancer and
-------
-5-
cardiovascular disease (4,11). It is also noteworthy that water
supplies have been found to be polluted in a growing number of areas
(Figure 6), owing to contamination by metals, toxic wastes,
pesticides, agricultural chemicals, and products of chlorination or
ozonization.
The health impacts of small quantities of chemicals in drinking
water cannot be assessed precisely on the basis of existing knowledge.
Research is needed to elucidate the relevant causal relationships and
to clarify the pathways through which compounds affecting human health
may enter the water supply (17).
Food
There is some truth to the adage, "you are what you eat". Overall
health is undoubtedly influenced by the total intake of calories in
the diet, the relative intakes of different types of foods (protein,
fat, carbohydrates), the nutritional value of the various foods that
are ingested, the presence in food of certain naturally occurring
constituents or contaminants, and the presence of man-made additives
or pollutants (18). In general, more is known about the nutritional
requirements for normal growth, maturation, and reproduction than
about the optimal diet for long life and vigor.
In the case of cancer, for example, there appear to be many ways
through which the diet may affect the probability of the disease
(Table 2); however, the relative contributions of any of these
-------
-6-
hypochetical mechanisms co che pathogenesis of a particular form of
human cancer remains to be established (18). In this connection it is
noteworthy that some dietary factors may exert protective effects
which are of equal or greater importance than the carcinogenic effects
of others. Hence, the net effects of the diet may reflect the balance
between the two types of influences.
Because of the importance attributed to the diet in the
pathogenesis of cancer, heart disease, and other leading causes of
death in the modern world, the role of dietary factors strongly merits
further study.
Occupation
As noted above, occupational exposure to diverse physical and
chemical agents at relatively high dose levels has been observed to
cause various diseases. Collectively, the health impacts of these
agents and of work-related stresses may approach those caused by
occupattonally-related accidents (5).
Occupational diseases are also significant in pointing to risk
factors that may affect other populations at lower levels of exposure.
In addition, occupational disease represents a category of health
effects that is relatively amenable to preventive strategies. To
lessen the health impacts of occupational risk ractors, research of
several types deserves further emphasis: 1) more systematic and
quantitative monitoring of physical and chemical agents in the
workplace; 2) more complete surveillance and recording of work-related
health effects; and 3) development of clinical and laboratory tests
for ascertaining prior exposure to disease causing agents, for
-------
-7-
identifying high-risk groups, and for dececcing work-related diseases
at early stages, when they are most readily arrested, or reversed (4).
Toxic Wastes
Love Canal and Times Beach, to mention only two of many recent
examples (Tables 3 and 4), testify to the need for more adequate
disposal of toxic wastes. Although it is clear that disposal
practices have been deficient in many instances, the development of
optimally safe and cost-effective techniques will require further
research, as will precise assessment of the magnitude of the risks
posed by prevailing levels of contamination around existing dump sices
(21-23).
The assessment of risks cannot depend on epidemiological
approaches alone. This would be tantamount to making guinea pigs of
exposed populations. Instead, comparative toxicological methods
involving laboratory assays and animal models must be exploited
insofar as possible in view of the paucity of toxicological data tor
most chemicals in the human environment (Figure 7). This will
necessitate research to advance the state-of-the-art, in view of
existing uncertainties about species differences and the interactive
effects of the many chemicals that are characteristically present at
dump sites.
IlEXJHAMISilS OF TOXICITY: IMPLICATIONS FUK RISK ASSESS! IB IT
AND DIStiASE PREVENTION
Toxicological Research
As noted above, many of the impacts of environmental agents result
-------
-8-
t'rotn Che combined effects of mulcipLe factors, each of which may
contribute differently to the total. Furthermore, the effects of a
given agent, or combination of agents, may vary, depending on the
conditions of exposure as well as the dose. In addition, although
some chemicals exert their effects directly, many act indirectly,
through the formation of biological active metabolites or through
effects on the metabolism of other substances (4). Because ot the
complexity of these processes, it is difficult or impossible to assess
the effects of a given agent without some understanding of its
metabolism and mode of action. Knowledge of the comparative
toxicology and mechanisms of action of a substance Is also essential
in assessing its potential risks for humans on the basis ot
extrapolation from its observed effect? in laboratory animals, since
choice of the appropriate extrapolation model cannot be made without
assumptions about the relevant dose-effect relationships and
mechanisms of action (25-26).
With respect to the dose-effect relationship, it must not be
forgotten that for some types of environmental insults no thresholds are
known or presumed to exist. These include the mutagenic, carcinogenic,
and some of the teratogenic effects of ionizing radiation (14) and certain
chemicals (
-------
-9-
In addition, since it is noc always feasible co eliminate a toxic
agent from the environment, the most practical approach for mitigating
its noxious effects may be to arrest or reverse them in exposed
individuals. For this purpose knowledge of the mechanisms of such
effects may be crucial, as well as the ability to identify affected
individuals at early enough stages for effective protective
intervention, Methods tor monitoring exposed populations, as well as
for monitoring the environment, are thus needed.
Social and Behavioral Factors
Any consideration of the role of environmental factors in health
should not neglect the influence of social and behavioral
influences (28). Among these, socio-economic status is one of the
most important since-it nay affect many, if not all, other
environmental influences, directly or indirectly. Mortality from many
of the common causes ot death tends to vary inversely with
socio-economic and educational levels (29). The poor who live in
urban ghettos exemplify the problem in their high incidence of
malnutrition, congested and stressful living conditions, vermin
infestation, chronic exposure to dusts and other air pollutants, and
relegation to hazardous working conditions,. Poverty also breeds
deviant behavior, including alcoholism, drug addiction, and crime,
which have enormous impacts on health.
The importance of wholesome daily living habits in those who are
not economically disadvantaged also deserves comment. Such simple
-------
-10-
faccors as regularity of meals, physical exercise, adequate hours of
sleep, control of body weight, abstinence from smoking, and avoidance
of excessive intake of alcohol are correlated with marked reductions
in overall morbidity (30). In Mormons (31) and Seventh Day Advent is ts
(32), who generally practice these habits, mortality from cancer and
many other diseases is appreciably lower than in the population at
large.
Also noteworthy is the inverse correlation between level of
educational attainment and cigarette consumption (33), which points to
the importance of education in motivating people not to smoke or to
stop smoking. The large numbers of people at all educational levels
who continue to smoke, however, attest to the need for further efforts
to solve the problem completely.. The cigarette problem — which
accounts for more than 300,000 deaths per year in the U. S. from
cancer, respiratory ailments, and cardiovascular disease (33) —
exemplifies the importance of behavioral factors, socio-economic
influences , and political forces in shaping the environment for
better or for worse.
UNDER-RECOGNITION AND INDER-DIAGNOSIS OF ENVIKUMlhNTAL DISEASE
As noted above, environmental diseases encompass an extremely
broad range of human illnesses. They include, for example, emphysema
in persons chronically exposed to acid air pollution, leukemia in
persons exposed to benzene, lung cancer and mesothelioma in
individuals exposed to asbestos, chronic kidney disease and neurologic
-------
-11-
impairment in persons exposed to solvents, impairment: of brain
development in children exposed early in life Co lead, heart disease
in individuals exposed to carbon monoxide, and impairment of
reproductive function in men and women exposed to lead and certain
pesticides. Such illnesses afflict millions of persons in the United
States.
Because such environmental diseases arise from man-made
conditions, they can be prevented through the elimination or reduction
of hazardous exposures at the source; i.e., through primary
prevention. They are also amenable to secondary prevention — i.e.,
early detection in prasymptomatic stages when they can still be
controlled or cured; this- depends, however, on efficiently and
effectively identifying populations at high risk. Finally, their
impacts may be lessened by tertiary prevention; i.e., the prevention'
of complications or disability by application of appropriate
diagnostic and treatment strategies. Prevention at all three levels
requires adequate information about the effects of specific
environmental exposures and adequate data on the places and
populations affected.
Laws enacted in the past two decades are intended to prevent
environmentally-induced disease. These include, for example, the
Clean Air Act, the Safe Drinking Water Act, the Resource Conservation
and Recovery Act, and the Superfund legislation. In spite of this
legislation, however, environmentally-induced disease remains
widespread in American society. Given that such illnesses are
-------
-12-
importanc and highly preventable, why do chey scill persist? A series
of factors interact to maintain this situation.
1. Despite at least two decades of regulatory and scientific
awareness and effort, relatively little is known about the
potential health effects of most synthetic chemicals. Host
attention and research have been focused on a small number of
relatively well known hazards, such as asbestos and lead, and
their associated diseases. Virtually no information is available
on the toxicity of approximately 8U percent of the 48,000 chemical
substances in commercial use (Figure 7). Even for groups of
substances which are most closely regulated and about which most
is known — drugs and fooas — reasonably complete information on
possible untoward effects is available for only a minority of
agents (Figure 7). Premarket evaluation of new chemical products
is notably inadequate.
2. Physicians are not trained to suspect the environment as a cause
of disease. Host physicians do not routinely obtain histories of
environmental exposure for their patients, which would allow them
to identify an environmental origin of disease. Recent surveys
indicate that environmental histories are recorded on fewer than
10 percent of hospital charts (34). In consequence, many diseases
of environmental origin are mistakenly assigned to other causes,
such as old age or cigarette smoking, and opportunities for early
prevention or treatment are lost. This problem of inaccuracy in
-------
-13-
diagnosis is compounded by Che face chat disease of environmental
origin are typically not clinically or pathologically different
from those caused by lifestyle and other factors.
3. Physicians do not receive adequate training in environmental
medicine. Very little time is devoted in American medical schools
to teaching physicians in training to recognize the symptoms of
known toxins, or to understand the known associations between
environmental exposures and disease outcomes. The average
American medical student receives only four hours of training in
environmental and occupational health during the four years of
medical school (34).
4. Persons are typically exposed to more than one toxic substance in
the environment and often do not realize that they have been
exposed at all. Further, the symptoms of many environmental
conditions develop only many years after onset of exposure during
this long latency, (incubation). Persons may change addresses, may
be exposed to a variety of environmental exposures, may suffer
various environmental exposures, and finally may forget exposures
which they had many years ago. All of these issues compound the
difficulty that physicians and environmental scientists face in
attempting to deduce the etiology of environmentally induced
illness.
-------
-14-
5. The U. S. Environmental Protection Agency (USEPA) and State
environmental agencies are empowered to investigate hazardous and
environmental conditions; however, severe resource limitations
have reduced the capacity of these agencies to undertake necessary
inspection and enforcement actions.
6. Fragmented, unreliable and outdated surveillance systems for
environmentally related disease produce significant underestimates
of the actual number of cases of environmentally induced illness
in our society. As a result, the picture they plroduce does not
convey an appropriate sense of urgency about reducing the burden
of environmental disease.
In summary, a profound lack of information on the toxicity of the
majority of commercial chenicals, insufficient and inappropriate
education of physicians, and inadequate surveillance impede all
efforts to reduce the impact of environmentally induced disease in the
United States. A coherent plan to improve the surveillance,
prevention, diagnosis, and treatment of environmental disease is
sorely needed. Models which have recently been developed for the
detection, treatment and prevention of occupational disease in states
such as New York, New Jersey, and California might serve as userul
models for undertaking such an effort (35).
INADEQUACY OF RESEARCH SUPPORT
From the foregoing it is evident that much of the burden of illness in
-------
-15-
are produced by man and/or subject: Co his modification. Although some
such agents produce adverse effects only at high dose levels, others
nay cause effects at lower dose levels, conceivably without a
threshold. In practice, furthermore, the observed impacts on human
health frequently result from the cumulative effects of combinations
of agents, in which additive or multiplicative interactions among
causal agents are involved, hence, although environmental factors
have been implicated as major causes of disease, the precise role of
any one causal tactor in the occurrence of a particular disease cannot
always be specified, by the same token, it is difficult to predict
the potential risks to health that may result from a given agent at
any particular dose level. In our present state of knowledge,
assessment of such risks is especially uncertain when direct human
•evidence is lacking and estimates must be based on extrapolation from
observations in laboratory animals or other assay systems.
To advance our understanding of the role of environmental factors
in health and disease, priority must be given to research on the
following: 1) more systematic monitoring and characterization of the
human environment; 2) more adequate recording of human morbidity and
mortality rates, with record-linkage systems to enable the frequency
of specific disease to be related to possible environmental causes; 3)
further development of methods for detecting indices of exposure to
toxicants and for identifying high-risk subgroups; 4) refinement of
laboratory tests for characterizing the biological activity of
chemical and physical agents, especially at low doses and in
-------
-16-
corabinacions; 5) improvenenc in techniques for human risk assessment
with particular reference to comparative toxicological methods and
extrapolation from animal data; and 6) better understanding of the
mechanisms of environmentally-related health effects, as needed for
improvements in risk assessment and in the primary and secondary
prevention of environmentally-related diseases. In addition, more
vigorous efforts should be directed toward the application of existing
knowledge, through: 1) public and professional education, 2)
standards-setting, 3) implementation of new and existing legislation,
4) law enforcement, and 5) research to evaluate the efficacy of such
measures. In pursuit of its mission EPA in coordination with other
agencies and institutions must have a long-range research strategy
addressing each of the above needs.
-------
-17-
REFERENCES
1. Jones, H.B. A Special Consideration of the Aging Process, Disease.
and Life Expectancy.University of California, Radiation
Laboratory,.Berkeley, California, 1955.
2. Fries, J.F. Aging, Natural Death, and the Compression of
Mortality. New England J. tied. 303:130-135. 1980.
3. Donabedian, A., Axelrod, S.J., Swearingen, C., and Jameson, J.
Medical Care Chart Book, 5th Edition. Bureau of Public Health
Economics, University of Michigan, Ann Arbor, Michigan, 1972.
4. Second Task Force for Research Planning in Environmental Health
Science. Human Health and the Environment - Some Research Needs.
U. S. Department of health., Education, ana Welfare, CHEW Publ. No.
NIH 77-127, Washington, D. C., 1977.
5. Levy, B.S. and Wegman, D.H. (Editors) Occupational Health;
Recognizing and Preventing Work-Related Disease. Little, brown
and Co., Boscon, 1983.
6. Cairns, J. The Cancer Problem. Sci. Amer. 233:64-78, 1975.
7. Doll, R. An Epidemiological Perspective of the Biology of Cancer.
Cancer Res. 38:3573-3583. 1978.
8. Inhaber, h. Risfc of Energy Production. ACEG-1119/REV-1, Atomic
Energy Control Board, Ottawa, Canada, May 1978.
9. Morgan, M.G., Morris, S.C., Henrion, II., Amaral, D.A.L., and Rish,
W.R. Technical Uncertainty in Quantitative Policy Analysis - A
Sulfur Air Pollution Example. Risk Analysis 4:201-216. 1984.
10. Morgan, M.G., Henrion, M., Morris, S.C., and Amaral, D.A.L.
Uncertainty in Risk Assessment. Environ. Sci. Technol.
19:662-667, 1985.
11. Letter to the Editor. Chernobyl Public health Effects. Science
238:10-11, 1987.
12. Task Force on Environmental Cancer and Heart and Lung Disease.
Environmental Cancer and Heart and Lung Disease, 3rd Annual Report
to Congress.U. S. Environmental Protection Agency,
Washington, D. C., 1980.
13. Committee on Indoor Pollutants. Indoor Pollutants. National
Academy of Sciences, Washingcon, D. C., L981.
-------
-18-
14. Advisory Commie tee on the Biological Effects of Ionizing
Radiation. The Effects on Population of Exposure to Low Levels of
Ionizing Radiation.National Academy of Sciences,
Washington, D.C., 1980.
15. National Council on Radiation Protection and Measurements.
Evaluation of Occupational and Environmental Exposures to Radon
and Radon Daughters in the United States. NCKF Report No. 78.
National Council on Radiation Protection and Measurements,
Washington, D. C., 1984.
16. Dickson, D. Toxic Synopses: United States. Lessons on Love Canal
Prompt Clean Up.. Ambio 11;46-50, 1982.
17. Committee on Safe Drinking Water. Drinking Water and health. Vol.
4. National Academy of Sciences, Washington, D. C., 1982.
18. Committee on Diec, Nutrition, and Cancer. Diet, Nutrition, and
Cancer. National Academy of Sciences, Washington, D.' C., 1982.
19. Doll, R. The Epidemiology of Cancer. In: Accomplishments in
Cancer Research. 1979 Prize Year General ttotors
dancer Research
Foundation. edited by J.G. Former and J.&. Rhoads, J. B.
Lippincott Co., Philadelphia, 1979, pp. 103-121.
20. Weiss, B. and Clarkson, T. Toxic Chemical Disasters and the
Implications of Bhopal for Technology Transfer. Milbank Quarterly
64:216-240, 1986.
21. Office of Technology Assessment. Technologies and Strategies for
hazardous Waste Control. U. S. Congress, Washington, D. C., 1983.
22. National Materials Advisory Board. Management of Hazardous
Industrial Vfastes. National Academy ot Sciences,
Washington, D. C., 1983.
23. Committee on Response Strategies to Unusual Chemical Hazards.
Assessment of Ilultichemical Contamination. National Academy of
Sciences, Washington, D. C., 1981.
24. National Academy of Sciences - National Research Council. Toxicity
Testing. Strategies to Determine Needs and Priorities. National
Academy Press, Uashington, D. C., 1984.
25. Committee on Chemical Environmental Mutagens. Identifying and
Estimating the Genetic Impact ot Chemical hutagens. National
Academy of Sciences, Washington, D. C., 1982.
-------
26. Ctnenn, G.S. Environment Risk Assessment: Relation to ttutagenesis,
Teratogenesis, and Reproductive Effects. J. Amer. Coll. Toxicol.
2: 113-123, 1983.
27. Bellinger, D., Leviton, A.. Waternaux, C., Neddleman, H. , and
Rabinowitz, *M. Longitudinal Analyses of Prenatal and Postnatal
Lead Exposure and Early Cognitive Development. N. Engl. J. Med.
316:1037-1043, 1987.
28. Eisenbud, M. Environment, Technology, and health, human Ecology
in historical Perspective! New York University Press, New YorkT
T578I
29. Occupational Mortality; England and Wales, 1970-1972. (Decemial
Supplement) . Her- Majesty's Stationery Office, London, 1978.
30. Belloc, N.b. Relationship of Health Practices and Mortality.
Prev. Med. <2:67-81, 1973.
31. Lyon, J.L. Gardner, J.W. , and West, D.U. Cancer Incidence in
Itornons and non-itormons in Utah During 1967-1975. J. Natl. Cancer
Inst. 65:1055-1062, 1980.
32. Phillips, R.L. , Garfinkel, L. , Kuzcia, J.U. , Beeson, N.L. ,
Lotz, T. , and Brin, B. tortality Among California Seventh-Day
Adventists for Selected Cancer Sites. J. Natl. Cancer Inst.
65:1097-1108, 1980.
33. Surgeon General. Smoking and Health. Department of health,
Education, and Welfare, Washington, D. Ci , 1979.
3t. Levy, B.S. The Teaching of Occupational Health in United States
Medical Schools: Five-Year Fol low-Up of An Initial Survey. Amer.
J. Public Health 75:79-80. 1985.
35. Harkowitz, S.B. and Landrigan, P.J. Occupational Disease in New-
York State. Mount Sinai School of Itedicine, New York, 1987.
36. Institute of Iledicine. Costs of Environment-Related Health
Effects ; A Plan for Continuing Study. National Academy Press,
Washington, D. C. ,
37. National Institutes of Health. Data book. U. S. Printing Office,
Washington, D. C. , 1986.
-------
-20-
Table 1. Some Selected Acute Air Pollution Episodes
Estimated Nos,
of Attributed
Place Date Excess Deaths
Meuse Valley, Belgium December 1930 63
Donora, Pennsylvania October 1948 20
London, England December 1952 4,000
New York, New York November 19S3 200
London, England December 1962 700
-------
-21-
Table 2. Ways in Which Diet I lay Affect Incidence of Cancer
1. By providing source of carcinogens or precarcinogens:
— Natural components of plants
— Products of chemical, bacterial or rungal action during
processing or starage
— Products of cokking
— Contaminants (products of fuel combustion, pesticide
residues)
2. By affecting formation of carcinogdens:
— Provision of substances for formation of nitrosamines
(secondary amines, nitrates, nitrites)
— Inhibitionof formation of nitrosamines as in stomach
(Vitamin C)
— Alteration of excretion of bile salts and cholesterol into
large bowel (fat)
— Alteratio no-f metabolism of carcinogens (enzyme induction
by meat, fat, indoles in vegetables, antioxidants)
— Alterationof enzyme formaiton (trace elements)
— Affect on formation of estrogen (fats, total calories)
3. By modifying effects of carcinogens:
— Through transport (alcohol, fiber)
— Through effect on concentration in bowel (fiber)
— Inhibition of promotion (Vitamin A, beta-carotene)
(From Reference 19)
-------
-22-
Table 3.
Toxic Pollutant
'Mercury
PCBa
PBBa
Lead
Dioxina
DBCPa
Kepone
Multiple Agents
Dioxin
Dioxin
Some Acute Environmental Pollution Episodes
Location Year
Miniraata Bay, Japan 1959
Kyushu, Japan 1968
St. Louis, Michigan 1973
Kellogg. Idaho 1976
Seveso, Italy 1976
Lathrop, California 1977
Hopewell, Virginia 1978
Love Canal, New York 1978
Tines Beach, Missouri 1983
Newark, New Jersey 1983
aPCB defined as polychlorinated biphenyls, PBB as polybrominateo
biphenyls, dioxin as 2,3,7,8-tetrachlorodibenzo-p-dioxin, and DHCP as
1,2-dibromo-3-chloropropane.
-------
-23-
Table 4. Examples of Outbreaks or Mass Hunan Poisoning From
Toxic Chemicals
Date3 Location.
1930 U.S.A.
1934 Detroit
1952 London
1952 Japan
1952 Moringa (Japan)
1955 Minamata (Japan)
1956 Turkey
1958 Kerala (India)
1959 Morocco
1960 . Iraq
1964 Niggata (Japan)
19b7 Qatar
19b8 Japan
1971 Iraq
1976 Pakistan
1981 Spain
1984 Bhopal
Chemical
Triorthocresylphosphate
Lead
Air pollutants
Parathion
Arsenic
ttetnylmercury
Hexachlorobenzene
Parathion
Triorthocresylphosphate
Ethylmercury
Methylmercury
Endrin
Poiychlorinated biphenyls
Ilethylwercury
Malathion
Toxic oil
Dimethylisocyanate
No. Affected
16,000
4,000
4,000
1,800
12,159b
1,000
4,000
828
10,000
1,022
646
691
1,665
50,000
7,500
12,600
2,000C
aYear of onset.
blhese were the estimated number of exposed babies. It was stated
Chat several thousand were poisoned and 131 died.
GDeaths. The full scale of lingering and permanent morbidity remains
unknown.
(From Reference 20)
-------
-24-
PER Dr. McKinney - do not use chis page
Figure Legends
Figure 1 Age-Specific Death Bates in Various Countries and
Years (From Reference 1).
Figure 2 The Increasingly Rectangular Survival Curve in the
U.S. About 80 percent (stippled area) of the
difference between the 1900 curve and the ideal
curve (stippled area plus hatched area) had been
eliminated by 1980. Trauma is now the dominant
cause of death in early life. (From Reference 2).
Figure 3 Leading Causes of Death in the United States, 1967,
as Compared witn 190U. (From Reference 3).
Figure 4 Time-Trends in Lung Cancer Mortality and Cigarette
Consumption in England and Wales. (From
Reference 33).
Figure 5 Incidence of Lung Cancer in Regular Cigarette
Smokers in Relation to Number of Cigarettes Smoked
Per Day. (From Reference 6).
Figure 6 Drinking Water Problem Areas (As Identified by
Federal and State Regional Study Teams). Source:
U. S. Water Resource Council. (From Reference 16).
Figure 7 Adequacy of Available Data on Chemicals of
Different Categories for Health-hazard Assessments.
(From Reference 23).
-------
-25-
1000
500
— i I
r i i i i i r
a ina«. 1900
• Mexico. i9dO
• Tfxjiiond. 1947
10
20
3O 4Q 50 6O
YEARS OF AGE
70
80
Figure 1
Age-Specific Death Rates in Various Countries and
Years (From Reference 1).
-------
-26-
100
O
2 75 h
|
CO
H
LU
CJ
cc
Ul
50 -
25 -
10 20 30 40 50 60 70 80 90 100
AGE
Figure 2
The Increasingly Rectangular Survival Curve in the
U.S. About 80 percent (stippled area) of the
difference between the 1900 curve and the ideal
curve (stippled area plus hatched area) had been
eliminated by 1980. Trauma is now the dominant
cause of death in early life. (From Reference 2).
-------
-27-
Percent of all deaths
1900 1967
Cause of death 10 (
Diseases of the heart £"'"-'•
Malignant neoplasms El
\f i • tt
All accidents t.
) 10 20 30 40
f\' .V . '-. ' . V*»V- VvV*>>>.WV*-.,'V,X«\'-1>J
iVvN V^V'vwf
. \ '.' *• \*i
n . *.'.]
r/ --1 1900
3
Diseases of early infancy
General arteriosclerosis
Diabetes mellitus
Other diseases of
circulatory system
Other broncho-
pulmonic diseases
Tuberculosis
Castries, etc.
Chronic nephritis
Diphtheria
a
3
Figure 3
Leading Causes of Death in the United States, 1967,
as Compared with 1900. (From Reference 3).
-------
-28-
5,000
4,000
3,000
< 2,000
1,000
1900
200
- 150
1920
1940
8
I
- 100
1960
1980
Figure 4
Time-Trends in Lung Cancer Mortality and Cigarette
Consumption in England and Wales. (From Reference
6).
-------
-29-
500
ANNUAL 40°
INCIDENCE
standardized
for age
PER 100000 30°
MEN
(x)
200
i
100
T
X
10 20 30 40
DOSE RATE (cigarettes smoked per day)
50
Figure 5
Incidence of Lung Cancer in Regular Cigarette
Smokers in Relation to Number of Cigarettes Smoked
Per Day. (From Reference 7).
-------
-30-
Shaded areas
Open areas
Reported pollution areas
Areas that may not be problem-free, but the problem 1s
not considered major.
Industrial chemicals other than chlorinated hydrocarbons
Heavy metals, such as mercury, zinc, copper, cadmium and lead
Chlorinated hydrocarbons from treatment processes 4 energy
development
Coll form and other bacteria
Saline water
General municipal and Industrial waste
Figure 6
Drinking Water Problem Areas (As Identified by
Federal and State Regional Study Teams). Source:
U. S. Water Resource Council. (From Reference 16)
-------
-31-
5L»« •<
Caccterr
UclMcid Hua P.rcrat
la th« S«l«ct Untv«ra«
Puclcld.i «a4 la.rt
Ci »l Tt»titidu
Catmutle lajr»«l«t»
Dru|. «nd txclplmci
VM.4 ia Dru« rervul^ciac.
3.330
3..10
1.113
food A441:l»««
I.U7
1^^
31
J 14 1
.»
Crt«c«r clua at
1.000.000 Ib/rr
td«n
1.000.000 Ib/rr
tO
C\»icfti« la Caaacre.:
Praduecloa taih
taace..«lbl*
13. »11
U.7JJ
11 U
?^V ^JJS'
12 12
10 I
71
Black bars = Complete health hazard assessment possible
Dotted bars * Partial health hazard assessment possible
Slanted line bars * Minimal toxiclty Information available
Horizontal line bars a Some toxlcity information available
(but below minimal)
Open space bars » No toxicity information available
Figure 7
Adequacy of Available Data on Chemicals of Different
Categories for Health-Hazard Assessments. (From
Reference 24).
-------
-32-
Chaycer 2
KINDS OF LUNG-TEKM RESEARCH
James Fouts
LGNG-TEKM HEAIJH EFFECTS RESEARCH SUPPORTIVE OF EPA PROGKAll NEEDS
I. Basic Research
Basic research needed in EPA programs may or may not be
directed specifically ac supporc of certain applied research programs.
Such basic research may seek only co understand deeper levels of the
general universe of problems attacked in the specific, discrete
long-term, applied researches (such as described just below). The
general basic research philosophy is that understanding more about the
ways chemicals cause disease can lead to earlier detection or better
tests for adverse health effects (and better designs of epidemiology
studies), better analytical methods, etc. All of this can and often
does lead to better bases for regulation and, thus, better regulation.
(See Section III below)
Some of this basic research can be directed at using some of
the "new biology" to advance our ability to assess exposure or to better
identify and quantity speciric bad effects of (or bad actors in) complex
mixtures of chemicals occurring "naturally". Overall though, the
distinguishing feature of this basic research is that it addresses more
"generic" issues, and that it not necessarily be tied into any one
specific problem nor seek "quick" answers. As such, it must be
supported for several years to be effective and to give the kinds of
findings that will be most useful to many "applied" research programs.
-------
-33-
Ic is, however, true chac often the most useful tacts and new approaches
needed in resolving any environmental emergency have come from turning
to laboratories doing good basic research.
There are many examples here of the kind of research that
probes deeper (and is more risky) than any of the applied programs.
Some of these might be:
A. New methods to detect and quantity dioxins
Basic research has identified and characterized an
intracellular "receptor" for dioxins and related compounds. Studies
carried out over many years resulted in the partial purification of this
receptor, and better understanding of the mechanisms, and specificity of
several of the biological effects of dioxins. Recently, using the "new
biology" techniques, this "dioxin" receptor has been cloned and can now
be made available to "methods" (and other) research. Basic research in
such areas/uses has pointed to one possible application of these
studies—the use of this cloned dioxin receptor to isolate/separate and
identify small amounts of dioxins and, at least, some dioxin-like
materials in complex mixtures—particularly of the dioxins in soils,
waste site effluents, etc.
The research on the "dioxin" receptor is the kind of
basic research effort which may now coine to "fruition" (e.g., in the new
methods for assaying dioxins in mixtures), but it has stretched over
many years, and although never without some merit to the raosc
practical/applied of objectives, has not been of immediate value to most
of the EPA needs.
-------
-34-
B. New methods for detecting exposure to some toxic chemicals
The cytochrome P-450s are a component of steroid,
lipid, and xenobiotic metabolizing enzyme systems found in a variety of
living systems (from yeast to humans). Much basic research over at
least 35 years has led to some understanding of the diversity and
responsiveness of these systems in many species. The "new biology"
again has given us some new tools for quantifying and identifying many
of these pigments. It is now possible to "fingerprint" the kinds and
amounts of many different isozymes of P-450 in tissues of many animals
(including humans) and plants. Basic research has described in some
detail the responsiveness of these P-450s to various environmental
stresses (including chemical exposures). Taken all together then, this
long-continuing, basic-research program may now be giving us tools for
looking at the exposures of plants, animals, and humans to many
environmental chemicals—e.g., the amounts and types of P-450s seem to
reflect exposures to things like pesticides, smoke, solvents, etc.
Further, basic research work (especially in pharmacokineti.es) may even
give us a tool for assessing both acute and cumulative/chronic toxic
exposures (of species ranging from fish to humans) using these
monoclonal antibodies for specific P-450s.
II. Applied Research
There are several types of research activity which have
application to specific problems and specific settings, but which must
be carried on over a period of several years. These can be divided into
3 major categories: 1) research programs with discrete and sequential
-------
-35-
parcs/seeps—where one pare muse usually be done before another can be
initiated/planned, 2) research programs that often take a long time, but
parts of which can be carried on concurrently, and 3) methods
development and validation.
A. Long-term research programs best done in sequential steps
This is usually a series of several, discrete projects,
each of which generates data useful/needed in other related
projects—either in their design or execution. There are many examples
here, but the key feature in each is that this is a long-lasting program
with several stages, and each stage feeds into/sets up the next action:
1. The ozone layer and ozone depletion
This is a program which has continued for many years.
The human and ecological health effects implications of this are
enormous. Human health effects of the ozone-layer depletion include
possibly large increases in UV light-induced cancers and other serious
skin diseases. Ecological effects on agriculDare/crops may be equally
human life-threatening, though less direct. There have been many stages
in this overall program:
a. The first studies looked at the issue—Is there
any evidence that we are actually losing stratospheric ozone? The
answer to this (data supporting this) is still being gathered and
debated (at least in some quarters), but the first indications were chat
evidence existed to suggest a loss; therefore, step 2 was needed.
b. The second step seemed to be: What might be
causing this loss of ozone? Is there any human contribution (e.g.,
chemical) which can destroy ozone and which is likely to get to the
-------
-36-
ozone layer? Daca about cne chemistry and interactions of light, ozone,
and hydrocarbons had to be generated here first. Some experiments are
still being carried out at this stage.
c. The next step was to gather data about the
presence of ozone-destroying materials/chemicals (e.g., halogenated
hydrocarbons) in the upper layers of the atmosphere. New methods for
measurement, collection of samples etc. had to be developed, validated
and used.
d. Then real-life sources of these hydrocarbons had
to be sorted out and evaluated for their possible contributions to the
problem.
e. Then decisions as to which steps would be most
effective in changing the amount of hydrocarbons at the ozone layer had
to be decided.
Thus, many types of research were/are involved
here—chemistry, biochenistry, ecology, climate, stratospheric,
marketing, sociological/ psychological, and political, however, the
steps to be taken next in the overall strategy of dealing with this
problem depended on the outcome of those studies made just before and on
most of those preceding.
2. The ecologic and health effects of acid rain
A number of issues have been raised here, but they all
concern whether acid rain or another source ofc pollution has caused the
effects, and what these effects really are. Acid rain is believed to be
formed primarily from industrial sources, though others are also
possible and constitute another subset of evolving issues. Cne example
in this problem area is whether the damage to trees (and other flora,
-------
-37-
nere and in Europe) is due to acid rain from factories and electric
power generation or is caused by pollutants from cars/traffic, etc. A
series of studies" has been made and others are continuing. It is
becoming obvious from some of the results that the answer is "yes" to
both; tree damage (and crop effects/human health effects) may result
from acid rain and car exhaust. This answer comes from a series of
sequential and evolving researches carried out over several/many years.
One of the most recent reports on.all this (including some limited
assessment of human health effects of acid aerosols) is probably the
National Acid Precipitation Assessment Program report issued in
September 1987. human health effects of acid aerosols were recently
re-assessed at an EPA-NIEHS sponsored symposium -held at NIEHS in October
1987. The report of this will be published in Environmental health
Perspectives in 1988. This research effort in both ecology and human
health effects of acid rain has gone on for years /decades, and some
answers are only now becoming barely visible.
B. Long-term studies with concurrent steps
These are studies that just take a long time—the
objectives are such that the study just can't be done in short time
frames. I1any "purely" epidemiology studies fall here—where the
questions concern health effects of low level, chronic exposures or seek
to determine endpoints resulting only years after exposure or in
populations that inust "age" to have detectable effects, Itost studies on
possible causes of cancer or on carcinogenic effects of chemicals are
here. So are evaluations of the causes of many other slowly developing
effects/diseases (e.g., emphysema, kidney failures, liver damage, and
CVS or CNS effects). These evaluations involve multiple studies done at
-------
-38-
the same tine but continuing for a Long time on the sane populations.
Chronic toxicity studies in animals are a subset of this kind of
approach. There are many examples here:
1. The "Six Cities Studies" of health effects of air
pollution—comparing various indices of health in persons living in 6
cities of widely varying degrees of pollution. This Study has been
going on for years now. Some part of the increasing clarity in this
Study results from-more data—accumulated now over more than 10 years,
but some part is the adding of new tests and better data analysis to the
screens for health effects. The point is that this Study required/used
repeated studies of the same populations/regions over several years to
establish effects and to clearly associate these health effects with the
changes in air pollution (which occurred during the years of the study)
in these 6 "regions". The principal effects now being seen are those on
the lung (lung function decrements), but other systems (e.g., kidney,
CVS) may be shown to be affected as these studies continue.
2. The effects of maternal polychlorinated biphenyl (PCb)
exposures on childhood development. This began with several accidents
both in the U.S. and elsewhere (e.g., cooking oil contaminations in
Taiwan and Japan and accidents like the dumping of waste oil
contaminated with PCbs along highways, and the exposures of persons
living near, or walking along these highways). From both short-and
long-term animal studies it was known that many serious effects of PCbs
were not seen acutely but were instead delayed in onset an*d subtle.
Therefore, several epidemiology studies were begun to follow (for
several years) health in populations of PCB-exposed persons and
especially in any children they might have. The effects of various
-------
-39-
levels of macernal exposure co PCBs on childhood development are now
being described in some detail buc only because these
accidentally-exposed populations and a large number of "less-exposed"
and "normal"/unexposed women and children were followed for many years.
3. The effects of polybrominated biphenyl (PBB) exposure.
Again, this began with an accident—the mixing of PBBs into animal feed
and the spread of this chemical/mixture among many farms and into many
parts of the food chain in Michigan. Heavily-exposed persons are still
being monitored for effects, since again, animal studies show that these
effects are delayed and subtle.
C. Development and validation of test methods
In many cases the. methods for detecting and quantifying
new environmental toxins/problem chemicals do not exist at the time such
"problems" are first discovered. This set of "long-term" research
activities is vital in any program seeking to understand and affect
environmental health hazards. There are many examples here, but only a
few can be given:
1. Dioxins (PCDDs) and dibenzofurans (PCDFs)
Chemical methods for detecting, separating, and
quantifying these "families" of toxic materials did not exist when the
first "poisoning" episodes in humans occurred. The amounts of these
materials present in samples from most accidents is very small, and yet,
in animals, these chemicals show toxic effects at extremely low
concentrations. We are only now getting the methods needed to detect,
quantify and selectively identify and separate the wide variety of these
chemicals found in most real life exposures. Some of the newest in
analytical techniques were developed to meet this problem/ series of
-------
-40-
problems. The best of separation and analytical methods were required
Co identify che dibenzofurans as contaminants of the PCBs and dioxin
mixtures and also as contributors
to some of the toxicological effects/problems associated with these
mixtures. This long-term research has stretched over at least twenty
years and is not ended yet. Validation of all these methodological
advances is still occurring.
2. Lead
With/ in several environmental problems we need some
measure of the toxic material in "deep" body tissues. Getting at these
without painful surgery/biopsy or the use of autopsy material is a must
if the amounts of information we need are to be generated—particularly
for long-terra studies, or for uncovering chronic effects (although this
information may also be vital for acute emergencies). Lead, like
several other metals, tends to stay only briefly in readily accessible
body tissues and fluids. Stores of lead and several other chemicals
occur in relatively inaccessible tissues like bone, teeth (or deep fat,
etc.). Methods to measure these "deep" stores of toxic chemical are
urgently needed. Non-invasive methods are especially useful/attractive
for screening/repeated measurements. Newer methods for this in the case
of lead may be possible now with X-ray fluoroscopy. Validation of this
method is now taking place—total time from concept to use will be about
ten years if all goes well—a long-term effort typical here ot several
others.
III. How £PA Uses/Depends on Basic Research Conducted by Other
Federal Agencies
Health research within the EPA is ultimately directed toward
the regulatory mission of the Agency. While such research is often of
-------
-41-
an "applied" and/or "inmediace" nature which answers specific problems
that the Agency must deal with in an expeditious manner, sound basic or
•fundamental research is the only method of improving the scientific
rationale underlying regulatory decisions. It is vital that the EPA
scientific staff maintain current awareness of relevant basic research
by performing such research within the Health Effects Research
Laboratory and by closely following the latest developments in
toxicological research. The Agency cannot effectively accomplish its
research mission without scientists who have competence in and knowledge
of the tools of basic research. Vdthout this competence and knowledge
health scientists within the Agency would be unable to effectively
translate the findings of fundamental research into the applied research
areas most supportive of the Agency's regulatory mission. However,
since basic research performed by EPA represents only a small fraction
of that which is necessary to support its regulatory mission, the Agency
must rely heavily on basic research information developed by other
Federal agencies particularly by the various Institutes of the
Department of Health and Human Services. These organizations have been
reponsible for many of the scientific breakthroughs in molecular
biology, genetics, biochemistry, immunology, and cancer research that
have enabled development of applied methods for exposure monitoring,
dosimetry, toxicological testing, and biochemical epidemiology.
Basic research performed through programs developed at the
National Institutes of Health has substantially impacted the Agency's
regulatory approaches and policies. Research on the molecular basis of
mutation, xenobiotic metabolism, pharmacokinetics, and molecular
dosimetry performed at the National Institute of Environmental Health
-------
-42-
Sciences has found applications at EPA in genetic bioassay development
and improved metabolic activation systems for in vitro test systems,
molecular techniques for exposure monitoring, and advanced methods for
human biochemical epidemiology. Fundamental research by the National
Cancer Institute on mechanisms of carcinogenesis and immune surveillance
has contributed directly to the development of toxicological test
methods and guidelines for cancer risk assessment promulgated by the EPA
Office of Health and Environmental Assessment. EPA is benefiting
directly from widely and federally-funded basic research in the area of
neurotoxicology. The discovery of biochemical differences among various
cell types within the central nervous system (and their concomitant
differential vulnerability) is leading to an. improved understanding of
mechanisms of neurotoxicity and improved methods for the assessment- of
adverse neurotoxicologic responses. These methods will undoubtedly
contribute to future Agency guidelines for neurotoxicity testing.
In addition to the use which the Agency makes of basic
research information generated by other Federal agencies through
indirect means (information appearing in the literature and discussed at
scientific forums), the Agency also depends upon active research
collaborations which take advantage of basic findings and/or expertise.
EPA scientists frequently engage in collaborative studies with
scientists in other governmental agencies as well as their colleagues in
academia who may be funded by these agencies. These research efforts
often take advantage of expertise in new technologies and new findings
that may have applications to the regulatory mission of the Agency. As
an example, research on mechanisms involved in the successful
fertilization of the oocyte has led to interagency collaborative
-------
-43-
research co improve methods for the evaluation of male fertility. Uther
research efforts delineating the fundamental factors involved in dermal
absorption have led to joint interagency research projects centered on
the development of improved methodologies for the assessment of the
kinetics of such exposure.
Clearly, it would be possible to extend this list of relevant
examples since much of the scientific information utilized by the Agency
for regulatory decision-making and guidelines formulation rests on a
foundation of basic research.
-------
- 44 -
Chapter 3
RESEARCH ADVANCES IN UlL TUXICOLUGY OF LEAD
Kathryn Hahaffey
PREAMBLE
The place of and necessity for long-sustained basic research activity
in the development of a foundation for constructive action in important
problems in environmental health could be illustrated by reference to any
of several current problem areas. We have chosen the story about lead and
its dangers or toxicity to serve this purpose. Lead as a public health
problem has been recognized for years (if not centuries). Yet how, what,
and when to do something about both preventing its health effects and
treating those not prevented have been obvious only recently, and only as a
result of long-continuing basic research. For one thing, only long-range,
multidisciplinary, continuing basic research has given us the varied tools
we need to detect some of the more subtle (yet extremely important) effects
of lead. We have moved from counting dead bodies to worrying about things
like changed behavior and nerve damages in lead-exposed children—but only
because we now have some good tests for such'effects of lead. This then is
the story of an environmental health research success—made possible only
because such slow-moving (and sometimes hard to explain) studies were
pursued and supported by tar-sighted people who believed that long-range
research was and would continue to be extremely cost-benefit positive.
-------
-45-
Background
Understanding Che range of adverse health effects produced by lead
exposure has advanced markedly in this century. Research into the toxic
effects of lead provides a paradigm that has guided the entire discipline
of clinical and laboratory toxicology for the past five decades.
Fundamental uiultidisciplinary laboratory research in such areas as
biochemistry and physiology has been a major key to this progress.
Lead has long been recognized to be acutely toxic at high-dose
exposure. In addition, we now recognize, based on reearch findings in the
1970's and 1980's, that lead toxicity reflects two patterns of lead
exposure. Adverse neurobehaviorial effects of lead on infants occur at
levels within one standard deviation of the mean concentration of the
United States population. Superimposed on the general population lead
exposure is an additional severe problem of high-level lead exposure
concentrated among young children from lower socioeconomic families,
particularly those from urban areas.
In children, high-dose exposure to lead, such as results from ingestion
of lead-based paint, has been shown to cause a profound neurologic syndrome
characterized by coma, convulsions, and in severe cases death. In adults
with high-dose exposure to lead, abdominal cramping, a syndrome termed
"wrist and ankle drop," and end-stage renal disease are the well-recognized
consequences.
The challenge has been to understand that the range of health problems
caused by lead was much more extensive than the clinically-obvious disease.
-------
-46-
What has made this challenge especially difficult is thac environmental
lead pollution has been at very high levels, producing an elevated body
burden of lead in a sizable portion of the population. During the 197U's
in metropolitan areas, young children frequently had blood lead
concentrations greater than 40jug/dl—; a concentration now associated with
several neuropsychological impairments. The challenge is to perceive the
etiology and severity .of health problems that are so common they are
considered "normal." In the paradigm of lead public health and preventive
medicine have progressed from enumerating mortality and morbidity (i.e.,
case reports) to understanding the disease process. This progress reflects
and has been possible only because of long-range support of environmental
research.
Among the most exciting recent findings with respect to understanding
of the toxicology of lead is the realization that lead is capable of
producing toxic effects in adults and children at relatively low levels of
exposure, i.e., levels that are insufficient to produce grossly clinical
symptoms. Only a decade ago such levels of lead exposure were considered
"safe". Lead is now recognized to produce a syndrome of subclinical
toxicity.
Recent research has demonstrated that this subclinical toxicity of lead
is a many-faceted syndrome involving multiple organ systems. The
developing red blood cells, the nervous system, and the kidneys are the
organ systans in which these toxic effects have been more intensively
studied.
-------
-47-
In the early 1900's lead exposures were so high that occupational
records routinely reported lead-induced mortality statistics. For example,
Hoffman (1935) reported that the nuaber of deaths attributed to lead
poisoning for the United States registration area between 1900 and 1933 was
in excess of 3400. The nunber of deaths among children, who are more
susceptible to the effects of lead exposure, remains largely unknown. In
the 1940*s through the 1960's descriptive reports ot clinical aspects of-
the disease dominated the literature. Prior to the introduction of
chelation therapy, severe lead poisoning with encephalopathy had a
mortality rate of 65% (NRG, 1972).
Among survivors of lead poisoning profound neurological damage is the
predominant, reported effect. For example, Byers and Lord (1943) and other
clinicians showeci long-term residual sequelae of acute pediatric lead
poisoning which included mental retardation, seizures, optic atrophy,
sensory motor deficits, and behavioral dysfunctions. PerIstein and Attala
(1966) reported such sequelae in 37% of children who suffered lead
Through screening programs to identify children with lead toxicity
before they become symptomatic, and through legal requirements to monitor
occupational exposures of workers to lead, severe clinical cases of lead
toxicity have been brought under a degree of control; however, they have
not been eliminated. These case reports, describing clinical aspects of
intoxication, have identified which organ systems are most affected at
high-dose exposures. The limited reversibility or irreversibility has been
-------
documenced in many of che clinically-reported, neurologic effaces. Using
chese clinical studies as a guide, long-range, multidisciplinary research
'has extended the understanding of lead toxicity to the current emphasis on
biomarkers of exposure, dose-response relationships for specific effects,
and identification of susceptible subgroups for these effects.
Research Findings in the 1970's and 1980's
The general picture of adult and pediatric lead poisoning has changed
in recent decades. The overall pattern is identification of significant
adverse health effects at progressively lower exposures. These can be
arbitrarily separated into neurobehavioral, hematopoietic, renal/endocrine,
and reproductive effects. As a part of this effort, differential
sensitivity of various subpopulacions has been revealed. Identification ot
effects occurring at environmental exposures once considered "normal" has
coincided with reducing environmental exposures to lead. Only through
reduced exposures can the results given by toxicology and epidemiology
research be evaluated in general human populations.
I. Neurobehavioral Effects
Kecognition that neurobehavioral effects in children are produced
by lead exposures considered "normal" in earlier decades (e.g., blood lead
concentrations of 20-50 ;ug/dl) has been among the most significant research
findings in the 1970's and 1980's. Longitudinal studies during the past
10-15 years built upon early case reports and cross-sectional studies. The
longitudinal prospective designs have permitted gathering improved
-------
-49-
information on exposure hiscories. Information on exposure levels and
patterns is clearly important in assessing effects of a cumulative toxicant
on endpoints such as neurobehavioral function that may reflect changes
induced at far earlier, but critical, developmental periods.
The most consistent finding of the prospective studies is that an
association exists between low-level lead exposures during developmental
periods (especially prenatally) and later deficits in neurobehavioral
performance. This latter is reflected by indices such as the Bayley Mental
Development Index, a well-standardized test for infant intelligence. Blood
lead concentrations of 10-15 ug/dl constitute a level of concern for these
effects (EPA, 1986). In addition, impaired iieurophysiological function has
been associated with increasing blood lead concentrations among children.
These functional deficits include changes in the auditory brainstem evoked
potentials and evidence of lead-related reduced hearing acuity (Robinson et
al., 1985, 1987). These subclinical toxic effects of lead on the central
nervous system are generally considered to be permanent and irreversible,
and they are associated with permanent loss of intelligence and
irreversible alteration in patterns of behavior.
Bellinger et al (1987) reported significantly lower post-natal
development scores on the Mental Development Index of infants from an
upper-middle class population when maternal blood lead levels were in che
ranage of 10-25 ug/dl. Among adult women ages 20-40 years mean, blood lead
levels were between 10 and 12 ug/dl based on the NHAHES II general
population data for the period 1976-1980 (tlahatfey et al, 1982). Thus, it
-------
-50-
musc be emphasized that these neurobehavioral changes are associated with
blood lead levels within one standard deviation of the mean blood lead
level of the United States' population reported in the NhANES II data.
The peripheral nervous system is also affected by lead.
Typically, adults are likely to demonstrate peripheral rather than central
nervous system effects. In the early 1960's investigators began to call
attention to "subclinical" neuropathy manifested by changes in peripheral
nerve conduction velocity in lead workers not having overt neurological
involvement (Sessa et al., 1965). In the 1970's Seppalainen et al. (1972,
1975) reported the slowing of the maximal motor conduction velocity ofc the
median and ulnar nerves and other electromyelographic abnormalities in
workers whose blood lead concentrations never exceeded 70 ug/dl.
Investigations ot the behavioral effects of lead uncovered an increased
hearing threshold, decreased eye-hand coordination, and other physiological
and psychological changes in workers with blood lead concentrations below
80 ug/dl (Repko et al., 1975;.
II. Hematopoiesis
Anemia has been a symptom of severe clinical lead poisoning in
both children and adults. Anemia (increased prevalence of hemotocrit
values below 35%) is now recognized to become evident in one-year-old
children at blood values of 30 ug/dl. Lead interferes with synthesis of
heme and the formation of hemoglobin at a number of metabolic steps. In
the developing red blood cells lead inhibits the enzyme -aminolevulinic
acid dehydratase to increase levels of erythrocyte protoporphyrin in
-------
-51-
children. The threshold for this effecc in children is associated with a
blood lead concentration of 15-18 ug/dl (Piomelli et al.. 1982).
Impaired heme biosynthesis produces effects in addition to anemia.
The accumulation of protoporphyrin IX (measured as zinc protoporphyrin or
as protoporphyrin in erythrocytes) is not only an indicator of diminished
heme biosynthesis but also signals general mitochondria! injury. The final
step of heme biosynthesis occurs in the mitochondria. Such injury to the
mitochondria can impair a variety-of subcellular processes including energy
metabolism and homeostasis. Health implications of such impairment
include: reduced transport of oxygen to all tissues; impaired cellular
energetics; disturbed immunoregulatory role of calcium; disturbed calcium
metabolism; disturbed role in hematogenesis control; impaired
detoxification of xenobiotics; and impaired metabolism of endogenous
agonists (e.g., metabolism of tryptophan).
III. Renal Effects
Acute high-dose lead exposure in children produces a Fanconi-type
syndrome with glucosuria, phosphaturia and aminoaciduria secondary to
poisoning of the proximal convoluted tubule. High-dose exposure to lead in
childhood has been associated with gloraerular nephritis and renal disease
in adults. Among occupationally-exposed adults, an increased rate for
mortality from all causes, from all neoplasms (specifically, cancers of the
stomach, liver, and lungs), from chronic nephritis, and from other
hypertensive disease (i.e., hypertension due to kidney damage and not hearc
-------
-52-
disease) were observed in a longitudinal study of workers in lead battery
planes and lead smelters (Copper, 1985).
A statistically-significant relationship has been reported between
increases in systolic and diastolic blood presures and increases in blood
lead among 40-to-59 year old, white males from the NHANES II survey
population (Pirkle et al, 1985).
Impairment of the endocrine functions of the kidney have been
reported to occur at much lower lead exposures. Recognition ot these
effects required development of several areas of research:
A. Understanding the metabolic activation of Vitamin D to
1,25-dihydroxyvitamin D. This metabolite is critical to regulation of
calcium metabolism.
B. Recognition that lead impairs various steps in both
biosynthesis and function of 1,25-dihydroxyvitamin D.
Currently, the most studied site at which these metabolic pathways
converge is the proximal convoluted tubule of the kidney. Here
25-hydroxyvitamin b, formed in liver from Vitamin D, undergoes a second
hydroxylation which is catalyzed by the enzyme 1, o,25-hydroxyvitamin D
hydroxylase. Kesearch using ^n vitro techniques (following in vivo
exposure of chickens to lead) has demonstrated that lead inhibits the
activity ot this enzyme. Findings from a clinical investigation among
young children indicated that plasma 1,25-dihydroxyvitamin D levels were
depressed in proportion to blood lead concentration, delation therapy co
reduce body burden of lead, resulted in»increasing serum concentrations of
-------
-53-
1,25-dihydroxyvitamin D up co Levels similar to chose present in children
serving as controls (Rosen et al., 1980). Additional epidemiological
research has shown that 1,25-dihydroxyvitamin D concentrations were
decreased with increasing blood lead concentration over a broad range of
blood lead concentrations, 12 to 120 ug/dl (Mahaffey et al., 1982b).
IV. Reproductive Effects
Early in the century a number of adverse effects of lead on
reproduction were reported among women-with occupational lead exposures.
These included increased spontaneous abortion rate, increased still-birth
rate, and a higher, post-natal and early childhood mortality rate among
children of such exposed women. Exposures associated with these adverse
outcomes were very high. However, longitudinal, prospective studies,
designed to evaluate neuropsychological effects of lead, have provided
important information on reproductive effects at the upper range of current
environmental levels. McHichaels et al. (1986) found that the incidence of
preterra deliveries (before the 37th week of pregnancy) were significantly
related to maternal blood lead at delivery. When late fetal deaths were
excluded, the strength of the asocial:ion increased. The relative risk of
preterra delivery at exposure levels reflected in blood lead concentrations
of 14 ug/dl or higher was 8.7 times the risk at blood lead concentrations
up to 8 ug/dl. Reduction in gestational age at delivery with increasing
blood lead concentrations were also reported by Dietrich et al. (1986),
Bellinger et A!. (1984), tbore et al. (19b2), and bornschein et al. (I987a, b).
-------
-54-
The data from Bornschein indicate than for each 10 ug/dl increase in
blood lead concentrations birth weight decreased between 58 and 601 grams
depending on the age of the mother.
The findings of McMichaels et al. (1986) also identified an excess
in miscarriages and still births in the high-lead exposure areas. In
contrast, data fcrora this study show that average, maternal blood lead
concentration was lower for still births than for live births. Placental
response to lead remains an unanswered question.
Basic research in the toxic effects of lead at low doses is of profound
importance for the fields of preventive medicine and public health. Until
recently, blood lead concentrations of 25 ug/dl and below were considered
safe, and indeed, only five years ago the Centers for Disease Control (CDC)
stated that 25 ug/dl should constitute a threshold level indicative of
increased lead absorption in children. Now, on the basis of recent
research it is evident that lead produces toxic effects-in children at
levels below this guideline. Thus, recent research into the toxicity of
lead at low doses is about to force a total re-evaluation of current
standards for assessing the exposure of American children to lead.
The importance of these basic research findings stems from the fact
chat lead exposure retrains extremely widespread among children in the
United States. Data from the Second National Health and Nutrition
Examination Survey (NHANES) indicated that in 1980 9.1% of all preschool
children in the United States -1.5 million children - had blood lead
concentrations of 25 ug/dl or more (tlahaf fey et al., 1982a). Among
-------
-55-
black preschool children che prevalence of Increased lead absorption (high
blood lead concentrations) was 25%.
These findings on the high prevalence of increased lead absorption
(high blood lead concentrations), when taken in conjunction with the data
on subclinical lead toxicity, carry a message of chilling significance.
These findings suggest that 9% of all children in this nation, and 25% of
minority children, may be suffering irreversible neurologic, intellectual,
and behavioral impairment as the result of chronic, low-dose exposure to
lead. The implications of these basic research data for public health and
environmental medicine are enormous.
This then has been a very condensed story about one of the many
pervasive and important environmental health hazards. It is a story that
continues beyond the present findings and their implications. It will
reach even more successful conclusions only, if the kind of studies which,
brought us to this stage are continued. Continued long-range and basic
research investigations on lead toxicity are at one and the same time
perhaps among the more justifiable and yet less supportable of such
activities in the entire field of environmental health sciences. So ranch
has been done before in lead research that in comparison, no other (few at
least) of all the current health hazards has received chis much emphasis.
Yet it is obvious that this sustained effort in Lead research lias paid off
handsomely and is still needed. It is this "apology" for long-range, basic
research that we teel can stand cor the entire field of environmental
health science, whatever may be the specific stage ot development of this
research for any one hazard.
-------
- 56 -
Chapeer 4
NEWER BASIC/LONG-TERM RESEARCH
WITH
APPLICATION TO ENVIRON iBITAL HEALTH PROBLEMS
PREAMBLE
"In this Chapter a number of authors discuss some of the newer
basic/long-terra research with possible applications to current
environmental health problems (especially in humans). This does not
represent the whole universe of possible basic/long-range research which
will or could be of great, benefit to such environmental issues. It is,
however, an attempt at careful choices of those studies which have required
such long-term support for the reaching of this stage where their
applications could have great impact on environmental health. AS such
then, this is a look at the many and more probable benefits of supporting
such long-range research more adequately than has been done in the past."
-------
-57-
ACTIVATION OF PROTO-ONCGGENES bY CHEMICALS
Ilarshall Anderson
INTRODUCTION
Proco-oncogenes are cellular genes chat are expressed during normal
growth and development processes. These genes were initially discovered as
the transduced oncogenes of acute transforming retroviruses (1). Recent
studies have established that proto-oncogenes can also be activated to
cancer causing oncogenes by mechanisms independent of retroviral
involvement (2-4). These mechanisms include point mutations or gross DMA
rearrangements such as translocations or gene amplifications. The
activation of proto-oncogenes by genetic alterations results in altered
levels of expression of the normal protein product, or in normal or altered
levels of expression of an abnormal protein.
ACTIVATION OF PROTOONCUGEWES
The activation of proto-oncogenes in spontaneous and chemically-induced
rodent tumors and in human tumors has been studied in great detail during
the past several years. Investigations in rodent models for chemical
carcinogenesis imply that certain types of oncogenes are activated by
carcinogen treatment and that this activation process is an early event in
tumor induction (5-6). Alternatively, analysis of some human and rodent
tumors suggests that oncogene activation is involved in neoplastic
progression (7-9). The number of proto-oncogenes that must be activated in
the multistep process of neoplasia is unclear at present. The concerted,
-------
-58-
low level expression of ac lease two oncogenes, ras and myc, are needed tor
the partial transformation of primary rodent cells in vitro (10).
Furthermore, in addition to the activation of proto-oncogenes, the loss of
specific regulatory functions such as tumor suppressor genes may be a
distinct step in neoplastic transformation (11). The implication of
activated oncogenes in rodent tumor will be discussed in terms of
extrapolation of rodent carcinogenic data to human risk assessment.
The activation of ras proto-oncogencs appears to represent one step ii
the multistep process of carcinogenesis for a variety of rodent and human
tumors (5,6). The activation of ras by point mutations is probably an
early event in turaorigenes is and may be the "initiation" event in some
cases. Thus, a chemical that induces rodents tumors by activation of ras
proto-oncogenes can potentially invoke one step of the neoplastic process
in humans exposed to the chemical. Is this property alone enough to
classify the chemical as a potential human carcinogen? Dominant
transforming oncogenes other than ras have also been detected in
chemical-induced rodent tumors (6). The involvement of these oncogenes in
the development of human tumors is unclear at present, as well as whether
the non-ras genes detected in human tumors can be activated by chemicals or
radiation (6).
ONCOGENE ANALYSIS
ttost chemicals are classified as potentially hazardous to humans on die
basis of long-term carcinogenesis studies in rodents. While these rodent
carcinogenesis studies are often designed to mimic the route of human
exposure in the environment or workplace, the dose of a given chemical is
-------
-59-
usually higher Chan chat which actually occurs in human exposure. Coupled
with Che appearance of species- and strain-specific spontaneously occurring
tumors in vehicle-treated rodents, this complicates the extrapolation of
rodent carcinogenic data to human risk. Cncogene analysis of tumors from
spontaneous origin and from long-term carcinogenesis studies should help
determine the mechanisms of tumor formation at a molecular level. For
instance, the finding of activating mutations in different codons of the
H-ras gene in furan-induced liver tumors versus finding activating
mutations in only one codon of the n-ras gene in spontaneous liver tumors
suggest that the chemical itself activated the H-ras proto-oncogene by a
genotoxic event (12). In general, comparison of patterns of oncogene
activation in spontaneous versus chemically-induced rodent tumors, together
with cytotoxic information, should be helpful in determining whether the
chemical in question is mutagenic, cytotoxic, has a receptor mediated
mechanism of promotion, or some combination of these (and other) modes of
action. This type of analysis might be of particular importance for
compounds such as furan and furfural (12,13) which are negative for
mutagenicity in short-term bioassays.
APPLICATION TO STUDY UF CARCINOGEN ICITY
Another approach which should be helpful in species-to-species
extrapolation of risk from carcinogenic data is to examine oncogene
activation and expression in tumors from different.species induced by the
same agent. For example, K-ras oncogenes with the activating lesion in
codon 12 were observed in both rat and mouse lung tumors induced by
tetranitromethane (14). Even though little is known about the DMA damaging
-------
-60-
properties of this chemical, these aaca suggest t±iac this compound is
acting in the same manner to induce tumors in both rats and mice.
The role of chemicals and radiation in the activation of proto-oncogenes by
gene amplification, chromosomal trans location, and other mechanisms which
can alter gene expression, is currently being investigated by several
groups. Also, as human life span increases, it becomes'more important to
study cnemical-induced enhancement of the progression of benign to
malignant tumors. These and similar approaches to explore the mechanisms
by which chemicals induce tumors in animal model systems may remove sorae of
the uncertainty in risk analysis of rodent carcinogenic data.
-------
-bl-
REFERtNCES
1. Bishop JM. 1985. Viral Oncogenes. Cell; 42:23-38.
2. Varmus HE. 1984. The Molecular'Genetics of Cellular'Oncogenes.
Annual Rev Genet; 18:553-612.
3. Weinberg RA. 1985. The Action of Oncogenes in the Cytoplasm and
Nucleus. Science; 230:770-776.
4. Bishop JM. 1987. The Moleuclar Genetics of Cancer. Science;
235:305-311.
5. Barbacid M. 1987. Ras Genes, Aim Rev of Biochem 56, in press.
6. Anderson M, Reynolds S. Activation of Oncogenes by Cheaical
Carcinogens in: The Pathology of Neoplasia. A Sirica, ed, Plenum
Press, N.Y., N.Y. (In press 1988).
7. Brodeur oil, Seeger RC, Schwab M, Varmus HE, Bishop JM. 1984,
Amplification of N-rayc in Untreated Neuroblastoraas Correlated with
Advances Disease Stage; Science 224:1121-1124.
8. Seeger RC, Brodeur Gl, Sather H, Dalton A, Siegel SE, Wong KY,
Hammond D. 1985, Association of Multiple Copies of the N-rayc
Oncogene With Rapid Progression of Neuroblasts, The New England
Journal of Medicine; 313:1111-1116.
9. Slamon DJ, Clark CM, vJbng SG, Levin UJ, Ullrich A, McGuire WL. 1987,
Human Breast Cancer: Correlation of Relapse and Survival With
Amplification of the HER-2/neu Oncogene; Science 235:117-182.
10. Land H, Parada LF, Weinberg RA. 1983. Tumorigenic Conversion of
Primary Einbryofibroblasts Requires at Least Two Cooperating
Oncogenes; Nature (London) 304:596-602.
11. Barrett JC, Oshimura M, Koi M. 1987. Role of Oncogenes and Tumor
Supressant Genes in a Multistep Model of Carcinogenesis, In:
Symposium on Fundamental Cancer Research. Volume 38 (F. Becker,
ed.,), in press.
12. Reynolds SH, Stovers SJ, Patterson R, Maronpot RK, Aaronson SA,
Anderson MW. 1987. Activated Oncogenes in B6C371 Mouse Liver
Tumors: Implications for Risk Assessment, Science 237:1309-1316.
13. Tennant NW, Margolin BH, Shelby MD, Zeiger E, Haseman JK, Spalding J,
Caspary W, Resnick il, Stasiewica S, Anderson B, Minor R. 1987.
Prediction of Chemical Carcinogenicity in Rodents from In Vitro
Genetic Toxicity Assays, Science 236:933-941.
14. S towers SJ, Glover PL, Boone LR, Maronpot RR, Reynolds SH,
Anderson MW. 1987. Activation of the K-ras Proto-oncogene in Rat and
Mouse Lung Tumors Induced by Chronic Exposure to Tetranitromethane,
Cancer Res 47:3212-3219.
-------
-b2-
CAKCINOGEN-DNA ANb PROTEIN ADDUCTS: RESEARCH PERSPECTIVES
Frederica P. Perera
INTRODUCTION
Advances in basic research in molecular biology and biochemistry have
permitted the development of innovative methods applicable to studies of
human populations exposed to chemical carcinogens. These highly sensitive
techniques can detect and sometimes quantify the internal dose of
carcinogens (the amount of the carcinogen or its metabolite in body tissues
and fluids) or the biologically effective dose (the amount that lias
interacted with cellular macromolecules such as DNA, KNA or protein) in
target tissue or a surrogate. This latter type of dosimetry data could be
especially valuable in studies of cancer etiology by providing a
mechanistically relevant link between external exposure data on the one
hand and clinical disease on the other. Comparable molecular dosimetry
data in rodents and humans have the potential to improve interspecies
extrapolation of risk in addition to providing early warning of a
carcinogenic hazard to humans. Successful applications of such "adducts
research" could directly address major programs/needs at EPA for better
estimates of exposure and risk to humans.
Various methods are available to monitor chemical-speciric lesions
(such as immunoassays for DNA and protein adducts) as well as non-chemical
specific biologic alterations (such as cytogenetic effects or somatic cell
mutations). Table 1 gives examples of currently available methods for
measuring the biologically erfective dose of carcinogens. As can readily
be seen, all pertain to endpoints associated with carcinogens that exert
genetic toxicity. Moreover, almost all available methods depend on readily
-------
-03-
available cells or fluids as surrogates for the actual target tissue
itself. Despite these limitations, biological markers have significant
potential usefulness in cancer etiology and risk assessment.
Table 1. Examples of Human biologic Monitoring Methods(a)
Sites &
End Point
Method
Biologically effective dose
Adduces (DMA)
escence specCroraetry
Adducts (protein)
Excised adducts
UDS
SCE
ItLcronuclei
Chromosomal aber-
rations
Soiiiatic cell mutation
(HGPKT)
Somatic cell mutation
(glycophorin A)
Sperm quality
Immunoassay, postlabeling, fluor-
Mass spectroraetry, ion-exchange
atnino acid analysis, HPLC, gas
chroraatography
HPLC, fluorescence
Cell culture, thymidine incorpor-
ation
Cytogenetic
Cytogenetic
Cytogenetic
Autoradiography, light microscopy
Iimunoassay
Analyses of count, morphology,
motility
FluidsHigh Performance Liquid
Chromatography; SCE=Sister Chromatid Exchange; HGPKT=Hypoxanthine
Guanine. Phosphoribosyl Transferase
-------
-64-
ADDUCTS
Carcinogen-DNA and carcinogen-protein adduces have been Che focus of
considerable research in Che pasc 5 years and illustrace a number of
sCrengChs and limications common co biological markers in general (2,3).
Biological Basis
The biologic rationale for measuring DMA adduces is chat these lesions,
if unrepaired, can produce a gene ciucation. There is considerable evidence
that gene mutation in somatic cells "initiates" Che mulciscage process of
carcinogenesis (4,5); but it may also result in conversion of tumors to the
malignant state (6,7). Carcinogen-DMA adduces resulting in gene mutation
may also activate certain oncogenes instrumental in carcinogenesis
(8,9,10).
Protein such as hemoglobin can, in theory, act as a more readily
available surrogate for DMA. Proportionality between protein and DNA
binding has been demonstrated for a number of carcinogens (11,12,13).
Adduces are generally monitored in peripheral blood cells rather than
target tissue. However, for only a few carcinogens (e.g., benzo(a)pyrene
and cis platinum) is there accual experimencal and/or human evidence ehac
comparable levels are formed at both sices (14,15).
IlEIHODS
Techniques Co measure carciaog-^-DNA adduces include immunoassays using
adduce-specific polyclonal or monoclonal ancibodies, synchronous
fluorescence spectroscopy, HPLC fluorescence speccrophoCoraecry, and
32p-posclabelling. Carcinogen-procein adduces may be decermined using
-------
-65-
antibodies and gas chroraatography-raass spectroraecry. The sensicivicy of
the DNA Co adduce methods is in the range of one adduct per 106-1010
nucleotides. Those methods aimed at carcinogen-protein adduct
quantification also appear to have adequate sensitivity for environmental
studies (16). However, unambiguous identification of particular DNA
adducts at low levels is difficult with present analytical methods.
Moreover, cross-reactivity of antibodies (such as the BPDE-I-DNA antibody
which also detects closely related polycyclic aromatic hydrocarbon (PAH-
adducts) presents problems in definitive characterization of adducts (17).
ANII1AL AND HUMAN STUDIES
Experimental studies involving acute and/or chronic exposure to diverse
carcinogens have shown that the relationship between administered dose and
macromolecular binding is generally linear-with few exceptions
(12,2,18,3). With respect to humans, carcinogen-DNA and -protein adducts
have been investigated in human populations with exposures such as
cigarette smoke, PAHs, tobacco and betel nut, dietary aflatoxin and
N-nitrosamines, cis platinum, psoralen, 4-aminobiphenyl, propylene oxide,
vinyl chloride and ethylene oxide (3).
While results thus far support the feasibility and adequate sensitivity
of the methods in terms of human studies, they are frequently limited by
technical variability in the assays, small sample size, lack of appropriate
controls, and inadequate data about exposure. However, they consiste-if.ly
illustrate that there is significant variability in the formation ot
carcinogen-DNA and -protein adducts between individuals with comparable
exposure or administered dose (15,19-25). Another consistent finding in
-------
-66-
the human studies involving environmental exposure, is chac measurable
levels of adduces are seen even in so-called "unexposed controls"
(19-20,26-29). Both of these observations have obvious implications for
risk assessment.
Although still largely in the validation stage, methods to monitor DMA
and protein adducts in experimental animals and humans have considerable
potential in a number of areas. These include: hazard identification,
understanding of mechanisms involved in carcinogenesis, interspecies risk
extrapolation and improving the power and timeliness of epidemiology
(19,26,30-32).
Research Needs
Research is needed in the following areas:
A. Interlaboratory validation of methods as has' been undertaken
recently for PAH-DNA inmunoassays (33).
B. Research on the stabililty of adducts in stored tissues.
C. Investigation-of intra-and inter-individual variation in adduce
levels.
D. Research on the persistence of adducts in various cells and
tissues.
£. Comparison of adduce levels in DMA versus protein as well as in
surrogate versus target tissue for a number of different classes of
compounds.
F. Identification of critical sites or "hot spots" on DMA with
respect to the carcinogenic effectiveness of adducts.
-------
-67-
G. Incerspecies comparisons of DNA and protein adduce formation
(e.g., humans and rodents with acute and chronic exposure to the sane
corapound(s)).
H. Experimental and human studies on die relationship between
adduct formation, gene mutation, and oncogene activation.
I. Longitudinal studies (experimental and human) on the relationship
between adduct levels and tumor incidence/cancer risk. Examples would be
molecular epidemiological studies in model populations (such as patients
exposed to high dose chemotherapy and who experience a high rate of
secondary cancer, or heavily-exposed worker groups). Biologic samples
could be drawn at the outset and stored for future analysis.
J. Sample banks to serve as archives of human blood, urine, and
tissue for retrospective analysis.
-------
-68-
REFERhNCES
1. Perera F. 1987. Molecular Cancer Epidemiology: A New Tool in
Cancer Prevention, J Nad Cancer Insc, 78, 887-898.
2. Wogan GN, Gorelictc NJ. 1985. Chemical and Biochemical Dosimetry of
Exposure co Genotoxic Chemicals. Environ Health Perspecc, 62, 5-18.
3. Perera F. The Significance of DMA ana Protein Adducts in Human
Biomoni coring Studies, Mut Res (in press).
4. Weinstein IB, Gattoni-Celli S, Kirschmeier P, Lambert M, Hsiao U,
Backer J, Jeffrey A. 1984. Multistage Carcinogenesis Involves
Multiple Genes and Multiple Mechanisms, Cancer cells 1. The
Transformed Phenotype, Cold Spring Harbor Laboratory. -New York, pp.
229-237.
5. Harris CC. 1985. Future Directions in the Use of DNA Adducts as
Internal Dosimeters for Monitoring Hunan Exposure to Environmental
Mutagens and Carcinogens. Environ Health Perspec, 62, 185-191.
6. Hennings H, Shores R, Wenk ML, Spangler EFr Tarone R, Yuspa SH 1983.
Malignant Conversion of House Skin Tumors is Increased by Tumor
Initiators and unaffected by Tumor Promoters. Nature .(London), 304,
67-69.
7. Scherer E. 1984. Neoplastic Progression in Experimental
Hepatocarcinogenesis. Biochim Biophys Acta, 738, 219-236.
8. Beland FA, Kadlubar Ft1. 1985. Formation and Persistence of Arylamine
DNA Adducts In Vivo. Environ Health Perspect, 62, 19-30.
9. Marshall CJ, Vousden KH, Phillips DH. 1984. Activation of c-Ha-ras-1
Proto Oncogene by In Vitro Modification with the Chemical Carcinogen,
Benzo(a)pyrene Diol-epoxide, Nature (London), 310, 586-589.
10. Hemminki K, Forsti R, Mustonen R, Savela K. 1986. DMA Adducts in
Experimental Cancer Hn^-arch. J. Cancer Res Clin Oncol, 112,131-188.
11. Ehrenberg L, Moustacchi E, Osterman-Golkar, Eknan G. 1983. Dosimetry
of Genotoxic Agents and Dose Response Relationships of Their Effects.
Mutation Res 123, 121-182.
12. Neuman HG. 1984a. Dosinetry and Dose- response Relationships, in:
Berlin A, Draper I!, Heriminki K, Vsainio li (£ds.)f Monitoring Hunan
Exposure to Carcinogenic and Mutagenic Agent, IARC Sci, Publ No 59,
Lyon, pp. 115-126.
13. Neunan HG. 1984b. Analysis of Hemoglobin as a Dose !tonitor for
Alkylating and Arylating Agents, Arch Toxicol, 56, 1-6.
-------
-69-
14. Stowers SJ, Anderson Mv. 1985. Formation and Persistence of
Benzo(a)pyrene Metabolite-DMA Adducts. Environ Health Perspect, 62,
31 -39.
15. Reed E, Yuspa SH, Zwelling lA, Ozols RF, Poirier MP. 1986.
Quant i tat ion of Cis-diamminedichloroplatinum II (cis platin)
-DNA-intrastrand Adducts in Testicular and Ovarian Cancer Patients
Receiving Cisplatin Chemotherapy, J Clin Invest, 77, 545-550.
16. Tannenbaura SR, Skipper PL. 1984. Biological Aspects to the Evaluation
of Risk: Dosiinetry of Carcinogens in Man. tund Appl Toxicol, 4,
S367-S370.
17. Santella RM. Application of New Techniques for Detection of
Carcinogen Adducts to Human Population Monitoring. Mutation Res
(in press).
18. Poirier MC, Beland FA. 1987. Determination of Carcinogen- induced
Macromolecular Adducts in Animals and Humans , Prog Exp Tumor Res , 31 ,
1-10.
19. Perera F, Santella R, Fischman HK, Munshi AR, Poirer 11. Brenner D,
Mehta H, Van Ryzin J. 1987a. DMA Adducts, Protein Adducts and Sister
Chromatid Exchange .in Cigarettee Smokers and Nonsmokers. J Natl
Cancer Inst, 79:449-456.
20. Perera F, Hemminki K, Young TL, Brenner D, Kelly G, Santell RM.
1987b. Detection of Polycyclic Aromatic hydrocarbon-DNA Adducts in
White Blood Cells of Foundry Workers. (Accepted).
21. Shamsuddin AKM, Sinopoli K, hemminki, Boesch RR, Harris CC. 19*55.
Detection of Benzo(a)pyrene-UNA Adducts in Human V/hite Blood Cells.
Cancer Res, 45, 66-68.
22. Haugen A, Becher G, Benestad C, Vahakangas K, Trivers GE, Newman til,
Harris CC. 1986. Determination of Polycyclic Aromatic Hydrocarbons in
the Urine, Benzo[a]pyrene Diol Epoxide-DNA Adducts in Lymphocyte DNA,
and Antibodies to the Adducts in Sera from Coke Oven Workers Exposed
to Measured Amounts of Polycyclic Aromatic Hydrocarbons in the Vtork
Atmosphere. Cancer Res 46, 4178-4183.
23. Bryant MS, Skipper PL, Tannebaum SR, Maclure M. 1987. Hemoglobin
Adducts of 4-aminobiphenyl in Smokers and Nonsmokers.
Cancer &»<* 47, 602-608.
24. Dunn BP, Stich HF. 1986. 32p Posclabeling Analysis ot" Aro«.iatic DNA
Adducts in Human Oral Mucosal Cells. Carcinogenesis 7, 111-5-1120.
25. Phillips DH, Hewer A, Grover PI. 1986. Aromatic DNA Adducts in
Human Bone Marrow and Peripheral Blood Leukocytes, Carcinogenesis 7,
2071-2075.
26. Bridges BA. 1980. An Approach to the Assessment of the Risk to Man
from DNA Damaging Agents. Arch Toxicol, Suppl 3:271-281.
-------
-70-
27. Wright AS. 1983. Itolecular Dosimetry Techniques in Human Risk
Assessment:: An Industrial Perspective, in: Hayes AW, Schnell RC,
Miya TS (Eds.). Developments in the Science and Practice of
Toxicology. Elsevier, Amsterdam, pp. 311-318.
28. Tornqvist M, -Qsterraan-Golkar S, Kautiainen A, Jensen A, Farmer
PB, Ehrenberg L. 1986. Tissue Doses of Ethylene Oxide in Cigarette
Smokers Determined from Adduct Levels in Hemoglobin,
Carcinogenesis, 7, 1519-1521.
29. Everson RB, Randerath Rll, Santella RM, Cefalo RC, Avitts TA,
Randerach R 1986. Detection of Smoking Related Covalent DNA Adduces
in Human Placenta. Science 231, 54-57.
30. Bridges BA, Butterworth BE, We ins Ce in IB. Banbury Report 1982.
Indicators of Genotoxic Exposure; Report No. 13. Cold Spring Harbor
Lab, Cold Spring Harbor, NY.
31. MAS Briefing Panel. 1983. Report on Human Effects of Hazardous
Chemical Exposures. National Acad Sci, Washington, DC.
32. Sobsel FH. 1982. The Parallelogram: An Indirect Approach for the
Assessment of Genetic Risks from Chemical Mutagens. In:-Progress in
! tutat ion Research (Bora KG, Douglas GR, Nestmann ER. eds.). Elsevier,
Amsterdam, pp. 323-327.
33. Santella Km, V/eston A, Perera F, et al. 1987. Interlaboratory
Comparison on Antibodies and Immunoassays for Benzo[a]pyrene Diol
Epoxide-1 Modified DMA. (Submitted).
-------
-71-.
NEURQTUXICOLOGY
Lawrence Reiter
INTRODUCTION
Epidemiclogical studies in Europe indicate that long-term exposure to
solvents can produce neurobehavioral disorders which, depending on the
length and severity of exposure, can range from loss* of concentration and
memory impairment to mood and personality changes to severe and apparently
irreversable dementia. Indeed, cognitive impairment appears to be an early
sign of solvent neurotoxicity. These studies have led the international
neurotoxicology community to call for improved laetiuxi.s 'for identifying and
characterizing solvent neurotoxicity both in animal models and in human
clinical populations.
NEUROBIOLJGY OF LEARNING AND MEhORY
An area of long-term research which promises to produce powerful
applications to this problem is the neurobiology of learning and memory.
The goal of this field is to understand how normal memory function is
carried out by the nervous system as well as how various neuropathological
conditions, such as Alzheimer's disease nnd Xorsakoff's syndrome, produce
cognitive dysfunction. Interest in this area of neuroscience research is
very intense. By some estimates, fully a quarter of all research in the
basic neurosciences is concerned with the neurobiology of learning and
memory. It is not surprising then that progress in this ^rea is occurring
at a very rapid rate. This paper will briefly highlight some specific
recent developments in this field which should have a major future impact
on neurotoxicological assessment.
-------
-72-
Analysis of Che neurobiology of learning has been organized around
three general areas: (1) key brain regions i.e., which brain regions are
essential for different forms of memory; (2) memory "circuits" in the
brain, i.e., the delineation of neural pathways through which sensory
information results in the production of learned behavioral, r.j^onses; and
(3) synaptic neohanisns, i.e., the nature of the synaptic changes that
occur during learning, and the biochemical and cellular processes which
underline them. The first of these areas has had, as one of its major
concerns, the problem of how to extrapolate from animal models of cognitive
dysfunction to human dementia. The latter two areas have been concerned
primarily with analyzing the animal model systems at more molecular levels.
In the past 5-7 years, dramatic discoveries have been made in all three of
these areas.
NEUFOTQXICOUX^UJAL ASSESSMENT
Attempts in the area of extrapolation have taken two forms. Oae has
oeen to develop behavioral tests in animals which are more analogous to
those which are used to assess cognitive function in humans. The other
form, and the one which we will emphasize here, has been to apply
behavioral tests to humans which are analogous to those which are well
understood, both behaviorally and neurohiologically, in animals. For
example, it has recently been shown that delayed-non-inatching-to-sample, a
task which is a sensitive indicator of memory impairment associated with
lijnbic system and frontal cortical damage in rats .-IK! ^cin-ites, is also a
sensitive indicator of dementia associated with similar neuropathology in
human clinical populations. Another example is the successful use of human
-------
-73-
eyeblink conditioning Co detect learning deficits, associated with aging
and Alzheimer's disease, which were predicted by neurobiological studies of
eyeblink conditioning in rabbits. These recent developments in basic
behavioral neuroscience establish a direct neurobehavioral link between the
experimental analysis of cognitive dysfunction in animals and its
assessment to humans. Neurotoxicological research, aimed at validating the
application of these new animal models to the problem of risk assessment
will substantially advance progress on the question of how animal studies
can be used to characterize risk to human populations, following exposure
to solvents and other environmental pollutants.
The second important development which could greatly increase the
sophistication of neurotoxicological assessment is the identification of
neural circuits subserving learning. The best example of this is the
neurobiological study of rabbit eyeblink conditioning. This Pavlovian
conditioning preparation has many advantages for neurotoxicological
assessments, including: (a) the wealth of knowledge of its behavioral
properties, which makes it possible to study anything from simple
associative reflexes to complicated cognitive-perceptual processes in a
single experimental preparation; (b) the ability to directly compare
quantitative measures of both learned and unlearned behavior, on-line and
in real time; (c) the ability to directly compare the same type of
conditioning in animals and humans; (d) the ease of arranging concurrent
electrophysiological recording from discrete brain loci (or, in the human,
brain EEC activity recorded from scalp electrodes). However, the most
important advantage offered by this recent research development is the
wealth of knowledge that we now have about its essential neural circuitry
-------
-74-
in Che brain stem and cerebellum. We also know a good deal about the
effect of pharmacological agents on this type of conditioning and this
greatly improves bur ability to integrate the various aspects of
neurotoxicological assessment. If an unknown compound produces a
behavioral effect, we have a good idea of where to look for its
neurochemical and neuroanatomical effects, and ultimately its mechanism(s)
of action. Gonversly, if a compound produces an effect on a neurochemical
or neuroanatomical system, we know what functional consequences to look for
in terms of the types of behavioral or cognitive processes which might be
impaired. Some investigators have already begun to use Pavlovian
techniques of this kind as animal mooels in the neurotoxicological
assessment process. Just this year (1987), the rabbit eyeblink preparation
has been applied to the study of dementia associated with aluminum
toxicity.
One final development which is worth mentioning is the use of the in
vitro brain slice technique to study neural plasticity.
Electrophysiological studies of hippocampal slices have uncovered a
phenomen, termed long term potentiation (LTP), which has become very
influential as an experimental model for studying the synaptic mechanisms
of learning. In LTP there is, in effect, an increase in synaptic efficacy
that occurs with repeated use. Investigations of the cellular and
biochemical mechanisms of LTP have revealed a special role of a particular
receptor type (the N-methyl-D-aspartate or NMDA receptor). Pharmacological
antagonists of the NMDA receptor may prevent the induction of LTP, and may
disrupt cognitive function in rats. What is true of drugs may also be true
of other compounds with neurotoxic potential (eg., environmental
-------
-75-
chemicals). Ic is likely chat with continued research in this area,
hippocampal slice preparations may be used as a means of screening unknown
compounds for their potential ability to produce cognitive dysfunction, and
of characterizing the neurobiological mechanisms of any neurotoxic effects
which are found.
SUMMARY
In summary, these three general areas of long-term research in
behavioral neuroscience create a framework for the analysis of
neurobehavioral function which is integrated at both a conceptual and,
perhaps more importantly, a practice level. With this framework, it is
possible' to use information from diverse scientific subdisciplines,
including cell biology, neurocheuiistry, neuroanatomy, neurophysiology, and
both animal and human psychology, in a very direct and real way to either
(a) identify the risk that compounds with neurotoxic potential may pose to
normal cognitive function or (b) characterize the risk of classes of
compounds, such as the solvents, which are known to produce memory loss,
dementia and other neurobehavioral disorders.
-------
-76-
USE OF IDNOCLDNAL ANTIBODIES IN NEUROTOXICOLOGY
Monoclonal antibodies provide another example of long-term research
which has promise for application to a wide variety of environmental
problem (See Chapter 2 for some others). This section will describe some
new applications in neurotoxicity.
Background
Exposure to a foreign substance often elicits an inmune response
characterized by production of antibodies. Antibodies are serum proteins
tnat react with antigens (antigens are foreign substances capable of
inducing antibody formation). Such antigenic substances can include
viruses, bacteria, .proteins, or even complex molecules like environmental
chemicals. Antigen-antibody reactions are highly specific, indeed, among
the most specific known to biology. It is this specificity of the
antigen/antibody complex that has been exploited by the biomedical
scientist with applications ranging from curing Polio to understanding the
molecular basis of enzyme catalysis.
Antibodies are produced in the body by B lymphocytes (B-cells), each of
which produces its own unique antibody. In theory, as many as 10 million
antibodies can be produced by a mouse in response to a single antigen.
Each antibody reacts with a unique antigenic site (termed an epitope) and
each antigen contains several epitopes. Because one B-cell can form
antibodies against only one epitope but there are many B-cells producing
antibodies -against each epitope, this is referred to as a polyclonal (many
cells) antibody.
The lymphocyte fusion technique of Kohler and Milstein, for which they
received the 1984 Nobel Prize, was designed to overcome the limitations
-------
-77-
associated with the use of polyclonal antibodies (e.g., contamination,
heterogeneity, limited supply). The antibodies produced by Kohler and
Milstein were referred to as monoclonal because they were produced by a
single (mono) B-cell line (clone). Monoclonal antibodies have several
advantages including: 1) inherent specificiy (each clone produces only one
specific antibody); 2) unlimited supply (clones produce large amounts of
antibody and can be kept indefinitely); and 3) purified antigens are not
required for the production of pure antibodies (monoclonals by definition
recognize only a single antigenic determinant).
ttonoclonals have been used to detine, localize, purify, quantify, and
modify antigens. The main distinction between the use of monoclonals, as
opposed to polyclonal antibodies, is that monoclonals confer far greater
precision and accuracy and are available as essentially immortal, off the
shelf reagents. Thus, it is now possible to define antigens with a greater
degree of certainty than ever before. This inherent trait of monoclonals
has made it far easier to identity rare antigens both in vivo and in vitro
(e.g., nervous tissue cell types and tissue typing in cell culture). One
example of the application of monoclonals that is relevant to the EPA is
the use of specific monoclonals to identify dioxin congeners in
contaminated soils. True purfication of antigens from heterogeneous
sources (e.g., serum, tissue) also is now possible with monoclonals. Thus,
rare factors or hormones, such as interferon, can now be easily obtained in
bulk pure form. Likewise, quantification of antigens in complex mixtures
is also easier to achieve with monoclonals than with polyclonal antibodies,
an example being human chorionic gonadatrophin for pregnancy tests. By
targeting specific antigens with monoclonals, modification of toxicity or
-------
-78-
disease states also may be realized. Examples are treatment of digoxin
overdose (with antibody to digoxin), and cancer therapy with anticancer
agents linked to monoclonals targeted to tumor cell antigens.
Applications of ttonoclonals to Neuroscience/neurotoxicology
The years of research on monoclonal antibodies that followed Kohler and
Milstein's original report in 1975 are now beginning to revolutionize
neurobiology by providing the tools with which to understand the complex
cellular and subcellular organization of the nervous system. Thus, the
major impact of monoclonal antibody technology on neuroscience has been the
unambiguous identification of different cell classes in the nervous system.
Indeed, monoclonals have now been produced which identify previously
unknown subsets of neurons and glia (the major cell types of nervous
tissue) which otherwise would not appear to be different using classical
techniques of light or. electron microscopy, ttonoclonals- have also proved
crucial for the identification and characterization of unique
macroroolecules, and have been even shown to reveal important differences
within the same molecule. For example, monoclonal antibodies have now been
produced that reveal phosphate-containing versus nonphosphate-
containing neurofilaments, the major structural (filament) component of all
neurons. The significance of this subtle difference, i.e., the absence or
presence of a single phosphate, is that this substitution may be related to
a variety of neurological disease states, including Alzheimer's disease,
and also may represent a general response to injury of the nervous system.
In neurotoxicology, it is known that toxicant-induced injury to the
developing or mature nervous system often is manifested by alterations in
the cytoarchitecture of specific neuroanatomical regions.
-------
-79-
Furtherraore, within an affected region, che response co injury may encompass
several cell types. Because antigens that distinguish the diverse cell
types comprising the mammalian nervous system have been revealea by
monoclonal antibodies, these same antibodies can be used to detect,
localize and characterize cellular responses to neurotoxic exposures.
This can be accomplished by a technique known as iramunohistochemistry,
where antibodies are used as specific probes for microscopically localizing
specific antigens witnin tissue obtained from toxicant-exposed animals.
Quantitative data are obtained with the same antibodies by using monoclonal-
based radioimmunoassays. Thus, through the use of monoclonal antibodies
an integrated morphological/biochemical evaluation of neurotoxicity may
eventually be achieved. The possibility also exists that the sensitivity
and specificity of monoclonal antibodies can be applied to the detection
and measurement of antigens released into the cerebrospinal fluid and
blood as a consequence of neurotoxic exposures. Theoretically, it would
then become possible to develop inexpensive monoclonal-antibody based
test kits for detecting neurotoxicity in the exposed human population.
In summary, it is clear that current advances in the neurosciences will
continue to reveal the extensive cellular and subcellular heterogeneity ot
the nervous system based on the use of monoclonal antibodies. The EPA, by
actively participating in long-range research, will benefit by having the
tools with which to assess and predict environmentally-induced
neurotoxicity.
-------
-80-
tlAGNETIC RESONANCE IMAGING
ttorrow Thompson
INTRODUCTION
A major problem in environmental health sciences is Che non-invasive
detection of small adverse effects or adverse effects at early stages.
Research applications ot magnetic resonance imaging hold promise for just
such advances. In the few years since Lauterbur's (1) paper was published,
magnetic resonance (Ilk) imagine has evolved rapidly into an accepted
clinical technique and, also, a research tool of enormous potential.
Systems with high field, superconducting magnets (1.5 co <+.7 Tesla) are
available commercially and are designed for human beings and laboratory
animals (separate systems). Sophisticated techniques that modulate the
effects of proton density, relaxation times, and motion permit the
acquisition of 3-dimensional images that optimize dirferences between
normal tissue types, define pathologic structures of areas, and allow the
measurement of blood flow or perfusion (2-5). For reasons of abundance and
signal intensity, the hydrogen nucleus (proton) is probed for the
production of practically all Ilk images. The abilities to image alternate
nuclei (e.g. 23fja) and chemically shifted nuclei (e.g. 1H in water versus
fat) have been demonstrated and show the versatility and undeveloped
potential of the technology.
Present day proton ilR images of human beings and laboratory annuals
contain superb anatomic detail that, in some applications (biologic
specimens and small animals), approaches microscopic levels. In recent
publications (6,7), images of frog eggs and plant stems have been shown
with volume elements (voxels) of 0.2 and 12.0uL, respectively. Perhaps
-------
-81-
more impressive are experiments being conducted at Duke University
in which chemically induced hepatic lesions as small as 100uL in
volume have been imaged in rats. The ability to detect such small lesions
in live animals requires long imaging sessions (as long as 6 hours),
strong magnetic fields and gradients, sophisticated pulse sequences, and
little or no relative motion. Because respiratory motion is transferred
through the diaphram to the liver, the last issue (no motion) is accomplished
by intubating the animal, using a gaseous anesthetic, and synchronizing
signal acquisition to respiratory motion (8,9).
Some of the advantages of MR imaging are common to those of other
techniques, and other advantages are unique. Similar to computerized
tomography (CT) scans, MR imaging is non-invasive and may be performed
multiple times on the same animal or patient. In toxicology experiments,
for example, the incorporation of MR imaging of a group ot animals could
provide important information concerning target organs, time to lesion
(e.g., tumor) development, and response to continued or modified treatment
(e.g., progression or regression of lesions). MR imaging uses fewer
animals per exepriment compared with conventional means tor gathering
similar information.
While imaging techniques based on ionizing radiation are well
established, rapidly produced (a distinct advantage compared to MR imaging
at its present state of development), and excellent for demonstrating some
anatomic structures or abnormalities (e.g., bone lesions containing calcium
deposits, recent hemorrhage), l^R imaging has some distinct and important
advantages. With current and anticipated magnetic fields, gradients, and
RF signals, and with the proper precautions MR imaging is considered safe
-------
-82-
for patients and technicians (10). Additionally, the MR signal, unlike the
penetrating beams of ionizing radiation, contains information in addition
to that of tissue (in this case, proton) density. The signal is also
determined by the rates at which protons relax in relationship to the
molecular lattice (T1, spin-lattice, longitudinal relaxation) and to each
other (12, spin-spin, transverse relaxation). Because these time constants
are influenced by the chemical composition of the tissue (probably by the
amount and motional freedom of water molecules), the resulting image can
permit distinction of tissues that are similar in proton density but differ
in relaxation times.
Although not a consistent finding, malignant tumors frequently have T1
and 12 relaxation tiroes greater than those of benign tumors or normal
tissue. Recent disappointments concerning the apparent inability of iiR
imaging (relaxation times) to distinguish between pathologic entities have
been expressed (11). This may be partially related to the acquisition of
the signal from tissue slices that, because of slice thickness, include
degenerative and normal areas within and adjacent to the lesion of
interest. In animal experiments at Duke University, this possibility is
being explored by excising very thin (only 1.25 mm thick) tissue slices in
rats. While signals from such thin slices are weak and imaging sessions
are relatively long, the thin sections with high resolution greatly improve
the selectivity, and, hopefully, the discriminating ability of the method.
CURRENT AND FUTURE APPLICATIONS
In clinical medicine, IiR imaging compliments and frequently exceeds the
performance of other imaging methods. IIR imaging excells in demonstrating
-------
-83-
neoplastic, demyelinating, and degenerative processes of che central
nervous system. Because of the suscetibility of the thyroid and
parathyroid glands to ionizing radiation, MR imaging is a preferred method
for examination of these tissues. Respiratory and cardiac gating have been
used to produce excellent diagnostic images of the heart, thoracic blood
vessels, and lungs. MR images of liver, kidney, reproductive organs, and
pelvis routinely demonstrate a variety of neopLastic and non-neoplastic
processes. Current and future developments will incorporate the use of
faster scanning sequences, 3-dimensional imaging, measurement of perfusion
and flow, contrast agents imaging combined with in vitro spectroscopy of
different nuclei (e.g., ^P, 13c, 23^ja> 19p)f chemical shift imaging (e.g.,
permitting separate proton images of ^H in water verses fat), and,
possibly, multinuclear imaging (e.g., ^P, 23f4). These developments, in
addition to improving the sensitivity of detecting lesions, will allow
imaging to be combined with in vivo metabolic studies that can characterize
biochemical activities in a region of interest.
In toxicologic experiments, techniques have been developed that allow
prolonged anesthetization of rats (as long as 6 hours) associated with
respiratory and cardiac scan synchronization for thoracic and abdominal
imaging. High field systems (300 IlHz, 7 Tesla) are being developed and
tested chat have a theoretical resolution of 10uti. Areas of active
research include the improvement of RE' coil designs, and the use of
stronger field gradients, surface and implaneed coils, and contrast agents.
Within a few years, increases in resolution should permit, for example, the
visualization of renal glomeruli, preneoplastic hepatocellular foci, and
nuclei in the brain. With such developments, Lauterbur's closing statement
-------
-84-
in his 1973 paper would seem remarkably prophetic, "Zeusinatographic
techniques should find many useful applications in studies of the internal
structures, states, and compositions of microscopic objects."
-------
-85-
REFERENCES
1. Lauterbur PC. Image Formation by Induced Local Interactions:
Examples Employing Nuclear Magnetic Resonance. Nature 1973;
242:190-1.
2. Morgan CJ, Hendee WR. The Evolution of Nuclear Magnetic Resonance.
In: Introduction to Magnetic Resonance Imaging. Denver: Multi-Media
Publishing, Inc. 1984:1-12.
3. Andrew ER. A Historical Review of NMR and Its Clinical Applications.
Br Med Rev 1984;40:115-9.
4. Damadian R. Tumor Detection by Nuclear Magnetic Resonance. Science
1971 ;171:1151-3.
5. Lauterbur PC. Cancer Detection by Nuclear Magnetic Resonance
Zeugmatographic Imaging. Cancer 1986;57:1899-1904.
6. Aguayo JB, Blackband SJ, Schoeniger J, Mattingly MA, Hintermann M.
Nuclear Magnetic Resonance Imaging of a Single Cell. Nature
1986 ;322:190-1.
7. Johnson GA, Brown J., Kramer PJ. Magnetic Resonance Microscopy of
Changes in Water Content in Stems of Transpiring Plants. Proc ' Natl
Acad Sci USA 1987;84:2752-5.
8. Hedlund L, Dietz J, Nassar R, Herfkens R, et al. A Ventilator for
Magnetic Resonance Imaging. Invest Radiol 1986:21:18-23.
9. Hedlund L, Johnson GA, Mills GI. Magnetic Resonance Microscopy ot the
Rat Thorax and Abdomen. Invest Radiol 1986;21:843-6.
10. Saunders RD, Smith H. Safety Aspects of MIR Clinical Imaging. Br tied
Bull 1984;40:148-54.
11. Johnston DL, Liu P, Wismer GL, Rosen BR, et al. Magnetic Resonance
Imaging: Present and Future Applications. Can Med Assoc J
1985;132:765-77.
-------
-Se-
ll 1MUNOTOXICOLOGY
Michael Luster
In a broad sense immunotoxicology can be defined as Che study of
adverse (inadvertent) effects of environmental chemicals, therapeutics or
biologicals on the immune system. The types of effects that may occur
include immunoraodulation (i.e., suppression or enhancement),
hypersensitivity (allergy) and, in rare instances, autoimmunity. A large
body of information has developed over the past 10 years that exposure to
certain chemicals or therapeutics can produce immune dysfunction and alter
host resistance in experimental animals following acute and subchronic
exposure. Examples of these are listed in the attached table. The most
extensively studied-class of environmental chemicals is the polyhalogenated
aromatic hydrocarbons (PHAs), including polychlorinated biphenyls,
polybrominated biphenyls, chlorinated dibenzofurans and the prototype of
this class, chlorinated dibenzo-p-dioxins.
Despite the species variability associated with the toxic manifestation
of these compounds, studies in laboratory animals exposed during neonatal
or adult life with PAhs and, in particular, dibenzo-p-dioxins have
indicated that the immune system is one of the most sensitive targets for
toxicity. These effects are characterized by thymic atrophy and severe and
persistent suppression of cell-mediated (T cell) immunity and share many
feanires of neonatal thymectomy. Laboratory studies have further indicated
that the target cell for immunosuppression by PhAs is the chymic epithelium
which is necessary for T cell maturation. Although only a limited number
of reports indicate immune dysfunction following human exposure to PHAs,
the effects have been found to be remarkedly similar to these which occur
-------
-87-
in animals. For example, suppression of a delayed hypersensitivity
response and increased susceptibility to respiratory infections have been
found in patients who accidentally ingested polychlorinated
biphenyl/dibenzofuran-contamined rice oil. Another example of this immune
dysregulation by PHAs has been reported in Michigan farm residents who
inadvertently ingested polybrominated biphenyls. These individuals also
demonstrated persistent suppression of cell-mediated immunity with
increased numbers of null cells, possible reflecting the presence of
immature cells. Although long-term deleterious consequences of
polybrominated biphenyls remain to be determined in humans, early data
indicate a correlation between immune alterations and increased tumor
incidence.
Thus, it appears that early laboratory studies in rodents have provided
a very accurate account of the iramunological dysfunction that, is observed
in humans following inadvertent exposure to these compounds.
-------
-88-
EXAMPLES OF IMMUNOLOGICAL ABNORMALITIES ASSOCIATED
HITH
CHQ1ICAL EXPOSURE IN RODENTS AND HUtlANS
Chemical
Class
Polyhalogenated
[27?%
Aromatic
Hydrocarbons
Heavy Metals
Aromatic Hydro-
carbons
(Solvents)
Example
TCDD
PCS
PBB
to
Lead
Caomium
Methyl ?tercury
Benzene
Toluene
Laboratory
Immune
Abnormality
Human Immune
Abnormality
N.S.
.N.S
Polycyclic
Aromatic
Hydrocarbons
Pesticides
CiffiA
BaP
MCA
Trimethyl Phospho-
rothioate
N.S.
N.S.
N.S.
N.S.
Organotins
Aromatic Amines
Oxidant Gases
(Air Pollutants)
Others
Carbofuran +
Chlordane +
DUTC +
DSTC +
Benzidine +
N02 +
S02 i-
Asbestos +
DMN +
N.S.
N.S.
N.S.
N.S.
+
N.S.
N.S.
N.S.
N.S. = Not studied; + = Positive and negative findings have been reported.
-------
-89-
HLMAN OiORIONIC GUNADOTiOPIN (HCG)
Donald Matt is on and Alan Wile ox
BACKGROUND
Public health scientists have long been concerned about the many
possible reproductive hazards of environmental pollution. One unanswered
and troubling question has been whether there are effects of toxins on the
earliest stages of pregnancy. This can include environmentally-induced
very early abort ions/fetal wastage. If there were some way to detect die
earliest stages of pregnancy, then perhaps such effects of occupational,
environmental, or drug exposures could be more easily defined and
addressed. It is known that about 15% of clinically-recognized pregnancies
end in recognized loss (spontaneous abortion). The risk of such loss has
been found to be higher in some populations with occupational,
environmental, etc., exposures, however, clinical losses don't tell the
whole story; clinically-recognized losses represent only a portion of all
pregnancy losses. There are at least twice as .many earlier losses as
recognized spontaneous abortions. Thus, a technique which could detect
pregnancy very early and define its ending precisely could help pinpoint
whether chemical or other environmental exposures might have been involved
in such ending. The application of new researches with human chorionic
gonadotrophin offer such possibilities.
tiETHOU
Determination of very early pregnancy loss requires sensitive and
specific methods for identifying pregnancy. The recent development ot
-------
-90-
antibodies to one component: of the beta subunit of HCG has vascly improved
Che capacity of HCG assays to detect early pregnancy. HCG is produced by
the conceptus starting at about the seventh day after fertilization. hCG
is quickly excreted in the mother's urine and is detectable by imtnunometric
assays. For this reason, HCG assays are the mainstay of studies of early
pregnancy. This immunoradiometric assay is reactive to the unique
carboxyterminal peptide of the HCG molecule. The assay is up to one
hundred times more, sensitive than any previously available assay. This
added sensitivity has proved to be important because up to three-quarters
of early pregnancy losses never reach a level of HCG secretion that could
have been detected by previous assays.
IMPLICATIONS
Early pregnancy loss may be one of the earliest signs of human exposure
to mutagens or other toxins that damage human reproduction. Ib should be
possible to streamline this type of study, collecting urines only on days
when early loss is most likely to be detected. This approach could be
extended to high-risk groups of women in occupational or other settings
where toxic effects on reproduction are suspected. These assays are now
able to measure HCG in urine down to the background levels that occur in
healthy non-pregnant persons. These assays are just now beginning to be
applied in epidemicLogic studies for the detection of very early pregnancy
loss. This is an exciting new applied research area in environmental
medicine which is the direct result of very basic research in reproductive
biology. This may be a model for future research ard suggests chat basic
and clinical studies are essential if we are Co make progress in
-------
-91-
underscanding human reproductive vulnerability co environmental chemical
exposure.
Further basic and applied research is needed in this area — as a high
priority — because of existing data which suggest that there are indeed
exposures which can increase the rate of clinically-recognized, spontaneous
abortion. These may include various segments of the chemical industry and
the microelectronics industry.
-------
-92-
Chapcer 5
ESTIMATION OF POPULATION RISKS
David Hoel/Michael Hogan
ANIMAL MODELS AND RISK ESTIMATION
Since relevant epidemiclogic ana clinical information are often
lacking on the potential health hazards associated with exposure to a
specified agent or chemical, laooratory animal data usually constitute
the primary basis for both qualitative and quantitative human risk
estimation. The majority of animal-based, human risk estimation is
qualitative in nature. That is, laboratory or experimental
identification of a given exposure source as a potential human
health hazard is often sufficient, in and of itself, to control or
even prevent future exposure of the general public to the agent or
chemical in question, ana no determination of the magnitude of the
risk involved in the anticipated exposure may be required (e.g.,
regulation of potentially carcinogenic food additives under the
Delaney Amendment). Nevertheless, it is the role of animal data in
the quantification of possible human health risks that is of
greater scientific interest and debate.
Animal-based, quantitative risk estimation almost always involves
ti^o separate issues or problems that must be addressed: low-dose
extrapolation, necessitated by the high uose levels typically employed
in laboratory animal studies and, of course, species extrapolation,
since the ultimate concern is with the risk posed to humans. Perhaps
-------
-93-
Che single roost itnporcant issue involved in low-dose extrapolation is
the choice of the specific mathematical model or extrapolation
procedure to be" used in determining the low-dose risk or the
acceptable exposure level for the agent under consideration. In
carcinogenesis, mathematical modeling may have progressed as far as is
possible or defensible without further insights into the mechanisms
underlying the carcinogenic process. Certainly the need for greater
emphasis on the meaningful incorporation of molecular and biochemical
data into risk models is well recognized, and it offers an important
research opportunity to those interested in the quantification of
potential human risk based on animal data. For noncarcinogenic
outcomes or endpoints there is definitely a need to reevaluate the
"safety factor" approach to risk determination, which has been the
regulatory standard since the mid-50's, and, in some -instances, to
promote the development of quantitative models similar to those used
in carcinogenesis.
Regardless of the toxicologic response of interest, however, it is
clear that, increasingly, attention will be focused on making the
selected model or extrapolation procedure more closely reflect the
underlying biological mechanises. For example, in carcinogenesis che
question of "primary" versus "secondary" or "indirect" modes of action
and their 'potential impact on the risk assessment process is sometimes
raised with those who assume the latter mechanism often arguing
against traditional low-dose extrapolation models (1).- On the other
hand, those, who out of convenience or convention, have relied on the
safety factor approach for determining permissible exposure levels for
-------
-94-
noncarcinogenic coxicancs, may need Co reconsider the biological
issues chat underlie ics use, giving particular attention to the
question of thresholds. For example, if one argues that a threshold
mechanism is present, does Che chreshold represenC a Crue
(biological),"no effect" level or merely imply a dose or exposure
level where Che observable effects are minimal? Does ic apply to the
population as a whole or vary from individual to individual? (In the
latter instance the population dose-response may be indistinguishable
from one for which no threshold exists. That is, if threshold levels
vary among individuals, then the "population" threshold level would
correspond to the threshold for the roost sensitive individual in that
population, which, for all practical purposes, might be
indistinguishable from a zero exposure level.) Another issue of
concern is whether there is a biological (as opposed to traditional)
basis for the selection of any given safety factor to be used with an
observed/estimated threshold value in generating estimates ot
acceptable human exposure levels (2).
The question of species extrapolation may well generate as much
scientific debate as the selection of the most appropriate low-dose
extrapolation procedure. Certainly, the utility of the laboratory
animal model for identifying potential human health risks is broadly
recognized within Che scientific conmunicy [e.g., see che IAKC
Preamble (3) regarding Che interpretation of experimental results with
regard co human carcinogenic risk when epideraiologic or clinical daca
are noc available]. However, chere is no universally accepCed means
of quantitatively scaling Che results observed in laboratory animals
-------
-95-
co humans, irtiac is usually done is Co assume chac animals and humans
have equivalent risks when risk is expressed in terms of the
appropriate dosage scale. Yet, human risk estimates based, e.g., on
mouse data can vary by as much as 40-fold (4) depending on whether
they are expressed in terras of an average lifetime daily tag/kg dose or
a total acculumated rag dose, standardized (divided) by body weight.
Furthermore, even though necessity may force one to rely on nothing
more than a common dosage scale as the basis for extrapolating risk
estimates across species, such an approach is only an approximate
adjustment for the variety of factors that can contribute to
interspecies differences in response (e.g., differences in lifespan,
body size, kinetic profile, genetic homogeneity, general environment,
etc.). Improvements in the quantitative extrapolation of toxicologic
responses across species will require greater emphasis on the use of
molecular and biochemical data. For example, the use ot
pharmacokinetics or molecular dosimetry, when scientifically feasible,
to estimate the "biologically effective dose" could significantly
reduce the uncertainty associated with interspecies extrapolation
of observed toxicologic responses.
HUMAN STUDIES
Mathematical dose-response models for quantitative risk estimation
have been and are increasingly being applied to epideniologic data as
well as to laboratory animal results, particularly in the area of
carcinogenesis. Some of the better known examples include Peto's
fitting of the multistage model to Doll's smoking data (6), Day and
-------
-96-
Brown's use of che sane model co assess whether a number of human
cancer risk factors such as smoking, asbescos and radiation atfected
early, late or both early and late stages of the carcinogenic process
(7), BEIR Ill's (6) use of absolute and relative risk models to
characterize the time-related distribution of site-specific tumors
among Japanese A-bomb survivors, and their use of linear,
linear-quadratic and quadratic models to predict low-dose cancer risk
associated with ionizing radiation. While the use of epideniologic
data obviously eliminates the need for species extrapolation, such
data may not be sufficiently sensitive to allow one to chose among
competing dose-response models or, in some instances, even to
determine if any health risk appears to be associated
with low or'moderate levels of exposure.
A number of procedures may be employed to increase the sensitivity
of the available epidemiologic data. For instance, initial attempts
at human risk identification and estimation could be focused on
sensitive subgroups within the general population under study, such as
the very old or young, individuals with insufficient immune response,
individuals suffering from concurrent disease or inherited
deficiencies, and individuals also exposed to other known risk factors
for the toxicologic endpoint .or health effect of interest.
Recently, a new speciality has emerged in the tield of
epideniology, which is commonly known as molecular or biochemical
epidemiology. One of the primary purposes of molecular epidemiology
is to adapt laboratory procedures for the identification and
characterization of biochemical markers to epidemiclogic field
-------
-97-
studies, so as to clarify Che nature of underlying dose-response
relationships, i.e., relationships between exposure and disease or
toxicologic effect (8). Specifically, biochemical markers may provide
quantitative evidence of generalized exposure (e.g., blood lead
levels), organ specific exposure (e.g., DNA adduce formation),
biologic change, and early or frank disease to replace the more
subjective and qualitative measures that have often been used in
epideroiologic investigations (e.g., determining exposure histories
through questionnaire data and then classifying study subjects as
being either "exposed" or "unexposed".)
While interest in and and even application of biochemical markers
is increasing rapidly, validation of their use for epidemiology is
currently a major research endeavor, and it is likely to continue to
be so in the future. [Among the issues that should be considered in
any validation exercise are the determination of marker sensitivity,
specificity, predictivity, range of normal or baseline values, and
whether the marker is reflecting current or cumulative exposures,
average or peak exposures, and cumulative or noncumulacive biological
effects (8).]
POPULATION RISKS
The last step in the quantitative risk assessment process is the
determination of the overall risk for the population of interest or,
alternatively, the selection of an acceptable exposure level for ttiat
population. Some of the uncertainties involved in using experimental
animal or epidemiologic data in hazard identification and,
-------
-98-
particularly, in dose-response modeling and low-dose risk escimacion
have already been enumerated. If a strong case can be presented for
the presence of-a threshold phenomenon and a safety factor
approach is elected, then it is important to remember that failure to
compensate adequately for the unknown, underlying threshold can result
in a proportion of the exposed population having their individual
threshold values falling below the estimated acceptable exposure level
in some instances (2).
Another significant factor that must be addressed in developing
population risk estimates is the determination or estimation of
exposure levels within the population under evaluation. There are a
number of potential problems or uncertainties typically involved in
the estimation of population exposure levels. Exposures may vary
considerably among individuals or even for a single individual across
time, so that the use of average exposure levels may not be very
representative of the exposure histories of individual population
members. While use of worst-case exposures may provide an upperbound
on the actual levels of exposure encountered, it can also lead to an
overestimate of the population's health risks and certainly engenders
a great deal of uncertainty about such estimates. The uncertainty is
compounded when average or worst-case exposure estimates are
multiplied by the estimated average risk per unit dose to obtain an
overall estimate of population risk. For example, even though
worst-case exposure estimates may overestimate the actual exposure
experience of much or possibly all of the population of interest,
"average" risk per unit dose estimates may significantly underestimate
-------
-99-
Che risks of Che most susceptible subsets of that population.
Some argue that the uncertainties involved in quantitative risk
estimation and concern for the health of the exposed population have
often led to the overuse of worst-case or upperbound assumptions in
quantitative risk estimation—assumptions that result in what they
regard as unduly conservative estimates of the population risks.
However, there are other investigators (9) who fear that national
concern about the assessment of human health risks has tended to
be focused almostly exclusively on cancer risk, and that as a result,
other (perhaps less quantifiable) forms of human disease or
dysfunction may have received insufficient attention: (See Appendix).
If this is the case, then, in any specific situation the estimated
"acceptable", "virtually safe" or "minimal risk" dose for
carcinogenesis may still entail an unreasonable level of risk of other
adverse health outcomes, even when the estimation process has been
based on conservative assumptions.
The OSTP cancer document (10) and other science policy reports
have stressed the need for qualitative and quantitative
characterization of the uncertainties of specific risk estimates
(e.g., consideration of the impact of model selection, the use of one
set of laboratory data over another, the choice ofc a particular
species as being most representative of humans, etc.). Also important
are considerations and specification of the assumptions underlying a
particular risk assessment (e.g., the construct of an estimated
lifetime average daily dose rate so that animals continuously dosed at
a constant rate throughout their lifetimes might be used to estimate
-------
-100-
the risk in humans who may have received intermittent exposures at
varying doses for only a portion of their lifespan). The continued
attention to/stress on such descriptions of specific uncertainties and
assumptions involved in any given risk assessment and to their
potential impact on the estimation of risks has been most helpful to
those charged with regulatory responsibilities for more rational and
reasonable decisions about the proper fate of the agent/chemical under
consideration.
-------
-101-
1. Heel, D.G., Haseman, J.K., Hogan, M.D., Huff, J., and McConnell,
U.E. The Impact of Toxicity on Carcinogenicity Studies:
Implications for Risk Assessment. (Submitted for Publication)
2. Portier, C., and Hogan, M. (1987). An Evaluation of the Safety
Factor Approach in Risk Assessment. In: McLachlan, J.A., Pratt,
R.M., and Markert, C.L. eds. Banbuxy Report 26: Developmental
Toxicology: Mechanisms and Risk. Cold Spring Harbor Laboratory,
New York.
3. International Agency for Research on Cancer (1985). Preamble
(p. 20). In: Volume 35: Polynuclear Aromatic Compounds, Part 4,
Bitumens, Coal-tars and Derived Products, Shale-oils and Soots.
IARC, Lyon, France.
4. Office of Technology Assessment (1981). Assessment of Technology
for Determining Cancer Risks from the Environment. Washington,
D.C.: Government Printing Office.
5. Doll, R., and Peto, R. (1978). Cigarette Smoking and Bronchial
Carcinoma: Dose and Time Relationships Among Regular Smokers and
Life-Long Non-Smokers. J. Epid. Coram. Health 32: 303-313.
6. Day, N.E., and Brown, C. C. (1980). Multistage todels and Primary
Prevention of Cancer. JNCI 64: 977-989.
7. National Academy of Sciences, Committee on the Biological Effects
of Ionizing Radiations (1980). The Eftects of Populations of
Exposure to Low Levels of Ionizing Radiation: 1980. Washington,
D. C.: National Academy Press.
8. Schulte, P.A. (1987). Methodologic Issues in the Use of Biologic
Markers in Epidemiologic Research. Am. J. Epid. 126: 1006-1016.
9. Silbergeld, E.K. (1987). Letters: Risk Assessment. Science 237:
1399.
10. U. S. Interagency Staff Group on Carcinogens (1986). Chemical
Carcinogens: A Review of the Science and Its Associted
Principles. EHP 67: 201-282.
-------
-102-
APPENUIX
Balance of Cancer and Non-Cancer Endpoints
Neil Chernoff and Stephen Nesnow
The balance of basic research on cancer and non-cancer endpoints
within any Federal organization is dependent upon a variety of tactors
such as Congressional mandates, the given organization's operational
policy, public perceptions and concern, ongoing identification of
potential data gaps and to some extent, the range of disciplines
represented by the organization's scientific staff. All of these
forces have an impact on scientific managers in their designing and
implementing a basic research program to meet their organization's
needs now and in the future.
Generally, in enacting legislative authority Congress exhorts EPA
to evaluate a broad range of potential health effects associated with
exposure to environmental chemicals and insults, barely does
legislation require specific health endpoints to be addressed over
other endpoints. It is, therefore, EPA's policy which directs
attention to specific endpoints of concern for environmental
exposures. Being a public institution, EPA is influenced by the
perceptions and concerns of the public and industrial sectors
regarding adverse health effects of chemicals. Consequently, EPA
molds its administrative and regulatory policy to balance these
concerns.
For many years the primary environmental health concern, as
perceived by the public, was the possibility of chemically-induced
cancer. The reasons for this concern include the prevalence and
-------
-103-
general irreversibilicy of hce disease, its pocencial for debilicacion
and eventual lethality, and the knowledge that many chemicals to which
there is prevalent human exposure can cause cancer in laboratory
animals. As a result of these concerns and the body of data that has
been generated over the years, the EPA's (as well as other regulatory
agencies) regulatory policy has been largely driven by cancer as the
health endpoint of greatest severity.
Over .the past several decades, however, it has become increasingly
apparent that there are nany ocher adverse health endpoints which may
be and have been induced by exposure to environmental agents. The
methyl mercury-induced epidemic of birth defects in Japan, the
incidents of delayed neuropathy in the Middle East, and the occurrence
of male sterility in workers occupationally exposed to chemicals in
the USA. all have served to alert the public that the potential risk of
exposure to environmental agents may require consideration of nany
health endpoints. The outcome of this realization has been a
broadening of the areas of concern and a simultaneous commitment of
resources to these additional research areas. Along with this
commitment there has been an increasing tendency to consider these
health endpoints during, the formulation of regulatory policy.
Toxicologists in both the public and private sectors have also
identified other organ systems and susceptible populations that: are at
potential risk from exposure Co environmental agents. These
realizations have lead to considerable support for research in other
areas such as immunotoxicology, heritable disease, and prepubertal and
geriatric populations. Additionally, scientists and the public have
-------
-104-
becorne increasingly concerned about the toxic pocenciai of lifetime
exposure co relatively small amounts of a multitude of xenobiotics.
Increasing resources have been allocated to gain a better
understanding of the potential risks from these types of exposures as
well as the most accurate ways to measure such risk.
Finally, there are two additional concerns of the general public
and the scientific community that influence allocation of resources.
The translation of laboratory data into human health risk assessments
is an extremely difficult process. While a safe environment is
desirable, the regulation and removal of chemicals based upon faulty
assumptions may lead to undesirable results (including their
substitution with potentially more hazardous compounds) entailing a
reduction in the quality of life through increased expense, disease
prevalence, and/or reductions of food and other material production.
Therefore, as the preceding Chapter indicates, considerable research
resources are now and in the future will be devoted to increasing the
scientific basis and accuracy of risk estimates. The development of a
better understanding of the basic mechanisms responsible for cancer
and non-cancer responses is ultimately the most rational way in which
to formulate regulatory policy. This obviously leads to a continuing
requirement for long-term basic research.
The second factor <<£iich influences. resources allocations concerns
the need for simpler, less expensive, and less whole-anoal-oriented
forms of testing. The nunber of agents and complex mixtures o£
potential concern is far in excess of our ability to test for toxic
potential by standard methodologies. Concerns raised by the public
-------
-105-
about the use of laboratory animals in such studies have been a
further impetus to the development of alternative test methods.
The EPA research efforts in non-cancer endpoints have greatly
increased over the last decade for the reasons listed above. Whether
this increase has led to a proper balance between cancer and
non-cancer endpoints is impossible to say, since there are so many
competing factors that go into the composition of this balance.
Certainly, a resource allocation to both cancer and non-cancer
endpoints has enabled the Agency to utilize a broad base of health
endpoints in the formulation of regulatory policy.
Long-term basic research into both cancer and non-cancer endpoints
is recognized as being essential if the Agency is to formulate a broad
regulatory policy in the most accurate manner possible. Bather than
consider cancer and non-cancer effects separately, research' in the
future will evaluate multiple toxicological responses from the same
exposure. Issues of adversity and severity ot effects over time will
be given greater attention. Efforts will be made to capture and
analyze toxicological data in a more systematic fashion. These data
will form the basis of improved structure-activity and pharmacokinetic
modelling, test battery design, and dose-response evaluation of cancer
as well as non-cancer endpoints.
-------
-106-
Chapter 6
SUMMARY
This century has seen Che emergence of abnormalities in early growth
and development, chronic degenerative diseases, and cancer as the major
causes of human morbidity and mortality in the industrialized nations of
the world. Initially, these diseases were often viewed as being the result
of heredity or the natural consequence of the aging process, I tore
recently, however, there has been a growing recognition that they
frequently have important environmental components or risk factors in their
etiology.
Many of these environmental risk factors are either produced directly
by humans or subject to their manipulation. They include chemical and
physical agents in the air, water, food supply, drugs, consumer products,
home and workplace. Uiile detailed estimates of the impact of these risk
factors are difficult to generate or verify, it has been variously
postulated that a significant number of the two million individuals who die
each year in the United States may have had their lives shortened to some
degree by the effects of air pollution; that pollutants in our drinking
water systems may play a role in the onset of cancer and heart disease,
which are the two leading causes of death in this country; and that the
collective effects of work-related disease and stress ;nay now be
approaching a level of impact more typically associated with workplace
-------
-107-
accidents. Therefore, federal health researchers and regulators are
increasingly being challenged Co identity these environmental risk factors
and reduce or eliminate their deleterious effects.
Attention has been focused on some of the technical problems that can
be encountered when one attempts to assess the true impact of environmental
exposures on human health. Often, relevant epidemiologic and clinical
information on the potential health hazard associated with exposure to a
specific chemical or physical agent will not be available. Even when such
data is available, however, it may not be sufficiently sensitive or
specific to allow an investigator to choose among competing mathematical
models that attempt to characterize the unknown, underlying relationship
between exposure and dose. In some instances, the available human data may
not even permit one to determine if any health risk appears to be
associated with low or moderate levels of exposure. As a result,
laboratory animal data will often constitute the primary basis for both
qualitative (i.e., hazard identification) and quantitative human risk
estimation.
Because of the high (often maximally tolerated) doses typically
employed in laboratory animal screening studies, quantitative risk
estimation based on laboratory data involves two separate issues that oust
be addressed: low-dose extrapolation and species extrapolation. In some
instances (e.g., when the agent of concern is a carcinogen or mutagen)
mathematical modeling will be employed to generate low-dose risk estimates,
and choice of a particular model may have a significant impact on the
-------
-108-
magnicude of Che estimated risk. In other cases a threshold phenomenon may
be assumed and a safety factor approach used to determine acceptable
exposure levels. This approach also clearly suffers from a number of
methodological uncertainties and problems.
Ideally, the problem of species extrapolation should be addressed by
taking into consideration all of the various species-specific factors that
could contribute to interspecies differences in response to the exposure of
interest. Instead, the conventional approach to this issue is to assume
that humans and the test animal in question will have equivalent responses
when comparisons are maae on an appropriately-chosen dosage scale.
unfortunately, choice of the most appropriate-dosage scale cannot always be
justified on biological grounds, and significant differences can result in
the projected human risk estimate depending on the decision reached. AS
our reliance on quantitative risk estimation/assessment continues to
increase, more and more importance is being attached to the need to
characterize the uncertainties associated with these risk assessments and
to reduce these uncertainties by improving the biological basis upon which
die risk assessment process is based.
In addition to technical problems related to the risk assessment
process itself, which have complicated and on occasion frustrated our
efforts to evaluate adequately the potential risks posed by various
environmental hazards, there are also a number of additional factors that
have hampered our attempts to reduce the impact of environmentally-related
disease. Among these are the lack of substantive toxicologic information
-------
-109-
on the majority of commercial chemicals that have been introduced into the
human environment, the insufficient and sometimes inappropriate training of
our nation's physicians with respect to environmental issues, and the
inadequate surveillance of populations exposed or potentially exposed to
environmental hazards.
Priority must be given to research in a number of important areas if we
are to resolve these problems and advance our understanding of the role of
environmental factors in human health and diseases. For example, more
emphasis needs to be given to the development and refinement of procedures
(particularly non-invasive procedures) for measuring low levels of human
exposure to toxic environmental agents. Similarly, we need to develop a
better understanding of the biological mechanisms that underlie
environmentally-related health effects to improve both the quantitative
assessment of human health risks, and the primary/secondary prevention of
environmentally-related diseases. A number of examples of long-term, basic
research activities in these areas that either have or may utlimately have
direct application to the types of environmental health problems that EPA
and other regulatory agencies must address on an ongoing basis are cited in
this document.
Comparison of patterns of proto-oncogene (i.e., cellular genes
expressed during normal growth and developnent processes) activation in
spontaneous and chemically-induced rodent tumors may provide insight into
the mechanisms of tumor formation at the molecular level. In addition,
some of the uncertainty involved in species-to-species extrapolation of
-------
-110-
carcinogenic risk estimates may eventually be removed by interspecies
comparisons of oncogene activation and expression.
Recent advances in biochanistry and molecular biology have led to the
development of highly sensitive techniques which may allow the
quantification of the internal dose of carcinogens or in some cases the
biologically effective dose in target tissues. This ability to express
external exposure or administered dose levels on a more
biologically-relevant basis should eventually lead to a clearer
understanding of the relationships between exposure and disease or
toxicologic effect for many health hazards in the human environment.
Recognition of the potential usefulness of these biochemical markers has
led to the emergence of a new field of epidemiology, known as molecular or
biochemical epidemiology, that has as one of its major goals the adaptation
of these laboratory procedures into epidemiologic field studies.
In the fields of neurotoxicology and immunotoxicology new methodologies
promise to enable toxicologists to greatly improve our ability to assess
both central nervous and immune system deficits. The utilization of novel
techniques in molecular biology (e.g., monoclonal antibodies to specific
critical chemical components of these systems) promises to allow improved
evaluations of potential disfunctions.
In the area of human reproduction one of the most important questions
involves the potential of environmental agents to affect pre-impIantation
loss. Researchers have recently identified an antibody to a subunit ot the
hormone human chorionic gonadotropin. This advance enables the
-------
-111-
identification of spontaneous abortions ac an earlier stage and with
greater accuracy than was previously possible and may significantly
improve our monitoring capabilities.
In addition to identifying specific examples of long-term research
activities that either are generating or nay generate results directly
applicable to the environmental health issues that EPA must address from a
regulatory viewpoint, this document also attempts to describe the
relationship between long-term -and short-term (or immediate)
"problem-solving" research and to put it in perspective. For example, it
is noted that the general philosophy underlying basic health research is
that understanding more about the biologic mechanisms by which
environmental hazards such as toxic chemicals induce adverse effects will
lead, ultimately, to earlier detection of such effects, more sensitive
analytical methods for fully characterizing their potential impact on human
health, and a better understanding of how to eliminate or, at least, reduce
that impact. The distinguishing characteristic of this basic research is
that it typically addresses "generic" scientific issues and is not focused
on a specific problem or immediate concern. Furthermore, it must usually
be supported for a period of several years before it produces results that
may have a direct application to regulatory needs or problems.
The environmental health problem with the toxic metal leart is used to
illustrate the necessity of and role for long-term research activities in
the development of a sound, scientific foundation necessary for
constructive actions dealing with public health problems. While lead
-------
-112-
an important public health concern, ongoing, long-term basic research has
only recently given us the technical tools to detect some of the more
subtle yet extremely important effects of low-level lead exposure.
Even when research is more focused on a specific issue or health
concern, it may need to be sustained for a considerable length ot tine
before any practical results or applications can be produced. Research is
most often sequential with each new phase of the overall effort dependent
on the results from the preceding phase(s). Alternatively, even if the
research is focused and the required course of action clearly delineated
before any effort is expended, a considerable investment of time and effort
may be required before the project is completed. Certainly, this is the
case with prospective cohort studies in epidemiology and to a lesser extent
with laboratory-based, lifetime carcinogenidity screening experiments.
It seems clear, theretore, that while many of the health effects (or
possible health eftects) issues that confront EPA require an expeditious if
not -immediate response, the most appropriate and in many cases the only
approach to formulating such responses will be to draw on the experience
and insights gained from long-term research. This certainly has been the
experience in dealing with most environmental crises to date. The only
approach that will enable us to engage in sucn long-term research is to
provide stable, consistent support for such a program. With continued
support and in-house expertise EPA can directly address applied research
issues with which it is particularly concerned and effectively apply boch
its own long-terra findings and those of other public and private
institutions to the solution of critical environmental health problems.
------- |