RESEARCH NEEDS FOR RISK ASSESSMENT
OF INHALED PARTICULATE MATTER
Report of a Workshop Sponsored by the
Office of Health and Environmental Assessment,
U.S. Environmental Protection Agency
Ritz Carlton Hotel, McLean, Virginia
March 10-11,1992
Chairmen:
EPA Project Officers and Workshop Chairs,
William Pepelko, PhJX, and Chao Chen, Ph.D.
Authors:
Gunter Oberdorster, PhJD., and
William Pepelko, Ph,,D.
Technical Support and Conference Management:
Susan J. Brager,
Eastern Research Group
EPA/600/R-93/104
June 1993
Printed on Recycled Paper
-------
DISCLAIMER
The views expressed in this report are those of the authors and do not necessarily
reflect the views or policies of the U.S. Environmental Protection Agency or other sponsors.
Mention of trade names or commercial products does not constitute endorsement by the
Agency or recommendation for use.
11
-------
TABLE OF CONTENTS
Preface iv
1. EXECUTIVE SUMMARY 1
2. INTRODUCTION 3
3. SUMMARIES OF INDIVIDUAL PRESENTATIONS 5
3.1. Overview, Gunter Oberdorster 5
3.2. DNA Adducts, James Bond 6
3.3. Particle-Induced Macrophage Function, Kevin Driscoll 7
3.4. Biology of Macrophages, Bruce Lehnert 7
3.5. Clearance of Inhaled Particles, Timothy Gerrity 9
3.6. Dosimetry Models, Chia Ping Yu 10
3.7. Risk Assessment Methodology for Particles, Roy Albert 10
3.8. Additional Comments by the Workgroup 11
4. WORKSHOP RECOMMENDATIONS 13
4.1. General Outline 13
4.2. Specific Recommendations 14
4.2.1. Particles Are Promoters 15
4.2.2. Differentiation of Adducts 15
4.2.3. Mutation Spectra 16
4.2.4. Dose-Time Response for Radical Production
by Macrophages 16
4.2.5. Dose-Time Response for Cell Proliferation 17
4.2.6. Particle-Macrophage-Cytokine-Mutation Links 17
4.2.7. Examination of Archival Tissue 18
4.2.8. Particle-Associated Compromises in Tumor
Defense Mechanisms 18
5. RESEARCH PLANS RESULTING FROM WORKSHOP
RECOMMENDATIONS 20
5.1. Goals of the Research 20
5.2. Work Plan 20
5.3. Time Schedule 21
5.4. Utilization of Research Data 21
List of Panelists 23
in
-------
PREFACE
Recent studies have shown that insoluble, biochemically inert particles, small enough
to deposit in the deep lung, are capable of inducing carcinogenic as well as pathological
effects. These findings have important implications for risk assessment, especially since
particulate matter of this type, carbon black for example, has been considered in the past to
be relatively benign. On March 10-11, 1992, the U.S. Environmental Protection Agency's
(EPA) Office of Health and Environmental Assessment sponsored a workshop in McLean,
Virginia, on "Research Needs for Risk Assessment of Inhaled Particulate Matter." During
the IV^-day workshop, the 14 expert panelists discussed the current state of the art regarding
pathological effects of inhalable particulate matter. They also developed a list of research
recommendations aimed at improving risk assessment in this area. A number of observers
also were present to witness and join in the discussion. This report summarizes the
proceedings of the workshop.
IV
-------
1. EXECUTIVE SUMMARY
At a workshop sponsored by the U.S. Environmental Protection Agency's Office of
Health and Environmental Assessment, expert panelists discussed research needed to support
the development of improved risk assessment methodologies for inhalable, biochemically
inert, insoluble particulate matter. During the IVi-day workshop, the panelists exchanged
current data concerning the effects and mechanisms of action of particles, identified data
gaps, and provided research recommendations. While both carcinogenic and noncarcinogenic
pathological effects were discussed, the panel focused on carcinogenic endpoints.
The panelists agreed that the first step in the pathological and/or carcinogenic process
is the result of particle ingestion by macrophages residing in the alveolar regions. Following
particle uptake, the macrophages are induced to secrete a variety of mediators (growth
factors, oxidants, proteolytic enzymes, etc.). These mediators diffuse to the target cells
lining the alveoli, are taken up, and induce pathological effects. The panelists also agreed
that the combination of mediators responsible for effects and mechanisms of action are still
somewhat uncertain. There was less agreement about the lung burdens of particulate matter
required to induce harmful effects. Some believed a minimal particle load is required to
induce secretion of mediators. Others pointed out that, for any level of exposure, particle
load varied among macrophages. Thus, no clear threshold for overload activation of
particles is likely.
Other questions arose about the characteristics of particles responsible for activation
of macrophages and the relevance of animal data to human responses. There appeared to be
a general agreement that inflammatory and subsequent cellular proliferative responses may be
the most important events leading to fixation of induced mutations and, finally, particle-
induced lung tumors.
As a result of the discussions, a list of specific recommendations for research was
developed. These included (1) determination of the ability of particles to induce tumor
1
-------
promotion, (2) determination of the specific DNA adducts induced by particles,
(3) determination as to whether deposited particles induce specific mutational changes or
whether spontaneous mutations become "fixed" as a result of proliferative responses,
(4) determination of dose-time response for radical production by macrophages,
(5) determination of dose-time response for cell proliferation, (6) determination of linkages
between macrophage release of cytokine/growth factors and mutational events, (7) use of
archival tissue from previous animal experiments and autopsy material to aid in determining
mechanisms of action, and (8) evaluation of particle-associated compromises in tumor
defense mechanisms.
-------
2. INTRODUCTION
During March 1992, a workshop sponsored by the U.S. Environmental Protection
Agency's Office of Health and Environmental Assessment convened to discuss research needs
supporting the development of risk assessment methods for inhaled particulate matter, with a
primary focus on insoluble, biochemically inert particles. During the IVa-day workshop, 25
participants from academia, industry, and regulatory agencies reviewed current data on the
toxicology, carcinogenicity, and mechanisms of action of inhaled particles; identified data
gaps; and formulated general and specific research recommendations. It is expected that
identification of knowledge gaps and research needs ultimately will benefit EPA's risk
assessment process and also will improve our understanding of mechanistic links between
particle exposure and lung injury, especially tumorigenesis.
The need for health risk assessment and regulation in this area is based on the results
of a number of recent long-term inhalation studies in rats. These studies have shown that
exposure to different particles, including coal dust, titanium dioxide (TiO2), toner, carbon
black, and diesel exhaust, can result in serious lung injury. The injuries are characterized by
functional impairment of alveolar macrophages, chronic inflammation, cell proliferation,
fibrotic reactions, and lung tumors. All of the particles used in these studies are highly
insoluble and have a low intrinsic cytotoxicity. Apparent threshold levels varied with the
agent tested, but effects generally occurred at lung particle burdens exceeding about 1-2 mg/g
of lung. Since lung injury and impaired clearance appeared to be associated with excessive
accumulation of particles in the lung macrophages, the term "particle overload of alveolar
macrophages" was used subsequently to associate these effects with excessive particle
accumulation in alveolar macrophages. It is uncertain at this time, however, whether lung
injury occurs only with lung overload, or if lung overload is a necessary precursor of lung
injury.
A number of questions arose from these studies, such as: Are these effects specific to
the rat? Are these generic particle effects that will occur with any type of highly insoluble
-------
particle of low cytotoxicity? Does potency vary with the particle? If effects vary
qualitatively and quantitatively among particles, what are the physicochemical characteristics
of a particle that are important for its effects? Are these studies relevant to human
exposures? Can we extrapolate these results to humans? Can we extrapolate these results to
low lung particle burdens due to low exposure concentrations or short exposures? Is there a
threshold for toxic effects? If so, is it related to particle overload of macrophages? Can we
identify biologically plausible mechanisms of action(s) to explain these results? What data
gaps do we need to fill to gain a better understanding of mechanistic events? While these
questions were addressed by the participants, time did not allow discussing each of them
exhaustively.
Presentations were made by seven speakers who addressed issues of particle-induced
lung tumors, the role of DNA adducts, the role of mediators (oxidants, cytokines, growth
factors, chemotactic factors) released by alveolar macrophages and other cells, dosimetric
considerations with respect to particle deposition and clearance, and particle-alveolar
macrophage interactions as well as issues of risk extrapolation. The presentations were
followed by intensive discussions. Short summaries of the presentations and some points of
the discussions and individual recommendations are given in the pages that follow.
It was agreed that not enough data are available to improve currently inadequate
procedures for risk assessment of particle-induced lung tumors. For such improvement, the
most urgent need is for mechanistic data. The group, therefore, focused on mechanisms of
particle-induced lung tumors.
-------
3. SUMMARIES OF INDIVIDUAL PRESENTATIONS
3.1. Overview, Gunter Oberdorster
An overview of particle-induced pulmonary effects was presented, which included a
discussion of the sequence of events originating from the particle-macrophage encounter and
leading to chronic inflammatory, cell proliferative, fibrotic, and, in some cases, tumorigenic
effects. The question was raised whether fibrotic alterations are a necessary prerequisite for
subsequent induction of tumors. It was also pointed out that different pathophysiological
changes due to particle "overload" effect appeared to correlate with different dose
parameters. Alveolar macrophage functional changes might be correlated best with a dose
expressed as the phagocytized particle volume in macrophages, whereas inflammatory and
proliferative responses appeared to be better correlated with the particle dose expressed as
surface area of the retained particles. The importance of the pulmonary inflammatory
response in particle overload situations was emphasized, and several research needs were
presented with the ultimate goal of replacing the present simplistic approach of dosimetric
extrapolation modeling with a mechanistically oriented: dosimetric model.
Needed mechanistic data on cellular and molecular mechanisms include cell
proliferative responses, involvement of cytokines and growth factors, the development of
fibrosis, and the formation of DNA and protein adducts. Such data can be collected by using
results of new studies as well as by using archival tissues. It also would be useful to search
for biomarkers or functional assays of "overload," for example, analyzing lavaged lung cells
and fluid for biological mediators released by overloaded macrophages or perhaps measuring
particle clearance in vivo. Known species differences between rats and hamsters, with
respect to tumor induction and other long-term effects in the lung, also should be exploited to
address the question of extrapolation to humans. Dosimetric mechanisms with respect to
particle uptake and retention (e.g., why do alveolar macrophages migrate up the mucociliary
ladder and what retards their movement when they are overloaded?) need further
investigation.
-------
3.2. DNA Adducts, James Bond
Data were presented and discussed indicating that although total adducts induced in
the lungs of rats either by carbon black or by diesel exhaust are quantitatively similar, there
might be differences in specific adduct formation. Preliminary data mentioned in the
discussion showed that adducts elicited by exposure to carbon black may represent polar
adducts, possibly including adducts induced by reactive oxygen species, whereas diesel
exhaust-induced adducts are more specific for the organics present in diesel exhaust. The
carbon core of diesel particles is similar to that of carbon black. However, whereas carbon
black has no more than a trace of adsorbed organic material, numerous organic agents are
adsorbed to the surface of the diesel particle, including carcinogens such as benzo(a)pyrene
and nitropyrenes. These organic agents are thought to be taken up and to act directly on the
target cells. Particles, on the other hand, induce macrophages to secrete mediators, which
then migrate to and are taken up by the target cells. It is therefore not surprising that
different adducts are formed.
It was also pointed out that DNA repair is an important issue and that repair of DNA
adducts could be very different for different regional lung tissues. The specificity of DNA
adducts is a very important issue that needs to be investigated.
Specific recommendations for future research to fill present data gaps include:
1. Ascertainment of the effect of carrier particles on the delivery of adsorbed
compounds to specific regions of the respiratory tract.
2. Ascertainment of rates of desorption of compounds from particles after
deposition.
3. Better elucidation of the actual toxic effects of increased lung retention of inhaled
particles.
4. More complete characterization of respiratory tract metabolism of inhaled
particle-associated xenobiotics (with emphasis on lung metabolism).
5. Identification of specific biomarkers of exposure to inhaled particles.
6
-------
3.3. Particle-Induced Macrophage Function, Kevin Driscoll
In this presentation it was pointed out that releases of certain cytokines and growth
factors by alveolar macrophages following particle uptake are important events in the
particle-induced inflammatory reaction. Fibronectin, for example, may prove to be a
biomarker of fibrotic responses occurring at a later time point. Fibronectin release could be
determined in cultured lavaged macrophages after in vivo particle exposure. The importance
of the inflammatory cytokine tumor necrosis factor (TNF-a) for the influx of
polymorphonuclear leukocytes (PMN) was discussed, but it was also shown that the release
of cytokines (IL-1, TNF, fibronectin) is not necessarily the same after in vivo or in vitro
particle exposure of alveolar macrophages. Although the in vitro responses of cytokine
release appear to be lower, alveolar macrophages can be primed by 7-interferon to release
TNF also in vitro. Thus, important mechanistic interactions with respect to cytokine release
are occurring and need further study. Research needs with respect to mechanisms of
particle-induced inflammation were outlined, including development of biological markers;
influence of the pulmonary extracellular environment in the lung's response to particles; the
role of particle surface, mass, volume, and numbers in the inflammatory response;
contribution of other macrophage populations to particle responses; species differences in
responses; the role of macrophage interaction with other cells, e.g., structural cells;
contribution of fibroblasts, epithelial cells, and endothelial cells, as initial effectors of
response since these cells in turn can, via an intricate cytokine networking, interfere with
each other in the overall inflammatory response.
3.4. Biology of Macrophages, Bruce Lehnert
An overview of the biology of macrophages was presented. Heterogeneity of the
alveolar macrophages in different lung regions was discussed. The possibility was brought
up that this heterogeneity could indicate development stages.
Macrophage-particle interactions in different compartments of the lung, including the
redistribution of phagocytized particles within the macrophages of the alveolar space over
time, were reviewed. As a result of either cell death or cell division, the particle load per
-------
macrophage decreases following cessation of exposure. Presumably fewer of the cells are
then overloaded. In spite of a decrease in the alveolar macrophage load, impaired clearance
was still observed. The underlying mechanisms are still unclear. In this context, the
importance of the pores of Kohn for alveolar clearance was mentioned, as preliminary
evidence has suggested that these pores may be grown over by epithelial cells when lung
burdens are excessive. A pathway is therefore occluded that might otherwise be important
for clearance of macrophages.
Research needs were outlined with respect to basic pulmonary macrophage biology
and the interaction of pulmonary macrophages and particles. Needs include the following:
Origins, translocation pathways and transit times, proliferative characteristics,
and lifetimes of pulmonary subpopulations.
Factors controlling the steady-state regulation of the sizes of the pulmonary
macrophage subpopulations.
Bases of phenotype and functional heterogeneity and changes following particle
deposition.
Identification of macrophage-derived growth factors, their target cells, and
cytokine networks.
Influence of prior phagocyte history on subsequent chemotactic responsiveness
and phagocyte function; in vivo factors that regulate macrophage-particle
encounters and rates of phagocytosis.
Mechanism(s) by which alveolar macrophages encounter and become coupled
to the mucociliary apparatus for tracheobronchial transport.
Roles of the airway intraluminal macrophages in particle clearance and airway
disease.
Factors that govern the fate of free particles that enter the lung's interstitial
compartments.
Functional characteristics and translocation/clearance pathways of pulmonary
interstitial macrophages.
-------
Mechanism(s) responsible for the formation of aggregates of alveolar
macrophages in the particle overload condition.
Mechanisms underlying the "particle redistribution phenomenon" and how they
are affected by differing lung burdens, particles of differing size, and particle
cytotoxicity.
Comparative studies of human alveolar macrophages and alveolar macrophages
from other laboratory species as to their relative abilities to extract/metabolize
procarcinogens and carcinogens from particles and the effect of particle load
on this process.
Characterizations of the abilities of other endocyte cell types, including
polymorphonuclear leukocytes, interstitial macrophages, and type I epithelial
cells to extract/metabolize procarcinogens and carcinogens from engulfed
particles.
Kinetics of particle endocytosis by airway/alveolar epithelial cells, and how
particle size, numbers, and surface characteristics affect the endocytic process,
and the subsequent transcellular passage of particles into subepithelial
compartments.
Measure of reactive oxygen species, proinflammatory mediators, and
promitogenic cytokines as a result of cell particle interactions.
3.5. Clearance of Inhaled Particles, Timothy Gerrity
The first point made was a reminder of the tremendous species differences in
retention half-times of highly insoluble particles deposited in the lungs. Underlying
mechanisms for these differences in clearance from the alveolar region require further
evaluation. It was also emphasized that retention of particles in the conducting airways may
not be as short as generally thought, since newer studies seem to indicate that there is a long-
term component involved, which leads to a significantly longer retention of particles in this
region. Research needs include the study of very basic mechanisms of macrophage-mediated
particle clearance, including the basic question—why do macrophages move toward the
mucociliary ladder for effective clearance of phagocytized material? With respect to
overload-induced retardation of clearance, we need to find out whether this is an irreversible
event or to what degree recovery does occur.
-------
3.6. Dosimetry Models, Chia Ping Yu
The deposition and clearance behavior of particles in humans was reviewed. It was
pointed out that the relative total deposition of inhaled particles in humans of different ages is
very similar. Across species, deposition in different airway regions can differ significantly.
For example, deposition in the head region is significantly different in different animal
species (e.g., rats vs. guinea pigs vs. man). With respect to research needs in the area of
particle deposition, more details are needed on airway geometry and local deposition. With
respect to particle clearance, the relationship between clearance rate and cell function needs
to be evaluated. Deposition and clearance models likewise need to be refined to include and
study regional nonuniformity of these events. With respect to dosimetry, models are needed
to include children and diseased lungs, and a larger intersubject variability also has to be
incorporated. Finally, the basis for an interspecies extrapolation method needs to be
developed.
3.7. Risk Assessment Methodology for Particles, Roy Albert
The complexity of the problem was stressed when dealing with animal data and
attempting to make an assessment of human risk. Furthermore, using possibly available data
from human occupational studies to extrapolate to the environmental situation is a difficult
undertaking, potentially involving orders of magnitude downward extrapolation. In rats, we
are almost exclusively dealing with bronchoalveolar tumors, whereas in humans the observed
tumors are more of bronchial origin. This represents an additional difficulty for using
animal studies to assess human risks and requires a more detailed analysis of the exposure
and dosimetry of airways and bronchi. The need to know more about the contribution of the
alveolar particle dose to bronchial cells during subsequent clearance in order to assess the
significance of bronchogenic carcinogenesis was also stressed. Recommendations for future
research included the evaluation of the involvement of oxygen radicals in adduct formation.
It was further suggested that use be made of human populations heavily exposed to particles,
such as coal miners; possibly, pulmonary lavage could be used to assess DNA adducts in
lavageable cells. Pulmonary tissues from animals exposed to different particles, such as
diesel exhaust, carbon black, and TiO2, could be used to examine more closely the
10
-------
relationship of particle exposure and adduct formation,, and alveolar macrophages could be
used to determine if a correlation exists between tissue adducts and macrophage adducts. In
vitro studies with alveolar macrophages to evaluate DNA adduct formation after different
particle exposures also would be useful in this respect. The basic question concerning the
usefulness of this marker needs to be addressed.
3.8. Additional Comments by the Workgroup
Several issues already presented were reemphasized, such as the necessity and
methodology of extrapolating to low-dose levels, the question of human risk of exposure to
particle "X," the biological plausibility of mechanisms,, and the question of whether there are
multiple mechanisms of action, some occurring at high levels as opposed to others operating
at low levels of exposure. Extrapolation from laboratory animal data to man requires the
incorporation of both extrapolation across materials, i.e., carbon black, TiO2, diesel exhaust,
as well as across exposure levels when multiple mechanisms of action may take place, i.e., at
high exposure levels and others operating at low exposure levels. Exposure-dose-response
relationships for a number of studies are needed, including in vitro cell/tissue culture studies,
in vivo short-term studies, and in vivo long-term studies, to address the issues of high to low
exposure, and possibly even of in vitro studies of human exposure. It might be useful to
search for human data that may serve as a "gold standard" with which animal data could be
compared (diesel studies, coal dust?). Other questions to be addressed relate to whether
nonmutagenic materials have a threshold and what the basis is for a threshold generally
believed to be present for mutagenic materials. In any type of research, dose-response
relationships are important components, where the dose might be represented as particle
burden or formation of adducts and the response in terms of tumors, mutational events, DN\
damage, etc. Possible mechanisms for particle-induced lung injury include direct and
indirect mutational events, cell proliferative responses, oxidative damage, and others. These
mechanistic aspects were discussed with respect to the needs outlined in Figure 1 and
detailed in the first part of the summary report.
11
-------
I
•SJ
O
5
l
I-
Q.
X
111
C
3
I
.a
I
I
o
"S
•a
C
I
<-t-<
(D
O
1
f
S
I
I
13
o
C
I
i
ffi
£>JO
12
-------
4. WORKSHOP RECOMMENDATIONS
4.1. General Outline
As a result of these discussions, a scheme was developed during the last session of the
workshop, which outlined a number of mechanisms of particle-induced effects in the lung
and which served as a basis for discussing and recommending research needs. Figure 1
depicts this scheme, which tries to incorporate in a very general way major mechanistic
pathways that may lead to particle-induced lung tumors. This scheme does not imply that all
of the indicated intermediate steps occur with any particle, but rather it demonstrates some of
the different interactions in the lung that may occur with highly insoluble particles. Although
the group discussion in this last session focused exclusively on the issue of mechanisms of
particle-induced tumorigenesis (see Figure 1), also discussed was the possibility that these
tumors are the result of several intermediate events involving chronic inflammatory,
proliferative, and possibly fibrotic reactions in the lung.. Although fibrosis is a dominant and
equally serious endpoint of particle-induced pulmonary toxicity, the workshop dealt primarily
with tumorigenicity of insoluble particles. However, the interplay of several features of
Figure 1 may be subject to incorporation into a comparable schema for pulmonary fibrosis.
The major pathways and events in Figure 1 are as follows: Particles deposited in the
lung will mostly be phagocytized by alveolar macrophages, which in turn may be activated to
release mediators including reactive oxygen species, cytokines, and growth factors.
Inflammatory cells such as PMN elicited into the lung also will contribute to the release of
mediators including reactive oxygen species. Reactive oxygen species could interact with
DNA of target cells and induce specific DNA adducts such as 8-hydroxyguanine adducts,
which may in turn lead to mutations and genetic changes, including activation of oncogenes
or inactivation of suppressor genes, which may eventually result in an increased tumor
incidence. Particles also may interact directly with other cells in the lung, such as epithelial
cells and fibroblasts, which may then be stimulated to release mediators. One expected
outcome of particle-cell interactions is fibrosis. Particles also may have a physical direct
interaction during cell division, which, for example, has been demonstrated for asbestos
13
-------
fibers interfering with the spindle apparatus of dividing cells, and may also lead to mutational
events.
An important pathway derived from the interaction of particles with macrophages and
other cell types involves the release of growth factors and cytokines, which leads to an
intricate network of cell-cell interactions resulting in stimuli for proliferative responses that,
as indicated in the scheme, can contribute to the manifestation of tumors at several crucial
points. Specific particles such as diesel soot containing organic materials also may lead to
direct organic-specific adduct formations, which are different from those induced by reactive
oxygen species. The importance of these adducts in particle-induced lung tumor formation,
however, needs further investigation. A conclusive pathogenic role could not be established
for organic-induced DNA adducts or oxygen radical-induced DNA adducts.
At several points in the scheme of Figure 1, physiological defense mechanisms are
indicated that may affect lung injury and reduce tumor formation. For example, in the first
case, cell-particle interactions can lead to elaboration of many factors that affect proliferation
and are important in long-term effects including fibrosis. In the latter case, antioxidant
systems are present in most cell systems in the lung; antimitogenic factors in addition to
promitogenic factors also can significantly alter the proliferative processes occurring at
different levels during pathogenesis of tumor formation. Tumor defenses are also present,
including natural killer cell activity released from alveolar macrophages. All these could
become effective in this general scheme of particle-induced lung injury and tumor formation.
It was generally agreed that we do not need more studies just to demonstrate that inhaled
particles can induce lung tumors, but that research should focus on mechanistic links. This
does not preclude the need for bioassays of additional materials to determine their
carcinogenic potency.
4.2. Specific Recommendations
Several research recommendations were made aimed at filling knowledge gaps in our
understanding of the different pathways outlined in the figure. Research may start at
14
-------
different points of the scheme, involving a multidisciplinary approach. While it is
recognized that particle-adsorbed organics may be important for certain particulate
compounds (diesel), their contribution to tumor development via specific adduct formation
and mutational events might only be minimal and was thought, therefore, not to be as
important in contrast to other more generic effects of particles. Ongoing research in this
area by several groups should clarify the significance of particle-adsorbed organics, but this
should not be high on EPA's priority list of research needs for particle-induced tumors.
Research involving insoluble particulate matter such as silica and freshly fractured coal dust,
which are thought to have additional toxic properties due to the presence of oxygen radicals
on the particle surfaces, also should be lower on EPA's priority list of research needs for
inert particles. Mechanistic studies pertaining to the surface-associated radicals, however,
may aid in our overall understanding of particle-induced toxicity.
The pathway in Figure 1, starting with direct particle-macrophage interactions,
includes the most important events, and the group agreed on several recommendations to
investigate specific mechanisms along this pathway.
4.2.1. Particles Are Promoters
Particles deposited in the lung may have a promoting effect that can result in
increased proliferative responses. Research in this area should include initiation-promotion
studies in which an initiating pulmonary carcinogen would be combined with administration
of particles in the lung. However, since target cells may be different for the initiating agent
on one hand and particles on the other (particles so far have induced peripheral lung tumors
in rats whereas other initiators may target more central lung regions), one must be careful in
interpreting observed effects. Study of proliferative responses induced by particles for
different cell types in the lung would be of high importance in these studies.
4.2.2. Differentiation of Adducts
DNA adduct formation has been observed in the lung after inhalation of diesel exhaust
(particles with adsorbed organic compounds) and after inhalation of carbon black (three
15
-------
orders of magnitude less organic content). However, whereas inhaled diesel exhaust particles
induced organic-specific adducts, preliminary evidence appears to suggest that inhalation of
carbon black as well as of TiO2 (ultrafme particle size) led to the formation of polar adducts
(oxygen radical induced?), which have yet to be specified. Whether, indeed, diesel exhaust
and carbon black induce different adducts needs to be verified. Since the tumor response
was not different between the two particle types, the role of adduct formation for particle
tumorigenesis needs to be established. Studies are needed to investigate the importance of
the adducts and correlation of the adduct formation with the particle-induced tumors. The
specificity of adducts in different regions of the lung and for different particle types needs to
be studied. However, results on DNA adduct formation should be interpreted very
cautiously, in particular with respect to linking them to specific mutations. DNA adduct
formation also has been found in alveolar macrophages of humans who have been exposed to
coal or smoke particles, and the question is whether and how these adducts, which could
conveniently be determined from lavage samples, might be correlated to adducts induced in
other cell types and regions of the lung.
4.2.3. Mutation Spectra
Mutational changes may occur at different gene locations, and nothing is known as to
whether specific mutations are related to specific particle exposures or not. Thus, analysis of
specific gene mutations needs to be performed in studies using different particle types and
comparing those to spontaneously occurring mutations. The fact that chronic inhalation of
particles at high concentrations can lead to lung tumors in rats implies that mutations have
occurred. However, it needs to be investigated whether these are just spontaneous mutations
that have become "fixed" due to a high proliferative response in the lung or whether these
are new mutations with a very different spectrum.
4.2.4. Dose-Time Response for Radical Production by Macrophages
Oxygen-derived radicals may play an important role both for the formation of adducts
(see paragraph 4.2.2.) and for causing cell injury leading to cell death. The lack of
knowledge of dose-response curves and of the time course of production of oxygen-derived
16
-------
radicals from inflammatory cells and other cells was seen as an area warranting further
research.
4.2.5. Dose-Time Response for Cell Proliferation
Proliferative events are important aspects in the scheme outlined in Figure 1. There
is an urgent need to know more about such proliferative responses induced by inhaled
particles in an exposure/dose-dependent relationship pertinent to fibrosis and tumor
development. Although proliferative lesions have been observed in essentially all of the
chronic high-exposure particle inhalation studies in rats, no data to quantitate these responses
have been reported pertinent to fibrosis and tumor development. It would be of particular
importance to obtain such information in response to particle exposure during as well as after
cessation of such exposures.
4.2.6. Particle-Macrophage-Cytokine-Mutation Links
This is perhaps the most important area for research needs-combining the particle-
induced inflammatory responses of alveolar macrophages to the release of cytokines/growth
factors and their linkages to both mutational events and fibrotic changes (upper left portion of
Figure 1). Understanding the mechanisms that eventually lead to lung injury and mutations
after the initial macrophage-particle encounter will give a better basis and justification for
extrapolating results of animal studies to humans. Knowledge about the intricate networking
of released cytokines and growth factors, not only from alveolar and interstitial macrophages
but also from other cell types such as type 1 and type 2 epithelial cells and fibroblasts, is
only slowly emerging now. We need more data related specifically to particle effects in this
respect to better understand interactions of pro- and antimitogenic events in the lung and their
implications for the fixation of mutations. The inflammatory response resulting from the
activation of macrophages and the subsequent release of mediators from these cells and other
cells is of crucial importance in the initial phase because this may eventually lead to the
formation of tumors. Cellular proliferative responses as a consequence of inflammatory
events were considered an important area requiring further research.
17
-------
4.2.7. Examination of Archival Tissue
Techniques have been developed and are being developed further to use fixed tissues
from previous animal experiments or from autopsy material for a detailed analysis of some of
the mechanistic events outlined in Figure 1. For example, in situ hybridization techniques
are available to determine whether and which types of growth factors or cytokines might
have been involved in a specific response. Methodologies to examine cell proliferation
events using proliferating cell nuclear antigens (PCNA) are being refined for use in archival
tissues to quantitate proliferative responses of specific cell types after particle exposure.
Therefore, archival tissues are an important resource for further investigations, thereby
reducing the need for some new and costly experiments.
4.2.8. Particle-Associated Compromises in Tumor Defense Mechanisms
It is unknown whether excessive particle burdens actually initiate or promote the
emergence of tumors or whether transformed cells, which may normally occur otherwise,
successfully progress because of a failure in tumor surveillance and killing activities by cells
such as macrophages. The "tumor defenses" shown in Figure 1 refer to this mechanism.
Many of the research needs outlined in this section can and should make use of both
in vivo and in vitro studies. In vitro studies also should incorporate the use of human
alveolar macrophages and possibly other human pulmonary cells and compare their responses
with those of the respective cells of rats and other animal species; that is, studying species-
specific responses is very important. The influence of culture conditions must be
appropriately assessed. It was also emphasized that experimental studies using particles
should be performed over a wide range of doses. There was a consensus that the research
outlined above must include studies in different species, which might uncover important
mechanistic differences in particle-induced lung tumors.
The different pathways proposed for carcinogenesis converge at the point of mutations
in Figure 1. Thus, it was suggested that the basic underlying questions can be summarized
in a "convergence" approach by examining (1) whether particles cause increased or different
18
-------
mutations and (2) whether particle-induced changes in "'cancer genes" are different from
those occurring spontaneously.
Although prioritization of the research recommendations was not attempted at the
workshop, there appeared to be a general consensus that inflammatory and subsequent
cellular proliferative responses may be the most important events leading to the fixation of
induced mutations and finally particle-induced lung tumors.
19
-------
5. RESEARCH PLANS RESULTING FROM WORKSHOP RECOMMENDATIONS
5.1. Goals of the Research
EPA funding to support research in this area is limited. To provide answers to some
of the most pressing questions, an experiment was designed that will be carried out with the
aid and cooperation of other government and private institutions. The study is designed to
focus on the following issues:
1. Species differences in responsiveness to inhaled particulate matter.
2. The relationship between lung particle burdens and release of mediators by
pulmonary macrophages. Is there a threshold?
3. Relationship between the output of mediators and pathological responses in the
lungs.
4. Relationship between the output of mediators and cell proliferation.
5. Relationship between the lung burden of particulate matter, production of
mediators, and formation of DNA adducts.
5.2. Work Plan
Animals will be exposed for periods of up to 90 days to several concentrations of
carbon black. Additional groups exposed to cristobalite and amorphous ultrafine silica will
serve as positive controls along with one group of sham-exposed animals.
Fischer 344 rats will be used. Groups of animals will be sacrificed after 45 and 90
days of exposure as well as 6 months and 12 months post exposure. Exposure durations of 7
to 8 hours per day, 5 days per week, are planned. Whole animal exposure chambers will be
used.
20
-------
The following endpoints will be examined:
• Particle lung burdens
• Analysis of bronchoalveolar lavage fluid
• Cell proliferation assays
• Growth factor production by macrophages
• Cytokines and chemotactic mRNA
• Oxidant-induced DNA adducts in lung epithelium
• Mutation assay in alveolar epithelial cells
• Lung histology-morphology
5.3. Tune Schedule
These studies are planned to begin during the third quarter of 1993 and are expected
to be completed near the end of 1994.
5.4. Utilization of Research Data
The EPA and other regulatory agencies are charged with assessing health risk effects
of a variety of types of inhalable particulate matter. The EPA, for example, is developing
unit risk estimates for diesel exhaust. Although a large number of organic compounds,
including some that are carcinogenic, are adsorbed to the surface of diesel exhaust particles,
available evidence, nevertheless, indicates that pathological and carcinogenic effects of diesel
exhaust are primarily a function of the insoluble core of the diesel particle. Certain
manmade mineral fibers, which are insoluble as well as biochemically inert, also have been
shown to be toxic and carcinogenic. Insoluble, biochemically inert polymers are regulated
by EPA's Office of Prevention, Pesticides, and Toxic Substances. Other insoluble particulate
matter of concern to EPA and other regulatory agencies includes silica and coal dust.
Although the latter two agents have additional toxic properties due to the presence of oxygen
radicals on the particle surface, particle effects may still contribute to responses. The
research data that are obtained are expected to be used to develop updated risk estimates of
exposure to particulate matter agents such as the ones mentioned above.
21
-------
The results of planned research should allow the development of mechanistically
based dose-response models, in the hope that a major uncertainty in quantitatively assessing
risk of exposure to particles, the shape of the low-dose extrapolation curve, will be
eliminated or at least decreased. It may be possible to partition the relative effects of
particles, surface-adsorbed organics, particle-associated oxygen radicals, or vapor phase
compounds. The ultimate goal in collecting these data is to reduce uncertainty in assessing
risk of exposure to the agents listed above.
22
-------
LIST OF PANELISTS
Roy Albert
University of Cincinnati Medical Center
3223 Eden Avenue
Cincinnati, OH 45221
James Bond
Chemical Industry Institute of Toxicology
P.O. Box 12137
Research Triangle Park, NC 27709
Daniel Costa
U.S. Environmental Protection Agency
MD-82
Research Triangle Park, NC 27711
Bruce Lehnert
Group LS-1 (MS-M888)
Los Alamos National Laboratory
Los Alamos, NM 87545
Joellen Lewtas
U.S. Environmental Protection Agency
MD-68A
Research Triangle Park, NC 27711
Joe Mauderly
Lovelace Inhalation Toxicology Research Institute
P.O. Box 589
Albuquerque, NM 87185
Roger McClellan
Chemical Industry Institute of Toxicology
P.O. Box 12137
6 Davis Drive
Research Triangle Park, NC 27709
Paul Morrow
University of Rochester Medical Center
Box BPHYS
Rochester, NY 14642
23
-------
Kathleen M. Nauss
Health Effects Institute
141 Portland Street
Cambridge, MA 02139
Gunter Oberdorster
University of Rochester Medical Center
Box EHSC
Rochester, NY 14642
Andrew Sivak
Health Effects Institute
141 Portland Street
Cambridge, MA 02139
Werner Stober
Chemical Industry Institute of Toxicology
P.O. Box 12137
6 Davis Drive
Research Triangle Park, NC 27709
Ronald Wolff
Lilly Research Laboratories
P.O. Box 708
Greenfield, IN 46140
Chia Ping Yu
Department of Mechanics and Aerospace Engineering
State University of New York at Buffalo
314 Jarvis Hall
Amherst, NY 14642
24
v.s. oommsxr muriNC OFFICE: 1993—750-002 / sozsi
------- |