&EPA
United States
Environmental Protection
Agency
Office of Health and
Environmental 'Assessment
Washington DC 20460"
EPA/600/8-90/057A
July 1990
Workshop Review Draft
Research and Development
Health Assessment
Document for
Diesel Emissions
Sections: 1 Thru 10
Workshop
Review
Draft
(Do Not
Cite or Quote)
Notice
This document is a preliminary draft. It has not been formally
released by EPA and should not at this stage be construed to
represent Agency policy. It is being circulated for comment on its
technical accuracy and policy implications
-------
EPA/600/8-90/057A
July 1990
Workshop Review Draft
Health Assessment Document for
Diesel Emissions
Sections: 1 Thru 10
This document is an internal draft for review purposes
only and does not constitute Agency policy. Mention of
trade names or commercial products does not constitute
endorsement or recommendation for use.
U.S. Environmental Protection Agency
Office of Reasearch and Development
Office of Health and Environmental Assessment
Environmental Criteria and Assessment Office
Research Triangle Park, NC 27711
-------
NOTICE
This document is an internal draft for review purposes
only and does not constitute Agency policy. Mention of
trade names or commercial products does not constitute
endorsement or recommendation for use.
ii
-------
TABLE OF CONTENTS
LIST OF TABLES ix
LIST OF FIGURES xii
1. SUMMARY 1-1
1.1. INTRODUCTION 1-1
1.2. COMPOSITION OF DIESEL EXHAUST 1-4
1.2.1. Introduction 1-4
1.2.2. Overview of Pollutants and Pollution Formation 1-4
1.2.3. Emission Factors and Inventories 1-5
1.2.4. Emission Controls-Now and Projected 1-5
1.2.5 Conclusions 1-5
1.3. CONCENTRATIONS OF DIESEL-DERIVED POLLUTANTS IN
AIR, THEIR TRANSPORT AND TRANSFORMATIONS 1-6
1.3.1. Fundamental Nature of Diesel Pollutants in Air 1-6
1.3.2. Dispersion of Primary Emissions in Air 1-6
1.3.3. Atmospheric Transformations 1-6
1.3.4. Mutagenicity of Ambient Air Particles 1-6
1.4. NONCANCER HEALTH EFFECTS OF DIESEL EXHAUST 1-7
1.4.1. Toxic Effects of Diesel Exhaust on Humans 1-7
1.4.2. Toxic Effects of Diesel Exhaust on Animals 1-8
1.4.3. Interactive Effects of Diesel Exhaust
Components 1-10
1.4.4. Comparison of the Effects of Gasoline and
Diesel Exhaust 1-11
1.5. MUTAGENICITY OF DIESEL ENGINE EMISSIONS 1-12
1.6. METABOLISM AND MECHANISM OF ACTION OF DIESEL
EMISSIONS INDUCED CARCINOGENICITY 1-12
1.7. CARCINOGENICITY OF DIESEL ENGINE EMISSIONS IN
LABORATORY ANIMALS 1-13
1.8. PHARMACOKINETIC CONSIDERATIONS IN THE PULMONARY
CARCINOGENICITY OF DIESEL ENGINE EMISSIONS 1-15
iii
-------
TABLE OF CONTENTS (CONT)
1.9. EPIDEMIOLOGY OF DIESEL EMISSIONS
CARCINOGENICITY 1-16
1.10. QUANTITATIVE ESTIMATE OF UNIT RISK 1-20
1.12. REFERENCES 1-21
2. DIESEL EMISSIONS 2-1
2.1. INTRODUCTION 2-1
2.1.1. Diesel Engine - What it is and How it is Used 2-1
2.2. OVERVIEW OF POLLUTANTS AND POLLUTANT
FORMATION 2-1
2.2.1. Gas Phase Emissions 2-1
2.2.2. Carbon Formation and Emission 2-1
2.2.3. Gas-to-Particle Conversions 2-1
2.2.4. Mutagens 2-2
2.3. EMISSION FACTORS AND INVENTORIES 2-2
23.1. Existing Data 2-2
23.2. Models 2-2
2.4. EMISSION CONTROLS: NOW AND EXPECTED 2-2
2.4.1. Engine Modifications 2-2
2.4.2. Add-On Devices: Descriptions and Performance 2-2
2.4.3. Alternative Fuels: Performance 2-2
2.5. CONCLUSIONS 2-2
2.6. REFERENCES 2-2
3. DIESEL DERIVED POLLUTANTS: ATMOSPHERIC CONCENTRATIONS,
TRANSPORT, AND TRANSFORMATIONS 3-1
1.0. INTRODUCTION 3-2
2.0. PRIMARY DIESEL EMISSIONS 3-4
2.1. Gaseous Emissions 3-4
2.1.1. Inorganic Gases 3-4
2.1.2. Organic Gases 3-6
iv
-------
TABLE OF CONTENTS (CONT.)
2.2. Particulate Emissions 3-9
2.2.1. Diesel Particulate Matter 3-9
2.2.2. Particulate Phase Matter 3-14
2.2.3. Particulate Phase Organic Compounds 3-14
2.3. Factors Influencing Emissions of PAH and Nitro-PAH 3-25
2.4 Gaseous/Particulate Phase Emission Partitioning 3-28
3.0. ATMOSPHERIC TRANSFORMATIONS OF PRIMARY DIESEL
EMISSIONS 3-32
3.1. The Fate of Primary Diesel Emissions and the Long Range
Transport 3-32
3.2. Chemical Transformations 3-35
3.2.1. Gas-phase Reactions 3-35
3.2.2. Particulate-Phase Reactions 3-51
3.3. Physical Removal Processes 3-59
33.1. Dry Deposition 3-59
33.2. Wet Deposition 3-61
4.0. ATMOSPHERIC CONCENTRATIONS OF PRIMARY DIESEL
EMISSIONS AND THEIR TRANSFORMATION PRODUCTS . . 3-65
4.1. Volatile Organic Compounds (VOC) Attributable to Traffic 3-66
4.2. Polycyclic Aromatic Hydrocarbons 3-68
4.3. Nitroarene Concentrations in Ambient Air 3-72
4.4. The Need for Atmospheric Tracers of Diesel Emissions .... 3-78
5.0. MUTAGENICITY OF RESPERABLE AMBIENT PARTICLES .. 3-84
5.1. Bioassay Directed Chemical Analysis 3-85
5.2. Contribution of Nitroarenes to Ambient Air 3-88
6.0. SUMMARY 3-93
7.0. REFERENCES 3-95
v
-------
TABLE OF CONTENTS (CONT.)
4. NONCANCER HEALTH EFFECTS OF DIESEL EXHAUST 4-1
4.1. HEALTH EFFECTS OF WHOLE EXHAUST 4-1
4.1.1. Human Data 4-1
4.1.2. Animal Studies 4-8
4.2. COMPARISON OF HEALTH EFFECTS OF FILTERED AND
UNFILTERED EXHAUST 4-44
4.3. INTERACTIVE EFFECTS OF DIESEL EXHAUST
COMPONENTS 4-47
4.4. COMPARISON OF THE EFFECTS OF DIESEL EXHAUST AND
GASOLINE EXHAUST 4-50
4.5. SUMMARY AND DISCUSSION 4-53
4.5.1. Toxic Effects of Diesel Exhaust on Humans 4-53
4.5.2. Toxic Effects of Diesel Exhaust on Animals 4-54
4.5.3. Interactive Effects 4-62
4.5.4. Comparisons with Gasoline Exhausts 4-63
4.6. REFERENCES 4-65
5. MUTAGENICITY 5-1
5.1. GENE MUTATIONS 5-1
5.2 CHROMOSOME EFFECTS 5-3
5.3 OTHER GENOTOXIC EFFECTS 5-5
5.4 SUMMARY 5-5
5.5 REFERENCES 5-6
6. METABOLISM AND MECHANISM OF ACTION IN DIESEL EMISSION-
INDUCED CARCINOGENESIS 6-1
6.1 METABOLISM CONSIDERATIONS 6-1
6.1.1. Metabolism and Disposition of B[a]P 6-2
6.1.2. Metabolism and Disposition of 1-Nitropyrene 6-7
vi
-------
TABLE OF CONTENTS (CONT.)
6.2. CARCINOGENIC MECHANISM OF PAH COMPONENTS OF
DIESEL EXHAUST 6-9
6.2.1. Carcinogenic Mechanism of B[a]P 6-11
6.2.2. Carcinogenic Mechanism of Nitropyrenes 6-14
6.3. CARCINOGENICITY OF ALDEHYDES 6-17
6.3.1. Metabolism and Carcinogenicity of Formaldehyde 6-17
6.3.2. Metabolism and Carcinogenicity of Acrolein 6-18
6.4. POTENTIAL INVOLVEMENT OF PULMONARY LEUKOCYTES
IN THE DEVELOPMENT OF LUNG TUMORS 6-19
6.5. SUMMARY OF METABOLISM AND MECHANISM OF ACTION
OF CARCINOGENIC COMPONENTS OF DIESEL EXHAUST . 6-23
6.6. REFERENCES 6-26
7. CARCINOGENICITY OF DIESEL EMISSIONS IN LABORATORY
ANIMALS 7-1
7.1. INTRODUCTION 7-1
7.2. CARCINOGENICITY STUDIES IN LABORATORY ANIMALS .. 7-2
7.2.1. Long-term Inhalation Studies 7-2
7.2.2. Short-term Inhalation and Intratracheal
Instillation Studies 7-38
7.2.3. Dermal Application, Subcutaneous Injection,
and Intraperitoneal Injection Studies 7-44
7.2.4. Summary of Animal Carcinogenicity Studies 7-50
7.3. REFERENCES 7-55
8. PHARMACOKINETIC CONSIDERATIONS IN THE PULMONARY
CARCINOGENICITY OF DIESEL ENGINE EMISSIONS 8-1
8.1. INTRODUCTION 8-1
8.2. REGIONAL LUNG DEPOSITION OF INHALED
PARTICLES BY HUMANS AND ANIMALS 8-1
8.3. TRACHEOBRONCHIAL CLEARANCE OF
PARTICULATE MATTER 8-2
vii
-------
TABLE OF CONTENTS (CONT.)
8.4. CLEARANCE FROM DEEP LUNG REGIONS 8-3
8.4.1. Species Variability in Pulmonary Clearance
Processes 8-10
8.4.2. Role of Alveolar Macrophages in the Clearance
of Particulate Matter 8-12
8.4.3. Summary: Pulmonary Clearance of Diesel Exhaust
Particulate Matter 8-37
8.5. DESORPTION OF CONSTITUENTS FROM DIESEL EXHAUST
PARTICLES 8-38
8.5.1. Bioavailability of Agents Adsorbed to Diesel
Exhaust Particles 8-38
8.5.2. Extraction of Carcinogens from Particles by
Alveolar Macrophages and Other Cell Types 8-42
8.5.3. Bioavailability of Adsorbed Compounds as a
Function of Particle Clearance Rates and
Extraction Rates of Adsorbed Compounds 8-48
8.5.4. Summary: Bioavailability of Particle-Adsorbed
Agents 8-50
8.6. INHIBITION OF RESPIRATION BY HIGH CONCENTRATIONS
OF NOXIOUS AGENTS 8-51
8.7. CONSIDERATIONS FOR DOSIMETRY MODELING 8-52
8.8. SUMMARY 8-53
8.9 REFERENCES 8-55
9. EPIDEMIOLOGY STUDIES 9-1
9.1. EPIDEMIOLOGIC STUDIES OF THE CARCINOGENICITY OF
EXPOSURE TO DIESEL EMISSIONS 9-1
9.2 COHORT STUDIES 9-2
9.2.1. Waller (1981): Trends in lung cancer in London
Relation to expsoure to diesel fumes 9-2
9.2.2. Howe et al. (1983): Cancer mortality (1965-1977)
in relation to diesel fume and coal exposure in
a cohort of retired railroad workers 9-4
9.2.3. Rushton et al. (1983): Epidemiological survey
of maintenance workers in the London transport
executive bus garages and Chiswick works 9-6
viii
-------
TABLE OF CONTENTS (CONT.)
9.2.4. Wong et al. (1985): Mortality among members of a
heavy construction operators union with potential
exposure to diesel exhaust emissions 9-7
9.2.5. Edling et al. (1987): Mortality among personnel
exposed to diesel exhaust 9-12
9.2.6. Boffetta and Stellamn (1988): Diesel exhaust
exposure and mortality among males in the American
Cancer Society prospective study 9-14
9.2.7. Garschick et al. (1988): A retrospective cohort
study of lung cancer and diesel exhaust exposure
in railroad workers 9-16
9.3 CASE CONTROL STUDIES OF LUNG CANCER 9-20
9.3.1. Williams et al. (1977): Associations of cancer
site with occuaption and industry from the third
national cancer survey interview 9-20
9.3.2. Hall and Wynder (1984): A case-control study of
diesel exhaust exposure and lung cancer 9-21
9.3.3. Damber and Larsson (1987): Occupation and male
lung cancer: A case control study in northern
Sweden 9-23
9.3.4. Lerchen et al. (1987): Lung cancer and occupation
in New Mexico 9-25
9.3.5. Garschick et al. (1987): A case-contol study of
lung cancer and diesel exhaust exposure in railroad
workers 9-27
9.3.6. Benhamou et aL (1988); Occupational risk factors
of lung cancer in a French case-control study 9-30
9J. 7. Hayes et aL (1988): Lung cancer in motor exhaust-related
(MER) occupations 9-32
9.4 CASE-CONTROL STUDIES OF BLADDER CANCER 9-34
9.4.1. Howe et aL (1980): Tobacco use, occupation,
coffee, various nutrients, and bladder cancer 9-34
9.4.2. Wynder et al. (1985): A case-control study of
diesel exhaust exposure and bladder cancer 9-37
9.4.3. Hoar and Hoover (1985): Truck driving and
bladder cancer mortality in rural New England 9-38
ix
-------
TABLE OF CONTENTS (CONT.)
9.4.4. Iscovich et al. (1987): Tobacco smoking,
occupational exposure and bladder cancer in
Argentina 9-40
9.4.5. Steenland and Burnett (1987): A case-control
study of bladder cancer using city directories
as a source of occupational data 9-43
9.5 SUMMARY 9-45
9.6 REFERENCES 9-51
10. UNIT RISK ESTIMATES 10-1
10.1. INTRODUCTION 10-1
10.2. REVIEW OF PREVIOUS RISK ESTIMATES 10-4
10.3. RISK CALCULATIONS BASED IN ANIMAL BIOASSAY
DATA 10-10
10.3.1. Data Available for Risk Calculations 10-10
10.3.2. Calculations of Unit Risk 10-10
10.4. DISCUSSION 10-20
10.5 WEIGHT OF EVIDENCE 10-23
10.6. SUMMARY 10-27
10.7 REFERENCES 10-29
APPENDIX A: Emission Standards, Exhaust Components, Emissions Models ... A-l
APPENDIX B: Animal Carcinogenicity Summary Table B-l
APPENDIX C: Models for Calculating Lung Burdens C-l
APPENDIX D: Extrapolation Dosimetry Model D-l
x
-------
LIST OF TABLES
TABLE 2-1 Levels of Emissions from Current Diesel and
Gasoline Engines (FTP cycle only) 3-5
TABLE 2-2 Emission Rates of Volatile Organic Compounds (VOC)
from Diesel and Gasoline Engines 3-7
TABLE 2-3 Particulate Matter Emission Rates and Their
Distribution Between Total Carbon (TC) and Organic
Carbon (OC) for Heavy- and Light-Duty Diesel and
Gasoline Engines 3-13
TABLE 2-4 Summary of Composition and Emission Rates (mg/km) of
Airborne Particulate Matter from On-Road Vehicles,
Tuscarora Mountain Tunnel 1977 Experiment 3-15
TABLE 2-5 Classes of Organic Compounds Identified in Particulate-
Phase Combustion Emissions 3-17
TABLE 2-6 Polycyclic Aromatic Hydrocarbons Identified and
Quantified in Extracts of Diesel Particles 3-19
TABLE 2-7 Emission Rates of Particle-Bounded PAH from Heavy-
and Light-Duty Diesel and Gasoline Engines 3-21
TABLE 2-8 Concentrations of Nitro-PAH Identified in a
LDD Particulate Extract 3-24
TABLE 2-9 Factors Affecting Rate of Emission of Polycyclic
Aromatic Hydrocarbons in jig/mile from Diesel
Engine Exhaust and Mutagenicity 3-27
TABLE 2-10 Vapor Pressures at 25 • C for a Series of PAH 3-29
TABLE 3-1 Calculated Atmospheric Lifetimes for Gas-Phase
Reactions of Selected Compounds Present in Automotive
Emissions with Atmospherically Important Reactive
Species 3-37
xi
-------
LIST OF TABLES (CONT.)
TABLE 3-2 Summary of the Nitroarenes Produced from the Gas-Phase
OH Radical-Initiated and N2Os Reactions and
Electrophilic Nitration of PAH 3-52
TABLE 3-3 Average Atmospheric Lifetimes of Particles Due to
Dry Deposition 3-60
TABLE 3-4 Examples of Dry Deposition Velocities 3-61
TABLE 3-5 Mean Particle, Gas, and Overall Scavenging Ratios
for Neutral Organic Compounds 3-64
TABLE 4-1 Concentrations of Individual Hydrocarbons and
Aldehydes Measured in Raleigh, NC 3-67
TABLE 4-2 Particle- and Vapor-Phase PAH Concentrations for
Baltimore Harbor Tunnel Samples 3-69
TABLE 4-3 Average Ambient Concentrations of PAH Measured in
Glendora, CA 3-71
TABLE 4-4 The Maximum Concentrations of Nitrofluoranthene (NF)
and Nitropyrene (NP) Isomers Observed at Three South
Coast Basin Sampling Sites 3-73
TABLE 5-1 Contribution of Nitrofluoranthene (NF) Isomers to the
Direct Mutagenicity of Ambient Particulate Extracts 3-91
TABLE 4-1 Composition of the exposure atmospheres in the
EPA studies 4-11
TABLE 4-2 Composition of the exposure atmospheres in the
EPA studies 4-13
TABLE 4-3 Composition of exposure atmospheres in the
Lovelace studies 4-15
TABLE 4-4 Composition of exposure atmospheres in the
Ishinishi study 4-17
TABLE 4-5 Composition of the atmosphere in the Lewis study 4-19
xii
-------
LIST OF TABLES (CONT.)
TABLE 4-6 Composition of the exposure atmosphere in the
Heinrich 1982 study 4-22
TABLE 4-7 Composition of the exposure atmosphere in the
Heinrich 1986 study 4-22
TABLE 4-8 Composition of exposure atmospheres in the
Mauderly et al., 1987a study 4-28
TABLE 4-9 Composition of atmosphere in Campbell et al. study 4-34
TABLE 4-10 Emission rates for diesel and gasoline engines 4-51
TABLE 4-11 Exposure atmosphere for EPA chronic toxicity study 4-53
TABLE 4-12 Short-term toxicity of diesel exhaust to laboratory
animals 4-56
TABLE 7-1 Gas-phase components of control and diesel exhaust
exposure atmospheres 7-3
TABLE 7-2 Tumor and Lung particulate matter (soot) burden in male
and female F344 rats exposed to diesel exhaust for
24 mo 7-5
TABLE 7-3 Gas-phase components of control (clean air) and Diesel
exhaust atmospheres 7-6
TABLE 7-4 Gas-phase components of control and diesel exhaust
atmospheres 7-8
TABLE 7-5 Tumor incidence in female F344 rats, NMRI mice, and
male and female Syrian golden hamsters following
long-term (120 70 140 weeks) inhalation exposure to
total diesel exhaust (4 mg particles/m3), filtered
exhaust, or clean air 7-10
TABLE 7-6 Gas-phase components of diesel exposure atmospheres 7-12
TABLE 7-8 Tumor incidence in male and female F344 rats following
long-term (30 mo) inhalation exposure to diesel
exhaust 7-14
xiii
-------
LIST OF TABLES (CONT.)
TABLE 7-9 Tumor incidence in ICR and C57B6 mice following
long-term (24 mo) inhalation exposure to diesel
exhaust and in rats exposed to diesel exhaust with
or without DIPN treatment 7-18
TABLE 7-10 Pulmonary lesion incidence in female Wistar rats
following long-term (24 to 30 mo) inhalation exposure
to unfiltered diesel exhaust 7-20
TABLE 7-11 Tumor incidence in male and female F344 rats and
Syrian golden hamsters following long-term inhalation
exposure to filtered or unfiltered diesel exhaust
with or without DEN pretreatment 7-23
TABLE 7-12 Analysis of fresh air control (FA), diesel exhaust
(DE), coal dust (CD), and coal dust plus diesel exhaust
(DEO)) atmospheres 7-26
TABLE 7-13 Concentration of gas-phase components and particles
in exposure chambers 7-29
TABLE 7-14 Lung tumor incidence in strain A mice exposed for 8
weeks to raw or irradiated exhaust at a particle
concentration of 6 mg/m3 7-31
TABLE 7-15 Lung tumor incidence in strain A mice exposed to diesel
exhaust at a particle concentration of 6 mg/m3 until
9 mo of age 7-32
TABLE 7-16 Effects of inhalation exposure to diesel exhaust on
lung tumor incidence in male and female SENCAR mice 7-34
TABLE 7-17 Lung tumor incidence in strain A mice exposed to a
particle concentration of 12 mg/m3 7-35
TABLE 7-18 Tumor incidence in male and female F344 rats
chronically exposed to whole exhaust from LD and HD
engines 7-37
TABLE 7-19 Tumor incidence and survival time of rats treated
with fractions from diesel exhaust condensate
(35 rats/groups) 7-39
xiv
-------
LIST OF TABLES (CONT.)
TABLE 7-20 Pulmonary adenoma response of strain A/J mice exposed
to diesel exhaust (1500 Mg/m3 particles)) for
three mo 7-42
TABLE 7-21 Tumor incidence in female Wistar rats treated with
diphenylnitrosamine (DNP) and exposed to clean air
(CA), filtered diesel exhaust (FDE), or whole diesel
exhaust (DE) 7-43
TABLE 7-22 Tumorigenic effects of dermal application of acetone
extracts of diesel exhaust 7-47
TABLE 7-23 Dermal, tumorigenic and carcinogenic effects of various
emission extracts 7-51
TABLE 10-1 Particulate lung burden and lung tumors in rats
following a 2-year exposure to titanium dioxide (Ti02)
and diesel exhaust 10-2
TABLE 10-2 Estimated lifetime risk of cancer from inhalation of
1 ng/m3 diesel particulate matter 10-9
TABLE 10-3 Incidence of lung tumors in Fisher 344 rats (males and
females combined) exposed to diesel exhaust in air 10-11
TABLE 10-4 Incidence of lung tumors in Fisher 344 rats (males and
females combined) exposed to heavy duty engine
exhaust 10-12
TABLE 10-5 Incidence of lung tumors in Fisher 344 rats (males
and females combined) exposed to diesel exhaust in air
(Brightwell et alM 1986) 10-13
TABLE 10-6 Combined dose-response data from four studies
conducted on Fisher 344 rats 10-14
xv
-------
LIST OF FIGURES
FIGURE 2-1 Typical Size Distribution of Diesel Exhaust Particles 3-11
FIGURE 2-2 Example of Human Respiratory System Aerosol Particle
Deposition Curve 3-12
FIGURE 2-3 Vapor/Particle Phase PAH Distribution in Samples
Collected in Baltimore Harbor Tunnel 3-31
FIGURE 3-1 Diesel-Derived Pollutants: Emission-to-Deposition
Atmospheric Cycle 3-33
FIGURE 4-1 Mass Chromatograms of the Molecular Ion of the
NitroQuoranthenes (NF) and Nitropyrenes (NP) Formed
from the Gas-phase Reaction of Fluoranthene and Pyrene
with the OH Radicals and Present in Ambient Particulate
Sample Collected at Torrance, CA 3-75
FIGURE 4-2 Mass Chromatograms of the Molecular Ion of the
Nitrofluoranthenes (NF) and Nitropyrenes (NP) Present
in Ambient Particulate Samples Collected in Torrance,
CA and Claremont, CA 3-77
FIGURE 5-1 Protocol for Bioassay-Directed Chemical Analysis 3-86
FIGURE 5-2 Distribution of Direct-Acting Mutagenicity (TA98,
-S9) Between Moderately Polar and Polar Fractions of
Extracts of Particulate Matter in Ambient Air, Wood
Smoke, and in the Exhaust from Heavy-Duty and Light-Duty
Motor Vehicles 3-89
FIGURE 6-1 Metabolic pathway for B[a]P and formation of ultimate
carcinogenic intermediate 6-3
FIGURE 6-2 Possible metabolic pathways for in vivo
biotransformation of 1-NP to 6-hydroxy-N-
acetyl-l-aminopyrine 6-10
FIGURE 6-3 Proposed pathway for cellular activation of
1-nitropyrene (1-NP) 6-16
xvi
-------
FIGURE 6-4 Electron micrograph of an aggregate of particle-filled
AM in an alveolus on day 106 following the deposition
of 6.8x10* microspheres 6-24
FIGURE 8-1 Clearance of insoluble particles deposited in
tracheobronchial and pulmonary regions 8-4
FIGURE 8-2 Short-term thoracic clearance of inhaled particles as
determined by model prediction and experimental
measurement 8-5
FIGURE 8-3 Lung retention kinetics of 2.13 mm diam. polystyrene
microspheres after instillation into rat lungs 8-14
FIGURE 8-4 Estimated number of AM with an indicated particle load
of microspheres over a 174 d period following the
instillation of 1.6xl07 microspheres 8-16
FIGURE 8-5 Relationship between the slope (exponents) of later
phase AM disappearance relative to the particle loads
in the AM 8-17
FIGURE 8-6 Relationship of the rates (exponents) of early
component disappearance of AM relative to particle
loads in the AM 8-18
FIGURE 8-7 Proportions of AM with a given load of particles that
disappeared from the total AM population by the early
disappearance components as of a theoretical day 0
postinstillation time and as of day 7 after the particles
were administered 8-19
FIGURE 8-8 Percentage distribution of the retained lung burden
in the various particle-AM categories over the course
of lung clearance of the 1.6xl07 microsphere initial
burden 8-21
FIGURE 8-9 Estimated numbers of AM with a given particle burden
over the course of clearance of 2.0x10s
microspheres 8-23
xvii
-------
FIGURE 8-10 Percentage distributions of retained lung burdens in
the various particle-AM categories during the clearance
of the 2.0x10s initial lung burden of
microspheres 8-24
FIGURE 8-11 Estimated numbers of AM in the indicated particle-
containing categories following the deposition of
6.8x10s microspheres into the rat's lung 8-25
FIGURE 8-12 Electron micrograph of an aggregate of particle
filled AM in an alveolus on day 106 following the
deposition of 6.8x10s microspheres 8-26
FIGURE 8-13 Percentages of the retained lung burden contained in
the various particle-AM categories following the
deposition of 6.8x10s microspheres 8-27
FIGURE 8-14 Micrograph of cells lavaged from a lung 86 d after
the deposition of 6.8x10s microspheres 8-29
FIGURE 10-1 Lung burden (organic material, mg) in rats, under exposure scenario
(7 h/d.5 d/week) in Mauderly et al. (1987) 10-15
FIGURE 10-2 Cumulative lung burden (particles, mg) in rats, under exposure
scenario (7 h/d,5d/week) in Mauderly et al. (1987) 10-17
FIGURE 10-3 Lung burden (organics, mg) in human adults over years after
exposure 10-18
FIGURE 10-4 Effect of increasing the mitotic rate of initiated cells on
the tumor incidence rate over age when other parameters are held
constant 10-22
xviii
-------
Principal authors of this document are:
Dr. Chao Chen
Carcinogen Assessment Statistics and Epidemiology Branch
U.S. Environmental Protection Agency
Washington, D.C.
Dr. Aparna Koppikar
Carcinogen Assessment Statistics and Epidemiology Branch
U.S. Environmental Protection Agency
Washington, D.C.
Dr. Bruce E. Lehnert
Pulmonary Biology-Toxicology Section
Los Alamos National Laboratory
Los Alamos, NM
Dr. Kumar Menon
Albert Einstein Medical Center
Philadelphia, PA
Dr. Gunther Oberdoerster
Department of Biophysics
University of Rochester
Rochester, NY
Dr. Dennis Opresko
Health and Safety Research Division
Biomedical and Environmental Information Analysis
Oak Ridge National Laboratory
Oak Ridge, TN
Dr. William Pepelko
Carcinogen Assessment Toxicology Branch
U.S. Environmental Protection Agency
Washington, D.C
Dr. Larry Valcovic
Genetic Toxicology Assessment Branch
U.S. Environmental Protection Agency
Washington, D.C.
xix
-------
Dr. Ronald K. Wolff
Lilly Research Laboratories
Greenfield Laboratories
Greenfield, IN
Dr. Robert A. Young
Health and Safety Research Division
Biomedical and Environmental Information Analysis
Oak Ridge National Laboratory
Oak Ridge, TN
Dr. C. P. Yu
Department of Mechanical and Aerospace Engineering
Sate University of New York
Buffalo, NY
Project Managers:
Mr. William Ewald
Environmental Criteria and Assessment Office
U.S. Environmental Protection Agency
Research Triangle Park, NC
Dr. William Pepelko
Carcinogen Assessment Toxicology Branch
U.S. Environmental Protection Agency
Washington, D.C
xx
-------
1. SUMMARY
1.1. INTRODUCTION
Diesel engines are used to power ships, heavy machinery, locomotives, and heavy duty
trucks, as well as a small number of light-duty passenger cars and trucks. Cuddihy et al.
(1984) estimated that about 2 percent of the light-duty vehicles in the United States were
equipped with diesel engines.
Since passage of the Clean Air Act of 1963 and the Motor Vehicle Air Pollution
Control Act of 1965, efforts have been made by the federal government to limit pollutant
emissions from mobile sources to levels that will be protective of human health and the
environment. Regulation of emissions from gasoline- and diesel-powered vehicles falls under
the authority of EPA's Office of Air and Radiation, Office of Mobile Sources. EPA
particulate emission standards for light-duty diesel (LDD) vehicles and trucks were
promulgated in 1980 and became effective in 1982 (U.S. EPA, 1980). In 1987, the
particulate emission standard for LDD vehicles became 0.20 g/mi, and the standard for LDD
trucks became 0.26 g/mi. For heavy-duty diesel (HDD) vehicles, a particulate emission
standard of 0.6 gram/brake horsepower-hour (g/bhp-h) took effect in 1988. In 1991, the
standard becomes 0.10 g/bhp-h for urban buses and 0.26 g/bhp-h for all other HDD vehicles.
Effective in 1994, the emission standard for all HDD vehicles will be 0.10 g/bhp-h (Carey,
1987). A summary of gaseous and particulate emission standards for new LDD and HDD
vehicles is given in Appendix A.
The continued use of diesel vehicles has raised concerns regarding the potential
health and environmental effects associated with diesel exhaust Polycyclic aromatic
hydrocarbons (PAH) are adsorbed onto the diesel exhaust particles, and several of these,
such as benzo[a]pyrene and 1-nitropyrene, are known or potential human carcinogens. As
early as 1955, it was reported that repeated skin applications of organic solvent extracts of
exhaust particles resulted in an increased incidence of skin tumors in mice. In 1979, such
extracts were shown to be mutagenic in the Ames Salmonella assay, a useful bioassay for
June, 1990 1-1 DRAFT - DO NOT QUOTE OR CITE
-------
indicating possible carcinogenicity. Since then, the genotaxicity of exhaust extracts has been
demonstrated in a number of in vitro and in vivo assays. Much of the reported mutagenicity
of diesel exhaust extracts has been attributed to nitrated PAHs, and particularly to
nitropyrenes, dinitropyrenes, and nitrohydroxypyrenes.
Suggestive evidence for a carcinogenic effect of inhaled diesel exhaust in animals was
first reported in 1981, in studies demonstrating an increased incidence of pulmonary
adenomas in mice treated with the tumor promotor urethane and subsequently exposed to
diesel exhaust Recent animal studies have since demonstrated a clear association between
chronic inhalation of diesel exhaust and increased incidences of lung tumors.
Evidence for the potential carcinogenicity of diesel exhaust in humans is limited;
however, a few recent studies have indicated a small but significant increased risk of lung
cancer in occupationally exposed workers.
In addition to the potential carcinogenicity of diesel exhaust, there has also been
some concern that diesel particulate matter may contribute to other health problems,
especially those associated with the respiratory tract Respirable particles such as those in
diesel exhaust have been implicated as etiological factors in various types of chronic lung
disease. They may also increase the lung's susceptibility to bacterial and viral infections,
aggravate preexisting diseases such as bronchitis or emphysema, or aggravate specific
respiratory conditions such as bronchial asthma (U.S. EPA, 1986). There is also some
evidence for adverse behavioral and neurological effects. Other components of diesel
exhaust, such as sulfur dioxide, nitrogen dioxide, formaldehyde, acrolein, and sulfuric acid
may contribute to some of these potential health effects.
Several EPA analyses have been conducted to assess the economic, environmental,
and health impacts of the current and proposed changes in the diesel emissions standards,
especially as they pertain to particulate matter (U.S. EPA, 1983; Carey, 1987; Carey and
Somers, 1988). These analyses indicate that emissions of diesel particulate matter in the
United States in 1986 totaled 274,000 metric tons, or about 3.9 percent of the total
suspended particulate emissions from all sources (Carey, 1987). As a result of new and
proposed mobile source emission standards, diesel particulate emissions were projected to
drop to 125,000 to 154,000 metric tons/year in 1995. The annual mean exposure of the U.S.
June, 1990 1-2 DRAFT - DO NOT QUOTE OR CITE
-------
population to diesel particulate matter in 1986 was estimated to be 2.6 Mg/m3 (Carey, 1987).
The exposure level for 1995 was projected to drop to 1.2 to 1.6 jig/m3 as a result of the new
and proposed emission standards.
EPA has evaluated the carcinogenic risks associated with the continued use of
diesel-powered motor vehicles. The calculations were based on the exposure estimates given
above and also on estimates of the carcinogenicity of diesel exhaust In 1983, EPA
calculated the carcinogenic potency of diesel particulate matter by the comparative potency
method (Albert et al., 1983; see also U.S. EPA, 1983). The upper confidence limit on unit
risk was found to be 0.2 to l.OxlO"4 per jig/m3 of particulate matter for a lifetime exposure.
Using these data, EPA estimated that the annua] lung cancer incidence in the U.S.
population in 1986 due to diesel exhaust particulate matter ranged from 178 to 860 (Carey,
1987). It was also estimated that the incidence rate in 1995 would fall between 92 and 443.
The latter figure took into account the more stringent emissions standards and also allowed
for a range of small to large increases in the use of diesel engines in light-duty vehicles.
Since EPA's initial evaluation of the carcinogenic risks associated with diesel exhaust,
additional animal carcinogenicity data have become available. The purpose of this report
is to reevaluate the carcinogenic potency of diesel particulate matter in light of the new data
from these animal studies, as well as to reexamine the evidence available from human
epidemiological studies. Additional issues that are addressed include the interrelationship
between carcinogenic effects and rates of deposition and clearance of the exhaust particles
from the lungs, and the significance of potentially carcinogenic organic compounds adsorbed
to the exhaust particles and their subsequent desorption and bioavailability. Also included
is an overview of the potential noncarcinogenic health effects associated with exposure to
diesel exhaust
June, 1990
1-3
DRAFT - DO NOT QUOTE OR CITE
-------
12. COMPOSITION OF DIESEL EXHAUST
The purpose of this chapter is to provide a basic understanding of how the Diesel
engine contributes to the atmospheric cheical inventory attributed to mobile sources. Its five
elements will provide information which will serve to put those chapters that follow in the
proper context
1.2.1. INTRODUCTION
1.2.1.1. Diesel Engine — What It Is and Why It Is Used.
1. Design Aspects—Two and four stroke cycle; naturally aspirated
and turbocharged; fuel and lubricant systems
2. Typical Uses — Light-duty vehicles; line—haul trucks; buses;
off—road equipment; locomotives
1.2.1.2. Fundamentals of Diesel Combustion — Heterogeneous with Excess
Air.
1. Geometric Variables — Compression ratio; direct and indirect
injection; injection timing
2. Operating Variables — Load; speed; lubricant; fuel
1.2.2. OVERVIEW OF POLLUTANTS AND POLLUTANT FORMATION
1.2.2.1. Gas Phase Emissions
1. Oxides of Nitrogen Formation
2. Hydrocarbon and Carbon Monoxide Formation
1.2.2.2. Carbon Formation and Emission
1. In Flames
2. Physical Characteristics of Emitted Particles
1.2.23. Gas—to—Particle Conversions
1. Organics > Condensation — Absorption
2. Acids > Oxidation — Condensation
June, 1990
1-4
DRAFT - DO NOT QUOTE OR CITE
-------
1.22.4. Mutagens
1. Role of Nitrogen and its Oxides
2 Identification of Emitted Mutagens
1.23. EMISSION FACTORS AND INVENTORIES
1.23.1. Existing Data
1. Regulation Packages
2 Open Literature
1.23.2 Models
1. Mobil—4
2. Part Ext
1.2.4. EMISSION CONTROLS—NOW AND PROJECTED
1.24.1. Engine Modifications
1. Combustion Chamber
2 Intake Port and Valves
3. Piston and Rings
4. Air Handling; Intercooling; Aftercooling
5. Electronic Control of Fuel Management
1.2.4.2 Add—on Devices—Description and Performance
1. Traps
2 Catalysts
1.24.3. Alternative Fuels — Performance
1. Natural Gas
2 Methanol
3. Reformed Diesel Fuel Oil
1.23. Conclusions
June, 1990
1-5
DRAFT - DO NOT QUOTE OR CITE
-------
13. CONCENTRATIONS OF DIESEL-DERIVED POLLUTANTS IN AIR,
THEIR TRANSPORT AND TRANSFORMATIONS
13.1. Fundamental Nature of Diesel Pollutants in Air
1.3.1.1. Gaseous pollutants
13.1.1.1. Inorganic gases
NOd CO, and S02 derived from diesel engines; relative
contribution of diesel vehicles compared to gasoline-
fueled vehicles.
13.1.1.2. Organic gases
Hydrocarbons and derivatives, PAH and derivatives,
aldehydes, etc. Diesel compared to gasoline as a source
of olefinic hydrocarbon in air.
1.3.1.2. Particulate phase pollutants
1.3.1.2.1. Vapor- and particle-phase distribution
Semi-volatile organic compounds (SOC), factors
controlling vapor-to-particle ratio.
13.1.2.2. Particulate phase
PAH and derivates
PAH, nitroarenes, polar PAH.
13.1.23. Particulate carbon in air OC/EC, particulate size
distribution, etc.
13.1.2.4. Other particulate phase pollutant sulfate and trace metals
1.3.2. Dispersion of Primary Emissions in Air
13.2.1. Aliphatic and aromatic hydrocarbons
1.3.2.2. Polycyclic Aromatic Hydrocarbons
1.3.23. Nitroarenes
13.3. Atmospheric Transformations in Air
1.3.3.1. Physical removal processes; wet and dry deposition, atmospheric
transport
13.3.2. Nitroarene formation
133.3 Reactions on particles
13.4. Mutagenicity of ambient air particles Bioassay directed fractionation, etc.
June, 1990
1-6
DRAFT - DO NOT QUOTE OR CITE
-------
1.4. NONCANCER HEALTH EFFECTS OF DIESEL EXHAUST
1.4.1. Toxic Effects of Diesel Exhaust on Humans
Symptoms of acute exposure to high levels (i.e., above ambient) diesel exhaust
include mucous membrane and eye irritation, headache, light-headedness, nausea, vomiting,
heartburn, weakness, numbness and tingling in extremities, chest tightness, and wheezing.
Exhaust odors can cause nausea, headache, and loss of appetite.
The effects of short-term exposures to diesel exhaust have been investigated primarily
in occupationally exposed workers. In studies on underground miners, bus garage workers,
dock workers, and locomotive repairmen, changes in respiratory symptoms and pulmonary
function over the course of a workshift were generally found to be minimal and not
statistically significant. However, in one study of bus garage workers, an increased frequency
of symptoms of cough, labored breathing, itching, burning and watering of the eyes, chest
tightness, and wheezing was observed. In addition, in some studies, reductions in forced vital
capacity (FVQ and forced expiratory volume at 1 s (FEY^ were measured in the exposed
workers. The latter effects were attributed, in part, to nitrogen dioxide or to nitrogen
dioxide adsorbed onto particulate matter.
Chronic effects of diesel exhaust exposure have been evaluated in epidemiological
studies of occupationally exposed workers including miners, railroad yard workers,
stevedores, and bus garage mechanics. In most cases the data have been insufficient to
establish clear correlations between effects and exposure level or length of employment.
Such correlations may be complicated by various factors including smoking habits,
occurrence of allergies, frequency of virally induced respiratory problems, variations in
exhaust composition and workplace ventilation, and the possible presence of other workplace
air pollutants. In a few of these studies, a higher prevalence of respiratory symptoms,
primarily cough and phlegm, were observed among the exposed workers. These symptoms
were usually not accompanied by significant changes in pulmonary function; however,
reductions in FVC and FEV^ and occasionally in forced expiratory flow (FEFM) and FEF75
have been reported. In one study, exposure to diesel exhaust was associated with chronic
respiratory disease.
June, 1990
1-7
DRAFT - DO NOT QUOTE OR CITE
-------
1.4.2. Toxic Effects of Diesel Exhaust on Animals
Animal studies on the toxic effects of diesel exhaust have involved acute, subchronic, and
chronic exposure regimens. In acute exposure studies, toxic effects have been associated
primarily with high concentrations of carbon monoxide, nitrogen dioxide, and aliphatic
aldehydes. In short-term and chronic exposure studies, toxic effects have been associated
with high concentrations of particulate matter. The data for short-term exposures (see
Table 5-12) indicate minimal effects on pulmonary function even at exhaust concentrations
sufficiently high to cause histological and cytological changes in the lungs. Exposures for
several months or longer to levels far above ambient concentrations resulted in accumulation
of particles in the lungs, increase in lung weight, increases in macrophages and leukocytes,
macrophage aggregation, hyperplasia of alveolar epithelium, and thickening of the alveolar
septa. Similar histological changes, as well as reductions in growth rates and alterations in
parameters of pulmonary function, have been observed in chronic exposure studies. Several
animal studies have also demonstrated reductions in resistance to respiratory tract infections
in animals exposed to diesel exhaust. In addition, there are limited animal data associating
minor behavioral and hematological changes, and cytological alterations in the liver with
exposure to diesel exhaust
In animals chronically exposed to diesel exhaust reductions in growth rates have been
observed most often in studies in which the exhaust was diluted to produce concentrations
of particulate matter of at least 2 mg/m3 and in which the exposure periods lasted for 16 or
more hours per day (see Table 5-13). In studies in which the daily exposures were only 6
to 8 h per day, no effects on growth were seen at particle levels of 6 to 8 mg/m3.
Alterations in pulmonary function parameters that have been observed in animals
chronically exposed to diesel exhaust include decreases in vital capacity, residual lung
volume, diffusing capacity, dynamic lung compliance, and expiratory flow rates, as well as
increases in airway resistance (Table 5-14). These effects were not reported in all exposure
studies or for all species tested. They generally appeared only after prolonged exposures.
The lowest exposure levels (expressed in terms of the concentration of exhaust particulate
matter) that resulted in changes pulmonary function parameters occurred at 2 mg/m3 in
June, 1990
1-8
DRAFT - DO NOT QUOTE OR CITE
-------
cynomolgus monkeys, 11.7 mg/m3 in cats, 1.5 and 3.47 mg/m3 in rats, and 4.24 and 6 mg/m3
in hamsters.
Typical histological findings in the lungs of animals chronically exposed to diesel
exhaust include alveolar histiocytosis, macrophage aggregation, tissue inflammation, increase
in polymorphonuclear leukocytes, hyperplasia of bronchiolar and alveolar Type II epithelial
cells, thickened alveolar septa, edema, emphysema, and fibrosis (see Table 5-15). Lesions
in the trachea and bronchi have also been observed in some studies. The exposure levels
(exhaust particle concentrations) at which one or more of these effects occurred were 3.9,
4.24, and 6 mg/m3 in hamsters, 1, 2, 2.2, 3 J, 4.24, 6.0, and 83 mg/m3 in rats, 1.5 mg/m3 in
guinea pigs, and 4.24 mg/m3 in mice. Associated with these histopathological findings are
various histochemical changes in the lung, including increases in lung DNA, total protein,
alkaline and acid phosphatase, glucose-6-phosphate dehydrogenase (G6P-DH), increased
synthesis of collagen, and release of inflammatory mediators such as leukotriene LTB and
prostaglandin PGF^.
Histopathological and histochemical changes appear to be dependent on the
concentration of the exhaust particulate matter, and such changes have not been observed
at low particle concentrations (i.e., 0.7 mg/m3 in rats and 0.25 mg/m3 in guinea pigs).
The histopathological changes caused by prolonged exposure to diesel exhaust appear
to be strongly dependent on relative rates of pulmonary deposition and clearance. At
particle concentrations of 1 mg/m3 or above, pulmonary clearance becomes reduced and
focal aggregations of particle-laden macrophages appear in the lungs, particularly in the
alveolar and peribronchiolar regions, as well as in the hilar and mediastinal lymph nodes.
With prolonged exposures, the continued presence of aggregations of soot-laden
macrophages results in an inflammatory tissue response, septal thickening, epithelial
proliferation, and fibrosis. The critical factor in this process is the minimal exposure, in
terms of the concentration of the particulate matter, as well as in the length and frequency
of the exposure periods, at which focal aggregations of soot-laden macrophages occurs. This
exposure level varies from species to species and is likely to be dependent on rates of
uptake, deposition, clearance, relative size of the macrophage population per unit of lung
June, 1990
1-9
DRAFT - DO NOT QUOTE OR CITE
-------
tissue, rates of recruitment of macrophages and leukocytes, and relative efficiency of the
mucociliary and lymphatic transport systems.
1.43. Interactive Effects of Diesel Exhaust Components
The physiological and toxicological effects of diesel emissions can be enhanced or
altered as a result of the combined actions or interactions of the various exhaust
components. Nitrogen dioxide, sulfuric acid, SO^ and aliphatic aldehydes are chemical
irritants, and one or more of these components may contribute to the eye and respiratory
tract irritation observed in acute exposures to diesel exhaust. Several of these exhaust
components, including NOj, sulfuric acid, SOj, and aliphatic aldehydes, are
bronchoconstrictors. Sulfur dioxide and aliphatic aldehydes cause reductions in respiratory
rates, while N02 may cause an increase. On an individual basis, any of these compounds
could be expected to affect pulmonary function. However, the responses seen in short-term
exposure studies with diesel exhaust are minimal, suggesting that concentrations are too low
or that the methods used are not sensitive enough.
Impaired resistance to respiratory tract infections has been reported in some diesel
exhaust studies. This may come about as a result of reduced mucociliary clearance,
diminished effectiveness of alveolar macrophages, or reduced immunological competence.
Nitrogen dioxide, acrolein, formaldehyde, and sulfuric acid may contribute to this effect as
a result of their immediate cytotoxic activity and the resulting impairment of the mucociliary
clearance mechanism in the respiratory tract
The histological and cytological changes seen in the lungs of animals exposed to
diesel exhaust have been attributed primarily to the accumulation and aggregation of exhaust
particles in lung macrophages. Other exhaust components, particularly NO^ aliphatic
aldehydes, and sulfuric acid are also known to cause histopathological effects in the lungs.
Most effects are directed at the bronchial and bronchiolar epithelial cells, with resulting
hypertrophy and hyperplasia. Exposures to high levels of N02 can also lead to
emphysematous lesions, which have also been reported in some diesel studies.
June, 1990
1-10
DRAFT - DO NOT QUOTE OR CITE
-------
Adsorption of inorganic or organic compounds on diesel particulate matter may alter
the chemical composition and toxicological effects of the exhaust. Sulfur dioxide and
nitrogen dioxide can be adsorbed onto exhaust particles. The S02 could be catalytically
converted to sulfuric acid. Attached to the exhaust particles, these substances may be
carried deeper into lungs where they might have a more direct and potent effect on
epithelial cells or on alveolar macrophages engulfing the particles. In addition, adsorption
of these compounds may alter the acidity of the particles and thereby cause physicochemical
changes affecting the poly cyclic aromatic hydrocarbon fraction of the particles. Binding of
the hydrocarbons and, consequently, bioavailablity may be altered by changes in pH, and this
may result in more direct toxicological activity at the level of lung macrophages or lung
epithelial surface.
1.4.4. Comparison of the Effects of Gasoline and Diesel Exhaust
Animal studies indicate that prolonged exposure to gasoline engine exhaust can result
in histopathological changes in the lungs of exposed animals. These changes consist
primarily of atypical bronchiolar epithelial hyperplasia causing, in some areas, partial
occlusion of the bronchiolar lumina. Hyperplastic changes involving bronchiolar epithelial
tissue also occur in chronic exposures to diesel exhaust. However, in the latter case there
is usually extensive involvement of Type II cells in the alveolar region. This suggests that
the mechanisms involved in the lung response to the two types of exhaust are similar but
that the target tissue may be slightly altered due to the different physical and/or chemical
composition of the exhausts. The higher levels of particulate matter in diesel exhaust may
allow for a greater concentration of the active components of the exhaust in the deeper
regions of the lungs. The effectiveness of diesel exhaust in inducing these tissue responses
might also be enhanced by the presence of other exhaust components.
June, 1990
Ml
DRAFT - DO NOT QUOTE OR CITE
-------
1.5. MUTAGENICITY OF DIESEL ENGINE EMISSIONS
Extensive studies with Salmonella have unequivocally demonstrated direct-acting
mutagenic activity in both particulate and gaseous fractions of diesel exhaust. The induction
of gene mutations has been reported in several in vitro mammalian ceil lines after exposure
to extracts of diesel particles.
Dilutions of whole diesel exhaust did not induce sex-linked recessive lethals in Drosophila
or specific-locus mutations in male mouse germ ceils. Structural chromosome aberrations
and sister chromatid exchanges (SCE) in mammalian cells have been induced by particulates.
Whole exhaust induced micronuclei, but not SCE or structural aberrations, in bone marrow
of male Chinese hamsters exposed to whole diesel emissions for 6 mo. In 7-week exposures,
neither micronuclei nor structural aberrations were increased in bone marrow of female
Swiss mice. Likewise whole diesel exhaust did not induce dominant lethals or heritable
translocations in male mice exposed for 7.5 and 4.5 weeks, respectively.
1.6. METABOLISM AND MECHANISM OF ACTION OF DIESEL
EMISSIONS-INDUCED CARCINOGENICITY
Several studies affirm the bioavailability from inhaled diesel exhaust particles of
compounds such as B[a]P and 1-nitropyrene (1-NP) which are known to be carcinogenic or
mutagenic. Biotransformation of B[a]P, 1-NP, and some of the dinitropyrenes to reactive
intermediates following inhalation of diesel exhaust particles has been verified. Furthermore,
several reports have provided data indicating the formation of DNA adducts, considered an
underlying mechanism of carcinogenicity, following administration of these compounds. The
development of lung tumors in experimental laboratory animals following chronic exposures
to particulate diesel exhaust occurs under conditions in which alveolar macrophage-mediated
particle clearance from the lung is compromised. Although tumors have also been found
to develop with other types of particles (e.g., titanium dioxide) when this clearance
mechanism is diminished, tumors developing in the lungs of diesel emissions-exposed rats
with less lung mass or comparatively less volume burden of diesel particles suggest that the
June, 1990 1-12 DRAFT - DO NOT QUOTE OR CITE
-------
carcinogenic response is not exclusively related to an overabundance of particles in the lungs
per se. Therefore, the organic components on diesel particles may be importantly involved
in the development of lung tumors. The lung's pulmonary macrophages, which phagocytize
deposited diesel particles, probably participate in the gradual in situ extraction and
metabolism of procarcinogens associated with the diesel particles. Additionally, the normal
tumoricidal activities of the pulmonary macrophages may be compromised upon interaction
with excessive numbers of diesel particles, and diesel particle-macrophage interactions could
lead to the generation of reactive oxygen species that have been shown to be at least
mutagenic. Processes and potential mechanisms discussed herein have largely been derived
from animal data, and further research is required to determine how the activities of human
pulmonary macrophages in response to particulate diesel exhaust compare with pulmonary
macrophages from experimental animals. Most importantly, valid dosimetry for the human
condition will require the elucidation of the underlying mechanisms involved in the
development of lung tumors following chronic exposure to whole diesel exhaust
1.7. CARCINOGENICITY OF DIESEL ENGINE EMISSIONS IN
LABORATORY ANIMALS
For human exposure considerations, inhalation is the most relevant mode of
exposure. Only within the last 10 years have extensive, animal studies been conducted
assessing chronic inhalation exposure of diesel exhaust. Studies employing rats and an
adequate experimental design were nearly all positive in demonstrating diesel exhaust-
induced increases in tumorigenicity. The 9.5 percent increase in tumor incidence for female
Wis tar rats exposed to whole diesel exhaust (4 mg/m3) is supported by data showing a 3.6
percent and 12.8 percent tumor incidence in F344 rats following chronic exposure to diesel
exhaust at particle concentrations of 3.5 mg/m3 and 7.0 mg/m3, respectively. Another study
affirmed the observations of potential carcinogenicity of diesel exhaust by providing evidence
for DNA damage in rats. Similarly, diesel exhaust-induced tumorigenicity was observed in
rats exposed to an exhaust particle concentration of 4.9 mg/m3, although the sample size was
small. This study also reported development of a splenic lymphoma which represents the
June, 1990 1-13 DRAFT - DO NOT QUOTE OR CITE
-------
only non-pulmonary tumor resulting from inhalation exposure to diesel exhaust. A long-term
study showed a greater incidence of carcinomas (6.5 percent) in rats following 30 mo
exposure to diesel exhaust at 4 mg/m3 but not at lower (0.4,1, or 2 mg/m3) exposure levels.
However, a dose-dependent increase in tumor incidence was reported for male and female
F344 rats exposed to unfiltered diesel exhaust (five 16-h periods per week) at concentrations
as low as 2.2 mg/m3 and also at 6.6 mg/m3. This study indicated that the tumor incidence
was higher for female than for male rats. A more recent study reported a significant 6.5
percent increase in pulmonary carcinomas in male and female F344 rats chronically exposed
to whole diesel exhaust (3.7 mg/m3). Thus, these studies demonstrated carcinogenic effects
in rats at exposure levels ranging from 2.2 to 7.0 mg/m3. Additional studies also provided
evidence of carcinogenicity in mice.
Two studies provided negative results for tumorigenicity of diesel exhaust in hamsters,
a species known for its resistance to tumor induction. Negative results were also presented
by several other investigators but these studies tended to employ inadequate exposure
durations, low exposure concentrations, or inadequate animal numbers per group. Similarly,
the studies using monkeys and cats were of inadequate duration (2 years) for these longer-
lived species.
Alternate exposure routes including dermal exposure, skin painting, and subcutaneous
injection provided additional evidence for tumorigenic effects of diesel exhaust. Evidence
for tumorigenicity was observed for mice to which an acetone extract of diesel exhaust
particles was applied dermally. Additionally, it has been shown that extracts from some
diesel engines were tumorigenic following dermal application to rodents. A significant
increase in the incidence of subcutaneous tumors in female C57BL mice was reported for
subcutaneous administration of light-duty diesel exhaust tar extract at doses of 500 mg/kg.
Additional data for this study identified an increased tumor incidence in the mice following
injection of light-duty engine exhaust extract at doses of 100 and 500 mg/kg. Negative results
were reported for skin painting studies using mice and acetone extracts of diesel exhaust
particle suspensions, but particle collection procedures may have minimized the condensation
of vapor-phase organics and, therefore, the availability of potentially carcinogenic
components. Intraperitoneal injection studies were generally negative.
June, 1990 1-14 DRAFT - DO NOT QUOTE OR CITE
-------
Experiments using tumor initiators such as DEN, B[a]P, DPIN, or DBahA did not
provide conclusive results regarding the carcinogenic potential of filtered versus whole diesel
exhaust. Some of these studies, however, strongly imply the importance of adsorbed organic
compounds for the carcinogenic effect and data are available regarding the carcinogenic
activity of several compounds known to be components of diesel exhaust
Although the tumorigenicity of the gaseous fraction is presently unresolved and
experiments using filtered exhaust were negative, most of these experiments did not provide
definitive evidence that a mayimiim tolerated dose was achieved. It has not yet been
unequivocally determined that the carbon core of the exhaust particle is not without
carcinogenic potential.
Although uncertainties exist regarding the tumorigenic potential of the gaseous
component and the carbon core component of diesel exhaust, it is clear that diesel exhaust
is carcinogenic in animals. This contention is supported by positive results in numerous,
independent studies, in male and females of at least two species, and by several routes of
administration including, inhalation, intratracheal administration, skin painting, and
subcutaneous injection.
1.8. PHARMACOKINETIC CONSIDERATIONS IN THE PULMONARY
CARCINOGENICITY OF DIESEL ENGINE EMISSIONS
An understanding of the pharmacokinetics associated with pulmonary deposition of
diesel exhaust particles and their adsorbed organics is critical in understanding the
carcinogenic potential of diesel engine emissions. The pulmonary clearance of diesel exhaust
particles is multiphasic and involves several processes including a relatively rapid mucociliary
transport and slower macrophage-mediated processes. The observed dose-dependent
increase in the particle burden of the lungs is due, in part, to an overloading of alveolar
macrophage function. The resulting increase in particle retention has been shown to
increase the bioavailability of particle adsorbed mutagenic and carcinogenic components such
as B[a]P and 1-NP. Experimental data also indicate alveolar macrophage-mediated
metabolism and phagolysosomal solubilization of particle-adsorbed components. Although
June, 1990
1-15
DRAFT - DO NOT QUOTE OR CITE
-------
macromolecular binding of diesel exhaust particle-derived PAH and the formation of DNA
adducts following exposure to diesel exhaust have been reported, a quantitative relationship
between these and increased carcinogenicity is not available.
In addition to the aforementioned points, one must also consider the fact that other
compounds (e.g gas-phase chemical irritants) may alter respiratory rate and, therefore the
actual inhaled dose of potentially toxic components. Moreover, a better knowledge of
particle dissolution rate and particle removal rate is necessary for more accurately assessing
bioavailability of potentially carcinogenic components of diesel exhaust.
1.9. EPIDEMIOLOGY OF DIESEL EMISSIONS CARCINOGENICITY
A major difficulty with the epidemiological studies regarding exposure to diesel
engine emissions was the measurement of the actual diesel exhaust exposure. Most studies
compared persons in job categories with presumably some exposure to diesel exhaust to
either standard populations (presumably no exposure to diesel exhaust) or with men in other
job categories from industries with little or no potential for diesel exhaust exposure. A few
studies have included measurements of diesel fumes, but there is no standard method for
the measurement. Neither is any attempt made to correlate these exposures with the
cancers observed in any of these studies nor is it clear exactly which extract should be
measured to assess the occupational exposure to diesel exhaust. The occupations involving
potential exposure to diesel exhaust are miners, truck drivers, transportation workers,
railroad workers, and heavy equipment operators.
With the exception of one study, there have been no known studies of miners to
assess whether diesel exhaust is associated with lung cancer. Virtually all miners (metal and
coal) use diesel equipment, which was introduced in the early 1960s. Estimates of how many
of these miners are exposed to diesel fumes are not available. The cohort studies mainly
demonstrated an increase of lung cancer. Studies of bus company workers failed to
demonstrate any statistically significant excess risk of lung cancer, but these studies have
certain methodological problems such as small sample sizes, short follow-up periods, lack of
June, 1990
1-16
DRAFT - DO NOT QUOTE OR CITE
-------
information on confounding variables and lack of analysis by duration of exposure or latency
that preclude their use in determining the carcinogenicity of diesel exhaust
A mortality study of heavy equipment operators demonstrated a significant increased
risk of liver cancer in total and in various subcohorts. The same analysis also showed
statistically significant deficits in cancers of the large intestine and rectum. Metastasis from
the cancers of these sites in the liver probably got misclassified as primary liver cancer
leading to observed excess risk of liver cancer. This study also found a non-significant
positive trend for lung cancer with increasing length of membership and latency. However,
due to flaws in methodology and data analysis, this study can not be used to support or
refute a causal association between exposure to diesel exhaust and occurence of cancers.
After controlling for age and smoking, a two-year mortality analysis demonstrated an
excess risk of lung cancer in certain occupations with potential exposure to diesel exhaust.
These excesses were statistically significant among miners (RR = 2.67, 95 percent CI = 1.63
- 4.37) and heavy equipment operators (RR = 2.6, 95 percent CI = 1.12 - 6.06). The
elevated risks were nonsignificant in railroad workers (RR = 1.59) and truck drivers (RR
= 1.24). A dose response was also observed for the truck drivers. Despite the methodologic
limitations such as the lack of representation of the study population comprised of volunteers
only and the questionable reliability of exposure data based on self administered
questionnaires which were not validated, this study is suggestive of a causal association
between exposure to diesel exhaust and excess risk of lung cancer.
Two mortality studies were conducted on railroad workers, one in Canada and one
in the United States. The Canadian study found relative risks of 1.2 (p <0.01) and 1.35
(p <0.001) among "possibly" and "probably" exposed groups, respectively. The trend test
showed a highly significant dose response relationship with exposure to diesel exhaust and
the risk of lung cancer. However, several limitations including lack of detail in work
descriptions, confounding factors of simultaneous coal dust exposure, asbestos exposure, and
smoking habits makes interpretation of this study difficult
The U.S. study provided evidence for linking diesel exhaust exposure to lung cancer.
Relative risks of 1-57 (95 percent CI = 1.19 to 2.06) and 1.34 (95 percent CI 1.02 to 1.76)
were found for ages 40 to 44 and 45 to 49, respectively, after the exclusion of workers
June, 1990 1-17 DRAFT - DO NOT QUOTE OR CITE
-------
exposed to asbestos. This study also found that risk of lung cancer increased with increasing
duration of employment. This large cohort study, with lengthy follow up and adequate
analysis including dose response (based on duration of employment as a surrogate) as well
as adjustment for other confounding factors such as asbestos and smoking, make the
observed association between increased lung cancer and exposure to diesel exhaust more
meaningful.
Among the seven lung cancer case-control studies reviewed in this document, a study
adjusting for age and smoking did not find an increased risk of lung cancer due to diesel
fume exposure. A major limitation of this study was the lack of adequate exposure data
derived from the job titles obtained from occupational histories. On the other hand,
statistically non-significant excess risks were observed for diesel exhaust exposure in three
different studies. The first study found the lung cancer excesses in railroad workers (OR =
1.4) and truck drivers (OR = 1.34). The second study found the excess for workers who
were exposed to diesel exhaust versus who were not (OR = 1.4 and 1.7 with two different
criteria), and the third study found the excess for professional drivers (OR = 1.2). These
rates were adjusted for age and smoking. However, the first two studies had non
participation rates of 47 percent and 36 percent respectively. In the second study the use
of self reported exposures were not validated, which also had low power to detect excess risk
of lung cancer for specific occupations.
After adjusting for smoking, significantly increased risks of lung cancer were found
among French motor vehicle drivers (RR = 1.42) and transport equipment operators (RR
= 1.35). The main limitation of this study was inability to separate the exposures to diesel
exhaust from that of gasoline exhausts since both motor vehicle drivers and transport
equipment operators probably were exposed to the exhausts of both types of vehicles.
A group of investigators combined data from three studies (conducted in 3 different
states) used to increase the power to detect an association of lung cancer with different
occupations which had high potential for exposure to diesel exhaust. They found that truck
drivers employed for more than 10 years had a significantly increased risk of lung cancer
(OR = 1.5, 95 percent CI = 1.1 to 1.9). This study also found a significant trend of
increasing risk of lung cancer with increasing duration of employment among truck drivers.
June, 1990 1-18 DRAFT - DO NOT QUOTE OR CITE
-------
The relative odds were computed by adjusting for birth cohort, smoking, and state of
residence. The main limitation of this study is again the mixed exposures to diesel and
gasoline exhausts since information on type of engine was lacking. Potential bias may have
been introduced because the ascertainment of cause of death for the selection of cases
varied in three studies. The methods used to classify the occupational categories in these
studies are different and, therefore, probably lead to incompatibility of occupational
categories.
The most convincing evidence comes from the case-control study among U.S. railroad
workers. After adjustment for asbestos and smoking, the relative odds for continuous
exposure were 1.39 (95 percent CI = 1.05 to 1.83). Among the younger workers with longer
diesel exhaust exposure, the risk of lung cancer increased with the duration of exposure after
adjusting for asbestos and smoking. Even after the exclusion of recent diesel exhaust
exposure (5 years before death), relative odds increased to 1.43 (95 percent CI = 1.06 to
1.94). This study appears to be a well conducted and well analyzed nested case-control study
with reasonably good power. Potential confounders were controlled adequately and
interactions between diesel exhaust and other lung cancer risk factors were tested.
Of the five bladder cancer case-control studies, three studies found increased risk in
occupations with a high potential diesel exhaust exposure. A significantly increased risk of
bladder cancer was found in Canadian railroad workers (RR = 9.0, 95 percent CI = 12 to
349.5) and in Argentinean truck and railroad drivers (RR = 4.31, p <0.002). Significantly
increased risks were observed with increasing duration of employment of £20 years in truck
drivers (OR = 12,/? = 0.01) and railroad industry workers (OR = 2.21,p <0.05) in the third
study. No significant increased risk was found for any diesel related occupations in the
remaining two studies but both had several limitations such as inadequate characterization
of diesel exhaust exposure, lack of validation of surrogate measures of exposure, presence
of other confounding factors (urinary retention, concentrated smoke within the truck cab
etc), most of them had small sample sizes, and none presented any latency analysis.
An excess risk of lung cancer was observed in three out of seven cohort studies and
six out of seven case-control studies. Of these studies, two cohort and two case-control
studies observed a dose response relationship using duration of employment as a surrogate
June, 1990 1-19 DRAFT - DO NOT QUOTE OR CITE
-------
for dose. However, due to the lack of actual data on exposure to diesel exhaust in these
studies and other methodologic limitations such as lack of latency analysis etc. the evidence
of carcinogenicity in humans is considered to be limited for diesel exhaust exposure.
June, 1990
1-20
DRAFT - DO NOT QUOTE OR CITE
-------
1.10. WEIGHT OF EVIDENCE CLASSIFICATION FOR CARCINO-
GENICITY AND QUANTITATIVE ESTIMATE OF UNIT RISK
On the basis of limited evidence for carcinogenicity of diesel engine emissions in
humans, supported by adequate evidence in animals and positive mutagenicity data, diesel
engine emissions are considered to best fit the weight-of-evidence category Bl. Agents
classified into this category are considered to be probable human carcinogens.
Two different approaches and several different equivalence dose assumptions are
used to calculate unit risk estimates. In approach 1, deposition efficiency and a retardation
of particle clearance from the alveolar region at high inhaled concentration, due to particle
overload effect are taken into account by using mathmatical models. Approach 2 assumes
that the lung burden is linearly proportional to the inhaled concentration in respective of the
level of exposure. The approach that does not take into account the particle overload effect
tends to predict higher risk when compared to those estimates where the overload effect is
eliminated by either using the mathematical model or by using only data from low exposure
groups in different studies (i.e., pooled data). When the particle overload effect is
eliminated (or at least minimized when pooled data are used), the unit risk estimates are as
follows: (Unit risk is equal to lifetime exposure to 1 /xg/m3)
• 3.0x1c6, on organics per lung weight basis, by Approach 1*
1.3xl0"\ on organics per lung surface basis, by Approach 1*
l.OxlO"6, on organics per lung weight basis, pooled data, by Approach 1.
- 4.0x1c6, on organics per lung surface basis, pooled data, by Approach 1.
l.OxlO"6, on body weight basis, pooled data, by Approach 2
• 23xl0"5, on air concentration basis, pooled data, by Approach 2
• 6.1xl0"5, on body surface basis, pooled data, by Approach 2
* Geometric mean unit risks from 3 studies
Of these unit risk estimates, the first two are considered to be the most reasonable
because the effects of particle overload upon the dose of bioavailable, particle associated
June, 1990
1-21
DRAFT - DO NOT QUOTE OR CITE
-------
organics are accounted for in the calculations. Although pooled data provide comparable
risk estimates, it is questionable that pooling data from different studies is appropriate
because these studies were not conducted under the same protocol, and secondly because
the responses were quite small and variable. Of the first two estimates, the one based on
organics per unit lung surface area (13 x 10~s/Mg/m3) is considered to be the most accurate
because it most closely reflects concentration at the critical target organ. Uncertainties
about these risk estimates and research needs to improve the dose-response modeling are
elaborated in the discussion Section 7 Chapter 10.
June, 1990
1-22
DRAFT - DO NOT QUOTE OR CITE
-------
1.11. REFERENCES
Albert, R_ E.; Lewtas, J.; Nesnow, S.; Thorslund, T. W.; Anderson, E. (1983) Comparative potency method
for cancer risk assessment: Application to diesel particulate emissions. Risk Anal 3; 101-117.
Carey, P.M. (1987) Air toxic emissions from motor vehicles. Ann Arbor, MI: Office of Mobile Sources, U.S.
Environmental Protection Agency. EPA Report no. EPA-AA-TSS-PA-86-5.
Carey, P. M.; Somers, J. H. (1988) Air toxic emissions from motor vehicles. Ann Arbor, MI: Office of Mobile
Sources, US. Environmental Protection Agency.
Cuddihy, R. G.; Griffith, W. C; McClellan, R. O. (1984) Health risks from light-duty diesel vehicles.
Environ. ScL TecimoL 18: 14-21A.
U.S. Environmental Protection Agency (1980) Standard for emission of particulate regulation for
diesel-fueled light-duty vehicles and light-duty trucks. Fed. Register 45: 14496.
U.S. Environmental Protection Agency (1983) Diesel particulate study. Office of Mobile Sources, Ann Arbor,
ML
U.S. Environmental Protection Agency (1986) Second addendum to air quality criteria for particulate matter
and sulfur oxides (1982): Assessment of newly available health effects information. Environmental
Criteria and Assessment Office. Research Triangle Park, NC EPA report no. 600/8-86-020F.
June, 1990
1-23
DRAFT - DO NOT QUOTE OR CITE
-------
NOTE: Chapter 2 is in preparation. However, an outline reflecting major topics is provided.
2. DIESEL EMISSIONS (OUTLINE)
The purpose of this chapter is to provide a basic understanding of how the Diesel engine
contributes to the atmospheric cheical inventory attributed to mobile sources. Its five elements
will provide information which will serve to put those chapters that follow in the proper context
2.1. INTRODUCTION
2.1.1. Diesel Engine — What It Is and Why It Is Used.
1. Design Aspects — Two and four stroke cycle; naturally aspirated and
turbocharged; fuel and lubricant systems
2. Typical Uses — Light-duty vehicles; line—haul trucks; buses; off—road
equipment; locomotives
2.1.1. Fundamentals of Diesel Combustion — Heterogeneous with Excess Air.
1. Geometric Variables — Compression ratio; direct and indirect injection;
injection timing
2. Operating Variables — Load; speed; lubricant; fuel
22. OVERVIEW OF POLLUTANTS AND POLLUTANT FORMATION
2-2.1. Gas Phase Emissions
1. Oxides of Nitrogen Formation
2. Hydrocarbon and Carbon Monoxide Formation
2.2.2. Carbon Formation and Emission
1. In Flames
2. Physical Characteristics of Emitted Particles
2.2.3. Gas—to—Particle Conversions
1. Organics > Condensation — Absorption
2. Acids > Oxidation — Condensation
-------
22.4. Mutagens
1. Role of Nitrogen and its Oxides
2. Identification of Emitted Mutagens
23. EMISSION FACTORS AND INVENTORIES
23.1. Existing Data
1. Regulation Packages
2. Open Literature
23.2. Models
1. Mobil—4
2. Part Ext
2.4. EMISSION CONTROLS—NOW AND PROJECTED
2.4.1. Engine Modifications
1. Combustion Chamber
2. Intake Port and Valves
3. Piston and Rings
4. Air Handling; Intercooling; Aftercooling
5. Electronic Control of Fuel Management
2.4.2. Add—on Devices—Description and Performance
1. Traps
2. Catalysts
2.43. Alternative Fuels — Performance
1. Natural Gas
2. Methanol
3. Reformed Diesel Fuel Oil
2~5. Conclusions
-------
NOTE: Chapter 3 has not been formatted for integration into the final document. The text
headers, table numbers, and figure numbers arc not consistent with the total document-
Chapter 3
DIESEL-DERIVED POLLUTANTS: ATMOSPHERIC CONCENTRATIONS,
TRANSPORT. AND TRANSFORMATIONS
Draft Report
DRI Document No. 8622.D1
Prepared for:
U.S. Environmental Protection Agency
Research Triangle Park. NC 27711
Prepared by:
Dr. B. Zielinska
Desert Research Institute
Energy and Environmental Engineering Center
P.O. Box 60220
Reno, NV 89506
June. 1990
June, 1990
3-1
DRAFT DO NOT QUOTE OR CITE
-------
L.O INTRODUCTION
Combustion of fuel in an engine resulcs in the formation of a complex
mixture of gaseous and particulate exhaust. As discussed in detail in the
previous chapter, the concentration of a chemical in the exhaust is a function
of a number of factors, such as engine type, fuel composition, fuel and oil
consumption, and engine operating conditions. During the late 1970s and early
1980s, because of a projected increasing usage of diesel engines in passenger
cars and growing concerns over possible health effects associated with diesel
particulate matter emissions, measurements were made to chemically characterize
in detail the exhausts from light-duty and, co a lesser extent, heavy-duty diesel
engines. Most of these measurements were of primary pollutants, chat is, gases
and particulate matter emitted directly into the air from their sources. In
section 2 of this chapter, che available literature data concerning the chemical
composition of primary diesel emissions and their distribution between gas- and
parcicle-phase are reviewed.
However, it is important co realize chat, once emitted, che primary
pollutants are subjected co dispersion and transport and. at the same time, co
chemical and physical transformations into secondary pollutants, which may be
more harmful than cheir precursors. The time scales of these atmospheric
transformations and physical loss processes vary widely, with atmospheric
lifetimes ranging from < 1 min for some highly reactive organic compounds to
months or even years for other much more inert constituents of direct emissions.
Thus, to assess the impact of diesel emissions on human health and welfare, it
is necessary to determine che chemical and physical changes chat primary diesel
emissions undergo during their transport thcough the atmosphere. Sections 3 and
u of this chapter review che literature data concerning che concentration of
June, 1990
3-2
DRAFT DO NOT QUOTE OR CITE
-------
diesel-derived pollucancs in ambienc air and the atmospheric processes leading
co their formation and disappearance.
June, 1990
3-3
DRAFT DO NOT QUOTE OR CITE
-------
2.0 PRIMARY DIESEL EMISSIONS
Diesel engine exhausts are complex mixtures containing thousands of
chemical compounds in the particulate and gaseous phases. Sections 2.1 and 2.2
of this chapter review data concerning gaseous and particulate diesel engine
emissions, respectively, and compare these data with those of spark-ignition
engine emissions. Section 2.3 briefly reviews factors influencing emissions of
certain organic compounds (mainly polycyclic aromatic hydrocarbons and their
nitrated derivatives). The partitioning of the emitted organics between the
gaseous and particulate phases is discussed in Section 2.4.
It has to be pointed out that rhe detailed chemical characterization of
diesei engine emissions was performed mostly in the late 1970s and early 1980s.
Since chat time substantial changes have occurred in engine and emission control
technologies and additional changes are to be expected in the future (see Chapter
II) . It is therefore reasonable to assume chat the emissions from the currently
manufactured diesel vehicles may not be the same as those measured and reported
earlier. When possible, the latest data were used; however, the data presented
in this chapter should not be considered as fully representative of either
currenc emissions from the wide range of diesel engines presently used or of
chose chat may have occurred in the past.
2.1 Gaseous Emissions
2.1.1 Inorganic Gases
The major products of the complete combustion of petroleum-based fuels in
diesel and gasoline engines are carbon dioxide and water, with nitrogen from air
constituting most of the remaining exhaust. A very small portion of the nitrogen
is converted to nitrogen oxides and some nitrated hydrocarbons (IARC Monographs.
June, 1990
3-4
DRAFT DO NOT QUOTE OR CITE
-------
1989). Diesel exhausts also contain some sulfur dioxide, due co the presence of
sulfur in the dlesel fuel (on the average 0.2 vc % ). Following combustion,
approximately 98% of the sulfur is emitted as S02 and 2% as particulate sulfate
(Pierson et al., 1978; Pierson et al., 1979; Truex et al., 1980). Most of the
sulfate is in the form of H2S04 (Truex et al. , 1980) .
Combustion of dlesel fuel and incomplete combustion of gasoline results in
the emission of CO and many gaseous- and particulate-phase organic compounds
originating from unbumt fuel and lubricating oil. Table 2-1 shows the
comparison of the average emission level of inorganic gases from light-duty
diesei (LDD). heavy-duty diesel (HDD) and light-duty gasoline engines, vith and
without catalytic converter.
Table 2-1
Levels of Emissions from Current Diesel and
Gasoline Engines (FTP cycle only)
Gases Gasoline Car
g/milefg/krn) HDD LDP Noneat. Catalyst
CO* 10 (6.25) 3 (1.9) 15 (9.<0 5 (3.1)
NOx* 28 (17.5) 1 (0.6) U (2.5) 2 (1.2)
S02b 1.6 (1.0) .53 (.33)e .1(.06) ,07(.05)
• From Schuetzle and Frazier (1986).
b From Pierson et al., 1979.
e From Truex et al., 1980.
The concentration of nitric acid (HH03) in light-duty diesel exhausts has
been reported to be negligible in comparison to other anthropogenic sources
(Okaaoto et al., 1983; Karris et al. , 1987). A range of concentrations from -100
ppbv <-250 Mg m"3 ) to -800 ppbv (- 2 mg m*3) [Harris et al., 1987] and an
emission rate of - 1.3 mg/km were reported (Okamoto et al. , 1983).
June, 1990
3-5
DRAFT DO NOT QUOTE OR CITE
-------
A small amount of ammonia was also dececced in diesel engine exhausts
(Plerson and Brachaczek, 1983b). The highest value of 25 mg/kn of NH3 was
reported for heavy-duty diesels, - 4 mg/km for light -duty diesel and -10 mg/km
and -5 rag/km for gasoline-powered vehicles, vich and without catalyst,
respectively (Pierson and Brachaczek, 1983b, and references therein). It was
concluded that motor vehicles are an insignificant source of atmospheric NH3.
2.1.2 Organic Cases
The emissions of gaseous organic compounds from diesel. engines and spark-
Lgnicion engines are qualitatively similar (e.g. , similar chemical components are
present in both exhausts), although there are significant quantitative
differences. Although the same federal emission standard of 0.41 g/mile applies
to exhaust hydrocarbons for both diesel and spark-ignition light-duty vehicles,
diesel engines emit more photochemically reactive hydrocarbons chat are
precursors of smog (National Research Council, 1982). In theory, new spark-
lgnition vehicles equipped with catalytic converters emit almost no reactive
hydrocarbons in cheir exhausts. However, catalyst deterioration over the
Lifetime of the vehicle and the evaporative and refueling emissions cause an
increase in the amount of reactive material released.
Table 2 shows the comparison of the emission rates of some representative
alkanes, alkenes, aromatic hydrocarbons, and aldehydes from heavy- and light-duty
diesel and gasoline engines, with and without a catalytic converter. Data on
catalyst equipped gasoline vehicles are averaged from 46 in-use passenger cars,
1975-1982 models, selected to be representative of vehicles actually driven by
June, 1990
3-6
DRAFT DO NOT QUOTE OR CITE
-------
Table 2-2
Emission Races of Volatile Organic Compounds (VOC)
from Diesel and Gasoline Engines
VOC in
e/fu fs/Tgn)
THC
Me chane
Ethylene
Acetylene
Propylene
n-Pentatie
iso-Pencane
n-Decane
n-Dodecane
Benzene
Toluene
Xylenes
Ethyl benzene
.•aphthalene
Formaldehyde
Acetaldehyde
Acrolein
Benzaldehyde
Total. aldehyde
JU2&.
01( .007)d
. 027 ( . 017)14
.01(.006)r
.005(.003)d
LPP*
. 23(.14)s
.01(.008)
.04(.03)
01(.008)
Gasoline Car
.02(.015)£
. 006(,004)«
.002(.001)«
.001(.0006)®
.003(.002)*
.02(.01)
.007 C.004)"
.01(.006)
. 03(.02)®
Noncac.
5.4 (3.4)
.27 (.17)
.3 (.2)
.26 (.16)
.15 (.09)
.09 (.06)
27 (.17)
.003
.003
.31 (.19)
.7 (.45)
.96 (.6)
.21 (.13)
.06(.04)'
Catalyst'
1.8 (1-2)
.26 (.16)
.14 (.09)
.04 (.02)
.04 ( 03)
.03 ( 02)
.07 ( 04)
(.0016)*
(.002)"
.06 ( 04)
.1 (.07)
.08 (.05)
.02 (.01)
,025(.015)
.01 (.007)
.002(.001)
,003(.002)
.04 (.03)
* From National Research Council (1982), except as indicated.
b From Bailey ec al., 1990, except as indicated.
c From Sigsby et al., 1987.
d From Hampton et al. , 1983, data £rom Allegheny Mountain Tunnel.
* As above, no differentiation between vehicles with and without catalyst.
1 From Schuetzle and Frazier, 1986.
8 From Smith, 1989, and Smith and Paskind, 1989; four-cycle FTP test, 1986
Mercedes Benz.
June, 1990
3-7
DRAFT DO NOT QUOTE OR CITE
-------
che U.S. public (Sigsby et al. , 1987). Table 2 shows the data from the Federal
Test Procedure (FTP) only (which attempts to simulate a typical urban driving
pattern with average speed of - 20 mph), although two other driving cycles
(Crowded Urban Expressway and the New York City cycle) are reported in che
original publication.
The emission data on non-catalyst gasoline vehicles shown in Table 2 are
averaged from 25 in-use passenger cars, representing late 1980s vehicle design
in intensive use in the U.K. (Bailey et al., 1990). To allow direct comparison,
che urban driving data (- 13.5 mph average speed) are given in Table 2.
The data on light-duty diesel emissions are from che National Research
Council Report (1982) and from che recencly published (Smith, 1989; Smith and
PasKind. 1989) studies concerning che evaluacion of particle crap efficiencies
for cwo light-duty diesel passenger cars, Mercedes Benz. 1986 model, and
Volkswagen Jetta, prototype model. However, little has been published recently;
most of che available data are from lace 1970s and early 1980s and refer mainly
co light-duty diesel. Quantitative data on emissions from heavy-duty vehicles
are relatively sparse.
In addition, low concentrations of phenols were reported m heavy- and
light-duty diesel emissions (Hare and Baines, 1979a; Hare and Baines, 1979b).
Aliphatic carboxylic acids (mainly formic, acetic, propionic and benzoic acids)
were also reported in vehicle exhausts (Kavaoura et al., 1985), however no data
exist concerning diesel engine emissions of these acids.
Table 2 lists che emission rates for exhaust pipe emissions only.
Currently, fuel evaporation (e.g.. from fuel lines and carburetors) accounts for
30 - 60% of the total hydrocarbon emissions from passenger gasoline vehicles with
and without catalytic converter (IARC Monographs, 1989). The vapor pressure of
June. 1990
3-8
DRAFT DO NOT QUOTE OR CITE
-------
most current diesel fuels under ambient conditions is so lov chat emission due
co evaporation of diesel fuels is noc significant. However, for both diesel and
gasoline engine emissions, methane, ethylene, and acethylene originate strictly
from tail pipe emissions.
It has to be stressed that individual differences between vehicles of the
same category in the quantity of emitted material are very significant and such
differences arise from many factors of vehicle design, fuel control, engine
conditions, and the general condition of the vehicle at the time of test. The
differences between test parameters (e.g., speed, cold or hot start, fuel
composition, etc.) make the comparison of the data given in Table 2 more
uncertain. However, it is clear from these data, chat the emission profile of
gaseous organic compounds is different for diesel and spark-ignition vehicles;
the aromatic hydrocarbons and low molecular weight alkanes (< C9 ) are more
characteristic of spark-ignition vehicle emissions, while the heavier alkanes
(>C1Q) are more characteristic of diesel emissions.
2.2 Particulate Emissions
2.2.1 Diesel Particulate Hatter
The formation of particles is fundamental to the combustion process in
diesel engines; it begins when fuel is injected into the combustion chamber and
continues during and after the dilution of the exhaust into the atmosphere. As
discussed in detail in the previous chapter, smaller primary spheres, formed
during incomplete combustion of diesel fuel within the combustion cylinder, grow
by agglomeration and by acting as nuclei for the condensation of organic
compounds. The first steps of the process (injection, mixing, ignition, and
combustion events) occur rapidly, within milliseconds. whereas the final steps.
June, 1990
3-9
DRAFT DO NOT QUOTE OR CITE
-------
caking place afcer dilution of che raw exhausc into che atmosphere, extend to
hours or days.
Diesel particles are aggregates of spherical primary particles and 75-95%
of che particulate mass is in the accumulation mode centered about an aerodynamic
diameter of 0.2 urn. Figure 2-1 shows a typical size distribution of diesel
exhaust particles (from National Research Council, 1982). The size distribution
is important, because transport of the particles in the atmosphere (see Section
3.3) and deposition in the human respiratory tract depend essentially on
aerodynamic diameter. Figure 2-2 shows the human respiratory system particle
deposition efficiency curve (Heyder ec al., 1986). The most important feature
of chis curve is che minimum in deposition in the range of 0.2 to 0.5
particles of chis size range are not retained effeccively by che upper
respiracory tract and can penetrate deep into the lung.
Diesel particulate matter is generally defined as any material that is
collected on a filtering medium at a temperature 52 *C or less after dilution of
che raw exhausc gases. Diesel engines produce 2 co 40 times and 2 Co 10 times
more particulate emissions than gasoline engines with and without catalytic
converters, respectively (National Research Council, 1982). The main constituent
of diesel particles is carbon, which accounts for - 80% of total particle mass.
Approximately 70% of this so-called total carbon (TC) occurs in the form of soot
or elemental carbon (EC); che rest is in the form of organic compounds and is
called organic carbon (OC). Table 2-3 compares the emission rates of particulate
matter and its distribution between total and organic carbon for heavy- and
light-duty diesel and gasoline engines. Data on heavy- and light-duty diesel and
June, 1990
3-10
DRAFT DO NOT QUOTE OR CITE
-------
z
Accumulation Mod*
75-95* of Particulate
Milt
m
oc
01
»-
Nuclei Mod*
0-10% ol
Particulate Mats
CoarM Particle Mode
5>15%ol Particulate
Mai*
10.0
0.01
3.0
0.03
0.1
1.0
PARTICLE DIAMETER Uim)
Figure 2-1 Typical size distribution of diesel exhaust particles (from National
Research Council, 1982).
June, 1990
3-11
DRAFT DO NOT QUOTE OR CITE
-------
100
80
60
40
20
0
PARTICLE OIAMETEh
Figure 2-2 Example of human respiratory syscem aerosol parcicle deposition
curve, adapted from Heyder et al., 1986. The dominant particle
deposition mechanism on the left side of the curve is brownian
diffusion; on the right side, it is impaction and sedimentation.
June, 1990
3-12
DRAFT DO NOT QUOTE OR CITE
-------
Table 2-3
Particulate Matter Emission Rates and Their Distribution Between
Total Carbon (TC) and Organic Carbon (OC) for
Heavy' and Light-Duty Diesel and Gasoline Engines
Emission
Particulate
natter in
g/mile (g/km)
TC (% w/w)
OC (% of TC)
HDD
3.2 (2)*
1.4 (.87)d
78*
76*
58*
30*
IQg
.6 (.37)*
80*
36'
Noncac.
CatfrlYSE
.1 (.07)b .02 (.01)c
.04 (.025)"
31
87'
67'
* From Williams et al., 1989b.
b From Uilliams et al., 1989a.
c From Schuetzle and Frazier, 1986.
d From Pierson and Brachaczek, 1983a.
* From Szkarlat and Japar. 1983.
gasoline engines without catalytic converter are from the recently published
survey of 13 HDD vehicles, 19 LDD vehicles, and 22 spark-ignition vehicles
currently in intensive use in Sydney, Australia (Williams et al. , 1989a; Williams
ec al. . 1989b). The vehicles were tested on a dynamometer using several cest
procedures, but Table 2-1 lists the data from ADR 37 cycle, which is essentially
identical to the 1979 U.S. FTP cycle.
Since the dynamometer studies, are not fully representative of the
conditions that exist on the road, the data obtained from the field experiments
in two highway tunnels, the Allegheny and Tuscaroxa Mountain Tunnels of the
Pennsylvania Turnpike (Pierson and Brachaczek, 1983a; Szkarlat and Japar, 1983),
are also given in Table 2-3 for comparison.
June, 1990
3-13
DRAFT DO NOT QUOTE OR CITE
-------
2.2.2 Particulate Phase Inorganics
As discussed above, carbon, both organic and elemental, accounts for - 80%
of total particulate matter mass. The remaining 20% is composed of sulfate
(mainly H2S0t) [Pierson and Brachaczek, 1983a] and some inorganic additives and
adventitious components of fuel and motor oil. Table 2-4 gives the average
compositions of inorganic constituents of airborne particulate matter associated
with vehicles on the road (from Pierson and Brachaczek, 1983a). All airborne
constituents of particulate matter associated with vehicle traffic (other than
atmospheric transformation products of primary emissions) are included, whether
emitted from the exhaust or not (e.g., originated from tire wear debris and soil
dust).
2.2.3 Particulate Phase Organic Compounds
Carbonaceous, diesel-emitted particles have high specific surface areas of
30 to SO m2/g (Frey and Com, 1967). Due to this high surface area, diesel
particles are capable of adsorbing relatively large quantities of organic
material originating from unburned fuel and lubricating oil and from
pyrosynthesis occurring during combustion of fuel (see Chapter II). After
removal of extractable organic material, the surface area of diesel particles
increases up to 90 m2/g (Pierson and Brachaczek, 1976) .
The extractable fraction of diesel particles is typically in the range of
20-30%, but it may be as high as 90% (Williams et al., 1989b) depending on
vehicle type and operating conditions. In general, if a diesel engine is running
under low load, the incomplete combustion leads to a relatively low particle
concentration and a higher proportion of organic compounds associated with the
particles (Dutcher et al., 1984).
June, 1990
3-14
DRAFT DO NOT QUOTE OR CITE
-------
Table 2-4
Summary of Composition and Emission Races (mg/km) of
Airborne Particulate Matter from On-Road Vehicles,
Tuscarora Mountain Tunnel 1977 Experiment
(from Pierson and Brachaczek, 1983a)
Gasoline*
Dieselb
Constituent
rae/kra (% of total mass}
ra^/kra (% of total mass)
H
5 ± 4 (10 ± 6)
47 ± 11 (5 ± 1)
£
.04 ± .6 (.07 ± .11)
1.14 ± .16 (.13 ± .02)
C
34 ± 21 (67 ± 42)
725 ± 117 (84 ± 14)
N
1.1 ± .8 (2 ± 2)
16 ± 2 (1.9 ± .3)
Nac
.09 ± .37 (.2 ± .7)
6.6 ± 1.0 ( .8 ± .1)
Kgc
7 ± .3 (1.3 ± .6)
8 ± 1 (.9 ± .15)
Ale
2 ± 5 (.3 ±. 9)
8.5 ± 1 (1.0 ± .2)
Sid
.5 ± .7 (1.0 ± 1.3)
L4 ± 2 (1.6 ± .2)
P£
07 ± .06 (.13 ±.11)
1.3 ± .2 (.15 ± .02)
S[S04-j
.4(3.4 ±.91 (.9(7+3])
23(42 ± 51 (2.7(4.9±.91)
ci*
.8 ± .4 (1.6 ± .8)
0 (0)
Kd
.17 ± .08 (.3 ± .2)
1.5 ± .2 (.17 ± .03)
Cae
1.3 ± .3 (2.5 ± .7)
5.8 ± 1.4 (.7 ± .2)
Tid
.006 ± .01 (.01 ± .02)
.12 ± .03 (.014 ± .004)
Mne,#
.08 ± .01 (.16 ± .025)
.34 ± .04 (.04 ± .004)
Fee
.32 ± .32 (.6 ± .6)
5.0 ± .9 (.6 ± .1)
Cu
.04 ± .02 (.07 ± .03)
.22 ± .09 (.025 ± .01)
2nr
.04 ± .04 (.08 ± .08)
1.4 ± .1 (.16 ± 01)
Br*
5.75 ± .45 (11.2 ± .9)
0 (0)
Ba-
.03 ± .01 (.07 ± .02)
.66 ± .03 (.08 ± .033)
Pb"
12.4 ± 1.6 (24)
11.5 ± 3 (1.3 ± .3)
* Mostly passenger cars, no distinction between catalytic and noncatalytic
vehicles.
b Mostly heavy-duty diesel trucks, average weight - 30 ton.
e Partially attributable to soil dust.
6 Wholly attributable to soil dust.
* Attributable to fuel additives.
1 Attributable to motor oil.
June, 1990
3-15
DRAFT DO NOT QUOTE OR CITE
-------
The extractable portion of total carbon, although commonly used as a
measure of organic compound concents, is not totally equivalent to the organic
carbon (OC) fraction as measured by the thermal-optical technique. The average
ratio of organic carbon to extractable mass was shown to be 0.70 ± 0.0S, when
toluene/propanol-1 mixture was used as an extraction solvent (Japar et al. ,
1984) , and this ratio was probably due to the presence of both oxygenated organic
compounds and inorganic sulfates in the extracted mass.
2.2 3 1 Extraction and fractionation techniques. A variety of solvents and
extraction techniques have been used in past for the separation of organic
comoounds from diesel particles (Levsen. 1988, and references therein). Although
the reports on the extraction efficiencies are in part contradictory, it appears
that Soxhlet extraction and che binary solvent system composed of aromatic
solvent and alcohol gave the higher recovery of mass, as determined by 14C-BaF
(benzo(a)pyrene1 spiking experiments (Schuetzle and Perez, 1981). Direct
chemical analysis of the entire extractable fraction of diesel particulate matter
is not generally possible because a large number of compounds of different
polarity are present. The separation of diesel particulate organic matter (POM)
into various fractions according to chemical functionalities is a necessary
preliminary step to chemical identification of individual compounds. Open-column
liquid chromatography (LC) and liquid-liquid separation procedures have been the
most widely used fractionation methods (Lee and Schuetzle, 1983). Open-column
LC is very often followed by normal-phase high performance liquid chromatography
(HPLC), if the identification of less abundant components is required.
June, 1990
3-16
DRAFT DO NOT QUOTE OR CITE
-------
2.2.3 2 Chemical composition. Table 2-5 lists the general classes of organic
compounds identified in particles from combustion emissions, including diesel
emissions (from Schuetzle, 1988).
Table 2-5
Classes of Organic Compounds Identified in
Particulate-Phase Combustion Emissions
Soluble Organic Material Insoluble Organic Material
Hydrocarbons Polymers
Derivatives* of hydrocarbons Metal organic complexes
Polvcvclic aromatic hydrocarbons (PAH) Hydrocarbons (>C40)
Derivatives* of PAH
Multifunctional derivatives* of PAH
Heterocyclic compounds
Derivatives* of heterocyclics
Multifunctional derivatives* of
heterocyclic compounds
* Derivatives include: acids, alcohols, aldehydes, esters, ketones, nitrates,
sulfonates.
Liquid chromatography methods usually divide the complex environmental
mixtures of organic compounds into nonpolar, moderately polar and polar
fractions. This separation is achieved by the proper selection of solvents (or
solvent mixtures) for the elution of compounds from chromatographic columns.
Schuetzle and coworkers (1985) proposed that standard chemical compounds be
selected for establishing reference points for the fractionation of diesel
particulate organic matter (POM) into nonpolar, moderately polar, and polar
fractions by normal phase HPLC. They proposed that the elution of 1-
nitronaphthalene would define the end of the elution of nonpolar compounds and
the beginning of the moderately polar fractions. In a similar way, 1,6-pyrene
June, 1990
3-17
DRAFT DO NOT QUOTE OR CITE
-------
quinone would define che end of elucion of moderacely polar compounds and Che
beginning of che polar region.
For llghc-ducy diesel, - 57% of che excracc mass is concained in che
nonpolar fraccion (Schueczle, 1983). Abouc 90% of chis fraccion consists of
aliphacic hydrocarbons from approximacely Cu co abouc C3S (Schueczle. 1983), vich
a carbon number maximum ac C& co C26 (Black and High, 1979) . Polycyclic aromacic
hydrocarbons (PAH) and alkyl-subscicuced PAH accounc for che remainder of che
nonpolar mass.
The moderacely polar fraccion (- 9% v/v of excracc) consiscs mainly of
oxygenaced PAH species and nicraced PAH. The polar fraccion (- 32% w/v of
excracc) is composed mainly of carboxylic and dicarboxylic acids of PAH. hydroxy-
PAH, hydroxymcro-PAH, nicraced N-concaining hecerocyclic compounds, ecc.
(Schueczle. 1983; Schueczle ec al.. 1985).
Limiced recovery scudies have shown chac chere is litcle degradacion or
loss of diesel POM on che HPLC column. Creacer chan 90% of che mass and 70-100%
of che Ames Salmonella cyphimurium-active macerial injecced onco che column have
been recovered (Schueczle ec al., 1985).
2.2 3.3 Polvcycllc aromacic hydrocarbon. Parcicle-bound PAH and cheir
derivacives (mainly nitraced PAH) accracced considerable accencion relacively
early due co cheir mucigenic and, in some cases, carcinogenic propercies
(Nacional Research Council, 1982). The mosc widely used tnechods of PAH analysis
included chin layer chromacography (TLC), capillary gas chromacography (GC), gas
chromacography/mass speccromecry (GC/MS), and HPLC wich UV or fluorescence
dececcion (Levsen, 1988). Table 2-6 liscs che PAH and chioarenes idencified and
June, 1990
3-18
DRAFT DO NOT QUOTE OR CITE
-------
Table 2-6
Polycyclic Aromatic Hydrocarbons Identified and
Quantified In Extracts of Diesel Particles
(from Tong and Karasek, 1984)
ConiDound**b
mol vt
Concentration
^ne/me extractle,d
Acenaphthylene
152
30
Trimechylnaphchalene
170
140
200
Fluorene
166
100
168
Dimechylbipheny1
1B2
30
91
C4-Naphthalene
184
285
351
Trimethylbiphenyl
196
50
Dibenzothiophene
184
129
246
Phenanchrene
178
2.186
- 4,883
Anthracene
178
L55
356
Hethyldlbenzothiophene
198
520
772
He thy1phenanchrene
192
2,028
- 2,768
Methylanthracene
192
517
- 1,522
Echylphenanthrene
206
388
464
4H-Cyclopenta(de f)phenanthrene
190
517
- 1,033
Ethyldlbenzothiophene
212
151
179
2 -Phenylnaphthalene
204
650
- 1,336
DlmechyKphenanchrene/anchracene)
206
1.296
- 2,354
Fluoranthene
202
3,399
- 7,321
Benzo(def)dibenzothiophene
208
254
333
Benzacenaphthylene
202
791
- 1,643
Pyrene
202
3,532
- 8,002
Ethylmethyl(phenanthrene/anthracene)
220
590
717
Methyl(fluoranthene/pyrene)
216
1,548
- 2,412
Benzo(a)Cluorene/benzo(b)fluorene
216
541
990
Benzo(b)naphtho(2,l-d)thiophene
234
30
53
Cyclopentapyrene
226
869
- 1,671
Benzo(ghi)fluoranthene
226
217
418
Benzonaphthothiophene
234
30
126
Benz(a)anthracene
228
463
- 1,076
Chrysene or triphenyLene
228
657
- 1,529
1,2-Binapthyl
254
30
50
He thylbenz(a)anthracene
242
30
50
3 -He thy1chrys ene
242
50
192
June, 1990
3-19
DRAFT DO NOT QUOTE OR CITE
-------
Table 2-6 (continued)
Polycyciic Aromatic Hydrocarbons Identified and
Quantified in Excraccs of Diesel ?articles
(from Tong and Karasek, 1984)
Concentration
Coiroound*,b
mol vt
(nv/me excract)e,d
Phenyl(phenanthrene/anthracene
254
210
559
Benzo(j)fluoranchene
252
492
- 1,367
Benzo(b)fluoranchene
252
421
- 1,090
Benzo(k)fluoranchene
252
91
289
Benzo(e)pyrene
252
487
946
3enzo(a)pyrene
252
208
558
Benzo(ah)anthracene
278
50
96
Indeno(1.2.3-cd)pyrene
276
30
93
Benzo(ghi)perylene
276
443
- 1.050
Dibenzopyrene
302
136
254
* Compounds are arranged according co increasing CC retention times.
b Isomeric alkyl derivatives are not listed separately.
e Concentration range as found in the particulate extract of 3 VW passenger
cars .
J SoLuble organic fractions accounted for 11.1%, 12.1% and 14.7% of total
particulate matter (v/v) for these three diesel samples.
3-20 DRAFT DO NOT QUOTE OR CITE
-------
quantified by GC/MS in three light-duty diesel particulate matter extracts (Tong
and Karasek, 1984) . Data listed in this table reveal the presence of a large
number of alkyl derivatives of PAH, which are sometimes more abundant than the
parent PAH.
Table 2-7 compares the emission rates of several representative PAH from
heavy- and light-duty diesel and gasoline (with and without catalytic converter)
engines.
Table 2-7
Emission Rates of Particle-Bound PAH from
Heavy- and Light-Duty Diesel and Gasoline Engines
(from Schuetzle and Frazier. 1986, except as indicated)
PAH in /ig/mile
HDD
LDP
Gasoline Cars
Noncat.
CataLvst
Pyrene
Fluoranthene
1182 (739)
933 (583)
Benzo(a)pyrene 54 (34)
284 (177)
39 (24)*
66 (42)e
224 (140)
50 (31)e
13 (8.1)
1 (.6)'
NDe
431 (269) 7 (4.4)
45 (28)b
340 (212)
32 (20)b
20 (12.5)
3.2 (2)b
5 (3.1)
4 (.25)
Benzo(e)pyrene 64 (40)
15 (9.4)
3 (1.9)«
23 (14) .4 (.25)
4.8 (3)b
* From Schuetzle and Frazier, 1986, but with 22% fuel aromacity.
b From Alsberg et al., 1985.
e From Smith, 1989, four-cycle FTP test, 1986 Mercedes fienz. ND - none
detected.
June, 1990
3-21
DRAFT DO NOT QUOTE OR CITE
-------
2.2,3.4 Nitrated polvcvcllc aromatic hydrocarbon. Nitro-PAH (nicroarenes) have
been shown to be present in diesel particulate extracts, however in much lower
concentration than the parent PAH (Schuetzle et al. , 1981; Schuetzle et al..
1982; Paputa-Peck et al., 1983). Since many nitroarenes are potent direct-acting
(e.g., without metabolic activation) mutagens in the Ames assay using Salmonella
cyphiourium strains (Rosenkranz and Mermelstein, 1983, and references therein),
the analysis of nitro-PAH in diesel particulate organic matter attracted
considerable attention in the early 1980s.
A large number of nitro-PAH were identified in light-duty diesel
particulate extracts using capillary CC with thermionic nitrogen-phospnorus (NPD)
detector (Paputa-Peck et al.. 1983). Positive isomer identification for 16
nitro-PAH has been made utilizing che CC retention times of authentic standards
and low* and high-resolution mass spectra as identification criteria. These
include 1-nitropyrene, 2-methyl-l-nitronaphthalene, 4-nitrobiphenyl, 2-nitro-
fluorene, 9-nitroanthracene, 9-methyl-10-nitroanthracene. 2-nitroanthracene, 2-
nitrophenanthrene, 1-methyl-9-mtroanthracene , 1-methyl-3-nitropyrene. 1-methyl-
6-nitropyrene. 1-methyl-8-nitropyrene, 1,3-.1,6- and 1,8-dinitropyrene. and 6-
nitrobenzo(a)pyrene. In addition two nitrated heterocyclic compounds were
identified, namely 5- and 8-nitroquinoline. Forty-five additional nitro-PAH were
tentatively identified in this diesel particulate extract (Paputa-Peck et al.t
1983).
The concentration of nitro-PAH adsorbed on diesel particles varies
substantially from sample to sample. Usually 1-nitropyrene is the predominant
component, and concentrations ranging from -7 to -165 Mg/g of particles are
reported (Levsen, 1988).
June. 1990
3-22
DRAFT DO NOT QUOTE OR CITE
-------
Table 2-8 gives the approxi&ace concentrations of several more abundant
nitro-PAH In light-duty diesel particulate extracts (from Campbell and Lee, 1984,
unless noted otherwise).
2.2.3.5 Oxygenated PAH. The moderately polar fraction of diesel particulate
extract contains a variety of oxy-PAH (in particular aldehydes, ketones, quinones
and acid anhydrides) in much higher amounts than nitro-PAH, which elute in the
same fraction. Oxy-PAH are nonmutagenic or very weakly mutagenic, which explains
the relatively low interest in this group of compounds. The most detailed study
of oxy-PAH was puolished by Schuetzle ec al. (1981). who identified more than 100
compounds. A large number of oxy-PAH in the molecular weight range of 182-272
were also identified by Tong ec al. (1984). The main components identified by
Tong and coworkers in particulate matter extracts of three light-duty diesel cars
(W) were: 9-fluorenone, anthraquinone, 4H- cyclopenta(def)phenanthrene-^-one,
9-phenanthrene aldehyde, benzo(de)anthracene-7-one, and benzo(cd)pyrene-6-one.
These components are present at concentrations of 30-300 ng/g particles. Some
of these oxy-PAH are formed during sampling (Levsen, 1988).
2.2.3 6 Polar PAH derivatives. Uhile 65-75% of the directly-acting mutagenicity
(as tested by Ames Salmonella cyphiouriua assay) for light-ducy diesel
particulate extracts is associated with the fraction of moderate polarity, more
than 65% of the mutagenic activity for heavy-duty diesel particulate extract is
concentrated in the most polar fraction (Schuetzle et al., 1985). However, due
to the serious analytical difficulties, only preliminary data exist on the
identification of compounds which are responsible for this mutagenic activity (so
June, 1990
3-23
DRAFT DO NOT QUOTE OR CITE
-------
Table 2-8
Concentrations of Nitro-PAH Identified
in a LDD Particulate Extract
Concencracion
Nitro-PAH' fiig/g of particles)
4-nitrobiphenyl 2.2
2-nitrofluorene 1.8
2-nitroanthracene 4.4
9-nicroanthracene 1.2
9-mtrophenanthrene 1.0
3-nitrophenanthrene 4.1
2-methyl-1-nitroanthracene 8.3
1 - mtrofluoranthene 1.8
7-nitrofluoranthene 0.7
3 - mtrofluoranthene 4.4
8-nitrofluoranthene 0.8
1 -nitropyrene 18.9; 75b
6-nitrobenzo(a)pyrene 2.5
1.3-dinitropyreneb 0.30
1.6-dinitropyreneb 0.40
1.8-dinitropyreneb 0.53
2.7-dinitrofluorenec 4.2; 6.0
2,7-dinicro-9-fluorenone' 8.6; 3.0
4 From Campbell and Lee. 1984, unless noted otherwise. Concentrations
recalculated from yg/g of extract to pg/g of particles using a value of 44%
for excractable material (v/v)
3 From Paputa-Peck et al.. 1983.
c From Schuetzle, 1983.
June, 1990
3-24
DRAFT DO NOT QUOTE OR CITE
-------
called "polar mutagens"). Schueczle and co-workers (Schueczle ec al. , 1985)
employed the concept of "bioassay directed chemical analysis" (see Section 5 of
this chapter) for the isolation and identification of polar PAH derivatives from
che extracts of heavy-duty particulate matter (MIST standard reference material,
SRM 1650). Several hydroxynitro-PAH, hydroxy-PAH, and nitrated heterocyclic
compounds were tentatively identified in the polar fraction.
In another study (Bayona et al.. 1988) , the polar HPLC fractions of the
same NIST SRM 1650 were analyzed by fused silica capillary GC with low- and high-
resolution mass spectrometry (MS), using electron impact (El) and negative ion
cnemical ionization (N1CI) techniques In addition, direct-probe EI andNICI-MS
analyses were performed. Over 80 polycyclic aromatic compounds (PAC) belonging
to several different chemical classes (anhydrides, carboxaldehydes, diazaarenes,
cyclic imides, hydroxynitro-PAH. nitroaza-PAC, nitrolactones, and quinones) were
tentatively identified. Ten of them were positively identified by comparison of
retention times with authentic standards. Among them, phenazme and phthalic
anhydride were positively identified for the first time in dieseL exhaust
particles. In addition, cycLic imides and their alkylated derivatives were
tentatively identified.
2.3. Factors Influencing Emissions of PAH and Nltro-PAH
The influence of diesel engine type and make, diesel fuel aromaticity, and
engine operating conditions (engine timing and load) on the emission level of
pyrene, benzo(a)pyrene, benzo(e)pyrene, and 1-nitropyrene and mutagenicity
(Salmonella cyphitmirium strain TA98 with and without metabolic activation) of
diesel particulate extracts has been investigated (Schuetzle and Frazier, 1986;
June, 1990
3-25
DRAFT DO NOT QUOTE OR CITE
-------
Schueczle and Perez. 1983). Table 2-9 summarizes the resulcs (from IARC
Monographs. 1989) of these experiments.
Particulate samples were collected from four light-duty diesel vehicles
produced by four major manufacturers. The emission rates of the four test
compounds varied by no more than factor of three (see Table 2-9). Also, the
differences in mutagenicities of the total exhaust particulate extracts were not
very significant.
Table 2-9 also shows that the emission of PAH and' total mutagenicity
increased by a factor of 3-4 when the fuel aromaticity was increased from 22% to
55% The 22% and 55% fuels contained 2-24 and 2-60 mg/1 of pyrene. respectively.
The amount of pyrene in the particulate emissions was not related to the pyrene
content in the fuel, indicating that PAH are formed primarily as a result of the
combustion process and are not due to the unburned fuel in the exhaust. In
addition, fuel aromaticity had no effect on the 1-nitropyrene emission. These
facts suggested that the presence of N0X species and not pyrene content is the
limiting factor in the chemical formation of 1-nitropyrene.
Changes in engine timing (see Table 2-9) had little effect on the PAH
emissions but both 1-nitropyrene and mutagenicity increased by several fold as
the timing was advanced, correlating with the increase in emissions of nitrogen
oxides. Engine speed and load had the most significant effect on the emission
of PAH and 1-nitropyrene (Schuetzle and Perez. 1983). High load and speed raise
engine and exhaust temperatures enhancing partial oxidation of PAH and nitro-PAH,
thereby resulting in reduction of PAH and nitro-PAH emissions relative to the
partially oxidized PAH derivatives.
June. 1990
3-26
DRAFT DO NOT QUOTE OR CITE
-------
n
lable 7-9
Factors Aflccllng Rale of falsi Ion of Polycycllc Arun.il ic Hydrocarbons In
lig/otle (pg/lun) frtn Diesel Inglne Eihdusl and Mutagenicity*
U»
I
to
o
o
2
O
H
O
a
o
a
§
Vehic leb
Fuel aromatic H v®
[fialne conditions'1
Retarded
Standard
Advanced
fi
fi
i
e
MJL
SSI
(Hik)
1 imlnq
i1BlP9
Pyrene
39*16
62* 1S
29.IS
24*11
39118
I2S*39
31*20
39*18
35*22
(24tl0|
(19*9)
(18*9)
(15*7)
(24i 11)
('8*24)
(19*13)
(24*11)
(22*14)
Ben*o(a)pyrene
1 3*0 9
1 9*0 1
1 6i0 2
0 6*0 1
1 3i0 5
7 1*3 6
1 7*1 I
t 3*0 S
1 StO.6
(0 8t0 6)
(1 2*0 06)
(1*0 1)
(0 4*0 06)
|0 8i0 3)
(4 4*2 3|
(1 1*0 7)
(0 8*0 3)
(0 9t0 4)
6ento(e]pyrene
2 iti.2
5 1*0 2
3 0*0 4
1 3*0 4
3 Oil 1
10 3*4 1
3 6*2 1
3 0*1 1
4.2*0 S
(1 4*0.8|
I3t0.ll
(1 9*0 3)
(0 8*0 3)
|l 9*0 7|
(6 4*2 6)
12 3*1 3)
(1 9*0.7)
(3.0 3)
l-Nitropyrene
J 0(1.0
7 8*2 2
1 8*0 9
3 8*1 4
4 1*1 9
3 7*1 6
2 3*0 S
4 111 9
15 5*7 7
(1 9*0.6)
(4 9*1 4|
(1 1*0 6)
(2.4*0 9)
(2 6*1 2)
(2 3.1)
(1.4*0 3)
12 6*1 2)
(10iS)
Nitrogen oxides In g/nile
-
-
-
-
-
-
0 9*0 02
1 0*0 0J
1 3*0 1
(g/kn)
(0 6*0 01)
(0 6*0 006)
(0 8*0 06)
Mutagenicity In I04 rev/Bile
(10 rev/kis)
IA98(w1thout actlvatlon|
1 0*0 4
1 3*0 3
0 8*0 3
0 8*0 2
0 99*0 35
2 9*0 80
2 2*1 6
3 4*1 S
6 4*2 7
(0 6i0 2)
(0 8*0 2|
|0 SiO 2)
(0 StO 1)
(0 6*0 2)
(1 8*0 S)
(1.4*1.0)
(2 1*0 9)
14.0*1.7)
IA98 (with activation)
0 StO 2
0 7*0 2
0 610 2
0 StO 1
0 61*0 18
i 1*0 49
1 0*0 S
1 8*0 7
2. St 1 1
(0.3*0.1)
(0.4*0 1)
(0 4*0 1)
(0 3*0 06)
(0 4*0 1)
(1 3*0 3)
(0.6*0 3)
(1 ItO 4)
(1.6*0.6)
"from IMC Monographs (1989), bated on da I* given by Schuelile and Frailer (1986)
''Duplicate tests for each vehicle (Colunns A, 6. C. D) run on four different fuels of 22X aromatic composition
'Duplicate tests on four vehicles run at standard timing on four different fuels
duplicate tests on two vehicles.
-------
2.4 Gaseous/Particulate Phase Emission Partitioning
The distribution of the emissions between the gaseous and particulate
phases is determined by the vapor pressure of the individual species, by the
amount and type of the particulate matter present (adsorption surface available),
and by the temperature (Ligocki and Pankow, 1989). Table 2-10 gives the vapor
pressures at 25 *C of some representative PAH ranging from naphthalene to
benzo(a)pyrene.
The factor of -107 in the range of vapor pressures is reflected in the fact
chat, at equilibrium at ambient temperature, naphthalene exists almost entirely
i.n che gas phase, while SaP, other five-ring, and higher-ring PAH are
predominantly adsorbed on particles. The intermediate three- and four-ring PAH
are distributed between cwo phases.
However, the vapor pressures of these intermediate PAH can be significantly
reduced by their adsorption on various type of surfaces. Because of this
phenomenon, the amount and type of particulate matter present plays an important
role, together with temperature, in the vapor-particle partitioning of
semivolatile organic compounds (SOC).
The measurements of gas/particulate phase distribution are often
accomplished by using a high-volume filter followed by an adsorbent such as
polyurethane foam (PUF), Tenax. or XAD-2 (Cautreels and Van Cauwenberghe, 1978;
Thrane and Hikalsen, 1981; Yamasaki et al., 1982). However, the pressure drop
behind a hi-vol filter or cascade impactor contributes to volatilization of the
chree- to five-ring PAH. to a degree reflecting their vapor pressures. The
magnitude of this "blow-off" artifact depends on a number of factors including
sampling temperature and the volume of air sampled (Van Vaeck et al. , 1984;
Coutant et al., 1988). Despite these problems from volatilization, measurements
June, 1990
3-28
DRAFT DO NOT QUOTE OR CITE
-------
Table 2-10
Vapor Pressures ac 25"C
for a Series of PAH*
Vapor Pressure
ac 298 K
PAH (norr)
Naphthalene
8.0
X
10*2
Acenaphchylene
6.7
X
10*3
Acenaphchene
2.2
X
10°
Fluorene
6.0
X
10*4
Phenanchrene
1.2
X
10**
Anthracene
6.0
X
10"6
Fluoranchrene
9.2
X
10*6
Pyrene
4.5
X
•
O
Benzo[a]anthracene
2.1
X
10*7
Benzo(a]pyrene
5.6
X
10*9
Chryseneb
6.4
X
10*9
'SonnereId ec al., 1983, except as indicaced.
hYanasaki ec al.. 1984.
June, 1990
3-29
DRAFT DO NOT QUOTE OR CITE
-------
with the hi-vol filters followed by a solid adsorbent have provided most
estimates of vapor-parcicle partitioning of SOC in ambient air, as well as
insights into the factors influencing SOC adsorption onto aerosols.
Average distributions of PAH between hi-vol filter and PUF plugs
(positioned downstream of the filter) in samples collected in a heavily traveled
roadway tunnel (Baltimore Harbor Tunnel), are shown in Figure 2-3 (from fienner
et al., 1989). As discussed above, the "blow-off" from the filter preclude
detailed quantitative interpretation. However, it can be seen from this figure
that significant fractions of phenanthrene, anthracene, and their alkylated
derivatives, along with fluoranthene and pyrene. exist in the gas phase. No PAH
less volatile than pyrene were observed in any of the PUF samples. Comparison
of the observed vapor-to-particle PAH ratios and those calculated based on the
relationship derived by Yamasaki and co-workers (1982) , generally agreed within
a factor of two (Benner et al., 1989).
June. 1990
3-30
DRAFT DO NOT QUOTE OR CITE
-------
Particle L_J Vapor Phase
phenanthrene
anthracene
3-methylphenanlhrene
2-methylanthracene
1 -melhylphenanthrene
C2phenanthrene(1)
C2-phenanlhrene(2)
C2-phenanthrene(3)
C2-phenanlhrene(4)
C2-phenanlhrene(5)
C2-phenanthrene(6)
lluoranthene
pyrene
0%
.
1 • vL . I.!-- if. StVzV v ¦•••=. :
• • :
j ii ' i
lltiliiMiltj
1
H ^
•
•
v:':! • vi: \
1 •
lit j4l' if-V'i 2'
¦ .
v • •;*¦' $:£ '<:¦¦ j >>:•; i\ • ::j . ><
:
25% 50%
% in particulate phase
75%
100%
Figure 2-3. Vapor/particle phase PAH distribution in samples collected in
Baltimore Harbor Tunnel
-------
3.0 ATMOSPHERIC TRANSFORMATIONS OF PRIMARY DIESEL EMISSIONS
3.1 The Fate of Primary Diesel Eaissions and the Long Range Transport
Once released into the atmosphere .primary diesel eaissions (or any other
direct emissions) are subjected to dispersion and transport and, at the same
time, to various physical and chemical processes that determine their ultimate
environmental fate. The role of the atmosphere may be compared in some way to
that of a giant chemical reactor in which materials of varying reactivity are
mixed together, subjected to chemical and/or physical processes and finally
removed (Schroeder and Lane, 1988) The main features of the atmospheric cycle
for primary diesel emissions, beginning with emission and ending with deposition
co the Earth's surface, are shown on Figure 3-1 (from Schroeder and Lane, 1988).
Initial mixing describes the physical processes that act on pollutants
immediately after their release from an emission source. The dilution of diesel
exhaust under roadway conditions is an important factor to consider; whereas a
dilution factor of 10 is typical of many dilution tunnels used in dynamometer
studies of automobile exhaust, a dilution factor of -103 is more realistic under
roadway conditions. This discrepancy may lead to slightly different particle
size distributions under real driving conditions than those predicted from
laboratory data (Klttelson and Dolan, 1980); e.g., due co slower coagulation
processes, more particles in the Aitken nuclei range (£ 0.08 pm diameter) may be
expected under typical roadway conditions.
Diffusion and transport processes occur simultaneously in the atmosphere
and account for the dispersion of emissions. The actual distance travelled by
gaseous- and particulate-phase pollutants depends on the amount of time a
specific pollutant resides in the atmosphere and is available for dispersion
(Schroeder and Lane. 1988). As was discussed in the previous section, primary
June, 1990
3-32
DRAFT DO NOT QUOTE OR CITE
-------
Cheat cat
Transformation
Concentration
initial
uixint
Tranioort and
Diffusion
Sicsel
Wet Depot it Ion
Eal>•ion>
Depot it Ion
wtt
Transformation
1
t
Scavtniioi
¦
t
Figure 3-1. Dl<
catsslon-to-deposItIon ataotpnerlc crcte.
June, 1990
3-33
DRAFT DO NOT QUOTE OR CITE
-------
diesel emissions are a very complex mixture containing thousands of organic and
inorganic constituents in the gas and particulate phases. These compounds have
different chemical reactivities and are removed by dry and wet deposition
processes with different rates as discussed more thoroughly in Sections 3.2 and
3.3. As a result, more reactive compounds with short lifetimes will be removed
from the atmosphere relatively quickly, whereas more stable pollutants can be
transported over greater distances. Clearly, a knowledge of the atmospheric loss
processes and lifetimes for automotive emissions is important, since these
lifetimes determine the geographic extent of the influence of these emissions.
It is now well established that man-made pollutants can travel througn the
atmosphere over long distances. In particular the long-range transport of S02
and its transformation to SO" have been studied extensively (see, for example,
Calloway and Uhelpdale, 1980; Lowenthal and Rahn, 1985). The organic pollutants,
particularly those adsorbed on carbonaceous particles, are also subjected to
long-range transport. As will be discussed in Section 3.2.2, organic compounds,
such as polycyclic aromatic hydrocarbons (PAH), adsorbed on diesel particulate
matter are generally more resistant to atmospheric reactions than those in the
gas phase. In addition, particles of smaller diameter (< 1 nm), such as diesel
particulate matter (see Figure 2-1, Section 2.2.1) are removed less efficiently
by wet and dry deposition (see Section 3.3), thus having longer atmospheric
residence times.
It has been reported (Laflamme and Hites, 1978; Hites et al. , 1980) that
PAH and their alkyl homologs are distributed in sediments throughout the world
and that the PAH patterns are similar to each other and to air particulate matter
for most of the locations studied. Furthermore, the quantities of PAH increase
June 1990
3-34
DRAFT DO NOT QUOTE OR CITE
-------
with proximity to urban areas. This suggests anthropogenic combustion sources
and long*range atmospheric transport of PAH.
Conclusive evidence for Long-range transport of PAH was also reported from
the recent measurements of PAH in Siskiwit Lake, located on a wilderness island
in northern Lake Superior (McVetty and Hites, 1988). Because of its remote
location, any PAH found in this lake must have originated exclusively from
atmospheric transport.
Earlier studies by Bjerseth and coworkers (Bjerseth and Olufsen, 1983, and
references therein) showed that PAH are transported from Great Britain and from
the European continent to remote Locations in Norway and Sweden. The specific
sources of PAH emissions could not be identified, however. The authors
speculated that combustion engines were not the major sources of PAH, since the
amounts of BaP found in the samples collected in Norway and Sweden were higher
than could be accounted for from gasoline and diesel fuel consumption in Great
Britain and since the coronene/BaP ratios in the samples were lower than usually
found in gasoline and diesel exhaust. However, due to the lack of PAH profiles
specifically for combustion engines (or. as a matter of fact, any other specific
cracer) the problem of relative contribution of gasoline and diesel vehicle
exhausts to long-range transport of organics cannot be solved at the present
time. The need for such "organic profiles" is addressed in Section A.4. of this
Chapter.
3.2 Chemical Transformations
3.2.1 Gas-phase Reactions
For gas-phase compounds, the following chemical processes contribute to
their removal from the atmosphere (Atkinson, 1988):
June, 1990
3-35
DRAFT DO NOT QUOTE OR CITE
-------
• PhotoLysis during daylight hours.
• Reaction with hydroxyl (OH) radicals during daylight hours.
• Reaction with ozone (03) during daytime and nighttime.
• Reaction with hydroperoxyl (H02) radicals during,
typically, late daytime and early nighttime hours.
• Reaction with gaseous nitrate (N03) radicals during nighttime hours.
• Reaction with dinitrogen pentoxide (N203) during nighttime hours1.
• Reaction with N02 during daytime and nighttime hours.
• Reaction with gaseous nitric acid (HN03) and other species such as
nitrous (HNO:) acid and sulfuric acid (HS04).
The reactive gaseous species, such as OH radicals. N03 radicals. H02
radicals, and ozone, are present in the atmosphere either during the daytime (OH
radicals) or nighttime (N203 and N03 radicals) hours or both time periods (ozone,
N02) . For the routes of formations of these species and their concentrations in
the troposphere, see Finlayson-Pitts and Pitts, 1986.
Table 3-1 gives che calculated atmospheric lifetimes for some selected
compounds present in automotive gas-phase emissions due to known tropospheric
chemical removal reactions (from Atkinson, 1988). These lifetimes (i.e.. the
time for che compound to decay to 1/e or 37% percent of its original
concentration) are calculated from the corresponding measured reaction rate
constants and the average ambient concentration of che tropospheric species
involved.
It has been shown recently (Atkinson et al. , 1990) chat the N203 reactions
with polycyclic aromatic hydrocarbons proceed by initial N03 addition to form an
N03-PAH adduct which either dissociates back to reactants or reacts exclusively
with N02 to form nitroarenes and other products. Since under atmospheric
conditions, where N203, N03 radicals, and NO, are in equilibrium, these reactions
are kinetically equivalent to a reaction with N203 with an effective N20? reaction
rate constant, we will further refer to chese reactions as N203 reactions.
June. 1990
3-36
DRAFT DO NOT QUOTE OR CITE
-------
Table 3*1
Calculated Atmospheric Lifetimes for Gas-Phase Reactions
of Selected Compounds Present in Automotive Emissions
with Atmospherically Important Reactive Species
(from Atkinson, 1988, unless noted otherwise)
Comoound
Atmosoheric Lifetime Due to
OH'
JkL-
TOl'
N02
2 days
12 hr
1 hr
NO
4 days
1 min
3 min
HNO,
180 days
-
-
l/l
o
26 days
>200 yr
>4x10* yr
NHj
140 days
-
-
Propane
19 days
>7,000 yr
-
n-Butane
9 days
>4,500 yr
9 yr
n-Octane
3 days
•
3 yr
Ethylene
3 davs
9 days
3 yr
Propylene
11 hr
1.5 days
15 days
Acetylene
30 days
6 yr
>14 yr
FormaLdehyde
3 days
>2x10* yr
210 days
Acetaldehyde
1 day
>7 yr
50 days
Benzaldehyde
2 days
-
60 days
Acrolein
1 day
60 days
-
Formic acid
50 days
-
-
Benzene
18 days
600 yr
>16 yr
Toluene
4 days
300 yr
9 yr
m-Xylene
11 hr
75 yr
2 yr
Phenol
10 hr
-
20 min
Naphthalene
1 day
>80 days
2-Methylnaph-
5 hr
>40 days
. t
chalene
2.3-dimethyl-
4 hr
>40 days
.£
naphthalene
Acenaphthene
2 hr
>30 days
-3 hr
Acenaphthylene
2 hr
-50 min
13 min
Phenanthrene
9 hr
-
-
Anthracene
2 hr
•
•
Fluoranthene*
6 hr
-
mt
Pyrene*
6 hr
-
mt
H0-»d
2 hr
20 min
>600 yr
hK*
2 mm
23 days
4 hr
60 hr
For 12-hr average concentration of OH
For 24-hr average 03 concentration of
radical of 1 x 10s molecule/cm3.
7 x 1011 molecule/cm3.
For 12-hr average NO, concentration of 2 x 10® molecule/cm3.
For 12-hr average H02 concentration of 108 raolecule/cm3.
For solar zenith angle of 0 ".
Lifetimes due to gas-phase reactions with a 12-hr average concentration of N203
of 2 x 101C molecule/cm3 are: naphthalene -80 days, 2-methylnaphthalene -35
days, 2,3-dimethylnaphthalene -20 days, fluoranthene -64 days, and pyrene -20
days.
Lifetimes calculated from kinetic data given in Atkinson et al.f 1990.
June, 1990
3-37
DRAFT DO NOT QUOTE OR CITE
-------
Although che individual race cons cants are known co a reasonable degree of
accuracy (in general, co vichin a faccor of two) , che tropospheric concencracions
of chese key reactive species are much more uncercain. For example, che ambient
concentrations of OH radicals at any given time and/or location are uncertain to
a factor of at least 5, and more likely 10 (Atkinson, 1988). The cropospheric
diurnally and annually averaged OH radical concentrations are more certain, to
possibly a factor of two. For this reason, che calculated lifetimes listed in
Table 3-1 are approximate only and are valid for those reactive species
concentrations which are listed in che foocnoces. However, chese daca perraic one
co escimace che concribucion of each of chese acmospheric reaccion co che overall
races of removal of mosc pollucancs from che acmosphere.
As can be seen from Table 3-1, che major acmospheric loss process for most
of che automotive emission constituents listed is by daytime reaction with OH
radicals. For some pollutants, photolysis, reactions with ozone, and reactions
with N03 radicals during nightcime hours are also imporcanc removal rouces.
The acmospheric lifetimes do noc cake into consideration che pocencial
chemical or biological imporcance of che produces of chese various reaccions.
For example, che reaccion of gas-phase PAH with N205 appears to be of minor
significance as a PAH loss process, but, as will be discussed later, more
important as a route of formation of mutagenic nitro-PAH.
3.2.1.1 Reaccions of NOx. Only the major atmospheric reactions of oxides of
nitrogen are considered here: for detailed discussion of che chemistry of chese
imporcanc species, see Finlayson-Pitts and Pitts, 1986.
June, 1990
3-38
DRAFT DO NOT QUOTE OR CITE
-------
Oxides of nitrogen emitted, by diesel engines include mainly NO vich lesser
amounts of N02. NO is easily oxidized to N0Z in the reactions with H02 radicals
and alkylperoxy radicals (the reaction of NO with Oz is too slow at typical
ambient concentrations of HO). In addition, NO reacts rapidly (Table 3-1) with
ozone via reaction (1):
NO + Oj - N02 + 02 (1)
N02 is photolysed rapidly at wavelengths of <430 run:
NOj + hy - NO + 0(3P) (2)
The oxygen atom produced in this reaction reacts with 02 forming ozone:
Oj + 0(JP) - 0, (3)
The photolysis of N02 is the only knovn anthropogenic source of ozone, produced
via reaction (2) and (3). With this series of reactions, NO, N02 and 03 are in
a photostationary state:
2 + 0Z
hy
with
kj [N02]
[03]
ki (NO]
—or?)—
June, 1990
3-39
DRAFT DO NOT QUOTE OR CITE
-------
where kj and k2 are Che race constants for" Che reaccions (1) and (2),
respectively, and brackets [ ] signify concentrations. This photostationary
state is strongly affected by NO to N02 conversions caused by reactions involving
organic compounds.
The important atmospheric reactions of N02 also include formation of N03
radicals and N203;
N02 + 0, - N03 + 02
NO, + NO, ^ N2°s
v/ith N;Oj being in equilibrium with NO; and NO, radicals.
The other important atmospheric reactions of NO and N02 include nitrous and
nitric acid formation, respectively, by reaction with OH radicals:
NO + OH - HN02
N02 + OH - HNO,
3.2 1.2 Reactions of SO,. Cas-phase reaction with OH radical is the dominant
S02 atmospheric reaction process (Stockwell and Calverc, 1983):
S0Z + OH - HS03
June, 1990
3-40
DRAFT DO NOT QUOTE OR CITE
-------
followed by the formation of H02 radicals and H2S0t:
HSOj + 02 ~ HOj + S03
| H20
h2so4
Since S02 is soluble in uacer, it undergoes scavenging by fog, cloud water,
and raindrops. In aqueous system, SO, is readily oxidized to sulfate (Calvert
and StockweLl, 1983).
3 2 13 Reactions of alkanes. Only a brief overview of the most important
atmospheric reactions of alkanes is presented here; for detailed discussion
consult Finlayson-Pitts and Pitts, 1986; Atkinson, 1988; Atkinson. 1990, and
references therein.
Under atmospheric conditions, alkanes react with OK radicals during the
daytime and with NO, radicals during nighttime hours:
RH + OH - R- + H20
RH t NOj ' R- t- HNO3
Alkyl radical, R-. reacts with 0Z forming an alkylperoxy radical:
R- + 02 - R02-
which, under polluted urban atmospheric conditions, reacts predominantly with NO
by two pathways: (a) oxidation of NO to N02 and formation of alkoxy radical (R0-)
June, 1990
3-41
DRAFT DO NOT QUOTE OR CITE
-------
-- chis is che only significanc pach for the smaller (< C4) radicals; and (b) che
addition reaction co form scable alkyl nicraces -• chis pachvay becomes
significanc for larger alkyl peroxy radicals:
p— R0- + N02
R02- + NO
I— RONOj
b
Alkoxv radical (RO) reacts essentially by three routes; (a) with 0;, by
aostraction of H acom from the neighboring carbon and formation of stable
carbonyl compound and H02 radical, (b) unimolecular decomposition to form stable
carbonyl compounds and free radicals, which will react further, probably by
analogous rouces as discussed for alkyl radicals above; (c) unimolecular
isomerizacion by 1,4- or 1,5-hydrogen shifc (if a hydrogen acom in appropriace
position is available), forming a- or subscicuced alkyl radical, which will
reacc wich 02 as discussed above.
For example, for 2-pencoxy radical:
CH3CH2CH2CHCH3
0-
e 1,5-H
Shift
CHjCHO + CH3CH2CH2-
ch2ch2ch2chch3
June, 1990
3-42 DRAFT DO NOT QUOTE OR CITE
-------
For Che simplest alkoxy radicals, reaction with 02 (pathway a) is
predominant. For Larger radicals, however, isomerisation (c) and decoaposicion
(b) may become significant, the relative importance of these two pathways
depending on the structure of the radical. In all three cases free radicals are
produced which then will carry on the chain reactions. The first-generation
products include aldehydes, ketones, and alkyl nitrates, which can react further
under atmospheric conditions.
3 2 14 Reactions of Alkene Lower molecular weight alkenes. such as ethylene.
proDvlene. and isomeric bucenes are present in exhaust from gasoLirve and
(although in lower amount) diesel engines (see Table 2-2, Section 2.1.2). Gas-
phase alkenes are removed from the troposphere by reaction with OH radicals. NO,
radicals, and 03 (Finlayson-Pitts and Pitts, 1986; Atkinson, 1988; Atkinson and
Carter, 1984). Reactions with OH radicals are rapid (see Table 3-1) and proceed
by OH radical addicion to the double bond. For example, for propylene:
OH
I
CH3CH-CH2 + OH - CHjCHCHjOH + CH,CHCH, •
(a) -65%
-------
CH3CHCH2OH + 02
NO
~
NOj
?•
CH3CHCH2OH
U-hydroxyalkoxy radical is che major produce. This radical, as discussed above
for alkanes, can (a) reacc with 02 or (b) decompose (isomenzation is noc
important in che case of smaller alkenes):
CH,CHCH,OH
The available daca suggest thac che decomposicion (b) dominaces over reaccion
wich 02. Thus, OH radical reaccions wich alkenes lead co che formacion of
aldehyde and/or ketones.
The NO] radical reaccion becomes imporcanc under acmospheric conditions
only for alkenes more reaccive chan propylene (see Table 3-1). Similarly co che
OH radical reaccion, chis reaccion proceeds chrough NO, radical addicion co che
double bond, followed by reaccion wich 02 (see Finlayson-Pitcs and Pitts, 1986;
Atkinson 1988). Carbonyl compounds are formed as major products, but minor
products (possibly dinitrates) are not well defined (Atkinson, 1988).
The reaccions of alkenes wich ozone compece wich che daycime OH radical
reaccion (see Table 3*1). These reaccions proceed by addicion to the double
bond, followed by rapid decomposition of so-called "ozonide" or "raolozonide" into
a carbonyl compound and an energy-rich biradical:
June, 1990
3-44
DRAFT DO NOT QUOTE OR CITE
-------
xc -
+ o.
n^n
0 0 _
Rl\l l/R'
;c — c
r2 nr*
R^RjCOO + R3R4C-0
R^C-O + R,R4COO
The importance of pathvays (a) and (b) are assunea to be equal (Atkinson and
Career, 1984).
The major uncertainty concerns the atmospheric face of the energy rich-
incermediate, the so-called Criegee biradical. It is believed thac some of the
Criegee intermediates contain sufficient excess energy co spontaneously
decompose. The remaining "thermalized" species are expected co react in the
atmosphere with NO, NQZl S02, H20 vapor, CO, and carbonyl compounds, although the
rate constants and mechanisms of these reactions are presently uncertain
(Finlayson-Picts and Pitts, 1986).
Reactions of orretn containing organics. Diesel and gasoline vehicle
exhauscs contain aldehydes (mainly formaldehyde and acetaldehyde) and (most
probably) carboxylic acids (See Section 2.1.2).
The major loss processes for aldehydes involve photolysis and reaction with
OH radicals (see Table 3-1). Photolysis is an important loss process for
fomaldehyde, forming H atom and HCO radical in the first seep:
June, 1990
3-45
DRAFT DO NOT QUOTE OR CITE
-------
—~ H + HCO
HCHO + hv
H,
CO
The rapid reaction of H acom and HCO radicals wich 02 produces H02 radicals:
H + 02 - H02
HCO + 02 - HO, + CO
The nigher aldehydes also phocoaissociate, ultimately yielding H0; radicals.
The OH radical reaction with formaldehyde yields CO and H02:
HCHO + OH - H20 + HCO-
1 °2
HOj + CO
For higher aldehydes, che RCO radicals initially formed add 02 to yield
acylperoxy radical, which can reacc further vich NO (pathway a) and N02 (pathway
b):
RCHO + OH - RCO* + H20
1 o2
RC(0)00-
NO I * NOj
RC(0)0- + N02 RC(0)00N02
R* + CO,
(peroxyacyl nitrates)
June, 1990
3-46
DRAFT DO NOT QUOTE OR CITE
-------
Acecaldehyde forms peroxyacecyl nicrace (PAN) which has been shown co be a
direct-acting mutagen toward Ames Salmonella, cyphiouri.ua strain TA100
(Kleindiensc ec al., 1985) and is phycocoxic.
Benzaldehyde, the simplest aromatic aldehyde, forms peroxybenzoyl nicrace
by che series of reaccions analogous co pachway (b), above, and nicropnenols as
a resulc of che reaccion with NO, analogous to pachway (a), above:
Jlf2 [b) C6H3C(0)00N02
C8HjC(0)00-
NO (a)
r~
C.HS • + CO,
N02
NO N02
C6Hs- t- 02 - CgHjOO • J CeHj0 • » N02C6H40H
N02 JH02 (o- and p-)
CflHsOH + 02
Carboxylic acids react vith OH radicals under atmospheric conditions (see
Table 3-1). For formic acid, hydrogen atoms are produced (Atkinson 1988):
HC(0)0H + OH - HjO + C02 + H
For che higher carboxylic acids, reaccion produces are presencly unknown.
3.2 16 Reactions of monocyclic aromatic compounds. The monocyclic aromacic
compounds are removed from che acmosphere solely by reaccions wich OH radicals
(see Table 3-1). These reaccions proceed by cwo pachways (Finlayson-Piccs and
Piccs. 1986; Ackinson, 1988):
June, 1990
3-47
DRAFT DO NOT QUOTE OR CITE
-------
(a) a major pathway by OH radical addicion co che aromacic ring;
(b) a minor pachvay by H acom abstraction, either from che aromacic ring (in
che case of benzene) or from an alkyl group (in che case of alkyl-
substituted aromatics). This Laccer pathway leads co che formation of
aromacic aldehydes and is analogous co OH radical reaccion of alkanes.
For example, for coluene:
The produces arising from che OH radical addicion pachway (a) are noc veil
known. Reaccion with 02, again occurring by cvo pachways, is expecced to
predommace:
Pachway (a) yields phenolic compounds and, for coluene, accouncs for -20% of che
overall reaction yield (Atkinson, 1988). The major reaccions involve ring
June, 1990
3-48 DRAFT DO NOT QUOTE OR CITE
-------
cleavage (opening), leading to a variety of bifunccional produces (Finlayson-
Pitts and Pitts, 1986).
For phenolic compounds, in addition to OH radical reaction (proceeding
mainly by initial OH radical addition to the ring), the N03 radical reaction
yielding nitrophenols, appear to be important (Atkinson. 198B):
QH O- OH
3 2 17 Reactions of polvcvclic aromatic compounds CPAH). As was discussed in
Section 2.U, cvo- to four-ring PAH emitted from diesel and spark-ignition engines
are distributed between gas and particle phases. For those PAH present in the
gas phase, the reaction with OH radical is predominant, leading to atmospheric
lifetimes of a few hours or less (see Table 3-1). The nighttime gas-phase
reaction with N203 is of minor significance as a PAH loss process, but (as will
be discussed below) may be important as a formation route of mutagenic mtro-PAH.
In addition, for the PAH containing cyclopenta-fused ring, such as acenaphthene,
acenaphthylene, and acephenanthrylene, the N0S radical reaction can be an
important gas-phase loss process during nighttime hours.
Relatively few product data are available concerning these gas-phase
reactions. It has recently been shovn that, in the presence of N0S, the OH
radical reactions with naphthalene, 1- and 2-methylnapnthalene, acenaphthylene,
biphenyl, fluoranthene, pyrene, and acephenanthrylene lead to the formation of
nitroarenes (Arey at al., 1986; 1989; Sweetman et al. . 1986; Atkinson et al.,
1987; Zielinska et al., 1986; 1988; 1989a). The postulated reaction pathway
+ o-isomer
NOj
June, 1990
3-49
DRAFT DO NOT QUOTE OR CITE
-------
involves initial OH radical addition to the most reactive ring position; for
example, C-3 position for fluoranthene (Pitts ec al., 1985a):
8 9
6
5
3
4
H OH
followed by N02 addition in the C-2 position. Subsequent elimination of water
results in 2-nicrofluoranthene formation:
NO
HO,
H OH
H OH
7 \ 7 \
\=/ NOj \=/ -H20 V=/
jjj ^jTjL Cif
aah
H OHH OH
The analogous reaction sequence for pyrene produces 2-nitropyrene (Pitts
ec al. . 1985a). In contrast, che eleccrophilic nitration reactions of
fluoranthene, and pyrene, involving NO^ ion, produces mainly 3-nitrofluoranthene
from fluoranthene and 1-nitropyrene from pyrene.
The gas-phase reactions of N20s with naphthalene, 1- and 2-
methylnaphthalene, acenaphthene, phenanthrene, anthracene, fluoranthene, and
pyrene yield, in general, che same nitro-PAH isomers as OH radical reaction, but
with different yields. For example, the same 2-nitrofluoranthene is produced
from boch OH radical and N203 gas-phase fluoranthene reactions, but the reaction
with N203 produces a much higher yield. The postulated reaction pathway involves
initial NOj radical addition to the most reactive position, followed by N02
June. 1990
3-50
DRAFT DO NOT QUOTE OR CITE
-------
addicion co che neighboring position and elimination of HN03 (Zielinska ec al.,
Table 3-2 summarizes che mtroarene produce daca for OH radical initiated
and NzOj gas-phase reactions with several PAH studied to date in environmental
chambers (Arey et al., 1989; Zielinska et al., 1990). As will be discussed in
Section 4.3 of this Chapter, generally the same nitro-PAH isomers as those formed
from OH radical and N203 reactions are observed in ambient air samples.
3.2.2 Particulate-Phase Reactions
As discussed in Section 2.2.3 of this Chapter, thousands of organic
compounds present in diesel exhausts are partitioned essentially totally into the
particulate phase under atmospheric conditions. The following chemical processes
are likely to contribute to the degradation of these compounds in the troposphere
(Atkinson, 1988):
• photolysis
• reaction with 03
• reaction with N203 during nighttime hours
® reaction with N02 during nighttime and daytime hours
1986):
June, 1990
-------
lable 3 2
<—«
c
a
P Surmary of the Nitroarcnes Produced from the Gas-Plusu Oil
Radical-Initiated and N20j React Ions and Clectrophlllc Nitration of PAII
PAH
Structure
Position of HltratIon tTlcldl In Reaction with
OH
-*295-
Posltlon of Clectrophlllc
Nitration
Naphthalene
CO
1(0 3X); 2(0 3X)
1-ll/X). 2-(;x)
l->>2-
I-HethyInaphthalene
o5'
5><->6-»3->/--2-»>8-
Total yield (-0.4X)
3->S->4->8--6->;->2-
Total yield (-30X)
4-»2-»5->0-»;--3->6-
2-HethyInaphthalene
OCrCB'
Total yield (-0.2X)
4->|-5->8--3--7-6-
Total yield (-30X)
l-»»8-»4-»6->5-»3->7-
Acenaphthylene
d
o
^ Acenaphthene
o
H
O
e
o
BIphenyl
O
2 3
4- I2X)
5-»3-»4-
Total yield (-0 ?X)
3- (5X)
Hone observed
4-(4OX). 3 (?X). S-(-2XJ
No reaction observed
I-
3-; 5-
2-; 4-
-------
c
a
o
Table 3-2 (continued)
Stannary of the Nitroarenes Produced fran the Gas-Phase OH
Radlcal-tnit lated and N^Oj React Ion* and C lectrophi I ic' Nitration of PAH
m
Stru«;tMrg
Pot It ton of Nitration (Tteld) In Reaction with
OH
-im-
position of Electrophilie
Nitration
Phenanthrene
Imo Isomers (not
9-nitrophenanthrene)
lot a 1 yield (fO IX)
Four isomers (including
9-nItrophenanthrene)
Total yield («IX)
9-»3-; 2-; I-
9 I
Anthracene
u>
I
LA
u>
Fluoranthene
OCO'
0&
I . 2
Total yield (-0 2X)
2-(3X); 7(0.ISX);
8-1-0.ISX)
I-. 2-
Total yield (
-------
• reaccion with H202
• reaccion vich HN03, HNOz and H2SOt
However. che acmospheric lifetimes of particle-bound organic compounds are
noc well known. This is parcially due co the face chac these chemical processes
are dependent upon the nature of the substrate (see, for example, Behymer and
Hites, 1985; 1988) and that many of the laboratory studies have been done using
atmospherically unrealistic adsorbents, such as glass fiber and Teflon-coated
glass-fiber filters, silica gel, and alumina. The extrapolation of sometimes
contradictory results reported by different Laboratories to atmospherically
realistic conditions presents major problems.
The atmospheric face of particle-bound PAH has received much attention
since their potential toxicity was first realized. In their recent publication,
Behymer and Hites (1988) define two opposite schools of thought on this subject.
One school, which they call the "Fast School," says chat particle-bound PAH
degrade quickly in che acmosphere with lifecimes as short as few hours (see, for
example Kamens et al., 1988; Nielsen 1988; and references ciCed in Behymer and
Hices. 1988). The ocher school, called che "Slow School" (which che auchors
represenc), says chac PAH degrade slowly, if ac all. in che atmosphere and
eventually deposic on soil or water. The "Slow School" conclusion is supported
by the studies of marine and lacustrine sediments (the ultimate environmental
sinks of PAH) which have shown that the relative abundances of PAH. even at the
most remote locations, are similar co chose in combuscion sources and in air
parciculace maccer (Laflamme and Hices, 1978; HiCes ec al., 1980; McVecty and
Hices, 1988).
June, 1990
3-54
DRAFT DO NOT QUOTE OR CITE
-------
3.2.2.1 Photooxidation of particulate PAH. Laboratory studies of photolysis of
PAH adsorbed on 18 different fly ashes, carbon black, silica gel. and alumina
(Behymer and Hites, 1985; 1988) and several coal stack ashes (Yokley et al.,
1986; Dunstan et al. , 1989) showed that the extent of PAH photodegradation
depended very much on the nature of the substrate to which they are adsorbed.
The dominant factor in the stabilization of PAH adsorbed on fly ash was the color
of the fly ash, which is related to the amount of black carbon present. It
appeared that PAH were stabilized if the carbon content of the fly ash was
greater than -5%. On black substrates, half-lives of PAH studied were on the
order of several days (Behymer and Hites. 1988).
Similar conclusions were reached from studies of photolysis of PAH adsorbed
on coal stack ashes (Yokley et al., 1986; Dunstan et al.. 1989). The relative
quantity of carbon in coal ash was the main factor determining the extent of
photochemical degradation of pyrene and benzo(a)pyrene adsorbed on the surface.
In addition, in coal ashes that contained a relatively large quantity of iron,
the magnetic particles played a minor role in stabilizing adsorbed pyrene coward
photodegradation (Dunstan et al., 1989).
On the other hand, che environmental chamber studies of Kamens and co-
workers (1988) on the daytime decay of PAH present on residential wood smoke
particles and on gasoline internal combustion emission particles showed PAH half-
lives of the order of 1 hr at moderate humidities and temperatures. At very low*
angle sunlight, very low water-vapor concentration, or very low temperatures, PAH
daytime half-lives increased to a period of days.
In addition, the atmospheric studies of Nielsen (1988), carried out in
rural areas during the winter and early spring when ambient temperatures and
concentrations of N02 and 03 were low, showed evidence for atmospheric decay of
June. 1990
3-55
DRAFT DO NOT QUOTE OR CITE
-------
more reactive PAH, such as benzo(a)pyrene and cyclopenteno(cd)pyrene. Although
no escimacion of chese PAH lifetimes was given, the author concluded that the
decay appeared to be relatively fast.
Due to the limited understanding of the mechanisms of these complex
heterogenous reactions, it is presently impossible to draw any firm conclusion
concerning the photostability of particle-bound PAH in the atmosphere. Since
diesel particulate matter contains a relatively high quantity of elemental carbon
(see Section 2.2.1), it is reasonable to assume that PAH adsorbed onto chese
particles should be relacively scable under scandard atmospheric condicions.
Clearly, additional comprehensive and syscemacic investigation of adsorbed-phase
reactions of PAH is needed.
3.2.2.2 Nitration of particulate PAH under simulated atmospheres. Since 1978
when Pitts and co-workers (Pitts et al., 1978) first demonstrated chat
benzo(a)pyrene (BaP) deposited on glass-fiber filters and exposed co air
containing 0.2S ppm of N02 with traces of HN03 formed nitro-BaP, numerous studies
of che heterogenous nitration reactions of PAH adsorbed on a variety of
substrates in different simulated atmospheres have been carried out (see, for
example. Finlayson-Pitts and Pitts. 1986). It has been shown that the rate of
PAH nitration in a NO^/HNOj system depends on the substrate to which the PAH are
adsorbed (JAger and Hanus, 1980) and on the reactivity of particular PAH
(Nielsen, 1984).
It has also been shown that certain PAH deposited on glass-fiber and
Teflon-impregnated glass-fiber filters react with gaseous N203 yielding their
nitro derivatives (Pitts et al.. 1985b, 1985c). The nitro-PAH isomers formed
from the parent PAH are the same as those formed from electrophilic nitration
June, 1990
3-56
DRAFT DO NOT QUOTE OR CITE
-------
reactions involving N02 ions. Thus, the most abundant isomers formed were 1-
nitropyrene from pyrene, 6-nitro-B(a)P from BaP, 3-nitroperylene from peryLene.
For fluoranthene, 3-, 8-, 7- and 1-nitrofluoranthene isomers were formed in
approximately equal amounts in N205 reactions, whereas the nitrofluoranthene
isomer distribution from electrophilic nitration reaction is 3-> 8-» 7-fc 1-
nitrofluoranthene. However, no 2-nitrofluoranthene (the sole isomer formed from
gas-phase N205 reaction) was produced from this adsorbed-phase reaction (Pitts
ec al. , 1985c). It was speculated chat N,0S becomes ionized on the filter
surface prior to the reaction with fluoranthene. but the resulting N07 ion is not
"free" nitronium ion, i.e., not completely dissociated (Zielinska et al., 1986).
Based on these laboratory studies, it has been proposed that some nitro-PAH
detecced in ambient particles may be formed from the reaction of the parent PAH
with gaseous co-pollutants in the atmosphere, or during the collection of
particulate matter, or both (Pitts et al.. 1978, 198Sa; JAger and Hanus, 1980;
Brorscrom et al. , 1983). However, the extrapolation of the data obtained under
laboratory conditions to the ambient atmosphere requires several major
assumptions. These include, for example, the assumptions that substrate effect,
PAH concentration, the presence of co-pollutants, relative humidity, etc., have
no major impact on PAH nitration reactivities. If so, the available data
indicate that the nitration of particle-bound PAH with HO2/HNO3 and N203 is
probably not significant under atmospheric conditions (Pitts et al. , 1985c).
However, this may not always be the case in airsheds with high N02 and nighttime
N20s concentrations.
The formation of nitro-PAH during sampling may be an important problem for
diesel particulate matter collection, since R02 and HN03 are present in
relatively high concentrations. However, Schuetzle (1983) concluded that the
June, 1990
3-57
DRAFT DO NOT QUOTE OR CITE
-------
artifactual formation of 1-nitropyrene during dilution tube sampling accounts for
less than 10-20% of the total 1-nitropyrene present in diesel particles, if the
sampling time is less than 23 min (one FTP cycle) and if the sampling temperature
is not higher than 63 *C.
The formation of nitroarenes during ambient hi-vol sampling conditions has
been reported to be minimal, at least for the most abundant nitropyrene and
nitrofluoranthene isomers (Arey ec al., 1988a).
3 2 2 3 Ozonolvsis of particulate PAH. Numerous laboratory studies have shown
chat PAH deposited on combustion-generated fine particles and on model substrates
undergo reaction with ozone (see, for example. Katz et al.. 1979; Pitts et al. ,
1980, 1986; Van Vaeck and Van Cauvenberghe, 1984, Finlayson-Pitts and Pitts,
1986. and references therein). The kinetics of the dark reaction of several PAH
deposited on model substrates toward 03 has been shown to be relatively fast
under simulated atmospheric conditions (Katz et al., 1979; Pitts et al., 1980,
1986) Half-lives of the order of one to several hours were reported for the
more reactive PAH, such as BaP, anthracene, and benz(a)anthracene (Katz ec al..
1979)
The reaction of PAH deposited on diesel particles with 1.5 ppm of 03 under
hi-vol sampling conditions was shown to be relatively fast, and the approximate
half-lives of the order of 0.5 to 1 h were reported for most PAH studied (Van
Vaeck and Van Cauwenberghe, 1986). The most reactive PAH included BaP. perylene,
benz(a)anthracene, cyclopenta(cd)pyrene, and benzo(ghi)perylene. The
benzofluoranthene isomers are the least reactive of the PAH studied and
benzo(e)pyrene is less reactive than its isomer BaP. The implications of this
study for the hi-vol sampling of ambient POM are important; reaction of PAH with
June. 1990
3-58
DRAFT DO NOT QUOTE OR CITE
-------
0} could possibly occur under hi-vol sampling conditions during severe
photochemical snog episodes, when the ambient level of 03 is high. However, the
magnitude of this artifact is difficult to assess based on available data.
Exposures of PAH adsorbed on filters and ambient particulate matter to
ambient levels of 03 in an environmental chamber under "passive" conditions, more
nearly resembling atmospheric transport (as opposed to filtration type
experiments analogous to hi*vol sampling of particles), were also carried out
(Pitts et al., 1986). These experiments shoved that significant degradation of
more reactive PAH adsorbed on ambient particulate matter, such as BaP, pyrene,
and benr (aianthracene , may occur azone-polluted atmospheres .
3.3 Physical Removal Processes
3.3.1 Dry Deposition
Dry deposicion is the removal of particles and gases from the atmosphere
through the delivery of mass to the surface by nonprecipitation atmospheric
processes, and the subsequent physical attachment to, or chemical reaction with,
che surfaces, (e.g., vegetation, soil, water, or the built environment) (Dolske
and Gatz. 1985]. It should be noted that the surface icself may be wet or dry;
the term "dry deposition" refers to the mechanism or transport to the surface,
noc to the nature of the surface itself. Dry deposition plays an important role
as a removal mechanism of pollutant in the absence of precipitation. Even in
remote locations such as Slskivic Lake, located on a wilderness island in
northern Lake Superior, the dry deposition of aerosol was found to exceed the wet
removal mechanism by an average ratio of 9:1 (McVeety and Hites, 1988).
June, 1990
3-59
DRAFT DO NOT QUOTE OR CITE
-------
Dry deposition is usually characterized by a deposicion velocicy, Vd, which
is defined as che flux (F) , or deposicion race, of che species S Co Che surface
divided by che concencracion [S] ac some reference height (generally 1 m):
F
The amount of species deposiced per unic area per second in a given
geographical locacion, chac is, che flux, can be eicher calculated, if che
deposition velocicy and the pollutant concencracion are known, or measured
expenaencally. The deposicion velocicy depends on che specific gaseous or
particle species, the properties of the surface co which che species is being
deposited, and che reference heighc. Ic also depends on a microaeceorological
process chac cransporcs che species co che surface (see, for example, Finlayson-
Piccs and Piccs. 1986, and references therein, for more decailed discussion).
For parcicles, che deposicion velocicies depend on che parcicle size,
exhibicing a minimum for parcicles of mean diamecer of -0.1 co -1 urn. Table 3-3
gives che average calculated lifetimes of atmospheric parcicles as a funccion of
their diamecer (from Ackinson. 1988):
[S]
Table 3-3
Average Atmospheric Lifetimes of
Particles Due co Dry Deposition*
Diameter (um)
Lifetime fdavs)
0.002
0.02
0.2
2
0.01
1
10
10
1
20
200
0.01
•From Ackinson (1988); see chis for original references.
June, 1990
3-60
DRAFT DO NOT QUOTE OR CITE
-------
Table 3*4 gives some selected examples of dry deposition velocity for
several inorganic and organic species (from Dolske and Cacz, 1985, with grass as
the surface, except as noted).
Table 3-4
Examples of Dry Deposition Velocities*
PepppUlne Species
Mean Deposition
Velocity (em/sec)
Ozone
Particulate sulfur
Particles:
0.18 um median diameter
0.25 mo median diameter
S02
HN03
Benzeno(ghi)peryleneb
Irideno(l,2,3-cdj pyreneb
0.49
0.17
0.16
0.35
2.1
2.5
0.99
0.99
'From Dolske and Catz, 1985, with grass as the surface, except as noted.
bFrom McVeety and Hites, 1988, with water as the surface.
However, due to the differences in meteorology, nature of surfaces, diurnal
variation, and measurement uncertainties, the reported values for deposition
velocities of a given species can differ by more than an order of magnitude. It
is important to note, that for certain chemicals that have relatively slow gas-
phase chemical removal rates, such as S02 and KN03, dry deposition can be the
major loss process under typical atmospheric conditions.
3.2.2 Vet Deposition
Uet deposition encompasses all processes by which airborne pollutants are
transported to the Earth's surface in aqueous form (i.e. , in rain, snow, or fog).
June, 1990
3-61 DRAFT DO NOT QUOTE OR CITE
-------
The mechanisms of wee removal from the atmosphere may be very different for
particle-associated compounds and for gas-phase compounds. However, since many
organic compounds are partitioned between che aerosol and vapor phase (as
discussed in Section 2.4), both processes of gas and particle scavenging may be
important for a given compound (Ligocki ec al. , 1985a, 1985b; Bidleman, 1988).
When there is no exchange of material between the particulate and dissolved
phases in the rain, che total scavenging of a given compound can be expressed as
(Pankow ec al., 1984):
U - Ug (1-rf) + Up*
where W is the overall scavenging ratio:
[rain,total]
W -
[air.total]
Ug is che gas scavenging racio:
[rain.dissolved]
w» "
(air.gas]
Up is the particle scavenging ratio:
[rain,particulate]
W -
p
[air,particulate j
and 4> is the fraction of the atmospheric concentration which is associated with
particles.
June, 1990
3-62
DRAFT DO NOT QUOTE OR CITE
-------
Parcicle scavenging is a complex process which depends upon che
meteorological conditions in che cloud as well as che size and chemical
composition of the aerosol particles. The simplest model for in-cloud particle
scavenging involves nucleacion scavenging followed by coalescence of the cloud
droplet into raindrops. As many as -10® cloud droplets of -10 diameter must
combine to form one 1 mm raindrop. Hence, scavenging ratios under these
conditions are expected to be of the order 106. However, this process alone
seldom produces precipitation.
In cold clouds, ice crystals grow by vapor accretion and by collection of
supercooled droplets (riming) Scavenging ratios may be considerably lower than
10® under these conditions. In the case of below-cloud scavenging, Wp values
have been estimated to be 10J - 10s for 0.01 -1.0 /jm particles (Slinn et al. ,
1978). From these data one may expect to observe overall particle scavenging
ratios in the range of 10J - 106.
Ligocki and co-workers (Ligocki et al., 1985a, 1985b) measured gas- and
particle - scavenging ratios for a number of organic compounds, including PAH and
their derivatives. Table 3-5 gives mean gas, particle and overall scavenging
ratios for measured neutral organic compounds. It can be seen from this table
that particle scavenging ratios range from 102 to 103, whereas gas scavenging
ratios range from 22 to 10s. Cas scavenging dominates over parcicle scavenging
for compounds of lower molecular weights (mw < 252 for PAH) . Particle scavenging
dominates for the alkanes, which are essentially insoluble in water.
The complexity of liquid-phase inorganic acid formations from gaseous
precursors and the problems of acid rain and acid fog are beyond the scope of
this chapter and are not discussed here (see, for example Finlayson-Pitts and
Pitts, 1986, and references therein for more information).
June, 1990
3-63
DRAFT DO NOT QUOTE OR CITE
-------
Table 3-5
Mean Particle. Gas, and Overall Scavenging Ratios
for Seucral Organic Compounds*
Compound
Mean <3b Mean Mean W Mean W
Dominant
scav mech.d
Toluene
0.0
0.0
22
22±5
g
1,2,4-Trimethylbenzene
0.0
0.0
27±9
27±9
g
Ethylbenzene
0.0
0.0
27±11
27±11
g
ratp-Xylene
0.0
0.0
33±17
33±17
g
o-Xylene
0.0
0.0
35±15
35±15
g
Naphthalene
0.0
0.0
250±73
250±73
g
2-Methvlnaphthalene
0.0
0.0
250±78
250±78
g
I-Methvlnaphthalene
0.0
0.0
330±100
330±100
g
Diethvlphthalate
0.0
NA
20,000
20.000
g
Dibenzofuran
0.008
11,000
930
1000
g
Fluorene
0.009
15,000
1500
1600
g
Phenanthrene + Anthracene
0.011
17,000
3300
3500
g
O-Fluorenone
0.021
15,000
11,000
11,000
g
Methylphenanthrenes
0.027
13,000
2500
2800
g
Fluoranthene
0.053
11,000
6300
6600
g
Pyrene
0.071
9300
5900
6100
g
Eicosane
0.14
AO,000
NA
5600
P
9,10-Anthracenedione
0.21
2400
27,000
22,000
g
Dioctvlphthalate
0.56
36,000
20,000
30,000
P
Docosane
0.61
27,000
NA
17,000
P
Chrvsene
0.71
2600
18,000
7000
g
Benz[a j anthracene
0.75
1300
12,000
4000
g
Benzole jpyrene
0.97
2000
5800
2100
P
Benzo(a]pyrene
0.98
1700
NA
1700
P
Benzo(b+j +k]fluoranthene
0.98
2200
7400
2300
P
Perylene
1.0
1800
NA
1800
P
Tricosane
1.0
22,000
NA
22,000
P
Teracosane
1.0
16,000
NA
16,000
P
Benzo[ghi]perylene
1.0
3100
NA
3100
P
Coronene
1.0
5900
NA
5900
P
'From Ligocki et al. (1985a, 1985b).
b0 - (aerosol]/[vapor + aerosol).
CU - tfpP + W,(l-P).
dg - gas.
p - particle.
NA - not available.
June, 1990
3-64
DRAFT DO NOT QUOTE OR CITE
-------
4.0 ATMOSPHERIC CONCENTRATIONS OF PRIMARY DIESEL EMISSIONS AND THEIR
TRANSFORMATION PRODUCTS
Hose of Che data collected on vehicle emissions are from laboratory studies
using dynamometer/dilution tube measurements. The relevance of chese
measurements to the atmosphere is always a question, since emissions from
vehicles on the road have much higher dilution ratios (-103 versus 10), are
collected at lover temperature, are composed of a large number of individual
vehicle exhausts, have longer residence times (seconds to days, versus -5
seconds) and, as discussed in Section 3, interact with ambient air pollutants.
Pierson and co-workers (Pierson ec al., 1983; Salmeen ec al.. 1985)
conducted field experiments in che Allegheny Mountain Tunnel of the Pennsylvania
Turnoike, co address this problem. They found that che diesel- produced
particulate matter at tunnels was, in general, very similar to that encountered
in dilution-cube scudies with respecc co cocal parciculace raaccer emission races,
percencage excraccables, hydrocarbon molecular weighc distribution, HPLC
profiles, particle size distribution, elemental compositions and extract
mutagenicities. However, chese findings did not preclude the possibility of
substantial differences in detailed chemical compositions. Indeed, the
concentracion of 1-nitropyrene in the extract of parciculace samples collecced
in che Allegheny Mouncain Tunnel was reporced co be lower Chan would be predicced
on che basis of laboratory dilucion cube measuremencs either for diesel or spark-
ignition vehicles (Corse ec al., 1983).
The data on organic compound concentrations in airsheds heavily impacted
by motor vehicle emissions (tunnels, roadsides, etc.) are reviewed below. The
possibilicy of using chese daca co discinguish emissions from different sources
is discussed in Seccion 4.4.
3-65
DRAFT DO NOT QUOTE OR CITE
-------
4.1 Volatile Organic Compounds (VOC) Attributable to Traffic
Individual volatile hydrocarbons and aldehydes were measured along a
section of U.S. Highway 70 near Raleigh, NC (Zveidinger et al., 1988). Traffic
volume during sampling was determined by visual counting (- 1050 ± 10% vehicles/h
in each direction) and was classified into four groups: light-duty, including
gasoline and diesel vehicles through class 2 trucks; heavy-duty gasoline, heavy-
duty diesel; and motorcycles. Typical distributions were 91.5%, 3.2%, 5.1% and
0.2%, respectively.
Table 4-1 lists the mean concentrations from four roadsides for selected
hydrocarbons and aldehydes. expressed in ppbC and as a percent contribution of
individual hydrocarbons and aldehydes to total nownethane hydrocarbons (TNMHC)
and total aldehydes, respectively.
The roadside VOC distribution was compared to dynamometer/dilution tube
results on in-use vehicles, which were weighted in an attempt to reflect the same
model year distribution as observed on the roadway (Sigsby et al., 1987, see also
Table 2-2, Section 2.1.2). The two sets of data were similar in that the
different driving cycles, like the different sampling sites, generally show no
significant differences in the distribution of hydrocarbons or aldehydes on a
percent of total basis. There were, however, differences observed between the
sets of data, particularly for the contribution of combustion products (i.e.,
hydrocarbons below C4 and aldehydes). For the roadside study, ethylene,
formaldehyde, and acetaldehyde were lower, while acetylene was higher than in the
dynamometer study. However, noncatalyst vehicles, which constituted 15% of all
light-duty vehicles in the roadside study, as well as light- and heavy-duty
diesel vehicles and trucks, were not included in the dynamometer study.
June, 1990
3-66
DRAFT DO NOT QUOTE OR CITE
-------
Table 4-1
Concentrations
of Individual Hydrocarbons and
Aldehydes Measured
in the Raleigh (NC)
Roadside Study*
Concentration
Percent
Conraound
foobCl
Contribution®
Hydrocarbons
Ethane
16.3
1.81
Ethylene
64.3
7.15
Acetylene
50.9
5.65
Propane
7.9
0.88
Propylene
22.6
2.51
n-Butane
15.8
1.75
L -Butene
5.7
0.64
n-Pentane
25.4
2.82
Lso-Pentane
53
5.89
Methylcyclopentane
10.4
1.15
Methvlcyclohexane
4.7
0.53
n-Decane
3.0
0.33
Benzene
29
3.23
Toluene
59.3
6.60
ra- and p-Xylenes
53.1
5.90
o-Xylene
12.7
1.41
Ethylbenzene
12.0
1.33
TNMHCb
900
100
Total paraffins
369.2
41.0
Total olefins
164.5
18.2
Total aromatics
252.0
28.0
Total unidentified NMHC
63.6
7.1
Aldehydes
Formaldehyde
6.7
41.05
Acetaldehyde
3.0
18.4
Acrolein
1.2
7.3
Benzaldehyde
2.3
13.88
Total aldehydes
16.38
100
* From Zwevdinger ec al., 1988.
b TNMHC: total nonmechane hydrocarbons.
c Percencs based on ppbC.
d Percent contribution of individual hydrocarbons to TNMHC and of individual
aldehydes to total aldehydes.
June, 1990
3-67
DRAFT DO NOT QUOTE OR CITE
-------
4.2 Polycyclic Aromatic Hydrocarbons
Particulate and vapor phase samples were collected from the traffic passing
through the Baltimore Harbor Tunnel and analyzed for PAH and related compounds
(Benner et al. , 1989). High-volume air samplers equipped with Teflon filters
backed by polyurethane foam (PUF) plugs were used for sample collection. There
was no breakdown of traffic into numbers of diesel- and gasoline-fueled vehicles.
The range of particle-phase PAH concentrations and the mean particle- and
vapor-phase PAH concentrations for 48 samples collected in the Baltimore Tunnel
are tabulated in Table 4-2. The ratios of mean particle-phase PAH concentrations
co chat of benzo(e)pyrene iBeP), which is considered to be a nonreactive PAH. are
also given in this table.
As can be seen from chis cable, alkyl-substituted phenanthrenes in the
Tunnel samples had relatively high concentrations compared with those of the
parent compound. This suggests a significant contribution from diesel vehicle
emissions (particularly diesel-fueled trucks), since extracts of diesel
particulate matter are known to have significant concentrations of methyl and
dimethylphenanthrenes (see Table 2-5, Section 2.2.3, and also Yu and Hites,
1981).
Factor analysis was applied to the Tunnel data in an attempt to identify
factors associated with different types of vehicles, with two factors being
obtained. The alkylated phenanthrenes loaded significantly on factor 1,
suggesting the diesel vehicles as the source of these compounds. Several of the
higher molecular weight PAH loaded onto factor 2, which may be associated with
the contribution of gasoline-fueled emissions in the Tunnel.
June, 1990
3-68
DRAFT DO NOT QUOTE OR CITE
-------
Table 4-2
Particle- and Vapor-Pbase PAH Concentrations for
Baltimore Harbor Tunnel Samples*
Concentration, ng/m3
Range,
Mean,
Mean.
Ratio*
ComDound
Darticles
Darticles
vaDorb
co BeP
Phenanthrene
4.3 -
56
18
132
4.3
Anthracene
0.6 -
12
2.9
18
0.6
3 -Methy1-
3.9 -
58
13.9
70
3.3
phenanthrene
2-Methyl-
5.3 -
74
19
-
4.6
phenanthrene
2-Methylanthracene
0.6 -
12
3.0
5.3
0.7
9-and 4-Methyl
4.7 -
50
12.9
71
3.0
phenanthrene and
4H-cvclopenta(def) -
pnenanthrene
1-Mechvlphenanchrene
2.6 -
43
9.8
43
2.3
2,6 - Dimethyl-
4.7 -
62
14
30
3.4
phenanthrene
2,7-Dimethyl-
3.4 -
38
9.2
16
2.2
phenanthrene
1.3-, 2,10-, 3,9-
9.5 -
119
26
61
6.3
and 3,10-Dimethyl-
phenanthrene
1,6- and 2.9-Di-
4.5 -
63
14
27
3.3
rae thylphenanthrene
1,7-Dimethyl-
3.9 -
41
10.2
20
2.4
phenanthrene
2,3-Dimethyl-phenanthrene
3.5 -
41
9.3
16
2.2
Fluoranchene
6.4 -
69
20
16
4.5
Pyrene
9.7 -
76
27
26
6.3
Benzo(ghi)-fluoranthene
3.2 -
26
9.6
NDC
2.1
Cyclopenta(cd)-pyrene
7.6 -
65
20
ND
4.6
Benz(a)anthracene
1.9 -
29
7.6
ND
1.5
Chrysene/Tri-
2.9 -
47
12
ND
2.4
phenylene
Benzofluoran-
2.2 -
44
10.6
ND
2.1
thenes (b,j,+k)
Benzo(e)pyrene
1.5 -
19
5.0
ND
1
Benzo(a)pyrene
1.3 -
26
5.8
ND
1.1
Indeno[l,2.3-cd]•
0.3 -
15
4.6
ND
0.9
pyrene
Benzo(ghi)perylene
1.8 -
18
8.0
ND
1.6
Coronene
1.0 -
10
4.7
ND
0.9
From Benner et al., 1989.
b Mean concentrations of PAH collected on PUF plug (calculated from data given
in Table III of Benner et al., 1989).
c None detected.
d Mean ratios.-
June, 1990 3-69 DRAFT DO NOT QUOTE OR CITE
-------
Ambient air sampling for PAH was also conducted during a summertime
photochemical air pollution episode in Glendora, CA, at a site situated less than
1 km from the heavily traveled 1-210 freeway and generally downwind of Los
Angeles, therefore affected by motor vehicle emissions (Atkinson et al., 1988).
Samples were collected by means of hi-vol samplers equipped with Teflon-
impregnated glass fiber filters followed by PUF plugs. Table 4-3 shows the
average (from three daytime and three nighttime samples) concentrations of PAH
measured and the ratios of their concentrations to that of BeP. Unfortunately,
no alkylated phenanthrenes were measured.
As can be seen from the comparison of Tables 4-2 and 4-3, the
concentrations of all PAH measured in Glendora were much Lower than those
measured in the Tunnel, as would be expected. However, the ratios of the
concentrations of particle-bound PAH to that of BeP were also different for the
two sites, usually much lower for the Glendora site (with the exception of higher
molecular weight PAH, indeno[ 1,2,3-cdjpyrene , benzo(ghi)perylene, and coronene).
This may indicate either contributions from sources other than motor vehicles in
the Glendora study or PAH photochemical transformations occurring on particles
prior to or during hi-vol sample collections (or both). The generally higher PAH
concentrations during the nighttime sampling periods than during the daytime
sampling periods (Atkinson et al., 1988) seem to support the latter possibility
(however, the influence of meteorology cannot be excluded). This conclusion is
also consistent with high levels of photochemical pollutants observed in
Glendora; for example the daily maxima of ozone concentrations (which occurred
always between 2 and 5 pm) ranged from 160 to 240 ppb throughout the entire nine
days of the study.
June, 1990
3-70
DRAFT DO NOT QUOTE OR CITE
-------
Table 4-3
Average Ambient Concentrations of
PAH Measured in Glendora, CA*
PAH
Total Concn., ng/m3
Particle concn.)
Ratio to
Phenanthreneb
20
Anthraceneb
1.0
-
Fluoranthene
5.6 (0.26)e
0.27
Pyrene
4.1 (0.35)e
0.37
Benzo(ghi)-
fluoranthene
0.26
0.28
Cyclopenca(cd)-
pyrene
0.09
0.1
Benz(a)anthracene
0.2
0.22
Chrysene/Tri-
phenylene
1.0
1.1
Benz o fluo ranthene s
[b.J + kj
1.6
1.7
Benzo(e)pyrene
0.94
1
Benzo(a)pyrene
0.33
0.35
Indeno[l,2.3-cd]•
pyrene
1.6
1.7
Benzo(ghi)perylene
3.8
4.0
Coronene
2.8
3.0
* From Atkinson et al., 1988.
b Fhenanchrene and anthracene were not presenc on filters, only on FUF plugs.
c Fluoranthene and pyrene are distributed between gas and particulate phases;
numbers in parentheses represent particle concentrations.
June, 1990
3-71
DRAFT DO NOT QUOTE OR CITE
-------
4.3 Nitroarene Concentrations in Ambient Air
As discussed in Section 2.2.3, diesel particulate matter contains a variety
of nitroarenes, with 1-nitropyrene being the most abundant among identified
nitro-PAH. The concentration of 1-nitropyrene was measured in the extract of
particulate samples collected at the Allegheny Mountain Tunnel on the
Pennsylvania Turnpike (Gorse et al., 1983). This concentration was 2.1 ppn and
< 5 ppm (by mass) of the extractable material from diesel and spark-ignition
vehicle particulate matter, respectively. These values are much lower than would
be predicted on the basis of laboratory dilution tunnel measurements either for
diesel or for spark-ignition engines. Unfortunately, there are no published data
on other Chan 1-nitropyrene nitroarene concentrations in tunnels or roadsides.
Several nitroarene measurements were conducted in airsheds heavy impacted
by motor vehicle emissions (Arey et al., 1987Atkinson et al., 1988; Zielinska
et al., 1989a, 1989b; Ciccioli et al., 1989). For example, ambient particulate
matter samples were collected at three sites (Claremont, Torrance, and Clendora)
situated in the Los Angeles Basin, with the Claremont and Clendora sites being
-30 km and - 20 km, respectively, northeast and the Torrance site -20 km
southwest of downtown Los Angeles (Arey et al., 1987; Atkinson et al., 1988;
Zielinska et al, 1989a, 1989b). The sampling was conducted during two summertime
periods (Claremont, September 1985 and Clendora, August 1986) and one wintertime
period (Torrance, January-February 1986). Table 4-4 lists the maximum
concentrations of nitropyrene and nitrofluoranthene isomers observed at these
three sites during the daytime and nighttime sampling periods.
As can be seen from this table, 1-nitropyrene (1-NP), the most abundant
nitroarene emitted from diesel engines, is not the most abundant nitroarene
observed in ambient particulate matter collected at three sites heavy impacted
June, 1990
3-72
DRAFT DO NOT QUOTE OR CITE
-------
Table 4-4
The Maximum Concentrations of
Nitrofluoranthene (NF) and Nltropyrene (NF) Isomers Observed
at Three South Coast Air Basin Sampling Sices
Nitroarene,
Claremonc*'b
Collection Period
Glendora6-"
Torrance*-*
2-NF, day
AO
350
410
2-NF, night
1,700
2,000
750
3-NF, day
3
NDf
-3
3-NF, night
-3
ND
70
8-NF, day
2
3
8
8-NF, night
2
4
50
1-NP. day
3
15
60
L-NP, night
10
15
50
2-NP, day
1
14
50
2-NP, night
8
32
60
* From Zielinska ec al., 1989b.
b Daytime sample collected from 1200 to 1800 hr and nighttime sample from 1800
co 2400 hr on 9/13/85.
c From Atkinson ec al., 1988.
d Daytime sample collected from 0800 co 2000 hr on 8/20/86 and nighttime sample
from 2000 to 0800 hr on 8/20-21/86.
* Daytime sample collected from 0500 to 1700 hr on 1/28/86 and nighttime sample
from 1700 to 0500 hr on 1/27-28/B6.
1 ND: none detected.
June, 1990
3-73
DRAFT DO NOT QUOTE OR CITE
-------
by mocor vehicle emissions. Of che cvo nitropyrene isomers present,
2-nicropyrene (2-NP), the main nitropyrene isomer formed from the gas-phase OH
radical initiated reaction with pyrene (see Section 3.2.1), is sometimes more
abundant. 2-Nitrofluoranthene (2-NF) was always the most abundant nitroarene
observed in ambient particulate matter collected at these three sites and this
nitrofluoranthene isomer is not present in diesel and gasoline vehicle emissions
(see Table 2-7, Section 2.2.3, and, for example, Ciccioli et al., 1989).
2-Nitrofluoranthene is the only nitroarene produced from the gas-phase OH radical
initiated and NZ0S reactions with fluoranthene (see Sections 3.2.1 and 3.2.2),
vhereas mainly 3-nitrofluoranthene. followed by lesser amounts of 1-, 7-, and 8-
nitroisomers are present in diesel particulate matter and are produced from the
electrophilic nitration reactions of fluoranthene.
Figure 4-1 shows the comparison of the nitroarenes formed from the OH
radical-initiated reaction of fluoranthene and pyrene in an environmental chamber
(upper trace) with the ambient samples collected at Torrance (lower trace) [from
Arey et al. , 1989). The presence of N203 during the nighttime winter collections
in Torrance was very unlikely, given the high level of NO present at sunset.
More likely a relatively high level of OH radicals was present due to the
measured high concentration of HN02, which photolyzes to yield OH radicals. This
suggests that all isomers observed in Figure A -1 (lower trace), with the
exception of 1-nitropyrene, are the product of the OH radical-initiated reactions
of the parent PAH. Direct emissions may account for the 1-nitropyrene (and 3-
nitrofluoranthene) observed at relatively low levels in these ambient samples
(see Zielinska et al., 1989b, for full discussion of all the mw 247 nitroarenes
observed in ambient particles).
June, 1990
3-74
DRAFT DO NOT QUOTE OR CITE
-------
2-NF
100-
m/z 247
PRODUCTS OF THE
OH RADICAL REACTION
WITH FLUORANTHENE
AND PYRENE
2-NP
8-NF a
I * A
AMBIENT SAMPLE
7-NF
2-NF
m/z 247
2-NP
8-NF
7-NF
22 23 24
RETENTION TIME (min)
Figure 4-1. Mass chromacograos of che molecular ion of the
nicrofluoranchenes (NF) and nicropyrenes (NP) formed from che
gas-phase reaction of fluoranchene pyrene with che OH
radicals (cop) and presenc in aabienc particulate saople
collected at Tnrranp. rA /hnreon)
June, 1990 3-75 DRAFT DO NOT QUOTE OR CITE
-------
Although the reaction with OH radicals is the major atmospheric loss
process for gas-phase fluoranthene and pyrene (see Table 3-1, Section 3.2.1),
evidence for atmospheric formation of 2-nitrofluoranthene from N203 reaction with
fluoranthene has also been reported (Zielinska et al. , 1989b). Since the 2-
nitrofluoranthene/2-nitropyrene yield ratio for N205 reactions, observed from
environmental chamber experiments, is > 100, compared to -10 for the OH radical
reaction (see Table 3-2, Section 3.2.1), the high 2-nitrofluoranthene/2-
nitropyrene concentration ratio in ambient samples suggests a contribution fron
the N20j reaction with fluoranthene. Figure 4-2 shows a comparison of a
wintertime sample collected in Torrance (upper trace) with a summertime sample
collected in Claremont (lower trace). The 2-nitrofluoranthene/2-nitropyrene
ratio reached - 200 for the summer night sample. The N205 concentration was
calculated to be - 5 ppb for this night, which supports the suggested formation
route of 2-nitrofluoranthene via reaction with NzOs (Zielinska et al. , 1989b).
The evidence presented above, as well as the observation that 2-
nitrofluoranthene has been the most abundant mw 247 nitroarene in ambient samples
collected worldwide (Ramdahl et al. , 1986), strongly suggests that the
atmospheric formation from the parent PAH, not the direct automotive emissions,
is the major source of these nitroarenes in ambient air. However, under certain
sampling conditions, when ambient particulate matter is collected very close to
emission sources, the mw 247 nitroarene profile may be different. For example,
in urban samples collected during wintertime rush hours at a central square of
Rome, Italy, at a height of 1.5 m above street level, 2-nitrofluoranthene and 2-
nitropyrene were not observed (Cicciolli et al., 1989).
June, 1990
3-76
DRAFT DO NOT QUOTE OR CUE
-------
lOOr
LU
o
2
<
Q
Z)
CD
<
LlJ
>
LU
a:
0
100
m/z 247
2-NF
WINTER NIGHT
7-NF
2-NP
-NP
SUMMER NIGHT
2-NF
m/z 247
| 1 "NJ* 2-NP
I A A
Figure {*•!.
RELATIVE RETENTION TIME
Mass chroaatograas of the molecular loa of the
tiitrofluoranchenofl (NF) and nicropyrenes (HP) present Ln
ambient particulate samples collected in Torrance, CA (top)
and Claremont, CA (batfoml
June, 1990
3-77
DRAFT DO NOT QUOTE OR CITE
-------
4.4 The Need for Atmospheric Tracers of Diesel Emissions
Recepcor source-apportionment models assist In the identification of the
principal sources of airborne pollutants and in the determination of source
contributions to ambient aerosol mass concentrations (or gas- or particulace-
phase species concentrations, light extinction, etc.). The technique ideally
used would require the measurement at receptor sites of selected "tracers," i.e.,
constituents which can be attributed to specific emission sources. Trace
elements and particle mass are most widely used as tracers. However, emissions
from certain sources are difficult to distinguish by this means. For example,
potassium, which is widely used as a wood smoke tracer, is also abundant in
resuspended soil and cigarette smoke. Due to increasing use of unleaded
gasoline, the ambient concentrations of traditional motor vehicle tracers. Pb and
Br, are diminishing and there is a need to identify alternative tracers for
mobile sources. In addition, some sources of toxic air pollutants do not emit
trace metals.
Unique tracers, which could be used to distinguish diesel emissions from
those of spark-ignition engines, have not yet been identified in diesel exhausts.
However, as has been demonstrated by an experiment conducted in Vienna, Austria
(Horvath et al., 1988), such tracers can be added deliberately. In the Vienna
case, the rare-earch element dysprosium was added in Che form of an organo-
metallic compound to the entire diesel fuel supply in Vienna and the vicinity.
By measuring the amount of this tracer in the atmospheric samples, the
contribution of diesel vehicle emissions to the particulate pollutants in Vienna
was determined (it ranged from - 12 to 33%). This approach, however, is costly
and not always practical.
June, 1990
3-78
DRAFT DO NOT QUOTE OR CITE
-------
A second, much more approximate approach used in the past for estimating
diesel exhaust contributions to ambient aerosol was based on the fact that
diesel- and gasoline-fueled vehicles coexist on the highways. A "highway aerosol
signature" was constructed as an emissions-weighted average of the chemical
composition of the aerosol from gasoline-powered automobiles and trucks, diesel-
powered automobiles and trucks, tire dust, and brake dust (Cass and MacRae,
1983). The lead content of the leaded gasoline burned by the gasoline-powered
vehicles was used to determine the amount of "highway" aerosol.in the ambient
air. The diesel exhaust contribution was estimated in proportion to the relative
contribution of diesel exhaust in che highway emission profile (which was based
on source test data). However, as mentioned above, the lead tracer approach will
not be possible in the near future, as lead is being removed from gasoline. In
addition, this method does not address the emissions from diesel engines used in
railroad locomotives, ships, off-highway construction equipment, etc.
Finally, it has been suggested that, since diesel particulate emissions are
enriched in elemental carbon (EC), the elemental carbon content of an ambient
particle, when scaled in proportion to the EC content of diesel exhaust, places
an upper limit on the amount of diesel exhaust aerosol that can be present in an
ambient sample (Cass, 1990). However, the reported values for elemental carbon
in diesel exhausts range from 10 to 90% (see, for example, Villiams et al.,
1989b), although in most cases they are roughly 70% (see Table 2-3, Section
2.2.1). In addition, other combustion sources may contribute to the EC content
of ambient particles.
Since organic compounds are emitted from all combustion sources, the
potential exists for their use in receptor modeling. Reviews of existing
literature data (see, for example, Daisey et al., 1986) revealed the potential
June, 1990
3-79
DRAFT DO NOT QUOTE OR CITE
-------
usefulness of certain classes of organic compounds in developing "fingerprints"
for specific emission sources. Ac Che same time, ic is clear thac che existing
source emission data are not sufficient for receptor modeling purposes, and much
more experimental data are required to develop consistent emission profiles for
specific emission sources.
The class of organic compounds most suitable for serving as a source tracer
should be:
• emitted in relatively high concentration, to allow small sample
sizes and short sampling times;
• relacively easy co separate from ocher classes of organic compounds;
• relatively easy to identify and quantify on che basis of che
chromatographic and spectral properties of its members;
• chemically scable co assure che same composition in source and
receptor sices, i.e., it should not undergo atmospheric
transformations.
In addition, the composition pattern (fingerprint) for this class should differ
between different emission sources to assist in distinguishing among them.
Polycyclic aromatic hydrocarbons were advocaced as potential tracers of
various cypes of combuscion (Daisy ec al. , 1986). PAH are presenc in all
combuscion sources and cheir relative proportions in emissions from a given
source type frequently vary over several orders of magnitude. In addition, good
sampling and analytical methods already exist for this class of compounds.
However, although che PAH concentrations in motor vehicle emissions were
frequently measured in the pasc, che purpose of chese measurements has been co
decermine an emission rate under a given set of conditions rather than co
establish a PAH "fingerprint" for receptor source modeling. Therefore there is
an apparent lack of compatibility among available motor vehicle PAH emission
profiles.
June, 1990
3-80
DRAFT DO NOT QUOTE OR CITE
-------
Ac the presenc cime, ic is noc possible co make any firm judgmenc as co che
possibilicies for discinguishing diesel emissions from chose of spark-ignicion
engines. However, some PAH compounds appear co be promising. As mencioned in
Seccion 2.2.3 of chis Chapcer, diesel exhausc is enriched in alkyl-subscicuced
PAH, parcicularly alkyl-phenanchrenes, relacive co parenc compounds. Ic is known
chac low-temperacure formacion (pecrogenesis) of PAH produces a mixcure enriched
in alkyl-subscicuced PAH and ocher kinecically favored compounds, whereas high-
cemperacure processes (combuscion and pyrolysis) favor che generacion of
unsubscicuced compounds. In agreemenc wich chis, ic has been reporced chac che
cocal concencracion of alkvl-PAH in diesel emissions increased, as che cylinder
exhausc cemperacure decreased (Jensen and Hices, 1983).
Benner and co-workers (1989) reporced chac che high concencracions of
mechyl- and dimechylphenanchrenes measured in che Balcimore Harbor Tunnel suggesc
che concribucion of diesel emission sources in Che Tunnel (see also Seccion 4.2
of chis Chapcer). However, alkyl subscicuced PAH are also abundanc in coal and
coal-derived macerial (see, for example. Whice, 1983) and shale oil (Garrigues
ec al. , 1987) . More daca on alkylaced PAH concencracions in differenc combuscion
sources are clearly needed.
The mosc imporcanc limicacion for PAH being used as emission markers is
cheir relacively high chemical reaccivicy (see Seccion 4 of chis Chapcer). Thus,
che PAH profile decermined ac che emission source may differ considerably from
che source PAH profile as ic exiscs in che ambienc acmosphere. However, ic has
been suggesced (Miguel and Pereira, 1989) chac some presumably more scable
parcicle-bound PAH, such as benzo(k)fluoranchene, benzo(ghi)perylene, and
ideno(l,2,3*cd)pyrene can be used as cracers of aucomocive emissions ac recepcor
sices which have no ocher major sources of PAH.
June, 1990
3-81
DRAFT DO NOT QUOTE OR CITE
-------
Taking chemical reactivity and ambient abundance into account, alkanes seem
to be more suitable for tracing motor vehicle emissions than PAH. The lower
homologs which exist entirely in the gas-phase react slowly with the OH radicals
and their atmospheric lifetimes are on the order of several days (see Table 3-1,
Section 3.2.1). The higher, mostly particle-associated homologs (C > 20) are
relatively unreactive. The n-alkanes originating from natural sources (e.g.,
plant waxes) could be distinguished from those originating from anthropogenic
sources on the basis of the odd-to-even carbon number preference. Also, the
ratio of normal to branched isomers is lower in emissions from fossil fuel
combustion than for biogenetically derived alkane aerosol (Simoneit. 1984). In
addition, the literature data indicate chat a homologous series of
alkylcyclohexanes (C^H^ with n ranging from n - 16 to 29) are characteristic for
spark-ignition vehicle emissions, but are only trace components of diesel exhaust
(Simoneit, 1984). It has been also suggested that it may be possible to
distinguish emissions from diesel and spark-ignition engines based on the ratios
of methylated isomers to total branched isomers (Boone et al., 1987).
However, the very limited data sets available for source emissions indicate
that the relative concentrations of the particulate C-24 to C-36 alkanes vary
only one order of magnitude among different source types. It is possible that
the inclusion of the semivolatile alkanes would extend the range of relative
concentrations. Clearly, much more measurement of alkanes in source emissions
is needed to allow comparison among sources.
In summary, the existing data indicate that it may be possible to use
organic compound profiles, perhaps in combination with trace element tracers, to
assist in distinguishing among diesel and other particulate pollutant sources.
However, the determination of organic and inorganic compositions of emissions
June, 1990
3-82
DRAFT DO NOT QUOTE OR CITE
-------
from a number of sources, using sampling and analytical methods appropriate for
the purpose of receptor source apportionment modeling, is clearly necessary.
June, 1990
3-83
DRAFT DO NOT QUOTE OR CITE
-------
5.0 MUTAGENICITY OF RESPXEABLE AMBIENT PARTICLES
In 1976, Tokiwa and co-workers reported chat organic extracts of amblenc
particles collected in Japan were active in the Ames Salmonella cyphimuriua
assay, when tested in the presence of homogenized rat liver tissue, colloquially
known as S9 mix (Tokiwa et al., 1976; 1977). This was not surprising, since it
was known then that ambient particulate matter contained promutagenic PAH. Soon
thereafter, a direct mutagenic activity (i.e., in the absence of S9 mix) of
extracts of ambient particles collected in major cities throughout the world was
reporced (see, for example, Pitts et al., 1977; Talcott and tfei, 1977; Tokiwa ec
al. , 1980; Lofroth. 1981; Finlayson-Pitts and Pitts, 1986, and references
therein). It has been shown chat this direct activity was primarily associated
with organic species present in respirable particles of < 2 diameter.
Reports of the presence of direct-acting mutagenic species in extracts of
ambient particulate organic matter (POM) and vehicle emissions (see previous
chapter) resulted in extensive worldwide investigations, which ultimately
resulted in the identification of a number of PAH derivatives, mostly in diesel
particulate extracts (see, for example, Schuetzle, 1983). The use of short-term
bacterial bioassay in conjunction with analytical procedures, so-called "bioassay
directed chemical analysis," proved to be extremely useful in identification of
the key biologically active compounds in complex environmental mixtures (see
Schuetzle and Levcas, 1986, for a recent review). This technique is briefly
described in Section 5.1 below.
The contribution of mutagenic species present in diesel particulate
extracts to the mutagenic activities of ambient particulate extracts has been the
subject of several studies (see, for example, Gibson, 1983; Tokiwa et al., 1983;
Siak et al., 1985). The contribution of atmospherically formed nitroarenes to
June, 1990
3-84
DRAFT DO NOT QUOTE OR CITE
-------
Che mutagenic activities of particulate samples collected in southern California
has been also recently assessed (Axey et al., 1988b). These studies are briefly
discussed in Section 5.2 below.
5.1 Bioassay Directed Chemical Analysis
Host environmental samples are complex mixtures and comprise thousands of
chemical compounds. Identification of the biologically active compounds, often
present in minute quantities, by traditional analytical methods, would present
an enormous if not impossible cask. It became apparent in the late 1970s that
the short-term bioassays could be used in combination with chemical fractionation
co simplify greatly the process of identifying significant mutagens in complex
environmental samples, such as diesel or ambient particulate extracts (Schuetzle
and Lewcas, 1986).
The Ames Salmonella bacterial strains, used with and without S9 mix,
provide information about the general classes of chemicals causing mutagenic
response (e.g., frameshift versus base pair substitutions, promutagen or direct-
acting mutagen). More recently, tester strains chat are sensitive to certain
classes of compounds have been developed. For example, strain TA98NR, developed
by Rosenkranz and co-workers (see, for example, McCoy et al. , 1981), is deficient
in nitroreductase enzymes and therefore gives a reduced response to nitro-PAH.
Strain TA98 in conjunction with strain TA98NR are most frequently used for assays
of environmental samples (see Rosenkranz and Mermelstein, 1983, for more detailed
discussion).
Figure 5-1 illustrates the principles of bioassay-directed chemical
analysis (from Schuetzle and Lewcas, 1986). Proper sampling, storage, and
June, 1990
3-85
DRAFT DO NOT QUOTE OR CITE
-------
Saapllai
Extraction
Preparative
Fractionation
(
Haas/
Mutagenicity
Recovery?
VH
T ««
< Mutagenicity \ No
HigbT J »
" Tee
Level 1 Fractionation
(
Hats/
Mutagenicity
Recovery?
>1
~ Te*
/ Mutagenicity \ No
( High? J >
1
' Tee
Level 2 Fractionation
< Mutagenicity \ No
MigtiT J >
' Tee
Cheaical
>
s
•1
m
t
(Mutagenicity \ No
HlgbT 1 >
Synthesize Selected
laeaers
(
Mutagenicity
HigbT
fercent Contribution
To Total
Saaple Mutagenicity
Percent Contribution
To
Fraction Mutagenicity
Te*
Coapound
Quantitation
Modify
Modify
Procedure
Figure 5-1. Protocol for btoassaydirected chemical analysis (from Schuetzle and Leatas. 1986).
June, 1990 3-86 DRAFT DO NOT QUOTE OR CITE
-------
extraction of environmental samples are crucial parts of the analysis.
Sequential extraction with increasingly polar solvents or binary solvents is most
frequently used to separate organic material from particles (see Section 2.2.3).
The following step, preparative fractionation, achieves crude separation of
extract into several less complicated fractions. The two most widely used
prefractionation techniques are chromatography on an open silica column to
separate groups of compounds on the basis of polarity and separation of compounds
into acidic, basic, and neutral fractions (Schuetzle and Lewtas, 1986). These
fractions contain hundreds of compounds and are still too complicated to be
characterized chemically. However, the bioassay analysis is used to decide which
fractions should be analyzed further.
Normal-phase high performance liquid chromatography (HPLC) is usually
employed as a level 1 fractionation step (see Figure 5-1). The reference
compounds 1-nitronaphthalene and 1,6-pyrenequinone were proposed (see Section
2.2.3) as chemical markers to designate the separation of samples into nonpolar,
moderately polar, and polar fractions. It was found that the nonpolar fractions
accounted for less than 2-3% of the total extract mutagenicity and that the
distribution of mutagenicity between moderately polar and polar fractions was
dependent on the sample origin (Schuetzle et al., 1985).
Although a multitude of compounds were identified or tentatively identified
from the chemical analysis of fractions from level 1 fractionation (see Section
2.2.3), it soon became obvious that these fractions were still too complex to
allow identification of many less abundant chemical mutagens. Further separation
of each fraction into subfractions using level 2 (see Figure 5-1) fractionation
was necessary. This fractionation is usually achieved by employing reversed-
June, 1990
3-87
DRAFT DO NOT QUOTE OR CITE
-------
phase HPLC or normal-phase HPLC with a different solvent system and/or
chromatographic column than that used in level 1 fractionation.
The analytical techniques most frequently used for the characterization of
HPLC fractions include high resolution capillary column gas chromatography (CC)
with selective detectors and/or coupled with mass spectrometry (GC/MS).
Comparison of the results from different mass spectrometric ionization techniques
(i.e., electron impact versus chemical ionization) may be helpful in the
identification of individual compounds. However, very polar or labile compounds
cannot be analyzed with GC/MS techniques due to losses resulting from adsorption,
thermal decomposition and chemical interactions occurring on GC columns. Direct-
insertion probe coupled with high-resolution MS or MS/MS techniques have been
used as screening techniques for tentative identification of polar compounds.
One of the most promising analytical techniques for the analysis of polar PAH
appears to be supercritical fluid chromatography and HPLC coupled with mass
spectrometry (see, for example, Schuetzle et al., 1985, for more detailed
discussion of new analytical techniques for the identification of polar
mutagens).
5.2 Contribution of Nitroarenes to Ambient Mutagenicity
The distribution of direct-acting mutagenicity between the moderately polar
(MP) and polar (P) fractions for particulate samples of different origins is
shown in Figure 5-2 (from Schuetzle et al., 1985). As mentioned above, the
nonpolar fractions account for less than 2-3% of the total mutagenicity of
extracts; thus, they are not shown in this figure.
According to Figure 5-2, for light-duty diesel particulate extracts, most
of the direct-acting mutagenicity (65-75%) is associated with the moderately
June, 1990
3-88
DRAFT DO NOT QUOTE OR CITE
-------
AIROORNE POIYCYCLIC ORGANIC MATTER. MUTAGENIC DERIVATIVES OF PAH
100
Ambient air HO diesel ID diesei Wood smoke
Figure 5-2. Distribution of direct-acting mutagenicity (TA98, -S9) between
moderately polar (MP) and polar (P) fractions of extracts of
particulate natter in ambient air, wood smoke, and in the exhaust
from heavy duty (HD) and light duty (LD) motor vehicles. Adapted
from Schuetzle et al. (198S).
June, 1990
3-89
DRAFT DO NOT QUOTE OR CITE
-------
polar fraction. In contrast, the polar fractions of extracts of ambient
particles, heavy-duty diesel particles, and wood smoke particles are said to
contain more than 65% of the total extract mutagenicity. As mentioned above (see
also, Section 2.2.3), the compounds responsible for the mutagenicity of these
polar fractions have not been yet identified.
Up to 40% of the direct-acting mutagenicity of total extracts of light-duty
diesel particles can be accounted for by six nitroarenes (1-nitropyrene, 3- and
8-nitrofluoranthene, and 1,3-, 1,6-, and 1,8-dinitropyrene), eluting in the
moderately polar fraction (Salmeen et al. , 1984). In contrast, these nitroarenes
accounted for not nore than 3% of the total mutagenic activity of ambient
particulate samples collected at urban and suburban sices (Siak et al., 1985).
The contribution of atmospherically formed nitroarenes, 2-
nitrofluoranthene, and 2-nitropyrene (see Section 3.2), to the mutagenicity of
ambient particulate samples collected in southern California has been recently
assessed (Arey et al., 1988b). This contribution could be compared to that of
1-nitropyrene and 3- and 8-nitrofluoranthene, regarded as direct emissions from
various combustion sources.
Table 5-1 gives the mutagen density (revertants/m1) of eight ambient
samples along with the calculated percent contributions of the measured
nitroarenes to this mutagenicity. The samples were collected in Torrance, CA,
during wintertime and in Claremont, CA, during summertime (see Section 4.4 for
more detailed description of sampling sites and maximum nitroarene
concentrations).
Although 2-nitrofluoranthene was always the most abundant nitroarene
measured in these ambient samples, the high mutagenic activity of 8-
nitrofluoranthene resulted, in one instance, in this much less abundant isomer
June, 1990
3-90
DRAFT DO NOT QUOTE OR CITE
-------
Table 5-1
Contribution of Nitrofluoranthene (NF) Isomers co the
Direct Mutagenicity of Ambient Particulate Extracts*
Collection Collection Mutagen X Contribution to Mutagenicity
Date Time Oensityb
and Location fh) (rev »'] 2-HF 3-NF 8-NF 1-NP 2-NP Total
1/27-28/86 1700-0500 120 2.6 1.8 3.1 0.1 0.8 8.4
Torrance. CA
1/28/86 0500-1700 120 1 4 -e 0.5 0.1 0.7 2.7
Torrance. CA
2/24-25/86 1800-0600 34 3 9 - - 0 2 1 4 S 5
Torrance. CA
2/25/86 0600-1800 73 I 6 - 0 1 0.9 2.6
Torrance. CA
9/14/85 0600-1200 35 1.0 - - 0.2 0.1 1.3
Claremont, CA
9/14/85 1200-1800 15 0.8 • • 0.2 0.1 1.1
Claremont, CA
9/14/85 1800-2400 40 5.2 0.4 0.4 0.2 0.2 6.4
Claremont. CA
9/15/85 0000-0600 20 2.1 - 0.3 0.1 0.2 2.7
Claremont. CA
'From Arey et al.. 1988b.
bTested on strain TA98I-S9).
eNot quantified or only an upper limit determined.
June, 1990
3-91
DRAFT DO NOT QUOTE OR CITE
-------
contributing a greater fraction of the ambient activity than did 2-
nitrofluoranthene. In all remaining cases, the contribution of 2-
nitrofluoranchene to the ambient mutagenicity was higher than that of the other
measured nitroarenes and ranged from -1% co -5%. In contrast, 1-nitropyrene
never contributed more than 0.2%.
Thus, although nitroarenes directly emitted from combustion sources may,
in some cases, contribute significantly to ambient mutagenicity (see, for example
Gibson, 1983), the contribution of atmospherically formed nitroarenes should be
also recognized.
June, 1990
3-92
DRAFT DO NOT QUOTE OR CITE
-------
6.0 SUMMARY
Major research programs were carried ouc in the lace 1970s and early 1980s
co underscand che physical and chemical charaeceriscics of emissions from diesel
engines and che biological effeccs of chese emissions. Alchough new control
technologies are being incroduced inco currently manufactured diesel vehicles,
che effect of chese changes on diesel emissions is likely co be visible in che
fucure. Ac che presenc time, diesel vehicles manufaccured in che lace 1970s and
early 1980s are scill on che road and, in chis sense, daca collected from thac
time period are scill valid.
However, many of chese daca were collected using laboratory dynamometers
with selected new, or ac least "well-tuned" to manufacturers specificacions,
vehicles. The well-controlled conditions of the dynamometer tests have many
benefics but do not necessarily represenc vehicle emissions under real on-road
conditions. The small number of vehicles cesced in che laboratory is not truly
representative of the distribution within the on-road vehicle fleet. Although
several roadway and tunnel emission measurements were performed in the past, che
daca base on mobile sources emission races, necessary co assess che role of
vehicle emissions in air pollution problems, is still not sufficient. More
measurements carried ouc under realiscic on-road condicions are necessary, in
parcicular for gaseous- and particulate-phase organic compounds presenc in
vehicle emissions.
Once released into che atmosphere, diesel emissions are subjecced co
dispersion and cransporc and, at the same time, co chemical and physical
cransformacions into secondary pollutants, which may be more harmful than their
precursors. Thus, a knowledge of diesel emissions at or near their sources is
no longer sufficient to assess fully the impact of these emissions on human
June, 1990
3-93
DRAFT DO NOT QUOTE OR CITE
-------
health and welfare. The understanding of physical and chemical changes chac
primary diesel emissions undergo during their transport through the atmosphere
is equally important. As a result of the last two decades of laboratory and
ambient experiments and computer modeling, a comprehensive set of data now exist
concerning the atmospheric loss processes and transformation of automotive
emissions. However, our knowledge concerning the products of these chemical
transformations is still very limited. Study is required to determine the
products from the OH radical-initiated reactions of the aromatic and aliphatic
hydrocarbons, the major components of automobile emissions. The atmospheric
transformation produces of PAH and their oxygen-, sulfur-, and nitrogen-
containing analogs require study in the gaseous and adsorbed phases. In
particular, the reactions occurring in adsorbed phases on atmospherically
relevant surfaces are poorly understood and require further study. In addition,
gas-to-particle conversion processes and the chemical processes that lead to
aerosol formation should be further investigated.
The quantitation of the contribution of diesel emissions to total ambient
aerosol mass concentration is not possible without developing a specific profile
for diesel emissions, a "fingerprint" which may be used in receptor source
apportionment models. The existing data indicate that it may be possible to use
PAH and/or alkylated PAH, alkanes, and possibly certain unique compounds to
assist in distinguishing between diesel and other pollutant sources. However,
the available data are not adequate for use in receptor modeling and study is
required to determine the profile of diesel emissions, using sampling and
analytical methods appropriate to the purpose of receptor modeling.
June, 1990
3-94
DRAFT DO NOT QUOTE OR CITE
-------
7.0 REFERENCES
Alsberg, T. , U. Stanberg, R. Westerholm, M. Strandell, U. Rannung, A. Sundvall,
L. Romert, V. Bernson, B. ectersson, R. Toftgard, B. Franzen, H. Jansson,
J.A. Guscafsson, K.E. Egeback, and G. Tejle (1985). Chemical and
biological characcerizacion of organic macerial from gasoline exhausc
particles. Environ. Sci. Technol., 19, 43-50.
Arey, J., B. Zielinska, R. Atkinson, A.M. Winer, T. Ramdahl, and J.N. Pitts, Jr.
(1986). The formation of nitro-FAH from the gas-phase reactions of
fluoranthene and pyrene with the OH radical in the presence of NOx.
Acaos. Environ., 20, 2339-2345.
Arey, J., B. Zielinska, R. Atkinson, and A.M. Viner (1987). Polycyclic aromatic
hydrocarbon and nitroarene concentrations in ambient air during a
wintertime high NOx episode in the Los Angeles basin. Acaos. Environ.,
21. 1437-1444.
Arey, J., B. Zielinska, R. Atkinson, and A.M. Viner (1988a). Formation of
nitroarene during ambient high-volume sampling. Environ. Sci. Technol.,
22, 457-462
Arey J., B. Zielinska, W.P. Harger, R. Atkinson, and A.M. Winer (1988b). The
contribution of nitrofluoranthenes and nitropyrenes to the mutagenic
activity of ambient particulate organic matter collected in southern
California, rtutac. Res., 207, 45-51.
Arey, J., B. Zielinska, R. Atkinson, and S.M. Aschmann (1989). Nitroarene
products from the gas-phase reactions of volatile polycyclic aromatic
hydrocarbons with OH radical and N203. Inc. J. Chem. Kin., 21, 775-799.
Atkinson, R. (1988). Atmospheric transformations of automotive emissions. In
Air Pollution, che Aucoaobile and Public Healch, Health Effects Institute,
National Academy Press, Washington, DC.
Atkinson, R. (1990). Gas-phase tropospheric chemistry of organic compounds: a
review. Atmos. Environ., 24A, 1-41.
Atkinson, R., and U.P.L. Carter (1984). Kinetics and mechanisms of the gas-
phase reactions of ozone with organic compounds under atmospheric
conditions. Chem. Rev., 84, 437-470.
Atkinson, R., J. Arey, B. Zielinska, and S.M.Aschmann (1987). Kinetics and
products of the gas-phase reactions of OH radicals and N203 with
naphthalene and biphenyl. Environ. Sci. Technol., 21, 1014-1022.
Atkinson, R., J. Arey, A.M. Winer, and B. Zielinska (1988). A survey of ambient
concentrations of selected polycyclic aromatic hydrocarbons (PAH) at
various locations in California. Final Report, prepared under Contract
No- A5-185-32, for the California Air Resources Board, Sacramento, CA.
June, 1990
3-95
DRAFT DO NOT QUOTE OR CITE
-------
Atkinson, R., J. Arey, B. Zielinska, and S.M. Aschmann (1990). Kinecics and
nicro-produccs of che gas-phase OH and N03 radical-initiated reactions of
naphthalene-d8, fluoranthene-d10 and pyrene. Inc. J. Chea. KLnec., in
press.
Bailey, J.C., B. Schmidl, and M.L. Williams (1990). Speciated hydrocarbon
emissions from vehicles operated over the normal speed range on the road.
ACflios. Environ., 24A, 43-52.
Bayona, J.M., K.E. Markides, and M.L. Lee (1988). Characterization of polar
polycyclic aromatic compounds in a heavy-duty diesel exhaust particulate
by capillary column gas chromatography and high-resolution mass
spectrometry. Environ. Sci. Techno1., 22, 1440-1447.
Behymer, T.D., and R.A. Hites (1985). Photolysis of polycyclic aromatic
hydrocarbons adsorbed on simulated atmospheric particulates. Environ.
Sci. Technol., 19, 1004-1006.
Behymer. T.D., and R.A. Hites (1988). Photolysis of polycyclic aromatic
hydrocarbons adsorbed on fly ash. Environ. Sci. Technol., 22, 1311-1319.
Benner, B.A., G.E. Gordon, and S.A. Wise (1989). Mobile sources of atmospheric
PAH: a roadway tunnel study. Environ. Sci. Technol., 23, 1269-1278.
Bidleman, T.F. (1988). Atmospheric processes. Environ. Sci. Technol., 22, 361-
367.
Bj^rseth, A., and B.S. Olufsen (1983). Long-range transport of polycyclic
aromatic hydrocarbons. In Handbook of Polycyclic Aromatic Hydrocarbons,
A. Bj^rseth, ed. Marcel Dekker, Inc., New York, NY.
Black, F. , and L. High (1979). Methodology for determining particulate and
gaseous diesel hydrocarbon emissions. SAE Technical Paper 790422.
Boone, P.M., and E.S. Macias (1987). Methyl alkanes in atmospheric aerosols.
Environ. Sci. Technol., 21, 903-909.
Brorstrbm, E., P. Grennfelt, and A. Lindskog (1983). The effect of nitrogen
dioxide and ozone on the decomposition of particle-associated polycyclic
aromatic hydrocarbons during sampling from the atmosphere. Aeoos.
Environ., 17, 601-605.
Calvert, J.C., and W.R. Stockwell (1985). Acid generation in the troposphere by
gas-phase chemistry. Environ. Sci. Technol., 17, 428A.
Campbell, R.M., and M.L. Lee (1984). Capillary column gas chromatographic
determination of nitro polycyclic aromatic compounds in particulate
extracts. Anal. Chea., 56, 1026-1030.
Cass, G.R., and G.J. MacRae (1983). Source-receptor reconciliation of routine
air moniotoring data for trace metals: an emissions inventory assisted
approach. Environ. Sci. Technol., 17, 129-139.
June, 1990
3-%
DRAFT DO NOT QUOTE OR CITE
-------
Cass, G.R. (1990). Personal communicacion
Caucreels, W. , and K. Van Cauvenberghe (1978). Experimencs on the distribution
of organic pollutants between airborne particulate matter and the
corresponding gas phase. Aaaos. Environ., 12, 1133-1141.
Ciccioli, A., A. Cecinato, E. Brancaleoni, R. Draisci, and A. Liberti (1989).
Evaluation of nitrated PAH in anthropogenic emission and air samples.
Aerosol Sci. TechnoL., 10, 296-310.
Coutant, R.VJ. , L. Brown, J.C. Chuang, R.H. Riggin, and R.G. Levis (1988). Phase
distribution and artifact formation in ambient air sampling for
polynuclear aromatic hydrocarbons. Atmos. Environ., 22, 403-409.
Daisey, J.M., J.L. Cheney, and P.J. Lioy (1986). Profiles of organic
particulate emissions from air pollution sources: status and needs for
receptor source apportionment modeling. JAPCA, 36, 17-33.
Dolske, D.A. , and D.F. Gatz (1985). A field intercomparison of methods for the
measurement of particle and gas dry deposition. J. Ceophys. Res., 90,
2076-2084.
Dunstan, T.D.J., R.F. Mauldin, Z. Jinxian, A.D. Hipps, E.L. Wehry, and G.
Hamantov (1989). Adsorption and photodegradation of pyrene on magnetic,
carbonaceous, and mineral subfractions of coal stack ash. Environ. Sci.
TechnoL., 23, 303-308.
Dutcher J.S., J.D. Sun. J.A. Lopez, I. Wolf, R.K. Wolf, and R.O.McClean (1984).
Generation and characterization of radiolabeled diesel exhaust. Am. Ind.
Hyg. Assoc. J., 45, 491-498.
Finlayson-Pitts, B.J., and J.N. Pitts, Jr. (1986). Acmospheric Chemistry:
Fundamentals and Experimental Techniques. John Wiley & Sons, New York, NY.
Frey, J.W., and M. Corn (1967). Diesel exhaust particulates. Nature, 216, 615-
616.
Galloway, J.N., and D.M. Whelpdale (1980). An atmospheric sulfur budget for
eastern North America. Acoos. Environ., 14, 409-417.
Garrigues P., E. Parlanti, M. Radke, J. Bellocq, H. Willsch, and M. Ewald
(1987). Identification of alkylphenanthrenes in shale oil and coal by
liquid and capillary gas chromatography and high-resolution
spectrofluorimetry (Shpol'skii effect). J. Chromatog., 395, 217-228.
Gibson, T.L. (1983). Sources of direct-acting nitroarene mutagens in airborne
particulate matter. Mutat. Res., 122, 115-121.
Gorse, R.A., Jr., T.L. Riley, F.C. Ferris, A.M. Pero, and L.M. Skewes (1983).
1-Nitropyrene concentration and bacterial mutagenicity in on-road vehicle
particulate emissions. Environ. Sci. Technol., 17, 198.
June, 1990
3-97
DRAFT DO NOT QUOTE OR CITE
-------
Hampton, C.V., W.R. Pierson, D. Schuetzle, and M. Harvey (1983). Hydrocarbon
gases emlcced from vehicles on Che road. 2. Determination! of emission
races from diesel and spark-ignicion vehicles. Environ. Sci. Technol.,
17, 699-708.
Hare. C.T., and T.M. Baines (1979a). Characterization of particulate and gaseous
emissions from cwo diesel aucomobiles as functions of fuel and driving
cycle. Paper 790424 presented at SAE Automotive Engineering Congress and
Exposicion, Cobo Hall, Decroic, February 26- March 2, 1979.
Hare, C.T., and T.M. Baines (1979b). Characterization of heavy-duty diesel
gaseous and parciculace emissions and effect of fuel composition. Paper
790490 presented at SAE Automotive Engineering Congress and Exposition,
Cobo Hall, Detroit, February 26- March 2, 1979.
Harris, G.V., G.I. Mackay, T.Iguchi, H.I. Schiff, and D. Schuetzle (1987).
Measurements of N02 and HN03 in diesel exhaust gas by tunable diode laser
absorption spectrometry. Environ. Sci. Technol., 21, 299-304.
Heyder, J., J. Gebhart, G. Rudolf, C.F. Schiller, and W. Stahlhofen (1986).
Deposition of parcicles in the human respiratory tract in the size range
0.005 - 15 fim. J. Aerosol Sci., 5, 811-825.
Hites, R.A., R.E. Laflaome, and J.G. Windsor, Jr. (1980). Polycyclic aromatic
hydrocarbons in marine-aquatic sediments: their ubiquity. Advan. Chem.
Series, 185, 289- 311.
Horvath, H., I. Kreiner, C. Norek, 0. Preining, and B. Georgi (1988). Diesel
emissions in Vienna. Acmos. Environ., 22, 1255-1269.
IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Vol. 46,
Lyon, 1989.
Japar, S.M., A.C. Szkarlat, R.A. Gorse, Jr., E.K. Heyerdahl, R.L. Johnson, J.R.
Rau, and J.J. Huntzicker (1984). Comparison of solvent extraction and
thermal-optical carbon analysis mechods: applicacion to diesel vehicle
exhaust aerosol. Environ. Sci. Technol., 18, 231*234.
JAger, J., and V. Hanus (1980). Reaction of solid carrier-adsorbed polycyclic
aromatic hydrocarbons with gaseous lov-concentrated nitrogen dioxide. J.
Hyg. Epidemio. Hlcrobio. Immunol., 24, 1-12.
Jensen, T.E., and R.A. Hites (1963). Aromatic diesel emissions as a function of
engine conditions. Anal. Chem., 55, 594-599.
Kamens, R.M., Z. Guo, J.N.Fulcher, and D. A. Boll (1988). Influence of
humidity, sunlight and temperature on the daytime decay of polyaromatlc
hydrocarbons of polyaromatlc hydrocarbons on atmospheric soot particles.
Environ. Sci. Technol., 22, 103-108.
June, 1990
3-98
DRAFT DO NOT QUOTE OR CITE
-------
Katz, M. , C. Chan, H. Tosine, and T. Sakuma (1979). Relacive races of
photochemical and biological oxidation (in vicro) of polycyclic aromatic
hydrocarbons. In Polynuclear Aromatic Hydrocarbons, P.W. Jones and P.
Leber, eds. Ann Arbor Science Publisher, Ann Arbor, HI.
Kawaoura, K., L-L. Ng, and I.R. Kaplan (198S). Determination of organic acids
(Ct -C10) in the atmosphere, motor exhausts and engine oil. Environ. Sci.
Techno1., 19, 1092-1086.
Kittelson, D.B., and D.F. Dolan (1980). Diesel exhaust aerosol. In (Feneration
of Aerosol and Facilities for Exposures Experiments, K. Willeke, ed. Ann
Arbor Science Publisher, Ann Arbor, MI, pp. 337-359.
Kleindienst, T.E., P.B. Shepson, E.O. Edney, and L.D. Claxton (1985).
Peroxyacetyl nitrate: measurements of its mutagenic activity using the
Salmonella/mammalian microsome reversion assay. Hutat. Res., 157, 123-
128.
Laflamme, R.E., and R.A. Hites (1978). The global distribution of polycyclic
aromatic hydrocarbons in recent sediments. Geochim. Cosmochim. Acta, 42,
289-310.
Lee, F.S.C., and D. Schuetzle (1983). Sampling, extraction and analysis of PAH
from internal combustion engines. In Handbook of Polycyclic Aromatic
Hydrocarbons, A. Bjeorseth, ed.. Marcel Dekker, New York, NY, Chapter II.
Levsen, K. (1988). The analysis of diesel particulate. Fresenius Z. Anal.
Chem., 331, 467-478.
Ligocki. M.P., C. Leuenberger, andJ.F. Pankov (1985a). Trace organic compounds
in rain. II. Gas scavenging of neutral organic compounds. Atmos.
Environ., 19, 1609-1617.
Ligocki, M.P., C. Leuenberger, andJ.F. Pankov (1985b). Trace organic compounds
in rain. III. Particle scavenging of neutral organic compounds. Acnos.
Environ., 19, 1619-1629.
Ligocki, M. , and J.F. Pankow (1989). Measurements of the gas/particle
distributions of atmospheric organic compounds. Environ. Scl. Technol.,
23, 75-83.
Lofroth, G. (1981). Comparison of the mutagenic activity in carbon particulate
matter and in diesel and gasoline engine exhaust. In Shore-Term Bioassays
in the Analysis of Complex Environmental Mixtures, II, M.D. Waters, S.
Sandhu, J. Lewtas-Huisingh, L. Claxton, andS. Nesnow, eds. Plenum Press,
New York, NY.
Loventhal, D.H., and K.A. Rahn (1985). Regional sources of pollution aerosol at
Barrow, Alaska during Winter 1979-80 as deduced from elemental tracers.
Atmos. Environ., 19, 2011-2024.
June, 1990
3-99
DRAFT DO NOT QUOTE OR CITE
-------
McCoy E.C., H.S. Rosenkranz, and R. Mermelstein (1981). Evidence for the
existence of a family of bacterial nitroreductases capable of activating
nitrated polycyclics to mutagens. Environ. Mutagen.. 3, 421-427.
McVeety, B.D., and R.A. Hites (1988). Atmospheric deposition of polycyclic
aromatic hydrocarbons to water sourfaces: a mass balance approach. Acoos.
Environ., 22, 511-536.
Miguel, A.H., and P.A.P. Pereira (1989). Benzeno(k)fluoranthene,
benzo(ghi)perylene and indeno(l,2,3-cd)pyrene: new tracers of automotive
emissions in receptor modeling. Aerosol Sci. Techno1., 10, 292-295.
National Research Council (1982). Diesel Cars. Benefits, Risks and Public
Policy. Final Report of the Diesel Impacts Study Committee, National
Academy Press, Washington, DC.
Nielsen, T. (1984). Reactivity of polycyclic aromatic hydrocarbons coward
nitrating species. Environ. Sci. Techno!., 18, 157-163.
Nielsen, T. (1988). The decay of benzo(a)pyrene and cyclopenteno(cd)pyrene in
the atmosphere. Atoos. Environ., 22. 2249-2254.
Okamoto, U.K., R.A. Corse, Jr., and W.R. Pierson (1983). Nitric acid in diesel
exhaust. JAPCA, 33, 1098-1100.
Pankow, J.F., L.M. Isabelle, and U.E. Asher (1984). Trace organic compounds in
rain-I. Sampler design and analysis by adsorption/thermal desorption
(ATD). Environ. Sci. Technol., 18, 310-316.
Paputa-Peck, M.C. , R.S. Marano, D. Schuetzle, T.L. Riley, C.V. Hampton, T.J.
Prater, L.M. Skewes, T.E. Jensen, P.H. Ruehle, L.C. Bosch, and W.P. Duncan
(1983). Determination of nitrated polynuclear aromatic hydrocarbons in
particulate extracts by capillary gas chromatography with nitrogen
selective detector. Anal. Chem., 55, 1946-1954.
Pierson, W.R., and W.W. Brachaczek (1976). Particulate matter associated with
vehicles on the road. Society of Automotive Engineers Technical Paper
760039.
Pierson, W.R., and W.W. Brachaczek (1983a). Particulate matter associated with
vehicles on the road. II. Aerosol Sci. Technol., 2, 1-40.
Pierson W.R., and W.W. Brachaczek (1983b). Emissions of ammonia and amines from
vehicles on the road. Environ. Sci. Technol., 17, 757-760.
Pierson, W.R., W.W. Brachaczek, R.H. Hammerle, D.E. McKee, and J.W. Butler
(1978). Sulfate emissions from vehicles on the road. JAPCA, 28, 123-132.
Pierson, W.R., W.W. Brachaczek, and D.E. McKee (1979). Sulfate emissions from
catalyst-equipped automobiles on the highway. JAPCA, 29, 255-257.
June, 1990
3-100
DRAFT DO NOT QUOTE OR CITE
-------
Pierson W.R., R.A. Gorse, Jr.. A.C. Szkarlac, W.W. Brachaczek, S.M. Japar, F.
S.-C. Lee, R.B. Zweidinger, and L.D. Claxcon (1983). Mutagenicicy and
chemical characteristics of carbonaceous particulate matter from vehicles
on the road. Environ. Sci. Technol., 17, 31-44.
Pitts, J.N., Jr., D. Grosjean, T.M. Mischke, V.F. Simon, and D. Poole (1977).
Mutagenic activity of airborne organic pollutants. Toxicol. Lett., 1, 65-
70.
Pitts, J.N., Jr., K.A. Van Cauvenberghe, D. Grosjean, J.P. Schmid, D.R. Fitz,
W.R. Besler, Jr., G.B. Knudson, and P.M. Hynds (1978). Atmospheric
reactions of polycyclic aromatic hydrocarbons: facile formation of
mutagenic derivatives. Science, 202, 515-519.
Pitts, J.N., Jr., D.M. Lokensgard, P.S. Ripley, K.A. Van Cauvenberghe, L. Van
Vaek, S.D. Shaffer, A.J. Thill. and W.L. Belser, Jr. (1980).
"Atmospheric" epoxidation of benzo(a)pyrene by ozone: formation of the
metabolite benzo(a)pyrene-4,5-oxide. Science. 210, 1347-1349.
Pitts, J.N., Jr., J.A. Sweetman. B. Zielinska, A.M. Winer. and R. Atkinson
(1985a). Determination of 2-nitrofluoranthene and 2*nitropyrene in
ambient particulate organic matter: evidence for atmospheric reactions.
Acaos. Environ., 19, 1601-1608.
Pitts, J.N., Jr., B. Zielinska, J.A. Sweetman, R. Atkinson, and A.M. Winer
(1985b). Reaction of adsorbed pyrene and perylene with gaseous N203 under
simulated atmospheric conditions. Acrnos. Environ., 19, 911-915.
Pitts, J.N., Jr., J.A. Sweetman, B. Zielinska, R. Atkinson, A.M. Winer, and W.P.
Harger (1985c) . Formation of nitroarenes from the reaction of polycyclic
aromatic hydrocarbons with dinitrogen pentoxide. Environ. Sci. Technol.,
19, 1115-1121.
Pitts, J.N., Jr., H-R. Paur, B. Zielinska. J.Arey, A.M. Winer, T. Ramdahl, and
V. Mejia (1986). Factors influencing the reactivity of polycyclic
aromatic hydrocarbons adsorbed on filters and ambient POM with ozone.
Chemosphere, 15, 675-685.
Ramdahl, T., B. Zielinska, J. Arey, R. Atkinson, A. Winer, and J.N. Pitts, Jr.
(1986). Ubiquitous occurrence of 2-nitrofluoranthene and 2-nitropyrene in
air. Nature, 231, 425-427.
Rosenkranz, H.S., and R. Mermelstein (1983). Mutagenicity and genotoxicity of
nitroarenes: all nitro-containing chemicals were not created equal. Hucac.
Res., 114, 217-267.
Salmeen, I.T., A.M. Pero, R. Zator, D. Schuetzle, and T.L. Riley (1984). Ames
assay chromatograms and the identification of mutagens in diesel particle
extracts. Environ. Sci. Technol., 18, 375-382.
June, 1990
3-101
DRAFT DO NOT QUOTE OR CITE
-------
Salmeen, I.T., R.A. Gorse, and W.R. Pierson <1985). Ames assay chromacograss of
extracts of diesel exhaust particles from heavy-duty trucks on the road
and from passenger cars on a dynamometer. Environ. Sci. Technol., 19,
270-273.
Schroeder, W.H., and D.A. Lane (1988). The fate of toxic airborne pollutants.
Environ. Sci. Technol., 22, 240-246.
Schuetzle, 0. (1983). Sampling of vehicle emissions for chemical analysis and
biological testing. Environ. Health Perspectives, 47, 65-80.
Schuetzle, D. (1988). Organic tracers for source apportionment of combustion
emissions to ambient air particulates. In Proceedings of the Sixth
Symposium on Environmental Analytical Chemistry and Source Apportionment
in Ambient and Indoor Atmospheres, June 27-29, Brigham Young University,
Provo, Utah.
Schuetzle, 0., and J.M. Perez (1981). A CRC cooperative comparison of
extraction and HPLC techniques for diesel particulate emissions. Paper
81-56.4, in Proceedings of the 74th Annual Meeting of the Air Pollution
Control Association. Philadelphia, PA, June 16-21, 1981.
Schuetzle, 0., and J.M. Perez (1983). Factors influencing the emissions of
nitrated-polynuclear aromatic hydrocarbons (nitro-PAH) from diesel
engines. JAPCA, 33, 751-755.
Schuetzle, D. , and J. Levtas (1986). Bioassay-directed chemical analysis in
environmental research. Anal. Cham., 58, 1060A-1075A.
Schuetzle, D., and J.A. Frazier (1986). Factors influencing the emission of
vapor and particulate phase compounds from diesel engines. In Carcinogenic
Effects of Diesel Engine Exhaust, N. Ishinishi, K. Koizumi, R.0.
McClellan, and W. Stdber, eds. Elsevier, Amsterdam, The Netherlands,
pp.41-43.
Schuetzle, D., F.S.C. Lee, T.J. Prater, and S.B. Tejada (1981). The
identification of polynuclear aromatic hydrocarbons (PAH) derivatives in
mutagenic fractions of diesel particulate extracts. Inter. J. Environ.
Anal. Chem. , 9, 93-144.
Schuetzle, D., T.L. Riley, T.J. Prater, T.H. Harvey, and D.F. Hunt (1962).
Analysis of nitrated polycyclic aromatic hydrocarbons in diesel
particulates. Anal. Chem., 54, 265-271.
Schuetzle, D., T.E. Jensen, and J.C. Ball (1985). Polar PAH derivatives in
extracts of particulates: biological characterization and techniques for
chemical analysis. Environ. Intern., 11, 169-181.
Siak J., T.L. Chan, T.L. Gibson, and G.T. Wolff (1985). Contribution to
bacterial mutagenicity from nitro-PAH compounds in ambient aerosols.
Atmos. Environ., 19, 369-376.
June, 1990
3-102
DRAFT DO NOT QUOTE OR CITE
-------
Sigsby, J.E., S. Tejada, W. Ray, J.M. Lang, and J.W. Duncan (1987). Volatile
organic compound emissions from 46 in-use passenger cars. Environ. Scl.
Technol., 21, 466-475.
Simoneic, B.R. (1984). Organic matter of the troposphere••III.
Characterization and sources of petroleum and pyrogenic residues in
aerosols over the western United States. Aeaos. Environ., 18, 51-67.
Slinn, W.G.N. , L. Hasse, B.B. Hicks, A.W. Hogan, D. Lai, P.S. Liss, K.O.
Munnich, G.A.. Sehmel, and 0. Vittori (1978). Some aspects of the transfer
of atmospheric trace constituents past the air-sea interface. Aeaos.
Environ., 12, 2055- 2087.
Smith, L.R. (1989). Characterization of exhaust emissions from trap-equipped
light-duty diesels. Final Report, prepared under Contract No. A5-159-32,
for California Air Resources Board, Sacramento, CA.
Smith, L.R., and J. Paskind (1989). Characterization of exhaust emissions from
trap-equipped light-duty diesels. Paper No. 891972, SAE Technical Paper
Series.
Sonnefeld, W.J., W.H. Zoller, and W.E. May (1983). Dynamic coupled-column
liquid chromatographic determination of ambient temperature vapor
pressures of polynuclear aromatic hydrocarbons. Analy. Chea., 55, 275-
280.
Stockwell, W.R. , and J.G. Calvert (1983). The mechanism of the H0-S02 reaction.
Aeaos. Environ., 17, 2231-2235.
Sweetman. J.A., B. Zielinska, R. Atkinson, T. Ramdahl. A.M. Winer, and J.N.
Pitts, Jr. (1986). A possible formation pathway for the 2-
nitrofluoranthene observed in ambient particulate organic matter. Atmos.
Environ., 20, 235-238.
Szkarlat, A.C., and S.M. Japar (1983). Optical and chemical properties of
particle emissions from on-road vehicles. JAPCA, 33, 592-597.
Talcott, R. , and E. Wei (1977). Airborne mutagens bioassayed in Salmonella
eyphiauriua. J. Natl. Cancer Insc., 58, 449-451.
Thrane, K.E., and A. Mikalsen (1981). High-volume sampling of airborne
polycyclic aromatic hydrocarbons using glass fibre filters and
polyurethane foam. Aeaos. Environ., 15, 909-918.
Tokiwa H. , H. Takeyoshi, K. Morita, K. Takahashi, N. Saruta, and Y. Ohnishi
(1976). Detection of mutagenic activity in urban air pollutants. Hueae.
Res., 38, 351.
Tokiwa H., K. Morita. H. Takeyoshi. K. Takahashi, and Y. Ohnishi (1977).
Detection of mutagenic activity in particulate air pollutants. Mucac.
Res., 48, 237-248.
June, 1990
3-103
DRAFT DO NOT QUOTE OR CITE
-------
Tokiwa, H. , S. Kitamori, K. Takahashi, and Y. Ohnishi (1980). Mutagenic and
chemical assay o£ excracc of airborn particulates, Hucac. Res., 77, 99.
Tokiwa, H. , S. Kitamori, R. Nakagawa, K. Horikawa, and L. Matamala (1983).
Demonstration of a powerful mutagenic dinitropyrene in airborne
particulate matter. Hucac. Res., 121, 107-116.
Tong, H.Y., and F.W. Karasek (1984). Quantitation of polycyclic aromatic
hydrocarbons in diesel exhaust particulate matter by high-performance
liquid chromatography fractionation and high resolution gas
chromatography. Anal. Chea., 56, 2129-2134.
Tong, H.Y., J.A. Sweetoan, F.W. Karasek, E. Jellua, and A.K. Thorsrud (1984).
Quantitative analysis of polycyclic aromatic compounds in diesel exhaust
particulate extracts by combined chromatographic techniques. J.
Chromacogr., 312, 183-202.
Truex, T.J., VJ.R. Pierson, and D.E. McKee (1980). Sulfate in diesel exhaust.
Environ. Sci. Technol., 14, 1121-1124.
Van Vaeck, L. , and K. Van Cauwenberghe (1984). Conversion of polycyclic
aromatic hydrocarbons on diesel particulate matter upon exposure to ppm
levels of ozone. Acmos. Environ., 18. 323-328.
Van Vaeck. L., K. Van Cauwenberghe, and J.Janssens (1984). The gas-particle
distribution of organic aerosol constituents: measurements of the
volatilisation artifact in hi-vol cascade impactor sampling. Acmos.
Environ., 18, 417-430.
White, C.H. (1983). Determination of polycyclic aromatic hydrocarbons in coal-
derived materials. In Handbook of Polycyclic Aromatic Hydrocarbons, A.
Bj^rseth, ed. Marcel Dekker, Inc., New York, NY.
Williams, D.J., J.W. Milne, D.B. Roberts, and M.C. Kimberlee (1989a).
Particulate emissions from "in-use" motor vehicles-I. Spark ignition
vehicles. Acmos. Environ., 23. 2639-2645.
Williams. D.J., J.W. Milne, S.M. Quigley, D.B. Roberts, and M.C. Kimberlee
(1989b). Particulate emissions from "in-use" motor vehicles-II. Diesel
vehicles. Aeoos. Environ., 23, 2647-2661.
Yamasaki. H., K. Kuwata, and H. Miyamoto (1982). Effects of ambient temperature
on aspects of airborne polycyclic aromatic hydrocarbons. Environ. Scl.
Technol., 16, 189-194.
Yamasaki, H. , K. Kuwata, and Y. Kuge (1984). Determination of vapor pressure of
polycyclic aromatic hydrocarbons in the supercooled liquid phase and their
adsoprtion on airborne particulate matter. J. Chem. Soc. Japan, 8, 1324-
1329.
June, 1990
3-104
DRAFT DO NOT QUOTE OR CITE
-------
Yokley, R.A. , A.A. Garrison, E.L. Wehry, and G. Mamancov <1986). Photochemical
transformation of pyrene and benzo(a)pyrene vapor-deposited on eight coal
stack ashes. Environ. Sci. Technol., 20, 86-90.
Yu, M.-L., and R.A. Hires (1981). Identification of organic compounds on diesel
engine soot. Anal. Chea., S3, 951-954.
Zielinska, B. , J. Arey, R. Atkinson, T. Ramdahl, A.M. Winer, and J.N. Pitts, Jr.
(1986). Reaction of dinitrogen pentoxide with fluoranthene. J. Am. Chea.
Soc., 108, 4126-4132.
Zielinska, B. , J. Arey, R. Atkinson, and P.A. McElroy (1988). Nitration of
acephenanthrylene under simulated atmospheric conditions and in solution
and the presence of nitrophenanthrylene(s) in ambient particles. Environ.
Sci. Technol., 22, 1044-1048.
Zielinska, B. . J. Arey, R. Atkinson, and P.A. McElroy (1989a). Formation of
methylnitronaphthalenes from the gas-phase reactions of 1- and 2-
methylnaphthalene with OH radicals and N20s and their occurrence in ambient
air. Environ. Sci. Technol., 23, 723-729.
Zielinska, B.. J. Arey, R. Atkinson, and A.M. Winer (1989b). The nitroarenes of
molecular weight 247 in ambient particulate samples collected in Southern
California. Ataos. Environ., 23, 223-229.
Zielinska B., J. Arey, and R. Atkinson (1990). The atmospheric formation of
nitroarenes and their occurrance in ambienc air. Proceedings of the
Fourth International Conference on N-Substituted Aryl Compounds:
Occurence, Metabolism and Biological Impact of Nitroarenes, Cleveland, OH,
15-19 July 1989, in press.
Zweidinger, R.B., J.E. Sigsby, Jr., S.B. Tejada, F.D. Stump, D.L. Dropkin, F.D.
Ray, and J.W. Duncan (1988). Detailed hydrocarbon and aldehyde mobile
source emissions from roadway studies. Environ. Sci. Technol., 22, 956-
962.
June, 1990
3-105
DRAFT DO NOT QUOTE OR CITE
-------
4. NONCANCER HEALTH EFFECTS OF
DIESEL EXHAUST
The objective of this chapter is to evaluate the potential noncarinogenic health effects
of diesel exhaust. Data pertaining to exposures to whole diesel exhaust will be presented
first, followed by a comparison of the effects of filtered and unfiltered exhaust. Filtered
exhaust consists of the gaseous components of the exhaust without the associated particulate
matter.
4.1. HEALTH EFFECTS OF WHOLE EXHAUST
4.1.1. Human Data
4.1.1.1. Short-Term Exposures
Kahn et al. (1988) reported on the occurrence of 13 cases of acute overexposure to
diesel exhaust. Twelve cases included symptoms of mucous membrane irritation, headache,
and lightheadedness. Eight victims reported nausea, four reported a sensation of unreality,
four reported heartburn, three reported weakness, numbness and tingling in extremities, and
vomiting, two reported chest tightness, and two others reported wheezing (no other data
were given).
4.1.1.1.1. Diesel exhaust odor
Diesel exhaust odor is generally considered a nuisance pollutant; however, at high
intensities it may produce physiological and psychological effects to warrant concern. Strong
unpleasant odors may cause nausea, headache, loss of appetite, and psychological stress.
The intensity of the odor is an exponential function of the concentration such that a tenfold
change in the concentration of the odorants will alter the intensity of the odor by one unit.
Diesel odor is measured in terms of a 12-step (D-units) odor intensity scale (Turk, 1967) or
in terms of a 7-step Total Intensity of Aroma (TIA) scale (CRC, 1979; Levins, 1981).
May 1990
4-1
DRAFT - DO NOT QUOTE OR CITE
-------
Surveys have been undertaken to evaluate the general public's response to diesel
exhaust odor. Hare and Springer (1971; Hare et al., 1974) found that at a D rating of about
2 (TIA = 0.9), about 90 percent of the participants perceived the odor and almost 60
percent found it objectionable. At a D rating of 3.2 (TIA = 1.2), about 95 percent
perceived the odor and 75 percent objected to it, and at a D rating of 5 (TIA = 1.8) about
95 percent objected to it. Linnell and Scott (1962) evaluated the odor threshold for diesel
exhaust and found that a dilution factor of 140 to 475 was needed to reduce the odor
intensity to the threshold level.
4.1.1.1.2. Pulmonary and respiratory effects
Battigelli (1965) conducted tests on 13 human volunteers and found that 15-min to
1-h exposures to diesel exhaust did not have a significant effect on pulmonary resistance.
The exhaust was generated from a single-cylinder, four-cycle, 7-hp engine (no other data
were given). Pulmonary resistance was measured through the simultaneous recording of
esophageal pressure and flow obtained through electrical differentiation of the volume signal
from the spirometer of a plethysmograph. Three dilutions of exhaust were used. The
concentration of the exhaust particles was not reported. The three exhaust dilutions
contained on the average 1.3, 2.8, and 4.2 ppm NO:; 0.2, 0.5, and 1 ppm SO:; <20, 30, and
55 ppm CO; and < 1.0, < 1 to 2, and 1 to 2 ppm total aldehydes, respectively. The two
higher concentrations of exhaust were strongly irritating to the eyes within 10 min.
A number of studies have evaluated changes in pulmonary function occurring over
a workshift in workers occupationally exposed to diesel exhaust (specific time period not
always reported but assumed to be 8 h). In a study of coal miners, Reger et al. (1978) found
that both forced expiratory volume in 1 s (FEVJ and forced vital capacity (FVC) decreased
by 0.05 L in 60 diesel exhaust-exposed miners, an amount not substantially different from
the reductions seen in miners not exposed to diesel exhaust (0.02 and 0.04 L, respectively).
Miners with a history of smoking had more flow rate decrements over the shift than did
nonsmokers. No information was given on pollutant concentrations.
Except for differences related to age, 120 underground iron ore miners exposed to
diesel exhaust exhibited no significant workshift changes in FVC and FEVj when compared
May 1990 4-2 DRAFT - DO NOT QUOTE OR CITE
-------
with 120 matched controls (Jorgensen and Svensson, 1970). Both groups were subdivided
with regard to smoking habit. Among the underground miners there was a higher frequency
of bronchitis, but this was seen primarily in smokers and those in older age groups.
Jorgensen and Svensson estimated that the miners could have been exposed to 0.5 to 1.5
ppm N02 and between 3 and 9 mg/m3 of paniculate matter, with 20 to 30 percent of the
particles <5 /xg. The majority of the panicles were magnetic ore.
In tests on salt miners exposed to diesel exhaust, Gamble et al. (1978, as reponed in
Gamble et al., 1987a) found small changes in pulmonary function over the course of a
workshift. These changes were not statistically significant, and they appeared to be related
to the concentration of NO, (mean value 1.5 ppm) and not to the concentration of
paniculate matter (mean value 0.7 mg/m3).
Ames et al. (1982) compared the pulmonary function of 60 coal miners exposed to
diesel exhaust with a control group of 90 miners not exposed to exhaust. Measurements
were made before and after an 8-h workshift. The workplace air concentrations of the
primary exhaust pollutants were 2.0 mg/m3 particulate matter (panicle size not reponed),
0.2 ppm N03 12 ppm CO, and 0.3 ppm formaldehyde. In terms of FVC, FEV„ and forced
expiratory flow rate at 50 percent FVC (FEFjq) there were no significant differences
between the two groups. However, coal miners exposed to diesel exhaust who were also
smokers or ex-smokers did have significant reductions in FEV, and FEFjo. These reductions
were similar to those seen in the nonexposed group who were also smokers.
In a study of 232 diesel bus garage workers, Gamble et al. (1987a) found some
evidence of health effects associated with exposure to diesel exhaust. The prevalence of
symptoms of cough, labored breathing, itching, burning and watering of eyes, chest tightness,
and wheezing were higher (p <0.05) than in a comparison population of 248 battery
workers. Comparisons were made without adjustment for age or smoking history. Eye
irritation appeared to be the most sensitive indicator of exposure. There was no detectable
association between the daily exposure and changes in pulmonary function. Mean reductions
in FVC, FEV„ peak flows, and flows at 50 percent and 75 percent FVC were not different
from zero. Workers who had respiratory symptoms had slightly greater, but statistically
insignificant reductions in FEVj and FEFsq. Although smokers had the greatest reductions
May 1990 4-3 DRAFT - DO NOT QUOTE OR CITE
-------
in pulmonary function over the workshift, these were not significantly different from those
of nonsmokers or exsmokers. The concentration of particulate matter (particle size not
reported) averaged 0.24 mg/m3 (maximum mean value 0.61 mg/m3). The concentration of
N02 averaged < 1J ppm.
Ulfvarson et al. (1987) evaluated workshift changes in pulmonary function in 15 bus
garage workers, 25 crew members of two car ferries, and in 37 workers on roll-on roll-off
ships. The latter group was exposed primarily to diesel exhaust, while the first two groups
were exposed to both gasoline and diesel exhaust. In the diesel-exposed group, average S-h
exposures were 0.13 to 1.0 mg/m3 of particulate matter, 0.02 to 0.8 mg/m3 (0.016 to
0.64 ppm) NO, 0.06 to 23 mg/m3 (0.03 to 1.2 ppm) NCX, 1.1 to 5.1 mg/m3 (0.96 to 4.45 ppm)
CO, and up to 0.5 mg/m3 (0.4 ppm) formaldehyde. FVC and FEV, were significantly
reduced over the workshift (p <0.01 and p <0.001, respectively). There was no difference
between smokers and nonsmokers. Maximal midexpiratoiy flow, closing volume expressed
as the percentage of expiratory vital capacity, and alveolar plateau gradient were not
affected by the exposure to the diesel exhaust. Pulmonary function returned to normal after
3 d of nonexposure. No exposure-related correlation was found between the observed
effects and concentrations of NO, NOj, formaldehyde, or carbon monoxide; however, it was
suggested that NOz adsorbed onto the exhaust particles may have contributed to the overall
dose of N02 to the lungs. In a related study, in which six volunteers were exposed to diesel
exhaust diluted to a level such that the particle concentration was 0.6 mg/m3 and the NO?
concentration was 3.8 mg/m3 (2.0 ppm), there were no changes in pulmonary function. The
exhaust was generated by a 6-cylinder, 2.38-L diesel vehicle operated at constant speed,
equivalent to 60 km/h, and at about one-half full engine load.
4.1.1.2. Long-Term Exposures
Several epidemiological studies have evaluated the effects of chronic exposure to
diesel exhaust on occupationally exposed workers. Battigelli et al. (1964) measured several
indices of pulmonary function, including vital capacity, FEV„ peak flow, nitrogen washout,
and diffusion capacity in 210 locomotive repairmen. When compared with a control
population of comparable age and smoking history (154 railroad yard workers), no significant
May 1990 4-4 DRAFT - DO NOT QUOTE OR CITE
-------
differences were found in the exposed workers (statistical data not given). The frequency
of symptoms of respiratory distress (i.e., dyspnea, cough, sputum) was not significantly
different between the two groups; however, in both groups the majority of complaints came
from smokers. For the exposed group, maximum workplace concentrations of air pollutants
were 1.8 ppm NOj, 1.7 ppm total aldehydes, 0.15 ppm acrolein, 4.0 ppm SOj, and 5.0 ppm
total hydrocarbons. The concentration of paniculate matter was not reported.
Reger et al. (1982) reported a higher incidence of symptoms of cough and phlegm,
fewer symptoms of moderate to severe dyspnea, and generally lower pulmonary function
(lower FVC, FEV„ and flow rates at lower lung volumes) in underground coal miners
exposed to diesel exhaust than in a group of matched controls (coal miners not exposed to
diesel exhaust). However, there was no significant trend between health characteristics and
years of exposure. When analyzed by age and smoking status, the data showed no consistent
interacting effects. Mean concentrations of NO, at the six mines ranged from 0 to 0.6 ppm
NO, for short-term area samples, 0.13 to 0.28 ppm for full shift personal samples, and 0.03
to 0.80 ppm for full shift area samples. Concentrations of particulate matter averaged 0.93
to 2.73 mg/m3 for personal samples and 0 to 16.1 mg/m3 for area samples. Five hundred and
fifty underground miners and 273 surface workers were included in the study.
Attfield et al. (1982) evaluated 630 potash miners who were exposed to diesel exhaust
and found a higher prevalence of symptoms of cough and phelgm, but no differences in
pulmonary function (FVC and FEV,) when compared with the predicted values derived from
a logistics model based on data from blue collar workers not exposed to dust. Symptom
prevalence was higher for smokers than nonsmokers. Environmental monitoring at the six
mines indicated that NO, concentrations ranged from 0.1 to 3.3 ppm, aldehyde
concentrations from 0.1 to 4.0 ppm, carbon monoxide from 5 to 9 ppm, and total dust
concentrations from 9 to 23 mg/m3. The ratio of total to respirable dust ranged from
2 to 11.
Gamble et al. (1983) evaluated symptoms of respiratory stress and pulmonary
function in 259 salt miners exposed to diesel exhaust. After adjustment for age and smoking,
the salt miners showed no increased prevalence of cough, phelgm, dyspnea, or airway
obstruction when compared with coal miners, potash miners, and blue collar nonminers.
May 1990 4-5 DRAFT - DO NOT QUOTE OR CITE
-------
Mean predicted lung function values for the salt miners were lower than all comparison
populations. Reductions of 2 to 4 percent in FEV, and FVC, 7 to 13 percent in FEFj©, and
18 to 22 percent in FEF75 were reported. These reductions were statistically significant
(p <0.05) except for FVC in blue collar workers. There was, however, no association
between predicted pulmonary function and estimated cumulative exposure to particulate
matter or estimated cumulative exposure to NOj. The highest average exposure to
particulate matter was 1.4 mg/m3 (panicle size not reported, measurement includes NaCl).
Mean NOz exposure was 1.3 ppm, with a range of 0.17 to US ppm. In a continuation of
these studies, Gamble and Jones (1983) assessed the effect of exposure level on prevalence
of respiratory symptoms and pulmonary function. The workers were grouped into low,
intermediate, and high exposure categories; the average concentration of paniculate matter
being 0.40, 0.64, and 0.82 mg/m3. respectively. A statistically significant association was
found for the prevalence of phelgm in the miners and exposure to diesel exhaust
(p <0.0001) and a similar, but nonsignificant trend for cough and dyspnea. Changes in
pulmonary function showed no association with exposure. In a comparison with a control
group of nonexposed workers, and after adjustment for age and smoking, the overall
prevalence of cough and phelgm, but not dyspnea, was elevated. FEVt and FVC were
within 4 percent of expected, which was considered to be within the normal range of
variation for a nonexposed population.
In a study of 283 diesel bus garage workers, Gamble et al. (1987b) found some
evidence of chronic respiratory health effects associated with exposure to diesel exhaust.
When compared with a control population (716 blue collar workers), and after indirect
adjustment for age, race, and smoking, the exposed workers had a higher incidence of cough,
phlegm, and wheezing; however, there was no correlation between symptoms and length of
employment. Dyspnea showed a exposure-response trend but no apparent increase in
prevalence. Mean pulmonary function was not reduced in the exposed population when
compared with a control population, but reductions in FEVj of about 13 mL/yr for each year
of employment were seen in workers with 10 or more years of tenure. There was a 2.5
percent incidence of pneumoconiosis in the exposed workers, but data were insufficient for
funher analysis.
May 1990 4-6 DRAFT - DO NOT QUOTE OR CITE
-------
Ames et al. (1984) conducted a 5-yr prospective study of 1118 underground coal
miners to determine whether occupational exposure to diesel exhaust led to chronic
respiratory effects. Changes in pulmonary function and the occurrence of symptoms of
respiratory effects were evaluated in 280 miners exposed to diesel exhaust and in 838 control
miners. Comparisons were controlled for age, years of underground mining, and for smoking
status. In one group of 200 diesel-exposed miners from western states, forced expiratory
flow rate at 50 percent FVC was significantly lower than that in the controls (p <0.001);
however, FVC, FEV„ and chronic respiratory symptoms (i.e., chronic cough, chronic phlegm,
and breathlessness) were not significantly different between the exposed and control groups.
An internal analysis by cumulative years of exposure did not reveal any association between
respiratory symptoms or pulmonary function parameters and exposure level.
In a preliminary study of 129 employees of a bus company, Edling and Axelson (1984)
found a fourfold higher risk of mortality due to cardiovascular disease, even after adjusting
for smoking history and allowing for at least 10 yr of exposure and 15 or more years of
induction-latency time. However, In a later more comprehensive epidemiological study,
Edling et al. (1987) evaluated mortality data covering a 32-yr period for a cohort of 694 bus
garage workers and found no significant differences between observed and expected number
of deaths due to cancer or cardiovascular disease. Information on exposure levels was not
reported.
Purdham et al. (1987) evaluated respiratory symptoms and pulmonary function in 17
stevedores employed in car ferry operations and in a control group of 11 onsite office
workers. Twenty-four percent of the exposed group and 36 percent of the controls were
smokers. If a particular symptom was considered to be influenced by smoking, smoking
status was used as a covariate in the logistics regression analysis. The frequency of
respiratory symptoms was not significantly different between the two groups; however,
baseline lung function measurements were significantly different. The stevedores had
significantly lower FEV„ FEV,/FVC, FEFjq, and FEF73 (p <0.021, p <0.023yp <0.001, and
p <0.008, respectively). Purdham et al. (1987) concluded that the lung function
measurements provided evidence for an obstructive ventilatory defect. The stevedores were
exposed to both diesel and gasoline exhaust. Concentrations of S02 were below detection
May 1990 4-7 DRAFT - DO NOT QUOTE OR CITE
-------
limits and only traces of NO: (0.5 ppm) were present in the work environment.
Concentrations of CO exceeded 50 ppm for short periods of time. The stevedores were
exposed to significantly (p <0.04) higher concentrations of particulate matter (0.06 to 1.72
rag/m3; mean 0.50 mg/m3) than the controls (0.13 to 0.58 mg/m3).
4.1.2. Animal Studies
4.1.2.1. Acute Exposures
The acute toxicity of undiluted diesel exhaust to rabbits, guinea pigs, and mice was
assessed by Pattle et al. (1957). The engine used was a single-cylinder, air-cooled diesel
engine with a rating of 3.75 to 6.0 (brake horsepower) at 1200 to 1800 rpm. Four different
engine operating conditions were used, the exposures were for 5-h periods, and four rabbits,
10 guinea pigs, and 40 mice were tested under each exposure condition (no controls were
used). Mortality was assessed up to 7 d after exposure. With the engine operating under
light load, the exhaust was highly irritating but not lethal to the test species and only mild
tracheal and lung damage was observed in the exposed animals. The exhaust contained 74
mg/m3 of particulate matter (particle size not reported), 0.056 percent CO, 23 ppm NO* and
16 ppm aldehydes. Exhaust containing 53 mg/m3 of particulate matter, 0.038 percent CO,
43 ppm NO,, and 6.4 ppm aldehydes resulted in low mortality rates (mostly below 10
percent) and moderate lung damage. Exhaust containing 122 mg/m3 of particulate matter,
0.0418 percent CO, 51 ppm NO,, and 6.0 ppm aldehydes produced high mortality rates
(mostly above 50 percent) and severe lung damage. Exhaust containing 1070 mg/m3 of
particulate matter, 0.17 percent CO, 12 ppm NOj, and 154 ppm aldehydes resulted in 100
percent mortality in all three species. High CO levels, which resulted in a
carboxyhemoglobin value of 60 percent in mice and 50 percent in rabbits and guinea pigs,
was considered to be the main cause of death in the latter case. High NOz levels were
considered to be the cause of the lung damage and mortality seen in the other three tests.
Aldehydes and N02 were considered to be the main irritants in the light load test.
May 1990
4-8
DRAFT - DO NOT QUOTE OR CITE
-------
4.1 7 ?.. Short-Term Exposures
EPA-Sponsored Studies. 0*Neil et al. (1980) evaluated pulmonary function in nine
mice exposed to light duty diesel exhaust for 3 mo (the exposure parameters, exhaust
generating system, and concentrations of the exhaust components were not reported). When
compared with five control mice, there were no significant changes in lung volume, vital
capacity, residual volume, or diffusing capacity. However, the exposure did cause an
accumulation of particles in alveolar macrophages and in the interstitium.
EPA conducted several short-term diesel exhaust toxicity studies using a 2.2-L Nissan
CN6-33 engine coupled to a Chrysler Torque-Flite automatic transmission, Model A 727, and
operated under a modified "California Cycle". In one study the effects of diesel exhaust
exposures on the pulmonary function of cats was evaluated (Pepelko et al., 1980a). Eighteen
male cats were exposed 20 h/d for 28 d. After a 14:1 dilution, the air in the exposure
chambers contained an average of 6.4 mg/m3 of paniculate matter, 14.6 ppm CO, 2.13 ppm
NOj, 2.1 ppm SOj, 0.577 mg/m3 sulfates, and 3136 ppm total hydrocarbons. With the
exception of a decrease in maximum expiratory flow rate at 10 percent vital capacity, there
were no significant changes in pulmonary function. Although there was no evidence of
serious lung damage, a predominantly peribronchiolar localization of soot-laden macrophages
within the alveoli resulted in focal pneumonitis or alveolitis.
In another EPA study the effects of a short-term diesel exhaust exposure on arterial
blood gases, pH, blood buffering, body weight changes, lung volumes, and deflation PV
curves of rats was evaluated (Pepelko et alM 1982b). Groups of 15 Sprague-Dawley rats
were exposed to diluted (14:1) irradiated exhaust, to nonirradiated exhaust, or to clean air.
Exposures were 20 h/d for 28 d. The exhaust was irradiated in a chamber illuminated with
artificial sunlight to generate photochemical reactants. Concentrations of the exhaust
components in the exposure chambers are given in Table 4-1. Body weight gains were
significantly (p <0.05) reduced in the exposed animals. Those animals exposed to irradiated
exhaust showed a greater weight reduction than those exposed to nonirradiated exhaust.
Arterial blood gases and standard bicarbonate were unaffected, but arterial blood pH was
significantly (p <0.05) decreased in animals exposed to irradiated exhaust. Residual volume
May 1990 4-9 DRAFT - DO NOT QUOTE OR CITE
-------
and wet lung weight were not altered by either exposure, but vital capacity and total lung
capacity were increased significantly (p <0.05) following exposure to the nonirradiated
exhaust.
In related studies, Wiester et al. (1980) evaluated pulmonary function in 4-d old male
and female Hartley guinea pigs exposed for 28 d to irradiated diesel exhaust (IE). The test
animals (18 or 19 per test group) were exposed for 20 h/d, 7 d/week to the exhaust after
dilution with clean air. The concentration of the exhaust components in the test chambers
are given in Table 4-1. In comparison to controls, animals exposed for 28 d to the irradiated
exhaust exhibited a 35 percent increase in pulmonary airflow resistance and a significant
decrease in heart rate (p <0.002). Dynamic compliance, minute volume, respiratory rate,
and tidal volume were not affected. Histopathological changes in the lungs were evaluated
in test animals exposed for 56 days to irradiated exhaust, to diluted raw exhaust (RE), or to
clean air. The exposure resulted in a significant increase in the ratio of lung to body weight
in both exposed groups (0.82 percent increase for RE and 0.78 percent for IE compared
with 0.68 percent for controls). Heart/body weight ratio was not altered in either group.
In both groups of exposed animals, black-pigmented macrophages accumulated in the alveoli
and black particles were also seen in focal areas of the tracheobronchial tree and in
bronchial and carinal lymph nodes. Hypertrophy of goblet cells in the tracheobronchial tree
and focal hyperplasia of alveolar lining cells were also observed.
GM-Sponsored Studies. In studies conducted on male Fischer 344 rats (16 exposed
and 8 controls), White and Garg (1981) found that short-term exposures (20 h/d, 5.5 d/week
for up to 9 weeks) to diesel exhaust diluted to a particle concentration of 6.0 mg/m3 resulted
in an increase in soot-laden macrophages in the lungs after 6 h, an increase in Type II
pneumocytes after 24 h, and an accumulation of inflammatory cells within the septa after
4 weeks. After 9 weeks of exposure, focal aggregations of soot-laden macrophages
developed near terminal bronchi, along with an influx of polymorphonuclear leukocytes,
Type II cell proliferation, and thickening of the alveolar walls. The affected alveoli occurred
in clusters which for the most part were located near the terminal airways, but occasionally
were focally located in the lung parenchyma. The exhaust was generated with a 5.7-L, four-
May 1990
4-10
DRAFT - DO NOT QUOTE OR CITE
-------
cycle, indirect injection engine operated at a steady speed and load (1350 rpm, 96 N*m,
simulating 65 km/h) (Schreck et al., 1980).
Kaplan et al. (1982) evaluated the effects of subchronic exposure to diesel exhaust
on Fischer 344 male rats, Syrian golden hamsters, and strain A/J mice. The exhaust was
diluted to a particle concentration of 1500 jxg/m3, and the exposures were 20 h/d and
TABLE 4-1. COMPOSITION OF THE EXPOSURE ATMOSPHERES
IN THE EPA STUDIES'
Particles
CO
NO,
NO
so2
Sulfate
THC
Ozone
Exposure
(mg/m3)
(ppm)
(ppm)
(ppm)
(ppm)
(mg/m3)
(ppm)
(ppm)
Control
0.00
2.0
0.07
0.11
0.0
0.0
0.0
0.0
Nomrradiated
632
17.4
22
5.9
2.1
0.57
31.6
0.0
Irradiated
6.83
16.7
2.9
5.0
1.9
0.57
26.1
<0.01
'Mean values.
'Total hydrocarbons.
Source: Wiester et al., 1980.
7 d/week. The hamsters were exposed for 86 d and the rats and mice for 90 d. Exhaust was
generated by a 5.7-L Oldsmobile engine operated continuously at the equivalent of 40 mph.
The concentrations of the gaseous components of the exhaust were not reported. There
were no significant differences in mortality or growth rates between exposed and control
animals. Lung weight relative to body weight of 15 rats exposed for 13 weeks was
significantly higher (p < 0.05) than that for a group of control rats. Histological examination
of tissues of all three species indicated particle accumulation in the lungs and mediastinal
lymph nodes. A small increase occurred in the thickness of the alveolar walls without major
histopathological changes. After 6 mo recovery, considerable clearance of the diesel
particles from the lungs occurred in all three species.
May 1990
4-11
DRAFT - DO NOT QUOTE OR CITE
-------
4.1.23. Chronic Exposures.
4.L23.1. Effects on Growth and Longevity
Changes in growth, body weight, absolute or relative organ weights, and longevity can
be measurable indicators of chronic toxic effects. Such effects have been observed in some,
but not all, of the long-term studies conducted on laboratory animals exposed to diesel
exhaust.
EPA-Sponsored Studies. As part of an EPA-sponsored series of studies, Vinegar et
al. (1980, 1981a, 1981b) investigated the effects of 6-mo exposure to diesel exhaust on adult
male Chinese hamsters. The test animals (8 or 9 per group) were exposed 8 h/d, 7 d/week,
for 6 mo to one of two dilutions of exhaust. One dilution (18:1) resulted in a panicle
concentration of 6 mg/m3 and the other (9:1) resulted in a particle concentration of 12
mg/m3. Neither exposure regime affected body weight gain of the hamsters; however, lung
weights and lung/body weight ratios of the exposed animals were significantly greater than
the controls (p <0.01). The exhaust was generated by a 3.24-L Nissan engine operated in
the Federal Short Cycle. The concentrations of exhaust components in the exposure
chambers over the 6-m period were not reported, but concentrations for 2.5-yr-long tests
conducted with the same exposure facilities and with the same exhaust dilution factors were
reported by Pepelko (1982a). These are given in Table 4-2.
NIOSH-Sponsored Studies. As part of a series of studies evaluating the combined
effects of diesel exhaust and coal dust on laboratory animals. Lewis et al. (1986) exposed rats
and monkeys to diesel exhaust alone and followed changes in body weight over a 2-yr period.
The exhaust was generated by a 425-in3 displacement, four-cycle, water-cooled, naturally
aspirated diesel engine (Caterpillar Model 3304) which developed 100 bhp (brake
horsepower) at 2200 rpm. The engine was operated in an 8-mode (load/haul/dump) duty
cycle typically used in underground coal mining. Idling ccnstitued 60 percent of the cycle.
The exhaust was passed through a Wagner water scrubber and then diluted with air by a
factor of 27:1 before entering the exposure chambers. The particle concentration, measured
with Real Time Aerosol Monitors, was about 2 mg/m3.
May 1990
4-12
DRAFT -- DO NOT QUOTE OR CITE
-------
TABLE 4-2. COMPOSITION OF THE EXPOSURE
ATMOSPHERE IN THE EPA STUDIES *
Time
Particles
CO
NO2
NO
SO2
Acrolein
Form.
THC"
Period
(mg/m3)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
Control'
0.00
22
0.03
0.05
0.03
0.00
0.00
2.82
Weeks 1 to 61
634
20.17
2.68
11.64
2.12
0.025
0.106
7.93
Weeks 61 to 124 11.70
3330
437
1939
5.03
0.034
0.251
11.02
'Mean values.
Total hydrocarbons.
'Weeks 1 to 124.
Source: Pepelko, 1982a.
The size distribution of the particles was determined with a Thermal-Systems Model 3030
electrical aerosol size analyzer. The mass median aerodynamic diameter of the particles was
0.23 and 0.36 jxm by the instrumental and scanning electron microscope techniques,
respectively. The geometric standard deviations were 2.5 and 2.0, respectively. Exposures
were 7 h/d, 5 d/week. The test animals were either male and female Fischer 344 rats (72
or 144 per test) or male cynomolgus monkeys (15 per test). The concentrations of gaseous
components of the exhaust were 11.5 ppm CO; 8.7 ppm NO; 1.5 ppm NO,; 0.81 ppm SO;;
60.2 ppb acrolein; 38.3 ppb formaldehyde; 29.0 jig/m3 sulfates; and 7.5 ppm total
hydrocarbons. After 24 mo of exposure, there were no significant changes in weight gain
for either rats or monkeys.
DOE-Sponsored Studies. Karagianes et al. (1981) reported that Wistar rats (40 per
exposure group and 40 controls) exposed for 20 mo to diluted diesel exhaust (particle
concentration 8.3±2.0 mg/m3, 0.71 /im MMAD (mass median aerodynamic diameter), 2.1
GSD) exhibited no significant changes in body weight or mortality pattern when compared
with controls. Similar results were obtained with animals exposed to coal dust and diesel
exhaust or to coal dust (14.9±6.2 mg/m3). The test animals were exposed for 6 h/d and 5
May 1990 4-13 DRAFT -- DO NOT QUOTE OR CITE
-------
d/week. The exhaust was generated from a 43-hp diesel engine operated under conditions
simulating the use of underground mining vehicles. The concentration of CO was
maintained at 50 ppm by using a 35:1 dilution with air. The NO, concentration ranged from
4 to 6 ppm, and the SOz and aliphatic aldehyde levels were below detection limits (1 ppm).
Lovelace Studies. Studies conducted at the Lovelace Inhalation Toxicology Research
Institute on the chronic exposure of rats to diesel exhaust were reported on by McGellan
et al. (1986). The exhaust was generated by a 5.7-L engine operated under the FTP urban
driving cycle. Male and female F344 rats (initially 288 per test group) were exposed to three
dilutions of diesel exhaust for 7 h/d, 5 d/week for 30 mo. The three exhaust dilutions were
designed to achieve nominal particle concentrations of 0.35, 3.5, and 7 mg/m3 (MMAD 0.23
to 0.26 nm). The concentrations of the other exhaust components in the test chambers are
given in Table 4-3. The exposures had no apparent adverse effects on body weight or
median life span of the rats.
GM-Sponsored Studies. In chronic exposure studies conducted at the General
Motors Research Laboratories (GMR), the general health and body weight gain of male
Fischer 344 rats and Hartley guinea pigs was not affected by 2-yr exposures to diesel exhaust
diluted to achieve particle concentrations of 0.258, 0.796, or 1.533 mg/m3 (0.1 to 0.3 Mm
MMAD) (Schreck et al., 1981). The exposures were for 20 h/day and 5.5 d/week. The
exhaust was generated with a 5.7-L, four-cycle, indirect injection engine operated at a steady
speed and load (1350 rpm, 96 N*m, simulating 65 kpm). Concentrations of the gaseous
components of the exhaust were 3.4,5.3, and 7.9 /ig/m3 CO, and 2.1, 5.0, and 9.2 /ig/m3 NOr
for dilution levels of 240:1, 80:1, and 40:1, respectively. In related studies in which rats were
exposed for up to 36 weeks to 0.25 or 1.5 mg/m3, lung wet weight, normalized to body
weight, was significantly higher than control values (p <0.01) after 12 weeks of exposure
(Misiorowski et al., 1980).
Fraunhofer Institut Studies. Heinrich et al. (1986a, 1986b; see also Stoeber, 1986)
investigated the effects of long-term exposure to diesel exhaust on male and female Syrian
golden hamsters, female SPF Wistar rats, and female NMRI mice. The exhaust was
generated by a 1.6-L engine operated continuously according to the U.S. 72 (FTP)
May 1990
4-14
DRAFT - DO NOT QUOTE OR CITE
-------
TABLE 4-3. COMPOSITION OF THE EXPOSURE
ATMOSPHERES IN THE LOVELACE STUDIES'
Particles
CO
NO,
NO
Ammonia
Hydrocarbons
(mg/mJ)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
7.08
30
0.7
10
0.7
13
3.47
17
0.3
6
0.9
9
035
3
0.1
0.7
1.4
4
0.013b
1
0
0
1
3
'Mean of weekly values for 24 mo of exposure.
'Control.
Source: Henderson et al.. 1988.
test driving cycle. The test fuel had a 0.36 percent sulfur content. The test animals (96 per
exposure group and 96 per control group) were exposed for 19 h/d, 5 d/week. The
maximum exposure period was 120 weeks for hamsters and mice, and 140 weeks for rats.
After a 14:1 dilution with filtered air, the exhaust contained 4.24 mg/m3 of particulate matter
(0.3 tim MMAD), 2.5 ppm CO, 1.12 ppm SO,, 1.5 ppm NO,, and 11.4 ppm NOr Body
weights of hamsters were not affected by exposure to diesel exhaust. Body weights of
exposed rats and mice were reduced about 25 percent compared to controls, panicularly in
the second year of exposure. After 2 yr of exposure, higher mortality rates occurred in mice,
and the wet and dry weights of the lungs of the exposed mice and rats were 2 to 3 times
greater than the controls (p <0.05). Lung weights of the hamsters increased 70 percent for
me same time period.
CCMC Studies. The Committee of Common Market Automobile Constructors
(CCMC) sponsored long-term rodent inhalation studies comparing the effects of diesel
exhaust with those of gasoline exhaust. As part of this study, Fischer 344 rats were exposed
for 2 yr (five 16-h periods per week) to diesel exhaust diluted to achieve particle
May 1990
4-15
DRAFT - DO NOT QUOTE OR CITE
-------
concentrations of 0.7, 2.2, and 6.6 mg/m3 (Brightwell et al., 1986). The exhaust was
generated with a 1.5-L engine operated under the U.S. 72 (FTP) test driving cycle. At the
highest exposure the NO, concentration was 8±1 ppm, and the CO concentration was 32±11
ppm (concentrations of other gaseous pollutants and for the other exposure levels were not
given). At the two higher exposures, significant reductions were seen in the body weights
of the exposed animals when compared with control animals. A significant, exposure-related
increase in lung weight also occurred. Lung weights in males were 2.7 times the control
values; those of females were 3.9 times the control values.
Tshinkhi Studies. Ishinishi et al. (1986b) exposed male and female Fischer 344 rats
to diesel exhaust from light-duty (LD) and heavy-duty (HD) diesel engines. The test animals
(64, 95, or 120 per test group) were exposed for 16 h/d, 6 d/week, for 30 mo. The LD
exhaust was generated from 1.8-L, 4-cylinder (swirl chamber) engines operated at 1200 rpm.
The HD exhaust was generated from 11-L direct injection engines operated at 1700 rpm.
The exhaust was diluted to achieve nominal particle concentrations of 0.1 (LD only), 0.4 (LD
and HD), 1 (LD and HD), 2 (LD and HD), and 4 mg/m3 (HD only). (Concentrations of
other components of the diluted exhaust are given in Table 4-4). An exposure-related
decrease in body weight was seen in both series. The survival rates in all exposed groups
decreased between 24 and 30 mo.
4.1.23.2. Effects on pulmonary junction
The effect of long-term exposure to diesel exhaust on pulmonary function has been
evaluated in laboratory studies on hamsters, rats, guinea pigs, cats, and monkeys.
EPA-Sponsored Studies. Pepelko et al. (1980b, 1981; see also Pepelko, 1982a) and
Moorman et al. (1985) measured lung function in adult male cats chronically exposed to
diesel exhaust. The exhaust was generated with a four-cycle, 3.24-L, naturally aspirated
Nissan engine operated in the 9-mode Federal Short Cycle and with No. 2 diesel fuel
containing 0.15 percent sulfur (Hinners et aU 1980). During the first year of the tests, the
exhaust dilution ratio was 18:1 and the resulting particle concentration was 6.34 mg/m3 (by
mass, 90 percent <1 nm, 50 percent <0.3 jim). The dilution ratio was decreased 10 9:1
during the second year resulting in a particle concentration of 11.70 mg/m3.
May 1990 4-16 DRAFT - DO NOT QUOTE OR CITE
-------
TABLE 4-4. COMPOSITION OF THE EXPOSURE
ATMOSPHERES IN THE ISHINISHI STUDY
Diesel
Panicles
CO
no2
NO
SO,
HTHC
LTHC
Formal
so<
type
(mg/m3)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
(jig/mJ)
LD*
0.003
0.80
0.01
0.03
0.06
2.20
117
0.002
0.41
0.11
1.23
0.08
1.16
038
2.44
221
0.01
18.8
0.41
112
0.26
3.81
1.06
2.93
151
0.03
62.4
1.08
3.96
0.70
9.44
2.42
3.82
2.87
0.07
151
231
7.10
1.41
18.93
4.70
5.49
3.57
0.13
315
HD"
0.002
0.63
0.02
0.04
0.06
2.43
3.50
0.003
0.49
0.46
2.65
0.46
5.71
0.98
4.63
4.27
0.05
62.9
0.96
4.85
1.02
12.11
1.79
7.15
5.16
0.11
111
1.84
7.75
1.68
19.99
182
9.94
5.90
0.18
198
3.72
12.91
3.00
34.45
4.57
15.65
7.62
0.29
361
'Light-duty diesel engine.
"Heavy-duty diesel engine.
Source: Ishinishi el aL 1986b.
Concentrations of other components of the diluted exhausts are given in Table 4-2. The test
animals (21 per test group) were exposed to the exhaust 8 h/d and 7 d/week for up to 2.25
yr. Results were generally negative after the first year of exposure. After the second year,
vital capacity was 369±42 mL in the exposed group ana 410±5S mL in the controi group,
total lung capacity was 428±56 mL in the exposed group and 484±68 mL in the control
group, and diffusing capacity was 0.90±0.27 mL/min/mm Hg in the exposed group and
1.01±0.14 mL/min/mm Hg in the control group. These differences were statistically
significant at p <0.05. Residual volume, maximum expiratory flow, compliance, resistance,
and closing volume were not significantly different.
May 1990 4-17 DRAFT--DO NOT QUOTE OR CITE
-------
Vinegar et al. (1980, 1981a, 1981b) investigated the effects of diesel exhaust on lung
function in adult male Chinese hamsters. The test animals (8 or 9 per group) were exposed
8 h/d, 7 d/week, for 6 mo to either of two dilutions of exhaust. One dilution (18:1) resulted
in a particle concentration of 6.4-mg/m3 and the other (9:1) resulted in a particle
concentration of about 12 mg/m3. The exhaust was generated by a 3.24-L Nissan engine
operated under the Federal Short Cycle. The concentrations of exhaust components in the
exposure chambers over the 6-mo period were not reported, but concentrations for
23-yr-long tests conducted with the same exposure facilities and with the same exhaust
dilution factors were reported by Pepelko (1982a). These are given in Table 4-2. Vital
capacity, vital capacity/lung weight ratio, residual lung volume by water displacement, and
carbon dioxide diffusing capacity decreased significantly (p <0.01) in the animals exposed
to the 6.4 mg/m3 level of particulate matter. Static deflation lung volume, expressed as
percent vital capacity, was increased at 0 and 5 cm H;0. Similar effects were reported to
occur at the 9:1 dilution (Vinegar et al., 1981a), but detailed information was not provided.
NIOSH-Sponsored Studies. As part of a NIOSH-sponsored study evaluating the
combined effects of diesel exhaust and coal dust on laboratory animals, Green et al. (1983)
and Lewis et al. (1986) measured changes in pulmonary function in rats and monkeys
exposed to diluted diesel exhaust in the absence of coal dust. The exhaust was generated
by a 7.0-L, four-cycle Caterpillar engine using fuel with <0.5 percent sulfur. The engine was
equipped with a water scrubber and operated in a repetitive 8-mode coal mining duty cycle.
About 55 percent of the cycle was in the idle mode. Exposures were for 7 h/d, 5 d/week for
up to 2 yr. Male and female Fischer 344 rats (72 or 144 per test group) and male
cynomolgus monkeys (72 per test group) were used in the tests. The exhaust was diluted
with air (27:1) to achieve a particle concentration of 2 mg/m3 (0.23 and 0.36 pm MMAD by
the instrumental and scanning electron microscope techniques; the geometric standard
deviations were 25 and 2.0, respectively). The concentrations of other exhaust pollutants
in the test chamber are given in Table 4-5. Results of the physiological tests conducted on
the rats (10 per test group) exposed to diesel exhaust alone revealed no evidence of
impairment of pulmonary function after 12 mo (Green et al., 1983). After 24 mo of
exposure to the diesel exhaust in the absence of coal dust, there were no significant
May 1990 4-18 DRAFT - DO NOT QUOTE OR CITE
-------
TABLE 4-5. COMPOSITION OF THE EXPOSURE
ATMOSPHERE IN THE LEWIS STUDY
Particles
(mg/mJ)
ol
"5
NO,
(ppm)
NO
(ppm)
SOj
(ppm)
Sulfates
(pg/m1)
Ammonia
(ppm)
Acrolein
(ppm)
Formal,
(ppm)
THC
(ppm)
2.01*
12.7
1.6
9.7
0.83
1.13
control
23
0.04
0.07
0.01
—
0.63
—
—
—
1.95b
11.5
1.5
8.7
0.81
29.0
0.64
0.06
0.038
7.5
control
22
0.06
0.08
**
0.003
0.008
4.1
'Average values for 12 mo (Green et aL, 1983).
"Average values for 24 mo (Lewis et at., 1986).
changes in residual capacity, airway resistance, and alveolar morphometry (Lewis et al.,
1986). In tests on 15 monkeys, forced expiratory flow at 50 percent of vital capacity/forced
vital capacity was significantly lower in animals exposed to diesel exhaust alone for 18 mo
(p = 0.02) compared with controls. Forced expiratory flow at 40 percent of total lung
capacity/total lung capacity was significantly lower in animals exposed for 24 mo (p = 0.05).
Lovelace Studies. Diesel exhaust studies conducted at Lovelace Inhalation
Toxicology Research Institute included an evaluation of the effects of chronic inhalation
exposures on pulmonary function of male and female Fischer 344 rats (Mauderly et al.,
1985a, 1985b; see also McClellan et al., 1986). In these studies the exhaust was generated
by a 5.7-L engine operated under the FTP urban driving cycle. Three exhaust dilutions were
used which were designed to achieve nominal particle concentrations of 0.35, 3.5, and 7
mg/m3 (MMAD 0.23 to 0.26 m). The concentrations of other exhaust components in the
test chambers are given in Table 4-3. Exposures were for 7 h/d, 5 d/week, for up to 30 mo.
After 12 mo of exposure to the highest concentration of diesel exhaust, the exposed rats (n
- 22) had lower total lung capacity (TLC), dynamic lung compliance (DLC), FVC, and CO
diffusing capacity than controls (n = 23). After 30 months of exposure to 7 mg/m1, mean
TLC, DLC, quasistatic chord compliance, and CO diffusing capacity were significantly lower
than control values. Nitrogen washout and percent of FVC expired in 0.1 s were
May 1990
4-19
DRAFT ~ DO NOT QUOTE OR CITE
-------
significantly greater than control values (p <0.05). There was no evidence of airflow
obstruction. The functional alterations were attributed to focal fibrotic, proliferative lesions,
and thickened alveolar membranes which were observed by histological examination.
GM-Sponsored Studies. The effect of diesel exhaust on the pulmonary function of
25 male Fischer 344 rats was investigated in a series of chronic toxicity tests conducted at
the General Motors Research Laboratories. The exposure system, engine operating
conditions, and characterization of the exposure atmospheres were described by Schreck et
al. (1980, 1981). The exhaust was generated with a 5.7-L, four-cycle, indirect injection
engine operated at a steady speed and load (1350 rpm, 96 N»m) simulating 65 km/h. The
pipes in the exhaust delivery system were heated to 100±15*C to simulate typical tailpipe
conditions. The exhaust was then diluted with clean air to achieve a particle concentration
of 1.5 mg/m3 (0.2 #xm MMAD). The diluted exhaust also contained 7.9 /ig/m3 (7 ppm) CO
and 9.2 jxg/m3 NOr Exposures were for 20 h/d and 5.5 d/week for seven exposure periods
up to 612 d. No significant alterations in pulmonary function occurred after 1 y (Gross,
1981a); however, after 612 d of exposure, 6 of the 18 measured parameters showed
statistically significant elevations in the exposed animals when compared with controls.
Functional residual capacity and its component volumes, expiratory reserve and residual
volume, maximum expiratory flow (MEF) at 40 percent FVC, MEF at 20 percent FVC and
FEV01, were significantly greater in the exposed animals (Gross, 1981b).
Fraunhofer Institut Studies. Heinrich et al. (1982) determined the effects of diesel
exhaust on the pulmonary function of female Wistar rats. The exhaust was generated with
a 2.4-L Daimler-Benz engine operated at a constant load of 16 kW and a uniform speed of
2400 rpm. The diesel fuel contained 0.36 percent sulfur. The exhaust was diluted to achieve
a particle concentration of 3.9 mg/m3 (0.1 #im MMAD). The concentrations of the other
components of the diluted exhaust, as well as the pollutant levels in test chambers using
filtered exhaust (discussed in Section 5.2.1) are given in Table 4.6. The test animals (number
per group not given) were exposed for 2 years. The exposure period was 7 to 8 h/d, 5
d/week. When compared with a control group, no significant changes in respiratory rate,
respiratory minute volume, or compliance and resistance occurred in the exposed group.
May 1990
4-20
DRAFT - DO NOT QUOTE OR CITE
-------
Additional studies were conducted by Heinrich et al. (1986, see also Stoeber, 1986);
on the effects of long-term exposure to diesel exhaust on pulmonary function of male and
female Syrian golden hamsters and female SPF Wistar rats. The exhaust was generated by
a 40-kW, 1.6-L engine operated under the U.S. 72 (FTP) test driving cycle. The fuel
contained 0.36 percent sulfur. The exhaust was diluted to achieve a particle concentration
of 4.24 mg/m3 (0.35 nm MMAD). The concentration of the other components of the diluted
exhaust are given in Table 4.7. The animals (96 per test and control groups) were exposed
for 19 h/d, 5 d/week, for a maximum of 120 weeks (hamsters) or 140 weeks (rats). After 1
yr of exposure to the unfiltered diesel exhaust, the hamsters exhibited a significant increase
in airway resistance, and a nonsignificant reduction in lung compliance. For the same time
period, rats exhibited increased lung weights, a significant decrease in dynamic lung
compliance, and a significant increase in airway resistance. These parameters did not show
any further changes after 2 yr of exposure. After 2 yr, dynamic lung compliance was
0.190±0.02 mL/cm H;0 in the exposed group and 0.281±0.09 mL/cm H20 in the controls
(p <0.05), and airway resistance was 0.628±0.10 cm H20/mL/s in the exposed group and
0.544±0.10 cm H20/mL/s in the controls (p <0.05).
CCMC Studies. Brightwell et al. (1986) evaluated the effect of chronic inhalation
of diesel exhaust on the pulmonary function of Syrian hamsters and Fischer 344 rats. The
test animals (54 male and 54 female hamsters and 72 male and 72 female rats per test
group) were exposed for five 16-h periods per week for 2 yr. The exhaust was generated
with a 1.5-L engine operated under the U.S. 72 (FTP) test driving cycle. Three exhaust
dilutions were tested, producing particle concentrations of 0.7, 2.2, and 6.6 mg/m3. At the
highest exposure the NOx concentration was 8±1 ppm, and the CO concentration was 32±11
ppm (concentration of other gaseous pollutants and for the other exhaust dilutions was not
reported). There were no treatment-related changes in pulmonary function in the hamsters.
Rats exposed to the highest concentration of exhaust exhibited changes in respiratory
physiology (data not reported) which were reported to be indicative of restrictive and
obstructive airway disease.
May 1990
4-21
DRAFT - DO NOT QUOTE OR CITE
-------
TABLE 4-6. COMPOSITION OF THE EXPOSURE
ATMOSPHERE IN THE HEINRICH 1982 STUDY*
Panicles CO NOj NO S02 THC"
(mg/mJ) (ppm) (ppm) (ppm) (ppm) (ppm)
Unfiltered 3.9 18-5 1.2 16_5 3.1 93
Filtered - 18.0 1.0 17.2 2.8 7.9
'Mean values.
Total hydrocarbons.
Source: Heinnch et at.. 1982.
TABLE 4-7. COMPOSITION OF THE EXPOSURE
ATMOSPHERE IN THE HEINRICH 1986 STUDY*
Particles CO NO, NO SO, THC
(mg/m3) (ppm) (ppm) (ppm) (ppm) (ppm)
Unfiltered 4.24 12^ 1.5 10.0 1.12 5.5
Filtered - 11.1 1.2 8.7 1.02 5.2
Control — 0.16 — — — 3_5
'Mean values.
Total hydrocarbons.
Source: Heinrich et al., 1986a.
May 1990
4-22
DRAFT - DO NOT QUOTE OR CITE
-------
4.1J233. Lung morphology, biochemistry, and huig lavage analysis
Several series of investigations have examined the morphological, histological, and
histochemical changes occurring in the lungs of animals chronically exposed to diesel
exhausts.
EPA-Sponsored Studies. EPA has conducted chronic diesel exhaust studies on several
species of laboratory animals including hamster, cats, rats, and mice. In these studies the
exhaust was generated with a four-cycle, 3.24-L, naturally aspirated Nissan engine operated
in the 9-mode Federal Short Cycle, with No. 2 diesel fuel containing 0.15 percent sulfur.
The test animals were exposed for 8 h/d or 20 h/d for 7 d/week, and the exhaust was diluted
with clean air at ratios of 18:1 or 9:1 to achieve nominal particle concentrations of 6.0 or 12
mg/m3 (by mass 90 percent < 1 fim. 50 percent <0.3 »im). The composition of the exposure
atmospheres is shown in Table 4-2.
In studies conducted on Chinese hamsters, Pepelko (1982a) found that animals
exposed 8 h/d, 7 d/week, for six mo to panicle levels of 6 or 12 mg/m3 exhibited lungs
characterized by large numbers of black alveolar macrophages in the alveolar spaces;
thickening of the alveolar lining; hyperplasia of Type II alveolar cells; and edema.
In an EPA study conducted on cats, the exposure level (exhaust particle
concentration) was increased from 6.34 mg/m3 during the first year of the tests to 11.70
mg/m3 during the second year (Plopper et al., 1983; Hyde et al., 1985). The 27-mo-long
exposure resulted in an accumulation of exhaust particles in macrophages m the alveolar
sacs and interstitial spaces. Type II cell hyperplasia was seen in the proximal interalveolar
septa (Plopper et al.. 1983). Biochemical analysis indicated increased synthesis of lung
collagen. There was a 2.7-fold increase in the thickness of the peribronchial connective
tissue space in the proximal acinar region. Significant (p <0.05) increases were seen in
numbers of extracellular fibers, interstitial macrophages, and lymphocytes. The
peribronchiolar fibrosis increased in severity even after 6 mo of recovery in clean air (Hyde
et al., 1985).
In related EPA-sponsored studies utilizing the same exposure system as that described
above, Bhatnagar et al. (1980; see also Pepelko, 1982a) evaluated changes in the
biochemistry of lung connective tissue of exhaust-exposed male Sprague-Dawley CD rats and
May 1990 4-23 DRAFT - DO NOT QUOTE OR CITE
-------
A/HEJ mice. The mice were exposed for 8 h/d and 7 d/week, and the rats for 20 h/d and
7 d/week for up to 9 mo, to exhaust containing 6.34 mg/m3 particulate matter. Total lung
protein in rats exposed for 42 d increased about 40 percent over that of controls. In vivo
leucine incorporation was decreased, suggesting a decrease in overall protein synthesis. In
vivo proline incorporation, an estimate of collagen synthesis, was not affected by the
exposure. Prolyl-hydroxylase activity was increased in rats exposed 42 d and in rats exposed
in utero, suggesting increased collagen synthesis. In mice exposed to diesel exhaust for up
to 9 mo, large increases in lung protein content and collagen synthesis were found, but
overall protein synthesis decreased. The increase in collagen synthesis suggested
proliferation of connective tissue and possible fibrosis (Pepelko, 1982a).
NIOSH-Sponsored Studies. The chronic effects of diesel exhaust on laboratory
animals was investigated as as part of a NIOSH-sponsored study evaluating the combined
effects of diesel exhaust and exposure to coal dust (Green et al., 1983; Castranova et ah,
1985; Fedan et ah, 1985; and Lewis et al., 1986). Exposures were for 7 h/d, 5 d/week, for
up to 2 yr. The test animals were either male and female Fischer 344 rats (72 or 144 per
test) or male cynomolgus monkeys (15 per test). The diesel exhaust was generated by a
four-cycle, Caterpillar engine equipped with a water scrubber and operated in a repetitive
8-mode coal mining duty cycle. The concentration of particulate matter was about 2 mg/m3
(0.23 and 0.36 ;im MMAD by the instrumental and scanning electron microscope techniques,
with corresponding GSDs of 2.5 and 2.0, respectively). The concentrations of other exhaust
components, as reported in the separate studies were: 10.5 to 11.5 ppm CO; 7.8 to 9.7 ppm
NO; 1.5 to 1.6 ppm NO:; 0.6 to 0.83 ppm SO:; 60.2 ppb acrolein; and 38.3 ppb
formaldehyde. Histological examination of rat lung tissue showed that particles accumulated
in macrophages in the alveolar sacs after 3 mo of exposure to the diesel exhaust in the
absence of coal dust. Accumulations increased up to 12 mo, at which time the particles were
also being incorporated into the walls of the respiratory bronchioles. No changes were seen
in cytochrome P450 levels in the lung microsomes of exposed rats. The physiological
viability of alveolar macrophages of exhaust-exposed rats was assessed after 2 yr of exposure
by Castranova et al. (1985). The exposure had little effect on the following: viability, cell
May 1990
4-24 DRAFT -- DO NOT QUOTE OR CITE
-------
number, oxygen consumption, membrane integrity, lysosomal enzyme activity, and protein
content of the macrophages. A slight decrease in cell volume, a decrease in
chemiluminescence, and a decrease in ruffling of the cell membrane was observed. The
investigators attributed the latter effects to an overall reduction in macrophage activity.
Histopathological studies of rats exposed for 24 mo revealed multifocal histiocytosis
(pulmonary lipoidosis) occurring primarily subpleurally and composed of collections of
degenerating foamy macrophages, as well as amorphous granular material together with
fibrosis and chronic inflammatory cells. These lesions involved both the interstitium and
alveolar spaces. Hyperplasia of alveolar epithelial Type II cells and degeneration of Type
I cells was also observed. In the exposed monkeys, there was perivascular, peribronchiolar
and alveolar accumulation of particles, but no evidence of fibrosis, focal emphysema, or
inflammation. Exposure of rats and monkeys to coal dust alone (4.98±0.82 mg/m3, total
particles, and 2.00±0.41 mg/m3 respirable particles) resulted in histopathological changes
similar to those seen in the diesel exhaust assays; however, there was less evidence of
lipoidosis. Coal dust increased the number of alveolar macrophages obtained by lavage,
enhanced the secretion of reactive forms of oxygen, and increased cellular surface area. In
contrast, diesel exhaust caused a reduction in the secretion of reactive forms of oxygen and
diminished surface ruffling. In animals exposed to both diesel exhaust and coal dust there
was no evidence of synergistic effects.
DOE-Sponsored Studies. As part of another study examining the combined effects
of diesel exhaust and coal dust, Karagianes et al. (1981) investigated the effects of inhaled
diesel exhaust (in the absence of coal dust) on the lungs of 18-week-old, male Wistar rats.
The test animals (6 per test group) were exposed for 6 h/d, 5 d/week, for 4, 8, 16 or 20 mo.
The exhaust was diluted to achieve a panicle concentration of 8.3±2.0 mg/m3 (0.71 nm
MMAD, 2.1 GSD). Histological examination of lung tissue indicated focal aggregation of
soot-laden alveolar macrophages, alveolar histiocytosis, interstitial fibrosis, and alveolar
emphysema (methodology not described). Lesion severity was related to length of exposure.
Statistical analysis of lesion severity after 20 mo indicated no significant difference (p = 0.05)
between diesel exhaust, diesel exhaust plus coal dust (5.8+3.5 mg/m3), or high concentration
(14.9±6.2 mg/m3) coal dust exposure groups. The exhaust was generated from a 43-bhp
May 1990 4-25 DRAFT - DO NOT QUOTE OR CITE
-------
diesel engine operated under conditions simulating the use of underground mining vehicles.
The concentration of CO was maintained at 50 ppm by using a 35:1 dilution with air. The
concentration of N02 was 4 to 6 ppm, and the levels of S02 and aliphatic aldehydes were
below the detection limit of 1 ppm.
Lovelace Studies. Diesel exhaust studies conducted on rats and mice at the Lovelace
Inhalation Toxicology Research Institute included analyses of biochemical and cytological
changes in lung tissue and bronchoalveolar lavage fluid (BALF). Results of these studies
appeared in several reports (Henderson et al., 1983; McQellan et al., 1986; Mauderly et al.
1987b; Henderson et al., 1988). The exhaust was generated by a 5.7-L engine operated
under the FTP urban driving cycle. Three exhaust dilutions were selected to achieve
nominal panicle concentrations of 0.35, 3.5, and 7 mg/m3 (0.23 to 0.26 ^m MMAD). The
concentrations of other exhaust components in the test chambers are given in Table 4-3.
The test animals were male and female Fischer 344 rats (initially 288 per test group) and
CD-I mice (initially 360 per test group). Exposures were for 7 h/d, and 5 d/week. The test
period was 2 yr in the rat study and 18 mo in the mouse study. At the lowest exposure level,
no biochemical or cytological changes occurred in BALF or in lung tissue in either species.
A chronic inflammatory response was seen at the two higher exposure levels in both species,
as evidenced by increases in inflammatory cells (macrophages and neutrophils), cytoplasmic
and lysosomal enzymes (lactate dehydrogenase, glutathione reductase, and B-glucuronidase),
and protein (hydroxyproline) in BALF. Analysis of lung tissue indicated similar changes in
enzyme levels as well as an increase in total lung collagen content. After 1 yr of exposure
at 7 mg/m3, the lungs of the rats exhibited focal areas of fibrosis, whereas the lungs of the
mice exhibited only a fine fibrillar thickening of a few alveolar septa. Foci of epithelial
metaplasia occurred uniformly in the alveoli and terminal bronchioles of rats exposed to 7.0
mg/m3 and were more scattered in the lungs of rats exposed to 3.5 mg/m3. Hats exposed
for 24 mo to exhaust containing a particle concentration of 3.5 mg/m3 had a five-fold
increase in the bronchoconstrictive prostaglandin PGF2a and twofold increases in the
inflammatory leukotriene LTB4 (Henderson et al., 1985). In similarly exposed mice, there
was a twofold increase in both parameters. The investigators concluded that the release of
May 1990
4-26 DRAFT - DO NOT QUOTE OR CITE
-------
larger amounts of these mediators of inflammation from the alveolar phagocytic cells of the
rats accounted for the greater fibrogenic response seen in that species.
In a continuation of the diesel exhaust studies carried out at the Lovelace Inhalation
Toxicology Research Institute, Mauderly et al. (1987a) evaluated the relative susceptiblity
of developing lungs to damage by diesel exhaust and nitrogen dioxide. Male Fischer 344 rats
(48 per test group) were exposed to either diesel exhaust or NOz (9.5 ppm). Exposures
were 7 h/d and 5 d/week through gestation to the age of 6 mo, or from the age of 6 to 12
mo. The exposure atmospheres are given in Table 4-8. Comparative studies were
conducted on respiratory function, immune response, lung clearance, airway fluid enzymes,
protein and cytology, lung tissue collagen and proteinase, lung burdens of diesel soot, and
lung morphology and histopathology. Exposure to diesel exhaust altered airway fluid
enzyme levels, as well as lung collagen and proteinases in both age groups. After the 6-mo
exposure, adult rats, when compared with controls, exhibited: (1) focal aggregates of
soot-containing macrophages in the alveolar ducts near the terminal bronchioles; (2) a sixfold
increase in the neutrophils (as percentage of total leukocytes) in the airway fluids; (3) a
significantly (p <0.05) higher number of total lymphoid cells in the pulmonary lymph nodes;
(4) delayed clearance of diesel soot and radiolabeled particles (tH = 90 d vs. 47 d for
controls); and (5) increased lung weights. These effects were not seen in the young rats.
On a weight for weight basis, diesel soot accumulation in the lungs was similar in young and
adults immediately after the exposure. During a 6-mo postexposure period, soot clearance
was much more rapid in the young rats. During this time, soot-laden macrophages became
aggregated in the young rats but these aggregations were located primarily in a subpleural
position. Mauderly et al. (1987a) concluded that, in terms of pulmonary function, structure,
and biochemistry, developing rats were not affected more severely than the adult rats by the
diesel exhaust exposures.
GM Studies. Studies on the effects of chronic inhalation of diesel exhaust on male
Fischer 344 rats and male Hartley guinea pigs were conducted at the General Motors
Research Laboratories. The exposures were 20 h/d, 5 J d/week, for up to 2 yr. The exhaust
was generated with a 5.7-L, four-cycle, indirect injection engine operated at a steady speed
and load (1350 rpm, 96 N • m), simulating 40 mph (Schreck et al., 1981). The test fuel had
May 1990 4-27 DRAFT - DO NOT QUOTE OR CITE
-------
TABLE 4-8. COMPOSITION OF EXPOSURE ATMOSPHERES
IN THE MAUDERLY et al., 1987a STUDY*
Particles
NO,
NO,
CO
HC*
(mg/m3)
(ppm)
(ppm)
(ppm)
(ppm)
Exhaust study:
develop.
3.5
52
—
7.9
4.6
adults
3.55
5.0
—
7.9
4.7
NOj study.
develop.
—
—
9 JS
—
—
adult
—
9.5
—
—
Control
0.006
—
—
—
—
'Mean values.
"Hydrocarbon vapors.
Source: Mauderly et al., 1987a.
a sulfur content of 0.27 percent. The pipes in the exhaust delivery system were heated to
100±15*C to simulate typical tailpipe conditions. The exhaust was then diluted with clean
air (dilution ratios 240:1, 80:1, and 40:1) to achieve nominal particle concentrations of 0.25,
0.75, and 1.5 mg/m3. The mean concentrations for the entire 2-yr period were 0.258. 0.796,
and 1.533 mg/m3 (0.2 Mm MMAD, with 90 percent of the mass associated with panicles less
than 1.0 nm). The exposure chambers also contained 3.4 Mg/m3 CO and 2.1 Mg/m3 NO, at
0.258 mg/m3, 5.3 Mg/m3 CO and 5.0 Mg/m3 NO, at 0.796 mg/m3, and 7.9 Mg/m3 CO and 9.2
Mg/m3 NO, at 1.533 mg/m3.
Histological changes in the lungs of the guinea pigs used in the GM tests were
reported on by Barnhart et al., 1981; 1982). In a preliminary study exposures to 0.75 and
1.5 mg/m3 for 2 weeks to 6 mo resulted in uptake of exhaust particulate matter by three
alveolar cell types (alveolar macrophages, epithelial Type I cells, and interstitial
macrophages) and also by granulocytic leukocytes (eosinophils). The alveolar-capillary
boundary increased in thickness as a result of an increase in absolute tissue volume of
interstitium and epithelial Type II cells. In a continuation of these studies guinea pigs were
May 1990
4-28 DRAFT -- DO NOT QUOTE OR CITE
-------
exposed to diesel exhaust (including exhaust diluted to achieve a particle concentration of
6.0 mg/m3) for 2 yr (Barnhart et a]., 1982). A minimal tissue response occurred at the
concentration of 0.25 mg/m3 (i.e., after 9 mo, epithelial Type I and II cells, endothelial cells,
and interstitial cells increased 30 percent over concurrent age-matched controls (p <0.05);
by 24 mo only macrophages and epithelial Type II cells were significantly increased). As in
the earlier study, ultrastructural evaluation indicated that epithelial Type I cells and
macrophages in the alveolar parenchyma and granulocytic leukocytes (eosinophils)
phagocytized the exhaust panicles. Exposure to 0.75 mg/m3 for 6 mo resulted in histological
evidence of fibrosis in areas of macrophage clustering and in focal epithelial Type II cell
proliferation. With increasing exposure levels, epithelial Type II cell clusters occurred in
some alveoli. Excessive secretory products were seen in the alveolar spaces and were
particularly prominent after exposure to 1.5 and 6.0 mg/m3.
As part of the GM studies, Wallace et al. (1987) determined the cellular makeup of
the intravascular and interstitial spaces of the lungs of exposed guinea pigs and rats. The
test animals (three per group) were exposed for 2, 6, 10, or 72 weeks. After 18 mo,
significant increases occurred in alveolar macrophages and in mononuclear interstitial cells
at all exposure levels. From the 6th week through the 18th mo of exposure, both species
had slight increases in fibroblasts in the interstitial areas.
Another part of the GM studies involved the evaluation of the exhaust effects of the
exhaust on the types and numbers of cells lavaged from the lungs of the exposed animals.
In one series of tests, rats and guinea pigs were exposed to exhaust particle levels of 0.25,
1.5, or 6.0 mg/m3 for up to 12 mo (Chen et al., 1980). The exposures resulted in increases
in the size and relative surface area of lavaged alveolar macrophages, and in an apparent
decrease in phagocytizing ability. Exudative leukocytes (eosinophils in guinea pigs and
neutrophils in rats) appeared in the lavage fluid in numbers which increased with exposure
level and duration. Lavage fluid from exposed guinea pigs also contained increased numbers
of "reactive" monocytes, macrophages characterized by atypically lobulated or multiple
nucleii. In related studies, rats were exposed for 26 and 48 weeks to 0.25 and 1.5 mg/m3, or
for 24J, 28, and 52 weeks to 0.75 mg/m3 (Vostal et al., 1981, 1982; Strom, 1984).
Exposure-related changes occurred in alveolar macrophages, polymorphonuclear leukocytes
May 1990 4-29 DRAFT - DO NOT QUOTE OR CITE
-------
(PMN), and lymphocytes in the lung lavage fluid. At a panicle concentration of 0.25 mg/m\
cell counts of alveolar macrophages were not different from control values, at 0.75 mg/m3,
the counts were 50 percent higher than the controls; and at 1.5 mg/m3, counts were twice
as high as the controls. The data suggested that the response of the alveolar macrophages
was concentration dependent. Polymorphonuclear leukocytes were also present in the lung
lavage fluid of animals exposed to 0.75 and 1.5 mg/m3. Mononuclear leukocytes (which
appeared to be lymphocytes) were found in the fluid after 1 yr of exposure. Protein,
B-ghicuronidase activity, and acid phosphatase activity were significantly elevated in lavaged
cells from animals exposed for 6 mo to 0.75 and 1.5 mg/m3. In related tests on rats and
guinea pigs, Weller et al. (1981) found that acid phosphatase activity in lavaged alveolar
macrophages from exposed animals was reduced in an exposure-dependent (0.25 and 1.5
mg/m3) and time-dependent (1 d, 2 weeks, and 12 mo) manner.
In the same series of GM tests, the lungs of animals exposed to the exhaust were also
analyzed for biochemical changes (Misiorowski et al., 1980; Misiorowski et al., 1981;
Eskelson et al., 1981). In most cases, exposures to 0.25 mg/m3 did not cause any significant
changes. DNA content in lung tissue and the rate of collagen synthesis were significantly
increased at the highest exposure (1.5 mg/m3) after 6 mo. Collagen deposition was not
affected. Total lung collagen content increased in proportion to the increase in lung weight.
The activity of prolyl hydroxylase was significantly increased only at 12 weeks at 0.25 and 1.5
mg/m3; it then decreased with age. Lysal oxidase activity did not change. After 9 mo of
exposure, there were significant increases in lung phospholipids in animals exposed to 0.75
mg/m3 and in lung cholesterol in animals exposed to 1.5 mg/m3. Pulmonary prostaglandin
dehydrogenase activity in guinea pigs was stimulated by an exposure for 6 weeks to 0.25
mg/m3, but was not affected by exposure to 1.5 mg/m3 (Chaudhari et al., 1980, 1981).
Exposures for 12 or 24 weeks resulted in a concentration-dependent lowering cf this enzyme
activity. Schneider and Felt (1981) reported that similar exposures did not substantially
change adenylate cyclase and guanylate cyclase activities in lung or liver tissue of the
exposed rats and guinea pigs.
In related GM-sponsored studies, Chaudhari and Dutta (1982) investigated whether
diesel exhaust-induced lung damage would result in the release of angiotensin converting
May 1990 4-30 DRAFT - DO NOT QUOTE OR CITE
-------
enzyme (ACE) from lung endothelial tissue as is the case for pulmonary edemagenic agents
such as paraquat. Lung ACE activity decreased significantly after 8 weeks of exposure to
6.0 mg/m3. ACE activity in serum increased significantly after 4 weeks of exposure and
showed a much greater increase after 8 weeks. Chaudhari and Dutta (1982) suggested that
the changes in ACE activity in the lung and serum were indicative of damage to pulmonary
endothelial cells.
Fraunhofer Institut Studies. Chronic inhalation toxicity studies on diesel exhaust have
been conducted at the Fraunhofer Institut fOr Toxikologie und Aerosolforschung. The
results of these studies had been reported by Heinrich et al. (1982,1985,1986a, 1986b). In
one set of tests, the exhaust was generated with a 2.4-L Daimler-Benz engine operated at
a constant load of 16 kW and a uniform speed of 2400 rpm. The diesel fuel contained 0.36
percent sulfur. The concentrations of the other components of the diluted exhaust are given
in Table 4-6. The test animals were female Syrian golden hamsters (48 per test group).
Exposures were 7 to 8 h/d, 5 d/week, to diesel exhaust diluted to achieve a particle
concentration of 3.9 mg/m3 (0.1 jim MMAD). Histological examination of the respiratory
tract revealed discrete proliferative changes in the lung (Heinrich et al., 1982,1985). Sixty
percent of the changes were described as adenomatous proliferations. Both control and test
animals exhibited increasingly serious degrees of amyloidosis of the kidneys, adrenals, liver,
and spleen. Cysts were also observed in the liver.
Heinrich et al. (1986a; see also Stoeber, 1986) conducted additional long-term
inhalation studies on hamsters, mice, and rats. The exhaust was generated by a 40-kW, 1.6-L
engine operated under the US. 72 (FTP) test driving cycle. The fuel contained 0.36 percent
sulfur. The concentrations of the other components of the diluted exhaust are given in
Table 4-7. The test animals consisted of 96 male and female Syrian golden hamsters, 96
female NMR1 mice, and 96 female SPF Wistar rats per test group. Exposures were for 19
h/d, 5 d/week for up to 2 years. The exhaust was diluted to achieve a particle concentration
of 4.24 mg/m1 (0.35 nm MMAD). The maximum exposure period was 120 weeks for
hamsters and mice and 140 weeks for rats. Histopathological examination indicated that
different levels of response occurred in the thise species. In hamsters, the exhaust caused
thickened alveolar septa, bronchiolo-alveolar hyperplasia, and what were termed
May 1990 4-31 DRAFT - DO NOT QUOTE OR CITE
-------
emphysematous lesions. In mice, bronchiolo-alveolar hyperplasia occurred in 64 percent of
the animals exposed to the exhaust and in 5 percent of the controls. Multifocal alveolar
lipoproteinosis occurred in 71 percent and multifocal interstitial fibrosis occurred in 43
percent of the animals exposed to exhaust, but in only 4 percent or less of the controls. In
the exposed rats there were severe inflammatory changes in the lungs, as well as thickened
septa, foci of macrophages, and hyperplastic and metaplastic lesions. Biochemical studies
of lung lavage fluids revealed that the exposures caused significant (p £0.05) increases in
lactate dehydrogenase, alkaline phosphatase, acid phosphatase, glucose-6-phosphate
dehydrogenase (G6P-DH), total protein, protease (pH 5.1), and collagen in the hamsters and
rats. Cytological studies revealed a significant increase in macrophages in lung lavage fluid
of rats and hamsters after 1 yr of exposure, but not after 2 yr. Total leukocytes increased
significantly in both species after 1 and 2 yr of exposure (p <0.05). In hamsters, significant
increases in granulocytes and lymphocytes in the lung lavage fluid were seen after 2 yr of
exposure (p <0.05).
Ishmishi Studies. Histopathological effects of diesel exhaust on the lungs of rats have
been investigated by Ishinishi et al. (1986b) who compared the effects of both light-duty
(LD) and heavy-duty (HD) diesel engines. The LDD exhaust was generated from 1.8-L 4-
cylinder (swirl chamber) engines operated at 1200 rpm. The HDD exhaust was generated
from 11-L direct injection engines operated at 1700 rpm. The exhaust was diluted to achieve
nominal panicle concentrations of 0.1 (LD only), 0.4 (LD and HD), 1 (LD and HD), 2 (LD
and HD), and 4 mg/m3 (HD only). Particle size was not reported. Concentrations of other
components of the diluted exhausts are given in Table 4-4. The test animals (male and
female F344 rats; 64, 95, or 120 per test group) were exposed for 16 h/d, 6 d/week, for 30
mo. No histopathological changes were observed in the lungs of rats exposed to particle
concentrations less than 0.4 mg/m3. At concentrations above 0.4 mg/m3, severe
morphological changes were seen. These changes consisted of marked hyperplasia of Type
II epithelial cells and development of lung adenomas and carcinomas. Pathological lesions
consisting of shortened and absent cilia in the tracheal and bronchial mucosal epithelium
were also observed in the exposed animals. These lesions appeared to increase in seventy
with increase in exhaust concentration and period of exposure.
May 1990 4-32 DRAFT - DO NOT QUOTE OR CITE
-------
Iwai Studies. Histopathological effects of diesel exhaust on the lungs of rats have also
been investigated by Iwai et al. (1986). The exhaust was generated from a 2.37-L diesel
engine operated at 1000 rpm and 80 percent engine load. The exhaust was diluted (1:10)
with clean air so that the exposure atmosphere contained 4.9±1.6 mg/m3 of particulate
matter, 30.9±10.9 ppm NO,, 1.8±1.8 ppm NO^ 13.1±3.6 ppm SOj, and 7.0±1.4 ppm CO.
The test animals were Fischer 344 female rats. The exposures were 8 h/d, 7 d/week for 24
mo. At 1 and 3 mo of exposure, there were only minimal histological changes in the lungs
of the exposed animals. After 6 mo of exposure, there were noticeable numbers of particle-
laden macrophages distributed irregularly throughout the lung and a proliferation of Type
II alveolar epithelial cells with adeomatous metaplasia in areas where the macrophages had
accumulated. After 1 yr of exposure, foci of heterotropic hyperplasia of ciliated or
nonciliated bronchiolar epithelium on the adjacent alveolar walls were more common, the
quantity of deposited particulate matter increased, and there was an increase in the number
of degenerated alveolar macrophages and in the number of proliferative lesions of Type II
alveolar or bronchiolar epithelial cells. After 2 yr of exposure, there was a fibrous thickening
of the alveolar walls and mast cell infiltration with epithelial hyperplasia in areas where the
macrophages had accumulated. Neoplastic changes occurred, resulting in two types of lung
carcinoma, adenocarcinoma and squamous or adenosquamous carcinoma.
4.1.23.4. Effects on the liver
Meiss et al. (1981) investigated the effects of diesel exhaust on the liver of Syrian
golden hamster exposed 7 to 8 h/d, 5 d/week for 5 mo to exhaust diluted 1:5 (or 1:10) with
clean air. The diluted exhaust contained 11 (4) mg/m3 of particulate matter, 25 (12) ppm
CO, 43 (23) ppm NO, 1.5 (0.5) ppm NO2, 7 (3) ppm SO^ and 12 (8) ppm hydrocarbons.
The livers of animals exposed to both dilutions exhibited moderate dilatation of the sinusoids
with activation of the Kupffer cells and slight changes in nuclei. Fatty deposits were
observed in the sinusoids, and small fat droplets were occasionally observed in the peripheral
hepatocytes. Mitochrondria often showed a loss of cristae and exhibited pleomorphic
character. Giant microbodies were seen in the hepatocytes which were moderately enlarged,
and gap junctions between hepatocytes exhibited a wide range in structural diversity.
May 1990 4-33 DRAFT -- DO NOT QUOTE OR CITE
-------
417 TS Effects on defense mechanisms
A number of laboratory studies have demonstrated that exposure to diesel exhaust
can alter resistance to respiratory tract infections.
EPA-Sponsored Studies. Campbell et al. (1980, 1981) recorded the mortality
response of mice exposed to diesel exhaust following infectious challenge with Salmonella
typhimurium, Streptococcus pyogenes, or A/PR8-3 influenza virus. The test animals (20 per
group) were young adult female, CR/CD-1 albino mice. The exposure period was 8 h/d, 7
d/week, for up to 321 d. The exhaust dilution ratio was 18:1, resulting in a particle
concentration of about 6 mg/m3. The exhaust was generated by a Nissan CN-6 engine
operated on the Federal Short Cycle. Characteristics of the exposure atmosphere are given
in Table 4-9. Exposure to the diesel exhaust resulted in enhanced susceptibility to the lethal
effects of S. pyogenes infection at all exposure durations (2 and 8 h; 8, 15, 16, 307, and
321 d). Tests with Salmonella were inconclusive due to the high mortality rates in the
controls. The mice exposed to diesel exhaust exhibited no enhanced susceptibility to
infection when challenged with the influenza virus.
TABLE 4-9. COMPOSITION OF EXPOSURE ATMOSPHERE
IN CAMPBELL et aL, STUDY"
Particles
CO
NO,
NO
SO,
HC
(mg/mJ)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
Test
53-7.9
19-22
1.8-3.6
10-16
0.9-2.8
7-10
Control
—
1.2-2.7
0.02-0.13
0.03-0.17
0.04-0.94
2.9-4.7
'Range of mean values.
Source: Campbell et aL, 1981.
Hatch et al. (1985) reported that in vitro tests on rabbit alveolar macrophages
exposed to diesel exhaust resulted in significant decreases in viability and cellular ATP. In
contrast, in vivo tests on CD-I mice revealed no changes in susceptibility to Streptococcus
May 1990
4-34 DRAFT - DO NOT QUOTE OR CITE
-------
infection following intratracheal injection of 100 ng of diesel exhaust particles suspended in
unbuffered saline. The recorded excess mortality was 14.4±11.3 percent (S.E), and not
significantly different from zero. Hatch et al. (1985) suggested that any exhaust-induced
effects on macrophage viability in the in vivo test might have been compensated for by other
factors such as an increased influx of phagocytically active leukocytes from the blood.
NIOSH-Sponsored Studies. The effects of diesel exhaust on the immune system of
male Fischer 344 rats were investigated by Mentnech et al. (1984). Exposures were for 7
h/d, 5 d/week for up to 2 yr. The diesel emissions were generated by a four-cycle,
Caterpillar engine equipped with a water scrubber. The 24-mo mean concentration of
particulate matter was 1.95±0.25 mg/m3. As indicated by Lewis et al. (1986), the mass
median aerodynamic diameter of the particles was 0.23 and 0.36 nm by the instrumental and
scanning electron microscope techniques: the geometric standard deviations were 2.5 and
2.0, respectively. The 24-mo mean concentrations of other exhaust components were 11.5
ppm CO; 8.7 ppm NO; 1.5 ppm NO:; 0.81 ppm S02; 60.2 ppb acrolein; and 38.3 ppb
formaldehyde (as given in Lewis et al., 1986). Hie exposed animals were tested for
immunocompetency by determining antibody-producing cells in the spleen 4 d after
immunization with sheep erythrocytes. Proliferative response of splenic T-lymphocytes to
the mitogens concanavalin A and phytohemagglutinin was used as the monitoring technique.
There were no statistically significant differences between the test and control animals after
12 and 24 mo of exposure.
In related NIOSH-sponsored studies, the effects of diesel exhaust on interferon
production and susceptibility of mice to viral infection was invesigated (Hahon et al., 1982,
1985). In in vitro studies using mammalian LLC-MK; cells, viral induction of interferon was
depressed by about 60 percent following exposure of the cells to diesel exhaust particles
'Hahon et al., 1982). Influenza virus growth in pretreated cell monolayers was 2 to 3 times
higher than that in controls. In in vivo studies conducted on Ao/PR/8/34 influenza
virus-infected CD-I Swiss mice, 3- and 6-mo exposures to diesel exhaust resulted in
significant increases in the percentage of animals showing lung consolidation, higher virus
growth, depressed IFN levels, and a fourfold reduction in hemagglutinin antibody levels
(Hahon et al., 1985). These effects were not seen after 1-mo exposure. The test animals
iMay 1990 4-35 DRAFT - DO NOT QUOTE OR CITE
-------
were exposed for 7 h/d, 5 d/week, and the exhaust was diluted to achieve a particle
concentration of 2 mg/m3.
Lovelace Studies. Bice et al. (1985) evaluated the development of the immune
response in Fischer 344 rats and CD-I mice exposed to diesel exhaust in the Lovelace
inhalation toxicity studies. The test animals were exposed to three dilutions of diesel exhaust
for 7 h/d, 5 d/week, for 30 mo. The three exhaust dilutions resulted in nominal particle
concentrations of 0.35, 3 J, and 7 mg/m3 (0.23, 0.25, and 0.26 jim MMAD, respectively).
Concentrations of other exhaust components are given in Table 4-3. Test animals and
controls were immunized by intratracheal instillation of sheep red blood cells (SRBC) after
6, 12, 18, and 24 mo of exposure. No suppression in the immune response occurred in
either species. After 12, 18, and 24 mo of exposure, the total number of anti-SRBC IgM
antibody forming cells (AFC) was elevated in rats, but not in mice, exposed to the two
highest levels of diesel exhaust; after 6 mo of exposure, only the highest exhaust
concentration caused this effect in rats. The number of AFC per 106 lymphoid cells in
lung-associated lymph nodes and the level of specific IgM, Igl, or IgA in rat sera were not
significantly altered. The investigators concluded that the exposures had only a minimal
effect on the immune and antigen filtration function of the lymphoid tissue.
GM-Sponsored Studies. As part of the GM-sponsored series of studies on diesel
exhaust, Dziedzic (1981) investigated the effects of exhaust on the number of T, B, and null
lymphocytes in tracheobronchial lymph nodes and blood and spleen of male Hartley guinea
pigs. The exhaust was generated with a 5.7-L, four-cycle, indirect injection engine operated
at a steady speed and load (1350 rpm, 96 N*m), simulating 40 mph. The test fuel had a
sulfur content of 0.27 percent. The pipes in the exhaust delivery system were heated to
100±15 *C to simulate typical tailpipe conditions. The exhaust was then diluted with clean
air (dilution ratio 40:1) to achieve a particle concentration of 1.533 mg/m3 (0.2 ^m MMAD)
(Schreck et ah, 1981). The diluted exhaust also contained 7.9 /xg/m3 CO and 9.2 ^g/m3 NOr
The animals were exposed for 20 h/d, 5.5 d/week, for 4 and 8 weeks. There were no
significant exposure-related changes in any of the measured parameters, and cell viability was
comparable to the control value.
May 1990
4-36 DRAFT - DO NOT QUOTE OR CITE
-------
Other Studies. Takafuji et al. (1987) evaluated the IgE antibody response of BDF,
mice innoculated intranasally at intervals of 3 weeks with varying doses of a suspension of
diesel exhaust particles in ovalbumin solution. Antiovalbumin IgE antibody titres, assayed
by passive cutaneous anaphylaxis, were enhanced by doses as low as 1 fig particles.
Responses were higher than those in animals immunized with ovalbumin alone.
4.1.23.6. Blood and Cardiovascular Systems
Several studies have evaluated the effects of diesei exhaust exposure on hematological
and cardiovasular parameters of laboratory animals.
EPA-Sponsored Studies. Wiester et al. (1980) studied the effects of diesel exhaust
on the hean of male and female Hartley guinea pigs. The test animals were exposed 20 h/d
for 28 d to irradiated and nonirradiated diesel exhaust containing 6 mg/m3 of exhaust
particles (other characteristics of the exposure atmospheres are given in Table 4-1). No
effect on hean mass and no changes in ECG were observed. Animals exposed to irradiated
exhaust did exhibit a small but significant decrease in heart rate.
NIOSH-Sponsored Studies. The effects of diesel exhaust on the cardiovascular
system of Fischer 344 rats were investigated by Vallyathan et al. (1986) in NIOSH-sponsored
studies evaluating the combined effects of diesel exhaust and coal dust exposure. The test
animals (72 males and 72 females per test group) were exposed 7 h/d, 5 d/week for 2 yr.
The exposure atmosphere contained 2 mg/m3 of particulate matter (0.23 MMAD). The
diesel emissions were generated by a four-cycle, Caterpillar engine equipped with a water
scrubber. The fuel contained <0.5 percent sulfur by mass. As indicated by Lewis et al.
(1986), the concentrations of other exhaust components were 11.5 ppm CO; 8.7 ppm NO;
1.5 ppm NOz; 0.81 ppm SO^ 60.2 ppb acrolein; and 38.3 ppb formaldehyde. Body weights,
heart weights, right and left ventricular wall thickness, severity of cardiomyopathy, and
changes in the small pulmonary arteries were evaluated. No statistically significant
differences were seen in heart weight, heart wall thickness, or pulmonary artery wall
thickness between control groups and those exposed to diesel exhaust alone, coal dust alone,
or diesel exhaust and coal dust; however, there was a trend toward increased pulmonary
arterial wall thickness in the diesel exhaust-exposed animals.
May 1990 4-37 DRAFT - DO NOT QUOTE OR CITE
-------
DOE-Sponsored Studies. As pan of a study examining the combined effects of
inhaled diesel exhaust and coal dust in male Wistar rats, Karagianes et al. (1981) measured
hematological parameters in animals exposed to diesel exhaust alone. The test animals (6
per test group) were exposed 6 h/d, 5 d/week, for 4, 8, 16, or 20 mo. The exhaust was
diluted to achieve a particle concentration of 8.3±2.0 mg/m3. The concentration of CO was
maintained at 50 ppm, and that of NO: at 4 to 6 ppm. The exposure had no effect on body
weight, mortality pattern, hematocrit, or differential leukocyte count. A 3 to 5 percent
increase in percent COHb saturation was observed in the exhaust-exposed animals.
GM-Sponsored Studies. As pan of the GM-sponsored studies, Penney et al. (1981)
examined hematological and cardiovascular parameters in Fischer 344 rats (32 per test
group) and Hanley guinea pigs (about 36 guinea pigs per test group) exposed to diesel
exhaust 20 h/d, 5.5 d/week, for 78 weeks. Particle concentrations were 0.25, 0.75, and 1.5
mg/m3. At the highest exposure, the concentration of CO was 4.3 ppm and NOx was 5.9
ppm (concentrations at other exposure levels and for other components were not reported).
Morphometric analysis revealed no significant alterations in heart mass at any exposure level.
In addition, hemoglobin, hematocrit, and RBC levels were not significantly different from
control values in either species.
Fraunhofer Institut Studies. Heinrich et al. (1982, 1985) measured hematological
parameters in female Syrian golden hamsters exposed to diesel exhaust. The test animals
(48 per test group) were exposed 7-8 h/d, 5 d/week, to exhaust diluted to achieve a particle
concentration of 3.9 mg/m3 (0.1 nm MMAD). Concentrations of other exhaust components
are given in Table 4-6. Blood analyses conducted after 29 weeks of exposure indicated a low
erythrocyte count accompanied by an increase in erythrocyte volume and a reduced
leucocyte count in the exhaust-exposed animals.
Xshinishi Studies. Hematological effects of diesel exhaust in male and female Fischer
344 rats was evaluated by Ishinishi et al. (1986b) as part of a long-term inhalation exposure
study. The rats (64, 95, or 120 per test group) were exposed to diluted exhaust from
light-duty (LD) and heavy-duty (HD) diesel engines. Exposures were 16 h/d, 6 d/week, for
30 mo. Nominal concentrations of particulate matter in the separate tests were 0.1, 0.4,1,
2, and 4 mg/m3. Exposure conditions and exhaust component concentrations are given in
May 1990 4-38 DRAFT - DO NOT QUOTE OR CITE
-------
Table 4-4. No significant changes in hematological parameters occurred in the first 24 mo
of exposure. At 30 mo there were slight increases in red blood cell count, hemoglobin
concentration, and hematocrit in some exposed groups but not in others (no other data
reported). The platelet count was significantly decreased in males exposed to 4 mg/m3.
CCMC Study. Brightwell et al. (1986) evaluated the hematological effects of filtered
and unfiltered diesel exhaust on male and female Fischer 344 rats and Syrian hamsters. The
animals (144 rats and 312 hamsters per test group) were exposed for five 16-h periods per
week for 2 yr. The exhaust was generated with a 1.5-L engine operated under the U.S. 72
(FTP) test driving cycle. Three exhaust dilutions were tested, producing particle
concentrations of 0.7, 2.2, and 6.6 mg/m3. At the highest exposure the NO, concentration
was 8±1 ppm. and the CO concentration was 32±11 ppm (concentration of other gaseous
pollutants and for the other exposure levels were not given). Rats exposed to the highest
levels of unfiltered diesel exhaust exhibited a significant reduction in lymphocyte
concentrations. Significant increases in red blood cell count, hemaglobin, hematocrit, white
blood cell count, segmented neutrophils, and prothrombin time were also reported.
4.1.23.7. Serum Chemistry
As part of chronic inhalation exposure studies, Ishinishi et al. (1986b) examined the
effects of diesel exhaust on serum biochemistry in male and female Fischer 344 rats. The
rats (64,95, or 120 per test group) were exposed to diluted exhaust from light-duty (LD) and
heavy-duty (HD) diesel engines. Exposures were 16 h/d, 6 d/week, for 30 mo. Nominal
concentrations of exhaust particles in the separate tests were 0.1, 0.4, 1, 2, and 4 mg/m3.
Exposure conditions and exhaust component concentrations are given in Table 4-4.
Compared to control values, significant decreases were seen in the activity of cholinesterase
and in the concentrations of free cholesterol and phospholipid in HD males exposed to 4
mg/m3 (no other data given). The latter two parameters were also reduced in HD males
exposed to 2 mg/m3.
Brightwell et al. (1986) measured changes in serum biochemical parameters in male
and female Fischer 344 rats and Syrian hamsters exposed to both filtered and unfiltered
diesel exhaust. The animals (144 rats and 312 hamsters per test group) were exposed for
May 1990 4-39 DRAFT -- DO NOT QUOTE OR CITE
-------
five 16-h periods per week for 2 yr. Three diesel particle concentrations were used; 0.7, 2.2,
and 6.6 mg/m3. At the highest exposure, the CO and NOx concentrations were 32 ppm and
8 ppm, respectively (concentrations of CO and NO, at the other two exhaust levels were not
reported). Rats exposed to the highest levels of unfiltered diesel exhaust showed significant
reductions in blood glucose level, total blood proteins, triglycerides, and cholesterol.
Significant increases in blood urea nitrogen, alkaline phosphatase, alanine aminotransferase,
and aspartate aminotransferase were also reported. These changes were attributed to
possible reduced kidney function, impaired liver function, and hypoxia. Kidney or liver
damage was not confirmed by histological examination. In hamsters, some blood chemistry
changes, such as an increase in gamma-glutamyl transpeptidase were thought to be
suggestive of liver damage.
4.1.23.8. Effects on Enzyme Activity
Several studies have examined the effects of diesel exhaust exposure on tissue
enzymes associated with the metabolism and possible activation of polynuclear aromatic
hydrocarbons (PAHs). As part of the EPA studies on diesel exhaust, Lee et al. (1980)
measured the activities of aryl hydrocarbon hydroxylase (AHH) and epoxide hydrase (EH)
in liver, lung, testis, and prostate gland of adult male CD Sprague-Dawley rats exposed to
the exhaust. The test animals were exposed to the diluted exhaust (13:1 dilution ratio)
20 h/d, for 42 d. The exposure atmosphere, as reported in Lee et al. (1978), included 6.32
mg/m3 of suspended panicles, 2.19 ppm NO* 2.13 ppm SO* 15.7 ppm CO, and 15.6 ppm
total hydrocarbons (as carbon). After 33 d, AHH activity (pmol/min/mg microsomal protein)
was 0.28±0.16 in the prostate, 4.19±0.32 in the lung, 117.5±4.6 in the liver, and 1.54±0.23 in
the testis. Except for the testis, these values were significantly higher (p <0.05) than
corresponding activity in control animals. Epoxide hydrase activity was not significantly
different from control values for any of the organs tested.
In the NIOSH-sponsored diesel exhaust studies, Rabovsky et al. (1984) investigated
the effects of chronic inhalation of diesel exhaust and coal dust on microsomal cytochrome
P450-associated benzo[a]pyrene hydrolase and 7-ethoxycoumarin deethylase activities in rat
lung and liver. Male Fischer 344 rats were exposed 7 h/d, 5 d/week, for 24 mo to diesel
May 1990 4-40 DRAFT - DO NOT QUOTE OR CITE
-------
exhaust consisting of 2 mg/m1 particles, 10.5 ppm CO, 1.5 ppm NOj, and 0.6 ppm SO,. The
exposure had no effect on B[a]P hydrolyase or 7-ethoxycoumarin deethylase activities in lung
or liver. In related studies, Rabovsky et al. (1986) examined the effects of diesel exhaust on
virally induced enzyme activity and interferon production in female Swiss CD-I mice. The
mice were exposed 7 h/d, 5 d/week, for 1 mo to diesel exhaust diluted to achieve a particle
concentration of 2 mg/m3. Following the exposure, the mice were innoculated intranasally
with influenza virus, and changes in serum levels of interferon and activity levels of three
liver microsome enzymes associated with the cytochrome P450 pathway were recorded. In
the absence of viral innoculation, exposure to diesel exhaust had no effect on activity levels
of the three liver microsome enzymes; however, following viral innoculation, the temporal
pattern of enzyme activity was altered by the exhaust.
Effects of diesel exhaust on enzyme activities in male Fischer 344 rats were studied
by Chen and Vostal (1981) and Navarro et al. (1981) as part of the General Motors diesel
toxicity studies. Chen and Vostal (1981) measured the activity of aryl hydrocarbon
hydroxylase (AHH) and the content of cytochrome P450 in lungs and liver of animals
exposed by inhalation or intraperitoneal (i.p.) injection (of dichloromethane extracts of the
particulate matter). In the inhalation exposures, the exhaust was diluted to achieve particle
concentrations of 0.75 or 1.5 mg/m3, and the exposure protocol was 20 h/d, 5.5 d/week, for
up to 9 mo. The concentration of total hydrocarbons, and particle-phase hydrocarbons was
not reported. Parenteral administration involved repeated i.p. injections at several dose
levels for 4 d. Inhalation exposures had no significant effect on liver microsomal AHH
activity; however, lung AHH activity was slightly reduced after 6 mo exposure to 1.5 mg/m3.
An i.p. dose of diesel exhaust extract, estimated to be equivalent to the inhalation exposure,
had no effect on AHH activity in liver or lungs. No changes were observed in cytochrome
P450 content in lungs or liver following either inhalation exposure or i.p. treatment.
In related studies, Navarro et al. (1981), evaluated the effect of diesel exhaust
exposure on hepatic and pulmonary microsomal enzyme activities in Fischer 344 rats. The
same exposure regime was used (20 h/d, 5.5 d/week, for up to 1 yr), and the exhaust was
diluted to achieve particle concentrations of 0.25 and 1.5 mg/m3 (a few tests were also
conducted at 0.75 mg/m3). After 8 weeks of exposure, there was no evidence for the
May 1990 4-41 DRAFT - DO NOT QUOTE OR CITE
-------
induction of cytochrome P450, cytochrome P448, or NADPH-dependent cytochrome c
reductase in rat liver microsomes. One year of exposure had little, if any, effect on hepatic
metabolism of B[a]P. However, 1 yr of exposure to 0.25 and 1.5 mg/m3 impaired the ability
of lung microsomes to metabolize B[a]P (0.015±0.05 and 0.12±0.01 nmole/30 min/mg protein,
respectively, vs. 0.32±0.05 nmole/30 min/mg protein for the controls p <0.05).
4.L2J.9. Effects on Behavior
In EPA studies Laurie et al. (1978) and Laurie et al. (1980) examined behavioral
alterations in adult and neonatal Sprague-Dawley rats exposed to diesel exhaust. Laurie et
al. (1978, as reported in Laurie et al., 1980) reported that rats exposed 20 h/d, 7 d/week, for
6 weeks, to exhaust containing 6 mg/m3 of particulate matter exhibited a significant reduction
in adult spontaneous locomotor activity (SLA) and in neonatal pivoting (no other data
given). In a follow-up to this study, Laurie et al. (1980) found that shorter exposures (8 h/d)
also resulted in a reduction in SLA in adult rats. The exhaust contained 5.97 mg/m3 of
particulate matter, 19.2 ppm CO, 7.29 ppm hydrocarbons, 2J1 ppm NO^ and 1.82 ppm SOj.
Laurie et al. (1980) conducted additional behavioral tests on adult rats that had been
exposed during the neonatal period. For two of three exposure situations (20 h/d for 17 d
postpaturition, or 8 h/d for the first 28 or 42 d postpaturition), significantly lower SLA was
observed in the majority of the tests conducted on the adults after week 5. When compared
with control animals, adult 15-mo-old rats that had been exposed as neonates (20 h/d for 17
d) also exhibited a significantly slower rate of acquistion of a bar pressing task to obtain
food. The investigators noted that the evidence was insufficient to determine whether the
differences were due to a learning deficit or to some other cause (e.g., motivational or
arousal differences).
4.L23.10. Neurophysiologies Effects
Neurophysiological effects of diesel exhaust in male Sprague-Dawley rats were
investigated by Laurie and Boyes (1980, 1981). Rat pups were exposed to a 18:1 dilution
of diesel exhaust (6 mg/m3 particle concentration) for 8 h/d and 7 d/week (total exposure
period not reported). Somatosensory evoked potential, as elicited by a 1 mA electrical pulse
May 1990 4-42 DRAFT - DO NOT QUOTE OR CITE
-------
to the tibial nerve in the left hind limb, and visual evoked potential, as elicted by a flash of
light, were the end points tested. An increased pulse latency was reported for the diesel-
exhaust-exposed rats, and this was thought to be due to a reduction in the degree of nerve
myelinization.
4.1.2.4. Effects on Reproduction and Development
EPA-sponsored studies on diesel exhaust have included an evaluation of potential
teratogenicity in rats and rabbits (Werchowski et al., 1980a, 1980b). Twenty albino Spague-
Dawley rats were exposed in inhalation chambers to an atmosphere containing 10 percent
diesel exhaust and 90 percent clean air (no other data given). Exposures lasted 8 h/d and
were conducted on days 6 through 15 of gestation. There were no signs of maternal toxicity
or decreased fertility. No malformations or other teratogenic effects were observed in
20-d-old fetuses (Werchowski et al., 1980a). In a second study, 42 New Zealand albino
rabbits were exposed to the same exhaust concentration as rats on day 6 through 18 of
gestation. The exposures were for 8 h/d. No adverse effects on body weight gain or fertility
were seen in the animals exposed to diesel exhaust. No visceral or skeletal developmental
abnormalities were observed in the fetuses (Werchowski et al., 1980a).
Several studies have evaluated the effect of diesel exhaust exposure on sperm
morphology. Pereira et al. (1981a) reported a significant (p = 0.0024) 2.67-fold increase in
sperm abnormalities in male Chinese hamsters exposed for 8 h/d for 6 mo to diesel exhaust
(concentrations of particulate matter and gaseous components were not reported). In
another study in which A/Strong mice were exposed to diesel exhaust (particle concentration
6.0 mg/m3) for 8 h/d for 31 or 38 weeks, no significant differences were observed in sperm
morphology between exposed and control animals (Pereira et al., 1981b). It was noted,
however, that there was a very high rate of spontaneous sperm abnormalites in this strain
of mice, and this may have masked any small positive effect. Quinto and DeMarinis (1984)
reported a statistically significant and dose-related increase in sperm abnormalities in male
(C57Bl/6xC3H)F, mice dosed intraperitoneally for 5 d with 50,100, and 200 mg/kg of diesel
particulate matter suspended in corn oil. A significant decrease in sperm number was seen
at the highest dose, but testicular weight was not affected by the treatment.
May 1990 4-43 DRAFT - DO NOT QUOTE OR CITE
-------
4.2. COMPARISON OF HEALTH EFFECTS OF
FILTERED AND UNFILTERED EXHAUST
In several chronic toxicity studies on diesel exhaust, the experimental protocol
included exposing some of the test animals to filtered exhaust containing no paniculate
matter. Comparisons were then made between the effects caused by the whole, unfiltered
exhaust and those caused by the gaseous components of the exhaust.
In Fraunhofer Institut studies, Heinrich et al. (1982) compared the toxic effects of
whole and filtered diesel exhaust in female Syrian golden hamsters and female Wistar rats.
The exhaust was generated with a 2.4-L Daimler-Benz engine operated at a constant load
of 16 kW and a uniform speed of 2400 rpm. The diesel fuel contained 0.36 percent sulfur.
The exhaust was diluted to achieve a particle concentration of 3.9 mg/m3 (0.1 nm MMAD).
The concentrations of the other components of the exposure atmospheres are given in Table
4-6. The exposures were for 7 to 8 h/d and 5 d/week. Rats exposed for 24 mo to either
whole or filtered exhaust exhibited no significant changes in respiratory rate, respiratory
minute volume, compliance, or resistance. In the hamster tests, histological changes
(adenomatous proliferations) were seen in the lungs of animals exposed to both whole and
filtered exhaust; however, in all groups exposed to the whole exhaust, the number of animals
exhibiting such lesions was significantly higher than for the corresponding groups exposed
to filtered exhaust or clean air. Both control and test animals exhibited increasingly serious
degrees of amyloidosis of the kidneys, adrenals, liver, and spleen. In addition, cysts were
observed in the liver of both control and treated animals.
In another Fraunhofer Institut study, Heinrich et al. (1986a, see also Stoeber. 1986)
compared the toxic effects of whole and filtered diesel exhaust in hamsters, rats, and mice.
The exhaust was generated by a 40-kW, 1.6-L engine operated under the U.S. 72 (FTP) test
driving cycle. The fuel contained 0.36 percent sulfur. The exhaust was diluted to achieve
a particle concentration of 4.24 mg/m3 (0.35 nm MMAD). The concentration of the other
components of the exposure chambers are given in Table 4-7. The test animals (96 per test
group) were exposed for 19 h/d, 5 d/week, for 120 (hamsters and mice) or 140 (rats) weeks.
Body weights of hamsters were not affected by either exposure. Body weights of rats and
mice were reduced by the whole exhaust, but not by the filtered exhaust. Exposure-related
May 1990 4-44 DRAFT - DO NOT QUOTE OR CITE
-------
higher mortality rates occurred in mice after 2 yr of exposure to whole exhaust. After 1 yr
of exposure to the whole exhaust, hamsters exhibited increased lung weights, a significant
increase in airway resistance, and a nonsignificant reduction in lung compliance. For the
same time period, rats exhibited increased lung weights, a significant decrease in dynamic
lung compliance, and a significant increase in airway resistance. Test animals exposed to
filtered exhaust did not exhibit such effects. Histopathological examination indicated that
different levels of response occurred in the three species. In hamsters, filtered exhaust
caused no significant histopathological effects in the lungs; whole exhaust caused thickened
alveolar septa, bronchiolo-alveolar hyperplasia, and emphysematous lesions, but no lung
tumors. In mice, whole exhaust, but not filtered exhaust, caused multifocal
bronchiolo-alveolar hyperplasia, multifocal alveolar lipoproteinosis, and multifocal interstitial
fibrosis. The incidence of adenocarcinomas was significantly increased in both groups (the
control group had a very high incidence rate for spontaneous tumors). In rats, there were
no significant morphological changes in animals exposed to filtered exhaust and no increase
in lung tumor incidence. In rats exposed to whole exhaust there were severe inflammatory
changes in the lungs, thickened alveolar septa, foci of macrophages, crystals of cholesterol,
hyperplastic and metaplastic lesions, and a significant increase in lung tumors
(bronchiolo-alveolar adenomas and squamous cell tumors). Heinrich et al. (1986a)
concluded that rats appeared to be more sensitive to the tumorigenic effects of diesel
exhaust than mice and hamsters. Biochemical studies of lung lavage fluids of hamsters and
rats indicated that exposure to filtered exhaust caused fewer changes than whole exhaust.
The latter produced significant increases in lactate dehydrogenase, alkaline phosphatase, acid
phosphatase, glucose-6-phosphate dehydrogenase (G6P-DH), total protein, protease (pH
5.1), and collagen. The filtered exhaust had a slight, but nonsignificant effect on G6P-DH,
total protein, and collagen. Similarly, cytological studies showed that while the filtered
exhaust had no effect on differential cell counts, the whole exhaust resulted in an increase
in leucocytes (161±43.3/jiL vs. 55.7±12.8//aL in the controls), a decrease in macrophages
(30.0±12.5 vs. 51.3±12.5/#iL in the controls), and a increase in granulocytes (125±39.7 vs.
1.23±1.14/^L in the controls). In all three cases the differences were significant at p <0.05.
There was also a slight increase in lymphocytes (5.81±4.72 vs. 3.01±1.23/mL in the controls).
May 1990 4-45 DRAFT -- DO NOT QUOTE OR CITE
-------
In CCMC-sponsored studies, Brightwell et al. (1986) evaluated the toxic effects of
whole and filtered diesel exhaust on male and female Fischer 344 rats and male and female
Syrian hamsters. The exhaust was generated with a 1.5-L engine operated under the U.S.
72 (FTP) test driving cycle. Three exhaust dilutions were tested, producing panicle
concentrations of 0.7, 12, and 6.6 mg/m3. At the highest exposure the NO, concentration
was 8±1 ppm, and the CO concentration was 32±11 ppm (concentrations of other gaseous
pollutants and for the other exposure levels were not given). The test animals (144 rats and
312 hamsters per exposure group) were exposed for 5 16-h periods per week for 2 yr. The
major histopathological finding was an increased incidence of lung tumors in rats exposed
to the two higher levels of unfiltered diesel exhaust, as compared to controls (see Chapter
7). Rats exposed to the lowest concentration of particulate matter and those exposed to
filtered diesel exhaust did not exhibit an increase in pulmonary tumors. Rats exposed to
whole exhaust also exhibited marked obstructive and restrictive airway disease. There was
no evidence of any toxicological effects in either rats or hamsters exposed to filtered diesel
exhaust
Iwai et al. (1986) exposed SPF Fischer 344 rats (24 per group) to whole or filtered
diesel exhaust for 8 h/d, 7 d/week, for 24 mo. The whole exhaust was diluted to achieve a
particle concentration of 4.9±1.6 mg/m3. The exposure atmosphere also contained 30.9±10.9
ppm NOr 1.8±1.8 ppm NO^ 13.1±3.6 ppm SO^, and 7.0±1.4 ppm CO. Body weights in both
exposed groups began to decrease after 18 mo while that of the controls did not. Lung-to-
body weight ratios of the animals exposed to the total exhaust showed a significant (p <0.01)
increase after 12 mo in comparison to control values. Spleen-to-body weight ratios in both
exposed groups were higher than control values after 24 mo. After 6 mo exposure to whole
exhaust, particles accumulated in macrophages, and type II cell hyperplasia was observed.
After 2 yr exposure, the alveolar walls had become fibrotic with mast cell infiltration and
epithelial hyperplasia. In rats exposed to filtered exhaust, after 2 yr there were only minimal
histologic changes in the lungs with slight hyperplasia and stratification of bronchiolar
epithelia and infiltration of atypical lymphocytic cells in the spleen. Lung tumors occurred
in animals exposed to whole exhaust, but not in those exposed to filtered exhaust. Malignant
lymphoma of the spleen was the main cause of death in animals exposed to whole exhaust
May 1990 4-46 DRAFT - DO NOT QUOTE OR CITE
-------
and those exposed to filtered exhaust, and the incidence rate for both groups was
significantly higher than that for the controls (p <0.05).
43. INTERACTIVE EFFECTS OF DIESEL EXHAUST COMPONENTS
This section briefly discusses the potential toxicity of the various components of diesel
exhaust and how they may interact to enhance or modify the overall effects of exposure to
the exhaust. For more specific information the reader is referred to the following review
and criteria documents. Health effects information concerning sulfur dioxide and particulate
matter is available in an EPA Air Quality Criteria Document (U.S. EPA, 1982a) and in two
addenda to that document (U.S. EPA, 1982b, 1986b). Information on acid aerosols can be
found in an EPA Issue Paper (U.S. EPA, 1989a). Information on oxides of nitrogen is
reviewed in a 1982 EPA Criteria Document (U.S. EPA 1982c). A revision of the Criteria
Document is currently in the draft stage (U.S. EPA, 1989b). Health effects associated with
carbon monoxide exposure are reviewed in an Air Quality Criteria Document (US. EPA,
1979) and in an addendum to that document (U.S. EPA, 1984b). A revision of the Criteria
Document is currently in the draft stage (U.S. EPA, 1989c). Information on formaldehyde
can be found in an EPA Health and Environmental Effects Profile (U.S. EPA, 1985), and
information on acrolein in a Health Assessment Document (U.S. EPA, 1986a). Health
effects information on polycyciic aromatic hydrocarbons is available in an EPA Drinking
Water Criteria Document (U.S. EPA, 1987).
Case studies indicate that acute exposure to diesel exhaust can produce eye and
respiratory tract irritation, headache, light-headedness, nausea, heartburn, numbness and
tingling in extremities, chest tightness, and wheezing. Several components of the exhaust,
such as NO;*, sulfuric acid, formaldehyde, and acrolein are irritants, and exposure to any one
of these compounds, at sufficiently high concentrations, could result in burning and itching
sensations in the eyes, nose, and throat, and may also produce wheezing and coughing.
When these irritants are present together, as in a diesel exhaust, their combined effect may
produce a response at concentrations below the thresholds for the individual compounds.
May 1990
4-47 DRAFT - DO NOT QUOTE OR CITE
-------
One physiological response to respiratory irritation is a reduction in the respiratory
rate. According to Kane and Alarie (1978), the two major aldehyde components of diesel
exhaust, formaldehyde and acrolein, affect the same respiratory receptor site and, therefore,
exhibit competitive agonism in this regard. Sulfur dioxide' also induces a reduction in
respiratory rates; however, in the presence of acrolein, the effect may be altered or blocked,
possibly due to the formation of an acrolein-bisulfite adduct (Kane and Alarie, 1979). Other
components of diesel exhaust such as N02 and CO, may act as respiratory stimulants and
thereby counteract the effects of the aldehydes and SO2.
Diesel exhaust can affect pulmonary function as indicated by reductions in vital
capacity, dynamic compliance, diffusing capacity, and forced expiratory flow and increases
in airway resistance. Nitrogen dioxide, sulfur dioxide, sulfuric acid, and formaldehyde have
been identified as bronchoconstrictors and one or more of these compounds may contribute
to the increases in airway resistance and associated changes in lung function parameters.
Animals exposed to diesel exhaust develop deficiencies in pulmonary defense
mechanisms. This may come about as a result of reduced mucociliary clearance, diminished
effectiveness of alveolar macrophages, or reduced immunological competence. Several
exhaust components can affect mucociliary clearance. Nitrogen dioxide levels as low as 1
to 2 ppm decrease ciliary beating in the respiratory epithelium, and a level of 4 ppm may
affect the phagocytic ability of the alveolar macrophages. Animals exposed to NO-, have
exhibited an enhanced susceptibility to bacterial and viral infections. Sulfuric acid aerosols
also affect pulmonary clearance mechanisms. Fine aerosols (4 nm) of sulfuric acid generally
have an inhibitory effect on clearance rates; however, low concentrations (0.1 mg/m3) of
large particles (7.5 pm) accelerate clearance, particularly in the upper airways. Aliphatic
aldehydes also cause reductions in mucociliary clearance.
Very limited data suggest that exposure to diesel exhaust may cause minor
neurobehavioral effects such as decreased activity in laboratory animals. Of the components
in diesel exhaust, carbon monoxide has been suspected of causing neurobehavioral effects
as a result of interference with the oxygen-carrying capacity of hemoglobin. Another
compound in diesel exhaust with similar properties is nitric oxide (NO). Carbon monoxide
forms carboxyhemoglobin (CoHb) in the blood and NO forms nitrosylhemoglobin (NOHb)
May 1990 4-48 DRAFT - DO NOT QUOTE OR CITE
-------
which rapidly oxidizes to methemogloblin (MetHb). In diesel exhaust studies, concentrations
of carbon monoxide have averaged 10 to 40 ppm, and those of nitric oxide 10 to 20 ppm.
Studies on mice indicate that a continuous exposure to 10 ppm NO can result in NOHb and
MetHb levels of 0.13 percent and 0.2 percent (Oda et al., 1976). In comparison, modeling
studies indicate that 8-h exposures to 10 ppm CO may lead to 1 to 2 percent COHb in
humans (U.S. EPA, 1984b). Carboxyhemoglobin levels of 5 percent and higher have been
associated with neurobehavioral effects.
The histopathological changes seen in the lungs of animals exposed to diesel exhaust
consist primarily of macrophage aggregation, alveolar epithelial Type II cell proliferation,
fibrosis, and, in some cases, emphysematous-like changes. Histopathological effects reported
for the various components of diesel exhaust vary in the sites and types of cells affected.
Acute exposures to NO, result in death and desquamation of bronchiolar epithelial cells and
Type I alveolar cells and proliferation of Gara cells and Type II epithelial cells. Chronic
exposures can lead to bronchiolitis and emphysematous-like lesions in the lungs. Acute
exposures to acid aerosols such as sulfuric acid result in bronchial and bronchiolar epithelial
desquamation, alveolitis, and edema. Chronic exposures cause hypertrophy and hyperplasia
of bronchial and bronchiolar epithelial secretory cells and may result in chronic bronchitis.
Subchronic exposure to fine aerosols of ammonium sulfate (0.3 jim MMAD) have been
reported to cause lung structural changes in hamsters. Of the aldehydes in diesel exhaust,
formaldehyde induces hyperplastic lesions in the epithelium of the upper respiratory tract,
and acrolein causes increased mucus secretion and bronchial epithelial proliferation which
may lead to bronchitis and bronchiolitis.
The effects reported above may be altered in the presence of diesel paniculate
matter as a result of the adsorption of the compounds onto the particles. When adsorbed
onto particles, the compounds can be transported and deposited deeper into the lungs, and
being more concentrated on the particle surface, cytotoxic effects or physiological responses
may be enhanced. Boren (1964) reported that NO, adsorbed onto carbon particles caused
parenchymal lesions in mice, while the NOz alone produced edema and inflammation, but
no lesions. Last et al. (1983) reported that collagen synthesis in lung tissue was highei in
animals exposed to N02 and ammonium sulfate aerosols than in those exposed to N02
May 1990 4-49 DRAFT - DO NOT QUOTE OR CITE
-------
alone. Kilburn and McKenzie (1978) reported that exposure to formaldehyde and acrolein
adsorbed onto carbon particles (1 to 4 nm) resulted in recruitment of polymorphonuclear
leukocytes to tracheal and intrapulmonaiy epithelial tissue but not when the aldehydes were
tested alone. In the case of SO^ particulate matter acts as a catalytic surface increasing the
rate of conversion of S02 to the more toxic sulfuric acid. The bronchoconstrictive effects
of S02 are considerably enhanced in the presence of aqueous aerosols or particulate matter.
Adsorption of hydrocarbons, such as PAHs, onto diesel exhaust particulate matter
could affect the bioavailability of these compounds. The size, effective surface area, and
chemical composition of the particles, as well as the presence of other adsorbed compounds
such as sulfuric acid, may have an effect on the rate of dissolution, metabolic degradation,
or toxicological activity of the PAHs at the level of the lung macrophages or lung epithelium
surface.
Evidence of the importance of the particulate matter in enhancing the toxic effects
of diesel exhaust is shown by chronic animal studies comparing the toxicity of whole diesel
exhaust with that caused by filtered exhaust (see Section 4.2). As indicated by pulmonary
function parameters, biochemical and histopathological changes in lung tissue, and the
occurrence of lung tumors, these studies demonstrated that animals exposed to whole diesel
exhaust exhibit more severe responses or a greater incidence of adverse responses than
animals exposed to filtered diesel exhaust.
4.4. COMPARISON OF THE EFFECTS OF DIESEL EXHAUST AND
GASOLINE EXHAUST
Diesel and gasoline engines differ considerably in the composition of their exhaust.
Diesel engines generally have higher levels of particulate matter while gasoline encines have
higher levels of carbon monoxide (Table 4-10).
The concentrations of nitrogen oxides and hydrocarbons in the exhaust vary more
with the size of the engine than with the type of engine. A greater proportion of the exhaust
hydrocarbons in diesel exhaust is associated with the particulate matter.
May 1990
4-50 DRAFT -- DO NOT QUOTE OR CITE
-------
TABLE 4-10. EMISSION RATES FOR DIESEL'
AND GASOLINE ENGINES
Panicles
CO
NOx
Hydrocarbons
Vehicle type
(g/km)
(g/km)
(g/km)
(g/km)
Light-duty diesel
0.1-0.4
OJ-3
0.5-2
0.05-0.8
Heavy-duty diesel
0.5-4
5-50
3-20
0.9-6
Light-duty gasoline
0.001-0.004
1-3
0.2-1
0.08-0.5
Heavy-duty gasoline
0.004-0.2
10-200
1-11
2-20
'Federal test procedures, hot start.
Source: Cuddihy et al.. 1981.. as adapted by Cuddihv ei al., 1984.
The different compositions of diesel and gasoline engine exhaust can result in
different toxic effects, and several studies have been undertaken to evaluate these
differences.
Brightwell et al. (1986) compared the toxic effects of filtered (FDE) and unfiltered
diesei exhaust (DE), gasoline exhaust (GE), and gasoline exhaust from an engine equipped
with a catalytic converter (CGE). The test animals, male and female Fischer 344 rats and
Syrian hamsters, were exposed for 5 16-h periods per week for 2 yr. Three diesel panicle
concentrations were used: 0.7, 2.2, and 6.6 mg/m3. At the highest diesel exposure, the CO
concentration was 32 ppm and the NO, concentration was 8 ppm (concentrations of CO and
NO, at the other two exhaust levels were not reported). The concentration of CO was 224
ppm in the GE test and 21 ppm in the CGE test. The concentration of NO, was 49 ppm
m the GE test and 7 ppm in the CGE test. The diesel-exposed rats exhibited alterations in
respiratory physiology indicative of obstructive and restrictive airway disease (specific data
not given). These changes were not seen in rats exposed to gasoline exhaust. Rats exposed
to DE and FDE had elevated levels of RBC, hemoglobin, and hematocrit, suggestive of
hypoxia. Gasoline exhaust-exposed rats had similar hematological changes as well as
significantly enlarged hearts. These changes were thought to be due to hypoxia brought
May 1990 4-51 DRAFT - DO NOT QUOTE OR CITE
-------
about by high CO levels. The blood carboxyhemoglobin level was 20 percent. The major
histopathologic^ finding was an increased incidence of lung tumors in rats exposed to the
two highest levels of unfiltered diesel exhaust. Rats exposed to the lowest concentration of
particles, those exposed to filtered diesel exhaust and those exposed to the gasoline exhausts,
did not exhibit an increase in pulmonary tumors. No other histological data were presented.
The EPA conducted a 5-yr long study on the toxic effects of gasoline engine exhaust
on beagle dogs (Stara et al., 1980). The test animals were exposed to irradiated (I) or
nonirradiated (R) exhaust with or without concurrent exposure to a mixture of sulfur dioxide
and sulfuric acid. The exposures were for 16 h/d and 7 d/week. The composition of the
exposure atmospheres is given in Table 4-11. Postexposure studies indicated that the
exhaust-exposed animals had higher levels of hemoglobin and hematocrit. Red blood cell
counts were also elevated, but not significantly. There were no effects on white blood cells
or clinical chemistry parameters (Orthoefer et al., 1980), and no significant effects on lung
collagen content (Bhatnagar, 1980). At 18 and 36 mo of exposure, pulmonary function
parameters were not significantly altered in the exposed animals. After 61 mo, dogs exposed
to irradiated exhaust had higher total expiratory resistances than controls (Lewis and
Moorman, 1980). Some evidence of right ventricular hypertrophy was also observed in dogs
exposed to irradiated exhaust. Studies conducted 2 yr after the exposures ended revealed
deficiencies in pulmonary function in dogs exposed to the nonirradiated exhaust (Gillespie,
1980). These deficiencies included abnormalities in ventilatory values, airway resistance,
blood-gas exchange, and lung volumes. Animals exposed to irradiated exhaust exhibited
similar, but less severe, changes in pulmonary function. Histological studies revealed
substantial atypical bronchiolar epithelial hyperplasia in both the R and I groups as well as
in the groups also exposed to SO, (Hyde et al., 1980). The hyperplastic lesions were derived
from nonciliated epithelial cells. Also seen were aggregations of inflammatory cells in the
distal regions of the terminal bronchioles. Data on the degree of peribronchiolar fibrosis
was not presented. The R+SOx group showed the greatest severity of bronchiolar
hyperplasia, the highest level of pulmonary resistance, and a significant increase in squamous
metaplasia in the trachea and bronchi.
May 1990
4-52 DRAFT - DO NOT QUOTE OR CITE
-------
TABLE 4-11. EXPOSURE ATMOSPHERE FOR EPA
CHRONIC TOXICITY STUDY*
Exhaust
CO
(ppm)
no2
(ppm)
NO
(ppm)
so2
(rag/m1)
Sulfates
(ppm)
Acrolein
(ppm)
Formal
(ppm)
Hydrocarbons
(ppm)
Control
4.9
0.04
0.04
0.03
—
—
—
2.7
Nomrrad. (R)
97.5
0.05
1.45
—
037
—
—
27.5
Irrad. (I)*
94.5
0.94
0.19
—
5.9
0.0198
0.57
23.9
R+SO,
98.4
0.05
1.51
0.48
21.0b
—
—
27.4
I'+SO,
94.8
0.89
0.19
0.42
99.2s
—
—
23.9
'Mean values.
"0.09 mg/mJ sulfuric aad.
*0.11 mg/m3 sulfunc aad.
tfOx (as 03) s 0.20 ppm.
Source: Malanchuk et al., 1980.
4.5. SUMMARY AND DISCUSSION
4.5.1. Toxic Effects of Diesel Exhaust on Humans
Symptoms of acute exposure to diesel exhaust include mucous membrane and eye
irritation, headache, light-headedness, nausea, vomiting, heartburn, weakness, numbness and
tingling in extremities, chest tightness, and wheezing. Exhaust odors can cause nausea,
headache, loss of appetite, and psychological stress. In one study in which a bus engine was
operated under varying load conditions, a dilution factor of 140 to 475 was needed to reduce
the exhaust odor to a threshold level.
The effects of short-term exposures to diesel exhaust have been investigated primarily
in occupationally exposed workers. In studies on underground miners, bus garage workers,
dock workers, and locomotive repairmen, changes in respiratory symptoms and pulmonary
function over the course of a workshift were generally found to be minimal and not
May 1990 4-53 DRAFT -- DO NOT QUOTE OR CITE
-------
statistically significant. However, in one study of bus garage workers, an increased frequency
of symptoms of cough, labored breathing, itching, burning and watering of the eyes, chest
tightness, and wheezing was observed. In addition, in several other studies, reductions in
forced vital capacity (FVC) and forced expiratory volume at 1 s (FEV,) were measured in
the exposed workers. The latter effects were, in some cases, attributed to nitrogen dioxide
or to nitrogen dioxide adsorbed onto paniculate matter.
Chronic effects of diesel exhaust exposure have been evaluated in epidemiological
studies of occupationally exposed workers including miners, railroad yard workers,
stevedores, and bus garage mechanics. In most cases the data have been insufficient to
establish clear correlations between effects and exposure level or length of employment.
Such correlations may be complicated by various factors including smoking habits,
occurrence of allergies, frequency of virallv induced respiratory problems, variations in
exhaust composition and workplace ventilation, and the possible presence of other workplace
air pollutants. In a few of these studies a higher prevalence of respiratory symptoms,
primarily cough and phelgm, were observed among the exposed workers. These symptoms
were usually not accompanied by significant changes in pulmonary function. However,
reductions in FVC and FEV1 were reported in a few studies. Reductions in FEFjq and
FEF75 have also been reported. In only one case was there sufficient evidence to indicate
that exposure to diesel exhaust was associated with chronic respiratory disease.
4.5.2. Toxic Effects of Diesel Exhaust on Animals
Animal studies on the toxic effects of diesel exhaust have involved acute, subchronic,
and chronic exposure regimes. In acute exposure studies, toxic effects have been associated
primarily with high concentrations of carbon monoxide, nitrogen dioxide, and aliphatic
aidenydes. In short-term ana chronic exposure studies, toxic effects have oeen associated
with exposure to exhaust particulate matter. Toxic effects of diesel exhaust in various animal
species are summarized in Tables 4-12 to 4-15. The data are arranged on the tables in
general order of increasing concentration of the exhaust particulate matter. The data for
short-term exposures (Table 4-12) indicate minimal effects on pulmonary function even at
May 1990
4-54 DRAFT - DO NOT QUOTE OR CITE
-------
exhaust concentrations sufficiently high to cause histological and cytological changes in the
lungs. Exposures for several months or longer resulted in accumulation of particles in the
lungs, increase in lung weight, increases in macrophages and leukocytes, macrophage
aggregation, hyperplasia of alveolar epithelium, and thickening of the alveolar septa. Similar
effects have also been observed in studies in which animals have been chronically exposed
to diesel exhaust. In addition, chronic exposures have also resulted in reduced growth rates
and altered pulmonary function. Exposures to diesel exhaust have also been associated with
increased susceptibility to respiratory tract infections, neurological or behavioral changes,
some hematological changes, and morphological alterations in the liver.
4.5.2.1. Effects on Growth
The data presented in Table 4-13 show that growth of laboratory animals can be
impaired by chronic exposures to diesel exhaust, particularly when the exhaust dilution
results in particle concentrations of 2 mg/m3 or higher and the exposure periods are 16 or
more h/d. In studies in which the daily exposures were only 6 to 8 h/d, no effects on growth
were seen even at particle levels of 6 to 8 mg/m3.
4-5.2.2. Effects on Pulmonary Function
Alterations in pulmonary function that have been observed in animals chronically
exposed to diesel exhaust include decreases in vital capacity, residual lung volume, diffusing
capacity, dynamic lung compliance, and expiratory flow rates, as well as increases in airway
resistance (Table 4-14). These effects were not reported in all exposure studies or for all
species tested. They generally appeared only after prolonged exposures. The lowest
exposure levels (expressed in terms of the diesel particulate concentration) that resulted in
impairment of pulmonary function occurred at 2 mg/m3 in cynomolgus monkeys, 11.7 mg/m3
in cats, 1-5 and 3.47 mg/m3 in rats, and 4.24 and 6 mg/m3 in hamsters.
4.5.23. Histopathological and Histochemical Effects
Histological studies have demonstrated that chronic exposure to uiesel exhaust can
result in adverse effects on respiratory tract tissue (Table 4-15). Typical findings include
May 1990 4-55 DRAFT ~ DO NOT QUOTE OR CITE
-------
TABLE 4-12. SHORT-TERM TOXICITY OF DIESEL EXHAUST TO LABORATORY ANIMALS
Species
Exposure
Period
Panicles
(mg/mJ)
CO
(ppm)
NO,
(ppm)
SO,
(ppm)
Effects
References
Rat,
mouse,
hamster
20 h/d
7 d/week
86-90 d
1.5
—
—
—
Increase in lung weight.
Increase in thickness of
alveolar walls.
Kaplan et al., 1982.
Kilt,
mouse
7 h/d
5 d/week
19 weeks
4.38
No effects on lung function.
Increase in neutrophils
and lung proteases.
Ntacrophjge aggregation.
Mauderly et al,
1981
Kai
20 h/d
5 5 d/week
9 weeks
60
Macrophage aggregation. Incr.
in polymorphonuclear
leukocytes.
Type II cell proliferation.
Iliickened alveolar walls.
White and G;irg,
1981
Guinea pig
20 h/d
7 d/week
56 d
6.32
174
2.3
2 1
Incr. in relative lung wt
Macrophage aggregation
1 lypertrophy of goblet cells.
1 ocal hyperplasia of alveolar
epithelium
Wiester et al., 1980
Git
20 h/d
28 d
6.4
14 6
2 1
2 1
Few effects on lung function,
l ocal pneumonitis or
alveolitis.
I'cpclko et al,
1980a
Guinea pig
20 h/d
7 dMeek
28 d
6 83*
167
2.9
1 9
Increase in pulmonary now
resistance. Bradycardia
Wiester et al, 1980
'Irradiated exhaust.
-------
TABLE 4-13. EFFECIS OF CHRONIC EXPOSURES TO DIESEL
EXHAUST ON GROWTH OF LABORATORY ANIMALS
1 Exposure
Particles
(()
no2
SO J
Specie*
period
mg/mJ
(ppm)
(ppm)
(PPn>)
bf feels
References
Rat,
20 h/d
025
2 7b
01*
_
Reduced body weight in
Schreck el al, 1981
guinea pig
5 dAveek
075
4 4*1
0 27°
-
rats al 1.5 mg/m3.
106 weeks
1 5
7 lb
0 5e
-
K.«t.
7 h/d
2
II s
1 5
0 81
No effects on growth No
Ix-wu el al, 1986
monley
5 dMeck
hematological effects
24 mo
R.u*
16 h/d
ou
1 23
008
038
Dose-dependent decrease in
Miinishi et al, 1986b
5 JAveek
041
2 12
0 26
1 06
growth Decreased survival
30 mo
108
3%
0 70
2 42
between 24 to 30 nto
231
7 10
1 41
4 70
Rat,
19 h/d
4 24
i: 5
1 5
I |
Reduced body weights In-
llcinnch et al, 1986a
mouse
5 d/weck
creased mortality in mite
24 mo
Rat
16 h^eriod
07
-
_
-
Growth reduced at 2 2 and
Bnghtwell et al, 1986
S periods/week
22
-
-
-
6 6 mg/mJ.
24 mo
66
3'>7
068
-
Rat
7 ll/d
035
2't
005
-
No effect on growth or
McCellan el al, 1986
SdAwek
347
165
034
-
mortality rates
30 mo
708
yn
068
-
Rat
6 h/d
83
5i»
44
-
No effect on growth or
Karagianesct al, 1981
5 d/weck
mortality rates
20 mo
Vinegar et al, 1981a
Hamster
8 h/d
6
-
-
-
No effect on growth Sig-
Vinegar el al, 1981b
7 days/week
12
nificant Increase in lung
6 mo
wt and lung/body wt ratio
'D.it.i (or tests with light-duty engine Similar result! Willi heavy-duly engine
bhstimated from graphically depicted m.iss concentration data
'Estimated from graphically presented mass concentration ddla fur NO, (.issuming 90 percent NO and 10 percent NOJ
-------
TABLE 4-14. EFFECTS OF DIESEL EXHAUST ON
PULMONARY FUNCTION OF LABORATORY ANIMALS
Species
Exposure
penod
Parades
(mg/tn1)
CO
(ppm)
NO;
(ppm)
SO:
(Ppm)
Effects
Reference
Rat
20 h/d
5.5 d/week
24 mo
15
7.0
-
—
Increase in FRC,*
ER." RV.e MEF<>d
MEF^b FEVai*
Grass, 1981b
Rat
7 h/d
5 d/week
24 mo
11.5
15
0.8
No effect on pul-
monary function.
Lewis et aL 1986
Monkey
7 h/d
5 dAneek
24 mo
2
IU
15
0.8
Significant decrease
in FEFjq/FVC and
FEF«/TLC.'
Lewis et aL 1986
Rat
7-8 h/d
S d/week
24 mo
3.9
185
1 2
3.1
No effect on minute
volume, compliance,
or romance.
Heinnch et al.. 1982
Rat
19 h/d
5 d/week
140 weeks
4.24
12-5
;5
1.1
Decrease in dynamic
lung compliance.
Increase in airway
Heinnch et al.. 1986a
Hamster
19 h/d
5 d/week
120 weeks
4.24
125
15
1.1
Significant increase
in auway resistance.
Decrease in lung
compliance.
Heinnch et aL 1986a
Hamster
8 h/d
7 d/week
6 mo
6
12
—
—
—
Decrease in vital
capacity. RV,C and
diffusing capacity.
Increase m static
deflation lung
volume.
Vinegar et aL 1981a. b
Cat
8 h/d
7 day/week
2L25 years
6J*
11.7°
20.2
33J
2.7
44
2.1
5.0
Decrease in vital
capacity, total lung
capacity, and
diffusing capacity
after 2 yr.
Pepelko et al. 1980b.
1981; Moorman et aL
198S
Rat
16 h/penod
5 penodsAveek
24 mo
0.7
22
6.6
32
—
—
Adverse pulmonary
effects at 6.6 mg/m3
(oo other data).
Bnghtwell et aL 1986
Rat
7 h/d
5 dAweek
30 mo
035
3.47
7.08
Z9
16.5
39.7
0.05
034
0.68
—
Diffusing capacity,
lung compliance
reduced at 35 and 7
mg/to'
McClellan et aL 1986
'Functional residual capacity.
^Expiratory reserve.
cRciudual lung volume.
dMax. expiratory flaw at 40 percent VC
'Forced expiratory volume in 0.1 s.
'Total lung capacity.
tFint year of exposure.
h Second year of exposure.
i»-58
-------
TABLE 1-15. HIS rOI'A'I'l IOLOGICAL EFFECI'S OF DIESEL EXHAUST
IN THE LUNGS OF LABORATORY ANIMALS
Species
Exposure
period
Tarliclcs
(mg/m1)
CO
(l'P,n)
NO,
(ppm)
SO,
(ppm)
Effects
References
It.il,
mouse,
li.misler
20 h/d
7 d/week
80 90 d
I 5
Increase in lung weight Increase Kaplan el al,
in thickness or alveolar waits 1982
Kit
7 li/J
5 d/week
2-1 mo
II 5
I 5
08
Mullifoc.il hislocyhais Inll.im- lrwisetal,
malory changes T\pe II cell prolif- 19B6
eralion Fibrosis
Monkey
It.il
7 h/d
5 d/week
2-1 mo
16 h/d
J d/week
30 mo
Oil
041
108
2 31
II 5
1 23
2 12
396
710
15
008
0 26
0 70
1 41
OH
038
1 06
2 42
4 70
Macrophage aggregation No Lewis el al,
fibrosis, inflammation, or em- 1986
physenu
"type II cell hyperplasia and lung lshinishi cl al,
tumors seen al >04 mg/mJ. 1986b
I *sions in trachea and bronchi
ll.imstcr
7 -8 h/d
5 d/week
120 weeks
39
IK 5
12
31
b0 percent adenomatous cell
proliferation.
lleinnch et at,
1982, 1985
K.il
Hamster
19 h/d
5 d/week
140 weeks
19 h/d
5 dAveck
120 weeks
4 24
4 24
\Z5
125
I 5
IS
I I
'thickened alveolar septa
Macrophage aggregation
Inflammatory changes
Hyperplasia Lung tumors
Thickened alveolar sepia
Bronchioloalveolar hyperplasia.
Emphysema
I leinnch el at,
1986a
lleinnch el al.
1986a
-------
TABLE 4-15. (continued)
Exposure Particles CO NOz SOt
Specks Period (mg/mJ) (ppm) (ppm) (ppm) Effect* References
Mouse
19 h/d
Sd/week
120 weeks
4 24
125
IS
11
llnmchiolo-atveolar hyper-
plasia Alveolar lipopro-
Iciiiiisis lllirosis.
Ileinrich el al,
1986a
(itnnea Pig
*r
I
O
o
K.il
K.il
Hat
r.ii
I l.imster
20 h/d
SSd/Week
24 iihi
7 h/d
5 dAveek
30 mo
16 h/d
5 penodsAvecL
24 mo
6 h/d
Sd/week
20 mo
8 h/d
7 day/wcck
2 yr
8 h/d
7 dAveek
6 mo
025
0 75
15
60
0 35
3 47
708
07
22
66
83
634
11.70
6
12
29
lo5
39 7
12
5u
20 2
313
005
034
068
4-6
2.7
44
2 I
SO
Minimal response al 0 25 and
ultrastnidural changes at 0 75
mg/m1. Thickened alvtol.ir
membranes, cell proliferation, and
fibrosis al 6 0 mg/mJ Increase in
PMN leukocytes al 0 75 mg/m1 and
I 5 mg/mJ
Alveolar and bronchioldr epithelial
metaplasia at 3J and 7 0 mg/m1.
Fibrosis at 7 0 mg/mJ Inflam-
matory changes
Increased incidence of lung tumois
(no other data)
Macrophage aggregation Alveolar
cell hypertrophy. Inlenlitial
fibrosis. Emphysema
Bamhart el al,
1982
VibI.iI cl al,
1981, 1982
McClcllan el al,
1986
Brighlwell et al,
1986
Karagianes et
al., 1981
Macrophage aggregation. Plopper cl al,
Bronchiolar epithelial metaplasia 1983
Type II cell hyperplasia llyde el al,
Peribronchiolar fibrosis 1985
Macrophage accumulation Pepclko, 1982b
Thickened alveolar lining. Type II
cell hyperplasia Edema Increase
in collagen
-------
alveolar histiocytosis, macrophage aggregation, tissue inflammation, increase in polymorpho-
nuclear leukocytes, hyperplasia of bronchiolar and alveolar Type II epithelial cells, thickened
alveolar septa, edema, emphysema, and fibrosis. Lesions in the trachea and bronchi have
also been observed in some studies. The exposure levels (exhaust particle concentrations)
at which one or more of these effects occurred were 3.9, 4.24, and 6 mg/m3 in hamsters, 1,
2,2.2,3.5, 4.24,6.0, and 8.3 mg/m3 in rats, 1~5 mg/m3 in guinea pigs, and 4.24 mg/m3 in mice.
Associated with these histopathological findings were various histochemical changes in the
lung, including increases in lung DNA, total protein, alkaline and acid phosphatase,
glucose-6-phosphate dehydrogenase (G6P-DH), increased synthesis of collagen, and release
of inflammatory mediators such as leukotriene LTB and prostaglandin PGR..
While the overall laboratory evidence is that prolonged exposure to diesel exhaust
paniculate matter results in histopathological changes in the lungs of exposed animals,
several studies have also demonstrated that histopathological changes do not occur when
particle concentrations are low, even following lifetime exposures. These no-effect levels
were reported to be 2 mg/m3 for cynomologous monkeys, 0.7 mg/m3 for rats, and 0.25 mg/m3
for guinea pigs.
The histopathological effects of diesel exhaust particulate matter appear to be
strongly dependent on relative rates of pulmonary deposition and clearance. Gearance of
panicles from the lungs is a multiphasic process involving phagocytosis by alveolar
macrophages. At panicle concentrations of 1 mg/m3 or above, pulmonary clearance
becomes reduced and focal aggregations of panicle-laden macrophages appear in the lungs,
particularly in the alveoli and peribronchiolar regions, as well as in the hilar and mediastinal
lymph nodes. With prolonged exposures, the continued presence of aggregations of
soot-laden macrophages appears to result in an inflammatory tissue response, septal
thickening, epithelial proliferation, and fibrosis. The critical factor in this process appears
to be the minimal exposure, in terms of the concentration of the paniculate matter, as well
as in the length and frequency of the exposure periods, at which focal aggregations of
soot-laden macrophages occurs. This exposure level may vary from species to species
depending on rate of uptake and pulmonary deposition, pulmonary clearance rates, the
relative size of the macrophage population per unit of lung tissue, the rate of recruitment
May 1990 4-61 DRAFT - DO NOT QUOTE OR CITE
-------
of macrophages and leukocytes, and the relative efficiency of the mucociliary and lymphatic
transport systems.
4.5.2.4. Effects on Defense Mechanisms
Several animal studies have indicated that exposure to diesel exhaust can reduce an
animal's resistance to respiratory tract infections. This effect, which can occur even after
only a few hours of exposure, does not appear to be caused by direct impairment of the
lymphoidal or splenic immune systems; however, in one study, interferon levels and
hemagglutinin antibody levels were adversely affected in the exposed animals.
45J2~5. Neurological and Behavioral Effects
Behavioral effects have been observed in a few studies of animals exposed to diesel
exhaust. These effects include reductions in spontaneous locomotor activity (SLA), as well
as diminished learning ability. Neurophysiological studies have shown possible impairment
of nerve impulse transmission.
4.5.3. Interactive Effects
The physiological and toxicological effects of diesel emissions can be enhanced or
altered as a result of the combined actions or interactions of the various exhaust
components. Nitrogen dioxide, sulfuric acid, SO^ and aliphatic aldehydes are chemical
irritants, and one or more of these components may contribute to the eye and respiratory
tract irritation observed in acute exposures to diesel exhaust. Several of these exhaust
components, including NCK, sulfuric acid, SO,, and aliphatic aldehydes, are broncho-
constrictors. Sulfur dioxide and aliphatic aldehydes cause reductions in respiratory rates,
while NO, may cause an increase. On an individual basis, any of these compounds could be
expected to affect pulmonary function. However, the responses seen in short-term exposure
studies with diesel exhaust are minimal, suggesting that concentrations are too low or that
there are interactive effects diminishing the response.
May 1990
4-62 DRAFT - DO NOT QUOTE OR CITE
-------
Impaired resistance to respiratory tract infections have been reported in some diesel
exhaust studies. This may come about as a result of reduced mucociliary clearance,
diminished effectiveness of alveolar macrophages, or reduced immunological competence.
Nitrogen dioxide, acrolein, formaldehyde, sulfuric acid, and inorganic sulfates may contribute
to this effect as a result of their immediate cytotoxic activity and the resulting impairment
of the mucociliary clearance mechanism in the respiratory tract.
The histological and cytological changes seen in the lungs of animals exposed to diesel
exhaust have been attributed primarily to the accumulation and aggregation of exhaust
particles in lung macrophages. Other exhaust components, panicularly NO^ aliphatic
aldehydes, and sulfuric acid are also known to cause histopathological effects in the lungs.
Most effects are directed at the bronchial and bronchiolar epithelial cells, with resulting
hypertrophy and hyperplasia. Exposures to NO; can also lead to emphysematous lesions
which have also been reported in some diesel studies.
Adsorption of inorganic or organic compounds on diesel particulate matter may alter
the chemical composition and toxicological effects of the exhaust. Sulfur dioxide and
nitrogen dioxide can be adsorbed onto exhaust particles. The S02 could be catalytically
converted to sulfuric acid. Attached to the exhaust particles, these substances may be
carried deeper into lungs where they might have a more direct and potent effect on
epithelial cells or on alveolar macrophages engulfing the particles. In addition, adsorption
of these compounds may alter the acidity of the particles and thereby cause physiochemical
changes affecting the hydrocarbon fraction of the particles. Binding of the hydrocarbons
and, consequently, bioavailablity may be altered by changes in pH, and this may result in
more direct toxicological activity at the level of lung macrophages or lung epithelium surface.
4.5.4. Comparisons with Gasoline Exhausts
Laboratory studies indicate that prolonged exposure to gasoline engine exhaust can
result in histopathological changes in the lungs of exposed animals. These changes consist
primarily of atypical bronchiolar epithelial hyperplasia causing, in some areas, partial
occlusion of the bronchiolar lumina. Hyperplastic changes involving bronchiolar epithelial
May 1990
4-63 DRAFT - DO NOT QUOTE OR CITE
-------
tissue are also a characteristic feature of chronic exposures to diesei exhaust. However, in
the latter case there can aiso be extensive involvement of Type II cells in the alveolar region.
This suggests that the mechanisms involved in the lung response to the two types of exhaust
are similar but that the target tissue may be slightly altered due to the different physical
and/or chemical composition of the exhausts. The higher levels of particulate matter in
diesei exhaust may allow for a greater concentration of the active components of the exhaust
in the deeper regions of the lungs. The effectiveness of diesei exhaust in inducing these
tissue responses might also be enhanced by the presence of other exhaust components. The
sulfur dioxide and sulfuric acid in diesei exhaust are known to induce pulmonary hyperplastic
lesions and these components may have additive or synergistic effects in exposures to diesei
exhaust.
May 1990 4-64 DRAFT -- DO NOT QUOTE OR CITE
-------
4.6. REFERENCES
Ames, R. G.; Attfield, M. D.; Hankinson, J. L.; Hearl, F. J.; Reger, R. B. (1982) Acute respiratory effects of
exposure to diesel emissions in coal miners. Am. Rev. Resp. Dis. 125(1): 39-42.
Ames, R. G.; Hall, D. S.; Reger, R. B. (1984) Chronic respiratory effects of exposure to diesel emissions in
coal mines. Arch. Environ. Health 39(6): 389-394.
Attfield, M D. (1978) The effect of exposure to silica and diesel exhaust in underground metal and
non-metal miners. In: Industrial Hygiene for Mining and Tunneling. Proceedings of an ACGIH topical
symposium, Nov. 6-7, 1978. As reported in Ames et al., 1984.
Attfield, M. D.; Trabant, G. D.; Wheeler, R. W. (1982) Exposure to diesel fumes and dust in a potash mine.
Ann. Indust. Hyg. 26: 817-831.
Barnhart, M. I.; Chen. S.-T.; Salley, S. O.; Puro, H. (1981) Ultrastructure and morphometry of the alveolar
lung of guinea pigs chronically exposed to diesel engine exhaust: six month's experience. J. Appl.
Toxicol. 1: 88-103.
Barnhart, M. I.; Salley, S. O.; Chen, S.-T.; Puro, H. (1982) Morphometry ultrastructural analysis of alveolar
lungs of guinea pigs chronically exposed by inhalation to diesel exhaust (DE). In: Lewtas, J., ed.
Toxicological Effects of Emissions from Diesel Engines. New York, NY: Elsevier, pp. 183-200.
Battigelli, M. C; Mannella, R.; Hatch, T. F. (1964) Environmental and clinical investigation of workmen
exposed to diesel exhaust in railroad engine houses. Indust. Med. Surg. 33: 121-124.
Battigelli, M. C (1965) Effects of diesel exhaust. Arch. Environ. Health 10: 165-167.
Bhatnagar, R. S (1980) Collagen and prolyl hydroxylase levels in lungs in dogs exposed to automobile
exhaust and other noxious gas mixtures. In: Stara. J. F.; Dungworth, D. L.; Orthoefer. J. G.; Tyler, W.
S., eds. Long-Term Effects of Air Pollutants in Canine Species. Cincinnati. OH: U.S. Environmental
Protection Agency, Environmental Criteria and Assessment Office. EPA report No.
EPA-600/8-80-014.
Bhatnagar, R. S.; Hussain, M. Z.; Danner, R. M.; McMillan, L.; Lee, S. D. (1980) Biochemical alterations in
lung connective tissue in rats and mice exposed to diesel emissions. In: Pepelko, W. E.; Danner, R.
M.; Clarke, N. A., eds. Health Effects of Diesel Engine Emissions. Cincinnati, OH: U.S.
Environmental Protection Agency, Office of Research and Development. EPA report No. EPA-
600/9-80-057b. pp. 557-570.
Bice. D. E.; Mauderly, J. L.; Jones, R. K.; McClellan, R. O. (1985) Effects of inhaled diesel exhaust on
immune responses after lung immunization. Fund. AppL Toxicol. 5: 1075-1086.
Boren, H. G. (1964) Carbon as a carrier for irritant gases. Arch. Environ. Health 8: 119-124.
Brightwell, J.; Fouillet, X.; Cassano-Zoppi, A. L.; Gatz, R.; Duchosal, F. (1986) Neoplastic and functional
changes in rodents after chronic inhalation of engine exhaust emissions. In: Ishinishi, N.; Koizumi, A.;
McClellan, R. O.; Stoeber, W., eds. Carcinogenic and Mutagenic Effects of Dioei Engine Exhaust.
Amsterdam: Elsevier, pp. 471-485.
May 1990
4-65
DRAFT-DO NOT QUOTE OR CITE
-------
Campbell. K. L; George. E. L.; Washington. L S. Jr. (1980) Enhanced susceptibility to infection in mice after
exposure to dilute exhaust from light duty diesel engines. In: Pepelko, W. E.; Danner, R. M.; Clarke,
N. A., eds. Health Effects of Diesel Engine Emissions. Cincinnati, OH: U.S. Environmental
Protection Agency, Office of Research and Development. EPA report No. EPA-600^-80-057b. pp.
772-785.
Campbell, K. L: George, E. L.; Washington, I. S. Jr. (1981) Enhanced susceptibility to infeaion in mice after
exposure to dilute exhaust from light duty diesel engines. Environ. InternatL 5: 377-382.
Castranova, V.; Bowman, L.; Reasor, M. J.; Lewis, T.; Tucker, J.; Miles, P. R. (1985) The response of rat
alveolar macrophages to chronic inhalation of coal dust and/or diesel exhaust Environ. Res. 36(2):
405-419.
Chan, T. L.; Lee, P. S.; Henng, W. E. (1981) Deposition and clearance of inhaled diesel exhaust particles
in the respiratory tract of Fischer rats. J. AppL Toxicol. 1: 177-182.
Chaudhari, A.; Farrer, R. G.; Dutta, S. (1980) Effect of exposure to diesel exhaust on pulmonary
prostaglandin dehydrogenase (PGHD) activity. In: Pepelko. W. E.; Danner, R. M.; Clarke, N. A., eds.
Health Effects of Diesel Engine Emissions. Cincinnati. OH: U.S. Environmental Protection Agency,
Office of Research and Development. EPA report No. EPA-600/9-80-057b. pp. 532-537.
Chaudhari, A.; Farrer, R. G.; Dutta, S. (1981) Effea of exposure to diesel exhaust on pulmonary
prostaglandin dehydrogenase (PGHD) activity. J. Appl. Toxicol. 1(2): 132-134.
Chaudhari, A.; Dutta, S. (1982) Alterations in tissue glutathione and angiotens in convening enzyme due
to inhalation of diesel engine exhaust J. Toxicol. Environ. Health 9: 327-337.
Chen, K. C (1986) Induction of aryl hydrocarbon hydroxylase in rat tissue following intratracheal instillation
of diesel paniculate extract and benzo[a]pyrene. J. Appl. Toxicol. 6: 259-262.
Chen, K. C; Vostal, J. J. (1981) Aryl hydrocarbon hydroxylase activity induced by injected diesel panicle
extract vs. inhalation of diluted diesel exhaust J. Appl. Toxicol. 1: 127-131.
Chen. S.; Weller. MA; Barn hart, M. I. (1980) Effects of diesel engine exhaust on pulmonary alveolar
macrophages. Scanning Electron. Microsc. 3: 327-338.
CRC Coordinating Research Council. (1979) Development and evaluation of a method for the measurement
of diesel exhaust odor using a diesel odor analysis system (DOAS). CRC Repon No. 507.
Cuddihy, R. G.; Griffith, W. C; dark, C R_; McQellan, R. O. (1981) Potential health and environmental
effects of light-duty diesel vehicles. Albuquerque, NM: Inhalation Toxicology Research Institute.
Repon No. LMF 89.
Cuddihy, R. G.; Griffith, W. C; McQellan, R. O. (1984) Health risks from light-duty diesel vehicles.
Environ. Sci. Technol. 18: 14-21A.
Dziedzic, D. (1981) Differential counts of B and T lymphocytes in the lymph nodes, circulating blood and
spleen after inhalation of high concentration of diesel exhaust J. Appl. Toxicol. 1: 111-115.
May 1990
DRAFT-DO NOT QUOTE OR CITE
-------
Ediing, G; Axelson, O. (1984) Risk factors of cornary heart disease among personnel in a bus company. Int.
Arch Occup. Environ. Health 54: 181-183.
Ediing, C; Anjou, C-G.; Axelson, O.; Kling, H. (1987) Mortality among personnel exposed to diesel exhaust
Int. Arch Occup. Environ. Health 59: 559-S65.
Eskelson, C D.; Strom, K. A.; Vostal, J. J.; Misiorowski. R. L.; Chrapil, M. (1981) Lipids in the lung and lung
lavage fluid of animals exposed to diesel paniculate. Toxicologist 1: 74.
Fedan, J. S.; Frazer, D. G.; Moorman, W. J.; Attfield. M. D.; Franczak, M. S.; Kosten, C J.; Cahill, J. F.;
Lewis, T. R.; Green, F. H. Y. (1985) Effects of a two-year inhalation exposure of rats to coal dust
and/or diesel exhaust on tension response of isolated airway smooth muscle. Am. Rev. Resp. Dis.
131(4): 651-655.
Forrest, L; Lee, W. B.; Smalley, W. M.; Strum, J. C (1980) Assessment of environmental impacts of light
duty vehicle dieselization. Aerospace Corporation. Report No. DOT-TSC-NHTSA-80-5.
Gamble, J.; Jones. W. (1983) Respiratory effects of diesel exhaust in salt miners. Am. Rev. Respir. Dis. 128:
389-394.
Gamble. J.; Jones, W.; Hudak. J. (1983) An epidemiological study of salt miners in diesel and nondiesel
mines. Am. J. Ind. Med. 4: 435-458.
Gamble, J.; Jones, W.; Hudak, J.; Merchant, J. (1978) Acute changes in pulmonary function in salt miners.
ACGIH Conference on Mining and Tunneling. Denver, Colorado, November, 1978. pp. 119-128.
As reported in Gamble and Jones, 1983a.
Gamble. J.; Jones, W.; Minshall, S. (1987a) Epidemiological-environmental study of diesel bus garage
workers: acute effects of NOz and respirable particulate matter on the respiratory system. Environ.
Res. 42: 201-214.
Gamble. J.; Jones, W.; Minshall, S. (1987b) Epidemiological-environmental study of diesel bus garage
workers: chronic effects of diesel exhaust on the respiratory system. Environ. Res. 44: 6-17.
Garg, B.D. (1983) Histologic quantitation of macrophage aggregates in the lungs of diesel exhaust exposed
rats. Acta StereoL 2 (suppl. I): 235-238.
Gillespie, J. R. (1980) Review of the cardiovascular and pulmonary function studies on beagles exposed for
69 months to auto exhaust and otheT air pollutants. In: Stara. J. F.; Dungworth, D. I_; Orthoefer, J.
G.; Tyler, W. S., eds. Long-Term Effects of Air Pollutants in Canine Species. Cincinnati, OH: U.S.
Environmental Protection Agency, Environmental Criteria and Assessment Office. EPA report No.
EPA-600/8-80-014.
Green, F. H. Y.; Boyd, R- Danner-Rabovsky, J.; Fisher, M. J.; Moorman, W. J.; Ong, T.-M.; Tucker, J.;
Vallyathan, V.; Ehong, W.-Z.; Zoldak, J,; Lewis, T. (1983) Inhalation studies of diesel exhaust and
coal dust in rats. Scan. J. Work Environ. Health 9: 181-188.
Gross. K. B. (1981a) Pulmonary function testing of animals chronically exposed to diluted diesel exhaust for
267 days. Environ. Intl. 5: 331-337.
May 1990
4-67
DRAFT-DO NOT QUOTE OR CITE
-------
Gross, K. B. (1981b) Pulmonary function testing of animals chronically exposed to diluted diesel exhaust.
J. AppL ToxicoL 1(2): 116-123.
Hahon, N.; Booth, J. A.; Wheeler. R. (1982) Aaivity of diesel engine emission particulates on the interferon
system. Environ. Res. 28: 443-455.
Hahon, N.; Booth, J. A.; Green. F.; Lewis, T. R. (1985) Influenza virus infection in mice after exposure to
coal dust and diesel engine emissions. Environ. Res. 37: 44-60.
Hare, C T.; Springer, K. J. (1971) Public response to diesel engine exhaust odors. Final report to EPA.
Contract No. CPA-70-44.
Hare, C T.; Springer, K. J.; Somers, J. H.; Huls, T. A. (1974) Public opinion of diesel odor. Paper No.
740214, presented at the Annual Meeting of the Society of Automotive Engineers.
Hatch, G. E.; Boykin, E.; Graham, J. A; Lewtas, J.; Pott, F.; Loud, K.; Mumford, J. L. (1985) Inhalable
particles and pulmonary host defense: In vitro and in vivo effects of ambient air and combustion
particles. Environ. Res. 36: 67-80.
Heinnch, U.; Peters. L.; Funcke, W.; Pott, F.; Mohr, U.; Stoeber, W. (1982) Investigation of toxic and
carcinogenic effects of diesel exhaust in long-term inhalation exposure of rodents. In: Lewtas, J., ed.
Toxicological Effects of Emissions from Diesel Engines. New York, NY: Elsevier, pp. 225-242.
Heinrich, U.; Pott, F.; Mohr, U.; Stoeber, W. (1985) Experimental methods for the detection of the
carcinogenicity and/or cocarcinogenicity of inhaled polycyclic-aromatic-hydrocarbon-containing
emissions. In: Mass, M. J.; Kaufman, D. G.; Siegfried, J. M.; Steele, V. E.; Nesnow, A_ eds.
Carcinogenesis A Comprehensive Survey, vol. 8. Cancer of the Respiratory Tract, Predisposing
Factors. New York, NY: Raven Press, pp. 131-146.
Heinrich, U.; Muhle, H.; Takenaka, S.; Ernst, H.; Fuhst, R.; Mohr, U.: Pott F.; Stoeber, W. (1986a) Chronic
effects on the respiratory tract of hamsters, mice and rats after long-term inhalation of high
concentrations of filtered and unfiltered diesel engine emissions. J. Appl. Toxicol. 6(6): 383-395.
Heinnch, U.; Pott, F.; Rittinghausen, S. (1986b) Comparison of chronic inhalation effects in rodents after
long-term exposure to either coal oven flue gas mixed with pyrolized pitch or diesel engine exhaust.
In: Ishinishi, N.; Koizumi, A.; McCIellan, R.; Stober, W., eds. Carcinogenic and Mutagenic Effects of
Diesel Engine Exhaust. Amsterdam: Elsevier, pp. 441-457.
Henderson, R. F.; Sun, J. D.; Jones, R- K.; Mauderly, J. L.: McQellan. R. O. (1983) Biochemical and
cytological response in airways of rodents exposed in life span studies to diesel exhaust. Am. Rev.
Resp. Dis. 127: 164.
Henderson, R. F.; PickrelL, J. A.; Jones. R. K.; Sun, J. D.; Benson, J. M.; Mauderly, J. L_; McClcllan, R. O.
(1988) Response of rodents to inhaled diluted diesel exhaust: biochemical and cytological changes
in broncho-alveolar lavage fluid and in lung tissue. Fund. Appl. Toxicol. 11: 546-567.
Hinners, R. G.; Burkart, J. K.; Malanchuk, M.; Wagner, W. D. (1980) Facilities for diesel exhaust studies.
In: Pepelko, W. E; Danner, R. M; Clarke, N. A., eds. Health Effects of Diesel Engine Emissions.
Cincinnati, OH: U.S. Environmental Protection Agency, Office of Research and Development EPA
report No. EPA-600/9-SO-057b. pp. 681-697.
May 1990 4-68 DRAFT-DO NOT QUOTE OR CITE
-------
Hyde, D.; Orthoefer, J. G.; Dungworth, D.; Tyler, W.; Carter, R.; Lum, H. (1980) Morphometric and
morphologic evaluation of pulmonary lesions in beagle dogs chronically exposed to high ambient
levels of air pollutants. In: Stara, J. F.; Dungworth, D. L.; Orthoefer, J. G.; Tyler, W. S., eds.
Long-Term Effects of Air Pollutants in Canine Species. Cincinnati. OH: U.S. Environmental
Protection Agency, Environmental Criteria and Assessment Office. EPA report No. EPA-
600/8-80-014.
Hyde, D. M.; Plopper, C G.; Weir, A. J.; Murnane, R. D.; Warren, D. L.; Last, J. A.; Pepelko, W. E. (1985)
Peribronchiolar fibrosis in lungs of cats chronically exposed to diesel exhaust Lab. Invest. S2(2):
195-206.
Ingats, M. N. (1985) Improved mobile source exposure estimation. Ann Arbor, MI: U.S. Environmental
Protection Agency, Office of Mobile Source Air Pollution Control. EPA report No. 460/3-85-002.
Ishinishi, N.; Koizumi, A.; McClellan, R.; Stoeber, W., eds. (1986a) Carcinogenic and Mutagenic Effects of
Diesel Engine Exhaust. Amsterdam: Elsevier.
Ishinishi. N.; Kuwabara, N.; Nagase, S.; Suzuki, T.; Ishiwata, S.; Kohno, T. (1986b) Long-term inhalation
studies on effects of exhaust from heavy and light duty diesel engines on F344 rats. In: Ishinishi, N.;
Koizumi. A; McClellan. R. O.; Stoeber. W., eds. Carcinogenic and Mutagenic Effects of Diesel Engine
Exhaust. Amsterdam: Elsevier, pp. 329-348.
Iwai. K.; Udagawa, T.; Yamagishi, M.; Yamada, H. (1986) Long-term inhalation studies of diesel exhaust on
F344 SPF rats. In: Ishinishi, N.; Koizumi, A.; McClellan. R. O.; Stoeber, W., eds. Carcinogenic and
Mutagenic Effects of Diesel Engine Exhaust. Amsterdam: Elsevier, pp. 349-360.
Jorgensen. H.; Svensson. A. (1970) Studies on pulmonary function and respiratory tract symptoms of workers
in an iron ore mine where dieset trucks are used underground. J. Occup. Med. 12: 348-354.
Kahn, G.; Orris, P.; Weeks, J. (1988) Acute overexposure to diesel exhaust: report of 13 cases. Am.
J. IndusL Med. 13(3): 405-406.
Kane, L. E.; Alarie. Y. (1978) Evaluation of sensory irritation from acrolein-formaldehyde mixtures. J.Am.
Ind. Hyg. Assoc. 39(4): 270-274.
Kane, L. E.; Alarie. Y. (1979) Interactions of sulfur dioxide and acrolein as sensory irritants. Toxicol. Appl.
Pharmacol. 48(2): 305-315.
Kaplan, H. L.; Mackenzie, W. F.; Springer, K. J.; Schreck, R. M.; Vostal, J. J. (1982) A subchronic study of
the effects of exposure of three species to diesel exhaust In: Lewtas, J., ed. Toxicological Effects of
Emissions from Diesel Engines. New York, NY: Elsevier, pp. 161-182.
Karagianes, M. T.; Palmer, R. F.; Busch, R. H. (1981) Effects of inhaled diesel emissions and coal dust in
rats. Am. IndusL Hyg. Assoc J. 42: 382-391.
Kilburn, K. H-; McKenzie, W, N. (1978) Leukocyte recruitment to airways by aldehyde-carbon combinations
that mimic cigarette smoke. Lab. Invest. 38: 134-142.
Last, J. A.; Gemets, J. E.; Hyde, D. M. (1983) Synergistic effects on rat lungs of mixtures of oxidant air
pollutants (ozone or nitrogen dioxide) and resptrable aerosols. Am. Rev. Respir. Dis. 128: 539-544.
May 1990
4-69
DRAFT-DO NOT QUOTE OR CITE
-------
Laurie, R. Lewkowski, J. P.; Cooper, G. P.; Hastings, L. (1978) Effects of diesel exhaust on behavior of
the rat. Presented at the Annual Meeting of the Air Pollution Control Association. Huston, TX, June
25-29, 1978. As reported in Laurie et aL, 1980.
Laurie, R. D.; Boyes, W. K.; Wessendarp, T. (1980) Behavioral alterations due to diesel exhaust exposure.
In: Pepelko, W. E; Danner, R. M; Clarke, N. A, eds. Health Effects of Diesel Engine Emissions.
Cincinnati, OH: U.S. Environmental Protection Agency, Office of Research and Development EPA
report No. EPA-600/9-80-057b. pp. 698-712.
Laurie, R. D.; Boyes, W. K. (1980) Neurophysiology alterations due to diesel exhaust exposure during the
neonatal life of the rat In: Pepelko, W. E; Danner, R. M; Clarke, N. A, eds. Health Effects of
Diesel Engine Emissions. Cincinnati. OH: U.S. Environmental Protection Agency, Office of Research
And Development. EPA report No. EPA-600^-80-057b. pp. 713-726.
Laurie, R. D.; Boyes, W. K. (1981) Neurophysiological alterations due to diesel exhaust exposure during the
neonatal life of the rat Environ. InU. 5: 363-368.
Lee, I. P.: Suzuki, K.; Lee. S. D.; Dixon. P. L. (1980) Aryl hydrocarbon hydroxylase induction in rat lung,
liver, and male reproductive organs following inhalation exposure to diesel emission. ToxicoL Appl.
Pharmacol. 52: 181-184.
Lee, S. D.; Campbell, K. L; Lauri, D.; Hinners. R. G.; Malanchuk. M.: Moore, W.; Bhatnagar, R. J.; Lee, S. D.
(1978) Toxicological assessment of diesel emissions. Presented at the 71st Annual Meeting of the
Air Pollution Control Association, Houston TX, June 25-30,1978.
Levins, P. L. (1981) Review of diesel odor and toxic vapor emissions. U. S. Department of Transportation,
Report No. DOT-HS-805831. Cambridge, MA: Arthur D. Little, Tech. Report,
DOT-TSC-NHTSA-81-9.
Lewis. T. R.; Campbell, K. L; Vaughn, T. R. (1969) Effects on canine pulmonary function via induced NO:
impairment, particulate interaction, and subsequent SO*. Arch. Environ. Health 18: 596-601. As
reported in Alarie et aL, 1972.
Lewis. T. R.; Moorman. W. J. (1980) Pulmonary and cardiovascular physiology studies during exposure. In:
Stara. J. F.; Dungworth, D. L»; Orthoefer, J. G.; Tyler. W. S.. eds. Long-Term Effects of Air Pollutants
in Canine Species. Cincinnati, OH: U.S. Environmental Protection Agency, Environmental Criteria
and Assessment Office. EPA report No. EPA-600/8-80-014.
Lewis, T. R_; Green, F. H. Y.; Moorman, W. J4 Burg, J. A R.; Lynch, D. W. (1986) A chronic inhalation
toxicity study of diesel engine emissions and coal dust, alone and combined. In: Ishinishi, N.;
Koizumi, A; McClellan, R. O^ Stoeber, W., eds. Carcinogenic and Mutagenic Effects of Diesel Engine
Exhaust Amsterdam: Elsevier, pp. 361-380.
Linnell, R. H^ Scott, W. E. (1962) Diesel exhaust competition and odor studies. J. Air Pollut Control
Assoc 12: 510-515.
Malanchuk, M. (1980) Exposure chamber atmospheres. Sampling and analysis. In: Stara, J. F.; Dungworth,
D. L.; Orthoefer, J. G.; Tyler, W. S., eds. Long-Term Effects of Air Pollutants in Canine Species.
Cincinnati, OH: U.S. Environmental Protection Agency, Environmental Criteria and Assessment
Office. EPA report No. EPA-600/8-80-014. pp. 41-54.
May 1990
4-70
DRAFT-DO NOT QUOTE OR CITE
-------
Mauderly, J. 1_; Benson, J. M.; Bice, D. E.; Henderson, R. F.; Jones, R. K.; McClelian. R. O.; Mokler, B. V.;
Pickrell. J. A; Redman, H. C; Wolff, R. K. (1981) Observations on rodents exposed for 19 weeks
to diluted diesel exhaust. Inhalation Toxicology Research Institute. Annual Report 1980-1981.
Albuquerque, NM: Lovelace Biomedical and Environmental Research Institute. Annual Report No.
LMF-91. pp. 305-311.
Mauderly, J. L; Benson, J. M.; Bice, D. E.; Carpenter, R I»; Evans, M. J.; Henderson, R. F.; Jones, R. 1C;
McClelian, R. O.; PickrelL, J. A; Redman, H. C; Shami, S. G.; Wolff, R. K. (1983) Lifespan study
of rodents inhaling diesel exhasut: results through 30 months. Inhalation Toxicology Research
Institute. Annual Report, 1982-1983. Albuquerque, NM: Lovelace Biomedical and Environmental
Research Institute. Report No. LMF-107. pp. 305-316.
Mauderly, J. L; Jones, R. F.; Henderson, R. F.; Wolff, R. K.; Pickrell, J. A; McClelian, R. O.; Gillett, N. A
(1985a) Relationship of lung structural and functional changes to accumulation of diesel exhaust
panicles. Proceeding of the Sixth International Symposium on Inhaled Particles. Cambridge,
England, September 1985. As reported in McClelian et aL, 1986.
Mauderly, J. L_; Gillett, N. A; Jones, R. F.: Henderson. R. F.; McClelian. R. O. (1985b) Structure function
correlates in diseased lungs of rats chronically exposed to diesel exhaust. Am. Rev. Resp. Dis. 42:184.
As reported in McClelian et aL. 1986.
Mauderly, J. L.; Bice, D. E.; Carpenter. R. L.; Gillett, N. A; Henderson. R. F.; Pickrell, J. A; Wolff, R. K.
(1987a) Effects of inhaled nitrogen dioxide and diesel exhaust on developing lung. Health Effects
Research Institute. Research report No. 8. 43 pp.
Mauderly, J. L_; Jones. R. K.; Griffith, W. G; Henderson, R. F.; McClelian, R. O. (1987b) Diesel exhaust is
a pulmonary carcinogen in rats chronically exposed by inhalauoa. Fundam. AppL ToxicoL 9: 208-221.
McClelian, R. O.; Bice, D. E.; Cuddihy, R. G.; Gillett, N. A; Henderson, R. F.; Jones, J.; Mauderly, J.;
Pickrell, J. A; Shami, S.; Wolff. R. K. (1986) Health effects of diesel exhaust. In: Lee, S.; Schneider,
T.; Grant, L; Verteck. P., eds. Aerosols: Research. Risk Assessment, and Control Strategies. Chelsea,
MI: Lewis Publishing pp. 2122-2140.
Meiss. R.; Robenek. H.; Schubert, M.; Themann. H.; Heinnch. U. (1981) Ultrastructural alterations in the
Uvers of golden hamsters following experimental chronic inhalation of diluted diesel exhaust emission.
Int. Arch. Occup. Environ. Health 48: 147-157.
Mentnech, M. Lewis. D. M.; Olenchock. S. A; Mull. J. C; Koller, W. A; Lewis. T. R. (1984) Effects of
coal dust and diesel exhaust on immune competence in rats. J. Toxicol. Environ. Health 13: 31-41.
Misiorwski, R. L.; Strom, K. A; Vostal, J. J.; ChviapiL, M. (1980) Lung biochemistry of rats chronically
exposed to diesel particulates. In: Pepelko, W. E; Danner, R. M: Clarke, N. A, eds. Health Effects
of of Diesel Engine Emissions. Cincinnati, OH: U.S. Environmental Protection Agency, Office of
Research and Development EPA report No. EPA-600/9-8(M)57b. pp. 465-480.
Misiorwski. R. L; Strom, K. A; Vostal, J. J.; Tillema, L; Chvapil, M. (1981) Collagen parameters in the lung
of rats chronically exposed to diesel particulate. Toxicologist 1: 75.
Moorman, W. J.; Clark, J. C; Pepelko, W. E; Mattox, J. (1985) Pulmonary function responses in cats
fallowing long-term expsoure to diesel exhaust. J. Appl. Toxicol. 5(5): 301-305.
May 1990
4-71
DRAFT-DO NOT QUOTE OR CITE
-------
Navarro, C; Charboneau, J.; McCauley, R. (1981) The effect of in vivo exposure to diesel exhaust on rat
hepatic and pulmonary microsomal activities. J. AppL Toxicol. 1: 124-126.
O'Neil J. J.; Hu, P.; Miller, F. J.; Carson, J. L.; Collier, A. M.; Garland. D. E. (1980) Functional and
morphological consequences of diesel exhaust inhalation in mice. In: Pepelko, W. E.; Danner, R. M.;
Clarke, N. A., eds. Health Effects of Diesel Engine Emissions. Cincinnati. OH: U.S. Environmental
Protection Agency, Office of Research and Development EPA report No. EPA-600/9-80-057b. pp.
766-771.
Orthoefer, J. G.; Stara, J. F.; Yang, Y. Y.; Campbell, K. I. (1980) Clinical study of beagle study. Physical
examinations, ocular examinations, hematology examination (post-exposure) and blood chemistries.
In: Stara, J. F.; Dungworth, D. L.; Orthoefer, J. G.; Tyler, W. S., eds. Long-Term Effects of Air
Pollutants in Canine Species. Cincinnati, OH: U.S. Environmental Protection Agency, Environmental
Criteria and Assessment Office, EPA report No. EPA-600/8-80-014. pp. 55-70.
Pattle, R. E.; Stretch, H.; Burgess, F.; Sinclair, K.; Edginton, J. A. (1957) The toxicity of fumes from a diesel
engine under different running conditions. Brit J. Indust Med. 14: 47-55.
Penney. D. G.; Bavlenan. M. S.; Fanning K. E.; Thill. J. E.: Yedavally, S.: Fanning, G M. (1981) A study of
hean and blood of rodents inhaling diesel engine exhaust particulates. Environ. Res. 26: 453-462.
Pepelko. W. E. (1982a) Effects of 28 davs exposure to diesel engine emissions in rats. Environ. Res. 27: 16-
23.
Pepelko, W. E. (1982b) EPA studies on the lexicological effects of inhaled diesel engine emissions. In:
Lewtas, J., ed. Toxicological Effects of Emissions from Diesel Engines. New York, NY: Elsevier, pp.
121-142.
Pepelko, W. E.; Mattox, J.; Yang, Y. Y.; Moore, W. (1980a) Pulmonary function and pathology in cats
exposed 28 days to diesel exhaust J. Environ. Path. Toxicol. 4: 449-458.
Pepelko, W. E4 Mattox, J.; Moorman, W. Clark. J. G (1980b) Pulmonary function evaluation of cats after
one year of exposure to diesel exhaust In: Pepelko. W. E; Danner. R. M: Clarke, N. A., eds. Health
Effects of Diesel Engine Emissions. Cincinnati. OH: U.S. Environmental Protection Agency, Office
of Research and Development EPA report No. EPA-600A>-80-057b. pp. 757-765.
Pepelko, W. E.; Mattox. J.; Moorman, W. J.; Clark. J. G (1981) Pulmonary function evaluation of cats after
one year of exposure to diesel exhaust Environ. IniL 5: 373-376.
Pereira, M. A.; Sabharwal, P. S.; Kaur, P.; Ross, G B.; Choi, A.; Dixon. T. (1981a) In vivo detection of
mutagenic effects of diesel exhaust by short-term mammalian assays. Environ. Int 5: 439-443.
Pereira, M. Aa Sabharwal. P. S.; Kaur, P.; Ross. G 3.; Choi, A.; Dixon. T. (1981b) The effect of diesel
exhaust on sperm-shape abnormalities in mice. Environ. Int 5: 459-460.
Plopper, GC4 Hyde, D. M.; Weir, A. J. (1983) Centriacinar alterations in lung of cats chronically exposed
to diesel exhaust J. Lab. Invest 49(4): 391-399.
I
Purdham, J. T.; Holness, D. L; Pillger, G W. (1987) Environmental and medical assessment of stevedores
employed in ferry operations. Appl. Ind. Hyg. 2(3): 133-139.
May 1990
4-72
DRAFT-DO NOT QUOTE OR CITE
-------
Quinio, J.; DeMannis, E. (1984) Sperm abnormalities in mice exposed to diesel particulate. Muuu Res. 130;
240.
Rabovsky, J.; Petersen, M. R.; Lewis, T. R.; Marion, K. 1_; Groseclose, R. D. (1984) Chronic inhalation of
diesel exhaust and coal dust: effect of age and exposure on selected enzyme activities associated with
microsomal cytochrome P-450 in rat lung and liver. J. Toxicol. Environ. Health 14: 655-666.
Rabovsky, J.; Judt, D. J.; Rodak. D. J.; Petersen. M. (1986) Influenza virus induced alterations of cytochrome
P-450 enzyme activnes following exposure of mice to coal and diesel particulates. Environ. Res. 40:
136-144.
Reger, R_; Hancock. J.; Hankinson. J.; Hearl F.; Merchant, J. (1982) Coal miners exposed to diesel exhaust
emissions. Ann. Occup. Hyg. 26: 799-815.
Reger, R.; Hankinson, J.; Merchant, J. (1978) Ventilatory function changes over a work shift for coal miners
exposed to diesel emissions. In: Bridbord. K.; French. J. eds. Toxicological and Carcinogenic Health
Hazards in the Workplace. Proceedings of the First NIOSH Scientific Symposium. Morgantown,
WV: National Institute for Occupational Safety and Health. Appalachian Laboratory for
Occupational Safety and Health.
Schneider. D. R.: Felt. B. T. (1981) Effect of diesel paniculate exposure on adenylate and guanylate cyclase
of rat and guinea pig liver and lung. J. Appi. Toxicol. 1(2): 135-139.
Schreck, R. M.; Soderholm, S. C; Chan, T. 1.; Henng, W. E.; D'Arcy, J. B. Smiler, K. L. (1980) Introduction
to and experimental conditions in the OMR chronic inhalation studies of diesel exhaust. In: Pepelko,
W. E.; Danner. R. M.: Clarke, N. A., eds. Health Effects of Diesel Engine Emissions. Cincinnati,
OH: U.S. Environmental Protection Agency, Office of Research and Development EPA report No.
EPA-600/9-80-057b. pp. 573-591.
Schreck, R. M.; Soderholm, S. C; Chan, T. L.; Smiler, K. L.; D'Arcy, J. B. (1981) Experimenul conditions
in GMR chronic inhalation studies of diesel exhaust. J. Appl. Toxicol. 1(2): 67-76.
Stara. J. F.; Dungworth, D. L.: Orthoefer, J. G.; Tyler, W. S., eds. (1980) Long-Term Effects of Air Pollutants
in Canine Species. Cincinnati, OH: U.S. Environmental Protection Agency. Environmental Criteria
and Assessment Office. EPA report No. EPA-600/8-80-014.
Stoeber, W. (1986) Experimental induction of tumors in hamsters, mice and rats after long-term inhalation
of filtered and unfiltered diesel engine exhaust. In: Ishinishi. N.; Koizumi. A.; McCIellan. R. O.;
Stoeber, W„ eds. Carcinogenic and Mutagenic Effects of Diesel Engine Exhaust. Amsterdam: Elsevier,
pp. 421-439.
Strom, K. A. (1984) Response of pulmonary cellular defenses to the inhalation of high concentrations of
diesel exhaust. J. Toxicol. Enwron. Health 13(4): 919-944.
Takafuju S.; Suzuki. S.; Koizumi. K.; Tadokoro, K.; Miyamoto, T.; Ikemori, R.; Muranaka, M. (1987) Diesel-
exhaust particulates inoculated by the intranasal route have an adjuvant activity for IgE production
in mice. J. Allerg. Clin. ImmunoL 79: 639-645.
Turk, A. (1967) Selection and training of judges for sensory evaluation of the intensity and character of
diesel exhaust odors. Washington, DC: U.S. Public Health Service, PHS Publ. No. 999-AP-32.
May 1990
4-73
DRAFT--DO NOT QUOTE OR CITE
-------
Ulfvarson. Alexandersson. R.; Annger, I_; Svensson, E.; Hedenstierna, G.; Hogstedt, C; Holmbcrg, B.;
Roses, G.; Sorsa, M. (1987) Effects of exposure to vehicle exhaust on health. Scand. J. Work
Environ. Health 13: 505-512.
U.S. Environmental Protection Agency. (1979) Air quality criteria for carbon monoxide. Research Triangle
Park, NC: Environmental Criteria and Assessment Office. EPA report No. EPA-600/8-79-022.
U.S. Environmental Protection Agency. (1982a) Air quality criteria for particulate matter and sulfur oxides.
Research Triangle Park, NC: Environmental Criteria and Assessment Office. EPA report No. EPA-
600/8-82-029aF-cF.
U.S. Environmental Protection Agency. (1982b) Air quality criteria for particulate matter and sulfur oxides:
v. l addendum. Research Triangle Park, NC: Environmental Criteria and Assessment Office, pp.
A1-A15. EPA report No. EPA-600/8-82-029aF.
U.S. Environmental Protection Agency. (1982c) Air Quality Criteria for Oxides of Nitrogen. Research
Triangle Park. NC Environmental Criteria and Assessment Office. EPA report
No. EPA 600/8-82-026F.
U.S. Environmental Protection Agency. (1983) Diesel paniculate study. Ann Arbor. MI: Office of Mobile
Sources.
U.S. Environmental Protection Agency. (1984a) Health Assessment Document for Sulfuric Acid. Cincinnati,
OH: Environmental Criteria and Assessment Office. EPA report No. EPA-540/I-86AJ31.
VS. Environmental Protection Agency. (1984b) Revised evaluation of health effects associated with carbon
monoxide exposure: An addendum to the 1979 EPA air quality criteria document for carbon
monoxide. Research Triangle Park, NC: Environmental Criteria and Assessment Office. EPA report
No. EPA-600/8-83-033F.
U.S. Environmental Protection Agency. (1985) Health and Environmental Effects Profile for Formaldehyde.
Cincinnati, OH: Environmental Criteria and Assessment Office. EPA Report No. EPA-600/X-85/362.
U.S. Environmental Protection Agency. (1986a) Health Assessment Document for Acrolein. U.5.
Environmental Protection Agency. Research Triangle Park. NC: Environmental Criteria and
Assessment Office. EPA report No. EPA-600/8-86-014A. (Draft report).
U.S. Environmental Protection Agency. (1986b) Second addendum to air quality criteria for particulate
matter and sulfur oxides (1982): assessment of newly available health effects information. Research
Triangle Park, NC Environmental Criteria and Assessment Office. EPA report No. EPA-
600/8-86-020F.
U.S. Environmental Protection Agency. (1987) Drinking water criteria document for polycyclie aromatic
hydrocarbons. Cincinnati, OH: Environmental Criteria and Assessment Office. EPA report
No. ECAO-CIN-DO1110.
U.S. Environmental Protection Agency. (1989a) Acid Aerosols Issue Paper. Washington. D.C: Office of
Health and Environmental Assessment. EPA report No. EPA-600/8/88-005A.
May 1990
4-74
DRAFT-DO NOT QUOTE OR CITE
-------
U.S. Environmental Protection Agency. (1989b) Oxides of Nitrogen Criteria Document. Research Triangle
Park, NC: Environmental Criteria and Assessment Office. EPA report No. ECAO-R-082. (Draft
report).
U.S. Environmental Protection Agency. (1989c) Air quality criteria for carbon monoxide. Research Triangle
Park. NC: Environmental Criteria and Assessment Office. (Draft report), pp. 33-41.
Vallyathan, V.; Virmani, R.; Rochlani. S.; Green. F. R; Lewis. T. (1986) Effect of diesel emissions and coal
dust inhalation on heart and pulmonary anenes of rats. J. ToxicoL Environ. Health 19(1): 33-41.
Vinegar, A; Carson. A. L; Pepelko, W. E. (1980) Pulmonary function changes in Chinese hamsters exposed
six months to diesel exhaust. In: Pepelko, W. E.; Danner, R. M; Clarke, N. A, eds. Health Effects
of Diesel Engine Emissions. Cincinnati, OH: US. Environmental Protection Agency, Office of
Research and Development EPA report No. EPA-600/9-80-057b. pp. 749-756.
Vinegar, A; Carson, A. I.; Pepelko, W. E.; Orthoefer, J. G. (1981a) Effects of six months of exposure to two
levels of diesel exhaust on pulmonary function of Chinese hamsters. Fed. Proc 40: 593.
Vinegar, A: Carson. A 1.; Pepelko. W. E. (1981b) Pulmonary function changes in Chinese hamsters exposed
six months to diesel exhaust. Environ. IntL 5: 369-371
Vostal, J. J.; Chan. T. L.; Garg, B. D.; Lee. P. S.; Stom. K. A (1981) Lymphatic transport of inhaled diesel
panicles in the lungs of rats and guinea pigs exposed to diluted diesel exhaust Environ. Intl. 5:
339-347.
Vostal, J. J.; White, H. J.; Strom, K. A; Siak, J.-S.; Chen, K.-C.; Dziedzic, D. (1982) Response of the
pulmonary defense system to diesel particulate exposure. In: Lewtas, J., ed. Toxicological Effects of
Emissions from Diesel Engines. New York, NY: Elsevier, pp. 201-221.
Wallace, M. A; Salley, S. O.; Barnhart, M. I. (1987) Analysis of the effects of inhaled diesel exhaust on the
alveolar intravascular and interstitial cellular components of rodent lungs. Scanning Microscopy 1:
1387-1395.
Waller. R. E; Commins. B. T., Lawther. P. J. (1961) Air pollution in road tunnels. Bnt J. Ind. Med.
18: 250-259.
Weller, M. A: Chen, S.-T.; Barnhart. M. I. (1981) Acid phosphatase in alveolar macrophages exposed in vivo
to diesel engine exhaust. Micron. 12: 89-90.
Werchowski. K. M.; Chaffee, V. W.; Briggs, G. B. (1980a) Teratologic effects of long-term exposures to diesel
exhaust emissions (rats). Cincinnati, OH: U.S. Environmental Protection Agency, Health Ffferts
Research Laboratory. EPA report No. EPA-600/1-80-010.
Werchowski, K. M.; Henne, S. P4 Briggs, G. B. (1980b) Teratologic effects of long-term exposures to diesel
exhaust emissions (rabbits). Cincinnati, OH: U.S. Environmental Protection Agency, Health Effects
Research Laboratory. EPA report No. EPA-600/1-80-11.
White, R J.; Garg, B. D. (1981) Early pulmonary response of the rat lung to inhalation of high
concentrations of diesel panicles. J. Appl. ToxicoL 1(2): 104-110.
May 1990
4-75
DRAFT-DO NOT QUOTE OR CITE
-------
Wiester, M. J.; litis, R.; Moore, W. (1980) Altered function and histology in guinea pigs after inhalation of
diesei exhaust. Environ. Res. 22: 285-297.
Wright, E. S. (1986) Effects of short-term exposure to diesei exhaust on lung cell proliferation and
phospholipid metabolism. Exp. Lung Res. 10: 39-55.
May 1990
4-76
DRAFT-DO NOT QUOTE OR CITE
-------
5. MUTAGENICITY
Since 1978 over 100 publications have appeared in which genotoxicity assays have
been used to analyze diesel emissions, volatile and particulate extracts, or individual
chemicals found in diesel emissions. While most of the early papers dealt with the question
of whether diesel particulates exhibited mutagenic activity in the system of choice, most of
the studies in recent years have employed bioassays (most commonly Salmonella TA98
without S9) to evaluate: (1) extraction procedures, (2) fuel modifications, (3) bioavailability
of chemicals from particulates, and (4) exhaust filters or other modifications and other
variables associated with diesel emissions. The use of bioassay-directed fractionation and
chemical analyses to identify genotoxic components of diesel exhaust has been reviewed by
Schuetzle and Lewtas (1986). Some of these studies are discussed in other sections of this
document that deal with chemical identification of diesel emissions, metabolism, and
mechanisms of carcinogenesis. This section will deal primarily with the aforementioned
qualitative genotoxicity studies. Also, because of the large number of reports, this discussion
will focus on key references. Several symposia on the health effects of diesel emissions have
been held, and the published proceedings (U.S. EPA, 1980; Lewtas, 1982; Ishinishi et alM
1986) and review articles (Gaxton, 1983; Pepelko and Peraino, 1983; LARC. 1989) on
various aspects of the mutagenicity of diesel exhaust contain detailed discussions of
individual reports.
5.1. GENE MUTATIONS
Huisingh et al. (1978) demonstrated that dichloromethane extracts from diesel
particulates were mutagenic in strains TA1537, TA1538, TA98, and TA100 of S.
typhimurium, both with and without rat liver S9 activation. This first report contained data
from several different fractions as well as particulates from different vehicles and different
fuels. Similar results with diesel extracts from various engines and fuels have been reported
May 1990 5-1 DRAFT-DO NOT QUOTE OR CITE
-------
by a number of investigators using the Salmonella frameshift-sensitive strains TA1537,
TA1538, and TA98 (Siak et al., 1981; Qaxton, 1981; Dukovich et al., 1981; Brooks et al.,
1984). Similarly, mutagenic activity was observed in Salmonella forward mutation assays
measuring 8-azaguanine resistance (Qaxton and Kohan, 1981) and in E. coli mutation assays
(Lewtas, 1983). Qaxton (1983) has provided a thorough discussion of the early studies using
the Salmonella system, especially relating to some of the factors, such as fuel, type of
combustion, environmental and atmospheric conditions that modify the generation of
mutagens in diesel emissions.
In that review Qaxton also noted that at least 184 chemicals had been identified in
diesel exhaust. Of these, 44 had been evaluated in at least one mutagenicity assay with 21
giving clear results. Indeed, a large number of the reported studies using Salmonella have
focused on the mutagenic activity of individual chemical components of diesel exhaust
particulates. Several investigators, generally using TA98 without S9 activation and expressing
mutagenic activity as activity per distance driven or mass of fuel consumed, have estimated
that the nitroarenes (both mono and dinitropyrenes) contribute a significant amount of the
total mutagenic activity of the whole extract (Nishioka et al., 1982; Salmeen et al., 1982;
Nakagawa et al., 1983).
More recently, Matsushita et al. (1986) tested particulate-free diesel exhaust gas and
a large number of benzene nitroderivatives (some of which have been identified as
components of diesel exhaust gas). The exhaust gas and 61 of the 94 chemicals were
mutagenic-Most more active in TA100 and tending to be very active without S9.
Mitchell et al. (1981) reported mutagenic activity of particle extracts of diesel
emissions in the mouse lymphoma L5178Y mutation assay. Positive results were seen both
with and without S9 activation in extracts from several different vehicles, with mutagenic
activity only slightly lower in the presence of S9. These findings have been confirmed in a
number of other mammalian cell systems using several different genetic markers. Casto et
al. (1981), Qiescheir et al. (1981), Li and Royer (1982), and Brooks et al. (1984) all reported
positive responses at the HGPRT locus in CHO cells. Morimoto et al. (1986) used the
APRT and Ouar loci in CHO; Curren et al. (1981) used Ouar in Balb/c 3T3 cells. In all of
these studies, mutagenic activity was observed without S9 activation. Liber et al. (1981) used
May 1990 5-2 DRAFT-DO NOT QUOTE OR CITE
-------
the thymidine kinase (TK) locus in the TK6 human lymphoblast cell line and observed
induced mutagenesis only in the presence of rat liver S9 when testing a methylene chloride
extract of diesel exhaust. Barfnecht et al. (1982) also used the TK6 assay to identify some
of the chemicals responsible for this activation-dependent mutagenicity. They suggested that
fluoranthene, 1-methylphenanthrene, and 9-methylphenanthrene could account for over 40
percent of the observed activity.
Morimoto et al. (1986) injected diesel particulates (2 to 4 g/kg) into pregnant Syrian
hamsters and detected increased APRT mutants in embryo cells cultivated 11 d after
injection. Both light-duty and heavy-duty tars were effective. Schuler and Niemeier (1981)
reported that an 8-h inhalation exposure of Drosophila to a fivefold dilution of exhaust from
a diesel engine did not result in an increase in sex-linked recessive lethals. Sperifio
locus mutations were not induced in (C3Hxl01)Ft male mice exposed to diesel exhaust 8
h/d, 7 d/week for either 5 or 10 weeks (Russell et al., 1980). The exhaust was a 1:18
dilution, and the average particle concentration was 6 mg/m3. After exposure, males were
mated to T-stock females, and matings continued for the reproductive life of the males. The
results were unequivocally negative; no mutants were detected in 10,635 progeny derived
from postspermatogonial cells or in 27,917 progeny derived from spermatogonia! cells.
5.2. CHROMOSOME EFFECTS
Mitchell et al. (1981) and Brooks et al. (1984) reported increases in sister chromatid
exchanges (SCE) in CHO cells exposed to paniculate extracts of emissions from both light-
duty and heavy-duty diesel engines. Morimoto et al. (1986) observed increased SCE from
both light-duty and heavy-duty tar samples in PHA-stimulated human lymphocyte cultures.
Tucker et aL (1986) exposed human peripheral lymphocytes from four donors to direct
diesel exhaust for up to 3 h. Cell cycle delay was observed in all cultures along with
increased SCE levels in two of the four cultures. Structural chromosome aberrations were
induced in CHO cells by particulate extracts from a Nissan diesel engine (Lewtas, 1983) but
not by similar extracts from an Oldsmobile diesel engine (Brooks et al., 1984).
May 1990
5-3
DRAFT-DO NOT QUOTE OR CITE
-------
Pereira et al. (1981a) exposed female Swiss mice to diesel exhaust 8h/d, 5 d/week for
1, 3, and 7 weeks. The incidence of micronuclei and structural aberrations were similar in
bone marrow cells of both control and treated mice. Increased incidence of micronuclei, but
not SCE, were observed in bone marrow cells of male Chinese hamsters after 6 mo exposure
to diesel exhaust (Pereira et al., 1981b).
Guerrero et al. (1981) observed a linear dose-related increase in SCE in lung cells
cultured after intratracheal instillation of diesel exhaust particulates at doses up to 20
mg/hamster. However, they did not observe any increase in SCE after 3 mo of inhalation
exposure to diesel exhaust particulates (6 mg/m3).
Pereira et al. (1982) measured SCE in embryonic liver cells of Syrian hamsters.
Pregnant females were exposed to 12 mg/m3 diesel particulates for 9 d or injected with diesel
particulates or extracts 18 h before sacrifice. No increased level of SCE was observed in the
inhalation experiment. The authors reported that the injection of the extract but not the
particulate resulted in a dose-related increase in SCE.
Russell et al. (1980) reported no increase in either dominant lethals or heritable
translocations in males of T-stock mice exposed by inhalation to diesel emissions. In the
dominant lethal test T-stock males were exposed for 7.5 weeks and immediately mated to
females of different genetic backgrounds [T-stock; (C3Hxl01); (C3HxC57Bl/6);
(SECxC57Bl/6)]. There were no differences from controls in any of the parameters
measured in this assay. For heritable translocation analysis, T-stock males were exposed for
4.5 weeks, mated to (SECxC57Bl/6) females, and the F, males were tested for the presence
of heritable translocations. No translocations were detected among 358 progeny tested.
53. OTHER GENOTOXIC EFFECTS
Pereira et al. (1981c) exposed males of strain A mice to diesel exhaust particulates for
31 or 39 weeks using the same exposure regimen as noted in the previous section. Analyses
of caudal sperm for sperm head abnormalities were conducted independently in three
separate laboratories. Despite clear differences in scoring criteria between the laboratories,
May 1990
5-4
DRAFT-DO NOT QUOTE OR CITE
-------
all were in agreement that the exposed incidence was not above the control level.
Conversely, male Chinese hamsters exposed for 6 mo (Pereira et aU 1981b) exhibited almost
a threefold increase in sperm head abnormalities. However, the control incidence in strain
A mice was in excess of 9 percent but less than 0.5 percent in the Chinese hamsters.
5.4. SUMMARY
Extensive studies with Salmonella have unequivocally demonstrated direct-acting
mutagenic activity in both particulate and gaseous fractions of diesel exhaust. The induction
of gene mutations has been reported in several in vitro mammalian cell lines after exposure
to extracts of diesel particulates. Dilutions of whole diesel exhaust did not induce sex-linked
recessive lethals in Drosophila or specific-locus mutations in male mouse germ cells.
Structural chromosome aberrations and SCE in mammalian cells have been induced by
particulates. Whole exhaust induced micronuclei, but not SCE or structural aberrations, in
bone marrow of male Chinese hamsters exposed to whole diesel emissions for 6 mo. In
shorter exposure (7 weeks), neither micronuclei nor structural aberrations were increased
in bone marrow of female Swiss mice. Likewise whole diesel exhaust did not induce
dominant lethals or heritable translocations in male mice exposed for 7.5 and 4.5 weeks,
respectively.
May 1990
5-5
DRAFT-DO NOT QUOTE OR CITE
-------
5.5 REFERENCES
Barfnecht, T. R.; Hites, R. A.; Cavaliers. E. L.; Thilly W. G. (1982) Human cell mutagenicity of polycyclic
aromauc hydrocarbon components of diesel emissions. In: Lewtas, J., ed. Toxicological Effects of
Emissions from Diesel Engines. New York: Elsevier, pp. 277-294.
Brooks, A. L.; Li, A. P.; Dutcber, J. S.; Clark, C R.: Rothenberg, S. J.; Kiyoura, R.; Bechtold, W.E.;
McClellan, R. O. (1984). A comparison of genotoxicity of automobile exhaust particles from
laboratory and environmental sources. Environ. Mutagen. 6: 651-668.
Casto. B. C; Hatch, G. G.; Huang, S. L; Huisingh, J. L.; Nesnow, S.; Waters, M. D. (1981) Mutagenic and
carcinogenic potency of extracts of diesel and related environmental emissions: In vitro mutagenesis
and oncogenic transformation. Environ. Int. 5: 403-409.
Chescheir. G. M. Ill; Garrett, N. E.; Shelburne, J. D.; Huisingh, J. L.; Waters, M. D. (1981) Mutagenic effects
of environmental particulates in the CHO/HGPRT system. In: Waters, M. D.; Sandhu, S. S.;
Huisingh, J. L.; Claxton, L. C; Nesnow, S.. eds. Application of short-term bioassays in the
Fractionation and Analysis of Complex Environmental Mixtures. New York: Plenum Press, pp.
337-350.
Claxton, L. D. (1981) Mutagenic and carcinogenic potency of diesel and related environmental emissions:
Salmonella bioassay. Environ. Int. 5: 389-391.
Claxton, L. D. (1983) Characterization of automotive emissions by bacterial mutagenesis bioassay: A review.
Environ. Mutagen. 5: 609-631.
Claxton, L. C; Kohan, M. (1981) Bacterial mutagenesis and the evaluation of mobile-source emissions. In:
Waters, M. D.; Sandhu, S. S.; Huisingh, J. L; Claxton, L. C; Nesnow, S., eds. Application of Short-
term Bioassays in the Fractionation and Analysis of Complex Environmental Mixtures. New York:
Plenum Press, pp. 299-317.
Curren. R. D.; Kouri, R. E; Kim. D. M.; Schectman, L. M. (1981) Mutagenic and carcinogenic potency of
extracts from diesel related environmental emissions: Simultaneous morphological transformation and
mutagenesis in BALB/c 3T3 cells. Environ. Int. 5: 411-415.
Dukovich, M.; Yasbin, R. E.; Lestz, S. S.; Risby, T. H.; Zweidinger. R. B. (1981) The mutagenic and SOS-
induang potential of the soluble organic fraction collected from diesel particulate emissions. Environ.
Mutagen. 3: 253-264.
Guerrero, R. R.; Rounds, D. E.; Orthoefer, J. (1981) Sister chromatid exchange analysis of Syrian hamster
lung cells treated m vivo with diesel exhaust particulates. Environ. Int. 5: 445-454.
Huisingh, J.; Bradow, R.; Jungers, R.; Claxton, L_; Zweidinger, R.; Tejada, S.; Bumgarner, J.; Duffield, F.;
Waters, M.; Simmon, V.F.; Hare, C; Rodngues, C; Snow, L. (1978) Applicauon of bioassay to the
characterization of diesel particle emissions. Waters, M. O.; Nesnow, S.; Huisingh, J. L; Sardhu, S.
S.; Claxton, L., eds. Application of Short-term Bioassays in the Fractionation and Analysis of
Complex Environmental Mixtures. New York: Plenum Press, pp. 381-418.
International Agency for Research on Cancer (LARC) (1989) Diesel and gasoline engine exhausts. LARC
Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans 46: 41-185.
May 1990
5-6
DRAFT-DO NOT QUOTE OR CITE
-------
Ishinishi, N.; Koizumi, A.; McClellan. R. O.; Stober, W., eds. (1986) Carcinogenic and mutagenic effects of
diesel engine exhaust. New York: Elsevier Science Publishers.
Lewtas, J. (1982) Mutagenic activity of diesel emissions. In: Lewtas J. ed. Toxicological Effects of Emissions
from Diesel Engines, New York: Elsevier; pp. 243-264.
Lewtas, J. (1983) Evaluation of the mutagenicity and carcinogenicity of motor vehicle emissions in short-term
bioassays. Environ. Health Perspect 47: 141-152.
Li, A. P.; Royer, R. E. (1982) Diesel-exhaust-panicle extract enhancement of chemical-induced mutagenesis
in cultured Chinese hamster ovary cells: possible interaction of diesel exhaust with environmental
chemicals. MutaL Res. 103: 349-355.
Liber, R L.; Andon, B. M.; Hites. R. A.; Thilly, W. G. (1981) Diesel soot: Mutation measurements in
bacterial and human cells. Environ. Int. 5: 281-284 .
Matsushita, R; Goto, S.; Endo, O.; Lee. J.; Kawai, A. (1986) Mutagenicity of diesel exhaust and related
chemicals. In: Ishinishi, N.; Koizumi, A.; McClellan. R. O.; Stober, W., eds. Caranogenic and
Mutagenic Effects of Diesel Engine Exhaust. New York: Elsevier, pp. 103-118.
Mitchell. A. D.; Evans. E. L_; Jotz. M. M.; Riccio, E. S.; Mortelmans, K. E.; Simmon. V.F. (1981) Mutagenic
and caranogenic potency of extracts of diesel and related environmental emissions: In vuro
mutagenesis and DNA damage. Environ. Int. 5: 393-401.
Monmoto, K.; Kondo, R; Kitamura, M.; Koizumi, A. (1986) Genotoxiaty of diesel exhaust emissions in in
vuro short-terra assays. In: Ishinishi, N.; Koizumi, A.; McClellan. R.O.; Stober, W., eds. Carcinogenic
and Mutagenic Effects of Diesel Engine Exhaust. New York: Elsevier, pp. 85-102.
Nakagawa, R.; Kitamori, S.; Horikawa, K.; Nakasbima, K. Tolciwa, R (1983) Identification of dinitropyrenes
in diesel-exhaust particles: Their probable presence as major mutagens. Mutai Res. 124: 201-211.
Nishioka, M. G.; Petersen, B. A.; Lewtas, J. (1982) Comparison of nitro-aromatic content and direct-acting
mutagenicity of diesel emissions. In: Cooke, M.; Dennis. AJ.; Fisher. G.L., eds. Polynuclear Ajomauc
Hydrocarbons: Physical and Biological Chemistry. Columbus, OH: Battelle Press, pp. 603-613.
Pepelko, W. E.; Peraino, W. B. (1983) Health effects of exposure to diesel emissions: A summary of animal
studies conducted by the US Environmental Protection Agency's Health Effect Research Laboratory
at Cincinnati, OR J. Am. ColL ToxicoL 2: 253-306.
Pereira, M. A.; Connor, T. R; Meyne, J.; Legator, M. S. (1981a) Metaphase analysis, micronucleus assay and
urinary mutagenicity assay of mice exposed to diesel emissions. Environ. Int. 5: 435-438.
Pereira, M. A_; Sabharwal, P. S.; Gordon, L^ Wyrobek, A. J. (1981b) The effea of diesel exhaust on sperm-
shape abnormalities in mice. Environ. Int 5: 459-460.
Pereira, M. A.; McMillan, L; Kaur, P.; Gulati, D. K.; Sabharwal, P. S. (1982) Effea of diesel exhaust
emissions, particulates, and extract on sister chromatid exchange in transplacental^ exposed fetal
hamster liver. Environ. Mutagen. 4: 215-220.
Pereira, M. A.; Sabharwal, P. S.; Kaur, P.; Ross, C B.; Choi, A.; Dixon, T. (1981c) In vivo detection of
mutagenic effects of diesel exhaust by short-term mammalian bioassays. Environ. Int 5: 439-443.
May 1990
5-7
DRAFT--DO NOT QUOTE OR CITE
-------
Russell, L. B.; Generoso, W. M.; Oakberg, E. F.; Russell W. L; Bangham. J. W.; Stelzner, K. F. (1980) Tests
for hentable effects induced by diesel exhaust in the mouse. ORNL-5685. Martin Manetta Energy
Systems, Inc. Oak Ridge NatL Lab.
Salmeen, L; Durisin, A. M.; Prater, T. J.; Riley, T.; Schuetzle, D. (1982) Contribution of 1-nitropyrene to
direct-acting Ames assay mutagenicities of diesel paniculate extracts. Mutat. Res. 104: 17-23.
Schuetzle, D.; Lewtas, J. (1986) Bioassay-directed chemical analysis in environmental research. AnaL Chem.
58: 1060A-1075A.
Schuler. R. L; Niemeier, R. W. (1981) A study of diesel emissions on Drosophila. Environ. Int. S: 431-434.
Siak, J. Chan, T. L; Lees, P. S. (1981) Diesel particulate extracts in bacterial test systems. Environ. InL
5: 243-248.
Tucker, J. D.; Xu, J.; Stewart, J.; Baciu, P. C; Ong, T. (1986) Detection of sister chromatid exchanges induced
by volatile genotoxicants. Teratogen. Carcinogen. Mutagen. 6: 15-21.
U.S. Environmental Protection Agency (U.S. EPA). (1980) Health effects of diesel engine emissions:
Proceedings of an international symposium. EPA 600/9-80-057b. Office of Health and Environmental
Assessment, Environmental Criteria and Assessment Office, Cincinnati, OR
May 1990
5-8
DRAFT-DO NOT QUOTE OR CITE
-------
6. METABOLISM AND MECHANISM OF ACTION
IN DIESEL EMISSION-INDUCED CARCINOGENESIS
The metabolism and proposed carcinogenic mechanism of action of chemicals are
intrinsically related. It is difficult to obtain an understanding of one without consideration
of the other and, therefore, both of these processes are being considered in this chapter.
Specifically, emphasis will be given to polycyclic aromatic hydrocarbons (PAH) and
aldehydes. These components are of greatest concern due to their demonstrated or
suspected activity as carcinogens and/or procarcinogens. and their universal occurrence in
diesel emissions. It is also well known that PAH biological reactivity and carcinogenicity are
dependent upon metabolic processes (reviewed by Conney, 1982). Relative to diesel exhaust
carcinogenicity, discussions in this chapter are limited to B[a]P, 1-nitropyrene, several
dinitropyrenes, and aldehydes, all of which are present in diesel emissions.
6.1. METABOLISM CONSIDERATIONS
Several long-term inhalation studies have provided evidence for carcinogenicity and
tumorigenicity of whole diesel exhaust in animals (Heinrich et al., 1986; Iwai et al., 1986:
Mauderly et al., 1987). Over 100 carcinogenic or potentially carcinogenic components have
been specifically identified in diesel emissions, including various PAHs. The PAH
compounds are adsorbed to the carbon core of the particulate phase of the exhaust, and.
as described in this chapter and in Chapter 7, this particle association is important relative
to the biological effects of these components. Among the PAH compounds identified from
diesel exhaust are benzo[a]pyrene (B[a]P), dibenz(a,h)anthracene, pyrene, chrysene, and
nitroaromatics such as 1-nitropyrene (1-NP), 1,3-dinitropyrene, 1,6-dinitropyrene, and 1,8-
dinitropyrene, all of which are mutagenic, carcinogenic, or implicated as procarcinogens or
cocarcinogens (Stenback et al., 1976; Weinstein and Troll, 1977; Thyssen et al., 1981; Pott
May 1990
6-1
DRAFT-DO NOT QUOTE OR CITE
-------
and Stober, 1983; Howard et al., 1983; Hirose et al., 1984; Nesnow et al_, 1984; El-Bayoumy
et al., 1988).
6.1.1. Metabolism and Disposition of B[a]P
It is generally recognized that B[a]P is an activation-dependent carcinogen with the
activated metabolites forming covalent DNA adducts (Boyland, 1980). The reactions
responsible for this activation are mediated by the cytochrome P-450 complex and are known
to occur in multiple tissues and in different species. The activation proceeds through Phase
I oxidative and hydrolytic reactions which result in the formation of the ultimate carcinogenic
metabolite, B[a]P 7,8-dihydrodiol 9,10-epoxide. Specifically, B[a]P undergoes a mixed
function oxidase (MFO)-mediated epoxidauon to form B[a]P-7,8-oxide which, in turn, is
subjected to an epoxide hydrolase-mediated hydrolysis resulting in the stereoisomer^ diols,
(+)-B[a]P 7,8-dihydrodiol and (-)-B[a]P 7,8-dihydrodiol. The diasterioisomeric forms of
B[a]P 7,8-diol 9,10-epoxide are derived following another P-450-mediated reaction. This
metabolic pathway is summarized in Figure 6-1.
Several studies have examined the carcinogenic potential of diesel exhaust by
evaluating the metabolism and disposition of constituents such as B[a]P. The principle
source of environmental B[a]P is its association with airborne particles such as those
generated by diesel engines and coke ovens (U.S. EPA. 1985).
Mitchell (1982), subjected 24 male F344 rats to nose-only inhalation of 3HB[a]P
aerosol (500 jxg/L of air) for 60 min. High levels of radiolabel were detected in the trachea,
lungs and turbinates. A biphasic clearance was noted with tI/2 values of 2 to 3 h and 25 to
56 h. Absorption from the lungs and systemic distribution was demonstrated by the presence
of radiolabel in soft tissues such as the liver, kidney, gastrointestinal tract, spleen, brain, and
testes. The majority of the radiolabel in these tissues was removed after 2 d and the major
route of excretion was in the feces. The significance of this study is the demonstration of
rapid absorption and systemic distribution of B[a]P
May 1990
6-2
DRAFT-DO NOT QUOTE OR CITE
-------
roioi:
010101
*- r k
MitrJ I mlka
(>uIm Sj^tra
I'fariir ItjdnUm „ 1 ^
0' "0
(i) iq>|i* . ,
ii u.ir i s .i.i.|.ii«»i»J i
Mi«4 Cuu.. I,w
Oilfat S^ilra
Dc
it 'C
- j; K
Miri I'lMi
OuJ
I'funMt llydrolMc -i0 I
0 5
O'
()l Ml' 7.1 .aide
(i) iq.|r iiJi^Jiixibj
mi
() I1»|r iij.j ». in i, |
j^lf
Oil
(•) H»|l' J»4 9|0r|..lll|r |
0.
0c
MO T
OH
( I M ^ni| 9,111 rfxuiJc 2
Figure 61. Metabolic pathway for B[a]P and formation of ultimate carcinogenic
intermediate.
Adapted from: Levin et ;il., 1980.
-------
following inhalation exposure, the ramifications of which may include altered metabolism due
to circumventing the first-pass metabolism effect of the gastrointestinal tract or liver.
Metabolism of intratracheally instilled B[a]P (1.0 ng) in strain A/J mice exposed to
diluted diesel exhaust (8 h/d, 7 d/week for 9 mo) was reported by Tyrer et al. (1981). The
radiolabel (,4C or 3H) was rapidly distributed throughout the body within 2 h. The highest
levels were detected in the lungs, liver, and gastrointestinal tract. Only trace levels were
detected in the gastrointestinal tract 168 h after administration. A companion study
(Cantrell and Tyrer, 1980) examined the effects of diesel exhaust exposure on in vivo BMP
metabolism in the aforementioned mice. Homogenates of lung, liver, and testes were
obtained from five mice sacrificed at 2, 24, or 168 h after B[a]P instillation. HPLC analysis
detected free B[a]P, nonconjugated primary metabolites, and sulfate, glucuronide and
glutathione conjugates. Based on the concurrence of pnmarv and secondary metabolites in
each tissue. B[a]P metabolism was verified in all three tissues. The major hepatic metabolite
was 3-hydroxy-B[a]P. No significant difference in metabolite profiles was noted between the
diesel exhaust-exposed and the control (clean air) mice with the exception that diesel
exhaust-exposed mice retained greater levels of parent compound in the lungs suggesting
that B[a]P was adsorbed to the exhaust particles.
Leung et al. (1985) studied the role of microsomes in the removal and metabolism
of B[a]P from diesel exhaust particles. Hepatic and lung microsomal preparations were
made from 3-methylcholanthrene-mduced F344 rats. uCarbon-B[a]P was adsorbed to diesel
exhaust panicles (0.49 jxCi/mg) and incubated with the microsomal preparations. Results
indicated that both lung and liver microsomes were capable of removing B[a]P from the
exhaust particles and that this capacity was dependent upon the lipid content of the
microsomes. The microsomes metabolized free B[a]P to B[a]P-9,10-diol and B[a]P-7,8,9,10-
tetrol but did not metabolize particle-associated B[a]P.
Bond et al. (1984), however, demonstrated metabolism of particle-associated B[a]P
and free B[a]P by alveolar macrophages (AM). AM isolated from beagle dogs were
incubated with 1 jiM ,4C-B[a]P in solution or with 1 /xM ,4C-B[a]P coated onto diesel exhaust
particles obtained from the exhaust of an Oldsmobile engine operated on the U.S. EPA
driving cycle. Incubations were conducted at 37 * C in 5 ml of Dulbecco's modified essential
May 1990 6-4 DRAFT-DO NOT QUOTE OR CUE
-------
medium (2.0x10* cells/mL) for 6, 12, 24, 36, or 48 h. The B[a]P concentrations used were
not cytotoxic nor were they rate limiting. B[a]P-9,10-diol and B[a]P-7,8-diol were identified
in the culture media and B[a]P-7,8-diol and B[a]P-4,5-diol were detected in the cellular
extracts. Additionally, small amounts of B[a]P phenols and B[a]P quiniones were detected
in both the cells and the media. The total amount of metabolites from both the cells and
media were increased with increasing incubation time up to 48 h. However, use of B[a]P
in solution or B[a]P coated onto diesel exhaust particles did not alter the total amount of
metabolites produced by the macrophages over a 24-h incubation period. AM-mediated
metabolism of particle-associated B[a]P is especially relevant considering that macrophages
are instrumental in sequestering and transporting diesel exhaust particulate matter in the
lungs. Although this investigation points to the ability of the AM to metabolize B[a]P
associated with diesel panicles, Chen and Vostal (1982) have reported that aryi hydrocarbon
hydroxylase in AM is decreased after in vivo exposure to diesel exhaust. Whether such
diesel-associated decreases in AM enzymatic activity is counterbalanced by increases in the
AM population size in response to diesel panicle deposition (White and Garg, 1981) is
unknown. Moreover, further studies are required to determine the metabolic activity of AM
as a function of cellular burdens of diesel panicles; preferably, these studies should be
conducted using human AM inasmuch as metabolic pathways present may qualitatively and
quantitatively differ among species. Although it is known that human AM contain aryl
hydrocarbon hydroxylase (AHH) activity (McLemore et al., 1981), and that they can
metabolize B[a]P (Harris et al., 1978), comparative studies of the AHH activities in rat,
hamster, and human AM could contribute toward determining the relationship such activity
may have on the development of lung tumors.
Even though the AM appear to contain the bulk of diesel panicles deposited in the
lung during chronic exposures, other cell type may also panicipate in the metabolic
activation of carcinogenic agents. Significant metabolism of B[a]P by rat type II alveolar
lung epithelial cells was reponed by Bond et al. (1983). In this study a lung epithelial cell
line (LEC) was shown to metabolize B[a]P to B[a]P-7,8-diol and B[a]P-9,10-diol. the latter
accounting for 80 percent of the total B[a]P metabolites. Small quantities of glucuronide
conjugates of B[a]P-7,8-diol and 9-hydroxy-B[a]P. Pre-exposure of the cells to diesel exhaust
May 1990 6-5 DRAFT-DO NOT QUOTE OR CITE
-------
particle extract, benz[a]anthracene or coal gas condensate increased rates of B[a]P
metabolism two- to fivefold. It was also found that pretreatment of LEC with diesel exhaust
particle extract produced a threefold increase in covalent binding of 14C-B[a]P. Compared
to the AM that were examined in the aforementioned study, the rat Type II cells showed
approximately 10 times greater activity in their ability to metabolize B[a]P.
Under healthy conditions, the Type II pneumocytes represent about 12 to 16 percent
of all cells in the alveolar region of mammalian lungs and account for approximately 4 to 9
percent of the cells in the lungs (Crapo et al., 1983). AM, on the other hand, account for
approximately 4 to 9 percent of the cells in the alveolar region (Crapo et al., 1983). In
terms of their relative availability and existing information on their relative abilities to
metabolize B[a]P, the Type II epithelial cells may play an even more important role in
metabolically activating PAH than the AM, assuming PAH as a substrate is available to
them (e.g., extraction of PAH from diesel panicles by AM and the subsequent release of
PAH or metabolically susceptible metabolites of PAH at Type II sites). The Type II cell
hyperplasia observed after the deposition of diesel particles, and other types of particles
(White and Garg, 1981; Lee et al., 1986; Lee et al., 1988; Plopper et al., 1983), seemingly
would favor a prominent role for these cells in producing activated PAH metabolites.
Another cell type that may be importantly involved in the metabolism of PAH to ultimate
carcinogens is the Clara cell. This lung cell type is relatively rich in chemical metabolizing
enzymes. The pulmonary P-450 system, for example, is present in Type II cells but it is not
as concentrated in this epithelial cell type as it is in the Type II and Clara cell (Boyd, 1984).
It is worthy to note that bronchoalveolar adenomas following diesel exposure have been
found to resemble both Type II and Clara epithelial cells (Mauderly et al., 1987). Like the
Type II pneumocyte, the Clara cells are viewed as not being importantly involved in the
phagocytosis of particles that deposit in the lung. As previously indicated, any metabolism
of procarcinogens by these cells probably involves the preextraction of carcinogen(s) in the
extracellular lining fluid (ELF) and/or in other endocytic cells.
It is evident from the preceding studies that B[a]P adsorbed to diesel exhaust particles
that are deposited in the respiratory tract is readily distributed throughout much of the
organism via absorption from the lung and transport by the mucociliary escalator to the
May 1990 6-6 DRAFT-DO NOT QUOTE OR CITE
-------
gastrointestinal tract. The present data base appears to support the contention that particle-
associated B[a]P can ultimately be metabolized by AM and or Type II cells to reactive
intermediates following deep lung deposition of these particles.
6.1.2. Metabolism and Disposition of 1-Nitropyrene
1-Nitropyrene (1-NP), a genotoxic and carcinogenic nitro-substituted PAH, is a
particle associated component of diesel exhaust (Pitts et al., 1982; Schuetzle et al., 1982;
King, 1988). As with B[a]P, several investigators have studied the metabolism and
disposition of 1-NP both in free form and in association with diesel exhaust particles.
Bond and Mauderly (1984) quantitated 1-NP metabolism and covalent binding in the
isolated-perfused rat lung. The study verified oxidation, reduction, acetylation, and
conjugation biotransfomation of 1-NP by the lung with oxidation being the major process.
The overall metabolism of 1-NP was increased by prior exposure of the rats to the MFO
inducer 3-methyl cholanthrene (3-MC) but not to phenobarbital. Based on the absence of
a concomitant increase in DNA-adducts with the increased covalent binding, it was uncertain
if this covalent binding was due to formation of mutagenic, reactive intermediates.
A more recent study (Bond et al., 1986) examined the metabolism and deposition of
free and particle-associated 1-NP in F344 rats. Results of the work indicated that the
urinary and fecal excretion of "C-l-NP was not altered by exposure to the pure form or to
that adsorbed on diesel exhaust particles. Pure 1-NP, possibly because of its smaller
aerodynamic diameter, was more efficiently absorbed in the lung than was 1-NP coated onto
diesel exhaust particles and, therefore, greater lung retention was noted for panicle-adsorbed
1-NP. However, no significant difference between the two forms of 1-NP was noted for
nonpulmonary tissue distribution or metabolic profiles. Analysis of excreta and tissues
indicated that 1-NP is rapidly metabolized by the lungs or metabolized by other tissues
following translocation from the lungs. For both 1-NP forms, small amounts of 6- and 8-
hydroxyacetylaminopyrene were detected in the lungs suggesting pulmonary oxidation,
reduction, and conjugation of the parent compound. The demonstration of pulmonary
metabolism of 1-NP and greater retention of 1-NP when adsorbed to diesel exhaust particles
May 1990
6-7
DRAFT-DO NOT QUOTE OR CITE
-------
may be of significance regarding the dose to the lungs of both parent compound and
metabolites.
Whole body exposure of rats to diesel exhaust (7.4 mg/m3) 7 h/d, 5 d/week for 4
weeks resulted in twofold increases in the rates of nitropyrene metabolism in nasal tissue and
in isolated perfused lungs from these animals (Bond et al., 1986). HPLC analysis of ethyl
acetate-extractable 1-[,4C]NP metabolites indicated that the major metabolites were 3-, 6-,
and 8-hydroxy-l-aminopyrene and 4,5-dihydro-4,5-dihydroxy-l-nitropyrene. Furthermore, a
fourfold increase in 14C covalently bound in the lungs of these rats was detected. The
increase in 1-NP metabolism was not observed for rats of lower (0.35 or 33 mg/m3) exposure
groups or clean air controls. The data from this study indicate that exposure to diesel
exhaust paniculate matter at concentrations of 7.4 mg/m3 significantly alters the metabolism
and subsequent covalent binding of nitropyrene.
Ball and King (1985) administered [UC]1-NP to rats intraperitoneal!^ orally or by
intratracheal instillation of vapor-phase-coated diesel exhaust particles (380 ng [wC]l-NP/g;
5 mg/rat). Over 59 percent of the radiolabel was recovered within 24 h regardless of the
route of administration. The metabolic profile and elmination kinetics were similar for all
routes of administration. Of the recovered radiolabel, 20 to 30 percent was in the urine and
40 to 60 percent was in the feces. The principle urinary metabolite (representing 15 to 25
percent of the total urinary 14C) was 6-hydroxy-jV-acetyl- 1-aminopyrene (6-OH-NAAP), a
compound with demonstrated S-9 dependent mutagenic activity in Salmonella strain TA98.
Gut flora was shown to be necessary for the formation of 6-OH-NAAP and for the observed
enterohepatic circulation of first pass hepatic metabolites. Accumulation of 14C and diesel
exhaust particles was detected in the lungs and gastrointestinal tract 24 h after intratracheal
adminsitration, thereby attesting to the importance of mucociliary transport and distribution
of particles and their adsorbed components. Based on these results and previous in vitro
studies (King et al., 1983), demonstrating 1-NP binding to lung macromolecules, the authors
note the possible risk to these organs relative to 1-NP. A proposed pathway for in vivo
biotransformation of 1-NP is presented in Figure 6-2.
Howard et al. (1986) studied the binding of intratracheally instilled nitropyrenes and
B[a]P to mouse lung DNA following preexposure to intratracheally instilled doses of the
May 1990 6-8 DRAFT--DO NOT QUOTE OR CITE
-------
putative inducing agents, B[a]P, diesel exhaust dichloromethane extract, or 1-NP. The results
indicated 1-NP to be a potent DNA binding agent even in the absence of enzyme induction
and that this potency was increased following B[a]P exposure. Dinitropyrene (a mixture of
the 1,3-, 1,6-, and 1,8- isomers) was also a potent lung DNA binding agent with and without
the inducers. B[a]P was not as potent a binding agent. Preexposure to the diesel exhaust
extract but not to B[a]P resulted in increased DNA binding of B[a]P. Pretreatment with the
dichloromethane extract of diesel exhaust failed to increase the DNA binding of the
nitropyrenes. The significance of this report is the demonstration that exposure to diesel
exhaust may potentiate the DNA binding of some of its components.
King (1988) provided information relating the metabolism of nitropyrenes and their
carcinogenic potential. Briefly, dinitropyrenes were found to be much more carcinogenic
than 1-NP. The cytosolic enzymes of the rat mammary gland activated dintiropyrenes by
monoreduction to hydroxylamine followed by O-acetylation. Reduction and acetylation
pathways for dinitropyrenes and subsequent DNA adduct formation were detected in intact
cells. However, the intact mammary cells metabolized 1-NP primarily through oxidative
pathways.
6.2. CARCINOGENIC MECHANISM OF PAH
COMPONENTS OF DIESEL EXHAUST
The mechanism of action of known PAH carcinogens has been attributed to the
reactivity of certain metabolic intermediates with cellular macromolecules and the
subsequent formation of DNA adducts. As described in Section 8.5, B[a]P and 1-NP
adsorbed to diesel exhaust particles can become available for biotransformation to known
reactive intermediates, and macromolecular binding of these metabolites has been
demonstrated.
The available data base does not allow for a definitive discussion of the carcinogenic
mechanism action for these compounds as it specifically relates to diesel exhaust, but rather
must be addressed from the standpoint of the chemicals per se. These data are derived
primarily from in vitro studies that were not specifically concerned with the potential
May 1990 6-9 DRAFT-DO NOT QUOTE OR CITE
-------
Oxidised l»y
liver
NO,
IIO
I Excreted in llilc .1$
O-Cliiciirouiilc
llydrnlysed jimI
rilui cil liy Cut Flora
Nil,
IIO
Acctylaled liy
liver or Kidney
Reabsorbed
(Ciilerolicpalic
Cir< ulalinn)
NO,
Itnliued liy
. Iiver
l-Nilropyrene
(J
B
IIO
ll>drolysed by
(iul flora
IIO
Fxcreled in Bile as
N-Cliicuronidc
Reabsorbed (H'O
Oxidised by
Nil, liver
Excrcled in ('line as
O-CIih mnnide
Figure 6-2. Possible metabolic pathways for in vivo biotransformation of 1-NP lo 6-hydroxy-
/V-acetyl-1 -aminopyrene.
Source: Uall and King, 1985
-------
carcinogenicity of diesel exhaust but are relevant because the compounds of concern are
known components of diesel emissions.
6.2.1. Carcinogenic Mechanism of B[a]P
As a result of PAH metabolism studies such as those conducted on B[a]P, theories
have been proposed regarding the molecular mechanism by which activated intermediates
express their genotoxic effects. B[a]P served as the model for the "bay-region" concept
summarized by Jerina et al. (1980). This proposed mechanism would also be applicable to
such compounds as benz[a,h]anthracene which is a potent carcinogen also known to occur
as a diesel exhaust combustion product.
Briefly, this concept states that compounds derived from an angular
benz(a)anthracene nucleus may undergo epoxidation and if the resulting epoxide is located
in the "bay region" will be better alkylating agents and. therefore, have greater genotoxic
potential. The chemical reactivity of these "bay region" epoxides is positively correlated with
biological reactivity of these compounds.
Based on the assumption that DNA adduct formation is a critical step in the initiation
of carcinogenesis (Harris, 1985), increases in adducts provide a plausible mechanism by
which association with particles might increase the carcinogenicity of organic compounds in
long-term inhalation studies. Various studies have examined this process in relation to
panicle-associated B[a]P.
An experiment was undertaken to test the hypothesis that inhalation of B[a]P
associated with carbon black (CB) particles would increase the levels of DNA adducts
compared to inhalation of pure Bfa]P (Wolff et al.. 1989). DNA modification was measured
using the 32P-postlabeling method recently developed by Randerath et al. (1985). The high
sensitivity of this technique w l adduct in 10l° bases (Reddy and Randerath, 1986) allowed
measurement of the low levels of DNA adducts resulting from repeated inhalation exposures
to l4C-B[a]P aerosols (2 mg/m3), ,4C-B[a]P (2 mg/m3) adsorbed to carbon black particles (97
mg/m3) (B[a]P/CB), or filtered air. Total UC levels in the lung, an indicator of reactive B[a]P
May 1990
6-11
DRAFT-DO NOT QUOTE OR CITE
-------
metabolites, were 100-fold greater following exposure to B[a]P/CB than following exposure
to B[a]P alone.
The levels of total DNA adducts in the lung were not significantly different whether
the rats were exposed to pure B[a]P or B[a]P/CB. Exposure to B[a]P appeared to be the
dominant factor in producing DNA damage with little influence of adsorption of B[a]P onto
CB particles in terms of total DNA adducts and the B[a]P diol-epoxide(BPDE)-DNA adduct.
However, association of B[a]P with CB resulted in the formation of lung adducts that were
not seen in DNA from lungs of rats exposed to pure B[a]P. These adducts were not
identified. It is possible that these adducts seen only in the B[a]P/CB exposures may play
an important role in the potential tumorigenic effect of panicle-associated B[a]P. Possible
reasons for the discrepancy between panicle effects on total DNA adducts and retention of
l4C include the possibility that the kinetics for formation and decline of DNA adducts are
different than those of total bound MC. As a consequence, long-term retention of total
B[a]P and metabolites in the lung may not be a good marker for genotoxic damage.
There were clear differences in the kinetics of the buildup and decline of DNA
adduct levels and total ,4C for rats exposed to B[a]P/CB. The half-time for the decline of
total l4C was approximately tenfold greater than that for the decline in levels of DNA
adducts for rats exposed to B[a]P/CB. Previous work has shown that at 1 d or later after
the end of single exposures to B[a]P or B[a]P/CB, most of the UC present was bound to total
macromolecules (Sun et al., 1988), presumably largely, non-DNA protein. This information
in combination with the present data thus suggest that decline or repair of DNA adducts is
considerably faster than that of protein turnover. This effect would lead to increased
buildup of l4C in the lung relative to DNA adducts following repeated exposures, as was
observed. The half-times for decline in DNA adducts observed in the present work are
similar to the half-times of approximately 4 weeks reported for B[a]P metabolite-DNA
adducts in the lungs of A/HeJ and C57B1/6J mice (Stowers and Anderson, 1985). Protein
turnover is generally on longer time scales.
It appears that long-term retention of l4C radiolabel in the lung may not be as
imponant as previously suspected, at least with respect to indicating DNA damage. 14C
binding levels and DNA adducts were not closely related and it is clear from these results
May 1990 6-12 DRAFT-DO NOT QUOTE OR CITE
-------
that DNA adduct levels cannot be predicted form total ,4C levels. This observation is
consistent with the work of Morse and Carlson (1985) who observed that binding levels of
3H with lung protein were greater than levels of 3H to lung DNA 6 h after administration
of oral H-B[a]P to mice. They also found that 3H binding to protein was more persistent
than 3H bound to DNA.
Caution should be used in interpreting the results from the present experiment in
terms of possible implications for long-term exposures when carcinogenicity might be
observed. The present experiment was of short duration (12 weeks); the same pattern of
results might not continue after many months of exposure.. The adduct levels were higher
in the rats exposed to B[a]P/CB than B[a]P after 12 weeks of exposure and so it is possible
that this difference might become greater with continued exposure. In addition, the different
adduct patterns between the B[a]P/CB and B[a]P exposures may indicate that other adducts,
besides the BPDE-DNA adduct, are important in potential carcinogenic effects of B[a]P/CB
exposures. Another factor to consider is the possible influence of a chronic inflammatory
response, cell injury and cell proliferation which accompanies long-term exposures to inhaled
insoluble particles (Morrow, 1986). Such responses are generally greater after prolonged
exposure than those in the present 12-week exposure. These responses might be factors in
progression to tumors in long-term inhalation exposures of rodents when large lung burdens
of particles accumulate (Morrow, 1986), and in the increased incidence in tumors when
B[a]P is merely mixed with Fe^ particles as opposed to being adsorbed onto the particle
(Saffioti et al., 1965).
The potential interaction of diesel exhaust components may also be of concern
regarding carcinogenicity. Laskin et al. (1976) (as cited in Lippmann. 1980) reported that
S02 can act as a cofactor in the development of bronchial carcinomas. Compared to B[a]P
exposure alone, combined S02 and B[a]P exposures resulted in a significant increase in the
incidence of squamous cell carcinomas in rats.
May 1990
6-13
DRAFT-DO NOT QUOTE OR CITE
-------
6.2-2. Carcinogenic Mechanism of Nitropyrenes
Subcutaneous injection (4.0 rag total dose) of 1,3-NP, 1,6-NP, or 1,8-NP to male F344
rats resulted in a 100 percent (10 of 10) sarcoma incidence at the injection site. After 320 d
observation, the dimethyl sulfoxide controls exhibited a tumor incidence of 0 percent
(0 of 20). No tumors were found in rats receiving 4- or 40 mg 1-NP. Transforming activity
was observed for the DNA from 4 of 11 dinitropyrene-induced tumors suggesting oncogene
activation (Sato et al., 1986).
The absence of tumor formation by 1-NP and the positive results for the dinitro-
pyrenes from the previous study is consistent with the findings of King (1988). In this study,
female CD and F344 rats were adminstered 1.3-, 1.6% or 1,8-NP. or 1-NP subcutaneously,
intraperitoneallv or intragastrically. Based on incidences of malignant histiocytomas,
mammary gland tumors and leukemias, the dinitropyrenes were much more carcinogenic
than 1-NP. Specifically, the potency order was: 1,6 > 1,8 > 1,3 dinitropyrene > 1-NP. Oral
intubation was relatively ineffective in tumor induction by these compounds. Cytosolic
enzymes of the rat mammary gland were capable of activating the dinitropyrenes to reactive
intermediates that formed tRNA adducts. The adduct formation was of the same relative
order as the carcinogenic potential and was catalyzed by rat mammary cytosol. In addition
to demonstrating the potential of dinitropyrenes to induce three tumor types in rats, this
study provided data affirming the importance of metabolic activation of the dinitropyrenes,
the quantitative variability among these isomers relative to their carcinogenic activity, and
the susceptibility of a tissue distant from the site of administration to the carcinogenic
potential of these compounds following biotransformation. The data support the conclusion
that the dinitropyrenes (especially 1,6- and 1,8-NP) are reduced by mammalian enzymes to
reactive hydroxylamines which are in turn activated by acetyl CoA to potential carcinogens.
Maher et al. (1988) examined the metabolism of 1-NP and dinitropyrenes in cultured
human fibroblasts. The results of their experiments using these diploid cells indicated that
1-NP underwent bioactivation to a form that produced stable covalent DNA adducts. The
mutagenic effect of 1-NP was correlated with the cytotoxic effect which in turn correlated
May 1990
6-14
DRAFT-DO NOT QUOTE OR CITE
-------
directly with the number of DNA adducts. It was also reponed that 1-nitrosoaminopyrene
(1-NOP), a major metabolite of 1-NP was similar in its mutagenic response and extent of
DNA adduct formation to 1-NP. The number of 1-NP or 1-NOP adducts per 106 DNA
nucleotides to reduce survival of the normal (cells that were not repair deficient) cells to 37
percent was 25. However, comparisons with repair-deficient cells did indicate that
nucleotide excision repair protected against the mutagenic and cytotoxic effects of both 1-NP
and 1-NOP. The mutagenic potency of both 1-NP and 1-NOP was intermediate between
that of N-acetoxy-2-acetylaminofluorene and benzo[a]pyrene-7,8-diol-9,10-epoxide. Based
on Salmonella typhimurium assays, Heflich et al. (1985) reported that the mutagenicity of 1-
NOP was 20-fold greater than that of 1-NP. However, if the number of DNA adducts
formed by 1-NP or 1-NOP were considered, the frequency of mutation was similar for both
compounds. A proposed pathway for cellular activation of 1-NP and DNA adduct formation
is presented in Figure 6-3.
El-Bayoumy et al. (1988) investigated the comparative tumorigenicity of 1-NP and its
reduced derivatives, 1-nitrosopyrene and 1-aminopyrene. Results of their tests using
Sprague-Dawley rats indicated that 1-NP was considerably more carcinogenic than its
reduced derivatives relative to production of mammary adenocarcinomas following
administration of these compounds by gavage. These results appear to be in conflict with
those of Wislocki et al. (1986) where 1-nitrosopyrene produced a greater incidence of
hepatic tumors in newborn mice. However, the possibility of complete reduction of 1-NP
to 1-aminopyrene by gastrointestinal and/or hepatic metabolism pnor to its reaching the
mammmary glands could explain this discrepancy.
Djuric et al. (1988) also provided data affirming the correlation between DNA
binding of 1-NP or 1,6-dinitropyrene and the relative tumorigenicity of the two compounds.
Following i.p. injections of the radiolabeled test compounds, covalent DNA binding was not
detected for 1-NP treated animals but the DNA adduct, JV-(deoxyguanosin-8-yl)-l-amino-6-
nitropyrene was detected in the livers, kidneys, urinary bladder, and mammary glands of rats
given 1,6-dinitropyrene. Induction of nitroreductases failed to increase the DNA binding by
1,6-dinitropyrene suggesting that additional factors such as O-acetylation may affect the
observed DNA binding. This and the other described studies emphasize the importance of
May 1990 6-15 DRAFT-DO NOT QUOTE OR CITE
-------
ACTIVATION CF 1-NITROPYRENE
NOe
1-NITROPYRENE
2 c"
JSC
iyj©i
2 c"
1-NITH050I'YHENE
0
II
h NH
/N'\ J J.
r m m
N ~ri NHj
I
DNA ADUUCT (C-U GUANINE )
N-HYDROXY-
I - AMINOPYRENE
NC-0
Figure 6-3. Proposed pathway for cellular activation of 1-nnropyrene H-NP).
Source: Maher et alM 1988.
May 1990
6-16
DRAFT-DO NOT QUOTE OR CITE
-------
metabolism in determining the formation and subsequent carcinogenic activity of nitro PAH-
nucleotide adducts.
63. CARCINOGENICITY OF ALDEHYDES
Regarding the carcinogenic potential of diesel emissions, most emphasis has been
placed on the PAH components. However, various aldehydes, including acrolein and
formaldehyde, are also components of diesel emissions. In addition to being primary
irritants, both formaldehyde and acrolein are known carcinogens. Although the aldehydes
are primarily gas-phase components of diesel emissions, they will be given consideration in
this chapter.
6.3.1. Metabolism and Carcinogenicity of Formaldehyde
The pharmacokinetics of formaldehyde are reviewed in U.S. EPA (1985).
Formaldehyde is readily absorbed in the upper and lower respiratory tract, and via most
other routes of exposure. The metabolism of formaldehyde is complex due to its natural
occurrence in the body, but is basically subjected to oxidation by aldehyde oxidase and
conjugation with glutathione. The major end-products of formaldehyde metabolism are
formic acid and CO: the production of these products having been verified in a wide variety
of tissues and species.
Inhalation exposure to formaldehyde at 8 and 15 ppm produced tumors of the
nasopharynx in rats (Clary et al., 1983). Similarly, clear evidence of carcinoma development
ui the nasal cavity and upper trachea was seen following inhalation exposure of rats to 14.5
ppm formaldehyde for 104 weeks (Kerns et al., 1983). A study by Albert et al. (1982) also
indicated a significant increase in squamous cell carcinomas in rats discontinuously exposed
to 14.1 to 14.3 ppm formaldehyde for 588 d. Other inhalation studies (Horton, et al., 1963;
Dalbey, 1982) were also negative with respect to carcinogenic effects.
May 1990
6-17
DRAFT-DO NOT QUOTE OR CITE
-------
Epidemiologic studies regarding increased cancer incidence from formaldehyde
exposure have not provided definitive findings. An analysis of carcinogenic risk to humans
from exposure to formaldehyde is provided by Squire and Cameron (1984).
The carcinogenic mechanism of formaldehyde is unclear. Interactions with nucleic
acids and other proteins has been demonstrated by several studies. The formation of
reversible adducts and irreversible crosslinks with nucleic acids and proteins were detected
by several investigators (French and Edsall, 1945; Feldman, 1973). The formation of DNA
adducts and DNA-protein crosslinks in the nasal mucosa of rats was reported by Heck and
Casanova-Schmitz (1983). Casanova-Schmitz et al. (1984) also found covalent binding to
DNA in the olfactory mucosa of rats exposed to formaldehyde. More recently, Grafstrom
et al. (1987) demonstrated formaldehyde-induced DNA single strand breaks and DNA
protein crosslinks in cultured human bronchial cells.
63.2. Metabolism and Carcinogenicity of Acrolein
Definitive studies regarding the in vivo metabolism of acrolein are lacking. In vitro
studies indicate that acrolein is metabolized by oxidative mechanisms to acrylic acid or
glycidaldehyde (Patel et al., 1980), which may be followed by glutathione conjugation.
Adduct formation with nucleic acids and proteins has been demonstrated (Chung et al.,
1984V Acrolein-induced DNA protein crosslinking as a result of covalent binding has also
been reported (Lam et al., 1985), and Munsch et al. (1974) provided in vivo data for
covalent binding of acrolein to RNA, DNA, and protein in rat liver. Generally, sufficient
evidence exists to verify the reactivity of acrolein with biological macromolecules.
Although there are no studies that provide adequate evidence for the carcinogenicity
of acrolein following inhalation or oral exposure, evidence that glycidaldehyde, an acrolein
metabolite, was carcinogenic in female ICR/Ha mice following subcutaneous injection was
provided by Van Duuren et al. (1966).
May 1990
6-18
DRAFT-DO NOT QUOTE OR CITE
-------
6.4. POTENTIAL INVOLVEMENT OF PULMONARY LEUKOCYTES
IN THE DEVELOPMENT OF LUNG TUMORS.
Phagocytic leukocytes have been shown by numerous investigators to be toxic to
tumor cells in vivo, and increasing evidence suggests that cells of the mononuclear phagocyte
series in particular may be of pivotal importance in providing protection against malignancy
in situ. This protective function may be due, at least in part, to their ability to produce a
tumor necrosis factor (TNF) (Urban et al., 1986). Whether the tumor surveillance and
tumoricidal activities of AM (Hengst et al., 1978; Sone et al., 1983; Sone, 1986; Kan-Mitchell
et al., 1985) are compromised or otherwise modified when they are engorged with even
relatively benign panicles has not been experimentally evaluated. Some existing information
indicates that these AM activities can be modified by environmental factors (Sone et al.,
1983; Flick et al., 1985). Additionally, the possibility remains that diesel and other types of
panicles at high lung burdens result in decreases in NK cell functional activities in providing
defense against tumor formation either by direct particle-cell interactions or by altering the
ability of AM to influence NK cell-mediated host defense against metastatic tumor cells
(Sone, 1986). Moreover, the tumoricidal function of cytotoxic T lymphocytes (Sone, 1986)
may be directly or indirectly compromised by the presence of high lung burdens of particles
in the lungs. These possibilities, accordingly, require further investigation.
In addition to the release of TNF. another mechanism by which phagocytes may
provide protection against tumor cells is via the production of toxic reactive species of
oxygen (Qark and Klebanoff, 1975). Yet, this same process may mechanistically be a two-
edged sword. Under some conditions the production of oxygen species may afford
protection against emerging tumor cells by killing the cells, while under other conditions the
production of reactive oxygen products conceivably may actually contribute to the
development of neoplastic cells. Potential macrophage and polymorphonuclear neutrophil
(PMN) involvement in the development of lung tumors in laboratory rats administered high
lung burdens of diesel panicles (Mauderly et al., 1987), or the inhalation of particles that
are generally considered to have low to no cytotoxic potential [e.g. titanium dioxide (Lee et
al., 1986)], over a prolonged period of time may be related to the ability of the lung free
May 1990 6-19 DRAFT-DO NOT QUOTE OR CITE
-------
cells to produce reactive oxygen metabolites during phagocytic oxidative metabolism (Hatch
et al., 1980). Whereas products of phagocytic oxidative metabolism, including superoxide
anion, hydrogen peroxide, and hydroxy! radical can kill tumor cells (Klebanoff and Clark,
1978), and the reactive oxygen species can peroxidize lipids to produce cytotoxic metabolites
such as malonyldialdehyde, some products of oxidative metabolism apparently can also
interact with DNA and produce mutations. Along this line, Weitzman and Stossel (1981)
found that human peripheral leukocytes were mutagenic in the Ames assay. This mutagenic
activity was related to polymorphonuclear leukocytes and blood monocytes; blood
lymphocytes alone were not mutagenic. These investigators speculated that the mutagenic
activity of the phagocytes was due to their ability to produce reactive oxygen metabolites
inasmuch as blood leukocytes from a patient with chronic granulomatous diseases, a disease
:n which neutrophils have a defect in the NADPH oxidase generating system (Klebanoff and
Clark, 1978), were less effective in producing mutations than were normal leukocytes. Of
related significance in terms of oxygen species being able to cause genetic disturbances,
Phillips et al. (1984) demonstrated that the incubation of Chinese hamster ovary cells with
xanthine plus xanthine oxidase (a system for enzymatically generating active oxygen species)
resulted in genetic damage hallmarked by extensive chromosomal breakage and sister-
chromatid exchange, and produced an increase in the frequency of thioguanidine-resistant
cells (HGPRT test). Aside from interactions of oxygen species with DNA, increasing
evidence also points to an important role of phagocyte-derived oxidants and/or oxidant
products in the metabolic activation of procarcinogens to their ultimate carcinogenic form
(Kensler et al., 1987).
Hatch and co-workers (1980) have demonstrated that interactions of guinea pig AM
with a wide variety of particles including silica, metal oxide-coated fly ash, polymethyl-
methacrylate beads, chrysotile asbestos, fugative dusts, polybead carboxylate microspheres,
glass and latex beads, uncoated flyash, and fiberglass increase the production of reactive
oxygen species. Similar findings have been reported by numerous investigators for human,
rabbit, mouse, and guinea pig AM (Drath and Karnovsky, 1975; Allen and Loose. 1976;
Beall et 2d., 1977; Lowne and Aber, 1977; Miles et al., 1977; Rister and Baehner, 1977;
Hoidal et al., 1978; Drath, 1985). As well, polymorphonuclear leukocytes (PMN) are also
May 1990 6-20 DRAFT-DO NOT QUOTE OR CITE
-------
known to increase production of superoxide radical (02 ), hydrogen peroxide (H20;), and
hydroxy! radical (OH-) in response to membrane-reactive agents and particles (Goldstein
et al„ 1975; Weiss et aU 1978; Root and Metcalf, 1977). Thus, it appears that phagocytes
from a variety of species produce elevated levels of oxidant reactants in response to
challenges such as phagocytic stimuli, with the physicochemical characteristics of a
phagocytized particle being a major factor in determining the magnitude of the oxidant-
producing response.
It is well recognized that the deposition of particles in the lung can result in the efflux
of PMN from the vascular compartment into the alveolar space compartment in addition to
expanding the AM population size. Following acute exposures, the outpouring of the PMN
is transient and generally over in a few days (Adamson and Bowden, 1978, 1981; Bowden
and Adamson, 1978; Lehnert et al., 1988). The recruitment of this cell type in response to
particle deposition has been related to chemotactic factors derived from the activation of
complement (Warheit et al., 1985), the necessary components of which presumably are
sufficiently available in the lung's ELF, and chemotactic factors that are elaborated by AM
by phagocytic stimuli (Adamson and Bowden, 1982). StTom (1984) has reported that PMN
become abnormally abundant following chronic exposures to particulate diesel exhaust. In
the study by Strom (1984), the numbers of PMN lavaged from the lungs of diesel-exposed
rats generally increased with increasing exposure duration and inhaled exposure mass
concentration. Strom (1984) also found that PMN in diesel-exposed lungs remained
persistently elevated for at least 4 mo after cessation of exposure. The duration of time over
which AM continue to produce chemotactic factors after they have phagocytized particles
has not been well characterized. However, a potential mechanism for the appearance of
PMN at times well after particles have been deposited in the lungs may be related to an
ongoing release of previously phagocytized particles by alveolar phagocytes that engulfed
them shortly after deposition. Evidence in support of this possibility has been obtained by
Lehnert et al. (1989). In the previously described study in which rats were intratracheally
instilled with 85 /ig, 1.06 mg, or 3.6 mg of polystyrene particles, PMN in lavaged lung free
ceil populations during the alveolar clearance of the three lung burdens of particles were
quantitated. PMN were not found to be abnormally abundant during the clearance of the
May 1990 6-21 DRAFT-DO NOT QUOTE OR CITE
-------
two lower lung burdens, but they did become progressively elevated in the lungs of the
animals in which alveolar phase clearance was restored. Moreover, the particle burdens in
the PMN became progressively greater over time. Such findings are consistent with an
ongoing particle relapse process, given the relatively short lifespan of PMN. As previously
indicated, lung tumors develop in the rat at lesser lung burdens of diesel exhaust particles
than with a particle like TiO,. PMN characteristically are abnormally increased in the lung
by diesel exhaust exposure, but their presence in the lungs does not appear to be excessive
following the pulmonary deposition of even high lung burdens of Ti02 (Strom, 1984; Lee et
al., 1986). Thus, the generation of reactive oxygen species by both AM and PMN should be
considered as one potential factor of what probably is a multistep process that culminates
in the development of lung tumors in response to chronic paniculate diesel exhaust
deposition.
Another characteristic of the AM and PMN that may contribute to the pathologic
process leading to lung tumor development following paniculate diesel exhaust deposition
is that these phagocytes are known to release a variety of potentially destructive hydroiytic
enzymes. This release process is known to occur simultaneously with the phagocytosis of
particles (Sandusky et al., 1977). The essentially continual release of such enzymes during
chronic particle deposition and phagocytosis in the lung may be detrimental to the alveolar
epithelium, especially Type I pneumocytes. Evans et al. (1986) showed that injury to Type
I cells is followed shortly thereafter by a proliferation of Type II pneumocytes. Type II cell
hyperplasia is a generally common feature observed in the lungs of animals that have
received high lung burdens of various types of particles, including simple polystyrene
micropsheres (see Figure 6-4). Exaggerated proliferation as a repair or defensive response
to diesel deposition may have the effect of amplifying the likelihood of neoplastic
transformation in the presence of carcinogens beyond that which would normally occur with
lower rates of proliferation, assuming an increase in the cell cycling of target cells and the
probability of a neoplastic-associated genomic disturbance.
The proliferative response of Type II cells following the deposition of paniculate
diesel or other types of particles, however, has yet to be directly related to a Type I cell
destruction by proteolytic enzymes released by lung phagocytes or to a direct action of
May 1990 6-22 DRAFT-DO NOT QUOTE OR CITE
-------
particles on the proliferation kinetics of the Type II cells. The production of reactive oxygen
species or products therefrom could also be involved in the process. Whatever the stimulus,
it remains possible that the lung's AM population may play a role aside from any
responsibility for Type I pneumocyte damage. AM have the ability to release several other
effector molecules or cytokines that can regulate numerous functions of other lung cells,
including their rates of proliferation (Bitterman et al., 1983; Jordana et al_, 1988; Denholm
and Phan, 1989). Am-derived mediators that may stimulate Type II cell hyperplasia
following particle deposition in the lung, however, remain to be identified, if in fact the AM
play a regulatory role in the Type II ceil proliferative response.
6.5. SUMMARY OF METABOLISM AND MECHANISM OF ACTION
OF CARCINOGENIC COMPONENTS OF DIESEL EXHAUST
Several studies affirm the bioavailability from inhaled diesel exhaust particles of
compounds such as B[a]P and 1-NP which are known to be carcinogenic or mutagenic.
Furthermore, that xenobiotics can undergo biotransformation to reactive intermediates
following their entry into the body via inhalation of diesel exhaust particles has been
demonstrated for B[a]P, 1-NP, and some of the dinitropyrenes. It is generally accepted that
the underlying mechanism of carcinogenesis involves the formation of covalent adducts with
DNA resulting in the alteration of cellular genetic information. Several reports have
provided data indicating the formation of such adducts following administration of these
compounds. A complete understanding of the carcinogenic potential of these compounds
following diesel exhaust exposure appears to be dependent upon a thorough knowledge of
metabolism and kinetic parameters relative to adduct formation and the involvement of
genetic repair processes.
The development of lung tumors in experimental laboratory animals following chronic
exposures to particulate diesel exhaust occurs under conditions in which alveolar
macrophage-mediated particle clearance from the lung is compromised. Tumors have also
been found to develop with other types of panicles when this clearance mechanism is
diminished. Thus, reductions in the functional activity of the lung's alveolar macrophage
May 1990 6-23 DRAFT-DO NOT QUOTE OR CITE
-------
Figure 6-4. Electron micrograph of an aggregate of particle-filled AM in an alveolus on day
106 following the deposition of 6.8x10s microspheres. Arrows point to Type II pneumocytes
which appear to be present in greater than normal abundance. The microspheres are also
found in Type I epithelial cells.
May 1990
6-24 DRAFT-DO NOT QUOTE OR CITE
-------
population in the clearance process generally appear to be intimately related to the
carcinogenic response to high lung burdens of panicles. Findings that tumors develop in the
lungs of laboratory rats at lesser lung mass or volume burdens of diesel particles than with
a substance like titanium dioxide suggest that the carcinogenic response, however, is not
exclusively related to an over abundance of panicles in the lungs per se. Most likely, the
organic components on diesel panicles, many of which have demonstrated carcinogenic
activity, are importantly involved in the development of lung tumors. The lung's pulmonary
macrophages, which phagocytize deposited diesel particles, probably panicipate in the
gradual in situ extraction and metabolism of procarcinogens associated with the diesel
panicles. Additionally, the normal tumoricidal activities of the pulmonary macrophages may
be compromised upon interaction with excessive numbers of diesel panicles, and diesel
panicle-macrophage interactions could lead to the generation of reactive oxygen species that
have been shown to be at least mutagenic. Caution must be exercised in extrapolating
observations made in animal models to the human condition. Processes and potential
mechanisms discussed herein have largely been derived from animal data. Further research
is required to determine how the activities of human pulmonary macrophages in response
to paniculate diesel exhaust compare with pulmonary macrophages from experimental
animals. Most imponantly, valid dosimetry for the human condition requires the elucidation
of the underlying mechanisms involved in the development of lung tumors following chronic
exposure to whole diesel exhaust.
May 1990
6-25
DRAFT-DO NOT QUOTE OR CITE
-------
6.6. REFERENCES
Adamson. L Y. R.; Bowden. D. H. (1978) Adaptive response of the pulmonary macrophagic system to carbon.
IL Morphologic studies. Lab. Invest. 38: 430-438.
Albert. R. £4 Sellakumar, A. R.; Laslun. S.; Kuschner, M.; Nelson. N.; Snyder, C A. (1982) Gaseous
formaldehyde and hydrogen chlonde induction of nasal cancer in the rat J. NatL Cancer InsL 68:597-
603.
Allen, R. C; Loose, L. D. (1976) Phagocytic activation of a luminol-dependent
chemiluminescence in rabbit alveolar and peritoneal macrophages. Biochem. Biophys. Commun. 69:
245-252.
Ball, L. M.; King, L. C (1985) Metabolism, mutagenicity, and activation of 1-nitropyrene in vivo and in vuro.
Environ. Inter. 11: 355-362.
Beall. G. D.; Repine, J. E.; Hoidal. J. R.; Raps, F. L. (1977) Chemilummscence by human alveolar
macrophages: Stimulation with heat killed bacteria or phorbol mvnstate acetate. Infect. Immunol. 17:
117-120.
Bitterman, P. B.; Aselberg, S.; Crystal. R. G. (1983) Mechansim of pulmonary fibrosis: Spontaneous release
of the alveolar macrophage-denved growth factor in interstitial lung disorders. J. Clin. Invest. 72:
1801-1813.
Bond, J. A.; Mauderly, J. L. (1984) Metabolism and macromolecular covalent binding of [>4C]-l-nitropyrene
in isolated perfused and ventilated rat lungs. Cancer Res. 44: 3924-3929.
Bond, J. A.; Mitchell, C E; Li, A. P. (1983) Metabolism and macromolecular covalent binding of
benzo(a)pyrene in cultures Fischer-344 rat lung type II epithelial cells. Biochem. Pharmacol. 32:3771-
3776.
Bowden. D. H.; Adamson, I. Y. R. (1978) Adaptive responses of the pulmonary macrophagic system to carbon
I. Kinetic Studies. Lab. Invest. 38: 422-438.
Boyd, M. R. (1984) Metabolic activation and lung toxicity: A basis for cell-selective pulmonary damage by
foreign chemicals. Environ. Health Perspect 55: 47-51.
Boyland, E. (1980) The history and future of chemical carcinogenesis. Br. Med. Bull. 36: 5-10.
Cantrell, E. T.; Tyrer, H. W.; Peirano, W. B.; Danner, R. M. (1980) Benzo(a)pyrene metabolism in mice
exposed to diesel exhaust: IL Metabolism and excretion. In: Pepelko, W. E.; Danner, R. M.; Clarke.
N. A^ eds. Health Effects of Diesel Engine Emissions. Health Effects Research Laboratory,
U.S. EPA, Cincinnati, OH, EPA-600/9-80-057a.
Casanova-Schmitz, M.; Starr, T.; Heck, Hd'A. (1984) Differentiation between metabolic incorporation and
covalent binding in the labeling of macromolecules in the rat nasal mucosa and bone marrow by
inhaled [l4C]- and [^-formaldehyde. Toxicol. Appl. Pharmacol. 76: 26-44.
Chen, K. -C; Vostal, J. J. (1982) Aryl hydrocarbon hydroxylase induction in rat pulmonary alveolar
macrophages (PAM) by diesel particulates or their extracts. Pharmacologist 24: 151.
May 1990
6-26
DRAFT-DO NOT QUOTE OR CITE
-------
Chung, F. Young, R.; Hecht, S. S. (1984) Formation of cyclic 1, NJ-propanodeoxyguanos ine adducts in
DNA upon reaction with acrolein or crotonaldehyde. Cancer Res. 44: 990-995.
Clark, R. A.; Klebanoff, S. J. (1975) Neutrophil-mediated tumor cell cytotoxicity: Role of the peroxidase
system. J. Exp. Med. 141: 1442-1447.
Clary, J. J.; Gibson, J. E; Wantz, R. S., eds. (1983) Formaldehyde: toxicology, edpidemiology, mechanisms.
New York, Marcel Dekker.
Conncy, A. K. (1982) Induction of microsomal enzymes by foreign chemicals and carcinogenesis by polycyclic
aromatic hydrocarbons: G. H. A. Clowes Memorial Lecture. Cancer Res. 42: 4875-4917.
Crapo, J. D.; Young, S. L-; Fram, E K.; Pinkerton, K. E.; Barry, B. E; Crapo, R. O. (1983) Morphometric
characteristics of cells in the alveolar region of mammalian lungs. Am. Rev. Respir. Dis. 128: S42-S46.
Dalby, W. E. (1982) Formaldehyde and tumors in hamster respiratoiy tract. Toxicology 24: 9-14.
Denholra, E M.; Phan. S. H. (1989) The effects of bleomyan on alveolar macrophage growth factor secretion.
Am. J. Pathol. 134: 355-363.
Drath. D. B.; Karnovsky. M. L. (1975) Superoxide production bv phagocytic leukocytes. J. Exp. Med. 141:
257-261
Djunc, Z.; Fifer, E K.; Yamazoe, Y.; Beiand. F. A. (1988) DNA binding by 1-mtropyrene and
1,6-diniiropyrene in vuro and in vivo: effects of nitroreductase induction. Carcinogenesis 9: 357-364.
El-Bayoumy, K.; Rivensoo, A.; Johnson, B.; DiBello, J.; Little, P.; Hecht, S. S. (1988) Comparative
tumongemcity of 1-nitropyrene, 1-narosopyrene, and 1-aminopyrene administered by gavage to
Sprague-Dawley rats. Cancer Res. 48: 4256*4260.
Evans, Z.; Shami, S. G.; Martinez, L. A. (1986) Enhanced proliferation of pulmonary alveolar macrophages
after carbon instillation in mice depleted of blood monocytes bv sironiium-89. Lab. Invest. 54: 154-
159.
Feldman. M. Y. (1973) Reactions of nucleic acids and nudeoproteins with formaldehyde. Proe. Nucl. Acid
Res. MoL BioL 13: 1-49.
Flick, D. A.; Gonzalez-Rothi. R. J.; Harris, J. O.; Gilford, G. E (1985) Rat lung macrophage tumor cytotoxxn
production: Impairment by chronic in vivo cigarette smoke exposure. Cancer. Res. 45: 5225-5229.
French, D.; Edsall. J. T. (1945) The reaction of formaldehyde with amino acids and proteins. Adv. Protein
Chem. 2: 277.
Goldstein, L M^ Roos, D.; Kaplan, H. B.; Weissman, G. (1975) Complement and immunoglobulins stimulate
superoxide production by human leukocytes independently of phagocytosis. J. Clin. Invest 56: 1155-
1163.
Grafstrom, R. C; Dypbukt, J. M.; Sundqvist, K.; Edman, C C; Atzori, L. (1987) Effects of automotive
emission-related aldehydes in cultured human bronchial cells. Health Effects Institue. Fourth Ann.
Conf. p. 46.
May 1990
6-27
DRAFT-DO NOT QUOTE OR CITE
-------
Harris, C C (1985) Future directions in the use of DNA adducts as internal dosimeters for monitoring human
exposure to environmental mutagens and carcinogens. Environ. Health Perspect. 62:185-191.
Hatch, G. E.; Gardner, D. E.; Menzel, D. B. (1980) Stimulation of oxidant production in alveolar macrophages
by pollutant and latex panicles. Environ. Res. 32: 121-136.
Heck. Hd'A.; Casanova-Schmitz, M. (1983) Biochemical toxicology of formaldehyde. Rev. Biochem. Toxicol
6: 155-189.
Heinnch, U.; Muhle, R; Takenaka, S.; Ernst, R; Fuhst, R.; Mohr, U.; Pott, F.; Stober, W. (1986) Chronic
effects on the respiratory tract of hamsters, mice and rats after long-term inhalation of high
concentrations of filtered and unaltered diesel engine emissions. J. Appl. ToxicoL 6: 383-395.
Hengst, J. C D.; Kan-Mitchell. J.; Kempf, R. A.; Strumpf, I- J.; Shanna, O. P.; Hibbs, J. B.; Chapman, R
A. Jn Weinberg, J. B. (1978) The macrophage as an antineoplastic surveillance cell: Biological
perspectives. J. ReticulendotheL Soc. 24: 549-570.
Hirose. M.; Lee, M-S.; Wang, C Y.: King, C M. (1984) Induction of rat mammary gland tumors by 1-
nitropyrene, a recently recognized environmental mutagen. Cancer Res. 44: 1158-1161
Hoidal. J. R.: Repine, J. E.; Beall, G. D.; Rasp, F. L; Whne, J. G. (1978) The effect of phorbol myrotate
acetate on metabolism and ultrastructure of human alveolar macrophages. Amer. J. Pathol. 91: 469-
482.
Horton, A. W.; Tye, R.; Steinner, K. L. (1963) Experimental carcinogenesis of the lung.Inhalauon of gaseous
formaldehyde or an aerosol of coal tar by C3H mice. J. NatL Cancer Inst 30: 31-43.
Howard, P. C; Gerrard, J. A.; Milo, G. E.; Fu, P. P.; Beland, F. A^ Kadlubar, F. F. (1983) Transformation
of normal human skin fibroblasts by 1-nitropyrene and 6-nitrobenzo(a)pyrene. Carcinogenesis 4:353-
355.
Howard. A. J.; Mitchell, C E.; Dutcher, J. S.; Henderson, T. R-; McClellan, R. O. (1986) Binding of
mtropyrenes and benzo(a]pyrene to mouse lung deoxyribonucleic acid after preireaiment with
inducing agents. Biochem. PharamcoL 35: 2129-2134.
Iwai, K.; Udagawa, T.; Yamagishi, M.; Yamada, H. (1986) Long-term inhalauon studies of diesel exhaust on
F344 SPF rats. Incidence of lung cancer and lymphoma. In: Ishinishi, N.; Koizumi, A.; McClellan, R.;
Stober, W., eds. Carcinogenic and mutagenic effects of diesel engine exhaust. Amsterdam: Elsevier,
pp. 349-360.
Jenna, D. Sayer, J. M.; Thakker, D. et aL (1980) Caxanogeniaty of polycyclic aromatic hydrocarbons:
the bay-region theory. In: Pullman, B.; Ts'o, P. O. P.; Gelboin, R Eds. Carcinogenesis: Fundamental
Mechanisms and Environmental Effects, Hingham. MA, D. Reidel Publishing Co., pp. 1-12.
Jordana, M.; Richards, C; Irving, I. B. Gauldie, J. (1988) Spontaneous in vino release of alveolar macrophage
cytokines after the intratracheal instillation of bleomycin in rats: Characetrization and kineuc studies.
Am. Rev. Respir. Dis. 137: 1135-1140.
Kan-Mitchell, IC; Hengst, J. C D.; Kempf, R. A; Rothbart, R. K.; Simons, S. M.; Brooker, A. S.; Kortes, V.
L; Mitchell, M. S. (1985) Cytotoxic activity of human pulmonary alveolar macrophages. Cancer, Res.
45: 453-458.
May 1990
6-28
DRAFT-DO NOT QUOTE OR CITE
-------
Kensler, T. W.; Egner, P. A.; Moore, K. G.; Taffe, B. G.; Twerdok. L. E.; Thrush. M. A. (1987) Role of
inflammatory cells in the metabolic act of polycyciic aromatic hydrocarbons in mouse skin. Toxicol.
AppL Pharm. 90: 337-346.
Kerns. W. D.; Pavkov, K. L.; Donofrio. D. J.; Gralla, E. J.; Swenberg, J.A. (1983) Carcinogenicity of
formaldehyde in rats and mice after long-term inhalation exposure. Cancer Res. 34: 4382-4392.
King, C M. (1988) Metabolism and biological effects of nitropyrene and related compounds. Health Effects
Research Institute, Research Report No. 16. pp. 1-31.
King, L. C; Jackson, M.; Ball. L. M.; Lewtas, J. (1983) Binding of l-nitro[l4C]pyrene to DNA and protein in
cultured lung macrophages and respiratory tissues. Cancer Lett. 19: 241-246.
Klebanoff, S. J.; Clark, R. A. (1978) In: The Neutrophil: Function and Clinical Disorders. North-Holland,
Amsterdam.
Lam, C. W.; Casanova. M.; Heck. Hd'A (1985) Depletion of nasal glutathione by acrolein and enhancement
of formaldehyde-inducca DNA-protein cross-linking by simultaneous exposure to acrolein. Arch.
Toxicol. 58: 67-71.
Laskin. S.; Kuschner. M.; Sellekumar. A; Katz, G. V. (1976) Combined carcmogen-irniant animal mhalatton
studies. In: Aharson, E. F.; Ben-David, A.; Klingberg, M. A. Eds. Air Pollution and the Lung. Wiley,
Jerusalem. 313 pp.
Lee, JC P.; Henry III, N. W.; Trochimowicz, H. J. Reinhardt, C F. (1986) Pulmonary response to impaired
lung clearance in rats following excessive TiOz dust deposition. Environ. Res. 41: 144-167.
Lee, K. P.; Ulrich, C E.; Geil, R. G.; Trochimowicz. H. J. (1988) Effects of inhaled chromium dioxide dust
on rats exposed for two years. Fund. AppL ToxicoL 10: 125-145.
Lehnert, B. E.; Valdez, Y. E.; Bomalaski. S. H. (1988) Analyses of particles in the lung free cell, tracheo-
bronchial lymph nodal, and pleural space compartments following their deposition in the lung as
related to lung clearance mechanisms. Ann. Occup. Hyg. 32: 125-140.
Lehnert. B. E: Ortiz, J. B.; London. J. E.; Cline A F. (1989) Migratory behavior of alveolar macrophages
during clearance of light to heavy lung burdens of particles. 1989 Annual Meeting of the American
Thoracic Society, Cinnanati, Ohio, May 14-17.
Leung, H. W.; Henderson, R. F.; Bond, J. A. (1985) Facilitation of removal and metabolism of benzo(a)pyrene
from diesel particles by rat lung and liver microsomes. Lovelace Biomedical and Environmental
Research Institute. Annual Report LMF-114 pp. 135-138.
Levin, W.; Buening, M. K.; Wood, A. W.; et aL (1980) An enantiomeric interaction in the metabolism and
tumongeniaty of (+) and (-)-benzo(aJpyrene 7,8-oxide. J. Biol. Chem. 255: 9067-9074.
Lippmann, M. (1980) Health significance of exposures to sulfur oxide air pollutants. In: Shnner. D. S.;
Richmond, C. R.; Lindberg. S. E.. Eds. Atmospheric Sulfur Deposition Environmental Impact and
Health Effects. Ann Arbor, MI. Ann Arbor Science Publishers. Inc. pp. 85-97.
Lowne, D. B.: Aber, V. R. (1977) Superoxide production by rabbit alveolar macrophages. Life Sci. 21: 1575-
1584.
May 1990
6-29
DRAFT-DO NOT QUOTE OR CITE
-------
Maher, V. M.; Patton, J. D.; McCormick, J. J. (1988) Studies on the metabolism and biological effects of
mtropyrene and related mtro-polycyclic aromauc compounds in diploid human fibroblasts. Health
Effects Research Institute. Research Report No. 17. p. 16.
Mauderly, J. L-; Jones, R. K.; Griffith, W. C; Henderson, R. F.; McQellan, R. O. (1987) Diesel exhaust is a
pulmonary carcinogen in rats exposed chronically by inhalation. Fund. Appl. Toxicol 9: 208-221.
McLemore, T. L.; Martin, R. R-; Wray, N. P.; CantrelL, E. T.; Busbee, D. L.(1981) Reassessment of the
relauonship between aryl hydrocarbon hydroxylase and lung cancer. Cancer 48: 1438-1443.
Miles, P. R-; Lee, P.; Thrush, M. A.; Van Dyke, K. (1977) Chemiluminescence associated with phagocytosis
of foreign particles tn rabbit alveolar macrophages. Life ScL 20: 165-170.
Mitchell, C. E. (1982) Distribution and retention of benzo(a)pyrene in rats after inhalation. Toxicol. Leu. 11:
35-42.
Morrow, P. E. (1986) The setting of paniculate exposure levels for chronic inhalation toxicity studies. J. Am.
Coll. Toxicol 5: 533-544.
Morse. M. A; Carlson. G. P. (1985) Distribution and macromoiecular binding of benzo(a)pyrene in Senear
and BALB/c mice following topical and oral administration. J. Toxicol. Environ. Health 16: 263-276.
Munsch, N.; deRecondo, A M.; Frayssinet, C (1974) In vuro binding of 3H-acrolein to regenerating rat liver
DNA polymerase. Expermentxa 30: 1234-1236.
Nesnow, S.; Triplett, L. Slaga, T. J. (1984) Tumor initiating activities of 1-nitropyrene and its nitrated
products in Senear mice. CanceT Lett. 23: 1-8.
Patel, J. M; Wood, J. C; Leibman, K. C (1980) The biotransformation of allyl alcohol and acrolein in rat liver
and lung preparations. Pharmacologist 20: 181.
Phillips, B. J.; James, T. EL B.; Anderson, D. (1984) Genetic damage in CHO cells exposed to enzymatically
generated active oxygen soecies. Mut. Res. 126: 265-271.
Pitts. J. N.. Jr. Lokensgard. D. M.: Harger, W.; et al. (1982) Mutagens in diesel exhaust particulates,
identification and direct activities of 6-nitrobenzo(a)pyrene, 9-mtroanthracene. 1-nitropyrene. and JH-
phenantbra(4,5-bcd)pyran-5-one. Mutat. Res. 103: 241-249.
Plopper, C G.; Hyder, D. M.; Weir, A J. (1983) Centnaanar alteration in lungs of cats chronically exposed
to diesel exhaust. Lab. Invest. 49: 391-399.
Pott, F.; Stober, W. (1983)Carcinogenicity of airborne combustion products observed in subcutaneous tissue
and lungs of laboratory rodents. Environ. Health Perspea. 47: 293-303.
Randerath, K.; Randerath, E.; Agarwal. H. P.; Gupta, R. C; Schurdak, M. E.; Reddy, M. V. (1985)
Postlabeling methods for carcinogen-DNA addua analysis. Environ. Health Perspect. 62: 57-65.
Reddy, M. V.; Randerath. K. (1986) Nuclease Pl-mediated enhancement of sensitivity of phosphorus-32-
postlabeling test for structurally diverse DNA adducts. Carcinogenesis (London) 7: 1543-1551.
May 1990
6-30
DRAFT-DO NOT QUOTE OR CITE
-------
Rister, M.; Baehner. R. L. (1977) Effect of hyperoxia on superoxide anion and hydrogen peroxide production
of polymorphonuclear leukocytes and alveolar macrophages. Brit J. HaematoL 36: 241-248.
Root, R. K.; Metcalf, J. A. (1977) H2Oz release from human granulocytes during phagocytosis. J. Clin. Invest.
60: 1266-1279.
Saffioti, U.; Cefin. F.; Kolb. L. R; Shubik, P. (1965) Experimental studies of the conditions of exposure to
carcinogens for lung cancer induction. J. Air Pollut. Corn. Assoc 15: 23-25.
Sandusky, C B.; Cowden, M. W.; Schwartz, S. L. (1977) Effea of panicle size on regurgitive exocytosis by
rabbit alveolar marcophages. In: Pulmonary Macrophages and Epithelial Cells. CONF-760927,
Nauonal Technical Information Service, U.S. Dept. of Commerce, pp. 85-105.
Sato, S.; Ohgald, R; Takayama, S.; et aL (1986) Carcinogenicity of dinitropyrenes in rats and hamsters. In:
Ishinishi, N.; Koizumi, A.; McClellan. R. O.; Stober, W„ eds. Carcinogenic and Muatgeruc Effects of
Diesel Engine Exhaust. Amsterdam, Elsevier, pp. 271-277.
Schuetzle. D.; Riley, T.L.. Prater, T. J.; Harvey, T. M.: Hunt. D. F. (1982) Analysis of nitrated polycyclic
aromatic hydrocarbons in diesel particulates. Anal. Chem. 54: 265-271.
Sone, S. (1986) Role of alveolar macrophages in pulmonary neoplasias. Biochim. Biophys. Acta 823: 227-245.
Sone, S.; Brennan, L. M. Creasia, D. A. (1983) In vivo and ui vitro NO, exposures enhance phagocytic and
tumoncidal activities of rat alveolar macrophages. J. Toxicol. Envnon. Health 11: 151-163.
Squire, R. A.; Cameron, L. L. (1984) An analysis of potential carcinogenic risk from formaldehyde. Regulatory
ToxicoL Pharmacol. 4: 107-129.
Stenback. F.; Rowland, Sellakumar, A. (1976) Caranogenicity of benzo(a)pyrene and dusts in the hamster
lung (instilled intratrachealiy with titanium oxide, aluminum oxide, carbon and feme oxide). Oncology
33: 29-34.
Stowers. S. J.: Anderson. M. W. f 1985^ Formation and persistence of benzofa)nyrene metaboltte-DNA
adducts. Envnon. Health Perspect- 62: 31-39.
Strom. K. A. (1984) Response of pulmonary cellular defenses to the inhalation of high concentrations of diesel
exhausL J. ToxicoL Environ. Health 13: 919-944.
Sun. J. D.; Wolff, R. K.; Maio, S. M.; Barr, E B. (1988) The influence of adsorption to carbon black panicles
on the retention and metabolic activation of benzo(a)pyrene in rat lungs following inhalation
exposure or intratracheal instillation. Inhalation Toxicology, In Press.
Thyssen, J.; Althoff, J.; Kimmerle, G.; Mohr, U. (1981) Inhalation studies with benzo[a]pyrene in Synan
golden hamsters. J. Natl. Cancer Inst 66: 575-577.
Urban, J. L.; Shepard, H. M.; Rothstein, J. L.; Sugarman, B. J. (1986) Tumor necrosis factor: A potent
molecule for tumor cell killing by activated macrophages. Proc. Natl Acad. Sci. U.S.A. 83:5233-5237.
May 1990
6-31
DRAFT-DO NOT QUOTE OR CITE
-------
U.S. Environmental Protection Agency (1985) Health and Environmental Effects Profile for Formaldehyde.
Prepared by Syracuse Resaerch Corporation under Contract No. 68-03-3228 for Environmental
Criteria and Assessment Office, Office of Health and Environmental Assessment, Cincinnati, OH.
ECAO. EPA report No. EPA/600/X-85/362.
Van Duuren, B. L.; Langseth, L.; Orris, L.; Teebor, G.; Nelson, N.; Kuschner, M. (1966) Carcinogenicity of
epoxides, lactones, and peroxy compounds. IV. Tumor response in epithelial and connective tissue
in mice and rats. JNCI J. NatL Cancer Inst 37: 825-838.
Warheit, D. B.; George, G.: Hill. LH.; Snydennan, R.; Brody, A. R. (1985) Inhaled asbestos activates a
complement-dependent chemoattractant for macrophages. Lab. Invest. 52: 505-514.
Weinstein, L B.; Troll, W. (1977) National Cancer Institute Workshop on tumor promotion and cofactors in
carcinogenesis. Cancer Res. 37: 3461-3463.
Weiss, S. J.; Rustagei, P. K.; LoBuglio. A. F. (1978) Human granulocyte generauon of hydroxyl radical. J. Exp.
Med. 147: 316-323.
Weitzman. S. A.. Stossel. T. P. (1981) Mutation caused by human phagocytosis. Science 212: 546-547.
White. H. J.. Garg, B. D. (1981) Early pulmonary response of the rat lung to inhalation of high concentration
diesel particles. J. Appl. Toxicol. 1: 104-110.
Wislocki, P. G.; Bagan, E. S.; Lu, A. Y. H.; et al. (1986) Tumorigemcity of nitrated derivatives of pyrene.
benz(a]anthracene, chrysene, and benz(a] pyrene in the newborn mouse assay. Carcinogenesis
(London) 7: 1317-1322.
Wolff, R. K.: Bond, J. A.; Sun, J. D.; et aL (1989) Effects of adsorption of benzo[a]pyrene onto carbon black
particles on levels of DNA adducts in lungs of rats exposed by inhalation. Toxicol. Appl. Pharmacol.
97: 289-299.
May 1990
6-32
DRAFT--DO NOT QUOTE OR CITE
-------
7. CARCINOGENICITY OF DIESEL EMISSIONS
IN LABORATORY ANIMALS
7.1 INTRODUCTION
Significant concern has been expressed regarding the potential carcinogenicity of
diesel exhaust emissions, with special emphasis being placed on the particulate phase of the
emissions. This particulate phase is composed of a carbon core representing about 80
percent of the total particle mass are adsorbed to this carbon core. A great variety of
organic compounds, including polvcyclic aromatic hydrocarbons (PAHs), Benzo[a]pyrene and
1-nitropyrene are exhaust components that have received special attention regarding
carcinogenicity. These organics may be strongly or weakly bound to the carbon core and
represent varying amounts of the total particle mass. Qualitative and quantitative
relationships for these organics depend on such variables as fuel quality, engine design, and
engine operating conditions. The primary particle diameter ranges from 10 to 80 nm, but
may form aggregates with mass median diameters of 0.2 to 0.3 nm (Vuk et al., 1976;
Carpenter and Johnson, 1979). The ease with which these particles and their associated
organics can be respired provides a basis for health hazard concerns. This fact and the
reported mutagenicity (Huisingh et al., 1978) and carcinogenicity (Kotin et al., 1955) of
solvent extracts of diesel soot have provided a rationale for examining the potential
carcinogenicity of diesel emissions. Zamora et al. (1983) provided evidence that diesel
exhaust particle extracts contained components that acted as weak tumor promoters in vitro.
Recently, emphasis has been directed to assessing the carcinogenic potential of the panicle-
associated organics using whole animal studies and to understanding the mechanics and
ramifications of deposition, retention, and clearance of the particle phase of diesel exhaust.
May 1990
7-1
DRAFT-DO NOT QUOTE OR CITE
-------
7.2 CARCINOGENICITY STUDIES IN LABORATORY ANIMALS
This chapter summarizes studies assessing the carcinogenic potential of diesel exhaust
in laboratory animals. Experimental protocol for the inhalation studies consisted of exposure
to diluted exhaust in whole animal exposure chambers (usually of chronic duration) using
rats, mice, and hamsters as the model species. Some of these studies used both filtered
(free of particulate matter) diesel exhaust and unfiltered (whole) diesel exhaust to
differentiate gaseous phase effects from effects induced by the particulate matter and its
adsorbed components. Diesel exhaust paniculate matter concentrations ranged from 0.35
to 7.0 mg/m3. Clean air was used in the control group exposures. Studies providing both
positive and negative results as well as those indicating inconclusive findings have been
reviewed.
Also included are studies that assessed the carcinogenic and/or tumongenic effects
of diesel exhaust particles and solvent extracts of these particles following dermal
application, (s.c.) subcutaneous injection, (i.p.) intraperitoneal injection, or (itr.) intratracheal
instillation in rodents.
7.2.1 Long-term Inhalation Studies
Mauderly et al. (1987) provided data affirming the carcinogenicity of automotive
diesel engine exhaust on F344/Crl rats following chronic inhalation exposure. Male and
female rats were exposed to diesel engine exhaust at nominal paniculate matter concen-
trations of 0.35 (n = 366), 3.5 (n = 367), or 7.0 (n = 364) mg/m3 for 7 h/d, 5 d/week up to
30 mo. Sham-exposed (n = 365) controls breathed filtered room air. The exhaust was
generated by 1980 Model 5.7-L Oldsmobile V-8 engines operated through continuously
repeating U.S. Federal Test Procedure (FTP) urban certification cycles. The engines were
equipped with automatic transmissions connected to eddy-current dynamometers and
flywheels simulating resistive and inenial loads of a mid-size passenger car. The D-2 diesel
control fuel (Phillips Chemical Co.) met U.S. EPA certification standards and contained
approximately 30 percent aromatic hydrocarbons and 0.3 percent sulfur. Following passage
through a standard automotive muffler and tail pipe, the exhaust was diluted 10:1 with
May 1990
7-2
DRAFT-DO NOT QUOTE OR CITE
-------
filtered air in a dilution tunnel and serially diluted to the final concentrations. The primary
dilution process was such that particle coagulation was retarded, and the serial dilution
procedure was chosen because of a need for greater precision and stability of exhaust
concentration. Flow through the 2.2-m3 exposure chambers (H-200, Hazleton Systems,
Aberdeen, MD) was set at 0.4 m3/min, which provided an average residence time of 5.5 min
in the chamber. Mokler et al. (1984) provides a detailed description of the exposure system.
The gas-phase components of the diesel exhaust atmospheres are presented in Table 7-1.
No exposure-related changes in body weight or life span were noted for any of the exposed
TABLE 7-1. GAS-PHASE COMPONENTS OF CONTROL
AND DIESEL EXHAUST EXPOSURE ATMOSPHERES'
Component
Control
Low
Medium
High
Carbon monoxide (ppm)
1(1)
3(1)
17(7)
30(13)
Nitric oxide (ppm)
0
0.6 (0.3)
5.4 (1.5)
10.0 (2.6)
Nitrogen oxide (ppm)
0
0.1 (0.1)
03 (0.2)
0.7 (0.5)
Hydrocarbon vapor (ppm)
3(1)
4(1)
9(5)
13 (8)
Carbon dioxide (percent)
0.2 (0.04)
0.2 (0.03)
0.4 (0.06)
0.7 (0.1)
The mean (SE) of weekly mean values.
Source: Adapted from Mauderly et al, 1987.
animals, nor were there any signs of overt toxicity. Collective lung tumor incidence was
greater (z statistic, p <0.05) in the high (7.0-mg/m3) and medium (3.5-mg/m3) exposure
groups (12.8 percent and 3.6 percent, respectively) vs. the controls and low (0.35-mg/m3)
exposure groups and (0.9 percent and 1.3 percent, respectively). Bronchoalveolar adenomas,
adenocarcinomas, and squamous cysts (these were considered benign except for two that
were classified as squamous cell carcinomas because of the presence of less differentiated
cells and invasion of blood and lymph vessels) were identified. Using the same statistical
May 1990
7-3
DRAFT-DO NOT QUOTE OR CITE
-------
analysis on specific tumor types, adenocarcinoma plus squamous cell carcinoma and
squamous cyst incidence was significantly greater in the high exposure group, and the
incidence of adenomas was significantly greater in the medium exposure group. A significant
(p <0.001) dose-response relationship was obtained for tumor incidence relative to exposure
concentration and lung burden of particulate matter (soot). These data are summarized in
Table 7-2. A logistic regression model estimating tumor prevalence as a function of time,
dose (lung burden of soot), and sex indicated a sharp increase in tumor prevalence for the
high dose level at about 800 d after the commencement of exposure. A less pronounced,
but definite increase in prevalence with time was predicted for medium and low dose levels.
The model estimates closely agreed with the observed data. The particulate matter burden
(mg/lung) of rats exposed to 0.35. 3.5, or 7.0 mg soot/m3 for 24 mo was 0.6, 11.5, and 20.8,
respectively, and served to affirm the greater accumulation that was due to decreased
clearance following high exposure conditions. The greater incidence of lung DNA adducts
for high-level exposed rats and the progressive accumulation of particulate matter in the
lungs of these rats were used to support both genetic and epigenetic etiology of the observed
lung tumors.
In summary, this study demonstrates a dose-dependent pulmonary carcinogenicity of
diesel exhaust particulate matter following long-term inhalation exposure, and increased
lung particulate matter burden resulting from this high-level exposure with subsequent
decreased clearance. Furthermore, the logistic regression model indicated that both lung
particulate matter burden and exposure concentration are useful as dose terms.
Wong et al. (1986) reported an increase in DNA adduct formation in male and
female F344 rats exposed to whole diesel exhaust (7.1 mg particles/m3) for 7 h/d
(0800 - 1500 h), 5 d/week for 31 mo. The diesel exhaust was generated by a 5.7-L
Oldsmobile engine burning U.S. EPA-certified fuel connected to an automatic transmission
and an eddy-current dynamometer and was operated according to the (FTP) urban driving
cycle (same emission generating system as described under Mauderly et al., 1987). The mass
median diameter of the diesel exhaust particles was 0.23 nm. The gas-phase components
of the diesel exhaust exposure and control atmospheres are listed in Table 7-3. Phosphorus-
32 postlabeling was applied to DNA that was extracted from six control and six
May 1990 7-4 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-2. TUMOR AND LUNG PARTICULATE MATTER (SOOT) BURDEN IN MALE
AND FEMALE F344 RATS EXPOSED TO DIESEL EXHAUST FOR 24 MO
Exposure
concentration
(mg/m')
Lung soot burden
(mg/lung)
Percentages of rats with tumors
Adenomas
Adenocarcinomas
+ squamous cell
carcinomas
Squamous cysts
All tumors
7.0 (high)
208*0.8
04
7.5*
4.9*
12 8*
3.5 (medium)
11.5i0.5
2.3*
05
09
3 6*
0.35 (low)
0.6t0.02
0
1 3
0
1.3
0 (control)
0
0
0.9
0
0.9
'Significant increase relative lo control at p <0.05, z-test, 364 to 367 rats per group.
Source: Adapted from Maudcrly el al., 1987.
-------
TABLE 7-3. GAS-PHASE COMPONENTS OF CONTROL
(CLEAN AIR)AND DIESEL EXHAUST ATMOSPHERES
Components
Exposed rats
Control rats
X
SD
X
SD
Particles ng/m3
7082
808
13
6
Carbon dioxide ppm
6643
1320
2005
290
Carbon monoxide ppm
29.7
12.9
1.0
0.7
Hydrocarbon vapors ppm
13.4
8.3
2.6
0.6
Nitrogen dioxide ppm
0.68
0.48
0*
or
Nitrogen oxide ppm
10.0
2.6
0*
or
Ammonia ppm
0.7
0.6
1.1
3.0
'Below detection limits.
Source: Wong et aL, 1986.
exhaust-exposed rats (males and females). Characterization of the adducts and identification
of the exhaust components responsible for their formation were not within the scope of the
study. The lungs of exhaust-exposed rats were darkly pigmented and contained diesel
panicle-laden macrophages. Aggregates of these macrophages were frequently associated
with alveolar wall fibrosis, bronchiolar metaplasia, and, occasionally, squamous metaplasia.
Lungs from control rats were not darkly pigmented and had relatively unaltered airway and
alveolar macrophage structure. Autoradiographic analysis revealed a higher intensity and
elevated levels of DNA adducts in the exhaust-exposed rats. The authors indicated that
quantitative and qualitative data regarding DNA adducts resulting from diesel exhaust
exposure may be useful for extrapolation to potential effects in humans.
Lifetime inhalation exposure (19 h/d, 5 d/week) of female Wistar rats and male and
female NMRI mice and Syrian golden hamsters to filtered and unaltered (total) diesel
May 1990 7-6 DRAFT-DO NOT QUOTE OR CITE
-------
exhaust (particulate matter concentration of 4 mg/m3) was studied by (Heinrich et aL,
1986a,b). Animals were exposed for a maximum of 2~5 yr. The exposure system as
described by Heinrich et al. (1986b), used a 40 kw 1.6-L diesel engine operated continuously
under the U.S. 72 FTP driving cycle. The engines used European Reference Fuel with a
sulfur content of 0.36 percent. Filtered exhaust was obtained by passing engine exhaust
through a Luwa FP-65 HT 610 particle filter heated to SO* C and a secondary series of filters
(Luwa FP-85, Luwa NS-30, and Drager CH 63302) at room temperature. The filtered and
unfiltered exhausts were each diluted 1:17 with filtered air and passed through respective 12-
m3 exposure chambers. Mass median diameter of the diesel exhaust particulate matter was
0.35±0.10 nm. The gas-phase components of the diesel exhaust atmospheres are presented
in Table 7-4.
Hamsters exposed to total or filtered diesel exhaust exhibited no significant difference
in body weight gain relative to the control animals. Similarly, there was no significant change
in body weight gain of rats or mice exposed to filtered exhaust. However, after approxi-
mately 480 d of exposure to total (unfiltered) diesel exhaust, rats and mice exhibited a
decrease in weight gain and a weight loss, respectively, compared to the control animals and
those exposed to filtered exhaust. For all three species, mortality (as determined by the
median experimental lifetime) was not affected by either of the exposure protocols.
Lung weights (wet and dry) for all three species were significantly (p <0.05) increased
in the total exhaust exposure groups, but not in those animals exposed to filtered exhaust
or clean air. Sex-dependent differences were not detected. Mechanical lung function
parameters including tidal volume, minute volume, compliance, resistance, resistance with
acetylcholine, functional residual capacity, and respiratory rate were also assessed. A
significant (p <0.05) reduction in airway resistance in hamsters was detected after 1 y of
exposure, and significant reductions in resistance and compliance were noted for rats after
2 y exposure to total diesel exhaust. No significant changes in lung mechanics were reported
for animals exposed to filtered exhaust or clean air. The lung mechanic tests were not
performed using mice.
The total exhaust exposure protocol resulted in a 16-percent tumor incidence for
female SPF Wistar rats. In addition to the bronchio-alveolar adenomas and squamous cell
May 1990 7-7 DRAFT-DO NOT QUOTE OR CITE
-------
TABIJi 7-4. GAS-PHASE COMPONENTS OF CONTROL AND DIESEL EXHAUST ATMOSPHERES*
Exhaust dilution (exhaust/exhaust + air) 1/17
Component Control (filtered air) Filtered cxluust Total exhaust
CO (ppm)
016i0 27
II 111 92
12 5 ±2 18
CO, percent
010x0 01
0 35 id 05
0.38i0.05
SO, (ppm)
—
I02t062
1.1210 89
NO, (ppm)
--
9.911 K()
11.4 ±2.09
NO (ppm)
—
8 711 8-1
10.0 ±2 09
NO, (ppm)
—
1.2 i() 26
1.5 ±0.33
QH,,, (ppm)
3.510 29
5 2 tO 65
5.5 iO 69
CH, (ppm)
2 3 ill 17
2-1.0 2(1
2.6iO.I9
CnHm without CH, (ppm)
1.3 j().I5
2 9i0 5(1
3.110.53
'MeaniSD.
Source: Adapted from Hcinrich et al., 1986b.
-------
tumors, there was a high incidence of bronchio-alveolar hyperplasia (99 percent) and
metaplasia of the bronchio-alveolar epithelium (65 percent). After 3 and 8 mo of exposure,
these rats (n = 95) also exhibited a 2.5-fold increase in clearance halftime for a radioactive
test aerosol (59Fe203). This decreased clearance, however, was partially ameliorated after
12 and 19 mo. Filtered diesel exhaust was not tumorigenic to female Wistar rats (n = 92).
Control rats (n = 96) exhibited no spontaneous tumor incidence. Following lifetime (19 h/d,
5 d/week, maximum of 120 weeks) exposure to filtered (n = 93) and unfiltered (n = 76)
diesel exhaust, female NMRI mice exhibited a tumor incidence of 31 percent (12 percent
adenomas, 19 percent adenocarcinomas) and 32 percent (15 percent adenomas, 17 percent
adenocarcinomas), respectively (Heinrich et al., 1986b; Stober, 1986). Control mice (n =
84) had tumor incidences of 11 percent and 2.4 percent for adenomas and adenocarcinomas,
respectively. When considering the incidence of only malignant tumors in the controls (2.4
percent), filtered exhaust (19 percent), and unfiltered exhaust (17 percent) groups, the
effects of exhaust exposure are more dramatic. These data are summarized in Table 7-5.
No pulmonary tract tumors were observed for any of the 288 exhaust-exposed Syrian
golden hamsters. Prior treatment with 4.5 mg diethylnitrosamine/kg, s.c^ or 20 weekly
intratracheal instillations of 250 ^g B[a]P produced no significant evidence of syncarcino-
genicity in hamsters exposed to diesel exhaust. Similar studies using mice (50 or 100 jzg
B[a]P intratracheal^ for 10 or 20 weeks, respectively, or 50 /ig dibenz[a,h]anthracene
intratracheally for 10 weeks) were inconclusive regarding the syncarcinogenicity of diesel
exhaust. Rats pretreated with 25 weekly s.c. injections of 250 or 500 ng dipentylnitrosamine
(DPN)/kg and exposed to unfiltered diesel exhaust exhibited a significantly increased
incidence of squamous cell carcinomas of the lungs when compared to clean air controls.
The rate of DPN-induced kidney and liver tumors was not influenced by exposure to total
diesel exhaust. The study identified the rat as a sensitive model for diesel exhaust-induced
tumorigenicity, but indicated that further investigations were required to ascertain the precise
mechanism of tumorigenicity.
A long-term inhalation study (Ishinishi et al., 1986) examined the effects of emissions
from light-duty (LD) and heavy-duty (HD) diesel engines on male and female Fisher 344/Jcl
rats. The LD engines were 1.8-L, 4-cylinder, swirl chamber type power plants, and the HD
May 1990 7-9 DRAFT-DO NOT QUOTE OR CITE
-------
engines were 11-L, 6-cylinder, direct injection type power plants. The engines were
connected to eddy-current dynamometers and operated at 1200 rpra (LD engines) and 1700
rpm (HD engines). Nippon Oil Co. JIS No. 1 or No. 2 diesel fuel was used. The 30-mo
whole body exposure protocol (16 h/d, 6 d/week) employed diesel exhaust particle
concentrations of 0, 0.4, 1, 2, or 4 mg/m3 from HD engines and 0, 0.1, 0.4, 1, or 2 mg/m3
from LD engines. Benzo(a)pyrene concentrations were reported as 4.4 and 2.8 ng/g of
particulate matter, and 1-nitropyrene concentrations were 57.1 and 153 Mg/g of particulate
matter for the LX> and HD engines, respectively. An analysis of gas-phase components are
presented in Table 7-6. Inhalation chambers were maintained on a 12 h light-dark cycle at
23±2*C and 55 percent relative humidity. The animals inhaled the exhaust emissions from
1700 to 0900 h (although not specifically stated, it is assumed that this represents approxi-
mately 10 to 12 h of exposure during darkness). One hundred twenty male and 95 female
rats were used for each exposure concentration.
The long-term exposure to diesel exhaust from LD and HD engines did not signifi-
cantly affect the survival of the rats. Body weight data, presented for only female rats,
indicated a 15- to 20-percent decrease in mean body weight relative to the controls.
Indication of statistical significance was not provided for these data. Light microscopic
examination revealed a dose- and time-dependent pulmonary burden of panicles that was
evident after 6 mo exposure to diesel emissions. Following 24 mo of exposure, the platelet
count was significantly decreased in male rats of the 4-mg/m3 exposure group. Significant
decreases in serum cholinesterase, free cholesterol, and phospholipids were detected in rats
exposed to 4 mg/m3 of HD engine exhaust.
For the experiments using the LD series engines, the incidence of carcinomas was
dose-independent with the highest incidence (4.1 percent) being in the 1-mg/m3 exposure
group. No significant difference was found among the exposure groups for the LD series
engines. For the experiments employing the HD series engines the carcinoma incidence was
dose-dependent with the highest incidence (6.5 percent) occurring in the 4-mg/m3 exposure
group. The only statistically significant difference reported was that between the 4-mg/m3
and 0 mg/m3 exposure groups for the HD series engines. Tumor incidence data for this
May 1990
7-11
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-6. GAS-PHASE COMPONENTS OF DIESEL EXPOSURE ATMOSPHERES
Gaseous concentration
Group
NO.
NO
NO,
CO
CO,
SO,
1ITMC
LTHC
HCHO
SO,1
o,
Experiment
mg/m'
ppm
ppm
ppm
ppm
%
ppm
ppmc
ppmc
ppm
Mg/m,
%
Carcinogenicity
4
37 45
34 45
3 00
1291
0 360
4 57
15 65
7.62
0.29
361
20.4
& long-term
2
2167
19.99
1 68
7.75
0215
2 82
9 94
5.90
0 18
198
206
inhalation
1
13 13
12.11
1.02
4 85
0140
1.79
7.15
5.16
Oil
111
20.7
(HD)
0.4
6 17
5.71
0 46
2 65
0084
098
4 63
4.27
005
62.9
20.8
0
0061
0.042
0 021
0.63
0.035
0 06
2 43
3.50
0.003
0.49
20.8
Carcinogenicity
2
18 93
20.34
1 41
7.10
0.418
4 70
5.49
3.57
0.13
315
20.3
& long-term
1
9 44
10.14
0 70
3.96
0219
2.42
3 82
2.87
0.07
151
20.5
inhalation
04
381
406
0 26
2 12
0.105
106
2 93
2.51
0.03
62.4
20.7
(LD)
0.1
1 16
1.24
008
1.23
0.050
0.38
2 44
2.27
0.01
18.8
20 8
0
0 033
0.044
0.011
080
0.026
006
2.20
2.17
0.002
0.41
20.8
Source: Adapted from khinishi el al, 1986
-------
experiment are presented in Table 7-7. No information was reported regarding a correlation
between carcinoma incidence and concentrations of benzo(a)pyrene and 1-nitropyrene.
Iwai et al. (1986) also examined the long-term effects of diesel exhaust inhalation on
female F344 rats. The exhaust was generated by a 0.37-L displacement truck engine. The
exhaust was diluted 10:1 with clean air at 20 to 25 *C and 50 percent relative humidity in a
620-cm-long dilution tunnel with an inside diameter of 45 cm. The engines were operated
at 1000 rpm with an 80 percent engine load. These operating conditions were found to
produce exhaust with the highest particle concentration and lowest N02 and SO; content.
The exposure chambers were ventilated 10 times per hour and maintained at -5 to -7 mm
H,0. For those chambers using filtered exhaust, proximally installed HEPA filters were
employed. Three groups of 24 rats each were exposed to unfiltered diesel exhaust, filtered
diesel exhaust, or filtered room air 8 h/d, 7 d/week for 24 mo. Particle concentration was
4.9-mg/m3 for unfiltered exhaust. Concentrations of gas-phase exhaust components were
30.9 ppm NOr 1.8 ppm NO^ 13.1 ppm SO^ and 7.0 ppm CO.
No lung tumors were found in the 2-yr control (filtered room air) rats, although one
adenoma was noted in a 30-mo control rat providing a spontaneous tumor incidence of 4 J
percent. No lung tumors were observed for rats exposed to filtered diesel exhaust. Four
of 14 rats exposed to unfiltered diesel exhaust for 2 yr developed lung tumors, two of these
tumors being specified as malignant. Five rats of this 2-vr exposure group were subsequently
placed in clean room air (3 to 6 mo) with four eventually (time not specified) exhibiting lung
tumors (one malignancy). Thus, the lung tumor incidence for rats exposed to whole diesel
exhaust was 42.1 percent (8/19) with 26.3 percent (5/19) being malignant. The tumors types
identified were adenomas (3/19), adenocarcinomas (1/19), adenosquamous carcinoma (2/19),
squamous carcinoma (1/19), and large-cell carcinoma (1/19). The lung tumor incidence in
rats exposed to whole diesel exhaust was significantly greater than that of controls (p <0.01).
Tumor data are summarized in Table 7-8. Malignant, splenic lymphomas were detected in
37.5 percent of the rats in the filtered exhaust group and in 25.0 percent of the rats in the
unfiltered exhaust group, these values being significantly (p <0.05) greater than the 8.2-
percent incidence noted for the control rats. The study demonstrates production of lung
cancer in rats following 2-yr exposure to unfiltered diesel exhaust. Additionally, the
May 1990 7-13 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-7. TUMOR INCIDENCE IN MALE AND FEMAIJi 1-344 RATS FOLLOWING
LONG-TERM (30 MO) INHALATION EXPOSURE TO DIESEL EXHAUST
Percentages of Rals Wilh Tumors
Exposure
concentration
mg/mJ Lung soot burden Adenomas Carcinoma/ All tumors Total Tor group
l.D
20 M NP 0(0/64) 31(2/64) 31 M 24
F NP 1 7 (1/60) 0 (0/60) I 7 F
10 M NP 0(064) 4 7 (3/64) 47 M 4 1
F NP 0(0/59) 3 4 (2/59) 34 F
04 M NP 16(1/64) 0(0/64) I 6 M 08
F NP 0(0/61) 0(0/61) 0 1-
01 M NP 0(0/64) 16 (1/64) I6M 2 4
F NP 1 7(1/59) 1 7 (1/59) 34 F
Control
M NP 0(0/64) 3 1 (2/64 ) 3 1 M 3 3
F NP 1 7(1/59) 1 7 (1/59) 34 F
III)
40 M NP 0(1/64) 78(5/64) 78 M 65b
F NP 0(0/60) 5 0(3/60) 50 F
20 M NP 0(0/64) 4 7(3/64) 47 M 3 3
F NP 0(0/59) 1 7 (1/59) 1.7 F
10 M NP 0 (0/64) 0 (0/64) 0 M 0
F NP 0(0/64) 0(0/64) OF
04 M NP 0(0/64) 1 6(1/64) I6M 08
F NP 0(0/59) 0(0/59) OF
Control
M NP 0(0/64) 0(0/64) OM 0 8
F NP 0 (0/59) I 7 (1/59) 1.7 F
*Si|uamouj cell carcinomas, adi'iinsqu.imnus c.ircinom.is, adcn
-------
TABLE 7-8. TUMOR INCIDENCE IN FEMALE F344 RA1S FOLIjOWING LONG-TERM (24 MO)
EXPOSURE TO FILTERED OR UNFILTERED DIESEL EXHAUST, OR CLEAN AIR
Percentage of nils with tumors
Exposure
concentration Lung soot Adenosquamous Squamous Large cell Splenic
(mg/m1) burden Adenomas Adenocarcinomas carcinomas carcinomas carcinomas lymphomas All tumors'
4 9 NR 15.8(3/19)
Filtered — 0
Clean air — 0C
5 3 (1/19) 10.5 (2/19) 5 3 (1/19)
0 0 0
I) 0 0
5.3 (1/19) 25 (6/24) 42.1 (8/l9)b
0 37.5 (9/24) 0
0 8 3 (2/24) (I
'Lung tumors only.
''Significant relative to controls at p <0.01 (with Yates correction); 26.3 percent (5/19) malignancies, significant relative to control at p <0.05 (with
Yates correction).
cOne microscopic adenoma found in 30-mo control rat during the last 6-mo observation period.
Source: Adapted from Iwui el al., 1986.
-------
occurrence of splenic, malignant lymphomas was attributed to exposure to filtered and
unfiltered diesel exhaust. It was not possible to ascertain if the splenic malignancies were
due to the particulate phase of the diesel exhaust or to adsorbed organic compounds. It is
possible that the gas phase of the exhaust was responsible for these lymphomas because they
were detected in filtered exhaust. Other long-term studies apparently did not examine
splenic tissue or did not detect such lesions.
A chronic (to 28 mo) inhalation exposure study by Takemoto et al. (1986) was
conducted to determine the effects of diesel exhaust, di-isopropanol-nitrosamine (DIPN), and
diesel exhaust following DIPN treatment on female F344 rats. Male and female C57BI/6N
mice and ICR mice were also used to determine the effects of diesel exhaust but were not
treated with DIPN. DIPN was administered i.p. at 1 mg/kg weekly for 3 weeks to two
groups of rats. A control group received neither treatment. Treatment protocol consisted
of exposure to diesel exhaust for 4 h/d, 4 d/week. The diesel exhaust was generated by a
269-cc displacement engine operated at an idle state (1600 rpm). The exhaust was directed
into a 104-L stainless steel storage chamber and diluted 1: to 1:4 in a 92.8-L PVC mixing
chamber. The diluted exhaust then entered the inhalation chamber (376 L) at a flow rate
of 120 IVmin Pressure in the chamber was maintained at 5 mmAq. Gas-phase components
of the exhaust were 40 to 100 ppm CO, 3 to 6 ppm NO, 2 to 4 ppm NO,, and <0.5 ppm
SOj. Diesel exhaust particle concentration in the exposure chamber was 2 to 4 mg/m3. The
mean diameter of the panicles was 0.32 ^m. B[a]P and 1-nitropyrene concentrations were
0.85 and 93 #ig/g panicle, respectively.
After 1 week of exposure to diesel exhaust, deposition of black panicles in alveolar
spaces and bronchioles of the exposed animals was noted. At 1 mo of exposure the particle
deposition extended to the alveolar walls and the peritracheal lymph nodes, and by 3 mo
deposition beneath the pleura was observed. Further exposure resulted in greater particle
deposition, but no quantitative data were provided. Rats exhibited a uniform particle
deposition in all lobes of the lung, and mice exhibited regional accumulations of panicles.
Pneumonia, bronchitis, alveolar hyperplasia, and pituitary tumors were observed in rats of
all treatment groups, but not in the controls, however, no quantitative data were provided
relative to these findings.
May 1990 7-16 DRAFT-DO NOT QUOTE OR CITE
-------
No significant evidence of carcinoma formation was noted in any of the rats exposed
to only diesel exhaust. One adenocarcinoma was observed in a control ICR mouse at 24 mo
and a slight, but insignificant increase in adenocarcinomas (2/11 males, 1/11 females) was
noted in diesel exhaust-exposed mice at 18 to 24 mo. For C57B1 mice, adenomas were
observed in 2/40 males and 2/39 females at 13 to 18 mo and in 4/33 males and 4/38 females
at 19 to 24 mo. Adenocarcinomas were noted for 2/33 males and 1/38 females at 19 to 24
mo. This was in contrast to one adenocarcinoma in a male control mouse. A slight but
insignificant increase in tumor incidence was observed for both strains of mice exposed to
diesel exhaust. The authors suggested, however, that mouse strains known for low
spontaneous lung tumor occurrence may be less sensitive and thus be less likely to respond
to tumorigenic agents. Tumor incidence data for the various treatment protocols are
presented in Table 7-9. It was also noted that the small diesel engine used in this study had
different operating characteristics than that of a diesei-powered automobile. A long-term
inhalation study using female rats was conducted by Mohr et al. (1986) in which the effects
of inhalation exposure to unfiltered diesel exhaust and PAH-enriched coal oven flue gas
were compared. The diesel exhaust generating and exposure systems were those used by
Heinrich et al., 1986a; 1986b). Briefly, exhaust was produced by a 40-kW, 1.6-L diesel
engine operated continuously according to the U.S. 72 (FTP) test driving cycle and using
European Reference Fuel with a 0.36 percent sulfur content. The exhaust was diluted 1:17
and delivered to the exposure chambers at a panicle concentration of 3.9 mg/m3. The mass
median diameter of the diesel exhaust particles was 0.35±0.10 /im. The dilution and filtering
system are described fully in the Heinrich et al., 1986a study descriptions. An analysis of the
gas-phase components of the diesel exhaust was not provided but assumed to be similar or
identical to that reported by Heinrich et al. (1986a,b). Groups of 96 female Wistar rats were
exposed to unfiltered diesel exhaust, filtered diesel exhaust, or clean air for 18 h/d. 5 d/week
for 2.5 yr.
Substantial alveolar deposition of carbonaceous panicles was noted for rats exposed
to the unfiltered diesel exhaust. Squamous metaplasia was observed in 65.3 percent of the
rats breathing unfiltered diesel exhaust but not in any of the control rats. Bronchio-alveolar
neoplasms and squamous cell tumors were noted in 8 of 15 and 9 of 15 diesel
May 1990 7-17 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-9. TUMOR INCIDENCE IN ICR AND C57B6 MICE FOLLOWING LONG-TERM
(24 MO) INHALATION EXPOSURE TO DIESEL EXHAUST AND IN RATS
EXPOSED TO DIESEL EXHAUST WITH OR WITHOUT DIPN TREATMENT
Percentage of animals with tumors
Exposure group
Ad
Ad-Ca
Ca
Ad
Ad-Ca
Ca
Ad
Ad-Ca
Ca
F 344 RaU
Diesel
(2-4 mg/m3)
DIPN
Diesel + DIPN
Control
Percent
0 (0/4)
0(0/3)
0(0/5)
0(0/3)
6 mo
Percent
0 (0/4)
0(0/3)
0 (0/5)
0(0/3)
Percent
0(0/7)
25(2/8)
67 (12/18)
0(0/3)
12 to 17 mo
Percent
0(0/7)
0(0/8)
17 (3/18)
0(0/3)
Percent
0 (0/15)
48(10/21)
67(12/18)
0(0/3)
18 to 24 mo
Percent
0 (OAS)
19 (4/21)
39 (7/18)
0(0/3)
ICR Mice
Diesel ,
(2-4 aig/m*)
Control
7 -12 mo
13-18 mo
19-28 mo
M
F
M
F
0(0/29)
0(0/20)
0 (0/19)
0 (0/17)
0(0/29)
0(0/20)
0 (0/19)
0 (0/12)
13 (2/15)
11 (2/19)
II (2/19)
5 (1/19)
7(1/151
0 (0/19)
0 (0/19)
0 (0/19)
27 (3/11)
27 (3/11)
10 (1/10)
17 (2/12)
18(2/11)
9(1/11)
0(0/10)
C57B6
Diesel
(2-4 mg/m})
Control
M
F
M
F
0(0/28)
0 (0/10)
0(0/14)
0 (0/14)
0(0/28)
0 (0/10)
0 (0/4)
0 (0/4)
5(2/40)
5 (2/39)
0(0/7)
0(0/12)
3 (1/40)
3(1/39)
0(0/7)
0(0/12)
12 (4/33)
11 (4/38)
6(1/17)
0 (0/15)
6 (2/33)
3(1/38)
0(0/1
0
Ad: Adenoma.
Ad-Ca: Adenocarcinoma.
Ca: Carcinoma.
DIPN. di-Uopropanol nitrosamine.
Source
adapted from Takemoto et «!., 1986.
-------
exhaust-treated rats, respectively. One of the nine squamous cell tumors was characterized
as a Grade I carcinoma, while the remaining eight were classified as benign, keratinizing,
cystic tumors. Although both the diesel exhaust-exposed and the coal oven flue gas-exposed
rats exhibited a greater incidence of neoplastic and non-neoplastic lesions compared to the
control animals (Tables 7-5 and 7-10), statistical analysis was not provided for these data.
Rats exposed to the diesel exhaust had a greater incidence of non-neoplastic lesions than
did the flue gas-exposed rats with both exhibiting greater incidence than control rats, but
again statistical evaluation of the datawas not provided.
For comparison, rats were also exposed to coal oven flue gas mixed with pitch fumes
that had been pyrolized under nitrogen. Of 120 rats exposed to PAH-enriched coal oven
flue gas, there were 16 benign, keratinizing, cystic tumors and 2 Grade-II carcinomas. The
concomitance of severe inflammatory changes such as hyperplasia and metaplasia
development of squamous tumors in the lungs of diesel exhaust-exposed rats, and the
absence of such a relationship in rats exposed to PAH-enriched coal oven flue gas, provided
the basis for the authors' contention that diesel exhaust may have more of a promotional
effect than a complete carcinogen effect. However, diesel fumes may have direct-acting
carcinogenic properties in addition to those inducing lung pathology so this study does not
offer proof of Heinrichs theory. This study, apparently a phase of the research reported by
Heinrich et al. (1986a,b), provides more detailed histological interpretation of the lesions
produced following inhalation exposure to diesel exhaust and PAH-enriched coal oven flue
gas. Additionally, it provides evidence for induction of neoplastic lesions following long-term
inhalation exposure to unflltered diesel exhaust, but does not provide information linking the
tumorigenic effects to either the particulates or adsorbed organic compounds.
Brightwell et al. (1986) studied the effects of gasoline engine exhaust and of filtered
and unfiltered diesel exhaust on male and female Syrian golden hamsters and F344 rats.
The hamsters of each treatment group were also pretreated with a single administration of
4.5 mg diethylnitrosamine (DEN) 3 d prior to exhaust exposure. The diesel exhaust was
generated by a 1.5-L (manufacturer not provided) engine, computer-operated according to
the U.S. 72 (FTP) driving cycle, and the gasoline engine exhaust was produced by a 1.6-L
engine. The engine emissions were diluted by conditioned air delivered at 800 m3/h to
May 1990 7-19 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-10. PULMONARY LESION INCIDENCE IN FEMALE WISTAR RATS FOLLOWING
LONG-TERM (24 to 30 MO) INHALATION EXPOSURE TO UNFILTERED DIESEL EXHAUST
Treatment
Percentages of animals with non-ncoplast lesions
Diesel (3.9 mg/m3)
exhaust
Bronchio-a|vcol;ir
hyperplasia
98 9 (94/95)
43.1 (50/116)
1.0(1/96)
Sental thickening
97.9 (93/95)
26.7 (31/116)
8 4 (8AJ6)
Squamous metaplasia
65.3 (62/95)
30.2 (35/116)
0 (0/96)
Source: Adapted from Molir ci al, 1986.
-------
produce the high exposure (6.6 rag/m3, diesel exhaust) atmosphere. Further dilutions of 1:3
and 1:9 produced the medium (2.2 mg/m3, diesel) and low (0.7 mg/m3, diesel) exposure
atmospheres. Filtered diesel exhaust was generated by a similar engine equipped with a
particle filter. The CO and NOx concentrations were 32±11 and 8+1 for the unfiltered diesel
exhaust (high exposure concentration chamber) and 32±11 and 8±1 for the filtered diesel
exhaust. CO and NOx concentrations for the gasoline engine exhaust were 224±32 and 49±5
ppm, respectively. Passage of this exhaust through a catalytic converter reduced these values
to 21+10 and 7±3 ppm, respectively. The inhalation exposures were conducted overnight
with five 16-h periods per week for two yr.
The body weights of rats were significantly lower (p value and statistical test not
indicated although t-tests were used for analysis of other parameters) in both medium and
high diesel exhaust exposure groups. These weight changes were evident shortly after the
beginning of exposure and became more significant with increasing exposure duration. The
raw data were not presented for these findings, nor were there any comments regarding
body weights of the hamsters. None of the exposure treatments resulted in significant
changes in urinalysis parameters. Blood chemistry analyses for rats indicated significant (p
values ranging from 0.05 to 0.001, t-test) elevations of blood urea nitrogen, alkaline
phosphatase, and aspanate aminotransferase in male rats and significant increases in blood
urea nitrogen and alkaline phosphatase in female rats. Significant decreases in glucose, total
proteins, triglycerides, and cholesterol were reported for female rats exposed to diesel
exhaust for 18 to 24 mo. For male rats, decreases were noted for glucose, cholesterol, and
cholinesterase. These blood chemistry values were for 18- and 24-mo high exposure animals
only. Blood chemistry values for hamsters following 16 mo exposure to the high concentra-
tion diesel exhaust indicated significant elevations in albumin and gamma-glutamyl
transpeptidase in female hamsters, but only gamma-glutamyl transpeptidase was elevated in
males. Significant decreases in potassium, lactate dehydrogenase, alpha-hydroxybutyric
dehydrogenase, and aspartate aminotransferase were detected in female but not in male
hamsters. The high exposure concentration of both diesel engine and gasoline engine
exhaust elevated red blood cell counts, hemoglobin content, and hematocrit values in the
rats. However, the diesel exhaust also caused an increase in white blood cells and
May 1990
7-21
DRAFT-DO NOT QUOTE OR CITE
-------
segmented neutrophils, and a decrease in lymphocytes in the rats. Only hemoglobin and
hematocrit values were significantly increased in hamsters (males only) exposed for 16 mo
to the high concentration of diesel exhaust.
A statistically significant increase in lung weight was noted for rats and hamsters
exposed to diesel exhaust, but was more pronounced and progressive with exposure level and
duration in the rats. The greatest increase in lung weight was observed in high exposure
male rats at 30 mo (includes a 6-mo observation period) and the high exposure female rats
at 24 mo at which times the lung weights were, respectively, 2.8-fold and 3.9-fold greater
than the controls. Exposure of rats to the high level of diesel exhaust resulted in statistically
significant (p values not provided), dose-related obstructive and restrictive airway effects.
No further details of these pulmonary effects were provided in the cited report.
The cardiovascular system was also affected by the exhaust exposures. The heart-to-
body weight ratio was significantly (p <0.001) increased in rats exposed to gasoline engine
exhaust. Left ventricular contractility was significantly (p <0.05) lower, and right
ventricular-heart weight ratio was significantly (p <0.001) greater in diesel exhaust - exposed
rats when compared to controls. Right ventricular weight-body weight ratio and total heart
weight-body weight ratio were significantly (p <0.001 and 0.01, respectively) greater in rats
exposed to diesel exhaust than in controls.
Tumor incidence data for this study are presented in Table 7-11. A 4-percent (3/72)
and 23-percent (16/71) incidence of primary lung tumors occurred in male F344 rats
following exposure to 2.2 and 6.6 mg diesel soot/m3, respectively. The tumor incidence in
female rats was 15 percent (11/72) for the 2.2-mg/m3 exposure group and 54 percent (39/72)
for the 6.6-mg/m3 exposure group. No significant increase was reported for male (1 percent,
1/72) or female (0 percent, 0/72) rats in the low exposure (0.7 mg/m3) group compared with
the control animals (male: 2 percent. 3/140: female: 1 percent. 1/142). Diesel exhaust-
induced tumor incidence in rats was dose-dependent and higher in females than males.
Similar to other studies, tumor incidence in rats was related to exposure to unfiltered as
opposed to filtered exhaust. It must be noted, however, that the exposure during darkness
(when increased activity would result in greater lespiratory exchange and greater dose) could
account, in part, for the high response reported for the rats. A detailed histological analysis
May 1990 7-22 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-11. TUMOR INCIDENCE IN MAIJi AND FEMALE F344 RATS AND SYRIAN GOLDEN HAMSTERS
FOLLOWING LONG-TERM INHALATION EXPOSURE TO FILTERED OR UNFILTERED DIESEL EXHAUST
WITH OR WITHOUT DEN' PRETREATMENT
Percentages of animals with lumorsb
23 (16/72)
54 (39/72)
4 (3/72)
15 (11/72)
1 11/72)
0 (0/72)
2 (3/-I0)
1 (1/42)
Respiratory tract tumors rare in non-DEN
prctrcatcd hamsters and were not related
to exhaust exposure (t-lcst). Increased
incidence or tracheal papillomas in all
DEN-nrctrcaicd groups, but incidence was
not related to exhaust exposures.
Treatment
Lung soot burden
Rats
6.6 mg mJ M
F
2.2 mg/mJ M
F
0.7 nig/m' M
F
Control
M
F
Hamsters
6 6 mg/m1
6 6 mg/m1
2.2 mg/mJ
2.2 mg/m3 DEN
0.7 mg/mJ
0.7 mg/mJ DEN
Control
Control + DEN
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
'DEN: dimclhylmirosaminc 4.5 mg/kg, sc., 3 d prior to exposure
'Primary lung tumors; specific liistopathology not reported
NR: not reported
Source: Ad.ipied lioni Uiiglilwcll el al, l'J,S6
-------
of tumor types was not provided. Although DEN-pretreated hamsters exhibited an increase
in tracheal papillomas in all treatment and control groups when compared to non*DEN
pretreated hamsters, there was no statistically significant (t-test) relationship between tumor
incidence and exhaust exposure.
Karagianes et al. (1981) exposed male Wistar rats (40 per group) to diesel engine
exhaust (8.3±2.0 mg soot/m3), room air, diesel engine exhaust plus low-concentration coal
dust, low-concentration coal dust only, or high-concentration coal dust 6 h/d, 5 d/week for
up to 20 mo. The exhaust source was a 3-cylinder, 43-hp diesel engine using ASTM D-975
grade No. 2-D diesel fuel oil and Federal Specification W-F-800 A grade DF-2 diesel fuel
oil. This type of engine is normally used in mining situations and, as such, was connected
to an electnc generator and operated at varying loads and speeds to simulate operating
conditions in an occupational situation. Controlling the CO concentration at 50 ppm, the
exhaust was diluted 35:1 with compressed air. The exposure chambers were 3000-L
plexiglass spheres of with diameters of 178 cm. The exposure chamber temperatures were
25 to 26.6 * C, relative humidity was 45 to 80 percent, and total flow through the chambers
was 50 L/min. The mass median diameter of the diesel exhaust particles was 0.71 fim. The
N02 concentration ranged from 4 to 6 ppm, the CO level was maintained at 50±3 ppm, and
the SO, and aliphatic aldehyde levels remained below detection limits (1 ppm).
Exposure to diesel exhaust or coal dust alone, or in combination, failed to cause
significant alteration of body weight or mortality in the rats. Blood samples of rats sacrificed
at 4, 8, or 16 mo of exposure revealed no significant changes in hematocrit, erythrocyte, or
differential leucocyte counts or in blood protein profiles. At 4, 8, 16 and 20 mo, a 4- to 5-
percent increase in carboxyhemoglobin levels was noted in rats exposed to diesel exhaust or
diesel exhaust plus coal dust. Histological examination of rats that died during the study
and for those sacrificed at 4, 8, 16. or 20 mo of exposure revealed interstitial fibrosis,
alveolar histiocytosis, and alveolar emphysema, the severity of which was related to exposure
duration. There was no statistically significant (p <0.05) difference in severity of lesions
among the various exposure groups. One bronchiolar adenoma each was found in the diesel
exhaust and diesel exhaust plus coal dust exposure groups; however, the occurrence was not
statistically significant. Particle accumulation occurred at the terminal bronchioles,
May 1990 7-24 DRAFT-DO NOT QUOTE OR CITE
-------
respiratory bronchioles, alveolar ducts, and within alveolar sacs. Alveolar macrophages were
heavily burdened with panicles. No malignant respiratory tract tumors were found in any
of the rats in any of the exposure groups, but soot samples taken from the diesel exhaust
chamber and other system locations were mutagenic in the Ames assay (no specific data
presented).
Male cynomolgus monkeys (15) and male (151) and female (72) F344 rats were
exposed to diesel exhaust (2 mg/m3) for 7 h/d, 5 d/week for 24 mo (Lewis et al., 1986). The
same number of animals was also exposed to coal dust (2 mg/m3), diesel exhaust plus coal
dust (1 mg/m3 for each component), or filtered air. The diesel exhaust was produced by a
7.0-L, four-cycle, water-cooled Caterpillar Model 3304 engine using No. 2 diesel fuel (<0.5
percent sulfur by mass). The exhaust was passed through a Wagner water scrubber, which
lowered the exhaust temperature and quenched engine backfire. The engine was connected
to an Eaton-Dynamic Model 758-DG dynamometer and operated by a KIM-1 computer
simulating the activity of a diesel-powered coal mining tram. The duty cycle included eight
modes, of which idling constituted approximately 60 percent. The exhaust was diluted 27:1
prior to entering the exposure chambers. An analysis of the exposure atmospheres is
presented in Table 7-12.
None of the monkeys exposed to diesel exhaust exhibited a significantly increased
incidence of neoplastic lesions. It should be noted, however, that the 24-mo time frame
employed in this study may not allow for prediction of tumors in primates. Tumors and
premalignant conditions were not observed in the upper respiratory tract of treated rats, nor
did the incidence of neoplasms in 50 tissues differ significantly from the control rats. Focal
deposition of diesel exhaust particles in the lungs of both species was reported. Alveolar
type II cell hyperplasia and lipidosis were demonstrated in the rats exposed to diesel exhaust.
Mixed function oxidase activity of the lungs and liver was not altered by the exposure to
diesel exhaust. Pulmonary maximum expiratory flow rate was significantly reduced in
monkeys exposed to diesel exhaust for 12, 18, and 24 mo. Oven signs of toxicity were not
demonstrated by body weight changes, organ/body weight ratios, or clinical chemistry indices.
Pepelko and Peirano (1983) summarized an extensive series of studies on the health
effects of diesel emissions. The exposure system consisted of twenty-four 2.8-m3 stainless
May 1990 7-25 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-12. ANALYSIS OF FRESH AIR CONTROL (FA), DIESEL EXHAUST (DE), COAL
DUST (CD), AND COAL DUST PLUS DIESEL EXHAUST (DECD) ATMOSPHERES
Component Unit FA CD DE DIZCD
Total particles
mg/mJ
a
4 981() 82b
a
3.23*0.60
Rcspirablc particles
mg/mJ
a
2 OOiO 41
1 95*0.25
2.02*0.30
CO,
percent
(• 08 *0.02
0.09*005
(12010.06
0 20*0 07
CO
PPm
2 2i09
2 2 ill9
1 5i3 1
109*28
NO
ppnt
(IO81O 14
OOSiO 29
8.7i3 6
8 3*3.2
NO,
ppm
0 06i004
00Oi0l)5
1.51.0.5
1.6*0.5
THC (cold)c
ppm C
-I I ± 19
a
7.512 2
7.4*2.0
THC (hot)
ppm C
a
;i
a
8 3*1.9
SO,
ppm
a
001*007
081*0.38
06110.29
NHj
ppm
(t 521() 28
0 57*0 52
0 6410.71
0 48*0.55
Total aliphatic aldehydes
ppm
0 02*001
002i()0l
012*006
O.I2iU.05
SO,2
ppb
a
168117 9
29.0*24.9
42.3i33.8
nH,4
pph
a
6Sil43
27.0*30.7
16 5i233
Acrolein
ppb
1 <1*3 3
6 2i4 7
60 2124.5
57 8*205
Acctaldehydc
ppb
1 5t3 5
0.912 5
38.7*15.3
37.7114 0
-------
TABLE 7-12. (Continued)
Component
Unit
FA
CD
DE
DECD
Formaldehyde
Mfi/m1
7 6 ±3.5
7.4t4 1
38.3 ±23.0
37.4 ±26 6
Silica
|ig/mJ
ND*
21 9±I0 4
ND
ND
Silicon
lig/m1
ND
22. It 10 3
ND
ND
Bcnzo(a)anthraccnc
Mg/mJ
a
3.2±2 2
196±9.9
11.2 ±5.2
Bcnzo(k)fluuranilicne
Mg/mJ
a
j
5 6±2.3
3 6±2 4
Bcnzo(a)pyrene
Mfi/m'
a
a
13 5 ±6 8
10 2±6 5
Flouranthene
Mg/mJ
a
26 51115
139.3 ±98 1
67.5 ±52 4
Pyrene
|ig/mJ
a
32 3 ± 15 1
123 4±72.2
60.0 ±36 6
'Not measured.
bMean±SD.
"Total hydrocarbons.
ND = not delected; limits of detection for silica and silicon were 12 |ig/m} and II Mg/m1, respectively.
Source: Adapted from Lewis cl al., 1986.
-------
steel exposure chambers. Exhaust was provided by two Nissan CN 6-33, 6-cylinder, 3.24-L
diesel engines coupled to a Chrysler A-272 automatic transmission and Eaton model 758-DG
dynamometer. The exhaust was mixed with filtered, conditioned air in a dilution tube and
further mixed in a large-volume mixing chamber. The flow rate through the exposure
chamber allowed for 15 changes per hour. Sixty-day pilot studies were conducted where a
near-maximum tolerated exposure of diesel exhaust (1:14 dilution) provided 6-mg/m3 particle
concentrations. The engines were operated using the Modified California Cycle. These 20-
h/d, 7-d/week exposures using rats, cats, guinea pigs and mice produced decreases in weight
gain and food consumption. Therefore, the long-term studies reduced the exposure time to
8 h/d at an exhaust panicle concentration of 6 mg/m3 followed by another 12-mo period of
exposure using a panicle concentration of 12 mg/m3. For these studies the engines were
operated using the Federal Shon Cycle. Analysis of the exhaust components are
summarized in Table 7-13.
Some shon-term studies (Onhoefer et al., 1981) also employed exposure of mice to
irradiated exhaust (to simulate sunlight exposure). Groups of 25 male Strong A mice were
exposed to irradiated or nonirradiated diesel exhaust for 20 h/d, 7 d/week for 7 weeks. The
exhaust was produced by a 198-in3 Nissan automotive diesel coupled to a Chrysler Torque-
Flite transmission and absorption dynamometer. No. 2 diesel fuel was used by the engines.
The engine was operated continuously for 20 h daily for the irradiated exhaust exposure and
for 8 h daily for the nonirradiated exhaust exposures. Exhaust was diluted 1:13 with filtered
air. The diluted exhaust then passed through a large-volume mixing chamber and into the
exposure chambers; some of the exhaust from the mixing chamber was directed into a
dynamic flow irradiation chamber to simulate sunlight.
The diesel exhaust exposures did not produce any significant signs of gross toxicity
or affect the growth rates of the mice. Exposures to irradiated exhaust, nonirradiated
exhaust, or control atmospheres were not significantly different relative to lung tumor
incidence. The authors considered the results of the diesel exhaust exposure to be negative,
although the following tumor incidence data were provided: control female mice (4/58);
historical controls male mice (iC/60); urethane-treated mice (9/52); diesel exposed (14/56);
diesel plus urethane (22/59).
May 1990 7-28 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-13. CONCENTRATION OF GAS-PI IASE COMPONENTS AND PARTICLES IN EXPOSURE CHAMBERS
Clean air
c hambcrs
Exhaust chambers
Component
Units
Weeks 1-124
Weeks 1-61
Weeks 62-124
CO,
percent
004i0()02*
0 30 ±004
0 52i()04
CO
PPm
2.20i0.50
211 17 ±3.01
33.30 ±2.94
Total Hydrocarbons (cold)
ppm
2.82 tO 51)
7 93tl <12
ll.02i 1 (M
NO
ppm
005 ±004
II <>4 ±2 3-4
19 49±3 80
NO,
ppm
003±0 03
2 <>8 ±0 80
4 3711.19
SO,
ppm
0.03 ±0.02
2. !2i0 58
5 0311.03
Dilution factor
DF
—
18 I6il 72
9.3711.13
Particulate mass
mg/mJ
0.00
6 34±081
Il.70i0.99
Acrolein
ppm
—
0 025 ±0.003*
0 034 i0(X)9
Formaldehyde
ppm
—
0.106 ±0.029
0.25 Ii0 059
Total aliphatic aldehydes
ppm
—
0 177i0.043
0.338 tO 057
'Standard deviation of weekly means.
Source: Adapted from I'cpclko and Pcirano, 1983
-------
Small and nonsignificant increases in tumor incidence were found for male and
female Strong A and Jackson A mice following 8-week inhalation exposure to whole exhaust
or irradiated whole exhaust at a panicle concentration of 6 mg/m3 (Table 7-14) followed by
holding in clean air until sacrifice at about 9 mo of age (Orthoefer et al., 1981 as reported
in Pepelko and Peirano, 1983). Male and female Strain A mice were exposed to diesel
exhaust (6 mg/m1) from 8 weeks to 9 mo of age. Sixty animals each from the exposure and
control groups also received a single i.p. dose of 1 mg urethane, a carcinogen initiator.
Relative to respective clean air controls, females exposed to diesel exhaust or diesel exhaust
and urethane pretreatment showed a slight but significant (p <0.01 for exhaust alone; p
<0.02 for exhaust plus urethan pretreatment) increase in tumor incidence (Table 7-15).
However, if these results are combined with the data for the much larger groups of male
mice, which showed no increase in tumor incidence following any of the treatments, the
controls and treatment groups are not significantly different.
Furthermore, it was noted that the tumor incidence (0.09 tumors per mouse) in the
control group was less than the expected value of 0.25 tumors per mouse. Personal
communication with the author indicated that in the case of the Strong A mice (1-mg
urethane study), genetic drift had occurred in the main colony and the background tumor
incidence and presumably sensitivity had decreased, therefore suggesting that the increased
incidence in the female mice may, in fact, have been real.
Pepelko and Peirano (1983) described a two-generation study using Senear mice
exposed to diesel exhaust with and without pretreatment with tumor initiators. Exposure
conditions were as described previously for Pepelko and Peirano (1983). Male and female
mice were exposed to a diesel exhaust particle concentration of 6 mg/m3 prior to and
throughout mating. The dams continued exposure through gestation, birth, and weaning.
Groups of offspring (130 males and 130 females) received an i.p. injection of butylated
hydroxytoluene (BHT) (300 mg/kg for week 1, 83 mg/kg for week 2. and 150 mg/kg from
week 3 to 1 year), a single i.p injection of 1 mg urethan, or no injections of tumor initiators.
The exhaust exposure was increased to a particle concentration of 12 mg/m3 when the off-
spring were 12 weeks of age and was maintained until termination of the experiment when
the mice were 15 mo old. The incidence of pulmonary adenomas was significantly (p <0.02)
May 1990 7-30 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-14. LUNG TUMOR INCIDENCE IN STRAIN A MICE
EXPOSED FOR 8 WEEKS TO RAW OR IRRADIATED EXHAUST
AT A PARTICLE CONCENTRATION OF 6 mg/m3
Survivors/ Mice With Tumors/
Treatment Sex iniual tumors mouse
STRONG A MICE
Clean air
M
22/25
3'
0.13
Raw exhaust
M
19/25
7
0.63
Irradiated
exhaust
M
22/25
JACKSON A MICE
6
0.27
Clean air
M
18/20
5
033
Clean air
F
18/20
11
0.66
Gean air
M & F
36/40
16
0.50
Raw exhaust
M
16/20
5
031
Raw exhaust
F
18/20
6
0 50
Raw exhaust
M & F
34/40
11
0.41
The percentage of mice with tumors was compared using X2 testing. No significant differences were detected
among groups.
Source: Adapted from Pepelko and Pierano, 1983 tumor initiators.
May 1990
7-31
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-15. LUNG TUMOR INCIDENCE IN STRAIN A MICE EXPOSED TO
DIESEL EXHAUST AT A PARTICLE CONCENTRATION OF 6 MG/M3
UNTIL 9 MO OF AGE
Treatment
Sex
Survivors/
initial
Mice with
tumors
P'
Tumors/
mouse
Clean air
F
58/60
4
0.09
Exhaust
F
S6/60
14
<0.01
032
<0.01
Clean air +
1 mg urethan
F
52/60
9
0.25
Exhaust +
1 mg urethan
F
59/60
22
<0.02
0.39
<0.01
Clean air
M
403/429
73
0.23
Exhaust .
M
368/430
66
0.20
•X1
Source: Adapted from Pepelko and Pcirano. 1983
May 1990
7-32
DRAFT-DO NOT QUOTE OR CITE
-------
increased in the non-injected female mice exposed to diesel exhaust, but the exhaust
exposure reduced (p <0.01) the tumor incidence in female mice receiving BHT. Males were
unaffected by either condition. All pulmonaiy tumors (adenomas and carcinomas) were
significantly increased in female mice {p <0.05) and male and females combined (p <0.05).
These results provided the earliest evidence for cancer induction following inhalation
exposure to diesel exhaust. The lack of a clearly defined response may well have been
influenced by the relatively early sacrifice times of the mice. These data are summarized
in Table 7-16.
Additional studies in the Pepelko and Peirano report examined the effects on Strain
A mice of diesel exhaust (particle concentration of 12 mg/m3) following pretreatment with
5 mg urethan, the effects of exposure during night vs. daylight, and the effects of increasing
exposure (12 mg particles/m3) to 12 mo of age rather than 9 mo. These results are
presented in Table 7-17. Tumor incidence was significantly reduced in the diesei
exhaust/urethan treatment groups compared to animals receiving 5 mg urethan only.
Exposure to diesel exhaust during the dark ponion of the photoperiod or until 12 mo of age
also reduced the tumor incidence when compared to control groups. The reasons for a
decreased response to diesel exhaust exposure are unknown, although these greater exposure
levels may have induced a greater degree of cell death in the cancerous than in normal cells.
General Motors Research Laboratories performed chronic inhalation studies using
male F344 rats and Hartley guinea pigs exposed to diesel exhaust panicle concentrations of
0.25,0.75, or 1.5 mg/m1 (Schreck et al., 1981). The exposure protocol was 20 h/d, 5.5 d/week
for 24 mo. Exhaust was generated by a 5.7-L Oldsmobile engine (four different engines used
throughout the experiment) operated at a steady speed and load simulating a 40 mph driving
speed of a full-size passenger car. AMOCO Type 2D federal compliance fuel with a 0.27
percent sulfur content and AMOCO 200 30W oil were used. The mass median aerodynamic
diameter of the exhaust particles was 0.2 nm with 90 percent of the mass associated with
particles of less than 1.0 fim diam. Exhaust was directed into stainless steel lines and heated
to 100±15*C to simulate typical tail pipe temperatures. It was diluted with clean air and
then entered the large volume (12.6 m3) exposure chambers, which contained two cage racks
May 1990
7-33
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-16. EFFECTS OF INHALATION EXPOSURE TO DIESEL EXHAUST
ON LUNG TUMOR INCIDENCE IN MALE AND FEMALE SENCAR MICE
Treatment
group
Sex
Percent
adenomas
Percent
carcinomas
Percent all
lung tumors
CA
DE
CA
DE
CA
DE
Uninjected
M
3.8
4.0
0.0
2.0
3.8
5.9
F
63
16J*
0.9
0.0
7.2
16J"
M & F
5.1
10.2"
0.5
1.0
5.6
112'
BHT
M
8.5
8.8
0.0
2.9
8J
lli
F
16.7
3.9
1.5
2.6
18.1
6J"
M & F
12.2
5.4
1.7
2.7
12.8
8.1
Urethan
M
9.2
9.0
0.0
1.1
9.2
10.1
F
7.0
8.4
1.8
3.7
8.7
12.1
M & F
8.1
8.7
0.9
2.6
9.0
11.2
All treatments
M
7.1
6.7
0.0
1.7*
7.1
8.0
F
8.9
10.1
1.4
2.1
103
122
M & F
8.0
8.6
0.7
2.0
8.7
102
*p = 0.02
"pi 0.05
CA = Clean air.
DE - Diesel exhaust
3HT = Butylated hydroxyioluene.
Source: Adapted from Pepelko and Pcirano. 1983.
May 1990
7-34
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-17. LUNG TUMOR INCIDENCE IN STRAIN A MICE
EXPOSED TO A PARTICLE CONCENTRATION OF 12 mg/m5
Treatment
Age
(mo)
Illumination
Sex
Survivors/
initial
Muc with
luinurs
Tumors/mouse
Clean .nr
light
M
I
M X F
44/45
43/45
87/90
in
II
21
0 23
0.15
(129
1-jkhnusl 9 light
Clc.in air + 9 light
5 mg urcthan
Exhaust + 9 light
5 mg urethan
Clean air 9 dark
Exhaust 9 dark
Clean air 12 light
lixhaust 12 light
•X;
''Student's t-icst.
Source' Ad.iplcd Irom I'cpclko ;nul I'eininn, 1981
M .17/45 "»
F 43/45 4
M \ F 80/90 III
M 38/45 12
F 37/45 .14
M & F 75/90 M
M 39/45 20
F 36/45 16
M A F 75/90 42
M 97/108 28
F 140/142 31
M .V F 237/250 59
M Ill/IIS 13
I' 139/143 9
M A F 250/258 22
M 38/ 22
M 44/ 11
0 19
<0 05 0 09 <0 05
<0 05 014 <0 10
2 17
3 24
2 80
<010 103 <0001
<0001 0 86 <00001
<0001 095 <00001
0 32
0 23
0 27
<001 0 14 <001
<0001 0 07 <0001
<0001 0 10 <0 0001
0 68 <0001
<001 0 25
-------
with up to 144 rats or 72 guinea pigs. Exposure chambers were maintained at 22±2*C at
50±20 percent relative humidity. Control animals received clean air.
Animals exposed to the diesel exhaust had discolored fur throughout the exposure
period. Rats of the high exposure (1.5 mg/m3) group exhibited body weights that were
significantly less (p <0.05, t-test) than that of controls. One group of guinea pigs exposed
to the low (0.25 mg/m3) exhaust particle concentration had body weights significantly lower
(p <0.05, t-test) than respective controls; however, this effect was not observed in animals
of the higher exposure groups. No exposure-related overt signs of toxicity were reported for
the test animals.
Ultrastructural changes in the lungs of the exposed guinea pigs were reported in
Barnhart et al. (1981). No evidence was found for neoplastic changes, nor were pathological
changes indicative of fibrosis or emphysema noted. After only 2 weeks of exposure and
throughout the exposure period, significant uptake of exhaust particles by alveolar
macrophages, epithelial type 1 cells, and interstitial macrophages was noted.
Takaki et al. (1989) exposed male and female F344 rats to whole exhaust from HD
(11-L) and LD (1.8-L) engines using 13:15 and 10:15 dilution ratios, respectively. The rats
(64 males and 59 females per group) were exposed to particle concentrations of 3.7,1.8,1.0
or 0.5 mg/m3 (HD) or 0.1, 0.4, 1.1, or 2.3 mg/m3(LD) for 16 h/d (1700-0900 h), 6 d/week for
approximately 30 mo. Control groups were exposed to room air. Carbon particle deposition
occurred primarily in the alveolar cavity for low panicle concentration exposures, but was
more prominent in the interstitium and lymphoid tissue at higher particle concentration
exposures. Although the incidence of adenomatous hyperplasia increased with particle
concentration for both the LD and HD groups, a significant difference in tumor incidence
was noted only for the 3.7-mg/m3 HD group (Table 7-18). For the LD groups, tumor
incidence for the control, 0.1-. 0.4-. 1.1-. and 2.3-mg/m3 exposure groups was 3.3.2.4, 0.8.4.1,
and 2.4 percent, respectively. For the HD groups tumor incidence of 0.8, 0.8, 0.0, 3.3. and
6.5 was reported for the control, 0.5-, 1.0-, 1.8-, and 3.7-mg/m3 groups, respectively.
May 1990
7-36
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-18. TUMOR INCIDENCE IN MALE AND FEMALE F344 RATS
CHRONICALLY EXPOSED TO WHOLE EXHAUST FROM I D AND HD ENGINES
Carcinomas
Treatment*
(mg/m})
Hyperplastic
lestons
Adenomas
Adeno
Adcno -
squamous
Squamous
Carcinoma
incidence (%)
Total incidence6
(%)
LD
0
4
1
0
2
1
24
8(6 5)
U.I
4
1
1
0
1
16
7 (5.7)
0.4
6
1
0
0
0
n
7(56)
1.1
12
0
5
0
0
4 1
17 (13 K)
2.3
87
1
1
1
0
1.6
90 (72.6)
HI)
0
1
0
1
0
(1
0.8
2(1.6)
05
3
0
0
0
1
0 8
4(3 3)
1.0
7
1)
0
0
(1
0
7(5 6)
1.8
14
0
3
1
0
3.3
18(14.5)
3.7
25
0
5
1
2
6 5*
33 (26.6)
'n = 123 to 125 ruts per group
bAII tumors plus hyperplastic lesions.
'Significantly diflcrcnl from controls (p not spccilicd)
Source: Adapted from lukaki ci al, 1989.
-------
7.2.2 Short-Term Inhalation and Intratracheal Instillation Studies
Grimmer et aL (1987), using female Osborne Mendel rats (35 per treatment group),
provided evidence that the carcinogenic potential of diesei exhaust resided in the polycyclic
aromatic hydrocarbons (PAH) that consist of four or more rings. Condensate was obtained
from the whole exhaust of a 3.0-L passenger car diesei engine connected to a dynamometer,
the operation of which simulated city traffic driving conditions. This condensate was
separated by liquid-liquid distribution into hydrophilic and hydrophobic fractions representing
25 percent and 75 percent of the total condensate, respectively. The hydrophilic,
hydrophobic, or reconstituted hydrophobic fractions were implanted into the lungs of the
rats. Untreated controls, vehicle (beeswax/trioctanoin) controls, and positive
(benzo[a]pyrene) controls were also included in the protocol (Table 7-19). Results of the
various treatments are presented in Table 7-19. Fraction lib (made up of PAHs with 4 to
7 rings), which accounted for only 0.8 percent of the total weight of diesei exhaust
condensate, produced the highest incidence of lung carcinomas following implantation into
the rat lungs. A carcinoma incidence of 17.1 percent was observed following implantation
of 0.21 mg lib/rat, whereas the nitro-PAH fraction (lid) at 0.18 mg/rat accounted for only
a 2.8 percent carcinoma incidence. Hydrophilic fractions of the diesei exhaust particulate
extracts, vehicle (beeswax/trioctanoin) controls, and untreated controls failed to exhibit
carcinoma formation. Administration of all hydrophobic fractions (Ila-d) produced a
carcinoma incidence (20 percent) similar to the summed incidence of fraction lib
(17.1 percent) and lid (2.8 percent). The benzo[a]pyrene positive controls (0.03, 0.1, 0.3
mg/rat) yielded a carcinoma incidence of 8.6 percent, 31.4 percent, and 77.1 percent, respec-
tively. The study showed that the tumorigenic agents were primarily 4 to 7 ring PAHs and
to a lesser extent nitroaromatics. It did not, however, provide any information regarding the
bioavailability of the panicle-associated PAHs assumed to be responsible for carcinogenicity.
Kunitake et al. (1986) reported an insignificant (2J3 percent) increase in tumor
incidence in female Syrian hamsters (tt = 62) to which a total dose of 7.5 mg of a methylene
chloride extract of diesei exhaust tar (DET) had been instilled intratracheal^ over 15 weeks.
Addition of 7.5 mg B[a]P to a total DET dose of 1.5 mg resulted in a tumor incidence of
May 1990
7-38
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-19. TUMOR INCIDENCE AND SURVIVAL TIME OF RATS TREATED
WITH FRACTIONS FROM DIESEL EXHAUST CONDENSATE (35 RATS/GROUP)
Median survival Carcinoma
Material portion Dose lime in weeks No. No. incidence
by weight (%) (mg) (range) of carcinomas' of adenomas* (%)
Hydrophilic fraciion(l) (25)
Hydrophobic Traction (II) (75)
nonaromalics +
PAC 2 + 3 rings ((la) (72)
PAH 4 to 7 rings (lib) (0.8)
polar PAC (lie) (1.1)
nitro-PAH (lid) (0.7)
Reconstituted hydrophobics
(la), b, c, d) (74.5)
Control, unrelated
Control (becswax/triocianoin)
Bcnzo|a|pyrene
6 7(1
20.01)
19 22
0.21
0.29
0.19
19.91
0.3
01
0.03
97(24-139)
99(50-139)
103(25-140)
102(50-140)
97(44-138)
106(32-135)
93(46*136)
110(23-138)
103(51-136)
69(41-135)
98(22-134)
97(32-135)
0
5
I)
6
0
1
I)
II
27
II
3
I
0
0
(I
0
1
0
0
0
0
14.2
(I
17.1
0
2.8
200
0
771
31.4
8.6
'Squamous cell carcinoma
"Bronchlolar/alveolar adenoma.
Source: Adapted Irom Grimmer el al., 1987.
-------
91.2 percent with 88 percent of these being characterized as malignant. B[a]P (7.5 mg over
15 weeks) alone produced a tumor incidence rate of 88.2 percent (85 percent of these being
malignant), which was not significantly different from the DET + B[a]P group. This
subchronic study demonstrated a lack of detectable interaction between DET and B[a]P, the
failure of DET to induce carcinogenesis, and the propensity for respiratory tract carcinogen-
esis following intratracheal instillation of B[a]P. Intratracheal instillation of 0.03 jxg B[a]P,
the equivalent content in 15 mg DET, failed to cause a significant increase in tumor
incidence in rats. Ishinishi et aL (1988) also reported this study. The effects on hamsters
of intratracheally instilled whole diesel exhaust suspension, diesel particles with Fe,03 as a
carrier dust or diesel particle extract with Fe20 as the carrier were studied by Shefner et
al., 1982. The diesel exhaust component in each of the treatments was administered at
concentrations of 1.25, 2.5, or 1.25 mg/week for 15 weeks to groups of 50 male Syrian
Golden hamsters. The total volume instilled was 0.2 mL. The diesel panicles and the
dichloromethane extracts were suspended in physiological saline with gelatin (0.5 percent
w/v), gum arabic (0.5 percent w/v), and propylene glycol (10 percent by volume). Fe^j
concentration, when used, was 1.25 mg/0.2 mL of suspension. Controls received vehicle and,
where appropriate, carrier particles without the exhaust component. Two replicates of the
experiments were performed.
At the time of the report, only preliminary results were available. A decrease in body
weight gain was noted for those animals in the highest diesel exhaust particle and panicle
extract treatment groups. At the 12-mo sacrifice time (6 mo after the last treatment),
alveolar macrophage phagocytosis of the lung particle burden was not yet complete.
Adenomatous hyperplasia was reponed as being most severe in those animals treated with
the diesel exhaust panicles or the diesel exhaust panicles plus FejOj panicles, and least
severe in those animals receiving the diesel panicle extract suspension. Of the two lung
adenomas detected microscopically, one was in a high-dose diesel particle-treated animal and
the other was in a high-dose diesel extract suspension-treated animal.
Kaplan et aL (1982) conducted a 90-d inhalation exposure study using A/J mice. The
animals were exposed to diluted diesel exhaust (1J±0.14 mg/m3) or filtered air for 20 ii/d,
7 d/week. As described by Schreck et al. (1981), the exhaust was produced by a 5.7-L
May 1990 7-40 DRAFT-DO NOT QUOTE OR CITE
-------
Oldsmobile engine operated continuously (steady-state mode) 20 h/d, 7 d/week at the
equivalent of 40 mph. The engine exhaust passed into a stainless steel line which was heated
to 100±15*C to simulate tail pipe exhaust temperatures of motor vehicles and then rapidly
diluted with clean air near the top of the exposure chambers. The stainless steel and glass
exposure chambers (14.5 m3) contained four cage racks each holding 14 cages. The
temperature in the exposure chambers was 22±2 *C with a relative humidity of 50±10
percent At the end of the exposure period, 30 control and 30 experimental animals were
necropsied. Additional groups of control and experimental animals were used for analysis
of pulmonary ultras tructure, morphometry, and proliferative changes. At 9 mo of age, 500
control and 500 diesel exhaust-exposed mice were used to study the pulmonary adenoma
response. Mice injected with urethane (1 mg/kg, i.p.) at 2 mo of age and sacrificed four
months later served as positive controls for pulmonary adenoma response. Six months after
exposure, 30 control and 30 exposed mice were sacrificed for histopathology analysis.
No change in mortality, growth patterns, or pulmonary adenoma tumor incidence
relative to control groups was observed. Tumor prevalence in the positive control A/J mice
(1.0 mg urethane/kg, i.p.) was 100 percent. For diesel exhaust-exposed mice and controls,
tumor incidence was 34.2 percent and 31.4 percent, respectively (Table 7-20).
The effect of chronic (2 to 2J yr) diesel exhaust exposure on the tumor-inducing
effect of diphenylnitrosamine (DPN) was examined using female Wistar rats (Heinrich et al.,
1989). Groups of rats (45 to 48 per group) were exposed to clean air, diluted (1:17) filtered
diesel exhaust, or whole diesel exhaust (particle concentration of 4.24 mg/m3) and
administered 250 or 500 mg DPN/kg/week during the first 25 weeks of exposure. The total
DPN dose was 6.25 or 12J g/kg of body weight. Exposure was for 19 h/d and 5 d/week.
The exhaust was generated by engines (make not specified) that were operated using the
U.S. 72 (FTP) driving cycle. The concentrations of B[a]P, B[e]P, and chrysene in the diesel
exhaust were 13, 21, and 76 ng/m3, respectively. The overall tumor rate in the lungs of
DPN-treated rats was not affected by the exposure to either filtered or whole diesel engine
exhaust. However, when only pulmonary squamous cell carcinomas were considered, the
exposure to whole diesel exhaust significantly (p <0.05) increased the tumor incidence (Table
7-21). Conversely, nasal tumor incidence was significantly decreased in the rats exposed to
May 1990 7-41 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-20. PULMONARY ADENOMA RESPONSE OF STRAIN A/J MICE EXPOSED
TO DIESEL EXHAUST (1500 jig/m3 PARTICLES) FOR THREE MO
Number of mice according lo number
of lung lumorAnnuse
Treatment
Number
of mice
0
1
2
3
4
13
>13
lot al
number
Mean number of
lung tumonAnouse*
Percent
prevalence
Control
458
114
116
24
4
0
0
0
I7(i
(13810 01
314
Imposed
48S
119
133
26
S
1
1
0
217
0 4StQ(M
34 2
IJrethane
(1 mg/fi)
18
0
0
0
(1
0
0
18
4(17
22 At 190
100
'McaniS E
Source: Kaplan ct at, 1982.
-------
TABLE 7-21. TUMOR INCIDENCE IN FEMALE WBTAR RATS TREATED WITH
DIPHENYLNITROSAMINE (DNP) AND EXPOSED TO CLEAN AIR (CA),
FILTERED DIESEL EXHAUST (FDE), OR WHOLE DIESEL EXHAUST (DE)
Tumor incidence (percent)
Treatment
Squamous cell
carcinomas (lung)
All tumors
(lung)
All tumors
(nasal cavity)
6*25 DPN/kg
CA
FDE
DE
12-5g DPN/kg
CA
FDE
DE
4.4
4.4
46
16.7
14.6
31 J*
84.8
67.4
S3.0
93.8
89.6
89.6
283
4.4'
8.7*
52.1
313'
22.9'
'p iO.Q5 relative to clean air (CA) controls.
Source: Adapted from Heinnch et aL, 1989.
May 1990
7-43
DRAFT-DO NOT QUOTE OR CITE
-------
the diesel engine emissions. These data are consistent with the diesel exhaust-induced 16
percent increase in pulmonary squamous cell carcinoma in rats previously reported by
Heinrich et aL (1986a)
7.23 Dermal Application, Subcutaneous Injection, and Intraperitoneal
Injection Studies
Kunitake et al. (1986) reported a dose-related response in squamous cell papilloma
formation for DET (acetone extract of diesel exhaust tar) and DET + B[a]P applied
dermally to female ICR mice (50/group). The diesel exhaust tar was obtained from a heavy-
duty (HD) 11-L V-6 engine. A tumor incidence rate of 18 percent was noted for DET (45
tng) + B[a]P (1.8-ug), and an incidence rate of 8 percent was noted for DET (45 mg) alone.
No skin cancer was observed for any of the treated mice. In subcutaneous injection
experiments, five subcutaneous injections of DET (500 rag/kg per injection) resulted in a
significant (p <0.01) increase (22.7 percent) in subcutaneous tumors for female C57B1 mice
(n = 22). Five subcutaneous doses of DET at 10, 25, 30, 100, or 200 mg/kg failed to
produce a significant increase in tumor incidence. One of 12 female mice (83 percent) and
4 of 12 male mice (333 percent) developed malignant lymphomas following neonatal
subcutaneous administration of 10 mg DET/mouse. The malignant lymphoma incidence for
the male mice was statistically significant at (p <0.05) from controls. Dermal painting of
DET did not induce skin cancer, although multiple subcutaneous administrations of DET
at high dose levels (500 mg/kg) produced a significant increase in subcutaneous tumors in
mice. Additional studies using LD engine particulate extract dermally applied to female ICR
mice (50/group) and subcutaneous injection studies of HD and LD extracts using female
ICR and nude mice (BALB/c/cA/JCL-nu) were also reported (Kunitake et aL, 1988). In the
LD extract dermal exposure experiments, groups of 50 mice were treated with total doses
of 45, 15, 5, or 0.5 mg of LD particulate extract and one control group was given acetone.
The LD extract served as an initiator and was applied 10 times every other day. A promoter
[2J ng of tetradecanoyl phorbol acetate (TPA)] was applied three times a week for 25
consecutive weeks following application of the LD extract. No significant difference was
noted between the control group and treatment groups for the incidence of malignant
May 1990 7-44 DRAFT-DO NOT QUOTE OR CITE
-------
tumors, papillomas, papillomatous lesions, and acanthosis. In the subcutaneous injection
studies, groups of 30 ICR and nude mice were given a single subcutaneous injection of 10
mg HD extract, 10 mg HD + 50 fig TP A, 10 mg LD extract, 10 mg LD extract + 50 yg
TP A, or 50 fig TP A. No malignant tumors or papillomas were observed. One papilloma-
tous lesion was observed in an ICR mouse receiving LD extract + TP A, and acanthosis was
observed in one nude mouse receiving only TP A.
In what appears to be an extension of the Kunitake et aL (1986) subcutaneous
injection studies, Takemoto et aL (1988) presented additional data for subcutaneously
administered diesel paniculate extracts from HD and LD diesel engines. In this report, the
extracts were administered to 5-week old and neonatal (<24-h old) C57B1 mice of both
sexes. Exhaust particle extract from HD or LD engines was administered weekly for 5 weeks
at doses of 10,25, 50, 100, 200, or 500 mg/kg with group sizes ranging from 15 to 54 animals.
After 20 weeks, comparison with a control group indicated a significant increase in the
incidence of subcutaneous tumors for the 500-mg/kg HD group [5 of 22 mice (22.7 percent),
p <0.01], the 100 mg/kg LD group [6 of 32 (18.8 percent),/? <0.01], and the 500 mg/kg LD
group [7 of 32 (21.9 percent), p <0.01] in the adult mouse experiments. The tumors were
characterized as malignant fibrous histiocytomas and no tumors were observed in other
organs. In the neonate experiments, there was a significantly higher incidence of malignant
lymphomas in males receiving HD extract at 10 mg/kg, of lung tumors for males given 15
mg HD extract/kg, and for males given 5 mg and females given 10 mg LD extract/kg. A
dose-related, but not significant, trend was observed for the incidences of liver tumors for
both the HD- and LD-treated neonatal mice. The incidence of mammary tumors in female
mice and multiple organ tumors in male mice was also greater for some extract-treated mice
but were not dose related. The report concluded that LD particulate extract showed greater
carcinogenicity than did HD particulate extract.
In addition to inhalation studies, Orthoefer et al. (1981) also tested the effects of i.p.
injections of diesel exhaust particulate matter on male Strong A mice. Three groups of 30
mice were injected with 0.1 mL of suspension (particles in distilled water) containing 47,117,
or 235 fig of diesel exhaust particles collected from Fluoropore filters in the inhalation
exposure chambers. Vehicle controls received injections of particle suspension made from
May 1990 7-45 DRAFT-DO NOT QUOTE OR CITE
-------
particulate matter from control exposure filters, positive controls received 20 mg of urethan,
and negative controls received no injections. Injections were made three times weekly for
8 weeks resulting in a total diesel particle dose of 1.1, 2.8, and 5.6 mg for the low, medium
and high dose groups and 20 mg of urethane for the positive control group. These animals
were sacrificed after 26 weeks and examined for lung tumors. For the low, medium, and
high diesel exhaust particle dose groups, the tumor incidence was 2/30, 10/30, and 8/30,
respectively. The tumor incidence or number of tumors per mouse for the diesel exhaust-
treated animals was not significantly different from vehicle controls (8/30) or negative
controls (7/28). The urethane-treated (positive) controls had a tumor incidence of 29/29.
The effects of dermal application of extract from diesel exhaust particles was
examined by Kotin et al. (1955). Acetone extracts were prepared from a blocked diesel
engine (type and size not provided) operated at warm-up mode and under load. These
extracts were applied dermally, three times weekly to male and female C57 and Strain A
mice. Results of these experiments are summarized in Table 7-22. In the initial experiments
using 52 (12 male, 40 female) C57 mice treated with exhaust extract from an engine
operated in a warm-up mode, two papillomas were detected after 13 mo. Four tumors in
8 surviving of 50 exposed male Strain A mice treated with exhaust extract from an engine
operated under full load were detected 16 mo after the start of treatment. For female
Strain A mice treated with extract from an engine operated under full load, 17 tumors were
detected in 20 of 25 mice surviving longer than 13 mo. This provided a significantly
increased tumor incidence of 85 percent.
Pepelko and Peirano (1983) described a series of experiments wherein male and
female Strain A mice were injected with diesel emission particles, particle extracts, or various
environmental mixtures of known carcinogenicity including cigarette smoke condensate, coke
oven emissions, and roofing tar emissions. Injection of urethan or DMSO served as positive
or vehicle controls, respectively. Diesel exhaust was produced by Nissan engines as
described for Pepelko and Peirano (1983). An 8-cylinder Oldsmobile engine operated at the
equivalent of 40 mph was also used for comparison of emission effects from different makes
and models of diesel engine. The mice were sacrificed at 9 mo of age and lungs examined
for histopathological changes. The only significant findings other than for positive controls
May 1990 7-46 DRAFT-DO NOT QUOTE OR CUE
-------
TABLE 7-22. TUMORIGENIC EFFECTS OF DERMAL APPLICATION OF
ACETONE EXTRACTS OF DIESEL EXHAUST
Number of
Dale started animab SirainAcx
3/5/52 52 C57/40F,
12M
5/19/52 50 A/M
2/13/53 25 A/F
Dale of
appearance of
Sample material first tumor
Extract of paniculate material 4/1/53
obtained during warm-up
Extract of particulate material 9/2/53
obtained during
full load
Lxtract of particulate material .1/20/54
obtained during
full load
Survivors at Date
time of experiment
first tumor Total tumors terminated
33 2 1/4/54
8 4 4/19/54
20 17 7/23/54
Source: Kotkin el ill, 1955.
-------
were increases in numbers of lung adenomas per mouse, and small increases in tumor
incidence in male mice receiving Nissan diesel engine exhaust extract and in female mice
receiving coke oven extract injections. The number of mice with tumors was not, however,
significantly changed in these groups. The overall negative findings of this study may
indicate that the lung was not the target organ for the carcinogens present, the active
component(s) was not reaching the target organ (lung), or the mouse strain used was not
a sensitive test for the type of carcinogens present.
Depass et al. (1982) examined the potential of diesel exhaust particles and
dichloromethane extracts of diesel exhaust panicles to act as complete carcinogens,
carcinogen initiators, or carcinogen promoters. In skin-painting studies, the exhaust material
was obtained from an Oldsmobile 350D engine operated under constant load at 65 km/h.
The exhaust panicles were collected at a temperature of 100±10*C. Groups of 40 C3H/HeJ
mice were used because of their low spontaneous tumor incidence. For the complete
carcinogenesis experiments, diesel exhaust panicles were applied as a 5- or 10 percent
suspension in acetone. Dichloromethane extract was applied as S-, 10-, 25-, or 50 percent
suspensions. Negative controls received acetone, and positive controls received 0.2 percent
B[a]P. For tumor promotion experiments, a single application of 1.5 percent B[a]P was
followed by repeated applications of 10 percent diesel particle suspension, 50 percent diesel
particle extract, acetone only (vehicle control), 0.0001 percent phorbol 12-myristate 13-
acetate (PMA) as a positive control, or no treatment (negative control). For the tumor
initiation studies, a single initiating dose of 10 percent diesel panicle suspension, 50 percent
diesel particle extract, acetone, or PMA was followed by repeated applications of 0.0001
percent PMA. Following 8 mo of treatment, the PMA dose in the initiation and promotion
studies was increased to 0.01 percent Animals were treated three times per week in the
complete carcinogenesis and initiation experiments and five times per week in promotion
experiments. All test compounds were applied to a shaved area on the back of the mouse.
In the complete carcinogenesis experiments, one mouse receiving the high
(50 percent) suspension developed a squamous cell carcinoma after 714 d of treatment.
Tumor incidence in the B[a]P group was 100 percent and no tumors were observed in any
of the other groups. Squamous cell carcinomas with pulmonary metastases were identified
May 1990 7-48 DRAFT-DO NOT QUOTE OR CITE
-------
in one mouse of the 50 percent diesel exhaust particle extract group. One mouse in the 25
percent extract group developed a grossly diagnosed papilloma. Nineteen positive control
mice possessed tumors (11 papillomas, 8 carcinomas). No tumors were observed for any of
the other treatment groups. Three tumors (two papillomas and one carcinoma) were
identified in the group receiving diesel particle suspension and three tumors (two papillomas
and one fibrosarcoma) were found in the diesel exhaust particle extract group. These
finding were reported to be statistically insignificant using the Breslow and Mantel-Cox tests.
The data from this study indicated that diesel exhaust particles and dichloromethane
extracts of these particles are not effective with regard to tumor promotion or initiation.
Although these findings were not consistent with those of Kotin et aL (1955), the occurrence
of a single carcinoma in a strain known to have an extremely low spontaneous tumor
incidence may be of importance. Furthermore, a comparison between studies employing
different strains of mice with varying spontaneous tumor incidences may result in erroneous
assumptions.
Nesnow et al. (1982) studied the formation of dermal papillomas and carcinomas
following dermal application of dichloromethane extracts from coke oven emissions, roofing
tar, diesel engine exhaust, and gasoline engine exhaust. Diesel exhaust from five different
engines including a pre-production Nissan 220C, a 350 in3 Oldsraobile, a prototype VW
Turbo Rabbit, a Mercedes 300D, and a heavy-duty Caterpillar 3304 were used for various
phases of the study. Male and female Senear mice (40 per group) were used for tumor
initiation, tumor promotion, and complete carcinogenesis studies. For the tumor initiation
experiments, the diesel exhaust extracts were topically applied in single doses of 100, 500,
1000 or 2000 fig/mouse. The high dose (10,000 jig/mouse) was applied in five daily doses
of 2000 fig. One week later, 2 ng of the tumor promoter, teradecanoylphorbol-acetate
(TPA), was applied topically, twice weekly. The tumor promotion experiments used mice
treated with 50.5 ng of B[a]P followed by weekly (twice weekly for high dose) topical
applications (at the aforementioned doses) of the extracts. For the complete carcinogenesis
experiments, the test extracts were applied weekly (twice weekly for the high doses) for 50
to 52 weeks. Only extracts from the Nissan, Oldsmobile, and Caterpillar engines were used
in the complete carcinogenesis experiments.
May 1990 7-49 DRAFT-DO NOT QUOTE OR CITE
-------
In the tumor initiation studies, both B[a]P alone and the application of the Nissan
engine exhaust extract followed by TP A treatment produced a significant increase in tumor
(dermal papillomas) incidence at 7 to 8 weeks postapplication. By IS weeks, the tumor
incidence was greater than 90 percent for both groups. No significant carcinoma formation
was noted for mice in the tumor initiation experiments following exposure to the extracts of
the other diesel engines, although the Oldsmobile engine exhaust extract at 2.0 mg/mouse
did produce a 40-percent papilloma incidence in male mice at 6 mo. This effect was,
however, not dose-dependent
Results of the tumor promotion experiments indicated that coke oven extract and
roofing tar were as effective as the TPA positive control at promoting B[a]P-initiated
papillomas. No diesel exhaust extracts were tested for tumor promotion activity.
For complete carcinogenesis activity, B[a]P (50.5 Mg/week), coke oven extract (at 1.0,
2.0, or 4.0 mg/week), and the highest dose of roofing tar extract (4.0 mg/week) all tested
positive. Exhaust extracts from only the Nissan, Oldsmobile, and Caterpillar engines were
tested for complete carcinogenic potential, and all three proved to be negative using the
Senear mouse assay.
The results of the dermal application experiments by Nesnow et al. (1982) are
presented in Table 7-23. The tumor promotion-initiation assay was considered positive if
a dose-dependent response was obtained and if at least two doses provided a papilloxna-per-
mouse value that was three times or greater than that of the background value. Based on
these criteria, only emissions from the Nissan were considered positive. Tumor initiation and
complete carcinogenesis assays required that at least one dose produce a tumor incidence
of at least 20 percent. None of the diesel exhaust samples yielded positive results based on
this criterion.
7.2.4 Summary of Animal Carcinogenicity Studies
As early as 1955, Kotkin et al. (1955) provided evidence for tumorigenicity and
carcinogenicity of acetone extracts of diesel exhaust and also provided data suggesting a
difference in this potential depending on engine operating mode. Until the mid 1980s, no
May 1990
7-50
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 7-23. DERMAL TUMORIGENIC ANI> CARCINOGENIC EFFECTS Op VARIOUS EMISSION EXTRACTS
Tumor Initiation Complete carcinogenesis Tumor promotion
Sample Papillomas* Carcinomas* Carcinomas* l'apilomas*
Dcnzofajpyrene
+/+'
+/+
+/+
+/+
Topside coke oven
+/+
-/+
NDd
NO
Cuke oven main
+/~
+/+
+/+
+/+
Roofing lar
+/+
+/+
+/+
+/+
Nissan
+7+
+/+
ND
Oldsmobile
+/+
-/-
ND
VW Rabbit
r
ND
Mercedes
+/-
¦I-
ND
ND
Caterpillar
•A
-/-
ND
Residential furnace
-/-
ND
ND
Mustang
+/+
-/+
Nl)
ND
'Scored oi 6 months
'Cumulative score at I yr.
cMalc/Temalc.
dND = Not determined.
*1 = Incomplete.
Source: Ncsnow ci ;il, 1982
-------
chronic studies assessing inhalation of diesel exhaust, the relevant mode for human exposure,
had been reported, and it is this route of exposure with which the most extensive, recent
studies were concerned. Studies employing rats and an adequate experimental design were
nearly all positive in demonstrating diesel exhaust-induced increases in tumorigenicity. The
9.5 percent increase in tumor incidence for female Wis tar rats reported by Heinrich et aL
(1986a) is supported by the repon by Mauderly et al. (1987), which reported a 3.6-percent
and 12.8-percent tumor incidence in F344 rats following chronic exposure to diesel exhaust
at particle concentrations of 3 J and 7.0 mg/m3, respectively. However, only one of the
squamous cell tumors reported by Heinrich et aL (1986a) was classified as a carcinoma. In
the Mauderly et aL (1987) study, the carcinoma incidence was 0.9, 1.3, 0.5, and 7.5 percent
for the control, low, medium, and high exposure groups, respectively. The study by Wong
et al. (1986) affirms observations of potential carcinogenicity of diesel exhaust by providing
evidence for DNA damage in rats. Similarly, Iwai et al. (1986), demonstrated diesel exhaust-
induced tumorigenicity in rats exposed to an exhaust particle concentration of 4.9 mg/m3,
although the sample size was small. This study also reported development of a splenic
lymphoma which represents the only nonpulmonaiy tumor resulting from inhalation exposure
to diesel exhaust. The long-term study by Ishinishi et aL (1986) showed a greater incidence
of carcinomas (6.5 percent) in rats following 30-mo exposure to diesel exhaust at 4 mg/m3,
but not at lower ( 0.4-, 1-, or 2-rag/m3) exposure levels. However, Brightwell et al. (1986)
demonstrated a dose-dependent increase in tumor incidence for male and female F344 rats
exposed to unfiltered diesel exhaust (five 16-h periods per week) at concentrations as low
as 2.2 mg/m3 and also at 6.6 mg/m3. This study indicated that the tumor incidence was
higher for female (0 percent, 15 percent, or 54 percent at 0.0, 2J2, or 6.6 mg/m3) than for
male rats (1 percent, 4 percent, or 23 percent for 0.0, 22, or 6.6 mg/m3). Thus, these studies
demonstrated carcinogenic effects in rats at exposure levels ranging from 2.2 to 7.0 mg/m3.
Heinrich et aL (1986a,b) and Pepelko and Peirano (1983) also provided evidence of
carcinogenicity in mice.
Both the Heinrich et al. (1986a) and Brightwell et al. (1986) studies provided negative
results for tumorigenicity of diesel exhaust in hamsters, a species known for its resistance to
tumor induction. Negative results were also presented by several other investigators
May 1990 7-52 DRAFT-DO NOT QUOTE OR CITE
-------
(Takemoto et aL, 1986; Schreck et aL, 1981; Karagianes et aL, 1981), but these studies
tended to employ inadequate exposure durations, low exposure concentrations, or
inadequate animal numbers per group. Similarly, the studies using monkeys and cats were
of inadequate duration (2 yr) for these longer-lived species.
Alternate exposure routes including dermal exposure, skin painting, and subcutaneous
injection provided additional evidence for tumorigenic effects of diesel exhaust Evidence
for tumorigenicity was demonstrated by Kotin et aL (1955) for mice to which an acetone
extract of diesel exhaust panicles was applied dermally. Nesnow et al. (1982) also showed
that extracts from some diesel engines were potentially tumorigenic following dermal
application to rodents. A significant increase in the incidence of subcutaneous tumors in
female C57B1 mice was reported by Kunitake et al. (1986) for subcutaneous administration
of light-duty diesel exhaust tar extract at doses of 500 mg/kg. Doses at or below 200 rag/kg,
however, were negative. Takemoto et al. (1988) provided additional data for this study and
reported an increased tumor incidence in the mice following injection of light-duty engine
exhaust extract at doses of 100 and 500 mg/kg. Negative results were reported by Depass
et al. (1982) for skin-painting studies using mice and acetone extracts of diesel exhaust
particle suspensions. However, in this study the exhaust particles were collected at
temperatures of 100 * C, a temperature that would minimize the condensation of vapor-
phase organics and, therefore, reduce the availability of potentially carcinogenic compounds
that might normally be present on diesel exhaust particles. Intraperitoneal injection studies
were generally negative.
Experiments using tumor initiators such as DEN, B[a]P, DPIN, or DBahA (Brightwell
et aL, 1986; Heinrich et aL, 1986a; Takemoto et aL, 1986) did not provide conclusive results
regarding the carcinogenic potential of filtered vs. whole diesel exhaust. Some of these
studies, however, strongly imply the importance of adsorbed organic compounds for the
carcinogenic effect, and data are available regarding the carcinogenic activity of several
compounds known to be components of diesel exhaust.
Diesel exhaust is composed of gaseous and particulate phases and is known to be a
complex mixture containing verified and potential carcinogens. A study by Grimmer et aL
(1987) demonstrated that a whole exhaust condensate fraction (lib) containing PAHs with
May 1990 7-53 DRAFT-DO NOT QUOTE OR CITE
-------
4 to 7 rings produced a high tumor incidence when implanted into rat lungs. It was also
noted that this fraction represented only 0.8 percent of the total weight of the exhaust
condensates, and that some tumorigenicity was also associated with nitroaromatic fractions.
The PAH traction produced a tumor incidence similar to that of a low concentration of
BMP.
An additional point of concern, as of yet unresolved, is the role of the diesel exhaust
particle in carcinogenicity. Several of the previously discussed studies indicated that only
whole (unfiltered) diesel exhaust is tumorigeruc or carcinogenic and that these properties are
eliminated or greatly minimized in filtered diesel exhaust exposure. Although the
tumorigenicity of the gaseous fraction is presently unresolved and experiments using filtered
exhaust were negative, most of these experiments did not provide definitive evidence that
a maximum tolerated dose was achieved. Furthermore, it has not yet been unequivocally
determined that the carbon core of the exhaust particle is not without carcinogenic potential.
The fact that allegedly inert, "noncarcinogenic" dusts such as titamum dioxide, shale dust, and
quartz dust have been shown to induce lung cancer at very high concentrations is of concern
in this respect. That the carbon core alone may be carcinogenic or may have promotional
effects in initiated cells remains to be determined. Studies currently in progress may resolve
this question.
Although uncertainties exist regarding the tumorigeric potential of the gaseous
component and the carbon core component of diesel exhaust, it is clear that diesel exhaust
is carcinogenic in animals. This contention is supponed by positive results in numerous,
independent studies in male and females of at least two species and by several routes of
administration, including inhalation, intratracheal administration, skin painting, and
subcutaneous injection.
A summary of studies assessing the tumorigenic and carcinogenic effects in laboratory
animals following inhalation exposure to diesel exhaust is presented in Table B-l of
Appendix B.
May 1990
7-54
DRAFT-DO NOT QUOTE OR CITE
-------
73 REFERENCES
Barnhart, M. L; Chen, S4 Salley, S. O.; Puro, R (1981) Ultras tructure and morphometry of the alveolar long
of guinea pigs chronically exposed to diesel engine exhaust: Six months experience. J. AppL ToncoL 1:
88-103.
Brightwell, J.; Fouillet, Xa Cassano-Zoppu A.-L; Gatz. R-; Duchosal, F. (1986) Neoplastic and functional
changes in rodents after chronic inhalation of engine exhaust emissions. In: Ishinishi, Koizumi, A4
McQelian, R4 Stober, W„ eds. Carcinogenic and Mutagenic Effects of Diesel Engine Exhaust.
Amsterdam: Elsevier, pp. 471-485.
Carpenter, IC; Johnson, J. R (1979) Analysis of the physical characteristics of diesel particulate matter using
transmission electron microscope techniques. SAE Transactions 88: 2743.
Cuddihy, R. G4 Griffith, V. G; McClellan, R. O. (1983) Health risks from light-duty diesel vehicles. Environ.
ScL TechnoL 18: 14-21A.
Depass, L. R_; Chen. K. C: Peterson. L. C. (1982) Dermal carcinogenesis bioassays of diesel particulates and
dichloromeihane extraa of diesel paniculates in C3H mice. In: Lewtas, J., ed. Toxicological Effects of
Emissions from Diesel Engines. Amsterdam: Elsevier, pp. 321-327.
Grimmer, G4 Brune, PL; Deutsch-Wenzel, R_; Dettbarn, Jacob. J.; Naujack, K.-W.; Mohr, U4 Ernst, R
(1987) Contribution of polycyclic aromatic hydrocarbons and nitro-derivatives to the carcinogenic impact
of diesel exhaust condensate evaluated by implantation into the lungs of rats. Cancer Lett 37:173-180.
Heinrich, U4 Muhle, R; Takenaka, S.; Ernst, R; Fuhst, R4 Mohr, U.; Pott, F.; Stober, W. (1986a) Chronic
effects on the respiratory tract of hamsters, mice and rats after long-term inhalation of high concentrations
of filtered and unfiltered diesel engine emissions. J. AppL ToxicoL 6: 383-395.
Heinnch, U4 Pott, F.; Rittinghausen, S. (1986b) Comparison of chronic inhalation effects in rodents after
long-term exposure to ether coal oven flue gas mixed with pyrolized pitch or diesel engine exhaust. In:
Ishinishi. N4 Koizumi. A.; McQelian. R.; Stober, W., eds. Carcinogenic and Mutagenic Effects of Diesel
Engine Exhaust. Amsterdam: Elsevier, pp. 441-457.
Heinnch, U4 Peters, L; Furst. R^ Mohr, U. (1989) The effect of automotive exhaust exposure on the
carcinogenicity of diphenylnitrosamine (DPN) in the respiratory tract of rats. Exp. PathoL 37: 51-55.
Huisingh, J.; Bradow, R_; Jungers, R4 Claxton, L; Zweidinger, R-; Tejada, S. et aL (1978) Application of
bioassay to the characterization of diesel panicle emissions. In: Waters, M. D4 Nesnow, S4 Huisingh, J.
L.; Sandhu, S^; Claxton, L-. eds. Application of shon-term bioassay in the fractionation and analysis of
complex mixtures. New York: Plenum pp. 381-418.
Ishinishi N4 Inamasu, T., Hisanaga, A.; Tanaka, Aa Hirata, M.; Ohyama, S. (1988) Intratracheal instillation
study of diesel particulate extracts in hamsters. In: Diesel Exhaust and Health Risk, Research Committee
for HERP Studies, Ibaraki, Japan, pp. 209-216.
Ishinishi, N.; Kuwabara, N.; Nagase, S.; Suzuki, T.; Ishiwaia. S.; Kohno, T. (1986) Long-term inhalation studies
on effects of exhaust from heavy and light duty diesel engines on F344 rats. In: Ishinishi, N.; Noizumi, A4
McClellan, R.; Stober. W., eds. Caranogenicity and mutagenicity of diesel engine exhaust. Amersterdam:
Elsevier, pp. 329-348.
May 1990
7-55
DRAFT-DO NOT QUOTE OR CITE
-------
Iwai, K.; Udagawa, Yamaguhi, M; Yamada, R (1986) Long-ierm inhalation studies of diesel exhausi on
F344 SPF rats. Incidence of lung cancer and lymphoma. In: Ishinishi, N.; Koizumi, A.; McClellan, R^
Stober, W., eds. Carcinogenic and mutagenic effects of diesel engine exhaust. Amsterdam: Elsevier, pp.
349-360.
Kaplan. H. L-; Mackenzie. W. F.; Springer, K. J.; Schreck. R. M.; VostaL J. J. (1982) A subchronic study of
the effects of exposure of three species of rodents to diesel exhaust. In: Lewtas, J., ed. Toxicological
Effects of Emissions from Diesel Engines. New York: Elsevier Biomedical, pp. 161*220.
Karagianes, R. F4 Palmer, R. F.; Bosch. R. H. (1981) Effects of inhaled diesel emissions and coal dust in rats.
Am. Ind. Hyg. Assoc J. 42:382-391.
Kotin, P4 Falk, R; Thomas. M. (1955) Aromatic hydrocarbons. IIL Presence in the paniculate phase of diesel
engine exhausts and the carcinogenicity of exhaust extracts. Arch. Ind. Health 11: 113-120.
Kumtake, E; Shimamura, 1C; Katayama, H.; Takemoto, K.; Yamamoto, A.; Hisanaga. A.; Ohyama, S.;
Ishinishi. N. (1986) Studies concerning carcinogenesis of diesel paniculate extracts following intratracheal
instillation, subcutaneous injection, or skin application. In: Ishinishi. N.; Koizumi. A.; McClellan. Ra
Stober. W.. eds. Carcinogenic and mutagenic effects of diesel engine exhaust. Amsterdam: Elsevier, pp.
235-252.
Kumtake, E.; Imase, A.; Shimamura, K4 Ishinishi. N.; Hisanga. A.; Tanaka. A. (1988) Skin application and
subcutaneous injection experiments of diesel paniculate extracts using ICR mice and nude mice. In: Diesel
Exhaust and Health Risks. Research Committee for HERP Studies, Ibaraki. Japan, pp. 217-225.
Lewis, T. R-; Green, F. K; Moorman, W. Burg, J. A4 Lynch, D. W. (1986) A chronic inhalation toxicity
study of diesei engine emissions and coal dust alone and combined. In: Ishinishi, N4 Koiznma, A^
McQellan, R^ Stober. W., eds. Caranogenic and mutagenic effects of diesel engine exhaust. Amsterdam:
Elsevier, pp. 361-380.
Mauderly. J. L-; Jones, R. KU Griffith, W. C; Henderson. R. F.; McClellan. R. O. (1987) Diesel exhaust is
a pulmonary carcinogen in rats exposed chronically by inhalation. Fundam. AppL ToxicoL 9: 208-221.
Mohr, U.; Takenaka. S4 Dungwonh. D. L. (1986) Morphologic effects of inhaled diesel engine exhaust on
lung? of rats: Comparison with effects of coal oven Que gas mixed with pyrolized pitch. In: Ishinishi. N.;
Koizumi. McClellan. R-; Stober, W„ eds. Caranogenic and Mutagenic Effects of Diesel Engine
Exhaust Amsterdam: Elsevier, pp. 459-470.
Mokler, B. V.; Archibeque, F. A^ Beethe, R. Kelly, C P. J.; Lopez, J. L-; Stafford. D. L. (1984) Diesel
exhaust exposure system for animal studies. Fundam. AppL ToxicoL 4:270-277.
Nesncrw, Evans, C; Stead, A; Creason, J4 Slaga, T. J.; Triplett, L. L. (1982) Skin carcinogenesis studies
of emission extracts. In: Lewtas, J., ed. Toxicological Fffms of Emissions from Diesel Engines.
Amsterdam: Elsevier, pp. 295-320.
Orthoefer, J. Gu Moore, W.; Kraemer, D. (1981) Carcinogenicity of diesel exhaust as tested in strain a mice.
Environ. Int. 5: 461-471.
Pepelko, W. E; Peirano. W. B. (1983) Health effects of exposure to diesel engine emissions. J. Amer. ColL
ToxicoL 2: 253-306.
May 1990
7-56
DRAFT-DO NOT QUOTE OR CUE
-------
Shefner, A. M.; Collins, S. C; Doolcy, Fiks, A.; Graf, J. Preache, M. M. (1982) Repiratory
carcinogenicity of diesel fuel emissions interim results. .In: Lewtas, J., ed. Toxicological Effects of
Emissions from Diesel Engines. Amsterdam: Elsevier, pp. 329-350.
Schreck, R. M.; Soderholm, S. C; Chan, T. L-; Smiler, K. L; D'Arcy, J. B. (1981) Experimental conditions in
GMR chronic inhalation studies of diesel exhaust. J. AppL ToxicoL 1: 68-76.
Stober, W. (1986) Experimental induction of tumors in hamsters, mice and rats after long-term inhalation of
filtered and unfiltered diesel engine exhaust In: Ishinishi, N4 Koizumi, Aa McClellan, R4 Stober, W_, eds.
Carcinogenic and Mutagenic Effects of Diesel Engine Exhaust. Amsterdam: Elsevier, pp. 421-429.
Takaki, Y4 Kitamura, S4 Kuwabara, N.; Fukuda, Y. (1989) Long-term inhalation studies of exhaust from <<«>«»>
engine in F-344 rats: The quantitative relationship between pulmonary hyperplasia and anthracosis. Exp.
PathoL 37:56-62. Report of poster presentation.
Takemoto, IC; Yoshimura. R; Katayama, R (1986) Effects of chronic inhalation exposure to diesel exhaust
on the development of lung tumors in di-isopropanol-nitrosamine-treated F344 rats and newborn C57BL
and ICR mice. In: Ishinishi, N.; Koizumi, A.; McClellan, R4 Stober, W., eds. Carcinogenic and Mutagenic
Effects of Diesel Engine Exhaust. Amsterdam: Elsevier, pp. 311-327.
Takemoto, K.; Katayama, R; Kuwabara, T.; Hasumi, M. (1988). Carcinogenicity by subcutaneous
administration of diesel particulate extracts in mice. In: Diesel Exhaust and Health Risks. Research
Committee for HERP Studies, Ibarald, Japan, pp. 227-234.
Vuk. C T4 Jones, M. A.; Johnson, J. R (1976) The measurement and analysis of the physical character of
diesel paniculate emissions. SAE Transactions 8S: 556.
Wong, D4 Mitchell, C E.; Wolf, R. IC; Mauderiy, J. L; Jeffrey, A. M. (1986) Identification of DNA damage
as a result of exposure of rats to diesel engine exhaust. Carcinogenesis 7: 1595-1597.
Zamora, P. O.; Gregory, R. E.; Brooks, A. L. (1983) In vitro evaluation of the tumor-promoting potential of
diesel-exhaust-panicle extracts. J. Toxicol. Environ. Health 11: 188-197.
May 1990
7-57
DRAFT-DO NOT QUOTE OR CITE
-------
8. PHARMACOKINETIC CONSIDERATIONS IN THE
PULMONARY CARCINOGENICITY OF DIESEL
ENGINE EMISSIONS
8.1. INTRODUCTION
Analysis of whole animal studies examining the carcinogenicity of diesel exhaust
revealed the importance of understanding dosimetric factors in evaluating the carcinogenic
potential of diesel emissions. Several studies addressed the effects of diesel exhaust
particulate matter concentration on the pulmonary retention, clearance, and the potential
for pulmonary overload of these particles. Although many other factors, such as metabolism
considerations and target organ sensitivity, are important in obtaining an accurate assessment
of risk, understanding aspects of pulmonary retention, clearance and bioavailability of
potentially carcinogenic compounds are also quite relevant The following sections
summarize those studies examining pharmacokinetic parameters such as exposure
concentration, clearance rates and compartmental considerations, and alveolar macrophage
functions as well as factors possibly affecting these processes and the ramifications of these
regarding the bioavailability of particle-associated organics. Additionally, some discussion
of factors deemed to be important in extrapolation modeling to humans will be discussed
to the extent feasible.
&2. REGIONAL LUNG DEPOSITION OF INHALED
PARTICLES BY HUMANS AND ANIMALS
The site of deposition of particulate matter in the respiratory tract is dependent upon
a number of factors including respiratory tract anatomy (airway dimensions and branching
configurations) and physiology (breathing mode and rate, pulmonary volumes and capacities,
and clearance mechanisms) and the physicochemical characteristics (particle size, shape,
density, and solubility) of the inhaled particles. Deposition of particles may occur through
several different processes or combinations of these, including diffusion, sedimentation,
June 1990
8-1
DRAFT-DO NOT QUOTE OR CITE
-------
interception, electrostatic precipitation, and impaction. Obviously, the anatomical and
physiologic factors are paramount in determining dissimilar regional deposition, especially
among species. A large airway diameter creates a greater turbulent flow velocity than does
a small airway diameter. Assuming the physicochemical characteristics of the particle to be
constant, it follows that the larger diameter airways of humans would create a turbulent flow
and the smaller diameter airways of laboratory animals such as rodents would exhibit a
laminar flow. Furthermore, when compared with humans, most laboratory animals have
tracheas that are longer relative to the diameter. The flow patterns resulting from these
anatomical differences would greatly affect most of the deposition processes. For example,
increased flow through the larger diameter primary passages of humans would result in
greater deposition of panicles in the bronchi. Humans also tend to have a more
symmetrical branching pattern than do animals, a condition that results in greater impaction
deposition at the carinae of the bronchi and bronchioles. Factors affecting pulmonary
deposition are reviewed in U.S. EPA (1988) and for more detailed consideration of these
processes, the reader is referred to Raabe (1979), Phalen and Oldham (1983), and
Lippmann and Schlesinger (1984).
83. TRACHEOBRONCHIAL CLEARANCE
OF PARTICULATE MATTER
Upon inhalation of paniculate matter, deposition will initially occur in the
nasopharyngeal and tracheobronchial regions of the respiratory tract. Due to high air flow
velocities and abrupt directional changes in these regions, inertial impaction and sedimenta-
tion are the primary deposition mechanisms (especially for particles larger than 23 pm
aerodynamic equivalent diameter). Mechanical processes such as sneezing and nose wiping
remove particles from the nasopharyngeal region. Gearance of paniculate matter from the
tracheobronchial region is mediated primarily by mucociliary transport, and is usually a more
rapid clearance process than those operating in deep lung regions. Mucociliary transpon
(often referred to as the mucociliary escalator) is accomplished by the mucus-covered,
ciliated epithelia which line the respiratory tract from the trachea through the terminal
bronchioles. Gearance rate by this system is determined primarily by the flow velocity of
May 1990
8-2
DRAFT-DO NOT QUOTE OR CITE
-------
the mucus, which is greater in the proximal airways and decreases distally. These rates also
exhibit interspecific variability. The half-times for tracheobronchial clearance are usually on
the order of hours while those for pulmonary clearance may be hundreds of days or years.
The relationship between the early clearance of insoluble particles from tracheobronchial
regions and their long-term clearance from pulmonary regions is illustrated in Figure 8-1.
Cuddihy and Yeh (1986) reviewed respiratory tract clearance of particles inhaled by
humans. Depending on the type of particle (ferric oxide, teflon discs, albumin
microspheres), the technique employed, and the anatomic region (midtrachea, trachea, main
bronchi), particle velocity (moved by mucociliary transport) ranged from 2.4 to 21.5 mm/min.
The highest velocities were recored for midtracheal transport, and the lowest were for main
bronchi. In one study, an age difference was noted for tracheal mucociliary transport
velocity (5.8 mm/min for individuals less than 30 years of age, and 10.1 mm/min for
individuals over 55 years of age).
The Cuddihy and Yeh (1986) report described salient points to be considered when
estimating particle clearance velocities from tracheobronchial regions and included:
respiratory tract airway dimensions, calculated inhaled particle deposition fractions for
individual airways, and thoracic clearance measurements. Predicted clearance velocities for
the trachea and main bronchi were found to be similar to those experimentally determined
for inhaled radiolabeled particles but not for intratracheally instilled particles. Figure 8-2
illustrates a comparison of the short-term clearance of inhaled panicles by human subjects
and the model predictions for this clearance.
8.4. CLEARANCE FROM DEEP LUNG REGIONS
Several studies have investigated the effects of exposure concentration and lung
burden on the pulmonary clearance of diesel exhaust particles by laboratory animals. These
factors are important for establishing dose-response relationships and determining the
bioavailability of organic compounds adsorbed to these particles.
Wolff et aL (1986, 1987) provided clearance data (t,^) and lung burden values for
F344 rats exposed to diesel exhaust for 7 h/d, 5 d/week for 24 mo. Exposure
May 1990
8-3
DRAFT-DO NOT QUOTE OR CITE
-------
TRACHEOBRONCHIAL
DEPOSITION
HOURS AFTER INHALATION
Figure 8*1. Clearance of insoluble panicles deposited in tracheobronchial and
pulmonary regions.
Source: Cuddihy and Yeh, 1986.
May 1990
DRAFT-DO NOT QUOTE OR CITE
-------
0
&
01
o
o
o
<
s
o
—«*nae or three
I MEASUREMENTS _
04:
0.6 -
1 0.*-
i
u. _
O 0.2 -
z
o
—MODEL PROJECTION
SAME AS LOWER
LBOT OP RANGE
O
<
e
20 40 80 80 100
HOURS AFTER INHALATION
120
Figure 8-2. Short-term thoracic clearance of inhaled panicles as determined by model
prediction and experimental measurement
Source: Cuddihy and Yeh. 1986 (from Stahlhofen et al., 1980)
May 1990
8-5
DRAFT-DO NOT QUOTE OR CITE
-------
concentrations of 035, 3.5, and 7.0 mg soot/m3 were employed in this whole body-inhalation
exposure experiment. Intermediate clearance of 67Ga203 particles (one half-hour, nose-only
inhalation) was assessed after 6t 12, 18, and 24 mo of exposure at all of the diesel exhaust
concentrations. A two-component function described the clearance of the administered
radiolabel:
F[t) " A exp(-0.693 //T,)+fl exp(-0.693 th2) ,
where F(t) was the percentage retained, A and 8 were the magnitudes of the two
components [nasal, pulmonary conducting airways, and gastrointestinal clearance], and
intermediate lung clearance, respectively], and r, and r: the half-times for the A and B
components, respectively. The early clearance half-times (r,), representing clearance from
primary, ciliated conducting airways, were similar for rats in all exposure groups at all time
points except for those in the high exposure (7.0 mg/m3) group following 24 mo of exposure
where the clearance time was faster than that of the controls. Significantly longer B
component clearance half-times, representing the more rapid clearance from nonciiiated
passages such as bronchioles and alveoli, were noted after as few as 6 mo exposure to diesel
exhaust at 7.0 mg/m3 and 18 mo exposure to 3 J mg/m3. Nose-only exposure to l34Cs fused
aluminosilicate particles (FAP) were used to assess long-term clearance. Following 24-mo
exposure to diesel exhaust, long-term clearance of ,wCs-FAP was significantly (p <0.01)
altered in the 3.5 and 7.0 mg/m1 exposure groups (t1/4 of 264 and 240 d. respectivelyj relative
to the 0.35 mg/m3 and control groups (t% of 81 and 79 d, respectively). Long-term clearance
represents the slow component of particle removal from alveoli. The decreased clearance
correlated well with increased lung burden in the 3 J and 7.0-mg/m3 exposure groups.
The Heinrich et al. (1986) and Stober (1986) studies reported increased tumor
incidence following lifetime inhalation exposure to diesel exhaust (4 mg/m3) for 19 h/d,
5 d/week. The impaired pulmonary clearance observed in the rats exposed to total diesel
exhaust may have allowed for greater desorption of particle-bound PAHs, thus accounting
for the increased tumor incidence. Based on the analysis of the diesel exhaust used in the
experiments, values for rat ventilatory functions, and estimates of deposition and clearance,
May 1990 8-6 DRAFT-DO NOT QUOTE OR CITE
-------
bioavailability of particle-bound organic material was estimated at 6 to 15 mg for a 2-5 yr
period.
Chan et al. (1981) subjected 24 male F344 rats to nose-only inhalation of diesel
exhaust (6 mg/m3) labeled with 13lBa or UC for 40 to 45 minutes, and assessed total lung
deposition, retention, and elimination. The deposition efficiency based on radiolabel
inventory was 15 to 17 percent. Quantitation of 13,Ba label in the feces during the first 4 d
following exposure indicated that 40 percent of the diesel particulates were eliminated via
mucociliary clearance. Gearance of the particles from the lung followed a two-phase
elimination described by:
R(t) = AeP'+Be-**
where the first term represents a rapid (t^ of 1 d) mucociliary transport elimination, and the
second term represents a slower (t1yi of 62 d) macrophage-mediated alveolar clearance. The
effects of prolonged exposure and overloading were not addressed in this study.
Accumulated lung burden of diesel soot particles following 18 mo, 7 h/d, 5 d/week
exposure to diesel exhaust was reported by Griffis et al. (1983). Male and female F344 rats
exposed to 0.15, 0.94, or 4.1 mg soot/m3 were sacrificed at 1 d, and 1,5, 15,33, and 52 weeks
after exposure, and diesel soot extracted from lung tissue dissolved in tetramethylammonium
hydroxide. Following centnfugation and washing of the supernatant, diesel soot content of
the tissue was quantitated using spectrophotometric techniques. The analytical procedure
was verified by comparing results to recovery studies using known amounts of diesel soot in
lungs of unexposed rats. Lung burdens of the 0.15, 0.94, and 4.1 mg soot/m3 exposure
groups were 35, 220, and 1890 jxg/g lung, respectively, 1 d after the 18-week exposure.
Long-term clearance rates for the 0.15 and 0.94 mg/m3 groups had estimated half-times of
87±28 and 99±8 d, respectively. The clearance half-time for the 4.1 mg/m3 exposure group
was 165±8 d which was significantly (p <0.0001) greater than those of the lower exposure
groups. This study provided evidence for an overload phenomenon and possible threshold
relative to clearance of diesel exhaust particulates from the lungs of rats.
May 1990
8-7
DRAFT-DO NOT QUOTE OR CITE
-------
In a subsequent study (Lee et aL, 1983) a three-phase model was used to describe
the clearance of diesel exhaust particles (7 mg/m3 for 45 min or 2 mg/m3 for 140 min) by
F-344 rats (24 per group) exposed by nose-only inhalation. The exposure protocols provided
comparable total doses based on "C radiolabeL UC02 resulting from combustion of 14C-
labeled diesel fuel was removed by a diffusion scrubber to avoid erroneous assessment of
UC intake by the animals. Retention of the radiolabeled particles was determined up to
335 d after exposure, and resulted in the derivation of a three phase clearance of the
particles. The resulting clearance half-times for the three-phases were 1, 6, and 80 d. The
three clearance phases are taken to represent removal of tracheobronchial deposits by the
mucociliary escalator, removal of particles deposited in the proximal respiratory bronchioles,
and alveolar clearance, respectively. Species variability in clearance of diesel exhaust
particles was also demonstrated by the fact that Hanley guinea pigs exhibited negligible
clearance from day 10 to day 432 following a 45 min exposure to a diesel particle
concentration of 7 mg/m3. Initial deposition efficiency (20±2 percent) and short-term
clearance were, however, similar to the rats.
Lung clearance in male F344 rats pre-exposed to diesel exhaust at 0.25 or 6 mg/m3,
20 h/d, 7 d/week for periods lasting from 7 to 112 d was studied by Chan et al. (1984).
Following this pre-exposure protocol, rats were subjected to 45 min nose-only exposure to
,4C-diesel exhaust, and pulmonary clearance of radiolabel monitored for up to 1 yr. First
order clearance for the two compartment model, R(t) = Ae'^'+Be'"21, yielded alveolar
clearance half-times of 166 and 562 d for rats pre-exposed to 6.0 mg/m3 for 7 and 62 d.
respectively. These values were significantly (p <0.05) greater than the clearance half-time
of 77±17 d for control rats. The same clearance values for rats of the 0.25-mg/m3 groups
were 90±14 and 92±15 d, respectively. Estimated lung burdens of diesel exhaust particles
for controls, and for those exposed for 7 d and 62 d to 6.0 mg/m3 were 0, 0.7 and 6 J mg,
respectively. Lung burden of particulates resulting from pre-exposure to 6.0 mg/m3 for 112 d
was 11.8 mg. No alveolar clearance was observed for these rats. The two-compartment
model, however, represents overall clearance of the tracer particles and does not account
for a slower clearing compartment for panicles sequestered in alveolar macrophage
aggregates. A lung retention model for pre-exposed rats accounts for mucociliary clearance
May 1990 8-8 DRAFT-DO NOT QUOTE OR CITE
-------
(1st term), an active clearance phase by alveolar macrophages (2nd term), and a residual
fraction representing macrophage aggregates with limited clearance capabilities (3rd term)
R(t) - Ae-u*Be'u**R0 .
The magnitude of R, was shown to be dependent on the initial particulate burden and,
therefore, would increase in higher exposure scenarios. This study emphasizes the
importance of particle overloading of the lung, and the ramifications on clearance of
particulates and associated organics.
Morrow (1988) provided additional information on the significance of alveolar
macrophage-mediated clearance of particulates from the lungs and possible mechanisms by
which this process becomes overloaded- Based on current data for exposure, deposition
rate, pulmonary retention half-time, and ventilatory parameters, kinetic descriptions of
panicle buildup and steady-state phases are provided. These kinetic models, along with
experimentally determined alveolar macrophage pool size for rat lungs (2Jxl07 cells) and
panicle size data, provide a basis for describing the overloading of alveolar macrophages
when rats are used as the model system. Since histological data indicate that the phagocytic
functions of alveolar macrophages are not impaired in the presence of high paniculate
concentrations, it is postulated that impaired translocation of panicle-bound macrophages
may be the critical factor in reduced pulmonary clearance of paniculates. The precise
mechanism by which macrophage translocation is impaired is uncenain but may involve
alterations of chemoattractants, a critical volume overload of the macrophage, or alterations
of activators and modulators of alveolar macrophage function.
Lehnen (1988), in studying particle-alveolar macrophage relationships during
clearance of insoluble, noncytotaxic particles, provided results indicating that phagocytized
particles were gradually redistributed among the total lung macrophage population during
the clearance process. This type of translocation would affect the ultimate clearance of the
paniculate matter and, at least in part, may explain long-term retention of paniculate
matter.
May 1990
8-9
DRAFT-DO NOT QUOTE OR CITE
-------
Oberdtterster et aL (1988), using F344 rats receiving intratracheal instillations of 33
Mm polystyrene microspheres labeled with ,4,Ce and 103 jim particles labeled with MNb,
followed lung clearance of these panicles for 200 d after administration of a high dose
(100 Mg small particles + 100 jig large particles) or a low does (40 Mg small + 10 jig large
particles). Only minimal clearance was detected for the large particles at both doses,
suggesting inability of alevolar macrophages containing one large particle to clear from the
lung. The researchers hypothesized that the reported minimal clearance of the large
particles was analogous to a chronic overloading with small particles. It was noted that the
volume (600 nm3) of one large particle (103 Mm) represented 60 percent of the volume of
an average alveolar macrophage and that this critical level could be obtained by chronic
overloading with many small panicles that would provide an equivalent volume. The study
indicated that impairment of alveolar macrophage-mediated clearance of panicles would
occur when this critical phagocytized volume is maintained for extended periods.
8.4.1. Species Variability in Pulmonary Clearance Processes
The delivered dose of potentially carcinogenic agents in diesel exhaust is dependent
upon a number of factors. The net dose received will be dependent upon the minute
ventilation and, therefore, a smaller animal may receive a higher total dose of panicles
because of a higher ventilation rate-to-body weight ratio. Additionally, factors such as
pulmonary deposition of paniculate matter, total lung burden of panicles and their clearance
irom the lungs, and dissolution of organic compounds from the paniculate matter are
instrumental in determining the bioavailable dose. Of special significance in this respect is
species variability in pulmonary clearance of particle burden and its ramifications in
extrapolating from animal data to human risk assessment This emphasis is due, in part, to
the fact that longer clearance times may be associated with increased bioavailability of
chemicals adsorbed to the paniculate matter. The effects of interspecific variability in long-
term pulmonary clearance of paniculate matter on dose adjustment have been reviewed by
Pepeiko (1987).
May 1990
8-10
DRAFT-DO NOT QUOTE OR CITE
-------
As indicated in the previous section, pulmonary clearance of particulate matter may
occur as a rapid phase (mediated primarily by pulmonary alveolar macrophages) and as a
long-term phase. The latter is a function of a several factors total lung burden of particles
and overloading, dissolution and absorption, and physical transport. The physical transport
processes are predominant in the conducting airways and are discussed more fully in the
section on tracheobronchial clearance. The long-term clearance processes are concerned
primarily with clearance from alveolar regions of the lung, and it is these processes that may
exhibit the greatest interspecific variability.
Pepelko (1987) noted that alveolar clearance rates were similar for dogs and humans
but at least four times more rapid in rats. Most other laboratory species were similar to rats
with the exception of guinea pigs which cleared aluminosilicate panicles at a rate similar to
that of humans (Snipes and McClellan, 1984 as reported in Pepelko. 1987).
A wide range of clearance rates for paniculate matter has been reponed for several
animal species and humans (Pepelko, 1987). A review of the data for pulmonary clearance
of various compounds by dogs and rats indicated, overall, that the clearance half-time (t%)
for dogs was longer than that for rats. Based on the first few months of pulmonary
clearance of aluminosilicate reponed by Snipes et al. (1983) for rats (t,^ of 35 d) and by
Bailey et al. (1982) for humans (t,^ of 420 d), it appears that the clearance rate for rats may
be considerably more rapid. In comparing clearance rates of humans with those of dogs, the
t1/2 values are similar for iron oxide, but for manganese oxide the t,^ for humans was
approximately twice that of dogs (Pepelko, 1987). Additionally, intraspecific vanability was
also pronounced, especially when comparing panicles of different chemical composition and
size. A review of clearance half-times for rats indicated that soluble particles were cleared
more rapidly than insoluble particles. For example, the t^ for the relatively soluble
compounds, vanadium pentoxide and barium sulfate, was 1.8 to 3.0 d, while the insoluble
titanium oxide and carbon core diesel particles had a t1yi ranging from 60 to 87 d.
May 1990
8-11
DRAFT-DO NOT QUOTE OR CITE
-------
8.4.2. Role of Alveolar Macrophages in the Clearance of Particulate
Matter
8A2.L Alveolar Macrophage-Mediated Clearance of Insoluble Particles
from the Lung
Alveolar macrophages (AM) constitute an important, first-line cellular defense
mechanism against inhaled particles that deposit in the peripheral regions of the lung. It is
well established that a host of diverse materials, including particulate diesel exhaust, are
phagocytized by the AM shortly after deposition (White and Garg, 1981; Lehnert and
Morrow, 1985), and, in this manner, such cell-contained particles are generally rapidly
sequestered from the extracellular fluid lining in the alveolar region and from the potentially
sensitive alveolar epithelial cells. In addition to this role in compartmentalizing particles
from other lung constituents, the AM are also believed to be prominently involved in
mediating the clearance of relatively insoluble panicles from the air spaces (Lehnert and
Morrow, 1985). Although the details of the actual process have not been delineated, AM
with their particulate burdens gain access and become coupled to the "mucociliary escalator"
and are subsequently transported from the lung via the conducting airways. While
circumstantial in nature, numerous lines of evidence indicate that such AM-mediated particle
clearance is normally the predominant mechanism by which relatively insoluble particles are
removed from the lungs (Lehnert and Morrow. 1985; Powdrill et al.. 1989: Gibb and
Morrow, 1962; Ferin, 1982).
The removal characteristics for panicles deposited in the lung's alveolar region have
been descriptively represented by numerous investigators as a multicompartment or
multicomponent process with each component following simple first-order kinetics (Snipes
and Gem, 1981; Snipes et aL, 1988; Lee et al., 1983). While often times suitable as a
mathematical approach for describing lung retention data, actual physiologic mechanisms
that form the bases for differing empirically resolvable compartments or components of
alveolar phase clearance [i.e., early, more rapid component(s)], and slower, longer term
component(s) have not been well characterized Regardless, it seems reasonable to assume
that the lung's population of AM must in some manner play a central role in the overall
May 1990
8-12
DRAFT-DO NOT QUOTE OR CITE
-------
retention characteristics of insoluble particles that deposit on the alveolar surface.
Delineation of this role under conditions of lung burdens commensurate with typical
environmental exposures to particles and following particulate exposure levels that lead to
diminutions of the AM-mediated clearance process should provide basic information on the
phenomenon of particle overload that has been associated with the development of lung
tumors following the deposition of high lung burdens of diesel (Mauderly et aL, 1987) as well
as other types of particles; for example, titanium dioxide (Lee et alM 1985), silica (Holland
et al., 1985), and chromium dioxide (Lee et aL, 1988). Additionally, lung dosimetric models
could incorporate this information for the purpose of extrapolating from existing animal data
involving diesel particle exposures to the human condition.
Lehnert and co-workers (1988,1989, and presently unpublished observations) recently
performed a study using laboratory rats to examine particle-AM relationships over the
course of alveolar clearance of low to high lung burdens of noncytotoxic polystyrene
microspheres (2.13 mm diam.) as an approach to obtain information on potential
AM-related mechanisms that form the underlying bases for kinetic patterns of alveolar
clearance as a function of particulate lung burdens. The intratracheally instilled lung
burdens studied were 1.6xl07 panicles (-85 /xg) for the low lung burden, 2.0x10s panicles
(-1.06 mg) for a medium range lung burden, and 6.8x10* particles (-3.6 mg) for the
highest lung burden. The retention kinetics of the panicles following the deposition of these
three lung burdens (Figure 8-3) were described using a two-companment model of the
following form: LB(I) = Ae *!t+Be -1, where LB(t) is the lung burden of panicles present
at a postinstillation time (t in d), and A and B are the coefficients associated with the
retention rates and respectively. The half-times (t,A) of both the early and latter
components of clearance were virtually identical following the deposition of the 85 ng and
1.06 mg lung burdens. The coefficients A and B were also closely similar. [Statistical
analyses revealed that the rate constants and the coefficients A and B for the low- and
medium-range lung burdens were not significantly different from one another]. For the
highest lung burden (3.6 mg), significant prolongations in the rates of the early, more rapid
component and the longer term, slower
May 1990
8-13
DRAFT-DO NOT QUOTE OR CITE
-------
II
10
m
Ml
S
2
lM>ll UTt *"*
IL
O
B
w
B
• TOt 41
im e* Mmeui ixti nu
n ciui
0» MltKUl ¦
10'
er
ee
r i«
1T8
DAYS AFTER MSTBXATION
U»H> CIUI
IOI
I
fc
»t II
••
am irm «mumoN
!
Ul BITi
UfAMD euii
its Of P*m'
it «•~*»
mot
MB O* HltlCUl Ml
T %4
IAVI 4ITII BITLUVM
Figure 8-3: Lung retention kinetics of 2.13 mm diam. polystyrene microspheres after
instillation into rat lungs. A: 1.6xl07 particles; B: 2.0x10" particles; C: 6.8x10s panicles.
May 1990
8-14
DRAFT-DO NOT QUOTE OR CITE
-------
component of alveolar clearance were found. The percentages of the particle burden
associated with the earlier and later components, however, were similar to those of the
lesser lung burdens. Particle-AM relationships (based on analyses of lavaged AM) during
the clearance of the low lung burden were expressed in terms of the estimated total numbers
of AM in the lung that contained a given load of the particles (Figure 8-4). Over a 176-d
period following the deposition of the low lung burden, the apparent overall rate(s) of
disappearance of AM with engulfed panicles was found to increase with increasing particle
burdens in the macrophages. For particle burdens up to 13 to 14 microspheres per AM, the
disappearance of the AM followed a biphasic pattern, which, like the lung retention data,
could be described as a two-component, negative exponential equation. A plot of the later
component exponents (d in Figure 8-5) relative to AM-particle burdens showed a linear
trend between the rate of later phase AM disappearance and particle burden per cell (i.e..
the magnitudes of the exponents decreased linearly as the number of microspheres defining
an AM category increased). Analyses of the early components of AM disappearance also
pointed to decreases in the magnitude of the exponents with increasing burdens in the AM
(6 in Figure 8-6). The proportions of the AM with a given particle burden that disappeared
from the total AM population via an apparent early component are graphically summarized
in Figure 8-7. Unlike the lower particle burden per AM categories, the disappearance rates
for AM in the higher particle categories were found to be best fitted with single exponential
functions, the exponents of which appeared to be more akin to the later component rates
of the other particle-AM categories than they were to the early component rates of AM with
less than 15 particles.
The simplest explanation for the "disappearance" characteristics of AM with differing
particle loads seen in the above study is that the varying rates by which AM seemingly vanish
from the lung's AM population during the course of alveolar clearance of the panicles
denotes the actual removal kinetics of these cells from the lung via thetracheobronchial
route. This interpretation, accordingly, implies that the translocation of AM from the
alveolar space by way of the conducting airways is fundamentally influenced by the particle
burdens in the cells so translocated. On the other hand, the
May 1990
8-15
DRAFT-DO NOT QUOTE OR CITE
-------
Ui
a
ui
o o
UI C
0 S
UI ^
& 2
1 O
H <
CO X
uj a
O
101
10' r
^ a
U. -9
O uj
se > s
ui 5 io9
s <
10* r
10
|3 L
10s
(9-10)
O NO PARTICLES
O 1-2 PARTICLES
A 9-4 PARTICLES
O S-8 PARTICLES ,
v 7-» PARTICLES
~ »-10 PARTICLES
¦ 11-12 PARTICLES
O 13*14 PARTICLES
10
10'
- 10'
104
(11-12)1 10
^(15.16):
• 1 s.l• PARTICLES 1(>16) J
o >it PARTICLES "•
' '
714
57
—-L_
85
176
10'
DAYS AFTER INSTILLATION
Figure 8-4: Estimated number of AM with an indicated particle load of microspheres over
a 174 d period following the instillation of 1.6x10' microspheres.
May 1990
8-16
DRAFT-DO NOT QUOTE OR CITE
-------
1*2 3*4 M 7-$ MO 11*12 13*14 1»*1» »1t
PARTICLE BURDEN/AM
Figure 8-5: Relationship between the slope (exponents) of later phase AM disappearance
relative to the panicle loads in the AM.
o
a.
x
May 1990
8-17
DRAFT-DO NOT QUOTE OR CITE
-------
PARTICLE BURDEN/AM
Figure 8-6: Relationship of the rates (exponents) of early
relative to particle loads in the AM.
component disappearance of AM
Mav 1990
8-18 DRAFT-DO NOT QUOTE OR CITE
-------
DAY 0
/
0.0
1-2 3-4 5-6 7-8 9-10 11-12 13-14
PARTICLE BURDEN CATEGORY
Figure 8-7: Proportions of AM with a given load of particles that disappeared from the total
AM population by the early disappearance components as of a theoretical day 0
postinstillation time and as of day 7 after the particles were administered.
May 1990
8-19
DRAFT-DO NOT QUOTE OR CITE
-------
disappearance of AM from the various particle categories as observed in the low lung
burden component of the above study may not exclusively reflect the translocation of AM
from the lung. The observations are also consistent with a gradual redistribution of retained
particles among the lung's AM concurrent with the removal of particle-containing AM via
the conducting airways perse. Consistent with a gradual release of previously endocytosed
particles, Lehnert and Morrow (1985) found that a small percentage of a low lung burden
of relatively insoluble iron oxide was not associated with lavaged, intact lung free cells over
the course of alveolar phase clearance. Potential mechanisms for such a redistribution of
retained particles among the AM include: (1) the in situ division of particle-containing AM
and allocations of the parent AM's particles to daughter cells, (2) the in situ autolysis of
particle-containing AM and the uptake of the panicles by other AM with pre-existing or no
panicle burdens, (3) the exocytosis of panicles by AM and the subsequent phagocytosis of
the particles by other AM, (4) the phagocytosis of effete panicle-containing AM by other
AM, and/or, perhaps, 5) the direct transfer of panicles from one AM to another.
Experimental suppon suggestive of at least some of these potential processes comes from
a variety of investigations involving AM and other endocytic cell types (Heppleston and
Young, 1974; Evans et aL, 1986; Aronson, 1963; Sandusky et aL, 1977; Heppleston, 1961;
Riley and Dean, 1978). Regardless, an outcome of the apparent differing disappearance
rates of AM from the panicle categories shown in Figure 8-4 resulted in a relative
redistnbution in the fractions of the retained lung burdens in the various AM-panide
categories over time (Figure 8-8). For example, only ~ 33 percent of the lung burden of
microspheres was in AM that contained 1 to 2 particles as of 7 d after they were instilled,
whereas by post-instillation day 176, this particle category accounted for ~ 65 percent of the
retained particles. Nevertheless, if the differing disappearance rates for AM as a function
of particle load is due at least in pan to particle burden-dependent differences in transloca-
tion rates of the AM out of the respiratory tract, the following findings from the above
investigation may point to an underlying basis for a slowing of lung clearance at high panicle
burdens in the lung: (1) as AM reach a given load of particles, a diminishing fraction
May 1990
8-20
DRAFT-DO NOT QUOTE OR CITE
-------
is
II
§«
en
x
411
o
Iu
60 r
50 r
a. 40
Oui
3 (9
• 1-S PARTICLES
O t-4 PARTICLES
O PARTICLES
¦ 7-t PARTICLES
# t-10 PARTICLES
O 1 1-12 PARTICLES
A 1S-14 PARTICLES
A IS-lt PARTICLES
V » It PARTICLES
67 85
DAYS AFTER INSTILLATION
iJ
m
P 1
176
Figure 8-8: Percentage distribution of the retained lung burden in the various particle-AM
categories over the course of lung clearance of the 1.6xl07 microsphere initial burden.
May 1990
8-21
DRAFT-DO NOT QUOTE OR CITE
-------
of them is removed via an early component, and (2) AM with high particle loads disappear
from the lung essentially by a slower, long-term component.
Similar patterns of AM disappearance as a function of particle load were observed
following the deposition of the 2.0x10* lung burden, Figure 8-9. That is, AM with the higher
initial burdens of particles decreased in numbers more rapidly following the deposition of
the particles than did AM with lesser particulate loads. Again, over time, increasing
fractions of the retained lung burden became progressively contained in AM with relatively
lower particulate burdens per cell, Figure 8-10. The possibility that particles are gradually
redistributed among the AM during the course of alveolar phase clearance is further
supponed by apparent increases in the number of AM containing 1 to 4 particles at the end
of the study relative to the number of AM with this particle load as of day 7 after the
panicles were instilled (Figure 8-9). In the case of the particle overload condition caused by
the instillation of the 6.8x10s microsphere burden, AM numbers in the higher particle
categories showed little change over the course of the 168-d postinstiilation period,
suggesting that the translocation of AM with the heaviest cellular burdens of panicles (Le^
> ~ 100 microspheres per AM) out of the respiratory tract via the tracheobronchial route
was compromised (Figure 8-11). This observation could be related to aggregates of
particle-laden AM in the alveoli (Figure 8-12) which, in turn, could contribute to diminished
lung clearance of the microspheres, and represent, at least in part, the particle "sequestra-
tion" compartment described by other investigators following the deposition of high lung
burdens of diesel and other types of panicles. Over time, greater percentages of the
retained lung burden were contained in AM with the highest particle burdens (Figure 8-13).
Further evidence of a gradual release of particles from AM during the course of diminished
lung clearance was again observed by increases in AM numbers in the lowest panicle
categoiy over the study period (Figure 8-9) and the appearance of particle-containing
polymorphonuclear leukocytes (Lehncrt et aL, 1989) and blood monocyte interstitial
macrophage-like cells among the lavaged free cell populations (Figure 8-14).
Although lung burdens in the above study were achieved by intratracheal instillations
and not by chronic aerosol exposures scenarios, which have led to tumor formation
May 1990
8-22
DRAFT-DO NOT QUOTE OR CITE
-------
S103
I 1 I
• •MflClIlN
c « »••• »niCkii
• •¦TaClll * NIIIClll
OMttiiieui ftit-MMitieui
•I MIIKUI VlfUfllflCKI
7 14
58 84
DAYS AFTER INSTILLATION
17*
Figure 8-9: Estimated numbers of AM with a given particle burden over the course of
clearance of 2.0x10s microspheres.
May 1990
8-23
DRAFT-DO NOT QUOTE OR CITE
-------
* 5
9 o
m -»
x o
g 5 20
o Z
z °
s s
< u
»• *
S s
10
» O
III Ul
S3
5 x
58
84
DAYS AFTER INSTILLATION
Figure 8-10: Percentage distributions of the retained lung burdens in the various
particle-AM categories during the clearance of the 2.0x10s initial lung burden of
microspheres.
May 1990
8-24
DRAFT-DO NOT QUOTE OR CITE
-------
•»
u >
10
5x10
55
ae
DAYS AFTER INSTILLATION
16C
Figure 8-11: Estimated numbers of AM in the indicated panicle-containing categories
following the deposition of 6.8x 108 microspheres into the rat's lung.
May 1990
8-25
DRAFT-DO NOT QUOTE OR CITE
-------
Figure 8-12: Electron micrograph of an aggregate of panicle filled AM in an alveolus on
day 106 following the deposition of 6.8x10s microspheres. Arrows point to Type II
pneumocytes, which appear to be present in greater than normal abundance. The
microspheres are also found in Type I epithelial cells.
May 1990
8-26
DRAFT-DO NOT QUOTE OR CITE
-------
7 14 56 86 lee
OATS AFTER INSTILLATION
Figure 8-13: Percentages of the retained lung burden contained in the vanous particle-AM
categones following the deposition of 6.8x10s microspheres.
May 1990 8-27 DRAFT-DO NOT QUOTE OR CITE
-------
in response to a variety of types of materials, the general outcomes may, in principle, be
applicable to lung clearance mechanisms at play in the latter condition. If so, two
mechanisms that may play major roles in lung retention kinetics following the deposition of
aerosolized particles are: (1) the removal of AM from the respiratory tract as a particle-de-
pendent process and (2) the gradual redistribution of particles among the lung's AM
population. Factors that may contribute to the phenomenen of particle release by AM,
including particle size, shape, composition, and surface features, have not been delineated.
Additionally, it remains obscure as to how the magnitude of particulate burdens in AM may
influence the particle redistribution process. Regardless, a lack in the fidelity between
engulfed panicles and AM that initially phagocytized the particles suggests that panicles
released by AM could be available for interactions with other extra-AM lung cells prior to
their endocytosis. Of relevance, the in situ disintegration of AM as one source of particle
release in the alveoli has been observed following exposure to paniculate diesel exhaust
(White and Garg, 1981). Experimental information about factors underlying panicle release
by AM, the kinetics of the release process, and the kinetics and efficiency of particle
re-uptake by phagocytes and other lung cells, inter alia, is required to establish the
cumulative fraction of a lung burden that is available to interact directly with other lung cells
over time. Such information would be of use for cellular (epithelial cells) dosimetry
estimates while furthering the current understanding of particle-AM interactions that may
contribute to the phenomenen of particle overload in the lung.
8.4.2.2. Translocations of Particles to Extra-Alveolar Macrophage
Compartment Sites
While the phagocytosis of particles by lung free cells and the mucociliary clearance
of the cells with their particulate burdens represent the most prominent mechanisms that
govern the fate of particles that deposit in the alveoli, other mechanisms exist that can
impact both on the retention characteristics of relatively insoluble particles in the lung as
well as on lung clearance pathways for the particles. One mechanism is the endocytosis
May 1990
8-28
DRAFT-DO NOT QUOTE OR CITE
-------
I
aS*.
>0
4
0
0
»
f»
•
Figure 8-14: Micrograph of cells lavaged from a lung 86 d after the deposition of 6.8x10s
microspheres.
May 1990
8-29
DRAFT-DO NOT QUOTE OR CITE
-------
of particles by the Type I pneumocytes (Sorokin and Brain, 1975; Adamson and Bowden,
1978; Adamson and Bowden, 1981) that normally cover >90 percent of the surface area of
the alveoli in the lungs of a variety of mammalian species (Crapo et al., 1983). This process
may be related to the size of particles that deposit in the lungs and the numbers of particles
that are deposited. Adamson and Bowden (1981) found with increasing loads of carbon
particles (0.03 mm) instilled into the lungs of mice, more free particles were observed in the
alveoli within a few d thereafter; the relative abundance of particles endocytosed by Type
I epithelial cells also increased with increasing lung burdens of the particles. These same
investigators demonstrated that instilled latex particles with a diameter of 0.1 mm also were
endocytosed by Type I pneumocytes when delivered at high lung burdens (4 mg), but latex
microspheres with a diameter of 1.0 mm (4 mg) were rarely observed to be engulfed by the
alveolar epithelial cells. Although the endocytic ability of Type I pneumocytes must have
an as yet to be determined upper panicle size limitation, the infrequent appearance of the
1.0 mm spheres in the Type I cells may have been primarily related to the numbers of
particles instilled and not necessarily to their size. A 4 mg burden of 0.1 mm diam latex
particles is equivalent to 8xl012 particles whereas a 4 mg burden of 1.0 mm diam latex
particles is composed of 8xl09 particles. Regardless, diesel particles with volume median
diameters between 0.05 and 0.3 mm (Frey and Corn, 1967; Kittleson et al., 1978) would be
expected to be within the size range for engulfment by Type I cells should suitable
encounters occur. Indeed, it has been demonstrated that diesel particles are endocytosed
by Type I pneumocytes in vivo (White and Garg, 1981). Adamson and Bowden (1981)
postulated that when acutely administered high lung burdens of particles exceed the lungs
capacity to recruit for additional phagocytes into the alveolar space compartment to
phagocytize the delivered burden, the presence of free particles increases the likelihood of
particle encounters with Type I cells. A hypothetical extension of this postulate is: some
fraction of particles which are deposited in the alveoli gain entry into Type I epithelial cells
even when deposited lung burdens are low or when the mechanism of particle phagocytosis
by AM is not overwhelmed. Alternatively, the process of particle phagocytosis (i.e., particle
encounters and phagocytosis), by AM seemingly would have to be kinetically accomplished
with 100 percent efficiency within a postdeposition time span that would preclude particle
May 1990 8-30 DRAFT-DO NOT QUOTE OR CITE
-------
uptake by the Type I cells. Unfortunately, information on the kinetics of particle endocytosis
by Type I cells relative to that by AM is scant. However, the above extension of the
Adamson and Bowden postulate is consistent with some existing observations. For example,
even when relatively low burdens of particulate agents are deposited in the lungs, some
fraction of the particles usually appear in the regional lymph nodes (Ferin and Feldstein,
1978; Lehnert, 1989); as will be discussed, endocytosis of particles by Type I pneumocytes
is an initial, early step involved in the passage of particles to the lymph nodes. Assuming
particle phagocytosis is not sufficiently rapid or perfectly efficient, increasing numbers of
particles would be expected to gain entry into the Type I epithelial cell compartment during
chronic aerosol exposures. Additionally, if panicles are released on a continual basis by
alveolar phagocytes that initially sequestered them after lung deposition, some fraction of
the "free" panicles so released could also undergo passage from the alveolar space into the
Type I cells.
As previously indicated, the endocytosis of particles by Type I pneumocytes represents
only the initial stage of a process that can lead to the accumulation of particles in the lung's
interstitial compartment and the subsequent translocation of particles to the regional lymph
nodes. As shown by Adamson and Bowden (1981), a vesicular transport mechanism in the
Type I pneumocyte can transfer particles from the air surface of the alveolar epithelium into
the lung's interstitium where particles may be phagocytized by interstitial macrophages or
they may remain in a "free" state for a poorly defined period of time that may be dependent
on the physicochemical characteristics of the particle. The lung's interstitial compartment,
accordingly, represents an anatomical site for the retention of particles in the lung. Whether
or not AM and perhaps PMN that have gained access to the alveolar space compartment
and phagocytize particles there also contribute to the particle translocation process into the
lung's interstitium remains a controversial issue. However, it is widely believed that once
AM, at least, assume occupancy in the alveoli, they do not reenter the lung's interstitium
(Roser, 1970; Adamson and Bowden, 1978; Brain et al., 1977). Adamson and Bowden
(1978,1981), for example, found no ultrastructural evidence that AM migrate into the lungs
intersitium after the acute deposition of a heavy burden of carbon or of latex microspheres
into the lungs of mice. Likewise, Lauweryns and Baert (1977) also found no morphologic
May 1990 8-31 DRAFT-DO NOT QUOTE OR CITE
-------
evidence for an AM migration through the alveolar epithelium after intratracheal instillations
of ferritin or carbon. On the other hand, Cony and co-workers (1984) intratracheal^
instilled radiolabeled AM into syngeneic guinea pigs and found radioactivity in the hilar
lymph nodes as well as the presence of radiolabeled cells in autoradiographic preparations
made from nodal cell suspensions obtained over a 24 to 72 h period after the instillations.
Lehnert and co-workers (1986) also examined mononuclear phagocytes from the
tracheobronchial lymph nodes of rats for the presence of particles after the intratracheal
instillation of polystyrene microspheres. These investigators found a disconcordance between
the frequency distributions of the microspheres in the nodal macrophages over a 30 d period
after the particles were administered relative to the particle distributions in lavaged AM (i.e.,
the nodal phagocytes contained lesser panicle cellular burdens than did the AM). These
latter findings suggest that if AM do translocate across the alveolar epithelial barrier, the
process may be limited by the magnitude of their particulate burdens. These collective
findings also suggest that at least some, if not most, of the macrophages observed to contain
particles in the lungs interstitium after particle deposition represent resident interstitial and
perhaps newly recruited blood monocytes that have phagocytized particles translocated
across the alveolar epithelial barrier (Adamson and Bowden, 1978; Bowden and Adamson,
1984). It should be pointed out, however, that the migration of AM into the interstitium
may be species dependent. Evidence that such migration of AM may importantly contribute
to the passage of particles to the interstitial compartment and also be involved in the subse-
quent translocation of particles to draining lymph nodes has been obtained with the dog
model (Harmsen et al., 1985).
The fate of particles once they enter the lung's interstitial spaces remains unclear.
Some particles, as previously indicated, are phagocytized by interstitial macrophages whereas
others apparently can remain in a free state in the interstitium for some time without being
engulfed by interstitial phagocytes. It is currently unknown what fraction of the interstitial
macrophages may subsequently enter the alveoli with their engulfed burdens of particles and
thereby contribute to the size of the resident AM population over the course of lung
clearance. Moreover, no investigations have been conducted to date to assess the influence
May 1990
8-32
DRAFT-DO NOT QUOTE OR CITE
-------
that an interstitial macrophage's burden of particles may have on its ability to migrate into
the alveolar space compartment.
It appears evident that at least some particles that gain entry into the interstitial
compartment can further translocate to the extrapulmonary regional lymph nodes. This
process apparently can involve the passage of free particles as well as particle containing
cells via lymphatic channels in the lungs (Harmsen et al., 1985; Ferin and Feldstein, 1978;
Lee et al., 1985). It is conceivable that the mobility of the interstitial macrophages could be
particle burden limited, and under conditions of high cellular burdens, a greater fraction of
particles that accumulate in the lymph nodes under conditions of high lung burdens may
reach these sites as free particles. Whatever the process, existing evidence indicates that
when lung burdens of panicles result in the panicle overload condition, panicles accumulate
both more rapidly and abundantly in lymph nodes that receive lymphatic drainage from the
lung (Ferin and Feldstein, 1978; Lee et al., 1985).
8.4.23. Some Potential Mechanisms that May Underlie the Appearance
of an Alveolar Macrophage "Sequestration" Compartment for Particles
During Particle Overload
Animal studies have revealed that impairment of alveolar clearance can occur
following chronic exposure to particulate diesel exhaust (Griffis et al., 1983; Wolff et al.,
1987; Vostal, 1982; Lee et al., 1983). Several other experimental animal studies involving
chronic aerosol exposures to a variety of other diverse materials have also demonstrated that
the lung retention of panicles can be increased following the deposition of excessive lung
burdens of panicles (Lee et al., 1986; Lee et al., 1988; Ferin and Feldstein, 1978). In that
high lung burdens of different types of panicles result in diminutions of normal lung
clearance kinetics, or result in what is now called panicle overloading, this outcome appears
to be more related to the mass and/or volume of panicles in the lung than to the nature of
the particles perse. Regardless, as pointed out by Morrow (1988), particle overloading in
the lung modifies the dosimetry for particles in the lung and thereby can alter toxicologic
responses to the particles. Unfortunately, information on particle burden and particle
May 1990
8-33
DRAFT-DO NOT QUOTE OR CITE
-------
overload relationships in the human lung are not available as with experimental animal
models.
The mechanisms that underlie particle-load dependent retardations in the rate of
removal of particles from the lung and the corresponding functional appearance of an
abnormally slow clearing compartment, or particle "sequestration" compartment, are likely
numerous. As previously discussed, one potential site for particle sequestration is the
extra-macrophagic containment of particles in the lung's Type I epithelial cells. Information
on the retention kinetics for particles in the Type I epithelial cells is nonexistent, and no
information as to how the vesicular transport of particles across the Type I cell may be
exhausted or otherwise modified during particle overload is currently available. Also, no
morphometric analyses have been performed to date to estimate what fraction of a retained
lung burden may be contained in the lung's Type I pneumocyte population during the
condition of lung overloading with particulate diesel exhaust or other types of particles.
Another anatomical region in the lung that may be a slowly clearing site is the lung's
interstitial compartment. Little is known about the kinetics of removal of "free" particles or
particle-containing macrophages from the lung's interstitial spaces or what fraction of a
retained burden of particles is contained in the lung's interstitium during particle overload.
The gradual accumulation of particles in the regional lymph nodes and the appearance of
particles and cells with associated particles in lymphatic channels and in peribronchial and
perivascular lymphoid tissue (Lee et al., 1985; White and Garg, 1981) suggest that the
mobilization of particles from interstitial sites via local lymphatics is a continual process.
Indeed, it is clear from histologic observations of the lungs of animals chronically
exposed to diesel particles that Type I pneumocytes, the interstitium, the lymphatic channels,
and pulmonary lymphoid tissues are sites that could represent subcompartments of a more
generalized particle sequestration compartment. Strom (1984) has reported that lesser
percentages of retained lung burdens of diesel particles are lavageable from lungs of animals
chronically exposed to diesel particles relative to those percentages of the retained lung
burdens that are lavageable following acute exposures. The basis for this observation may,
at least in part, be related to the relative fraction of a chronically administered burden of
particulate diesel that is contained in extra-AM sites that are not susceptible to lavage.
May 1990 8-34 DRAFT-DO NOT QUOTE OR CITE
-------
Although the above mentioned sites must be considered as potential contributors to
the diminutions of lung clearance during particle overload, disturbances in particle associated
AM-mediated clearance is undoubtedly the predominant cause inasmuch as the AM are the
primary reservoirs of containment of deposited particles. The mechanism underlying a
failure of AM to translocate from the alveolar space compartment in lungs with high
particulate burdens remains obscure, although a hypothesis concerning the process have
recently been offered. Morrow (1988) has proposed that the condition of dust overloading
in the lungs is caused by a loss in the mobility of particle engorged AM and that such an
impediment is related to the cumulative volumetric load of particles in the AM. Thus,
according to this hypothesis, the essentially continuous phagocytosis of particles by AM
during chronic exposures to high mass concentrations of particles gradually results in a
volume load in the AM that in some manner results in their immobilization. Morrow (1988)
has further estimated that the clearance of an AM is impaired when the particulate burden
in the AM is of a volumetric size equivalent to — 60 percent of the normal volume of the
AM. Hence, a primary determinate of an incapacitation of AM-mediated particle clearance
may not be due to cumulatively retained lung mass burden, but rather, the emergence of the
particle overload condition and the extent of diminished lung clearance may more directly
relate to the volumetric load of a retained burden in the lung and its distribution among
members of the lung's AM population. Some recent evidence is consistent with the
hypothesis that an AM's volume load of particles can decreased the removal of AM from
the alveoli. Specifically, Oberdorster and co-workers (1988) assessed the alveolar clearance
of smaller (33 mm dia.) and larger (10.3 mm dia.) polystyrene particles, the latter ofwhich
are volumetrically equivalent to — 60 percent of the average normal volume of an AM, after
they were intratracheally instilled into the lungs of rats. Whereas both sizes of particles were
found to be phagocytized by AM as of one d after deposition, only minimal lung clearance
of the larger particles was observed over an ~ 200-d postinstillation period.
Other processes may also be involved in preventing AM from leaving the alveolar
compartment with their phagocytized particulate burdens under conditions of particle
overload in the lung. Ousters or aggregates of particle-laden AM in the alveoli are typically
found in the lungs of experimental animals that have received high lung burdens of a variety
May 1990 8-35 DRAFT-DO NOT QUOTE OR CITE
-------
of types of panicles (Lee et ah, 1985), including particulate diesel exhaust (White and Garg,
1981; McClellan et al., 1982), see Figure 8-14. The aggregation of AM may represent a
major, if not the primary explanation for the reduced clearance of particle-laden AM during
particle overload. The raechanism(s) responsible for the agglutination of the AM has not
been elucidated to date. Conceivably, the clustering of AM may be related to the Type II
hyperplasia observed in lungs that have received high lung burdens of panicle and an
abnormal abundance of Type II cell-derived products, phospholipids, in the alveoli; after
shon term exposure to diesel exhaust, at least, lavageable phospholipid has been found to
increase (Wright, 1986). Support for this possibility comes from a study by Ferin (1982) in
which the lung clearance of Ti02 was shown to cease when a condition of phospholipidosis
was experimentally induced. Also, the appearance of large "foam" cells that are observed
following the deposition of high lung burdens of a variety of panicle types is consistent with
this possibility. Another underlying factor that may be involved in the clustering effect is a
continual alteration in the permeability of the alveolar epithelial barrier under conditions of
particle overload (Creutzenberg et al., 1989). In this case, "sticky" factors derived from the
blood compartment, such as fibronectin, could contribute to AM adhering to one another.
Still another potential mechanism for the agglutination of AM could be related to cytokines
released from lymphocytes (Garcia-Moreno and Myrvik, 1977). Some evidence, however,
suggests that lymphokines (e.g., macrophage migration inhibition factor) may not be the
dominant mechanism responsible for AM aggregation in the air spaces. While lung lympho-
cytes do increase following diesel exposure (Strom, 1984), AM aggregation following TiO:
deposition in the lung occurs with the general absence of excessive lung lymphocytes (Lee
et al., 1986). Whatever the underlying mechanism(s) for the AM agglutinating response, it
is worthy to note that AM lavaged from the lungs of diesel exhaust-exposed animals continue
to demonstrate a propensity to aggregate (Strom, 1984). This observation suggests that the
surface characteristics of the AM are fundamentally altered in a manner that promotes their
adherence to one another in the alveolar region and that AM aggregation may not simply
be due directly to their abundant accumulation as a result of a particulate volumetric load
immobilization.
May 1990
8-36
DRAFT-DO NOT QUOTE OR CITE
-------
8.4.3. Summary: Pulmonary Clearance of Diesel Exhaust Particulate Matter
The animal studies examined indicate that exposure to diesel exhaust unequivocally
increases lung particle burden in a dose-dependent fashion with high exposure levels (3.5 to
7.0 mg/m3) resulting in compromised clearance of particulate matter. An initial, relatively
rapid clearance phase is mediated by mucociliary transport. This process is especially
important for removal of particles from the major airways and terminal bronchioles. A
slower clearance phase involving aggregates of alveolar macrophages is very important in the
pulmonary clearance of particulate matter. The role of alveolar macrophage activity and the
potential for overloading of this function are of concern relative to removal of particles from
the lung. The development of lung tumors in experimental laboratory animals following
chronic exposures to particulate diesel exhaust occurs under conditions in which alveolar
macrophage-mediated particle clearance from the lung is compromised. Tumors have also
been found to develop with other types of particles when this clearance mechanism is
diminished. Thus, reductions in the functional activity of the lung's alveolar macrophage
population in the clearance process appears to be intimately related to the carcinogenic
response to high lung burdens of particles, generally. Findings that tumors develop in the
lungs of laboratory rats at lesser lung mass or volume burdens of diesel particles than with
a substance like titanium dioxide suggest that the carcinogenic response, however, is not
exclusively related to an over abundance of particles in the lungs per se. Most likely, the
organic components on diesel panicles, many of which have demonstrated carcinogenic
activity, are importantly involved in the development of lung tumors. The lung's pulmonary
macrophages, which phagocytize deposited diesel particles, probably participate in the
gradual in situ extraction and metabolism of procarcinogens on diesel particles. The
extraction of potentially carcinogenic compounds from particles is addressed more fully in
the following scetion. Additionally, the normal tumoricidal activities of the pulmonary
macrophages may be compromised upon interaction with excessive numbers of diesel
particles, and diesel particle-macrophage interactions could lead to the generation of reactive
oxygen species that have been shown to be at least mutagenic. Caution must be exercised
in extrapolating observations made in animal models to the human condition. Processes and
May 1990
8-37
DRAFT-DO NOT QUOTE OR CITE
-------
potential mechanisms discussed herein have largely been derived from animal data. Further
research is required to determine how the activities of human pulmonary macrophages in
response to particulate diesel exhaust compare with pulmonary macrophages from
experimental animals. Most importantly, valid dosimetry for the human condition requires
the elucidation of the underlying mechanistic bases involved in the development of lung
tumors following chronic particulate diesel exhaust exposure.
The extrapolation of the reported pulmonary retention and clearance data to
humans appears to be tenuous at present due to the high exposure levels used in the animal
studies, and interspecific variability in physiological function and anatomy of the respiratory
tract which may alter the threshold for clearance overloading. Additionally, humans are
exposed to much lower particle concentrations, and thus not likely to exhibit impaired
clearance. Partially offsetting the impaired clearance during lung overloading in rats is the
normally greater long-term clearance times for humans. Bailey et al. (1982) reported t,^
values of nearly 1 yr.
&5. DESORPTION OF CONSTITUENTS FROM DIESEL EXHAUST
PARTICLES
8.5.1. Bioavailability of Agents Adsorbed to Diesel Exhaust Particles
Bioavailability of toxic organics in diesel exhaust appears to be dependent on
the dissociation of these compounds from diesel particles (Vostal, 1983) and on the
persistence of the particulate matter which is a function of deposition and pulmonary
clearance. That toxicity of the organic compounds is increased by the presence of particles
has been demonstrated for benzo[a]pyrene (Saffiotti et al., 1964; Creasia and Nettesheim,
1974; Henry et al., 1975). These studies indicated that a higher incidence of lung carcinomas
was observed when B[a]P was adsorbed onto iron oxide particles than when intratracheally
instilled in pure form. However, the mechanism of this greater toxicity has not been
elucidated. Several studies have provided evidence that polycyclic aromatic hydrocarbons
(PAHs) including 1-nitropyrene and benzo[a]pyrene may be released from diesel exhaust
May 1990
8-38
DRAFT-DO NOT QUOTE OR CITE
-------
particles, thus demonstrating the potential bioavailability of carcinogenic agents from diesel
exhaust. Pepelko and Peirano (1983) in summarizing U.S. EPA studies on health effects of
diesel engine emissions described a study designed to assess the effect of inhaled diesel
exhaust upon the bioavailability, distribution, metabolism, and excretion of intratracheal
instilled B[a]P. Results of this phase of the study indicated that exposure to diesel exhaust
had no significant effect on the uptake or distribution of the administered B[a]P. More
recent studies, described below, have also studied the relationship between the panicle phase
of diesel exhaust and the bioavailability of the associated organics.
Sun et al. (1984) provided additional information comparing the disposition of
particle-adsorbed B[a]P (0.1 percent by weight) and pure B[a]P following 30 min nose-only
inhalation by F344 rats. Long-term lung retention (percentage retained after 7 d) of
particle-adsorbed 3H-B[a]P was approximately 230-fold greater than that for pure 3H-B[a]P.
Pulmonary clearance of particle-associated 3H was biphasic with an initial t1/& of 1 h and a
second phase tlyi of 18 d, the latter representing clearance of 50 percent of the initially
deposited radiolabel. Clearance of pure B[a]P aerosol was >99 percent within 2 h and was
apparently due to pulmonary and mucous membrane absorption into the blood rather than
by mucociliary clearance and subsequent ingestion (Sun et al., 1982). Of the radiolabel
retained in the lungs, 65-76 percent was B[a]P, 13 to 17 percent was B[a]P-phenol, and 5 to
18 percent was B[a]P-quinone. Tissue distribution and excretion of 3H were similar to
previous studies (Sun et al., 1982) where B[a]P was adsorbed onto gallium oxide (Ga:03)
panicles, thus suggestmg that adsorption to the carbonaceous diesel exhaust panicles
prolongs the release of B[a]P.
Although the Sun et al. (1984) study demonstrated the bioavailability and subsequent
biotransformation of B[a]P to several metabolites, the epoxide intermediates known to be
carcinogenic (Sims et al., 1974; Slaga et al., 1976) were not identified. However, B[a]P-
phenol metabolites are reponed to be mutagenic (Glatt and Oesch, 1976; Wislocki et al.,
1976; Wood et al., 1976).
A companion study (Bond et al., 1986) examined the biological fate of UC-1-
nitropyrene (,4C-NP) both in pure form and adsorbed to diesel exhaust panicles following
1 h nose-only inhalation by male F344 rats. Concentrations of UC-NP ranged from 50 to
May 1990 8-39 DRAFT-DO NOT QUOTE OR CITE
-------
1100 ng/L air, and diesel particle concentrations, where utilized, ranged from 3.7 to 6.1 jxg/L
air. The results indicated that long-term lung retention of UC-NP adsorbed onto diesel
exhaust particles was 80-fold greater than that for pure ,4C-NP. These studies also
demonstrated that adsorption onto the diesel exhaust particles prolongs the release of the
PAHs.
In these studies retention of ,4C-B[a]P and ,4C-NP was less than that for diesel
particles alone, thus release from the particles was occurring. The contention that
adsorption to particles affects the bioavailability of diesel exhaust PAHs is affirmed by these
studies.
Bond (1987) summarized the importance of particle association in the clearance of
adsorbed organic compounds, and emphasized the necessity for understanding the physical
and chemical interactions between the particle and these organic compounds. Also
emphasized was the significance of alveolar macrophage activity relative to deep lung
clearance of panicles and potential for metabolism of the organics to reactive intermediates.
Ball and King (1985) studied the disposition and metabolism of 14C-labeled 1-
nitropyrene (>99.9 percent purity) coated onto diesel exhaust particles. A single dose of 1-
NP (380 /xg/g particle) was intratracheal^ administered (in 0.2-mL buffered saline) at a
particle dose of 5 mg/rat. Another group of rats (number not specified) received the labeled
1-nitropyrene in 0.5 mL intragastrically. Additional groups of AGUS strain rats raised
conventionally or germ-free received intraperitoneal injections of 14C-NP to determine the
role of gastrointestinal flora on the metabolism of 1-NP.
Regardless of the route of administration, greater than 50 percent of the ,4C was
excreted within the first 24 h, 20-30 percent appeared in the urine, and 40 to 60 percent was
excreted in the feces. Intestinal flora was essential for the formation of the metabolite, 6-
hydroxy-N-acetyl- 1-aminopyrene, which was mutagenic (with S9) in Salmonella typhimurium
strain TA98. The intestinal flora were also required for making first-pass biliary metabolites
available (by conjugation-hydrolysis reactions) for enterohepatic circulation. The MC
excretion pattern for intratracheal^ instilled compound was nearly identical to that of orally
administered compound. For these animals, 16 to 38 percent of the unexcreted dose was
in the gastrointestinal tract and 5 to 8 percent remained in the lungs. Traces of radiolabel
May 1990 8-40 DRAFT-DO NOT QUOTE OR CITE
-------
were detected in the trachea and esophagus. Five to 12 percent of the radiolabel in the lung
copurified with the protein fraction, thus indicating protein binding of the 1-NP-derived 14C.
However, the corresponding DNA fraction contained no l4C above background levels. This
study demonstrated that the excretion kinetics and the metabolic profiles for 1-NP are
similar for oral and pulmonary routes of administration.
As a means of assessing bioavailability of particle-associated compounds, a
comparison of the mutagencity of compounds extracted from diesel exhaust particles by
dichloromethane or by biological fluids (serum, lavage fluid) was conducted by Brooks et al.
(1981). Diesel exhaust particles were obtained from a 5.7-L engine operated at idle and
from a 2.1-L engine operating on the FET cycle. Particles from these sources were
extracted using dichloromethane or were subjected to incubation with the biological fluids
for 6, 24, 48, 72, or 120 h at 37 *C. Mutagenicity was assessed using S. typhimurium strains
TA98 and TA100 with and without S9. For the diesel exhaust particles incubated with
biological fluids, there was no increase in mutagenic activity with increased incubation time.
Based on the concentration response curves for the mutagenicity tests, the biological fluids
extracted only 3 to 10 percent of the activity extracted by dichloromethane. The authors
qualified the findings, however, by indicating that the time frame of the experiments does
not reflect that of the long-term pulmonary clearance processes, and that the biological fluids
may bind or inactivate the mutagenic compounds in vitro, but interactions with cells in vivo
may still occur. The binding of compounds by biological fluids may not actually inactivate
the compounds, but rather allow transport to target cells and subsequent activation. The
potential for in vivo extraction, detoxification or activation of particle-associated compounds
makes predictive use of in vitro systems tenuous until additional research combining in vivo
and in vitro tests is conducted.
Similar findings were reported by King et al. (1981). In this study, lung lavage fluid
and lung cytosol fluid extracts of diesel exhaust panicles were not mutagenic. Serum extracts
of diesel particles did exhibit some mutagenic activity but this was considerably less than that
for organic solvent extracts. Furthermore, the mutagenic activity of the solvent extract was
significantly reduced when combined with serum or lung cytosol fluid suggesting protein
binding or biotransformation of the mutagenic components.
May 1990 8-41 DRAFT-DO NOT QUOTE OR CITE
-------
The involvement of alveolar and interstitial macrophages in determining the
biovailability of adsorbed organics is demonstrated by several studies. As indicated by
Lehnert (1988) and Godleski et al. (1988), translocation and phagolysosomal solubilization
of particles may be instrumental in affecting pulmonary clearance of the particles and
biovailability of the particle-associated constituents.
Bond et aL (1984) provided evidence that alveolar macrophages from beagle dogs
metabolized B[a]P in solution as well as B[a]P coated on diesel exhaust particles. Culture
media incubations produced B[a]P-9,10-diol and B[a]P-7,8-diol and incubations using the
pulmonary macrophages produced B[a]P-7,8-diol and B[a]P-4,5-diol. Both incubation
mixtures formed small quantities of B[a]P phenols and quinones. The significance of these
findings is the demonstration of formation by pulmonary macrophages of B[a]P metabolites
that are proposed proximate carcinogens and that these metabolites are derived from diesel
exhaust particle-associated B[a]P. These findings confirmed an earlier report by King et al.
(1983) indicating that pulmonary macrophages metabolized mutagenic nitroaromatics found
in diesel exhaust particles.
8.5.2. Extraction of Carcinogens from Particles by Alveolar Macrophages
and Other Cell Types
High lung burdens of chronically administered particles such as titanium dioxide
(TiO;) have been shown to produce lung tumors in laboratory animals (Lee et al.. 1986).
Experimental results such as these, as well as other findings (Ishinishi, 1986), suggest that
an over abundance of particles in the lung, however innocuous the particle type, can lead
to the development of lung tumors. However, some existing evidence suggests that the
development of tumors in diesel exhaust-exposed animals (Mauderly et al., 1987) may not
be related exclusively to a high lung burden of particles per se. Specifically, the lung
burdens of Ti02 that have been associated with the production of lung tumors in laboratory
rats (Lee et al., 1986) substantially exceed lung burdens of diesel particles that have been
associated with tumor development (Mauderly et al., 1987). Still other data suggests that
the quality (e.g., physicochemical characteristics), of a particle, or, in other words, a pure
May 1990
8-42
DRAFT-DO NOT QUOTE OR CITE
-------
particle effect may be of primary importance in producing lung tumors. For example,
Kawabata et al (1986) found that the intratracheal instillation of repeated doses of carbon
black was as effective in producing lung tumors as was the same mass of instilled particulate
diesel exhaust. Thus, it remains possible that a particle composed of carbon may be
inherently more carcinogenic than a particle with the physicochemical characteristics of a
material like TiOy Even so, approximately 10 to 20 percent of the particulate mass of diesel
particles is extractable by organic solvents (Frey and Corn, 1967). Among the extractable
compounds are numerous mutagenic and carcinogenic chemicals; for example, polycyclic
aromatic hydrocarbons (PAH) (Barth and Blacker, 1974; Huisingh et al., 1978; Lee et al.,
1980). Experimental findings that metabolites of benzo(a)pyrene (B[a]P) and DNA adducts
are formed in the lungs of rats following exposure to diesel exhaust and B[a]P adsorbed onto
carbon suggest that the PAH in diesel particles may contribute significantly to the
carcinogenic response (Wolff et al., 1989). Thus, the extraction of procarcinogenic
compounds from diesel particles that deposit in the lung may be one underlying factor in the
development of lung tumors associated with the pulmonary deposition of diesel particles.
Such a process would potentially provide a means for the bioavailability of cancer producing
agents to sensitive target cells. One mechanism by which carcinogenic agents could be
eluted from diesel particles is by extraction into extracellular fluids lining the respiratory
tract. Siak et al (1980) assessed the mutagenicity of material extracted from diesel particles
by bovine serum albumin in solution, simulated lung surfactant, fetal calf serum (FCS), and
physiologic saline. Only FCS was found to extract some mutagenic activity from the diesel
particles. These investigators concluded that the mutagens in diesel particles would not be
readily available in vivo. Similarly, Brooks et al. (1981) found only a small amount of
mutagenic activity in serum or lavage fluid that was incubated with diesel exhaust particles.
The conclusion that the mutagens in diesel particles would not be readily available in vivo
(Siak et al., 1980) via particle extraction by the lung's extracellular lining fluid (ELF) lacks
definitive experimental proof. The ELF is a complex mixture of constituents that
undoubtedly have a broad range in hydrophobicity (George and Hook, 1984; "Wright and
Clements, 1987), and it fundamentally differs from serum in terms of chemical composition
(Gurley et al., 1988). Moreover, assessments of the ability of lavage fluids, which actually
May 1990 8-43 DRAFT-DO NOT QUOTE OR CITE
-------
represent substantially diluted ELF, to extract mutagenic activity from diesel particles clearly
do not reflect the in vivo condition. However, as will be discussed hereafter, some existing
evidence does suggest that extraction of organic constituents from diesel particles by ELF
prior to particle phagocytosis by lung phagocytes may be of relatively minor importance in
mobilizing carcinogenic agents from the particles in vivo.
Another, more likely mechanism by which organic carcinogens, (e.g., polycyclic
aromatic hydrocarbons) may be extracted from diesel particles in the lung is by particle
dissolution within the phagosomes of alveolar macrophages (AM). Dissolution in
macrophages presupposes that the particles are internalized by these phagocytes. Normally,
the lung's AM can phagocytize particles, including diesel panicles, that deposit in the
peripheral air spaces within hours (Lehnert and Morrow, 1985; White and Garg, 1981).
Some in vitro evidence (Strom, 1984) suggests that AM in the lungs of rats chronically
exposed to diesel exhaust may not be markedly impaired in terms of phagocytic activity. On
the other hand, Castranova et al. (1985) contend that chronic exposure to diesel exhaust
depresses AM phagocytic activity, this latter interpretation, however, was based on the
results of a chemiluminescence assay and not on direct quantitation of phagocytosis per se.
Regardless, histologic observations of the lungs of animals subchronically or chronically
exposed to diesel particles indicate that most of the retained particles are sequestered in
particle-laden macrophages (Wiester et al., 1980; Plopper et al., 1983). Thus, a mechanism
by which diesel particles can gain entry into the AM, and subsequently undergo dissolution
therein, is available, at least under some exposure conditions. However, as high lung
burdens of diesel particles are attained, such as during chronic, high atmospheric mass
concentration exposures to diesel particles, AM that become heavily laden may reach their
phagocytic capacity, and in this case, the abilities of the AM to further phagocytize deposited
particles may be reduced. Additionally, the decreased mobility of AM that are filled with
particles may diminish their migratory activities (Morrow, 1988; Lehnert et aL, 1989) on the
alveolar surface and their migration from one alveolus to another (Ferin, 1982).
Furthermore, AM that reside as aggregates in the alveoli following chronic diesel exposure
(White and Garg, 1981) would appear to have essentially no migratory activity. Decreases
in particle-macrophage interactions would be an expected outcome of reduced levels of AM
May 1990 8-44 DRAFT-DO NOT QUOTE OR CITE
-------
migratory activity. Under these conditions, an increasing fraction of deposited particles
could escape the phagocytic mechanism and thereby be relatively more abundantly available
over time in the ELF prior to: (1) their removal from the lung by extra-macrophagic
clearance via the tracheobronchial route, (2) their subsequent engulfment by newly recruited
phagocytes, and/or 3) their engulfment by alveolar epithelial cells (i.e., Type I pneumocytes).
Specific details about the physicochemical conditions of the intraphagolysosomal
environment where particle dissolution in AM presumably occurs in vivo have not been well
characterized. However, it is known that the phagolysosomes constitute an acidic (pH 4 to
5) compartment in macrophages (Nilsen et al., 1988; Ohkuma and Poole, 1978). The
relatively low pH in the phagolysosomes has been associated with the dissolution of some
types of inorganic panicles (e.g., some metals) by macrophages (Marafante et al., 1987;
Lundborg et al., 1984), but quantitative information as to how organic constituents of diesel
particles (e.g., B[a]P) may be extracted in the phagolysosomes is limited (Bond et al., 1983).
Possibly, the containment of diesel particles in macrophages may in fact limit the
solubilization of organic components while promoting their retention in the lung, with the
end result being a prolonged exposure of the respiratory epithelium to the gradual release
of carcinogenic agents. For example, Creasia and co-workers (1976) found that when
crystalline B[a]P was instilled into the lungs of mice, it was removed from the respiratory
tract with a half time (t1/a) of ~ 1.5 h, but when the B[a]P was adsorbed to 0.5 to 1.0 fim
carbon panicles (a panicle size range that would be expected to be readily phagocvtized by
resident AM shortly after deposition), its t1yi in the respiratory tract increased to - 36 h.
Hence, the adsorption of BaP to the carbon particles increased the lung retention of the
B[a]P greater than 20-fold. Similar results have also been obtained with BaP adsorbed onto
other particle types, including insoluble Ga203 (Sun et al., 1982), insoluble ferric oxide
(Saffiotti et al., 1964), and diesel particles (Sun et al., 1984). Consistent with a gradual
elution of the B[a]P in AM, Creasia and co-workers (1976) also found that the removal of
B[a]P when bound to the carbon was faster than the lung clearance of carbon particles only,
which had a clearance t% of about 7 d. When the B[a]P was adsorbed onto carbon particles
larger than would be expected to be easily phagocytized by AM (15 to 30 nm), the rates of
elimination of the B[a]P and the larger particles from the lung were virtually identical. This
May 1990 8-45 DRAFT-DO NOT QUOTE OR CITE
-------
latter finding suggests that the elution of organics like B[a]P from particles by ELF may play
a less important role in the extraction process than do the lung phagocytes.
Several investigators have shown that crystalline B[a]P has a relatively low
carcinogenicity (Henry and Kaufman, 1973; Kuschner, 1968; Henry et al., 1973; Saffiotti et
al., 1964), which may be due to its rapid elimination when unbound to particles. Henry and
Kaufman (1974) also reported that the lung retention of B[a]P is prolonged as the carrier
particle size is increased, and such increases in particle-associated retention of B[a]P have
been correlated with enhanced carcinogenicity (Farrell and Davis, 1974). Such findings have
been interpreted to suggest that the retardation in the lung clearance of an organic
carcinogen is one of the major mechanisms by which particles can enhance their tumor-
genicity (Creasia et al., 1976). Yet, Farrell and Davis (1974) found that a BaP-carbon
particle combination in the particle size range of 0.5-1.0 mm was more tumongenic than
when the B[a]P was adsorbed to 15 to 30 nm size carbon panicles; again, panicles in this
latter larger size range would not be readily endocytosed by resident AM, and, accordingly,
would essentially be constantly in the presence of ELF. This observation indicates that the
pulmonary retention of B[a]P per se is not a firm predictor of tumorigenicity. Rather, the
above lines of information suggest that relatively continual availability of B[a]P via an
extraction mechanism (i.e., extraction in the phagosomes of lung free cells) may be a primary
determinant of carcinogenicity. As argued by Saffiotti et al. (1972), a carcinogen can be
much more effective in producing a tumongenic response when given in small divided doses
over a period of time than when it is delivered as a single equivalent dose.
Quantitative predictions as to how readily carcinogenic organics may be extracted
from diesel particles in human lung cannot be reliably extrapolated from existing animal
data. As shown by Creasia et al (1976), the polycyclic aromatic hydrocarbon B[a]P
adsorbed onto carbon particles is removed from the mouse lung at a rate that is — 5 times
faster than the rate of clearance of the carbon particles only. For the rat, Sun and
co-investigators (1984) have reported that the t% for the lung clearance of B[a]P adsorbed
onto diesel particles over a period of time consistent with alveolar phase clearance was —18
d. This latter value is similar to the t1/a for the removal of innocuous panicles from the rat's
lung during the early, more rapid component of alveolar phase clearance (Snipes et al., 1988;
May 1990 8-46 DRAFT-DO NOT QUOTE OR CITE
-------
Snipes et al., 1981; Ferin, 1982; Lehnert, 1988; Lehnert et al., 1989; Lehnert, 1989). These
findings collectively suggest that the extraction of organic components from carrier particles
by AM may differ among species, although differences in the lung burdens administered in
the above investigations may have influenced their outcomes.
Comparative studies of human AM and AM from other species used in previous
diesel exposure studies as to their relative abilities to extract carcinogens from diesel
particles may provide useful information for human dosimetry estimates. Such investigations
would ideally examine how the extraction process relates to the level of diesel particulate
load in the AM. A subsequent dosimetry model would take into account, then, the kinetics
of carcinogen mobilization by AM as a function of particulate burden, the kinetics of the
metabolic fates of procarcinogens associated with diesel particles a la Bond et al (1984), and
existing information that relates the amount of a material like B[a]P in the respiratory tract
to its rate of removal (Medinsky and Kampcik, 1985). Additionally, studies directed to the
relative abilities of human polymorphonuclear leukocytes (PMN) to extract organic
constituents from diesel particles are also warranted in that diesel-containing PMN can
become a major component of the lung's free cell population following chronic exposure to
particulate diesel exhaust (Strom, 1984; White and Garg, 1981). In this same regard, further
information as to how other lung cell types that can engulf deposited diesel particles (i.e.,
Type I pneumocytes, interstitial macrophages, and newly recruited blood monocytes) may
kinetically extract procarcinogenic agents from the particles is required in order to better
establish existing cellular bases for an extraction process. Such data could then be used in
conjunction with cell type/estimated lung cell number data obtained from morphometric
analyses of lungs exposed to diesel particles as an approach to quantitatively model potential
cellular sites of extractions of organics from diesel particles. This approach, however, is
further complicated by the likelihood that some fraction of the particulate burdens in lung
cells may include particles that were previously contained in still other cells; hence,
extractable organics in this fraction presumably would be less than that for particles that had
no history of prior cell containment.
It should also be pointed out that studies designed to examine the extraction of
organics by AM have generally focused on B[a]P as a representative procarcinogen
May 1990 8-47 DRAFT-DO NOT QUOTE OR CITE
-------
associated with diesel particles. Numerous other agents with carcinogenic activity are also
associated with diesel particles; these chemical constituents may be extracted from
particulate diesel exhaust with in vivo kinetics that differ more or less from those of B[a]P.
Thus, existing dosimetry models that incorporate the dissolution of B[a]P from diesel
particles as a representative organic constituent (Yu and Yoon, 1988) may not suitably
reflect the actual bioavailability of other procarcinogenic agents on diesel particles that may
contribute to the development of lung neoplasms.
833. Bioavailability of Adsorbed Compounds as a Function of Particle
Gearance Rates and Extraction Rates of Adsorbed Compounds
The bioavailability of the organic compounds adsorbed to diesel soot particles is a
function of desorpuon from the panicles which is dependent, at least in part, upon the
retention time of these particles in the alveolar region. The precise quantitative relation-
ships between bioavailability and rates of clearance and desorption are unclear at present.
Sun et al. (1984) exposed F344 rats to 30 min, nose-only inhalation of diesel exhaust
particles to which JH-B[a]P had been adsorbed (0.1 percent by mass). The rate limiting step
for bioavailability of B[a]P was suggested to be the rate at which it was desorbed from the
particle, and that the composition of the carrier particle and the associated physico-chemical
interactions between the particle and the adsorbed organic were involved in this process.
This postulation was justified by the fact that lung retention times for diesel exhaust particles
and insoluble gallium oxide particles is quite long (62±14 and 65±16 d, respectively), yet lung
retention of 3H-B[a]P was much greater when associated with diesel exhaust particles.
Also of importance in assessing the rate of biovailability would be the rate and extent
of phagolysosomal solubilization of adsorbed compounds. In analyzing phagolysosomal
dissolution of various ions from particles deposited in the lungs of Syrian golden hamsters,
Godleski et al. (1988) demonstrated that solubilization did not necessarily result in clearance
of the ions and that binding of the solubilized components to cellular and extracellular
structures occurred. It is reasonable to assume that phagocytized diesel soot particles may
be subject to similar processes, and that these processes would be Important in determining
the rate of bioavailability of the particle-bound constituents of diesel exhaust. Perhaps
May 1990
8-48
DRAFT-DO NOT QUOTE OR CITE
-------
because of longer residence times for panicle-adsorbed PAHs, the macrophages are better
able to carry out some of the metabolic activation steps.
Pepelko (1987) in reviewing species variability in alveolar clearance of particulate
matter provided an expression for the relationship among overall clearance rate (t^,), the
dissolution half-time (tyj, and the particle removal half-time (t^):
, hrit^xnp
'» " 7—7t— '
l/2» lflp
That lung clearance rate is the result of the combined effects of dissolution and AM-
mediated transport from the lungs is indicated by this expression. Hence, for insoluble
particles it follows that the half-time for AM-mediated transport would be similar to the
actual lung clearance half-time. If particle clearance half-time (t^) is not experimentally
derived, it can be estimated by rearrangement of the equation:
»0p b t -t '
U2m 41/26
If the dissolution half-time (t^) and the particle removal half-time (t^,) are sufficient for
accounting for clearance, then pulmonary retention can be calculated by first-order kinetics:
v o dRIdt a kR ,
where v is velocity. R is particulate matter concentration in the lungs, and k is the
elimination rate constant (equal to In 2/t,^). Integration of this equation yields:
R - V"* .
where Rq is the initial paniculate matter content of the lungs and t is time. Since dissolution
of paniculate matter and mechanical transpon are concurrent, the retention can be
expressed as:
lOOe
-<*,V
May 1990
8-49
DRAFT-DO NOT QUOTE OR CITE
-------
This expression does not account for delayed absorption following dissolution, transport to
lymph nodes, reactions with macroroolecules of the respiratory tract lining, or clearance by
more than an exponential rate. However, if these factors are adjusted for or are not of
significance, then the percentage of the remaining dose solubilized and absorbed each day
can be obtained:
A = lOOke'^'^ ,
where A is the daily absorption at time t. The total bioavailable dose can then be
approximated by
lOQfc .w
% absorption = i 1 -e m ' .
k +k
p '
&5.4. Summary: Bioavailability of Particle-Adsorbed Agents.
The available studies provide animal data attesting to the importance of exhaust
particles for increasing absorption potential of PAHs and the subsequent increase in
bioavailability of these compounds over time due to retention of particles and panicle
overload conditions. The total exposure of the biological system to panicle-associated
organics will be a function of exposure duration, deposition and pulmonary clearance of the
particles, the desorption of these organics from the particles, and the subsequent absorption
of the organic compounds into the blood and tissues. Sun et al. (1984) suggested this
process to be dependent upon the physicochemical interactions between the particle and its
associated organics. Furthermore, the profile of absorbed organic constituents may be
affected by alveolar macrophage-mediated biotransformation of these particle-associated
organic compounds. Alveolar macrophage activity may also alter the potential bioavailable
dose by relocation of ingested particles and their adsorbed organics to other pulmonary
regions or lymphatic tissue. It is also important to note that macrophage-mediated
metabolism of panicle-associated organics is no doubt dependent upon the accumulation of
May 1990 8-50 DRAFT-DO NOT QUOTE OR CITE
-------
the particles by the macrophage. This may, in part, explain the greater effects noted for
particle-associated vs. free PAHs. When a particle overload situation exists (as demonstrat-
ed by the studies summarized in Chapter 7), the time frame in which these processes can
occur would also be increased. There do not appear to be any studies that provide definitive
quantitative relationships among these processes.
8.6. INHIBITION OF RESPIRATION BY HIGH CONCENTRATIONS OF
NOXIOUS AGENTS
Exposure to whole diesel exhaust will result in inhalation of components such as
formaldehyde, acrolein, and sulfur dioxide, all of which have been demonstrated
to be sensory irritants. The fact that these irritants also affect respiratory rates (Kane and
Alarie, 1978, 1979) is of concern when considering inhalation exposure to diesel emissions.
The sensory irritant-induced reduction of respiratory rate is mediated through stimulation
of free nerve endings of the afferent trigeminal nerve (Ulrich et al., 1972). This physiologic
reflex response has been shown to be a dose-dependent response (Alarie, 1966,1973; Kane
and Alarie, 1978).
A rapid (within 1 to 2 min), concentration-dependent decrease in respiratory rate
followed by a rapid (within 12 min), concentration-dependent accommodation was reported
by Alarie et al. (1973) for mice exposed to sulfur dioxide at concentrations of 17, 62, 123,
or 298 ppm. This return to normal respiratory rates suggested that the sensory irritation
response resulting in respiratory rate depression was undergoing a desensitization despite
continued exposure. It was hypothesized that the SOz dissociated into S032' and HS03
anions, both of which stimulate the free nerve endings of the trigeminal nerve. This
concentration range encompasses the 20 to 40 ppm S02 content reported for diesel exhaust
(see Chapter 2).
Kane and Alarie (1978) reported similar findings for acrolein-exposed mice, although
the response to the irritant was slower. This varied response pattern initiated further
experiments that examined the effects of mixtures of acrolein and sulfur dioxide (Kane and
Alarie, 1979) and acrolein and formaldehyde (Kane and Alarie, 1978). In the 1978 study,
May 1990
8-51
DRAFT-DO NOT QUOTE OR CITE
-------
evaluation of the response patterns (rapidity, magnitude and duration of response, and
accommodation period) indicated that each compound produced a characteristic response.
Application of a mathematical model indicated that these interactions exhibited a
competitive agonism between acrolein and sulfur dioxide. In the 1979 study, exposures to
sulfur dioxide alone or to acrolein alone produced a maximum, but transient, reduction in
respiratory rate of 36.1 percent and 51.6 percent, respectively. Each of these compounds
produced a characteristic response pattern. However, exposure to mixtures of sulfur dioxide
and acrolein produced response patterns indicating that either irritant could alter or
completely negate the effect of the other. Generally, the recovery pattern following
exposure to the mixtures was slower than that for either irritant alone.
Alarie (1973) using data of Murphy et al. (1963) and Amdur and Mead (1955) noted
that nitro-olefins (possible reaction products of nitrogen oxides and olefinic hydrocarbons
of exhaust gases) are more potent (estimated at 40 to 100-fold) sensory irritants than is
sulfur dioxide, and that these considerations may be relevant in assessing effects of these
pollutants on sensory irritation and alteration of respiratory rate.
8.7. CONSIDERATIONS FOR DOSIMETRY MODELING
It is evident from the material reviewed in this and preceding chapters that derivation
of a dosimetry model for assessing the carcinogenic potential of diesel exhaust must consider
many factors. A primary point of concern in this respect is the difficulty inherent in
demonstrating evidence of carcinogenicity of a weak carcinogen, especially one that is
delivered at very low exposure concentrations. Epidemiologic studies provide only limited
and often controversial evidence. Experimental animal studies, however, have demonstrated
a relationship between diesel exhaust exposure and increased tumor incidence, but employ
higher exposure concentrations than would be normally encountered by humans.
Furthermore, the uncertainty of animal-to-human extrapolation and interspecific variability
in physiologic (respiratory rates, pulmonary capacities and volumes, respiratory tract
anatomy, etc.) and response parameters (pulmonary clearance mechanisms and panicle
overloading, tumor incidence) must be considered in dosimetry modeling.
May 1990 8-52 DRAFT-DO NOT QUOTE OR CITE
-------
It is also necessary to consider if the observed increases in tumor incidences reported
in animal studies are attributable to the organic constituents only, or if the particles to which
they are adsorbed are also functionally involved and not acting merely as a vehicle. The
function of the particulate matter may be twofold: (1) increasing lung epithelial cell
proliferation rates or (2) increasing the retention of organic compounds. This latter function
may be instrumental in explaining the increased tumor incidence in rats exposed to particle
adsorbed organics (B[a]P) (Saffiotti et al., 1964, 1965; Henry and Kaufman, 1973; Henry et
al., 1975) where a slow release of compound dose is delivered to the cells as opposed to
administration of pure B[a]P which may overwhelm the cells' metabolizing capacities thus
limiting the production of the reactive B[a]P metabolite. These factors would translate into
a situation of competing risks in the dosimetry model (dose delivery rate vs. chronicity of
exposure).
8.8. SUMMARY
Recent studies have provided evidence for tumorigenicity and carcinogenicity in
laboratory animals following chronic inhalation exposure to whole diesel exhaust at particle
concentrations as low as 3.5 mg/m3. The pulmonary clearance of diesel exhaust particles is
multiphasic and involves several processes including a relatively rapid mucociliary transport
and slower macrophage-mediated processes. The observed dose-dependent increase in the
particle burden of the lungs is due, in part, to an overloading of alveolar macrophage
function. The resulting increase in particle retention has been shown to increase the
bioavailability of particle adsorbed mutagenic and carcinogenic components such as B[a]P
and 1-NP. Experimental data also indicate alveolar macrophage-mediated metabolism and
phagolysosomai solubilization of particle-adsorbed components. Although macromolecular
binding of diesel exhaust particle-derived PAH and the formation of DNA adducts following
exposure to diesel exhaust have been reported, a quantitative relationship between these and
increased carcinogenicity is not available.
In addition to the aforementioned points, one must also consider the fact that other
compounds (e.g., gas-phase chemical irritants) may alter respiratory rate and, therefore, the
May 1990 8-53 DRAFT-DO NOT QUOTE OR CITE
-------
actual dose of potentially toxic components. Moreover, a better knowledge of particle
dissolution rate and particle removal rate is necessary for more accurately assessing
bioavailability of potentially carcinogenic components of diesel exhaust
May 1990
8-54
DRAFT-DO NOT QUOTE OR CITE
-------
8.9. REFERENCES
Adamson, L Y. R.; Bowden, D. H. (1978) Adaptive responses of the pulmonary macrophagic system to carbon.
IL Morphologic studies. Lab. Invest. 38: 43Q438.
Adamson, L Y. R.; 2 Bowden, D. H. (1981) Dose response of the pulmonary macrophagic system to various
particulates and its relationship to transepithelial passage of free particles. Exp. Lung Res. 2:165-175.
Alarie, Y. (1966) Irritating properties of airborne materials to the upper respiratory tract. Arch. Environ.
Health 13: 433-449.
Alarie, Y. (1973) Sensory irritation to the upper airways by airborne chemicals. ToxicoL Exp. Pharmacol. 24:
279-297.
Amdur, M. O.; Mead, J. (1955) A method for studying the mechanical properties of the lungs of unanesthe-
tized animals: Application to the study of respiratory irritants. Proc. 3rd Nat. Air Pollut Symp. pp.
150-159.
Aronson, M. (1963) Bridge formation and cytoplasmic flow between phagocytic cells. J. Exp. Med. 118:
1084-1102.
Bailey, M. R.; Fry, F. A.; James, A. C (1982) The long-term clearance kinetics of insoluble panicles from the
human lung. Ann. Occup. Hyg. 26: 273-271.
Ball, L. M.; King L. C (1985) Metabolism, mutagenicity, and activation of 1-nitropyrene in vivo and in vitro.
Environ. Internat. 11: 355-361.
Barth, D. S.; Blacker, S. M. (1974) The EPA program to assess the public health significance of diesel
emissions. Environ. Lett 6: 241-251.
Bond, J. A.; Mitchell, C E; Li, A. P (1983) Metabolism and macromolecular covalent binding of
benzo(a)pyrene in cultures of Fischer-344 rat lung Type II epithelial cells. Biochem. Pharmacol. 32:
3771-3776.
Bond. J. A.; Butler, M. M.; Medinsky, M. A.: Muggenburg, B. A.; McClellan, R. O. (1984) Dog pulmonary
macrophage metabolism of free and panicle-associated l4C-benzo(a)pyrene. J. ToxicoL Environ. Health
14: 181-189.
Bond, J. A.; Sun, J. D.; Medinsky, M. A.; Jones, R. K.; Yeh, R C (1986) Deposition, metabolism, and
excretion of l-[MC]nitropyrene and l-[l4C]nitropyrene coated on diesel exhaust particles as influenced
by exposure concentration. ToxicoL AppL PharmacoL 85:102-117.
Bond, J. A. (1987) Disposition of inhaled pure and particle-associated organic chemicals. Diesel emissions and
health effects. Founh Annual HEI Conference, p. 61.
Bowden, D. R; Adamson, L Y. R. (1984) Pathways of cellular efflux and paniculate clearance after carvon
instillation to the lung. J. Pannacol. 143: 117-125.
Brain. J. D.; Godleski, J. J.; Sorokin, S. P. (1977) Quantification, ongin, and fate of pulmonarv macrophages.
In: Respiratory Defense Mechanisms, Pan IL New York, Marcel Dekker, pp. 849-889.
May 1990
8-55
DRAFT-DO NOT QUOTE OR CITE
-------
Brooks, A. L.; Wolff, R. K4 Royer, R. E.; Clark, C R.; Sanchez, A.; McClellan, R. O. (1981) Deposition and
biological availability of diesel panicles and their associated mutagenic chemicals. Environ. Internat.
5: 263-267.
Castranova, V.; Bowman, L.; Reasor, M. J.; Lewis, T.; Tucker, J.; Miles, P. R. (1985) The response of rat
alveolar macrophages to chronic inhalation of coal dust and/or diesel exhaust Environ. Res. 36:
405-419.
Chan, T. L; Lee, P. S.; Hering, W. E (1981) Deposition and clearance of inhaled diesel exhaust particles in
the respiratory tract of Fischer rats. J. AppL ToxicoL 1: 78-82.
Chan, T. L.; Lee, P. S.; Hering, W. E. (1984) Pulmonary retention of inhaled diesel panicles after prolonged
exposures to diesel exhaust. Fundam. AppL ToxicoL 4: 624-631.
Cony, D.; Kulkarni, P.; Lipscomb, M. F. (1984) The migration of bronchoalveolar macrophages into hilar
lymph nodes. Am. J. PathoL 115: 321-328.
Crapo, J. D., Young, S. L., Frara. E K., Pinkenon, K. E., Barry, B. E., Crapo, R. O. (1983) Morphometric
characterisucs of cells in the alveolar region of mammalian lungs. Am. Rev. Respir. Dis. 128:
S42-S46.
Creasia, D. A.; Nettesheim, P. (1974) Respiratory cocarcinogenesis studies with ferric oxide: a test case of
cunent experimental models. In: Karbe, E. and Park, J. F., eds. Experimental lung cancer,
carcinogenesis and bioassays. Frankfun: Springer-Verlag. p 234.
Creasia, D. A.; Poggenburg, J. K.; Nettesheim, P., Jr. (1976) Elution of Benzo[a]pyrene from carbon particles
in the respiratory tract of mice. J. ToxicoL Environ. Health 1: 967-775.
Creutzenberg, O. Muhle, H. Bellmann, B., Kilpper, R., Mermestein, R., Morrow, P. (1989) Reversability of
biochemical and cytological alterations in bronchoalveolar lavagate upon cessation of dust exposure.
The Toxicologist 9: A305.
Cuddihy, R. G.; Yeh, H. C (1986) Model analysis of respiratory tract clearance of panicles inhaled by people.
In: Muggenburg, B. A.; Sun, J. D. eds. Annual Repon of the Inhalation Toxicology Research Institute.
Lovelace Biomedical and Environmental Research Institute. LMF-115. Albuquerque, NM. pp. 140-
147.
Evans, Z., Shami, S. G., Martinez, L A. (1986) proliferation of pulmonary alveolar macrophages
after carbon instillation in mice depleted of blood monocytes by stronuum-89. Lab. Invest. 54:
154-159.
Farrell, R. L.; Davis, G. W. (1974) Effect of paniculate benzo(a)pyrene carrier on carcinogenesis in the
respiratory tract of hamsters. In: E. Karbe; J.F. Park, eds. Experimental Lung Cancer, Carcinogenesis,
and Bioassays. Germany, Springer-Verlag, pp. 186-192.
Ferin, J. (1982) Pulmonary alveolar pores and alveolar macrophage-mediated particle clearance. Anat Rec
203: 265-272.
Ferin, J., Feldstein, M. L. (1978) Pulmonaiy clearance and hilar lymph node content in rats after panicle
exposure. Environ. Res. 16: 342-352.
May 1990
8-56
DRAFT-DO NOT QUOTE OR CITE
-------
Frey, J. W.; Corn, M (1967) Physical and chemical characterization of particles in diesel exhaust Amer. Ind.
Hyg. Assoc J. 28: 468-478.
Garcia-Moreno, L. F.; Myrvik, Q. N. (1977) Macrophage-agglutinating factor produced in vitro by
BCG-sensitized lymphocytes. Infect. Immun. 17: 613-620.
George, G.; Hook, G. E. R. (1984) Hie pulmonary extracellular lining. Environ. Health Perspectives 55:
227-237.
Gibb, F. R.; Morrow, P. E. (1962) Alveolar clearance in dogs after inhalation of an iron 59 oxide aerosoL J.
AppL PhysioL 17: 429-432.
Glatt, H. R.; Oesch, F. (1976) Phenolic benzo(a)pyrene metabolites are mutagenic. Mutat. Res. 36: 379-384.
Godleski, J. J.; Stearns, R. C; Katler, M. R.; Brain, J. D. (1988) Particle dissolution in alveolar macrophages
awssari by electron energy loss analysis using the Zeiss CEM902 electron microscope. J. Aerosol
Med. 1: 198.
Griffis, L. C; Wolff, R. K.; Henderson. R. F.; Griffith, W.C; Mokler, B.V.; McCIellan. R. O. (1983) Clearance
of diesel soot particles from rat lung after subchronic diesel exhaust exposure. Fund. Appt. Toxicol.
3: 99-103.
Gurley, G. R.; Spall, W. D.; Valdez, J. G.; London, J. E.; Dethloff, L. A.; Lehnert, B. E. (1988) An HPLC
procedure for the analysis of proteins in lung lavage Quid. Anal. Biochem. 172: 465-478.
Hannsen, A. G.; Muggenburg, B. A.; Snipes, M. B.; Bice, D. B. (1985) The role of macrophages in particle
translocation from lungs to lymph nodes. Science 230: 1277-1280.
Heinrich, U.; Muhle, R; Takenaka. S.; Ernst, H.; Fuhst, R.; Mohr, U.; Pott, F.; Stober, W. (1986) Chronic
effects on the respiratory tract of hamsters, mice and rats after long-term inhalation of high
concentrations of filtered and unfiltered diesel engine emissions. J. AppL ToxicoL 6: 383-395.
Henry, M. G; Kaufman, D. G. (1973) Clearance of benzo[a]pyrene from hamster lungs after administration
on coated particles. J. NatL Cancer Inst. 51: 1959-1964.
Henry, M. C; Port, C; Kaufman, D. G. (1974) Role of panicles in respiratory carcinogenesis. In:
Experimental Lung Cancer, Carcinogenesis and Bioassays. (E. Karbe and J. F. Park, eds.) Germany,
Springer-Verlag, p. 173.
Heniy, M. C; Port, C D.; Kaufman, D. G. (1975) Importance of physical properties of benzo(a]pyrene-fenic
oxide mixtures in lung tumor induction. Cancer Res. 35: 208-217.
Heppleston, A. G. (1961) Observations on the disposal of inhaled dust by means of the double exposure
technique. In: Inhaled Particles and Vapors. C N. Davies, ed., Pergamon, NY, pp. 320-325.
Heppleston, A. G.; Young, A. E. (1974) Uptake of inert particulate matter from alveolar cells: An
ultrastructural study. J. PathoL 111: 159-164.
Holland, L. M.; Wilson, J. S.; Tillery, M. L; Smith, D. M. (1985) Lung cancer in rats exposed to fibrogenic
dusts. In: Silica, Silicosis and Cancer. (D. Goldsmith, D. Winn, and C Shy, eds.) Praeger, New
York, pp. 267-279.
May 1990
8-57
DRAFT-DO NOT QUOTE OR CITE
-------
Huisingb, J.; Bradow, R.; Jungers, R.; Oaxton, L; Rodriguez, C; Snow, L. (1978) Applicauon of bioassay to
the characterization of diesel particle emissions. EPA-600/9-78-027, pp. 1-32.
Ishinishi, N. (ed.) (1986) Carcinogenic and Mutagenic Effects of Diesel Exhaust. Elsevier, B. V.
Kane, L. E.; Alarie, Y. (1978) Evaluation of sensory irritation from acrolein-formaldehyde matures. Am. Ind.
Hyg. Assoc J. 39: 270-274.
Kane, L. E.; Alarie, Y. (1979) Interactions of sulfur dioxide and acrolein as sensory irritants. ToxicoL AppL
Pharmacol. 48: 305-315.
Kawabata, Y. (1986) Effects of diesel soot on unscheduled DNA synthesis of tracheal epithelium and lung
tumor formation. In: Carcinogenic and Mutagenic Effects of Diesel Engine Exhaust (Ishinishi,
N., ed.) Elsevier, B. V., pp. 213-222.
King, L. C; Kohan, M. J.; Austin, A. C; Claxion, L. D.; Huisingh, J. L. (1981) Evaluation of the release of
mutagens from diesel particles in the presence of physiological fluids. Environ. Mutagen 3: 109-121.
King, L. C; Loud, K.; Tejada, S. B.; Kohan, M. J.; Lewtas, J. (1983) Evaluation of the release of mutagens
and 1-nitropyrene from diesel panicles in the presence of lung macrophages in culture. Environ.
Mutagen. 5: 577-588.
Kittelson, D. B.; Dolan, D. F.; Verrant, J. A. (1978) Investigation of a diesel exhaust aerosol. Soc. Auto
Engineers Technical Paper No. 780109 (Transactions of SAE).
Kuschner, M. (1968) The causes of cancer. Am. Rev. Respir. Dis. 98: 573-590.
Lauweryns, J. M.; Baert, J. R (1977) Alveolar clearance and the role of pulmonary lymphatics. Am. Rev.
Respir. Dis. 115: 625-683.
Lee, F. S. C; Pierson, W. R.; Ezike, J. (1980) The problem of PAH degradation of airborne particulates: An
evaluation of several commonly used filter media. In: Polynuclear Aromauc Hydrocarbons: The
Fourth International Symposium. Battelle Press, pp. 543-564.
Lee, P. S.; Chan, T. L.; Henng, W. E. (1983) Long-term clearance of inhaled diesel exhaust particles in
rodents. J. ToxicoL Environ. Health 12: 801-813.
Lee, K. P., Henry III; N. W., Trochimowicz; R J., Reinhardt; C F. (1986) Pulmonary response to impaired
lung clearance in rats following excessive TiO: dust deposition. Environ. Res. 41: 144-167.
Lee, K. P.; Trochimowicz; R J„ Reinhardt; C F. (1985) Transmigration of titanium dioxide (TiOj) particles
in rats after inhalation exposure. Exp. Molecular Pathol. 42: 331-343.
Lee, K. PM Ulrich, C EL, Ceil, R. G., Trochimowicz, R J. (1988) Effects of inhaled chromium dioxide dust
on rats exposed for two years. Fund. AppL ToxicoL 10: 125-145.
Lehnert, B. E. (1988a) Distribution of particles in alveolar macrophages during lung clearance. J. Aersol Med.
1:206.
Lehnert, B. E. (1988b) Some points of discussion for session 5: lung clearance ana retention of particles:
mechanisms involving pulmonary macrophages. J. Aerosol Med. 1: 214.
May 1990
8-58
DRAFT-DO NOT QUOTE OR CITE
-------
Lehnert, B. E. (1988) Distributions of particles in alveolar macrophages during lung clearance. J. Aerosol
Med. 1: A210-21L
Lehnert, B. E. (1989) Rates of disappearance of alveolar macrophages during lung clearance as a function of
phagocytized paniculate burden. 1989 Annual Meeting of the American Thoracic Society, Cincinnati,
Ohio, May 14-17.
Lehnert, B. Cline, A.; London, J. E (1989) Kinetics of polymorphonuclear leukocytes and their particle
burdens during the alveolar clearance of a high lung burden of panicles. The Toxicologist 9: A307.
Lehnert, B. E.; Morrow, P. E.(1985) Association of ^iron oxide with alveolar macrophages dunng alveolar
clearance. Exp. Lung Res. 9: 1-16.
Lehnert, B. E.; Valdez, Y. E.; Stewart, C. C (1986) Translocation of particles to the tracheobronchial lymph
nodes after lung deposition: Kinetics and particle-cell relationships. Exp. Lung Res. 10: 245-266.
Lippmann, M.; Schlesinger, R. B. (1984) Interspecies comparisons of panicle deposition and mucociliary
clearance in tracheobronchial airways. J. Toxicol. Environ. Health 13: 441-469.
Lundborg, M.; Lind, B.; Camner, P. (1984) Ability of rabbit alveolar macrophages to dissolve metals. Exp.
Lung Res. 7: 11-22.
Marafante, E.; Lundborg, M.; Valuer, M.; Camner. P. (1987) Dissolution of two arsenic compounds by rabbit
alveolar macrophages in vitro. Fund. AppL Toxicol 8: 382-388.
Mauderty, J. L.; Jones, R. K.; Griffith, W. C; Henderson, R. F.; McClellan, R. O. (1987) Diesel exhaust is a
pulmonary carcinogen in rats exposed chronically by inhalation. Fund. AppL ToxicoL 9: 208-221.
McClellan, R. O.; Brooks, A. L.; Cuddihy, R. Jones, R. K.; Mauderty, J. L.; Wolff, R.K. (1982) Inhalation
toxicology of diesel exhaust panicles. In: Toxicological Effects of Emissions from Diesel Engines (J.
Lewtas, ed.) Elsevier/Nonh Holland Amsterdam, pp. 99-120.
Medinsky, M^.; Kampcik. S. J. (198S) Pulmonary retention of f"C]benzo[a]pyrene in rats as influenced by
the amount instilled. Toxicology 35: 327-336.
Morrow, P. E. (1988) Possible mechanisms to explain dust overloading of the lungs. Fund. Appl. Toxicol. 10.
369-384.
Murphy, S. D.; Ulrich, C E^ Leng, J. K. (1963) Altered function in animals inhaling conjugated nitro-olefins.
ToxicoL AppL PharmacoL 5: 319-330.
Nilsen, A.; Nyberg, K.; Camner, P. (1988) Intraphagosomal pH in alveolar macrophages after phagocytosis of
fluorescein-labelled yeast panicles. Exp. Lung Res. 14: 197-207.
OberdOrster, G.; Ferin, J.; Morse. P.; Corson, N. M.; Morrow, P. E. (1988) Volumetric alveolar macrophage
(AM) burden as a mechanism of impaired panicle clearance dunng chronic dust overloading of the
lung. J. Aerosol Med. 1: A207.
Ohkuma, S.; Poole, B. (1978) Fluorescence probe measurement of the intralysosomal pH in living cells and
the p^nurbauon of pH by various agents. Proc. NatL Acad. ScL USA 75: 3327-3331.
May 1990
8-59 DRAFT-DO NOT QUOTE OR CITE
-------
Patra, A. L. (1986) A morphometry comparison of the nasopharyngeal airway of laboratory animals and
humans. Anat Rec. 215: 42-50.
Pepelko, W. E. (1987) Feasibility of dose adjustment based on differences in long-term clearance rates of
inhaled particulate matter in humans and laboratory animals. Regulatory ToxicoL Pharmacol. 7:236-
252.
Pepelko, W. E.; Peirano, W. B. (1983) Health effects of exposure to diesel engine emissions. J. Amer. Coll.
ToxicoL 2: 253-306.
Phalen, R. F.; Oldham, M. J. (1983) Tracheobronchial airway structure as revealed by casting techniques. Am.
Rev. Respir. Dis. 128: S1-S4.
Plopper, C G4 Hyder, D. M.; Weir, A J. (1983) Centnacmar alteration in lungs of cats chronically exposed
to diesel exhaust Lab. Invest. 49: 391-399.
Powdrill, J., Buckley, C, Valdez. Y. E, Lehnert, B. E(1989) Airway intra-luminal macrophages: Origin and
role in lung clearance. The ToxiCQlogist 9: A306.
Raabe, O. G. (1979) Deposition and clearance of inhaled aerosols. Washington, DC: U.S. Department of
Energy, Report No. UCD-472-503.
Riley, P. A., Dean, R. T. (1978) Phagocytosis of latex particles in relation to the cell cycle in 3T3 cells. Exp.
Cell Biol. 46: 367-373.
Roser, B. (1970) The origins, kinetics, and fate of macrophages populations. J. ReticuloendotheL Soc. 8:
139-161.
Saffiotti, U.; Borg, S. A.; Grote, M. L; Karp, D. B. (1964) Retention rates of particulate carcinogens in lungs.
Chicago Med. Scholl Quart. 2.24: 10-17.
Saffiotti, U4 Borg, S. A.; Grote, M. I.; Karp, D. B. (1965) Experimental studies of the conditions of exposure
to carcinogens for lung cancer induction. JAPCA 15: 23-25.
Saffiotti, U.; Montesano. R.; Sellakumar. A. R.; Cefis, F.; Kaufman. D.G. (1972) Respiratory tract
carcinogenesis in hamsters induced by benzo(a)pyrene and ferric oxide. Cancer Res. 32: 1073-1081.
Sandusky, C B.; Cowden, M. W.; Schwartz, S. L (1977) Effect of particle size on regurgiuve exocytosis by
rabbit alveolar macrophages. In: Pulmonary Macrophage and Epithelial Cells. CONF-760927,
National Technical Information Service, U.S. Dept. of Commerce, pp. 85-105.
Sims, P.; Graver, P. L; Swaisland, A; Pal, K.; Hewer, A. (1974) Metabolic activation of benzo(a)pyrene
proceeds by a diol-cpoxide. Nature (London) 252: 326-328.
Siaga, T. J.; Viage, A.; Berry, D. L.; Bracken, W. (1976) Skin tumor initiating activity of benzo(a)pyrene, 4,5-
7,8- and 7,8-
-------
Snipes, M. B.; Boecker, B. B.; McClellan, R. O. (1983) Retention of monod is perse or polydisperse
aluminosilicate panicles inhaled by dogs, rats, and mice. Toxicol. Appl. Pharmacol. 69: 345-362.
Snipes, M. B.; Clem, M. F. (1981) Retention of microspheres in the rat lung after intratracheal instillation.
Environ. Res. 24: 33-41.
Snipes, M. B.; McClellan, R. O. (1984) Disposition and retention of monodispeise aluminosilicate pantiles
inhaled by guinea pigs. Inhalation Toxicology Research Institute Annual Report, LMF-113,
Albuquerque, NM. pp. 80-83.
Snipes, M B.; Olson, T. R.; Yeh, H. C (1988) Deposition and retention patterns for 3-, 9-, and 15-mm latex
microspheres inhaled by rats and guinea pigs. Exp. Lung Res. 14: 37-50.
Sorokin, S. P., Brain, J. D. (1975) Pathways of clearance in mouse lungs exposed to iron oxide aerosols. Anat.
Rec. 181: 581-625.
Stahlhofen, W.; Cebhan, J.; Heyder, J. (1980) Experimental determination of the regional deposition of
aerosol particles in the human respiratory tract. Am. Ind. Hyg. Assoc J. 41: 385-398.
Stober, W. (1986) Experimental induction of tumors in hamsters, mice and rats after long-term inhalation of
filtered and unfiltered diesel engine exhaust. In: Ishinishi, N.; Koizumi, A.; McClellan, R.; Stober, W.,
eds. Carcinogenic and mutagenic effects of diesel engine exhaust. Amsterdam: Elsevier, pp. 421-429.
Strom, K. A. (1984) Response of pulmonary cellular defenses to the inhalation of high concentrations of diesel
exhaust. J. Toxicol. Environ. Health 13: 919-944.
Sun, J. D.; Wolff, R. K.; Kanapilly, G. M. (1982) Deposition, retention, and biological fate of inhaled
benzo(a)pyrene adsorbed onto ultra fine particles and as pure aerosol. Toxicol. Appl. Pharmacol. 65:
231-244.
Sun, J. D.; Wolff, R. K.; Kanapilly, G. M.; McClellan, R. O. (1984) Lung retention and metabolic fate of
inhaled benzo(a)pyrene associated with diesel exhaust particles. Toxicol. Appl. Pharmacol. 73: 48-59.
Ulrich, C; Haddock, M.; Alane, Y. (1972) Airborne chemical irritants: role of the trigeminal nerve. Arch.
Environ. Health 24: 37-42.
U.S. EPA (1988) Interim Methods for Development of Inhalation Reference Doses. Environmnetal Criteria
and Assessment Office, Office of Health and Environmental Assessment. Cincinnati, OH, and
Research Triangle Park, NC EPA/600/8-88/066.
Vostal, J. J., Schreck, R. M., Lee, P. S., Chan, T. L», Soderholm, S. C (1982) Deposition and clearance of
diesel particles from the lung. In: Toxicological Effects of Emissions from Diesel Engines (J. Lewtas,
ed.), Elsevier, NY. pp. 143-159.
Vostal, J. J. (1983) Bioavailability and biotransformation of the mutagenic component of particulate emissions
present in motor exhaust samples. Environ. Health Perspectives 47: 269-281.
White, H. J.; Garg, B. D. (1981) Early pulmonary response of the rat lung to inhalation of high concentration
diesel particles. J. Appl. ToxicoL 1: 104-110.
May 1990
8-61
DRAFT-DO NOT QUOTE OR CITE
-------
Wiester, M J4 Ilitis, R.; Moore, W. (1980) Altered function and histology in guinea pigs after inhalation of
diesel exhaust. Environ. Res. 22: 285-297.
Wislocki, R. G.; Wood, A. W.; Chang, R. L.; et aL (1976) Mutagenicity and cytotoxicity of benzo(a)pyrene
arene oxides, phenols, quinones, and dihydrodiols in bacterial and mammalian cells. Cancer Res. 36:
3350-3357.
Wolff, R. K.; Henderson, R. F.; Snipes, M. B.; Sun, J. D.; Bond, J. A.; Mitchell, C E.; Mauderty, J. L.;
McClellan, R. O. (1986) Lung retention of diesel soot and associated organic compounds. In: Ishinfchi,
N4 Koizuma, A.; McClellan, R.; Stober, W„ eds. Carcinogenic and mutagenic effects of diesel engine
exhaust. Amsterdam: Elsevier, pp. 199-211.
Wolff; R. K; Henderson, R. F.; Snipes, M. Griffith, W. C; Mauderly, J. L.; Cuddihy, R. G.; McClellan, R.
O. (1987) Alterations in particle accumulation and clearance in lungs of rats chronically exposed to
diesel exhaust Fundam. AppL Toxicol. 9: 154-166.
Wolff, R. K. (1989) Effects of adsorption of benzo(a)pyrene onto carbon black particles on levels of DNA
adducts of rats exposed by inhalation. Toxicol Appl. Pharmacol. 97: 289-299.
Wood, A. W.; Levin, W.; Lu, A. Y. H.; Yagi, H.; Hernandez, O.; Jerina, D. M.; Conney, A. H. (1976)
Metabolism of benzo(a)pyrene and benzo(a)pyrene derivatives to mutagenic products by highly
purified hepatic microsomal enzymes. J. BioL Chem. 251: 4882-4890.
Wright, E. (1986) Effects of short-term exposure to diesel exhaust on lung cell proliferation and phospholipid
metabolism. Exp. Lung Res. 10: 39-55.
Wright, J. R4 Clements, J. A. (1987) Metabolism and turnover of lung surfactant. Am. Rev. Respir. Dis. 135:
426-444.
Yu, C P.; Yoon, JC J. (1988) Determination of lung doses of diesel exhaust particulates. 5th Annual
Conference of HEI, Colorado Springs. CO.
May 1990
8-62
DRAFT-DO NOT QUOTE OR CITE
-------
9. EPIDEMIOLOGY STUDIES
9.1. EPIDEMIOLOGIC STUDIES OF THE CARCINOGENICITY
OF EXPOSURE TO DIESEL EMISSIONS
Emissions of diesel exhaust are made up of toxicants that include oxides of nitrogen
and sulfur, carbon monoxide, and particulate matter consisting of a carbon core with many
organic compounds, especially the polycyclic aromatic hydrocarbons absorbed on the surface.
These emissions contain about 100 times more particulate matter than gasoline engine
exhaust; the implications of this for human health are still unknown. To evaluate the risk
assessment of the diesel exhaust emissions mortality and morbidity, studies on health effects
of diesel emission exposures are reviewed here.
Three types of studies on the health effects of diesel emission exposures are
considered: (1) cohort studies, (2) case-control studies of lung cancer, and (3) case-control
studies of bladder cancer. Although an attempt was made to cover all the relevant studies,
there are certain studies that are not included in this report for various reasons. Since the
change from steam to diesel engines in locomotives was about 95 percent complete by 1959
and since diesel buses were introduced about the same time, studies conducted prior to 1959,
when diesel exhaust exposure was less common, are not considered here (Raffle, 1957;
Kaplan, 1959). As more definitive results were available in railroad workers, the pilot study
in those workers (Schenker et al., 1984) is not reviewed here. Studies by Waxweiler et al.
(1973), Gustafsson et al. (1986), Stern et al. (1981), and Garland et al. (1988) were excluded
either because of uncertainty of exposure to diesel exhaust or because exposure was to both
diesel and gasoline exhaust. For the first type of studies the cohorts of heavy construction
equipment operators, railroad and locomotive workers, and bus garage employees were
retrospectively studied to determine increased mortality and morbidity resulting from
exposures to varying levels of diesel emissions at the work place.
The hypothesis-generating study by Buiatti et al. (1985) was excluded because it would
have needed confirmation by a definitive study. The study by Coggon et al. (1984) (both
case-control study of lung cancer as well as of bladder cancer) also was not included, as the
June 1990 9-1 DRAFT - DO NOT QUOTE OR CITE
-------
occupational information was abstracted from death certificates, that had not been validated;
this would have resulted in limited information. A total of seven lung case control studies
are considered in this report.
Case-control studies of bladder cancer by Vineis and Magnani (1985), Silverman et
al. (1986), Jensen et al. (1987), and Risch et al. (1988) were not considered in this report
either because the exposure was categorized as motor exhaust or because the occupations
such as taxi, bus, and truck drivers were considered together as one job category and it was
difficult to determine whether the observed effect was because of exposure to diesel or
gasoline exhaust
A total of five bladder cancer case-control studies are described in this report.
9.2. COHORT STUDIES
9.2.1. Waller (1981): Trends in Lung Cancer in London in
Relation to Exposure to Diesel Fumes
A nonconcurrent prospective mortality study of a cohort of London transport workers
was conducted to determine if there was an excess of deaths from lung cancer that could be
attributed to diesel exhaust exposure. A total of nearly 20,000 male employees aged 45 to
64 were followed for a 25-yr period between 1950 and 1974 constituting a total of 420,700
man-years at risk. These were distributed among five categories: drivers, garage engineers,
conductors, motormen or guards, and engineers (works). Most of them lived in the greater
London area. Lung cancer cases occurring in this cohort were ascertained only from death
certificates of individuals who died while still employed, or if retired, following diagnosis.
Expected death rates were calculated by applying greater London death rates to the
population at risk within each job category. Data were calculated in 5-yr periods and 5-yr
age ranges, finally combining the results to obtain the total expected deaths in the required
age range of 45 to 64 for the calendar period of 1950 to 1974. There were 667 cases of lung
cancer reported compared with 849 expected to give a mortality ratio of 79 percent. In each
of the five job categories, the observed numbers were below those expected. Engineers in
garages had the highest mortality ratio (90 percent), but this did not differ significantly from
June 1990
9-2
DRAFT - DO NOT QUOTE OR CITE
-------
the other job categories. Environmental sampling was done at one garage, on 1 d, for
benzo(a)pyrene concentrations and was compared with corresponding values recorded in
1957. Concentrations of benzo(a)pyrene recorded in 1957 were at least 10 times greater
than those measured in 1979.
This study has several methodologic limitations. The lung cancer deaths ascertained
for the study were the ones that occurred while in service (worker either died of lung cancer
or retired after lung cancer was diagnosed). Although man-years at risk were based on the
entire cohort, no attempt was made to trace or evaluate the individuals who had resigned
from the London transport company for any other reason. Hence the information on
resignees and lung cancer deaths among them was not available for analysis. This fact may
have lead to a dilution effect, resulting in under- ascertainment of observed lung cancer
deaths and underestimation of mortality ratios. Eligibility criteria for inclusion in the cohort,
such as starting date and length of service with company were not specified. Because an
external comparison group was used to obtain expected number of deaths, the resulting
mortality ratios were less than 1; this may be a reflection of the "healthy worker effect".
Investigators also did not categorize the five job categories by levels of diesel exhaust
exposure nor did they use an internal comparison group to derive risk estimates.
The age range considered for this study was limited to 45 to 64 yr for the period
between 1950 and 1964. It is not clear whether this age range was applied to calendar year
1950 or 1964 or at the mid-point of this 25-yr follow-up period. The rates were calculated
by 5-year grouping but were not presented by 5-yr age group, which would have been more
meaningful. No analyses were presented either by latency or by duration of employment
(surrogate for exposure). The environmental survey based on benzo(a)pyrene concentra-
tions suggests that the cohort in their earlier years were exposed to much higher concentra-
tions of environmental contaminants than current concentrations. It would have been useful
to examine lung cancer deaths at different time periods. Lastly, no data were collected on
smoking habits.
June 1990
9-3
DRAFT - DO NOT QUOTE OR CITE
-------
9.2.2. Howe et al. (1983): Cancer Mortality (1965-1977) in Relation to
Diesel Fume and Coal Exposure in a Cohort of Retired Railroad
Workers
This is a retrospective cohort study of the mortality experience of 43,826 male
pensioners of the Canadian National Railroad (CNR) between 1965 and 1977. Members
of this cohort consisted of male CNR pensioners who had retired before 1965 and who were
known to be alive at the start of that year, as well as those who retired between 1965 and
1977. The records were obtained from a computer file, which is regularly updated, that is
used by the company for payment of pensions. To receive a pension each pensioner must
provide, on a yearly basis, evidence to the effect that he is alive. Specific cause of death
among members of this cohort was ascertained by linking these records to the Canadian
Mortality Data Base which contains records of all deaths registered in Canada since 1950.
Of the 17,838 deaths among members of the cohort between 1965 and 1977, 16,812 (94.4
percent) were successfully linked to a record on the mortality file. A random sample manual
check on unlinked data revealed that failure to link was mainly due to some missing
information on the death records.
Occupation at time of retirement was used by the Department of Industrial Relations
to classify workers into three diesel fume and coal dust exposure categories. Nonexposed,
possibly exposed, and probably exposed. Person-years of observation were calculated and
classified by age at observation in 5-yr age groups (35-39, 40-44 — 80-84, >85). The
observed deaths were classified by age at death for different cancers, for all cancers
combined, and for all causes of death combined. Standard mortality ratios (SMR) were then
calculated using rates of the Canadian population for the period between 1965 to 1977.
Both total mortality (SMR = 0.95, p <0.001) and all cancer deaths (SMR = 0.99,
p >0.05) were close to that expected for the entire cohort. Analysis by exposure to diesel
fume levels in the three categories (nonexposed, possibly exposed, and probably exposed)
revealed an increased relative risk for lung cancer among workers with increasing exposure
to diesel fumes. The relative risk for nonexposed workers was presumed to be 1.0; for those
possibly exposed the relative risk was elevated to 1.2, which was statistically significant
June 1990
9-4
DRAFT - DO NOT QUOTE OR CITE
-------
(p = 0.013); for those probably exposed it was elevated to 135 which was statistically highly
significant (p = 0.001). The corresponding rates for exposure to varying levels of coal dust
were very similar at 1.00, 1.21 (p = 0.012), and 1.35 (p = 0.001), respectively. The trend
tests were highly significant for both exposures (p <0.001). Since there was considerable
overlap between occupations involving probable exposure to diesel fumes and probable
exposure to coal dust and since most members of the cohort were employed during the years
in which the transition from coal to diesel occurred, it was difficult to distinguish whether
lung cancer was associated with exposure to coal dust or diesel fumes.
Although this study showed a highly significant dose-response relationship with diesel
fumes and lung cancer, it has several methodological limitations. There were concurrent
exposures to both diesel fumes and coal dust during the transition period, therefore,
misclassification of exposure may have occurred, since only occupation at retirement was
available for analysis. It is possible that the elevated response observed for lung cancer was
due to the combined effect of exposure to both coal dust and diesel fumes and not just one
or the other. No information was provided on duration of employment in either diesel work
or the coal dust-related jobs for other than those jobs held at retirement. Therefore, it was
not possible to evaluate whether this omission would have led to an under- or overestimate
of the true relative risk. Furthermore, a lack of information on potential confounders such
as asbestos exposure and smoking makes the interpretation of the excess risk of lung cancer
even more difficult. Information on cause of death was acquired from the mortality data
linkage. There is a possibility that the cause of death may have been misclassified because
of miscoding of the underlying cause of death. Further, it is not clear whether or not the
investigators verified the cause of death which could also lead to misclassification. Again,
it is difficult to evaluate whether this may have lead to under- or overestimation of the risk.
Because of the above limitations, the findings from this study, at best, are only suggestive
and not confirmatory of an excess risk of lung cancer from occupational exposure to diesel
fumes.
June 1990
9-5
DRAFT - DO NOT QUOTE OR CITE
-------
9.23. Rushton et aL (1983): Epidemiological Survey of Maintenance
Workers in the London Transport Executive Bus Garages and
Chiswick Works
This is a nonconcurrent mortality cohort study of male maintenance workers
employed for at least one continuous year between January 1,1967 and December 31,1975,
at 71 London transport bus garages (also known as rolling stock) and at Chiswick Works.
For all men the following information was obtained from computer listings: surname with
initials, date of birth, date of joining company, last or present jobs, and location of work.
For those individuals who left their job, date of and reason for leaving were also obtained.
For those who died in service or after retirement and for men who had resigned, full name
and last known address were obtained from an alphabetical card index in the personnel
department. Additional tracing of individuals who had left was carried out through social
security records. The area of their residence was assumed to be close to their work;
therefore their place of work was coded as their residence. There were 100 different job
titles that were coded into 20 broader groups. The reason for leaving was coded as died in
service, retired, or other. The underlying cause of death was coded using the 8th revision
of the International Classification of Dieases (ICD). Person-years were calculated from date
of birth and dates of entry to and exit from the study using the man-years computer
language program. These were then subdivided into 5-yr age and calendar period groups.
The expected number of deaths were calculated by applying the 5-yr age and calendar
period death rates of the comparison population to the person-years of corresponding
groups. The mortality experience of the male population in England and Wales was used
as the comparison population. Significance values were calculated for the difference
between the observed and expected deaths, assuming a Poisson distribution.
There were a total of 50,008 person-years of observation in the study with a mean
follow-up of 5.9 yr. Only 2.2 percent of the men were not traced. Observed deaths from
all causes were significantly lower than expected (observed/expected = 495/607.46,
p <0.001). The observed deaths from all neoplasms and cancer of the lung were
approximately the same as those expected. The only significant excess observed for cancer
of the liver and gall bladder at Chiswick Works was based on four deaths (p <0.05). A few
June 1990
9-6
DRAFT - DO NOT QUOTE OR CITE
-------
job groups showed a significant excess of risks for various cancers. Cancer of the liver and
gall bladder (observed — 2,p <0.01) and cancer of the bladder (Observed = 3, p <0.01)
were significant in the job categories of inspector and progress hand, respectively. Welders
showed an excess of lung cancer (Observed = 3,p <0.05), bus mechanics had an excess of
cancer of the brain (Observed = 4,p <0.04), and painters had an excess of bladder cancer
(Observed = 2>p <0.02). All the excess deaths observed for the various job groups were
based on very small numbers (usually smaller than 5) and merited cautious interpretation.
This is an exploratory mortality study of London transport maintenance workers that
did not demonstrate any cancer excesses based on a large number of cases, which needs
further exploration. Its several limitations include the small sample size, short duration of
follow-up (average of only 6 yr), and lack of long enough latency period. Thus, the number
of deaths by different causes and among the various job groups was too small to allow any
meaningful conclusions. Details of work history were not obtained to permit any analysis
by diesel exhaust emission exposure. Death information was ascertained from death
certificates with inherent problems of inaccuracy, misdiagnosis, and errors in coding. It was
not known whether a trained nosologist coded the death certificates or not. No adjustments
were made for the confounding effects of smoking and socioeconomic factors.
9.2.4. Wong et aL (1985): Mortality Among Members of a Heavy
Construction Operators Union With Potential Exposure to Diesel
Exhaust Emissions
This is a historical prospective mortality study conducted on a cohort of 34,156 male
members of a heavy construction equipment operators union with potential exposure to
diesel exhaust emissions. Study cohort members were identified from records maintained
at Operating Engineers' Local Union No. 3-3A in San Francisco. This union has maintained
both work and death records on all its members since 1964. Individuals with at least 1 yr
of membership in this union between January 1, 1964, and December 31, 1978, were
included in the study. Work histories of the cohort were obtained from job dispatch
computer tapes. The study follow-up period was from January 1964 to December 1978.
Death information was obtained from a trust fund, which provided information on
retirement dates, vital status, and date of death for those who were entitled to retirement
June 1990
9-7
DRAFT - DO NOT QUOTE OR CITE
-------
and death benefits. The average duration of membership was 15 yr. As of December 31,
1978, there were 15,735 cohort members who were active members of the union, 10,505
(30.8 percent) who were inactive or terminated, 6670 (19.6 percent) retired, and 1238 (3.6
percent) who had died while active or retired. Vital status of 10,505 members who left the
union was ascertained from the Social Security Administration. Death certificates were
obtained from appropriate state health departments. Altogether, 3345 deaths in the cohort
were coded using the ICD-7. Expected deaths and SMRs were calculated using the U.S.
national age-sex-race cause-specific mortality rates for 5-yr time periods between 1964 and
1978. The entire cohort population contributed to 372,525.6 person-years in this 14-yr study
period. Death certificates could not be obtained for 102 individuals; these individuals were
included in the calculation of the SMR for all causes of death but were deleted from the
cause-specific SMR analyses.
A total of 3345 deaths was observed compared with 4109 expected. The correspond-
ing SMR for all causes was 81.4 (p = 0.01), which confirmed the "healthy worker effect".
Eight hundred seventeen deaths were attributed to malignant neoplasms, slightly fewer than
the expected 878.34 based on the U.S. white male cancer mortality rates (SMR = 93.0,
p = 0.05). Mostly there were SMR deficits for cause-specific cancers, including lung cancer
for the entire cohort (SMR = 98.6, observed = 309). The only significant excess of SMR
was observed for cancer of the liver (SMR = 166.7, observed = 23, p <0.05). There were
a couple of statistically nonsignificant cancer excesses observed for cancer of the bladder
(SMR = 118.1, observed = 27) and cancer of the stomach (SMR = 117.6, observed = 44).
For nonmalignant causes statistically significant (p <0.01) excesses were observed for
emphysema (SMR = 165.3, observed = 116) and accidental deaths (SMR = 127.0, observed
= 348).
Analysis by length of union membership as a surrogate of duration for potential
exposure showed statistically significant (p <0.01) increases in SMRs of cancer of the liver
(SMR = 424.1) in the 10 to 14-yr membership group, emphysema in the 15 to 19 and 20+
year membership groups, SMRs being 174.5 and 174.8, respectively, and accidental deaths
in the 10 to 14-yr membership group (SMR = 127.7). In addition to these increases, a
statistically significant (p <0.05) increase was also observed for cancer of the stomach
June 1990
9-8
DRAFT - DO NOT QUOTE OR CITE
-------
(SMR = 248.4) in the 5 to 9-yr membership group. For emphysema a positive trend was
observed with increasing length of membership. Although the SMR for cancer of the lung
had a statistically significant deficit in the less than 5-yr duration group, it showed a positive
trend with increasing length of membership which, leveled off after 10 to 14 yr.
Cause-specific mortality analysis by latency period showed a positive trend for SMRs
of all causes of death, although all of them were statistically significant deficits, reflecting the
diminishing "healthy worker effect". This analysis also demonstrated statistically significant
SMR excesses for cancer of the liver (10 to 19 yr group, SMR = 257.9), emphysema (20+ yr
group, SMR = 199.4), and accidents (10 to 19 yr group, SMR = 140.6). A positive trend
with increasing latency was observed for emphysema. The SMR for cancer of the lung
showed a statistically significant deficit for the 10-yr follow-up, but showed a definite positive
trend with increasing latency.
In addition to these analyses of the entire cohort, similar analyses were carried out
in various subcohorts. Analyses of retirees, 6678 individuals contributing to 32,670.1 person-
years, showed statistically significant increases (p <0.01) in SMR for all cancers (SMR =
145 J, observed = 389), all causes of death (SMR = 114.5, observed = 1,345), cancer of the
digestive system (SMR = 142.4, observed = 103), cancer of the large intestine (SMR =
181.8, observed = 46), cancer of the respiratory system (162.4, observed = 161), cancer of
the lung (SMR = 164.1, observed = 155), emphysema (SMR = 277.3, observed = 75), and
cirrhosis of the liver (SMR = 173.5, observed = 38). The other two significant excesses
(p <0.01) were for lymphosarcoma and reticulosarcoma (SMR = 231.2, observed = 10) and
nonm align ant respiratory diseases (SMR = 129.0, observed = 112). Further analysis of the
4075 retirees (18,677.8 person-years) who retired at age 65 or who retired earlier but had
reached the age of 65, revealed statistically significant SMR increases for all cancers
(SMR = 114.7, observed = 224, p = 0.05), cancer of the lung (SMR = 130, observed = 86,
p <0.05), and lymphosarcoma and reticulosarcoma (SMR = 266.5, observed = 8,p <0.05).
To analyze cause-specific mortality by job held (potential exposure to diesel exhaust
emissions), 20 functional job titles were used, which were further grouped into three
potential categories: (1) high exposure, (2) low exposure, and (3) unknown exposure. A
person was classified in a job title if he ever worked on that job. Based on this classification
June 1990
9-9
DRAFT - DO NOT QUOTE OR CITE
-------
system, if a person had ever worked in a high-exposure job title he was included in that
group, even though he may have worked for a longer time in a low-exposure group or in an
unknown exposure group. Information on length of work in any particular job, hence
indirect information on potential length of exposure, was not available either.
For the high-exposure group a statistically significant excess was observed for
bulldozer operators who had 15 to 19 yr of membership and 20+ yr of follow-up for cancer
of the lung (SMR = 343.4,/? <0.05). This excess was based on 5 out of 495 deaths observed
in this group of 6712 individuals who contributed 80,327.6 person-years of observation.
The cause-specific mortality analysis in the lowdiesel exhaust exposure group revealed
statistically significant SMR excesses in individuals who had ever worked as engineers. These
excesses were for cancer of the large intestine (SMR = 807.2, observed = 3, p <0.05)
among those with 15 to 19 yr of membership and length of follow-up of at least 20 yr, and
cancer of the liver (SMR = 871.9, observed = 3, p <0.05) among those with 10 to 14 yr
membership and length of follow-up of 10 to 19 yr. There were 7032 individuals who
contributed to 78,402.9 person-years of observation in the low diesel exposure group.
For the unknown exposure group a statistically significant SMR was observed for
motor vehicle accidents only (SMR = 173.3, observed = 21, p <0.05). There were 3,656
individuals who contributed to 33,388.1 person-years of observation in this categoiy.
No work histories were available for those who started their jobs before 1967 and for
those who held the same job prior to and after 1967. This constituted 9,707 individuals (28
percent of the cohort) contributing to 104,447.5 person-years. Statistically significant SMR
excesses were observed for all cancers (SMR =112, observed = 339, p <0.05) and cancer
of the lung (SMR = 1193, observed = 141,/? <0.01). A significant SMR elevation was also
observed for cancer of the stomach (SMR = 199.1, observed = 30,p <0.01). There were
statistically significant SMR elevations at p <0.05 for 3 non-carcinogenic causes, benign
tumors, emphysema, and non-motor-vehicle accidents.
This study demonstrates a statistically significant excess for cancer of the liver, but
also shows statistically significant deficits in cancers of the large intestine and rectum. It may
be, as the authors suggested, that the liver cancer cases were actually cases resulting from
metastases from the large intestine and/or rectum, since tumors of these sites will frequently
June 1990
9-10
DRAFT - DO NOT QUOTE OR CITE
-------
metastasize to the liver. The excess in liver cancer mortality and the deficits in mortality that
are due to cancer of the large intestine and rectum could also, as the authors indicate, be
due to misclassification. Both possibilities have been considered by the investigators in their
discussion.
An important finding of the study is the positive trend by latency period as well as
by length of membership for SMR of emphysema. Emphysema is usually associated with
smoking, and with exposure to heavy dust, fumes, and gases. As most of these heavy
construction equipment workers work outdoors at construction sites, there is'a lot of
occupational exposure to various kinds of heavy dusts as well as exhaust fumes, and thus it
is not known what the contribution is, if any, of diesel exhaust to emphysema risk. The
confounding by smoking cannot be ruled out either, although a smoking habits survey was
carried out. This smoking survey has some limitations, however, which are described at the
end of the study.
Cancer of the lung also showed a positive trend with length of membership as well
as with latency although none of the SMRs were statistically significant except for the
workers without any work histories. The individuals without any work histories may have
been the ones who were in their jobs for the longest period of time, since workers without
job histories included those who had the same job before and after 1967 and thus may have
worked 12 to 14 yr or longer. If they had belonged to the category where heavy exposure
to diesel exhaust emissions was very common for this prolonged time, then the increase in
lung cancer as well as stomach cancer might be linked to diesel exhaust. Further
information on those without work histories should be obtained if possible since such
information may be quite informative with regard to the evaluation of the carcinogenicity
of diesel exhaust.
The study design is adequate, covers a long observation period, has a large enough
population, and is appropriately analyzed; however, it has too many limitations to permit any
conclusions. First, there are no exposure histories available. One has to make do with the
job histories which provide limited information on exposure level. Any person who ever
worked at the job or any person working at the same job over any period of time are in-
cluded in the same category; this would have a dilution effect, since extremely variable
June 1990
9-11
DRAFT - DO NOT QUOTE OR CITE
-------
exposures were considered in the study. The length of time worked in any particular job is
not available. Work histories were not available for 9707 individuals who contributed
104,447.5 person-years, a large proportion of the study cohort (28 percent). These
individuals happen to show the most evidence of a carcinogenic effect. Confounding by
alcohol consumption for cancer of the liver and smoking for emphysema and cancer of the
lung were not ruled out. Though 34,156 members were eligible for the study, the vital status
of 1765 individuals was unknown. Nevertheless, they were still considered in the
denominator of all the analyses. The investigators fail to mention how the person-year
calculation for these individuals was handled. Also, some of the person-years might have
been overestimated, as people may have paid the dues for a particular year and then left
work. These two causes of overestimation of the denominator may have resulted in some
or all the SMRs to be underestimated.
As for the smoking survey the investigators took a very small sample (133 out of
34,156, which was not even 1 percent). Of 133, only 107 (80 percent) participated. It was
a systematic sample, but the authors have neglected to mention how the list was prepared
for this systematic sample. Hence, the sample may not be representative of the study
population, and, with a small sample size the results are not generalizable. Lastly, the
questionnaire asked only for the current smoking histoiy. No detailed histoiy was obtained
for the amount smoked or length of smoking histoiy, both of which have bearing on
emphysema as well as lung carcinoma.
9.25. Edling et aL (1987): Mortality Among Personnel Exposed to Diesel
Exhaust
This is a cohort study of bus company employees, which investigated a possible
increased mortality in cardiovascular diseases and cancers from diesel exhaust exposure. The
cohort comprised all males employed in five different bus companies in southeastern Sweden
between 1950 and 1959. The total cohort of 694 men (after loss of 5 men to follow-up)
were divided into three exposure categories: clerks with the lowest exposure, bus drivers
with moderate exposure, and bus garage workers with highest exposure. Using information
from personnel registers, individuals were classified into one or more categories and could
June 1990
9-12
DRAFT - DO NOT QUOTE OR CITE
-------
have contributed person-years at risk in more than one exposure category. The study period
was from 1951 to 1983; information was collected from the National Death Registry, and
copies of death certificates were obtained from the National Bureau of Statistics. Workers
who died after age 79 were excluded from the study because diagnostic procedures were
likely to be more uncertain at higher ages (according to investigators). The cause-, sex-, and
age-specific national death rates in Sweden were applied to the 5-yr age categories of
person-years of observation to determine expected deaths for all causes, malignant diseases,
and cardiovascular diseases. A Poisson distribution was used to calculate />-values and
confidence limits for the ratio of observed to expected deaths.
The 694 men provided 20,304 person-years of observation with 195 deaths compared
to 237 expected. A deficit in cancer deaths largely accounted for this lower- than-expected
mortality in total cohort. Among subcohorts, no difference between observed and expected
deaths for total mortality, total cancers, or cardiovascular causes was observed for clerks
(lowest diesel exposure), bus drivers (moderate diesel exposure), and garage workers (high
diesel exposure). The risk ratios for all three categories were less than 1 except for
cardiovascular diseases among bus drivers which was 1.1.
When the analysis was restricted to members who had at least a 10-yr latency period
and either any exposure or an exposure exceeding 10 yr, similar results were obtained with
fewer neoplasms than expected while cardiovascular diseases showed risk around or slightly
above unity.
The small size of the cohort and poor data on diesel exhaust exposure are among the
major limitations of this study. Although lifetime occupational history was available, no
industrial hygiene data were presented to validate the classification of workers into low,
moderate, and high exposure to diesel exhaust based on job title. The power of the present
study was estimated to be 80 percent to detect a relative risk of 1.2 for cardiovascular
diseases and 1.4 for cancers, but for specific cancer sites, the power was much lower than
this. No information was available on confounding effects of smoking and asbestos exposure
at the work sites. In view of these serious limitations, there is insufficient evidence from the
study to confirm or refute the hypothesis that exposure to diesel exhaust is associated with
an increased risk of cancer and cardiovascular diseases.
June 1990
9-13
DRAFT - DO NOT QUOTE OR CITE
-------
9.2.6. Boffetta and Stellman (1988): Diesel Exhaust Exposure and
Mortality Among Males in the American Cancer Society Prospective
Study
Boffetta and Stellman conducted a mortality analysis of 46,981 males whose vital
status was known at the end of the first 2-yr follow-up. The analysis was restricted to males
aged 40 to 79 yr in 1982 who enrolled in the American Cancer Society's prospective
mortality study of cancer. Mortality was analyzed in relation to exposure to diesel exhaust
exposure and to employment in selected occupations related to diesel exhaust exposure. In
1982, more than 77,000 American Cancer Society volunteers enrolled over 1,200,000 men
and women from all SO states, the District of Columbia, and Puerto Rico in a long-term
cohort study, the Cancer Prevention Study II (CPS-II). Enrollees were usually friends,
neighbors, or relatives of the volunteers; enrollment was by family groups with at least one
person in the household 45 yr of age or older. Subjects were asked to fill out a four-page
confidential questionnaire and return it in a sealed envelope. The questionnaire included
history of cancer and other diseases; use of medications and vitamins; menstrual and
reproductive history; occupational history; and information on diet, drinking, smoking and
other habits. There were three questions on occupation: (1) current occupation, (2) the last
occupation, if retired, and (3) the job held for the longest period of time, if different from
the other two. Occupations were coded to an ad hoc two-digit classification in 70 categories.
Exposures at work or in daily life to any of the 12 groups of substances were also
ascertained. These included diesel engine exhausts, asbestos, chemicals/acids/solvents, dyes,
formaldehyde, coal or stone dusts, and gasoline exhausts. Volunteers checked whether their
enrollees were alive or dead and recorded the date and place of all deaths every second year
through. Death certificates were then obtained from state health departments and coded
according to a system based in the ICD-9 by a trained nosologist.
The data were analyzed to determine the mortality for all causes and lung cancer
in relation to diesel exhaust exposure, mortality for all causes and lung cancer and
employment in selected occupations with high diesel exhaust exposure, and mortality from
other causes in relation to diesel exhaust exposure. The incidence-density ratio was used as
a measure of association, and test-based confidence limits were calculated by the Miettinen
June 1990
9-14
DRAFT - DO NOT QUOTE OR CITE
-------
method. For stratified analysis, the Mantel-Haenszel method was used for testing linear
trends. Data on 476,648 subjects comprising 939,817 years of person risk were available for
analysis. Three percent of the subjects (14,667) had not given any smoking history, and 20
percent (98,026) of them did not give information on diesel exhaust exposure and were
therefore excluded from the main diesel exhaust analysis. Among individuals who had
provided diesel exhaust exposure history, 62,800 were exposed and 307,143 were not
exposed. Comparison of the population with known information on deisel exhaust exposure
with the excluded population with no information on deisel exhaust exposure showed that
the mean ages were 54.7 and 57.7 yr, the non-smokers were 72.4 percent and 73.2 percent,
and the total mortality rates/1000/yr were 23.0 percent and 28.8 percent, respectively.
The all cause mortality was elevated among railroad workers [RR (relative risk)
= 1.43, 95 percent CI = 1.2, 1.72], heavy equipment operators [RR = 1.7, 95 percent
CI (confidence internal) = 1.19,2.44], miners (RR = 1.34, 95 percent CI =* 1.06,1.68), and
truck drivers (RR = 1.19, 95 percent CI = 1.07, 1.31). For lung cancer mortality the risks
were significantly elevated for miners (RR = 2.67, 95 percent CI = 1.63, 4.37) and heavy
equipment operators (RR = 2.60, 95 percent CI = 1.12,6.06). Risks were also elevated but
not significantly for railroad workers (RR = 1.59, 95 percent CI=0.94, 2.69) and truck
drivers (RR = 1.24,95 percent CI=0.93,1.66). These risks were calculated according to the
Mantel-Haenszel method, controlling for age and smoking. Although relative risk was
nonsignificant for truck drivers, a small dose-response effect was observed when duration of
diesel exhaust exposure for them was examined. For drivers who worked for 1 to 15 yr RR
was 0.87, while for drivers who worked for more than 16 yr the RR was 1.33 (95 percent CI
= 0.64, 2.75). Relative risks for lung cancer were not presented for other occupations.
Mortality analysis for other causes and diesel exhaust exposure showed a significant excess
of deaths (p <0.05) in the following categories: cerebrovascular disease, arteriosclerosis,
pneumonia, influenza, cirrhosis of the liver, and accidents.
The two main methodologic concerns in this study are the representativeness of
the study population and the quality of information on exposure. The sample, though very
large, was comprised of volunteers thus the cohort, was healthier and less frequently exposed
to important risk factors such as smoking and alcohol. Self-administered questionnaires were
June 1990
9-15
DRAFT - DO NOT QUOTE OR CITE
-------
used to obtain data on occupation and diesel exhaust exposure. None of this information
was validated. Nearly 20 percent of the individuals had an unknown exposure status to
diesel exhaust, and they experienced a higher mortality for all causes and lung cancer than
both the diesel exhaust exposed and unexposed groups. This could have introduced a
substantial bias in the estimate of the association. While only 0.8 percent of the subjects
were lost to follow-up, the use of death certificates alone as a source of medical information
poses problems in accuracy and coding. But the authors report that cancer deaths are
routinely checked by histological confirmation from physicians or cancer registries. Given
the fact that all the diesel exhaust exposure occupations such as heavy equipment operators,
truck drivers, and railroad workers showed elevated lung cancer risk this study suggests a
causal association between the two.
9.2.7. Garshick et al. (1988): A Retrospective Cohort Study of Lung
Cancer and Diesel Exhaust Exposure in Railroad Workers
An earlier case-control study of lung cancer and diesel exhaust exposure in U.S.
railroad workers by these investigators had demonstrated a relative odds of 1.41 (95 percent
CI = 1.06, 1.88) for lung cancer with 20 yr of work in diesel exhaust exposure jobs. To
confirm these results a large retrospective cohort study was conducted by the same
investigators. Data sources for the study were the work records of the U.S. Railroad
Retirement Board (RRB). The cohort was selected based on job titles in 1959, which was
the year by which 95 percent of the locomotives in the United States were diesel powered.
Diesel exhaust exposure was considered to be a dichotomous variable depending on yearly
job codes between 1959 and death or retirement through 1980. Industrial hygiene
evaluations and descriptions of job activities were used to classify jobs as exposed or
unexposed to diesel emissions. A questionnaire survey of 534 workers at one of the
railroads was used to validate this classification. Employed workers 40 to 64 yr of age
starting work between 1939 and 1949 in the sampled job codes in 1959 and eligible for
railroad benefits were selected for the study. To qualify for benefits, a worker must have
10 yr or more of service with the railroad and should not have worked for more than 2 yr
in a non-railroad job after leaving railroad work. Workers with recognized asbestos exposure
June 1990
9-16
DRAFT - DO NOT QUOTE OR CITE
-------
because of passenger car, steam pipe, and railroad building construction and repairs were
not included among the job categories selected for study. However, a few jobs with some
potential for asbestos exposure were included in the cohort, and the analysis was done with
and without them.
The death certificates on all subjects identified in 1959 and reported by the RRB to
have died through 1980 were searched. Twenty-five percent of them were obtained from
the RRB and the remainder from the appropriate state department of health. Coding of
cause of death was done without knowledge of exposure history, according to the ICD-8.
If the underlying cause of death was not lung cancer, but was mentioned on the death
certificate, it was assigned as a secondary cause of death so that the ascertainment of all
cases was complete. Workers not reported by the RRB to have died by December 31,1980,
were considered to be alive. Deceased workers for whom death certificates had not been
obtained, or if obtained did not indicate cause of death, were assumed to have died of
unknown causes.
Proportional hazard models were fitted that provided estimates of relative risk for
death caused by lung cancer using the partial likelihood method described by Cox, and 95
percent confidence intervals were constructed using the asymptotic normality of the
estimated regression coefficients of the proportional hazards model. Exposure was analyzed
by diesel exhaust-exposed jobs in 1959 and by cumulative number of years of diesel exhaust
exposure through 1980. Directly standardized rate ratios for deaths from lung cancer were
calculated for diesel exhaust exposed compared with unexposed for each 5-yr age group in
1959. The standardized rates were based on the overall 5-yr person-year time distribution
of individuals in each age group starting in 1959. The only exception to this was between
1979 and 1980, when 2-yr person-year distribution was used. The Mantel-Haenszel analogue
for person-year data was used to calculate 95 percent confidence intervals for the
standardized rate ratios.
The cohort consisted of 55,407 workers; there were 19,396 deaths by the end of 1980.
Death certificates were not available for 11.7 percent of all deaths. Of 17,120 deaths for
whom death certificates were obtained, 48.4 percent were attributable to diseases of the
circulatory system while 21 percent were attributable to all neoplasms. Of all neoplasms 8.7
June 1990
9-17
DRAFT - DO NOT QUOTE OR CITE
-------
percent (1694 deaths) were due to lung cancer. A higher proportion of workers in the
younger age groups, mainly brakemen and conductors, were exposed to diesel exhaust while
a higher percentage of workers in the older age groups were potentially exposed to asbestos.
In a proportional hazards model, analyses by age in 1959 found a relative risk of 1.45
(95 percent CI = 1.11, 1.89) among the age group 40 to 44 yr and a relative risk of 1.33 (95
percent CI = 1.03, 1.73) for the age group 45 to 49 yr. Risk estimates in the older age
groups 50 to 54,55 to 59, and 60 to 64 yr were 1.2,1.18, and 0.99, respectively, and were not
statistically significant The two youngest age groups in 1959 had workers with the highest
prevalence and longest duration of diesel exhaust exposure and lowest exposure to asbestos.
When potential asbestos exposure was considered as a confounding variable in a
proportional hazards model, the estimates of relative risk for asbestos exposure were all near
null value and not significant. Analysis of workers exposed to diesel exhaust in 1959 (n =
42,535) excluding the workers with potential past exposure to asbestos, yielded the relative
risks of 1.57 (95 percent CI = 1.19, 2.06) and 1.34 (95 percent CI = 1.02, 1.76) in 40 to
44 yr and 45 to 49 yr age groups in 1959. Directly standardized rate ratios were also
calculated for each 1959 age group based on diesel exposure in 1959. The results obtained
confirmed those obtained by using the proportional hazards model.
Relative risk estimates were then obtained using duration of diesel exhaust exposure
as a surrogate for dose. In a model that used years of exposure up to and including
exposure in the year of death, no exposure duration-response relationship was obtained.
When analysis was done by disregarding exposure in the year of death and 4 yr prior to
death, the risk of dying from lung cancer increased with the number of years worked in a
diesel exposed job. In this analysis, diesel exposure was analyzed by exposure duration
groups and in a model entering age in 1959 as a continuous variable. The workers with
greater than 15 yr of exposure had a relative risk of lung cancer of 1.72 (95 percent CI =
1.27, 233). The risks for 1 to 4 yr of cumulative exposure was 1.20 (95 percent CI = 1.01,
1.44), for 5 to 9 yr of cumulative exposure it was 1.24 (95 percent CI = 1.06, 1.44), and for
10 to 14 yr of cumulative exposure it was 1.32 (95 percent CI = 1.13, 1.56). Directly
standardized rate ratios were also calculated for each 1959 age group based on diesel
June 1990
9-18
DRAFT - DO NOT QUOTE OR CITE
-------
exposure in 1959. The results obtained were confirmatory to those obtained by using the
proportional hazards model.
The results of this study demonstrating a positive association between diesel exhaust
exposure and increased lung cancer are consistent with the results of the case-control study
conducted by the same investigators in railroad workers dying of lung cancer from March
1981 through February 1982. This cohort study has addressed many of the weaknesses of
the other epidemiologic studies. The large sample size (60,000) allowed sufficient power to
detect small risks and also permitted the exclusion of workers with potential past exposure
to asbestos. The stability of job career paths in the cohort ensured that of the workers 40
to 44 yr of age in 1959 classified as diesel exhaust exposed, 94 percent of the cases were still
in diesel exhaust exposed jobs 20 yr later.
One of the limitations of this study was the lack of qualitative and quantitative data
on exposure. Number of years exposed to diesel exhaust was used as a surrogate for dose.
Another limitation of this study was the inability to examine the effect of years of exposure
and latency. The use of death certificates and the changes in diagnosis of lung cancer from
1959 through 1980 was a methodologic consideration left unaddressed by the authors.
Despite these limitations, the results of this study support the hypothesis that occupational
exposure to diesel exhaust is associated with a modest risk (1.5) of lung cancer.
93. CASE CONTROL STUDIES OF LUNG CANCER
93.1. Williams et al. (1977): Associations of Cancer Site With Occupation
and Industry From the Third National Cancer Survey Interview
This paper reports findings of the analysis of the Third National Cancer Survey
(TNCS). The lifetime histories, occupations, and industries were studied for associations
with specific cancer sites and types after controlling for age, sex, race, education, use of
cigarettes or alcohol, and geographic location. A total of 7,518 incident invasive cancers
occurring in the three years surveyed by the TNCS were interviewed. These comprised 57
percent of those eligible to participate. The interview included items on use of tobacco and
alcohol (by type, amount, and duration), family income, patient education, and employment
June 1990 9-19 DRAFT - DO NOT QUOTE OR CITE
-------
history. Actual descriptions of the occupation and industry were recorded by interviewers,
and according to the 1970 Census Coding Scheme were coded separately for main lifetime
employment, recent employment, and other jobs held. Occupations or industries were
combined to form larger groups. Coding of occupational and industrial labels in meaningful
job categories was done by one of the authors. Of the 3539 interviewed males and 3937
interviewed females 95 percent and 84 percent listed some main employment, respectively.
The basic analysis consisted of an intercancer comparison and involved comparing the
proportions of specific main lifetime industries and occupations among patients with cancer
at one site with those of patients having cancer at other sites combined as a control group,
and this was done using a series of Mantel-Haenszel stratified contingency table analyses to
yield odds ratios and Chi-Square values. Odds ratios (OR) were computed controlling for
age, race, education, tobacco, alcohol, and geographic location, and these were done
separately for males and females.
An excess risk for males for lung cancer was observed for the following main
industrial groups; mines [OR = 1.21], construction (OR = 1.24), transportation (OR =
1.17), utility and sanitary services (OR = 2.79,p <0.05), and professional (OR = 1.41). An
excess of bladder cancer was reported for the mining industry (OR = 1.61). For females
an excess of lung cancer was detected for the transportation industry (OR = 1.96); finance
and retail industry (OR = 1.73); and the business, car repair, and miscellaneous service
industry (OR = 2.29). None of these excesses were statistically significant. The transporta-
tion industry for males and females also showed a nonsignificant excess risk for cancers of
the liver and gall bladder ducts. When the analysis was done for specific lifetime industries,
railroad workers had an OR of 1.40 and truck drivers an OR of 1.34 for lung cancer for
males. Both these excesses were statistically nonsignificant. Other major findings of interest
were an excess of leukemia and multiple myeloma among sales personnel of both sexes and
an excess of women with lymphomas and Hodgkin's disease among the health care industry.
Cancer of the rectum in both men and women was associated with the retail trade, and
cancer of the prostate was more common among ministers, farmers, plumbers, rubber
workers, and coal miners.
June 1990
9-20
DRAFT - DO NOT QUOTE OR CITE
-------
The several strengths of the TNCS interview data set are its large size, histological
confirmation of nearly 95 percent of diagnoses, available information of occupation, and
details of confounding variables obtained by personal interview and ability to control for
them. Among its weaknesses are a 47 percent nonresponse rate and the fact that the
population surveyed came from predominantly urban areas which did not represent many
industries. Also, most of the associations observed did not achieve statistical significance
because they were based on small numbers of patients who had both specific cancers and
specific types of employment. Further, when multiple comparisons are made, some
significant associations arise by chance. The authors note that these results should only be
used as a research source for hypothesis generation and follow-up studies. This analysis does
suggest an association with lung cancer for three industries with potential diesel exhaust
exposure. These were trucking, railroading, and mining.
93.2. Hall and Wynder (1984): A Case-Control Study of Diesel Exhaust
Exposure and Lung Cancer
Hall and Wynder conducted a case-control study of 502 male lung cancer cases and
502 controls without tobacco-related diseases that examined an association between
occupational diesel exhaust exposure and lung cancer. Histologically confirmed primary lung
cancer patients who were 20 to 80 yr old were ascertained from 18 participating hospitals
in 6 U.S. cities, 12 mo prior to the interview. Eligible controls comprised of patients at the
same hospitals without tobacco-related diseases and were matched by age (±5 yr), race,
hospital, and hospital room status. Five hundred two male lung cancer subjects (64 percent
of those who met the study criteria for eligibility) were interviewed. Of the remaining 36
percent, 8 percent refused, 21 percent were too ill or had died, and 7 percent were
unreliable. Seventy-five percent of eligible controls completed interviews. Of these
interviewed controls, 49.9 percent were from the all cancers category, while 50.1 percent
were from the all-noncancers category. All interviews were obtained in hospitals to gather
detailed information on smoking history, coffee consumption, artificial sweetener use,
residential history, and abbreviated medical history as well as standard demographic
variables. Occupational information was elicited by a question on the usual lifetime
June 1990
9-21
DRAFT - DO NOT QUOTE OR CITE
-------
occupation and was coded by the abbreviated list of the U.S. Bureau of Census Codes. The
odds ratios were calculated to evaluate the association between diesel exhaust exposure and
risk of lung cancer incidence. Summary odds ratios were computed by the Mantel-Haenszel
method after adjusting for potential confounding by age, smoking, and socioeconomic class.
Two-sided, 95 percent confidence intervals were computed by Woolf s method. Occupation-
al exposure to diesel exhaust was defined by two criteria. First, occupational titles were
coded "probably high exposure" as defined by the industrial hygiene standards established
for the various jobs. The job titles included under this category were warehousemen, bus
and truck drivers, railroad workers, and heavy equipment operators and repairmen. The
second method used to analyze occupations by diesel exposure used the National Institute
for Occupational Safety and Health (NIOSH) criteria. In this method, the estimated
proportion of exposed workers was computed for each occupational category by using the
NIOSH estimates of the exposed population as the numerator and the estimates of
individuals employed in each occupational category from the 1970 census as the denomina-
tor. Occupations estimated to have at least 20 percent of their employees exposed to diesel
exhaust were defined as having "high exposure". Those with 10 to 19 percent of their
employees exposed were defined as having "moderate exposure" and those with less than 10
percent of their empolyees exposed as "low exposure".
Cases and controls were compared with respect to exposure. The odds ratio was 2.0
(95 percent CI = 1.2, 3.2) for those workers who were exposed to diesel exhaust versus
those who were not. The risk, however, decreased to a nonsignificant 1.4 when the data
were adjusted for smoking. Analyses by NIOSH criteria found a nonsignificant relative risk
of 1.7 in the highest probable exposure. There were no significantly increased cancer risks
by occupation either by the first method or by the NIOSH method. In order to assess any
possible synergism between diesel exhaust exposure and smoking, the lung cancer risks were
calculated for different smoking categories. No synergistic effects were observed.
The major strength of this study lies in the availability of detailed smoking history.
However, this is more than offset by the lack of diesel exhaust exposure measurements, lack
of consideration of latency period, and use of poor surrogate for exposure. Information was
collected on only one major lifetime occupation, and it is likely that those workers who had
June 1990
9-22
DRAFT - DO NOT QUOTE OR CITE
-------
more than one major job may not have reported the occupation, with the heaviest diesel
exhaust exposures. Occupational histories were obtained from self report and were not
validated with work records. This could have resulted in recall bias and misclassification of
exposure status.
933 Damber and Larsson (1987): Occupation and Male Lung Cancer
A Case-Control Study in Northern Sweden
A case-control study of lung cancer was conducted in northern Sweden to determine
the occupational risk factors that could explain the large geographic variations of lung cancer
incidence in that country. The study region comprised the three northern most counties of
Sweden with a total male population of about 390,000. The rural municipalities with IS to
20 percent of the total population have forestry and agriculture as dominating industries and
the urban areas have a variety of industrial activities (mines, smelters, steel factories, paper
mills, and mechanical workshops). All male cases of lung cancer reported to the Swedish
Cancer Registry during the 6-yr period between 1972 and 1977 who had died before the start
of the study, were selected. Of 604 eligible cases, 5 did not have microscopic confirmation
and in another 5 the diagnosis was doubtful, but these cases were included nevertheless.
Cases were classified as small carcinomas, squamous cell carcinomas, adenocarcinomas, and
other types. For each case a dead control was drawn from the National Death Registry
matched by sex, year of death, age, and municipality. Deaths in controls classified as lung
cancer and suicides were excluded. A living control matched to the case by sex, year of
birth, and municipality was also drawn from the National Population Registry. Postal
questionnaires were sent to close relatives of cases and dead controls, and to living controls
themselves to collect data on occupation, employment, and smoking habits. Replies were
received from 589 cases (98 percent), 582 dead controls (96 percent), and 453 living controls
(97 percent).
Occupational data were collected on occupations or employment held for at least 1
yr and included type of industry, company name, task, and duration of employment.
Supplementary telephone interviews were performed if occupational data were lacking for
any period between age 20 and time of diagnosis. Data analysis involved calculation of the
June 1990
9-23
DRAFT - DO NOT QUOTE OR CITE
-------
odds ratio by the exact method based on the hypergeometric distribution and the use of a
linear logistic regression model to adjust for the potential confounding effects of smoking.
Separate analyses were performed with dead and living controls, and on the whole there was
good agreement between the two control groups. A person who had been active for at least
1 yr in a specific occupation was in the analysis assigned to this occupation.
Using dead controls, the odds ratios adjusted for smoking were 1.0 (95 percent CI =
0.9,1.9) and 2.7 (95 percent CI = 1.2, 6.0) for professional drivers (>1 yr employment), and
for underground miners (>1 yr employment) respectively. For 20 or more years of
employment in those occupations, the odds ratios adjusted for smoking were 1.2 (95 percent
CI = 0.9, 2.6) and 2.7 (95 percent CI = 1.2, 6.0). These were the only two occupations
listed with potential diesel exhaust exposure. An excess significant risk was detected for
copper smelter workers, plumbers and electricians, as well as concrete and asphalt workers.
Occupational asbestos exposure was also associated with an elevated odds ratio of 2.6 (95
percent CI = 1.6, 3.6). All the odds ratios were calculated by adjusting for age, smoking,
and municipality.
This study did not detect any excess risk of lung cancer for professional drivers who,
amongst all the other occupations listed, have the most potential for exposure to motor
vehicle exhaust. However, we do not know if these drivers were exposed exclusively to
gasoline exhaust, diesel exhaust, or varying degrees of both. An excess risk was detected for
underground miners but it is not known if this is due to diesel emissions from engines or
from radon daughters in poorly ventilated mines. Although a high response rate (98
percent) was obtained for the postal questionnaires, the use of surrogate respondents is
known to lead to misclassification errors that can bias the odds ratio to 1.
93.4. Lerchen et al. (1987): Lung Cancer and Occupation in New Mexico
This is a population-based case-control study conducted in New Mexico which
examines the association between occupation and occurrence of lung cancer in Hispanic and
non hispanic whites. Cases involved residents of New Mexico, age 25 through 84 yr
diagnosed between January 1, 1980, and December 31, 1982, with primaiy lung cancer,
excluding bronchioalveolar carcinoma. Cases were ascertained through the New Mexico
June 1990
9-24
DRAFT - DO NOT QUOTE OR CITE
-------
Tumor Registry which is a member of the Surveillance Epidemiology and End Results
(SEER) Program of the National Cancer Institute. Controls were chosen by randomly
selecting residential telephone numbers and for those over 65 years of age, from the Health
Care Financing Administration's roster of Medicare participants. They were frequency-
matched to cases for sex, ethnicity, and 10 yr age category with a ratio of 1.5 controls per
case. The 506 cases (333 males and 173 females) and 771 controls (499 males and 272
females) were interviewed with a nonresponse rate of 11 percent. Next of kin provided
interviews for 50 percent and 45 percent for male and female cases respectively, while for
controls the corresponding interviews were provided by 2 percent for each sex. Data were
collected by personal interviews conducted by bilingual interviewers in the participants'
homes. A lifetime occupational history and a self reported history of exposure to specific
agents were obtained for each job held for at least 6 mo from age 12. Questions were asked
about the title of the position, duties performed, location and nature of industry, and time
at each job title. A detailed smoking history was also obtained. The variables on
occupational exposures were coded according to the Standard Industrial Classification
scheme by a single person and reviewed by another. To test the hypothesis about the high-
risk jobs for lung cancer, a priori listing of suspected occupations and industries was created
by a two-step process involving a literature review for implicated industries and occupations
by the principal investigator. The appropriate Standard Industrial Classification and
Standard Occupational Codes associated with job titles were also determined by the principal
investigator. For four agents — asbestos, wood dust, diesel exhaust, and formaldehyde — the
industries and occupations determined to have exposure were identified, and linking of
specific industries and occupations was based on literature review and consultation with local
industrial hygienists.
The relative odds were calculated for suspect occupations and industries classifying
individuals as ever employed for at least 1 yr in an industry or occupation and defining the
reference group as those subjects never employed in that particular industry or occupation.
Multiple logistic regression models were used to control simultaneously for age, ethnicity,
and smoking status. For occupations with potential diesel exhaust exposure the analysis
showed no excess risks for diesel engine mechanics and auto mechanics. Similarly, when
June 1990
9-25
DRAFT - DO NOT QUOTE OR CITE
-------
analysed by exposure to specific agents, the odds ratio adjusted for age, smoking and
ethnicity was not elevated for diesel exhaust fumes (OR = 0.6, 95 percent CI = 0.2, 1.6).
Elevated odds ratios were found for uranium miners (OR — 2.8, 95 percent CI = 1.0, 7.7),
underground mining (OR = 2.4, 95 percent CI = 1.2, 4.4), for construction painters (OR
= 2.4, 95 percent CI = 0.6, 9.6) and welders (OR = 4.3, 95 percent CI = 1.6, 11.0). No
excess risks were detected for the following industries: shipbuilding, petroleum refining,
construction, printing, blast furnace, and steel mills, and for the following occupations;
construction workers, painters, plumbers, paving equipment operators, roofers, engineers and
firemen, woodworkers, and shipyard workers. Females were excluded from detailed analysis
because none of the Hispanic female controls had been employed in high-risk jobs; among
the non-Hispanic white controls, employment in a high-risk job was recorded for at least five
controls for only two industries, construction and painting, for which the odds ratios were
not significantly elevated. Data on 333 male cases and 499 controls were therefore analysed.
Among the many strengths of this study are its population-based design, high
participation rate, detailed smoking histoiy, and the separate analysis done for the two ethnic
groups, southwestern Hispanic and non-Hispanic white males. The major limitations pertain
to the occupational exposure date. Job titles obtained from occupational histories were used
as proxy for exposure status, but these were not validated. Further, for nearly half the cases,
next of kin provided occupational histories. The authors acknowledge the above sources of
bias but state without substantiation that these biases would not strongly affect their results.
They also did not use a job exposure matrix to link occupations to exposures and do not
provide details on the method they used to classify individuals as diesel exhaust-exposed
based on reported occupations. The observed absence of an association for exposure to
asbestos, a well-established lung carcinogen, may be explained by the misclassification errors
in exposure status or by sample size constraints (not enough power). Likewise the
association for diesel exhaust reported by only 7 cases and 17 controls also may have gone
undetected because of low power. In conclusion, there is insufficient evidence from this
study to confirm or refute an association between lung cancer and diesel exhaust exposure.
June 1990
9-26
DRAFT - DO NOT QUOTE OR CITE
-------
93.5. Garshick et aL (1987): A Case-Control Study of Lung Cancer and
Diesel Exhaust Exposure in Railroad Workers
An earlier pilot study of the mortality of railroad workers by the same investigators
(Schenker et al., 1984) found a moderately high risk of lung cancer among the workers who
were exposed to diesel exhaust as compared to those who were not. This study was designed
to evaluate the feasibility of conducting a large retrospective cohort study. On the basis of
these findings the investigators conducted a case-control study of lung cancer in the same
population. The population base for this case-control study of lung cancer and diesel
exhaust was approximately 650,000 active and retired male U.S. railroad workers with 10 yr
or more of railroad service who were born in 1900 or after. The U.S. Railroad Retirement
Board (RRB), which operates the retirement system, is separate from the Social Security
System and to qualify for the retirement or survivor benefits the workers had to acquire the
service of 10 yr or more. Information on deaths that occurred between March 1, 1981, and
February 28, 1982, was obtained from RRB. For 75 percent of the deceased population
death certificates were obtained from RRB and for the remaining 25 percent they were
obtained from the appropriate state department of health. Cause of death was coded
according to the ICD-8. The cases were selected from deaths with primary lung cancer,
which was the underlying cause of death in most cases. Each case was matched to two
deceased controls whose dates of births were within 2.5 yr of the date of birth of the case
and whose dates of deaths were within 31 d of the date of death noted in cases. Controls
then were selected randomly from workers who did not have cancer noted anywhere on their
death certificates and who did not die of suicide or of accidental or unknown causes.
Each subject's work history was determined from a yearly job report filed by his
employer with RRB from 1959 until death or retirement. The year 1959 was chosen as the
effective start of diesel exhaust exposure for this study, since by this time 95 percent of the
locomotives in the United States were diesel powered. Investigators acknowledge that since
the transition to diesel powered engines took place in the early 1950s there were some
workers who had additional exposure prior to 1959; however, if a worker had died or retired
prior to 1959 he was considered unexposed. Exposure to diesel exhaust was considered to
be dichotomous for this study; which was assigned based on an industrial hygiene evaluation
June 1990
9-27
DRAFT - DO NOT QUOTE OR CITE
-------
of jobs and work areas. Selected jobs with and without regular diesel exhaust exposure were
identified by a review of job title and duties. Personal exposure was assessed in 39 job
categories representative of workers with and without diesel exhaust exposure. Those jobs
for which no personal sampling was done were considered exposed or unexposed on the
basis of similarities in job activities and work locations and by degree of contact with diesel
equipment. Asbestos exposure was categorized on the basis of jobs held in 1959 or on the
last job held if the subject retired before 1959. Asbestos exposure in railroads occurred
primarily during the steam engine era and was related mostly to the repair of locomotive
steam boilers that were insulated with asbestos. Smoking history information was obtained
from the next of kin.
Death certificates were obtained for approximately 87 percent of the 15,059 deaths
reported by RRB from which 1374 cases of lung cancer were identified. Fifty-five cases of
lung cancer were excluded from the study for either incomplete data (20) or refusal of
permission by two states to use information on death certificates to contact the next of kin.
Successful matching to at least 1 control with work histories was achieved for 335 (96
percent) cases <64 yr of age at death and 921 (95 percent) cases £65 yr of age at death. In
both age groups 90 percent of the cases were matched with 2 controls. There were 2385
controls in the study, 98 percent were matched within ±31 days of the date of death while
the remaining 2 percent were matched within 100 days. Deaths from diseases of the
circulatoiy system predominated among controls. Among the younger workers approximate-
ly 60 percent had exposure to diesel while among older workers only 47 percent were
exposed to diesel.
Analysis by a regression model, in which years of diesel exhaust exposure was the sum
total of the number of years in diesel-exposed jobs, used as a continuous exposure variable,
yielded an OR of lung cancer to be 1.39 (95 percent CI = 1.05,1.83) for over 20 yr of diesel
exaust exposure in the younger age group. After adjustment for asbestos exposure and
lifetime smoking (pack-years), the OR was 1.41 (95 percent CI = 1.06, 1.88). Increasing
years of diesel exhaust exposure categorized as £20 diesel years and 5 to 19 diesel-years with
0 to 4 yr as the referent group showed significantly increased risk in the younger age group
after adjusting for asbestos exposure and pack-year category of smoking for individuals who
June 1990
9-28
DRAFT - DO NOT QUOTE OR CITE
-------
had £20 yr of diesel exhaust exposure (OR = 1.64; 95 percent CI = 1.18, 2.29). Subjects
with 5 to 19 diesel years had a relative odds of 1.02 (95 percent CI = 0.72, 1.45). In the
older age group, only 3 percent of the workers were exposed to diesel exhaust for more than
20 yr. Relative odds for 5 to 19 yr and >20 yr of diesel exposure were less than 1 (p >0.01),
after adjusting for smoking and asbestos exposure.
Alternate models to explain post asbestos exposure were tested. These were variables
for regular and intermittent exposure groups and an estimate of years of exposure based on
estimated years worked prior to 1959. No difference in results were seen. The interaction
between diesel exhaust exposure and the three pack-year categories <50, >50, and missing
pack-years, were explored. The cross-product terms were not significant. A model was also
tested that excluded recent diesel exhaust exposure occurring within the 5 yr before death
and gave an odds ratio of 1.43 (95 percent CI = 1.06, 1.94) adjusted for cigarette smoking
and asbestos exposure.
The results of this study support the hypothesis that occupational exposure to diesel
exhaust increases lung cancer risk. The relative odds obtained for this association after
adjusting for smoking and asbestos exposure are low, less than 1.5 for the highest exposure
categoiy, but the 95 percent confidence interval is narrow.
The many strengths of the study are consideration of confounding factors such as
asbestos exposure and smoking; classification of diesel exhaust exposures by job titles and
industrial hygiene sampling; exploration of interactions between smoking, asbestos exposure,
and diesel exhaust exposure; and high ascertainment (87 percent) of death certificates from
the 15,059 deaths reported by the RRB.
The investigators also recognized and reported the following limitations: overestima-
tion of cigarette consumption by surrogate respondents that may have exaggerated the
contribution of smoking to lung cancer risk, and use of the ICC job classification as a
surrogate for exposure which may have lead to misclassification of diesel exhaust exposure
jobs with low intensity and intermittent exposure, such as railroad police, and bus drivers,
as unexposed. These two limitations would result in the underestimation of the lung cancer
risk. This source of error could have been avoided if diesel exhaust exposures were
categorized by a specific dose associated with a job title which could have been classified as
June 1990
9-29
DRAFT - DO NOT QUOTE OR CITE
-------
heavy, medium, low, and zero exposure instead of a dichotomous variable. The use of death
certificates to identify cases and controls may have resulted in misclassification. Controls
may have had undiagnosed primary lung cancer, and lung cancer cases might have been
secondary lesions misdiagnosed as primary lung cancer. However, the investigators quote
a third National Cancer Survey report in which the death certificates for lung cancer were
coded appropriately in 95 percent of the cases. Lastly, as in all previous studies, there is
lack of data on the contribution of unknown occupational or environmental exposures and
passive smoking. In conclusion, this study, compared with previous studies (on diesel
exposure and lung cancer risk), provides the most valid evidence that occupational diesel
exhaust emission exposure increases the risk of lung cancer.
9.3.6. Benhamou et aL (1988): Occupational Risk Factors of Lung Cancer
in a French Case-Control Study
This is a case-control study of 1625 histologically confirmed cases of lung cancer and
3091 matched controls, conducted in France between 1976 and 1980. This study was part
of an international study to investigate the role of smoking and lung cancer. Each case was
matched with one or two controls whose diseases were not related to tobacco use, sex, age
at diagnosis (±5 yr), hospital of admission, or interviewer. Information was obtained from
both cases and controls on place of residence since birth, educational level, smoking, and
drinking habits. A complete occupational history was obtained by asking participants to give
their occupations from the most recent to the first. Women were excluded because most
of them had no occupation. Men who smoked cigars and pipes were excluded because there
were very few in this category. Thus the study was restricted to nonsmokers and cigarette
smokers. Cigarette smoking exposure was defined by age at the first cigarette (nonsmokers,
£20 yr, or >20 yr), daily consumption of cigarettes (nonsmokers, <20 cigarettes a day, and
>20 cigarettes a day), and duration of cigarette smoking (nonsmokers, <35 yr, and >35 yr).
The data on occupations were coded by a panel of experts according to their own chemical
or physical exposure determination. Occupations were recorded blindly using the
International Standard Classification of Occupations. Data on 1260 cases and 2084 controls
June 1990
9-30
DRAFT - DO NOT QUOTE OR CITE
-------
were available for analysis. Of these, 365 cases and 1007 controls were excluded because
they did not satisfy the required smoking status criteria.
A matched logistic regression analysis was performed to estimate the effect of each
occupational exposure after adjusting for cigarette status. Matched relative risk ratios were
calculated for each occupation with the baseline category, which consisted of patients who
had never been engaged in that particular occupation. The matched relative risk ratios
adjusted for cigarette smoking for the major groups of occupations showed that the risks
were significantly higher for production and related workers, transport equipment operators,
and laborers (RR = 1.24,95 percent CI = 1.04,1.47). For occupations with potential diesel
emission exposure a borderline significant excess risk was found for motor vehicle drivers
(RR = 1.42, 95 percent CI = 1.07, 1.89) and transport equipment operators (RR = 1.35,
95 percent CI = 1.05, 1.75). No interaction with smoking status was found in any of the
occupations. The only other significant excess was observed for mines and quariymen (RR
= 2.14, 95 percent CI = 1.07, 4.31). None of the significant associations showed a dose-
response relationship with duration of exposure.
This study was designed primarily to investigate the relationship between smoking
(not occupations and environmental exposures) and lung cancer. While an attempt was
made to obtain complete occupational histories, the authors did not clarify if in the logistic
regression analysis, they used the subjects first occupation, predominant occupation, last
occupation, or ever worked in that occupation as the risk factor of interest. The most
important limitation of this study is that the occupations were not coded into exposures for
different chemical and physical agents thus precluding the calculation of relative risks for
diesel exposure. Using occupations as surrogate measures of diesel exposure, an excess
significant risk was obtained for motor vehicle drivers and transport equipment operators,
but not for motor mechanics. However, it is not known if subjects in these occupations
worked with diesel engines or nondiesel engines.
June 1990
9-31
DRAFT - DO NOT QUOTE OR CITE
-------
93.7. Hayes et aL (1989): Lung Cancer in Motor Exhaust-Related
(MER) Occupations
This study reports the findings from an analysis of pooled data from three lung cancer
case-control studies that examine in detail the association between employment in motor
exhaust related occupations and lung cancer risk adjusted for confounding by smoking and
other risk factors. The three studies were carried out by the National Cancer Institute in
Florida (1976 - 79), New Jersey (1980 - 81) and Louisiana (1979 - 83). These three studies
were selected because the combined group would provide a sufficient sample to detect a risk
of lung cancer in excess of SO percent among workers in motor exhaust-related occupations.
The analyses were restricted to males who had given occupational history. The Florida study
was hospital-based with cases ascertained through death certificates. Controls were
randomly selected from hospital records and death certificates, excluding psychiatric diseases
matched by age and county. The New Jersey study was population-based with cases
ascertained through hospital records, cancer registry, and death certificates. Controls were
selected from among the pool of New Jersey licensed drivers and death certificates. The
Louisiana study was hospital-based (it is not specified how the cases were ascertained), and
controls were randomly selected from hospital patients, excluding lung diseases and tobacco
related cancers.
A total of 2291 cases of male lung cancers and 2570 controls were eligible, and the
data on occupations were collected by next-of-kin interviews for all jobs held for 6 mo or
more, including the industry, occupation, and number of years employed. The proportion
of next-of-kin interviews varied by site between 50 percent in Louisiana to 85 percent in
Florida. The coding schemes were reviewed to identify motor exhaust-related occupations,
which included truck drivers, heavy equipment operators (cranes, bulldozers, and graders);
bus drivers, taxi drivers, chauffeurs, and other motor vehicle drivers; and automobile and
truck mechanics. Truck drivers were classified as routemen and delivery men and other
truck drivers. All jobs were also classified with respect to potential exposure to known and
suspected lung carcinogens. Odds ratios were calculated by the maximum likelihood method
June 1990
9-32
DRAFT - DO NOT QUOTE OR CITE
-------
adjusting for age by birth year, usual amount smoked, and study area. Logistic regression
models were used to examine the interrelationship of multiple variables.
A statistically significant excess risk was detected for employment of 10 yr or more
for all motor exhaust-related occupations (except truck drivers) adjusted for birth cohort,
usual daily cigarette use, and study area. The odds ratio for lung cancer using data gathered
by direct interviews was 1.4 (95 percent CI = 1.1, 2.0), allowing for multiple motor exhause,
related (MER) employment and 2.0 (95 percent CI = 13, 3.0) excluding individuals with
multiple MER employment. Odds ratios for all MER employment except truck drivers who
were employed for less than 10 yr, were 1.3 (95 percent CI = 1.0, 1.7) and 1.3 (95 percent
CI = 0.9, 1.8) for including and excluding multiple MER employment, respectively. No
association was observed with employment of less than 10 yr in these occupations. Odds
ratios were then derived for specific MER occupations, and, to avoid the confounding effects
of multiple MER job classifications, analyses were also done excluding subjects with multiple
MER job exposures. Truck drivers employed for more than 10 yr had an odds ratio of 1.5
(95 percent CI = 1.1, 1.9). A similar figure was obtained excluding subjects with multiple
MER employment. An excess risk was not detected for truck drivers employed less than ten
yr. The only other job category which showed statistically significant excess for lung cancer
was the one that included taxi drivers and chauffeurs who worked multiple MER jobs for
less than 10 yr (OR = 2.5, 95 percent CI - 1.4, 4.8). For the same category the risk for
individuals working in that job for more than 10 yr was 1.2 (95 percent CI = 0.5, 2.6). A
statistical significant trend (p <0.05) <2 yr, 2, 9 yr, 10 to 19 yr, and 20+ yr of employment
was observed for truck drivers but not for other MER occupations. A non-statistically
significant excess risk was also observed for heavy equipment operators, bus drivers, taxi
drivers and chauffeurs, and mechanics employed for 10 yr or more. All the above-
mentioned odds ratios were derived adjusted for birth cohort, usual daily cigarette use, and
state. Exposure to other occupational suspect lung carcinogens did not account for the
excess risks detected.
Results of this large study provide evidence that workers in motor exhaust related
jobs are at an excess risk of lung cancer that is not explained by their smoking habits or
exposures to other lung cancers. Since no information on type of engine use had been
June 1990
9-33
DRAFT - DO NOT QUOTE OR CITE
-------
collected, it was not possible to determine if the excess risk was due to diesel exhaust or
gasoline exhaust exposure or the mixture of the two. Among its limitations are possible bias
due to misclassification of jobs reported by the large proportion of next-of-kin interviews and
the problems in classifying individuals into uniform occupational groups based on the pooled
data in the three studies, which used different occupational classification schemes.
9.4. CASE-CONTROL STUDIES OF BLADDER CANCER
9.4.1. Howe et aL (1980): Tobacco Use, Occupation, Coffee, Various
Nutrients, and Bladder Cancer
This is a Canadian population-based case-control study conducted in the provinces
of British Columbia, Newfoundland, and Nova Scotia. These areas were selected because
they had cancer registries and were believed not to have concentrations of high-risk
industries. All patients with newly diagnosed bladder cancer occurring in the three provinces
between April 1974 and June 1976 were identified, and 77 percent of them were interviewed
at home. A total of 480 male case-control pairs and 152 female case-control pairs were
available for analysis. For each case one neighborhood control, matched by age (±5 yr) and
sex, was also interviewed at home to obtain data on smoking, occupation, dietary sources of
nitrites and nitrates which convert to nitrosamines (non-public water supply and preserved
meat products), and beverage consumption, including a detailed history of coffee
consumption. A detailed smoking history was obtained. The occupational history included
a chronologic account of all jobs and the number of years and months during which the
respondent had worked in each job, experience in industries that were suspected a priori to
increase the risk of bladder cancer, and exposure to any jobs that involved exposure to dust
and fumes at the work place. Relative risk estimates were computed using the linear logistic
model applied to individually matched case-control pairs.
A base-line comparison of cases and controls showed that while male patients were
similar to controls on income and education, there was an excess of female cases with low
family incomes and low levels of educational attainment. For both sexes the mean ages for
cases and controls did not differ and the times required for the interview were similar. An
June 1990
9-34
DRAFT - DO NOT QUOTE OR CITE
-------
analysis by the a priori suspect industries showed elevated risks for a number of industries
for males. These included chemical (RR — 7.5, 95 percent CI = 1.7, 67.6), rubber
(RR = 5.0, 95 percent CI = 0.6, 236.5), petroleum (RR = 53, 95 percent CI = 1.5, 28.6),
medicine (RR = 2.6, 95 percent CI = 0.9, 9.3), and the spray painting industry (RR = 1.8,
95 percent CI = 0.7, 4.6). The excess risks were statistically significant only for the
petroleum and chemical industry. The estimates did not change when the analysis was done
separately for subjects who reported only one exposure and for those who reported exposure
to more than one suspect industry. The estimates also remained unchanged after controlling
for smoking. Too few females reported working in the a priori suspect industries to make
any meaningful contribution to the analysis. A statistically nonsignificant excess risk was
observed for tanning, electric cable, photographic, commercial paint, tailoring, medicine,
food processing, and agricultural industries. The analysis by exposure to dust and fumes in
occupations other than those in the a priori suspect list detected the relative risks for diesel
and traffic fumes (RR = 2.8, 95 percent CI = 0.8, 11.8). Statistically significant excess risks
were observed for railroad workers (RR = 9.0, 95 percent CI = 1.2, 394.5) and welding
(RR = 2.8, 95 percent CI = 1.1, 8.8). For occupations other than those on the a priori list
for males and females, statistically significant excesses were detected for metal machinists
(RR = 2.7,95 percent CI = 1.1, 7.6), metal recorders (RR = 2.6,95 percent CI = 1.0, 7.3),
and nurseiy men (RR = 5.5, 95 percent CI = 1.2, 51.1). Statistically nonsignificant excesses
were also detected for exposure to two chemicals: benzidine and its salts, RR=1.3, and
bischoloromethyl ether, RR = 5.0. A detailed analysis was done for cigarette smoking,
which demonstrated statistically significant increasing bladder cancer risk with increasing
duration of smoking, total lifetime consumption of packs, and average frequency of cigarettes
per day. In males the highest significant risk was observed for latency of less than 35 yr.
Then the risk reduced slightly with increasing latency. In females the highest significant risk
was for more than 35 yr of latency. Risks were elevated for males consuming all types of
coffee and for females consuming instant coffee. Hair dye usage in females and phenacetin
usage in males and females carried no risk. Significant risks for use of artificial sweeteners
and use of nonpublic water supplies (nitrate and nitrite) were found among males only.
June 1990
9-35
DRAFT - DO NOT QUOTE OR CITE
-------
This study was mainly designed to evaluate the various risk factors for bladder cancer
such as smoking, coffee consumption, nitrates and nitrites in diet, etc. The major limitation
of this study, as the authors noted, was that the three selected provinces did not have high
concentrations of industries suspected to be linked to bladder cancer. An excess risk was,
however, detected for railroad workers and for those in the "exposed to diesel and traffic
fumes category." Risks for those exposed to "diesel fumes only" were not available nor do
we know the exact job title of the railroad workers and the type of engines they were
operating. The authors also did not detail the method by which they coded the information
given by respondents in response to questions on exposure to dust and fumes into the
various categories they used in the analysis. These analyses were done for subjects reported
having "ever been" exposed vs. "never been exposed" to these fumes, and, although detailed
chronological work histories were obtained, no attempt was made to develop a lifetime
cumulative exposure index to diesel fumes. In multiple logistic regression models authors
used the a priori high-risk occupations, hence nothing can be concluded about exposure to
diesel exhaust for occupations that were not a part of that list. The authors provide no
explanation on possible selection bias as only 77 percent of the eligible population was
included in the study. The nonstatistically significant excesses observed for exposure to
diesel and traffic fumes suggest the association that warrants further study.
9.4.Z Wynder eL al. (1985): A Case-Control Study of Diesel Exhaust
Exposure and Bladder Cancer
A case-control study of diesel exhaust exposure and bladder cancer risk was
conducted by Wynder et al. (1985). Cases and controls were obtained from 18 hospitals
located in 6 U.S. cities between January 1981 and May 1983. Cases were individuals with
histologically confirmed primary cancer of the bladder, diagnosed within 12 mo prior to the
interview. Controls were individuals with non-tobacco-related diseases and matched to the
case by age (within 8 yr), race, year of interview, and hospital of admission. Women were
excluded from the study since the focus was on male-dominated occupations. A structured
questionnaire was administered in the hospital to cases and controls to elicit information on
June 1990
9-36
DRAFT - DO NOT QUOTE OR CITE
-------
usual occupation, smoking history, alcohol and coffee consumption, as well as other
demographic factors.
Two methods were used to define occupational exposure to diesel exhaust. First,
occupational titles defined by the industrial hygiene standards as probable high exposure
were classified as exposed or not exposed to diesel exhaust. The probable high exposure
category consisted of bus and truck drivers, heavy equipment repairmen and operators,
railroad workers, and warehousemen. In the second method, guidelines set by NIOSH were
used to classify occupations based on the diesel exhaust exposures. In this method the
estimated proportion of exposed workers was computed for each occupational category by
using the NIOSH estimates of the exposed population as the numerator and the estimates
of individuals employed in each occupational category from the 1970 census as the
denominator. Occupations estimated to have at least 20 percent of their employees exposed
to diesel exhaust were defined as having "high exposure," those with 10 percent to 19 percent
of their employees exposed as having "moderate exposure," and those with less than 10
percent of their empolyees exposed as "low exposure." The odds ratio was used as a
measure of association to assess the relationship between bladder cancer and diesel exhaust
exposure. The overall participation among those eligible and available for interview was 75
percent and 72 percent in cases and controls, respectively.
One hundred ninety-four bladder cancer cases and 582 controls were examined, and
the two groups were found to be comparable by age and education. Except for railroad
workers who had relative odds of 2.0 (95 percent CI = 0.34, 11.61), the relative odds were
less than 1 for other occupations. None of the occupations with diesel exposure such as
truck drivers, railroad workers, and heavy equipment workers were found to be associated
with bladder cancer. When the risk was examined using the NIOSH criteria for high,
moderate, and low exposure, relative odds were 1.68 and 0.16 for high and moderate,
respectively, with low as the referent group, neither was statistically significant. Cases and
controls were compared by smoking status. Cases were more likely to be current cigarette
smokers than controls. Current smokers of 1 to 20 cigarettes/d had relative odds of 3.64 (95
percent CI = 2.04, 6.49), current smokers of 21+ cigarettes/d had relative odds of 3.51 (95
percent CI = 2.00, 6.19), while ex-smokers had relative odds of 1.72 (95 percent CI = 1.01,
June 1990
9-37
DRAFT - DO NOT QUOTE OR CITE
-------
2.92). After controlling for smoking there was no significant increase in the risk of bladder
cancer for occupations with diesel exhaust exposure compared to occupation without diesel
exhaust exposure. A synergistic effect between the two was also not detected.
There are two major methodologic limitations to this study, both pertaining to
exposure classification. First, the use of "usual" occupation may have lead to misclassifica-
tion of those individuals who had held a previous job with diesel exhaust exposure that was
not their usual occupation; this may have resulted in reduced power to detect weak
associations. Second, since there was no information on amount or duration of diesel
exhaust exposure, no analysis on dose-response relationship could be done. Also, no
information was available on other confounding risk factors of bladder cancer such as urinary
retention, amphetamine abuse, and smoking within the confined space of a truck cab, all of
which are life-style factors specific to the truck driving occupation.
9.4.3. Hoar and Hoover (1985): Truck Driving and Bladder Cancer
Mortality in Rural New England
This study investigated the relationship between the occupation of truck driving and
bladder cancer mortality in a case-control study in New Hampshire and Vermont. Cases
included all white residents of New Hampshire and Vermont who died from bladder cancer
ICD-8 between 1975 and 1979. Death certificates were provided by the Vital Records and
Health Statistics office of the two states, and the next of kin were traced and interviewed in
person. Two types of controls were selected for each case. One control was randomly
selected from all other deaths excluding suicides and matched on state, sex, race, age (±2 yr),
and year of death. The second control was selected with the additional matching criteria of
county of residence. Completed interviews were obtained from 325 (out of 410) next of kin
for cases and 673 (out of 923) controls. The odds ratio was calculated to ascertain a
measure of association between truck driving and bladder cancer. Since separate analyses
of the two control series gave similar results, the two control series were combined. Also,
since matched analyses yielded results similar to those provided by the unmatched analyses,
results of the unmatched analyses were presented.
June 1990
9-38
DRAFT - DO NOT QUOTE OR CITE
-------
Sixteen percent (35) of cases and 12 percent (53) of controls had been employed as
truck drivers yielding an OR of 1.5 (95 percent CI = 0.9, 2.6) after adjustment for county
of residence and age at death. For New Hampshire, the OR was 13 (95 percent
CI = 0.7, 2.3) and for Vermont the OR was 1.7 (95 percent CI = 0.8, 3.4). For a large
number of subjects the next of kin were unable to give the durations of truck driving, and
there was an inconsistent positive association with years of truck driving. Crude relative
odds were not altered after adjustment for coffee drinking, cigarette smoking, and education
as a surrogate for social class. Little variation in risks was seen when the data were analyzed
by the industry in which the men had driven trucks. No relationship was seen between age
at which employment as a truck driver started and occurrence of bladder cancer. Analysis
by duration of employment as a truck driver and bladder cancer showed a positive trend
showing increasing relative odds with increasing duration of employment. The odds ratio
was statistically significant for the 5 to 9 yr employment category only (RO = 2.9,95 percent
CI = 1.2, 6.7). Similarly, analysis by calendar year first employed showed a statistically
significant odds ratio for 1930 to 1949 (RO = 2.6, 95 percent CI = 1.3, 5.1), while relative
odds were not significant if they were employed prior to 1929 or after 1950.
The effects of reported diesel exhaust exposure from fuel or engines in truck driving
or other occupations were then analyzed. An OR of 1.8 (95 percent CI = 0.5, 7.0) was
derived for those who were exposed to diesel exhaust during their truck driving jobs as
compared to an OR of 1.5 (95 percent CI = 0.8, 2.7) for those not reporting diesel exhaust
exposure. Although none of the individual ORs in the duration categories (0,1 to 19 yr, 20
to 29 yr, 30 to 39 yr, and 40+ yr) were statistically significant, analysis by self reported
duration of exposure to diesel fuel or engines in any occupation and bladder cancer showed
a statistically significant positive trend.
This study has investigated an association between truck driving and bladder cancer
and was conducted in an attempt to understand the reasons for the high rates of bladder
cancer in a rural area of New Hampshire and Vermont. Although an elevated odds ratio
for bladder cancer (not statistically significant) was observed for reported truck driving
occupations there was insufficient evidence from this study to conclude that excess risk of
bladder cancer was due to exposure to diesel emissions. This is because the excess bladder
June 1990
9-39
DRAFT - DO NOT QUOTE OR CITE
-------
cancer risk was observed for all truck drivers irrespective of their exposure to diesel
emissions. Also, no information is provided on the confounding effects of other aspects of
the road environment such as urinary retention, amphetamine abuse, and concentrated
cigarette smoke within the truck cab. Other limitations of this study include the use of next
of kin for occupational histories who may either under- or overreport occupations and the
use of death certificates with their inherent problems of misclassification.
9.4.4. Iscovich et al. (1987): Tobacco Smoking, Occupational Exposure
and Bladder Cancer in Argentina
This is a hospital based case-control study of bladder cancer conducted in La Plata,
Argentina, to estimate the risk of bladder cancer associated with different types of tobacco,
beverages, and occupational exposures. Bladder cancer is one of the most common cancers
among males in the La Plata area.
Cases were selected from patients with a histologically confirmed diagnosis of bladder
cancer (transitional, squamous-cell, or nonspecific cell type) admitted to the 10 general
hospitals in the greater La Plata area (population in 1980 = 580,000) between March 1983
and December 1985. Cases with true bladder papilloma and individuals who were residents
of Greater La Plata <5 yr were excluded. Of the 120 cases who were eligible to participate,
1 died prior to the interview, 2 refused to participate and the remaining 117 cases,
representing 60 percent of the incident cases registered in the registry, were interviewed.
Two control groups (117 neighborhood and 117 hospital controls) were matched by sex and
age (±5 yr). Of one hundred seventeen, 99 were males and 18 were females. Hospital
controls were selected from the hospital as the case hospitalized for the first time within 3
mo of diagnosis of the case. Twelve percent of the hospital controls had illnesses known to
be associated with tobacco smoking. Neighborhood controls were sampled from among
persons living in the same block. The interviewer proceeded north in the street block where
the case resided and selected the first control who met the matching criteria. Seven hospital
controls could not be interviewed because of their poor physical health and were replaced.
Three neighborhood controls refused to participate and were replaced. Cases and hospital
controls were interviewed at the hospital and the neighborhood controls at their homes to
June 1990
9-40
DRAFT - DO NOT QUOTE OR CITE
-------
collect data on demographic, socio-economic, and medical variables, detailed smoking habits,
and alcoholic and other beverages consumed.
The interviews were done by trained interviewers, two physicians, and a social worker.
The cases and hospital controls were interviewed in the hospital by the physicians, hence,
the interviews could not be conducted "blind." A detailed occupational history was obtained
for the three occupations of longest duration and the most recent one. For each job title,
the activity of the plant and type of production was also ascertained. Job titles were coded
according to the International Labor Union (ILO) 1970 classification. Plant activity and type
of production were coded according to the United Nations 1975 classification categories.
Information was also collected on exposure to 33 chemical and physical agents, which
included confirmed or suspected bladder carcinogens. A detailed history of smoking habits
was also obtained, and individuals were categorized as current smokers if they were smoking
at present or if they had stopped smoking less than two yr previously. Ex-smokers were
those who ceased smoking at least or more than 2 yr previously. For each subject a
cumulative lifelong consumption of cigarettes by type was estimated and an average
consumption of cigarettes per day was computed.
Relative risks were computed for occupational factors using the unconditional logistic
regression method adjusting for age and tobacco smoking. These risks were derived for
those who were ever employed in that occupation vs. those who were never employed in that
occupation. Socioeconomic status of cases and neighborhood control was similar but there
were fewer professionals and more manual workers among hospital controls compared with
cases. Occupational variables included job title and type of activity of the plant. Significant
excess risks were observed for truck and railroad drivers (RR = 4.31, p <0.002) and oil
refinery workers (RR = 6.22, p <0.02). The risk for truck and railroad drivers was reduced
after adjusting for smoking while that for oil refinery workers increased after adjusting for
smoking. The adjusted RRs were not reported. A positive but nonsignificant association
was observed for printers (RR = 2.6,p <0.77).
This study identified smoking and coffee drinking as the major risk factors for bladder
cancer in this area. The overall age-adjusted RR in males for current smokers relative to
nonsmokers was 7.2 (95 percent CI = 3.0, 20.1) with dose-response relationships observed
June 1990
9-41
DRAFT - DO NOT QUOTE OR CITE
-------
for the average daily amount as well as for duration of smoking. A strong dose-response
relationship was also observed for coffee drinking in males with a RR of 12 (95 percent CI
= 4.3, 33.31) for those drinking more than three cups of coffee per day after adjusting for
the effect of smoking. No association was found with use of saccharin in males. No results
were presented for females for these risk factors.
This case-control study was conducted primarily to determine the reasons for the high
rates of bladder cancer in the La Plata region of Argentina. Only 60 percent of the cases
registered in the cancer registry were interviewed, and no information was provided for the
remaining 40 percent eligible nonrespondents to determine if the study sample was
selectively biased in any way. The sample size of 117 was small to start with. The analysis
of males reduced it to 99. While the use of two different types of control groups is a
strength of this study, none of the interviews were done "blind," and it appears that the
hospital interviews were done by the physicians and the neighborhood interviews were done
by the social worker. Job titles were used as surrogates of exposure, but the authors state
that although they did attempt to analyze by an exposure index derived from a job exposure
matrix (details not provided), they found no difference in exposure between cases and
controls. This explanation is ambiguous. The authors also grouped truck and railroad
drivers together for reasons not mentioned and did not present separate risk estimates. A
table of distribution of cases and controls for selected activities or professions did not state
if the data pertain to both sexes or males only, and the text did not clarify that either. The
reported significant excess risks for truck and railroad drivers were reduced after adjusting
for smoking but it was not known if the statistical significance persisted. No analysis was
provided for the data collected in the interviews on exposures to the 33 chemical and
physical agents and it was not known if the truck and railroad drivers were operating diesel
engines or not. Although rare in the La Plata area, the authors acknowledge the occupa-
tions known to be associated with bladder cancer (dye, rubber, leather, and textile workers).
June 1990
9-42
DRAFT - DO NOT QUOTE OR CITE
-------
9.4.5. Steenland and Burnett (1987): A Case-Control Study of Bladder
Cancer Using City Directories as a Source of Occupational Data
The primary objective of the study was to test the usefulness of city directories as a
source of occupational data in epidemiologic studies of illness and occupational exposure.
Commercial city directories indicating individual data on occupations and employer are
compiled from a door-to-door yearly census of all residents 18 yr old and older. They are
available in most medium-size cities in the United States. A unique advantage of city
directory data is that they identify specific employers and as the authors suggest may be
better than death certificates for rapid, inexpensive, record-based, epidemiologic studies.
A case-control study of male bladder cancer deaths in Hamilton County (including
Cincinnati), Ohio, was conducted. This county was selected because it is in an industrialized
area with high bladder cancer rates and also because city directories cover approximately
85 percent of the people living in this county. A computerized list of all male bladder
cancer deaths (n = 731) and all other male deaths (n = 95,057) with the exclusion of deaths
from urinary tract tumors and pneumonia, which occurred between 1960 and 1982 was
obtained from the Ohio Department of Vital Statistics. Death certificates had been coded
by a nosologist according to the ICD code in use at the time of death. A pool of six controls
was created for each case matched on sex, residence in Hamilton County at time of death,
year of death, age at death (±5 yr), and race. Two types of analysis were performed, one
based on city directory data and the other based on death certificate data. In the former,
cases and controls were restricted to individuals who had at least one yearly directory listing
with some occupational data. The first two controls from the pool of six who met the
requirements were selected. This analysis involved 648 cases (627 cases had 2 controls and
21 cases had only 1 control) and 1275 controls.
The death certificate analysis involved all 731 cancer deaths, with 2 controls per case.
In most cases, the same two controls were used in this analysis too. The usual lifetime
occupation and industry on the death certificate was abstracted from them. Data on
occupation and industry were coded with a 3-digit U.S. census code using the method
adopted by the U.S. Bureau of Census. Five percent of the occupational data were recorded
June 1990
9-43
DRAFT - DO NOT QUOTE OR CITE
-------
for occupation and industry by a second coder, with a high degree of reproducibility. Odds
ratios were calculated for bladder cancer using a Mantel-Haenszel procedure.
The city directory data identified four employers significantly associated with bladder
cancer deaths; only one of them was identified by the death certificate data, which provided
only lifetime type of industry rather than the name of a specific employer. The industries
represented by the four employers were a chemical plant, printing company, valve company,
and machinery plant. The city directory data analysis demonstrated significant positive
associations for quite a few occupations. The occupations that had at least 10 cases or more
were engineers (OR = 3.00, p = 0.01), carpenters (OR = 2.36, p <0.01), tailors (OR =
2.56, p <0.01), and furnace operators (OR = 2.5, p = 0.03). Relative odds were increased
significantly with increased duration of employment (£20 yr) for truck drivers (OR = 12,
p = 0.01) and furnace operators (based on 4 cases and no controls, p — 0.05). For
occupations in which subjects had ever been employed a significant increase in the relative
odds with increased duration of employment was observed for the railroad industry (>20 yr
of employment, OR = 2,21, p <0.05). Both truck driving and railroad industry occupations
involve diesel emission exposures. The analysis of the death certificate data yielded
associations in the same direction for most of the occupations. A check of the validity of
city directory data indicated that 77 percent of the listings agreed with the social security
earnings report for the employer in any given year. A comparison of city directory and
death certificate information on occupations indicated a match for occupation between at
least one city directory listing and occupation on death certificate for 68.3 percent of the
study subjects.
This study demonstrated that city directories are a relatively inexpensive and rapidly
accessible source of occupational data in epidemiologic studies. Its many limitations include
the problem in tracing women because of the change from maiden to married name, and
the availability of data for only the year of residence in the city. They are superior to death
certificates in being able to identify high risk employers in specific geographic sites. While
death certificate data reflect usual lifetime occupation, city directories yield data on short-
term jobs, some of which may involve critical exposure. Thus, a combination of the two
approaches may be most productive in record-based hypothesis-generating studies. In the
June 1990
9-44
DRAFT - DO NOT QUOTE OR CITE
-------
context of the mentioned pros and cons of using city directories, this study found an excess
risk for bladder cancer among two occupations with potential diesel exposure: truck drivers
and railroad workers. Two sources of bias that may have influenced these findings are
selection bias arising from the use of deaths instead of incident cases since survival for
bladder cancer is high and the absence of data on confounding factors such as smoking,
beverage consumption, and medication use.
9.5 SUMMARY
Certain extracts of diesel exhaust have been demonstrated to be both mutagenic and
carcinogenic in animals and in humans. Further, animal data are suggestive that diesel
exhaust is a pulmonary carcinogen among rodents exposed by inhalation to high doses over
long periods of time. It is presumed that diesel engine production will increase if another
fuel crisis requiring alternative fuels should occur in the future, since diesel engines are more
fuel efficient. This may result in even more excessive levels of diesel exhaust in the ambient
air. Since large working populations are currently exposed to diesel exhaust and since
nonoccupational exposures, currently are of concern as well, the possibility that exposure to
this complex mixture may be carcinogenic to humans has become an important public health
issue.
A major difficulty with the occupational studies considered here was the measurement
of the actual diesel exhaust exposure. Most studies compared men in job categories with
presumably some exposure to diesel exhaust with either standard populations (presumably
no exposure to diesel exhaust) or with men in other job categories from industries with little
or no potential for diesel exhaust exposure. A few studies have included measurements of
diesel fumes, but there is no standard method for the measurement. No attempt is made
to correlate these exposures with the cancers observed in any of these studies, nor is it clear
exactly which extract should be measured to assess the occupational exposure to diesel
exhaust. The occupations involving potential exposure to diesel exhaust are miners, truck
drivers, transportation workers, railroad workers, and heavy equipment operators.
June 1990
9-45
DRAFT - DO NOT QUOTE OR CITE
-------
With the exception of the study by Waxweiler et al. (1973), there have been no
known studies of miners to assess whether diesel exhaust is associated with lung cancer.
Virtually all miners (metal and coal) use diesel equipment, which was introduced in the early
1960s. Estimates of how many of these miners are exposed to diesel fumes are not
available.
The cohort studies mainly demonstrated an increase of lung cancer. Studies of bus
company workers by Waller (1981), Rushton et al. (1983), and Edling et al (1987) failed to
demonstrate any statistically significant excess risk of lung cancer, but these studies have
certain methodological problems such as small sample sizes, short follow-up periods, lack of
information on confounding variables and lack of analysis by duration of exposure or latency
that preclude their use in determining the carcinogenicity of diesel exhaust. Although the
Waller (1981) study had a 25-yr follow-up period the cohort was restricted to only employees
(ages 45 to 64) currently in service. Employees who left the job earlier, as well as those who
were still employed after age 64 and who may have died from cancer, were excluded.
Wong et al.,(1985) conducted a mortality study of heavy equipment operators that
demonstrated a significant increased risk of liver cancer in total and in various subcohorts.
The same analysis also showed statistically significant deficits in cancers of the large intestine
and rectum. Metastasis from the cancers of the large intestine and rectum in the liver
probably got misclassified as primary liver cancer which lead to an observed excess risk.
This study did demonstrate a nonsignificant positive trend for cancer of the lung with length
of membership and latency. Individuals without work histories who started work prior to
1967 when records were not kept may have been the ones who were in the same job for the
longest period of time. The workers without job histories included those who had the same
job before and after 1967 and thus may have worked about 12 to 14 yr longer, these workers
exhibited significant excess risks of lung cancer and stomach cancer. If this assumption
about their jobs is correct, then these site- specific causes can be linked to diesel exhaust
exposure. However, this study has quite a few methodological limitations such as the
absence of detailed work histories for 30 percent of the cohort and the availability of only
partial work histories for the remaining 70 percent, thus jobs were classified and ranked
according to presumed diesel exposure. Information is lacking regarding duration of
June 1990
9-46
DRAFT - DO NOT QUOTE OR CITE
-------
employment in the job categories (used for surrogate of exposure), and other confounding
factors (alcohol consumption, cigarette smoking, etc.). Thus, this study can not be used to
support a causal association or to refute the same between exposure to diesel exhaust and
lung cancer.
A twoyr mortality analysis of the American Cancer Society's prospective study by
Boffetta et al. (1988), after controlling for age and smoking, demonstrated an excess risk of
lung cancer in certain occupations with potential exposure to diesel exhaust These excesses
were statistically significant among miners (RR = 2.67, 95 percent CI = 1.63 to 4.37) and
heavy equipment operators (RR = 2.6, 95 percent CI = 1.12 - 6.06). The elevated risks
were nonsignificant in railroad workers (RR = 1.59) and truck drivers (RR = 1.24). A dose
response was also observed for the truck drivers. With the exception of miners, exposure
to diesel exhaust occurred in the three other occupations showing an increase in the risk of
lung cancer. Despite the methodologic limitations such as the lack of representiveness of
the study population composed of volunteers only and the questionable reliability of
exposure data based on self-administered questionnaires which were not validated, this study
is suggestive of a causal association between exposure to diesel exhaust and excess risk of
lung cancer.
There were two mortality studies conducted on railroad workers by Howe et al.
(1983) in Canada and Garshick et al. (1988) in the United States. The Canadian study
found relative risks of 1.2 (p <0.01) and 1.35 (p <0.001) among "possibly" and "probably"
exposed groups, respectively. The trend test showed a highly significant dose response
relationship with exposure to diesel exhaust and the risk of lung cancer. The main limitation
of the study was the inability to separate overlapping exposures of coal dust and diesel
fumes. Information on jobs was available at retirement only. There was also insufficient
detail on the classification of jobs by diesel exhaust exposure. The exposures could have
been nonconcurrent or concurrent, but, since the data are lacking, it is possible that
observed excess could be due to the effect of both coal dust and diesel fumes and not due
to just one or the other. However, it should be noted that, so far, coal dust has not been
demonstrated to be a pulmonary carcinogen in studies on coal miners. But lack of data on
June 1990
9-47
DRAFT - DO NOT QUOTE OR CITE
-------
confounders such as asbestos and smoking makes interpretation of this study difficult The
findings of this study are at the best suggestive of diesel exhaust being a lung carcinogen.
The most definitive evidence for linking diesel exhaust exposure to lung cancer comes
from a railroad worker study conducted in the United States. Relative risks of 1.57 (95
percent CI = 1.19 to 2.06) and 134 (95 percent CI 1.02 to 1.76) were found for ages 40 to
44 and 45 to 49, respectively, after the exclusion of workers exposed to asbestos. This study
also found that risk of lung cancer increased with increasing duration of employment. This
large cohort study with lengthy follow up and adequate analysis, including dose response
(based on duration of employment as a surrogate) as well as adjustment for other
confounding factors such as asbestos and smoking, makes the observed association between
increased lung cancer and exposure to diesel exhaust more meaningful.
Among the seven lung cancer case-control studies reviewed in this document, the
study by Lerchen et al. (1987) did not find increased risk of lung cancer, after adjusting for
age and smoking, for diesel fume exposure. The major limitation of this study was lack of
adequate exposure data derived from the job titles obtained from occupational histories.
Next of kin provided the occupational histories for 50 percent of the cases which were not
validated. The power of the study was small (analysis done on males only, 333 cases). On
the other hand statistically nonsignificant excess risks were observed for diesel exhaust
exposure by Williams et al. (1977) in railroad workers (OR = 1.4) and truck drivers (OR
= 1.34), by Hall and Wynder (1984) for workers who were exposed to diesel exhaust versus
who were not (OR = 1.4 and 1.7 with two different criteria), and by Damber and Larsson
(1987) in professional drivers (OR = 1.2). These rates were adjusted for age and smoking.
Both Williams et al. (1977) and Hall and Wynder (1984) had high nonparticipation rates of
47 percent and 36 percent respectively. In addition, the self-reported exposures used in the
the study by Hall and Wynder (1984) were not validated. This study also had low power to
detect excess risk of lung cancer for specific occupations.
The study by Benhamou et al. (1988), after adjusting for smoking, found significantly
increased risks of lung cancer among French motor vehicle drivers (RR = 1.42) and
transport equipment operators (RR = 1.35). The main limitation of the study was inability
to separate the exposures to diesel exhaust from those of gasoline exhausts since both motor
June 1990
9-48
DRAFT - DO NOT QUOTE OR CITE
-------
vehicle drivers and transport equipment operators probably were exposed to the exhausts
of both types of vehicles. Hayes et al. (1989) combined data from three studies (conducted
in three different states) to increase the power to detect an association of lung cancer with
different occupations that had high potential for exposure to diesel exhaust. They found that
truck drivers employed for more than 10 years had a significantly increased risk of lung
cancer (OR = 1.5, 95 percent CI = 1.1 to 1.9). This study also found a significant trend of
increasing risk of lung cancer with increasing duration of employment among truck drivers.
These relative odds were computed by adjusting for birth cohort, smoking, and state of
residence. The main limitation of this study is again the mixed exposures to diesel and
gasoline exhausts, since information on type of engine was lacking. Potential bias may have
been introduced because the way in which the cause of death was ascertained for the
selection of cases varied in the three studies. The methods used in these studies to classify
the occupational catagories are different hence probably leading to incompatability of
occupational catagories.
The most convincing evidence comes from the Garshick et al. (1987) case-control
study among railroad workers. After adjustment for asbestos and smoking, the relative odds
for continuous exposure were 1.39 ( 95 percent CI = 1.05 to 1.83). Among the younger
workers with longer diesel exhaust exposure, the risk of lung cancer increased with the dura-
tion of exposure after adjusting for asbestos and smoking. Even after the exclusion of recent
diesel exhaust exposure (5 yr before death) relative odds increased to 1.43 (95 percent
CI = 1.06 - 1.94). This study appears to be well conducted and well- analyzed case-control
study with reasonably good power. Potential confounders were controlled adequately, and
interactions between diesel exhaust and other lung cancer risk factors were tested.
Of the five bladder cancer case-control studies, three studies found increased risk in
occupations with a high potential diesel exhaust exposure. A significantly increased risk of
bladder cancer was found in Canadian railroad workers (RR = 9.0, 95 percent CI = 1.2 to
349.5; in Howe et al., 1980) and in Argentinean truck and railroad drivers (RR = 4.31,
p <0.002; Iscovich et al., 1987). Significantly increased risks were observed with increasing
duration of employment of £20 yr in truck drivers (OR =12,p= 0.01) and railroad industry
workers (OR = 2.21 ,p <0.05; Steenland and Burnett, 1987). No significant increased risk
June 1990
9-49
DRAFT - DO NOT QUOTE OR CITE
-------
was found for any diesel-related occupations in studies by Wynder et al. (1985) and by Hoar
and Hoover et al. (1985). All these studies had several limitations including inadequate
characterization of diesel exhaust exposure, lack of validation of surrogate measures of ex-
posure, and presence of other confounding factors (urinary retention, concentrated smoke
within the truck cab, etc.); most of them had small sample sizes, and none presented any
latency analysis.
In summary, an excess risk of lung cancer was observed in three out of seven cohort
studies and six out of seven case-control studies. Of these studies, two cohort and two case-
control studies observed a dose-response relationship using duration of employment as a sur-
rogate for dose. However, because of the lack of actual data on exposure to diesel exhaust
in these studies and other methodologic limitations such as lack of latency analysis, etc. the
evidence of carcinogenicity in humans is considered to be limited for diesel exhaust exposure.
June 1990
9-50 DRAFT - DO NOT QUOTE OR CITE
-------
9.6 REFERENCES
Benhamou, S.; Benhamou, E.; Flamant, R. (1988) Occupational risk factors of lung cancer in a French
case-control study. Br. J. Ind. Med. 45: 231-233.
Boffetta, P.; Stellman, S. D. (1988) Association between diesel exhaust exposure and multiple myeloma
an example of confounding. Prev. Med. 17: 236-237.
Damber, L A.; Larsson, L G. (1987) Occupation and male lung cancer A case-control study in northern
Sweden. Br. J. Ind. Med. 44: 446-453.
Edling, O.; Anjou, C-G.; Axelson, O.; Wing, H. (1987) Mortality among personnel exposed to diesel
exhaust Int. Arch. Occup. Environ. Health 59: 559-565.
Garland, F. C.; Gorham, E. D.; Garland, C. F.; Ducatman, A. M. (1988) Testicular cancer in US Navy
personnel. Am. J. Epidemiol. 127: 411-414.
Garshick, E.; Schenker, M. B.; Munoz, A.; Segal, M.; et aJ. (1987) A case control study of lung cancer and
diesel exhaust exposure in railroad workers. Am. Rev. Respir. Dis. 135: 1242-1248.
Garshick, E.; Schenker, M. B.; Munoz, A.; Segal, M.; et al. (1988) A retrospective cohort study of lung
cancer and diesel exhaust exposure in railroad workers. Am. Rev. Respir. Dis. 124: 820-825.
Gustafsson, L; Wall, S.; Larsson, L-G.; Skog, B. (1986) Mortality and cancer incidence among Swedish
dock workers - a retrospective cohort study. Scand. J. Work Environ. Health 12: 22-26.
Hall, N.E.; Wynder, E. (1984) Diesel exhaust exposure and lung cancer. Environ. Res. 34: 77-86.
Harris, J.E., (1983) Diesel emissions and lung cancer. Risk Analysis: Vol. 3, No. 2.
Hayes, R. B.; Thomas T.; Silverman, D.T.; et al. (1989) Lung cancer in motor exhaust occupations Am. J.
Ind. Med. 16: 685-695.
Hoar, S.; Hoover, R. (1985) Truck driving and bladder cancer mortality in Rural New England. J. Natl.
Cancer Inst. 74: 771-774.
Howe, G. R.; Burch, J. D.; Miller, A. B.; et al. (1980) Tobacco use, occupation, coffee, various nutrients,
and bladder cancer. J. Natl. Cancer Inst 64: 701-713.
Howe, G.R.; Fraser, D.; Lindsay, J.; et al. (1983) Cancer mortality (1965-77) in Relation to diesel fume and
coal exposure in a cohort of retired railway workers: 3NCI 70: 1013-1019. J. Natl. Cancer Inst
Iscovich, J.; Castelletto, R.; Esteve, J.; et al. (1987) Tobacco smoking, occupational, exposure and bladder
cancer in Argentina Int J. Cancer 40: 734-740.
Jensen, O. M.; Wahrendorf, J.; Knudson, J. B.; Sorensen, B. L (1987) The Copenhagen case-referent
study on bladder cancer. Risks among drivers, painters, and certain other occupations. Scand.
J. Work Environ. Health 13:129-134.
June 1990
9-51
DRAFT - DO NOT QUOTE OR CITE
-------
Lerchen, M. L; Wiggins, C. L; Samet, J. M. (1987) Lung cancer and occupation in New Mexico. J. Natl.
Cancer Inst 79: 639-645.
Kaplan, I. (1959) Relationship of noxiuos gases to carcinoma of the lung in railroad workers. J. Am. Med.
Assoc. 171: 97-101.
Raffle, P. A. B. (1957) The health of the worker. Br. J. Ind. Med. 14:73-80.
Risch, H. A.; Burch, J. D.; Miller, A. B.; Hill, G. B.; Steele, R.; Howe, G. R. (1988) Occupational factors and
the incidence of cancer of the bladder in Canada Br. J. Ind. Med. 45: 361-367.
Rushton, L; Alderson, M. R.; Nagarajah, C. R. (1983) Epidemiological survey of maintenance workers in
London Transport executive bus garages at Chiswick works. Br. J. Ind. Med. 40: 340-345.
Schenker, M. B.; Smith, T.; Munoz, A.; et al. (1984) Diesel exposure and mortality among railway workers;
results of a pilot study. Br. J. Ind. Med. 41: 320-327.
Silverman, D. T.; Hoover, R. N.; Mason, T. J.: Swanson, G. M. (1986) Motor exhaust-related occupations
and bladder cancer. Cancer Res. 46: 2113-2116.
Steenland, K.; Burnett, C.; Osoria, A. M. (1987) A case-control study of bladder cancer using city
directions as a source of occupational data Am. J. Epidemiol. 126: 247-257.
Stern, F. B.; Curtis, R. A.; Lemen, R. A. (1981) Exposure of motor vehicle examiners to carbon monoxide:
A historical prospective mortality study. Arch. Environ. Health 36: 59-66.
Vine is, P.; Magnani, C. (1985) Occupational and bladder cancer in males: A case-control study. Int. J.
Cancer 35: 599-606.10
Waller, R. E. (1981) Trends in lung cancer in London in relation to exposure to diesel fumes. Environ. Int.
5: 479-483.
Waxweiler, R. J.; Wagoner J. K.; Archer, V. E. (1973) Mortality of potash workers. J. Occup. Med. 15: 486-
489.
Williams, R. R.; Stegens, N. L; Goldsmith, J. R. (1977) Associations of cancer site and type with
occupation and industry from the Third National Cancer Survey interview. J. Natl. Cancer Inst 59:
1147-1185.
Wong, O.; Morgan, R. W.; Kheyts, L; et al. (1985) Mortality among members of a heavy construction
equipment operators' union with potential exposure to diesel exhaust emissions. Br. J. Ind. Med.
42: 435-448.
Wynder, L E.; Dieck, G.; Hall, N. E. (1985) A case control study of diesel exhaust exposure and bladder
cancer. Environ. Res. 34: 475-489.
June 1990
9-52
DRAFT - DO NOT QUOTE OR CITE
-------
10. WEIGHT OF EVIDENCE FOR CARCINOGENICITY
AND UNIT RISK ESTIMATES
10.1. INTRODUCTION
This chapter provides a quantitative assessment of carcinogenicity risk and calculates the unit
risk estimate of diesel engine emissions based on animal inhalation bioassay data. (Unit risk is a
characterization of the carcinogenic potency for a compound.) The unit risk estimate for an air
pollutant is defined as the increased lifetime cancer risk for an individual who is continuously exposed
to an air pollutant at a concentration of 1 jig/m3 in ambient air. We are now in the process of
acquiring human epidemiological data (Garshick et al., 1987, 1988) to perform cancer risk
calculations. The unit risk estimate calculated from these human data may provide a useful
comparison with that calculated from animal data.
Animal studies show a definite relationship between exposure to diesel engine emissions and the
induction of tumors, but issues remain as to their relevance to humans. Although several inhalation
studies of rats (Ishinishi et aL, 1986) reported a relationship between exposure to diesel exhaust and
induction of lung tumors, these effects have been interpreted (VostaJ, 1986) as being produced by
"particle overload" due to inhibition of particle clearance from the lungs at high exposure
concentrations. However, since diesel exhaust contains many known carcinogens, such as polycyclic
aromatic hydrocarbons (PAHs), long-term inhalation of such compounds could induce lung tumors
even without particle overload. Nevertheless, the potential effects of particles must be considered.
First of all, the site of particle deposition will determine, at least to some extent, the target site. The
delivered dose will vary with particle deposition efficiency. The delivered rate of particle absorbed
organics will differ from exposure to pure organics. Finally, particle-containing macrophages may
release factors capable of acting as either cocarcinogens or promoters.
The exact role of particles in the formation of lung tumors is not well understood. A pure
particle overload effect (Morrow, 1988) on the induction of lung tumors occurs in rats only when the
animals inhale "nuisance dust"[e.g., titanium dioxide (TiO^] at very high concentrations (Lee et aL,
1986). The particle burdens in the lung which led to lung tumor formation in the Ti02-exposed
animals were 30 to 60 times higher than those found in the groups in which lung tumors were
produced by high diesel exposure. The incidence of lung tumors did not increase when rats had
particulate lung burdens of that were 11 times greater than those producing lung tumors from diesel
exhaust soot (Table 10-1). Even the volumetric burden was higher by a factor of 11 to 20
May, 1990 10-1 DRAFT - DO NOT QUOTE OR CITE
-------
TABLE 10-1. PARTICULATE LUNG BURDEN AND LUNG TUMORS IN RATS
FOLLOWING A 2-YR EXPOSURE TO TITANIUM DIOXIDE (Ti(k) AND
DIESEL EXHAUST (MEAN FOR MALE AND FEMALE RATS)
Exposure
concentration
(mgftn3)
Panicle load
mg/g lung mmJ/g lung
Lung tumors
percent of rats
TiO,
0
IS
10
25.5
6.0
2.1
50
124
29.1
13
250
665
156.1
26*
Diesel
0
.
0.9
exhaust
0.35
ij.6
•} +
1.3
3.5
11.5
• —
3.6'
7.0
20.5
13.7
12.8*
' Significantly different from controls.
SOURCE: Lee et aL, 1985. 1986; Wolff et al.. 1987.
(156.1/13.7 to 156.1/7.7) in Tiot-exposed animals with lung tumors than in tumor-bearing
animals exposed to diesel exhaust. (Volumetric burden may be important in the functional
impairment of macrophages [Morrow. 1988]). A volumetric lung burden of Tio: particles
that was almost founold higner (29.1/7.7) than in tumor-Dearing animais exposed :o uiesei
exhaust did not lead to tumor formation.
This comparison suggests that the panicle overload effect on lung minor induction
depends heavily on the presence of organic material absorbed to the panicles. That is. the
panicle load in lungs in the rats of the diesel studies could have svnergistically influenced
the carcinogenic potency of the inhaled organic compounds associated with panicles by
increasing lung epithelial cell proliferation rates that are due to the release of mediators
from activated inflammatory cell, or by increasing the retention of organic compounds.
Recent results on the formation of benzo[a]pyrene (B[a]P) metabolites and DNA
adducts in lungs of rats after exposure to diesel exhaust and to B[a]P adsorbed onto carbon
black panicles suggest that inhaled organic compounds are related to the initiation of lung
-------
tumors (Wolf et al., 1989). Panicles may play an important role in the carcinogenic
potential of inhaled diesel exhaust in so far as they prolong the retention of the associated
organics. Indeed, early experiments by Saffiotti et al. (1964, 1965) and by Henry and
Kaufman (1973) and Henry et al. (1975) indicated that lung tumor incidences were increased
when B[a]P was intratracheally instilled adsorbed on panicles (Fe203 or carbon) as opposed
to instillation of the pure compound.
When administered alone, the high dose of B[a]P amving at a target cell over a short
period of time may overwhelm the cell's ability to metabolize B[a]P and to form carcinogenic
metabolites. In this case, much of the B[a]P is quickly cleared via the blood circulation
fWolff et al.. 1989). On the other hand, the slower release of B[a]P (i.e., a slower dose rate
delivered to target cells, over a longer penod of time in the case of B[a]P administered with
panicles) may allow more metabolites to form which are responsible for tumor induction.
Results on the formation of B[a]P metabolites after inhalation of the pure compound
vs. inhalation of B[a]P plus panicles show that the latter led to the formation of more
metabolites in the lung (Sun et aU 1984; Bond et al., 1986). In these studies, covalent
binding of B[a]P with macromolecules was increased 10- to 20-fold when B[a]P was adsorbed
to carbon black compared to inhalation of the pure compound. In addition, prolonged
delivery of B[a]P to target cells when inhaled with panicles imDlies that oossiblv more cells
m their sensitive phase can be exposed to the carcinogen, assuming a given cell proliferation
rate. This means that the duration of exposure is imponam in tumor induction. This issue
needs further investigation.
Understanding the roles played by panicles and organic compounds is essential for
biologically based dose-response modeling. In the absence of any scientific information, a
reasonable approach for the risk estimate is to assume that panicles only prolong the
retention of the associated organic compounds. It does not seem appropriate to assume that
panicles could act as a complete carcinogen. In this report, several approaches, with and
without adjusting for the overload effect, are used to calculate unit risk estimates.
Since little is known about the carcinogenic mechanism of inhaled diesel exhaust, we
cannot extrapolate with confidence the positive results of the rat inhalation studies to
May. 1990 IU-3 DRAb I — l>0 i\G7 QUOTE OR CITE
-------
humans. However, it may indicate the magnitude of risk associated with environmental or
occupational exposure of humans to diesei exhaust, assuming that humans respond like rats.
10.2. REVIEW OF PREVIOUS RISK ESTIMATES
Early attempts to quantitatively assess the carcinogenicity of diesei engine emissions
were hindered by a lack of well-controlled epidemiologic studies and long-term animal
cancer bioassavs. In attempts to surmount these obstacles, investigators used the
"comparative potency'1 method (Albert et al.. 1983) and an approach that involved estimating
:;sk from a nonoositive epidemiologic study (Hams. 1983).
In the comparative potency metnoa. a combustion product is selected that has a
previously determined potency estimate based upon epidemiologic data. The ratios of the
potency of this agent (e.g., coke oven emissions) to diesei paniculate matter extract in a
variety of mutagenicity and carcinogenicity tests are then multiplied by the epidemiology-
based potency estimate for coke oven emissions and averaged. If epidemiology-based
estimates from more than one pollutant are used, the derived potencies are generally
averaged to obtain an overall mean.
The comparative potency estimate of Albert et al. (1983) is probably best known.
Their results were obtained using epidemiology-based unit cancer risk estimates for coke
oven emissions, cigarette smoke condensate, and roofing tar. Samples of paniculate matter
were collected from three light duty-engines (a Nissan 220 C. an Oldsmobile 350. and a
Volkswagen turbocharged Rabbit), all run on a highway fuel economy test cycle and a heavy
duty engine (Caterpillar 3304), run under steady-state, low-load conditions.
The particulate matter was extracted with dichloromethane and tested in a variety of
assays. Those selected for development of comparative potency estimates included:
• Ames Salmonella typhimwium (TA98) reverse mutation:
• gene mutation in L5178Y mouse lymphoma cells:
o viral enhancement of chemical transformation in Synan hamster embryo cells: and
• Senear mouse skin tumor initiation.
May. 1990
10-4
DRAFT - DO NOT QUOTE OR CITE
-------
The first two studies were earned out in the presence ot metabolic activators. The potency,
defined as the slope of the dose-response curve, was measured for each sample m each
short-term assay.
The skin tumor initiation test gave definitive positive results only for the Nissan engine.
Since this was considered to be the most biologically relevant tesi. it was used to derive
potency estimates for the Nissan engine. This was achieved by multiplying the epidemiology-
based potency estimates for each of the three agents (coke oven emissions, roofing tar. and
cigarette smoke condensate) by the ratios of their potencies in the skin tumor initiation test
to that of the Nissan diesel. According to this method, the estimated lifetime cancer
risk/jig/m3 of extractable organic matter is as rollows:
Coke oven Rooting :ar Cigarette smoke
emissions condensate
2.6xlOa 5.2x10"* 5.4x10-*
The average of the three equals 4.4x10"*.
The potency of the other diesel emission samples was not estimated directly because
of the weak response in the skin tumor initiation test. Instead, their potency relative to the
Nissan engine was estimated as tne arithmetic mean c: :r.e:r roter.cy relative *" *-e Nissan
in the Salmonella assay m strain TA 98. the sister chromatid exchange (SCEX: assay in
Chinese hamster ovary (CHO) cells, and the mutation assay in mouse lvmpnoma ceils. The
estimated lifetime cancer risk per Mg/m3 extract was:
Volkswagen Oldsmobile Caterpillar
1.3x10"* 1.2x10"* 6.6x10*
To convert these values to a lifetime risk per ueym3 cf total particulate
matter, it was necessary to multiply these results by the fraction of extractable organic matter
in the panicles. These fractions were:
May.
iUo DRArT -- JO NOT OUUlt uK ^:IE
-------
Nissan Volkswagen Oldsmobile Caterpillar
0.08 0.18 0.17 0.27
After this adjustment, the resulting estimated potencies were equal to 3.5x10s. 2.3x10 s,
2.0xl0'5, and 1.8xl0"VMg/m3 of inhaled diesel paniculate matter.
Harris (1983) analyzed the data on the four engines used by Albert et al.
(1983), but used only two epidemiology-based potency estimates, those for coke oven
emissions and roofing tar. He used three of the same assays as did Albert et al. (1983) the
Senear mouse skin tumor initiation assay, enhancement of viral transformation in hamster
embryo cells, and the L5178 mouse lymphoma test. The third assay, however, was used both
•vith and without metabolic activation, while the Salmonella assay was not used. Hams also
obtained slightly different potency values trom the bioassavs. even though most of the same
data sets were used.
The diesel potency estimates were then derived by multiplying the potencies
for both coke oven and roofing tar by the ratio of their potencies compared to diesel in each
of the four bioassays. For example, the epidemiology-based potency for coke oven
emissions was estimated to equal 4.4xl0~7jig/m3. In the skin tumor initiation test, 2.1
papillomas per mouse were reported for the coke oven sample, compared with 0.53 for the
Nissan engine extract. The benzene extractable fraction was 0.06. The diesel potency
estimate using this comparison is then
(0.53/2.1)x0.06x4 4x 10"7jxg/m3
or 6.6xl0"6/Mg/rn3 paniculate matter. A total of eight comparisons were made for each
engine, four bioassays times two epidemiology-based potency estimates.
The Hams estimates are not comparable to those of Albert et al. without
adjustment. The unit risk estimates of Albert et al. are based upon absolute risk during
lifetime exposure, while Hams reported his values in terms at' relative risk per year of
exposure. To
,\lav. i#H) LU-o L)KAi*i - uJ
-------
adjust this to lifetime extra risk for continuous exposure,:: is necessary to multiply Harris's
values by a factor of
8.2 [= ' 24 h/8 h)x70x0.039] ,
where (24/8) adjusts from the 8-h work day to 24-h continuous exposure, 70 reflects the
lifetime exposure (70 yr), and 0.039 is the lifetime lung cancer mortality rate in the U.S.
population.
The range of unadjusted potencies included 0.02-0.06x10"* for the Nissan
sample. 0.01-0.24x 10"* for the Oldsmobile 350.0.02-2.78x 10"* for the Volkswagen Rabbit, and
0.01-0.25x 10"7/ig/nr paniculate matter tor the Cateirmiar sample. Hams (1983^ derived an
overall mean relative risk value of 3.5x iO'Yug/m*' for the three hght-dutv engines. Individual
mean values for each engine were not reported. .After conversion to a unit risk, the
estimated potency of the paniculate matter from the three engines was equal to 2.9x10'
7ug/m3.
McClellan (1986), Cuddihy et al. (1981, 1984), and Cuddihy and McQellan
(1983) reponed a potency factor of about 7.0xl0"s/Mg/m3 using a comparative potency
method similar to those reported above. The data base was similar to that used by Alben
st al. (1983) and Harns (1983), although there were some differences. In addition to the
epidemiology-based potency estimates tor coke oven emissions, roofing tar. ana cigarette
smoke condensate, the authors derived a human potency estimate for uroan soot. This
estimate was based upon an average air concentration of 0.06 mg/m3 and an annual cancer
incidence of 7/100,000. Interestingly, this is equivalent to a lifetime risk of 8.0x10'5//ig/m\
a value very similar to estimates denved for diesel paniculate matter.
Harris (1981) also assessed the risk of exposure to diesel engine emissions
using data from the London Transport Worker Study reported by Waller (1981). Five
groups of employees from the London Transport Authority (LTA) were used. These
included bus garage engineers, bus drivers, bus conducters. engineers in central works,
motormen. and guards. The first group was considered the highest exposed, the next two
intermediate, and the last two unexposed.
May. 1990 10-7 DRAFT -- JO NOT (JLU1 £ vjK U7E
-------
When cancer death rates for the high exposure group were compared with
those for London males, the observed to expected (O/E) ratios were less than 1. This was
thought to be due to a "healthy worker effect." Hams therefore merged the three exposed
groups and compared them with the two groups considered to be unexposed. An adjustment
was made for the estimated greater exposure levels of the garage engineers as compared to
bus drivers and conductors. The relative risk was greater than one but was statistically
significant only for garage engineers exposed from 1950 to 1960.
Hams (1983) identified a variety of uncertainties relative to potency
assessment based on this study. These included:
• Small unobserved differences m smojcine incidences among groups, which could have
a significant effect on lung cancer rates.
• Uncertainty about the magnitude of exposure in the exposed groups.
• Uncertainty regarding the extent of change in exposure conditions over time.
• Random effects arising from the stochastic nature of cancer incidence.
• Uncertainty in the mathematical specification of the model.
Taking these uncertainties into account, he derived a maximum likelihood relative risk
estimate of 1.23x10"* with a 95-percent upper confidence limit of 5xlOa'ug/m3 ^articulate
matter/yr. These estimates are equal to l.OxlO'3 and 4.1xl0'\ respectively, when convened
;o an absolute risk for lifetime exposure to 1 Mg/nr pamcuiate matter.
The potency estimates derived using various approaches are summarized in Table 10-2.
Estimates for the Caterpillar engine were not included, since it was not run under normal
use conditions. The largest estimate, 4.1xl0"J//ig/ni3, was derived by Harris (1983), using data
from the London Transpon Workers study. The smallest estimate. 2.5xl0"5/ng/m3. is the
mean from three light-duty engines evaluated by AJben et al. (1983) using the comparative
potency method.
Some of the imponant potential sources of error associated with the comparative
potency method include:
May. 1990
10-8
DRAFT -- DO NOT QUOTE OR CITE
-------
• Uncertainty regarding the composition of roofing tar and coke emissions.
• Omission of the possible effects of gaseous phase emissions are not included
in the potency estimates;
• Uncertainty regarding the epidemiology based potency estimates.
• Uncertainty regarding the bioavailabilty of the panicle adsorbed organics
following extraction with an organic solvent than in vivo.
• The assumption that short-term tests define carcinogenicity both
quantitatively and qualitatively.
TABLE 10-2.
ESTIMATED LIFETIME RISK OF CANCER FROM INHALATION OF
1 jig/m3 DIESEL PARTICULATE MATTER
Basis
Lifetime Risk
Comments
Reference
Comparative
3.5x10-5
Nissan
Albert et al..
potency
engine
1983
Comparative
16x10-5
Average of
Albert ct al..
potency
three engines
1983
Comparative
7.0x10-5
Lizht-Uutv
McClcllan ci
potency
engines
.lL :9S6
Comparative
19x10-4
Averaee oi
Hams. 1983
potency
three engines
Epidemiology
1.0x10-3
London irans-
Harris. 1983
port study
May. 1990
10-9
DRAFT -- DO NOT QUOTE OR CITE
-------
10.3. RISK CALCULATIONS BASED ON ANIMAL BIOASSAY DATA
10.3.1. Data Available for Risk Calculations
As reviewed in Chapter 7, five bioassay studies showed positive lung tumor response
in rats (Brightwell et al.. 1986; Ishinishi et al.. 1986: Iwai et aL 1986: Mauderly et aL 1987;
StOber, 1986). Only data from three studies (Tables 10-3 to 10-5) are selected for risk
calculations because each study consists of multiple exposure groups and thus is more
appropriate for risk calculations. The times to event (i.e.. death with or without tumors) are
available for the Mauaerlv et al. study. These time-to-event data are used in all risk
calculations, except when data from low exposure groups (i.e.. groups with the adjusted dose
less than 0.73 mg/mJ) from different studies are pooled (Table 10-6). The pooling of the low
exposure groups provides a data base tor calculating unit risk estimates for which the
panicle overload effect is minimized to the extent possible without relying on mathematical
modeling. The risk estimates calculated from these data serve a useful basis for comparison
with ones calculated with more mathematical sophistication.
10.3.2. Calculation of Unit Risks
The linearized multistage model is used to calculate unit risk estimates, using various
dose equivalence assumptions. Two approaches are used to calculate unit risks: :ne first
approach considers both the particle overload effect ana the amount of panicie-assoaated
organics in the lung, while the second approach uses the conventional approach to calculate
lung burden without considering either amount of particle associated organics in the lung or
the particle overload effect.
May. i990
ilMU DRAFT - DO NOT QUOTE OR CITE
-------
TABLE 10-3
INCIDENCE OF LUNG TUMORS IN FISHER 344 RATS (MALES AND FEMALES
COMBINED) EXPOSED TO DIESEL EXHAUST IN AIR
Experimental
Adjusted dose-1
Orgamcs dose
Lung tumor
dose (mg/mJ)
mg/mJ
mg/kg/d
(mg/cnr)b
incidence
rate®
0
0
0
0
2/230
033
0.073
0.049
2Jxl0*
3/223
3.47
0.73
0.48
3 6x10'
8/222
7.08
1.43
0.98
"3x105
29/227
'Dose (mg/kg/d) = adjusted dose (mg/mJ) x (0.2 m;/d)/0.3 ke. wnere 0.2 mJ is assumed to be the
daily air intake volume tor a 0.3-kg rai. and where adjusted dose = experimental dose x 7/24 x 5/7, reflecting
that animals were exposed only 7 h/d. 5 d/wk for life. These doses are used in the Approach 2 calculations.
They are calculated by using mathematical models and are used in the Approach 1 calculations.
cm1 is a lung surface area.
Ttme-io-tumor data are used in all risk calculations except for the pooled data calculation.
SOURCE; Mauderty et aL. 1987.
May. 1990
iU-ll
DRAFT -- DO NOT QUOTE OR CITE
-------
TABLE 10-4. INCIDENCE OF LUNG TUMORS IN FISHER 344 RATS (MALES
AND FEMALES COMBINED) EXPOSED TO HEAVY-DUTY ENGINE EXHAUST
Experimental Adjusted dose* Organics dose Lung tumor
dose (mg/m3) mg/m3 mg/kg/d (mg/cnr)b incidence rate'
0
0
0
0
1/123
0.46
0.26
0.18
9.6x1 cr
1/123
0.96
0.55
0.37
2.8x10-'
0/125
1.84
1.05
0.70
5.3x101
4/123
3.72
113
1 42
1.1x10"
8/124
' Dose (mg/kg/d) = adjusted dose (ragfin3)x(0.2 mJ/d)yO-3 kg, where 0.2 m3 is assumed to be the daily air
intake volume for a 03-kg rat. and where adjusted dose = experimental dose x 16/24 x 6/7. reflecting that
animal were exposed only 16 h/d, 6 d/weelc for life. These doses are used in Approach 2 calculations.
0 They are calculated by using mathematical models and are used in the Approach 1 calculations.
SOURCE: Isinishi et aL. 1986.
May. 1990
10-12 DRAFT -- DO NOT QUOTE OR LITE
-------
TABLE 10-5. INCIDENCE OF LUNG TUMORS IN FISHER 344 RATS (MALES
AND FEMALES COMBINED) EXPOSED TO DIESEL EXHAUST IN AIR
Experimental Adjusted dose. Orgamcs dose Lung tumor
dose (mg/ta3) mg/nr1 mg/kg/d (mg/cm:)b incidence
rate*
0
0
0
0
4/282
0.7
0.33
0.22
1.2x10°
1/144
22
1.05
0.70
5.2xl0"s
14/144
6.6
3.14
2.10
1.6x10-
55/143
1 Dose (mg/kg/d) = adjusted dose img/mn x 10.2 m\'d)/0.3 kg, wnere 0.2 is assumed to be the dailv air
intake volume for a 0.3-kg rat. and where adjusted dose = experimental dose x 16/24 % 5/7. reflecting that
animals were exposed only 16 h/d, 5 d/week for Itfe. These doses are used in the Approach 2 calculations.
3 They are calculated by using mathematical models and are used in the Approach 1 calculations.
SOURCE; Brightweil et aL. 1986.
May. 1990
10-13
DRAFT -- DO NOT QUOTE OR CITE
-------
TABLE 10-6. COMBINED DOSERESPONSE DATA FROM FOUR STUDIES
CONDUCTED ON FISHER 344 RATS'
Experimental
dose (mg/m3)
Adjusted dose
(mg/m11
Lung
tumors
Remarks/References
0
0.11
0
0.063
11/773
3/123
All control groups
Ishinishi et al.,1986
(light-duty engine)
035
0.41
0.073
0.23
3/223
1/125
Mauderly et al, 1987
Ishinishi et aL,1986
(light-duty engine)
0.46
0.26
1/123
Ishinishi, et al, 1986
(heavy-duty engine)
0.70
0.96
033
0.55
1/144
0/125
Brightwell et al, 1986
Ishinishi et al, 1986
(heavy-duty engine)
1.08
0.62
5/123
Ishinishi, et al, 1986
(light-duty engine)
3.47
0.73
8/221
Mauderly, et al, 1987
¦ Only those groups with an adjusted dose less than 0.73 rag/m3 are included in the table.
103.2.1. Approach 1
The deposition-clearance-retention model for inhaled diesel particles for the rat
(described in Appendix C) takes into account deposition efficiency and a retardation of
particle clearance from the alveolar region at high inhaled concentrations due to a particle
overload effect. This overload effect accounts for the difference (as shown in Figure 10-1)
between high and low exposure concentrations with respect to the organics lung burden (mg)
per unit of air concentration. When this approach is used, a retained dose (particles or
particle associated organics) per unit of epithelial surface area (or weight) of the lung can
be calculated for each exposure situations in different rat studies. In these calculations, the
mass fraction of the total absorbed organics in a diesel particle is assumed to be 20 percent,
May, 1990
10-14
DRAFT - DO NOT QUOTE OR CITE
-------
DRAFT -- DO NOT QUOTE OR CITE
C
O
c .—
® E
u
C 3
0 o
° O)
1 E
CO ^
C £
0) c
•o
3
43
O)
C
a
0.05'
0.04-
:.c3-
0C2-
0.01
0.00-
•0 01
3.47 mg/cu m
7.08 mg/cu m
a a a a a
0.01 ana 0.35 mg;cu m
Organics
=0 30 "00 "20 '40
Weeks after exposure
Figure 10-1. Lung burden (organic material, mg) in rats, under exposure scenano (7 h/d.
5 d/week) in Mauderiy et al. (1987).
ylav
lu*15 DRAFT -- DO NOT QUOTE OR CITE
-------
with half of this mass composed of slowiv ciearea argarucs and the other half composed of
quickly cleared organics.
As shown in Figures 10-1 and 10-2. the steady-state lung burdens are reached much
faster far organics than for the carbon core. The steady state in the lung of the retained
dose is reached very quickly for organics and wiil not change during further continuous
exposure over 2 yr. However, it is very likely that this continuing exposure after steady state
is reached will have a major influence on tumor induction. For panicles inhaled at a
concentration above a few mg/mJ. retarded clearance will become an increasingly important
factor for their accumulated dose over an extended exposure period. Since lung burden of
organics reaches steady state very quickly after exposure, the steady state value is used as
the target' dose m nsk calculations. The use of sucn a dosimetry is practically equivalent
:o the use of the area under the curve: a dosimetry often used in risk assessment.
For the reasons discussed previously in the introduction, it is reasonable to assume,
at present, that only organic material is carcinogenic. Assuming that the same accumulated
doses of the absorbed organics per lung epithelial surface area (pig/cm2) are equivalent
between animals and humans, unit risks are calculated, using tumor incidence data from
each of the three bioassays (Brightwell et alM 1986; Ishinishi et al., 1986: and the time to
event data from Mauderly et al., 1987) and the corresponding doses in terms of organics per
'une surface area. The resultant unit nsk estimates using the linearized multistage model
are given in Table 10-7, ranging from 5.5x10* to 2.6x10 * with a geometric mean of 1.3x10°.
In these calculations, the relationship between air concentration (jig/nr5) ana Jung ourden
(mg) in humans (Figure 10-3) is used to determine human lung burden due to 1 jig/m* of
diesel engine emissions. For instance, when data from Ishinishi et al. (Table 10-4) are used,
the carcinogenic slope (i.e. 95 percent upper confidence of the linear coefficient in the
multistage model) expressed in term of equivalent dose (jig/crrr) is 0.64 per ng/cm1. Since
1 fig/at3 of diesel engine emissions in air corresponds to a steady-state lung burden of 5.4 ng
per lung, or equivalently 8.6x10"* ng/crrr (5.4 ^g/627.000 cnr lung surface) of organics. the
lifetime risk due to 1 ug/m3 of diesel engine emissions in the ambient air is 5.5x10*
(0.64x8.6x10*).
.Vlay. 1990
10-16
DRAFT » DO NOT QUOTE OR CITE
-------
DRAFT -- DO NOT QUOTE OR CITE
Carbon
7.08 mg;cu m
3.47 mg/cu m
O)
0.01 and 0.35 mg/cu m
o
-0
= 3
Weeks after exposure
Figure 10-2- Cumulative lung burden (particles, mg) in rats, under exposure scenario C7 h/d.
5 d/week) in Mauderly et aL (1987).
May. 1990
10-17
DRAFT -
- JO NOT OL'OTE uK ciTE
-------
DRAFT -- 00 NOT G'JOTE OR CITE
6
1 mg/cu m
0-001. 0.01 & 0.1 mg/cu m
5
Organics
(A) 24 hrs^day, 7 days/week
4
(B) 12 hrsjday, 7 days/week
iC) 8 hrs./day. 5 davs/weeK
» 3
1 mg/cu m
ch
0.01 & 0.1 mg/cu m
2
1
0.01, 0.1 & 1 mg/cu m
0
2
0
4
6
8
10
Year
Figure 10-3. Lung burden (organics. mg) in human adults over years after exposure.
May. 1990
10-18
DRAFT -- DO NOT QUOTE OR CITE
-------
TABLE 10-7. UNIT RISK ESTIMATES PER ug/m3 OF DIESEL EXHAUST
Equivalent dose assumption
Air Body Organic per
Data base concentration surface* lung surfaceb
Mauderly et al., 1987
1.4x10-*
3.7x10"*
2.6x10 s
Ishinish et al., 1986
3.2xl0's
8.5x10 s
5.5x10s
Brightwell et al., 1986
7.3x10 s
1.9X104
1.6xl0's
Geometric mean (rats)
6.9xl05
1.8x10"*
1.3x10 s
Pooled Fischer rat data'
2.3x10'5
6.1x10 s
4.0X10"6
* If body weight, instead of body surface, basis is used, the unit risk estimates will be
reduced by a factor of 6.
b If organic per lung weight, instead of organics per surface, is used as equivalent dose,
those risk estimates would be reduced by a factor of 4.
c Data from exposure groups in all studies in which the adjusted concentration was equal
to or less than 0.73 mg/m3 were used.
May, 1990
10-19
DRAFT - DO NOT QUOTE OR CITE
-------
103.22. Approach 2
Risk estimates are calculated under three dose equivalence assumptions: namely, mg/m3
(referred to as air concentration equivalence), mg/kg per day (referred to as body weight
equivalence), and mg/W20 per day (referred to as body surface equivalence), where w is
body weight in kg. Under a dose equivalence assumption, the same unit of dose (or
exposure) is assumed to be equally potent in inducing tumor responses between animals and
humans. For instance, when mg/kg per day (i.e., body weight basis) is assumed to be an
equivalent dose, it would require (as to be illustrated below) a 2.3 times higher air
concentration for humans than for animals to induce the same tumor responses. When the
air concentration is 1 Mg/m3* the body burdens for rats and humans are calculated to be:
1 #ig/m3xl0'3 mg/ng x (0.20 m3/d)/0.30 kg = 0.00067 mg/kg per day,
where 0.20 m3/d is the daily air intake volume for a 0.30 kg-rat and,
1 jig/m3xl0'3 mg/jig*(20 m3/d)/70 kg = 0.00029 mg/kg per day,
where 20 m3/d is the daily air intake volume for a 70-kg man.
Thus, under the assumption that mg/kg per day is an equivalent dose, 23
(0.00067/0.00029) times more tumors would be induced in rats than in humans when both
rats and humans were exposed to diesel emissions at the same air concentration. Note that,
in these calculations, the particle overload effect is not considered and tumor response is
assumed to be linearly proportional to the body dose.
The unit risk estimates based on the air concentration and the body surface dose
equivalence assumptions are listed in Table 10-7. The unit risk estimates on the body weight
basis (not presented here) can be derived from those based on the body surface equivalence
by dividing by a factor of 6. One of the primary potential sources of error in estimating risk
based upon animal exposure to inhaled particulate matter involves slowing of clearance due
to lung overloading. The longer residence time can increase bioavailability by allowing for
greater desorption of the organic compounds from the particles. The higher concentrations
used in these studies were in the range where slower lung clearance would be expected. To
partially avoid this source of error and to obtain a better-defined upper-bound estimate, lung
May, 1990 10-20 DRAFT - DO NOT QUOTE OR CITE
-------
rumor respohses were pooled from all groups in all the rat studies in which the adjusted
exposure concentration was equal to or less than 0.73 mg/m3. The pooled estimate of
2.3x10 s (under the air concentration equivalence assumption) and 6.1xl0"5 (under the body
surface equivalence assumption) are based upon data from 773 control and 1208 exposed
animals.
The geometric mean of the risk estimates calculated by Approach 1 is comparable
to that calculated by Approach 2 when pooled data are used and air concentration dose
equivalence is assumed (Table 10-7). In general, Approach 1 produces smaller risk
estimates than those calculated by Approach 2, with the largest difference (by about an
order of magnitude) occurring when body surface dose equivalence is assumed and non-
pooled data are used. The pooled data produce comparable risk estimates to those
calculated by Approach 1 because the overload effect is considered in both situations.
In the Approach 1 calculations, the amount of organics per lung surface (ng/cm2) is
assumed to be dose equivalent. If, instead, the amount of organics per lung weight (Mg/g)
is assumed to be equivalent, the unit risk estimate will be reduced by a factor of about 4,
the ratio of rat surface area (4090 cm^/human surface area (627,000 cm2) and rat lung
weight (1.6 gram)/human lung weight (1065 gram).
10.4. DISCUSSION: UNCERTAINTIES AND RESEARCH NEEDS
An ideal approach to estimate cancer risk due to diesel exhaust exposure would be to
consider the delivered dose rate of organic material together with particles as well as the
effects of these compounds on initiation, cell proliferation, and progression rates. If such
information is available, the models proposed by Moolgavakar and Knudson (1981) and
Chen and Farland (1989) can be used to calculate the dose-response relationship. The
observation from dose and time to tumor relationship from Mauderly et al. (1987), where
tumor incidence increases drastically over time after exposure in the high exposure group
but not in the low- and mid-exposure groups, suggests that diesel emissions increase cell
proliferation rate with an increase of exposure concentration. (As shown in Table 10-3, the
high dose group has a significantly higher proportion of animals that developed tumors than
May, 1990 10-21 DRAFT - DO NOT QUOTE OR CITE
-------
the other treatment groups. More than 90 percent of tumors in the high dose group were
observed after 700 d.) This speculation is based on a theoretical demonstration
(Figure 10-4) that, when the mitotic rate for initiated cells is doubled from 0.14/d to 0.28/d
and other mode] parameters remain constant, tumor incidence is drastically increased over
time under a two-stage model of carcinogenesis proposed by Chen and Farland (1989),
which is a modified version of the one proposed by Moolgavkar and Knudson (1981). The
increase of proliferation rate may be due to organic material and particles acting individually
or jointly. Since particles probably cannot act as complete carcinogens, organic material is
used as the "target" dose in the risk calculations, assuming that particles play a role in the
carcinogenic potential of inhaled diesel exhaust only as a vehicle that allows organics to be
slowly released (Approach 1). It would be interesting to conduct an initiation-promotion-
progression type of study to determine whether or not particles can promote organics-
initiated cells and whether the rate of progression is affected by the particle-associated
organics.
A prolonged retention of particles occurring during impaired lung clearance does not
prolong the retention of the adsorbed PAH because it is cleared from particles and absorbed
quickly compared to the long retention half time of particles. Thus, the rate of delivery of
the PAH to target cells may be of importance to metabolize these substances. If this delivery
rate is temporarily very high, interaction of the PAH with target cells may become less
effective. However, more PAH can be metabolized when the same total amount is delivered
at a slower rate over a longer time (i.e., when it is slowly released from particles). Studies
have shown that B[a]P in association with particles forms more metabolites and induces
more lung tumors than does B[a]P alone. These findings are consistent with the concept of
rate-dependent absorption.
Several other data gaps must be filled to improve the risk extrapolation from animals
to humans. Experimental studies are needed in particular on mechanistic aspects of diesel
exhaust carcinogenicity, since extrapolation modeling from one species to another can best
be performed when the underlying mode of action is known. The question of tumor sites
and their histologic differences (cells of origin) between rats and humans needs to be
addressed. The delivery rate of carcinogens to specific cell types needs to be estimated, and
May, 1990 10-22 DRAFT - DO NOT QUOTE OR CITE
-------
m: mitotic rate,
per cell per day
06
Q.
04
m = 0.14
02-
0
150
300
450
600
Time, t days
Figure 10-4 Effect of increasing tne mitotic rate of tnitiatea ceils on tne tumor incidence rate over age
wnen other parameters are held constant a hypothetical calculation based on a two-stage model
proDOsed by Chen and Fariand (1989)
May, 1990
10-23
DRAFT - DO NOT QUOTE OR CITE
-------
normal and initiated cell proliferation rates as well as an effect of exposure on the
proliferation rates needs to be investigated. Since we know from the rat studies that alveolar
macrophages are activated and that their ability to clear panicles is impaired at high inhaled
concentrations, other important cellular defense mechanisms might be affected already at
low concentrations (e.g., killer cell activity). The generation of mediators from inflammatory
cells (e.g., oxygen radicals) could be an important mechanistic aspect in lung tumor
induction.
Ongoing studies with inhaled pure carbon black particles will shed light on the role
of carbonaceous particles perse in lung tumor induction. The potential of diesel exhaust to
act as a promoter on both normal and initiated cells deserves attention, and studies on DNA
interactions (DNA-adduct formation) and on effects of normal cellular repair mechanisms
will be important. Since rats, mice, and hamsters vary in their sensitivity to carcinogens,
comparative studies would be useful to elucidate underlying mechanisms. Knowledge of
these carcinogenic mechanisms will eventually lead to an improved extrapolation model for
the pulmonary carcinogenic effect of inhaled diesel exhaust from rat to humans. Finally,
some of the rate constants used in this paper (Appendix C) may need to be confirmed,
although the model adequately predicts available data on the lung burden of particles.
10.5 Weight-of-Evidence for Carcinogenicity of Diesel Engine Emissions
and Recommended Unit Risk Estimates
Inhalation of whole diesel exhaust resulted in the induction of lung tumors in F344
rats (Mauderly et al., 1987; Ishinishi et al., 1986; Iwai et aL, 1986; Brightwell et aL 1986; in
Wistar rats Heinrich et al., 1986b; in NMRI mice (Heinrich et al., 1986b); and in Senear
mice (Pepelko and Peirano, 1983). Lung tumors were also induced by implantation of diesel
exhaust condensate in Osborne Mendal rats (Grimmer et al., 1987). Skin painting of diesel
particle extracts induced dermal tumors in strain "A" mice (Kotin et al., 1955) and in Senear
mice following promotion with TPA (Nesnow et al., 1982). Extensive studies using
Salmonella have unequivocally demonstrated direct-acting mutagenic activity in both
particulate and gaseous fractions of diesel exhaust. Positive results have also been reported
May, 1990
10-24
DRAFT - DO NOT QUOTE OR CITE
-------
for gene mutations and chromosome effects in mammalian cell systems after exposure to
diesel products.
Based upon the induction of lung tumors via inhalation in two strains of rats and two
strains of mice; the induction of lung tumors by implantation of diesel exhaust condensate,
subcutaneous tumors following injection of exhaust particule extracts, and skin tumors
following dermal application of exhaust extracts; and supported by positive mutagenicity
results, the evidence for carcinogenicity of diesel exhaust in animals is considered to be
sufficient.
Among human populations potentially exposed to diesel engine exhaust, an excess
risk of lung cancer was observed in three mortality studies (Garshick et al., 1988; Howe et
al., 1983 and Boffetta et al, 1988) and in six case-control studies (Garshick et al., 1987;
Benhamon et al.. 1988: Hayes et al., 1989; Damber and Larsson, 1987; Hall and Wynder,
1984 and Williams et al., 1977). A dose response was observed in two of the cohort and two
of the case-control studies. Although an increased risk of lung cancer was not detected in
a few studies (Waller et al., 1981; Rushton et al., 1983; Edling et al., 1987; Wong et al., 1985
and Lerchen et al., 1987), these studies had several methodologic limitations such as small
sample sizes, short follow up, lack of adjustment for confounding factors, etc.
Collectively, the epidemiologic studies show a positive assocation between diesel
exhaust exposure and cancer. However, because of the uncertainties created by lack of
actual diesel exhaust exposure data in these populations, the evidence for carcinogenicity of
diesel engine emissions in humans is considered to be limited.
On the basis of limited evidence for carcinogenicity in humans, diesel engine
emissions are classified as category B1 according to U.S. EPA cancer assessment guidelines.
This classification is supported by sufficient evidence in animals and positive results in
mutagenicity studies and is consistent with the presence of known carcinogens such as
benzo(a)pyrene on diesel particles. Chemicals classified in category B1 are considered to
be probable human carcinogens.
Previous quantitative risk estimates for diesel engine emissions were based upon
either the comparative potency method or upon epidemiology data. In the comparative
potency method, the effectiveness of diesel exhaust was compared to a related environmental
May, 1990 10-25 DRAFT - DO NOT QUOTE OR CITE
-------
pollutant, i.e., coke oven emissions, using short term in vivo and in vitro tests. The relative
potency of diesel exhaust under these test conditions was then multiplied by the
epidemiology based cancer unit risk estimate for the related pollutant. Confidence in this
method is limited by the aggregate uncertainty of the epidemiology based assessments for
the related pollutants, by the possibility that the relative potency of the agents compared
may be different in short-term tests than in chronically exposed humans, and by the use of
extracts rather than whole particles in most of the short-term tests. Confidence in the Harris
(1983) epidemiology-based cancer potency estimate is limited because deaths from cancer
in exposed workers did not significantly exceed those of the general population, because of
limited data regarding exposure levels and changes in exposure with time, because of
possible group differences in smoking rates, because of random effects, and because of
uncertainties regarding mathematical specification of the model used.
In the derivation of newer potency estimates, several potential uncertainties may still
remain, even with the availability of adequately designed and conducted chronic bioassays.
One present in all animal data-based assessments is the potential difference in sensitivity of
animals vs. humans. Of special concern during diesel exhaust exposure is the inhibition of
particle clearance at the high concentrations used in the animal bioassays. This results in
greater lung particle burden/exposure concentration ratios than at lower exposure levels.
Since the particles could conceivably act as tumor promoters or even as cocarcinogens, unit
risk estimates derived from high dose bioassays may overestimate cancer potency at ambient
concentrations. Slower particle clearance rates can also result in slightly increased
bioavailability of organics even though the half-time for desorption of organics is less than
that for normal particle clearance. Finally, the unit risk is based upon particle
concentrations and does not account for possible effects of the gaseous fraction of exhaust.
Even though carcinogenic effects of the fraction have not been detected, the gases could
influence the potency of the particulate matter fraction in some as yet undetected manner.
Two approaches were used in an attempt to circumvent as many of these sources of
error as possible. In one approach, inhibition of clearance at high exposure concentrations,
with an accompanying excess particulate matter lung burden, was to some degree mitigated
by the use of pooled, relatively low level exposure data. The responses at low exposure
May, 1990 10-26 DRAFT - DO NOT QUOTE OR CITE
-------
concentrations, however, were small and showed considerable variation. The other approach
utilized a sophisticated lung dosimetry model that accounted for human-animal differences
in regional lung deposition, lung surface area, and effects of panicle clearance rates upon
the bioavailability of the organic fraction. The precision of the estimate was enhanced
further by the use of time-to-tumor data in one of the studies. The two approaches were
also combined in that the dosimetry model was used with the pooled, low-level exposure
data.
Any of the three unit risk estimates derived from the Mauderly et al. (1987), Ishinishi
et al. (1986), and Brightwell et al. (1986) studies, using the dosimetry model (Approach 1),
based upon the concentration of organics per unit lung surface area, are considered
acceptable. These estimates vary from 5.5x10"" to 2.6x 10'\ with a geometric mean of 1.3x10
5. They are considered to be more appropriate than the body surface area method used
by EPA in the past, because they provide a more accurate estimate of the delivered dose
of organics per surface area of lung rather than lung weight-based estimates and are
recommended because they most closely reflects dose at the critical target organ. The
estimate based on pooled data from low-level exposures is not recommended because of the
low and variable responses in the different studies. Among these estimates, the geometric
mean value of 1.3xl0*s is considered slightly superior because it represents the largest data
base. Among those based on individual studies, the one using Mauderly et al (1987) has the
advantages of the longest exposure duration as well as data allowing use of a time-to-tumor
model. The estimate of 2.6x10 s based on this data set was also, perhaps fortuitously, exactly
the same as the comparative potency estimate reported by Albert et ai. (1983).
Whereas estimates based upon the recommended method do not account for
overload of the insoluble carbon particle core, there are little data that would allow
adjustments to be made. It is even possible that only minimal adjustment would be
necessary. Pott and Stober (1983), for example, reported that intratracheal instillation of
particulate matter did not enhance the carcinogenic potency of benzo(a)pyrene. Moreover,
since mechanical clearance of panicles from the alveolar regions is normally much slower
in humans than in rats, lung burdeniexposure concentration ratios for humans exposed at
low doses may not differ that greatly from rats exposed at much higher concentrations.
May, 1990 10-27 DRAFT - DO NOT QUOTE OR CITE
-------
Confidence in the recommended unit risk estimates are increased by the facts that
(1) They are based upon several well designed and conducted chronic animal bioassays
(2) time-to-tumor data were available in at least one of the bioassays (3) the major factors
responsible for the internal dosimetry of the putative carcinogenic organics are accounted
for (4) the risk estimate is based upon whole particulate matter rather than extracts.
Because of remaining uncertainties including possible differing sensitivities between
humans and rats, the possible effects of the inorganic fraction of the inhaled paniculate
matter, and possible effects of the gaseous components, confidence in the recommended unit
risk estimate as an upper bound is considered to be medium rather than high.
10.6. SUMMARY
On the bases of limited evidence for carcinogenicity of diesel engine emissions in
humans, supported by adequate evidence in animals and positive mutagenicity data, diesel
engine emissions are considered to best fit the weight-of-evidence category Bl. Agents
classified into this category are considered to be probable human carcinogens.
Two approaches and several different equivalence dose assumptions are used to
calculate unit risk estimates. Approach one accounted for animal-human differences in
particle deposition efficiency, lung surface area, as well as normal and impaired clearance
rates upon the lung burden of organics. In approach two, the same exposure level based
upon either air concentration, body weight or body surface area was considered to be equally
potent for inducing tumors. When particle overload effects of impaired clearance upon lung
burden of organics, by using doses at which overload is minimized (pooled data), or by doing
both, the following unit risk estimates were calculated.
3.0X10"6, on organics per lung weight basis, by approach 1"
1.3x10 s, on organics per lung surface basis, by approach 1*
l.OxlO"6, on organics per lung weight basis, pooled data, by approach 1
4.0x10"6, on organics per lung surface basis, pooled data, by approach 1
l.OxlO"6, on body weight basis, pooled data, by approach 2
May, 1990
10-28
DRAFT -- DO NOT QUOTE OR CITE
-------
2.3x10 s, on air concentration basis, pooled data, by approach 2
6.1xl0'5, on body surface basis, pooled data, by approach 2
•Geometric mean of unit risks from three studies
Of these unit risk estimates, the first two are the most reasonable because effects of particle
overload upon the dose of bioavailable. particle associated-organics are accounted in the
calculations. Although pooled data provide comparable risk estimates, it is questionable that
pooling data from different studies is appropriate because these studies were not conducted
under the same protocol, and secondly because the responses were quite small and variable.
Of the first two estimates, the one based on organics per unit lung surface area (1.3 x 10"
5/Mg/m3 is considered to be the most accurate because it most closely reflects concentration
at the critical target organ.
May, 1990
10-29
DRAFT -- DO NOT QUOTE OR CITE
-------
10.7 REFERENCES
Albert, R. E.; Lewtas, J.; Nesnow, S.; Thorslund, T. W.; Anderson. E. (1983) Comparative potency method
for cancer risk assessment: Application to diesel particulate emissions. Risk Anal. 3: 101-117.
Benhamou, S.; Benhamou, £.; Flamant. R. (1988) Occupational nsk factors of lung cancer in a
French case-control study. Br. J. Ind. Med. 45: 231-233.
Boffetta, P.; Stellman, S. D. (1988) Association between diesel exhaust exposure and multiple
myeloma an example of confounding. Prev. Med. 17: 236-237.
Bond, J. A.; Sun, J. D.; Medinsky, M. A.; Jones. R. K.; Yeh, H. C (1986a) Deposition, metabolism and
excretion of l-[HC]nitropyrene and l-(HC]nitropyrene coated on diesel exhaust particles as influenced
by exposure concentration. Toxicol Appl. Pharmacol. 85: 102-117.
Bond. J. A.: Sun, J. D.; Mitchell, C E-; Dutcher. J. S.: Wollf. R. K.; McClellan, R. O. (1986b) Biological fate
of inhaled organic compounds associated with paniculate matter. In: Lee. S.D.: Schneider. T.; Grant,
L. D.; Verker, P. J.; eds. Aerosols. Lewis Publisher, pp. 570-592.
Brightwell, J.; Fouillet, X.; Cassano-Zoppi, A. L.; Gatz, R.; Duchosal, F. (1986) Neoplastic and functional
changes in rodents after chronic inhalation of engine exhaust emissions. Ishimshi. N.; Koizumi, A.;
McClellan, R. O.; Stober, W., eds. Caranogemc and Mutagenic Effects of Diesel Engine Exhaust.
Amsterdam: Elsevier, pp. 471-485.
Chen C; Far land, W. (1989) lncorporatmg cell proliferation in quantitative cancer risk assessment:
approaches, issues, and uncertainties. To appear in proceedings of the conference Chemically Induced
Cell Proliferation—Implications for Risk Assessment University of Texas. M. D. Anderson Cancer
Center.
Cuddihy, R. G.; McClellan, R. (1983) Evaluating lung cancer risks from exposure to diesel engine exhaust.
Risk AnaL 3(2): 119-124.
Cuddihy, R. G.; Griffith, W. C; Clark, C F. R.; McClellan. R. O. (1981) Potential health ana environmental
effects of light duty diesel vehicles. II(LMF-89). Springfield. VA: National Technical Information
Service, U.S. Department of Commerce.
Cuddihy, R. G.; Griffith, W. C; McQellan, R. O. (1984) Health risks from light-duty diesel vehicles.
Environ. Sci. TechnoL 18: 14A-21A.
Damber, L. A.; Larson. L. G. (1987) Occupation and male lung cancer: a case-control study m
northern Sweden. Br. J. Ind. Med. 44: 446-453.
Edling, O.; Anjou, C-G.; Axelson. O.; Kling, H. (1987) Mortality among personnel exposed to
diesel exhaust. Int. Arch Occup. Environ. Health 59:559-565.
Garshick, E.; Schenker, M. B.; Munoz. A. (1987) A case-control study of lung cancer and diesel exhaust
exposure in railroad workers. Am. Rev. Resp. Dis. 135: 1242-1248.
May, 1990
10-30
DRAFT - DO NOT QUOTE OR CITE
-------
Garshick, E.: Schenker, MB.; Munoz, A.; Segal, M.; ei al. (1988) A retrospeaive cohort study
of lung cancer and diesel exhaust exposure in railroad workers. Am. Rev. Respir. Dis. 124:820-825.
Grimmer, G.; Brune, R; Deutsch-Wenzel, R.; Dettbarn. G.; Jacob, J.; Naujack, K.-W.; Mohr,
U.; Ernst, R (1987) Contribution of polycyclic aromatic hydrocarbons and mtro-derrvatives to the
carcinogenic impact of diesel exhaust condensate evaluated by implantation into the lungs of rats.
Cancer Letters 37: 173-180.
Hail, N.E.; Wynder, E. (1984) Diesel exhaust exposure and lung cancer. Env. Research
34:77-86.
Hanis, J. (1981) Potential risk of lung cancer from diesel engine emissions. Washington, D.C. National
Academy Press.
Harris, J. (1983) Diesel emissions and lung cancer. Risk Anal. 3(2): 83-100.
Hayes, R. B.; Thomas T.; Silverman, D.T.; et al. (1989) Lung cancer in motor exhaust occupations Am. J. lnd.
Med. 16: 685-695.
Heinnch. U.: Pott. F.; Rittinghausen. S. (1986a) Comparison ot chronic inhalation effects in rodents after
long-term exposure to either coal oven flue gas mixed with pyrolized pitch or diesel engine exhaust.
In: Ishimshi, N.; Koizumi. A.; McClellan. R.; Stober, W.. eds. Carcinogenic and Mutagenic Effects
of Diesel Engine Exhaust. Amsterdam: Elsevier, pp. 441-457.
Heinnch, Muhle, H.; Takenaka, S.; Ernst, R; Fohst, R.; Mohr, U.; Pott, F.; Stober, W. (1986)
Chronic effects on the respiratory tract of hamsters, mice and rats after long-term inhalation
of high concentrations of filtered, and unfiltered diesel engine emissions. J. AppL ToxicoL
6(6): 383-395.
Henry, M. C; Kaufman, D. G. (1973) Clearance of benzo(a)pyrene from hamster lungs after administration
on coated parucles. J. Natl. Cancer Inst 51: 1961-1964.
Henry, M. C; Port, C; Kaufman, D. (1975) Importance of physical properties of benzo(a)pyrene-iemc oxide
mixture in lung tumor induction. Cancer Res. 35: 207-217.
Howe, G.R.; Fraser, D.; Lindsay, J.; et aL (1983) Cancer mortality (1965-77) in Relation to
diesel fume and coal exposure in a cohort of retired railway workers: 3NCI 70:1013-1019.
Ishinishi, N.; Kuwabara, N.; Nagase, S.; Suzuki, T.; Ishiwata, S.; Kohno, T. (1986) Effects of diesel soot on
unscheduled DNA synthesis of trachealepithelium and lung tumor formation. In: Ishinishi, N.;
Koizumi, A.; McClellan R.; Stober, W., eds. Carcinogenic and Mutagenic Effects of Diesel Engine
Exhaust. Amsterdam: Elsevier, pp. 213-221
Iwai, K.; Udagawa, T.; Yamagishi, M.; Yamada, H. (1986) Long-term inhalation studies of diesel exhaust on
F344 SFP rats. Incidence of lung cancer and lymphoma. Ishinishi, N.; Koizumi, A.; McClellan. R.O.;
StOber, W„ eds. Carcinogenic and Mutagenic Effects of Diesel Engine Exhaust. Amsterdam:
Elsevier, pp. 349-360.
Kotin, P.; Falk, R; Thomas, M. (1955) Aromatic hydrocarbons. IIL Presence in the particulate
phase of diesel engine exhausts and the carcinogenicity of exhaust extracts. Arch. Ind. Health 11:113-
120.
May, 1990
10-31
DRAFT - DO NOT QUOTE OR CITE
-------
Lee, K. P.; Trochimowicz. H. J.; Reinhardt, C F. (1985) Transmigration of titanium dioxide (HO,)
particles in rats after inhalation exposure. Exp. Molecular Pathol. 42: 331-343.
Lee, K. P.; Henry III, N. W.; Trochimowicz, H. J.; Reinhardt, C F. (1986) Pulmonary reponse to
impaired lung clearance in rats following excessive TiO: dust deposition. Environ. Res.
41: 144-167.
Lerchen. M. L.. Wiggins, C L.; Samet, J. M. (1987) Lung cancer and occupation in New
Mexico. J. Natl. Cancer Inst. 79: 639-645.
Mauderly. J. L.; Jones, R. K.; Griffith, W. C.; Henderson, R. F.; McQellan, R. O. (1987) Diesel exhaust is
a pulmonary carcinogen in rats exposed chronically by inhalation. Fundam. Appl. Tox. 9: 208-221.
McQellan, R. (1986) Health effects of diesel exhaust: a case study in risk assessment. Amer. Indust. Hygiene
Assoc J. 47: 1-13.
Moolgavkar, S. H.; Knudson, A G. (1981) Mutation and cancer: a model for human carcinogens. J. Natl.
Cancer Inst. 66: 1037-1052.
Morrow. P. E. (1988> Possible mechanisms to explain dust overloading of the lungs. Fund. Appl. Tox.
10: 369-384.
Nesnow, S„ Evans, C.; Stead, A; Creason, J.; Slaga, T. J.; Triplett, L. L_ (1982) Skin
carcinogenesis studies of emission extracts. In: Lewtas, J., ed. Toxicological Effects of Emissions from
Diesel Engines, Amsterdam, Elsevier, pp. 295-320.
Pepelko, W. E-; Peirano, W. B. (1983) Health effects of exposure to diesel engine emissions. J.
Amer. Coll. Toxicol. 2: 253-306.
Pott, F.; Stober, W. (1983) Carcinogenicity of airborne combustion products observed in
subcutaneous tissue and lungs of laboratory rodents. Environ. Health Perspec 47: 293-303.
Pott, F.; Heinnch. U. (1987) Dieselmotorabgas and Lungenk auf die Gefanhrdung des Menschen. In:
Umwelthygiene, Vol. 19. Med. Instutut. f. Umwelthygiene. Annual Report 1986/87. Dusseldorf.
F.R.G., pp. 130-167.
Rushton, L_; Alderson. M. R.; Nagarajah, C R. (1983) Epidemiological survey of maintenance
workers in London Transport executive bus garages at Chiswick works. Br. J. Ind. Med. 40: 340-345.
Saffiotti. U.; Borg, S. A.; Grote, M. I.; Karp, D. B. (1964) Retention rates of paniculate carcinogens in lungs.
Chicago Med. Sch. Quart 24: 1(M7.
Saffiotti. U.; Cefin, F.; Kolb. L. H.; Shubik, P. (1965) Expenmental studies of the conditions of exposure to
carcinogens for lung cancer obduction. J. Air Pollut. Control Assoc. 15: 23-25.
StOber. W. (1986) Expenmental induction of tumors in hamsters, mice and rats after long-term inhalation
of filtered and unfiltered diesel engine exhaust. Ishinishi. V.; Koizumi, A.; McClellan, R. O.; Stober.
W., eds. Carcinoeenic and Muta Genie Effects of Diesel Engine Exhaust. Amsterdam: Elsevier, pp.
421-429.
May, 1990
10-32
DRAFT - DO NOT QUOTE OR CITE
-------
Sua, J*. D.; Woff, R. K.; Kanapilly, G. M.; McClellan, R. O. (1984) Lung retenuon and metabolic fate of
inhaled benzo(a)pyrene associated with diesel exhaust particles. Toxicol. Appl. Pharmacol. 73: 48-59.
Vostal, J. J. (1986) Factors limiting the evidence for chemical carcinogenicity of diesei emissionsin long term
inhalation experiments. In: Ishinishi, N.; Koizumi, A.; McClellan, R.; Stober, W., eds. Carcinogenic
and Mutagenic Effects of Diesel Engine Exhaust. Amsterdam: Elseveier, pp. 381-396.
Waller, R. (1981) Trends in lung cancer in London in relation to exposure to diesel fumes. Environ. Int.
5: 479-483.
Williams, R. R.; Stegens, N. L.; Goldsmith, J. R. (1977) Associations of cancer site and type
with occupation and industry from the Third National Cancer Survey interview. J. Natl. Cancer Inst.
59: 1147-1185.
Wolff, R. K.; Bond, J. A.: Sun, J. D.; Sun, J.; Henderson, R.; Harkema, J.; Griffith, W.; Mauderly, J.;
McClellan, R. (1989) Effects of adsorption of benzo(a)pyrene onto carbon black panicles on levels
of DNA adducts of rats exposed by inhalation. Toxicol. Appl. Pharmacol. 97: 289-299.
Wolff, R. K.; Henderson, R. F.; Shipes, M. D.; Griffith, W. C.; Mauderly, J. L.; Cuddihy, R. G.: McClellan,
R. O. (1987) Alternations in panicle accumulation and clearance in lungs of rats chronically exposed
to diesel exhaust. Fundam. Appl. Toxicol. 9. 154-166.
Wong, O.; Morgan, R.W., Kheyts, L.; et al. (1985) Mortality among members of a heavy
construction equipment operators' union with potential exposure to diesel exhaust emissions. Br. J.
Ind. Med. 42:435-448.
May, 1990
10-33
DRAFT -- DO NOT QUOTE OR CITE
------- |