EPA # 9345.4-06
October 2003
FINAL DRAFT:
TECHNICAL SUPPORT DOCUMENT FOR A
PROTOCOL TO ASSESS ASBESTOS-RELATED
RISK
Prepared for:
Office of Solid Waste and Emergency Response
U.S. Environmental Protection Agency
Washington, DC 20460
-------
NOTICE
This document provides guidance to EPA staff. It also provides guidance to the public and to the
regulated community on how EPA intends to exercise its discretion in implementing the National
Contingency Plan. The guidance is designed to implement national policy on these issues. The
document does not, however, substitute for EPA's statutes or regulations, nor is it a regulation
itself. Thus, it cannot impose legally-binding requirements on EPA, States, or the regulated
community, and may not apply to a particular situation based upon the circumstances. EPA may
change this guidance in the future, as appropriate.
-------
U.S. Environmental Protection Agency
Authors
D. Wayne Berman
Aeolus, Inc.
751TaftSt.
Albany, CA 94706
Kenny S. Crump
Environ Corporation
602 East Georgia Ave.
Ruston, LA 71270
Contributors/Reviewers
Affiliation: McGill University
Name: Bruce Case, Associate Professor
Location: 462 Argyle Avenue Westmount, Quebec
H3Y 3B4 Canada
Phone: 514-398-7192 #00521
Fax: 514-398-7446
Email: bruce.case@mcgiIl.ca
Affiliation: Pathology & Physiology Research Branch
National Institute for Occupational Safety & Health
Name: Vincent Castranova, Chief
Location: 1095 Willowdale Road (L 2015)
Morgantown, WV 26505
Phone: 304-285-6056
Fax: 304-285-5938
Email: vicl@cdc.gov
Affiliation: Department of Medicine National Jewish
Medical Research Center
Name: James Crapo, Chairman
Location: 1400 Jackson Street Denver, CO 80206
Phone: 303-398-1436
Fax: 303-270-2243
Email: crapoi@nic.org
Affiliation: Medical University of South Carolina
Name: David Hoel, Professor
Location: 36 South Battery Charleston, SC 29401
Phone: 843-723-1155
Fax: 843-723-7405
Email: whitepoint@aol.com
Affiliation: New York University School of Medicine
Name: Morton Lippmann, Professor
Location: 57 Old Forge Road Tuxedo, NY 10987
Phone: 845-731-3558
Fax: 845-351-5472
Email: liDpmann@env.med.nvu.edu
Affiliation: Toxicology & Human Health Risk
Analysis
Name: Roger McClellan, Advisor
Location: 13701 Quaking Aspen Place, NE
Albuquerque, NM 87111
Phone: 505-296-7083
Fax: 505-296-9573
Email: roeer.o.mcclellan@att.net
Affiliation: Price Associates, Inc.
Name: Bertram Price
Location: 1 North Broadway - #406 White Plains, NY
10601
Phone: 914-686-7975
Fax: 914-686-7977
Email: bprice@priceassociatesinc.com
Affiliation: California Environmental Protection Agency
Name: Claire Sherman, Biostatistician
Location: 1515 Clay Street, 16th Floor Oakland, CA
94612
Phone: 510-622-3214
Fax: 510-622-3211
Email: csherman@ochha.ca.gov
Affiliation: Risk Evaluation Branch National Institute for
Occupational Safety & Health
Name: Leslie Thomas Stayner, Chief
Location: Robert Taft Laboratories, C15 4676 Columbia
Parkway Cincinnati, OH 45226
Phone: 513-533-8365
Fax: 513-533-8224
Email: Its2@cdc.gov
Affiliation: Rollins School of Public Health, Emory
University
Name: Kyle Steenland, Professor
Location: 1518 Clifton Road Atlanta, GA 30322
Phone: 404-712-8277
Email: nsteenl@sph.emorv.edu
Affiliation: Exponent, Inc.
Name: Mary Jane Teta, Principal Epidemiologist
Location: 234 Old Woodbury Road Southbury, CT
06488
Phone: 203-262-6441
Fax: 203-262-6443
Email: iteta@exponent.com
-------
ACKNOWLEDGMENTS
We wish to acknowledge the guidance and support for this project provided by Aparna
Koppikar, Richard Troast, Chris Weis, and Paul Peronard (all of U.S. EPA).
We wish to gratefully acknowledge the assistance of Nick de Klerk, FDK Liddell, J. Corbett
McDonald, Terri Schnoor, and John Dement for graciously providing the raw epidemiology data
from several key studies, without which we could not have completed our analysis.
We also wish to thank John Addison, John Dement, Agnes Kane, Bruce Case, Michel Camus,
and Vanessa Vu for their input and assistance.
Finally, we wish to acknowledge Eric Hack and Tammie Covington for their assistance with
calculations and data management and Mark Follansbee and Joanne White for their assistance
with technical editing and managing production of the document.
IV
-------
ACRONYMS
AM alveolar macrophages
AP alkaline phophatase
ATSDR American Toxic Substances
Disease Registry
BAL bronchio-alveolar lavage
BrdU bromodeoxyuridine
CalEPA California Environmental
Protection Agency
CFE colony-forming efficiency
CHO Chinese hamster ovary
EDXA energy dispersive X-ray analysis
EGFR Epithelial Growth Factor
Receptor
ERK EGFR-regulated kinase
ESR electron spin residence
FBP's fibrin breakdown products
HAF human amniotic fluid
HBE human bronchiolar epithelial
HNE human neutrophil elastase
HTE hamster tracheal epithelial
IPF idiopathic pulmonary fibrosis
IRIS Integrated Risk Information
System
KGF kertinocyte growth factor
LDH lipid dehydrogenase
LPS lipopolysaccharide
MAPK mitogen activated protein kinase
MI midget impinger
MLE maximum likelihood estimate
MMVF's man-made vitreous fibers
MnSOD manganese-containing
superoxide
mpcf millions of particles per cubic
foot
mppcf millions of dust particles per
cubic foot
MSHA Mine Safety and Health
Administration
bronchioepithelial
NHIS National Health Interview
Survey
NIOSH National Institute for
Occupational Safety and Health
OSHA Occupational Safety and Health
Agency
PARS poly-ADP-ribosyl transferase
PCM phase contrast microscopy
PE pulmonary epithelial
PMN polymorphonucleocyte
RBC's red blood cells
RCF refractory ceramic fiber
RNS reactive nitrogen species
ROS reactive oxygen species
RPM rat pleural mesothelial
RR relative risk
SAED selected area electron diffraction
SEM scanning electron microscopy
SHE Syrian Hamster Embryo
SMG small mucous granule
SMRs standardized mortality ratios
THE tracheal epithelial cells
TEM transmission electron
microscopy
TGF-P transforming growth factor beta
TNF-oc tumor necrosis factor alpha
uPA urokinase-type plasminogen
activator
uPAR urokinase-type plasminogen
activator
U.S. EPA U.S. Environmental Protection
Agency
VCAM-1 vascular cell adhesion molecule
NAC N-acetylcysteine
NHBE normal human
-------
TABLE OF CONTENTS
1.0 EXECUTIVE SUMMARY 1.1
2.0 INTRODUCTION 2.1
3.0 OVERVIEW 3.1
4.0 BACKGROUND 4.1
4.1 THE MINERALOGY OF ASBESTOS 4.2
4.2 MORPHOLOGY OF ASBESTOS DUSTS 4.3
4.3 CAPABILITIES OF ANALYTICAL TECHNIQUES USED TO MONITOR
ASBESTOS 4.4
4.4 THE STRUCTURE AND FUNCTION OF THE HUMAN LUNG 4.12
4.4.1 Lung Structure 4.12
4.4.2 The Structure of the Mesothelium 4.16
4.4.3 Cytology , 4.17
4.4.4 Implications 4.18
5.0 THE ASBESTOS LITERATURE 5.1
5.1 HUMAN EPIDEMIOLOGY STUDIES 5.2
5.2 HUMAN PATHOLOGY STUDIES 5.6
5.3 ANIMAL STUDIES 5.8
5.4 IN VITRO STUDIES 5.9
6.0 SUPPORTING EXPERIMENTAL STUDIES 6.1
6.1 FACTORS AFFECTING RESPIRABILITY AND DEPOSITION 6.2
6.1.1 Respirability of Spherical Particles 6.4
6.1.2 Respirability of Fibrous Structures 6.5
6.1.3 The Effects of Electrostatic Charge on Particle Respirability 6.10
6.1.4 General Conclusions Concerning Particle Respirability 6.11
6.2 FACTORS AFFECTING DEGRADATION, TRANSLOCATION, AND
CLEARANCE 6.12
6.2.1 Animal Retention Studies 6.17
6.2.1.1 Studies involving short-term exposures 6.18
6.2.1.2 Studies involving chronic or sub-chronic exposures 6.28
6.2.2 Animal Histopathological Studies 6.36
6.2.3 Human Pathology Studies 6.41
6.2.4 Studies of Dissolution/BioDurability 6.46
6.2.5 Dynamic Models 6.50
6.2.6 General Conclusions Regarding Deposition, Translocation, and
Clearance 6.53
6.3 FACTORS GOVERNING CELLULAR AND TISSUE RESPONSE 6.59
6.3.1 The Current Cancer Model 6.64
6.3.2 Evidence for Transformation 6.66
6.3.3 Evidence that Asbestos Acts As a Cancer Initiator 6.79
6.3.3.1 Interference with mitosis 6.79
VI
-------
6.3.3.2 Generation of reactive oxygen species (ROS) 6.82
6.3.3.3 Generation of reactive nitrogen species (RNS) 6.89
6.3.3.4 Conclusions concerning asbestos as a cancer initiator ... 6.91
6.3.4 Evidence that Asbestos Acts As a Cancer Promoter 6.92
6.3.4.1 Asbestos-induced proliferation 6.93
6.3.4.2 Asbestos induced cell signaling 6.97
6.3.4.3 Asbestos-induced apoptosis 6.102
6.3.4.4 Asbestos-induced cytotoxicity 6.104
6.3.4.5 Association between fibrosis and carcinogenicity 6.106
6.3.4.6 Interaction between asbestos and smoking 6.107
6.3.4.7 Conclusions concerning asbestos as a promoter 6.108
6.3.5 Evidence that Asbestos Induces an Inflammatory Response 6.109
6.3.6 Evidence that Asbestos Induces Fibrosis 6.109
6.3.7 Evidence that Asbestos Mediates Changes in Epithelial Permeability 6.109
6.3.8 Conclusions Regarding the Biochemical Mechanisms of Asbestos-Related
Diseases 6.109
6.4 ANIMAL DOSE RESPONSE STUDIES 6.111
6.4.1 Injection-Implantation Studies 6.111
6.4.2 Animal Inhalation Studies 6.115
6.4.3 Supplemental Inhalation Study 6.117
6.4.4 Conclusions Concerning Animal Dose-Response Studies 6.126
6.5 CONCLUSIONS FROM AN EVALUATION OF SUPPORTING
STUDIES 6.127
7.0. EPIDEMIOLOGY STUDIES 7.1
7.1 APPROACH FOR EVALUATING THE EPIDEMIOLOGY LITERATURE . 7.1
7.2 LUNGCANCER 7.4
7.2.1 The Adequacy of the Current U.S. EPA Model for Lung Cancer 7.4
7.2.1.1 Exposure Dependence 7.5
7.2.1.2 Time Dependence 7.8
7.2.1.3 Smoking-Asbestos Interaction With Respect to Lung
Cancer 7.12
7.2.1.4 Conclusions Concerning the Adequacy of the U. S. EPA Lung
Cancer Model 7.14
7.2.2 Estimating KL values from the Published Epidemiology Studies .... 7.15
7.3 MESOTHELIOMA 7.23
7.3.1 The Adequacy of the Current U.S. EPA Model for Mesothelioma ... 7.23
7.3.1.1 Time Dependence 7.28
7.3.1.2 Exposure Dependence 7.28
7.3.1.3 Discussion of Adequacy of Mesothelioma Model 7.31
7.3.2 Estimating KM Values from Published Epidemiology Studies 7.33
7.4 EVALUATION OF ASBESTOS EXPOSURE INDICES 7.35
7.4.1 Fiber Type and Size Distribution Data Available for Deriving Exposure
Indices 7.36
7.4.2 Modification of Existing KL and KM to Conform to a New Exposure
Index 7.38
7.4.3 Derivation of an Improved Exposure Index for Asbestos 7.42
7.4.3.1 Optimizing the Exposure Index for Lung Cancer 7.43
vii
-------
7.4.3.2 Optimizing the Exposure Index for Mesothelioma 7.48
7.5 THE OPTIMAL EXPOSURE INDEX 7.50
7.5.1 Definition of the Optimal Index and the Corresponding Exposure-
Response Factors 7.50
7.5.2 Evaluation of the Optimal Exposure Index 7.50
7.6 GENERAL CONCLUSIONS FROM QUANTITATIVE ANALYSIS OF
HUMAN EPIDEMIOLOGY STUDIES 7.59
8.0 DISCUSSION, CONCLUSIONS, AND RECOMMENDATIONS 8.1
8.1 DISCUSSION AND CONCLUSIONS 8.1
8.1.1 Addressing Issues 8.1
8.1.2 Comparison with Other, Recent Risk Reviews 8.5
8.2 RECOMMENDATIONS FOR ASSESSING ASBESTOS-RELATED
RISKS 8.7
8.3 RECOMMENDATIONS FOR FURTHER STUDY 8.13
9.0 REFERENCES 9.1
APPENDIX A: UPDATE OF POTENCY FACTORS FOR LUNG CANCER (KL) AND
MESOTHELIOMA (KM) A.I
A.I LUNG CANCER MODEL A.I
A.2 MESOTHELIOMA MODEL A.2
A.3 STATISTICAL FITTING METHODS A.3
A.4 SELECTION OF A "BEST ESTIMATE" OF KL AND KM A.4
A.5 UNCERTAINTY IN KL AND KM A.4
A.5.1 The Factor Fl A.6
A.5.2 The Factor F2 A.7
A.5.3 The Factor F3 A.8
A.5.4 The Factor F4L for Lung Cancer and F4M for mesothelioma A.8
A.5.5 Combining Individual Uncertainty Factors into an Overall "Uncertainty
Range" A.8
A.6 ANALYSIS OF INDIVIDUAL EPIDEMIOLOGY STUDIES A.8
APPENDIX B: REPORT ON THE PEER CONSULTATION WORKSHOP TO DISCUSS A
PROPOSED PROTOCOL TO ASSESS ASBESTOS-RELATED RISK ... B.I
APPENDIX C: COMPENDIUM OF MODEL FITS TO ANIMAL INHALATION DATA IN
SUPPORT OF THE BERMAN ET AL. (1995) STUDY AND POST-STUDY
WORK C.I
APPENDIX D: THE VARIATION IN KL VALUES DERIVED FOR CHRYSOTILE MINERS
AND CHRYSOTILE TEXTILE WORKERS D-l
APPENDIX E: CALCULATION OF LIFETIME RISKS OF DYING OF LUNG CANCER OR
MESOTHELIOMA FROM ASBESTOS EXPOSURE E-l
VIM
-------
LIST OF TABLES
Table 4-1. Capabilities and Limitations of Analytical Techniques Used for Asbestos
Measurements 4.5
Table 4-2. Comparison of Applicable Methods For Measuring Asbestos in Air 4.6
Table 6-1. Estimation of Lung Volume and Lung Surface Area Loading Rates for Rats and
Humans 6.10
Table 6-2. Relative Rates, Half-lives for Particles Cleared by the Varous Operating
Mechanisms of a Healthy Lung 6.13
Table 6-3. Fraction of Fibers Retained Following Chronic Exposure 6.31
Table 6-4. Measured in vitro Dissolution Rates for Various Fibers 6.48
Table 6-5. Putative Mechanisms by Which Asbestos May Interact with Lung Tissue to
Induce Disease Following Inhalation 6.60
Table 6-6a. Sources of Various Cytokines and Other Chemical Transmitters 6.67
Table 6-6b. Effects of Various Cytokines and Other Chemical Transmitters 6.74
Table 6-7. Summary Data for Animal Inhalation Experiments Conducted by Davis and
Coworkers 6.119
Table 7-1. Fit of EPA Lung Cancer Model to Observed Lung Cancer Mortality Among
Wittenoom, Australia Miners (Deklerk 2001) Categorized by Cumulative
Exposure Lagged 10 Years 7.6
Table 7-2. Fit of EPA Lung Cancer Model to Observed Lung Cancer Mortality Among
South Carolina Textile Workers (Schnoor 2001) Categorized by Cumulative
Exposure Lagged 10 Years 7.7
Table 7-3. Fit of EPA Lung Cancer Model to Observed Lung Cancer Mortality Among
Wittenoom, Australia Miners (Deklerk 2001) Categorized by Years Since Last
Exposure 7.9
Table 7-4. Fit of EPA Lung Cancer Model to Observed Lung Cancer Mortality Among
South Carolina Textile Workers (Schnoor 2001) Categorized by Years
Since Last Exposure 7.10
Table 7-5. Fit of EPA Lung Cancer Model to Observed Lung Cancer Mortality Among New
Jersey Factory Workers (Siedman et al. 1986) Categorized by Years Since First
Exposure 7.11
IX
-------
Table 7-6. Lung Cancer Exposure-Response Coefficients (KL) Derived from Various
Epidemiological Studies 7.17
Table 7-7. Fit of EPA Mesothelioma Model to Observed Mesothelioma Mortality Among
Wittenoom, Australia Miners (Deklerk 2001) Categorized by Years Since First
Exposure 7.24
Table 7-8. Fit of EPA Mesothelioma Model to Observed Mesothelioma Mortality Among
Wittenoom, Australia Miners (Deklerk 2001) Categorized by Average Value of
Integral (Equation 6-12) 7.25
Table 7-9. Mesothelimoa Exposure-Response Coefficients (KM) Derived from Various
Epidemiological Studies 7.27
Table 7-10. Fit of EPA Mesothelioma Model to Observed Mesothelioma Mortality Among
Wittenoom, Australia Miners (Deklerk 2001) Categorized by Years Since Last
Exposure 7.29
Table 7-11. Fit of EPA Mesothelioma Model to Observed Mesothelioma Mortality Among
Quebec Miners (Liddell 2001) in Each of Three Mining Areas, Categorized by
Years Since Last Exposure 7.30
Table 7.12. Comparison of Wittenoom, Australia (DeKlerk 2001) Mesothelioma Deaths to
Predicted Deaths Assuming Risk Varies Linearly with Exposure Intensity After
Controlling for Years Since First Exposure and Duration of Exposure 7.31
Table 7-13. Correlation Between Published Quantitative Epidemiology Studies and Available
Tern Fiber Size Distributions 7.37
Table 7-14. Representative KL and KM Values Paired with Averaged TEM Fiber Size
Distributions From Published Papers 7.39
Table 7-15. Estimated Uncertainty Assigned to Adjustment for Fiber Size 7.40
Table 7-16. Estimated Fraction of Amphiboles in Asbestos Dusts 7.44
Table 7-17. Results from Fitting Exposure Indices Defined by Equation 7.12 and Pcme to
Lung Cancer and Mesothelioma Exposure-response Coefficients Estimated from
Different Environments 7.47
Table 7-18. Optimized Dose-Response Coefficients for Pure Fiber Types 7.50
Table 7-19. Study Specific K,, K,,,, K,., Km., Kla, K^, Klc, and K^ Values 7.57
Table 7-20. Comparison of Spread in Range of Original and Adjusted K, and K,,, Values for
Specific Fiber Types 7.59
Table 8-1. Conservative Risk Coefficients for Pure Fiber Types 8.7
-------
Table 8-2. Estimated Additional Deaths from Lung Cancer or Mesothelioma per 100,000
Persons from Constant Lifetime Exposure to 0.0001 TEM f/cc Longer than 10 um
and Thinner than 0.4 um - Based on Optimum Risk Coefficients
(Table 7-18) 8.9
Table 8-3. Estimated Additional Deaths from Lung Cancer or Mesothelioma per 100,000
persons from Constant Lifetime Exposure to 0.0001 TEM f/cc Longer than 10 um
and Thinner than 0.4 um - Based on Conservative Risk Coefficients
(Table 8-1) 8.9
XI
-------
LIST OF FIGURES
Figure 4-1. The Structure of Lung Parenchyma Showing Alveoli and Alveolar Ducts ... 4.13
Figure 4-2. The Structure of the Inter-Alveolar Septa 4.14
Figure 4-3. The Detailed Structure of an Alveolar Wall that is Part of an Inter-Alveolar\
Septum 4.15
Figure 6-1. Fractions of Respirable Particles Deposited in the Various Compartments of the
Human Respiratory Tract as a Function of Aerodynamic Equivalent
Diameter 6.5
Figure 6-2. Fractions of Respirable Particles Deposited in the Various Compartments of the
Human Respiratory Tract as a Function of the True Diameter of Asbestos
Fibers 6.7
Figure 6-3. [Chrysotile/Lung Burden Concentration] vs. Time of Exposure 6.32
Figure 6-4. Key for Putative Mechanisms for Clearance and Translocation of Fibers in the
Lung 6.54
Figure 6-5. Fit of Model. Tumor Incidence vs. Structure Concentration by TEM (Length
Categories 5-40 urn, >40nm, Width Categories: <0.3 nm and >5 u.m) .... 6.122
Figure 6-6. Fit of Model. Tumor Incidence vs. Structure Concentration by TEM (Length
Categories 5-40 u,m, >40u.m, Width Categories: <0.4 uin) 6.125
Figure 7-1. Plot of Estimated KL Values and Associated Uncertainty Intervals by Study
Environment 7.20
Figure 7-2. Plot of Estimated KM Values and Associated Uncertainty Intervals by Study
Environment 7.34
Figure 7-3. Plot of Estimated (Adjusted) KL Values and Associated Uncertainty Intervals by
Study Environment 7.52
Figure 7-4. Plot of Estimated (Adjusted) KM Values and Associated Uncertainty Intervals by
Study Environment 7.53
Figure 7-5. Plot of Estimated KLA KLC Values and Associated Uncertainty Intervals by Study
Environment 7.54
Figure 7-6. Plot of Estimated KMA KMC Values and Associated Uncertainty Intervals by Study
Environment 7.55
XII
-------
1.0 EXECUTIVE SUMMARY
The purpose of this report is to provide a foundation for completing a state-of-the-art-protocol to
assess potential human-health risks associated with exposure to asbestos. Such a protocol is
intended specifically for use in performing risk assessments at Superfund sites, although it may
be applicable to a broad range of situations.
The current report is a revision to a version originally submitted on September 4, 2001 (Berman
and Crump 2001), which was the subject of a peer-review consultation held in San Francisco on
February 25-26, 2003. In general, the expert panel endorsed the overall approach to risk
assessment proposed in this report, although they highlighted areas where controversies persist.
The current report incorporates the changes recommended by the peer review consultation panel
to correct minor problems with internal consistency and the overall transparency of the
discussion that are needed to improve readability. Although some of the research and analyses
recommended by the peer consultation panel are not complete, it is anticipated that the current
document can be distributed for broader review and comment. Thus, the recommended approach
to risk assessment can be considered for use in the interim, while the additional research and
analyses recommended by the expert panel are completed. At that point, a final revision of this
document will be developed and it is expected to serve as a component of a broader effort by the
U.S. Environmental Protection Agency (U.S. EPA) to revise the Agency's current approach for
assessing asbestos-related risks.
The approach currently employed at the U.S. EPA to evaluate asbestos-related risks (IRIS 1988)
is based primarily on a document completed in 1986 (U.S. EPA 1986) and has not been changed
substantially in the past 15 years, despite substantial improvements in asbestos measurement
techniques and in the understanding of the manner in which asbestos exposure contributes to
disease. Therefore, this document provides an overview and evaluation of the more recent
studies and presents proposed modifications to the protocol for assessing asbestos-related risks
that can be justified based on the more recent work.
As reported in several recent technical meetings and reinforced by information gleaned from the
literature, the following were identified as issues that need to be addressed to develop a protocol
for evaluating asbestos-related risk:
• whether the exposure-response models currently in use by the U.S. EPA for
describing the incidence of asbestos-related diseases adequately reflect the time-
and exposure-dependence for the development of these diseases;
• whether different potencies need to be assigned to the different asbestos mineral
types to adequately predict risk for the disease endpoints of interest;
• to the extent that different asbestos mineral types are assigned distinct potencies,
whether the relative in vivo durability of different asbestos mineral types
determines their relative potency;
1.1
-------
• whether the set of minerals included in the current definition of asbestos
adequately covers the range of minerals that potentially contribute to asbestos-
related diseases;
• whether the analytical techniques and methods currently used for determining
asbestos concentrations adequately capture the biologically relevant
characteristics of asbestos (particularly with regard to the sizes of the structures
counted using the various analytical methods) so that they can be used to support
risk assessment; and
• whether reasonable confidence can be placed in the cross-study extrapolation of
exposure-response relationships that are required to assess asbestos-related risks
in new environments of interest.
These outstanding issues (and other related considerations) are addressed in this document to
provide a foundation for proposing a new approach for assessing asbestos-related risks.
Although the objective of this evaluation was to identify the single best procedure, when current
knowledge is inadequate for distinguishing among alternatives, options are presented along with
a discussion of their relative advantages and limitations. In a few cases, limited and focused
additional research studies are recommended, which may enhance the current state of knowledge
sufficiently to resolve one or more of the important, remaining issues.
Background
Inhalation of asbestos dusts has been linked to several adverse health effects including primarily
asbestosis, lung cancer, and mesothelioma (U.S. EPA 1986). Asbestosis, a chronic, degenerative
lung disease, has been documented among asbestos workers from a wide variety of industries.
Although asbestosis cases have been observed at some locations of current interest to the U.S.
EPA, the disease is generally expected to be associated only with the higher levels of exposure
commonly found in workplace settings and is not expected to contribute substantially to
potential risks associated with environmental asbestos exposure. Therefore, asbestosis is only
considered in this document to the extent required to address its putative association with lung
cancer. Overall, the majority of evidence indicates that lung cancer and mesothelioma are the
most important risks associated with exposure to low levels of asbestos.
The Asbestos Literature
A variety of human, animal, and tissue studies have provided insight into the nature of the
relationship between asbestos exposure and disease. Ideally, human epidemiology studies are
employed to determine the quantitative exposure-response relationships and the attendant risk
coefficients for asbestos exposure. Exposure-response coefficients have been estimated for
asbestos from approximately 20 epidemiology studies for which adequate exposure-response
data exist. Such coefficients vary widely, however, and the observed variation has not been
reconciled. Among the objectives of this study is to evaluate and account for the sources of
uncertainty that contribute to the variation among the exposure-response coefficients derived
from the literature so that these estimates can be reasonably interpreted and recommendations for
their use in risk assessment developed.
1.2
-------
Animal and tissue studies indicate that asbestos potency is a complex function of several
characteristics of asbestos dusts including fiber size and fiber type (i.e., fiber mineralogy).
Moreover, the influence of fiber size is a complex function of both diameter and length.
Therefore, whenever the goal is to compare across samples with differing characteristics, it is not
sufficient to report asbestos concentrations simply as a function of mass (or any other single
measure), which is in stark contrast to the treatment of chemical toxins. It has generally been
difficult to distinguish among the effects of fiber size and type in many studies because such
effects are confounded and the materials studied have not been adequately characterized.
The Epidemiology Studies
The existing epidemiology studies provide the most appropriate data from which to determine
the relationship between asbestos exposure and response in humans. As previously indicated,
however, due to a variety of methodological limitations, the ability to compare and contrast
results across studies needs to be evaluated to determine the confidence with which results from
existing epidemiology studies may be extrapolated to new environments where risk needs to be
assessed. This requires both that the uncertainties contributed by such methodological
limitations and that several ancillary issues be addressed.
Briefly, the major kinds of limitations that potentially contribute to uncertainty in the available
epidemiology studies include:
• limitations in air measurements and other data available for characterizing
historical exposures;
• limitations in the manner that the character of exposure (i.e., the mineralogical
types of fibers and the range and distribution of fiber dimensions) was delineated;
• limitations in the accuracy of mortality determinations or incompleteness in the
extent of tracing of cohort members;
• limitations in the adequacy of the match between cohort subjects and the selected
control population; and
• inadequate characterization of confounding factors, such as smoking histories for
individual workers.
The existing asbestos epidemiology database consists of approximately 150 studies of which
approximately 35 contain exposure data sufficient to derive quantitative exposure/response
relationships. A detailed evaluation of 20 of the most recent of these studies, which includes the
most recent follow-up for all of the cohorts evaluated in the 35 studies, was completed. The
following conclusions result from this evaluation:
(1) To study the characteristics of asbestos that relate to risk, it is necessary to
combine results (i.e., in a meta analysis) from studies of environments having
asbestos dusts of differing characteristics. More robust conclusions regarding risk
1.3
-------
can be drawn from an analysis of the set of epidemiology studies taken as a whole
than results derived from individual studies.
(2) By adjusting for fiber size and fiber type, the existing database of studies can be
reconciled adequately to reasonably support risk assessment.
(3) The U.S. EPA models for lung cancer and mesothelioma both appear to track the
time-dependence of disease at long times following cessation of exposure.
However, the relationship between exposure concentration and response may not
be adequately described by the current models for either disease. There is some
evidence that these relationships are supra-linear.
(4) Whereas the U.S. EPA model for lung cancer assumes a multiplicative
relationship between smoking and asbestos, the current evidence suggests that the
relationship is less than multiplicative, but possibly more complex than additive.
However, even if the smoking-asbestos interaction is not multiplicative as
predicted by the U.S. EPA model, exposure-response coefficients estimated from
the model are still likely to relate to risk approximately proportionally and,
consequently, may be used to determine an exposure index that reconciles
asbestos potencies in different environments. However, adjustments to the
coefficients may be required in order to use them to estimate absolute lung cancer
risk for differing amounts of smoking. This issue needs to be investigated further
in the next draft of this document.
(5) The optimal exposure index that best reconciles the published literature assigns
equal potency to fibers longer than 10 urn and thinner than 0.4 u.m and assigns no
potency to fibers of other dimensions.
(6) The optimal exposure index also assigns different exposure-response coefficients
for chrysotile and amphibole both for lung cancer and mesothelioma. For lung
cancer the best estimate of the coefficient (potency) for chrysotile is 0.27 times
that for amphibole, although the possibility that chrysotile and amphibole are
equally potent cannot be ruled out. For mesothelioma the best estimate of the
coefficient (potency) for chrysotile is only 0.0013 times that for amphibole and
the possibility that pure chrysotile is non-potent for causing mesothelioma cannot
be ruled out by the epidemiology data.
(7) Using the approach recommended in the U.S. EPA (1986) update, the lung cancer
exposure-response coefficients (KL values) estimated from 15 studies vary by a
factor of 72 and these values are mutually inconsistent (based on non-overlap of
uncertainty intervals). Using the approach based on the optimal exposure index
that is recommended herein, the overall variation in KL values across these studies
is reduced to a factor of 50.
(8) Using the approach recommended in the U.S. EPA (1986) update, the
mesothelioma exposure-response coefficients (KM values) estimated from 10
studies vary by a factor of 1,089 and these values are likewise mutually
1.4
-------
inconsistent. Using the approach based on the optimal exposure index that is
recommended herein, the overall variation in KM values across these studies is
reduced to a factor of 30.
(9) The exposure index and exposure-response coefficients embodied in the risk
assessment approach developed in this report are more consistent with the
literature than the current U.S. EPA approach. In particular, the current approach
appears highly likely to seriously underestimate risk from amphiboles, while
possibly overstating risk from chrysotile. Furthermore, most of the remaining
uncertainties regarding the new proposed approach also apply to the current
approach. Consequently, it is recommended that the proposed approach begin to
be applied in assessment of asbestos risk on an interim basis, while further work,
as recommended below, is conducted to further refine the approach.
(10) The residual inconsistency in both the lung cancer and mesothelioma potency
values is primarily driven by those calculated from Quebec chrysotile miners and
from South Carolina chrysotile textile workers. The difference in the lung cancer
potency estimated between these studies has long been the subject of much
attention. A detailed evaluation of the studies addressing this issue, the results of
our analysis of the overall epidemiology literature, and implications from the
broader literature, indicate that the most likely cause of the difference between
these studies is the relative distribution of fiber sizes in the two environments. It
is therefore likely that the variation between these studies can be further reduced
by developing improved characterizations of the dusts that were present in each of
these environments (relying on either archived samples, or newly generated
samples using technologies similar to those used originally).
Recommendations for Risk Assessment
Although gaps in knowledge remain, a review of the literature addressing the health-related
effects of asbestos (and related materials) provides a generally consistent picture of the
relationship between asbestos exposure and the induction of disease (lung cancer and
mesothelioma). Therefore, the general characteristics of asbestos exposure that drive the
induction of cancer can be inferred from the existing studies and were applied to define
appropriate procedures for evaluating asbestos-related risk.
Optimum values for exposure-response coefficients for lung cancer and mesothelioma were
derived in this analysis and can be combined with appropriately defined exposure estimates as
inputs for the U.S. EPA lung cancer and mesothelioma models (respectively) to assess risk.
Although these values are optimized within the constraints of the current analysis and reduce the
apparent variation across published studies substantially, the need to manage and minimize risk
when developing a general approach for assessing risk, is also recognized. Thus, to reduce the
chance of under-estimating risks, a conservative set of potency estimates were also developed
and are also presented. To assess risk, depending on the specific application, either the best-
estimate risk coefficients or the conservative estimates can be incorporated into procedures
described herein for assessing asbestos-related risks.
1.5
-------
Tables are also provided that present estimates of the additional risk of death from lung cancer,
from mesothelioma, and from the two diseases combined that are attributable to lifetime,
continuous exposure at an asbestos concentration of 0.0001 f/cm3 (for fibrous structures longer
than 10 u.m and thinner than 0.4 urn) as determined using TEM recommended methods. The risk
estimates in these tables can be combined with appropriately determined estimates of exposure
to develop estimates of risk in environments of interest.
Recommendations for Limited, Further Study
The two major objectives identified for further study are:
(1) to evaluate a broader range of exposure-response models in fitting the observed
relationship between asbestos exposure and lung cancer or mesothelioma,
respectively. For lung cancer models, this would also include an attempt to better
account for the interaction between asbestos exposure and smoking; and
(2) to develop the supporting data needed to define adjustments for exposure-
response coefficients that will allow them to be used with an exposure index that
more closely captures the criteria that determine biological activity. Among other
things, this work should focus on obtaining data that would permit more complete
reconciliation of the exposure-response coefficients derived for Quebec miners
and South Carolina textile workers.
1.6
-------
2.0 INTRODUCTION
The purpose of this report is to provide a foundation for completing a state-of-the-art-protocol to
assess potential human-health risks associated with exposure to asbestos. Such a protocol is
intended specifically for use in performing risk assessments at Superfund sites, although it may
be applicable to a broad range of situations.
The current report is a revision to a version originally submitted on September 4, 2001 (Berman
and Crump 2001), which (among other things) includes both an extensive review of the general
literature and a detailed analysis of the existing epidemiology studies. These are reproduced in
the current report in Chapter 6 and Chapter 7/Appendix A (respectively).
The September 4, 2001 version was also the subject of a peer-review consultation held in San
Francisco on February 25-26, 2003. The comments of the expert panel convened to conduct the
peer review are included in this report as Appendix B.
In general, the expert panel endorsed the overall approach to risk assessment proposed in this
report, although they highlighted areas where controversies persist. They also suggested
additional research and analyses to attempt to resolve some of the outstanding controversies and
to refine several of the details of the approach. In addition, they offered recommendations for
modifications to improve the overall transparency and readability of the earlier version of this
report.
The current report incorporates the changes recommended by the peer review consultation panel
to correct minor problems with internal consistency and the overall transparency of the
discussion that are needed to improve readability. Although some of the research and analyses
recommended by the peer consultation panel are not complete, it is anticipated that the current
document can be distributed for broader review and comment. Thus, the recommended approach
to risk assessment can be considered for use in the interim, while the additional research and
analyses recommended by the expert panel are completed. At that point, a final revision of this
document will be developed and it is expected to serve as a component of a broader effort by the
U.S. Environmental Protection Agency (U.S. EPA) to revise the Agency's current approach for
assessing asbestos-related risks.
The approach currently employed by U.S. EPA to evaluate asbestos-related risks (IRIS, 1988) is
based primarily on a document completed in 1986 (U.S. EPA 1986) and has not changed in the
past 15 years, despite substantial improvements in asbestos measurement techniques and in the
understanding of the manner in which asbestos exposure contributes to disease. Therefore,
among other things, this document provides an overview and evaluation of more recent studies
and presents proposed modifications to the current approach for assessing asbestos-related risks
that can be justified based on the more recent work.
In May 2001, the U.S. EPA along with the California Environmental Protection Agency
(CalEPA), the National Institute for Occupational Safety and Health (NIOSH), the American
Toxic Substances Disease Registry (ATSDR), and the Mine Safety and Health Administration
(MSHA) hosted an international conference on asbestos in Oakland, California that was attended
2.1
-------
by leading international experts on asbestos. The state of knowledge concerning such issues as
the nature of asbestos, the measurement of asbestos, and the relationship between asbestos
exposure and the induction of disease was reviewed during this conference. Particular emphasis
was placed on identifying important knowledge gaps in these areas.
By coupling the outstanding issues identified at the Oakland meeting with additional information
gleaned from the literature, the following set of issues was identified as risk-specific issues of
current interest:
• whether the exposure-response models currently in use by the U.S. EPA for
describing the incidence of asbestos-related diseases adequately reflect the time-
and exposure-dependence for the development of these diseases;
• whether different potencies need to be assigned to the different asbestos mineral
types to adequately predict risk for the disease end points of interest;
• to the extent that different asbestos mineral types are assigned distinct potencies,
whether the relative in vivo durability of different asbestos mineral types
determines their relative potency;
• whether the set of minerals included in the current definition of asbestos
adequately covers the range of minerals that potentially contribute to asbestos-
related diseases;
• whether the analytical techniques and methods currently used for determining
asbestos concentrations adequately capture the biologically relevant
characteristics of asbestos (particularly with regard to the sizes of the structures
counted using the various analytical methods) so that they can be used to support
risk assessment; and
• whether reasonable confidence can be placed in the cross-study extrapolation of
exposure-response relationships that are required to assess asbestos-related risks
in new environments of interest.
These outstanding issues (and other related considerations) are addressed in this document to
provide a foundation for proposing a new approach for assessing asbestos-related risks.
Compared to the current U.S. EPA approach, it is shown that the new approach better predicts
risks among the environments in which asbestos-related risks have been previously evaluated
(i.e., the published epidemiology studies) so that the new approach can be used to predict risks in
unstudied environments of interest with greater confidence than predictions based on the current
approach. Moreover, completing the additional research and analysis recommended by the
expert panel (Appendix B) should facilitate further refinement while providing additional
opportunities to better evaluate and validate the proposed approach.
2.2
-------
The remainder of this document is divided into 6 chapters:
• Chapter 3 presents an overview of the general considerations that need to be
addressed to assess asbestos-related risks (including considerations associated
with the manner in which asbestos exposures are characterized, the manner in
which risk is modeled from existing data, and the manner that risk models are
then applied to estimate risk in new environments). The nature of the diseases
commonly attributed to asbestos exposure are also briefly described;
• Chapter 4 presents a background discussion that addresses the definition of
asbestos, the mineralogy of asbestos, the morphology of asbestos-containing dusts
to which people are typically exposed, the capabilities and limitations of
analytical techniques and methods used to determine airborne asbestos
concentrations, and the structure and function of the human lung;
• Chapter 5 provides a description of the kinds of literature studies that are
commonly used to support development of a protocol to assess risk, with
particular emphasis on identifying their relative strengths and weaknesses;
• Chapter 6 presents a review of the literature with particular emphasis on studies
published since the Health Effects Assessment Update (U.S. EPA 1986).
Combined with a description of supporting analyses, the review is focused on
reconciling apparently conflicting studies and hypotheses (when possible) and
identifying the best candidate procedures for assessing asbestos-related risks. To
reconcile studies, the strengths and weaknesses common to various types of
studies are explicitly considered;
• Chapter 7 presents a reevaluation of the published epidemiology studies that,
among other things, is designed to address and (when possible) resolve the
outstanding issues of current interest; and
• Chapter 8 presents a proposed, new approach for assessing asbestos-related risks.
Regarding Chapter 8, although the objective of this document was to identify the single best
procedure, when current knowledge is inadequate for distinguishing among alternatives, options
are presented along with a discussion of their relative advantages and limitations. A few limited
and focused additional research studies are recommended, which have the potential to enhance
the current state of knowledge sufficiently to resolve one or more of the important, remaining
issues. These recommended studies parallel those identified by the peer review panel (Appendix
B).
This report is part of a series of documents developed as part of a multi-task project to develop a
set of mutually consistent methods for determining asbestos concentrations in a manner useful
for assessing risk and a companion protocol for conducting such risk assessments. A method for
the determination of asbestos in air (Chatfield and Berman 1990) and a companion technical
background document (Berman and Chatfield 1990) were published by the U.S. EPA in 1990.
The air method has since been superceded (improved) by the ISO Method (ISO 1995). A
2.3
-------
method for the determination of asbestos in soils and bulk materials (Berman and Kolk 1997)
was also published by the U.S. EPA and the draft of an improved version was also recently
completed (Berman and Kolk 2000). The recommendations in this document should serve as the
basis for development of the companion risk-assessment protocol.
2.4
-------
3.0 OVERVIEW
Inhalation of asbestos dusts has been linked to several adverse health effects including primarily
asbestosis, lung cancer, and mesothelioma (U.S. EPA 1986). The kinds of lung cancers linked to
asbestos exposure are similar to those induced by smoking and a greater-than-additive effect has
been observed from combined exposure (see, for example, Liddell and Armstrong 2002).
Mesothelioma is a rare cancer of the membranes that line the pleural cavity (containing the heart
and lungs) and the peritoneal cavity (i.e., the gut). Although there is some evidence of a low
background incidence of spontaneous mesotheliomas, this cancer has been associated almost
exclusively with exposure to asbestos and certain other fibrous substances (HEI-AR, 1991).
Asbestosis, a chronic, degenerative lung disease, has been documented among asbestos workers
from a wide variety of industries. Although asbestosis cases have been observed at some
locations of current interest to the U.S. EPA, the disease is generally expected to be associated
only with the higher levels of exposure commonly found in workplace settings and is not
expected to contribute substantially to risks potentially associated with environmental asbestos
exposure. Therefore, asbestosis is only considered in this document to the extent required to
address its putative association with lung cancer. Overall, the majority of evidence indicates that
lung cancer and mesothelioma are the most important risks associated with exposure to low
levels of asbestos.
The primary route of exposure of concern in association with asbestos is
inhalation. There is little evidence that ingestion of asbestos induces disease (see,
for example, IRIS 1988; U.S. EPA 1986). Therefore, this study is focused on
inhalation hazards, and other routes of exposure are not addressed.
Gastrointestinal cancers and cancers of other organs (e.g., larynx, kidney, and ovaries) have also
been linked with asbestos exposures (by inhalation) in some studies. However, such associations
are not as compelling as those for lung cancer and mesothelioma and the potential risks from
asbestos exposures associated with these other cancers are much lower (U.S. EPA 1986).
Consequently, by addressing the more substantial asbestos-related risks associated with lung
cancer and mesothelioma, the much more moderate risks potentially associated with cancers at
other sites are also addressed by default. Therefore, this document is focused on the risks
associated with lung cancer and mesothelioma.
A variety of human, animal, and tissue studies have provided insight into the nature of the
relationship between asbestos exposure and disease. Ideally, human epidemiology studies are
employed to determine the quantitative exposure/response relationships and the attendant risk
coefficients for asbestos exposure. Risk coefficients have been estimated for asbestos from
approximately 20 epidemiology studies for which adequate exposure-response data exist.
However, such coefficients vary widely (for lung cancer, coefficients vary by more than a factor
of 70 and, for mesothelioma, they vary by more than a factor of 1,000) and this variation has not
been reconciled. Among the objectives of this study, one is to evaluate and account for the
sources of uncertainty that contribute to the variation among the risk coefficients derived from
the literature so that these estimates can be reasonably interpreted and recommendations for their
use in risk assessment developed.
3.1
-------
Animal and tissue studies indicate that asbestos potency is a complex function of several
characteristics of asbestos dusts including fiber size and fiber type (i.e., fiber mineralogy).
Moreover, the influence of fiber size is a complex function of both diameter and length.
Therefore, whenever the goal is to compare across samples with differing characteristics, it is not
sufficient to report asbestos concentrations simply as a function of mass (or any other single
parameter), which is in stark contrast to the treatment of chemical toxins. It has generally been
difficult to distinguish among the effects of fiber size and type in many studies because such
effects are confounded and the materials studied have not been adequately characterized.
The influence of different characteristics of asbestos dusts upon risk cannot be adequately
evaluated in the existing epidemiological studies because the analytical techniques used to
monitor asbestos exposure in these studies are not capable of resolving all of the characteristics
of asbestos dusts that other types of studies indicate are important. Moreover, the exposure
indices (the range of structure sizes and shapes used to characterize an asbestos dust) that are
employed in the existing epidemiology studies may not correspond with the characteristics of
asbestos that best relate to biological activity. This hinders the ability to reconcile risk
(exposure-response) coefficients derived from different studies. It also limits the confidence
with which risk coefficients derived from the existing epidemiology studies can be applied to
assess risks from asbestos exposure in other environments. Such limitations are explored in this
report, along with potential remedies.
Based on the current approach for evaluating asbestos-related cancer risk (U.S. EPA 1986), risk
is estimated as the product of a risk coefficient and a mathematical function that depends on the
level of exposure, the duration of exposure, and time. The risk coefficient for lung cancer is
generally denoted as, "KL" and the one for nasothelioma as "K M".A detailed description of
both the current lung cancer and mesothelioma models is provided in Chapter 7. The models
differ depending on whether lung cancer or mesothelioma is being considered.
For lung cancer, the model estimates relative risk, which means that the increase in lung cancer
incidence that is attributable to asbestos exposure is proportional to the background lung cancer
incidence in the exposed population. The background cancer incidence is the rate of lung cancer
that would be expected to occur in the population in the absence of asbestos exposure. In other
words, background lung cancer incidence is the lung cancer rate for the exposed population that
is attributable to all causes other than asbestos.
Among the implications of the lung cancer model is that the combined effects of asbestos
exposure and smoking is multiplicative and, until recently, the majority of studies have
suggested such a multiplicative relationship (see, for example, Hammond et al. 1979). However,
newer studies (for example, Liddell and Armstrong 2002) suggest a complex relationship that is
closer to additive than multiplicative. Such considerations are addressed further in Chapter 7.
The current EPA model for mesothelioma is an absolute risk model. This means that the
increase in mesothelioma risk attributable to asbestos is independent of the background rate of
mesothelioma, which is negligible in the general population.
3.2
-------
Ideally, the risk coefficients derived from the existing epidemiology studies can be combined
with measurements from other exposure settings to estimate lung cancer and mesothelioma risks
in these other exposure settings. However, such risk estimates are only valid if both of the
following conditions are met:
(1) asbestos is measured in the exposure setting of interest in the identical manner in
which it was measured in the study from which the corresponding risk
coefficients are derived; and
(2) such measurements reflect the characteristics of asbestos exposures that
determine risk.
A growing body of evidence indicates that the way in which asbestos concentrations were
measured in the existing epidemiology studies do not reflect the characteristics of asbestos
exposure that determine risk. Therefore, measuring asbestos concentrations in the same way in
exposure settings of interest may not be sufficient to assure the validity of risk estimates derived
using the published risk coefficients (and the corresponding models). This is because the second
of the above-listed conditions would not be satisfied.
Considerations necessary to compare risk coefficients derived in different exposure settings (or
to apply a coefficient to predict risk in a setting different from the one in which the coefficient
was derived) have been elucidated clearly in a mathematical model (Chesson et al. 1990). The
consequences of the model indicate that adjusting the existing risk coefficients so that they
reflect asbestos characteristics that determine biological activity requires knowledge of the fiber
size distributions of the dusts studied in the original epidemiology studies. To the extent they
exist, such data may be used to normalize each of the published risk coefficients so that they
relate to a common exposure index reflecting asbestos characteristics that determine biological
activity.
Among the goals of this evaluation is to explore the possibility of defining an improved exposure
index (that better reflects biological activity) and to use this index to reconcile the epidemiology
data (see Chapter 7). We also evaluated improved ways of simultaneously accounting for the
effects of both fiber size and type.
Unfortunately, some of the issues that need to be resolved to support development of a protocol
for assessing asbestos-related risks cannot be entirely resolved with existing data. Therefore, in
later chapters (i.e., Chapters 6, 7, and 8), we have attempted to identify such issues, to assess
their relative importance, and, when deemed appropriate, to propose limited and focused
research projects designed to provide the data required to reduce the impacts of such knowledge
gaps.
3.3
-------
4.0 BACKGROUND
Asbestos is a term used to describe the fibrous habit of a family of hydrated metal silicate
minerals. The most widely accepted definition of asbestos includes the fibrous habits of six of
these minerals (IARC 1977). The most common type of asbestos is chrysotile, which is the
fibrous habit of the mineral serpentine, a magnesium silicate. The other five asbestos minerals
are all amphiboles (i.e., all partially hydrolyzed, mixed-metal silicates). These are: fibrous
reibeckite (crocidolite), fibrous grunerite (amosite), anthophyllite asbestos, tremolite asbestos,
and actinolite asbestos.
All six of the minerals whose fibrous habits are termed asbestos occur most commonly in non-
fibrous, massive habits. While unique names have been assigned to the asbestiform varieties of
three of the six minerals (noted parenthetically above) to distinguish them from their massive
forms, such nomenclature has not been developed for anthophyllite, tremolite, or actinolite.
Therefore, when discussing these latter three minerals, it is important to specify whether a
massive habit of the mineral or the fibrous (asbestiform) habit is intended.
Although other minerals may also occur in a fibrous habit, they are not generally included in the
definition of asbestos either because they do not exhibit properties typically ascribed to asbestos
(e.g., high tensile strength, the ability to be woven, heat stability, and resistence to attack by acid
or alkali) or because they do not occur in sufficient quantities to be exploited commercially.
The first four of the six asbestos minerals listed above have been exploited commercially (IARC
1977). Of these, chrysotile alone accounts for more than 90% of the asbestos found in
commercial products.
Importantly, it is neither clear whether the term asbestos maps reasonably onto the range of
fibrous minerals that can contribute to asbestos-like health effects nor whether individual
structures of the requisite mineralogy must formally be asbestiform to contribute to such health
effects.
Regarding whether the term asbestos is a useful discriminator for health effects, it is well
established that erionite (a fibrous zeolite not related to asbestos) is a potent inducer of
mesothelioma (Baris et al. 1987), which is one of the two primary asbestos-induced cancers (see
Chapter 3). It is therefore possible that the fibrous habits of at least some other minerals not
formally included in the current definition of asbestos may contribute to the induction of
asbestos-related diseases. Therefore, an efficient procedure is needed for separating potentially
hazardous materials from those that are most likely benign.
There are two issues related to the question of whether fibers must formally be asbestiform to
contribute to health effects. The first involves the relationship between fiber structure and
disease induction and the second involves measurement. Although the evidence is
overwhelming that the size and shape of a fiber affects the degree to which it contributes to the
induction of disease (this is addressed in detail in Chapter 6), it does not appear that sizes
inducing biological activity are well distinguished by criteria that define the asbestiform habit.
Therefore, depending on the definition employed for the fibers (or fibrous structures) that are
4.1
-------
counted during an analysis, it may or may not be necessary to distinguish formally between
asbestiform and non-asbestiform structures for the concentrations derived from such a count to
adequately reflect biological activity.
The dimensions of an asbestiform fiber are determined by the manner in which the fiber grows
(Addison 2001). In contrast, the massive forms of various minerals, when cleaved, also form
elongated particles (termed "cleavage fragments") and, depending on the definition employed for
fibrous structures during an analysis, such cleavage fragments may or may not be included along
with asbestiform fibers in a count (see Section 4.3). Although it is clear from the manner in
which they are each formed that the surface properties of asbestiform fibers and cleavage
fragments are likely to be very different (for example, the latter will have many "unsatisified"
chemical bonds), the degree to which such differences affect the toxic potency for comparable
sized structures is not currently known.
Although it is beyond the scope of this document to present a detailed treatise on asbestos
mineralogy, the morphology of asbestos dusts, or the nature and limitations of analytical
techniques and methods used to determine asbestos concentrations, a brief overview of these
topics is presented in the following sections both to identify issues that need to be addressed as
part of the development of an appropriate protocol for assessing asbestos risks and to provide the
background required to facilitate evaluation of the relevant issues. In that regard, a section on
lung physiology and function is also provided.
4.1 THE MINERALOGY OF ASBESTOS
As previously indicated, the six asbestos minerals can be divided into two general classes.
Chrysotile is the fibrous habit of the mineral serpentine (Hodgson 1965). The smallest fibrils of
chrysotile occur as rolled sheets or hollow tubules of this magnesium silicate mineral. The larger
fibers of chrysotile form as tightly packed bundles of the unit fibrils.
Chrysotile fibrils typically range from 20 nm to approximately 300 or 400 nm (0.02 to 0.3 or
0.4 fim) in diameter. Although slightly thicker fibrils may occasionally occur, at some point the
curvature induced by the mismatch between the magnesium and silicon layers of the fibril
becomes thermodynamically unstable, so that production of thicker fibrils is prohibited (Addison
2001).
Chrysotile bundles are held together primarily by Van der Waals forces and will readily
disaggregate in aqueous solutions containing wetting agents (e.g., soap). They will also readily
disaggregate in lung surfactant (Addision, 2001).
The general chemical composition of serpentine is reported as Mg3(Si2O5)(OH)4 (Hodgson
1965). However, the exact composition in any particular sample may vary somewhat from the
general composition. For example, aluminum may occasionally replace silicon and iron, nickel,
manganese, zinc or cobalt may occasionally replace magnesium in the crystal lattice of
chrysotile (serpentine).
4.2
-------
The five other common varieties of asbestos are all fibrous forms of amphibole minerals
(Hodgson 1965). These are ferro-magnesium silicates of the general composition:
A2.3B5(Si,Al)8022(OH)2
where:
A = Mg, Fe, Ca, Na, or K; and
B = Mg, Fe, or Al.
Some of these elements may also be partially substituted by Mn, Cr, Li, Pb, Ti, or Zn.
The fibrous habits of the amphibole minerals tend to occur as extended chains of silica tetrahedra
that are interconnected by bands of cations (Hodgson 1965). Each unit cell typically contains
eight silica tetrahedra and the resulting fibers tend to be rhomboid in cross-section. Amphibole
fibers are generally thicker than chrysotile fibrils and may typically range from approximately
100 nm to 700 or 800 nm in diameter (Addison 2001). Substantially thicker fibers have also
been observed.
4.2 MORPHOLOGY OF ASBESTOS DUSTS
Structures comprising the fibrous habits of the asbestos minerals come in a variety of shapes and
sizes. Not only do single, isolated fibers vary in length and thickness, but such fibers may be
found combined with other fibers to form bundles (aggregates of closely packed fibers arranged
in parallel) or clusters (aggregates of randomly oriented fibers) or combined with equant
particles to form matrices (asbestos fibers embedded in non-asbestos materials). Consequently,
dusts (even of one mineral variety) are complex mixtures of structures. For precise definitions of
the types of fibrous structures typically found in asbestos dusts, see ISO (1995).
Detailed descriptions of the characteristics of dusts typically encountered at environmental and
occupational asbestos sites have been reported in the literature and the following summary is
based on a previously published review (Berman and Chatfield 1990). Typically, the major
components of the dust observed in most environments are non-fibrous, isometric particles.
Fibrous structures consistently represent only a minor fraction of total dust. Asbestos structures
represent a subset of the fibrous structures.
The magnitude of the fraction of total dust represented by fibers and the fraction of fibers
composed of asbestos minerals vary from site to site. However, the fraction of asbestos in total
dusts has been quantified only in a very limited number of occupational and environmental
settings (see, for example, Cherrie et al. 1987 or Lynch et al. 1970).
The gross features of structure size distributions appear to be similar among asbestos dusts
characterized to date (Berman and Chatfield 1990). The major asbestos fraction of all such dusts
are small fibrous structures less than 5 um in length. Length distributions generally exhibit a
mode (maximum) between 0.8 and 1.5 |im with larger fibers occurring with decreasing
frequency. Fibrous structures longer than 5 |im constitute no more than approximately 25% of
total asbestos structures in any particular dust and generally constitute less than 10%.
In some environments, the diameters of asbestos fibers exhibit a narrow distribution that is
4.3
-------
largely independent of length. In other environments, diameters appear to exhibit a narrow
distribution about a mean for each specific length. In the latter case, both the mean and the
spread of the diameter distribution increases as the length of the structures increase. The
increase in diameter with length appears to be more pronounced for chrysotile than for the
amphiboles, presumably due to an increase in the fraction of chrysotile bundles contributing to
the overall distribution as length increases.
Only a few studies have been published that indicate the number of complex structures in
asbestos size distributions. The limited data available indicate that complex structures may
constitute a substantial fraction (up to one third) of total structures, at least for chrysotile dusts
(see, for example, Sebastien et al. 1984). Similar results were also obtained during a re-analysis
of dusts generated from the asbestos samples evaluated in the animal inhalation studies
conducted by Davis et al. (Berman et al., in preparation). This is the same re-analysis used to
support a study to identify asbestos characteristics that promote biological activity (Berman et al.
1995), which is discussed further in Section 6.4.3.
Historically, fibrous structures have been arbitrarily defined as structures exhibiting aspect ratios
(the ratio of length to width) greater than 3:1 to distinguish them from isometric particles
(Walton 1982). However, alternate definitions for fibers have also been proposed, which are
believed to better relate to biological activity (see, for example, Berman et al. 1995 or Wylie
et al. 1993). The degree to which fibers are combined within complex structures in a particular
dust may also affect the biological activity of the dust (Berman et al. 1995). Therefore, proper
characterization of asbestos exposure requires that the relative contributions from each of many
components of exposure be simultaneously considered. Factors that need to be addressed
include the distribution of structure sizes, shapes, and mineralogy in addition to the absolute
concentration of structures. Such considerations are addressed further in Chapter 6. Thus,
unlike the majority of other chemicals frequently monitored at hazardous wastes sites, asbestos
exposures cannot be adequately characterized by a single concentration variable.
4.3 CAPABILITIES OF ANALYTICAL TECHNIQUES USED TO MONITOR
ASBESTOS
Due to a complex history, a range of analytical techniques and methods have been employed to
measure asbestos in the various studies conducted over time (Walton 1982). Use of these
various methods has affected the comparability of results across the relevant asbestos studies
(Berman and Chatfield 1990). Therefore, the relative capabilities and limitations of the most
important methods used to measure asbestos are summarized here. Later sections of this report
incorporate attempts to reconcile effects that are attributable to the limitations of the different
methods employed in the various studies evaluated.
Analytical techniques used to measure airborne asbestos concentrations vary greatly in their
ability to fully characterize asbestos exposure. The capabilities and limitations of four analytical
techniques (midget impinger [MI], phase contrast microscopy [PCM], scanning electron
microscopy [SEM], and transmission electron microscopy [TEM]) are described here. A general
comparison of the relative capabilities and limitations of the analytical techniques introduced
above is presented in Table 4-1.
4.4
-------
Table 4-1. Capabilities and Limitations of Analytical Techniques Used for Asbestos
Measurements9
Parameter
Range of Magnification
Particles Counted
Midget
Impinger
100
All
Phase Contrast
Microscopy
400
Fibrous
Structures5
Scanning
Electron
Microscopy
2,000-10,000
Fibrous
Structures5
Transmission
Electron
Microscopy
5,000-20,000
Fibrous
Structures'^
Minimum Diameter (size) 1 |im
Visible
0.3
0.1 urn
<0.01
Resolve Internal Structure
Distinguish Mineralogy*1
No
No
No
No
Maybe
Yes
Yes
Yes
"The capabilities and limitations in this table are based primarily on the physical constraints of the indicated
instrumentation. Differences attributable to the associated procedures and practices of methods in common use
over the last 25 years are highlighted in Table 4-2.
bFibrous structures are defined here as particles exhibiting aspect ratios (the ratio of length to width) greater than
3 (see Walton 1982).
TEM counts frequently resolve individual fibrous structures within larger, complex structures. Based on internal
structure, several different counting rules have been developed for handling complex structures. See the
discussion of methods presented below.
dMost SEM and TEM instruments are equipped with the capability to record selected area electron diffraction
(SAED) spectra and perform energy dispersive X-ray analysis (EDXA), which are used to distinguish the
mineralogy of structures observed.
MI and PCM are the two analytical techniques used to derive exposure estimates in the majority
of epidemiology studies from which the existing risk factors are derived. SEM is an analytical
technique that has been employed in several key animal studies. TEM provokes interest because
it is the only analytical technique that is potentially capable of distinguishing all of the
characteristics of asbestos that potentially affect biological activity.
Although PCM was (and still is) widely used to characterize occupational exposures, its inability
to distinguish between asbestos and non-asbestos and its lack of sensitivity limits its usefulness
in environmental settings (Berman and Chatfield 1990). In fact, PCM analyses and TEM
analyses showed no correlation among measurements collected during the cleanup of the 1991
Oakland Hills fire (Berman, unpublished data). Such lack of correlation is expected to be
observed generally whenever measurements are collected at sites where asbestos concentrations
are low enough that a substantial fraction of the structures counted by PCM are not asbestos.
Consequently, TEM is the technique that has been recommended for use at Superfund sites
(Berman and Kolk 1997; Chatfield and Berman 1990).
Importantly, the physical limitations of the various analytical techniques is only part of the
problem. To provide reproducible results that can be compared meaningfully to other analyses
in other studies, one must also consider the choice of procedures (methods) that address
everything from sample collection and preparation to rules for counting and quantifying asbestos
structures.
4.5
-------
Table 4-2. Comparison of Applicable Methods For Measuring Asbestos in Air
Analytical Technique
Preparation
Methodology
Magnification
Dimensions Counted
Length (L):
Width (W):
Aspect Ratio (AR):
Sensitivity:
s/cm3
s/mm2
Mineralogy
Determined
Maximum Number
Counted
NIOSH 7400s
PCM
Direct (no transfer)
450x
• -
L > 5 um
'> . . - ' ; "
W> 0.25 (im
AR>3
'
No
100 structures '
NIOSH 7402b
TEM
Direct
10,000x
L> 1 um
3.0 > W > 0.04 urn
AR>3
Yes
100 structures
YAM ATE"
TEM
Direct
(Indirect Optional)
20,000x
L > 0.06 urn
' "
>
, '
W> 0.02 urn
AR>3 - ',
- '<
» -
•
Yes, except matrix
particles , . ,
100 structures
AHERAd
TEM
Direct
15,000x-20,000x
L > 0.5 um
W > 0.02 um
AR>5
0.005
70
Yes, except matrix
particles
50 structures
' ' 1
ISOef
TEM
Direct
(Indirect Optional)
20,000x (total
structures) ' -
lO.OOOxXstructures'
longer than 5 u.m)
L > 0.5 Jim, total
structures
L > 5 \im, long - •
structures • 1 .
/W<3.o'jim ,
'- (respirability)
AR.>5
V
- Adjustable :
: 10 for total structures
, 0. 1 for long structures
Yes
100 total structures
100 long structures
4.6
-------
Table 4-2. Comparison of Applicable Methods For Measuring Asbestos in Air (continued)
Maximum Area
Scanned
Statistically Balanced
Counting8
Counting Rules
Structures
NIOSH7400"
100 fields
~ - * ~,
-"No - ' -: '- - t
!/;<"> -
Count all structures
exhibiting L > 5 Jim,
W< 3.0 |ira, and
AR > 3. >• '•' . - ,
*• / •<. j- f \.
s / ^ \
. - -
' ^ ' " ' /
' -
NIOSH 7402b
100 grid openings
No
Count all structures
exhibiting L > 1 |im, W
<3.0 jim, and AR>3.
Note PCMEh fraction
within count.
YAMATE0
10 grid openings
No
' '
Count all structures
^exhibiting an AR > 3. ,,
', I
s «• ^ *
" ' - v ,
AHERAd
Blanks: 10 openings
Samples: 10 openings
(assuming defined
sensitivity is achieved
with the collected air
volumes).
No
Count all structures
with L > 0.5 urn
exhibiting an AR > 5.
Record individual
fibers within all
groupings with fewer
than 3 intersections.
Count structures with L
> 5 jim separately
(PCME8).
ISOef
Adjustable
A \ ',,
yes. :;.;:: :;,;;;;
. ; ? : ;.;>/••*!
H'r ' - I'l "'••"• :- 1
. Count all structures «
with L > 0.5 pm or A >. ••
containing components
\vtth L > 0.5 (|m that ^ •.
also exhibit an AR > 5,.
Separately identifiable ;
components of parent; ,
structures that satisfy ;
dimensional criteria are
also separately
enumerated; '-, ,
" Condudt a similar
count to that indicated
above for structures - :
* with L > 5 jim. '
4.7
-------
Table 4-2. Comparison of Applicable Methods For Measuring Asbestos in Air (continued)
Bundles
NIOSH 7400"
Bundles meeting
overall dimensional
criteria generally
counted as single fibers
- unless up to 10
individual fiber ends
can be distinguished
within the bundle
(representing 5"
individual fibers).
-
'
Clusters
v * .
.
Within a cluster, count
up to JO individual
fiber ends from (up to . ,
5) fibers that meet the
overall dimensional
criteria. Otherwise, <
count a cluster as a
single entity.
,
; < • . , - . , ,
' - *• f
;•••••
NIOSH 7402b
Bundles meeting
overall dimensional
criteria generally
counted as single
fibers.
YAMATEC
Bundles meeting
overall dimensional
criteria generally
counted as single
entities and noted as
bundles on the.count
f sheet. -
I - '..
Within a cluster, count
up to 3 individual
fibers that meet the
overall dimensional
criteria. Otherwise,
clusters that contain
more than 3 fibers that
meet the overall
dimensional criteria are
counted as single
clusters.
- -
„ /
>
- >
Within a cluster,' count
up to 3 individual
fibers that meet the
overall dimensional
criteria. Otherwise,
clusters that contain ',
more than 3 fibers that
meet the overall
dimensional criteria are
counted as single
clusters.
, . . .
AHERAd
Bundles of 3 or more
fibers that meet the
overall dimensional
criteria are counted as
single entities and
noted as bundles on the
count sheet.
A collection of fibers
with more than 2
intersections where at
least one individual
projection meets the
overall dimensional
criteria is counted as a
single cluster.
ISO"
Count parent bundles
with L > 0.5 Jim
containing at least one
component fiber that
exhibits an AR> 5.
Qualifying bundles that
are components of
other parent1 structures
are also separately *
- enumerated. ?
For counts of structures
with L > 5 um, include
only bundles longer
than 5 (im.
Distinguish "disperse" - |
and "compact" clusters!
Count all clusters - -
containing at least one
- component fiber or
• bundle satisfying - •• -<
appropriate \
dimensional criteria.
Separately enumerate
up to 10 component
| structures satisfying
appropriate
dimensional criteria. .-
4.8
-------
Table 4-2. Comparison of Applicable Methods For Measuring Asbestos in Air (continued)
Matrices
NIOSH 7400"
Count up to 5 fibers
emanating from a
, clump (matrix). Each
individual fiber must
meet the .dimensional. .
' criteria. • ' ' s \H ',',
•'-, - '~>'-* 'ff ^ ,' • ~- & J* ,» ,
•i, •<* •, ' / r .-*•' ^ /• '-
;,«v>i.|,. : .«>- . -^ •* . -a '•,
,4^-t- ,' ,,-'f' ; f " '- " • ;
NIOSH 7402b
Count individually
identifiable fibers
within a matrix. Fibers
must individually meet
the dimensional
criteria.
YAMATEC
Count as a single
matrix, all matrices
with at least one
protruding or •.',-.,
' embedded fiber that'
;< meets the dimensional
' criteria, j ' • ;\' ; £
; ....,* ~ - , ^ -j- •"- • f.
• ' *,?'" "i. •" •', • ';'*'•• . |s '- ''• " 'I
AHERAd
Count as a single
matrix, all matrices
with at least one
protruding fiber such
that the protruding
section meets the
dimensional criteria.
ISOe^
Distinguish "disperse"
- and "compact"
matrices.. Count all
matrices containing at ' '
least one ^component , ;
fiber or bundle .Z >• • [ •
Satisfying appropriate .!
-'dimensional criteria. •[ '' i
^Separafely enumerate ;
up to 10 component - -
.structures satisfy ihg, r'',.
• appropriate : • ^ ••: 4. ; ;
dimensional' criteria/^
National Institute for Occupational Safety and Health (1985). Method for Determination of Asbestos in Air Using Positive Phase Contrast Microscopy.
NIOSH Method 7400. NIOSH, Cincinnati, Ohio, U.S.A.
''National Institute for Occupational Safety and Health (1986). Method for Determination of Asbestos in Air Using Transmission Electron Microscopy.
NIOSH Method 7402. NIOSH, Cincinnati, Ohio, U.S.A.
Tamate, G., Agarwal, S.C., and Gibbons, R.D. (1984). Methodology for the Measurement of Airborne Asbestos by Electron Microscopy. U.S. EPA Report
No. 68-02-3266. U.S. Environmental Protection Agency, Washington, D.C., U.S.A.
dU.S. Environmental Protection Agency (1987). Asbestos Hazard Emergency Response Act: Asbestos-Containing Materials in Schools. Final Rule and
Notice (Appendix A: AHERA Method). Federal Register, 40 CFR 763, Vol. 52, No. 2, pp. 41826-41903, October.
'Chatfield, E.J. (1995). Ambient Air: Determination of Asbestos Fibres, Direct Transfer Transmission Electron Microscopy Procedure. Submitted to the
International Standards Organization: ISO/TC 10312.
fNote that the ISO Method is a successor to the Interim Superfund Method: Chatfield, E.J. and Herman, D.W. (1990). Superfund Methodfor the
Determination of Asbestos in Ambient Air. Part 1: Method Interim Version.. Prepared for the U.S. Environmental Protection Agency, Office of Emergency
and Remedial Response, Washington, D.C. EPA/540-2-90/005a. May.
Statistically balanced counting is a procedure incorporated into some asbestos methods (e.g. the Superfund Methods and the ISO Methods) in which long
structures (typically longer than 5 \im) are counted separately during a lower magnification scan than used to count total structures (which are predominantly
short). This procedure assures that the relatively rare longer structures are enumerated with comparable precision to that of the shorter structures.
hPCME stands for phase contrast microscope equivalent and indicates the fraction of structures observed by transmission electron micrscopy that would also
be visible by phase contrast microscopy.
4.9
-------
Multiple methods have been published for use in conjunction with several of the analytical
techniques mentioned above (particularly TEM). Such methods differ in the procedures
incorporated for sample preparation and for the manner in which asbestos structures are counted.
The sample preparation requirements, conditions of analysis, and structure counting rules for
several of the most commonly employed methods are presented in Table 4-2 to illustrate how the
choice of method can result in substantially different measurements (even on duplicate or split
samples).
The second column of Table 4-2 describes the specifications of the PCM method currently
mandated by the Occupational Safety and Health Agency (OSHA) for characterizing asbestos
exposure in occupational settings. Although this method is in common use today, several
alternate methods for counting fibrous structures by PCM have also been used historically.
Therefore, PCM measurements reported in earlier studies (including the available epidemiology
studies) may not be comparable to PCM results collected today.
The last four columns of Table 4-2 describe TEM methods that are in current use. Comparison
across these methods indicates:
• the shortest lengths included in counts using these methods vary between 0.06
and 1 urn Given that structures shorter than 1 ^m represent a substantial fraction
of total asbestos structures in almost any environment (Section 4.2), this
difference alone contributes substantially to variation in measurement results
across methods;
• the definitions and procedures for counting complex structures (i.e., bundles,
clusters, and matrices) vary substantially across methods, which further contribute
to variation in measurement results. For example, the ISO Method requires that
component fibers of clusters and matrices be counted separately, if they can be
readily distinguished. In contrast, clusters are counted as single structures under
the AHERA Method; and
• although all of the methods listed incorporate sample preparation by a direct
transfer process (in which the fibers are counted in their original position on the
filter), several of the methods have also been paired with an optional indirect
transfer process (which involves ashing the original air filter, mixing the residue
in water, sonicating, and re-suspending the fibers on a new filter). Measurements
derived from split samples that are prepared, respectively, by direct and indirect
transfer, can vary by factors as large as several 100 (Berman and Chatfield 1990).
Typically, counts of asbestos structures on samples prepared by an indirect
transfer procedure are greater than those derived from directly prepared samples
by factors of between 5 and 50.
4.10
-------
Given the combined effects from the physical limitations of the various techniques employed to
analyze for asbestos and the varying attributes of the methods developed to guide use of these
techniques, the relative capabilities and limitations of asbestos measurements derived,
respectively, from paired methods and techniques in common use can be summarized as follows:
• all four techniques are particle counting techniques;
• neither MI nor PCM are capable of distinguishing asbestos from non-asbestos
(i.e., they are incapable of determining structure mineralogy);
• counting rules used in conjunction with MI do not distinguish isometric particles
from fibers;
• counting rules used in conjunction with PCM limits counting to fibrous structures
longer than 5 |o,m with aspect ratios greater than 3:1;
• the range of visibility associated with PCM limits counting to fibers thicker than
approximately 0.3 \im;
• under conditions typically employed for asbestos analysis, the range of visibility
associated with SEM limits counting to fibers thicker than approximately 0.1 ^m,
which is only marginally better than PCM;
• SEM is capable of distinguishing asbestos structures from non-asbestos
structures;
• TEM is capable of resolving asbestos structures over their entire size range (down
to thicknesses of 0.01 urn);
• TEM is capable of distinguishing the internal components of complex asbestos
structures; and
• TEM is capable of distinguishing asbestos structures from non-asbestos
structures.
More detailed treatments of the similarities and differences between asbestos techniques and
methods can also be found in the literature (see, for example, Herman and Chatfield 1990).
Due to the differences indicated, measurements from a particular environment (even from
duplicate samples) that are derived using different analytical techniques and methods can vary
substantially and are not comparable. In fact, results can differ by two or three orders of
magnitude (Berman and Chatfield 1990). More importantly, because the relative distributions of
structure sizes and shapes vary from environment to environment, measurements derived using
different analytical techniques and methods do not even remain proportional from one
environment to the next. Therefore, the results from multiple asbestos studies can only be
meaningfully compared if the effects that are attributable to use of differing analytical techniques
and methods can be quantified and reconciled. Few of the existing studies, however, document
analytical procedures in sufficient detail to reconstruct exactly what was done.
4.11
-------
4.4 THE STRUCTURE AND FUNCTION OF THE HUMAN LUNG
4.4.1 Lung Structure
The lungs are the organs of the body in which gas exchange occurs to replenish the supply of
oxygen and eliminate carbon dioxide. To reach the gas exchange regions of the lung, inhaled air
(and any associated toxins) must first traverse the proximal conducting (non-respiratory) airways
of the body and the lung including the nose (or mouth), pharynx, larynx, trachea, and the various
branching bronchi of the lungs down to the smallest (non-respiratory) bronchioles. Air then
enters the distal (respiratory) portion of the lung, where gas exchange occurs.
In humans, the respiratory portion of the lungs are composed of the respiratory or aveolarized
bronchioles, the alveolar ducts, and the alveoli (or alveolar sacs). There are approximately 3x108
alveoli in human lungs (about 20 per alveolar duct) with a cumulative volume of 3.9x103 cm3
(Yeh and Harkema 1993). This represents approximately 65% of the total volume capacity of
human lungs at full inspiration.
Yeh and Harkema (1993) also report that the ratio of lung volume to body weight is
approximately constant across a broad range of mammals (from a shrew, 0.007 kg to a horse,
500 kg). Human lung volumes average a little more than 5 L.
Each human alveolus has a diameter of approximately 0.03 cm (300 urn) and a length of
0.025 cm (Yeh and Harkema 1993). The typical path length from the trachea to an alveolus is
approximately 25 cm. The bronchi leading to each alveolus have branched an average of 16
times from the trachea (with a range of 9 to 22 branches). Importantly, the mean path length, the
number of branches between trachea and alveolus, and the detailed architecture (branching
pattern) of the respiratory region of the lung vary across mammalian species. For example, rats
and mice lack respiratory bronchioles while macaque monkeys exhibit similar numbers of
respiratory bronchiole generations as humans (Nikula et al. 1997). Furthermore, branching in
humans tends to be symmetric (each daughter branch being approximately the same size and the
angle of branching for each is similar but not quite equal) while rodents tend to exhibit
monopodal branching in which smaller branches tend to come off at angles from a main trunk
(Lippmann and Schlesinger 1984).
The gas exchange regions of the lung are contained within the lung parenchyma, which
constitute approximately 82% of the total volume of the lungs (Gehr et al. 1993). Importantly,
the lung parenchyma is not a "portion" of the lung; it fills virtually the entire volume of the
organ traditionally visualized as the "whole" lung. Embedded within the parenchyma are the
larger conducting airways of the lungs and the conducting blood vessels that transport blood to
and from the capillaries that are associated with each alveolus. In the human lung,
approximately 213 ml of blood are distributed over 143 m2 of gas-exchange (alveolar) surface
area (about the size of a tennis court). The gas-exchange surface area of lungs scale linearly with
body weight over most mammalian species. The slope of the "reduced" line (where the Y-Axis is
the ratio of the surface area to the surface area in a reference species and the X-Axis is the ratio
of the body mass to the body mass of the same reference species) is 0.95. Figure 4-1 is a
photomicrograph showing two views of a portion of lung parenchyma in the vicinity of a
terminal bronchiole and an avleolar duct, which are labeled. The circular spaces in the left
4.12
-------
portion of the figure and the cavities in the right portion of the figure are alveoli. Note the
thinness of the walls (septa) separating alveoli.
Figure 4-1. The Structure of Lung Parenchyma Showing Alveoli and Alveolar Ducts
(Source: St. George et al. 1993)
FIG. 10. LM and SEM comparison of the centnacinar region with two different organizations
and B: The short or poorly developed respiratory bronchiole of the rat; C and D, the w
alveolarized respiratory bronchiole of the cat. Terminal bronchiole (TB), respiratory bronchi
(RB). and alveolar duct (AD). For details see ref 1O6.
Confidential: Need Permission to Reproduce this Figure
4.13
-------
Alveoli are separated from each other by alveolar septa that average only a few micrometers in
thickness (Gehr et al. 1993). Gas-exchange capillaries run within these septa and the air-blood
barrier, which averages only 0.62 p.m in thickness, is composed of three layers: the alveolar
epithelium, the interstitium, and endothelium. The epithelium is described in detail below. The
interstitium is primarily composed of a collagenous, extracellular matrix, which constitute about
two-thirds of the interstitial volume. There is also a collection of fibroblasts, macrophages, and
other cells interspersed within the matrix (Miller et al. 1993). The cells of the interstitium
constitute about one third of its volume. The endothelial layer is composed of the smooth
muscle cells and connective tissue that constitute the walls of vascular capillaries. The amount
of connective tissue in the septa between alveoli also varies between animal species; small
rodents have less and primates more (Nikula et al. 1997).
Figures 4-2 and 4-3 show, respectively, a typical alveolar septum and a closeup of one portion of
such a septum. In Figure 4-2, one can see that the septa themselves are thin and are filled almost
entirely with capillaries. Figure 4-3 shows that the epithelial lining of an alveolus is no more
than 1 (j.m thick, that the underlying interstitium is no thicker, and that the endothelium of the
adjacent capillary is similarly thin. These three layers constitute the major tissues of the air-
blood barrier (the rest of the barrier includes the limited quantity of blood plasma between the
endothelial wall of a capillary and a red blood cell and the outer membrane of the red blood cell
itself).
Figure 4-2. The Structure of the Inter-Alveolar Septa (Source: Gehr et al. 1993)
FIG. 1O. Transmission electron micrograph of an alveolus from a dog lung fixed by intravascular
perfusion. The lung was inflated with air at a pressure of 5 cm of water on the deflation limb of the
last of three hysteresis cycles (approximately 6O% TLC). It shows interalveolar septa that are
folded at this degree of inflation. The surface lining layer (SLL) smoothes out every depression of
the alveolar surface (arrows). A, alveoli; C, capillary. x2,1OO. Inset: High power view/ of an
epithelial depression filled with a fluid surface lining layer (SLL). x 1 4.6OO.
Confidential: Need Permission to Reproduce this Figure
4.14
-------
Figure 4-3. The Detailed Structure of an Alveolar Wall that is Part of an Inter-Alveolar
Septum (Source: Gehr et al. 1993)
!
FIG. 19. Fraction of square lattice test grid superimposed on alveolar capillary to measure inter-
cept lengths in tissue (I,) and plasma (lp) for the calculation of the harmonic mean barrier thick-
ness. Note the erythrocyte (EC) in capillary (C) and that the barrier separating blood from alveo-
lar air (A) is made of the three layers epithelium (EP), endothelium (EN), and interstitium (IN).
x 29,000. (From ref. 80.)
Confidential: Need Permission to Reproduce this Figure
4.15
-------
Gehr et al. (1993) also report that interalveolar septa constitute approximately 14% of the
volume of the gas-exchange region of the lung (i.e., the lung parenchyma). Of this tissue mass,
approximately 20% is endothelium, 55% is interstitium, and the rest is composed of cells
associated with the alveolar epithelium. The remainder of the gas-exchange region is air space.
Gehr et al. (1993) also report that the interalveolar septa fold to accommodate changes in lung
volume during respiration, although the major contribution to lung volume changes (over the
range of normal inspiration) appears to be the collapsing (folding) of alveolar ducts (Mercer and
Crapo 1993).
According to Gehr et al. (1993), Type I epithelial cells, which constitute approximately 95% of
the surface area of alveolar epithelium, are flat and platy (squamous), and average less than 1 (am
in thickness (except where cell nuclei exist, which are approximately 7.5 u.m in diameter and
protrude into the alveolar space). Type II epithelial cells, which are cuboidal, constitute no more
than 5% of the epithelial surface. Despite the small fraction of surface that they occupy, Type II
cells serve to maintain the integrity of the overall epithelial lining so that, for example, they limit
the tissue's permeability and control/prevent transport of macromolecules from the interstitium
to the alveolar space, or the reverse (Leikauf and Driscoll 1993). Type II cells also secrete lung
surfactant. The basement membrane of alveolar epithelium is a collagenous structure.
In contrast, the epithelial cells lining the trachea and bronchi (including the respiratory
bronchioles) are ciliated and columnar and averages between 10 and 15 urn in thickness (based
on photomicrographs presented in St. George et al. 1993). Tracheobronchial epithelium
reportedly contains at least 8 distinct cell types (St. George et al. 1993): ciliated epithelium,
basal cells (small flat cells situated along the basal lamina and not reaching the luminal surface),
mucous goblet cells, serous cells, nonciliated bronchiolar (clara) cells, small mucous granule
(SMG) cells, brush cells, and neuroendocrine cells. The relative abundance of the various cells
varies across mammalian species as well as across the various airway generations (branches) and
even the opposing sides of specific airways. The number of cells per length of basal lamina also
varies across mammalian species.
4.4.2 The Structure of the Mesothelium
The mesothelium is a double layered membrane and each layer is a single-cell thick. The two
layers of the mesothelium are separated by a space (e.g., the pleural space), which contains
extracellular fluid and free macrophages (Kane and McDonald 1993). The pleural space is
drained at fixed locations by lymphatic ducts. Each layer of the mesothelium overlies a
collagenous basement membrane containing dispersed spindle cells. Depending on the location
of the mesothelium, the basement membrane may overlie the skeletal muscle of the diaphragm or
the rib cage (in the case of the parietal pleura, which is the outer layer). For the inner layer or
visceral pleura, the basement membrane overlies visceral organs (including the lungs) within the
rib cage. Healthy mesothelium is quiescent, meaning that cells are not actively dividing.
The relative size and thickness of mesothelial tissue varies across mammalian species (Nikula
et al. 1997). For example, rats have relatively thin pleura with limited lymphatic ducts. In
contrast, nonhuman primates have thicker pleura with greater lymphatic drainage than rats and
humans have even thicker pleura and relatively abundant lymphatic ducts.
4.16
-------
4.4.3 Cytology
Alveolar Epithelium. In the respiratory region of the lung, Type II epithelial cells are
progenitor cells for Type I epithelium (Leikauf and Driscoll 1993). Type I epithelial cells are
not proliferation competent (Nehls et al. 1997). After injury to Type I cells, Type II cells
proliferate and reestablish the continuous epithelial surface. Type II cells also secrete surfactant.
It appears that the identity and location of the progenitor cells for Type II epithelial cells are not
currently known.
Injury or alteration of Type II cell function are associated with several diseases included
idiopathic pulmonary fibrosis (Leikauf and Driscoll 1993). Also, crystalline silica and other
toxic agents have been shown to directly modify Type II cellular activity. For example,
crystalline silica (at sub-cytotoxic levels, <100 ug/ml) stimulates Type II proliferation in tissue
culture. In contrast, neither titanium dioxide nor aluminum coated silica induce proliferation at
corresponding concentrations.
In culture, Type II epithelial cells transform slowly into Type I cells and thus have limited
population doubling capacity (Leikauf and Driscoll 1993). Rarely can the number of Type II
cells expand past 3-10 passages (20-30 doublings). During this time, cells terminally
differentiate, develop cross-linked envelopes, and appear squamous, enlarged, and
multinucleated. The process is noted to be accelerated by the presence of transforming growth
factor beta (TGF-P).
Macrophages. Alveolar macrophages (the largest population of macrophages in the lung) are
mobile, avidly phagocytic, present antigens, and release cytokines that trigger various other
immune responses (Leikauf and Driscoll 1993). They also initiate inflammatory responses and
other repair mechanisms that are designed to restore tissue homeostasis.
The next largest population of macrophages in the lung are the interstitial macrophages (Leikauf
and Driscoll 1993). These are localized in the peribronchial and perivascular spaces, the
interstitial spaces of the lung parenchyma, the lymphatic channels, and the visceral pleura. The
various populations of macrophages in the lung express different surface proteins, show different
proliferative capacity, and show differences in metabolism.
The sizes of alveolar macrophages varies substantially across mammalian species (Krombach
et al. 1997). Krombach and co-workers provide a table summary of the relative sizes across
several species of interest:
Animal Mean diameter (u.m) Mean Volume (\nm3)
Rats
Syrian Golden Hamsters
Cyanomolgus Monkeys
Healthy Humans
13.1±0.2
13.6±0.4
15.3±0.5
21.2±0.3
1166±42
1328±123
4990±174
Note: as indicated later (Section 6.2), the relative size of various macrophages has direct
implications regarding the dependence of clearance mechanisms on fiber size.
4.17
-------
Mesothelium. Importantly, mesothelial cells are proliferation competent and may be their own
progenitor cells (Kane and McDonald 1993). It is also possible, however, that as yet-to-be-
identified progenitor cells are located along opposite walls of the pleura or at other locations
within the pleural space.
Tracheo-bronchiolar epithelium. As previously indicated, tracheo-bronchiolar (i.e., non-
respiratory) epithelium is composed primarily of ciliated, columnar cells that are 10 to 15 urn
thick. Although some report that Clara cells serve as progenitor cells for tracheo-bronchiolar
epithelium (Finkelstein et al. 1997), others report that both Clara cells and bronchiolar
epithelium are proliferation competent (Nehls et al. 1997). It appears that the identity and
location of the progenitor cells for Clara cells are not currently known.
4.4.4 Implications
The potential implications of the above observations concerning lung structure and cytology are:
• that the thicknesses of Type I epithelial cells, endothelial cells, and the
interstitium in the alveolar septa are all very small relative to the lengths of the
putative asbestos fibers that contribute to disease;
• that an entire alveolus is only 300 um across and a typical Type I cell is 46 ^m in
radius by <1 \im thick;
• that distances across alveolar septa are only on the order of a few |im and such
septa contain both the endothelium and the interstitum. Thus, the distances that
have to be traversed to get to these structures are also small relative even to the
length of a fiber;
• that the alveolar septa and the walls of the alveolar ducts fold during respiration,
which may provide mechanical forces that facilitate movement of fibers into and
through the alveolar epithelium;
• that Type I epithelium do not proliferate so they cannot be the cells that lead to
cancer. It is the Type II epithelial cells (and potentially macrophages, basal cells,
or endothelial cells) that contribute to cancer in the pulmonary portion of the lung.
Type II cells eventually terminally differentiate to Type I cells;
• that other cells in the lung that have variously been reported to be proliferation
competent (so that they potentially serve as target cells for the induction of
cancer) include Clara cells and bronchiolar epithelial cells;
• that the distance from the most distal airways and alveoli to the pleura is small
relative to the lengths of a fiber; and
• that mesothelial cells are proliferation competent and thus serve as potential
targets for the induction of cancer.
4.18
-------
5.0 THE ASBESTOS LITERATURE
This is a description of the common types of studies in the asbestos literature and an overview of
the sources of potential uncertainty typically associated with each. Such limitations must be
considered when drawing conclusions from these studies and, more importantly, when deriving
inferences based on cross-study comparisons. Throughout this document, we have endeavored
to identify the major sources of uncertainty in the studies we examined and have endeavored to
account for such uncertainties during our evaluation and interpretation of study results.
The types of studies available for examining relationships between risk and asbestos exposure
include human epidemiology studies, human pathology studies, a broad variety of animal
studies, and a broad variety of in vitro studies in both tissue cultures and cell-free systems. To
properly compare and contrast the results from such studies:
• the method(s) employed for asbestos characterization in each study need to be
reconciled;
• the procedures employed for evaluating study endpoints need to be compared and
contrasted;
• the relationship between the route of exposure employed in each type of study and
the exposure route of interest (i.e., human inhalation) needs to be examined; and
• other major, study-specific sources of uncertainty need to identified and
addressed.
Among study conditions and procedures that must be considered before evaluating study
conclusions, it is particularly important to address the analytical methodologies employed to
characterize the nature of exposures (or doses) in each study and such considerations are
common to virtually all of the various types of studies of interest.
As indicated in Section 4.3, the only instrument capable of completely delineating asbestos
structure-size distributions is TEM (or TEM combined with other techniques). Thus, for
example, conclusions regarding variations in biological effects due to differences in such things
as fiber size must be viewed with caution when fiber sizes are characterized using only PCM,
SEM, or other, cruder analytical techniques. Similarly, the ability to adequately determine fiber
mineralogy (fiber type), particularly of what may be minor constituents of various dusts or bulk
materials, also depends strongly on the instrumentation employed for analysis as well as the
strategy for sampling and for conducting the actual structure count. All of these factors must be
considered.
Not only does the specific instrumentation (analytical technique) that is employed in an asbestos
measurement affect the outcome of that measurement, but the particular method employed to
guide the measurement affects the outcome. As previously indicated (Section 4.3), details
concerning the definition of structures to be included in a count, the strategy for counting, and
the minimum number of specific types of structures to be included in a count (all features that
5.1
-------
vary across analytical methods) affect the precision with which fiber concentrations (particularly
of longer and thinner fibers) are delineated. It is not uncommon, for example, that asbestos
concentrations measured in the same sample may vary by several orders of magnitude, due
simply to a difference in the analytical method employed for the analysis (even when the same
analytical instrument is employed, see Section 4.3).
Other important sources of uncertainty tend to be study-type specific and are thus addressed
separately below.
5.1 HUMAN EPIDEMIOLOGY STUDIES
A good overview of the kinds of limitations that contribute to uncertainty in the available
epidemiology studies was presented in the Health Effects Assessment Update (U.S. EPA, 1986).
As described in Appendix A of this document, while evaluating exposure-response factors
derived from the human epidemiology studies, an attempt was made to address most of the major
sources of uncertainty commonly associated with such studies, which are described briefly
below.
Epidemiology studies, which track the incidence of disease (or mortality) within a defined group
(cohort) sharing comparable exposures, have been performed on cohorts of workers exposed to
asbestos and other mineral fibers in a variety of occupational and environmental settings.
Among these, studies that include quantification of exposures are particularly useful for
evaluating exposure-response relationships and deriving risk factors.
Generally, the most severe limitations in an epidemiology study involve the exposure data. Both
estimates of the level of exposure and determination of the character of exposure are affected by
such limitations. Regarding the character of exposure, because exposure measurements from
most of the available quantitative epidemiology studies are based on MI measurements or PCM
measurements, detailed characterization of the size distribution or the mineral type of fibrous
structures (particularly of minor constituents) that contributed to exposure in such studies is
generally lacking (Appendix A). This is particularly important because of the evidence that
neither MI nor PCM are capable of providing measurements that remain proportional (across
study environments) to the biologically-relevant characteristics of an asbestos dust (Berman et
al. 1995). This limits the ability both to compare results across the existing studies and to
extrapolate such results to new environments for which risks need to be estimated. At the same
time, effects on the ability to observe exposure-response trends within a single study are not
typically impaired.
Samples collected prior to the mid-1960s were often analyzed by measuring total dust in units of
millions of particles per cubic foot (mpcf) using impingers or thermal precipitators. A
description of the relative strengths and weaknesses of these techniques is provided in
Section 4.3. The fibrous portion of the dust was not monitored. Impinger measurements are
sometimes related to fiber counts (based on PCM) using side-by-side measurements of total dust
and fiber counts collected during a relatively brief period of time (e.g., Dement et al. 1983a;
McDonald et al. 1980b). However, the correlation between fiber counts and total dust is
sometimes poor within a plant (i.e., a single study environment) and generally poor between
plants (see, for example, U.S. EPA 1986). Thus, conversions based on limited sets of paired
5.2
-------
measurements are of questionable validity. In some studies (e.g., McDonald et al. 1983b) the
only available measurements are MI measurements (in mpcf) and these have been related to f/ml
by PCM using conversion factors derived in other plants, which raises further questions
concerning validity.
Even if all measurements could be adequately converted to PCM, this may still not be adequate
for assessing risk in a manner that allows extrapolation across exposure environments or studies.
Comparing exposure-response factors derived in different exposure environments (or
extrapolating to new environments to predict risk) requires that asbestos measurements reflect
the characteristics of asbestos structures (size, shape, mineralogy) that determine biological
activity. If surrogate measures are employed (e.g., measures of asbestos structures displaying
characteristics other than those that determine biological activity), there is no guarantee that
concentrations of such surrogate measures and the true biologically active structures will remain
proportional from one environment to the next. As a consequence, the relationship between
exposure (measured by surrogate) and risk may not remain constant from one environment to the
next. Importantly, several studies suggest that PCM may, at best, be no more than a surrogate
measure (see, for example, Berman et al. 1995). Moreover, the technique was adapted to
asbestos in an ad hoc fashion with only limited thought given to biological relevance (Walton
1982).
Use of surrogate measures of asbestos exposure may be less of a problem within a single
exposure environment (where airborne asbestos structures likely have been generated in a similar
manner from similar source material). Thus, surrogate measures of asbestos exposure may
remain approximately proportional to the true biologically active structures, which suggests why
monotonically increasing exposure-response relationships have likely been observed with PCM-
measured concentrations in single exposure environments. In different exposure environments,
however, the distribution of fiber sizes and types of airborne asbestos structures are likely
different, since they are generated in different processes from different source material. There is
thus little reason to expect surrogate measures of exposure to remain proportional across such
environments.
None of the published epidemiology studies incorporate TEM measurements of asbestos and
such measurements are not widely available in occupational settings (Appendix A). However,
TEM is the method currently used (and recommended) to assess exposure in environmental
settings, due both to questions concerning biological relevance (Berman et al. 1995 and
addressed in detail in Chapter 6) and to problems with measuring environmental asbestos
concentrations by PCM (Section 4.3).
In some cases, the limited exposure characterization presented in specific epidemiology studies
can be augmented by pairing such studies with published TEM characterizations of dusts from
the same or similar exposure settings, to the extent the appropriate supplemental studies are
available. In fact, this is the procedure adopted in this document to adjust the existing risk
factors to exposure indices that are thought to better relate to biological activity (described in
detail in Section 7.4). Such an approach is limited, however, to the extent that the published
asbestos characterizations actually represent exposure conditions in the corresponding
epidemiology studies. To the extent reasonable, the limitations of this approach have been
5.3
-------
addressed in this study by assigning and incorporating additional factors into the calculation of
uncertainty intervals (defined in Appendix A) that are associated with the adjusted potency
factors.
Regarding levels of exposure in the epidemiology studies, in most cases, air measurements were
collected only infrequently and measurements may be entirely lacking from the earliest time
periods, when exposures may have been greatest. In such cases, exposures are typically
estimated either by extrapolation from periods when measurements are available or by expert
judgement based on personal accounts and records of changes in plant operations, industrial
hygiene procedures, air standards, etc. Moreover, the majority of exposure measurements used
in these occupational studies are based on area (ambient) rather than personal samples.
Typically, only a few areas of a plant have been sampled so that levels in other areas must be
approximated using expert judgement by persons familiar with operations at the plant.
It is difficult to judge the degree that available asbestos concentration measurements are
representative of actual exposures in the existing studies. In some cases, it seems likely that
operations were shut down or otherwise modified in preparation for sampling. Likewise, in
some operations there are brief episodes of very intense exposure and it is questionable whether
such episodes are adequately represented in the available data.
Most of the asbestos measurements used in the published epidemiology studies were collected
for insurance or compliance purposes. They were not intended to provide representative
estimates of the direct level of exposure to workers. Some of the published epidemiology studies
lack any direct exposure data. For example, exposures were estimated for the cohort studied by
Seidman (1984) based on conditions simulated many years later in a similar plant to the one
from which Seidman studied the original cohort. In fact, the equipment in the plant from which
Seidman obtained exposure estimates came originally from the plant where Seidman studied the
workers; it was purchased and moved. Recently, an epidemiology study was also completed for
a cohort working at that new plant (Levin et al. 1998).
In addition to problems with the actual analysis of asbestos concentrations, individual exposures
are generally estimated in the existing epidemiology studies by relating ambient asbestos
measurements to job descriptions and integrating the duration of exposure over the recorded time
that each worker spent in each job category. However, sometimes there are no records of
specific areas in which an employee worked, so that work areas must be assumed based on job
title. Some types of workers (e.g., maintenance workers) may have spent time in many different
areas of a plant so their exposure varies from what might otherwise be assumed.
Although the greatest problems with the data in existing epidemiology studies likely lies within
the estimates of exposure, problems with disease-response data also exist. Mesothelioma is rare
and this disease may have been under-reported as a cause of death in older studies. This is
probably less of a problem in more recent studies, since the association of mesothelioma with
asbestos exposure is now well known. In fact, the opposite tendency (over-reporting) may now
be occurring because of increased sensitivity by examiners (an asbestos worker with
mesothelioma is now more likely to be eligible for compensation). Some studies have
re-diagnosed causes of death from all of the available data (e.g., Selikoff et al. 1979); however,
5.4
-------
this creates the problem of lack of comparability to control populations (for which such re-
diagnosis is not generally performed).
The choice of an appropriate control population is also an important consideration. Local cancer
rates may differ substantially from regional or national rates and the choice of an appropriate
control is not always clear. A related problem is the lack of smoking data in many of the studies.
Because of the interrelation between smoking and asbestos in lung cancer, errors could occur in
lung cancer risk estimates if the smoking patterns of the cohort are substantially different from
those of the control population.
In some of the studies, a substantial portion of the population is lost to follow-up (e.g.,
Armstrong 1988), and this adds additional uncertainty to the analysis. Also, the effect of
exposure may be inaccurately evaluated if the follow-up of the population is too brief. This may
be a limitation, for example, of the Levin et al. (1998) study.
Another problem frequently associated with these studies is that available data are not reported
in a form that is well-suited to risk assessment. The EPA lung cancer model, for example,
requires that exposure be estimated as cumulative exposure in f/ml-years excluding the most
recent 10 years (U.S. EPA 1986, also described Section 7.2); generally the data are not published
in this form. The data are also frequently not available in a form that permits study of the shape
of the lung cancer exposure-response curve, so it is not possible to determine how well the EPA
model describes the data. The reporting of the mortality data for mesothelioma is generally even
less appropriate for risk assessment. Ideally, what is needed is the incidence of mesothelioma
subdivided according to exposure level, age at beginning of exposure, and duration of exposure
(U.S. EPA 1986, also described in Section 7.3). Such data are almost never available in
published studies and crude approximations must be made to account for this lack of
information.
It is important to understand the type and magnitude of effect that each of these sources of
uncertainties are likely to have on the distribution of potency estimates derived from the set of
available studies for lung cancer and mesothelioma, respectively. Some of the above-described
limitations likely introduce random errors that simply decrease the overall precision of a potency
estimate. However, other types of limitations may cause systematic errors in particular studies,
which potentially bias the potency estimate either high or low. Some of the limitations may only
affect between-study comparisons and some may introduce a systematic bias between either
industry types or fiber types. Examples of some of these types of variation are provided in
Section 7.1.
We also note that, although individual estimates of potency factors from individual studies may
be highly uncertain, by combining results across multiple studies while properly addressing such
uncertainties, it may be possible to draw conclusions with greater precision than reasonable for
any individual study. This is the essential advantage of the type of meta analysis discussed in
this document (see Chapter 7).
5.5
-------
5.2 HUMAN PATHOLOGY STUDIES
Human pathology studies provide a characterization of disease morphology and correlations
between causes of death and the types of asbestos fibers retained in the lungs and other bodily
tissues. These studies generally involve microscopic examination of tissue samples for
indications of morphologic changes characteristic of disease and/or microscopic examination of
digested tissue specimens to characterize the mineral fibers extracted from such tissue.
The results of human pathology studies need to be evaluated carefully by addressing effects that
are attributable to:
• the way tissue samples are fixed for preservation;
• the way tissue samples are prepared for analysis (e.g., ashing, bleach digestion,
digestion in alkali, or some combination);
• the choice of methods employed for characterization of asbestos; and
• the choice of locations within tissues from which samples are collected for
analysis.
Because tissue samples obtained from deceased individuals are typically stored for long periods
of time before they may be analyzed as part of a human pathology study, such samples are
commonly fixed by treatment with chemical preservatives prior to storage. However, Law et al.
(1991) studied the effects of two common fixatives (Karnovsky's fixative and formalin fixative)
on asbestos fibers and concluded that such fixatives degrade and dissolve asbestos fibers
(including both chrysotile and crocidolite) at measurable rates. Therefore, particularly for
samples that are stored "wet" (as opposed, for example, to storage in paraffin blocks), the
concentrations and character of the tissue burden of asbestos may be altered during storage.
Even for studies in which relative (as opposed to absolute) concentrations are being compared,
alterations associated with preservation may limit the ability to make such comparisons,
particularly among samples stored for widely disparate periods of time or stored using widely
disparate procedures.
Fiber concentrations in tissue samples have also been shown to vary as a function of the method
employed for preparing such samples for analysis. Historically, samples that are digested in
bleach or alkali have tended to exhibit lower recovery of asbestos fibers than samples that are
ashed. However, more recent studies suggest that improving technique has narrowed these
differences so that this is no longer a major consideration . Thus, when comparing results across
studies, due consideration needs to be given for the time frame during which such studies were
conducted and the comparability/differences in the techniques employed for tissue sample
preparation.
Once prepared, both the character and the concentration of the tissue burden measured in a tissue
sample will also depend heavily on the particular analytical method employed to characterize
asbestos and differences attributable to such techniques must be reconciled before measurements
across samples or conclusions across studies can be reasonably compared. A more detailed
5.6
-------
description of the effects attributable to asbestos measurement was presented in the previous
section on human epidemiology studies (Section 5.1) and the same issues obtain for human
pathology studies. Unless measurements are made using comparable instrumentation with
comparable methodology, comparisons across such measurements can be very misleading.
Perhaps the biggest limitation hindering the kinds of evaluation that can be conducted based on
human pathology studies is that due to the strong dependence of asbestos concentrations on the
specific location within a tissue from which a sample is obtained. Numerous authors have
reported that asbestos is non-uniformly distributed in lung parenchyma and other tissues
following exposure (see, for example: Bignon et al. 1979; Davis et al. 1986a; Pooley 1982). The
incidence of lesions and other pathological effects attributed to asbestos exposure
correspondingly exhibit a non-uniform distribution.
For lung tissue samples (which tend to be among the primary interests in human pathology
studies) the relationship between sample location and asbestos concentration is particularly
important. To sample deep lung tissue reproducibly, it has been shown necessary to select a
specific section of lung parenchyma from a defined portion of the bronchio-alveolar tree.
Pinkerton et al. (1986) showed that the deposition of asbestos in the lungs is an inverse function
both of the path length and the number of bifurcations between the trachea and the site. Thus,
analyses of samples from different animals of the same species can only be compared
meaningfully if the samples are collected from identical locations in the bronchio-alveolar tree.
Similar, nonuniform depositional patterns have also been observed in humans (Raabe 1984).
Furthermore, due to the complex branching and folding pattern of the lung, adjacent sections of
lung parenchyma frequently represent disparate portions of the bronchio-alveolar tree (Brody
et al. 1981; Pinkerton et al. 1986). Consequently, lung burdens derived even from adjacent
samples of lung parenchyma can show broadly varying concentrations (differing by orders of
magnitude).
Unfortunately, however, tissue samples that are available for analysis in support of a human
pathology study are typically "opportunistic" samples, which means that they were selected and
stored for an entirely different purpose than the study at hand and, although there may sometimes
have been attempts to sample comparable locations across lungs in a general way, this is not
adequate for assuring that comparable portions of the respiratory tree are being sampled. It-is
therefore seldom possible to address the effects of sample location directly. Consequently,
comparisons of tissue burden concentrations across samples from different individuals in a
human pathology study are at best qualitative and may only be useful when averaged over large
numbers of individuals and only when large differences in concentrations (several orders of
magnitude) are being distinguished. Moreover, because the parts of a tissue that undergoes
morphologic changes induced by asbestos typically corresponds to the parts of a tissue where
asbestos burdens are the highest, even comparison of morphologic effects across tissue samples
requires proper consideration of the effects of the locations from which such tissue samples were
derived.
5.7
-------
5.3 ANIMAL STUDIES
In animal studies, members of one of various species (generally rodents) are exposed to
measured doses of size-selected mineral fibers and the resultant biological responses are
monitored. Animals may be dosed either by inhalation, ingestion, intratracheal installation,
implantation, or injection (U.S. EPA 1986). Such studies are conducted for several purposes.
As with human pathology studies, animal pathology studies are those in which the transport of
asbestos structures is tracked through the various organs and tissues of the animal and the
attendant cellular and molecular changes are characterized. In parallel with quantitative
epidemiology studies, animal dose/response studies track the incidence of disease among a
population that has been exposed in a controlled manner. One of the advantages of animal
dose/response studies over epidemiology studies is that exposures are controlled and can be well
characterized. The major disadvantage is that there are many uncertainties introduced when
extrapolating the results of animal data to predict effects in humans. Therefore, attempts to
adapt such things as animal-derived dose-response factors to humans are not generally
recommended.
As with human epidemiology and pathology studies, the validity of conclusions drawn from
animal studies depends strongly on the techniques and methods used to characterize and quantify
asbestos structures either in the delivered dose or in the tissues of the dosed animals (see
Section 5.1). The ability to reconcile conclusions derived from many animal studies with the
rest of the asbestos literature is limited because SEM was commonly employed to measure
asbestos in animal studies, but not other kinds of studies. Even many of those studies in which
TEM was employed for asbestos analysis suffer from use of non-standard methods that cannot
be easily reconciled with the more traditional TEM methods, particularly because such
specialized methods are seldom adequately documented to allow comparison.
As with human pathology studies, the location of a tissue sample excised for analysis is a critical
factor that also governs the quality of an animal pathology study (Section 5.2). However, one
potential advantage frequently available in animal pathology studies over human studies is the
ability to carefully identify and select the precise tissue samples to be analyzed. The extent that
a particular animal pathology study exploits this capability can affect the overall utility of the
study. Thus, such issues need to be addressed carefully when evaluating and comparing the
results across animal pathology studies.,
The route of exposure employed in a particular animal study is also important to consider. Each
of the routes of exposure commonly employed in these studies (inhalation, ingestion,
intratracheal installation, and injection or implantation) delivers different size fractions of
asbestos to a target tissue with varying efficiencies. For example, injection or implantation
studies deliver 100% of all size categories of structures to the target tissue. However, the
efficiency that each size category is delivered by inhalation is a function of the aerodynamic
properties of the asbestos structures and the air flow characteristics of the lungs (see, for
example, Yu et al. 1991, 1994). Thus, the relationship between dose and exposure depends upon
the route of exposure employed. Importantly, because the ultimate goal is to understand the
effects that inhaled fibers have on humans, differences between the character of the delivered
dose in an animal study and the character that such a dose would have, had it originally been
inhaled, typically need to be addressed.
5.8
-------
Regarding the measurement of health effects, many of the results in animal studies suffer from a
lack of statistical significance because of small numbers of observed tumors. Consequently,
trends cannot be established conclusively.
For animal inhalation studies, meaningful comparison of the relative deposition of asbestos dusts
across species is not direct. To extrapolate results across species, the detailed differences in the
physiology of the respiratory tracts between the species need to be addressed (Section 4.4).
However, if measurements are available for both species, differences in physiology are
addressed, and the manner in which tissue burdens are analyzed is considered, it may be possible
to compare relative tissue doses (mass of asbestos per mass of tissue) across species.
5.4 IN VITRO STUDIES
A broad range of in vitro studies provide useful insight on the effects of asbestos. These include,
for example, studies in cell-free systems (which have been used to evaluate such things as
asbestos dissolution rates or the kinetics of free radical formation on the surface of asbestos
fibers) and studies of the effects of asbestos on cultures of a broad variety of cell types and
tissues.
As with other studies, the potential limitations and sources of uncertainty associated with in vitro
studies need to be considered when evaluating the validity of study results or comparing such
results to those of other studies, particular studies of varying type. Also, as with other studies,
among the primary sources of uncertainty that need to be addressed for in vitro studies is the
manner in which asbestos doses are characterized and quantified (Section 5.1). For in vitro
studies (as with animal studies dosed by routes other than inhalation), this also extends to the
need to consider the relationship between the character of the asbestos dose applied in vitro and
the character that a similar exposure might possess following inhalation exposure in vivo
(Section 5.3). This can be particularly problematic for studies in tissue cultures because it is not
clear how the application of a suspension of fibers (with known concentration) to a dish
containing cultured cells can be related to doses that reach corresponding tissues following
administration to whole animals.
In vitro studies, of necessity, represent isolated components of living systems observed under
conditions that may vary radically from those under which such components operate in vivo.
Consequently, the behavior of such components may also vary radically from the behavior of the
same components in vivo. Therefore, additional study-specific considerations (concerning the
design of a study and the conditions under which a study is conducted) also need to be addressed
before evaluating the validity or relevance of the results from an in vitro study to what might
otherwise be observed in a whole animals. Examples of such considerations include:
For cell-free systems
• whether conditions under which the study is conducted are sufficiently similar to
conditions in vivo to expect that the observed effect is likely to occur in vivo; and
5.9
-------
• if the observed variables describing the nature or magnitude of the effect are also
likely to reflect what may occur in vivo;
For tissue cultures
• whether conditions under which the study is conducted are sufficiently similar to
conditions in vivo to expect that the observed effect is likely to occur in vivo;
• whether responses by specific tissues or cells in culture are likely to behave
similarly in vivo where their behavior may be suppressed, enhanced, or modified
in some other manner due to additional stimuli provided by responses of other
tissues and cells that are components of a complete organism but that may be
lacking in culture; and
• whether the conditions required to establish and maintain a tissue culture for
experimentation sufficiently alters the characteristics and behavior of the cells
being studied to minimize the relevance of results from such a study to conditions
in vivo.
Among the most important examples of the last consideration relates to the general need to
create immortalized cells to maintain tissue cultures. Thus, questions must always be raised
concerning whether the alterations required to create immortalized cells for culture (from what
are normally mortal cells in vivo) also alter the nature of the responses being studied.
Note, many studies in the current literature also incorporate combined aspects of several of the
four general study types described in this chapter. For these studies, a corresponding
combination of the considerations described must therefore be addressed when evaluating such
studies and comparing their results with inferences derived from the rest of the literature.
5.10
-------
6.0 SUPPORTING EXPERIMENTAL STUDIES
To evaluate the plausibility of cancer risk models for asbestos it is useful to examine the current
state of knowledge regarding (1) the mechanisms that facilitate the transport of asbestos to the
various target tissues of interest (i.e., the lungs and mesothelium) and (2) the mechanisms that
contribute to the development of cancer in these target tissues. Accordingly, a review of the
relevant literature is provided in this chapter to assure that the quantitative analysis in Chapter 7,
and the proposed approach for assessing asbestos-related risks in Chapter 8, are qualitatively
consistent with general implications from the broader literature. This review, although
extensive, is not exhaustive. A much larger number of studies was reviewed than are actually
cited. Studies that are not cited are largely confirmatory or redundant. In selecting studies for
review, special effort was expended to assure that opposing views (particularly for controversial
issues) were adequately represented.
Although much progress has been made over the last decade toward elucidating the fiber/particle
mechanisms that contribute to transport and subsequent cancer induction, at least two critical
data gaps remain:
• no one has yet been able to track a specific lesion induced by asbestos in a
specific cell through to the development of a specific tumor. There have been
experiments that show altered DNA and other types of cellular and tissue damage
that are produced in association with exposure to asbestos. Other studies have
demonstrated that various tumors of the kinds that result from asbestos exposure
exhibit patterns of DNA alteration (or other kinds of cellular damage) that are
sometimes (but not always) consistent with the earlier cellular changes associated
with asbestos exposure. There are also studies that show that exposure to
asbestos can lead ultimately to development of tumors. However, these types of
studies have yet to be linked; and
• the specific target cells that serve as precursors to tumors in various target tissues
are not known with certainty.
Because of the first of the above limitations, researchers have tended to report on a broad range
of tissue and cellular effects induced by asbestos that may lead generally to various kinds of
cellular damage or injury. Cytotoxicity, for example, is one of the endpoints typically tracked as
a marker for asbestos-induced injury. However, not all of these effects necessarily contribute
(either directly or indirectly) to the development of cancer. Therefore, one of the goals of the
following discussion is to distinguish among effects that likely contribute to the development of
cancer from those that are less likely or unlikely to contribute. Of course, delineating such
distinctions are subject to the limitations of the current state of knowledge.
In addition, because the relative effects of fiber size, shape, and mineralogy need to be elucidated
to better indicate how asbestos concentrations should be characterized to support risk
assessment, studies that address these topics are highlighted. Of particular interest are studies
that (1) contrast the effects of different sized fibers, (2) contrast the effects of fibers and non-
6.1
-------
fibrous particles of similar mineralogy, and (3) contrast the effects of fibers of comparable
morphology (size and shape), but differing mineralogy.
The types of studies that have contributed to the state of knowledge of the effects of asbestos (in
addition to the human epidemiology studies that are evaluated in Chapter 7) include:
whole animal inhalation studies;
whole animal instillation studies;
whole animal injection, implantation studies;
human pathological studies;
in vitro studies in cell cultures; and
in vitro studies in cell-free systems.
Depending on the outcome(s) monitored, the animal studies may alternately be categorized as
retention studies, histopathology studies, or dose-response studies.
Each type of study possesses certain advantages and exhibits certain limitations, which have
previously been described (Chapter 5) along with descriptions of the nature of each of these
study types. In addition to the advantages and limitations that are attributable to the type of
study, the quality of the characterization of asbestos (or other particulate matter) determines the
utility of the study for addressing issues associated with fiber morphology and mineralogy.
Unfortunately, for many published studies, both the characterization of the asbestos (or other
particulate matter) and descriptions of the manner in which such materials were handled are
insufficient to establish the detailed morphology or mineralogy. Such limitations need to be
considered when comparing across study results or evaluating the validity of study conclusions.
The rest of this chapter is divided into separate sections that address the set of factors that have
been previously identified (Chapter 3) as those that determine the biological activity of inhaled
asbestos (which is the exposure route of primary concern for humans). These are:
• the extent that asbestos structures are respirable and the pattern of deposition of
inhaled structures;
• the extent that deposited structures are subsequently cleared or degraded;
• the extent that deposited structures are transported or migrate to the various target
tissues; and
• the extent that retained structures induce a biological response in each target
tissue.
6.1 FACTORS AFFECTING RESPIRABILITY AND DEPOSITION
Discounting systemic effects resulting from other forms of exposure, factors affecting
respirability are common to all of the toxic endpoints associated with asbestos exposure
considered in this study (asbestosis, lung cancer, and mesothelioma). Moreover, respirability is
common to the factors affecting the toxicity of inhaled, insoluble particles in general. To be
6.2
-------
respirable, an inhaled particle must pass the blocking hairs and tortuous passageways of the nose
and throat and be deposited in the lungs. Particles deposited in the naso-pharyngeal portion of
the respiratory tract are not considered respirable.
Not all of the inhaled particles that reach the lungs will be deposited. Small particles may not
impact lung surfaces during inhalation and are subsequently exhaled. Once a particle impacts on
a surface, however, it is likely to remain because the surfaces of the lungs are wetted with a
surfactant (Raabe 1984).
Adverse health effects potentially result when particles that are deposited in the lungs remain in
contact with the tissues in the lung for a sufficient period of time to provoke a biological
response. To affect the mesothelium, an offending particle may also need to migrate or be
transported from the lung to this surrounding tissue. However, due to the proximity of the
mesothelium to peripheral portions of the lung parenchyma (which include locations where
particles are typically deposited), it is also possible that diffusable molecules produced in lung
tissue in response to deposited particles can have an adverse effect on the mesothelium (see, for
example, Adamson 1997). Such effects are considered among the mechanisms of disease
induction addressed in the discussion of biological responses (Section 6.3).
The interplay between deposition and removal (clearance) of inhaled particles is an important
determinant of biological activity and separating the influence of these two processes in the
pathology of asbestos-induced disease is difficult. The term "retention" is used here to represent
the fraction of particles remaining in the lungs beyond the time frame over which only the most
rapid removal processes (i.e., muco-ciliary clearance) are active. The factors affecting retention
are addressed further in Section 6.2.
Published inhalation studies divide the respiratory tract into three units (see, for example, Raabe
1984). The naso-pharyngeal portion of the respiratory tract extends from the nares in the nose
through the entrance to the trachea. The tracheo-bronchial portion of the respiratory tract
includes the trachea and all of the branching bronchi down to the terminal bronchioles. The
respiratory bronchioles and the alveoli, which are collectively referred to as the "deep lung", are
the bronchio-alveolar (or pulmonary) portion of the respiratory tract. For a more detailed
description of the features of the respiratory tract, see Section 4.4.
The dimensional requirements for respirability have been studied and reviewed in several studies
(see, for example, Raabe 1984 or U.S. EPA 1986). A more recent review is also presented by
Stober et al. (1993). Much of the data reviewed in these studies is based on earlier studies in
which researchers exposed animals or human volunteers to a series of monodisperse spherical
particles (although Stober et al. also reviews fiber experiments). In this manner, the impact of
the diameter of spherical particles on respirability was elucidated. The respirability of fibrous
materials (such as asbestos) tends to be described in terms closely associated with those
employed for spherical particles, but with adjustments for density and shape. Importantly,
because respirability is a mechanical process: the size, shape, and density of a particle (or fiber)
determine its respirability along with the morphometry of the airways through which the particle
passes (Stober et al. 1993). Other than affecting the particle's density or the distribution of fiber
shapes, the chemical composition (mineralogy) of a particle (or fiber) does not influence
respirability.
6.3
-------
The respirability of particles and fibers by humans, and a variety of other mammals of
experimental interest, has also been the subject of increasingly sophisticated modeling efforts
(Stober et al. 1993). The latest refinements of such models predict particle deposition with a
degree of accuracy that is beyond what can be validated with existing, experimental data. The
application of several of these models to asbestos (and other fibrous materials) are considered
throughout this chapter. However, a detailed overview of the state-of-the-art of such modeling is
beyond the scope of this document. Such an overview is presented by Stober et al. (1993).
6.1.1 Respirability of Spherical Particles
Spherical particles larger than 10 \im in diameter are considered non-respirable because virtually
all particles in this size range are trapped in naso-pharyngeal passageways and blocked from
entering the lungs. As the diameter of the particles fall, an increasing fraction traverses the nose
and throat and may be deposited in the lungs. About half of particles 5 (am in diameter are
blocked before entering the lungs. Virtually all particles smaller than 1 urn enter the lungs,
although other factors determine whether they are in fact deposited or simply exhaled.
Figure 6-1 (Raabe 1984) is a representation of the relative deposition in the various
compartments of the respiratory tract as a function of particle diameter.
Within the lungs (Figure 6-1), the greatest fraction of respirable particles (over the entire range
of diameters down to <0.01 (j,m) are deposited in the deep lung (the broncho-alveolar portion of
the respiratory tract), primarily at alveolar duct bifurcations (see, for example, Brody et al. 1981;
Davis et al. 1987; Johnson 1987; Sussman et al. 1991a). These studies also indicate that
biological responses appear to be initiated where deposition is heaviest. Generally, the fraction
of particles deposited in the deep lung increases regularly with decreasing diameter until a
maximum of 60% deposition in the deep lung is reached at about 0.1 um diameter.
As indicated in Figure 6-1, a transition occurs at particle diameters between 0.5 and 1 urn. For
particles in this range and smaller, deposition in the deep lung competes primarily with
deposition in the tracheo-bronchial tree and with exhalation; smaller particles have an increasing
probability of being exhaled without ever impacting the surface of an air passageway. For
particles larger than this transition range, broncho-alveolar deposition is limited chiefly by the
fraction of particles that are removed from the air stream prior to reaching the deep lung (either
by deposition in the naso-pharyngeal or the tracheo-bronchial portions of the respiratory tract).
The transition between naso-pharyngeal competition with deep-lung deposition and competition
from other removal processes is important because, during mouth breathing, a process that
bypasses the tortuous pathways of the nose and throat, it has been observed that larger particles
(up to several micrometers in diameter) may be deposited in the deep lung (Raabe 1984).
Studies of the effects of mouth breathing are also reviewed by Stober et al. (1993). Because
most people spend at least small amounts of time mouth breathing, especially during exertion or
while snoring, this mechanism for allowing larger particles to settle in the deep lung should not
be ignored.
6.4
-------
Figure 6-1. Fractions of Respirable Particles Deposited in the Various Compartments of
the Human Respiratory Tract as a Function of Aerodynamic Equivalent Diameter8'11
i i in 1 i .( i/i
0.01
0,1
i i Yiftfir 'MI
1.0
10.0
Diameter of Spherical Particle with Density Equal to One g/cm (um) c
"Source: Raabe 1984
bAssumes a typical tidal value of 1,450 cm3 and a rate of 15 breaths a minute
'Aerodynamic equivalent diameter
Confidential: Need Permission to Reproduce this Figure
A diameter of 0.5 urn also happens to represent the transition between the regime where inertial
flow becomes the major factor controlling deposition in the lungs and the regime where
diffusional flow dominates. Below the 0.5 |j,m transition, the diffusional diameter becomes more
important in determining deposition than the aerodynamic equivalent diameter (defined below).
6.1.2 Respirabiliry of Fibrous Structures
Several authors have investigated the effect of the shape of non-spherical particles (including
fibers) on respirability and deposition (see, for example, Harris and Timbrell 1977; Strom and
Yu 1994; Sussman et al. 1991a,b; Yu et al. 1995a,b). It has been found that the behavior of non-
spherical particles can be related to the behavior of spherical particles by introducing a concept
known as the aerodynamic equivalent diameter. The aerodynamic equivalent diameter is the
6.5
-------
diameter of a hypothetical spherical particle of unit density that would exhibit the same settling
velocities and aerodynamic behavior as the real, non-spherical particle of interest. Factors that
affect the aerodynamic equivalent diameter are density, true diameter, true length (for elongated
particles such as fibers), and the regularity of the particle shape.
Harris and Timbrel! (1977) Findings. Because fibrous particles tend to align primarily along
the axis of travel under the flow conditions found in the lungs, respirability is predominantly a
function of the diameter of a fiber and the effect of length is secondary (Harris and Timbrel!
1977). Fibrous structures with aspect ratios (ratio of length to width) >3:1 behave like spherical
particles (of similar density) with diameters up to 3 times larger and exhibit only a very weak
dependence on length. As previously indicated, however, the aerodynamic equivalent diameter
of a fibrous structure must also be adjusted for the effects of density. This is demonstrated in
Figure 6-2 where the true diameter of a fiber is graphed on the top horizontal axis against
spherical (aerodynamic equivalent) diameters on the bottom horizontal axis. Figure 6-2 is an
overlay of Figure 6-1. Note that, to adjust for the density of asbestos, the true diameters listed in
the figure have been shifted to the right of where they would appear if the relationship was
exactly 1/3 of the aerodynamic equivalent diameter.
Two vertical dashed lines in Figure 6-2 represent effective limits to the range of respirable
asbestos. The line on the left side in the figure represents the limiting diameter of the smallest
chrysotile fibril (about 0.02 urn true diameter) and thus represents a lower limit to the diameter
that is of concern when considering asbestos. The vertical line to the right represents the cutoff
where deposition in the deep lung becomes unimportant due to removal of such particles by the
naso-pharyngeal passageways. This latter cutoff corresponds to a true fiber diameter of 2.0 \im,
which theoretically represents the upper limit to the size of asbestos that is respirable. As
indicated in the figure, however, deposition in the deep lung drops precipitously for fibers
thicker than about 0.7 [im so that no more than a few percent of asbestos fibers thicker than
approximately 1 urn actually reach the deep lung.
Harris and Timbrell (1977) also evaluated the relationship between the overall shape of a particle
and the extent of deposition. Over the range of diameters that potentially represent the range of
asbestos fibers likely to be encountered, pulmonary deposition decreases with increasing
complexity of shape beyond simple cylinders (such as clusters and matrices (see Section 4.2) at
the expense of increasing naso-pharyngeal or tracheo-bronchial deposition. This change also
becomes increasingly important as the length of the structure increases. For structures <25 pm
in length, the difference in deposition between simple fibers and complex clusters or matrices
may vary by up to a factor of 2 with the complex structures being more likely to be removed in
the naso-pharyngeal portion of the respiratory tract and the fibers more likely to be deposited in
the deep lung. At 100 jam lengths, the fraction of complex structures that survive passage
through the nose and throat in comparison with simple fibers may vary by a factor of 5. This
means that large structures become relatively less respirable as their complexity increases.
However, during mouth breathing large clusters and matrices may enter the deep lung.
6.6
-------
Figure 6-2. Fractions of Respirable Particles Deposited in the Various Compartments of
the Human Respiratory Tract as a Function of the True Diameter of Asbestos Fibers"'M
.002
The Diameter of Asbestos Fiber
01 2 .05 .1 ,2 .5
t,o
XJ
o>
'ro
o
fj
CO
0,01
1.0 10.0
Diameter of Spherical Particle wfth Density Equal to One o/cm (urn)c
"Source of original: Raabe 1984
bAssumes a typical tidal value of 1,450 cm3 and a rate of 15 breaths per minute
The relationship between true diameters and aerodynamic equivalent diameters derived from
Harris and Timbrell (1977). Diameters adjusted for shape and density of asbestos fibers.
dAerodynamic equivalent diameter
Confidential: Need Permission to Reproduce this Figure
When all of the factors that Harris and Timbrell (1977) addressed are considered, the efficiency
of the deposition of asbestos structures in the deep lung is maximal for short, thin, single fibers
(<10 |im in length with a true diameter <0.7 urn). The efficiency decreases slowly with
increasing length (up to an effective limit of 200 |im), moderately with increasing complexity of
shape, and rapidly with increasing diameter (up to an effective limit of 2.0 urn, true diameter).
Thinner fibers, down to the lower limit of the range for asbestos fibers (0.02 urn, true diameter),
are deposited with roughly the same efficiency. Approximately 20-25% of the fibers between
0.7 and 0.02 |j.m in diameter (and <10 (im in length) are deposited in the deep lung.
6.7
-------
Sussman et al. (1991a,b) Findings. Based on a series of experiments on human tracheal
bronchial casts, Sussman et al. (1991a,b) also developed models of fiber deposition in the human
lung. Such experiments are in fact illustrative of several research groups who have developed
deposition models based on results from experiments on airway casts (for a review, see Stober et
al. 1993).
The results reported by Sussman et al. (1991a,b) appear to be generally consistent with the
results reported by Harris and Timbrell (1977) and Yu and coworkers (described below),
although the manner in which their results are reported make them somewhat less directly
comparable. Briefly, Sussman et al. (1991a,b) report that deposition increases along most
generations of the bronchial tree with increasing fiber length and increasing airflow rate for any
fixed aerodynamic diameter. This increased deposition efficiency is demonstrated for airway
generations at least through the ninth bifurcation and is implied to continue through to airway
generations that would be representative of the respiratory (pulmonary) portion of the lung (i.e.,
airway bifurcations greater than approximately 16 to 22). For definitions and a description of
airway generations, see Section 4.4.
Findings of Yu and Coworkers. In a series of studies, Yu and coworkers combined an
improved model of human lung physiology (Asgharian and Yu 1988) with a series of more
rigorous equations to describe fiber mobility (Chen 1992) and used these to evaluate the
deposition of various types of fibrous materials in the lung. The trends indicated in their studies
show general agreement with those reported by Harris and Timbrell (1977), but with several
notable refinements.
In a study of refractory ceramic fibers (Yu et al. 1995a), a maximum deposition efficiency of
15% is reported for fibers that are approximately 6 urn long and approximately 1 jim in diameter.
This is close to the fiber size at which maximal deposition is reported by Harris and Timbrell
(1977). As with Harris and Timbrell (1977), Yu et al. (1995a) also report that deposition
efficiency decreases precipitously as diameter increases beyond 1 \im and decreases more slowly
as diameter decreases below 1 |im. For thinner structures, deposition efficiency increases with
both decreasing width and length. As fibers get longer, optimum deposition occurs with
decreasing thickness. Thus, for example, a maximum deposition rate of 10% occurs for fibers
that are 20 urn long at a thickness of 0.8 um.
In a study of silicon-carbide whiskers (Strom and Yu 1994), the deposition model is extended to
fiber widths as narrow as 0.01 urn. Results from this study indicate that fibers between 0.01 and
0.1 (am in thickness are deposited with a minimum efficiency of 5% up to lengths of
approximately 40 urn before efficiency drops below 5%. For thin fibers (thinner than 0.5 fim),
shorter fibers tend to be deposited in the deep lung much more efficiently than longer fibers.
More than 25% of thin fibers shorter than 1 urn are deposited in the deep lung following
inhalation. Strom and Yu (1994) report that the efficiency of deposition in the deep lung of long
structures increases substantially during mouth breathing.
Comparing the results reported for refractory ceramic fibers (density=2.7 g/cm3) and silicon-
carbide whiskers (density=3.2 g/cm3), it also appears that the efficiency of deep-lung deposition
increases for thinner and for longer structures as the density of the structures increases. Given
the observed density effect, chrysotile fibers that are longer than approximately 6 urn and thinner
6.8
-------
than 1 \im would be deposited in the deep lung less efficiently than (denser) amphibole fibers of
the same size. However, shorter and thicker chrysotile structures would be deposited somewhat
more efficiently than similarly sized amphiboles. This suggests that a greater fraction of the
mass of chrysotile that gets deposited in the deep lung will be composed of very short fibers and
somewhat longer bundles than the mass fraction of short fibers or longer bundles in the air
breathed. Also, to the extent that chrysotile fibers are curved, these would be deposited
somewhat less efficiently than straighter (amphibole) fibers of comparable size.
Based on the deposition efficiencies predicted by Yu and coworkers, fibrous structures that reach
the deep lung in humans are effectively limited to those thinner than approximately 1 urn. Given
that fibrous structures have traditionally been defined as particles exhibiting aspect (length to
width) ratios >3:1 (Walton 1982), it is clear that only particles shorter than 3 p.m could
potentially be respirable and still be excluded from the definition of a fibrous structure based on
aspect ratio. Therefore, the thickness constraint for all longer structures is best described as a
maximum width (rather than an aspect ratio) when defining the range of structures that
potentially contribute to biological activity.
Rats versus Humans. Yu and coworkers also modified their models to evaluate the rates that
fibrous materials are deposited in rat lungs and compared these with results for humans. Such
comparisons have implications for the manner in which results from animal inhalation studies
are extrapolated to humans.
Results from Yu et al. (1994) suggest that pulmonary deposition of all fibrous structures with
lengths between about 1 and 100 jam and thinner than approximately 1 (im occurs at much higher
rates in rats than in humans. Fibers as long as 90 um are deposited in rat lungs at efficiencies
exceeding 20% while fewer than 5% of structures this long are deposited in the pulmonary
region of human lungs. In fact, it is only structures between 1 and about 20 \nm within a very
narrow range of thicknesses (centered around 1 urn) that are deposited more efficiently in the
deep lungs of humans than in rats.
Yu et al. (1995a) also indicate that, even when deposition efficiencies are comparable in rats and
humans, due to differences in the total lung mass and breathing dynamics across species, the
resulting lung burdens (i.e., the mass or number of structures per mass of lung tissue) are 5-10
times higher in the rat than in humans for any given exposure. Lung burden per lung surface
area are also higher in the rat than in humans.
To illustrate, assume rats and humans are similarly exposed to a concentration of 0.1 f/cm3
(100 f/L) of some fibrous material with a length at which both species retain approximately 10%
of the fibers inhaled. Table 6-1 then indicates the calculations required to determine the relative
rates at which the lung (volume and surface area) burdens in each species would develop.
6.9
-------
Table 6-1. Estimation of Lung Volume and Lung Surface Area Loading Rates for Rats
and Humans
Species
Human
Rat
Species
Human
Rat
Body
Weight (kg)
70
0.15
Breathing
Rate
(L/min)
21.7
0.13
Lung Volume
(L)
5
0.01
No. Fibers
Inhaled per
Minute
(f/min)
2170
13.3
Lung Surface
Area (m2)
140
0.4
No. Fibers
Deposited per
Minute
(f/min)
217
1.3
Rest Breaths per
Minute (bpm)
15
70
Lung Volume
Loading Rate
(f/L-min)
43.4
130
Tidal Lung
Volume (L)
1.5
0.0019
Lung Surface
Area Loading
Rate
(f/m2-min)
1.6
3.3
From Table 6-1, it is clear that rats exposed to comparable airborne concentrations as humans
will increase their loading of fibers per volume (or mass) of lung at a rate that is approximately
3 times that of humans (for fibers in sizes that are deposited with 10% efficiency in both
species). Similarly, the fiber load per surface area of lung will increase in rats at a rate that is
approximately twice that of humans. Moreover, even higher relative mass or surface area
loading rates are expected for the rat than shown in the table, due to the greater efficiency with
which most fiber sizes are deposited in rat lungs. Data used to compute the loading rates in the
table (which are also presented) are derived from Gehr et al. (1993) and supplemented with
information from Stober et al. (1993). A more detailed description of this information is
provided in Section 4.4.
6.1.3 The Effects of Electrostatic Charge on Particle Respirability
Electrostatic charge has been shown to affect the retention of particles within the lungs (see, for
example, Vincent 1985). Since processes that generate airborne particles generally involve some
form of abrasion, airborne dust particles frequently exhibit varying degrees of electrostatic
charge. Although this potentially leads to variation in the efficiency of particle retention in the
lungs as a function of the source of the dust, a detailed relationship between surface charge and
retention was not described in this paper. A more detailed and quantitative treatment was
developed by Chen and Yu (1993) and the implications of the Chen and Yu model are described
below (following discussion of the results of Davis et al. 1988a). Davis et al. (1988a) report that
animals exposed to dusts containing fibrous chrysotile, whose surface charge is reduced with a
beta minus source, retain significantly less chrysotile than animals dosed with dusts containing
particles whose surface charge has not been reduced. However, the magnitude of the difference
in the mass of fibers retained is less than a factor of 2, implying that the absolute variation due to
this effect may be small. Further research in this area is needed.
Chen and Yu (1993) report that, based on modeling of lung deposition, overall deposition
increases with increasing charge density on the particles inhaled. However, due to the pre-
6.10
-------
iltering by the naso-pharyngeal and tracheo-bronchial portions of the respiratory tract, the effects
of electrostatic charge on deep lung deposition appear to be only slight to modest.
Given the results of the above studies, the overall effects of electrostatic charge on particle
deposition in the deep lung appear to be relatively minor. Therefore, such effects do not need to
be considered explicitly when evaluating the health consequences of asbestos.
6.1.4 General Conclusions Concerning Particle Respirability
Based on the information provided in the last several sections, it is apparent that in humans:
• deposition of asbestos fibers in the pulmonary portion of the lung occurs
primarily at alveolar duct bifurcations;
• electrostatic effects on pulmonary deposition are likely minor;
• fibers that are deposited in the pulmonary portion of the lung are largely thinner
than approximately 0.7 urn and virtually all are thinner than 1 u.m (except during
mouth breathing, when thicker and more complex structures may be respired);
• the length of a fiber has limited impact on respirability up to a length of
approximately 20 (am, but the efficiency of deposition of longer fibers decrease
slowly with increasing length for longer fibers;
• as the length of the fibers that are inhaled increases, the thinner fibers are
deposited with greater efficiency. Thus, the longer the fibers inhaled, the thinner
the fibers retained;
• due to differences in density, shorter and thicker chrysotile structures will be
deposited more efficiently in the pulmonary portion of the lung than
corresponding amphibole structures and longer and thinner amphibole structures
will be deposited more efficiently than corresponding chrysotile structures;
• curly chrysotile structures are less likely to reach the pulmonary portion of the
lung than straight amphibole (or chrysotile) structures;
• except for a very narrow range of fiber sizes (centered around 6 u.m in length and
1 urn in diameter), virtually all size fibers are deposited with greater efficiency in
rat lungs than human lungs;
• due to body morphology and the dynamics of breathing, rats exposed to similar
air concentrations will accumulate fiber burdens both per mass (volume) of lung
tissue and per lung surface area at a rate that is several times the rate such burdens
accumulate in humans; and
• the dynamics of fiber lung deposition can now be accurately predicted in great
detail using currently available models.
6.11
-------
6.2 FACTORS AFFECTING DEGRADATION, TRANSLOCATION, AND
CLEARANCE
Degradation and clearance mechanisms compete with deposition to determine the fraction of
asbestos that is retained in the lungs. Other (translocation) mechanisms mediate the movement
of asbestos from sites of initial deposition to various target tissues within the lung and
mesothelium. These factors affect all of the toxic endpoints of interest. Studies indicating the
dependence of the various contributing mechanisms on fiber size and mineralogy are
highlighted, as well as studies indicating differences between mechanisms in humans and
laboratory animals.
The three units of the respiratory tract defined in the last section (naso-pharyngeal, tracheo-
bronchial, and bronchio-alveolar units) differ primarily by the types of clearance (and
translocation) mechanisms operating in each unit (Raabe 1984). These are summarized in
Table 6-2 along with rough estimates of the time frames over which each mechanism may
operate (to the extent that such estimates are available in the literature).
Briefly, the structures of the nose and throat are bathed in a continual flow of mucous, which is
ultimately swallowed or expectorated. The mucous traps deposited particles and carries them
out of the respiratory tract. The air channels of the tracheo-bronchial section of the respiratory
tract are lined with cilia and mucous secreting cells. As in the nose and throat, the mucous traps
particles deposited in these air pathways and the ciliary escalator transports the mucous up to the
throat where it may be swallowed or expectorated. Neither the alveolar ducts nor the alveoli of
the pulmonary compartment of the lung are ciliated (inferred from St. George et al. 1993).
Therefore, particles deposited in this section of the respiratory tract can only be cleared by the
following mechanisms:
• if the deposited particles are soluble, they may dissolve and be transported away
from the lungs in blood or lymph; or
• if they are sufficiently compact, they may be taken up by alveolar macrophages
and transported outward to the muco-ciliary escalator of the tracheo-bronchial
portion of the respiratory tract.
Due to a combination of chemical and physical stresses in the environment of the lung, deposited
asbestos structures may degrade by splitting. Longitudinal splitting, primarily of bundles,
produces thinner structures and transverse splitting produces shorter structures. In both cases,
the number of structures produced may be larger than the number of structures initially
deposited.
By changing the size and number of structures that were initially deposited in the lungs, splitting
may affect the rates and efficiency with which the various other degradation and clearance
mechanisms operate.
6.12
-------
Table 6-2. Relative Rates, Half-lives for Particles Cleared by the Varous Operating Mechanisms of a Healthy Lung
Tissue/Lung Regime
Species
Fiber
Type
Particle
Half-life3
(days)
Kinetic Mineralogical Size
Order Effects Effects
Reference
Mechanisms
(Component Mechanisms)
Nasal-pharyngeal
Expectoration and Swallowing
Muco-ciliary Transport
Human
Human
Particles
Minimal
0.0028
Zero No Effect No Effect
Raabe 1984
Tracheo-bronchial
Muco-ciliary Transport
Human
Particles
0.021-0.21
Zero No Effect No Effect
Raabe 1984
Pulmonary (Bronchio-alveolar)
AM Phagocytosis, Transport to
MC Escalator
Dissolution in Extracellular
Fluid
Transport to the Interstitium
Rat
Rat
In-vitro
In-vitro
Rat
Particles
Short
Chr
Chr
Crc
Particles
49
14
180
11,000
2.3
Ps - First No Effect Inhibited
by length;
cone.
Fibers
<4 um
Zero Affects rate Diameter
determines
lifetime
_ _ _
Stoberetal. 1990
in Stober et al.
(1993)
Yuetal. 1990
Hume and
Rimstidt 1992
Zoitusetal. 1997
Stoberetal. 1990
in Stober et al.
(1993)
(Component Mechanisms)
(Phagocytosis and expulsion by epithelial cells)
(AM phagocytosis, transport through epithelium)
6.13
-------
Table 6-2. Relative Rates, Half-lives for Particles Cleared by the Varous Operating Mechanisms of a Healthy Lung
(continued)
Tissue/Lung Regime
Species
Fiber
Type
Particle
Half-life3
(days)
Kinetic Mineralogical Size
Order Effects Effects
Reference
Mechanisms
(Diffusive transport through the epithelium)
(Forced mechanical transport through the epithelium)
Sequestration
(Phagocytosis and internalization by epithelial cells)
(AM phagocytosis, immobilization due to overload)
Pulmonary (Interstitial)
IM Phatogyctosis, Transport to Rat
Lymphatics
Dog
Particles 2,300
Ams
Diffusive Fluid Transport to
Lymph
Dissolution in Extracellular
Fluid
Transport to Endothelium, Pleura
(IM phagocytosis, transport through interstitium)
(Diffusive transport through the interstitium)
(Forced mechanical transport through the interstitium)
Sequestration
(Encapsulation in granulomatous tissue)
(Internalization by interstiial/endothelial cells)
Unspecified
negative effect
Inhibited
by length;
cone.
2,200
2,200
(Same as Above)
Stober et al.
(1990)inStoberet
al. (1993)
Oberdorster et al.
(1988)
Churg 1994
6.14
-------
Table 6-2. Relative Rates, Half-lives for Particles Cleared by the Varous Operating Mechanisms of a Healthy Lung
(continued)
Tissue/Lung Regime
Fiber
Species Type
Particle
Half-life"
(days)
Kinetic
Order
Mineralogical
Effects
Size
Effects
Reference
Mechanisms
The Pleura
PM Phatogyctosis, Transport to Lymphatic Stomata
Dissolution in extracellular fluid
Sequestration
(Encapsulation by granulomatous tissue)
(Phagocytosis by mesothelial cells)
Tor zero order mechanisms, half-lives reported are half of the time required for complete clearance for the process that is constant with time.
For first order mechanisms, the true half-lives (i.e., the time required for half of the initial population to disappear) is reported.
6.15
-------
Particles and fibers that are deposited in the pulmonary portion of the lung may also be
transported by a variety of mechanisms into and through the epithelium lining, the alveolar
ducts, and alveoli to the underlying interstitium and endothelium that are located within the
interalveolar septa (see Section 4.4). In those portions of the lung parenchyma that lie proximal
to the pleura, such mechanisms may also facilitate transport to the mesothelium. Putative
mechanisms by which such transport may occur include:
• if particles are sufficiently compact to be phagocytized by alveolar macrophages,
they may be transported within macrophage "hosts" through the epithelium to the
interstitium;
• if particles are sufficiently compact to be phagocytized by the epithelial cells
lining the air passageways of the deep lung, they may be transported into cell
interiors or transported through to the basement membrane, the interstitium, the
endothelium, and (eventually) the pleura;
• particularly when associated biological effects that cause changes in the
morphology of epithelial cells, particles may diffuse between the cells of the
epithelium to underlying tissues; and/or
• particles may be transported through respiratory epithelium mechanically due to
physical stresses associated with respiration within the lung.
Although the transport of fibers and particles from airway lumena to the interstitium is apparent
in many studies (see below), the precise mechanisms by which such transport actually occurs has
yet to be delineated with certainty.
Particles deposited in the interstitium can also be cleared and the processes by which these
particles are ultimately cleared are similar to, but may be substantially slower than, the
mechanisms by which particles deposited in airway spaces can be cleared. Such mechanisms
include:
• if the deposited particles are soluble, they may dissolve and be transported away
from the lungs in blood or lymph; or
• if the particles of the interstitium are sufficiently compact to be phagocytized by
interstitial macrophages, they may be taken up and transported to the lymphatic
system for removal.
The mechanisms by which particles that reach the pleura and mesothelium may be cleared are
also similar to those operating in the interstitium:
• if the deposited particles are soluble, they may dissolve and be transported away
from the lungs in blood or lymph; or
6.16
-------
• if the particles that reach the pleura are sufficiently compact to be phagocytized
by pleural macrophages, they may be taken up and transported to the lymphatic
system for removal.
Particles cleared from the pleura by macrophages appear subsequently to be deposited at sites of
lymphatic drainage along the pleura (i.e., at lymphatic ducts) from where they are ultimately
cleared in lymph (Kane and MacDonald 1993).
The various degradation, clearance, and transport mechanisms that affect the retention of
asbestos in the lung and other target tissues (identified above) exhibit disparate kinetics that may
be further altered by the size, shape, mineralogy, and concentration of the particles affected.
Therefore, the kinetics of these mechanisms are considered below. The mechanisms evaluated
include:
• dissolution;
• muco-ciliary transport;
• macrophage phagocytosis and transport; and
• diffusional transport.
Evidence for the existence of these mechanisms and inferences concerning their kinetics derive
primarily from retention studies, which may include both studies of retained structures in
animals following either short-term or chronic exposure, or human pathology studies in which
the lung burdens of deceased individuals are correlated with their exposure history. Other
information also comes from in vitro studies. Various, increasingly sophisticated models have
also been developed to predict the individual and combined effects of these mechanisms.
6.2.1 Animal Retention Studies
Retention studies track the time-dependence of the lung burden of asbestos or other paniculate
matter (i.e., the concentration of particles in the lung) during or following exposure. Thus, such
studies are designed to indicate the degree to which inhaled structures are retained. Depending
on the time frame evaluated, however, effects due to deposition and those due to clearance may
not easily be distinguished in such studies. Moreover, due to the near impossibility of isolating
the various compartments of the lung when preparing for quantitative analysis of tissue burden
(e.g., the pure respiratory components vs. the larger airways or the tissues directly associated
with airway lumena vs. the underlying interstitium or endothelium), it is nearly impossible to
separate the effects of the various clearance mechanisms, which typically operate over vastly
different time scales (Table 6-2). This is why modeling has proven so important to
distinguishing effects attributable to individual mechanisms.
Results from retention studies must be evaluated carefully. In addition to the limitations
highlighted above, the lung burden estimates from such studies may be affected by the manner in
which asbestos is isolated from lung tissue for measurement and the manner in which the
concentration of asbestos is quantified (Chapter 5). For example, lung burden estimates may
vary substantially depending on what portions of lung parenchyma are sampled or whether
whole lungs are homogenized. Results may also vary depending on whether lung tissue is ashed
or dissolved in bleach during sample preparation. More importantly, because several clearance
6.17
-------
mechanisms are affected by the size and even the mineralogy of the structures being cleared,
studies (particularly older studies) that track lung burden by mass or by total fiber number may
not adequately capture such distinctions.
6.2.1.1 Studies involving short-term exposures
The latest retention studies tend to focus on the fate of long fibers (typically those longer than
20 u.m) in support of the generally emerging recognition that these are the fibers that cannot be
readily cleared from the pulmonary compartment of the lung and that, not coincidentally,
contribute most to disease (further addressed in Section 6.4).
Hesterberg et al. (1998a), for example, tracked the time-dependent retention in rats of two fiber
categories: (1) WHO fibers' and (2) WHO fibers longer than 20 \im for a range of man-made
vitreous fibers (MMVF's), a refractory ceramic fiber (RCFla), and amosite following a 5-day
(6 hr/day), nose-only exposure. Rats were sacrificed at intervals up to a year following
exposure. The amosite was size-selected to contain a high proportion of fibers longer than 20
um. Aerosol concentrations were also adjusted to maintain target concentrations of 150 f/cm3
for long fibers for each sample tested. Airborne mass concentrations varied between 17 mg/m3
for amosite to as much as 60 mg/m3 for the other fiber types. Lungs (without trachea or main
bronchi) were weighed and stored frozen. For analysis, each lung was dried to constant weight,
minced, and a portion was ashed. The ashed portion was further washed with filtered, household
bleach, then filtered and applied to an SEM stub. Fiber numbers and dimensions (in both
aerosols and tissue) were determined by SEM with a minimum of 200 fibers counted. In
addition, analysis continued until a minimum of 30 fibers longer than 20 u.m were counted.
In their study, Hesterberg et al. (1998a) tracked the ratios of retained fiber concentrations with
time to the concentration retained 1-day following cessation of exposure. The observed time-
dependent decay in these ratios were then fit to one-pool (single first order decay) or two-pool
(weighted sum of two first order decays) models. With zero time assumed to be the time
immediately following cessation of exposure. The authors recognize that at least some clearance
likely takes place during the 5 days of exposure so they expected the assumption that retained
concentrations at the end of exposure on day 5 to be equal to deposited concentrations would
cause their analysis to slightly underestimate clearance rates. They also recognized that waiting
24 hours after cessation of exposure to measure retention allows some short-term clearance of
upper airways, so that they expected their analysis would better focus on slower clearance from
deeper in the lung.
Results reported by Hesterberg et al. (1998a) indicate that the dimensions and concentrations of
fibers in aerosols from the five synthetic fibrous materials were all similar, but that the amosite
aerosol contained a substantially greater number of fibers (including the longest fibers) and that
the fibers, on average, were somewhat shorter and substantially thinner than the other aerosols.
Of the fibers initially deposited in the lung (based on measurements made 1 day following
cessation of exposure), comparable fiber numbers of long (>20 urn) fibers were retained across
WHO fibers are those longer than 5 \itn, thinner than 3 |j.m with an aspect (length to width) ratio greater
than 3 (WHO 1985).
6.18
-------
all six fiber types. Deposited concentrations of fibers 5-20 um in length were more variable, but
values within one standard deviation still overlapped. About 6 times as many short amosite
fibers (<5 u.m) were initially deposited than for any of the other fiber types. The authors also
indicate that the dimensions of retained fibers were generally shorter and thinner than the
original aerosol and were much more similar across retained fiber types than the original
aerosols.
Clearance of long fibers (>20 (am) for all six fiber types could best be described using a two-pool
model. The first pool cleared relatively rapidly (within the first 90 days) and represented a
minimum of 65% of the lung burden observed 1 day following exposure. The second pool
cleared much more slowly. For amosite fibers in the second pool, during the approximately 275
days of clearance, retention was only reduced to 80% of the 90-day value. In contrast, all five of
the synthetic fibers were reduced to less than 30% of their 90-day value during this period. For
amosite, the first pool decayed with a half-life of 20 days (90%CL: 13-27) and all of the other
fibers with half-lives of 5-7 days (with varying confidence bounds). For the slower pool,
amosite fibers exhibited a half-life of 1,160 days (90% CL: 420-°°) with the other fibers showing
half-lives varying between 24 and 179 days. The combined, weighted half-life for amosite was
418 days (90%CL: 0-1060). The authors also note that data reanalyzed from an earlier study
(Hesterberg et al. 1996) indicate a corresponding weighted half-life for crocidolite of 817 days
(246-^)'and indicate that this was best fit using a single exponential (a one-pool model).
Hesterberg et al. also indicate that in this and previous studies approximately 20-60% of long
fibers typically clear from the lung within 2 weeks post-exposure. They further suggest that this
rapid clearance may be attributable to muco-ciliary clearance from the upper respiratory tract.
They further report from the present study that short amosite fibers cleared much more rapidly
than long fibers. Fibers <5 ^m in length were reduced by 90% in the first 90 days (in
comparison to 65% for long fibers). However, from 90 to 365 days, little or no clearance was
observed for amosite fibers of any length.
For four of the synthetic fibers, long fibers cleared at the same rate as short fibers (all more
rapidly than amosite) and the authors report that the data suggest transverse breakage for these
fibers. Moreover, they attribute the more rapid clearance of long fibers among the MMVF's to
dissolution, since these fibers exhibit in vitro dissolution rates that are rapid relative to the time
scale of macrophage clearance. One synthetic fiber MMVF34, which is a stonewool,
disappeared much more rapidly than any other fiber and the long fibers disappeared more rapidly
than the short fibers. MMVF34 shows the greatest in vitro dissolution rate at neutral pH for any
of the fibers tested in ths study and dissolves particularly rapidly at pH 4.5 (the pH found in the
phagosomes of macrophages). The authors postulate that clearance of all of the synthetic fibers
are enhanced over amosite by dissolution and breakage.
In summary, Hesterberg et al. (1998a) observed that:
• multiple clearance mechanisms (operating over multiple time scales) contribute to
clearance;
• for sufficiently soluble fibers, long fibers clear more rapidly than short fibers;
6.19
-------
• for insoluble fibers, a subset of long fibers clears rapidly within the first few
months following exposure and the remaining long fibers clear only extremely
slowly, if at all;
• short fibers of all types are cleared at approximately the same rate (much more
rapidly than long, insoluble fibers);
• a small, residual concentration of short fibers may not always clear and may
remain in the lungs (sequestered in alveolar macrophages) for extended periods;
and
• in this study, there is some suggestion that short amosite fibers clear somewhat
more slowly than short fibers of the other, non-asbestos mineral types studied.
Regarding the last observation, whether this is attributable to differences in fiber thicknesses
among the various mineral types, due to partial contributions (even among short structures) to
dissolution, or due to a unique, toxic effect of amosite is unclear. However, the likeliest of these
candidate hypotheses is that the effect is due to partial dissolution.
This general pattern of observations are consistent with the findings of most, recent retention
studies following short-term exposure.
In an earlier study of similar design, Bernstein et al. (1996) evaluated the deposition and
clearance of a series of 9 glass and rock wools. These authors similarly found that clearance
could be modeled using a double exponential for all length fibers (in similar length categories of
<5, 5-20, and >20 um) and that for soluble fibers, long fibers clear more rapidly than short fibers
(with the intermediate length fibers in between).
For the Bernstein et al. (1996) study, if one assumes that the pool of longer-lived fibers is
representative of macrophage clearance, this suggests that the efficiency of clearance by
macrophages decreases with increasing fiber length and that the longest structures are not
phagocytized at all, so that they remain exposed to the extracellular medium where dissolution
occurs. Lending further support to this interpretation, the authors also report that the clearance
rate for long fibers correlate with measured in vitro dissolution rates at neutral pH while the
clearance rates for short fibers neither correlate with in vitro dissolution rates at neutral pH or at
pH 4.5. Although the latter pH corresponds to the pH found in the phagosomes of macrophages,
there is likely too little fluid available in such organelles to support efficient dissolution. The
authors also indicate that a sufficient number of fibers were counted during the study to suggest
that breakage is not playing a role in clearance (except at very early times) and that the clearance
rate for short fibers appears to be the same or slower than that observed for nuisance dusts.
In another, earlier study of similar design Bastes and Hadley (1995) evaluated four types of
MMMF's and crocidolite. All of the samples (including crocidolite) had been size-selected to
assure a large fraction of fibers longer than 5 um. Unfortunately, due to differences in reporting,
it is not possible to compare the initial loading of crocidolite fibers to those reported for amosite
in the Hesterberg et al. (1998a) study. However, results from this study further support the
physical interpretation of clearance suggested in the studies discussed above. In fact, the authors
6.20
-------
report that the time-dependent size distribution of retained fibers observed in this study agree
well with a computer simulation of fiber clearance. The simulation assumes that long fibers
dissolve at the rate measured for such fibers in vitro and that short fibers of every type are
removed at the same rate as short fiber crocidolite (which is practically insoluble). This is strong
evidence that short fibers are cleared by macrophage phagocytosis and that long fibers cannot be
cleared by macrophages, but may dissolve in extracellular fluid provided that they are
sufficiently soluble.
Regarding crocidolite, the data from the Bastes and Hadley (1995) study suggest that short
crocidolite fibers appear to clear at a rate that is somewhat slower than observed for any of the
short MMVF fibers. Importantly, however, the interpretation of short fiber clearance in this
paper is somewhat confounded because, unlike the studies discussed above, short fibers in this
paper are defined as all fibers <20 ^im, so there may be some confounding with MMVF fibers
that are dissolving. As previously indicated, the Hesterberg et al. (1998a) work also suggests
that short asbestos (amosite) fibers may clear more slowly than short fibers of differing
mineralogy and Hesterberg et al. only includes fibers <5 u.m in their definition. Nevertheless, it
is still possible that some effects due to dissolution may still be affecting the clearance of these
shorter fibers.
Surprisingly, a visual inspection of the data presented in Bastes and Hadley (1995) table suggests
a lack of any long-term clearance for long fiber crocidolite (>20 urn). Yet, the authors model
long fiber crocidolite clearance using a single exponential (suggesting no rapidly clearing
compartment). The long-term half-life reported for crocidolite in this study is approximately
220 days (with estimated CIs of 165-566 days). This overlaps with the long-term clearance half-
life reported by Hesterberg et al. (1996) for crocidolite of approximately 820 days (246-«°).
Equally surprising, Hesterberg et al. (1998a) also modeled crocidolite clearance as a single
exponential, which might suggest better penetration to the deep lung by crocidolite, less
clearance by muco-ciliary transport or alveolar macrophage transport, or better penetration to the
interstitium than other fibers. More likely, however, it may simply indicate that the two-pool
model does not represent a statistically significant improvement in model fit over the one-pool
model. However, relative size distributions would need to be evaluated carefully before drawing
any such conclusions. Bastes and Hadley (1995) also report clearance of short fiber crocidolite
is modeled as a double exponential with short and long half-lives of 25 and 112 days,
respectively. Since this fiber category contains fibers up to 20 urn in length (in this study only),
this does suggest at least some contribution from muco-ciliary and alveolar macrophage
mediated clearance for crocidolite.
In two studies, Coin et al. (1992, 1994) evaluated the fate of chrysotile fibers in rats exposed for
3 hours to 10 mg/m3 (reportedly containing >5,000 fibers longer than 5 urn/cm3). For lung
analysis, the left lung was separated into peripheral and central regions under a dissecting
microscope. Slices of peripheral and central portions were separately weighed and minced.
Tissue was digested in sodium hypochlorite and then filtered. A quality control test indicated
that the digestion process caused a slight (-10%) decrease in fiber number and slight decreases
in fiber diameter and fiber length. Fiber-size distributions were evaluated by SEM. A stratified
counting procedure was employed to assure equal precision for each length category of interest.
Measurements for each category were then converted to mass equivalents.
6.21
-------
Results from the Coin et al. (1992, 1994) studies indicates no difference between deposition in
central or peripheral regions of the lung. They also confirm that chrysotile splits longitudinally
in the lungs with a half-life that is competitive with the clearance rates measured in this study.
Clearance was found to be very length-dependent, so that rates decrease from a half-life of about
10 days for fibers about 4 urn in length, through 30 days for fibers 8 urn, to 112 days (which is
no different from zero) for fibers longer than 16 um (all after adjusting for longitudinal splitting).
Importantly, the brief follow-up period (30 days) is too short to provide an adequate evaluation
of the longer term clearance pools observed in other studies and certainly too short to evaluate
any effects potentially associated with chrysotile dissolution. Also, that the decay curves for
clearance were'limited to four points, makes evaluation of the slopes for these curves highly
uncertain.
Coin et al. (1992, 1994) also report that the mass of chrysotile deposited during these short
exposures (i.e., no more than 20 ug) is very small compared to levels at which overload has been
reported to occur (approximately 1 mg, see, for example, Yu and Yoon 1991) and that the
volume of the 16 um fibers, which have an average diameter of 0.2 um and therefore a mean
volume of 0.5 um3, is small relative to the volume at which macrophage clearance of non-fibrous
particles is reported to be hindered (Morrow 1988). Thus, the authors conclude that fiber length
presents an additional constraint on macrophage clearance, independent of any other overload.
They also indicate that inhibition of clearance due to fiber length is independent of fibrosis.
Coin et al. (1992, 1994) also discuss the effect of fibrosis on clearance. They indicate that,
although increased concentrations of short fibers are observed in focal areas of fibrosis, it is
more likely that such fibers accumulate because clearance is hindered by fibrosis in these areas
than the hypothesis that the short fibers are causing fibrosis. This is because, as they point out,
there are too many studies demonstrating the lack of ability of short fibers to induce fibrosis.
Evidence in the Coin et al. (1992, 1994) studies suggests that no translocation from central to
peripheral regions of the lung were detected. An upper bound rate that is about 20% of clearance
is reported. However, the short follow-up time in this study would have precluded slower
processes from being detected. Despite the lack of evidence of translocation, the authors report
that duct bifurcations in peripheral regions of the lung where fibers are deposited are no more
than 1-2 mm from the visceral pleura. In fact, in the 1994 study, the authors show that 50% of
the primary duct bifurcations in the peripheral portion of the rat lung occur within 1 mm of the
visceral pleura and some occur as close as 220 um. Deposited fibers may also affect the pleura
by inducing generation of diffusable, inflammatory agents.
A short-term inhalation study by Warheit et al. (1997) evaluated retention of chrysotile and
aramid fibers. In this study, rats (and hamsters) were exposed, nose only, for 6 hours/day,
5 days/week for 2 weeks by inhalation to UICC chrysotile and p-aramid fibers (each at two doses
of 460 or 780 fibers/ml, although the size range of these fibers is not stated nor is the manner in
which they were analyzed). Fixed lungs were digested in chlorox during preparation for
asbestos analysis. Animals were followed for up to a year post-exposure.
As in studies described above, results from the Warheit et al. (1997) study indicate rapid
clearance of short chrysotile fibers, but slow to non-existent clearance of fibers longer than
20 um. In contrast, aramid fibers apparently degrade and are subsequently cleared fairly rapidly
6.22
-------
in vivo. Based on the data provided in figures, although the reported concentrations of chrysotile
and aramid fibers to which animals were exposed were equivalent, at both the lower and higher
concentrations, it appears that rats initially retain 3 to 5 times as many aramid fibers as chrysotile
fibers (at least for the size range counted, which was not reported). For both fiber types,
clearance appears rapid for an initial period of approximately 90 days post-exposure. During
this time, the mean length of chrysotile fibers also appears to increase steadily, which suggests
rapid, preferential clearance of short structures. After the initial period, it appears that (as the
authors suggest), a residual concentration of longer fibers are cleared only very slowly, if at all.
Oberdorster et al. (1988) instilled a 3 ml suspension of irradiated amosite into the bronchio-
alveolar space of the right diaphragmatic lobe of the lungs of dogs to evaluate clearance and
transport. The amosite used was modified by sedimentation from UICC amosite to contain only
fibers shorter than 20 urn. One dog also had unmodified UICC amosite instilled directly into a
lymph node in the thigh. The dogs had been cannulated to allow collection of lymph from the
right lymph duct-RLD and the thoracic duct-TD (both in the neck).
Results from Oberdorster et al. (1988) indicate that within 4 hours following instillation in the
lung, low activity was noted in postnodal lung lymph, but not in either the RLD or TD. Within
24 hours, however, activity and fibers (determined by SEM) were observed in both the RLD and
the TD. The median length of fibers observed in the lymph were significantly longer than the
instilled material, although there appeared to be a cutoff length of 16 um in fibers observed at
nodes and 9 |o,m in fibers observed directly in lymph. Fibers recovered from lymph were also
significantly thinner and appeared to exhibit an absolute cutoff at a maximum width of 0.5 (am.
Fibers recovered from the TD and RLD in the dog that had unmodified UICC amosite instilled
directly into leg lymph were all short (with a maximum length of 6 |im). Since collection times
were all short, the authors indicate that it is unknown whether longer fibers would have been
observed at later times. The authors also note the almost total absence of fibers shorter than
1 um in lymph, which they assume are cleared rapidly and efficiently by alveolar macrophages.
Oberdorster et al. (1988) also report that a rough calculation, based on the fraction of the
material originally instilled that was recovered in the first 24 hours, it would take approximately
6 years to clear all of the instilled asbestos (assuming no other clearance mechanisms were
active).
Everitt et al. (1997) performed a short-term inhalation study that is interesting, particularly,
because it focused on pleural (as opposed to lung) fiber burden. The authors exposed rats and
hamsters to one type of refractory ceramic fiber (RCF-1) by nose-only inhalation for periods of
0, 4, and 12 weeks and animals were held for observation for up to an additional 12 weeks post-
exposure. Exposures were conducted for 4 hours/day, 5 days/week, at 45.6±10 mg/m3. Groups
of 6 animals were held for 0,4, 12, and 24 weeks to determine pleural fiber burden. An agarose
casting method was reportedly used to recover fibers from the pleura. Analysis was by electron
microscopy. Fibers were observed in the pleura at each time point examined (including samples
from rats sacrificed immediately following the last day of a 5-day exposure). Fibers were all
reported to be short and thin (geometric mean length: 1.6 u.m with GSD: 1.8, geometric mean
diameter: 0.1 |im with GSD: 1.5). Concentrations averaged approximately 40,000 fibers (per
whole pleura, units not reported). The authors indicate that such fibers would not typically be
6.23
-------
visible by optical microscopy. They also indicate that use of casts may be a more efficient
method of recovering fibers from the pleura.
Everitt et al. (1997) indicate that observation of rapid translocation of short, thin fibers to the
pleura has also been observed in studies of chrysotile so these results are not unique. Although it
is stated that the mechanisms facilitating translocation are currently unknown, the authors
indicate that their finding of site-specific mesothelial proliferation supports observations by
Boutin et al. (1996) that asbestos fibers accumulate in the parietal pleura of humans at sites
associated with lymphatic drainage. Kane and MacDonald (1993) have suggested that fibers are
transported to these locations by pleural macrophages. However, the mechanisms by which
fibers are transported from the lung to the pleura are still unconfirmed.
Older Retention Studies. Although the older retention studies generally support the results of
newer studies (such as those cited above), older studies are sometimes limited by such things as
the tracking of lung burden in terms of fiber mass or use of analytical techniques such as infrared
spectroscopy for detection of asbestos, which are neither capable of distinguishing individual
fibers nor provide any information on their sizes. Tracking of lung burdens in terms of mass
may not reflect the fate of long, thin fibers, which (by increasing concurrence) appear to be the
legitimate focus of studies evaluating biological hazards attributable to asbestos.
In two studies (Roggli and Brody 1984 and Roggli et al. 1987), Roggli and coworkers tracked
the behavior of chrysotile (not UICC) and UICC crocidolite in rats following 1 hour exposure by
inhalation to 3.5-4.5 mg/m3 dusts. The authors indicate that this results in deposition of
approximately 21 ug of dust. Portions of the lower lung lobes of selected rats were collected and
digested for asbestos analysis using a scheme that was shown to be representative. To evaluate
size distributions, more than 400 fibers from each sample were characterized by SEM. Fiber
dimensions were then used to estimate total fiber mass.
Based on their study, Roggli and coworkers indicate that similar fractions of inhaled chrysotile
and crocidolite dust are deposited in the lung during inhalation (23 and 19%, respectively). The
authors therefore concluded that respirability and deposition do not depend on fiber type.
Importantly, however, the manner in which this study was conducted does not facilitate
distinguishing deposition in the deep lung from deposition in the upper respiratory tract.
Roggli and coworkers further indicate that clearance rates for the two fiber types appear
comparable. Of the chrysotile initially deposited, they report that 81% of this material is cleared
after 4 weeks. Similarly, 75% of the crocidolite is cleared. Importantly, because this is based on
total mass (estimated by summing volume contributions from observed fibers), it may not reflect
the specific behavior of long, thin structures. Therefore, it is difficult to compare such results
with those of more recent studies. However, the authors do report that short structures are
cleared more readily than long structures and that chrysotile is observed to split longitudinally
in vivo (based on observation that the total number of chrysotile structures initially increases and
the mean length increases). The authors further conclude that clearance rates appear to be
independent of fiber type.
In another short-term study, Kauffer et al. (1987) report that the average length of retained
chrysotile structures increases in rat lungs following 5-hours inhalation of chrysotile dust. Based
6.24
-------
on their results, the authors report that fibers shorter than approximately 8 urn are preferentially
cleared. Kauffer and coworkers also confirm that chrysotile fibers split longitudinally in the
lung. In fact, several other studies (Kimizuka et al. 1987;Le Bouffant 1980) also provide
supporting observations that chrysotile fibers (or bundles) split longitudinally in the lung.
In two studies (Morgan et al. 1978,1980), Morgan and coworkers report on the fate of fibers
following short-term inhalation of radio-labeled fibers by rats. In the first study (Morgan et al.
1978), rats inhaled UICC anthophyllite at 35 mg/m3 for a total of 8.4 hours spread over 3 days.
The authors report that the rats retained approximately 190 jig of dust at the end of exposure,
mostly in the alveolar region; the authors assumed that conducting airway clearance is
sufficiently rapid to clear this portion of the lung within a few days. Beginning about 7 days
following exposure, the rats were then sacrificed serially for a period up to 205 days following
exposure. Because anthophyllite fibers are relatively thick, fibers were analyzed by optical
microscopy. Fibers were determined both in free cells (mostly macrophages) recovered in
bronchopulmonary lavage and in lung tissue. Tissue samples and cells were digested in KOH
and peroxide in preparation for fiber analysis.
Based on this first study, Morgan et al. (1978) report that anthophyllite lung content declined
steadily by a process that could be described as a simple first order decay with a half-life of
approximately 76 days. Free macrophages recovered by lavage, initially contained about 8 pg
and this too declined steadily with a half-life of about 49 days. The authors further indicate that,
if the number of macrophages remains constant with time (i.e., they are replaced at the same rate
they are cleared), then the decay of the load in the macrophages should match what is observed
in the rest of the lung. They suggest that the discrepancy may be due either to an influx of an
increasing number of macrophages in response to injury with time and/or to transfer of some
fibers through the alveolar wall. They also cite unpublished work indicating that uptake of fibers
by alveolar macrophages is essentially complete within hours after cessation of exposure.
The authors also report that, initially, the lengths of fibers recovered in lung lavage was greater
than in the original aerosol, but that the prevalence of the longest fibers decreased after the first
7 days. In lung tissue, however, the fraction of longer fibers (among total fibers) steadily
increased with time. This suggests rapid clearance of naked fibers by muco-ciliary transport
(which is a length independent process) with later times dominated by slower clearance in the
deep lung by alveolar macrophages, which is a length-dependent process.
Morgan and coworkers also radiometrically determined the fraction of fibers in rat feces (prior to
sacrifice). They assumed that after 14 days, this would represent the fraction of asbestos cleared
primarily from the alveolar region of the lung. Initially 1.4% of lung anthophyllite content was
excreted daily, but this fell to 0.5% after 120 days. The authors indicate that this suggests that
the elimination from asbestos in lung tissue cannot be described by a single exponential because
multiple processes are involved and that, over longer periods of time, the slower processes
become increasingly important.
In the second study, Morgan et al. (1980) track the fate of several size-selected radiolabeled
glasses in rats, again following short-term inhalation. From an analysis of the size dependence
of deposited fibers in this study, the authors suggest that alveolar deposition in the rat is limited
to structures with aerodynamic equivalent diameters less than about 6 urn and that deposition in
6.25
-------
this region of the lung falls precipitously for fibers with thicknesses between about 2 and 3
(aerodynamic equivalent diameter). For fibers that are the density of asbestos, this represents an
upper bound limit to alveolar deposition for the absolute thickness of a fiber of approximately
1.5 u.m with fibers deposition of fibers thicker than approximately 0.7 u.m being drastically
reduced. This is in concordance with conclusions concerning deposition provided in
Section 6.1.4. Alveolar deposition efficiency is also shown to decrease with increasing fiber
length, at least for fibers longer than approximately 8 ^m, also in concordance with findings
presented in Section 6.1.4.
Intratracbeal Instillation. The fate of fibers following intratracheal instillation into the lungs
has also proven informative in some studies. For example, Wright and Kushner (1975)
intratracheally instilled paired samples each of several types of glass fibers, fluoramphibole, and
crocidolite into guinea pigs. For each mineral tested, a sample with predominantly short
structures (25 mg total dose for crocidolite, reportedly 99% <5 ^m) and another with
predominantly long structures (4 mg total dose for crocidolite, reportedly 80% >10 u.m) were
evaluated. Unfortunately, the authors do not report how fibrous structures were characterized.
Results in the cited paper report observations only after 2 years following the last injection.
Wright and Kuschner (1975) report that long structures uniformly caused fibrosis (primarily
involving the respiratory bronchioles and alveoli and abutting the terminal bronchioles) while
the short structures were uniformly phagocytized and generally removed to thoracic lymph
nodes. Among other things, clearance to lymph suggests that fibers reached the interstitium
(Section 6.2.5). It is interesting that, even after 2 years of recovery, the authors observe elevated
levels of macrophages in the alveoli of animals dosed with short structures. Based on the
relative size distributions of the samples analyzed, the authors report that structures up to 10 [im
in length appear to be efficiently scavenged by macrophages. Based on the observation of larger
numbers of short structures than expected in comparison with their fractions in the original
samples, the authors further conclude that glass structures underwent biodegradation so that
longer structures broke down into shorter structures that could be phagocytized.
Wright and Kuschner (1975) also report that long fibers are occasionally visible within the
fibrotic interstitium of dosed animals. The long-fiber dosed animals also show macrophages in
hilar lymph nodes containing fibers that are too small to resolve and all of them are short. In
short-fiber dosed animals, some fibers are seen to remain in the lung within aggregates of
macrophages, both in alveoli and the interstitium. Short-fiber dosed animals also show many
more macrophages within the hilar lymph nodes than long-fiber dosed animals.
In two reports of the same study (Bellman et al. 1986, 1987), Bellman and coworkers followed
the fate of UICC chrysotile, UICC crocidolite, several fibrous glasses, and other manmade
mineral fibers following a single intratracheal instillation of 0.3 ml of fibrous material in rats.
Groups of rats were then sacrificed at 1, 6, 12, 18, and 24 months following instillation. Lungs
were low temperature ashed and the resulting, filtered suspension analyzed by transmission
electron microscopy. Some of the fiber types were also acid treated with 0.1 M oxalic acid for
24 hours prior to instillation.
Bellman and coworkers (1987) report that short fibers (<5 u.m) from all of the fiber types were
shown to be cleared from the lungs with half-lives of approximately 100 days, with the asbestos
6.26
-------
varieties tending to exhibit slightly longer half-lives than the other fibers. Short crocidolite
fibers exhibited a half-life of 160 days. The half-life for clearance of short chrysotile was
reported to be 196 days (the longest of all). However, this was attributed to positive
contributions from breakage of longer fibers.
Bellman and coworkers (1987) report that the behavior of the different long fibers (>5 urn) for
the different fiber types was radically different. The authors report no observed net decline in
long crocidolite fibers over the 2 years of follow-up. The also report no observable changes in
width of these fibers with time. In contrast, long chrysotile fibers increased in number with time
throughout the 18-month follow-up period and this was attributed to longitudinal splitting. The
width of these fibers reportedly decreased with time.
Bellman et al. (1987) also report that a more detailed examination of the time dependence of the
width of chrysotile fibers indicates a rapid increase in the number of thin fibrils (<0.05 [im in
width) and thin bundles (<0.1 |im in width) within 100 days (at the expense of thicker bundles).
The authors suggest that this would result in rapid decrease in the number of chrysotile structures
visible by optical microscopy and, possibly, increased clearance of the thinnest fibrils by
dissolution, but this study shows no increased rate of clearance for thinner chrysotile structures
compared to thicker structures (when viewed by electron microscopy). In contrast, long
chrysotile fibers that were acid-leached prior to instillation reportedly disappeared with a half-
life of 2 days.
Generally, the rate of clearance of the long fractions of the other fibers reported in the Bellman et
al. (1986, 1987) papers varies as a function of solubility and overall thickness. Importantly, all
half-lives are reported to have high standard errors in this study, due to the small number of
animals included for examination.
In summation, virtually all short-term retention studies indicate that:
• fibers retained in the lung tend to be shorter and thinner than the aerosols from
which they derive and the size distributions of retained structures tend to be more
similar overall than the size distributions observed in the original aerosols;
• chrysotile asbestos undergoes rapid, longitudinal splitting in the lung while
amphiboles do not;
• by mass, chrysotile and amphibole asbestos are deposited in the lung with
comparable efficiencies, although it is not clear whether chrysotile dusts tend to
contain sufficient numbers of curly fibers to limit deposition in the deep lung;
• multiple clearance processes operate over different time frames and some of these
processes are strongly length-dependent. Fibers shorter than approximately
10 urn appear to be cleared rapidly relative to longer fibers and those longer than
approximately 20 urn are not cleared efficiently at all (if the fibers are insoluble).
The Bellman et al. (1986, 1987) studies appear to contrast with other studies in
this regard in that they suggest fibers longer than 5 (im do not readily clear;
6.27
-------
• the quickest clearance process (presumably muco-ciliary clearance) is not
dependent on length; and
• the effects of fiber diameter on clearance have not been well delineated overall,
although fibers that reach the deep lung appear to be largely limited to those
thinner than approximately 0.7 u.m.
These findings are in addition to those mentioned previously from the newer studies:
• multiple clearance mechanisms (operating over multiple time scales) contribute to
clearance;
• for sufficiently soluble fibers, long fibers clear more rapidly than short fibers;
• for insoluble fibers, a subset of long fibers clears rapidly while the remaining long
fibers clear only extremely slowly, if at all;
• short fibers of all types are cleared at approximately the same rate (much more
rapidly than long, insoluble fibers);
• a small fraction of short fibers may be retained for long periods under certain
circumstances (sequestered in alveolar macrophages) despite overall rapid
clearance of these structures; and
• there is some suggestion that short asbestos fibers clear somewhat more slowly
than short fibers of the other, non-asbestos mineral types studied.
Regarding specifically the clearance of long fibers, it appears that a component of all such fibers
clears rapidly within the first 2 weeks and this likely represents muco-ciliary clearance. A
second component (representing as much as 60% of the fibers) clears within 90 days and this
likely represents clearance by alveolar macrophages. The remaining long fibers are cleared only
very slowly, if at all, and this likely represents fibers that are sequestered in granulomas or that
escape into the interstitium.
6.2.1.2 Studies involving chronic or sub-chronic exposures
Although the results of older retention studies following longer term (sub-chronic or chronic)
exposure were difficult to reconcile with the results following shorter-term exposures, newer
studies suggest greater consistency and a clearer picture of the fate of fibers in the lung.
Moreover, although there are further suggestions of mineralogy (fiber type) dependent effects
with some clearance mechanisms, it is important that size effects be considered simultaneously,
if the dynamics of these processes are to be understood.
In some of the latest studies, for example, Hesterberg et al. (1993, 1995, and 1998b) exposed rats
(nose only) by inhalation to a series of man-made vitreous fibers (including a variety of fibrous
glasses, rock wools, and refractory ceramic fibers) and two kinds of asbestos: chrysotile
(intermediate length NIEHS fiber) and crocidolite (size selected). Animals were dosed for 6
6.28
-------
hours/day, 5 days/week for up to 2 years at target concentrations of 10-60 mg/m3. The target
concentration for chrysotile and crocidolite was 10 mg/m3. Animals were periodically sacrificed
during the exposure regimen to determine the character of the retained fibers. Vitreous fiber
aerosols were characterized by PCM, SEM or, for chrysotile, by TEM. The right accessory lung
lobe of sacrificed animals was tied off, frozen, and stored for lung burden analysis.
For analysis, lung lobes were dried to constant weight, ashed, the residue suspended in distilled
water, and then filtered on Millipore filters (for examination by optical microscopy) or
Nuclepore filters (for analysis by SEM or TEM for chrysotile). Approximately 100 fibers were
reportedly characterized to establish fiber size distributions. However, this is problematic for
this study because chrysotile asbestos concentrations in the aerosols to which the animals were
exposed contained approximately 100 times as many fibers as the other aerosols. Thus, although
no fibers longer than 20 \im were observed during characterization of the chrysotile, the
concentration of such long fibers could still have been larger in this aerosol than the other
aerosols and it would not necessarily have been observed. This is also true of lung burden
analyses especially because indirect preparation tends to magnify the number of short chrysotile
structures observed in a sample.
A comparison of the retention patterns of chrysotile and RCF-1 from the Hesterberg et al.
(1998b) study is particularly instructive. First, it should be noted that, in contrast to the values
reported by the authors of this study, chrysotile and RCF-1 in fact appear to exhibit comparable
in vitro dissolution rates (12.7 vs. 8 ng/cm2-hr, respectively) when rates are measured using
comparable techniques (see discussion in Section 6.2.4). The dissolution rates quoted in the
Hesterberg et al. study are not derived in comparable studies.
Although a full set of time-dependent analyses are apparently not available for chrysotile, it is
reported that approximately 14% of those chrysotile WHO fibers observed to be retained after
104 days of exposure continue to be retained after 23 days of recovery. Under the same
conditions, it is reported that 43% of RCF-1 fibers are retained, which suggests more rapid
clearance for chrysotile. Even adjusted for the fraction of RCF-1 structures that are longer than
20 |o.m (and that are presumably cleared even more slowly), approximately 37% of the RCF-1
WHO fibers (<20 jxm) are apparently retained over this period, which is still more than twice the
rate reported for chrysotile. Still, more detailed characterization of the size distributions of these
two fiber types would need to be evaluated before it could be concluded with confidence that
chrysotile is cleared more rapidly than RCF-1 or that dissolution plays a role. In fact, dissolution
would tend to cause more rapid clearance only of the longest fibers (i.e., the ones that cannot be
cleared by macrophages [see Section 6.2.1.1]), which would further reduce the apparent
retention rate of the shorter RCF fibers, making it even more comparable to the chrysotile
number.
Based on these studies, Hesterberg et al. (1998b) report that fibers deposited and retained in the
lung tend to be shorter and thinner on average than the sizes found in the original aerosol. It is
also apparent from their data that long RCF-1 fibers clear more rapidly than short RCF-1 fibers
(although a small fraction of long structures are retained at all time points following a recovery
period after cessation of exposure), which is consistent with observations in other studies for
fibers that dissolve at moderate rates. In the short-term study performed by the same laboratory,
Hesterberg et al. (1998a), long fiber (>20 |im) RCF-la appears to clear at approximately the
6.29
-------
same rate as the shorter structures (<5 \im), although the scatter in the data (and an unexplained
initial rise in long-fiber RCF) prevent a more careful comparison. Similar results are also
apparent in the data presented for MMVF21. It should be noted that the dissolution rates for
RCF-1 and MMVF21 bracket the estimated dissolution rate for chrysotile asbestos (when the
three are derived from comparable studies [see Section 6.2.4]).
The data presented in Table 3 of Hesterberg et al. (1998b), which are reproduced in Table 6-3,
can also be used to evaluate the time-trend of retention during chronic exposure. The values
presented in Columns 2, 4, and 6 of Table 6-3 present, respectively, measurements of the lung
burden for chrysotile WHO fibers, RCF-1 WHO fibers, and RCF-1 long WHO fibers (>20 |im)
in animals sacrificed immediately following cessation of exposure for the time period indicated
in Column 1. Unlike results reported in some earlier chronic studies based on mass (see below),
there is no evidence from this table (based on fiber number) that chrysotile lung burdens reach a
plateau. Rather chrysotile lung burdens (as well as RCF-1 lung burdens) continue to increase
with increasing exposure.
The data presented in Table 6-3 can also be used to gauge the relative efficiency with which the
chrysotile and RCF fibers are retained. Considering that the number of fibers inhaled (Ninh) over
the period of exposure would be equal to the product of the aerosol concentration (Cair in f/cm3),
the breathing rate of the exposed animal (RB in cmVweek), and time (in weeks):
N^KWt (Eq-6-1)
and the efficiency of retention is simply equal to the quotient of the number of fibers retained
(N,^) and the total number inhaled: Nlung/Ninh, then the efficiency of retention is estimated by the
following simple relationship:
Efficiency of retention=Nlung/(Cair*RB*t) (Eq. 6-2)
By rearranging Equation 6-2, one obtains:
Efficiency of retention*t=(Nlung/(Cair)*(l/RB) (Eq. 6-3)
6.30
-------
Table 6-3. Fraction of Fibers Retained Following Chronic Exposure8
Chrysotile
RCF-1
Exposure
Period
Weeks
WHO
Fibers
f/lung x
10A6
Lung/
Aerosol
Ratio
WHO
Fibers
f/lung x
10A6
Lung/
Aerosol
Ratio
Long
WHO
fibers
f/lung x
10A6
Lung/
Aerosol
Ratio
0.0357
13
26
52
78
104
Aerosol
(f/ml)
"Source:
250
180
1020
853
1600
Concentration
10600
Hesterberg et al.
0.024
0.017
0.096
0.080
0.151
(1998b)
0.009
39
56
119
173
143
187
4.81E-05
0.209
0.299
0.636
0.925
0.765
0.002
3
6
20
21
25
101
1.98E-05
0.030
0.059
0.198
0.208
0.248
Because the breathing rate for the rats in the Hesterberg et al. (1998b) study can be considered a
constant for all experiments, Equation 6-3 indicates that the slope of a plot of Nlung/Cair versus
time should yield estimates of the relative efficiency of retention for each of the fiber types
evaluated. The plot for chrysotile is presented in Figure 6-3. Results from this plot and similar
plots for RCF-1 WHO fibers and long WHO fibers (data not shown), result in the following
estimates of the relative efficiencies of retention (along with the corresponding R2 value for the
fit of the linear trend line):
chrysotile WHO fibers: 0.0014, R2=0.856
RCF-1 WHO fibers: 0.0095, R2=0.694
RCF-1 long, WHO fibers: 0.0026, R2=0.880
Thus, it appears that chrysotile WHO fibers are retained somewhat less efficiently than either
RCF-1 WHO fibers or RCF-1 long WHO fibers. However, whether this is due to less efficient
deposition or more efficient clearance cannot be determined from this analysis. It is also not
possible to determine whether such differences are due to the effects of differences in size
distributions among the various fiber types. Interestingly, based on the data presented by
Hesterberg et al., which indicates that long RCF-1 WHO fibers clear more rapidly than regular
RCF-1 WHO fibers, the differences in the relative retention of these two length categories of
fibers is due primarily to relative efficiency of clearance.
6.31
-------
o
1
41
"c "o1
is
30
CD
O)
c
3
Figure 6.3:
[Chrysotile Lung Burden/Aerosol
Concentration] vs. Time of Exposure
0.16
0.14
0.12-
0.1 -
0.08-
0.06-
0.04-
0.02
0
y=0.0014x
R2 = 0.8556
0 20 40 60 80
Time of Exposure (weeks)
100
120
In a similar study involving chronic exposure to Syrian golden hamsters (Hesterberg et al. 1997),
fiber retention and biological effects associated with exposure to amosite and a series of MMVFs
were evaluated. The amosite was size selected and hamsters were exposed to one of three levels
(0.8±0.2, 3.7±0.6, and 7.3±1.0 mg/m3). Amosite lung burdens were shown to increase regularly
with dose and time of exposure. The time dependence for accumulation of some of the MMVF's
was more complicated. None of the animals were apparently followed for any recovery periods
following cessation of exposure. The authors also indicate that the severity of the effects
observed (inflammation, cellular proliferation, fibrosis, and eventually several mesotheliomas),
appear to correlate well with the concentration of fibers longer than 20 [am.
Earlier Studies. Earlier studies, in which asbestos concentrations tend to be monitored as total
mass tend commonly to show that chrysotile asbestos is neither deposited as efficiently as
various amphibole asbestos types nor is it retained as long (i.e., it is cleared much more rapidly
from the lungs). In fact, several such studies tend to show that chrysotile asbestos concentrations
eventually reach a plateau despite continuing exposure, which suggests that clearance and
deposition come into balance and a steady state is reached. In contrast, amphibole asbestos
concentrations continue to rise with increasing exposure, even at the lowest exposure levels at
which experimental animals have been dosed. Such observations do not appear to be entirely
consistent with those reported in newer studies (see above) that track fiber number
concentrations (in specific size categories). In these newer studies, chrysotile retention is not
observed to level off, but continues to increase in a manner paralleling amosite or other fibers.
As indicated below, however, the limitations associated with these older studies suggest that,
6.32
-------
although it may not be easy to reconcile them quantitatively with the newer studies, results from
these studies are not necessarily inconsistent with those of the newer studies. Moreover, the
trends observed in the newer studies are likely more directly relevant to issues associated with
the induction of asbestos-related disease. The problems with the older studies are:
• the trends seen in the older studies (based on mass) may mask the more important
trends associated with deposition and retention of long, thin fibers. Thus, results
from such studies may not be directly relevant to considerations of risk; and
• the observed differences between chrysotile and the amphiboles may be attributed
to differences in size distribution (among other possibilities). Thus, lacking
detailed information on size distributions, it is difficult to reconcile the results
from the older studies with results from the newer studies, which explicitly track
specific size ranges of fibers.
Given these limitations, the earlier studies are only mentioned briefly.
Middleton et al. (1979) tracked the fate of asbestos (as mass measured by infrared spectroscopy)
in rats following inhalation of several asbestos aerosols (UICC chrysotile A, UICC amosite, and
UICC crocidolite) at multiple concentrations (reported at 1, 5, or 10 mg/m3). To account for
possible differences in the nocturnal (vs. daytime) activity level of rats, several groups of rats
were also exposed in a "reversed daylight" regimen (in which cages were darkened during the
real day and bathed in light during the real night). Acclimatized rats in these groups were thus
dosed at times corresponding to their night. Exposure continued for 7 hours/day, 5 days/week,
for 6 weeks.
Results from the Middleton et al. (1979) study were fit to a three compartment model (originally
proposed by Morgan et al. 1978) and the authors concluded that clearance was independent of
fiber type, but that the initial deposition of fibers was very dependent on fiber type. This was
indicated by a "K-factor" representing the efficiency of initial deposition. Chrysotile showed K
factors that range between 0.17 and 0.36 and vary inversely with the initial exposure
concentration. In contrast, amosite exhibits a K factor of 0.69 and crocidolite a K factor of 1.0
and both are independent of exposure level. Although the design of this experiment precluded
fitting of the shortest two compartments of the model (with half-lives of 0.33 and 8 days, from
Morgan et al. (1978), they did optimize the half-life of the longest compartment. Fibers in this
compartment were cleared with a half-life of 170 days.
In a series of studies, Davis and coworkers (1978, 1980, 1988a, and 1988b) report that retention
of asbestos (measured in terms of mass) appears to be a function of fiber type and surface charge
in addition to fiber size. With regard to fiber type, for example, Davis et al. (1978) report that
substantially more amphibole (amosite) asbestos appears to be deposited and retained in the
lungs of exposed rats than chrysotile. Chrysotile is also apparently cleared more readily than
amosite. However, mineralogical effects should only be judged after adjusting for fiber size.
Rats in the studies by Davis and coworkers were dosed at 6 hours/day, 5 days/week for up to
1 year at dust concentrations of 2, 5, or 10 mg/m3 (depending on the specific experiment). Right
lungs (used for determining lung burden) were ashed and the residue was washed in distilled
6.33
-------
water and filtered. The residue was formed into a potassium bromide disc and asbestos (mass)
content was determined by infrared spectroscopy.
Jones et al. (1988) report that the lung-tissue concentration of amosite increases continually with
exposure (at 7 hours/day, 5 days/week for up to 18 months) and the rate of increase is
proportional to the level of exposure. A leveling off of amphibole concentrations in lung tissue
was not observed in this study as long as exposure continued, even for the lowest level of
exposure (0.1 mg/m3) studied. The lowest exposure concentration evaluated in this study is only
1% of the concentration at which chrysotile lung burdens were shown to reach equilibrium in
other retention studies (see below). Importantly, however, these are only the older studies in
which fiber burden is tracked by mass. The newer studies don't show this effect.
The authors also report lack of any apparent change in size distribution with time among the
fibers recovered from the animal's lungs, which suggests lack of substantial clearance even of
short fibers. However, the longest recovery period following the cessation of exposure evaluated
in this study is only 38 days, which may be too short to allow evidence for differential clearance
as a function of size to become apparent (at least in a chronic study; the time dependence in
chronic studies such as this are more complicated than for short-term studies). Moreover, the
apparent inclusion of lymph nodes as part of the lung homogenate may have caused short fibers
initially cleared from the lung to be added back in. In this study, lungs were recovered intact
including the associated mediastinal and hilar lymph nodes, which were ashed in toto. Ash
residue was washed in acid and water, ultrasonicated and filtered for electron microscope
analysis. Note that such a procedure would include any fibers cleared to local lymph nodes.
In a widely cited study, Wagner et al. (1974), report that amphibole lung burdens increase
continually as long as exposure to amphiboles continues and that amphibole concentrations in
lung tissue decrease only slowly following cessation of exposure. In contrast, chrysotile lung
burdens reach a plateau despite continued exposure. Importantly, asbestos content was estimated
by determining total lung silica content and adjusting for similar analysis on filtered samples of
the original aerosols. Thus, in addition to suffering from the limitations associated with tracking
fiber burden by mass, there are questions concerning the validity of using total silica to represent
asbestos content. Therefore, for these reasons and the additional reason of the lack of controlling
for fiber size, the ability to interpret this study and reconcile its conclusions with those of newer
studies is severely limited.
Chronic Inhalation of Non-fibrous Particulate Matter. A recent study involving chronic
inhalation of non-fibrous materials is helpful at elucidating the relative localization of particles
in rats and primates. Nikula et al. (1997) studied lung tissue from a 2-year bioassay, in which
Cynomolgus monkeys and F344 rats were exposed to filtered, ambient air or air containing one
of three particulate materials: diesel exhaust (2 mg/m3), coal dust (2 mg/m3, particles <7 urn in
diameter), or a 50/50 mix of diesel exhaust and coal dust (combined concentration: 2 mg/m3).
Results from Nikula et al. (1997) indicate that responses to all three particulate materials were
similar. The particles tended to localize in different compartments of the lung in a species-
specific manner:
6.34
-------
• 73% of particles remain in the alveolar lumen of rats, but only 43% in monkeys.
The remainder can be found in the interstitium;
• in both the alveolar lumen and in the interstitium, virtually all of the particles are
observed to be isolated within macrophages; and
• the particles in the interstitium reside in macrophages within the alveolar septa,
the interstitium of respiratory bronchioles, the adventitia and lymphatic capillaries
surrounding arterioles and veins of pulmonary parenchyma, or in the pleura.
It is not known whether free particles penetrate the epithelial lining of the airway lumena and
escape into the interstitium or whether such particles are first engulfed by macrophages and then
transported in their macrophage "hosts" into the interstitium.
Importantly, even after 2 years of exposure, the particles in the interstitium do not appear to have
elicited a tissue response. Also, the aggregates of particle-laden macrophages observed in
alveolar lumena elicited significantly less of a tissue response im monkeys than in rats. Such
responses included: alveolar epithelial hyperplasia, inflammation, and focal septal fibrosis.
The authors further indicate that "epithelial hyperplasia concomitant with aggregation of
particle-laden macrophages in alveolar lumen is a characteristic response to many poorly soluble
particles in the rat lung, both at exposure concentrations that result in lung tumors and at
concentrations below those resulting in tumors. Such a response, however, was not
characteristic of what was observed in monkeys. Among other things, these differences in
responses suggest that rats may not represent a good model for human responses to inhalation of
poorly soluble particulate matter. It would also have been interesting had they tested a "benign"
dust such as TiO2.
In summation:
results of (newer) sub-chronic and chronic retention studies are generally
consistent with those of retention studies that track lung burden following short-
term exposure (Section 6.2.1.1);
there is some indication in these sub-chronic and chronic studies that chrysotile
asbestos may not be retained as efficiently as amphibole asbestos. It is likely,
however, that such distinctions are due more to fiber size than fiber type so that
definitive conclusions concerning such effects cannot be reached until better
studies that properly account for both size and type are conducted; and
although earlier studies that track mass instead of fiber number suggest otherwise,
chrysotile and amphibole asbestos concentrations (when measured by fiber
number) continue to increase with time as long as exposure continues. Due to a
lack in the ability to distinguish among size-dependent effects when lung burdens
are tracked by mass, the results of the earlier studies are not necessarily
inconsistent with the results of the later studies.
6.35
-------
6.2.2 Animal Histopathological Studies
Studies in which the lungs of dosed animals are examined to determine the fate and effects of
inhaled asbestos are helpful for understanding the movement and distribution of retained
particles within the lung and surrounding tissue. Both the newest retention studies and the older
retention studies tend to include at least some of this type of information. They also tend to
indicate a consistent picture of the fate and effects of asbestos. While such studies tend to
confirm that translocation in fact occurs, they are less helpful for elucidating the specific
mechanisms by which translocation occurs.
Newer Studies. Ilgren and Chatfield (1998) studied the biopersistence of three types of
chrysotile ("short chrysotile" from Coalinga in California, "long" Jeffrey fiber from the Jeffrey
mine in Asbestos, Quebec, and UICC-B (Canadian) Chrysotile, a blend from several mines in
Quebec). Both the Coalinga-fiber and the Jeffrey-fiber were subjected to further milling prior to
use. In this study, rats were exposed via inhalation for 7 hours/day, 5 days/week for up to
2 years. Concentrations were: 7.78±1.46 mg/m3 for Coalinga-fiber, 11.36±2.18 mg/m3 for
Jeffrey-fiber, and 10.99±2.11 mg/m3 for the UICC-B fiber. An additional group of rats was also
dosed for a single 24-hour period with Jeffrey-fiber at a concentration of 5,000 f/ml >5 ^m.
Estimates of lung content of chrysotile were based on measurements of total silica content. The
character of the three chrysotile types evaluated in this study was previously reported (Campbell
et al. 1980; Pinkerton et al. 1983). The animal studies were conducted previously with the
overall approach reported by McConnell et al. (1983a,b) and Pinkerton et al. (1984). Based on
the characterization presented in these papers, Coalinga-fiber is short, but not as extremely short
as suggested by the authors:
Ratio of fibers longer than: 5 |im: 10 u.m: 20 urn
Coalinga-fiber: 200: 78: 0.98
Jeffrey-fiber: 591: 220: 78
Such calculations also suggest that the single, "high" dose in this experiment was equivalent
only to a concentration that is approximately 10 times the other concentrations studied, so that it
is equivalent only to a 10-day exposure and small relative to the longer term (up to 2 years)
exposures considered in this study.
Results reported by Ilgren and Chatfield (1998) indicate that the lung burden for short fiber
chrysotile initially increases with exposure, reaches a steady state, and then decreases steadily
following cessation of exposure. Approximately 95% of this material is cleared within 2 years.
Thus, short fibers appear to exhibit the trend suggested by the older chronic retention studies that
tracked burden by mass and is consistent with the newer studies indicating rapid clearance of
short structures (Section 6.2.1.2).
In the studies reported by Ilgren and Chatfield (1998), most short fibers are found initially within
alveolar macrophages and, while concentrations in Type I epithelium, interstitial cells, and
interstitial matrix increase with time, little appears to be taken up by Type II epithelium. Small
6.36
-------
amounts of short fiber chrysotile are also observed to be taken up by endothelial cells. There is
also little sign of inflammation or fibrosis following exposure to the Calidria-chrysotile.
Jeffrey-chrysotile also initially appears to be taken up primarily by alveolar macrophages, but
later becomes most prevalent in the interstitial matrix and, to a lesser extent, in interstitial cells.
Substantial numbers of fibers are also taken up by Type I epithelium and small amounts by
endothelial cells. Similar, but slightly delayed effects were seen for UICC-B chrysotile (which
has a smaller fraction of long fibers and therefore, the authors suggest, takes longer to
accumulate). Note that, by 12 months, the majority of long fibers from both these types were
found in interstitial matrix while the majority of Calidria-material was still found in alveolar
macrophages. This is consistent with observations concerning behavior between short and long
fibers reported by Wright and Kushner (1975), Section 6.2.1.1. At this point, the Jeffrey-
material also caused substantial thickening of the basement membrane and most fibers in the
interstitium were trapped within the collagenous matrix. The authors note that some of the most
severe interstitial changes occurred adjacent to areas of bronchiolar metaplasia. Such effects
were not seen with Calidria-exposure.
Thickened basement membranes, calcium deposits, metaplastic changes, and structural
abnormalities were all observed with long fiber exposure, but not with Calidria-exposure.
Interstitial macrophages also showed morphological changes following phagocytosis of fibers.
While Calidria-material was about evenly distributed between interstitial matrix and cells, the
vast majority of long fiber material was found in the matrix. Movement into the matrix was also
observed to increase even after exposure ceased. With time, the number of long fibers in
interstitial cells declined modestly, but declined precipitously for Calidria-material.
Jeffrey-fibers accumulated in Type I epithelial cells during exposure and then levels decreased
slowly after exposure ceased. UICC-B fibers accumulated more slowly, never reaching the same
levels as for Jeffrey and decreased more rapidly. Concentrations of Calidria-fibers in Type I
epithelial cells was low at all time points.
Type II epithelial cells accumulated very few fibers of any type (although they took up slightly
more Jeffrey-fiber than the others). All three fiber types caused substantial increases in
interstitial cells (mostly macrophages) at 3 months and this increase persisted for the Jeffrey-
fiber, but decreased to background after 24 months for the other two fiber types. Fibroblast
numbers also increased with the long fiber types, but not Calidria-chrysotile.
Type II epithelial cells showed decreases in volume and number that persisted until exposure to
long fiber ceased and these cells displayed dramatic structural aberrations despite absence of a
fiber load. One possible explanation for the observed changes in Type II cells, especially the
reduction in their number, is that they were undergoing terminal differentiation to Type I cells
(see Section 4.4). In fact, the apparent absence of fibers observed within Type II cells might be
explained by such cells taking up fibers, but being induced to terminally differentiate, once fibers
are accumulated. Overall, Type II cells displayed greater cellular response than Type I cells
(which might also suggests a role for cytokines). All effects were observed to be fiber length
dependent and all were exaggerated following exposure to long fiber material.
6.37
-------
Rats exposed to long fiber had numerous accumulations of dust-laden interstitial macrophages
and/or small focal accumulations of dust within the interstitium at the end of the lifetime study,
but such changes were not observed for Calidria-exposed animals.
The lung burden for rats exposed to the single, "high" Jeffrey-fiber exposure (based on total
silica) at 12 months (i.e., 12 months post-exposure) was not different from controls. Therefore,
the authors conclude that short-term, "high" exposures are rapidly cleared (even exposures
containing substantial quantities of long fibers). Other changes induced by the single, short-term
high exposure of Jeffrey-fiber that was followed for 24 months also showed reversion to close to
background status. Importantly, these observations are not based on quantitation of fiber burden
in lung tissue. Rather they are inferred by observing the effects caused by the presence of fibers.
This may suggest, for example, that more than 10 day's worth of exposure would be required at
this level of exposure before irreversible lesions develop.
In the study by Hesterberg et al. (1997), which was previously discussed (Section 6.2.1.2), the
authors note (among dosed hamsters) that the magnitude of cellular effects appeared to differ
among the fibrous glasses as a function of their relative biodurability. For animals dosed with
the least biopersistent glass, only transient effects (influx of macrophages, development of
microgranulomas) were observed and these did not progress further. For the more persistent
glasses, injury progressed through more intense inflammation, interstitial fibrosis, pleural
collagen deposition, mesothelial hypertrophy and hyperplasia and eventually, mesothelioma.
The authors also suggest that amosite appears to be more potent than chrysotile, even when
aerosols contain comparable numbers of fibers longer than 20 \im.
Choe et al. (1997) exposed rats to chrysotile and crocidolite (both NIEHS samples) by inhalation
for 6 hours/day, 5 days/week, for 2 weeks. The rats were then sacrificed and their pleural
cavities lavaged. Results indicate that significantly more pleural macrophages were recovered in
plueral lavage fluid at one and 6 weeks following exposure than sham exposed rats. The
centrifuged pellet from pleural lavage fluid from one of four rats also exhibited long (>8 u.m),
thin (<0.5 |im) crocidolite fibers (1 week following exposure). The concentration of fibers in
this pellet suggested approximately 1 f per 4,000 cells in the pleura. Note that chrysotile rats
were not examined for fiber content.
Older Studies. In the series of studies by Davis et al. (1978, 1980, 1985, 1986a), the authors
generally report similar histopathological observations that emphasizes a marked distinction
between effects from long and short fibers. From the 1986 study, for example, Davis et al.
report that at the end of 12 months of exposure, rats exposed to long fibers (amosite in this case)
exhibited deposits of granulation tissue around terminal and respiratory bronchioles. They
further indicate that the granulation tissue consists primarily of macrophages and fibroblasts with
occasional foreign body giant cells.
As the animals aged, there was increased evidence of collagen deposits in these lesions and the
oldest lesions consist mainly of acellular, fibrous tissue. The alveolar septa in these older
animals showed progressive thickening. Initially, this was apparently due primarily to
hyperplasia of Type II epithelial cells, but with time was increasingly due, first, to reticulin and,
later, to collagenous deposits in the septal walls. Asbestos dust was frequently visible in these
deposits. Epithelial cells lining alveoli adjacent to the oldest lesions also tended to become
6.38
-------
cuboidal in shape. As the animals aged, these areas of interstitial fibrosis became more
extensive.
In contrast, animals exposed to short fibers (also amosite in this case) showed no such lesions
(peribronchial fibrosis) at any point in time. At the end of exposure, the lungs of these animals
contained large numbers of pulmonary macrophages packed with fibers, but these cells remained
free in the avleolar spaces. The authors report that large numbers of laden macrophages
sometimes aggregated in alveoli close to respiratory bronchioles, but that there would be no
formation of granulation tissue or thickening of alveolar septa at these locations. Thus, with the
exception of the presence of dust-laden macrophages, the structure of the lung of these rats was
not altered.
In the Davis et al. (1987) study of chrysotile, a slight variation of the above scenario is worth
noting. In this study, the development of peribronchial fibrosis was reported for animals dosed
both with the long fiber material and with the short fiber material. However, the authors also
report that the short fiber chrysotile in this study in fact contains a sizable fraction of longer
structures and this finding was corroborated by more formal size characterization (Berman et al.,
unpublished) later conducted in support of a study to evaluate the effects of size (Berman et al.
1995).
In this study, Davis et al. also reported observations on the morphological changes observed in
the mesothelium during these studies. The authors indicate that the older animals in these
studies exhibit "areas of vesicular pleural metaplasia consisting of loose, fibrous tissue
containing large vesicular spaces lined with flattened cells". The authors also report that
examination in previous studies indicates that these cells are of a mesothelial type.
Davis et al. report that, "...occasionally the walls between vesicular spaces were so thin that they
consisted of two closely opposed layers of extended and flattened cells with no basement
membrane in between them. Where cells were supported by areas of fibrous tissue, a basement
membrane was present. While no method for the direct quantification of this pleural metaplasia
has been developed, its occurrence is closely related to the presence of advanced interstitial
fibrosis or adenomatosis in the lung tissue and it is particularly common where patches of this
type of parenchymal lesion have reached the surface. It is not known whether such lesions are
precursors to mesothelioma. Davis et al. also note that neither of the two mesotheliomas
observed in this study showed histological patterns consistent with the observed vesicular
hyperplasia.
Brody et al. (1981) tracked the distribution of chrysotile following inhalation by rats. Asbestos
was initially deposited almost exclusively at alveolar duct bifurcations. In agreement with
Pinkerton et al. (1986), the degree of deposition appeared to be an inverse function of the path
length and bifurcation number for each alveolar duct. Uptake by macrophages and type 1
epithelial cells were observed following deposition. Asbestos was observed both in lipid
vesicles and free in the cytoplasm of type 1 cells. After 8 days, alveolar duct bifurcations
became thickened with an influx of macrophages. Asbestos was also observed in basement
membrane below the epithelium. Apparently, structures had been transported through type 1
cells to the basement membrane. Once in the basement membrane, asbestos may enter the
interstitium. Predominantly short structures were monitored in this study. Long structures were
6.39
-------
not readily observed (but this is likely a counting problem; under such circumstances, short
structures may serve as surrogates for the presence of other structures).
Intratracheal Instillation. Bignon et al. (1979) studied the rate of translocation of various
materials in rats. Chrysotile, crocidolite, and glass fibers were intrapleurally injected into rats
and their concentration was monitored as a function of time in lung parenchyma and other tissues
removed from the pleura. Within 1 day following injection, asbestos was detectable in lung
parenchyma. After 90 days, asbestos was found in all of the tissues analyzed. Based on the rate
of translocation to the lung, crocidolite migrates about 10 times more rapidly than chrysotile (on
a mass basis). The rate of migration of glass is in between the two asbestos types. Structures
initially found in the lung were significantly shorter than the average size of structures injected.
After 7 months, however, the average lengths of structures in all tissues monitored were longer
than the average length of structures originally injected. Thus, short structures migrate more
rapidly than longer structures (possibly by a different mechanism), but long structures eventually
translocate as well. Within a target tissue, preferential clearance of short structures also
contributes to observed increases in the average length of the structures with time.
Studies of Non-Fibrous Particulate Matter. Studies of the fate and effects of respirable, non-
fibrous particulate matter provide evidence for at least one mechanism by which particles (and
fibers) may be transported to the interstitium.
Li et al. (1997) evaluated the effect of urban PM 10, carbon black, and ultrafine carbon black on
rats following intratracheal instillation (0.2 ml volume instilled containing between 50 and 125
|ig of particles). After 6 hours, there was a noted influx of neutrophils (up to 15% of total cells
observed in bronchioalveolar lavage-BAL fluid) and increases in epithelial permeability was
surmised based on increased total protein (including increases in levels of lactate dehydrogenase,
which is a marker for cell membrane damage) in BAL fluid.
Conclusions. Overall, observations among both the newer and the older studies (including the
study by Wright and Kuschner 1975, see Section 6.2.1.2) tend to be highly consistent,
particularly with regard to the distinction between the effects of short and long fibers. Typically,
long fibers initially produce substantial inflammation characterized by an influx of macrophages
and other inflammatory cells. Ultimately, exposure to such fibers cause thickening of alveolar
septa (particularly near avleolar duct bifurcations) due to a combination of epithelial hyperplasia
and deposition of reticulin and, later, collagen resulting in interstitial fibrosis. In contrast, short
fibers cause an initial influx of macrophages and, long-term, show persistent accumulations of
fiber-laden macrophages both in alveolar lumena and in pulmonary lymph nodes, but otherwise
no structural changes are observed in the lung tissue of these animals.
These studies also provide ready evidence of the effects of fiber translocation, but generally offer
only limited evidence for elucidating the mechanisms by which such translocation occurs. There
is evidence that Type I (and possibly Type II) epithelium phagocytize particles and fibers and it
is possible that such fibers may be passed through to the basement membrane and the
interstitium. For Type I cells, the distance between the alveolar lumen and the basement
membrane averages less than 1 ^m in any case (Section 4.4). Certainly fibers are also observed
in the interstitium. Particulate studies also indicate that oxidative stress induced by particulate
matter and fibers may cause morphological changes in Type II cells with consequent loss of
6.40
-------
integrity of the epithelium, which increases it's permeability overall and may also allow
diffusional passage of particles and fibers.
6.2.3 Human Pathology Studies
Human pathology studies provide additional information concerning the nature of asbestos
deposition, clearance, and retention. These are the studies in which lung burdens are measured
in samples of lung tissue and correlated with the exposures received by the individuals from
which the lung samples derive.
Among the advantages of human pathology studies is that they provide direct insight into the
behavior of asbestos in humans. They are also limited, however, by the lack of ability to obtain
time-dependent estimates of lung burden (because samples are derived from deceased
individuals), by the manner in which lung tissue is stored (several of the fixatives employed to
store tissue samples have been shown to enhance dissolution of asbestos (Law et al. 1990, 1991),
by the manner in which samples are prepared for asbestos analysis, by the manner in which
asbestos is analyzed, and by the limited ability to re-construct the uncontrolled exposures
experienced by study subjects (Section 5.2).
Perhaps most importantly, the ability to construct anything but the coarsest quantitative
comparisons across subjects is also typically limited by use of "opportunistic" tissue samples
(i.e., use of samples that happened to have been collected and stored during autopsy or necropsy)
because such samples are not controlled for location on the respiratory tree (i.e., the linear
distance and branch number from the trachea) that is represented by the sample. Because it has
previously been shown that deposition is a strong function of such location (see, for example, Yu
et al. 1991), comparisons across lung samples not controlled for these variables are problematic.
It has also been shown that samples collected from adjacent locations in lung parenchyma can in
fact exhibit strikingly different fiber concentrations due specifically to the differences in the
location of the respiratory tree represented by the alveoli and respiratory bronchioles in the
spatially adjacent samples (Brody et al. 1981; Pinkerton et al. 1986 [Section 5.2]).
Despite the above indicated cautions, when interpreted carefully, human pathology studies can
provide useful evidence regarding fiber deposition, clearance, and retention in the human lung.
Newer Studies. Among the most recent studies, Finkelstein and Dufresne (1999) evaluated
trends in the relationship between lung burdens for different fiber types and different size ranges
as a function of historical exposure, the duration of such exposure, the time since last exposure,
and other variables. The analyses were performed among 72 cases from which tissue samples
could be obtained (including 36 asbestosis cases, 25 lung cancer with asbestosis cases, and 11
mesothelioma cases).
Due to the excessive scatter in the data, most of the analyses presented depend on "Lowess
Scatterplot Smoothers". Moreover, although not stated, it is likely that the tissue samples
obtained were "opportunistic" in that they were not matched or controlled for relative position in
the respiratory tree.
6.41
-------
Finkelstein and Dufresne (1999) employed the multi-compartment model developed by Vincent
et al. (1985) to evaluate trends in their data. The features of this model include:
• a compartment representing conducting airways that are cleared within minutes to
hours by muco-ciliary transport;
• a compartment representing the subset of fibers reaching the pulmonary portion of
the lung that are cleared by alveolar macrophages and transported to the muco-
ciliary escalator. This type of clearance is also considered relatively rapid with
half-lives of no more than several days to several weeks. Macrophage clearance
is also considered size-dependent and long fibers are cleared less efficiently than
short fibers;
• when sufficient dust is inhaled (or dust is sufficiently cytotoxic) to impair the
motility of macrophages (either by volumetric overload or by toxicity), a
sequestration compartment forms that consists of laden, but immobile,
macrophages. Although this compartment may ultimately be cleared to lymphatic
drainage, such clearance is assumed to be slow and size dependent (with half-
lives of 2 or 3 years for short fibers and 8 years for fibers longer than 10 urn); and
• once the macrophage system is overloaded, fibers may cross the alveolar
epithelium and reach the interstitium and this compartment must be cleared by
transport to lymphatic drainage, which is assumed to be an extremely slow
process.
Finkelstein and Dufresne (1999) indicate that chrysotile splits both longitudinally and
transversely in the lung and that chrysotile lung burdens decrease significantly with time since
last exposure (with short fibers clearing even faster than long fibers), while tremolite burdens do
not appear to decrease with time since last exposure. They also suggest that smoking does not
appear to affect clearance rates.
Finkelstein and Dufresne (1999) indicate that these type of studies are not useful for examining
the behavior in rapidly clearing compartments of the lung, but they may provide insight
concerning the more slowly clearing compartments. Based on their modeling, they suggest that
tremolite is transferred to the sequestration compartment at rates that are 6-20 times that of
chrysotile (which they indicate is comparable to what was found for crocidolite by de Klerk).
The authors suggest that retained chrysotile concentrations tend to plateau after accumulation of
about 35 years of exposure, while tremolite concentrations continue to increase. They also
suggest, however, that chrysotile concentrations may begin to increase again after 40 years
(suggesting the overload is eventually reached for chrysotile as well). Reported half-lives from
the long-term compartment are:
6.42
-------
Chrysotile
fibers shorter than 5 urn 3.8 years
fibers 5-10 \im in length 5.7 years
fibers longer than 10 urn 7.9 years
Tremolite
fibers shorter than 5 urn 14.3 years (not different from °°)
fibers 5-10 urn in length 15.8 years (not different from °°)
fibers longer than 10 urn 150 years (not different from ~)
In a case-control study, Albin et al. (1994) examined the lung burdens of deceased workers from
the asbestos-cement plant previously studied for mortality (Albin et al. 1990). In this study,
details of the procedures used to prepare lung tissue for analysis were not provided. It is also
assumed that available tissue samples were "opportunistic" in that they were not matched or
controlled for relative position in the respiratory tree.
Results from Albin et al. (1994) are consistent with (but do not necessarily support) the
hypothesis that chrysotile is cleared more readily from a long-term sequestration compartment
than amphiboles. The authors also report that chrysotile fibers observed in this study are much
shorter than the amphibole fibers observed, so that differences in clearance rates might be
attributable to size differences. The authors also suggest that clearance is impaired by fibrosis.
Studies of Quebec Miners. Several authors also studied the lung content of various groups of
deceased chrysotile miners in Quebec and found, despite overwhelming exposure to chrysotile
from the ore (which contains only trace quantities of tremolite, see Case et al. 2000 and
Sebastien et al. 1986), a substantial number of fibers (in some cases the majority of fibers)
observed in the lungs of deceased miners from this area are tremolite. Thus, for example:
• in a study of lung burdens in 6 mesothelioma victims, Churg et al. (1984) showed
that amphiboles structures were 5-15 times as plentiful as chrysotile despite the
predominantly chrysotile exposure;
• in a study of lung tissue from 20 asbestosis cases, Pooley (1976) found substantial
concentrations of tremolite in the lungs of deceased Quebec chrysotile workers;
• in a much larger study comparing lung burdens of Quebec workers with those
from the South Carolina textile mill, which has also been extensively studied for
asbestos-related mortality (Section 7.2.3), Sebastien et al. (1989) examined 161
lung tissue samples (89 from the Quebec mines). Results from this study indicate
that geometric mean tremolite fiber concentrations were more than 3 times mean
chrysotile fiber concentrations (18.4 vs. 5.3 f/ng dry lung tissue) among the
deceased Quebec miners evaluated. It was also found that, despite these
differences, the overall size distributions of tremolite and chrysotile fibers
observed in lung tissue were approximately the same, although this conclusion is
suspect. A more detailed discussion of the results of this study is provided in
Section 7.2.3; and
6.43
-------
• in a more focused study using a subset of the lung samples evaluated by Sebastien
et al. (1989), Case et al. (2000) found that the majority of long fibers (longer than
18 urn) in the lungs samples from the deceased Quebec miners that he examined
were in fact composed of tremolite. These authors also found substantial
concentrations of long tremolite fibers (relative to chrysotile fibers) in deceased
workers from South Carolina and even higher concentrations of commercial
amphibole fibers (amosite and crocidolite) in the lungs of these workers. Thus, in
addition to suggesting the relative persistence of amphibole asbestos compared to
chrysotile in vivo, this finding also suggests that the accepted notion that the
South Carolina cohort studied by Dement et al. and McDonald et al. (see
Appendix A) was exposed almost exclusively to chrysotile may not be correct.
This study is discussed more fully in Section 7.2.3.
These observations provide evidence that either amphibole (tremolite) asbestos is deposited
more efficiently in the compartments of the lung where clearance is slow or chrysotile asbestos
is cleared more rapidly and efficiently from even the slowest clearing compartments of the lung
(or both). Moreover, this conclusion appears to apply similarly to both short and long fibers.
In another study, McDonald et al. (1993) suggest more specifically that lung burden data from
Quebec indicate little evidence of decreasing chrysotile concentration with time since last
exposure. Rather they suggest simply that tremolite is initially deposited in the deep lung more
efficiently. These authors also indicate that tremolite fibers are mostly optical while chrysotile
are mostly "Stanton" or thinner.
McDonald et al. (1993) report good correlation of both tremolite and chrysotile with estimated
past exposures, which contrasts with the findings of the evaluation we conducted on the data
from Sebastien et al. (Section 7.2.3). In McDonald et al.(1985), reported geometric mean
measurements for the various fibers in lung are: tremolite: Ixl06-18.2xl06, chrysotile :
1.5xl06-15.7xl06, respectively, when exposure varied from <30 mpcf to >300 mpcf. However,
note that, if these are PCM measurements, this may not be telling the whole story. The authors
also report that 66% of those who died 10 years since first exposure and half of those who died
30 years since first exposure showed high chrysotile concentrations in their lungs.
Unfortunately, without access to the raw data from this study, it is not possible to identify the
route of the apparent discrepancy between the findings of this study and those reported above for
ostensibly similar studies.
Some of the pathology studies that have been published suggest that at least some clearance
mechanisms show a dependence on fiber size, which is consistent with what is observed in
animal studies (Section 6.2.1). Notably, for example, Timbrell (1982) studied deceased workers
and relatives from the Paakkila anthophyllite mine in Finland. He found that structures shorter
than 4 urn and less than 0.6 p,m in diameter are completely cleared from healthy lungs. The
efficiency of clearance decreases slowly with increasing size. Structures longer than 17 \im and
thicker than 0.8 |im in diameter are not significantly cleared. The study is based on a
comparison of structure size distributions in lungs compared to the structure size of the material
in the original dust exposure. Timbrell also noted that asbestosis suppresses the removal
process.
6.44
-------
When considering the dependence of clearance on size (particularly via mechanisms involving
phagocytosis), it is necessary to address differences in human and animal physiology. Due to
differences in the morphology, for example, human macrophages have been shown capable of
phagocytizing larger particles and longer fibers than macrophages found in mice and rats
(Krombach et al. 1997 [for details, see Section 4.4]). Thus, the range of fibrous structures that
are efficiently cleared from human lungs is expected to include longer fibers than the range
efficiently cleared in mice or rats. Unfortunately, given the limited precision of the available
data, the size ranges that are reported to be cleared efficiently in rats and humans, respectively,
cannot be easily distinguished.
Several human pathology studies also support observations from animal studies indicating that
clearance may be inhibited by the development of fibrosis (Albin et al. 1994;Churg et al. 1990;
Morgan and Holmes 1980) or by heavy smoking. However, other studies do not indicate such
hindered clearance either with smoking (Finkelstein and Dufresne 1999) or with fibrosis.
Older Studies. Morgan and Holmes (1980) examined tissue samples from 21 patients in
England (10 who died of mesothelioma, 3 who died of lung cancer, and 8 who died of other
causes). In this study, formalin-fixed tissue samples were digested with hypochlorite. The
residue was then rinsed, diluted, and an aliquot filtered. The filter was mounted on a microscope
slide and clarified for analysis by phase contrast optical microscopy. Importantly, the authors
note that chrysotile fibers were ignored in this study because they would not generally have been
detected by this technique. Portions of the filters were also carbon coated and prepared for TEM
analysis. Based on the observation that only 19% of the fibers observed in this study were
between 2.5 and 5 (j.m, when the authors expect airborne distributions to contain closer to 90% of
the fibers within this size range, the authors conclude that short fibers are preferentially cleared
from the lung. They also conclude, based on one subject with asbestosis whose lung tissue
exhibited 72% short fibers, that asbestosis hinders clearance. The authors also note that fewer
than 1% of ferrugenous bodies (iron-coated asbestos bodies) are <10 urn in length, which
indicates (in agreement with previously published work) that such bodies seldom form on short
fibers. They also suggest that virtually all fibers longer than 20 (im tend to be coated in the
distributions they observe.
Le Bouffant (1980) studied the concentrations, mineralogy, and size distributions of asbestos
fibers found in the lungs and pleura of deceased asbestos workers. Based on the analysis,
Le Bouffant (1980) found that the average ratio of chrysotile fiber concentrations found in the
lung versus the pleura is 1.8 while for amosite the ratio is 34. This indicates that chrysotile
migrates from the lung to the pleura more rapidly than amphiboles resulting in a higher fraction
of total fibers in the pleura being composed of chrysotile (3% in the lungs versus 30% in the
pleura). With regard to size, the researchers found the size distribution of amosite is virtually
identical in the lung and pleura while chrysotile fibers found in the pleura are much shorter than
chrysotile fibers found in lung tissue. This suggests that the movement of chrysotile is a result of
a combination of translocation and degradation to shorter fibers (or that tissue samples have been
contaminated with environmentally ubiquitous short, chrysotile structures). The authors indicate
that chrysotile fibers apparently degrade to shorter fibers more rapidly than amosite and
translocate to the pleura more rapidly than amosite. Thus, a greater fraction of chrysotile fibers
(albeit short fibers) reach the pleura than amosite fibers over fixed time intervals. However, the
6.45
-------
results of this study also confirm that the longer amosite fibers do eventually translocate,
although on a much more extended time scale than the translocation of chrysotile.
Importantly, the results of this study need to be evaluated carefully. Boutin et al. (1996) showed
that the majority of asbestos fibers in the pleura (particularly the long fibers) are aggregated in
localized "black spots" (which surround the sites of lymphatic drainage). Thus, if the tissue
samples analyzed by Le Bouffant (1980) do not contain representative sets of such spots, the
conclusions drawn by Le Bouffant (1980) may be subject to question.
In summation, human pathology studies tend generally to support the findings of other studies
regarding the size effects of asbestos (i.e., short fibers tend to clear more rapidly than long fibers,
which can be retained in pulmonary tissues for extended periods). They also appear to highlight
drastically different behavior between chrysotile and the amphibole asbestos types (particularly
tremolite) regarding the heavily favored retention of the latter, which has also been indicated in
animal studies. Unfortunately, the ability to draw quantitative conclusions from human
pathology studies is hampered by the severe limitations of these studies (Section 5.2).
6.2.4 Studies of Dissolution/BioDurability
Although asbestos minerals are relatively insoluble in vivo in comparison, for example, to
various fibrous glasses or other man-made mineral fibers (see, for example, Hesterberg et al.
1998a or Bastes and Hadley 1996), they do eventually dissolve in the body. Therefore, this
pathway may contribute importantly to the overall biological clearance of asbestos. Moreover, it
has been suggested by several researchers (see, for example, McDonald 1998a and other
references cited below) that differences in biodurability between chrysotile and the amphiboles
may at least partially explain the disparate potencies observed for these fiber types toward the
induction of mesothelioma and, potentially, lung cancer (see Sections 7.2.4.2 and 7.3.3.2).
Note that the term "biodurability" is used here to indicate the persistence of a
particle or fiber attributable specifically to solubility (in the absence of other
clearance or degradation mechanisms). In contrast, the term "biopersistence" is
used to indicate the overall persistence of a particle or fiber in the body
attributable to the combined effect of all mechanisms by which it might be
removed. Thus, for example, while biopersistence can be evaluated in vivo,
biodurability can best be inferred from in vitro dissolution studies so that effects
from other clearance mechanisms can be eliminated.
Several studies further indicate that both the in vivo biopersistence and the bio-activity
(including carcinogenicity) of various fiber types may be linked to their observed, in vitro
dissolution rates (Bernstein et al. 1996; Bastes and Hadley 1995, 1996; Hesterberg et al. 1998a,
1998b). Such studies, however, typically involve fiber types with dissolution rates that are rapid
relative to the rates of clearance by other mechanisms (Sections 6.2.1.1 and 6.2.1.2). In such
studies, moreover, the various types of asbestos are typically employed as negative (insoluble)
controls. In fact, most of these studies are based on experiments with rats and the 2-year lifetime
of a rat is comparable to the anticipated lifetime of chrysotile asbestos in the body and short
compared to the anticipated lifetimes for the amphiboles (see below). Therefore, such studies
are not particularly sensitive to differences in the relative biodurability of the different asbestos
6.46
-------
types. In fact, in the majority of these studies, the dissolution rates reported for asbestos were
derived indirectly by analogy with other minerals or are quoted from other studies that derive
rates similarly and may therefore be somewhat unreliable. Nevertheless, a review of a subset of
these studies is instructive.
Bastes and Hadley (1995 and 1996) report a simple model that reasonably predicts the relative
fibrogenicity and tumorigenicity for a range of synthetic fibers based on the dissolution rates of
the fibers measured in vitro. The authors found that they could explain observations by
assuming that the effects of the various fibers are a function of an adjusted dose that accounts for
biodurability. Thus,
F=f(ax)
where "F" is the observed incidence of the endpoint, "f' is the dose-response function proposed
for the effect, "x" is the measured dose, and "a" is an adjustment factor that accounts for
durability.
In the model, "a" is determined simply as "td/tL" where td is the time that a fiber of diameter, "D"
remains in the lung and tL is the lifetime of the exposed animal (e.g., 2 years for rats). This
simple model reasonably reconciles the results observed in animal inhalation and injection
studies of MMVF's, RCF's, and asbestos for endpoints including lung tumors, degree of fibrosis,
and (for intrapleural injection studies) mesothelioma. Based on a chi-square test, the simple
model is shown to adequately fit the data to a number of databases reviewed. In contrast, the
unadjusted doses do not. Importantly, the dissolution rates used for the various asbestos
minerals in this study were estimated by analogy with similar minerals and therefore may be
unreliable.
In studies comparing in vivo biopersistence with dissolution rates measured in vitro, Bernstein et
al. (1996) and Hesterberg et al. (1998a), indicate that it is necessary to consider only long fibers
(typically longer than 20 |im), because shorter structures are typically cleared by other
mechanisms. They also indicate, at least for this type of study in which whole lungs were
homogenized and dissolved prior to preparation for asbestos analysis,2 that clearance is initially
rapid. This reflects muco-ciliary clearance from the upper respiratory tract. Therefore, it is
clearance of long fibers from a longer term pool that tracts in vitro dissolution rates.
These authors also report that long, soluble fibers (longer than 20 |im) are actually cleared more
rapidly than short fibers in these studies. They indicate that this is likely due to long fibers being
too long to be effectively phagocytized by macrophages so that they are left to dissolve in the
extracellular fluid at neutral pH. Shorter fibers are effectively taken up by macrophages so that
dissolution is hindered by the more acidic environment of the phagosomes (pH 4.5) and by the
limited volume of fluid within which to dissolve.
When whole lungs are homogenized to determine lung burden, this includes the largest airways, which
initially contain substantial concentrations of material that is rapidly cleared by the muco-ciliary escalator.
However, because this material dominates the quantity of material observed, such studies are not useful for tracking
the longer term clearance processes that occur in the deep lung.
6.47
-------
Law et al. (1991) studied the dissolution of a range of fibers in solutions used as common
fixatives for biological samples. The authors report that chrysotile and crocidolite, as well as
many other fibers, dissolve at measureable rates in the fixatives studied (Karnovsky's fixative
and formalin fixative). They therefore recommend that fiber concentrations and size
distributions obtained from tissue samples stored in such fixatives should be evaluated carefully
to account for the possible effects of the fixatives.
Although Coin et al. (1994) reported seeing no effective reduction in long fiber (>16 |o.m)
chrysotile (nor other evidence of dissolution) in their study of fiber biopersistence, the limited
time frame of this study (30 days) may have been too short to allow detectable changes to
accumulate.
The most consistent data for the comparative biodurability of chrysotile and the amphiboles
(specifically crocidolite) is found in two in vitro studies of the dissolution rates of fibers that
were conducted under comparable conditions. In the first of these studies, Hume and Rimstidt
(1992) measured the dissolution rate for chrysotile asbestos at neutral pH under conditions
analogous to biological systems. The dissolution rate that they report for chrysotile converts to:
Kdjss=12.7 ng/cm2-hour and this is reportedly independent of pH. In a comparable study Zoitus
et al. (1997) report the following dissolution rate for crocidolite: Kdiss=0.3 ng/cm2-hour, which is
40 times slower than for chrysotile. Dissolution rates for several MMVF's and RCF-1 are also
reported in the latter paper, which are listed from fastest dissolving to slowest in Table 6-4.
Table 6-4. Measured in vitro Dissolution Rates for Various Fibers8
Fiber Type
MMVF 10
MMVF 1 1
MMVF 22
MMVF 21
Chrysotile
RCF 1
Crocidolite
KdiM (ng/cm'-hr)
259
142
119
23
12.7b
8
0.3
"Source: Zoitus et al. (1997)
bSource: Hume and Rimstidt (1992)
Note that dissolution rates for other amphiboles, such as amosite are probably no more than a factor of
two or three different than that reported above for crocidolite (see, for example, Hesterberg et al.
1998a,b).
6.48
-------
To compare the effect of biodurability on the in vivo biopersistence of asbestos and other fiber
types, both the detailed kinetics of dissolution and the distribution of fiber sizes must be
considered.
As reported by Zoitus et al. (1997), at a sufficiently high rate of fluid flow, the rate of mass loss
from a fiber is proportional to its surface area, A. Thus:
dM/dt=-kA. (Eq. 6-4)
This means that for a uniform mass fiber dissolving congruently:
l-(M/M0)05=2kt/D0p. (Eq. 6-5)
where:
M is the mass at time t;
M0 is the initial mass at time t=0;
D0 is the initial diameter of the fiber; and
p is the density of the fiber.
Substituting the equation relating the mass and the diameter of a fiber (M=p7id2h/4) into the
above equation, cancelling terms, and rearranging indicates that (during dissolution) the diameter
of a fiber decreases linearly with time:
D=D0 -2kt/p. (Eq. 6-6)
where:
D is the diameter at time t; and
all other terms have been previously defined.
Furthermore, the rate of reduction in radius is given by: k/p. Based on the dissolution rates given
above for chrysotile and crocidolite, the radius reduction rates (vrad) for these fiber types are
determined to be: 1.26xlO"8 urn/sec and 2.6x10"'° u.m/sec, respectively. Thus, the dissolution of
each fiber is a zero order process (i.e., the rate is constant with time and independent of
concentration). Given these rates, a chrysotile fiber 1 (am in diameter will disappear in
approximately a year (3.9xl07 sec) and a crocidolite fiber of the same diameter in approximately
60 years (1.9x109 sec).
The number rate of disappearance of a population of fibers due to dissolution is a function of the
rate of radial reduction for the fiber type and the distribution of fiber diameters in the population.
The time at which the entire population finally dissolves can be estimated simply by dividing the
radius of the largest fiber by the radius reduction rate, vrad, that is appropriate for the fiber type.
The number of fibers remaining from the population at time t will be equal to the number of
fibers in the original distribution with radii larger than vradt for the reduction rate that is
appropriate for the fiber type.
6.49
-------
Note that dissolution will not cause an immediate reduction in fiber concentration.
The number of fibers will not begin to decrease until sufficient time has elapsed
for the thinnest fibers to completely dissolve. Bastes and Hadley (1994) therefore
recommend tracking the time dependence of the mode of the distribution of fiber
diameters to best gauge the effects of dissolution in vivo.
Importantly, fibers in vivo will only dissolve at the rates predicted by the above equations if the
fluid in which they are dissolving flows past the fibers sufficiently rapidly to prevent saturation
from limiting the rate (Mattson 1994). Especially for slow dissolving materials of limited
solubility like asbestos, it is expected that the observed dissolution rate in vivo will generally be
slower than the rates predicted based on in vitro measurements. Even for more rapidly
dissolving fibers like most fibrous glasses and manmade mineral fibers, dissolution is hindered
in compartments of the body in which the volume of available solute is limited.
In summary, dissolution is a zero-order (i.e., constant with time, independent of concentration)
clearance mechanism that is dependent on fiber mineralogy, that the effect it has on fiber
populations (concentrations) is a function of the distribution of fiber diameters within the
population, and that the theoretical rate of dissolution may not be achieved in all tissues in all
compartments of the lung or mesothelium due to limits in the rate of in vivo solute flow.
6.2.5 Dynamic Models
Unlike particle deposition in the lungs, which is an entirely mechanical process, clearance,
transport, and degradation mechanisms tend to be complex biochemical processes. Due to the
incomplete understanding of such processes, state-of-the-art modeling of degradation and
clearance is not as advanced as that for deposition. Even the most sophisticated of degradation
and clearance models remain semi-empirical. Although, current models in this area show
general agreement with the sparse, available data, there are clearly areas of weakness that require
additional research. Nevertheless, the models provide a good indication of the kinds of
processes that are important in the body and their overall constraints. It is also noted that models
for degradation and clearance in humans tend to be better developed than those for animals,
primarily due to confounding uncertainties associated with animal ventilation rates. An
overview of the state of the art, which was current as of the date of publication, can be found in
Stoberetal. (1993).
According to Stober et al. (1993), the general conclusions that can be drawn from the current
models are that:
• clearance from ciliated airways is rapid, independent of particle/fiber type,
apparently independent of particle size, and can be described as the sum of two,
weighted exponentials (i.e., an assumed combination of two first-order decay
processes), although the process may in fact be zero order (i.e., the reduction in
concentration with time is constant and independent of concentration) with rates
that differ primarily by the distance that a particle must traverse to return to the
trachea.
6.50
-------
Based on studies of particle clearance reported by Raabe (1984), muco-ciliary
transport in the nose and throat generally exhibits a half-life for clearance of
4 minutes. Clearance of the tracheo-bronchial section of the respiratory tract is a
function of the distance from the trachea and generally varies from a half-life of
30 minutes for the largest bronchi to approximately 5 hours for the smallest and
most remote bronchi. In healthy humans, material deposited in this region is
generally cleared within 24 hours. In contrast, the clearance mechanisms
operating in the deep lung, beyond the muco-ciliary escalator, operate over time
frames of many days to years (see Table 6-2);
• clearance of insoluble particles from the pulmonary portion of the lung occurs
primarily by macrophage transport and such transport has several components.
One component represents the population of "free" macrophages located within
the alveoli that engulf particles and transports them to the muco-ciliary escalator.
Macrophages are also renewed at some rate of recruitment that may be dependent
on particle concentrations. In fact, numerous studies have demonstrated that
macrophage recruitment is induced by the deposition of asbestos and other
particles in the lung (Section 6.3.5). Particles also migrate into the interstitium
where another population of macrophages clears these particles to lymph. This
second component (interstitial clearance) is much slower than the first (see
Table 6-2);
• each macrophage can carry a maximum load and the mobility of each macrophage
decreases with increasing load. At sufficient loading, macrophages become
immobile and aggregates of overloaded macrophages in the alveoli may then
sequester particles for some period of time as this clearance mechanism is shut
down. In the interstitium, masses of immobile macrophages may trigger
development of granulomas that sequester particles for extended periods of time
by effectively preventing clearance of the particles within such tissue, at least
until or unless the granulomas resolve. Thus, these models incorporate overload
mechanisms and the incorporation of such overload mechanisms are required to
explain observed trends in experimental results; and
• in various published studies, overload (immobilization of laden macrophages) has
been modeled as dependent on the total volume or mass of phagocytized material
(for compact particles) and (additionally) on the length of phagocytized material
(for fibers). It is also possible that the motility of macrophages and the
consequent overall rate of this clearance process is additionally a function of fiber
diameter and/or particle toxicity (the latter for special cases).
Interestingly, while it is reported that large, spherical particles are not readily cleared by this
mechanism, the range of sizes over which clearance becomes hindered corresponds reasonably
well to the limits of overall respirability. In contrast, fibers that are clearly too long to be cleared
by macrophages, if they are sufficiently thin, are quite respirable. Thus fibrous materials present
a unique challenge to the respiratory tract based solely on the dimensions of these materials.
6.51
-------
Stober et al. (1993) also notes that many models incorporate the assumption that most clearance
processes are first order (i.e., that the rate of reduction of mass or fiber number is proportional to
the remaining mass or fiber number, respectively, and independent of other factors). Thus, the
combined effects of multiple clearance processes can be expressed as a weighted sum of
exponentials and this approach has been fairly successful at mimicking actual processes. This
means, however, that the half-lives "t1/2 's" attributed to the various first order decays are
empirical and do not necessarily correspond to any specific physiological or biochemical
features of the processes being modeled. Depending on the specific process, clearance rates may
be zero order or may be a more complicated function of multiple variables than can be described
by a first order decay. Nevertheless, models incorporating these simplifying assumptions have
shown good success at adequately describing observed effects.
Note that half-lives for first order decay processes represent the time required for half of the
initial mass to decay (or be transported or whatever) and can be estimated as: ti/2=(ln2)/k with k
being the first order rate coefficient or proportionality constant between rate and mass. This is
why so many of the retention studies cited above provide estimates of a series of decay constants
or half-lives that are assumed to correspond approximately to the major clearance processes
contributing to the observed, overall reduction in concentration.
Due to the complexity of the processes involved, only a small number of dynamic models for
fiber retention have been developed. Interpretation of the results of these models requires that
the meaning of the term "retention" first be reconciled across studies.
Dement and Harris (1979) report that, based on a mathematical model, the fraction of structures
retained in the deep lung is unlikely to vary by more than a factor of 2 for different asbestos
mineral types. In this study, however, the term retention appears to refer primarily to a very
short time period that primarily includes consideration of deposition, but not clearance processes.
Using a definition for retention that reflects long-term residence in the lung, Yu et al. (1990)
developed a model of chrysotile retention that explicitly incorporates longitudinal splitting,
dissolution, and size-dependent clearance. Time-dependent lung burden estimates derived using
the model were shown to compare reasonably well with published data (Abraham et al. 1988, as
cited by Yu et al. 1990) both in terms of fiber concentrations and fiber size-distributions.
In a later modification of their retention model for chrysotile, Yu et al. (1991) also considered
the effect of airway asymmetry on fiber retention. In this version of the model, Yu and
coworkers incorporated information concerning the geometry of the bronchio-alveolar tree
(including mean distance and the mean number of airway bifurcations between the trachea and
the alveoli in each section of the lung) and studied the effects of such considerations. The
modified model predicts a non-uniform distribution of the asbestos that is retained in the lung
and the predictions reasonably reproduce the distributions observed by various researchers and
measured formally by Pinkerton et al. (as cited by Yu et al. 1991).
Yu and Asgharian (1990) also modeled the long-term retention of amosite in rat lungs. In
contrast to the models employed for chrysotile, the model presented for amosite incorporates a
term for the clearance rate that is not a constant but, rather, is a function of the lung
concentration of asbestos, which was adapted from an earlier model for diesel soot. This
6.52
-------
modification was incorporated to adequately mimic the suppression of clearance with increasing
lung burden that has been observed by several research groups (e.g., Davis et al. 1978; Wagner
et al. 1974) for amphiboles. Conditions under which elevated asbestos (or dust) concentrations
are observed to reduce clearance are referred to as "overload" conditions. Model predictions
were shown to reasonably reproduce the time-dependence of amosite lung burdens (in terms of
mass) in several studies.
Importantly, the overload conditions addressed in the Yu and Asgharian (1990) model were
primarily observed among older retention studies where lung burden was tracked as total
asbestos mass (Section 6.2.1). Such studies tend to suggest a difference in the behavior of
chrysotile and the amphiboles. As indicated in Section 6.2.1, however, later retention studies,
which track lung burden as a function of fiber number (in specific size categories) tend to show
this effect is a function of fiber length more than fiber type and newer models may need to
incorporate such factors that indicate reduced macrophage motility as a function of fiber length.
Moreover, it is important to consider the major, confounding effects, if the goal is to develop a
model that not only reproduces the time-dependence of clearance, but also captures relevant
physical phenomena. Thus, for example, Yu et al. (1994) were able to reproduce the time-
dependence of the retention in rats of inhaled RCF-1 (as a function of fiber size) using a model
in which macrophage motility was limited only by total lung burden and not dependent on fiber
size. However, these authors also failed to consider that long RCF-1 fibers in fact dissolve at
rates competitive with the clearance of short fibers (see Section 6.2.1), which is probably why
they did not find a dependence on length; the two effects cancelled out.
6.2.6 General Conclusions Regarding Deposition, Translocation, and Clearance
The current literature on deposition, translocation, and clearance paint a consistent picture of the
fate of fibrous structures in the lung. The ultimate fate of biodurable fibers depends
overwhelmingly on their size. Although there may be additional effects due to mineralogy
(addressed further in Section 6.3) and for rare, special cases this may be important, generally
such effects appear to be minor.
The primary effect attributable to mineralogy that is important to consider in relation to
clearance is that associated with biodurability. Fibers that dissolve in the lung at rates that are
competitive with the other clearance mechanisms described below may be cleared sufficiently
rapidly to preclude adverse effects, even when such fibers are too long to be cleared efficiently
by macrophages (see Section 6.2.1).
A schematic representation of the complex set of mechanisms that contribute to the translocation
and clearance of fibrous structures that have been deposited in the deep lung is presented in
Figure 6-4. This description was developed based on the complete spectrum of observations
reported in each of the previous sections of this chapter including, primarily, the descriptions of
the most sophisticated of the models reviewed by Stober et al. (1993).
6.53
-------
FIGURE 6-4:
PUTATIVE MECHANISMS FOR CLEARANCE AND TRANSLOCATION
Deposition
Alveolar
Lumen
• tO
muco-clllary
escalator
K8 = f(?7?'
f(slze)
f(size)
Al'
acrophaj
Type 1 Epithelium
. K8 = f(?77)
Interstitium
= f(size, cone, toxicity)
Immobile Macrophage
Granuloma
to Lymph
K11 = f'slze.conc. loxiclly)
Capillary
Lumen
isothelium
P)eura,V16o^ze'™nc'toxicity)
Drainage
Figure 6-4. Key for Putative Mechanisms for Clearance and Translocation of Fibers in the
Lung
K, = rate constant for phagocytosis of fibers by alveolar macrophages. This mechanism is an
inverse function of fiber length and, likely, diameter. This mechanism is likely pseudo
first order (assuming sufficient numbers of macrophages, the rate will be proportional to
the number of fibers);
K2 = rate constant for phagocytosis of fibers by Type I epithelial cells. This mechanism is an
inverse function of fiber length and, likely, diameter. This mechanism is likely pseudo
first order;
K3 = rate constant for diffusion of fibers from the alveolar air space (lumen) through the
epithelial lining to the underlying interstitium. This mechanism is diffusion limited and
likely independent fiber size or type. This mechanism likely parallels the behavior of
diffusion in through a finite, column of fixed diameter.
K4 = rate constant for phagocytosis of fibers by Type II epithelial cells. This mechanism is an
inverse function of fiber length and, likely, diameter. This mechanism is likely pseudo
first order;
6.54
-------
Figure 6-4 Key for Putative Mechanisms for Clearance and Translocation of Fibers in the
Lung (continued)
K5 = rate constant for transport by macrophage to the muco-ciliary escalator. This mechanism
is likely an inverse function of fiber length and the volume (or mass) of the fibers
phagocytized. Macrophages that become immobilized tend to aggregate in alveolar
lumena. For macrophages with fixed loads, this mechanism may be first order or zero
order.
K6 = rate constant for putative discharge to interstitium of phagocytized fibers by Type I
epithelial cells. There has been no direct verification of this mechanism;
K7 = rate constant for phagocytosis of fibers by endothelium. This mechanism is an inverse
function of fiber length and, likely, diameter. This mechanism is likely pseudo first
order;
K8 = rate constant for putative mechanism in which fibers internalized by macrophages are
transported through the epithelial lining of the alveolar space to the underlying
interstitium. There has been no direct verification of this mechanism;
K9 = rate constant for diffusion of fibers from the interstitium through the endothelial lining to
the enclosed, capillary lumen. This mechanism is diffusion limited and
likely independent fiber size or type. There has been no independent verification of this
mechanism;
K10 = rate constant for phagocytosis of fibers by interstitial macrophages. This mechanism is
an inverse function of fiber length and, likely, diameter. This mechanism is likely pseudo
first order (assuming sufficient numbers of macrophages, the rate will be proportional to
the number of fibers);
K,, = rate constant for transport by macrophage from the interstitium to the lymphatic system.
This mechanism is likely an inverse function of fiber length and the volume (or mass) of
the fibers phagocytized. Macrophages that become immobilized tend to induce
formation of interstitial granuloma. For macrophages with fixed loads, this mechanism
may be first order or zero order.
K12 = rate constant for putative discharge to capillary lumena of phagocytized fibers by
endothelial cells. There has been no direct verification of this mechanism;
K,3 = rate constant for phagocytosis of fibers by mesothelial cells of fibers transported to the
mesothelium. This mechanism is an inverse function of fiber length and, likely,
diameter. This mechanism is likely pseudo first order;
6.55
-------
Figure 6-4 Key for Putative Mechanisms for Clearance and Translocation of Fibers in the
Lung (continued)
K14 = rate constant for putative mechanism in which fibers internalized by macrophages are
transported from the interstitium through intervening tissue and the mesothelium to the
pleural space. There has been no direct verification of this mechanism;
K15 = rate constant for phagocytosis of fibers by pleural macrophages. This mechanism is an
inverse function of fiber length and, likely, diameter. This mechanism is likely pseudo
first order (assuming sufficient numbers of macrophages, the rate will be proportional to
the number of fibers);
K16 = rate constant for transport by macrophage to sites of lymphatic drainage (lymphatic
ducts) along the pleura. This mechanism is likely an inverse function of fiber length and
the volume (or mass) of the fibers phagocytized. For macrophages with fixed loads, this
mechanism may be first order or zero order.
Not shown: apparently fibers that are too large to pass through lymphatic ducts attract
accumulation of macrophages at sites of lymphatic drainage (Kane and MacDonald
(1993).
R, = rate constant for the renewal of the alveolar macrophage population. While the there is
likely a steady state rate for background renewal, given that the average life of an
alveolar macrophage is reported to be on the order of 7 days, this rate is also stimulated
in response to insult by foreign substances in the lung;
R2 = rate constant for the renewal of the interstitial macrophage population. While the there is
likely a steady state rate for background renewal, this rate is also expected to be
stimulated in response to insult by foreign substances in the interstitial space;
R3 = rate constant for the renewal of the pleural macrophage population. While the there is
likely a steady state rate for background renewal, this rate is also expected to be
stimulated in response to insult by foreign substances in the pleura;
6.56
-------
As shown in Figure 6-4, briefly, the first reaction to the introduction of fibers (or other
particulate matter) into the alveolar lumen is scavenging by alveolar macrophages. It has been
reported that the initial uptake by macrophages is a rapid process that is essentially complete
within hours after initial deposition. Rates for several of the mechanisms depicted in Figure 6-4
have been estimated in the literature and are summarized in Table 6-2.
The rate of removal (to the muco-ciliary escalator) by alveolar macrophages is then determined
by a variety of effects. Macrophage motility is a size dependent-process so that only fibers that
are sufficiently compact (<~20 ^m) can be removed from the lung. The rate of removal by this
process may also be suppressed both for fibers of intermediate lengths (10-20 ^m, which are
short enough to be phagocytized, but long enough to suppress macrophage motility) and by the
overall mass/volume of particles deposited (and, proportionally, taken up by each macrophage).
Note that the dimensions provided are the ones that are apparently appropriate for humans. For
rats, the corresponding dimensions may be somewhat smaller.
Likely competing with scavenging by macrophages are (1) phagocytosis by the epithelial cells
lining the alveolus and (2) diffusive transport to the intersititium. Both Type I and Type II
epithelial cells appear to phagocytize fibers. Although relatively few fibers are observed to be
taken up by Type II cells, as previously discussed (Section 6.2.2), one possible explanation for
the limited observation of fibers in Type II cells is that uptake of fibers induces terminal
differentiation to Type I cells. It is expected that phagyctosis by epithelial cells is a size-
dependent process.
Especially when the presence of fibers (or other particulate matter) induces morphological
changes in Type II cells that increase the overall permeability of the epithelial lining (Section
6.3.7), fibers can apparently diffuse into the interstitium. This process, potentially supplemented
with expulsion of phagocytized fibers by epithelial cells and/or transport of fiber laden
macrophages through the epithelial lining, represents the set of putative mechanisms by which
fibers may reach the interstitium. It is expected that these processes are somewhat slower than
uptake by alveolar macrophages (see Table 6-2).
Therefore, if this latter mechanism is operating at peak efficiency, relatively few fibers may
reach the interstitium. Note that diffusive transport is likely independent of fiber length, but may
be dependent on fiber width with thinner fibers more rapidly diffusing to the interstitium.
Fibers reaching the interstitium are likely cleared primarily by interstitial macrophages, which
phagocytize the fibers and transport them to the lymphatic system. Both the efficiency of
phagocytosis and motility of macrophage transport in the interstitium likely depend on fiber size
and the total volume/mass of fibers in the same way described above for transport by alveolar
macrophages. However, all such mechanisms are substantially slower in the interstitium than in
the alveolar lumen (see Table 6-2).
Macrophages that have been immobilized (due to fiber size or volume/mass) in the interstitium
tend to aggregate and induce formation of granulomas, which may sequester the fibers in these
cells. Although there is less evidence for this, fibers free in the interstitial matrix might also
trigger such a process. Such fibers would typically be too large to have been effectively
phagocytized by any of the cells of the interstitium.
6.57
-------
Fibers may also reach the endothelium and be taken up by endothelial cells lining the capillaries
of the deep lung. Because fibers have also been observed in capillary lumena, mechanisms
similar to those described for transport through the alveolar epithelium to the interstitium may be
operating to transport fibers into capillary lumena. While it is expected that such mechanisms
will also show size dependence similar to that previously described, little is known about the
details or the rates of such processes.
Also by mechanisms similar to some or all of the putative mechanisms described for
translocation of fibers from the alveolar lumen to the interstitium, fibers may reach the pleura.
Whether fibers can also reach the pleura via transport in blood or lymph has not been definitively
determined. Fibers reaching the pleura may be phagocytized by mesothelial cells or may pass
through such cells to the pleural cavity. Fibers reaching the pleural cavity are apparently
phagocytized by pleural macrophages (probably showing a similar size or volume/mass effect as
described above for similar mechanisms) and are apparently transported to (and deposited at)
sites of lymphatic drainage along the pleura. If such fibers are then too large to pass through the
lymphatic ducts, they may trigger accumulation of additional macrophages and other
inflammatory cells.
Overall, the effects of size appear to be:
• few fibers thicker than approximately 0.7 urn and virtually none thicker than
approximately 1.5 u.m appear to reach the deep lung. Of these, the longer the
fiber, the thinner the fiber. Importantly, the distribution of sizes of structures
deposited in the deep lung tend to be much more similar across studies than the
distributions in the aerosols originally inhaled. Thus, deposition is a very
effective filtering process; and
• short fibers (or compact particles) that are shorter than somewhere between 10
and 20 u.m tend to be taken up almost entirely by macrophages and are either
cleared via the muco-ciliary escalator, isolated in immobilized macrophages that
remain within alveolar lumena, or transported to lymphatics (presumably after
first reaching the interstitum). These processes appear to be efficient for the
shortest of the fibers in this range (and all shorter fibers) so that no further effects
are manifest. In contrast, longer fibers, which are not efficiently cleared or
isolated by macrophages either in the alveolar lumen or the interstitium appear to
trigger a range of additional responses, some of which appear to lead to disease.
Therefore, based on deposition, translocation, and clearance, it is fibers thinner than
approximately 0.7 um and longer than a minimum of approximately 10 urn (with relative
contributions increasing with increasing length up to at least 20 |im) that likely contribute to
disease. Modifications to this range of structure sizes due to effects attributable to direct
biological responses are addressed further in Section 6.3.
Regarding putative differences in behavior between chrysotile asbestos and amphibole asbestos,
such effects are adequately addressed by the unifying discussion provided above. To make sense
of such differences, the effects of fiber size must first be explicitly considered. Thus, chrysotile
fibers may not be as efficiently deposited in the deep lung to the extent that they are curlier or
6.58
-------
occur in thicker bundles than amphibole fibers. The overall load of chrysotile deposited in the
deep lung may also be cleared more rapidly than amphiboles to the extent that (1) short, thin
fibrils ultimately represent a greater fraction of the total load of chrysotile than the amphiboles
and (2) a subset of long chrysotile fibers, not sequestered in an environment with limited fluid
flow, may be cleared more rapidly than similarly long amphibole fibers due to contributions
from dissolution.
Importantly, the concentrations of asbestos to which humans are exposed are much lower than
the concentrations to which animals were exposed in the various literature studies cited.
Moreover, as indicated in Section 6.1.2, for virtually any exposure of interest, the resulting
(volumetric or surface area) lung burden will be substantially higher in rats than in humans.
Therefore, overload conditions or other processes that might impede or alter the clearance
mechanisms described above, will never occur in humans without first having affected the
results of the animal studies reported. Thus, conclusions concerning size-dependence and related
effects (except to the extent that they need to be adjusted for cross-species differences) should
remain valid when extrapolated between animals and humans.
6.3 FACTORS GOVERNING CELLULAR AND TISSUE RESPONSE
For inhaled structures that survive degradation and clearance, a series of complex reactions
between the structures retained in the lung and surrounding tissue may induce a biological
response. Asbestosis (fibrosis), pulmonary carcinomas, or mesotheliomas may result.
Mesotheliomas are likely associated with structures that are translocated from the lung to the
mesenchyme, although diffusable growth promoters and other chemical signals produced by
asbestos exposed cells in lung tissue immediately proximal to the mesothelium may also play a
role (see Adamson 1997, as described in Section 6.3.4.1).
That the specific biochemical triggers for asbestos-related diseases have not been definitively
delineated as of yet is not surprising. Despite great progress in elucidating candidate
mechanisms, the number of candidate mechanisms is large and confounded by "cross-talk"
between mechanisms. Moreover, similar toxic endpoints may result from entirely independent
mechanisms that exhibit disparate dose-response characteristics but, nevertheless, may be
triggered by the same or similar toxins. In such cases, however, the relative contributions from
each mechanism to a particular endpoint may vary substantially. Unfortunately, the ability to
compare results across studies of different mechanisms is currently limited due to the inability to
reconcile the quantitative effects of dose and response across dissimilar studies.
Table 6-5 illustrates the range and complexity of the biological responses triggered by asbestos
in lung tissue. The table was developed based on the information gleaned from the studies
described in this section. Importantly, not all of the mechanisms listed contribute equally to the
toxic endpoints that are attributable to asbestos, but their relative importance has yet to be
entirely delineated. The toxic endpoints of potential interest to which each of the listed
mechanisms potentially contribute are indicated in bold italics.
6.59
-------
Table 6-5. Putative Mechanisms by Which Asbestos May Interact with Lung Tissue to
Induce Disease Following Inhalation
Asbestos (in vivo)
Generates reactive oxygen species (ROS)
• may affect neighboring cells and tissues
Interacts with macrophages
Interacts with lung epithelium
Asbestos interacts with macrophages
Induces generation and release of reactive oxygen species/reactive nitrogen species (ROS/RNS)
• may affect neighboring cells and tissues
• may induce inflammatory response (which may promote cancer)
• may induce fibrogenesis (which may promote cancer)
• induces signaling cascades in macrophages and neighboring tissues
- mediates apoptosis (which may regulate cancer)
- mediates proliferation (which may promote cancer)
• causes genotoxic effects in neighboring tissues
- may cause cancer initiation
- may induce arrest of cell cycle
- induces signaling cascades
• mediates apoptosis (which may regulate cancer)
• mediates proliferation (which may promote cancer)
• depletes reserves of antioxidants in macrophages and neighboring tissues
- may induce cytotoxicity (which may promote cancer, by inducing proliferation)
- may increase susceptibility to insult by other toxic agents (which may promote cancer)
Induces release of various cytokines
• affects neighboring cells and tissues
• mediates inflammatory response (which may promote cancer)
• mediates fibrogenesis (which may promote cancer)
• induces signaling cascades in macrophages and neighboring tissues
6.60
-------
Table 6-5. Putative Mechanisms by Which Asbestos May Interact with Lung Tissue to
Induce Disease Following Inhalation (continued)
- mediates apoptosis (which may regulate cancer)
- mediates proliferation (which may promote cancer)
• mediates proliferation in neighboring tissues (which may promote cancer)
Induces signaling cascades and mediates apoptosis in macrophages (which may regulate cancer)
At high enough concentrations, promotes cytotoxic cell death (which may promote cancer by
inducing proliferation)
Asbestos interacts with Lung Epithelium
Induces release of reactive oxygen species/reactive nitrogen species (ROS/RNS)
• may affect neighboring cells and tissues
• may induce inflammatory response (which may promote cancer)
• may induce fibrogenesis (which may promote cancer)
• induces signaling cascades in epithelium and neighboring tissues
- mediates apoptosis (which may regulate cancer)
- mediates proliferation (which may promote cancer)
• causes genotoxic effects in epithelium and neighboring tissues
- may cause cancer initiation
- may induce arrest of cell cycle
- induces signaling cascades
• mediates apoptosis (which may regulate cancer)
• mediates proliferation (which may promote cancer)
• depletes reserves of antioxidants in epithelium and neighboring tissues
- may induce cytotoxicity (which may promote cancer by inducing proliferation)
- may increase susceptibility to insult by other toxic agents (which may promote cancer)
Induces release of various cytokines
• affects neighboring cells and tissues
• mediates inflammatory response (which may promote cancer)
• mediates fibrogenesis (which may promote cancer)
• induces signaling cascades in epithelium and neighboring tissues
- mediates apoptosis (which may regulate cancer)
6.61
-------
Table 6-5. Putative Mechanisms by Which Asbestos May Interact with Lung Tissue to
Induce Disease Following Inhalation (continued)
- mediates proliferation (which may promote cancer)
• mediates proliferation in epithelium and neighboring tissues (which may promote cancer)
Causes genotoxic effects
• may cause cancer initiation
• may induce arrest of cell cycle
• induces signaling cascades
- mediates apoptosis (which may regulate cancer)
- mediates proliferation (which may promote cancer)
Induces signaling cascades
• mediates apoptosis (which may regulate cancer)
• mediates proliferation (which may promote cancer)
Increases epithelial permeability
• encourages fibrosis (which may promote cancer)
• facilitates translocation
At high enough concentrations, promotes cytotoxic cell death (which may promote cancer by
inducing proliferation)
6.62
-------
As indicated previously (Section 6.0), although much has been learned about specific
components of the underlying mechanisms by which asbestos causes the above-listed diseases,
substantial knowledge gaps remain. Moreover, because of these gaps, multiple candidate effects
have been explored as potential contributors to carcinogenicity (or fibrogenicity) and one of the
goals of this document is to distinguish among those effects that are likely to contribute to the
induction of cancer from those that are less likely or unlikely to contribute (given the current
state of knowledge). Accordingly, an overview of recent studies is presented below following a
brief description of the current model of the general mechanism for cancer. Note that, due to the
availability of several recent reviews (including, for example, Economou et al. 1994; Floyd
1990; Kamp et al. 1992; Kane and MacDonald 1993; Mossman and Churg 1998; Mossman et al.
1996; Oberdorster 1994; Robledo and Mossman 1999), only the most recent primary articles are
included in the following review.
Also, as indicated below, many of the biological responses provoked by retained asbestos are in
fact dependent on fiber size and type. Therefore, studies that distinguish among effects induced
by different size fibers (or fibers and non-fibrous particles) of the same mineralogy and studies
that distinguish among effects induced by comparably sized fibers (or non-fibrous particles), but
differing mineralogy are highlighted. However, due to the limits with which fibrous materials
have tended to be characterized in many of these studies, the database from which to distinguish
among the effects of size and mineralogy are limited and conclusions from differing studies must
be compared with caution.
A large body of evidence supports the conclusion that it is primarily (if not exclusively) long
fibers (those longer than a minimum of 5-10 ^m) that contribute to disease (see Sections 6.2.1,
6.2.2, and 6.4). Much evidence also indicates that the potency of long fibers increases with
length at least up to a length of approximately 20 urn. Therefore, because short and long fibers
are both respirable (for fibers that are sufficiently thin, less than approximately 0.7 u.m in
diameter), differences in the ultimate response to short and long fibers must be attributable to
differences in tissue and cellular responses to the retained fibers in each size range.
At least at a histopathological level, clear differences have been observed in the responses
evoked by short and long structures (see Sections 6.2.1 and 6.2.2). It is a goal of this section to
determine whether the biochemical triggers that mediate the disparate responses to short and
long fibers can also be identified. Unfortunately, while a large body of knowledge has been
amassed, definitive conclusions are not yet possible. This is because the specific mechanisms by
which asbestos acts have still not been definitively determined, although many candidate
mechanisms have been elucidated (see above). However, important inferences can still be
gleaned from the available studies.
Evidence for the relationship between fiber diameter and disease is somewhat less clear.
Although there appears to be a fairly sharp cutoff in the diameter of fibers that are respirable (see
Section 6.1.4), several studies suggest that the most potent fibers are substantially thinner than
the sharp cutoff in respirability (see Section 6.4). If these latter observations are valid then, as
with length, differences in the ultimate response to thick and thin fibers must be attributable to
differences in the tissue and cellular responses elicited by retained fibers of each width. As
indicated with length, however, definitive conclusions regarding biochemical triggers and the
effect of size on such triggers are not yet possible because the specific triggers that lead to
6.63
-------
specific asbestos-related diseases have not been definitively identified. Still, useful inferences
can be developed.
6.3.1 The Current Cancer Model
The following description of the current model for cancer is derived from the ideas presented in
Moolgavkar et al. 1988, Mossman 1993, and Economou et al. 1994.
In the current model of cancer, normal cells must accumulate specific, multiple mutations before
a tumorigenic cell is created that can lead to the development of cancer. Each successive
mutation produces an initiated cell (a cell that is transformed from normal cells because it
incorporates one or more of the requisite mutations, but that has not yet acquired all of the
changes needed to produce cancer). Each initiated cell may then proliferate to generate a
population of similarly initiated cells, which increases the probability that other events will lead
to further mutation in at least one of these cells. Individual mutations may occur spontaneously
or may be induced by exposure to mutagens.
Generally, the minimum, heritable changes required before a normal cell is transformed into a
metastatic tumor cell include, but may not be limited to (see, for example, Hei et al. 1997 or
Kravchenko et al. 1998):
• escape from terminal differentiation or programmed cell death (especially in
response to DNA damage);
• escape from anchorage/neighbor dependent growth inhibition;
• development of self-promoting growth and proliferation; and
• active expression of cytokines needed to promote angiogenesis and allow tissue
invasion.
It is not clear whether the mutations associated with these changes need to occur in a particular
order, although the first of the above-listed changes would facilitate accumulation of all later
changes. It is also unclear which of the above changes contribute to the time dependence of
tumor development and this may vary among tumor types. It is likely that only a subset of the
required mutations determine the ultimate time development of the associated cancer. For
example, once a cell begins self-promoting growth, later mutations (even relatively rare ones)
may become incorporated rather quickly. Also, some mutations may be rare or may require
intervention by a toxic agent, while other mutations may occur spontaneously and may thus
occur frequently, once a sufficient number of initiated precursor cells are generated. This is one
of the reasons that models with as few as one or two stages have proven successful at predicting
the time course of many types of cancer (see, for example, Moolgavkar et al. 1988).
6.64
-------
To produce cancer, the mutations that occur must also be "heritable" meaning that they must be
preserved and passed on to daughter cells during mitosis. Thus, it is not only necessary to cause
alteration in the DNA of a cell (genetic damage), but the mechanisms by which the cell
subsequently repairs such changes or prevents cell division (e.g., arrest of the cell cycle,
programmed cell death, terminal differentiation) in the presence of such changes must also be
defeated. Generally, if a cell proceeds through DNA synthesis (in preparation for mitosis) before
accumulated alterations to DNA are repaired, the sites of such alterations can lead to errors in
replication during synthesis, which in turn result in permanent, heritable mutations in one or both
of the daughter cells that are created from mitosis.
Toxic agents that (directly or indirectly) cause DNA alterations may contribute to the
development of cancer by inducing one or more of the set of requisite mutations required for
cancer to develop. In traditional parlance, such agents are termed "initiators". In addition, any
toxic agent that enhances proliferation also facilitates the development of cancer both by
increasing the probability of creating spontaneous mutations (due to errors that infrequently, but
unavoidably, occur whenever DNA is synthesized in support of mitosis) and by increasing the
numbers of any initiated cells that may be present, which may then serve as additional targets for
initiators or may incorporate additional spontaneous mutations. Agents that facilitate the
development of cancer by inducing proliferation are traditionally termed "promoters."
Multiple mechanisms have been identified by which both initiators and promoters may act.
Initiators, for example, may react directly with DNA to cause genetic damage, or may induce
generation of other reactive species (such as reactive oxygen species or reactive nitrogen
species) that, in turn, react with DNA to cause genetic damage. In addition, fibrous materials
may uniquely damage chromosomes by interfering with mitosis causing aneuploidy
(incorporation of an incorrect number of chromosome copies in cells) and/or various clastogenic
changes (alterations in the organization and structure of specific chromosomes).
Promoters may also induce proliferation via a variety of mechanisms. Cytotoxic agents, for
example, may induce proliferation in a tissue by damaging or killing cells and thereby induce
stem cells to proliferate to replace the damaged cells. Promoters may also induce release of
various growth factors that, in turn, induce proliferation in targeted tissues. This may occur, for
example, as part of the inflammatory response to tissue insult.
Promoters that are biopersistent (such as long asbestos fibers) or promoters that are continually
reintroduced (through chronic, external exposure) may also chronically up regulate (or down
regulate) certain cell signaling cascades that may contribute to cancer development in a variety
of ways including: (1) activation of genes that mediate proliferation or production of various
growth factors (including any of various oncogenes) or (2) suppression of genes that inhibit
proliferation or growth (including any of various tumor-suppressor genes).
There is growing evidence that all varieties of asbestos fibers (and certain other fibrous
materials) can act both as cancer initiators and promoters. However, the biological responses to
these materials appear to vary in different tissues so that it may be important to separately
evaluate the behavior of asbestos in specific tissues. Biological responses to varying fiber types
also appear to vary, particularly in relation to a fiber's biopersistence (Sections 6.2.1 and 6.2.4).
Accordingly, an overview of the generic evidence that specific types of asbestos may act as an
6.65
-------
initiator and, separately, as a promoter is reviewed below, followed by consideration of the more
limited data suggesting tissue-specific variation in biological responses.
6.3.2 Evidence for Transformation
Several recent studies provide evidence that specific types of asbestos can induce transformation
of cells in specific tissues (both lung epithelium and mesothelium) to create tumor cells. This
provides further, confirmatory support for the whole animal studies in which cancer is induced
by exposure to asbestos (see Sections 6.4).
An immortalized, but non-tumorigenic, cell line of human bronchial epithelial cells (BEP2D
cells) was transformed by a single exposure to 4 ug/cm2 UICC chrysotile B (Rhodesian).
Surviving cells (the treatment caused 18% cell death) went through several transformations
including: altered growth kinetics, resistence to serum induced terminal differentiation, and loss
of anchorage dependent growth, before becoming tumorigenic (Hei et al. 1997). Tumorigenicity
developed in the various exposed cell lines over a period of several to 11 weeks following
exposure. When injected into nude mice, secondary tumors developed with a latency of 8-10
weeks.
The authors indicate that there were no mutations in these cells at either codon 12 or 13 or the
ras gene (mutations that have sometimes been observed in asbestos-induced lung cancers. Also,
because this cell line already contains alterations in genes for p53 (a protein that plays a role in
cell-cycle control, among other things, see Table 6-6) and Rb (Table 6-6), the authors speculate
that such changes are not rate controlling for transformation to cancer.
It should also be noted that cultures of Type II cells have been particularly difficult to maintain
due to the tendency of these cells to undergo terminal differentiation (to Type I cells) once they
are removed from their natural environment in the epithelial lining of lung alveoli (see Leikoff
and Driscoll 1993, as described in Section 4.4).
Kravchenko et al. (1998) indicates that, unlike cultures of lung epithelial cells, in vitro cultures
of rat mesothelial cells tend to transform spontaneously to tumorigenic cells. The major changes
that occur with time include: altered response to epithelial growth factor (from growth-
proliferation inhibition by this factor to growth-proliferation stimulation), morphological
changes (from polygonal epithelial-like cells to elongated fibroblast-like cells and, eventually,
polynucleated cells exhibiting broad polymorphism), and multi-layered growth and an ability to
grow as colonies and masses in semi-solid agar. Eventually, these cells become immortal and
induce cancer when harvested and injected into other rats. The authors indicate that asbestos-
induced transformations in such cultures proceed through identical stages, but that they occur
much more rapidly. For example, incorporation of the stimulating response to EOF occurs
spontaneously at passages 9 or 10, but at passages 6 or 7 when induced by asbestos. Asbestos
was applied at 5 u.g/cm2, which was noted to be sub-lethal (95% cell survival was noted at this
rate of application).
6.66
-------
Table 6-6a. Sources of Various Cytokines and Other Chemical Transmitters
Chemical Species
Abbrev.
Sources
Reference
Effects
Reference
Signal Transmitters
Activator Protein-1 AP-1
Fifth Component
of Compliment
Cyclin Dependent
Kinases
Cyclin Dependent
Kinase Inhibitors
Epithelial Growth
Factor
Epithelial Growth
Factor Receptor
BAX
Bcl-2
C5
CDK's
GDI's
CINC
EOF
EGFR
Induced by high
concentrations of p53
Lechneretal. 1997
A transcription factor
Binds the DNA promoter
region of genes governing
inflammation, proliferatin, and
apoptosis
Induces apoptosis
Inhibits apoptosis
Mediates asbestos-induced
fibrosis
Mediates cell cycle
Mossman and
Churg 1998 (Citing:
Angel and Karin
1991)
Lechneretal. 1997
Lechneretal. 1997
McGavran et al.
1989
Lechneretal. 1997
Inhibits advance of cell cycle Lechner et al. 1997
Inhibits proliferation of
mesothelial cells
Stimulates proliferation of
mesothelioma tumor cells
Mediates MAPK signaling
Kravchenko et al.
1998
Goodglick et al.
1997
Mossman et al.
1997
6.67
-------
Table 6-6a. Sources of Various Cytokines and Other Chemical Transmitters (continued)
Chemical Species
Abbrev.
Sources
Reference
Effects
Reference
EGFR regulated
kinase
ERK
Granulocyte- GM-CSF
macrophage colony
Stimulating Factor
GRP78
Human neutrophil HNE
elastase
Intercellular
adhesion
molecule-1
Interleukin-1
Interleukin-6
Interleukin-8
HSP72/73
ICAM-1
IL-1
IL-6
IL-8
Released from asbestos
activated PMN
Kampetal. 1993
Released by lung Nario and Hubbard
epithelial and 1997
endothelial cells
following exposure to
silica but not TiO2
Enhanced expression by Choe et al. 1997
RPM cells exposed to
chrys or crc
Induced by ROS in
Type II epithelial cells
Luster and
Simeonova 1998
Mediates apoptosis
Mossman et al.
1997
By itself, increases pulmonary Kamp et al. 1993
epithelial cell detachment in
culture
With asbestos, increases Kamp et al. 1993
pulmonary epithelial cell lysis
Facilitates migration of
leukocytes out of blood to
sites of injury
Nario and Hubbard
1997
Is a pleiotropic cytokine
Luster and
Simeonova 1998
Is a neutrophil chemoattractant Luster and
Simeonova 1998
6.68
-------
Table 6-6a. Sources of Various Cytokines and Other Chemical Transmitters (continued)
Chemical Species
Inducible Nitric
Oxid Synthase
Keritonizing
Growth Factor
Mitogen Activated
Protein Kinase
Macrophage
Inflammatory
Protein- 1 alpha
Macrophage
Inflammatory
Protein-2
Matrix Metallo-
proteinases
Abbrev. Sources Reference
iNOS Expressed by activated Jesch et al. 1997
macrophages, but not in
primates
KGF Derived from fibroblasts Adamson 1997
MAPK
MIP-1
alpha
MIP-2
MMP's Relative expression Morimoto et al.
affected by cigaret 1 997
Effects
Induces transient proliferation
of mesothelial cells
Mediates transcription of
AP-1
Mediates ERK signaling
A chemoattractant for PMN's,
but may not be involved with
inflammatory response
Reference
Adamson 1997
Mossman et al.
1997
Mossman et al.
1997
Osier etal. 1997
smoke and fiber
exposure
Neu
Nuclear Factor-KB NF-KB
A transcription factor that
regulates genes involved with
the inflammatory response,
cell adhesion, and growth
control.
Barchowsky et al.
1998
Nitric Oxide
Synthase
NOS
6.69
-------
Table 6-6a. Sources of Various Cytokines and Other Chemical Transmitters (continued)
Chemical Species
Abbrev.
Sources
Reference
Effects
Reference
Ornithine ode
Decarboxylase
Platelet-Derived PDGF
Growth Factor
Activated macrophages Bauman et al. 1990
Elevated expression in
mesothelioma cells
Asbestos and iron
treated AM and
Interstitial macrophages
release PDGF
Lechneretal. 1997
Osornio-Vargas
etal. 1993
Promoted by AP-1
Encodes a key enzyme for the
biosynthesis of polyamines
Induces proliferation of lung
fibroblasts
Induce proliferation of
mesothelial cells
Is a chemotactic attractant for
rat lung fibroblasts
Mossman and
Churg 1998
Mossman and
Churg 1998
Bauman etal. 1990
Brodyetal. 1997
Osornio-Vargas
etal. 1993
P53
Poly(ADP-
ribosyl)polymerase
Rb
Tissue inhibitors of TIMP's
MMP's
Induced by DNA
damage
Inhibited by
3-aminobenzamide
Relative expression
affected by cigaret
smoke and fiber
exposure
Lechner et al. 1997 At low cone, arrests cell cycle Unfried et al. 1997
at Gl/S
At higher cone, induces BAX
protein, which induces
apoptosis
Broaddus et al. 1997
6.70
-------
Table 6-6a. Sources of Various Cytokines and Other Chemical Transmitters (continued)
Chemical Species
Tumor Necrosis
Factor Alpha
Abbrev.
TNF-
alpha
Sources Reference Effects
Some particles induce Driscoll etal. 1997 Induces JNK arm of MAPK
activated macrophages cascade
to release this (shown
both in vitro and in
vivo)
Reference
Mossman et al.
1997
Transforming
Growth Factor
Alpha
TGF-
alpha
Regulated by oxidant-
sensitive transcription of
NF-KB
Macrophages in BAL
from asbestosis and
idiopathic interstitial
fibrosis patients release
TNFalpha
Expressed in BA
epithelium exposed to
asbestos
Driscoll etal. 1997
Zhang etal. 1993
Brodyetal. 1997
Is involved in the recruitment Driscoll et al. 1997
of inflammatory cells
Stimulates macrphages, Driscoll et al. 1997
epithelial cells,endothelial
cells, fibroblasts to secrete
chemokines and adhesion
molecules
Can induce apoptosis among Leigh et al. 1997
neutrophils
Can induce ROS production in Kaiglova et al. 1999
leukocytes
Intradermal inj stimulates Zhang et al. 1993
local accumulation of
fibroblasts and collagen
In vitro, stimulates fibroblast Zhang et al. 1993
DNA synthesis and
proliferation
Is a mitogen for epithelial cells Brody et al. 1997
6.71
-------
Table 6-6a. Sources of Various Cytokines and Other Chemical Transmitters (continued)
Chemical Species
Transforming
Growth Factor
Beta
Abbrev. Sources
TGF-beta Expressed in BA
epithelium exposed to
asbestos
Macrophage stimulated
Type II epi cells
produce both TGF-beta 1
and TGF-beta2
Reference
Brodyetal. 1997
Brodyetal. 1997
Effects
Inhibits fibroblast proliferation
but stimulates formation of
extracellular matrix
Can induce apoptosis in
different types of cells
Reference
Brodyetal. 1997
Leigh etal. 1997
Urokinase-type
Plasminogen
Activator
uPA
All three isoforms
expressed in the fibrotic
lesions (by hyperplatic
type II cells) of
asbestosis and pleural
fibrosis patients from
Quebec
Mesothelial tumor cells
expressed only TGF-
betal and 2
Also type II cells of
silicosis patients express
all three forms
Asbestos and iron
treated AM and
Interstitial macrophages
release TGF-beta
Jagiirdir et al. 1997
Jagiirdir et al. 1997
Jagiirdir et al. 1997
Osornio-Vargas
etal. 1993
Does not appear to induce
chemotaxis of rat lung
fibroblasts
Induces chemotaxis of rat
mononuclear leukocytes
Associated with increased
pericellular protyolytic
activity in endothelial tissue
Osornio-Vargas et
al. 1993
Osornio-Vargas et
al. 1993
Barchowsky et al.
1998
6.72
-------
Table 6-6a. Sources of Various Cytokines and Other Chemical Transmitters (continued)
Chemical Species
Abbrev.
Sources
Reference
Effects
Reference
Vascular Cell
Adhesion Molecule
VCAM-1
WT1
Enhanced expression by
RPM cells exposed to
chrys or crc
Choeetal. 1997
Forms a heterodimer w p53
and alters behavior
Unfriedetal. 1997
Enzymes
Mn-dependent
SuperOxide
Dismutase
Mn-SOD
Gene Products
c-fos
c-jun
mdm2
c-myc
Linked to apoptosis
Transcription factor that
activates the TGF-betal
promoter
Proteins of c-fos dimerize with
c-jun to create activator
protein-1 (AP-1)
Linked to proliferation
Transcription factor that
activates the TGF-betal
promoter
Transcription factor that
activates the TGF-betal
promoter
A protococoncogene that
inhibits p53
Timblinetal. 1998a
Jagirdar et al. 1997
Mossman and
Churg 1998
Timblinetal. 1998a
Jagirdar et al. 1997
Mossman and
Churg 1998
Lechneretal. 1997
6.73
-------
Table 6-6b. Effects of Various Cytokines and Other Chemical Transmitters
Chemical Species
Abbrev.
Effects
Reference
Signal Transmitters
Activator Protein-1
AP-1
Fifth Component of Compliment
Cyclin Dependent Kinases
BAX
Bcl-2
C5
CDK's
Cyclin Dependent Kinase Inhibitors GDI's
Cytokine-induced Neutrophil CINC
Chemoattractant
Epithelial Growth Factor
EGF
Epithelial Growth Factor Receptor EGFR
EGFR regulated kinase ERK
Granulocyte-macrophage colony GM-CSF
Stimulating Factor
Human neutrophil elastase
GRP78
HNE
A transcription factor
Binds the DNA promoter region of genes
governing inflammation, proliferatin, and
apoptosis
Induces apoptosis
Inhibits apoptosis
Mediates asbestos-induced fibrosis
Mediates cell cycle
Inhibits advance of cell cycle
Inhibits proliferation of mesothelial cells
Stimulates proliferation of mesothelioma tumor
cells
Mediates MAPK signaling
Mediates apoptosis
By itself, increases pulmonary epithelial cell
detachment in culture
Mossman and Churg 1998
(Citing: Angel and Karin 1991)
Lechneretal. 1997
Lechneretal. 1997
McGavran et al. 1989
Lechneretal. 1997
Lechneretal. 1997
Kravchenko et al. 1998
Goodglick et al. 1997
Mossman et al. 1997
Mossman etal. 1997
Kampetal. 1993
6.74
-------
Table 6-6b. Effects of Various Cytokines and Other Chemical Transmitters (continued)
Chemical Species
Abbrev.
Effects
Reference
Intercellular adhesion molecule-1
Interleukin-1
Interleukin-6
Interleukin-8
Inducible Nitric Oxid Synthase
Keritonizing Growth Factor
Macrophage Inflammatory
Protein-1 alpha
Macrophage Inflammatory
Protein-2
Matrix Metalloproteinases
Nuclear Factor-KB
Nitric Oxide Synthase
Ornithine Decarboxylase
HSP72/73
ICAM-1
IL-1
IL-6
IL-8
iNOS
KGF
Mitogen Activated Protein Kinase MAPK
MIP-1
alpha
MIP-2
MMP's
Neu
NF-KB
NOS
ode
With asbestos, increases pulmonary epithelial cell Kamp et al. 1993
lysis
Facilitates migration of leukocytes out of blood to Nario and Hubbard 1997
sites of injury
Is a pleiotropic cytokine
Is a neutrophil chemoattractant
Induces transient proliferation of mesothelial
cells
Mediates transcription of AP-1
Mediates ERK signaling
A transcription factor that regulates genes
involved with the inflammatory response, cell
adhesion, and growth control.
Promoted by AP-1
Luster and Simeonova 1998
Luster and Simeonova 1998
Adamson 1997
Mossmanetal. 1997
Mossmanetal. 1997
A chemoattractant for PMN's, but may not be Osier et al. 1997
involved with inflammatory response
Barchowsky et al. 1998
Mossman and Churg 1998
6.75
-------
Table 6-6b. Effects of Various Cytokines and Other Chemical Transmitters (continued)
Chemical Species
Abbrev.
Effects
Reference
Platelet-Derived Growth Factor
PDGF
P53
Poly(ADP-ribosyl)polymerase
Tissue inhibitors of MMP's
Tumor Necrosis Factor Alpha
TIMP's
TNF-alpha
Transforming Growth Factor Alpha TGF-alpha
Transforming Growth Factor Beta TGF-Beta
Encodes a key enzyme for the biosynthesis of
polyamines
Induces proliferation of lung fibroblasts
Induces proliferation of mesothelial cells
Is a chemotactic attractant for rat lung fibroblasts
At low concentration, arrests cell cycle at Gl/S
At higher concentration, induces BAX protein,
which induces apoptosis
Induces JNK arm of MAPK cascade
Is involved in the recruitment of inflammatory
cells
Stimulates macrphages, epithelial
cells,endothelial cells, fibroblasts to secrete
chemokines and adhesion molecules
Can induce apoptosis among neutrophils
Can induce ROS production in leukocytes
Intradermal inj stimulates focal accumulation of
fibroblasts and collagen
In vitro, stimulates fibroblast DNA synthesis and
proliferation
Is a mitogen for epithelial cells
Inhibits fibroblast proliferation but stimulates
formation of extracellular matrix
Mossman and Churg 1998
Baumanetal. 1990
Brodyetal. 1997
Osornio-Vargas et al. 1993
Unfriedetal. 1997
Mossman etal. 1997
Driscoll et al. 1997
Driscoll et al. 1997
Leigh etal. 1997
Kaiglova et al. 1999
Zhang etal. 1993
Zhang etal. 1993
Brodyetal. 1997
Brodyetal. 1997
6.76
-------
Table 6-6b. Effects of Various Cytokines and Other Chemical Transmitters (continued)
Chemical Species
Abbrev.
Effects
Reference
Urokinase-type Plasminogen
Activator
uPA
Can induce apoptosis in different types of cells Leigh et al. 1997
Does not appear to induce chemotaxis of rat lung Osornio-Vargas et al. 1993
fibroblasts
Induces chemotaxis of rat mononuclear Osornio-Vargas et al. 1993
leukocytes
Associated with increased pericellular protyolytic Barchowsky et al. 1998
activity in endothelial tissue
Vascular Cell Adhesion Molecule
VCAM-1
WT1 Forms a heterodimer w p53 and alters behavior
Untried etal. 1997
Enzymes
Mn-dependent SuperOxide
Dismutase
Mn-SOD
Gene Products
c-fos
Linked to apoptosis
Timblinetal. 1998a
Transcription factor that activates the TGF-betal
promotor
Proteins of c-fos dimerize with c-jun to create
activator protein-1 (AP-1)
c-jun Linked to proliferation
Transcription factor that activates the TGF-betal
promotor
Transcription factor that activates the TGF-betal
promotor
mdm2 A protococoncogene that inhibits p53
Jagirdar et al. 1997
Mossman and Churg 1998
Timblinetal. 1998a
Jagirdar et al. 1997
Mossman and Churg 1998
Lechneretal. 1997
6.77
-------
Table 6-6b. Effects of Various Cytokines and Other Chemical Transmitters (continued)
Chemical Species Abbrev. Effects Reference
c-myc
H-RAS
K-RAS
6.78
-------
In a study in which p53 deficient mice were intrapleurally injected with UICC crocidolite (200 (ig/week),
Marsella et al. (1997) showed that p53 deficient mice exhibited substantially increased susceptibility to
development of mesothelioma. In this study, 12.5% of homozygous mice (p53 deficient) developed
mesothelioma and the rest died of lymphomas or hemangiosarcomas that develop spontaneously in these
mice; 76% of heterozygous mice died of mesothelioma, and only 32% of wild-type mice (p53 competent)
died of mesothelioma. The authors suggest that p53 deficient mice are susceptible to excess proliferation
induced by crocidolite due to loss of control at the Gl/S check point that is normally mediated by p53.
In further confirmation of this hypothesis, Marsella et al. (1997) report in the same study that 7.5 ug/cm2
of crocidolite applied to wild murine mesothelial cells in culture induced substantial apoptosis while
p53-deficient cells were resistant to apoptosis. The authors also note that most of the p53-deficient cells
are tetraploid (suggesting a loss of a spindle check point) while the wild type cells are all diploid.
Note that, although these studies suggest that asbestos alone can induce complete transformation of both
lung epithelial cells and mesothelial cells, such studies must be evaluated with caution. In the case of the
Hei et al. (1997) study, for example, the effects of the (asbestos-independent) mutations required to
initially establish the immortal line of lung epithelial cells are not entirely known. Therefore, the
response to asbestos of epithelial cells in vivo may be substantially different.
In the case of the Kravchenko et al. (1998) study, it is clear that mesothelial cells in vitro do not behave in
the same manner as those in vivo; in vitro, they spontaneously transform to tumorigenic cells. This
suggests that one or more growth inhibitory signals exist in vivo (which are absent in vitro) that are
critical to maintaining the health of the mesothelium
6.3.3 Evidence that Asbestos Acts As a Cancer Initiator
As indicated in a review by Jaurand (1997), historically, the status of asbestos as a mutagen was
questioned, due primarily to the failure to produce detectable gene mutations in short-term assays.
However, more recent studies provide clear evidence that asbestos (and other fibrous materials) can
produce mutations. Moreover, fibrous materials may induce mutations by multiple mechanisms
including, for example:
• direct interference with mitosis;
• production and release of reactive oxygen species (ROS); or
• production and release of reactive nitrogen species (RNS).
Of these, the most consistent, positive evidence that asbestos can act as a cancer initiator (i.e., a genotoxin
or mutagen) has been from assays designed to detect the kinds of genetic damage that result from
interference with mitosis (Jaurand 1997). Although generation of ROS and RNS plays a clear role in
mediating asbestos-induced disease, the direct link to a role of asbestos as an initiator is somewhat more
tenuous (see Sections 6.3.3.2 and 6.3.3.3).
6.3.3.1 Interference with mitosis
Apparently, the physical presence of asbestos fibers can interfere with proper spindle formation and the
function of other structural units required for mitosis (Jaurand 1997). This typically leads to aneuploidy
(an incorrect number of copies of the chromosomes contained within a cell), development of micronuclei
(fragments of chromosomes enclosed in a membrane that are isolated from the main nucleus of the cell),
and has also been shown to lead to clastogenic effects (changes in the organization and structure of the
6.79
-------
chromosomes). Assays for these kinds of genetic alterations have consistently shown asbestos capable of
inducing these effects.
Jaurand (1997) also indicates that:
• fibers must be phagocytized by the target cells before they can interfere with mitosis;
• once phagocytized, all asbestos types are observed to interfere with mitotic activity; and
• samples enriched in long, thin fibers enhance these effects. In contrast, short fibers do
not appear to contribute to these effects.
Jaurand (1997) notes, however, that results related to fiber size have not been entirely consistent,
primarily because not all studies have rigorously controlled for or even adequately characterized the sizes
of the fibrous structures in test materials. Jaurand (1997) also notes that this mechanism is not dependent
on the formation of reactive oxygen species or any other reactive free radicals.
Among studies that suggest that surface chemistry (or presumably fiber type) may not be an important
factor in determining the degree to which asbestos (or other fibrous materials) interfere with mitosis,
Keane et al. (1999), exposed cultured V79 cells (hamster lung fibroblasts) to untreated and HCI-treated
chrysotile asbestos. The acid treatment substantially reduces the magnesium content on the surface of the
chrysotile. The authors also noted a "small" effect on fiber size due to treatment (treated fibers showed a
20% excess of short fibers). Cells were exposed to doses ranging between 0.4 and 12.7 (ig/cm2 (each
dose left in place for 24 hours and then rinsed off). Cells were harvested after an additional 24 hours and
evaluated for cytotoxicity and the presence of micronuclei.
Results from the Keane et al. (1999) study indicate that untreated fibers were shown to be slightly more
cytotoxic than treated fibers, but both treated and untreated fibers were observed to increase the
abundance of micronuclei in a similar, dose-dependent fashion. The induction of micronuclei appeared to
saturate at approximately 35/1,000 cells observed at an applied asbestos concentration of 40 ng/cm2. In
contrast, substantial cytotoxicity was only observed at the highest doses employed. The authors thus
concluded that surface chemistry (at least in terms of magnesium content) does not appear to have a major
affect on induction of micronuclei and that the observed genetic damage and cytotoxicity appear to occur
through entirely independent mechanisms.
In other studies, cultured cells were assayed for a variety of genotoxic effects following exposure to a
range of fibrous materials.
For example, Dopp and Shiffmann (1998) dosed human amniotic fluid (HAF) cells or Syrian Hamster
Embryo (SHE) cells with UICC amosite, Rhodesian chrysotile, crocidolite, or ceramic fibers at
concentrations of 0.5, 1.0, 5.0, and 10.0 jig/cm2 and assayed the cultures for formation of micronuclei and
a variety of clastogenic effects.
Based on their study, Dopp and Shiffmann (1998) report that all asbestos types induced formation of
micronuclei in SHE cells in a dose-dependent fashion at rates that were significantly elevated over control
animals. The effect appeared to saturate at doses between 1 and 5 ug/cm2 and rates appeared to peak at
between 48 and 66 hours post-exposure. Ceramic fibers, which were noted to be longer but thicker than
the asbestos fibers tested, also showed significantly increased induction of micronuclei, but at rates less
pronounced than for asbestos. However, it is difficult to judge whether this is due to differences in fiber
type because the data in the paper are not adequate to distinguish effects of fiber size and number from
effects of fiber type. Similar results were obtained with HAF cells, but overall rates were about one third
those observed in SHE cells.
6.80
-------
The authors also note observing various disturbance of chromatin structure during interphase. They
report observing formation of chromatin bridges and chromosome displacements in meta and anaphases
and impaired chromatin separation in mitosis. They also report that cytokinesis was frequently blocked.
In addition to the effects that they observed that are attributable to fiber interference with mitosis, Dopp
and Shiffmann report observing a variety of clastogenic effects. In all cases where the authors labeled
specific regions of specific chromosomes, fiber-exposed cells showed significantly greater frequencies of
DNA breaks over controls or gypsum-exposed animals. Different regions of various chromosomes also
showed significantly different frequencies of breakage with patterns that were specific to the different
fiber types. The authors hypothesize that the observed clastogenic effects may be due to production of
reactive oxygen species, to formation of some type of clastogenic factor, or to the direct interactions
between fibers and chromosomes during mitosis that are the apparent cause of the disturbances discussed
in the previous paragraphs. However, It was not possible to distinguish among hypothetical causes of the
observed clastogenic effects in this study.
Kodama et al. (1993) exposed cultures of human bronchiolar epithelial (HBE) cells to asbestos (chrysotile
at 0 to 4 ug/cm2 and crocidolite at 0 to 300 ug/cm2). They then examined cells at 24, 48, 72, and 96 hours
following exposure for cytotoxic effects and cytogenetic effects. Results indicate that both fiber types
induced concentration-dependent inhibition of cell proliferation and colony-forming ability, but chrysotile
was 100 to 300 times more toxic. The authors report this translates to a 40-fold increase in toxicity on a
fiber for fiber basis (although the range of sizes included in this count is not indicated).
Kodama et al. (1993) also report that, at 72 hours, chrysotile (4 ug/cm2) caused a 2.7-fold increase in
binuclei and a 1.6-fold increase in micronuclei. Over the same time interval, at 300 ug/cm2, crocidolite
caused a 1.9-fold increase in binuclei, but did not cause micronuclei. They also report that chrysotile
failed to produce significant numerical chromosome changes in HBE cells and increased structural
aberrations only at the 24-hour time point. The frequency of neither aneuploidy nor polyploidy was
increased at any time point following exposure to asbestos in this study. The authors indicate that this
contrasts with observations of relatively high incidences of asbestos-induced chromosome changes
observed in some rodent cell cultures and clastogenic effects observed in human mesothelial cell lines.
They further speculate that phagocytic cells with high mitogenetic activity are likely most susceptible to
the effects of asbestos, which primarily interferes with mitosis. However, they suggest that epithelial
cells that are exposed to fibers may undergo terminal differentiation (from Type II to Type 1 cells) and
thus cease mitosis. This would effectively prevent adverse genetic effects from asbestos exposure. Such
pathways are not available to mesothelial cells.
Hart et al. (1994) studied the effects of a range of fiber types (with varying size distributions) on Chinese
hamster ovary (CHO) cells. The authors indicate that such cells are very different from cells in
pulmonary tissues in that they are immortalized, aneuploid, undifferentiated, and preneoplasic. They also
note that the responses observed in these cells differs from responses observed in cells from pulmonary
tissues. Nevertheless, the implications concerning fiber size are instructive. Fibers evaluated included
long, medium, short, and UICC crocidolite and chrysotile along with a range of MMVF's, RCF's, and
fibrous glasses. Exposure concentrations ranged between 1 0 and 225
Results from the Hart et al. (1994) study indicate that all of the fibers caused qualitatively similar toxic
effects: concentration-dependent reduction in cell numbers and an increase in the incidence of abnormal
nuclei with little or no loss in viability. Fiber-induced cell death in CHO cells appears to be minor, even
at relatively high exposure concentrations. Based on mean dimensions (which is problematic), the
diameter dependence on the observed toxic effects, particularly on the formation of aberrant nuclei, was
slight or absent. However, the effect with length was striking. For lengths up to at least 20 u.m, potency
toward both cytotoxicity and the induction of aberrant nuclei increased dramatically with increasing
length. The authors also note that the lack of an observed fiber composition associated effect on the
6.81
-------
toxicity of CHO cells does not correlate with findings from recent rodent inhalation studies using the
same test fibers. The authors therefore speculate that CHO cells may not represent an appropriate in vitro
model for fiber effects. However, it may also be that effects in vitro occur over time scales that are rapid
relative to those that occur in vivo so that biodurability is not important in vitro.
In a study by Takeuchi et al. (1999) cultured human mesothelioma cells (MSTO211H) and, separately,
cultured human promyelocytic leukemia cells (HL60), which are not phagocytic, were dosed with
crocidolite between 0.6 and 6.6 ug/cm2 (no size range indicated). Studies with latex beads confirmed that
the mesothelioma cells are actively phagocytic, but that the leukemia cells are not. The authors indicate
that dosed mesothelioma cells showed significantly increased numbers of polynucleated cells, tetraploid
cells, and cells with variable DNA content at the GO/G1 transition in the cell cycle and that the extent of
effects was dose-dependent. Leukemia cells showed no such effects.
The authors further indicate that, when the mesothelioma cells were sorted by fiber content, those with the
highest fiber content showed the greatest effects. The authors also indicate that cells are stimulated
neither to release superoxide nor NO at the concentrations of crocidolite studied. However, they
hypothesize that intracellular ROS may have been generated because they report finding increased levels
of 8-OH-dGua (and oxidized form of one of the DNA bases) following crocidolite exposure.
Nevertheless, the authors conclude that the mechanisms by which crocidolite induce cytotoxicity and,
potentially, carcinogenicity is related to phagocytosis. Importantly, the effects described in the Takeuchi
et al. study are entirely consistent with effects attributable to interference with mitosis, despite any
speculation by the authors.
6.3.3.2 Generation of reactive oxygen species (ROS)
ROS have been implicated as mediators in a variety of toxic effects (including cancer initiation)
associated with a broad range of toxins (see, for example, Floyd 1990). Moreover, substantial evidence
indicates that asbestos can induce generation of ROS by several mechanisms and that asbestos-induced
ROS play a role in several of the toxic effects attributable to asbestos (see below). However, whether
ROS play an important role in asbestos-associated cancer initiation is less clear and needs to be evaluated
carefully. Therefore, evidence that asbestos induces the production of ROS and that ROS contribute to
the adverse health affects attributable to asbestos are reviewed below with particular attention to effects
that may contribute to the initiation of cancer. Contributions by ROS to other asbestos-related toxic
effects are also evaluated in later sections of this chapter.
Note, that generation of certain reactive nitrogen species (specifically the peroxinitrite ion) is closely
associated with ROS generation so that evidence for the generation of certain reactive nitrogen species
(Section 6.3.3.3) constitutes additional evidence for the generation of ROS.
Asbestos-induced Generation of ROS. Asbestos has been shown capable of generating a variety of
reactive oxygen species (ROS) including hydrogen peroxide (H2O2), superoxide (O2~), and hydroxyl
radical (OH«) via several mechanisms (see, for example, Fubini 1997; Jaurand 1997; Kamp et al. 1992)
including:
• catalytic production of superoxide from oxygen in aqueous solution;
• catalytic production hydrogen peroxide from oxygen in aqueous solution;
• catalytic production of hydroxyl radical by the Fenton reaction (degradation of hydrogen
peroxide catalyzed by iron on the surface of fibers or mobilized from the surface of
fibers);
6.82
-------
• catalytic production of several ROS by redox cycling of iron on the surface of fibers or
mobilized from the surface of fibers (Haber-Weiss type reactions);
• catalytic production of ROS by release of heme and heme protein from various cellular
components;
• by inducing release of various ROS species from phagocytes during "frustrated"
phagocytosis; and
• by binding to cell receptors or other features of surface membranes that trigger signaling
cascades mediating the production and release of various ROS (and RNS).
For a general review of the chemistry involved in these processes, see Floyd (1990).
Fenton, Haber-Weiss, and Related Reactions. Most of the evidence for Fenton and Haber-Weiss
reactions (and related free radical generating reactions) that take place on the surface of asbestos fibers
comes from experiments in cell-free systems (see below). Therefore, their relevance to the conditions
found in vivo may be limited. Moreover, the size of the fibers (especially in terms of their cumulative
surface area) and the history of their surfaces (in terms of metal contaminants or coatings that might be
present) may substantially alter the effects of such experiments (Fubini 1997). Unfortunately, however,
few of the available studies report characterization of fiber sizes or surface conditions in sufficient detail
to judge the importance of these effects.
For fibers. Zalma et al. (1987) evaluated a range of fibers (UICC crocidolite, UICC amosite, UICC
Canadian and Zimbobwean chrysotile, industrial chrysotile, and magnetite) for their ability to produce
free radicals by the direct reduction of oxygen in aqueous solution. In some cases, hydrogen peroxide
was also added to the solution. Results indicate that all of the fibers tested were able to generate hydroxyl
radicals (even in the absence of hydrogen peroxide), but that the efficiency of production was a strong
function of the activation (by grinding) or pacification (by coating with benzene or other agents) of the
fiber surface. Chrysotile was reported to be the most efficient at generating radicals and the authors
assumed that this is due to iron contamination on the surface (since iron is not a component of the "ideal"
chrysotile fiber). However, such conclusions are difficult to evaluate in the absence of simultaneous
consideration of fiber size.
Governa et al. (1998) evaluated the ability of wollastonite fibers to generate ROS in both a cell free
system and a suspension of polymorphonucleocytes (PMN's). The fibers were observed to produce ROS
in both systems and that ground wollastonite produced substantially more ROS than unground. The
efficiency of ROS generation in PMN suspension is also reported to be greater for wollastonite than for
either chrysotile or crocidolite (tested in previous studies). However, no size information is given. Based
on additional work with various inhibitors added to the system, the authors also indicate that only a
fraction of the ROS generated was composed of hydroxyl radicals.
Brown et al. (1998) subjected several different fiber types (amosite, silicon carbide whiskers, RCF-1, and
various fibrous glasses) to two standard chemical assays for free-radical production (in cell-free systems).
The authors indicate that, of the fiber types tested, only amosite showed free radical activity significantly
above controls in both assays and only RCF-1 additionally showed significantly elevated free radical
activity in one assay. However, there is not enough information provided in this study to determine
whether the observed differences are due to differences in fiber sizes, sample preparation (i.e., surface
condition), or fiber type. In apparent contrast, for example, Gold et al. (1997) report that amosite and
crocidolite produce few free radicals in cell free systems, unless they are ground.
6.83
-------
Weitzman and Graceffa (1984) indicate that chrysotile, crocidolite, and amosite are all capable of
catalyzing the generation of hydroxyl radicals and superoxide from hydrogen peroxide in vitro and that,
based on experiments with various iron chelators, these reactions are iron dependent. The authors further
indicate that hydrogen peroxide is produced in large quantities as a normal bi-product of tissue
metabolism, but that it is effectively scavenged by various enzymes. The authors speculate that, by
physically damaging cell membranes, asbestos may allow release of the precursor hydrogen peroxide
before it can be scavenged.
For particles. Silica, residual fly ash, and ambient air dusts also can create ROS in vitro and the
efficiency of production correlates with ionizable concentrations of various transition metals complexed
on the dust (Martin et al. 1997). In vivo, such metal containing particles also cause release of ROS from
macrophages (Martin et al. 1997). Additionally, binding of silica to plasma membranes of airway lung
cells and phagosomes provokes generation of ROS.
Castronova et al. (1997) further indicate that it is the concentration of contaminating iron on the surface
of freshly fractured quartz that enhances free radical production in aqueous solution (in cell free systems).
These authors also showed that high iron-containing (430 ppm) quartz dust inhaled by rats (at 20 mg/m3
for 5 hours/day for 10 days) induced 5 times more leukocyte recruitment, 2 times more lavageable red
blood cells, 30-90% increase in macrophage production of ROS, 71% increase in nitric oxide production
by macrophages, and 38% increase in lipid peroxidation of lung tissue than observed in rats exposed to
low iron-containing (56 ppm) quartz.
Although iron is required for the reactions considered here, studies indicate that the iron content of the
fiber itself is not a good indicator of reactivity (Gold et al. 1997). Studies also indicate that the iron that
participates in these reactions need not originate with the fiber (Fubini 1997; Jaurand 1997) and
biological systems contain abundant sources of iron. Therefore, both iron-containing fibers and iron-free
fibers have been shown to participate in these reactions in vivo.
Release ofheme and hemeprotein. At least one research group (Rahman et al. 1997) indicates that heme
and heme protein cause extensive DNA damage in the presence of asbestos in vitro and, based on
previous studies, that this may involve heme catalyzed production of ROS following asbestos-induced
release ofheme from cytochrome P-450, from prostaglandin H synthetase (or perhaps from other heme
containing proteins). Importantly, the authors indicate that such observations relate to a nuclear pool of
heme, which suggests that ROS generation via this mechanism may occur in the immediate vicinity of
DNA. The work by this group suggests at least one additional pathway by which asbestos may induce the
production of ROS and by which ROS-mediated damage to DNA might occur.
Frustrated phagocytosis. Numerous studies indicate that long asbestos fibers (longer than somewhere
between 10 and 20 urn) cannot be efficiently phagoycitized by macrophages (see, for example, Sections
4.4 and 6.2) and that macrophages that are damaged by such "frustrated" phagocytosis release ROS (see,
for example, Kamp et al. 1992; Mossman and Marsh 1991). Due to the differences in the size of
macrophages across species (see discussion of Krombach et al. [1997] in Section 4.4). The minimum
length beyond which phagocytosis may become frustrated may differ in different animals. However, it is
clearly longer fibers that contribute to this mechanism for generating ROS. Shorter fibers (<10 um) are
unlikely to cause frustrated phagocytosis in any of the mammalian species of potential interest.
Lim et al. (1997) showed that alveolar macorphages in culture (after stimulation with
lippopolysaccharide) generated free radicals (ROS) when subsequently exposed to chrysotile, crocidolite,
or amosite (all UICC). They found chrysotile to be the most potent inducer of free radical activity (which
is not surprising given that UICC chrysotile contains the highest fraction of long fibers of the UICC
samples tested (Berman, unpublished). Based on tests with various inhibitors, the authors indicated that
6.84
-------
the free radicals generated by the alveolar macrophages occurred through a pathway mediated by protein
tyrosine kinase, phospholipase C, and protein kinase C and that the effects are dose-related.
Kostyuk and Potapovich (1998) cultured peritoneal macrophages and showed that treatment with
chrysotile asbestos (1 jig, no size data given) resulted in production of frustrated phagocytosis and cell
injury (the latter as evidenced by release of lipid dehydrogenase, LDH, a marker for membrane damage).
By working with various chelators and flavonoids (natural plant products, some of which quench
superoxide and some of which chelate iron), the authors indicate that cell injury was likely induced by
superoxide and that the superoxide was likely produced by an iron-dependent mechanism. Note that this
contrasts with the above studies that suggests production of radicals by frustrated phagocytosis in culture
is an iron-independent mechanism.
At least for one kind of phagocyte: polymorphonucleocytes (PMN's), a study by Ishizaki et al. (1997)
suggests that crocidolite and erionite may induce production of ROS from PMN's by each of two
mechanisms. The first requires phagocytosis and may represent the traditional, "frustrated" phagocytosis
pathway indicated above. The second pathway is triggered by an interaction between the fiber and the
cell surface and is mediated by NADPH. The authors also cite evidence that chrysotile may similarly act
through both of these pathways.
Afaq et al. (1998) cultured alveolar macrophages and peripheral red blood cells (RBC's) that were
harvested from rats 30-days following a single 5 mg intratracheal instillation of UICC crocidolite, UICC
chrysotile, or ultrafine titanium dioxide. The authors indicate that populations of alveolar macrophages
were significantly increased (over sham-exposed rats) for all three particle types and that acid
phosphatase and lipid dehydrogenase (LDH), which are markers of cell membrane damage, were
observed in cell-free lung lavage from animals exposed to all three particle types. Both alveolar
macrophages taken from asbestos-exposed animals (both types) showed significantly elevated lipid
peroxidation and hydrogen peroxide production over titanium dioxide exposed animals. However, the
latter also showed elevated peroxidation and peroxide production that were significantly elevated over
sham-exposed animals. Similar results were observed among RBC's from asbestos-exposed animals, but
not from titanium dioxide-exposed animals. Note, it is possible that ROS production induced by Ti02
occurs by a different mechanism (or set of mechanisms) than that for asbestos (see, for example, Palekar
et al. 1979, Section 6.3.4.4).
In vivo Evidence for Asbestos-generated ROS. Several studies involving whole animals also indicate
that asbestos exposure induces the generation of ROS. Importantly, in such studies, evidence for
generation of ROS is generally determined based on observation of the putative effects of ROS, rather
than ROS directly.
Ohio et al. (1998) intratracheally instilled 500 jig of crocidolite (NIEHS) into rats. This was observed to
induce a neutrophilic inflammatory response within 24 hours (in contrast to saline-exposed rats). The
authors collected chloroform extracts from exposed lungs and subjected them to electron spin residence
(ESR) spectrometry. Results indicate the presence of a carbon-centered radical adduct that has a structure
consistent with products of lipid-peroxidation. The radical signal was only observed in asbestos-exposed
animals and persisted even after one-month following exposure. The authors also report that depletion of
neutrophils did not affect the signal and that dextrin-induced inflammation did not produce the signal.
Yamaguchi et al. (1999) studied effects in rat lung tissue at 1, 3, 5, 7, and 9 days following a single
intratracheal instillation of 2mg of either glass fiber or UICC crocidolite. The authors indicate
significantly increased levels of 8-OH-Guanine (an oxidized form of the DNA base Guanine) one day
after crocidolite instillation and increasing repair activity for this oxidized form of guanine with time that
became significant at days 7 and 9 following instillation. Glass fibers (noted to be non-fibrogenic and
6.85
-------
non-tumorigenic) did not produce either increases in 8-OH-Gua or its repair activity. The effects
associated with crocidolite were all noted to be dose related.
Several of these studies also suggest distinctions in ROS generation (either the absolute generation of
ROS or generation of specific ROS species by specific tissues) due to differences in fiber (or particle) size
or type.
Nehls et al. (1997) intratracheally instilled rats either with quartz (2.5 mg) or corundum (2.5 mg). The
latter mineral is reportedly non-tumorigenic. Results indicate that lung epithelial cells in quartz exposed
rats exhibited increased 8-oxo-Guanine levels (a DNA adduct generated by reaction with ROS, see above)
that persisted for up to 90 days post-exposure. Elevated levels of the DNA adduct appeared in all cell
types in all areas of the lung. The authors suggest that the observed persistence of the elevated levels of
8-oxo-Guanine suggests that it was produced at a rate in excess of the lung's capacity for repair. The
authors also report enhanced and persistent inflammation, cell proliferation, and an increase in neutrophil
population in bronchio-alveolar lavage (BAL) fluid and an increase in tumor necrosis factor alpha
(TNF-a) in BAL fluid in the quartz exposed animals. TNF-a is a cytokine linked to a variety of effects
including the recruitment of inflammatory cells (see Table 6-6). In contrast, exposure to corundum
produced none of the effects observed with quartz.
Timblin et al. (1998b) report that ROS induced responses by rat lung epithelial cells vary depending on
whether exposure is to crocidolite, hydrogen peroxide, or cadmium chloride (CdCl2). In response to ROS
generation induced by the first two agents increased levels of cJun protein (a protooncogene, Table 6-6) is
observed. Further, crocidolite, but not hydrogen peroxide, causes elevation in the levels of manganese-
containing superoxide (MnSOD) dismutase (an enzyme that catalyzes dismutation of superoxide, Table
6-6). Neither of these agents affect levels of either of two common stress proteins (Table 6-6): GRP78 or
HSP72/73, nor do they affect cellular glutathione levels. In contrast, cadmium chloride does not alter
MnSOD levels, but increases levels of GRP78 and HSP72/73 in addition to cJun protein. Therefore,
ROS-related mechanisms may be complex and may be toxicant-specific. Thus, it may not be correct to
assume that all fibers and particles act through common ROS-related pathways.
In contrast to the results of the above study (which showed no affect on glutathione levels), Golladay et
al. (1997) showed that human lung epithelial cells (cultured A549 cells) exposed to crocidolite (NIEHS
sample) showed substantial reduction in intracellular levels of glutathione (without increases in the
oxidized forms of glutathione). Rather, an associated increase in extracellular, reduced glutathione was
observed, suggesting that crocidolite induces release of glutathione from the interior of these cells to the
environment. The authors also indicate that, given that the half-life for reduced glutathione outside of
cells is on the order of an hour, while extracellular reduced glutathione levels remained elevated for more
than 24 hours following exposure, cells must have been releasing reduced glutathione continuously.
Because no concomitant release of LDH or labeled adenine was observed (despite loading of cells with
labeled adenine prior to the experiment), the authors conclude that release of glutathione is not associated
with membrane disruption or apoptosis (which is induced to some degree by exposure to crocidolite).
Also, all of the effects described above were similarly associated with exposure to de-ironized crocidolite.
Thus, the iron content of the fibers does not play a role in this process.
Note that the apparent difference in the reported effect of crocidolite exposure on intracellular glutathione
levels in the above two studies might be due (individually or in combination) to differences in the cell-
types studied, differences in the size distribution of the crocidolite employed, differences in study design,
or other factors. Insufficient information is available to distinguish among these possibilities.
Kaiglova et al. (1999) intrapleurally injected rats with 10 mg of long amosite and collected bronchio-
alveolar lavage (BAL) fluid 24 hours following exposure and at later time intervals. They indicate that
total protein and alkaline phophatase (AP) were both elevated in BAL 24 hours after exposure and that
6.86
-------
AP remained elevated for at least 3 months following exposure. They also noted increased levels of lipid
peroxides in BAL at 24 hours, but not 3 months following exposure. The authors indicate that
antioxidants were significantly decreased following exposure: glutathione was significantly decreased in
lung tissue at both 24 hours and 3 months following exposure, but was normal in BAL fluid at all time
points; a-tokopherol and retinol were significantly decreased at 3 months in lung tissue; and ascorbic acid
was significantly decreased in both lung tissue and BAL at 24 hours and remained low at 3 months. The
authors indicate that decreases in antioxidants implies a role for ROS (or RNS) in lung tissue injury. It is
also possible that the varied responses of specific antioxidants may suggest a role for toxin- and injury-
specific ROS/RNS.
Conclusions Concerning Generation of ROS. Except for frustrated phagocytosis (which is unique to
long fibers), ROS generation by the mechanisms discussed above are considered a common response to
respiration of particles in general (see, for example, Martin et al. 1997 who suggest that ROS "...may be a
global signaling mechanism mediating response to particulate insult mostly by activation of kinases and
transcription factors common to many response genes." They further indicate that if the load of ROS
generated is too great, or the airway in which it is generated has been previously impaired, "...these same
mechanisms can result in deleterious respiratory lesions and outright pathology"). However, not all non-
fibrous particles are similarly capable of inducing production of ROS. As indicated above, for example,
while crystalline silica is a potent inducer of ROS production, carundum is not (Nehls et al. 1997).
Moreover, the spectrum of ROS (set of species) that are induced by particular toxicants are generally
specific to the offending toxicant (Timblin et al. 1998b). Therefore, the generic grouping of ROS
mediated pathways by particles and, especially, by particles and fibers, does not appear justified. These
mechanisms are more complex and individualized than such generic grouping suggests.
ROS can be generated by multiple pathways that are variously dependent on particle size, whether a
particle is a fiber, fiber size, and particle or fiber type (i.e., chemical composition). The primary
mechanism(s) through which ROS are generated in response to one type of particle or fiber may be very
different than that through which ROS generation is induced by another and the resulting suite of ROS
(set of species) may also differ. Importantly, the relationship between dose and response for each
mechanism may also differ (see, for example, Palekar et al. 1979, Section 6.3.4.4).
Given the above, comparing among the ability of fibrous materials and non-fibrous analogs to induce
generation of ROS requires that such analogs be properly matched before valid conclusions can be drawn.
Thus, for example, the appropriate non-fibrous analog for crocidolite is the massive habit of reibeckite
and the appropriate analog for chrysotile is the massive habit of antigorite or lizardite. Due to differences
in both chemistry and crystal structure, crystalline silica is not an appropriate non-fibrous analog for any
of the asbestos types. Moreover, because ROS can be generated by different mechanisms, critical
comparison across analogs requires more than the simple confirmation that ROS are generated or even
whether the relative efficiency with which ROS are generated is comparable. It is also necessary to
contrast the specific complexion of ROS (set of species) generated and the specific tissue/cellular
environments (i.e., locations within a cell) in which they are generated in response to each analog.
It is also clear that ROS are generated by both iron-dependent and iron-independent pathways. Even for
the iron-dependent pathways, however, the source of iron need not derive from the fibers or particles
themselves. Therefore, since iron is abundant in vivo (and the environment), both iron-containing and
iron-free fibers (or particles) can potentially participate in both the iron-independent and the iron-
dependent pathways.
Effects Mediated by ROS. ROS have been implicated as mediators in a variety of toxic effects
(including cancer initiation) associated with a broad range of toxins (see, for example, Floyd 1990; Martin
et al. 1997). Cellular and tissue effects that have been associated with the effects of ROS include:
6.87
-------
enhancement of overall uptake of particles by epithelium;
stimulation of inflammatory responses;
stimulation of various signaling cascades and production of cytokines;
inducement of apoptosis;
cytotoxicity;
mediation of cell proliferation;
formation of oxidized macromolecule (including DNA) adducts; and
induction of DNA strand breaks.
However, only the last two of the above list of ROS effects are potentially relevant to the initiation of
cancer (the topic of this section). The other effects in the above list likely contribute to the induction of
other asbestos-related diseases and may even promote (but not initiate) asbestos-related cancer. Thus,
they are addressed further in later sections of this chapter (see below).
Although, ROS generation is associated with exposure to various particles and fibers (including all forms
of asbestos), generation of ROS does not necessarily imply carcinogenesis. For example, Zhu et al.
(1998) indicate that ROS generation is induced in response to exposure to asbestos, crystalline silica, and
coal mine dust. However, based on an extensive record of human exposure, the latter (coal mine dust) is
not carcinogenic in humans. Therefore, evidence related to the last two of ROS-associated effects listed
above, are examined in more detail below.
Several studies indicate that, once generated by exposure to asbestos, ROS can interact with DNA in vitro
and in vivo to produce oxygenated adducts, primarily 8-oxo-Guanine. Of the studies reviewed above, for
example, Yamaguchi et al. (1999) observed that crocidolite, but not (non-tumorigenic) glass produced
dose-dependent increases in 8-oxo-Guanine in rat lung tissue following intratracheal instillation. Also,
Leanderson et al. (1988), Park and Aust (1998), Keane et al. (1999) indicate that asbestos induces
formation of oxidized DNA adducts in in vitro assays. Brown et al. (1998) indicate that asbestos induces
ROS-mediated DNA strand breaks.
However, not all DNA strand breaks attributable to asbestos occur through pathways involving ROS.
Ollikainen et al. (1999) exposed cultures of human mesothelial cells (MeT-5A, transfected with SV-40,
but nontumorigenic) to hydrogen peroxide (100 uM) or crocidolite (2-4 u.g/cm2), either alone or in
combination with TNF-cc and performed assays for DNA strand breaks. The authors note that the
concentrations of asbestos evaluated are well below those that have been associated with cytotoxic
effects. Crocidolite alone was shown to produce DNA strand breaks at the concentrations tested. The
authors also note that, at lower concentrations, only reversible effects were observed (presumably
indicating DNA repair). Co-exposure to crocidolite and TNF-a increased the observed incidence of DNA
damage, but the effect was less than additive. The authors indicate that additional studies with
antioxidants indicate that the DNA damage induced by crocidolite in this study occurs through a
mechanism that does not involve ROS. In fact, a potentially much more substantial mechanism by which
asbestos may induce DNA breaks and various clastogenic alterations involves its ability to interfere with
mitosis (Section 6.3.3.1).
There is also evidence that different tissues may respond to asbestos-induced ROS generation differently.
For example, Zhu et al. (1998) indicate that MnSOD is found in the mitochondria of Type II epithelial
cells of rats exposed to crocidolite. The authors further indicate that, because fibroblasts, alveolar
macrophages, or endothelial cells do not display this protein when stimulated by exposure, this suggests a
difference in the susceptibility of epithelial cells to certain types of asbestos-induced injury.
Importantly, although these studies provide evidence that indicates asbestos is capable of producing
oxidized DNA adducts (or strand breaks) through ROS mediated processes, they do not address the
question of whether such adducts can lead to heritable mutations in DNA in vivo nor do they indicate
6.88
-------
whether such adducts lead to tumor production. Therefore, such studies should only be construed to
suggest a potential that asbestos can act as a cancer initiator through ROS-mediated pathways.
As previously indicated, not all mechanisms involving the generation or effects of ROS are similarly
fiber-size or fiber-type dependent and those that are do not necessarily depend on these variables in the
same way. At this time, it is not possible to distinguish among the relative importance of the different
mechanisms, so that it is difficult to judge the relative importance of the different effects. However, the
overall general implication (with the few exceptions noted) is that ROS generation more likely contributes
to other asbestos-related diseases and to cancer promotion than to cancer initiation.
One final note concerning the effects of ROS that specifically involves the behavior of the hydroxyl
radical is also warranted. The hydroxyl radical is an extremely reactive species. Whether in vitro or in
vivo, this species will react with virtually 100% efficiency with every organic molecule it encounters.
Therefore the effects attributable to the hydroxyl radical are limited to those involving reactions in the
immediate vicinity of the location at which it is generated. Thus, unless asbestos-induced generation of
this radical occurs within the nucleus and in the immediate vicinity of susceptible strands of DNA, it is
unlikely that these radicals are the direct cause of DNA damage.
Rather, hydroxyl radicals tend to react with cell membranes and other cellular components to produce
further intermediate radicals (primarily lipid peroxides), which are much more stable than the hydroxyl
radical and may migrate substantial distances before having an effect. It is likely that these intermediate
radicals are ultimately responsible for any ROS-mediated DNA damage that may be attributed to
asbestos.
6.3.3.3 Generation of reactive nitrogen species (RNS)
Nitric oxide (NO) is produced ubiquitously in biological systems and serves many functions (Zhu et al.
1998). It is highly reactive and, therefore, generally short-lived in vivo. However, nitric oxide has been
shown to react with superoxide (O2~) to form the peroxynitrite ion (ONOO") at near diffusion-limited rates
(see, for example, Zhu et al. 1998). This the peroxynitrite ion (an RNS) may represent the primary
species responsible for the effects attributed to ROS (at least when the primary ROS formed is
superoxide).
Asbestos-induced Generation of RNS. Alveolar macrophages, lung endothelial and epithelial cells, and
alveolar epithelium (in both rats and humans), when stimulated by inflammatory agents, generate both
superoxide and over-produce nitric oxide. These then combine to produce the perxoynitrite ion (Martin et
al. 1997; Zhu et al. 1998). These cells up-regulate production of NO when stimulated by various
cytokines, lipopolysaccharide, and interferon y. Because the peroxynitrite ion is a strong oxidant and
nitrating agent and is extremely reactive, evidence for its production is generally indicated in most studies
by the presence of nitrotyrosine, the stable product of tyrosine nitration, (Zhu et al. 1998). Evidence for
production of NO is frequently indicated by observation of nitrite. Numerous studies also provide
evidence of nitric oxide and peroxynitrite ion production specifically in response to exposure to asbestos
in various tissues.
Both chrysotile and crocidolite up-regulate the production of nitric oxide by alveolar macrophages in the
presence of interferon-y and the interaction between asbestos and interferon is synergistic. Non-
fibrogenic carbonyl iron did not induce nitric oxide formation (Zhu et al. 1998). These authors also cite a
study in which intratracheal instillation of silica and coal mine dust caused more inflammation and nitric
oxide formation than TiO2 or carbonyl iron (on an equal particle basis). This suggests both the geometry
and chemical composition of particles determine their ability to up-regulate nitric oxide.
6.89
-------
Inhalation of chrysotile or crocdolite induces secretion of both TNF-a and nitric oxide by pleural
macrophages (Tanaka et al. 1998). Tanaka et al. (1998) studied the effects of RNS in rats exposed by
inhalation to 6 to 8 mg/m3 crocidolite or chrysotile (both NIEHS samples) for 6 hours/day, 5 days/week
for 2 weeks. Rats were sacrificed at 1 and 6 weeks following exposure. The authors indicate that
asbestos induces formation of stable products of nitric oxide in cells obtained by lung lavage 1 week after
asbestos exposure. Nitrotyrosine (a marker for ONOO' formation) was also observed. Also, a greater
number of alveolar macrophages and pleural macrophages were shown to express iNOS protein (the
inducible form of nitric oxide synthase, Table 6-6) than sham exposed animals. Exposed rats showed
significantly elevated immuno-staining for nitrotyrosine in the region of thickened duct bifurcations as
well as within bronchiolar epithelium, alveolar macrophages, and mesothelial cells of both the visceral
and parietal pleura. Nitrotyrosine staining was persistent, being observed at both 1 and 6 weeks following
exposure.
Quinlin et al. (1998) studied the production of nitric oxide in rats exposed to crocidolite or chrysotile
asbestos (both NIESH reference samples). Rats were exposed by inhalation at 6 hours/day, 5 days/week
and lavaged at 3, 9, and 20 days. Lavaged macrophages showed significantly increased nitrite/nitrate
(indicating production of nitric oxide) and this was suppressed with inhibitors to iNOS. Thus, nitric oxide
is produced via an iNOS pathway. The authors also note that nitric oxide production correlated
temporally with neutrophil influx in the lavage fluid. They also indicate that asbestos exposed animals
showed a 3- to 4-fold increase in iNOS positive macrophages in their lungs.
Quinlin et al. (1998) also exposed cultured murine alveolar macrophages (RAW 264.7 cells) to
crocidolite, riebeckite, and crstobalite silica in vitro. These cells showed increased iNOS mRNA
following exposure to asbestos and even greater increases if the cells were also stimulated with
lipopolysaccharide (LPS). Both crocidolite and riebeckite (but not cristobalite silica) stimulated increased
iNOS promoter activity when applied in combination with LPS. Thus, in this case, their appears to be a
mechanism that is sensitive to particle composition, but not size.
Park and Aust (1998) treated cultures of human lung epithelial (A549) cells with crocidolite and observed
induction of iNOS and reduction of intracellular glutathione (GSH) levels. Based on studies with
inhibitors, the authors further indicate that iron mobilized from crocidolite was required both for
formation of nitric oxide and to generate 2'-deoxy-7-hydro-8-oxoguanosine, but not for the observed
decrease in intracellular glutathione. The authors note that approximately 5 times as much chrysotile
(containing approximately 3% iron) as crocidolite was required to produce the same level of nitric oxide
formation. Importantly, these experiments were conducted in vitro in serum-free medium, so that no
extra-biological source of iron was present.
Choe et al. (1998) dosed cultured rat pleural mesothelial cells with either chrysotile or crocidolite (both
NIEHS samples) at concentrations between 1.05 and 8.4 |ig/cm2 with or without co-stimulation with 50
ng/ml of interleukin-lp (IL-lp). The authors report that mRNA for iNOS in asbestos and IL-lp dosed
cells increased progressively from 2 to 12 hours following exposure. Both types of asbestos also
stimulated production of nitric oxide (measured as nitrite) in IL-lp stimulated cells in a dose- and time-
dependent fashion. Both types of asbestos also induced expression of iNOS protein and formation of
nitrotyrosine (based on nitrate detection) in IL stimulated cells. In contrast, carbonyl iron particles did
not induce any of the effects observed for asbestos in IL stimulated cells. Thus, formation of RNS
appears to be either fiber size dependent (not induced by particles) or mineralogy-dependent (or both).
As with the production of ROS, RNS production is apparently a function of multiple, complex
mechanisms. Also, as with production of ROS, the dose-response characteristics of the various
mechanisms differ. Several of the mechanisms show a strong dependence on fiber size and some
dependence on fiber type. However, there are other mechanisms that are dependent primarily on fiber (or
particle) type, but may not be dependent on size (or at least not dependent on fiber size). At this point in
6.90
-------
time, it is not possible to gauge the relative importance of the various mechanisms by which RNS may be
generated, except that it is likely that the importance of the various mechanisms likely differ in different
cells and tissues and likely differ as a function of the specific toxin whose presence is inducing RNS
production.
There are also indications that production of RNS may be (animal) species specific. For example, Jesch
et al. (1997) report that alveolar macrophages harvested from rats expressed iNOS when stimulated with
either LPS (lippopolysaccharide) or interferon-y. In contrast, iNOS expression could not be induced in
hamster, monkey or human macrophages.
Effects Mediated by RNS. Based on Zhu et al. (1998), over-production of nitric oxide can:
• inactivate critical enzymes;
• cause DNA strand breaks that result in activation of poly-ADP-ribosyl transferase
(PARS);
• inhibit both DNA and protein synthesis; and
• form peroxynitrite by reaction with superoxide.
In turn, peroxynitrite ion may:
• initiate iron-dependent lipid peroxidation;
• oxidize thiols;
• damage the mitochondria! electron transport chain; and
• nitrate phenolics (including tyrosine).
Also, some of the damage to alveolar epithelium and pulmonary surfactant system previously attributed to
reactive oxygen species may actually be caused through ROS generation of peroxynitrite. For example,
Chao et al. (1996) report that crocidolite treatment of human lung epithelial cells (A549 cells) results in
formation of 8-hydroxy-2'-deoxyguanosine (8-OHdG) in DNA and synthesis of mRNA for iNOS. An
iNOS inhibitor reduces intracellular nitrite and eliminates production of 8-OHdG. Addition of
independent NO donor, recovers production of 8-OhdG. Thus, production of the oxygenated DNA
adduct in this case appears to be generated by reaction with RNS.
As with ROS, it is primarily the potential for RNS to contribute to DNA damage (e.g., strand breaks or
generation of oxygenated adducts) that represent the primary pathways by which RNS might participate
in cancer initiation (as opposed to cancer promotion or other asbestos-related diseases). It appears that
RNS-mediated DNA damage is closely associated with ROS generation and mediation of DNA damage
(see Section 6.3.3.2). Thus, there is little to add here.
6.3.3.4 Conclusions concerning asbestos as a cancer initiator
The strongest, most consistent evidence that asbestos can act as a cancer initiator relates to the tendency
of asbestos to interfere with mitosis. Although there is evidence that asbestos may induce production of
DNA adducts and DNA strand breaks (through ROS and RNS mediated pathways), whether such adducts
or breaks ultimately lead to permanent, heritable changes to DNA remain to be demonstrated. The
relative importance of the ROS/RNS mediated pathways compared to the pathway involving interference
with mitosis also remains to be determined. As indicated in later sections, however, ROS/RNS mediated
pathways may play substantial roles in cancer promotion and induction of other asbestos-related diseases.
Regarding the primary mechanism by which asbestos may initiate cancer (interference with mitosis), the
pathway is length dependent (short fibers do not appear to contribute to the effect). Further, although
6.91
-------
there may also be a dependence on fiber type (chemical composition), it is apparent that all types of
asbestos can act through this pathway. The dependence of this pathway on fiber diameter is less clear.
Thus, other than to suggest that fibers must be respirable (and therefore thinner than approximately 0.7
urn, Section 6.1) to have an opportunity to act, whether further diameter constraints are associated
specifically with the mechanism of cancer initiation is not known.
For pathways involving generation of ROS or RNS, some mechanisms (such as those associated with
frustrated phagocytosis) are length dependent, others are not. Although there appears to be some
dependence on fiber type for several mechanisms and a general dependence of several of these
mechanisms on surface chemistry, the relative importance of fiber type to the overall contributions from
these pathways remains to be determined.
Importantly, although crystalline silica may also act to produce some of the same effects as asbestos
(including potentially induction or promotion of cancer), there is substantial evidence that this material
does not act through the same pathways and that the characteristics of its respective dose-response
relationships may differ. Thus, for example, while asbestos likely initiates cancer through a mechanism
that favors long (and potentially thin) fibers, silica more likely acts through a mechanism that is
dependent on total surface area, with freshly and finely ground material likely being the most potent. In
contrast, grinding asbestos fibers tends to lesson carcinogenicity overall. Due to differences in chemistry
and crystallinity (reinforced by studies indicating a lack of correspondence in behavior) crystalline silica
does not appear to be an appropriate analog for any of the asbestos types. Rather, for example, the
appropriate non-fibrous analog for crocidolite is riebeckite and the appropriate non-fibrous analog for
chrysotile is antigorite or lizardite.
There is also evidence that the relative importance of asbestos as a cancer initiator may differ in differing
tissues. For example, asbestos can only interfere with mitosis in cells that actively phagocytize fibers and
not all cell types actively phagocytize particles (although both mesothelial cells and pulmonary epithelial
cells appear to actively phagocytize fibers, Section 6.3.3.1). However, there is also evidence that
pulmonary epithelial cells (Type II) may undergo terminal differentiation to Type I cells and thus escape
potential cancer initiation by asbestos (Section 6.3.3.1). Such a pathway is not available to mesothelial
cells. To the extent that pathways involving generation of ROS or RNS contribute to cancer initiation, as
indicated throughout this section, the rates of generation and the spectrum of the species generated varies
as a function of cell type.
6.3.4 Evidence that Asbestos Acts As a Cancer Promoter
Primarily, asbestos may promote cancer by facilitating tissue proliferation. However, additional
mechanisms associated with the observed interaction between asbestos exposure and smoking (Section
6.3.4.6) also need to be considered.
Substantial evidence exists indicating that asbestos induces proliferation in target tissues associated with
lung cancer and mesothelioma and this is summarized below followed by an overview of studies that
suggest the various mechanisms by which asbestos may facilitate such proliferation. Evidence suggesting
the various mechanisms related to the interaction between smoking and asbestos exposure are also briefly
reviewed.
There are numerous mechanisms by which asbestos may facilitate proliferation including:
• direct cell signaling to induce proliferation. This may occur by:
- direct interactions between fibers and receptors on the cell surface;
- interactions between phagocytized fibers and intracellular components of
6.92
-------
signaling cascades; and
- or interactions between cells and intermediate species (e.g., ROS or RNS)
whose generation and release has been induced by asbestos; or
• response to induced cell death in target tissues, which then stimulates stem cells to
proliferate to replace killed cells. Cell death may be induced either through:
- inducing apoptosis (programmed cell death); or
- direct cytotoxic effects, which leads to necrosis.
These pathways are summarized in Table 6-5, which provides a perspective on the complexity of the
interactions between asbestos and the cells and tissues of the body.
Regarding asbestos-induced cell death, asbestos-induced apoptosis (and all of the other effects described
above) typically occurs at exposure concentrations that are much lower than required to induce frank
cytotoxic effects (Sections 6.3.4.3 and 6.3.4.4). Therefore, it is primarily the former that is of potential
interest in terms of implications for asbestos-induced diseases in humans. Due to the high exposure
concentrations typically required, the importance of contributions from frank cytotoxicity to human
disease is unclear (Section 6.3.4.4).
6.3.4.1 Asbestos-induced proliferation
Numerous in vivo studies indicate that all types of asbestos induce proliferation in target tissues relevant
to lung cancer and mesothelioma. Such proliferation is also suggested by the animal histopathological
observations previously described (Section 6.2.2). Moreover, many of these studies suggest that the
underlying mechanisms may be fiber size- and fiber-type specific. Responses may also be species-
specific.
Importantly, there are some studies (primarily in vitro studies) that suggest asbestos acts to inhibit (rather
than induce) proliferation. Although it is likely that the contrasting observations found in such studies are
due to differences in conditions, timing, dose, or the type of asbestos employed, the underlying reason for
the contrasting observations is not always apparent.
The evidence for proliferation is summarized by tissue below.
In lung epithelium. Brody et al. (1997) showed that rats and mice exposed for a brief (5 hour) period to
chrysotile asbestos at 1,000 fibers/cm3 (no size indicated) exhibited focal scarring at bronchio-alveolar
duct junctions that are identical to those seen in asbestos-exposed humans. After 3-consecutive
exposures, the lesions persisted for 6 months. In regions where fibers are deposited, macrophages are
observed to accumulate, epithelium is injured, and proliferation is observed to occur. In this study, the
authors also showed by immunohybridization staining that the four genes required to express three
peptide growth factors (TGF-a, TGF-P, and the A and B chains of PDGF) and the proteins themselves are
expressed in bronchio-alveolar tissue within 24 hours of exposure. PDGF is expressed almost
immediately and expression remains elevated for 2-weeks post-exposure, but only in regions where fibers
are deposited.
The authors report that PDGF is a potent growth factor for mesenchymal cells, TGF-a is a potent mitogen
for epithelial cells, and TGF-P inhibits fibroblast proliferation, but stimulates synthesis of extra-cellular
matrix (Table 6-6). The authors also report additional experiments with knockout mice indicate that
TNF-a is required to induce the early stages of proliferation. The authors also indicate that Type II cells
produce TGF-pl and TGF-P2 (two of three isoforms of this protein) and that they are stimulated to do so
when co-cultured with macrophages.
6.93
-------
Adamson (1997) intratracheally instilled size-separated (by sedimentation) long and short crocidolite
fibers into rats (0.1 mg in a single dose) and noted that the long fibers damaged the bronchiolar
epithelium and that fibers were incorporated into the resulting connective tissue; granulomas formed with
giant cells containing fibers). The long fibers also appeared to escape into the interstitium. Labeled
thymidine uptake (which indicates DNA synthesis and suggests proliferation) following long fiber
exposure was seen in lung epithelial cells, fibroblasts, and pleural mesothelial cells. Such labeling peaked
at 2% in mesothelial cells and 3% in epithelial cells within one week following exposure. Proliferation
appeared to end shortly beyond one week. Short fibers were observed to have been efficiently
phagocytized by alveolar macrophages and a small increase in macrophage population appeared to have
been induced. Otherwise, none of the other effects attributable to long fibers (described above) were
observed with short fibers. Note that such observations are entirely consistent with those reported for the
range of studies described in Section 6.2.2.
McGavran et al. (1989) exposed both normal (C5+) mice and compliment deficient (C5-) mice to asbestos
(at 4 mg/m3) by inhalation and observed proliferation of bronchio-alveolar epithelium and interstitial cells
at alveolar duct bifurcations (based on incorporation of tritiated thymidine) between 19 and 72 hours after
a single, 5-hour exposure. Sham exposed rats showed fewer than 1% of epithelial and interstitial cells at
alveolar duct bifurcations incorporate labeled thymidine. In contrast, thymidine uptake in asbestos
exposed animals is significantly elevated for the first few days, begins to decrease at 8 days, and returns
to normal by one month following exposure. Both C5+ and C5- mice show similar increases in volume
density of epithelial and interstitial cells at 48 hours post-exposure. However, one month following
exposure C5+ mice developed fibrotic lesions while C5- mice were no different than controls. The
authors conclude that the depressed macrophage response in C5- mice does not appear to change the early
mitogenic (and proliferative) response to asbestos, but apparently attenuates later fibrogenesis.
Chang et al. (1989) describes the morphometric changes observed in rats following 1 hour inhalation
exposure to chrysotile (at 13 mg/m3). Within 48 hours following exposure: the volume of the epithelium
increased by 78% and the interstitium by 28% at alveolar duct bifurcations relative to sham-exposed
animals. Alveolar macrophages increased 10-fold and interstital macrophages 3-fold. Numbers of Type I
and Type II epithelial cells increased by 82% and 29%, respectively. At 1 month following exposure, the
numbers of Type I and Type II pneumocytes were still elevated, but not significantly. However, the
volume of the interstitium had increased by 67% accompanied by persistently high numbers of interstitial
macrophages, accumulation of myofibroblasts, smooth muscle cells, and an increased volume of
interstitial matrix.
In lung endothelium. In addition to the general evidence for proliferation of epithelial cells provided by
Adamson (1997), McGavran et al. (1989), and Chang et al. (1989), as cited above, a more detailed
description of the nature of asbestos-induced proliferation of endothelial cells is also available.
Proliferation of endothelial cells and smooth muscle cells of arterioles and venules near alveolar duct
bifurcations is induced in rats inhaling chrysotile (no size information given) for 5 hours at 4 mg/m3
(McGavran et al. 1990). This is based on the observed uptake of labeled thymidine (which indicates
DNA synthesis and suggests proliferation) that is significantly increased over controls between 19 and 72
hours following exposure. Twenty-eight percent of vessels near bifurcations exhibited labeled cells 31
hours after exposure. One month following exposure, the thickness of the smooth muscle layers around
these blood vessels is significantly increased (doubled). In contrast, labeling of these same endothelial
and smooth muscle cells in sham exposed rats is zero. The authors indicate that endothelial cells and
smooth muscles associated with pulmonary blood vessels are normally quiescent with turnover rates on
the order of years.
6.94
-------
In mesothelium. In the second part of the study addressing epithelial proliferation, Adamson (1997)
reports that rats instilled with 0.5 mg of unmodified, UICC crocidolite were sacrificed at 1 week and
6 weeks following exposure and subjected to bronchiolar lavage and pleural cavity lavage. Lavaged
alveolar macrophages were observed to contain fibers, but pleural macrophages did not. At 1 week,
collected pleural macrophages were shown to induce proliferation of fresh mesothelial cells in culture and
pleural lavage fluid showed an even greater effect. No effects were observed at 6 weeks. Further work
with anti-bodies to various cytokines indicated that early, transient proliferation of mesothelial cells was
dependent on kertinocyte growth factor (KGF), but not on PDGF, FGF, or TNF-ot (Table 6-6). This
suggests that early, transient proliferation is induced by diffusing cytokines rather than direct fiber
exposure. Adamson further reports that KGF is a fibroblast-derived cytokine that acts on epithelial cells
so that it's up-regulation likely results from epithelial injury with penetration of asbestos to the
interstitium (where fibroblasts are found, Section 4.4). A similar transient proliferative response in
mesothelial cells was also observed following exposure to chrysotile and in response to crystalline silica
exposure. Thus, it appears that the mesenchymal proliferative response may be mineralogy specific for
particles and is size-specific for fibers.
Everitt et al. (1997) exposed rats and hamsters by inhalation to RCF-1 (45 mg/m3-650 f/cm3) for 12
weeks and then let them recover for up to an additional 12 weeks prior to sacrifice. The authors indicate
that both rats and hamsters showed qualitatively similar levels of inflammation at time examined (4 weeks
and 12 weeks). They also indicate the mesothelial cell proliferation was observed in both animal types,
but was more pronounced in hamsters at all time points examined. The greatest proliferation in both
species was in the parietal pleura lining the diaphragm. The authors also report that fibers (primarily
short and thin) were also observed in the pleural cavities of both species at all time points.
Several in vitro studies provide evidence that asbestos either induces or inhibits proliferation in lung
tissues of interest and that at least some of the mechanisms involved are fiber size and mineralogy
dependent. As previously indicated, the specific reasons for the apparent contrast between results
observed in vivo (where asbestos consistently promotes proliferation) and the inhibition sometimes
observed in in vitro studies is not always apparent. However, it must be due to the special conditions that
must be created to conduct in vitro studies, which may not support certain mechanisms that are important
in vivo.
Timblin et al. (1998a) completed a study in which rat pleural mesothelioma (RPM) cells in culture were
dosed with crocidolite or various cation-substituted erionites. The expression of several gene and gene
products were then tracked. Cultures in this study were exposed to 1, 5, or 10 ug/cm2 of the various
fibrous materials. Analysis of the fibrous materials indicated that crocidolite contained many more fibers
per gram of material (probably because they are thinner) and that the preparation contains somewhat
longer fibers than the erionites evaluated. In crocidolite, for example, 88% of the fibers are longer than 5
um, 68% longer than 10 um, and 37.5% longer than 20 urn. All of the cation substituted erionites showed
approximately the same size distribution: 50% longer than 5 um, 10-20% longer than 10 um, and 1-5%
longer than 20 um.
Results indicate that the various cation substituted erionites behave differently and that the Na substituted
erionite shows the largest overall potency, at least for some endpoints, but not others. Fe, Na, and Ca
substituted erionite all appear to induce c-fos expression in a dose-dependent fashion (increasing regularly
among the 1, 5, and 10 ug applications). K-erionite may also show the same pattern, but the changes
were not indicated as significant over controls (apparently due to greater variability). Only Na-erionite
showed significantly increased expression of c-jun (at 1 and 5 jig applications, but not significantly at 10
ug, apparently due to greater variability. However, the mean result for 10 ug shows a consistent trend
with the lower concentration application results).
6.95
-------
Na-erionite induces c-fos at the same or greater rates as crocidolite asbestos for the same mass application
(but not the same fiber number). Crocidolite also appears to show a dose-response trend for c-jun
expression, but only the result for the highest application (10 ug) is significantly different from controls.
In contrast, Na-erionite appears to show greater induction of c-jun expression at lower dose than
crocidolite, but the increase with increasing dose is much lower for Na-erionite. Crocidolite also appears
to induce substantial apoptosis (even at the low dose of 5 ug) and that the induction is dose-dependent. In
contrast, non-fibrous riebeckite does not appear to induce apoptosis. Comparison between crocidolite
dosed cultures and Na-erionite dosed cultures indicate that crocidolite induces substantial apoptosis at all
time periods following application, but that Na-erionite induces little apoptosis even at higher mass dose
and longer time periods than crocidolite. The authors also indicate that crocidolite and Na-erionite appear
to stimulate DNA-synthesis, which appears to be a compensating mechanism to fiber cell toxicity. The
authors indicate that chemistry is important in fiber toxicity as Na-erionite was a strong inducer of c-jun,
even at relatively low concentrations, but several of the other cation substituted erionites (including
Fe-erionite) were not. They further suggest that, given the difference in the fiber lengths of the
crocidolite samples and Na-erionite samples, that fiber length may be a less important consideration than
fiber surface chemistry. However, despite the author's assertion, considering that non-fibrous riebeckite
does not induce any of the effects observed for crocidolite, there appears to be a clear size effect. It may
simply require that a defined, minimum length is necessary to induce the effect.
The authors also indicate that balance between proliferation and apoptosis is required to maintain
homeostasis in healthy tissue. They further indicate that other studies suggest that c-Jun expression is
linked to proliferation and induction of cancer, while c-fos expression is linked to apoptosis. Thus,
suppression of c-fos may be linked to carcinogenesis by allowing establishment or maintenance of a
transformed cellular phenotype. This is in fact an early step in carcinogensis. Many environmental
agents stimulate both apoptosis and proliferation and, depending on the degree, may cause imbalances
that lead to disease. Relative stimulation of c-fos and c-jun may reflect some of these pathways. Since
crocidolite induces both c-fos and c-jun in this study, the implication is that it mediates both apoptosis
and proliferation.
Wylie et al. (1997) dosed hamster tracheal epithelial (HTE) cells and rat pleural mesothelial (RPM) cells
with various asbestos and talc samples and-evaluated proliferation based on a colony-forming efficiency
(CFE) assay. The samples were NIEHS crocidolite and chrysotile and three different talcs. Samples were
characterized in the paper by mineralogical composition, surface area, and size distributions.
The authors indicate that both asbestos samples increased colony formation of HTE cells (suggesting
induction of proliferation), but talc samples did not. RPM cells, in contrast, showed only dose-dependent
decreases in colony forming efficiency for all samples, which the authors indicate is a sign of
cytotoxicity. The authors report that all samples show corresponding effects when concentrations are
expressed as fibers longer than 5 um or by total surface area. They also suggest that the "unique"
proliferative response by HTE cells could not be explained by either fiber dimension or surface area and
suggests a mineralogical effect.
Barchowsky et al. (1997) dosed cultured (low passage) endothelial cells to NIEHS chrysotile, crocidolite,
or RCF-1. After 1 to 3 hours exposure to 5 ng/cm2 (non-lethal concentrations), asbestos (but not RCF-1)
causes changes in cell morphology (cells elongate), increases in cell motility, and increases in gene
expression. Further work by the authors indicate that these effects are mediated by interaction between
asbestos and the receptor for urokinase-type plasminogen activator (uPAR). The authors also suggest that
attachment of asbestos to cell membranes, internalization of asbestos fibers by the cells, and the
morphological changes induced by asbestos are each mediated by different proteins.
Examples of in vitro studies that indicate asbestos (and other fibrous materials) may inhibit proliferation
in culture include the following.
6.96
-------
In contrast to the above, Levresse et al. (1997) found that chrysotile and crocidolite act to inhibit
proliferation in cultured rat pleural mesothelioma (RPM) cells. In this study, RPM cells (diploid, no more
than 25 passages) were dosed with either UICC crocidolite or NIEHS Zimbabwean Chrysotile at
concentrations varying between 0.5 and 20 ng/cm2. The authors also note that the chrysotile sample
contains approximately 4 times the number of fibers as the crocidolite samples. Cells were then examined
at 4, 24, and 48 hours following treatment. In untreated cultures, the number of cells in replicative phase
decrease with time, which indicates that such cells are headed for confluence (completion of a monolayer
on the culture medium). At 48 hours, for example, 10% of untreated control cells were observed to be in
replicative stage. Chrysotile (but not crocidolite) decreased the fraction of cells in replicative phase in a
time- and dose-dependent manner. At 48 hours, for example, cells treated with 10 ng/cm2 chrysotile
showed only 1.5% in replicative phase. Further tests confirmed that this was due to blockage of cells at
the Gl/S boundary of the cell cycle. Both chrysotile and crocidolite appears to induce a time-dependent
increase in the number of cells at G2/M in the cell cycle, although this effect was not observed to be dose-
dependent for chrysotile. Even on a fiber-number basis, chrysotile appears to be elicit a greater response
(arrest a greater percentage of cells) than crocidolite.
The authors also indicate that chrysotile caused nuclear-localized, time-dependent increases in p53
concentrations. Crocidolite produced much lower levels that were not detectable in the nucleus.
Chrysotile was also observed to produce blockage at the GO/G1 transition of the cell cycle, but crocidolite
did not. They also note that p53 is known to mediate arrest at this stage in the cycle so observing that
chrysotile induces arrest at this transition in the cell cycle may be consistent with the observed increased
expression of p53. The authors also note that chrysotile triggers apoptosis in this study and that
crocidolite shows a smaller, but detectable effect. Spontaneous apoptosis in untreated cultures ran
between 0.5 and 1% at 24-48 hours whereas chrysotile induced 4% apoptosis, peaking at 72 hours
following exposure to 10 ng/cm2. The authors indicate that the lower level of effects observed with
crocidolite could be due to the substantially smaller number of long fibers in UICC crocidolite compared
to NIEHS chrysotile.
The previously reviewed (Section 6.3.3.1) study by Hart et al. (1994) also suggests that long, medium,
short, and UICC crocidolite and chrysotile along with a range of MMVF's show a dose-dependent
inhibition on proliferation of cultured CHO cells and that potency toward the effect is a direct function of
fiber length.
Several of the above-described studies, in addition to providing evidence that asbestos induces
proliferation in various lung tissues, also suggests certain mechanisms. Asbestos may induce
proliferation, for example, by inducing production of specific cytokine growth factors (Adamson 1997;
Brody et al. 1997), or by inducing certain signaling cascades (Barchowsky et al. 1997, Timblin et al.
1998b). It is also possible that the two effects may be related (i.e., that stimulation of a particular
signaling cascade may result in production of certain growth-stimulating cytokines). Other mechanisms
may also be important (Table 6-5).
6.3.4.2 Asbestos induced cell signaling
Asbestos has been shown to induce a variety of cell signaling cascades in a variety of target cell and
tissue types. Such signaling may then trigger effects in the stimulated cells that may include:
• proliferation;
• morphological changes;
• generation and release of various cytokines, enzymes, or extracellular matrix; or
• programmed cell death.
6.97
-------
Note, due to the large number of chemical species that need to be considered in this discussion, Table 6-6
provides a summary of the sources of such species (Table 6-6A) and the effects attributable to such
species (Table 6-6B).
In specific cases, asbestos may initiate cell signaling by interacting directly with receptors on the cell
surface, by causing generation and release of intermediate species (e.g., ROS or RNS) that trigger cell
signaling, or (for phagocytized fibers) by interacting with intracellular components of a particular
signaling cascade. The specific responses to cell signaling induced by asbestos are frequently cell- or
tissue-type specific. Moreover, depending on the specific mechanism, cell signaling by asbestos may be
dependent on fiber size and/or type.
Barchowsky et al. (1998) showed in a set of studies that long chrysotile and long crocidolite, but not
RCF-1 fibers (at concentrations between 1 and 10 ug/cm2), which is reportedly below levels that typically
induce cytotoxic effects) induced up-regulation of urokinase-type plasminogen activator (uPA) and its
receptor (uPAR) in both lung endothelial cells (vascular cells) and lung epithelial cells. They also
showed that the increased pericellular protolytic activity (requiring cleavage of plasminogen to plasmin)
that is induced by asbestos in these cells is mediated by uPA.
In prior studies, chrysotile has been shown to cause endothelial cells to elongate and increase expression
of adhesion molecules for phagocytes. They also show enhanced proteolytic activity and matrix
interactions. Chrysotile has also been shown to stimulate fibrinolytic activity in lung epithelial cells and
extravasating macrophages. All such stimulation appears to result from up-regulation of uPA. Thus, this
mechanism may explain the observed asbestos-induced changes in lung endothelial and epithelial cells
including vascular remodeling, development of vascularized granular tissue, increased matrix turnover,
and leukocyte extravasation (which in turn may be caused by cell activation and elaboration of proteases
and adhesion molecules). The authors suggest that asbestos-induced up-regulation of uPA and uPAR
expression may represent a global mechanism for pulmonary toxicity and fibrosis induced by crystalline
fibers. Importantly, chrysotile was shown to induce uPA and uPAR expression in the absence of serum,
so the effect is apparently due to direct binding of fibers to cell-surface receptors.
Mossman et al. (1997) observed that concentrations of 1.25-5 ug/cm2 of crodidolite (sample from TIMA)
caused expression of c-jun and AP-1 in both cultured hamster tracheal epithelial (THE) cells and cultured
rat pleural mesothelial (RPM) cells. Crocidolite was also observed to trigger the EGFR-regulated kinase
(ERK) and mitogen activated protein kinase (MAPK) pathways in RPM cells. The authors indicate that
these pathways are also stimulated by hydrogen peroxide and that NF-KP induction stimulated by
crocidolite is also stimulated by crystalline silica (although silica stimulation of the MAPK pathway was
not investigated). They also note that the non-fibrous analog to crocidolite, riebeckite does not elicit
these activities.
The authors indicate that induction of the NF-icp cascade was inhibited by excess glutathione, which is
stimulated by N-acetylcysteine (NAC), suggesting that this pathway is induced by asbestos-caused
oxidative stress (perhaps through ROS or RNS intermediates). Application of NAC also diminished
crocidolite induced c-fos and c-jun RNA levels and inhibited activation of the ERK-MAPK cascade.
Further work suggests that asbestos triggers the MAPK pathway by interaction with the Epithelial Growth
Factor Receptor (EGFR), either directly or by phosphorylation of this receptor by ROS. At the
concentrations examined, crocidolite induces substantial apoptosis (apparently through activation of the
ERK-MAPK cascade). In contrast, the authors note that TNF-a induces the JNK arm of the ERK-MAPK
cascade, which leads to proliferation. Asbestos does not elevate JNK over the time-period of the study.
The authors note that it has been shown in some studies that inhibition of ERK in some cells, also inhibits
asbestos-induced apoptosis. Importantly, given that these processes are induced by crocidolite, but not by
its non-fibrous analog, riebeckite, induction of the ERK-MAPK cascade appears to be a fiber-size
dependent process.
6.98
-------
In a related study to that conducted by Mossman et al. (1997), Zanella et al. (1999) report that the (TIMA)
crocidolite (at concentrations of 2.5-10 ng/cm2, but not its non-fibrous analog, riebeckite, eliminated
binding of EOF to its receptor EGFR. Because EOF does not bind to crocidolite in the absence of
membrane, this is not simply a case of crocidolite tying up ligand. Crocidolite also induces a greater than
2-fold increase in steady-state message and protein levels of EGFR.
The authors also note that the tyrphostin, AG-1478 (which specifically inhibits the tyrosine kinase
activity of EGFR), significantly mitigated asbestos-induced increases in mRNA levels of c-fos, but not
c-jun, and that the asbestos action was not blocked by a non-specific tyrphostin, AG-10. Moreover,
pretreatment of RPM cells with AG1478 significantly reduced asbestos-induced apoptosis. Therefore, the
authors concluded that asbestos-induced binding to EGFR initiates signaling pathways responsible for
increased expression of the protooncogene c-fos and the development of apoptosis. This apparently
occurs through the EGFR-extracellular signaling regulated kinase (ERK). It is hypothesized that asbestos
may induce dimerization and activation (phosphorylation) of EGFR, which also prevents binding of EGF.
Asbestos apparently serves the same role as EGF in that it promotes aggregation of EGFR, which in turn
promotes binding to the extracellular domain of tyrosine kinase receptors and the activation of their
intracellular kinases. The authors also indicate that other work suggests that crocidolite fiber exposure
leads to aggregation and accumulation of EGFR at sites of fiber contact and that asbestos also stimulates
biosynthesis of the EGFR and activates ERK in an EGFR-dependent manner.
The authors speculate that asbestos binding may not be ligand-site specific, but may be charge related or
may induce EGFR phosphorylation by local production of ROS, which has previously been demonstrated
to cause EGFR activation. As previously indicated, that this effect is driven by crocidolite exposure, but
not by exposure to riebeckite indicates that this mechanism is fiber-size specific.
Johnson and Jaramillo (1997) showed that UICC crocidolite, but not JM-100 glass, applied to a culture of
immortalized human Type II epithelial (A549) cells at non-cytotoxic concentrations (for 20 hours) results
in increased expression of p53, Cipl, and GADD153 in a dose- and time-dependent fashion. Expression
was observed to be maximum at 18 hours. The crocidolite treatment was also shown to cause an increase
in the number of cells arrested in Stage G2 of the cell cycle (with a persistent decrease in the number of
cells in Gl). This was considered surprising because both p53 and Cipl are known to mediate arrest in
Stage G1. The authors suggest that these findings indicate a strong dependence on both fiber type and
fiber size (JM-100 glass contains substantially more long fibers than crocidolite). However, it is not
possible to separate the effects of fiber type versus fiber size in this study.
Luster and Simeonova (1998) indicate that at high concentrations, ROS may induce frank cytotoxicity.
At low or moderate levels, ROS are more likely to induce cell signaling cascades that may, in turn,
contribute to asbestos-related disease. The authors dosed cultures of immortalized human Type II
epithelial (A549) cells, originally derived from a lung carcinoma, and normal human bronchioepithelial
(NHBE) cells with long (Certain-Teed supplied) crocidolite (reported mean length: 19 urn) at
concentrations ranging between 0 and 24 ng/ml- Results indicate that secretion of both Interluken-8
(IL-8) and IL-6 was stimulated by crocidolite exposure in a dose-dependent manner. In contrast,
increases in LDH levels (which indicate cell damage) was only detected at the highest exposure
concentrations tested. Further work indicates that stimulation of IL-6 and IL-8 secretion occurs through
ROS that are generated in an iron-dependent process (that may also include NF-icp induction). Note that
the trend of cell signal induction at low and moderate levels of asbestos exposure with evidence for
cytotoxicity observed only at the highest exposure concentrations is common to many of these kinds of
studies.
6.99
-------
Choe et al. (1999) conducted a combined in vivo/in vitro study of the effects of low level exposure to
chrysotile or crocidolite at inducing leukocyte attachment to rat pleural mesothelial cells. The authors
note that similar populations of rat pleural leukocytes (74% macrophages, 2% neutrophils, 10% mast
cells, and 10% eosinophils) were observed in both asbestos-exposed and unexposed rats.
In the second part of the study, cultured RPM cells were exposed to either crocidolite or chrysotile (both
NIEHS samples) at concentrations ranging between 1.25 and 10 ng/cm2, which was noted to be below
levels at which substantial cytotoxicity is observed. Attachment of rat pleural leukocytes to RPM cells
was then observed to increase with increasing dose of asbestos to the RPM cells. In contrast, carbonyl
iron (a non-fibrous particle) also induced enhanced attachment, but at much lower levels and the effect
was not dose-dependent. Further analysis indicated that asbestos-induced adhesion is mediated by up-
regulation of IL-lp (but not dependent on TNF-a or nitric oxide production, although it is noted that
TNF-
-------
IL-6 is elevated in lavage fluids following exposure to Ni2S3, a suspected human carcinogen, but not
following exposure to Ti02 orNiO. In vitro studies indicate that release of IL-lp and TNF-a by Type II
cells occurred only following exposure to crocidolite or ultrafine TiO2, but not pigment grade TiO2. The
authors also indicate that protein C and Clara cell secretory protein were both expressed in their
respective source cells following exposure only to the fibrogenic of the above particles. They also report
that crystalline silica has been shown to promote cytokine release and hypertrophy in Type II cells.
Jagirdar et al. (1997) used immunohistochemistry in a study to show that all three isoforms of TGF-0
(1,2, and 3) are expressed in the fibrotic lesions of asbestosis and pleural fibrosis patients from the
Quebec mines, primarily by Type II pneumocytes. The cases examined averaged 38 years of exposure to
the Quebec chrysotile. The authors also indicate that the hyperplastic epithelium of silicosis patients also
show elevated expression of all three isoforms. They further indicate from previous studies that
mesothelioma tumor cells frequently express TGF-p2 while the cells in the stroma of such tumors
frequently express TGF-(J1. It is also noted in this study that jun and fos are both transcription factors
that activate the TGF-pl promoter.
Zhang et al. (1993) indicate that macrophages obtained in BAL fluid from idiopathic pulmonary fibrosis
(IFF) patients and asbestosis patients show significantly increased secretion of TNF-a and asbestosis
patients also showed significantly increased secretion of IL-lp. Macrophages and monocytes obtained
from both kinds of patients also show elevated expression of mRNA for these cytokines. In an in vitro
part of this study, Zhang and coworkers, showed that chrysotile, crocidolite, amosite, and crystalline silica
all stimulated IL-lp and TNF-a release and up-regulated their respective mRNA in both macrophages and
monocytes. The authors also report that these two cytokines have been shown to up-regulate collagen
Types I and III and fibronectin gene expression in human diploid lung fibroblasts after short-term, serum
free exposure in vitro.
Holian et al. (1997) exposed cultures of normal human alveolar macrophages (AM) (obtained by lavage)
to varying concentrations (up to 25 ug/ml) of short chrysotile, UICC crocidolite, ground silica,
wollastonite, and titanium dioxide to determine whether these materials cause a phenotypic shift in
macrophage populations by inducing selective apoptosis. The authors indicate that normal lungs contain:
40-50% RFD1+7+ suppressor AM, and 5-10% RFD1+ immune activator. In this study, the fibrogenic
subset of the particles tested (not wollastonite or titanium dioxide) increased the ratio of
activator/suppressor AM by a factor of 4 within a few hours and the effect was seen to increase with time.
The authors also note that fibrogenic particles decrease the abundance of RFD7+ AM (phagocytic), but
the consequences of this phenotypic shift are unclear.
The authors indicate that AM taken from fibrotic patients release a variety of proinflammatory mediators
capable of stimulating fibroblast proliferation and collagen synthesis. Even in the absence of evidence of
fibrosis, workers who have been heavily exposed to asbestos yield similarly activated AM. In contrast,
they also note that, in vitro studies in which AM are stimulated with fibrogenic particles, while such AM
are activated to release inflammatory cytokines, such releases are orders of magnitude less than that seen
from AM derived from fibrotic patients. The authors indicate that the apoptosis-driven phenotypic shift
in AM that is indicated by this study may explain the apparent discrepancy.
In a previously described study, Timblin et al. (1998a), see Section 6.3.4.1, showed that crocidolite
asbestos and several cation substituted erionites all stimulate c-jun and c-fos in rat pleural mesothelial
cells, but to varying degrees depending on fiber chemistry. The effect also appears to be size dependent
as crocidolite, but not its non-fibrous analog riebeckite induces the effect.
In general, the kinds of signaling cascades that are potentially stimulated by exposure to asbestos are
important due to their potential to contribute to the promotion of cancer. Such pathways, for example,
may mediate proliferation or may suppress apoptosis. Alternately, they may mediate an inflammatory
6.101
-------
response that in turn may lead to proliferation or to production and release of other, mutagenic agents
(e.g., ROS or RNS). Pathways that facilitate development of fibrosis may also contribute to cancer
promotion, given the apparent link between fibrosis and the development of lung cancer, which may
relate (among other possibilities) to inhibition of fiber clearance (Section 6.3.4.5).
The ERK-MAPK signaling pathway evaluated in multiple studies by Mossman and coworkers (Mossman
et al. 1997; Timblin et al. 1998a,b; Zanella et al. 1999) in rat pleural mesothelial (RPM) cells is a case in
point (see above). These studies suggest that crocidolite stimulates the ERK-MAPK cascade through
interaction with the EOF receptor. This ultimately leads to transcription of mRNA for c-fos. Crocidolite
has also been shown to induce c-jun (apparently through a separate mechanism) and the balance between
c-jun and c-fos has been implicated in guiding a cell toward either proliferation or apoptosis. Although
the direct connection between c-jun/c-fos and apoptosis has not been established, it is observed that
crocidolite induces substantial apoptosis in RPM cells at the same concentrations at which it induces
substantial expression of c-fos and c-jun. The link is also implied because inhibition of the ERK pathway
has been shown in some studies to inhibit asbestos-induced apoptosis. Na-erionite has also been shown to
induce c-fos at levels comparable or higher than crocidolite for comparable exposures (at least on a mass
basis) and induce c-jun at higher levels. However, it is not known whether Na-erionite and crocidolite act
via the same pathways. Potentially due to the increased, relative expression of c-jun induced by
Na-erioinite, increased apoptosis is not observed in association with exposure to Na-erionite. However,
the link between increased c-jun expression and inhibition of apoptosis has not been demonstrated
explicitly.
Both crocidolite and Na-erionite were also shown by Mossman and coworkers to induce uptake of
bromodeoxyuridine (BrdU) by RPM cells in these same experiments. Uptake of BrdU is an indicator of
DNA synthesis. Since it has also been shown that crocidolite is capable of damaging DNA via ROS and
other pathways (Sections 6.3.3) and both crocidolite and erionite are known to induce mesotheliomas in
any case, the balance between proliferation and apoptosis in this cell population that is struck by exposure
to these toxins may very well determine whether development of cancers are promoted or prevented. Of
course, both proliferation and apoptosis may also be mediated by other pathways independent of the ones
described here.
The problem is that the range of responses that are induced by asbestos in the lung are varied and
complex (see Table 6-5) so that it has not yet been possible to definitively identify the biochemical
triggers that lead to lung cancer or mesothelioma. It is even likely, for example, that different
mechanisms (or combinations of mechanisms) predominate under different exposure conditions or in
association with differing fiber types or particle sizes. Still, examination of the dependence of candidate
mechanisms on fiber type and particle size can be instructive, especially to the degree that such
indications are consistent with observations in whole animal studies (see, for example, Section 6.2.2). For
the signaling cascade described above by Mossman and coworkers, for example, the effects attributable to
crocidolite are clearly dependent on fiber size because the non-fibrous analog to crocidolite, riebeckite
does not induce any of the effects. It also appears that the chemistry of the fibers is important, given the
observed differences in responses among the various, substituted erionites.
6.3.4.3 Asbestos-induced apoptosis
Apoptosis (programmed cell death) is generally triggered when a cell accumulates certain types of genetic
damage, when cell signaling cascades are triggered by external stimuli that may occur, for example, as
part of the need to maintain tissue homeostasis or to cause a phenotypic shift in response to toxic
challenge (see, for example, Holian et al. 1997, Section 6.3.4.2), or when a cell has completed a pre-
programmed number of divisions. Asbestos can induce apoptosis in a variety of cells by several
mechanisms including primarily:
6.102
-------
• by causing sufficient genetic damage to trigger apoptosis; or
• by triggering a signal cascade that leads to apoptosis.
As previously indicated, asbestos may trigger signaling cascades by interacting directly with receptors on
the cell surface, or by inducing production of intermediate species (such as ROS or RNS) that may in turn
induce cell signaling.
Some of the mechanisms by which asbestos may act to induce apoptosis may be fiber size- or type-
dependent. Also, responses may vary in different target tissues.
Fibers. As indicated in Section 6.4.3.2, crocidolite (but not the non-fibrous analog riebeckite) induces
apoptosis in hamster tracheal epithelial cells and rat pleural mesothelial cells when applied at non-
cytotoxic concentrations (Mossman et al. 1997). Results from the study also indicate that apoptosis is
triggered in this case by inducing an ERK-M APK signaling cascade as a consequence of interaction with
EOF receptors on the cell surface. The interaction may be direct or may be caused by asbestos-induced
ROS. In a related study (also previously summarized, Section 6.3.4.1), Timblin et al. (1998a) indicate
that the asbestos-induced apoptosis reported in the Mossman et al. (1997) work is fiber-type specific and
the pathway involved appears to stimulate expression of c-fos.
In a study previously reported in greater detail (Section 6.3.4.1), Levresse et al. (1997) indicate that
chrysotile induces apoptosis in cultured rat pleural mesothelioma cells with the effect peaking at 4% at 72
hours following exposure to 10 ng/cm2. Although the authors observed a much smaller effect with
crocidolite, they indicate that the difference is likely due to the much smaller number of long fibers in the
particular crocidolite sample evaluated.
Broaddus et al. (1997) indicates that crocidolite (not UICC), but not wollastonite, glass beads, or non-
fibrous riebeckite cause substantial apoptosis in rabbit pleural mesothelioma cells in culture in a dose-
dependent fashion. The extent of apoptosis induced was inhibited by treatment with catalase and by
3-minobenzamide (an inhibitor of poly(ADP-ribosyl) polymerase. The former indicates a role for ROS
mediation and the latter indicates that this enzyme, which mediates DNA repair, also mediates asbestos-
induced apoptosis (perhaps triggered by asbestos-induced DNA damage). Asbestos induced apoptosis
was also inhibited by treatment with desfeoxamine, but effects were restored by adding iron to the
medium. The authors note that in other studies, crocidolite has been shown to induce DNA strand breaks
within 2 hours after exposure and induces unscheduled DNA synthesis within 24 hours following
exposure. Asbestos also induces production of poly(ADP-ribosyl) polymerase.
Non-fibrous particles. Leigh et al. (1997) intratracheally instilled rats with silica (a non-fibrous particle)
at doses varying between 2 and 22 mg. They then collected cells by bronchio-alveolar lavage (BAL) 10
days after instillation. The authors observed large numbers of apoptotic cells in BAL fluid and that the
number of such cells was dose-dependent. The dead cells were primarily neutrophils (so that this might
represent some type of mechanism to restore homeostasis). Engulfment of apoptotic cells by
macrophages was also observed. The authors report that, 56 days after instillation, apoptotic cells were
observed in granulomatous tissue within the lungs of rats exposed to silica. This suggests that apoptosis
may also occur in response to chronic inflammation. The authors conclude that silica induces apoptosis
among granulomatous cells and alveolar cells and that such apoptosis and the subsequent engulfment of
apoptotic cells by macrophages may play a role in the evolution of silica-related disease. The authors
also note that granuloma formation is a hyperplasia-related event.
At least some of the mechanisms suggested above for asbestos-induced apoptosis are dependent on fiber
size (the non-fibrous analog of crocidolite does not induce the effect) and dependent on the chemistry of
the fibers involved (various, cation-substituted analogs of erionite exhibit disparate ability to induce the
6.103
-------
effect). Although non-fibrous particles (such as crystalline silica) may also induce apoptosis, as
previously suggested, this may be through separate mechanisms from those responsible for asbestos-
related effects, even if the same endpoint results.
6.3.4.4 Asbestos-induced cytotoxicity
While there is ample evidence from various in vitro studies that asbestos is cytotoxic, such effects are
observed almost exclusively at the highest concentrations evaluated in an experiment (for example, Luster
and Simeonova [1998], Section 6.3.4.2 and Choe et al. [1999], Section 6.3.4.2). Many experiments are
conducted at concentrations below those for which cytotoxicity is important because the other toxic
effects attributable to asbestos occur at substantially lower exposure levels and researchers prefer to study
such effects in the absence of potentially confounding cytotoxicity. For in vitro studies, for example,
non-cyotoxic effects are typically studied at concentrations less than approximately 10 ug/cm2 (or
20 ug/ml) while substantial cytotoxicity is not typically observed until exposure concentrations are
several times higher.
Because most of the other effects attributable to asbestos occur at concentrations that are substantially
lower, this begs the question as to whether frank cytotoxicity is an effect that is relevant to human
exposures. There is also substantial evidence that the mechanisms associated with asbestos-induced
cytotoxicity are separate from the mechanisms that mediate most of the other asbestos-related effects of
interest.
Kamp et al. (1993) dosed cultured pulmonary epithelial (PE) cells with UICC amosite asbestos. In some
studies, polymorphonuclear leukocytes (PMN) were also added to the culture. Typical doses in this
experiment were on the order of 250 ug/cm2, which is quite high for these types of studies. For example,
compare this level with the levels reported for studies of ROS/RNS generation (Section 6.3.3), cell
signaling (Section 6.3.4.2), proliferation (Section 6.3.4.1) or apoptosis (Section 6.3.4.3).
Kamp and coworkers indicate that the effect of amosite exposure on cultured PE cells (at the
concentrations studied) was to induce substantial cell lysis (cytotoxicity) and little cell detachment (from
the culture medium), which would indicate increased cell motility. Addition of PMN to the culture
resulted in both increased cell lysis and cell detachment for comparable exposures to amosite. The
observed cell detachment was mitigated in a dose-dependent fashion by adding protease inhibitors.
Further work indicated that asbestos induces release of human neutrophil elastase (HNE), which may
mediate the combined effects with PMN. PE cell exposure to HNE alone causes increases in cell
detachment in a dose-dependent fashion. However, when combined with asbestos exposure, cell lysis
increases at the expense of cell detachment. The authors suggest that HNE becomes bound to asbestos,
which also becomes bound to PE cells and this facilitates augmented cytotoxicity by proteases that are
secreted by PMN's.
Blake et al. (1998) studied the effect of fiber size on the cytotoxicity of alveolar macrophages in vitro.
Cultured cells were dosed with concentrations varying between 0 and 500 ug/ml of each of 5 different
length preparations of JM 100 glass fiber. Cytotoxicity was monitored by assays for extracellular LDH
and by chemiluminescence following zymosan addition. The latter assy is intended to show macrophage
stimulation. Results indicate that all samples showed dose-dependent increases in toxicity (i.e.,
increasing LDH and decreasing chemiluminescence). Comparing across samples, relatively long fibers
(mean=17 urn) showed the greatest toxicity. The authors further indicate that microscopic examination
suggests that frustrated phagocytosis plays a role in cytotoxicity.
Goodglick and Kane (1990) studied the effect of three different length preparations of crocidolite (long,
short, and UICC) on elicited macrophages (stimulated initially with thioglycolate) in vitro and in vivo.
6.104
-------
The long and short samples were reportedly prepared from the UICC sample by repeated centrifugation.
Goodglick and Kane (1990) report that all three types of crocidolite stimulated release of ROS from
macrophages. At sufficiently high concentrations, all three also caused substantial cytotoxicity, although
apparently due to the longer time required to settle in culture, the full effects from short fibers take longer
to develop. They suggest that, on a total fiber number or surface area basis, long and short crocidolite
appear to exhibit approximately equal potency toward the cytotoxicity of macrophages. Further work
with various inhibitors indicates that cytotoxicity is mediated by production of ROS and that ROS are
produced via an iron-dependent pathway. They also indicate that, among the effects of crocidolite
exposure is that macrophage mitochondrial membranes are depolarized.
Goodglick and Kane (1990) also evaluated the effects of long and short crocidolite in vivo. This was
done by evaluating the effects of intraperitoneal injection of the various samples (long, short, or mixed
crocidolite or titanium dioxide particles) in C57B1/6 mice. Results indicate that a single injection of long
crocidolite (480 jig) induced an intense inflammatory response, leakage of albumin, and fibers observed
scattered across the diaphragm. In contrast, a single injection of short crocidolite (600 |ig) induced only a
relatively mild inflammatory response and only limited clusters of fibers observed on the surface of the
diaphragm.
To test whether short fibers would show a greater response, if they were not cleared more readily than
long fibers, Goodglick and Kane (1990) also subjected mice to 5 consecutive, daily injections of 120 ng
of short crocidolite and noted more substantial aggregations of fiber clusters along the diaphragm as well
as a more pronounced inflammatory response. Cell injury was also assessed by Trypan blue staining
(which indicates cell death). All mice singly or multiply injected with mixed or long crocidolite showed
marked Trypan blue staining. Single injections of short structures showed only limited Trypan blue
staining. However, following 5 daily injections of short fibers, multiple Trypan blue stained cells were
observed on the diaphragm in the vicinity of the locations were clusters of fibers were also observed. The
authors also indicate (in contrast) that neither single injections of 160 or 800 ug nor 5 consecutive (160
ug) injections of titanium dioxide produced any Trypan blue staining.
The authors conclude from this study that both short and long crocidolite fibers appear to be cytotoxic to
macrophages while titanium dioxide particles are not (suggesting that not only fiber length, but fiber type
is important to cytotoxicity). They further suggest that, while short fibers tend to be cleared rapidly in
vivo, when such clearance mechanisms are overwhelmed (such as by repeated insult through repeated,
daily injections in this study), then the toxic effects of short structures becomes apparent. As indicated in
other studies, however, there may be multiple mechanisms working to produce similar responses, that
such mechanisms may exhibit varying dose-response characteristics, and that cytotoxicity may not
generally be directly related to mechanisms that contribute to carcinogenesis. Moreover, there almost
certainly are at least some longer fibers in the short fiber preparation and extended analysis to determine
their relative concentration with adequate precision would be helpful to see if the relative magnitude of
the observed effects correlate.
Palekar et al. (1979) studied the ability of four different samples of commingtonite-grunerite, each also
subjected to varying degrees of grinding, to induce hemolysis of mammalian erythrocytes and
cytotoxicity to Chinese hamster ovary (CHO) cells. The samples studied include: UICC amosite
(4.13 m2/g surface area/mass), which is denoted as "asbestiform grunerite; "semi-asbestiform"
commingtonite (3.88 m2/g); acicular commingtonite (ground to three particle sizes: 3.76, 2.45, and 0.82
m2/g), and acicular grunerite (2.82 m2/g).
Results from this study indicate that amosite induced the greatest hemolysis of erythrocytes by far
(approximately 50%) while acicular, unground grunerite caused no hemolysis. However, grinding the
acicular grunerite to increasingly smaller particle sizes and greater surface area ultimately results in some
hemolysis. Both semi-asbestiform and acicular, unground commingtonite show hemolytic activity
6.105
-------
between amosite and unground, acicular grunerite and grinding acicular commingtonite also increased its
hemolytic activity.
Similar results were also observed for cytotoxicity. Amosite was by far the most cytotoxic and the effect
was dose-dependent. A dose of 0.05 mg/ml caused approximately 75% cell death for CHO cells. At 0.2
mg/ml, only 1% of cells survived. Acicular grunerite was nontoxic even at 0.5 mg/ml. With grinding,
acicular grunerite cytotoxicity increased, albeit only slowly. The most heavily ground sample killed
fewer than 25% of cells at 0.2 mg/ml and killed only 65% at 0.5 mg/ml. The cytotoxicity of semi-
asbestiform commingtonite was substantially less than amosite, but greater even than ground, acicular
grunerite. For this material, 0.2 mg/ml killed approximately 65% of cells and 0.5 mg/ml killed
approximately 90%. Interestingly, approximately the same dose-response curve for cytotoxicity was
observed for the 3.88 and 1.61 specific surface area samples of this material. The 1.21 samples was
somewhat less cytotoxic. Acicular commingtonite was somewhat less cytotoxic (for corresponding
doses) than semiasbestiform commingtonite at the highest specific surface area (3.76) and its toxicity
decreased with decreasing surface area.
The authors also indicate that neither surface charges on crystal particles nor Magnesium ion content
appear to correlate with biological activity. The authors conclude that the degree of "asbestiform"
character of a mineral has a dominant effect on biological activity. Moreover, although non-fibrous
particles may also be biologically active (and their activity increases with increasing specific surface
area), the effects of particles and fibers lie along entirely separate dose-response curves. The biological
activity of fibrous materials does not appear to depend directly on specific surface area.
Importantly, the results of the Palekar et al. (1979) study are also consistent with the possibility that
fibrous structures within a specific range of sizes and shapes contribute strongly to biological activity
while largely non-fibrous particles act through a separate mechanism that depends primarily on total
surface area, but that particle-for-particle elicits a substantially lower overall response than the
mechanism by which fibers act. Such a scenario is supported by several studies. Jaurand (1997), for
example, indicate that ROS are implicated in the cytotoxicity of long, but not short fibers on tracheal
epithelial cells. Although the evidence for distinct mechanisms for fibers and particles discussed here is
specific to cytotoxic and hemolytic effects, evidence in other studies suggest similar scenarios for other
toxic endpoints (potentially including endpoints that contribute to carcinogenicity).
Comparisons of the rate and extent of effects observed in epidemiology studies, whole animal dose-
response studies, and in vitro studies suggests that cytotoxicity may not be important to human exposures.
Unfortunately, however, there is currently insufficient information to compare doses and exposures across
these studies in a more quantitative fashion. Therefore, the importance of cytotoxicity to human asbestos
exposure cannot be definitively determined at this time.
6.3.4.5 Association between fibrosis and carcinogenicity
The hypothesis that lung tumor induction is associated with the fibrosis has been examined by several
authors. There appears to be a debate as to whether fibrosis is a necessary precursor for development of
lung cancer (associated with exposure to fibers), whether the presence of fibrosis is an additional factor
contributing to increased risk for lung cancer, or whether the two diseases are largely unrelated. This is
an important consideration because the characteristics of the exposure-response relationship between
asbestos and lung cancer or asbestosis (fibrosis) apparently differ (Sections 6.3.6 and 6.4).
Based on animal studies, Davis and Cowie (1990) found that rats that developed pulmonary tumors
during inhalation experiments exhibited a significantly greater clinical degree of fibrosis than rats that did
not develop tumors. Furthermore, Davis and Cowie (1990) reported suggestive evidence that the
6.106
-------
pulmonary tumors that did develop in the dosed rats tended to develop within portions of the rat's lungs
that were- alr^arlv «r.arrpH hv fiVirn«i
-------
• smoke-product induced inhibition of clearance of asbestos fibers and/or asbestos induced
inhibition of clearance of smoke products (see, for example, Mossman et al. 1996); and
• smoke-product induced facilitation of uptake of asbestos by lung epithelium (see, for
example, Mossman et al. 1996).
Although much progress has been made at elucidating the nature of these mechanisms and other candidate
mechanisms, at this point in time it is possible to indicate definitively neither what mechanisms are
important to any observed interaction between smoking and asbestos exposure nor to indicate the relative
magnitude of the contributions from such mechanisms. Moreover, a detailed review of such mechanisms
is beyond the scope of this document.
6.3.4.7 Conclusions concerning asbestos as a promoter
There is strong evidence that asbestos acts as a promoter for cancer. While this may primarily involve
mechanisms the contribute to the induction of proliferation, mechanisms associated with the an
interaction between smoking and exposure to asbestos (for the induction of lung cancer; smoking does not
appear to affect mesothelioma) are also important. There also appears to be an association between the
development of fibrosis and an increased risk for lung cancer.
Evidence indicates that multiple mechanisms may be involved with asbestos-induced cancer promotion
and that such mechanisms may be complex and interacting. The different mechanisms also appear to
exhibit dose-response relationships with differing characteristics. While there are indications that the
most important among these mechanisms may be strong functions of fiber size (with long fibers
contributing most to the induction of disease), mechanisms that depend primarily on surface area or total
fiber (particle) number (for any size range) may also contribute to overall cancer promotion. Importantly,
these latter mechanisms also appear to be strongly associated with the composition of fibers (particles)
and may therefore contribute more substantially to the disease induction of agents that have been shown
to be particularly toxic (such as crystalline silica), as opposed to particles, fibers, or asbestos in general.
At this point in time, the available data may not be sufficient to distinguish among the relative
contributions from the various mechanisms to the overall promotion of cancer, at least in terms of the
mechanistic data itself. Importantly, however, the mechanistic data should not be considered to be
inconsistent with the results from whole animal studies, where there are clearer indications that fiber size
plays a major role in carcinogenicity and fiber (particle) type is also important (see Sections 6.2 and 6.4).
Such studies indicate, for asbestos (and other biodurable fibers) that:
• short fibers (less than somewhere between 5 and 10 urn) do not appear to contribute to
disease;
• potency likely increases regularly for fibers between 10 urn and a minimum of 20 |im
(and, perhaps, continues to increase up to lengths of at least 40 urn); and
• fiber type may be important primarily in determining biodurability.
They further indicate that particularly (or uniquely) toxic particles (such as crystalline silica) may act
through a different set of mechanisms that are not dependent on fiber length, but that induce toxic
endpoints paralleling those observed for asbestos.
Importantly, the mechanisms by which asbestos may act as a promoter appear to occur in cell lines that
may contribute both to the induction of lung cancer and mesothelioma.
6.108
-------
6.3.5 Evidence that Asbestos Induces an Inflammatory Response
There is ample evidence that asbestos induces an inflammatory response in pulmonary tissues and the
pleura (see, for example, Sections 6.2.2 and 6.3). Moreover, there appears to be multiple biochemical
triggers that mediate this response and various mechanisms may be fiber size- and/or fiber type-specific
(Table 6-5). Because the role that inflammation plays in the induction of cancer has been addressed
elsewhere (Sections 6.3.3 and 6.3.4), it is beyond the scope of this document to provide a detailed review
of the mechanisms that lead specifically to inflammation.
6.3.6 Evidence that Asbestos Induces Fibrosis
There is ample evidence that asbestos induces fibrosis in pulmonary tissues (see, for example, Sections
6.2.2 and 6.3). Moreover, there appears to be multiple biochemical triggers that mediate this response
and various mechanisms may be fiber size- and/or fiber type-specific (Table 6-5). Because the role that
fibrosis plays in the induction of cancer has been addressed elsewhere (Sections 6.3.3 and 6.3.4), it is
beyond the scope of this document to provide a detailed review of the mechanisms that lead specifically
to fibrosis. Such mechanisms have also been the subject of recent reviews (see, for example, Mossman
and Churg 1998; Robledo and Mossman 1999).
6.3.7 Evidence that Asbestos Mediates Changes in Epithelial Permeability
As previously indicated (Section 4.4), maintaining the overall integrity of the epithelial surface of the
lung is among the various functions of Type II epithelial cells (Leikauf and Driscoll 1993). It has been
shown that asbestos induces changes in the morphology of Type II epithelial cells (see, for example,
Ilgren and Chatfield 1998), which has the effect (among others) of increasing the overall permeability of
lung epithelial tissue to various macromolecules and, potentially to asbestos fibers themselves. The
former plays a role in asbestos-induced fibrosis (by allowing cytokines that stimulate fibroblast
proliferation or stimulate fibroblasts to generate extracellular matrix to pass through the epithelium and
reach the underlying fibroblasts, Section 6.3.6). The latter may be important to facilitating transport of
asbestos from the alveolar lumen to the interstitium (see, for example, Lippmann 1994).
Changes in epithelial permeability may be triggered by cytokines released from other cells or by the
action of asbestos fibers on epithelial cells directly. Moreover, some of the mechanisms that mediate this
response may be sensitive to fiber size and/or fiber type. For example, Gross et al. (1994) showed that
monolayers of human bronchial epithelial cells cultured over a porous medium and exposed to
cryogenically ground chrysotile (average length: 1 urn, average aspect ratio: 14 at 15 jig/culture plate)
became permeable to fibrin breakdown products (FBP's). The cultures were grown over human serum
with labeled fibrinogen. This was based on observed increased concentrations of FBP's (double in 24
hours) in the ablumenal chambers of exposed cells compared to cells in control cultures. Because the
epithelium showed greater permeability to all concentrations, the increased concentrations were not due to
increased breakdown. The observed FDP flux was not vectoral, not saturable, and required neither
proteolytic processing nor active transport. Thus, asbestos increases the paracellular flux of intact FDP
across airway epithelium.
6.3.8 Conclusions Regarding the Biochemical Mechanisms of Asbestos-Related Diseases
That the specific biochemical triggers for asbestos-related diseases (particularly, the asbestos-related
cancers) have not been definitively delineated as of yet is not surprising. The detailed interactions
between fibers (and particles) and the cells and tissues of the lung are complex and there are complex,
multiple, interacting mechanisms by which such interactions may contribute to disease. Despite great
progress in elucidating candidate mechanisms, the number of candidate mechanisms is large and
6.109
-------
distinguishing among their relative contributions has been difficult. This is because, among other things,
the ability to compare results across studies of different mechanisms is currently limited due to the
inability to reconcile the quantitative effects of dose and response across dissimilar studies.
Nevertheless, a number of important implications can be gleaned from the available literature. First, it
appears that asbestos can function both as a cancer initiator and a promoter. It also appears that both the
initiation and promotion of cancer may occur through more than one mechanism.
Regarding cancer initiation, asbestos likely acts primarily through a mechanism involving interference
with mitosis. By this mechanism, asbestos fibers are phagocytized by target cells, migrate to perinuclear
locations, and interact with the spindle apparatus and other cell assemblages required to complete mitosis.
This tends to result in aneuploidy and may cause various clastogenic effects. This mechanism is driven
by long fibers; short fibers do not appear to contribute to the effect. It also appears that all asbestos fiber
types (and potentially other durable fibers with sufficient dimensions) cause genetic damage via this
mechanisms. If there are effects due to fiber type, they appear only to play a secondary role.
Although there is also evidence that asbestos may induce production of DNA adducts and DNA strand
breaks (through ROS and RNS mediated pathways), whether such adducts or breaks ultimately lead to
permanent, heritable changes to DNA remain to be demonstrated. The relative importance of ROS/RNS
mediated pathways for initiating cancer, compared to the pathway involving interference with mitosis,
also remains to be determined.
There is also some evidence that the relative importance of asbestos as a cancer initiator may differ in
different tissues. Lung epithelial cells, for example, appear to be relatively resistant to the mechanisms by
which asbestos may initiate cancer. Mesothelial cells are not. Among several possibilities, this may be
due to the ability of proliferation-competent lung epithelial cells (Type II cells) to undergo terminal
differentiation when challenged with certain toxins and this is a pathway not available to mesothelial
cells.
The mechanisms by which asbestos may promote cancer primarily involve mechanisms that contribute to
the induction of proliferation, although mechanisms associated with an interaction between smoking and
exposure to asbestos to induce lung cancer are also important. There also appears to be an association
between the development of fibrosis (including asbestosis) and an increased risk of lung cancer.
Evidence indicates that multiple mechanisms may be involved with asbestos-induced cancer promotion
and that such mechanisms may be complex and interacting. The different mechanisms also appear to
exhibit dose-response relationships with differing characteristics. While there are indications that the
most important of these may be strong functions of fiber size (with long fibers contributing the most to
carcinogenicity), mechanisms that depend primarily on surface area or total fiber number (for any size
range) may also contribute to overall cancer promotion. These latter mechanisms also appear to be
strongly associated with the composition of fibers and may therefore contribute more substantially to the
disease induction of agents that have been shown to be particularly toxic, as opposed to particles, fibers,
or asbestos in general.
Although crystalline silica may act to produce some of the same effects as asbestos (including
carcinogenicity), there is substantial evidence that this family of materials do not act through the same
pathways and that the characteristics of their respective dose-response relationships may differ. Thus, for
example, while asbestos likely induces cancer through mechanisms that favor long (and potentially thin)
fibers, silica more likely acts through a mechanism that is dependent on total surface area, with freshly
and finely ground material likely being the most potent. In contrast, grinding asbestos fibers tends to
lesson its carcinogenicity overall. Due to differences in chemistry and crystallinity (reinforced by studies
indicating a lack of correspondence in behavior), crystalline silica does not appear to be an appropriate
6.110
-------
analog for any of the asbestos fiber types. Rather, for example, the appropriate non-fibrous analog for
crocidolite is riebeckite and the appropriate non-fibrous analog for chrysotile is antigorite or lizardite.
6.4 ANIMAL DOSE RESPONSE STUDIES
Ideally, human epidemiology studies (reviewed in Chapter 7) provide the best data from which to judge
the effects of asbestos in humans and from which to derive exposure-response factors for humans.
However, animal dose-response studies have proven useful for elucidating certain features of the
relationship between asbestos dose and response that cannot be adequately explored in the human studies,
primarily due to limitations in the manner that exposures were characterized in the human studies (see
Chapter 5).
Unlike human epidemiology studies, exposures in animal studies are controlled and better quantified.
Frequently, the characteristics (in terms of fiber size, shape, and type) of such exposures have also been
better quantified and this has allowed exploration of the effects that such characteristics (fiber size, shape,
type) have on disease response. Accordingly, an overview of animal dose-response studies is provided in
this section. Both injection-implantation studies and inhalation studies are reviewed. Particular attention
is also focused on a "supplemental" animal inhalation study that we conducted with the specific aim of
identifying the characteristics of asbestos that best relate to risk. The strengths and limitations of these
kinds of studies are described in Chapter 5.
6.4.1 Injection-Implantation Studies
Because the fibrous materials in injection and implantation studies are placed immediately against the
target tissue, the effects of processes associated with inhalation, retention, and translocation are avoided.
The only active mechanisms that need to be considered in these studies are those that occur directly in the
target tissue (including degradation, clearance, and biological responses of the types described in the
previous sections of this chapter). Fibrous materials placed against the tissue surface are subject to
dissolution, phagocytosis by macrophages, and phagocytosis by the cells of the target tissue. These
mechanisms are described in greater detail in Section 6.2. A range of biologic responses have also been
observed (described in Section 6.3).
Numerous researchers have performed these types of studies.
The Work of Stanton and Coworkers. In a series of studies, Stanton and coworkers (1972, 1977, 1981)
implanted fibrous materials and induced mesotheliomas in rats. In the studies, a pledgette composed of
coarse glass is loaded with hardened gelatin containing sample material and is surgically implanted
immediately against the left pleura of the rats. Control studies demonstrate that the coarse glass of the
pledgette does not induce significant tumors in the absence of other tumorigenic agents in the gelatin.
Although the mass dose of material implanted was the same for all experiments (40 mg), the observed
incidence of mesothelioma varied among samples. By characterizing the dimensions of fibrous structures
in the samples using a microscope, the researchers were able to explore the relationship between fiber size
and the incidence of mesothelioma. By studying a wide range of fibrous materials, Stanton and his
coworkers concluded that the induction of mesothelioma is determined primarily by the physical
dimensions of fibers and that mineral composition is secondary. Further, potency appears to increase
with the length and decrease with the diameter of fibrous structures. The researchers also concluded that
the incidence of malignant tumors correlates with the degree of fibrosis induced by the presence of the
fibrous materials. This does not necessarily imply, however, that fibrosis is a necessary step in the
induction of asbestos-induced tumors (see Section 6.3.4.5).
6.111
-------
Conclusions from the Stanton et al. (1972, 1977, 1981) studies indicating that mineralogy is not a factor
in biological response conflicts with evidence provided in Chapter 7 and implications gleaned from
mechanism studies presented in Section 6.3. However, the studies by Stanton and coworkers have been
shown to suffer from certain methodological limitations (Berman et al. 1995) so that results from these
studies should be considered more qualitative than quantitative.
Due to limitations in the ability to produce samples composed of uniform fibers, quantitative relationships
between size and potency were explored by Stanton and coworkers using a regression analysis.
Structures longer than 8 (im with diameters less than 0.25 urn or longer structures with diameters less than
1.5 (im were found to represent the range of sizes that best correlate with carcinogenicity. It was further
stated that such correlations did not eliminate the possibility that other size ranges also contribute to
potency, only that the two size ranges identified appear to correlate best. Samples that varied
significantly from the reported correlations were attributed to errors in the characterization of structure
size distributions in those samples. However, other methodological limitations might also have
contributed to the observed deviations or such "outliers" may also suggest evidence for a mineralogical
effect that is similar to what is reported in other studies (see Section 6.3 and Berman et al. 1995).
The precision of estimates for the ranges of sizes that contribute to biological activity that are derived
from the Stanton and coworkers (1972, 1977, 1981) studies is limited so that such estimates should also
be considered qualitative. Size distributions were determined by characterizing 200 to 1,000 structures
using TEM and there is no indication that statistically balanced counting rules were employed (Section
4.3). Under such conditions, counts of structures longer than 8 [im are likely small and subject to large
uncertainties for most of the samples characterized. Confidence intervals are not provided for any of the
exposure values presented in these studies.
Potentially larger errors in the studies by Stanton and coworkers could have been introduced by the
method employed to relate fiber counts to sample mass. As indicated in Chapter 5, estimating
contributions to mass by sizing total particles and assuming that this is proportional to total sample mass
is subject to error from the limit to the precision of characterizing structure dimensions (particularly
diameter) and by not accounting for nonasbestos (and possibly nonfibrous) material in the samples. Thus,
for example, there is no discussion of the precision with which the cut point of 8 urn was determined in
these studies.
Re-analysis and Extension of the Stanton Studies. Several researchers have re-evaluated data from the
implantation studies to test additional hypotheses. Using the Stanton and coworkers (1972, 1977, 1981)
data, Bertrand and Pezerat (1980) examined the relationship between mesothelioma incidence and several
characteristics not evaluated by Stanton and coworkers including: average fiber length, average fiber
diameter, average fiber aspect ratio, total fiber surface area and total fiber volume. Results from the
regression analysis indicate that potency varies directly with average length and inversely with average
diameter, but that neither parameter is a good indicator alone. Combining the effects of length and
diameter, average aspect ratio is highly correlated with potency. Biological activity does not correlate
highly with structure count, surface area or volume except when fiber sizes are restricted to the long, thin
structures that Stanton and coworkers defined. Results of this study are not inconsistent with those
originally presented by Stanton and coworkers except that they emphasize a set of characteristics that
relate parametrically to biological activity rather than expressing exposure as a single restricted size range
of structures.
Bertrand and Pezerat (1980) were able to find good correlations between response and specific "average"
characteristics of the samples that are not proportional to the quantity of the material present in the sample
("intensive" characteristics). Such intensive characteristics as average aspect ratio, average length, or
average diameter are properties that are independent of the mass of material in a sample. Since response
6.112
-------
must be a function of the quantity of sample present, intensive characteristics should have to be multiplied
by characteristics that are proportional to the mass of a sample (e.g., fiber number, sample mass, or
sample volume) in order to relate them to response. Properties that vary with the mass of a sample are
termed "extensive" properties.
The correlations between intensive properties and response reported by Bertrand and Pezerat (1980)
likely succeed within the Stanton and coworkers (1972, 1977, 1981) database because a constant sample
mass (40 mg) was employed for all of the implantation experiments. However, to apply dose-response
relationships that are dependent only on intensive characteristics beyond the data presented by Stanton
and coworkers (where mass dose will not be constant), it is necessary to pair intensive characteristics with
extensive characteristics (such as mass or number of fibers per sample). Therefore, it is unclear how the
conclusions from this paper may be generalized to other data sets.
In a similar study, Bonneau et al. (1986) also examined parametric relationships between structure
characteristics and mesothelioma induction. The paper examined specifically correlations between
carcinogenicity and dose in terms of two specific relationships: dose expressed as fibers longer than 8 urn
that are thinner than 0.25 um ("Stanton" fibers) and dose expressed as mean aspect ratio. The researchers
conclude that mean aspect ratio provides an excellent indication of carcinogenicity for individual fiber
types, but that each fiber type must be treated separately. Poorer correlations are found for the
relationship between the concentration of "Stanton" fibers and mesothelioma, even when fiber types are
considered independently. Although these results appear to be consistent with findings reported from
mechanistic studies (Sections 6.2 and 6.3) in that they posit a role for fiber mineralogy, the relationships
evaluated by Bonneau et al. (1986) also suffer from the limitation of expressing dose only in terms of
intensive quantities, as discussed above. Direct comparison with other studies is therefore difficult.
Following up on the reported problems of the studies by Stanton and coworkers in characterizing
crocidolite, Wylie et al. (1987) reanalyzed seven crocidolite samples originally studied by Stanton et al.
She and coworkers then used the new size distributions to reevaluate the "Stanton hypothesis" (that the
concentration of "Stanton" fibers in a sample correlates with carcinogenicity). Wylie and coworkers note
that substantial deviations from the Stanton hypothesis occur for specific samples. They conclude that a
specific structure size range alone is not sufficient to characterize biological activity and that a parametric
relationship with other structure characteristics (potentially including mineral type) may be necessary to
sufficiently describe biological activity.
Conclusions from the Wylie et al. (1987) paper must be interpreted carefully because the researchers
evaluated only the relationship between carcinogenicity and the single specific size range indicated
("Stanton" fibers). Thus, the possibility that improved correlations exist between biological activity and
different size ranges or a combination of size ranges cannot be ruled out. Qualitatively, conclusions
presented in this paper are not inconsistent with the conclusions reported by Stanton and coworkers
regarding the general relationship between response and fiber dimensions.
The Wylie et al. (1987) study appears to suffer from several methodological problems. These relate to the
manner in which the sample reanalysis was performed. The drop method for preparing electron
microscopy grids (used in this study) is not satisfactory for preparing grids. In fact, as reported in the
study itself, grids prepared as duplicates by this method were shown to be non-uniform at the 95%
confidence interval using a chi-square test. In addition, only 100 to 300 fibers were counted for each
sample. Since there is no indication that statistically balanced counting was performed, the uncertainty
associated with counts of "Stanton" fibers may be substantial. Such errors would be further multiplied by
uncertainty introduced during the sizing of total particles to determine the number of fibers per unit mass.
In a later study, Wylie et al. (1993) examined the effect of width on fiber potency. In this latter study,
results from animal injection and implantation studies were pooled and subjected to regression analyses to
6.113
-------
identify correlations between exposure and tumor incidence. The animal studies selected for inclusion in
this analysis were performed on a variety of tremolite samples exhibiting a range of morphological and
dimensional characteristics.
In their regression analyses, Wylie et al. (1993) evaluated a range of exposure indices that emphasize
different morphological or size characteristics to help elucidate the characteristics of asbestos that induce
a biological response. Because all of the animal studies included in their analyses involved tremolite,
mineralogy was not an issue. Results from this study suggest that fibers longer than 5 um and thinner
than 1 (im best correlate with tumor incidence among the animal injection and implantation studies
examined. Further, they suggest that a width limit, rather than a limit on aspect ratio, better reflects the
bounds of the asbestos characteristics that determine biological activity. They also suggest that complex
structures (bundles and clusters) need to be evaluated as part of the determination of exposure because
such structures can breakdown and contribute to the population of thinner fibers.
Although the results of the Wylie et al. study tend to support the general conclusions in this document
related to width, if not to length (see Section 6.5 and Appendix A), as the authors themselves indicate,
such results should be considered qualitative due to the limitations imposed on their study by the
methodology employed. Their study was conducted by:
• combining results from multiple studies without careful consideration of variation
introduced by methodological differences across the studies;
• employing asbestos concentrations determined by SEM and without careful consideration
of differences in the counting methodologies employed by differing research groups
across studies; and
• considering injection and implantation studies, which do not account for mechanisms
related to inhalation and deposition that affect the exposure-response relationship in
humans.
The limitations imposed by the above constraints are highlighted in Chapters 4 and 5 of this report.
Other Injection Studies. A series of injection studies were conducted by several research groups. In
these studies, fibrous materials were suspended in saline and injected into rats immediately adjacent either
to the pleura or peritoneum. A large number of fibrous materials have now been studied by this process,
as reported by: Bolton et al. (1982, 1984, 1986); Davis et al. (1985, 1986a,d, 1987, 1988a); Muhle et al.
(1987); Pott et al. (1974, 1976, 1978, 1982, 1987); and Wagner et al. (1976, 1982, 1985). Newer studies
are also discussed in Section 6.2. Results confirm that it is the fibrous nature of the materials that is the
primary factor leading to the induction of tumors and that potency appears to depend directly on length
and inversely on diameter.
The authors of these studies tend to indicate that, except where fibers are not persistent in vivo, due to
solubility or other degradation processes, the mineralogy of the fibers appears to play only a secondary
role in determining disease incidence. Researchers conducting injection experiments also tended to report
a correlation between tumor incidence and the degree of fibrosis induced by the sample. These
observations are consistent with the ideas originally articulated by Stanton.
Pott developed Stanton's ideas further by suggesting that carcinogenicity is a continuous function of fiber
dimensions, which decreases rapidly for lengths less than 10 |im and also decreases with increasing
diameter. The possibility was also raised that the apparent inverse dependence on diameter may be an
artifact due to the limited number of thick fibers that can be injected in a sample of fixed mass.
6.114
-------
Although the published injection studies indicate that potency decreases with decreasing length,
researchers in these earlier studies were reluctant to identify a length below which contributions to
carcinogenicity can be considered inconsequential. This may be due in part to the skewed distribution of
fiber sizes typical of asbestos dusts. Thus, for example, even if structures less than 5 urn are only 1% as
potent as structures longer than 5 urn, they may be as much as 100 times as plentiful in some asbestos
dusts, so that the total contribution to potency would be equal for both size fractions.
Reasonable dose-response curves have been generated using various sample masses of a single material in
some of these studies. This has been demonstrated for UICC crocidolite and UICC chrysotile "A"
(Bolton 1984). Results indicate that the relationship between tumor incidence and the log of the dose
may be linear and there is no effective threshold. A consistent difference between the two dusts is
apparent; the points lie along separate curves and chrysotile appears to be more potent per unit sample
mass.
In general, the analytical techniques used for quantifying size distributions in these studies are not fully
documented. To the extent that they are, it appears that similar approaches were adopted to those
described for the implantation studies above. Consequently, similar limitations apply to the interpretation
of results. Briefly, large uncertainties are likely associated with counts of long fibers and estimates of the
number of fibers per unit sample mass. Counts in several of the studies also suffer from limitations in the
ability of SEM or PCM to detect thin^fibers (Section 4.3); whenever SEM or PCM was employed, the
thinner fibers were likely under-represented in reported fiber size distributions.
Because samples are placed against mesenchyme in the published implantation and injection studies,
results of these studies most directly represent processes associated with the induction of mesothelioma.
Assuming, however, that clearance and degradation processes are similar in the deep lung, once a fiber
reaches a target tissue, results from the implantation and injection studies may also provide a model of
biological response in lung tissue and the factors that lead to the induction of pulmonary tumors. Such a
model must be considered qualitative at best, however, because it has been shown that the mechanisms of
tissue response to the presence of asbestos in lung parenchyma and in the mesenchyme differ in detail
(Section 6.3). The time periods over which the various clearance mechanisms operate in the deep lung
and the mesenchyme also differ (Section 6.2), although, it is apparent that the general nature of the
clearance and degradation processes in the two tissue types are generally similar.
6.4.2 Animal Inhalation Studies
Animal inhalation studies measure response to exposure in controlled systems that model most of the
relevant variables associated with asbestos disease mechanisms in humans (including respirability,
retention, degradation, clearance, translocation, and tissue-specific response). Thus, the available
inhalation studies are the best database from which to evaluate the integrated effects that lead to the
development of asbestos-related disease. Such studies can be used both to identify the characteristics of
asbestos that determine biological activity and to qualitatively elucidate the nature of the corresponding
relationship between exposure via inhalation and the induction of disease.
In this section, the existing animal inhalation studies are reviewed. In the following section, a project
undertaken to overcome the limitations of the existing animal inhalation studies is described (the
supplemental inhalation study). Because this latter project was specifically designed to support the risk
protocol presented in this document, the nature and results of this project are described in detail.
The existing animal inhalation database consists of approximately 30 studies of which approximately 20
contain dose-response information based on lifetime monitoring of exposed animals, including the work
by: Bellman et al. (1986, 1995); Bolton et al. (1982); Davis et al. (1978, 1980, 1985, 1986a,d, 1988a,b);
6.115
-------
Goldstein et al. (1983); Le Bouffant et al. (1987); Lee et al. (1981); McConnell et al. (1982); Muhle et al.
(1987); Platek et al. (1985); Smith et al. (1987); and Wagner et al. (1974, 1982, 1985, 1987). The studies
are similar in overall design, although differences in experimental details potentially affect the
comparability of results from separate studies.
In the inhalation studies, plugs formed from bulk samples of fibrous asbestos and related materials are
placed in a dust generator and aerosolized. The generators (Beckett 1975), usually a modified version of
the apparatus originally designed by Timbrell et al. (1968), consist of a rotating brush that sweeps over an
advancing plug of bulk material and liberates fibers that are entrained in the controlled air flow passing
through the device. The airborne dust is then passed either into a delivery system for nose-only exposure
or into an exposure chamber where animals are kept for fixed periods of time (usually 7 hours per day) on
a weekly routine (typically 5 days per week). The exposure routine is continued for as long as 2 years in
some of the studies. In some, but not all of the studies, fiber-containing air is passed through a cyclone or
elutriator prior to the exposure chamber so that exposure consists primarily of particle sizes within the
respirable range.
Asbestos concentrations in the animal inhalation experiments are monitored by a combination of
techniques. The concentration of total dust in the chamber is generally monitored gravimetrically.
Simultaneously, membrane filter samples are collected and fibers counted by PCM. The quotient of these
two measurements yields the number of (PCM) fibers per unit mass of dust (Section 4.3). The
distribution of fiber sizes within the dusts introduced into the animal: exposure chambers may also be
determined in these studies by any of a variety of methods. As indicated previously (Section 4.3),
however, the utility of such measurements depends on the precise manner in which they are derived.
To derive fiber size distributions, dust samples from these studies have generally been collected on
polycarbonate filters for analysis by SEM. However, such distributions suffer both from the limitations
of SEM (Section 4.3) and from the manner in which they are tied to the inhalation experiments (see
Chapter 5).
Theoretically, the dose of any fiber size fraction can be estimated in a two-step process. The procedure
incorporates consideration of a size fraction termed the PCM-equivalent fraction (PCME), which is the
fraction of structures measured by SEM (or TEM) that correspond to the size range of structures known to
be visible and therefore countable by PCM. First, the concentration of the PCM-equivalent fraction of the
fiber size distribution (measured by SEM) is normalized by dividing its value by the PCM-measured
concentration per unit dust mass observed in the inhalation experiment. This ratio is then multiplied by
the fractional concentration of any specified size range of interest within the distribution (measured by
SEM) to determine the exposure level for that size fraction. However, because bivariate (length by
diameter) size distributions have not typically been developed in the available studies and because the
number of total fibers longer than 5 urn observed by SEM (without adjustment for width) does not
correspond to the number of total fibers longer than 5 urn observed by PCM, it is not possible to derive a
true PCME fraction from the SEM data. Therefore, the theoretical approach described above for
estimating exposure to specific size fractions cannot generally be applied in the existing studies.
Note that SEM analyses are typically conducted on limited dust samples only to provide information on
size distributions. SEM is not used routinely to monitor daily asbestos concentrations in these
experiments. Therefore a procedure like that described above is required to link the absolute
concentrations to which rats are exposed to the measured and relative size distributions that are
determined by SEM.
As indicated above, the data within the published animal inhalation studies are further constrained by the
limitations of the analytical methods employed to generate the data (Section 4.3). Comparison of data
between studies is also hindered by the lack of sufficient documentation to indicate the specific methods
6.116
-------
and procedures employed in each study. Frequently, for example, it is unclear whether respirable dusts or
total dusts have been monitored. Also, several studies fail to report one or both of two critical pieces of
information: fiber-number-to-mass conversion factors and fiber size distributions. In addition, few
studies indicate the precise counting rules employed for generating size distributions.
When structure-number-to-mass conversion factors are provided, unless the conversion factor is derived
by counting fibers in a specific size range in a known mass of sample, and fiber concentrations in other
size ranges are normalized to this count, several types of error may be introduced. For example, if total
sample mass is assumed proportional to calculated mass derived from volume characterizations of the
particles counted, unless isometric particles are sized along with fibers and both asbestos and nonasbestos
particles are included in the count, a bias will be introduced in the conversion factor because total sample
mass will have been under-represented to the extent that such particles are ignored in the estimation of
fiber mass. Even if such particles are included, significant uncertainty may result from estimating the
volumes of irregular particles and the limited precision associated with the count of the largest particles
(due to their limited number). The uncertainty in the measurement of a fiber's diameter is squared in
contributing to the uncertainty associated with a mass estimate.
Among reported variations in study design, differences in the detailed design and operation of the
aerosolization chamber and the frequency and duration of exposure also potentially contribute to variation
in results between studies. Also, use of differing animal strains and species across the various studies
suggest the possibility that physiological differences may contribute to the observed variation in study
results. Such differences are discussed further in Chapter 5.
A small subset of the asbestos dusts evaluated in the animal inhalation studies have been analyzed by
TEM. However, even the published fiber size distributions from these TEM studies are subject to
variation from differences in procedures used for sample preparation, from differences in counting rules,
and from precision limitations due to the limited number of fibers actually characterized (Section 4.3).
This latter limitation particularly affects the precision with which longer fibers are counted.
Although fiber-size distributions are primarily based on SEM analyses rather than TEM analyses in the
existing animal inhalation studies, results generally echo the results of the injection and implantation
studies. Thus, longer fibrous structures are observed to contribute most to asbestos biological activity, at
least qualitatively. For example, dusts containing predominantly long amosite or long chrysotile fibers
induce far more pulmonary tumors than samples containing predominantly short structures (Davis et al.
1986a,b). However, dusts evaluated in the existing inhalation experiments have not been characterized
sufficiently to distinguish the dependence of biological activity on fiber diameter. Neither are the existing
studies sufficient to evaluate the importance of mineralogy (or other potentially important asbestos
characteristics) in determining risk.
6.4.3 Supplemental Inhalation Study
Given the problems with the existing animal inhalation studies, a project was undertaken to overcome
some of the attendant limitations (Herman et al. 1995). To control for effects from variation in study
design and execution (including choice of animal strain, animal handling procedures, equipment design,
sample handling procedures, dosing regimen, and pathology protocols), the project focused on a set of
studies generated from a single laboratory (i.e., the studies published by Davis and coworkers).
Ultimately, the results from six studies covering nine different asbestos samples (including four types of
asbestos with samples exhibiting multiple size distributions for two asbestos types) and a total of 13
separate experiments (some samples were studied at multiple exposure levels or in duplicate runs) were
pooled for analysis. The database of experiments employed in the project is described in Table 6-7.
6.117
-------
To overcome the limitations in the Davis et al. studies associated with the characterization of asbestos
itself, the dusts studied in the thirteen experiments listed in Table 6-7 were regenerated by the same group
who performed the original studies, from the same starting materials, using the same equipment, and
reproducing the same conditions under which the original studies were conducted. Samples of the
regenerated dusts were then collected and analyzed by TEM using a modified version of the Superfund
air method (Chatfield and Berman 1990) to generate bi-variate size distributions that also include detailed
characterization of the shapes and complexity of fibrous structures observed.
The total mass concentration of the regenerated dusts and fiber measurements by PCM were also
collected to provide the data required to link size distributions in the regenerated dusts to absolute
structure concentrations in the original inhalation experiments. The manner in which such calculations
are performed has been published (Berman et al. 1995).
The concentration estimates (for asbestos structures exhibiting a range of characteristics of interest) that
were derived from the TEM analyses of the regenerated dusts were then combined with the tumor
response data from the set of inhalation experiments listed in Table 6-7 and a statistical analysis was
completed to determine if a measure of asbestos exposure could be identified that satisfactorily predicts
the lung tumor incidence observed. A more limited analysis was also performed to address
mesothelioma; the small number of mesotheliomas observed Davis et al. studies constrained the types of
analyses that could be completed for this disease. The detailed procedures employed in this analysis and
the results from the first part of the study have been published (Berman et al. 1995). These are
summarized below along with results from the parts of the study that remain to be published.
6.118
-------
Table 6-7. Summary Data for Animal Inhalation Experiments Conducted by Davis and Coworkers°>b
Fiber
Type
Chrysotile
Chrysotile
Chrysotile
Chrysotile
Chrysotile
Chrysotile
Chrysotile
Amosite
Amosite
Amosite
Crocidolite
Crocidolite
Tremolite
None
None
None
None
None
Description
UICC-A
UICC-A
Long
Short
UICC-A
UICC-A (Discharged)0
WDC Yarnd
UICC
Long
Short
UICC
UICC
Korean
Control
Control
Control
Control
Control
Abbreviations
UC
uc
LC
SC
UC
DC
we
UA
LA
SA
UR
UR
KT
C
C
C
C
C
Mass
Concentration
(mg/m3)
2
10
10
10
9.9
9.9
3.6
10
10
10
4.9
10
10
0
0
0
0
0
PCM
f/ml
390
1,950
5,510
1,170
2,560
2,670
679
550
2,060
70
430
860
1,600
Number
of
Animals
42
40
40
40
36
39
41
43
40
42
43
40
39
20
36
61
64
47
Number of
Benign
Pulmonary
Tumors
6
7
8
1
6
4
5
2
3
0
2
1
2
0
0
1
1
1
Number of
Malignant
Pulmonary
Tumors
2
8
12
6
8
6
13
0
8
0
0
0
16
0
0
1
1
1
Total
Number of
Pulmonary
Tumors
8
15
20
7
14
10
18
2
11
0
2
1
18
0
0
2
2
2
Meso-
theli-
omas Reference
1
0
3
1
0
1
0
0
3
1
1
0
2
0
0
0
0
0
Davis et al.
Davis et al.
Davis et al.
Davis et al.
Davis et al.
Davis et al.
Davis et al.
Davis et al.
Davis et al.
Davis et al.
Davis et al.
Davis et al.
Davis et al.
Davis et al.
Davis et al.
Davis et al.
Davis et al.
Davis et al.
1978
1978
1988a
1988a
1988b
1988b
1986b
1978
1986a
1986a
1978
1978
1985
1978
1985
1986a
1986b
1988a
'Source: Berman et al. 1995
'Exposure occurred for 7 hours per day, 5 days per week for 1 year.
CUICC-A chrysotile in this experiment was treated with mixed polarity air (produced with a source of beta radiation) following generation to reduce the surface charge on
individual particles within the dust.
dChrysotile samples used for dust generation in this experiment were obtained from material treated by a commercial wet dispersion source.
6.119
-------
In the statistical analysis performed in this study, the individual dose-response profiles from each
of the two data sets were fit to a linear dose-response model:
(Eq. 6-7)
where:
"Pj" is the probability of inducing pulmonary tumors observed in the "ith" study;
Q0" is a parameter that accounts for the background incidence of pulmonary tumors
(assumed to be the same in all studies);
Xjj" is the concentration of the "jth" size fraction of fibers in the "ith" study;
3j" is the coefficient of potency for the "jth" size fraction of fibers; and
II- M
"bj" is a coefficient that represents the absolute potency of asbestos. In some analyses
this coefficient is allowed to assume different values for different types of
asbestos (e.g., chrysotile vs. amphibole), or other differences in experimental
conditions.
The "aj"s in this analysis are constrained to be positive because it is assumed that no fiber
prevents cancer. The "a^'s are also constrained to sum to 1.0, which means that they represent
relative potency rather than absolute potency. Asbestos size fractions evaluated represent
disjoint (mutually exclusive) sets.
The model (Equation 6-7) allows separate potency coefficients to be assigned to individual size
fractions in a dose-response relationship that depends on multiple size fractions.
Simultaneously, the "b coefficients allows separate potencies to be assigned to different fiber
types or to results from different studies performed under different experimental conditions.
Several investigators (Bertrand and Pezerat 1980; Bonneau et al. 1986; Stanton et al. 1977;
Wylie et al. 1987) have used a logit curve to investigate the dose-response relating various
measures of asbestos exposure to tumor response. The logit formula specifies that the tumor
probabilities satisfy the relation:
log[P/(l-P)]=a + b-log x (Eq. 6-8)
where log x is some measure of asbestos exposure, such as log of concentration of fibers in some
size range. In some instances, the logit model was expanded by replacing b-log x with a term
representing a linear combination of exposure indices, so that multiple exposure indices could be
explored simultaneously. The models were fit using standard linear regression based on normal
theory.
6.120
-------
An equivalent form for the logit model is:
P=eaxb/(l + e"xb) (Eq. 6-9)
Written in this form, it is clear that this model does not permit a background response (i.e., P=0,
whenever x=0). This is not a serious limitation when there are no tumors in control animals,
such as was the case in Stanton et al. (1977). However, the model will not adequately fit data in
which tumors are found in control animals. This was one reason for adopting the linear model
(Equation 6-7) used in the investigation of the animal data reported in the study described here.
There is no evidence from this study that the linear model is inadequate. For cases in this study
in which the fit between exposure and response is shown to be inadequate, the lack of fit is
typically observed to be due to an inconsistent (non-monotonic) dose-response curve so that
there is no indication that a non-linear model, such as the logit, would provide a better fit.
The linear model (Equation 6-7) used in this study was fit using a maximum likelihood (Cox and
Lindley 1974) approach that utilizes the actual underlying binomial probabilities. This is a more
efficient estimation method than use of regression methods based on normal theory, which was
the fitting method used in the earlier studies (described above). In addition, the regression
procedures applied in the earlier studies indicate only whether the exposure measures that were
studied are significantly correlated with tumor response. In contrast, statistical goodness of fit
tests were applied in this study to determine whether exposures that are described by a particular
characteristic (or combination of characteristics) satisfactorily predict the observed tumor
incidence. To illustrate, it is apparent from Text Figure 2 of Stanton et al. (1981) that the
exposure measure they identify as being most highly correlated with tumor incidence (fibers
longer than 8 urn and thinner than 0.25 urn) does not provide an acceptable fit to the observed
tumor incidence. Similarly, although all of the univariate exposure measures listed in Table 2 of
Berman et al. (1995) are highly correlated with tumor incidence, none of them adequately
describe (fit) lung tumor incidence.
To test for goodness-of-fit in this study, each relationship was subjected to a chi-square
goodness-of-fit test in which the fit of the model was rejected if the corresponding p-value was
less than 0.05, indicating that the true model would provide a worse fit only 5% of the time.
Among models that were not rejected based on a goodness-of-fit test, several hypotheses
concerning the relative merit of the various models were also examined using the method of
maximum likelihood (Cox and Lindley 1974).
An example of an adequate fit to the tumor response data is provided in Figure 6-5 (Figure 3 of
Berman et al. 1995). Note that Figures 2 and 3 of the original paper were inadvertently switched
during publication; the correct Figure 3 is reproduced here. The exposure index plotted in
Figure 6-5 is the sum:
Exposure=a, Xj, + <^ x12 + a3 Xi3=0.0017C, + 0.853C2 + 0.145C3 (Eq. 6-10)
6.121
-------
Figure 6-5. Fit of Model. Tumor Incidence vs. Structure Concentration by TEM
(Length Categories 5-40 iim, >40|im, Width Categories: <0.3 |j,m and >5 \im)
70-
60
50-
f 4D-
I '
3O-
10-
bars Incfeatt 90% corr9d«f>c» Warvait
fcx deta a» exptelinwd fn T«t*» 1. ' •
Less compfsx ctusters and matrtcss feplaeed toy ocmporwitta,
Ortgtnal concentrations estim^od by muttlp!y<"0 oxcmtreSJom ki recor»3t(uaed du»53 by rato o< PCM
•f'-'t i-:-i--"t j •'•) •••<•*• <• ; f » f- '-»•••< ; >-* t i-'
3 4 f «
WfllQhted Afttsonw Conc«ntr«t3or>
Length Categories: S/tfn-40^m, i
Widti Categories: < 0.3^m and
10
6.122
-------
where:
"C," (= Xj,) is the concentration of structures between 5 and 40 |im in length that are
thinner than 0.3 u.m;
"C2" (= X;2) is the concentration of structures longer than 40 jim that are thinner than
0.3 jim; and
"C3" (= Xj3) is the concentration of structures longer than 40 urn that are thicker than 5
jim.
This index of exposure represents one of the optimum indices reported in Berman et al. 1995.
As is clear from the figure, when exposure is expressed in the manner described above, the
tumor responses observed in the 13 separate experiments that were evaluated increase
monotonically with increasing exposure. It is also apparent that the data points representing
each study fall reasonably close to the line representing the optimized model for this exposure
index. This was verified by the fact that the corresponding goodness-of-fit test p-value was
greater than 0.05. Thus, exposure adequately predicts response.
Results obtained from completing more than 200 statistical analyses to determine whether
various measures of asbestos exposure adequately predict lung tumor response (Berman et al.
1995) indicate that:
• neither total dust mass nor fiber concentrations determined by PCM adequately
predict lung tumor incidence;
• no univariate measure of exposure (i.e., exposure represented by the
concentration of a single size category of structures as measured by TEM) was
found to adequately predict lung tumor incidence. Of the univariate measures of
exposure examined, the concentration of total structures longer than 20 jim
provides the best fit (although still inadequate); and
• lung tumor incidence can be adequately predicted with measures of exposure
representing a weighted sum of size categories in which longer structures are
assigned greater potency than shorter structures.
The set of analyses completed in support of this work are summarized in Appendix C.
One example of an exposure measure that adequately describes lung tumor incidence is
presented in Figure 6-5. Another exposure index shown to provide an adequate fit is:
0.0024Ca + 0.9976Cb (Eq. 6-11)
6.123
-------
where:
"Ca" is the concentration of structures between 5 and 40 u.m in length that are thinner
than 0.4 urn; and
"Cb" is the concentration of structures longer than 40 um that are thinner than 0.4 um.
The fit of this index is depicted in Figure 6-6.
In addition to the above, a series of hypotheses tests were also conducted to test such questions
as: whether fiber type affects potency or whether the component fibers in complex clusters and
matrices should be counted individually. Questions concerning whether mesothelioma incidence
can be adequately described by the same measure(s) of exposure that describe lung tumor
incidence were also addressed. Taken as a whole, the results presented in Berman et al. (1995)
support the following general conclusions:
structures contributing to lung tumor incidence are thin (<0.5 um) and long (>5
fim) with structures longer than 20 urn being the most potent;
the best estimate is that short structures (<5 (im) are non-potent. There is no
evidence from this study that these structures contribute anything to risk;
among long structures, those shorter than 40 (im appear individually to contribute
no more than a few percent of the potency of the structures longer than 40 um;
lung tumor incidence is best predicted by measurements in which the component
fibers and bundles of complex structures are individually counted;
at least for lung tumor induction in rats, the best estimate is that chrysotile and the
amphiboles are equipotent;
for equivalent size and shape structures, amphiboles are more potent toward the
induction of mesothelioma than chrysotile; and
after adjusting for the relative potencies of fiber type, the size categories that
contribute to lung tumor incidence appear also to adequately describe
mesothelioma incidence.
6.124
-------
Figure 6-6. Fit of Model. Tumor Incidence vs. Structure Concentration by TEM
(Length Categories 5-40 \an, >4Q\un, Width Categories: <0.4 jim)
5 10 15
Weighted Airborne Concentration
Length Categories: 5pm-40 \im, ar 40 jjm
Width Categories: < 0.4
20
6.125
-------
A number of supplemental analyses were also conducted, primarily to identify optimal
procedures for performing asbestos analysis and for estimating concentrations. These analyses
have not yet been published, but they are included in the summaries in Appendix C. The most
important results of the supplemental analyses are that:
• tumor incidence can only be adequately fit by data derived from TEM analysis of
samples prepared by a direct transfer procedure. Measurements derived from
indirectly prepared samples could not be fit to lung tumor incidence in any
coherent fashion; and
• it was not possible to identify an exposure measure in which potency is expressed
in terms of a single, continuous function of structure length.
Regarding the last point, although we were not able to identify a continuous function of length
that provides an adequate fit to the tumor incidence data, the general results from the above
analysis are not inconsistent with the hypothesis that potency is a continuous function of length
(i.e., the Pott hypothesis, Pott 1982). The Pott hypothesis suggests that relative potency is low
for short fibers, rises rapidly over an intermediate range of length, and approaches a constant for
the longest fibers.
6.4.4 Conclusions Concerning Animal Dose-Response Studies
Results from our evaluation of the animal dose-response data for asbestos (including the existing
injection/implantation studies, the existing inhalation studies, and our supplemental study)
indicate that:
• short structures (less than somewhere between 5 and 10 um in length) do not
appear to contribute to cancer risk;
• beyond a fixed, minimum length, potency increases with increasing length, at
least up to a length of 20 p.m (and possibly up to a length of as much as 40 urn);
• the majority of structures that contribute to cancer risk are thin with diameters
less than 0.5 \im and the most potent structures may be even thinner. In fact, it
appears that the structures that are most potent are substantially thinner than the
upper limit defined by respirability;
• identifiable components (fibers and bundles) of complex structures (clusters and
matrices) that exhibit the requisite size range may contribute to overall cancer risk
because such structures likely disaggregate in the lung. Therefore, such structures
should be individually enumerated when analyzing to determine the concentration
of asbestos;
• for asbestos analyses to adequately represent biological activity, samples need to
be prepared by a direct-transfer procedure; and
6.126
-------
• based on animal dose-response studies alone, fiber type (i.e., fiber mineralogy)
appears to impart only a modest effect on cancer risk (at least among the various
asbestos types).
Regarding the last of the above bullets, that only a modest effect of fiber mineralogy was
observed in the available animal dose-response studies (when large effects are observed among
human studies, Chapter 7), may be due at least in part to the limited lifetime of the rat relative to
the biodurability of the asbestos fiber types evaluated in these studies, although it is also possible
that different mechanisms drive the effects observed in the animal studies than those that
dominate for asbestos-induced cancers in humans and that such mechanisms depend more
strongly on mineralogy. Other explanations are also possible. Issues relating to both fiber
mineralogy and fiber size are addressed further in Section 6.5 and Chapter 7.
6.5 CONCLUSIONS FROM AN EVALUATION OF SUPPORTING STUDIES
Although gaps in knowledge remain, a review of the literature addressing the health-related
effects of asbestos (and related materials) provides a generally consistent picture of the
relationship between asbestos exposure and the induction of lung cancer and mesothelioma.
Therefore, the general characteristics of asbestos exposure that drive the induction of cancer can
be inferred from the existing studies and can be applied to define appropriate procedures for
evaluating asbestos-related risk. Furthermore, although it would be helpful to definitively
identify the underlying biochemical triggers and associated mechanisms that drive asbestos-
induced cancer, this is not an absolute prerequisite for the development of a technically sound
protocol for assessing asbestos-related risk.
As previously indicated (Sections 6.3.8), the biochemical mechanisms that potentially contribute
to the induction of asbestos-induced cancer are complex and varied. Moreover, different
mechanisms appear to exhibit differing dose-response characteristics (i.e., the various
mechanisms do not all show the same kind of dependence on fiber size or fiber type). Some
mechanisms, for example, suggest that fiber length is important and that only structures that are
sufficiently long induce a response. In contrast, other mechanisms suggest that fibers (and even
non-fibrous particles) may all contribute to response and that the magnitude of the response is a
function of the total surface area of the offending fibers (or particles). Among these
mechanisms, additionally, some suggest that fiber type (i.e., mineralogy) is not an important
determinant of potency while other mechanisms indicate that fiber type is an important
determinant of potency.
The existing studies are not currently adequate to support definitive identification of the specific
mechanisms that drive the induction of asbestos-related cancer (versus other mechanisms that
may contribute only modestly or not at all). However, whatever mechanisms in fact contribute
to the induction of disease, they must be consistent with the gross characteristics of exposure that
are observed to predict response in the available whole-animal dose-response studies and human
epidemiology studies. Therefore, the implications from these latter studies regarding the
dependence of asbestos-induced cancers on fiber size and type are reviewed here in some detail.
Further, in Chapter 8, they are used to support development of a protocol for evaluating
asbestos-associated risks.
6.127
-------
Fiber Length. Fibers less than a minimum length between 5 and 10 |im do not appear to
contribute to risk. This is supported both by the results of our re-analysis of the animal
inhalation studies conducted by Davis and coworkers (Section 6.4.3 and Berman et al. 1995), in
which this hypothesis was tested formally, and by inferences from the broader literature. As
long as fiber size is adequately characterized, the animal inhalation studies (Section 6.4.2) and
injection/implantation (Section 6.4.1) studies consistently indicate lack of ability of short
structures to contribute to the induction of cancer. Furthermore, animal retention studies
(Section 6.2.1) and histopathology studies (Section 6.2.2) provide strong mechanistic evidence
that explains the lack of potency for short structures; they are readily cleared from the respiratory
tract. Even when sequestered in large numbers in macrophages within the lung, there is little
indication that such structures induce the kinds of tissue damage and related mechanisms that
appear to be closely associated with the induction of cancer.
Although there are mechanism studies that may suggest a role for short fibers in the induction of
asbestos-related disease (see, for example, Goodglick and Kane 1990, Section 6.3.4.4), such
studies do not track cancer as an endpoint. Therefore, the relationship between the toxic
endpoint observed and the induction of cancer needs to be adequately addressed before it can be
concluded definitively that short structures can contribute to cancer. Moreover, such a
conclusion would be surprising given the substantial evidence that exists to the contrary.
Beyond the minimum length below which structures may be non-potent, potency appears to
increase with increasing length, at least up to a length of 20 (im and potentially up to a length of
40 urn. The latter limit is suggested by our re-analysis of the Davis et al. studies (Section 6.4.3)
in which it was also found that structures longer than 40 u.m may be as much as 500 times as
potent as those between 5 and 40 |im in length. The former limit is suggested by broader
inferences from the literature that suggest the cutoff in the length of structures that are at least
partially cleared by macrophages from the lung may lie close to 20 u.m and that the efficiency of
clearance likely decreases rapidly for structures between 10 and 20 jam in length (Section 6.2).
Such inferences are further reinforced by measurements of the overall dimensions of
macrophages in various mammals by Krombach et al. (1997), as reported in Section 4.4.
Importantly, the inferences that potency increases for structures longer than 10 urn (up to some
limiting length) from these various studies are strongly reinforcing, even though the upper limits
to the points at which potency stops increasing do not precisely correspond. Furthermore, that
the longest structures are substantially more potent than shorter structures (and that the shortest
structures are likely non-potent) dictates that asbestos analyses performed in support of risk
assessment need to provide adequate sensitivity and precision for counts of the longest
structures.
Fiber Diameter. Because fibers that contribute to the induction of cancer must be respirable,
they must also be thin. The studies reviewed in Section 6.1 indicate that respirable fibers are
thinner than 1.5 ^m and the vast majority of such structures are thinner than 0.7 ^m. In fact, the
results of injection, implantation, and inhalation studies reviewed in Sections 6.4.1 and 6.4.2 and
the results of our supplemental re-analysis of the Davis et al. studies (Section 6.4.3) indicate that
the fibers that contribute most to the induction of asbestos-related cancers are substantially
thinner than the limit suggested by respirability alone.
6.128
-------
Importantly, the results of all of the studies cited above indicate that it is a cutoff in absolute
width that defines the bounds of biological activity rather than a cutoff in aspect ratio (the ratio
of length to width) that has been used to define fibrous structures heretofore. That is why the
exposure index recommended based on our review of the studies by Davis and coworkers
incorporates a maximum width as a cutoff, rather than a minimum aspect ratio.
Fiber Complexity. In our supplemental evaluation of the Davis et al. studies (Section 6.4.3), the
tumor incidence data from the animal inhalation studies were best fit (predicted) by exposure
indices in which the component fibers and bundles of complex structures (clusters and matrices)
were separately enumerated and included in the exposure index used to represent concentration
(Section 6.4.3). The appropriateness of such an approach is further supported by the observation
that loosely bound structures (including, for example, chrysotile bundles) readily disaggregate
in vivo (Section 6.2). Therefore, it is recommended in this report that those components of
complex structures that individually exhibit the required dimensional criteria be individually
enumerated and included as part of the count during analyses to determine the concentration of
asbestos in support of risk assessment.
Fiber Type (Mineralogy). The magnitude of any effect of mineralogy upon cancer risk in
rodents appears to be modest at best. On the other hand, mineralogy appears to be an important
determinant for cancer risk in human epidemiology studies (Chapter 7), with chrysotile
appearing less potent than amphibole for inducing mesothelioma and (with lesser certainty) lung
cancer. This difference may be due to differences in the life spans of rats and humans compared
to the differential biodurability of the different fiber types. It must also be emphasized that, due
to confounding, the effects of fiber size and fiber mineralogy need to be addressed
simultaneously, if one is interested in drawing useful conclusions concerning fiber mineralogy.
Results from some (but not all) of the animal injection, implantation, and inhalation studies
previously reviewed (Sections 6.4.1 and 6.4.2) suggest that mineralogy plays an important role
in determining biological activity. However, the nature of the effects of mineralogy are not
easily separated from size effects, due to the methodological limitations of the studies cited.
Therefore, the evidence from these studies can be considered ambiguous. Formal hypothesis
testing during the re-analysis of rat inhalation studies (Section 6.4.3) indicates that, when size
effects are addressed, chrysotile and the amphiboles exhibit comparable potency toward the
induction of lung cancer. In contrast, amphiboles were estimated to be approximately 3 times
more potent than chrysotile toward the induction of mesothelioma, once fiber size effects are
addressed.
Several of the human pathology studies cited previously (Section 6.2.3) suggest that mineralogy
is an important factor in determining cancer risk, but these studies similarly suffer from
methodological difficulties that introduce ambiguity into the inferences drawn. However, it is
clear from the human epidemiology data (Chapter 7) that mineralogy plays a substantial role in
the determination of risk for human cancer (primarily, mesothelioma).
The underlying cause(s) for the observed difference in potency between chrysotile and the
amphiboles may relate to differences in fiber durability (Section 6.2), to size/shape related
differences in fibers that are a function of mineralogy and that cause differences in deposition,
retention, or translocation (Sections 6.1 and 6.2), and/or to the dependence on mineralogy of the
6.129
-------
specific mechanisms underlying the biological responses of specific tissues (Section 6.3). The
relative magnitudes of such effects on animal and human pathology also need to be considered,
if the observed differences in potency among animal and human studies, respectively, is to be
reconciled. Such considerations are addressed further in Chapter 8.
Importantly, whether the observed differences in the role of mineralogy toward animal and
human pathology can be reconciled, the effects of mineralogy can be adequately addressed when
assessing asbestos-related cancer risk for humans by incorporating dose-response coefficients
explicitly derived from the human epidemiology data. Because this is the approach proposed in
this document, effects due to mineralogy are properly addressed.
An Appropriate Exposure Index for Risk Assessment. The optimum exposure index defined
based on the re-analysis of the animal inhalation studies conducted by Davis and coworkers
(Section 6.4.3) is a weighted sum of the concentrations of (1) structures between 5 and 40 ^m in
length that are thinner than 0.4 urn, and (2) structures longer than 40 (am that are thinner than 0.4
u.m (Equation 6-11).
This index was shown to adequately fit (predict) the tumor incidence data across the 13 separate
animal inhalation experiments evaluated (P=0.09). Whether the exposure index defined in
Equation 6-11 is also optimal for capturing the relevant characteristics of fibers that contribute to
the induction of human cancer is an open question. Because it captures the major characteristics
(concerning length and diameter) identified above that are indicated to be important for human
exposures, it represents a promising candidate. Unfortunately, however, the data required to
match this index to a set of human-derived exposure-response coefficients does not currently
exist (Section 7.4). Therefore, compromises are required to apply the general conclusions of this
Chapter to the human data. These are addressed further in Section 7.4.1.
6.130
-------
7.0. EPIDEMIOLOGY STUDIES
The existing epidemiology studies provide the most appropriate data from which to determine
the relationship between asbestos exposure and response in humans. As previously indicated,
however, due to a variety of methodological limitations (Section 5.1), the ability to compare and
contrast results across studies needs to be evaluated to determine the confidence with which risk
may be predicted by extrapolating from the "reference" epidemiology studies to new
environments where risk needs to be assessed. Reliable extrapolation requires both that the
uncertainties contributed by such methodological limitations and that several ancillary issues
(identified in Chapter 2) be adequately addressed.
A detailed discussion of the methodological limitations inherent to the available epidemiology
studies was provided in the Health Effects Assessment Update (U.S. EPA 1986) and additional
perspectives are provided in this document (Section 5.1). The manner in which the uncertainties
associated with these limitations are addressed in this document are described in Appendix A.
As previously indicated (Chapter 2), the ancillary issues that need to be addressed include:
(1) whether the models currently employed to assess asbestos-related risk adequately
predict the time and exposure dependence of disease;
(2) whether different mineral types exhibit differential potency (and whether any
differences in potency relate to the relative in vivo durability of different asbestos
mineral types);
(3) whether the set of minerals included in the current definition for asbestos
adequately covers the range of minerals that potentially contribute to asbestos-
related diseases; and
(4) whether the analytical techniques and methods used to characterize exposures in
the available epidemiology studies adequately capture the characteristics of
exposure that affect biological activity.
All but the third of the above issues are addressed in this chapter. Currently, the third issue can
best be addressed by evaluating inferences from the broader literature (see Chapter 6). The
remaining issues are addressed separately for lung cancer and mesothelioma following a brief
overview of the approach adopted for evaluating the epidemiology literature.
7.1 APPROACH FOR EVALUATING THE EPIDEMIOLOGY LITERATURE
To develop exposure-response relationships (and corresponding exposure-response coefficients)
for use in risk assessment from epidemiological data, two basic types of information are
necessary: information on the disease mortality experienced by each member of the study
population (cohort) and information on the asbestos exposure experienced by each member of
the cohort. So that disease mortality attributable to asbestos can be distinguished from other
(background) causes of death, it is also necessary to have knowledge of the rates of mortality
7.1
-------
that would be expected in the study population, absent exposure. Normally, such information
must be determined based on a "reference" or "control" population.
Ideally, one would like to have complete knowledge of exposure at any period of time for each
individual in the cohort and complete access to the data to fit different types of
exposure/response models to the data so that the approach for evaluating the relationship
between exposure and response can be optimized. In most instances, unfortunately, the data
suffer from multiple limitations (see Section 5.1 and Appendix A) and the analysis is further
constrained by less than complete access to the data.
Briefly, the major kinds of limitations that potentially contribute to uncertainty in the available
epidemiology studies (and the effect such limitations likely produce in estimates of exposure-
response coefficients) include:
• limitations in the manner that exposure concentrations were estimated
(contributing to variation across studies);
• limitations in the manner that the character of exposure (i.e., the mineralogical
types of fibers and the range and distribution of fiber dimensions) was delineated
(contributing to systematic variation between industry types and, potentially,
between fiber types);
• limitations in the accuracy of mortality determinations or incompleteness in the
extent of tracing of cohort members (contributing to variation across studies);
• limitations in the adequacy of the match between cohort subjects and the selected
control population (contributing to variation across studies and may have a
substantial effect on particular studies); and
• inadequate characterization of confounding factors, such as smoking histories for
individual workers (contributing to variation across studies and may have a
substantial effect on particular studies).
More detailed discussion of the above limitations is provided in Section 5.1. The manner in
which these limitations are being addressed in this evaluation are described briefly below and in
more detail in Appendix A.
The existing asbestos epidemiology database consists of approximately 150 studies of which
approximately 35 contain exposure data sufficient to derive quantitative exposure/response
relationships. A detailed evaluation of 20 of the most recent of these studies, which includes the
most recent follow-up for all of the cohorts evaluated in the 35 studies, based on the
considerations presented in this overview, is provided in Appendix A.
This new analysis of the epidemiology database differs from the evaluation conducted in the
1986 Health Effects Assessment Update (U.S. EPA 1986) in several ways. It incorporates new
studies not available in the 1986 update that contain information on exposure settings not
previously evaluated as well as more recently available follow-up for exposure settings
7.2
-------
previously evaluated. It also incorporates new features in the manner in which the analysis was
conducted. These new features include:
• estimation of "uncertainty" bounds for the exposure-response coefficients
(potency factors) derived from each study; and
• for the lung cancer model, introducing a parameter, a, which accounts for the
possibility that the background lung cancer mortality rate in the asbestos-exposed
cohort differs systematically from the rate in the control population.
The uncertainty bounds were developed to account for uncertainty contributed by the manner
that exposure was estimated, by the manner that work histories were assigned, by limitations
imposed by the manner in which results were reported in published papers, and by limitations in
the accuracy of follow-up, in addition to accounting for the statistical uncertainty associated with
the observed incidence of disease mortality. A detailed description of how the uncertainty
bounds were constructed is provided in Appendix A.
Exposure-response coefficients were estimated for each cohort both by requiring that oc=l (the
approach followed in the 1986 Health Effects Assessment Update, U.S. EPA 1986) and by
allowing a to vary while fitting the lung cancer model to data. Allowing a to vary addresses
potential problems due to differences in background lung cancer rates between the cohort and the
control population due, for example, to differences in smoking habits in the two populations. As
indicated in Appendix A, this adjustment has a substantial effect on the fit of the U.S. EPA
model to the data for several specific cohorts and a corresponding effect on the estimates of the
lung cancer exposure-response coefficients for those cohorts.
We were also able to obtain the original, raw data for selected cohorts from a limited number of
the more important of the published epidemiology studies. This allowed us to more formally
evaluate the appropriateness of the existing U.S. EPA models for lung cancer and mesothelioma
(Sections 7.2 and 7.3).
Exposure-response coefficients, and corresponding risk estimates derived therefrom, must be
based upon an "exposure index" that expresses the relative potency of asbestos fibers of different
dimensions. For example, the exposure index utilized in the 1986 Health Effects Assessment
Update (U.S. EPA 1986) assigns equal potency to all fibers longer than 5 u.m that exhibit an
aspect ratio >3 and a thickness >0.25 urn, regardless of type of asbestos, and assigns zero
potency to shorter, squatter, or thinner fibers. In this update, we evaluated a range of such
exposure indices, both with respect to agreement with evidence from the literature on the relative
potency of asbestos structures of differing types and dimensions (Section 7.4) and with respect to
overall agreement across the exposure-response coefficients derived from the available
epidemiology studies and adjusted for exposure index. This analysis led to proposed new
exposure indices that better reflect the evidence from the literature on the relative potency of
different structures and provide improved agreement among exposure-response coefficients
estimated from different environments. Such improvement in agreement across studies
correspondingly increases the confidence with which the exposure-response factors derived from
the existing studies can be applied to new environments.
7.3
-------
7.2 LUNG CANCER
The 1986 U.S. EPA lung cancer model (U.S. EPA 1986) assumes that the relative risk, (RR), of
mortality from lung cancer at any given age is a linear function of cumulative asbestos exposure
in units of fiber-years/ml (f-y/ml) as measured by PCM, disregarding any exposure in the most
recent ten years. This exposure variable is denoted by CE10) and its use embodies the assumption
that asbestos exposures during the most recent 10 years do not affect current lung cancer
mortality risk. The mathematical expression for this model is:
RR=1+KL*CEIO (Eq. 7-1)
where the linear slope, KL, is termed the "lung cancer exposure-response coefficient." This
parameter is generally estimated by fitting the model to data from an occupational mortality
cohort study consisting of observed and expected numbers of cancer deaths categorized by
cumulative exposure, with the expected numbers determined from age- and calendar-year-
specific lung cancer mortality rates from an appropriate control population (e.g., U.S. males).
This model predicts that the mortality rate in an asbestos-exposed population is the product of
the mortality rate in an unexposed, but otherwise comparable, population, and the RR.
Consequently the excess mortality due to asbestos is the product of the background mortality rate
and the excess RR, KL*CE10. Since smokers have a higher background mortality from lung
cancer, the model predicts a higher excess asbestos-related lung cancer mortality in smokers than
in non-smokers.
To account for the possibility that an occupational cohort may have a different background
mortality rate of lung cancer than the control population (e.g., due to different smoking habits or
exposures to other lung carcinogens), in the present analysis Equation 7-1 is expanded to the
form,
RR = cc*(l+KL*CE10) (Eq.7-2)
where a is the RR in the absence of asbestos exposure relative to the control population. This
form of the model contains two parameters, the background RR, a, and the lung cancer
exposure-response coefficient, KL.
7.2.1 The Adequacy of the Current U.S. EPA Model for Lung Cancer
Access to the raw epidemiology data from two key studies allowed us to evaluate the adequacy
of the U.S. EPA model (Equation 7-2) for describing the time and exposure dependence for lung
cancer in asbestos-exposed cohorts. For this analysis, the raw data for the cohort of crocidolite
miners in Wittenoom, Australia was graciously provided by Dr. Nick de Klerk (de Klerk 2001)
and the raw data for the cohort of chrysotile textile workers (described by Dement et al. 1994)
was graciously provided by Terri Schnoor of NIOSH (Schnoor 2001). The Wittenoom cohort
was originally described by Armstrong et al. (1988), but the data provided by de Klerk includes
additional follow-up through 1999.
7.4
-------
7.2.1.1 Exposure Dependence
To evaluate the adequacy of the linear exposure response relationship assumed by the U.S. EPA
lung cancer model, the lung cancer model (Equation 7-2) was fit to the raw data from both the
South Carolina and Wittenoom cohorts. In these analyses, each person-year of follow-up was
categorized by cumulative exposure defined using a lag of 10 years. The data were then grouped
into a set of cumulative exposure categories and the observed and expected numbers of lung
cancers were computed for each category. For South Carolina, expected numbers were based on
sex-race-age- and calendar-year-specific U.S. rates. Separate analyses were conducted for white
males, black males, and white females, as well as for the combined group. For Wittenoom
expected rates were based on age- and calendar-year-specific rates for Australian men. The
categorized data and the resulting fit of the model to the data from Wittenoom and South
Carolina are presented in Tables 7-1 and 7-2, respectively.1
For the Wittenoom data (Table 7-1), the fit with ce=l is poor (p<0.0001), as the model
overpredicts the number of cancers in the highest exposure category and underpredicts at lower
exposures. By contrast the fit of the model with a variable is adequate (p=0.1). This model
predicts a relatively high background of lung cancer in this cohort relative to Australian men in
general (a=2.1) and a correspondingly shallow slope, with the RR increasing only from 2.1 at
background to 3.6 in the highest exposure category. A test of the hypothesis that a=l for this
cohort is rejected (p<0.01) and the model fit to the data (with a variable) predicts K^O.0047
(f/ml-yr)'1.
The fit of the U.S. EPA model to the South Carolina lung cancer data categorized by cumulative
exposure is shown in Table 7-2. The model with a=l cannot be rejected both when the model is
applied to white males only (p=0.54) or with all data combined (p=0.92). Since the values a and
KL estimated from white males only are similar to those estimated using the complete cohort
(black and white males and white females), the fit to the complete data is emphasized in this
analysis. This fit predicts a=1.2 and 1^=0.021 (f-y/ml)"1. A test of the hypothesis that a=l for
this cohort cannot be rejected (p=0.21), and in this fit KL=0.028 (f-y/ml)'1
'The fitting of the lung cancer model to the cohort data was carried out by assuming that the observed
numbers of cancers in different exposure categories were independent, with each having a Poisson distribution with
expectation equal to the expected number based on the control population times the relative risk predicted by
Equation 7-2. In the computation of the relative risk for a cumulative exposure category, the person-year-weighted
average cumulative exposure (lagged 10 years) for the category was used to represent the exposure in that category.
With these assumptions, a and KL were estimated by the method of maximum likelihood, confidence intervals were
constructed using the profile likelihood method, and likelihood ratio tests were used to test hypotheses (Cox and
Oakes 1984; Venson and Moolgavkar 1988).
7.5
-------
Table 7-1. Fit of EPA Lung Cancer Model to Observed Lung Cancer Mortality Among
Wittenoom, Australia Miners (Deklerk 2001) Categorized by Cumulative Exposure
Lagged 10 Years
Cumulative Exposure Lagged
10 Years (f-y/ml)
Range
0
0-0.4
0.4-1.0
1.0-2.3
2.3-4.5
4.5-8.5
8.5-16
16-28
28-60
60+
Total
Test of H0: a=l
p<0.01
Estimates of KL
Average
0
0.19
0.69
1.6
3.3
6.2
11.8
21.5
41.1
142.0
(f-y/ml)-1
Observed
Deaths
5
27
11
22
28
38
31
21
25
43
251
Goodness of Fit
Expected
Deaths
4.6
7.9
8.2
11.6
12.9
14.3
13.2
9.2
11.6
11.6
105.1
P-value
Predicted Deaths
by Model
(0=1)
4.6
8.0
8.3
12.1
14.0
16.7
17.4
14.5
24.5
56.5
176.6
O.0001
(0=2.1)
9.8
17.0
17.6
24.9
27.9
31.4
29.8
21.6
29.6
41.6
251.0
0.1
KL=0.027
90% CI: (0.020,0.035)
(a=2.1 [MLE])
KL=0.0047
90% CI: (0.0017, 0.0087)
7.6
-------
Table 7-2. Fit of EPA Lung Cancer Model to Observed Lung Cancer Mortality Among
South Carolina Textile Workers (Schnoor 2001) Categorized by Cumulative Exposure
Lagged 10 Years
Cumulative Exposure Lagged
10 Years (f-y/ml)
Range
0
0-0.5
0.5-1.0
1.0-2.5
2.5-4.5
4.5-8.5
8.5-16
16-28
28-60
60-80
80-110
110+
Total
Test of H0: cc=l
p=0.21
Average
0
0.35
0.75
1.7
3.4
6.2
11.8
21.3
41.5
69.2
93.3
173
Observed
Deaths
0
4
7
10
13
11
11
8
14
9
10
25
122
Goodness of Fit
Expected
Deaths
0.9
2.6
5.5
9.7
7.9
7.7
7.2
5.3
6.3
3.0
2.9
4.3
63.4
P-value
Predicted Deaths
by Model
(0=1)
0.9
2.7
5.6
10.1
8.6
9.1
9.6
8.4
13.8
9.0
10.8
25.6
114.1
0.92
(o=1.2)
1.1
3.2
6.8
12.1
10.2
10.5
10.9
9.2
14.3
9.0
10.5
24.2
122.0
0.96
Estimates of KL (f-y/ml)-1
KL=0.028
90% CI: (0.021,0.037)
(a=1.2[MLE])
KL=0.021
90% CI: (0.012,0.034)
Appendix A contains fits of the model to published exposure-response data from 18 studies (all,
with the exception of Wittenoom and South Carolina, obtained from the published literature). In
each of these 18 cases the linear model (Equation 7-2) provides an adequate description of the
exposure-response. However, a was estimated as >1.0 in 15 of these cases and statistically
significantly so in six cases, compared to only one case where a was found to be significantly
<1.0. If the true background lung cancer rates in these 15 cohorts with cc>l .0 are equal to that in
the corresponding control population (i.e., so that a fit with cc=1.0 is appropriate), the exposure-
response would appear to be supra-linear. At the same time, the data in 11 of the 18 studies
(more than half) can be adequately fit with
-------
model (Equation 7-2), further evaluation of the lung cancer exposure-response relationship
appears to be warranted.
7.2.1.2 Time Dependence
The lung cancer model (Equation 7-2) was next evaluated to determine whether it adequately
describes the time-dependence of the lung cancer mortality observed in the Wittenoom and
South Carolina cohorts. Of particular interest is whether the model accurately predicts lung
cancer mortality many years after exposure has ceased. The model predicts that RR increases
linearly with cumulative exposure lagged 10 years until 10 years following the end of exposure,
after which it remains constant from that time forward. However, it has been suggested (e.g.,
Walker 1984) that RRs for lung cancer eventually decline after cessation of exposure. Any
investigation of this issue should control for exposure level, since exposures can also fall with
increasing time due to higher death rates in more heavily exposed subjects. The availability of
the raw data from the South Carolina and Wittenoom cohorts provides an opportunity to explore
this issue in more depth.
To investigate the assumption inherent in the lung cancer model (Equation 7-2) that the RR
remains constant following the secession of exposure, bivariate tables were constructed from
both the Wittenoom and South Carolina data in which observed and expected numbers of lung
cancers were cross-classified by both cumulative exposure (using the same exposure categories
as in Tables 7-1 and 7-2) and time since last exposure categorized using 5-year intervals. The
lung cancer model (Equation 7-2) was then fit to these bivariate data. A formal statistical test
was conducted of whether the lung cancer RR changed following the end of exposure. This test
consisted of appending the multiplicative factor, exp(-K.TSLE), to the lung cancer model
(Equation 7-2), where TSLE is time since last exposure, and conducting a likelihood ratio test of
whether the estimated parameter, K, is statistically significantly different from zero. For
presentation the bivariate tables were collapsed into univariate tables (Tables 7-3 and 7-4)
categorized only by time since last exposure, by summing observed and expected cancers over
cumulative exposure categories and computing person-year-weighted averages of cumulative
exposure and times since last exposure.
Table 7-3 shows the resulting fit of the U.S. EPA lung cancer model to the Wittenoom lung
cancer data categorized by time since last exposure. The RRs rise to a maximum between 10
and 30 years from last exposure and then decline thereafter. However, a very similar pattern is
seen with cumulative exposure, which peaks between 10 and 15 years from last exposure, and
then declines due to higher mortality among more heavily exposed workers. The U.S. EPA lung
cancer model (which does not assume a decrease in RR with time, but does account for any
decrease in exposure with increasing time since last exposure) provides an adequate fit to these
data (p=0.13, a estimated), and there is no apparent tendency for the predicted deaths to fall
below observed at the longest times since last exposure. To the contrary, the model-predicted
number of lung cancer deaths more than 35 years since last exposure (53.3) is very close to, but
slightly larger than, the observed number (51). Likewise, the parameter K was not significantly
different from zero (p=0.16). Thus, the lung cancer model (Equation 7-2) provides a good
description of the Wittenoom lung cancer data categorized by time since last exposure and there
is no indication of a drop in the RR up to 45 or more years after exposure has ended that cannot
be accounted for by reduced exposures in the longest time categories. It should also be noted
7.8
-------
that the values of a (2.1) and KL (0.0051 f/ml-y) estimated from this bivariate analysis are very
similar to those estimated from the univariate analysis summarized in Table 7-1.
Table 7-3. Fit of EPA Lung Cancer Model to Observed Lung Cancer Mortality Among
Wittenoom, Australia Miners (Deklerk 2001) Categorized by Years Since Last Exposure
Average
Years Since Last Exposure Exposure Predicted Deaths
Range
0-1
1-5
5-10
10-15
15-20
20-25
25-30
30-35
35^0
40^5
45+
Total
Lagged 10 Years Observed
Average ,, ... _
(f-y/ml) Deaths
0.27
3.0
7.5
12.5
17.5
22.5
27.4
32.3
37.1
43.3
51.2
1.2
1.2
7.7
24.8
24.1
22.8
22.0
21.7
20.0
16.7
13.0
0
1
8
19
26
42
54
50
31
20
0
251
Expected
Deaths
0.6
1.6
3.9
7.0
11.4
16.4
20.1
20.3
13.9
9.6
0.4
105.1
Relative
Risk
0
0.6
2.0
2.7
2.3
2.6
2.7
2.5
2.2
2.1
0
by Model
(a=2.1)
1.5
3.9
10.3
19.2
29.1
39.8
47.2
46.6
31.4
21.1
0.8
250.9
Goodness of Fit P-value
0.19
Table 7-4 shows the corresponding fit of the U.S. EPA lung cancer model to the South Carolina
lung cancer data. This table indicates a marked decrease in RR with increasing time since last
exposure. However, there is a concomitant decrease in cumulative exposure. The U.S. EPA
lung cancer model provides an adequate fit to these data (p=0.31, a estimated), and the estimates
of a (1.3) and KL (0.20 (f-y/ml)"1) are very similar to those obtained from the univariate analysis
(Table 7-2). There is no obvious tendency for the model to underestimate risk at the longest
times since last exposure, and the value of K estimated for this cohort is not significantly
different from zero (p=0.12).
7.9
-------
Table 7-4. Fit of EPA Lung Cancer Model to Observed Lung Cancer Mortality Among
South Carolina Textile Workers (Schnoor 2001) Categorized by Years
Since Last Exposure
Average
Years Since Last Exposure Exposure Predicted Deaths
Lagged 10 Years Observed Expected Relative by Model
(f-y/ml) Deaths Deaths Risk (
-------
only 1.5 years, time since first exposure approximates time since last exposure. Because
Seidman et al. (1986) did not include a lag in their exposure estimates, only data for times since
first exposure >10 years are included in our analysis, since (in this range) lagged and unlagged
exposures would be very similar.
Table 7-5. Fit of EPA Lung Cancer Model to Observed Lung Cancer Mortality Among
New Jersey Factory Workers (Siedman et al. 1986) Categorized by Years Since First
Exposure
Years Since
First Exposure
10-14
15-19
20-24
25-29
30-34
35+
Total
Observed
Deaths
13
20
17
22
21
16
109
Expected
Deaths
8.1
8.8
9.0
8.6
6.1
4.2
44.7
Relative Risk
1.6
2.3
1.9
2.5
3.5
3.8
Predicted Deaths by Model
(0=3.4)
16.2
18.8
20.1
20.8
17.8
15.3
109.0
Goodness of Fit P-value 0.55
The RR of lung cancer increases with time since the beginning of exposure in the Seidman
cohort. Nevertheless, the fit of the U.S. EPA model to these data with a estimated is very good
(p=0.55) and the numbers of cancer deaths predicted by the model agree closely with the
observed numbers, even up through 35+ years from the beginning of exposure. Time since last
exposure (estimated in each cell as time since first exposure minus average duration of exposure)
is not a significant predictor of lung cancer mortality (K is not significantly different from zero,
p=0.33). The values of a (3.4) and KL (0.01 (f-y/ml)"1) estimated from this analysis agree closely
with those estimated in Appendix A by a different approach (cf. Table A-13 in Appendix A).
The analysis in Appendix A was also repeated after eliminating workers who worked longer than
2 years (i.e., workers for whom time since first exposure was not a good approximation to time
since last exposure). Results obtained in this analysis are very similar to those shown in Table
7-5.
These analyses of the relationship between lung cancer mortality and time after exposure ends
are based on cohorts exposed to relatively pure asbestos fiber types: crocidolite (Wittenoom,
Table 7-3), chrysotile (South Carolina factory, Table 7-4) and amosite (Patterson, New Jersey
factory, Table 7-5). All three of these were consistent with the assumption inherent in the U.S.
EPA lung cancer model (Equation 7-2) that RR of lung cancer mortality remains constant after
10 years past the end of exposure. Thus, the U.S. EPA model appears to adequately describe the
time-dependence of lung cancer mortality in asbestos exposed cohorts.
7.11
-------
7.2.1.3 Smoking-Asbestos Interaction With Respect to Lung Cancer
The existence of an interaction between smoking and asbestos in causing lung cancer has been
known for many years. For example, Hammond et al. (1979), in a study of U.S. insulation
workers, found a multiplicative relationship between smoking and asbestos exposure in causing
lung cancer. A multiplicative relationship for the interaction between smoking and asbestos is
also the relationship inherent in the U.S. EPA lung cancer model (Equations 7-1 and 7-2), by
virtue of the fact that increased risk from smoking in reflected in the background risk, which is
multiplied by the asbestos-induced RR. More recent work, however, suggests that the
interaction between smoking and asbestos exposure may be more complicated.
As described by Liddell (2001), many reviews of the interaction between smoking and asbestos
exposure have tended to conclude that the interaction is multiplicative due primarily to (1) the
conclusions of Hammond et al. (1979), which was the largest of the early studies and was thus
most heavily weighted, and (2) lack of viable alternatives among the options considered. As
Liddell (2001) explains, commonly, reviewers examined fits of a simple multiplicative model or
a simple additive model and found that (of these two) the multiplicative model tended to provide
a better fit, although the fits of either model were sometimes relatively poor. Moreover, only
limited sets of mortality data from asbestos exposed cohorts could be used to evaluate these
model fits because only limited smoking histories had been collected among such cohorts and,
among the data sets studied, the evaluation was typically limited to distinguishing effects among
dichotomous groups (i.e., non-smokers and smokers, the latter of which were treated as a single,
collective group).
In a more recent analysis of cigarette smoking among the cohort of Quebec miners and millers
exposed to chrysotile (which had been previously described in numerous studies, see Appendix
A), Liddell and Armstrong (2002) found that the interaction between smoking and asbestos
exposure was complex: certainly less than multiplicative, but not quite linear either.
To get some idea of the possible impact of a non-multiplicative asbestos-smoking interaction
upon our work to reconcile lung cancer exposure-response coefficients calculated from different
environments, consider the generalized model for RR,
RR = 8 + ps.S + pA.A + Y.A.S (Eq. 7-3)
where S is a measure of the amount smoked, A is a measure of asbestos exposure, and 8, ps, PA
and y are parameters. If Y=0, the model predicts an additive smoking/asbestos interaction, and if
Y=PS * PA / 8, the model becomes
RR = 8 * [1 + (ps / 8). S] * [1 + (pA /8). A] (Eq. 7-4)
which is a multiplicative smoking/asbestos interaction. Thus, this model generalizes both
additive and multiplicative interaction, and has been used by a number of researchers, including
Liddell and Armstrong (2002), to study smoking/asbestos interactions.
7.12
-------
Note that the generalized model Equation 7-4 can be written in the form
ps.S)] (Eq. 7-5)
Further, note that Equation 7-5 is of the form: const, * (l+constj* [asbestos exposure]), which is
the same form as the U.S. EPA lung cancer model (Equation 7-2), except that in Equation 7-5
const, (=6+ps.S) and const2 (=(pA+Y.S)/(5+ps,S)) depend upon the amount smoked, while in the
U.S. EPA lung cancer model (Equation 7-2) const, (=a) likewise depends upon the amount
smoked, but the const2 (=KL) does not. Thus even with this generalized model, the U.S. EPA
model would still apply, although the KL estimated would depend upon the smoking habits of the
underlying cohort. That the generalized model Equation 7-3 is at least approximately correct is
supported by the fact that the U.S. EPA model (Equation 7-2), which we have noted is
comparable to Equation 7-3 except that the KL value may depend upon smoking, provided a
valid fit to data from all of the 20 studies to which it was applied.
This at least suggests that, even if the multiplicative interaction assumed by the U.S. EPA model
(Equation 7-2) is not correct, as long as the smoking habits in different cohorts don't differ
extremely, the KL value from different cohorts will be estimates of nearly the same quantity. In
particular, the differences in the estimated KL produced by differences in smoking habits in
different cohorts is likely to be a relatively small component of the very large variation in the KL
computed from different studies (see, e.g., Table A-l in Appendix A). In that case, KL computed
from different environments using the U.S. EPA model (Equation 7-2) would still be comparable
and it would be meaningful to search for measures of asbestos exposure ("exposure indices")
that better rationalize KL values (calculated using Equation 7-2) from different studies.3
Although the above argument suggests that efforts to reconcile KL values from different cohorts
is still a valid and useful exercise, even if the smoking/asbestos interaction is not multiplicative,
as assumed by the U.S. EPA model, further evaluation of the interaction between smoking and
asbestos exposure, including evaluation of different exposure-response models, is clearly
warranted. However, it must be recognized that the ability to conduct such evaluation will be
limited by the small number of asbestos-related epidemiology studies in which smoking data are
available, and possibly further limited by the ability to gain access to the raw data from these
studies. It should also be recognized that smoking data are not available for most of the studies
currently available (see, e.g., Table A-l in Appendix A), and, for most of those studies, the
published data are probably not sufficient to allow fitting of a model that incorporates any
smoking/asbestos interaction other than multiplicative.
3However, if the interaction is not multiplicative as assumed by the EPA model, it could be problematic to
use a KL value estimated from the model to quantify absolute asbestos risk (e.g., from lifetime exposure) while
taking into account smoking habits.
7.13
-------
7.2.1.4 Conclusions Concerning the Adequacy of the U.S. EPA Lung Cancer
Model
Three principle assumptions inherent to the U.S. EPA lung cancer model were considered in this
study to evaluate the overall adequacy of the model for describing the manner in which asbestos
exposure induces lung cancer. These are:
• that lung cancer RR is proportional to cumulative exposure (lagged 10 years);
• that risk remains constant following 10 years after the cessation of exposure; and
• that the interaction between smoking and asbestos exposure is multiplicative.
The findings from this evaluation are summarized briefly below. For convenience, timed-
dependence is addressed first.
In Section 7.2.1.2, the time response predicted by the lung cancer model (Equation 7-2) was
evaluated using the Wittenoom and South Carolina data, along with the data from a New Jersey
amosite factory (Seidman et al. 1986). Follow-up in the Wittenoom and South Carolina was
sufficient to permit evaluation of the adequacy of the model 40-45 years past the end of
exposure, and follow-up in the New Jersey factory permitted the model to be evaluated up
through 35+ years past the end of exposure.
The data from all three studies were consistent with the assumption inherent in the lung cancer
model that RR of lung cancer mortality remains constant after 10 years past the end of exposure.
That this result holds equally for a cohort exposed to chrysotile (South Carolina) as for cohorts
exposed to crocidolite (Wittenoom) and amosite (New Jersey factory) is particularly noteworthy,
because chrysotile fibers have been reported to be much less persistent than amphibole fibers
in vivo (see Section 6.2.4). The fact that the lung cancer risk in South Carolina remained
elevated 40^45 years following the cessation of exposure, along with the fact that the KL from
the South Carolina cohort is the largest KL value obtained from our review of 20 studies (see
Table A-l in Appendix A) suggests that carcinogenic potency of asbestos is not strongly related
to durability, at least for lung cancer (see Section 6.2).
Based on the results indicated in Section 7.2.1.1, the linear, RR exposure-response model for
lung cancer (Equation 7-2) provides an adequate description of the exposure-response
relationship from 20 studies. There is little evidence that exposure-response is sub-linear or
"threshold-like". However, in a number of these cases the fitted model predicts a background
response that is higher than that in the control population. If the background rate in the control
population was actually appropriate in these studies, the exposure-response relationship in these
studies would appear in some cases to be supra-linear. These results further reinforce the
recommendation of the expert panel (Appendix B) that evaluation of a broader range of
exposure-response models (including those that incorporate smoking-asbestos interactions that
are other than multiplicative) for lung cancer is appropriate.
7.14
-------
Although further evaluation of alternate models is warranted, the specific studies for which
oc=l .0 does not provide an adequate fit to the data do not appear to be related by mineral type,
industry, or size of study cohort and such studies total fewer than half of the studies evaluated.
Coupled with the additional evidence that the linear, RR model for lung cancer (Equation 7-2)
provides a good description of the time-dependence of disease mortality during and following
exposure and that it is not likely to grossly underestimate exposure despite limitation in the
manner that smoking is addressed (Section 7.2.1.3), this suggests that use of this model may be
adequate for estimating exposure-response factors for the existing studies, at least unless and
until a model that provides a superior fit to the data is identified.
Further support for the U.S. EPA lung cancer model is also provided by Stayner et al. (1997).
Stayner et al. evaluated the relative fit of a variety of additive, multiplicative, and more complex,
empirical models to the same raw data from the South Carolina textile plant that we evaluated
and conclusions drawn by these researchers generally parallel and reinforce those reported here.
Stayner et al. (1997) found a highly significant exposure-response relationship for lung cancer
and that a linear, RR model (similar to that described by Equation 7-1, except that they initially
assumed a lag of 15 years) provided the best overall fit to the data (among the additive and
multiplicative models evaluated). Moreover, the fit to the data was not significantly improved
by adding additional parameters, nor was there any indication of a significant interaction with
any of the covariates evaluated (age, race, sex, or year). These researchers did find an
interaction with time such that by applying different slopes for different latencies (15-29 years,
30-39 years, and >40 years), they were able to obtain a significantly better fit. However,
Stayner et al. (1997) did not evaluate a RR model with a single slope using a lag of 10 years.
Therefore, their analysis cannot be used to evaluate the effect of assuming different slopes for
different latencies with respect to the U.S. EPA model, which assumes a lag of 10 years.
Because we found that mortality among the South Carolina cohort is adequately described with
oc=1.0, the Stayner et al. (1997) study does not directly address questions concerning supra
linearity that have been suggested in our findings. Nevertheless, taken as a whole, the evidence
evaluated here suggests that the model described in Equation 7-2 may be adequate for evaluating
the relationship between asbestos exposure and lung cancer mortality among the existing
epidemiology studies. Certainly, no clearly superior model has yet been identified. Therefore,
the U.S. EPA lung cancer model was employed in this study for evaluating lung cancer risk
while we also recognize that further evaluation of alternate models is warranted. The primary
obstacle to more fully evaluating alternate models has been and continues to be lack of access to
the raw data for a greater number of cohorts.
7.2.2 Estimating KL values from the Published Epidemiology Studies
The U.S. EPA model for lung cancer (Equation 7-2) was applied to each of the available
epidemiology data sets to obtain study-specific estimates for the lung cancer exposure-response
coefficient, KL. Based on the results presented in Section 7.2.1.4, while there is some evidence
that models in addition to the U.S. EPA lung cancer model should be explored, there is little
indication that the current model does not provide an adequate description of lung cancer
mortality that is sufficient to support general risk assessment. The set of KL values derived from
7.15
-------
available epidemiology studies are presented in Table 7-6, which is a reproduction of Table A-l
in Appendix A.
In Table 7-6, Column 1 lists the fiber types for the various studies, Column 2 lists the exposure
settings (industry type), and Column 3 indicates the specific locations studied. Column 4
presents the best estimate of each KL value derived for studies, as reported in the original 1986
Health Effects Update and Column 5 presents the reference for each respective study. Columns
6, 7, and 8 present, respectively: the best-estimates for the KL values derived in our evaluation
for all of the studies currently available (including studies corresponding to those in the 1986
Health Effects Update); a statistical confidence interval for each KL value (derived as described
in Appendix A); and an uncertainty interval for each KL value (also derived as described in
Appendix A). The reference for the respective study from which the data were derived for each
KL estimate is provided in the last column of the table. To assure comparability across studies,
values for all studies (even those that have not been updated since their inclusion in the 1986
Health Effects Update) were re-derived using the modified procedures described in Appendix A.
As explained in Appendix A, the uncertainty intervals for KL values (and corresponding intervals
for KM values, the exposure-response coefficients for mesothelioma) are intended to reflect, in
addition to statistical variation, other forms of uncertainty that are difficult to quantify, such as
model uncertainty and uncertainty in exposure estimates. We interpret these informally as
providing a range of KL values that are reasonable, based on the data available for a given study.
Accordingly, if uncertainty intervals for two KL values do not overlap, these two underlying sets
of data are considered to be incompatible. Potential reasons for such incompatibility include the
possibility that the KL values are based on an exposure measure that does not correlate well with
biological activity of asbestos. One of the goals of this report is to determine an exposure index
that correlates better with biological activity, and consequently brings the KL values into closer
agreement. The degree of overlap of the corresponding uncertainty intervals provides an
indication of the extent to which a groups of KL values are in agreement.
The KL values derived in this study and the corresponding values derived in the original 1986
Health Effects Update generally agree at least to within a factor of 3; a couple vary by a factor of
4; one varies by a factor of 5; and one varies by a factor of about 15. However, in every case the
KL from the 1986 update lies within the uncertainty interval for the KL derived in the current
update.
Perhaps the most interesting of the changes between the 1986 KL value estimates and the current
KL value estimates involves the friction products plant in Connecticut (McDonald et al. 1984).
Although a relatively small, positive exposure-response was estimated from this study in the
1986 Health Effects Update, the best current estimate is that this is essentially a negative study
(no excess risk attributable to asbestos). The difference derives primarily from allowing a to
vary in the current analysis. The exposure groups in this cohort do not exhibit a monotonically
increasing exposure-response relationship and, in fact, the highest response is observed among
the group with the lowest overall exposure. As indicated in Appendix A, lack of a
monotonically increasing exposure-response relationship is a problem observed in several of the
studies evaluated.
7.16
-------
Table 7-6. Lung Cancer Exposure-Response Coefficients (KL) Derived from Various Epidemiological Studies
Fiber Type
Chrysotile
Crocidolile
Amosite
Tremolite
Mixed
Operation
Mining and
Milling
Friction
Products
Cement
Manufacture
Textiles
Mining and
Milling
Insulation
Manufacture
Vermiculite
Mines and Mills
Friction
Products
Cohort
Quebec mines
and mills
Italian mine
and mill
Connecticut
plant
New Orleans
plants
South
Carolina plant
Wirtenoom
Patterson, NJ
factory
Tyler, Texas
factory
Libby,
Montana
British factory
EPA (1986)
KL*100 Reference
0.06 McDonald
etal. 1980b
0.17 Nicholson et al.
1979
0.081 Piolatto et al.
1990
0.01 McDonald
etal. 1984
2.8 Dement et al.
1983b
2.5 McDonald
etal. 1983a
4.3 Seidman 1984
0.058 Berry and
Newhouse
1983
This Update
KL*100
0.029
0.051
0
0.25
2.1
1
0.47
1.1
0.13
0.51
0.39
0.058
90%
Confidence
Interval
(0.019,0.041)
(0, 0.57)
(0,0.17)
(0, 0.66)
(1.2,3.4)
(0.44, 2.5)
(0.17,0.87)
(0.58,1.9)
(0, 0.6)
(0.11,2.0)
(0.067, 1.2)
(0, 0.8)
Uncertainty
Interval'
(0.0085,
0.091)
(0,1.1)
(0, 0.62)
(0, 1.5)
(0.81,5.1)
(0.22, 4.9)
(0.084, 1.7)
(0.17,6.6)
(0, 1.8)
(0.049, 4.4)
(0.03, 2.8)
(0, 1.8)
Reference
Liddell et al.
1997
Piolatto et al.
1990
McDonald
etal. 1984
Hughes et al.
1987
Dement et al.
1994b
McDonald
1983a
de Klerk et al.
1994°
Seidman et al.
1986
Levin et al.
1998
Amandus and
Wheeler 1987
McDonald
etal. 1986
Berry and
Newhouse
1983
7.17
-------
Table 7-6. Lung Cancer Exposure-Response Coefficients (KL) Derived from Various Epidemiological Studies (continued)
Fiber Type Operation
Cement
Manufacture
Factory workers
Insulation
Application
Textiles
Cohort
Ontario
factory
New Orleans
plants
Swedish plant
Belgium
factory
US. retirees
U.S.
insulation
workers
Pennsylvania
plant
Rochedale
plant
EPA (1986)
KL*100
4.8
0.53
0.49
0.75
1.4
1.1
Reference
Finkelstein
1983
Weill 1979,
1994
Henderson and
Enterline 1979
Seilkoffetal.
1979
McDonald
etal. 1983b
Peto 1980a
This Update
KL*100
0.29
0.25
0.067
0.0068
0.11
0.18
1.8
0.41
90%
Confidence
Interval
(0, 3.7)
(0, 0.66)
(0, 3.6)
(0,0.21)
(0.041,0.28)
(0.065, 0.38)
(0.75, 4.5)
(0.12,0.87)
Uncertainty
Interval"
(0, 22)
(0, 1.5)
(0, 14)
(0, 0.84)
(0.011,1.0)
(0.012,2.1)
(0.2, 16)
(0.046, 2.3)
Reference
Finkelstein
1984
Hughes et al.
1987
Albin et al.
1990
Laquet et al.
1980
Enterline et al.
1986
Seilkoffand
Seidman 1991
McDonald
etal. 1983b
Peto et al.
1985
"Uncertainty Interval formed by combining 90% confidence interval with uncertainty factors in Table A-3.
bWith supplemental raw data from Terri Schnorr (NIOSH) and Dement
'With supplemental unpublished raw data with follow-up through 2001
7.18
-------
Among the KL values derived in the current study, the lowest and highest of the best-estimate
values differ by a factor of 300 (excluding the negative study of the Connecticut friction
products plant, which would make the spread even larger) and several pairs of uncertainty
intervals have no overlap. For example, the KL uncertainty interval for the chrysotile miners in
Quebec lies entirely below the corresponding intervals for chrysotile textile workers (in either of
the two studies for South Carolina, which are of highly redundant cohorts in the same plant), for
textile workers in the Pennsylvania plant, and for amosite insulation manufacturers (in the
Seidman study (1986).
The KL values and the associated uncertainty intervals are plotted in Figure 7-1. Each exposure
environment is plotted along the X-axis of the figure and is labeled with a 4-digit code that
indicates fiber type (chrysotile, mixed, crocidolite, or tremolite), industry (mining, friction
products, asbestos-cement pipe, textiles, insulation manufacturing, or insulation application); and
a 2-digit code indicating the study from which the data were derived. A key is also provided. In
Figure 7-1, the chrysotile studies are grouped on the left, amphibole studies are grouped on the
right, and mixed studies are in the middle.
For studies conducted at the same facility (generally among highly overlapping cohorts), such as
the Dement et al. (1994) and the McDonald et al. (1983a) studies of the same South Carolina
textile facility, a single study was selected for presentation in Figure 7-1. Thus, for South
Carolina, the Dement et al. (1994) study is presented because we had access to the raw data for
this study. It is also a newer study. Similarly, the Amandus and Wheeler (1987) study was
selected to represent the Libby Vermiculite site over the other study at this facility (McDonald et
al. 1986). The effects of such selection is expected to be small in any case because the KL values
estimated for the individual studies in each pair vary only by a factor of 2.
Comparisons of KL values across the available studies are instructive. Within chrysotile studies
alone (and excluding the negative friction products study), lowest and highest KL values vary by
approximately a factor of 70. Moreover, as previously indicated, the uncertainty intervals for the
lowest (non-zero) value (for Quebec miners) and the highest value (textile workers) have no
overlap. Uncertainty intervals for the negative friction product study and the other estimates for
chrysotile do overlap, primarily due to the wide confidence interval associated with the negative
study.
Among the apparent variations, differences in lung cancer potency observed among Quebec
miners versus that observed among South Carolina textile workers has been the subject of much
discussion and evaluation, which is worthy of review (Appendix D). The inability to reconcile
these differences, appears to be among the biggest obstacles to reliably estimating an overall
chrysotile dose-response coefficient for lung cancer.
As indicated in Appendix D, the leading hypothesis for the apparent differences in lung cancer
risk per unit of exposure observed between chrysotile mining and textile manufacturing is the
relative distribution of fiber sizes found in dusts in these industries. Evidence from several
studies indicates that textile workers were exposed to dusts containing substantially greater
concentrations of long structures than dusts to which miners were exposed. Thus, the effects of
fiber size is considered further in Section 7.4.
7.19
-------
Figure 7-1:
Plot of Estimated KL Values and Associated Uncertainty Intervals by Study Environment
1000,,
100,
CM
o
0.1,
0.01,
1E-3
T- co ^- 10 CD r- co o o •<- co
5 ^U_QL|— u_ o_ o_
O Q0o
IO CO Is- 00 O> O
T^ T- •«- T- ^ ^
^ 11 ?>
-------
Key for Figures 7-1 through 7-6 Code
Fiber Types Study Environment
(First Digit of Code) (Second Digit of Code)
A = amosite A = insulation application
C = chrysotile F = friction products manufacturing
M = mixed fibers I = insulation manufacturing
R = crocidolite M = mining
T = tremolite (in vermiculite) P = ac pipe manufacturing
T = textile manufacturing
X = misc. products manufacturing
Study Cohorts
(Last 2 digits)
1 = Quebec miners (Liddell et al. 1997)
2 = Quebec miners (Liddell et al. 1997, raw data)
3 = Italian miners (Piolatto et al. 1990)
4 = Connecticut friction product workers (McDonald et al. 1984)
5 = New Orleans ac pipe manufacturers (Hughes et al. 1987)
6 = South Carolina textile manufacturers (Dement et al. 1994, raw data)
7 = British friction product manufacturers (Berry and Newhouse 1983)
8 = Ontario ac pipe manufacturers (Finkelstein 1984)
9 = New Orleans ac pipe manufacturers (Hughes et al. 1987)
10 = Swedish ac pipe manufacturers (Albin et al. 1990)
11 = Belgium ac pipe manufacturers (Laquet et al. 1980)
13 = Retired factory workers (Enterline et al. 1986)
14 = Factory workers (Liddell et al. 1997)
15 = Insulation appliers (Selikoff and Seidman 1991)
16 = Pennsylvania textile workers (Mcdonald et al. 1983b)
17 = British textile workers (Peto et al. 1985)
18 = Australian crocidolite miners (de Klerk, unpublished, raw data)
19 = New Jersey insulation manufacturers (Seidman et al. 1986)
20 = Texas insulation manufacturers (Levin et al. 1998)
21 = Libby vermiculite miners (Amandus and Wheeler 1987)
7.21
-------
Ignoring the negative Connecticut friction products study, the range of KL values observed
across chrysotile studies (70) appears to be substantially narrower than the range observed across
all studies (300). However, if the nearly negative study of the Belgium asbestos-cement pipe
manufacturers is also removed, the range observed for chrysotile studies is almost identical to
the range observed across all studies. This is because the KL values for Quebec represents the
low extreme of both ranges and South Carolina represents the high extreme of both ranges.
Among "pure" amphibole studies, the lowest and highest of the best-estimate KL values vary by
a factor of approximately 9 and the two extremes both derive from within the same industry
(amosite insulation). The two studies in question are of amosite insulation manufacturing plants
(one in Patterson, New Jersey, and one in Tyler, Texas) that utilized the same equipment
(literally) and apparently had similar sources of asbestos (South Africa). Despite the 9-fold
difference in KL values, the uncertainty intervals for these two estimates have substantial
overlap. If the arithmetic mean of the values for the two amosite insulation studies is used, KL
values estimated, respectively, for crocidolite mining, amosite insulation manufacture, and
mining of vermiculite contaminated with tremolite vary by less than a factor of 2. However, this
is possibly fortuitous, given the magnitude of the associated uncertainty intervals (Figure 7-1).
As indicated in Appendix D, for example, it is possible that mining studies tend to exhibit low
KL values relative to studies of asbestos products industries. This is due to the presence of large
numbers of cleavage fragments in the dusts that may not contribute to biological activity
(because the majority of these may not exhibit the requisite size to be biologically active) but
which, nevertheless, are included in estimates of asbestos concentrations in the original
epidemiology studies. If the KL estimates for Libby and Wittenoom could be adjusted for this
effect, they might be closer in value to that obtained from the Seidman study.
It is also instructive to compare variation within and between industries. Within industries
(especially for a single fiber type), the data are limited. The studies from the two chrysotile
mines (Quebec and Italy) show remarkably close agreement, varying by less than a factor of 2.
The three studies involving mining of amphibole (Wittenoom and the two studies of Libby) also
vary by less than a factor of 2. However, the mean of the amphibole mining group is
approximately 10 times the mean of the chrysotile mining group. Moreover, based on inspection
of their respective uncertainty intervals, the KL values for chrysotile mining in Quebec and
Crocidolite mining in Wittenoom appear to be incompatible.
Across all asbestos types (including mixed), the asbestos-cement pipe industry shows the
greatest variation, including a nearly negative study (best estimate KL=0.000068) and four more
studies with KL values that range up to 0.0029 producing a variation within this industry of a
factor of 40. The friction products industry includes one negative and one positive study. Better
agreement is observed among textiles. The two mixed textile plants show KL values that vary by
no more than a factor of 5 from each other and from the KL for the South Carolina chrysotile
textile plant.
7.22
-------
7.3 MESOTHELIOMA
The model proposed in the Airborne Health Assessment Update (U.S. EPA 1986) to describe the
mortality rate from mesothelioma in relation to asbestos exposure assumes that the mortality rate
from asbestos-induced mesothelioma is independent of age at first exposure and increases
according to a power of time from onset of exposure, as described in the following relationship:
IM = KM f [(T-10)3-(T-10-d)3] forT>10 + d (Eq. 7-6)
= Kwf(T-10)3 forlO + d>T>10
= 0 forlO>T
where:
IM is the mesothelioma mortality rate at T years from onset of exposure to
asbestos for duration d and concentration f;
KM is the proportionality constant between exposure and mesothelioma
response and represents the potency of asbestos;
A more general expression that holds for variable exposure is given by
IM = SKW-IO-tfdx (Eq.7-7)
where f(x) is the concentration of fibers at time x following the beginning of exposure. This
expression reduces to Equation 7-6 when exposure is constant (see Appendix A).
7.3.1 The Adequacy of the Current U.S. EPA Model for Mesothelioma
Access to the raw epidemiology data from a few key studies allowed us to evaluate the adequacy
of the U.S. EPA model (Equation 7-6) for describing the time-dependence for mesothelioma in
asbestos-exposed cohorts. For this analysis, the raw data from a cohort of chrysotile miners in
Quebec was graciously provided by Drs. Douglass Liddell and Corbett McDonald (described in
Liddell et al. 1997), the raw data for the cohort of crocidolite miners in Whittenoom, Australia
was graciously provided by Dr. Nick de Klerk (unpublished) and the raw data for the cohort of
chrysotile textile workers (described by Dement et al. 1994) was graciously provided by Ms.
Terri Schnoor of NIOSH and Dr. John Dement of Duke University. The Whittenoom cohort was
originally described by Armstrong et al. (1988), but the data provided by Dr. de Klerk included
additional follow-up through 1999.
To identify potential effects due to varying statistical procedures, different methods for fitting
the U.S. EPA mesothelioma model to epidemiological data were evaluated. In this evaluation,
three methods were used to fit the U.S. EPA mesothelioma model to data from Wittenoom.
7.23
-------
In the first approach the data were categorized in a manner often available in published form, so
this method mimics the method generally used when raw data are not available. The observed
mesotheliomas and person-years of observation were categorized by time since first exposure,
and the mean exposure level and duration of exposure were calculated for each such category.
The U.S. EPA model was then applied to such data using the approach for the typical situation
(as described in Appendix A) and results for the Wittenoom cohort are presented in Table 7-7.
The KM value estimated for Wittenoom using this approach is 7.15xlO'8 (90% CI: 6.27x10'8,
8.1 IxlO"8). The fit of this model to data categorized by time since first exposure is good
(p=0.65).
For most of the published epidemiology data sets, the average level and duration of exposure for
individual time-since-first-exposure categories are not provided and have to be estimated from
cruder data representing cohort-wide averages. Thus for most of the epidemiology data sets, the
calculation of KM is based on cruder information than the calculation presented in Table 7-7.
Table 7-7. Fit of EPA Mesothelioma Model to Observed Mesothelioma Mortality Among
Wittenoom, Australia Miners (Deklerk 2001) Categorized by Years Since First Exposure
Years Since First
Range
0-5
5-10
10-15
15-20
20-25
25-30
30-40
40-100
Total
Estimates of KM
KM=7.15xlO-8
90% CI: (6.27x10''
Exposure
Average
0
0.19
0.69
1.6
3.3
6.2
11.8
21.5
!, S.llxlO'8)
Average
Duration
(years)
0.643
0.91
0.958
0.970
0.953
0.957
1.05
1.13
Average
Concentration
(f/ml)
30.2
30.2
30.0
29.7
29.5
29.3
29.0
28.1
Goodness of Fit
Observed
Deaths
0
0
1
5
20
25
90
23
164
P-value
Predicted Deaths
by Model
0.0
0.0
0.6
6.6
17.6
30.1
78.1
31.1
164.0
0.65
A second approach to fitting the U.S. EPA model to epidemiology data exploits the fact that the
mesothelioma model (Equation 7-7) expresses the mesothelioma mortality rate as the product of
KM and an integral involving the exposure pattern, the time of observation, and the 10-year time
lag, but not any parameters that require estimation.
The value of this integral was calculated for each year of follow-up of each subject. Person-
years of follow-up and mesothelioma deaths were then categorized according to the values of the
7.24
-------
integral, and the average value of the integral determined for each category. Results are
nrsspntfvl in Tahlp 7-8
presented in Table 7-8.
Table 7-8. Fit of EPA Mesothelioma Model to Observed Mesothelioma Mortality Among
Wittenoom, Australia Miners (Deklerk 2001) Categorized by Average Value of Integral
(Equation 6-12)
Average Value of Integral
0
28.4
206.3
732.3
2,038.1
5,153.7
12,385.8
30,639.7
144,801
Total
Observed Deaths
0
0
1
4
10
23
32
27
67
164
Predicted Deaths by Model
0.0
0.0
0.3
1.0
2.9
6.8
13.8
25.8
113.3
164.0
Goodness of Fit P-value O.OOO 1
Estimates of KM
KM=9.00xlO'8
90% CI: (7.89x10-8, 10.2x10'8)
KM was estimated by maximum likelihood from Table 7-8 assuming that the observed numbers
of cancer deaths in the different categories of the integral were independently Poisson distributed
with a mean equal to the mean value of the integral for that category times KM. The value of KM
obtained in this fashion was KM=9.00xl(r8 (90% CI: 7.89xlQ-8,10.2xlO-8). This estimate is
about 25% larger than obtained with the data categorized by time since first exposure, although
confidence intervals obtained from the two procedures overlap. The fit of the model to data
categorized by the integral is poor (p<0.00001), as the model predicts too few mesotheliomas for
small values of the integral and too many mesotheliomas for large values of the integral.
A third method of estimating KM (termed the "exact method") employs a likelihood that does not
involve any categorization of data. With this method, the hazard function, h(t)=IM(t) and the
corresponding survival function (probability of surviving to age t without death from
mesothelioma in the absence of competing causes of death), S(t)=exp(-J0' h(s)ds), are computed
for time at the end of follow-up. The contribution to the likelihood of a subject who died of
mesothelioma t years after beginning of follow-up is h(t)*S(t), and the contribution of a subject
whose follow-up was not terminated by death from mesothelioma is S(t). The complete
likelihood is the product of such terms over all members of the cohort. The estimate of KM
7.25
-------
obtained by maximizing the logarithm of this likelihood was 7.95x10"8 (90% CI: 7.0x10"8,
9.0x10'8).
We consider the exact method of computing KM to be the most accurate and results from this
method are reported in the summary table for mesothelioma (Table 7-9, which is a reproduction
of Table A-2). The Quebec and South Carolina data sets were thus evaluated using the exact
method. However, it is noteworthy that the two other methods described above, one of which is
often applied to published data, give similar estimates of KM, at least for this data set.
The Quebec cohort was subdivided into three subcohorts, believed to correspond to differing
amounts of amphibole exposure due to tremolite contamination of the ore at different mining
locations, and to use of some imported commercial amphibole at one factory location (Liddell et
al. 1997). Location I consisted of workers at the mine at Asbestos where the ore reportedly had
less tremolite contamination. Locations 3 and 4 consisted of workers at the large central mine
and at smaller mines, respectively, near Thetford, where the ore was more heavily contaminated
with tremolite. Location 2 consisted of workers at an asbestos products factory at Asbestos,
which processed some commercial amphibole fibers in addition to chrysotile. The exact method
of calculating KM produced the following estimates: Location 1 (8 cases): KM=1.3xlO"'° (90%
CI: O.SxlO'10, 4.9x10-'°); Location 2 (5 cases): KM=9.2xlO-'°, (90% CI: 2.0x10-'°, 35x10-'°);
Locations 3 and 4 (22 cases): KM=2.1xlO-'° (95% CI: 0.65x10-'°, 6.5X10'10).
The relative magnitudes of these estimates track with the relative amounts of amphibole
exposure estimated for these locations (Liddell et al. 1997), which is consistent with the
hypothesis that the mesothelioma risk in this cohort is due, at least in large measure, to exposure
to amphiboles.
There were only two confirmed mesothelioma deaths in the South Carolina cohort and four
additional suspected deaths. These were too few to permit detailed analysis. Based on both
confirmed and suspected mesothelioma deaths, the exact method of analysis gave an estimate of
KM=0.43xlQ-8, 90% CI: (0.20xlQ-8, 0.79xlO'8). Using only the two confirmed mesotheliomas,
the same analysis yielded KM=0.14xlO-8, 90% CI: (0.034xlO'8, 0.38xlO'8). Very similar
estimates were obtained by estimating KM from data categorized by time since first exposure and
fitting a linear model to the categorized value of the integral in the definition of the U.S. EPA
model. Thus, for this cohort comparable KM values are estimated no matter which of the three
methods described above are used for fitting the U.S. EPA mesothelioma model to the
epidemiology data.
7.26
-------
Table 7-9. Mesothelimoa Exposure-Response Coefficients (KM) Derived from Various Epidemiological Studies
Fiber Type
Chrysotile
Crocidolile
Amosite
Mixed
Operation
Mining and Milling
Friction Products
Cement
Manufacture
Textiles
Mining and Milling
Insulation
Manufacture
Cement
Manufacture
Factory Workers
Insulation
Application
Textiles
EPA
(1986)
Cohort KM*100 Reference
Asbestos, Quebec
Thedford Mines
Connecticut plant
New Orleans
plant
South Carolina
plant
Wittenoom
Patterson, NJ 3.2 Seidman 1984
factory
Ontario factory 12 Finkelstein 1983
New Orleans
plant
Asbestos, Quebec
U.S. insulation 1.5 Seilkoff et al.
workers 1979
Pennsylvania
plant
Rochedale plant 1 Peto 1980; Peto
etal. 1982
This
Update
KM*100
0.013
0.021
0
0.2
0.25
0.088
7.9
3.9
18
0.3
0.092
1.3
1.1
1.3
90%
Confidence
Interval
(0.0068, 0.022)
(0.014, 0.029)
(0,0.12)
—
(0.034, 0.79)
(0.0093, 0.32)
(7,9)
(2.6, 5.7)
(13, 24)
(0.04,0.18)
(1.2, 1.4)
(0.76, 1.5)
(0.74,2.1)
Uncertainty
Interval"
(0.003, 0.049)
(0.0065, 0.065)
(0, 0.65)
(0.033, 1.2)
(0.023, 1.2)
(0.0025, 1.2)
(3.5, 18)
(0.74, 20)
(2, 160)
(0.089, 1)
(0.018,0.39)
(0.25, 6.5)
(0.17,6.6)
(0.28, 5.6)
Reference
Liddell et al. 1997b
Liddelletal. 1997b
McDonald et al. 1984
Hughes et al. 1987
Dement etal. 1994°
McDonald et al.
1983a
de Klerk etal. 1994"
Seidman et al. 1986
Finkelstein 1984
Hughes et al. 1987
Liddell et al. 1997"
Seilkoff and Seidman
1991
McDonald et al.
1983b
Peto etal. 1985
"Uncertainty Interval formed by combining 90% confidence interval with uncertainty factors in Table A-3.
bWith supplemental raw data from Liddell
°With supplemental raw data from Terri Schnorr (NIOSH) with Dement
dWith supplemental unpublished raw data with follow-up through 2001
7.27
-------
7.3.1.1 Time Dependence
The U.S. EPA mesothelioma model (Equation 7-6 or 7-7) was next evaluated to determine
whether it adequately describes the time-dependence of mesothelioma mortality following
cessation of exposure in the Wittenoom and Quebec cohorts. The small number of
mesotheliomas observed among the South Carolina cohort precluded a meaningful evaluation of
this issue for that cohort.
For times since first exposure longer than 10 years past the end of exposure the mesothelioma
model (Equation 7-6) can be rewritten as
IM = 3 * KM * f * d * (t - 10)2 * {1 - 3 * [d / (t - 10)] + [d / (t - 10)]2} (Eq. 7-8)
From this expression we see that, when time since first exposure lagged 10 years, (t-10), is large
compared to duration of exposure, (d), the model predicts that the mesothelioma mortality rate is
approximately proportional to the product of cumulative exposure (the exposure level, f, (f/ml)
times the duration of exposure, (d) and the square of time since first exposure lagged 10 years.
Thus, the model predicts that the mesothelioma mortality will increase indefinitely with age as
the square of time since first exposure lagged 10 years. The availability of raw data from the
Wittenoom and Quebec cohorts provides an opportunity to evaluate this assumption.
Table 7-10 shows the fit of the U.S. EPA mesothelioma model to Wittenoom data characterized
by time since last exposure, based on the KM estimated from the exact analysis. There is no
indication from this table that the mesothelioma mortality rate declines after the cessation of
exposure, or that the model over-predicts the mesothelioma risk at long times after the cessation
of exposure. In fact, the model under-predicts the number of deaths at the longest times, as it
predicts 76.8 deaths after 30 years from the end of exposure, whereas 96 were observed.
Table 7-11 shows the observed number of mesothelioma deaths in the three separate locations at
Quebec, categorized by time since last exposure, and compared with the predicted numbers
obtained using the KM values obtained using the exact fitting method. From all three locations
combined the model predicts 10.6 deaths from mesothelioma after more than 30 years following
the cessation of exposure, whereas 10 were observed. Although the small numbers of
mesotheliomas make it difficult to draw definite conclusions about the adequacy of the model,
there is little evidence that the model under or over-predicts the numbers of mesotheliomas at
long periods after the end of exposure.
7.3.1.2 Exposure Dependence
The lack of fit of the mesothelioma model (Equations 7-7 and 7-8) to the Wittenoom data
categorized by the value of the integral in Equation 7-8 (Table 7-8) suggests that the Wittenoom
data may not be consistent with the assumption that the mortality rate is linear in the intensity of
exposure (f/ml). Specifically, the mesothelioma model predicts that for fixed time since first
exposure and duration of exposure the mesothelioma mortality rate varies linearly with f, the
asbestos air concentration. To test this prediction, expected numbers of mesothelioma deaths
were calculated for each of four categories of asbestos air concentrations while controlling for
7.28
-------
both time since first exposure and duration of exposure.4 Person-years in the first 10 years
following the beginning of exposure were ignored since no mesothelioma deaths occurred in this
time interval, as predicted by the model. The relatively few workers who were employed for
longer than 5 years were also excluded from this analysis (exposure durations were generally
quite short in this cohort, with the average employment duration being <1 year), which means
that all follow-up in this analysis occurred after exposure had ended. Results of this analysis are
shown in Table 7-12.
Table 7-10. Fit of EPA Mesothelioma Model to Observed Mesothelioma Mortality
Among Wittenoom, Australia Miners (Deklerk 2001) Categorized by Years Since Last
Exposure
Years Since
Range
0-1
1-5
5-10
10-20
20-30
30^0
40+
30+
Total
Last Exposure
Average
0.3
3
7.5
14.9
24.7
34
43.6
Average
Value
of Integral
35
72
327
4025
18058
39802
57952
Observed
Deaths
0
0
0
10
58
79
17
96
164
Predicted
Deaths
by Model
0
0.1
0.6
14.4
52.9
61.3
15.5
76.8
144.9
Observed/
Predicted
0
0
0
0.7
1.1
1.3
1.1
1.3
Goodness of Fit P-value
0.21
4In this analysis the Wittenoom data were categorized by time since first exposure (10-20,20-30, 30-40,
and 40+ years), exposure intensity (0-15, 15-30, 30-60, and 60+ f/ml), and duration of exposure (0-1 and 1 + years).
This categorization was facilitated by the facts that in the subcohort being analyzed, exposure had ended prior to the
beginning of follow-up, and in the Wittenoom data base exposure intensity was assumed to be constant throughout
employment. Within each of the eight [time since first exposure] x [duration of exposure] categories the total
number of mesothelioma deaths were allocated to the various exposure intensity sub-categories in proportion to the
product of the average exposure intensity times the person-years of observation in each sub-category, thereby
producing the expected number of deaths in each sub-category under the assumption that the response was linear in
exposure intensity within each [time since first exposure] x [duration of exposure] category, as predicted by the
mesothelioma model. These expected deaths and the corresponding numbers of observed deaths were summed
across time-since-first-exposure and duration categories to yield observed and expected deaths categorized only by
exposure intensity
7.29
-------
Table 7-11. Fit of EPA Mesothelioma Model to Observed Mesothelioma Mortality
Among Quebec Miners (Liddell 2001) in Each of Three Mining Areas, Categorized by
Years Since Last Exposure
Years
Since Last
Exposure
Average
Value
of Integral
Person-Years
Observed
Deaths
Predicted
Deaths
by Model KM
Location 1
0-10
10-20
20-30
30^0
40-50
50+
30+
Total
94,267.5
119,190
119,971
181,470
278,118
339,160
77,008
27,225
22,831
18,321
11,846
6,673
3
2
0
1
2
0
3
8
3.1 1.34x10-'°
1.4
1.2
1.4
1.4
1.0
3.7
9.3
Location 2
0-10
10-20
20-30
30-40
40-50
50+
30+
Total
56,377.5
56,445.7
58,307.7
78,685.8
88,541.2
59,590.4
15,065
4,835
3,991
3,129
1,844
795
0
2
0
2
0
1
3
5
2.5 9.55x10-'°
0.8
0.7
0.7
0.5
0.1
1.4
5.4
Locations 3 and 4
0-10
10-20
20-30
30^0
40-50
50+
30+
Total
194,913
262,898
179,052
251,766
348,264
298,743
95,299
23,885
18,777
14,311
8,800
4,451
13
5
0
0
1
3
4
22
12.7 2.18x10-'°
4.3
2.3
2.5
2.1
0.9
5.5
24.8
7.30
-------
Table 7.12. Comparison of Wittenoom, Australia (DeKlerk 2001) Mesothelioma Deaths
to Predicted Deaths Assuming Risk Varies Linearly with Exposure Intensity After
Controlling for Years Since First Exposure and Duration of Exposure
Intensity (f/ml) Mesothelioma Deaths
Range
0-15
15-30
30-60
>60
Total
Average
9.7
17.0
50.3
100.2
Person-Years
50,736
23,881
18,166
13,353
Observed
32
51
27
40
150
Predicted
19.5
22.4
41.9
66.3
150.0
Goodness of Fit P-value <0.0001
Table 7-12 shows that the assumption that the mesothelioma risk varies linearly with exposure
intensity leads to a 2-fold under-prediction of risks for exposure intensities below 60 f/ml (83
deaths compared to only 41.8 predicted) and a corresponding over-prediction for exposure
intensities above 60 f/ml (only 67 deaths compared to 108.2 predicted). Thus, instead of risk
varying linearly with exposure intensity, Table 7-12 indicates that the exposure response is
supra-linear, with lower fiber intensities being more potent per f/ml.
7.3.1.3 Discussion of Adequacy of Mesothelioma Model
The mesothelioma model (Equations 7-6 and 7-7) provides adequate fits to each of the three data
sets evaluated (Wittenoom, Quebec and South Carolina) when the data are categorized by time
since first exposure. The value of KM estimated from the cohort of crocidolite miners
(Wittenoom) was largest and was about 60-fold larger than the KM estimated from the South
Carolina chrysotile textile workers (based only on confirmed mesotheliomas in South Carolina)
and more than 100-fold larger than the estimates obtained from the Quebec chrysotile miners
who did not work in the factory that utilized crocidolite. This is consistent with numerous
indications from the literature that crocidolite is more potent than chrysotile in causing
mesothelioma. The relative magnitudes of the KM for the Quebec data estimated from three
locations track with the relative amounts of amphibole exposure estimated for these locations,
which is also consistent with the hypothesis that the mesothelioma risk is greater from amphibole
exposure than from chrysotile exposure. Despite the very few mesotheliomas in the South
Carolina cohort, the KM estimated from these data is larger than those from Quebec, although the
discrepancy is not as large as estimated for lung cancer.
The Wittenoom and Quebec data were evaluated further to see if they were consistent with the
prediction of the mesothelioma model that risk continues to increase indefinitely after exposure
has ceased. By comparing the observed number of mesothelioma deaths to the number predicted
by the mesothelioma model at various times since the cessation of exposure, no evidence was
found that mesothelioma risk dropped off below that predicted by the model, at least up to 40 to
50 years after the cessation of exposure. To the contrary, in Wittenoom there was some evidence
that the model under-predicted at the longest times since the end of exposure, as past 30 years
7.31
-------
from the cessation of exposure there were 96 observed mesothelioma deaths compared to only
76.8 predicted by the model.
Chrysotile is much more soluble than crocidolite and consequently a chrysotile fiber exhibits a
much shorter residence time in the body than a comparable-sized crocidolite fiber. Based on
in vitro studies a chrysotile fiber with a diameter of 1 ^m will dissolve in body fluid in
approximately 1 year whereas a 1-um crocidolite fiber will take 60 years to dissolve (see Section
6.2.4). It is noteworthy that despite the short residence time of chrysotile fibers, mesotheliomas
deaths have occurred in the Quebec cohort more than 50 years following the cessation of
exposure. This suggests that either these mesotheliomas are the result of amphibole
contamination of the ore, or else long residence times for inhaled fibers are not necessary for the
production of mesotheliomas.
The mesothelioma model predicts that risk is proportional to the intensity of exposure (Equation
7-6) and, at long times past the end of exposure, to cumulative exposure (Equation 7-8). Two
analyses of the Wittenoom mesothelioma data suggest that the assumption of a linear exposure-
response may not be valid. First, whereas the model is linear in the value of the integral in
Equation 7-7, a very poor fit was obtained when the data were categorized according to the value
of the integral (Table 7-8). Second, an analysis that categorized the data by intensity of
exposure, while controlling for both duration of exposure and time since last exposure, also
provided a poor fit (Table 7-12). Both of these analyses exhibit a supra-linear exposure-response
in which less intense lower exposures are more potent per f/ml than more intense ones.
In light of these findings, it is interesting that supra-linearity in the exposure-response
relationship for lung cancer has also been suggested by fits of data in several studies (see Section
7.2.1.1). Although the effects for lung cancer are difficult to separate from the confounding
effects from smoking, common suggestions of supra linearity for both disease endpoints
certainly indicate a need to evaluate the nature of the exposure-response models for asbestos in
greater detail. This is consistent with the recommendations of the expert panel (Appendix B)
that evaluation of a broader range of exposure-response models for mesothelioma is appropriate.
This analysis of the Wittenoom data appear to be one of very few that provide information on the
shape of the exposure-response relationship for mesothelioma. It is possible that systematic
errors in the exposure data for Wittenoom could have resulted in an apparent supra-linear
exposure response. Like most asbestos-exposed cohorts, the estimated exposures for Wittenoom
are uncertain. If higher estimated intensities are overestimates, a linear exposure response would
appear to be supra-linear, and a linear fit to such data would underestimate the true KM.
In addition, errors in exposure measurement even if unbiased, can tend to make a linear dose-
response appear supra-linear (Crump 2003). Even if the supra-linear exposure response in the
Wittenoom data is real, the exposure-response at lower doses is likely to be linear, but with a
larger KM value than was obtained in the fit to the complete data set. If this is the case, our
analysis (Table 7-12) suggests that the KM for the lower exposures is about a factor of two larger
than the value estimated from the complete cohort. A factor of 2 is not extremely large
compared to the other sources of uncertainty in the analysis. Provisionally, the value of KM
estimated from the complete Wittnoom cohort will be applied in our analysis. In revisions to this
7.32
-------
document, it will be important to attempt to evaluate the exposure-response for mesothelioma
using data from additional other cohorts.
Given the importance of these two issues: (1) the relative potencies of chrysotile and the
amphiboles and (2) the adequacy of U.S. EPA models for predicting the time and exposure
dependence of disease, limited analysis of raw data from a small number of additional cohorts is
warranted. Recommendations for further research along these lines are discussed in the
conclusions to this Chapter (Section 7.6) and parallel the recommendations of the expert panel
(Appendix B).
7.3.2 Estimating KM Values from Published Epidemiology Studies
At the time that the Health Effects Update was published (U.S. EPA 1986), four studies were
found to provide suitable quantitative data for estimating a value for KM and six additional
studies provided corroborative support for the mesothelioma model applied. Currently, there are
14 published studies with adequate data for deriving an estimate of KM (including updates to all
four of the quantitative studies evaluated in 1986).
The U.S. EPA mesothelioma model (Equations 7-6 and 7-7) was applied to each of these data
sets to obtain study-specific estimates for the mesothelioma dose-response coefficient, KM. The
resulting set of KM values are presented in Table 7-9. The format for this table is identical to that
described in Section 7.2.2 for Table 7-6. As with the KL values in Table 7-6, the KM values for
all studies presented in Table 7-9 (including those studies that have not been updated since their
inclusion in the 1986 Health Effects Update) were re-derived using the modified procedures
described in Appendix A. Uncertainty intervals in Table 7-9 for each estimated KM were derived
using the method described in Appendix A.
As Table 7-9 indicates, the KM values derived in this study and the corresponding values derived
in the original 1986 Health Effects Update are in close agreement, none vary by more than a
factor of 1.5. Among the KM values derived in the current study, the lowest and highest of the
best-estimate values differ by a factor of approximately 1,400 (excluding the one negative study
of Connecticut friction product manufacturers) and many of the pair-wise sets of uncertainty
intervals do not overlap. For example, none of the uncertainty intervals for the KM values
derived for any of the environments involving exposure to chrysotile overlap the uncertainty
intervals associated with the KM values derived for either crocidolite mining or asbestos-cement
manufacture using mixed fibers at the Ontario plant (Finkelstein 1984). Furthermore, neither of
the uncertainty intervals for the KM values derived for chrysotile mines in Quebec (Asbestos and
Thetford) overlap the intervals around KM values for any of the amphibole environments or any
of the mixed environments, except the Quebec factory that is associated with the Asbestos mine
(Liddell et al. 1997).
The KM values and the associated uncertainty bounds derived in the current study are plotted in
Figure 7-2. Each exposure environment is plotted along the X-axis of the figure and is labeled
with a 4-digit code that indicates fiber type (chrysotile, mixed, crocidolite, or tremolite), industry
(mining, friction products, asbestos-cement pipe, textiles, insulation manufacturing, or insulation
application); and a 2-digit numeric code indicating the study from which the data were derived.
7.33
-------
The key for Figure 7-1 also applies to this figure. In Figure 7-2, the chrysotile studies are
grouped on the left, amphibole studies are grouped on the right, and mixed studies are in the
middle. As in Figure 7-1, data from the Dement et al. (1994) study are used to represent the
South Carolina textile cohort and data from the Amandus and Wheeler (1987) are used to
represent the Libby mine cohort in Figure 7-2. Also, the estimated KM values for the Quebec
miners (Liddell raw data) from Asbestos and Thetford Mines, respectively, are averaged in this
figure.
Figure 7-2:
Plot of Estimated Ky Values and Associated Uncertainty Intervals by Study Environment
1000^
100.
10,
CO
O
0.01,
1E-3-
CN -sj- IO CD
2 U- Q- I-
o ooo
oo a>
D_ Q.
in CD
oo
7.34
-------
Figure 7-2 indicates that, within chrysotile studies alone, lowest and highest KM values vary by
approximately a factor of 15 (excluding the negative friction products study), which is minimal
variation relative to the spread across the values from all of the studies in the table. Also, the
corresponding uncertainty intervals have considerable overlap.
The KM values for the two "pure" amphibole studies (crocidolite mining and amosite
manufacturing) agree to within a factor of 2 and, with the exception of the value from the
Ontario asbestos-cement plant (MP8 in Figure 7-2 and the ninth study listed in Table 7-9) and
the factory plant associated with the Asbestos, Quebec mines (MF14 in Figure 7-2 and the 11th
study listed in Table 7-9), the mixed exposure KM values lie between the high values for the
"pure" amphibole exposures and the lower values reported for all of the chrysotile sites.
Although the KM value for the asbestos-cement plant (MP8) appears high, its corresponding
uncertainty interval overlaps those of the other "pure" amphibole studies as well as those of a
number of the other mixed exposures. The KM value reported for the factory associated with the
Asbestos, Quebec mine (MF14) is the lowest of those reported for mixed exposures, but its
uncertainty interval overlaps those for several of the other mixed exposures, as well as all of
those involving only chrysotile exposures.
As with KL values, the asbestos-cement pipe industry shows the greatest variation in KM values
across all asbestos types, with a range of more than a factor of 90. Moreover, the uncertainty
intervals for the largest and smallest values in this industry do not overlap. Within the textile
industry, the KM value for the South Carolina chrysotile plant is only an eighth of the values
reported for the two textile plants using mixed fibers, although there is considerable overlap in
their uncertainty intervals. The potential sources of variation across all of the KM values are
likely attributable to the same sources of variation identified for the KL values and previously
described in detail (Section 7.22).
7.4 EVALUATION OF ASBESTOS EXPOSURE INDICES
As indicated in Chapter 3, for exposure-response coefficients derived from one environment to
be applicable in a different environment requires that both of the following two conditions be
satisfied:
• that asbestos be measured in both environments in an identical manner; and
• that such measurements reflect (or at least remain proportional to) the
characteristics of asbestos exposure that determine biological activity.
When these two conditions are met, it is possible to define an "exposure index" that accurately
reflects biological activity, and consequently an exposure-response coefficient based upon such
an exposure index derived in one environment can confidently be applied in a different
environment. Such an index can be defined as a weighted sum of concentrations of categories of
structures of different asbestos types and sizes, where the weights reflect the relative
carcinogenic potencies of the different type and size categories. For example, as described in
Section 6.4.3, Herman et al. (1995) derived an optimal exposure index from an analysis of rat
inhalation studies involving exposures to different types of asbestos and fibrous structures of
differing dimensions. The optimum index (defined in Equation 6-11) consists of a weighted sum
7.35
-------
of the air concentrations of structures between 5 and 40 urn in length and >40 \im in length (all
thinner than 0.4 urn).
There is considerable evidence that the manner in which asbestos is quantified in the available
epidemiology studies (i.e., PCM) may not adequately reflect the characteristics that relate to
biological activity (Sections 6.4 and 6.5). Therefore, the second of the above two criteria may
not be satisfied when exposure-response coefficients (i.e., KL and KM values) derived from these
studies are used to predict risk in new environments. We therefore investigated the possibility of
adjusting the KL and KM values so as to apply to a measure of exposure that better reflects
biological activity. Such an adjustment requires both (1) data from each studied environment on
fiber size and asbestos type needed to adjust the corresponding KL and KM and (2) evidence
regarding what measure of exposure better reflects biological activity.
7.4.1 Fiber Type and Size Distribution Data Available for Deriving Exposure Indices
Since the range of possible adjustments to the KL and KM values is constrained by the data on
fiber size and type available from each studied environment, we first consider the characteristics
of the data currently available for making such adjustments. The data considered to be pertinent
consisted of TEM analyses of samples conducted in the same environment in which an
epidemiological study was conducted or from an environment involving a similar operation (e.g.,
mining, textile manufacture, etc.). Table 7-13 lists available fiber size distributions obtained
from a search of the literature and categorized by fiber type and type of operation. The
epidemiological studies from which KL or KM have been calculated are categorized accordingly.
Assuming that the available TEM size distributions are representative of dust characteristics for
the exposure settings (industries) studied, these were paired with corresponding epidemiology
studies. The TEM size distributions were then used to convert the exposure measurements used
in the epidemiology studies to the new exposure index that potentially better reflects biological
activity. Studies were paired as indicated in Table 7-13.
Only a subset of the TEM size distributions listed in Table 7-13 were actually employed in the
effort to normalize KL and KM values. To minimize variability resulting from differences in
TEM analysis methodology employed by different authors, it was decided to employ
distributions from common studies conducted by common groups of researchers, to the extent
that this could be accomplished without reducing the number of "size-distribution-
epidemiological study" pairs available for inclusion in the analysis. Also, studies containing the
best documented procedures were favored. Ultimately, with one exception the size distributions
selected for use came from only two studies, which were reported in three publications: Dement
and Harris (1979), Gibbs and Hwang (1980), and Hwang and Gibbs (1981). In the case of the
one exception, size distributions for Libby (tremolite asbestos in vermiculite) were derived from
unpublished TEM data recently acquired directly from the site.
7.36
-------
Table 7-13. Correlation Between Published Quantitative Epidemiology Studies and
Available Tern Fiber Size Distributions
Fiber Type Exposure Setting
Distribution Reference
Epidemiology Reference
Chrysotile Textiles
Chrysotile
and
Crocidolite
Crocidolite
Amosite
Friction Products
Mining and Milling
Asbestos Cement
Manufacturing
Asbestos Cement
Manufacturing
Dement and Harris 1979
Cherrie et al. 1979
Dement and Harris 1979
Marconi et al. 1984
Winer and Cossett 1979
Roberts and Zumwalde 1982
Rood and Scott 1989
Gibbs and Hwang 1980
Winer and Cossett 1979
Dement and Harris 1979
Snyderetal. 1987
Winer and Cossett 1979
Hwang and Gibbs 1981
Mining and Milling Gibbs and Hwang 1980
Hwang and Gibbs 1981
Insulation Manufacturing Dement and Harris 1979
Insulation Application
Insulation Clearance
Tremolite Vermiculite Mining
Snyderetal. 1987
Cherrie et al. 1979
U.S. EPA, unpublished
Dement etal. 1994, 1983b
McDonald et al. 1983a,b
Peto 1980a; Peto et al. 1985
Berry and Newhouse 1983
McDonald et al. 1984
Liddelletal. 1997
McDonald et al. 1980b
Nicholson et al. 1979
Piolatto et al. 1990
Hughs etal. 1987
Finkelstein 1984
Finkelstein 1983
Hughs etal. 1987
Weilletal. 1979
Weilletal. 1994
Albinetal. 1990
Armstrong et al. 1988
de Klerk, unpublished data
Levin etal. 1998
Seidmanetal. 1986
Seidman 1984
Selikoffetal. 1979
McDonald et al. 1986
Amandus and Wheeler 1987
7.37
-------
Table 7-14 presents the resulting bivariate fiber size distributions derived from the published
TEM data that are paired with representative KL and KM values from the corresponding
epidemiology studies. This table shows 17 KL values and 11 KM values matched with a fiber size
distribution from the literature. In this table some length and width categories from some
published distributions have been combined, so that, for the most part, only those categories
available for all of the epidemiological studies are presented. The column labeled "PCME"
("PCM-equivalent") provides the relative concentrations of asbestos fibers that would have been
identified by PCM (^5 |im in length and 2:0.2 um in width).
The fiber size distributions in Table 7-14 are not all of equal relevance to the respective
epidemiological studies to which they were paired. As indicated in Table 7-15, some of the
distributions are based on data collected at the same facility, others are based on data collected at
a similar facility, still others are based on a combination of data from similar facilities, etc. The
uncertainty factors listed in Table 7-15 were developed to quantify the relevance of each fiber
size distribution to its paired epidemiological study, where larger factors indicate a less certain
relevance. How these factors were used is explained below. It should also be kept in mind that,
whereas these fiber distributions were based on air samples collected over a fairly narrow time
range, they are used to represent the fiber size distributions throughout the exposure period,
which in most of the epidemiological studies covers many years.
As indicated in Tables 7-14 and 7-15, for the two environments for which multiple studies were
available (i.e., the South Carolina textile plant and the Libby, Montana vermiculite mine), a
single study was selected to represent each environment. For the South Carolina textile plant the
Dement et al. (1994) study was selected and for Libby, the Amandus and Wheeler (1987) study
was selected.
7.4.2 Modification of Existing KL and KM to Conform to a New Exposure Index
The fiber size distribution data in Table 7-14 are used to transform the existing KL and KM values
(which are defined in terms of PCM measurements) so they conform to a different exposure
index based on TEM. To see how this is accomplished, consider a KL value pertaining to a
specific environment, let CPCM be an air concentration from that environment measured by PCM,
let Cnew be the concentration in the same air measured by a new exposure index using TEM (e.g.,
perhaps defined as a weighted sum of TEM concentrations in various length, width and asbestos
type categories), and let KL* be the adjusted exposure-response coefficient corresponding to the
new exposure index. It is clear that for the U.S. EPA lung cancer model (Equation 7-1 or 7-2) to
estimate the same risk from the given air concentration using either exposure index, it is
necessary that
(KL)(CPCM) = (KL*)(CNEW) (Eq.7-9)
Using CpcME (the air concentration of PCM-equivalent fibers - fibers measured by TEM that
would be identified by PCM) as a replacement for CPCM, we get
KL*=KL(CPCME/CNEW) (Eq.7-10)
7.38
-------
Table 7-14. Representative KL and KM Values Paired with Averaged TEM Fiber Size Distributions From Published Papers
Environment
Quebec mines and
mills
Quebec mines
Italian mine and mill
Connecticut plant
New Orleans plants
South Carolina plant
Wittenoom,
Australia
Patterson, NJ factory
Tyler, Texas factory
Libby, Montana
Study
Code
CM1
CM2
CM3
CF4
CP5
CT6
RM18
AI19
AI20
TM21
(xlOO) (x!08)
0.029
0.0165
0.051
0 0
0.25 0.2
1.6 0.17
0.47 7.9
1.1 3.9
0.13
0.45
PCME
00.14
0.014
0.014
0.07235
0.05071
0.12963
0.01167
0.35198
0.35198
4
w<0.2
L<5
0.93545
0.93545
0.93545
0.76359
0.77469
0.65629
0.89017
0.17109
0.17109
50.3 L w>0.4 L
5
-------
Table 7-15. Estimated Uncertainty Assigned to Adjustment for Fiber Size
Study Location
Quebec mines and
mills
Quebec mines
Italian mine and
mill
Connecticut plant
New Orleans plants
South Carolina
plant
British factory
Ontario factory
New Orleans plants
Swedish plant
Belgium factory
Estimated
Study Uncertainty
Code Factor Explanation KL Reference KM Reference
CM1
CM2
CMS
CF4
CP5
CT6
MF7
MP8
MP9
MP10
MP11
1 Location common to epidemiology Liddell et al. 1997
study and size study
1 Location common to epidemiology Liddell et al. 1997 (raw data
study and size study Loc. 1,3,4)
1.75 Same industry, separate locations Piolatto et al. 1990
for epidemiology and size studies
1 .25 Epidemiology location one of McDonald et al. 1984 McDonald et al. 1984
several combined for size study
1 .25 Epidemiology location one of Hughes et al. 1 987 Hughes et al. 1 987
several combined for size study
1 .25 Epidemiology location one of Dement et al. 1994 (raw data) Dement 2001 (personal
several combined for size study communication)
1.5 Same industry, separate locations Berry and Newhouse 1983
for epidemiology and size studies
1.5 Epidemiology location probably Finkelstein 1984 Finkelstein 1984
one of several combined for size
study
2 Same industry, separate locations, Hughes et al. 1987 Hughes etal. 1987
mixed exposures
2 Same industry, separate locations, Albin etal. 1990
mixed exposures
2 Same industry, separate locations, Laquet et al. 1980
mixed exposures
7.40
-------
Table 7-15. Estimated Uncertainty Assigned to Adjustment for Fiber Size (continued)
Study Location
U.S. retirees
Asbestos, Quebec
U.S. insulation
workers
Pennsylvania plant
Rochedale,
England plant
Whitenoom,
Australia
Patterson, NJ
factory
Tyler, Texas
factory
Libby, Montana
Estimated
Study Uncertainty
Code Factor Explanation
MX13
MX14
MI15
MT16
MT17
RM18
AI19
AI20
TM21
2
2
2
1.75
1.25
1.25
1.75
Generally similar industries
studied for epidemiology and size
Same industry, separate locations,
mixed exposures
Same industry, separate locations,
mixed exposures
Same industry, separate locations
for epidemiology and size studies
Epidemiology location one of
several combined for size study
Epidemiology location one of
several combined for size study
Extrapolated from limited,
marginally associated air data
KL Reference
Enterline et al. 1986
Selikoff and Seidman 1991
McDonald et al. 1983b
Petoetal. 1985
DeKlerk, unpublished data
Seidman et al. 1986
Levin et al. 1998
Amandus and Wheeler 1987
KM Reference
Liddell et al. 1997 (raw data)
Selikoff and Seidman 1991
McDonald et al. 1983b
Petoetal. 1985
DeKlerk, unpublished data
Seidman etal. 1986
7.41
-------
In the actual calculation of KL*, this equation is applied with the ratio of air concentrations
appearing on the right side of this expression replaced by the equivalent ratio of fiber proportions
from Table 7-14.
As an example of the calculation of a KL*, an earlier draft of this report proposed use of an
exposure index defined as the weighted sum,
0.997 CL>10;W<0.4 + 0.003 C5
-------
In view of the limitations of the fiber distribution data, it was decided, as an interim measure, to
adopt a maximum fiber width of 0.4 (am in the proposed exposure index for both lung cancer and
mesothelioma. This is the width indicated by the animal data (Herman et al. 1995). Subject to
that decision, we then derived separate exposure index indices for lung cancer and
mesothelioma, respectively, that are each optimal (as defined below) with respect to fiber type
(chrysotile and amphibole) and fiber length (5-10 \im and >10 ^m). Based on results of Herman
et al. (1995) and lack of compelling evidence elsewhere in the literature (assuming that size
effects are adequately addressed, see Chapter 6), it was assumed that all similarly-sized
amphibole fibers are equipotent.
To develop separate potency estimates for chrysotile and amphibole fibers (adjusted for fiber
size) it was necessary to estimate the relative amounts of chrysotile and amphibole in each
environment. Table 7-16 presents our estimates of the fraction of exposure in each study
environment contributed by amphibole asbestos, based on information on each environment
available in the literature. The source of the information used to develop each estimate as well
as a brief description of how the estimate was developed is also provided.
7.4.3.1 Optimizing the Exposure Index for Lung Cancer
A statistical model was fit to the KL values in Table 7-14 and results from the fitting were used to
estimate separate potencies for amphibole and chrysotile, to estimate relative potencies of fibers
of different sizes, and to test certain hypotheses. In this model Ln(KL) (the log transform of a KL
value in Table 7-14) was assumed to be normally distributed with mean equal to
Ln{KLa* . [f^ + rpc . (1 - f^)]. [q . C5.IO + (1 - q) . C>10] / CPCME} (Eq. 7-12)
In this expression C5.,0, and C>10 are the fractions of fibers among those thinner than 0.4 urn that
are between 5 and 10 nm in length or longer than 10 u.m, respectively (from Table 7-14), CpcME
is the fraction of PCME fibers (also from Table 7-14), and f^,, is the fraction of amphiboles
estimated for each environment (from Table 7-16). In addition there are four parameters that are
estimated by fitting the model to the KL values:
q - the relative potency of fibers thinner than 0.4 \im and between 5 and 10 urn in length,
relative to fibers thinner than 0.4 urn and >10 urn in length;
KLa* - the potency (KL value) of pure amphibole (based upon the exposure index defined by q);
rpc -the relative potency of chrysotile, relative to amphibole
The fourth parameter, o, is described below.
Note that the part of Equation 7-12 that is inside the curly brackets is like Equation 7-9 solved
for KL but with an additional term to account for the relative amounts of chrysotile and
amphibole.
7.43
-------
Table 7-16. Estimated Fraction of Amphiboles in Asbestos Dusts
Fraction of Amphiboles
Study Location
Quebec mines and
mills
Quebec mines
Italian mine and
mill
Connecticut plant
New Orleans
plants
South Carolina
plant
British factory
Ontario factory
New Orleans
plants
Study Best
Code Estimate (%)
CM1 1
CM2 1
CM3 0.3
CF4 0.5
CP5 1
CT6 0.5
MF7 0.5
MP8 30
MP9 5
Estimated
Range (%)
0-4
0-4
0.1-0.5
0-2
0-2
0-2
0-2
10-50
2-15
Source of Estimate
. Sebastien et al. 1986,
extrapolated from air data
Sebastien et al. 1986,
extrapolated from air data
McDonald et al. 1984,
extrapolated from plant history
Hughs etal. 1987,
extrapolated from plant
history.
Sebastien et al. 1989 (based
on), extrapolated from Quebec
source material
Berry andNewhouse 1983,
extrapolated from plant history
Finkelstein 1984, extrapolated
from plant history
Hughes etal. 1987,
extrapolated from plant history
KL Reference
Liddelletal. 1997
Piolatto et al. 1990
McDonald et al. 1984
Hughes etal. 1987
Dement etal. 1994
(raw data)
Berry and Newhouse
1983
Finkelstein 1984
Hughes etal. 1987
KM reference
Liddelletal. 1997 (aw
dataLoc. 1,3,4)
McDonald et al. 1984
Hughes etal. 1987
Dement 2001 (personal
communication)
Finkelstein 1984
Hughes etal. 1987
Swedish plant MP10
0-6 Albin et al. 1990, extrapolated Albin et al. 1990
from plant history
7.44
-------
Table 7-16. Estimated Fraction of Amphiboles in Asbestos Dusts (continued)
Fraction of Amphiboles
Study Location
Study Best
Code Estimate (%
Estimated
Range (%)
Source of Estimate
KL Reference
KM reference
Belgium factory MF11
U.S. retirees MX13
Asbestos, Quebec MX 14
Laquetetal. 1980
Enterline et al. 1986
Liddelletal. 1997 (raw
data)
U.S. insulation
workers
Pennsylvania plant
Rochedale,
England plant
Whitenoom,
Australia
Patterson, NJ
factory
Tyler, Texas
factory
Libby, Montana
MI15
MT16
MT17
RM18
AI19
AI20
TM21
50
8
5
97
97
97
95
25-75
3-15
2.5-15
95-100
95-100
95-100
90-100
Guess estimate for broad
industry
McDonald et al. 1983b,
extrapolated from plant history
Peto et al. 1985, extrapolated
from plant history
General estimate"
General estimate"
General estimate"
General estimate3
Selikoff and Seidman
1991
McDonald et al. 1983b
Peto etal. 1985
DeKlerk, unpublished
data
Seidman etal. 1986
Levin etal. 1998
Amandus and Wheeler
1987
Selikoff and Seidman
1991
McDonald et al. 1983b
Peto et al
DeKlerk,
data
Seidman
. 1985
unpublished
et al. 1986
"Allows for the possibility of some foreign material. Practical effect for the analysis is nil. Might just as well assume 100%.
7.45
-------
With this formalism, the potency, KLc*, of pure chrysotile is defined by the product, rpc. KLa*.
Thus rpc=l corresponds to equal potency of amphibole and chrysotile and rpc=0 corresponds to
chrysotile being non-potent for causing lung cancer. Similarly, q=l corresponds to fibers
between 5 and 10 um in length having the same potency as fibers longer than 10 urn and q=0
corresponds to such fibers being non-potent for causing lung cancer.
The variance of Ln(KL) was assumed to be composed of two components. The first component,
O;, was calculated so as to reflect the uncertainty in the KL values as reflected by both the
uncertainty intervals reported in Table 7-6 and the uncertainty in the relevance of the size
distributions applied to each environment, as reported in Table 7-15. Specifically, the upper
bound of the uncertainty interval for KL in Table 7-6 was modified by multiplying it by the
uncertainty factor in Table 7-15. This modified upper bound was then divided by the best
estimate of KL from Table 7-6, and then divided by 2.0. The log transform of the result was
defined as Oj. A second component, a, of the standard deviation, assumed to be constant for all
studies, was also estimated. This component may be thought of as representing the uncertainty
in the KL estimate resulting from random variation that is not represented in the QJ. The overall
standard deviation of the Ln(KL) from the study was assumed to be (oi2+o2)'/1.
Using the model described by Equation 7-12, parameters (q, KLa*, rpc, and o) were estimated by
maximum likelihood and likelihood tests were used to test the hypotheses that chrysotile was
non-potent (rpc=0) or equally potent (rpc=l) with amphibole (Wilks 1963). Results from the
analysis are summarized in Table 7-17. Shaded values in the table indicate parameter values that
were fixed rather than estimated for each particular model run. By holding certain parameters
fixed, we evaluated the fit of a range of exposure indices, defined as indicated. This table also
contains results from a similar analysis of mesothelioma dose-response coefficients (KM), which
are discussed in Section 7.4.3.2.
The results of fitting the model defined by Equation 7-12 to the KL values in Table 7-14 are
shown in the columns of Table 7-17 labeled "Equation 7-12". The first column, labeled
"optimized values" contains the resulting parameter estimates and log-likelihood with all four
parameters estimated. Note that the estimate of q is q=0. Since q represents the potency of
fibers between 5 and 10 |im in length, relative to fibers >10 urn in length (considering only fibers
thinner than 0.4 p,m), the model predicts that fibers between 5 and 10 \im are non-potent in
causing lung cancer.
The estimate of rpc for the optimized model run of Equation 7-12 is rpc=0.266, which predicts
that chrysotile is about 25% as potent as amphibole in causing lung cancer (after adjusting for
fiber size). The fourth and fifth columns of Table 7-17 contain results of fitting the model with
rpc fixed at either rpc=0 or rpc=l (both with q fixed at q=0). These results are used to conduct
likelihood ratio tests of the hypotheses that rpc=0 and rpc=l. The resulting test for rpc=0 is
highly significant (p=0.007). Thus, with this formulation the hypothesis that chrysotile is non-
potent in causing lung cancer can be rejected. The test for rpc=l is non-significant
(p=0.54),indicating that even though the best estimate is that chrysotile is only one-fourth as
potent in causing lung cancer as amphibole, the hypothesis that chrysotile and amphibole are
equally potent cannot be rejected.
7.46
-------
Table 7-17. Results from Fitting Exposure Indices Defined by Equation 7.12 and Feme to
Lung Cancer and Mesothelioma Exposure-response Coefficients Estimated from
Different Environments
Variable
Index
Defined
by
Equation
7.11
Equation 7.12
Optimize
d
Values
RPC=1
RPC=0
Optimize
d
Values
PCME
RPC=1
RPC=0
Lung Cancer (N=16)
RFC
100*(KLA)
s
Log-
likelihood
q
Hypothesi
s tests
100*(KLC)
0.267
2.3
1.007
-17.1022
0.003
0.61
0.266
2.34
1.004
-17.0833
0
0.61
1
0.953
1.092
-17.2659
0
H0:
RPC=1
p=0.20
0.953
0
15.80
1.730
-20.6989
0
H0:
RPC=0
p=0.001
0
0.469
0.48
1.050
-17.3451
NA
0.23
1
0.29
1.070
-17.6271
NA
HO:
RPC=1
p=0.54
0.29
0
4.42
1.915
-21.8543
NA
H0:
RPC=0
p=0.007
0
Mesothelioma (N=ll)
RFC
108*(KMA)
s
Log-
likelihood
q
Hypothesi
s tests
108*KMr)
0.0013
26.78
0.6062
-9.33248
0.003
0.035
0.0013
26.99
0.6038
-9.31812
0
0.035
1
2.54
1.903
-16.7403
0
H0:
RPC=1
p=0.0007
2.54
0
28.8
0.6099
-9.35267
0
H0:
RPC=0
p=0.61
0
0.0033
7.69
0.7605
-10.4599
NA
0.025
1
0.737
1.805
-16.16
NA
H0:
RPC=1
p=0.0001
0.74
0
8.87
0.7782
-10.5931
NA
H0:
RPC=0
p=0.79
0
Notes: Shaded areas indicate values that were fixed in advance of the analyses.
"NA" means not applicable.
Also shown in Table 7-17 are the results of applying the exposure index proposed in the earlier
draft of this report (Equation 7-11). This index assigns a small relative potency of 0.003 to
fibers between 5 and 10 ^m in length, compared to the fully optimized model, which, as noted
above, assigns zero potency to these fibers. Table 7-17 indicates that both the quality of fit (by
comparison of likelihoods) and the resulting parameter estimates are virtually identical.
Accordingly, unless the ratio of fibers between 5 and 10 \im in length to those longer than 10 \im
(among those thinner than 0.4 jam) in an environment is extremely large (e.g., >300-fold), the
two indices will provide practically equivalent results.
7.47
-------
As previously indicated, the current approach for estimating asbestos-related risk (U.S. EPA
1986) uses (effectively) PCME as the exposure index. To allow comparison with this approach,
analyses were also conducted using PCME as the exposure index, rather than the size range of
fibers considered heretofore (i.e., Equation 7-12 was simplified to Ln{Ku* , [f^,, + rpc . (1 -
famph)])- Results of this analysis are shown on the right side of Table 7-17. With this exposure
index, the best estimate is that chrysotile is about one-half as potent as amphibole, the hypothesis
that chrysotile is non-potent can be rejected (p=0.001), and the hypothesis that chrysotile and
amphibole are equally potent cannot be rejected (p=0.20).
Based upon a comparison of either the residual variance, o, or the likelihood, it appears that the
exposure index defined by fibers longer than 10 u.m and thinner than 0.4 |im (corresponding to
q=0 in Table 7-17) provides at most a very marginal improvement in fit over use of PCME as the
exposure index for lung cancer when rpc is held at 1. Similarly, adjusting for fiber type but not
size (i.e., optimizing rpc to reflect separate potencies for chrysotile and the amphiboles) also
provides at best a marginal improvement over the current approach (PCME with rpc=l).
However, when the exposure index is adjusted for both fiber size and type, (comparing the
optimized values for Equation 7-12 to PCME with rpc=l), a small improvement is apparent. The
log-likelihood increases by half a unit and the spread in the estimated KL values (represented by
a) decreases by about 7%. Moreover, as discussed in the following section, the improvement for
mesothelioma is substantial..
In addition to the analyses reported in Table 7-17, other analyses were conducted in which an
additional term was added to the linear combination of fiber lengths appearing in Equation 7-12
to represent fibers shorter than 5 \im (still considering only fibers thinner than 0.4 urn). In this
analysis, the best estimate of both the potencies of fibers shorter than 5 ^m, and between 5 and
10 urn in length, was zero.
Discussion of the results from the analysis of the mesothelioma values (also presented in Table
7-17) is provided in Section 7.4.3.2. Based on this analysis and the rest of the evaluation
described in this chapter, a recommended set of lung cancer and mesothelioma exposure-
response coefficients is presented in Section 7.5.
7.4.3.2 Optimizing the Exposure Index for Mesothelioma
Concomitant with the analysis reported in Section 7.4.3.1 regarding development and evaluation
of an improved exposure index for lung cancer exposure-response coefficients (KL values) that
correlates better with biological activity, a parallel analysis was performed for the mesothelioma
exposure-response coefficients (KM values) presented in Table 7-14. This table presents data
from 11 environments in which KM values are paired with fiber size distribution data. The
corresponding uncertainty intervals for these 11 KM values are provided in Table 7-9. The same
relationship defined by Equation 7-9 was used to adjust these KM values to a different exposure
index and the same statistical model (Equation 7-12) was applied both to evaluate different
adjustments and to develop an adjustment that was optimal for the available data. The results of
this analysis of KM values are presented in the bottom half of Table 7-17, which also contains the
results of the comparable analysis of KL values.
7.48
-------
The results of fitting the model defined by Equation 7-12 to the KM values in Table 7-14 are
shown in the columns of Table 7-17 labeled "Equation 7-12". The first of these columns, labeled
"optimized values" contains the resulting parameter estimates and log-likelihood with all four
parameters estimated. Just as was the case with the analysis of KL values, the best estimate of q
is q=0. Since q represents the potency of fibers between 5 and 10 [im in length, relative to fibers
>10 \i.m in length (considering only fibers thinner than 0.4 um), just as was the case for lung
cancer, the model predicts that fibers between 5 and 10 [im are non-potent in causing
mesothelioma.
For mesothelioma, the best estimate of rpc is rpc=0.0013, which predicts that chrysotile is only
0.13% as potent as amphibole in causing mesothelioma (after adjusting for fiber size). This
small estimate for rpc is not significantly different from rpc=0 (p=0.79). Consequently, in this
analysis, the data are consistent with the hypothesis that all of the mesotheliomas occurring in
cohorts exposed primarily to chrysotile are due to small amounts of amphibole contamination
within the chrysotile. Moreover, the hypothesis that chrysotile and amphibole are equally potent
in causing mesothelioma (the assumption inherent in the U.S. EPA (1986) asbestos health effect
document) is clearly rejected (p=0.0001).
Results using PCME as the exposure index for mesothelioma (the bottom right side of Table
7-17) are similar. The best estimate is that chrysotile is only 0.0033 as potent as amphibole, the
hypothesis that chrysotile is non-potent cannot be rejected (p=0.61), and the hypothesis that
chrysotile and amphibole are equally potent is definitely rejected (p=0.0007).
Based upon a comparison of either the residual variance, o, or the likelihood, it appears that the
exposure index defined by fibers longer than 10 um and thinner than 0.4 um (corresponding to
q=0 in Table 7-17) provides an improvement in fit over use of PCME as the exposure index for
mesothelioma. Even after adjusting for the effects of fiber type (i.e., comparing the o values
estimated for the optimized values with PCME and Equation 7-12 as the exposure index,
respectively), the variation across KM values appears to decrease by more than 20% when values
are adjusted to the index of fibers that are longer than 10 |j,m and thinner than 0.4 um.
Moreover, comparing o values between the index in current use (PCME with rpc=l) (U.S. EPA
1986) and the optimized index of longer and thinner fibers (with rpc=0.0013), use of the latter
exposure index results in a 67% reduction in variation across KM values.
As was also the case for the lung cancer analysis, the fit based on the exposure index defined by
Equation 7-11, which was proposed in an earlier version of this report (with q, the relative
potency of fibers between 5 and 10 ^m compared to fibers longer than 10 um fixed at q=0.003),
is virtually identical to the fully optimized fit (which predicts q=0).
In addition to the analyses reported in Table 7-17, another analysis for mesothelioma were
conducted in which an additional term was added to the linear combination of fiber lengths
appearing in Equation 7-12 to represent fibers shorter than 5 |im (still considering only fibers
thinner than 0.4 um). In this analysis, the best estimate of the potency of fibers shorter than 5
m was also zero.
7.49
-------
Based on results in this section and the corresponding evaluation of lung cancer (Section 6.4.3.1)
a recommended set of lung cancer (and mesothelioma) exposure-response coefficients are
developed and presented in Section 7.5.
7.5 THE OPTIMAL EXPOSURE INDEX
7.5.1 Definition of the Optimal Index and the Corresponding Exposure-Response Factors
Table 7-17 presents the results of fitting a statistical model to asbestos exposure-response
coefficients to estimate the relative potencies for fibers of various sizes and types that define an
optimized exposure index for asbestos. Although the coefficients for lung cancer (the KL values)
and mesothelioma (the KM values) were separately evaluated, the optimal index for each
incorporates the same size range of fibers (at least based on the limited range of options
evaluated). These are fibers longer than 10 urn and thinner than 0.4 urn
Results in Table 7-17 also differentiate between the potency of chrysotile and amphibole for both
lung cancer and mesothelioma. Amphibole is estimated as being about four times as potent as
chrysotile for lung cancer (although the difference is not significant) and about 800 times as
potent as chrysotile for mesothelioma (a highly significant difference). Moreover, the data are
consistent with the hypothesis that chrysotile has zero potency toward the induction of
mesothelioma.
The optimized dose-response coefficients (rounded up) from Table 7-17 for pure fiber types
(chrysotile or amphibole) are summarized in Table 7-18. These coefficients apply to exposures
quantified in terms of concentrations (in f/ml) of fibers longer than 10 (im and thinner than
0.4 um.
Table 7-18. Optimized Dose-Response
Coefficients for Pure Fiber Types
Fiber Type
Chrysotile
Amphiboles
KL. x 100
0.6
3
KM.xl08
0.04
30
7.5.2 Evaluation of the Optimal Exposure Index
The optimal coefficients presented in Table 7-18, result from adjustments for both fiber type and
fiber size to the KL and KM values obtained directly from the literature (Tables 7-6 and 7-9),
which are linked to PCM measurements. To get some idea of the relative effects of adjustments
for fiber size (from PCM to fibers longer than 10 um and thinner than 0.4 u.m) and fiber type
toward the goal of rationalizing the KL and KM values obtained from different environments, the
effect of the two adjustments are considered sequentially—first the effect of adjusting for fiber
size is considered and next the added effect of adjusting for fiber type is evaluated.
7.50
-------
To compare the effects of adjusting KL and KM values for fiber size, the KL. and KM. values,
which are adjusted to the optimal exposure index (using Equation 7-9) are plotted (along with
their associated uncertainty intervals) in Figure 7-3, and 7-4. These figures are in a format
identical to that of Figures 7-1 and 7-2 of the untransformed values. The "key" provided for
Figure 7-1 is also directly transferable to Figures 7-2, 7-3, and 7-4.
Note that one of the points plotted in Figures 7-1 and 7-2 was omitted from Figures 7-3 and 7-4
(for the Enterline study [1986] of retired factory workers: study MX13) because it was felt that
none of the available size distributions were suitably applicable for this study site, so no
conversions were possible. Also, the confidence intervals are larger in Figures 7-3 and 7-4
because, as previously indicated, we have attempted throughout our analysis of the epidemiology
data to account for major sources of uncertainty. Thus, the confidence intervals depicted in
Figure 7-3 and 7-4 are modified from those depicted in Figures 7-1 and 7-2 to account for the
uncertainty of making the adjustment for fiber size using paired data from published size
distributions. The intervals were adjusted by multiplying the upper bound and dividing the
lower bound by a factor thought to represent the relative contribution to uncertainty contributed
by the need to match data from separate studies to perform the conversions. The factors
employed (along with the rationale for selecting the value of each factor) are provided in
Table 7-15.
The visual impressions from a comparison between Figures 7-1 and 7-3 for lung cancer and
between Figures 7-2 and 7-4 for mesothelioma are that the changes resulting from adjusting for
fiber size are subtle. This visual impression is reinforced by the relative similarity of the fits
(based on comparison of o and log-likelihoods) reported in Table 7-17 for PCME and the
optimal exposure index, particularly for lung cancer.
Figures 7-5 and 7-6 show the effects of adjusting exposure-response coefficients for both fiber
size and type. To develop Figure 7-5 for lung cancer, a size-adjusted KL, as plotted in Figure
7-3, was further adjusted to correspond to pure amphibole by dividing it by the factor [f^,, + rpc
. (1 - f^ph)], where f^ is the proportion of amphibole fibers estimated for a given environment
(as listed in Table 7-16) and rpc=0.267, the optimal value from Table 7-17. The corresponding
adjusted factors corresponding to pure chrysotile can be calculated simply by multiplying the
amphibole value by rpc=0.267. Since the study-specific values for pure amphibole all differ
from the corresponding value for pure chrysotile by a multiplicative constant, values for both
types of asbestos are plotted simultaneously in Figure 7-5 by using a different scale for
amphibole and chrysotile. The same method was used to develop Figure 7-6 for mesothelioma
exposure-response coefficients, from the size-adjusted KM values plotted in Figure 7-4, the only
difference being that the mesothelioma rpc=0.0013 (Table 7-17) was used for this latter
conversion. Comparing Figure 7-5 to 7-1 suggests that the adjustments for size and type resulted
in as somewhat better reconciliation of the dose-response coefficients for lung cancer, although
the improvement is still somewhat subtle. However, a comparison of Figures 7-6 to 7-2
indicates a much more dramatic improvement in the case of mesothelioma. Comparisons of
Figures 7-2, 7-4, and 7-6 indicate this is primarily due to the adjustment for fiber type, although
the subsequent adjustment for fiber size provides further, noticeable improvement.
7.51
-------
Figure 7-3:
Plot of Estimated (Adjusted) KL. Values and Associated Uncertainty Intervals by Study Environment
1000-a
100^
10 -.
^ 1 ~*
T3
3
Tn
.2, 0.1 -
0.01
1E-3
o -1- o
in (D
Q. I-
O O
h- oo cn o ^ co
u. Q. Q. •«- *~ •«-
in
-------
Figure 7-4:
Rot of Estimated (Adjusted) K^ Values and Associated Uncertainly Intervals by Study Environment
1000-a
100-
00
O
0)
3
10-
1-
0.1-
0.01-
1E-3
^ li- Q.I-
O 000
oo co
0. 0-
-------
Rgue7-5:
Rot of Estimated K^ard K^ Values axl /^seriated Uhoertainty Intervals by Study ErMnonmBrt
1000 1
100:
CN
o :
X 10-
0>
o
.c
B" ^
C :
>5-
*
^f5 0.1,
B
U
'•6' 0.01-
1FV^-
(
t
1
\
I
1
(
(
\
|
1
1
i
i
(
(
^
(
i
i
(
i
i
i
\
i
i
i
i
.
-
i
_
i
.
-.
-
100
10 S
^
5T
1 *r^
o
^,
r-
0
Q1 ^
•X
0
^
Q01 ^
(/)
.2,
^^
1E3
o
into h»-coo)OT-co IOCDI^- oooo
CM
7.54
-------
Rgire7-6:
Rot of Estimated K,^ and ^ Vaues and /Associated UhcErtairtylrteivalsbySlLriyEn^iTXTrEnt
10000-
1000-
100-
10-
CL
E
t/>
0.01-
CM ^ IO CD 00 O
5 U- Q. I- Q- Q.
o ooo ss
co i^ oo O)
ii P
10
00
o
Q1
Q01 w
^
o
1E-4 OJ
^
7.55
-------
In addition to the graphical comparisons, it is also instructive to consider numerical comparisons.
Table 7-19 reproduces values for all of the original (study-specific) KL and KM (from Tables 7-6
and 7-9, respectively) along with all the corresponding values for the adjusted KL, and KM,.
Study-specific estimates for the all the corresponding KLA, KLC, KMA, and KMC are also presented.
Table 7-20 presents the magnitude of the spread in the range of original and adjusted KL and KM
values estimated as the quotient of the maximum and minimum values of each range. Note that,
of necessity, such an analysis requires that the zero values obtained for the Connecticut friction
products plant (CF4) be omitted. Note further that the data sets evaluated in Table 7-20 for the
original and adjusted values are identical (i.e., the one study for which no suitable fiber size
distribution could be found was excluded).
In Table 7-20, for mesothelioma, the spread in unadjusted values among "pure" fiber types (i.e.,
chrysotile only or amphibole only) are both substantially smaller than those for mixed data sets
(containing both fiber types). This provides further evidence of differences in the potencies of
each fiber type toward mesothelioma. It is also apparent that adjusting for fiber size decreases
the range within pure fiber types. Moreover, by simultaneously adjusting for both fiber size and
type (as illustrated by the column labeled, "K,^"), the range of variability across the entire data
set of 10 mesothelioma studies is reduced from almost 1,100 to a factor of 30, which is a clear
and substantial improvement.
For lung cancer, the results presented in Table 7-20 are a bit more complicated. While there is a
substantial reduction in the spread of unadjusted values among "pure" amphibole environments
when mixed environments are excluded, the spread among "pure" chrysotile environments is no
different than that observed for the entire data set. This is because, as previously described
(Section 7.2.2) the difference between the KL values observed among chrysotile miners in
Quebec and chrysotile textile workers in South Carolina represent the extremes of the entire
range of reported KL values. Adjusting these exposure-response coefficients for size provides
some improvement in agreement across these two environments. This reinforces the finding
from Appendix D that, if size adjustments incorporating cutoffs for longer fibers could be
incorporated into a new exposure index for asbestos, the apparent discrepancy between the
exposure-response observed among Quebec miners and South Carolina textile workers (and,
thus, among KL values as a whole) is likely to be further reconciled. The impressions from the
figures discussed above and from Table 7-20 confirm the conclusions concerning the
quantitative improvement in agreement across exposure-response coefficients highlighted in
Sections 7.4.3.1 and 7.4.3.2 and reinforce the conclusion that, especially with regard to
mesothelioma, adjusting exposure-response coefficients for fiber size and type leads to
substantially improved agreement across studies.
7.56
-------
Table 7-19. Study Specific K,, Km, K,., Km., Kla, Kma, Klc, and Kmc Values
Amphiboles
Chrysotile
Study
KL'
K,
K
MA
K
LC
K,
•MC
Environment Code KL Reference KM Reference (xlOO) (xlO8) (xlOO) (xlO8) (xlOO) (xlOO) (xlO8) (xlO8)
Quebec mines CM1 Liddell et al.
and mills 1997
0.029
Quebec mines
Italian mine
and mill
Connecticut
plant
New Orleans
plants
South
Carolina plant
Wittenoom,
Australia
Patterson, NJ
factory
Tyler, Texas
factory
Libby,
Montana
British factory
Ontario
factory
CM2
CMS
CF4
CP5
CT6
RM18
AI19
AI20
TM21
MF7
MP8
Piolatto et al.
1990
McDonald et
al. 1984
Hughes et al.
1987
Dement et al.
1994 (raw data)
DeKlerk,
unpublished
data
Seidman et al.
1986
Levin et al.
1998
Amandus and
Wheeler 1987
Berry and
Newhouse
1983
Finkelstein
1984
Liddell et al.
1997 (raw data)
McDonald et al.
1984
Hughes et al.
1987
Dement 2001
(person, comm.)
DeKlerk,
unpublished
data
Seidman et al.
1986
Finkelstein 1984
0.051
0
0.25
2.1
0.47
1.1
0.13
0.45
0.058
0.29
0.0165
0.108
0.131
0.062
0.399
0.716
5.51
0.106
0.190
0.00716
0.2 0.864 0.432 1.979 38.3
0.526 0.0498
0.25 6.38 0.660 20.7
106
5.50
7.9 0.87
14.43
0.9
15.08 0.24
0.137
0.02
3.9 17.36074 26.04111 7.526804 26.9044 2.00213 0.034976
0.868037
2.44
0.134276
0.889531
1.868577
0.498301
0.236615
0.497041
0.132548
18 3.246555 48.69833 1.63756 164.2291 0.435591 0.213498
7.57
-------
Table 7-19. Study Specific K,,
Environment
New Orleans
plants
Swedish plant
Belgium
factory
U.S. retirees
Asbestos,
Quebec
U.S. insulation
workers
Pennsylvania
plant
Rochedale,
England plant
Study
Code
MP9
MP10
MP11
MX13
MX14
MI15
MT16
MT17
KL Reference
Hughes et al.
1987
Albin et al.
1990
Laquet et al.
1980
Enterline et al.
1986
Selikoffand
Seidman 1991
McDonald et
al. 1983b
Petoetal. 1985
KM Reference
Hughes et al.
1987
Liddell et al.
1997 (raw data
Ioc2)
Selikoffand
Seidman 1991
McDonald et al.
1983b
Petoetal. 1985
Km, K,,,
KL
(xlOO)
0.25
0.067
0.0068
0.11
0.18
1.8
0.41
Km., Kb, Kma, Klc, and Kmc Values (continued)
Amphiboles Chrysotile
KM KL' KM' K^ KMA KLC KMC
(xlO8) (xlOO) (xlO8) (xlOO) (xlOO) (xlO8) (xlO8)
0.3 1.082185 0.811639 2.267472 16.07566 0.603147 0.020898
0.189382 0.638655 0.169882
0.092
1.3 7.411067 24.08597 5.331754 48.68671 1.418246 0.063293
1.1 4.781501 2.922028 14.72572 35.9891 3.917041 0.046786
1.31 2.39075 3.47987 3.598193 67.9231 0.957119 0.0883
7.58
-------
Table 7-20. Comparison of Spread in Range of Original and Adjusted K, and Km Values
for Specific Fiber Types
Fiber Type
Ranges of Values
Number in KL, Number in
KL.,KLX Sets KL KL. KLX KM, KM,,KMX KM KM. KMX
Sets
All fiber types
combined
Chrysotile only
(excluding
mixed)
Chrysotile only
(also excluiding
textiles)
Chrysotile and
mixed settings
Textiles only
Amphibole and
mixed settings
Amphiboles
only (excluding
mixed)
15 72 67 52 10 1,089 795 30
4 72 51 52 3 15 11 19
3 8.6 5.0 5.0 2 12 7.0 7.0
11 72 51 52 8 1,089 794 30
3 5.1 5.1 5.7 3 5.2 5.2 2.9
11 31 55 30 7 60 60 11
4 8.5 8.5 8.5 2 2.0 1.7 1.7
7.6 GENERAL CONCLUSIONS FROM QUANTITATIVE ANALYSIS OF HUMAN
EPIDEMIOLOGY STUDIES
The following conclusions result from our evaluation of the available epidemiology studies.
(1) To study the characteristics of asbestos that relate to risk, it is necessary to combine results
(i.e., in a meta analysis) from studies of environments having asbestos dusts of differing
characteristics. More robust conclusions regarding risk can be drawn from an analysis of
the set of epidemiology studies taken as a whole than results derived from individual
studies.
(2) By adjusting for fiber size and fiber type, the existing database of studies can be reconciled
adequately to reasonably support risk assessment.
(3) The U.S. EPA models for lung cancer and mesothelioma both appear to track the time-
dependence of disease at long times following cessation of exposure, however, the
relationship between exposure concentration and response may not be adequately
described by the current models for either disease. There is some evidence that these
relationships are supra-linear.
7.59
-------
(4) Whereas the U.S. EPA model for lung cancer assumes a multiplicative relationship
between smoking and asbestos, the current evidence suggests that the relationship is less
than multiplicative, but possibly more complex than additive. However, even if the
smoking-asbestos interaction is not multiplicative as predicted by the U.S. EPA model,
exposure-response coefficients estimated from the model are still likely to relate to risk
approximately proportionally and, consequently, may be used to determine an exposure
index that reconciles asbestos potencies in different environments. However, adjustments
to the coefficients may be required in order to use them to estimate absolute lung cancer
risk for differing amounts of smoking. This issue needs to be investigated further in the
next draft of this document.
(5) The optimal adjustment found for fiber size and type that best reconciles the published
literature assigns equal potency to fibers longer than 10 |im and thinner than 0.4 u.m and
assigns no potency to fibers of other dimensions. Different exposure-response coefficients
for chrysotile and amphibole are assigned both for lung cancer and mesothelioma. For
lung cancer the best estimate of the coefficient (potency) for chrysotile is 0.27 times that
for amphibole, and for mesothelioma the best estimate of the coefficient (potency) for
chrysotile is only 0.0013 times that for amphibole.
(6) Without adjustments, the lung cancer exposure-response coefficients (KL values) estimated
from 15 studies vary by a factor of 72 and these values are mutually inconsistent (based on
non-overlap of uncertainty intervals). By simultaneously adjusting these values for fiber
size and type, the overall variation in KL values across these studies is reduced to a factor
of 50.
(7) Without adjustments, the mesothelioma exposure-response coefficients (KM values)
estimated from 10 studies vary by a factor of 1,089, and these values are likewise mutually
inconsistent. By simultaneously adjusting these values for fiber size and type, the overall
variation in KM values across these studies is reduced to a factor of 30.
(8) The exposure index and exposure-response coefficients embodied in the risk assessment
approach developed in this report are more consistent with the literature than the current
U.S. EPA approach. In particular, the current approach appears highly likely to seriously
underestimate risk from amphiboles, while possibly overstating risk from chrysotile.
Furthermore, most the remaining uncertainties regarding the new proposed approach also
apply to the current approach. Consequently, we recommend that the proposed approach
begin to be applied in assessment of asbestos risk on an interim basis, while further work,
as recommended below, is being conducted to further refine the approach.
(9) The residual inconsistency in both the lung cancer and mesothelioma potency values is
primarily driven by those calculated from Quebec chrysotile miners and from South
Carolina chrysotile textile workers. The difference in the lung cancer potency estimated
between these studies has long been the subject of much attention. A detailed evaluation
of the studies addressing this issue, the results of our analysis of the overall epidemiology
literature, and implications from the broader literature, indicate that the most likely cause
of the difference between these studies is the relative distribution of fiber sizes in the two
environments. It is therefore likely that the variation between these studies can be further
7.60
-------
reduced by developing improved characterizations of the dusts that were present in each of
these environments (relying on either archived samples, or newly generated samples using
technologies similar to those used originally).
Recommendations for Limited, Further Study
The two major objectives identified above for further study are:
(1) to evaluate a broader range of exposure-response models in fitting the observed
relationship between asbestos exposure and lung cancer or mesothelioma, respectively.
For lung cancer models, this would also include an attempt to better account for the
interaction between asbestos exposure and smoking; and
(2) to develop the supporting data needed to define adjustments for exposure-response
coefficients that will allow them to be used with an exposure index that more closely
captures the criteria that determine biological activity (see Section 6.5). Among other
things, this work should focus on obtaining data that would permit more complete
reconciliation of the exposure-response coefficients derived for Quebec miners and South
Carolina textile workers.
The first of the above objectives requires access to raw data from a small number of selected,
additional epidemiology studies. The best candidate studies include: (for chrysotile) the lung
cancer data from Quebec (best) or, potentially, from the New Orleans asbestos-cement pipe plant
studied by Hughes et al. For amphiboles, the best candidate studies include: the lung cancer and
newest mesothelioma data from Libby, or, potentially the lung cancer and mesothelioma data
from the Paterson, New Jersey insulation manufacturing plant studied by Seidman et al. (1986).
The possibility of obtaining some or all of these data sets needs to be further explored.
The second of the above objectives requires more detailed size characterization data for the
environments of interest. Although archived air samples do not appear to be available from any
of the study locations of interest (except South Carolina), we believe that suitable data can be
developed from appropriate bulk samples. Thus, for example, it would be useful to obtain
samples of the raw ore from Libby, Wittenoom, and Quebec and the textile, asbestos-cement
pipe, and friction-product grade products from Quebec.
Results from our review of the supporting literature suggest that the optimum cutoff for
increased potency occurs at a length that is closer to 20 jam than 10 um, (the latter of which is
the cutoff in the exposure index provided in this study). Data do not currently exist to improve
on this latter cutoff. However, provided that study-specific size distribution data could be
obtained as indicated above, with the appropriate analyses, it will be possible to develop the size
distributions necessary to evaluate a range of considerations including:
(1) delineation of size fractions among individual length categories out to lengths as long as
30 or 40 um;
(2) determination of the relative presence and importance of cleavage fragments (ofnon-
biologically relevant sizes) in mine ores vs. finished fibers; and
7.61
-------
(3) the relative fraction of fibrous material vs. non-fibrous particles in the various exposure
dusts of interest.
7.62
-------
8.0 DISCUSSION, CONCLUSIONS, AND
RECOMMENDATIONS
Although gaps in knowledge remain, a review of the literature addressing the health-related
effects of asbestos (and related materials) provides a generally consistent picture of the
relationship between asbestos exposure and the induction of disease (lung cancer and
mesothelioma). Therefore, the general characteristics of asbestos exposure that drive the
induction of cancer can be inferred from the existing studies and can be applied to define
appropriate procedures for evaluating asbestos-related risk. Moreover, such procedures provide
substantial improvement in the confidence that can be placed in predicting risks in exposure
environments of interest compared to risk predictions based on procedures in current use.
Following a general discussions of the findings of this study, specific recommendations for
finalizing a protocol for assessing asbestos-related risks using the procedures identified in this
document are provided in Section 8.2. Recommendations are described in Section 8.3 for
limited, focused, additional studies to:
(1) settle a small number of outstanding issues (concerning whether better models
exist than the current U.S. EPA models for tracking the time and concentration
dependence of the exposure-response relationships for asbestos-induced lung
cancer and mesothelioma);
(2) improve the manner in which smoking is addressed in the modeling of asbestos-
induced lung cancer; and
(3) provide the data required to fully optimize the approach recommended in this
document (i.e., reconciling the published epidemiology studies by addressing the
effects of fiber size and type).
Moreover, these recommendations parallel those of the expert panel convened to peer-review the
previous version of this report (Appendix B).
8.1 DISCUSSION AND CONCLUSIONS
8.1.1 Addressing Issues
The issues identified in the introduction (Chapter 2) as part of the focus of this study can now be
addressed. These are:
• adequacy of existing models: whether the exposure-response models currently in
use by the U.S. EPA for describing the incidence of asbestos-related diseases
adequately reflect the time- and exposure-dependence for the development of
these diseases;
8.1
-------
• relative potency for different mineral types: whether different potencies need to
be assigned to the different asbestos mineral types to adequately predict risk for
the disease endpoints of interest;
• biodurability: to the extent that different asbestos mineral types are assigned
distinct potencies, whether the relative in vivo durability of different asbestos
mineral types determines their relative potency;
• minerals of concern: whether the set of minerals included in the current definition
of asbestos adequately covers the range of minerals that potentially contribute to
asbestos-related diseases;
• analytical methods: whether the analytical techniques and methods currently used
for determining asbestos concentrations adequately capture the biologically-
relevant characteristics of asbestos (particularly with regard to structure sizes), so
that they can be used to support risk assessment; and
• extrapolation of risk coefficients: whether reasonable confidence can be placed in
the cross-study extrapolation of exposure-response relationships that are required
to assess asbestos-related risks in new environments of interest.
The Adequacy of Existing Models. Regarding the first of the above-listed issues, both the U.S.
EPA lung cancer model and mesothelioma model appear to adequately reflect the time-
development of asbestos-induced lung cancer. For lung cancer, the assumption in the model that
risk remains constant with time following the end of exposure was confirmed for cohorts
exposed, respectively, to chrysotile, to crocidolite, and to amosite (Section 7.2.1). A similar
analysis for mesothelioma suggests that, as that model predicts, risk for mesothelioma continues
to increase with the square of time since the end of exposure (Section 7.3.1).
Regarding exposure concentration, we did find some evidence suggesting that, for both lung
cancer and mesothelioma, the relationship between exposure concentration and risk may not be
linear, but rather supra-linear (Sections 7.2.1.1 and 7.3.1.2). If confirmed, this would contradict
the assumed strictly linear relationship in both the current lung cancer and mesothelioma models.
For 15 of the 18 studies that were fit using the lung cancer model (Equation 7-2) in our analysis
(Appendix A and Section 7.2.2), the parameter, a, (which indicates differences in background
lung cancer incidence between cohorts and controls) was greater than 1.0 and significantly so in
six cases. In these cases, if it is assumed instead that the background lung cancer mortality rate
applied to the cohort is appropriate, the correct fitting would be with a=1.0, in which case the
exposure-response would appear supra-linear. Similarly, an analysis conducted using the raw
data from Wittenoom (Section 7.3.1.2) in which exposure is categorized by intensity while
controlling for both time since the end of exposure and duration of exposure suggests a supra-
linear relationship between exposure concentration and mesothelioma as well.
Due to these observations and additional concerns about the relationship between smoking and
asbestos exposure toward the induction of lung cancer (Section 7.2.1.3), evaluation of the fit of a
broader range of models to the available lung cancer and mesothelioma data is recommended.
8.2
-------
Importantly, because any such evaluation would be greatly enhanced by broadening the number
of data sets utilized, it is further recommended that all means be explored for acquiring raw data
sets for cohorts from additional epidemiology studies.
At the same time, although there are suggestions from our analysis that models other than the
current models might better describe the relationship between exposure and risk for both
mesothelioma and lung cancer any limitation associated with use of the current models for lung
cancer and mesothelioma would be common to the procedures recommended in this report and
the existing U.S. EPA approach for assessing asbestos-related risks. Therefore, given the degree
of improvement demonstrated for the proposed approach over the current U.S. EPA approach,
there appears to be little reason not to adopt the proposed approach as an interim measure, while
further research is conducted to address the remaining outstanding issues highlighted by the
expert panel (Appendix B) and highlighted in this report.
Relative Potency for Different Mineral Types. As indicated in Sections 7.4.3.2 and 7.5.2, our
analysis indicates a substantial difference in the relative potency of amphiboles and chrysotile
toward the induction of mesothelioma, with amphiboles estimated as almost 1,000 times more
potent than chrysotile (fiber-for-fiber). Moreover, this difference was shown to be highly
statistically significant. When fiber size and type are simultaneous addressed, variation across
the 10 published epidemiology studies included in our analysis drops from a factor of more than
1,000 to a factor of 30 (Table 7-19). This, coupled with the growing evidence in the literature
supporting this difference in potency among fiber types, provides strong support for defining
distinct risk coefficients for chrysotile and the amphiboles to assess the risk of mesothelioma.
The situation with lung cancer is less clear. Although our analysis suggests that the best estimate
is that (fiber-for-fiber) amphiboles are about 4 times more potent than chrysotile toward the
induction of lung cancer, this difference was not found to be statistically significant (Sections
7.4.3.1 and 7.5.2). This issue also remains unresolved in the wider literature. It is also possible
that the confounding effects of smoking, coupled with the lack of adequate data for properly
assessing the effects of smoking may be limiting our ability to address this question. Thus, this
is one of the issues that would likely benefit from additional research.
At the same time, when a small difference in potencies is incorporated into our meta analysis of
the epidemiology studies we evaluated, variation across the data set is reduced by about 30%
(Table 7-20). This suggests that incorporating a small difference in lung cancer risk coefficients
for chrysotile and the amphiboles is reasonable.
Biodurability. Because the in vitro dissolution rate for chrysotile in biological fluids is
substantially greater than for crocidolite and, likely, other amphiboles (Section 6.2.4), effects
potentially attributable to differences in the relative biodurability of these asbestos types are
addressed in several sections of this document. It is possible that such differential biodurability
at least partially explains the clear difference in potency between chrysotile and the amphiboles
toward mesothelioma (along with the possible, albeit smaller, difference toward lung cancer).
However, that no difference was observed in the time-development of either lung cancer or
mesothelioma following exposure to chrysotile or amphibole asbestos (respectively) suggests
that any relationship that exists between potency and biodurability must be more subtle and
complicated than the obvious effect on internal dose. At the same time, there is ample literature
8.3
-------
evidence that some kind of relationship in fact exists (Section 6.2.4). While more research into
this relationship may prove interesting (and may be useful for assessing effects of less durable
fibers), it is unlikely to have a direct impact on procedures for assessing asbestos-related risk.
Minerals of Concern. Regarding the range of fibrous minerals that potentially contribute to
lung cancer and mesothelioma, available evidence (Sections 6.2 and 6.4) suggests that:
• several minerals and the most biodurable among synthetic fibers (such as erionite
or refractory ceramic fibers) in addition to those included strictly within the
definition for asbestos have been shown capable of inducing lung cancer and/or
mesothelioma (as long as the corresponding fibers fall within the appropriate size
range);
• fibrous minerals that differ radically in chemical composition or crystal structure
(such as erionite, chrysotile, and the amphibole asbestos types) appear to exhibit
substantially different potencies (even after adjusting for size); and
• fibrous minerals that exhibit closely related chemical compositions and crystal
structures (such as the family of amphibole asbestos types) appear to exhibit
relatively similar potencies (once effects are adjusted for size). To illustrate this
point for mesothelioma, consider that the range of variation in estimated KM's
over 10 studies (which include studies of exposures to tremolite, amosite, and
crocidolite in addition to chrysotile) is reduced to a factor of 30 (from almost
1,100), once the effects of size and differences between the amphiboles and
chrysotile are accounted for (Table 7-20). Moreover, among the 7 of these
studies reflecting exposure exclusively to amphiboles (which still includes
exposures to tremolite1, crocidolite, and amosite), the range of variation is
reduced to a factor of 11 (Table 7-20). Regarding lung cancer, the four studies of
"pure" amphibole environments (which includes exposures to tremolite,
crocidolite, and amosite) vary by only a factor of 8.5 (Table 7-20) and this range
is bounded by studies of the same mineral: amosite, which certainly suggests that
mineralogical differences do not drive the observed variation (Section 7.2.2).
Given the above-described observations, it is clear that fibrous minerals beyond those included
in the definition for asbestos can contribute to lung cancer and mesothelioma. It is also likely
that potencies for minerals that exhibit similar chemistry and crystal structure (and which
therefore likely exhibit similar physical-chemical properties) also exhibit corresponding potency
(for similarly sized fibers). However, the carcinogenicity of fibers exhibiting radically different
chemical compositions and crystal structures than those already identified as carcinogenic should
be evaluated on a case-by-case basis.
Two additional considerations may be useful for focusing such evaluations. First, the size
distribution for fibers composed of the mineral of concern should be shown to include substantial
Formally, the exposures at Libby are to the amphibole mineral richterite. However, this mineral is closely
related to tremolite and has sometimes been called "sodium tremolite" because it contains a greater fraction of
sodium than the composition range that is commonly termed tremolite.
8.4
-------
numbers within the range of structures that potentially contribute to biological activity (e.g., that
fall within the size range defined by the improved index recommended in this study, i.e.,
structures longer than 10 p,m and thinner than 0.4 (im, see Section 7.5). Second, such fibers
should also be shown to be relatively biodurable (i.e., that they exhibit dissolution rates less than
approximately 100 ng/cm2-hour, Lippmann 1999).
Analytical Methods. Given the need to detect the thinnest fibers and the need for reliable
measurements in outdoor settings, the only analytical technique that appears to be capable of
providing quantitative data useful for supporting risk assessment is TEM (Sections 4.3 and 7.6).
Further, given the specific size range of structures that need to be evaluated and the specific
manner in which they need to be counted (to assure both cross-study comparability and
compatibility with the recommended dose-response coefficients), analyses should be performed
using the specific analytical methods recommended in this document. These are ISO 10312
(modified to focus on interim index structures) for air and the Modified Elutriator method
(Berman and Kolk 2000) for soils or bulk materials. However, on a study-specific basis, other
methods may be shown to provide comparable results so that they can also be used as part of a
properly integrated investigation.
8.1.2 Comparison with Other, Recent Risk Reviews
Although several other reviews have also recently been published that nominally address risk-
related issues for asbestos (including questions concerning the identification of an appropriate
exposure index and the relative potency of varying fiber types), these studies are either
qualitative or involve analysis of data in a manner that does not allow formal evaluation of the
nature of specific exposure-response relationships for the various diseases. Therefore, they are
not well suited to support development of a protocol for conducting formal assessment of
asbestos-related risks. Nevertheless, the general conclusions from these reviews are not
inconsistent with our findings.
Hodgson and Darnton (2000). Hodgson and Darnton (2000) conducted a comprehensive
quantitative review of potency of asbestos for causing lung cancer and mesothelioma in relation
to fiber type. They concluded that amosite and crocidolite were, respectively, on the order of
100 and 500 times more potent for causing mesothelioma than chrysotile. They regarded the
evidence for lung cancer to be less clear cut, but concluded nevertheless that amphiboles
(amosite and crocidolite) were between 10 and 50 times more potent for causing lung cancer
than chrysotile. In reaching this latter conclusion they discounted the high estimate of chrysotile
potency obtained from the South Carolina cohort. Hodgson and Darnton concluded that inter-
study comparisons for amphibole fibers suggested non-linear exposure-response relationships for
lung cancer and mesothelioma, although a linear relationship was possible for pleural
mesothelioma and lung tumors, but not for peritoneal mesothelioma.
The Hodgson and Darnton study was based on 17 cohorts, 14 of which were among the 20
included in the present evaluation. This study had different goals from the present evaluation
and used different methods of analysis. Hodgson and Darnton did not use the exposure-response
information within a study. Instead, lung cancer potency was expressed as a cohort-wide excess
mortality divided by the cohort mean exposure. Likewise, mesothelioma potency was expressed
as the number of mesothelioma deaths divided by the expected total number of deaths,
8.5
-------
normalized to an age of first exposure of 30, and by the mean exposure for the cohort. These
measures have the advantage of being generally calculatable from the summarized data available
from a study. However, since they are not model-based, it is not clear how they could be used to
assess lifetime risk from a specified exposure pattern, which is an important goal of the current
project. Use of average cohort exposure could cause biases in the estimates, if, e.g., a large
number of subjects were minimally exposed. Also, the recognized differences between studies
in factors, in addition to level of exposure, that may affect risk will also affect the reliability of
conclusions concerning the dose response shape based on comparisons of results across studies.
Lippmann (1994,1999). In the most recent of these reviews, Lippmann (1999) reinforces our
general findings that it is longer fibers (those longer than a minimum of approximately 5 \im)
that contribute to lung cancer and mesothelioma. He further indicates that, based primarily on
the limits observed for fibers that can be phagocytized, fibers that contribute most to lung cancer
are likely longer than a minimum of 10 urn. In his review, based on a series of comparisons of
mean and median dimensions reported for the relevant exposures across a broad range of studies,
Lippmann draws several fairly specific conclusions on the ranges of fiber sizes that may
contribute to various diseases (i.e., that the minimum length fibers that contribute to asbestosis,
lung cancer, and mesothelioma are 2, 5, and 10 u.m, respectively). He also suggests that fibers
that contribute to mesothelioma may need to be thinner than 0.1 urn while those that contribute
to lung cancer may need to be thicker than 0.15 |im. While it is not clear that drawing such
specific conclusions can be firmly supported by the kinds of qualitative comparisons across
reported mean and median dimensions for exposures in various studies that are described in this
paper, the author indicates that further, more formal study of the dose-response relationships that
he posits is warranted. It is noted that many of the studies reviewed by Lippmann (1999) are
also incorporated in our analysis.
In the earlier review, Lippmann (1994) plots lung tumor incidence as a function of inhaled
animal dose for data from a series of broadly varying studies based, respectively, on fibers
longer than 5, 10, and 20 ^m (no widths considered) and suggests that the quality of the fits are
comparable. The author further suggests, based on this evaluation, that PCM seems to provide a
reasonable index of exposure. However, no formal goodness-of-fit tests were performed in this
analysis and, based on visual inspection, none of the plots would likely show an adequate fit.
Moreover, the plot of the tumor response vs. dose as a function of fibers longer than 5 urn
appears to be substantially worse than the other two plots; if one removes the single highest point
in this plot, it appears that any correlation will largely disappear.
Stayner et al. 1996. In the context of evaluating the "amphibole hypothesis", Stayner et al.
(1996) computed the excess relative risk of lung cancer per fiber/ml/year from 10 studies
categorized by the fiber types to which the cohort was exposed. Each of these studies was also
included in the present evaluation. Both the lowest and highest excess relative risks came from
cohorts exposed exclusively to chrysotile. Based on their evaluation, they concluded that the
epidemiologic evidence did not support the hypothesis that chrysotile asbestos is less potent than
amphibole for inducing lung cancer. However, based on a review of the percentage of deaths in
various cohorts from mesothelioma, they concluded that amphiboles were likely to be more
potent than chrysotile in the induction of mesothelioma. They also noted that comparison of the
potency of different forms of asbestos are severely limited by uncontrolled differences in fiber
sizes. None of these conclusions are inconsistent with our general findings.
8.6
-------
8.2 RECOMMENDATIONS FOR ASSESSING ASBESTOS-RELATED RISKS
The optimum values for the risk coefficients for lung cancer and mesothelioma (the adjusted KL
and KM values) derived in our analysis are provided in Table 7-18. Although these values are
optimized (within the constraints of the current analysis) and use of these values reduces the
apparent variation across published studies substantially (Section 7.5.2), the need to manage and
minimize risk when developing a general approach for assessing risk, is also recognized. Thus,
to reduce the chance of under-estimating risks, a conservative set of potency estimates were
developed (Table 8-1) by adjusting upward the best-estimate potency coefficients listed in
Table 7-18 to provide additional health protectiveness. The manner in which this was
accomplished is described in the following paragraphs.
Table 8-1. Conservative Risk Coefficients for
Pure Fiber Types
Fiber Type
Chrysotile
Amphiboles
KL. \ 100
5.5
20
KM.xlO"
0.15
100
Reviewing the KLc and KLa dose response coefficients values listed in Table 7-19, it is apparent
that the maximum values for lung cancer derive from the Dement et al. (1994) study of the South
Carolina textile plant. As indicated in Appendix A, the South Carolina study is a high quality
study. Moreover, we were able to obtain the raw data from this study and have analyzed these
data in detail. We found, among other things, a well-behaved (i.e., monotonic) exposure-
response trend in this study for lung cancer. Therefore, because this study is of high quality and
provides the largest values of KLc and KLa, the values from this study were selected for our
conservative estimates of the corresponding potency coefficients.
Reviewing the KMa and KMc values listed in Table 7-19, it is apparent that the maximum values
derive from the Finkelstein study (1984) of the Ontario asbestos-cement factory. As indicated in
Appendix A, the data from this study appear problematic. Among other things, the exposure-
response relationships observed in this study are not well-behaved (i.e., not monotonic).
Moreover, the value for a estimated for this study is the highest of any study we evaluated.
Possible reasons for such a large a include large discrepancies between the background
incidence of lung cancer between cohort and controls in this study and/or serious errors in
exposure estimates. Given the potential problems associated with this study, (which suggests
that the potency coefficients estimated from this study may be less reliable than for many of the
other available studies), we decided to bypass this study and select the next highest values in
Table 7-19 for KMa and KMc as the conservative estimates to be recommended in this study.
Interestingly, these also turn out to be from the South Carolina textile study. Note that the
difference in the estimates from these two studies vary by less than a factor of two in any case.
Based on the above evaluation, conservative estimates for the various potency coefficients
recommended in this report are presented in Table 8-1.
8.7
-------
Importantly, a measure of the degree of reconciliation among the results of the published
epidemiology studies that has been accomplished by the analysis presented in this report is
indicated by the ratios of the values presented in Tables 7-18 and 8-1, respectively. The ratios
between the corresponding coefficients in Tables 7-18 and 8-1 are no more than 10 for the lung
cancer potency coefficients and no more than 4 for the mesothelioma potency coefficients.
Moreover, the bounding study for the values presented in Table 8-1 is once again the South
Carolina textile study, which further reinforces earlier discussions identifying the particular need
to reconcile this study with the other chrysotile studies (particularly the Quebec mining study).
To assess risk, depending on the specific application, either the best-estimate risk coefficients
presented in Table 7-18 or the conservative estimates presented in Table 8-1 can be incorporated
into procedures described below for assessing asbestos-related risks.
Tables 8-2 and 8-3 present estimates of the additional risk of death from lung cancer, from
mesothelioma, and from the two diseases combined that are attributable to lifetime, continuous
exposure at an asbestos concentration of 0.0001 f/cm3.(for fibrous structures longer than 10 ^m
and thinner than 0.4 ^m) as determined using TEM methods recommended herein. Table 8-2
was developed using the best-estimate values for risk coefficients defined in Table 7-18, and
Table 8-3 was developed using the conservative estimates defined in Table 8-1. Separate risk
estimates are provided for males and females and for smokers and non-smokers. The method
used to construct these tables is described in detail in Appendix E.
Separate estimates are presented for smokers and nonsmokers because the lifetime asbestos-
induced risk of both lung cancer and mesothelioma differ between smokers and non-smokers.
The asbestos-induced risk of lung cancer is higher among smokers because the lung cancer
model (Equation 7-2) assumes that the increased mortality rate from lung cancer risk due to
asbestos exposure is proportional to background lung cancer mortality, which is higher among
smokers. Note that, while this is consistent with a multiplicative effect between smoking and
asbestos exposure that has been reported by several researchers (see, for example, Hammond et
al. 1979), some of the latest studies of the interaction between smoking and asbestos exposure
suggest a more complicated relationship (Section 7.2.1.3). This issue needs to be addressed
more fully in future analyses of these data. However, we believe the effects of such
considerations on the overall accuracy of asbestos-related risk estimates is likely to be small
relative to other sources of error.
8.8
-------
Table 8-2. Estimated Additional Deaths from Lung Cancer or Mesothelioma per
100,000 Persons from Constant Lifetime Exposure to 0.0001 TEM f/cc Longer than 10
um and Thinner than 0.4 urn - Based on Optimum Risk Coefficients (Table 7-18)
Chrysotile
NonSmokers
Lung Cancer
Mesothelioma
Combined
Males
0.185
0.0836
0.269
Females
0.207
0.096
0.303
Smokers
Males
1.6
0.0482
1.65
Females
1.5
0.0702
1.57
Amphibole
Lung Cancer
Mesothelioma
Combined
0.2
62.7
62.9
0.286
72.3
72.5
2.22
36.1
38.3
2.47
52.7
55.1
Table 8-3. Estimated Additional Deaths from Lung Cancer or Mesothelioma per
100,000 persons from Constant Lifetime Exposure to 0.0001 TEM f/cc Longer than 10
um and Thinner than 0.4 um - Based on Conservative Risk Coefficients (Table 8-1)
Chrysotile
Non-Smokers
Lung Cancer
Mesothelioma
Combined
Males
1.7
0.314
2.02
Females
1.9
0.361
2.26
Smokers
Males
14.7
0.181
14.9
Females
13.8
0.263
14
Amphibole
Lung Cancer
Mesothelioma
Combined
3.77
209
213
4.41
241
245
34.1
120
154
33.2
175
209
8.9
-------
If there is a desire to generate population averaged risks, this can also be accomplished using the
data in Tables 8-2 or 8-3. For such a case, simply choose the appropriate row (for lung cancer,
mesothelioma, or combined risk) from either Table 8-2 or 8-3 and substitute the four values
given in the row into the following equation to derive a single, population averaged risk:
Ravg = 0.5*[0.214*(MS + FS) + 0.786*(MNS + FNS) (Eq. 8-1)
Where:
Ravg is the population averaged risk for the chosen disease endpoint;
MS is the corresponding risk for male smokers;
FS is the corresponding risk for female smokers;
MNS is the corresponding risk for male non-smokers; and
FNS is the corresponding risk for female non-smokers.
Note that Equation 8-1 is derived based on the assumption that 21.4% of the general population
smokes (see Appendix E).
The asbestos-induced risk of mesothelioma is smaller among smokers because the mesothelioma
model (Equation 7-6) assumes that risk from constant exposure increases rapidly with age, with
the result that the predicted mortality rate is highest among the elderly. Thus, since smokers
have a shorter life span than non-smokers, their risk of dying from mesothelioma is also
predicted to be smaller.
Risks from lifetime exposures to asbestos levels other than 0.0001 may be estimated from the
appropriate entry in Tables 8-2 or 8-3 by multiplying the value in the selected cell from the table
by the airborne asbestos concentration of interest and dividing by 0.0001 (i.e., by assuming that
the additional risk is proportional to the asbestos exposure level). Airborne asbestos
concentrations to be used in this manner must be estimates of lifetime average exposure and must
be expressed as structures longer than 10 [am and thinner than 0.4 um derived as described
below.
Importantly, the risks provided in Tables 8-2 and 8-3 relate to exposure estimates expressed in
terms of the interim exposure index (i.e., estimates including only asbestos structures longer than
10 |im and thinner than 0.4 |j,m). Only exposures expressed in terms of the same exposure index
can be used to adjust the risk estimates to other levels of exposure (in the manner described
above). Use of exposures expressed in terms of any other index of exposure (such as the PCME
index in current use) will result in invalid estimates of risk.
The procedure described above for estimating risks using Tables 8-2 or 8-3 should provide good
approximations as long as the projected risk is no greater than 1,000 per 100,000. Risks greater
than 1,000 per 100,000 (i.e., 1 in 100) that are derived from the tables are likely to be over-
estimated. However, for risks associated with short-duration exposures or exposures that differ
radically over time, it may be better to use a lifetable analysis or a modified version of Tables
8-2 or 8-3 that reflect the differences in exposure duration and frequency. This is to avoid
substantially under- or over-estimating risk (depending on how the table might otherwise be
applied).
Tables 8-2 and 8-3 were derived using the approach described in Appendix E by incorporating
8.10
-------
the age-, sex-, and smoking-specific death rates reported for the general U.S. population and
assuming that exposure is constant and continuous at the level indicated in the table. The
underlying models are provided in Chapter 7 for cases in which exposure is not constant
throughout life and for which sufficient data exist to characterize the time-dependence of such
exposure. If available, there may also be cases in which it is advantageous to employ mortality
data from a control population that better matches the exposed population of interest than the
U.S. population as a whole.
For the interim, it is recommended that asbestos-related risks from constant low-level exposures
be estimated using Tables 8-2 and 8-3. Although it is possible also to use the tables to estimate
risk from short-term exposures by applying the corresponding long-term average exposure
(derivedfrom appropriate measurements, as described below), this method can result in
significant errors in some cases. It is anticipated that a flexible and user-friendly software
package for evaluating risk will eventually be developed to supplement this document. Such a
package will be capable of accurately implementing the calculation method presented in
Appendix E to calculate risks from general exposure patterns, rather than from constant
exposures only.
Requirements for Asbestos Measurements. One additional advantage of the approach for
evaluating asbestos-related risks recommended in this document (in comparison to the current
approach) is that the procedure for assessing risks is tied unambiguously to a specific index for
measuring and expressing exposure (i.e., the index of all fibrous structures longer than 10 urn
and thinner than 0.4 urn, as defined in Section 7.5) and this, in turn, is tied unambiguously to
requirements for analyzing asbestos samples.
Estimates of airborne asbestos concentrations that are required to support risk assessment can be
derived either by extrapolation from airborne measurements or by modeling release and
dispersion of asbestos from sources (soils or other bulk materials). In either case, exposure
estimates must be representative of actual (time-dependent) exposure and must provide
measurements that include only fibrous structures satisfying the dimensional criteria listed in the
last paragraph. Additional considerations that need to be addressed to assure the validity of risk
estimates derived using this protocol include:
• the array of samples collected for estimating airborne asbestos concentrations
must be representative of the exposure environment;
• the time variation of airborne asbestos concentrations must be properly addressed;
• airborne samples must be collected on membrane filters that are suitable for
preparation for analysis by transmission electron microscopy (TEM).
Appropriate rocedures for sample collection are described in Chatfield and
Berman (1990) or the ISO Method (ISO 10312)2;
2Note that the ISO Method (ISO 10312) is a refinement of the method originally published as the Interim
Superfund Method (Chatfield and Berman 1990). It incorporates improved rules for evaluating fiber morphology.
Both methods derive from a common development effort headed by Eric Chatfield.
8.11
-------
• sample filters must be prepared for analysis using a direct transfer procedure (e.g.,
ISO 10312). Should indirect preparation be required (due, for example, to
problems with overloading of sample filters), a sufficient number of paired
samples will need to be collected and analyzed to establish a site-specific
correlation between directly and indirectly prepared samples;
• samples must be analyzed by TEM;
• samples must be analyzed using the counting and characterization rules defined in
ISO 10312 and the structures used to determine exposure concentrations for use
with this protocol need to satisfy the dimensional criteria defined in Section 7.5
(i.e., structures longer than 10 um and thinner than 0.4 jim). Importantly, ISO
Method rules require separate enumeration and characterization of component
fibers and bundles that are observed within more complex clusters and matrices.
Such components, if they meet the dimensional criteria defined above must be
included in the structure count;
• when risks are estimated using the risk tables (Tables 8-2 or 8-3) the risk values
selected from the tables must be appropriate for the fiber type (i.e., chrysotile or
amphibole) and the disease endpoint (i.e., lung cancer or mesothelioma) relevant
to the situation of interest; and
• to use Tables 8-2 or 8-3, the concentration of total asbestos structures longer than
10 um and thinner than 0.4 u.m must be derived, divided by 0.0001, and
multiplied by the risk estimate listed in the appropriate cell of the selected table to
generate the risk estimate of interest.
Considerations that need to be addressed to assure the validity of risk estimates derived from soil
or bulk measurements combined with release and transport modeling include:
• the array of samples collected for estimating source concentrations must be
representative of the surface area or volume of source material from which
asbestos is expected to be released and contribute to exposure;
• samples must be prepared and analyzed using the Modified Elutriator Method for
soils and bulk materials (Herman and Kolk 1997, 2000), which is the only method
capable of providing bulk measurements that can be related to risk;
• membrane filter samples prepared using the tumbler and vertical elutriator per the
Superfund method must themselves be prepared for TEM analysis using a direct
transfer procedure;
• TEM analysis must be conducted using the counting and characterization rules
defined in the ISO Method (ISO 10312) in precisely the same manner that is
described above for air measurements. Also, the same size categories need to be
evaluated (in the same manner described above) to estimate exposures for use
with this protocol to assess risk; and
8.12
-------
• release and dispersion models that are selected for assessing risks must be
appropriate to the exposure scenario and environmental conditions of interest.
Such models must also be adapted properly so that they accept input estimates
expressed in terms of fiber number concentrations. Procedures suggested for
adapting such models are illustrated in a recent publication (Berman 2000).
Note, if new analytical procedures can be designed to focus on long structures, risks can be
evaluated more cost-effectively. The alternate approach of spending a large proportion of
available resources counting many (potentially non-potent or marginally potent) short structures,
while not characterizing longer structures with adequate sensitivity or precision, leaves open the
possibility of missing potentially serious hazards because a small population of extremely potent,
long fibers were missed in a particular environment. Moreover, any potential contribution to risk
by shorter structures will be incorporated to some extent by default, i.e., to the extent that similar
proportions of such structures were also present in the environments from which the exposure-
response coefficients were derived and such structures are known to have been ubiquitous in
these environments (see, for example, Dement and Harris 1979; Gibbs and Hwang 1980; Hwang
and Gibbs 1981).
8.3 RECOMMENDATIONS FOR FURTHER STUDY
A small number of limited and focused studies (described previously) are recommended in this
document because they are likely to provide very cost-effective improvements to the quality of
this document and may support substantial improvement to the recommended procedures for
assessing asbestos-related risks. The recommended studies are:
(1) a focused study to expand our evaluation of the current U.S. EPA models to
include other candidate models that might better track the exposure dependence
of asbestos-related disease (Sections 7.2.2 and 7.3.2). Such models should also
be used to explore and better represent the relationship between smoking and
asbestos exposure (to the extent that data suitable for supporting such an analysis
can be acquired); and
(2) a focused study to develop the supporting data needed to define adjustments for
potency factors that will allow them to be used with an exposure index that even
more closely captures asbestos characteristics that determine biological activity
than the currently proposed index (Section 7.5).
Note that, by properly designing the second of the above-listed studies, it may also be possible to
further address another outstanding issue that was previously identified: the question of whether
exposure-response coefficients derived from mining studies are under-estimated relative to
studies involving asbestos products because exposures in the mining studies may contain large
numbers of non-asbestos particles contributed by the disturbance of host rock (Appendix D).
8.13
-------
9.0 REFERENCES
Aalto M; Heppleston AG. Fibrogenesis by Mineral Fibres: An In-Vitro Study of the Roles of the
Macrophage and Fibre Length. British Journal of Experimental Pathology. 65:91-99. 1984.
Adamson IYR. Early Mesothelial Cell Proliferation After Asbestos Exposure: In Vivo and
In Vitro Studies. Environmental Health Perspectives. 105(Suppl 5): 1205-1208. September.
1997.
Addison J. Consultant Mineralogist, John Addison Consultancy, Cottingham, East Yorkshire,
UK. 2001. (private communication)
Afaq F; Abidi P; Matin R; Rahman Q. Activation of Alveolar Macrophages and Peripheral Red
Blood Cells in Rats exposed to Fibers/Particles. Toxicology Letters. 99:175-182. 1998.
Albin M; Jakobsson K; Attewell R; Johansson L; Welinder H. Mortality and Cancer Morbidity
in Cohorts of Asbestos Cement Workers and Referents. British Journal of Industrial Medicine.
79(9):602-610. September. 1990.
Albin M; Pooley FD; Stromberg U; Attewell R; Mitha R; Johansson L; Welinder H. Retention
Patterns of Asbestos Fibres in Lung Tissue Among Asbestos Cement Workers. Occupational
and Environmental Medicine. 51:205—211. 1994.
Albin M; Magnani C; Krstev S; Rapiti E; Shefer I. Asbestos and Cancer: An Overview of
Current Trends in Europe. Environmental Health Perspectives. 107(2):289-298. May. 1999.
Amandus HE; Wheeler R. The Morbidity and Mortality of Vermiculite Miners and Millers
Exposed to Tremolite-Actinolite: Part II. Mortality. American Journal of Industrial Medicine
11:15-26. 1987.
Amandus HE; Wheeler R; Jankovic J; Tucker J. The Morbidity and Mortality of Vermiculite
Miners and Millers Exposed to Tremolite-Actinolite: Part I. Exposure Estimates. American
Journal of Industrial Medicine. 11:1-14. 1987.
Armstrong BK; de Klerk NH; Musk AW; Hobbs MST. Mortality in Miners and Millers of
Crocidolite in Western Australia. British Journal of Industrial Medicine. 45:5-13. 1988.
Asgharian B; Yu CP. Deposition of Inhaled Fibrous Particles in the Human Lung. Journal of
Aerosol Medicine. 1:37-50. 1988.
Barchowsky A; Lannon BM; Elmore LC; Treadwell MD. Increased Focal Adhesion Kinase-
and Urokinase-Type Plasminogen Activator Receptor-Associated Cell Signaling in Endothelial
Cells Exposed to Asbestos. Environmental Health Perspectives. 105:(Suppl )5:1131-1137.
September. 1997.
9.1
-------
Barchowsky A; Roussel RR; Krieser RJ; Mossman BT; Treadwell MD. Expression and Activity
of Urokinase and its Receptor in Endothelial and Pulmonary Epithelial Cells Exposed to
Asbestos. Toxicology and Applied Pharmacology. 152(2):388-396. 1998.
Baris YI; Simonato L; Artvinli M; Pooley F; Saracci R; Skidmore J; Wagner C.
Epidemiological and Environmental Evidence of the Health Effects of Exposure to Erionite
Fibres: A Four-Year Study in the Cappadocian Region of Turkey. InternationalJournal of
Cancer. 39:10-17. 1987.
Bauman MD; Jetten AM; Bonner JC; Kumar RK; Bennett RA; Brody AR. Secretion of a
Platelet-Derived Growth Factor Homologue by Rat Alveolar Macrophages Exposed to
Particulates In Vitro. European Journal of Cell Biology. 51:327-334. 1990.
Beckett ST. The Generation and Evaluation of UICC Asbestos Clouds in Animal Exposure
Chambers. Annals of Occupational Hygiene. 18:187-198. 1975.
Bellman B; Konig H; Muhle H; Pott F. Chemical Durability of Asbestos and of Man-Made
Mineral Fibres In Vivo. Journal of Aerosol Science. 17(3):341-345. 1986.
Bellman B; Muhle H; Pott F; Konig H; Kloppeel H; Spurny K. Persistence of Man-Made Fibers
(MMF) and Asbestos in Rat Lungs. Annals of Occupational Hygiene. 31:693-709. 1987.
Berman DW. Asbestos Measurement in Soils and Bulk Materials: Sensitivity, Precision, and
Interpretation - You Can Have It All. In Advances in Environmental Measurement Methods for
Asbestos. ASTM STP 1342. Beard ME; Rook HL (eds.). American Society for Testing and
Materials. 2000.
Berman DW. President, Aeolus, Inc., Albany, CA. (unpublished data)
Berman DW Chatfield EJ. Interim Superfund Method for the Determination of Asbestos in
Ambient Air. Part 2: Technical Background Document. Office of Solid Waste and Remedial
Response. U.S. Environmental Protection Agency, Washington, D.C. EPA/540/2-90/005b.
May. 1990.
Berman DW and Crump KS. Methodology for Conducting Risk Assessments at Asbestos
Superfund Sites. Part 2: Technical Background Document. Prepared for: Kent Kitchingman,
U.S. Environmental Protection Agency, Region 9. Work Assignment No. 59-06-D800 under
Contract No. 68-W9-0059. 1999.
Berman DW; Crump KS. Technical Support Document for a Protocol to Assess Asbestos-
Related Risk. Final Draft. Prepared for the Volpe Center, U.S. Dept. of Transportation,
Cambridge, MA and the U.S. EPA, Region 8, Denver, CO. Prepared under Contract No.
DTRS57-01-C-10044. September. 2001.
Berman DW; Kolk AJ. Superfund Method for the Determination of Asbestos in Soils and Bulk
Materials. Office of Solid Waste and Emergency Response. U.S. Environmental Protection
Agency, Washington, D.C., EPA 540-R-97-028. 1997.
9.2
-------
Berman DW; Kolk AJ. Draft: Modified Elutriator Method for the Determination of Asbestos in
Soils and Bulk Materials. Aeolus, Inc. 751 Taft St., Albany, CA 94706. 2000. (unpublished)
Berman DW; Crump KS; Chatfield EJ; Davis JMG; Jones AD. The Sizes, Shapes, and
Mineralogy of Asbestos Structures that Induce Lung Tumors or Mesothelioma in AF/HAN Rats
Following Inhalation. Risk Analysis. 15(2):181-195. 1995.
Berman DW; Crump KS; Chatfield EJ; Davis JMG; Jones AD. The Size Distribution of Fibers
and Particles in Airborne Dust Generated for Selected Animal Inhalation Studies, (in
preparation)
Bernstein DM; Morscheidt C; Grimm H-G; Thevenaz P; Teichert U. Evaluation of Soluble
Fibers Using the Inhalation Biopersistence, a Nine-Fiber Comparison. Inhalation Toxicology.
8:345-385. 1996.
Berry G; Newhouse ML. Mortality of Workers Manufacturing Friction Materials Using
Asbestos. British Journal of Industrial Medicine. 40:1-7. 1983.
Bertrand R; Pezerat H. Fibrous Glass: Carcinogenicity and Dimensional Characteristics. In
Biological Effects of Mineral Fibres. Wagner JC (ed.). IARC Scientific Publications.
pp. 901-911. 1980.
Bignon J; Monchaux G; Sebastien P; Hirsch A; Lafuma J. Human and Experimental Data on
Translocation of Asbestos Fibers Through the Respiratory System. Annals of New York
Academy of Sciences. 330:745-750. 1979.
Blake T; Castranova V; Schwegler-Berry D; Baron P; Deye GL; Li C; Jones W. Effect of Fiber
Length on Glass Microfiber Cytotoxity. Journal of Toxicology and Environmental Health.
54(Part A):243-259. 1998.
Bolton RE; Davis J; Donaldson K; Wright A. Variation in the Carcinogenicity of Mineral
Fibres. Annals of Occupational Hygiene. 26(l-4):569-582. 1982.
Bolton RE; Davis JMG; Miller B; Donaldson K; Wright A. The Effect of Dose of Asbestos on
Mesothelioma Production in the Laboratory Rat. Proceedings of the Vfh International
Pneumonoconiosis Conference. 2:1028—1035. 1984.
Bolton RE; Addison J; Davis JMG; Donaldson K; Jones AD; Miller BG; Wright A. Effects of
the Inhalation of Dusts from Calcium Silicate Insulation Materials in Laboratory Rates.
Environmental Research. 39:26-43. 1986.
Bonneau L; Malard C; Pezerat H. Studies on Surface Properties of Asbestos. Environmental
Research. 41:268-275. 1986.
Boutin C; Dumortier P; Rey F; Viallat JR; De Vuyst P. Black Spots Concentrate Oncogenic
Asbestos Fibers in the Parietal Pleura: Thoracoscopic and Mineralogic Study. American Journal
of Respiratory and Critical Care Medicine. 153:444-449. 1996.
9.3
-------
Brass DM; Hoyle GW; Poovey HG; Liu J-Y; Brody AR. Reduced Tumor Necrosis Factor-
Alpha and Transforming Growth Factor-Beta-1 Expression in the Lungs of Inbred Mice That
Fail to Develop Fibroproliferative Lesions Consequent to Asbestos Exposure. American Journal
of Pathology. 154(3)853-862. 1999.
Broaddus VC; Yang L; Scabo LM; Ernst JD; Boylan AM. Crocidolite Asbestos Induces
Apoptosis of Pleural Mesothelial Cells: Role of Reactive Oxygen Species and Poly(ADP-
ribosyl) Polymerase. Environmental Health Perspectives. 105(Suppl 5): 1147-1152.
September. 1997.
Brody AR; Hill LH; Adkins B; O'Connor RW. Chrysotile Asbestos Inhalation in Rats:
Deposition Pattern and Reaction of Alveolar Epithelium and Pulmonary Macrophages.
American Review of Respiratory Diseases. 123:670-679. 1981.
Brody AR; Liu J-Y; Brass D; Corti M. Analyzing the Genes and Peptide Growth Factors
Expressed in Lung Cells In Vivo Consequent to Asbestos Exposure. Environmental Health
Perspectives. 105(Suppl 5):1165-1171. September. 1997.
Brown DM; Fisher C; Donaldson K. Free Radical Activity of Synthetic Vitreous Fibers: Iron
Chelation Inhibits hydroxyl Radical Generation by Refractory Ceramic Fiber. Journal of
Toxicology and Environmental Health Part A. 53(7):545-561. 1998.
Campbell WJ; Huggins CW; Wylie AG. Chemical and Physical Characterization of Amosite,
Chrysotile, Crocidolite, and Nonfibrous Tremolite for Oral Ingestion Studies by the National
Institute of Environmental Health Sciences. Report of Investigations 8452. National Library of
Natural Resources, U.S. Department of the Interior. 1980.
Case BW; Dufresne A. Asbestos, Asbestosis, and Lung Cancer: Observations in Quebec
Chrysotile Workers. Environmental Health Perspectives. 105(Suppl 5):1113-1120. September.
1997.
Case BW; Dufresne A; McDonald AD; McDonald JC; Sebastien P. Asbestos Fiber Type and
Length in Lungs of Chrysotile Textile and Production Workers: Fibers Longer than 18 urn.
Inhalation Toxicology. l(Suppl 1):411-418. 2000.
Castranova V; Vallyathan V; Ramsey DM; McLaurin JL; Pack D; Leonard S; Barger MW; Ma
JUC; Dalai NS; Teass A. Augmentation of Pulmonary Reactions to Quartz Inhalation by Trace
Amounts of Iron-Containing Particles. Environmental Health Perspectives. 105(Suppl
5): 1319-1324. September. 1997.
Chang L-Y; Overby LH; Broday AR; Crapo JD. Progressive Lung Cell Reactions and
Extracellular Matrix Production After a Brief Exposure to Asbestos. American Journal of
Pathology. 131(1):156-170. 1989.
Chao C-C; Park S-H; Aust AE. Participation of Nitric Oxide and Iron in the Oxidation of DNA
in Asbestos-Treated Human Lung Epithelial Cells. Archives of Biochemistry and Biophysics.
326(1):152-157. 1996.
9.4
-------
CHAP (Chronic Hazard Advisory Panel on Asbestos). Report to the U.S. Consumer Product
Safety Commission. July. 1983.
Chatfield EJ. Ambient Air: Determination of Asbestos Fibres, Direct Transfer Transmission
Electron Microscopy Procedure. Submitted to the International Standards Organization: ISO/TC
10312. 1995.
Chatfield EJ; Berman DW. Interim Superfund Method for the Determination of Asbestos in
Ambient Air. Part 1: Method. Office of Solid Waste and Remedial Response. U.S.
Environmental Protection Agency, Washington, D.C. EPA/540/2-90/005a. May. 1990.
Chen YK. Ph.D. Dissertation. Mechanical and Aerospace Engineering, State University of New
York at Buffalo. 1992.
Chen YK; Yu CP. Deposition of Charged Fiber in the Human Lung. Journal of Aerosol
Science. 24(Suppl 1):5459-5460. 1993.
Cherrie JW; Dodgson J; Groat S; Carson M. Comparison of Optical and Electron Microscopy
for Evaluating Airborne Asbestos. Institute of Occupational Medicine, Edinburgh, (U.S.
Department of Commerce). 1979.
Cherrie JW; Dodgson J; Groat S; Carson M. Comparison of Optical and Electron Microscopy
for Evaluating Airborne Asbestos. Institute of Occupational Medicine, Edinburgh, U.S.
Department of Commerce. 1987.
Chesson J; Rench JD; Schultz BD; Milne KL. Interpretation of Airborne Asbestos
Measurements. RiskAnalysis. 10(3):437-47. 1990.
Choe N; Tanaka S; Xia; W; Hemenway DR; Roggli VI; Kagan E. Pleural Macrophage
Recruitment and Activation in Asbestos-Induced Pleural Injury. Environmental Health
Perspectives. 105(Suppl 5):1257-1260. September. 1997.
Choe N; Tanaka S; Kagan E. Asbestos Fibers and Interleukin-1 Upregulate the Formation of
Reactive Nitrogen Species in Rat Pleural Mesothelial Cells. American Journal of Respiratory
Cell and Molecular Biology. 19(2):226-236. 1998.
Choe N; Zhang J; Iwagaki A; Tanaka S; Hemenway DR; Kagan E. Asbestos Exposure
Upregulates the Adhesion of Pleural Leukocites to Pleural Mesothelial Cells via VCAM-1.
American Journal of Physiology. 277(2:Part 1):L292-L300. 1999.
Churg A. Deposition and clearance of chrysotile asbestos. Annals of Occupational Hygiene.
38(4):625-633. 1994.
Churg A; Wright JL; Stevens B. Exogenous mineral particles in the human bronchial mucosa
and lung parenchyma. I. Nonsmokers in the general population. Experimental Lung Research.
16:159-175. 1990.
9.5
-------
Churg A; Wiggs B; Depaoli L; Kampe B; Stevens B. Lung Asbestos Content in Chrysotile
Workers with Mesothelioma. American Review of Respiratory Disease. 130:1042-1045. 1984.
Coin PG; Roggli VL; Brody AR. Deposition, Clearance, and Translocation of Chrysotile
Asbestos from Peripheral and Central Regions of the Rat Lung. Environmental Research.
58:970-116. 1992.
Coin PG; Roggli VL; Brody AR. Persistence of Long, Thin Chrysotile Asbestos Fibers in the
Lungs of Rats. Environmental Health Perspectives. 102(Suppl 5): 197-199. 1994.
Costa DL; Dreher KL. Bioavailable Transition Metals in Paniculate Matter Mediate
Cardiopulmonary Injury in Healthy and Compromised Animal Models. Environmental Health
Perspectives. 105(Suppl 5):1053-1060. September. 1997.
Cox D; Lindley D. Theoretical Statistics. Chapman and Hall, London. 1974.
Cox D; Oakes DV. Analysis of Survival Data. Cox DR; Hinkley DV (eds.). Chapman and Hall,
London. 1984.
Crump KS. The Effect of Random Error in Exposure Measurement upon the Shape of the Dose
Response. Nonlinearity in Biology, Toxicology and Medicine. 2003. (pending publications)
Crump KS. Asbestos Potency Assessment for EPA Hearing. Prepared for Asbestos Information
Association/North America. 116pp. 1986.
Davis JMG; Cowie HA. The Relationship Between Fibrosis and Cancer in Experimental
Animals Exposed to Asbestos and Other Fibers. Environmental Health Perspectives.
88:305-309. 1990.
Davis JMG; Jones AD. Comparisons of the Pathogenicity of Long and Short Fibres of
Chrysotile Asbestos in Rats. British Journal of Exposure Pathology. 69(5):171-738. 1988b.
Davis JMG; Beckett ST; Bolton RE; Collings P; Middleton AP. Mass and Number of Fibres in
the Pathogenesis of Asbestos-Related Lung Disease in Rats. British Journal of Cancer.
37:673-688. 1978.
Davis JMG; Beckett ST; Bolton RE; Donaldson K. A Comparison of the Pathological Effects in
Rats of the UICC Reference Samples of Amosite and Chrysotile with Those of Amosite and
Chrysotile Collected from the Factory Environment. In Biological Effects of Mineral Fibres.
Wagner JC (ed.). IARC Scientific Publications, pp. 288-292. 1980.
Davis JMG; Addison J; Bolton RE; Donaldson K; Jones AD; Miller BG. Inhalation Studies on
the Effects of Tremolite and Brucite Dust in Rats. Carcinogenesis. 6(5):667-674. 1985.
Davis JMG; Addison J; Bolton R; Donaldson K; Jones AD; Smith T. The Pathogenicity of Long
Versus Short Fibre Samples of Amosite Asbestos Administered to Rats by Inhalation and
Intraperitoneal Injection. British Journal of Experimental Pathology. 67:415-430. 1986a.
9.6
-------
Davis JMG; Gylseth G; Morgan A. Assessment of Mineral Fibres from Human Lung Tissue.
Thorax. 41:167-175. 1986b.
Davis JMG; Bolton RE; Brown D; Tully HE. Experimental Lesions in Rats Corresponding to
Advanced Human Asbestosis. Exposure Molecular Pathology. 44(2):207-221. 1986c.
Davis JMG; Addison J; Bolton RE; Donaldson K; Jones AD. Inhalation and Injection Studies in
Rats Using Dust Samples from Chrysotile Asbestos Prepared by a Wet Dispersion Process.
British Journal of Pathology. 67:113-129. 1986d.
Davis JMG; Jones AD; Smith T. Comparisons of the Pathogenicity of Long and Short Fibres of
Chrysotile Asbestos in Rats. Institute for Research and Development of Asbestos, Montreal
(ed.). Institute of Occupational Medicine. Report No. TM-87/08. 1987.
Davis JMG; Bolton RE; Douglas AN; Jones AD; Smith T. Effects of Electrostatic Charge on the
Pathogenicity of Chrysotile Asbestos. British Journal of Industrial Medicine. 45(5):292-309.
1988a.
Davis JMG; Addison J; Mclntosh C; Miller BG; Niven K. Variations in the Carcinogenicity of
Tremolite Dust Samples of Differing Morphology. Annals New York Academy of Sciences.
473-490. 1991.
de Klerk N. Unpublished Raw Data Provided to Dr. Wayne Berman by Dr. Nick de Klerk from
Study of Crocidolite Miners in Wittenoom, Australia, originally described in Armstrong et al.
(1988) but with followup extended through 1999. 2001.
de Klerk NH; Musk AW; Armstrong BK; Hobbs MST. Diseases in Miners and Millers of
Crocidolite from Wittenoom, Western Australia: A Further Followup to December 1986. Annals
of Occupational Hygiene. 38(Suppl l):647-655. 1994.
Dement JM. Estimation of Dose and Evaluation of Dose-Response in a Retrospective Cohort
Mortality Study of Chrysotile Asbestos Textile Workers. Ph.D. Thesis. The University of North
Carolina at Chapel Hill. 1980.
Dement JM; Brown DP. Lung Cancer Mortality Among Asbestos Textile Workers: A Review
and Update. Annals of Occupational Hygiene. 38(4):525-532. 1994.
Dement JM; Brown DP. Cohort Mortality and Case-Control Studies of White Male Chrysotile
Asbestos Textile Workers. Journal of Clean Technology, Environmental Toxicology, and
Occupational Medicine. 7:1052-1062. 1998.
Dement JM; Harris RL. Estimates of Pulmonary and Gastrointestinal Deposition for
Occupational Fiber Exposure. NTIS PB80-149644. U.S. HEW Contract #78-2438. 1979.
Dement JM; Harris RL; Symons MJ; Shy CM. Estimates of Dose-Response for Respiratory
Cancer Among Chrysotile Asbestos Textile Workers. Annals Occupational Hygiene.
26(l-4):869-887. 1982.
9.7
-------
Dement JM; Harris RL; Symons MJ; Shy CM. Exposures and Mortality Among Chrysotile
Workers. Part I: Exposure Estimates. American Journal of Industrial Medicine. 4:399-419.
1983a.
Dement JM; Harris RL; Symons MJ; Shy CM. Exposures and Mortality Among Chrysotile
Workers. Part II: Mortality. American Journal of Industrial Medicine. 4:421-433. 1983b.
Dement JM; Brown DP; Okun A. Follow-up Study of Chrysotile Asbestos Textile Workers:
Cohort Mortality and Case-Control Analysis. American Journal of Industrial Medicine.
26:431-447. 1994.
Doll R; Peto R. Cigarette Smoking and Bronchial Carcinoma: Dose and Time Relationships
Among Regular Smokers and Lifelong Non-Smokers. Journal of Epidemiology and Community
Health. 32:303-313. 1978.
Doll R; Peto J. Asbestos: Effects on Health of Exposure to Asbestos. Health and Safety
Commission, London, United Kingdom. 1985.
Dopp E; Schiffmann D. Analysis of Chromosomal Alterations Induced by Asbestos and
Ceramic Fibers. Toxicology Letters. 96-97:155-162. 1998.
Driscoll KE; Carter JM; Hassenbein DG; Howard B. Cytokines and Particle-Induced
Inflammatory Cell Recruitment. Environmental Health Perspectives. 105(Suppl 5): 1159-1164.
September. 1997.
Eastes W; Hadley JG. A Mathematical Model of Fiber Carcinogenicity and Fibrous in
Inhalation and Intraperitoneal Experiments in Rats. Inhalation Toxicology. 8:323-343. 1994.
Eastes W; Hadley JG. Dissolution of Fibers Inhaled by Rats. Inhalation Toxicology.
7:179-196. 1995.
Eastes W; Hadley JG. A Mathematical Model of Fiber Carcinogenicity and Fibrous in
Inhalation and Intraperitoneal Experiments in Rats. Inhalation Toxicology. 8:323-343. 1996.
Economou P; Samet JM; Lechner JF. Familial and Genetic Factors in the Pathogenesis of Lung
Cancer. In: Epidemiology of Lung Cancer. Chapter 14. Samet JM (ed.). Marcel Dekker, Inc.,
New York. 1994.
Enterline PE; Harley J; Henderson V. Asbestos and Cancer ~ A Cohort Followed to Death.
Graduate School of Public Health, University of Pittsburgh. 1986.
Everitt JI; Gelzleichter TR; Bermudez E; Mangum JB; Wong BA; Janszen DB; Moss OR.
Comparison of Pleural Responses of Rats and Hamsters to Subchronic Inhalation of Refractory
Ceramic Fibers. Environmental Health Perspectives. 105(Suppl 5):1209-1213. September.
1997.
9.8
-------
Finkelstein MM. Mortality Among Long-Term Employees of an Ontario Asbestos-Cement
Factory. British Journal of Industrial Medicine. 40:138-144. 1983.
Finkelstein MM. Mortality Among Employees of an Ontario Asbestos-Cement Factory.
American Review of Respiratory Disease. 129:754-761. 1984.
Finkelstein MM; Dufresne A. Inferences on the Kinetics of Asbestos Deposition and Clearance
Among Chrysotile Miners and Millers. American Journal of Industrial Medicine.
35(4):401-412. April 1999.
Finkelstein JN; Johnston C; Barrett T; Oberdorster G. Particulate-Cell Interactions and
Pulmonary Cytokine Expression. Environmental Health Perspectives. 105(Suppl
5):1179-1182. September. 1997.
Floyd RA. Role of Oxygen Free Radicals in Carcinogenesis and Brain Ischemia. The FASEB
Journal. 4:2587-2597. June. 1990.
Fubini B. Surface Reactivity in the Pathogenic Response to Particulates. Environmental Health
Perspectives. 105(Suppl 5): 1013-1020. September. 1997.
Gehr P; Geiser M; Stone KC; Crapo JD. Morphometric Analysis of the Gas Exchange Region of
the Lung. In Toxicology of the Lung, 2nd edition. Gardner DE; Crapo JD; McClellan RO (eds.).
Raven Press, New York. 1993.
Ghio AJ; Kadiiska MB; Xiang Q-H; Mason RP. In Vivo Evidence of Free Radical Formation
After Asbestos Instillation: An ESR Spin Trapping Investigation. Free Radical Biology and
Medicine. 24(1):11-17. 1998.
Gibbs GW; Hwang CY. Physical Parameters of Airborne Asbestos Fibres in Various Work
Environments - Preliminary Findings. American Industrial Hygiene Association Journal.
36(6):459-466. 1975.
Gibbs GW; Hwang CY. Dimensions of Airborne Asbestos Fibers. In Biological Effects of
Mineral Fibers. Wagner JC (ed.). IARC Scientific Publication, pp. 69-78. 1980.
Gilbert O. Statistical Method for Environmental Pollution Monitoring. Van Nostrand Reinhold,
New York. 1987.
Gold J; Amandusson H; Krozer A; Kasemo B; Ericsson T; Zanetti G; Fubini B. Chemical
Characterization and Reactivity of Iron Chelator-Treated Amphibole Asbestos. Environmental
Health Perspectives. 105(Supple 5):1021-1030. 1997.
Goldstein B; Rendall R; Webster I. A Comparison of the Effects of Exposure of Baboons to
Crocidolite and Fibrous-Glass Dusts. Environmental Research. 32:334-359. 1983.
9.9
-------
Golladay SA; Park S-H; Aust AE. Efflux of Reduced Glutathione after Exposure of Human
Lung epithelial Cells to Crocidolite Asbestos. Environmental Health Perspectives. 105(Suppl
5):1273-1278. 1997.
Goodglick LA; Kane AB. Cytotoxicity of Long and Short Crocidolite Asbestos Fibers In Vitro
and lin Vivo. Cancer Research. 50:5153-5163. 1990.
Governa M; Camilucci L; Amati M; Visona I; Valentino M; Botta GC; Campopiano A; Fanizza
C. Wollastonite Fibers In Vitro Generate Reactive Oxygen Species Able to Lyse Erythrocytes
and Activate the Complement Alternate Pathway. Toxicological Sciences. 44(l):32-38. 1998.
Gross TJ; Cobb SM; Peterson MW. Asbestos Exposure Increases Paracellular Transport of
Fibrin Degradation Products Across Human Airway Epithelium. American Journal of
Physiology. 266(3):L287-295. March. 1994.
Hammond EC; Selikoff IJ; Seidman H. Asbestos Exposure, Cigarette Smoking and Death Rates.
Annals New York Academy of Sciences. 330:473^90. 1979.
Harris RL; Timbrell V. The Influence of Fibre Shape in Lung Deposition - Mathematical
Estimates. Inhaled Particles IV. Walton WH (ed.). Pergamon Press. 1977.
Hart GA; Kathman LM; Hesterberg TW. In Vitro Cytotoxicity of Asbestos and Man-Made
Vitreous Fibers: Roles of Fiber Length, Diameter and Composition. Carcinogenesis.
15(5):971-977. May. 1994.
Health Effects Institute - Asbestos Research (HEI-AR). Asbestos in Public and Commercial
Buildings: A Literature Review and Synthesis of Current Knowledge. HEI-AR, 141 Portland
St., Suite 7100, Cambridge, MA. 1991.
Health Effects Institute (HEI). Asbestos in Public and Commercial Buildings: A Literature
Review and Synthesis of Current Knowledge. 1991.
Hei TK; Wu LJ; Piao CQ. Malignant Transformation of Immortalized Human Bronchial
Epithelial Cells by Asbestos Fibers. Environmental Health Perspectives. 105(Suppl
5): 1085-1088. September. 1997.
Heidenreich WF; Luebeck EG; Moolgavkar SH. Some Properties of the Hazard Function of the
Two-Mutation Clonal Expansion Model. Risk Analysis. 17:391-399. 1997.
Henderson VL; Enterline PE. Asbestos Exposure: Factors Associated with Excess Cancer and
Respiratory Disease Mortality. Annals New York Academy of Sciences. 330:117-126. 1979.
Hesterberg TW; Miller WC; McConnell EE; Chevalier J; Hadley JG; Bernstein DM; Thevenaz
P; Anderson R. Chronic Inhalation Toxicity of Size-Separated Glass Fibers in Fischer 344 Rats.
Fundamental and Applied Toxicology. 20:464-476. 1993.
9.10
-------
Hesterberg TW; Miiller WC; Thevenaz P; Anderson R. Chronic Inhalation Studies of Man-
Made Vitreous Fibres: Characterization of Fibres in the Exposure Aerosol and Lungs. Annals of
Occupational Hygiene. 39(5):637-653. 1995.
Hesterberg TW; Miiller WC; Musselkman RP; Kamstrup O; Hamilton RD; Thevenaz P.
Biopersistence of Man-Made Vitreous Fibers and Crocidolite Asbestos in Rat Lung Following
Inhalation. Fundamentals and Applied Toxicology. 29:267-279. 1996.
Hesterberg TW; Axten C; McConnell EE; Oberdorster G; Everitt J; Miller WC; Chevalier J;
Chase GR; Thevenaz P. Chronic Inhalation Study of Fiber Glass and Amosite Asbestos in
Hamsters: Twelve-Month Preliminary Results. Environmental Health Perspectives. 105(Suppl
5): 1223-1230. September. 1997.
Hesterberg TW; Chase G; Axten C; Miller WC; Musselman RP; Kamstrup O; Hadley J;
Morscheidt C; Bernstein DM; Thevenaz P. Biopersistence of Synthetic Vitreous Fibers and
Amosite Asbestos in the Rat Lung Following Inhalation. Toxicology and Applied
Pharmacology. 151:262-275. 1998a.
Hesterberg TW; Hart GA; Chevalier J; Miller WC; Hamilton RD; Bauer J; Thevenaz P. The
Importance of Fiber Biopersistence and Lung Dose in Determining the Chronic Inhalation
Effects of X607, RCF1, and Chrysolite Asbestos in Rats. Toxicology and Applied
Pharmacology. 153(l):68-82. 1998b.
Hodgson AA. Fibrous Silicates. Lecture Series No. 4. The Royal Institute of Chemistry,
London, United Kingdom. 1965.
Hodgson J; Darnton A. The Quantitative Risk of Mesothelioma and Lung Cancer in Relation to
Asbestos Exposure. Annals of Occupational Hygiene. 44(8): 5 65-601. 2000.
Holian A; Uthman MO; Goltsova T; Brown SD; Hamilton RF. Asbestos and Silica-Induced
Changes in Human Alveolar Macrophage Phenotype. Environmental Health Perspectives.
105(Suppl 5): 1139-1142. September. 1997.
Hughes JM; Weill H. Asbestos Exposure: Quantitative Assessment of Risk. American Review
of Respiratory Disease. 133:5-13. 1986.
Hughes JM; Weill H; Hammad YY. Mortality of Workers Employed at Two Asbestos Cement
Plants. British Journal of Industrial Medicine. 44:161-174. 1987.
Hume LA; Rimstidt. The Biodurability of Chrysotile Asbestos. American Mineralogist.
77:1125-1128. 1992.
Hwang CY; Gibbs GW. The Dimensions of Airborne Asbestos Fibres —I. Crocidolite from
Kuruman Area, Cape Province, South Africa. Annals Occupational Hygiene. 24(1):23-41.
1981.
9.11
-------
Ilgren E; Chatfield E. Coalinga Fibre: A Short, Amphibole-Free Chrysotile. Part 3: Lack of
Biopersistence. Indoor Built Environment. 7:98-100. 1998.
International Agency for Research on Cancer (IARC). Monographs on the Evaluation of
Carcinogenic Risks to Man. Volume 14. IARC Scientific Publications. Lyon, France. 1977.
Integrated Risk Information System (IRIS). Toxicological Review of Asbestos. U.S.
Environmental Protection Agency. Office of Research and Development, National Center for
Environmental Assessment. Washington, D.C. http://www.epa.gov/iris/subst/0371.htm. 1998.
Ishizaki T; Yano E; Evans PH. Cellular Mechanisms of Reactive Oxygen Metabolite Generation
from Human Polymorphonuclear Leukocytes Induced by Crocidolite Asbestos. Environmental
Research. 75(2): 135-140. 1997.
International Organization for Standardization (ISO). Ambient Air-Determination of Asbestos
Fibres - Direct-Transfer Transmission Electron Microscopy Method. ISO 10312. 1995.
Jagirdar J; Lee TC; Reibman J; Gold LI; Aston C; Begin R; Rom WN. Immunohistochemical
Localization of Transforming Growth Factor Beta Isoforms in Asbestos-Related Diseases.
Environmental Health Perspectives. 105(Suppl 5): 1197-1203. September. 1997.
Jaurand MC. Observations on the Carcinogenicity of Asbestos Fibers. Annals New York
Academy of Science. 643:258-70. 1991.
Jaurand MC. Mechanisms of Fiber-Induced Genotoxicity. Environmental Health Perspectives.
105(Suppl 5): 1073-1084. September. 1997.
Jesch NK; Dorger M; Enders G; Rieder G; Vogelmeier C; Messmer K; Krombach F. Expression
of the Inducible Nitric Oxide Synthase and Formation of Nitric Oxide by Alveolar Macrophages.
Environmental Health Perspectives. 105(Suppl 5):1297-1300. September. 1997.
Johnson NF. Asbestos-Induced Changes in Rat Lung Parenchyma. Journal of Toxicology and
Environmental Health. 21:193-203. 1987.
Johnson NF; Jaramillo RJ. P53, Cip 1, and Gadd 153 Expression Following Treatment of A549
Cells with Natural and Man-Made Vitreous Fibers. Environmental Health Perspectives.
105(Suppl5):l 143-1145. September. 1997.
Jones AD; McMillan CH; Johnston AM; Mclntosh C; Cowie H; Bolton RE; Borzuki G; Vincent
JH. Pulmonary Clearance of UICC Amosite Fibres Inhaled by Rats During Chronic Exposure at
Low Concentrations. British Journal of Industrial Medicine. 45:300-304. 1988.
Kaiglova A; Hurbankova M; Kovacikova Z. Impact of Acute and Subchronic Asbestos
Exposure on Some Parameters of Antioxidant Defense System and Lung Tissue Injury.
Industrial Health. 37(3):348-351. July. 1999.
9.12
-------
Kamp DW; Graceffa P; Pryor WA; Weitzman SA. The Role of Free Radicals in Asbestos-
Induced Diseases. Free Radical Biology & Medicine. 12:293-315. 1992.
Kamp DW; Dunne M; Dykewicz MS; Sbalchiero JS; Weitzman SA; Dunn MM. Asbestos-
Induced Injury to Cultured Human Pulmonary Epithelial-Like Cells: Role of Neutrophil
Elastase. Journal of Leukocyte Biology. 54:73-80. July. 1993.
Kamp DW; Greenberger MJ; Sbalchierro JS; Preusen SE; Weitzman SA. Cigarette Smoke
Augments Asbestos-Induced Alveolar Epithelial Cell Injury: Role of Free Radicals. Free
Radical Biology & Medicine. 25(6):728-739. 1998.
Kane AB; MacDonald JL. Mechanisms of Mesothelial Cell Injury, Proliferation, and Neoplasia
Induced by Asbestos Fibers. Chapter 14. Fiber Toxicology, pp. 323-347. 1993.
Kauffer E; Vigneron JC; Hesbot A; Lemonnier M. A Study of the Length and Diameter of
Fibres, in Lung and in Broncho-Alveolar Lavage Fluid, Following Exposure of Rats to
Chrysotile Asbestos. Annals of Occupational Hygiene. 31(2):233-240. 1987.
Keane MJ; Miller WE; Ong T; Stephens JW; Wallace WE; Zhong B-Z. A Study of the Effect of
Chrysotile Fiber Surface Composition on Genotoxicity In Vitro. Journal of Toxicology and
Environmental Health Part A. 57(8):529-541. August. 1999.
Kimizuka G; Wang N; Hayashi Y. Physical and Microchemical Alterations of Chrysotile and
Amosite Asbestos in the Hamster Lung. Journal of Toxicology and Environmental Health.
21:251-264. 1987.
Kodama Y; Boreiko CJ; Maness SC; Hesterberg TW. Cytotoxic and Cytogenetic Effects of
Asbestos on Human Bronchial Epithelial Cells in Culture. Carcinogenesis. 14(4):691-697.
1993.
Kostyuk VA; Potapovich AI. Antiradical and Chelating Effects in Flavonoid Protection Against
Silica-Induced Cell Injury. Archives of Biochemistry and Biophysics. 355(l):43-48. 1998.
Kravchenko IV; Furalyov VA; Vasylieva LA; Pylev LN. Spontaneous and Asbestos-Induced
Transformation of Mesothelial Cells In Vitro. Teratogenesis Carcinogenesis andMutagenesis.
18(3):141-151. 1998.
Krombach F; Munzing S; Allmeling AM; Gerlach JT; Behr J; Dorger M. Cell Size of Alveolar
Macrophages: An Interspecies Comparison. Environmental Health Perspectives. 105(Suppl
5): 1261-1263. September. 1997.
Lacquet LM; VanderLinden L; Lepoutre J. Roentgenographic Lung Changes, Asbestosis and
Mortality in a Belgian Asbestos-Cement Factory. In Biological Effects of Mineral Fibres,
Wagner JC (ed.). IARC Sci Publ. pp. 783-793. 1980.
Law BD; Bunn WB; Hesterberg TW. Solubility of Polymeric Organic Fibers and Manmade
Vitreous Fibers in Gambles Solution. Inhalation Toxicology. 2:321-339. 1990.
9.13
-------
Law BD; Bunn WB; Hesterberg TW. Dissolution of Natural Mineral and Man-Made Vitreous
Fibers in Karnovsky's and Formalin Fixatives. Inhalation Toxicology. 3:309-321. 1991.
Leanderson P; Soderkvist P; Tagesson C; Axelson O. Formation of 8-hydroxydeoxyguanosine
by Asbestos and Man-Made Mineral Fibres. British Journal of Industrial Medicine.
45:309-311. 1988.
Le Bouffant L. Physics and Chemistry of Asbestos Dust. In Biological Effects of Mineral
Fibres. Wagner JC (ed.). IARC Scientific Publications, pp. 15-34. 1980.
Le Bouffant L; Daniel H; Heninn JP; Martin JC; Normand C; Tichoux G; Trolard F.
Experimental Study on Long-Term Effects of Inhaled MMF on the Lungs of Rats. Annals of
Occupational Hygiene. 31(4B):765-790. 1987.
Lechner JF; Tesfaigzi J; Gerwin BI. Oncogenes and Tumor-Suppressor Genes in Mesothelioma
-A Synopsis. Environmental Health Perspectives. 105(Suppl 5):1061-1067. September.
1997.
Lee KP; Barras CE; Griffith RS; Waritz RS; Lapin CA. Comparative Pulmonary Responses to
Inhaled Inorganic Fibers with Asbestos and Fiberglass. Environmental Research. 24:167-191.
1981.
Leigh J; Wang H; Benin A; Peters M; Ruan X. Silica-Induced Apoptosis in Alveolar and
Granulomatous Cells In Vivo. Environmental Health Perspectives. 105(Suppl 5):1241-1246.
September. 1997.
Leikoff G; Driscoll K. Cellular Approaches in Respiratory Toxicology. In Toxicology of the
Lung, 2nd edition. Gardner DE; Crapo JD; McClellan RO (eds.). Raven Press, New York. 1993.
Levin JL; McLarty JW; Hurst GA; Smith AN; Frank AL. Tyler Asbestos Workers: Mortality
Experience in a Cohort Exposed to Amosite. Occupational and Environmental Medicine.
55:155-160. 1998.
Levresse V; Reiner A; Fleury-Feith J; Levy F; Moritz S; Vivo C; Pilatte Y; Jaurand M-C.
Analysis of Cell Cycle Disruptions in Cultures of Rat Pleural Mesothelial Cells Exposed to
Asbestos Fibers. American Journal of Respiratory Cell and Molecular Biology. \ 7(6):660-671.
1997.
Li XY; Gilmour PS; Donaldson K; MacNee W. In Vivo and In Vitro Proinflammatory Effects of
Particulate Air Pollution (PM10). Environmental Health Perspectives. 105(Suppl
5): 1279-1284. September. 1997.
Liddell FDK. The Interaction of Asbestos and Smoking in Lung Cancer. Annals of
Occupational Hygiene. 45:341-56. 200la.
9.14
-------
Liddell FDK. Unpublished raw mesothelioma data provided to Dr. Wayne Berman by Dr. FDK
Liddell from multiple studies of the 1891-1920 Birth Cohort of Quebec Chrysotile Miners and
Millers most recently described in Liddell et al. 1997. 200Ib.
Liddell FDK; Armstrong BG. The Combination of Effects on Lung Cancer of Cigarette
Smoking and Exposure in Quebec Chrysotile Miners and Millers. Annals of Occupational
Hygiene. 46(1):5-13. 2002.
Liddell FDK; McDonald AD; McDonald JC. The 1891-1920 Birth Cohort of Quebec Chrysotile
Miners and Millers: Development From 1904 and Mortality to 1992. Annals of Occupational
Hygiene. 41:13-36. 1997.
Lim Y; Kim S-H; Kim K-A; Oh M-W; Lee K-H. Involvement of Protein Kinase C,
Phospholipase C, and Protein Tyrosine Kinase Pathways in Oxygen Radical Generation by
Asbestos-Stimulated Alveolar Macrophage Environmental Health Perspectives. 105(Suppl
5): 1325-1328. September. 1997.
Lippmann M. Deposition and Retention of Inhaled Fibres: Effects on Incidence of Lung Cancer
and Mesothelioma. Occupational and Environmental Medicine. 51:793-798. 1994.
Lippmann M. Asbestos and Other Mineral and Vitreous Fibers. In Environmental Toxicants:
Human Exposures and Their Health Effects. Lippman M (ed). Wiley-Interscience; 2nd edition.
December. 1999.
Lippmann M; Schlesinger RB. Interspecies Comparisons of Particle Deposition and Mucociliary
Clearance in Tracheobronchial Airways. Journal of Toxicology and Environmental Health.
3:441 [261]-469[289]. 1984.
Luster MI; Simeonova PP. Asbestos Induces Inflammatory Cytokines in the Lung Through
Redox Sensitive Transcription Factors. Toxicology Letters (Shannon). 102-103:271-275. 1998.
Lynch JR; Ayer HE; Johnson DJ. The Interrelationships of Selected Asbestos Exposure Indices.
American Industrial Hygiene Association Journal. 31(5):598-604. 1970.
Marconi A; Menichini E; Paoletti L. A Comparison of Light Microscopy and Transmission
Electron Microscopy Results in the Evaluation of the Occupational Exposure to Airborne
Chrysotile Fibres. Annals of Occupational Hygiene. 28(3):321-331. 1984.
Marsella JM; Liu BL; Vaslet CA; Kane AB. Susceptibility of p53-Deficient Mice to Induction
of Mesothelioma by Crocidolite Asbestos Fibers. Environmental Health Perspectives.
105(Suppl5):1069-1072. September. 1997.
Martin LD; Krunkosky TM; Dye JA; Fischer BM; Jiang NF; Rochelle LG; Akley NJ; Dreher
KL; Adler KB. The Role of Reactive Oxygen and Nitrogen Species in the Response of Airway
Epithelium to Particulates. Environmental Health Perspectives. 105(Suppl 5): 11301-1308.
September. 1997.
9.15
-------
Mattson SM. Glass Fiber Dissolution in Simulated Lung Fluid and Measures Needed to
Improve Consistency and Correspondence in In-Vitro Studies. Presented at the IARC Conf.
Biopersistence of Respirable Synthetic Fibres and Minerals. Lyon, France, September 7-9.
1992. Environmental Health Perspectives. 1994.
McConnell E; Wagner J; Skidmore J; Moore J. A Comparative Study of the Fibrogenic and
Carcinogenic Effects of UICC Canadian Chrysotile B Asbestos and Glass Microfibre (JM 100).
Biological Effects of Man-Made Fibres. Proceedings of a WHO/IARC Conference, pp.
234-252. 1982.
McConnell EE; Rutter FLA; Ulland BM; Moore JA. Chronic Effects of Dietary Exposure to
Amosite Asbestos and Tremolite in F344 Rats. Environmental Health Perspectives. 53:27-44.
1983.
McConnell EE; Adkins B. Studies on the Chronic Toxicity (Inhalation) of Wollastonite in
Fischer 344 Rats. Inhalation Toxicology. 3:323-337. 1991.
McConnell EE; Mast RW; Hesterberg TW; Chevalier J; Kotin P; Bernstein DM; Thevenez P;
Glass LR; Anderson R. Chronic Inhalation Toxicity of a Kaolin-Based Refractory Ceramic
Fiber in Syrian Golden Hamsters. Inhalation Toxicity: 7:503-532. 1995.
McDonald AD; Fry JS; Wooley AJ; McDonald JC. Dust Exposure and Mortality in an
American Chrysotile Textile Plant. British Journal of Industrial Medicine. 39:361-367. 1983a.
McDonald AD; Fry JS; Woolley AJ; McDonald JC. Dust Exposure and Mortality in an
American Factory Using Chrysotile, Amosite, and Crocidolite in Mainly Textile Manufacture.
British Journal of Industrial Medicine. 40:368-374. 1983b.
McDonald AD; Fry JS; Woolley AJ; McDonald JC. Dust Exposure and Mortality in an
American Chrysotile Asbestos Friction Products Plant. British Journal of Industrial Medicine.
41:151-157. 1984.
McDonald JC. Mineral Fibre Persistence and Carcinogenicity. Industrial Health.
36(4):372-375. October. 1998a.
McDonald JC. Invited Editorial: Unfinished Business - The Asbestos Textiles Mystery. Annals
of Occupational Hygiene. 42(l):3-5. 1998b.
McDonald JC; Gibbs GW; Liddell FDK. Chrysotile Fibre Concentration and Lung Cancer
Mortality: A Preliminary Report. In Biological Effects of Mineral Fibres. Wagner JC (ed).
IARC Scientific Publications, pp. 811-817. 1980a.
McDonald JC; Liddell FDK; Gibbs GW; Eyssen GE; McDonald AD. Dust Exposure and
Mortality in Chrysotile Mining, 1910-1975. British Journal of Industrial Medicine. 37:11-24.
1980b.
9.16
-------
McDonald JC; McDonald AD; Armstrong B; Sebastien P. Cohort Study of Mortality of
Vermiculite Miners Exposed to Tremolite. British Journal of Industrial Medicine. 43:436-444.
1986.
McDonald JC; Liddell FDK; Dufresne A; McDonald AD. The 1891-1920 Birth Cohort of
Quebec Chrysotile Miners and Millers: Mortality 1976-1988. British Journal of Industrial
Medicine. 50:1073-1081. 1993.
McGavran PD; Butterick CJ; Brody AR. Tritiated Thymidine Incorporation and the
Development of an Interstitial Lesion in the Bronchiolar-Alveolar Regions of the Lungs of
Normal and Complement Deficient Mice After Inhalation of Chrysotile Asbestos. Journal of
Environmental Pathology, Toxicology, and Oncology. 9(5-6):377-391. December. 1989.
McGavran PD; Moore LB; Brody AR. Inhalation of Chrysotile Asbestos Induces Rapid Cellular
Proliferation in Small Pulmonary Vessels of Mice and Rats. American Journal of Pathology.
136(3):695-705. 1990.
Mercer RR and Crapo JD. Three-Dimensional Analysis of Lung Structure and its Application to
Pulmonary Dosimetry Models. In Toxicology of the Lung, 2nd edition, Gardner DE; Crapo JD;
McClellan RO (eds.). Raven Press, New York. 1993.
Middleton AP; Beckett ST; Davis JMG. Further Observations on the Short-Term Retention and
Clearance of Asbestos by Rats, Using UICC Reference Samples. Annals of Occupational
Hygiene. 22:141-152. 1979.
Miller FJ; Overton JH; Kimbell JS; Russell ML. Regional Respiratory Tract Absorption of
Inhaled Reactive Gases. In Toxicology of the Lung, 2nd edition, Gardner DE; Crapo JD;
McClellan RO (eds.). Raven Press, New York. 1993.
Moolgavkar SH; Dewanji A; Venzon DJ. A Stochastic Two-Stage Model for Cancer Risk
Assessment. 1: The Hazard Function and the Probability of Tumor. Risk Analysis.
8(3):383-392. 1988.
Moolgavkar SH; Luebeck EG; De Gunst M; Port RE; Schwarz M. Quantitative Analysis of
Enzyme-Altered Foci in Rat Hepatocarcinogenesis Experiments. Single Agent Regimen.
Carcinogenesis. 11(8):1271-1278. 1990.
Moolgavkar SH; Luebeck EG; Krewski D; Zielinski JM. Radon, Cigarette Smoke, and Lung
Cancer: A Re-Analysis of the Colorado Plateau Uranium Miners' Data. Epidemiology.
4:204-217. 1993.
Morgan A; Holmes A. Concentrations and Dimensions of Coated and Uncoated Asbestos Fibres
in the Human Lung. British Journal of Industrial Medicine. 37:25—32. 1980.
Morgan A; Talbot RJ; Holmes A. Significance of Fibre Length in the Clearance of Asbestos
Fibres from the Lung. British Journal of Industrial Medicine. 35:146-153. 1978.
9.17
-------
Morgan A; Black A; Evans N; Holmes A; Pritchard JN. Deposition of Sized Glass Fibres in the
Respiratory Tract of the Rat. Annals of Occupational Hygiene. 23:353-366. 1980.
Morrow PE. Issues: Possible Mechanisms to Explain Dust Overloading of the Lungs.
Fundamental and Applies Toxicology. 10:369-384. 1988.
Mossman BT. Mechanisms of Asbestos Carcinogenesis and Toxicity: The Amphibole
Hypothesis Revisited. British Journal of Industrial Medicine. 50:673-676. 1993.
Mossman BT; Churg A. Mechanisms in the Pathogenesis of Asbestosis and Silicosis. American
Journal of Critical Care Medicine. 157:1666-1680. 1998.
Mossman BT; Marsh JP. Role of Active Oxygen Species in Asbestos-Induced Cytotoxicity, Cell
Proliferation, and Carcinogenesis. In Cellular and Molecular Aspects of Fiber Carcinogenesis,
Cold Spring Harbor Laboratory Press. 0-87969-361-4/91. pp. 159-168. 1991.
Mossman BT; Bignon J; Corn M; Seaton A; Gee JBL. Asbestos: Scientific Developments and
Implications for Public Policy. Science. 247:294-301. 1990.
Mossman BT; Kamp DW; Sigmund A; Weitzman SA. Mechanisms of Carcinogenesis and
Clinical Features of Asbestos-Associated Cancers. Cancer Investigation. 14(5):466-480. 1996.
Mossman BT; Faux S; Janssen Y; Jimenex LA; Timblin C; Zanella C; Goldberg J; Walsh E;
Barchowsky A; Driscoll K. Cell Signaling Pathways Elicited by Asbestos. Environmental
Health Perspectives. 105(Suppl 5): 1121-1125. September. 1997.
Muhle H; Bellman B; Takenata S; Ziem Y. Inhalation and Injection in Rats to Test the
Carcinogenicity of MMMF. Annals Occupational Hygiene. 31(4B):755-764. 1987.
Nario RC; Hubbard AK. Localization of Intercellular Adhesion Molecule-1 (ICAM-A) in the
Lungs of Silica-Exposed Mice. Environmental Health Perspectives. 105(Suppl 5):1183-1190.
September. 1997.
Nehls P; Seiler F; Rehn B; Greferath R; Bruch J. Formation and Persistence of 8-Oxoguanine
Rat Lung Cells as an Important Determinant for Tumor Formation following Particle Exposure.
Environmental Health Perspectives. 105(Suppl 5):1231-1240. September. 1997.
Nicholson WJ. Part III. Recent Approaches to the Control of Carcinogenic Exposures. Case
Study 1: Asbestos - The TLV Approach. AnnalsNew YorkAcademy of Science. 271:152-169.
1976.
Nicholson WJ; Selikoff IJ; Seidman H; Lilis R; Formby P. Long-Term Mortality Experience of
Chrysotile Miners and Millers in Thetford Mines, Quebec. Annals New YorkAcademy of
Sciences. 330:11-21. 1979.
9.18
-------
Nikula KJ; Avila KJ; Griffith WC; Mauderly JL. Sites of Particle Retention and Lung Tissue
Responses to Chronically Inhaled Diesel Exhaust and Coal Dust in Rats and Cynomolgus
Monkeys. Environmental Health Perspectives. 105(Suppl 5):1231-1240. September. 1997.
National Institute for Occupational Safety and Health (NIOSH). Method for Determination of
Asbestos in Air Using Positive Phase Contrast Microscopy. NIOSH Method 7400. NIOSH,
Cincinnati, Ohio, U.S.A. 1985.
National Institute for Occupational Safety and Health (NIOSH). Method for Determination of
Asbestos in Air Using Transmission Electron Microscopy. NIOSH Method 7402. NIOSH,
Cincinnati, Ohio, U.S.A. 1986.
Oberdorster G. Macrophage-Associated Responses to Chrysotile. Annals of Occupational
Hygiene. 38(4):601-615. 1994.
Oberdorster G; Morrow PE; Spurny K. Size Dependent Lymphatic Short Term Clearance of
Amosite Fibres in the Lung. Annals of Occupational Hygiene. 32(Suppl 1): 149-156. 1988.
Ollikainen T; Linnainmaa K; Kinnula VL. DNA Single Strand Breaks Induced by Asbestos
Fibers in Human Pleural Mesothelial Cells In Vitro. Environmental and Molecular Mutagenesis.
33(2):153-160. 1999.
Ontario Royal Commission. Report of the Royal Commission on Matters of Health and Safety
Arising form the Use of Asbestos in Ontario. Volume 3. 1984.
Osier M; Baggs RB; Oberdorster G. 1997. Intratracheal Instillation vs. Intratracheal Inhalation:
Influence of Cytokines on Inflammatory Response. Proceedings of the Sixth International
Meeting on the Toxicology of Natural and Man-Made Fibrous and Non-Fibrous Particles.
Environmental Health Perspective. 105(Suppl 5):1265-1271.
Osornio-Vargas AR; Kalter VG; Badgett A; Hernandez-Rodriguez N; Aguilar-Delfin I; Brody
AR. Rapid Communication: Early-Passage Rat Lung Fibroblasts do not Migrate In Vitro to
Transforming Growth Factor-Beta. American Journal of Respiratory Cell Molecular Biology.
8(5):468-471. May. 1993.
Palekar LD; Spooner CM; Coffin DL. Influence of Crystallization Habit of Minerals on In Vitro
Cytotoxicity. Annals New York Academy of Sciences, pp. 673-687. 1979.
Park S-H; Aust AE. Regulation of Nitric Oxide Synthase Induction by Iron and Glutathione in
Asbestos-Treated Human Lung Epithelial Cells. Archives of Biochemistry and Biophysics.
360(l):47-52. 1998.
Peto J. Lung Cancer Mortality in Relation to Measured Dust Levels in an Asbestos Textile
Factory. In Biological Effects of Mineral Fibres. Wagner JC (ed.). IARC Scientific
Publications, pp. 829-836. 1980a.
9.19
-------
Peto J. The Incidence of Pleural Mesothelioma in Chrysotile Asbestos Textile Workers. In
Biological Effects of Mineral Fibres. Wagner JC (ed.). IARC Scientific Publications, pp.
703-711. 1980b.
Peto J; Doll R; Howard SV; Kinlen LJ; Lewinsohn, HC. A Mortality Study Among Workers in
an English Asbestos Factory. British Journal of Industrial Medicine. 34:169-173. 1977.
Peto J; Seidman H; Selikoff IJ. Mesothelioma Mortality in Asbestos Workers: Implications for
Models of Carcinogenesis and Risk Assessment. British Journal of Cancer. 45:124-135. 1982.
Peto J; Doll R; Hermon C; Binns W; Clayton R; Goffe T. Relationship of Mortality to Measures
of Environmental Asbestos Pollution in an Asbestos Textile Factory. Annals of Occupational
Hygiene. 29(3):305-355. 1985.
Pinkerton KE; Brody AR; McLauren DA; Adkins B Jr; O'Connor RW; Pratt PC; Crapo JD.
Characterization of the Three Types of Chrysotile Asbestos After Aerosolization.
Environmental Research. 31:32-35. 1983.
Pinkerton KE; Pratt PC; Brody AR; Crapo JD. Fiber Localization and its Relationship to Lung
Reaction in Rats After Chronic Inhalation of Chrysotile Asbestos. American Journal of
Pathology. 117:484-498. 1984.
Pinkerton KE; Plopper CG; Mercer RR; Rogli VL; Patra AL; Brody AR; Crap JD. Airway
Branching Patterns Influence Asbestos Fiber Location and the Extent of Tissue Injury in the
Pulmonary Parenchyma. Laboratory Investigation. 55(6):688-695. 1986.
Piolatto G; Negri E; LaVecchia C; Pira E; Decarli A; Peto J. An Update of Cancer Mortality
Among Chrysotile Asbestos Miners in Balangero, Northern Italy. British Journal of Industrial
Medicine. 47:810-814. 1990.
Platek SF; Groth DH; Ulrich CE; Stettler LE; Finnell MS; Stoll M. Chronic Inhalation of Short
Asbestos Fibers. Fundamental and Applied Toxicology. 5:327-340. 1985.
Pooley FD. An Examination of the Fibrous Mineral Content of Asbestos Lung Tissue from the
Canadian Chrysotile Mining Industry. Environmental Research. 12:281-298. 1976.
Pooley FD. Tissue Burden Studies. In Short and Thin Mineral Fibres: Identification, Exposure,
and Health Effects. Chatfield EJ (ed). pp. 96-129. 1982.
Pott F. Some Aspects on the Dosimetry of the Carcinogenic Potency of Asbestos and Other
Fibrous Dusts. Staub-Reinhalt. 38(12):486-490. 1978.
Pott F. Animal Experiments with Mineral Fibers. In Short and Thin Mineral Fibers:
Identification, Exposure, and Health Effects. Chatfield EJ (ed.). pp. 133-161. 1982.
Pott F; Huth F; Friedrichs KH. Tumorigenic Effect of Fibrous Dust in Experimental Animals.
Environmental Health Perspectives. 9:313-315. 1974.
9.20
-------
Pott F; Huth F; Friedrichs KH. Results of Animal Carcinogenesis Studies After Application of
Fibrous Glass and Their Implications Regarding Human Exposure. Occupational Exposure to
Fibrous Glass. U.S. HEW Publication No. 76-151. pp. 183-191. 1976.
Pott F; Ziem U; Reiffer RJ; Huth F; Ernst H; Mohr U. Carcinogenicity Studies on Fibres, Metal
Compounds, and Some Other Dusts in Rats. Experimental Pathology. 32:129-152. 1987.
RaabeO. Deposition and Clearance of Inhaled Particles. In Occupational Lung Diseases. Gee
JB; et al. (eds.). Raven Press, pp.1047. 1984.
Quinlan TR; Berube KA; Hacker MP; Taatjes DJ; Timblin CR; Goldberg J; Kimberley P;
O'Shaughnessy P; Hemenway D; Torino J; Jimenez LA; Mossman BT. Mechanisms of
Asbestos-Induced Nitric Oxide Production by Rat Alveolar Macrophages in Inhalation and
In Vitro Models. Free Radical Biology and Medicine. 24(5):778-788. 1998.
Rahman Q; Mahmood N; Khan SG; Arif JM; Athar M. Mechanism of Asbestos-Mediated DNA
Damage: Role of Heme and Heme Proteins. Environmental Health Perspectives. 105(Suppl
5): 1109-1112. September. 1997.
Roberts DR; Zumwalde RD. Industrial Hygiene Summary Report of Asbestos Exposure
Assessment for Brake Mechanics. NIOSH Reports No. IWS-32-4A, Industrial Hygiene Section,
NTIS#PB 87-105433. 1982.
Robledo R; Mossman B. Cellular and Molecular Mechanisms of Asbestos-Induced Fibrosis.
Journal of Cellular Physiology. 180:158-166. 1999.
Roggli VL; Brody AR. Changes in Numbers and Dimensions of Chrysotile Asbestos Fibers in
Lungs of Rats Following Short-Term Exposure. Experimental Lung Research. 7:133-147.
1984.
Roggli VL; George MH; Brody AR. Clearance and Dimensional Changes of Crocidolite
Asbestos Fibers Isolated from Lungs of Rats Following Short-Term Exposure. Environmental
Research. 42:94-105. 1987.
Rood AP; Scott RM. Size Distributions of Chrysotile Asbestos in a Friction Products Factory As
Determined by Transmission Electron Microscopy. Annals of Occupational Hygiene.
33(4):583-590. 1989.
Rubino GF; Piolatto GW; Newhouse ML; Scansetti G; Aresini GA; Murray R. Mortality of
Chrysotile Asbestos Workers at the Balangero Mine, Northern Italy. British Journal of
Industrial Medicine. 36:187-194. 1979.
St. George JA; Harkema JR; Hyde DM; Plopper CG. Cell Populations and Structure/Function
Relationship of Cells in the Airways. In Toxicology of the Lung, 2nd edition. Gardner DE; Crapo
JD; McClellan RO (eds.). Raven Press, New York. 1993.
9.21
-------
Sanden A; Jarvholm B; Larsson S; Thiringer G. The Risk of Lung Cancer and Mesothelioma
After Cessation of Asbestos Exposure: A Prospective Cohort Study of Shipyard Workers.
European Respiratory Journal. 5:281-285. 1992.
Schnoor T. Unpublished Raw Data Provided to Dr. Wayne Berman by Ms. Terri Schnoor of
NIOSH from Study of South Carolina Textile Workers (Dement et al. 1994). 2001.
Sebastien P; Plourde M; Robb R; Ross M. Ambient Air Asbestos Survey in Quebec Mining
Towns - Part 1, Methodological Study. Environmental Protection Service, Quebec Region,
3/AP/RQ/1E. pp. 1-41. 1984.
Sebastien P; Plourde M; Robb R; Ross M; Nadon B; Wypruk T. Ambient Air Asbestos Survey
in Quebec Mining Towns. Part II: Main Study. Environmental Protection Service, Environment
Canada. EPS 5/AP/RQ/2E. July. 1986.
Sebastien P; McDonald JC; McDonald AD; Case B; Harley R. Respiratory Cancer in Chrysotile
Textile and Mining Industries: Exposure Inferences from Lung Analysis. British Journal of
Industrial Medicine. 46:180-187. 1989.
Seidman H. Short-Term Asbestos Work Exposure and Long-Term Observation ~ July 1984
Update. Department of Epidemiology, American Cancer Society. 1984.
Seidman H; Selikoff IJ; Hammond EC. Short-Term Asbestos Work Exposure and Long-Term
Observation. Annals New YorkAcademy of Sciences. 330:61-89. 1979.
Seidman H; Selikoff IJ; Gelb SK. Mortality Experience of Amosite Asbestos Factory Workers:
Dose-Response Relationships 5 to 40 Years After Onset of Short-Term Work Exposure.
American Journal of Industrial Medicine. 10(5/6):479-514. 1986.
Selikoff IJ; Seidman H. Asbestos-Associated Deaths among Insulation Workers in the United
States and Canada, 1967-1987. Annals of the New YorkAcademy of Sciences. 643:1-14. 1991.
Selikoff IJ; Hammond EC; Seidman H. Mortality Experience of Insulation Workers in the
United States and Canada 1943-1976. Annals New YorkAcademy of Sciences. 330:91-116.
1979.
Smith DM; Oritiz LW; Archuleta RF; Johnson NF. Long-Term Health Effects in Hamsters and
Rats Exposed Chronically to Man-Made Vitreous Fibres. Annals of Occupational Hygiene.
34(4B):731-754. 1987.
Snyder J; Virta R; Segreti J. Evaluation of the Phase Contrast Microscopy Method for the
Detection of Fibrous and Other Elongated Mineral Particulates by Comparison with a STEM
Technique. American Industrial Hygiene Association Journal. 48(5):471-477. 1987.
Stanton M; Wrench C. Mechanisms of Mesothelioma Induction with Asbestos and Fibrous
Glass. Journal of the National Cancer Institute. 48:797-821. 1972.
9.22
-------
Stanton M; Layard M; Tegeris A; Miller E; May M; Kent E. Carcinogenicity of Fibrous Glass:
Pleural Response in the Rat in Relation to Fiber Dimension. Journal of the National Cancer
Institute. 58(3):587-597. 1977.
Stanton M; Layard M; Tegeris A; Miller E; May M; Morgan E. Relation of Particle Dimension
to Carcinogenicity in Amphibole Asbestos and Other Fibrous Minerals. Journal of the National
Cancer Institute. 67(5):965-975. 1981.
Stayner L; Smith R; Bailer J; Gilbert S; Steenland K; Dement J; Brown D; Lemen R. Exposure-
Response Analysis of Risk of Respiratory Disease Associated with Occupational Exposure to
Chrysotile Asbestos. Occupational and Environmental Medicine. 54:646-652. 1997.
Stayner LT; Dankovic DA; Lemen RA. Occupational Exposure to Chrysotile Asbestos and
Cancer Risk: A Review of the Amphibole Hypothesis. American Journal of Public Health.
86(2): 176-186. February. 1996.
Stober W; McClellen RO; Morrow PE. Approaches to Modeling Disposition of Inhaled
Particles and Fibers in the Lung. In Toxicology of the Lung, 2nd edition. Gardner DE; Crapo JD;
McClellan RO (eds.). Raven Press, New York. 1993.
Strom KA; Yu CP. Mathematical Modeling of Silicon-Carbide Whisker Deposition in the Lung-
Comparison Between Rats and Humans. Aerosol Science and Technology. 24(3): 193-209.
1994.
Sussman RG; Cohen BS; Lippmann M. Asbestos Fiber Deposition in a Human
Tracheobronchial Cast. I. Experimental. Inhalation Toxicology. 3:145-160. 199la.
Sussman RG; Cohen BS; Lippmann M. Asbestos Fiber Deposition in a Human
Tracheobronchial Cast. II. Empirical Model. Inhalation Toxicology. 3:161-179. 1991b.
Takeuchi T; Nakajima M; Morimoto K. A Human Cell System for Detecting Asbestos
Cytogenotoxicity In Vitro. Mutation Research. 438(1):63-70. 1999.
Tanaka S; Choe N; Hemenway DR; Zhu S; Matalon S; Kagan E. Asbestos Inhalation Induces
Reactive Nitrogen Species and Nitrotyrosine Formation in the Lungs and Pleura of the Rat.
Journal of Clinical Investigation. 102(2):445-454. 1998.
Timblin CR; Guthrie GD; Janssen YWM; Walsh ES; Vacek P; Mossman BT. Patterns of C-Fos
and C-Jun Proto-Oncogene Expression, Apoptosis, and Proliferation in Rat Pleural Mesothelial
Cells Exposed to Erionite or Asbestos Fibers. Toxicology and Applied Pharmacology.
151(l):88-97. 1998a.
Timblin CR; Janssen YMW; Goldberg JL; Mossman BT. GRP78, HSP72/73, and CJUN Stress
Protein Levels in Lung Epithelial Cells Exposed to Asbestos, Cadmium or H2O2. Free Radical
Biology & Medicine. 24(4):632-642. 1998b.
9.23
-------
Timbrell V. Deposition and Retention of Fibres in the Human Lung. Annals of Occupational
Hygiene. 26(l-4):347-369. 1982.
Timbrell V; Hyett AW; Skidmore JW. A Simple Dispenser for Generating Dust Clouds From
Standard Reference Samples of Asbestos. Annals of Occupational Hygiene. \ 1:273-281. 1968.
Untried K; Kociok N; Roller M; Pott F; Dehnen W. P53 mutations in tumours induced by
intraperitoneal injection of crocidolite asbestos and benzo[a]pyrene in rats. Experimental
Toxicololgoy and Pathology. 49:181-187. 1997.
U.S. EPA (U.S. Environmental Protection Agency). Airborne Asbestos Health Assessment
Update. Report 600/8-84-003F, U.S. Environmental Protection Agency. 1986.
U.S. Environmental Protection Agency (EPA). Asbestos Hazard Emergency Response Act:
Asbestos-Containing Materials in Schools. Final Rule and Notice (Appendix A: AHERA
Method). Federal Register, 40 CFR 763, Vol. 52, No. 2, pp. 41826-41903. October. 1987.
Venzon D; Moolgavkar S. A Method for Computing Profile-likelihood-based Confidence
Intervals. Applied Statistics. 37:87-94. 1988.
Vincent JH. On the Practical Significance of Electrostatic Lung Deposition of Isometric and
Fibrous Aerosols. Journal Aerosol Science. 16(6):511-519. 1985.
Vincent JH; Johnston AM; Jones AD; Bolton RE; Addison J. Kinetics of Deposition and
Clearance of Inhaled Mineral Dusts During Chronic Exposure. British Journal of Industrial
Medicine. 42:707-715. 1985.
Wagner JC. Opening Discussion ~ Environmental and Occupational Exposure to Natural
Mineral Fibres. In Biological Effects of Mineral Fibres. Wagner JC (ed.). IARC Scientific
Publications, pp. 995-998. 1980.
Wagner JC; Berry G; Skidmore JW; Timbrell V. The Effects of the Inhalation of Asbestos in
Rats. British Journal of Cancer. 29:252-269. 1974.
Wagner JC; Berry G; Skidmore JW. Studies of the Carcinogenic Effects of Fiber Glass of
Different Diameters Following Intrapleural Innoculation in Experimental Animals. NIOSH
76-151. pp.193-197. 1976.
Wagner JC; Berry G; Hill R; Munday D; Skidmore J. Animal Experiments with MMM(V)F -
Effects of Inhalation and Intrapleural Inoculation in Rats. In Biological Effects of Man-Made
Fibres - Proceedings of a WHO/IARC Conference, Copenhagen, pp. 209-233. 1982.
Wagner JC; Skidmore JW; Hill RJ; Griffith DM. Erionite Exposure and Mesotheliomas in Rats.
British Journal of Cancer. 51:727-730. 1985.
Wagner JC; Griffiths DM; Munday DE. Experimental Studies with Palygorskite Dusts. British
Journal of Industrial Medicine. 44(ll):749-763. 1987.
9.24
-------
Walker AM. Declining Relative Risks for Lung Cancer After Cessation of Asbestos Exposure.
Journal of Occupational Medicine. 26(2):422-426. 1984.
Walton WH. The Nature, Hazards, and Assessment of Occupational Exposure to Airborne
Asbestos Dust: A Review. Annals of Occupational Hygiene. 25:117-247. 1982.
Warheit DB; Snajdr SI; Hartsky MA; Frame SR. Lung Proliferative and Clearance Responses to
Inhaled para-Aramid RFP in Exposed Hamsters and Rats: Comparisons with Chrysotiles
Asbestos Fibers. Environmental Health Perspectives. 105(Suppl 5):1219-1222. September.
1997.
Weill H. 1994. Cancer Mortality in Chrysotile Mining and Milling: Exposure-Response.
Asbestos-Cement. Annals of Occupational Hygiene. 38(4):412. 1994.
Weill H; Hughes J; Waggenspack C. Influence of Dose and Fibre Type on Respiratory
Malignancy Risk in Asbestos Cement Manufacturing. American Review of Respiratory Disease.
120:345-354. 1979.
WilksSS. Mathematical Statistics. 2nd edition. Wiley Publication, New York. 1963.
Winer AA; Cossette M. The Effect of Aspect Ration on Fiber Counts: A Preliminary Study.
Annals New York Academy of Sciences. 330:661-672. 1979.
Weitzman SA; Graceffa P. Communication: Asbestos Catalyzes Hydroxyl and Superoxide
Radical Generation from Hydrogen Peroxide. Archives of Biochemistry and Biophysics.
228(l):373-376. 1984.
World Health Organization (WHO). Reference Methods for Measuring Airborne Man-Made
Mineral Fibers (MMMF). 1985.
Wright GW; Kuschner M. The Influence of Varying Lengths of Glass and Asbestos Fibres on
Tissue Response in Guinea Pigs. In Inhaled Particles: Part 2. Walton WH; McGovern B (eds.).
Pergamon Press Oxford, pp. 455-474. 1975.
Wylie AG; Virta RL; Segretti JM. Characterization of Mineral Population by Index Particle:
Implications for the Stanton Hypothesis. Environmental Research. 43:427-439. 1987.
Wylie AG; Bailey KF; Kelse JW; Lee RJ. The Importance of Width in Asbestos Fiber
Carcinogenicity and its Implications for Public Policy. American Industrial Hygiene Association
Journal. 54:239-252. 1993.
Wylie AG; Skinner HCW; Marsh J; Snyder H; Garzione C; Hodkinson D; Winters R; Mossman
BT. Mineralogical Features Associated with Cytotoxic and Proliferative Effects of Fibrous Talc
and Asbestos on Rodent Tracheal Epithelial and Pleural Mesothelial Cells. Toxicology and
Applied Pharmacology. 147:143-150. 1997.
9.25
-------
Yamaguchi R; Hirano T; Ootsuyama Y; Asami S; Tsurudome Y; Fukada S; Yamato H; Tsuda T;
Tanaka I; Kasai H. Increased 8-Hydroxyguanine in DNA and Its Repair Activity in Hamster and
Rat Lung After Intratracheal Instillation of Crocidolite Asbestos. Japanese Journal of Cancer
Research. 90(5):505-509. 1999.
Yamate G; Agarwal SC; Gibbons RD. Methodology for the Measurement of Airborne Asbestos
by Electron Microscopy. U.S. EPA Report No. 68-02-3266. U.S. Environmental Protection
Agency, Washington, D.C., U.S.A. 1984.
Yeh H-C; Harkema JR. Gross Morphometry of Airways. In Toxicology of the Lung, 2nd edition.
Gardner DE; Crapo JD; McClellan RO (eds.). Raven Press, New York. 1993.
Yu CP; Asgharian B. A Kinetic Model of Alveolar Clearance of Amosite Asbestos Fibers from
the Rat Lung at High Lung Burdens. Journal of Aerosol Science. 21:21-27. 1990b.
Yu CP; Yoon KJ. Investigator's Report: Retention Modeling of Diesel Exhaust Particles in Rats
and Humans. 1991.
Yu CP; Asgharian B; Abraham JL. Mathematical Modeling of Alveolar Clearance of Chrysotile
Asbestos Fibers from the Rat Lungs. Journal of Aerosol Science. 21:587-594. 1990a.
Yu CP; Asgharian B; Pinkerton KE. Intrapulmonary Deposition and Retention Modeling of
Chrysotile Asbestos Fibers in Rats. Journal of Aerosol Science. 22(6):757-761. 1991.
Yu CP; Zhang L; Oberdoster G; Mast RW; Glass LR; Utell MJ. Deposition Modeling of
Refractory Ceramic Fibers in the Rat Lung. Journal of Aerosol Science. 25(2):407--417. 1994.
Yu CP; Zhang L; Oberdoster G; Mast RW; Maxim D; Utell MJ. Deposition of Refractory
Ceramic Fibers (RCF) in the Human Respiratory Tract and Comparison with Rodent Studies.
Aerosol Science and Technology. 23(3):291-300. 1995a.
Yu CP; Zhang L; Oberdoster G; Mast RW; Glass LR; Utell MJ. Clearance of Refractory
Ceramic Fibers (RCF) from the Rat Lung - Development of a Model. Environmental Research.
65(2):243-253. 1995b.
Zalma R; Bonneau L; Guignard J; Pezerat H; Jaurand M-C. Formation of Oxy Radicals by
Oxygen Reduction Arising from the Surface Activity of Asbestos. Canadian Journal of
Chemistry. 65:2338-2341. 1987.
Zanella CL; Timblin CR; Cummins A; Jung M; Goldberg J; Raabe R; Tritton TR; Mossman BT.
Asbestos-Induced Phosphorylation of Epidermal Growth Factor Receptor is Linked to c-fos and
apoptosis. American Journal of Physiology. 277(4:Part 1):L684-L693. 1999.
Zhang Y; Lee TC; Guillemin B; Yu M-C; Rom WN. Enhanced IL-lBeta and Tumor Necrosis
Factor-Alpha Release and Messenger RNA Expression in Macrophages from Idiopathic
Pulmonary Fibrosis or After Asbestos Exposure. Journal of Immunology. 150(9):4188-4196.
May. 1993.
9.26
-------
Zhu S; Manuel M; Tanaka S; Choe N; Kagan E; Matalon S. Contribution of Reactive Oxygen
and Nitrogen Species to Particulate-Induced Lung Injury. Environmental Health Perspectives.
106(5):1157-1163. 1998.
Zoitus BK; De Meringl A; Rouyer E; Thelohan S; Bauer J; Law B; Boymel PM; Olson JR;
Christensen VR; Guldberg M; Koenig AR; Perander M. In Vitro Measurement of Fiber
Dissolution Rate Relevant to Biopersistence at Neutral pH: An Interlaboratory Round Robin.
Inhalation Toxicology. 9:525-540. 1997.
9.27
-------
APPENDIX A:
UPDATE OF POTENCY FACTORS FOR
LUNG CANCER (KL) AND MESOTHELIOMA (KM)
Estimates of risk of dying of lung cancer or mesothelioma from asbestos exposure are quantified
by means of mathematical models that express risk as a function of exposure. The models
utilized in the 1986 U.S. EPA Airborne Asbestos Health Assessment Update (U.S. EPA 1986)
contain parameters (KL for lung cancer and KM for mesothelioma) that gauge the potency of
asbestos for causing these health effects. USEPA calculated KL and KM values from a number of
studies. In this section these KL and KM calculations are revised using the same models as in the
U.S. EPA (1986) update, but incorporating newer data from more recent publications. Since the
1986 update, additional cohorts have been studied from several new exposure settings and the
followup periods have been extended for several of the previously studied cohorts.
In the 1986 update, KM values were not calculated from all of the available studies, perhaps
owing to the limited number of mesotheliomas observed in some of these studies. In this update,
an attempt has been made to utilize any study with suitable health and exposure data, regardless
of the number of mesotheliomas reported, and to quantify the statistical uncertainty attributable
to small numbers using statistical confidence limits. Since the present work utilizes somewhat
different methods from the 1986 update, for consistency, all of the KL and KM values were
recalculated, even from studies for which no new data were available. Table A-l contains a
summary of the new values for KL and Table A-2 contains the new values for KM. The original
values from the 1986 update are also provided for comparison. These tables also contain
statistical confidence limits and ad hoc "uncertainty limits" for KL and KM. The derivation of
these limits will be described in detail in subsequent sections.
A.I LUNG CANCER MODEL
The 1986 U.S. EPA lung cancer model (U.S. EPA 1986) assumes that the relative risk, RR, of
mortality from lung cancer at any given age is a linear function of cumulative asbestos exposure
(fiber-years/ml, or f-y/ml, as measured by PCM), omitting any exposure in the most recent
10 years. This exposure variable is denoted by CE10. The 10-year lag embodies the assumption
that exposures during the most recent 10 years do not affect current lung cancer mortality risk.
The mathematical expression for this model is
RR=l-t-KL*CE10, (Eq. A-l)
where the linear slope, KL, is the "lung cancer potency factor." To make allowance for the
possibility that the background lung cancer risk in the exposed population differs from that of the
comparison population, the model is expanded to the form,
RR = a * (1 + KL * CE10). (Eq. A-2)
A.1
-------
With this form of the model the relative risk at zero exposure is a rather than 1.0. Both KL and a
are estimated by fitting the model to data. The type of data usually available for applying this
model are from cohort studies in which observed and expected (based on an appropriate
comparison population, e.g., U.S. males) numbers of lung cancers are categorized by cumulative
exposure incorporating a 10 year lag. To explore the adequacy of the model, it is useful to have
the data cross-classified by one or more other variables, such as latency.
Frequently the cumulative exposure variable available from the published report of a study does
not incorporate a lag (or, less frequently, incorporates a lag of less than 10 years). In this report,
rather than attempting an ad hoc correction, no correction for lag has been made. Although this
tends to cause KL values to be slightly underestimated, this is unlikely to be a serious problem.
For most cohorts, exposures decreased significantly over time. Also, in many studies, followup
didn't begin until several years after the start of exposure and the bulk of the lung cancers
occurred at older ages. All of these factors tend to mitigate the error created from use of data
with no lag. Moreover, use of an ad hoc correction for lag could hinder comparisons of KL
values among studies that do not employ a lag (which includes the majority of studies).
A.2 MESOTHELIOMA MODEL
The 1986 U.S. EPA mesothelioma model (U.S. EPA 1986) can be derived by assuming that the
mortality rate at time t after the beginning of exposure can be calculated by summing the
contributions from exposure at each increment of time, du, in the past. The contribution to the
mortality rate at time t from exposure to E(u) f/ml (as measured by PCM) at time u is assumed to
be proportional to the product of the exposure rate, E(u), and (t-u-10)2, the square of the elapsed
time minus a lag of 10 years. Thus, as with the lung cancer model, the mesothelioma model
assumes a 10-year lag before exposure has any effect upon risk. With the additional assumption
that the background rate of mesothelioma is zero, the mesothelioma mortality rate at time t since
the beginning of exposure is given by
IM(t)= B^'EM'tf-U-lOrdu, (Eq.A-3)
where t and u are in years, and IM(t) is the mortality rate per year at year t after the beginning of
exposure. The proportionality factor, KM, is called the "mesothelioma potency factor." The
factor of "3" is needed to retain the same meaning of KM as defined by U.S. EPA (1986).
If exposure is at a constant level, E, for a fixed duration, DUR, this model can be written as
0 Osts 10
IM(t) = KM*E*(t-10)3 10s t<; 10 + DUR (Eq. A-4)
KM * E * [(t - 10)3 - (t - 10 - DUR)3] DUR ^ t
The genesis of this model and its agreement with data were discussed in U.S. EPA (1986).
Through the courtesy of Dr. Corbett McDonald, Professor Douglass Liddell, Dr. Nicholas
de Klerk, Dr. John Dement, and the National Institute for Safety and Health (NIOSH), raw data
on mesothelioma mortality were obtained from a cohort of Quebec chrysotile miners and millers
A.2
-------
(Liddell et al. 1997; McDonald et al. 1980a), a cohort of Wittenoom, Australia crocidolite miners
and millers (Armstrong et al. 1988; de Klerk et al. 1994), and a cohort of workers from a plant in
Charleston, South Carolina that manufactured textiles from chrysotile (Dement et al. 1983a,b,
1994; Dement and Brown 1998). These data were used to calculate KM values in a more
accurate manner for these cohorts (using the "exact" approach described below) and to explore
the potential magnitude of the errors incurred by the crude application of cohort-wide averages
when fitting the mesothelioma model.
A.3 STATISTICAL FITTING METHODS
The method of maximum likelihood (Cox and Oakes 1984; Venzon and Moolgavkar 1988) was
used herein to fit the lung cancer and mesothelioma models to data and to estimate KL and KM.
The profile likelihood method was used to calculate statistical confidence intervals and
likelihood ratio tests were used to assess goodness-of-fit and test hypotheses.
Typically the data for calculating a lung cancer potency factor, KL, consist of observed and
expected (based on an external control group, such as U.S. males) numbers of cancer deaths
categorized by cumulative exposure. The likelihood of these data is determined by assuming
that the deaths in different exposure categories are independent and that the number of deaths in
a particular category has a Poisson distribution with expected number given by the expected
number predicted by the external control group times the relative risk given by either expression
(Eq. A-l or A-2).
In the typical situation, the published data most useful for calculating the mesothelioma potency
factor, KM, consist of the number of mesothelioma deaths and person-years of observation
categorized by time since first exposure. The likelihood of these data is determined by assuming
statistical independence of the number of mesothelioma deaths in different categories and that
the number of mesothelioma deaths in a category has a Poisson distribution with mean equal the
number of person-years in the category times expression (Eq. A-4), using average values for E,
DUR, and t appropriate for that category.
The fitting of the mesothelioma model (Eq. A-3) to raw (unsummarized) mesothelioma data is
accomplished using an "exact" maximum likelihood method. The cumulative mesothelioma
hazard is defined as
H(t)=foX(u)dU. (Eq.A-5)
The contribution to the likelihood of a person whose followup terminated at t is
S(t) = exp[ -H(t) ] if the followup did not terminate in death from mesothelioma, and IM(t) * S(t)
if the person died of mesothelioma The complete likelihood was defined as the product of these
individual contributions. The integrals in expressions (Eq. A-3 and A-5) were evaluated
numerically.
A.3
-------
A.4 SELECTION OF A "BEST ESTIMATE" OF KL AND KM
For each study for which a KL or KM is estimated, a "best estimate" is provided. For lung cancer,
the best estimate of KL (Table A-l) was generally assumed to be the maximum likelihood
estimate (MLE) obtained with a estimated. For mesothelioma, the best estimate of KM (Table
A-2) is generally the maximum likelihood estimate derived from the best-fitting model in the
form (Eq. A-3) for raw data and (Eq. A-4) for published data. As described in the descriptions of
the individual studies, in a few cases these general rules had to be adapted to fit the particular
form of the data available.
A.5 UNCERTAINTY IN KL AND KM
Statistical uncertainty in KL and KM estimates is expressed using 95% upper and lower statistical
confidence limits. These limits (summarized in Table A-l for lung cancer and Table A-2 for
mesothelioma) were computed using the profile likelihood method and (for KL) with a estimated.
However, non-statistical sources of uncertainty, such as model uncertainty and uncertainty in
exposure, are also likely to be very important. Although these uncertainties are difficult to
quantify, it is important to attempt quantification, since presentation of statistical uncertainty
alone may provide a misleading picture of the reliability of the estimates. Consequently, an ad
hoc approach to quantifying non-statistical uncertainty was adopted in this report. In this
approach, the primary sources of uncertainty are identified. Then, for each study, a factor was
selected for each uncertainty source using guidelines that will be described in this appendix. The
individual factors were combined with the statistical confidence bounds to arrive at an
"uncertainty range" for KL or KM for each particular cohort. These ranges are described in detail
in following sections and are summarized in Table A-l for lung cancer and Table A-2 for
mesothelioma.
Because the most serious uncertainties among published epidemiology studies are often
attributable to the estimation of exposure, three factors (Fl, F2, and F3) were defined to address
distinct sources of uncertainty associated with exposure. Two additional factors (F4L and F4M)
were defined to account for uncertainty due to special limitations that had to be addressed to
facilitate estimation of exposure-response factors from specific studies for lung cancer and
mesothelioma, respectively.
To define the factors we used to address uncertainty associated with exposure, we first
considered that, ideally, cumulative exposure would be estimated in an epidemiology study by:
• continuously monitoring the concentrations to which the worker is exposed over
their entire working life;
• measuring such concentrations using personal monitors (samplers worn by
workers with sampling ports placed within a few inches of the breathing zone of
the worker); and
A.4
-------
• analyzing samples in a manner appropriate for determining the concentration of
the specific range of structures of interest1.
In practice, however, measurements are collected only periodically at fixed locations considered
representative of worker exposures for jobs performed at that location (local operations).
Moreover, measurements were frequently derived using analytical methods that report results in
units different from those of interest, so that some type of conversion is required. Then,
typically, cumulative exposures are estimated for individual workers as the sum (over the set of
jobs held by that worker) of the product of the mean exposure concentration for each job and the
duration over which that job is performed. Thus:
uA (Eq. A-6)
j
where:
C is the cumulative exposure experienced by a worker to PCM fibers (f-years/ml);
Q is a factor used to convert concentration measurements in a particular study to
PCM fiber concentrations whenever the measurements in the study were collected
using a different method (usually dust concentrations determined by midget
impinger, in which case the units of Q are f/ml/mppcf);
CLO is the concentration estimated for a particular "local operation" typically derived
by a combination of measurement and extrapolation; and
Dj is the duration of time that the worker spent working in local operation "j".
Note that, because exposure concentrations at specific locations have generally been observed to
decrease over time due to changes in process, introduction of dust control equipment, and other
factors, cumulative annual exposures are generally estimated for workers in the manner
described above and the annual exposures are then summed. However, this does not change the
general applicability of Equation A-6.
Based on Equation A-6, a factor, Fl, is defined to account for uncertainty introduced in the
manner that the CLO are determined in specific epidemiology studies; a factor, F2, is used to
address uncertainty associated with the determination of the conversion factors, Q, for specific
studies; and F3 is defined to represent uncertainty in the manner that job matrices are developed
Most comparisons of epidemiology studies involve converting estimates of cumulative exposures to fiber
concentrations as determined by phase contrast microscopy (PCM) using the "membrane filter method". Thus, for
the discussion above, the range of structures (exposure index) of interest would be those determined using the
membrane filter method. Importantly, however, discussions in other portions of this document deal with determining
asbestos concentrations using an exposure index representing the specific range of structures that contribute directly
to biological activity, which should not be confused with the exposure index reported using the membrane filter
method.
A.5
-------
in specific studies to assign workers to specific local operations over specific durations. The
manner in which values were assigned for each uncertainty factor is described more fully below.
A.5.1 The Factor Fl
As indicated above, the factor, Fl, represents the uncertainty in concentration estimates to which
workers are exposed (in whatever units of exposure that are reported in a particular study). In
addition to analytical uncertainty, considerations addressed when assigning values for Fl for
specific epidemiology studies include:
• to what extent exposure concentrations were directly determined from
measurements collected at the locations and times that worker exposures actually
occurred; and
• whether measurements were derived from personal monitoring or from area
monitoring (sampling a general area that is assumed representative of exposure
conditions associated with jobs performed within the local area).
Regarding the latter consideration, exposure concentrations estimated in the published
epidemiology studies were almost universally determined by area, rather than personal
monitoring. As has been reported in several of these studies (see, for example, McDonald et al.
1983b), area monitoring can miss short-term, high-level exposures contributed by the personal
actions being performed by a worker. Moreover, certain periodic activities potentially
associated with extremely high exposure (typically involving cleanup) were not performed
during time periods when work areas were routinely monitored.
Regarding the first bullet above, published epidemiology studies differ in the frequency and time
period over which sampling was conducted. With few exceptions, little or no sampling was
conducted prior to the 1950's when exposure concentrations are thought generally to be higher
than those monitored more recently, due to lack of use of dust control equipment and procedures
that were introduced only later. For many studies, therefore, early exposures had to be estimated
by extrapolation from later measurements and the care with which such extrapolations were
performed also varies from study to study.
Studies vary in the degree to which the range of local operations associated with a particular
facility were individually sampled. Exposure conditions attendant to jobs performed in
association with local operations not sampled directly would then be extrapolated from
measurements collected for other local operations assumed to be associated with "comparable
exposures." As with extrapolations in time, the care with which such spatial extrapolations were
performed varies from study to study.
Values assigned for Fl vary between 1.5 and 4 (all to the nearest 0.5). The most typical value
assigned is 2.0 for studies in which additional uncertainty is introduced due to use of area
samplers rather than personal samplers, lack of measurements representative of episodic but
high-exposure jobs (usually associated with cleanup), and lack of direct measurements from the
earliest periods of exposure (when dust control equipment and procedures were absent). To be
assigned a value of 2.0, however, authors must have had access to substantial numbers of
A.6
-------
samples representative of the majority of the local operations of interest, must have described a
systematic procedure for extrapolating exposure estimates to less well studied local operations,
and must have described a systematic procedure for extrapolating exposure estimates to earlier
times when measurements were lacking. The logic used to assign Fl values (and values for the
other uncertainty factors) for individual studies is described for each study in Section A. 6 of this
appendix.
A.5.2 The Factor F2
F2 is a factor used to characterize the uncertainty introduced in deriving conversion factors to
convert from the exposure indices measured in a particular study to the exposure index typically
reported using the membrane filter method (as determined by PCM). In about half of the studies,
concentrations are estimated in millions of dust particles per cubic foot (mppcf) as determined by
midget impinger (see Section 4.3). The uncertainty introduced by such conversions varies from
study to study because:
• for a small number of studies, the majority of measurements were performed by
the membrane filter method so that conversion was unnecessary;
• for some studies, conversion factors were derived from a statistical analysis of a
set of side-by-side measurements determined, respectively, using the membrane
filter method and the other method from which measurements need to be
converted (typically the midget impinger method);
• for some studies, lack of side-by-side measurements required expert judgement
for comparing across samples collected at different times and locations; and
• for some studies, conversion factors were not derived at all, but were adapted
from other studies of similar processes.
Moreover, as has been demonstrated in several studies, the factors used to convert other
measurements (primarily midget impinger) to the exposure index determined by PCM vary as a
function of study environment, local operation, and time. For example, the ratio of PCM to
midget impinger derived from side-by-side measurements in a single study reportedly varied
between 0.3 and 30 (McDonald et al. 1980a).
Note that, given the above, the factors used to convert measured concentrations to exposure
concentrations in units of interest (Q in Equation A-6) ideally should be brought into the sum on
the right and determined individually for each local operation. However, with the exception of
the South Carolina study by Dement and coworkers (Dement et al. 1994; Dement and Brown
1998), only average (study-wide) conversion factors are typically estimated in any particular
study.
Values for F2 assigned to particular studies vary between 1.0 and 3.0. Studies in which
conversions were not required (due to routine use of PCM) or studies in which conversion
factors were determined for specific operations were assigned an F2 value of 1.0. Studies in
which a study-wide conversion factor was determined from paired measurements are assigned a
A.7
-------
value of 1.5. Studies in which conversion factors were adapted from other studies or for which
authors did not define a conversion factor were assigned larger values for F2.
A.5.3 The Factor F3
The factor, F3, is used in this study to represent the uncertainty attributable to the manner in
which job-exposure matrices were constructed in the various published epidemiology studies.
Authors for some studies had detailed work histories that could be used to identify the complete
set of specific jobs that each worker performed over their working life and the duration of time
spent on each job. Authors from other studies did not have access to individual work histories so
that crude estimates of average duration was applied to all members of the cohort. The factor,
F3, is used to account for conditions in which less than optimal job histories were used to
identify the set of jobs performed by each worker and the duration that each worker spent
performing each such job.
A.5.4 The Factor F4L for Lung Cancer and F4M for mesothelioma
An additional factor is included (F4L for lung cancer) and (F4M for mesothelioma) to account
for uncertainties in mortality data (e.g., when diagnosis is uncertain for a substantial fraction of
potential mesothelioma cases) or when approximations or assumptions are required because the
data are not presented in the form needed for fitting the exposure-response models. Two
assigned F4L values are greater than 1.0 (1.5 and 2.0), and six F4M values are greater than 1.0;
these six values range from 2.0 to 5.0.
A.5.5 Combining Individual Uncertainty Factors into an Overall "Uncertainty Range"
As indicated above, in addition to statistical confidence intervals, four uncertainty factors have
been proposed: Fl: exposure, general; F2: exposure conversion factor; F3: lack of individual
work histories; and F4L (lung cancer) and F4M (mesothelioma): non-exposure related. Since it
is unlikely that all of the uncertainty sources would cause errors in the same direction in the
same study, rather than multiplying the uncertainty factors, an overall uncertainty factor, F, was
calculated as:
F = exp{ [ Ln2(Fl) + Ln2(F2) + Ln2(F3) + Ln2(F4) ]* },
where 1.0 is the default value for any factor not explicitly provided. The overall "uncertainty
range" for KL or KM was calculated by dividing the statistical 95% lower bound by F and
multiplying the 95% upper bound by F.
A.6 ANALYSIS OF INDIVIDUAL EPIDEMIOLOGY STUDIES
Predominately Chrysotile Exposure
Quebec Mines and Mills. Liddell et al. 1997 extended the followup into 1992 of a cohort of
about eleven thousand workers at two chrysotile asbestos mines and related mills in Quebec that
had been studied earlier by McDonald et al. 1980b (follow-up through 1975) and McDonald
et al. 1993 (follow-up through 1988). Production at the mines began before 1900. The cohort
A.8
-------
consisted of workers who worked ^ 1 month and who were born between the years of 1 891 and
1920. Follow-up began for each individual after 20 years from first employment. The most
recent follow-up (Liddell et al. 1997) traced 9,780 men through May 1992, whereas 1,138 (10%)
were lost to view, most of whom worked for only a few months prior to 1935. Of those traced,
8,009 (82%) were deceased as of 1992.
Estimates of dust levels in specific jobs were made from some 4,000 midget impinger
measurements collected systematically starting in 1948 and periodically in the factory beginning
in 1944. Estimates for the period prior to 1949 utilized interviews with long-term employees and
comparison with more recent conditions. These dust-level estimates were matched to individual
job histories to produce estimates of cumulative exposure for each worker (mppcf-years).
Conversions between dust levels and PCM concentrations were derived from side-by-side
samples. On the basis of over 600 side-by-side midget impinger and optical microscopy
measurements, it was estimated that 3.14 fibers/ml was, on the average, equivalent to 1 .0 mppcf
(McDonald et al. 1980b).
Liddell et al. (1997) categorized cancer deaths after age 55 from of lung, trachea, and bronchus
by cumulative asbestos exposure to that age (Liddell et al. 1997, Table 8). Standardized
mortality ratios (SMRs) were calculated based on Quebec rates from 1950 onward, and
Canadian, or a combination of Canadian and Quebec rates, for earlier years. Table A-4 shows
the fit of the lung cancer model to these data. Although the models both with cc=l and a variable
provided reasonably adequate fits to the data, the hypothesis a=l can be rejected (p=0.014). The
model with a estimated yields a best estimate of KL of 0.00029 (f-y/ml)'1, 90% CI: (0.00019,
0.00041). With ce=l, the estimate was 1^=0.00041 (f-y/ml)-1, 90% CI: (0.00032, 0.00051).
Smoking history was obtained in 1970 by a questionnaire administered to current workers, and
to proxies of those who had died after 1950. Although no analyses of lung cancer and asbestos
exposure were presented for the 1992 follow-up (Liddell et al. 1997) that controlled for smoking,
such an analysis was conducted for the follow-up that continued through 1975 (McDonald et al.
1980a). Table 9 of McDonald et al. (1980a) contained data on lung cancer categorized jointly by
cumulative exposure to asbestos and by smoking habit. Two models were fit to these data: the
multiplicative model for relative risk
= a*(l+b*d)*(l + c*x),
and the additive model
RR=a*(l+b*d + c*x),
where d is cumulative exposure to asbestos to age 45, x is number of cigarettes smoked per day,
and a,b, and c are parameters estimated from the data. The multiplicative model fit the data
well, but the fit of the additive model was inadequate. This corroborates the multiplicative
interaction between smoking and asbestos exposure in causing lung cancer (Hammond et al.
1979). The estimate of potency using the multiplicative model was 0.00051 (f-y/ml)-1, which
was very close to that of 0.00045 (f-y/ml)-1 estimated from Table 5 of McDonald et al. (1980a),
which did not utilize smoking data. This suggests that the association between lung cancer and
asbestos exposure is not strongly confounded with smoking in this cohort.
A. 9
-------
By 1993, 38 deaths from mesothelioma had occurred in this cohort (Liddell et al. 1997).
Through the courtesy of Dr. Corbett McDonald and Professor Douglass Liddell, the underlying
mesothelioma data from this study were provided for additional analysis (Liddell 2001). These
data contained the following information on each worker: the date of birth, asbestos exposure
history, last date of follow-up, whether follow-up ended as a result of death from mesothelioma,
location of first employment, and whether a worker had been employed at more than one
location.
Nine distinct locations for first employment were coded. Locations 5-9 referred to small
operations, some having very heterogeneous exposures, and were omitted from the analysis.
Also, workers who worked at more than one location were omitted. After these exclusions, there
remained 9,244 workers who worked at Locations 1-4, and among whom 35 deaths from
mesothelioma occurred. Location 1 (4,195 men, 8 deaths from mesothelioma) was the mine and
mill at the town of Asbestos. Location 2 (758 men, 5 deaths) was a factory at the town of
Asbestos that, in addition to processing chrysotile, had also processed some crocidolite.
Location 3 (4,032 men, 20 deaths) comprised a major mining and milling company complex near
Thetford Mines. Location 4 (259 men, 2 deaths) comprised a number of smaller mines and mills
also in the vicinity of Thetford Mines. Because of the small number of workers at Location 4,
the fact that both locations were near Thetford Mines, and the fact that the separate KM values
obtained from Locations 3 and 4 were similar, data from these locations were combined. The
remaining groups were analyzed separately, because of the crocidolite used at Location 2, and
because of evidence of greater amounts of tremolite in the ore at Thetford Mines that at Asbestos
(Liddell etal. 1997).
The availability of the raw data from this study made it possible calculate KM from this study
using an "exact" likelihood approach based on expression (Eq. A-3) that did not involve any
grouping of data, or use of average values. For Location 1 (Asbestos mine and mill),
K^O.ODxlO-8, 90% CI: (0.0068xlO'8, 0.022xlO'8). For Location 2 (Asbestos factory),
KM=0.092x10-8, 90% CI: (0.040x1O'8, 0.18x10'8). For Locations 3 and 4, KM= 0.021xlO'8, 90%
CI: (0.014xlO'8, 0.029xlO"8). The KM estimate from Location 1 (whose ore was reported to have
a lower tremolite content) was about one-half that from Locations 3 and 4, although this
difference was not significant (p=0.22). The KM estimated from Location 2, the mill where
substantial crocidolite was used, was 4-7 times higher than the KL estimated from Location 1
and Locations 3 and 4.
For comparison purposes, KM were also calculated using grouped data and applying expression
(4), since this is the method that must be used with most studies. For Location 1 (3 and 4) the
KM estimate based on the "exact" analysis was 34% (25%) higher than that based upon grouped
data This suggests that reliance upon published data for calculating KM may introduce some
significant errors in some cases. Such errors may be further compounded by the failure of some
studies to report the needed data on levels and durations of exposure in different categories of
time since first exposure.
For this study Fl is set equal to 2.0. This study is the paradigm used to define the typical case
(see Section A.5.1) in which increased uncertainty can be attributed to use of area rather than
personal samplers, lack of measurements early in the study, and lack of direct measurements
from certain episodic but high-exposure operations. At the same time, the authors of this study
A.1O
-------
appear to have used the available data in a systematic and objective manner to address the issues
raised by the lack of sampling.
The uncertainty factor F2 is set equal 1.5 for this study to reflect use of a conversion factor that
is derived from paired samples, but that is based on a project wide average, rather than
addressing variation for specific, local operations.
All other uncertainty factors are set equal to 1.0 for this study due to lack of remarkable
distinctions. Thus:
Fl = 2.0
F2=1.5
F3 = 1.0
F4L = 1.0
F4M=1.0
These uncertainty factors, when coupled with the statistical confidence limits, resulted in the
uncertainty ranges for KL and KM shown in Tables A-l and A-2.
Italian Mine and Mill. Piolatto et al. (1990) conducted additional follow-up of workers at a
chrysotile mine and mill in Italy that was earlier studied by Rubino et al. (1979). The cohort
consisted of 1058 workers with at least 1 year of employment between 1946 and 1987. Follow-
up extended from 1946 through 1987, which is 12 more years of follow-up than in Rubino et al.
(1979). Lung cancer mortality was compared to that of Italian men.
As described in Rubino et al. (1979), fiber levels were measured by PCM in 1969. In order to
estimate earlier exposures, information on daily production, equipment changes, number of
hours worked per day, etc. were used to create conditions at the plant during earlier years. PCM
samples were obtained under these simulated conditions and combined with work histories to
create individual exposure histories.
Piolatto et al. (1990) observed 22 lung cancers compared to 11 in the earlier study (Rubino et al.
1979). Lung cancer was neither significantly in excess nor significantly related to cumulative
asbestos exposure. Piolatto et al. (1990, Table 1) presented observed and expected lung cancers
(based on age- and calendar-year-specific rates for Italian men) categorized by cumulative
exposure in f-y/ml. The lung cancer model with fixed a provided a good fit to these data (Table
A-5, p=0.75) and allowing a to vary did not significantly improve the fit. The KL estimate with
-------
Fl=2.0
F2 = 1.0
F3 = 1.0
F4L=1.0
These uncertainty factors, when coupled with the statistical confidence limits, resulted in the
uncertainty range for KL shown in Table A-l.
Connecticut Friction Product Plant. McDonald et al. (1984) evaluated the mortality of
workers employed in a Connecticut plant that manufactured asbestos friction products. The
plant began operation in 1913 and used only chrysotile until 1957, when a little anthophyllite
was used. Also, a small amount of crocidolite (about 400 pounds) was handled experimentally
between 1964 and 1972. Brake linings and clutch facings were made beginning in the 1930s,
and production of automatic transmission friction materials, friction disks and bands was begun
in the 1940s.
The cohort was defined to include any man who had been employed at the plant for at least
1 month before 1959, omitting all that had worked at a nearby asbestos textile plant that closed
in 1939. This cohort consisted of 3,515 men, of whom 36% had died by the end of follow-up
(December 31,1977). Follow-up of each worker was only begun past 20 years from first
employment.
Information on dust levels from impinger measurements were available for the years 1930,1935,
1936, and 1939. There was little other exposure information available until the 1970s. An
industrial hygienist used these measurements and information on processes and jobs,
environmental conditions and dust controls to estimate exposures by process and by period in
units of mppcf. No conversion from mppcf to f/ml value was suggested by the authors, a
conversion factor or between 1.4 and 10 is suggested by other studies. The most common value
seems to be around 3 f/ml per mppcf, which has been observed in diverse environments such as
mining and textile manufacture. This value was provisionally applied to this cohort, although
this conversion has considerable uncertainty associated with it.
Total deaths and deaths from most individual causes investigated were elevated; these elevations
were due primarily to increased deaths in the group working for <1 year. This pattern holds for
lung cancer in particular; the SMR for lung cancer was highest (180) for persons exposed for
<1 year. A similar pattern holds when the analysis was carried out by cumulative exposure
(Table A-6); the SMR in the lowest exposure category is higher than in any other category. The
linear relative risk lung cancer model provided a poor fit (p=0.01) to these data when the
Connecticut rates were assumed to be appropriate for this cohort (fixing the parameter a=l); use
of U.S. rates gave similar results. However, the fit was adequate (p=0.28) if the background
response is allowed to rise above that of Connecticut men (allowing the parameter a to vary).
Although the reason for this increased response in persons that worked for a short period or have
low exposures is not clear, the analysis in which the background response is allowed to vary
appears to be the most appropriate. This analysis yields an estimate of 1^=0.0 (f-y/ml)"1, 90%
CI: (0, 0.0017). The analysis with a=l yielded KL=0.0019 (f-y/ml)-1, 90% CI: (0, 0.0061).
A.12
-------
McDonald et al. did not find any mesotheliomas in this cohort. It is useful to determine the
range of mesothelioma risk that is consistent with this negative finding. Although McDonald et
al. do not furnish data in the form needed for this calculation, these data can be approximated
from Table 1 of McDonald et al. (1984). In this table they list 511 deaths occurring after age 65.
Assuming that the overall SMR of 108.5 held for persons over 65 years of age, the expected
number of deaths is 511/1.085 = 471. The death rate in U.S. white males between 65 and 75
years of age is approximately 0.050 per year (from 1971 vital statistics). Therefore the number
of person years observed in persons post 65 years of age is estimated as 471/0.050=9,420.
A lower bound on the person-years of follow-up between ages 45 and 65 can be estimated by
assuming that follow-up was complete for this age group. First we estimate the number of
persons that would have had to have been in the cohort to experience the observed deaths.
Assuming that x persons in the cohort are alive at age 45, we have the following estimates of the
number entering each successive five-year age interval and the corresponding number of deaths
(based on death rates in 1,971 white males).
Age
45-50
50-55
55-60
60-65
65+
TOTALS
Number Entering Interval
X
x(l-0.00638)5=0.97x
0.97x(l-0.01072)5=0.92x
0.92x(l-0.01718)5=0.84x
0.84x(l-0.02681)5=0.73x
Number of Deaths
in Interval
0.032x
0.052x
0.076x
0.1 Ix
0.27x
Person-Years in
Interval
4.9x
4.7x
4.4x
3.9x
18.0x
Since there were 616 deaths in men between the ages of 45 and 65, the expected number of
deaths is estimated as 616/1.085=567.7 expected deaths between ages of 45 and 60, the number
of persons entering this age interval is estimated as x=567.7/0.27=2,100. The person-years is
then estimated as (2,100)(17.964)=38,000.
Using the average age of beginning work of 30.95 years (McDonald et al. [1984], Table 3) yields
the data in Table A-7. Moreover, the average duration of exposure in this cohort was 8.04 years
and the average exposure level was 1.84 mppcf (McDonald et al. [1984], Table 3), which is
equivalent to 1.84x3=5.52 fibers/ml. These data yields an estimate of KM=0.0 and a 90% upper
bound of KM=1.2xlO-9.
The best estimate of KM was assumed to be zero. For uncertainty factors, Fl is assigned a value
of 2.0 for reasons similar to those described for Quebec. F2 is assigned a value of 3.0 for this
study because there is no conversion factor reported by the authors so that an average value of 3
for the range of conversion factors observed among the available studies (U.S. EPA 1986) was
selected for use with this study. To derive an exposure-response factor for mesothelioma from
this study, an upper bound had to be estimated by reconstructing the data because the authors do
not provide the data in a form suitable for performing the required calculation. Therefore, F4M
is assigned a value of 3 for this study. Thus:
A.13
-------
Fl=2.0
F2 = 3.0
F3 = 1.0
F4L = 1.0
F4M = 3.0
These values, when coupled with the statistical confidence limits, resulted in the uncertainty
ranges for KL and KM shown in Table A-l and A-2, respectively.
New Orleans Asbestos-Cement Plants. Hughes et al. (1987) report on follow-up through 1981
of a cohort of Louisiana workers from two asbestos cement plants studied previously by Weill et
al. (1979). Although chrysotile, amosite and crocidolite were used at these plants, a group of
workers at one of the plants were only exposed to chrysotile. The cohort contained 6,931
workers, of whom 95% were traced, compared to a 75% success in tracing by Weill et al. (1979).
This improved trace was the result both of greater access to Social Security Administration
records and greater availability of computerized secondary information sources (Dr. Hughes,
personal communication).
Both of the plants have operated since the 1920s. Chrysotile was used predominantly in both
plants. Some amosite was used in Plant 1 from the early 1940s until the late 1960s, constituting
about 1% of some products, and crocidolite was used occasionally for approximately 10 years
beginning in 1962. Plant 2 utilized only chrysotile, except that pipe production, which began in
1946 and was housed in a separate building, produced a final product that contained about 3%
crocidolite. Since the total percentage of asbestos fiber in most asbestos cement products ranges
from 15 to 28%, it is estimated that crocidolite constituted between 10 and 20% of the asbestos
used to make cement pipe (Ontario Royal Commission 1984). Workers from Plant 2 that did not
work in pipe production were exposed only to chrysotile.
Estimates of airborne dust levels were made for each job by month and year from midget
impinger measurements initiated in the early 1950s. Levels estimated from initial samples in the
1950s were also assumed to hold for all earlier periods because no major dust control measures
had been introduced prior to that time. New exposure data from Plant 2 became become
available after the earlier study (Weill et al. 1979) was completed, and these, along with a
complete review of all the exposure data, were used to revise the previous estimates of exposure.
In Plant 1 the earlier and revised estimates were reasonably similar, but in Plant 2, the revised
estimates tended to be about one-third of the previous estimates through the 1940s and about
one-half the previous estimates thereafter. Based on 102 side-by-side measurements by midget
impinger and PCM in various areas of one of the plants, Hammond et al. (1979) estimated an
overall conversion factor of 1.4 fibers/ml per mppcf. There were substantial variations in this
factor among different areas of the plant.
The principal cohort studied consisted of all workers who, according to company records, were
employed for at least one month prior to 1970, had a valid Social Security number, and were first
employed in 1942 or later (Plant 1), or in 1937 or later (Plant 2). Mortality experience was
compared with that expected based on Louisiana rates.
A.14
-------
Hughes et al. found no significant difference between the exposure responses for lung cancer in
Plant 2 among workers exposed to chrysotile only and those who were also exposed to
crocidolite in pipe production. A single lung cancer exposure response model adequately
describes the lung cancer data from Plants 1 and 2 combined (p^O.42, Table A-8). The fit of this
model is good when Louisiana men are assumed to be an appropriate control group (fixing the
parameter a=l). This fit provides an estimate of 1^=0.0040 (fiber-y/ml)-1, 90% CI: (0.0015,
0.0070). With a allowed to vary, the estimate is 0.0025 (fiber-y/ml)-1, 90% CI: (0, 0.0066 ).
Six mesotheliomas were identified in the primary cohort studied by Hughes et al., two in Plant 1
and four in Plant 2. Four other mesotheliomas are known to have occurred, one among those
initially employed in Plant 2 before 1937 and three among Plant 2 workers shortly after follow-
up ended in 1981. A case control analysis conducted among Plant 2 workers found a
relationship between mesothelioma risk and length of employment and proportion of time spent
in the pipe area after controlling for length of exposure, which is consistent with a greater risk of
mesothelioma from crocidolite exposure.
Data were not presented in the paper in the form required for estimating KM. However, Hughes
and Weill (1986) present estimates of mesotheliomas potency from several data sets, including
the cohort studied in Hughes et al. and containing six mesotheliomas, but using a model slightly
different from the 1986 EPA model (3). Estimating KM by multiplying the potency estimated by
the Hughes and Weill (1986) model by the ratio of the potency values estimated for another
study using the 1986 U.S. EPA model and the Hughes-Weill (1986) model yielded the following
estimates of KM for the Hughes et al. (1987) data: 0.25x10'8 (Selikoff et al. 1979); 0.21xlQ-8
(Dement et al. 1983b); 0.27xlO-8(Seidman et al. 1979); and 0.43xlO'8 (Finkelstein 1983). Based
on these calculations, KM=0.30xlO'8 seems to be a reasonable estimate for the Hughes et al.
cohort.
It would be worthwhile to estimate mesothelioma risk using additional follow-up that included
the three cases that occurred shortly after follow-up ended. However, such an estimate should be
no larger than about KM=0.45xlO'8. This is because, since there were six mesotheliomas in the
cohort studied by Hughes et al., even if the additional person years of follow-up post-1981 is not
taken into account, the three additional mesotheliomas would increase the estimate of KM by
only about 50%.
The finding by Hughes et al. (1987) of an association with crocidolite exposure implies that a
smaller KM would correspond to the chrysotile-only exposed group in Plant 2. Although Hughes
et al. didn't furnish the data needed for precise estimation of KM from this cohort, it is possible to
make some reasonable approximations to this KM. Since none of the six mesotheliomas occurred
among workers exposed only to chrysotile, KM=0 would be the point estimate derived from the
data used by Hughes et al.
However, one mesothelioma was discovered in a person whose employment began in 1927 and
thus was not eligible for inclusion in the cohort. This person was employed continuously for 43
years in the shingle production area, where only chrysotile was used. In an attempt to compute
an alternative KM using this one case, it was noted that the duration of observation of the Hughes
et al. cohort was roughly equivalent to that of the Dement et al. (1983b) cohort. If the person-
years from this cohort, categorized by years since first exposure, are adjusted by the ratio of the
A.15
-------
sizes of Dement et al. and the Hughes et al. non-crocidolite-exposed cohort from Plant 2, one
mesothelioma is assumed to occur (in 30+ years from first exposure category) and the average
duration of exposure (2.5 years) and fiber level (11.2 fibers/ml) appropriate for the Hughes et al.
cohort are applied to these data, a KM=0.2xlO'8 is obtained.
The best estimate of KM was assumed to be 0.2x10"8 for workers exposed only to chrysotile and
0.3x10"8 for workers exposed to both chrysotile and amphibole. For uncertainty factors, Fl is
assigned a value of 2.0 for reasons similar to those described for Quebec. F2 is assigned a value
of 1.5 because most early measurements were collected by midget impinger and the authors
report using a conversion factor of 1.4 derived from paired measurements. Due to the lack of
adequate data for estimating both the overall mesothelioma rate and a confidence interval for
such rates and the consequent need to reconstruct the data (incorporating numerous assumptions)
to be able to obtain the needed estimates, a value of 5.0 was assigned to the factor F4M for
chrysotile exposures and 2.5 for mixed exposures. Thus:
Fl=2.0
F2=1.5
F3 = 1.0
F4L = 1.0
F4M = 5.0
These values, when coupled with the statistical confidence limits, resulted in the uncertainty
ranges for KL and KM shown in Tables A-l and A-2.
South Carolina Textile Factory. Dement and coworkers (Dement et al. 1994; Dement and
Brown 1998) conducted a retrospective cohort study of employees of a chrysotile textile plant in
South Carolina. In an earlier study of this plant (Dement et al. 1982,1983a,b), the cohort was
defined as all white male workers who worked for one or more months between 1940 and 1965,
and follow-up was through 1975. Dement et al. (1994) expanded the cohort to include black
male and white female workers who met the entrance requirements, and extended follow-up
through 1990, an additional 15 years. This expanded cohort included 1,247 white males (2.8%
lost to follow-up), 1,229 white females (22.8% lost to follow-up) and 546 black males (7.8% lost
to follow-up). A total of 1,259 deaths were identified, and a death certificate was located for all
but 79 (6.2%) of the deaths.
Based on data from 5,952 air samples taken at the plant between 1930 and 1975, linear statistical
models were used to reconstruct exposure levels, while taking into account textile processes,
dust control methods, and job assignments (Dement et al. 1983a). For each worker, time spent in
each job was multiplied by the estimated exposure level for that job to estimate cumulative
exposure (f/ml-days). Based on regression analyses applied to 120 side-by-side particle and
fiber counts, Dement (1980) estimated a f/ml to mppcf ratio of 2.9, 95% CI: (2.4, 3.5). Also,
between 1968 and 1971 both impinger and PCM samples were collected (a total of 986 samples).
Based upon a regression analysis of these data, Dement (1980) determined that a common
conversion factor could be used for jobs except fiber preparation. For fiber preparation, a
conversion factor of 7.8 was found, 95% CI: (4.7-9.1). For all other operations, a value of 2.5,
95% CI: (2.1-3.0) was calculated. Based on this information, Dement et al. (1983a) concluded
A.16
-------
that a conversion factor of 3 was appropriate for all operations except preparation, for which a
factor of 8 was adopted.
The underlying data for this cohort were obtained from the National Institute for Safety and
Health (NIOSH). These data consisted of a work history file and a file with exposure levels by
job category and time period. The work history file contained codes for race, sex, month and
year of birth, vital status, month and year of death, and the department, operation, start date, and
stop date for each job worked. The exposure level file contained the exposure start and stop
dates and the exposure level (fiber/ml) by the plant code, the department code, and the operation
code.
The cohort was defined as the white and black males and the white females who met the
employment requirements described above. This cohort included 1,244 white males (1.5% lost
to follow-up), 550 black males (7.5% lost to follow-up), and 1,228 white females (22.1%lostto
follow-up).
Table A-9 shows observed and expected deaths for lung cancer among white males, black males
and white females, categorized by cumulative exposure. This table shows an excess of lung
cancers that exhibited an exposure response relationship. U.S. rates were used for calculating
expected deaths, whereas South Carolina lung cancer rates are higher for white men, but slightly
lower for white women and black men. Whereas twelve categories of cumulative exposure were
used for fitting the model, these were been combined into seven categories for display in Table
A-9. The model with a=l and a variable fit the data well (ps 0.8), and the hypothesis that a=l
cannot be rejected (p=0.19). The estimate of KL with o=l was 0.028 (f-y/ml)-1, 90% CI: (0.021,
0.037), and the estimate with a variable was KL=0.021 (f-y/ml)-1, 90% CI: (0.012, 0.034). An
analysis applied to white men alone gave somewhat higher estimates (KL=0.040 (f-y/ml)-1 with
a=l, and KL=0.026 (f-y/ml)-1 with a variable).
Two deaths were certified as due to mesothelioma on the death certificates. In addition, Dement
et al. (1994) considered four other deaths as likely due to mesothelioma. The availability of the
raw data from this study made it possible calculate KM from this study using an "exact"
likelihood approach based on Equation A-3 that did not involve any grouping of data, or use of
average values. Using the six confirmed and suspected mesotheliomas, KM= 0.43x10'8, 90% CI:
(0.20xlQ-8, 0.79xlQ-8). Using the two confirmed mesotheliomas, KM=0.14xlO'8, 90% CI:
(0.034x10'8, 0.38x10'8).
For comparison purposes, KM were also calculated using grouped data and applying Equation
A-4, since this is the method that must be used with most studies. The data were divided into 10
categories by the tabulated values of Equation A-4. The KM estimate based on the "exact"
analysis was 2% greater than that based upon grouped data.
The best estimate of KM was assumed to be the geometric mean of the MLE estimates computed
using either confirmed or both confirmed and suspected mesotheliomas (0.25xlO"8). The
statistical lower bound used for this estimate was the one based on confirmed cases and the
upper bound used was the one based on confirmed and suspected cases.
A.17
-------
Regarding uncertainty factors, Fl is assigned a value of 1.5 for this study to give credit for the
reasonably complete sampling coverage of exposures by a combination of midget impinger and
extensive PCM, and the formal statistical evaluation conducted to derive job-specific exposure
estimates. However, the exposure estimates are still based on analyses of area rather than
personal samples. Because multiple factors were used to convert midget impinger measurements
to PCM based on side-by-side samples collected from specific areas (associated with specific
operations) within the plant, a value of 1.0 is assigned for F2 for this study. The treatment of
statistical confidence limits described above was considered adequate to account for the
uncertainty in the number of mesotheliomas, and a value of KM=1 was assigned. In summary:
Fl = 1.5
F2 = 1.0
F3 = 1.0
F4L = 1.0
F4M=1.0
These uncertainty factors, when coupled with the statistical confidence limits, resulted in the
uncertainty ranges for KL and KM shown in Tables A-l and A-2.
McDonald et al. (1983a) conducted a cohort mortality study in the same South Carolina textile
plant that was studied by Dement et al. (1994). Their cohort consisted of all men employed for
at least 1 month before 1959 and for whom a valid social security record existed. This cohort
consisted of 2,410 men, of whom 36% had died by the end of follow-up (December 31,1977).
Follow-up of each worker was begun past 20 years from first employment.
McDonald et al. (1983a) had available the same exposure measurements as Dement et al.
(1983b) and used these to estimate cumulative exposures for each man in mppcf-y. In their
review of the environmental measurements in which both dust and fiber concentrations were
assessed, they found a particle to fiber conversion range of from 1.3 to 10.0 with an average of
about 6 fibers/ml per mppcf. This value, which is intermediate between the values of 3 and 8
found by Dement et al. (1983b) for different areas of the same plant, will be used in the
calculations involving the McDonald et al. (1983a) study.
McDonald et al. describe two practices at the plant that entailed very high exposures and which
were not reflected in either their's or Dement and coworkers estimates: cleaning of burlap bags
used in the air filtration system by beating them with buggy whips during the years 1937-1953,
and the mixing of fibers, which was carried out between 1945 and 1964 by men with pitch forks
and no dust suppression equipment.
A strong exposure response for lung cancer was observed (Table A-10), which parallels the
results of Dement et al. (1994). Unlike Dement et al., McDonald et al. used South Carolina men
as the control group rather than U.S. men. Use of this control group provided an adequate
description of the data and lung cancer potency values estimated both with oc=l and allowing a
to vary provided excellent descriptions of the data (p^0.88) and the hypothesis a=l could not be
rejected (p=0.80). Assuming o=l resulted in KL=0.012 (f-y/ml)'1, 90% CI: (0.0075, 0.016), and
when a was allowed to vary, KL=0.010 (f-y/ml)'1, 90% CI: (0.0044, 0.025). These results are
A.18
-------
reasonably consistent with the potency estimated from Dement et al. (1994), and the differences
can be largely accounted for by the different assumptions regarding the fiber/particle ratio.
McDonald et al. (1983a) found one case of mesothelioma in this cohort, apparently the same one
discovered by Dement et al. (1983b): a man born in 1904 who died in 1967 and worked at the
plant for over 30 years. Since this study was conducted exactly as McDonald et al. (1984), the
same method used there to reconstruct person-years by years from first exposure can be applied
to this cohort as well. The reconstructed data are listed in Table A-l 1. The estimated potency
MLE is KM=0.088 xlO'8, with a 90% CI: (0.0093xlO'8, 0.32x10'8).
For uncertainty factors, Fl is assigned a value of 2.0 for reasons similar to those described for
Quebec. F2 is assigned a value of 1.0 because McDonald essentially used the same data that
Dement and coworkers used to estimate conversion factors (see above), although they favored a
slightly higher mean value. We used the values favored by Dement when evaluating this study.
Unlike the study by Dement and coworkers (for which we received the raw data so that we
could calculate the exposure-response factor and the attendant confidence interval for
mesothelioma directly), the mesothelioma data published in the McDonald study of this facility
was not suitable to estimating confidence bounds. Thus the data had to be reconstructed, which
required incorporation of numerous assumptions. To account for the uncertainty associated with
the reconstruction, F4M is assigned a value of 3 for this study. Thus:
Fl=2.0
F2=1.0
F3 = 1.0
F4L = 1.0
F4M - 3.0
These uncertainty factors, when coupled with the statistical confidence limits, resulted in the
uncertainty ranges for KL and KM shown in Tables A-l and A-2.
Predominant Crocidolite Exposure
Wittenoom, Australia Mine and Mill, de Klerk et al. (1994) followed a cohort of 6,904 men
and women employed at a crocidolite mine and mill in Wittenoom, Australia. This cohort was
followed through 1999 and the raw data were obtained through the courtesy of Dr. de Klerk.
The data consisted of a record number, date of birth, sex, employment start date, total days of
employment, average exposure level (f/ml), cumulative exposure (f-year/ml), date of last contact,
ICD code for cause of death, indicator variable for mesothelioma death, and date of death if
applicable.
A number of subjects from the full cohort were removed from the analysis reported herein: 412
because the sex was not designated as male; one because the date of last contact was missing;
1,275 subjects because the follow-up period was <5 years; 41 because the number of days
worked was 0 or missing. After these subjects were removed, the cohort consisted of 5,173 men
who were employed at Wittenoom Gorge between 1943 and 1966.
A.19
-------
The concentrations of fibers greater than 5 \im in length as measured by PCM were measured at
various work sites in a survey conducted in 1966. Job category data were obtained from
employment records and supplemented by records from the Perth Chest Clinic and the Western
Australian Mineworkers Relief Fund. The concentration measurements and job category
information were used to estimate the exposure level for each subject in the cohort (de Klerk
et al. 1994). The exposure levels were high with a median of 17.8 (fiber/ml). The durations of
employment were low with a median of 128 days.
There were 251 lung cancer deaths in the cohort. Table A-12 shows the observed, expected, and
predicted lung cancer deaths among the males categorized by cumulative exposure (fiber-
year/ml). The number of expected lung cancer deaths are based on Australian lung cancer
mortality rates. With no allowance for difference between the background lung cancer death
rates among Australia and the members of this cohort (a=l), the fit of the model is poor
(p<0.01). Allowing for difference in the background lung cancer death rates (a variable), the
model provides a reasonably good fit to the data (p=0. 10) and estimates KL=0.0047 (fiber-
year/ml)-1, 90% CI: (0.0017, 0.0087). The hypothesis a=l can be rejected with high confidence
There were 165 mesotheliomas in the cohort. The availability of the raw data from this study
made it possible calculate KM from this study using an "exact" likelihood approach based on
Equation A-3 that did not involve any grouping of data, or use of average values. With this
approach, KM=7.95xlQ-8, 90% CI: (6.97x1 O'8, 9.01xlO'8).
For comparison purposes, KM were also calculated using grouped data and applying Equation
A-4, since this is the method that must be used with most studies. The KM estimate based on the
"exact" analysis was 12% lower than the estimate based upon grouped data.
Regarding uncertainty factors, Fl is assigned a value 2.0 for this study for reasons similar to
those described for Quebec. F2 is assigned a value of 1.0 because measurements were conducted
using PCM so that conversion is unnecessary for this study. All other factors are also assigned a
value of 1.0 because there are no other unique limitations. Thus:
Fl=2.0
F2=1.0
F3 = 1.0
F4L = 1.0
F4M=1.0
These uncertainty factors described earlier, when coupled with the statistical confidence limits,
resulted in the uncertainty ranges for KL and KM and shown in Table A-l and A-2.
Predominant Amosite Exposure
Patterson, N.J. Insulation Factory. Seidman et al. (1986) studied a cohort of 820 men (mostly
white) who worked at an amosite asbestos factory that operated in Patterson, New Jersey from
1941 through 1954. The men began work between 1941 and 1945 and follow-up was through
1982. The follow-up of a worker began 5 years following the beginning of employment.
A.20
-------
Workers who had prior asbestos exposure were not included in the cohort, and follow-up was
stopped when a worker was known to have begun asbestos work elsewhere (6 men). Exposures
were generally brief, as 76% were exposed for
-------
Seidman et al. (1986) discovered 17 deaths from mesothelioma in this population. Table HI of
Seidman et al. categorized mesothelioma deaths and person-years of observation by years since
onset of work. In order to apply the 1986 U.S. EPA mesothelioma model it is necessary to have
estimates of the duration of exposure and level of exposure for each category. Using the
categorization of the members of the cohort by duration of work in Table XXIII of Seidman
et al., it was estimated that the mean duration of work was 1.5 years. Using data from Seidman
et al. Table XIV, an average cumulative exposure was for each category of time from onset of
exposure by weighting exposures according to the expected total number of deaths. These
averages were divided by 1.5 years to obtain the average fiber concentrations in Table A-14.
The estimated exposure levels decrease with time since onset, which is consistent with higher
mortality among more heavily exposed workers.
The 1986 mesothelioma model provided an adequate fit to these data (p=0.35), although it over-
predicted somewhat the number of cases in the highest latency category (>35 years). The
estimate of KM was 3.9xlO-8, 90% CI: (2.6xlQ-8, 5.7x10-*).
Regarding uncertainty factors, Fl is assigned a value of 3.5 for this study because exposure
concentrations were not measured at this facility at all. Rather exposures were estimated (as
described in Lemon et al. [1980]) based on measurements collected at another facility in Tyler,
Texas (see below) that manufactured the same products from the same source of raw materials
using some of the same equipment, which was moved from the Patterson plant to the plant in
Tyler. Because the measurements collected in Tyler were analyzed by PCM, no conversion
factor is required. Thus, F2 is assigned a value of 1.0 for this study. All other factors are also
assigned a factor of 1.0 due to lack of other remarkable limitations. Thus:
Fl=3.5
F2=1.0
F3 = 1.0
F4L = 1.0
F4M-1.0
These uncertainty factors, when coupled with the statistical confidence limits, resulted in the
uncertainty ranges for KL and KM shown in Table A-l and A-2.
Tyler, Texas Insulation Factory. Levin et al. (1998) studied the mortality experience of 1,121
men who formerly worked at a plant in Tyler, Texas that manufactured asbestos pipe insulation.
The plant operated from 1954 through February 1972. The plant used the same raw materials
and some of the same equipment that was used in the Patterson, New Jersey plant that was
studied by Seidman et al. (1986). The asbestos used was amosite from the Transvaal region of
South Africa. The insulation was manufactured from a mixture that contained 90% amosite
asbestos.
Environmental surveys were conducted at the plant in 1967,1970, and 1971, with average fiber
concentrations ranging from 15.9 through 91.4 f/ml. An average exposure of 45 f/ml is assumed
for this plant, which is near the middle of this range obtained in the three surveys. It is also
consistent with average levels assumed for the Patterson, New Jersey plant, which operated
under very similar conditions.
A.22
-------
The cohort consisted of 744 whites, 305 non-white (mostly black), and 72 with missing race
(assumed to be white, based on hiring practices at that time). For the entire cohort, the median
age of first employment was 25 years, and the mean duration of employment was 12.7 months
(range of one day to 17.3 years). Follow-up was through 1993. Death certificates were obtained
for 304 of the 315 men known to be dead. In the mortality analysis only white men were
evaluated and follow-up started 10 years after first employment. After additional exclusions of
men with missing birth dates or missing employment information, the cohort analyzed in the
mortality analysis consisted of 753 former workers, among whom 222 deaths were recorded.
These deaths were compared with those expected based on age, race and sex-specific U.S. rates.
There was an excess of deaths from respiratory cancer (SMR=277, based on 36 deaths, not
including four deaths from mesothelioma). Table A-15 contains observed and expected numbers
of deaths from respiratory cancer, categorized by duration of exposure. Cumulative exposure in
f-y/ml was estimated by multiplying the duration of exposure times the assumed average fiber
level of 45 f/ml. There was an excess of lung cancer deaths in the lowest exposure group
(23 observed, 8.9 expected), and consequently the model with a=l did not fit these data
(p<0.01), and the hypothesis ot=l could be rejected (p<0.01). The KL with a variable was
KL=0.0013, 90% CI: (0, 0.0060). With o=l, K^O.013 (f-y/ml)-1, 90% CI: (0.0055, 0.022).
Four mesotheliomas were reported in this study. However, the data are not presented in a form
that would permit application of the U.S. EPA 1986 mesothelioma model.
Regarding uncertainty factors, Fl is assigned a value of 3.0 for this study because, although
exposure concentrations were measured at this facility, the data are sparse so that only an overall
average concentration for the entire plant could be derived. Because the measurements collected
were analyzed by PCM, no conversion factor is required. Thus, F2 is assigned a value of 1.0 for
this study. All other factors are also assigned a factor of 1.0 due to lack of other remarkable
limitations. Thus:
Fl=3.0
F2 = 1.0
F3 = 1.0
F4L = 1.0
These uncertainty factors, when coupled with the statistical confidence limits, resulted in the
uncertainty range for KL shown in Table A-l.
Predominant Tremolite-Actinolite Exposure
Libby, Montana Vermiculite Mine. Amandus and Wheeler (1987) conducted a retrospective
cohort study of 575 men who were exposed to tremolite-actinolite while working at a vermiculite
mine and mill in Libby, Montana. A dry mill began operation in 1935 and a wet mill began
operating in the same building as the dry mill in 1950 (Amandus et al. 1987).
A total of 376 impinger samples were available that had been collected during 1950-1969,
although only 40 of these were collected prior to 1965. In addition 4,118 PCM samples were
available from the period 1967-1982. Exposure estimates for years later than 1968 were based
A.23
-------
on historical measures of fiber concentrations (f/ml), and those for earlier years were based on
concentrations measured by midget impinger (mppcf) and converted to f/ml assuming a
conversion ratio of 4 f/ml per mppcf. This conversion factor was derived from 336 impinger
samples collected during 1965-1969 and 81 filter samples collected during 1967-1971.
Individual cumulative fiber exposure estimates (f-y/ml) were computed from job-specific
exposure estimates and work histories (Amandus et al. 1987).
The cohort consisted of all men hired prior to 1970 and employed for at least 1 year in either the
mine or the mill. Follow-up was through December 31,1981. The vital statuses of 569 of the
men (99%) were determined and death certificates were obtained for 159 of the 161 who were
deceased.
Smoking information was available for 161 men employed between 1975 and 1982 and with at
least 5 years of tenure. The proportion of these workers who smoked (current or former) was
84% compared to 67% among U.S. white males during the same time period.
A total of 20 deaths from lung cancer were observed (9 expected, SMR=223.2, using U.S. white
males as the comparison population). Table A-16 (based on Amandus and Wheeler 1987, Table
II) shows that the excess occurred mainly in workers whose cumulative exposure exceeded 400
f-y/ml (10 observed, 1.7 expected). The 1986 U.S. EPA lung cancer model fit these data
adequately (p^0.25) both with a=l and a variable, and the hypothesis a=l could not be rejected
(p=0.8). With o=l, KL was estimated as 0.0061 (f-y/ml)-1, 90% CI: (0.0029, 0.010), and with a
variable, KL= 0.0051 (f-y/ml)-1, 90% CI: (0.0011, 0.020).
Amandus and Wheeler (1987) observed 2 deaths from mesothelioma in this cohort. However,
information on these cases was not sufficient to permit application of the 1986 U.S. EPA
mesothelioma model.
For uncertainty factors, Fl is assigned a value of 2.0 for reasons similar to those described for
Quebec. F2 is assigned a value of 1.5 because most early measurements were collected by
midget impinger and the authors report using a conversion factor of 4 derived from temporally
overlapping (but not paired) measurements. All other uncertainty factors were assigned a value
of 1.0. Thus:
Fl=2.0
F2 = 1.5
F3 = 1.0
F4L = 1.0
These uncertainty factors, when coupled with the statistical confidence limits, resulted in the
uncertainty range for KL shown in Table A-l.
McDonald et al. (1986) also conducted a cohort study of workers at the Libby, Montana
vermiculite mine and mill. Their cohort was composed of 406 workers employed prior to 1963
for at least 1 year. Follow-up was until July 1983. Vital status was determined for all but one
man and death certificates were obtained for 163 of the 165 men who had died. Cumulative
exposures (f-y/ml) were estimated for each worker using work histories based on 42 job
A.24
-------
categories, and 1,363 environmental measurements, including samples analyzed by PCM (f/ml)
and by midget impinger (mppcf).
A total of 23 deaths from lung cancer were observed (SMR=303, based on Montana rates).
Table A-17 shows these deaths categorized by cumulative exposure (based on Table 4 of
McDonald et al. 1986). Both the models with a=l and a variable fit these data adequately
(p^ 0.16) although the test of a=l was marginally significant (p=0.11). The estimate of KL with
a=l was 0.011, (f-y/ml)-1, 90% CI: (0.0055, 0.017), and with a variable, KL= 0.0039 (f-y/ml)-1,
90% CI: (0.00067, 0.012).
McDonald et al. (1986) observed 2 deaths from mesothelioma. However, information on these
cases was not sufficient to permit application of the 1986 U.S. EPA mesothelioma model.
Because this study and the Amandus study used virtually the same data and very similar
approaches to analysis, the same values are assigned to uncertainty factors for this study that are
assigned for the Amandus and Wheeler study. These factors, when coupled with the statistical
confidence limits, resulted in the uncertainty range for KL shown in Table A-l.
Exposure to Mixed Fiber Types
British Friction Products Factory. Berry and Newhouse (1983) conducted a mortality study of
13,460 workers in a factory in Britain that manufactured brake blocks, brake and clutch linings,
and other friction materials. Only chrysotile was used at the plant except for two relatively short
periods before 1945 when crocidolite was used in the production of railway blocks.
The cohort studied consisted of all men or women employed at the plant between 1941 and 1977.
Follow-up was to the end of 1979 and the mortality experience was examined after 10 years
from first exposure. Airborne dust measurements were only available from 1967 onward and
these were made using the PCM method. Fiber concentrations in earlier years were estimated by
reproducing earlier working conditions using knowledge of when processes were changed and
exhaust ventilation introduced.
Deaths from all causes were less than expected both prior to 10 years from first employment
(185 observed versus 195.7 expected) and afterward (432 observed versus 450.8 expected).
There was no indication of an effect of employment at the plant upon lung cancer; there were 51
lung cancers >10 years from first employment compared to 47.4 expected. A significant deficit
of gastrointestinal cancers was observed after 10 years from first employment (25 observed
versus 35.8 expected, p=0.04).
A linear exposure response model relating cumulative exposure and lung cancer was fit to case-
control data presented by Berry and Newhouse. The resulting KL was 0.00058 (f-y/ml)"1 and the
95% upper limit was 0.0080 (f-y/ml)"1. This estimate was used as the best estimate of KL, and
the lower confidence bound was assumed to be zero.
A case control study on mesothelioma deaths showed that 8 of the 11 cases had been exposed to
crocidolite and another possibly had intermittent exposure to crocidolite. The other two had
been employed mostly outside the factory and possibly had other occupational exposures to
A.25
-------
asbestos. The case control analysis showed that the distribution of cases and controls in respect
to exposure to crocidolite was quite unlikely assuming no association with crocidolite. This
indicates that some, and possibly all, of the eight mesotheliomas with crocidolite exposure were
related to this exposure. The data were not presented in a form that permitted a quantitative
estimate of mesothelioma risk.
Regarding uncertainty factors, Fl is assigned a value of 2.0 for this study because, although the
manner in which unmeasured exposure was estimated in this study is different than for that
reported for the majority of other studies (see, for example, Quebec), it is unlikely to introduce
greater uncertainty. Rather than extrapolating measured estimates to earlier times based on
expert judgements, judgements were used to simulate earlier conditions at the plant and
exposures were measured directly. Because the measurements collected were analyzed by PCM,
no conversion factor is required. Thus, F2 is assigned a value of 1.0 for this study. An
uncertainty factor F4L=1.5 was included to account for the fact that a was not estimated. F3 was
assigned a factor of 1.0 due to lack of other remarkable limitations. Thus:
Fl=2.0
F2=1.0
F3 = 1.0
F4L = 1.5
These factors, when coupled with the statistical confidence limits, resulted in the uncertainty
range for KL shown in Table A-l.
Ontario Asbestos-Cement Plant. Finkelstein (1984) studied mortality among a group of 535
exposed and 205 unexposed employees of an Ontario asbestos-cement factory who had been
hired before 1960 and who had been employed for at least 1 year. This cohort contained the
cohort studied by Finkelstein (1983) and which required at least 9 years of employment for
membership. Follow-up continued until 1977 or 1981.
The plant produced asbestos cement pipe from 1948, asbestos cement board from 1955-1970,
and manufacture of asbestos insulation materials was added in 1960. Both chrysotile and
crocidolite were used in each batch processed in the pipe process, but only chrysotile was used in
the cement board operation. Crocidolite constituted approximately 20% of the asbestos used in
the pipe process (Ontario Royal Commission 1984).
Fiber concentrations in various work areas and for various epochs were estimated from
membrane filter samples taken after 1969, impinger measurements taken during 1949, 1954,
1956,1957, and semiannually during the 1960s, and information on changes in dust control
methods. Finkelstein judged that the resulting exposure estimates were "probably accurate to
within a factor of three or five." Exposures of maintenance workers were not estimated, and the
exposure response analysis consequently involved only the unexposed workers (N=205) and the
production workers (N=428).
Only 21 deaths from lung cancer were observed among production workers. Based on these
deaths, Finkelstein compared age-standardized lung cancer mortality rates in production workers
after a 20-year latency, categorized into five groups according to their cumulative exposure
A.26
-------
through 18 years from date of first employment (Finkelstein 1984, Table 7). Mortality rates
were standardized with respect to age and latency using the man-years distribution in the cohort
as a whole as the standard. Using similarly standardized mortality rates in Ontario males as the
comparison population, lung cancer rates were elevated in all five categories, and Finkelstein
found a significant exposure-response trend. However, the trend was not monotone, as rates
increased up to the middle exposure category and decreased thereafter (Table A-18).
These data may be put into a form roughly equivalent to the more conventional age-adjusted
comparison of observed and expected lung cancer deaths by dividing the rates in the exposed
group by that of Ontario men. (The rate for unexposed workers was not used because it was
based on only 3 deaths.) The results of this are shown in Table A-18, which also shows the
results of fitting the 1986 U.S. EPA lung cancer model both assuming the Ontario rates were
appropriate for this cohort (fixing the parameter a=l) and not making this assumption (allowing
the parameter a to vary). Neither approach provided an adequate fit to these data (p^O.05) and
the test of a=l was marginally significant (p=0.07). The maximum likelihood estimate of a was
4.26, which seems too large to be due to differences in smoking habits. The KL estimate with
a=l was 0.048 [f-y/ml]'1, 90% CI: (0.028, 0.074). With a allowed to vary the estimate was
KL=0.0029 [f-y/ml]'1, 90% CI: (0, 0.037). The fact that the lower limit was zero indicates that
the exposure-response trend was not significant when the background was allowed to vary.
Based on a "best evidence" classification of cause of death, Finkelstein identified 17 deaths from
mesothelioma among production workers. Table 3 of Finkelstein (1984) gives these
mesotheliomas categorized by years since first exposure. This table also provides the mortality
rate, from which can be calculated the person-years of observation. Finkelstein states that the
average cumulative exposure for production workers was about 60 f-y/ml, but does not provide
information for determining duration and level of exposure separately. CHAP (1983) used an
average exposure of 9 f/ml for a subcohort of production workers, although they provided no
support for this assumption. If this value is assumed to be appropriate for the expanded cohort,
the average duration is estimated as about 60/9=6.7 years. However these values are uncertain.
Table A-19 presents the result of applying the 1986 U.S. EPA mesothelioma model to the
Finkelstein (1984) data based on these assumptions. The mesothelioma model describes these
data adequately (p=0.26) and provides an estimate of KM=18xlO'8, 90% CI: (13xlO'8, 24xlO'8).
Regarding uncertainty, Fl is assigned a value of 4 because Finkelstein indicates that exposure
estimates derived for this study are probably good to within a factor of 3 or 4. Findlestein also
notes that many of the assumptions employed to extrapolate exposures were only weakly
supported by limited, earlier impinger measurements. The source of the conversion factor
employed to link impinger measurements and PCM measurements in this study is unclear.
Therefore a value of 3.0 is assigned to F2. Because data for evaluating mesothelioma incidence
was not provided in a format suitable for deriving confidence intervals, so that some
reconstruction was required, a value of 2.0 is assigned for F4M. All other factors are also
assigned a factor of 1.0 due to lack of other remarkable limitations. Thus:
A.27
-------
Fl=4.0
F2 = 3.0
F3 = 1.0
F4L-1.0
F4M = 2.0
These factors, when coupled with the statistical confidence limits, resulted in the uncertainty
ranges for KL and KM shown in Tables A-l and A-2.
Swedish Asbestos-Cement Plant. Albin et al. (1990) studied workers at a Swedish plant that
operated from 1907 to 1978 and produced various asbestos cement products, including sheets,
shingles, and ventilation pipes. The asbestos handled was mainly chrysotile (>95%).
Crocidolite was used before 1966, but never exceeded 3-4% of the total asbestos. Amosite was
used for a few years in the 1950s but never exceeded 18% of the total asbestos used. Fiber
length classes were the commercial grades 3-7, and all asbestos was milled prior to
incorporation into products.
Impinger and gravimetric dust measurements were available for 1956-1969, and PCM
measurements after 1969. These data, along with information on production and dust control,
were used to estimate exposures for different jobs and periods of time.
The cohort contained 2,898 men and was defined as all male employees who worked for at least
3 months between 1907 and 1977. A reference cohort was composed of 1,233 men who worked
in other industries in the region and who were not known to have worked with asbestos. Vital
status of both groups was determined through 1986. Follow-up of both began after 20 years
from first employment.
Excluding mesothelioma, other respiratory cancers were not significantly increased. Albin et al.
present relative risks of these respiratory cancers and corresponding 95% CIs for three categories
of cumulative exposure (Table A-20), based on Poisson regression with control for age and
calendar year. In order to obtain crude estimates of the range of KL that are consistent with these
data, the 1986 U.S. EPA lung cancer relative risk model was fit, assuming that the Ln (RR) were
normally distributed with fixed variances computed from the reported confidence intervals for
the RR. Although elevated, the RR did not exhibit an exposure response, and the hypothesis a=l
was not rejected (p=0.13). In this analysis KL was not significantly different from zero,
regardless of whether a was fixed at 1.0 or estimated. With a=l the estimate of KL was 0.019
(f-y/ml)-1, 90% CI: (0, 0.065), and KL=0.00067 (f-y/ml)-1, 90% CI: (0, 0.036) with a estimated.
Thirteen mesotheliomas were identified among exposed workers and one in the referent
population, and a significant exposure response was observed with increasing cumulative
exposure. Unfortunately, the mesothelioma data were not presented in a format that would
permit application of the 1986 U.S. EPA mesothelioma model.
Regarding uncertainty, Fl is assigned a value of 4 due to the sparsity of data and the need to
extrapolate. Several assumptions were incorporated into the extrapolations performed that were
A.28
-------
based, among other things, on the scarcity of raw-material asbestos during World War II. All
other factors are also assigned a factor of 1 .0 due to lack of other remarkable limitations. Thus:
Fl=4.0
F2=1.0
F3 = 1.0
F4L = 1.0
These factors, when coupled with the statistical confidence limits, resulted in the uncertainty
range for KL shown in Table A-l.
Belgium Asbestos-Cement Plant. Lacquet et al. (1980) conducted a roentgenologic, asbestosis,
and mortality study in a Belgium asbestos cement factory employing about 2,400 employees that
annually processed about 39,000 tons of asbestos, of which 90% was chrysotile, 8% crocidolite,
and 2% amosite. The mortality study considered male workers who worked in the factory for at
least 12 months during the 15-year period 1963-1977. Apparently no minimal latency was
required before follow-up began.
Fiber counts were available for the years 1970-1976; fiber levels were estimated for as far back
as 1928, but these estimates were considered to be "only good guesses at best." Individual
exposures were estimated in fiber-years from work histories and estimated yearly concentrations
in four work areas.
The incidence of respiratory cancer was very close to that which was expected in a Belgium
population of matched age and sex (Table A-21). The models with ot=l (p=0.51) and a variable
(p=0.39) gave similar results and the hypothesis cc=l was not rejected (p=0.77). With a=l, the
estimate of KL was 0.0 (f-y/ml)'1, 90% CI: (0, 0.0010). With a estimated, KL=6.8xlQ-5 (f-y/ml)'1,
90% CI: (0, 0.0021).
One death was due to pleura! mesothelioma. Unfortunately, the data were not presented in a way
that allowed the estimation o
Regarding uncertainty, Fl is assigned a value of 4 due to the sparsity of data and the need to
extrapolate. Much of the data appear to be based on PCM, so that conversion is not necessary.
All other factors are also assigned a factor of 1.0 due to lack of other remarkable limitations.
Thus:
Fl=4.0
F2=1.0
F3 = 1.0
F4L = 1.0
These factors, when coupled with the statistical confidence limits, resulted in the uncertainty
range for KL shown in Table A-l .
A.29
-------
Retirees from U.S. Asbestos Products Company. Enterline et al. (1986) extended follow-up
through 1980 for a cohort of U.S. retirees from a large asbestos products company that had been
the subject of an earlier report (Henderson and Enterline 1979). Products manufactured by the
company included textiles, cement shingles, sheets, insulation and cement pipe. Exposure was
predominately to chrysotile in most operations, although amosite predominated in insulation
production, and crocidolite in manufacture of cement pipe. Each worker's exposure was
estimated from dust measurements in mppcf obtained from environmental surveys that started in
the mid-1950's and were extrapolated back in time by the company industrial hygienist. No data
are provided for conversion from mppcf to PCM in f/ml. Given the wide range of products
manufactured, this conversion likely varied according to operation. Conversions calculated in
different environments have ranged from 1.4 to 10, the most common value being around 3 f/ml
per mppcf, which has been observed in diverse environments such as mining and textile
manufacture. This value was provisionally applied to this cohort.
The cohort consisted of 1,074 white males who retired from the company during 1941-1967, and
who were exposed to asbestos in production or maintenance jobs. The average duration of
employment was 25 years. Follow-up started at age 65 or at retirement if work continued past
age 65. By the end of follow-up in 1980, 88% were deceased.
Overall, respiratory cancer was significantly increased (SMR=258 in comparison to U.S. rates,
based on 79 observed deaths). Enterline et al. (1986) categorized lung cancer deaths by
cumulative exposure (their Table 4). Results of applying the 1986 U.S. EPA lung cancer model
to these data are shown in Table A-22. Although both the model with a=l and a variable fit the
data adequately (p^0.75), the test of oc=l was not rejected (p=0.24). With oc=l the estimate of KL
was 0.0021 (f-y/ml)-1, 90% CI: (0.0015, 0.0027). With a variable, KL=0.0011 (f-y/ml)'1, 90% CI:
(0.00041,0.0028).
From the death certificates Enterline et al. identified eight deaths from mesothelioma. These
data were not presented in a form that permitted application of the 1986 U.S. EPA mesothelioma
model.
Regarding uncertainty, Fl is assigned a value of 2.0 for this study for reasons similar to those
described for Quebec. Because the manner employed for deriving the conversion factor used to
convert impinger counts to fiber concentrations is not documented, a value of 3.0 is assigned to
F2 for this study. All other factors are also assigned a factor of 1.0 due to lack of other
remarkable limitations. Thus:
Fl=2.0
F2 = 3.0
F3 = 1.0
F4L = 1.0
These factors, when coupled with the statistical confidence limits, resulted in the uncertainty
range for KL shown in Table A-l.
A.3O
-------
U.S. Insulation Applicators. Selikoff and Seidman (1991) reported on follow-up through 1986
of a cohort of 17,800 asbestos insulation applicators that had been followed through 1976 by
Selikoff et al. (1979). The cohort consisted of men enrolled as members of the insulator's union
in the United States and Canada. Deaths were classified both based on the information the death
certificate, and using "best evidence," in which death certificate information was augmented by
clinical data, histopathological material and X-rays.
Based on the composition of insulation material, it seems likely that these workers were exposed
to substantial amounts of chrysotile and amosite. Data on insulator's exposures were reviewed
by Nicholson (1976), who concluded that average exposures of insulation workers in past years
could have ranged 10-15 f/ml and could have been 15-20 f/ml in marine construction. U.S.
EPA (1986) assumed a value of 15 f/ml as an overall average, with an associated 3-fold
uncertainty. This estimate of 15 f/ml will be used provisionally here as well.
The form of the data provided in Selikoff and Seidman (1991) is not particularly suitable for
calculating KL. Table 4 of Selikoff and Seidman (1991) contain observed and expected deaths
from lung cancer (determined from either death certificates or best information) categorized by
years from first exposure (<15, 15-19, 20-24,..., 50+). Death certificate information was
utilized herein to facilitate comparisons with expected deaths (based on the mortality experience
of U.S. white males), which were also based on death certificates. Lung cancer was significantly
increased over expected, except for the category of <15 years from onset of exposure. Selikoff
and Seidman did not provide information on the duration of exposure. The U.S. EPA (1986,
page 90) assumed an average exposure duration of 25 years. Assuming that all workers worked
exactly 25 years and were exposed to 15 f/ml, the data in Table 4 of Selikoff and Seidman can be
used to categorize lung cancer deaths by cumulative exposure lagged 10 years. The result is
shown in Table A-23. The 1986 U.S. EPA lung cancer model provided a reasonable fit to these
data with a variable (p=0.12), but not with a=l (pO.Ol). Also, the hypothesis that a=\ could be
rejected (p<0.01). The estimate of KL with a variable was 0.0018 (f-y/ml)'1, 90% CI: (0.00065,
0.0038). With o=l, 1^=0.0087 (f-y/ml)'1, 90% CI: (0.0081, 0.0093).
Based on best evidence, Selikoff and Seidman (1991) found 458 mesotheliomas in this cohort.
Table A-24 shows these deaths categorized by years from onset (based on Selikoff and Seidman
1991, Tables 5 and 6). Table A-24 also shows the results of fitting the 1986 U.S. EPA
mesothelioma model to these data, assuming, as above, that workers worked for 25 years and
were exposed to 15 f/ml. The 1986 U.S. EPA mesothelioma model provided a poor fit to these
data (p<0.01), as it overestimates by more than a factor of 2 the number of mesothelioma deaths
after 50+ years from first exposure. The estimate of K^ was IJxlO'8, 90% CI: (1.2xlQ-8,
1.4xlQ-8).
Regarding uncertainty, Fl is assigned a value of 4.0 for this study because data employed to
estimate exposure is not facility-specific, but represents general, industry-wide exposure
estimates derived from limited data. F3 is assigned a value of 2 for this study because the study
provides no information on worker histories. F4L is assigned a value of 2 for this study because
the data presented in the study were not provided in a form suitable for fitting the lung cancer
model. Thus, the data had to be partially reconstructed. Other factors are also assigned a factor
of 1.0 due to lack of other remarkable limitations. Thus:
A.31
-------
Fl=4.0
F2=1.0
F3 = 2.0
F4L = 2.0
F4M=1.0
These factors, when coupled with the statistical confidence limits, resulted in the uncertainty
ranges for KL and KM shown in Tables A-l and A-2.
Pennsylvania Textile Plant. McDonald et al. (1983b) report on mortality in an asbestos plant
located near Lancaster, Pennsylvania that produced mainly textiles, but also some friction
materials. About 3,000 to 6,000 tons of chrysotile were processed annually at the plant, which
began operation in the early 1900s. Crocidolite and amosite were used from 1924 onward; about
3-5 tons of raw crocidolite were processed annually and the use of amosite reached a peak of
600 tons during World War II.
The cohort consisted of all men employed for at least 1 month prior to 1959 and who had a valid
record with the Social Security Administration. This group consisted of 4,022 men, of whom
35% had died by the end of follow-up (December 31,1977). Follow-up of each worker was only
begun past 20 years from first employment.
To estimate exposures, McDonald et al. had available reports of surveys conducted by the
Metropolitan Life Insurance Company during the period 1930-1939, Public Health Service
surveys conducted during 1967 and 1970, and company measurements made routinely from
1956 onward. These data were used to estimate by department and year in units of mppcf.
The lung cancer mortality in this cohort exhibited a significant exposure response trend (Table
A-25), which was partially due to a deficit of cancers in the group exposed to <10 mppcf-y (21
with 31.4 expected). A survey of those employed in the plant in 1978 revealed a larger per cent
of nonsmokers (25%) than were found in the other plants studied by these researchers
(McDonald et al. 1983a, 1984), although this finding was based on a sample of only 36 workers.
Regardless of the reason for this shortfall in the number of lung cancers, it appears that the most
appropriate analysis is that in which the background is allowed to vary; this analysis fits the data
well (p>0.7), whereas the analysis which assumes the Pennsylvania rates are appropriate
provides a marginal fit (p=0.08). The hypothesis a=l was rejected (p=0.01). Consequently, the
former analysis is judged to be the most appropriate (allowing the parameter a to vary).
McDonald et al. (1983b) did not provide a factor for converting from mppcf to f/ml. Assuming
that 3 f/ml is equivalent to one mppcf, the resulting estimate of lung cancer potency with a
variable was 0.018 (f-y/ml)'1, 90% CI: (0.0075, 0.045). With ct=l, KL=0.0057 (f-y/ml)'1, 90%
CI: (0.0027, 0.0094).
A diagnosis of mesothelioma was specified on 14 death certificates (ten pleura! and four
peritoneal). Thirty other deaths were given the ICD code 199 (malignant neoplasms of other and
unspecified sites) and the diagnosis given in many of these cases was said to be consistent with
an unrecognized mesothelioma. McDonald et al. (1983b) Table 3 lists the average age at
beginning of employment as 28.92 and the average duration of employment as 9.18 years, and
their Table 1 lists 191, 667, and 534 deaths as occurring before age 45, between 45 and 65, and
A.32
-------
after 65 years of age, respectively. Assuming that Vi of the deaths given the ICD code 199 might
have been due to mesotheliomas, the total number of mesotheliomas in this cohort is estimated to
be 23. Proceeding as in the mesothelioma analysis carried out for the McDonald et al. (1984)
data, the data in Table A-26 were generated. Noting that the age since first exposure categories
in which the mesotheliomas occurred is irrelevant as far as estimating KM is concerned, the
estimate of KM is l.lxlO'8, 90% CI: (0.76x1O'8, LSxlO'8). These estimates are uncertain due to
the uncertainty regarding the number of mesotheliomas in the cohort.
Regarding uncertainty, Fl is assigned a value of 2.0 for this study for reasons similar to those
described for Quebec. Because the manner employed for deriving the conversion factor used to
convert impinger counts to fiber concentrations is not documented, a value of 3.0 is assigned to
F2 for this study. A factor of 2.0 is assigned for F4M because the number of mesotheliomas
observed in this study are reported to be estimates expected to be good to within a factor of 2.
Thus:
Fl=2.0
F2 = 3.0
F3 = 1.0
F4L = 1.0
F4M = 2.0
These factors, when coupled with the statistical confidence limits, resulted in the uncertainty
ranges for KL and KM shown in Tables A-l and A-2.
Rochdale, England Textile Factory. Peto et al. (1985) studied a textile factory in Rochdale,
England that has been the subject of a number of earlier reports (Peto et al. 1977; Peto 1980a,b).
Peto et al. (1985) has the most complete follow-up (through 1983) and emphasizes assessment of
risk. The factory, which began working with asbestos in 1879, used principally chrysotile, but
approximately 5% crocidolite was used between 1932 and 1968.
Quantitative estimates of risk were based on a subgroup of Peto et al. (1985) "principal cohort"
consisting of all men first employed in 1933 or later who had worked in scheduled areas or on
maintenance and had completed 5 years of service by the end of 1974. In the analyses of interest
relating to lung cancer, follow-up only begins 20 years after the beginning of employment and
exposure during the last 5 years of follow-up is not counted.
Routine sampling using a thermal precipitator began at 23 fixed sampling points in 1951.
Comparisons of particle counts and fiber counts taken in 1960 and 1961 were used to convert
between particles/ml and f/ml. Dust levels prior to 1951 were assumed to be the same as those
observed during 1951-1955 for departments for which no major changes had been made. In
departments in which conditions had improved, higher levels were assigned. These levels and
work histories were used to assign individual exposure estimates. A conversion factor of 34
particles/ml per f/ml was determined by comparing average results obtained by the Casella
thermal precipitator (particles/ml) with Ottway long running thermal precipitator (f/ml) at the
same sampling point during 1960 and 1961. However, a conversion factor of 35.3 was used by
Peto et al. (1985) for the sake of consistency with earlier work, and this factor will be used here
as well.
A.33
-------
After 20 years from first employment, there were 93 lung cancer deaths with only 64.6 expected.
Using a lung cancer model essentially the same as the 1986 U.S. EPA model, Peto et al.
estimated KL=0.0054 (f-y/ml)-1 for the entire cohort, and KL=0.015 (f-y/ml)-1 when the analysis
was restricted to men first employed in 1951 or later. Peto et al. felt that the most plausible
explanation for this difference was that it was largely due to chance and also possibly to the
chance that exposure to the most carcinogenic fibers was not reduced as much as changes in
particle counts from 1951 to 1960 would suggest.
Table A-27 displays the exposure response data based on men first employed in 1933 or later for
lung cancer based on shows that the excess occurred mainly in workers whose cumulative
exposure exceeded 400 f-y/ml (10 observed, 1.7 expected). The 1986 U.S. EPA lung cancer
model fit these data adequately (p^0.63) both with a=l and a variable, and the hypothesis a=l
could not be rejected (p=0.57). With a=l, K^ was estimated as 0.0052 (f-y/ml)-1, 90% CI:
(0.0028, 0.0079), and with a variable, KL=0.0041 (f-y/ml)-1, 90% CI: (0.0012, 0.0087).
Ten mesotheliomas were observed in the cohort used by Peto et al. for quantitative analysis (an
1 1th case who was exposed for 4 months and died 4 years later was omitted because the short
latency made it unlikely that this case was related to exposure at the factory). Observed
mesotheliomas and corresponding person-years of observation by duration of service and years
since first employment (Peto et al. 1985, Table 8) are shown in Table A-28. An overall average
exposure was estimated by applying the Peto mesothelioma model to the data in Table A-28 with
a single exposure estimate selecting the value that gave the smallest least squares fit of this
model to the mesothelioma data. The fitting was carried out both unweighted and by weighting
by the person years, with resulting estimates of 360 and 322 particles/ml, respectively; the latter
value was the one selected. Using the conversion factor of 35.3 particles/ml per f/ml, the
estimated average exposure is 322/35.2=9.1 f/ml. The 1986 U.S. EPA mesothelioma model fit
these data well, and the resulting estimate of mesothelioma potency (Table A-28) was
8, 90% CI: (0.74x10-8, 2.1xlQ-8).
Regarding uncertainty, Fl is assigned a value of 2.0 for this study for reasons similar to those
described for Quebec. Because a conversion factor was derived for measurements collected
using Otway long-running thermal precipitators and PCM measurements based on measurements
of each collected under similar conditions (but not side-by-side), a value of 2 is assigned to F2.
Thus:
Fl=2.0
F2 = 2.0
F3 = 1.0
F4L = 1.0
F4M = 1.0
These factors, when coupled with the statistical confidence limits, resulted in the uncertainty
ranges for KL and KM shown in Tables A-l and A-2.
A.34
-------
REFERENCES
Albin M; Jakobsson K; Attewell R; Johansson L; Welinder H. Mortality and Cancer Morbidity
in Cohorts of Asbestos Cement Workers and Referents. British Journal of Industrial Medicine.
79(9):602-610. September. 1990.
Amandus HE; Wheeler R. The Morbidity and Mortality of Vermiculite Miners and Millers
Exposed to Tremolite-Actinolite: Part II. Mortality. American Journal of Industrial Medicine
11:15-26. 1987.
Amandus HE; Wheeler R; Jankovic J; Tucker J. The Morbidity and Mortality of Vermiculite
Miners and Millers Exposed to Tremolite-Actinolite: Part I. Exposure Estimates. American
Journal of Industrial Medicine. 11:1-14. 1987.
Armstrong BK; de Klerk NH; Musk AW; Hobbs MST. Mortality in Miners and Millers of
Crocidolite in Western Australia. British Journal of Industrial Medicine. 45:5-13. 1988.
Berry G; Newhouse ML. Mortality of Workers Manufacturing Friction Materials Using
Asbestos. British Journal of Industrial Medicine. 40:1-7. 1983.
CHAP (Chronic Hazard Advisory Panel on Asbestos). Report to the U.S. Consumer Product
Safety Commission. July. 1983.
Cox D; Oakes DV. Analysis of Survival Data. Cox DR; Hinkley DV (eds.). Chapman and Hall,
London. 1984.
de Klerk NH; Musk AW; Armstrong BK; Hobbs MST. Diseases in Miners and Millers of
Crocidolite from Wittenoom, Western Australia: A Further Follow-up to December 1986.
Annals of Occupational Hygiene. 38(Suppl l):647-655. 1994.
Dement JM. Estimation of Dose and Evaluation of Dose-Response in a Retrospective Cohort
Mortality Study of Chrysotile Asbestos Textile Workers. Ph.D. Thesis. The University of North
Carolina at Chapel Hill. 1980.
Dement JM; Brown DP. Cohort Mortality and Case-Control Studies of White Male Chrysotile
Asbestos Textile Workers. Journal of Clean Technology, Environmental Toxicology, and
Occupational Medicine. 7:1052-1062. 1998.
Dement JM; Harris RL; Symons MJ; Shy CM. Estimates of Dose-Response for Respiratory
Cancer Among Chrysotile Asbestos Textile Workers. Annals Occupational Hygiene.
26(l-4):869-887. 1982.
Dement JM; Harris RL; Symons MJ; Shy CM. Exposures and Mortality Among Chrysotile
Workers. Part I: Exposure Estimates. American Journal of Industrial Medicine. 4:399-419.
1983a.
A.35
-------
Dement JM; Harris RL; Symons MJ; Shy CM. Exposures and Mortality Among Chrysotile
Workers. Part II: Mortality. American Journal of Industrial Medicine. 4:421-433. 1983b.
Dement JM; Brown DP; Okun A. Follow-up Study of Chrysotile Asbestos Textile Workers:
Cohort Mortality and Case-Control Analysis. American Journal of Industrial Medicine.
26:431-447. 1994.
Enterline PE; Harley J; Henderson V. Asbestos and Cancer — A Cohort Followed to Death.
Graduate School of Public Health, University of Pittsburgh. 1986.
Finkelstein MM. Mortality Among Long-Term Employees of an Ontario Asbestos-Cement
Factory. British Journal of Industrial Medicine. 40:138-144. 1983.
Finkelstein MM. Mortality Among Employees of an Ontario Asbestos-Cement Factory.
American Review of Respiratory Disease. 129:754-761. 1984.
Hammond EC; Selikoff IJ; Seidman H. Asbestos Exposure, Cigarette Smoking and Death Rates.
Annals New York Academy of Sciences. 330:473-490. 1979.
Henderson VL; Enterline PE. Asbestos Exposure: Factors Associated with Excess Cancer and
Respiratory Disease Mortality. Annals New York Academy of Sciences. 330:117-126. 1979.
Hughes JM; Weill H. Asbestos Exposure: Quantitative Assessment of Risk. American Review
of Respiratory Disease. 133:5-13. 1986.
Hughes JM; Weill H; Hammad YY. Mortality of Workers Employed at Two Asbestos Cement
Plants. British Journal of Industrial Medicine. 44:161-174. 1987.
Lacquet LM; VanderLinden L; Lepoutre J. Roentgenographic Lung Changes, Asbestosis and
Mortality in a Belgian Asbestos-Cement Factory. In Biological Effects of Mineral Fibres,
Wagner JC (ed.). IARC Sci Publ. pp. 783-793. 1980.
Lemon RA; Dement JM; Wagoner JK. Epidemiology of asbestos-related diseases.
Environmental Health Perspective. 34:1-11. 1980.
Levin JL; McLarty JW; Hurst GA; Smith AN; Frank AL. Tyler Asbestos Workers: Mortality
Experience in a Cohort Exposed to Amosite. Occupational and Environmental Medicine.
55:155-160. 1998.
Liddell FDK. Unpublished raw mesothelioma data provided to Dr. Wayne Berman by Dr. FDK
Liddell from multiple studies of the 1891-1920 Birth Cohort of Quebec Chrysotile Miners and
Millers most recently described in Liddell et al. 1997. 2001.
Liddell FDK; McDonald AD; McDonald JC. The 1891-1920 Birth Cohort of Quebec Chrysotile
Miners and Millers: Development From 1904 and Mortality to 1992. Annals of Occupational
Hygiene. 41:13-36. 1997.
A.36
-------
McDonald JC; Liddell FDK; Gibbs GW; Eyssen GE; McDonald AD. Dust Exposure and
Mortality in Chrysotile Mining, 1910-1975. British Journal of Industrial Medicine. 37:11-24.
1980a.
McDonald JC; Gibbs GW; Liddell FDK. Chrysotile Fibre Concentration and Lung Cancer
Mortality: A Preliminary Report. In Biological Effects of Mineral Fibres. Wagner JC (ed).
IARC Scientific Publications, pp. 811-817. 1980b.
McDonald AD; Fry JS; Wooley AJ; McDonald JC. Dust Exposure and Mortality in an
American Chrysotile Textile Plant. British Journal of Industrial Medicine. 39:361-367. 1983a.
McDonald AD; Fry JS; Woolley AJ; McDonald JC. Dust Exposure and Mortality in an
American Factory Using Chrysotile, Amosite, and Crocidolite in Mainly Textile Manufacture.
British Journal of Industrial Medicine. 40:368-374. 1983b.
McDonald AD; Fry JS; Woolley AJ; McDonald JC. Dust Exposure and Mortality in an
American Chrysotile Asbestos Friction Products Plant. British Journal of Indus trial Medicine.
41:151-157. 1984.
McDonald JC; McDonald AD; Armstrong B; Sebastien P. Cohort Study of Mortality of
Vermiculite Miners Exposed to Tremolite. British Journal of Industrial Medicine. 43:436-444.
1986.
McDonald JC; Liddell FDK; Dufresne A; McDonald AD. The 1891-1920 Birth Cohort of
Quebec Chrysotile Miners and Millers: Mortality 1976-1988. British Journal of Industrial
Medicine. 50:1073-1081. 1993.
Nicholson WJ. Part III. Recent Approaches to the Control of Carcinogenic Exposures. Case
Study 1: Asbestos - The TLV Approach. Annals New York Academy of Science. 271:152-169.
1976.
Nicholson WJ; Selikoff IJ; Seidman H; Lilis R; Formby P. Long-Term Mortality Experience of
Chrysotile Miners and Millers in Thetford Mines, Quebec. Annals New York Academy of
Sciences. 330:11-21. 1979.
Ontario Royal Commission. Report of the Royal Commission on Matters of Health and Safety
Arising form the Use of Asbestos in Ontario. Volume 3. 1984.
Peto J. Lung Cancer Mortality in Relation to Measured Dust Levels in an Asbestos Textile
Factory. In Biological Effects of Mineral Fibres. Wagner JC (ed.). IARC Scientific
Publications, pp. 829-836. 1980a.
Peto J. The Incidence of Pleural Mesothelioma in Chrysotile Asbestos Textile Workers. In
Biological Effects of Mineral Fibres. Wagner JC (ed.). IARC Scientific Publications, pp.
703-711. 1980b.
A.37
-------
Peto J; Doll R; Howard SV; Kinlen LJ; Lewinsohn, HC. A Mortality Study Among Workers in
an English Asbestos Factory. British Journal of Industrial Medicine. 34:169-173. 1977.
Peto J; Seidman H; SelikoffU. Mesothelioma Mortality in Asbestos Workers: Implications for
Models of Carcinogenesis and Risk Assessment. British Journal of Cancer. 45:124-135. 1982.
Peto J; Doll R; Hermon C; Binns W; Clayton R; Goffe T. Relationship of Mortality to Measures
of Environmental Asbestos Pollution in an Asbestos Textile Factory. Annals of Occupational
Hygiene. 29(3):305-355. 1985.
Piolatto G; Negri E; LaVecchia C; Pira E; Decarli A; Peto J. An Update of Cancer Mortality
Among Chrysotile Asbestos Miners in Balangero, Northern Italy. British Journal of Industrial
Medicine. 47:810-814. 1990.
Rubino GF; Piolatto GW; Newhouse ML; Scansetti G; Aresini GA; Murray R. Mortality of
Chrysotile Asbestos Workers at the Balangero Mine, Northern Italy. British Journal of
Industrial Medicine. 36:187-194. 1979.
Seidman H. Short-Term Asbestos Work Exposure and Long-Term Observation -- July 1984
Update. Department of Epidemiology, American Cancer Society. 1984.
Seidman H; SelikoffU; Hammond EC. Short-Term Asbestos Work Exposure and Long-Term
Observation. Annals New York Academy of Sciences. 330:61-89. 1979.
Seidman H; SelikoffU; Gelb SK. Mortality Experience of Amosite Asbestos Factory Workers:
Dose-Response Relationships 5 to 40 Years After Onset of Short-Term Work Exposure.
American Journal of Industrial Medicine. 10(5/6):479-514.1986.
SelikoffU; Seidman H. Asbestos-Associated Deaths among Insulation Workers in the United
States and Canada, 1967-1987. Annals of the New York Academy of Sciences. 643:1-14. 1991.
SelikoffU; Hammond EC; Seidman H. Mortality Experience of Insulation Workers in the
United States and Canada 1943-1976. Annals New York Academy of Sciences. 330:91-116.
1979.
U.S. EPA (U.S. Environmental Protection Agency). Airborne Asbestos Health Assessment
Update. Report 600/8-84-003F, U.S. Environmental Protection Agency. 1986.
Venzon D; Moolgavkar S. A Method for Computing Profile-likelihood-based Confidence
Intervals. Applied Statistics. 37:87-94. 1988.
Weill H. 1994. Cancer Mortality in Chrysotile Mining and Milling: Exposure-Response.
Asbestos-Cement. Annals of Occupational Hygiene. 38(4):412. 1994.
Weill H; Hughes J; Waggenspack C. Influence of Dose and Fibre Type on Respiratory
Malignancy Risk in Asbestos Cement Manufacturing. American Review of Respiratory Disease.
120:345-354. 1979.
A.38
-------
Table A-l. Lung Cancer Exposure-Response Coefficients (KjJ Derived from Various Epidemiological Studies
Fiber Type
Chrysotile
Crocidolile
Amosite
Tremolite
Mixed
Operation
Mining and
Milling
Friction
Products
Cement
Manufacture
Textiles
Mining and
Milling
Insulation
Manufacture
Vermiculite
Mines and Mills
Friction
Products
Cohort
Quebec mines
and mills
Italian mine
and mill
Connecticut
plant
New Orleans
plants
South
Carolina plant
Wittenoom
Patterson, NJ
factory
Tyler, Texas
factory
Libby,
Montana
British factory
EPA (1986)
KL*100 Reference
0.06 McDonald
etal. 1980b
0.17 Nicholson et al.
1979
0.081 Piolatto et al.
1990
0.01 McDonald
etal. 1984
2.8 Dement et al.
1983b
2.5 McDonald
et al. 1983a
4.3 Seidman 1984
0.058 Berry and
Newhouse
1983
This Update
KL*100
0.029
0.051
0
0.25
2.1
1
0.47
1.1
0.13
0.51
0.39
0.058
90%
Confidence
Interval
(0.019, 0.041)
(0, 0.57)
(0,0.17)
(0, 0.66)
(1.2,3.4)
(0.44, 2.5)
(0.17,0.87)
(0.58, 1.9)
(0, 0.6)
(0.11,2.0)
(0.067, 1.2)
(0, 0.8)
Uncertainty
Interval3
(0.0085,
0.091)
(0, 1.1)
(0, 0.62)
(0, 1.5)
(0.81,5.1)
(0.22, 4.9)
(0.084, 1.7)
(0.17, 6.6)
(0, 1.8)
(0.049, 4.4)
(0.03, 2.8)
(0, 1.8)
Reference
Liddell et al.
1997
Piolatto et al.
1990
McDonald
et al. 1984
Hughes et al.
1987
Dement et al.
1994b
McDonald
1983a
de Klerk et al.
1994C
Seidman et al.
1986
Levin et al.
1998
Amandus and
Wheeler 1987
McDonald
et al. 1986
Berry and
Newhouse
1983
A.39
-------
Table A-l. Lung Cancer Exposure-Response Coefficients (KL) Derived from Various Epidemiological Studies (continued)
Fiber Type Operation
Cement
Manufacture
Factory workers
Insulation
Application
Textiles
Cohort
Ontario
factory
New Orleans
plants
Swedish plant
Belgium
factory
US. retirees
U.S.
insulation
workers
Pennsylvania
plant
Rochedale
plant
EPA (1986)
KL*100
4.8
0.53
0.49
0.75
1.4
1.1
Reference
Finkelstein
1983
Weill 1979,
1994
Henderson and
Enterline 1979
Seilkoffetal.
1979
McDonald
etal. 1983b
Peto 1980a
This Update
KL*100
0.29
0.25
0.067
0.0068
0.11
0.18
1.8
0.41
90%
Confidence
Interval
(0, 3.7)
(0, 0.66)
(0, 3.6)
(0, 0.21)
(0.041, 0.28)
(0.065, 0.38)
(0.75, 4.5)
(0.12,0.87)
Uncertainty
Interval"
(0, 22)
(0, 1.5)
(0, 14)
(0, 0.84)
(0.011, 1.0)
(0.012,2.1)
(0.2, 16)
(0.046, 2.3)
Reference
Finkelstein
1984
Hughes et al.
1987
Albin et al.
1990
Laquet et al.
1980
Enterline et al.
1986
Seilkoffand
Seidman 1991
McDonald
etal. 1983b
Peto et al.
1985
"Uncertainty Interval formed by combining 90% confidence interval with uncertainty factors in Table A-3.
bWith supplemental raw data from Terri Schnorr (NIOSH) and Dement
cWith supplemental unpublished raw data with follow-up through 2001
A.4O
-------
Table A-2. Mesothelimoa Exposure-Response Coefficients (KM) Derived from Various Epidemiological Studies
Fiber Type
Chrysotile
Crocidolile
Amosite
Mixed
Operation
Mining and Milling
Friction Products
Cement Manufacture
Textiles
Mining and Milling
Insulation
Manufacture
Cement Manufacture
Factory Workers
Insulation
Application
Textiles
EPA
(1986)
Cohort KM* 100 Reference
Asbestos,
Quebec
Thedford Mines
Connecticut
plant
New Orleans
plant
South Carolina
plant
Wittenoom
Patterson, NJ 3.2 Seidmanl984
factory
Ontario factory 12 Finkelstein 1983
New Orleans
plant
Asbestos,
Quebec
U.S. insulation 1.5 Seilkoff et al.
workers 1979
Pennsylvania
plant
Rochedale plant 1 Peto 1 980; Peto
etal. 1982
This
Update
KM' 100
0.013
0.021
0
0.2
0.25
0.088
7.9
3.9
18
0.3
0.092
1.3
1.1
1.3
90%
Confidence
Interval
(0.0068, 0.022)
(0.014, 0.029)
(0,0.12)
—
(0.034, 0.79)
(0.0093, 0.32)
(7,9)
(2.6, 5.7)
(13, 24)
—
(0.04,0.18)
(1.2,1.4)
(0.76, 1.5)
(0.74,2.1)
Uncertainty
Interval"
(0.003, 0.049)
(0.0065, 0.065)
(0, 0.65)
(0.033,1.2)
(0.023,1.2)
(0.0025,1.2)
(3.5, 18)
(0.74, 20)
(2, 160)
(0.089, 1)
(0.018, 0.39)
(0.25, 6.5)
(0.17,6.6)
(0.28, 5.6)
Reference
Liddelletal. 1997b
Liddelletal. 1997"
McDonald et al. 1984
Hughes etal. 1987
Dement etal. 1994C
McDonald et al.
1983a
de Klerk etal. 1994"
Seidmanetal. 1986
Finkelstein 1984
Hughes etal. 1987
Liddelletal. 1997b
Seilkoff and Seidman
1991
McDonald et al.
1983b
Peto etal. 1985
'Uncertainty Interval formed by combining 90% confidence interval with uncertainty factors in Table A-3.
"With supplemental raw data from Liddell
°With supplemental raw data from Terri Schnorr (NIOSH) with Dement
dWith supplemental unpublished raw data with follow-up through 2001
A.41
-------
Table A-3. Uncertainty Factors Used to Develop Uncertainty Intervals for Exposure-Response Coefficients (KL's and KM's)
Fiber Type
Operation
Chrysotile
Mining and
Milling
Friction
Products
Cement
Manufacture
Textiles
Crocidolile
Mining and
Milling
Amosite
Insulation
Manufacture
Uncertainty Factors for Estimating
Exposure
Combined
Special Uncertainty Factors Uncertainty
Uncertainty Uncertainty Uncertainty Uncertainty Uncertainty
Estimating Converting Assigning for Special for Special
Exposure to PCM Job Lung Cancer Mesothelioma
Concentrations F2 Histories Limitations Limitations Lung Meso-
Cohort Fl F3 F4 F4M Cancer thelioma Reference
Quebec 2 1.5
Asbestos, 2 1.5
Quebec
Thedford 2 1.5
Mines
Italian mine 2
and mill
Connecticut 2 3
plant
New Orleans 2 1.5
plant
South 1.5
Carolina plant
South 2
Carolina plant
Wittenoom 2
Patterson, NJ 3.5
factory
2.2 2.2a Liddelletal. 1997
NR 2.2" Liddelletal. 1997
NR 2.2" Liddelletal. 1997
2.0 ND Piolattoetal. 1990
3 3.7 5.5 McDonald etal. 1984
5 2.2 6.0 Hughes etal. 1987
1.5 1.5 Dement et al. 1994b
3 2.0 3.7 McDonald et al. .
1983a
2.0 2.0 de Klerk et al. 1994C
3.5 3.5 Seidman etal. 1986
A.42
-------
Table A-3. Uncertainty Factors Used to Develop Uncertainty Intervals for Exposure-Response Coefficients (KL's and KM's) (continued)
Uncertainty Factors for Estimating Combined
Exposure Special Uncertainty Factors Uncertainty
Fiber Type
Operation
Tremolite
Vermiculite
Mines and
Mills
Mixed
Friction
Products
Cement
Manufacture
Factory
Workers
Insulation
Application
Uncertainty Uncertainty Uncertainty Uncertainty Uncertainty
Estimating Converting Assigning for Special for Special
Exposure to PCM Job Lung Cancer Mesothelioma
Concentrations F2 Histories Limitations Limitations Lung Meso-
Cohort Fl F3 F4 F4M Cancer thelioma
Tyler, Texas 3
factory
Libby, 2
Montana
Libby, 2
Montana
British 2
factory
Ontario 4
factory
New Orleans 2
plant
Swedish plant 4
Belgium 4
factory
U.S. retirees 2
Asbestos, 2
Quebec
U.S. 4
insulation
workers
3.0 ND
1.5 2.2 ND
1.5 2.2 ND
1.5 2.2 ND
3 2 5.9 6.7
1.5 2.5 2.2 3.4
4.0 ND
4.0 ND
3 3.7 ND
1.5 NR 2.2a
2 2 5.5 4.7
Reference
Levin etal. 1998
Amandus and
Wheeler 1987
McDonald et al. 1986
Berry and Newhouse
1983
Finkelstein 1984
Hughes etal. 1987
Albinetal. 1990
Laquetetal. 1980
Enterline et al. 1986
Liddelletal. 1997
Seilkoff and Seidman
1991
A.43
-------
Table A-3. Uncertainty Factors Used to Develop Uncertainty Intervals for Exposure-Response Coefficients (KL's and KM's) (continued)
Uncertainty Factors for Estimating Combined
Exposure
Special Uncertainty Factors
Uncertainty
Uncertainty Uncertainty Uncertainty Uncertainty Uncertainty
Estimating Converting Assigning for Special for Special
Fiber Type
Operation
Textiles
Cohort
Pennsylvania
plant
Rochedale
plant
Exposure
Concentrations
Fl
2
2
to PCM
F2
3
2
Job Lung Cancer Mesothelioma
Histories Limitations Limitations
F3 F4 F4M
2
Lung
Cancer
3.7
2.7
Meso-
thelioma
4.4
2.7
Reference
McDonald et al.
1983b
Petoetal. 1985
"With supplemental raw data from Liddell et al. 1997 for mesothelioma
"With supplemental raw data from Terri Schnorr (NIOSH) with Dement
'With supplemental unpublished raw data with follow-up through 2001
NOTES:
Values for uncertainty factors not listed in the table are assumed to be equal to one.
A description of the manner in which each of the values presented in this table was assigned is presented under the descriptions of individual studies in Appendix A.
NR means no raw data. These are the data sets from Quebec for which we had access only to raw data for mesothelioma. Thus, lung cancer rates could not be determined.
NR means not determined. These are the data sets for which mesothelioma data were either lacking or were unuseable.
A.44
-------
Table A-4
Cancer of Lung, Trachea, or Bronchus by Cumulative Exposure
Level among Workers in Quebec Chrysotile Mines and Mills
Liddelletal. (1997)
mpcf-yr
Range Mean
[0,3)
[3, 10)
[10,30)
[30, 60)
[60, 100)
[100,200)
[200, 300)
[300, 400)
[400, 1000)
>= 1000
1.5
6.5
20
45
80
150
250
350
700
1500
(f-yr)/ml
Mean
4.71
20.41
62.8
141.3
251.2
471
785
1099
2198
4710
SMR
1.12
1.27
1.03
1.32
1.45
1.27
1.1
1.46
1.84
2.97
Expected
67.0
50.4
59.2
45.5
42.1
52.8
31.8
19.9
47.8
15.8
Observed
75
64
61
60
61
67
35
29
88
47
Predicted
a = 1 a = 1.15
67.1
50.8
60.8
48.1
46.4
63.0
42.1
28.8
91.1
46.5
76.9
58.2
69.2
54.3
51.8
68.8
44.8
30.1
89.9
43.0
Totals
(90% Confidence Interval)
Goodness of Fit P-value
Test of H: a = 1 P-value
a = 1 (fixed)
0.041
(0.032, 0.051)
0.18
0.014
432.2 587
a = 1.15(MLE)
0.029
(0.019,0.041)
0.58
544.7
587.0
A.45
-------
Table A-5
Lung Cancer Mortality among Chrysotile Asbestos
Miners in Balangero, Northern Italy
Piolatto et al. (1990)
f-y/m I
Range Mean
< 100 50
[100,400) 250
>= 400 600
Totals
KL* 100
(90% Confidence Inte
Goodness of Fit
Test of H0: a = 1
Observed
4
8
10
22
rval)
P-value
P-value
Expected
5.1
6.1
8.7
19.9
a = 1 (fixed)
0.035
(0, 0.15)
0.75
0.88
Pred
a = 1
5.2
6.6
10.5
22.3
a = 0
icted
a = 0.937
4.9
6.4
10.7
22.0
.937 (MLE)
0.051
(0, 0.57)
0.45
A.46
-------
Table A-6
Lung Cancer Mortality among Workers in a Chrysotile
Asbestos Friction Products Plant in Connecticut
McDonald et al. (1984)
m pcf-yr
Range
< 10
[ 10 ,20 )
[20 ,40 )
[40 , 80 )
> = 80
Totals
KL* 100
(f-yr)/ml
I SMR E
xpected Observed
Mean Mean
5
15
30
60
110
(90% Confidence
Goodness.
Test of H0:
of Fit
a = 1
15
45
90
180
330
Interval)
P-value
P-value
167.4
101.7
105.4
162.8
55.22
a = 1 (fixed)
0.19
(0, 0.61)
0.01
0.001
32.9
5.9
4.7
3.7
1.8
49.0
55
6
5
6
1
73
a = 1.49 (MLE)
0
(0, 0.17)
0.28
Predicted
a = 1
33.8
6.4
5.5
4.9
2.9
53.6
a = 1.49
49.0
8.8
7.1
5.5
2.7
73.0
A.47
-------
Table A-7
Mesothelioma Mortality among Connecticut Friction Product Plant Workers
McDonald et al. (1984)
Years After
Range
[14,34)
>=34
First Exposure
Mean
22
39
Duration of
Exposure
8.04
8.04
f/ml
5.52
5.52
Person
Years
37742
9420
Observed
0
0
Predicted
0.0
0.0
Totals
47162
0
0.0
KM MO8
(90% Confidence Interval)
Goodness of Fit P-Value
0
(0,0.12)
1.00
A.48
-------
Table A-8
Lung Cancer Mortality among Workers Employed in Two Asbestos
Cement Manufacturing Plants in New Orleans, Louisiana
Hughes et al. (1987)
m pcf-yr
Range Mean
Plant 1 Em ployees
( < 6 )
( 6 - 24 )
( 25 - 49 )
( 50 - 99 )
( >= 100) 1
Plant2 Employees
( < 6 )
( 6 - 24 )
( 25 - 49 )
( 50 - 99 )
( >= 100) 1
Totals
KL * 100
4
13
35
74
83
3
12
36
71
64
(90% Confidence Inte
Good ness of F it
Test of H0: a = 1
(f-yr)/ml Ob
Mean
5.6
1 8.2
49
103.6
256.2
4.2
16.8
50.4
99.4
229.6
a =
rval) (0.
P-value
P-value
served
3
9
2
3
5
20
19
12
10
12
95
1 (fixed
0.4
15, 0.7)
0.44
0.18
Expected
2.9
8
3.7
3.8
4.1
18.9
14.5
6
5.5
5.2
72.6
) a
P red icted
a = 1 a =
3.0
8.6
4.4
5.4
8.3
19.2
15.5
7.2
7.7
9.9
89.0
= 1.14 (M LE)
0.25
(0, 0.66)
0.42
1.14
3.4
9.6
4.8
5.5
7.7
21 .8
17.3
7.7
7.9
9.4
95.0
A.49
-------
Table A-9
Lung Cancer Mortality by Cumulative Exposure among Chrysotile
Asbestos Textile Workers in Charleston, South Carolina
Dement et al. (1994) -- based on raw data provided
by Terri Schnorr (NIOSH)
f-y/ml Observed Expected Predicted
Range Mean a = 1 a = 1.22
< 0.8
[0.8, 2 )
[2,4)
[4, 10)
[10,35)
[ 35, 85 )
>= 85
Totals
KL* 100
0.14
1:33
2.9
6.53
19.35
54.73
143.35
1
7
11
12
19
19
21
33
22
a
(90% Confidence Interval)
Goodness
Test of H0:
of Fit P-value
a = 1 P-value
6.8
9.3
9.2
11
11.9
8.5
6.6
63.3
= 1 (fixed)
2.8
(2.1,3.7)
0.81
0.19
6.8
9.7
10.0
13.0
18.4
21.7
33.5
113.1
a
8.3
11.6
11.8
15.1
20.2
22.1
31.9
121.1
= 1.22 (MLE)
2.1
(1.2,3.4)
0.93
A.5O
-------
Table A-10
Lung Cancer Mortality among Workers in a Chrysotile
Asbestos Textiles Plant in South Carolina
McDonald et al. (1983a)
mpcf-yr
Range Mean
< 10
[ 10 - 19 ]
[20 - 39 ]
[ 40 - 79 ]
>= 80
5
15
30
60
110
(f-yr)/ml
Mean
30
90
180
360
660
SMR
143.1
182.7
304.2
419.5
1031.9
Expected
21.7
2.7
2.6
1.7
0.8
Observed
31
5
8
7
8
Predicted
a = 1 a = 1.07
29.2
5.6
8.1
8.6
6.7
30.4
5.7
8.0
8.4
6.5
Totals
KL * 100
(90% Confidence Interval)
Goodness of Fit P-value
Test of H0: a = 1 P-value
a = 1 (fixed)
1.2
(0.75, 1.6)
0.95
0.80
29.5 59
a = 1.07 (MLE)
1
(0.44, 2.5)
0.88
58.1
59.0
A.51
-------
Table A-11
Mesothelioma Mortality among South Carolina Textile Plant Workers
McDonald et al. (1983a)
Years After First
Range
( 19-39 )
(>39)
Exposure
Mean
28
44
Duration
10
10
f/ml
5.4
5.4
Person
Years
26280
2787
Observed
0
1
Predicted
0.7
0.3
Totals
29067
1.0
KM* 10"
(90% Confidence Interval)
Goodness of Fit P-Value
0.088
(0.0093, 0.32)
0.14
A.52
-------
Table A-12
Lung Cancer Mortality Among
Asbestos Workers in Wittenoom, Australia
DeKlerk et al. (1994) -- supplemented with unpublished raw
data with follow-up through 2001
(f-yr)/ml
Range Average
0
0-0.4
0.4 - 1
1 - 2.3
2.3- 4.5
4.5 -8.5
8.5 - 16
16 -28
28 -60
60 +
Totals
0
0.19
0.69
1.59
3.27
6.19
11.81
21.53
41.07
142.28
Expected
4.6
7.9
8.2
11.6
12.9
14.3
13.2
9.2
11.6
11.6
105.1
KL* 100
(90% Confidence Interval)
Goodness of Fit P-value
Test of H0:
a = 1
P-value
Observed
5
27
11
22
28
38
31
21
25
43
251
a = 1 (fixed)
2.7
(2,3.5)
< 0.001
< 0.001
Predicted
a = 1 a = 2.13
4.6
8.0
8.3
12.1
14.0
16.7
17.4
14.5
24.5
56.5
176.6
9.8
17.0
17.6
24.9
27.9
31.4
29.8
21.6
29.6
41.6
251.0
a = 2.13 (MLE)
0.47
(0.17, 0.87)
0.10
A.53
-------
Table A-13
Lung Cancer Mortality by Cumulative Exposure among Amosite
Asbestos Factory Workers in Paterson, New Jersey
Seidman et al. (1986)
(f-yr)/ml
Range
<6
6-12
12-25
25-50
50-100
100-150
150-250
250+
Totals
KL * 100
Average
3
9
18.5
37.5
75
125
200
375
SMR
2.8
4.2
4.4
4.7
7.1
6.0
11.4
16.0
(90% Confidence Interval)
Goodness
Test of H0:
of Fit
a = 1
P-value
P-value
Expected
5.3
2.9
3.4
2.8
2.4
1.5
1.3
0.9
20.5
a = 1 (fixed)
6.2
(5,7.6)
< 0.001
< 0.001
Observed
15
12
15
13
17
9
15
15
111
a
Predicted
a = 1 a
6.3
4.5
7.3
9.3
13.5
13.1
17.7
22.9
94.5
= 3.32 (MLE)
1.1
(0.58, 1.9)
0.90
= 3.32
18.2
10.5
13.5
13.0
14.3
11.7
13.9
15.8
111.0
A.54
-------
Table A-14
Mesothelioma Mortality among Amosite Insulation Workers in New Jersey
Seidman et al. (1986)
Years After First Exposure
Range
(5-9)
( 10-14 )
( 15-19 )
( 20-24 )
( 25-29 )
( 30-34 )
( 35-39 )
Totals
KM MO8
(90% Confidence
Goodness of Fit
Mean
7.5
12.5
17.5
22.5
27.5
32.5
37.5
Interval)
P-value
Duration
1.5
1.5
1.5
1.5
1.5
1.5
1.5
f/ml
46.9
48.3
44.1
43.2
40.3
33.5
31.1
3.9
(2.6, 5.7)
0.35
Person Observed
Years
3952
3628
3198
2656
2094
1576
1086
18190
0
0
0
2
5
8
2
17
Predicted
0
0.1
1.1
2.8
4.2
4.4
4.3
17.0
A.55
-------
Table A-15
Lung Cancer Deaths among Asbestos Workers in Tyler, Texas
Levin et al. (1998)
Du
Range
( < 0.5
( 0.5-1
(1 -5
(> 5)
)
)
)
ration
Mean
0.
0.
7
25
,75
3
.5
f/ml
45
45
45
45
f-y/m I
11.25
33.75
135
337.5
Expected
8.9
1.1
1.8
1.5
Observed
23
3
4
6
Pred
a = 1
10
1.
4.
7.
.2
6
8
8
icted
a = 2.48
22
2.
5.
5.
.4
9
3
4
Totals
KL * 100
(90% Confidence Interval)
Goodness of Fit P-value
Test of H0: a = 1 P-value
a = 1 (fixed)
1.3
(0.55, 2.2)
0.004
< 0.001
13.3 36
a = 2.48 (MLE)
0.13
(0,0.6)
0.81
24.4
36.0
A.56
-------
Table A-16
Lung Cancer Mortality by Cumulative Exposure Among
Vermiculite Mine and Mill Workers Near Libby, Montana
Am and us and Wheeler (1987)
(f-yr)/ml
Range
(<50)
( 50-99 )
( 100-399 )
( >=400 )
Totals
KL* 100
Average
25
75
250
600
SMR
1.5
1.5
1.1
5.8
(90% Confidence Interval)
Goodness
of Fit
P-value
Expected
4.0
1.4
1.9
1.7
9.0
a = 1 (fixed)
0.61
(0.29, 1)
0.41
Observed
6
2
2
10
20
a
Predicted
a = 1 a
4.6
2.0
4.8
8.1
19.5
= 1.13 (MLE)
0.51
(0.11,2)
0.25
= 1.13
5.0
2.1
4.8
8.0
20.0
Test of H0: a = 1 P-value
0.80
A.57
-------
Table A-17
Lung Cancer Mortality by Cumulative Exposure Among
Vermiculite Miners Near Libby, Montana
McDonald et al. (1986)
(f-yr)/ml
Range Average
(0-25)
( 25-200 )
( 200-500 )
( > = 500 )
12.5
77.3
332.4
836.1
SMR
2.04
1.97
7.53
5.58
Expected
3.4
2.5
0.9
0.7
Observed
7
5
7
4
Pred
a = 1
3.9
4.6
4.2
7.0
icted
a = 1.91
6.9
6.3
4.1
5.8
Totals
KL* 100
(90% Confidence Interval)
Goodness of Fit P-value
Test of H0: a = 1 P-value
7.6
a = 1 (fixed)
1.1
(0.55, 1.7)
0.16
0.11
23 19.7
a = 1.91 (MLE)
0.39
(0.067, 1.2)
0.26
23.0
A.58
-------
Table A-18
Lung Cancer Mortality by Cumulative Exposure Among
Ontario Asbestos Cement Plant Workers
Finkelstein (1984)
(f-yr)/ml
Range Average
(
(
( < = 30 )
( 30-75 )
75-105 )
105-150 )
( >150 )
15
52.5
90
127.5
200
2
6
1
2
SMR
.307692
.153846
2.07692
9
.692308
Expected
1.3
1.0
0.4
0.6
0.7
Observed
3
6
5
5
2
Mortality
Rate
3
8
15.7
11.7
3.5
a :
2
3
2
4
7
Pred
= 1
.2
.4
.2
.0
.9
icted
a =
5
4
2
3
5
4.26
.8
.8
.2
.2
.0
Totals
KL* 100
(90% Confidence Interval)
Goodness of Fit P-value
Test of H0: a = 1 P-value
4.0
a = 1 (fixed)
4.8
(2.8, 7.4)
0.03
0.07
21 41.9
a = 4.26 (MLE)
0.29
(0,3.7)
0.05
19.7
21.0
A.59
-------
Table A-19
Mesothelioma Mortality among Employees of an Ontario Asbestos Cement Factory
Finkelstein (1984)
Years After First Exposure
Range Mean
(10-14) 12
(15-19) 17
( 20 - 24 ) 22
( 25 - 29 ) 27
( 30 - 34 ) 32
Totals
KM*108
(90% Confidence Interval)
Goodness of Fit P-value
Duration
6.7
6.7
6.7
6.7
6.7
f/ml
9
9
9
9
9
18
(13,24)
0.26
Person
Years
2500
2500
2963
2063
625
10651
Observed
1
1
8
13
6
29
Predicted
0.03
1.4
7.6
12.8
7.2
29.0
A.6O
-------
Table A-20
Lung Cancer Mortality among Asbestos Cement Workers in Sweden
Albin etal. (1990)
(f-yr)/ml
Relative
RR
Risk (R
Lower
Bound
3.1
25.6
88.2
1.8
1.9
1.9
0.8
0.7
0.5
R) of Dying
Upper
Bound
3.9
5.3
7.1
of Lung C
St. Dev.
0.39
0.52
0.67
ancer
Predicted
a = 1 a =
1.1
1.5
2.7
1.82
1.8
1.8
1.9
Totals
KL* 100
(90% Confidence Interval)
Goodness of Fit P-value
Test of H0: a = 1 P-value
a = 1 (fixed)
1.9
(0, 6.5)
0.32
0.13
5.2
a = 1.82 (MLE)
0.067
(0, 3.6)
0.95
5.6
A.61
-------
Table A-21
Lung Cancer Mortality among Belgian Asbestos-Cement Factory Workers
Laquetetal. (1980)
(f-yr)/ml
Range
( 0 - 49 )
( 50 - 99 )
( 100-199)
(200-399 )
( 400 - 799 )
(800- 1599 )
( 1600-3200)
Totals
KL* 100
(90% Confidence
Goodness of Fit
Test of H0: a = 1
Average
25
75
150
300
600
1200
2400
Interval)
Expected Observed
5.2
2.4
4.6
7.5
2.0
0.6
0.2
22.4
P-value
P-value
6
3
5
4
1
2
0
21
a = 1 (fixed)
0
(0,0.1)
0.51
0.77
Predicted
a = 1
5.2
2.4
4.6
7.4
1.9
0.5
0.2
22.1
a = 0.924
4.8
2.3
4.3
7.0
1.9
0.6
0.2
21.0
a = 0.924(MLE)
0.0068
(0,0.21)
0.39
A.62
-------
Table A-22
Lung Cancer Mortality among Retirees from a US Asbestos Company
Enterline et al. (1986)
mppcf-y
Range Mean
(< 125 )
( 125 - 249
( 250 - 499
( 500 - 749
( >= 750 )
62
182
352
606
976
f-y/m I
Mean
186
546
1056
1818
2928
SMR Observed
182.3
203.1
322
405
698.7
23
14
24
10
8
Expected
12.6
6.9
7.5
2.5
1.1
Pred
a = 1
17.5
14.7
23.7
11.7
8.1
icted
a = 1.43
21.8
15.9
23.4
10.8
7.1
Totals
KL* 100
(90% Confidence Interval)
Goodness of Fit P-value
Test of H0: a = 1 P-value
79
a = 1 (fixed)
0.21
(0.15, 0.27)
0.75
0.24
30.6 75.6
a = 1.43 (MLE)
0.11
(0.041, 0.28)
0.92
79.0
A.63
-------
Table A-23
Lung Cancer Deaths among Insulation Workers in the United States and Canada
Selikoff and Seidman (1991)
Years After First ExpcDuration
Range Mean
(<15)
( 15-19 )
( 20-24 )
( 25-29 )
( 30-34 )
( 36-39 }
( 40-44 >
( 45-49 )
{$0+}
(35+)
12.5
17.5
22.5
27.5
32.5
' 35
35
35
35
35
2.5
7.5
12.5
17.5
22.5
25
25
25
25
25
Person f-y/ml
Years
61655.4
52709.5
57595.4
50518.6
37165.8
20340
. 1Q200.5
5256,5
6151
41948
37.5
112.5
187.5
262.5
337.5
376
375
376
375
375
Observed
7
34
85
172
252
1$3
129
66
71
459
Expected Predicted
a = 1 a = 2.39
3.9
11.6
27.5
46.6
57.5
46.7
30.9
ia,8
25.4
121.8
5.1 9.9
23.0 33.4
72.4 88.2
153.1 164.8
226.5 222.3
These categories
combined irrtcTthe
Over 35 Years
category
519.0 490.4
Totals
Exposure Concentration is 15 f/ml
KL*100
(90% Confidence Interval)
Goodness of Fit P-value
Test of H0: a = 1 P-value
a = 1 (fixed)
0.87
(0.81, 0.93)
0.002
< 0.001
1468 390.6
a = 2.39 (MLE)
0.18
(0.065, 0.38)
0.12
519.0
490.4
A.64
-------
Table A-24
Mesothelioma Deaths among Asbestos Insulation Workers
Selikoff and Seidman (1991)
Years After
Range
(<15)
( 15-19 )
( 20-24 )
(25-29 )
( 30-34 )
(35-39 )
( 40-44 )
( 45-49 )
(50+)
First Exposure
Mean
12.5
17.5
22.5
27.5
32.5
37.5
42.5
47.5
55
Person
Years
61655
52710
57595
50519
37166
20340
10201
5257
6151
Pleura!
0
2
10
33
40
33
17
27
11
Observed
Peritoneal
0
3
8
40
65
58
42
31
38
Total
0
5
18
73
105
91
59
58
49
Predicted
0.2
4.6
23.4
56.3
88.0
87.9
71.9
55.5
106.3
Totals 301593 173 285
Duration = 25 Years and Exposure Concentration = 15 f/ml
KM*108 1.3
(90% Confidence Interval) (1.2, 1.4)
Goodness of Fit P-value < 0.001
458
494.1
A.65
-------
Table A-25
Lung Cancer Mortality among Workers in a Pennsylvania Textile Factory
McDonald et al. (1983b)
mppcf-y
Range Mean
(< 10)
( 10 - 20 )
( 20 -40 )
( 40 - 80 )
(>=80)
5
15
30
60
110
f-y/m I
15
45
90
180
330
SMR
66.9
83.6
156
160
416.1
Observed
21
5
10
6
11
Expected
31.4
6.0
6.4
3.8
2.6
Predicted
a = 1 a = 0.519
34.1
7.5
9.7
7.6
7.6
20.7
5.6
8.8
8.3
9.6
Totals
KL * 100
(90% Confidence Interval)
Goodness of Fit P-value
Test of H0: a = 1 P-value
53
a = 1 (fixed)
0.57
(0.27, 0.94)
0.08
0.01
50.2 66.4
a = 0.519 (MLE)
1.8
(0.75, 4.5)
0.76
53.0
A.66
-------
Table A-26
Mesothelioma Mortality among Pennsylvania Textile Plant Workers
McDonald etal. (1983b)
Years After
First Exposure
15.5
24
41
Duration
9.18
9.18
9.18
f/ml
6.96
6.96
6.96
Person
Years
17179
40868
9840
Observed
6
10
7
Predicted
0.2
8.2
14.6
Totals
67887
23
23.0
KM* 10"
(90% Confidence Interval)
Goodness of Fit P-value
1.1
(0.76, 1.5)
< 0.001
A.67
-------
Table A-27
Lung Cancer Mortality among Rochdale Asbestos Textile Factory
Peto et al. (1985)
particle-yr/ml
Range Mean
( < 1000 )
( 1000 - 2000 )
( 2000 - 3000 )
( 3000 -4000 )
( 4000 - 5000 )
( >= 5000 )
209
1409
2511
3474
4551
9057
f-y/ml Observed Expected
5.92
39.92
71.13
98.41
128.92
256.57
34
8
11
6
10
24
29.5
7.7
6.6
5.7
4.3
10.8
Predicted
a = 1 a = 1.10
30.4
9.2
9.0
8.5
7.2
25.2
33.2
9.8
9.4
8.8
7.2
24.6
Totals
KL*100
(90% Confidence Interval)
Goodness of Fit P-value
Test of H0: a = 1 P-value
93 64.6
a = 1 (fixed)
0.52
(0.28, 0.79)
0.72
0.57
89.6 93.0
a = 1.10 (MLE)
0.41
(0.12, 0.87)
0.63
A.68
-------
Table A-28
Mesothelioma Mortality among Rochdale Asbestos Textile Factory
Petoetal. (1985)
Years After First Exposure
Range
(0-19)
( 20-24 )
( 25-29 )
( 30-34 )
( 35-39 )
(>=40)
(0-19)
( 20-24 )
( 25-29 )
( 30-34 )
( 35-39 )
(>=40)
(0-19)
( 20-24 )
( 25-29 )
( 30-34 )
( 35-39 )
(>=40)
(0-19)
( 20-24 )
( 25-29 )
( 30-34 )
( 35-39 )
(>=40)
( 20-24 )
( 25-29 )
( 30-34 )
( 35-39 )
(>=40)
( 30-34 )
( 35-39 )
( >=40 )
Totals
KM MO8
(90% Confidence
Mean
11.5
22.5
27.5
32.5
37.5
42
11.5
22.5
27.5
32.5
37.5
42
11.5
22.5
27.5
32.5
37.5
42
11.5
22.5
27.5
32.5
37.5
42
22.5
27.5
32.5
37.5
42
32.5
37.5
42
Interval)
Duration
0.5
0.5
0.5
0.5
0.5
0.5
3
3
3
3
3
3
7.5
7.5
7.5
7.5
7.5
7.5
15
15
15
15
15
15
25
25
25
25
25
35
35
35
Goodness of Fit P -value
f/ml
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
9.12
1.3
(0.74, 2.1)
0.80
Person
Years
28015
4668
3470
2041
840
402
4786
877
632
421
238
148
8521
1417
1104
707
383
249
4814
1423
870
470
204
102
848
935
600
257
122
86
107
103
69861
Observed
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
3
0
1
1
1
2
1
0
0
0
0
10
Predicted
0.01
0.2
0.3
0.3
0.2
0.1
0.003
0.2
0.3
0.3
0.3
0.2
0.01
0.5
0.9
1.1
0.9
0.9
0.003
0.5
0.9
1.0
0.7
0.5
0.3
1.0
1.3
1.0
0.8
0.2
0.4
0.7
16.1
A.69
-------
APPENDIX B:
REPORT ON THE PEER CONSULTATION WORKSHOP
TO DISCUSS A PROPOSED PROTOCOL TO ASSESS
ASBESTOS-RELATED RISK
-------
Report on the Peer Consultation Workshop to Discuss a
Proposed Protocol to Assess Asbestos-Related Risk
Preparedfor:
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Washington, DC 20460
EPA Contract No. 68-C-98-148
Work Assignment 2003-05
Prepared by:
Eastern Research Group, Inc.
110 Hartwell Avenue
Lexington, MA 02421
FINAL REPORT
May 30, 2003
-------
NOTE
This report was prepared by Eastern Research Group, Inc. (ERG), an EPA contractor, as a general
record of discussion for the peer consultation workshop on a proposed protocol to assess asbestos-
related risk. This report captures the main points of scheduled presentations, highlights discussions
among the panelists, and documents the public comments provided at the meeting. This report does not
contain a verbatim transcript of all issues discussed, and it does not embellish, interpret, or enlarge upon
matters that were incomplete or unclear. EPA will use the information presented during the peer
consultation workshop to determine whether the proposed risk assessment methodology can be used to
support decisions at asbestos-contaminated sites. Except as specifically noted, no statements in this
report represent analyses by or positions of EPA or ERG.
-------
CONTENTS
List of Abbreviations iii
Executive Summary v
1. Introduction 1-1
1.1 Background 1-1
1.2 Scope of the Peer Consultation Workshop 1-2
1.2.1 Activities Prior to the Peer Consultation Workshop 1-2
1.2.2 Activities at the Peer Consultation Workshop 1-3
1.2.3 Activities Following the Peer Consultation Workshop 1-4
1.3 Report Organization 1-5
2. Background on the Proposed Protocol to Assess Asbestos-Related Risk 2-1
3. Comments on Topic Area 1: Interpretations of the Epidemiology
and Toxicology Literature 3-1
3.1 Lung Cancer 3-1
3.1.1 Lung Cancer and Fiber Type: Inferences from 1he Epidemiology Literatufb-1
3.1.2 Lung Cancer and Fiber Type: Inferences from Animal Toxicology
and Mechanistic Studies 3-6
3.1.3 Lung Cancer and Fiber Dimension: Inferences from the
Epidemiology Literature 3-7
3.1.4 Lung Cancer and Fiber Dimension: Inferences from Animal
Toxicology and Mechanistic Studies 3-9
3.1.5 Other Issues Related to Lung Cancer 3-10
3.2 Mesolhelioma 3-12
3.2.1 Mesothelioma and Fiber Type: Inferences from the Epidemiology Literahde2
3.2.2 Mesolhelioma and Fiber Type: Inferences from Animal
Toxicology and Mechanistic Studies 3-14
3.2.3 Mesothelioma and Fiber Dimension: Inferences from the
Epidemiology Literature 3-16
3.2.4 Mesolhelioma and Fiber Dimension: Inferences from Animal Toxicology and
Mechanistic Studies 3-17
3.3 Exposure Estimates in the Epidemiology Literature 3-18
-------
CONTENTS (Continued)
4. Comments on Topic Area 2: The Proposed Exposure Index 4-1
4.1 Responses to Charge Question 4 4-1
4.2 Responses to Charge Question 5 4-3
4.3 Responses to Charge Question 6 4-4
5. Comments on Topic Area 3: General Questions 5-1
5.1 Responses to Charge Question 7 5-1
5.2 Responses to Charge Question 8 5-2
5.3 Responses to Charge Question 9 5-3
5.4 Responses to Charge Question 10 5-4
5.5 Responses to Charge Question 12 5-6
6. Comments on Topic Area 4: Conclusions and Recommendations 6-1
6.1 Responses to Charge Question 11 6-1
6.2 Development of Final Conclusions and Recommendations 6-6
7. References 7-1
Appendices
Appendix A
Appendix B
Appendix C
Appendix D
Appendix E
Appendix F
List of Expert Panelists
Premeeting Comments, Alphabetized by Author (includes bios of panelists and the
charge to the reviewers)
List of Registered Observers of the Peer Consultation Workshop
Agenda for the Peer Consultation Workshop
Observer Comments Provided at the Peer Consultation Workshop
Observer Post-Meeting Comments
-------
LIST OF ABBREVIATIONS
ATSDR Agency for Toxic Substances and Disease Registry
EPA U.S. Environmental Protection Agency
ERG Eastern Research Group, Inc.
IARC International Agency for Research on Cancer
IRIS Integrated Risk Information System
NIOSH National Institute for Occupational Safety and Health
PCM phase contrast microscopy
SEM scanning electron microscopy
SVF synthetic vitreous fibers
TEM transmission electron microscopy
micrometers
111
-------
EXECUTIVE SUMMARY
Eleven expert panelists participated in a peer consultation workshop to review a proposed protocol to
assess asbestos-related risks. The protocol is documented in the report, 'Technical Support Document
for a Protocol to Assess Asbestos-Related Risk, Parts I and H" (Berman and Crump 1999,2001). At
the end of the 254-day workshop, which was open to the public, the expert panelists drafted the
following summary of their findings:
The peer consultation panel strongly endorsed the conceptual approach of developing an updated
cancer risk assessment methodology that takes into account fiber type and fiber dimension. The
opportunity is at hand to use substantial new information from epidemiology, experimental toxicology,
and exposure characterization on what continues to be an extremely important societal issue—assessing
the health risks associated with environmental and occupational exposures to asbestos. The panel
recommended that EPA proceed in an expeditious manner to consider the panelists' conclusions and
recommendations with a goal of having an updated asbestos risk assessment methodology. It is
important that EPA devote sufficient resources so that this important task can be accomplished in a
timely and scientifically sound manner. The panel urges that additional analyses underpinning the
document, preparation of documentation, and further review be carried out in an open and transparent
manner.
Prior to the workshop, the participants received draft copies of the "Methodology for Conducting Risk
Assessments at Asbestos Superfund Sites Part 1: Protocol" and "Part 2: Technical Background
Document" The panelists generally found that these documents did not provide a complete and
transparent description of how the data were analyzed to support the conclusions presented. The
incomplete documentation of methodology precluded the replication of the findings, in advance of the
meeting, by several panelists. The methodology used was clarified by the comprehensive presentations
that Drs. Berman and Crump made at the workshop. However, future drafts of these documents must
-------
clearly describe the methodologies and include sufficient data, perhaps in appendices, such that the
findings can be replicated.
The panelists made the following conclusions and recommendations:
Measurement methods. Continuing advances have been made in the application of exposure
measurement technology for asbestos fibers during the past two decades. These advances
include the use of transmission electron microscopy (TEM) and allied techniques (e.g., energy
dispersive x-ray detection, or EDS) as an alternative to phase contrast microscopy (PCM),
thereby allowing the bivariate (i.e., length and width) characterization of fibers and fiber type.
The proposed risk assessment methodology incorporates these advances in the development of
an exposure index. The panel was in agreement that this aspect of the new risk assessment
methodology represents a substantial advance over the existing methodology.
Integration of exposure and risk assessment models. A key aspect of the proposed risk
assessment methodology is a linking of specific exposure characterization methodology with
exposure-response coefficients. It has been emphasized that any change in the exposure
characterization metrics must be accompanied by changes in the exposure-response coefficients
of the risk assessment models. This was emphasized in the report and the panelists endorsed this
view.
Access to additional raw data sets. The panelists strongly recommended that EPA make
every attempt to acquire and analyze raw data sets from key human epidemiological studies.
Where possible, it would also be desirable to obtain bivariate (i.e., length and diameter) fiber
exposure information for these re-analyses. Several panelists believed that review of additional
data sets offers substantial opportunity for improving the proposed risk assessment
methodology. In the event that raw data cannot be obtained due to confidentiality reasons or
other restrictions, the panelists suggested that the authors consider asking those who have
access to the data to conduct the necessary statistical analyses and communicate their results
directly to EPA for further consideration.
Fiber diameter. The proposed risk assessment methodology uses a diameter cut-off of 0.5
micrometers (urn) for considering fibers. The report states that fibers 0.7 urn in diameter can
reach the respiratory zone of the lung. A few panel members indicated that the fiber diameter
cut-off could be as high as 1.5 |im during oral breathing. The 0.4 ^m cut-off came from rat data,
but larger diameters would be expected to be respirable in humans. There was general
agreement that the diameter cut-off should be between 0.5 and 1.5 urn. This issue is deserving
of further analysis.
VI
-------
Fiber length. The Herman and Crump analyses made a significant contribution by obtaining and
analyzing membrane filters from the animal inhalation studies in Edinburgh and conducting
quality-assured bivariate length and distribution analyses by TEM — thereby greatly reducing the
uncertainty of the exposure side of the exposure-response relationship for chronic fiber exposure
in rats. Unfortunately, correspondingly detailed information on bivariate size distribution is not
available for humans. This leads to the need to use the animal data, although one must always
recognize the uncertainties associated with interspecies extrapolations such as anatomic
characteristics and respirability between species. Future analyses may benefit from using other
available laboratory animal data sets and human data sets.
The fiber length distributions for the human cohort exposures are much more uncertain. For the
Wittenoom, Quebec, and South Carolina cohorts, there are limited fiber length distribution data
based on TEM analysis from historic membrane filter samples, but only fiber categories longer
than 5 ^m and longer than 10 |im were counted. For all other cohorts, the measurements were
limited to PCM fiber counts for all fibers greater than 5 |im in length in some, and particle counts
(lOx objective) on midget impinger samples in others. Both methods do not measure thin fibers,
do not discriminate between asbestos and other mineral particles, and provide no information on
the concentrations of fibers longer than 10, 20, or 40 urn, or inter-laboratory variations in
optical resolution and counting rules. As one approach to addressing the varying uncertainty in
assessing exposure in the different studies, Berman and Crump used the available information to
make adjustments to the uncertainty ranges in the exposure-response coefficients. The
workshop panel welcomed this initiative but suggested alternative approaches (see "Methods,"
below).
Some panelists felt that an Exposure Assessment Workshop, with participants having a broad
range of expertise, could evaluate the uncertainties in historic occupational data sets' exposure
measurements. They felt such a workshop could result in a more confident assessment of
exposure-response relationships for populations exposed to a variety of amphiboles, chrysotile,
and mixtures. With incorporation of other available knowledge on fiber type, process, smoking
(if available), and the relative number of excess lung cancer and mesothelioma, it may well be
possible to gain a much clearer understanding of the roles of these variables as causal factors for
these asbestos-associated cancers. In addition, the workshop would prove valuable in further
discussion of mineralogical, geological, and industrial hygiene issues with regard to application of
the model to risk assessment in environmental sites of concern.
The Berman and Crump index assigns zero risk to fibers less than 5 ^m in length Fibers
between 5 and 10 |im are assigned a risk that is one three-hundredth of the risk assigned to
fibers longer than 10 nm Panelists agreed that there is a considerably greater risk for lung
cancer for fibers longer than 10 ^m. However, the panel was uncertain as to an exact cut size
for length and the magnitude of the relative potency. The panelists also agreed that the available
vu
-------
data suggest lhat the risk for fibers less than 5 nm in length is very low and could be zero. This
specific issue was addressed by an expert panel convened by the Agency for Toxic Substances
and Disease Registry (ATSDR) in October 2002. Some panelists suggested that, for
mesothelioma, greater weight should perhaps be assigned to fibers in the 5 to 10 ^m length
range and to thinner fibers.
Fiber type. For mesothelioma, the panelists supported the use of different relative carcinogenic
potencies for different fiber types. The panelists unanimously agreed that the available
epidemiology studies provide compelling evidence that the carcinogenic potency of amphibole
fibers is two orders of magnitude greater than that for chrysotile fibers. There was some
discussion about the precise ratio expressed due to questions about the availability of exposure
data in existing studies (e.g., Wittenoom). There was recognition that time since first exposure is
an important factor in determining risk for mesothelioma and some discussion is needed on the
importance of duration and intensity of exposure.
For lung cancer, the panelists had differing opinions on Ihe inferences that can be made on the
relative potency of chrysotile and amphibole fibers. Some panelists supported the finding that
amphibole fibers are 5 times or more potent for lung cancer than are chrysotile fibers. Other
panelists did not think the statistical analyses in Ihe draft methodology document supports this
relative potency and wondered if additional review of the epidemiological data might identify
factors other than fiber type (e.g., industry considered) that provide further insights on the
matter. These other factors can then be considered when the risk assessment is applied.
or
assume
Cleavage fragments. The panel knew of little data to directly address the question as to
whether cleavage fragments of equal durability and dimension as fibers would have similar
dissimilar potency for lung cancer. The general view is that data indicate that durability and
dimension are critical to pulmonary pathogenesis. Therefore, it is prudent at this time to assume
equivalent potency for cancer in the absence of other information to the contrary. Consideration
of conducting a rat inhalation study using tremolite cleavage fragments was recommended to
address this issue. For mesothelioma, it was viewed that thin fibers greater than 5 (im in length
are more important. Cleavage fragments that do not meet these criteria would not contribute to
risk of mesothelioma
Other amphiboles. The panel agreed with the report's conclusion that the potency of currently
regulated and unregulated amphibole fibers should be considered equal based on the reasoning
that similar durability and dimension would be expected to result in similar pathogenicity.
Methods. The panelists extensively discussed the approach to conducting the meta-analysis of
the large number of epidemiological studies. A number of the panelists urged that consideration
be given to using more traditional approaches that would include development and application of
specific criteria for inclusion of studies into the exposure-response analysis, examination of
Vffl
-------
heterogeneity and sources of Ihe heterogeneity, and the use of sensitivity analysis to identify
influential studies.
The panelists also urged, in the study-specific analysis, exploration of alternative exposure-
response models other than the lung cancer and mesothelioma risk models EPA has been using
since 1986. This would possibly include non-linear response models (e.g., log-linear models),
examination of separate effects for concentration and duration, time since first exposure, time
since cessation of exposure, possibly dropping the "a factor," and different methods for
measurement error. The adequacy of different models should be examined using goodness of fit
statistics across all studies. The possibility of internal analyses should be re-examined (i.e., it may
be possible to obtain partial data, such as age-specific person years data, from authors).
Exploration of non-linearity should also include shape of the curve in the low exposure area
The panelists also urged alternative approaches to meta-analyses. In particular, panelists
recommended meta-regression using original (untransformed) exposure-response coefficients, in
which predictor variables include the estimated percentage of amphiboles, percentage of fiber
greater than 10 urn, and categorical grouping of studies according to quality. Original exposure-
response coefficient variances should be used in conjunction with random effects models in
which residual inter-study variation is estimated. Analyses restricted to long latency and a
predictor variable for industry type should be considered. A priori distribution for inter-study
residual variance might also be considered. Mela-regression will allow simple inspection of
likelihoods to consider the importance of different predictor variables. Sensitivity analyses should
be conducted in which the inclusion or exclusion of specific studies or groups of studies is
evaluated.
Cigarette smoking. Most panelists felt strongly that future analyses need to pay more attention
to the effects of smoking on the lung cancer exposure-response model and extrapolations to
risk. However, the current data sets have variable and limited information available on smoking.
The panelists noted that smoking is the primary cause for lung cancer, but the lung cancer dose-
response relationship for smoking is complex due to the effects of smoking duration, intensity,
and cessation.
The impact of smoking has effects on both the estimation and the application of the model for
projecting risk of lung cancer due to asbestos exposure. This may be an especially critical issue
for low-exposure extrapolation. With respect to estimation, accepting the form of the proposed
model, the effect of smoking may require different KL values for smokers and non-smokers. The
panelists recognized that there is limited epidemiologic data to address this issue, but
recommend that it be investigated. With respect to applying the model to make risk projections
for any future cohort, the background rate of lung cancer employed in the model needs to be
carefully determined to capture the smoking behavior of the cohort.
-------
• Localized tremolite exposures. During the course of public comments, the panel received
input from several individuals who expressed concerns about environmental exposures to
tremolite asbestos from localized geologic formations in California The individuals suggested
that inadequate attention had been given to characterization of the exposures to residents of
Ihese communities. While the panel was not in a position or charged with the evaluation of this
issue, the panel did feel that this was a potentially serious matter deserving of attention by the
appropriate public health authorities. Evaluation of these kinds of situations would benefit from
the use of Ihe improved risk assessment methodology being considered.
The remainder of Ibis report summarizes the discussions and observations that led to Ihese findings,
reviews the panelists' comments on many topics not listed in this executive summary, and documents
the observer comments provided at the workshop.
-------
1. INTRODUCTION
This report summarizes a peer consultation by 11 expert panelists of a proposed protocol to assess
asbestos-related risks. Contractors to the U.S. Environmental Protection Agency (EPA) developed the
proposed protocol, which is documented in a report titled: 'Technical Support Document for a
Protocol to Assess Asbestos-Related Risk" (Herman and Crump 2001). The purpose of the peer
consultation workshop was to provide EPA feedback on the scientific merit of the proposed protocol.
The peer consultation workshop took place in a meeting open to the public on February 25-27,2003,
in San Francisco, California
This report summarizes the technical discussions among the expert panelists and documents comments
provided by observers. These discussions largely focused on three topic areas: interpretations of the
epidemiology and toxicology literature, the proposed exposure index, and general questions about key
assumptions and inferences in the protocol. The remainder of this introductory section presents
background information on the protocol (Section 1.1), describes the scope of the peer consultation
workshop (Section 1.2), and reviews the organization of this report (Section 1.3).
1.1 Background
EPA's current assessment of asbestos toxicity is based primarily on an asbestos review completed in
1986 (EPA 1986) and has not changed substantially since that time. The 1986 assessment considers six
mineral forms of asbestos and all asbestos fiber sizes longer than 5 micrometers (nm) to be of equal
carcinogenic potency. However, since 1986, asbestos measurement techniques and the understanding
of how asbestos exposure contributes to disease have improved substantially. To incorporate the
knowledge gained over the last 17 years into the agency's toxicity assessment for asbestos, EPA
•
contracted with Aeolus, Inc., to develop a proposed methodology for conducting asbestos risk
assessments. The proposed methodology distinguishes between fiber sizes and fiber types in estimating
1-1
-------
potential health risks related to asbestos exposure. The methodology also proposes a new exposure
index for estimating carcinogenic risk
As a key step in determining the scientific merit of the proposed risk assessment methodology, EPA
decided to obtain expert input on the draft report through a peer consultation workshop. The purpose
of the workshop was to obtain feedback from subject-matter experts during the development stage of
the proposed risk assessment methodology; the workshop was not an official peer review. Eastern
Research Group, Inc. (ERG), organized and implemented the peer consultation workshop under a
contract to EPA.
1.2 Scope of the Peer Consultation Workshop
The peer consultation involved many activities before the workshop (see Section 1.2.1), at the
workshop (see Section 1.2.2), and after the workshop (see Section 1.2.3). The following subsections
describe these activities.
1.2.1 Activities Prior to the Peer Consultation Workshop
This section describes the major activities ERG and the expert panelists conducted prior to the peer
consultation workshop:
Select expert panelists. ERG selected the expert panelists for the peer consultation workshop.
ERG sought to compile a panel of experts with broad experience and expertise in the following
disciplines: toxicology, epidemiology, biostatistics, asbestos sampling and analytical methods,
EPA's human health risk assessment guidelines, and asbestos-related environmental and
occupational health issues. Appendix A lists the expert panelists ERG selected, and Appendix B
includes brief biographies that summarize the panelists' areas of expertise.
Every panelist is either a senior scientist, physician, or researcher with extensive experience in
the aforementioned fields, as demonstrated by peer-reviewed publications, awards, and service
1-2
-------
to relevant professional societies. To ensure the peer consultation offered a balanced
perspective, ERG intentionally selected expert panelists with a broad range of affiliations (e.g.,
academia, consulting, state and federal agencies). When searching for panelists, ERG asked all
candidates to disclose real or perceived conflicts of interest.
Prepare a charge to the expert panelists. ERG worked with EPA to prepare written
guidelines (commonly called a "charge") for the peer consultation workshop. The charge
includes 12 specific questions, organized into 4 topic areas. Discussions at the workshop largely
addressed Ihe technical issues raised in the charge, but the expert panelists were encouraged to
discuss other relevant matters that were not specifically addressed in the charge questions. A
copy of the charge is included in Appendix B.
Distribute review documents and other relevant information. Several weeks prior to the
peer consultation workshop, ERG sent every panelist copies of the charge and the proposed
risk assessment methodology (Berman and Crump 2001). These items formed tiie basis of the
technical discussions at the workshop. In addition, ERG distributed several additional
publications on related topics (see Table 1, at the end of this section, for list of the publications).
The supplemental publications were provided largely in response to panelists' requests for
further background information on selected issues. The panelists also circulated publications
amongst themselves on specific topics. Finally, one of Ihe meeting chairs noted for Ihe record
that, upon arriving in San Francisco, he also received a memo and copies of many abstracts and
other information from Gate Jenkins of EPA. The meeting chair offered to share these materials
wilh other panelists during Ihe workshop.
Obtain and compile the panelists' premeeting comments. After receiving the workshop
materials, Ihe panelists were asked to prepare their initial responses to the charge questions.
Booklets containing the premeeting comments were distributed to the expert panelists before Ihe
workshop and were made available to observers at the workshop. These initial comments are
included in this report, without modification, as Appendix B. It should be noted lhat the
premeeting comments are preliminary in nature. Some panelists' technical findings may have
changed after the premeeting comments were submitted.
1.2.2 Activities at the Peer Consultation Workshop
The 11 expert panelists and approximately 75 observers attended the peer consultation workshop,
which was held at the Westin St. Francis Hotel in San Francisco, California, on February 25-27,2003.
The workshop was open to the public, and the workshop dates and times were announced in the
1-3
-------
Federal Register. Appendix C lists the observers who confirmed their attendance at the workshop
registration desk. The workshop schedule generally followed Ihe agenda, presented here as Appendix
D.
The workshop began with introductory remarks from Ms. Jan Connery (ERG), the facilitator of the
peer consultation. Ms. Connery welcomed the expert panelists and observers, stated the purpose of the
workshop, identified the document being reviewed, and explained the procedure for observers to make
comments. Mr. Richard Troast (EPA) then provided background information on the review document
and EPA's ongoing efforts to assess asbestos toxicity (see Section 1.1). Mr. Troast identified the main
differences between EPA's existing asbestos risk assessment methodology (EPA 1986) and the
proposed methodology (Berman and Crump 2001). Mr. Troast noted that the expert panelists'
feedback will ultimately help EPA complete its update of asbestos health risks for the Integrated Risk
Information System (IRIS); he clarified that the final IRIS update will be subject to peer review or
Science Advisory Board review before being implemented. Following these opening remarks, Dr.
Wayne Berman and Dr. Kenny Crump—the authors of the proposed methodology—presented
detailed information on the review document; Section 2 of this report summarizes their presentations.
After the background presentation, Dr. Roger McClellan and Dr. Leslie Stayner chaired the technical
discussions that followed. For the remainder of the meeting, the panelists engaged in free-flowing
discussions when answering the charge questions and addressing additional topics not specified in the
charge. Observers were given Ihe opportunity to provide verbal comments three different times during
the workshop; these observer comments are documented in Appendix E. Representatives from EPA
and the document authors provided clarifications on Ihe proposed methodology periodically throughout
the 2l/2-day workshop.
1.2.3 Activities Following me Peer Consultation Workshop
1-4
-------
The primary activity following the peer consultation workshop was preparing this summary report. A
technical writer from ERG who attended the meeting prepared a draft of this report, which ERG
distributed to the 11 expert panelists and asked them to verify that the draft accurately reflects the tone
and substance of the panelists' discussions at the workshop. After incorporating the panelists'
suggested revisions to the draft report, ERG submitted the final report (i.e., this report) to EPA.
1.3 Report Organization
The structure of this report follows the order of the technical discussions during the meeting. Section 2
summarizes Dr. Berman and Crump's background presentations. Sections 3 through 6 are records of
the panelists' discussions on the four main topic areas: interpretations of the epidemiology and
toxicology literature (Section 3), the proposed exposure index (Section 4), general questions (Section
5), and conclusions and recommendations (Section 6). Finally, Section 7 provides references for all
documents cited in the text.
The appendices to this report include background information on the peer consultation workshop. This
information includes items that were on display at the workshop and items generated since the
workshop (e.g., a final list of attendees). The appendices contain the following information:
• List of the expert panelists (Appendix A).
• The panelists' premeeting comments, the charge to the reviewers, and brief bios of the expert
panelists (Appendix B).
• List of registered observers of the peer consultation workshop (Appendix C).
• Agenda for the peer consultation workshop (Appendix D).
• Observer comments provided at the peer consultation workshop (Appendix E).
• Observer post-meeting comments (Appendix F).
1-5
-------
Table 1
References ERG Provided to the Expert Panelists
Herman, DW and Crump K. 1999. Methodology for Conducting Risk Assessments at Asbestos
Superfund Sites; Part 1: Protocol. Final Draft. Prepared for U.S. Environmental Protection Agency.
February 15,1999.
Berman, DW and Crump K. 2001. Technical Support Document for a Protocol to Assess
Asbestos-Related Risk. Final Draft. Prepared for U.S. Department of Transportation and U.S.
Environmental Protection Agency. September 4,2001.
Berman, DW, Crump, K., Chatfield, E., Davis, J. and A. Jones. 1995. The Sizes, Shapes, and
Mineralogy of Asbestos Structures that Induce Lung Tumors or Mesothelioma in AF/HAN Rats
Following Inhalation. Risk Analysis. 15:2,181-195.
Berman, DW. 1995. Errata Risk Analysis. 15:4, 541.
Committee on Nonoccupational Health Risks of Asbestiform Fibers. Breslow, L., Chairman. 1984.
Asbestiform Fibers Nonoccupational Health Risks. Washington, DC: National Academy Press.
EPA 1986. Airborne Asbestos Health Assessment Update. U.S. Environmental Protection Agency.
EPA 600/8-84-003F. 1986.
NIOSH taterdivisional Fiber Subcommittee Report. Prepared by the NIOSH Interdivisional Fiber
Subcommittee. 1999.
1-6
-------
2. BACKGROUND ON THE PROPOSED PROTOCOL
TO ASSESS ASBESTOS-RELATED RISK
This section summarizes presentations given by the principal authors of the proposed risk assessment
methodology. These presentations were given because several panelists asked ERG, prior to Ihe peer
consultation workshop, if the authors would provide detailed background information on how the
methodology was developed. This section reviews the major presentation topics, but does not present
the panelists' comments on the proposed protocol. Sections 3 through 6 document the expert panelists'
technical feedback on the protocol.
Motivation for developing the proposed protocol Dr. Herman identified several reasons for
developing the updated protocol for assessing asbestos-related risks. These reasons include
EPA's existing asbestos models being inconsistent with inferences from the scientific literature,
Ihe need for having uniformly-applied sampling and analytical procedures to measure asbestos
characteristics most predictive of risk, and the belief that EPA's current asbestos risk
assessment methodology may not be adequately protective in some circumstances. To improve
upon the current methodology, the authors intended to develop a risk assessment model that
adequately predicts cancer risk in all studied environments and can therefore be applied with
much greater confidence to environments that have not been studied. Dr. Berman outlined the
general approach taken to develop the proposed protocol, as summarized in the following
bulleted items.
Dr. Berman provided background information on and definitions for asbestos, other fibrous
structures, asbestos morphology, and cleavage fragments. He also described the capabilities and
limitations of the analytical techniques that have been used to characterize asbestos exposures,
such as midget impingers, phase contrast microscopy (PCM), scanning electron microscopy
(SEM), and transmission electron microscopy (TEM). Dr. Berman explained how differences in
these analytical techniques must be critically evaluated when comparing results reported in all
epidemiological and other types of studies that examine asbestos exposure. Dr. Berman also
stressed that it is not just differences in analytical techniques, but choice of specific methods for
each analytical technique that affects results. Further information on these topics is included in
Chapter 4 of the proposed protocol (Berman and Crump 2001).
Re-analysis of human epidemiological data. Dr. Crump described how the authors
evaluated the human epidemiological data He displayed a list of the studies that were
considered, noting that he had access to raw, individual-level data for three occupational
cohorts: chrysotile textile workers in South Carolina, United States; crocidolite miners in
2-1
-------
Wittenoom, Australia; and chiysotile miners and millers in Quebec, Canada All data sets with
exposure data were considered in the analysis, and criteria were not established for selecting
studies. Dr. Crump ihen presented findings for asbestos-related risks for lung cancer and
mesothelioma
For lung cancer, Dr. Crump first reviewed EPA's existing lung cancer model for asbestos
exposure (see equation 6.1 in the proposed protocol), which relates the relative risk of lung
cancer mortality linearly to cumulative asbestos exposure, with a 10-year lag time. Dr. Crump
noted that the model predicts that relative risk for developing lung cancer remains constant after
asbestos exposure ceases—an assumption he showed was reasonably consistent with findings
from epidemiological studies. Dr. Crump also discussed how the model assesses interactions
between exposures to cigarette smoke and to asbestos—an issue the panelists revisited several
times later in the workshop (e.g., see Section 3.1.1 and the executive summary). Dr. Crump
presented a series of tables and figures demonstrating the adequacy of multiple lung cancer
models: first using EPA's existing lung cancer model, next using a modified version of the model
that accounts for differences in the background rates of lung cancer, and finally using the
proposed lung cancer model, which considers an exposure index that assigns greater
carcinogenic potency to amphibole fibers and to longer fibers.
Similarly, Dr. Crump reviewed the performance of EPA's mesolhelioma model for asbestos
exposures (see equation 6.11 in the proposed protocol), which predicts that mesothelioma risks
vary linearly with the average asbestos exposure and increase quadratically with time from onset
of exposure. Dr. Crump presented several tables and graphs indicating how well EPA's existing
model and the proposed protocol fit the human epidemiological data He made several
conclusions about the existing risk model, including lhat mesothelioma risk coefficients varied
considerably across the cohorts and the risk coefficients were generally higher for cohorts
exposed primarily to amphibole fibers, compared to those exposed primarily to chiysotile fibers.
Dr. Crump also noted that the data did not support consideration of a sub-linear or threshold
dose-response relationship. This latter point generated considerable discussion later in the
workshop (e.g., see Section 4.3).
Dr. Crump then described the meta-analysis the authors conducted to evaluate the relative
potency of amphibole and chrysotile fibers. First, he explained how the authors weighted Ihe
different studies in Ihe meta-analysis, based on uncertainty factors assigned to the individual
studies. Dr. Crump identified Ihe four uncertainty factors and described generally how each
factor was assigned. Sources of uncertainty included representativeness of air sampling data, the
availability of conversion factors to express exposures in terms of PCM concentrations, and
whether data on exposure duration were available. Dr. Crump then highlighted the main
conclusions from the meta-analysis. For lung cancer, the meta-analysis suggested that amphibole
fibers are approximately five times more potent than are chrysotile fibers, but the difference in
potency was not statistically significant (i.e., the authors could not reject the hypothesis that
2-2
-------
chiysotile fibers and amphibole fibers are equally potent). For mesothelioma, the meta-analysis
suggested that chrysotile fibers are 0.002 times as potent as amphibole fibers, and the difference
in potency was statistically significant
Inferences drawn from the broader literature. Dr. Berman described how the authors
incorporated inferences from the broader scientific literature into the proposed protocol. He
reviewed key findings on how various mechanisms are biologically related to how asbestos
causes disease. These mechanisms included respiration, deposition, degradation, clearance,
translocation, and tissue-specific biological responses. Chapter 7 of the review document
provides detailed information on the relevance of Ihese mechanisms, with emphasis on the
influence of fiber type and fiber dimension.
Derivation of the exposure index. Dr. Berman explained how the authors derived the
exposure index, which is largely based on an earlier re-analysis (Berman et al. 1995) of six
animal inhalation studies conducted by a single laboratory. That re-analysis found that lung tumor
incidence is adequately predicted using an exposure index that assigns no carcinogenic potency
to fibers shorter than 5 urn, relatively low carcinogenic potency to fibers with lengths between 5
and 40 ^m and diameters less than 0.4 nm, and the greatest carcinogenic potency to fibers
longer than 40 ^m and thinner than 0.4 ^im However, these findings could not be applied
directly to the human epidemiological data, because the epidemiological studies do not include
exposure measurements that quantify the relative amounts of asbestos fibers shorter and longer
than 40
Dr. Berman noted that the proposed protocol includes an ad hoc assumption lhat the fiber size
weighting factors optimized from the laboratory animal studies can be applied to humans, but
with a length cut-off of 1 0 nm in the exposure index, ralher than a cut-off of 40 ^m Dr. Berman
emphasized that this assumption was made to model the critical characteristics of asbestos in a
manner that reasonably captures cancer risks observed across multiple epidemiological studies.
He acknowledged that asbestos potency is likely a continuous function of fiber length, but the
exposure measurements from Ihe available animal and epidemiological studies do not support
incorporating such a continuous function in the exposure-response model. The panelists
commented on Ihe proposed exposure index when discussing topic area 3 (see Section 4).
Dr. Berman also noted that the authors selected a conservative set of dose-response coefficients
(see Table 6-30 of the review document), ralher than using Ihe optimized ones from the animal
studies (see Table 6-29). However, the conservative and optimized dose-response coefficients
were reasonably consistent: none of the conservative coefficients differed by more than a factor
of 4 from the corresponding optimized ones.
Conclusions regarding proposed protocol Dr. Berman indicated that the proposed protocol
is substantially more consistent with inferences documented in the scientific literature (i.e., that
2-3
-------
long, thin structures contribute most to risk) than EPA's existing risk assessment methodology.
Further, the proposed protocol provides a better fit to cancer risks observed in the human
epidemiological studies than does EPA's existing model, and the proposed protocol appears to
underestimate risks of lung cancer and mesothelioma less frequently and to a lesser degree than
the existing approach. Finally, by recommending use of a standardized analytical method that
links directly to the exposure index, the proposed protocol will help ensure that future risk
assessments are conducted in a consistent fashion and their results can be readily compared
from one study to the next.
2-4
-------
3. COMMENTS ON TOPIC AREA 1: INTERPRETATIONS OF THE
EPIDEMIOLOGY AND TOXICOLOGY LITERATURE
This section summarizes the panelists' discussions on the interpretations of the epidemiology and
toxicology literature. The meeting co-chairs—Dr. McClellan and Dr. Stayner—facilitated the
discussions on this topic area, which focused first on lung cancer (see Section 3.1) and then on
mesothelioma (see Section 3.2). This section presents a record of discussion of the topics mentioned
during the workshop. Several panelists referred to their premeeting comments (see Appendix B) for
additional suggestions for how the review of epidemiology and toxicology literature can be improved.
3.1 Lung Cancer
The panelists discussed at length whether the epidemiology and toxicology literature support the
proposed protocol's finding for how lung cancer potency varies with fiber type and fiber length. This
section summarizes these discussions, first on fiber type (Sections 3.1.1 and 3.1.2) and then on fiber
length (Sections 3.1.3 and 3.1.4). General issues regarding the lung cancer evaluation are presented in
Section 3.1.5.
3.1.1 Lung Cancer and Fiber Type: Inferences from the Epidemiology Literature
According to the proposed risk assessment methodology, amphibole fibers have a 5-fold greater lung
cancer potency than do chrysotile fibers. The panelists had differing opinions on whether this finding is
consistent with the epidemiology literature. On the one hand, some panelists indicated that the
epidemiology literature is consistent with amphibole fibers being more potent for lung cancer, though the
magnitude of this increase may not be known precisely. One panelist noted, for example, that multiple
analyses (e.g., Hodgson and Damton 2000, Berman and Crump 2001, and the statistical analyses a
panelist presented during this discussion) all point to a consistent increased lung cancer potency for
amphibole fibers compared to chrysotile fibers, albeit a small increase. On the other hand, other
3-1
-------
panelists did not believe Ihe epidemiology literature supports this conclusion, for reasons staled below.
Finally, olher panelists were not convinced lhat the epidemiology literature supports the higher lung
cancer potency for amphibole fibers, but they believed the difference in potency seems likely based on
evidence from the animal toxicology studies (see Section 3.1.3) and lung burden studies. A summary of
the panelists' discussion on this topic follows:
Comments on specific publications. Several panelists cited specific studies to support their
positions on the relative lung cancer potency of chrysotile and amphibole fibers, but the panelists
often had differing opinions on the inferences that should be drawn. The panelists mentioned the
following specific studies:
>• Some panelists noted that a recent re-analysis of 17 cohorts (Hodgson and Damton
2000) indicates that the lung cancer potency for amphibole fibers is 10 to 50 times
greater than that for chrysotile fibers. One panelist did not agree with this finding, due to
the crude approach the article uses to characterize relative potency. Specifically, this
panelist noted that carcinogenic potency was calculated by dividing the overall relative
risk for a given cohort by the average exposure for the entire cohort, even for cohorts
where the data support more sophisticated exposure-response modeling. He was
particularly concerned about the authors' decision to omit the cohort of South Carolina
textile workers from the meta-analysis. This decision was apparently based on the
South Carolina cohort being an outlier, due to its much higher lung cancer potency
when compared to other studies. The panelist noted, however, that the lung cancer risk
for the South Carolina cohort is not unusually high when compared to other cohorts of
textile workers. The panelist was concerned that omitting this study might have biased
the article's finding regarding relative lung cancer potency. No other panelists discussed
the review article.
«• One panelist cited a study of Quebec chrysotile miners and millers (Liddell et al. 1997,
1998) that reports that increased lung cancer risk was limited to the mining region with
the highest level of tremolite asbestos, after correction for smoking and exposure. The
article was distributed to the panelists on the first day of the workshop, but no panelists
commented further on the study.
" One panelist noted that his review of multiple textile cohorts (Stayner, Dankovic, and
Lemen 1996) found relatively small differences in lung cancer potency, even though
some of the cohorts were exposed to asbestos mixtures containing different proportions
of amphibole fibers.
3-2
-------
•• One panelist indicated that further evidence on how fiber types relates to lung cancer
potency can be gleaned from epidemiological studies that were not included in the
meta-analysis due to inadequate exposure data for exposure-response modeling.
Examples include a study of non-occupationally exposed women from two chrysotile
asbestos mining regions (Camus et al. 1998) and a study of railroad workers employed
by shops lhat processed different proportions of amphibole fibers (Ohlson et al. 1984).
Both studies, she noted, provide evidence that amphibole fibers exhibit greater lung
cancer potency. This panelist added that studies of auto mechanics have provided no
convincing evidence of increased lung cancer due to chrysotile exposure, though she
acknowledged that Ihe absence of an effect might reflect the short fiber length in the
friction brake products. One panelist cautioned about inferring too much from these
studies regarding fiber type because they were not controlled for other factors, such as
fiber length and level of exposure.
>• One panelist added that a recent study of a cohort of Chinese asbestos plant workers
(Yano et al. 2001) should be considered in future updates to the proposed protocol;
the workers in the cohort had increased risks for lung cancer and were reportedly
exposed to "amphibole-free" chrysotile asbestos. However, another panelist cited a
publication (Tossavainen et al. 2001) that indicates that asbestos from many Chinese
chrysotile mines actually does contain varying amounts of amphibole fibers.
•• Several panelists noted that the proposed protocol's meta-analysis found a 5-fold
difference in lung cancer potency between amphibole and chrysotile fibers. However,
other panelists indicated that the reported difference was not statistically significant.
Some panelists had additional reservations about the authors' meta-analysis, as
summarized in the following bulleted items.
Comments on the meta-analysis approach. Several panelists commented on alternate
approaches the authors could have used to conduct their meta-analysis of the epidemiology
studies. One panelist noted that the lung cancer potencies reported by the various studies exhibit
considerable heterogeneity. In such cases, meta-regression is conventionally used to identify
which factors account for the variability in the results (i.e., in the lung cancer potencies). This
panelist suggested that the meta-analysis should have considered other factors in addition to
fiber type and dimension; such other factors could include industry, follow-up time for the
cohort, and estimated percentage of amphibole fibers in the exposures, to the extent that data on
these other factors are available.
To demonstrate how more detailed investigation might reveal further insights, one panelist
presented his own initial statistical analysis of the epidemiological studies. This analysis used a
fixed effects model and a random effects model, both inverse weighted by the variance of the
studies. His analysis examined how industry and fiber type contribute to the heterogeneity
3-3
-------
observed among the cohorts and found that the industry of the cohort appears to be a stronger
predictor than fiber type. The panelist explained that the purpose of displaying his statistical
analysis was to highlight how other approaches to conducting meta-analysis can offer different
insights on the epidemiological data This panelist recommended that the authors conduct similar
meta-regression analyses to investigate the importance of various variables on the lung cancer
potency.
This panelist also demonstrated how a sensitivity analysis might yield additional information on
influential studies. Using a fixed effects model, the panelist first showed how lung cancer potency
factors (KL) vary with exposure to chrysotile fibers, amphibole fibers, and mixed fiber types.
When all epidemiological studies were considered in his analysis, the amphibole fibers were
found to be three times more potent than the chrysotile fibers. When the cohort of chrysotile
miners and millers from Quebec was omitted from this analysis, however, the amphibole fibers
were found to be nearly two times less potent than the chrysotile fibers. Conversely, when the
cohort of textile workers from South Carolina was omitted, the amphibole fibers were found to
be more than ten times more potent than the chrysotile fibers. Given that the conclusions drawn
about the relative potency of chrysotile and amphibole fibers appear to be highly sensitive to
whether single studies are omitted from the analysis, this panelist was more skeptical about
whether the increased potency of amphibole fibers is a robust finding. He recommended that the
authors, when completing the proposed protocol, conduct similar sensitivity analyses to help
reveal the factors or studies that appear to contribute most to lung cancer.
Another panelist agreed with this feedback, and provided further comments on the meta-
analysis, noting that these analyses typically start with establishing criteria for study inclusion.
After selecting studies to evaluate, she said, various statistical analyses can be used to test
hypotheses and to understand the concordance and disparity among the individual studies. The
panelist thought such an approach is needed to help understand the variability in potency factors
observed across the multiple studies and to identify for further analysis the studies found to be
most descriptive of exposure-response. To clarify the authors' approach, Dr. Herman indicated
that the meta-analysis considered any published epidemiological study with sufficient quantitative
exposure data that allowed for a reasonable estimate of the exposure-response relationship;
uncertainty factors were than assigned to give greatest weight to the most robust studies. In
response, additional panelists concurred with the original comment that meta-analyses
conventionally begin with establishing explicit study inclusion criteria These panelists clarified
that they are not advocating removing a majority of studies currently considered in the proposed
protocol, but rather being more judicious in selecting the studies to evaluate.
One panelist offered additional comments on the meta-analysis. He supported, for instance, the
use of sensitivity analyses, and encouraged the authors to conduct additional analyses to identify
influential studies, factors that contribute to risk, and the impact of different weighting factors.
The panelist also noted that more sophisticated statistical methodologies (e.g., Bayesian
3-4
-------
modeling, Markov Monte Carlo) can be used to generate distributions of outputs, rather than
discrete values, which might offer greater understanding of the inferences that can be drawn from
the epidemiological studies.
Disparate findings from the South Carolina and Quebec cohorts. Multiple panelists noted
that the issue of the relative lung cancer potency of chrysotile and amphibole fibers depends
largely on how one interprets the disparate findings from the cohort of textile workers in South
Carolina and 1he cohort of chrysotile miners and millers in Quebec. Two of these panelists
indicated that the relative potency issue likely will not be resolved until Ihe underlying reasons for
the differences between these two studies are better understood. The other panelist viewed Ihe
difference in potency observed across industries (i.e., mining versus textile) as a more important
matter than the difference between 1he two specific cohorts. When discussing these studies, two
panelists indicated that the increased lung cancer risk for the South Carolina cohort might be
attributed to exposure to amphibole fibers, which are known to be found in trace levels in
commercial chrysotile.
Relevance of fiber durability. One panelist noted that the issue of fiber durability often enters
the debate on the relative lung cancer potency of chrysotile and amphibole fibers. Though he
agreed that the animal toxicology data indicate lhat amphibole fibers are more persistent than
chrysotile fibers, the panelist noted that trends among the human epidemiological
data—particularly the fact that lung cancer risk does not appear to decrease with time since last
exposure, even for chrysotile—suggest that the lower durability of the chrysotile fibers might not
be important.
Influence of smoking. The panelists had differing opinions on how the proposed protocol
should address cigarette smoking. In terms of inferences drawn from the epidemiological
literature, two panelists noted lhat very limited data are available on smoking, making
quantitative analysis of its interactions wilh asbestos exposures difficult Specifically, only one
study includes detailed information on smoking, but that study found no difference in lung cancer
potency between smokers and non-smokers. During this discussion, Dr. Berman explained that
the proposed protocol assumes a multiplicative interaction between smoking and asbestos
exposure, consistent with EPA's 1986 model. Dr. Berman noted that a multiplicative factor in
the model, a, represents the background risk in the studied cohort relative to the risk in the
comparison population, and both groups include smokers; he added that the influence of
smoking is addressed implicitly in the model because it is a relative risk model in which the effect
of asbestos is multiplied to the background risk lhat is present. A panelist clarified, however, that
neither the potency factors nor a were derived based on observations of smoking prevalence in
the epidemiological studies.
One panelist emphasized that the confounding effects of smoking greatly complicates the analysis
of lung cancer potency. He noted that the relative lung cancer risk from asbestos exposure is
3-5
-------
considerably lower than that for cigarette smoking. As a result, the panelist wondered how the
meta-analysis can truly discern the relative potency of the asbestos fiber types from studies that
present no information on cigarette smoking. This panelist provided an example to illustrate his
concern: if a given cohort has between 5 and 10% more smokers than the typical population,
this increased prevalence of smoking alone could totally confound relative risks attributed to
asbestos. The panelist indicated lhat all future analyses of epidemiological data will suffer from
similar limitations, so long as detailed information on smoking is not available.
• General comments. During this discussion, some panelists offered several general comments
that apply to the entire proposed protocol. These comments included concerns about the
transparency of the analyses, questions about data tables being inconsistent with text in the body
of 1he report, and some panelists' inability to reproduce certain findings from the available data
These general comments are reflected in the executive summary of this report
3.1.2 Lung Cancer and Fiber Type: Inferences from Animal Toxicology and
Mechanistic Studies
The panelists offered varying insights on the inferences that can, or should, be drawn from animal
toxicology studies and mechanistic studies regarding the relative lung cancer potency for chrysotile and
amphibole fibers.
Citing various publications (e.g., Lippmann 1994), multiple panelists noted that the animal toxicology
studies do not support the 5-fold difference in lung cancer potency between chrysotile and amphibole
fibers. Two panelists added that the absence of different potencies might result from the animal studies
being of too short duration (typically no longer than 2 years) for the greater dissolution of chrysotile
fibers to be an important factor. Another panelist added that exposure levels in some animal studies are
not relevant to human exposures; as an example, he noted that a recent rat inhalation study (Hesterberg
et al. 1998) involved exposure levels at 11,000 fibers per cubic centimeter. These panelists indicated
that the animal studies are generally more informative of how lung cancer potency varies with fiber
length (see Section 3.1.4), and are less informative on how potency varies with fiber type.
3-6
-------
The panelists noted that in vitro studies exhibit various findings, depending on the study design and
endpoint assessed. One panelist, for instance, indicated that some in vitro studies suggest that
chrysotile fibers are actually more potent than amphibole fibers. Other panelists added that many in
vitro studies show crocidolite being considerably more toxic than chrysotile. These panelist cautioned
against drawing firm conclusions from the in vitro studies, however, given that the study duration is far
too short for any impact of dissolution to be observed. Finally, another panelist referred to the
International Agency for Research on Cancer (IARC) consensus statement on fiber carcinogenesis for
an overview of inferences that can be drawn from mechanistic studies: "Overall, the available evidence
in favor of or against any of these mechanisms leading to the development of lung cancer and
mesothelioma in either animals or humans is evaluated as weak" (IARC 1996).
Based on Ihe previous comments, the panelists cautioned about attempting to draw inferences from the
animal toxicology for several reasons. One panelist indicated that the animal studies have limited utility
because lung cancer in humans results from a complex set of exposures, including cigarette smoke, and
because rats, when compared to humans, develop different types of tumors at different sites. Another
panelist reiterated that the duration of most animal studies precludes one from observing dissolution
effects. Given these limitations, two panelists emphasized that conclusions should be based primarily on
the epidemiological data, especially considering the volume of human data that are available. Though
not disagreeing with this recommendation, one panelist noted that the exposure index—one of the
major outcomes of the proposed protocol—is, in fact, based on observations from animal studies.
3.1.3 Lung Cancer and Fiber Dimension: Inferences from the Epidemiology
Literature
The panelists made several observations regarding what can be inferred from the epidemiology
literature on how lung cancer potency varies with fiber dimension, though they first noted that most
published epidemiology studies do not include detailed data on the distribution of fiber dimensions to
which cohorts were exposed. Overall, the panelists generally agreed that indirect evidence from the
3-7
-------
epidemiological studies supports the proposed protocol's finding that longer fibers have greater
carcinogenic potency for lung cancer. They added, however, that the epidemiology literature provides
no evidence to support or refute the magnitude of the relative potencies used in the proposed protocol
(i.e., fibers longer than 10 ^m being 300 times more potent than those with lengths between 5 and 10
nm). The panelists made no comments about fiber diameter when discussing this matter. Specific
discussion topics follow:
Observations from the epidemiology literature. The panelists identified several studies that
provide general insights on the role of fiber size in lung cancer. One panelist, for instance, noted
that cohorts of textile workers, which were believed to be exposed to relatively longer asbestos
fibers, exhibit higher lung cancer relative risks than do cohorts of miners or cement product
workers. Another panelist indicated that studies of taconite miners from Minnesota (Cooper et
al. 1988) and gold miners from South Dakota (McDonald et al. 1978) found no increased lung
cancer risks among the cohorts, which were known to be exposed primarily to fibers shorter
than 5 |im (see Dr. Case's premeeting comments for further information on these studies). This
panelist added that the Minnesota Department of Health is currently updating the study on
taconite miners and a publication is pending. Another panelist added that epidemiology studies
of workers exposed to asbestos from friction brake products show no clear evidence of
increased lung cancer. This panelist acknowledged that these epidemiology studies do not
include exposure measurements, but other studies of this work environment have indicated that
the asbestos fibers in friction brake products are predominantly short chrysotile fibers.
Relevance of fibrous structures shorter than 5 \im. Some panelists noted that no
epidemiology studies have examined the relative potency specifically of fibrous structures shorter
than 5 urn, thus no conclusions could be drawn from the epidemiology studies alone. While not
disagreeing with this observation, one panelist reminded panelists that airborne particles and
fibers have a broad distribution of fiber lengths, with a clear majority (75-90%) of fibrous
structures being shorter than 5 urn. This panelist added that indirect inferences can be drawn
from the epidemiology studies listed in the previous bulleted item. Another panelist noted that the
fibrous structures shorter than 5 ^m behave more like particles rather than fibers, at least in
terms of lung deposition and clearance patterns. Finally, two panelists indicated that an ATSDR
expert panel recently evaluated the issue of relative potency of fibers shorter than 5 \im,
however, the final report from that expert panel meeting was not available until after the peer
consultation workshop. The final report has since been released, and a conclusion from that
panel was that "there is a strong weight of evidence that asbestos and synthetic vitreous fibers
shorter than 5 ^m are unlikely to cause cancer in humans" (ERG 2003).
3-8
-------
Statistical analyses in the proposed protocol. As indirect evidence lhat longer fibers have
greater carcinogenic potency, one panelist indicated that the exposure-response modeling by
Drs. Berman and Crump showed an improved fit to the observed relative risk from
epidemiology studies when using an exposure index that assigns greater weight to longer fibers
and no risk to fibers shorter than 5 nm Another panelist concurred, but added lhat the authors
could have attempted to determine the specific weighting (i.e., between longer and shorter
fibers) that would optimize the fit to the epidemiological studies.
3.1.4 Lung Cancer and Fiber Dimension: Inferences from Animal Toxicology and
Mechanistic Studies
The panelists generally agreed that the animal toxicology studies and mechanistic studies indicate that
fiber dimension—especially fiber length—plays an important role, both in terms of dosimetry and
pathogenesis. However, panelists had differing opinions on the specific cut-offs that should be used for
fiber diameters and lengths in the exposure-response modeling (though panelists generally concurred
that fibers shorter than 5 ^m should be assigned zero potency).
Fiber length. Multiple panelists noted that the animal toxicology studies provide compelling
evidence that lung cancer potency increases with fiber length. Another panelist agreed, but had
reservations about assigning no potency to fibrous structures shorter than 5 |im, based on a
recent study of refractory ceramic fibers (Bellman et al. 2001) that found that the incidence of
inflammation and fibrosis appears to be related to the presence of small fibers in the lung. This
panelist indicated that exposure to small fibers likely has some bearing on the oxidative stress
state and inflammation in the lung, and he suspected that the exposure-response relationship for
long fibers might depend on co-exposures or past exposures to shorter fibers. Based on these
observations, the panelist was hesitant to exclude fibrous structures shorter than 5 urn from the
proposed risk assessment methodology. On the other hand, another panelist added that animal
toxicology studies have shown that fibrosis endpoints are strongly related to fiber length, wilh
exposures to shorter fibers showing less evidence of fibrosis or lung damage. The panelists
revisited the significance of fibers shorter than 5 p.m when discussing the proposed exposure
index (see Section 4).
Fiber diameter. The panelists offered several comments on the role of fiber diameter in the
proposed protocol. Noting that fibers with diameters up to 1.5 |j,m are capable of penetrating to
sensitive portions of the lung during oral inhalation, one panelist indicated that this range of fiber
diameters should not be excluded from future risk assessments. Other panelists shared the
3-9
-------
concern of assigning no lung cancer potency to respirable fibers with diameters greater than 0.5
urn, especially considering that respirabilily patterns in laboratory animals differ from those in
humans (i.e., thicker fibers are more likely to deposit in the human lung than they are in the rat
lung).
The panelists also discussed a statement in the proposed protocol that "few fibers thicker lhan
0.7 |im appear to reach the deep lung." First, one panelist indicated that the proposed protocol
includes outdated information on fiber deposition patterns; he recommended that the authors
obtain more current insights from specific publications (e.g., Lippmann 1994) and from the latest
lung dosimetry model developed by the International Commission on Radiological Protection.
Second, another panelist questioned the relevance of deposition in the deep lung, because
humans tend to develop bronchogenic carcinomas, while rats develop bronchoalveolar
carcinomas. Another panelist cautioned against inferring that asbestos fibers must deposit on
bronchial airways to cause lung cancer in humans, noting lhat significant accumulation of
asbestos fibers does not occur in Ihe airways where carcinomas develop in humans, due
primarily to mucociliary clearance; this panelist suspected that deposition of fibers in the deep
lung is likely related to lung cancer formation in humans, though the mechanisms of
carcinogenesis are not fully understood.
3.1.5 Other Issues Related to Lung Cancer
The panelists discussed several additional issues related to the proposed protocol's evaluation of lung
cancer potency. Most of the discussion focused on the utility of non-linear exposure-response
modeling, but other topics were also addressed:
Consideration of non-linear exposure-response models. The panelists had differing
opinions on the extent to which Ihe proposed protocol should consider non-linear exposure-
response modeling. On the one hand, one panelist strongly recommended that EPA consider
exploring Ihe applicability of non-linear exposure-response models, given his concerns with
linear low-exposure extrapolation. This panelist acknowledged that the revised linear model in
the proposed protocol clearly provides an improved statistical fit to the epidemiological data
when compared to EPA's 1986 lung cancer model, but he advocated more detailed exploration
of non-linear cancer risk models, particularly to account for observations of cohorts with low
exposures. This panelist was particularly concerned about the cancer risks that would be
predicted for low exposures: because the slope in any linear lung cancer model will be
determined largely by highly-exposed individuals, he questioned whether the slope derived from
3-10
-------
high exposures truly applies to lowly-exposed individuals. To demonstrate his concern, Ihis
panelist indicated that the epidemiologjcal studies consistently show that cohorts (or subsets of
cohorts) with low exposure generally exhibit no increased lung cancer risk (standardized
mortality ratios not statistically different from 1.0). To account for the possibility of a threshold
or non-linearity in the exposure-response relationship, this panelist recommended lhat EPA
investigate alternate exposure-response models, such as linear-linear models (i.e., models with
two linear exposure-response regions having different slopes) or log-linear models.
Other panelists generally supported these comments. One panelist, for instance, noted that
EPA's Draft Revised Guidelines for Carcinogen Risk Assessment indicates that exposure-
response relationships should first be evaluated over the range of exposure observations, and
then various approaches to extrapolate to exposure levels outside (i.e., below) this range should
be investigated. Another panelist added that some studies finding no evidence of lung cancer
risks among large cohorts with low exposures should factor into the decision of whether the lung
cancer model should include thresholds; he cited a study of non-occupationally exposed women
from chrysotile mining regions in Canada (Camus et al. 1998) to illustrate his concern. Other
panelists noted that the utility of Ihis study is limited, because exposures were not measured for
individuals; further, a panelist clarified that approximately 5% of the individuals considered in this
study were occupationally exposed. Finally, one panelist indicated lhat evidence from the
epidemiology literature strongly suggests there are asbestos exposure levels below which lung
cancer will not occur; this panelist added that he is unaware of any epidemiological study that
has found evidence of lung cancer risk at exposure levels below 25 fiber-years. He
recommended that the proposed protocol at least acknowledge the lowest exposure level at
which lung cancer effects have been demonstrated.
On Ihe other hand, some panelists were not convinced of the utility of conducting detailed
analyses at low exposures and investigating possible thresholds. One panelist, for instance,
indicated that a meaningful quantitative analysis of potential thresholds will not be possible, so
long as the authors do not have access to raw data from additional epidemiological studies.
Further, this panelist suspected that the protocol authors would find considerable heterogeneity
among exposure-response slopes for low exposures, and he questioned what conclusions could
be drawn by focusing exclusively on the low exposure region Another panelist agreed, adding
lhat Ihe failure to find significantly increased cancer risks among lowly-exposed cohorts very
likely results from poor statistical power and other uncertainties, and not necessarily from the
presence of an actual exposure threshold for asbestos-related lung cancer. Finally, one panelist
indicated that the National Institute for Occupational Safety and Health (NIOSH) previously
examined a threshold model for Ihe cohort of South Carolina textile workers, and that analysis
revealed that the best fit of the exposure-response data was a threshold of zero (i.e., the best fit
indicated that there was no threshold).
3-11
-------
Consideration of cigarette smoking. Several times during the workshop, the panelists
debated the ability of the proposed risk assessment model to address interactions between
cigarette smoking and asbestos exposure. One panelist recommended that the authors review a
recent study that examined the role of cigarette smoking on lung cancer among chrysotile miners
and millers in Quebec, Canada (Liddell and Armstrong 2002). Although the panelists generally
agreed that smoking is an important consideration for developing and applying the model, some
panelists were not convinced that the available data are sufficient to develop an exposure-
response model that accurately portrays the interactive effects of asbestos exposure and
smoking. The panelists further discussed this issue further later in the workshop.
Transparency of the proposed protocol Several panelists indicated that the review of
epidemiological data in the proposed protocol is not presented in a transparent fashion. One
panelist, for instance, sought more information on the uncertainty factors used in the meta-
analysis, such as what ranges of factors were considered, what criteria were used to assign the
factors, and a table of the factors that were eventually applied. This panelist also recommended
that the proposed protocol identify the a-values that were determined for each epidemiological
study and provide explanations for any cases when these values are unexpectedly large. Another
panelist indicated that the proposed protocol should more clearly differentiate conclusions that
are based on a meta-analysis of many epidemiological studies from conclusions that are based
on a detailed review of just one or two studies.
The need to obtain additional raw data sets. The panelists unanimously agreed that EPA
should make every effort to try to obtain additional raw data sets for the epidemiology studies,
such that the authors can further test how adequately the proposed risk assessment model
predicts risk. The executive summary of this report presents the panelists' specific
recommendation on this issue.
3.2 Mesothelioma
The following paragraphs document the panelists' responses to charge questions regarding inferences
from the epidemiology and toxicology literature on how mesothelioma potency varies with fiber type
(Sections 3.2.1 and 3.2.2) and fiber length (3.2.3 and 3.2.4).
3.2.1 Mesothelioma and Fiber Type: Inferences from the Epidemiology Literature
3-12
-------
The expert panelists unanimously agreed that the epidemiology literature provides compelling evidence
that amphibole fibers have far greater mesothelioma potency than do chrysotile fibers—a finding
reported both in the review document (Berman and Crump 2001) and a recent re-analysis of 17 cohort
studies (Hodgson and Damton 2000) that reported at least a 500-fold difference in potency. Two
panelists commented further that the epidemiology literature provides no scientific support for chrysotile
exposures having a role in causation of mesothelioma—an observation that is generally consistent with
the meta-analysis in the proposed protocol, which failed to reject the hypothesis that chrysotile fibers
have zero potency for mesothelioma
The most notable response to this charge question was the agreement among most panelists that
amphibole fibers are at least 500 times more potent than chrysotile fibers for mesothelioma, as
supported by two separate reviews of epidemiologjcal studies. The panelists made additional comments
on specific matters when responding to this question, as summarized below, but the key point in this
discussion was the agreement that chrysotile is a far less important cause of mesothelioma than are
amphiboles.
Relative roles of chrysotile and amphibole. One panelist indicated that cohort studies with
individual-level exposure-response data and the broader epidemiology literature both provide no
evidence of increased mesothelioma risk due to chrysotile exposure. Further, this panelist noted
that 33 of 41 mesothelioma cases previously identified as occurring among workers primarily
exposed to chrysotile fibers (Stayner et al. 1996) were later reported as likely resulting from
exposures to tremolite fibers found in the chrysotile mines (McDonald et al. 1997). This panelist
noted that a recent finding of a small mesothelioma risk from chrysotile (Hodgson and Damton
2000) results entirely on the assumption that the 33 mesothelioma cases mentioned above result
entirely from chrysotile exposures. Based on these observations, this panelist indicated that the
literature suggests that chrysotile exposures have limited, if any, role in causing mesothelioma He
nonetheless supported the relative potency attributed to chrysotile in the proposed protocol as a
conservative measure in the overall risk assessment process.
Specific comments on the Connecticut friction products workers. Another panelist
commented on an epidemiological study of a cohort of workers employed at a friction products
plant in Connecticut The panelist noted that the original study (McDonald et al. 1984) did not
identify any deaths frommesolhelioma, but review of the state cancer registry (Teta et al. 1983)
3-13
-------
revealed that Ihree Connecticut residents who died of mesothelioma were employed by the
same friction products company. One of these employees had amphibole exposures during the
time he worked for a textile plant that was under the same parent company that owned and
operated the friction products plant The other two cases, the panelist noted, were females who
indeed worked at the friction products plant. A pathology review found that one of these cases
was a woman with probable pleura! mesothelioma and 5 years of exposure; the other case was
a peritoneal mesothelioma in a woman who also had asbestosis, and worked as a clerk for 30
years. This panelist noted that it was questionable to attribute the latter two mesothelioma
diagnoses to Ihe chrysotile exposures at the friction products plant, though she added that this
possibility cannot be definitively ruled out This panelist encouraged that future review of this
epidemiologjcal study should be revised given Ihis new information.
Comments on the proposed 500-fold difference in relative potency. The panelists had
several comments on the finding in the proposed risk assessment methodology that amphibole
fibers are 500 times more potent for mesothelioma than are chrysotile fibers. Several panelists
noted lhat this finding is consistent with lhat of a recent re-analyses of 17 epidemiological studies
(Hodgson and Damton 2000). Though not disagreeing lhat amphibole fibers are clearly more
potent, one panelist was concerned lhat Ihe risk coefficients (KM) were largely derived from
data sets with inadequate exposure-response information for mesothelioma, and assumptions
had to be made to determine critical inputs to the mesothelioma model (e.g., average exposure,
duration of exposure).
Other panelists commented on specific sections in the proposed protocol. One panelist, for
example, recommended that the authors check the accuracy of data presented in Table 6-16
and Table 6-29 of the report, which are not reported consistently. Another panelist suggested
that the authors better explain why separate risk coefficients for amphiboles and chrysotile were
calculated for some cohorts (e.g., Hughes et al. 1987) but not for others (e.g., Berry and
Newhouse 1983), even though the exposure information available for the studies appears to be
comparable. Finally, one panelist recommended that the authors of the proposed protocol
consider questions recently raised (Rogers and Major 2002) about the quality of the exposure
data originally reported for the Wittenoom cohort (De Klerk et al. 1989) when evaluating
exposure-response relationships for mesothelioma
3.2.2 Mesothelioma and Fiber Type: Inferences from Animal Toxicology and
Mechanistic Studies
The panelists discussed the inferences provided by animal toxicology data and mechanistic data
regarding relative mesothelioma potency of different asbestos fiber types. Overall, two panelists
3-14
-------
commented that Ihe human epidemiological data clearly establish that exposures to amphibole asbestos
fibers pose a greater mesothelioma risk than do exposures to chrysotile fibers. They added that the
animal toxicology data are generally supportive of this finding, but the animal data suffer from some
limitations. Two panelists, for instance, noted that Ihe utility of animal toxicology studies is limited by the
fact that rodents are rather insensitive to mesothelioma These panelists added that the animal
toxicology studies involving intra-tracheal instillation or peritoneal injection are not directly relevant to
the inhalation exposures that occur in humans. These limitations notwithstanding, the panelists raised the
following points when discussing the animal toxicology and mechanistic studies:
One panelist referred to one of his earlier publications (Lippmann 1994) for further insights on the
occurrence of mesolhelioma in animal studies. At that time, this panelist noted, the animal inhalation
studies found fewer than 10 cases of mesothelioma, and the number of cases appeared to be greatest
among animals that were exposed to mixtures containing higher proportions of amphibole fibers. He
found this consistent with the influence of fiber type observed in the human epidemiological data (see
Section 3.2.1).
During this discussion, one panelist reviewed a publication (Suzuki and Yuen 2001) that was mentioned
earlier in the workshop. The publication documents the amounts and types of asbestos fibers measured
in samples of pleura! plaques and tumor tissue collected for legal cases. These analyses reportedly
found relatively large amounts of short, thin chrysotile fibers in the pleura, suggesting that these fibers
should not be excluded from the group of fibers believed to induce mesolhelioma The panelist had
several criticisms of the study. First, he indicated that the samples were analyzed using a non-standard
technique, without any controls. Second, he questioned the major finding of fibers being detected in the
pleura, because most of the samples analyzed were actually tumor tissue, in which he would not expect
to find fibers. The panelist suspected that the chrysotile fibers reportedly found in the study likely result
from specimen contamination—a bias that would have been more apparent had rigorous quality control
procedures been followed. Finally, the panelist noted that a more rigorous study (Boutin et al. 1996) of
3-15
-------
asbestos fibers in the parietal pleura found a mixture of fibers, including long amphibole fibers, among
living patients with asbestos-related conditions. Based on these concerns, the panelist concluded that
the publication of concern (Suzuki and Yuen 2001) is seriously flawed and its recommended should be
excluded from EPA's analyses.
A specific issue raised regarding the analytical technique in the study (Suzuki and Yuen 2001) was that
water was used during the digestion process. Noting that water may contain large amounts (>30,000
fibers/L) of small asbestos fibers, another panelist suspected that Hie fibers detected in the study might
have resulted from contamination introduced during the digestion process. Because control samples
were not analyzed, Ihe panelist said the study offers no evidence that the fibers detected truly were in
the original pleura! plaques or tumor tissues. He added that studies of lung-retained asbestos fibers
routinely detect primarily short, chrysotile fibers, and that the presence of the short fibers in the pleura!
tissue—even if the measurements from the study are valid—would not necessarily prove that short
fibers cause mesolhelioma
3.2.3 Mesothelioma and Fiber Dimension: Inferences from the Epidemiology
Literature
The panelists commented briefly on how the human epidemiological data characterize the role of fiber
size on mesothelioma risk Noting that exposure measurements in most every epidemiological study do
not characterize fiber length distribution, one panelist indicated that these studies provide no direct
evidence of how fiber length is related to mesolhelioma He added that the studies offer conflicting
indirect evidence of the role of fiber length. Specifically, the higher mesolhelioma risk coefficient among
textile workers in South Carolina, when compared to that for the chrysotile miners and millers in
Quebec, could be supportive of longer fibers being more potent, since exposures in South Carolina had
a larger percentage of long fibers. However, a cohort of cement plant workers in New Orleans was
found to have a higher mesolhelioma risk coefficient than that of the South Carolina cohort, even though
the South Carolina workers were exposed to higher percentages of long fibers. Finally, as indirect
3-16
-------
evidence that carcinogenic potency increases with fiber length, this panelist noted that the mesolhelioma
risk model using the proposed exposure index, which is heavily weighted by long fibers, provided-a
considerably improved fit to the epidemiological data
Hie panelists briefly revisited the inferences that can be drawn from studies of lung-retained fibers. One
panelist again commented that results from a recent study (Suzuki and Yuen 2001) should be viewed
with caution. He added that several other lung pathology studies (e.g., McDonald et al. 1989, Rogers et
al. 1991, Rodelsperger et al. 1999) have been conducted using more rigorous methods, such as using
appropriate controls for age, sex, and hospital. These studies all showed that risk of mesothelioma was
considerably higher for individuals with larger amounts of long fibers retained in their lungs.
One panelist indicated that results from a study of lung-retained fibers (Timbrell et al. 1988) suggest
fiber diameter plays a rule in mesothelioma risk: the study observed no mesothelioma cases among a
population highly exposed to anthophyllite fibers, which tend to be thicker fibers. Citing his earlier
review of mesothelioma cases (Lippmann 1988), the panelist also noted that crocidolite fibers are both
thinner than and more potent than amosite fibers, which further supports the hypothesis that
carcinogenic potency for asbestos decreases with increasing fiber diameter.
3.2.4 Mesothelioma and Fiber Dimension: Inferences from Animal Toxicology and
Mechanistic Studies
The panelists made few observations on findings from animal toxicology studies regarding mesolhelioma
and fiber length. One panelist indicated that findings from the animal toxicology studies generally
support the overall finding that mesolhelioma risks are greatest for long, thin fibers. However, another
panelist noted that his earlier review of mesothelioma risks (Lippmann 1988) hypothesized that the
critical fibers for mesolhelioma induction are those with lengths between 5 and 10 (im This panelist
added that fibers of this dimension are more likely to translocate to the pleura than are longer fibers, but
3-17
-------
he acknowledged that it is unclear whether fibers must first translocate to the pleura in order to cause
mesothelioma
Some panelists indicated that fiber durability likely plays a role in inducing mesothelioma, based on the
fact that mesothelioma is more easily induced in animals using administration methods (e.g., peritoneal
injection) that remove Ihe importance of dissolution.
3.3 Exposure Estimates in the Epidemiology Literature
0
The panelists raised numerous issues when responding to the third charge question: 'To what extent are
Ihe exposure estimates documented in the asbestos epidemiology literature reliable?" Recognizing that
the exposure estimates from the epidemiology studies are critical inputs to Ihe exposure-response
assessment, the panelists expressed concern about the exposure data: few studies provide detailed
information on fiber size distribution; many studies report exposures using outdated sampling and
analytical methodologies (e.g., midget impinger); individual-level data are not available for most studies;
and many studies do not report detailed information on parameters (e.g., exposure levels, exposure
duration) needed to evaluate exposure-response relationships, particularly for mesolhelioma Their
specific concerns on these and olher matters follow:
Concerns regarding exposure estimates in specific studies. Some panelists expressed
concern about the assumptions made to interpret the exposure data originally reported in the
epidemiology studies. One panelist reviewed specific examples of these concerns:
> The original study of workers at a Connecticut friction products plant (McDonald et al.
1984) reports exposures measured by midget impingers (in units of mmpcf), with no
information on how to convert this to PCM measurements, and the original publication
includes limited data on exposure duration.
•• The original study of workers at a New Jersey insulation factory (Seidman et al. 1986)
did not report any exposure measurements from the factory studied, and data collected
3-18
-------
from another plant with similar operations were used to characterize exposure-response
for Ihis cohort
>• The original study of workers at a Texas insulation factory (Levin et al. 1998) reported
a range of exposure levels (15-91 fibers/mL), and the authors of the proposed protocol
assigned an average exposure level (45 fibers/mL) to the entire cohort.
>• The original study of U.S. insulation applicators (Selikoff and Seidman 1991) has no
information on exposure. The proposed protocol assumes lhat all workers were
exposed to 15 fibers/mL for 25 years, based on a separate review of exposures among
insulation workers (Nicholson 1976).
•• The original study of retirees from the U.S. Asbestos Products Company (Enterline et
al. 1986) reported exposures based on midget impinger sampling, with no information
on how to convert these exposures to PCM measurements.
>• According to a recent letter to the editor (Rogers and Major 2002), the original study
of the Wittenoom cohort (De Klerk et al. 1989) might have overestimated exposures,
possibly by as much as a factor of 10.
The previous comments led to a discussion on whether certain studies should be excluded from
the meta-analysis used in the proposed protocol (see next bulleted item). Prior to this discussion,
one panelist expressed concern about being overly critical of the exposure estimates used for
many of the studies listed above; he emphasized that all exposure estimates appear to be based
on a critical review of the literature, and no estimates are completely arbitrary, as some of the
panelists' comments implied.
Comments on using study inclusion criteria for the meta analysis. Given the concerns
about the quality of exposure data reported in some epidemiology studies, the panelists debated
whether future revisions of the proposed protocol should exclude certain studies from the
exposure-response analysis. The panelists were divided on this matter.
On the one hand, several panelists recommended that the authors develop and apply study
inclusion criteria in the exposure-response evaluation, as is commonly done when conducting a
meta-analysis. One panelist, for instance, recommended assessing exposure-response
relationships for only those studies found to have adequate exposure data, and then using a
sensitivity analysis to examine the effect of excluding studies with inadequate exposure data
These panelists clarified that they are not advocating disregarding the majority of studies; rather,
they are suggesting simply that the authors of the proposed protocol use study inclusion criteria
and sensitivity analyses to ensure that the conclusions are based on the best available exposure
data.
3-19
-------
On the other hand, several panelists supported the current approach of using as many studies as
possible and accounting for the quality of the exposure measurements in the uncertainty factors.
One panelist, for example, commended the authors for being as inclusive as possible when
reviewing the studies; he supported the approach of recognizing the limitations of the available
exposure data and accounting for these limitations in the uncertainty factors that were ultimately
used to weight the studies in Ihe meta-analysis. This panelist acknowledged that the exposure
estimates in some of the epidemiological studies might be rough estimates, but he emphasized
that the estimates are not worthless and should not be discarded. Olher panelists concurred with
these comments, and did not support applying overly restrictive study inclusion criteria
Comments on the uncertainty factors assigned to each study. The panelists made several
comments on the uncertainty factors that the authors assigned to each study. Dr. Berman first
explained the four uncertainty factors: the first factor (Fl) characterizes the confidence in
exposure estimates; the second factor (F2) represents the confidence in the conversion to PCM
measurements from other exposure metrics (typically midget impinger analyses); the third factor
(F3) characterizes the confidence the authors had on worker history data; and the fourth factor
(F4) was a non-exposure related factor to account for other uncertainties (e.g., lack of
information on confounders, incomplete or inaccurate mortality ascertainment). Dr. Berman
described generally how the individual uncertainty factors were assigned and noted that each
factor could range from 1 to 5.
The panelists' comments primarily focused on the transparency of how uncertainty factors were
presented and incorporated into the meta-analysis. Multiple panelists, for instance,
recommended that future revisions to the proposed protocol include a table that lists the
uncertainty factors assigned to each study. Further, one panelist suggested that the revised
protocol describe the assumptions inherent in the uncertainty factor weighting approach, such as
explaining why some factors are assigned values over a broader range than others (e.g., why Fl
values span a broader range than F4 values) and describing why the individual uncertainty
factors have equal weights in generating the composite uncertainty factor. Another panelist
agreed, and added that the revised protocol should more explicitly describe how the uncertainty
factors were combined into the composite factor and how this composite factors affects the
weighting of studies in the meta-analysis. Expanding on this point, another panelist suggested that
the final document more clearly explain that the final estimates of cancer risk coefficients (KL*
and KM*) are actually weighted averages of the epidemiological studies, with the weights
assigned to each study being a function of that study's uncertainty. This panelist also
recommended that the revised document clearly state how, if at all, the fraction of amphibole
fibers and the fraction of fibers longer than 10 |im are reflected in the uncertainty factors.
Some panelists debated the utility of alternate approaches that could be used to assign
uncertainty factors. Two panelists noted that the approach used to assigning uncertainty factors
is somewhat subjective, because different groups of analysts would likely assign different
3-20
-------
uncertainty factors. To avoid the appearance of arbitrariness, these panelists suggested using
alternate meta-analysis approaches that do not require using uncertainty factors. They noted, for
example, that the authors could use a random effects model in which residual inter-study
variation is estimated. Another suggestion was to conduct sensitivity analyses examining the
effects of including or excluding studies, depending on the uncertainty factors assigned to them
Another panelist disagreed with these comments and supported the analyses in the proposed
protocol; this panelist indicated that the authors had no choice but to make judgments based on
the information documented in the epidemiology literature. He suggested that EPA consider
convening a separate expert panel to assign uncertainty factors, if panelists do not support those
selected by Drs. Berman and Crump.
Assumptions made to convert exposure estimates from midget impinger sampling.
Several panelists noted that the original publications for many epidemiology studies document
exposure estimates based only on midget impinger sampling and do not include any information
on how to convert these exposures to levels that would be measured by more modem methods
(e.g., PCM, TEM). The panelists noted that the conversion factor (from mmpcf to fibers/mL)
can vary considerably from one occupational setting to the next
Interpretations of the study of South Carolina textile workers. The panelists had different
opinions on interpretations of the study of South Carolina textile workers (Dement et al. 1994).
One panelist, for instance, found this particular study to be an outlier among the other
epidemiological studies, and he recommended that the authors exclude this study from the
exposure-response analysis until the causes for the increased relative risks observed for this
cohort are better understood. Another panelist suggested that the proposed protocol should
classify the South Carolina cohort as being exposed to mixed asbestos fibers, rather than being
exposed to chrysotile fibers. He indicated that some workers in the cohort were exposed to
amosite and crocidolite, in addition to being exposed to chrysotile.1
Other panelists, however, did not think the South Carolina study should be excluded from
EPA's analysis. One panelist was troubled about criticisms of the exposure estimates for this
cohort, given that this is one of few studies in which co-located samples were collected and
analyzed using different methods, thus providing site-specific data for converting midget impinger
1 After reviewing a draft of this report, one panelist indicated that it is important to note that exposure data
for the South Carolina cohort are available from more than just one reference (Dement et al. 1994). He suggested that
EPA use data from studies conducted by McDonald in the 1980s of a parallel cohort in the same plant. However, he
cautioned EPA against treating multiple studies of the same relatively small group of workers as separate studies,
considering the large overlap of workers studied by the two groups of investigators. This panelist encouraged EPA
to consider other data sources for this cohort, given that a recent re-analysis of epidemiological studies (Hodgson
and Darnton 2000) severely criticized the data source EPA uses (Dement et al. 1994), to the point of those data being
dropped from the recent re-analysis altogether.
3-21
-------
sampling results to PCM measurements. Another panelist challenged suggestions lhat the South
Carolina study is an outlier; he indicated that the South Carolina study is one of the more
rigorous epidemiology studies available for asbestos exposures, and he found no valid scientific
reasons for discarding it. During this discussion, one panelist point out in response that the South
Carolina study is indeed an outlier among the textile cohorts, with a slope which is higher than
either of the two textile cohorts; this panelist did acknowledge that Ihe lung cancer risk among
the textile cohorts is greater than that among the mining cohorts. This panelist added that
scientists need a better explanation for why the lung cancer risk among the South Carolina
cohort is greater than that of other cohorts before the South Carolina study can achieve
credibility, especially considering that exposures in South Carolina were supposedly to "pure"
chrysotile.
3-22
-------
4. COMMENTS ON TOPIC AREA 2: THE PROPOSED EXPOSURE INDEX
This section summarizes Ihe panelists' responses to the charge questions pertaining to the proposed
exposure index. Section 4.1,4.2, and 4.3 document the panelists' responses to charge questions 4, 5,
and 6, respectively.
4.1 Responses to Charge Question 4
Charge question 4 asks: "The proposed exposure index does not include contributions from fibers
shorter than 5 ^im Please comment on whether the epidemiology and toxicology literature support the
conclusion that asbestos fibers shorter than 5 |im present little or no carcinogenic risk." The panelists
discussed this matter earlier in the workshop (see Sections 3.1.3 and 3.1.4 for these comments), and
provided additional insights on the matter. Overall, the panelists agreed that carcinogenic potency
increases with fiber length, particularly for lung cancer. Most panelists supported assigning no potency
to fibrous structures smaller than 5 ^m Some panelists agreed that the short fibrous structures are
clearly less potent than long fibers, but they had reservations about assigning zero potency to the
structures smaller than 5 ^m; these panelists acknowledged that the toxicity of the short fibrous
structures might be adequately addressed by EPA's air quality standards for particulate matter. Specific
comments on this charge question follow:
Reference to ATSDR's expert panel workshop on the role of fiber length. Two panelists
noted that ATSDR convened an expert panel in October 2002 to discuss the role of fiber length
on toxicity, and much of that discussion specifically addressed fibrous structures smaller than 5
nm A main conclusion of that panel was that there is "a strong weight of evidence that asbestos
and synthetic vitreous fibers shorter than 5 jam are unlikely to cause cancer in humans" (ERG
2003). The panelists encouraged EPA to review the summary report prepared for that
workshop, which was officially released on March 17,2003, and is available on-line at:
www. atsdr. cdc.gov/HAC/asbestospanel.
Evidence from epidemiological studies. One panelist indicated that the epidemiological
studies do not provide direct evidence of Ihe role of fibrous structures shorter than 5
4-1
-------
However, the panelist indicated that a growing body of evidence suggests that the cohorts
predominantly exposed to shorter fibers (e.g., friction brake workers, gold miners, taconite
miners) do not have statistically significant increased cancer risks. This panelist added that the
mechanistic studies provide the strongest evidence for assigning no potency to fibrous structures
(see next bulleted item). Another panelist agreed with these statements, and added that his
interpretation of data compiled by the National Cancer Institute provide additional indirect
evidence of short fibrous structures presenting little or no carcinogenic risk (see page 102 of the
premeeting comments in Appendix B).
The panelists briefly revisited the findings from a recent publication (Suzuki and Yuen 2001) that
reported finding relatively large amounts of short, thin chrysotile fibers in malignant mesothelioma
tissue. Several panelists encouraged that these findings not be considered in the risk assessment
methodology for reasons cited earlier in the workshop (see Section 3.2.2).
Evidence from mechanistic studies. The panelists offered different interpretations of
mechanistic studies. One panelist indicated that mechanistic studies have shown that shorter
fibers are cleared more readily than long fibers from the alveolar region of the lung by
phagocytosis, and therefore provide supporting evidence that short fibers play little or no role in
carcinogenic risk. This panelist acknowledged that extremely high doses of particular matter and
other non-fibrous structures can generate biological responses (e.g., inflammation), but he
doubted that such "overload" conditions would be relevant to the environmental exposures that
the proposed protocol will be used to evaluate.
Another panelist agreed that long fibers are clearly more potent than short fibrous structures, but
he questioned the conclusion that short fibrous structures have no impact on carcinogenic risk.
This panelist noted that mechanistic studies have demonstrated that short fibrous structures and
spherical particles, like silica, can elicit the same toxic responses (e.g., generate reactive species,
stimulate proliferative factors) identified for asbestos fibers. This panelist added, referring to his
premeeting comments, that exposure to short fibers could cause inflammation and generation of
oxidative species that might increase the response to long fibers (see Bellman et al. 2001).
Overall, this panelist acknowledged that long fibers are more persistent than short fibers in the
lung and should be weighted more heavily in the exposure index, but he was hesitant to assign
the short fibrous structures zero potency.
Implications on sampling and analytical methods. One panelist commented on the
practical implications, from a sampling perspective, of any changes to the exposure index. This
panelist indicated that measuring all fibers (including structures shorter than 5 urn) in
environmental samples would not only be expensive, but also would compromise the sensitivity
of measuring the longer fibers that are most predictive of cancer risk. This panelist
acknowledged that human exposure is predominantly to fibrous structures less than 5 urn, but he
noted that the amounts of short fibrous structures retained by the lung tend to be very strongly
4-2
-------
correlated with the amounts of long fibers retained by the lung. Due to this correlation, this
panelist noted that measuring long fibers with sufficient accuracy would allow one to estimate
amounts of short fibrous structures in a sample. This panelist added, however, that he sees no
benefit of characterizing exposures to fibrous structures smaller than 5 |im, given ihe conclusion
that such fibers do not cause cancer (ERG 2003).
4.2 Responses to Charge Question 5
Charge question 5 asks: "The proposed exposure index is weighed heavily by fibers longer than 10
Specifically, Equation 7.13 suggests that the carcinogenic potency of fibers longer than 10 ^m is more
than 300 times greater than that of fibers with lengths between 5 and 10 ^im How consistent is this
difference in carcinogenic potency with the epidemiology and toxicology literature?" The panelists'
responses to this question follow:
Consistency with epidemiological literature. The panelists noted that the original
epidemiology studies did not collect exposure information that provides direct evidence of the
relative potency assigned to the two different fiber length categories: fibers longer than 10 nm,
and fibers with lengths between 5 and 10 (im During this discussion, one panelist recommended
that EPA consider the results of a case-control study (Rogers et al. 1991) that suggests that
mesolhelioma risks are greater for individuals with larger amounts of the shorter fibers (i.e.,
between 5 and 10 |im) retained in their lungs. Another panelist was not convinced of the findings
from this study, due to possible biases from selection of controls not matched for hospital of
origin. This panelist encouraged EPA to refer to more rigorous lung-retained fiber studies (e.g.,
McDonald et al. 1989, Rodelsperger et al. 1999) that have found that the maj ority of cancer
risk for mesothelioma is attributed to exposures to longer fibers, even when measurements of
short fibers are taken into account.
Questions about the fiber length-dependence used for mesothelioma. Some panelists
were not convinced that the relative potencies assigned to different fiber lengths were
appropriate for mesothelioma. One panelist, for instance, noted that his previous review of the
literature (Lippmann 1994) suggests that cancer risk for mesothelioma is most closely associated
with exposure to fibers between 5 and 10 fj.m long. He indicated that this assessment is
consistent with other human lung evaluations (e.g., Timbrell et al. 1988), which have reported
that fibers retained by the lung tend to be longer than fibers that translocate to the pleura This
panelist added that the epidemiology literature clearly suggests that lung cancer and
4-3
-------
mesotheliomahave different risk factors, as the relative amounts of lung cancer and
mesothelioma cases vary considerably from one cohort to the next. Based on these concerns,
this panelist suggested that EPA consider developing separate fiber length weighting schemes for
lung cancer and mesolhelioma
Another panelist indicated that the epidemiology studies provide indirect evidence that
carcinogenic potency appears to increase with fiber length. Specifically, he noted that the studies
consistently show that mesothelioma has a very long latency period—a trend that suggests that
the most durable fibers (i.e., the longer fibers) are the most potent. The panelist added that the
analyses in the proposed protocol provide further indirect evidence of mesothelioma risks
increasing with fiber length: when the exposure index was used in the mesolhelioma model, the
proposed risk assessment methodology generated an improved fit to the epidemiological data
During Ihis discussion, a panelist cautioned about inferring lhat only those fibers that reach the
pleura are capable of causing mesothelioma, because researchers have not determined the exact
mechanisms by which mesolhelioma is induced. Further, he cautioned about inferring too much
from a single study (Timbrell et al. 1988), given that many additional studies are available on
lung-retained fibers.
Questions about the relevance of animal toxicology data. Some panelists expressed
concern about basing the proposed weighting factors for different fiber lengths on observations
from animal data First, one panelist noted that the weighting factors were derived strictly based
on lung cancers observed in laboratory animals, and he questioned whether one can assume that
the weighting factors can be defensibly applied to mesolhelioma Second, other panelists noted
that extrapolating the weighting factors from rodents to humans also involves uncertainly, due to
inter-species differences in respiratory anatomy, macrophage sizes, and sites of lung cancers.
Suggested follow-up analyses. Given the concerns about basing the proposed exposure index
entirely on data from animal toxicology studies, two panelists recommended that EPA attempt to
optimize the weighting factors applied to different fiber length categories using the available
human epidemiological data One panelist suggested that this optimization could be performed
using the data compiled in Table 6-15 in the proposed protocol, which presents estimates of the
fiber length distribution for different occupational cohorts. A panelist also suggested that EPA
consider deriving separate weighting factors for lung cancer and mesothelioma, rather than
assuming the same fiber length dependence for both outcomes.
4.3 Responses to Charge Question 6
4-4
-------
Charge question 6 asks: "Please explain whether the proposed exposure index will allow meaningful
comparisons between current environmental exposures to asbestos and historical exposures to asbestos
that occurred in the work place." The panelists discussed several topics when addressing the question,
because some panelists had different impressions of what the question was asking. Some panelists
viewed the question as asking about the validity of low-dose linear extrapolations (see Section 3.1.5 for
more information on this topic), and others viewed the question as asking about whether the proposed
methodology is an improvement over EPA's current risk assessment model. A summary of the
panelists' specific responses follows:
Is the proposed exposure index an improvement to asbestos risk assessment? When
answering this charge question, multiple panelists focused on whether the proposed exposure
index is an improvement over EPA's 1986 asbestos risk models. These panelists agreed that the
proposed approach is more consistent with the overall literature on health risks from asbestos,
which show that cancer risks vary with fiber type and fiber dimension. Two panelists were
hesitant to call the proposed approach an improvement for evaluating mesolhelioma risks,
because the fiber length weighting factors are based entirely on lung cancer data in animals.
These panelists were particularly concerned that the proposed methodology might assign lower
risks for mesothelioma in certain circumstances, because the fiber-length dependence in Ihe
methodology is not based on any lexicological or epidemiologjcal studies of mesothelioma
Does the proposed risk assessment model support extrapolation from occupational
exposures to environmental exposures? Some panelists commented on the applicability of
the proposed risk assessment model to exposure doses below the ranges considered in the
occupational studies. Referring to observer comments provided earlier in the workshop, two
panelists indicated that some environmental exposures in areas with naturally-occurring asbestos
do not appear to be considerably lower than those experienced by occupational cohorts.
Another panelist agreed, and cautioned about distinguishing environmental exposures from
occupational exposures; he instead encouraged EPA and the panelists to focus on the exposure
magnitude, regardless of whether it was experienced in an occupational or environmental setting.
One panelist recommended that EPA investigate how cancer risks for lung cancer and
mesothelioma vary between EPA's 1986 model and the proposed risk assessment
methodology: for different distributions of fiber types and dimensions, does the proposed
methodology predict higher or lower risks than the 1986 model? Dr. Berman indicated that the
proposed methodology, when compared to EPA's 1986 model, generally predicts substantially
higher risks for environments with longer, thinner fibers and environments with larger amounts of
4-5
-------
amphibole fibers and predicts somewhat lower risks for environments with shorter, thicker fibers
and environments that contain only chrysotile fibers. One panelist recommended that future
revisions to the proposed protocol include sample calculations, perhaps in an appendix, for
several hypothetical environments to demonstrate how estimated cancer risks compare between
the new methodology and the 1986 model.
4-6
-------
5. COMMENTS ON TOPIC AREA 3: GENERAL QUESTIONS
This section summarizes the panelists' responses to charge questions 7-10 and 12. Responses to
charge question 11 are included in Section 6, because this charge question sought the panelists' overall
impressions of the proposed risk assessment methodology, rather than focusing on any one specific
issue.
5.1 Responses to Charge Question 7
This charge question asks: "The proposed risk assessment approach assigns carcinogenic potency to
individual fibers and to cleavage fragments (or 'bundles that are components of more complex
structures'). Please comment on whether cleavage fragments of asbestos are as lexicologically
significant as fibers of the same size range." The panelists raised the following points when responding:
Terminology used in the charge question. One panelist took strong exception to the
wording in this question (see pages 30-33 in Appendix B) and strongly recommended that the
panelists use correct terminology during their discussions. This panelist noted, for instance, that
cleavage fragments are not equivalent to bundles, nor do cleavage fragments meet the regulatory
definition of asbestos, as the charge question implies. He clarified that he defines cleavage
fragments as non-asbestiform amphiboles that are derived from massive amphibole structures.
This panelist was concerned that none of the panelists at the workshop has the mineralogical
expertise needed to address issues pertaining to cleavage fragments. Another panelist echoed
these concerns and agreed that this charge question raises complex issues.
Significance of cleavage fragments with respect to human health effects. The previous
concerns notwithstanding, several panelists commented on the role of cleavage fragments in the
proposed risk assessment methodology. One panelist, for example, indicated that there is no
reason to believe that cleavage fragments would behave any differently in the human lung than
asbestiform fibers of the same dimensions and durability; he added that this conclusion was also
reached by the American Thoracic Society Committee in 1990 (Weill et al. 1990). This panelist
acknowledged, however, that expert mineralogists have differing opinions on the role of
cleavage fragments. Several other panelists agreed that it is reasonable to assume that cleavage
fragments and asbestos fibers of the same dimension and durability would elicit similar toxic
responses.
5-1
-------
Review of selected epidemiological and toxicological studies. The panelists briefly
discussed what information has been published on the toxicity of cleavage fragments. One
panelist indicated that Appendix B in the proposed protocol (see pages B-3 through B-10)
interprets results from an animal study (Davis et al. 1991) that evaluated exposures to six
tremolite samples, including some that were primarily cleavage fragments. This panelist noted
that the study provides evidence that cleavage fragments can cause mesolhelioma in animals.
Another panelist, however, cautioned against inferring too much from this animal study for
several reasons: the study was not peer reviewed; the fiber measurements in the study reportedly
suffered from poor reproducibility; and the mesotheliomas observed in the study might have
reflected use of intra-peritoneal injection model as the dose administration method. This panelist
recommended that EPA conduct a more detailed review on the few studies that have examined
the toxicity of cleavage fragments, possibly considering epidemiological studies of taconite
miners from Minnesota (Higgins et al. 1983) and cummingtonite-grunerite miners from South
Dakota (McDonald et al. 1978); he noted that a pending publication presents updated risks
among the taconite miners.
Practical implications of measuring cleavage fragments in environmental samples. One
panelist added, and another agreed, that measuring cleavage fragments in environmental samples
presents some challenges, because microscopists cannot consistently distinguish cleavage
fragments from asbestiform fibers, even when using TEM.
5.2 Responses to Charge Question 8
Charge question 8 asks: "Please comment on whether the proposed cancer assessment approach is
relevant to all amphibole fibers or only to the five types of amphibole fibers (actinolite, amosite,
anthophyllite, crocidolite, tremolite) designated in federal regulations." The panelists made the following
general comments in response:
Review of evidence from toxicological and epidemiological studies. The panelists
identified few studies that address the toxicity of amphibole fibers other than actinolite, amosite,
anlhophyllite, crocidolite, and tremolite. One panelist indicated that animal toxicology studies
have demonstrated that synthetic vitreous fibers with differing chemistry, but having similar
durability and dimensions, generally exhibit similar potency for fibrosis, lung cancer, and
mesothelioma Another panelist added that lung cancer and mesothelioma exposure-response
5-2
-------
relationships for a cohort of vermiculite miners from Libby, Montana, have been published for
both asbestiform richterite and winchite.
Appropriateness of applying the model to non-asbestiform amphiboles. Several panelists
agreed that the proposed risk assessment methodology is relevant to amphibole fibers other than
those listed in the federal regulations. The panelists noted that, in the absence of more detailed
information on the matter, it is prudent to assume that fibers of similar dimension and durability
will exhibit similar toxic effects. Two panelists expressed some hesitation on applying the
proposed model to the non-asbestiform amphiboles: one panelist asked how confidently one can
apply the cancer risk coefficients to amphibole fibers that have not been studied, and another
panelist indicated he was not convinced that the model should be applied to the other
amphiboles, let alone for the amphiboles that are listed in the federal regulations.
Given the amount of naturally occurring amphiboles in the Earth's crust, one panelist suggested
that the proposed protocol clearly state that the non-asbestiform amphiboles being evaluated are
only those with the same dimensional characteristics and biodurability as the corresponding
asbestiform amphiboles.
5.3 Responses to Charge Question 9
Charge question 9 asks: "The review document recommends that asbestos samples be analyzed by
transmission electron microscopy (TEM) and count only those fibers (or bundles) longer than 5 (im
Such counting practices will provide no information on the amount of asbestos fibers shorter than 5 |im
To what extent would data on shorter fibers in samples be useful for future evaluations (e.g., validation
of the cancer risk assessment methodology, assessment of non-cancer endpoints)?"
The panelists expressed varying opinions on this matter: some panelists saw no benefit of measuring
fibrous structures shorter than 5 ^m, based on responses to earlier charge questions (see Sections
3.1.3,3.1.4, and 4.1); other panelists indicated that there is some utility to collecting information on
shorter fibrous structures, particularly if the incremental analytical costs are not prohibitively expensive
and if counting short fibers does not compromise accurate counts of longer fibers. The panelists raised
the following specific issues when discussing measurement methods:
5-3
-------
Support for using TEM in future sampling efforts. The panelists unanimously supported the
recommendation in the proposed protocol of using TEM, rather than PCM or some other
method, to characterize exposures in future risk assessments. The panelists also emphasized that
future measurement methodologies must focus on generating accurate counts of the most
biologically active fibers, or fibers longer than 5
Practical implications of counting fibers shorter than 5 \im. One panelist indicated that
analyzing samples for fibrous structures shorter than 5 |im would compromise analysts' ability to
accurately count the amounts of longer fibers that are of greater biological concern. Some
panelists and an observer further discussed the costs associated with counting fibers in multiple
length categories, including shorter than 5 jim The panelists did not cite firm cost figures for
these analyses. However, noting that environmental samples typically contain more than 90%
short fibrous structures, one panelist suspected that counting the shorter structures would
considerably increase the time a microscopist needs to analyze samples, and therefore also
would considerably increase the cost of the analysis. A panelist indicated that the costs and
benefits of counting fibers shorter than 5 ^m might be more appropriately debated between
microscopists and risk assessors, with inputs from industrial hygienists and mineralogists.
Relevance of fibers shorter than 5 \imfor non-cancer endpoints. One panelist noted that
exposures to fibrous structures shorter than 5 urn can contribute to asbestosis in occupationally
exposed individuals (Lippmann 1988), but he doubted that the exposure levels found to be
associated with asbestosis would be experienced in non-occupational settings. Another panelist
added that the role of shorter fibrous structures for other non-cancer endpoints is not known,
such as the pleura! abnormalities and active pleura! fibrosis observed in Libby, Montana No
panelists were aware of any authoritative statements made on the role that short fibers play, if
any, on these other non-cancer endpoints. During this discussion, one panelist indicated that the
toxicity of fibrous structures shorter than 5 ^m might be adequately addressed by EPA's
particulate matter standards.
5.4 Responses to Charge Question 10
Charge question 10 asks: "The proposed risk assessment methodology suggests that exposure
estimates should be based only on fibers longer than 5 nm and thinner than 0.5 nm. Is this cut-off for
fiber diameter appropriate?" Before the panelists responded to the question, Dr. Herman first clarified
that the exposure index optimized from the animal studies (see Equation 7.12 in the proposed protocol)
5-4
-------
assigns a far greater carcinogenic potency to fibers longer than 40 ^m, with diameters less than 0.4
he noted that the proposed diameter cut-off (0.5 ^m) was based on an ad hoc adjustment.
The panelists agreed that the proposed cut-off for fiber diameter (0.5 urn) would likely include most
fibers of health concern; however, they also unanimously agreed that the exposure index should not
exclude thicker fibers that are known to be respirable in humans. The main argument given for
increasing the cut-off is that fibers with diameters as large as 1.5 (im (or with aerodynamic diameters as
large as 4.5 ^m) can penetrate to small lung airways in humans. Other panelists provided additional
specific comments, generally supporting inclusion of thicker fibers in the proposed exposure index. One
panelist, for example, advised against basing the fiber diameter cut-off strictly on observations from rat
inhalation studies, due to inter-species differences in respirability. Further, noting that the proposed cut-
off for fiber diameter would likely exclude some amosite fibers and a considerable portion of tremolite
fibers with known carcinogenic potency, another panelist encouraged that the proposed exposure index
include contributions from thicker fibers.
The panelists noted that consideration of fibers thicker than 0.5 |im was viewed as being most
important for the lung cancer risk assessment model, as risks for mesothelioma appear to be more
closely linked to exposures to long, thin fibers (see Section 3.2.3). Further, some panelists suspected
that increasing the fiber diameter cut-off in the exposure index should be accompanied by changes to
the exposure-response coefficients in the risk assessment models, but the panelists did not unanimously
agree on this issue.
5-5
-------
5.5. Responses to Charge Question 12
Charge question 12 asks: "Section 8.2 of the review document presents three options for assessing
cancer risks from asbestos exposure. Please comment on the technical merit of the proposed risk
assessment options." The panelists briefly reviewed the strengths and weaknesses of the three options
presented in the proposed protocol for assessing asbestos-related cancer risks. The panelists agreed
that the first option—direct use of EPA's lung cancer and mesothelioma risk assessment
models—allows for the greatest flexibility in evaluating site-specific exposure scenarios, particularly
those with time-varying exposures. Dr. Crump indicated that he envisioned this option being coded into
a computer program, into which users enter their site-specific exposure information. Most panelists
endorsed developing such a program. The panelists did not reject use of the second and third options,
provided that EPA ensures that all three options generate equivalent risk estimates for the same
exposure scenario.
The one issue discussed in greater detail was how sensitive predictions using the first option are to the
mortality rates used in the evaluation Noting that mortality rates as functions of age and sex differ from
one location to the next, this panelist encouraged EPA to consider carefully whether nationwide
mortality estimates would be programmed into the risk assessment model or whether risk assessors
would have the option of entering site-specific mortality rates. The panelist also suggested that the
authors of the risk assessment conduct sensitivity analyses to quantify how strongly the mortality data
affect cancer risk estimates. These comments also raised questions about the fact that two populations
with different underlying mortality rates could have different cancer risks, even though their asbestos
exposure levels are equivalent
5-6
-------
6. COMMENTS ON TOPIC AREA 4: CONCLUSIONS AND RECOMMENDATIONS
This section reviews the panelists' individual conclusions and recommendations regarding the proposed
protocol (Section 6.1), as well as how the panelists developed their overall conclusions and
recommendations that appear in the executive summary of this report (Section 6.2).
6.1 Responses to Charge Question 11
Charge question 11 asks: "Discuss whether the proposed cancer assessment approach, as a whole, is a
reasonable evaluation of the available health effects data What aspects of the proposed cancer
assessment approach, if any, are inconsistent with the epidemiology or toxicology literature for
asbestos?" The panelists offered individual summary statements, which were not discussed or debated
among the panel. Following is a summary of the panelists' individual summary statements in the order
they were given:
Dr. Lippmann's summary statement Dr. Lippmann commended Drs. Herman and Crump
on developing the proposed risk assessment protocol and supported use of a model that
accounts for the factors (e.g., fiber type and dimension) that are most predictive of cancer risk.
Dr. Lippmann supported the authors' attempt to make full use of the existing data and to
interpret the results from the epidemiological studies. He strongly recommended that EPA make
every effort to obtain individual-level data from additional epidemiological studies. Dr. Lippmann
suggested that a follow-up workshop with experts in exposure assessment could help EPA
evaluate the uncertainties in exposure measurements from historic occupational data sets. Dr.
Lippmann supported an observer's suggestion to conduct an animal inhalation study using
tremolite cleavage fragments to help resolve the issue of these fragments' carcinogenic potency.
Overall, he encouraged that future work on the proposed protocol continue, through use of
additional expert panels, to make more informed usage of the human exposure data
Dr. Teta's summary statement. Dr. Teta indicated that the proposed protocol is an impressive
integration of the animal toxicology data and the human epidemiology data She commended the
authors for developing a scientific methodology that successfully reduces the variability in results
across the epidemiological studies, suggesting that the studies might be more consistent than
were previously thought. Dr. Teta recommended improvements to the meta-analysis of
6-1
-------
epidemiological studies, such as establishing and applying criteria for use of human data in
characterizing exposure-response relationships. Overall, Dr. Teta found no inconsistencies
between the proposed protocol and the larger body of epidemiology literature, including studies
of cohorts (e.g., gas mask workers, railroad workers, friction brake workers) that do not have
well-defined exposure information. Though not disagreeing with the utility of other panelists'
recommendations, such as re-analyzing data from additional epidemiological studies and
convening additional expert panels, Dr. Teta encouraged EPA to move forward expeditiously
with completing the proposed protocol and discouraged implementing additional steps lhat might
delay the overall project
Dr. Hoel's summary statement. Dr. Hoel encouraged the use of more sophisticated modeling
that incorporates data on exposure-response (including non-linear models), duration of
exposure, cessation of exposure, and uncertainty in exposure. Dr. Hoel also1 strongly
recommended that EPA attempt to obtain individual-level data from additional epidemiology
studies, or at least obtain partial data sets. He encouraged Drs. Herman and Crump to use more
sophisticated uncertainty analysis techniques, such as generating prior and posterior distributions
of uncertainly. To ensure lhat the lung cancer model is not confounded by cigarette smoking, Dr.
Hoel recommended that Drs. Berman and Crump more closely evaluate all available data on the
interactions between asbestos exposure and cigarette smoking.
Dr. Steenland's summary statement Dr. Steenland indicated that the proposed protocol is a
step forward in asbestos risk assessment; however, he had several recommendations for
improving the analysis of epidemiological studies. For instance, Dr. Steenland suggested that ihe
authors conduct meta-regression analyses using Ihe original exposure-response coefficients, in
which predictor variables include fiber size, fiber type, the estimated percentage of amphiboles,
percentage of fiber greater than 10 (im, and categorical grouping of studies according to quality.
He indicated that these factors can be examined using both fixed effects and random effects
models. Dr. Steenland recommended that the proposed protocol explicitly state and defend the
basis for choosing the 10 p.m cut-off for fiber length in the exposure index. He suggested that
EPA should consider using Bayesian techniques or other methods to determine which relative
potencies assigned to different fiber lenglh categories optimize the model's fit to the
epidemiological data
Focusing on specific topics, Dr. Steenland indicated that he disagrees with Ihe approach of
assigning amphibole fibers five times greater lung cancer potency lhan chrysotile fibers,
especially considering that Ihe statistical analysis in the proposed protocol could not reject Ihe
hypothesis that amphibole fibers and chrysotile fibers are equally potent. Further, he advocated
suggestions of exploring Ihe adequacy of other exposure-response models (e.g., non-linear
models). Finally, Dr. Steenland suspected that cigarette smoking likely will not be a confounding
factor in exposure-response analyses for two reasons. First, he noted that differences in smoking
practices between working populations and general populations typically do not cause
6-2
-------
substantial differences in standardized mortality ratios. Second, he indicated that it is highly
unlikely that prevalence of smoking varies with workers' exposure levels. Dr. Steenland
encouraged that EPA refer to a recent publication (Liddell and Armstrong 2002) for similar
insights on interactions between asbestos exposure and cigarette smoking.
Dr. Crapo's summary statement Dr. Crapo complimented Drs. Herman and Crump on
preparing the cancer risk assessment methodology, and he supported the general approach of
expressing cancer risk as a function of asbestos fiber type and fiber dimension. Dr. Crapo
indicated that the proposed protocol reaches several defensible conclusions, such as assigning
greater mesothelioma potency to amphibole fibers and to longer fibers while assigning no risk to
fibers less than 5 |im in length. However, he was concerned about some specific issues that are
not yet adequately resolved. For instance, Dr. Crapo felt additional data are needed to
rigorously define how mesothelioma potency varies with fiber length (i.e., fibers longer than 10
urn being 300 times more potent than fibers with lengths between 5 and 10 ^m). Dr. Crapo
recommended that EPA, when revising the proposed protocol, explore more sophisticated
modeling techniques, including non-linear exposure-response models and consideration
threshold effects. He supported more detailed analyses of interactions between asbestos
exposure and cigarette smoking, again through the use of non-linear models.
Dr. Sherman's summary statement Dr. Sherman first indicated that she concurred with
several recommendations made by Drs. Hoel and Steenland. She focused her summary
statements on the proposed exposure index, recommending that Drs. Herman and Crump use
the epidemiology data to further investigate other formulations of an exposure index. Dr.
Sherman recommended, for example, examining the goodness of fit of other formulations of the
exposure index (e.g., assigning zero potency to all fibers shorter than 10 |im). Further, she
recommended that the authors attempt to optimize the potency weighting factors in the exposure
index to the epidemiological data Finally, given that panelists expressed concern regarding how
potency varies with fiber length for mesothelioma, Dr. Sherman suggested that Drs. Herman and
Crump consider developing two different exposure indexes—one optimized for lung cancer, and
the other for mesothelioma Dr. Sherman added that she generally supported the lung cancer
and mesothelioma exposure-response models, and questioned whether using more complicated
models would necessarily lead to a better understanding of the data
Dr. Castranova's summary statement Dr. Castranova concluded that the proposed protocol
is a significant advance in asbestos risk assessment methodology. He strongly supported the
recommendation that future measurements be performed using TEM, rather than PCM. Dr.
Castranova also supported the approach of assigning equal carcinogenic potency to cleavage
fragments and asbestos fibers of similar dimension—a finding, he noted, that could be tested in
an animal inhalation study. Further, Dr. Castranova agreed that non-asbestiform amphiboles and
asbestos amphiboles of the same dimension should be assigned equal carcinogenic potency. Dr.
Castranova indicated that the epidemiology and toxicology literature clearly indicate that
6-3
-------
mesolhelioma potency varies with fiber type, but he was not convinced that this literature
supports a difference in lung cancer potency between amphibole and chrysotile fibers.
Dr. Price's summary statement Dr. Price found the proposed protocol to be an impressive
compilation of the epidemiology and toxicology literature into a cancer risk assessment model
that addresses most, but not all, risk factors debated since EPA's 1986 model. Dr. Price urged
EPA to explore exposure-response models other than the models that involve linear, low-dose
extrapolations, which he viewed as being inconsistent with the epidemiology literature. Dr. Price
indicated that future revisions to the protocol should definitely consider non-linear models and
threshold effects.
As an additional comment, Dr. Price emphasized that the two main elements of the
protocol—the proposed exposure index and the exposure-response analysis—are closely inter-
related and subsequent changes to the proposed exposure index could affect the robustness of
the overall modeling effort. As an example of his concern, Dr. Price noted that increasing the
fiber diameter cut-off in the exposure index from 0.5 jim to 1.5 ^m could (according to an
observer comment) lead to dramatic differences in the number of cleavage fragments counted in
environment samples; however, he indicated that the animal studies used to derive the original
exposure index did not include cleavage fragments. Such scenarios raise questions about using
an exposure index derived from very specific exposure conditions in animal studies to evaluate
human health risks associated with exposures of an entirely different character. Dr. Price
encouraged further study of cleavage fragments, perhaps in an animal inhalation study, to resolve
the role of cleavage fragments.
Dr. Case's summary statement. Dr. Case congratulated Drs. Berman and Crump for
compiling what he viewed as a reasonable evaluation of the available toxicology and
epidemiology literature, and he strongly supported the general approach of factoring fiber type
and fiber dimension into cancer risk assessment. Dr. Case indicated that he agreed with the
finding that amphibole fibers have slightly greater lung cancer potency than do chrysotile fibers,
although he believed that fiber dose, fiber length, and especially smoking history and type of
industry have greater importance in this regard. Dr. Case recognized that how one views the
differences between the Quebec and South Carolina cohorts affects the conclusions drawn on
this issue, and he encouraged EPA to classify the cohort of South Carolina textile workers as
being exposed to mixed asbestos fibers, rather than being exposed to only chrysotile fibers.2
When presenting the summary statements, one panelist (LS) indicated that NIOSH is re-analyzing filters
that were collected in the 1960s from the South Carolina textile plant, and these re-analyses should indicate the
distribution of fiber types in this cohort's exposures. Another panelist (BC) noted that these re-analyses will not
characterize earlier exposures to amosite fibers, which are believed to have occurred primarily before 1950 (based on
findings from studies of lung-retained fibers).
6-4
-------
Dr. Case made several recommendations for further evaluating the existing epidemiological data
and for collecting additional data First, Dr. Case indicated that it is critically important for any
lung cancer risk model to consider confounding effects of cigarette smoking, and he encouraged
EPA to incorporate interactions with cigarette smoking into the lung cancer model to the greatest
extent possible. Second, Dr. Case supported Dr. Lippmann's recommendation of convening an
additional expert panel workshop to critically review inferences that should be drawn from the
exposure measurements made in the epidemiological studies; such a panel, Dr. Case noted,
would require inputs from experts in mineralogy, industrial hygiene, and measurement
methodologies. Third, he supported comments recommending that EPA examine non-linear and
threshold exposure-response models. Finally, Dr. Case agreed that conducting an animal
inhalation study is probably the best way to examine whether tremolite cleavage fragments
produce lung cancer, but did not advocate using rat inhalation studies to examine whether these
fragments induce mesothelioma, because results from rat inhalation studies have been shown to
be a poor model for mesolhelioma in humans. He added, however, that it would quite probably
be impossible to design an experiment in which rats were exposed only to "cleavage fragments"
or "non-asbestiform fibers" with no asbestiform fibers present at all.
Dr. Stayner's summary statement. Dr. Stayner supported the general concept of
incorporating fiber type and fiber dimension into cancer risk assessment, but he recommended
that additional work be conducted before EPA accepts the proposed protocol as anew risk
assessment paradigm. Dr. Stayner indicated that his confidence in the proposed protocol varies
between the lung cancer and mesothelioma models.
For lung cancer, Dr. Stayner indicated that the available epidemiological data should be able to
support a new risk assessment model, but he recommended that EPA consider the panelists'
many recommendations for how the meta-analysis can be improved (e.g., using different
statistical models, developing and applying minimal study inclusion criteria, conducting additional
sensitivity analyses). Concurring with Dr. Steenland's summary statement, Dr. Stayner added
that cigarette smoking is very unlikely to be a confounding factor in Ihe lung cancer model and he
questioned whether the available data would support a quantitative assessment of the interaction
effects. While Dr. Stayner supported the recommendation for evaluating non-linear exposure-
response models, he noted that the individual-level data needed to construct these models are
not available for most epidemiological studies. Dr. Stayner added that obtaining raw data from
additional occupational cohorts would provide the best opportunity for more detailed
exploration of non-linear exposure-response relationships.
Dr. Stayner expressed greater concern about the foundation of the mesothelioma risk model. He
indicated, for instance, that the relative potencies included in the proposed exposure index are
based entirely on toxicology studies for lung cancer, and not on any epidemiology or toxicology
studies specific to mesolhelioma Despite these concerns about the biological basis for the
proposed mesothelioma model, Dr. Stayner noted that the proposed model does provide an
6-5
-------
improved fit to the findings from the epidemiological studies. He recommended that EPA
consider optimizing the relative potencies in the exposure index to the human data, especially if
EPA can access raw data from additional occupational cohorts to evaluate how exposure-
response varies with fiber size and fiber type.
• Dr. McClellan's summary statement Dr. McClellan congratulated Drs. Herman and Crump
for integrating the lexicological and epidemiological data into a reasonable evaluation of asbestos
cancer risks. Overall, Dr. McClellan found the proposed protocol to be a substantial
improvement over EPA's 1986 models and urged EPA to continue to move forward with
completing the protocol based on the panelists' feedback. Though he found the presentation of
information in the draft document to lack transparency on many important matters, Dr.
McClellan indicated that the authors' presentations at the workshop addressed many of his
concerns regarding the transparency of how the proposed model was developed. One
suggested improvement to the protocol's transparency was to clearly describe what literature
were reviewed and to specify what studies actually factored into the quantitative analyses.
Addressing specific topics, Dr. McClellan indicated that the analyses in the proposed protocol
adequately characterize the general roles that fiber type and fiber dimension play in cancer risk.
He supported suggestions for involving additional experts, perhaps in another expert panel
review, to further review interpretations of the epidemiological studies. Further, Dr. McClellan
agreed with other panelists' recommendation that EPA explore the utility of non-linear
exposure-response models, consistent with the agency's proposed revised Cancer Risk
Assessment Guidelines. If linear, low-dose extrapolation models are ultimately used, he
suggested that EPA explicitly acknowledge the uncertainties associated with such an approach.
Dr. McClellan indicated that obtaining raw data from additional epidemiological studies might be
particularly helpful in the exposure-response modeling. Finally, Dr. McClellan emphasized that
the exposure characterization in the proposed protocol is closely linked to the exposure-
response assessment; thus, the authors must carefully consider how revisions to the exposure
characterization affect the assumptions in the exposure-response assessment, and vice versa
6.2 Development of Final Conclusions and Recommendations
After presenting their individual conclusions and recommendations, the panelists worked together to
draft summary statements for the peer consultation workshop. Every panelist was asked to write a brief
synopsis of a particular topic debated during the workshop. These draft statements were then displayed
to the entire panel and observers, edited by the panelists, and then compiled into this document's
6-6
-------
executive summary, which should be viewed as the expert panel's final conclusions and
recommendations regarding the proposed protocol.
6-7
-------
7. REFERENCES
B Bellmann, H Muhle, O Creutzenberg, et al. 2001. Effects of nonfibrous particles on ceramic fiber
(RCF1) toxicity in rats. Inhalation Toxicology 13(10):877-901.
DW Berman, KS Crump, EJ Chatfield, JMG Davis, AD Jones. 1995. The Sizes, Shapes, and
Mineralogy of Asbestos Structures that Induce Lung Tumors or Mesothelioma in AF/HAN Rats
Following Inhalatioa Risk Analysis 15(2).
DW Berman and KS Crump 1999. Methodology for Conducting Risk Assessments at Asbestos
Superfund Sites; Part 1: Protocol. Final Draft. Prepared for U.S. Environmental Protection Agency.
February 15,1999.
DW Berman and KS Crump 2001. Technical Support Document for a Protocol to Assess Asbestos-
Related Risk. Final Draft. Prepared for U.S. Environmental Protection Agency and U.S. Department of
Transportation. September 4, 2001.
G Berry and ML Newhouse. 1983. Mortality of Workers Manufacturing Friction Materials Using
Asbestos. British Journal of Industrial Medicine 40:1-7.
C Boutin, P Dumortier, F Rey, et al. 1996. Black spots concentrate oncogenic asbestos fibers in the
parietal pleura American Journal of Respiratory Critical Care and Medicine 153:444-449.
M Camus, J Siematycki, B Meek. 1998. Nonoccupational exposure to chrysotile asbestos and the risk
of lung cancer. New England Journal of Medicine 338:1565-1571.
WC Cooper, 0 Wong, and R Graebner. 1988. Mortality of workers in two Minnesota taconite mining
and milling operations. Journal of Occupational Medicine 30(6):506-511.
JMG Davis, J Addison, C Mclntosh, BG Miller, and K Niven. 1991. Variations in the Carcinogenicity
of Tremolite Dust Samples of Differing Morphology. Annals of the New York Academy of Sciences,
473-490.
PE Enterline, J Harley, and V Henderson. 1986. Asbestos and Cancer—A Cohort Followed to Death.
Graduate School of Public Health, University of Pittsburgh.
N De Klerk, B Armstrong, A Musk, M Hobbs. 1989. Cancer mortality in relation to measures of
exposure to crocidolite at Wittenoom Gorge in Western Australia British Journal of Industrial Medicine
46:529-536.
7-1
-------
JM Dement, DP Brown, A Okun. 1994. Follow-up Study of Chrysotile Asbestos Textile Workers:
Cohort Mortality and Case-Control Analysis. American Journal of Industrial Medicine 26:431-447.
EPA 1986. Airborne Asbestos Health Assessment Update. U.S. Environmental Protection Agency.
EPA 600/8-84-003F. 1986.
ERG. 2003. Report on the Expert Panel on Health Effects of Asbestos and Synthetic Vitreous Fibers:
The Influence of Fiber Length. Prepared by Eastern Research Group, Inc., for 1he Agency for Toxic
Substances and Disease Registry. March 17, 2003.
TW Hesterberg, GA Hart, J Chevalier, et al. 1998. The importance of fiber biopersistence and lung
dose in determining the chronic inhalation effects of X607, RCF1, and chrysotile asbestos in rats.
Toxicology and Applied Pharmacology 153:68-82.
ITT Higgins, JH Glassman, MS Oh, and RG Cornell. 1983. Mortality of reserve mining company
employees in relation to taconite dust exposure. American Journal of Epidemiology 118(5):710-719.
J Hodgson and A Damton. 2000. The Quantitative Risk of Mesotiielioma and Lung Cancer in Relation
to Asbestos Exposure. Annals of Occupational Hygiene 44(8):565-201.
JM Hughes, H Weill, YY Hammad. 1987. Mortality of Workers Employed at Two Asbestos Cement
Plants. British Journal of Industrial Medicine 44:161-174.
IARC. 1996. AB Kane, P Boffetta, R Saracci, and JD Wilboum. Mechanisms of Fibre
Carcinogenesis. International Agency for Research on Cancer. Oxford University Press 140:1-9.
JL Levin, JW McLarty, GA Hurst, AN Smith, and AL Frank. 1998. Tyler Asbestos Workers:
Mortality Experience in a Cohort Exposed to Amosite. Occupational and Environmental Medicine
55:155-160.
FD Liddell and BG Armstrong. 2002. The combination of effects on lung cancer of cigarette smoking
and exposure in Quebec chrysotile miners and millers. Annals of Occupational Hygiene 46(1):5-13.
FD Liddell, AD McDonald, and JC McDonald. 1997. The 1891-1920 Birth Cohort of Quebec
Chrysotile Miners and Millers: Development from 1904 and Mortality to 1992. Annals of Occupational
Hygiene 41:13-36.
FD Liddell, AD McDonald, and JC McDonald. 1998. Dust exposure and lung cancer in Quebec
chrysotile miners and millers. Annals of Occupational Hygiene 42(1):7-20.
M Lippmann. 1988. Review: Asbestos exposure indices. Environmental Res 46:86-106.
7-2
-------
M Lippmann. 1994. Deposition and retention of fibers: Effects on incidence of lung cancer and
mesothelioma Occupational and Environmental Medicine 51:793-798.
JC McDonald, GW Gibbs, FD Liddell, and AD McDonald. 1978. Morality after long exposure to
cummingtonite-grunerite. American Review of Respiratory Disease 118(2):271-277.
AD McDonald, JS Fry, AJ Woolley, and JC McDonald. 1984. Dust Exposure and Mortality in an
American Chrysotile Asbestos Friction Products Plant British Journal of Industrial Medicine
41:151-157.
JC McDonald, B Armstrong, B Case, et al. 1989. Mesothelioma and asbestos fiber type. Evidence
from lung tissue analysis. Cancer 63:1544-1547.
AD McDonald, BW Case, A Churg, A Dufresne, GW Gibbs, P Sebastien, and JC McDonald. 1997.
Mesothelioma in Quebec chrysotile miners and millers: epidemiology and aetiology. Annals of
Occupational Hygiene 41(6):707-719.
JC McDonald, J Harris, and B Armstrong. 2002. Cohort mortality study of vermiculite miners exposed
to fibrous tremolite: an update. Annals of Occupational Hygiene 46(S1): 93-94.
WJ Nicholson. 1976. Part HI: Recent Approaches to the Control of Carcinogenic Exposures. Case
Study 1: Asbestos—The TLV Approach. Annals of New York Academy of Science 271:152-169.
C-G Ohlson, T Rydman, L Sundell, et al. 1984. Decreased lung function in long-term asbestos cement
workers: A cross-sectional study. American Journal of Industrial Medicine 5:359-366.
K Rodelsperger, H-J Woitowitz, B Briickel, et al. 1999. Dose-response relationship between
amphibole fiber lung burden and mesothelioma Cancer Detection and Prevention 23(3): 183-193.
AJ Rogers, J Leigh, G Berry, et al. 1991. Relationship between lung asbestos fiber type and
concentration and relative risk of mesothelioma Cancer 67:1912-1920.
A Rogers and G Major. 2002. Letter to the Editor: The Quantitative Risks of Mesothelioma and Lung
Cancer in Relation to Asbestos Exposure: The Wittenoom Data Annals of Occupational Hygiene
46(1):127-129.
H Seidman, IJ Selikoff, and SK Gelb. 1986. Mortality Experience of Amosite Asbestos Factory
Workers: Dose-Response Relationships 5 to 40 Years after onset of Short-Term Work Exposure.
American Journal of Industrial Medicine 10:479-514.
7-3
-------
IJ Selikoffand H Seidman. 1991. Asbestos-Associated Deaths among Insulation Workers in the
United States and Canada, 1967-1987. Annals of the New York Academy of Sciences 643:1-14.
LT Stayner, DA Dankovic, RA Lemen. 1996. Occupational Exposures to Chrysotile Asbestos and
Cancer Risk: A Review of the Amphibole Hypothesis. American Journal of Public Health
86(2): 176-186.
V Timbrell, T Ashcroft, B Goldstein, et al. 1988. Relationships between retained amphibole fibers and
fibrosis in human lung tissue specimens. In: Inhaled Particles VI. Annals of Occupational Hygiene
32(S1):323-340.
Y Suzuki and S Yuen. 2001. Asbestos Tissue Burden Study on Human Malignant Mesothelioma
Industrial Health 39:150-160.
MJ Teta, HC Lewinsohn, JW Meigs, et al. 1983. Mesothelioma in Connecticut: 1955-1977:
Occupational and geographic associations. Journal of Occupational Medicine 25(10):749-756.
A Tossavainen, M Kotilainen, K Takahashi, G Pan, and E Vanhala 2001. Amphibole Fibers in
Chinese Chrysotile Asbestos. Annals of Occupational Hygiene 45(2):145-152.
H Wall, JL Abraham, JR Balmes, B Case, AM Churg, J Hughes, M Schenker, and P Sebastien.
1990. Health Effects of Tremolite. Official statement of the American Thoracic Society. American
Review of Respiratory Disease 142(6): 1453-1458.
E Yano, ZM Wang, XR Wang, MZ Wang, and YJ Lan. 2001. Cancer Mortality among Workers
Exposed to Amphibole-free Chrysotile Asbestos. American Journal of Epidemiology 154(6): 53 8-543.
7-4
-------
The following Appendices:
Appendix A - List of Expert Panelists
Appendix B - Consultants' Premeeting Comments
Appendix C - List of Registered Observers
Appendix D - Agenda
Appendix E - Observer Comments Provided at the Peer Consultation Workshop
Appendix F - Observer Post-Meeting Comments
Are available at: http://www.epa.gov/superfund/programs/risk/asbestos/index.htm
-------
Appendix A
List of Expert Panelists
-------
Appendix B
Premeeting Comments, Alphabetized by Author
(includes bios of panelists and the charge to the panelists)
Note: This appendix is a copy of the booklet of the premeeting comments that ERG distributed at Ihe
peer consultation workshop. One panelist (Dr. Bruce Case) submitted an edited form of his
premeeting comments to ERG at the workshop. That edited version appears in this appendix.
-------
Appendix C
List of Registered Observers of the Peer Consultation Workshop
-------
Appendix D
Agenda for the Peer Consultation Workshop
-------
Appendix E
Observer Comments Provided at the Peer Consultation Workshop
Note: The peer consultation workshop included three observer comment periods, one on the first day
of the workshop and two on Ihe second day of the workshop. This appendix includes verbatim
transcripts (to the extent that specific remarks were audible from recordings) of the observer
comments, in the order the comments were given.
-------
Appendix F
Observer Post-Meeting Comments
-------
APPENDIX C:
COMPENDIUM OF MODEL FITS TO ANIMAL
INHALATION DATA IN SUPPORT OF THE HERMAN
ET AL. (1995) STUDY AND POST-STUDY WORK
The attached tables are a compendium of raw outputs for the fits of various (exposure index)
models to the Davis et al. animal inhalation studies. Each entry lists the date of the run, the size
categories included in the run, the maximum likelihood estimate for the run, the degrees of
freedom, the P-value for the fit, and the coefficients representing the relative potency assigned to
each size category for the model.
C.1
-------
"05/29/1992"
.7472 5.
"05/29/1992"
.0000 3.
"05/29/1992"
5.1206E-03 6
"05/29/1992"
.2085
"05/29/1992"
2.0226E-11 4
"05/29/1992"
.0000
"05/29/1992"
.0000
"05/29/1992"
.7619 4.
"05/29/1992"
.0000
"05/29/1992"
.0000
"05/29/1992"
.0000
"05/29/1992"
.0000 3.
"05/29/1992"
.8224
"05/29/1992"
1.2200E-02
"05/29/1992"
.0000 9.
"05/29/1992"
.1732
"05/29/1992"
.0000
"05/29/1992"
.0000
"05/29/1992"
1.0128E-11 4
"05/29/1992"
3.1396E-03
"17:52:38"
2207E-03 6
"17:52:42"
7528E-02
"17:53:01"
.2222E-04
"17:53:06"
0000 4
"17:53:16"
.2164E-03
"17:53:23"
0000
"17:53:42"
0000
"17:54:06"
7177E-03 7
"17:54:11"
0000 2
"17:55:00"
0000 1
"17:55:34"
0000
"17:55:52"
6828E-03 3
"17:56:15"
0000
"17:56:20"
.2542
"17:56:25"
7985E-03
"17:56:50"
1111
"17:56:54"
0000 4
"17:57:02"
0000
"17:57:16"
.0590E-03
"17:57:20"
.0000
Equation?
"PS PCM <5, 5-10, 10-20, >20
.7064E-05
"SC PCM 20-30, >30, <0.1, 0.1-0.2, 0.2-0.3
.0000 .0000 .0000 .5381
"SC PCM 20-30, 30-40 >40
"SC PCM 20-30, >30, AR<1060
1.8379E-04
"SC PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50,
.0000 .0000 .7552 .0000
, 0.3-0.4, >0.4
3.7282E-11 4.
5-60, >60, <0.
.0000
3022E-03 5
2, >0.2
2448
.1364E-03
.0000
"SC PCM 10-20, 20-30, 30-40, >40, AR<100, 100-200, AR>200
.0000 .2265 8.0142E-02 .0000
"SC PCM <5, 5-10, 10-20, >20
.9308E-05
"PS PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50,
.7635E-02 4.6469E-04 .5748 .0000
.4841
50-60, >60, <0
.3849 1.
0000 -6
.2, >0.2
2251E-02
"PS(no C or M)PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50, 50-60, >60, <0.
.7508E-02 4.3916E-02 .6920 .2466
"SC(no C or M) PCM <5, 5-10, 10-2, 20-30, 30-40
.0000 7.8183E-03 .6538 .0000
"SC PCM 20-30, 30-40, 40-50, <0.15, 0.15-0.25
.8012E-05 .0000 .0000 1.4520E-02
"PS PCM 20-30, 30-40, 40-50, <0.15, 0.15-0.25
.0000 .0000 .0000 .1776
"PS(no C or M) PCM 20-30, 30-40, 40-50, <0.15,
.1467 .0000 .0000 .0000
"SC(no C or M) PCM 20-30, 30-40, 40-50, <0.15,
.0000 .0000 .0000 1.5661E-03
"PS PCM 20-30, 30-40, 40-50, <0.2, 0.2-0.3,
.0000 .0000 6.4495E-02 .5343
"SC PCM 20-30, 30-40, 40-50, <0.2, 0.2-0.3,
.9612E-02 .1239 .0000 1.2642E-10
"SC PCM <5, 5-10, 10-20, >20, <0.25, >0.25
.0000 2.5766E-02 2.3627E-10 4.0282E-03
"SC PCME(SC) <5, 5-10, 10-20, >20
3.4073E-05
"FBC PCM <5, 5-10, 10-20, >20, <0.25, >0.25
4.8616E-02 .0000 5.2996E-03 4.2436E-03
.0000
, 40-50, 50-60
.0000
, 0.25-0.35,
.0000
, 0.25-0.35,
.0000
0.15-0.25, 0
.1467
0.15-0.25, 0
.0000
0.3-0.4
7.5079E-03 9.
0.3-0.4
5.0996E-03 1.
1.0973E-03
1.6013E-03
0000
, >60, <0.2
3383
>0.35
3824 -1
>0.35
0000 4
.25-0.35,
.1467
.25-0.35,
4401 -3
4402E-03
1890E-02
.0482E-08
.0000
2, >0.2
.0000
, >0.2
.0000
.7330E-08
.3492E-10
>0.35
3.5740E-11
>0.35
.8840E-08
MLE
" -62.1949
" -60.5687
" -60.7224
" -60.9782
" -61.1790
11 -60.1211
.0000
" -59.8032
4.1242E-03
" -61.7831
" -59.6135
.0000
" -59.5425
.0000
" -60.5424
.0000
" -59.1475
4.4091E-03
" -59.3960
4.2216E-03
" -59.8255
3.4723E-03
" -58.9800
4.1576E-03
" -59.6295
" -59.0333
" -61.4230
" -60.4908
" -61.5539
ChiSquare DF
18.14 11
13.56 8
13.89
14.77
15.12
•13.15
.0000
12.80
S.7554E-03
17.00
12.25
.0000
13.53
.0000
14.20
.0000
11.74
1.9861E-02
12.11
2.6240E-03
13.54
4.9456E-03
11.67
2.1421E-02
12.32
11.22
15.59
14.03
15.72
10
10
11
10
.0000
8
11
7
.0000
8
.0000
9
.0000
6
10
5
7
7
8
10
11
9
P-Value Coefficients?
7.8315E-02 .0000
9.4038E-02 .0000
.1782 7.
.1407
.1772
.2155
1.3025E-10
.1189
.1080
9.2727E-02
-1.5614E-08
9.4748E-02
1.9732E-10
.1155
-2.8277E-10
5.7916E-02
.2776
1.8790E-02
.1120 9.
9.0528E-02
.1897 9.
.1119
.2312
7.3068E-02
6677E-02 3.
0000
0000 1
0000
2.9064E-03
0000
0000
0000
5.6502E-03
0000
4.4931E-03
0000
3.8208E-03
1315
0000
1467
8451E-02
0000 5.
2764E-02
0000
0000
0000
0000
2042
2929E-13
7915
.000
0000
2.4831E-03
2092
0000
0000
9.6600E-03
0000
1.5296E-03
0000
1.8162E-03
0000
0000
0000
0000
2419E-02
0000
0000
0000
0000
C.2
-------
"05/29/1992" "17:57:37" "SC PCM 5-10, 10-20, 20-30, >30
4.2799E-13 4.2822E-03 2.1632E-04
"05/29/1992" "17:57:42" "SC PCM, AR>10, 5-10, 10-20, 20-30, >30
4.2799E-13 4.2822E-03 2.1632E-04
"05/29/1992" "17:57:46" "SC PCM, AR>10, 5-10, 10-20, >20, <0.25, > 0.25
.0000 2.5766E-02 1.9480E-11 4.0282E-03 1.0973E-03
"05/29/1992" "17:57:54" "SC PCM, 10-20, >20, <0.1, 0.1-0.2, 0.2-0.5, >0.5
.0000 3.9377E-02 .3581 .0000 -2.6301E-10 3.7544E-03 1.9807E-03
-61.6197 16.18 10 9.4624E-02 .0000 1.7645E-02
-61.6197 16.18 10 9.4624E-02 .0000 1.7645E-02
-61.4230 15.59 10 .1119 .0000 .0000
-60.8243 15.06 9 8.9374E-02 .6025 .0000
C.3
-------
"06/04/1992"
.0000
"06/04/1992"
.0000
"06/04/1992"
1.0642E-02
"06/04/1992"
.0000
"06/04/1992"
1.8243E-03
"06/04/1992"
9.7585E-03
"06/04/1992"
1.1391E-02
"06/04/1992"
1.0616E-02
"06/04/1992"
1.9830E-02
"06/04/1992"
9.7585E-03
"06/08/1992"
9.7772E-07
"06/08/1992"
3.5523E-07
"06/08/1992"
.0000
"06/08/1992"
.0000
"06/08/1992"
.0000
"06/08/1992"
7.9477E-02
"06/08/1992"
6.5890E-02
"06/08/1992"
8.9823E-02
"06/08/1992"
.0000
"06/08/1992"
"15:59:
0000
"16:00:
0000
"16:00:
.0000
"16:09:
0000
"16:09:
.0000
"16:10:
.0000
"16:25:
.0000
"16:25:
.0000
"16:25:
.0000
"17:01:
.0000
"14:23:
"14:23:
"15:17:
3697
"15:17:
8647
"15:17:
9434
"15:17:
.0000
"15:18;
.0000
"15:18:
.0000
"15:18:
0000
"15:18:
53
06
34
39
53
22
02
16
29
04
24
27
:31
38
43
:50
Equation?
" "PS M <5, 5-10, 10-20, 20-30, 30-40, 40-50, 50-60, >60
.3350 .0000 .6647 3.3093E-02 2.7314E-02
" "PS M <5, 5-10, 10-20, 20-30, 30-40, 40-50, 50-60, >60 (without)
.3131 .0000 .6868 2.3751E-02 3.0330E-02
" "PS PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50, 50-60, >60
.1925 .0000 .3962 2.6431E-02 1.3497E-02
" "SC M <5, 5-10, 10-20, 20-30, 30-40, 40-50, 50-60, >60
.4181 .0000 .5818 3.3111E-02 3.1739E-02
" "SC M <5, 5-10, 10-20, 20-30, 30-40, 40-50, 50-60, >60 (without)
.3599 .0000 .6382 2.3345E-02 3.4731E-02
" "SC PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50, 50-60, >60
.1659 .0000 .1266 2.5546E-02 1.7443E-02
" "FBC M <5, 5-10, 10-20, 20-30, 30-40, 40-50, 50-60, >60
.0000 .4578 .5308 3.2501E-02 7.0673E-02
" "FBC M <5, 5-10, 10-20, 20-30, 30-40, 40-50, 50-60, >60 (without)
.0000 .4798 .5095 2.3348E-02 7.4414E-02
" "FBC PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50, 50-60, >60
.0000 .9802 2.1638E-13 2.4967E-02 2.1750E-02
" "SC PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50, 50-60, >60
.1659
" "SC PCM,
" "SC PCM,
" "SC PCM
.1637
" "SC PCM
1.5222E-11
" "SC PCM
6.3283E-14
" "SC PCM
.0000
USING SUM OF
USING SUM OF
10-20, 20-30,
2.7318E-02
10-20, 20-30,
2.4334E-02
10-20, 20-30,
2.5224E-02
10-20, 20-30,
-7.6669E-10 2.4802E-02
:00
:16
33
48
" "SC PCM
.0000
" "SC PCM
.9063
" "SC PCM
.0000
" "SC PCM
10-20, 20-30,
4.0257E-02
10-20, 20-30,
.0000
10-20, 20-30,
.0000
10-20, 20-30,
.6978 2.5546E-02
AR INSTEAD OF
THE SUM OF
(L~2/W) INSTEAD OF
30-40, 40-50,
8.6314E-02
30-40, 40-50,
2.8640E-02
30-40, 40-50,
2.2209E-02
30-40, 40-50,
2.8345E-02
30-40, 40-50,
.0000
30-40, 40-50,
.0000
30-40, 40-50,
.8731
30-40, 40-50,
50-60
50-60
50-60
50-60
50-60
.0000
50-60
.0000
50-60
,0000
50-60
1.7443E-02
THE NUMBER OF STRUCTURES
THE SUM OF THE NUMBER OF STRUCTURES
, >60,
, >60,
, >60,
, >60,
, >60,
, >60,
, >60,
, >60,
AR>100
AR>50
AR>30
AR>10
<0.1, >0.1
.0000 .0000 .3413
<0.2, >0.2
.0000 .0000 3.9421E-11
<0.3, >0.3
.0000 1.6883E-02 .1024
<0.5, >0.5
MLE
" -325.
" -289.
" -279.
" -324.
" -288.
" -279.
" -327.
" -292.
" -280.
" -279.
" -320.
" -304.
" -282.
" -282.
" -284.
" -282.
" -277.
410
768
861
088
190
958
491
240
941
958
998
439
.372
195
342
527
.397
2.6604E-02
" -274.
.192
2.4470E-02
" -273.
570
2.5219E-02
" -277.
042
ChiSquare
72.32
40.91
24.45
69.36
37.50
24.27
76.80
48.03
27.21
24.27
128.1
92.38
31.86
29.24
33.22
28.52
21.06
3.2997E-02
14.21
9.5869E-02
13.47
7.8163E-02
18.60
DF P-Value Coefficients?
12 .0000 2.5437E-04 .0000
10 .0000 1.2826E-04 .0000
9 2.8979E-03 .0000 .0000
12 .0000 1.2222E-04 .0000
9 .0000 2.5206E-05 .0000
9 3.1396E-03 .0000 .0000
12 .0000 .0000 .0000
10 .0000 .0000 .0000
10 1.5744E-03 .0000 .0000
9 3.1396E-03 .0000 .0000
11 .0000 .5002 .1054
11 .0000 .5002 4.9910E-02
9 .0000 .3233 .1434
9 .0000 2.9482E-02 .1058
9 .0000 2.5570E-02 3.1076E-02
9 2.6697E-05 1.0006E-02 .0000
8 5.1426E-03 .5520 5.5765E-04
9 .1142 .0000 3.8278E-03
9 .1417 7.5561E-03 .0000
7 8.7967E-03 7.1726E-03 7.7336E-03
C.4
-------
1.8013E-02
"06/08/1992"
.0000
"06/08/1992"
.4112
"06/09/1992"
.0000
"06/09/1992"
.0000
"06/09/1992"
4.9270E-02
"06/09/1992"
.1396
"06/09/1992"
8.9222E-02 2
"06/09/1992"
9.9701E-02
"06/09/1992"
.2282 2.
"06/09/1992"
.2383
"06/09/1992"
.0000
"06/09/1992"
.0000
"06/09/1992"
.0000
"06/09/1992"
1.3378E-04
"06/09/1992"
.0000
"06/09/1992"
.0000
"06/10/1992"
.0000
"06/10/1992"
.0000
"06/10/1992"
.0000
"06/10/1992"
.0000
"15:19:
0000
"15:19:
0000
"09:36:
0000
"09:36:
9754
"09:36:
.0000
"09:37:
0000
"09:51:
15"
23"
4
19"
44"
5
50"
02"
9
30"
.5453E-04
"09:51:
.0000
"09:52:
9401E-03
"09:52:
0000
"10:49:
0000
"10:49:
0000
"10:50:
0000
"10:42:
.0000
"10:43:
0000
"10:43:
0000
"09:42:
0000
"10:46:
0000
"10:47:
0000
"10:47:
53"
08"
36"
30"
54"
29"
08"
04"
44"
5
52"
34"
3
09"
9
42"
3.6580E-02 .0000
"SC PCM 10-20, 20-30,
.1691 .1142
"SC PCM 10-20, 20-30,
.5280E-02 .1237
"SC PCM 10-20, 20-30,
.0000 .0000
"SC PCM 10-20, 20-30,
.0787E-11 2.4409E-02
"SC PCM 10-20, 20-30,
4.6592E-02 2.4754E-02
"SC PCM 10-20, 20-30,
.8428E-02 2.5738E-02
"PS PCM 10-20, 20-30,
.9088 .0000
"FBC PCM 10-20, 20-30
.8806 .0000
"PS PCM 10-20, 20-30,
.3567 .0000
"FBC PCM 10-20, 20-30
.0000 .2039
"SC PCM <5, >5, <0.1,
.0000 .0000
"SC PCM <10, >10, <0.
.0000 .0000
"SC PCM <20, >20, <0.
.0000 .0000
"SC PCM <30, >30, <0.
.0000 .0000
"SC PCM <40, >40, <0.
.0000 .0000
"SC PCM <50, >50, <0.
.0791E-04 .0000
.7005
30-40, 40-50,
.0000
30-40, 40-50,
.2896
30-40, 40-50,
.9036
30-40, 40-50,
2.0780E-02
30-40, 40-50,
2.5388E-02
30-40, 40-50,
1.9572E-02
30-40, 40-50,
.0000
, 30-40, 40-50
1.5773E-02
30-40, 40-50,
.0000
, 30-40, 40-50
.5428
0.1-0.2, 0.
.0000
1, 0.1-0.2,
.0000
1, 0.1-0.2,
.0000
1, 0.1-0.2,
.0000
1, 0.1-0.2,
2.8445E-04
1, 0.1-0.2,
.0000
.0000
50-60,
.0000
50-60,
50-60,
.0000
50-60,
50-60,
50-60,
50-60,
.0000
, 50-60
.0000
50-60,
.0000
, 50-60
.0000
2-0.3,
.0000
0.2-0.3
.7228
0.2-0.3
.3188
0.2-0.3
.0000
0.2-0.3
.0000
0.2-0.3
.0000
"SC PCMIdifferent order)<40, >40, <0.1, 0.
.0000 .0000
2.8445E-04
.0000
>60,
>60,
>60,
>60,
>60,
>60,
>60,
, >60
>60,
, >60
.0000
<1, >1
.0000
AR>200
<0.4, >0.4
.0000 3.
AR>20
AR>5
AR>3
<0.2, >0.2
.0000
, <0.2, >0.2
.0000
<0.3, >0.3
.0000
, <0.3, >0.3
.0000
0.3-0.5, 0.5-1,
, 0.
, 0.
, 0.
, 0.
, 0.
1-0.2
.0000
3-0.5, 0.5-1
.0000
3-0.5, 0.5-1
.5973
3-0.5, 0.5-1
.4087
3-0.5, 0.5-1
.0000
3-0.5, 0.5-1
.0000
, 0.2-0.3, 0.
.0000
.0000
0000
9774E-02
.0000
.0000
0000
0000
1-2, >2
0000
, 1-2, >2
0000
, 1-2, >2
0000
, 1-2, >2
.4290
, 1-2, >2
9300
, 1-2, >2
9601
3-0.5, 0.
9300
.2300
.7058
4.6900E-02
-2.0902E-08
-7.8124E-08
.4087
-2.1203E-06
1.0368E-02
.0000
4.5484E-02
5.4099E-02
.0000
.0000
5-1, 1-2, >2
.0000
"SC M 10-20, 20-30, 30-40, >40, AR<100, 100200
.6258E-03 .0000
.0000
.0000
.7186 2.
4978E-03
.1761
"SC M 10-20, 20-30, 30-40, >40, AR-C100, 100200 (without)
.5295E-03 .0000
"SC PCM 30-40, 40-50,
.0000
<0.2, 0.2-0.
.0000
3
.7246 4.
7535E-03
.1814
2.3656E-02
" -276.804
2.5210E-02
" -314.551
" -273.752
2.4767E-02
" -282.130
" -280.428
" -280.621
" -273.982
2.4482E-02
" -276.214
2.5377E-02
" -276.577
2.5828E-02
" -285.851
1.7593E-02
" -306.621
.2379
" -288.155
8.3293E-02
" -277.141
.0000
" -278.516
.0000
" -274.113
.0000
" -285.544
.0000
" -274.113
.0000
" -301.303
2.6786E-02
" -271.722
2.3813E-02.
" -328.264
3.0589E-0:
18.60
1.7712E-02
102.3
13.77
5.8149E-02
28.46
24.35
25.34
14.15
.2017
18.66
.1086
19.01
2.8763E-02
39.03
2.3597E-02
96.83
.3233
44.97
.1248
18.94
3.8390E-02
21.96
.1080
13.89
S.9566E-02
42.63
3.9410E-02
13.89
S.9566E-02
21.78
.6618
9.472
.6408
132.1
9 2.8109E-02 1.0869E-02 .0000
9 .0000 .0000 3.2936E-02
8 8.7091E-02 7.5744E-03
10 S.8359E-04 2.4648E-02
9 3.0278E-03 1.0293E-02
9 1.8654E-03 9.1834E-03
8 7.7092E-02
8 1.5954E-02
8 1.3996E-02
8 .0000
.0000
.0000
.0000
.0000
.0000
9 .0000
5.4390E-02 9.2242E-04
9 .0000 .0000
2.6367E-02 9.8343E-03
9 2.4998E-02 .0000
2.2548E-02 4.7097E-02
8 4.1566E-03 .0000
2.6150E-02 5.0063E-02
9 .1256 1.4143E-04
2.6644E-02 9.1119E-02
10 .0000 .0000
3.3774E-02 .1998
.1256
1.4143E-04
10 1.5441E-02 .0000
8 .3034 .0000
10 .0000 .4994
2.1524E-03
.0000
.0000
.0000
1.7037E-03
3.9259E-03
3.5134E-03
1.4957E-02
.0000
.0000
.0000
.0000
.0000
.0000
.0000
9.9225E-02
7.9714E-02
.0000
C.5
-------
.3992
"06/10/1992
.4357
"06/10/1992
.1084
"06/10/1992
.1009
"06/10/1992
.1226
"06/10/1992
1.7299E-04
"06/10/1992
4.5248E-02
"06/10/1992
5.9914E-02
"06/10/1992
6.6660E-02
"06/10/1992
.1224 2
"06/10/1992
.1842 2
"06/10/1992
-1.5012E-12
"06/10/1992
1.7035E-02
"06/10/1992
.2507 2
"06/10/1992
-7.5830E-12
"06/10/1992
.0000
.1441 .1207
" "10:47:44" "SC PCM 30-40, >40, <0.2, 0.2-0.3
.1441 .1141
"10:47:46" "SC PCM 30-40 AND <0.2, 40-50 AND 0.2-0.3
"10:47:47" "SC PCM 30-40 AND <0.2, >40 AND 0.2-0.3
"10:47:49" "SC PCM <40, >40, 0.2-0.3
"10:47:55" "SC PCM <40, >40, <0.1, 0.1-0.2, 0.2-0.3, 0.3-0.4, 0.4-0.6, 0.6-1, >1
.0000 .0000 .0000 1.6964E-04 .0000 .0000 .7574 .2283
" "15:23:45" "SC PCM <40, >40, <0.3, >2
2.5944E-02 9.8491E-02
" "15:24:05" "SC PCM <40, >40, <0.4, >2
2.7518E-02 4.8515E-02
" "15:24:25" "SC PCM <40, >40, <0.3, >3
2.5536E-02 9.0266E-02
" "15:24:36" "SC PCM <40, >40, <0.3, >5
.6666E-02 8.0706E-02
" "15:24:48" "SC PCM <40, >40, <0.3, >10
.9171E-02 8.2945E-02
" "15:25:11" "SC PCM 20-40, >40, <0.3, >2
2.5325E-02 8.2712E-02
"15:25:17" "SC PCM 20-40, >40, <0.4, >2
2.4296E-02 3.6304E-02
" "15:25:24" "SC PCM 20-35, >35, <0.3, >2
.6694E-02 2.8991E-02
" "15:25:27" "SC PCM 20-45, >45, <0.3, >2
2.4133E-02 .2121
" "15:25:34" "SC PCM <20, 20-40, >40, <0.3, >2
.0000 4.5312E-02 2.5936E-02 9.8583E-02
"06/11/1992" "15:31:51" "SC PCM <40, >40
5.9240E-03
"06/11/1992" "15:31:56" "SC PCM <20 20-40, >40, <0.3, >5
1.0718E-02 2.5619E-02 .1163 2.5487E-02 7.6466E-02
"06/11/1992" "15:32:29" "SC PCM <10 10-40, >40, <0.3, >2
2.4070E-03 .0000 4.5564E-02 2.5144E-02 9.3422E-02
"06/11/1992" "15:32:48" "SC PCM <10 10-40, >40, <0.3, >5
5.1824E-03 5.4184E-03 .1346 2.4875E-02 7.0329E-02
"06/11/1992" "15:33:31" "SC PCM <20 20-40, >40, with AR>100 or w>5
2.4093E-02 1.6167E-02
328.
328.
328.
277.
•274.
0000
274.
275.
•274.
274.
274.
•283.
•276.
282.
•286.
•274.
289.
•273.
•274.
•273.
279.
362
264
362
294
426
857
922
995
006
370
427
234
898
727
850
830
833
688
349
057
132
132
132
19.
.3
.1
.3
14
14.08
1.3866E-02
14.75
17.
14.
13.
14.
31.
17.
30.
36.
14.
50.
13.
14.
12.
23.
20
95
41
68
70
96
25
13
73
85
04
64
40
32
10
11
11
11
.0000
.0000
.0000
5.7976E-02
7 4.9035E-02
2.6012E-02 .1065
9 9.7473E-02
9
9
9
9
9
9
10
9
9
11
8
9
8
10
4.4919E-02
9.1591E-02
.1442
9.9294E-02
.0000
3.4871E-02
.0000
.0000
9.7945E-02
.0000
.1098
.1005
.1337
8.7947E-03
.4486
.4443
.4927
.9997
5.2545E-05
5.4764E-05
1.0871E-04
5.9029E-05
7.1369E-05
7.5858E-05
3.7466E-02
9.9336E-02
.1302
2.3230E-02
5.4613E-05
.9995
7.4307E-05
4.4249E-05
4.5873E-05
.1284
.0000
.1441
.1441
2.7463E-02
.0000
2.5762E-04
1.0373E-03
3.1457E-03
4.9037E-03
1.2275E-02
9.1336E-02
7.5388E-02
.0000
3.3283E-02
2.7873E-04
4.0751E-02
.0000
.0000
.0000
.4497
C.6
-------
"06/11/1992"
5.4238E-02 2
"06/11/1992"
.0000
"06/11/1992"
.0000
"06/11/1992"
.0000
"06/11/1992"
.0000
"06/11/1992"
1.4206E-02
"06/11/1992"
4.3934E-02
"06/12/1992"
1.0976E-07
"06/12/1992"
2.5583E-08
"06/12/1992"
.0000 3.
"06/12/1992"
7.5561E-03
"06/12/1992"
7.5744E-03 2
"06/12/1992"
.0000
"06/12/1992"
.0000
"06/12/1992"
3.9080E-02 3
"06/12/1992"
5.5882E-04 3
"06/12/1992"
1.7176E-03
"06/12/1992"
2.6148E-02 2
"06/12/1992"
7.5831E-03 5
"06/12/1992"
.2053
"06/12/1992"
"15:33:37
.0059E-02
"16:05:03
0000
"16:06:22
0000
"16:06:57
0000
"16:08:07
0000
"16:08:44
.0000
"16:09:50
.0000
"12:45:29
"12:46:51
"14:25:51
7448E-03
"14:26:39
.0000
"14:27:33
.1524E-03
"14:28:40
0000
"14:30:01
0000
"14:31:34
.7648E-02
"14:32:09
.2760E-04
"14:33:06
.0000
"14:33:22
.3396E-02
"14:33:37
.6428E-02
"14:33:58
0000
"14:34:10
" "SC PCM <20 20-40
" "SC M <10, 10-20,
.0000 .0000
, >40, with AR>200 or w>5
20-30, 30-40, 40-50, 50-60, >60, <0.4, >0
2.0091E-02
" "SC M <20, >20, <0.2, 0.2-0.3, 0
.0000 .0000
.0000
" "SC M <30, >30, <0.2, 0.2-0.3, 0
.0000 .0000
.0000
" "SC M <40, >40, <0.2, 0.2-0.3, 0
.0000 .0000
" "SC M <10, 10-20,
.0000 .0000
" "SC M <10, 10-20,
.0000 .0000
.0000
.0000
.3-0.4, 0.4-2
.0000
.3-0.4, 0.4-2
.8490
.3-0.4, 0.4-2
.0000
3376
, 2-5, 5-8,
0000
, 2-5, 5-8,
0000 9
, 2-5, 5-8,
5169
20-30, 30-40, 40-50, 50-60, >60, <0.2, >0
.9699
1.3334E-02
.0000
20-30, 30-40, 40-50, 50-60, >60, <0.3, >0
.4960
9.9726E-02 5
.7563E-02
.4
.0000
>8 (without)
.4395 1.
>8 (without)
.9544E-02
>8 (without)
.4323
.2
.0000
.3
.0000
6323
5005E-02
0000
0000
.0000
.0000
" "SC PCM, USING SUM OF (AR)"1.8 INSTEAD OF THE SUM OF THE NUMBER OF STRUCTURES
" "SC PCM MIMICS
RJ LEE
" "SC PCM <10, 10-20, 20-30, 30-40,
8.7979E-02 .0000
.8868
" "SC PCM <10, 10-20, 20-30, 30-40,
.0000 .0000
.0000
" "SC PCM <10, 10-20, 20-30, 30-40,
.0000 .0000
" "SC PCM <20, >20
2.8541E-03 .0000
" "SC PCM <30, >30
.0000
and <0.2, 0.2-0.
.0000
and <0.2, 0.2-0.
3.1038E-03 8.3599E-02 .0000
" "SC PCM <40, >40
.0000 .0000
" "SC PCM <50, >50
.0000 .0000
" "SC PCM <5, 5-40,
and <0.2, 0.2-0.
.0000
and <0.2, 0.2-0.
.0000
>40 and <0.3, >5
40-50, 50-60,
.0000 2.
40-50, 50-60,
.0000
40-50, 50-60,
.0000
3, 0.3-0.4,
.3170 8.
3, 0.3-0.4,
.0000 1.
3, 0.3-0.4,
.0000
3, 0.3-0.4,
.0000
>60 and ,0.
1443E-02
>60 and ,0.
.8731
>60 and ,0.
.9036
0.4-2, 2-5,
8019E-03
0.4-2, 2-5,
5371E-02
0.4-2, 2-5,
.0000
0.4-2, 2-5,
.0000
2, >0.2
.0000
3, >0.3
.0000
4, >0.4
.0000
5-8, >8
.2178
5-8, >8
.5256
5-8, >8
.0000
5-8, >8
.7592
0000
.0000
.0000
0000
0000
.0000
.0000
" -287.783
" -279.728
.0000
" -290.051
.0000
" -278.334
1.6664E-04
" -277.968
.0000
" -280.135
.0000
" -298.689
.0000
" -316.759
" -314.904
" -274.192
.0000
" -273.570
1.6883E-02
" -273.752
3.9774E-02
" -273.104
.0000
" -274.535
.0000
" -318.182
.0000
" -282.887
.0000
" -272.935
57.24
23.25
9.9335E-03
43.31
.5454
19.29
9.9650E-04
19.74
5.0666E-02
24.37
2.5917E-03
63.37
.3028
120.5
103.8
14.22
4.0984E-06
13.47
.1024
13.77
4.6900E-02
12.30
.4536
15.28
.3723
153.4
.9233
38.07
.2399
12.19
10 .0000 4.7695E-02
8 2.2192E-03 3.6878E-05
2.1964E-02 .3872
9 .0000 1.0731E-04
2.3980E-02 3.3505E-02
7 5.5724E-03 5.7529E-05
2.1948E-02 .3498
9 1.8859E-02 8.5360E-05
2.2914E-02 .2941
8 1.1420E-03 5.8051E-06
2.1348E-02 .3422
8 .0000 .0000
4.6309E-02 3.6312E-02
11 .0000 .5002
11 .0000 .4998
8 7.5497E-02 .0000
2.4473E-02 9.7935E-02
9 .1417 .0000
2.5219E-02 7.8163E-02
8 8.7091E-02 .0000
2.4767E-02 5.8149E-02
8 .1378 .0000
2.4370E-02 2.3120E-02
8 5.3177E-02 .0000
2.5364E-02 3.0722E-02
10 .0000 .0000
3.8196E-02 1.0287E-03
9 .0000 .0000
3.7498E-02 6.3278E-02
10 .2716 .0000
.6415
.0000
.0000
.0000
.0000
.0000
.0000
8 .6106E-02
.1020
.0000'
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.1453 2.5612E-02 7.0424E-02
" "SC PCM <5, 5-40,
>40 and <0.4, >5
" -301.510
78.09
9 .0000 .0000
.0000
S.1184E-02 6.3939E-02 5.3121E-03
" "SC PCM <10, 10-40, >40 and <0.4,
>5
" -273.662
13.27
8 .1023 2.6086E-05
.0000
3.3751E-14 2.4070E-02 4.5074E-02
" "SC PCM <20, 20-40, >40 and <0.4,
>5
" -337.543
281.4
10 .0000 2.7954E-02
.0000
.7667 3.7371E-02 3.2598E-04
" "SC PCM <20, 20-40, >40 and <0.3,
>8
" -345.304
353.5
9 .0000 .2255
.0000
C.7
-------
5.5410E-02
"06/12/1992"
6.0514E-02 1
"06/12/1992"
.1835 2.
"06/12/1992"
6.8066E-02
"06/12/1992"
.2309
"06/12/1992"
.3874 4.
"06/12/1992"
.2309
"06/12/1992"
.2258
"06/12/1992"
2.2757E-02 2
"06/12/1992"
7.5406E-02 2
"06/15/1992"
3.1957E-02
"06/15/1992"
4.7120E-02
"06/15/1992"
9.6472E-04
"06/15/1992"
.0000
"06/15/1992"
.0000
"06/15/1992"
.0000
"06/15/1992"
3.3972E-04
"06/15/1992"
3.2317E-04
"06/15/1992"
8.7093E-02
"06/15/1992"
5.6553E-02
"06/15/1992"
.0000
.0000
"14:34:49
.3755E-02
"14:36:04
0935E-02
"14:36:54
.0000
"14:36:59
0000
"14:37:04
7852E-02
"14:
0000
"14:
0000
"14:
37:11
37:15
37:21
.5509E-02
"16:
47:32
.7083E-02
"11:
.0000
"11:
.0000
"11:
.0000
"11:
0000
"11:
0000
"11:
0000
"11:
.0000
"11:
.0000
"12:
.0000
"12:
.0000
"12:
0000
06:50
08:19
08:49
10:43
12:36
14:25
16:10
18:03
16:10
16:21
16:44
5.1022E-02 3
" "SC PCM <20,
.1895 2
" "SC PCM <20,
.4459 2.
" "SC PCM <10,
.9319 2
" "SC PCM <10,
.3555 2.
" "SC PCM <10,
.2899 5.
" "SC AR>=100
.3555 2.
" "FBC AR>=100
.3623 2.
" "SC PCM <40,
9.4588E-02
" "SC PCM <20,
.7421 2
" "SC PCM <5,
.0000
" "SC PCM <5,
2.9985E-02
" "SC PCM <5,
4.3486E-03
" "SC PCM <5,
1.3893E-03
" "SC PCM <5,
3.7626E-03
" "SC PCM <5,
7.6795E-03
" "SC PCM <3,
4.1296E-03
" "SC PCM <5,
2.2255E-03
" "SC PCM <5,
8.7093E-02
" "SC PCM <5,
.2414
" "SC PCM <5,
1.3801E-02
.8161E-02 4
20-50, >50
.4667E-02 5
20-50, >50
9928E-02 2.
10-40, >40,
.4370E-02 1
10-40, >40,
5908E-02 3.
10-40, >40,
2688E-02 4.
OR (C
and CS
5908E-02 3.
OR SC(C and
5958E-02 3.
>40,
and <0
20-40, >40
.5097E-02 o
5-10,
.0000
5-10,
.0000
5-10,
.0000
5-10,
0000
5-10,
0000
5-10,
0000
3-10,
.0000
5-8,
.0000
5-10,
.0000
5-10,
.0000
5-10,
0000
10-20,
4
10-20,
10-20,
10-20,
5.
10-20,
3.
10-20,
3.
10-20,
.0910E-05
and <0.3, >5
.5813E-02
and <0.4, >5
6012E-05
with AR> = 50
.4888E-02
with AR>=100
4688E-02
with AR>=200
5503E-02
only) WIDTH>
4688E-02
or WIDTH>=5
or WIDTH>=5
or WIDTH> = 5
=5, PCM
CS only) WIDTH>=5,
4841E-02
.3, >0.3
and >8, <0.3
.9122E-02
20-30, 30-40
.3341E-02
20-30, 30-40
.0000
20-30, 30-40
.0000
20-30, 30-40
, 40-50
0000
, 40-50
0000
, 40-50
0000
, 40-50
2414E-02 .0000
20-30, 30-40
, 40-50
8885E-02 .0000
20-30, 30-40
, 40-50
4904E-03 .0000
20-30, 30-40
.0000
8-20, 20-30, 30-40,
10-20,
10-20,
10-20,
3.
.0000
20-30, 30-40
.1419
20-30, 30-40
.0000
20-30, 30-40
, 40-50
0000
, <10
(6 categories)
(6 categories)
(6 categories)
, 10-40, >40
PCM, <10, 10-40
, >50
, >50
, >50
, >50
and <0.2,
6315
and <0.3,
0000
and <0.4,
0000
and <0.5,
S.0335E-02
, >50
and <0.6,
.1015
, >50
and <0. 8,
5.0160E-02
, >50
and <0.3,
0000
, >40
>8
1.6485E-03
>8
.0000
>8
.1754
>8
.3280
>8
.4142
>8
.3190
>8
4.2215E-02
40-50, >50 and <0.3, >8
0000
, 40-50
0000
, 40-50
0000
, 40-50
7693E-02 .0000
, >50
, >50
, >50
0000
6.5672E-02
with AR>=200 or W>=l
0000
with AR>=
0000
with AR>=
5.3310E-02
.2037
100 or W>
.2577
50 or W> =
.3375
=1
8
" -277.024
" -349.979
" -274.917
" -275.017
" -286.452
" -275.017
" -275.121
" -276.137
" -273.463
18.84
404.9
15.55
15.94
56.19
15.94
16.15
17.09
12.58
"
.0000
"
.0000
"
.7476
"
.0000
"
.0000 7
"
.0000
"
.4837
"
.3431
(14 cat. ) "
8.7093E-02
(14 cat. ) "
.0000
(14 cat.) "
.0000
-296.708
.0000
-301.032
.0000
-272.385
.0000
-272.858
.3594
-273.029
. 0023E-02
-273.276
.2732
-272.209
.0000
-272.224
.0000
-285.792
.0000
-272.200
.2205
-273.057
.3226
70.84
.2916
82.76
.9229
11.16
7.1640E-02
11.87
.1984
12.11
.3717
12.53
.3365
10.93
.1544
11.04
8.4681E-02
53.75
.1092
11.02
2.2922E-02
12.26
.2351
8 1.4921E-02 6.7283E-05
9 .0000 .3496
11 .1582
10 .1006
9 .0000
10 .1006
.0000
.4137
.0000
.4137
10 9.4586E-02 .4119
.0000
.0000
.0000
.0000
.0000
.0000
.0000
9 4.6680E-02 1.6030E-05 1.8285E-04
8 .1262 .0000 8.1316E-05
8 .0000 .0000 .0000
2.8411E-02 1.6238E-02
10 .0000 .0000 .0000
2.5358E-02 5.3733E-03
8 .1922 .0000 .0000
2.5317E-02 6.4120E-02
7 .1043 .0000 .0000
2.4085E-02 4.6887E-02
7 9.6208E-02 .0000 .0000
2.4532E-02 3.5003E-02
7 8.3861E-02 .0000 .0000
2.4465E-02 3.7891E-02
7 .1412 .0000 .0000
2.5667E-02 9.9115E-02
6 8.6318E-02 .0000 S.2016E-07
2.5656E-02 .1429
4 .0000 8.7093E-02 8.7093E-02
5.5273E-02 .1341
5 5.0335E-02 4.9881E-02 4.9881E-02
2.5260E-02 9.4228E-02
6 5.5674E-02 .0000 1.3179E-07
2.3665E-02 4.4856E-02
C.8
-------
"06/15/1992" "12:18:06
.0000 .0000
"06/15/1992" "12:19:28
.0000 .0000
"06/15/1992" "12:20:57
.0000 .0000
"06/15/1992" "12:22:28
.0000 .0000
"06/15/1992" "12:24:05
.0000 .0000
"06/15/1992" "15:17:15
.0000 6.4848E-04
"06/16/1992" "15:57:13
.0000 4.7435E-03
"06/16/1992" "15:57:21
.0000 2.0072E-03
"06/16/1992" "15:57:28
1.068SE-02 8.2511E-03
"06/16/1992" "16:20:15
1.1243E-02 5.4704E-02
"06/16/1992" "16:20:42
7.3058E-02 .1151
"06/16/1992" "15:58:04
.1135 4.1740E-02
"06/16/1992" "15:58:14
9.8803E-02 3.6726E-02
"06/16/1992" "16:37:21
.1482 .1710
"06/16/1992" "15:58:39
.2372 .1225
"06/16/1992" "15:58:46
9.3549E-02 2.4707E-02
"06/16/1992" "15:58:54
.0000 4.5497E-04
"06/16/1992" "15:59:39
.0000 7.6856E-03
"06/17/1992" "14:24:32
.0000 6.4848E-04
"06/17/1992" "14:25:51
" "SC PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50,
4.9098E-03 .0000 1.9981E-02 .0000
" "SC PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50,
6.9703E-03 .0000 .0000 .0000
" "SC PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50,
8.1095E-03 .0000 .0000 .0000
" "SC PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50,
5.0759E-03 .0000 4.7770E-03 .0000
" "SC PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50,
5.8288E-03 .0000 7.2078E-04 .0000
" "SC PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50,
.0000 3.2479E-03 .0000 .0000
" "SC PCM <5, 5-30, >30 and >8, <0.3
.5686 2.5768E-02 2.5232E-02
" "SC PCM <5, 5-40, >40 and >8, <0.3
.7917 2.7267E-02 5.5788E-02
" "SC PCM <8, 8-40, >40 and >8, <0.3
.7160 2.5565E-02 5.4117E-02
" "SC PCM <10, 10-40, >40 and <0.3, >8
.2171 2.4137E-02 5.6716E-02
" "SC PCM <15, 15-40, >40 and <0.3, >8
.3195 2.3806E-02 3.7739E-02
" "SC PCM <10, 10-30, >30 and >8, <0.3
.2284 2.3632E-02 1.8589E-02
" "SC PCM <10, 10-50, >50 and >8, <0.3
1.3111E-11 2.4535E-02 2.2019E-02
" "SC PCM > 50, 20-50, <20 and>8, <0.3
1.8609E-04 2.3804E-02 2.4382E-02
>50 with AR>
5.4082E-03
>50 with AR>
3.0064E-02
>50 with AR>
2.5890E-02
>50 with AR>
.0000
>50 with AR>
1.0573E-02
>50 and >8,
4. 9158E-02
=30 or W>=8
.2060
•=20 or W>=8
.2439
=10 or W>=8
.2019
=5 or W>=8
.1090
=3 or W>=8
.1299
<0.3
.0000
(14 cat.)
.0000
(14 cat.)
.0000
(14 cat.)
.0000
(14 cat.)
.0000
(14 cat.)
.0000
.0000
" "SC PCM <10, 10-40, >40 and AR>=100 or W>=8 (6 categories)
.3545 2.4533E-02 3.6979E-02
" "SC PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50,
.2394 .0000 .0000 2.3420E-02
" "SC PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50,
.0000 2.2876E-03 .0000 .0000
" "SC PCM <10, 10-40, >40 and >8, <0.3 (with AR>
.7128 2.6190E-02 5.5721E-02
" "SC PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50,
.0000 3.2479E-03 .0000 .0000
" "SC PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50,
>50 and AR>=150 or W>=8
8.5612E-02
>50 and >8,
.0000
=3)
>50 and >8,
4.9158E-02
>50 and >8,
7.6892E-02
<0.3 (with
.0000
(14 cat.)
.1257
AR>=3)
.0000
<0.3 and AF!> = 10
.0000
.0000
<0.3 and AR>=3
" -273.
.6920
" -273.
.5410
" -273.
.4075
" -274.
.7382
" -274.
.6530
" -272.
.4088
" -275.
" -272.
" -272.
" -272.
" -272.
" -274.
" -274.
" -275.
" -274.
" -272.
.632
7.
.360
.866
167
231
.198
.834
.871
.805
.895
886
890
.641
.523
.361
.330
8.5612E-02
" -272.
.2874
" -272.
" -272.
.4088
" -272.
,197
.968
.198
.198
13.
,16
.0740E-02
12.
.1781
13.
.3565
14.
.1430
14.
.2000
10.
.4276
18.
12,
12,
12,
12.
15.
15.
16,
15,
11,
.1054
10.
.5971
11,
10,
.4276
10.
,73
.72
.40
.54
.89
.83
.51
.41
.21
.13
.82
.42
.03
,16
.22
.83
.90
,89
.89
7 5.7616E-02
2.3745E-02 6.6264E-02
8 .1208
2.4490E-02 5.1541E-02
8 8.8611E-02
2.5024E-02 3.8497E-02
8 7.1042E-02
2.5059E-02 5.8887E-02
7 4.1630E-02
2.5215E-02 5.1238E-02
7 .1430
2.5808E-02 .1227
10 4.1618E-02
10 .2515
9 .1906
8 .1414 1.
8 .1446 2.
9 7.0146E-02
9 7.9344E-02
8 4.1166E-02
9 8.5993E-02
1 5.8425E-05 5.
2.7771E-02 .1274
8 .2109
2.6318E-02 .1745
9 .2183
7 .1430
2.5808E-02 .1227
7 .1430
0000
0000
0000
0000
0000
0000
0000
0000
0000
5538E-05
1089E-05
0000
0000
6671
2858
6473E-02
0000
0000
0000
0000
.0000
.0000
.0000
.0000
.0000
.0000
..0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
5.6473E-02
.0000
3.9534E-05
.0000
.0000
C.9
-------
"06/17/1992"
2
0000
"06/17/1992
0000
"06/17/1992
0000
"06/17/1992
0000
"06/17/1992
.0354E-02
"06/17/1992
0000 3
"06/17/1992
1034 1
"
"
"
"
"
"14:
0000
"14:
0000
"14:
0000
"14:
0000
"14:
.2327
"14:
28:
32:
33:
33:
33:
34:
12
50
09
41
51
07
"
"
"
7
"
"
"
.9432E-02 8
"
"14:
34:
13
It
.4351E-02
"06/17/1992"
3
.7724E-04
"06/17/1992
0000 1
"06/17/1992
0000 1
"06/17/1992
0000
"
"14:
.0000
"15:
45:
45:
23
33
"
"
.5961E-03
"
"15:
45:
33
"
.5961E-03
"
"06/17/1992"
0000
"06/17/1992
1094 3
"06/17/1992
1094 2
"
"15:
0000
"15:
0000
"15:
45:
45:
47:
50
50
24
"
7
"
7
"
.2465E-02 8
"
"15:
47:
24
.0771E-02
"06/22/1992"
9
.7750E-04
"06/22/1992"
3
6
8
.3816E-03
"06/22/1992
0000
"06/22/1992
.0850E-04
"06/22/1992
.4954E-04
"06/22/1992
0000
"
"
"
"
"06/22/1992"
"15:
.0000
"15:
.0000
"15:
0000
"15:
.0000
"15:
.0000
"15:
0000
"15:
33:
35:
36:
37:
40:
43:
46:
41
19
20
36
55
58
08
"
6
"SC PCM <5,
.2027
"SC PCM <5,
.0000
"SC PCM <40
.7411E-03
"SC PCM <40
.0000
"SC PCM <10
.5564
"SC PCM <10
.5786E-02 2
"SC PCM <10
.6578 2
"SC PCM <5,
S.8825E-04
"SC PCM <5,
.0000 4
"SC PCM <5,
.0000 4
"SC PCM <40
.8049E-04
"SC PCM <40
.8049E-04
"SC PCM <10
.7587E-02
"SC PCM <10
.6119E-02
5-10, 10-20
.0000
5-10, 10-20
.0000 9
, >40 and <0
.0000
, >40 and <0
.0000 3
, 10-40, >40
3.6817E-02
, 10-40, >40
.3356E-02 8
, 10-40, >40
.4091E-02 3
5-10, 10-20
.0000
5-10, 10-20
.1331E-02
5-10, 10-20
.1331E-02
, >40 and <0
.0000 4
, >40 and <0
.0000 4
, 10-40, >40
, 10-40, >40
, 20-30, 30-40, 40-50, >50 and
.0000 .0000 .3859
, 20-30, 30-40, 40-50, >50 and
.3260E-02 .0000 .0000
.2, 0.2-0.3, 0.3-0.4, 0.4-2
.4238 .0000 .0000
.2, 0.2-0.3, 0.3-0.4, 0.4-2
.1406E-02 .0000 .8748
and <0.3, >8 (chrysotile)
4.5930E-05
and<0.3, >8 (amphiboles)
.8092E-02
and >8, <0.4
.4458E-02
, 20-30, 30-40, 40-50, >50 and
7.1379E-03 .0000 .0000
, 20-30, 30-40, 40-50, >50 and
.0000 .0000 .0000
, 20-30, 30-40, 40-50, >50 and
.0000 .0000 .0000
.2, 0.2-0.3, 0.3-0.4, 0.4-2
.1073E-02 .0000 .8914
.2, 0.2-0.3, 0.3-0.4, 0.4-2
.1073E-02 .0000 .8914
and <0.3, >8 (2 studies)
and <0.3, >8 (2 studies)
>8
>8
, <0.3 (chrysotile)
.2964 .0000
, <0.3 (amphiboles)
5.5979E-02 .0000
, 2-5, 5-8, >8 (chrysotile)
, 2
<0
>8
>8
, 2
, 2
" "SC M(14) <5, 5-10, 10-20, 20-30, 30-40, 40-50, >50 and
.0000
.0000
.0000 .0000 .0000
" "PS M(14) <5, 5-10, 10-20, 20-30, 30-40, 40-50, >50 and
"
"
"
"
"
.0000
"FBC M(14)
.0000
"SC PCM <5,
3.1381E-03
"SC PCM <5,
2.3397E-03
"SC PCM <5,
.0000
"SC PCM <40
.0000
1.6148E-02 .0000 .0000
<5, 5-10, 10-20, 20-30, 30-40, 40-50, >50
.8194 2
5-10, 10-20
.0000
5-10, 10-20
.0000
5-10, 10-20
.1688
, >40 and <0
.2740E-02 1.290
, 20-30, 30-40, 40-50, >50 and
.0000 .0000 .0000
, 20-30, 30-40, 40-50, >50 and
.0000 .0000 .0000
, 20-30, 30-40, 40-50, >50 and
.0000 .0000 .2492
.2, 0.2-0.3, 0.3-0.4, 0.4-2
and
<0
<0
<0
.0000 5.5665E-02
-5, 5-8, >8 (amphiboles)
.0000 .0000
.3, >8 (with AR>=10)
.0000 .2738
, <0.3 (2 studies)
.0000 -2.1416E-05
, <0.3 (2 studies)
.0000 -2.1416E-05
-5, 5-8, >8 (2 studies)
.0000 4.1637E-02
-5, 5-8, >8 (2 studies)
.0000 4.1637E-02
<0.3, >8
.0000 .2541
<0.3, >8
.0000 .0000
<0.3, >8
.3 and AK>=20, >8
5.6291E-02 .3825
.3 and AR>=30, >8
7.1845E-02 .3003
.3, >8 (chrysotile)
.5106 .0000
, 2-5, 5-8, >8 (chrysotile)
-199.146
.0000
-100.336
.7515
-199.147
.0000
-100.379
.0000
-231.202
-100.344
-273.352
8.429
1.1885E-11
1.656
7.0580E-11
8.432
.5028
1.697
9.3811E-02
165.7
1.655
13.00
2 1.3952E-02 .0000
3.2157E-02 7.5085E-02
1 .1976 .0000
2.3542E-02 .1025
3 3.7153E-02 .0000
3.2162E-02 2.7220E-02
3 .6374 .0000
2.3351E-02 8.8094E-02
3 .0000 .1906
.6468
.1619
" -275.830 18.52 7 9.0747E-03
.0000 -1.0060E-10 2.5949E-02 .3142
" -301.025 12.62
3.1086E-02 1.8208E-02
" -301.025 12.62
2.1682E-02 2.9493E-02
" -300.792 12.06
3.2830E-02 .1416
" -300.792 12.06
2.0805E-02 8.7989E-02
11 -300.139 11.32
-300.139
11.32
.0000
.0000
.0000
8 .1249 .0000
8 .1249 .0000
6 5.9787E-02 2.6205E-05
6 5.9787E-02 2.6205E-05
8 .1837 .0000
8 .1837 .0000
-277.733 20.96
.7450 1.9854E-10
-277.324 19.37
.9455 3.4942E-02
-282.927 31.48
9 1.2108E-02
2.1741E-02 .4278
9 2.1452E-02
2.2094E-02 .3327
9 .0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
2.0191E-06 2.8155E-04
-272.197
.0000
-272.256
.0000
-174.591
.0000
-174.591
10.92
9.9690E-02
11.12
5.9915E-02
3.392
7.1357E-02
3.395
5 5.2268E-02 1.7829E-05 1.3304E-09
2.5851E-02 .1301
5 4.8301E-02 8.7621E-05 1.3116E-06
2.5516E-02 .1679
2 .1827 .0000 .0000
3.0137E-02 9.0414E-02
2 .1825 .0000 .0000
C.10
-------
.0000 .0000 5.5468E-03 .0000 .3469 .0000 .0000 .0000 .0000 .0000 .3974 3.0140E-02 3.8470E-02
"06/22/1992" "15:55:24" "SC PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50, >50 and<0.3, >8 (chrysotile) (avg)" -126.207 1.2755E-15 0 .0000 .0000
3.9108E-03 .0000 1.7452E-02 .0000 1.0756E-02 .0000 .0000 .0000 .0000 .0000 .9679 2.6668E-02 2.1269E-02
"06/22/1992" "16:30:35" "SC PCM <40, >40 and<0.2, 0.2-0.3, 0.3-0.4, 0.4-2, 2-5, 5-8, >8 (chrysotile) (avg)" -126.207 2.9177E-13 -1 .0000 .0000
.0000 .0000 1.4347E-02 .0000 .0000 .0000 .0000 .0000 .8395 .1459 -1.9596E-10 2.6668E-02 1.3955E-02
"06/22/1992" "15:49:00" "PS PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50, >50 and<0.3, >8 " -272.711 12.25 8 .1395 .0000
8.6521E-03 .1246 .0000 .0000 .0000 .0000 .1227 .4444 .0000 .0000 .2996 2.5616E-02 3.4242E-02
"06/22/1992" "15:50:08" "FBC PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50, >50 and<0.3, >8 " -275.871 18.62 8 1.6231E-02 .0000
7.0798E-04 7.2592E-03 .0000 .7180 2.5916E-02 .3137
.0000
2.8851E-04
.0000
3.7774E-04
-127.858
-248.223
"06/23/1992" "14:32:26" "SC PCM <20, >20 (chrysotile only - averaged K013)
2.0218E-03
"06/23/1992" "15:20:51" "SC PCM <5,5-10,10-20,20-30,30-40,40-50,>50 with AR>=100 or W>=8 (14 CAT.)(no discharged)"
5.3902E-02 .0000 .2203 .0000 .0000 .0000 .0000 .2507 .0000 .0000
"06/23/1992" "15:21:06" "SC PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50, >50 with <0.3, >8 (no discharged) " -247.916
7.7664E-04 .0000 3.3297E-03 .0000 1.1683E-02 .0000 .0000 .1023 .3292 .0000
"06/23/1992" "15:21:59" "SC PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50, >50 with <0.4, >8 (no discharged) " -247.977
9.2037E-04 .0000 5.3676E-03 .0000 2.3489E-02 .0000 .0000 .2528 .5300 .0000
"06/23/1992" "15:22:39" "SC PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50, >50 with <0.5, >8 (no discharged) " -248.237
.0000 .0000 4.4608E-03 .0000 7.6930E-02 .0000 3.6633E-02 .4443 .0000 .0000
"06/23/1992" "15:43:52" "SC PCM <5,5-10,10-20,20-30,30-40,40-50,>50 with <0.5 and AR>=3,>8(no discharged)" -248.237
.0000 .0000 4.4608E-03 .0000 7.6930E-02 .0000 3.6633E-02 .4443 .0000 .0000
"06/23/1992" "15:50:53" "SC PCM <5, 5-10,10-20,20-30,30-40,40-50,>50 with >8,<0.5and AR>=10(no discharged)
.0000 .0000 .0000 .0000 .0000 .1014 .4356 6.4002E-03 .0000
"06/23/1992" "15:45:06" "SC PCM <5,5-10,10-20,20-30,30-40,40-50,>50 with <0.5and AR>=20,>8(no discharged)
.0000 .0000 .0000 .0000 .1019 .0000 .0000 .4509 .0000
"06/23/1992" "16:04:30" "SC PCM <20, >20 with <0.3, >8 (no discharged)
.6923 2.1630E-02 1.6030E-02
"06/23/1992" "16:04:37" "SC PCM <40, >40 with <0.3, >8 (no discharged)
.1952 2.8847E-02 7.4959E-02
"06/23/1992" "16:04:49" "SC PCM <40, >40 with <0.3, >5 (no discharged)
.1310 2.4517E-02 7.8915E-02
"06/23/1992" "16:39:42" "SC PCM >10 (all with AR >= 3) (no WDC chrysotile or tremolite)
4.4491E-03
"06/23/1992" "16:39:44" "SC PCM <10, 10-20, >20 (all with AR >= 3) (no WDC chrysotile or tremolite)
2.5926E-02 3.8301E-04
"06/23/1992" "16:39:46" "PS PCM >10 (all with AR >= 3) (no WDC chrysotile or tremolite)
4.6048E-03
"06/23/1992" "16:40:17" "PS PCM <10, 10-20, >20 (all with AR >= 3) (no WDC chrysotile or tremolite)
2.6163E-02 4.7643E-04
"06/23/1992" "16:39:49" "SC(no C,CS,M,or MS) PCM >10 (all with AR >= 3) (no WDC chrysotile or tremolite)
1.1936E-03
3.449
.3270
1.000
-247.715 6.231 5 .2839
6.1001E-02 2.4264E-02 .1097
6.347 5 .2733 .0000
.1214 2.3640E-02 .1100
6.302 6 .3897 .0000
.1875 2.3356E-02 5.1364E-02
6.697 7 .4607 .0000
4377 2.2722E-02 3.3184E-02
6.697 7 .4607 .0000
4377 2.2722E-02 3.3184E-02
6.690 6 .3499 .0000
0000 -3.7956E-06 2.2698E-02 3.2980E-02
-248.139 6.630 8 .5767 4.1766E-04 .0000
0000 .4468 2.2774E-02 3.3466E-02
-250.437
2.9297E-02
1653 .1653
.0000
.0000
.0000
.0000
1.0298E-04
-251.384
-250.175
-220.082
-219.772
-220.532
-220.468
-220.925
9.979
12.81
9.585
16.49
16.07
17.26
17.36
18.92
.3517
.1178
.2948
9 5.6634E-02
9 S.4832E-02
9 4.4107E-02
9 4.2634E-02
9 2.5190E-02
3.1512E-04 .0000
9.4338E-05 3.9873E-03
7.0228E-05 1.5561E-02
.9283
.0000
.9300
.0000
.6933
2.5391E-02
3.9618E-12
2.5829E-02
.1894
2.6422E-02
C.11
-------
"06/23/1992" "16:39:51" "SC(no C,CS,M,or MS) PCM <10,10-20,>20 (all with AR >= 3) (no WDC or tremolite) " -220.428 17.89 9 3.5777E-02
2.6733E-02 4.2818E-04
"06/24/1992" "14:59:32" "SC PCM >10 (all with AR>=3 and W>=0.2) (no tremolite or WDC chrysotile)
1.0459E-02
"06/24/1992" "14:59:34" "SC PCM <10,10-20,>20(all with AR>=3 and W>=0.2)(no tremolite or WDC chrysotile) '
2.5513E-02 4.4249E-04
"06/24/1992" "14:59:36" "FBC PCM >10 (all with AR>=3 and W>=0.2) (no tremolite or WDC chrysotile)
1.0404E-02
"06/24/1992" "14:59:38" "FBC PCM <10,10-20,>20(all with AR>=3 and W>=0.2)(no tremolite or WDC chrysotile)'
2.6426E-02 4.8211E-04
"06/24/1992" "14:59:41" "SC PCM <5, 5-40, >40, wlthW<0.5, >5
4.6480E-03 .0000 .8946 2.9880E-02 1.5430E-02
"06/30/1992" "10:00:07" "SC PCM 5-40, >40, with W >5, <0.3
.8530 2.5612E-02 7.0424E-02
"06/24/1992" "14:59:58" "SC PCM 5-40, >40, with W <0.5, >5
.8946 2.9880E-02 1.5430E-02
"06/24/1992" "15:00:08" "SC PCM 5-40, >40, with W <0.3, >5 (and all structures AK>=3)
.1517 2.5388E-02 7.8347E-02
"06/24/1992" "15:00:17" "SC PCM 5-40, >40, with W <0.3, >5 (and all structures AR>=-5)
.1902 2.5505E-02 8.6887E-02
"06/30/1992" "10:00:12" "SC PCM 5-40, >40, with W >5, <0.3 and AR>=3
.8530 2.5612E-02 7.0424E-02
"06/30/1992" "10:00:17" "SC PCM 5-40, >40, with W >5, <0.3 and AR>=5
.8530 2.5612E-02 7.0424E-02
"06/24/1992" "15:00:35" "SC PCM 5-40, >40, with W <0.5 and AR>=5, >5
.8946 2.9880E-02 1.5430E-02
"06/24/1992" "15:00:46" "SC PCM 5-40, >40, with W <0.5 and AR>=10, >5
.8946 2.9880E-02 1.5430E-02
"06/24/1992" "15:00:56" "SC PCM 5-40, >40, with W <0.5 and AR>=20, >5
.9347 2.8885E-02 1.5186E-02
"06/24/1992" "15:01:00" "SC PCM 5-40, >40, with W <0.5 and AR>=30, >5
.9274 2.7747E-02 1.5152E-02
"06/24/1992" "15:01:03" "SC PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50, >50 (no discharged chrysotile)
2.3031E-02 9.2428E-02 4.8158E-02 .8364 2.4594E-02 7.7840E-03
"06/24/1992" "15:01:13" "SC PCM 5-40, >40, with W <0.3, >5 (and all structures AR>=3) (no discharged)
.2102 2.4254E-02 7.1338E-02
"06/24/1992" "15:01:18" "SC PCM ^-40, >40, with W <0.5 and AR>=20, >5 (no discharged chrysotile)
.9926 2.6968E-02 1.6049E-02*
"06/25/1992" "10:28:00" "SC M(16) 5-40, >40 and <0.3, >5 " -320.420 65.65 12 .0000
5.5987E-02 2.9951E-02 .2335
.0000
5.4710E-11
•219.
•219.
•220.
•220.
•279.
•272.
•279.
•273.
•274.
•272.
•272.
•279.
•279.
•278.
•277,
•255,
-248.
•253.
.932
,402
855
072
.469
.935
.469
,227
,576
,935
,935
,469
,469
.500
,782
,492
.899
,154
16
15
18
16
27
12
27
12
15
12
12
27
27
24
22
20
8.
18
.02
.14
.70
.90
.19
.19
.19
.61
.49
.19
.19
.19
.19
.89
.81
.01
027
.13
9
9
9
8
10
10
10
10
10
10
10
10
10
10
10
8
9
10
S.5704E-02
8.6504E-02
2.7090E-02
3.0312E-02
1.5875E-03
.2716
1.5875E-03
.2459
.1144
.2716
.2716
1.5875E-03
1.5875E-03
4.7281E-03
1.0646E-02
9.4712E-03
.5311
5.2069E-02
.9655
.0000
.9594
1.9053E-03
.0000
.0000
4.6480E-03
1.4845E-03
1.4548E-03
.0000
.0000
4.6480E-03
4.6480E-03
6.8172E-03
1.2458E-02
.0000
1.7527E-03
7.4254E-03
2.5115E-02
2.4662E-11
2.6243E-02
1.9158E-12
.0000
1.7176E-03
.0000
.0000
.0000
1.7176E-03
1.7176E-03
.0000
.0000
.0000
.0000
.0000
.0000
.0000
1.2664E-03 .0000
C.12
-------
"06/25/1992" "10:28:07" "SC M(14) 5-40, >40 and <0.3, >5
7.1229E-02 2.3175E-02 .2276
"06/25/1992" "10:28:13" "SC M(16) 5-40, >40 and <0.3, >5 (no discharged chrysotile)
.1014 2.6934E-02 .2021
"06/25/1992" "10:28:17" "SC M(14) 5-40, >40 and <0.3, >5 (no discharged chrysotile)
.1164 2.1490E-02 .1998
"06/25/1992" "10:28:21" "SC PCM <5, 5-10, 10-20, 20-30, 30-40, 40-50, >50, and <0.3, >5
4.2994E-04 .0000 1.7850E-03 .0000 .0000 .0000 .0000 3.3391E-02 .2812
"06/25/1992" "10:30:25" "SC PCM 5-40, >40 and 0.1-0.3, >5
.1407 2.5473E-02 7.0764E-02
"06/25/1992" "10:30:30" "SC PCM 5-40, >40 and 0.2-0.3, >5
.1429 2.5396E-02 6.8760E-02
"06/25/1992" "10:53:03" "SC PCM 5-40, >40 with >5, <0.3 and AR>=10
.8530 2.5612E-02 7.0424E-02
"06/25/1992" "10:30:41" "SC PCM 5-40, >40 with <0.3 and AR>=20, >5
.1478 2.5544E-02 6.9820E-02
"06/25/1992" "10:30:47" "SC PCM 5-40, >40 with <0.4 and AR>=5, >5
.1005 2.5652E-02 4.2275E-02
"06/25/1992" "10:30:58" "SC PCM 5-40, >40 with <0.4 and AR>=20, >5
.1296 2.5808E-02 4.0611E-02
"06/25/1992" "10:31:15" "SC PCM 5-40, >40 with <0.3
.1140
"06/25/1992" "10:31:18" "SC PCM 5-40, >40 with >5
1.5134E-02
"06/25/1992" "10:53:08" "PS PCM 5-40, >40 and <0.3, >5
.9573 2.7270E-02 1.2536E-02
"06/25/1992" "10:32:46" "FBC PCM 5-40, >40 and <0.3 (there are no fibers >5)
.1139
"06/25/1992" "10:32:49" "SC PCM 5-40, >40 and <0.3, >5 (no discharged chrysotile}
.1770 2.4251E-02 6.5809E-02
"06/25/1992" "10:32:56" "PS PCM 5-40, >40 and <0.3, >5 (no discharged chrysotile)
.9285 2.4971E-02 1.2388E-02
"06/25/1992" "10:32:58" "FBC PCM 5-40, >40 and <0.3 (there are no fibers >5) (no discharged chrysotile)
.1133
"06/25/1992" "10:33:01" "SC PCM(16) 5-40, >40 and <0.3, >5
.1332 3.9379E-02 6.9539E-02
"06/25/1992" "10:53:11" "SC PCM(16) 5-40, >40 with >5, <0.3 (no discharged chrysotile)
.8193 3.7799E-02 6.2884E-02
"06/26/1992" "15:45:42" "SC PCM <5, >5 with W <0.3, >5
.9799 3.5380E-02 4.1829E-03.
"06/26/1992" "15:45:47" "SC PCM <5, >5 with W <0.3 and AR>=20, >5
286.
294.
260.
272.
0000
272.
272.
272.
•272.
•274.
•274.
•277.
•299.
•276.
•277.
•248.
•252.
•255.
•309.
•285.
•294.
•294.
209
603
648
188
938
985
935
902
247
223
690
923
320
767
966
063
100
395
380
370
170
40
56
32
.07
.81
.11
10.78
4.0372E-02
12.17
12
12
12
14
14
22
81
19
22
8.
14
21
49
45
61
61
.21
.19
.14
.82
.82
.79
.71
.51
.97
273
.90
.82
.04
.29
.95
.28
10
11
9
.0000
.0000
.0000
7 .1480
2.5783E-02 .1943
10 .2734
10
10
10
9
9
11
11
10
11
8
9
10
12
11
11
11
.2706
.2716
.2751
9.5423E-02
9.5218E-02
1.8207E-02
.0000
3.3427E-02
1.7084E-02
.4067
9.3166E-02
1.5245E-02
.0000
.0000
.0000
.0000
1.1853E-03
1.6400E-03
1.4970E-03
.0000
1.8856E-03
3.4336E-03
.0000
1.8812E-03
2.2016E-03
2.8966E-03
.9986
.7287
2.1545E-02
.9986
1.9186E-03
1.7562E-02
.9985
1.7750E-03
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
1.7176E-03
.0000
1.5835E-02
1.3154E-02
2.8284E-02
5.5902E-02
2.1116E-02
2 .8292E-02
4.6766E-03
5.3984E-02
2 .7767E-02
.0000
2.1511E-03
.0000
.0000
C.13
-------
.9179 3
"06/26/1992
.9929 2
"06/26/1992
.0000
"06/26/1992
.1453 2
"06/26/1992
7.8526E-02
"06/26/1992
7.2764E-02
"06/26/1992
2.7287E-02
"06/26/1992
.1014 2
"06/26/1992
7.7677E-02
"06/26/1992
7.2156E-02
"06/26/1992
3.3975E-02
"06/26/1992
.1018 2
"06/26/1992
8.6976E-02
"06/26/1992
7.5613E-02
"06/26/1992
.9546 3
"06/26/1992
.8983 2
"06/26/1992
9.5056E-02
"06/26/1992
7.3280E-02
"06/26/1992
.7876 3
"06/26/1992
9.3412E-02
"06/26/1992
7.2605E-02
"06/26/1992
.4927E-02 4.1831E-03
" "15:45:49" "SC PCM 5-40, >40 with AR>=20, >5 (4 categories)
.7947E-02 1.4448E-02
" "15:45:52" "SC PCM 5-40, >40
.8530 .1453 2.5612E-
" "15:46:03" "SC PCM 5-40, >40
.5612E-02 7.0424E-02
" "15:46:08" "SC PCM 5
with <0.3, 0.3-5, >5
•02 7.0424E-02
with <0.3, >5 (but exclude all M and MS)
"15:46:08" "SC PCM 5-
" "15:46:20"
3.2171E-02
"15:46:30"
.4772E-02 8.
" "15:46:32"
"SC PCM 5-
.1319
"SC PCM 5-
5164E-02
"SC PCM 5-
"15:46:32" "SC PCM 5-
15:46:42"
3.2170E-02
" "15:46:50"
.4718E-02 8.
" "17:18:46"
"SC PCM 5-
.1252
"SC PCM 5-
4964E-02
"SC PCM 5-
"17:18:46" "SC PCM 5-
"17:19:12"
.2171E-02
" "17:19:48"
.4743E-02 8.
" "17:19:51"
"SC PCM 5-
1304
"SC PCM 5-
5096E-02
"SC PCM 5-
"17:19:51" "SC PCM 5-
" "17:20:27"
.0145E-02 6,
" "15:47:32"
" "15:47:32"
" "15:47:47"
"SC PCM 5-
0929E-02
"SC PCM 5-
"SC PCM 5-
"SC PCM 5-
40, >40 with <0.3, >5 (2 studies)
40, >40 with <0.3, >5 (2 studies)
40, >40 with <0.3, >5 (chrysostile only)
40, >40 with <0.3, >5 (amphiboles only)
40, >40 with <0.3 and AR>=20, >5 (2 studies)
40, >40 with <0.3 and AR>=20, >5 (2 studies)
40, >40 with <0.3 and AR>=20, >5 (chrysotile only)
40, >40 with <0.3 and AR>=20, >5 (amphiboles only)
40, >40 with >5 (C and CS only), <0.3 (FBC only) (2 studies)
40, >40 with >5 (C and CS only), <0.3 (FBC only) (2 studies)
40, >40 with >5 (C and CS only), <0.3 (FBC only) (chrysotile only)
40, >40 with >5 (C and CS only), <0.3 (FBC only) (amphiboles only)
40, >40 with >5, <0.3 (no discharged) (2 studies)
40, >40 with >5, <0.3 (no discharged) (2 studies)
40, >40 with >5, <0.3 (chrysotile only - no discharged)
40, >40 with <0.3 and AR>=20, >5 (no discharged) (2 studies)
40, >40 with <0.3 and AR>=20, >5 (no discharged) (2 studies)
40, >40 with <0.3 and AR>=20, >5 (chrysotile only - no discharged)
C.14
•277.
•272.
•272.
•300.
•300.
•199.
•100.
•300.
•300.
•199,
•100,
-300,
•300,
•199,
-100,
-276,
•276
-174
•276,
-276
•174
958
935
935
,312
,312
.150
,725
.286
,286
.150
.723
.289
,289
.150
.724
.253
.253
.593
.253
.253
.593
23
12
12
11
11
8.
2.
11
11
8.
2.
11
11
8.
2.
7.
7.
3.
7.
7.
3.
.15
.19
.19
.46
.46
438
344
.41
.41
438
337
.34
.34
438
341
248
248
399
262
262
397
11
10
10
9
9
3
3
9
9
3
3
8
8
3
3
7
7
2
7
7
2
1.6070E-02
.2716
.2716
.2451
.2451
3.7044E-02
.5038
.2481
.2481
3.7048E-02
.5051
.1824
.1824
3.7044E-02
.5043
.4031
.4031
.1821
.4017
.4017
.1823
7.1418E-03
1.7176E-03
1.7176E-03
.1256
.1256
7.3631E-04
2.2569E-04
.1283
.1283
8.5321E-04
2.6046E-04
.8903
.8903
1.5135E-02
.0000
.8852
.8852
3.4305E-02
.1027
.1027
2.0719E-03
3,
2,
1.
3.
2,
1.
3,
2.
7.
2.
3,
2.
1,
3.
2.
3.
,0000
.0000
.0000
.3596E-02
.1933E-02
. 5232E-02
,0000
, 3499E-02
. 1909E-02
.4942E-02
.0000
. 3130E-02
.1784E-02
. 5254E-04
.3382E-04
.1198E-02
.1220E-02
.5901E-03
.1189E-02
, 1218E-02
.6781E-02
-------
.2255 3.0158E-02 5.1641E-02
"06/26/1992" "15:48:06" "SC PCM 5-40, >40 with AR>=20, >5 (4 categories) (no discharged) (2 studies)
1.3384E-02
"06/26/1992" "15:48:06" "SC PCM 5-40, >40 with AR>=20, >5 (4 categories) (no discharged) (2 studies)
1.9317E-02
"06/26/1992" "15:48:13" "SC PCM 5-40, >40 with AR>=20, >5 (4 categories) (chrysotile only - no discharged)"
2.1254E-11 3.0232E-02 1.2891E-02
"06/26/1992" "15:48:16" "SC PCM 5-40, >40 with AR>=20, >5 (4 categories) (amphiboles only)
.9863 1.9872E-02 1.5860E-02
"06/26/1992" "15:48:17" "SC PCM 5-40, >40 with AR>=3, >5 (4 categories) (amphiboles only)
.9919 2.0367E-02 1.3821E-02
"06/26/1992" "15:48:19" "SC PCM 5-40, >40 with AR>=5, >5 (4 categories) (amphiboles only)
.9915 2.0225E-02 1.3745E-02
"06/26/1992" "15:48:21" "SC PCM 5-40, >40 with AR>=10, >5 (4 categories) (amphiboles only)
.9908 2.0025E-02 1.4364E-02
"06/26/1992" "15:48:23" "SC PCM 5-40, >40 with AR>=30, >5 (4 categories) (amphiboles only)
.9770 2.0138E-02 1.6260E-02
"06/26/1992" "15:48:24" "SC PCM 5-40, >40 with AR>=50, >5 (4 categories) (amphiboles only)
.9403 2.0684E-02 1.7099E-02
"06/26/1992" "15:48:26" "SC PCM <5, 5-40, >40 with <0.3, >5 (no discharged chrysotile)
1.9186E-03 4.6766E-03 .1770 2.4251E-02 S.5809E-02
"06/26/1992" "15:48:36" "SC PCM 5-40, >40 with <0.3, 0.3-5, >5 (no discharged chrysotile)
4.6766E-03 .8164 .1770 2.4251E-02 5.5809E-02
"06/26/1992" "18:00:06" "SC PCMQ >10 (all structures AR>=3,W>0.2) (no WDC chrysotile or tremolite)
1.0063E-02
"06/26/1992" "18:10:08" "SC PCMQ <10,10-20,>20 (all structures AR>=3,W>0.2) (no WDC chrysotile/tremolite)"
2.3184E-02 4.3772E-04
"06/30/1992" "10:44:26" "PS PCM, USING SUM OF SURFACE AREA INSTEAD OF THE SUM OF THE NUMBER OF STRUCTURES"
1.0475E-06
"06/30/1992" "10:44:27" "PS PCM, USING SUM OF VOLUME INSTEAD OF THE SUM OF THE NUMBER OF STRUCTURES
1.0362E-06
"07/01/1992" "15:49:07" "SC PCM 5-40, >50 and W>5, <0.3 (all structures AR>=20) (no discharged) "
.3941 2.7504E-02 .2366
"07/02/1992" "11:56:05" "SC PCM <5, 5-10, 10-20, 20-30, 30-40, >=40 and W <0.3, >=5
1.1629E-03 .0000 5.3055E-03 .0000 .0000 .0000 .0000 6.1057E-02 .1258
"07/02/1992" "11:57:25" "SC PCM<5, 5-10, 10-20, 20-30, 30-40, >=40 andw<0.3, >=5 (no discharged)
9.5393E-04 .0000 5.7145E-03 1.9149E-02 .0000 .0000 .0000 8.2033E-02 .1382
"07/02/1992" "13:14:24" "SC PCM >=5 and W <0.3, >=5 (no discharged chrysotile)
5.0646E-03
"07/02/1992" "13:14:26" "SC PCM 5-40, >=40 (no discharged chrysotile)
-279.340
-279.340
-174.600
-103.580
-105.610
-105.251
-104.569
-103.778
-103.554
-248.966
-248.966
-223.907
-221.562
-285.826
-291.704
-253.886
-272.586
2.4594E-02
-248.218
2.2930E-02
-267.012
-261.069
13.11
13.11
3.396
6.615
10.10
9.464
8.284
6.865
6.499
8.273
8.273
24.09
19.60
39.16
55.77
19.50
11.47
6.9386E-02
6.779
6.7034E-02
49.63
31.75
9
9
2
4
4
4
4
4
4
8
8
9
8
11
11
9
8
6
10
10
.1571
.1571
.1823
.1569
3.7986E-02
4.9692E-02
8.0934E-02
.1425
.1641
.4067
.4067
3.4107E-03
1.1142E-02
.0000
.0000
2.0509E-02
.1757
.3412
.0000
.0000
.9916
.9916
1.1242E-03
1.3691E-02
8.0680E-03
8.4672E-03
9.2017E-03
2.3034E-02
5.9731E-02
.0000
1.9186E-03
.9521
.1793
.5002
.5002
.0000
.0000
.0000
.9815
.9907
3.1665E-02
2 .1102E-02
.3689
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
2.3551E-02
3.3351E-13
2.9496E-02
4.7021E-02
7.6482E-04
.0000
.0000
3.0991E-02
2.8947E-02
C.15
-------
5.4757E-03
"07/02/1992
7.0975E-03
"07/02/1992
4.4280E-03
"07/02/1992
2.6253E-03
"07/02/1992
1.9186E-03
"07/02/1992
2.8947E-02
"07/02/1992
.9815 3
"07/02/1992
2.0883E-05
"07/02/1992
8.8654E-05
"07/02/1992
1.6960E-05
"07/02/1992
1.6962E-02
"07/02/1992
.1069
"07/02/1992
.0000
.0000
"07/02/1992
.0000
.0000
"07/02/1992
.0000
.0000
"07/02/1992
.0000
.0000
"07/02/1992
.0000
.0000
"07/02/1992
.0000
.0000
"07/02/1992
"13:14:29" "SC PCM >=5 (no discharged chrysotile)
"13:14:31" "SC PCM 5-40 and W <0.3, >=5 (no discharged chrysotile)
"13:14:33" "SC PCM 5-40 (no discharged chrysotile)
"13:14:50" "SC PCM <5, 5-40, >=40 and W <0.3, >=5 (no discharged chrysotile)
4.6766E-03 .1770 2.4251E-02 S.5809E-02
13:15:02" "SC PCM <5, 5-40, >=40 (no discharged chrysotile)
5.4757E-03
" "13:14:35" "SC PCM <5, >=5 and W <0.3, >=5 (no discharged chrysotile)
.0991E-02 5.0646E-03
" "13:14:39" "SC PCM <5 and W <0.3 (there are no fibers W >= 5) (no discharged chrysotile)
" "13:14:41" "SC PCM <5, >=5 (no discharged chrysotile)
" "13:14:43" "SC PCM <5 (no discharged chrysotile)
" "13:14:45" "SC PCM 5-40, >=40 and W >=5 (no discharged chrysotile)
" "13:14:46" "SC PCM 5-40, >=40 and W <0.3 (no discharged chrysotile)
" "14:50:19" "SC PCM <5,5-10,10-20,20-40,>=40 and <0.15,0.15-0.3,0.3-1,1-5,>=5 (no discharged)
.0000 .0000 .0000 2.5839E-03 .0000 .0000 .0000 .1706
.0000 6.0103E-02 .0000 .5370 .0000 .2046 2.2673E-02 5.5023E-02
" "15:42:07" "SC M(14) <5,5-10,10-20,20-40,>=40 and <0.15,0.15-0.3,0.3-1,1-5,>=5
.0000 .0000 1.5838E-02 .0000 .0000 .0000 .0000 .0000
.0000 .0000 .0000 .8857 2.5155E-02 7.3297E-02 2.1648E-02 .2244
" "15:43:31" "SC PCM <5, 5-10,10-20,20-40,>=40 and <0.15,0.15-0.3,0.3-1,1-5,>-5
.0000 .0000 .0000 1.8178E-03 .0000 .0000 .0000 3.3464E-02
.0000 1.7003E-02 .0000 .7188 .0000 .1269 2.4549E-02 7.7920E-02
" "16:13:33" "SC M(14) <5, 5-10, 10-20, 20-40, >=40 and >=5, 1-5, 0.3-1, 0.15-0.3, <0.15
.0000 .0000 .0000 .0000 .0000 .0000 1.5838E-02 .0000
.0000 .0000 7.3297E-02 .0000 2.5155E-02 -1.2727E-09 2.1648E-02 .2244
" "16:30:29" "SC PCM <5, 5-10, 10-20, 20-40, >=40 and >=5, 1-5, 0.3-1, 0.15-0.3, <0.15
.0000 .0000 .0000 .0000 .0000 1.8178E-03 .0000 .0000
.0000 .1020 .1269 .0000 .0000 -1.9608E-09 2.4549E-02 7.7920E-02
" "17:20:59" "SC PCM <5, 5-10, 10-20, 20-40, >=40 and <0.3, 0.3-1, 1-5, >=5
.0000 1.2014E-03 .0000 .0000 .0000 5.4344E-03 .0000 .0000
.1431 2.4572E-02 6.7829E-02
" "17:18:45" "SC PCM <5, 5-10, 10-20, 20-40, >=40 and <0.3, 0.3-1, 1-5, >=5 (with CI)
-285.
-283.
-286.
-248.
-261.
-267.
-301.
-285.
-301.
-267,
-260.
-248.
.0000
-279,
.0000
-272,
.0000
-279.
.0000
-272.
.0000
-272,
.0000
.450
,414
.783
.966
,069
.012
,986
,450
,288
.213
.510
,178
.614
.616
,614
.616
.680
103.2
99.33
107.2
8.273
31.75
49.63
132.5
103.2
132.0
60.09
33.58
6.674
.0000
23.57
.0000
11.60
.0000
23.57
.0000
11.60
.0000
11.64
.0000
10
10
10
8
10
10
10
11
10
10
10
6
.0000
9
.0000
7
.0000
8
.0000
6
.0000
8
.0000
.0000
.0000
.0000
.4067
.0000
.0000
.0000
.0000 1
.0000
.0000
.0000
.3515
.0000
4.2818E-03
.0000
.1137
.0000
1.8615E-03
.0000
7.0614E-02
3.3464E-02
.1674
.0000
9875 5.
9753 5.
9671 5.
0000
0000
0000
6775
.000 5.
6657
9371 5.
9982 3.
0000
2.5171E-02
0000
.0000
0000
.1020
0000
.0000
0000
1.7003E-02
0000
1.1976E-02
7855E-02
5205E-02
2602E-02
0000
9907
0000
1194
7855E-02
1169
1633E-02
3808E-02
0000
.0000
0000
.0000
0000
.0000
0000
.0000
0000
.0000
0000
.8382
" -272.680
11.64
.1674
.0000
.0000
C.16
-------
.0000 .0000
.0000 .1431
"07/06/1992" "09:
5.7162E-02
"07/06/1992" "10:
1.3791E-05
"07/06/1992" "10:
2.5174E-03
"07/06/1992" "11:
1.3791E-05
"07/06/1992" "11:
2.5174E-03
"07/06/1992" "11;
.0000 .0000
.0000 .0000
"07/06/1992" "13:
.0000 .0000
.0000 .0000
"07/06/1992" "15;
.0000 .0000
.0000 .0000
"07/07/1992" "13:
.0000 .0000
.0000 .0000
"07/07/1992" "14:
.0000 .0000
.0000 .0000
"07/07/1992" "15:
.0000 .0000
.0000 .0000
"07/07/1992" "15:
.0000 .0000
1,
2,
:32:37"
:57:33"
:57:42"
:18:57"
:19:00"
:33:31"
7,
:44:41"
:20:12"
1,
:28:40"
:46:55"
:06:25"
1,
:24:54"
.4496 -1.8055E-07 2,
"07/07/1992" "15:
.0000 .0000
;30:50"
.3578 -7.4710E-09 2.
"07/07/1992" "15:
.0000 .0000
:50:02"
.0000 1.4806E-10 2,
"07/07/1992" "16:
:31:01"
,2014E-03
.4572E-02
"SC PCM
"SC PCM
"SC PCM
"SC PCM
"SC PCM
"PS PCM
.0000
.3795E-02
.0000 .0000 .0000 5.4344E-03
6.7829E-02
>-40 and W <0.3, >=5
(no length or width categories)
length >= 5
(no length or width categories)
length >= 5
<5, 5-10, 10-20, 20-40, > = 40 and W <0.15, 0.15-0
5.4514E-03 .0000 .0000 .0000
.0000 .0000 .0000 .4188
.0000
.3, 0.3-1,
.0000
2.3729E-02
"PS M(14) <5, 5-10, 10-20, 20-40, >=40 andW<0.15, 0.15-0.3, 0.3-1
.0000
.0000
"SC PCM
.0000
.2887E-02
"SC PCM
.0000
.1038
"PS PCM
.0000
,0000
"PS PCM
,0000
,0417E-02
"PS PCM
.0000
.6883E-02
"PS PCM
,0000
.4136E-02
"PS PCM
.0000
.4043E-02
"SC PCM
7.9976E-02 .0000 .0000 .0000
.0000 .0000 .2160 .3401
<5, 5-10, 10-20, 20-40, >=40 andW<0.1, 0.1-0.3
.0000 1.7628E-03 .0000 .0000
.0000 .7253 .0000 .1256
<5, 5-10, 10-20, 20-40, >=40 and Complex, >=1, 0.
.0000 .0000 .0000 1.2681E-03
4.4982E-02 .0000 6.8896E-03 -1.8720E-10
<5, 5-10, 10-20, 20-40, >=40 and Complex, >-l, 0.
.0000 .0000 .0000 .0000
.0000 .0000 .0000 -1.0440E-08
<5, 5-10, 10-20, 20-40, >=40 and W <0.15, 0.15-0
.0000 .0000 .0000 .0000
.0000 .0000 .3577 -1.1622E-08
5-10, 10-20, 20-40, >=40 andW<0.15, 0.15-0.3,
.2655 .0000 .0000 .2317
3.3488E-02
5-10, 10-20, 20-40, > = 40 and Complex, >=1, 0.3-1
2.8399E-03 .2072 .0000 .0000
4.6405E-02
5-10, 10-20, 20-40, > = 40 and Complex, >=1, 0.4-1
3.1200E-03 5.6519E-02 .0000 .0000
.1039
5-10, 10-20, 20-40, >=40 and Complex, >=1, 0.4-1
.0000
2.1931E-02
.0000
1-5, >=5
.4190
2.3807E-02
, 1-5, >=5
.3587
8.7607E-02
, 0.3-1, 1-5, >=5
.0000
2.4639E-02
3-1, 0.15-0
.0000
2.3811E-02
3-1, 0.15-0
.0000
3.4370E-02
.3, 0.3-1,
.0000
2.4136E-02
0.3-1, >=1,
4.4614E-04
, 0.15-0.3,
.4218
, 0.2-0.4,
.0000
, 0.2-0.4,
3.8041E-02
7.9469E-02
.3, <0.15
.0000
9.7120E-02
.3, <0.15
.0000
2.4421E-03
>=1, Complex
.4217
4.6418E-02
Complex
.0000
<0.15
1.0419E-02
<0.2
.0000
<0.2
.0000
" -310.173
" -322.668
" -309.474
11 -322.668
" -309.474
" -273.275
3.9527E-02
" -275.607
5.1851E-03
" -272.609
.0000
" -274.051
7.6933E-03
" -294.903
.0000
" -274.044
.0000
" -275.101
.0000
" -274.044
.0000
" -273.078
.0000
" -272.998
.0000
118.6
131.4
107.2
131.4
107.2
12.67
.0000
15.73
.0000
11.58
.0000
14.09
.0000
61.42
.2615
14.95
.0000
17.10
2.5093E-02
14.95
.0000
12.73
.0000
12.43
.0000
11
11
11
11
11
7
.0000
8
.0000
7
.0000
6
.0000
10
.0000
6
.2071
6
.0000
7
.0000
8
.0000
6
.0000
.0000
.0000
.0000
.0000
.0000
7.9868E-02
4.3412E-02
4.5597E-02
.0000
.1146
.0000
2.7817E-02
.0000
.0000
.0000
1.9811E-02
2.8387E-03
8.0952E-03
2.7684E-02
3.5898E-02
.0000
.1208
.2888
5.2239E-02
1.1976E-02
2645 9.
6354
9691 5.
6354
9691 o.
0000
.0000
0000
.0000
0000
9.6416E-02
0000
.0000
0000
.7385
0000
2.8200E-04
0000
.0000
0000
.0000
0000
.0000
0000
.8382
1781E-02
1125
3273E-02
1125
3273E-02
0000
.0000
0000
.0000
0000
.0000
0000
.0000
0000
.0000
0000
.0000
0000
.0000
0000
.0000
0000
.0000
0000
C.17
-------
.0000 1.
.0000 1.
"07/07/1992"
.0000
"07/08/1992"
.0000
.0000 -7.
"07/08/1992"
.0000
.0000
"07/08/1992"
.0000
.0000 -4.
"07/08/1992"
.0000
.0000 -2.
"07/08/1992"
.0000
.9329 2.
"07/08/1992"
.0000
.0000 -1.
"07/08/1992"
.0000
"07/08/1992"
.0000
"07/08/1992"
.0000
"07/08/1992"
.0000
5.9101E-03
"07/08/1992"
.0000
.0000
"07/08/1992"
.0000
.0000
"07/08/1992"
.0000 6.
.0000
"07/08/1992"
5795E-03 .0000 .0000 4.9393E-02 .0000 1.1903E-02 .0000
5002E-10 2.4148E-02 3.9445E-02
"16:56:39" "PS PCM 5-10, 10-20, 20-40, >=40 and Complex, >=1, 0.4-1, <0.4
0000 .0000 .4346 .0000 .0000 .0000 .0000
"08:36:46" "PS PCM <5, 5-10, 10-20, 20-40, >=40 and W <0.4, 0.4-1, >=1, Complex
0000 .0000 .0000 .0000 .0000 .0000 .0000
3093E-07 2.5732E-02 7.1660E-02
"08:42:38" "PS PCM <5, 5-10, 10-20, 20-40, >=40 and Complex, >=1, 0.4-1, <0.4
0000 .0000 .0000 .0000 .0000 3.9837E-03 3.0906E-02
9329 2.4067E-02 9.1268E-02
"09:04:18" "PS PCM >=40, 20-40, 10-20, 5-10, <5 and Complex, >=1, 0.4-1, <0.4
0000 .6995 .0000 .0000 .0000 .0000 .0000
5399E-07 3.0207E-02 2.7264E-03
"09:09:10" "PS PCM >=40, 20-40, 10-20, 5-10, <5 andW<0.4, 0.4-1, >=1, Complex
0000 .0000 .0000 .0000 .9993 .0000 .0000
5203E-07 4.4540E-02 2.2204E-03
"09:16:06" "PS PCM <5, 5-10, 10-20, 20-40, >=40 and Complex, <0.4, 0.4-1, >=1
0000 .0000 .0000 .0000 .0000 3.9837E-03 .0000
6822E-10 2.4067E-02 9.1268E-02
"09:24:17" "PS PCM <5, 5-10, 10-20, 20-40, >=40 and Complex, <0.4, >=1, 0.4-1
0000 .0000 .0000 .0000 .0000 .0000 .3006
2442E-06 3.0200E-02 2.7271E-03
"09:27:24" "PS PCM 5-10, 10-20, 20-40, >=40 and Complex, <0.4, 0.4-1, >=1
0000 3.9837E-03 .0000 .0000 3.0906E-02 .0000 3.2175E-02
"09:57:02" "PS PCM <5, 5-10, 10-20, 20-40, >=40 andW<0.4, 0.4-1, >=1 (no complex
0000 .0000 .0000 .0000 .0000 4.5807E-02 1.8949E-02
"10:17:31" "PS PCM<5, 5-10, 10-20, 20-40, >=40 andW>=l, 0.4-1, <0 . 4 (no complex
0000 .0000 .0000 4.5810E-02 .0000 .0000 .0000
"13:34:26" "PS PCM <5, 5-10, 10-20, 20-40,>=40 and >=1, 0.4-1, <0.4 and Complex with
0000 .0000 .0000 3.7007E-02 .0000 .0000 .0000
.0000 5.3636E-10 2.3957E-02 8.8626E-02
"14:44:47" "SC PCM <5, 5-10, 10-20, 20-40, >=40 and >=1, 0.4-1, <0.4 and Complex with
0000 .0000 .0000 .0000 5.4027E-05 .9999 .0000
0000 -9.9125E-08 1.4853E-02 9.5085E-04
"14:48:02" "SC PCM Complex with 6 lengths and <5, 5-10, 10-20, 20-40, >=40 and >=1, 0
0000 .0000 .0000 .0000 .0000 .0000 .0000
0000 -1.8353E-07 4.1522E-02 3.4939E-03
"14:58:17" "SC PCM <5, 5-10, 10-20, 20-40, >=40 and <0.4, 0.4-1, >=1 and Complex with
5666E-04 .0000 .0000 5.0087E-03 .0000 5.8557E-02 2.6025E-02
0000 9.1494E-02 2.4199E-02 3.6281E-02
.0000
"
.0000
»
S.1242E-02 8.
"
.0000
"
.0000
"
.0000
"
.0000 3.
"
.0000
"
.0000
structures) "
3.3914E-02
structures) "
3.3927E-02 1.
6 lengths "
.0000 3.
6 lengths "
.0000
.4-1, <0.4 "
.0000
6 lengths "
.0000
"15:13:14" "SC PCM <5, 5-10, 10-20, 20-30, >30 for<.3 and 5-10, 10-20, 20-30, 30-40, 40-50, >50 for>5"
.0000
-278,
.1909
-274.
.198
.519
.8340E-04
-273,
.0000
-292.
,3005
-298,
.0000
-273.
.814
.314
.842
,814
.0906E-02
-292,
.0000
-273.
.0000
-275,
.0000
-275,
.314
,814
.298
.298
.8952E-02
-273.
.774
,4493E-02
-311,
.0000
-293.
.0000
-272.
.0000
-274.
.176
.905
.749
,867
.0000
21.
.3745
14.
79
77
2.0129E-02
14.
.0000
55.
.0000
69.
.0000
14.
.0000
55.
.6994
14.
.0000
16.
.9000
16.
.0000
14.
.0000
143
.0000
58.
.0000
11.
.8173
15.
11
42
62
11
42
11
34
34
04
.3
17
85
39
2.1272E-03
9 8.
.0000
6 2.
5.4116E-02
9
.0000
10
.0000
10
.0000
8 7.
3.2175E-02
10
.0000
8 7.
. 9329
7 2.
-1.1481E-05
8 3.
.9000
7 4.
.0000
10
.0000
9
1.5413E-05
7
.0000
8 5.
.0000
8203E-03
-1.0420E-06
1279E-02
.0000
1177
.0000
0000
.0000
0000
.0000
8127E-02
.0000
0000
.0000
8127E-02
1.4118E-10
1478E-02
2.7020E-02
6935E-02
2.6966E-02
9819E-02
.9198
0000
.0000
0000
.1811
1049
.0000
1113E-02
3.
0000
2.
0000
1.
0000
3.
0000
0000
0000
0000
0000
2.
0000
8.
0000
8.
0000
0000
0000
0000
0000
0210E-02
2334E-02
2797E-02
2175E-02
0000
0000
0000
0000
4067E-02
1.
8065E-02
1.
8070E-02
0000
0000
0000
0000
3.
.0000
0000
1.3918E-02
0000
.8508
0000
.0000
0000
.0000
0000
6.9620E-04
0000
.0000
0000
.0000
0000
9.1268E-02
2928E-03
2930E-03
0000
2.8166E-03
0000
.0000
0000
.8189
0000
.0000
3674E-03
C.18
-------
.0000 7
"07/08/1992
.0000 6
.0000
"07/09/1992'
.0000 6
.0000
"07/09/1992'
.1140
"07/09/1992'
1.5134E-02
"07/09/1992'
3.3178E-03
"07/09/1992'
.6999 -1.
"07/09/19921
5.6375E-03
"07/09/1992'
.0000 8,
"07/09/1992'
.0000
"07/09/1992'
.1450 2.
"07/09/1992'
8.1133E-02
"07/09/1992'
7.1689E-02
"07/09/1992'
.0000
.0000
"07/09/1992'
.0000
.0000
"07/09/1992'
.0000
"07/09/1992'
.0000
.1045 9,
"07/10/19921
1.7176E-03
.4658E-02 .4261 .0000 .0000 .0000 .2326 .2633 2.3275E-02
" "15:45:37" "SC PCM <5,5-10,10-20,20-40,>=40 and <0.3, 0.3-1, >=1 and Complex with 6 lengths "
.3774E-04 .0000 .0000 4.1770E-03 .0000 1.8608E-02 .0000 .0000
.0000 7.4421E-02 2.4148E-02 7.5365E-02
1 "09:11:08" "SC PCM <5,5-10,10-20,20-40,>=40 and <0.4, 0.4-1, >=1 and C and CS with 6 lengths"
5666E-04 .0000 .0000 5.0087E-03 .0000 5.8557E-02 2.6025E-02 .0000
0000 9.1494E-02 2.4199E-02 3.6281E-02
1 "11:30:24" "SC PCM 5-40, >=40 and W <0.3 "
1 "11:30:28" "SC PCM 5-40, >=40 and W >5
1 "11:30:29" "SC PCM 5-40 and W<0.3, >5
"11:30:31" "SC PCM 5-40, >=40 and W <0.3, >5 and Length < 5 (no widths) (with 95% CI)
6245E-07 1.5694E-02 2.9785E-03
"12:05:26" "SC PCM <8, 8-15, 15-25, 25-40, >=40 and W <0.3, >=5
.0000 .0000 3.8505E-02 .0000 .0000 .1430 2.3986E-02 7.0573E-02
"12:05:47" "SC PCM 10-20, 20-40, >=40 and W <0.3, >=5
5144E-03 .1634 2.3902E-02 5.9924E-02
"14:31:48" "SC PCM <5, 5-10, 10-20, 20-40, >=40 and <0.4, 0.4-1 and Complex with 6 lengths "
0000 1.0729E-02 3.2242E-03 .0000 .0000 .8210 .0000 .0000
"14:55:48" "SC PCM 5-40, >=40 and W <0.3, >5 (multipy control animals/response by 10^6) "
6668E-02 7.0300E-02
"15:18:32" "SC PCM 5-40,>=40 and W<0.3,>5 (2 studies) (control animals/response * 10~6)
"15:18:32" "SC PCM 5-40,>=40 and W<0.3,>5 (2 studies) (control animals/response * 10~6) "
"15:46:24" "SC M(16) <5, 5-10, 10-20, 20-40, >=40 and <0.15, 0.15-0.3, 0.3-1, 1-5, >=5
0000 .0000 1.7778E-02 .0000 .0000 .0000 .0000 .0000
0000 .0000 .0000 .9076 1.9604E-02 5.5023E-02 2.2577E-02 .2350
"15:47:33" "PS M(16) <5, 5-10, 10-20, 20-40, >=40 and <0.15, 0.15-0.3, 0.3-1, 1-5, >=5
0000 .0000 .1318 .0000 .0000 .0000 .0000 .1955
0000 .0000 .0000 .0000 .1961 .3931 2.1022E-02 7.2003E-02
"16:06:08" "SC PCM Length < 5 (no widths) and 5-40, >=40 and W <0.3, >5 (with 95% CI)
1453 2.5612E-02 7.0424E-02
"17:22:46" "PS PCM(FsB)5-10,10-20,20-40,>40 and<.4,.4-1,>1 C only and CS with 6 length cat."
0000 2.1747E-03 7.1876E-02 5.3176E-02 .0000 .0000 .7673 .0000
8618E-04 .0000 .0000 .0000 -5.1981E-08 2.4947E-02 6.3726E-02
"09:34:47" "SC PCM Length < 5 (no widths) and 5-40, >=40 and W >5, <0.3 (with 95% CI) "
.8530 2.5612E-02 7.0424E-02
3.2685E-02
' -272.964 12.11
.0000 .8789
7 9.6267E-02
2.3299E-02 .0000
.0000 .0000
.0000 .0000
-272.749
.0000
-277.690
-299.923
-309.212
-318.153
-272.701
-273.565
-273.213
.0000
-2.774611E+07
-5.549198E+07
-5.549198E+07
-310.198
.0000
-304.903
.0000
-272.935
-272.433
.0000
11.
8173
22.
81.
107
159
11.
13.
12.
0000
12.
12.
12.
37.
0000
25.
0000
12.
11.
0000
85
79
71
.4
.1
78
25
51
11
10
10
43
27
19
32
7
.0000
11 1.
11
11
9
9
9
8
7.6252E-02
10
9
9
11
.0000
10 4.
.0000
10
6 7.
.0000
1049
.0000
8207E-02
0000
0000
0000 8.
2254
1508 1.
1290
8.8839E-02
2770 1.
2074
2074
0000
.0000
0203E-03
.0000
2716
8319E-02
.0000
0000
.0000
9986 2.
7287 5.
9697 5.
1546E-02
0000
4457E-02
0000
2.4809E-02
7130E-03
1240 2.
1240 2.
0000
.0000
0000
.0000
0000 1.
0000
.0000
0000
.0000
8284E-02
5902E-02
2884E-02
0000
0000
0000
0000
3.9978E-02
0000
6668E-02
6668E-02
0000
.0000
0000
8.3400E-02
7176E-03
0000
.0000
-272.935
12.19
10
.2716
.0000
.0000
C.19
-------
"07/10/1992"
.0000 1.
.4446
"07/10/1992"
.0000
.0000
"07/10/1992"
.0000
.0000
"07/10/1992"
.0000
.0000 1.
"07/10/1992"
.0000
.0000
"07/10/1992"
.0000
.0000
"07/10/1992"
.0000
.0000
"07/10/1992"
2.5406E-02 1
"07/10/1992"
2.6927E-02 1
"07/16/1992"
1.0898E-02 4
"07/16/1992"
.0000 6.
4.2294E-02
"07/16/1992"
.0000
"07/20/1992"
7.8262E-02 8
"07/20/1992"
3.6883E-02 9
"07/20/1992"
.1132 2.
"07/20/1992"
.1544
"10:28:12"
5443E-03
0000
"11:30:52"
0000 7.
0000
"12:20:58"
0000
0000
"12:15:35"
0000
0111E-02
"12:29:04"
0000
0000
"12:32:22"
0000
0000
"13:46:21"
0000
0000
"14:23:02"
.0021E-02
"15:20:11"
.2277E-02
"09:13:25"
.2320E-03
"09:14:13"
5172E-04
"16:17:33"
0000
"11:16:23"
.3237E-02
"11:16:24"
.8082E-03
"11:16:25"
6895E-02
"11:16:26"
1622
"SC PCM <5, 5-10, 10-20, 20-40, >=40 and <0.15,0
0000 .0000 .0000 .0000
5538 .0000 .0000 .0000
"SC PCM <5, 5-10, 10-20, 20-40, >=40 and >=5,l-5
4663E-05 7.7508E-02 .0000 .0000
0000 .0000 .0000 .0000
"SC PCM <5, 5-10, 10-20, 20-40, >=40 £ >-.3,.15-
.0000 .0000 5.4696E-02 .0000
,0000 -1.9458E-08 4.3684E-03 7.3528E-04
"SC PCM <5, 5-10, 10-20, 20-40, >=40 S <.15,.15-
.0000 .0000 .8618 .0000
.0000 -5.4691E-08 4.2054E-03 9.5365E-04
"SC PCM Complex w/6 lengths S <5, 5-10, 10-20,
.0000 .0000 9.2717E-05 .0000
.9970 -1.6111E-06 4.7046E-03 1.0201E-02
"SC PCM Complex w/6 lengths £ <5, 5-10, 10-20,
.0000 .0000 .0000 .0000
.0000 -1.5832E-08 4.6132E-03 2.6062E-04
"SC PCM Complex w/6 lengths £ <5, 5-10, 10-20,
.0000 .0000 7.8080E-05 .0000
.0000 .8977 4.6447E-03 1.1447E-02
"SC PCM >= 20
"SC PCM >= 20 and < 0.4
"SC PCM <5, 5-10, 10-20, 20-40, >=40 & <0.4
.8006 .0000 .0000 .0000
"SC PCM <5, 5-10, 10-20, 20-40, >=40 S <0.2,
.0000 1.4206E-02 .0000 .0000
"SC PCM <5, 5-10, 10-20, 20-40, > = 40 S, >-l.
.0000 .0000 5.0739E-02 .0000
"SC PCM Length >=20 and Width <0.2
"SC PCM Length >=30
"SC PCM Length >=30 and Width <0.4
"SC PCM Length >=30 and Width <0.2
.15-0. 3, 0.3-1, 1-5, >=5 (Mesotheliomas) "
.0000
-5.0341E-08
.0000
4.6448E-03
.0000
S.0542E-04
,0.3-1,0. 15-0. 3,<0. 15 (Mesotheliomas) "
.0000
4.2322E-10
.0000
1.4830E-03
.0000
1.1808E-02
.3,<.15 £ Complex w/6 lengths (Meso.)"
.8768
6.8520E-02
.0000
. 3,> = .3 £ Complex w/6 lengths (Meso.)"
3.1225E-02
20-40, >=40 £
.0000
20-40, >=40 £
.0000
20-40, >=40 £
.0000
and Complex
.1114
0.2-0.4 and
.0000
0.4-1, <0.4
1.1999E-02
.0000
<-15, .15-. 3
.0000
> = .3, .15-. 3
.0000
<.15,>=.3, .
.0000
only with 6
.0000
.0000
,>=.3 (Meso.)"
.0000
,<.15 (Meso.)"
.0000
15-. 3 (Meso.)"
.0000
"
"
lengths "
7.2753E-02
Complex only w/6 lengths"
.8095
(no complex
2.7987E-02
.0000
structures) "
.0000
-
"
"
"
-60.6932
.0000
-59.0535
.0000
-60.6227
.0000
-60.6067
9.6821E-02
-59.3875
.0000
-60.7657
.0000
-59.3238
.0000
-282.315
-285.310
-273.250
2.4602E-02
-273.150
.0000
-273.681
6.2087E-04
-313.069
-287.475
-316.358
-331.356
14.64
.0000
12.20
.0000
14.73
.0000
14.72
.0000
11.93
2.9416E-03
15.50
.1489
12.04
.0000
28.39
39.84
12.63
4.1308E-02
12.42
.0000
13.88
.0000
137.2
40.36
110.2
132.3
9
.0000
9
.0000
9
.0000
8
.0000
9
.0000
9
.0000
10
.0000
11
11
7
8
.1250
7
.0000
11
11
11
11
.1005
.0000
.2015
.0000
9.7866E-02
.0000
S.4088E-02
.0000
.2167
.0000
7.7300E-02
.5814
.2815
.0000
2.0676E-03
.0000
8.0924E-02
.1328
.0000
5.2712E-02
.9084
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.2697
.0000
.1022
.2055
.4983
.0000 1.
.0000
5.0575E-02
.0000 2.
2.4677E-02
.5000
.4976
.5702
.5175
0000
.0000
0000
.0000
0000
.0000
0000
.0000
0000
.0000
0000
.0000
0000
.0000
7945
5017
5099E-04
0000
2.4376E-02
4004E-04
4.0742E-02
5000
5024
4298
4825
C.20
-------
"07/20/1992" "11:16:27" "SC PCM Length >=20 and Width >=0.4 "
2.9009E-02 1.0020E-02
"07/20/1992" "11:16:28" "SC PCM Length >=10
3.0399E-02 1.0603E-02
"07/20/1992" "16:59:33" "SC PCM FSB with 5-40,<0.3, F&B with > = 40,<0.3, and Complex only with >=40,>=5 "
.1467 2.5581E-02 7.0175E-02
"07/20/1992" "17:02:26" "SC PCM F=>40,<.3, B=>40,<.3, Complex=>40,>5, F=5-40,<.3, B=5-40,<.3 "
5.9873E-02 8.8266E-06 8.3673E-04 2.9366E-02 .1525
"07/21/1992" "13:14:11" "SC PCM 5-40 S <0.3, >=40 S<0.3, and >=40 S >=5
.1453 2.5612E-02 7.0424E-02
"07/21/1992" "13:19:20" "SC PCM >=40 £ >=5, 5-40 £ <0.3, and >=40 £<0.3
.8530 2.5612E-02 7.0424E-02
"08/06/1992" "15:38:36" "SC PCM length >= 8 and width < 0.25
5.4171E-02 4.4551E-03
"12/14/1992" "16:30:46" "SC PCM Length < 5 (no widths) and 5-40, >=40 and W <0.3, >5 (with 95% CI)
.0000 .8530 .1453 2.5612E-02 7.0424E-02
•285.
•292.
272.
•272.
•272.
•272.
.590
.283
.933
.537
.935
.935
34.
56.
12.
11.
12,
12.
.13
.96
.20
.55
.19
.19
11
11
10
8
10
10
.0000
.0000
.2712
.1718
.2716
.2716
-306.530
" -272.935
"12/15/1992" "10:16:36" "(all except chrysotile) SC PCM <5,5-10,10-20,20-40,>=40 and <0.15,0.15-0.3,0.3-1,1-
.0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000
.0000 .0000 .0000 .0000 7.6546E-02 .0000 .9027 .0000 .0000 2,
"12/15/1992" "10:17:47" "(chrysotile only) SC PCM <5,5-10,10-20,20-40,>=40 and <0.15,0.15-0.3,0.3-1,1-5,>=5
.0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .1719
.0000 2.0494E-02 .0000 .5626 .0000 .0000 .2450 .0000 -6.1822E-05 3,
100.2
12.19
11 .0000
10 .2716
.3007 .6993
3.6654E-02 .9633
1.7376E-03 .8516
.4988 .4405
1.7176E-03 .8530
.1453 1.7176E-03
.3319
.0000
.6681
1.7176E-03
5,>=5
0000 .0000
0714E-02 2.2772E-02 E
" -199.
0000 .0000
2157E-02 3.7205E-02
-100.359 1.685 3 .6399
.0000 .0000 .0000 .0000
.6916E-02
146 8.430 1 2.9414E-03 1.9054E-08
.0000 .0000 .0000 .0000
"12/17/1992" "14:09:01" "(chrysotile only) SC PCM <5,5-10,10-20,20-40,>=40 and >=5,1-5,0.3-1,0.15-0.3,<0.15
.0000 .0000 .1402 9.5597E-03 .0000 .0000 .0000 .0000 .0000
.0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 1.2193E-07 5.
" -224.122 109.7
0000 .0000 .0000 .0000
9591E-02 1.1295E-04
3 .0000 .0000
.0000 .8502
"12/18/1992" "09:34:18" "(chrysotile only) SC PCM >=40,<5,5-10,10-20,20-40 and <0.15,0.15-0.3,0.3-1,1-5,>-5
.0000 .2450 .0000 .0000 .0000 .0000 .0000 .0000 .0000
.0000 .0000 .0000 .0000 .0000 .0000 2.0501E-02 .0000 .5625 3.
"10/08/1993" "16:02:37" "SC PCM 5-40, >=40 with >=5, < 0.3 (All except chrysotile)
9.8000E-02 .9019 2.6668E-02 8.6399E-02
"10/12/1993" "17:03:34" "SC PCM >=40, 5-40 with <0.3, >=5 (Chrysotile only) (dl*1000,d2*100,d4*30)
.2883 .2214 2.6668E-02 3.3475E-04
"03/23/1994" "14:46:29" "SC - Length: <5, 5-40; Width: <.3, >5
" -199.
0000 .0000
2158E-02 3.7206E-02
-2.774594E+07 2.321
-2.774604E+07 8.826
" -271.156
8.005
146
.0000
8.430
.0000
3 .5082
3 3.7196E-02 .0000
.1719 .0000
.0000
3 3.0963E-02 .3958
7.1027E-05
9.4552E-02
10 0.6280 2.5270E-03 O.OOOOE+00
C.21
-------
0.9161 8.1337E-02 2.3018E-02 8.4603E-02
"03/23/1994" "14:52:42" "SC - Length: 5-20, >20; Width: <-3, >5 " -276.444 17.53 10 S.2729E-02 1.5736E-02 O.OOOOE+00
0.3639 0.6204 2.1557E-02 1.0547E-02
"03/23/1994" "15:13:12" "SC - Length: 5-40, >40; Width: >5 " -303.315 90.11 11 O.OOOOE+00 0.3142 0.6858
6.1680E-02 1.6153E-02
"03/23/1994" "15:13:26" "SC - Length: 5-20, >20; Width: <.3 " -284.330 35.14 11 O.OOOOE+00 1.5189E-02 0.9848
2.2466E-02 1.0049E-02
"03/23/1994" "15:35:48" "SC - Length: 5-20, >20; Width: <.3 and Length: 5-40, >40; Width: >5 " -275.192 15.27 10 0.1217 1.1560E-02 0.2105
O.OOOOE+00 0.7780 2.2547E-02 1.5203E-02
"06/28/1994" "15:56:48" "SC (smooth): L: 5-10, 10-20, 20-40, >40; W: 0-.15, .15-.3, .3-1, 1-5, >5 " -274.350 13.01 8 0.1108 O.OOOOE+00 O.OOOOE+00
O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 9.3741E-02 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 1.7745E-02 O.OOOOE+00 2.8592E-02 O.OOOOE+00 O.OOOOE+00 0.5198 O.OOOOE+00
O.OOOOE+00 O.OOOOE+00 0.3401
2.1446E-02 6.0462E-02
"06/30/1994" "08:40:56" "SC (smooth): L: <5, 5-10, 10-20, 20-40, >40; W: 0-.15, .15-.3, .3-1, 1-5, >5 " -285.085 51.34 9 O.OOOOE+00 2.0498E-04 7.1561E-05
O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
0.7434 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 0.2563 4.5062E-02 6.3313E-02
"06/30/1994" "09:27:22" "SC (smooth): L: <5, 5-10, 10-20, 20-40, >40; W: 0-.15, .15-.3, .3-1, 1-5, >5 " -272.652 10.23 9 0.3313 O.OOOOE+00 O.OOOOE+00
O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 0.3393 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
2.2450E-02 O.OOOOE+00 O.OOOOE+00
O.OOOOE+00 0.1061 O.OOOOE+00 O.OOOOE+00 0.5322 2.1610E-02 2.6262E-02
"07/05/1994" "12:25:40" "SC (smooth): L: <5, 5-10, 10-20, 20-40, >40; W: 0-.15, .15-.3, .3-1, 1-5, >5 " -274.379 14.21 8 7.5695E-02 O.OOOOE+00 O.OOOOE+00
O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 8.6406E-04 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 9.4722E-02 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
1.8899E-02 O.OOOOE+00 O.OOOOE+00
O.OOOOE+00 0.4309 O.OOOOE+00 O.OOOOE+00 0.4546 2.3140E-02 2.9659E-02
"07/05/1994" "13:42:25" "SC (smooth) - Length: 5-40, >40; Width: <-3, >5 " -273.849 12.39 10 0.2589 5.4909E-03 O.OOOOE+00
0.6729 0.3216 2.1457E-02 3.1507E-02
"07/05/1994" "13:46:50" "SC (smooth) - Length: 5-40, >40; Width: <-3, >5 " -274.767 14.82 10 0.1382 3.1150E-03 O.OOOOE+00
0.6166 0.3803 2.2734E-02 3.0645E-02
"07/11/1994" "09:30:21" "Davis Studies - Using mass to calculate the constant for the dose " -271.857 9.198 7 0.2382 O.OOOOE+00 O.OOOOE+00
O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 5.9335E-02 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 7.7491E-02 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
O.OOOOE+00 0.3107 O.OOOOE+00
O.OOOOE+00 O.OOOOE+00 0.5525 O.OOOOE+00 -7.1447E-05 2.1965E-02 2.1994E-02
"07/11/1994" "09:48:26" "PS (smooth) L: <5, 5-10, 10-20, 20-40, >40; W: <.15, .15-.3, .3-1, 1-5, >5 " -274.561 15.30 8 5.2770E-02 O.OOOOE+00 O.OOOOE+00
O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 0.2206 O.OOOOE+00 O.OOOOE+00 1.5167E-03 0.1926 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
C.22
-------
O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
O.OOOOE+00 O.OOOOE+00 0.5853
O.OOOOE+00 -2.9093E-07 2.3400E-02 1.7669E-02
"09/06/1994" "14:33:11" " SC - Length: <5, 5-10, 10-20 20-40, >40; FSB, CSM
O.OOOOE+00 O.OOOOE+00 5.2769E-02 2.5228E-02 O.OOOOE+00 O.OOOOE+00 0.8311
"09/06/1994" "14:44:46" " PS - Length: <5, 5-10, 10-20 20-40, >40; FSB, CSM
O.OOOOE+00 O.OOOOE+00 5.4922E-02 1.4464E-02 O.OOOOE+00 O.OOOOE+00 0.7031
" -275.569
9.0947E-02 2.2160E-02 1.5963E-02
" -276.055
0.2175 2.2220E-02 1.4475E-02
"09/08/1994" "09:59:21" "(Indirect) PS - Length: <5, 5-10,
O.OOOOE+00 1.9526E-02 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
10-20 20-40, >40; FSB, CSM
O.OOOOE+00 O.OOOOE+00 0.9805
"09/08/1994" "10:40:42" "(Indirect) PS - Length: <5, 5-10, 10-20 20-40, >40; FSB, CSM
O.OOOOE+00 2.0334E-02 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 0.9797
"09/08/1994" "10:40:54" "(Indirect) SC - Length: <5, 5-10, 10-20 20-40, >40; FSB, CSM
" -281.736
2.7743E-02 0.1171
" -281.698
2.7748E-02 0.1125
" -301.014
1.1306E-03 0.6515
O.OOOOE+00 0.3474
O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 S.7921E-02 2.4401E-02
"09/09/1994" "15:25:33" "(Indirect) SC - Length: <5, 5-10, 10-20 20-40, >40; FSB (wC.3), CSM " -301.014
1.1306E-03 0.6515 O.OOOOE+00 0.3474 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 S.7921E-02 2.4401E-02
"09/09/1994" "15:25:54" "(Indirect) PS - Length: <5, 5-10, 10-20 20-40, >40; FSB (w<.3), CSM " -281.698
O.OOOOE+00 2.0334E-02 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 0.9797 2.7748E-02 0.1125
"09/09/1994" "15:28:54" " (Direct) PS - Length: <5, 5-10, 10-20 20-40, >40; FSB (w<.3), CSM
O.OOOOE+00 O.OOOOE+00 2.0403E-02 9.0581E-04 O.OOOOE+00 4.6108E-03 0.9375 3.6547E-02 2.2579E-02
"09/09/1994" "15:29:19" " (Direct) SC - Length: <5, 5-10, 10-20 20-40, >40; FSB (w<.3), CSM
1.7669E-03 O.OOOOE+00 3.8221E-03 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 0.9790 1.5448E-02 2.2709E-02
"09/15/1994"
1.6818E-04 5,
"09/15/1994"
O.OOOOE+00 2.
"09/15/1994"
1.3661E-03 0.
"09/15/1994"
8.6262E-04 0.
"09/29/1994"
O.OOOOE+00 0.
"09/29/1994"
O.OOOOE+00 0.
"09/29/1994"
O.OOOOE+00 0.
"09/29/1994"
O.OOOOE+00 0.
"15:28:41"
2736E-02
"15:29:00"
0334E-02
"15:29:18"
OOOOE+00
"15:29:34"
6668
"15:14:20"
1713
"15:14:30"
1713
"15:14:38"
OOOOE+00
"15:14:46"
OOOOE+00
1 "(Indirect) PS - Length: <5, 5-10,
O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
1 "(Indirect) PS - Length: <5, 5-10,
O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
"(Indirect) SC - Length: <5, 5-10,
O.OOOOE+00 0.9985 O.OOOOE+00
"(Indirect) SC - Length: <5, 5-10,
O.OOOOE+00 0.3324 O.OOOOE+00
10-20 20-40, >40; width: <.3, >= .3
O.OOOOE+00 O.OOOOE+00 0.9471 2.6545E-02
10-20 20-40, >40; FSB (wC.3), CSM (w>=.3)
O.OOOOE+00 O.OOOOE+00 0.9797 2.7748E-02
10-20 20-40, >40; width: <.3, >= .3
O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 8.5000E-02
10-20 20-40, >40; FSB (wC.3), CSM (w>=.3)
O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 8.1093E-02
'(Indirect) PS - Length: <5, 5-10, 10-20 20-40, >40; width: <-3, >= 1
O.OOOOE+00 0.1244
O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 0.7044
"(Indirect) PS - Length: <5, 5-10, 10-20 20-40, >40; width: <1, >= 1
O.OOOOE+00 0.1244
O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 0.7044
2.8964E-02
2.8964E-02
'(Indirect) SC - Length: <5, 5-10, 10-20 20-40, >40; width: <-3, >= 1
O.OOOOE+00 0.9859
1.1353E-02 O.OOOOE+00 O.OOOOE+00 -6.8408E-09 2.5704E-02
'(Indirect) SC - Length: <5, 5-10, 10-20 20-40, >40; width: <1, >= 1
O.OOOOE+00 0.9855
1.2036E-02 O.OOOOE+00 O.OOOOE+00 7.0771E-10 2.5422E-02
-277.173
0.1224
-272.162
0.1088
-282.522
0.1388
-281.698
0.1125
-312.657
1.2810E-02
-304.851
2.5418E-02
-278.476
1.1187E-02
-278.476
1.1187E-02
-273.829
1.2676E-02
-273.637
1.2459E-02
16.15 9 5.3160E-02 O.OOOOE+00 O.OOOOE+00
16.35 9 5.9294E-02 O.OOOOE+00 O.OOOOE+00
33.23 11 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
33.16 11 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
90.67 9 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
90.67 9 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
33.16 11 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
17.54 8 2.4088E-02 O.OOOOE+00 O.OOOOE+00
9.944 9 0.3545 O.OOOOE+00 O.OOOOE+00
33.70 10 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
33.16 11 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
116.4 9 O.OOOOE+00 1.1927E-04 O.OOOOE+00
103.5 8 O.OOOOE+00 2.1660E-07 O.OOOOE+00
26.59 10 2.1803E-03 O.OOOOE+00 O.OOOOE+00
26.59 10 2.1803E-03 O.OOOOE+00 O.OOOOE+00
15.15 8 5.5433E-02 5.3701E-05 2.6475E-03
14.50 8 5.8763E-02 5.1098E-05 2.3654E-03
C.23
-------
0
0
0
0
"10/08/1994
"
.9525 -4.
"10/08/1994
.9933
"10/08/1994
.3818
"10/08/1994
.9539
11
1 (
"
0.
"
1.
"10/08/1994"
0
0
0
0
0
.7112
"10/08/1994
.9238
"10/08/1994
0
"
2
"
.9586 -2
"10/08/1994
.9062
"10/08/1994
.1544
tt
8.
"
0,
"10/08/1994"
5
0
0
0
0
0
0
0
0
0
5
.6862E-02
"10/08/1994
3
"
.9525 -4
"10/08/1994
.9933
"10/08/1994
.3818
"10/08/1994
.9539
"10/08/1994
.7112
"10/08/1994
.9238
"10/08/1994
"
1
"
0.
"
1
"
0
"
2
"
.9586 -2
"10/08/1994
.9062
"10/08/1994
.1544
"10/08/1994
.6862E-02
"
8
"
0
"
3
"13:35:28"
. 4052E-13
"13:35:30"
. 9299E-13
"13:35:33"
.2283
"13:35:35"
.1028E-13
"13:35:38"
.1171
"13:35:41"
.0145E-02
"13:35:43"
.2348E-10
"13:35:46"
.4893E-02
"13:35:48"
.1680
"13:35:51"
.1167E-02
"15:07:24"
.4052E-13
"15:07:27"
.9299E-13
"15:07:29"
.2283
"15:07:32"
.1028E-13
"15:07:35"
.1171
"15:07:38"
.0145E-02
"15:07:41"
.2348E-10
"15:07:44"
.4893E-02
"15:07:47"
.1680
"15:07:49"
. 1167E-02
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
" (Direct)
.2537E-02
" (Direct)
.2452E-02
" (Direct)
.4778E-02
" (Direct)
.1312E-02
" (Direct)
.2471E-02
" (Direct)
.1867E-02
" (Direct)
.0451E-02
" (Direct)
.3156E-02
" (Direct)
" (Direct)
" (Direct)
.2537E-02
" (Direct)
.2452E-02
" (Direct)
.4778E-02
" (Direct)
.1312E-02
" (Direct)
.2471E-02
" (Direct)
.1867E-02
" (Direct)
.0451E-02
" (Direct)
.3156E-02
" (Direct)
" (Direct)
PS
4.
SC
2.
PS
1.
SC
2.
PS
2.
SC
1.
PS
3.
SC
1.
PS
SC
PS
4.
SC
2.
PS
1.
SC
2.
PS
2.
SC
1.
PS
3.
SC
1.
PS
SC
- Length:
9013E-02
- Length:
6220E-02
- Length:
6989E-02
- Length:
0133E-02
- Length:
1357E-02
- Length:
4365E-02
- Length:
4136E-02
- Length:
5083E-02
- Length:
- Length:
- Length:
9013E-02
- Length:
6220E-02
- Length:
6989E-02
- Length:
0133E-02
- Length:
1357E-02
- Length:
4365E-02
- Length:
4136E-02
- Length:
5083E-02
- Length:
- Length:
5-35,
5-35,
10-35
10-35
10-30
10-30
5-30,
5-30,
>40;
>35;
>35;
, >35;
, >35;
, >30;
, >30;
>30;
>30;
width:
>40; width:
5-35,
5-35,
10-35
10-35
10-30
10-30
5-30,
5-30,
>40;
>40;
>35;
>35;
, >35;
, >35;
, >30;
, >30;
>30;
>30;
width:
width:
width: <.4, >=.4
width: <.4, >=.4
width: <.4, >=.4
width: <-4, >=.4
width: <-4, >=.4
width: <.4, >=.4
width: <.4, >=.4
width: <.4, >=.4
<-4, >=.4
<.4, >=.4
width: <.4, >-.4
width: <.4, >=.4
; width: <.4, >=.4
: width: <-4, >=.4
: width: <.4, >=.4
: width: <.4, >=.4
width: <-4, >=.4
width: <.4, >=.4
: <.4, >=.4
"10/10/1994" "08:00:28" " (Direct) PS - Length: 5-35, >35; width: <.4, >=.4
•213.
•274.
•278.
•274.
•278.
•275.
•273.
•277.
•331.
•300.
•273.
•274.
•278.
•274.
-278,
•275,
•273,
-277.
•331,
-300,
526
549
769
,577
.770
,485
.993
,242
.352
.376
.526
.549
.769
,577
.770
.485
.993
.242
.352
.376
12.
13.
21.
13.
21.
15.
12.
19.
132
70.
12.
13.
21.
13.
21.
15.
12.
19.
132
70.
55
85
65
29
61
94
89
78
.3
55
55
85
65
29
61
94
89
78
.3
55
9
10
9
9
9
9
9
10
11
11
9
10
9
9
9
9
9
10
11
11
0.
0.
9.
0.
9.
5.
0,
3.
0.
0.
0.
0.
9.
0.
9.
5 ,
0.
3.
0,
0
1832
1792
.3099E-03
.1492
.4643E-03
.7477E-02
.1669
. 0632E-02
.OOOOE+00
.OOOOE+00
.1832
.1792
.3099E-03
.1492
.4643E-03
.7477E-02
.1669
.0632E-02
.OOOOE+00
.OOOOE+00
3.
6.
0.
2.
0.
3.
1.
8.
1
0.
3.
6.
0.
2.
0.
3.
1.
8.
1
0.
6855E-02
7383E-03
3404
7992E-02
1124
2623E-02
9502E-02
8853E-03
.000
7468
6855E-02
7383E-03
3404
7992E-02
1124
2623E-02
9502E-02
8853E-03
.000
7468
1.
0.
4 ,
1.
5.
2,
2.
0.
3.
0.
1.
0.
4.
1.
5.
2,
2.
0.
3,
0.
. 0638E-02
.OOOOE+00
. 9499E-02
.8079E-02
. 9273E-02
.3444E-02
.1946E-02
.OOOOE+00
.8068E-14
.2532
.0638E-02
.OOOOE+00
. 9499E-02
.8079E-02
. 9273E-02
.3444E-02
.1946E-02
.OOOOE+00
. 8068E-14
.2532
" -273.526
12.55
9 0.1832
3.6855E-02 1.0638E-02
C.24
-------
0.9525 1.
"10/10/1994"
0.9933 1.
"10/10/1994"
0.3817 0.
"10/10/1994"
0.9539 1.
"10/10/1994"
0.7113 0.
"10/10/1994"
0.9238 2.
"10/10/1994"
0.9586 2.
"10/10/1994"
0.9062 8.
"10/10/1994"
0.1544 0.
"10/10/1994"
5.6857E-02 3.
"10/18/1994"
0.9963 1.
"10/18/1994"
0.9684 3.
"11/10/1994"
0.9976 2.
2.
1095E-13 2
"08:00:31"
9284E-13 2
"08:00:33"
2283 2
"08:00:36"
0979E-13 2
"08:00:38"
1171 2
"08:00:41"
0172E-02
"08:00:44"
7876E-13 2
"08:00:47"
4873E-02 2
"08:00:49"
1680
"08:00:52"
1165E-02
"10:52:04"
2560E-03 2
"10:52:22"
0243E-02 2
"13:06:25"
7465E-02 4
2538E-02 4.9012E-02
(Direct) SC - Length: 5-35, >35; width: <.4, >=.4
2454E-02 2.6220E-02
(Direct) PS - Length: 10-35, >35; width: <.4, >=.4
4778E-02 1.6987E-02
(Direct) SC - Length: 10-35, >35; width: <.4, > = .4
1314E-02 2.0134E-02
(Direct) PS - Length: 10-30, >30; width: <-4, >=.4
2472E-02 2.1356E-02
(Direct) SC - Length: 10-30, >30; width: <.4, >=.4
1869E-02 1.4365E-02
(Direct) PS - Length: 5-30, >30; width: <-4, >=.4
0451E-02 3.4131E-02
(Direct) SC - Length: 5-30, >30; width: <.4, >=.4
3157E-02 1.5084E-02
(Direct) PS - Length: >40; width: <.4, >=.4
(Direct) SC - Length: >40; width: <.4, >=.4
' (Direct)
7391E-02
' (Direct)
5583E-02
' (Direct)
7575E-02
SC - Length: 5-40, >40; width: <.4, >=.4 (not adjusted)
4.7399E-02
SC - Length: 5-40, >40; width: <.3, >=.3 (not adjusted)
8.5973E-02
SC - L: 5, W: <.4; L:5, W: >=.4; L: >40, w: <.4 (not adjusted)
length-width category followed by 2 equation coefficients
"10/22/1996" "19:32:55" "PS PCM lengths <10, >=10
3.4161e-02 3.5653e-04
"10/22/1996" "19:32:56" "PS PCM lengths >=10
3.5653e-04
"10/22/1996" "19:32:56" "PS PCM lengths >=10 and widths < 0.3
2.6484e-03
"10/22/1996" "19:32:56" "PS PCM lengths >=10 and widths < 0.4
1.9852e-03
"10/22/1996" "19:32:57" "PS PCM lengths >=10 and widths < 0.5
1.4187e-03
"10/22/1996" "19:32:57" "PS PCM lengths >=10 and widths >=0.3
3.8243e-04
"10/22/1996" "19:32:57" "PS PCM lengths >=10 and widths >=0.4
4.0886e-04
-274
-278
-274
-278
-275
-273
-277
-331
-300
-275
-275
-275
Log-
-296
-296
-325
-312
-306
-297
-298
.549
.768
.577
.771
.485
.992
.241
.357
.376
.345
.510
.346
Like
.320
.320
.595
.334
.968
.031
.067
13.
21.
13.
21.
15.
12.
19.
132
70.
17.
17.
17.
Chi
69.
69.
129
110
96.
70.
72.
85
65
29
61
94
89
77
.3
55
84
22
86
-S
01
01
.6
.2
93
29
82
10
9
9
9
9
9
10
11
11
10
10
11
DF
12
12
12
12
12
12
12
0
9
0
9
5
0
3
0
0
5
S
8
.1793
.3202E-03
.1493
. 4607E-03
. 7504E-02
.1670
.0658E-02
.OOOOE+00
.OOOOE+00
.6943E-02
.8761E-02
.4167E-02
p-value
0
0
0
0
0
0
0
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
6.7384E-03
0.3404
2.7991E-02
0.1123
3.2620E-02
1.9478E-02
8.8850E-03
1.000
0.7468
2.4333E-03
1.3829E-03
2.4379E-03
coefficients
0. OOOOE+00
1.000
1.000
1.000
1.000
1.000
1.000
0
4
1
5
2
2
0
3
0
0
0
0
3
0
5
5
3
3
.OOOOE+00
.9500E-02
.8080E-02
.9294E-02
.3446E-02
.1954E-02
.OOOOE+00
.8065E-14
.2532
.OOOOE+00
.OOOOE+00
.OOOOE+00
for each
1.000
.4161e-02
.1064
.6970e-02
.5436e-02
.50806-02
.6791e-02
C.25
-------
"10/22/1996"
4.4693e-04
"10/22/1996"
0.6171 0.
"10/22/1996"
0.6218 0.
"10/22/1996"
0.5800 0.
"10/22/1996"
O.OOOOE+00 0.
"10/22/1996"
O.OOOOE+00 0.
"10/22/1996"
O.OOOOE+00 0.
•19:32:57" "PS PCM lengths >=10 and widths >=0.5
"19:32:58'
3829
"19:32:58'
3782
"19:32:58'
4200
"19:32:59'
OOOOE+00
"19:32:59'
OOOOE+00
"19:32:59'
OOOOE+00
"10/22/1996
3.0399e-02
"10/22/1996
3.8864e-04
"10/22/1996
1.2887e-03
"10/22/1996
8.0126e-04
"10/22/1996
5.9682e-04
"10/22/1996
5.3901e-04
"10/22/1996
7.0335e-04
"10/22/1996
1.0040e-03
"10/22/1996
O.OOOOE+00
"10/22/1996
0.1906
"10/22/1996
6.8701e-02
"10/22/1996
O.OOOOE+00
"10/22/1996
O.OOOOE+00
"10/22/1996
" "19:33:00"
3.8864e-04
19:33:01"
" "19:33:01"
19:33:01"
" "19:33:02"
"19:33:02"
"19:33:02"
" "19:33:03"
" "19:33:03"
1.000
" "19:33:03
0.8094
19:33:03
0.9313
" "19:33:04
O.OOOOE+00
19:33:04
O.OOOOE+00
" "19:33:05
"PS PCM lengths <10, >=10 and widths <0.3, >=0.3
3.4084e-02 8.9173e-04
"PS PCM lengths <10, >=10 and widths <0.4, >=0.4
3.3966e-02 8.64546-04
"PS PCM lengths <10, >=10 and widths <0.5, >=0.5
3.4167e-02 7.8386e-04
"PS PCM lengths <5, 5-10, >=10 and widths <0.3, >=0.3
0.6171 0.3829 3.4084e-02 8.9173e-04
"PS PCM lengths <5, 5-10, >=10 and widths <0.4, >=0.4
0.6218 0.3782 3.3966e-02 8.6454e-04
"PS PCM lengths <5, 5-10, >=10 and widths <0.5, >=0.5
0.5800 0.4200 3.4167e-02 7.8386e-04
"SC PCM lengths <10, >=10
"SC PCM lengths >-10
"SC PCM lengths >=10 and widths < 0.3
"SC PCM lengths >=10 and widths < 0.4
"SC PCM lengths >=10 and widths < 0.5
"SC PCM lengths >=10 and widths >=0.3
"SC PCM lengths >=10 and widths >=0.4
"SC PCM lengths >-10 and widths >=0.5
• "SC PCM lengths <10, >-10 and widths <0.3, >=0.3
3.0016e-02 5.3901e-04
* "SC PCM lengths <10, >=10 and widths <0.4, >-0.4
2.9857e-02 7.3280e-04
' "SC PCM lengths <10, >=10 and widths <0.5, >=0.5
2.8234e-02 9.6970e-04
' "SC PCM lengths <5, 5-10, >=10 and widths <0.3, >-0.3.
O.OOOOE+00 1.000 3.00166-02 5.3901e-04
1 "SC PCM lengths <5, 5-10, >=10 and widths <0.4, >=0.4
0.1906 0.8094 2.9857e-02 7.3280e-04
• "SC PCM lengths <5, 5-10, >=10 and widths <0.5, >=0.5
298.
•296.
•296.
•296.
•296.
•296.
•296
•292
•292
•301
•300
•299
•291
•291
•289
•291
•291
•289
•291
•291
•289
.522
.228
.124
.226
.228
.124
.226
.283
.283
.778
.564
.230
.282
.615
.538
.282
.297
.410
.282
.297
.410
73.
68.
68.
68.
68.
68.
68.
56.
56,
90,
87.
80.
53.
53.
47.
53,
53,
47.
53.
53.
47.
,82
.97
,75
95
97
,75
.95
,96
,96
.24
,30
,93
,08
,13
.01
,08
.07
.20
.08
.07
,20
12
11
11
11
11
11
11
12
12
12
12
12
12
12
12
12
11
11
12
11
11
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
OOOOE+00
OOOOE+00
, OOOOE+00
OOOOE+00
OOOOE+00
OOOOE+00
.OOOOE+00
, OOOOE+00
, OOOOE+00
, OOOOE+00
OOOOE+00
, OOOOE+00
, OOOOE+00
.OOOOE+00
.OOOOE+00
, OOOOE+00
.OOOOE+00
, OOOOE+00
, OOOOE+00
.OOOOE+00
, OOOOE+00
1.000
O.OOOOE+00
O.OOOOE+00
O.OOOOE+00
O.OOOOE+00
O.OOOOE+00
O.OOOOE+00
O.OOOOE+00
1.000
1.000
1.000
1.000
1.000
1.000
1.000
O.OOOOE+00
O.OOOOE+00
O.OOOOE+00
O.OOOOE+00
O.OOOOE+00
O.OOOOE+00
3
0
0
0
0
0
0
3
4
4
4
3
3
2
0
0
0
0
0
0
.7227e-02
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
1.000
.03996-02
.35486-02
.28306-02
.08936-02
.0016e-02
.06016-02
.8679e-02
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
C.26
-------
O.OOOOE+00 O.OOOOE+00
"10/28/1996"
2.5235e-02 5.
"10/28/1996"
5.1220e-04
"10/28/1996"
5.6629e-03
"10/28/1996"
3.4464e-03
"10/28/1996"
2.3840e-03
"10/28/1996"
5.4368e-04
"10/28/1996"
5.8117e-04
"10/28/1996"
6.2912e-04
"10/28/1996"
0.7938 0.
"10/28/1996"
0.7526 0.
"10/28/1996"
0.7528 0.
"10/28/1996"
O.OOOOE+00 0.
"10/28/1996"
O.OOOOE+00 0.
"10/28/1996"
O.OOOOE+00 0.
"10/28/1996"
2.3313e-02 5.
"10/28/1996"
5.5129e-04
"10/28/1996"
1.9603e-03
"10/28/1996"
1.2027e-03
"10/28/1996"
9.0725e-04
"10/28/1996"
"12:44:57"
12206-04
"12:44:57"
"12:44:58"
"12:44:58"
"12:44:58"
"12:44:58"
"12:44:59"
"12:44:59"
"12:44:59"
2062
"12:45:00"
2474
"12:45:00"
2472
"12:45:01"
OOOOE+00
"12:45:01"
OOOOE+00
"12:45:01"
OOOOE+00
"12:45:03"
51296-04
"12:45:03"
"12:45:03"
"12:45:04"
"12:45:04"
"12:45:04"
S.87016-02 0.9313 2.8234e-02 9.6970e-04
• "DPS PCM lengths <10, >=10
1 "DPS PCM lengths >=10
1 "DPS PCM lengths >=10 and widths < 0.3
1 "DPS PCM lengths >=10 and widths < 0.4
' "DPS PCM lengths >=10 and widths < 0.5
1 "DPS PCM lengths >=10 and widths >=0.3
' "DPS PCM lengths >=10 and widths >=0.4
1 "DPS PCM lengths >=10 and widths >=0.5
1 "DPS PCM lengths <10, >=10 and widths <0.3, >=0.3
2.46316-02 2.1051e-03
1 "DPS PCM lengths <10, >=10 and widths <0.4, >=0.4
2.4604e-02 1.6363e-03
1 "DPS PCM lengths <10, >=10 and widths <0.5, >=0.5
2.49066-02 1.46836-03
1 "DPS PCM lengths <5, 5-10, >=10 and widths <0.3, >=0.3
0.7938 0.2062 2.4631e-02 2.1051e-03
1 "DPS PCM lengths <5, 5-10, >=10 and widths <0.4, >=0.4
0.7526 0.2474 2.46046-02 1.63636-03
' "DPS PCM lengths <5, 5-10, >=10 and widths <0.5, >=0.5
0.7528 0.2472 2.4906e-02 1.4683e-03
"DSC PCM lengths <10, >=10
"DSC PCM lengths >=10
"DSC PCM lengths >=10 and widths < 0.3
"DSC PCM lengths >=10 and widths < 0.4
"DSC PCM lengths >=10 and widths < 0.5
"DSC PCM lengths >=10 and widths >=0.3
287.
287.
•316.
300.
•293.
•289.
291.
292.
286.
286.
286.
286.
286.
286.
283.
283.
292.
291,
288.
283.
.949
.949
.503
.922
.750
.658
.088
.360
.566
.586
.449
.566
.586
.449
.543
.543
.585
.994
.758
.168
44.
44.
110
82.
61.
48.
51.
54.
41.
41.
41.
41.
41.
41.
33.
33.
62.
61.
50.
31.
46
46
.2
23
78
05
44
33
86
99
81
86
99
81
35
35
75
26
76
45
12
12
12
12
12
12
12
12
11
11
11
11
11
11
12
12
12
12
12
12
0
0
0
0
0
0
0,
0.
0.
0.
0,
0.
0.
0.
1.
1.
0.
0.
0.
8,
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
,1074e-05
,1074e-05
.OOOOE+00
.OOOOE+00
.OOOOE+00
.3897e-04
O.OOOOE+00
1.000
1.000
1.000
1.000
1.000
1.000
1.000
O.OOOOE+00
O.OOOOE+00
O.OOOOE+00
O.OOOOE+00
O.OOOOE+00
O.OOOOE+00
O.OOOOE+00
1.000
1.000
1.000
1.000
1.000
2
1
4
3
2
2
2
0
0
0
0
0
0
2
2
2
2
2
1.000
.52356-02
.3622e-02
.41476-02
.5229e-02
.64656-02
.7837e-02
.86906-02
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
.OOOOE+00
1.000
.3313e-02
.93506-02
.9669e-02
.7648e-02
.39066-02
C.27
-------
7.5397e-04
"10/28/1996"
9.7866e-04
"10/28/1996"
1.3372e-03
"10/28/1996"
0.1662 0.
"10/28/1996"
0.2851 0.
"10/28/1996"
0.3050 0.
"10/28/1996"
O.OOOOE+00 0.
"10/28/1996"
O.OOOOE+00 0.
"10/28/1996"
O.OOOOE+00 0.
"12:45:05" "DSC PCM
"12:45:05" "DSC PCM
lengths >=10 and widths >=0.4
lengths >=10 and widths >=0.5
"12:45:05" "DSC PCM lengths <10, >=10 and widths <0.3, >=0.3
3338 2.3616e-02 8.4322e-04
"12:45:06" "DSC PCM lengths <10, >-10 and widths <0.4, >=0.4
7149 2.3447e-02 1.0489e-03
"12:45:06" "DSC PCM lengths <10, > = 10 and widths <0.5, >=0.5
6950 2.3103e-02 1.1895e-03
"12:45:06" "DSC PCM lengths <5, 5-10, >=10 and widths <0.3, >=0.3
OOOOE+00 0.1662 0.8338 2.3616e-02 8.4322e-04
"12:45:07" "DSC PCM lengths <5, 5-10, >=10 and widths <0.4, >=0.4
OOOOE+00 0.2851 0.7149 2.3447e-02 1.0489e-03
"12:45:07" "DSC PCM lengths <5, 5-10, >=10 and widths <0.5, >=0.5
OOOOE+00 0.3050 0.6950 2.3103e-02 1.1895e-03
-283.756 32.11 12 4.9080e-04 1.000 2.4780e-02
-284.839 33.82 12 O.OOOOE+00 1.000 2.5206e-02
-283.115 31.57 11 1.42706-04 O.OOOOE+00 O.OOOOE+00
-283.015 31.25 11 2.5114e-04 O.OOOOE+00 O.OOOOE+00
-283.016 31.11 11 3.0503e-04 O.OOOOE+00 O.OOOOE+00
-283.115 31.57 11 1.4270e-04 O.OOOOE+00 O.OOOOE+00
-283.015 31.25 11 2.5114e-04 O.OOOOE+00 O.OOOOE+00
-283.016 31.11 11 3.0503e-04 O.OOOOE+00 O.OOOOE+00
C.28
-------
APPENDIX D:
THE VARIATION IN KL VALUES DERIVED FOR
CHRYSOTILE MINERS AND CHRYSOTILE TEXTILE
WORKERS
The difference between the observed risk of lung cancer for comparable levels of chrysotile
exposure among Quebec miners (most recent followup: Liddell et al. 1997) and South Carolina
textile workers (Dement et al. 1994; McDonald et al. 1983a) has been the focus of much
attention. Reasonably good agreement between results from the Quebec studies and another
study of chrysotile miners in Italy (Piolatto et al. 1990) coupled with reasonably good agreement
between results from the South Carolina Plant and results from textile plants in Mannheim,
Pennsylvania (McDonald et al. 1983b) and in Roachdale, England (Peto 1980a,b; Peto et al.
1985) suggest that the difference between Quebec and South Carolina may reflect a general
difference between the two industries (see Table 7-6 and Section 7.2.2). This appears true
despite the fact, for example, that cohorts at two of the textile plants were apparently exposed to
significant amounts of amphibole in addition to chrysotile (see Appendix A and Section 7.2.2).
Three main hypotheses have been advanced to explain the difference in the risk per unit
exposure observed among miners and textile workers (see, for example, Sebastien et al. 1989).
These are:
(1) the low reliability of exposure estimates in the various studies;
(2) differences in fiber size distributions in the two industries (with textile-related
exposures presumably involving greater fractions of longer fibers); or
(3) simultaneous exposure to a co-carcinogen (i.e., oil that may have been sprayed on
the asbestos fibers) in the textile industry.
It has also been proposed that differences in the concentration of long tremolite (amphibole)
fibers in dusts from each of the two industries might represent an explanatory factor (see, for
example, McDonald 1998b). However, this would also require a large relative difference
between the potencies of tremolite (amphiboles) and chrysotile toward the induction of lung
cancer. This latter issue is addressed further in Sections 7.4-7.6. McDonald (1998b) also
presents an overview of the current status of each of the hypotheses described above.
In an attempt to distinguish among the above-listed hypotheses, Sebastien et al. (1989)
conducted a study to determine lung fiber concentrations in tissue samples from deceased
members of the cohorts studied from both the Quebec mines (specifically, from the Thetford
mine) and the South Carolina textile plant. These researchers ultimately analyzed tissue samples
from 72 members of the South Carolina cohort and 89 members of the Thetford (Quebec) cohort.
Because the tissue samples came from cohort members, they could be matched with estimates of
the exposure experienced by each of the individuals as well as details concerning the age at first
employment, the age at death, the years of employment, and the number of years following
employment until death.
D-1
-------
In the Sebastien et al. (1989) study, tissue samples were obtained in formalin-fixed or paraffin
blocks, which were then digested in bleach, filtered, and analyzed by TEM. Tissue samples
were apparently "opportunistic." Only fibers longer than 5 urn with an aspect ratio >3:1 were
included in the count. For consideration of the limitations associated with such preparations, see
Section 5.2.
Results from matching of tissue samples with the histories of corresponding cohort members
indicate that tissue samples obtained from each cohort covered a broad range of exposure levels,
duration of exposure, and years since the end of exposure. They also indicate that South
Carolina cohort members included in the Sebastien et al. (1989) study experienced, on average,
13.5 years of exposure with 18.1 years between the end of exposure and death. In contrast,
Thetford workers included in this study experienced an average of 32.6 years of exposure with
only 11.6 years between the end of exposure and death. Corresponding to differences in
exposure levels observed across the two cohorts in the original epidemiology studies, mean
exposure levels experienced by Thetford cohort members included in this study were about 10
times mean exposure levels experienced by South Carolina workers (19.5 mpcf vs. 1.9 mpcf).
Because Sebastien and coworkers recognized the general lack of a good model describing the
retention and clearance of asbestos fibers in the lungs at the time their study was conducted, they
performed most of their analyses either on pairs of members (one from each cohort) matched for
duration of exposure and time since end of exposure or on groups of members from each cohort
similarly stratified by duration of exposure and time since end of exposure.
Results from their study indicate that, overall, lung burdens observed among Thetford cohort
members are substantially higher than those observed among South Carolina cohort members.
Geometric mean lung chrysotile concentrations are reported to be 5.3 and 0.63 fibers/ng dry lung
tissue in Thetford workers and South Carolina workers, respectively. Furthermore, despite
tremolite representing only a minor contaminant in the chrysotile from Quebec and the dusts to
which the miners were exposed (Sebastien et al. 1986), the majority of fibers observed in the
lungs of Thetford miners were in fact tremolite (mean concentration 18.4 f/ng dry lung). Since
the raw material used in the South Carolina plant came largely from Quebec, tremolite was also
expected to be a minor contaminant in the dusts to which textile workers were exposed. Yet
among these workers also, tremolite represented a substantial fraction of the lung fibers observed
(mean concentration 0.36 f/u.g). Thus, the ratio of tremolite concentrations observed among
Thetford miners and that observed among South Carolina workers (18.4:0.36, or 51) is even
more extreme than the ratio observed for chrysotile (8.4).
To evaluate the first of the above-listed hypotheses, it is instructive to compare the ratios of
chrysotile or tremolite fibers observed in the lungs of deceased workers from Thetford and South
Carolina, respectively, with the overall exposures that each received. A rough estimate of
cumulative exposure for each set of workers in the Sebastien et al. (1989) study representing
each cohort can be derived as the product of the mean duration of exposure and the mean
intensity of exposure. Thus, for example, mean cumulative exposure in Thetford was 32.6
yearsx!9.5 mpcf or 635.7 mpcf-yrs. Similarly, for South Carolina, mean cumulative exposure
was 25.65 mpcf-yrs, which gives a Thetford/South Carolina ratio of 24.8. This presumably
represents the relative cumulative exposure to chrysotile. For tremolite, Sebastien and
coworkers report that, based on a regression analysis, the fraction of tremolite fibers among total
D-2
-------
asbestos fibers were likely only 0.4 times as much in South Carolina as in Thetford (where they
likely averaged 1% of total fibers). Therefore, the ratio of cumulative exposures to tremolite for
the sets of cohort members studied by Sebastien and coworkers is likely 62.
Comparing the ratio of Thetford: South Carolina lung burden estimates with the ratios of the
corresponding cumulative exposures, it appears that the chrysotile lung burden ratio (8.4) is only
a third of the ratio predicted based on cumulative exposure (24.8). However, the ratio of lung
tremolite concentrations (51) is much closer to the corresponding cumulative exposure ratio (62).
It thus appears that, although, airborne concentrations may not closely track the exposures that
led to the observed lung burdens for individuals (see below), the overall trend in exposures
predicted by airborne measurements is approximately correct. It is therefore likely that overall
exposure concentrations in Thetford were in fact substantially higher than in South Carolina (in
agreement with airborne measurements). Thus, we concur with Sebastien et al. that the
unreliability of exposure estimates in these two cohorts is unlikely to explain the observed
difference in the risk per unit of exposure observed for each cohort.
Importantly, although the general trend in relative overall exposure levels predicted by airborne
measurements between Thetford and South Carolina appear to have been confirmed by mean
lung fiber concentrations in the Sebastien et al. (1989) study, the estimated exposures correlate
poorly with lung burdens for any particular individual. To demonstrate this, we analyzed the
Thetford: South Carolina ratios of lung chrysotile concentrations and, separately, lung tremolite
concentrations reported by Sebastien et al. for their set of 32 matched pairs of cohort workers to
determine whether trends in these ratios adequately matched trends in the corresponding
estimated airborne exposure level ratios for the same matched pairs. To do this, we subjected the
ratios presented in Table 7 of the Sebastien et al. (1989) study to a Rank Von Neuman test
(Gilbert 1987). Results indicate that trends in neither lung chrysotile concentration ratios nor
lung tremolite concentration ratios can be predicted by the observed trend in the estimated
airborne concentration ratios among these 32 matched pairs.
There are numerous sources of potential uncertainty that may mask the relationship between
airborne exposure estimates and resulting lung burdens (Section 5.2). Potentially the largest of
these is the variation expected among lung burden estimates derived from use of "opportunistic"
tissue samples, which are not controlled for the portion of the respiratory tree represented by the
sample. Even for samples collected from adjacent locations in lung parenchyma, observed fiber
concentrations may vary substantially and such variation is magnified between samples taken
from different individuals at locations in the lung that may not in any way correspond to their
relative position in the respiratory tree.
Other potentially important sources of variation that may mask the relationship between airborne
exposure concentrations and resulting lung burden estimates may primarily involve limitations in
the degree to which the airborne estimates from an epidemiology study represent actual
exposures to the individual members of a study cohort (Section 5.1). The following factors may
all contribute to the uncertainty of exposure estimates:potential differences between individual
exposures versus area concentrations (which are what is typically measured), the adequacy of
extrapolation to the earliest exposures in a cohort (when measurements were generally not
available), or the adequacy of estimating job x time matrices for individual workers that can then
be integrated with work area exposure estimates to derive individual exposure estimates.
D-3
-------
The second of the above-listed hypotheses, involves potential differences in the size of structures
that may have been present in the airborne concentrations in Thetford and South Carolina, which
may not have been adequately represented by the exposure measurements. More generally, this
is a question of the degree to which measured exposures in the two environments adequately
reflect potential differences in the character of exposure that relate to biological activity.
Sebastien et al. (1989) considered this second hypothesis by generating and comparing size
distributions for the fibers observed in the lungs of workers from Thetford and, separately, South
Carolina. Importantly, the size distributions for each cohort were generated by including the first
five fibers observed from every member of that cohort, without regard to the duration of
exposure, level of exposure, or time since exposure experienced by each cohort member.
Therefore, the size distributions obtained are "averaged" over very different time frames during
which differing degrees of fiber retention and clearance will have taken place, each of which
potentially alters the distributions of fiber sizes (Section 6.2). Thus, the two distributions
generated are each actually collections of samples from multiple, varied size distributions (rather
than single distributions) and this likely masks distinctions between the two work environments.
It is therefore not surprising that the authors found relatively little differences in the two size
distributions.
The portion of the generated size distributions that are least likely to have been affected by the
limitations due to the manner in which they are generated (as Sebastien et al. suggest) is the
fraction of tremolite (amphibole) fibers longer than 20 \im. This is because (1) tremolite fibers
(unlike chrysotile) are biodurable and (2) biodurable fibers longer than approximately 20 \im
have been shown to clear from the lung only very slowly, if at all (Section 6.2). Thus, the
Thetford: South Carolina ratio of long tremolite fibers may provide the best indication of the
relative exposures to long fibers in the two environments.
Table D-l presents the estimated, relative concentrations of specific lengths of fibers observed in
lung tissue among Thetford miners and South Carolina workers, respectively. The length
category for various fibers is presented in the last column of the table. The estimated
concentrations, presented in Columns 2 (for Thetford) and 3 (for South Carolina) of this table
were derived as follows. For the first length category (L>5 p,m), concentrations are taken
directly from Table 5 of the Sebastien et al. (1989) paper (the geometric means are presented).
Concentrations for the remaining length categories were estimated by multiplying the
concentrations for this first length category by the fraction of the size distribution represented by
each succeeding length category (as provided in Table 4 of the Sebastien et al. paper). So that
the relative precision of these concentration estimates can be evaluated, an estimate of the
numbers of fibers included in each length category (from the total used to derive the size
distribution in Table 4 of Sebastien et al.) are provided in Columns 6 (for Thetford) and 7 (for
South Carolina), respectively. The Thetford: South Carolina ratios of the concentrations of fibers
in each length category (for each fiber type) are provided in Column 5 of the table.
D-4
-------
Table D-l. Estimated Concentrations of Sized Fibers Observed in the Lungs of Thetford
Miners and South Carolina Textile Workers"
MEAN LUNG
CONCENTRATION
NUMBER OF FIBERS
Fiber Type
Chrys
Trem
Chrys
Trem
Chrys
Trem
Chrys
Trem
Thetford
5.3
18.4
1.73
3.90
0.59
0.72
0.16
0.037
South
Carolina
0.63
0.38
0.17
0.091
0.070
0.024
0.031
0.008
Units
f/|j,g lung
f/jig lung
f/jig lung
f/Hg lung
f/jig lung
f/[ig lung
f/jig lung
f/Hg lung
Ratio:
Th/SC
8.41
48.42
10.00
42.95
8.41
30.46
5.15
4.40
Thetford
371
405
121
86
41
16
11
1
South
Carolina
226
175
62
42
25
11
11
4
Size Range of
Fibers6
Length>5 \im
Length>8 (xm
Length>13(j.m
Length>20p.m
"Derived from data presented in Tables 4 and 5 of Sebastien et al. (1989)
'Geometric mean
It is instructive to compare the ratios presented in Table D-l to the Thetford:South Carolina
ratios of mean cumulative exposures estimated above for chrysotile and tremolite among the
cohort members included in the Sebastien et al. (1989) study (24.8 and 62, respectively). As
indicated in Table D-l, for chrysotile, the ratio remains approximately constant at about 9
(varying only between 8.4 and 10) for all of the size ranges reported except the longest. For the
longest category (L>20), however, the ratio drops to 5. Because fibers longer than 20 jim are
expected to be the most persistent in the body (Section 6.2), it may be that the ratio of 5 best
represents the relative concentration of long chrysotile structures among the two sets of cohort
members.
Because this ratio (for the long fibers found in the lung) is only approximately 1/5 of the
estimated ratio for the cumulative exposure to chrysotile (24.8), this suggests that the South
Carolina cohort may indeed have been exposed to dusts enriched in long fibers relative to dusts
experienced at Thetford. Because the estimate of this ratio is based on counts of at least 11
fibers from Thetford and South Carolina, respectively, it is unlikely that this ratio will vary by
more than a factor of 2 or 3 (the 95% CI around 11 fibers, based on a Poisson distribution is
6-19).
The trend with tremolite is even more striking. Moreover, as previously indicated, because
tremolite fibers are biodurable, it is the tremolite fibers longer than 20 (im that may best
represent the ratio of long fibers to which these two groups of cohort members were exposed.
The ratios observed among tremolite fibers steadily decrease from approximately 50 for fibers
longer than 5 ^m to 4.4 for fibers longer than 20 [im, although this last value is uncertain (due to
it being based on only 1 fiber observed among Thetford-derived lungs and only 4 fibers among
South Carolina-derived lungs). In fact these data are statistically consistent even with a ratio
considerably less than 1, (i.e., with a considerably higher concentration of long tremolite fibers
in South Carolina than in Quebec). Given that the ratio of the original cumulative exposures for
tremolite was estimated to be 62, that the ratio of long tremolite fibers is only 4.4 suggests that
dusts in South Carolina may have been highly enriched in long fibers.
D-5
-------
Observations that the fibers to which textile workers were exposed were longer and thinner than
those found in mining are further supported by various published size distributions of fibers
determined in air samples collected in these environments (see, for example, Gibbs and Hwang
1975,1980). Also, as noted in Crump (1986), the raw fiber purchased by textile plants was
commonly described as the longest grade of product (see Table 22 of Crump). Size issues are
addressed further in Section 7.4.
The data in a more recent study by Case et al. (2000) demonstrates even more strongly that
South Carolina textile workers were exposed to fibers that were substantially longer than those
inhaled by Quebec miners and millers. In this study, lung fiber contents were determined for 64
deceased textile workers and 43 deceased chrysotile miners and millers, respectively, which
represent randomly selected subsets of the workers, miners, and millers for whom lung burdens
were previously described by Sebastien et al. (1989), as discussed above.
In the Case et al. (2000) study, analyses were conducted on sets of TEM specimen grids that had
originally been prepared in the Sebastien et al. (1989) study, thus selection of subjects and the
preparation of samples in this study is the same as described above for the Sebastien et al. study.
However, Case et al. focused specifically on the counting of fibers longer than 18 \im.
Results from the Case et al. (2000) study are summarized in Table D-2. As indicated in the
second column of Table D-2, the mean cumulative exposure to which the selected cohort
members from Quebec and South Carolina were exposed in this study was 186 and 3.63 mpcf-y
(millions of particles per cubic ft-years), respectively. This gives a Quebec/South Carolina ratio
of approximately 51. In contrast the Quebec/South Carolina ratios of the concentrations of
asbestos fibers observed in lungs among these selected cohort members are substantially smaller
(4.28 for long chrysotile, 12.04 for long tremolite, and 5.45 for long amphibole). This implies
that the lungs of South Carolina workers are substantially enriched in these long fibers relative to
the lungs of Quebec miners and millers. Moreover, because substantial numbers of long fibers
were counted in these analyses, the uncertainty of these ratios is relatively small.
TABLE D-2. ESTIMATED MEAN AIRBORNE EXPOSURE CONCENTRATIONS
AND ASSOCIATED LUNG FIBER BURDENS FOR A SELECTED SET OF TEXTILE
WORKERS, MINERS, AND MILLERS"
Location
Quebec Mining
SC Textiles
Ratio
Mean Airborne
Exposure
Concentration
(mpcfy)
186
3.63
51.24
Lung
Chrysotile
Content
(long fibers)
(f/|ig)
0.231
0.054
4.28
Lung
Tremolite
Content
(long fibers)
(f/Hg)
0.325
0.027
12.04
Lung Total
Amphibole
Content
(long fibers)
(f/lig)
0.349
0.064
5.45
"Derived from data presented in Table 2 of Case et al. (2000)
D-6
-------
If the estimated KL's derived for Quebec miners (0.00029) and South Carolina textile workers
(0.021), as reported in Table 7-6, are adjusted to account for the relative concentrations of long
fibers reported by Case et al. the disparity in these KL estimates effectively disappears. If
adjusted as described in Section 7.4.2, the new KL«'s for Quebec (0.234) and for South Carolina
(1.21) now differ by only a factor of 5 (rather than the original factor of 72). Thus, accounting
for long structures appears to reconcile these potency estimates.
The data presented by Case et al. (2000) also indicates that the lungs of textile workers in South
Carolina (but not those of Quebec miners) contain substantial concentrations of commercial
amphibole asbestos fibers (amosite and crocidolite) in addition to tremolite. In fact, the majority
of the amphibole fibers observed in lungs from South Carolina workers were composed of the
commercial amphibole types. This suggests, among other things, that the exposure environment
in South Carolina should actually be characterized as a mixed exposure environment rather than
a chrysotile exposure environment. As indicated in the following two paragraphs, however,
conclusions concerning the nature of the general exposure environments in Quebec mines or the
South Carolina textile mill that are based only on observations among the small subsets of these
cohorts examined by Case et al. may not be robust.
Importantly, Case et al. indicate in their paper that, because they observed substantially greater
absolute numbers of long fibers in the lungs of Quebec miners than in the lungs of South
Carolina workers, they conclude that (regardless of the above analysis), Quebec miners were still
exposed to a greater absolute number of long fibers than South Carolina workers. However, this
does not appear to be a valid conclusion that can be derived from the data provided in the paper.
We compared the mean exposure concentrations reported for the subset of Quebec miners and
South Carolina textile workers that Case et al. (2000) examined (Table D-2) to the distribution of
exposures reported among the entire cohorts in Quebec (Table A-2) and South Carolina (Table
A-8), respectively. Results suggest that, exposures for the subset of Quebec cohort members
included in the Case et al. study are higher than approximately 75% of the exposures
experienced by the overall cohort. In contrast, exposures for the subset of the textile worker
cohort examined by Case et al. are lower than approximately 50% of exposures experienced in
the overall cohort. Thus, given that the mean exposures experienced by the subsets of each
cohort examined by Case et al. do not reflect mean exposures for the respective cohorts as a
whole, it is not reasonable to compare absolute numbers of structures observed in the lungs of
these workers and draw general conclusions about the relative, absolute exposures among the
entire, respective cohorts.
At this point it is worth mentioning some of the potential differences in the characteristics of
mining dusts and textile mill dusts that may affect biological activity, but that may not be
adequately delineated when measuring exposures by PCM (in f/ml) and almost certainly not
delineated when exposures are measured by midget impinger (in mpcf), see Section 4.3. During
the mining of asbestos, only a small fraction of the rock (generally no more than 10%) that is
mined is typically composed of the fibers of interest.
While the host rock in a mine may be of similar chemical composition, it generally represents an
entirely different crystalline habit. Nevertheless, a large fraction of the dust that is created
during mining is likely composed of fragments from the host rock and many of these fragments
D-7
-------
will be of a size that would be included in the particles counted by midget impinger.
Furthermore, at least some fraction of the fragments created by the crushing and cutting of the
host rock will be elongated "cleavage" fragments (Section 4.0) so that at least some fraction of
these may be included even in PCM counts, despite many of them being either too thick to be
respirable, or too short or thick to be biologically active (see Section 6.2). Note, although
Sebastien et al. (1989) employed TEM to characterize fibers in the study, they apparently
employed a fiber definition that was sufficiently broad that they too would have counted large
numbers of structures that may be too short or too thick to contribute to biological activity.
In comparison, the dusts created in a textile factory are likely composed almost exclusively of
true asbestos fibers. The raw material received by the factory will already have been milled and
beneficiated to remove the vast majority of non-fibrous material. It is therefore, much less likely
that extraneous fragments (even cleavage fragments) exist that might be counted either by
midget impinger or PCM. We make this point because, if this represents the true situation, it
would be expected that risk per unit exposure estimates (i.e., exposure-response factors) derived
from any mining site, may be smaller than estimates derived for the same fiber type in
occupational environments where only finished fiber is used. Thus, another interpretation of the
variation observed among estimated KL values for amphiboles (reported in Section 7.2.2) is that
mining values are somewhat low. As later described (Section 7.3.2), the same may be true for
amphibole KM values. The implications of this possibility are discussed further in each
respective section.
Note, although the Sebastien et al. (1989) paper suggests that (mpcf) exposure estimates from
Thetford and South Carolina grossly suggest the relative range of lung burdens observed, there is
too much scatter in the data to determine how closely the air ratios track the lung burden ratios.
For example, ratios derived from arithmetic means (rather than the geometric means) for the
Sebastien et al. data are substantially different. Moreover, as indicated above, there may be
substantially different size distributions in the two environments, which might at least in part be
explained by the inclusion of large numbers of cleavage fragments (with dimensions
inappropriate for biological activity) in the mining environment.
Although the third of the above-listed hypotheses was not addressed by Sebastien and
coworkers, the question of whether a co-carcinogen contributes to the overall observed lung
cancer rate among textile workers has been considered by several other researchers. To test the
hypothesis of whether oils potentially contributed to disease in South Carolina, Dement and
Brown (1994) performed a nested case-control study among a subset of the cohort members
previously studied by Dement et al. (most recent update, 1994). In this analysis, Dement and
Brown qualitatively assessed the probability of mineral oil exposure for cases and controls based
on knowledge of historic descriptions of mineral oil use. The extent of such exposure was then
further categorized into three strata: none or little, moderate, or heavy, based on where each
worker was longest employed. Cases and controls were then further categorized based on years
at risk and level of asbestos exposure. Results from this nested analysis indicated no significant
change in the estimated exposure-response slope for asbestos after adjusting for mineral oil
exposure.
Additional, albeit qualitative, evidence that oils may not represent an adequate explanation for
the relative lung cancer risks observed in mining and textiles is provided by McDonald (1998b).
D-8
-------
McDonald suggests that oils were not used in the Roachdale plant until 1974. Therefore, due to
latency, it is unlikely that the use of such oils would have had a substantial impact on the
observed lung cancer cases at the point in time that the study was conducted (Peto 1980a,b; Peto
etal. 1985).
Taken as a whole, the evidence presented in this section suggests that the relative distribution of
fiber sizes found in dusts in the textile industry and the mining industry, respectively, may be the
leading hypothesis for explaining the observed differences in lung cancer risk per unit of
exposure between these two industries.
REFERENCES
Case BW; Dufresne A; McDonald AD; McDonald JC; Sebastien P. Asbestos Fiber Type and
Length in Lungs of Chrysotile Textile and Production Workers: Fibers Longer than 18 \im.
Inhalation Toxicology. l(Suppl 1):411-418. 2000.
Crump KS. Asbestos Potency Assessment for EPA Hearing. Prepared for Asbestos Information
Association/North America. 116pp. 1986.
Dement JM; Brown DP. Lung Cancer Mortality Among Asbestos Textile Workers: A Review
and Update. Annals of Occupational Hygiene. 38(4):525-532. 1994.
Dement JM; Brown DP; Okun A. Follow-up Study of Chrysotile Asbestos Textile Workers:
Cohort Mortality and Case-Control Analysis. American Journal of Industrial Medicine.
26:431^147. 1994.
Gibbs GW; Hwang CY. Physical Parameters of Airborne Asbestos Fibres in Various Work
Environments - Preliminary Findings. American Industrial Hygiene Association Journal.
36(6):459-466. 1975.
Gibbs GW; Hwang CY. Dimensions of Airborne Asbestos Fibers. In Biological Effects of
Mineral Fibers. Wagner JC (ed.). IARC Scientific Publication, pp. 69-78. 1980.
Gilbert 0. Statistical Method for Environmental Pollution Monitoring. Van Nostrand Reinhold,
New York. 1987.
Liddell FDK; McDonald AD; McDonald JC. The 1891-1920 Birth Cohort of Quebec Chrysotile
Miners and Millers: Development From 1904 and Mortality to 1992. Annals of Occupational
Hygiene. 41:13-36. 1997.
McDonald AD; Fry JS; Wooley AJ; McDonald JC. Dust Exposure and Mortality in an
American Chrysotile Textile Plant. British Journal of Industrial Medicine. 39:361-367. 1983a.
McDonald AD; Fry JS; Woolley AJ; McDonald JC. Dust Exposure and Mortality in an
American Factory Using Chrysotile, Amosite, and Crocidolite in Mainly Textile Manufacture.
British Journal of Industrial Medicine. 40:368-374. 1983b.
D-9
-------
McDonald JC. Invited Editorial: Unfinished Business - The Asbestos Textiles Mystery. Annals
of Occupational Hygiene. 42(l):3-5. 1998b.
Peto J. Lung Cancer Mortality in Relation to Measured Dust Levels in an Asbestos Textile
Factory. In Biological Effects of Mineral Fibres. Wagner JC (ed.). IARC Scientific
Publications, pp. 829-836. 1980a.
Peto J. The Incidence of Pleura! Mesothelioma in Chrysotile Asbestos Textile Workers. In
Biological Effects of Mineral Fibres. Wagner JC (ed.). IARC Scientific Publications, pp.
703-711. 1980b.
Peto J; Doll R; Hermon C; Binns W; Clayton R; Goffe T. Relationship of Mortality to Measures
of Environmental Asbestos Pollution in an Asbestos Textile Factory. Annals of Occupational
Hygiene. 29(3):305-355. 1985.
Piolatto G; Negri E; LaVecchia C; Pira E; Decarli A; Peto J. An Update of Cancer Mortality
Among Chrysotile Asbestos Miners in Balangero, Northern Italy. British Journal of Industrial
Medicine. 47:810-814. 1990.
Sebastien P; Plourde M; Robb R; Ross M; Nadon B; Wypruk T. Ambient Air Asbestos Survey
in Quebec Mining Towns. Part II: Main Study. Environmental Protection Service, Environment
Canada. EPS 5/AP/RQ/2E. July. 1986.
Sebastien P; McDonald JC; McDonald AD; Case B; Harley R. Respiratory Cancer in Chrysotile
Textile and Mining Industries: Exposure Inferences from Lung Analysis. British Journal of
Indus trial Medicine. 46:180-187. 1989.
D-10
-------
APPENDIX E:
CALCULATION OF LIFETIME RISKS OF DYING OF
LUNG CANCER OR MESOTHELIOMA FROM
ASBESTOS EXPOSURE
This appendix describes how additional lifetime risk of lung cancer and mesothelioma are
calculated from the estimated KL, the potency for lung cancer, and KM, the potency for
mesothelioma. Let SE(t | x) be the probability of surviving to age t, given survival to age x
-------
i"1 observational interval, given survival to the beginning of the interval, is calculated as ai*Ai
(risk per person-year times years of observation), and the probability of surviving this age
interval, given survival to the beginning of the interval, is calculated as S^l-b^A;. The
probability of surviving to the beginning of the 1th interval given survival to age Xj is calculated
recursively as
,_, (Eq. E-4)
n
where, by definition, S0=l. The probability PO(X!, x2) of dying of lung cancer between x, and x2,
given survival to xls is the sum over each observational interval of the probability of surviving to
the beginning of the age-interval times the probability of dying of lung cancer in the interval
given survival to the beginning of the interval, or
(Eq. E-5)
This expression represents a discrete approximation to the integral (Eq. E-2).
We now indicate how this expression is modified to account for exposure. First suppose the
exposure pattern E is a step function defined by constant exposure to f (in units of the optimal
exposure index) between ages el and e2, with no exposure at other ages. According to the lung
cancer model (Eq. A-l), in the presence of exposure the mortality rate ai for the 1th observational
interval is increased to ai*(l+KL*di), where di is the cumulative exposure lagged 10 years for
this interval. In the implementation of this algorithm, d; is calculated as
0,
ifm.
if e1+10
-------
account for the dose-related effects of both lung cancer and mesothelioma upon survival,
Si=l-bi*A i is replaced by
S,(E)= 1-fe -a, *KL *4 -KM *Q)* A,
(Eq. E-8)
Similarly, the probability of dying of mesothelioma from exposure pattern E between the ages of
Xj and x2, given survival to xl5 is calculated as
- (Eq.E-9)
i=l ;=0
The oldest (n*) age-interval is unbounded above. In the implementation of the algorithm, a
width of 1/bn is assigned to this interval, which is an estimate of the average survival time in this
age-interval. When, as is typical, the oldest interval is for ages ;> 85 years, this assignment only
affects the calculation when the followup period extends past 85 years (x2>85), and then only
minimally.
When used to estimate risk from continuous exposure (24 hours/day, 7 days/week), KL and KM
were adjusted upward by multiplying by 365/240 (to adjust from an assumed occupational
exposure of 240 days/year to 365 days/year) and by 2.0 (to adjust from an assumed exposure
during work hours to 24 hours/day, assuming that the amount of air breathed during 24 hours is
roughly double the amount breathed during a single work shift.
This algorithm is expanded to handle dose patterns composed of any linear combination of step
functions simply by replacing ds and Q; by the sum of the corresponding terms resulting from
each step function that composes the linear combination. Since any exposure pattern of interest
can be approximated to any degree or accuracy by a linear combination of step functions, the
algorithm can consequently estimate risk from any exposure pattern of interest.
Age-specific mortality rates for both lung cancer (a^ and all-causes (bj) are needed to calculate
asbestos-related risk using the above approach. In order to account for differences in asbestos-
related risk between males and females and — particularly for lung cancer — between smokers and
non-smokers, it is necessary to apply sex- and smoking-specific mortality rates. Lung cancer
and all-cause mortality rates for U.S. males and females for the year 2000 (CDC 2003) are
provided in Table E-l . Also provided in this table are corresponding rates for never-smokers
and current smokers, which were calculated from the U.S. 2000 rates, data on the effect of
smoking obtained from the Cancer Prevention Study II (CPS-II) of the American Cancer Society
(Thun et al. 1997a), and information on the prevalence of smoking obtained from the National
Health Interview Survey (NHIS) (Trosclair et al. 2002). The following paragraphs describe how
these smoking-specific rates were calculated.
E-3
-------
Table E-l. Mortality Rates for All Causes and Lung Cancer per 100,000 Population per Year
All Causes
Age
U.S. 2000
Non-
smokers
Lung Cancer
Non-
Smokers U.S. 2000 smokers Smokers
Males
1
5
130
288
15-20
20-25
25-30
30-35
35-40
40-45
45-50
50-55
55-60
60-65
65-70
70-75
75-80
80-85
85+
799.9
36.5
18.3
25.0
94.9
142.0
141.9
157.3
209.8
303.3
461.7
658.3
1007.5
1565.5
2399.3
3705.4
5591.2
8956.9
16605.4
799.9
36.5
18.3
25.0
94.9
93.8
93.7
103.9
138.4
192.5
314.9
430.1
671.4
1087.1
1690.8
2661.6
4334.9
7257.3
15651.1
799.9
36.5
18.3
25.0
94.9
281.4
281.2
311.7
416.3
623.7
886.0
1318.1
1979.1
2948.5
4447.7
6723.1
9223.2
13870.5
19364.3
0
0
0
0
0
0
0.3
0.8
3.3
10.4
26.0
54.6
118
206.2
327.5
444.0
507.3
549.6
499.0
0
0
0
0
0
0
0.1
0.2
0.9
2.7
5.8
8.1
15.7
23.6
42.6
59.7
80.9
128.4
144.1
0
0
0
0
0
0
0.9
2.5
10.4
32.6
84.4
189.1
413.8
734.2
1151.3
1555
1740
1767.3
1525.2
Females
1
5
130
288
15-20
20-25
25-30
30-35
35-40
40-45
45-50
50-55
55-60
60-65
65-70
70-75
75-80
80-85
85+
654.3
29.1
14.5
16.6
40.0
48.2
56.5
76.0
115.1
172.2
254.0
386.3
611.8
982.0
1527.5
2381.8
3812.6
6444.8
14768.6
654.3
29.1
14.5
16.6
40.0
48.2
56.5
76.0
112.7
171.7
207.6
324.1
486.0
774.5
1199.7
1943.6
3230.0
5753.4
13829.2
654.3
29.1
14.5
16.6
40.0
48.2
56.5
76.0
124.2
174.2
428.5
620.3
1085.1
1762.6
2760.7
4030.4
6004.2
9045.8
18302.3
0
0
0
0
0
0
0
0.9
2.6
8.1
16.6
35.1
70.9
122.3
181.6
238.7
268.6
272.8
213.5
0
0
0
0
0
0
0
0.3
0.8
2.6
3.2
11.2
16.5
32.1
41.4
81.6
79.6
118.0
84.7
0
0
0
0
0
0
0
3.2
9.3
29.0
67.2
125.0
275.4
461.5
709.2
829.6
979.6
855.3
698.2
CPS-II (Thun et al. 1997a) prospectively followed more than one million persons in the U.S.
E-4
-------
beginning in 1982. Subjects were recruited by volunteers and were 2:30 years of age at the time
of enrollment. Smoking status was determined by a questionnaire administered at the time of
enrollment. Never smokers were defined as persons who had never smoked any tobacco
product, and current smokers as persons who were cigarette smokers at the time of enrollment.
Although follow-up has now been extended, following a recommendation of Dr. Thun (Thun
2003), the present calculations are based upon follow-up through 1988 (Thun et al. 1997a).
There are two reasons for this: (1) follow-up past 1988 mostly involves older ages for which
sufficient numbers of deaths had already occurred prior to 1988 to insure adequate statistical
stability of mortality rates; and (2) more importantly, since smoking histories were not updated,
with longer follow-up there is greater misclassification of persons who were classified as current
smokers at time of enrollment, but who may have quit smoking during the follow-up.
Based on follow-up of the CPS-II cohort through 1988, Thun et al. (1997a) present age-specific
mortality rates for a number of causes of death, including lung cancer and all-cause mortality, by
5-year age-intervals beginning at age 30. Separate tabulations are provided for never smokers
and for current smokers in both males and females (reproduced in Table E-2). These rates are
not necessarily representative of the general U.S. population. For example, a member of the
CPS-II cohort is more likely to be college-educated, married, middle-class, and white (Thun et
al. 1997b). Note also, that, despite the fact that smoking is a well-documented health risk,
female smokers in CPS-II (Table E-2) had lower all-cause mortality than U.S. women in general
(Table E-l). Consequently, rather than applying the CPS-II rates directly to the U.S. population,
they are used only to estimate age- and sex-specific relative risks resulting from smoking. These
relative risks are used in conjunction with estimates of the current fraction of smokers to
partition the U.S. 2000 mortality rates between non-smokers and smokers. For a given age, sex
and mortality cause (lung cancer or all-cause mortality), we write
(Eq. E-10)
^2000
where r2000 is the U.S.
2000 mortality rate for the age, sex and cause category, pSM is the proportion of smokers in the
U.S. population, and RRSM is the relative risk for smoking obtained from the CPS-II data
(mortality rate in current smokers divided by mortality rate in non-smokers). The U.S. mortality
rate for non-smokers, %s, is estimated by solving this equation. The corresponding U.S. rate for
smokers is then estimated as the product of the rate in non-smokers and the relative risk for
smoking, rNS*RRSM.
E-5
-------
Table E-2. CPS-II Mortality Rates for All Causes and Lung Cancer
per 100,000 Population per Year8
All Causes
Age
Non-smokers
Smokers
Lung Cancerb
Non-smokers (adj)
Smokers (adj)
Males
25-30
30-35
35-40
40-45
45-50
50-55
55-60
60-65
65-70
70-75
75-80
80-85
85+
—
72.9
93.7
151.8
221.4
367.7
672.6
1096.7
1846.6
3441.2
5466.5
11141.6
—
219.3
303.6
427.1
678.5
1083.8
1824.2
2884.9
4664.5
7321.7
10447.8
13784.9
(0.02)
(0.2)
4.6(0.6)
0.0(1.5)
6.0(2.8)
5.5(4.9)
5.3(7.8)
11.6
21.5
34.9
52
89.2
86.8
(0.3)
(2.2)
5.9(7.4)
18.7(17.5)
41.4
115.3
206.1
361.1
581.6
909
1118.3
1227.7
919
Females
30-35
35-40
40-45
45-50
50-55
55-60
60-65
65-70
70-75
75-80
80-85
85+
80.6
109.3
122.4
182.1
268.2
411.4
666.5
1073.9
1838.7
3154.2
8069.2
88.8
110.9
252.6
348.5
598.8
936.3
1533.7
2227
3417.9
4959.2
10679.2
(0.03)
2.0(0.2)
0.0(0.8)
1.9(2.0)
5.8
7.2
12.3
16.7
30.5
32.5
57.6
60.6
(0.4)
4.0(2.8)
8.9(9.5)
42.4
64.7
119.9
176.6
286.3
310
400
417.6
499.6
Thun et al. (1997a).
""Adjusted rates (in parentheses) were used to calculate the rates in Table E-l.
See text for adjustment method.
E-6
-------
To implement this approach an estimate is needed for pSM, the proportion of smokers in the U.S.
population. Based on the NHIS administered in 2000 to a nationally representative sample of the
U.S. non-institutionalized population over 18 years of age, the proportion of current smokers was
0.257 among men and 0.210 among women. Smoking prevalence was fairly age-independent,
except among persons greater than 65 years of age. Among men the proportions of current
smokers was 0.285, 0.297, 0.264, and 0.102 among men aged 18-24, 25-44, 45-64 and ;>65,
respectively. The corresponding proportions for women were 0.251, 0.245, 0.216, and 0.093
(Trosclair et al. 2002). The oldest category likely includes a sizable percentage of former
smokers whose mortality rates are influenced by their former smoking habits. Because of this
and related problems, it was decided not to age-adjust smoking rates, but simply to apply the
overall rates from the NHIS survey. Consequently, the proportion of smokers was assume to be
pSM=0.257 in men and pSM=0.210 in women.
Smoking-specific mortality rates were not available from CPS-II below the age of 35.
Additionally, in both males and females the CPS-II lung cancer rates in the lowest age categories
were based on fewer than 10 deaths, and consequently quite uncertain. In these age categories,
the lung cancer rates were adjusted using a cubic function of (age less the oldest age at which the
2000 U.S. rate was zero), keeping the total expected number of lung cancer deaths in these
categories equal to the observed number. The resulting adjusted rates are shown in parentheses
in Table E-2. Equation E-10 was applied to these adjusted rates.
Turning now to all-cause mortality, for males between the ages of 35 and 60, the CPS-II all-
cause mortality rates in smokers were approximately three times the rates in non-smokers
(RR»3). Consequently, to estimate smoking-specific rates below the age of 35, equation E-10
was applied using RR=3 between the ages of 20 and 35 and RR=1 for earlier ages. For women,
since the all-cause mortality rates in smokers and non-smokers differed by less than 10%
between the ages of 35 and 45, the rates in smokers and non-smokers were assumed to be equal
below the age of 35. The resulting smoking-specific rates are shown in Table E-l. The
difference in estimated rates between smokers and non-smokers is not necessarily solely due to
smoking; other differences in lifestyle between smoker and non-smokers likely contributed,
particularly among males.
REFERENCES
Center for Disease Control National Center for Health Statistics (CDC). GMWK210. Death
Rates for 113 Selected Causes, by 5-year Age Groups, Race, and Sex: United States, 1999-2000.
http://www.cdc.gov/nchs/datawh/statab/iuipubd/mortabs/gmwk210 10.htm. 2003.
Thun MJ. Personal Communication to Dr. Kenny Crump. 2003.
Thun MJ; Day-Lally C; Myers DG; Calle EE; Flanders WD; Zhu BP; Namboodiri MM; Heath
CW. Trends in Tobacco Smoking and Mortality from Cigarette Use in Cancer Prevention
Studies, I (1959 through 1965) and II (1982 through 1988). In Changes in Cigarette-Related
Disease Risks and Their Implication for Prevention and Control. Burns D; Garfinkel L; Samet
J; (eds.). Smoking and Tobacco Control Monograph, No. 8, National Institutes of Health,
National Cancer Institute. Government Printing Office. Bethesda, MD. 1997a.
E-7
-------
Thun MJ; Peto R; Lopez AD; Monaco JH; Henley SJ; Heath CW; Doll R. Alcohol Consumption
and Mortality among Middle-aged and Elderly U.D. Adults. The New England Journal of
Medicine. 337(24):1705-1714. 1997b.
Trosclair A; Husten C; Pederson L; Dhillon I. Cigarette Smoking among Adults - United States,
2000. Morbidity and Mortality Weekly Report. Surveillance Summaries. 51(29): 642-645.
2002.
E-8
------- |