vvEPA
           United States
           Environmental Protection
           Agency
             Office of Research and
             Development
             Washington DC 20460
EPA/600/9-91/009
April 1991
ORD
Health  Biomarkers
Program

Research  Strategy
Document
                                      •ALLAS,
                  JU

-------
                                          EPA/600/9-91/009
                                          April 1991
                   ORD
     HEALTH BIOMARKERS
              PROGRAM

Research Strategy Document
                     by
               John R. Fowle HI
         Health Effects Research Laboratory
            Office of Health Research
         Office of Research and Development
        U.S. Environmental Protection Agency
    Research Triangle Park, North Carolina 27711
                    and
               Elizabeth Collins
             Eastern Research Group
               6 Whittemore St.
              Arlington, MA 02174
        Office of Research and Development
       U.S. Environmental Protection Agency
             Washington, DC 20460
                                           Printed on Recycled Paper

-------
Preface
   Targeted environmental health research, including both basic and applied research, is needed if the
Environmental Protection Agency (EPA) is to evaluate accurately the relationship between environmental
exposures and human health risks.  Over the past few years, Office of Research and Development (ORD)
scientists have worked to define how "Biomarker" measurements made on human tissues, fluids and excreta,
can be used to improve the assessment of environmental health risks. This document is the product of their
efforts.

   Initially, ORD co-sponsored a project with the National Academy of Sciences to develop biomarker
concepts and definitions, and to identify which markers are available for use now and which need additional
research before they can be applied to human  populations. Based on the results, ORD developed a draft
research strategy that was reviewed by the Environmental Health committee of EPA's Science Advisory
Board. Their suggestions helped to make this a better document.

   The import of the research described in  this document goes beyond ORD. The terms and concepts were
recommended by the International Programme for Chemical Safety (IPCS) as a model for using biomarker
data to evaluate health risks.
                                                    Ken Sexton, Sc.D.
                                                        Director
                                                Office of Health Research

-------
Disclaimer
This document has been subjected to the Agency's peer and
administrative review, and it has been approved for publica-
tion as an EPA document. Mention of trade names or com-
mercial  products  does  not  constitute  endorsement or
recommendation for use.
Acknowledgments
Contributors to this document include  Ralph  Cooper,
Richard Everson, Howard Kehrl, Hillel Koren,  Stephen
Nesnow, James O'Callaghan, Jane Ellen Simmons, and
Ralph Smialowicz, U.S. EPA Office of Health Research;
Gerald Akland, Gerald Blancato, and Charles Nauman,
U.S. EPA Office of Monitoring, Modeling Systems and
Quality Assurance; and Lorenz Rhomberg and David Reese,
U.S. EPA Office of Health and Environmental Assessment.
Life Systems and Eastern Research Group also contributed
to the writing, editing, and production of this document.
                                                                                                ill

-------
List of Acronyms
ATSDR    Agency for Toxic Substances and
           Disease Registry
CAA       Clean Air Act
CDC       Centers for Disease Control
CFR       Code of Federal Regulations
CERCLA   Comprehensive Environmental Response,
           Compensation and Liability Act
CSF       Cerebrospinal fluid
CWA      Clean Water Act
DNA       Deoxyribonucleic acid
EPA       Environmental Protection Agency
ETS       Environmental tobacco smoke
FIFRA     Federal Insecticide, Fungicide and
           Rodenticide Act
FR         Federal Register
HERL      Health Effects Research Laboratory
HIS        Health Interview Surveys
HIV       Human immunodeficiency virus
LOAEL    Lowest-Observed-Adverse-Effect Level
LH        Luteinizing hormone
MDSD     Monitoring and Data Support Division
NAAQS    National Ambient Air Quality Standards
NAL       Nasal lavage
NAS       National Academy of Sciences
NHANES   National Health and Nutrition Examination
           Surveys
NIEHS     National Institute of Environmental
           Health Sciences
NIOSH     National Institute for Occupational
           Safety and Health
NOX       Nitrogen oxide
NOAEL    No-Observed-Adverse-Effect Level
ODW      Office of Drinking Water
OHEA     Office of Health and Environmental
           Assessment
OHR       Office of Health Research
OMMSQA  Office of Modeling, Monitoring Systems and
           Quality Assurance
OPP       Office of Pesticide Programs
ORD       Office of Research and Development
OTS       Office of Toxic Substances
OWEP     Office of Water Enforcement and Permits
PCBs      Polychlorinated biphenyls
PHS       Public Health Service
RCRA     Resource Conservation and Recovery Act
RIHRA     Research to Improve Health Risk Assessment
SCE       Sister Chromatid Exchange
SDWA     Safe Drinking Water Act
TEAM     Total Exposure Assessment Methodology
TSCA     Toxic Substances Control Act
IV

-------
Contents
                                                                                     Page
SECTION ONE — INTRODUCTION                                                         1
   1.1  Overview                                                                          1
   1.2  Organization of This Document                                                         2
   1.3  Cascade of Events Between Exposure and Disease                                            2
   1.4  Risk Assessment Framework — The Scientific Rationale for Biomarker Research                    3
   1.5  Key Research Issues                                                                  4
   1.6  Specific Programmatic Needs — The Regulatory Rationale for Biomarker Research                  5
   1.7  Application of Biomarkers                                                             7

SECTION TWO — BIOMARKER RESEARCH STRATEGY                                      9
   2.1  Roles of Primary ORD Offices                                                           9
   2.2  Establishing a Biomarker Research Strategy — Criteria and Priorities                             9
   2.3  Program Issues                                                                     12

SECTION THREE — APPLICATION OF CURRENT BIOMARKERS                             15
   3.1  Exposure Biomarkers                                                                15
   3.2  Effects Biomarkers                                                                  15

SECTION FOUR — FUTURE RESEARCH DIRECTIONS                                       17
   4.1  Exposure Biomarkers                                                                17
   4.2  Effects Biomarkers                                                                  18

REFERENCES                                                                          23

-------
Section One
Introduction
1.1    OVERVIEW

The  U.S.  Environmental Protection  Agency  (EPA)  is
responsible for protecting public health from the adverse ef-
fects of exposures to environmental agents. In developing
regulations to protect public health, EPA currently relies on
quantitative assessments of the health risks associated with
different pollutants. However, the current lack of under-
standing about the underlying biological, chemical, and
physical processes that determine exposures and effects hob-
bles EPA's ability to make these assessments.  Often, quan-
titative risk assessment is entirely precluded because of the
paucity of appropriate data. In other cases, risk assessment
is possible, but numerous assumptions representing "fall
back" or "default" positions must be applied due to critical
data gaps.  Application  of these assumptions fosters enough
uncertainty regarding the interpretation of the available data
that diametrically opposed risk assessments can be made on
the basis of the same information by groups employing dif-
ferent default positions.

A stronger scientific basis for risk assessment  is the key to
EPA's making  more informed decisions  for  safeguarding
public health.  The current revolution in molecular science
has acted as a driving force in the development of this fun-
damental framework. Over the next decade, an  avalanche of
biologic measures bearing on chemical exposure and human
disease  is likely to flow from these scientific advances.
Chemical engineers, for example, are developing ever-more-
sensitive techniques for measuring chemical substances and
are coupling these techniques to sophisticated  instrumenta-
tion capable of detecting minuscule amounts of substances,
even in chemical mixtures such  as blood and other body
fluids.  With such  techniques, scientists can  explore life
processes at the most elemental levels — for example, how
proteins control the formation of other components of cells.

In addition, major advances  in toxicology and clinical/oc-
cupational medicine are fostering a better understanding of
the biological, chemical, and physical processes that are im-
portant in maintaining health  and in responding to chemical
and other challenges. One of the most important of these ad-
vances is an increased understanding of the human genome,
which molecular geneticists and other scientists are well on
their way to mapping within the next 15 years.  The fruits of
this achievement could be staggering: once scientists can
pinpoint the genes that cause various diseases such as some
forms of cancer, cystic fibrosis, and Down's syndrome, they
may be able to alter the course of the disease or to eliminate
it entirely.

This scientific revolution  is fueling the identification  of
biological markers, which are thought to offer great promise
in reducing the uncertainties associated with risk assessment.
"Biomarkers," as defined by  the National Academy  of
Sciences (NAS,  1989a), are indicators of variation in cel-
lular or physiological components or processes,  structures,
or functions that are measurable in a biological system or
sample.  For example, a high blood lead level is  a good in-
dicator of human exposure to lead and can be measured by
laboratory sampling. Ultimately, biomarkers will be used as
essential tools in monitoring and controlling environmental
exposure to a broad range of contaminants.

This document  outlines  the  framework  for developing,
validating,  and  applying  biomarkers  that ORD  uses  to
facilitate planning, budget  allocations, and collaboration in
biomarker research. As a framework for biomarker efforts
in ORD, this document is "a plan for a plan." Within the
framework, ORD evaluates EPA's  regulatory needs, its own
capabilities, and the   state-of-the-science.    During this
evaluation, ORD considers biomarker techniques as tools in
understanding life processes; thus,  rather than exploring
biomarkers  as ends  in  themselves,  ORD  incorporates
biomarker research efforts into ongoing and future research
programs.   Note that the research efforts described in this
strategy document cover only biomarkers for human health
effects, though ecological biomarkers can also be identified.
The document also defines terms and concepts used in this
research in an effort to standardize  their use across ORD
laboratories, in keeping with the Total Quality Management
philosophy adopted by EPA.

The ORD biomarker research program will position EPA at
the forefront of biomarker research, thus allowing the Agen-
cy to take full advantage of the new technology  as it evol-
ves. As part of this positioning, ORD must examine many
scientific and ethical issues raised by  biomarker research.
For example, how will  ORD recruit and employ molecular
biologists, as well as retrain chemists,  biologists,  and en-
gineers? What priorities will the Office set for work both in
the near and longer term? How will advances in science affect
the traditional roles of ORD offices (distinctions will tend to
blur as work takes on a greater molecular emphasis)?  What
implications do  scientific  advances  over the next decade
hold for the way ORD does business?

-------
Also, the increasing understanding of the genome raises a
number of difficult ethical issues.   People  with certain
genetic traits may have difficulty obtaining jobs or getting
insurance because of their susceptibility to sickness or death.
In addition, individuals carrying genes for diseases for which
no successful intervention procedure exists may suffer great
mental anguish.  (For more information on these concerns,
see Section 2.3.4.)

Besides this document, other strategy documents have been
or are being written.  They discuss and document specific
ORD plans for developing, validating,  and using human
biomarkers within ongoing ORD program areas. These in-
clude:

•  Cell Receptor-Xenobiotic Complexes as Exposure
   Biomarkers

•  Decision Model for the Development of Biomarkers of
   Exposure

•  Strategy for OMMSQA Biomarkers Research Program

•  Human Exposure Research Program: Strategy and Plan

•  Protein Adduct-Forming Chemicals for Exposure
   Monitoring: Literature Summary and Recommendations

•  Protein Adduct-Forming Chemicals for Exposure
   Monitoring: Chemicals Selected for Further Study

•  Selection of Adduct-Forming Chemicals for Human
   Monitoring Studies

•  Role of Pharmacokinetics and Biomarkers in Risk
   Assessment Implementation Plan: Part I. Exposure
   Assessment

•  Research to Improve Health Risk Assessments (RIHRA)
   Program (Topic III. Physiologically Based
   Pharmacokinetic Models)
1.2    ORGANIZATION OF THIS DOCUMENT
The remainder of Section One defines terms used by ORD to
describe biomarker efforts and outlines the framework  for
risk assessment as well as project office  biomarker needs.
Section Two discusses ORD's strategy for the biomarker re-
search program, outlining the criteria established for select-
ing projects and highlighting which research areas have been
given first priority for biomarker development One of  the
program's selection criteria is to identify and take advantage
of current opportunities  to  develop, validate, and apply
biomarkers in human populations.  Section Three describes
the program's philosophy for using currently  available  ex-
posure and effects biomarkers in  near-term human studies.
Section Four presents the program's directions for longer-
term research to identify and develop new biomarkers.
1.3   CASCADE OF EVENTS BETWEEN
      EXPOSURE AND DISEASE

Understanding the  human  health risk associated with en-
vironmental  exposures involves  defining the cascade of
events between exposure to an environmental agent and
resulting health effect. To cause a health effect, a pollutant
and/or its metabolite(s) must be absorbed into the body,
reach a target organ, and result in a biological change.  For
many environmental  pollutants,  little is known about this
flow  of  events between  exposure and  health  effect.
Biomarkers have the potential for shedding light on the fac-
tors influencing how much of a pollutant is absorbed and
reaches a target organ and the resulting biological effect of
that target dose.

Figure 1 shows the  cascade of events between exposure and
disease.    The  squares  represent potentially measurable
events that may be  defined by biomarkers. Biomarkers can
be structurally classified by the point in the cascade at which
a measurement is taken (vertical arrows, Fig. 1):

Exposure -  indicators of absorbed or target dose.  An ab-
sorbed pollutant, its metabolite(s), or products resulting from
interaction with endogenous substances measured in a body
tissue or fluid  is  a biomarker  of exposure.    These
biomarkers, such as blood or urinary lead, provide informa-
tion concerning the chemicals to which human populations
are exposed. Most exposure biomarkers are indicators of ab-
sorbed dose.

Effect - indicators  of biological effect.   These biomarkers
provide information concerning the likely health outcomes
associated with different target doses of environmental pol-
lutants or metabolite(s).

Susceptibility - indicators  of whether the individual or the
subpopulation is more or less sensitive to exposure to a par-
ticular environmental pollutant.   Sensitive subpopulations,
for example, can be pinpointed by an increased absorption
rate or a more severe biologic response.

These and other important terms  in biomarker research are
defined in Table 1-1.

Gold Standard
Ideally,  sufficient  information would be available on  the
links between events in the cascade so that  any biomarker
could be quantitatively related to either end of the spectrum.
In other words, given enough information on the underlying
processes, an "exposure" biomarker such as blood lead level
could be used to quantitatively estimate both the exposure
that must have occurred in the past and the  likelihood of a
specific health outcome in the future. For most environmen-
tal pollutants, the available data are and  will be insufficient
to allow such a broad application of a biomarker. However,
ORD's research programs  in pharmacokinetics  (the effects
of the duration, magnitude, and frequency of exposure on the

-------

















Susceptibility


Exposure

I
1




Absorbed
Dose







Delivered
or
Target Dose






Biolo
Eff

Exposure














Biomarkers






gical
ect







Health
Effect
t
1
Effect









Figure 1.   Utility of biomarkers as measures of exposure, susceptibility, and disease. The center row of boxes identifies key
           parameters in risk assessment. The upper and lower boxes identify how biomarkers can be used to provide data for
           assessing risk.
dose to the target organ) and dose-response (estimation of
the incidence of a specific health effect in human subpopula-
tions) are designed to help provide the necessary data on the
middle steps in the cascade for priority biomarkers. These
efforts will probably result in the development of an ap-
propriate suite  of biomarkers for evaluating the  cascade
from exposure to effect.

For general use,  the Agency has narrowed the NAS defini-
tion of biomarkers (see p. 2) as follows:

   A biomarker must be taken from material from an intact
   organism or involve  a functional evaluation of the or-
   ganism itself, and the measurement must serve as an in-
   dicator of susceptibility, exposure, and/or effect.

ORD's health biomarker program further limits the defini-
tion of biomarker to "measures of human susceptibility, ex-
posure, and/or effect"   (A separate document, Ecological
Biomarker Strategy for Research and Development, defines
the terms and concepts ORD will use in developing, validat-
ing, and applying biomarkers for ecological effects evalua-
tion (U.S. EPA, 1990g).)
1.4    RISK ASSESSMENT FRAMEWORK —
       THE SCIENTIFIC RATIONALE FOR
       BIOMARKER RESEARCH
Estimating the risk associated with exposure to a given pol-
lutant involves evaluating, for that chemical  or chemical
class, the likelihood  of resulting health outcomes and
themagnitude of such effects (i.e., does exposure result in a
cascade of events leading  to health effects?). To develop
these estimations, EPA has adopted a formal risk assessment
TABLE 1-1.  Key Structural Definitions in Biomarker
             Research
Term           Definition
Concentration   Amount of material (contaminant) per unit of
                volume or mass in an environmental sample
*Exposure       Contact between an environmental
                contaminant and a living organism(s)
                (e.g., human, indicator organism, ecosystem);
                this may result from a single challenge or
                from contact at a given concentration over
                time
* Absorbed dose  Amount of material that crosses one or
(Internal dose)   more of the body's boundaries; often
                absorbed dose is best measured by the area
                under the curve of intake versus time
*Delivered dose
(Target or
biologically
effective dose)

Body burden
                Amount of the absorbed dose and/or its
                metabolites that reaches the target
                (e.g., tissue, cell); often, delivered dose is
                best measured by the area under the curve
                of tissue concentration versus time
                Amount and distribution of material and/or
                its metabolites in the body
Biological effect A measurable response in a molecule, cell,
                tissue or fluid
Health effect     A biological effect that causes dysfunction,
                injury, illness, or death
Susceptibility    Increased or decreased resistance to
                absorption of and/or effect from chemical
                substances due to genetic predisposition,
                environmental or lifestyle factors
*Time component is not always critical.

-------
process, the steps of which mirror this flow of events.  This
process is based on recommendations from the National Re-
search Council of the National Academy of Sciences (NAS,
1983).  Given the inadequacy of current databases, these
steps  also  represent critical  questions  in  environmental
health research:

•  Exposure assessment - What exposures occur or are
   anticipated to occur in human populations?

•  Hazard identification - Is the agent capable of causing
   an adverse effect?

•  Dose-response assessment - What is the quantitative
   relationship between dose and effect in humans?

•  Risk characterization (based on syntheses of
   dose-response and exposure assessments) -What is the
   estimated health risk (occurrence/magnitude) from the
   anticipated human exposures?

The use of biomarkers  can improve the risk assessment
process.  First, the development and analysis of biomarkers
provides data needed to  construct the scientific framework
for risk assessment  Using biomarkers to monitor humans
for evidence of environmental exposure  and compromised
health will  highlight the factors involved in how chemical
substances  enter the body, reach target sites,  and elicit
biologic responses.  Second, biomarker analysis can make
the risk assessment process more realistic and more cost-ef-
fective.  For example, the level of a chemical or its metabo-
lite(s) in the blood is irrefutable evidence of exposure; and,
given data on  the  kinetic  behavior of the chemical in the
body, risk assessors can  accurately estimate the human ex-
posure to the chemical from that biomarker. This estimate
can be made without relying on traditional assessment  tech-
niques, including the use of assumptions based on data  from
laboratory test species.

When enough is known about the cascade of events follow-
ing exposure to a pollutant, a biomarker can be used to es-
timate both level of exposure and probable outcome. Blood
lead and carboxyhemoglobin levels can currently be used in
this way to link exposure to health effect. From a practical
perspective, however, such relationships  will be known for
very few biomarkers.  Thus, combinations of biomarkers
need to be developed for  use together in evaluating risk.

Trends in  biomarkers  can  also suggest changes in ex-
posure level.  These measurements,  when compared with
the level  of the  biomarker  in  control  populations,  can
provide strong evidence  of environmental contamination or
of exposures in an occupational setting.  In addition! when
these biomarker trends can be  associated with a specific en-
vironmental change, risk assessors and risk managers can
use these data to catalyze  changes in environmental policy
or regulations.
1.5    KEY RESEARCH ISSUES

The critical research questions for improving human health
risk assessments  (see Section 1.4) underlie  the  research
needs of all EPA regulatory program areas.  These questions
can be further subdivided into six long-standing research is-
sues  that cut across environmental media, scientific  dis-
ciplines, pollutant  classes,  and  regulatory  programs.
Handling these issues will require both basic and applied re-
search as well as a substantial commitment of time, effort,
and resources. Biomarkers, as indicated below,  can be im-
portant tools  in  facilitating research  addressing these dif-
ferent issues.

•  Exposure Assessment Research. Populations are
   exposed to a wide spectrum of chemicals. These
   environmental contaminants may enter the body through
   inhalation, ingestion, and/or dermal absorption.
   Exposure may be episodic (e.g., a chemical spill),
   intermittent, short-term chronic (e.g., rush hour traffic),
   or long-term (e.g., contaminated drinking water). Risk
   assessors need better information on actual human
   exposures, including magnitude, duration, frequency,
   and route of exposure. Researchers need to improve
   sampling and analytical methods to design and carry out
   population-based exposure measurement programs and
   to develop and validate appropriate exposure models.
   Biomarkers are needed that provide a memory of
   exposure from multiple routes or times.

•  Hazard Identification Research. To identify potential
   health hazards, EPA needs validated, short-term test
   methods (both in vitro and in vivo) for screening new and
   existing environmental agents.  Such tests will allow risk
   assessors to determine in a timely manner whether an
   environmental agent causes an adverse health outcome.
   Agents of current concern include biotechnology
   products released into the environment, alternative fuels
   (e.g., methanol), and increased ultraviolet radiation due
   to stratospheric ozone depletion. Scientists need effect
   biomarkers to identify exposures and possible health
   outcomes.

•  Dose-Response Research. Because of the limitations of
   current databases, risk assessors must usually estimate
   human response to a pollutant from data gathered
   through exposing laboratory animals to high pollutant
   exposures. The resulting effects must then be
   extrapolated to a different species and to different
   exposure conditions, because humans are usually
   exposed to low levels of pollutants and experience
   chronic rather than acute effects. Large uncertainties are
   associated with this extrapolation process. Improving
   the accuracy of risk assessments requires a better
   understanding of the physiologic and biochemical
   mechanisms of toxicity, including compensatory
   processes. In this area, researchers need homologous

-------
markers that can be used in different species to link
exposure to disease.

Risk Characterization Research.

Cancer and Noncancer Health Effects.  A variety of
health  outcomes may  arise from an environmental ex-
posure. In addition to carcinogenicity, adverse health
consequences may include various systemic (e.g., pul-
monary, cardiovascular) effects. Noncancer health risk
assessment involves determining which health outcomes
are important, for which  environmental contaminants,
and under what conditions.  The Agency currently relies
on a "safe dose" approach to noncancer effects (e.g., ref-
erence doses, maximum concentration limits).  However,
developing an appropriate  methodology for quantitative
risk assessment is critical for improving risk management
decisions in this area.  To do so, scientists must establish
criteria for defining the adversity and severity of effects
.and take into account the different effects induced in mul-
tiple organ systems.  Batteries of markers that could be
studied in parallel at different points in time following
a challenge would be key tools in these efforts.

Chemical Mixtures.  Humans are usually exposed to en-
vironmental contaminants in mixtures rather than singly.
Examples   include  emissions  from  hazardous  and
municipal  waste incinerators, urban air pollution, and
drinking water contaminants. Historically, in vitro and in
vivo tests have been developed to establish which com-
ponents of a mixture are most hazardous and to compare
the health  risks of different mixtures. Now, researchers
must determine whether the health risks of a pollutant
mixture can be reasonably assumed  to be  equivalent to
the sum of the risks associated with the specific mixture
components.  In this area, markers could be used to
determine the  relative importance of exposure with
respect to different pathways (e.g., air, water, soil) and
disease outcome. Markers can also be used to deter-
mine if exposure to multiple chemicals results in addi-
tive effects.

Evaluation of Human Populations.   Obtaining exposure,
dose, and effects data in human populations is key to as-
sessing the status of public health, identifying potential
problems,  and evaluating the  efficacy of risk reduction
measures.  This process is also important for identifying
and safeguarding population subgroups that may be at
elevated risk because of either increased susceptibility or
higher  exposures. Human data are also needed to assess
the comparability of  effects  observed in animals and
people. Researchers  need noninvasive,  inexpensive,
informative markers  for  gathering exposure and ef-
fects data  from human subjects.
1.6    SPECIFIC PROGRAMMATIC NEEDS —
       THE REGULATORY RATIONALE
       FOR BIOMARKER RESEARCH

EPA's authority to conduct environmental health research is
derived initially from the major federal  laws mandating
broad programs to protect public health and me environ-
ment.  Each  of these laws,  including the Clean Air Act
(CAA); Federal Insecticide, Fungicide and Rodenticide Act
(FTFRA); Toxic Substances  Control Act (TSCA); Clean
Water Act  (CWA); Safe  Drinking  Water Act  (SDWA);
Resource Conservation and Recovery Act (RCRA); and the
Comprehensive Environmental  Response, Compensation
and Liability Act (CERCLA), require  that EPA develop
regulatory programs to protect public health.  In performing
this task, EPA must analyze the  lexicological  effects of
specific pollutants in exposure media (e.g., air, water, soil,
food,  sludge) as well as develop health risk assessments.
Most  assessments to  date  are based on data from animal
toxicology studies or human  epidemiologic studies and es-
timates of environmental exposure.  Because they are tools
for  pinpointing  exposure,   susceptibility,  and   effect,
biomarkers are used to support the chemical-specific needs
of the program offices as well  as  the  longer-term estab-
lishment  of a stronger  scientific basis on which to build
health risk assessments.

This section highlights current uses of biomarkers in support
of the EPA program offices as well as potential  future ap-
plications.

1.6.1  Clean Air Act
Current uses of biomarkers to support CAA:  Low levels of
alpha-1-trypsin in the blood may indicate incipient or exist-
ing emphysema. Researchers also use blood lead and car-
boxyhemoglobin to indicate  exposure to lead and carbon
monoxide, respectively, as well as decrements in lung func-
tion to identify exposure to ozone.

Potential uses of biomarkers:  Pollutant-specific markers
could be used for monitoring air pollution trends on a local
and national level and as early warning signs for asbestos
toxicity, emphysema,  and lung cancer. Biomarkers can also
serve  as  tools  for epidemiological  research — especially
biomarkers for developmental neurotoxicity, immunotoxicity,
and reproductive  toxicity, which  can be used to assess
physiologic responses of the lung and to provide physiologi-
cal measures of effect  for  victims of the  sick-building
syndrome (due to polluted indoor air). Refining information
on cotinine in urine as a biomarker for exposure to environ-
mental tobacco smoke (ETS), for example, is much needed.

-------
1.6.2  Federal Insecticide, Fungicide, and
       Rodenticide Act
Current uses of biomarkers to support FIFRA: The Office of
Pesticide Programs (OPP) is using biomarkers now in risk
assessments  on pre-registered pesticides.  The office  uses
"classical" biomarkers such as measurement of cholinesterase
inhibition, liver enzyme changes, DNA adducts, and repair
synthesis.

Potential  uses of  biomarkers:   Markers for exposure,
reproductive function, genotoxicity, cancer, and pulmonary
effects could be used, as well as markers for outcomes as-
sociated with specific classes of pesticides  that can be used
to evaluate whether humans respond to various levels of pes-
ticide exposures. Such information would be used to refine
tolerance levels and evaluate human safety.

1.6.3  Toxic Substances Control Act
Current uses of biomarkers to support TSCA: The Office of
Toxic Substances (OTS) is using biomarkers to measure na-
tional trends of pesticide  and other chemical exposures.
Measurements of fat and blood levels for these chemicals are
being used to assess uptake on a national basis.

Potential uses of biomarkers:  The  development  of more
precise biomarkers would support the refinement of national
exposure and  uptake assessments.   Biomarkers of effect
would provide valuable information for extrapolating animal
test data to human outcomes. In particular, biomarkers that
are predictive  of neurotoxic endpoints and carcinogenesis
are needed.

1.6.4  Clean Water Act
Current uses of biomarkers to support CWA:  Biomarkers
are being used in water quality advisory and sludge criteria
development (e.g.,  cholinesterase inhibition).  The water
quality criteria are being developed on the basis of endpoints
such as lethality, reproductive impairment, carcinogenicity,
mutagenicity, and teratogenicity.

Potential uses of biomarkers: The Office of Water Enforce-
ment and Permits/Monitoring and Data Support Division
(OWEP/MDSD) needs biomarkers as screening  tools to
detect human and fish exposures to pollutants originating in
industrial/municipal effluents.  Tests are sought  that are
quick, inexpensive, and easily traced to a point source. As
an example, pollutant levels could be measured in blood or
urine to pinpoint a single industrial point source which dis-
charges to a river used for recreation, drinking  water, or fish
consumption.


1.6.5  Safe Drinking Water Act
Current uses of biomarkers to support SDWA:   Many drink-
ing water standards are based on exposure estimations based
on biomarker measurement, such as No-Observed-Adverse-
Effect Levels (NOAELs) or Lowest-Observed-Adverse-
Effect Levels (LOAELs).  Examples of chemical classes for
which these have been prepared include organophosphate
pesticides (biomarker: absence of cholinesterase inhibition);
metals (effect: liver or kidney damage; biomarker: level of
blood or urinary enzymes, blood lead level); and chlorinated
disinfectants (effect: long-term chronic; biomarker: serum
cholesterol levels, effects of immune toxicity).

Potential uses of biomarkers: Biomarkers will continue to
be used to implement regulations under the Office of Drink-
ing Water (ODW).  As with other programmatic areas, a key
issue  in using biomarkers to  develop drinking water stand-
ards and advisories is ensuring that a clear cause-and-effect
link can be made between the biomarker and exposure from
a specific chemical.

1.6.6  Resource Conservation and Recovery Act
Current uses of biomarkers  under RCRA have  not been
identified. In the future, however, biomarkers are needed in
two areas:   hazard assessments under Sections 3013 and
7003, and as a tool to assess the national or local success of
hazardous waste control and disposal management methods.
For example, biomarkers  could be used to define exposed
populations in the vicinity of RCRA-permitted facilities, and
to reassure enforcement staff that exposure is within an ac-
ceptable range. Also, these measurements could be used to
assess the risks associated with treatment variances and spe-
cial wastes from the oil and gas, coal-fired utility and cement
industries, as well as medical wastes, incinerator emissions,
and contaminated soil.

1.6.7  Comprehensive Environmental Response,
       Compensation and Liability Act
Current uses of biomarkers to support CERCLA: The Agen-
cy for Toxic  Substances and  Disease Registry (ATSDR)
uses a variety of biomarkers in human monitoring  studies at
Superfund sites.

Potential uses of biomarkers:  Exposure biomarkers are of
more  immediate interest to the Superfund program than ef-
fect measurements: Regulatory enforcement schedules can-
not wait for the results of the longer-term research required
to develop effects biomarkers and validate their predictive
value. Biomarkers are needed for national and local trend
analysis — for example, to assess whether people  are being
exposed to hazardous substances and, if so, by what routes.
Of particular concern to the Superfund program are noncan-
cer risk assessment methodologies, including  genetic, ner-
vous system, and reproductive system effects from chemical
exposure and combinations of effects from exposures
to  complex  chemical mixtures.  Chemical-specific  and
endpoint-specific biomarkers are needed to provide remedial
project managers with a more complete and realistic under-
standing of health risk.

-------
To summarize, the key needs for the regulatory offices
are cheap, quick, validated biomarker tools that provide
data about  exposure and key health effects, especially
cancer, reproductive, and neurotoxic effects.
1.7    APPLICATION OF BIOMARKERS
The risk assessment and lexicological needs of the various
program offices point to the connections between fundamen-
tal research, toxicity testing, and biomarker work.  For ex-
ample, biomarkers that can be used to evaluate exposures as
part  of EPA  risk assessments are often  tools used  in
mechanistic studies on membrane integrity, transport proces-
ses, and disease mechanisms.  Thus, research to develop
physiological models is closely integrated with efforts to
develop risk assessment models.

Biomarkers can be used in field studies examining a variety
of specific test populations. These efforts cover:

• Spills/accidents — situations that present opportunities
  to study short-term exposures to atypically high
  concentrations of chemicals.
•  Clinical drug testing — carefully controlled, measured
   exposures to specific substances.

•  Pollutant testing — controlled exposure studies to a
   known dose of a substance, such as ozone or NOX.

•  Case studies of disease incidence — cases in which
   health effects are known or suspected, but the cause(s) of
   the problem are unclear.

•  Occupational cases — situations in which the disease
   incidence appears to be related to a common occupation,
   or the suspect chemical substance/mixture is peculiar to
   an occupational setting.

•  Chronic nonoccupational — low-level exposures over
   long periods of time, such as residents downwind of an
   industrial facility or waste site.

•  Status and trends — work to identify trends in a large
   population at discrete points in time, possibly over a
   period of years.

The potential usefulness of biomarkers in such field studies
is  the primary reason the Agency  is  supporting research to
develop, validate, and apply these measurement tools.  For
more detail, see Section 3.1.

-------
Section Two
Biomarker Research  Strategy
This document focuses on the issues and priorities involved
in establishing an ORD biomarker strategy rather than on the
specific scientific approaches  used to  implement  that
strategy. For documentation concerning  ORD's plans for
using biomarkers, refer to the following documents: Role of
Pharmacokinetics and Biomarkers in Risk Assessment Im-
plementation  Plan: Part  I, Exposure Assessment (EPA,
1990e); the Human Exposure Research Program: Strategy
and  Plan (EPA, 1990b); and the OMMSQA Exposure
Research Strategy (EPA,  1990a).  In this section, the key
criteria for selecting ORD biomarker research projects are
outlined and key issues are introduced.
2.1   ROLES OF PRIMARY ORD OFFICES

Extensive planning and coordination among the ORD of-
fices will be required to implement an ORD-wide biomarker
program. Three offices will play key roles in this effort.
The Health Effects Research Laboratory (HERL) in the Of-
fice of Health Research (OHR) has the lead in conducting
laboratory  research  and developing the biology  for the
biomarker methods as well as in developing the associated
experimental data needed to develop the PB-PK  models.
The Office of Modeling, Monitoring Systems and Quality
Assurance (OMMSQA) has the lead in validating exposure
biomarkers  through  the  development  of measurement
devices and exposure models and the gathering of human
field data (in collaboration with HERL epidemiologists).
The Office of Health and Exposure Assessment (OHEA) has
the lead for risk assessment applications  and will  work
closely with OMMSQA and HERL scientists to develop risk
assessment   models  that synthesize   exposure,   phar-
macokinetics, and effects data.  These efforts help  identify
data and model limitations, which can then be addressed in
subsequent research projects.
2.2   ESTABLISHING A BIOMARKER
      RESEARCH STRATEGY — CRITERIA
      AND PRIORITIES
The criteria ORD will apply in selecting biomarker projects
are described here in priority order.
I.   Overarching Criteria

Projects must:

1.   Address the major scientific uncertainties in risk assess-
    ment:

    —High-to-low-dose extrapolation

    —Dose rate effects on tissue damage

    —Nonlinearities in pharmacodynamics

    —Validating route-to-route extrapolation

    —Elucidating mechanisms of toxicity

    —Explaining species, sex, and/or strain differences

    AND/OR

2.   Provide chemical-specific data needed by regulatory
    programs, either in current or future standard-setting ac-
    tivities.

    The biomarker program will  give first  priority to
    projects designed to advance knowledge on key issues
    in risk assessment.  To the extent possible within that
    context, projects will be selected that address chemicals
    of concern to program offices.  Any project that addres-
    ses a scientific uncertainty in risk assessment through
    work on a chemical of great regulatory concern will be
    given top priority in the program.

    Key scientific issues for risk assessment:

    —To what chemicals are people exposed?

    —What amount of a chemical is absorbed into the body
    across the skin, lung, or gut barriers?

    —Where do chemicals move in the body and what hap-
    pens to them there?  How does their fate depend on the
    route of entry?

    —What early biological responses occur following ex-
    posure to the target organ(s)?  Which of these can be
    measured in the blood or urine?

    —How do these early responses relate to health out-
    come?

-------
    Prioritization of chemical-specific efforts:

    Chemical-specific projects will be selected on the basis
    of these factors associated with a chemical:

    —Frequency of occurrence in the environment

    —Uniqueness of occurrence in the environment as op-
    posed to occurrence in medicines, the workplace, or
    food

    —Physical and chemical properties

    —Toxicological properties

    —Chemical specificity

    —Significant human exposure

3.   Provide near-term payoffs (i.e,, short-term applied re-
    search  and longer-term applied fundamental  research
    with interim outputs).

    Work  on exposure biomarkers will be given first
    priority  because  these efforts  will have the greatest
    near-term, practical impact on  improving risk assess-
    ments.  Questions these studies can answer include: To
    what chemicals is a population (individual) exposed?
    To how much of a  particular chemical is a population
    (individual) exposed?   Exposure biomarkers  are also
    often easier to develop  and use than effects biomarkers.
    Exposure markers can be measured by simple tests (e.g.,
    on blood, breath, and urine), while effects markers often
    involve techniques  that are more invasive and expen-
    sive.

    The state-of-science in a particular discipline, coupled
    with the  ease of developing biomarkers in that area to
    achieve near-term payoffs, determine the prioritization
    of work on  effects biomarkers.   Projects on effect
    biomarkers represent longer-term research, focusing on
    such questions as: What is the target organ dose?  What
    response is occurring  in  an  exposed  population (in-
    dividual)? What are the steps from exposure to effect?
    The highest-priority disciplines for development of ef-
    fects biomarkers are as follows:

    —Cancer.  Many  advances  have been  made  in this
    area, due to a strong research focus over the last decade.
    Emphasis will be placed on target organ and molecule
    dosimetry and on enzymatic and cell surface markers of
    neoplasia.

    —Pulmonary.   ORD  has a sophisticated  pulmonary
    program and  extensive working  knowledge  of pul-
    monary function.  Emphasis will be placed on validating
    techniques for use  in  clinical  studies to characterize
    dose-response relationships and on developing nonin-
    vasive techniques for field studies.
—Neurotoxicology.  This is  an important regulatory
endpoint and classical neurotoxicity techniques provide
a means for validating newly developed techniques.
Emphasis will be placed  on validating dose-response
relationships and on developing noninvasive techniques.

—Reproductive.   Serum, urine,  sperm,  and sputum
samples can provide insight into system function. Em-
phasis will be placed on techniques for assessing fer-
tility.

—Immunotoxicity. Many measures are available.  The
emphasis will be on evaluating the health significance
of these measures.

Within  these  priority  areas,   the   program  gives
precedence  to projects developing new measurement
techniques either for which no alternatives are available
or that are less expensive, quicker, and better than avail-
able techniques.

Second-priority topics include work in these disciplines:

—Developmental. Biomarker efforts in  this area re-
quire  invasive  techniques.   Furthermore, there  are
limited opportunities for studies of pregnant women fol-
lowing environmental exposures.

—Heritable Mutation. Presently available methodol-
ogy is inadequate to demonstrate chemically induced
heritable mutations in humans. Genetic variation is also
an unresolved issue.  Biochemical measures can be used
to identify altered phenotypes.

—Other Organ Systems. Other agencies are focusing
on this area.  The Agency for Toxic Substances and Dis-
ease Registry (ATSDR) and  the Centers  for Disease
Control (CDC), for example,  are studying kidney and
liver biomarkers (CEHIC/ATSDR, 1988).

Third-priority topic: Susceptibility

Much of the scientific basis needed to develop suscep-
tibility biomarkers is still not available, and this type of
biomarker also raises many ethical questions. For ex-
ample, should individuals in a test population be told
they  carry   a  biomarker  pointing  to  an  increased
likelihood of contracting a serious illness? Projects on
susceptibility biomarkers  score  low on the third
criterion for near-term results; however,  to the extent
that studies in various scientific disciplines develop data
on  susceptibility, this information will be applied to
biomarker work.

For exposure,  effects, and  susceptibility biomarkers
within each discipline, projects would  be  selected
through  a tiered approach appropriate to the state-of-
the-science  and the specific test population:   Tier 1,
quick, inexpensive, noninvasive studies designed to
determine the presence of exposure or health effects;
10

-------
    Tier 2, more costly and invasive studies to confirm the
    initial observation and begin to focus  on a particular
    aspect of either exposure or effect; and Tier 3, more nar-
    rowly focused and invasive examination to fully under-
    stand the exposure or effect.  In the design of biomarker
    studies, Tier 1 tests will be relied on as much as pos-
    sible. Animal studies and other nonhuman studies will
    be  used as appropriate, but human  studies will be
    needed  to validate the Tier  1 tests.  As  opportunities
    arise for using clinical and other invasive Tier 2 and
    Tier 3 techniques to obtain human tissues, these will be
    evaluated on a case-by-case basis.

II.  Importance of Flexibility: Response to Opportunity

4.  Identify and take advantage of opportunities to develop,
    validate, and apply biomarkers in humans.

    The biomarker program must be flexible in its selection
    of projects.  Opportunities to work with populations ex-
    posed to specific chemicals of interest are rare, and the
    program must  have the capability to respond quickly to
    these chances.  Section 3 discusses application of cur-
    rently available biomarkers and  details  how further
    studies in  test  populations would apply these measures
    of exposure and effect.

m. Other Key Criteria

    Three other key criteria  are also used  to prioritize
    biomarker projects. Selected projects must:

5.  Provide a practical tool at a modest cost for work with
    human subjects for EPA's purposes.

    To be practical, biomarkers must be:

    —Measurable in breath, fluids, or tissues from humans.

    —Measured using non- or minimally invasive techni-
    ques.

    —Interpretable in terms of human  susceptibility, ex-
    posure, and/or effect.

    —Indicate a significant and severe rather  than a minor
    health effect.

6.  Support competency building within ORD.  Innovative
    projects and programs will attract top-notch scientists to
    EPA. Biomarkers competency building will be struc-
    tured around pressing Agency biomarker needs.

7.  Be  leverageable  with  base  program and  other  key
    programs,   and  must be  integrated  across  ORD
    laboratories.

    Animal  research  conducted in  the base  and other
    programs is  key to the development and  validation of
    biomarkers (see IV below).  Various ORD laboratories
    are involved in the development, validation, and/or ap-
    plication of biomarkers, and their efforts  in exposure
    biomarkers are coordinated through a pharmacokinetics
    research  program  (see  Section  2.3.1).    In addition,
    biomarker efforts can vitally influence the success of
    other current  ORD  research initiatives,  such as in
    epidemiology.

IV. Type of Study: Human as Opposed to Animal

    Studies  on exposure  or effects biomarkers are  per-
    formed in either 1) laboratory test species or 2) humans.
    Work on experimental animals is performed under the
    base research program or under a program called Re-
    search to Improve Health Risk Assessments (RIHRA)
    (EPA, 1990f).  The biomarker program focuses on the
    development of measures (i.e., biomarkers) that can be
    applied in humans; RIHRA focuses on how to extrapo-
    late risk from animal data to human exposures.  As
    noted above  in  III 7,  work conducted  in the base
    program  and RIHRA will  be  leveraged  to develop
    human biomarkers to the fullest extent possible.

V.  Type of Human Study

    Biomarker work with human populations can be divided
    into three types of studies:  clinical, epidemiologic, and
    field.

    Clinical and epidemiologic  studies will probably be
    conducted only for chemicals of major environmental
    importance,  such as  National  Ambient  Air  Quality
    Standards (NAAQS) air  pollutants (e.g., ozone, carbon
    monoxide)  and  water  disinfectants  or  disinfectant
    byproducts (e.g.,  chloramine,  dichloroacetic  acid).
    Clinical studies  are  often  invasive,  expensive,  and
    designed to be helpful in  the treatment of individual
    patients.   Patients who expect  benefits  from  such
    markers will tolerate procedures they would not accept
    in  the field.  Epidemiologic studies can also be costly;
    however, biomarkers are a critical tool in improving ex-
    posure assessments in  these studies, and efforts will be
    made to use ORD biomarker research to support any in-
    house exposure initiative as  well as the work of other
    groups, such as the Centers for Disease Control, which
    runs large epidemiological studies.

    Field  studies,  which  are key  to the validation of
    biomarkers, will emphasize status and trends monitor-
    ing and total exposure assessment or assessment of per-
    sons in  chronic  nonoccupational settings (e.g., near
    specific industrial sources or waste sites, or close  to a
    chemical accident).
                                                                                                              11

-------
2.3   PROGRAM ISSUES

2.3.1 Coordinating ORD Research Efforts
As the state of environmental science has evolved over the
past  two decades, the distinctions between the missions of
engineering, monitoring, ecological and health laboratories
have blurred.  Now, all the laboratories are integrated into a
continuum (see Figure 1).  Measuring the release of a chemi-
cal substance into the environment is not sufficient as a basis
for sensible regulation. Decision makers also need informa-
tion on human exposure, health risk, and the efficacy of con-
trol technologies.  To efficiently use ORD resources,  ORD
laboratories need to collaborate, working to apply their ex-
pertise in part of the risk assessment continuum in concert
with others to address the full array of risk assessment con-
cerns.

The  biomarker  research  conducted   in  various   ORD
laboratories must be forged into  a  unified, cost-effective
program.  ORD  is composed of 14 laboratories  scattered
across the United States.  Some of the laboratories have al-
ready developed  independent biomarker research strategies
within their respective disciplines.  Furthermore, much re-
search is currently underway in these laboratories in  phar-
macokinetics, a research area in which exposure biomarkers
are a key tool. The first step in focusing these activities and
determining lead  responsibilities for exposure biomarkers
was  the  establishment of an  interlaboratory Working Group
in January 1990.  The coordinated pharmacokinetics strategy
developed by  this group will be phased in over the next two
to three years, beginning in the  1990 fiscal year.   The
strength of the biomarker research program will be greatly
enhanced by the continual exchange of ideas among the dif-
ferent laboratories and professional disciplines involved in
biomarker research.

2.3.2 Leveraging the Research Results of Other
      Federal Research
Given its limited biomarker research budget, ORD will focus
its efforts on pursuing unique and essential contributions to
the technology for exposure and effects biomarkers.  It will
leverage the  results  of  the  very substantial  research
laboratory developments  in  biomarker assays that are al-
ready in progress in HERL, the National  Institute for Oc-
cupational Safety and  Health  (NIOSH),  the  Agency  for
Toxic Substances and Disease Registry (ATSDR), the Na-
tional Institute of Environmental Health Sciences (NIEHS),
their extramural  grants programs,  and  others, bringing the
more promising biomarkers rapidly into field validations and
applications.  Early demonstrations of the utility and power
of the emerging biomarker technologies will further stimu-
late  additional research and applications in EPA programs
and the field in general. To date, ORD had conducted litera-
ture  evaluations of DNA adducts (EPA, 1990c), protein ad-
ducts (EPA,  1989c;   1990c),  and  chemically  exposed
populations (Uziel et al., 1989) to identify and assess the
work done by others.  (The chemically exposed population
survey  was conducted with  the  Oak  Ridge  National
Laboratory.)

ORD has also  taken advantage of opportunities to use
the substantial databases from the National Health
and Nutrition  Examination Surveys (NHANES) and
Health  Interview Surveys (HIS). NHANES has large
quality-assured databases on national population samples,
including trace metals  in  blood, serum chemistry,  tap
water concentration,  pulmonary  function, clinical data,
and household characteristics. These data, when tapped by
ORD epidemiological research, show that highly significant
associations can be made between high blood lead levels and
elevated blood pressure in  adults  and reduced stature and
hearing  acuity in  children.   In addition, during  the 1976-
1982 NHANES n data collection interval,  blood lead levels
dropped precipitously, in parallel  with the decline in lead
content  in motor vehicles fuel, thus showing the utility of
blood  lead  as an  exposure as well as an  effects
marker. Further use  of the NHANES I and NHANES II
databases will  yield other linkages between environmental
factors and human disease.

In addition, ORD commissioned the National Academy of
Sciences (NAS)  to   define the  state-of-the-science  for
biomarkers and to recommend which biomarkers are useful
now and which show promise for future development. The
ORD project  officer established  an advisory   group of
government experts to share information about their research
programs and to advise the NAS panels. ORD heavily relied
on these reports in the development of this document and is
taking the results from the NAS studies (NAS, 1989a, b) into
consideration in choosing research projects for funding.

2.3.3  Lack of Uniformity Among Scientific
       Disciplines
A third  issue is the lack of uniformity in the development —
and in the speed of development — of the different scientific
disciplines  associated  with biomarker  research.   This
variance is a factor in the types of studies that can be con-
ducted.   It also highlights  a need to define the level of
development in a discipline at which EPA should invest
resources  in biomarker research.   To date,  the state-of-
science  in each field has driven ORD biomarker research,
and efforts have focused on animal studies to develop tech-
niques for application in humans.  However, current criteria
for ORD biomarker research (see Section 2.2) place priority
on programmatic as well as scientific considerations. The
goal of the program is to develop the  techniques most
suitable for use in human studies that also  meet EPA's
regulatory needs.
 12

-------
23.4  Concerns Raised in Performing Studies on Human
       Populations
Performing biomarker studies with human populations raises
both ethical  and procedural issues.   For example,  what
obligation does  a researcher have to interpret the human
health  significance of a biomarker for members of a test
population in whom the biomarker was identified? Principal
investigators must clearly define the information that will be
provided before  the beginning of the study and ensure that
all participants understand these conditions.  Also, proce-
dures should be established for the different types of techni-
ques used to gather human data. These range from the use
of discarded body fluid specimens  obtained  as part of
routine medical care (e.g., human urine samples with no per-
sonal  identifiers)  to  relatively  invasive  procedures per-
formed specifically for a study (e.g., bronchoalveolar lavage
or fat biopsy). These procedures must follow  the guidelines
established under 45 CFR 46 as amended under the pending
Federal Policy for the Protection of Human  Subjects (see
53 FR 218,  Nov.  10,  1988).   In  addition, Institutional
Review Boards must review all such procedures (with cer-
tain  exceptions,  such as the use of  existing pathological
specimens obtained without identifiers).

Special rules apply to human immunodeficiency virus (HIV)
testing conducted or supported by the Public Health Service
(PHS).  Individuals whose test results are associated with
personal identifiers must be informed of these  results and
given appropriate counseling, if requested.  Exceptions to
this policy are defined in the National Institutes of Health,
"Office for Protection from Research Risks Reports," June
10,1988.

For each study, EPA will develop a comprehensive plan for
informing all  participants of biological monitoring  of  the
results.  This plan will cover the issues of confidentiality and
privacy.  In addition, EPA needs guidelines for anticipating
the ethical, legal, and social impacts of biological monitor-
ing.

2.3.5  Validation of Biomarkers
Once a biomarker is developed, it must be validated.  To
validate a biomarker, researchers must show that the meas-
urement can be used to reconstruct the level of exposure to a
chemical or predict a target dose or resulting health  effect.
In addition, the regulatory offices require biomarkers that
allow researchers to clearly identify the specific pollutants to
which a population is exposed, or the specific routes of ex-
posure involved (see  Section  1.6).  Validated biomarkers
also provide a basis for extrapolating risk  from animal data
to human conditions of exposure.  No biomarker will be
used alone in a human study unless it has been validated.

Several ORD laboratories (see Section 2.1) will be involved
in the validation of new biomarkers.  In addition, animal
data from research efforts in the base program will be used
to supplement data obtained in biomarker field studies.
Often, researchers  will employ  a putative  biomarker  in
parallel with a well-defined measure(s) of susceptibility,  ex-
posure, and/or effect.
                                                                                                                13

-------
Section Three
Application  of Current  Biomarkers in
Opportunistic Studies
If the opportunity arose  now to develop  or  validate a
biomarker in a human population exposed to a specific pol-
lutant, what course of action would researchers in the ORD
biomarker program follow? This section discusses how cur-
rently available exposure and effects biomarkers would be
used in such studies over the next three to five years.
3.1   EXPOSURE BIOMARKERS
In establishing field studies of exposure biomarkers, ORD
will critically evaluate the situation and design the effort ac-
cordingly on a case-by-case basis.  No two studies are exact-
ly the same. The kernel of the process, however, is provided
below for two types of field studies — source-specific ex-
posure modeling and status and trends monitoring.

Source-Specific Exposure Monitoring:

1.  Answer the questions: What regulatory/enforcement or
    scientific issue requires exposure information?  How
    will collection of biomarker data help the Agency?

2.  Identify sources and quantify emissions to the environ-
    ment.

3.  Trace movement of chemical substances from source to
    human(s).

4.  Screen the potentially affected humans with the aid of a
    questionnaire  and  measure  the  relevant validated
    biomarker of exposure, e.g., benzene  in breath, blood
    lead, etc. If high, go to step 5 or 6.

5.  Institute mitigation strategy.  End sequence, except for
    follow-up if necessary.

6.  Measure exposure pathways.

7.  Measure personal exposure levels.

8.  Relate the exposure levels, concentration in pathways
    and control group results.

9.  If high, go to step 5.  If low or further action is unwar-
    ranted, stop.
Current Status and/or Trends Monitoring:

1.   Identify chemical(s) of top priority to EPA.

2.   Identify exposed or potentially exposed human popula-
    tion.

3.   Select appropriate validated biomarker.

4.   Collect tissue/fluid/breath sample that will contain the
    biomarker.

5.   Analyze sample for the biomarker presence and con-
    centration.

6.   Apply pharmacokinetic models to these biomarker data
    to estimate target tissue dose as well as exposure con-
    centration.
3.2   EFFECTS BIOMARKERS
Similar to studies of exposure biomarkers, ORD will design
effects marker field studies on a case-by-case basis.  Again,
no two studies will be exactly the same. Often, EPA will
piggyback research onto human studies conducted primarily
by others (e.g., CDC). The process ORD will follow is out-
lined below.

Effects Evaluation (with a validated biomarker):

1.   Answer the questions: What regulatory/enforcement or
    scientific issue requires effects information?  How will
    the collection of biomarker data help the Agency?

2.   Identify chemicals of interest and their sources of ex-
    posure.

3.   Review literature and other databases, using structure-
    activity relationship (SAR) techniques as appropriate to
    identify likely health  out comes  associated  with  ex-
    posure to the chemicals/sources identified in  Step 2
    based on known or presumed links between  exposure
    and disease.

4.   Select appropriate validated biomarker.
                                                                                                    15

-------
5.   Collect tissue/fluid/ breath or other samples that will
    contain the biomarker.

6.   Analyze the sample for biomarker presence.

7.   Apply pharmacodynamic models to these data to predict
    health outcome.
When a human study is planned that incorporates proven
techniques   and  ORD  is  using  that  opportunity  to
develop/validate a new biomarker for either exposure or ef-
fect, the following steps will be taken:

1.   Identify chemical effects of concern.

2.   Develop the concept — identify the likely consequence
    of chemical exposure  that  might serve as a useful
    measure of exposure.

3.   Experimentally confirm  validity  of  the concept by
    evaluating  the  basic   mechanisms  by  which  the
    biomarker responds to chemical exposure.

4.   Develop method of measurement  by  identifying  a
    method for detecting changes in the biomarker at doses
    at or below those producing toxic effects.
5.  Determine if the biomarker is practical for the field by
    developing plausible field methodology and assessing
    the sensitivity of the biomarker in monitoring existing
    exposures.

6.  Establish a dose-response relationship.   Kinetics  of
    biomarker response must have a rational relationship to
    metabolism and pharmacokinetics of chemical.

7.  For each specific study in which biomarkers are applied,
    evaluate the population too.  Determine if individual
    variables preclude biomarker use.  Do lifestyle, genetic,
    disease/nutritional variables modify the response?

8.  Validate applicability in humans. Conduct a pilot study
    with a small group of humans with defined exposure
    gradients to the chemical of interest.   Examples of the
    types of endpoints that may be measured include:  can-
    cer, pulmonary, reproductive/developmental.

9.  Conduct a demonstration study.  Determine that  varia-
    tion in response can be accounted for  by exposure and
    previously identified variables.

10. Apply (validated) biomarker.
 16

-------
Section Four
Future  Research Directions
This section identifies the philosophy ORD will follow to
develop biomarkers that show promise and should be tar-
geted for further  development over  the  next  three to
five years (near-term research) and beyond. These tools and
techniques  will  best  improve  the  Agency's ability  to
evaluate chemical exposures and resultant health effects to
support its missions and responsibilities.
4.1    EXPOSURE BIOMARKERS

One of the major areas of concern in biomarker development
is the specificity of the tool.  Current regulations are chemi-
cal specific, and biomarkers are needed to unambiguously
identify and characterize exposures to specific chemicals.
To meet Agency needs, biomarker research efforts will en-
compass two types of markers:  those that are compound-
specific  for  chemicals   of  concern  (e.g.,  benzene,
trichloroethylene, acrylamide,  styrene, nicotine, lead)  and
those that are indicators of relevant classes of compounds
(e.g., dioxins, PCBs). In addition, the program will include
three related methods development activities:

1.  Evaluation of new scientific information with regard to
    newly developed biomarkers and biomarker technology,
    as well as detection method technology (i.e., can they be
    adapted for Agency use?).

2.  Laboratory  studies to ensure full  characterization of
    potentially useful biomarkers; refinement/development
    of  pharmacokinetic  models   for  the   analysis  of
    biomarker  data;  and  development  of  monitoring
    methods  and  devices  for detecting and  quantifying
    biomarkers.

3.  Field trials of markers, methods, devices, models,  and
    protocols.

The most promising biomarkers of exposure in the near term
are measures related to body burden because they provide
measurements of specific chemicals found in the blood,
urine, saliva, and breath, thus yielding data directly useable
in risk assessment models. Often a breath biomarker is the
specific  chemical substance under study and the biomarker
measurement serves to confirm exposure.   Other  (non-
specific) markers of exposure  are useful too because they
support  assumptions of exposure to  a specific chemical or
related substances.  They include conjugate complexes or
elevated or depressed enzyme levels.

Ongoing but longer-term research includes characterization
of DNA and protein adducts (including hemoglobin, al-
bumin, and membrane receptor proteins) to detect exposure
to toxic chemicals.   However, while these  measures  and
others, such as sister chromatid exchange (SCE), may be
used as first-tier  tests  to suggest that  exposure to some
chemical of concern has occurred,  follow-up, chemical-
specific tests must then be applied to provide data that are
useful for exposure assessment and risk characterization.

For reconstructive exposure assessment to be a practical
quantitative tool, a model based on the pharmacokinetics of
the particular chemicals in the individual or population of in-
terest is necessary.  These models, together with measure-
ments of the specific biomarker, will form the scientific
basis to link source, exposure, dose, and health effect.

The following efforts should have the highest priority in fu-
ture research:

•  Interpret/develop biomarkers for body burden (blood and
   urine sampling), including:

    —Identify  and  apply  markers   for   field  and
    epidemiological  studies on selected subject populations

    —Develop pharmacokinetics models for estimating tar-
    get tissue dose from surrogate tissue levels and for es-
    timating/reconstructing external exposures from  dose
    measures

    —Identify  appropriate  metabolites as  body  burden
    measures

•  Develop and validate new body burden markers for
   high-priority chemicals (e.g., from Total Exposure
   Assessment Methodology (TEAM) studies).  These
   markers would likely be:

    —Receptor-xenobiotic complexes

    —Protein adducts

    —DNA adducts
                                                                                                           17

-------
4.2    EFFECTS BIOMARKERS
EPA considers risk versus benefit in the vast majority of its
regulatory decisions.  The physiological responses that serve
as effects biomarkers are not disease endpoints. Unless the
relationship between  the biomarker and the disease can be
shown, the physiologic response will likely be considered
too trivial to form the basis for regulatory action.  As noted
in Section 4.1, most current Agency regulations are chemical
specific.  Thus, a major emphasis of effects biomarker re-
search at EPA will be to link exposure with health outcome.
For each of the priority areas, key needs are outlined below.

4.2.1  Cancer
Most major EPA regulations are based on cancer risk — a
framework with profound economic and social  implications.
Agency cancer risk  assessors  use a multistage model to
quantify the estimated lifetime cancer risk associated with
exposure to chemical carcinogens.  These assessments cur-
rently rely on a number of assumptions (see  Section 1.1),
and thus may result  in either over- or under-regulation of
economically useful chemical products.  Recent advances in
cancer biology, however,  are  fueling the  development of
more realistic cancer risk assessment models, which will in
turn facilitate the development of regulations that neither un-
derprotect the population nor unnecessarily hinder the  in-
dustrial use of chemicals.

Two  of  these advances are the identification of tumor
growth, or "onco," genes,  and  tumor anti-growth, or "sup-
pressor," genes. Studies on these genes in malignant cells of
certain tumor types,  such  as colon cancer, show that the
genome of these cells contain more than one genetic change,
and suggest that many mutations (perhaps 10-15) must occur
before cancer results (Cavanee et al., 1989). Improved
knowledge of the mechanism of cancer production following
chemical exposure will allow the development of measures
for evaluating whether humans have been exposed to car-
cinogens and for assessing risk. They also provide a basis for
intervention to mitigate exposures.

Scientists  currently evaluate exposure to carcinogens with
DNA and protein adducts because  of  their  chemical or
chemical-class  specificity.    Many  uncertainties  exist,
however, concerning the relationship of these  substances to
cancer, and many technical difficulties need to be resolved
in order to be able to expand promising detection techniques
(e.g., the   P-postlabeling  technique) to a wide variety of
chemical classes.  Initial efforts will focus on the develop-
ment and application of DNA adducts, stressing:

•  Lesion formation

•  Rate of repair

•  Tissue distribution

and working to:
•  Define the dose/response relationship between the adduct
  • and health effect

•  Define the relationship between measured adduct levels
   and target tissue levels if not measured in target tissue

•  Determine the lifetime of the adduct

•  Establish adduct reference standards

•  Extend the methodology to a wide variety of chemical
   classes

ORD researchers will study protein adducts in order to:

•  Relate adducts to target site dose/response

•  Develop and improve methodologies and techniques
   (preparation and identification of adducts)

•  Evaluate/refine applicability of biomarkers for field
   testing

A major emphasis for both DNA and protein adducts will be
to:

•  Develop adducts for evaluating exposure to complex
   mixtures (including reference adduct standards) and to
   identify patterns or arrays of adducts characteristic of
   chemical classes; these patterns or profiles will be stored
   on computer files

Regarding  the various genes that  control cell  growth,
proliferation, and differentiation,  ORD will apply  current
techniques   in  molecular   carcinogenesis  to   develop
biomarkers for carcinogenesis, including:

•  Alterations  in DNA:

    —Mutations in oncogenes and anti-oncogenes

    —Mutations in DNA repair genes

    —Chromosomal rearrangements including gene
    amplification

    —Multi-locus damage incorporating two-dimensional
    gel analysis

•  Alterations  in the expression or structure of proteins or
   enzymes:

    —By two-dimensional gel electrophoresis

    —By Western blots

    —By radioimmunoassay (RIA) techniques

    —By immunohistochemistry

•  Ultrastructural changes

•  Growth factors

•  Receptors/receptor-xenobiotic complexes:
 18

-------
     —Evaluate the  use and  interpretation  of receptor-
     xenobiotic complexes for specific chemicals

 The risk assessment process for  cancer is based on ex-
 trapolating human  risk from available animal data.   The
 quantified risk factor  is  generally  calculated  for the
 "average" human, even though the human population con-
 sists of a heterogeneous mix of diverse genetic backgrounds.
 This genetic diversity coupled with environmentally induced
 illness (infectious or chemically induced, such as  cigarette
 smoking) creates a stratified population with varying suscep-
 tibility to particular environmental  exposures.   Biomarkers
 will be developed to characterize the interindividual varia-
 tion in the human population and to identify susceptible sub-
 populations.

 4.2.2  Pulmonary
 Perhaps  the most significant environmental hazard facing
 citizens of all countries is air pollution, as noted by EPA in
 the Unfinished Business Report (EPA, 1987).  Automobile
 emissions and industrial outgassing result in  significant
 human exposure to ozone, nitrous oxides, sulfur oxides, and
 carbon  monoxide.    In addition,  household furnishings,
 cleaners, and  lifestyle habits (e.g., cooking, smoking, dry
 cleaning) can result in significant  exposure to many other
 substances.

 The lung is the primary route of entry  for air pollutants.
 Most of the regulations promulgated under the Clean Air
 Act are  based on  data showing effects to the  cardiopul-
 monary system. Because of the lung's relative  accessibility,
 EPA and others have extensively used biomarkers from lung
 fluids and tissue to assess human exposure to air pollutants.
 Most of these studies have been conducted in a clinical set-
 ting, often with invasive techniques.   A major goal of re-
 search on pulmonary  effects  biomarkers is  to  develop,
 validate, and apply less invasive markers on larger popula-
 tions than is now possible.

 The following pulmonary  biomarkers  have  the  highest
 priority for future research:

Near term:

•  Develop methods for assessing the effects from, and
   exposure to, acid aerosols

•  Develop new markers for use with urine and  blood

•  Develop markers for ozone degradation products  (DNA,
   protein products)

•  Strengthen interpretation of nasal lavage (NAL)
   biomarkers

•  Strengthen interpretation of biochemical/cellular markers

•  Strengthen interpretation of molecular biomarkers
Longer term:

•  Develop biomarkers for susceptibility (responders versus
   nonresponders)

4.2.3  Neurotoxicity
A recent Office of Technology and Assessment Report (U.S.
Congress,  1990) has identified neurotoxicity  as  a major
health outcome  following exposure to  environmental pol-
lutants. Many EPA regulations are based in whole or in part
on  measures of neurotoxicity (e.g., from organophosphate
pesticides, metals in drinking water, lead in gasoline). Many
of these measures, however,  are invasive or assess nonrever-
sible  effects  following  exposure.   ORD's  research in
neurotoxicological effects biomarkers focuses  on developing
relationships between exposure and outcome for cholinesterase
and neurotoxic esterase and developing  minimally invasive
techniques to detect nervous-system-specific proteins that point
to the early steps of neurotoxicity.

Research efforts will thus be designed to:

•  Determine the relationship between cholinesterase
   activity in blood, peripheral nervous system, and central
   nervous system.

•  Determine the relationship between the degree of
   cholinesterase inhibition, neurotoxic esterase inhibition,
   and neurotoxic outcome.

•  Determine whether age and/or species differences exist
   in the above relationship.

Other longer-term priority biomarkers under development
include:

•  Gel electrophoretic protein profiles from cerebrospinal
   fluid (CSF)

•  Presence of nervous system-specific proteins and
   degradation products in CSF, plasma and urine

•  Presence of immunoglobulins in serum directed at
   brain-derived antigens

•  Presence of neurotransmitter metabolites in CSF

•  Monitoring applications in blood and urine

•  Subcellular antigens in blood and urine pointing to the
   extent and localization of damage to the nervous and
   other organ systems
4.2.4  Reproductive/Developmental
Humans experience a significant reproductive failure rate.
Exposure  to environmental chemicals has been associated
with infertility, and EPA regulates a number of chemicals
(e.g., dibromochloropropane, nitrofen, dinoseb) thought to
cause this problem. In developing and using biomarkers to
                                                                                                                19

-------
support the Agency's regulatory activities, ORD researchers
will focus their efforts first on males, due to the relative ease
of obtaining samples from men of reproductive age. As they
become available, however, techniques will also be used for
better assessing female infertility.

The following reproductive/developmental biomarkers will
have the highest priority for future research:

Near term, male:

u  Improve in vitro tests of sperm function

•  Improve predictive value of semen analysis; also
   included are efforts to improve sampling techniques,
   standardize results, and develop suitable sample
   containers

•  Validate measurement of testosterone in saliva

•  Develop methods to assess the genetic integrity of sperm
   (e.g., DNA damage, chromatid structure) and generalized
   screening methods for genetic integrity

Near term, female:

•  Multiple sputum samples for steroids to validate the
   sensitivity of the approach

•  Validate measurement of LH  (ovarian cycle protocol) in
   mid-cycle urine

Near term, male and female:

•  Develop noninvasive techniques for human reproductive
   toxicology based on animal models

•  Identify better markers indicative of early changes in
   reproductive function

Longer termjemale:

•  Develop noninvasive methods for biomarkers of
   preimplantation development

Longer term, male:

•  Develop better markers of sperm membrane integrity and
   in vitro tests of sperm function

The highest priority for general research will be given to ef-
forts to:

•  Develop and validate a questionnaire and/or decision tree
   to identify and assess fertility problems in the male and
   female specific to EPA's needs; existing protocols will
   be standardized

•  Develop and validate noninvasive analytical markers
   from urine and saliva that would allow more frequent
   measures of endocrine control of reproductive function
•  Develop and validate biomarkers to monitor early
   pregnancy and to distinguish between pre- and
   post-implantation loss

•  Improve the predictive value of semen analysis,
   including improved methods of multivariant analysis and
   further identification of molecular and biochemical
   markers of sperm function

4.2.5  Immunotoxicity
Many in vivo and in vitro tests can be used to evaluate im-
mune system responses in humans.   Scientists have used
these tests to demonstrate allergic reactions to environmental
chemicals and  altered immune functioning following ex-
posure to environmental  pollutants.  However, because of
the complexity and interactive dynamics associated with im-
mune responses, it  has  been difficult to interpret these
responses for their adverse health impact on humans. Con-
sequently, with the exception of allergic responses, risk as-
sessors have had difficulty using immune system responses
as a basis for regulatory decision-making because these
responses are  not severe enough to be considered  a  sig-
nificant  adverse health  impact.   ORD  research  in  im-
munotoxicity will focus on interpreting the significance of
immune response parameters and on  using immune respon-
ses to link exposure to effects (in both immune and other tar-
get organs).

The  following  immunotoxicity biomarkers will have the
highest priority for future research:

•  Standardize procedures and techniques

•  Improve biomarker data interpretation; consider such
   improvements for their impact on the development and
   prioritization of other biomarkers

•  Improve the ability to extrapolate from animals to
   humans and determine species relevance for human
   hazard identification

•  Develop sensitive immunological markers (with
   identifiable responses at threshold levels)

•  Develop more quantitative markers of exposure for dose
   determination (long term and low probability of success,
   but potentially high payoff)

•  Improve the interpretability of biomarkers of
   immunotoxicity in order to predict susceptibility to
   disease

•  Develop and validate biomarkers that are indicators of
   hypersensitivity responses

•  Improve the understanding of the mechanism(s) of
   chemical-induced immune alterations  (i.e., immuno-
   suppression, hypersensitivity and autoimmunity) and
   the identification of biomarkers for such chemicals
20

-------
4.2.6  Heritable Genetic Mutations
Heritable mutation research will  utilize new recombinant
DNA techniques  to  evaluate mechanisms of mutation in
germ cells.  The significance of genetic recombination and
segregation events at meiosis leading to mutation  will be
analyzed to achieve a better understanding of germ-line
specific targets and mechanisms in the induction of heritable
mutations, in order to identify appropriate biomarkers.  The
greatest challenge will be to develop noninvasive techniques
for use on human subjects to identify mutant phenotypes. In
addition, the research projects will be undertaken to:

•  Evaluate the scientific literature and newly emerging
   epidemiology studies for evidence of environmentally
   induced heritable damage in humans

•  Evaluate and develop rodent models for induced
   heritable damage to identify new biomarkers

•  Utilize new molecular and cytogenetic techniques for
   monitoring offspring for newly induced genetic damage
   in humans

4.2.7  Hepatotoxicity
Because of the variety of serum markers of hepatic damage
available and the varying specificity and sensitivity of these
markers, near-term  research will focus on defining an
optimal battery of serum  biomarkers suitable for  routine
environmental monitoring. Results of interlaboratory proficien-
cy studies indicate marked variability in the  results of serum
enzyme tests (Blanchaert,  1987).   Research efforts  will
determine the causes  of this variability and suggest methods
to reduce its effect on the outcome of studies. Research is
also needed on noninvasive tests for hepatic fibrosis  and
cirrhosis; examples of research areas  include serum/urine
measurement of proline and hydroxyproline and serum assay
(immunoassay) of type III procollagen.
Epidemiologic studies of populations exposed to toxic waste
sites are limited by the technical and human problems out-
lined by Marsh and Caplan (1986), including small popula-
tion size, heterogeneity of the exposed population with
respect to characteristics that can influence health outcome,
poorly defined exposures, and health outcomes that are rare
or have long latency periods.  The ORD program will ex-
amine the influence of  these variables on the outcome of
biomarker monitoring studies.

In addition, priority will be given to efforts to:

•  Evaluate/develop noninvasive clearance techniques;
   examples of research areas include evaluation of
   usefulness of spectrophotometric measurements of dye
   clearance in ear veins for human population monitoring;
   aminopyrine clearance

•  Validate/develop markers for changes in hepatic
   cytochrome P-450 content and composition and hepatic
   glutathione content; research will interpret results,
   determine delivered dose/effect relationships, and define
   the chemical specificity of the methods

•  Validate/develop serum markers to determine the
   location of damage in liver acinus/lobule and subcellular
   location of damage

4.2.8 Other Organ Systems
For other organ systems, future research efforts will focus
on:
   Kidney biomarkers

   Cardiovascular biomarkers targeting morbidity/mortality
   factors

   Biomarkers for organ-specific gene expression
                                                                                                              21

-------
References
Blanchaert N.   1987.   Biochemical liver  tests:   recent
developments and challenges. In: Assessment and Manage-
ment of Hepatobiliary Disease.  Okolicsanyi L, Csomos G,
and Crepaldi G, eds. Springer-Verlag: New York.

Cavanee W, Hastle W, Stanbridge, E, eds.  1989. Current
Communications  in  Molecular Biology: Recessive  On-
cogenes  and Tumor  Suppression. Cold Spring  Harbor
Laboratory Press: Cold Spring Harbor, NY. p. 73.

CEHIC/ATSDR.  1988.  Center for Environmental Health
and Injury Control/Agency for Toxic  Substances and Dis-
ease Registry.   Preliminary draft:   summary  report for
biochemical and cellular markers for kidney, liver and im-
mune systems. October.

Guzelian PS.  1983. Research needs for hepatic injury due
to environmental  agents.  Environmental Health Perspec-
tives 48:73-76.

Marsh  GM and Caplan RJ.   1986. The feasibility of con-
ducting epidemiological studies of populations residing near
hazardous  waste  disposal   sites.   In:    Environmental
Epidemiology, Kopfler FC, Craun GF, eds.  Lewis Publish-
ers, Inc.

NAS.  1983.  National Academy of Sciences. Risk assess-
ment in the federal government: Managing the process.  Na-
tional Academy  Press: Washington, DC.

NAS.  1989a.  National Academy of Sciences.  Biologic
markers  in reproductive toxicology.  National Academy
Press:  Washington, DC.

NAS.  1989b.  National Academy of Sciences.  Biologic
markers in pulmonary toxicology. National Academy Press:
Washington, DC.

NIH. 1988. National Institutes of Health.  Office for protec-
tion  from research risks reports.  Office of the Director.
June 10.

U.S. Environmental Protection Agency.  1987. Unfinished
business:   A comparative  assessment  of  environmental
problems. Overview report and four appendices.
U.S. Environmental Protection Agency.  1989a.  Cell recep-
tor-xenobiotic  complexes  as exposure biomarkers. EPA
600/X-89/362.

U.S. Environmental  Protection Agency.  1989b. Decision
model for the development of biomarkers of exposure. EPA
600/X-89/163.

U.S. Environmental Protection Agency.  1989c.  Protein ad-
duct-forming chemicals for exposure monitoring:   Chemi-
cals selected for further study.  EPA 600/4-89/035.
U.S. Environmental  Protection  Agency.   1990a.
strategy for OMMSQA biomarkers research program.
Draft
U.S. Environmental  Protection  Agency.   1990b.  Draft
human exposure research program:  Strategy and plan.

U.S. Environmental Protection Agency. 1990c. Protein ad-
duct-forming chemicals for exposure monitoring: Literature
summary and recommendations. EPA 600/4-90/007.

U.S. Environmental Protection Agency. 1990d. Draft selec-
tion of adduct-forming  chemicals for human monitoring
studies.

U.S. Environmental Protection Agency.  1990e. Draft role
of pharmacokinetics and biomarkers in risk assessment.  Im-
plementation plan: Parti. Exposure assessment.

U.S. Environmental Protection Agency. 1990f. Research to
improve health risk assessments (RIHRA) program.

U.S. Environmental  Protection  Agency.   1990g.  Draft
ecological biomarkers research strategy.

Uziel M, Munro NB, Katz DS, Vo-Dinh T, and Zeighami
EA. 1989.  DNA adduct formation by  twelve chemicals
with  populations  potentially  suitable  for  molecular
epidemiological studies.   Oak Ridge National Laboratory.
ORNL/TM 108-49.
                                                          •&U.S. GOVERNMENT PRINTING OFFICE: 1991 - 548-187/25605
                                                                                                           23

-------