DRAFT EPA/60G/8-90/057D
DO NOT CITE OR QUOTE November 1999
SAB Review Draft
Health Assessment Document
for Diesel Emissions
NOTICE
TfflS DOCUMENT IS A PRELIMINARY DRAFT. It has not been formally
released by the Environmental Protection Agency 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.
National Center for Environmental Assessment-Washington Office
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
-------
DISCLAIMER
This document is a draft for review purposes only and does not constitute U.S. ^
Environmental Protection Agency policy. Mention of trade names or commercial products does
not constitute endorsement or recommendation for use.
11/5/99 -' ii DRAFT-DO NOT CITE
-------
CONTENTS
1. EXECUTIVE SUMMARY 1-1
2. DIESEL EMISSIONS CHARACTERIZATION, ATMOSPHERIC
TRANSFORMATION, AND EXPOSURES 2-1
2.1. INTRODUCTION 2-1
2.2. PRIMARY DIESEL EMISSIONS 2-3
2.2.1. Diesel Combustion and Formation of Primary Emissions 2-3
2.2.2. Diesel Emission Standards and Emission Trends Inventory 2-5
2.2.3. Engine Technology Description and Chronology 2-8
2.2.3.1. Injection Rate 2-9
2.2.3.2. Turbocharging, Charge-Air Cooling, and
Electronic Controls 2-10
2.2.3.3. Indirect and Direct Injection High-Speed Diesel Engines 2-13
2.2.3.4. Two-Stroke and 4-Stroke High-Speed Diesel Engines 2-14
2.2.3.5. Near-Term Diesel Emission Reduction Technologies 2-16
2.2.3.6. Future (2004+) Diesel Emission Reduction Technologies 2-18
2.2.4. History of Dieselization 2-24
2.2.4.1. Dieselization of the On-Road Fleet 2-24
2.2.4.2. Dieselization of Railroad Locomotive Engines 2-30
2.2.4.3. Historical Trends in Diesel Fuel Use and Impact of
Fuel Properties on Emissions 2-31
2.2.5. Chronological Assessment of Emission Factors 2-35
2.2.5.1. On-Road Vehicles 2-35
2.2.5.2. Locomotives 2-44
2.2.6. Physical and Chemical Composition of Particles 2-44
2.2.6.1. SOF and Elemental Carbon Content of Particles 2-47
2.2.6.2. PAHs and Nitro-PAH Emissions 2-51
2.2.6.3. Aldehyde Emissions 2-56
2.2.6.4. Dioxin and Furan Emissions 2-56
2.2.6.5. Particle Size 2-56
2.3. ATMOSPHERIC TRANSFORMATION OF DIESEL EXHAUST 2-60
2.3.1. Gas-Phase Diesel Exhaust 2-61
2.3.1.1. Organic Compounds 2-62
2.3.1.2. Inorganic Compounds 2-64
2.3.1.3. Atmospheric Transport of Gas-Phase Diesel Exhaust 2-65
2.3.2. Particle-Phase Diesel Exhaust 2-65
2.3.2.1. Particle-Associated PAH Photooxidation 2-67
2.3.2.2. Particle-Associated PAH Nitration 2-67
2.3.2.3. Particle-Associated PAH Ozonolysis 2-68
2.3.2 A. Atmospheric Transport of Diesel Exhaust Particle Matter 2-69
2.3.3. Diesel Exhaust Aging 2-69
2.4. AMBIENT DIESEL EXHAUST CONCENTRATIONS AND EXPOSURES .... 2-70
ST.
-------
CONTENTS (continued)
2.4.1. Diesel Exhaust Gases in the Ambient Atmosphere 2-70
2.4.2. Ambient Concentrations of Diesel PM 2-71
2.4.2.1. Receptor Modeling Estimates of Diesel PM 2-71
2.4.2.2. Elemental Carbon Surrogate for Diesel PM 2-74
2.4.2.3. Dispersion Modeling Results 2-77
2.4.3. Exposures to Diesel PM 2-78
2.4.3.1. Exposure Measurements 2-79
2.4.3.2. Modeling Exposures to Diesel PM 2-80
2.4.4. Ambient Diesel PM Summary 2-84
2.5. SUMMARY 2-86
2.6. REFERENCES 2-88
3. DOSIMETRY OF DIESEL EXHAUST PARTICLES IN THE
RESPIRATORY TRACT 3-1
3.1. INTRODUCTION 3-1
3.2. CHARACTERISTICS OF INHALED DPM AND RELATIONSHIP TO PM25.... 3-1
3.3. REGIONAL DEPOSITION OF INHALED DPM :. :,... 3-2
3.3.1. Deposition Mechanisms 3-3
3.3.1.1. Biological Factors Modifying Deposition 3-4
3.3.2. Particle Clearance and Translocation Mechanisms 3-7
3.3.2.1. ET Region 3-8
3.3.2.2. TB Region .V.:... 3-8
3.3.2.3. A Region .v...-.',-;" . 3-12
3.3.3. Translocations of Particles to Extra-alveolar Macrophage •*"
Compartment Sites .. 3-19
3.3.3.1. Clearance Kinetics ' 3-20
3.3.3.2. Interspecies Patterns of Clearance '. .-J.. *.-.- f.. 3-20
3.3.3.3. Biological Factors Modifying Clearance >-..' 3-21
3.3.3.4. Respiratory Tract Disease 3-21
3.4. PARTICLE OVERLOAD „...- ; 3-22
3.4.1. Introduction H^ I-..; *./!s*;3.. 3-22
3.4.2. Relevance lo Humans :ii.''.l\ . ./.. '.YJiTM'J.. 3-24
3.4.3. Potential Mechanisms for an AM Sequestration Compartment for •'
Particles During Particle Overload '...•.• ..-.'.'.... 3-26
3.5. MODELING THE DISPOSITION OF PARTICLES IN THE >, . .
RESPIRATORY TRACT , ,t................. :3-27
3.5.1. Introduction r... 3-27
3.5.2. Dosimetry Models for DPM , .;,;.... 3-27
3.5.2.1. Introduction ,.., ;.;..c.. 3-27
3.5.2.2. Deposition Models 3-28
3.5.2.3. Physiologically Based Models foi Clciuoiicc 3-29
3.5.2.4. Model Assumptions and Extrapolation to Humans 3-32
3.5.3. Deposition of Organics 3-34
11 /5/99 iv DRAFT-DO NOT CITE OR QUOTE.'
-------
CONTENTS (continued)
3.6. BIO AVAILABILITY OF ORGANIC CONSTITUENTS PRESENT ON
DIESEL EXHAUST PARTICLES 3-34
3.6.1. In Vivo Studies 3-35
3.6.1.1. Laboratory Investigations 3-35
3.6.1.2. Studies in Occupationally Exposed Humans 3-36
3.6.2. In Vitro Studies 3-36
3.6.2.1. Extraction of Diesel Particle-Associated Organics
By Biological Fluids 3-36
3.6.2.2. Extraction of Diesel Particle-Associated Organics by
Lung Cells and Cellular Components 3-37
3.6.3. Modeling Studies 3-38
3.7. SUMMARY 3-39
3.8. REFERENCES 3-41
4. MUTAGENICITY OF DIESEL EXHAUST 4-1
4.1. GENE MUTATIONS 4-1
4.2. CHROMOSOME EFFECTS 4-4
4.3. OTHER GENOTOXIC EFFECTS 4-6
4.4. SUMMARY 4-6
4.5. REFERENCES 4-7
5. NONCANCER HEALTH EFFECTS OF DIESEL EXHAUST 5-1
5.1. HEALTH EFFECTS OF WHOLE DIESEL EXHAUST 5-1
5.1.1. Human Studies 5-1
5.1.1.1. Short-Term Exposures 5-1
5.1.1.2. Long-Term Exposures 5-11
5.1.2. Laboratory Animal Studies 5-15
5.1.2.1. Acute Exposures 5-15
5.1.2.2. Short-Term and Subchronic Exposures 5-21
5.1.2.3. Chronic Exposures 5-26
5.2. COMPARISON OF HEALTH EFFECTS OF FILTERED AND
UNFILTERED DIESEL EXHAUST 5-78
5.3. INTERACTTVBEiTECTS OF DIESEL EXHAUST 5-82
5.4. COMPARATIVE RESPONSIVENESS AMONG SPECIES TO THE
PULMONARY EFFECTS OF DIESEL EXHAUST 5-84
5:5. DOSE-RATE AND PARTICIPATE CAUSATIVE ISSUES 5-85
5.6. SUMMARY AND DISCUSSION .; 5-89
5.6.1. Effects of Diesel Exhaust on Humans 5-89
5.6.2. Effects of Diesel Exhaust on Laboratory Animals 5-91
5.6.2.1. Effects on Survival and Growth 5-91
5.6.2.2. Effects on Pulmonary Function 5-92
5.6.2.3. Histopathological and Histochemical Effects 5-92
5.6.2.4. Effects on Airway Clearance .J 5-93
1 iy5/99 v DRAFT-DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
5.6.2.5. Neurological and Behavioral Effects ................... '. . . . 5-94
5.6.2.6. Effects on Immunity and Allergenicity ...................... 5-94
5.6.2.7. Other Noncancerous Effects .............................. 5-94
5.6.3. Comparison of Filtered and Unfiltered Diesel Exhaust ................. 5-94
5.6.4. Interactive Effects of Diesel Exhaust ............................... 5-95
5.6.5. Conclusions ............................................ ...... 5-95
5.7. REFERENCES [[[ 5-96
6. NONCANCER DOSE-RESPONSE EVALUATION: RfC DERIVATION ............ 6-1
6. 1 . INTRODUCTION-BACKGROUND OF THE INHALATION RfC
AND ORAL RfD [[[ 6-1
6.1.1. The Acceptable Daily Intake ...................................... 6-1
6. 1 .2. Oral RfD and Inhalation RfC— Dose-Response Assessments Inclusive of
Uncertainty Factors ............................................. 6-1
6. 1 .3. UFs— Designation and Application ................................. 6-2
6. 1 .4. Animal-to-Human Extrapolation Factor in the RfC— A Human Equivalent
Concentration ................................................. 6-3
6.1.5. Basic Procedures for Derivation of an RfC— Identification
of the Critical Effect, the Principal Study, Application of UF,
and Assignment of Confidence Level ............................... 6-4
6.2. ISSUES IN DERIVATION OF THE DIESEL RfC .......................... 6-5
6.2. 1 . Chronic Noncancer Effects in Humans— Relevancy of Rodent Data ....... 6-5
6.2.2. Pulmonary Pathology and Immunologic Effects as Critical Effects ........ 6-5
6.2.3. Application of UFs ......................................... ____ 6-5
6.2.4. Relationship of DPM to Ambient Levels of PM25 ...... , .............. 6-6
6.3. APPROACH FOR DERIVATION OF THE RfC FOR
DIESEL ENGINE EMISSIONS ........................................ 6-6
6.3.1. Consideration of Long-Term Inhalation Studies ....................... 6-6
6.3.2. Derivation of a HEC— Application of a Pharmacokinetic Model ......... 6-6
6.4. CHOICE OF THE CRITICAL EFFECT-RATIONALE AND
JUSTIFICATION ............. ..................... . ............... , . 6-8
6.4. 1 . Mode-of-Action and Candidate Effects .......... : . . ................. 6-8
6.5. PRINCIPAL STUDIES FOR INHALATION RfC DERIVATION ............. 6-10
6.6. SUPPORTING STUDIES FOR INHALATION RfC DERIVATION ........... 6-13
6.6= 1. Respiratory Tract Effects in Species Other Than the Rat ...... ......... 6-17
6.6.2. Application of the Benchmark Dose Approach to Derivation of the RfC ... 6-19
6.7. DERIVATION OF THE INHALATION RfC ............................. 6-20
6.7. 1 . The Effect Level— A NOAEL From a Chronic Inhalation Study ......... 6-20
6.7.2. Application of UFs— Animal-to-Human and Sensitive Subgroups ........ 6-21
6.7.3. Designation of Confidence Level ................................. 6-22
-------
CONTENTS (continued)
6.9. REFERENCES 6-23
7. CARCINOGENICITY OF DIESEL EXHAUST 7-1
7.1. INTRODUCTION 7-1
7.2 EPIDEMIOLOGIC STUDIES OF THE CARCINOGENICITY OF
EXPOSURE TO DIESEL EXHAUST 7-2
7.2.1. Cohort Studies 7-2
7.2.1.1. Waller (1981): Trends in Lung Cancer in London in Relation
to Exposure to Diesel Fumes 7-2
7.2.1.2. Howe et al. (1983): Cancer Mortality (1965 to 1977)
in Relation to Diesel Fumes and Coal Exposure in a Cohort of
Retired Railroad Workers 7-4
7.2.1.3. Rushton et al. (1983): Epidemiological Survey of
Maintenance Workers in the London Transport Executive
Bus Garages and Chiswick Works 7-5
7.2.1.4. Wong et al. (1985): Mortality Among Members of a
Heavy Construction Equipment Operators Union With Potential
Exposure to Diesel Exhaust Emissions 7-7
7.2.1.5. Edlingetal. (1987): Mortality Among Personnel
Exposed to Diesel Exhaust 7-10
7.2.1.6. Boffetta and Stellman (1988): Diesel Exhaust Exposure
and Mortality Among Males in the American Cancer Society
Prospective Study 7-11
7.2.1.7. Garshick et al. (1988): A Retrospective Cohort
Study of Lung Cancer and Diesel Exhaust Exposure in
Railroad Workers 7-13
7.2.1.8. Gustavsson et al. (1990): Lung Cancer and Exposure to
Diesel Exhaust Among Bus Garage Workers 7-16
7.2.1.9. Hansen(1993): A Followup Study on the Mortality
of Truck Drivers 7-18
7.2.2. Case-Control Studies of Lung Cancer 7-19
7.2.2.1. Williams et al. (1977): Associations of Cancer Site and
Type With Occupation and Industry From the Third National
Cancer Survey Interview 7-19
7.2.2.2. Hall and Wynder (1984): A Case-Control Study of Diesel
Exhaust Exposure and Lung Cancer -... 7-26
7.2.2.3. Damber and Larsson( 1987): Occupation and Male Lung
Cancer, a Case-Control Study in Northern Sweden 7-27
7.2.2.4. Lerchen et al. (1987): Lung Cancer and Occupation in
New Mexico 7-29
7.2.2.5. Garshick et al. (1987): A Case-Control Study of Lung Cancer
and Diesel Exhaust Exposure hi Railroad Workers 7-30
11/5/99 vii DRAFT-DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
7.2.2.6. Benhamou et al. (1988): Occupational Risk Factors of
Lung Cancer in a French Case-Control Study 7-33
7.2.2.7. Hayes et al. (1989): Lung Cancer in Motor
Exhaust-Related Occupations 7-35
7.2.2.8. Steenland et al. (1990): A Case-Control Study of
Lung Cancer and Truck Driving in the Teamsters Union 7-36
7.2.2.9. Steenland et al. (1998): Diesel Exhaust and Lung Cancer
in the Trucking Industry: Exposure-Response Analyses
and Risk Assessment 7-38
7.2.2.10. Boffetta et al. (1990): Case-Control Study on Occupational
Exposure to Diesel Exhaust and Lung Cancer Risk 7-40
7.2.2.11. Emmelin et al. (1993): Diesel Exhaust Exposure and
Smoking: A Case-Referent Study of Lung Cancer Among
Swedish Dock Workers 7-41
7.2.3. Case-Control Study of Prostate Cancer 7-43
7.2.3.1. Aronsen et al. (1996): Occupational Risk Factors
for Prostate Cancer: Results from a Case-Control
Study in Montreal, Quebec, Canada 7-43
7.2.4. Summaries of Studies and Meta-Analyses of Lung Cancer 7-48
7.2.4.1. Cohen and Higgins (1995): Health Effects of
Diesel Exhaust: Epidemiology 7-48
7.2.4.2. Bhatia et al. (1998): Diesel Exhaust Exposure and
Lung Cancer 7-49
7.2.4.3. Lipsett and Campleman( 1999): Occupational Exposure
to Diesel Exhaust and Lung Cancer: A Meta-Analysis 7-52
7.2.5. Case-Control Studies of Bladder Cancer 7-54
7.2.5.1. Howe et al. (1980): Tobacco Use, Occupation, Coffee,
Various Nutrients, and Bladder Cancer 7-54
7.2.5.2. Wynder et al. (1985): A Case-Control Study of Diesel
Exhaust Exposure and Bladder Cancer 7-56
7.2.5.3. Hoar and Hoover (1985): Truck Driving and Bladder
Cancer Mortality in Rural New England 7-58
*T ."» I- tf ri... , 1 - * ^. 1 ft f\>-*1\ A r-l ., . ,/-! ^ „ t •<•*, t C
/.^..j.t. oicciuitiiu CL Hi. (i^o/1. /\ oase-^oriutu iiuuy Ol
Bladder Cancer Using City Directories as a
Source of Occupational Data 7-59
7.2.5.5. Iscovich et al. (1987): Tobacco Smoking, Occupational
Exposures and Bladder Cancer in Argentina 7-61
7.2.5.6. Iyer et al. (1990): Diesel Exhaust Exposure and
Bladder Cancer Risk 7-63
7.2.5.7. Steineck et al. (1990): Increased Risk of Urothelial Cancer
in Stockholm From 1985 to 1987, After Exposure to
Benzene and Exhausts 7-65
11 /5/99 viii DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
7.2.6. Discussion and Summary 7-66
7.2.6.1. The Cohort Mortality Studies 7-72
7.2.6.2. Case-Control Studies of Lung Cancer 7-75
7.2.6.3. Reviews and Meta-analyses of Lung Cancer 7-77
7.2.6.4. Case-Control Studies of Bladder Cancer 7-78
7.2.6.5. Relevant Methodologic Issues 7-79
7.2.6.6. Criteria of Causal Inference 7-81
7.3. CARCINOGENICITY OF DIESEL EMISSIONS IN
LABORATORY ANIMALS 7-85
7.3.1. Inhalation Studies (Whole Diesel Exhaust) 7-86
7.3.1.1. Rat Studies 7-86
7.3.1.2. Mouse Studies : 7-102
7.3.1.3. Hamster Studies 7-105
7.3.1.4. Monkey Studies 7-106
7.3.2. Inhalation Studies (Filtered Diesel Exhaust) 7-106
7.3.3. Inhalation Studies (Diesel Exhaust Plus Co-Carcinogens) 7-107
7.3.4. Lung Implantation or Intratracheal Instillation Studies 7-108
7.3.4.1. Rat Studies 7-108
7.3.4.2. Syrian Hamster Studies 7-112
7.3.4.3. Mouse Studies 7-114
7.3.5. Subcutaneous and Intraperitoneal Injection Studies 7-114
7.3.5.1. Mouse Studies 7-114
7.3.6. Dermal Studies 7-116
7.3.6.1. Mouse Studies 7-116
7.3.7. Summary and Conclusions of Laboratory Animal Carcinogenicity Studies 7-121
7.4. MODE OF ACTION OF DIESEL EMISSION-INDUCED CARCINOGENESIS 7-126
7.4.1. Potential Role of Organic Exhaust Components in Lung Cancer
Induction 7-127
7.4.2. Role of Inflammatory Cytokines and Proteolytic Enzymes hi the
Induction of Lung Cancer by Diesel Exhaust 7-131
7.4.3. Role of Reactive Oxygen Species hi Lung Cancer Induction by
Diesel Exhaust 7-132
7.4.4. Relationship of Physical Characteristics of Particles to
Cancer Induction 7-135
7.4.5. Integrative Hypothesis For Diesel-induced Lung Cancer 7-136
7.4.6. Summary 7-138
7.5. CANCER WEIGHT-OF-EVIDENCE: HAZARD EVALUATION 7-139
7.5.1. Cancer Hazard Summary 7-139
7.5.2. Supporting Information 7-140
7.5.2.1. Human Data ' 7-140
7.5.2.2. Animal Data 7-141
7.5.2.3. Other Key Data 7-142
11/5/99 ix DRAFT-DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
7.5.2.4. Mode of Action ...................................... 7-142
7.6. DISCUSSION OF THE ROLE OF DIESEL EXHAUST IN THE OVERALL
PICTURE OF PM10 ................................................. 7-142
7.7. REFERENCES ............................ ........................ 7-143
8. CANCER DOSE-RESPONSE EVALUATION ................................. 8-1
8.1. INTRODUCTION [[[ 8-1
8.2. REVIEW OF PREVIOUS QUANTITATIVE RISK ESTIMATES .............. 8-1
8.2.1. Comparative Potency Method ..................................... 8-2
8.2.2. Suitability of Comparative Potency Approach ........................ 8-5
8.2.3. Animal Bioassay-Based Cancer Potency Estimates ..................... 8-6
8.2.4. Suitability of Laboratory Animal Bioassay Approach ................... 8-7
8.2.5. Epidemiology-Based Estimation of Cancer Potency ............ . ....... 8-8
8.2.6. Suitability of Using Epidemiologic Data ............................ 8-10
8.2.6.1. Railroad Worker Data .................................. 8-12
8.2.6.2. Teamster Truck Driver Data ............................. 8-12
8.3. OBSERVATIONS ABOUT RISK ...................................... 8-13
8.3.1. Perspectives .................................................. 8-13
8.4. SUMMARY OF CANCER DOSE-RESPONSE CONSIDERATIONS .......... 8-15
8.5. REFERENCES [[[ 8-16
9. CHARACTERIZATION OF HEALTH HAZARD AND DOSE-RESPONSE FOR
DIESEL ENGINE EXHAUST ......................... ..................... 9-1
9.1. INTRODUCTION [[[ 9-1
9.2. WHAT IS DIESEL EXHAUST IN A HEALTH HAZARD
ASSESSMENT CONTEXT? ........................................... 9-2
9.3. NONOCCUPATIONAL AND OCCUPATIONAL EXPOSURE ............. 9-5
9.4. HAZARD CHARACTERIZATION ..................................... 9-6
9.4.1. Health Effects Other Than Cancer: Acute Exposures ................... 9-6
9.4.2. Effects Other Than Cancer: Chronic Exposure ........................ 9-7
9.4.3. Health Effects Other Than Cancer: Derivation of Inhalation
Reference Concentration ......................................... 9-
1 7 TT *
9.5.1. Cancer Hazard ............................................... 9-11
9.6. CANCER DOSE-RESPONSE ASSESSMENT ............................ 9-14
-------
LIST OF TABLES
2-1. Emission standards: HD highway diesel engines 2-6
2-2. Emission standards: locomotives (g/bhp/hr) 2-7
2-3. Vehicle classification and weights for on-road trucks 2-25
2-4. Truck fleet results for 1992 from Census of Transportation (1995),
results in thousands 2-27
2-5. Emissions results from tunnel tests (adapted from Yanowitz et al., 1999b) 2-41
2-6. Remote sensing results for hd vehicles (Yanowitz et ai., 1999b) 2-42
2-7. Concentrations of nitro-polycyclic aromatic hydrocarbons identified in a LD diesel
paniculate extract 2-52
2-8. Comparison of PAH and nitro-PAH emissions for IDI naturally aspirated
engines and two DI turbocharged engines 2-54
2-9. Classes of compounds in diesel exhaust 2-61
2-10. Calculated atmospheric lifetimes for gas-phase reactions of selected compounds
present in automotive emissions with important reactive species 2-63
2-11. Major components of gas-phase diesel engine emissions and their known
atmospheric transformation products 2-64
2-12. Major components of particle-phase diesel engine emissions and their
known atmospheric transformation products 2-66
2-13. Ambient diesel PM concentrations reported from chemical mass
balance modeling 2-72
2-14. Diesel PM 2.5 concentrations in urban and rural locations using EC surrogate for
NESCAUM (1995) and IMPROVE (1992-1995) network sites 2-75
2-15. Modeled diesel PM2.1 for South Coast Air Basin in 1982 2-78
2-16. Diesel PM 1.0 exposures reported by Zaebst et al. (1991) and
calculated using the EC ratiometric approach 2-80
2-17. Annual average diesel PM exposures for 1990 in the general population and
among the highest exposed demographic groups hi
nine urban areas (on-road sources only) 2-82
2-18. Projected annual average diesel PM exposures from all on-road vehicles 2-83
2-19. Modeled and estimated concentrations of diesel PM hi microenvironments
(California EPA, 1998a) ; 2-84
2-20. Estimated indoor air and total air exposures to diesel PM in California hi 1990 2-85
3-1. Predicted doses of inhaled diesel exhaust particles per minute based on
total lung volume (M), total airway surface area (M,), or
surface area in alveolar region (M2) 3-7
3-2. Alveolar clearance in laboratory animals exposed to DPM 3-14
5-1. Human studies of exposure to diesel exhaust 5-16
5-2. Short-term effects of diesel exhaust on laboratory animals 5-22
5-3. Effects of chronic exposures to diesel exhaust on survival and growth of
laboratory animals 5-27
11/5/99 xi DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES (continued)
5-4. Effects of chronic exposures to diesel exhaust on organ weights and
organ-to-body-weight ratios 5-29
5-5. Effects of diesel exhaust on pulmonary function of laboratory animals 5-33
5-6. Histopathological effects of diesel exhaust in the lungs of laboratory animals 5-37
5-7. Effects of exposure to diesel exhaust on the pulmonary defense mechanisms of
laboratory animals 5-48
5-8. Effects of inhalation of diesel exhaust on the immune system of
laboratory animals 5-57
5-9. Effects of diesel particulate matter on the immune response of laboratory animals .. 5-61
5-10. Effects of exposure to diesel exhaust on the liver of laboratory animals 5-66
5-11. Effects of exposure to diesel exhaust on the hematological and cardiovascular
systems of laboratory animals 5-68
5-12. Effects of chronic exposures to diesel exhaust on serum chemistry of
laboratory animals 5-70
5-13. Effects of chronic exposures to diesel exhaust on microsomal enzymes of
laboratory animals 5-72
5-14. Effects of chronic exposures to diesel exhaust on behavior and neurophysiology ... 5-75
5-15. Effects of chronic exposures to diesel exhaust on reproduction and development in
laboratory animals 5-77
5-16. Composition of exposure atmospheres in studies comparing unfiltered and
filtered diesel exhaust 5-79
6-1. UFs and their default values used hi EPA's noncancer RfD and RfC methodology ... 6-3
6-2. Human equivalent continuous concentrations from the principal studies 6-14
6-3. Decision summary for the derivation of the RfC for diesel engine emissions 6-23
7-1. Epidemiologic studies of the health effects of exposure to diesel exhaust:
cohort mortality studies 7-20
7-2. Epidemiologic studies of the health effects of exposure to diesel exhaust:
case-control studies of lung cancer 7-44
7-3. Epidemiologic studies of the health effects of exposure to diesel exhaust:
case-control studies of bladder cancer 7-67
7-4. Summary of animal inhalation carcinogenicity studies 7-87
7-5. Tumor incidence and survival time of rats treated by surgical lung
implantation with fractions from diesel exhaust condensate (35 rats/group) 7-110
7-6. Tumor incidences in rats following intratracheal instillation of diesel exhaust
particles (DPM), extracted DPM, carbon black (CB), benzo[a]pyrene (BaP),
or particles plus BaP 7-113
7-7. Tumorigenic effects of dermal application of acetone extracts of diesei
particuiate matter (DPM) 7-117
7-8. Dermal tumorigenic and carcinogenic effects of various emission extracts 7-120
11/5/99 xii DRAFT-DO NOT CITE OR QUOTE
-------
LIST OF TABLES (continued)
7-9. Cumulative (concentration * time) exposure data for rats exposed to whole
diesel exhaust 7-122
8-1. Estimated 95% upper confidence limits of the lifetime risk of cancer
from inhalation of 1 jig/m3 diesel particulate matter (DPM) 8-3
11/5/99 xiii DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF FIGURES
2-1. A comparison of IDI (A) and DI (B) combustion systems of high-speed,
HD diesel truck engines. DI engines almost completely replaced IDI
engines for these applications by the early 1980s 2-4
2-2. Effect of turbocharging and aftercooling on NOX and PM (Mori, 1997) 2-11
2-3. An example of uniflow scavenging of a 2-stroke diesel engine with a positive
displacement blower (Adapted from Taylor, 1990) 2-14
2-4. NOx-storage catalyst operation under oxidizing and reducing conditions 2-19
2-5. A comparison of the NOX reduction efficiency over a range of temperature
conditions for the sulfur-intolerant NOX storage catalyst system and the more
sulfur-tolerant, active Pt-zeolite catalyst system 2-20
2-6. Schematic showing the operating principles of the continuously regenerating
trap (CRT) 2-22
2-7. Efficiency of NO to NO2 conversion over the oxidation catalyst component of the
CRT at different exhaust temperatures and at differing diesel fuel sulfur levels 2-23
2-8. Estimated sulfate (primarily H2SO4) PM emissions from a LD truck
equipped with a low-temperature Pt-zeolite lean-NOx catalyst system (Wall, 1998) . 2-24
2-9a. Number of HD diesel trucks sold in years 1957-1998 based on industry sales data .. 2-26
2-9b. Diesel truck sales (domestic) for the years 1939-1997 2-27
2-10a. Diesel truck sales as a percentage of total truck sales for the years 1957-1998 2-28
2-1 Ob. Diesel truck sales as a percentage of total truck sales for the years 1939-1997 2-29
2-11. Model year distribution of in-use truck fleet in 1992 2-29
2-12. Diesel fuel use since 1949 2-33
2-13. On-highway diesel fuel consumption since 1949, values in thousands of gallons .... 2-34
2-14. Model year trends in NOX emissions (g/mile) 2-37
2-15. Model year trends in PM emissions (g/mile) 2-38
2-16. Model year trends in HC emissions (g/mile) 2-39
2-17. Comparison of 2-stroke and 4-stroke engines PM emissions on a g/mi and
g/gal basis (low altitude data only) 2-43
2-18. Line-haul and switch emissions data 2-45
2-19. Comparison of SOF emissions for 2- and 4-stroke engines in g/mi and as a
-rxif/^art+on-a s>f4w+»1 'D\X ^ fit
LS^A W^AJLb4*&W VI" A fcV/ MIA JL XV A ................................................. ^—*TO
2-20. Trends in PM solids emissions with model year, a reasonable surrogate for
elemental carbon content 2-49
2-21. Parity plot showing approximate agreement between PM elemental carbon and
PM solids measurements in g/mi „.. , 2-50
2-22. 1-Nitropyrene emission rates from several HD diesel vehicles 2-55
2-23. Chassis dynamometer measurements of total aldehyde emissions from HD diesel
vehicles 2-57
2-24. Particle size distribution in diesel exhaust, taken from Kittelson (1998) 2-58
3-1. Modeled deposition distribution patterns of inhaled diesel exhaust
particles hi the airways of different species 3-6
11/5/99 xiv DRAFT-DO NOT CITE OR QUOTE
-------
LIST OF FIGURES (continued)
3-2. Modeled clearance of insoluble 4-um particles deposited in tracheobronchial and
.alveolar regions in humans 3-9
3-3. Short-term thoracic clearance of inhaled particles as determined by model
prediction and experimental measurement 3-11
3-4. Clearance from lungs of rats of 134Cs-FAP fused aluminosilicate tracer
particles inhaled after 24 months of diesel exhaust exposure at
concentrations of 0 (control), 0.35 (low), 3.5 (medium), and 7.0 (high) 3-22
3-5. Lung burdens of DPM within rats exposed to 0.35 (low), 3.5 (medium), and
7.0 (high) 3-23
7-1. Pooled relative risk estimates and heterogeneity-adjusted 95% confidence
intervals for all studies and subgroups of studies included in the meta-analysis 7-51
7-2. Pooled estimates of relative risk of lung cancer in epidemiological studies
involving occupational exposure to diesel exhaust (random-effects models) 7-53
7-3. Lung cancer and exposure to diesel exhaust in railroad workers 7-82
7-4. Lung cancer and exposure to diesel exhaust in truck drivers 7-83
7-5. Pathogenesis of lung disease in rats with chronic, high-level
exposure to particles 7-137
11/5/99 xv DRAFT-DO NOT CITE OR QUOTE
-------
PREFACE
This draft health risk assessment document was prepared by the National Center for
Environmental Assessment (NCEA), which is the health risk assessment program in EPA's
Office of Research and Development. The assessment has been prepared for EPA's Office of
Mobile Sources which requested advice regarding the potential health hazards associated with
diesel engine use. As diesel exhaust emissions also affect air toxics and ambient particulate
matter, other EPA air programs also have an interest hi this assessment. The previous draft of
this assessment was released for public comment in February 1998, and the Agency's Clean Air
Scientific Advisory Committee (CASAC) met in public session in May 1998 to review the draft.
This November 1999 draft is a revision of that 1998 draft, but also builds on the 1990-1999
history of the development of this diesel health risk assessment.
The scientific literature search for this assessment is generally current through January
1999, though a few more recent publications on key topics also have been included.
This November 1999 draft assessment will be reviewed by CASAC in December 1999,
and concurrently, public comments will be accepted for a limited time. Following the receipt of
comments from CASAC and the public, NCEA plans to finalize the assessment.
1175/99 xvi DRAFT-DO NOT CITE OR QUOTE
-------
AUTHORS AND CONTRIBUTORS
The National Center for Environmental Assessment (NCEA), within EPA's Office of
Research and Development (ORD), was responsible for the preparation of this document.
Authors and chapter managers for this draft health assessment document are listed below.
CHAPTER 1. EXECUTIVE SUMMARY
Authors
NCEA Diesel Team
CHAPTER 2. DIESEL EMISSIONS CHARACTERIZATION, ATMOSPHERIC
TRANSFORMATION, AND EXPOSURES
Chapter Manager/Author
Marion Hoyer, Office of Mobile Sources, U.S. Environmental Protection Agency, Ann Arbor,
MI.
Contributors
Chad Bailey, Office of Mobile Sources, U.S. Environmental Protection Agency, Ann Arbor, MI.
Tom Baines, Office of Mobile Sources, U.S. Environmental Protection Agency, Ann Arbor, MI.
David Cleverly, National Center for Environmental Assessment, U.S. Environmental Protection
Agency, Washington, DC.
William Ewald, National Center for Environmental Assessment, U.S. Environmental Protection
Agency, Research Triangle Park, NC.
Robert McCormick, Colorado School of Mines, Golden, CO
Joseph McDonald, Office of Mobile Sources, U.S. Environmental Protection Agency, Ann
Arbor, ML
11/5/99 xvii DRAFT-DO NOT CITE OR QUOTE
-------
Joseph Somers, Office of Mobile Sources, U.S. Environmental Protection Agency, Ann Arbor,
MI.
Janet Yanowitz, Colorado School of Mines, Golden, CO.
Barbara Zielinska, Desert Research Institute, Reno NV.
CHAPTER 3. DOSIMETRY OF DIESEL EXHAUST PARTICLES IN THE
RESPIRATORY TRACT
Authors
James McGrath, National Center for Environmental Assessment, U.S. Environmental Protection
Agency, Research Triangle Park, NC.
William Pepelko, National Center for Environmental Assessment, U.S. Environmental Protection
Agency, Washington, DC.
Contributor
Gary Foureman, National Center for Environmental Assessment, U.S. Environmental Protection
Agency, Research Triangle Park, NC.
CHAPTER 4. MUTAGENICITY OF DIESEL EXHAUST
Author
Lawrence Valcovic, National Center for Environmental Assessment, U.S. Environmental
Protection Agency, Washington, DC.
CHAPTER 5. NONCANCER HEALTH EFFECTS OF DIESEL EXHAUST
Authors
James McGrath, National Center for Environmental Assessment, U.S. Environmental Protection
11/5/99 xviii DRAFT-DO NOT CITE OR QUOTE
-------
Agency, Research Triangle Park, NC.
Contributor
Gary Foureman, National Center for Environmental Assessment, U.S. Environmental Protection
Agency, Research Triangle Park, NC.
CHAPTER 6. NONCANCER DOSE-RESPONSE
EVALUATION: RfC DERIVATION
Authors
Gary Foureman, National Center for Environmental Assessment, U.S. Environmental Protection
Agency, Research Triangle Park, NC.
Contributor
James McGrath, National Center for Environmental Assessment, U.S. Environmental Protection
Agency, Research Triangle Park, NC.
CHAPTER 7. CARCINOGENICITY OF DIESEL EXHAUST
Authors
Aparna Koppikar, National Center for Environmental Assessment, U.S. Environmental
Protection Agency, Washington, DC.
William Pepelko, National Center for Environmental Assessment, U.S. Environmental Protection
Agency, Washington, DC.
Contributors
Drew Levy, University of Washington, Seattle, WA
Robert Young, Oak Ridge National Laboratory, Oak Ridge, TN
11/5/99 xix DRAFT-DO NOT CITE OR QUOTE
-------
CHAPTER 8. CANCER DOSE-RESPONSE EVALUATION
Authors
Chao Chen, National Center for Environmental Assessment, U.S. Environmental Protection
Agency, Washington, DC.
William Pepelko, National Center for Environmental Assessment, U.S. Environmental Protection
Agency, Washington, DC.
Contributor
Charles Ris, National Center for Environmental Assessment, U.S. Environmental Protection
Agency, Washington, DC.
CHAPTER 9. CHARACTERIZATION OF HEALTH HAZARD AND
DOSE-RESPONSE FOR DIESEL ENGINE EXHAUST
Author
Charles Ris, National Center for Environmental Assessment, U.S. Environmental Protection
Agency, Washington, DC.
Contributors
NCEA Diesel Team
This document was preceded by three earlier drafts: a Workshop Review Draft
(EPA/600/8-90/057A, July 1990), an External Review Draft (EPA/600/8-90/057B, December
1994), and an SAB Review Draft (EPA/600/8-90/057C, February 1998). The Science Advisory
Board's Clean Air Scientific Advisory Committee (CASAC) reviewed the 1994 draft in public
sessions in May 1995 and the 1998 draft in May 1998. Public comment periods also were
conducted concurrently with the CASAC reviews. In addition, many reviewers both within and
outside the Agency provided assistance at various review stages.
11/5/99 xx DRAFT-DO NOT CITE OR QUOTE
-------
ACKNOWLEDGMENTS
Document Production
Terri Konoza
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
Kay Marshall
Clara Laucho
Eric Sorensen
The CDM Group, Inc.
Chevy Chase, MD
Printing and Distribution
Linda Bailey-Becht
Judy Theisen
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
11/5/99 xxi DRAFT-DO NOT CITE OR QUOTE
-------
1. EXECUTIVE SUMMARY
1 The Health Assessment Document for Diesel Emissions represents the Agency's first
2 comprehensive review of health effects from exposure to exhaust from diesel engines. In-depth
3 research on diesel exhaust (DE) started in the 1970s, and EPA began regulating emission levels
4 for certain types of diesel engines during the same period. EPA wanted to be aware of the
5 current health issues as it continues with Clean Air Act regulatory programs, hence the need for
6 this assessment. In nine chapters, this health assessment addresses key themes or questions such
7 as (1) the health effects of concern for humans, (2) the best insight as to the mode of action and
8 measure of dose/exposure for the toxic response(s), (3) what dose-response analysis suggests
9 about the possible impact/risk to a human population, and (4) the overall nature of the hazard and
10 the related confidence or uncertainties.
11 Diesel exhaust is a complex mixture of particles and gases with hundreds of chemical
12 compounds, including many organic compounds, present on the particles and in the gases. The
13 particles have an elemental carbon core, with individual particles being very small (a mean
14 aerodynamic diameter of about 0.2 um) and thus highly respirable. The small particles have a
1 5 large surface area upon which many organic compounds are adsorbed. The particle organics
16 generally contribute 10%-30% of particle weight and, for example, contain various types of
17 polyaromatic hydrocarbons (PAHs). The gases have both inorganic and organic constituents
18 (e.g., sulfur dioxide, nitrogen oxides, benzene, ethylene, toluene, aldehydes, olefins, and low-
19 molecular-mass PAHs). Both the particles and the numerous organic compounds of DE have
20 toxicological properties that are capable of influencing a toxic response in humans, though the
21 role of either or both in producing a toxic effect in humans is unknown.
22 DE particles contribute to ambient paniculate matter, e.g., PM2 5. Compared to other
23 sources of ambient PM, the elemental carbon core is nearly unique to DE, as are a few of the
24 adsorbed organic compounds. The DE gases are more ubiquitous in an urban environment.
25 Diesel engines may be on-road (vehicle engines) or off-road (many types of engines
26 powering equipment, machinery, railroad locomotives, and ships). Quantitatively, amounts of
27 specific emission constituents vary by type of engine and even within the same engine type.
28 Qualitatively, the basic composition is fairly consistent, for example, an elemental carbon core
29 particle with PAHs adsorbed to the particle and also present in the gases. Over the years, the
30 mass of particles emitted in engine exhaust has been reduced, as have the accompanying
31 organics.
i i /5/99 1 -1 DRAFT—DO NOT CITE OR QUOTE
-------
1 For years researchers have measured DE concentrations using particle mass per unit
0 volume, i.e., ug/m3 of diesel paniculate matter. This assessment adopts ug/m3 as a dosimeter and
^ further assumes that the important toxicologic agents in DE will be proportional to ug/m3. This
4 leads to some uncertainty, but the best dosimeter will not be known until the mode of action for
5 DE toxicity is better understood. Questions have been raised as to whether toxicological findings
6 generated from exposure to older engine exhaust can appropriately be applied to current-day
7 engine exhaust exposures. This question is not resolvable with present information, except to
8 note that available evidence does not point to significant shifts in DE composition relative to the
9 total organics over the years, and that organics are believed to be in relative proportion to the
10 mass of particles.
11 The primary chronic health concerns include nonmalignant respiratory effects and lung
12 carcinogenicity. The DE particulates can be a component of ambient PM2S. Compared to
13 ambient PM2 s with no DE component, DE is likely to have a higher proportion of fine and
14 ultrafine particulates and is likely to have a higher or at least a varied content of toxicologically
15 active organic compounds. Although some similarities exist between DE and ambient PM, the
16 differences are potentially significant. A comparison of the DE RfC and the PM2.5 standard has
17 considerable complexity. For ambient PM we see increased mortality and morbidity in human
^ studies from various forms of chronic respiratory disease. For DE we expect adverse respiratory
"W effects but have not clearly observed them in human studies, possibly because few such studies
20 have focused on respiratory effects. Animal studies conducted at higher than ambient exposure
21 levels, the most prominent being in the rat, provide the basis for the expectation of human
22 respiratory disease. A recommended human chronic exposure level without appreciable hazard
23 (i.e., inhalation Reference Concentration, RfC, 5 ug/m3) from adverse noncancer respiratory
24 effects is provided in the assessment. From an acute exposure standpoint, DE is an irritant to the
25 respiratory system given sufficient episodic exposure and may cause a variety of inflammation-
26 related symptoms (e.g., headache, eye discomfort, asthma-like reactions, nausea, etc.) depending
27 on individual susceptibility to the DE constituents. Data also suggest that DE is a factor in
28 exacerbating or initiating allergenic hypersensitivity; this is an emerging area of concern.
29 The carcinogenicity of DE also has been of research and public health interest. Diesel
30 engine exhaust is "highly likely" to be carcinogenic by the inhalation route of exposure,
31 according to EPA's 1996 Proposed Guidelines for Carcinogen Risk Assessment. This hazard is
32 viewed as being applicable to ambient (i.e., environmental) exposures. Many of the organics
33 present on the DE particles and in the gases, though in small quantities, are mutagenic and/or
1175/99 1 -2 DRAFT—DO NOT CITE OR QUOTE
-------
1 carcinogenic in their own right. DE shows a pattern of statistically increased lung cancer in more
2 than 20, but not all, human occupational studies where DE exposure is prominent. Lung cancer
3 increases are, on average, about 33-47% above background levels, though specific studies
4 suggest some modestly higher increases. There are some uncertainties about the magnitude of
5 the increase, because questions about exposure are almost always present in the human studies in
6 which the increases are seen, and with lung cancer, the question of confounding by cigarette
7 smoke is present. Nevertheless, analysis of the occupational studies shows that the pattern of
8 increased lung cancer remains after consideration of these issues. Bladder cancer also has been
9 elevated in some epidemiologic studies, though the totality of the evidence is too weak to form a
10 clear conclusion. Although rat inhalation cancer bioassays were once thought to be useful for
11 inferring a human cancer hazard or supporting human evidence, in recent years, the rat lung
1 2 cancer responses seen with DE exposure are thought to be less clear for human hazard prediction
13 and unsuitable for environmental exposure risk estimation. None of the available studies show
14 that the lung cancer hazard is present at environmental levels of exposure, although the margin
1 5 may be relatively small between some higher environmental exposures and occupational
1 6 exposures where lung cancer risks are thought to be present.
17 The plausibility of an environmental lung cancer hazard from DE by inhalation exposure
18 is supported by findings contained hi this assessment. Overall, the evidence for a likely human
19 lung cancer hazard by inhalation is persuasive, even though, hi the absence of complete data,
20 inferences and thus uncertainties are involved. Some of the key uncertainties include: (1)
21 methodologic limitations inherent in epidemiologic studies, as well as a lack of reliable historical
22 exposure data for occupationally exposed cohorts, (2) uncertainties regarding the extent of
23 bioavailability of organic compounds present on diesel particles and their impact on the
24 carcinogenic process, and (3) other uncertainties regarding the mode of action of DE on lung
2 5 cancer in humans.
26 A decision has been made in this assessment that, despite the finding that DE is best
27 characterized as highly likely to be a lung cancer hazard, the available data are currently
28 unsuitable to make a confident quantitative statement about the magnitude of the lung cancer nsk
29 attributable to DE at ambient exposure levels. Therefore, this assessment does not adopt or
30 recommend a specific cancer unit risk estimate for DE. However, information is provided to put
31 DE cancer hazard hi perspective and to assist decisionmakers and the public to make prudent
32 public health judgments in the absence of a definitive estimate of the upper bound on cancer risk.
33 Efforts to derive cancer risk estimates for environmental purposes continue, with the focus'being
11/5/99 1 -3 DRAFT—DO NOT CITE OR QUOTE
-------
1 on epidemiologic studies because the epidemiology-based estimates are always the ideal starting
2 point, while also recognizing that the rat inhalation studies are no longer favored and other
approaches identified to date have limitations.
4 There is no DE-specific information that provides direct insight to the question of
5 variable susceptibility within the population. Default approaches to account for uncertainty in
6 inter-individual susceptibility have been included in the derivation of the RfC. Individuals with
7 preexisting lung burdens of particulates may have less of a margin of safety from DE particulate-
8 driven hazards than might be inferred from incremental DE exposure analysis, although this
9 cannot be quantified. DE exposure could be additive to many other daily or lifetime exposures to
10 organics and PM. For example, adults who predispose their lungs to increased particle retention
11 (e.g., smoking or high participate burdens from nondiesel sources), have existing respiratory or
12 lung inflammation or repeated respiratory infections, or have chronic bronchitis, asthma, or
13 fibrosis could be more susceptible to adverse impacts from DE exposure. Although there is no
14 information from studies of DE, infants and children could have a greater susceptibility to the
15 acute/chronic toxicity of DE because they have greater ventilatory frequency, resulting in greater
16 respiratory tract particle deposition. The issue of DE impacts on allergenicity and potential onset
17 and exacerbation of childhood asthma is being actively investigated, but firm conclusions await
18 peer review and publication of ongoing work.
ll Another aspect of differential susceptibility involves subgroups that may receive
20 additional exposure to DE because of their proximity to DE sources. Those having outside time
21 in their daily routine and being near a diesel emission source would likely receive more exposure
22 than others in the population. The highest exposed are most likely the occupational subgroups
23 whose job brings them very close to diesel emission sources (e.g., trucking industry, machinery
24 operations, engine mechanics, some types of transit operations, railroads, etc.).
25 Ongoing analyses by EPA, other Federal agencies, and worldwide researchers are
26 expected to improve the existing epidemiology and related exposure databases. These will
27 provide new opportunities to evaluate the potential health effects of DE on the general population
28 and susceptible subgroups.
1175/99 1-4 DRAFT—DO NOT CITE OR QUOTE
-------
2. DIESEL EMISSIONS CHARACTERIZATION, ATMOSPHERIC
TRANSFORMATION, AND EXPOSURES
1 2.1. INTRODUCTION
2 The intent of this chapter is to provide background information relating to the diesel
3 engine, the pollutants it emits, the history of its use in highway vehicles and railroad
4 locomotives, diesel exhaust composition and emissions trends, and air pollution regulatory
5 standards for diesel engines in the United States. The chapter also provides specific information
6 about physical and chemical composition of diesel exhaust, descriptions of its atmospheric
7 transformations, observations of measured and modeled ambient concentrations (considered
8 alone and as a component of atmospheric particles in general), and some preliminary estimates of
9 population exposures. This information provides background information that is used in
10 conjunction with the toxicological and epidemiology data to formulate the conclusions about
11 human health hazards that are discussed in later chapters of this document. The exposure
12 information does not represent a formal or rigorous exposure assessment; it is only intended to
13 provide a context for the health effects data and health hazard findings.
14 The diesel engine was patented in 1892 by Rudolf Diesel, who conceived it as a prime
15 mover that would provide much improved fuel efficiency compared with spark-ignition engines.
16 To the present day, the diesel engine's excellent fuel economy remains one of its strongest selling
17 points. In the United States, the diesel engine is used mainly in trucks, buses, agricultural and
18 other off-road equipment, locomotives, and ships.
19 The chief advantages of the diesel engine over the gasoline engine are its fuel economy
20 and durability. Diesel engines, however, emit a higher mass of carbonaceous particulate matter
21 than do gasoline engines. Over the past decade, modifications of diesel engine components have
22 substantially reduced particle emissions (Hammerle et al., 1994; Sawyer and Johnson, 1995).
23 The diesel engine compresses air to high pressure and temperature. Fuel, when injected
24 imo this compressed air, autoignii.es, releasing its chemical energy. The resulting combustion
2G gases expand, doing work en the piston, before being exhausted tc the atmosphere. Power output
26 is controlled by the amount of injected fuel rather than by throttling the air intake. Compared to
27 its spark-ignited (SI) counterpart, the diesel engine's superior efficiency derives from a higher
28 compression ratio and no part-load throttling. To ensure structural integrity for prolonged
29 reliable operation at the higher peak pressures brought about by a higher compression ratio and
30 autoignition, the structure of a diesel engine generally is more massive than its SI counterpart.
31 Diesel engines (also called compression-ignition, Ui) may be broadly identified as being
32 either two- or four-stroke cycle, injected directly or indirectly, and naturally aspirated or
11/5/99 2-1 DRAFT-DO NOT CITE OR QUOTE
-------
1 supercharged. They also are classified according to service requirements such as light-duty (LD)
2 or heavy-duty (HD) automotive/truck, small or large industrial, and rail or marine.
L All diesel engines use hydraulic fuel injection in one form or another. The fuel system
F must meet four main objectives if a diesel engine is to function properly over its entire operating
5 range. It must: (1) meter the correct quantity of fuel, (2) distribute the metered fuel to the correct
6 cylinder, (3) inject the metered fuel at the correct time, and (4) inject the fuel so that it is
7 atomized and mixes well with the in-cylinder air. The first two objectives are functions of a
8 well-designed injection pump, and the last two are mostly functions of the injection nozzle. As a
9 part of the effort to obtain lower exhaust emissions without diminishing fuel efficiency, fuel
10 injection systems are moving toward the use of electronic components for more flexible control
11 than is available with purely mechanical systems.
12 Both the fuel and the lubricants that are used to service diesel engines are highly finished
13 petroleum-based products combined with chemical additives. Diesel fuel is a mixture of many
14 different hydrocarbon molecules from about C7, to about C35, with a boiling range from roughly
15 350 to 650 °F. Many of the fuel and oil properties, such as its specific energy content (which is
16 higher than gasoline), ignition quality, and specific gravity, are related to its hydrocarbon
. 17 composition. Therefore, fuel and lubricant composition affects many aspects of engine
18 performance, including fuel economy and exhaust emissions.
^^ Complete and incomplete combustion of fuel in the diesel engine results in the formation
^sv of a complex mixture of gaseous and particulate exhaust. Because of concerns over health
21 effects associated with diesel particulate emissions, EPA began regulating emissions from diesel
22 engines in 1970 (for smoke) and then added regulations for gaseous emissions. EPA first
23 regulated particulate emissions from HD diesels in 1988.
24 This chapter begins with background information regarding the formation of primary
25 emissions resulting from diesel combustion, a summary of EPA emission standards for on-road
26 and locomotive diesel engines, and a description of the national trends in emissions from on- and
27 off-road diesel sources. The chapter continues with a description of engine technologies and the
28 history of dieselization for on-road vehicles and locomotives, then provides a chronological
29 assessment of emission rates and the chemical and physical nature of emissions. The data
30 describing diesel engine emissions consider primary emissions, which undergo chemical and
31 physical transformations in the atmosphere. Since the atmospheric transformations potentially
32 have important impacts on environmental and human health, the available information regarding
33 these transformations is discussed. This chapter concludes with a summary of the available
34 literature regarding concentrations and exposures to diesel particulate matter (PM) in different
35 exposure settings.
11/5/99 2-2 -DRAFT-DO NOT CITE OR QUOTE
-------
1 2.2. PRIMARY DIESEL EMISSIONS
2 2.2.1. Diesel Combustion and Formation of Primary Emissions
3 A basic understanding of diesel combustion processes can assist in understanding the
4 complex factors that influence the formation of PM and other diesel exhaust emissions. Unlike
5 spark-ignition combustion, diesel combustion is a fairly nonhomogenous process. Fuel is
6 sprayed at high pressure into the compressed cylinder contents (primarily air with some residual
7 combustion products) as the piston nears the top of the compression stroke. The turbulent
8 mixing of fuel and air that takes place is enhanced by injection pressure, the orientation of the
9 intake ports (e.g., inducement of intake-swirl tangential to the cylinder wall), piston motion, and
10 piston bowl shape. In some cases fuel and air mixing is induced via injection of the fuel into a
11 turbulence-generating pre-chamber or swirl chamber located adjacent to the main chamber
12 (primarily in older, higher speed engines and some LD diesels). Examples of typical direct
13 injection (DI) and indirect injection (IDI) combustion systems are compared in Figure 2-1.
14 Diesel combustion can be considered to consist of the following phases (Heywood, 1988;
15 Watson and Janota, 1982):
16
17 1. An ignition delay period, which starts after the initial injection of fuel and continues
18 until the initiation of combustion. The delay period is governed by the rate of fuel
19 and air mixing, diffusion, turbulence, heat transfer, chemical kinetics, and fuel
20 vaporization. Fuel cetane rating is an indication of ignition delay.
21 2. Rapid, premixed burning of the fuel and air mixture from the ignition delay period.
22 3. Diffusion-controlled burning, in which the fuel burns as it is injected and diffuses
23 into the cylinder.
24 4. A very small amount of rate-controlled burning during the expansion stroke, after
25 the end of injection.
26
27 Engine speed and load are controlled by the quantity of fuel injected. Thus, the overall
28 fuel-io-aii iatio varies as engine speed and lead vary. On a macro scale the cylinder contents are
29 always fuel-lean. Depending on the time available for combustion and the proximity of oxygen,
30 the fuel droplets are either completely or partially oxidized. At temperatures above 1300 K,
31 unburned fuel that is not oxidized is pyrolized (stripped of hydrogen) to form elemental carbon
32 soot (Dec and Espey, 1995). Soot formation occurs primarily during the diffusion-bum phase of
33 combustion, and is highest during high load and other conditions consistent with high fuel-air
1175/99 2-3 DRAFT-DO NOT CITE OR QUOTE
-------
Single Hole
Injection Nozzle
Precombustion
Chamber
Four Hole
Injection Nozzle
Figure 2-1. A comparison of IDI (A) and DI (B) combustion systems of high-speed,
HD diesel truck engines. DI engines almost completely replaced IDI engines for
these applications by the early 1980s.
1 equivalence ratios. Most of the soot formed (80% to 98%) is oxidized during later stages of
2 combustion, most likely by hydroxyl (OH) radicals formed during combustion (Kittelson et al.,
3 1986; Foster and Tree, 1994). The remainder of the soot leaves as a component of PM emissions
4 from the engine.
5 During combustion, sulfur compounds present in the fuel are oxidized to sulfur dioxide
6 (SQ2). Approximately 1% to 4% of fuel sulfur is oxidized to SO3, which combines with water
7 vapor in the exhaust to form sulfuric acid (H2SO4) (Wall et al., 1987; Khatri et al., 1978;
8 Baranescu, 1988). Upon cooling, sulfuric acid and water condense into an aerosol that is
9 nonvolatile under ambient conditions. The mass of sulfuric acid PM is more than doubled by the
11/5/99
2-4
DRAFT-DO NOT CITE OR QUOTE
-------
1 mass of water associated with the sulfuric under typical PM measurement conditions (50%
2 relative humidity, 20-25 °C) (Wall et al., 1987).
3 Oxide of nitrogen (NOX) emissions from combustion engines, primarily (at least initially)
4 in the form of NO, are generally thought to be formed via the Zeldovich mechanism, which is
5 highly temperature dependent. High combustion temperatures cause reactions between oxygen
6 and nitrogen to form NO and some NO2. The majority of NO2 formed during combustion is
7 rapidly decomposed. NO can also decompose to N2 and O2 but the rate of decomposition is very
8 slow because of the rapidly decreasing temperatures from the expansion of combustion gases
9 during the expansion stroke (Heywood, 1988; Watson and Janota, 1982). Thus, most of the NOX
10 emitted is NO.
11 Some organic compounds from unbumed fuel and from lubricating oil consumed by the
12 engine can be trapped in crevices or cool spots within the cylinder and thus are not sufficiently
13 available to conditions that would lead to their oxidation or pyrolysis. These compounds are
14 emitted from the engine and either contribute to gas-phase organic emissions or to PM emissions,
15 depending on their volatility. Within the exhaust system, temperatures are sufficiently high that
16 these compounds are entirely present within the gas phase (Johnson and Kittelson, 1996). Upon
17 cooling and mixing with ambient air in the exhaust plume, some of the less volatile organic
18 compounds can adsorb to the surfaces of soot particles. Lacking sufficient soot adsorption sites,
19 the organic compounds may condense on sulfuric acid nuclei (Abdul-Khalek et al., 1999).
20 Metallic compounds from engine component wear, and from compounds in the fuel and
21 lubricant, contribute to PM mass. Ash from oil combustion also contributes trace amounts to PM
22 mass.
23
24 2.2.2. Diesel Emission Standards and Emission Trends Inventory
25 EPA set a smoke standard for on-road HD diesel engines beginning with the 1970 model
26 year, and then added a CO standard and a combined hydrocarbon (HC) and NOX standard for the
27 1974 model year, as detailed in Table 2-1. Beginning in the 1979 model year, the EPA added a
28 HC standard while retaining the combined HC and NOX standard. All of the testing for HC, CO,
29 and NOX was completed using a steady state test procedure. Beginning in the 1985 model year,
30 the EPA added a NOX standard, dropped the combined HC and NO,; standard, and converted from
31 steady-state to transient testing for HC, CO, and NOX emissions. EPA introduced a particulate
32 standard for the 1988 model year.
33 Since the 1985 model year, only the NOX and particuiate standards have been tightened
34 for diesel engines. For truck and bus engines, the particuiate standard was reduced in 1991, and
35 again in 1994 for truck engines. For urban bus engines, the particuiate standard was reduced in
11/5/99 2-5 DRAFT-DO NOT CITE OR QUOTE
-------
Table 2-1. Emission standards: HD highway diesel engines
Model
year
1970
1974
1979
1985C
1988
1990
1991
1993
1994
1996
1998
2004
Pollutant (g/bhp-hr)
HC
1.5
1.3
1.3
1.3
1.3
1.3
1.3
1.3
1.3
1.3
CO
40
25
15.5
15.5
15.5
15.5
15.5
15.5
15.5
15.5
15.5
NOX
10.7
10.7
6.0
5.0
5.0
5.0
5.0
4.0
—
HC + NOX
16b
10b
—
—
—
—
—
—
—
—
2.4 NMHCd
Participate (PM)
t=truck, b=bus,
ub=urban bus
—
0.60
0.60
0.25
0.25 t, 0.10 b
O.lOt, 0.07 ub
0. lOt, 0.05 ub
O.lOt, 0.05 ub
O.lOt, 0.05 ub
Smoke'
A:40%; L:20%
A:20%; L:15%; P:50%
A:20%; L:15%; P:50%
A:20%; L:15%; P:50%
A:20%; 'L:15%; P:50%
A:20%; L:15%; P:50%
A:20%; L:15%; P:50%
A:20%; L:15%; P:50%
A:20%; L:15%; P:50%
A:20%; L:15%; P:50%
A:20%; L:15%; P:50%
A:20%; L:15%; P:50%
'Emissions measured in percent opacity during different operating modes: A=Acceleration; L=Lug; P=Peaks
during either mode.
"Total HC.
°In 1985, test cycle changed from steady-state to transient operation for HC, CO, and NOX, measurement and in
1988 for PM.
dOr 2.5 plus a limit of 0.5 nonmethane hydrocarbon (NMHC).
1 1994 and again in 1996. The NOX standard was reduced in 1998 for all on-road diesel engines,
2 bus and truck. For 2004, the standards were further lowered in a 1997 rulemaking, with limits on
3 non-methane hydrocarbon (NMHC) and NOX combined, but no further reductions in CO,
4 paniculate matter, or smoke. These lower NMHC-plus-NOx levels will very likely be confirmed
5 in the "1999 technology review" of these standards. EPA is currently evaluating further
6 reductions in NOX and particulate matter for the post-2004 time frame.
7 In December 1997, the EPA adopted emission standards for NOX, HC, CO, PM, and
8 smoke for newly manufactured and remanufactured railroad locomotives and locomotive
9 engines. The rulemaking, which takes effect in the year 2000, applies to locomotives originally
10 manufactured from 1973, any time they are manufactured or remanufactured (locomotives
U originally manufactured before 1973 are not regulated). Three sets of emission standards have
11/5/99
2-6
DRAFT-DO NOT CITE OR QUOTE
-------
1 manufactured from 1973 through 2001 (Tier 0), from 2002 through 2004 (Tier 1), and in 2005
2 and later (Tier 2) (Table 2-2; see EPA web page at hrtp://w\\-w.epa.gov/omswww/ or
3 http://www.dieselnet.com/standards/ for current information on mobile source emission
4 standards). The emissions are measured over two steady-state test cycles which represent two
5 different types of service, including the line-haul (long-distance transport) and switch (involved
6 in all transfer and switching operations in switch yards) locomotives.
7 The EPA emission trends report (U.S. EPA, 1998a) provides emission inventories for
8 criteria pollutants (PM10, PM2.5, SO2, NOX, VOC, CO, Pb, and NH3) from point, area, and
9 mobile sources, which indicate how emissions have changed from 1970 to 1977. For the
10 purposes of this document, "primary and secondary emissions from diesel engines (on-
11 and off-road) are briefly discussed for PM10, sulfur dioxide (SO2), nitrogen oxides (NOX), and
12 volatile organic compounds (VOC).
13 Mobile-source paniculate emissions come from both gasoline- and diesel-powered
14 engines in on-road vehicles and from a number of nonroad sources. Nonroad sources include
15 aircraft, commercial boats (which are mainly diesel-powered), construction equipment,
16 agricultural equipment, lawn/garden equipment, and other sources. The EPA emission trends
17 report shows that among point, area, and mobile sources (excluding fugitive dust sources),
18 mobile sources are responsible for 24% of PM10 emissions, with stationary sources (fuel
19 combustion and industrial processes) responsible for the remainder.
20 Particulate emissions from diesels are much greater than those from gasoline-fueled
21 engines. Particulate emissions (PM10) from gasoline-fueled engines decreased dramatically in
22 1975 with the widespread introduction of unleaded gasoline. Particulate emissions from diesel
23 highway vehicles have decreased recently because of EPA emission standards for new model
Table 2-2. Emission standards: locomotives (g/bhp/hr)
Line-haul
Switch
Line-haul
Switch
Line-haul
Switch
Year1
1973-200 1 (Tier 0)
1973-2001 (TierO)
2002-2004 (Tier 1)
2002-2004 (Tier 1)
2005 + (Tier 2)
2005 + (Tier 2)
CO
5.0
8.0
2.2
2.5
1.5
2.4
HC
1.0
2.1
0.55
1.2
0.3
0.6
NOX
9.5
14.0
7.4
11.0
5.5
8.1
PM
0.6
0.72
0.45
0.54
0.20
0.24
•Date of engine manufacture.
11/5/99
2-7
DRAFT-DO NOT CITE OR QUOTE
-------
1 year HD diesel trucks that were first implemented in 1988 and became increasingly stringent in
2 1991 and 1994, as presented in Table 2-1 above.
& The EPA emission trends report indicates that annual on-road vehicle PM10 emissions
3 decreased from 397,000 tons to 268,000 tons from 1980 to 1997. Passenger car paniculate
5 emissions decreased from 120,000 to 56,000 tons (53%) in this time frame while diesel vehicle
6 emissions decreased much less, from 208,000 to 163,000 tons (22%). Nonroad diesel engine
7 paniculate emissions decreased from 439,000 tons in 1980 to 316,000 tons in 1997 (28%).
8 Emissions data for PM2.5 are available only for the period from 1990 to 1997, prohibiting an
9 analysis of emission trends over the same time period as the other pollutants. For comparison to
10 PM10, annual on-road diesel vehicle PM2.5 emissions were estimated at 144,000 tons in 1997
11 and nonroad diesel PM2.5 emissions in 1997 were 290,000 tons.
12 Diesel engines also contribute to secondary PM formation from NOX and SO2 emissions
13 that are converted to nitrate and sulfate, although the direct emission of carbonaceous diesel
14 particulates are much greater than secondary nitrate or sulfate formation. In 1997, about 50% of
15 total ambient NOX came from mobile sources, with diesels responsible for 26%, or approximately
16 half of the mobile source contribution. About 6% of SO2 came from mobile sources in 1997,
17 with diesels responsible for 80% of that total. VOC emissions from diesel engines in 1997 were
18 estimated at 4% of the total emissions from all sources.
•LQ
^ff 2.2.3. Engine Technology Description and Chronology
21 NOX emissions, PM emissions, and brake-specific fuel consumption (BSFC) are among
22 the parameters that are typically considered during the development of a diesel engine. Many
23 engine variables that decrease NOX can also increase PM and BSFC. One manifestation of the
24 interplay among NOX, PM, and BSFC is that an increase in combustion temperatures will tend to
25 increase NO formation via the Zeldovich mechanism, will often improve thermal efficiency, can
26 improve BSFC, and can increase the rate of PM oxidation, thus lowering PM emissions. One
27 example of this is the tradeoff of PM emissions and BSFC versus NOX emissions with fuel
28 injection timing. Many recent advances in reducing the engine-out emissions of diesel engines
29 are combinations of technologies that provide incremental improvements in the tradeoffs among
30 these different emissions and fuel consumption. The sum total, though, can be considerable
31 reductions in regulated emissions within acceptable levels of fuel consumption.
32 The majority of current HD diesel truck engines certified for use in the United States
33 utilize:
34
3_5
1175/99 2-8 DRAFT-DO NOT CITE OR QUOTE
-------
1 • a 4-stroke cycle;
2 • direct-injection, high-pressure (1200 bar to >2000 bar) fuel injection systems with
3 electronic control of injection timing and, in some cases, injection rate;
4 • centrally located multihole injection nozzles;
5 • 3 or 4 valves per cylinder;
6 • turbochargers;
7 • in many cases, air-to-air aftercooling; and
8 • in some cases, the use of an oxidation catalyst.
9
T
10 These features have phased into use with HD truck engines because they offer a relatively
11 good combination of fuel consumption, torque-rise, emissions, durability, and the ability to better
12 "tune" the engines for specific types of applications. Fuel consumption, torque-rise, and
13 drivability have been maintained or improved while emissions regulations have become more
14 stringent. Many Class 8a and 8b diesel truck engines are now capable of 700,000 to 1,000,000
15 miles of driving before their first rebuild, and can be rebuilt several times because of their heavy
16 construction and the use of removable cylinder liners. This is several times the regulatory
17 estimate of full useful life for HD engines (290,000 miles) previously used by EPA.
18 Current 4-stroke locomotive engines use engine technology similar to on-highway diesel
19 engines, except that electronic controls have only recently been introduced. It is difficult to
20 separate the components of current high-speed diesel engines for discussion of their individual
21 emissions effects. Most of the components interact in numerous ways that affect emissions,
22 performance, and fuel consumption.
23
24 2.2.3.1. Injection Rate
25 Decreasing the duration of diffusion combustion and promoting soot oxidation during the
26 expansion stroke can reduce formation of soot agglomerates (Stone, 1995). Both of these effects
27 are enhanced by increasing the fuel injection rate. The primary means of accomplishing this is
28 by increasing fuel injection pressure. Increased injection rate can significantly reduce soot
29 emissions, but it can also increase combustion temperatures and cause an increase in NOX
30 emissions (Springer, 1979; Watson and Janota, 1982; Stone, 1995). However, when combined
31 with turbocharging, aftercooling, and injection timing retard, low NOX, low PM, and relatively
32 good BSFC and brake mean engine pressure (BMEP) are possible.
33 In 1977 Robert Bosch introduced a new type of high-pressure pump capable of producing
34 injection pressures of 1700 bar at the nozzle (Voss et ai., 1977). This increased fuel injection
35 pressure by roughly a factor of 10. Unit injection, which combines each fuel injection nozzle
11 /5/99 2-9 DRAFT-DO NOT CITE OR QUOTE
-------
1 with individual cam-driven fuel pumps, can achieve very high injection pressures (>2000 bar).
2 The first combination of unit injectors with electronically controlled solenoids for timing control
&. was offered in the United States by Detroit Diesel Corporation in the 1988 model year (Hames et
V al., 1985). Replacement of the injection cam with hydraulic pressure, allowing a degree of
5 injection rate control, was made possible with the hydraulic-electronic unit injection (HEUI)
6 jointly developed by Caterpillar and Navistar, introduced on the Navistar T444E engine (and
7 variants) in 1993.
8 It is widely known that high fuel injection pressures have been used to obtain compliance
9 with the PM standards that went into effect in 1988 (Zelenka et al., 1990). Thus, it is likely that
10 a transition to this technology began in the 1980s, with the vast majority of new engine sales
11 employing this technology by 1991, when the 0.25 g/bhp-h Federal PM standard went into effect.
12 The use of electronic control of injection rate is rapidly increasing on medium-HD diesel
13 engines equipped with HEUI (currently available on Caterpillar 3126 and Navistar T444E,
14 DT466, and 530E engines). Engines are currently under development, perhaps for 2002-2004
15 introduction, that use common-rail fuel injection systems with even more flexible control over
16 injection pressure and timing than previous systems.
17
18 2.2.3.2. Turbocharging, Charge-Air Cooling, and Electronic Controls
12. Use of exhaust-driven turbochargers to increase intake manifold pressure has been
W applied to both IDI and DI diesel engines for more than 40 years. Turbocharging can decrease
21 fuel consumption compared to a naturally aspirated engine of the same power output.
22 Turbocharging utilizes otherwise wasted exhaust heat to generate intake boost. The boosted
23 intake pressure effectively increases air displacement and increases the amount of fuel that can be
24 injected to achieve a given fuel-air equivalence ratio. Turbocharging increases the power density
25 of an engine. Boosting intake pressure via turbocharging and reducing fuel-to-air ratio at a
26 constant power can significantly increase both intake temperatures and NOX emissions.
27 Increased boost pressure can significantly reduce ignition delay, which reduces VOC and PM
28 soluble organic fraction (SOF) emissions (Stone, 1995) and increases the flexibility in selection
29 of injection timing. Injection timing on turbocharged engines can be retarded further for NOX
30 emission control with less of an effect on PM emissions and fuel consumption. This allows a
31 rough parity in NOX emissions between turbocharged (non-aftercooled) and naturally aspirated
32 diesel engines (Watson and Janota, 1982).
33 Turbocharging permits the use of higher initial injection rates (higher injection pressure),
34 which can reduce particulate emissions. Although this may offer advantages for steady-state
35 operation, hard accelerations can temporarily cause overly fuel-rich conditions because the
1175/99 2-10 DRAFT-DO NOT CITE OR QUOTE
-------
1 turbocharger speed lags behind a rapid change in engine speed (turbo-lag). This can cause
2 significant increases in PM emissions during accelerations. Before the advent of electronic
3 controls, the effect of acceleration on PM emissions could be limited by mechanically delaying
4 demand for maximum fuel rate with a "smoke-puff eliminator." Since this device also limited
5 engine response, there was considerable incentive for the end-users to remove or otherwise
6 render the device inactive. Charge-air cooling, for example using an air-to-air aftercooler (air-
7 cooled heat exchanger) between the turbocharger compressor and the intake manifold, can
8 greatly reduce intake air and peak combustion temperatures. When combined with injection
9 timing retard, charge-air cooling allows a significant reduction in NOX emissions with acceptable
10 BSFC and PM emissions when compared to either non-aftercooled or naturally aspirated diesel
11 engines (Hardenberg and Fraenkle, 1978; Pischinger and Cartellieri, 1972; Stone, 1995) (Figure
12 2-2).
13 Electronic control of fuel injection timing allowed engine manufacturers to carefully
14 tailor the start and length of the fuel injection events much more precisely than through
15 mechanical means. Because of this, newer on-highway turbocharged truck engines have
16 virtually no visible smoke on acceleration. Electronic controls also allowed fuel injection retard
17 under desirable conditions for NOX reduction, while still allowing timing optimization for
18 reduced VOC emissions on start-up, acceptable cold-weather performance, and acceptable
19 performance and durability at high altitudes. Previous mechanical unit injected engines (e.g., the
20 1980s Cummins L10, the non-DDEC DDC 6V92) were capable of reasonably high injection
130 n
120
110 ]
100
90 -
80
70
Naturally Aspirated
Turbocharged/Aftercooled
60 65 70 75 80 85 90 95 100 105 110 115
%
Figure 2-2. Effect of turbocharging and aftercooling on NOX and PM (Mori, 1997).
11/5/99
2-11
DRAFT-DO NOT CITE OR QUOTE
-------
1 pressures, but had fixed injection timing that only varied based on the hydraulic parameters of
2 the fuel system. Many other engines with mechanical in-line or rotary injection pumps had only
t coarse injection timing control or fixed injection timing.
Precise electronic control of injection timing over differing operating conditions also
5 allowed HD engine manufacturers to retard injection timing for low NOX emissions during highly
6 transient urban operation similar to that found during emissions certification, and advance the
7 injection timing during less transient operation (such as freeway driving) for fuel consumption
8 improvements (-3% to 5%) at the expense of greatly increased NOX emissions (~3 to 4 times
9 regulated levels). This particular situation resulted in the recent consent-decree settlements
10 between the Federal Government and most of the HD engine manufacturers to assure effective
11 NOX control in all driving conditions.
12 Turbocharged engines entered the market very slowly beginning in the 1960s. During the
13 years 1949 to 1975 the total improvement in emissions for the on-road diesel fleet was
14 considerably less than 10%-20% for gaseous emissions and, for particulates, there was really no
15 change at all until the advent of particulate standards in 1988. Charge air cooling was introduced
16 during the 1960s and was initially performed in a heat exchanger using engine coolant. Cooling
17 of the charge air using ambient air as the coolant was introduced by Mack in 1977 with
18 production of the ETAY(B)673A engine (Heywood, 1988). Use of ambient air allowed cooling
.i9 of the charge air to much lower temperatures. Most HD diesel engines sold today employ some
"^O form of charge air cooling, with air-to-air aftercooling the most common. Johnson and co-
21 workers (Johnson et al., 1994) have presented a comparison of similar engines that differ in that
22 the charge air is cooled by engine coolant (1988 engine) and by ambient air with a higher boost
23 pressure for the second (1991 engine). The 1991 engine also used higher pressure fuel injectors.
24 The 1991 engine exhibited both lower PM (50%) and NOX emissions. Higher injection pressure
25 likely enabled the reduced PM emissions, while the lower charge air temperature and the ability
26 to electronically retard the injection timing under some conditions likely enabled the lower NOX
27 emissions.
28 It is apparent on the basis of both the literature and certification data that turbochargers
29 with aftercoolers can be used in HD engines in conjunction with other changes to result in a
30 decrease in emissions. NOX was probably reduced on the order of 10% to 30% in turbocharged
31 aftercooled engines with retarded injection timing. Prior to the late 1970s, only a portion of all
32 HD diesel engines were rurbocharged, so the total improvement in emissions that could be
33 associated with these changes was considerably less until more stringent emissions regulations
34 were implemented. The lowest combination of in-use NOX and PM emissions would likely be
35 for turbocharged aftercooled engines that used retarded, high-pressure unit injection without
1175/99 2-12 DRAFT-DO NOT CITE OR QUOTE
-------
1 electronic control in the early 1990s. Although tighter NOX standards phased in for model years
2 1994 and 1998, this is complicated by the instances of defeating NOX control during cruise
3 conditions by most engine manufacturers. Defeat of NOX control occurred to a different extent
4 with all Class 8 electronically controlled engines beginning with their introduction in 1988. PM
5 emissions were likely much lower for engines on which electronic controls were introduced, but
6 NOX emissions in-use were likely much higher than for early electronic or late mechanically
7 injected versions of the engines. Overall, it is expected that engines in the 1950 to 1980 time
8 frame would have PM emissions similar to those of the mid-1980 engines that were not yet
9 controlled for particulates, while later engines would have lower PM emissions.
10
11 2.2.3.3. Indirect and Direct Injection High-Speed Diesel Engines
12 Prior to the 1930s, diesel engine design was limited to relatively low-speed applications
13 because sufficiently high-pressure fuel injection equipment was not available. With the advent of
14 high-speed and higher pressure pump-line-nozzle systems, introduced by Robert Bosch in the
15 1930s, it became possible to inject the fuel directly into the cylinder for the first time, although
16 IDI diesel engines continued in use for many years. As diesels were introduced into the heavy
17 truck fleet in the 1930s through the 1950s, both IDI and DI naturally aspirated variants were
18 evident. A very low-cost, rotary injection pump technology was introduced by Roosa-Master in
19 the 1950s, reducing the cost of DI systems and allowing their introduction on smaller
20 displacement, higher speed truck engines.
21 DI diesel engines have now all but replaced IDI diesel engines for HD on-highway
22 applications'. IDI engines typically required much more complicated cylinder head designs, but
23 generally were capable of using less sophisticated, lower pressure injection systems with less
24 expensive single-hole injection nozzles. IDI combustion systems are also more tolerant of lower
25 grades of diesel fuel. Fuel injection systems are likely the single most expensive component of
26 many diesel engines. Caterpillar continued producing both turbocharged and naturally aspirated
27 IDI diesel engines for some on-highway applications into the 1980s. Caterpillar and Deutz still
28 produce engines of uiis type, primarily for use in -jndergrcund ninbg applications. IDI
29 combustion systems are still used in many small-displacement (<0.5 L/cylinder), very high-speed
30 (>3000 rpm rated speed) diesel engines for small offroad equipment (small imported tractors,
31 skid-steer loaders), auxiliary engines, and small generator sets.
'The GM Powertrain/AM General 6.5L electronically controlled, turbocharged IDI-
swirlchamber engine, certified as a light-HD diesel truck engine, is the last remaining HD on-
highway IDI engine sold in the United States.
1175/99 2-13 DRAFT-DO NOT CITE OR QUOTE
-------
1 IDI engines have practically no premixed burn combustion, and thus are often quieter and
2 have somewhat lower NOX emissions than DI engines. Electronic controls, high-pressure
injection (e.g., GM 6.5), and 4-valve/cylinder designs (e.g., the 6-cylinder Daimler LD engine)
can be equally applied to IDI diesel engines as with their DI counterparts, but negate any
5 advantages in cost over DI engines. DI diesel engines of the same power output consume 15%-
6 20% less fuel than IDI engines (Heywood, 1988). Considering the sensitivity of the HD truck
7 market to fuel costs, this factor alone likely accounts for the demise of IDI diesel engines in these
8 types of applications. Throttling and convective heat transfer through the chamber-connecting
9 orifice, and heat rejection from the increased surface area of IDI combustion systems, decreases
10 their efficiency and can cause cold-start difficulties when compared to DI designs. Most IDI
11 diesel engine designs require considerably higher than optimum (from an efficiency standpoint)
12 compression ratios to aid in cold starting (19:1 to 21:1 versus -15:1 to 17:1 for DI engines).
13 Because of the early introduction of DI technology into truck fleets, it is likely that by
14 end of the 1970s, only a small fraction of the HD diesel engines sold for on-highway use were
15 IDI engines. It is unlikely that the gradual shift from IDI to DI engine designs through the 1960s
16 and 1970s had any significant impact on emissions.
17
18 2.2.3.4. Two-Stroke and 4-Stroke High-Speed Diesel Engines
«A detailed discussion of the 2- and 4-stroke engine cycles can be found in Heywood,
Taylor, or Stone, and so will not be presented here (Heywood, 1988; Taylor, 1990; Stone, 1995).
21 Nearly all high-speed 2-stroke diesel engines utilize uniflow scavenging assisted by a positive
22 displacement blower (Figure 2-3). Unifiow-scavenged 2-stroke diesels use poppet exhaust
23 valves similar to those found in 4-stroke engines. The intake air enters the cylinder through a
24 pressurized port in the cylinder wall. A crankshaft-driven, positive-displacement blower (usually
25 a roots-type) pressurizes the intake port to ensure proper scavenging. A turbocharger may be
26 added to the system to provide additional boost upstream of the blower at higher speeds, and to
27 reduce the size and parasitic losses associated with the positive-displacement blower.
28 Two-stroke diesel engines can achieve efficiency comparable to 4-stroke counterparts and
29 have higher BMEP (torque per unit displacement) (Heywood, 1988). It is useful to note that the
30 2-stroke cycle fires each cylinder once every revolution, while the 4-stroke cycle fires every
31 other revolution. Thus, for a given engine size and weight, 2-strokes can produce more power.
32 However, 2-stroke diesel engines are less durable than their 4-stroke counterparts. Lubricating
33 oil is transferred from the piston rings to the intake port, which causes relatively high oil
11/5/99 2-14 DRAFT-DO NOT CITE OR QUOTE
-------
Engine
Exhaust
Air Intake
Ports
Intake Air
Positive
Displacement
Blower
Figure 2-3. An example of uniflow scavenging of a 2-stroke diesel engine with a positive
displacement blower (Adapted from Taylor, 1990). Scavenging is the process of
simultaneously emptying the cylinder of exhaust and refilling with fresh air.
1 consumption relative to 2-stroke designs. Durability and low oil consumption are desirable for
2 on-highway truck applications. This may be why 4-stroke engines have been favored for these '
3 applications since the beginning of dieselization in the trucking industry, with the notable
4 exception of urban bus applications. Although it is no longer in production, the Detroit Diesel
5 6V92 series of 2-stroke diesel engines is still the most popular for urban bus applications, where
6 the high power density allows the engine to be more easily packaged within space limitations.
7 The primary reason that 2-stroke engines like the 6V92 are no longer offered for urban bus
8 applications is PM emissions. The reduced lubricating oil control with 2-strokes tends to
9 increase VOC and organic PM emissions relative to 4-stroke designs. This was particularly
10 problematic for urban bus applications because urban bus engines must meet tighter Federal and
11 California PM emissions standards. The current urban bus PM standard (0.05 g/bhp-hr) is half of
12 the current on-highway HD diesel engine PM standard. No 2-stroke diesel engine designs have
13 been certified to meet the most recent urban bus PM emissions standards, and Detroit Diesel
14 Corporation has not certified a 2-stroke diesel engine for on-highway truck use since 1995.
15 A comprehensive review of emissions from hundreds of later model vehicles (1976-1998)
16 found no significant difference between 2- and 4-stroke vehicles (Yanowitz et al., 1999a).
17 Overall, regulated emissions changes due to changing proportions of 2- and 4-stroke engines in
18 the in-use fleet during the years 1949-1975 do not appear to be significant for HD truck and bus
11/5/99
2-15
DRAFT-DO NOT CITE OR QUOTE
-------
1 engines. Furthermore, it appears that the proportion of 2-stroke engines in the in-use fleet was
2 relatively constant until the 1980s.
T' 2.2.3.5. Near-Term Diesel Emission Reduction Technologies
5 2.2.3.5.1. Exhaust gas recirculation. Exhaust gas recirculation (EGR) (i.e., routing some of the
6 exhaust gas to the intake manifold) is widely used in LD SI gasoline engines to control NOX
7 emissions. Unlike most SI applications, use of EGR with diesels necessitates both electronic
8 control of the EGR as well as EGR cooling to limit the associated increase in PM. Because EGR
9 displaces part of the intake air, it can increase the overall fuel-to-air equivalence ratio to a point
10 that can lead to large increases in PM emissions. Hot EGR further exacerbates this problem by
11 increasing the temperature of the intake air. The increased temperature decreases air density and
12 further reduces the volume of intake air entering the engine.
13 Cooled EGR systems typically use an engine coolant heat exchanger to cool the
14 recirculated exhaust gases before mixing with the intake air. EGR cooling has the potential to
15 significantly reduce the increase in intake air temperature associated with EGR. This would
16 mitigate (though not eliminate) the PM emissions penalty associated with diesel EGR systems.
17 Though EGR cooling greatly extends the operational range over which EGR can be used,
18 electronic control of the EGR will be necessary to prevent large PM increases under hard
^k acceleration, near peak torque conditions, or at high altitudes. EGR cooling can also reduce
txr combustion temperatures beyond uncooled EGR, resulting in further decreases in NOX emissions
21 relative to uncooled EGR under certain conditions (Kakoi et al., 1998; Leet et al., 1998).
22 Cooled EGR is currently used with the relatively small number of LD diesel vehicles sold
23 for the U.S. market. Although it is not widely used in HD diesel engines today, many believe that
24 cooled EGR will be an important technology for future NOX reductions (Johnson et al., 1994;
25 Zelenka, 1990). Most, if not all, of the diesel engines that will meet either the 2002 consent
26 decree requirements (early compliance with 2004 standards) or the 2004 emissions standards for
27 HD trucks will incorporate some form of cooled EGR into their engine designs to meet the 2.5
28 g/bhp-hr NOX + NMHC standard.
29 Ladomatos et al. (1996-1997) have described the three mechanisms by which EGR is
30 thought to lead to reduced emissions of NOX:
31
32 • Dilution: Recirculating exhaust gas leads to a reduction in the oxygen content of
33 the intake charge. Although this increases ignition delay, it also reduces peak
34 temperature. With respect to NO formation, the increased ignition delay and
3JL premixed burn fraction are more than offset by the dilution effects on peak
11/5/99 2-16 DRAFT-DO NOT CITE OR QUOTE
-------
1 . temperature, resulting in a significant reduction in NOX emissions. The reduction
2 in oxygen content can also cause an increase in HC, CO, and PM emissions. The
3 dilution mechanism is thought to be, by far, the most important mechanism
4 affecting diesel engine emissions.
5 • Thermal: The recirculated exhaust gas contains CO2 and water vapor, which
6 increase the specific heat of the intake charge. This lowers peak temperature and
7 hence formation of NOX is decreased.
8 • Chemical: It is possible that endothermic dissociation of recirculated CO2 and
9 water lowers peak temperature, leading to a reduction in formation of NOX.
10
11 Early studies of uncooled diesel EGR were conducted by Pischinger and Cartellier (1972)
12 and Springer (1979). Although sub-4 g/bhp-hr NOX emissions levels were possible, the high PM
13 emissions associated with the NOX reductions delayed the introduction of EGR until fuel sulfur
14 levels were low enough to enable engine-coolant cooling of EGR. Theoretically, further cooling
15 of the EGR (for example, air cooling) would extend the range of engine operating conditions
16 under which EGR could be used, but is not possible at current fuel sulfur levels because of the
17 potential for very high levels of sulfuric acid condensation in the EGR cooler (McKinley, 1997;
18 Kreso et al., 1998a; Leet et al., 1998).
19 Johnson (1994) noted that engine durability is a serious concern with EGR because
20 recirculation of soot through the engine can increase wear. Kreso and co-workers (Kreso et al.,
21 1998b) recently examined a 1995 Cummins M-l 1 at two steady-state modes and two EGR rates.
22 The EGR system included cooling of the recirculated gas. EGR was effective at reducing NOX,
23 and reductions as high as 56% were observed under some conditions. Emissions of PM
24 increased by as much as 57%, while emissions of the SOF portion of PM were somewhat lower.
25 Examination of the mutagenicity of SOF with the Ames assay indicated that the SOF produced
26 by EGR was more mutagenic.
27 It is clear that any EGR system will require careful control so that EGR is applied only
28 under operating conditions where significant NO, reductions can be obtained without a major
29 increase in PM. There is little evidence to suggest that the character of PM, SOF, or gaseous
30 hydrocarbon emissions is dramatically altered by use of EGR.
31
32 2.2.3.5.2. Diesel oxidation catalysts (DOC). DOCs for HD diesel applications were originally
33 developed for underground mining equipment for exhaust odor and CO control (typically not
34 issues for uicscl cngiiics outside cf confined environments). The use cf early high platinum-
35 content DOCs was an issue for these applications because of their high levels of NO to NO2 and
1175/99 2-17 DRAFT-DO NOT CITE OR QUOTE
-------
1 SO2 to SO3 oxidation (McClure, 1992). McDonald et al. (1995) found that the SO3 oxidation rate
2 was sufficient to produce -0.2 g/bhp-nr sulfate PM emissions at high load conditions even with a
^ relatively low-sulfur diesel fuel (0.01% S).
"^ Later DOCs were developed that relied more on base metals and less on precious metals
5 for VOC oxidation (lowering SOF PM) while limiting high-temperature formation of sulfuric
6 acid PM. These types of catalysts were first applied to LD diesel vehicles in the 1980s, some
7 urban bus applications (1994 Cummins L10), and a number of medium-HD diesel engines after
8 1993 (Navistar T444E, some versions of the Caterpillar 3116 and 3126). There are also a
9 number of DOCs that are now being retrofitted to older urban buses as part of the EPA Urban
10 Bus Retrofit and Rebuild Program. Current DOCs oxidize more than 70% of the VOCs that
11 contribute to SOF PM, leading to a 15%-30% reduction in total PM emissions (Farrauto et al.,
12 1996; Brown and Rideout, 1996; Tamanouchi et al., 1998).
13 DOCs are highly effective at oxidizing lube oil components (Farrrauto et al., 1996) as
14 well as most PAHs (Mitchell et al., 1994; Pataky et al., 1994; Bagley et al., 1996; McDonald,
15 1997; Bagley et al., 1998). There are conflicting data as to whether DOCs catalyze the formation
16 or oxidation of nitro-PAH compounds. Bagley and co-workers (Bagley et al., 1998) and
17 McDonald (1997) found reductions in both PAHs and nitro-PAH and associated mutagenic
18 activity for a low-sulfate-forming base-metal/Pt/Pd oxidation catalyst that were statistically
1&. significant atp<.01, and found only one nitro-PAH (1-nitropyrene) above minimum detection
av limits in either catalyzed or uncatalyzed exhaust. Mitchell and co-workers (1994) found
21 decreases in PAHs with twofold increases in nitro-PAH (statistical significance is not known).
22 More comprehensive testing will be necessary to draw further conclusions about the effects of
23 DOCs on nitro-PAH.
24
25 2.2.3.6. Future (2004+) Diesel Emission Reduction Technologies
26 2.2.3.6.1. NOX storage catalysts. NOX storage catalysts currently under development might be
27 used to meet 2007 HD diesel engine standards if diesel fuel sulfur levels are considerably
28 reduced (0-30 ppm S fuel may be necessary). A generalized schematic of the operation of this
29 device is included in Figure 2-4. This catalyst system employs a high-platinum (Pt) content
30 catalyst for oxidation of NO to NO2 (in the absence of an oxidation catalyst, total NOX in diesel
31 exhaust is primarily NO [typically >90%] with lesser amounts of NO2). The NO2 is then stored,
32 using one of a number of barium compounds, as barium nitrate. For approximately 2-second
33 durations every 2 minutes, diesel fuel is either sprayed into the exhaust or injected into the
34 cylinder after combustion to provide the necessary hydrocarbons to remove the NOX from the
35 storage components. The NOX is then reduced over a standard three-way catalytic converter.
11/5/99 2-18 DRAFT-DO NOT CITE OR QUOTE
-------
c
NO
)xidizing (Lean) Conditions (4> < 1.00)
Pt Catalyst
NO2
Trap
(NO2 stored as
barium nitrate)
Reducing (Rich) Conditions (<{> = 1.02)
HC,
H20, CO2, N2
Approximately 2 seconds of stored nitrate regeneration needed for
every 2 minutes of operation.
Figure 2-4. NOx-storage catalyst operation under oxidizing and reducing conditions.
1 The average NOX reduction potential for this technology over the FTP is 50% to 75%, with a fuel
2 consumption penalty of approximately 3% to 5% (Wall, 1998). Figure 2-5 compares the NOX
3 reducing capabilities of a NOX storage catalyst system to a representative sulfur-tolerant NOX
4 catalyst system.
5
6 2.2.3.6.2. Lean-NOx catalysts. Various types of active (requiring a post-combustion fuel
7 injection event) and nonactive (no post-injection) lean-NOx catalysts are in production or are
8 under investigation for continuous reduction of NOX emissions in lean exhaust environments
9 such as those present in diesel exhaust. (These are continuous devices, as opposed to the cyclic
10 nature of NOX reduction using NOX storage catalysts.) Lean-NO- catalysts typically reduce NOX
11 efficiently over a very narrow range of exhaust temperatures. There are both high- and low-
12 temperature varieties of lean-NOx catalysts. Low-temperature, platinum-based lean-NOx
13 catalysts using zeolites for support, catalyst promotion, and adsorption of NOX and HC would be
14 typical of a lear;-NOx catalyst technology for medium and light-HD diesel applications.
15 High-temperature base-metal lean-NOx catalyst formulations (Cu-ZSM, for example) are under
16 investigation primarily for highly loaded HD diesel engine applications.
17 A number of new common-rail fuel injection systems are capable of injecting fuel after
combustion to provide additional hydrocarbons for use as an NOX reductant with lean-NOx
11/5/99
2-19
DRAFT-DO NOT CITE OR QUOTE
-------
Pt/Rh/Ba
NOx-Storage
Catalyst
Active Pt-Zeolite
Lean NOx Catalyst
200 300 400 500
Catalyst Inlet Temperature (°C)
Figure 2-5. A comparison of the NOX reduction efficiency over a range of temperature
conditions for the sulfur-intolerant NOX storage catalyst system and the more sulfur-
tolerant, active Pt-zeolite catalyst system. Although peak NO, reduction efficiencies for
various types of nonstorage Iean-NOx catalysts (similar to the Pt-Zeolite catalyst shown
here) approach 50%-60%, average reductions are 15% to 30% over various (FTP-75,
NEDC) driving cycles.
1 catalysts. Although active Pt-zeolite catalyst systems have higher NOX removal efficiencies than
2 similar nonactive catalyst systems, NOX removal efficiencies are still only in the range of 15% to
3 35% over the New European Drive Cycle (NEDC) (Peters et al., 1998; Engler et al., 1998) and
4 significantly below those of NOX storage catalyst systems (Figure 2-5). Newer systems use a
5 controlled fuel-exotherm over the platinum catalysts with feedback control to maintain a more
6 constant catalyst temperature, enabling higher NOX reduction efficiencies.
7
8 2.2.3.6.3. Selective catalytic reduction. Selective catalytic reduction (SCR) for NOX control is
9 currently available for stationary diesel engines, and has been proposed for mobile light- and
10 heavy-diesel applications. SCR uses ammonia as a reducing agent for NOX over a catalyst
I-*- composed of precious metals, base metals, and zeolite. The ammonia is supplied by introducing
11/5/99
2-20
DRAFT-DO NOT CITE OR QUOTE
-------
1 a urea/water mixture into the exhaust upstream of the catalyst. The urea/water mixture is
2 typically stored in a separate tank that must be periodically replenished. Ammonia has extremely
3 high selectivity as a reductant for NOX. NOX reductions of 70% to 90% over a broad range of
4 operating conditions are possible using such systems (Brown, 1998). These systems appear to be
5 tolerant of current U.S. on-highway diesel fuel sulfur levels for exhaust temperatures that are
6 consistent with heavier (Class 7, 8) HD on-highway applications and over the HD FTP test cycle
7 (40 CFR, Subpart N). NOX reduction efficiency drops considerably at exhaust temperatures less
8 than 200°C in the presence of SO2 in the exhaust. Therefore, the practical fuel sulfur limit for
9 LD diesel applications is probably somewhat less than 100 ppm. This reduced efficiency at low
10 temperatures and higher fuel sulfur levels may also have implications for "not-to-exceed" NOX
11 requirements for HD on-highway diesel engines introduced in the consent decrees and likely to
12 be a component of both the 2004 and 2007 HD diesel emissions standards.
13 Control of the quantity of urea injection into the exhaust, particularly during transient
14 operation, is an important issue with SCR systems. Injection of too large of a quantity of urea
15 leads to a condition of "ammonia slip," whereby excess ammonia formation can lead to both
16 direct ammonia emissions and oxidation of ammonia to produce (rather than reduce) NOX. There
17 are also a number of potential hurdles to overcome with respect to using a major emission control
18 system that requires frequent replenishing of a consumable fluid in order to function. This raises
19 issues related to supply, tampering, and the possibility of running the urea storage tank dry.
20 Packaging of the urea supply within the constraints of modern LD vehicles may also be
21 particularly challenging. Packaging of SCR systems does not appear to be a major problem for
22 HD truck or bus applications.
23
24 2.2.3.6.4. Continuously regenerating traps. One method of exhaust aftertreatment for
25 controlling diesel PM emissions is to pass diesel exhaust through a ceramic or metallic filter or
26 "PM trap" to collect the PM, and to use some means of burning the collected PM so that the trap
27 can be either periodically or continuously regenerated. Previous traps have used catalyzed
28 coatings, fuel additives, and electrical heating to assist trap regeneration. Failure to consistently
29 regenerate the trap can lead to plugging, excessive exhaust back-pressure, and eventually
30 overheating and permanent damage to the trap. Inconsistent regeneration due to the high
31 frequency of fairly low exhaust temperatures has been a particular problem in applying PM traps
32 to some lightly loaded diesel applications.
33 The recently developed continuously regenerating trap (CRT) has shown considerable
34 promise in a broad range of diesel applications because of its ability to regenerate even at fairly
35 low exhaust temperatures. The CRT uses nitrogen dioxide (NO2) to assist trap regeneration.
11 /5/99 2-21 DRAFT-DO NOT CITE OR QUOTE
-------
1 NO2 can oxidize soot collected within the trap at exhaust temperatures as low as 250°C (Hawker
2 et al., 1997), which is within the typical exhaust temperature range of many diesel vehicle and
truck applications (Liiders et al., 1997). The NO, is produced by oxidizing NO in the exhaust
using a high-platinum-content oxidation catalyst brick located immediately upstream of the
5 ceramic trap-filter. A general schematic of the CRT system is presented in Figure 2-6.
6 The CRT is capable of reducing PM emissions by more than 80% (Hawker et al., 1997;
7 Hawker et al., 1998). SO2 inhibition of NO oxidation effectively limits the CRT to use with
8 diesel fuel sulfur levels below 50 ppm.
9 In some cases, excess fuel can be used to induce an exotherm over the Pt-catalyst to
10 ensure that minimum soot oxidation temperatures are reached.
11 It appears likely that introduction of emission standards that would force the use of CRT
12 or similar catalyst/trap technologies would likely be accompanied by steep reductions in toxic
13 emissions. Hawker and co-workers found substantial reductions in gas- and PM-phase VOC,
14 soot, acetaldehyde, formaldehyde, and total particle number (Hawker, 1998). Considering that
15 the CRT incorporates a Pt-DOC, PAHs oxidation is probably also high, but this was not
16 determined in the study.
17
18 2.2.3.6.5. Possible effects of advanced aftertreatment systems. NO2 formation: One constant
among many of the various proposed diesel exhaust aftertreatment devices is the reliance on high
Pt content for some components. In some cases, lean reactions of NO to NO2 are integral to the
NO oxidizes
to NO-,
NO, PM other
exhuast constituents
Pt Catalyst
PM collects
in trap
PM
Trap
NO2 oxidizes PM,
forming NO, CO2, CO
Figure 2-6. Schematic showing the operating principles of the continuously regenerating
trap (CRT).
11/5/99
2-22
DRAFT-DO NOT CITE OR QUOTE
-------
1 design of the device (CRT, NOX storage catalyst). In the case of the CRT, >50% conversion of
2 NO to NO2 is desired for efficient trap regeneration (Figure 2-7). This could result in a
3 significant increase in direct NO2 emissions from diesel exhaust. In the case of NOX storage
4 catalysts and SCR, there may be less cause for concern because of the relatively high NOX
5 reduction efficiencies. Low-temperature lean-NOx catalysts have relatively high Pt contents, but
6 no data were found in the literature that quantified NO/NO2 emissions for these devices.
7 Sulfate PM: The relatively high conversion rates of fuel sulfur to sulfuric acid aerosol
8 possible with high-Pt content diesel exhaust aftertreatment systems are similar to those found
9 with early high-Pt DOCs (for example, a Pt lean-NOx catalyst in Figure 2-8), although it is likely
i
10 that broad introduction of advanced diesel exhaust aftertreatment systems through reductions in
11 standards for regulated emissions would be accompanied by significant fuel sulfur control.
12 Ammonia: Widespread use of urea-SCR catalyst systems could increase ammonia
13 emissions. Newer SCR designs are incorporating electronic control of urea injection and the use
14 of a "cleanup" catalyst for oxidation of excess ammonia to minimize ammonia emissions.
1 5 PMand VOC: Most of the aftertreatment systems under development are still too new to
16 have been subjected to comprehensive exhaust speciation analyses. The CRT has the additional
problem that PM emissions are so low that it is difficult to collect a large enough PM sample for
10 ppm
50 ppm
100 ppm
500 ppm
1500 ppm
150
250
350
450
550
Figure 2-7. Efficiency of JNU to JNO2 conversion over the oxidation catalyst component of
the CRT at different exhaust temperatures and at differing diesel fuel sulfur levels.
11/5/99
2-23
DRAFT-DO NOT CITE OR QUOTE
-------
0.1
0.09
0.08
0.07
0.08
O.OS
£ 0.0«
in
0.03
0.02
O.OT
Tier 1 LDV PM Standard
Proposed Ca-LEV IITLEV PM Standard
Proposed Ca-LEV n LEV PM Standard
100 200 300 400
Fuel Sulftir Content (ppm)
500
600
Figure 2-8. Estimated sulfate (primarily H2SO4) PM emissions from a LD truck equipped
with a low-temperature Pt-zeolite lean-NOx catalyst system (Wall, 1998).
2
3
4
5
6
7
8
9
10
11
12
13
14
15
compounds to be above their minimum detection limits. Because all of these devices incorporate
oxidation catalyst functions to some extent, some of the comments related to oxidation catalysts
also apply here. One major difference is that some of the aftertreatment devices rely on (lean-
NOX catalyst, NOx-storage catalyst), or are sometimes assisted by (CRT), the introduction of
additional fuel hydrocarbons, either as a reductant or to maintain a high exhaust temperature.
The possible species formed from the oxidation or partial oxidation of fuel hydrocarbons have
not been determined.
2.2.4. History of Dieselization
2.2.4.1. Dieselization of the On-Road Fleet
Understanding the prevalence of diesel engine penetration into the motor vehicle market
is an important aspect of estimating the potential health effects of diesel emissions, past and
present. Two classification systems based on rated gross vehicle weight are in use for trucks.
These are listed below (Table 2-3).
11/5/99
2-24
DRAFT-DO NOT CITE OR QUOTE
-------
Table 2-3. Vehicle classification and weights for on-road trucks
Class
3
4
5
6
7
8Aa
8Ba
Medium duty (MD)
Light-heavy duty (LHD)
Heavy-heavy duty (HHD)
Weight (Ib)
10,001-14,000
14,001-16,000
16,001-19,500
19,501-26,000
26,001-33,000
33,001-60,000
>60,000
10,001-19,500 (same as Classes
19,501-26,000 (same as Class
3-5)
6)
>26,001 (same as Class 7-8)
'Class 8A and Class 8B are often considered together.
1 New diesel vehicle sales data for weight classes 5-8 are shown in Figure 2-9 for the years
2 1957-1998. The number of Class 7 and 8 diesel trucks sold has increased steadily with time
3 while the number of smaller Class 5 and 6 trucks sold peaked in the 1960s and early 1970s and
4 has since decreased. Retail and factory sales data show an increase in the percentage of diesel
5 engines used in trucks sold in Classes 5-8. Using data from factory and retail sales, the
6 percentage of diesel trucks sold by class is shown for the years 1957-1998 in Figure 2-10. The
7 increase in the use of diesel relative to other fuels first occurred for Class 8 trucks. By 1983
8 more than 97% of the Class 8 trucks sold had diesel engines, according to Navistar and Motor
9 Truck Facts (Bunn, 1999; AAMA Motor Vehicle Facts & Figures, 1983). Before 1980, about
10 60% of the Class 7 trucks were diesel but very few of the Class 5 and 6 trucks were diesel (< 16%
11 combined). Use of diesel engines in these weight classes increased substantially in the 1980s,
12 with roughly 80% of Class 6 and 67% of Class 7 trucks sold in 1997 being diesel.
13 Additional insight into dieselization of the on-road fleet can be gained from the 1992 U.S.
14 Census of Transportation (1995). A summary of results is presented in Table 2-4. These data
15 indicate that in 1992 the Class 7 plus Class 8 fleet was 88% diese!. The data presented in Figure
16 2-10 for the combined fleet in 1992 are in agreement with this value. Dieselization for Class 6 in
17 1992 was only 37% and for Classes 3-5 only 26%.
18 The 1992 Census of Transportation also provides information on the model year
19 distribution for vehicles of various weight classes (Figure 2-11). A few 1993 model year
11/5/99
2-25
DRAFT-DO NOT CITE OR QUOTE
-------
300000
250000
200000
B 150000
100000
soooo
1950 IP55 I960 IMS 1970 197S 1980 1985 1990 1995 2000 2005
Year
Figure 2-9a. Number of HD diesel trucks sold in years 1957-1998 based on industry sales
data.
Source: Bunn(1999).
11/5/99
2-26 DRAFT-DO NOT CITE OR QUOTE
-------
2
o
1
3
z
180 000
160 000
140 000
120 000 -
100 000
80 000
60000
40 000
20 000
19
•
• .J
• ** _
_• • * m Q
• ta o
•" • " 0 °
^P O rt
IT <^p
o^rf^^p^cto^-- ^--^- £
30 1940 1950 1960 1970 1980 1990 20
j« Class
' .Class
i o Class
: aClass
8{!
Figure 2-9b. Diesel truck sales (domestic) for the years 1939-1997.
Source: AMA/AAMA Motor Track Facts.
Table 2-4. Truck fleet results for 1992 from Census of
Transportation (1995), results in thousands
Truck class
Class 3 , 4. and 5
(Medium duty)
Class 6 (Light
heavy-duty)
Class 7 and 8
(Heavy heavy-
duty)
1992 trucks
1,259.0
732.0
1,966.2
1992 diesel
326.3
269.7
1725.3
% Diesels
25.9
36.8
87.8
i
11/5/99
2-27
DRAFT-DO NOT CITE OR QUOTE
-------
100.0% -
x
«
o
40.0% •
20 0% •
ggBOUUOOBOOOOO]
oaHGlB
$O^*^^
X ..
ft
"ft
6a A
^
o° * A x
OD A
a A *
A X XX
A «D
A A A A
DO A AA A v
JJ A A
A A. .A «P<
. AA "
AAAA
X
X
^^^Jwcxjfl^^XxxxSf* w *
1940 1950 1960 1970 1980 1390
O Class 8 by factory sales
D Class 8 by retail sales
ACIass 7 by retail sales
X Class 6 by retail sales
X Class 5 by re tail sales
2000 20
Year
Figure 2-10a. Diesel truck sales as a percentage of total truck sales for the years 1957-1998.
Source: Bunn(1999).
11/5/99
2-28
DRAFT-DO NOT CITE OR QUOTE
-------
100.0%
80.0%
0.0% :
» Class 51|
.Class 6l!
o Class 7 j!
•Class 8 i!
1930 1940 1950 1960 1970 1980 1990 2000
Figure 2-10b. Diesel truck sales as a percentage of total truck sales for the years
1939-1997.
Source: AMA/AAMA Motor Truck Facts.
1000
300 •
800 •
700 -
M
i
I 600 H
£
s
5 500 -
I
fe 400
Z 300 -
200 -
100
OMD
LHD
nHHD
n
1993 1992 1991
1990 1989 1988 1987 198S 1985 1984 1983 Pre-1983
Mod*iY»»r
Figure 2-11. Model year distribution of in-use truck fleet in 1992.
Source: Census of Transportation, 1995.
11/5/99
2-29 DRAFT-DO NOT CITE OR QUOTE
-------
1 vehicles were in use in 1992. For all three vehicle classes in 1992 there were a large number of
2 vehicles more than 10 years old: 54% for medium duty, 60% for light HD, and 43% for heavy
HD. For heavy HD trucks (Classes 7 and 8) there are roughly 100,000 vehicles in each model
year from 1983 to 1993. Assuming this is also true before 1983, a large number of trucks
5 (400,000) must be more than. 14 years old. This suggests a truck life of around 18 years. The
6 EPA MOBILES model assumes a vehicle life of 20 years and MOBILE6 assumes a vehicle life
7 of 25 years, with aging vehicles having fewer vehicle miles traveled on an annual basis (U.S.
8 EPA, 1999a). Motor Truck Facts and later American Automobile Association "Facts and
9 Figures" (AMA, 1927-1975), indicate that 53% of trucks from model years 1947-1956 were still
10 on the road after 14 years and 55% of trucks from model years 1960-1969 were still on the road
11 after 14 years. The proportion of trucks in use after 14 years is 63% for model years 1974-1983,
12 suggesting that the lifespan of trucks built in later years is longer.
13 In the years since 1950 to 1990 and beyond, vehicle miles traveled by all types of
14 vehicles have increased significantly. For example, Department of Transportation Federal
15 Highway Administration statistics show that passenger car vehicle miles traveled increased from
16 about 400 billion in 1951 to 1,400 billion in 1995 and 1,500 billion in 1997, an increase of about
17 360% and 380% for these years compared to 1951. Meanwhile, vehicle miles traveled by
18 combination trucks increased from about 20 billion in 1951 to 94,000 billion in 1990 and
19 124,500 billion in 1997, a somewhat larger increase of 470% and 620% for these years compared
P to 1951. These data highlight the fact that combination truck usage has increased more than
21 passenger car usage from the early 1950s to the 1990 and 1997 time frame. The Department of
22 Transportation statistics are also available for other vehicle categories such as lighter trucks.
23 . The EPA MOBILES and PARTS models calculate that about 2.6% of total vehicle miles
24 traveled in the 1950 time frame came from diesels with a gross vehicle weight over 33,000
25 pounds (Classes 7 and 8). In 1990, about 3.3% of total vehicle miles traveled came from diesel
26 trucks in these weight categories. In the 1950-1990 time frames and beyond, diesel trucks are
27 responsible for an increasing fraction of the vehicle miles traveled.
28
29 2.2.4.2. Dieselization of Railroad Locomotive Engines
30 Early in the 20th century the political and economic pressure on the railroads to replace
31 steam locomotives was substantial. Railroads were losing business to other forms of transport.
32 The diesel-electric locomotive provided 90% in-service time compared to only 50% for steam
33 locomotives, and had three times the thermal efficiency (Klein, 1991; Kirkland, 1983).
34 Additionally, several cities had passed laws barring steam locomotives within the city limits
35 because the large quantities of smoke obscured visibility, creating a safety hazard. The first
11/5/99 2-30 DRAFT-DO NOT CITE OR QUOTE
-------
1 prototype diesel locomotive was completed in 1917. By 1924 General Electric was producing a
2 standard line of switching locomotives on a production basis. Electro-Motive Corporation was
3 founded the same year to produce diesel locomotives in competition with GE. This company
4 was purchased in 1929 by General Motors and became the Electro-Motive Division. After this
5 acquisition, GM began to develop the 2-stroke engine for this application. Up to this time, all
6 locomotive diesel engines were 4-stroke. Two-strokes offered a much higher power-to-weight
7 ratio and GM's strategy was to get a large increase in power by moving to the 2-stroke cycle.
8 The first true high-speed, 2-stroke diesel-electric locomotives were produced by GM in 1935.
9 However, because of the economic climate of the Great Depression few of these were sold until
10 after the Second World War.' At the end of the war most locomotives were still steam-driven but
11 were more than 15 years old, and the railroads were ready to replace the entire locomotive fleet.
12 Few if any steam locomotives were sold after 1945 as the entire fleet was converted to diesel
13 (Coifman, 1994).
14 The locomotive fleet has included significant percentages of both 2- and 4-stroke engines.
15 The 4-stroke diesel engines were naturally aspirated in the 1940s and 1950s. It is unlikely that
16 any of the 2-stroke engines used in locomotive applications were strictly naturally aspirated.
17 Nearly all 2-stroke diesel locomotive engines are uniflow scavenged, with a positive-
18 displacement blower for scavenging assistance. In 1975, it was estimated that 75% of the
19 locomotives in service were 2-stroke, of which about one-half used one or more turbochargers in
20 addition to the existing positive-displacement blower for additional intake boost pressure.
21 Almost all of the 4-stroke locomotive engines were naturally aspirated in 1975 (Hare and
22 Springer, 1972). Electronic fuel injection for locomotive engines was first offered in the 1994
23 model year (U.S. EPA, 1998b). All locomotive engines manufactured in recent years are
24 turbocharged, aftercooled or intercooled 4-stroke engines. In part, this is because of the
25 somewhat greater durability of 4-strokes, although impending emissions regulations may have
26 also been a factor in this shift. The typical lifespan of a locomotive has been estimated to be
27 more than 40 years (U.S. EPA, 1998b). Many of the smaller railroads ore sti!! using engines
20 built in the ^'ICs, although th.3 engines may hav
-------
1 time diesel fuel consumption increased from about 400 million gallons to 26 billion gallons per
2 year in the United States, an increase by a factor of more than 75 (Figures 2-12 and 2-13).
The chemistry and properties of diesel fuel have a direct effect on engine emissions.
Researchers have studied the effect of sulfur content, total aromatic content, polyaromatic
5 content, fuel density, T90/T95, oxygenate content, and cetane. Lee et al. (1998) have
6 comprehensively reviewed literature studies of the effect of these fuel properties on regulated
7 emissions. Their conclusions were based on fleet and multiple engine tests conducted over both
8 transient and steady-state cycles, and were limited to studies in which the effects of the various
9 fuel characteristics could be decoupled from each other. Sulfur content, cetane number, density,
I
10 total aromatics and polyaromatics content, as well as boiling point distribution can have an
11 impact on emissions. It was concluded that the effect of most fuel changes on modem engines is
12 less than the effect on older, higher emitting engines.
13 Most important for emissions, the chemical makeup of diesel fuel has changed over time,
14 in.part because of new regulations. EPA currently regulates diesel fuel and requires sulfur
15 content to be less than 500 ppm for on-road applications, and that cetane index (a surrogate for
16 actual measurements of cetane number) be greater than or equal to 40, or the maximum aromatic
17 content to be 35% or less (CFR 40:80.29). California has placed additional restrictions on the
18 cetane number and aromatic content of diesel fuel (California Code of Regulations, Title 13).
19 Prior to 1993, diesel fuel sulfur levels were not federally regulated in the United States.
^P Only recommended industry practices were in place (e.g., the ASTM D 975 specified 0.5% fuel
21 sulfur limit). During the years 1960 to 1986, fuel sulfur content showed no chronological
22 increasing or decreasing trends and ranged from 0.23-0.28wt%, while the average cetane number
23 of U.S. diesel fuel declined steadily from 50.0 to 45.1, or about 0.2 per year (NIPER, 1986).
24 Based on a linear regression analysis, the average cetane number was 52.2 in 1949 and 46.8 in
25 1976. This declining trend in cetane number was likely accompanied by an increase in aromatic
26 content and density (Lee et al., 1998). The reason for the decline was that as diesel demand
27 grew, straight-run diesel became a smaller part of the pool and light-cycle oil from catalytic
28 cracking became important. Light-cycle oil is high in aromatics. One study measuring the
29 impact of changes in cetane number and aromatic content found that increasing the aromatic
30 content from 20% to 40%, with an accompanying decrease in the cetane number from 53 to 44
31 resulted in a 4% increase in NOX and a 7% increase in PM (McCarthy et al., 1992). These values
32 can be considered reasonable upper bounds for the small effect changes in fuel quality likely had
33 on NOX and PM emissions during the years 1949-1975.
34
11/5/99 2-32 DRAFT-DO NOT CITE OR QUOTE
-------
Trends in Diesel Fuel Use (1949-1995)
X 20
S
•5 18
I
Q
t
14
"
10
INK
linn inn
IMIIIIII inn
.minium inn
i mini mi
linn i mini mi
illinium i mini mi
iimiiiiiiiiinmii i mini mi
Figure 2-12. Diesel fuel use since 1949.
Source: Federal Highway Administration, 1995.
11/5/99
2-33
DRAFT-DO NOT CITE OR QUOTE
-------
On-Highway Diesel Usage, thousands of gallons
30000000 n
25000000
20000000
|
o
« 15000000
10000000
5000000
...... Illlll
Figure 2-13. On-highway diesel fuel consumption since 1949, values in thousands of
gallons.
Source: Federal Highway Administration, 1995.
11/5/99
2-34 DRAFT-DO NOT CITE OR QUOTE
-------
1 ' In wintertime, on-road No. 2 diesel may contain some percentage (-15% or more) of No.
2 1 diesel to improve cold flow properties. Blending of No. 1 may also lower the aromatic content,
3 resulting in improved emissions performance. Thus, there may also be some small but
4 perceptible seasonal changes in emissions from diesel engines.
5 A maximum allowable fuel sulfur content in the United States for on-road diesel fuel was
6 established at 0.05 mass % in 1 993 in advance of a new 0.10 g/bhp-hr PM standard for HD on-
7 highway trucks. This reduced total PM emissions through reduction of sulfate PM (primarily
8 present as sulfuric acid). Approximately 1% to 4 % of fuel sulfur is oxidized to SO3, which forms
9 sulfuric acid in the presence of water vapor in the exhaust (Wall et al., 1987; Khatri et al., 1978;
i
10 Baranescu, 1988). Considerably higher sulfuric acid PM emissions are possible with diesel
1 1 exhaust aftertreatment systems containing precious metals (oxidation catalysts, lean NOX
1 2 catalysts, catalyzed PM traps). At temperatures over 350 to 500 °C (depending on device), SO2
13 in the exhaust can be oxidized to SO3 and increase sulfuric acid PM emissions (McClure et al.,
1 4 1992; McDonald et al., 1995; Wall, 1998). Sulfur content remains at unregulated levels for off-
1 5 highway diesel fuels. Nationally, on-road fuels averaged 0.032% sulfur in 1994 while off-
1 6 highway fuels averaged 0.322% (Dickson and Sturm, 1994).
17
18 2.2.5. Chronological Assessment of Emission Factors
19 2.2.5.1. On-Road Vehicles
20 Historically, measured emissions from HD diesel vehicles have been widely variable. -
2 1 However, certain chronological trends can be identified, driven primarily by tightening
22 regulatory standards since the mid-1970s. Prior to that time, changes in average fuel
23 composition and engine technologies were implemented for reasons other than emissions control.
24 Although there is a reasonable amount of data upon which to base an emission factor for late
25 1970s and later engines, there are virtually no transient test data available to EPA on engines
26 earlier than the mid- 1 970s. Nevertheless, there are some factors that help lead to conclusions
missions of these enines. Diese! truck enine technolo chaned little in this time
28 frame, and over tbi* whole ne^od used roughly tbe same means of tuning the engine's air-fuel
29 ratio; that is, tuning was done to not permit air-fuel ratios richer than the "smoke limit" of about
30 22: 1 . This tuning, in essence, formed an "upper limit" on particulate emissions and was done
3 1 before EPA smoke standards (for customer satisfaction reasons). There is only qualitative
32 correlation between smoke and particulate emissions over the transient driving cycle, but there is
33 semiquantitative con elation between smoke and particulates over steady-state operating modes
34 (ivicGuckm and Rykowski, 1981). Tlic fact tlidt engines, turbochargcd or net, were tuned tc
35 avoid smoky operation makes it reasonable to assume that they had emissions roughly at the
1 1/5/99 2-35 DRAFT-DO NOT CITE OR QUOTE
-------
1 mid-1970s level. Other than the increased use of turbochargers, HD diesel engine technology
2 was reasonably stable. Thus, it is reasonable to conclude that the emission factors developed
4 above for mid-1970s (and later) engines adequately represent the engines in use in the 1950-1970
time frame.
5 There have been numerous studies of emissions from in-use on-road HD (greater than
6 8,500 Ibs GVWR) diesel vehicles. Emissions of regulated pollutants from these studies have
7 been reviewed (Yanowitz et al., 1999b) and the review findings, which encompass vehicles from
8 model years 1976 to 1998, are summarized below.
9 Figures 2-14 through 2-16 below show chassis dynamometer data from more than 200
10 different vehicles, reported in 20 different published studies (approximately half of which are
11 from transit buses) (Yanowitz et al., 1999a; Warner-Selph and Dietzmann, 1994; Dietzmann et
12 al., 1980; Graboski et al, 1998a; McCormick et al., 1999; Clark et al., 1997; Data et al., 1992;
13 Brown and Rideout, 1996, Brown et al., 1997; Clark et al., 1995; Dunlap et al., 1993; Ferguson
14 et al., 1992; Gautam et al., 1992; Katragadda et al., 1993; Rideout et al., 1994; Wang et al., 1993;
15 Wang et al., 1994; Williams et al., 1989; Whitfield and Harris, 1998), as well as a large amount
16 of additional data collected by West Virginia University and available on the World Wide Web
17 at www.afdc.nrel.gov. The results from vehicles tested more than once using the same test cycle,
18 and without any additional mileage accumulated between tests, are averaged and reported as one
19 data point. Emissions results from vehicles tested under different test cycles or at different points
^P in the engine's life cycle have been reported as separate data points. Note that all NOX mass
21 emissions data are reported as equivalent NO2.
22 Figures 2-14 through 2-16 show emissions trends for NOX, PM, and HC in g/mile. A
23 least-squares linear regression is plotted on each graph and yields the following equations for
24 predicting emissions trends (applicable to the years 1976-1998):
25
26 Log NOX (g/mile) = (Model year * -0.008) +16.519 (2-1)
27 Log PM (g/mile) = (Model year * -0.044) + 88.183 (2-2)
28 Log HC (g/mile) = (Model year * -0.055) + 109.390 (2-3)
29
30 As shown in Figures 2-14,2-15, and 2-16, which include 95% confidence limits and
31 regression lines, changes in NOX emissions have been relatively small, with an emission rate
32 averaging about 26 g/mile. PM, CO, and THC emissions, though widely variable within any
33 model year, have shown a pronounced declining trend. PM emissions from chassis
34 dynamometer tests decreased from an average 3.0 g/mi in 1977 to 0.47 g/mi in 1997, suggesting
35 a decrease in PM emissions of a factor of 6. Although it is clear that emissions of CO, HC, and
11/5/99 2-36 DRAFT-DO NOT CITE OR QUOTE
-------
E
O)
CO
c
O
CO
CO
E
LLI
X
O
1 00
1 0 -
1975 1980 1985 1990 19.95
M odel Year
2000
Figure 2-14. Model year trends in NOX emissions (g/mile).
11/5/99
2-3 7 DRAFT-DO NOT CITE OR QUOTE
-------
)
U)
c
o
V)
tn
E
01
1 0
1 -
0.1 -
1975 1980
1985 1990
M o d eI Year
1995 2000
Figure 2-15. Model year trends in PM emissions (g/mile).
11/5/99
2-38
DRAFT-DO NOT CITE OR QUOTE
-------
D
O)
a>
c
o
CO
CO
E
LU
O
X
1 0 -
0.1 -
1 9
75 1980 1985 1990
M odel Year
1995
2000
Figure 2-16. Model year trends in HC emissions (g/mile).
11/5/99
2-39
DRAFT-DO NOT CITE OR QUOTE
-------
1 PM have declined significantly since the early 1970s, emissions of NOX have remained
2 approximately constant.
3 Other approaches for measuring emissions from in-use, on-road diesel vehicles include
^P remote sensing and tunnel tests. The literature reports of those studies are summarized in Tables
5 2-5 and 2-6. Gram-per-mile emission factors vary substantially for the various tunnels, with NOX
6 ranging from 9 to 24 g/mile, PM ranging from 0.6 to 1.8 g/mile, CO ranging from 6 to 14 g/mile,
7 and THC ranging from 0.16 to 2.55 g/mile. Remote sensing produces results in terms of
8 pollutant emissions per unit of fuel, not on a per-mile basis. On a g/gallon of fuel consumed
9 basis, agreement between the studies for NOX emissions is reasonably good, suggesting an
10 average level for the fleet of-about 130 g/gal for both tunnel tests and remote sensing,
11 comparable to the average emissions factor generated from chassis dynamometer studies.
12 Generally, chassis dynamometer tests and engine dynamometer test results are corrected for
13 ambient humidity in accordance with the Federal Test Procedure (CFR 40, Subpart N). Tunnel
14 tests and remote sensing tests have typically not included corrections for humidity. Appropriate
15 humidity corrections for NOX and PM can be greater than 20% and 10%, respectively (or a total
16 difference of more than 45% and 20% between low- and high-humidity areas), under normally
17 occurring climatic conditions. Additionally, the remote sensing literature has not addressed how
18 to determine the correct value for the NO/NOX ratio, and there is reason to believe that this value
19 may differ systematically from site to site although, again, most of the NOX is NO.
^B There are no reported instances of HD diesel PM measurement by remote sensing, but
21 there were several tunnel tests that measured PM. In addition to the humidity correction
22 discussed above, several factors must be taken into account when comparing PM measurements
23 from tunnel tests to the chassis dynamometer measurements (Yanowitz et al., 1999b): (1) Chassis
24 dynamometer testing measures only tailpipe emissions; tunnel tests include emissions from other
25 sources (tire wear, etc). (2) Tunnel tests typically measure emissions under steady-speed freeway
26 conditions, whereas most chassis dynamometer tests are measured on cycles that are more
27 representative of stop-and-go urban driving conditions. This latter limitation also applies to
28 remote sensing readings, which measure instantaneous emissions versus emissions over a
29 representative driving cycle. PM is emitted from HD vehicles at the greatest rate during
30 accelerations.
31 Because THC emissions for diesel vehicles are very low in comparison to gasoline
32 vehicles, tunnel test results for THC have a high degree of uncertainty. A regression analysis to
33 determine the contribution of the limited number of HD vehicles to THC emissions is somewhat
34 unstable, i.e., small errors in the total measurements can change estimates. Similarly, CO
35 emissions are comparable to automobile emissions on a per-vehicle-mile basis, but since there
1175/99 2-40 DRAFT-DO NOT CITE OR QUOTE
-------
Table 2-5. Emissions results from tunnel tests (adapted from Yanowitz et al., 1999b)
IX
Test
Pierson and
Brachaczecl;, 1983
Rogaketal., 1998
Miguel etal., 1998
Weingartner et al.,
1997
Pierson eta., 1996
Pierson eta ., 1996
Pierson eta., 1996
Kirchstetter et al.,
1999
Tunnel Location
Allegheny, 1 £70-74
Allegheny, IS 75
Allegheny, IS 76
Allegheny, 1976
Tuscarora, 1976
Tuscarora, 1976
Allegheny, IS 77
Allegheny , 1979
Allegheny, 1S79
Cassiar Tunnel,
1995, Vancouver
CaldecottTurnel,
1996, San Francisco
Gubrist Tunael,
1993, Zurich
Fort McHenry
Tunnel, downhill,
1992, Baltimore
Fort McHenry
Tunnel, uphil ,
1992, Baltimore
Tuscarora Tunnel
1992, Pennsylvania
Caldecott Tur nel,
1997, San Fraacisco
Fuel
efficiency
(mi/gal)
5.42b
8.03b
5.42C
5.60C
11.46"
5.42"
6.44"
5.42C
NO,"
(g/mi)
19.50+/-
4.22
23.82 +/-
4.17
9.66 +/-
0.32
22.50 +/-
1.00
1 9.46+7-
0.85
23.82 +/-
2.98
NMHC
(g/mi)
-0.16+/-
0.88
0.92 +/-
0.21
2.55 +/-
1.05
0.68 +/-
0.20
CO
(g/mi)
6.79 +/-
11.78
6.8 +/-
1.5
14.3 +/-
5.5
6.03 +/-
1.61
PM
(g/mi)
.90-1.80
1.75+/-.19
I.5+/-.IO
1.4+/-.07
1.3+/-.19
1.39+/-26
1.3+/-.08
1.2+/-.03
1.4+/-.04
1.67+/-
0.24d
0.62 +/-
0.02f
I.43+/-
0.128
CO2
(g/mi)
1280+/-
40
897 +/-
48
1897+/-
168
1596+/-
78
NO,"
(g/gal)
157+/-
34
129+/-
23
1 1 1 +/-
4
122 +/-
5
125 +/-
5
129+/-
16
NMHC
(g/gal)
-1 +1-1
1 1 +/- 2
14+/-6
4+/-1
CO
(g/gal)
55+/-
95
78+/-
17
78+/-
30
39+/-
10
PM
(g/gal)
4.9-9.8
9.49+/-1.03
8.I+/-.54
7.6 +/-.4
7.0+/-1.0
7.5-t/-1.40
7.0 +A.43
6.5 •*•/-. 16
7.6+/-.I9
9.0 +/-
1.3d
3.5 +/-
O.lr
7.7 +/-
0.6s
K)
I
o
z
o
H
G
H
tn
o
&
o
C
O
"NOX reported as NO2.
bCalculated from observed CO2 emissions assuming fuel density 7.1 Ib/gal and C is 87% of diesel fuel by weight.
cSince CO2 omisiiions not available, fuel efficiency assumed to be the same as in slightly uphill tunnel (Fort McHenry).
dReported a; black carbon, assumed that 50% of total PM emissions are BC.
'Slope of tunnel unknown, so used average fuel efficiency for the United States.
fPM,.
gPM2,.
hUncertaint> reported as +/-1.0 standerd deviation, except where literature report did not specify standard deviation; in those cases uncertainty listed as reported.
-------
Table 2-6. Remote sensing results for hd vehicles (Yanowitz et al., 1999b)
NOX
(reported
as NO,)
CO
THC
Reference
Jimenez et al., 1998
Cohen etal., 1997
Countess et al., 1999
Bishop etal., 1996
Cohen etal., 1997
Countess et al., 1999
Bishop etal., 1996
Cohen etal., 1997
Year
study
conducted
1997
1997
1998
1992
1997
1998
1992
1997
Emissions (g/gal)
150a,b,c
108a,b.c
187a.b,c
59b
54 b
85"
0.002 HC/CO2 mole ratio d
0.00073 HC/C02 mole ratio d
"Remote sensing measures NO. The reported value was corrected to a NOX (as NO2) value by
assuming 90% (mole fraction) of NOX is NO.
•"Emissions in g/gal calculated by assuming that fuel density is 7.1 Ib/gal and C is 87% by weight
of fuel.
Tslo humidity correction factor is included.
dln order to calculate emissions in g/gal, an average molecular weight is needed.
are generally many more automobiles than HD diesels in tunnel tests, CO measurements may be
associated with a high degree of uncertainty.
3 Given the variability between testing methods, and assuming the on-road fleet measured
4 in tunnel and remote sensing tests is primarily vehicles 0 to 10 years old, there is reasonable
5 comparability between chassis dynamometer results and average tunnel and remote sensing
6 results for PM, CO, and HC. Neither tunnel testing nor remote sensing results show any
7 pronounced chronological trends for these regulated pollutants, primarily because virtually all of
8 the testing was done over a short period in the 1990s. The average model year for HD vehicles
9 was reported in only one case (Kirchstetter et al., 1999) and was found to be 1988 for a tunnel
10 test conducted in 1997. However, even in the case of tunnel testing, measurements of PM from
11 the 1970s PM levels are not significantly different from what was measured in the 1990s.
12 The data on regulated emissions do allow a comparison of emissions from 2- and 4-stroke
13 engines for PM (Figure 2-17). It is clear that there is no significant difference in PM mass
14 emissions for 2- and 4-stroke engines over the time period covered. This is true even in 1993. In
15 model year 1994 there were no on-road 2-stroke engines. Similarly, no significant difference
16 was observed for emissions of HC, CO, or NOX.
17
11/5/99
2-42
DRAFT-DO NOT CITE OR QUOTE
-------
9 •
4 -
1
C 3-
0
'5?
0)
E 2-
Ul
S
o.
•\ •
n -
• 4-Stroke Engines
o 2-Stroke Engines
o
0
o
6 r
• 8 i
, n ft
0
§
o •
8 • •
f
88!°
88 I o
p n x £
i§B8i| 1.
ra
•5*
"5)
o
'3
.2
UJ
a.
1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998
Engine Emissions Model Year
10 •
14 •
12
10
8
6 •
4 -
2
n -
• 4-Stroke Engines
o 2-Stroke Engines
O
•
o o i
^ - •
°
8 i
8 E
O Q
r
, , , , — r>— r
w w •
0 2
8»1
0 • i
g y § c
O o p f
8 •
•
5
^ •
ii-
1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998
Engine Emissions Model Year
Figure 2-17. Comparison of 2-stroke and 4-stroke engines PM emissions on a
g/mi and g/gal basis (low altitude data only).
Source: Yanowitzetal., 1999b.
11/5/99
2-43
DRAFT-DO NOT CITE OR QUOTE
-------
1 2.2.5.2. Locomotives
2 Locomotive engines generally range from 1000 horsepower up to 6000 horsepower.
Similar to the much smaller truck diesel engines, the primary pollutants of concern are NOX, PM,
CO, and HC. Unlike truck engines, most locomotive engines are not mechanically coupled to
5 the drive wheels. Because of this decoupling, locomotive engines operate in specific steady-state
6 modes rather than the continuous transient operation normal for trucks. Because the locomotive
7 engines operate only at certain speeds and torques, the measurement of emissions is considerably
8 more straightforward for locomotive engines than for truck engines. Emissions measurements
9 made during the relatively brief transition periods from one throttle position to another indicate
t
10 that transient effects are very short and thus could be neglected for the purposes of overall
11 emissions estimates.
12 Emissions measurements are made at the various possible operating modes with the
13 engine in the locomotive, and then weighting factors for typical time of operation at each throttle
14 position are applied to estimate total emissions under one or more reasonable operating
15 scenarios. In the studies included in this analysis, two scenarios were considered: line-haul
16 (movement between cities or other widely separated points) and switching (the process of
17 assembling and disassembling trains in a switchyard).
18 The Southwest Research Institute made emissions measurements for three different
« engines in locomotives in 1972 (Hare and Springer, 1972) and five more engines in locomotives
using both low- and high-sulfur fuel in 1995 (Fritz, 1995). Two engine manufacturers (the
21 Electromotive Division of General Motors, or EMD, and the General Electric Transportation
22 Systems, or GETS) tested eight different engine models and reported the results to EPA (U.S.
23 EPA, 1998b). There are also additional data. All available data on locomotives are summarized
24 in the regulatory impact assessment and shown in Figure 2-18.
25
26 2.2.6. Physical and Chemical Composition of Particles
27 Diesel PM is defined by the measurement procedures summarized in Title 40 CFR, Part
28 86, subpart N. These procedures define PM emissions as the mass of material collected on a
29 filter at a temperature of 52 °C or less after dilution of the exhaust. As the exhaust is diluted and
30 cooled, nucleation, condensation, and adsorption transform volatile material to solid and liquid
31 PM. Diesel exhaust particles are aggregates of primary spherical particles consisting of solid
32 carbonaceous material and ash, and which contain adsorbed organic and sulfur compounds
33 (sulfate) combined with other condensed material. The organic material includes unburned fuel,
34 lube oil, and partial combustion and pyrolysis products. This is frequently quantified as the
35 soluble organic fraction, or SOF. The SOF can range from less than 10% to more than 90% by
1175/99 2-44 DRAFT-DO NOT CITE OR QUOTE
-------
20
15
0 10
5
0
C
Line-Haul Cycle Emissions Data
NOx and PM (g/bhp-hr)
o
— o ° qp
^op c. o o
I I I I
0.1 0.2 0.3 0.4 0.
PM
5
40
30
0 20
10
0
C
Switch Cycle Emissions Data
NOx and PM (g/bhp-hr)
0
—
o °
o
fl&8$
O ^^ ^^ O
i i i i
) 0.2 0.4 0.6 0.8
1
PM
Figure 2-18. Line-haul and switch emissions data.
Source: U.S. EPA, 1998a.
11/5/99
2-45 DRAFT-DO NOT CITE OR QUOTE
-------
1 mass, with the highest values occurring at light engine load where exhaust temperature is low
2 (Kittelson, 1998). The SOF fraction can also vary with engine design, with high lube oil
3 emitting older engines producing higher SOF. Sulfate depends on fuel sulfur content primarily.
Carbonaceous diesel particulate matter has a high specific surface area (30-50 m2/g) (Frey
5 and Com, 1967). Because of this high surface area diesel particles are able to adsorb large
6 quantities of organic materials. After removal of the organic material by extraction, the surface
7 area increases to as high as 90 m2/g (Pierson and Brachaczek, 1976). A variety of solvents have
8 been used to extract the SOF (Levson, 1988). Soxhlet extraction with a binary solvent consisting
9 of an aromatic and an alcohol appears to give the best recovery of PAHs, although
10 dichloromethane is also used. Some studies have then used liquid chromatography to separate
11 the extract into various fractions on the basis of chemical composition and polarity.
12 The distribution of emissions between the gas and particle phases in diesel exhaust
13 (gas/particle partitioning) is determined by the vapor pressure of the individual species, the type
14 and amount of particulate matter present (surface area available for adsorption), and the
15 temperature (Zielinska et al., 1998). Two-ring and smaller compounds exist primarily in the gas
16 phase while five-ring and larger compounds are completely adsorbed on the particles. Three-
17 and four-ring compounds are distributed between the two phases. Some studies use sampling
18 trains designed to collect both gas-phase and particle-phase compounds, while others simply
19 report the amount or emission of a given compound in the PM SOF. During the collection of
fe particulate and organic compounds, filter adsorption, blow-off (loss from the filter), and
21 chemical transformation of the semivolatile compounds have been reported to occur (Schauer et
22 al., 1999; Cantrell et al., 1986; Feilberg et al., 1999; Cautreels and Cauwenberghe, 1978).
23 For diesel engine emissions, approximately 57% of the extracted organic mass is
24 contained in the nonpolar fraction (Schuetzle, 1983). About 90% of this fraction consists of
25 aliphatic hydrocarbons from approximately C14 to about C40 (Black and High, 1979; Pierson et
26 al., 1983). Polycyclic aromatic hydrocarbons and alkyl-substituted PAHs account for the
27 remainder of the nonpolar mass. The moderately polar fraction (-9% w/w of extract) consists
28 mainly of oxygenated PAHs species, substituted benzaldehydes, and nitrated PAHs. The polar
29 fraction (-32% w/w of extract) is composed mainly of n-alkanoic acids, carboxylic and
30 dicarboxylic acids of PAHs, hydroxy-PAHs, hydroxynitro-PAHs, nitrated N-containing
31 heterocyclic compounds, etc. (Schuetzle, 1983; Schuetzle et al., 1985).
32 Rogge et al. (1993) reported the composition of the extractable portion of fine particulate
33 matter emitted from two HD diesel trucks (1987 model year). No HPLC separation step was
34 employed and the extract (hexane followed by benzene/2-propanol) was analyzed by capillary
35 gas chromatography/mass spectrometry (GC/MS) before and after derivatization. The
unresolved organic mass, which comprises 90% of the elutable organic mass, consists mainly of
11/5/99 2-46 DRAFT-DO NOT CITE OR QUOTE
-------
1 branched and cyclic hydrocarbons. From the mass fraction that is resolved as discrete peaks by
2 GC/MS, -42% was identified as specific organic compounds. Most of the identified resolved
3 organic mass (-60%) consists of n-alkane, followed by n-alkanoic acids (-20%). PAHs account
4 for -3.5% and oxy-PAHs (ketones and quinones) for another -3.3%. Taking into account the
5 differences in the analytical procedures and the percentage of identified peaks, this distribution is
6 roughly similar to those reported by Schuetzle (1983).
7
8 2.2.6.1. SOF and Elemental Carbon Content of Particles
9 Chassis dynamometer results indicate that SOF emissions have trended downward over
10 the years as engine manufacturers have tried to reduce oil consumption. This is shown in Figure
11 2-19, where the trend can be seen as both reduction in SOF weight percent and in SOF g/mi
12 emissions. The downward trend is driven primarily by the need to reduce oil consumption, and
13 thereby reduce engine wear and maintenance costs, as well as the need to meet PM emission
T4 standards. There is no significant difference in SOF emissions from 2- and 4-stroke engines in
15 later years, while 2-stroke engines in the 1970s tended to emit greater amounts of SOF compared
16 with typical 4-stroke engines. The downward trend in SOF as a percentage indicates that the
17 solid carbonaceous material as a percentage of PM has been increasing. Figure 2-20 shows the
18 PM solids g/mile emissions (TPM-SOF) for several vehicles tested on chassis dynamometers. A
19 decreasing trend in PM solids is also evident, consistent with the observed decline in total PM
20 emissions. It is tempting to assume that this solid carbonaceous portion is approximately the
21 same as elemental carbon (EC, a quantity not commonly measured in HD studies). This
22 assumption is validated, to a good approximation, by a study in which both SOF and EC were
23 measured (SOF measured by Yanowitz and co-workers [1999a] and EC measured on the same
24 samples by Zielinska and co-workers [1998]); a parity plot of these results is shown in Figure 2-
25 21. The data reported by Zielinska and co-workers currently appear to be the only measurements
26 of EC for in-use vehicles. These data are for 4-stroke engines. EC ranged from 31% to 84% of
27 PM. and averaged 63%. It is apparent from Figure 2-20 that g/mile EC emissions have declined
28 with model year.
29 Engine testing studies show SOF percentage to be highly variable, ranging from 20% to
30 60%, and exhibiting a declining trend with model year (McCarthy et al., 1992; Springer, 1979;
31 Johnson et al., 1994; Bagley et al., 1998; Tanaka et al., 1998; Rantanen et al., 1993; Mitchell et
32 al., 1994; Hansen et ai., 1994). On a g/bhp-h basis SOF emissions declined significantly in the
33 early 1990s and are typically in the range of 0.02-0.05 g/bhp-h. Engine dynamometer data
34 provide confrrm"tinn that total SOF and PM solids (or EC) emissions have declined. Note that
35 there are many more engine testing studies available, which this document does not attempt to
36 comprehensively review.
1175/99 2-47 DRAFT-DO NOT CITE OR QUOTE
-------
70 -
60-
50-
1 «-
LL 30-
20-
10-
0-
7
20 -|
1.5-
* 1.0-
0.5-
0.0-
7
9 • 4-strote
0 2-strake
°8 .
.°
• •
• °
. 1
• • • . • • •
0 • 1 • * • •
•8 ..-•«.
5 80 85 90 95
Model Year
• 4-stroke
0 • 0 2-s*c*e
O
O
0 •
• * 0
•o
° :°:-. .
'!•• o .
•
• •
5 80 85 . 90 95
Made! Year
Figure 2-19. Comparison of SOF emissions for 2- and 4-stroke engines in
g/mi and as a percentage of total PM.
Sources: Warner-Selph et al., 1984; Dietzmann et al., 1980; Graboski et al.,
1998b.
11/5/99
2-48
DRAFT-DO NOT CITE OR QUOTE
-------
10
o
CO
2 -
• 4-stroke
O 2-stroke
• 8 •
o • o
o o •
I •
75
80
85 90
Model Year
95
Figure 2-20. Trends in PM solids emissions with model year, a reasonable
surrogate for elemental carbon content.
Sources: Yanowitz et al., 1999a; Warner-Selph and Dietzmann, 1984; Dietzmann
et al., 1980; Rcgge et al., 1993.
11/5/99
2-49
DRAFT-DO NOT CITE OR QUOTE
-------
3.0
2.5 -
U? 2.0 -
O
0. 1.5 -
0.5 -
0.0
0.0 0.5 1.0 1.5 2.0 2.5
Elemental Carbon, g/mi
3.0
Figure 2-21. Parity plot showing approximate agreement between PM
elemental carbon and PM solids measurements in g/mi.
Sources: Yanowitz et al., 1999a; Zielinska et al., 1998.
11/5/99
2-50
DRAFT-DO NOT CITE OR QUOTE
-------
1 2.2.6.2. PAHs and Nitro-PAH Emissions
2 PAHs, nitro-PAHs, and oxidized derivatives of these compounds have attracted
3 considerable attention because of their known mutagenic and, in some cases, carcinogenic
4 character. Nitrated polycyclic aromatic compounds have caused lung cancer and remote
5 metastases in laboratory animals (HEI, 1995). 1 -Nitropyrene has been speculated to be the major
6 source of mutagenicity in diesel soot (Noorkhoek and Bos, 1995); however, a large number of
7 other nitro-PAHs are present (Paputa-Peck et al., 1983) and other studies suggest that it is the
8 oxygenated hitro-PAH species that are responsible (Schuetzle et al., 1981; Schuetzle et al., 1985;
9 Ciccioli et al., 1986). For example, Grosovsky and co-workers (1999) have shown that 2-
10 nitrodibenzopyranone (2NDBP) is highly mutagenic in human cells. 3-Nitrobenzanthrone has
11 been shown to be one of the most potent bacterial mutagens known (Enya et al., 1997). 3-
12 Nitrobenzanthrone is also known to be a component of diesel PM, while 2NDBP is proposed to
13 be an atmospheric transformation product of phenanthrene, but may also be present in diesel
14 exhaust.
15 A few engine and chassis studies have measured PAH emissions. Dietzmann and co-
16 workers (1980) examined four vehicles equipped with late 1970s turbocharged DI engines.
17 Emissions of benzo(a)pyrene from particle extracts were reported and ranged from 1.5 to 9
18 ug/mi. No gas-phase PAH measurements were reported. No correlation with engine technology
19 (one of the engines was 2-stroke) was observed. Table 2-7 gives the approximate concentrations
20 of several of the abundant nitro-PAHs quantified in early 1980s LD-diesel particulate extracts
21 (with the exception of 3-nitrobenzanthrone, reported recently), in ng/g of particles.
22 Concentrations for some of the nitro-PAHs identified range from 0.3 ug/g for 1,3-dinitropyrene
23 to 8.6 ug/g for 2,7-dinitro-9-fluorenone and 75 u^g for 1-nitropyrene. More recent nitro-PAH
24 and PAH data for HD diesel engines are reported in units of g/hgp-hr or mass/volume of
25 exhaust,making it impossible to compare them to the older data (Norbeck et al., 1998; Bagley et
26 al., 1996, 1998; Baumgard and Johnson, 1992; Opris et al., 1993; Hansen et al., 1994; Harvey et
27 al., 1994; Kantola et aL 1992; Kreso et al., 1998b; McClure et al., 1992; Pataky et al., 1994).
28 Rogge and co-workers (1993) reported PAH emissions from particle extracts for two
29 1987 model year trucks (averaged together, 4-stroke and turbocharged engines). They report
30 results for many specific PAH compounds with total PAHs summing to 0.43 mg/mi and
31 benzo(a)pyrene emissions of 2.7 |ig/mi. Particle-phase PAH was about 0.5% of total PM mass.
32 Schauer and co-workers (1999) have recently reported gas- and particle-phase PAH emissions for
33 a 1995 medium-duty turbocharged and intercooled truck. They also report results for a large
34 number of individual PAHs, but summed emissions were 6.9 mg/mi (gas phase) and 1.9 mg/mi
35 (particle phase). Particle-phase PAHs were about 0.7% of total PM mass. Emissions of
36 benzo(a)pyrene were not reported, but emissions of individual species of similar molecular
1175/99 2-51 DRAFT-DO NOT CITE OR QUOTE
-------
Table 2-7. Concentrations of nitro-polycyclic aromatic
hydrocarbons identified in a LD diesel participate
extract
Nitro-PAH'
4-nitrobiphenyl
2-nitrofluorene
2-nitroanthracene
9-nitroanthracene
9-nitrophenanthrene
3 -n itrophenanthrene
2-methyl-l-nitroanthracene
1 -nitrofluoranthene
7-n itrofluoranthene
3 -nitrofluoranthene
8-nitrofluoranthene
1-nitropyrene
6-nitrobenzo[a]pyrene
1 ,3-dinitropyreneb
l,6-dinitropyreneb
1 ,8-dinitropyreneb
2,7-dinitrofluorenec
2,7-dinitro-9-fluorenonec
3 - n i trobenzanthroned
Concentration
(Hg/g of particles)
2.2
-1.8
4.4
1.2
1.0
4.1
8.3
1.8
0.7
4.4
0.8
18.9;75b
2.5
0.30
0.40
0.53
4.2; 6.0
8.6; 3.0
0.6 to 6.6
"From Campbell and Lee (1984) unless noted otherwise. Concentrations recalculated from ug/g of
extract to ug/g of particles using a value of 44% for extractable material (w/w).
bFrom Paputa-Peck et al, 1983.
'From Schuetzle, 1983.
dFrom Enya et al., 1997 (Isuzu Model 6HEL 7127cc).
11/5/99
2-52
DRAFT-DO NOT CITE OR QUOTE
-------
1 weight were approximately 10 Jig/mi. Measurements of particle- and gas-phase PAHs conducted
2 for the Northern Front Range Air Quality Study (Zielinska et al., 1998) found the benzo(a)pyrene
3 emission rate to average 13 ug/mi for 15 vehicles ranging from 1983-1993 model years.
4 Summing of individual PAH emissions from this study yields a total PAH rate (combined gas
5 and particle phase) of 13.5 mg/mi.
6 Benzo(a)pyrene emissions were also reported in the engine dynamometer studies of
7 Springer (1979). A comparison of turbocharged and naturally aspirated engines (both about 1
8 p.g/bhp-h), and of DI and IDI engines (both about 0.15 ug/bhp-h) showed no significant effect of
9 these technology changes on emissions of this compound, as shown in Table 2-8. The difference
10 between 1 and 0.15 ug/bhp-h cannot be attributed to specific technology changes. The engines
11 were from different manufacturers. Bagley and co-workers (1998) studied a 1983 model year
12 IDI, naturally aspirated engine and observed emission levels listed in Table 2-8 from particulate
13 matter extracts. These results can be compared to data presented by Mitchell and co-workers
14 (1994) for two DI turbocharged engines. It is likely that there are also other differences in
15 technology between the 1983 and 1991 engines; however, it is clear that total PM emissions are
16 substantially lower for the newer engines. Results for SOF are inconclusive. Total PAH
17 emissions are the same for the 1983 and 1991 engines and range from 0.05% to 0.15% of the
18 total PM mass. 1-Nitropyrene emissions are near to detection limits and thus the apparent
19 differences are probably not significant.
20 On the basis of these limited data it is difficult to draw a precise, quantitative conclusion
21 regarding how PAH emissions have changed over time. However, it seems likely that total PAH
22 emissions have been in the range from 0.1 to 15 mg/mi from the early 1970s to the early 1990s.
23 PM-associated PAHs make up on the order of 0.1% of PM mass. Emissions of benzo(a)pyrene
24 were on the order of 10 ug/mi. It is also highly likely that PAH emissions have declined in
25 parallel with emissions of total PM and SOF, which have declined by a factor of approximately 6
26 over this time period. There is no evidence for a change in PAH emissions out of proportion to
27 the observed changes in mass emissions of PM or SOF.
28 One chassis study has reported emissions of 1-nitropyrene in PM extracts from 17 HD
29 diesel vehicles (Warner-Selph and Dietzmann, 1984). All engines were turbocharged and direct
30 injected. Results are shown in Figure 2-22; unfortunately there are no distinct trends in these
31 data and the data do not extend to the period of strict emission regulations in the late 1980s. The
32 results suggest, however, that the introduction of new technologies, which did occur to some
33 extent over the model years covered, has not produced dramatic changes in emissions of 1-
34 nitropyrene. Again, it seems likely that nitro-PAH emissions have declined in parallel with
35 decreasing emissions of total PM and SOF.
11/5/99 2-53 DRAFT-DO NOT CITE OR QUOTE
-------
Table 2-8. Comparison of PAH and nitro-PAH emissions for IDI naturally aspirated engines and two DI turbocEarged
engines
Emissions,
(ig/bhp-h
PM (g/bhp-h)
SOF (%)
PAH
Benzopyrene
Nitro-PAH
1-Nitro-
pyrene
1977 Mack
ETAY(B)673A
DI, turbo-
charged,
aftercooled
0.61
16
-
2.23
--
--
1977
Caterpillar
3206, DI,
turbocharged,
aftercooled
.35
18
--
0.15
-
~
1977
Caterpillar
3206, IDI,
turbocharged,
aftercooled
0.28
11
--
0.10
--
-
1977
Daimler-Benz
OM-352A,
DI, turbo-
charged,
aftercooled
0.56
34
--
0.87
--
-
1977
Daimler-
Benz OM-
352A, DI,
naturally
aspirated
0.99
29
-
1.07
--
-
1983
Caterpillar
3304 IDI,
naturally
aspirated
0.56
57
132.5
1.5
-
2.2
1991 DDC
Series 60 DI,
turbocharged,
aftercooled
0.12
26
131
-
0.18
0.06
1991 Navistar
DTA466, DI,
turbocharged,
aftercooled
0.1
55
145
0.05
0.65
0.32
to
s
o
z;
o
H
O
H-4
a
o
c!
Sources: Springer, 1979; Bagley et a!., 1998; Mitchell et al., 1994.
w
-------
'E
E
2
o
S
C/5
c
g
-------
1 2.2.6.3. Aldehyde Emissions
2 Many aldehydes of interest typically occur in the gas phase rather than the particle phase
3 of diesel exhaust. Some aldehydes are also known carcinogens and there are considerable data
on aldehyde emissions from diesel engines. Figure 2-23 reports mg/mile total aldehyde
5 emissions from chassis dynamometer studies (Warner-Selph and Dietzmann, 1984; Schauer et
6 al., 1999; Unnasch et al., 1993). The results indicate no difference between 2 and 4-stroke
7 engines, although aldehyde emissions appear to have declined substantially since 1980 on the
8 basis of a limited number of data points (only 2). Engine dynamometer studies show aldehyde
9 emission levels of 150-300 mg/bhp-h for late 1970s engines with no significant effect of
10 turbocharging, or IDI versus DI. High-pressure fuel injection may have resulted in a marginal
11 increase in aldehyde emissions (Springer, 1979). By comparison, 1991 model year engines (DI,
12 turbocharged) exhibited aldehyde emissions in the 30-50 mg/bhp-h range (Mitchell et al., 1994).
13 It seems likely that aldehyde emissions have declined by perhaps one order of magnitude since
14 about 1980, on average, in line with the decline in total PM and SOF emissions. Insufficient
15 information is available to determine the cause of this decline; however, more complete
16 combustion because of higher pressure fuel injection coupled with leaner operation because of
17 turbocharging with aftercooling is the most likely cause.
18
19 2.2.6.4. Dioxin and Fur an Emissions
^^ Dioxin and furan emissions from on-road HD diesel vehicles were measured in the Fort
21 McHenry Tunnel in Baltimore, MD (Gertler et al., 1998). For the limited range of vehicle
22 operating conditions in the tunnel, the average HD diesel emission factor was 0.28±0.13 ng-
23 TEQ/mi. This is a factor of 3 lower than the initial EPA estimate (Gertler et al., 1998). The
24 recent dynamometer measurements from a HD diesel engine (Cummins L10) showed negligible
25 dioxin and furan emission rates (Norbeck et al., 1998).
26
27 2.2.6.5. Particle Size
28 Figure 2-24 shows the general size distribution for diesel particulate based on mass and
29 particle number. Most of the mass is accumulation mode particles ranging in size from 50 to 700
30 nm and averaging about 200 run. Aggregated carbonaceous particles and absorbed organic
31 material are primarily in this mode. The nuclei mode consists of particles in the 5-50 nm range,
32 averaging about 20 nm. These are believed to form from exhaust constituents during cooling and
33 to consist of sulfuric acid droplets, ash particles, condensed organic material, and perhaps
34 primary carbon spherules (Abdul-Khalek et al., 1998; Baumgard and Johnson, 1996). The nuclei
35 mode typically contains from 1%-20% of particle mass and from 50%-90% of the particle
number.
11/5/99 2-56 DRAFT-DO NOT CITE OR QUOTE
-------
£
1000 -
800-
jg 600-
S
•s
T3
400 -
200 -
O
o 0
8°
o
• •
o o
75
80
85 90
Model Year
O 4-stroke
• 2-stroke
95
Figure 2-23. Chassis dynamometer measurements of total aldehyde
emissions from HD diesel vehicles.
Sources: Warner-Selph and Dietzmann, 1984; Schauer et al., 1999; Unnasch et
al., 1993.
11/5/99
2-57
DRAFT-DO NOT CITE OR QUOTE
-------
0.18
0.001
Nanoparticles
Dp < 50 nm
Fine Particles
Dp < 2.5 nm
H
0.010
Ultrafine Particles
Dp < 100 nm
/ Accumulation
Mode
PM10
Dp < 10 n
Coarse
Mode
0.100 1.000
Dlamattr (|im)
•Mass Weighting Number Weighting
10.000
Figure 2-24. Particle size distribution in diesel exhaust, taken from Kittelson
(1998).
1 Measurements made on diluted diesel exhaust typically show higher numbers of nuclei-
2 mode particles than do measurements made on raw exhaust because of condensation to form
3 nuclei mode aerosol upon cooling of the exhaust. Dilution ratio, sampling temperature, and other
4 sampling factors can therefore have a large impact on the number and makeup of nuclei mode
5 particles (Abdul-Khalek et al., 1999). Just as diesel exhaust particulate matter is defined by how
6 it is collected (i.e., on a filter at or below 52 °C); the size distribution of diesel exhaust is also
7 determined by how is it measured. Baumgardand Johnson (1996) have proposed that
8 accumulation-mode particles are formed in the combustion chamber whereas nuclei-mode
9 particles are formed during the dilution and measurement process. It seems likely that the .
10 situation is not nearly so clear-cut and that both accumulation and nuclei-mode particles are
11 formed in the combustion chamber, but that a large number of additional nuclei-mode particles
12 are formed during dilution.
13 Several groups have shown that decreasing sulfur content decreases the number of nuclei-
14 mode particles measured in the exhaust, assuming temperature is low enough and residence time
15 is long enough for nucleation and condensation of sulfate aerosol and water (Baumgard and
Fohnson, 1992; Opris et al., 1993; Baumgard and Johnson, 1996; Abdul-Khalek et al., 1999).
11/5/99 2-58 DRAFT-DO NOT CITE OR QUOTE
-------
1 The application of this finding to real-world conditions is difficult to predict, as the number of
2 nuclei-mode particles formed from sulfate and water in the atmosphere will be determined by
3 atmospheric conditions, not by dilution tunnel conditions.
4 More controversial is the suggestion that PM emission size distribution from newer
5 technology engines (1991 and later) may be shifted to have a much higher number concentration
6 of nuclei-mode particles, independent of fuel sulfur content (Kreso et al., 1998b; Abdul-Khalek
7 et al., 1998; Baumgard and Johnson, 1996; Bagley et al., 1996). For example,'in the study of
8 Kreso and coworkers (1998b), a comparison of emissions from a 1995 model year engine
9 measured in that work with measurements made on 1991 (Bagley et al., 1996) and 1988 (Bagley
10 et al., 1993) model year engines in earlier studies is presented. Dilution conditions (relatively
11 low temperature, low primary dilution ratio, long residence time of more than 3 seconds)
12 strongly favor the formation of nucleation products. The 1991 and 1988 engines were tested
13 with 100 ppm sulfur fuel while the 1995 engine was tested with 310 ppm sulfur fuel, which may
14 confound the results to some extent. Nuclei-mode particles made up 40% to 60% of the-number
15 fraction of PM emissions for the 198 8 engine and 97%+ of the PM from the 1991 and 1995
16 engines. Number concentrations were also roughly two orders of magnitude higher for the newer
17 engines. SOF made up 25%-30% of PM in the 1988 engine and 40%-80% of PM for the newer
18 engines. Total PM was significantly reduced for the newer engines. It was suggested that
19 increased fuel injection pressure leads to improved fuel atomization and evaporation, leading to
20 smaller primary carbonaceous particles, but there appears to be no more direct experimental or
21 computational evidence supporting this hypothesis. The high degree of SOF with the 1991
22 engine, particularly at high-load test modes, was also inconsistent with measured SOF values of
23 other engines using similar types of technology (Last et al., 1995, Ullman et al., 1995). Kittelson
24 (1998) notes that there is far less soot-type particulate and that higher number concentrations of
25 the small particles are formed from nucleation of VOC and sulfuric acid-type compounds.
26 At present, no conclusions can be made regarding the reported shift in size distribution
27 because:
28
29 • The result may simply be a sampling artifact because of the substantial effect of
30 dilution conditions on measured particle size distributions. The results may also be
31 a sampling artifact because no study has reported back-to-back testing of engines
32 with varying technology. All results are based on a comparison of results from
33 individual studies performed over several years. Only recently has the impact of
34 sampling conditions begun to be understood, and thus early results, and results that
35 are not clearly obtained under identical sampling conditions, may lead to erroneous
36 conclusions. Extensive research is underway to understand the factors in the
11/5/99 2-59 DRAFT-DO NOT CITE OR QUOTE
-------
1 sampling procedure that affect the PM size distribution and to determine the actual
2 size distribution (Kittelson et al., 1999).
3
• The result may not be relevant to the fate of engine exhaust in the atmosphere. It is
5 unclear what sampling conditions are appropriate for these studies. If
6 understanding the fate of the exhaust in the atmosphere is the intended use of the
7 study, then simulation of atmospheric dilution conditions is desirable. On the other
8 hand, some studies may be more interested in the impact of engine technology on
9 engine emissions with no sampling artifacts. Given an understanding of the
10 atmospheric chemistry and physics, engine emissions might also be used to predict
11 the formation of aerosol in the atmosphere. An understanding of these factors can
12 only come through knowledge of the chemical composition of nuclei-mode particles.
13 and through studies of the fate of diesel exhaust in the atmosphere (an alarmingly
14 complex situation).
15
16 • Particle sizing studies have been performed under steady-state conditions that are
17 probably not representative of how nearly all diesel particulate is actually formed in
18 use. Engine transients create temporary situations that favor PM production, and in
19 all likelihood more than 90% of diesel PM in use is generated under these
conditions.
21
22 2.3. ATMOSPHERIC TRANSFORMATION OF DIESEL EXHAUST
23 Primary diesel emissions are a complex mixture containing hundreds of organic and
24 inorganic constituents in the gas and particle phases, the most abundant of which are listed in
25 Table 2-9. The more reactive compounds with short atmospheric lifetimes will undergo rapid
26 transformation in the presence of the appropriate reactants, whereas more stable pollutants can be
27 transported over greater distances. A knowledge of the atmospheric transformations of gaseous
28 and particulate components of diesel emissions and their fate is important in assessing
29 environmental exposures and risks. This section describes some of the major atmospheric
30 transformation processes for gas-phase and particle-phase diesel exhaust, focusing on the
31 primary and secondary organic compounds that are of significance for human health. For a more
32 comprehensive summary of the atmospheric transport and transformation of diesel emissions see
33 Winer and Busby (1995).
34
1175/99 2-60 DRAFT-DO NOT CITE OR QUOTE
-------
Table 2-9. Classes of compounds in diesel exhaust
Participate phase
Gas phase
Heterocyclics, hydrocarbons (C]4-C}i), and
PAHs and derivatives:
Acids
Alcohols
Alkanoic acids
n-Alkanes
Anhydrides
Aromatic acids
Cycloalkanes
Esters
Halogenated cmpds.
Ketones
Nitrated cmpds.
Sulfonates
Quinones
Elemental carbon
Inorganic sulfates and nitrates
Metals
Water
Heterocyclics, hydrocarbons (C,-C,0), and
derivatives:
Acids Cycloalkanes, Cycloakenes
Aldehydes Dicarbonyls
Alkanoic acids Ethyne
n-Alkanes Halogenated cmpds.
n-Alkenes Ketones
Anhydrides Nitrated cmpds.
Aromatic acids Sulfonates
Quinones
Acrolein
Ammonia
Carbon dioxide, carbon monoxide
Benzene
1,3-Butadiene
Formaldehyde
Formic acid
Hydrogen cyanide, hydrogen sulfide
Methane, methanol
Nitric and nitrous acids
Nitrogen oxides, nitrous oxide
Sulfur dioxide
Toluene
Water
Source: Mauderly (1992), which summarized the work of Lies et al., 1986; Schuetzle and Frazier,
1986; Carey 1987; Zaebst et al., 1988; updated from recent work by Johnson, 1993; McDonald,
1997; Schauer et al., 1999.
1 2.3.1. Gas-Phase Diesel Exhaust
2 Gas-phase diesel exhaust contains of several organic and inorganic compounds which
3 undergo various chemical and physical transformations in the atmosphere depending on the
4 abundance of reactants and meteorological factors such as wind speed and direction, solar
5 irradiance, humidity, temperature, and precipitation. Gaseous diesel exhaust will primarily react
6 with the following species (Atkinson, 1988):
7
8 • Sunlight, during daylight hours;
9 • Hydroxyl radical (OH), during daylight hours;
10 • Ozone (O3), during daytime and nighttime;
11 • Hydroperoxyl radical HO2, typically during afternoon/evening hours;
12 • Gaseous nitrate radicals (NO3) or dinitrogen pentoxide (N2O5), during nighttime
13 hours; and
14 • Gaseous nitric acid (HNO3) and other species such as nitrous acid (MONO) and
15 sulfuric acid (H2SO4).
11/5/99
2-61
DRAFT-DO NOT CITE OR QUOTE
-------
1 The major loss process for most of the diesel exhaust emission constituents is oxidation,
2 which occurs primarily by daytime reaction with OH radical (Table 2-10). For some pollutants,
photolysis, reaction with ozone, and reactions with NO3 radicals during nighttime hours are also
important removal processes. The atmospheric lifetimes do not take into consideration the
5 potential chemical or biological importance of the products of these various reactions. For
6 example, the reaction of gas-phase PAHs with NO3 appears to be of minor significance as a PAH
7 loss process, but is more important as a route of formation of mutagenic nitro-PAHs. The
8 reaction products for some of the major gaseous diesel exhaust compounds are listed in Table 2-
9 11 and are discussed briefly below.
10
11 2.3.1.1. Organic Compounds
12 The organic fraction of diesel exhaust is a complex mixture of compounds, very few of
13 which have been characterized. The atmospheric chemistry of several organic constituents of
14 diesel exhaust (which are also produced by other combustion sources) has been studied. A few
15 of these reactions and their products are discussed below. For a complete summary of the
16 atmospheric chemistry of organic combustion products, see Seinfeld and Pandis (1998).
17 Acetaldehyde forms peroxyacetyl nitrate (PAN), which has been shown to be a direct-
18 acting mutagen toward S. typhimurium strain TA100 (Kleindienst et al., 1985) and is phytotoxic.
19 Benzaldehyde, the simplest aromatic aldehyde, forms peroxybenzoyl nitrate or nitrophenols
^P' following reaction with oxides of nitrogen (Table 2-11).
21 For those PAHs present in the gas phase, reaction with the hydroxyl radical is the major
22 removal route, leading to atmospheric lifetimes of a few hours. The gas-phase reaction of PAHs
23 containing a cyclopenta-fused ring, such as acenaphthene, acenaphthylene, and
24 acephenanthrylene with the nitrate radical may be an important loss process during nighttime
25 hours. Relatively few data are available concerning the products of these gas-phase reactions. It
26 has been shown that, in the presence of NOX, the OH radical reactions with naphthalene, 1- and
27 2-methylnaphthalene, acenaphthylene, biphenyl, fluoranthene, pyrene, and acephenanthrylene
28 lead to the formation of nitroarenes (Arey et al., 1986; Atkinson et al., 1986; 1990; Zielinska et
29 al., 1988; 1989a; Arey, 1998). In addition, in a 2-step process involving OH radical reaction and
30 NO2 addition, 2-nitrofluoranthene and 2-nitropyrene can be formed and eventually partition to
31 the particle phase, as will other nitro-PAHs.
32 The addition of the N03 radical to the PAH aromatic ring leads to nitroarene formation
33 (Sweetman et al., 1986; Atkinson et al., 1987,1990; Zielinska et al., 1989a). The gas-phase
34 reactions of NO3 radical with naphthalene, 1- and 2-methylnaphthalene, acenaphthene,
35 phenanthrene, anthracene, fluoranthene, and pyrene produce, in general, the same nitro-PAH
isomers as the OH radical reaction, but with different yields (Arey et al., 1989; Sweetman et al.,
1175/99 2-62 DRAFT-DO NOT CITE OR QUOTE
-------
Table 2-10. Calculated atmospheric lifetimes for gas-phase reactions of
selected compounds present in automotive emissions with important reactive
species
Compound
NO,
NO
HNO3
SO,
NH3
Propane
n-Butane
n-Octane
Ethylene
Propylene
Acetylene
Formaldehyde
Acetaldehyde
Benzaldehyde
Acrolein
Formic acid
Benzene
Toluene
m-Xylene
Phenol
Naphthalene f
2-Methylnaphthalene f
1-Nitronaphthalene f
Acenaphthene f
Acenaphthylene f
Phenanthrene f
Anthracene f
Fluoranthene f
Pyrene f
Atmospheric lifetime resulting from reaction with:
OHa
1 .3 days
2.5 days
110 days
16 days
90 days
12 days
5.6 days
1.9 days
1 .9 days
7h
19 days
1.9 days
0.6 day
1.2 days
0.6 day
31 days
1 1 days
2.5 days
7h
6h
6.8 h
2.8 h
2.3 days
1.5 h
1.3 h
11.2h
8.6 h
-2.9 h
-2.9 h
200 years
-
>7000 years
>4500 years
-
9 days
1 .5 days
6 years
>2- 104 years
>7 years
-
60 days
-
600 years
300 years
75 years
-
>80 days
>40 days
>28 days
>30 days
-43 min
41 days
-
-
-
NCV
24 min
1.2 min
-
> 1.4 xlO4 years
-
-
3.6 years
1.2 years
1 .2 years
6 days
>5.6 years
84 days
20 days
24 days
-
-
>6.4 years
3.6 years
0.8 years
8 min
1 .5 years
1 80 days
1 8 years
1.2 h
6 min
4.6 h
-
-1 year
- 120 days
H02d
2h
20 min
-
>600 years
-
-
-
-
-
-
-
23 days
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
hvl
2 min
-
-
-
-
-
-
-
-
-
-
4h
60 h
-
-
-
-
-
-
-
-
-
1.7 h
-
-
-
-
-
-
1 For i2-h average concentration of OH radical of 1.6*!06 molecule/cm3 (Prinn et a!., 1992)-
k For 24-h average G3 iuuCcuu~a.tion cf 7- 1C" mc!cculs/e:r.3.
c For 12-h average NO3 concentration of 5x 10* molecule/cm3 (Atkinson, 1991).
d For 12-h average HO2 concentration of 108 molecule/cm3.
e For solar zenith angle of 0°.
f Lifetimes from Arey (1998), for 12-h concentration of OH radical of 1.9x10s molecule/cm3.
Source: Winer and Busby (1995) unless noted otherwise.
11/5/99
2-63
DRAFT-DO NOT CITE OR QUOTE
-------
Table 2-11. Major components of gas-phase diesel engine emissions and their
known atmospheric transformation products
Emission component
Carbon dioxide
Carbon monoxide
Oxides of nitrogen
Sulfur dioxide
Hydrocarbons:
Alkanes (
-------
1 aqueous systems, SO2 is readily oxidized to sulfate by reaction with H2O2, O3, or O2 in the
2 presence of a metal catalyst (Calvert and Stockwell, 1983). Sulfur emitted from diesel engines is
3 predominantly (-98%) in the form of SO2, a portion of which will form sulfate aerosols by the
4 reaction described above. Off-road equipment, which typically uses fuel containing 3300 ppm
5 sulfate, emits more SO2 than on-road diesel engines, which use fuels currently containing an
6 average of 340 ppm sulfur because of EPA regulations effective in 1993 decreasing diesel fuel
7 sulfur levels. EPA (1998b) estimates that mobile sources are responsible for about 7% of
8 nationwide SO2 emissions, with diesel engines contributing 80% of the mobile source total (the
9 majority of the diesel SO2 emissions originate from nonroad engines) (U.S. EPA, 1998b).
1 0 Nitric oxide (NO) is also oxidized in the atmosphere to form NO2 and particulate nitrate.
1 1 The fraction of motor vehicle NOX exhaust converted to particulate nitrate in a 24-hour period
1 2 has been calculated using a box model to be approximately 3.5% nationwide, a portion of which
1 3 can be attributed to diesel exhaust (Gray and Kuklin, 1996). EPA estimates that in 1997, mobile
1 4 sources were responsible for about 50% of nationwide NOX emissions, with diesel engines being
1 5 responsible for approximately half of the mobile source total (U.S. EPA, 1998b).
16
17 2.3.1.3. Atmospheric Transport of Gas-Phase Diesel Exhaust
1 8 Gas-phase diesel exhaust can be dry deposited, depending on the deposition surface,
1 9 atmospheric stability, and the solubility and other chemical properties of the compound. Dry
20 deposition of organic species is typically on the order of weeks to months, with dry deposition
21 velocities of approximately 10"4 cm/sec (Winer and Busby, 1995). In contrast, inorganic species
22 such as sulfur dioxide and nitric acid have relatively fast .deposition rates (0.1-2.5 cm/sec) and
23 will remain in the atmosphere for shorter time periods compared with the organic exhaust
24 components. Some gas-phase species will also be scavenged by aqueous aerosols and potentially
25 deposited via precipitation. These processes can greatly reduce the atmospheric concentration of
26 some vapor-phase species. Atmospheric lifetimes for several gas-phase components of diesel
2*7 •*•*•» /•>»•» +l*A />»v4/a»« r\f It /M *«•<:• yxt* /io*rp y-1nr-if»rr ^Trr>f/^r\ 4*ivmA *a+f>t*-\OT^Ji^i-iio •Hiv^rviilo'npo pf^/1
I W./VllU>t*k>l Ci4W \JJLJL ULA.W V/A.UWA VJL JL iW fcii O VI. V*.t*Jr O, VI Mi. A~l JLg VVlA^Wi.1. IA.AAJ.V M.U.1..1.VS t> JLSAA^A. JIV »,V** O Vfc* WAiWt £**,*.**
7R advectinn ^an flisnpr^p the^e nolliitzmts wiHelv
_ , _ . ^, ^ ^
29
30 2.3.2. Particle-Phase Diesel Exhaust
3 1 Particle-associated diesel exhaust is composed of primarily carbonaceous material
32 (organic and elemental carbon) with a very small fraction composed of inorganic compounds and
33 metals. The organic carbon fraction adsorbed on diesel PM is composed of high-molecular-
34 weight cumpuuuus, sucii tu> FAris, wiutu tuc gcucioii^ iiiOic icaiSuilii tu auTLOSpiicnC reactions
35 than PAHs in the gas phase. The elemental carbon component of diesel exhaust is inert to
1175/99 2-65 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
»
5
6
7
8
9
10
11
12
13
14
15
16
atmospheric degradation, while the PAH compounds are degraded by reaction with the following
species:
• Sunlight, during daylight hours;
• Ozone (O3), during daytime and nighttime;
• Nitrate radical (NO3) and dinitrogen pentoxide (N205), during nighttime hours;
• Hydroxyl (OH) and hydroperoxyl radicals (HO2);
• NO2, during nighttime and daytime hours;
• Hydrogen peroxide (H2O2); and
• Gaseous nitric, acid (HNO3) and other species such as nitrous acid (HONO) and
sulfuric acid (H2SO4).
Since many of the PAH derivatives formed by reaction with some of the reactants listed
above have been found to be highly mutagenic, a brief discussion of PAH photolysis, nitration,
and oxidation follows. Some of the major degradation products from particulate diesel exhaust
are listed in Table 2-12.
Table 2-12. Major components of particle-phase diesel engine emissions and
their known atmospheric transformation products
Emission component
Elemental carbon
inorganic suiioic
Hydrocarbons (C14-C35)
PAHs (s4 rings) (e.g., pyrene,
benzo[a]pyrene)
Nitro-PAHs (*3 rings) (e.g.,
nitropyrenes)
Atmospheric reaction products i
i
i
__^__ i
i
i
Little information; j
possibly aldehydes, ketones, and alkyl !
nitrates i
Nitro-PAHs (;>4 rings); Nitro-PAH
lactones !
Hydroxylated-nitro derivatives i
Source: Adapted from Winer and Busby, 1995.
11/5/99
2-66
DRAFT—DO NOT CITE OR QUOTE
-------
1 2.3.2.1. Particle-Associated PAH Photooxidation
2 Laboratory studies of photolysis of PAHs adsorbed on 18 different fly ashes, carbon
3 black, silica gel, and alumina (Behymer and Hites, 1985, 1988) and several coal stack ashes
4 (Yokely et al., 1986; Dunstan et al., 1989) have shown that the extent of photodegradation of
5 PAHs depends very much on the nature of the substrate to which they are adsorbed. The
6 dominant factor in the stabilization of PAHs adsorbed on fly ash was the color of the fly ash,
7 which is related to the amount of black carbon present. It appears that PAHs were stabilized if
8 the carbon black content of the fly ash was greater than approximately 5%. On black substrates,
9 half-lives of PAHs studied were on the order of several days (Behymer and Hites, 1988).
10 The environmental chamber studies of Kamens et al. (1988) on the daytime decay of
11 PAHs present on residential wood smoke particles and on gasoline internal combustion emission
12 particles showed PAH half-lives of approximately 1 hour at moderate humidities and
13 temperatures. At very low-angle sunlight, very low water-vapor concentration, or very low
14' temperatures, PAH daytime half-lives increased to a period of days. The presence and
15 composition of an organic layer on the aerosol seems to influence the rate of PAH photolysis
16 (Jang and McDow, 1995; McDow et al., 1994; Odum et al., 1994).
17 Because of limited understanding of the mechanisms of these complex heterogeneous
18 reactions, it is currently impossible to draw any firm conclusion concerning the photostability of
19 particle-bound PAHs in the atmosphere. Because diesel particulate matter contains a relatively
20 high quantity of elemental carbon, it is reasonable to speculate that PAHs adsorbed onto these
21 particles might be relatively stable under standard atmospheric conditions, leading to an
22 anticipated half-life of 1 or more days.
23
24 2.3.2.2. Particle-Associated PAH Nitration
25 Since 1978, when Pitts et al. (1978) first demonstrated that benzo(a)pyrene deposited on
26 glass-fiber filters exposed to air containing 0.25 ppm NO2 with traces of HNO3 formed nitro-
27 benzo(a)pyrene, numerous studies of the heterogeneous nitration reactions of PAHs adsorbed on
28 a variety of substrates in different simulated atmospheres have been carried out (Finiayson-Pitts
29 and Pitts, 1986). PAHs deposited on glass-fiber and Teflon-impregnated glass-fiber filters react
30 with gaseous N2O5, yielding their nitro derivatives (Pitts et al., 1985b,c). The most abundant
31 isomers formed were 1-nitropyrene (1-NP) from pyrene, 6-nitro-benzo(a)pyrene from
32 benzo(a)pyrene, and 3-nitroperylene from perylene.
33 The formation of nitro-PAHs during sampling may be an important consideration for
34 diesel PM collection because of the presence of NO2 and HNO; (Feilberg et al,. 1999). However.
3 5 Schuetzle (1983) concluded that the artifact formation of 1 -NP was less than 10%-20% of the 1 -
36 NP present in the diesel particles if the sampling time was less than 23 min (one FTP cycle) and
11/5/99 2-67 DRAFT-DO NOT CITE OR QUOTE
-------
1 if the sampling temperature was not higher than 43 °C. The formation of nitroarenes during
2 ambient high-volume sampling conditions has been reported to be minimal, at least for the most
3 abundant nitropyrene and nitrofluoranthene isomers (Arey et al., 1988).
^P Diesel PM contains a variety of nitroarenes, with 1-NP being the most abundant among
5 identified nitro-PAHs. The concentration of 1-NP was measured in the extract of particulate
6 samples collected at the Allegheny Mountain Tunnel on the Pennsylvania Turnpike as 2.1 ppm
7 and ~5 ppm by mass of the extractable material from diesel and SI vehicle PM, respectively.
8 These values are much lower than would be predicted on the basis of laboratory measurements
9 for either diesel or SI engines (Gorse et al., 1983).
10 Several nitroarene measurements have been conducted in airsheds heavily affected by
11 motor vehicle emissions (Arey et al., 1987; Atkinson et al., 1988; Zielinska et al., 1989a,b;
12 Ciccioli et al., 1989, 1993). Ambient PM samples were collected at three sites in the Los
13 Angeles Basin during two summertime periods and one wintertime period. Concentrations of 1-
14 NP ranged from 3 pg/m3 to 60 pg/m3 and 3-NF was also present in diesel PM at concentrations
15 ranging from not detectable to 70 pg/m3.
16
17 2.3.2.3. Particle-Associated PAH Ozonolysis
18 Numerous laboratory studies have shown that PAHs deposited on combustion-generated
19 fine particles and on model substrates undergo reaction with O3 (Katz et al., 1979; Pitts et al,
^ 1980, 1986; Van Vaeck and Van Cauwenberghe, 1984; Finlayson-Pitts and Pitts, 1986). The
21 dark reaction toward O3 of several PAHs deposited on model substrates has been shown to be
22 relatively fast under simulated atmospheric conditions (Katz et al., 1979; Pitts et al., 1980, 1986).
23 Half-lives on the order of one to several hours were reported for the more reactive PAHs, such as
24 benzo[a]pyrene, anthracene, and benz[a]anthracene (Katz et al., 1979).
25 The reaction of PAHs deposited on diesel particles with 1.5 ppm O3 under high-volume
26 sampling conditions has been shown to be relatively fast, and half-lives on the order of 0.5 to 1
27 hour have been reported for most PAHs studied (Van Vaeck and Van Cauwenberghe, 1984).
28 The most reactive PAHs include benzo(a)pyrene, perylene, benz[a]anthracene,
29 cyclopenta[cd]pyrene, and benzo[ghi]perylene. The benzofluoranthene isomers are the least
30 reactive of the PAHs studied, and benzo[e]perylene is less reactive than its isomer
31 benzo[a]pyrene. The implications of this study for the high-volume sampling ambient POM are
32 important: reaction of PAHs with O3 could possibly occur under high-volume sampling
33 conditions during severe photochemical smog episodes, when the ambient level of O3 is high.
34 However, the magnitude of this artifact is difficult to assess from available data.
35
1115/99 2-68 DRAFT—DO NOT CITE OR QUOTE
-------
1 2.3.2.4. Atmospheric Transport of Diesel Exhaust Particle Matter
2 Ultrafine particles emitted by diesel engines undergo nucieation, coagulation, and
3 condensation to form fine particles. PM can be removed from the atmosphere by dry and wet
4 deposition. Particles of small diameter (<1 jam), such as diesel PM, are removed less efficiently
5 than larger particles by wet and dry deposition and thus have longer atmospheric residence times.
6 Dry deposition rates vary depending on the particle size. Because of their small size, diesel
7 exhaust particles will have residence times of several days (dry deposition velocities of
8 approximately 0.01 cm/sec) (Winer and Busby, 1995). Diesel particulates may be removed by
9 wet deposition if they serve as condensation nuclei for water vapor deposition, or are scavenged
10 by precipitation in- or below-cloud.
11 In a study designed to assess the atmospheric concentrations and transport of diesel
12 exhaust particles, Horvath et al. (1988) doped the sole source of diesel fuel in Vienna with an
13 organometallic compound of the heavy earth element dysprosium. The authors found that in
14 some of the more remote sampling areas, diesel PM comprised more than 30% of the particulate
15 mass, indicating that diesel PM can be dispersed widely.
16
17 2.3.3. Diesel Exhaust Aging
18 After emission from the tailpipe, diesel exhaust undergoes dilution, reaction, and
19 transport in the atmosphere. The primary emission is considered "fresh," while "aged" diesel
20 exhaust is considered to have undergone chemical and physical transformation and dispersion
21 over a period of a day or two. Laboratory dilution tunnel measurements represent a
22 homogeneous environment compared to the complex and dynamic system into which real-world
23 diesel exhaust is emitted. The physical and chemical transformation of diesel exhaust will vary
24 depending on the environment into which it is emitted. In an urban or industrial environment,
25 diesel exhaust may enter an atmosphere with high concentrations of oxidizing and nitrating
26 radicals, as well as nondiesel organic and inorganic compounds that may influence the toxicity,
27 chemical stability, and atmospheric residence time. In general, secondary pollutants formed in
28 an aged aerosol mass are more oxidized, and therefore have increased polarity and water
29 solubility (Finlayson-Pitts and Pitts, 1986). These oxidized compounds may be removed at rates
30 different from their precursor compounds and may exhibit different biological reactivities.
31 In addition, particle size distributions may vary depending on aggregation and
32 coagulation phenomena in the aging process. People in vehicles, near roadways (e.g., cyclists,
33 pedestrians, people in nearby buildings) and on motorcycles will be exposed to more fresh
34 exhaust than the general population. In some settings where emissions are entrained for long
35 periods through meteorological or other factors, exposures would be expected to include both
36 fresh and aged diesel exhaust. The complexities of transport and dispersion of emission arising
1115199 2-69 DRAFT-DO NOT CITE OR QUOTE
-------
1 from motor vehicles have been the subject of extensive modeling and experimental studies over
2 the past decades and have been summarized by Sampson (1988); exposures to diesel PM are
3 discussed in the next section of this chapter.
^P The major organic constituents of diesel exhaust and their potential degradation
5 pathways described above provide evidence for (1) direct emission of PAHs, (2) formation of
6 nitroarenes, and (3) secondary sulfate and nitrate formation. Since nitro-PAH products are often
7 more mutagenic than their precursors, the formation, transport, and concentrations of these
8 compounds in an aged aerosol mass are of significant interest.
9
10 2.4. AMBIENT DIESEL EXHAUST CONCENTRATIONS AND EXPOSURES
11 2.4.1. Diesel Exhaust Gases in the Ambient Atmosphere
12 Diesel exhaust gas is a complex mixture composed mainly of nitrogen, carbon dioxide,
13 carbon monoxide, sulfur dioxide, and volatile organic compounds including aldehydes, alkanes,
14 alkenes, and aromatic compounds such as benzene, toluene, 1,3-butadiene, naphthalene, and
15 other low-molecular-weight aromatics. The primary source of these gas-phase compounds is
16 incomplete fuel combustion and lubricating oil, with some contribution from compounds formed
17 during the combustion process or by reaction with catalysts (Johnson et al., 1994). While direct
18 emissions of several diesel exhaust components have been measured, few studies have attempted
19 to elucidate the contribution of diesel-powered engines to atmospheric concentrations of these
^fc components, most of which are emitted by several combustion sources.
21 The emission profile of gaseous organic compounds is different for diesel and SI
22 vehicles; the low-molecular-weight aromatic hydrocarbons and alkanes (C,0) and aromatic
24 hydrocarbons (such as naphthalene, methyl- and dimethyl- naphthalenes, methyl- and dimethyl-
25 indans) are more characteristic of diesel engine emissions. These differences were the basis for
26 apportionment of gasoline- and diesel-powered vehicle emissions to ambient nonmethane
27 hydrocarbon (NMHC) concentrations in the Boston and Los Angeles (South Coast Air Basin)
28 urban areas. The chemical mass balance receptor model (described below) was applied to
29 ambient samples collected in these areas, along with appropriate fuel, stationary, and area source
30 profiles (Fujita et al., 1997). The average of the sum of NMHC attributed to diesel exhaust,
31 gasoline-vehicle exhaust, liquid gasoline, and gasoline vapor was 73% and 76% for Boston and
32 the South Coast Air Basin (SoCAB), respectively. The average source contributions of diesel
33 exhaust to NMHC concentrations were 22% and 13% for Boston and the SoCAB, respectively.
34 The relative contribution of diesel exhaust will clearly depend on several factors including fleet
35 composition, sampling location (e.g., near a bus station vs. near a highway or other sources), and
the contribution from point and area sources. The source apportionment in the Fujita et al.
11/5/99 2-70 DRAFT-DO NOT CITE OR QUOTE
-------
1 (1997) study indicates that mobile vehicle-related emissions account for the majority of ambient
2 NMHC in the two urban areas studied and the results can likely be extrapolated to other urban
3 areas with similar source compositions.
4
5 2.4.2. Ambient Concentrations of Diesel PM
6 The EPA Office of Air Quality Planning and Standards report on national air pollutant
7 emission trends indicates that annual emissions of diesel PM less than 2.5 |am (PM2.5)
8 nationwide are 5.7% of the total PM2.5 inventory (21% excluding natural and fugitive dust
9 sources) (U.S. EPA, 1998b). The inventory includes on-road and off-road sources that have
10 specific local and regional distributions. As a result of inhomogeneous source distributions,
11 ambient diesel PM concentrations will vary by location. Only a few studies have been conducted
12 to assess diesel PM concentrations in urban and rural areas, local "hotspots," and the potential for
13 diesel PM episodes. The main approaches used to estimate the contribution of diesel exhaust to
14 ambient PM concentrations are receptor modeling, elemental carbon surrogate calculations, and
15 dispersion modeling. Studies conducted in Europe and Japan were reviewed, but, for the most
16 part, were not included because of questions surrounding the applicability of measurements in
17 locations that use different diesel technology and control measures from the United States.
18
19 2.4.2.1. Receptor Modeling Estimates of Diesel PM
2O Receptor models are used to infer the types and relative contributions of sources
21 impacting a receptor site on the basis of measurements made at the receptor site for the pollutants
22 of interest. As such, receptor models are referred to as "top-down" in contrast to "bottom-up"
23 methods, which use emission inventory data, activity patterns, and dispersion modeling from the
24 source to predict concentrations at a receptor site. Receptor models assume that the mass is
25 conserved between the source and receptor site and that the measured mass of each pollutant is a
26 linear sum of the contribution from each source.
27 The most commonly used receptor model for quantifying concentrations of diesel PM at a
28 receptor site is the chemical mass balance model (CMS). Input to the CMB model includes PM
29 measurements made at the receptor site as well as measurements made of each of the source
30 types suspected to impact the site. Because of problems involving the elemental similarity
31 between diesel and gasoline emission profiles and their co-emission in time and space, it is
32 necessary to carefully quantify chemical molecular species that provide markers for separation of
33 these sources (Lowentha! et a!., 1992). Recent advances in chemical analytical techniques have
34 facilitated the development of sophisticated molecular source profiles, including detailed
35 speciation of organic compounds, which allow the apportionment of PM to gasoline and diesel
11/5/99 2-71 DRAFT-DO NOT CITE OR QUOTE
-------
1 sources with increased certainty. Older studies that made use of only elemental source profiles
2 have been published and are summarized here, but are subject to more uncertainty.
The CMB model has been used to assess the contribution of diesel PM to total PM mass
in areas of California, Denver, Phoenix, and Manhattan (Table 2-13). Diesel PM concentrations
5 reported by Schauer et al. (1996) for data collected in 1982 ranged from 4.4 ug/m3 in west Los
6 Angeles to 11.6 ug/m3 in downtown Los Angeles. The average contribution of diesel PM to total
7 PM mass ranged from 13% in Rubidoux to 36% in downtown Los Angeles. It should be noted
8 that this model accounts for primary emissions of diesel PM only; the contribution of secondary
9 aerosol formation (both acid and organic aerosols) is not included. In sites downwind from
10 urban areas, such as Rubidoux in this study, secondary nitrate formation can account for a
Table 2-13. Ambient diesel PM concentrations reported from chemical mass
balance modeling
"PM10.
OC: Organic carbon.
EC: Elemental carbon.
NA: Not available.
Major ions: nitrate, sulfate, chloride and, in some cases, ammonium, sodium, potassium.
Author
Schauer et
al., 1996,
Southern
California
Chow et al.,
1991
California
EPA, 1998
Federal
Highway
Admin.,
1997b
NFRAQS,
1998
Year of
sampling,
no. days
1982,60
days (one
every sixth
day)
Winter,
1989-90,
NA
1988-92,
approx.
150 days
Spring,
1993,
3 days
Winter,
1996-97,
60 days
Location
West LA
Pasadena
Rubidoux
Downtown LA
Phoenix, AZ
area
1 5 Air basins
Manhattan, NY
Welby, CO
Brighton, CO
Location
type
Urban
Urban
Suburban
Urban
Urban
Rural-
urban
Urban bus
stop
Urban
Suburban
Source
profile used
OC species,
EC, elements
NA
EC, OC total,
Elements,
Major Ions
EC, OC total,
elements,
major ions
OC Species,
EC, elements,
major ions
Total
PM2.5
(stdev),
fig/m3
24.5 (2.0)
28.2(1.9)
42.1 (3.3)
32.5 (2.8)
NA
NA
35.8-83.0
16.7
12.4
Diesel
PM2.5
(stdev),
ug/m3
4.4 (0.6)
5.3 (0.7)
5.4 (0.5)
11.6(1.2)
4-22a
0.2-3.6"
13.2-46.7"
1.7
1.2
11/5/99
2-72
DRAFT-DO NOT CITE OR QUOTE
-------
1 substantial fraction of the mass (25% of the fine mass measured in Rubidoux was attributed to
2 secondary nitrate), a portion of which comes from diesel exhaust (Gray and Kuklin, 1996).
3 A wintertime study conducted in the Phoenix area by Chow et al. (1991) indicated that
4 diesel PM levels on single days can range from 4 ug/m3 in west and central Phoenix to 14 ug/m3
5 in south Scottsdale and 22 ug/m3 in central Phoenix. This apportionment, like the Schauer et al.
6 (1996) data, reflects direct emissions only. These data relied on source profiles and ambient
7 data collected prior to the introduction of technology to reduce PM emissions from diesel-
8 powered vehicles.
9 A second CMB study reported ambient diesel PM concentrations for California and used
10 ambient measurements from the San Joaquin Valley (1988-89), South Coast (1986), and San
11 Jose (winters for 1991 -92 and 1992-93) (California EPA, 1998a). The incorporation of sampling
12 data from later dates provides information regarding exposures more relevant to current levels.
13 The CMB in the California study (1998a) indicated that on an annual basis, basin-wide levels of
14 direct diesel PM emissions may be as low as 0.2 ug/m3 in the Great Basin Valleys and as high as
15 3.6 u,g/m3 in the South Coast basin.
16 The most recent study reporting diesel PM concentrations is from winter 1996-1997
17 sampling conducted in the Denver area as part of the Northern Front Range Air Quality Study
18 (NFRAQS), (NRC, 1998). Ambient levels of diesel PM in the urban core site at Welby averaged
19 1.7 ug/m3 over a 60-day winter period, and a slightly lower average concentration of 1.2 ug/m3
20 was measured at an urban downwind site in Brighton, CO. One of the major findings from this
21 study was a substantial contribution of elemental carbon from gasoline-powered vehicles. At the
22 Welby site, the contribution of diesel and gasoline emissions to elemental carbon measurements
23 was 52% and 42%, respectively. At the Brighton site, the contribution of diesel and gasoline
24 emissions to elemental carbon measurements was 71% and 26%, respectively. The findings
25 from the NFRAQS study are compelling and suggest the need for further investigations of this
26 type that specifically address high-emitting vehicles. Geographical and other site-specific
27 parameters that influence PM concentrations, such as altitude, must be considered when
28 extrapolating the NFRAQS findings to other locations.
29 Limited data are available to allow a characterization of diesel PM concentrations in
30 "hotspots" such as near heavily traveled roadways, bus stations, train stations, and marinas. One
31 "hotspot" study conducted in Manhattan reported diesel PM concentrations of-1-3.0 to 46.7 ug/m3
32 during a 3-day sampling period in the spring of 1993 (Federal Highway Administration, 1997b).
33 This study attributed, on average, 50% of the PM to diesel exhaust. The diesel PM
34 concentrations resulting from the source apportionment method used in this study require some
35 caution. The CMB model overpredicted PM10 concentrations by an average 30%, suggesting
36 that additional sources of the mass were not accounted for in the model. New advances in
11 /5/99 2-73 DRAFT—DO NOT CITE OR QUOTE
-------
1 organic carbon speciation, as has been noted above, are necessary to most appropriately
2 characterize gasoline and diesel PM sources to ambient PM measurements. The relevance of the
Manhattan bus stop exposure for large urban populations provides strong motivation for further
studies in the vicinity of such "hotspots."
5 In summary, recent source apportionment studies (California EPA, 1998a; NRC, 1998)
6 indicate that ambient diesel PM concentrations averaged over 2-12 month periods for
7 urban/suburban areas can range from approximately 1.2 ug/m3 to 3.6 ug/m3, while diesel PM
8 concentrations in more rural/remote areas are generally less than 1.0 ug/m3. In the vicinity of
9 "hotspots," or for short exposure times under episode-type conditions, diesel PM concentrations
10 are expected to be substantially higher than these levels; however, a thorough and replicated
11 characterization of these situations is not yet available. Two studies nearing completion by the
12 South Coast Air Quality Management District will shed some light on near-highway
13 concentrations of diesel PM (SCAQMD, 1999).
14
15 2.4.2.2. Elemental Carbon Surrogate for Diesel PM
16 Elemental carbon (EC) is a major component of diesel exhaust, contributing to
17 approximately 60%-80% of diesel particulate mass, depending on engine technology, fuel type,
18 duty cycle, lube oil consumption, and state of engine maintenance (Graboski et al., 1998; Zaebst
19 et al., 1991; Pierson and Brachaczek, 1983; Wamer-Selph et al., 1984). In most ambient
^P environments, diesel PM is one of the major contributors to EC, with other potential sources
21 including spark-engine exhaust; combustion of coal, oil, or wood; charbroiling; cigarette smoke;
22 and road dust. Gasoline combustion was recently found to be an important source of elemental
23 carbon in Denver (NRC, 1998).
24 Because of the large portion of EC in diesel PM, and the fact that diesel exhaust is one of
25 the major contributors to EC in most ambient environments, diesel PM concentrations can be
26 bounded using EC measurements. Source apportionment (NRC, 1998) indicates that diesel
27 exhaust comprises from 52% to 71% of the elemental carbon concentrations in ambient PM in
28 the Denver area for the winter of 1996-97. If we assume that gasoline and diesel exhaust
29 contributions to EC measured in Denver will be similar to other areas, a plausible estimate of
30 diesel particulate concentrations can be calculated by multiplying a measured EC concentration
31 by 64% and dividing by the fraction of diesel PM mass accounted for by EC (note: a middle-of-
32 the-range value of 70% was chosen for illustrative purposes), e.g., diesel PM concentration =
33 [(EC * 0.64)70.7]. This estimate uses an average of the diesel contributions to EC observed in
34 Denver with contributions from other, potentially site-specific sources of EC subtracted (e.g.,
35 wood smoke). An upper-bound estimate of diesel PM from EC measurements would attribute all
11/5/99 2-74 DRAFT-DO NOT CITE OR QUOTE
-------
1 ambient EC to diesel exhaust, e.g., diesel PM concentration = EC/0.7, which may be applicable
2 to some occupational exposures.
3 The surrogate diesel PM calculation is a useful approach for estimating diesel PM in the
4 absence of a more sophisticated receptor modeling analysis for locations where fine PM
5 elemental carbon concentrations are available. Table 2-14 provides diesel PM concentrations
6 that were calculated using the EC surrogate ratiometric approach. Under an EPA Research Grant
7 with the Northeastern States for Coordinated Air Use Management (NESCAUM), PM2.5
8 samples were collected in Boston (Kenmore Square), MA, Rochester, NY, Reading, MA,
9 Quabbin Reservoir, MA, and Brockport, NY (Salmon et al.. 1997). Samples were collected
10 every sixth day for one year (1995). Using the EC surrogate calculation described above, diesel
11 PM concentrations are estimated to range from 0.3 ug/m3 in Brockport, NY to 1.1 ug/m3 in
12 Boston.
13 The Interagency Monitoring of Protected Visual Environments (IMPROVE) project of
14 the National Park Service operates an extensive aerosol monitoring network in mainly rural or
15 remote areas of the country (National Parks, National Monuments, Wilderness Areas, National
16 Wildlife Refuges, and National Seashores). PM2.5 samples, collected twice weekly for 24-hour
17 duration, at 43 sites (some co-located in the same rural park area) were analyzed for a suite of
18 chemical constituents, including elemental carbon. The IMPROVE data in Table 2-14 represent
19 average values for the period from March 1992 through February 1995 (Sisler, 1996).
Table 2-14. Diesel PM 2.5 concentrations in urban and rural locations using
EC surrogate" for NESCAUM (1995) and IMPROVE (1992-1995) network
sites
Area
Annual average
PM2.5 ug/m3
Diesel PM2.5
pig/m3
Urban areas
Boston, MAb
Rochester, NY"
Washington, DCC
16.2
1.1
14.9 j 0.5
19.2
i <
i.\i
Nonurban areas
Quabbin, MAb
Reading, MAb
Brockport, NY6
IMPROVE sites
12.4
14.6
12.8
1.8-12.1
0.4
0.6
0.3
0.1-0.8
a Assumes 64% of EC mass is from diesel exhaust and 70% of diesel PM is elemental carbon.
b NESCAUM sites.
c IMPROVE sites.
11/5/99
2-75
DRAFT-DO NOT CITE OR QUOTE
-------
The estimated diesel PM concentrations for rural/remote areas from the IMPROVE
network ranged from 0.1 ug/m3 for Denali National Park, AK, to 0.8 ug/m3 for the Lake Tahoe,
CA, area (which includes on-road and off-road diesel emissions). The diesel PM concentrations
4 in rural areas of the northeastern States are similar to those calculated for rural areas in other
5 parts of the country, with all values less than 1.0 ug/m3. Among the urban areas included in the
6 NESCAUM and IMPROVE networks, Washington, DC, had the highest calculated diesel PM
7 concentration of 1.6 ug/m3. The annual average value for Washington, DC, is quite similar to the
8 wintertime diesel PM concentrations reported for Denver (NRC, 1998). Seasonally averaged
9 data for the Washington, DC, site indicates that EC concentrations, and by extension, diesel PM
10 concentrations at this site peak in the autumn and winter (1.9 and 1.8 ug/m3 diesel PM,
11 respectively).
12 Recently, EC measurements were reported for enclosed vehicles driving on Los Angeles
13 roadways (California EPA, 1998b). Applying the ratiometric approach for diesel PM
14 determinations from EC measurements, diesel PM concentrations in the vehicle ranged from
15 approximately 2.8 ug/m3 to 36.6 ug/m3 depending on the type of vehicle being followed (higher
16 concentrations were observed when the vehicle followed HD diesels). The California Air
17 Resources Board also collected EC near the Long Beach Freeway for 4 days in May 1993 and 3
IB days in December 1993 (California EPA, 1998a). Using emission estimates from their
^P EMFAC7G model, and elemental/organic carbon composition profiles for diesel and gasoline
20 exhaust, tire wear, and road dust, the California Air Resources Board estimated the contribution
21 of the freeway to diesel PM concentrations. For the 2 days of sampling in December 1993,
22 diesel exhaust from vehicles on the nearby freeway were estimated to contribute from 0.7 ug/m3
23 to 4.0 ug/m3 excess diesel PM above background concentrations, with a maximum of 7.5 ug/m3.
24 In two additional studies, EC concentrations were measured in Glendora, CA, during a
25 carbonaceous aerosol intercomparison study (Cadle and Mulawa, 1990; Hansen and Novakov,
26 1990). EC concentrations ranged from 5.0 ug/m3 to 6.4 ug/m3, corresponding to diesel PM
27 concentrations of 4.6 ug/m3 to 5.9 ug/m3 using the ratiometric approach described above. One
28 technique used during the study reported EC concentrations in 1-minute intervals, revealing the
29 impact from diesel vehicles 50 meters from the study site. The diesel vehicles were estimated to
30 contribute up to 5 ug/m3 EC above the background concentration.
31 In a study designed to investigate relationships between diesel exhaust exposure and
32 respiratory health of children in the Netherlands, EC measurements were collected in 23 schools
33 located from 47 to 377 meters from a freeway and in 8 schools located at a distance greater than
34 400 meters from a freeway (Brunekreef, 1999). EC concentrations in schools near freeways
ranged from 1.1 to 6.3 ug/m3, with a mean of 3.4 ug/m3, and EC concentrations in schools more
11/5/99 2-76 DRAFT-DO NOT CITE OR QUOTE
-------
1 than 400 meters from freeways ranged from 0.8 to 2.1 |ag/m3 with a mean of 1.4 ug/m3. Using
2 the EC surrogate calculation for diesel PM described above, the estimated average diesel PM
3 concentration in the schools near a freeway is 3.1 U£/m3 and the estimated average diesel PM
4 concentration in schools located more than 400 meters from a freeway is 1.3 ug/m3. Total
5 PM2.5 mass inside the schools averaged 23.0 ug/m3 while PM2.5 outside was only slightly
6 higher (24.8 ug/m3), suggesting extensive intrusion of outdoor air into the school environment.
7
8 2.4.2.3. Dispersion Modeling Results
9 Dispersion models estimate ambient levels of PM at a receptor site on the basis of
10 emission factors for the relevant sources and the investigator's ability to model the advection,
11 mixing, deposition, and chemical transformation of compounds from the source to the receptor
12 site. Cass and Gray (1995), Gray and Cass (1998), and Kleeman and Cass (1998) have applied
13 dispersion models to the South Coast Air Basin to estimate diesel PM concentrations. The
14 models used by these investigators applied emission factors from 1982 and consequently are
15 representative of concentrations prior to the implementation of diesel PM emission controls. In
16 addition to offering another approach for estimating ambient diesel PM concentrations, the
17 dispersion models summarized here provide the ability to distinguish on-highway from off-
18 highway diesel source contributions and have presented an approach for quantifying the
19 concentrations of secondary aerosols from diesel exhaust.
20 Cass and Gray (1995) used a Lagrangian particle-in-cell model to estimate the source
21 contributions to atmospheric fine carbon particle concentrations in the Los Angeles area,
22 including diesel emission factors from on-highway and off-highway sources. Their dispersion
23 model indicates that for 1982, the annual average ambient concentrations of diesel PM ranged
24 from 1.9 ug/m3 in Azusa, CA, to 5.6 ug/m3 in downtown Los Angeles (Table 2-15). The
25 contribution of on-highway sources to diesel PM ranged from 63.3% in downtown Los Angeles
26 to 89% in West Los Angeles. Of the on---high\vay diesel contribution, the model predicted that
27 for southern California, HD trucks c^mr»r'?pH the ma'ontv (85%) of the diesel PM inventory and
7 f ^ f i r v
28 overall they contributed 66% of the diesel PM in the ambient air. Off-road sources of diesel
29 exhaust include pumping stations, construction sites, shipping docks, railroad yards, and heavy
30 equipment repair facilities. Cass and Gray (1995) also report that wintertime peaks in diesel PM
31 concentrations can reach 10 ug/m3.
32 Kleeman and Cass (1998) developed a Lagrangian model that examines the size and
33 chemical evoiuiion of aerosols incluuiug gas-lo-particle conversion processes during transport.
34 This model was applied to one-well characterized episode in Claremont County, CA, on August
1175/99 2-77 DRAFT-DO NOT CITE OR QUOTE
-------
Table 2-15. Modeled diesel PM2.1 for South Coast Air Basin in 1982a
Location
Azusa
Pasadena
Anaheim
Long Beach
Downtown Los Angeles
Lennox
West Los Angeles
On-highway
diesel PM,
ug/m3
1.4
2.0
2.7
3.5
3.5
3.8
3.8
Total diesel
PM, ng/m3
1.9
2.5
3.5
4.6
5.6
4.7
4.3
% On-
highway
75
78
78
76
63
81
89
'Adapted from Cass and Gray (1995), modeling results of Gray (1986).
1
2
f
7
8
9
10
11
12
13
14
15
16
17
18
19
20
27-28, 1987. The model provided reasonable predictions of PM10 (overpredicting PM10 13%),
elemental and organic carbon, and adequately reconstructed the size distribution of the aerosols.
The model indicated that on August 27-28, 1987, the PM2.5 concentration was 76.7 ng/m3,
13.2% of which (10.1 ng/m3) was attributable to diesel engine emissions. This estimate includes
secondary aerosol formation for sulfate, ammonium, nitrate, and organic compounds, which
accounted for 4.9 ug/m3 of the total estimated diesel PM mass. The secondary organic aerosol
was estimated to be 1.1 fig/m3, or 31% of the total seconary aerosol mass, with the remainder
composed of nitrate, ammonium, and sulfate aerosols.
Dispersion modeling is also being conducted by EPA to estimate county-specific
concentrations of, and exposures to, several toxic species, including diesel PM. Results from this
model are expected in 2000.
2.4.3. Exposures to Diesel PM
Up to this point, the information on diesel PM has focused on estimates of concentrations
in outdoor environments. Ultimately, it is personal exposure that determines health impacts.
Personal exposures can be measured using surrogate chemical species such as EC, or exposures
can be modeled. Results of both exposure assessment methods are discussed below.
Occupational exposures to diesel PM were reported for long-distance truck drivers, local
drivers, mechanics, and dockworkers by Zaebst et al. (1991), and other occupational exposures
are summarized by Watts (1995) and Birch and Gary (1996). Two modeling efforts have been
developed to determine diesel PM exposures in the general population: the Hazardous Air
11/5/99
2-78
DRAFT-DO NOT CITE OR QUOTE
-------
1 Pollutant Exposure Model-Mobile Sources version 3 (HAPEM-MS3) and the California
2 Population Indoor Exposure Model (CPIEM).
3
4 2.4.3.1. Exposure Measurements
5 Occupational exposures to diesel PM were estimated by Zaebst et al. (1991), who
6 reported EC measurements from personal samplers worn by road drivers, local drivers,
7 dockworkers, and mechanics for 8-hour shifts at each of six large hub truck terminals.
8 Residential background and highway background samples at fixed sites were also collected.
9 Zaebst et al. (1991) reported warm- and cold- weather EC concentrations in residential
1 0 background and highway background environments, which ranged from 0.9 ug/m3 to 4.9 ug/m3.
1 1 Elemental carbon exposures for road and local truckers ranged from 2.0 u-g/m3 to 7.0 |ig/m3,
1 2 while exposure levels for mechanics and dockworkers were reported between 4.8 ug/m3 and 28.0
1 3 ug/m3.
1 4 The geometric mean of the EC concentrations reported by Zaebst et al. (1991) was
1 5 adjusted for the potential contribution of other EC sources using the ratiometric approach
1 6 described above. The estimated diesel PM exposures calculated range from 3.5 ng/m3 and 3.7
1 7 ug/m3 for road and local drivers, respectively, to 12.6 ug/m3 for mechanics (Table 2-16).
1 8 Important variables in this calculation include the potential contribution of 2-stroke diesel
1 9 engines (which generate PM with lower EC concentrations than 4-stroke diesel engines) and the
20 contribution of other EC sources such as cigarette smoke, wood smoke, or gasoline combustion
2 1 above the level accounted for. The exposure levels for road and local drivers to diesel PM
22 estimated from the Zaebst et al. (1991) study are a factor of 2 to 3 higher than recent ambient PM
23 levels reported for Denver, which is reasonable given that drivers are likely to be in closer
24 proximity to traffic than at either of the two Denver fixed sites.
25 Additional occupational exposures to EC have been reported for miners, fire engine
og c^sratcrs in engine houses automotive repair sho*"*s dedicated to d'ssel vehicles se^/i^e ^ay
?7 workers jp a public transit system, and aircraft ground crews (Birch and Gary. 1 996: Watts.
28 1995). Diesel PM exposures were calculated from the EC exposures using an upper-bound
29 estimate (e.g., EC concentration = diesel PM/0.7) because the workers were generally in confined
30 spaces in which diesel exhaust was the dominant source of EC. If this upper-bound estimate is
31 applied, the calculated occupational exposures to diesel PM range from 10 to 21 ug/m3 for an
32 aircraft ground crew, and up to 43 ug/m3, 1 1 3 ug/m3, and 140 ug/m3 for bus public transit areas,
~~ f f i * • ^i... ^_i- — i ------ --- 1 u — ^ ---- tj. — —-•- — i ____ _____ ___ i _ _____ A.; ___ i.. T_ _jj:*:^__i
OO lllCilgllLCii ill uic ai.ai.iyu injuot, ay pciauiui^i, iwDjp^v-u vt,i^ . 111 auumwiicu
34 occupational settings, breathing zone concentrations of EC have been reported by Birch and Gary
1175/99 2-79 DRAFT—DO NOT CITE OR QUOTE
-------
Table 2-16. Diesel PM1.0 exposures reported by Zaebst et al. (1991)
and calculated using the EC ratiometric approach0
Location/job type
Residential background
Highway background
Road drivers
Local drivers
Dockworkers
Mechanics
Number of
samples
23
21
72
56
75
80
Geom. mean
diesel PM
(stdev) ug/m3
1.1(2.0)
2.5 (2.4)
3.8 (2.3)
4.0 (2.0)
12.1 (3.7)
13.8(3.6)
Calculated
diesel PM%
jig/m3
1.0
2.3
3.5
3,7
11.1
12.6
1 Diesel PM=(EC*0.64)/0.7.
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
(1996) and the estimated maximum diesel PM concentrations from these measurements suggest
levels of 24 ug/m3 for a beverage distributor warehouse, 100 |ig/m3 for diesel automotive repair
crews, 286 ug/m3 for a front end loader operator in a confined space of a timber processing plant,
and 976 ug/m3 for a firehouse bay area.
Watts (1995) reports diesel PM sampling conducted in mines during significant diesel
activity, which does not represent personal exposures, but is a snapshot of short periods of
elevated concentration that comprise a portion of a worker's daily exposure. The levels of diesel
PM in four mines ranged from 850 ug/m3 to 3260 ug/m3. In a study of four railroads, Woskie et
al. (1988) reported concentrations of respirable dust (corrected for cigarette particulate) that
ranged from 17 jag/m3 for clerks to 130 ug/m3, 134 ug/m3, and 191 ug/m3 for supervisors,
electricians, and hostlers, respectively. Although these exposures may have included nondiesel
PM, the majority of the respirable PM is believed to have originated from the locomotive
emissions.
2.4.3.2. Modeling Exposures to Diesel PM
Modeled estimates of individual exposures to diesel PM must integrate exposure in the
various indoor and outdoor environments in which different individuals are active.
Consequently, the demographic distribution, time-activity patterns, and diesel PM concentrations
in the various environments, including job-related exposures, must all be taken into account.
11/5/99
2-80
DRAFT-DO NOT CITE OR QUOTE
-------
1 2.4.3.2.1. The Hazardous Air Pollutant Exposure Model for Mobile Sources, version 3. The
2 HAPEM-MS3 model is based on the carbon monoxide (CO) probabilistic NAAQS exposure
3 model (CO pNEM), which is used to estimate the frequency distribution of population exposure
4 to CO and the resulting carboxyhemaglobin levels (Law et al., 1997). The HAPEM model
5 simulates the CO exposure scenario of individuals in 22 demographic groups for 37
6 microenvironments. CO concentrations are based on ambient measurements made in 1990 and
7 are related to exposures of individuals in a 10 km radius around the sampling site. Diesel PM
8 (DPM) exposures are calculated as in Equation 2-4, using a ratiometric approach to CO.
9
10 DPM , , =(CO , , ICOelmi)^DPMalmi (2-4)
uglm ^ uglm g'1"' ' "'""
11
1 2 Input to the model includes CO monitoring data for 1 990, time-activity data collected in
1 3 Denver, CO, Washington DC, and Cincinnati, OH, from 1982 to 1985, microenvironmental data,
1 4 and 1 990 census population data. Motor vehicle diesel PM and CO emission rates reported by
1 5 EPA (1999b) are used to calculate mobile-source diesel PM exposures. While gasoline vehicles
1 6 emit the large majority of CO, gasoline and diesel highway vehicles travel on the same
1 7 roadways, albeit with different spatial and temporal patterns. Nevertheless, the assumption can
18 be made that the highway fleet (gasoline+diesel) emissions ratio of CO to diesel PM can be used
19 as an adjustment factor to convert known or estimated CO personal exposure to diesel. PM
20 exposure estimates.
2 1 This modeling approach was first used to estimate population average exposures among
22 22 demographic groups for 9 urban areas (U.S. EPA, 1999b). The exposures were calculated
23 based on air districts, which were defined as the population within the 10 km radius of the CO
24 monitor. Employing average CO exposures in this approach may underestimate by
25 approximately 30% the exposures experienced by the 98th percentile population (Law et al.,
28 1 997). In order to characterize exposures in these highly exposed populations, Brcdowicz ( 1 999)
27 i:sed CO concentratioiis relev?1?* to tb.« most highly exposed populations to determine diesel PM
28 exposures for different demographic groups within this population. Results for both the annual
29 average diesel PM exposures and exposures in the most highly exposed demographic groups are
30 given in Table 2- 17.
31 The annual average diesel PM exposures ranged from 0.6 |ig/m3 in Spokane, WA, to 1.7
32 ng/'m3 in New York (Table 2-17). The highest diesel PM exposures ranged from 0.9 pg/m3 for
• • • r* . T • * + I ^ f* j. y 1 * 1 1 _ " _ "XT X ^ 1 /'TP 1_ 1
33 Outdoor WOrKCrS in at. LOUIS 10 H.I ng/m iui uuuiuui cimuicii iii INCVV i OIK ^iauuC I.-L i ).
34 highest exposed demographic groups were those who spend a large portion of their time
35 outdoors. Overall, the highest exposed individuals experienced diesel PM levels that were on
1 1 75/99 2-81 DRAFT— DO NOT CITE OR QUOTE
-------
Table 2-17. Annual average diesel PM exposures for 1990 in
the general population and among the highest exposed
demographic groups in nine urban areas (on-road sources
only)
Urban area
Chicago, IL
Denver, CO
Houston, TX
Minneapolis, MN
New York, NY
Philadelphia, PA
Phoenix, AZ
Spokane, WA
St. Louis, MO
Population
average
exposure,
ug/m3
0:8
0.8
0.6
1.0
1.7
0.7
1.4
1.3
0.6
Highest diesel PM exposure,
ug/m3 (demographic group
experiencing this exposure)
1 .3 (outdoor workers)
1 .3 (outdoor workers)
0.9 (outdoor workers)
1 .6 (outdoor workers)
4. 1 (outdoor children)
1.3 (outdoor children)
2.6 (nonworking men 18-44)
2.0 (outdoor workers)
0.9 (outdoor workers)
1 average 1.7 times higher than the population average. Exposures to diesel PM in rural areas
2 nationwide were estimated by HAPEM-MS3 to be 0.5 ^g/m3. It is important to note that these
3 exposure estimates are lower than the total exposure to diesel PM because they reflect only diesel
4 PM from on-road sources (U.S. EPA, 1998b).
5 Diesel PM exposure projections were calculated using the HAPEM model for 1996,
6 2007, and 2020 with a base case and a case assuming increased penetration of diesel engines in
7 the LD truck fleet. The base case uses baseline fuels and emission rates, assuming the
8 implementation of a National Low-Emission Vehicle (NLEV) program. The increased
9 dieselization case uses baseline emission factors with Tier 2 standards and an assumed increase
10 in LD diesel truck implementation equivalent to 50% of the LD truck sales beginning in model
11 year 2004 (which is more aggressive than the likely increase in diesel engine market share).
12 Predicted diesel PM exposures for 2007 and 2020 decrease from 1996 levels by an
13 average 55% and range from a low of 0.3 ug/m3 in St. Louis to a high of 0.6 ug/m3 in Phoenix,
ffor both 2007 and 2020 (Table 2-18). The predicted decreases are a result of fleet turnover and
the full implementation of Federal regulations that are currently in place. If the modeled increase
1175/99 2-82 DRAFT-DO NOT CITE OR QUOTE
-------
Table 2-18. Projected annual average diesel PM exposures from all
on-road vehicles
Area
Chicago, IL
Denver, CO
Houston, TX
Minneapolis, MN
New York, NY
Philadelphia, PA
Phoenix, AZ
Spokane, WA
St. Louis, MO
Diesel PM exposure by calendar year, jig/m3
1996
0.6
0.7
0.8
0.9
1.1
0.6
1.2
1.0
0.5
2007
0.3
0.4
0.3
0.4
0.5
0.3
0.6
0.5
0.3
2020
0.3
0.4
0.3
0.4
0.5
0.2
0.6
0.5
0.2
1 in diesels in the LD truck fleet occurs, projected diesel PM exposures are expected to increase
2 38% on average over 1996 exposures. If diesel engines reached 50% of the light duty truck
3 sales in 2010, instead of 2004, the increase in diesel PM exposure would be about 30%.
4
5 2.4.3.2.2. The California Population Indoor Exposure Model The California Population
6 Indoor Exposure Model (CPIEM), developed under contract to the California Air Resources
7 Board, estimates Califomians' exposure to diesel PM using distributions of input data and a
8 Monte Carlo approach (California EPA, 1998a). This model uses population-weighted outdoor
9 diesel PM concentrations in a mass-balance model to estimate diesel PM concentrations in four
10 indoor environments: residences, office buildings, schools, and stores/retail buildings. The
11 model takes into account air exchange rates, penetration factors, and a net loss factor for
12 deposition/removal. In four additional environments (industrial plants, restaurants/lounges, other
13 indoor places, and enclosed vehicles), assumptions were made about the similarity of each of
14 these spaces to environments for which diesel PM exposures had been calculated^ Industrial
15 plants and enclosed vehicles were assumed to have diesel PM exposures similar to those in the
16 outdoor environment, restaurant/lounges were assumed to have diesel PM concentrations similar
17 to stores, and other indoor places were assumed to have diesel PM concentrations similar to
18 offices. The estimated diesel PM concentrations in the indoor and outdoor environments range
19 from 1.6 pig/m3 to 3.0 ug/m3 (Table 2-19).
11/5/99
2-83
DRAFT-DO NOT CITE OR QUOTE
-------
Table 2-19. Modeled and estimated concentrations of
diesel PM in microenvironments (California EPA,
1998a)
Environment
Residences
Offices
Schools
Stores/public/retail bldgs
Outdoor places
Industrial plants3
Restaurants/lounges2
Other indoor places"
Enclosed vehicles"
Estimated mean diesel
PM (stdev), ug/m3
1.9(0.9)
1.6(0.7)
1.9(0.8)
2.1 (0.9)
3.0(1.1)
3.0(1.1)
2.1 (0.9)
1.6(0.7)
3.0(1.1)
"Concentrations assumed based on similarity with modeled environments.
The diesel PM concentrations reported in Table 2-19 were used as input to the CPIEM
2 model, and time-activity patterns for children and adults were used to estimate total indoor and
3 total air exposures to diesel PM. Overall, total indoor exposures were estimated at 2.0 ± 0.7
4 ug/m3 and total air exposures (indoor and outdoor exposures) were 2.1 ± 0.7 ug/m3 (Table 2-20).
5 The South Coast Air Basin and the San Francisco Bay Area were also modeled using CPIEM,
6 where total air exposures to diesel PM were estimated to be 2.5 ± 0.9 ng/m3 and 1.7 ± 0.9 ug/m3,
7 respectively.
8 Exposure estimates were also made by California EPA (1998a) for 1995,2000, and 2010
9 using a ratiometric approach to 1990 exposures. Total air exposures reported for 1995 and
10 projected for 2000 and 2010 were 1.5 ug/m3, 1.3 ug/m3, and 1.2 |ig/m3, respectively.
11
12 2.4.4. Ambient Diesel PM Summary
13 It appears from the limited number of studies available that annually averaged diesel PM
14 concentrations at fixed sites, in urban and suburban areas in the 1980s ranged from approximately
15 4.4 ug/m3 to 11.6 ug/m3. CMB and dispersion modeling indicate that diesel PM concentrations
16 on some winter days may reach 22 ug/m3 and on episode days concentrations of 10 ug/m3 are
possible. CMB modeling results, which include emissions and measurements from 1990 and
11/5/99
2-84
DRAFT—DO NOT CITE OR QUOTE
-------
Table 2-20. Estimated indoor air and total air exposures to diesel PM in
California in 1990
Exposed population
All Califomians
•South Coast Air Basin
(
San Francisco Bay Area
Total indoor
exposure (stdev),
ug/m3
2.0 (0.7)
2.4 (0.9)
1.7(0.9)
Total air
exposure, (stdev),
ug/m3
2.1 (0.8)
2.5 (0.9)
1.7(0.9)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
on
^.A.
23
24
later, indicate that diesel PM concentrations in "hotspots" may reach 46.7 ug/m3; diesel PM
concentrations at fixed sites in urban and suburban areas range from approximately 1.2 ug/m3 to
3.6 ug/m3. Annual average diesel PM concentrations in rural and remote areas of the country are
less than 1.0 ug/m3.
Measurements of the diesel PM surrogate, EC, in rural and urban environments indicate
concentrations similar to those reported by CMB methods, with diesel PM concentrations less
than 1.0 ug/m3 in rural areas and concentrations from approximately 0.5 ug/m3 to 5.9 ug/m3 in
urban areas. The EC surrogate approach also provides some estimates of diesel PM in
microenvironments such as in-vehicle concentrations (2.8-36.6 ug/m3), near roadways with
diesel traffic (0.7-7.5 ug/m3 higher than background), and in schools (0.9-5.5 ug/m3).
Measurements of EC in occupational environments indicate that diesel PM exposures range from
approximately 3.5 ug/m3 for long distance diesel truck drivers to 140 ug/m3 for bus transit
service bay personnel.
It is noteworthy that the annually averaged concentrations and exposures will mask
potentially important ex.e-ursions experienced during episodic conditions, and/or seasonal
elevations that may have important associated health risks. Individuals for whom exposures are
equal to, or may greatly exceed, the annual average ambient diesel PM levels reported here may
include those who spend a significant portion of their day on or near roadways, such as sales
representatives, delivery personnel, construction workers, and individuals living in the vicinity of
"hotspots" (near highway, bus depot, or other transit facility).
The HAPEM-MS3 exposure model, which assesses exposures from en-road diesel
diesel PM from 0.6 ug/m3 to 1.7 ug/m3. Diesel PM exposures for the most highly exposed
individuals in urban areas are estimated by HAPEM-MS3 to range from 0.9 ug/m3 to 4.1 ug/m3,
11/5/99
2-85
DRAFT—DO NOT CITE OR QUOTE
-------
with individuals spending a large amount of time outside comprising the highest exposure group.
The California EPA (1998a) exposure model, which assesses diesel PM exposures from on- and
off-road sources, reports diesel PM exposures for Califomians ranging from 1.7 ug/m3 to 2.5
4 ng/m3. Projected diesel PM exposure levels are expected to decrease by 2007 in the absence of
5 an increase in LD diesel truck implementation.
6
7 2.5. SUMMARY
8 Dieselization of the trucking industry first occurred in the 1930s, and reached 40%-50%
9 of the market for Class 7 and 8 trucks (based on sales data) by early 1960. By the late 1980s,
10 more than 95% of heavy trucks used diesel engines. Dieselization of locomotives began at the
11 end of the Second World War and was completed rapidly, probably by the early 1950s.
12 Technology innovations that impact emissions have occurred in the years since 1960, in
13 particular the advent of turbocharging with charge air cooling, and DI engines. These advances
14 have tended to lower emissions, but until the late 1980s engines were optimized for performance
15 rather than emissions, so the effect was small. Overall, it is expected that engines in the 1950 to
16 1980s time frame would have PM emissions similar to those of the mid-1980 engines that were
17 not yet controlled for particulates.
The proportion of 2-stroke engines in the in-use truck fleet was in all likelihood 20%-
25% for most of the time from 1960 to 1985. Only in the late 1980s did 2-stroke engines begin
20 to decline. Overall, regulated emissions changes due to changing proportions of 2- and 4-stroke
21 engines hi the in-use fleet during the years 1949-1975 do not appear to be significant for HD
22 truck and bus engines. No significant difference in PM mass emissions between 2- and 4-stroke
23 vehicles are evident; however, 2-stroke engines emitted PM with a higher organic and higher
24 amounts of VOCs than did 4-stroke engines.
25 Regulated emissions of CO, HC, and PM have declined significantly for on-road trucks
26 since the mid-1970s. PM emissions appear to have decreased by a factor of 6 while emissions of
27 NOX have remained approximately constant. Emissions trends for earlier years are unknown;
28 however, given that there were no emissions regulations in effect until the early 1970s it is likely
29 that emissions were fairly constant during the 1960s. Little change in locomotive emissions from
30 the early 1970s to the 1990s is evident. It is likely that this trend can also be extrapolated back to
31 the mid-1950s.
32 Data on nonregulated emissions and particle size were reviewed. It is apparent that the
33 soluble organic fraction of particulate, as well as the solid portion, have declined during the past
34 two decades. EC content comprises the largest fraction of diesel PM and so has declined. There
is also evidence for a decrease in the percentage of organics adsorbed on the particulates over
11/5/99 2-86 DRAFT—DO NOT CITE OR QUOTE
-------
1 time. Emissions of PAHs and nitro-PAHs appear to have declined in parallel with emissions of
2 total PM and SOF. There is no evidence to suggest that lexicologically significant organic
3 components of DE (e.g., PAHs and nitroaromatics) have changed out of proportion to the change
4 in organic mass.
5 Particle size measurements suggest that the size distribution for current emissions may be
6 shifted more toward slightly higher number concentrations of nuclei-mode particles. However,
7 methodologies for assessing nuclei particles are in a very early stage of technical development
8 and no conclusions can be made at the current time.
9 The dilution of exhaust under roadway conditions is not well simulated by dynamometer
10 dilution tunnel tests (dilution ratios of approximately 1000 in the ambient environment,
11 compared to 10-fold dilution in laboratory tests). This discrepancy may lead to particle size
12 distribution and gas-particle phase distributions of semivolatile compounds under conditions
13 slightly different from those predicted from laboratory data. Diesel engines emit several
14 lexicologically important compounds, including nitroarenes and other PAH compounds. The
15 chemical and physical changes of diesel exhaust in the atmosphere have been extensively
16 explored, but knowledge concerning the products of these chemical transformations is still
17 limited and is challenging to predict from laboratory tests. In general, diesel exhaust components
18 will become more oxidized hi the atmosphere, making them more polar and therefore more
19 water-soluble. Secondary aerosols from diesel exhaust may be removed at rates different from
20 then* precursor compounds, and may exhibit different biological reactivities.
21 Diesel PM concentrations reported from chemical mass balance studies in the 1980s
22 suggest that annually averaged concentrations ranged from approximately 4.4 ug/m3 to 11.6
23 ug/m3. More recent analysis suggests that annually averaged ambient concentrations of diesel
24 PM range from 0.2 fig/m3 to 3.6 Hg/m3, with levels below 1.0 ug/m3 for the more rural/remote
25 areas. Chemical mass balance modeling and dispersion analysis suggest that in urban hot spots
26 and during episodic conditions, diesel PM concentrations may be as high as 10-47 fig/m3.
27 Exposure modeling has indicated that individuals from the general population in urban areas may
28 be exposed to 0.6-1.7 ug/m3 diesel PM, while those individuals who spend a large amount of
29 their time out of doors may have exposures ranging up to 4.1 ug/m3. Diesel PM exposures in
30 some occupational environments can exceed these levels by 2-3 orders of magnitude.
31
1115199 2-87 DRAFT-DO NOT CITE OR QUOTE
-------
2.6. REFERENCES
AAMA, American Automobile Manufacturers Association. (1975-1998) Motor vehicle facts & figures.
4
5 Abdul-Khalek, IS; Kittelson, DB; Graskow, BR; et al. (1998) Diesel exhaust particle size: measurement issues and
6 trends. SAE Technical Paper Ser. No. 980525.
7
8 Abdul-Khalek, IS; Kittelson, DB; Brear, F. (1999) The influence of dilution conditions on diesel exhaust particle
9 size distribution measurements. SAE Technical Paper Ser. No. 1999-01-1142.
10
11 AMA, Automobile Manufacturers Association. (1927-1974) Motor truck facts. Motor Truck Committee.
12
13 Arey, J. (1998) Atmospheric reactions of PAHs including formation of nitroarenes. In: The handbook of
14 environmental chemistry, Vol. 3, Part I. PAHs and related compounds. Nielsen, AH, ed. Berlin/Heidelberg,
15 Germany: Springer-Verlag, pp. 347-385.
16
17 Arey, J; Zielinska, B; Atkinson, R, et al. (1986) The formation of nitro-PAHs from the gas-phase reactions of
18 fluoranthene and pyrene with the OH radical in the presence of NOX. Atmos Environ 20:2339-2345.
19
20 Arey, J; Zielinska, B; Atkinson, R; et al. (1987) Polycyclic aromatic hydrocarbon and nitroarene concentrations in
21 ambient air during a wintertime high-NOx episode in the Los Angeles Basin. Atmos Environ 21:143 7-1444.
22
«Arey, J; Zielinska, B; Atkinson, R; et al. (1988) Formation of nitroarenes during ambient high-volume sampling.
Environ Sci Technol 22:457-462.
25
26 Arey, J; Zielinska, B; Atkinson, R; et al. (1989) Nitroarene products from the gas-phase reactions of volatile
27 polycyclic aromatic hydrocarbons with OH radical and N2OS. Int J Chem Kinet 21:775-799.
28
29 Atkinson, R. (1986) Kinetics and mechanisms of the gas-phase reactions of the hydroxyl radical with organic
30 compounds under atmospheric conditions. Chem Rev 86:69-201.
31
32 Atkinson, R. (1988) Atmospheric transformation of automotive emissions. In: Air pollution, the automobile and
33 public health. Watson, AY; Bates, RR; Kennedy,D; eds. Washington, DC: National Academy Press, pp. 99-132.
34
35 Atkinson, R. (1990) Gas-phase tropospheric chemistry of organic compounds: a review. Atmos Environ 24A:1-41.
36
37 Atkinson, R. (1991) Kinetics and mechanisms of the gas-phase reactions of the NO3 radial with organic compounds.
38 JPhys Chem Ref Data 20:459-507.
39
40 Atkinson, R; Arey, J. (1994) Atmospheric chemistry of gas-phase polycyclic aromatic hydrocarbons: formation of
41 atmospheric mutagen. Environ Health Perspect 102(Suppl. 4): 117-126.
42
43 Atkinson, R; Arey, J; Zielinska, B; et al. (1987) Kinetics and products of the gas-phase reactions of OH radicals and
^A N2O5 with naphthalene and biphenyl. Environ Sci Technol 21:1014-1022.
11/5/99 2-88 DRAFT—DO NOT CITE OR QUOTE
-------
1 Atkinson, R; Arey, J; Winer, AM; et al. ( 1 988) A survey of ambient concentrations of selected polycyclic aromatic
2 hydrocarbons (PAH) at various locations in California. Final report prepared under Contract No. A5-185-32 for
3 California Air Resources Board, Sacramento, CA.
4
5 Atkinson, R; Arey, J; Zielinska, B; et al. (1990) Kinetics and nitro-products of the gas-phase OH and NO3 radical-
6 initiated reactions of naphthalene-dg fluoranthene-d|0 and pyrene. Int J Chem Kinet 22:999-1014.
7
8 Bagley, ST; Gratz, LD; Leddy, DG; et al. (1993) Characterization of particle and vapor phase organic fraction
9 emissions from a HD diesel engine equipped with a particle trap and regeneration controls. Health Effects Institute,
1 0 Research Report No. 56. Cambridge, MA: HEI.
11
1 2 Bagley, ST; Baumgard, KJ; Gratz, LD; et al. (1996) Characterization of fuel and aftertreatment effects on diesel
1 3 emissions. HEI Report No. 76. Cambridge, MA: HEI.
14
1 5 Bagley, ST; Gratz, LD; Johnson, JH; et al. (1998) Effects of an oxidation catalytic converter and a biodiesel fuel on
1 6 the chemical, mutagenic, and particle size characteristics of emissions from a diesel engine. Environ Sci Technol
17 32:1183-1191.
18
1 9 Baines, TM; Somers, JH; Harvey, CA. (1979) Heavy duty diesel particulate emission factors. J Air Pollut Control
20 Assoc 29(6):6 16-621.
21
22 Baranescu, RA. (1988) Influence of fuel sulfur on diesel particulate emissions. SAE Technical Paper Ser. No.
23 881174.
24
25 Bata, R; Wang, W; Gautam, M; et al. (1992) Fleet-site measurements of exhaust gas emissions from urban buses.
2 6 ASME, ICE. New Dev Off-Highway Engines 1 8: 1 85- 1 96.
27
28 Baumgard, KJ; Johnson, JH. (1992) The effect of low sulfur fuel and a ceramic particle filter on diesel exhaust
29 particle size distributions. SAE Technical Paper Ser. No. 920566.
30
31 Baumgard, KJ; Johnson, JH. (1996) The effect of fuel and engine design on diesel exhaust particle size
3 2 distributions. SAE Technical Paper Ser. No. 960 1 3 1 .
33
34 Behyiner, TD; Hites, RA. (1985) Photolysis of polycydic aromatic hydrocarbons adsorbed on simulated
OO '
36
37 Behymer, TD; Hites, RA. (1988) Photolysis of polycyclic aromatic hydrocarbons adsorbed on fly ash. Environ Sci
38 Techno! 22:13 11-13 19.
39
40 Birch, ME; Gary, RA. (1996) Elemental carbon-based method for monitoring occupational exposures to particulate
4 1 diese! exhaust. Aerosol Sci Techno! 25 :2? ) -?4 1
42
43 Bishop, GA; McLaren, SE; Stedman, DH; et al. (1996) Method comparisons of vehicle emissions measurements in
44 the Fort McHenry and Tuscarora Mountain Tunnels. Atmos Environ 30:2307.
45
1 1 75/99 2-89 DRAFT-DO NOT CITE OR QUOTE
-------
Brodowicz, P. (1999) Determination of the demographic groups with the highest annual average modeled diesel
PM exposure. U.S. Environmental Protection Agency, National Fuel and Vehicle Emissions Laboratory
memorandum.
4
5 Brown, KF. (1998) Diesel aftertreatment technologies. In: Proceedings, SAE Topical Technical Seminar Diesel
6 Technology for the New Millennium. Schaumburg, IL, April 22, 1998.
7
8 Brown, KF; Hideout, GR; Turner, JE. (1996) Urban driving cycle results of retrofitted diesel oxidation catalysts on
9 heavy duty vehicles. SAE Technical Paper Ser. No. 960134.
10
11 Brown, KF; Hideout, GR; Turner, JE. (1997) Urban driving cycle results of retrofitted diesel oxidation catalysts on
12 heavy duty vehicles: one year later. SAE Technical Paper Ser. No. 970186.
13
14 Brunekreef, B. (1999) Environmental diesel exhaust exposure and respiratory health of children in the Netherlands.
15 Presented at the Health Effects Institute Diesel Workshop: Building a Research Strategy to Improve Risk
16 Assessment. Stone Mountain, GA, March 7-9, 1999.
17
18 Bunn, W. (1999) An analysis of the suitability of Steenland et al. for conducting quantitative risk assessment of
19 diesel exhaust. Navistar International Transportation Corp.
20
21 Cadle, SH; Mulawa, PA. (1990) Atmospheric carbonaceous species measurement methods comparison study:
22 General Motors results. Aerosol Sci Technol 12(1):128-141.
California Environmental Protection Agency. (1998a) Report to the Air Resources Board on the proposed
25 identification of diesel exhaust as a toxic air contaminant. Appendix HI, Part A: Exposure assessment. April 1998.
26
27 California Environmental Protection Agency. (1998b) Measuring concentrations of selected air pollutants inside
28 California vehicles. Final Report.
29
30 Calvert, JG; Stockwell, WR. (1983) Acid generation in the troposphere by gas-phase chemistry. Environ Sci Technol
31 17:428A-443A.
32
33 Campbell, RM; Lee, ML. (1984) Capillary column gas chromatographic determination of nitro polycyclic aromatic
34 compounds in paniculate extracts. Anal Chem 56:1026-1030.
35
36 Cantrell, BK; Salas, LJ; Johnson, WB; et al.(1988) Phase distributions of low volatility organics in ambient air.
37 U.S. EPA Contract 68-02-3748.
38
39 Carey, PM. (1987) Air toxics emissions from motor vehicles. EPA Office of Mobile Sources, Report No.
40 EPA/AA-TSS-PA-86-5.
41
42 Cass, GR; Gray, HA. (1995) Regional emissions and atmospheric concentrations of diesel engine paniculate matter:
43 Los Angeles as a case study. In: Diesel exhaust: a critical analysis of emissions, exposure, and health effects. A
^A Special Report of the Institute's Diesel Working Group. Cambridge, MA: Health Effects Institute, pp. 125-137.
11/5/99 2-90 DRAFT-DO NOT CITE OR QUOTE
-------
1 Cautreels, W; Van Cauwenberghe, K. (1978) Experiments on the distribution of organic pollutants between airborne
2 paniculate matter and the corresponding gas phase. Atmos Environ 12:1133-1141.
3
4 Chow, JC; Watson, JG; Richards, LW; et al. (1991) The 1989-1990 Phoenix PM10 study. Volume II: Source
5 apportionment. Final Report. DRI Document No. 8931.6F1, prepared for Arizona Department of Environmental Air
6 Quality, Phoenix, AZ, by Desert Research Institute, Reno, NV.
7
8 Ciccioli, P; Brancaleoni, E; Cecinato, A; et al. (1986) Fractionation of polar polynuclear aromatic hydrocarbons
9 present in _ industrial emissions and atmospheric samples and their determination by gas chromatograph-mass
10 spectrometry. J Chromatogr 351:451 -464.
11
12 Ciccioli, P; Cecinato, A; Brancaleoni, E; et al. (1989) Evaluation of nitrated PAH in anthropogenic emission and air
13 samples: a possible means of detecting reactions of carbonaceous particles in the atmosphere. Aerosol Sci Technol
14 10:296-310.
15
16 Ciccioli, P; Cecinato, A; Cabella, R; et al. (1993) The contribution of gas-phase reactions to the nitroarene fraction
17 of molecular weight 246 present in carbon particles sampled in an urban area of northern Italy. Atmos Environ
18 27A:1261.
19
20 Clark, N; Messer, JT; McKain, DL; et al. (1995) Use of the West Virginia University truck test cycle to evaluate
21 emissions from Class 8 trucks. SAE Technical Paper Ser. No. 951016.
22
23 Clark, N; Gautam, M; Lyons, D; et al. (1997) Natural gas and diesel transit bus emissions: review and recent data.
24 SAE Technical Paper Ser. No. 973203.
25
26 Clary, WB. (1989) The effect on diluted diesel paniculate emissions of two mining water scrubbers. MS Thesis,
2 7 University of Minnesota.
28
29 CFR, Code of Federal Regulations, 40 CFR Subpart N; Vol. 40.
30
31 CFR, Code of Federal Regulations, Title 40, Section 80.29.
32
33 Cohen, LH; Countess, RJ; Countess, SJ. (1997) Advanced remote sensing technology demonstration. Mobile
34 Sources Air Pollution Review Committee.
35
36 Coifrnan, B. (1994) Evolution of the diesel locomotive in the United States.
37 http://www.cyberus.ca/~yardlimit/guide/locopaper.html
38
39 Coordinating Research Council. (1987) Chemical methods for the measurement of unregulated diesel emissions.
40 CRC-APRAC Project No. CAPI-1 -64, August 1987.
41
42 Countess. R: Countess, S; Cohen, R. (1999) Remote sensing of heavy duty diesel trucks revisited. 9th CRC On-
43 Road Vehicle Emissions Workshop: San Diego, CA.
44
1175/99 2-91 DRAFT—DO NOT CITE OR QUOTE
-------
1 Dec, JE; Espey, C. (1995) Ignition and early soot formation in a diesel engine using multiple 2-D imaging
2 diagnostics. SAE Technical Paper Ser. No. 950456.
4 Dickson, CL; Sturm, GP. (1994) Diesel fuel oils, 1994. National Institute for Petroleum and Energy Research.
5 NIPER-187PPS94/5.
6
7 Dietzmann, HE; Pamess, MA; Bradow, RL. (1980) Emissions from trucks by chassis version of 1983 transient
8 procedure. SAE Technical Paper Ser. No. 801371.
9
10 Dunlap, LS; Pellegrin, V; Ikeda, R; et al. (1993) Chassis dynamometer emissions testing results for diesel and
11 alternative-fueled transit buses. SAE Technical Paper Ser. No. 931783.
12
13 Dunstan, TDJ; Mauldin, RF; Jinxian, Z; et al. (1989) Adsorption and photodegradation of pyrene on magnetic,
14 carbonaceous, and mineral subtractions of coal stack ash. Environ Sci Technol 23:303-308.
13
16 Engler, BH. (1998) Catalytic reduction of NO,, with hydrocarbons under lean diesel exhaust conditions. SAE
17 Technical Paper Ser. No. 980189.
18
19 Enya, T; Suzuki, H; Watanabe, T; et al. (1997) 3-Nitrobenzanthrone, a powerful bacterial mutagen and suspected
20 human carcinogen found in diesel exhaust and airborne particles. Environ Sci Technol 31 -.2772-2716.
21
22 Farrauto, RJ; Voss, KE. (1996) Monolithic diesel oxidation catalysts. Appl Catal B: Environ 10:29-51.
23
^H Federal Highway Administration. (1995) Motor fuel use, 1919-1995. U.S. Department of Transportation:
2 5 www.fhwa.dot.gov/ohim/Summary95/section 1 .html; Vol. MF221.
26
27 Federal Highway Administration. (1997a) Highway statistics summary to 1995. U.S. Department of Transportation.
28 FHWA-PL-97-009.
29
30 Federal Highway Administration. (1997b) Assessment of highway paniculate impacts: phase I, tasks B and C. Field
31 sampling and evaluation program. Prepared by E.H. Pechan & Associates Inc; Durham, NC.
32
33 Feilberg, A; Kamens, RM; Strommen, MR; et al. (1999) Modeling the formation, decay, and partitioning of
34 semivolatile nitro-polycyclic aromatic hydrocarbons (nitronaphthatenes) in the atmosphere. Atmos Environ
35 33:1231-1243.
36
37 Ferguson, DH; Gautam, M; Wang, WG; et al. (1992) Exhaust emissions from in-use heavy duty vehicles tested on a
38 transportable transient chassis dynamometer. SAE Technical Paper Ser. No. 922436.
39
40 Finlayson-Pitts, BJ; Pitts, JN, Jr. (1986) Atmospheric chemistry: fundamentals and experimental techniques. New
41 York: John Wiley & Sons.
42
43 Foster, DE; Tree, DR. (1994) Optical measurements of soot particle size, number density and temperature in a
^4 direct injection diesel engine as a function of speed and load. SAE Technical Paper Ser. No. 940270.
11/5/99 2-92 DRAFT-DO NOT CITE OR QUOTE
-------
1 Frey, JW; Corn, M. (1967) Diesel exhaust participates. Nature 216:615-616.
2
3 Fritz, SG.( 1995) Emission measurements-locomotives. Southwest Research Institute. EPA Contract 68-C2-0144.
4
5 Fujita, EM; Lu, Z; Sheetz, L; et al. (1997) Determination of mobile source emission source fraction using ambient
6 field measurements. Final report, prepared under CRC Project No. E-5-1 for Coordinating Research Council,
7 Atlanta, GA, by Desert Research Institute, Reno, NV.
8
9 Gautam, M; Ferguson, D; Wang, WG; et al. (1992) In-use emissions and performance monitoring of heavy duty
10 vehicles using a transportable transient chassis test facility. SAE Technical Paper Ser. No. 921751.
11
12 Gertler, AW; Sagebiel, JC; Dippel, WA; et al. (1998) Measurements of dioxin and furan emission factors from HD
13 diesel vehicles. J Air Waste Manage Assoc 48:276-278.
14
15 Gorse, RA, Jr.; Riley, TL; Ferris, FC; et al. (1983) 1-Nitropyrene concentration and bacterial mutagenicity in on-
16 road vehicle paniculate emissions. Environ Sci Technol 17:198-202.
17
'18 Graboski, MS; McCormick, RL; Allemanj T. (1998a) Testing of natural gas and diesel buses for comparison with
19 WVU mobile dynamometer. Colorado Institute for Fuels and High Altitude Engine Research, Colorado School of
20 Mines.
21
22 Graboski, MS; McCormick, RL; Yanowitz, J; et al. (1998b) HD diesel testing for the Northern Front Range Air
23 Quality Study. Colorado Institute for Fuels and Engine Research.
24
25 Gray, HA. (1986) Control of atmospheric fine primary carbon particle concentrations. Ph.D. thesis, California
26 Institute of Technology.
27
28 Gray, HA; Kuklin, A. (1996) Benefits of mobile source NOX related paticulate matter reductions. Systems
29 Applications International. Final report prepared for U.S. Environmental Protection Agency. SYSAPP-96/61.
30
31 Gray, HA; Cass, GR. (1998) Source contributions to atmospheric fine carbon particle concentrations. Atmos
32 Environ 32(22):3805-3825.
33
34 Grosovsky, AJ; Sasaki, JC; Arey, J; et a!. (1999) Evaluation of the potential health effects of the atmospheric
35 reaction products of polycyclic aromatic hydrocarbons. Health Effects Institute 84.
36
37 Hames, RJ; Straub, RD; Amann, RW. (1985) DDEC Detroit Diesel electronic control. SAE Technical Paper Series
38 No. 850542.
39
40 Hammerle, RH; Schuetzle, D; Adams, W. (1994) A perspective on the potential development of environmentally
41 acceptable light-duty diese! engines. Environ Health Perspect (Supp!.) 102:25-30.
42
43 Hansen, ADA; Novakov, T. (1990) Real-time measurement of aerosol black carbon during the carbonaceous
44 species methods comparison study. Aerosol Sci Technol 12(1):194-199.
45
11/5/99 2-93 DRAFT-DO NOT CITE OR QUOTE
-------
Hansen, KF; Bak, F; Andersen, EM; et al. (1994) The influence of an oxidation catalytic converter on the chemical
and biological characteristics of diesel exhaust emissions. SAE Technical Paper Ser. No. 940241.
4 Hardenberg, H; Fraenkle, G. (1978) The effect of charge air cooling on exhaust emissions and power output of
5 turbocharged engines. Inst Mech Engineers Paper No. C71/78.
6
7 Hare, CT; Springer, KJ. (1972) Exhaust emissions from uncontrolled vehicles and related equipment using internal
8 combustion engines. Part I, locomotive diesel engines and marine counterparts. Southwest Research Institute.
9 Contract No. EHS 70-108.
10
11 Harvey, GD; Baumgard, KJ; Johnson, JH; et al. (1994) Effects of a ceramic particle trap and copper fuel additive on
12 HD diesel emissions. SAE Technical Paper Ser. No. 942068.
13
14 Hawker, P; Meyers, N; Huethwohl, G; et al. (1993) Experience with a new particulate trap technology in Europe.
15 SAE Technical Paper Ser. No. 930735.
16
17 Hawker, P; Myers, N; Huethwohl, G; et al. (1997) Experience with a new particulate trap technology in Europe.
18 SAE Technical Paper Ser. No. 970182.
19
20 Hawker, P; Huethwohl, G; Henn, J; et al. (1998) Effect of a continuously regenerating diesel particulate filter on
21 non-regulated emissions and particle size distribution. SAE Technical Paper Ser. No. 980189.
22
Health Effects Institute. (1995) Diesel exhaust: a critical analysis of emissions, exposure, and health effects.
25 Heywood, JB. Internal combustion engine fundamentals. New York: McGraw-Hill, Inc., 1988.
26
27 Horvath, H; Kreiner, I; Norek, C; et al. (1988) Diesel emissions in Vienna. Atmos Environ 22:1255-1269.
28
29 Jang, M; McDow, SR. (1995) Benz[a]anthracene photodegradation in the presence of known organic constituents of
30 atmospheric aerosols. Environ Sci Technol 29:2654.
31
32 Japar, SM; Szkarlat, AC; Gorse, RA; et al. (1984) Comparison of solvent extraction and thermal-optical carbon
33 analysis methods: application to diesel vehicle exhaust aerosol. Environ Sci Technol 18:231-234.
34
35 Jimenez, J; McRae, G; Nelson, D; et al. (1998) Remote sensing of HD diesel truck NOX emissions using TILDAS.
36 8th CRC On-Road Vehicle Emissions Workshop: San Diego, CA, pp. 7-45 - 7-61.
37
38 Johnson, JE. (1993) Hydrocarbon oxidation in a diesel oxidation catalytic converter. University of Minnesota MS
39 Thesis.
40
41 Johnson, JE; Kittelson, DB. (1996) Deposition, diffusion and adsorption in the diesel oxidation catalyst. Appl Catal
42 B: Environ 10:117-137.
43
• Johnson, JH; Bagley, ST; Gratz, LD; et al. (1994) A review of diesel particulate control technology and emissions
effects-1992 Homing Memorial Award Lecture. SAE Technical Paper Ser. No. 940233.
11/5/99 2-94 DRAFT-DO NOT CITE OR QUOTE
-------
1 Kakoi, Y; Tsutsui, Y; Ono, N; et al. (1998) Emission reduction technologies applied to high-speed direct injection
2 diesel engine. SAE Technical Paper Ser. No. 980173.
3
4 Kamens, RM; Guo, Z; Fulcher, JN; et al. (1988) Influence of humidity, sunlight, and temperature on the daytime
5 decay of polyaromatic hydrocarbons on atmospheric soot particles. Environ Sci Technol 22:103-108.
6
7 Kantola, TC; Bagley, ST; Gratz, LD; et al. (1992) The influence of a low sulfur fuel and a ceramic particle trap on
8 the physical, chemical, and biological character of HD diesel emissions. SAE Technical Paper Ser. No. 920565.
9
10 Katragadda, S; Bata, R; Wang, WG; et al. (1993) A correlation study between two HD vehicle chassis dynamometer
11 emissions testing facilities. SAE Technical Paper Ser. No. 931788.
12
13 Katz, M; Chan, C; Tosine, H; et al. (1979) Relative rates of photochemical and biological oxidation (in vitro) of
14 polycyclic aromatic hydrocarbons. In: Polynuclear aromatic hydrocarbons: Third International Symposium on
15 Chemistry and Biology - Carcinogenesis and Mutagenesis, October 1978, Columbus, OH. Jones, PW; Leber, P. eds.
16 Ann Arbor, MI: Ann Arbor Science Publishers, Inc, pp. 171 -189.
17
18 Khatri, NJ; Johnson, JH; Leddy, DG. (1978) The characterization of the hydrocarbon and sulfate fractions of diesel
19 paniculate matter. SAE Technical Paper Ser. No. 780111.
20
21 Kirchstetter, TW; Harley, RA; Kreisberg, NM; et al. (1999) On-road measurement of fine particle and nitrogen
22 oxide emissions from light- and HD motor vehicles. Atmos Environ 33:2955-2968.
23
24 Kirkland, JF. (1983) Dawn of the diesel age. Glendale, CA: Interurban Press.
25
26 Kittelson, DB. (1998) Engines and nanoparticles: a review. J Aerosol Sci 29:575-588.
27
28 Kittelson, DB; Pipho, ML; Ambs, JL; et al. (1986) Particle concentrations in a diesel cylinder: comparison of theory
29 and experiment. SAE Technical Paper Ser. No. 861569.
30
31 Kittelson, DB; Winthrop, W; Baltensperger, R; et al. (1999) Diesel aerosol sampling methodology. University of
32 Minnesota Center for Diesel Research.
33
34 Kleeman, MJ; Cass, GR. (1998) Source contributions to the size and composition distribution of urban paniculate
35 air pollution. Atmos Environ 32(16):2803-2816.
36
37 Klein, M. (1991) The diesel revolution. Am Heritage Invention Technol 6.
38 ,
39 Kleindienst, TE; Shepson, PB; Edney, EO; et al. (1985) Peroxyacetyl nitrate: measurement of its mutagenic activity
40 using the Salmonella/mammalian microsome reversion assay. Mutat Res 157:123-128.
41
42 Kreso. AM: Johnson. JH; Gratz, LD; et al. (1998a) A study of the vapor- and particle-phase sulfur species in the HD
43 diesel engine EGR cooler. SAE Technical Paper Ser. No. 981423.
44
1175/99 2-95 DRAFT-DO NOT CITE OR QUOTE
-------
Kreso, AM; Johnson, JH; Gratz, LD; et al. (1998b) A study of the effects of exhaust gas recirculation on HD diesel
engine emissions. SAE Technical Paper Ser. No. 981422.
4 Ladomatos, N; Abdelhalim, SM; Zhao, H; et al. (1996-1997) The dilution, chemical, and thermal effects of exhaust
5 gas recirculation on diesel engine emissions-Parts 1,2, 3, and 4. SAE Technical Paper Series.
6
7 Last, RJ; Krueger, M; Duemholz, M. (1995) Emissions and performance characteristics of a 4-stroke, direct injected
8 diesel engine fueled with blends of biodiesel and low sulfur diesel fuel. SAE Technical Paper Ser. No. 950054.
9
10 Law, PL; Lioy, PJ; Zelenka, MP; et al. (1997) Evaluation of a probabilistic exposure model applied to carbon
11 monoxide (pNEM/CO) using Denver personal exposure monitoring data. J Air Waste Manage Assoc 47:491-500.
12
13 Lee, R; Hobbs, CH; Pedley, JF. (1998) Fuel quality impact on heavy duty diesel emissions: literature review. SAE
14 Technical Paper Ser. No. 982649.
15
16 Leet, JA; Friesen, T; Shadboume, A. (1998) EGR's effect on oil degradation and intake system performance. SAE
17 Technical Paper Ser. No. 980179.
18
19 Levson, K. (1988) The analysis of diesel paniculate. Fresenius Z Anal Chem 331:467-478.
20
21 Lies, KH; Hartung, A; Postulka, A; et al. (1986) Composition of diesel exhaust with particular reference to particle
22 bound orgahics including formation of artifacts. In: Carcinogenic and mutagenic effects of diesel engine exhaust.
^3 Elsevier Science Limited, U.K. pp. 65-82.
25 Ligocki, MP; Pankow, JF. (1989) Measurements of the gas/particle distributions of atmospheric organic
26 compounds. Environ Sci Technol 23:75-83.
27
28 Lowenthal, DH; Chow, JC; Watson, JG; et al. (1992) The effects of collinearity on the ability to determine aerosol
29 contributions from diesel- and gasoline-powered vehicles using the chemical mass balance model. Atmos Environ
30 26A(13):2341-2351.
31
32 Luders, H; Stommel, P; Backes, R. (1997) Applications for the regeneration of diesel paniculate traps by combining
33 different regeneration systems. SAE Technical Paper Ser. No. 970470.
34
35 Mauderly, J. (1992) Diesel exhaust. In: Environmental toxicants: human exposures and their health effects.
36 Lippmann, M, ed. New York: Van Nostrand Reinhold, pp. 119-162.
37
38 McCarthy, CI; Slodowske, WJ; Sienicki, EJ; et al. (1992) Diesel fuel property effects on exhaust emissions from a
39 heavy duty diesel engine that meets 1994 emissions requirements. SAE Technical Paper Ser. No. 922267.
40
41 McClure, BT; Bagley ST; Gratz, LD. (1992) The influence of an oxidation catalytic converter and the fuel
42 composition on the chemical and biological characteristics of diesel exhaust emissions. SAE Technical Paper Ser.
43 No. 920854.
1175/99 2-96 DRAFT-DO NOT CITE OR QUOTE
-------
1 McCormick, RL; Ryan, LBA; Daniels, TL; et al. (1998) Comparison of chassis dynamometer in-use emissions with
2 engine dynamometer FTP emissions for three HD diesel vehicles. SAE Technical Paper Ser No. 982653.
3
4 McCormick, RL; Graboski, MS; Alleman, TL; et al. (1999) In-use emissions from natural gas fueled HD vehicles.
5 SAE Technical Paper Ser. No. 1999-01-1507.
6
7 McDonald, J; Purcell, DL; McClure, BT. (1995) Methyl ester oxygenated fuels for diesel mining applications. In:
8 Proceedings of the 7th U.S. Mine Ventilation Symposium, Society for Mining, Metallurgy, and Exploration, Inc.
9
10 McDonald, JF. (1997) The emissions and combustion characteristics of a soy-methyl-ester biodiesel fuel in a
11 naturally aspirated indirect injection diesel engine. MS thesis, University of Minnesota.
12
13 McDow, SR; Sun, Q; Vartiainen, M; et al. (1994) Effect of composition and state of organic components on
14 polycyclic aromatic hydrocarbon decay in atmospheric aerosols. Environ Sci Technol 28:2147.
15
16 McGuckin, J; Rykowski, R. (1981) Correlation between smoke and paniculate emissions from HD diesel engines.
17 U.S. EPA, Office of Mobile Sources Technical Report. Washington, DC: U.S. Environmental Protection Agency.
18
19 McKinley, TL. (1997) Modeling sulfuric acid condensation in diesel engine EGR coolers. SAE Technical Paper
20 Ser. No. 970636.
21
22 Miguel, AH; Kirchstetter, TW; Harley, RA; et al. (1998) On-road emissions of particulate polycyclic aromatic
23 hydrocarbons and black carbon from gasoline and diesel vehicles. Environ Sci Technol 32:450-455.
24
25 Mitchell, K; Steere, DE; Taylor, JA; et al. (1994) Impact of diesel fuel aromatics on particulate, PAH, and
26 nitro-PAH emissions. SAE Technical Paper Ser. No. 942053.
27
28 Mori, K. (1997) Worldwide trends in HD diesel engine exhaust emission legislation and compliance technologies.
29 SAE Technical Paper Ser. No. 970753.
30
31 Nielsen, T; Ramdahl, T. (1986) Discussion on determination of 2-nitrofluoranthene and 2-nitropyrene in ambient
32 particulate matter: evidence for atmospheric reactions. Atmos Environ 20:1507.
33
34 NIPER, National Institute for Petroleum and Energy Research. (198 6 and earlier years) Diesel fuel oils, annual
3s survey.
36
37 Noorkhoek, J; Bos, RP. (1995) 1-Nitropyrene as a marker for the mutagenicity of diesel exhaust derived particulate
38 matter in workplace atmospheres. Environ Mol Mutagen 25:134.
39
40 Norbeck, JM; Truex, TJ; Smith, MR; et al. (1998) Evaluation of factors that affect diesel exhaust toxicity. Final
41 report, prepared under Contract No. 94-312, for California Air Resources Board, by the Center for Environmental
42 Research and Technology, College of Engineering, University of California, Riverside, CA.
43
1115199 2-97 DRAFT—DO NOT CITE OR QUOTE
-------
1 NRC, National Research Council. (1998) Research priorities for airborne paniculate matter I: immediate priorities
2 and a long-range research portfolio. Northern Front Range Air Quality Study, Colorado. January 1998, Volume I.
Washington, DC: National Academy Press.
5 Odum, JR; McDow, SR; Kamens, RM. (1994) Mechanistic and kinetic studies of the photodegradation of
6 benz[a]anthracene in the presence of methoxyphenols. Environ Sci Technol 28:1285.
7
8 Opris, CN; Gratz, LD; Bagley, ST; et al. (1993) The effects of fuel sulfur concentration on regulated and
9 unregulated HD diesel emissions. SAE Technical Paper Ser. No. 930730.
10
11 Paputa-Peck, MC; Marano, RS; Schuetzle, D; et al. (1983) Determination of nitrated polynuclear aromatic
12 hydrocarbons in paniculate extracts by capillary column gas chromatography with nitrogen selective detection. Anal
13 Chem 55:1946-1954.
14
15 Pataky, GM; Baumgard, KJ; Gratz, LD; et al. (1994) Effects of an oxidation catalytic converter on regulated and
16 unregulated diesel emissions. SAE Technical Paper Ser. No. 940243.
17
18 Peters, A; Langer, H-J; Joki, B; et al. (1998) Catalytic NOX reduction on a passenger car diesel common rail engine.
19 SAE Technical Paper Ser. No. 980191.
20
21 Pierson, WR; Brachaczek, WW. (1976) Particulate matter associated with vehicles on the road. SAE Trans 85:209-
22 227.
23
f Pierson, WR; Brachaczek, WW. (1983) Particulate matter associated with vehicles on the road. Aerosol Sci Tech
2:1-40.
26
27 Pierson, WR; Gertler, AW; Robinson, NF; et al. (1996) Real-world automotive emissions-summary of studies in the
28 Fort McHenry and Tuscarora mountain tunnels. Atmos Environ 30:2233-2256.
29
30 Pischinger, R; Cartellieri, W. (1972) Combustion system parameters and their effect on diesel engine exhaust
31 emissions. SAE Technical Paper Ser. No. 720756.
32
33 Pitts, JN, Jr; Van Cauwenberghe, KA; Grosjean, D; et al. (1978) Atmospheric reactions of polycyclic aromatic
34 hydrocarbons: facile formation of mutagenic nitro derivatives. Science 202:515-519.
35
36 Pitts, JN, Jr; Lokensgard, DM; Ripley, PS; et al. (1980) Atmospheric epoxidation of benzofa]pyrene by ozone:
37 formation of the metabolite benzo[a]pyrene-4,5-oxide. Science 210:1347-1349.
38
39 Pitts, JN, Jr; Sweetman, JA; Zielinska, B; et al. (1985a) Determination of 2-nitrofluoranthene and 2-nitropyrene in
40 ambient paniculate matter: evidence for atmospheric reaction. Atmos Environ 19:1601-1608.
41
42 Pitts, JN, Jr; Zielinska, B; Sweetman, JA; et al. (1985b) Reaction of adsorbed pyrene and perylene with gaseous
43 N2O5 under simulated atmospheric conditions. Atmos Environ 19:911-915.
44
1175/99 2-98 DRAFT-DO NOT CITE OR QUOTE
-------
1 Pitts, JN; Jr; Sweetman, JA; Zielinska, B; et al. (1985c) Formation of nitroarenes from the reaction of polycyclic
2 aromatic hydrocarbons with dinitrogen pentoxide. Environ Sci Technol 19:1115-1121.
3
4 Pitts, JN, Jr; Paur, HR; Zielinska, B; et al. (1986) Factors influencing the reactivity of polycyclic aromatic
5 hydrocarbons adsorbed on filters and ambient POM with ozone. Chemosphere 15:675-685.
6
7 Prinn, R; Cunnold, D; Simmonds, P; et al. (1992) Global average concentration and trend for hydroxyl radicals
8 deduced from ALE/GAGE trichloroethane (methyl chloroform data for 1978-1990). J Geophys Res 97:2445-2461.
9
10 Rantanen, L; Mikkonen, S; Nylund, L; et al. (1993) Effect of fuel on the regulated, unregulated and mutagenic
11 emissions of DI diesel engines. SAE Technical Paper Ser. No. 932686.
12
13 Rasmussen, RE. (1990) Effect of fuel properties on mutagenic activity in extracts of HD diesel exhaust particulate.
14 J. Air Waste Manage Assoc 40:1391-1396.
15
16 Rideout, G; Kirshenblatt, M; Prakash, C. (1994) Emissions form methanol, ethanol, and diesel powered urban
17 transit buses. SAE Technical Paper Ser. No. 942261.
18
19 Rogak, SN; Pott, U; Dann, T; et al. (1998) Gaseous emissions from vehicles in a traffic tunnel in Vancouver, British
20 Columbia. J Air Waste Manage Assoc 48:604-615.
21
22 Rogge, WF; Hildemann, LM; Mazurek, MA; et al. (1993) Sources of fine organic aerosol. 2. Noncatalyst and
23 catalyst-equipped automobiles and HD diesel trucks. Environ Sci Technol 27:636-651.
24
25 Ross, MM; Risby, TH; Steele, WA; et al. (1982) Physicochemical properties of diesel particulate matter. Colloids
26 Surf5:17-31.
27
28 Salmon, LG; Cass, GR; Pedersen, DU; et al. (1997) Determination of fine particle concentration and chemical
29 composition in the northeastern United States, 1995. Progress Report to Northeast States for Coordinated Air Use
30 Management (NESCAUM), October 1997.
31
32 Sampson, PJ. (1988) Atmospheric transport and dispersion of air pollutants associated with vehicular emissions. In:
33 Air pollution, the automobile and public health. Watson, AY; Bates, RR; Kennedy, D, eds. Washington, DC:
34 National Academy Press, pp. 77-97.
35
36 Sawyer, RF; Johnson, JH. (1995) Diesel emissions and control technology. In: Diesel exhaust: a critical analysis of
37 emissions, exposure, and health effects. Cambridge, MA: Health Effects Institute, pp. 65-81.
38
39 Schauer, JJ; Rogge, WF; Hildemann. LM; et al. (1996) Source apportionment of airborne particulate matter using
40 organic compounds as tracers. Atmos Environ 30(22):3837-3855.
41
42 S^aner, T.I; Kleeman. MJ; Cass. GR; et al. (1999) Measurement of emissions from air pollution sources. 2. C,
43 through C30 organic compounds from medium duty diesel trucks. Environ Sci Technol 33:1578-1587.
44
11/5/99 2-99 DRAFT-DO NOT CITE OR QUOTE
-------
^.
f
Schuetzle, D. (1983) Sampling of vehicle emissions of chemical analysis and biological testing. Environ Health
Perspect 47:65-80.
Schuetzle, D; Frazier, JA. (1986) Factors influencing the emission of vapor and paniculate phase components from
5 diesel engines. In: Carcinogenic and mutagenic effects of diesel engine exhaust. Elsevier Science Limited, U.K. pp.
6 41-63.
7
8 Schuetzle, D; Lee, FS-C; Prater, TJ. (1981) The identification of polynuclear aromatic hydrocarbon (PAH)
9 derivatives in mutagenic fractions of diesel particulate. Int J Environ Anal Chem 9:93-144.
10
11 Schuetzle, D; Jensen, TE; Ball, JC. (1985) Polar polynuclear aromatic hydrocarbon derivatives in extracts of
12 particulates: biological characterization and techniques for chemical analysis. Environ Int 11:169-181.
13
14 Seinfeld, JH; Pandis, SN. (1998) Atmospheric chemistry and physics: from air pollution to climate change. New
15 York: John Wiley & Sons, Inc.
16
17 Sisler, JF. (1996) Spatial and seasonal patterns and long term variability of the composition of the haze in the United
18 States: an analysis of data from the IMPROVE network. Cooperative Institute for Research in the Atmosphere.
19 Colorado State University. ISSN: 0737-5352-32.
20
21 South Coast Air Quality Management District. (SCAQMD) (1999) "Hot-spot" monitoring and multiple air toxics
22 exposure study (MATES II). http://www.aqmd.gov/news/mates.html. Accessed on July 20. 1999.
23
f Springer, KJ. (1979) Characterization of sulfates, odor, smoke, POM and particulates from light and heavy duty
engines-part IX. Southwest Research Institute. EPA/460/3-79-007.
26
27 Springer, KJ; White, JT; Domke, CJ. (1974) Emissions from in-use 1970-1971 diesel-powered trucks and buses.
28 SAE Technical Paper Ser. No. 741006.
29
30 Stone, R. (1995) Introduction to internal combustion engines. Warrendale, PA: Society of Automotive Engineers.
31
32 Sweetman, JA; Zielinska, B; Atkinson, R; et al.(1986) A possible formation pathway for the 2-nitrofluoranthene
33 observed in ambient particulate organic matter. Atmos Environ 20:235-238.
34
35 Tamanouchi, M; Morihisa, H; Araki, H; et al. (1998) Effects of fuel properties and oxidation catalyst on exhaust
36 emissions for heavy duty diesel engines and diesel passenger cars. SAE Technical Paper Ser. No. 980530.
37
38 Tanaka, S; Takizawa, H; Shimizu, T; et al. (1998) Effect of fuel compositions on PAH in particulate matter from DI
39 diesel engines. SAE Technical Paper Ser. No. 982648.
40
41 Taylor, CF. (1990) The internal combustion engine in theory and practice - volume 1. Cambridge, MA: MIT Press.
42
43 U.S. Bureau of the Census. (1995) Truck inventory and use survey. 1992 Census of Transportation, TC92-T-52.
44
U.S. EPA. (1998a) National air pollutant emission trends update, 1970-1997. EPA/454/E-98-007.
11 /5/99 2-100 DRAFT-DO NOT CITE OR QUOTE
-------
1 U.S. EPA. (1998b) Locomotive emission standards. Office of Mobile Sources Regulatory Support Document.
2
3 U.S. EPA. (1999a) Fleet characterization data for MOBILE6: development and use of age distributions, average
4 annual mileage accumulation rates and projected vehicle counts for use in MOBILE6. EPA/420/P-99-011.
5
6 U.S. EPA. (1999b) Estimation of motor vehicle toxic emissions and exposure in selected urban areas. Draft report
7 prepared by Sierra Research, Inc; Radian International Corp; and Energy & Environmental Analysis, Inc., for EPA,
8 Office of Mobile Sources, Assessment and Modeling Division, Ann Arbor, MI. Report No. EPA/420/D-99-002,
9 March 1999.
10
11 Ullman, TL; Spreen, KB; Mason, RL. (1995) Effects of cetane number on emissions from a prototype 1998 HD
12 diesel engine. SAE Technical Paper Ser. No. 950251.
13
14 Unnasch, S; Lowell, D; Lonyai, F; et al. (1993) Performance and emissions of clean fuels in transit buses with
15 Cummins L10 engines. SAE Technical Paper Ser. No. 931782.
16
17 Van Vaeck, L; Van Cauwenberghe, K. (1984) Conversion of polycyclic aromatic hydrocarbons on diesel particulate
18 matter upon exposure to ppm levels of ozone. Atmos Environ 18:323-328.
19
20 Voss, JR; Vanderpoel, RE. (1977) The shuttle distributor for a diesel fuel injection pump. SAE Technical Paper Ser.
21 No. 770083.
22
23 Wall, JC. (1998) Diesel fuel composition for future emissions regulations. Cummins Engine Co. panel discussion,
24 SAE International Fall Fuels and Lubricants Meeting and Exposition, October 21,1998.
25
26 Wall, JC; Shimpi, SA; Yu, ML. (1987) Fuel sulfur reduction for control of diesel particulate emissions. SAE
27 Technical Paper Ser. No. 872139.
28
29 Wang, W; Gautam, M; Sun, X; et al. (1993) Emissions comparisons of twenty-six HD vehicles operated on
30 conventional and alternative fuels. SAE Technical Paper Ser. No. 932952.
31
32 Wang, WG; Lyons, D; Bata, R; et al. (1994) In-use emissions tests of alternatively fueled HD vehicles by a chassis
33 dynamometer testing facility. SAE Technical Paper Ser. No. 945124.
34
3b Warner-Seipli, mA; Dictzrr»ann, IiE. (19S4) Chsrscteriz^tiori o*" wr> motor vehicle emissions under transient driving
36 conditions. Southwest Research Institute. EPA/600/3-84-104.
37
38 Watson, N; Janota, MS. (1982) Turbocharging the internal combustion engine. New York: John Wiley and Sons.
39
40 Watts, FW, Jr. (1995) Assessment of occupational exposure to diesel emissions. In: Diesel exhaust: a critical
41 analysis of emissions, exposure, and health effects. Cambridge, MA: Health Effects Institute, pp. 107-123.
4?
43 Weingartner, E; Keller, C; Stahel, WA; et al. (1997) Aerosol emissions in a road tunnel. Atmos tnviron 3 i(3).451-
44 462.
45
11/5/99 2-101 DRAFT-DO NOT CITE OR QUOTE
-------
1 Whitfield, JK; Harris, DB. (1998) Comparison of HD diesel emissions from engine and chassis dynamometers and
2 on-road testing. 8th CRC On-Road Vehicle Emissions Workshop: San Diego, CA.
4 Williams, DJ; Milne, JW; Quigley, SM; et al. (1989) Particulate emissions from in-use motor vehicles - II. Diesel
5 vehicles. Atmos Environ 23(12):2647.
6
7 Winer, AM; Busby, WF, Jr. (1995) Atmospheric transport and transformation of diesel emissions. In: Diesel
8 exhaust: a critical analysis of emissions, exposure, and health effects. Cambridge, MA: Health Effects Institute, pp.
9 84-105.
10
11 Woskie, SR; Smith, TJ; Hammond, SK; et al. (1988) Estimation of the diesel exhaust exposures of railroad workers:
12 I. Current exposures. Am J Ind Med 13:381 -394.
13
14 Yanowitz, J; Graboski, MS; Ryan, LBA; et al. (1999a) Chassis dynamometer study of emissions from 21 in-use HD
15 diesel vehicles. Environ Sci Technol 33:209-216.
16
17 Yanowitz, J; McCormick, RL; Graboski, MS (1999b) Critical review: in-use emissions from HD diesel vehicles.
18 Environ Sci Technol (submitted).
19
20 Yokely, RA; Garrison, AA; Wehry, EL; et al. (1986) Photochemical transformation of pyrene and benzo[a]pyrene
21 vapor-deposited on eight coal stack ashes. Environ Sci Technol 20:86-90.
22
23 Zaebst, DD; Blad, LM; Morris, JA; et al. (1988) Elemental carbon as a surrogate index of diesel exhaust exposure.
tin: Proceedings of the American Industrial Hygiene Conference, May 15-20, 1988, San Francisco, CA.
.
26 Zaebst, DD; Clapp, DE; Blake, LM; et al. (1991) Quantitative determination of trucking industry workers' exposures
27 to diesel exhaust particles. Am Ind Hyg Assoc J 52:529-541.
28
29 Zelenka, P; Kriegler, W; Herzog, PL; et al. (1990) Ways toward the clean HD diesel. SAE Technical Paper Ser. No.
30 900602.
31
32 Zielinska, B; Arey, J; Atkinson, R; et al. (1986) Reaction of dinitrogen pentoxide with fluoranthene. J Am Chem
33 Soc 108:4126-4132.
34
35 Zielinska, B; Arey, J; Atkinson, R; et al. (1988) Nitration of acephenanthrylene under simulated atmospheric
36 conditions and in solution and the presence of nitroacephenanthrylene(s) in ambient air. Environ Sci Technol
37 22:1044-1048.
38
39 Zielinska, B; Arey, J; Atkinson, R; et al. (1989a) Formation of methylnitronaphthalenes from the gas-phase
40 reactions of 1 - and 2-methylnaphthalene with OH radicals and N2O5 and their occurrence in ambient air. Environ
41 Sci Technol 23:723-729.
42
43 Zielinska, B; Arey, J; Atkinson, R; et al. (1989b) The nitroarenes of molecular weight 247 in ambient paniculate
44 samples collected in southern California. Atmos Environ 23:223-229.
11/5/99 2-102 DRAFT—DO NOT CITE OR QUOTE
-------
1 Zielinska, B; Arey, J; Atkinson, R. (1990) The atmospheric formation of nitroarenes and their occurrence in ambient
2 air. In: Proceedings of the Fourth International Conference on N-Substituted Aryl Compounds: Occurrence,
3 Metabolism and Biological Impact of Nitroarenes. Cleveland, OH, July 1989.
4
5 Zielinska, B; McDonald, J; Hayes, T; et al. (1998) Northern Front Range Air Quality Study, Volume B: Source
6 Measurements. Desert Research Institute.
1175/99 2-103 DRAFT-DO NOT CITE OR QUOTE
-------
3. DOSIMETRY OF DIESEL EXHAUST PARTICLES
IN THE RESPIRATORY TRACT
3.1. INTRODUCTION
2 This chapter presents the data and current scientific thought on the deposition and
3 clearance of diesel particulate matter (DPM) from biological systems as well as discussion of the
4 measures of DPM in tissues. The overall goal of the chapter is to address the issue of animal-to-
5 human extrapolation by integrating these data and thoughts into an estimate of a "human
6 equivalent concentration" (HEC), i.e., the concentration in humans corresponding to those used
7 in animal studies. Models that codify and integrate these data and thoughts to estimate the HEC
8 are described and evaluated. This information is needed to inform the dose-response and
9 extrapolation analyses in Chapter 6 and to facilitate understanding of animal carcinogenicity and
10 the bioavailability of particle organics hi the lung.
11 The major constituents of diesel engine exhaust and their atmospheric reaction products
12 are described in Chapter 2 and in the report on diesel exhaust issued by the Health Effects
13 Institute (Health Effects Institute, 1995). Diesel engine exhaust consists of a complex mixture of
14 typical combustion gases, vapors, low-molecular-weight hydrocarbons, and particles; it is the
15 particle phase that is of greatest health concern.
16 Because pulmonary toxiciry is the major focal point, dosimetric considerations are limited
^P to the lung. The dosimetric aspects of DPM to be considered in this chapter include the
18 characteristics of DPM, deposition of DPM in the conducting airways and alveolar regions,
19 normal DPM clearance mechanisms and rates of clearance in both regions, clearance rates during
20 lung overload, elution of organics from DPM, transport of DPM to extra-alveolar sites, and the
21 interrelationships of these factors in determining the target organ dose. Although assessment of
22 dose-response relationships may permit more advanced extrapolations from high experimental
23 exposure concentrations to ambient levels and from animal test species to humans, the question
24 of mechanistic similarities hi a tumorigenic response between rats and humans remains
25 unanswered and the relevance of the tumorigenic response hi rats to humans questionable.
26
27 3.2. CHARACTERISTICS OF INHALED DPM AND RELATIONSHIP TO PM2 5
28 The formation, transport, and characteristics of DPM are considered in detail hi Chapter 2
29 and in the report on diesel exhaust (Health Effects Institute, 1995). DPM consists of aggregates
30 of spherical carbonaceous particles (about 0.2 urn MMAD) to which significant amounts of
31 higher-molecular-weight organic compounds are adsorbed (Figure 2-1) as the hot engine exhaust
32 is cooled to ambient temperature. DPM has an extremely large surface area that allows for the
11/5/99 3-1 DRAFT—DO NOT CITE OR QUOTE
-------
1 adsorption of organic compounds. Typically, 10% to 40% of DPM mass consists of organic
2 compounds (Health Effects Institute, 1995). This figure compares with mass apportionment of
3 20.9% to organic compounds in PM2 5 samples collected at sites in the eastern United States
4 (U.S. EPA, 1996). These organic chemicals include high-molecular-weight hydrocarbons such as
5 the polyaromatic hydrocarbons (PAHs) and their derivatives. DPM also contains a sulfate
6 component that varies with the sulfur content of the fuel. DPM in areas such as Los Angeles and
7 Denver makes up about 7% and 10%, respectively, of the fine particulate matter (PM) fraction
8 (Health Effects Institute, 1995; Zielinska et al., 1998). In another study of fine particulate mass
9 concentration in southern California, the percentage apportioned to diesel exhaust was even
10 higher, ranging from 33% iri downtown Los Angeles to 14% in a suburban/rural area in
11 California (Schauer et al., 1996).
12
13 3.3. REGIONAL DEPOSITION OF INHALED DPM
14 This section discusses the major factors controlling the disposition of inhaled particles.
15 Note that disposition is defined as encompassing the processes of deposition, absorption,
16 distribution, metabolism, and elimination. The regional deposition of particulate matter in the
17 respiratory tract is dependent on the interaction of a number of factors, including respiratory tract
18 anatomy (airway dimensions and branching configurations), ventilatory characteristics (breathing
19 mode and rate, ventilatory volumes and capacities), physical processes (diffusion, sedimentation,
20 impaction, and interception), and the physicochemical characteristics (particle size, shape,
21 density, and electrostatic attraction) of the inhaled particles. Regional deposition of particulate
22 material is usually expressed as deposition fraction of the total particles or mass inhaled and may
23 be represented by the ratio of the particles or mass deposited in a specific region to the number or
24 mass of particles inspired. The factors affecting deposition in these various regions and their
25 importance in understanding the fate of inhaled DPM are discussed hi the following sections.
26 It is beyond the scope of this document to present a comprehensive account of the
27 complexities of respiratory mechanics, physiology, and toxicology, and only a brief review will
28 be presented here. The reader is referred to publications that provide a more in-depth treatment
29 of these topics (Weibel, 1963; Brain and Mensah, 1983; Raabe et al., 1988; Stober et ai., 1993;
30 U.S. EPA, 1996).
31 The respiratory tract in both humans and experimental mammals can be divided into three
32 regions on the basis of structure, size, and function (International Commission on Radiological
33 Protection, 1994): the extrathoracic (ET), the tracheobronchial (TB), and the alveolar (A). In
34 humans, inhalation can occur through the nose or mouth or both (oronasal breathing). However,
35 many animal models used in respiratory toxicology studies are obligate nose breathers.
36
11/5/99 3-2 DRAFT—DO NOT CITE OR QUOTE
-------
1 3.3.1. Deposition Mechanisms
2 This section provides an overview of the basic mechanisms by which inhaled particles
deposit within the respiratory tract. Details concerning the aerosol physics that explain both how
and why particle deposition occurs as well as data on total human respiratory tract deposition are
5 presented in detail in the earlier PM Criteria Document (U.S. EPA, 1996) and will only be briefly
6 reviewed here. For more extensive discussions of deposition processes, refer to reviews by
7 Morrow (1966), Raabe (1982), U.S. EPA (1982), Phalen and Oldham (1983), Lippmann and
8 Schlesinger (1984), Raabe et al. (1988), and Stober et al. (1993).
9 Particles may deposit by five major mechanisms (inertial impaction, gravitational settling,
10 Brownian diffusion, electrostatic attraction, and interception). The relative contribution of each
11 deposition mechanism to the fraction of inhaled particles deposited varies for each region of the
12 respiratory tract.
13 It is important to appreciate that these processes are not necessarily independent but may,
14 in some instances, interact with one another such that total deposition in the respiratory tract may
15 be less than the calculated probabilities for deposition by the individual processes (Raabe, 1982).
16 Depending on the particle size and mass, varying degrees of deposition may occur in the
17 extrathoracic (or nasopharyngeal), tracheobronchial, and alveolar regions of the respiratory tract.
18 Upon inhalation of particulate matter such as found in diesel exhaust, deposition will
19 occur throughout the respiratory tract. Because of high airflow velocities and abrupt directional
^B changes in the ET and TB regions, inertial impaction is a primary deposition mechanism,
21 especially for particles s 2.5 um dze (aerodynamic equivalent diameter). Although inertial
2 2 impaction is a prominent process for deposition of larger particles in the tracheobronchial region,
23 it is of minimal significance as a determinant of regional deposition patterns for diesel exhaust
24 particles, which have an dje < 1 um and a small aspect ratio (ratio of the length to diameter).
25 All aerosol particles are continuously influenced by gravity, but particles with a
26 dae > 0.5 um are affected to the greatest extent. A spherical, compact particle will acquire a
27 terminal settling velocity when a balance is achieved between the acceleration of gravity acting
28 on the particle and the viscous resistance of the air; it is this velocity that brings the particle into
29 contact with airway surfaces. Both sedimentation and inertial impaction cause the deposition of
30 many particles within the same size range. These deposition processes act together in the ET and
31 TB regions, with inertial impaction dominating in the upper airways and sedimentation becoming
32 increasingly dominant in the lower conducting airways, especially for the largest particles, which
33 can penetrate into the smaller bronchial airways.
34 As particle diameters become <1 um, the particles are increasingly subjected to diffusive
35 deposition because of random bombardment by air molecules, which results in contact with
airway surfaces. A dae of 0.5 um is often considered as a boundary between diffusion and
1175/99 3-3 DRAFT—DO NOT CITE OR QUOTE
-------
1 aerodynamic (sedimentation and impaction) mechanisms of deposition. Thus, instead of having
2 an aerodynamic equivalent diameter (dae), diffusive particles of different shapes can be related to
3 the diffusivity of a thermodynamic equivalent size based on spherical particles (Heyder et al.,
4 1986). Diffusive deposition of particles is favored in the A region of the respiratory tract by the
5 proximate surfaces and by relatively long residence times for particles.
6 Because their dae is generally £ 1 (im, diesel exhaust particles may deposit throughout the
7 respiratory tract. On the basis of animal data regarding the site of origin of diesel exhaust-
8 induced tumors, particle deposition in the alveolar region may be of greatest concern relative to
9 the carcinogenic potential of DPM and/or the adsorbed organics. However, such data for humans
10 are not available. As discussed above, deposition by diffusion would be especially prevalent in
11 the A region, whereas sedimentation would be less significant for such small particles.
12 Electrostatic precipitation is deposition related to particle charge. The electrical charge
13 on some particles may result hi an enhanced deposition over what would be expected from size
14 alone. This is due to image charges induced on the surface of the airway by these particles, or to
15 space-charge effects whereby repulsion of particles containing like charges results in increased
16 migration toward the airway wall. The effect of charge on deposition is inversely proportional to
17 particle size and airflow rate. A recent study employ ing hollow airway casts of the human
18 tracheobronchial tree that assessed deposition of ultrafine (0.02 urn) and fine (0.125 ^m)
19 particles found that deposition of singly charged particles was 5-6 times that of particles having
20 no charge, and 2-3 times that of particles at Boltzmann equilibrium (Cohen et al., 1998). This
21 suggests that within the TB region of humans, electrostatic precipitation may be a significant
22 deposition mechanism for ultrafine and some fine particles, the latter of which are inclusive of
23 DPM. Thus, although electrostatic precipitation is generally a minor contributor to overall
24 particle deposition, it may be important for DPM.
25 Interception is deposition by physical contact with airway surfaces and is most important
26 for fiber deposition; interception is described hi the 1996 CD.
27
28 3.3.1.1. Biological Factors Modifying Deposition
29 The available experimental deposition data in humans are commonly derived using
30 healthy adult Caucasian males. Various factors can act to alter deposition patterns from those
31 obtained in this group. The effects of different biological factors, including gender, age, and
32 respiratory tract disease, on particle deposition have been reviewed previously (U.S. EPA, 1996).
33 The various species that serve as the basis for dose-response assessment in inhalation
34 toxicology studies dr» not receive identical doses in a comparable respiratory tract region (ET,
35 TB, or A) when exposed to the same aerosol or gas (Brain and Mensah, 1983). Such interspecies
36 differences are important because the adverse toxic effect is likely more related to the
1175/99 3-4 DRAFT—DO NOT CITE OR QUOTE
-------
1 quantitative pattern of deposition within the respiratory tract than to the exposure concentration;
2 this pattern determines not only the initial respiratory tract tissue dose but also the specific
pathways by which the inhaled material is cleared and redistributed (Schlesinger, 1985).
Differences in patterns of deposition between humans and animals have been summarized (U.S.
5 EPA, 1996; Schlesinger et al., 1997). Such differences in initial deposition must be considered
6 when relating biological responses obtained in laboratory animal studies to effects in humans.
7 The deposition of inhaled diesel particles in the respiratory tract of humans and
8 mammalian species has been reviewed (Health Effects Institute, 1995). Schlesinger (1985)
9 showed that physiological differences in the breathing mode for humans (nasal or oronasal
10 breathers) and laboratory rats (obligatory nose breathers), combined with different airway
11 geometries, resulted in significant differences in lower respiratory tract deposition for larger
12 particles (> 1 um dae). In particular, a much lower fraction of inhaled larger particles is deposited
13 in the alveolar region of the rat compared with humans. However, relative deposition of the
14 much smaller diesel exhaust particles was not affected as much by the differences among species,
15 as was demonstrated in model calculations by Xu and Yu (1987). These investigators modeled
16 the deposition efficiency of inhaled DPM in rats, hamsters, and humans on the basis of
17 calculations of the models of Schum and Yeh (1980) and Weibel (1963). These simulations
18 (Figure 3-1) indicate relative deposition patterns in the lower respiratory tract (trachea =
19 generation 1; alveoli = generation 23) and are similar among hamsters, rats, and humans.
fc Variations in alveolar deposition of DPM over one breathing cycle in these different species were
21 predicted to be within 30% of one another. Xu and Yu (1987) attributed this similarity to the fact
22 that deposition of the submicron diesel particles is dominated by diffusion rather than
23 sedimentation or impaction. Although these data assumed nose-breathing by humans, the results
24 would not be very different for mouth-breathing because of the low filtering capacity of the nose
25 for particles in the 0.1 to 0.5 urn range.
26 However, for dosimetric calculations and modeling, it would be of much greater
27 importance to consider the actual dose deposited per unit surface area of the respiratory tract
28 rather than the relative deposition efficiencies per lung region. Table 3-1 compares the predicted
29 deposited doses of diesel exhaust particles inhaled in 1 min for the three species, based on the
30 total lung volume, the surface area of all lung airways, or the surface area of the epithelium of the
31 alveolar region only. In Table 3-1, the deposited dose, expressed as either mass/lung volume or
32 mass/surface area(s), is lower in humans than in the two rodent species as a result of the greater
33 respiratory exchange rate in rodents and smaller size of the rodent lung. Such differences in the
34 deposited dose in relevant target areas are important and have to be considered
1175/99 3-5 DRAFT—DO NOT CITE OR QUOTE
-------
10
-2
o
2
LL
C
mo
"3 4 A"
o 10
Q.
O
o
I-
10
Hamster /V
Fischer rat
Human
4 8 12 16 20
Generation Number
24
1
2
3
4
5
Figure 3-1. Modeled deposition distribution patterns of inhaled diesel exhaust particles in
the airways of different species. Generation 1-18 are TB; >18 are A.
Source: Xu and Yu, 1987.
whenextrapolating the results from diesel exhaust exposure studies in animals to humans.
Table 3-1 indicates that the differences (between humans to animals) are less on a surface area
basis (=3-fold) than on a lung volume basis (= 14-fold). This is due to larger alveolar diameters
and concomitant lower surface area per unit of lung volume in humans.
11/5/99
3-6
DRAFT—DO NOT CITE OR QUOTE
-------
Table 3-1. Predicted doses of inhaled diesel exhaust particles per minute based on total
lung volume (M), total airway surface area (M,), or surface area in alveolar region (M2)
M
M
M
M
Species (10° ug/mm/cm3) (KT6
Hamster 3.548
Fischer rat 3.434
Human 0.249
= mass DPM deposited in lune per minute
total lung volume
= mass DPM deposited in lune per minute
total airway surface area
= mass DPM deposited on the unciliated airways oer minute
M,
u.g/min/cm2)
3.088
3.463
1.237
M2
(1CT* ug/min/cm2)
2.382
2.608
0.775
surface area of the unciliated airways
Based on the following conditions: (1) MMAD = 0.2 |^m, o = 1.9, = 0.3, and p = 1.5 g/cm3; (2) particle
concentration = 1 mg/m3; and (3) nose-breathing.
Source: Xu and Yu, 1987.
1 The alternative, perhaps more accurate physiologically, is to consider deposition rate
2 relative to exposure concentration; the deposition rate will initiate particle redistribution
3 processes (e.g., clearance mechanisms, phagocytosis) that transfer the particles to various
1 subcompartments, including the alveolar macrophage pool, pulmonary interstitium, and lymph
5 nodes. Overtime, therefore, only small amounts of the original particle intake would be
6 associated with the alveolar surface.
7
8 3.3.2. Particle Clearance and Translocation Mechanisms
9 This section provides an overview of the mechanisms and pathways by which particles
10 are cleared from the respiratory tract. The mechanisms of particle clearance as well as clearance
11 routes from the various regions of the respiratory tract have been considered in the previous PM
12 Criteria Document (U.S. EPA, 1996) and reviewed by Schlesinger et al. (1997).
13 Particles that deposit upon airway surfaces may be cleared from the respiratory tract
14 completely, or may be translocated to other sites within this system, by various regionally distinct
15 processes. These clearance mechanisms can be categorized as either absorptive (i.e., dissolution)
16 or nonabsorptive (i.e., transport of intact particles) and may occur simultaneously or with
17 temporal variations. Particle solubility in terms of clearance refers to solubility within the
18 respiratory tract fluids and cells. Thus, an "insoluble" particle is one whose rate of clearance by
19 dissolution is insignificant compared to its rate of clearance as an intact particle (as is the case
11/5/99 3-7 DRAFT—DO NOT CITE OR QUOTE
-------
1 with DPM). For the most part, all deposited particles are subject to clearance by the same
2 mechanisms, with their ultimate fate a function of deposition site, physicochemical properties
3 (including any toxicity), and sometimes deposited mass or number concentration.
4
5 3.3.2.1. ETRegion
6 The clearance of insoluble particles deposited in the nonolfactory portion of nasal
7 passages occurs via mucociliary transport, and the general flow of mucus is backwards, i.e.,
8 towards the nasopharynx. Mucus flow in the most anterior portion of the nasal passages is
9 forward, clearing deposited particles to the vestibular region where removal is by sneezing,
10 wiping, or blowing.
11 Soluble material deposited on the nasal epithelium is accessible to underlying cells via
12 diffusion through the mucus. Dissolved substances may be subsequently translocated into the
13 bloodstream. The nasal passages have a rich vasculature, and uptake into the blood from this
14 region may occur rapidly.
15 Clearance of poorly soluble particles deposited in the oral passages is by coughing and
16 expectoration or by swallowing into the gastrointestinal tract.
17
18 3.3.2.2. TB Region
19 The dynamic relationship between deposition and clearance is responsible for
20 determining lung burden at any point in time. Clearance of poorly soluble particles from the
21 tracheobronchial region is mediated primarily by mucociliary transport and is a more rapid
22 process than those operating in alveolar regions. Mucociliary transport (often referred to as the
23 mucociliary escalator) is accomplished by the rhythmic beating of cilia that line the respiratory
24 tract from the trachea through the terminal bronchioles. This movement propels the mucous
25 layer containing deposited particles (or particles within alveolar macrophages [AMs]) toward the
26 larynx. Clearance rate by this system is determined primarily by the flow velocity of the mucus,
27 which is greater in the proximal airways and decreases distaily. These rates also exhibit
28 interspecies and individual variability. Considerable species-dependent variability m
29 tracheobronchial clearance has been reported, with dogs generally having faster clearance rates
30 than guinea pigs, rats, or rabbits (Felicetti et al., 1981). The half-time (t1/2) values for
31 tracheobronchial clearance of relatively insoluble particles are usually on the order of hours, as
32 compared to alveolar clearance, which is on the order of hundreds of days in humans and dogs.
33 The clearance of pa.rticu.late matter from the tracheobronchial region is generally recognized as
34 being hinha
-------
1 for humans may take much longer for a significant fraction of particles deposited in this region,
2 and may not be complete within 24 h as generally believed (Stahlhofen et al., 1990).
Although most of the paniculate matter cleared from the tracheobronchial region will
ultimately be swallowed, the contribution of this fraction relative to carcinogenic potential is
5 unclear. With the exception of conditions of impaired bronchial clearance, the desorption t,/2 for
6 particle-associated organics is generally longer than the tracheobronchial clearance times, thereby
7 making uncertain the importance of this fraction relative to carcinogenesis in the respiratory tract
8 (Pepelko, 1987). Gerde et al. (199la) showed that for low-dose exposures, particle-associated
9 PAHs were rapidly released at the site of deposition. The relationship between the early
10 clearance of insoluble particles (4 um aerodynamic diameter) from the tracheobronchial regions
11 and their longer-term clearance from the alveolar region is illustrated in Figure 3-2.
1.0
(A
o 0.8
2 0.6
o
Tracheobronchial
Deposition
Alveolar Deposition
40 60
Hours after Inhalation
80
100
Figure 3-2. Modeled clearance of insoluble 4-u.m particles deposited in tracheobronchial
and alveolar regions in humans.
Source: Cuddihy and Yeh, 1986.
11/5/99
3-9
DRAFT—DO NOT CITE OR QUOTE
-------
1 Cuddihy and Yeh (1986) reviewed respiratory tract clearance of particles inhaled by
2 humans. Depending on the type of particle (ferric oxide, Teflon discs, or albumin microspheres),
3 the technique employed, and the anatomic region (midtrachea, trachea, or main bronchi), particle
4 velocity (moved by mucociliary transport) ranged from 2.4 to 21.5 mm/min. The highest
5 velocities were recorded for midtracheal transport, and the lowest were for main bronchi. In one
6 study, an age difference was noted for tracheal mucociliary transport velocity (5.8 mm/min for
7 individuals less than 30 years of age and 10.1 mm/min for individuals over 55 years of age).
8 Cuddihy and Yeh (1986) described salient points to be considered when estimating
9 particle clearance velocities from tracheobronchial regions: these include respiratory tract airway
10 dimensions, calculated inhaled particle deposition fractions for individual airways, and thoracic
11 (ALV + TB) clearance measurements. Predicted clearance velocities for the trachea and main
12 bronchi were found to be similar to those experimentally determined for inhaled radiolabeled
13 particles, but not those for intratracheally instilled particles. The velocities observed for
14 inhalation studies were generally lower than those of instillation studies. Figure 3-3 illustrates a
15 comparison of the short-term clearance of inhaled particles by human subjects and the model
16 predictions for this clearance. However, tracheobronchial clearance via the mucociliary escalator
17 is of limited importance for long-term retention.
18 Exposure of F344 rats to whole DPM at concentrations of 0.35, 3.5, or 7.0 mg/m3 for up
19 to 24 mo did not significantly alter tracheal mucociliary clearance of "Tc-macroaggregated
20 albumin instilled into the trachea (Wolff et al., 1987). The. assessment of tracheal clearance was
21 determined by measuring the amount of material retained 1 h after instillation. The authors
22 stated that measuring retention would yield estimates of clearance efficiency comparable to
23 measuring the velocity for transport of the markers in the trachea. The results of this study were
24 in agreement with similar findings of unaltered tracheal mucociliary clearance in rats exposed to
25 DPM (0.21, 1.0, or 4.4 mg/m3) for up to 4 mo (Wolff and Gray, 1980). However, the 1980 study
26 by Wolff and Gray, as well as an earlier study by Battigelli et al. (1966), showed that acute
27 exposure to high concentrations of diesel exhaust soot (1.0 and 4.4 mg/m3 in the study by Wolff
28 and Gray [1980] and 8 to 17 mg/m3 in the study by Battigelli et al. [1966]) produced transient
29 reductions in tracheal mucociliary clearance. Battigelli et al. (1966) also noted that the
30 compromised tracheal clearance was not observed following cessation of exhaust exposure.
31 The fact that tracheal clearance does not appear to be significantly impaired or is impaired
32 only transiently following exposure to high concentrations of DPM is consistent with the absence
33 of pathological effects hi the tracheobronchial region of the respiratory tract in experimental
34 animals exposed to DPM However, the apparent retention of a fraction of the deposited dose in
35 the airways is cause for some concern regarding possible carcinogenic effects in this region,
36 especially in light of the results from simulation studies by Gerde et al. (1991b) that suggested
11/5/99 3-10 DRAFT—DO NOT CITE OR QUOTE
-------
1.0!
C "- V,
,o j / Range of Three
"
-------
1 Moreover, impairment of mucociliary clearance function as a result of exposure to
2 occupational or environmental respiratory tract toxicants or to cigarette smoke may significantly
3 enhance the retention of particles in this region. For example, Vastag et al. (1986) demonstrated
4 that not only smokers with clinical symptoms of bronchitis but also symptom-free smokers have
5 significantly reduced mucociliary clearance rates. Although impaired tracheobronchial clearance
6 could conceivably have an impact on the effects of deposited DPM in the conducting airways, it
7 does not appear to be relevant to the epigenetic mechanism likely present in diesel exhaust-
8 induced rat pulmonary tumors.
9 Poorly soluble particles (i.e., DPM) deposited within the TB region are cleared
10 predominantly by mucociliary transport, towards the oropharynx, followed by swallowing.
11 Poorly soluble particles may also be cleared by traversing the epithelium by endocytotic
12 processes, and enter the peribronchial region. Clearance may occur following phagocytosis by
13 airway macrophages, located on or beneath the mucous lining throughout the bronchial tree, or
14 via macrophages which enter the airway lumen from the bronchial or bronchiolar mucosa
15 (Robertson, 1980).
16
17 3.3.2.3. A Region
18 A number of investigators have reported on the alveolar clearance kinetics of human
19 subjects. Bohning et al. (1980) examined alveolar clearance in eight humans who had inhaled
20 <0.4 mg of 85Sr-labeled polystyrene particles (3.6 ± 1.6 jam diam.). A double-exponential model
21 best described the clearance of the particles and provided t,/2 values of 29 ± 19 days and 298 ±
22 114 days for short-term and long-term phases, respectively. It was noted that of the particles
23 deposited in the alveolar region, 75% ± 13% were cleared via the long-term phase. Alveolar
24 retention tm values of 330 and 420 days were reported for humans who had inhaled
25 aluminosilicate particles of MMAD 1.9 and 6.1 u,m (Bailey et al., 1982).
26 Quantitative data on clearance rates in humans having large lung burdens of particulate
27 matter are lacking. Bohning et al. (1982) and Cohen et al. (1979), however, did provide evidence
28 for slower clearance in smokers.; and Freedro^n ?>n
-------
1 increase the possibility of an enhanced particle accumulation effect resulting from exposure to
2 other particle sources such as diesel exhaust.
Normal alveolar clearance rates in laboratory animals exposed to DPM have been
reported by a number of investigators (Table 3-2). Because the rat is the species for which
5 experimentally induced lung cancer data are available and for which most clearance data exist, it
6 is the species most often used for assessing human risk, and reviews of alveolar clearance studies
7 have been generally limited to this species.
8 Chan et al. (1981) subjected 24 male F344 rats to nose-only inhalation of DPM (6 mg/m3)
9 labeled with 131Ba or 14C for 40 to 45 min and assessed total lung deposition, retention, and
10 elimination. Based on radiblabel inventory, the deposition efficiency in the respiratory tract was
11 15% to 17%. Measurement of I31Ba label in the feces during the first 4 days following exposure
12 indicated that 40% of the deposited DPM was eliminated via mucociliary clearance. Clearance
13 of the particles from the lower respiratory tract followed a two-phase elimination process
14 consisting of a rapid (ti/2 of 1 day) elimination by mucociliary transport and a slower (t\/2 of
15 62 days) macrophage-mediated alveolar clearance. This study provided data for normal alveolar
16 clearance rates of DPM not affected by prolonged exposure or particle overloading.
17 Several studies have investigated the effects of exposure concentration on the alveolar
18 clearance of DPM by laboratory animals. Wolff et al. (1986, 1987) provided clearance data (ti/2)
19 and lung burden values for F344 rats exposed to diesel exhaust for 7 h/day, 5 days/week for 24
P mo. Exposure concentrations of 0.35, 3.5, and 7.0 mg of DPM/m3 were employed in this whole
21 body-inhalation exposure experiment. Intermediate (hours-days) clearance of 67Ga2O3 particles
22 (30 min, nose-only inhalation) was assessed after 6,12, 18, and 24 mo of exposure at all of the
23 DPM concentrations. A two-component function described the clearance of the administered
24 radiolabel:
25
26 F(t) = ^exp(-0.693 t/r,) + 5exp(-0.693t/r2),
27 where F(t) was the percentage retained throughout the respiratory tract, A and B were the
28 magnitudes of the two components (component A representing the amount cleared from nasal,
29 lung, and gastrointestinal compartments and component B representing intermediate clearance
30 from the lung compartment), and t, and T2 were the half-times for the A and B compartments,
31 respectively. The early retention half-times (T,), representing clearance from primary, ciliated
32 conducting airways, were similar for rats in all exposure groups at all time points except for those
33 in the high-exposure (7.0 mg/m3) group following 24 mo of exposure, whose clearance rate was
34 faster than that of the controls. Significantly longer B compartment retention half-times,
1175/99 3-13 DRAFT—DO NOT CITE OR QUOTE
-------
Table 3-2. Alveolar clearance in laboratory animals exposed to DPM
OJ
o
2:
o
H
O
c
o
H
m
Species/set
Rats, F-344, M
Rats, F-34 I
Rats
Rats, F-34 4, MF
Rats, F-344;
Guinea pij;s,
Hartley
Rats, F-344
Exposure
technique
Nose only;
Radiolabeled DPM
Whole body;
assessed effect
on clearance of
67Ga203 particles
Whole body
Whole body
Nose-only;
Radiolabeled 14C
Exposure
duration
40-45 min
7h/day
5 days/week
24 mo
19h/day
5 days/week
2.5 years
7h/day
5 days/week
18 mo
45 min
140 min
45 min
20 h/day
7 days/week
7-1 12 days
Particles
mg/m3
6
0.35
3.5
7.0
4
0.15
0.94
4.1
7
2
7
0.25
6
Observed effects Reference
Four days after exposure, 40% of DPM eliminated by Chan et al. (1981)
mucociliary clearance. Clearance from lower RT was in
2 phases. Rapid mucociliary (t,/2 = 1 day; slower
macrophage-mediated (t,/2 = 62 days).
T, significantly higher with exposure to 7.0 mg/m3 for Wolff et al. (1986,
24 mo; T2 significantly longer after exposure to 7.0 mg/m3 1 987)
for 6 mo. and to 3.5 mg/m3 for 18 mo.
Estimated alveolar deposition = 60 mg; particle burden Heinrich et al.
caused lung overload. Estimated 6- 1 5 mg particle-bound ( 1 986)
organics deposited.
Long-term clearance was 87 ± 28 and 99 ± 8 days for Griffis et al. ( 1 983)
0.15 and 0.94 mg/m3 groups, respectively; t,/2 = 165 days
for 4. 1 mg/m3 group.
Rats demonstrated 3 phases of clearance with tl/2 = 1,6, Lee et al. ( 1 983)
and 80 days, representing tracheobronchial, respiratory
bronchioles, and alveolar clearance, respectively. Guinea
pigs demonstrated negligible alveolar clearance from
day 10 to 432.
Monitored rats for a year. Proposed two clearance models. Chan et al. (1 984)
Clearance depends on initial particle burden; t,/2 increases
with higher exposure. Increases in tl/2 indicate increasing
impairment of AM mobility and transition into overload
condition.
RT = respiratory tract.
AM = alveolar macrophagc.
T, = clearance from primary, ciliated airways.
t2 = clearance from non-ciliated passages.
-------
1 representing the early clearance from nonciliated passages such as alveolar ducts and alveoli,
2 were noted after as few as 6 mo exposure to DPM at 7.0 mg/m3 and 18 mo exposure to 3.5
mg/m3.
Nose-only exposures to l34Cs fused aluminosilicate particles (FAP) were used to assess
5 long-term (weeks-months) clearance. Following 24-mo exposure to DPM, long-term clearance
6 of l34Cs-FAP was significantly (p<0.01) altered in the 3.5 (cumulative exposure [C * T] of
7 11,760 mg-h/m3) and 7.0 mg/m3 C * T = 23,520 mg-h/m3) exposure groups (t,/2 of 264 and 240
8 days, respectively) relative to the 0.35 mg/m3 and control groups (ti/2 of 81 and 79 days,
9 respectively). Long-term clearance represents the slow component of particle removal from the
10 alveoli. The decreased clearance correlated with the greater particle burden in the lungs of the
11 3.5 and 7.0 mg/m3 exposure groups. Based on these findings, the cumulative exposure of
12 11,760 mg-h/m3 represented a particle overload condition resulting in compromised alveolar
13 clearance mechanisms.
14 Heinrich et al. (1986) exposed rats 19 h/day, 5 days/week for 2.5 years to DPM at a
15 particle concentration of about 4 mg/m3. This is equal to a C * T of 53,200 mg-h/m3. The
16 deposition in the alveolar region was estimated to equal 60 mg. The lung particle burden was
17 sufficient to result in a particle overload condition. With respect to the organic matter adsorbed
18 onto the particles, the authors estimated that over the 2.5-year period, 6-15 mg of particle-bound
19 organic matter had been deposited and was potentially available for biological effects. This
^B estimation was based on the analysis of the diesel exhaust used in the experiments, values for rat
21 ventilatory functions, and estimates of deposition and clearance.
22 Accumulated burden of DPM in the lungs following an 18-mo, 7 h/day, 5 days/week
23 exposure to diesel exhaust was reported by Griffis et al. (1983). Male and female F344 rats
24 exposed to 0.15, 0.94, or 4.1 mg DPM/m3 were sacrificed at 1 day and 1,5, 15, 33, and 52 weeks
25 after exposure, and DPM was extracted from lung tissue dissolved in tetramethylammonium
26 hydroxide. Following centrifugation and washing of the supernatant, DPM content of the tissue
27 was quantitated using spectrophotometric techniques. The analytical procedure was verified by
28 comparing results to recovery studies using known amounts of DPM with lungs of unexposed
29 rats. Lung burdens were 0.035, 0.220, and 1.890 mg/g lung tissue, respectively, in rats exposed
30 to 0.15, 0.94, and 4.1 mg DPM/m3. Long-term retention for the 0.15 and 0.94 mg/m3 groups had
31 estimated half-times of 87 ± 28 and 99 ± 8 days, respectively. The retention t,/2 for the
32 4.1-mg/m3 exposure group was 165 ± 8 days, which was significantly (pO.OOOl) greater than
33 those of the lower exposure groups. The 18-mo exposures to 0.15 or 0.96 mg/m3 levels of DPM
34 C x T equivalent of 378 and 2,368 mg-h/m3, respectively) did not affect clearance rates, whereas
35 the exposure to the 4.1 mg/m3 concentration C x T = 10,332 mg-h/m3) resulted in impaired
clearance.
11/5/99 3-15 DRAFT—DO NOT CITE OR QUOTE
-------
1 In a subsequent study (Lee et al., 1983), a three-phase model was used to describe the
2 clearance of DPM (7 mg/m3 for 45 min or 2 mg/m3 for 140 min) by F344 rats (24 per group)
3 exposed by nose-only inhalation with no apparent particle overload in the lungs. The exposure
4 protocols provided comparable total doses based on a I4C radiolabel. I4CO2 resulting from
5 combustion of HC-labeled diesel fuel was removed by a diffusion scrubber to avoid erroneous
6 assessment of 14C intake by the animals. Retention of the radiolabeled particles was determined
7 up to 335 days after exposure and resulted in the derivation of a three-phase clearance of the
8 particles. The resulting retention tm values for the three phases were 1, 6, and 80 days. The
9 three clearance phases are taken to represent removal of tracheobronchial deposits by the
10 mucociliary escalator, removal of particles deposited in the respiratory bronchioles, and alveolar
11 clearance, respectively. Species variability in clearance of DPM was also demonstrated because
12 the Hartley guinea pigs exhibited negligible alveolar clearance from day 10 to day 432 following
13 a 45-min exposure to a DPM concentration of 7 mg/m3. Initial deposition efficiency (20% ± 2%)
14 and short-term clearance were, however, similar to those for rats.
15 Lung clearance in male F344 rats preexposed to DPM at 0.25 or 6 mg/m3 20 h/day,
16 7 days/week for periods lasting from 7 to 112 days was studied by Chan et al. (1984). Following
17 this preexposure protocol, rats were subjected to 45-min nose-only exposure to 14C-diesel
18 exhaust, and alveolar clearance of radiolabel was monitored for up to 1 year. Two models were
19 proposed: a normal biphasic clearance model and a modified lung retention model that included
20 a slow-clearing residual component to account for sequestered aggregates of macrophages. The
21 first model described a first-order clearance for two compartments: R(t) = Ae"ult + Be""2'. This
22 yielded clearance t,/2 values of 166 and 562 days for rats preexposed to 6.0 mg/m3 for 7 and
23 62 days, respectively. These values were significantly (p<0.05) greater than the retention t1/2 of
24 77 ± 17 days for control rats. The same retention values for rats of the 0.25 mg/m3 groups were
25 90 ± 14 and 92 ± 15 days, respectively, for 52- and 112-day exposures and were not significantly
26 different from controls. The two-compartment model represents overall clearance of the tracer
27 particles, even if some of the particles were sequestered in particle-laden macrophages with
28 substantially slower clearance rates. For the second model, which excluded transport of the
29 residual fractions in sequestered macrophage aggregates, slower clearance was observed in the
30 group with a lung burden of 6.5 mg, and no clearance was observed in the 11.8 mg group.
31 Clearance was shown to be dependent on the initial burden of particles and, therefore, the
32 clearance t,/2 would increase in higher exposure scenarios. This study emphasizes the importance
33 of particle overloading of the lung and the ramifications on clearance of particles; the significant
34 increases in half-times indicate an increasing impairment of the alveolar macrophage mobility
35 and subsequent transition into an overload condition. Based on these data, a particle overload
11/5/99 3-16 DRAFT—DO NOT CITE OR QUOTE
-------
1 effect was demonstrated for both the high and low exposure levels (equivalent to C * T dose of
2 840 [transitional overload] and 7,440 mg-h/m3).
3 Long-term alveolar clearance rates of particles in various laboratory animals and humans
4 have been reviewed by Pepelko (1987). Although retention tI/2 varies both among and within
5 species and is also dependent on the physicochemical properties of the inhaled particles, the
6 retention t,/2 for humans is much longer (>8 mo) than the average retention t,/2 of 60 days for rats.
7 Clearance from the A region occurs via a number of mechanisms and pathways, but the
8 relative importance of each is not always certain and may vary between species. Particle removal
9 by macrophages comprises the main nonabsorptive clearance process in this region. Alveolar
10 macrophages reside on the epithelium, where they phagocytize and transport deposited material,
11 which they contact by random motion or via directed migration under the influence of local
12 chemotactic factors (Warheit et al., 1988).
13 Particle-laden macrophages may be cleared from the A region along a number of
14 pathways described in the 1996 CD. Undigested particles or macrophages hi the interstitium
15 may traverse the alveolar-capillary endothelium, directly entering the blood (Raabe, 1982; Holt,
16 1981); endocytosis by endothelial cells followed by exocytosis into the vessel lumen seems,
17 however, to be restricted to particles <0.1 //m diameter, and may increase with increasing lung
18 burden (Lee et al., 1985; Oberdorster, 1988). Once in the systemic circulation, transmigrated
19 macrophages, as well as uningested particles, can travel to extrapulmonary organs.
Alveolar macrophages constitute an important first-line cellular defense mechanism
21 against inhaled particles that deposit in the alveolar region of the lung. It is well established that
22 a host of diverse materials, including DPM, are phagocytized by AMs shortly after deposition
23 (White and Garg, 1981; Lehnert and Morrow, 1985) and that such cell-contained particles are
24 generally rapidly sequestered from both the extracellular fluid lining in the alveolar region and
25 the potentially sensitive alveolar epithelial cells. In addition to this role hi compartmentalizing
26 particles from other lung constituents, AMs are prominently involved in mediating the clearance
27 of relatively insoluble particles from the air spaces (Lehnert and Morrow, 1985). Although the
28 details of the actual process have not been delineated, AMs with their particle burdens gain
29 access and become coupled to the mucociliary escalator and are subsequently transported from
30 the lung via the conducting airways. Although circumstantial, numerous lines of evidence
31 indicate that such AM-mediated particle clearance is the predominant mechanism by which
32 relatively insoluble particles are removed from the lungs (Gibb and Morrow, 1962; Ferin, 1982;
33 Harmsen et al., 1985; Lehnert and Morrow, 1985; Powdrill et al., 1989).
34 The removal characteristics for particles deposited in alveolar region of the lung have
35 been descriptively represented by numerous investigators as a multicompartment or
multicomponent process in which each component follows simple first-order kinetics (Snipes
11/5/99 3-17 DRAFT—DO NOT CITE OR QUOTE
-------
1 and Clem, 1981; Snipes et al., 1988; Lee et al., 1983). Although the various compartments can
2 be described mathematically, the actual physiologic mechanisms determining these differing
3 clearance rates have not been well characterized.
4 . Lehnert (1988, 1989) performed studies using laboratory rats to examine particle-AM
5 relationships over the course of alveolar clearance of low to high lung burdens of noncytotoxic
6 microspheres (2.13 urn diam.) to obtain information on potential AM-related mechanisms that
7 form the underlying bases for kinetic patterns of alveolar clearance as a function of particle lung
8 burdens. The intratracheally instilled lung burdens varied from 1.6 x 107 particles (about 85 ^g)
9 for the low lung burden to 2.0 x 1Q8 particles (about 1.06 mg) for the mid-dose and 6.8 x 108
10 particles (about 3.6 mg) for the highest lung burden. The lungs were lavaged at various times
11 postexposure and the numbers of spheres in each macrophage counted. Although such
12 experiments provide information regarding the response of the lung to particulate matter,
13 intratracheal instillation is not likely to result in the same depositional characteristics as
14 inhalation of particles. Therefore, it is unlikely that the response of alveolar macrophages to
15 these different depositional characteristics will be quantitatively similar.
16 The t1/2 values of both the early and later components of clearance were virtually identical
17 folio wing deposition of the low and medium lung burdens. For the highest lung burden,
18 significant prolongations were found in both the early, more rapid, as well as the slower
19 component of alveolar clearance. The percentages of the particle burden associated with the
20 earlier and later components, however, were similar to those of the lesser lung burdens. On the
21 basis of the data, the authors concluded that translocation of AMs from alveolar spaces by way of
22 the conducting airways is fundamentally influenced by the particle burden of the cells so
23 translocated. In the case of particle overload that occurred at the highest lung burden, the
24 translocation of AMs with the heaviest cellular burdens of particles (i.e., greater than about
25 100 microspheres per AM) was definitely compromised.
26 On the other hand, analysis of the disappearance of AMs with various numbers of
77 particles indicates that the particles may not exclusively reflect the translocation of AMs from the
28 lung. The observations are also consistent with a gradual redistribution of retained particles
29 among the AMs in the lung concurrent with the removal of particle-containing AMs via the
30 conducting airways. Experimental support suggestive of potential processes for such particle
31 redistribution comes from a variety of investigations involving AMs and other endocytic cells
32 (Heppleston and Young, 1973; Evans et al., 1986; Aronson, 1963; Sandusky et al., 1977;
3 3 Heppleston, 1961: Riley and Dean, 1978).
34
35
36
11/5/99 3-18 DRAFT—DO NOT CITE OR QUOTE
-------
1 3.3.3. Translocations of Particles to Extra-alveolar Macrophage Compartment Sites
2 Although the phagocytosis of particles by lung-free cells and the mucociliary clearance of
3 the cells with their paniculate matter burdens represent the most prominent mechanisms that
4 govern the fate of particles deposited in the alveolar region, other mechanisms exist that can
5 affect both the retention characteristics of relatively insoluble particles in the lung and the lung
6 clearance pathways for the particles. One mechanism is endocytosis of particles by alveolar
7 lining (Type I) cells (Sorokin and Brain, 1975; Adamson and Bowden, 1978, 1981) that normally
8 provide >90% of the cell surface of the alveoli in the lungs of a variety of mammalian species
9 (Crapo et al., 1983). This process may be related to the size of the particles that deposit in the
10 lungs and the numbers of particles that are deposited. Adamson and Bowden (1981) found that
11 with increasing loads of carbon particles (0.03 um diam.) instilled in the lungs of mice, more free
12 particles were observed in the alveoli within a few days. The relative abundance of particles
13 endocytosed by Type I cells also increased with increasing lung burdens of the particles, but
14 instillation of large particles (1.0 (im) rarely resulted in their undergoing endocytosis. A 4 mg
15 burden of 0.1 um diameter latex particles is equivalent to 8 x 1012 particles, whereas a 4 mg
16 burden of 1.0 um particles is composed of 8 x 109 particles. Regardless, DPM with volume
17 median diameters between 0.05 and 0.3 um (Frey and Corn, 1967; Kittleson et al., 1978) would
18 be expected to be within the size range for engulfment by Type I cells should suitable encounters
19 occur. Indeed, it has been demonstrated that DPM is endocytosed by Type I cells in vivo (White
and Garg, 1981).
21 Unfortunately, information on the kinetics of particle engulfment (endocytosis) by Type I
22 cells relative to that by AMs is scanty. Even when relatively low burdens of particulate matter
23 are deposited in the lungs, some fraction of the particles usually appears in the regional lymph
24 nodes (Ferin and Fieldstein, 1978; Lehnert, 1989). As will be discussed, endocytosis of particles
25 by Type I cells is an initial, early step in the passage of particles to the lymph nodes. Assuming
26 particle phagocytosis is not sufficiently rapid or perfectly efficient, increasing numbers of
27 particles would be expected to gain entry into the Type I epithelial cell compartment during
28 chronic aerosol exposures. Additionally, if particles are released on a continual basis by AMs
29 that initially sequestered them after lung deposition, some fraction of the "free" particles so
30 released could also undergo passage from the alveolar space into Type I cells.
31 The endocytosis of particles by Type I cells represents only the initial stage of a process
32 that can lead to the accumulation of particles in the lung's interstitial compartment and the
33 subsequent translocation of particles to the regional lymph nodes. As shown by Adamson and
34 Bowden (1981), a vesicular transport mechanism in the Type I cell can transfer particles from the
35 air surface of the alveolar epithelium into the lung's interstitium, where particles may be
phagocytized by interstitial macrophages or remain in a "free" state for a poorly defined period
11/5/99 3-19 DRAFT—DO NOT CITE OR QUOTE
-------
1 - that may be dependent on the physicochemical characteristics of the particle. The lung's
2 interstitial compartment, accordingly, represents an anatomical site for the retention of particles
3 in the lung. Whether or not AMs, and perhaps polymorphonuclear neutrophils (PMNs) that have
4 gained access to the alveolar space compartment and phagocytize particles there, also contribute
5 to the particle translocation process into the lung's interstitium remains a controversial issue.
6 Evidence that such migration of AMs may contribute to the passage of particles to the interstitial
7 compartment and also may be involved in the subsequent translocation of particles to draining
8 lymph nodes has been obtained with the dog model (Harmsen et al., 1985).
3 The fate of particles once they enter the lung's interstitial spaces remains unclear. Some
10 particles, as previously indicated, are phagocytized by interstitial macrophages, whereas others
11 apparently remain in a free state in the interstitium for some time without being engulfed by
12 interstitial macrophages. It is unknown if interstitial macrophages subsequently enter the alveoli
13 with their engulfed burdens of particles and thereby contribute to the size of the resident AM
14 population over the course of lung clearance. Moreover, no investigations have been conducted
15 to date to assess the influence that the burden of particles may have on the ability of the
16 interstitial macrophage to migrate into the alveolar space compartment.
17 At least some particles that gain entry into the interstitial compartment can further
18 translocate to the extrapulmonary regional lymph nodes. This process apparently can involve the
19 passage of free particles as well as particle-containing cells via lymphatic channels in the lungs
20 (Harmsen et al., 1985; Ferin and Fieldstein, 1978; Lee et al., 1985). It is conceivable that the
21 mobility of the interstitial macrophages could be particle-burden limited, and under conditions of
22 high cellular burdens a greater fraction of particles that accumulate in the lymph may reach these
23 sites as free particles. Whatever the process, existing evidence indicates that when lung burdens
24 of particles result in a particle-overload condition, particles accumulate both more rapidly and
25 abundantly in lymph nodes that receive lymphatic drainage from the lung (Ferin and Feldstein,
26 1978; Lee etal., 1985).
27
28 3.3.3.1. Clearance Kinetics
29 The clearance kinetics of PM have been reviewed in the PM CD (U.S. EPA, 1996) and by
30 Schlesinger et al. (1997). Deposited particles may be completely or incompletely cleared from
31 the respiratory tract. However, the time frame over which clearance occurs affects the
32 cumulative dose delivered to the respiratory tract, as well as to extrapulmonary organs.
33
34 3.3.3.2, Interspecies Patterns of Clearance
35 The inability to study the retention of certain materials in humans for direct risk
36 assessment requires the use of laboratory animals. Since dosimetry depends on clearance rates
11/5/99 3-20 DRAFT—DO NOT CITE OR QUOTE
-------
1 and routes, adequate toxicological assessment necessitates that clearance kinetics in these
2 animals be related to those in humans. The basic mechanisms and overall patterns of clearance
3 from the respiratory tract are similar in humans and most other mammals. However, regional
4 clearance rates can show substantial variation between species, even for similar particles
5 deposited under comparable exposure conditions. This has been extensively reviewed in the
6 previous document (U.S. EPA, 1996) and in other papers (Schlesinger et al., 1997; Snipes et al.,
7 1989).
8 In general, there are species-dependent rate constant^ for various clearance pathways.
9 Differences in regional and total clearance rates between some species are a reflection of
10 differences in mechanical clearance processes. For consideration in assessing particle dosimetry,
11 the end result of interspecies differences in clearance is that the retention of deposited particles
12 can differ between species, which may result in differences in response to similar paniculate
13 exposure atmospheres.
14
15 3.3.3.3. Biological Factors Modifying Clearance
16 A number of host and environmental factors may modify normal clearance patterns.
17 These include age, gender, physical activity, respiratory tract disease, and irritant inhalation (U.S.
18 EPA, 1996).
19
20 3.33.4. Respiratory Tract Disease
21 Earlier studies reviewed in the PM CD (U.S. EPA, 1996) noted that various respiratory
22 tract diseases are associated with clearance alterations. The examination of clearance in
23 individuals with lung disease requires careful interpretation of results, since differences in
24 deposition of tracer particles used to assess clearance function may occur between normal
25 individuals and those with respiratory disease, and this would directly impact upon the measured
26 clearance rates, especially in the tracheobronchial tree. Prolonged nasal mucociliary clearance in
27 humans is associated with chronic sinusitis, bronchiectasis or rhinitis, and cystic fibrosis.
28 Bronchial mucus transport may be impaired in people with bronchial carcinoma, chronic
29 bronchitis, asthma, and various acute infections. In certain of these cases, coughing may enhance
30 mucus clearance, but it generally is effective only if excess secretions are present.
31 The rates of A region particle clearance were reduced in humans with chronic obstructive
. 32 lung disease and in laboratory animals with viral infections, while the viability and functional
33 activity of macrophages were impaired in human asthmatics and in animals with viral-induced
34 lung infections (U.S. EPA, 1996). However, any modification of functional properties of
35 macrophages appears to be injury specific, reflecting the nature and anatomic pattern of disease.
11/5/99 3-21 DRAFT—DO NOT CITE OR QUOTE
-------
1 3.4. PARTICLE OVERLOAD
2 3.4.1. Introduction
3 Some experimental studies using laboratory rodents employed high exposure
4 concentrations of relatively nontoxic, poorly soluble particles. These particle loads interfered
5 with normal clearance mechanisms, producing clearance rates different from those that would
6 occur at lower exposure levels. Prolonged exposure to high particle concentrations is associated
7 with what is termed particle overload. This is defined as the overwhelming of macrophage-
8 mediated clearance by the deposition of particles at a rate exceeding the capacity of that
9 clearance pathway.
10 Wolff et al. (1987) used I34Cs-labeled fused aluminosilicate particles to measure alveolar
11 clearance in rats following 24-mo exposure to low (L), medium (M), and high (H) concentrations
12 of diesel exhaust (targeted concentrations of DPMof 0.35, 3.5 and 7.0 mg/m3). The short-term
13 component of the multicomponent clearance curves was similar for all groups, but long-term
14 clearance was retarded in the M and H exposure groups (Figure 3-4). The half times of the long-
15 term clearance curves were 79, 81, 264, and 240 days, respectively, for the control, L, M, and
16 The observed lung burdens increased progressively, reaching levels of 11.5 and 20.5 mg
100
in
O
5
I
13
CD
O)
High
.2
'E
20
100 120
Time (Days)
140
160
180 200
Figure 3-4. Clearance from lungs of rats of 134Cs-FAP fused aluminosilicate tracer
particles inhaled after 24 months of diesel exhaust exposure at concentrations
of 0 (control) (•), 0.35 (low) (•), 3.5 (medium) (•), and 7.0 (high) mg DFM/nr
(A). Points on curves are means ± SE.
Source: Wolff et al., 1987.
11/5/99
3-22
DRAFT—DO NOT CITE OR QUOTE
-------
1 H exposure groups. Clearance was overloaded at M and H exposure levels, but not by the
2 L exposure level. Lung burdens of DPM were measured after 6, 12,18, and 24 mo of exposure.
3 DPM/lung, respectively, after 24 mo in the M and H exposed groups (Figure 3-5). The results
4 indicate that the clearance of freshly deposited particles was retarded after 24 mo of DPM
5 exposure at the two highest exposure levels, and that clearance had become overloaded at these
6 two exposures but not at the lowest exposure.
7 It has been hypothesized that overloading will begin in the rat when deposition
8 approaches 1 mg particles/g lung tissue (Morrow, 1988). When the concentration reaches 10 mg
9 particles/g lung tissue, macrophage-mediated clearance of particles would effectively cease. It is
10 a nonspecific effect noted in experimental studies, generally in rats, using many different kinds of
11 poorly soluble particles (including Ti02, volcanic ash, DPM, carbon black, and fly ash) and
12 results in A region clearance slowing or stasis, with an associated inflammation and aggregation
13 of macrophages in the lungs and increased translocation of particles into the interstitium (Muhle
14 et al., 1990; Lehnert, 1990; Morrow, 1994). Following overloading, the subsequent retardation
20-i
o.
a
-------
1 of lung clearance, accumulation of particles, inflammation, and the interaction of inflammatory
2 mediators with cell proliferative processes and DNA may lead to the development of tumors and
3 fibrosis in rats (Mauderly, 1996). The phenomenon of overload has been discussed in greater
4 detail in the previous PM CD (U.S. EPA, 1996).
5
6 3.4.2. Relevance to Humans
7 The relevance of lung overload to humans, and even to species other than laboratory rats
8 and mice, is not clear. While it is likely to be of little relevance for most "real world" ambient
9 exposures of humans, it is of concern in interpreting some long-term experimental exposure data
10 and perhaps for humans' occupational exposure. In addition, relevance to humans is clouded by
11 the suggestion that macrophage-mediated clearance is normally slower and perhaps less
12 important in humans than in rats (Morrow, 1994), and that there can be significant differences in
13 macrophage loading between species.
14 Particle overload appears to be an important factor in the diesel emission-induced
15 pulmonary carcinogenicity observed in rats. Studies described in this section provide additional
16 data showing a particle overload effect. A study by Griffis et al. (1983) demonstrated that
17 exposure (7 h/day, 5 days/week) of rats to DPM at concentrations of 0.15, 0.94, or 4.1 mg/m3 for
18 18 mo resulted in lung burdens of 0.035, 0.220, and 1.890 mg/g of lung tissue, respectively. The
19 alveolar clearance of those rats with the highest lung burden (1.890 mg/g of lung) was unpaired,
20 as determined by a significantly greater (pO.OOOl) retention t1/2 for DPM. This is reflected in
21 the greater lung burden/exposure concentration ratio at the highest exposure level. Similarly, in
22 the study by Chan et al. (1984), rats exposed for 20 h/day, 7 days/week to DPM (6 mg/m3) for
23 112 days had a total lung particle burden of 11.8 mg, with no alveolar particle clearance being
24 detected over 1 year.
25 Muhle et al. (1990) indicated that overloading of rat lungs occurred when lung particle
26 burdens reached 0.5 to 1.5 mg/g of lung tissue and that clearance mechanisms were totally
27 compromised at lung particle burdens ^ 10 mg/g for particles with a specific density close to 1.
28 Pritchard (1989), utilizing data from a number of diesel exhaust exposure studies,
29 examined alveolar clearance in rats as a function of cumulative exposure. The resulting analysis
30 noted a significant increase in retention t1/2 values at exposures above 10 mg/m3-h/day and also
31 showed that normal lung clearance mechanisms appeared to be compromised as the lung DPM
32 burden approached 0.5 mg/g of lung.
33 Morrow (1988) has proposed that the condition of particle overloading in the lungs is
34 caused by a loss in the mobility of particle-engorged AMs and that such an impediment is related
35 to the cumulative volumetric load of particles in the AMs. Morrow (1988) has further estimated
36 that the clearance function of an AM may be completely impaired when the particle burden in the
11 /5/99 3-24 DRAFT—DO NOT CITE OR QUOTE
-------
1 AM is of a volumetric size equivalent to about 60% of the normal volume of the AM. Morrow's
2 hypothesis was the initial basis for the physiology-oriented multicompartmental kinetic (POCK)
3 model derived by Stober et al. (1989) for estimating alveolar clearance and retention of
4 biologically insoluble, respirable particles.
5 A revised version of this model refines the characterization of the macrophage pool by
6 including both the mobile and immobilized macrophages (Stober et al., 1994). Application of
7 the revised version of the model to experimental data suggested that lung overload does not cause
8 a dramatic increase in the total burden of the macrophage pool but results in a great increase in
9 the particle burden of the interstitial space, a compartment that is not available for macrophage-
10 mediated clearance. The revised version of the POCK model is discussed in greater detail in the
11 context of other dosimetry models below.
12 Oberdorster and co-workers (1992) assessed the alveolar clearance of smaller (3.3 urn
13 diam.) and larger (10.3 (am diam.) polystyrene particles, the latter of which are volumetrically
14 equivalent to about 60% of the average normal volume of a rat AM. after intratracheal instillation
15 into the lungs of rats. Even though both sizes of particles were found to be phagocytized by AMs
16 within a day after deposition, and the smaller particles were cleared at a normal rate, only
17 minimal lung clearance of the larger particles was observed over an approximately 200-day
18 postinstillation period, thus supporting the volumetric overload hypothesis.
19 Animal studies have revealed that impairment of alveolar clearance can occur following
chronic exposure to DPM (Griffis et al., 1983; Wolff et al., 1987; Vostal et al., 1982; Lee et al.,
21 1983) or a variety of other diverse poorly soluble particles of low toxicity (Lee et al., 1986, 1988;
22 Ferin and Feldstein, 1978; Muhle et al., 1990). Because high lung burdens of insoluble,
23 biochemically-inert particles result in diminution of normal lung clearance kinetics or in what is
24 now called particle overloading, this effect appears to be more related to the mass and/or volume
25 of particles in the lung than to the nature of the particles per se. Particle overload only relates to
26 poorly soluble articles of low toxicity. It must be noted, however, that some types of particles
27 may be cytotoxic and impair clearance at lower lung burdens (e.g., silica may impair clearance at
28 much lower lung burdens than DPM). Regardless, as pointed out by Morrow (1988), particle
29 overloading in the lung modifies the dosimetry for particles in the lung and thereby can alter
30 toxicologic responses.
31 Although quantitative data are limited regarding lung overload associated with impaired
32 alveolar clearance in humans, impairment of clearance mechanisms appears to occur, and at a
33 lung burden generally in the range reported to impair clearance in rats. Stober et al. (1967), in
34 their study of coal miners, reported lung particle burdens of 2 to 50 mg/g lung tissue, for which
35 estimated clearance t,/2 values were very long (4.9 years). Freedman and Robinson (1988) also
reported slower alveolar clearance rates in coal miners, some of whom had a mild degree of
11/5/99 3-25 DRAFT—DO NOT CITE OR QUOTE
-------
1 pneumoconiosis. It must be noted, however, that no lung cancer was reported for those miners
2 with apparent particle overload.
3
4 3.4.3. Potential Mechanisms for an AM Sequestration Compartment for Particles During
5 Particle Overload
6 Several factors may be involved in the particle-load-dependent retardations in the rate of
7 particle removal from the lung and the corresponding functional appearance of an abnormally
8 slow clearing or particle sequestration compartment. As previously mentioned, one potential site
9 for particle sequestration is the containment of particles in the Type I cells. Information on the
10 retention kinetics for particles in the Type I cells is not currently available. Also, no
11 morphometric analyses have been performed to date to estimate what fraction of a retained lung
12 burden may be contained in the Type I cell population of the lung during lung overloading.
13 Another anatomical region in the lung that may be a slow clearing site is the interstitial
14 compartment. Little is known about the kinetics of removal of free particles or particle-
15 containing macrophages from the interstitial spaces, or what fraction of a retained burden of
16 particles is contained in the lung's interstitium during particle overload. The gradual
17 accumulation of particles in the regional lymph nodes and the appearance of particles and cells
18 with associated particles in lymphatic channels and in the peribronchial and perivascular
19 lymphoid tissue (Lee et al., 1985; White and Garg, 1981) suggest that the mobilization of
20 particles from interstitial sites via local lymphatics is a continual process.
21 Indeed, it is clear from histologic observations of the lungs of animals chronically
22 exposed to DPM that Type I cells, the interstitium, the lymphatic channels, and pulmonary
23 lymphoid tissues could represent subcompartments of a more generalized slow clearing
24 compartment.
25 Although these sites must be considered potential contributors to the increased retention
26 of particles during particle overload, a disturbance in particle-associated AM-mediated clearance
27 is undoubtedly the predominant cause, inasmuch as the AMs are the primary reservoirs of
28 deposited particles. The factors responsible for a failure of AMs to translocate from the alveolar
29 space compartment in lungs with high particuiate matter burdens remain uncertain, although a
30 hypothesis concerning the process has been offered involving volumetric AM burden (Morrow,
31 1988).
32 Other processes also may be involved in preventing particle-laden AMs from leaving the
33 alveolar compartment under conditions of particle overload in the lung. Clusters or aggregates of
34 particle-laden AMs in the alveoli are typically found in the lungs of laboratory animals that have
35 received large lung burdens of a variety of types of particles (Lee et al., 1985), including DPM
36 (White and Garg, 1981; McClellan et al., 1982). The aggregation of AMs may explain, in part,
37 the reduced clearance of particle-laden AM during particle overload. The definitive
11/5/99 3-26 DRAFT—DO NOT CITE OR QUOTE
-------
1 mechanism(s) responsible for this clustering of AMs has not been elucidated to date. Whatever
2 the underlying mechanism(s) for the AM aggregation response, it is noteworthy that AMs
3 lavaged from the lungs of diesel exhaust-exposed animals continue to demonstrate a propensity
4 to aggregate (Strom, 1984). This observation suggests that the surface characteristics of AMs are
5 fundamentally altered in a manner that promotes their adherence to one another in the alveolar
6 region and that AM aggregation may not simply be directly caused by their abundant
7 accumulation as a result of immobilization by large particle loads. Furthermore, even though
8 overloaded macrophages may redistribute particle burden to other AMs, clearance may remain
9 inhibited (Lehnert, 1988). This may, in part, be due to attractants from the overloaded AMs
10 causing aggregation of those that are not carrying a particle burden.
11
12 3.5. MODELING THE DISPOSITION OF PARTICLES IN THE RESPIRATORY
13 TRACT
14 3.5.1. Introduction
15 The biologic effects of inhaled particles are a function of their disposition. This, in turn,
16 depends on their patterns of both deposition (i.e., the sites within which they initially come into
17 contact with airway epithelial surfaces and the amount removed from the inhaled air at these
18 sites) and clearance (i.e., the rates and routes by which deposited materials are removed from the
19 respiratory tract). Removal of deposited materials involves the competing processes of
macrophage-mediated clearance and dissolution-absorption. Over the years, mathematical
21 • models for predicting deposition, clearance and, ultimately, retention of particles in the
22 respiratory tract have been developed. Such models help interpret experimental data and can be
23 used to make predictions of deposition for cases where data are not available. A review of
24 various mathematical deposition models was given by Morrow and Yu (1993) and in U.S. EPA
25 (1996).
26 Currently available data for long-term inhalation exposures to insoluble particles (e.g.,
27 TiO2, carbon black, and DPM) show that pulmonary retention and clearance of these particles are
28 not adequately described by simple first-order kinetics and a single compartment representing the
29 alveolar macrophage particle burden. Several investigators have developed models for
30 deposition, transport, and clearance of insoluble particulate matter in the lungs. All of these
31 models identify various compartments and associated transport rates, but empirically derived data
32 are not available to validate many of the assumptions made in these models.
33
34 3.5.2. Dosimetry Models for DPM
35 3.5.2.1. Introduction
« Diesel particles are irregularly shaped aggregates with a mass median aerodynamic
diameter (MMAD) of approximately 0.2 jim, formed from primary spheres 15-30 nm in
11/5/99 3-27 DRAFT—DO NOT CITE OR QUOTE
-------
1 diameter. The primary sphere consists of a dense carbonaceous core (soot) on which various
2 combustion-derived organic compounds, accounting for 10% to 30% of the particle mass, are
3 adsorbed.
4 The extrapolation of toxicological results from laboratory animals to humans requires the
5 use of dosimetry models for both species that include, first, the deposition of DPMs in various
6 regions of the respiratory tract, and second, the transport and clearance of the particles from their
7 deposited sites. Particles deposit by impaction, sedimentation, interception, and diffusion. The
8 contribution from each mechanism is a function of particle size, lung structure, and size and
9 breathing parameters. Because of the size of diesel particles, under normal breathing conditions
10 most of this deposition takes place by diffusion, and the fraction of the inhaled mass that is
11 deposited in the thoracic region is substantially similar for rats and humans. The clearance of
12 particles takes place (1) by mechanical processes: mucociliary transport in the ciliated conducting
13 airways and macrophage phagocytosis and migration in the nonciliated airways, and (2) by
14 dissolution. The removal of the carbonaceous soot is largely by mechanical clearance, whereas
15 the clearance of the adsorbed organics is principally by dissolution.
16
17 3.5.2.2. Deposition Models
18 Among deposition models that include aspects of lung structure and breathing dynamics,
19 the most widely used have been typical-path or single-path models (Yu, 1978; Yu and Diu,
20 1983). The single-path models are based on an idealized symmetric geometry of the lung,
21 assuming regular dichotomous branching of the airways and alveolar ducts (Weibel, 1963). They
22 lead to modeling the deposition in an average regional sense for a given lung depth. Although
23 the lower airways of the lung may be reasonably characterized by such a symmetric
24 representation, there are major asymmetries in the upper airways of the tracheobronchial tree that
25 in turn lead to different apportionment of airflow and particulate burden to the different lung
26 lobes. The rat lung structure is highly asymmetric because of its monopodial nature, leading to
27 significant errors in a single-path description. This is rectified in the multiple-path model of the
28 lung that incorporates asymmetry and heterogeneity in lung branching structure, and calculates
29 deposition at the individual airway level. This model has been developed for the rat lung by
30 Anjilvel and Asgharian (1995) and, in a limited fashion because of insufficient morphometric
31 data, for the human lung (Subramaniam et al., 1998; Yeh and Schun% 1980). Such models are
32 particularly relevant for fine and ultrafine particles. However, models for clearance have not yet
33 been implemented in conjunction with the use of the multiple-path model. Therefore, in this
34 report v/e use only the single-p?.tb model in deposition calculations, specifically the works by Yu
35 and Xu (1986) and Xu and Yu( 1987).
36
37
11/5/99 3-28 DRAFT—DO NOT CITE OR QUOTE
-------
1 3.5.2.3. Physiologically Based Models for Clearance
2 Several clearance models currently exist, and these differ significantly in the level of
3 physiological detail that is captured in the model and in the uncertainties associated with the
4 values of the parameters used. All of these models identify various compartments and associated
5 transport rates, but empirically derived data are not available to validate many of the assumptions
6 made in the models. We compare four of the most widely discussed models below.
7
8 3.5.2.3.1. Two-compartment model Currently available data for long-term inhalation exposures
9 to insoluble particles (e.g., TiO2, carbon black, and DPM) show that pulmonary retention and
i
10 clearance of these particles are not adequately described by simple first-order kinetics and a
11 single compartment representing the alveolar macrophage particle burden. A two-compartment
12 model was developed by Smith (1985) that includes alveolar and interstitial compartments. For
13 uptake and clearance of particles by alveolar surface macrophages and interstitial encapsulation
14 of particles (i.e., quartz dust), available experimental data show that the rate-controlling functions
15 followed Michaelis-Menton type kinetics, while other processes affecting particle transfer are
16 assumed to be linear. Although this model provides rate constants as functions that vary
17 depending on the conditions within the various compartments, most of the described functions
18 could not be validated with experimental data.
19
3.5.2.3.2. Multicompartmental models. Strom et al. (1988) developed a multicompartmental
21 model for particle retention that partitioned the alveolar region into two compartments on the
22 basis of the physiology of clearance. The alveolar region has a separate compartment for
23 sequestered macrophages, which corresponds to phagocytic macrophages that are heavily laden
24 with particles and clustered, and therefore have significantly lowered mobility. The model has
25 the following compartments: (1) tracheobronchial tree, (2) free particulate on the alveolar
26 surface, (3) mobile phagocytic alveolar macrophages, (4) sequestered particle-laden alveolar
27 macrophages, (5) regional lymph nodes, and (6) gastrointestinal tract. The model is based on
28 mass-dependent clearance (the rate coefficients reflect this relationship), which dictates
29 sequestration of particles and their eventual transfer to the lymph nodes. The transport rates
30 between various compartments were obtained by fitting the calculated results to lung and lymph
31 node burden experimental data for both exposure and postexposure periods. Since the number of
32 fitted parameters was large, the model is not likely to provide unique solutions that would
33 simulate experimental data from various sources and for different exposure scenarios. For the
34 same reason, it is not readily possible to use this model for extrapolating to humans.
35
« 3.5.2.3.3. POCKmodeL Stober and co-workers have worked extensively in developing models
for estimating retention and clearance of biologically insoluble, respirable particles in the lung.
11/5/99 3-29 DRAFT—DO NOT CITE OR QUOTE
-------
1 Their most recent work (1994), a revised version of the POCK (physiologically oriented
2 multicompartmental kinetic) model, is a rigorous attempt to incorporate most of the
3 physiologically known aspects of alveolar clearance and retention of inhaled insoluble particles.
4 Their multicompartmental kinetics model has five subcompartments. The transfer of particles
5 between any of the compartments within the alveolar region is macrophage-mediated. There are
6 two compartments that receive particles cleared from the alveolar regions: the tracheobronchial
7 tract and the lymphatic system.
8 The macrophage pool includes both mobile and particle-laden, immobilized
9 macrophages. The model assumes a constant maximum volume capacity of the macrophages for
10 particle uptake and a material-dependent critical macrophage load that results in total loss of
11 macrophage mobility. Sequestration of those macrophages heavily loaded with, a particle burden
12 close to a volume load capacity is treated in a sophisticated manner by approximating the particle
13 load distribution in the macrophages. The macrophage pool is compartmentalized in terms of
14 numbers of macrophages that are subject to discrete particle load intervals. Upon macrophage
15 death, the phagocytized particle is released back to the alveolar surface; thus phagocytic particle
16 collection competes to some extent with this release back to the alveolar surface. This recycled
17 particle load is also divided into particle clusters of size intervals defining a cluster size
18 distribution on the alveolar surface. The model yields a time-dependent frequency distribution of
19 loaded macrophages that is sensitive to both exposure and recovery periods in inhalation studies.
20 The POCK model also emphasizes the importance of interstitial burden in the particle
21 overload phenomenon and indicates that particle overload is a function of a massive increase in
22 particle burden of the interstitial space rather than total burden of the macrophage pool. The
23 relevance of the increased particle burden in the interstitial space lies with the fact that this
24 compartmental burden is not available for macrophage-mediated clearance and, therefore,
25 persists even after cessation of exposure.
26 While the POCK model is the most sophisticated in the physiological complexity it
27 introduces; it. suffers from a major disadvantage. Experimental retention studies provide data on
28 total alveolar and lymph node mass burdens of the particles as a function of time. The relative
29 fraction of the deposit between the alveolar subcompartments in the Stober model therefore
30 cannot be obtained experimentally; the model thus uses a large number of parameters that are
31 simultaneously fit to experimental data. Although the model predictions are tenable,
32 experimental data are not currently available to validate the proposed compartmental burdens or
33 the transfer rates associated with these compartments. Thus the over-parameterization in the
34 mode' leads to the problem that the model may not provide a unique solution that may be used
35 for a variety of exposure scenarios, and for the same reason, cannot be used for extrapolation to
36 humans. Stober et al. have not developed an equivalent model for humans; therefore the use of
37 their model in our risk assessment for diesel is not attempted.
11/5/99 3-30 DRAFT—DO NOT CITE OR QUOTE
-------
1 3.5.2.3.4. Yu-Yoon model Yuand Yoon( 1990), on the other hand, have developed a three-
2 compartment lung model that consists of tracheobronchial (T), alveolar (A), and lymph node (L)
3 compartments (Appendix B, Figure B-l) and, in addition, considered filtration by a
4 nasopharyngeal or head (H) compartment. Absorption by the blood (B) and gastrointestinal (0)
5 compartments, was also considered. While the treatment of alveolar clearance is physiologically
6 less sophisticated than that of the Stober et al. model, the Yu-Yoon model provides a more
7 comprehensive treatment of clearance by including systemic compartments and the head, and
8 including the clearance of the organic components of DPM in addition to the insoluble carbon
9 core.
10 The tracheobronchial compartment is important for short-term considerations, while
11 long-term clearance takes place via the alveolar compartment. In contrast to the Stober and
12 Strom approaches, the macrophage compartment in the Yu-Yoon model contains all of the
13 phagocytized particles; that is, there is no separate (and hypothetical) sequestered macrophage
14 subcompartment. Instead, in order to progress beyond the classical retention model
15 . (International Commission on Radiological Protection, 1979), Yu and Yoon have addressed the
16 impairment of long-term clearance (the overload effect) by using a set of variable transport rates
17 for clearance from the alveolar region as a function of the mass of DPM in the alveolar
18 compartment. A functional relationship for this was derived mathematically (Yu et al., 1989)
19 based upon Morrow's hypothesis for the overload effect that we discussed earlier in the section
on pulmonary overload. The extent of the impairment depends on the initial particle burden,
o
21 with greater particulate concentration leading to slower clearance.
22 DPM are treated as composed of three material components: an insoluble carbonaceous
23 core, slowly cleared organics (10% particle mass), and fast-cleared organics (10% particle mass).
24 Such a partitioning of organics was based on observations that the retention of particle-associated
25 organics in lungs shows a biphasic decay curve (Sun et al., 1984; Bond et al., 1986). For any
26 compartment, each of these components has a different transport rate. The total alveolar
27 clearance rate of each material component is the sum of clearance rates of that material from the
28 alveolar to the tracheobronchial, lymph, and blood compartments. In the Strom and Stober
29 models discussed above, the clearance kinetics of DPM were assumed to be entirely dictated by
30 that of the insoluble carbon core. For those organic compounds that get dissociated from the
31 carbon core, clearance rates are likely to be very different, and some of these compounds may be
32 metabolized in the pulmonary tissue or be absorbed by blood.
33 The transport rates were derived from experimental data for rats using several
34 approximations. The transport rates for the carbonaceous core and the organic components were
35 derived by fitting to data from separate experiments. Lung and lymph node burdens from the
§ experiment of Strom et al. (1988) were used to determine the transport rate of the carbon core.
The Yu-Yoon model incorporates the impairment of clearance by including a mass dependency
1175/99 3-31 DRAFT—DO NOT CITE OR QUOTE
-------
1 in the transport rate. This mass dependency is easily extracted because the animals in the
2 experiment were killed over varying periods following the end of exposure.
3 It was assumed that the transport rates from the alveolar and lymph compartments to the
4 blood were equal and independent of the paniculate mass in the alveolar region. The clearance
5 rates of particle-associated organics for rats were derived from the retention data of Sun et al.
6 (1984) for benzo[a]pyrene and the data of Bond et al. (1986) for nitropyrene adsorbed on diesel
7 particles.
8
9 3.5.2.4. Model Assumptions and Extrapolation to Humans
i
10 The Yu-Yoon approach takes the perspective that parsimonious models are to be
11 preferred in order to enable experimental validation and extrapolation from rats to humans.
12 Yu and Yoon make two important assumptions to carry out the extrapolation in the light
13 of inadequate human data. First, the transport rates of organics in the DPM do not change across
14 species. This is based upon lung clearance data of inhaled lipophilic compounds (Schanker et al.,
15 1986), where the clearance was seen to be dependent on the lipid/water partition coefficient. In
16 contrast, the transport rate of the carbon core is considered to be significantly species-dependent
17 (Bailey et al., 1982). DPM clearance fate is determined by two terms in the model (see equation
18 C-82). The first, corresponding to macrophage-mediated clearance, is a function of the lung
19 burden, and is assumed to vary significantly across species. The second term, a constant,
20 corresponding to clearance by dissolution, is assumed to be species-independent. The mass-
21 dependent term for humans is assumed to vary in the same proportion as in rats under the same
22 unit surface particulate dose. The extrapolation is then achieved by using the data of Bailey et al.
23 (1982) for the low lung burden limit of the clearance rate. This value of 0.0017/day was lower
24 than the rat value by a factor of 7.6. This is elaborated further in Appendix C. Other transport
25 rates that have lung burden dependence are extrapolated in the same manner.
26 The Bailey et al. experiment, however, used fused monodisperse aluminosilicate particles
27 of 1.9 and 6.1 um aerodynamic diameters. Yu and Yoon have used the longer of the half-times
28 obtained in this experiment; in using such data for diesel scot particles 0.2 um in diameter,, they
29 have assumed the clearance of insoluble particles to be independent of size over this range. Tnis
30 appears to be a reasonable assumption since the linear dimensions of an alveolar macrophage is
31 significantly larger, roughly 10 um (Yu et al., 1996), However, Snipes (1979) has reported a
32 clearance rate (we convert here from their half-life values) of 0.0022/day for 1 and 2 um particles
33 but a higher value of 0.0039/day for 0.4 um particles. In the absence of reliable data for 0.2 um
34 particles, clearance rate pertaining to a much larger particle size is being used. Although such a
35 choice may underestimate the correct clearance rate for DPM, the resulting error in the human
36 equivalent concentration is likely to be only more protective of human health. Long-term
37 clearance rates for particle sizes more comparable to DPM are available, e.g., iron oxide and
11/5/99 3-32 DRAFT—DO NOT CITE OR QUOTE
-------
1 polystyrene spheres (Waite and Ramsden, 1971; Jammet et al., 1978), but these data show a large
2 range in the values obtained for half-lives or are based upon a very small number of trials, and
therefore compare unfavorably with the quality of data from the Bailey experiment.
The deposition fractions of particulate matter in the pulmonary and tracheobronchial
5 regions of the human lung remain relatively unchanged over the particle size range between
6 0.2 and 1.0 um. Since the clearance of insoluble particles is also likely to remain the same over
7 this range, the dosimetry results in this report for the carbon core component of DPM could also
8 be extended to other particles in this size range within the PM2.5. For particle diameters
9 between 1.0 and 3.5 um, the deposition fraction in the pulmonary region increases significantly
10 (Yu and Diu, 1983), so the diesel model will not be applicable for particles in this range without
11 changing the value for the deposition fractions.
12 Although there was good agreement between experiment and calculated results, this
13 agreement follows a circular logic (as adequately pointed out by Yu and Yoon [L990]) because
14 the same experimental data figured into the derivation of transport rates used in the model.
15 Nevertheless, while this agreement is not a validation, it provides an important consistency check
16 on the model. Thus, the model awaits further experimental data for a reasonable validation.
17 The model showed that at low lung burdens, alveolar clearance is dominated by
18 mucociliary transport to the tracheobronchial region, and at high lung burdens, clearance is
19 dominated by transport to the lymphatic system. The head and tracheobronchial compartments
fl| showed quick clearance of DPM by mucociliary transport and dissolution. Lung burdens of both
21 the carbon core and organics were found to be greater in humans than in rats for similar periods
22 of exposure.
23 The Yu-Yoon publication provides a parametric study of the dosimetry model, examining
24 variation over a range of exposure concentrations, breathing scenarios and ventilation
25 parameters, particle mass median aerodynamic diameters, and geometric standard deviations of
26 the aerosol distribution. It examines how lung burden varies with age for exposure over a life
27 span, provides dosimetry extrapolations to children, and examines changes in lung burden with
28 lung volume. The results showed that children would exhibit more diminished alveolar clearance
29 of DPM at high lung burden than adults when exposed to the equal concentrations of DPM.
30 These features make the model easy to use in risk assessment studies. We refer the reader to
31 Appendix C for further details on the model and for analyses of the sensitivity of the model to
32 change in parameter values.
33 The Yu-Yoon model presents some uncertainties in addition to those discussed earlier in
34 the context of particle size dependence of clearance rate. The Yu and Yoon report underwent
35 extensive peer review; we list below the most important among the model uncertainties discussed
by the review panel. The experimental data used by the Yu-Yoon model for adsorbed organics
used passively adsorbed radiolabeled compounds as surrogates for combustion-derived organics.
11/5/99 3-33 DRAFT—DO NOT CITE OR QUOTE
-------
1 These compounds may adhere differently to the carbon core than those formed during
2 combustion. Yu and Yoon have estimated that slowly cleared organics represent 10% of the total
3 particle mass; the actual figure could be substantially less; the reviewers estimate that the amount
4 of tightly bound organics is probably only 0.1% to 0.25% of the particle mass.
5 The model was based upon the experimental data of Strom et al. (1988) where
6 Fischer-344 rats were exposed to DPM at a concentration of 6.0 mg/m3 for 20 hours/day and 7
7 days/week for periods ranging from 3 to 84 days. Such exposures lead to particle overload effects
8 in rats, whereas human exposure patterns are usually of much lower levels at which overload will
9 not occur. Secondly, human exposures are not likely to be continuous, but most likely over brief
1 0 periods of time. Parameters obtained by fitting to data under the conditions of the experimental
1 1 scenario for rats may not be optimal for the human exposure and concentration of interest.
1 2 The extrapolation of retained dose from rats to humans assumed that the macrophage-
1 3 mediated mechanical clearance of the DPM varies with the specific particulate dose to the
1 4 alveolar surface in the same proportion in humans and in rats, whereas clearance rates by
1 5 dissolution were assumed to be invariant across species. This assumption has not been validated.
16
17 3.5.3. Deposition of Organics
1 8 Using the data presented by Xu and Yu (1 987), it is possible to calculate the total mass of
1 9 DPM, as well as the total organic mass and specific carcinogenic PAHs deposited in the lungs of
20 an individual exposed to DPM. For example, the annual deposition of DPM in the lungs of an
21 individual exposed continuously to 1 |ig/m DPM can be estimated to be about 420 fig. About
22 0.7% of particle mass consists of PAHs (see Section 2.2.6.2, Chapter 2) for a total of 2.94 |ig.
23 Of this amount, the deposited mass of nitro-polycyclic aromatic compounds based on data by
24 Campbell and Lee (1984) would equal 37 ng, while the deposited mass of 7 PAHs that tested
25 positive in cancer bioassays (U.S. EPA, 1993), and measured by Tong and Karasek (1984),
26 would range from 0. 16 to 0.35 \ig. While these amounts are very small, exposure concentrations
27 are often greater than 1 jig/m3, and deposition in humans can be expected to be concentrated at
28 limited sites, especially at the biiurcations of the small bronchi.
29
30 3.6. BIOAVAILABILITY OF ORGANIC CONSTITUENTS PRESENT ON DIESEL
31 EXHAUST PARTICLES
32 Because it has been shown that DPM extract is not only mutagenic but also contains
33 known carcinogens, the organic fraction was originally considered to be the primary source of
35 lacking an organic component, is capable of inducing lung cancer at exposure concentrations
36 sufficient to induce lung particle overload. This suggested 'that the insoluble carbon core of the
37 particle may be of greater importance for the pathogenic and carcinogenic processes observed in
1 1/5/99 3-34 DRAFT— DO NOT CITE OR QUOTE
-------
1 the rat inhalation studies conducted at high exposure concentrations. (See Chapter 7 for a
2 discussion of this issue.) Nevertheless, because lung tumor induction was reported in
epidemiology studies at exposure levels unlikely to induce lung particle overload, it is reasonable
to assume that organic compounds play a role.
5 The bioavailability of toxic organic compounds adsorbed to particles can be influenced by
6 a variety of factors. Although the agent may be active while present on the particle, most
7 particles are taken up by AMs, a cell type not generally considered to be a target site. In order to
8 reach the target site, elution from the particle surface is necessary followed by diffusion and
9 uptake by the target cell. Metabolism to an active form by either the phagocytes or the target
10 cells is also required for activity of many of the compounds present.
11
12 3.6.1. In Vivo Studies
13 3.6.1.1. Laboratory Investigations
14 Several studies reported on the retention of particle-adsorbed organics following
15 administration to various rodent species. In studies reported by Sun et al. (1982, 1984) and Bond
16 et al. (1986), labeled organics were deposited on diesel particles following heating to vaporize
17 the organics originally present. Sun et al. (1982) compared the disposition of either pure or
18 diesel particle-adsorbed benzo[a]pyrene B[a]P following nose-only inhalation by F344 rats.
19 About 50% of particle-adsorbed B[a]P was cleared with a half-time of Ih predominantly by
^P mucociliary clearance. The long-term retention of particle-adsorbed 3H-B[a]P (18 days) was
21 approximately 230-fold greater than that for pure 3H-B[a]P (Sun et al., 1982). At the end of
22 exposure, about 15% of the 3H label was found in blood, liver, and kidney. Similar results were
23 reported in a companion study by Bond et al. (1986), and by Sun et al. (1984) with another PAH,
24 1-nitropyrene, except retention half-time was 36 days.
25 Ball and King (1985) studied the disposition and metabolism of intratracheally instilled
26 14C-labeled 1-NP (>99.9% purity) coated onto DPM. About 50% of the I4C was excreted within
27 the first 24 h; 20% to 30% of this appeared in the urine, and 40% to 60% was excreted in the
28 feces. Traces of radiolabel were detected in the trachea and esophagus. Five percent to 12% of
29 the radiolabel in the lung co-purified with the protein fraction, indicating protein binding of the
30 1-NP-derived 14C. However, the corresponding DNA fraction contained no 14C above
31 background levels.
32 Sevan and Ruggio (1991) assessed the bioavailability of B[a]P adsorbed to DPM from a
33 5.7-L Oldsmobile engine. In this study, exhaust particles containing 1.03 jig B[a]P/g particles
34 were supplemented with exogenous 3H-B[a]P to provide 2.62 jig B[a]P/g of exhaust particles.
35 In vitro analysis indicated that the supplemented B[a]P eluted from the particles at the same rate
as the original B[a]P. Twenty-four hours after intratracheal instillation in Sprague-Dawley rats,
68.5% of the radiolabel remained in the lungs. This is approximately a 3.5-fold greater
11/5/99 3-35 DRAFT—DO NOT CITE OR QUOTE
-------
1 proportion than that reported by Sun et al. (1984), possibly because smaller amounts of B[a]P •
2 adsorbed on the particles, resulting in stronger binding. At 3 days following administration, over
3 50% of the radioactivity remained in the lungs, nearly 30% had been excreted into the feces, and
4 the remainder was distributed throughout the body. Experiments using rats with cannulated bile
5 ducts showed that approximately 10% of the administered radioactivity appeared in the bile over
6 a 10-h period and that less than 5% of the radioactivity entered the feces via mucociliary
7 transport. Results of these studies showed that the retention of organics in the lungs is increased
8 considerably when organics are adsorbed to diesel particles. Because retention time is very short
9 following exposure to the pure compounds, it can be concluded that the increased retention time
10 is primarily the result of continued binding to the particles. The detection of labeled compounds
11 in blood, distant organs, urine, and bile as well as the trachea, however, provides evidence that at
12 least some of the organics are eluted from the particles following deposition in the lungs.
13
14 3.6.1.2. Studies in Occupationally Exposed Humans
15 DNA adduct induction in the lungs of experimental animals exposed to diesel exhaust
16 have been measured in a number of animal experiments (see World Health Organization [ 1996]
17 for a review). Such studies, however, provide limited information regarding bioavailability of
18 organics, since positive results may well have been related to factors associated with lung particle
19 overload. In fact, Bond et al. (1990) reported that carbon black, which is virtually devoid of
20 organics, is capable of inducing DNA adducts in rats at lung overload doses.
21 On the other hand, DNA adduct formation and/or mutations in blood cells following
22 exposure to DPM, especially at levels insufficient to induce lung overload, can be presumed to be
23 the result of organics diffusing into the blood. Hemminki et al. (1994) reported increased levels
24 of DNA adducts in lymphocytes of bus maintenance and truck terminal workers. Osterholm
25 et al. (1995) studied mutations at the hprt-locus of T-lymphocytes in bus maintenance workers.
26 Although they were unable to identify clearcut exposure-related differences in types of
27 mutations, adduct formation was significantly increased in the exposed workers. Nielsen et al.
28 (1996) reported significantly increased levels of lymphocyte DNA adducts, hydrcxyvaline
29 adducts in hemoglobin, and 1-hydroxypyrene in urine of garage workers exposed to diesel
30 exhaust.
31
32 3.6.2. In Vitro Studies
33 3.6.2.1. Extraction of Diesel Particle-Associated Organics by Biological Fluids
34 In vitro extraction of organics in biological fluids can be estimated by measurement of
35 mutagenic activity. Using this approach, Brooks et al. (1981) reported extraction efficiencies of
36 only 3% to 10 % that of dichloromethane following DPM incubation in lavage fluid, serum,
37 saline, albumin, dipalmitoyl lecithin, or dichloromethane. Moreover, extraction efficiency did
11/5/99 3-36 DRAFT—DO NOT CITE OR QUOTE
-------
1 not increase with incubation time up to 120 h. Similar findings were reported by King et al.
2 (1981). In the latter study, lung lavage fluid and lung cytosol fluid extracts of DPM were not
mutagenic. Serum extracts of DPM did exhibit some mutagenic activity, but considerably less
than that for organic solvent extracts. Furthermore, the mutagenic activity of the solvent extract
5 was significantly reduced when combined with serum or lung cytosol fluid, suggesting protein
6 binding or biotransformation of the mutagenic components. Siak et al. (1980) assessed the
7 mutagenicity of material extracted from DPM by bovine serum albumin in solution, simulated
8 lung surfactant, fetal calf serum (PCS), and physiologic saline. Only PCS was found to extract
9 some mutagenic activity from the DPM.
10 Despite the apparent inability of biological fluids to extract organics in vitro, their
11 effectiveness in vivo remains equivocal because of differing conditions. Extracellular lung fluid
12 is a complex mixture of constituents that undoubtedly have a broad range of hydrophobicity
13 (George and Hook, 1984; Wright and Clements, 1987), and it fundamentally differs from serum
14 in terms of chemical composition (Gurley et al., 1988). Moreover, assessments of the ability of
15 lavage fluids, which actually represent substantially diluted extracellular lung fluid, to extract
16 mutagenic activity from DPM clearly do not reflect the in vivo condition. Finally, except under
17 very high exposure concentrations, few particles escape phagocytosis and possible intracellular
18 extraction.
19
^P 3.6.2.2. Extraction of Diesel Particle-Associated Organics by Lung Cells and Cellular
21 Components
22 A more likely means by which organic carcinogens (e.g., PAHs) may be extracted from
23 DPM and metabolized in the lung is either particle dissolution or extraction of organics from the
24 particle surface within the phagolysosomes of AMs and[other lung cells. This mechanism
25 presupposes that the particles are internalized. Specific details about the physicochemical
26 conditions of the intraphagolysosomal environment, where particle dissolution in AMs
27 presumably occurs in vivo, have not been well characterized. However, it is known that the
28 phagolysosomes constitute an acidic (pH 4 to 5) compartment in macrophages (Nilsen et al.,
29 1988; Ohkuma and Poole, 1978). The relatively low pH in the phagolysosomes has been
30 associated with the dissolution of some types of inorganic particles (some metals) by
31 macrophages (Marafante et al., 1987; Lundborg et al., 1984), but few studies provide quantitative
32 information concerning how organic constituents of DPM (e.g., B[a]P) may be extracted in the
33 phagolysosomes (Bond et al., 1983). Whatever the mechanism, assuming elution occurs, the end
34 result is a prolonged exposure of the respiratory epithelium to low concentrations of carcinogenic
35 agents.
Early studies by King et al. (1981) found that when pulmonary alveolar macrophages
were incubated with DPM, amounts of organic compounds and mutagenic activity decreased
11/5/99 3-37 DRAFT—DO NOT CITE OR QUOTE
-------
1 measurably from the amount originally associated with the particles, suggesting that organics
2 were removed from the phagocytized particles. Leung et al. (1988) studied the ability of rat lung
3 and liver microsomes to facilitate transfer and metabolism of B[a]P from diesel particles. I4C-
4 B[a]P coated diesel particles, previously extracted to remove the original organics present, were
5 incubated with liver or lung microsomes. About 3% of the particle adsorbed B[a]P was
6 transferred to the lung microsomes within 2 h. Of this amount about 1.5% was metabolized, for
7 a total of about 0.05% of the B[a]P adsorbed to the DPM. While transformation is slow, because
8 of long retention half-lives of particles in humans the fraction eluted and metabolized may well
9 be significant.
10 In analyzing phagolysosomal dissolution of various ions from particles in the lungs of
11 Syrian golden hamsters, however, Godleski et al. (1988) demonstrated that solubilization did not
12 necessarily result in clearance of the ions and that binding of the solubilized components to
13 cellular and extracellular structures occurred. It is reasonable to assume that phagocytized DPM
14 particles may be subject to similar processes and that these processes would be important in
15 determining the rate of bioavailability of the particle-bound constituents of DPM.
16
17 3.6.3. Modeling Studies
18 Gerde et al. (1991 a,b) described a model simulating the effect of particle aggregation and
19 PAH content on the rate of PAH release in the lung. According to this model, particle
20 aggregation will occur with high exposure concentrations, resulting hi a slow release of PAHs
21 with prolonged exposure to surrounding tissues. However, large aggregates of inert dust are
22 unlikely to form at doses typical of human exposures. Inhaled particles, at low concentrations,
23 are more likely to deposit and react with surrounding lung medium without interference from
24 other particles. The model predicts that, under low-dose exposure conditions more typical in
25 humans, particle-associated organics will be released more rapidly from the particles because
26 they are not aggregated. Sustained exposure of target tissues to PAHs will result from repeated
7,7 exposures, not from increased retention due to association of PAHs with carrier particles. This
28 distinction is important because at low doses PAII exposure and lung tumor formation should
29 occur at sites of deposition rather than retention as occurs with high doses.
30 The site of release of PAHs influences effective dose to the lungs because, as noted
31 previously, at least some free organic compounds such as B[a]P deposited in the lungs are
32 rapidly absorbed into the bloodstream. Gerde ct al. (1991b) predicted lipophilic PAHs would be
33 retained in the alveoli less than 1 min, whereas they may be retained for hours in the bronchi.
34 These predictions were based on an average diffusion distance to capillaries of only about 0.5 (am
35 in the alveoli, whereas in the bronchi it probably exceeds 50 urn An experimental study by
36 Gerde et al. (1999) provided support for this prediction. Beagle dogs were exposed to 3H-B[a]P
37 adsorbed on the carbonaceous core of diesel particles at a concentration of 15 ug B[a]P/gm
11/5/99 3-38 DRAFT—DO NOT CITE OR QUOTE
-------
1 particles. A rapidly eluting fraction from particles deposited in the alveoli was adsorbed into the
2 bloodstream and metabolized in the liver. The rapidly eluting fraction from particles deposited in
the conducting airways, however, was to a large extent retained and metabolized in airway
epithelium.
5 Nikula et al. (1997) reported that 52% of DPM deposited in the lungs of Cynomolgus
6 monkeys chronically inhaling diesel exhaust were found in the interstitium of small airways,
7 compared with 27% in rats (Nikula et al., 1997). This is primarily due to a lack of respiratory
8 bronchioles in the rat. Because lung structure is similar in monkeys and humans, a significantly
9 greater percentage of DPM matter can be predicted to deposit in small airway branches of
10 humans. Overall, bioavailability of organics in humans is expected to be greatest at bifurcations
11 of small airways because they are a major site of particle deposition, have a longer residence time
12 for eluted organics in airways than alveoli, allow time for uptake and metabolism by airway
13 epithelium, and predict more rapid elution from particles at ambient exposure concentrations.
14 Overall, the results of studies presented in Section 3.6 provide evidence that at least some
15 of the organic matter adsorbed to DPM deposited in the lungs is eluted. However, the percentage
16 taken up and metabolized to an active form by target cells is uncertain. Organics eluted from
17 particles deposited in alveoli are likely to rapidly enter the bloodstream and pose little risk for
18 induction of lung pathology and/or cancer. Risk of harmful effects for particles deposited in the
19 small airways is predicted to be greater because solubilized organic compounds will be retained
^B longer, allowing for metabolism by epithelial cells lining the airways. Since the deposition in
21 small airways occurs primarily at bifurcations, localized higher concentrations will occur. At
22 present, unfortunately, the available data are insufficient to accurately model the
23 effective dose of organics in the respiratory tract of humans or animals exposed to diesel exhaust,
24 especially at specific target sites such as small airway branchings.
25
26 3.7. SUMMARY
27 The most consistent historical measure of dose for diesel emissions is DPM in units of jig
28 particles/m3. With the assumption that all components of diesel emissions (e.g., organics in the
29 form of volatilized liquids or gases) are present in proportion to the DPM mass, DPM can be
30 used as the basic dosimeter for effects from various scenarios including acute and chronic
31 exposures, as well as for different endpoints such as irritation, fibrosis, or even cancer. There is,
32 however, little evidence currently available to prove or refute DPM as being the most appropriate
33 dosimeter.
34 The DPM dose to the tissue is related to the extent of the deposition and clearance of
35 DPM. Diesel exhaust particles may deposit throughout the respiratory tract via sedimentation or
t diffusion, with the latter being prevalent in the alveolar region. Particles that deposit upon
airway surfaces may be cleared from the respiratory tract completely or may be translocated to
11/5/99 3-39 DRAFT—DO NOT CITE OR QUOTE
-------
1 other sites by regionally distinct processes that can be categorized as either absorptive (i.e.,
2 dissolution) or nonabsorptive (i.e., transport of intact particles via mucociliary transport). With
3 insoluble or poorly soluble particles such as DPM, clearance by dissolution is insignificant
4 compared to the rate of clearance as an intact particle. Another mechanism that can affect
5 retention of DPM is endocytosis by alveolar lining cells that, in turn, can lead to the
6 accumulation of DPM in the interstitial compartment of the lung and subsequent translocation of
7 DPM to lymph nodes. For poorly soluble particles such as DPM, species-dependent rate
8 constants exist for the various clearance pathways that can be modified by factors such as
9 respiratory tract disease.
10 In rats, prolonged exposure to high particle concentrations may be associated with what is
11 termed particle overload. This condition is defined as the overwhelming of macrophage-
12 mediated clearance by the deposition of particles at a rate exceeding the capacity of that
13 clearance pathway, occurring in rats when deposition approaches 1 mg particles/g lung tissue.
14 The relevance of lung overload to humans, and even to species other than laboratory rats and
15 mice, is problematic. Relevance to humans is further clouded by the suggestion that
16 macrophage-mediated clearance is normally slower and perhaps less important in humans than in
17 rats. Whereas such clearance is likely to be of little relevance for most "real-world" ambient
18 exposures of humans, it is of concern in interpreting some long-term experimental exposure data
19 and perhaps for some human occupational exposures.
20 Extrapolation of lexicological results from laboratory animals to humans to obtain an
21 HEC requires the use of a dosimetry model incorporating critical aspects of both species that
22 include (1) the deposition of DPM in various regions of the respiratory tract, and (2) the transport
23 and clearance of the particles from their deposited sites. Review and evaluation of the models
24 available led to the choice of the Yu and Yoon( 1990) model. This model has a three-
25 compartment lung consisting of tracheobronchial, alveolar, and lymph node compartments and,
26 in addition, considers filtration by a nasopharyngeal or head compartment. Absorption by the
27 blood and gastrointestinal compartments was also considered. In addition, the model treats DPM
28 as being composed of the insoluble carbonaceous core, slowly cleared organics, and fast-cleared
29 organics. Major assumptions made in this model include that transport rates of organics in DPM
30 do not change across species and that the transport rate of the carbonaceous core is species
31 dependent such that the clearance varies with specific dose to the alveolar surface in the same
32 proportion in humans and in rats. This model was used to project HECs from concentrations
33 used in experimental animal exposures. Use of HECs partially obviates the need for an animal-
34 to-human uncertainty factor, as explained in Chapter 9.
35 The degree of bioavailability of the organic fraction of DPM is still somewhat uncertain.
36 However, reports of DNA induction in occupationally exposed workers as well as results of
37 animal studies using radiolabeled organics deposited on diesel particles indicate that at least a
1175/99 3-40 DRAFT—DO NOT CITE OR QUOTE
-------
1 fraction of the organics present are eluted prior to particle clearance. In addition, data have been
2 presented indicating that a greater percentage of diesel particles are deposited in the branching of
small airways of laboratory primates, and presumably of humans, than in those of rats.
Carcinogenic organics eluted in this region remain in the lung long enough to be metabolized to
5 an active form. Some of the toxicologically significant compounds, however, do not require
6 metabolic activation. While adequate quantitative data are lacking, they do suggest the likely
7 involvement of particle-associated organics in the carcinogenic process.
8 3.8. REFERENCES
i
9 Adamson, IYR; Bowden, DH. (1978) Adaptive responses of the pulmonary macrophagic system to carbon: II.
10 Morphologic studies. Lab Invest 38:430-438.
11 Adamson, IYR; Bowden, DH. (1981) Dose response of the pulmonary macrophagic system to various particulates
12 and its relationship to transepithelial passage of free particles. Exp Lung Res 2:165-175.
13 Anjilvel, S; Asgharian, B. (1995) A multiple-path model of particle deposition in the rat lung. Fundam Appl Toxicol
14 28:41-50.
15 Aronson, M. (1963) Bridge formation and cytoplasmic flow between phagocytic cells. J Exp Med 118:1083-1088.
16 Bailey, MR; Fry, FA; James, AC. (1982) The long-term clearance kinetics of insoluble particles from the human
17 lung. Ann Occup Hyg 26:273-290.
Ball, LM; King, LC. (1985) Metabolism, mutagenicity, and activation of 1-nitropyrene in vivo and in vitro. Environ
Int 11:355-361.
20 Battigelli, MC; Hengstenberg, F; Mannela, RJ; et al. (1966) Mucociliary activity. Arch Environ Health 12:460-466.
21 Bevan, DR; Ruggio, DM. (1991) Bioavailability in vivo of benzo[a]pyrene adsorbed to diesel particulate. Toxicol
22 Ind Health 7:125-139.
23 Bohning, DE; Conn, SH; Lee, HD; et al. (1980) Two-phase deep-lung clearance in man. In: Sanders, CL; Cross,
24 FT; Dagle, GE; et al., eds. Pulmonary toxicology of respirable particles: proceedings of the nineteenth annual
25 Hanford life sciences symposium; October 1979; Richland, WA. Washington, DC: U.S. Department of Energy; pp.
26 149-161. Available from: NTIS, Springfield, VA; CONF-791002.
27 Bohning, DE; Atkins, HL; Conn, SH. (1982) Long-term particle clearance in man: normal and impaired. In: Walton,
28 WH, ed. Inhaled particles V: proceedings of an international symposium; September 1980; Cardiff, Wales. Ann
29 Occup Hyg 26:259-271.
30 Bond, JA; Mitchell, CE; Li, AP. (1983) Metabolism and macromolecular covalent binding of benzo[a]pyrene in
31 cultured Fischer-344 rat lung type II epithelial cells. Biochem Pharmacol 32:3771-3776.
32 Bond, JA; Sun, JD; Medinsky, MA; et al. (1986) Deposition, metabolism, and excretion of l-[MC]nitropyrene and
33 l-['4C]nitropyrene coated on diesel exhaust particles as influenced by exposure concentration. Toxicol Appl
34 Pharmacol 85:102-117.
35 Bond, JA; Mauderly, JL; Wolff, RK. (1990) Concentration- and time-dependent formation of DNA adducts in lungs
of rats exposed to diesel exhaust. Toxicology 60:127-135.
i
Brain, JD; Mensah, GA. (1983) Comparative toxicology of the respiratory tract. Am Rev Respir Dis 128:S87-S90.
11/5/99 3-41 DRAFT—DO NOT CITE OR QUOTE
-------
1 Brooks, AL; Wolff, RK; Royer, RE; et al. (1981) Deposition and biological availability of diesel particles and their
2 associated mutagenic chemicals. Environ Int 5:263-267.
3 Campbell, RM; Lee, ML. (1984) Capillaary column gas chromatographic determination of nitro polycyclic
4 aromatic compounds in paniculate extracts. Anal Chem 56:1026-1030.
5 Chan, TL; Lee, PS; Hering, WE. (1981) Deposition and clearance of inhaled diesel exhaust particles in the
6 respiratory tract of Fischer rats. J Appl Toxicol 1:77-82.
7 Chan, TL; Lee, PS; Hering, WE. (1984) Pulmonary retention of inhaled diesel particles after prolonged exposures to
8 diesel exhaust. Fundam Appl Toxicol 4:624-631.
9 Cohen, D; Arai, SF; Brain, JD. (1979) Smoking impairs long-term dust clearance from the lung. Science
10 (Washington, DC) 204:514-517.
11 Cohen, BS; Xiong, JQ; Fang, CP; et al. (1998) Deposition of charged particles on lung airways. Health Phys
12 74:554-560.
13 Crapo, JD; Young, SL; Fram, EK; et al. (1983) Morphometric characteristics of cells in the alveolar region of
14 mammalian lungs. Am Rev Respir Dis 128:S42-S46.
15 Cuddihy, RG; Yeh, HC. (1986) Model analysis of respiratory tract clearance of particles inhaled by people. In:
16 Muggenburg, BA; Sun, JD, eds. Annual report of the Inhalation Toxicology Research Institute. Albuquerque, NM:
17 Lovelace Biomedical and Environmental Research Institute; report no. LMF-115; pp. 140-147.
18 Dahl, AR; Schlesinger, RB; Heck, HD' A; et al. (1991) Comparative dosimetry of inhaled materials: differences
19 among animal species and extrapolation to man. Fundam Appl Toxicol 16:1-13.
20 Evans, MJ; Shami, SG; Martinez, LA. (1986) Enhanced proliferation of pulmonary alveolar macrophages after
21 carbon instillation in mice depleted of blood monocytes by Strontium-89. Lab Invest 54:154-159.
22 Felicetti, SA; Wolff, RK; Muggenburg, BA. (1981) Comparison of tracheal mucous transport in rats, guinea pigs,
2 3 rabbits, and dogs. J Appl Physiol: Respir Environ Exercise Physiol 51:1612-1617.
24 Ferin, J. (1982) Pulmonary alveolar pores and alveolar macrophage-mediated particle clearance. Anat Rec
25 203:265-272.
26 Ferin, J; Feldstein, ML. (1978) Pulmonary clearance and hilar lymph node content in rats after particle exposure.
27 Environ Res 16:342-352.
28 Freedman, AP; Robinson, SE. (1988) Noninvasive magnetopneumographic studies of lung dust retention and
29 clearance in coal miners. In: Frantz, RL; Ramani, RV, eds. Respirable dust in the mineral industries: health effects,
30 characterization, and control: proceedings of the international symposium on respirable dust in the mineral
31 industries; October 1986; University Park, FA. University Park, PA: Pennsylvania State University Press; pp.
32 181-186.
33 Frey, JW; Com, M. (1967) Physical and chemical characteristics of particulates in a diesel exhaust. -Am Ind Hyg
34 Assoc J 28:468-478.
35 George, G; Hook, GER. (1984) The pulmonary extracellular lining. Environ Health Perspect 55:227-237.
36 Gerde. P: Medinsky. MA: Bond. JA. (1991a) Particle-associated polycyclic aromatic hydrocarbons - a reappraisal of
37 their possible role in pulmonary carcinogenesis. Toxicol Appl Pharmacol 108:1-13.
38 Gerde, P; Medinsky, MA; Bond, JA. (1991b) The retention of polycyclic aromatic hydrocarbons in the bronchial
39 airways and in the alveolar region - a theoretical comparison. Toxicol Appl Pharmacol 107:239-252.
11/5/99 3-42 DRAFT—DO NOT CITE OR QUOTE
-------
1 Gerde, P; Muggenberg, BA; Dahl, AR. (1999) Bioavailability, absorption and metabolism of diesel soot-absorbed
2 benzo(a)pyrene after single-breath exposures in dogs. In: Relationships between acute and chronic effects of air
3 pollution: 7th international inhalation symposium; February; Hannover, Federal Republic of Germany; p. 76.
Gibb, FR; Morrow, PE. (1962) Alveolar clearance in dogs after inhalation of an iron 59 oxide aerosol. J Appl
Physipl 17:429-432.
6 Godleski, JJ; Stearns, RC; Katler, MR; et al. (1988) Particle dissolution in alveolar macrophages assessed by
7 electron energy loss analysis using the Zeiss CEM902 electron microscope. J Aerosol Med 1:198-199.
8 Griffis, LC; Wolff, RK; Henderson, RF; et al. (1983) Clearance of diesel soot particles from rat lung after a
9 subchronic diesel exhaust exposure. Fundam Appl Toxicol 3:99-103.
10 Gurley, LR; Spall, WD; Valdez, JG; et al. (1988) An HPLC procedure for the analysis of proteins in lung lavage
11 fluid. Anal Biochem 172:465-478.
12 Harmsen, AG; Muggenburg, BA; Snipes, MB; et al. (1985) The role of macrophages in particle translocation from
13 lungs to lymph nodes. Science (Washington, DC) 230:1277-1280.
14 Health Effects Institute. (1995) Diesel exhaust: a critical analysis of emissions, exposure, and health effects: a
15 special report of the Institute's Diesel Working Group. Cambridge, MA: Health Effects Institute.
16 Heinrich, U; Muhle, H; Takenaka, S; et al. (1986) Chronic effects on the respiratory tract of hamsters, mice, and
17 rats after long-term inhalation of high concentrations of filtered and unfiltered diesel engine emissions. J Appl
18 Toxicol 6:383-395.
19 Hemminki, K; So'derling, J; Ericson, P; et al. (1994) DNA adducts among personnel servicing and loading diesel
20 vehicles. Carcinogenesis 15:767-769.
Heppleston, AG. (1961) Observations on the disposal of inhaled dust by means of the double exposure technique.
In: Davies, CN, ed. Inhaled particles and vapours: proceedings of an international symposium; March-April 1960;
23 Oxford, United Kingdom. New York: Pergamon Press; pp. 320-326.
24 Heppleston, AG; Young, AE. (1973) Uptake of inert paniculate matter by alveolar cells: an ultrastructural study. J
25 Pathol 111:159-164.
26 Heyder, J; Gebhart, J; Rudolf, G; et al. (1986) Deposition of particles in the human respiratory tract in the size
27 range 0.005-15 "mu"m. J Aerosol Sci 17:811-825.
28 Holt, PF. (1981) Transport of inhaled dust to extrapulmonary sites. J Pathol 133:123-129.
29 International Commission on Radiological Protection. (1979) Limits for intakes of radionuclides by workers.
30 Oxford, United Kingdom: Pergamon Press; ICRP publication 30, part 1.
31 International Commission on Radiological Protection. (1994) Human respiratory tract model for radiological
32 protection: a report of a task group of the International Commission on Radiological Protection. Oxford, United
33 Kingdom: Elsevier Science Ltd. (ICRP publication 66; Annals of the ICRP: v. 24, nos. 1-3).
34 Iwai, K; Udagawa, T; Yamagishi, M; et al. (1986) Long-term inhalation studies of diesel exhaust on F344 SPF rats.
35 Incidence of lung cancer and lymphoma. In: Ishinishi, N; Koizumi, A; McClellan, RO; et al., eds. Carcinogenic and
36 mutagenic effects of diesel engine exhaust: proceedings of the international satellite symposium on toxicological
37 effects of emissions from diesel engines; July; Tsukuba Science City, Japan. Amsterdam: Elsevier Science
38 Publishers B. V.; pp. 349-360. (Developments in toxicology and environmental science: v. 13).
Jammet, H; Drutel, P; Parrot, R; et al. (1978) Etude de Pepuration pulmonaire chez Phomme apres administration
d'aerosols de particules radioactives [Study of pulmonary function in man after administration of radioactive
particulate aerosols]. Radioprotection 13:143-166.
11/5/99 3-43 DRAFT—DO NOT CITE OR QUOTE
-------
1 King, LC; Kohan, MJ; Austin, AC; et al. (1981) Evaluation of the release of mutagens from diesel particles in the
2 presence of physiological fluids. Environ Mutagen 3:109-121.
3 Kittelson, DB; Dolan, DF; Verrant, JA. (1978) Investigation of a diesel exhaust aerosol. Warrendale, PA: Society of
4 Automotive Engineers, Inc.; technical paper series no. 78109.
5 Lee, PS; Chan, TL; Hering, WE. (1983) Long-term clearance of inhaled diesel exhaust particles in rodents. J
6 Toxicol Environ Health 12:801-813.
7 Lee, KP; Trochimowicz, HJ; Reinhardt, CF. (1985) Transmigration of titanium dioxide (TiOJ particles in rats after
8 inhalation exposure. Exp Mol Pathol 42:331-343.
9 Lee, KP; Henry, NW, III; Trochimowicz, HJ.; et al. (1986) Pulmonary response to impaired lung clearance in rats
10 following excessive TiO2 dust deposition. Environ Res 41:144-167.
11 Lee, KP; Ulrich, CE; Geil, RG; et al. (1988) Effects of inhaled chromium dioxide dust on rats exposed for two
12 years. Fundam Appl Toxicol 10:125-145.
13 Lehnert, BE. (1988) Distributions of particles in alveolar macrophages during lung clearance. J Aerosol Med
14 1:206-207.
15 Lehnert, BE. (1989) Rates of disappearance of alveolar macrophages during lung clearance as a function of
16 phagocytized paniculate burden [abstract]. Am Rev Respir Dis 139(suppl.):A161.
17 Lehnert, BE. (1990) Alveolar macrophages in a particle "overload" condition. J Aerosol Med 3(suppl. 1): S9-S30.
18 Lehnert, BE; Morrow, PE. (1985) Association of 59iron oxide with alveolar macrophages during alveolar clearance.
19 Exp Lung Res 9:1-16.
20 Lehnert, BE; Valdez, YE; Bomalaski, SH. (1988) Analyses of particles in the lung free cell, tracheobronchial lymph
21 nodal, and pleural space compartments following their deposition in the lung as related to lung clearance
22 mechanisms. In: Dodgson, J: McCallum, RI; Bailey, MR; et al., eds. Inhaled particles VI: proceedings of an
23 international symposium and workshop on lung dosimetry; September 1985; Cambridge, United Kingdom. Ann
24 Occup Hyg 32 (suppl. 1):125-140.
25 Lehnert, BE; Cline, A; London, JE. (1989) Kinetics of appearance of polymorphonuclear leukocytes and their
26 particle burdens during the alveolar clearance of a high lung burden of particles. Toxicologist 9:77.
27 Leung, HW; Henderson, RF; Bond, JA; et al. (1988) Studies on the ability of rat lung and liver microsomes to
2 8 facilitate transfer and metabolism of benzo[a]pyrene from diesel particles. Toxicology 51:1-9.
29 Lippmann, M; Schlesinger, RB. (1984) Interspecies comparisons of particle deposition and mucociliary clearance in
30 tracheobronchial airways. J Toxicol Environ Health 13:441 -469.
31 Lundborg, M; Lind, B; Camner, P. (1984) Ability of rabbit alveolar macrophages to dissolve metals. Exp Lung Res
32 7:11-22.
33 Marafante, E; Lundborg, M; Vahter, M; et al. (1987) Dissolution of two arsenic compounds fay rabbit alveolar
34 macrophages in vitro. Fundam Appl Toxicol 8:382-388.
35 Mauderly, JL. (1996) Lung overload: the dilemma and opportunities for resolution. In: Mauderly, JL; McCunney,
36 RJ, eds. Particle overload in the rat lung and lung cancer: implications for human risk assessment. Proceedings of a
37 conference; March 1995; Cambridge, MA. Taylor & Francis.
38 McClellan, RO; Brooks, AL; Cuddihy, RG; et al. (1982) Inhalation toxicology of diesel exhaust particles. In:
39 Lewtas, J, ed. Toxicologicai effects of emissions from diesel engines: proceedings of the EPA diesel emissions
11/5/99 3-44 DRAFT—DO NOT CITE OR QUOTE
-------
1 symposium; October 1981; Raleigh, NC. New York: Elsevier Biomedical; pp. 99-120. (Developments in toxicology
2 and environmental science: v. 10).
Morrow, PE. (1966) International Commission on Radiological Protection (ICRP) task group on lung dynamics,
deposition and retention models for internal dosimetry of the human respiratory tract. Health Phys 12:173.
5 Morrow, PE. (1988) Possible mechanisms to explain dust overloading of the lungs. Fundam Appl Toxicol
6 10:369-384.
7 Morrow, PE. (1994) Mechanisms and significance of "particle overload." In: Mohr, U; Dungworth, DL; Mauderly,
8 JL; et al., Toxic and carcinogenic effects of solid particles in the respiratory tract: [proceedings of the 4th
9 international inhalation symposium]; March 1993; Hannover, Germany. Washington, DC: International Life
10 Sciences Institute Press; pp. 17-25.
11 Morrow, PE; Yu, CP. (1993) Models of aerosol behavior in airways and alveoli. In: Moren, F; Dolovich, MB;
12 Newhouse, MT; et al., eds. Aerosols in medicine: principles, diagnosis and therapy. 2nd rev. ed. Amsterdam:
13 Elsevier; pp. 157-193.
14 Muhle, H; Bellman, B; Creutzenberg, O; et al. (1990) Dust overloading of lungs after exposure of rats to particles
15 of low solubility: comparative studies. J Aerosol Sci 21:374-377.
16 Nielsen, PS; Andreassen, A; Farmer, PB; et al. (1996) Biomonitoring of diesel exhaust-exposed workers. DNA and
17 hemoglobin adducts and urinary 1-hydroxypyrene as markers of exposure. Toxicol Lett 86:27-37.
18 Nikula, KJ; Avila, K.J; Griffith, WC; et al. (1997) Sites of particle retention and lung tissue responses to chronically
19 inhaled diesel exhaust and coal dust in rats and cynomolgus monkeys. In: Driscoll, KE; Oberdorster, G, eds.
20 Proceedings of the sixth international meeting on the toxicology of natural and man-made fibrous and non-fibrous
21 particles; September 1996; Lake Placid, NY. Environ Health Perspect Suppl 105(5): 1231-1234.
Nilsen, A; Nyberg, K; Camner, P. (1988) Intraphagosomal pH in alveolar macrophages after phagocytosis in vivo
'3 and in vitro of fluorescein-labeled yeast particles. Exp Lung Res 14:197-207.
24 Oberd6rster, G. (1988) Lung clearance of inhaled insoluble and soluble particles. J Aerosol Med 1:289-330.
25 Oberdorster, G; Ferin, J; Morrow, PE. (1992) Volumetric loading of alveolar macrophages (AM): a possible basis
26 for diminished AM-mediated particle clearance. Exp Lung Res 18:87-104.
27 Ohkuma, S; Poole, B. (1978) Fluorescence probe measurement of the intralysosomal pH in living cells and the
28 perturbation of pH by various agents. Proc Nat. Acad Sci U S A 75:3327-3331.
29 Osterholm, AM; Fa'lt, S; Lambert, B; et al. (1995) Classification of mutations at the human hprt-Iocus in
30 T-lymphocytes of bus maintenance workers by multiplex-PCR and reverse transriptase-PCR analysis.
31 Carcinogenesis 16:1909-1995.
32 Pepelko, WE. (1987) Feasibility of dose adjustment based on differences in long-term clearance rates of inhaled
33 particulate matter in humans and laboratory animals. Regul Toxicol Pharmacol 7:236-252.
34 Phalen, RF; Oldham, MJ. (1983) Tracheobronchial airway structure as revealed by casting techniques. Am Rev
35 RespirDisl28:Sl-S4.
36 Powdrill, J; Buckley, C; Valdez, YE; et al. (1989) Airway intra-luminal macrophages: origin and role in lung
37 clearance. Toxicologist 9:77.
38 Pritchard, JN. (1989) Dust overloading causes impairment of pulmonary clearance: evidence from rats and humans.
Exp Pathol 37:39-42.
11/5/99 3-45 DRAFT—DO NOT CITE OR QUOTE
-------
1 Raabe, OG. (1982) Deposition and clearance of inhaled aerosols. In: Witschi, H, ed. Mechanisms in respiratory
2 toxicology. Boca Raton, FL: CRC Press; pp. 27-76.
3 Raabe, OG; Al-Bayati, MA; Teague, SV; et al. (1988) Regional deposition of inhaled monodisperse, coarse, and
4 fine aerosol particles in small laboratory animals. In: Dodgson, J; McCallum, RI; Bailey, MR; et al., eds. Inhaled
5 particles VI: proceedings of an international symposium and workshop on lung dosimetry; September 1985;
6 Cambridge, United Kingdom. Ann Occup Hyg 32(suppl. l):53-63.
7 Riley, PA; Dean, RT..(1978) Phagocytosis of latex particles in relation to the cell cycle in 3T3 cells. Exp Cell Biol
8 46:367-373.
9 Robertson, B. (1980) Basic morphology of the pulmonary defence system. Eur J Respir Dis 61(suppl. 107):21-40.
10 Sandusky, CB; Cowden, MW; Schwartz, SL. (1977) Effect of particle size on regurgitative exocytosis by rabbit
11 alveolar macrophages. In: Sanders, CL; Schneider, RP; Dagle, GE; et al., eds. Pulmonary macrophage and epithelial
12 cells: proceedings of the sixteenth annual Hanford biology symposium; September 1976; Richland, WA.
13 Washington, DC: Energy Research and Development Administration; pp. 85-105. (ERDA symposium series: no.
14 43). Available from: NTIS, Springfield, VA; CONF-760927.
15 Schanker, LS; Mitchell, EW; Brown, RA, Jr. (1986) Species comparison of drug absorption from the lung after
16 aerosol inhalation or intratracheal injection. Drug Metab Dispos 14:79-88.
17 Schauer, JJ; Rogge, WF; Hildemann, LM; et al. (1996) Source apportionment of airborne paniculate matter using
18 organic compounds as tracers. Atmos Environ 30:3837-3855.
19 Schlesinger, RB. (1985) Comparative deposition of inhaled aerosols in experimental animals and humans: a review.
20 J Toxicol Environ Health 15:197-214.
21 Schlesinger, RB; Ben-Jebria, A; Dahl, AR; et al. (1997) Disposition of inhaled toxicants. In: Massaro, EJ, ed.
22 Handbook of human toxicology. Boca Raton, FL: CRC Press; pp. 493-550.
23 Schum, M; Yen, HC. (1980) Theoretical evaluation of aerosol deposition in anatomical models of mammalian lung
24 airways. Bull Math Biol 42:1-15.
25 Siak, JS; Chan, TL; Lee, PS. (1980) Diesel paniculate extracts in bacterial test systems. In: Pepelko, WE; Danner,
26 RM; Clarke, NA, eds. Health effects of diesel engine emissions: proceedings of an international symposium, v. 1;
27 December 1979; Cincinnati, OH. Cincinnati, OH: U.S. EPA, Health Effects Research Laboratory; pp. 245-262;
28 EPA report no. EPA/600/9-80-057b. Available from: NTIS, Springfield, VA; PB81-173809.
29 Smith, TJ. (1985) Development and application of a model for estimating alveolar and interstitial dust levels. Ann
30 Occup Hyg 29:495-516.
31 Snipes, MB. (1979) Long-term retention of monodisperse and polydisperse particles inhaled by beagle dogs, rats
32 and mice. Albuquerque, NM. Lovelace Biomedical and Environmental Research Institute; Inhalation Toxicology
33 Research Institute annual report LF-69; pp. 420-423.
34 Snipes, MB; Clem, MF. (1981) Retention of microspheres in the rat lung after intratracheal instillation. Environ Res
35 24:33-41.
36 Snipes, MB; Olson, TR; Yeh, HC. (1988) Deposition and retention patterns for 3-, 9-, and 15-um latex
37 microspheres inhaled by tats and guinea pigs. Exp Lung Res 14:37-50.
38 Snipes, MB; McClellan, RO; Mauderly, JL; et al. (1989) Retention patterns for inhaled particles in the lung:
39 comparisons between laboratory animals and humans for chronic exposures. Health Phys 57(suppl. l):69-78.
40 Sorokin, SP; Brain, JD. (1975) Pathways of clearance in mouse lungs exposed to iron oxide aerosols. Anat Rec
41 181:581-625.
11/5/99 3-46 DRAFT—DO NOT CITE OR QUOTE
-------
1 Stahlhofen, W; Gebhart, J; Heyder, J. (1980) Experimental determination of the regional deposition of aerosol
2 particles in the human respiratory tract. Am Ind Hyg AssocJ 41:385-398a.
Stahlhofen, W; Koebrich, R; Rudolf, G; et al. (1990) Short-term and long-term clearance of particles from the upper
human respiratory tract as function of particle size. J Aerosol Sci 2I(suppl. 1):S407-S410.
5 Stober, W; Einbrodt, HJ; Klosterkotter, W. (1967) Quantitative studies of dust retention in animal and human lungs
6 after chronic inhalation. In: Davies, CN, ed. Inhaled particles and vapours II: proceedings of an international
7 symposium; September-October 1965; Cambridge, United Kingdom. Oxford, United Kingdom: Pergamon Press;
8 pp. 409-418.
9 StOber, W; Morrow, PE; Hoover, MD. (1989) Compartmental modeling of the long-term retention of insoluble
10 particles deposited in the alveolar region of the lung. Fundam Appl Toxicol 13:823-842.
11 StOber, W; McClellan, RO; Morrow, PE. (1993) Approaches to modeling disposition of inhaled particles and fibers
12 in the lung. In: Gardner, DE; Crapo, JD; McClellan, RO, eds. Toxicology of the lung. New York: Raven Press; pp.
13 527-601.
14 Stober, W; Morrow, PE; Koch, W; et al. (1994) Alveolar clearance and retention of inhaled insoluble particles in
15 rats simulated by a model inferring macrophage particle load distributions. J Aerosol Sci 25:975-1002.
16 Strom, KA. (1984) Response of pulmonary cellular defenses to the inhalation of high concentrations of diesel
17 exhaust. J Toxicol Environ Health 13:919-944.
18 Strom, KA; Chan, TL; Johnson, JT. (1988) Pulmonary retention of inhaled submicron particles in rats: diesel
19 exhaust exposures and lung retention model. In: Dodgson, J; McCallum, RI; Bailey, MR; et al., eds. Inhaled
20 particles VI: proceedings of an international symposium and workshop on lung dosimetry; September 1985;
21 Cambridge, United Kingdom. Ann Occup Hyg 32(suppl. l):645-657.
Subramaniam, RP; Miller, FJ; Asgarian, B. (1998) Numerical predictions of regional deposition of fine particles in
the human lung using a multiple-path model. Proceedings of the American Association of Aerosol Research, annual
24 meeting; Denver, CO.
25 Sun, JD; Wolff, RK; Kanapilly, GM. (1982) Deposition, retention, and biological fate of inhaled benzo(a)pyrene
26 adsorbed onto ultrafine particles and as a pure aerosol. Toxicol Appl Pharmacol 65:231-244.
27 Sun, JD; Wolff, RK; Kanapilly, GM; et al. (1984) Lung retention and metabolic fate of inhaled benzo(a)pyrene
28 associated with diesel exhaust particles. Toxicol Appl Pharmacol 73:48-59.
29 Tong, HY; Karasek, FW. (1984) Quantitation of polycyclic aromatic hydrocarbons in diesel exhaust paniculate
30 matter by high-performance liquid chromatography fractionation and high-resolution gas chromatography. Anal
31 Chem 56:2129-2134.
32 U.S. Environmental Protection Agency. (1982) Air quality criteria for paniculate matter and sulfur oxides. Research
33 Triangle Park, NC: Office of Health and Environmental Assessment, Environmental Criteria and Assessment
34 Office; EPA report no. EPA/600/8-82-029aF-cF. 3v. Available from: NTIS, Springfield, VA; PB84-156777.
35 U.S. Environmental Protection Agency. (1993) Provisional guidance for quantitative risk assessment of polycyclic
36 aromatic hydrocarbons. Office of Health and Environmental Assessment; EPA report no. EPA/600/R-93/089.
37 U.S. Environmental Protection Agency. (1996) Air quality criteria for paniculate matter. Research Triangle. Park,
38 NC: National Center for Environmental Assessment-RTF Office; report nos. EPA/600/P-95/001 aF-cF. 3 v.
39 Available from: NTIS, Springfield, VA; PB96-168224.
Vastag, E; Matthys, H; Zsamboki, G; et al. (1986) Mucociliary clearance in smokers. Eur J Respir Dis 68:107-113.
11/5/99 3-47 DRAFT—DO NOT CITE OR QUOTE
-------
1 Vostal, JJ; Schreck, RM; Lee, PS; et al. (1982) Deposition and clearance of diesel particles from the lung. In:
2 Lewtas, J, ed. Toxicological effects of emissions from diesel engines: proceedings of the 1981 EPA diesel
3 emissions symposium; October 1981; Raleigh, NC. New York: Elsevier Biomedical; pp. 143-159. (Developments in
4 toxicology and environmental science: v. 10).
5 Waite., DA; Ramsden, D. (1971) The inhalation of insoluble iron oxide particles in the sub-micron range. Part I.
6 Chromium-51 labelled aerosols. Winfrith, Dorchester, Dorset, United Kingdom: Atomic Energy Establishment;
7 report no. AEEW-R740.
8 Warheit, DB; Overby, LH; George, G; et al. (1988) Pulmonary macrophages are attracted to inhaled particles
9 through complement activation. Exp Lung Res 14:51-66.
10 Weibel, ER. (1963) Morphometry of the human lung. New York: Academic Press Inc.
11 White, HJ; Garg, BD. (1981) Early pulmonary response of the rat lung to inhalation of high concentration of diesel
12 particles. J Appl Toxicol 1:104-110.
13 Wolff, RK; Gray, RL. (1980) Tracheal clearance of particles. In: Diel, JH; Bice, DE; Martinez, BS, eds. Inhalation
14 Toxicology Research Institute annual report: 1979-1980. Albuquerque, NM: Lovelace Biomedical and
1 5 Environmental Research Institute; p. 252; report no. LMF-84.
16 Wolff, RK; Henderson, RF; Snipes, MB; et al. (1986) Lung retention of diesel soot and associated organic
17 compounds. In: Ishinishi, N; Koizumi, A; McClellan, R; et al. eds. Carcinogenic and mutagenic effects of diesel
18 engine exhaust: proceedings of the international satellite symposium on toxicological effects of emissions from
19 diesel engines; July; Tsukuba Science City, Japan. Amsterdam: Elsevier Science Publishers BV.; pp. 199-211.
20 (Developments in toxicology and environmental science: v. 13).
21 Wolff, RK; Henderson, RF; Snipes, MB; et al. (1987) Alterations in particle accumulation and clearance in lungs of
22 rats chronically exposed to diesel exhaust. Fundam Appl Toxicol 9:154-166.
23 World Health Organization. (1996) Diesel fuel and exhaust emissions. Geneva, Switzerland: World Health
24 Organization, International Programme on Chemical Safety. (Environmental health criteria 171).
25 Wright, JR; Clements, JA. (1987) Metabolism and turnover of lung surfactant. Am Rev Respir Dis 136:426-444.
26 Xu, GB; Yu, CP. (1987) Deposition of diesel exhaust particles in mammalian lungs: a comparison between rodents
27 and man. Aerosol Sci Technol 7:117-123.
28 Yeh, HC; Schum, GM. (1980) Models of human lung airways and their application to inhaled particle deposition.
29 Bull Math Biol 42:461-480.
30 Yu, CP. (19V8) Exact analysis of aerosol deposition during steady breathing. Pcv.-der Technc! 21:55-62.
31 Yu, CP; Diu, CK. (1983) Total and regional deposition ot inhaled aerosols in humans. J Aerosol Sci 5:593-609.
32 Yu, CP; Xu, GB. (1986) Predictive models for deposition of diesel exhaust particulates in human and rat lungs.
33 Aerosol Sci Technol 5:337-347.
34 Yu, CP; Yoon, KJ. (1990) Retention modeling of diesel exhaust particles in rats and humans. Amherst, NY: State
35 University of New York at Buffalo (Health Effects Institute research report no. 40).
36 Yu, CP; Chen, YK; Morrow, PE. (1989) An analysis of alveolar macrophage mobility kinetics at dust overloading
37 of the lungs. Fundam Appl Toxicol 13:452-45V.
38 Yu, CP; Ding, YJ; Zhang, L; et al. (1996) A clearance model of refractory ceramic fibers in the rat lung including
3 9 fiber dissolution and breakage. J Aerosol Sci 27:151 -160.
11/5/99 3-48 DRAFT—DO NOT CITE OR QUOTE
-------
1 Zielinska, B; Uberna, E; Fujita, EM; et al. (1998) Northern Front Range air quality study: the analysis of ambient
2 fine paniculate organic matter. Presented at: 91st annual meeting of the Air & Waste Management Association;
3 June; San Diego, CA. Pittsburgh, PA: Air & Waste Management Association; paper no. 98-RA89.03.
11 /5/99 3-49 DRAFT—DO NOT CITE OR QUOTE
-------
4. MUTAGENICITY OF DIESEL EXHAUST
f Since 1978, more than 100 publications have appeared in which genotoxicity assays were
used with diesel emissions, volatile and particulate fractions (including extracts), or individual
3 chemicals found in diesel emissions. Although most of the.studies deal with whether particulate
4 „ extracts from diesel emissions possess mutagenic activity in microbial and mammalian cell
5 assays, a number of studies in recent years have employed bioassays (most commonly
6 Salmonella TA98 without S9) to evaluate (1) extraction procedures, (2) fuel modifications, (3)
7 bioavailability of chemicals from diesel particulate matter (DPM), and (4) exhaust filters or other
8 modifications and other variables associated with diesel emissions. As indicated hi Chapter 2,
9 the number of chemicals in diesel emissions is very large. Many of these (e.g. PAHs and nitro-
10 PAHs) have been determined to exhibit mutagenic activity in a variety of assay systems.
11 Because of the limited and uncertain role of the individual chemicals in either the cancer or .
12 noncancer effects of diesel emissions, discussion of those data are not included. Also, several
13 review articles, some containing more detailed descriptions of the available studies, are available
14 (International Agency for Research on Cancer, 1989) (Claxton, 1983; Pepelko and Peirano,
15 1983; Shirname-More, 1995). The proceedings of several symposia on the health effects of
16 diesel emissions (U.S. EPA, 1980; Lewtas, 1982; Ishinishi et al., 1986) are also available. An
understanding of diesel exhaust mutagenicity is important to the cancer health effects and dose-
response chapters, Chapters 7 and 8, respectively.
19
20 4.1. GENE MUTATIONS
21 Huisingh et al. (1978) demonstrated that dichloromethane extracts from DPM were
22 mutagenic in strains TA1537, TA1538, TA98, and TA100 of S. typhimurium, both with and
23 without rat liver S9 activation. This report contained data from several fractions as well as DPM
24 from different vehicles and fuels. Similar results with diesel extracts from various engines and
25 fuels have been reported by a number of investigators using the Salmonella frameshift-sensitive
26 strains TA1537, TA1538, and TA98 (Siak et al., 1981; Claxton, 1981; Dukovich et al., 1981;
27 Brooks et al., 1984). Similarly, mutagenic activity was observed in Salmonella forward mutation
28 assays measuring 8-azaguanine resistance (Claxton and Kohan, 1981) and in E. coli mutation
29 assays (Lewtas, 1983).
30 One approach to identifying significant mutagens in chemically complex environmental
31 samples such as diesel exhaust or ambient particulate extracts is the combination of short-term
32 bioassays with chemical fractionation (Scheutzle and Lewtas, 1986). The analysis is most
frequently carried out by sequential extraction with increasingly polar or binary solvents.
11/5/99 4-1 DRAFT—DO NOT CITE OR QUOTE
-------
1 Prefractionation by silica-column chromatography separates compounds by polarity or into
2 acidic, basic, and neutral fractions. The resulting fractions are too complex to characterize by
3 chemical methods, but the bioassay analysis can be used to determine fractions for further
4 analysis. In most applications of this concept, Salmonella strain TA98 without the addition of S9
5 has been used as the indicator for mutagenic activity. Generally, a variety of nitrated polynuclear
6 aromatic compounds have been found that account for a substantial portion of the mutagenicity
7 (Liberti et al., 1984; Schuetzle and Frazer, 1986; Schuetzle and Perez, 1983). However, not all
8 bacterial mutagenicity has been identified in this way, and the identity of the remainder of the
9 mutagenic compounds remains unknown. The nitrated aromatics thus far identified in diesel
1 0 exhaust were the subject of review in the I ARC monograph on diesel exhaust (International
1 1 Agency for Research on Cancer, 1 989). In addition to the simple qualitative identification of
1 2 mutagenic chemicals, several investigators have used numerical data to express mutagenic
1 3 activity as activity per distance driven or mass of fuel consumed. These types of calculations
14 have been the basis for estimates that the nitroarenes (both mono- and dinitropyrenes) contribute
15 a significant amount of the total mutagenic activity of the whole extract (Nishioka et al., 1982;
16 Salmeen et al., 1982; Nakagawa et al., 1983). In a 1983 review, Claxton discussed a number of
1 7 factors that affected the mutagenic response in Salmonella assays. Citing the data from the
1 8 Huisingh et al. (1978) study, the author noted that the mutagenic response could vary by a factor
19 of 1 00 using different fuels in a single diesel engine. More recently, Crebelli et al. (1 995) used
20 Salmonella to examine the effects of different fuel components. They reported that while
21 mutagenicity was highly dependent on aromatic content, especially di- or triaromatics, there was
22 no clear effect of sulfur content. Later, Sjogren et al. (1996), using multivariate statistical
23 methods with 10 diesel fuels, concluded that the most influential chemical factors in Salmonella
24 mutagenicity were sulfur contents, certain polycyclic aromatic hydrocarbons (PAHs) (1 -
25 nitropyrene), and naphthenes.
26 Matsushita et al. (1 986) tested particle-free diesel exhaust gas and a number of benzene
27 nitro-derivatives and PAHs (many of which have been identified as components of diesel exhaust
- .•_ t.-^r. T A
uuui
29 activation. Of the 94 nitrobenzene derivatives tested, 61 were mutagenic, and the majority
30 showed greatest activity in TA100 without S9. Twenty-eight of 50 PAHs tested were mutagenic,
31 all required the addition of S9 for detection, and most appeared to show a stronger response in
32 TA100. When 1,6-dinitropyrene was mixed with various PAHs or an extract of heavy-duty (HD)
33 diesel exhaust, the mutagenic activity in TA98 was greatly reduced when S9 was absent but was
34 increased significantly when S9 was present. These latter results suggested that caution should
11/5/99 4-2 DRAFT—DO NOT CITE OR QUOTE
-------
1 be used in estimating mutagenicity (or other toxic effects) of complex mixtures from the specific
2 activity of individual components.
f Mitchell et al. (1981) reported mutagenic activity of DPM extracts of diesel emissions in
the mouse lymphoma L5178Y mutation assay. Positive results were seen both with and without
5 S9 activation in extracts from several different vehicles, with mutagenic activity only slightly
6 lower in the presence of S9. These findings have been confirmed in a number of other
7 mammalian cell systems using several different genetic markers. Casto et al. (1981), Chescheir
8 et al. (1981), Li and Royer (1982), and Brooks et al. (1984) all reported positive responses at the
9 HGPRT locus in Chinese hamster ovary (CHO) cells. Morimoto et al. (1986) used the APRT
10 and Ouar loci in CHO cells; Curren et al. (1981) used Ouar in BALB/c 3T3 cells. In all of these
11 studies, mutagenic activity was observed without S9 activation. Liber etal.(1981) used the
12 thymidine kinase (TK) locus in the TK6 human lymphoblast cell line and observed induced
13 mutagenesis only in the presence of rat liver S9 when testing a methylene chloride extract of
14 diesel exhaust. Barfknecht et al. (1982) also used the TK6 assay to identify some of the
15 chemicals responsible for this activation-dependent mutagenicity. They suggested that
16 fluoranthene, 1-methylphenanthrene, and 9-methylphenanthrene could account for more than
17 40% of the observed activity.
18 Morimoto et al. (1986) injected DPM extracts (250 to 4000 mg/kg) into pregnant Syrian
§ hamsters and measured mutations at the APRT locus in embryo cells cultivated 11 days after i.p.
injection. Neutral fractions from both light-duty (LD) and HD tar samples resulted hi increased
21 mutation frequency at 2000 and 4000 mg/kg. Belisario et al. (1984) applied the Ames test to
22 urine from Sprague-Dawley rats exposed to single applications of DPM administered by gastric
23 intubation, i.p. injection, or s.c. gelatin capsules. In all cases, dose-related increases were seen in
24 TA98 (without and with S9) from urine concentrates taken 24 h after particle administration.
25 Urine from Swiss mice exposed by inhalation to filtered exhaust (particle concentration 6 to 7
26 mg/m3) for 7 weeks (Pereira et al., 1981 a) or Fischer 344 rats exposed to DPM (2 mg/m3) for 3
27 months to 2 years was negative in Salmonella strains.
28 Schuler and Niemeier (1981) exposed Drosophila males in a stainless steel chamber
29 connected to the 3 m3 chamber used for the chronic animal studies at EPA (see Hinners et al.,
30 1980 for details). Flies were exposed for 8 h and mated to untreated females 2 days later.
31 Although the frequency of sex-linked recessive lethals from treated males was not different from
32 that of controls, the limited sample size precluded detecting less than a threefold increase over
33 controls. The authors noted that, because there were no signs of toxicity, the flies might tolerate
34 exposures to higher concentrations for longer time periods.
11/5/99 4-3 DRAFT—DO NOT CITE OR QUOTE
-------
1 Driscoll et al. (1996) exposed Fischer 344 male rats to aerosols of carbon black (1.1, 7.1,
2 and 52.8 mg/m3) or air for 13 weeks (6 h/day, 5 days/week) and measured hprt mutations in
3 alveolar type II cells in animals immediately after exposure and at 12 and 32 weeks after the end
4 of exposure. Both the two higher concentrations resulted in significant increases in mutant
5 frequency. While the mutant frequency from the 7.1 mg/m3 group returned to control levels by
6 12 weeks, the mutant frequency of the high exposure group was still higher than controls even
7 after 32 weeks. Carbon black particles have very little adsorbed PAHs, hence a direct
8 chemically-induced mechanism is highly unlikely. The authors suggested that the likely
9 explanation for the observed increases was persistent pulmonary inflamation and hyperplasia.
i
10 Specific-locus mutations were not induced in (C3H x 101 )F, male mice exposed to diesel
11 exhaust 8 h/day, 7 days/week for either 5 or 10 weeks (Russell et al., 1980). The exhaust was a
12 1:18 dilution and the average particle concentration was 6 mg/m3. After exposure, males were
13 mated to T-stock females and matings continued for the reproductive life of the males. The
14 results were unequivocally negative; no mutants were detected in 10,635 progeny derived from
15 postspermatogonial cells or in 27,917 progeny derived from spermatogonial cells.
16 Hou et al. (1995) measured DNA adducts and hprt mutations in 47 bus maintenance
17 workers and 22 control individuals. All were nonsmoking men from garages in the Stockholm
18 area and the exposed group consisted of 16 garage workers, 25 mechanics, and 6 other garage
19 workers. There were no exposure data, but the three groups were considered to be of higher to
20 lower exposure to diesel engine exhaust. Levels of DNA adducts determined by 32P-postlabeling
21 were significantly higher in workers than in controls (3.2 versus 2.3 * 10'8), but hprt mutant
22 frequencies were not different (8.6 versus 8.4 x 10"6). Both adduct level and mutagenicity were
23 highest among the 16 most exposed workers and mutant frequency was significantly correlated
24 with adduct level. All individuals were genotyped for glutathione transferase GSTM1 and
25 aromatic amino transferase NAT2 polymorphism. Neither GSTM1 nulls nor NAT2 slow
26 acetylators exhibited effects on either DNA adducts or hprt mutant frequencies.
27
28 4.2. CHROMOSOME EFFECTS
29 Mitchell et al. (1981) and Brooks et al. (1984) reported increases in sister chromatid
30 exchanges (SCE) in CHO cells exposed to DPM extracts of emissions from both LD and HD
31 diesel engines. Morimoto et al. (1986) observed increased SCE from both LD and HD DPM
32 extracts in PAH-stimulated human lymphocyte cultures. Tucker et al. (1986) exposed human
33 peripheral lymphocyte cultures from four donors to direct diesel exhaust for up to 3 b. Exhaust
^4- Wac ror»1<»H Kv rmmnino tVirrvnoVi si nlactic. tiihp afvvnt 90 fret Inner airflow wac 1ST /min
•w . •- —— J C f O ^ t ~~ ~ ~ ~ ~~ ~~ ~ "~" " ' t^? ~ ' " " — • — — • —————*
35 Samples were taken at 16, 48, and 160 min of exposure. Cell cycle delay was observed in all
11/5/99 4-4 DRAFT—DO NOT CITE OR QUOTE
-------
1 cultures; significantly increased SCE levels were reported for two of the four cultures. Structural
2 chromosome aberrations were induced in CHO cells by DPM extracts from a Nissan diesel
engine (Lewtas, 1983) but not by similar extracts from an Oldsmobile diesel engine (Brooks et
al., 1984).
5 Gu et al. (1992) reported that DEP dispersed in an aqueous mixture containing
6 dipalmitoyl lecithin (DPL), a component of pulmonary surfactant or extracted with
7 dichloromethane (DCM) induced similar responses in micronucleus tests in Chinese hamster
8 V79 and CHO cell cultures. After the samples were separated into supernatant and sediment
9 fractions, mutagenic activity was confined to the sediment fraction of the DPL sample and the
10 supernatant of the DCM sample. These findings suggest that the mutagenic activity of DEP
11 inhaled into the lungs could be made bioavailable through solubilization and dispersion nature of
12 . pulmonary surfactants, but the application of these in vitro findings to conditions in the human
13 lung remains to be studied.
14 Pereira et al. (1981a) exposed female Swiss mice to diesel exhaust 8 h/day, 5 days/week
15 for 1, 3, and 7 weeks. The incidence of micronuclei and structural aberrations was similar in
16 bone marrow cells of both control and exposed mice. Increased incidences of micronuclei, but
17 not SCE, were observed in bone marrow cells of male Chinese hamsters after 6 months of
18 exposure to diesel exhaust (Pereira et al., 1981 b).
§ Guerrero et al. (1981) observed a linear concentration-related increase in SCE in lung
cells cultured after intratracheal instillation of DPM at doses up to 20 mg/hamster. However,
21 they did not observe any increase in SCE after 3 months of inhalation exposure to diesel exhaust
22 particles (6 mg/m3).
23 Pereira et al. (1982) measured SCE in embryonic liver cells of Syrian hamsters. Pregnant
24 females were exposed to diesel exhaust (containing about 12 mg/m3 particles) from days 5 to 13
25 of gestation or injected intraperitoneally with diesel particles or particle extracts on gestational
26 day 13 (18 h before sacrifice). Neither the incidence of SCE nor mitotic index was affected by
27 exposure to diesel exhaust. The injection of DPM extracts, but not DPM, resulted in a dose-
28 related increase in SCE; however, the toxicity of the DPM was about twofold greater than the
29 DPM extract.
30 In the only studies with mammalian germ cells, Russell et al. (1980) reported no increase
31 in either dominant lethals or heritable translocations in males of T-stock mice exposed by
32 inhalation to diesel emissions. In the dominant lethal test, T-stock males were exposed for 7.5
33 weeks and immediately mated to females of different genetic backgrounds (T-stock; [C3H *
34 101]; [C3H * C57BL/6]; [SEC * C57BL/6]). There were no differences from controls in any of
35 the parameters measured in this assay. For heritable translocation analysis, T-stock males were
11/5/99 4-5 DRAFT—DO NOT CITE OR QUOTE
-------
1 exposed for 4.5 weeks and mated to (SEC * C57BL/6) females, and the F, males were tested for
2 the presence of heritable translocations. Although no translocations were detected among 358
3 progeny tested, the historical control incidence is less than 1/1,000.
4
5 4.3. OTHER GENOTOXIC EFFECTS
Q Pereira et al. (1981b) exposed male strain A mice to diesel exhaust emissions for 31 or 39
7 weeks using the same exposure regimen noted in the previous section. Analyses of caudal sperm
8 for sperm-head abnormalities were conducted independently in three separate laboratories.
9 Although the incidence of sperm abnormalities was not significantly above controls in any of the
10 three laboratories, there were extremely large differences in scoring (control values were 9.2%,
11 14.9%, and 27.8% in the three laboratories). Conversely, male Chinese hamsters exposed for 6
12 months (Pereira et al., 1981 c) exhibited almost a threefold increase in sperm-head abnormalities.
13 It is noted that the control incidence in the Chinese hamsters was less than 0.5%. Hence, it is not
14 clear whether the differing responses reflect true species differences or experimental artifacts.
15
16 4.4. SUMMARY
17 Extensive studies with Salmonella have unequivocally demonstrated mutagenic activity
18 in both particulate and gaseous fractions of diesel exhaust. In most of the studies using
19 Salmonella, DPM extracts and individual nitropyrenes exhibited the strongest responses in strain
20 TA98 when no exogenous activation was provided. Gaseous fractions reportedly showed greater
21 response in TA100, whereas benzo[a]pyrene and other unsubstituted PAHs are mutagenic only in
22 the presence of S9 fractions. The induction of gene mutations has been reported in several in
23 vitro mammalian cell lines after exposure to extracts of DPM. Note that only the TK6 human
24 cell line did not give a positive response to DPM extracts in the absence of S9 activation.
25 Mutagenic activity was recovered in urine from animals treated with DPM by gastric intubation
26 and i.p. and s.c. implants, but not by inhalation of DPM or diluted diesel exhaust. Dilutions of
27 whole diesel exhaust did not induce sex-linked recessive lethals in Drosophila or specific-locus
28 mutations in male mouse germ cells.
29 Structural chromosome aberrations and SCE in mammalian cells have been induced by
30 particles and extracts. Whole exhaust induced micronuclei but not SCE or structural aberrations
31 in bone marrow of male Chinese hamsters exposed to whole diesel emissions for 6 months. In a
32 shorter exposure (7 weeks), neither micronuclei nor structural aberrations were increased in bone
33 marrow of female Swiss mice. Likewise, whole diesel exhaust did not induce dominant lethais
34 or heritable translocations in male mice exposed for 7.5 and 4.5 weeks, respectively.
11 /5/99 4-6 DRAFT—DO NOT CITE OR QUOTE
-------
1 Application of mutagenicity data to the question of the potential carcinogenicity of diesel
2 emissions is based on the premise that genetic alterations are found in all cancers and that several
of the chemicals found in diesel emissions possess mutagenic activity in a variety of genetic
assays. These genetic alterations can be produce by gene mutations, deletions, translocations,
5 aneuploidy, or amplification of genes, hence no single genotoxicity assay should be expected to
6 either qualitatively or quantitatively predict rodent carcinogenicity. With diesel emissions or
7 other mixtures, additional complications arise because of the complexity of the material being
8 tested. Exercises that combined the Salmonella mutagenic potency with the total concentration
9 of mutagenic chemicals deposited in the lungs could not account for the observed tumor
10 incidence in exposed rats (Rosenkranz, 1993,Goldstein, et al. 1998). Additionally, it appears that
11 the some of constituents responsible for the mutational increases observed in bacteria are
12 different from those responsible for the observed increases in CHO cells (Li and Dutcher, 1983)
13 or in human hepatoma-derived cells (Eddy et al., 1986).
14
15 4.5. REFERENCES
16
17 Barfknecht, TR; Kites, RA; Cavaliers, EL; et al. (1982) Human cell mutagenicity of polycyclic aromatic
18 hydrocarbon components of diesel emissions. In: Lewtas, J, ed. Toxicological effects of emissions from diesel
19 engines: proceedings of the Environmental Protection Agency 1981 diesel emissions symposium; October 1981;
20 Raleigh, NC. (Developments on toxicology and environmental science: v. 10.) New York: Elsevier Biomedical; pp.
^ 277-294.
23 Belisario, MA; Buonocore, V; De Marinis, E; et al. (1984) Biological availability of mutagenic compounds
24 adsorbed onto diesel exhaust paniculate. Mutat Res 135:1-9.
25
26 Brooks, A; Li, AP; Dutcher, JS; et al. (1984) A comparison of genotoxicity of automobile exhaust particles from
27 laboratory and environmental sources. Environ Mutagen 6:651 -668.
28
29 Casto, BC; Hatch, GG; Huang, SL; et al. (1981) Mutagenic and carcinogenic potency of extracts of diesel and
30 related environmental emissions: in vitro mutagenesis and oncogenic transformation. Environ Int 5:403-409.
31
32 Chescheir, GM, III; Garrett, NE; Shelburne, JD; et al. (1981) Mutagenic effects of environmental particulates in the
33 CHO/HGPRT system. In: Waters, MD; Sandhu, SS; Huisingh, JL; et al., eds. Application of short-term bioassays in
34 the fractionation and analysis of complex environmental mixtures. New York: Plenum Press; pp. 337-350.
35
36 Claxton, LD. (1981) Mutagenic and carcinogenic potency of diesel and related environmental emissions:
37 Salmonella bioassay. Environ Int 5:389-391.
38
39 Claxton, LD. (1983) Characterization of automotive emissions by bacterial mutagenesis bioassay: a review. Environ
40 Mutagen 5:609-631.
41
42 Claxton, L; Kohan, M. (1981) Bacterial mutagenesis and the evaluation of mobile-source emissions. In: Waters,
43 MD; Sandhu, SS; Huisingh, JL; et al., eds. Short-term bioassays in the analysis of complex environmental mixtures
44 II: proceedings of the second symposium on the application of short-term bioassays in the fractionation and analysis
45 of complex environmental mixtures; March 1980; Williamsburg, VA. (Hollaender, A; Welch, BL; Probstein, RF,
eds. Environmental science research series: v. 22.) New York: Plenum Press; pp. 299-317.
11/5/99 4-7 DRAFT—DO NOT CITE OR QUOTE
-------
1 Crebelli, R; Conti, L; Crochi, B; et al. (1995) Carere A, Bertoli C, Delgiacomo N. The effect of fuel composition
2 on the mutagenicity of diesel engine exhaust. Mutat Res 346:167-172.
3
4 Curren, RD; Kouri, RE; Kim, DM; et al. (1981) Mutagenic and carcinogenic potency of extracts from diesel related
5 environmental emissions: simultaneous morphological transformation and mutagenesis in BALB/c 3T3 cells.
6 Environ Internal 5:411-415.
7
8 Driscoll, KE; Carter, JM; Howard, BW; et al. (1996) Pulmonary inflammatory, chemokine, and mutagenic
9 responses in rats after subchronic inhalation of carbon black. Toxicol Appl Pharmacol 136:372-380.
10
11 Dukovich, M; Yasbin, RE; Lestz, SS; Risby, TH; et al. (1981) The mutagenic and SOS-inducing potential of the
12 soluble organic fraction collected from diesel particulate emissions. Environ Mutagen 3: 253-264.
13
14 Eddy, EP; McCoy, EC; Rosenkranz, HS; et al. (1986) Dichotomy in the mutagenicity and genotoxicity of
15 nitropyrenes: apparent effect of the number of electrons involved in nitroreduction. Mutat Res 161:109- 111.
16
17 Goldstein, LS; Weyand, EH; Safe, S; et al. (1998) Tumors and DNA adducts in mice exposed to benzo[a]pyrene
18 and coal tars: implications for risk assessment. Environ Health Perspect 106(6): 1325-1330.
19
20 Gu, ZW; Zhong, BZ; Nath, B; et al. (1992) Micronucleus induction and phagocytosis in mammalian cells treated
2'1 with diesel emission particles. Mutat Res 279:55-60.
22
23 Guerrero, RR; Rounds, DE; Orthoefer, J. (1981) Sister chromatid exchange analysis of Syrian hamster lung cells
24 treated in vivo with diesel exhaust particulates. Environ Int 5:445-454.
25
26 Hinners, RG; Burkart, JK; Malanchuk, M. (1980) Facilities for diesel exhaust studies. In: Pepelko, WE; Danner,
27 RM; Clarke, NA, eds. Health effects of diesel engine emissions: proceedings of an international symposium;
28 December 1979. Cincinnati, OH: U.S. Environmental Protection Agency, Health Effects Research Laboratory; pp.
29 681-697; EPA report no. EPA-600/9-80-057b. Available from: NTIS, Springfield, VA; PB81-173817.
30
31 Hou, S; Lambert, B; Hemminki, K. (1995) Relationship between hprt mutant frequency, aromatic DNA adducts and
32 genotypes for GSTM1 and NAT2 in bus maintenance workers. Carcinogenesis 16:1913-1917.
33
34 Huisingh, J; Bradow, R; Jungers, R; et al. (1978) Application of bioassay to the characterization of diesel particle
35 emissions. In: Waters, MD; Nesnow, S; Huisingh, JL; et al., eds. Application of short-term bioassays in the
36 fractionation and analysis of complex environmental mixtures: [proceedings of a symposium; February;
37 Williamsburg, VA]. (Hollaender, A; Probstein, F; Welch, BL, eds. Environmental science research: v. 15.) New
38 York: Plenum Press; pp. 383-418.
39
40 International Agency for Research on Cancer. (1989) Diesel and gasoline engine exhausts and some nitroarenes.
41 (IARC monographs on the evaluation of carcinogenic risks to humans: v. 46). Lyon. France: World Health
42 Organization; pp. 41-185.
43
44 Ishinishi, N; Koizumi, A; McClellan, RO; et al., eds. (1986) Carcinogenic and mutagenic effects of diesel engine
45 exhaust: proceedings of the international satellite symposium on toxicological effects of emissions from diesel
46 engines; July; Tsukuba Science City, Japan. (Developments in toxicology and environmental science: v. 13.)
47 Amsterdam: Elsevier Science Publishers BV.
48
49 Lewtas, J. (1982) Mutagenic activity of diesel emissions. In: Lewtas, J, ed. Toxicological effects of emissions from
50 diesel engines: proceedings of the Environmental Protection Agency 1981 diesel emissions symposium: October
51 1981; Raleigh, NC. (Developments in toxicology and environmental science: v. 10.) New York: Elsevier
52 Biomedicai; pp. 243-264.
53
11/5/99 4-8 DRAFT—DO NOT CITE OR QUOTE
-------
1 Lewtas, J. (1983) Evaluation of the mutagenicity and carcinogenicity of motor vehicle emissions in short-term
2 bioassays. Environ Health Perspect 47:141-152.
3
4 Li, AP; Dutcher, JS. (1983) Mutagenicity of mono-, di-, and tri-nitropyrenes in Chinese hamster ovary cells. Mutat
5 Res 119:387-392.
6
7 Li, AP; Royer, RE. (1982) Diesel-exhaust-particle extract enhancement of chemical-induced mutagenesis in
8 cultured Chinese hamster ovary cells: possible interaction of diesel exhaust with environmental chemicals. Mutat
9 Res 103:349-355.
10
11 Liber, HL; Andon, BM; Hites, RA; et al. (1981) Diesel soot: mutation measurements in bacterial and human cells.
12 Environ Int 5:281-284.
13
14 Liberti, A; Ciccioli, P; Cecinato, A; et al. (1984) Determination of nitrated-polyaromatic hydrocarbons (nitro-PAHs)
15 in environmental samples by high resolution chromatographic techniques. J High Resolut Chromatogr Commun
16 7:389-397.
17
18 Matsushita, H; Goto, S; Endo, O; et al. (1986) Mutagenicity of diesel exhaust and related chemicals. In: Ishinishi, N;
19 Koizumi, A; McClellan, RO; Steber, W, eds. Carcinogenic and mutagenic effects of diesel engine exhaust:
20 proceedings of the international satellite symposium on toxicological effects of emissions from diesel engines; July;
21 Tsukuba Science City, Japan. (Developments on toxicology and environmental science: v. 13.) Amsterdam: Elsevier
22 Science Publishers BV; pp. 103-118.
23
24 Mitchell, AD; Evans, EL; Jotz, MM; et al. (1981) Mutagenic and carcinogenic potency of extracts of diesel and
25 related environmental emissions: in vitro mutagenesis and DNA damage. Environ Int 5:393-401.
26
27 Morimoto, K; Kitamura, M; Kondo, H; et al. (1986) Genotoxicity of diesel exhaust emissions in a battery of in-vitro
28 short-term bioassays. In: Ishinishi, N; Koizumi, A; McClellan, RO; Stober, W, eds. Carcinogenic and mutagenic
29 effects of diesel engine exhaust: proceedings of the international satellite symposium on toxicological effects of
}0 emissions from diesel engines; July; Tsukuba Science City, Japan. (Developments in toxicology and environmental
31 science: v. 13.) Amsterdam: Elsevier Science Publishers BV; pp. 85-102.
32
33 Nakagawa, R; Kitamori, S; Horikawa, K; et al. (1983) Identification of dinitropyrenes in diesel-exhaust particles:
34 their probable presence as the major mutagens. Mutat Res 124:201-211.
35
36 Nishioka, MG; Petersen, BA; Lewtas, J. (1982) Comparison of nitro-aromatic content and direct-acting
37 mutagenicity of diesel emissions. In: Cooke, M; Dennis, AJ; Fisher, GL, eds. Polynuclear aromatic hydrocarbons:
38 physical and biological chemistry. Columbus, OH: Battelle Press; pp. 603-613.
39
40 Pepelko, WE; Peirano, WB. (1983) Health effects of exposure to diesel engine emissions: a summary of animal
41 studies conducted by the U.S. Environmental Protection Agency's Health Effects Research Laboratories at
42 Cincinnati, Ohio. J Am Coll Toxicol 2:253-306.
43
44 Pereira, MA; Connor, TH; Meyne, J; et al. (198la) Metaphase analysis, micronucleus assay and urinary
45 mutagenicity assay of mice exposed to diesel emissions. Environ Int 5:435-438.
46
47 Pereira, MA; Sabharwal, PS; Gordon, L; et al. (1981b) The effect of diesel exhaust on sperm-shape abnormalities in
48 mice. Environ Int 5:459-460.
49
50 Pereira, MA; Sabharwal, PS; Kaur, P; et al. (1981c) In vivo detection of mutagenic effects of diesel exhaust by
51 short-term mammalian bioassays. Environ Int 5:439-443.
52
53 Pereira, MA; McMillan, L; Kaur, P; et al. (1982) Effect of diesel exhaust emissions, particulates, and extract on
sister chromatid exchange in transplacentally exposed fetal hamster liver. Environ Mutagen 4:215-220.
11/5/99 4-9 DRAFT—DO NOT CITE OR QUOTE
-------
1 Rosenkranz, HS. (1993) Revisiting the role of mutagenesis in the induction of lung tumors in rats by diesel
2 emissions. Mutat Res 303:91-95.
3
4 Russell, LB; Generoso, WM; Oakberg, EF; et al. (1980) Tests for heritable effects induced by diesel exhaust in the
5 mouse. Martin Marietta Energy Systems, Inc., Oak Ridge National Laboratory; report no. ORNL-5685.
6
7 Salmeen, I; Durisin, AM; Prater, TJ; et al. (1982) Contribution of 1-nitropyrene to direct-acting Ames assay
8 mutagenicities of diesel particulate extracts. Mutat Res 104:17-23.
9
10 Schuetzle, D; Perez, JM. (1983) Factors influencing the emissions of nitrated-polynuclear aromatic hydrocarbons
11 (nitro-PAH) from diesel engines. J Air Pollut Control Assoc 33:751 -755.
12
13 Schuetzle, D; Frazier, JA. (1986) Factors influencing the emission of vapor and particulate phase components from
14 diesel engines. In: Ishinishi, N; Koizumi, A; McClellan, RO; et al., eds. Carcinogenic and mutagenic effects of
15 diesel engine exhaust: proceedings of the international satellite symposium on toxicological effects of emissions
16 from diesel engines; July; Tsukuba Science City, Japan. (Developments in toxicology and environmental science: v.
17 13.) Amsterdam: Elsevier Science Publishers BV; pp. 41-63.
18
19 Schuetzle, D; Lewtas, J. (1986) Bioassay-directed chemical analysis in environmental research. Anal Chem
20 58:1060A-1076A.
21
22 Schuler, RL; Niemeier, RW. (1981) A study of diesel emissions on Drosophila. Environ Int 5:431-434.
23
24 Shirname-More, L. (1995) Genotoxicity of diesel emissions. Part I: Mutagenicity and other genetic effects. Diesel
25 exhaust: A critical analysis of emissions, exposure, and health effects. A special report of the Institute's Diesel
26 Working Group. Cambridge, MA: Health Effects Institute, pp. 222-242.
27
28 Siak, JS; Chan, TL; Lees, PS. (1981) Diesel particulate extracts in bacterial test systems. Environ Int 5:243-248.
29
30 Sjogren, M; Li, H; Banner, C; et al. (1996) Influence of physical and chemical characteristics of diesel fuels and
31 exhaust emissions on biological effects of particle extracts: a multivariate statistical analysis often diesel fuels.
32 Chem Res Toxicol 9:197-207.
33
34 Tucker, JD; Xu, J; Stewart, J; et al. (1986) Detection of sister chromatid exchanges induced by volatile
35 genotoxicants. Teratog Carcinog Mutagen 6:15-21.
36
37 U.S. Environmental Protection Agency. (1980) Health effects of diesel engine emissions: proceedings of an
38 international symposium. Cincinnati, OH: Office of Research and Development; EPA 600/9-80-057b.
11/5/99 4-10 DRAFT—DO NOT CITE OR QUOTE
-------
5. NONCANCER HEALTH EFFECTS OF DIESEL EXHAUST
The objective of this chapter is to report, evaluate, and interpret health effects other than
cancer that have been associated with inhalation exposure to diesel exhaust. Data on this class of
3 health effects of diesel exhaust have been obtained from diverse human, laboratory animal, and
4 in vitro test systems. The human studies comprise both occupational and human experimental
5 exposures, the former consisting of exposure to diesel exhaust in the occupational environment
6 and the latter consisting of exposure to diluted diesel exhaust or diesel paniculate matter (DPM)
7 under controlled conditions. The laboratory animal studies consist of both acute and chronic
8 exposures of laboratory animals to diesel exhaust or DPM. Diverse in vitro test systems
9 composed of human and laboratory animal cells treated with DPM or components of DPM have
10 also been used to investigate the effects of DPM at the cellular and molecular levels.
11 The noncancer health effects of ambient particulate matter, which is composed in part of DPM,
12 as well as the potential mechanisms underlying these effects, have been reviewed previously
13 (Health Effects Institute, 1995; U.S. EPA, 1996).
14
15 5.1. HEALTH EFFECTS OF WHOLE DIESEL EXHAUST
16 5.1.1. Human Studies
5.1.1.1. Short-Term Exposures
In a controlled human study, Rudell et al. (1990,1994) exposed eight healthy subjects in
19 an exposure chamber to diluted exhaust from a diesel engine for one hour, with intermittent
20 exercise. Dilution of the diesel exhaust was controlled to provide a median NO2 level of
21 approximately 1.6 ppm. Median particle number was 4.3 * 106/cm3, and median levels of NO
22 and CO were 3.7 and 27 ppm, respectively (particle size and mass concentration were not
23 provided). There were no effects on spirometry or on closing volume using nitrogen washout.
24 Five of eight subjects experienced unpleasant smell, eye irritation, and nasal irritation during
25 exposure. Brochoalveolar lavage was preformed 18 hours after exposure and was compared with
26 a control BAL performed 3 weeks prior to exposure. There was no control air exposure. Small
27 but statistically significant reductions were seen in BAL mast cells, AM phagocytosis of
28 opsonized yeast particles, and lymphocyte CD4/CD8 ratios. A small increase in recovery of
29 PMNs was also observed. These findings suggest that diesel exhaust may induce mild airway
30 inflammation in the absence of spirometric changes. This study provides an intriguing glimpse
31 of the effect of diesel exhaust exposure in humans, but only one exposure level was used, the
32 number of subjects was low, and a limited range of endpoints was reported, so the data are
11/5/99 5-1 DRAFT—DO NOT CITE OR QUOTE
-------
1 inadequate to generalize about the human response. To date, no well-controlled chamber study
2 has been conducted using methodologies for assessing subtle lung inflammatory reactions.
3 Rudell et al. (1996) exposed volunteers to diesel exhaust for 1 h in an exposure chamber.
4 Light work on a bicycle ergometer was performed during exposure. Exposures included either
5 diesel exhaust or exhaust with particle numbers reduced 46% by a particle trap. The engine used
6 was a new Volvo model 1990, a six-cylinder direct-injection turbo charged diesel with an
7 intercooler, which was run at a steady speed of 900 rpm during the exposures. Comparison of
8 this study with others was difficult because neither exhaust dilution ratios nor particle
9 concentrations were reported. Carbon monoxide concentrations of 27-30 ppm and NO of
10 2.6-2.7 ppm, however, suggested DPM concentrations may have equaled several mg/m3. The
11 most prominent symptoms during exposure were irritation of the eyes and nose and .an
12 unpleasant smell. Both airway resistance and specific airway resistance increased significantly
13 during the exposures. Despite the 46% reduction in particle numbers by the trap, effects on
14 symptoms and lung function were not significantly attenuated.
15 Kahn et al. (1988) reported the occurrence of 13 cases of acute overexposure to diesel
16 exhaust among Utah and Colorado coal miners. Twelve miners had symptoms of mucous
17 membrane irritation, headache, and lightheadedness. Eight individuals reported nausea; four
18 reported a sensation of unreality; four reported heartburn; three reported weakness, numbness,
19 and tingling in their extremities; three reported vomiting; two reported chest tightness; and two
20 others reported wheezing. Each miner lost time from work because of these symptoms, which
21 resolved within 24 to 48 h. No air monitoring data were presented; poor work practices were
22 described as the predisposing conditions for overexposure.
23 El Batawi and Noweir (1966) reported that among 161 workers from two garages where
24 diesel-powered buses were serviced and repaired, 42% complained of eye irritation, 37% of
25 headaches, 30% of dizziness, 19% of throat irritation, and 11% of cough and phlegm. Ranges of
26 mean concentrations of diesel exhaust components in the two diesel bus garages were as follows:
27 0.4 to 1.4 ppm NO2, 0.13 to 0.81 ppin SO2,0.6 to 44.1 pprn aldehydes, and 1.34 to 4.51 mg/m3 of
28 Drivl; the highest concentrations wer; obtained close to tb.e evb.avst systems o^tbe buses.
29 Eye irritation was reported by Battigelli (1965) in six subjects after 40 s of chamber
30 exposure to diluted diesel exhaust containing 4.2 ppm NO2,1 ppm SO2, 55 ppm CO, 3.2 ppm
31 total hydrocarbons, and 1 to 2 ppm total aldehydes; after 3 min and 20 s of exposure to diluted
32 diesel exhaust containing 2.8 ppm NO2, 0.5 ppm SO2, 30 ppm CO, 2.5 ppm total hydrocarbons,
33 and <1 to 2 ppm total aldehydes; and after 6 min of exposure to diluted diesel exhaust containing
34 1.3 ppm NO2, U.2 ppm SO2, <20 ppm CO, <2.0 ppm ioutl hydrocarbons, turn <1.0 ppm total
35 aldehydes. The concentration of DPM was not reported.
1175/99 5-2 DRAFT—DO NOT CITE OR QUOTE
-------
1 Katz et al. (1960) described the experience of 14 chemists and their assistants monitoring
2 the environment of a train tunnel used by diesel-powered locomotives. Although workers
complained on three occasions of minor eye and throat irritation, no correlation was established
with concentrations of any particular component of diesel exhaust.
5 The role of antioxidant defenses in protecting against acute diesel exhaust exposure has
6 been studied. Blomberg et al. (1998) investigated changes in the antioxidant defense network
7 within the respiratory tract lining fluids of human subjects following diesel exhaust exposure.
8 Fifteen healthy, nonsmoking, asymptomatic subjects were exposed to filtered air or diesel
9 exhaust (DPM 300 mg/m3) for 1 hr on two separate occasions at least three weeks apart. Nasal
10 lavage fluid and blood samples were collected prior to, immediately after, and 5 !/2 hr post
11 exposure. Bronchoscopy was performed 6 hr after the end of diesel exhaust exposure. Nasal
12 lavage ascorbic acid concentration increased 10-fold during diesel exhaust exposure, but returned
13 to basal levels 5.5 hr post-exposure. Diesel exhaust had no significant effects on nasal lavage
14 uric acid or GSH concentrations, and did not affect plasma, bronchial wash, or bronchoalveolar
15 lavage antioxidant concentrations, nor malondialdehyde or protein carbonyl concentrations. The
16 authors concluded that the physiological response to acute diesel exhaust exposure is an acute
17 increase in the level of ascorbic acid in the nasal cavity, which appears to be sufficient to prevent
18 further oxidant stress in the respiratory tract of healthy individuals.
19
^P 5.1.1.1.1. Diesel exhaust odor. The odor of diesel exhaust is considered by most people to be
21 objectionable; at high intensities, it may produce sufficient physiological and psychological
22 effects to warrant concern for public health. The intensity of the odor of diesel exhaust is an
23 exponential function of its concentration such that a tenfold change in the concentration will alter
24 the intensity of the odor by one unit. Two human panel rating scales have been used to measure
25 diesel exhaust odor intensity. In the first (Turk, 1967), combinations of odorous materials were
26 selected to simulate diesel exhaust odor; a set of 12 mixtures, each having twice the
27 concentration of that of the previous mixture, is the basis of the diesel odor intensity scale (D-
28 scale). The second method is the TIA (total intensity of aroma) scale based on seven steps,
29 ranging from 0 to 3, with 0 being undetectable, l/2 very slight, and 1 slight and increasing in one-
30 half units up to 3, strong (Odor Panel of the CRC-APRAC Program Group on Composition of
31 Diesel Exhaust, 1979; Levins, 1981).
32 Surveys, utilizing volunteer panelists, have been taken to evaluate the general public's
33 response to the odor of diesel exhaust. Hare and Springer (1971) and Hare et al. (1974) found
34 that at a D rating of about 2 (TIA = 0.9, slight odor intensity), about 90% of the participants
35 perceived the odor, and almost 60%.found it objectionable. At a D rating of 3.2 (TIA = 1.2,
1175/99 5-3 DRAFT—DO NOT CITE OR QUOTE
-------
1 slight to moderate odor intensity), about 95% perceived the odor, and 75% objected to it, and, at
2 a D rating of 5 (TIA = 1.8, almost moderate), about 95% objected to it.
3 Linnell and Scott (1962) reported odor threshold measurement in six subjects and found
4 that the dilution factor needed to reach the threshold ranged from 140 to 475 for this small
5 sample of people. At these dilutions, the concentrations of formaldehyde ranged from 0.012 to
6 0.088 ppm.
7
8 5.1.1.1.2. Pulmonary and respiratory effects. Battigelli (1955) exposed 13 volunteers to three
9 dilutions of diesel exhaust obtained from a one-cylinder, four-cycle, 7-hp diesel engine (fuel type
10 unspecified) and found that 15-min to 1-h exposures had no significant effects on pulmonary
11 resistance. Pulmonary resistance was measured by plethysmography utilizing the simultaneous
12 recording of esophageal pressure and airflow determined by electrical differentiation of the
13 volume signal from a spirometer. The concentration of the constituents in the three diluted
14 exhausts were 1.3, 2.8, and 6.2 ppm NO2; 0.2, 0.5, and 1 ppm SO2; <20, 30, and 55 ppm CO; and
15 <1.0, <1 to 2, and 1 to 2 ppm total aldehydes, respectively. DPM concentrations were not
16 reported.
17 A number of studies have evaluated changes in pulmonary function occurring over a
18 workshift in workers occupationally exposed to diesel exhaust (specific time period not always
19 reported but assumed to be 8 h). In a study of coal miners, Reger (1979) found that both forced
20 expiratory volume in 1 s (FEV,) and forced vital capacity (FVC) decreased by 0.05 L in
21 60 diesel-exposed miners, an amount not substantially different from reductions seen in non-
22 diesel-exposed miners (0.02 and 0.04 L, respectively). Decrements in peak expiratory flow rates
23 were similar between diesel and non-diesel exhaust-exposed miners. Miners with a history of
24 smoking had an increased number of decrements over the shift than nonsmokers did. Although
25 the monitoring data were not reported, the authors stated that there was no relationship between
26 the low concentrations of measured respirable dust or NO2 (personal samplers) when compared
27 with shift changes for any lung function parameter measured for the diesel-exposed miners. This
28 study is limited because results were preliminary (abstract) and there was incomplete information
29 on the control subjects.
30 Ames et al. (1982) compared the pulmonary function of 60 coal miners exposed to diesel
31 exhaust with that of a control group of 90. coal miners not exposed to diesel exhaust for evidence
32 of acute respiratory effects associated with exposure to diesel exhaust. Changes over the
33 workshift in FVC, FEV,, and forced expiratory flow rate at 50% FVC (FEF50) were the indices
34 for acute respiratory effects. The environmental concentrations of the primary pollutants were
35 2.0 mg/m3 respirable dust (<10 /um MMAD), 0.2 ppm NO2,12 ppm CO, and 0.3 ppm
1175/99 5-4 DRAFT—DO NOT CITE OR QUOTE
-------
1 formaldehyde. The investigators reported a statistically significant decline in FVC and FEV,
2 over the workshift in both the diesel-exposed and comparison groups. Current smokers had
,3 greater decrements in FVC, FEV,, and FEF50 than ex-smokers and nonsmokers. There was a
marked disparity between the ages and the time spent underground for the two study groups.
5 Diesel-exposed miners were about 15 years younger and had worked underground for 15 fewer
6 years (4.8 versus 20.7 years) than miners not exposed to diesel exhaust. The significance of
7 these differences between the populations studied on the results is difficult to ascertain.
8 Except for the expected differences related to age, 120 underground iron ore miners
9 exposed to diesel exhaust had no workshift changes in FVC and FEV, when compared with
10 120 matched surface miners (Jorgensen and Svensson, 1970). Both groups had equal numbers
11 (30) of smokers and nonsmokers. The frequency of bronchitis was higher among underground
12 workers, much higher among smokers than nonsmokers, and also higher among older than
13 younger workers. The authors reported that the underground miners had exposures of 0.5 to
14 1.5 ppm NO2 and between 3 and 9 mg/m3 particulate matter with 20 to 30% of the particles
15 <5 /am MMAD. The majority of the particles were iron ore; quartz was 6 to 7% of the fraction
16 <5 ^m MMAD.
17 Gamble et al. (1979) measured preshift FEV, and FVC in 187 salt miners and obtained
18 peak flow forced expiratory flow rates at 25,50, and 75% of FVC (FEF25, FEF50, or FEF75).
19 Postshift pulmonary function values were determined from total lung capacity and flows at
^P preshift percentages of FVC. The miners were exposed to mean NO2 levels of 1.5 ppm and mean
21 respirable particulate levels of 0.7 mg/m3. No statistically significant changes were found
22 between changes in pulmonary function and in NO2 and respirable particles combined. Slopes of
23 the regression of NO2 and changes in FEV,, FEF25, FEF50, and FEF7S were significantly different
24 from zero. The authors concluded that these small reductions in pulmonary function were
25 attributable to variations in NO2 within each of the five salt mines that contributed to the cohort.
26 Gamble et al. (1987a) investigated the acute effects of diesel exhaust in 232 workers in
27 four diesel bus garages using an acute respiratory questionnaire and before and after workshift
28 spirometry. The prevalence of burning eyes, headaches, difficult or labored breathing, nausea,
29 and wheeze experienced at work was higher in the diesel bus garage workers than in a
30 comparison population of lead/acid battery workers who had not previously shown a statistically
31 significant association of acute symptoms with acid exposure. Comparisons between the two
32 groups were made without adjustment for age and smoking. There was no detectable association
33 of exposure to NO2 (0.23 ppm ± 0.24 S.D.) or inhalable (less than 10 yum MMAD) particles
34 (0.24 mg/m3 ± 0.26 S.D.) and acute reductions in FVC, FEV,, peak flows, FEF50, and FEF75.
11/5/99 5-5 DRAFT—DO NOT CITE OR QUOTE
-------
1 Workers who had respiratory symptoms had slightly greater but statistically insignificant
2 reductions in FEV, and FEF50.
3 Ulfvarson et al. (1987) evaluated workshift changes in the pulmonary function of 17 bus
4 garage workers, 25 crew members of two types of car ferries, and 37 workers on roll-on/roll-off
5 ships. The latter group was exposed primarily to diesel exhaust; the first two groups were
6 exposed to both gasoline and diesel exhaust. The diesel-only exposures that averaged 8 h
7 consisted of 0.13 to 1.0 mg/m3 paniculate matter, 0.02 to 0.8 mg/m3 (0.016 to 0.65 ppm) NO,
8 0.06 to 2.3 mg/m3 (0.03 to 1.2 ppm) N02, 1.1 to 5.1 mg/m3 (0.96 to 4.45 ppm) CO, and up to
9 0.5 mg/m3 (0.4 ppm) formaldehyde. The largest decrement in pulmonary function was observed
10 during a workshift following no exposure to diesel exhaust for 10 days. Forced vital capacity
11 and FEV, were significantly reduced over the workshift (0.44 L and 0.30 L,p<0.01 and/><0.001,
12 respectively). There was no difference between smokers and nonsmokers. Maximal
13 midexpiratory flow, closing volume expressed as the percentage of expiratory vital capacity, and
14 alveolar plateau gradient (phase 3) were not affected. Similar but less pronounced effects on
15 FVC (- 0.16 L) were found in a second, subsequent study of stevedores (n = 24) only following
16 5 days of no exposure to diesel truck exhaust. Pulmonary function returned to normal after
17 3 days without occupational exposure to diesel exhaust. No exposure-related correlation was
18 found between the observed pulmonary effects and concentrations of NO, NO2, CO, or
19 formaldehyde; however, it was suggested that NO2 adsorbed onto the diesel exhaust particles
20 may have contributed to the overall dose of NO2 to the lungs. In a related study, six workers (job
21 category not defined) were placed in an exposure chamber and exposed to diluted diesel exhaust
22 containing 0.6 mg/m3 DPM and 3.9 mg/m3 (2.1 ppm) NO2. The exhaust was generated by a
23 6-cylinder, 2.38-L diesel engine, operated for 3 h and 40 min at constant speed, equivalent to
24 60 km/h, and at about one-half full engine load. No effect on pulmonary function was observed.
25
26 5.1.1.1.3. Immunological effects. The potential for DPM to cause immunologic changes has
27 been investigated in several studies. Wade and Newman (1993) reported that diesel exhaust can
28 induce reactive airway disease in railroad workers. In that study, three workers were identified
29 who developed asthma following either a single exposure or a series of short-term exposures to
30 high concentrations of diesel exhaust. Asthma diagnosis was based on symptoms, pulmonary
31 function tests, and measurement of airway hyperreactivity to methacholine or exercise. Exposure
32 occurred as a result of train crews riding in locomotive units trailing immediately behind the lead
33 engine. Although the individuals had worked for the railroad for many years and presumably
34 had been chronically exposed to lower levels of exhaust, the symptoms developed following
35 these subacute incidents. Unfortunately, exposure levels were not measured.
1175/99 5-6 DRAFT—DO NOT CITE OR QUOTE
-------
1 Salvi et al. (1999) exposed healthy human subjects to diluted diesel exhaust (DPM 300
2 y"g/m3) for 1 hr with intermittent exercise. Although there were no changes in. pulmonary
function, there were significant increases in neutrophils and B lymphocytes as well as histamine
and fibronectin in airway lavage fluid. Bronchial biopsies obtained 6 hr after diesel exhaust
5 exposure showed a significant increase in neutrophils, mast cells, CD4+ and CD8+ T
6 lymphocytes along with upregulation of the endothelial adhesion molecules ICAM-1 and
7 VCAM-1, with increases in the number of LFA-1 + in the bronchial tissue. Significant increases
8 in neutrophils and platelets were observed in peripheral blood following exposure to diesel
9 exhaust.
10 In an attempt to evaluate the potential allergenic effects of DPM in humans Diaz-Sanchez
11 and associates carried out a series of clinical investigations. In the first of these (Diaz-Sanchez
12 et al., 1994), healthy human volunteers were challenged by spraying either saline or 0.30 mg
13 DPM into their nostrils. This dose was considered equivalent to total exposure on 1-3 average
14 days in Los Angeles, but could occur acutely in certain nonoccupational settings such as sitting
15 at a busy bus stop or in an express tunnel. Enhanced IgE levels were noted in nasal lung lavage
16 cells in as little as 24 h, with peak production observed 4 days after DPM challenge. The effects
17 seemed to be somewhat isotype-specific, because in contrast to IgE results, DPM challenge had
18 no effect on the levels of IgG, IgA, IgM, or albumin. The selective enhancement of local IgE
production was demonstrated by a dramatic increase in IgE-secreting cells.
Although direct effects of DPM on B-cells have been demonstrated by in vitro studies, it
21 was considered likely that other cells regulating the IgE response may also be affected. Cytokine
22 production was therefore measured in nasal lavage cells from healthy human volunteers
23 challenged with DPM (0 or 0.15 mg in 200 fuL saline) sprayed into each nostril (Diaz-Sanchez et
24 al., 1996). Before challenge with DPM, most subjects' nasal lavage cells had detectable levels of
25 only interferon-y, IL-2, and IL-13 mRNA. After challenge with DPM, the cells produced readily
26 detectable levels of/wRNA for IL-2, IL-4, IL-5, IL-6, IL-10, IL-13, and interferon-y. In addition,
27 all levels of cytokine wRNA were increased. Although the cells in the nasal lavage before and
28 after challenge do not necessarily represent the same ones either hi number or type, the broad
29 increase in cytokine production was not simply the result of an increase in T cells recovered in
30 the lavage fluid. On the basis of these findings, the authors concluded that the increase in nasal
31 cytokine expression after exposure to DPM can be predicted to contribute to enhanced local IgE
32 production and thus play a role in pollutant-induced airway disease.
33 The ability of DPM to act as an adjuvant to the ragweed allergen Amb a I was also
34 examined by nasal provocation in ragweed-allergic subjects using 0.3 mg DPM, Amb a I, or both
(Diaz-Sanchez et al., 1997). Although allergen and DPM each enhanced ragweed-specific IgE,
1175/99 5-7 DRAFT—DO NOT CITE OR QUOTE
-------
1 DPM plus allergen promoted a 16-times greater antigen-specific IgE production. Nasal
2 challenge with DPM also influenced cytokine production. Ragweed challenge resulted in a weak
3 response, DPM challenge caused a strong but nonspecific response, and allergen plus DPM
4 caused a significant increase in the expression of mRNA for THO and TH2rtype cytokines (IL-4,
5 IL-5, IL-6, IL-10, IL-13), with a pronounced inhibitory effect on IFN-y gene expression. The
6 author concluded that DPM can enhance B-cell differentiation and, by initiating and elevating
7 IgE production, may be a factor in the increased incidence of allergic airway disease.
8
9 5.1.1.1.4. Human cell culture studies. The potential mechanisms by which DPM may act to
10 cause allergenic effects has been examined in human cell culture studies. Takenaka et al. (1995)
11 reported that DPM extracts enhanced IgE production from purified human B cells. Interleukin-4
12 plus monoclonal antibody-stimulated IgE production was enhanced 20% to 360% by the addition
13 of DPM extracts over a period of 10-14 days. DPM extracts themselves did not induce IgE
14 production or synergize with interleukin-4 alone to induce IgE from purified B cells, suggesting
15 that the extracts were enhancing ongoing IgE production rather than inducing germline
16 transcription or isotype switching. The authors concluded that enhancement of IgE production in
17 the human airway resulting from the organic fraction of DPM may be an important factor in the
18 increasing incidence of allergic airway disease.
19 Steerenberg et al. (1998) studied the effects of exposure to DPM on airway epithelial
20 cells, the first line of defense against inhaled pollutants. Cells from a human bronchial cell line
21 (BEAS-2B) were cultured in vitro and exposed to DPM (0.04-0.33 mg/mL) and the effects on
22 IL-6 and IL-8 production were observed. Increases in IL-6 and IL-8 production (11- and 4-fold,
23 respectively) were found after 24 or 48 hr exposure to DPM compared to the nonexposed cells.
24 This increase was lower compared to silica (17- and 3.3-fold) and higher compared to titanium
25 dioxide, which showed no increase for either IL-6 or IL-8. The study was extended to observe
26 the effects of DPM on inflammation-primed cells. BEAS-2B cells were exposed to TNF-a
27 followed by DPM. Additive effects on IL-6 and IL-8 production by BEAS-2B cells were found
28 after TNF-a priming and subsequent exposure to DPM only at a low dose of DPM and TNF-a
29 (0.05-0.2 ng/mL). The investigators concluded that BEAS-2B phagocytized DPM and produced
30 an increased amount of IL-6 and IL-8, and that in TNF-a-primed BEAS-2B cells DPM increased
31 interleukin production only at low concentrations of DPM and TNF-a.
32 Ohtoshi et al. (1998) studied the effect of suspended particulate matter (SPM), obtained
33 from high-volume air samplers, and DPM on the production of IL-8 and granulocyte-colony
34 stimuiating factor (GM-CSF) by human airway epithelial cells in vitro. Nontoxic doses of DPMs
35 stimulated production of IL-8 and GM-CSF by three kinds of human epithelial cells (nasal
1175/99 5-8 DRAFT—DO NOT CITE OR QUOTE
-------
1 polyp-derived upper airway, normal bronchial, and transformed bronchial epithelial cells) in a
2 dose- and time-dependent fashion. SPM had a stimulatory effect on GM-CSF, but not on IL-8
production. The effects could be blocked with a protein synthesis inhibitor, suggesting that the
process required de novo protein synthesis, and appeared to be due to an extractable component
5 because neither charcoal nor graphite showed such stimulatory effects. The authors concluded
6 that SPM and DPM, a major component of SPM, may be important air pollutants in the
7 activation of airway cells for the release of cytokines relevant to allergic airway inflammation.
8 The mechanisms underlying DPM-induced injury to airway cells were investigated in
9 human bronchial epithelial cells (HBEC) in culture (Bayram et al., 1998). HBEC from bronchial
10 explants obtained at surgery were cultured and exposed to DPM (10-100 //g/mL) or to a filtered
11 solution of DPM (50 /ug/mL), and the effects on permeability, ciliary beat frequency (CBF), and
12 release of inflammatory mediators were observed. DPM and filtered solution of DPM
13 significantly increased the electrical resistance of the cultures but did not affect movement of
14 bovine serum albumin across cell cultures. DPM and filtered DPM solution significantly
15 attenuated the CBF of these cultures and significantly increased the release of IL-8. DPM also
16 increased the release by these cultures of GM-CSF and soluble intercellular adhesion molecule-1
17 (sIC AM-1). The authors concluded that exposure of airway cells to DPM may lead to functional
18 changes and release of proinflammatory mediators and that these effects may influence the
§ development of airway disease.
Bayram et al. (1998) investigated the sensitivity of cultured airway cells from asthmatic
21 patients to DPM. Incubation with DPM significantly attenuated the CBF hi both the asthmatic
22 and nonasthmatic bronchial epithelial cell cultures. Cultured airway cells from asthmatic
23 patients constitutively released significantly greater amounts of IL-8, GM-CSF, and sICAM-1
24 than cell cultures from nonasthmatic subjects. Only cultures from asthmatic patients additionally
25 released RANTES. The authors concluded that cultured airway cells from asthmatic subjects
26 differ with regard to the amounts and types of proinflammatory mediators they can release and
27 that the increased sensitivity of bronchial epithelial cells of asthmatic subjects to DPM may
28 result in exacerbation of their disease symptoms.
29 Devalia et al. (1999) investigated the potential sensitivity of bronchial epithelial cells
30 (HBEC) biopsied from atopic mild asthmatic patients and non-atopic nonasthmatic subjects to
31 DPM. HBEC from asthmatic patients constitutively released significantly greater amounts of
32 IL-8, GM-CSF, and sICAM-1 than HBEC from nonasthmatic subjects. RANTES was only
33 released by HBEC of asthmatic patients. Incubation of the asthmatic cultures with 10 ug/mL
34 DPM significantly increased the release of IL-8, GM-CSF, and sICAM-1 after 24 h. In contrast,
35 only higher concentrations (50-100 ug/mL DPM) significantly increased the release of IL-8 and
11/5/99 5-9 DRAFT—DO NOT CITE OR QUOTE
-------
1 GM-CSF from HBEC of nonasthmatics. The authors conclude that the increased sensitivity of
2 the airways of asthmatics to DPM may be, at least in part, a consequence of greater constitutive
3 and DPM-induced release of specific pro-inflammatory mediators from bronchial epithelial cells.
4 Boland et al. (1999) compared the biological effects of carbon black and DPM collected
5 from catalyst- and noncatalyst-equipped diesel vehicles hi cultures of both human bronchial
6 epithelial cells (16HBE14o-) and human nasal epithelial cells. Transmission electron
7 microscopy indicated that DPM was phagocytosed by epithelial cells and translocated through
8 the epithelial cell sheet. The time and dose dependency of phagocytosis and its nonspecificity
9 for different particles (DPM, carbon black, and latex particles) were established by flow
10 cytometry. DPM also induced a time-dependent increase in interleukin-8, granulocyte-
11 macrophage colony-stimulating factor, and interleukin-lp release. The inflammatory response
12 occurred later than phagocytosis and, because carbon black had no effect on cytokine release, its
13 extent appeared to depend on the content of absorbed organic compounds. Furthermore,
14 treatment of the exhaust gas to decrease the adsorbed organic fraction reduced the DPM-induced
15 increase hi granulocyte-macrophage colony-stimulating factor release. These results indicate that
16 DPM can be phagocytosed by and induce a specific inflammatory response in airway epithelial
17 cells.
18
19 5.1.1.1.5. Summary. In the available exposure studies, considerable variability is reported in
20 diesel exhaust detection threshold. The odor scales described in some of these studies have no
21 general use at present because they are not objectively defined; however, the studies do clearly
22 indicate substantial interindividual variability in the ability to detect odor and the level at which
23 it becomes objectionable. Much of what is known about the acute effects of diesel exhaust
24 comes from case reports that lack clear measurements of exposure concentrations. The studies of
25 pulmonary function changes in exposed humans have looked for changes occurring over a
26 workshift or after a short-term exposure. The overall conclusion of these studies is that
27 reversible changes in pulmonary function in humans can occur in relation to diesel exhaust
28 sxpcsire, although it is net possible to re1a*e these changes to specific exposure levels. Based on
29 the report by Wade and Newman (1993), reversible airflow obstruction and a syndrome
30 consistent with asthma are possible following acute, high-level exposure to diesel exhaust. The
31 studies by Diaz-Sanchez and co-workers have provided data indicating that DPM is a likely
32 factor in the increasing incidence of allergic hypersensitivity. They have also shown that effects
33 are due primarily to the organic fraction and that DPM synergizes with known allergens to
34 increase their effectiveness. Results from thc.humaii cell culture indicate that DPM has the
1115/99 5-10 DRAFT—DO NOT CITE OR QUOTE
-------
1 potential to influence the development of airway disease through its adjuvant properties and by
2 causing the release of proinflammatory mediators.
5.1.1.2. Long-Term Exposures
5 Several epidemiologic studies have evaluated the effects of chronic exposure to diesel
6 exhaust on occupationally exposed workers.
7 Battigelli et al. (1964) measured several indices of pulmonary function, including vital
8 capacity, FEV,, peak flow, nitrogen washout, and diffusion capacity in 210 locomotive
9 repairmen exposed to diesel exhaust in three engine houses. The average exposure of these
10 locomotive repairmen to diesel exhaust was 9.6 years. When compared with a control group
11 matched for age, body size, "past extrapulmonary medical history" (no explanation given), and
12 job status (154 railroad yard workers), no significant clinical differences were found in
13 pulmonary function or in the prevalence of dyspnea, cough, or sputum between the diesel
14 exhaust-exposed and nonexposed groups. Exposure to diesel exhaust showed marked seasonal
15 variations because the doors of the engine house were open in the summer and closed in the
16 winter. For the exposed group, the maximum daily workplace concentrations of air pollutants
17 measured were 1.8 ppm NO2, 1.7 ppm total aldehydes, 0.15 ppm acrolein, 4.0 ppm SO2, and 5.0
18 ppm total hydrocarbons. The concentration of airborne particles was not reported.
19 Gamble et al. (1987b) examined 283 diesel bus garage workers from four garages in two
^P cities to determine if there was excess chronic respiratory morbidity associated with exposure to
21 diesel exhaust. Tenure of employment was used as a surrogate of exposure; mean tenure of the
22 study population was 9 years ±10 years S.D. Exposure-effect relationships within the study
23 population showed no detectable associations of symptoms with tenure. Reductions in FVC,
24 FEV,, peak flow, and FEFSO (but not FEF7S) were associated with increasing tenure. Compared
25 with a control population (716 nonexposed blue-collar workers) and after indirect adjustment for
26 age, race, and smoking, the exposed workers had a higher incidence of cough, phlegm, and
27 wheezing; however, there was no correlation between symptoms and length of employment.
28 Dyspnea showed an exposure-response trend but no apparent increase in prevalence. Mean
29 FEV,, FVC, FEF50, and peak flow were not reduced in the total cohort compared with the
30 reference population but were reduced in workers with 10 years or more tenure.
31 Purdham et al. (1987) evaluated respiratory symptoms and pulmonary function in
32 17 stevedores employed in car ferry operations who were exposed to both diesel and gasoline
33 exhausts and in a control group of 11 on-site office workers. Twenty-four percent of the exposed
34 group and 36% of the controls were smokers. If a particular symptom was considered to be
35 influenced by smoking, smoking status was used as a covariate in the logistic regression analysis;
11/5/99 5-11 DRAFT—DO NOT CITE OR QUOTE
-------
1 pack-years smoked was a covariate for lung function indices. The frequency of respiratory
2 symptoms was not significantly different between the two groups; however, baseline pulmonary
3 function measurements were significantly different. The latter comparisons were measured by
4 multiple regression analysis using the actual (not percentage predicted) results and correcting for
5 age, height, and pack-years smoked. The stevedores had significantly lower FEV,, FEV,/FVC,
6 FEF50, and FEF75 0<0.02 1 , p<0.023, /KO.OO 1 , and/K0.008, respectively) but not FVC. The
7 results from the stevedores were also compared with those obtained from a study of the
8 respiratory health status of Sydney, Nova Scotia, residents. These comparisons showed that the
9 dock workers had higher FVC, similar FEV,, but lower FEV,/FVC and flow rates than the
1 0 residents of Sydney. Based on these consistent findings, the authors concluded that the lower
1 1 baseline function measurements in the stevedores provided evidence of an obstructive ventilatory
1 2 defect but caution in interpretation was warranted because of the small sample size. There were
1 3 no significant changes in lung function over the workshift, nor was there a difference between
1 4 the two groups. The stevedores were exposed to significantly (p<0.04) higher concentrations of
1 5 paniculate matter (0.06 to 1 .72 mg/m3, mean 0.50 mg/m3) than the controls (0. 13 to 0.58 mg/m3,
1 6 mean not reported). Exposures of stevedores to SO2, NO2, aldehydes, and PAHs were very low;
1 7 occasional CO concentrations in the 20 to 100 ppm range could be detected for periods up to 1 h
1 8 in areas where blockers were chaining gasoline-powered vehicles.
1 9 Additional epidemiological studies on the health hazards posed by exposure to diesel
20 exhaust have been conducted for mining operations. Reger et al. (1982) evaluated the respiratory
2 1 health status of 823 male coal miners from six diesel-equipped mines compared with
22 823 matched coal miners not exposed to diesel exhaust. The average tenure of underground
23 work for the underground miners and their controls was only about 5 years; on average, the
24 underground workers in diesel mines spent only 3 of those 5 years underground in diesel-use
25 mines. Underground miners exposed to diesel exhaust reported a higher incidence of symptoms
26 of cough and phlegm but proportionally fewer symptoms of moderate to severe dyspnea than
27 their matched counterparts. These differences in prevalence of symptoms were not statistically
28 <;it The diesel-exnosed underground miners., on the average, had lower FVC, FEV,,
29 FEF50, FEF7S, and FEF90 but higher peak flow and FEF2S than then- matched controls. These
30 differences, however, were not statistically significant. Health indicators for surface workers and
3 1 their matched controls were directionally the same as for matched underground workers. There
32 were no consistent relationships between the findings of increased respiratory symptoms,
33 decreased pulmonary function, smoking history, years of exposure, or monitored atmosphere
34 pulluiaiiib (NGX, CC, poiticlcS, and aldehydes). Mean ccnccntrG.ticns cf NGX c.t the six mines
35 ranged from 0 to 0.6 ppm for short-term area samples, 0. 13 to 0.28 ppm for full-shift personal
1 1/5/99 5-12 DRAFT— DO NOT CITE OR QUOTE
-------
1 samples, and 0.03 to 0.80 for full-shift area samples. Inhalable particles (less than 10 /urn
2 MMAD) averaged 0.93 to 2.73 mg/m3 for personal samples and 0 to 16.1 mg/m3 for full shift
area samples. Ames et al. (1984), using a portion of the miners studied by Reger, examined
280 diesel-exposed underground miners initially in 1977 and again in 1982. Each miner in this
5 group had at least 1 year of underground mining work history in 1977. The control group was
6 838 miners with no exposure to diesel exhaust. The miners were evaluated for the prevalence of
7 respiratory symptoms, chronic cough, phlegm, dyspnea, and changes in FVC, FEV,, and FEF50.
8 No air monitoring data were reported; exposure to diesel exhaust gases and mine dust particles
9 were described as very low. These authors found no decrements in pulmonary function or
1 0 increased prevalence of respiratory symptoms attributable to exposure to diesel exhaust. In fact,
11 the 5-year incidences of cough, phlegm, and dyspnea were greater in miners without exposure to
1 2 diesel exhaust.
1 3 Attfield (1978) studied 2,659 miners from 21 mines (8 metal, 6 potash, 5 salt, and
14 2 trona). Diesels were employed in only 1 8 of the mines, but the 3 mines not using diesels were
1 5 not identified. The years of diesel usage, ranging from 8 in trona mines to 16 in potash mines,
1 6 were used as a surrogate for exposure to diesel exhaust. Based on a questionnaire, an increased
1 7 prevalence of persistent cough was associated with exposure to aldehydes; this finding, however,
1 8 was not supported by the pulmonary function data. No adverse respiratory symptoms or
19 pulmonary function impairments were related to CO2, CO, NO2, inhalable dust, or inhalable
^P quartz. The author failed to comment on whether the prevalence of cough was related to the high
21 incidence (70%) of smokers in the cohort.
22 Questionnaire, chest radiograph, and spirometric data were collected by Attfield et al.
23 (1982) on 630 potash miners from six potash mines. These miners were exposed for an average
24 of 10 years (range 5 to 14 years) to 0.1 to 3.3 ppm NO2, 0.1 to 4.0 ppm aldehyde, 5 to 9 ppm CO,
25 and total dust concentrations of 9 to 23 mg/m3. No attempt was made to measure diesel-derived
26 particles separately from other dusts. The ratio of total to inhalable (<10 /urn MMAD) dust
27 ranged from 2 to 1 1 . An increased prevalence of respiratory symptoms was related solely to
28 smoking. No association was found between symptoms and tenure of employment, dust
29 exposure, NO2, CO, or aldehydes. A higher prevalence of symptoms of cough and phlegm was
30 found, but no differences in pulmonary function (FVC and FEV,) were found in these
3 1 diesel-exposed potash miners when compared with the predicted values derived from a logistics
32 model based on blue-collar workers working in nondusty jobs.
33 Gamble et al. (1983) investigated respiratory morbidity in 259 miners from five salt
34 mines in terms of increased respiratory symptoms, radiographic findings, and reduced pulmonary
35 function associated with exposure to NO2, inhalable particles (<10 ^.m MMAD), or years worked
1 1 75/99 5-13 DRAFT— DO NOT CITE OR QUOTE
-------
1 underground. Two of the mines used diesel extensively; no diesels were used in one salt mine.
2 Diesels were introduced into each mine in 1956, 1957,1963, or 1963 through 1967. Several
3 working populations were compared with the salt miner cohort. After adjustment for age and
4 smoking, the salt miners showed no increased prevalence of cough, phlegm, dyspnea, or airway
5 obstruction (FEV,/FVC) compared with aboveground coal miners, potash miners, or blue-collar
6 workers. The underground coal miners consistently had an elevated level of symptoms. Forced
7 expiratory volume at 1 s, FVC, FEF50, and FEF75 were uniformly lower for salt miners in relation
8 to all the comparison populations. There was, however, no association between changes in
9 pulmonary function and years worked, estimated cumulative inhalable particles, or estimated
10 NO2 exposure. The highest average exposure to particulate matter was 1.4 mg/m3 (particle size
11 not reported, measurement includes NaCl). Mean NO2 exposure was 1.3 ppm, with a range of
12 0.17 ppm to 2.5 ppm. In a continuation of these studies, Gamble and Jones (1983) grouped the
13 salt miners into low-, intermediate-, and high-exposure categories based on tenure in jobs with
14 diesel exhaust exposure. Average concentrations of inhalable particles and NO2 were 0.40, 0.60,
15 and 0.82 mg/m3 and 0.64, 1.77, and 2.21 ppm for the three diesel exposure categories,
16 respectively. A statistically significant concentration-response association was found between
17 the prevalence of phlegm in the salt miners and exposure to diesel exhaust (p<0.0001) and a
18 similar, but nonsignificant, trend for cough and dyspnea. Changes in pulmonary function
19 showed no association with diesel tenure. In a comparison with the control group of
20 nonexposed, blue-collar workers, adjusted for age and smoking, the overall prevalence of cough
21 and phlegm (but not dyspnea) was elevated in the diesel-exposed workers. Forced expiratory
22 volumes at 1 s and FVC were within 4% of expected, which was considered to be within the
23 normal range of variation for a nonexposed population.
24 In a preliminary study of three subcohorts from bus company personnel (clerks [lowest
25 exposure], bus drivers [intermediate exposure], and bus garage workers [highest exposure])
26 representing different levels of exposure to diesel exhaust, Edling and Axelson (1984) found a
27 fourfold higher risk ratio for cardiovascular mortality in bus garage workers, even after adjusting
28 for smoking history and allowing for at least 10 years of exposure and 15 years or more of
29 induction-latency. Carbon monoxide was hypothesized as the etiologic agent for the increased
30 cardiovascular disease but was not measured. However, in a more comprehensive
31 epidemiological study, Edling et al. (1987) evaluated mortality data covering a 32-year period for
32 a cohort of 694 bus garage employees and found no significant differences between the observed
33 and expected number of deaths from cardiovascular disease. Information on exposure
34 components and their concentrations was not reported.
1175/99 5-14 DRAFT—DO NOT CITE OR QUOTE
-------
1 The absence of reported noncancerous human health effects, other than infrequently
2 occurring effects related to respiratory symptoms and pulmonary function changes, is notable.
f Unlike studies in laboratory animals to be described later in this chapter, studies of the impact of
diesel exhaust on the defense mechanisms of the human lung have not been performed.
5 No direct evidence is available in humans regarding doses of diesel exhaust, gas phase,
6 particulate phase, or total exhaust that lead to impaired particle clearance or enhanced
7 susceptibility to infection. A summary of epidemiology studies is presented in Table 5-1.
8 To date, no large-scale epidemiological study has looked for effects of chronic exposure
9 to diesel exhaust on pulmonary function. In the long-term longitudinal and cross-sectional
i
10 studies, a relationship was generally observed between work in a job with diesel exposure and
11 respiratory symptoms (such as cough and phlegm), but there was no consistent effect on
12 pulmonary function. The interpretation of these results is hampered by lack of measured diesel
13 exhaust exposure levels and the short duration of exposure in these cohorts. The studies are
14 further limited in that only active workers were included, and it is possible that workers who
15 have developed symptoms or severe respiratory disease are likely to have moved away from
16 these jobs. The relationship between work in a job with diesel exposure and respiratory
17 symptoms may be due to short-term exposure.
18
§5.1.2. Laboratory Animal Studies
Because of the large number of statistical comparisons made in the laboratory animal
21 studies and to permit uniform, objective evaluations within and among studies, data will be
22 reported as significantly different (i.e., /K0.05) unless otherwise specified. The exposure
23 regimens used and the resultant exposure conditions employed in the laboratory animal
24 inhalation studies are summarized in Appendix A. Other than the pulmonary function studies
25 performed by Wiester et al. (1980) on guinea pigs during their exposure in inhalation chambers,
26 the pulmonary function studies performed by other investigators, although sometimes
27 unreported, were interpreted as being conducted on the following day or thereafter and not
28 immediately following exposure.
29
30 5.1.2.1. Acute Exposures
31 The acute toxicity of undiluted diesel exhaust to rabbits, guinea pigs, and mice was
32 assessed by Pattle et al. (1957). Four engine operating conditions were used, and 4 rabbits,
33 10 guinea pigs, and 40 mice were tested under each exposure condition for 5 h (no controls were
34 used). Mortality was assessed up to 7 days after exposure. With the engine operating under light
35 load, the exhaust was highly irritating but not lethal to the test species, and only mild tracheal
11/5/99 5-15 DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-1. Human studies of exposure to diesel exhaust
Study
Description
Findings
Acute exposures
Kahn etal. (1988)
El Batawi and
Noweir(1966)
Battigelli(1965)
Katzetal. (1960)
Hare and Springer
(1971)
Hare etal. (1974)
Linnell and Scott
(1962)
Rudell etal. (1990,
1994)
Rudell etal. (1996)
Battigelli(1965)
Wade and Newman
(1993)
Diaz-Saiiehez el al.
(1994)
13 cases of acute exposure, Utah
and Colorado coal miners.
161 workers, two diesel bus
garages.
Six subjects, eye exposure
chamber, three dilutions.
14 persons monitoring diesel
exhaust in a train runnel.
Volunteer panelists who evaluated
general public's response to odor of
diesel exhaust.
Odor panel under highly controlled
conditions determined odor
threshold for diesel exhaust.
Eight healthy non-smoking subjects
exposed for 60 min in chamber to
diesel exhaust (3.7 ppm NO,
1.5 ppm NO2,27 ppm CO,
0.5 mg/m3 formaldehyde, particles
(4.3 * 106/cm3). Exercise, 10 of
each 20 min (75 W).
Volunteers exposed to diesel
exhaust for one hour while doing
•igui WOTiC. LirXpGSUrC
concentrations uncertain.
13 volunteers exposed to three
dilutions of diesel exhaust for
15 min to 1 h.
Three railroad workers acutely
exposed to diesel exhaust.
Voiunieers challenged by a nasal
spray of 0.30 me DPM.
Acute reversible sensory irritation, headache,
nervous system effects, bronchoconstriction were
reported at unknown exposures.
Eye irritation (42%), headache (37%), dizziness
(30%), throat irritation (19%), and cough and
phlegm (11%) were reported in this order of
incidence by workers exposed in the service and
repair of diesel-powered buses.
Time to onset was inversely related and severity
of eye irritation was associated with the level of
exposure to diesel exhaust.
Three occasions of minor eye and throat
irritation; no correlation established with
concentrations of diesel exhaust components.
Slight odor intensity, 90% perceived, 60%
objected; slight to moderate odor intensity, 95%
perceived, 75% objected; moderate odor
intensity, 100% perceived, almost 95% objected.
In six panelists, the volume of air required to
dilute raw diesel exhaust to an odor threshold
ranged from a factor of 140 to 475.
Odor, eye and nasal irritation in 5/8 subjects.
BAL findings small decrease in mast cells,
lymphocyte subsets and macrophage
phagocytpsis, small increase in PMNs.
Unpleasant smell along with irritation of eyes
and nose reported. Airway resistance increased.
Reduction of particle concentration by trapping
did not affect results.
No significant effects on pulmonary resistance
were observed as measured by plethysmography.
The workers developed symptoms of asthma.
Enhancement of IgE production reported due to a
dramatic increase in IsE-secretins cells.
11/5/99
5-16
DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-1. Human studies of exposure to diesel exhaust (continued)
Study
Description
Findings
Takenaka et al.
(1995)
Diaz-Sanchez et al.
(1996)
Diaz-Sanchez et al.
(1997)
Reger(1979)
Ames et al. (1982)
Jorgensen and
Svensson (1970)
Gamble et al. (1979)
Gamble etal. (1987a)
Volunteers challenged by a nasal
spray of 0.30 mg DPM.
Volunteers challenged by a nasal
spray of 0.30 mg DPM.
Ragweed-sensitive volunteers
challenged by a nasal spray of 0.30
mg DPM alone or in combination
with ragweed allergen.
DPM extracts enhanced interleukin-4 plus
monoclonal antibody-stimulated IgE production
as much as 360%, suggesting an enhancement of
ongoing IgE production rather than inducing
germline transcription or isotype switching.
A broad increase in cytokine expression
predicted to contribute to enhanced local IgE
production.
Ragweed allergen plus DPM-stimulated
ragweed-specific IgE to a much greater degree
than ragweed alone, suggesting DPM may be a
key feature in stimulating allergen-induced
respiratory allergic disease.
Studies of cross-shift changes
Five or more VC maneuvers by
each of 60 coal miners exposed to
diesel exhaust at the beginning and
end of a workshift.
Pulmonary function of 60 diesel-
exposed compared with 90 non-
diesel-exposed coal miners over
workshift.
240 iron ore miners matched for
diesel exposure, smoking, and age
were given bronchitis
questionnaires and spirometry pre-
and postworkshift.
200 salt miners performed before
and after workshift spirometry.
Personal environmental NO2 and
inhalable particle samples were
collected.
232 workers in four diesel bus
garages administered acute
respiratory questionnaire and
before and after workshift
spirometry. Compared to lead/acid
battery workers previously found to
be unaffected by their exposures.
FEV,, FVC, and PEFR were similar between
diesel and non-diesel-exposed miners. Smokers
had an increased number of decrements over
shift than nonsmokers.
Significant workshift decrements occurred in
miners in both groups who smoked; no
significant differences in ventilatory function
changes between miners exposed to diesel
exhaust and those not exposed.
Among underground (surrogate for diesel
exposure) miners, smokers, and older age
groups, frequency of bronchitis was higher.
Pulmonary function was similar between groups
and subgroups except for differences accountable
to age.
Smokers had greater but not significant
reductions in spirometry than ex- or nonsmokers.
NO2 but not paniculate levels significantly
decreased FEV1, FEF25, FEF50, and FEF75 over
the workshift.
Prevalence of burning eyes, headache, difficult
or labored breathing, nausea, and wheeze were
higher in diesel bus workers than in comparison
population.
11/5/99
5-17
DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-1. Human studies of exposure to diesel exhaust (continued)
Study
Description
Findings
Ulfvarson et al.
(1987)
Battigelli et al.
(1964)
Gamble etal. (1987b)
Purdham et al. (1987)
Reger etal. (1982)
Workshift changes in pulmonary
function were evaluated in crews of
roll-on/ roll-off ships and car
ferries and bus garage staff.
Pulmonary function was evaluated
in six volunteers exposed to diluted
diesel exhaust, 2.1 ppm NO2, and
0.6 mg/m3 paniculate matter.
Pulmonary function was affected during a
workshift exposure to diesel exhaust, but it
normalized after a few days with no exposure.
Decrements were greater with increasing
intervals between exposures. No effect on
pulmonary function was observed in the
experimental exposure study.
Cross-sectional and longitudinal studies
210 locomotive repairmen exposed
to diesel exhaust for an average of
9.6 years in railroad engine houses
were compared with 154 railroad
yard workers of comparable job
status but no exposure to diesel
exhaust.
283 male diesel bus garage workers
from four garages in two cities
were examined for impaired
pulmonary function (FVC, FEV,,
and flow rates). Study population
with a mean tenure of 9 ± 10 years
S.D. was compared to a
nonexposed blue-collar population.
Respiratory symptoms and
pulmonary function were evaluated
in 17 stevedores exposed to both
diesel and gasoline exhausts in car
ferry operations; control group was
11 on-site office workers.
Differences in respiratory
symptoms and pulmonary function
were assessed in 823 coal miners
from six diesel-equipped mines
compared to 823 matched coal
miners not exposed to diesel
exhaust.
No significant differences in VC, FEV,, peak
flow, nitrogen washout, or diffusion capacity or
in the prevalence of dyspnea, cough, or sputum
were found between the diesel exhaust-exposed
and nonexposed groups.
Analyses within the study population showed no
association of respiratory symptoms with tenure.
Reduced FEV, and FEF50 (but not FEF7S) were
associated with increasing tenure. The study
population had a higher incidence of cough,
phlegm, and wheezing unrelated to tenure.
Pulmonary function was not affected in the total
cohort of diesel-exposed but was reduced with
10 or more years of tenure.
No differences between the two groups for
respiratory symptoms. Stevedores had lower
baseline lung function consistent with an
obstructive ventilatory defect compared with
controls and those of Sydney, Nova Scotia, resi-
dents. Caution in interpretation is warranted due
to small sample size. No significant changes in
lung function over workshift or difference
between two groups.
Underground miners in diesel-use mines reported
more symptoms of cough and phlegm and had
lower pulmonary function. Similar trends were
noted for surface workers at diesel-use mines.
Pattern was consistent with small airway disease
but factors other than exposure to diesel exhaust
thought to be responsible.
11/5/99
5-18
DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-1. Human studies of exposure to diesel exhaust (continued)
Study
Description
Findings
Ames etal. (1984)
Airfield (1978)
Attfield etal. (1982)
Gamble etal. (1983)
Gamble and Jones
(1983)
Changes in respiratory symptoms
and function were measured during
a 5-year period in 280 diesel-
exposed and 838 nonexposed U.S.
underground coal miners.
Respiratory symptoms and function
were assessed in 2,659 miners from
21 underground metal mines (1,709
miners) and nonmetal mines (950
miners). Years of diesel usage in
the mines were surrogate for
exposure to diesel exhaust.
Respiratory symptoms and function
were assessed in 630 potash miners
from six potash mines using a
questionnaire, chest radiographs,
and spirometry. A thorough
assessment of the environment of
each mine was made concurrently.
Respiratory morbidity was assessed
in 259 miners in five salt mines by
respiratory symptoms,
radiographic findings, and
spirometry. Two mines used
diesels extensively, two had limited
use, one used no diesels in 1956,
1957, 1963, or 1963 through 1967.
Several working populations were
compared with the salt mine cohort.
Same as above. Salt miners were
grouped into low-, intermediate-,
and high-exposure categories based
on tenure in jobs with diesel
exposure.
No decrements in pulmonary function or
increased prevalence of respiratory symptoms
were found attributable to diesel exhaust. In fact,
5-year incidences of cough, phlegm, and dyspnea
were greater in miners without exposure to diesel
exhaust than in miners exposed to diesel exhaust.
Questionnaire found an association between an
increased prevalence of cough and aldehyde
exposure; this finding was not substantiated by
spirometry data. No adverse symptoms or
pulmonary function decrements were related to
exposure to NO2, CO, CO2, dust, or quartz.
No obvious association indicative of diesel
exposure was found between health indices, dust
exposure, and pollutants. Higher prevalences of
cough and phlegm but no differences in FVC and
FEV, were found in these diesel-exposed potash
workers when compared with predicted values
from a logistic model based on blue-collar staff
working in nondusty jobs.
After adjustment for age and smoking, salt
miners showed no symptoms, increased
prevalence of cough, phlegm, dyspnea, or air
obstruction (FEV,/FVC) compared with
aboveground coal miners, potash workers, or
blue-collar workers. FEV,, FVC, FEF50, and
FEF75 were uniformly lower for salt miners in
comparison with all the comparison populations.
No changes in pulmonary function were
associated with years of exposure or cumulative
exposure to inhalable particles or NO2.
A statistically significant dose-related association
of phlegm and diesel exposure was noted.
Changes in pulmonary function showed no
association with diesel tenure. Age- and
smoking-adjusted rates of cough, phlegm, and
dyspnea were 145, 169, and 93% of an external
comparison population. Predicted pulmonary
function indices showed small but significant
reductions; there was no dose-response
relationship.
11/5/99
5-19
DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-1. Human studies of exposure to diesel exhaust (continued)
Study
Description
Findings
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Edling and Axelson
(1984)
Pilot study of 129 bus company
employees classified into three
diesel-exhaust exposure categories:
clerks (0), bus drivers (1), and bus
garage workers.
Edling et al. (1987) Cohort of 694 male bus garage
employees followed from 1951
through 1983 was evaluated for
mortality from cardiovascular
disease. Subcohorts categorized by
levels of exposure were clerks (0),
bus drivers (1), and bus garage
employees (2).
The most heavily exposed group (bus garage
workers) had a fourfold increase in risk of dying
from cardiovascular disease, even after
correction for smoking and allowing for 10 years
of exposure and 14 years or more of induction
latency time.
No increased mortality from cardiovascular
disease was found among the members of these
five bus Companies when compared with the
general population or grouped as subcohorts with
different levels of exposure.
and lung damage was observed in the exposed animals. The exhaust contained 74 mg/m3 DPM
(particle size not reported), 560 ppm CO, 23 ppm NO2, and 16 ppm aldehydes. Exhaust
containing 5 mg/m3 DPM, 380 ppm CO, 43 ppm NO2, and 6.4 ppm aldehydes resulted in low
mortality rates (mostly below 10%) and moderate lung damage. Exhaust containing 122 mg/m3
DPM, 418 ppm CO, 51 ppm NO2, and 6.0 ppm aldehydes produced high mortality rates (mostly
above 50%) and severe lung damage. Exhaust containing 1,070 mg/m3 DPM, 1,700 ppm CO,
12 ppm NO2, and 154 ppm aldehydes resulted in 100% mortality in all three species. High CO
levels, which resulted in a carboxyhemoglobin value of 60% in mice and 50% in rabbits and
guinea pigs, were considered to be the main cause of death in the latter case. High NO2 levels
were considered to be the main cause of lung damage and mortality seen in the other three tests.
Aldehydes and NO2 were considered to be the main irritants in the light load test.
Kobayashi and Ito (1995) administered 1, 10, or 20 mg/kg DPM in phosphate-buffered
saline to the nasal mucosa of guinea pigs. The administration increased nasal airway resistance,
augmented increased airway resistance and nasal secretion induced by a histamine aerosol.
increased vascular permeability in dorsal skin, and augmented vascular permeability induced by
histamine. The increases in nasal airway resistance and secretion are considered typical
responses of nasal mucosa against allergic stimulation. Similar results were reported for guinea
pigs exposed via inhalation for 3 h to diesel exhaust diluted to DPM concentrations of either 1 or
3.2 mg/m3 (Kobayashi et al., 1997). These studies show that short-term exposure to DPM
augments nasal mucosal hyperresponsiveness induced by histamine in guinea pigs.
TVi^ ^•pFJa<-»tq r\"f T\P\A or»r1 itc ^nmr^Otl^T^tc /^vtt*a/*t^^ r\oH-ir»1^c or»H Y^arfi/^lo ovfra/^-l-eA r*»i +ViA
-.-.-.— _-.-._ —~ _ _ —— — -- — _____ „__ „•j, v..v^.-._ y —- - ——~ «v_. £^« ..-»-,U ».»•_ f^H_.**.W*.W W~».M.*»W »-*^ •W**. •.»•.—•
release of proinflammatory cytokines, interleukin-1 (IL-1), and tumor necrosis factor-a (TNF-a)
11/5/99
5-20
DRAFT—DO NOT CITE OR QUOTE
-------
1 by alveolar macrophages (AMs) were investigated by Yang et al. (1997). Rat AMs were
2 incubated with 0, 5, 10, 20, 50, or 100 ^g/106 AM/mL of DPM, methanol-extracted DPM, or
4 equivalent concentrations of DPM at 37 °C for 24 h. At high concentrations, both DPM and
DPM extracts were shown to increase IL-1-like activity secreted by AMs, whereas extracted
5 particles had no effect. Neither particles, particle extracts, or extracted particles stimulated
6 secretion of TNF-a. DPM inhibited lipid polysaccharide (LPS)-stimulated production of IL-1
7 and TNF-a. In contrast, interferon- (IFN) y stimulated production of TNF-a was not affected by
8 DPM. Results of this study indicate that the organic fraction of exhaust particles is responsible
9 for the effects noted. Stimulation of IL-1 but not TNF-a suggests that IL-1, but not TNF-a, may
10 play an important role in the development of DPM-induced inflammatory and immune
11 responses. The cellular mechanism involved in inhibiting increased release of IL-1 and TNF-a
12 by LPS is unknown, but may be a contributing factor to the decreased AM phagocytic activity
13 and increased susceptibility to pulmonary infection after prolonged exposure to DPM.
14 Takano et al. (1997) designed a study to evaluate the effects of DPM on the
15 manifestations of allergic asthma in mice, with emphasis on antigen-induced airway
16 inflammation, the local expression of IL-5, GM-CSF, IL-2 and IFN-y, and the production of
17 antigen-specific IgE and IgG. Male ICR mice were intratracheally instilled with ovalbumin
18 (OVA), DPM, and DPM+OVA. DPM was obtained from a 4JB1 -type, light-duty 2.74 L, four-
19 cylinder Izuzu diesel engine operated at a steady speed of 1,500 rpm under a load of 10 torque
^P (kg/m). The OVA-group mice were instilled with 1 ^g OVA at 3 and 6 weeks. The mice
21 receiving DPM alone were instilled with 100 yug DPM weekly for 6 weeks. The OVA + DPM
22 group received the combined treatment in the same protocol as the OVA and the DPM groups,
23 respectively. Additional groups were exposed for 9 weeks. DPM aggravated OVA-induced
24 airway inflammation, characterized by infiltration of eosinophils and lymphocytes and an
25 increase in goblet cells in the bronchial epithelium. DPM in combination with antigen markedly
26 increased IL-5 protein levels in lung tissue and bronchoalveolar lavage supernatants compared
27 with either antigen or DPM alone. The combination of DPM and antigen induced significant
28 increases in local expression of IL-4, GM-CSF, and IL-2, whereas expression of IFN-y was not
29 affected. In addition, DPM exhibited adjuvant activity for the antigen-specific production of IgG
30 and IgE.
31
32 5.1.2.2. Short-Term and Subchronic Exposures
33 A number of inhalation studies have employed a regimen of 20 h/day, 7 days/week for
34 varying exposure periods up to 20 weeks to differing concentrations of airborne particulate
35 matter, vapor, and gas concentrations of diluted diesel exhaust. Exposure regimens and
characterization of gas-phase components for these studies are summarized in Table 5-2.
11/5/99 5-21 DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-2. Short-term effects of diesel exhaust on laboratory animals
^
V
NJ
O
?
6
o
2:
o
H
o
tn
o
o
c:
o
H
m
Species/sex
Rat, F344, M;
Mouse, A/J, M; Hamster,
Syrian, M
Rat, F344, M, F; Mouse,
CD-I.M.F
Cat, Inbred, M
Rat, Sprague-
Dawley, M
Guinea Pig,
Hartley, M, F
Rat, F344,
M
Guinea Pig, Hartley, M
Guinea Pig, Hartley, M
Exposure period
20 h/day
7 days/week
10- 13 weeks
7 h/day
5 days/week
19 weeks
20 h/day
7 days/week
4 weeks
20 h/day
7 days/week
4 weeks
20 h/day
7 days/week
4 weeks
20 h/day
5.5 days/week
4 weeks
30 min
3h
Particles C * T
(ing/m1) (mg'h/rn3)
1.5 2,100(02,730
0.19/wnMMD
0.21 140
1.0 665
4.4 2,926
6.4 3,584
6.4 3,584
6.8' 3,808 '
6.8' 3,808
6.0 2,640
1-2 mg UPM -
Intranasally
1 ' 0.5
3.2 1.6
CO NO, SOj
(ppm) (ppm) (ppm) Effects
6.9 0.49 — Increase in lung wt; increase in
thickness of alveolar walls;
minimal species difference
— — — No effects on lung function in rats
— — — (not done in mice); increase in
— — — PMNs and proteases and AM
aggregation in both species
14.6 2.1 2.1 Few effects on lung function;
focal pneumonitis or alveolitis
16.9 2.49 2.10 Decreased body wt; arterial blood
16.1' 2.76' 1.86' pH reduced; vital capacity, total
lung capacities increased
(<0.0 1 ppm CM*
16.7 2.9 1.9 Exposure started when animals
were 4 days old; increase in
(<0.0 1 ppm O3)' pulmonary flow; bardycardia
— — — Macrophage aggregation; increase
in PMNs; Type H cell
proliferation; thickened alveolar
walls
— — — Augmented increases in nasal
airway resistance and vascular
permeability induced by a
histamine aerosol
5.9 1.4 0.13 Similar results to those reported in
12.9 4.4 0.34 the previous study using
intranasal challenge
Study
Kaplan el al. (1982)
Mauderly ct al. (1981)
Pepelkoetal. (1980a)
Pepelko(1982a)
Wiesteretal. (1980)
White and Garg( 1981)
Kobayashi and Ito
(1995)
Kobayashietal.(1997)
-------
Table 5-2. Short-term effects of diesel exhaust on laboratory animals (continued)
0
Species/sex
Guinea Pig, Hartley, M, F
Mouse ICR, M
Rat, Sprague-Dawley,
M
Exposure period
20 h/day
7 days/week
8 weeks
6 weeks
24 h
Particles C x T
(mg/m1) (mg-h/m1)
6.3 7,056
100/^gDPM -
intranasally
5-100 j/g/106 -
AM/mlofDPM
CO NO, SO,
(ppm) (ppm) (ppm) Effects
17.4 2.3 2.1 Increase in relative lung wt. AM
aggregation; hypertrophy of
goblet cells; focal hyperplasia of
alveolar epithelium
— — — DPM aggravated ovalbumin-
induced airway inflammation and
provided evidence that DPM can
enhance manifestations of allergic
asthma
— — — Unchanged, but not organic-free
DPM enhanced production of
proinflammatory cytokines
Study
Wiesteretal. (1980)
Takanoetal. (1997)
Yang etal.( 1997)
i "Irradiated exhaust.
to
u> PMN = Polymorphonuclear leukocyte.
AM = Alveolar macrophage.
n
O
o
o
H
m
-------
1 Pepelko et al. (1980a) evaluated the pulmonary function of cats exposed under these conditions
2 for 28 days to 6.4 mg/m3 DPM. The only significant functional change observed was a decrease
3 in maximum expiratory flow rate at 10% vital capacity. The excised lungs of the exposed cats
4 appeared charcoal gray, with focal black spots visible on the pleural surface. Pathologic changes
5 included a predominantly peribronchial localization of black-pigmented macrophages within the
6 alveoli characteristic of focal pneumoniris or alveolitis.
7 The effects of a short-term diesel exhaust exposure on arterial blood gases, pH, blood
8 buffering, body weight changes, lung volumes, and deflation pressure-volume (PV) curves of
9 young adult rats were evaluated by Pepelko (1982a). Exposures were 20 h/day, 7 days/week for
10 8 days to a concentration of 6.4 mg/m3 DPM in the nonirradiated exhaust (RE) and 6.75 mg/m3
11 in the irradiated exhaust (IE). In spite of the irradiation, levels of gaseous compounds were not
12 substantially different between the two groups (Table 5-2). Body weight gains were significantly
13 reduced in the RE-exposed rats and to an even greater degree in rats exposed to IE. Arterial
14 blood gases and standard bicarbonate were unaffected, but arterial blood pH was significantly
15 reduced in rats exposed to IE. Residual volume and wet lung weight were not affected by either
16 exposure, but vital capacity and total lung capacity were increased significantly following
17 exposure to RE. The shape of the deflation PV curves were nearly identical for the control, RE
18 and IE groups.
19 In related studies, Wiester et al. (1980) evaluated pulmonary function in 4-day-old guinea
20 pigs exposed for 20 h/day, 7 days/week for 28 days to IE having a concentration of 6.3 mg/m3
21 DPM. When housed in the exposure chamber, pulmonary flow resistance increased 35%, and a
22 small but significant sinus bradycardia occurred as compared with controls housed and measured
23 in control air chambers (/K0.002). Respiratory rate, tidal volume, minute volume, and dynamic
24 compliance were unaffected as were lead-1 electrocardiograms.
25 A separate group of adult guinea pigs was necropsied after 56 days of exposure to IE, to
26 diluted RE, or to clean air (Wiester et al., 1980). Exposure resulted in a significant increase in
9-7 *V.o .-o+Jo nf\nr*n «/e'"l>+ «•" Vu-k/Ur iiwMrrVit fC\ <80/£. fXf /»nntrn1?• (\ *7BO/. fpf 1"P orx-i O SO*^ fXr PTTN
£. • m*c JLl«fc&lS WX A»«» 1^ V* W&^AAV l.^ kyXS^AJT »<• **&^l*«. ^\*.WIS'W USA. VXSAlb*. VAI.*, V?.< V t V AW* *•*_», IMAtt* VS.Vu'V AXS& • - - • J-
?R Wp^rt/body weight ratios wsrs «ot effected by exposure. M'croscopic??Uy. tV»ere w?s 9. rn^rked
29 accumulation of black pigment-laden alveolar macrophages (AM) throughout the lung with a
30 slight to moderate accumulation in bronchial and carinal lymph nodes. Hypertrophy of goblet
31 cells in the tracheobronchial tree was frequently observed, and focal hyperplasia of alveolar
32 lining cells was occasionally observed. No evidence of squamous metaplasia of the
33 tracheobronchial tree, emphysema, peribronchitis, or peribronchiolitis was noted. White and
34 Garg (1981) sluuicu paimnugic aliciatiuiia in ilic lungs uf lots (16 cApuscu ami 8 cGiiuOia) afici
35 exposure to diesel exhaust containing 6 mg/m3 DPM. Two rats from the exposed group and one
11/5/99 5-24 DRAFT—DO NOT CITE OR QUOTE
-------
1 rat from the control group (filtered room air) were sacrificed after each exposure interval of 6 h
2 and 1,3,7, 14, 28,42, and 63 days; daily exposures were for 20 h and were 5.5 days/week.
Evidence of AM recruitment and phagocytosis of diesel particles was found at the 6-h sacrifice;
after 24 h of exposure there was a focal, scattered increase in the number of Type II cells. After 4
5 weeks of exposure, there were morphologic changes in size, content, and shape of AM, septal
6 thickening adjacent to clusters of AMs, and an appearance of inflammatory cells, primarily
7 within the septa. At 9 weeks of exposure, focal aggregations of particle-laden macrophages
8 developed near the terminal bronchi, along with an influx of polymorphonuclear Leukocytes
9 (PMNS), Type II cell proliferation, and thickening of alveolar walls. The affected alveoli
10 occurred in clusters that, for the most part, were located near the terminal bronchioles, but
11 occasionally were focally located in the lung parenchyma. Hypertrophy of goblet cells in the
12 tracheobronchial tree was frequently observed, and focal hyperplasia of alveolar lining cells was
13 occasionally observed. No evidence of squamous metaplasia of the tracheobronchial tree,
14 emphysema, peribronchitis, or peribronchiolitis was noted.
15 Mauderly et al. (1981) exposed rats and mice by inhalation to diluted diesel exhaust for
16 545 h over a 19-week period on a regimen of 7 h/day, 5 days/week at concentrations of 0, 0.21,
17 1.02, or 4.38 mg/m3 DPM. Indices of health effects were minimal following 19 weeks of
18 exposure. There were no significant exposure-related differences in mortality or body weights of
19 the rats or mice. There also were no significant differences in respiratory function (breathing
^P patterns, dynamic lung mechanics, lung volumes, quasi-static PV relationships, forced
21 expirograms, and CO-diffusing capacity) in rats; pulmonary function was not measured in mice.
22 No effect on trachea! mucociliary or deep lung clearances were observed in the exposed groups.
23 Rats, but not°mice, had elevated immune responses in lung-associated lymph nodes at the two
24 higher exposure levels. Inflammation in the lungs of rats exposed to 4.38 mg/m3 DPM was
25 indicated by increases in PMNs and lung tissue proteases. Histopathologic findings included
26 AMs that contained DPM, an increase in Type II cells, and the presence of particles in the
27 interstitium and tracheobronchial lymph nodes.
28 Kaplan et al. (1982) evaluated the effects of subchronic exposure to diesel exhaust on
29 rats, hamsters, and mice. The exhaust was diluted to a concentration of 1.5 mg/m3 DPM;
30 exposures were 20 h/day, 7 days/week. Hamsters were exposed for 86 days, rats and mice for 90
31 days. There were no significant differences in mortality or growth rates between exposed and
32 control animals. Lung weight relative to body weight of rats exposed for 90 days was
33 significantly higher than the mean for the control group. Histological examination of tissues of
34 all three species indicated particle accumulation in the lungs and mediastinal lymph nodes.
35 Associated with the larger accumulations, there was a minimal increase in the thickness of the
11/5/99 5-25 DRAFT—DO NOT CITE OR QUOTE
-------
1 alveolar walls, but the vast majority of the particles elicited no response. After 6 mo of recovery,
2 considerable clearance of the DPM from the lungs occurred in all three species, as evaluated by
3 gross pathology and histopathology. However, no quantitative estimate of clearance was
4 provided.
5 Toxic effects in animals from acute exposure to diesel exhaust appear to be primarily
6 attributable to the gaseous components (i.e., mortality from CO intoxication and lung injury
7 caused by cellular damage resulting from NO2 exposure). The results from short-term exposures
8 indicate that rats experience minimal lung function impairment even at diesel exhaust levels
9 sufficiently high to cause histological and cytological changes in the lung. In subchronic studies
1 0 of durations of 4 weeks or more, frank adverse health effects are not readily apparent and, when
1 1 found, are mild and result from exposure to concentrations of about 6 mg/m3 DPM and durations
1 2 of exposures of 20 h/day. There is ample evidence that subchronic exposure to lower levels of
1 3 diesel exhaust affects the lung, as indicated by accumulation of particles, evidence of
1 4 inflammatory response, AM aggregation and accumulation near the terminal bronchioles, Type II
1 5 cell proliferation, and thickening of alveolar walls adjacent to AM aggregates. Little evidence
1 6 exists, however, that subchronic exposure to diesel exhaust impairs lung function. Recent
1 7 studies have implicated the organic fraction of DPM in the induction of respiratory allergic
1 8 disease.
19
20 5.1.2.3. Chronic Exposures
21 5.1.2.3.1. Effects on growth and longevity. Changes in growth, body weight, absolute or
22 relative organ weights, and longevity can be measurable indicators of chronic toxic effects. Such
23 effects have been observed in some but not all of the long-term studies conducted on laboratory
24 animals exposed to diesel exhaust. There was limited evidence for an effect on survival in the
25 published chronic animal studies; deaths occurred intermittently early in one study in female rats
26 exposed to 3.7 mg/m3 DPM; however, the death rate began to decrease after 1 5 mo, and the
27 survival rate after 30 mo was slightly higher that?. tb?t of the cotrtro! group (Research Corornittee
28 for HERP Studies, 1988). Studies of the effects of chronic exposure to diesel exhaust on survival
29 and body weight or growth are detailed in Table 5-3.
30 Increased lung weights and lung-to-body weight ratios have been reported in rats, mice,
3 1 and hamsters. These data are summarized in Table 5-4. In rats exposed for up to 36 weeks to
32 0.25 or 1 .5 mg/m3 DPM, lung wet weights (normalized to body weight) were significantly higher
33 in the 1 .5 mg/in3 exposure group than control values after 12 weeks of exposure (Misiorowski et
r>_* ---- 1 o — : — i ---- *
CU1U LjrllCUl A1CU11OI
35 diesel exhaust diluted to achieve concentrations of 0.7, 2.2, and 6.6 mg/m3 DPM (Brightwell et
1 1 75/99 5-26 DRAFT— DO NOT CITE OR QUOTE
-------
Table 5-3. Effects of chronic exposures to diesel exhaust on survival and growth of laboratory animals
J\
B
o
j\
i
vj
,»
"^1
3
:d
^.
r
fl
T
3
D
2
•^
2
»]
^
j
— ^
3
D
c
D
— i
Species/sex
Rat, F344, M, F;
Monkey, cynomolgus, M
Rat, F344, M;
Guinea Pig, Hartley, M
Hamster, Chinese, M
Rat, Wistar, M
Rat, F344, M, F;
Mouse, CD- 1,M, F
Rat, Wistar, F;
Mouse, MMRI, F
Rat, F344
M,F
Rat*
F344/Jcl.
Rat, Wistar, F;
Mouse, NMR1, F
(7 mg/m' only)
Exposure
period
7 h/day
5 days/week
104 weeks
20 h/day
5 days/week
106 weeks
8 h/day
7 days/week
26 weeks
6 h/day
5 days/week
87 weeks
7 h/day
5 days/week
130 weeks
19 h/day
5 days/week
104 weeks
16 h/day
5 days/week
104 weeks
16 h/day
6 days/week
130 weeks
18 h/day
5 days/week
24 mo
Particles
(mg/m1)
2.0
0.23-0.36 urn MMD
0.25
0.75
1.5
6.0
12.0
8.3
0.71 urn MMD
0.35
3.5
7.1
0.25 Aim MMD
4.24
0.35 Aim MMD
0.7
2.2
6.6
0.11"
0.41"
1.08d
2.3111
3.72'
0.2-0.3 Aim MMD
0.84
2.5
6.98
(mg-h/rn1)
7,280
2,650
7,950
15,900
8,736
17,472
21,663
1,592
15,925
31,850
41,891
5,824
18,304
54,912
1,373
5,117
13,478
28,829
46,426
7,400
21,800
61,700
CO
(ppm)
11.5
2.7'
4.41
7.1'
_
—
50.0
2.9
16.5
29.7
12.5
—
—
32.0
1.23
2.12
3.96
7.10
12.9
2.6
8.3
21.2
NO,
(ppm)
1.5
0.1"
0.27b
0.5k
__
—
4.0-6.0
0.05
0.34
0.68
1.5
—
—
—
0.08
0.26
0.70
1.41
3.00
0.3
1.2
3.8
SO,
(ppm)
0.8
—
—
^
—
_
_
—
—
1.1
—
—
—
0.38
1.06
2.42
4.70
4.57
0.3
1.1
3.4
Effects
No effects on growth or survival
Reduced body weight in rats at
1.5 mg/mj
No effect on growth
No effect on growth or mortality
rates
No effect on growth or mortality
rates
Reduced body wts; increased
mortality in mice
Growth reduced at 2.2 and
6.6 mg/mj
Concentration-dependent
decrease in body weight; earlier
deaths in females exposed to
3.72 mg/m', stabilized by 15 mo
Reduced body weight in rats at
2.5 and 6.98 mg/m' and no effect
in mice
Study
Lewis et al.
(1989)
Schreck et al.
(1981)
Vinegar etal.
(1981a,b)
Karagianes
etal. (1981)
Mauderly et al.
(1984, 1987a)
Heinrich et al.
(1986a)
Brightwell et al.
(1986)
Research
Committee for
HERP Studies
(1988)
Heinrich et al.
(1995)
-------
(J\
K)
00
Table 5-3. Effects of chronic exposures to diesel exhaust on survival and growth of laboratory animals (continued)
Species/sex
Mice, NMRI, F;
C57BL/6N, F
Rats, F344, M
Mouse, CD-I,
M,F
Exposure
period
18h/day
5 days/week
13.5 mo
(NMRI)
24 mo
(C57BL/N)
16h/day
5 days/week
23 mo
7h/day
5 days/week
104 weeks
Particles
(mg/m1)
6.98
2.44
6.33
0.35
3.5
7.1
0.25 ^m MOD
CxT
(rng-h/m1)
35,500 -NMRI
38,300 -C57
19,520
50,640
1,274
12,740
25,844
CO
(ppm)
14.2
_
—
3
17
30
NO,
(ppm)
2.3
_
—
0.1
0.3
0.7
SO,
(ppm) Effects
2.8 Reduced body weight in NMRI
mice but not in C57BL/6N mice
— Reduced survival in 6.33 mg/m1
— after 300 days. Body weight
significantly lower at 6.33 mg/m'
— No effect on growth or mortality
— rates
—
Study .
1 Icinrich ct al.
(1995)
Nikulaet al.
(1995)
Mauderly et al.
(1996)
'Estimated from graphically depicted mass concentration data.
"•Estimated from graphically presented mass concentration data for NO2 (assuming 90% NO and 10% NO,).
cData for tests with light-duty engine; similar results with heavy-duty engine.
•"Light-duty engine.
'Heavy-duty engine
D
O
§
H
O
/O
d
o
H
tn
-------
Table 5-4. Effects of chronic exposures to diesel exhaust on organ weights and organ-to-body-weight ratios
S
1
to
D
§
V
J_,
O
r>
3
m
O
/O
C
3
m
Species/sex
Rat, F344, M;
Mouse, A/J, M;
Hamster, Syrian,
M
Rat, F344, M, F
Rat, F344, M
Rat, F344, F
Rat, F344; M
Guinea Pig,
Hartley, M
Hamster, Chinese,
M
Rat, Wistar, F;
Hamster, Syrian,
M, F
Mouse, NMRI, F
Rat, F344, M, F;
Hamster, Syrian,
M, F
Cat inbred, M
Mouse, NMRI, F
(7 mg/m' only)
Exposure
period
20 h/day
7 days/week
12-13 weeks
7 h/day
5 days/week
52 weeks
20 h/day
5. 5 days/ week
36 weeks
7 h/day
5 days/week
104 weeks
20 h/day
5. 5 days/ week
78 weeks
8 h/day
7 days/week
26 weeks
19 h/day
5 days/week
120-140 weeks
16 h/day
5 days/week
104 weeks
8 h/day
7 days/week
124 weeks
1 8 h/day
5 days/week
24 mo
Particles
(mg/m1)
1.5
0.19 ^m MMD
2.0
0.23-0.36 ^m
MMD
0.25
1.5
0.19^mMMD
2.0
0.23-0.36 ^m
MMD
0.25
0.75
1.5
0.19 urn MMD •
6.0
12.0
4.24
0.35 ^m MMD
0.7'
2.2b
6.6
6.01
12.0b
0.84
2.5
6.98
CxT CO
(mg-h/m1) (ppm)
2,520-2,730 -
3,640 12.7
990
5,940
7,280 II. 5
2,145 -
6,435 -
12,870
8,736 -
17,472 -
48,336-56,392 12.5
5,824 -
18,304
54,912 32.0
41,664 20.2
83,328 33.2
7,400 2.6
21,800 8.3
61,700 21.2
NO, SO,
(ppm) (ppm) Effects
— — No effect on liver, kidney, spleen, or
heart weights
1.6 0.83 No effects on weights of lungs, liver,
heart, spleen, kidneys, and testes
— — Increase in relative lung weight at
— — 1.5 mg/m1 only initially seen at
12 weeks
1.5 0.81 No effects on heart weights
— — No effects on heart mass
— —
— —
— — Increase in lung weight and lung/body
— — weight ratio
1.5 1.1 Increase in rat, mouse, and hamster
lung weight and dry weights
— — Increase in lung weight concentration
— — related in rats; heart weight/body
— — weight ratio greater at 6.6 mg/m'
2.7 2.7 Decrease in lung and kidney weights
4.4 5.0
0.3 0.3 Increased rat and mouse lung weight
1.2 I.I at 7 mg/m' from 6 mo and at 2.5
3.8 3.4 mg/mj at 22 and 24 mo
Study
Kaplan etal. (1982)
Green etal. (1983)
Misiorowski et al.
(1980)
Vallyathan etal. (1986)
Penney etal. (1981)
Vinegar et al. (1981 a,b)
Heinrich et al.
(1986a,b)
StOber(1986)
Brightwell etal. (1986)
Pepelkoetal. (1980b,
1981)
Moorman etal. (1985)
Heinrich etal. (1995)
-------
Table 5-4. Effects of chronic exposures to diesel exhaust on organ weights and organ-to-body-weight ratios (continued)
vo Species/sex
Mouse, NMR1, F;
C57BL/6N, F
Rats, F344, M
Rat
Mouse
Exposure
period
18h/day
5 days/week
13.5 mo (NMRI)
24 mo
(C57BL/N)
I6h/day
5 days/week
23 mo
Particles C * T
(mg/mj) (mg-h/in1)
6.98 35,500 -NMRI
38,300 -C57
2.44 19,520
6.33 50,640
0.8
2.5
6.98
6.98
4.5
CO NO, SO,
(ppm) (ppm) (ppm) Effects
14.2 2.3 2.8 Increased lung weight
— — — Increase in lung weight was
— — — significant at 2 and 6 mg/m!
Increased lung weight in rats and mice
at 3.5 and 7 mg/m1
Study
Heinrichetal.(1995)
Nikulactal. (1995)
Henderson etal. (1988)
•1 to 61 weeks of exposure.
Y1 b62 to 124 weeks (if exposure.
O
o
o
I
o
i—i
3
o
o
G
I
-------
1 al., 1986). At necropsy, a significant increase in lung weight was seen in both rats and hamsters
2 exposed to diesel exhaust compared with controls. This finding was more pronounced in the rats
3 in which the increase was progressive with both duration of exposure and particulate matter
^P level. The increase was greatest at 30 mo (after the end of a 6-month observation period in the
5 high-concentration male group where the lung weight was 2.7 times the control and at 24 mo in
6 the high-concentration female group [3.9 times control]). Heinrich et al. (1986a,b; see also
7 Stober, 1986) found a significant increase in wet and dry weights of the lungs of rats and mice
8 exposed at 4.24 mg/m3 DPM for 1 year in comparison with controls. After 2 years, the
9 difference was a factor of 2 (mice) or 3 (rats). After the same exposure periods, the hamsters
10 showed increases of 50 to 75%, respectively. Exposure to equivalent filtered diesel exhaust
11 caused no significant effects in any of the species. Vinegar et al. (1980,1981a,b) exposed
12 hamsters to two levels of diesel exhaust with resultant concentrations of about 6 and 12 mg/m3
13 DPM for 8 h/day, 7 days/week for 6 mo. Both exposures significantly increased lung weight and
14 lung weight to body weight ratios. The difference between lung weights of exposed and control
15 hamsters exposed to 12 mg/m3 DPM was approximately twice that of those exposed to 6 mg/m3.
16 Heinrich et al. (1995) reported that rats exposed to 2.5 and 7 mg/m3 DPM for 18 h/day,
17 5 days/week for 24 mo showed significantly lower body weights than control starting at day
18 200 in the high-concentration group and at day 440 in the low-concentration group. Body weight
19 in the low-concentration group was unaffected, as was mortality in any group. Lung weight was
^P increased in the 7 mg/m3 group starting at 3 mo and persisting throughout the study while the
21 2.5 mg/m3 group showed increased lung weight only at 22 and 24 mo of exposure. Mice (NMRI
22 strain) exposed to 7 mg/m3 in this study for 13.5 mo had no increase in mortality and
23 insignificant decreases in body weight. Lung weights were dramatically affected, with increases
24 progressing throughout the study from 1.5-fold at 3 mo to 3-fold at 12 mo. Mice (NMRI and
25 C57BL/6N strains) were also exposed to 4.5 mg/m3 for 23 mo. In NMRI mice, the body weights
26 were reported to be significantly lower than controls, but the magnitude of the change is not
27 reported so biological significance cannot be assessed. Mortality was slightly increased, but
28 statistical significance is not reported. The C57BL/6N mice showed minimal effects on body
29 weight and mortality, which were not statistically significant. Lung weights were dramatically
30 affected in both strains.
31 Nikula et al. (1995) exposed male and female F344 rats to DPM concentrations of 2.4 and
32 6.3 mg/m3 for 16 h/day, 5 days/week, for 23 mo in a study designed to compare the effects of
33 DPM with those of carbon black. Significantly reduced survival was observed in males exposed
34 to 6.3 mg/m3 but not in females or at the lower concentration. Body weights were decreased by
35 exposure to 6.3 mg/m3 DPM in both male and female rats throughout the exposure period.
11/5/99 5-31 DRAFT—DO NOT CITE OR QUOTE
-------
1 Significant increases in lung weight were first seen at 6 mo in the high-exposure group and at
2 12 to 18 mo in the low-exposure group.
3 No evidence was found in the published literature that chronic exposure to diesel exhaust
4 affected the weight of body organs other than the lung and heart (e.g., liver, kidney, spleen, or
5 testes) (Table 5-4). Morphometric analysis of hearts from rats and guinea pigs exposed to 0.25,
6 0.75, or 1 .5 mg/m3 DPM 20 h/day, 5.5 days/week for 78 weeks revealed no significant alteration
7 in mass at any exposure level or duration of exposure (Penney et al., 1981). The analysis
8 included relative wet weights of the right ventricle, left ventricle, combined atria, and ratio of
9 right to left ventricle. Vallyathan et al. (1986) found no significant differences in heart weights
1 0 and the ratio of heart weight to body weight between rats exposed to 2 mg/m3 DPM for 7 h/day,
1 1 5 days/week for 24 mo and their respective clean air chamber controls. No significant
1 2 differences were found in the lungs, heart, liver, spleen, kidney, and testes of rats exposed for
1 3 52 weeks, 7 h/day, 5 days/week to diluted diesel exhaust containing 2 mg/m3 DPM compared
1 4 with their respective controls (Green et al., 1983).
15
1 6 5.1.2.3.2. Effects on pulmonary function. The effect of long-term exposure to diesel exhaust
17 on pulmonary function has been evaluated in laboratory studies of rats, hamsters, cats, and
1 8 monkeys. These studies are summarized in Table 5-5, along with more details on the exposure
1 9 characteristics, in general order of increasing dose (C x T) of DPM. The text will be presented
20 using the same approach.
2 1 Lewis et al. (1989) evaluated functional residual capacity and airway resistance and
22 conductance in 10 control and 10 diesel-exposed rats (2 mg/m3 DPM, 7 h/day, 5 days/week for
23 52 or 104 weeks). At the 104-week evaluation, the rats were also examined for maximum flow
24 volume impairments. No evidence of impaired pulmonary function as a result of the exposure to
25 diesel exhaust was found in rats. Lewis et al. (1989) exposed male cynomolgus monkeys to
26 diesel exhaust for 7 h/day, 5 days/week, for 24 mo. Groups of 15 monkeys were exposed to air,
27 diesel exhaust (2 mg/m3), coal dust, ot combined coal dust and diesel exhaust. Pulmonary
29 including compliance and resistance, static and dynamic lung volumes, distribution of
30 ventilation, diffusing capacity, and maximum ventilatory performance. There were no effects on
3 1 lung volumes, diffusing capacity, or ventilation distribution, so there was no evidence of
32 restrictive disease. There was, however, evidence of obstructive airway disease as measured by
33 low maximal flows in diesel-exposed monkeys. At 1 8 mo of exposure, forced expiratory flow at
34 'iy/o ot vital capacity and torced expiratory tlow normalized to FVC were decreased. The
35 measurement of forced expiratory flow at 40% of total lung capacity was significantly decreased
1 1/5/99 5-32 DRAFT— DO NOT CITE OR QUOTE
-------
Table 5-5. Effects of diesel exhaust on pulmonary function of laboratory animals
^^.
^1
o
o
J\
1
jj
^J
rt
fij
>
T)
3
0
Z
3
-5
3
^
«
D
Species/sex
Rat, F344, M, F
Monkey, M,
Cynomolgus
Rat, F344, M
Rat, Wistar, F
Hamster, Chinese, M
Rat, F344,
M.F
Rat, F344, M, F;
Hamster Syrian, M, F
Hamster, Syrian, M, F
Rat, Wistar, F
Cat, inbred, M
Exposure
period
7 h/day
5 days/week
104 weeks
7 h/day
5 days/week
104 weeks
20 h/day
5.5 days/week
87 weeks
7-8 h/day
5 days/week
104 weeks
8 h/day
7 days/week
26 weeks
7 h/day
5 days/week
130 weeks
16 h/day
5 days/week
104 weeks
19 h/day
5 days/week
120 weeks
19 h/day
5 days/week
140 weeks
8 h/day
7 days/week
124 weeks
Particles
(mg/m1)
2.0
0.23-0.36 nm
MMD
2.0
0.23-0.36 ^m
MMD
1.5
0.19 Mm MMD
3.9
0.1 Mm MMD
6.0
12.0
0.35
3.5
7.1
0.23-0.26 Mm
MMD
0.7
2.2
6.6
4.24
0.35 Mm MMD
4.24
0.35 Mm MMD
6.0"
12.0"
CxT
(mg-h/m3)
7,280
7,280
14,355
14,196-16,224
8,736
17,472
1,593
15,925
31,850
5,824
18,304
54,912
48,336
56,392
41,664
83,328
CO
(ppm)
11.5
11.5
7.0
18.5
—
—
2.9
16.5
29.7
—
—
—
12.5
12.5
20.2
. 33.3
NO2
(ppm)
1.5
1.5
0.5
1.2
—
—
0.05
0.34
0.68
—
—
—
1.5
1.5
2.7
4.4
SO,
(ppm) Effects
0.8 No effect on pulmonary function
0.8 Decreased expiratory flow; no effect
on vital or diffusing capacities
— Increased functional residual capacity,
expiratory volume, and flow
3.1 No effect on minute volume,
compliance, or resistance
— Decrease in vital capacity, residual
— volume, and diffusing capacity;
increase in static deflation lung
volume
— Diffusing capacity, lung compliance
— reduced at 3.5 and 7 mg/m3
_
— Large number of pulmonary function
— changes consistent with obstructive
— and restrictive airway diseases at
6.6 mg/m1 (no specific data provided)
1 . 1 Significant increase in airway
resistance
i . 1 Decrease in dynamic lung
compliance; increase in airway
resistance
2. 1 Decrease in vital capacity, total lung
5.0 capacity, and diffusing capacity after
2 years; no effect on expiratory flow
Study
Lewis etal. (1989)
Lewis etal. (1989)
Gross (1981)
Heinrich etal. (1982)
Vinegar etal. (1980,
1981a,b)
Mauderly etal. (1988)
McClellan etal. (1986)
Brightwell etal. (1986)
Heinrich etal. (1986a)
Heinrich etal. (1986a)
Pepelkoetal. (I980b,
1981)
Moorman etal. (1985)
• 1 to 61 weeks exposure.
b62 to 124 weeks of exposure.
-------
1 at 12, 18, and 24 mo of exposure. The finding of an obstructive effect in monkeys contrasts with
2 the finding of restrictive type effects in other laboratory animal species (Vinegar et al., 1980,
3 1981 a; Mauderly et al., 1988; Pepelko et al., 1980b, 1981) and suggests a possible difference in
4 effect between primate and small animal respiratory tracts. In these monkeys there were no
5 specific histopathological effects reported (see next section) although particle aggregates were
6 reported in the distal airways, suggesting more small airway deposition.
7 Gross (1981) exposed rats for 20 h/day, 5.5 days/week for 87 weeks to diesel exhaust
8 containing 1.5 mg/m3 DPM. When the data were normalized (e.g., indices expressed in units of
9 airflow or volume for each animal by its own forced expiratory volume), there were no apparent
10 functionally significant changes occurring in the lungs at 38 weeks of exposure that might be
11 attributable to the inhalation of diesel exhaust. After 87 weeks of exposure, functional residual
12 capacity (FRC) and its component volumes (expiratory reserve [ER] and residual volume [RV]),
13 maximum expiratory flow (MEF) at 40% FVC, MEF at 20% FVC, and FEV0, were
14 significantly greater in the diesel-exposed rats. An observed increase in airflow at the end of the
15 forced expiratory maneuver when a decreased airflow would be expected from the increased
16 FRC, ER, and RV data (the typical scenario of human pulmonary disease) showed these data to
17 be inconsistent with known clinically significant health effects. Furthermore, although the lung
18 volume changes in the diesel-exposed rats could have been indicative of emphysema or chronic
19 obstructive lung disease, this interpretation was contradicted by the airflow data, which suggest
20 simultaneous lowering of the resistance of the distal airways.
21 Heinrich et al. (1982) evaluated the pulmonary function of rats exposed to a concentration
22 of 3.9 mg/m3 DPM for 7 to 8 h/day, 5 days/week for 2 years. When compared with a control
23 group, no significant changes in respiratory rate, minute volume, compliance, or resistance
24 occurred in the exposed group (number of rats per group was not stated).
25 Hamsters (eight or nine per group) were exposed 8 h/day, 7 days/week, for 6 mo to
26 concentrations of either about 6 mg/m3 or about 12 mg/m3 DPM (Vinegar et al., 1980,1981a,b).
27 Vital capacity, vital capacity/lung weight ratio, residual lung volume by water displacement- and
?8 CO2 d'.ffjisW capacity decreased siraiifieantiy in hamsters exposed to 6 mp7m3 DPM. Static
29 deflation volume-pressure curves showed depressed deflation volumes for diesel-exposed
30 hamsters when volumes were corrected for body weight and even greater depressed volumes
31 when volumes were corrected for lung weight. However, when volumes were expressed as
32 percentage of vital capacity, the diesel-exposed hamsters had higher lung volumes at 0 and 5 cm
33 K2O. In the absence of confirmatory histopathology, the authors tentatively concluded that these
34 elcvaicu iuug vuliiiiicS and the significantly reduced diffusing capacity in the same hamsters
11/5/99 5-34 DRAFT—DO NOT CITE OR QUOTE
-------
1 were indicative of possible emphysematous changes in the lung. Similar lung function changes
2 were reported in hamsters exposed at 12 mg/m3 DPM, but detailed information was not reported.
It was stated, however, that the decrease in vital capacity was 176% greater in the second
experiment than in the first.
5 Mauderly et al. (1988; see also McClellan et al., 1986) examined the impairment of
6 respiratory function in rats exposed for 7 h/day, 5 days/week, for 24 mo to diluted diesel exhaust
7 with 0.35, 3.5, or 7.1 mg/m3 DPM. After 12 mo of exposure to the highest concentration of
8 diesel exhaust, the exposed rats (n = 22) had lower total lung capacity (TLC), dynamic lung
9 compliance (Cdyn), FVC, and CO diffusing capacity than controls (n = 23). After 24 mo of
10 exposure to 7 mg/m3 DPM, mean TLC, Cdyn, quasi-static chord compliance, and CO diffusing
11 capacity were significantly lower than control values. Nitrogen washout and percentage of FVC
12 expired in 0.1 s were significantly greater than control values. There was no evidence of airflow
13 obstruction. The functional alterations were attributed to focal fibrotic and emphysematous
14 lesions and thickened alveolar membranes observed by histological examination. Similar
15 functional alterations and histopathologic lesions were observed in the rats exposed to 3.5 mg/m3
16 DPM, but such changes usually occurred later in the exposure period and were generally less
17 pronounced. There were no significant decrements in pulmonary function for the 0.35 mg/m3
18 group at any time during the study nor were there reported histopathologic changes in this group.
19 Additional studies were conducted by Heinrich et al. (1986a,b; see also Stober, 1986) on
^P the effects of long-term exposure to diesel exhaust on the pulmonary function of hamsters and
21 rats. The exhaust was diluted to achieve a concentration of 4.24 mg/m3 DPM; exposures were
22 for 19 h/day, 5 days/week for a maximum of 120 weeks (hamsters) or 140 weeks (rats). After
23 1 year of exposure to the diesel exhaust, the hamsters exhibited a significant increase in airway
24 resistance and a nonsignificant reduction in lung compliance. For the same time period, rats
25 showed increased lung weights, a significant decrease in Cdyn, and a significant increase in airway
26 resistance. These indices did not change during the second year of exposure.
27 Syrian hamsters and rats were exposed to 0.7,2.2, or 6.6 mg/m3 DPM for five 16-h
28 periods per week for 2 years (Brightwell et al., 1986). There were no treatment-related changes
29 in pulmonary function hi the hamster. Rats exposed to the highest concentration of diesel
30 exhaust exhibited changes in pulmonary function (data not presented) that were reported to be
31 consistent with a concentration-related obstructive and restrictive disease.
32 Pepelko et al. (1980b; 1981; see also Pepelko, 1982b) and Moorman et al. (1985)
33 measured the lung function of adult cats chronically exposed to diesel exhaust. The cats were
34 exposed for 8 h/day and 7 days/week for 124 weeks. Exposures were at 6 mg/m3 for the first 61
35 weeks and 12 mg/m3 from weeks 62 to 124. No definitive pattern of pulmonary function
11/5/99 5-35 DRAFT—DO NOT CITE OR QUOTE
-------
1 changes was observed following 61 weeks of exposure; however, a classic pattern of restrictive
2 lung disease was found at 124 weeks. The significantly reduced lung volumes (TLC, FVC, FRC,
3 and inspiratory capacity [1C]) and the significantly lower single-breath diffusing capacity,
4 coupled with normal values for dynamic ventilatory function (mechanics of breathing), indicate
5 the presence of a lesion that restricts inspiration but does not cause airway obstruction or loss of
6 elasticity. This pulmonary physiological syndrome is consistent with an interstitial fibrotic
7 response that was later verified by histopathology (Plopper et al., 1983).
8 Pulmonary function impairment has been reported in rats, hamsters, cats, and monkeys
9 chronically exposed to diesel exhaust. In all species but the monkey, the pulmonary function
1 0 testing results have been consistent with restrictive lung disease. The monkeys demonstrated
1 1 evidence of small airway obstructive responses. The disparity between the findings in monkeys
1 2 and those in rats, hamsters, and cats could be in part the result of increased particle retention in
1 3 the smaller species resulting from (1) exposure to diesel exhaust that has higher airborne
1 4 concentrations of gases, vapors, and particles and/or (2) longer duration of exposure. The nature
15 of the pulmonary impairment is also dependent on the site of deposition and routes of clearance,
1 6 which are determined by the anatomy and physiology of the test laboratory species and the
1 7 exposure regimen. The data on pulmonary function effects raise the possibility that diesel
1 8 exhaust produces small airway disease in primates compared with primarily alveolar effects in
1 9 small animals and that similar changes might be expected in humans and monkeys.
20 Unfortunately, the available data in primates are too limited to draw clear conclusions.
21
22 5.1.2.3.3. Lung morphology, biochemistry, and lung lavage analysis. Several studies have
23 examined the morphological, histological, and histochemical changes occurring in the lungs of
24 laboratory animals chronically exposed to diesel exhaust. The histopathological effects of diesel
25 exposure in the lungs of laboratory animals are summarized in Table 5-6, ranked in order of C *
26 T. Table 5-6 also contains an expanded description of exposures.
2 "7 l^o»Oof* ^t ol (\ QfiO^ TNOfHFXiTT^H rnp/*T/\p/^r\r\'iW* A.XS.**A..*W«* i.JL.kt*.-*'* VhSW£/*W &•..•.**. *.* «<•**. lu'^W^S^SA'w MIIOII aggicgauca unaivcuioi aiiu uiuiicmuiai auiiacca. me pai i
35 rnacrophages were often in masses near the entrances of the lymphatic drainage and respiratory
1 1/5/99 5-36 DRAFT— DO NOT CITE OR QUOTE
-------
Table 5-6. Histopathological effects of diesel exhaust in the lungs of laboratory animals
-^
,/!
0
•o
J)
1
>J
~J
73
>
m*
T]
3
D
2
H
2
_j
3
o
2
— i
Species/sex
Rat, F344, M;
Mouse, A/J, M;
Hamster, Syrian, M
Monkey,
Cynomolgus, M
Rat, F344, M, F
Rat, Sprague-Dawley,
M; Mouse, A/HEJ, M
Hamster, Chinese, M
Hamster, Syrian, M, F
Rat, Wistar, M
Rat, F344, F
Rat, F344, M, F;
Mouse, CD-I,
M, F
Exposure
period
20 h/day
7 days/week
12-13 weeks
7 h/day
5 days/week
104 weeks
7 h/day
5 days/week
104 weeks
8 h/day
7 days/week
39 weeks
8 h/day
5 days/week
26 weeks
7-8 h/day
5 days/week
120 weeks
6 h/day
5 days/week
87 weeks
8 h/day
7 days/week
104 weeks
7 h/day
5 days/week
130 weeks
Particles
(mg/m')
1.5
0.19 Mm MOD
2.0
0.23-0.36 Mm
MOD
2.0
0.23-0.36 Mm
MOD
6.0
6.0
12.0
3.9
0.1 Mm MOD
8.3
0.71 Mm MOD
4.9
0.35
3.5
7.1
0.23 Mm MOD
CxT
(mg-h/m3)
2,520-2,730
7,280
3,640
13,104
6,240
12,480
16,380-18,720
21,663
28,538
1,592
15,925
31,850
CO NO, SO,
(ppm) (ppm) (ppm) Effects
— — — Inflammatory changes; increase in lung
weight; increase in thickness of alveolar
walls
11.5 1.5 0.8 AM aggregation; no fibrosis,
inflammation, or emphysema
11.5 1.5 0.8 Multifocal histiocytosis; inflammatory
changes; Type II cell proliferation;
fibrosis
— — — Increase in lung protein content and
collagen synthesis but a decrease in
overall lung protein synthesis in both
species; prolylhydroxylase activity
increased in rats in utero
— — — Inflammatory changes; AM accumu-
— — — lation; thickened alveolar lining; Type II
cell hyperplasia; edema; increase in
collagen
18.5 1.2 3.1 Inflammatory changes, 60%
adenomatous cell proliferation
50.0 4.0-6.0 — Inflammatory changes; AM aggregation;
alveolar cell hypertrophy; interstitial
fibrosis, emphysema (diagnostic
methodology not described)
7.0 1.8 13.1 Type II cell proliferation; inflammatory
changes; bronchial hyperplasia; fibrosis
2.9 0.05 — Alveolar and bronchiolar epithelial
16.5 0.34 — metaplasia in rats at 3.5 and 7.0 mg/m';
29.7 0.68 — fibrosis at 7.0 mg/m' in rats and mice;
inflammatory changes
Study
Kaplan etal. (1982)
Lewis etal. (1989)
Bhatnagaret al.
(1980)
Pepelko(l982a)
Bhatnagar et al.
(1980)
Pepelko(l982a)
Pepelko(1982b)
Heinrich etal. (1982)
Karagianes et al.
(1981)
I wai etal. (1986)
Mauderly et al.
(1987a)
Henderson et al.
(1988)
-------
Table 5-6. Histopathological effects of diesel exhaust in the lungs of laboratory animals (continued)
— X.
J*
3
o
s\
i
jj
X
D
a
Tj
~1
1
^
D
_y
2
3
~>
-H
d
TJ
g
«
2
Species/sex
Rat, Wistar, F;
Mouse, NMRI, F
(7 mg/m1 only)
Mouse, NMRI, F;
C57BL/6N, F
Mouse
Rat, M, F,
F344/Jcl.
Hamster, Syrian, M, F
Mouse, NMRI, F
Rat, Wistar, F
Guinea Pig, Hartley, M
Exposure
period
18 h/day
5 days/week
24 mo
18 h/day
5 days/week
13.5 mo (NMRI)
24 mo
(C57BL/N)
16 h/day
6 days/week
1 30 weeks
19 h/day
5 days/week
120 weeks
19 h/day
5 days/week
120 weeks
19 h/day
5 days/week
140 weeks
20 h/day
5.5 days/week
104 weeks
Particles
(mg/m3)
0.8
2.5
6.98
6.98
4.5
0.11"
0.41"
1.081
2.311
3.72b
4.24
4.24
4.24
0.25
0.75
1.5
6.0
CxT
(mg-h/m1)
7,400
21,800
61,700
35,500 - NMRI
38,300 -C57
1,373
5,117
13,478
28,829
46,336
48,336
48,336
56,392
2,860
8,580
17,160
68,640
CO
(ppm)
2.6
8.3
21.2
14.2
1.23
2.12
3.96
7.10
12.9
12.5
12.5
12.5
_
—
—
—
NO,
(ppm)
0.3
1.2
3.8
2.3
0.08
0.26
0.70
1.41
3.00
1.5
1.5
1.5
_
—
—
—
SO,
(ppm)
0.3
1.1
3.4
2.8
0.38
1.06
2.42
4.70
4.57
1.1
1.1
1.1
_
—
—
—
Effects
Bronchioalveolar hyperplasia, interstitial
fibrosis in all groups. Severity and
incidence increase with exposure
concentration
No increase in tumors. Noncancer
effects not discussed
No increase in tumors
Noncancer effects not discussed
Inflammatory changes; Type 11 cell
hyperplasia and lung tumors seen at
>0.4 mg/m1; shortening and loss of cilia
in trachea and bronchi
Inflammatory changes; thickened
alveolar septa; bronchioloalveolar
hyperplasia; emphysema (diagnostic
methodology not described)
Inflammatory changes; bronchiolo-
alveolar hyperplasia; alveolai lipo-
proteinosis; fibrosis
Thickened alveolar septa; AM
aggregation; inflammatory changes;
hyperplasia; lung tumors
Minimal response at 0.25 and
ultrastructural changes at 0.75 mg/m';
thickened alveolar membranes; cell
proliferation; fibrosis at 6.0 mg/m1;
increase in PMN at 0.75 mg/m1 and
1 .5 mg/m'
Study
Heinrich et al. (1995)
Research Committee
for HERP Studies
(1988)
Heinrich et al.
(1986a)
Heinrich et al.
(1986a)
Heinrich et al.
(1986a)
Bamhartetal.(l98l,
1982)
Vostaletal. (1981)
-------
o
o
2
O
H
O
s
o
&
s
o
Table 5-6. Histopathological effects of diesel exhaust in the lungs of laboratory animals (continued)
Species/sex
Cat, inbred, M
Rat, F344, M
Mouse, CD- l.M.F
Exposure
period
8h/day
7 days/week
124 weeks
16h/day
5 days/week
23 mo
7h/day
5 days/week
104 weeks
Particles
(mg/m1)
6.0'
12.0J
2.44
6.33
0.35
3.5
7.1
0.25 /MI MOD
CxT
(mg-h/m1)
41,664
83,328
19,520
50,640
1,274
12,740
25,844
CO
(ppm)
20.2
33.2
—
—
3
17
30
NO,
(ppm)
2.7
4.4
_
—
0.1
0.3
0.7
SO,
(ppm) Effects
2. 1 Inflammatory changes; AM aggregation;
5.0 bronchiolar epithelial metaplasia;
Type II cell hyperplasia; peribronchiolar
fibrosis
— AM hyperplasia, epithelial hyperplasia,
— inflammation, septal fibrosis,
bronchoalveolar metaplasia
— Exposure-related increase in lung soot,
— pigment-laden macrophages, lung
— lesions.
Bronchiolization in alveolar ducts at
7.1 mg/mj
Study
Plopperetal.(l983)
Hyde et at. (1985)
Nikulaetal. (1995)
Maudcrly et al.
(1996)
'Light-duty engine.
'Heavy-duty engine.
cl to 61 weeks exposure.
d62 to 124 weeks of exposure.
AM = Alveolar macrophage.
PMN = Polymorphonuclear leukocyte.
-------
1 ducts. Associated with these masses was a minimal increase in the thickness of the alveolar
2 walls; however, the vast majority of the particles elicited no response. After 6 mo of recovery,
3 the lungs of all three species contained considerably less pigment, as assessed by gross
4 pathological and histopathological examinations.
5 Lewis et al. (1989; see also Green et al., 1983) performed serial histological examinations
6 of rat lung tissue exposed to diesel exhaust containing 2 mg/m3 DPM for 7 h/day, 7 days/week
7 for 2 years. Accumulations of black-pigmented AMs were seen in the alveolar ducts adjacent to
8 terminal bronchioles as early as 3 mo of exposure, and particles were seen within the interstitium
9 of the alveolar ducts. These macular lesions increased in size up to 12 mo of exposure. Collagen
10 or reticulum fibers were seen only rarely in association with deposited particles; the vast majority
1 1 of lesions showed no evidence of fibrosis. There was no evidence of focal emphysema with the
1 2 macules. Multifocal histiocytosis (24% of exposed rats) was observed only after 24 mo of
1 3 exposure. These lesions were most commonly observed subpleurally and were composed of
1 4 collections of degenerating macrophages and amorphous granular material within alveoli,
1 5 together with fibrosis and chronic inflammatory cells in the interstitium. Epithelial lining cells
1 6 adjacent to collections of pigmented macrophages showed a marked Type II cell hyperplasia;
1 7 degenerative changes were not observed hi Type I cells. Histological examination of lung tissue
1 8 from monkeys exposed for 24 mo in the same regimen as used for rats revealed aggregates of
1 9 black particles, principally in the distal airways of the lung. Particles were present within the
20 cytoplasm of macrophages in the alveolar spaces as well as the interstitium. Fibrosis, focal
2 1 emphysema, or inflammation was not observed. No specific histopathological lesions were
22 reported for the monkey.
23 Nikula et al. (1 997) reevaluated the lung tissue from this study. They concluded that
24 there were no significant differences in the amount of retained particulate matter between
25 monkeys and rats exposed under the same conditions. The rats, however, retained a greater
26 portion of the particulate matter in lumens of the alveolar ducts and alveoli than did the monkeys.
27 Conversely, monkeys retained a greater portion of tlic particulate material in the irueistitium than
29 signs of particle-associated inflammation in the monkeys. Minimal histopathologic lesions were
30 detected in the interstitium. Although the lungs of the monkeys showed a marginal and
31 significantly lesser inflammatory response than rats exposed to the same exposure regime, the
32 results should be interpreted with caution because 2 years is near the normal lifetime for rats, but
33 less than 10% of the normal lifespan of Cynomoigus monkeys.
34 Histopathological etlects of diesei exhaust on the lungs of rats have been investigated by
35 the Health Effects Research Program on Diesel Exhaust (HERP) in Japan. Both light-duty (LD)
1 1 75/99 5-40 DRAFT— DO NOT CITE OR QUOTE
-------
1 and heavy-duty (HD) diesel engines were used. The exhaust was diluted to achieve nominal
2 concentrations of 0.1 (LD only), 0.4 (LD and HD), 1 (LD and HD), 2 (LD and HD), and 4 (HD
only) mg/m3 DPM. Rats were exposed for 16 h/day, 6 days/week for 30 mo. No
histopathological changes were observed in the lungs of rats exposed to 0.4 mg/m3 DPM or less.
5 At concentrations above 0.4 mg/m3 DPM, severe morphological changes were observed. These
6 changes consisted of shortened and absent cilia in the tracheal and bronchial epithelium, marked
7 hyperplasia of the bronchiolar epithelium, and swelling of the Type II cellular epithelium. These
8 lesions appeared to increase in severity with increases in exhaust concentration and duration of
9 exposure.' There was no difference in the degree of changes in pulmonary pathology at the same
i
10 concentrations between the LD and the HD series.
11 Histological examination of the respiratory tract of hamsters revealed significantly higher
12 numbers of hamsters exhibiting definite proliferative changes in the lungs in the group exposed
13 to diesel exhaust than were observed in the group exposed to particle-free diesel exhaust or clean
14 air (Heinrich et al., 1982). Sixty percent of these changes were described as adenomatous
15 proliferations. Exposures were for 7 to 8 h/day, 5 days/week for 104 weeks to diesel exhaust
16 diluted to achieve a concentration of 3.9 mg/m3 DPM.
17 Heinrich et al. (1995) reported increased incidence and severity of bronchioloalveolar
18 hyperplasia in rats exposed to 0.8,2.5, and 7 mg/m3. The lesion in the lowest concentration
« group was described as very slight to moderate. Slight to moderate interstitial fibrosis also
increased in incidence and severity in all exposed groups, but incidences were not reported. This
21 chronic study also exposed NMRI mice to 7 mg/m3 for 13.5 mo and both NMRI and C56BL/6N
22 mice to 4.5 mg/m3 for 24 mo. Noncancer histological endpoints are not discussed in any detail in
23 the report, which is focused on the carcinogenicity on diesel as compared with titanium dioxide
24 and carbon black.
25 Iwai et al. (1986) performed serial histopathology on the lungs of rats at 1, 3, 6,12, and
26 24 mo of exposure to diesel exhaust. Exposures were for 8 h/day, 7 days/week for 24 mo; the
27 exposure atmosphere contained 4.9 mg/m3 DPM. At 1 and 3 mo of exposure, there were
28 minimal histological changes in the lungs of the exposed rats. After 6 mo of exposure, there
29 were particle-laden macrophages distributed irregularly throughout the lung and a proliferation of
30 Type II cells with adenomatous metaplasia in areas where the macrophages had accumulated.
31 After 1 year of exposure, foci of heterotrophic hyperplasia of ciliated or nonciliated bronchiolar
32 epithelium on the adjacent alveolar walls were more common, the quantity of deposited
33 particulate matter increased, and the number of degenerative AMs and proliferative lesions of
34 Type II or bronchiolar epithelial cells increased. After 2 years of exposure, there was a fibrous
11/5/99 5-41 DRAFT—DO NOT CITE OR QUOTE
-------
1 thickening of the alveolar walls, mast cell infiltration with epithelial hyperplasia in areas where
2 the macrophages had accumulated, and neoplasms.
3 Heinrich et al. (1986a; see also Stober, 1986) performed histopathologic examinations of
4 the respiratory tract of hamsters, mice, and rats exposed to diesel exhaust that had 4 mg/m3 DPM.
5 Exposures were for 19 h/day, 5 days/week; the maximum exposure period was 120 weeks for
6 hamsters and mice and 140 weeks for rats. Histological examination revealed different levels of
7 response among the three species. In hamsters, the exhaust produced thickened alveolar septa,
8 bronchioloalveolar hyperplasia, and what were termed emphysematous lesions (diagnostic
9 methodology not described). In mice, bronchoalveolar hyperplasia occurred in 64% of the mice
1 0 exposed to the exhaust and in 5% of the controls. Multifocal alveolar lipoproteinosis occurred in
11 71 % and multifocal interstitial fibrosis occurred in 43% of the mice exposed to exhaust but in
1 2 only 4% of the controls. In exposed rats, there were severe inflammatory changes in the lungs,
13 as well as thickened septa, foci of macrophages, and hyperplastic and metaplastic lesions.
1 4 Nikula et al. (1995) reported in detail the nonneoplastic effects in male and female
1 5 F344 rats exposed to 2.4 or 6.3 mg/m3 of DPM. At 3 mo in the low-concentration group,
1 6 enlarged particle-containing macrophages were found with minimal aggregation. With higher
1 7 concentration and longer duration of exposure, the number and size of macrophages and
1 8 aggregates increased. Alveolar epithelial hyperplasia was found starting at 3 mo and in all rats at
19 6 mo. These lesions progressed to chronic active inflammation, alveolar proteinosis, and septal
20 fibrosis at 12 mo. Other lesions observed late in the study included bronchiolar-alveolar
21 metaplasia, squamous metaplasia, and squamous cysts. This study reports hi detail the
22 progression of lesions in diesel exhaust exposure and finds relatively little difference between the
23 lesions caused by diesel exhaust exposure and exposure to similar levels of carbon black
24 particles.
25 The effects of diesel exhaust on the lungs of 1 8-week-old rats exposed to 8.3 ± 2.0 mg/m3
26 DPM were investigated by Karagianes et al. (1981). Exposures were for 6 h/day, 5 days/week,
27 for 4, 8, 16, or 20 mo. Histoiogicai examinations of lung tissue noted focai aggregation of
«^*i* .•" f I _ A "* if _l_1_^t*^.* __ ^ ___ ' *A~A*A**jC*^ __ " _ J 1 1
ZO pai LiciC-iiaucn y-vivia. aivcuicu jLuauuyy 1x1:513, uuciaiiiiai iiuiuala, cuixi cuVcvjicu
29 (diagnostic methodology not described). Lesion severity was related to length of exposure.
30 No significant differences were noted in lesion severity among the diesel exhaust, the diesel
3 1 exhaust plus coal dust (5.8 ±3.5 mg/m3), or the high-concentration (14.9 ± 6.2 mg/m3) coal dust
32 exposure groups following 20 mo of exposure.
33 Histological changes in the lungs of guinea pigs exposed to diluted diesel exhaust
34 containing either 0.25, 0.75, 1 .5, or 6.0 mg/m-i DPM were reported by Barnhart et al. (1 98 1 ;
35 1982). Exposures at 0.75 and 1 .5 mg/m3 for 2 weeks to 6 mo resulted in an uptake of exhaust
1 1/5/99 5-42 DRAFT— DO NOT CITE OR QUOTE
-------
1 particles by three alveolar cell types (AMs, Type I cells, and interstitial macrophages) and also
2 by granulocytic leukocytes (eosinophils). The alveolar-capillary membrane increased in
f thickness as a result of an increase in the absolute tissue volume of interstitium and Type II cells.
In a continuation of these studies, guinea pigs were exposed to diesel exhaust (up to 6.0 mg/m3
5 DPM) for 2 years (Barnhart et al., 1982). A minimal tissue response occurred at the
6 concentration of 0.25 mg/m3 After 9 mo of exposure, there was a significant increase, about
7 30%, in Type I and II cells, endothelial cells, and interstitial cells over concurrent age-matched
8 controls; by 24 mo only macrophages and Type II cells were significantly increased. As in the
9 earlier study, ultrastructural evaluation showed that Type I cells, AMs, and eosinophils
10 phagocytized the diesel particles. Exposure to 0.75 mg/m3 for 6 mo resulted in fibrosis in
11 regions of macrophage clusters and in focal Type II cell proliferation. No additional information
12 was provided regarding the fibrotic changes with increasing concentration or duration of
13 exposure. With increasing concentration/duration of diesel exhaust exposure, Type II cell
14 clusters occurred in some alveoli. Intraalveolar debris was particularly prominent after exposures
15 at 1.5 and 6.0 mg/m3 and consisted of secretory products from Type II cells.
16 In studies conducted on hamsters, Pepelko (1982b) found that the lungs of hamsters
17 exposed for 8 h/day, 7 days/week for 6 mo to 6 or 12 mg/m3 DPM were characterized by large
18 numbers of black AMs in the alveolar spaces, thickening of the alveolar epithelium, hyperplasia
of Type II cells, and edema.
Lungs from rats and mice exposed to 0.35, 3.5, or 7.1 mg/m3 (0.23 to 0.26 /^m mass
21 median diameter [MMD]) for 7 h/day and 5 days/week showed pathologic lesions (Mauderly et
22 al., 1987a; Henderson et al., 1988). After 1 year of exposure at 7.1 mg/m3, the lungs of the rats
23 exhibited focal areas of fibrosis; fibrosis increased with increasing duration of exposure and was
24 observable in the 3.5-mg/m3 group of rats at 18 mo. The severity of inflammatory responses and
25 fibrosis was directly related to the exposure level. In the 0.35 mg/m3 group of rats, there was no
26 inflammation or fibrosis. Although the mouse lungs contained higher burdens of diesel particles
27 per gram of lung weight at each equivalent exposure concentration, there was substantially less
28 inflammatory reaction and fibrosis than was the case in rats. Fibrosis was observed only in the
29 lungs of mice exposed at 7 mg/m3 and consisted of fine fibrillar thickening of occasional alveolar
30 septa.
31 Histological examinations were performed on the lungs of cats initially exposed to
32 6 mg/m3 DPM for 61 weeks and subsequently increased to 12 mg/m3 for Weeks 62 to 124 of
33 exposure. Plopper et al. (1983; see also Hyde et al., 1985) concluded from the results of this
34 study that exposure to diesel exhaust produced changes in both epithelial and interstitial tissue
compartments and that the focus of these lesions in the peripheral lung was the centriacinar
11/5/99 5-43 DRAFT—DO NOT CITE OR QUOTE
-------
1 region where the alveolar ducts join the terminal conducting airways. This conclusion was based
2 on the following evidence. The epithelium of the terminal and respiratory bronchioles in
3 exposed cats consisted of three cell types (ciliated, basal, and Clara cells) compared with only
4 one type (Clara cells) in the controls. The proximal acinar region showed evidence of
5 peribronchial fibrosis and bronchiolar epithelial metaplasia. Type II cell hyperplasia was present
6 in the proximal interalveolar septa. The more distal alveolar ducts and the majority of the rest of
7 the parenchyma were unchanged from controls. Peribronchial fibrosis was greater at the end of
8 6 mo in clean air following exposure, whereas the bronchiolar epithelial metaplasia was most
9 severe at the end of exposure. Following an additional 6 mo in clean air, the bronchiolar
10 epithelium more closely resembled the control epithelial cell population.
11 Wallace et al. (1987) used transmission electron microscopy (TEM) to determine the
12 effect of diesel exhaust on the intravascular and interstitial cellular populations of the lungs of
13 exposed rats and guinea pigs. Exposed animals and matched controls were exposed to 0.25,
14 0.75, 1.5, or 6.0 mg/m3 DPM for 2, 6, or 10 weeks or 18 mo. The results inferred the following:
15 (1) exposure to 6.0 mg/m3 for 2 weeks was insufficient to elicit any cellular response, (2) both
16 species demonstrated an adaptive multicelluiar response to diesel exhaust, (3) increased numbers
17 of fibroblasts were found in the interstitium from week 6 of exposure through month 18, and
18 (4) there was no significant difference in either cell type or number in alveolar capillaries, but
19 there was a significant increase at 18 mo in the mononuclear population in the interstitium of
20 both species.
21 Additional means for assessing the adverse effects of diesel exhaust on the lung are to
22 examine biochemical and cytological changes in bronchoalveolar lavage fluid (BALF) and in
23 lung tissue. Fedan et al. (1985) performed studies to determine whether chronic exposure of rats
24 affected the pharmacologic characteristics of rat airway smooth muscle. Concentration-response
25 relationships for tension changes induced with acetylcholine, 5-hydroxytryptamine, potassium
26 chloride, and isoproterenol were assessed hi vitro on isolated preparations of airway smooth
27 muscle (trachealis). Chronic exposure to diesel exhaust significantly increased the maximal
28 contractile responses to acetylcholine compared with control values; exposure did not alter the
29 sensitivity (EC50 values) of the muscles to the agonists. Exposures were to diesel exhaust
30 containing 2 mg/m3 DPM for 7 h/day, 5 days/week for 2 years.
31 Biochemical studies of BALF obtained from hamsters and rats revealed that exposures to
32 diesel exhaust caused significant increases hi lactic dehydrogenase, alkaline phosphatase,
33 glucose-6-phosphate dehydrogenase (G6P-DH), total protein, collagen, and protease (pH 5.1)
34 after approximately i year and 2 years of exposure (Heinrich et al., 1986a). These responses
11 /5/99 5-44 DRAFT—DO NOT CITE OR QUOTE
-------
1 were generally much greater in rats than in hamsters. Exposures were to diesel exhaust
2 containing 4.24 mg/m3 DPM for 19 h/day, 5 days/week for 120 (hamsters) to 140 (rats) weeks.
Protein, p-glucuronidase activity, and acid phosphatase activity were significantly
elevated in BALF obtained from rats exposed to diesel exhaust containing 0.75 or 1.5 mg/m3
5 DPM for 12 mo (Strom, 1984). Exposure for 6 mo resulted in significant increases in acid
6 phosphatase activity at 0.75 mg/m3 and in protein, p-glucuronidase, and acid phosphatase
7 activity at the 1.5 mg/m3 concentration. Exposure at 0.25 mg/m3 DPM did not affect the three
8 indices measured at either time period. The exposures were for 20 h/day, 5.5 days/week for 52
9 weeks. w
10 Additional biochemical studies (Misiorowski et al., 1980) were conducted on laboratory
11 animals exposed under the same conditions and at the same site as reported on by Strom (1984).
12 In most cases, exposures at 0.25 mg/m3 did not cause any significant changes. The DNA content
13 in lung tissue and the rate of collagen synthesis were significantly increased at 1.5 mg/m3 DPM
14 after 6 mo. Collagen deposition was not affected. Total lung collagen content increased in
15 proportion to the increase in lung weight. The activity of prolyl hydroxylase was significantly
16 increased at 12 weeks at 0.25 and 1.5 mg/m3; it then decreased with age. Lysal oxidase activity
17 did not change. After 9 mo of exposure, there were significant increases in lung phospholipids in
18 rats and guinea pigs exposed to 0.75 mg/m3 and in lung cholesterol in rats and guinea pigs
exposed to 1.5 mg/m3. Pulmonary prostaglandin dehydrogenase activity was stimulated by an
exposure at 0.25 mg/m3 but was not affected by exposure at 1.5 mg/m3 (Chaudhari et al., 1980,
21 1981). Exposures for 12 or 24 weeks resulted in a concentration-dependent lowering of this
22 enzyme activity. Exposure of male rats and guinea pigs at 0.75 mg/m3 for 12 weeks did not
23 cause any changes in glutathione levels of the lung, heart, or liver. Rats exposed for 2 mo at
24 6 mg/m3 showed a significant depletion of hepatic glutathione, whereas the lung showed an
2 5 increase of glutathione (Chaudhari and Dutta, 1982). Schneider and Felt (1981) reported that
26 similar exposures did not substantially change adenylate cyclase and guanylate cyclase activities
27 in lung or liver tissue of exposed rats and guinea pigs.
28 Bhatnagar et al. (1980; see also Pepelko, 1982a) evaluated changes in the biochemistry of
29 lung connective tissue of diesel-exposed rats and mice. The mice were exposed for 8 h/day and
30 7 days/week for up to 9 mo to exhaust containing 6 mg/m3 DPM. Total lung protein content was
31 measured as was labeled pro line and labeled leucine. Leucine incorporation is an index of total
32 protein synthesis, although collagen is very low in leucine. Proline incorporation reflects
33 collagen synthesis. Amino acid incorporation was measured in vivo in the rat and in short-term
34 organ culture in mice. Both rats and mice showed a large increase in total protein (41 to 47% in
11/5/99 5-45 DRAFT—DO NOT CITE OR QUOTE
-------
1 rats), while leucine incorporation declined and proline incorporation was unchanged. These data
2 are consistent with an overall depression of protein synthesis in diesel-exposed animals and also
3 with a relative increase in collagen synthesis compared to other proteins. The increase in
4 collagen synthesis suggested proliferation of connective tissue and possible fibrosis (Pepelko,
5 1982a).
6 A number of reports (McClellan et al., 1986; Mauderly et al., 1987a, 1990a; Henderson
7 et al., 1988) have addressed biochemical and cytological changes in lung tissue and BALF of
8 rodents exposed for 7 h/day, 5 days/week for up to 30 mo at concentrations of 0,0.35, 3.5, or
9 7.1 mg/m3 DPM. At the lowest exposure level (0.35 mg/m3), no biochemical or cytological
10 changes occurred in the BALF or in lung tissue in either Fischer 344 rats or CD-I mice.
11 Henderson et al. (1988) provide considerable time-course information on inflammatory events
12 taking place throughout a chronic exposure. A chronic inflammatory response was seen at the
13 two higher exposure levels in both species, as evidenced by increases in inflammatory cells
14 (macrophages and neutrophils), cytoplasmic and lysosomal enzymes (lactate dehydrogenase,
15 glutathione reductase, and p-glucuronidase), and protein (hydroxyproline) in BALF. Analysis of
16 lung tissue indicated similar changes in enzyme levels as well as an increase in total lung
17 collagen content. After 18 mo of exposure, lung tissue glutathione was depleted in a
18 concentration-dependent fashion in rats but was slightly increased in mice. Lavage fluid levels
19 of glutathione and glutathione reductase activity increased in a concentration-dependent manner
20 and were higher in mice than in rats. Rats exposed for 24 mo to diesel exhaust (3.5 mg/m3 DPM)
21 had a fivefold increase in the bronchoconstrictive prostaglandin PGF2cc and a twofold increase in
22 the inflammatory leukotriene LTB4. In similarly exposed mice, there was a twofold increase in
23 both parameters. These investigators concluded that the release of larger amounts of such
24 mediators of inflammation from the alveolar phagocytic cells of rats accounted for the greater
25 fibrogenic response seen in that species.
26 Biochemical analysis of lung tissue from cats exposed for 124 weeks and held in clean air
27 for an additional 26 weeks, indicated increases of lung coiiagen; this finding was confirmed by an
or> _i 1 ; :_ *_*_i i—_ ,—* ;—:_U4. j ; *± *: , szt , ,*;__„*,A
*-<-> uua&i vcu ujx-it-oot. iii IAJUXI imig wi/i wvigin aim 111 cuiuib^uvc uaouis uucia wDuiiia.ix.ii
29 morphometrically (Hyde et al., 1985). Exposures were for 7 h/day, 5 days/week at 6 mg/m3
30 DPM for 61 weeks and at 12 mg/m3 for weeks 62 to 124.
31 Heinrich et al. (1995) reported on bronchoalveolar lavage in animals exposed for 24 mo
32 and found exposure-related increases in lactate dehydrogenase, P-glucuronidase, protein, and
33 hydroxyproline in groups exposed to 2.5 or 7 mg/m3, although detailed data are not presented.
34 Lavage analyses were not carried out in concurrent studies in mice.
11/5/99 5-46 DRAFT—DO NOT CITE OR QUOTE
-------
1 The pathogenic sequence following the inhalation of diesel exhaust as determined
2 histopathologically and biochemically begins with the interaction of diesel particles with airway
epithelial cells and phagocytosis by AMs. The airway epithelial cells and activated macrophages
release chemotactic factors that attract neutrophils and additional AMs. As the lung burden of
5 DPM increases, there is an aggregation of particle-laden AMs in alveoli adjacent to terminal
6 bronchioles, increases in the number of Type II cells lining particle-laden alveoli, and the
7 presence of particles within alveolar and peribronchial interstitial tissues and associated lymph
8 nodes. The neutrophils and macrophages release mediators of inflammation and oxygen radicals
9 that deplete a biochemical defense mechanism of the lung (i.e., glutathione). As will be
10 described later in more detail, other defense mechanisms are affected, particularly the decreased
11 viability of AMs, which leads to decreased phagocytic activity and death of the macrophage.
12 The latter series of events may result in the presence of pulmonary inflammatory, fibrotic, or
13 emphysematous lesions. The data suggest that there may be a threshold of exposure to diesel
14 exhaust below which adverse structural and biochemical effects may not occur in the lung;
15 however, differences in the anatomy and pathological responses of laboratory animals coupled
16 with their lifespans compared with humans make a determination of human levels of exposure to
17 diesel exhaust without resultant pulmonary injury a difficult and challenging endeavor.
18
5.1.2.3.4. Effects on pulmonary defense mechanisms. The respiratory system has a number of
defense mechanisms that negate or compensate for the effects produced by the injurious
21 substances that repeatedly insult the upper respiratory tract, the tracheobronchial airways, and the
22 alveoli. The effects of exposure to diesel exhaust on the pulmonary defense mechanisms of
23 laboratory animals as well as more details on exposure atmosphere are summarized in Table 5-7
24 and ranked by cumulative exposure (C * T).
25 Several studies have been conducted investigating the effect of inhaled diesel exhaust on
26 the deposition and fate of inert tracer particles or diesel particles themselves. Lung clearance of
27 deposited particles occurs in two distinct phases: a rapid phase (hours to days) from the
28 tracheobronchial region via the mucociliary escalator and a much slower phase (weeks to mo)
29 from the nonciliated pulmonary region via, primarily but not solely, AMs. Battigelli et al. (1966)
30 reported impaired tracheal mucociliary clearance in vitro in excised trachea from rats exposed for
31 single or repeated exposures of 4 to 6 hours at two dilutions of diesel exhaust that resulted in
32 exposures of approximately 8 and 17 mg/m3 DPM. The exposure to 17 mg/m3 resulted in
33 decreased clearance after a single exposure as well as after a cumulative exposure of 34 or
1175/99 5-47 DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-7. Effects of exposure to diesel exhaust on the pulmonary defense mechanisms of laboratory animals
Ul
Species/sex
Exposure
period
Particles
(mg/m1)
CxT
(mg'h/m1)
CO
(ppm)
NO2
(ppm)
SO,
(ppm)
Effects
Study
Guinea Pig,
Hartley
Rat, F344, M
Rat, F344, M
20h/day
5.5
days/week
8 weeks
7h/day
5 days/week
104 weeks
20 h/day
5.5
days/week
26,48, or
52 weeks
0.25
1.5
0.19 Km MOD
2.0
0.23-0.36 nm
MOD
0.25'
0.75'
1.5"
0.19 Mm MOD
ALVEOLAR MACROPHAGE STATUS
220 2.9 - -
1,320 7.5
7,280
11.5
1.5
0.81
715-8,580
2.9
4.8
7.5
oo
No significant changes in absolute numbers
ofAMs
Little effect on viability, cell number,
oxygen consumption, membrane integrity,
lyzomal enzyme activity, or protein content
of AMs; decreased cell volume and ruffling
of cell membrane and depressed
luminescence of AM
AM cell counts proportional to
concentration of DPM at 0.75 and
1.5 mg/m5; AM increased in lungs in
response to rate of DPM mass entering
lung rather than total DPM burden in lung;
increased PMNs were proportional to
inhaled concentrations and/or duration of
exposure; PMNs affiliated with clusters of
aggregated AM rather than DPM
Chenet. al. (1980)
Castranovaelal. (1985)
Strom (1984)
Vostaletal. (1982)
•n
H
3
&
o
g
Rat F344/Crl,
M,F
Mouse, CD, M, F
Rat, Wistar, F
Rat, F344/Crl, M
7 h/day ,
5 days/week
104 weeks
(rat),
78 weeks
(mouse)
18 h/day
5 days/week
24 mo
7 h/day
5 days/week
24 mo
0.35
3.5
7.0
0.25 urn MOD
0.8
2.5
7.1
3.49
1,274C
12,740"
25,480C
7,400
21,800
61,700
12,704
2.9
16.5
29.7
2.6
8.3
21.2
9.8
0.05
0.34
0.68
0.3
1.1
3.4
1.2
— Significant increases of AM in rats and
— mice exposed to 7.0 mg/m5 DPM for 24
— and 18 mo, respectively, but not at
concentrations of 3.5 or 0.35 mg/m5 D'.'M
for the same exposure durations; PMNs
increased in a dose-dependent fashion in
both rats and mice exposed to 3.5 or
7.0 mg/m1 DPM and were greater in mice
than in rats
— Changes in differential cell counts in lung
— lavage
—
— Significantly reduced AM in lavage at 24
mo
Henderson etal. (1988)
Heinrichetal.(1995)
Mauderly etal. (1990a)
-------
Table 5-7. Effects of exposure to diesel exhaust on the pulmonary defense mechanisms of laboratory animals
(continued)
Species/sex
Exposure
period
Particles
(mg/m1)
CxT
(mg'h/rn1)
CO
(ppm)
NOj
(ppm)
SO,
(ppm)
Effects
Study
Rat, M, F
CLEARANCE
7h/day
5 days/week
12 weeks
0.2
1.0
4.5
0.25^m MOD
84
420
1,890
Evidence of apparent speeding of trachea!
clearance at the 4.5 mg/m1 level after 1
week of "Tc macroaggregated-albumin
and reduced clearance of tracer aerosol in
each of the three exposure levels at 12
weeks; indication of a lower percentage of
ciliated cells at the 1.0 and.4.5 mg/m1
levels
Wolff and Gray (1980)
Rat, Wistar, F
Rat, F344, M,
developing 0-6
mo
adult 6- 12 mo
Rat, F344, M, F
Rat, F344, M
Rat, Sprague-
Dawley, M
18h/day
5 days/week
24 mo
7h/day
5 days/week
6 mo
7h/day
5 days/week
18 weeks
7h/day
5 days/week
26-104
weeks
4-6 h/day
7 days/week
O.I to 14.3
weeks
0.8
2.5
7.1
3.55
0.15
0.94
4.1
<0.5 ^m MOD
2.0
0.23-0.36 urn
MOD
0.9
8.0
17.0
7,400 2.6
21,800 8.3
61,700 21.2
3,321 7.9
94.5 -
592 -
2,583 -
•
1,820-7,280 11.5
2.5-10,210 -
—
—
0.3
1.2
3.8
9.5
_
—
—
1.5
5.0
2.7
8.0
0.3 Significant increase in clearance half-time
1.1 of inhaled labeled aerosols in all groups at
3.4 3-18 mo.
Clearance of 2 ^m, aluminosilicate
particles. Half-time significantly increased
in adult, not different in developing rats
— Lung burdens of DPM were concentration-
— related; clearance half-time of DPM almost
— double in 4. 1 mg/m1 group compared to
0.15 mg/m1 group
0.8 No difference in clearance of "FeA
particles 1 day after tracer aerosol
administration; 120 days after exposure
tracer aerosol clearance was enhanced;
lung burden of DPM increased
significantly between 12 and 24 mo of
exposure
0.2 Impairment of tracheal mucocil iary
0.6 clearance in a concentration-response
1 .0 manner
Heinrichetal.(1995)
Mauderly et al. (I987b)
Griffisetal. (1983)
Lewis etal.( 1989)
Battigelli et al. (1966)
-------
NO
Table 5-7. Effects of exposure to diesel exhaust on the pulmonary defense mechanisms of laboratory animals
(continued)
Species/sex
Rat, F344,
M,F
Exposure
period
7h/day
5 days/week
130 weeks
Particles
(mg/ni1)
0.35
3.5
7.1
0.25 nm MDD
CxT
(mg'h/rn1)
1,593
15,925
31,850
CO
(ppm)
2.9
16.5
29.7
NO2
(ppm)
0.1
0.3
0.7
SO2
(ppm) Effects
— No changes in trachea! mucociliary
— clearance after 6, 12, 1 8, 24, or 30 mo of
— exposure; increases in lung clearance half-
times as early as 6 mo at 7.0 mg/m3 level
and 18 mo at 3.5 mg/m3 level; no changes
seen at 0.35 mg/m3 level; after 24 mo of
diesel exposure, long-term clearance
half-times were increased in the 3.5 and
7.0 mg/m3 groups
Study
Wolff et al. (1987)
Rat, F344/Crl, M
7h/day .
5 days/week
24 mo
3.49
12,704
9.8
1.2
MICROBIAL-INDUCED MORTALITY
Mice,CD-l,F
Mice CD-1,1;
7h/day
5 days/week
4, 12, or
26 weeks
2.0
0.23-0.36 urn
MDD
280-1,820
11.5
1.5
0.8
Doubling of long-term clearance half-time
for clearance of 1.0 urn alumino-silicate
particles. Less effect on clearance in
animals with experimentally induced
emphysema
No change in mortality in mice exposed
intratracheally to 100 ng of DPM prior to
exposure to aerosolized Streptococcus sp.
Mortality similar at each exposure duration
when challenged with Ao/PR/8/34
influenza virus; in mice exposed for 3 and
6 mo, but not 1 mo, there were increases in
the percentages of mice having lung
consolidation, higher virus growth,
depressed interferon levels, and a fourfold
reduction in hemagglutinin antibody levels
Mauderly et al. (I990a)
Hatch etal. (1985)
Hahonetal. (1985)
Mice.CR/CD-1,
F
8 h/day 5.3 to 7.9
7 days/week
2 h up to
46 weeks
11-20,350 19
to
22
1.8
to
3.6
0.9
to
2.8
Enhanced susceptibility to lethal effects of
5. pyogenes infections at all exposure
durations (2 and 6 h; 8, 15, 16, 307, and
321 days); inconclusive results with
5. typhimurium because of high mortality
rates in controls; no enhanced mortality
when challenged with A/PR8-3 influenza
virus
Campbell etal. (1980,
1981)
'Chronic exposure lasted 52 weeks.
'Chronic exposure lasted 48 weeks.
'Calculated for 104-week exposure.
DPM = Diesel paniculate matter.
AM = Alveolar macrophage.
PMN = Polymorphonuclear leukocyte.
-------
1 100 hours. Clearance was reduced to a lesser extent and in fewer tracheas from animals exposed
2 to 8 mg/m3 for a cumulative exposure of 40 hours. Lewis et al. (1989) found no difference in the
clearance of 59Fe3O4 particles (1.5 //m MMAD, og 1.8) 1 day after dosing control and diesel
exhaust-exposed rats (2 mg/m3, 7 h/day, 5 days/week for 8 weeks).
5 Wolff et al. (1987) and Wolff and Gray (1980) studied the effects of both subchronic
6 and chronic diesel exhaust exposure on the tracheal clearance of particles. Tracheal clearance
7 assessments were made by measuring the retention of radiolabeled technetium
8 macroaggregated-albumin remaining 1 h after instillation in the distal trachea of rats. In the
9 subchronic studies, rats were exposed to 4.5, 1.0, or 0.2 mg/m3 DPM on a 7 h/day, 5 days/week
10 schedule for up to 12 weeks. After 1 week there was an apparent speeding of tracheal clearance
11 at the 4.5 mg/m3 exposure level (p=0.10), which returned toward baseline after 6 weeks and was
12 slightly below the baseline rate at 12 weeks. In the 1.0 mg/m3 group, there was a progressive
13 significant reduction in the clearance rate at 6 and 12 weeks of exposure. There was a trend
14 toward reduced clearance in the 0.2 mg/m3 group. Scanning electron micrographs indicated
15 minimal changes in ciliary morphology; however, there was an indication of a lower percentage
16 of ciliated cells at the 1.0 and 4.5 mg/m3 levels. In the chronic studies, rats were exposed to 0,
17 0.35, 3.5, or 7.1 mg/m3 for 7 h/day, 5 days/week for 30 mo. There were no significant
18 differences in tracheal clearance rates between the control group and any of the exposure groups
19 after 6, 12, 18,24, or 30 mo of exposure. The preexposure measurements for all groups,
1 however, were significantly lower than those during the exposure period, suggesting a possible
21 age effect. The preexposure value for the 3.5-mg/m3 group was also significantly lower than the
22 control group.
23 There is a substantial body of evidence for an impairment of particle clearance from the
24 bronchiole-alveolar region of rats following exposure to diesel exhaust. Griffis et al. (1983)
25 exposed rats 7 h/day, 5 days/week for 18 weeks to diesel exhaust at 0.15,0.94, or 4.1 mg/m3
26 DPM. Lung burdens of the 0.15, 0.94, and 4.1 mg/m3 levels were 35, 220, and 1,890 ^g/g lung,
27 respectively, 1 day after the 18-week exposure. The clearance half-time of the DPM was
28 significantly greater, almost double, for the 4.1 mg/m3 exposure group than for those of the lower
29 exposure groups, 165 ± 8 days versus 99 ± 8 days (0.94 mg/m3) and 87 ± 28 days (0.15 mg/m3),
30 respectively.
31 Chan et al. (1981) showed a dose-related slowing of MC-diesel particle clearance in rats
32 preexposed to diesel exhaust at 0.25 or 6 mg/m3 particulate matter for 20 h/day, 7 days/week for
33 7 to 112 days. Clearance was inhibited in the 6 mg/m3 group when compared by length of
34 exposure or compared with the 0.25 mg/m3 or control rats at the same time periods.
11/5/99 5-51 DRAFT—DO NOT CITE OR QUOTE
-------
1 Heinrich et al. (1982) evaluated lung clearance in rats exposed for approximately 18 mo
2 at 3.9 mg/m3 DPM for 7 to 8 h/day, 5 days/week. Following exposure to 59Fe2O3-aerosol, the rats
3 were returned to the diesel exhaust exposure and the radioactivity was measured over the
4 thoracic area at subsequent times. The biological half-life of the iron oxide deposited in the rats'
5 lungs was nearly twice that of controls.
6 Heinrich also used labeled iron oxide aerosols to study clearance in rats exposed to 0.8,
7 2.5, or 7 mg/m3 diesel DPM for 24 mo (Heinrich et al., 1995). Clearance measurements were
8 carried out at 3, 12, and 18 mo of exposure. Half-times of clearance were increased in a
9 concentration- and duration-related way in all exposed groups, with a range of a 50% increase in
10 the 0.8 mg/m3 group at 3 mb to an 11-fold increase in the 7 mg/m3 group at 19 mo. The
11 differential cell counts in these animals were stated to have shown clear effects in the 2.5 and 7
12 mg/m3 groups, but specific information about the changes is not reported.
13 Wolff et al. (1987) investigated alterations in DPM clearance from the lungs of rats
14 chronically exposed to diesel exhaust at 0, 0.35, 3.5, or 7.1 mg/m3 DPM for 7 h/day, 5 days/week
15 for up to 24 mo. Progressive increases in lung burdens were observed over time in the 3.5 and
16 7.1 mg/m3 exposure groups. Levels of DPM in terms of milligrams per lung were 0.60, 11.5, and
17 20.5 after 24 mo of exposure at the 0.35, 3.5, or 7.1 mg/m3 exposure levels, respectively. There
18 were significant increases hi 16-day clearance half-times of inhaled radiolabeled particles of
19 67Ga2O3 (0.1 /am MMD) as early as 6 mo at the 7.1 mg/m3 level and 18 mo at the 3.5 mg/m3
20 level; no significant changes were seen at the 0.35 mg/m3 level. Rats inhaled fused
21 aluminosilicate particles (2 //m MMAD) labeled with 134Cs after 24 mo of diesel exhaust
22 exposure; long-term clearance half-tunes were 79, 81,264, and 240 days for the 0, 0.35, 3.5, and
23 7.1 mg/m3 groups, respectively. Differences were significant between the control and the 3.5 and
24 7.1 mg/m3 groups (p < 0.01).
25 Mauderly et al. (1987b) compared the effects of diesel exhaust hi the developing lung to
26 the adult lung by exposing groups of male F344 rats to 3.5 mg/m3 for 7 h/day, 5 days/week for 6
27 mo. One group (adult) was exposed between 6 and 12 mo of age, and the ether was exposed
28 beginning in utsrc cs:d 'jntil 6 rt:c cf age. Clearance of an inhaled moiiodispers5 2 /*!?.
29 aluminosilicate particle was measured after exposure for 6 mo. The clearance half-time of the
30 slow phase was found to be doubled in adult rats compared with age-matched controls and was
31 not significantly affected in developing rat lungs.
32 Mauderly et al. compared the effects of diesel exhaust hi normal lungs with rats in
33 which emphysema had been induced experimentally by instillation of elastase 6 weeks before
34 diesel exhaust exposures. The rats were exposed to 5.5 mg/'nr DFIvi for 7 h/'day, 5 days/week foi
35 24 mo. Measurements included histopathology, clearance, pulmonary function, lung lavage, and
11/5/99 5-52 DRAFT—DO NOT CITE OR QUOTE
-------
1' immune response. In the rats that were not pretreated with elastase, there was a significant
2 reduction in the number of macrophages recovered by pulmonary lavage in contrast to the
J3 increases in macrophages reported by Strom (1984) and Henderson et al. (1988). The half-time
of the slow phase of clearance of inhaled, 1 /urn, monodisperse particles was doubled in the
5 exposure animals without elastase pretreatment. The elastase pretreatment did not affect
6 clearance in unexposed animals but significantly reduced the effect of diesel. The clearance
7 half-time was significantly less in elastase-pretreated, diesel-exposed animals than in
8 diesel-exposed normal animals. Many other effects measured in this study were also less
9 affected by diesel exposure in elastase-treated animals. Measurements of lung burden of DPM
10 showed that elastase-pretreated animals accumulated less than half as much DPM mass as
11 normal animals exposed at the same time, suggesting that the difference in effect could be
12 explained by differences in dose to the lung.
13 Lewis et al. (1989) conducted lung burden and 59Fe3O4 tracer studies in rats exposed for
14 12 and 24 mo to 2 mg/m3 DPM (7 h/day, 5 days/week). The slope of the Fe3O4 clearance curve
15 was significantly steeper than that of the controls, indicating a more rapid alveolar clearance of
16 the deposited 59Fe3O4. After 120 days from the inhalation of the tracer particle, 19% and 8% of
17 the initially deposited 59Fe3O4 were present in the lungs of control and diesel exhaust-exposed
18 rats, respectively. The lung burden of DPM, however, increased significantly between 12 and
19 24 mo of exposure (0.52 to 0.97% lung dry weight), indicating a later dose-dependent inhibition
^^ of clearance.
21 Alveolar macrophages, because of their phagocytic and digestive capabilities, are one of
22 the prime defense mechanisms of the alveolar region of the lung against inhaled particles. Thus,
23 characterization of the effects of diesel exhaust on various properties of AMs provides
24 information on the integrity or compromise of a key pulmonary defense mechanism. The
25 physiological viability of AMs from diesel-exposed rats was assessed after 2 years of exposure
26 by Castranova et al. (1985). The 7 h/day, 5 days/week exposure at 2 mg/m3 DPM had little effect
27 on the following: viability, cell number, oxygen consumption, membrane integrity, lysosomal
28 enzyme activity, or protein content of the AMs. A slight decrease in cell volume, a decrease in
29 chemiluminescence indicative of a decreased secretion of reactive oxygen species, and a decrease
30 in ruffling of the cell membrane were observed. These findings could be reflective of an overall
31 reduction in phagocytic activity.
32 Exposure to diesel exhaust has been reported both to increase the number of recoverable
33 AMs from the lung (Strom, 1984; Vostal et al., 1982; Henderson et al., 1988) or to produce no
34 change in numbers (Chen et al., 1980; Castranova et al., 1985). Strom (1984) found that in rats
35 exposed to 0.25 mg/m3 DPM for 20 h/day, 5.5 days/week for 6 mo or 1 year, as well as in the
11/5/99 5-53 DRAFT—DO NOT CITE OR QUOTE
-------
1 controls, BAL cells consisted entirely of AMs, with no differences in the cell counts in the lavage
2 fluid. At the higher concentrations, 0.75 or 1.5 mg DPM/m3, the count of AM increased
3 proportionally with the exposure concentration; the results were identical for AMs at both 6 and
4 11 or 12 mo of exposure. The increase in AM counts was much larger after exposure to
5 1.5 mg/m3 DPM for 6 mo than after exposure to 0.75 mg/m3 for 1 year, although the total mass
6 (calculated as C * T) of deposited particulate burden was the same. These data suggested to the
7 authors that the number of lavaged AMs was proportional to the mass influx of particles rather
8 than to the actual DPM burden in the lung. These results further implied that there may be a
9 threshold for the rate of mass influx of DPM into the lungs of rats above which there was an
10 increased recruitment of AMs. Henderson et al. (1988) reported similar findings of significant
11 increases of AMs in rats and mice exposed to 7.1 mg/m3 DPM for 18 and 24 mo, respectively,
12 for 7 h/day, 5 days/week, but not at concentrations of 3.5 or 0.35 mg/m3 for the same exposure
13 durations. Chen et al. (1980), using an exposure regimen of 0.25 and 1.5 mg/m3 DPM for 2 mo
14 and 20 h/day and 5.5 days/week, found no significant changes in absolute numbers of AMs from
15 guinea pig bronchoalveolar lavage fluid (BALF), nor did Castranova et al. (1985) in rat BALF
16 following exposure to 2 mg/m3 DPM for 7 h/day, 5 days/week for 2 years.
17 A similar inflammatory response was noted by Henderson et al. (1988) and Strom
18 (1984), as evidenced by an increased number of PMNs present in BALF from rodents exposed to
19 diesel exhaust. Henderson et al. (1988) found these changes in rats and mice exposed to 7.1 and
20 3.5 mg/m3 DPM for 7 h/day, 5 days/week. Significant increases in BALF PMNs were observed
21 in mice at 6 mo of exposure and thereafter at the 7.1 and 3.5 mg/m3 exposure levels, but in rats
22 only the 7.1 mg/m3 exposure level showed an increase in BALF PMNs at 6 mo of exposure and
23 thereafter. Significant increases in BALF PMNs occurred in rats at 12,18, and 24 mo of
24 exposure to 3.5 mg/m3 DPM. Although increases in PMNs were usually greater in mice in terms
25 of absolute numbers, the PMN response in terms of increase relative to controls was only about
26 one-third that of rats. Strom (1984) reported that the increased numbers of PMNs in BALF were
27 proportional to the inhaled concentrations and/or duration of exposure. The PMNs also appeared
28 to be affiliated with clusters of aggregated AMs rather than to the diesel particles per se.
29 Proliferation of Type II cells likewise occurred in response to the formed aggregates of AMs
30 (White and Garg, 1981).
31 The integrity of pulmonary defense mechanisms can also be ascertained by assessing if
32 exposure to diesel exhaust affects colonization and clearance of pathogens and alters the response
33 of the challenged animals to respiratory tract infections. Campbell et al. (1980, 1981) exposed
34 mice to diesel exhaust followed by infectious challenge with Salmonella typhimurwm,
35 Streptococcuspyogenes, or A/PR8-3 influenza virus and measured microbial-induced mortality.
11/5/99 5-54 DRAFT—DO NOT CITE OR QUOTE
-------
1 Exposures to diesel exhaust were to 6 mg/m3 DPM for 8 h/day, 7 days/week for up to 321 days.
2 Exposure to diesel exhaust resulted in enhanced susceptibility to the lethal effects of S. pyogenes
infection at all exposure durations (2 h, 6 h; 8, 15,16, 307, and 321 days). Tests with S.
typhimurium were inconclusive because of high mortality rates in the controls. Mice exposed to
5 diesel exhaust did not exhibit an enhanced mortality when challenged with the influenza virus.
6 Hatch et al. (1985) found no changes in the susceptibility of mice to Group C Streptococcus sp.
7 infection following intratracheal injection of 100 t*g of DPM suspended in unbuffered saline.
8 Hahon et al. (1985) assessed virus-induced mortality, virus multiplication with
9 concomitant interferon (IFN) levels (lungs and sera), antibody response, and lung histopathology
10 in mice exposed to diesel exhaust prior to infectious challenge with Ao/PR/8/34 influenza virus.
11 Weanling mice were exposed to the diesel exhaust containing 2 mg/m3 DPM for 7 h/day,
12 5 days/week. In mice exposed for 1, 3, and 6 mo, mortality was similar between the exposed and
13 control mice. In mice exposed for 3 and 6 mo, however, there were significant increases in the
14 percentage of mice having lung consolidation, higher virus growth, depressed interferon levels,
15 and a fourfold reduction in hemagglutinin antibody levels; these effects were not seen after the
16 1-mo exposure.
17 The effects of diesel exhaust on the pulmonary defense mechanisms are determined by
18 three critical factors related to exposure: the concentrations of the pollutants, the exposure
19 duration, and the exposure pattern. Higher doses of diesel exhaust as determined by an increase
^P in one or more of these three variables have been reported to increase the numbers of AMs,
21 PMNs, and Type II cells in the lung, whereas lower doses fail to produce such changes. The
22 single most significant contributor to the impairment of the pulmonary defense mechanisms
23 appears to be an excessive accumulation of DPM, particularly as particle-laden aggregates of
24 AMs. Such an accumulation would result from an increase in deposition and/or a reduction in
25 clearance. The deposition of particles does not appear to change significantly following
26 exposure to equivalent diesel exhaust doses over time. Because of the significant nonlinearity in
27 particle accumulation between low and high doses of diesel exhaust exposure, coupled with no
28 evidence of increased particle deposition, an impairment in one or more of the mechanisms of
29 pulmonary defense appears to be responsible for the DPM accumulation and subsequent
30 pathological sequelae. The time of onset of pulmonary clearance impairment was dependent
31 both on the magnitude and on the duration of exposures. For example, for rats exposed for
32 7 h/day, 5 days/week for 104 weeks, the concentration needed to induce pulmonary clearance
33 impairment appears to lie between 0.35 and 2.0 mg/m3 DPM.
34
35 5.1.2.3.5. Effects on the immune system—inhalation studies. The effects of diesel exhaust on
^^ the immune system of guinea pigs were investigated by Dziedzic (1981). Exposures were to 1.5
37 mg/m3 DPM for 20 h/day, 5.5 days/week for up to 8 weeks. There was no effect of diesel
11/5/99 5-55 DRAFT—DO NOT CITE OR QUOTE
-------
1 exposure when compared with matched controls for the number of B and T lymphocytes and null
2 cells isolated from the tracheobronchial lymph nodes, spleen, and blood. Cell viability as
3 measured by trypan blue exclusion was comparable between the exposed and control groups. The
4 results of this study and others on the effects of exposure to diesel exhaust on the immune system
5 are summarized in Table 5-8.
6 Mentnech et al. (1984) examined the effect of diesel exhaust on the immune system of
7 rats. Exposures were to 2 mg/m3 DPM for 7 h/day, 5 days/week for up to 2 years. Rats exposed
8 for 12 and 24 mo were tested for immunocompetency by determining antibody-producing cells
9 in the spleen 4 days after immunization with sheep erythrocytes. The proliferative response of
10 splenic T-lymphocytes to the mitogens concanavalin A and phytohemagglutinin was assessed in
11 rats exposed for 24 mo. There were no significant differences between the exposed and control
12 animals. Results obtained from these two assays indicate that neither humoral immunity
13 (assessed by enumerating antibody-producing cells) nor cellular immunity (assessed by the
14 lymphocyte blast transformation assay) were markedly affected by the exposures.
15 Bice et al. (1985) evaluated whether or not exposure to diesel exhaust would alter
16 antibody immune responses induced after lung immunization of rats and mice. Exposures were
17 to 0.35,3.5, or 7.1 mg/m3 DPM for 7 h/day, 5 days/week for 24 mo. Chamber controls and
18 exposed animals were immunized by intratracheal instillation of sheep red blood cells (SRBC)
19 after 6, 12, 18, or 24 mo of exposure. No suppression in the immune response occurred in either
20 species. After 12, 18, and 24 mo of exposure, the total number of anti-SRBC IgM antibody
21 forming cells (AFCs) was elevated in rats, but not in mice, exposed to 3.5 or 7.1 mg/m3 DPM;
22 after 6 mo of exposure, only the 7.1 mg/m3 level was found to have caused this response in rats.
23 The number of AFC per 106 lymphoid cells hi lung-associated lymph nodes and the levels of
24 specific IgM, IgG, or IgA in rat sera were not significantly altered. The investigators concluded
25 that the increased cellularity and the presence of DPM in the lung-associated lymph nodes had
26 only a minimal effect on the immune and antigen filtration function of these tissues.
27 The effects of inhaled diesel exhaust and DPM have been studied in a murine model of
28 allergic asthma (Takano et al., 1998a,b). ICR mice were exposed for 12 h/day, 7days/week for
29 40 weeks to diesel exhaust (0.3,1.0, or 3.0 mg/m3). The mice were sensitized with ovalbumin
30 (OA) after 16 weeks exposure and subsequently challenged with aerosol allergen (1% OA in
31 isotonic saline for 6 min) at 3-week intervals during the last 24 weeks of exposure. Exposure to
32 diesel exhaust enhanced allergen-related eosinophil recruitment to the submucosal layers of the
33 airways and to the brcnchoalveclar space, and increased protein level s of granulocyte-colony
1175/99 5-56 DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-8. Effects of inhalation of diesel exhaust on the immune system of laboratory animals
Ul
O
O
Species/sex
Guinea Pig,
Hartley, M
Rat, F344, M
Rat, F344;
Mouse, CD-I
Mouse,
BALB/C, M
Exposure
period
20 h/day
5.5 days/week
4 or 8 weeks
7 h/day
5 days/week
52 or 104 weeks
7 h/day
5 days/week
104 weeks
12 h/day,
7 days/week,
Particles
(mg/m1)
1.5
0.19/iin
MOD
2.0
0.23-0.36
nmMDD
0.35
3.5
7.1
0.25 urn
MOD
3.0
6.0
CxT
(mg-h/m1)
660 or 7,280
3,640 or
7,280
1,274
12,740
25,480
756
1,512
CO
(ppm)
7.5
11.5
2.9
16.5
29.7
_
—
NO,
(ppm)
—
1.5
0.05
0.34
0.68
2.8
4.1
SO,
(ppm) Effects
— No alterations in numbers of B, T, and null
lymphocytes or cell viability among lymphocytes
isolated from tracheobronchial lymph nodes, spleen,
or blood
0.8 Neither humoral immunity (assessed by enumerating
antibody-producing cells) nor cellular immunity
(assessed by the lymphocyte blast transformation
assay) were markedly affected
— Total number of anti-sheep red blood cell IgM AFC
— in the lung-associated lymph nodes was elevated in
— rats exposed to 3.5 or 7.0 mg/m3 DPM (no such
effects in mice); total number of AFC per 106
' lymphoid cells in lung-associated lymph nodes and
level of specific IgM, IgG, or IgA in rat sera were not
altered
1.7 Spleen weights in mice exposed to dicsel exhaust
2.7 (6 mg/m3) increased significantly. Serum anti-OA
Study
Dziedzic
(1981)
Mentnech et
al. (1984)
Bice et al.
(1985)
Fujimaki et
al. (1997)
3 weeks
Mice administered OA
intranasally before,
immediately after, and
3 weeks after exposure
0.4 nm
IgE antibody tilers in mice exposed to 6 mg/m3
significantly higher than control. Antigen-stimulated
IL-4 and IL-10 production increased while IFN-g
production decreased significantly in spleen cells
from diesel exhaust-exposed (6 mg/m3) mice
stimulated with OA in vitro. Diesel exhaust
inhalation may affect antigen-specific IgE antibody
production through alteration of the cytokine
network.
O
O
G
O
H
m
Mouse,
C3H/Hen, M
12 h/day,
for 12 weeks. Before
exposure mice injected IP
with OA. After 3 weeks
and every 3 weeks
thereafter, mice
challenged with OA
aerosol.
1.0
3.0
1,008
3,024
1.42 0.87 Diesel exhaust + antigen challenge induced airway
4.02 1.83 hyperresponsiveness and inflamma lion with
increased eosinophils, mast cells, and goblet cells.
Diesel exhaust alone induced airway
hyperresponsiveness, but not eosinophilic infiltration
or increased goblet cells. Diesel exhaust inhalation
enhanced airway hyperresponsiveness and airway
inflammation caused by OA sensitization.
Miyabara et
al. (1998a)
-------
Ul
u\
00
Table 5-8. Effects of inhalation of diesel exhaust on the immune system of laboratory animals (continued)
Species/sex
Mouse,
C3H/HeN,
M
Exposure Particles
period .(mg/mj)
12h/dny, 3.0
for 5 weeks. After 7 days
mice injected IP with
OA. At end of exposure
mice challenged with OA
aerosol for IS minutes.
CxT CO NO, SO,
(mg-h/rn1) (ppm) (ppm) (ppm) Effects
1,260 — 4.08 1.26 Diesel exhaust alone increased neulrophils and
macrophages in BAL fluid; after diesel exhaust + OA
challenge eosinophils increased.
OA alone increased eosinophils but the increase was
enhanced by diesel exhaust.
Diesel exhaust + OA, but not diesel exhaust alone,
increased goblet cells, respiratory resistance,
production of OA-specific IgE and Igl in the serum,
and overexpression of IL-5 in lung tissue.
Study
Miyabara et
al.(1998b)
Mouse,
ICR
(murine model
of allergic
asthma)
12 h/day, 7days/week,
40 weeks.
After 16 weeks sensitized
to OA and challenged
with OA aerosol for
6 min, at 3-week
intervals during the last
24 weeks of exposure.
0.3
1.0
3.0
1,008
3,360
10,080
Diesel exhaust exposure enhanced allergen-related
recruitment to the submucosal layers of the airways
and the bronchoalveolar space, and increased GM-
CSF and IL-5 in the lung in a dose-dependent
manner. Increases in eosinophil recruitment and
local cytosine expression accompanied by goblet cell
proliferation in the bronchial epithelium and airway
hyperresponsiveness to inhaled acetylcholine Mice
exposed to clean air or DE without allergen
provocation showed no eosinophil recruitment to the
submucosal layers of the airways nor to the
bronchoalveolar space, and few goblet cells in the
bronchial epithelium. Daily inhalation of DE may
enhance allergen-related respiratory diseases such as
allergic asthma, and effect may be mediated by the
enhanced local expression of IL-S and GM-CSF.
Takanoet al.
(I998a)
DPM = Diesel paniculate matter.
AFC = Antibody-forming cells.
-------
1 stimulating factor (GM-CSF) and IL-5 in the lung hi a dose-dependent manner. In the diesel
2 exhaust-exposed mice, increases in eosinophil recruitment and local cytokine expression were
3 accompanied by goblet cell proliferation in the bronchial epithelium and airway
4 hyperresponsiveness to inhaled acetylcholine. In contrast, mice exposed to clean air or diesel
5 exhaust without allergen provocation showed no eosinophil recruitment to the submucosal layers
6 of the airways or to the bronchoalveolar space, and few goblet cells in the bronchial epithelium.
7 The authors concluded that daily inhalation of diesel exhaust can enhance allergen-related
8 respiratory diseases such as allergic asthma and this effect may be mediated by the enhanced
9 local expression of IL-5 and GM-CSF. The effects of DPM on a second characteristic of allergic
10 asthma, airway hyperresponsiveness, was examined by Takano et al. (1998b). Laboratory mice
11 were administered OA, DPM, or OA and DPM combined by intratracheal instillation for 6 wk.
12 Respiratory resistance (Rrs) after acetylcholine challenge was measured 24 h after the final
13 instillation. Rrs was significantly greater in the mice treated with OA and DPM than in the other
14 treatments. The authors concluded that DPM can enhance airway responsiveness associated with
15 allergen exposure.
16 In a series of inhalation studies following earlier instillation studies, Miyabara and
17 co-workers investigated whether inhalation of diesel exhaust could enhance allergic reactions in
18 laboratory mice. C3H/Hen mice were exposed to diesel exhaust (3 mg DPM/m3) by inhalation •
|9 for 5 weeks (Miyabara et al., 1998b) and, after 7 days of exposure, were sensitized to OA
0 injected intraperitoneally. At the end of the diesel exhaust exposure, the mice were challenged
21 with an OA aerosol for 15 minutes. Diesel exhaust caused an increase hi the numbers of
22 neutrophils and macrophages in bronchoalveolar lavage fluid independent of OA sensitization,
23 whereas a significant increase hi eosinophil numbers occurred only after diesel exhaust exposure
24 was combined with antigen challenge. While OA alone caused an increase hi eosinophil
25 numbers in lung tissue, this response was enhanced by diesel exhaust. Diesel exhaust exposure
26 combined with OA sensitization enhanced the number of goblet cells in lung tissue, respiratory
27 resistance, production of OA-specific IgE and Igl in the serum, and overexpression of IL-5 in
28 lung tissue. In a second study, C3H/Hen mice were sensitized with OA injected intraperitoneally
29 and then exposed to diesel exhaust by inhalation for 12 hours a day for 3 months (Miyabara et
30 al., 1998a). After 3 weeks of diesel exhaust exposure, and every 3 weeks thereafter, the mice
31 were challenged with an OA aerosol. Exposure to diesel exhaust with antigen challenge induced
32 airway hyperresponsiveness and airway inflammation, which was characterized by increased
33 numbers of eosinophils and mast cells in lung tissue. The increase in inflammatory cells was
34 accompanied by an increase in goblet cells in the bronchial epithelium. Airway
35 hyperresponsiveness, but not eosinophilic infiltration or increased goblet cells, was increased by
11/5/99 5-59 DRAFT—DO NOT CITE OR QUOTE
-------
1 diesel exhaust exposure alone. These workers concluded that inhalation of diesel exhaust can
2 enhance airway hyperresponsiveness and airway inflammation caused by OA sensitization in
3 mice.
4 The effects of diesel exhaust on IgE antibody production were investigated in BALB/c
5 mice sensitized with OA and exposed by inhalation to diesel exhaust (3.0 and 6.0 mg/m3) for 3
6 weeks (Fujimaki et al., 1997). The mice were sensitized by intranasal administration of OA
7 alone before, immediately after, and 3 weeks after diesel exhaust inhalation. While body and
8 thymus weights were unchanged in the diesel exhaust-exposed and control mice, spleen weights
9 in mice exposed to 6 mg/m3 diesel exhaust increased significantly. Anti-OA IgE antibody liters
10 in the sera of mice exposed to 6 mg/m3 diesel exhaust were significantly higher than control.
11 Total IgE and anti-OA IgG in sera from diesel exhaust-exposed and control mice remained
12 unchanged. Cytokine production was measured in vitro stimulated with OA in spleen cells from
13 mice exposed to diesel exhaust (6 mg/m3). Antigen-stimulated interleukin-4 (IL-4) and -10 (IL-
14 10) production increased significantly in vitro in spleen cells from diesel exhaust-exposed mice
15 compared to control, while interferon (IFN)-g production decreased markedly. The authors
16 concluded that diesel exhaust-inhalation in mice may affect antigen-specific IgE antibody
17 production through alteration of the cytokine network.
18
19 5.1.2.3.6. Effects on the immune system—noninhalation studies. The immune response of
20 laboratory animals to DPM has been studied in various non inhalation models and the results of
21 these studies are presented in Table 5-9. Takafuji et al. (1987) evaluated the IgE antibody
22 response of mice inoculated intranasally at intervals of 3 weeks with varying doses of a
23 suspension of DPM in ovalbumin. Antiovalbumin IgE antibody titers, assayed by passive
24 cutaneous anaphylaxis, were enhanced by doses as low as 1 /^g of particles compared with
25 immunization with ovalbumin alone.
26 The potential role of oxygen radicals in injury caused by DPM was investigated by
27 Sagai et al. (1996). These workers reported that repeated intratracheal instillation of DPM
28 (once/week for 16 weeks) in mice caused marked infiltration of inflammatory cells, proliferation
29 of goblet cells, increased mucus secretion, respiratory resistance, and airway constriction.
30 Eosinophils in the submucosa of the proximal bronchland medium bronchioles increased
31 eightfold following instillation. Eosinophil infiltration was significantly suppressed by
32 pretreatment with polyethyleneglycol-conjugated superoxide dismutase (PEG-SOD). Bound
33 sialic acid concentrations in bronchial alveolar lavage fluids, an index of mucus secretion,
34 increased with DPM, but were also suppressed by pretreatment with PEG-SOD. Goblet cell
35 hyperplasia, airway narrowing, and airway constriction also were observed with DPM.
11/5/99 5-60 DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-9. Effects of diesel particulate matter on the immune response of laboratory animals
Model
Treatment
Effects
Reference
TJ
H
O
O
2
O
H
O
I— t
a
O
O
H
M
Mouse,
BDFI, F
Mouse,
ICR, w/w-, M
Mouse,
A/J.M
Mouse,
DDF,, M
Mouse,
BALB/C,
nu/nu, F
Mouse,
BALB/cA, F
Intratracheal instillation of DPM, once/week
for 16 weeks
Mice immunized intranasally with Der f II +
pyrene, or Der f II + DPM 7 times at 2-week
intervals
Mice were administered 25 mg of each of
5 fine particles (Kanto loam dust, fly ash, CB,
DPM, and aluminum hydroxide [alum])
intranasally and exposed to aerosolized
Japanese cedar pollen allergens (JCPA) for
intervals up to 18 wk.
Inoculated OA with DPM or CB into hind
footpad measured response using popliteal
lymph node assay
Intranasal administration of DPM. Mice
immunized with OA or OA combined with
DPMorCB.
Intranasally delivered doses of DPM as low as 1 mg exerted an adjuvant activity for IgE
antibody production.
Infiltration of inflammatory cells, proliferation of goblet cells, increased mucus secretion,
respiratory resistance, and airway constriction. Increased eosinophils in the submucosa of the
proximal bronchi and medium bronchioles. Eosinophil infiltration suppressed by pretreatment
with PEG-SOD. Bound sialic acid, an index of mucus secretion, in bronchial alveolar lavage
fluids increased, but was suppressed by PEG-SOD. Increased respiratory resistance suppressed
by PEG-SOD. Oxygen radicals produced by instilled DPM may cause features characteristic of
bronchial asthma in mice.
IgE antibody responses to Der f II enhanced in mice immunized with Der f 11+ pyrene or Der f
II + DPM compared with Der f II alone. Response was dose related. DPM and pyrene
contained in DPM have adjuvant activity on IgE and IgGl antibody production in mice
immunized with house dust mite allergen.
Measurements were made of JCPA-specific IgE and IgG antibody liters, the protein-adsorbing
capacity of each type of particle, and nasal rubbing movements (a parameter of allergic rhinitis
in mice). The increases in anti-JPCA IgE and IgG antibody liters were significantly greater in
mice treated with particles and aerosolized JCPA than in mice treated wilh aerosolized JCPA
alone. In a subsequenl experiment, the mice received the particles as before, but about 160,000
grains of Japanese cedar pollen (JCP) were dropped onto the tip of the nose of each mouse
twice a week for 16 wk. After 18 wk Ihere were no significant differences in the anli-JCPA IgE
and IgG production, nasal rubbing, or hislopathological changes. The workers concluded that .
the nature of the particle, ihe ability of Ihe particle lo absorb anligens, and/or particle size is nol
related to the enhancement of IgE antibody production or symptoms of allergic rhinitis.
However, IgE antibody production did appear to occur earlier in mice treated with particles
than in mice immunized with allergens alone.
Increased response (increased weight, cell numbers, cell proliferation) and longer response
observed with DPM and OA, compared to DPM or OA alone. Response was specific and not
an unspecific inflammatory response. CB was slightly less potent than DPM. Nonextractable
carbon core contributes substantially to adjuvant activity of DPM.
Increased response to antigen in animals receiving DPM or CB. Increased number of
responding animals and increased serum anti OA IgE antibody. Both DPM and CB have
adjuvant activity for IgE production. DPM response more pronounced than CB, indicating
both organic matter adsorbed to DPM and the nonextractable carbon core responsible for
adjuvant activity.
Takafujietal. (1987)
Sagaietal. (1996)
Suzuki etal. (1996)
Maejima et at. (1997)
Leviketal. (1997)
Nilsenetal. (1997)
-------
~ Table 5-9. Effects of diesel particulate matter on the immune response of laboratory animals, (continued)
"••- — .... , . i _———
^g Model Treatment Effects Reference
Mouse, Intralracheal instillation of OA, DPM, or OVA Respiratory resistance (Rrs) measured 24 h after the final instillation. Rrs after acelylcholinc Takano et al. (I998b)
ICR, M and DPM combined, once/week for 6 wk. challenge was significantly greater in the mice treated with OVA and DPM than other
treatments. DPM can enhance airway responsiveness associated with allergen exposure.
OA- Ovalbumin
'DPM- diesel particulate matter
CB- carbon black
PEG-SOD- polyethyleneglycol-cpnjugated superoxide dismutase
IL-4-interleukin-4
IL-S- interleukin-S
IL-10-interleukin-10
\FH- interferon-g
GM-CSF -grunulocyte-colony stimulating factor
IP-intraperitoneally
S)
O
o
§
H
I
O
-------
1 Respiratory resistance to acetylcholine in the DPM-group was 11 times higher than in
2 controls, and the increased resistance was significantly suppressed by PEG-SOD pretreatment.
These findings indicate that oxygen radicals, caused by intratracheally instilled DPM elicits
responses characteristic of bronchial asthma.
5 Potential adjuvant effects of DPM on the response to the model allergen OA were
6 investigated in BALB/c mice using the popliteal lymph node (PLN) assay (L0vik et al., 1997).
7 DPM inoculated together with OA into one hind footpad gave a significantly augmented
8 response (increase in weight, cell numbers, and cell proliferation) in the draining popliteal lymph
9 node as compared to DPM or OA alone. The duration of the local lymph node response was also
10 longer when DPM was given with the allergen. The lymph node response appeared to be of a
11 specific immunologic character and not an unspecific inflammatory reaction. The OA-specific
12 response IgE was increased in mice receiving OA together with DPM as compared to the
13 response in mice receiving OA alone. Further studies using carbon black (CB) as a surrogate for
14 the nonextractable core of DPM found that while CB resembled DPM in its capacity to increase
15 the local lymph node response and serum-specific IgE response to OA, CB appeared to be
16 slightly less potent than DPM. The results indicate that the nonextractable particle core
17 contributes substantially to the adjuvant activity of DPM.
18 Nilsen et al. (1997) investigated which part of the particle was responsible, the carbon
*core and/or the adsorbed organic substances, for the adjuvant activity of DPM. Female Balb/cA
mice were immunized with OA alone or in combination with DPM or CB particles by intranasal
21 administration. There was an increased response to the antigen in animals receiving OA together
22 with DPM or CB, compared with animals receiving OA alone. The response was seen as both an
23 increased number of responding animals and increased serum antiOA IgE response. The
24 workers concluded that both DPM and CB have an adjuvant activity for specific IgE production,
25 but that the activity of DPM may be more pronounced than that of CB. The results suggest that
26 both the organic matter adsorbed to DPM and the non-extractable carbon are responsible for the
27 observed adjuvant effect.
28 Maejima et al. (1997) examined the potential adjuvant activity of several different fine
29 particles. These workers administered 25 jj.g of each of 5 particles (Kanto loam dust, fly ash,
30 carbon black, DPM, and aluminum hydroxide [alum]) intranasally in mice and exposed them to
31 aerosolized Japanese cedar pollen allergens (JCPA) for intervals up to 18 weeks. Measurements
32 were made of JCPA-specific IgE and IgG antibody titers, the protein-adsorbing capacity of each
33 type of particle, and nasal rubbing movements (a parameter of allergic rhinitis in mice). The
34 increases in anti-JPCA IgE and IgG antibody titers were significantly greater in mice treated with
35 particles and aerosolized JCPA than in mice treated with aerosolized JCPA alone. In a
11/5/99 5-63 DRAFT—DO NOT CITE OR QUOTE
-------
1 subsequent experiment, the mice received the particles as before, but about 160,000 grains of
2 Japanese cedar pollen (JCP) were dropped onto the tip of the nose of each mouse twice a week
3 for 16 wk. After 18 wk there were no significant differences in the anti-JCPA IgE and IgG
4 production, nasal rubbing, or histopathological changes. The workers concluded that the nature
5 of the particle, the ability of the particle to absorb antigens, and/or particle size is not related to
6 the enhancement of IgE antibody production or symptoms of allergic rhinitis. However, IgE
7 antibody production did appear to occur earlier in mice treated with particles than in mice
8 immunized with allergens alone.
9 Suzuki et al. (1996) investigated the effect of pyrene on IgE and IgGl antibody
10 production in mice to clarify the relation between mite allergy and adjuvancy of the chemical
11 compounds in DPM. The mite allergen was Der f II, one of the major allergens of house dust
12 mite (Dermatophagoid.es farinae). Allergen mice were grouped and immunized with Der f II (5
13 ,ug), Der f II (5 >ug) plus pyrene (200 /ug) and Der f II (5 /wg) plus DPM (100 //g) intranasally
14 seven times at 2-week intervals. The separate groups of mice were also immunized with Der f II
15 (10 ,ug) plus the same dose of adjuvants in the same way. The IgE antibody responses to Der f II
16 in mice immunized with Der f II plus pyrene or Der f II plus DPM were markedly enhanced
17 compared with those immunized with Der f II alone. The anti-Der f II IgE antibody production
18 increased with increasing the dose of Der f II from 5 /^g to 10 /zg in mice immunized with
19 Der f II plus the same dose of adjuvants. The IgGl antibody responses to Der f II in mice
20 immunized with Der f II (10 jug) plus pyrene (200 //g) or Der f II (10 /ug) plus DPM (100 ^ug)
21 were greater than those immunized with 10 /ug of Der f II alone. In addition, when peritoneal
22 macrophages obtained from normal mice were incubated with pyrene or DPM in vitro, an
23 enhanced IL-la production by the macrophages was observed. When spleen lymphocytes
24 obtained from the mice immunized with Der f II (10 yug) plus DPM (100 fj.g) or Der f II (10 //g)
25 plus pyrene (200 /ug) were stimulated with 10 ^g of Der f II in vitro, an enhanced IL-4
26 production of the lymphocytes was also observed compared with those immunized with Der f II
27 alone. This study indicates that DPM and pyrene contained in DPM have an adjuvant activity on
OO T«,T7 r~'-.'J T~O.t *~~+4-l\*r*A-*T ^*v>*4-."*^*i^^ ?*^ t^*?**^ '«V1^»0.,«~£'V'~,*1 ••»+*****-f>^*-11Tr T«n*t* ri l^^.-,-,^^, Am~n+ «.-^+~,
29 allergen.
30 Ormstad et al. (1998) investigated the potential for DPM, as well as other suspended
31 particulate matter (SPM) to act as carriers for allergens into the airways. These investigators
32 found both Can f 1 (dog) and Bet v 1 (birch pollen) on the surface of SPM collected in air from
33 different homes. In an extension of the study they found that DPM had the potential of binding,
34 in vitro, both ot these allergens as well as Fel d 1 (cat) and JUer p 1 (house mite). The authors
35 conclude that soot particles in indoor air house dust may act as carrier of several allergens in
36 indoor air.
11/5/99 5-64 DRAFT—DO NOT CITE OR QUOTE
-------
1 Knox et al. (1997) investigated whether free grass pollen allergen molecules, derived
2 from dead or burst grains and dispersed in microdroplets of water in aerosols, can bind to DPM
in air. Using natural highly purified Lol p 1, immunogold labeling with specific monoclonal
antibodies, and a high-voltage transmission electron-microscopic imaging technique, these
5 workers demonstrated, binding of the major grass pollen allergen, Lol p 1, to DPM in vitro.
6 These workers conclude that binding of DPM with Lol p 1 might be a possible mechanism by
7 which allergens can become concentrated in air and trigger attacks of asthma.
8 The inhalation of diesel exhaust appeared to have minimal effects on the immune status
9 of rats and guinea pigs. Conversely, intranasally delivered doses as low as 1 /ug of DPM exerted
10 an adjuvant activity for IgE antibody production in mice. Further studies of the effects of diesel
11 exhaust on the immune system are needed to clarify the impact of such variables as route of
12 exposure, species, dose, and atopy.
13
14 5.1.2.3.7. Effects on the liver. Meiss et al. (1981) examined alterations in the hepatic
15 parenchyma of hamsters by using thin-section and freeze-fracture histological techniques.
16 Exposures to diesel exhaust were for 7 to 8 h/day, 5 days/week, for 5 mo at about 4 or 11 mg/m3
17 DPM. The livers of the hamsters exposed to both concentrations of diesel exhaust exhibited
18 moderate dilatation of the sinusoids, with activation of the Kupffer cells and slight changes in the
»cell nuclei. Fatty deposits were observed in the sinusoids, and small fat droplets were
occasionally observed in the peripheral hepatocytes. Mitochondria often had a loss of cristae and
21 exhibited a pleomorphic character. Giant microbodies were seen in the hepatocytes, which were
22 moderately enlarged, and gap junctions between hepatocytes exhibited a wide range in structural
23 diversity. The results of this study and others on the effect of exposure of diesel exhaust on the
24 liver of laboratory animals are summarized in Table 5-10.
25 Green et al. (1983) and Plopper et al. (1983) reported no changes in liver weights of rats
26 exposed to 2 mg/m3 DPM for 7 h/day, 5 days/week for 52 weeks or of cats exposed to 6 to
27 12 mg/m3, 8 h/day, 7 days/week for 124 weeks. The use of light and electron microscopy
28 revealed that long-term inhalation of varying high concentrations of diesel exhaust caused
29 numerous alterations to the hepatic parenchyma of guinea pigs. A less sensitive index of liver
30 toxicity, increased liver weight, failed to detect an effect of diesel exhaust on the liver of the rat
31 and cat following long-term exposure to diesel exhaust. These results are too limited to
32 understand potential impacts on the liver.
33
34
35
11/5/99 5-65 DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-10. Effects of exposure to diesel exhaust on the liver of laboratory animals
Species/sex
Rat, F344.M.F
Hamster, Syrian
Cat, inbred, M
Exposure
period
7 h/day
5 days/week
52 weeks
7-8 h/day
5 days/week
22 weeks
8 h/day
7 days/week
124 weeks
Particles
(mg/m1)
2.0
0.23-0.36 A
-------
5.1.2.3.8. Blood and cardiovascular systems. Several studies have evaluated the effects of
diesel exhaust exposure on hematological and cardiovascular parameters of laboratory animals.
These studies are summarized in Table 5-11. Standard hematological indices of toxicological
effects on red and white blood cells failed to detect dramatic and consistent responses.
Erythrocyte (RBC) counts were reported as being unaffected in cats (Pepelko and Peirano, 1983),
rats and monkeys (Lewis et al., 1989), guinea pigs and rats (Penney et al., 1981), and rats
(Karagianes et al., 1981); lowered in rats (Heinrich et al., 1982); and elevated in rats (Research
Committee for HERP Studies, 1988; Brightwell et al., 1986). Mean corpuscular volume was 0
significantly increased in monkeys, 69 versus 64 (Lewis et al., 1989), and hamsters (Heinrich et
al., 1982) and lowered in rats (Research Committee for HERP Studies, 1988). The only other
parameters of erythrocyte status and related events were lowered mean corpuscular hemoglobin
and mean corpuscular hemoglobin concentration in rats (Research Committee for HERP Studies,
1988), a 3% to 5% increase in carboxyhemoglobin saturation in rats (Karagianes et al., 1981),
and a suggestion of an increase in prothrombin time (Brightwell et al., 1986). The biological
significance of these findings regarding adverse health effects is deemed to be inconsequential.
Three investigators (Pepelko and Peirano, 1983; Lewis et al., 1989; Brightwell et al.,
1986) reported an increase in the percentage of banded neutrophils in cats and rats. This effect
was not observed in monkeys (Lewis et al., 1989). The health implications of an increase in
abnormal maturation of circulating neutrophils are uncertain but indicate a toxic response of
leukocytes following exposures to diesel exhaust. Leukocyte counts were reported to be reduced
in hamsters (Heinrich et al., 1982); increased in rats (Brightwell et al., 1986); and unaffected in
cats, rats, and monkeys (Pepelko and Peirano, 1983; Research Committee for HERP Studies,
1988; Lewis et al., 1989). These inconsistent findings indicate that the leukocyte counts are
more indicative of the clinical status of the laboratory animals than any direct effect of exposure
to diesel exhaust.
An important consequence of particle retention in the lungs of exposed subjects can be
the development of pulmonary hypertension and cor pulmonale. Such pathology usually arises
from pulmonary fibrosis or emphysema obliterating the pulmonary vascular bed or by chronic
hypoxia. No significant changes in heart mass were found in guinea pigs or rats exposed to
diesel exhaust (Wiester et al., 1980; Penney et al., 1981; Lewis et al., 1989). Rats exposed to
diesel exhaust showed a greater increase in the medial wall thickness of pulmonary arteries of
differing diameters and right ventricular wall thickness; these increases, however, did not achieve
statistically significant levels (Vallyathan et al., 1986). Brightwell et al. (1986) reported
increased heart/body weight and right ventricular/heart weight ratios and decreased left
ventricular contractility in rats exposed to 6.6 mg/m3 DPM for 16 h/day, 5 days/week for
104 weeks.
11/5/99 5-67 DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-11. Effects of exposure to diesel exhaust on the hematological and cardiovascular systems of laboratory animals
^*
yi
s
L/
o
j\
^
»
3
>
T)
H
^
J
D
z!
H
2
j
3
•N
0
D
-4
3
rt
Species/sex
Monkey,
Cynomolgus, M
Rat, F344, M. F
Guinea Pig,
Hartley, M, F
Hamster, Syrian,
M, F
Rat, F344;
Guinea Pig,
Hartley
Rat, Wistar, M
Rat, F3444/JC),
M,F
Rat, F344
Cat, Inbred, M
Exposure
period
7 h/day
5 days/week
104 weeks
7 h/day
5 days/week
104 weeks
20 h/day
7 days/week
8 weeks
7-8 h/day
5 days/week
75 weeks
20 h/day
5.5 days/week
78 weeks
6 h/day
5 days/week
78 weeks
16 h/day
6 days/week
130 weeks
16 h/day
5 days/week
104 weeks
8 h/day
7 days/week
124 weeks
"Nonirradiated diesel exhaust.
blrr»HintpH rlipcpl pvltnnct
Particles
(mg/mj)
2
0.23-0.36 i^m MOD
2
0.23-0.36 (um MOD
6..T
6.8b
3.9
0.1/itnMDD
0.25
0.75
1.5
8.3
0.71 Mm MOD
0.11°
0.41'
1.08°
2.3 lc
3.72d
0.1 //m MOD
07
22
6.6
6.0'
12.0f
'Heavy-duty engine.
ct In fit w<>i>k« nf f.vm
C*T
(mg-h/rn3)
7,280
7,280
7,056
7,616
10,238-11,700
2,145
6,435
12,870
19,422
1,373
5,117
13,478
28,829
46,426
5,824
18,304
54,912
41,664
83,328
ttnre
CO
(ppm)
11.5
11.5
17.4
16.7
18.5
3.0
4.8
6.9
50.0
1.23
2.12
3.96
7.10
12.9
—
—
32.0
20.2
33.3
NO,
(ppm)
1.5
1.5
2.3
2.9
1.2
0.11
0.27
0.49
4-6
0.08
0.26
0.70
1.41
3.00
—
—
—
2.7
4.4
SO,
(ppm)
0.8
0.8
2.1
1.9
3.1
—
—
—
_
0.38
1.06
2.42
4.70
4.57
—
—
—
2.1
5.0
Effects
Increased MCV
Increase in banded neutrophils; no effect on
heart or pulmonary arteries
No effect on heart mass or ECG; small
decrease in heart rate (IE only)
At 29 weeks, lower erythrocyte count;
increased MCV; reduced leukocyte count
No changes in heart mass or hematology at
any exhaust level or duration of exposure in
either species
3% increase in COHb
At higher concentrations, RBC, Hb, Hct
slightly elevated; MCV and mean
corpuscular hemoglobin and concentration •
were lowered
Increases in RBC, Hb, Hct, and WBC,
primarily banded neutrophils; suggestion of
an increase in prothrombin time; increased
heart/body weight and right
ventricular/heart ratios and decreased left
ventricular contractility in 6.6 nig/m' group
Increases in banded neutrophils; significant
at 12 mo, but not 24 mo
Study
Lewis etal. (1989)
Lewis etal. (1989)
Vallyathan et al.
(1986)
Wiester etal. (1980)
Heinrich et al.
(1982)
Penney etal. (1981)
Karagianes ct al.
(1981)
Research Committee
for HERP Studies
(1988)
Brightwell et al.
(1986)
Pepelko and Peirano
(1983)
"Light-duty engine.
Key: MCV = Mean corpuscular volume.
'62 to 124 weeks of exposure.
-------
1 The effects of DPM on the endothelium-dependent relaxation (EDR) of vascular smooth
2 muscle cells has been investigated (Ikeda et al., 1995, 1998). Incubation of rat thoracic aortae
with suspensions of DPM (10-100 ,wg/mL) markedly attenuated acetyicholine-induced EDR.
The mechanism of this effect was studied further in cultured porcine endothelial cells (CPE).
5 A 10-min incubation of PEC with DPM (0.1-100 /ug/mL) inhibited endothelium-dependent
6 relaxing factor (EDRF) or nitric oxide (NO) release. A 10-min incubation of DPM with NO
7 synthase inhibited formation of NO2', a product of NO metabolism. The authors concluded that
8 DPM, at the concentrations tested, neither induced cell damage nor inhibited EDRF release from
9 PEC, but scavenged and thereby blocked the physiological action of NO.
10
11 5.1.2.3.9. Serum chemistry. A number of investigators have studied the effects of exposure to
12 diesel exhaust on serum biochemistry and no consistent effects have been found. Such studies
13 are summarized in Table 5-12.
14 The biological significance of changes in serum chemistry in female but not male rats
15 exposed at 2 mg/m3 DPM for 7 h/day, 5 days/week for 104 weeks (Lewis et al., 1989) is difficult
16 to interpret. Not only were the effects noted hi one sex (females) only, but the serum enzymes,
17 lactate dehydrogenase (LDH), serum glutamic-oxaloacetic transaminase (SGOT), and serum
18 glutamic-pyruvic transaminase (SGPT), were elevated in the control group, a circumstance
« contrary to denoting organ damage in the exposed female rats. The elevations of liver-related
serum enzymes in the control versus the exposed female rats appear to be a random event among
21 these aged subjects. The incidence of age-related disease, such as mononuclear cell leukemia,
22 can markedly affect such enzyme levels, seriously compromising the usefulness of a comparison
23 to historical controls. The serum sodium values of 144 versus 148 mmol/L in control and
24 exposed rats, respectively, although statistically different, would have no biological import.
25 The increased serum enzyme activities, alkaline phosphatase, SGOT, SGPT,
26 gamma-glutamyl transpeptidase, and decreased cholinesterase activity suggest an impaired liver;
27 however, such an impairment was not established histopathologically (Heinrich et al., 1982;
28 Research Committee for HERP Studies, 1988; Brightwell et al., 1986). The increased urea
29 nitrogen, electrolyte levels, and gamma globulin concentration and reduction in total blood
30 proteins are indicative of impaired kidney function. Again there was no histopathological
31 confirmation of impaired kidneys in these studies.
32 Clinical chemistry studies suggest impairment of both liver and kidney functions in rats
33 and hamsters chronically exposed to high concentrations of diesel exhaust. The absence of
34 histopathological confirmation, the appearance of such effects near the end of the lifespan of the
35 laboratory animal, and the failure to find such biochemical changes in cats exposed to a higher
1175/99 5-69 DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-12. Effects off chronic exposures to diesel exhaust on serum chemistry of laboratory animals
5)
0
0
Species/sex
Rat,F344,M,F
Exposure
peiriod
7 h/day
5 days/week
Particles
(nig/m1)
2.0
0.23
C*T
(mg-h/m1)
7,280
CO NO, SO,
(ppm) (ppm) (ppm) Effects
11.5 1.5 0.8 Decreased phosphate, LDH, SCOT, and SGPT;
increased sodium in females but not males
Study
Lewis el al.
(1989)
I04weaks 0 36 ;jtr MOD
j\
^
3
Jd
n
H
3
D
z!
D
3
fl
D
id
D
— I
3
Hamster, Syrian, M,
Rat, F344/JC!., M, F
Rat, F344; Hamster,
Syrian
Cat inbred, M
•Light-duty engine.
'Heavy-duty engine.
F 7-8 h/day
5 days/week (
75 weeks
1 6 h/day
6 days/week
130 weeks
C
16 h/day
3 days/week
104 weeks
8 h/day
7 days/week
124 weeks
3.0
.1 ^mMDD
0.11"
0.41'
1.08'
2.31'
3.72"
.19-0.28 A
-------
1 dose, however, tend to discredit the probability of hepatic and renal hazards to humans exposed
2 at atmospheric levels of diesel exhaust.
5.1.2.3.10. Effects on microsomal enzymes. Several studies have examined the effects of diesel
5 exhaust exposure on microsomal enzymes associated with the metabolism and possible
6 activation of xenobiotics, especially polynuclear aromatic hydrocarbons. These studies are
7 summarized in Table 5-13. Lee et al. (1980) measured the activities of aryl hydrocarbon
8 hydroxylase (AHH) and epoxide hydrase (EH) in liver, lung, testis, and prostate gland of adult
9 male rats exposed to 6.32 mg/m3 DPM 20 h/day for 42 days. Maximal significant AHH
10 activities (pmol/min/mg microsomal protein) occurred at different times during the exposure
11 period, and differences between controls and exposed rats, respectively, were as follows:
12 prostate 0.29 versus 1.31, lung 3.67 versus 5.11, and liver 113.9 versus 164.0. There was no
13 difference in AHH activity in the testis between exposed and control rats. Epoxide hydrase
14 activity was not significantly different from control values for any of the organs tested.
15 Pepelko and Peirano (1983) found no statistical differences in liver microsomal cytochrome
16 P448-450 levels and liver microsomal AHH between control and diesel-exposed mice either at
17 6 and 8 mo of exposure. Small differences were noted in the lung microsomal AHH activities,
18 but these were believed to be artifactual differences, due to increases in nonmicrosomal lung .
protein present hi the microsomal preparations. Exposures to 6 mg/m3 DPM were for 8 h/day,
7 days/week.
21 Rabovsky et al. (1984) investigated the effect of chronic exposure to diesel exhaust on
22 microsomal cytochrome P450-associated benzo[a]pyrene hydroxylase and 7-ethoxycoumarin
23 deethylase activities in rat lung and liver. Male rats were exposed for 7 h/day, 5 days/week for
24 104 weeks to 2 mg/m3 DPM. The exposure had no effect on B[a]P hydroxylase or
25 7-ethoxycoumarin deethylase activities in lung or liver. In related studies, Rabovsky et al.
26 (1986) examined the effects of diesel exhaust on vitally induced enzyme activity and interferon
27 production in female mice. The mice were exposed for 7 h/day, 5 days/week for 1 month to
28 diesel exhaust diluted to achieve a concentration of 2 mg/m3 DPM. After the exposure, the mice
29 were inoculated intranasally with influenza virus. Changes hi serum levels of interferon and
30 liver microsomal activities of 7-ethoxycoumarin, ethylmorphine demethylase, and nicotinamide
31 6 and 8 mo of exposure. Small differences were noted in the lung microsomal AHH activities,
32 but these were believed to be artifactual differences, due to increases hi nonmicrosomal lung
33 protein present in the microsomal preparations. Exposures to 6 mg/m3 DPM were for 8 h/day,
34 7 days/week.
35
11/5/99 5-71 DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-13. Effects of chronic exposures to diesel exhaust on microsomal enzymes of laboratory animals
5i
B
o
_j)
i
-j
o
3
>
T)
J
D
-?
— 1
H
-)
-H
1
t
3
3
•d
Exposure
Species/sex period
Rat, 044, m -
Mouse, cd-1, f 7 h/day
5 days/week
4 weeks
Rat, sprague- 20 h/day
dawley, m 7 days/week
1 -7 weeks
Ret, O44, m 20 h/day
5.5 days/week
4, 13, 26, or
39 wesks
20 h/day
5.5 days/week
4, 13, 26, or
39 weeks
Rat,.O44,f 7 h/day
5 dayc/week
12, 26, or
104 weeks
Rat, f344, m 20 h/day
5.5 days/week
8-53 weeks
Mouse, a/j, m 8 days/week
7 days/week
26 or 35 weeks
Ahh = ary . hydiocarbon hydroclase.
B[a]p = her zo[a]pyrene.
1'articles C x t Co
(mg/m1) (mg-h/m1) (ppm)
.. — —
2.0 280 11.5
0.2 -0.36 Minmdd
6.3 882-6,174 17.4
0.75 330-6,435 4.8
1.5 7.5
0. .9/,'.m mdd
0.75 330-6,435 4.8
1.5 7.5
0. i9;«nmdd
2.0 840-7,280 11.5
0.23' 0.36 A-rn mdd
0.21 220-8,745 2.9
1.5 7.5
0.19pm mdd
6.0 17.4 17.4
No, So,
(ppm) (ppm) Effects
— — Intratracheal administration of dpm extract required
doses greater than 6 mg/m1 before the lung ahh was
barely doubled; liver ahh activity was unchanged
1.5 0.8 Mice inoculated intranasally with influenza virus had
smaller increases in ethylmorphine demethylase
activity on days 2 to 4 postvirus infection and abolition
of day 4 postinfection increase in nadph-clependent
cytochrome c reductase
2.3 2. 1 Ahh induction occurred in lung, liver, and prostate
gland but not in testes; maximum significant activities
occurred at different times; liver has greatest overall
activity, percent increase highest in prostate; expoxide
hydrase activity was unaffected
— — Inhalation exposure had no significant effect on liver
— — ahh activity; lung ahh activity was slightly reduced
after 6-mo exposure to 1.5 mg/m5 dpm; an ip dose of
dp extract, estimated to be equivalent to inhalation
— — exposure, had no effect on ahh activity in liver and
— — lungs; cyt. P-50 was unchanged in lungs and liver
following inhalation or ip administration
1.5 0.8 No effect on b[o]p hydrolase or 7-exthoxycoumarin
deethylase activities in the liver
— — After 8 weeks, no induction of cyt. P-450, cyt. P-448,
— — or nadph-dependent cyt. c reductase; after 1 year of
exposure, liver microsomal oxidation of b[a]p was not
increased; 1 year of exposure to either 0.25 or
1.5 mg/m1 dpm impaired lung microsomal metabolism
ofb[a]p
2.3 2.1 No differences in lung and liver ahh activities and liver
p-448, p-450 levels
Study
Chen (1986)
Rabovsky el al. (1986)
Lee etal. (1980)
Chen and vostal ( 1981)
Rabovsky etal. (1984)
Navarro etal. (1981)
Pepelko and Peirano
(1983)
-------
1 Rabovsky et al. (1984) investigated the effect of chronic exposure to diesel exhaust on
2 microsomal cytochrome P450-associated benzo[a]pyrene hydroxylase and 7-ethoxycoumarin
3 deethylase activities in rat lung and liver. Male rats were exposed for 7 h/day, 5 days/week for 104
4 weeks to 2 mg/m3 DPM. The exposure had no effect on B[a]P hydroxylase or 7-ethoxycoumarin
5 deethylase activities in lung or liver. In related studies, Rabovsky et al. (1986) examined the effects
6 of diesel exhaust on vitally induced enzyme activity and interferon production in female mice. The
7 mice were exposed for 7 h/day, 5 days/week for 1 month to diesel exhaust diluted to achieve a
8 concentration of 2 mg/m3 DPM. After the exposure, the mice were inoculated intranasally with
9 influenza virus. Changes in serum levels of interferon and liver microsomal activities of 7-
10 ethoxycoumarin, ethylmorphine demethylase, and nicotinamide adenine dinucleotide phosphate
11 (NADPH)-dependent cytochrome c reductase were measured. In the absence of viral inoculation,
12 exposure to diesel exhaust had no significant effects on the activity levels of the two liver
13 microsomal monooxygenases and NADPH-dependent cytochrome c reductase. Exposure to diesel
14 exhaust produced smaller increases in ethylmorphine demethylase activity on days 2 to 4 postvirus
15 infection and also abolished the day 4 postinfection increase in NADPH-dependent cytochrome
16 c reductase when compared with nonexposed mice. These data suggested to the authors that the
17 relationship that exists between metabolic detoxification and resistance to infection in unexposed
18 mice was altered during a short-term exposure to diesel exhaust.
19 Chen and Vostal (1981) measured the activity of AHH and the content of cytochrome P450 in
the lungs and livers of rats exposed by inhalation or intraperitoneal (i.p.) injection of a
21 dichloromethane extract of DPM. In the inhalation exposures, the exhaust was diluted to achieve
22 concentrations of 0.75 or 1.5 mg/m3 DPM, and the exposure regimen was 20 h/day, 5.5 days/week
23 for up to 9 mo. The concentration of total hydrocarbons and particle-phase hydrocarbons was not
24 reported. Parenteral administration involved repeated i.p. injections at several dose levels for 4
25 days. Inhalation exposure had no significant effect on liver microsomal AHH activity; however,
26 lung AHH activity was slightly reduced after 6 mo exposure to 1.5 mg/m3. An i.p. dose of DPM
27 extract, estimated to be equivalent to the inhalation exposure, had no effect on AHH activity in liver
28 or lungs. No changes were observed in cytochrome P450 contents in lungs or liver following
29 inhalation exposure or i.p. treatment. Direct intratracheal administration of a dichloromethane
30 DPM extract required doses greater than 6 mg/kg body weight before the activity of induced AHH
31 in the lung was barely doubled; liver AHH activity remained unchanged (Chen, 1986).
32 In related studies, Navarro et al. (1981) evaluated the effect of exposure to diesel exhaust on
33 rat hepatic and pulmonary microsomal enzyme activities. The same exposure regimen was
34 employed (20 h/day, 5.5 days/week, for up to 1 year), and the exhaust was diluted to achieve
35 concentrations of 0.25 and 1.5 mg/m3 DPM (a few studies were also conducted at 0.75 mg/m3).
After 8 weeks of exposure, there was no evidence for the induction of cytochrome P450,
11/5/99 5-73 DRAFT—DO NOT CITE OR QUOTE
-------
1 cytochrome P448, or NADPH-dependent cytochrome c reductase in rat liver microsomes. One year
2 of exposure had little, if any, effect on the hepatic metabolism of B[a]?. However, 1 year
3 of exposure to 0.25 and 1 .5 mg/m3 significantly impaired the ability of lung microsomes to
4 metabolize B[a]P (0. 1 5 and 0.02 nmole/30 min/mg protein, respectively, versus
5 0.32 nmole/30 min/mg protein for the controls).
6 There are conflicting results regarding the induction of microsomal AHH activities in the
7 lungs and liver of rodents exposed to diesel exhaust. One study reported induced AHH activity in
8 the lungs, liver, and prostate of rats exposed to diesel exhaust containing 6.32 mg/m3 DPM for 20
9 h/day for 42 days; however, no induction of AHH was observed in the lungs of rats and mice
1 0 exposed to 6 mg/m3 DPM for 8 h/day, 7 days/week for up to 8 mo or to 0.25 to 2 mg/m3 for periods
11 up to 2 years. Exposure to diesel exhaust has not been shown to produce adverse effects on
1 2 microsomal cytochrome P450 in the lungs or liver of rats or mice. The weight of evidence suggests
1 3 that the absence of enzyme induction in the rodent lung exposed to diesel exhaust is caused either
1 4- by the unavailability of the adsorbed hydrocarbons or by their presence in insufficient quantities for
1 5 enzyme induction.
16
1 7 5.1.2.3.11. Effects on behavior and neurophysiology. Studies on the effects of exposure to diesel
1 8 exhaust on the behavior and neurophysiology of laboratory animals are summarized in Table 5-14.
1 9 Laurie et al. (1978) and Laurie et al. (1980) examined behavioral alterations in adult and neonatal
20 rats exposed to diesel exhaust. Exposure for 20 h/day, 7 days/week, for 6 weeks to exhaust
2 1 containing 6 mg/m3 DPM produced a significant reduction in adult spontaneous locomotor activity
22 (SLA) and in neonatal pivoting (Laurie et al., 1 978). In a follow-up study, Laurie et al. (1980)
23 found that shorter exposure (8 h/day) to 6 mg/m3 DPM also resulted in a reduction of SLA in adult
24 rats. Laurie et al. (1980) conducted additional behavioral tests on adult rats exposed during their
25 neonatal period. For two of three exposure situations (20 h/day for 1 7 days postparturition, or 8
26 h/day for the first 28 or 42 days postparturition), significantly lower SLA was observed in the
27 majority of the tests conducted on the adults after week 5 of measurement, "when compared with
ilj UU V JL\JL X / tOClY i3^ ttiOW
29 exhibited a significantly slower rate of acquisition of a bar-pressing task to obtain food. The
30 investigators noted that the evidence was insufficient to determine whether the differences were the
3 1 result of a learning deficit or due to some other cause (e.g., motivational or arousal differences).
32 These data are difficult to interpret in terms of health hazards to humans under ambient
33 environmental conditions because of the high concentration of diesei exhaust to which the .
34 laboratory rats were exposed. Additionally, there are no lurther concentration-response studies to
35 assess at what exposure levels these observed results persist or abate. A permanent alteration in
36 both learning ability and activity resulting from exposures early in life is a health hazard whose
37 significance to humans should be pursued further.
11/5/99 5-74 DRAFT— DO NOT CITE OR QUOTE
-------
Table 5-14. Effects of chronic exposures to diesel exhaust on behavior and neurophysiology
— •»»
-ft
0
o
J\
Ij
J\
!d
•-
>
n
3
z:
H
~>
Exposure Particles C x T
Species/sex period (mg/m1) (mg-h/ni3)
Rat, Sprague- 8 h/day 6 336-1,344
Dawley, M 7 days/week
1-4 weeks
Rat, Sprague 20 h/day 6 5,040
Dawley, F 7 days week
6 weeks
Rat, Sprague- 8 or 20 h/day 6 1,008-13,440
Dawley, F 7 days/week
3, 4, 6, or
16 weeks
SLA = Spontaneous locomotor activity.
CO NO, SO,
(ppm) (ppm) (ppm) Effects
19 2.5 1.8 Somatosensory and visual evoked
potentials revealed longer pulse
latencies in pups exposed neonatally
19 2.5 1 .8 Reduction in adult SLA and in
neonatal pivoting
19 2.5 1.8 Reduction in SLA 'in adults; neonatal
exposures for 20 or 8 h/day caused
reductions in SLA. Neonatal
exposures for 20 h/day for 1 7 days
resulted in a slower rate of a
bar-pressing task to obtain food
Study
Laurie and Boycs
(1980, 1981)
Laurie et al. (1978)
Laurie etal. (1980)
-------
1 Neurophysiological effects from exposure to diesel exhaust were investigated in rats by Laurie
2 and Boyes (1980, 1981). Rats were exposed to diluted diesel exhaust containing 6 mg/m3 DPM for
3 8 h/day, 7 days/week from birth up until 28 days of age. Somatosensory evoked potential, as
4 elicited by a 1 mA electrical pulse to the tibial nerve in the left hind limb, and visual evoked
5 potential, as elicited by a flash of light, were the end points tested. An increased pulse latency was
6 reported for the rats exposed to diesel exhaust, and this was thought to be caused by a reduction in
7 the degree of nerve myelinization. There was no neuropathological examination, however, to
8 confirm this supposition.
9 Based on the data presented, it is not possible to specify the particular neurological
1 0 impairment(s) induced by the exposure to diesel exhaust. Again, these results occurred following
1 1 exposure to a high level of diesel exhaust and no additional concentration-response studies were
1 2 performed.
13
1 4 5.1.2.3.12. Effects on reproduction and development. Studies of the effects of exposure to diesel
1 5 exhaust on reproduction and development are summarized in Table 5-15. Twenty rats were
1 6 exposed 8 h/day on days 6 through 15 of gestation to diluted diesel exhaust containing 6 mg/m3
1 7 DPM (Werchowski et al., 1980a,b; Pepelko and Peirano, 1983). There were no signs of maternal
1 8 toxicity or decreased fertility. No skeletal or visceral teratogenic effects were observed in 20-day-
1 9 old fetuses (Werchowski et al., 1980a). In a second study, 42 rabbits were exposed to 6 mg/m3
20 DPM for 8 h/day, on gestation days 6 through 18. No adverse effects on body weight gain or
21 fertility were seen in the does exposed to diesel exhaust. No visceral or skeletal developmental
22 abnormalities were observed in the fetuses (Werchowski et al., 1980b).
23 Pepelko and Peirano (1983) evaluated the potential for diesel exhaust to affect reproductive
24 performance in mice exposed from 100 days prior to exposure throughout maturity of the F2
25 generation. The mice were exposed for 8 h/day, 7 days/week to 12 mg/m3 DPM. In general,
26 treatment-related effects were minimal. Some differences hi organ and body weights were noted,
27 but overall fertility and survival rates were not altered by exposure to dicscl exhaust. The only
29 of anthracosis. These data denoted that exposure to diesel exhaust at a concentration of 12 mg/m3
30 did not affect reproduction. See Section 5.3, which reports a lack of effects of exposure to diesel
3 1 exhaust on rat lung development (Mauderly et al., 1987b).
32 Several studies have evaluated the effect of exposure to diesel exhaust on sperm. Lewis et al.
33 (1989) found no adverse sperm effects (sperm motility, velocity, densities, morphology, or
34 incidence of abnormal sperm) in monkeys exposed for 7 h/day, 5 days/ week, for 104 weeks to
35 2mg/m3 DPM. In another study in which A/Strong mice were exposed to diesel exhaust containing
1 1/5/99 5-76 DRAFT— DO NOT CITE OR QUOTE
-------
Table 5-15. Effects of chronic exposures to diesel exhaust on reproduction and development in laboratory animals
-ft
0
o
J\
ij
•^
30
>
n
3
D
2:
D
~i
3
rl
*J
•H
3
3
Species/sex
Mouse,
[C57BL]/
6XC3H]F,, M
Rat, Sprague-
Dawley, F
Rabbit, New
Zealand Albino,
F
Monkey,
Cynomolgus, M
Mouse,
A/Strong, M
Mouse, CD-I,
M,F
Exposure Particles C * T CO
period (mg/m1) (mg-h/m3) (ppm)
5 days 50, 100, or - -
200 mg/kg
in corn oil;
i.p. injection
8h/day 6 571 20
7
days/week
1 .7 weeks
8h/day 6 638 20
7
days/week
1 .9 weeks
7h/day 2 7,280 11.5
5
days/week
104 weeks
8h/day 6 10,416- 20
7 12,768
days/week
31 or
38 weeks
So/day 12 4,032-18,816 33
7
days/week
6 to 28
weeks
NO, SO,
(ppm) (ppm) Effects
— — Dose-related increase in
sperm abnormalities;
decrease in sperm number at
highest dose; testicular
weights unaffected
2.7 2.1 No signs of maternal toxicity
or decreased fertility; no
skeletal or visceral
teratqgenic effects in 20-day-
old fetuses
2.7 2.1 No adverse effects on
maternal weight gain or
fertility; no skeletal or
visceral teratogenic effects in
the fetuses
1 .5 0.8 No effects on sperm motility,
velocity, density,
morphology, or incidence of
abnormalities
2.7 2.1 No effect on sperm
morphology; high rate of
spontaneous sperm
abnormalities may have
masked small effects
4.4 5.0 Overall fertility and survival
rates were unaffected in the
three-generation reproductive
study; only consistent change
noted, an increase in lung
weights, was diagnosed as
anthracosis
Study
Quinto and De
Marinis(1984)
Werchowski et al.
(1980a)
Pepelko and Peirano
(1983)
Werchowski et al.
(1980a)
Pepelko and Peirano
(1983)
Lewis etal. (1989)
Pereiraetal. (1981)
Pepelko and Peirano
(1983)
-------
1 6 mg/m3 DPM for 8 h/day for 31 or 38 weeks, no significant differences were observed in sperm
2 morphology between exposed and control mice (Pereira et al.,)- It "was noted, however, that there
3 was a high rate of spontaneous sperm abnormalities in this strain of mice, and this may have
4 masked any small positive effect. Quinto and De Marinis (1984) reported a statistically significant
5 and dose-related increase in sperm abnormalities in mice injected intraperitoneally for 5 days with
6 50, 100, or 200 mg/kg of DPM suspended in corn oil. A significant decrease in sperm number was
7 seen at the highest dose, but testicular weight was unaffected by the treatment.
8 Watanabe and Oonuki (1999) investigated the effects of diesel engine exhaust on reproductive
9 endocrine function in growing rats. The rats were exposed to whole diesel engine exhaust (5.63
10 mg/m3 DPM, 4.10 ppm NO^, and 8.10 ppm NOX); a group exposed to filtered exhaust without
11 DPM; and a group exposed to clean air. Exposures were for 3 months beginning at birth (6 hr/day
12 for 5 days/week).
13 Serum levels of testosterone and estradiol were significantly higher and follicle-stimulating
14 hormone significantly lower in animals exposed to whole diesel exhaust and filtered exhaust
15 compared to controls. Luteinizing hormone was significantly decreased in the whole exhaust-
16 exposed group as compared to the control and filtered groups. Sperm production and activity of
17 testicular hyaluronidase were significantly reduced in both exhaust-exposed groups as compared to
18 the control group. This study suggests that diesel exhaust stimulates hormonal secretion of the
19 adrenal cortex, depresses gonadotropin-releasing hormone, and inhibits spermatogenesis in rats.
20 Because these effects were not inhibited by filtration, the gaseous phase of the exhaust appears
21 more responsible than particulate matter for disrupting the endocrine system.
22 No teratogenic, embryotoxic, fetotoxic, or female reproductive effects were observed in mice,
23 rats, or rabbits at exposure levels up to 12 mg/m3 DPM. Effects on sperm morphology and number
24 were reported in hamsters and mice exposed to high doses of DPM; however, no adverse effects
25 were observed in sperm obtained from monkeys exposed at 2 mg/m3 for 7 h/day, 5 days/week for
26 104 weeks. Concentrations of 12 mg/m3 DPM did not affect male rat reproductive fertility in the F0
27 and F, generation breeders. Thus, exposure to diesel exhaust would not appear to be a reproductive
23 ut developmental ha^aici.
29
30 5=2. COMPARISON OF HEALTH EFFECTS OF FILTERED AND UNFILTERED
31 DIESEL EXHAUST
32 In four chronic toxicity studies of diesel exhaust, the experimental protocol included exposing
33 test animals to exhaust containing no particles. Comparisons were then made between the effects
34 caused by whole, unfiltered exhaust and those caused by the gaseous components of the exhaust.
35 Concentrations of components of the exposure atmospheres in these four studies are given in Table
36 5-16.
11/5/99 5-78 DRAFT—DO NOT CITE OR QUOTE
-------
Table 5-16. Composition of exposure atmospheres in studies comparing unfiltered and filtered diesel exhaust8
~D
-ft
-J
O
*
>
T)
3
_^
z!
-i
•H
$
AJ
JO
D
— i
i
Species/sex
Rat wistar, f;
hamster, Syrian
Rat, f344, f
Rat, f344, m, f;
hamster, Syrian,
m, f
Rat, wistar, f;
hamster, Syrian, f;
Mouse nmri, f
Mouse, nmri, f,
c57bl/6n, f
Exposure11
period
7 h/day
5 days/week
104 weeks
8 h/day
7 days/week
104 weeks
16 h/day
5 days/week
104 weeks
19 h/day
5 days/week
120 to
140 weeks
18 h/day
5 days/week
23 mo
(nmri)
24 mo
(c57bl/6n)
Uf
F
C
Uf
P
C
Uf
Uf
Uf
f
C
Uf
F"
C
Uf
F
C
Particles
(mg/m1)
3.9
-
—
4.9
-
-
0.7
2.2
6.6
-
-
4.24
-
^~
4.5
0.01
0.0 1
CM
(mg-h/m3)
14,196
28,538
5,824
18,304
54,912
48,336
56,392
40,365
Co No, So,
(ppm) (ppm) (ppm)
18.5 1.2 3.1
18.0 1.0 2.8
_ _ _
7.0 1.8 13.1
_ _ _
- - -
_ _ _
_ _ _
32.0 - -
32.0 - -
1.0 - -
12.5 1.5 3.1
11.1 1.2 1.02
0.16 - -
14.2 2.3 2.8
14.2 2.9 2.4
0.2 0.01 0.1
Effects
No effect on pulmonary function or heart rate in rats; increases in
pulmonary adenomatous proliferations in hamsters, uf
significantly higher than f or c
Body weight decrease after 6 mo in uf, 18 mo in f; lung/body
rate weight rate higher in both groups at 24 mo; at 2 years,
fibrosis and epithelial hyperplasia in lungs of uf; nominal lung
and spleen histologic changes
Uf: elevated red and white cell counts, hematocrit and hemo-
globin; increased heart/body weight and right ventricular/heart
weight ratios; lower left ventricular contractility; changes in
blood chemistry; obstructive and restrictive lung disease; f: no
effects
Uf: decreased body wt in rats and mice but not hamsters;
increased mortality, mice only; decreased lung compliance and
increased airway resistance, rats and hamsters; species
differences in lung lavage enzymes and cell counts and lung
histopathology and collagen content, most pronounced in rats; f:
no effect on glucose-6-phosphate dehydrogenase, total protein,
and lung collagen
Uf: increased lung wet weight starting at 3 mo
F: no noncancer effects reported
Study
Heinrich et al.
(1982)
Iwaietal. (1986)
Brightwell et al.
(1986)
Heinrich et al.
(1986a)
Heinrich et al.
(1995)
'mean values.
buf = unfiltered whole exhaust, f = filtered exhaust, c = control.
'reported to have the same component concentrations as the unfiltered, except particles were present in undetectable amounts.
dconcentrations reported for high concentration level only.
-------
1 Heinrich et al. (1982) compared the toxic effects of whole and filtered diesel exhaust on
2 hamsters and rats. Exposures were for 7 to 8 h/day and 5 days/week. Rats exposed for 24 mo to
3 either whole or filtered exhaust exhibited no significant changes in respiratory frequency,
4 respiratory minute volume, compliance or resistance as measured by a whole-body
5 plethysmography, or heart rate. In the hamsters, histological changes (adenomatous proliferations)
6 were seen in the lungs of animals exposed to either whole or filtered exhaust; however, in all groups
7 exposed to the whole exhaust the number of hamsters exhibiting such lesions was significantly
8 higher than for the corresponding groups exposed to filtered exhaust or clean air. Severity of the
9 lesions was, however, not reported.
10 In a second study, Heinrich et al. (1986a, see also Stober, 1986) compared the toxic effects of
11 whole and filtered diesel exhaust on hamsters, rats, and mice. The test animals (96 per test group)
12 were exposed for 19 h/day, 5 days/week for 120 (hamsters and mice) or 140 (rats) weeks. Body
13 weights of hamsters were unaffected by either exposure. Body weights of rats and mice were
14 reduced by the whole exhaust but not by the filtered exhaust. Exposure-related higher mortality
15 rates occurred in mice after 2 years of exposure to whole exhaust. After 1 year of exposure to the
16 whole exhaust, hamsters exhibited increased lung weights, a significant increase in airway
17 resistance, and a nonsignificant reduction in lung compliance. For the same time period, rats
18 exhibited increased lung weights, a significant decrease in dynamic lung compliance, and a
19 significant increase in airway resistance. Test animals exposed to filtered exhaust did not exhibit
20 such effects. Histopathological examination indicated that different levels of response occurred in
21 the three species. In hamsters, filtered exhaust caused no significant histopathological effects in the
22 lung; whole exhaust caused thickened alveolar septa, bronchioloalveolar hyperplasia, and
23 emphysematous lesions. In mice, whole exhaust, but not filtered exhaust, caused multifocal
24 bronchioloalveolar hyperplasia, multifocal alveolar lipoproteinosis, and multifocal interstitial
25 fibrosis. In rats, there were no significant morphological changes in the lungs following exposure
26 to filtered exhaust. In rats exposed to whole exhaust, there were severe inflammatory changes in
27 the lungs, thickened alveolar septa, foci of macrophages, crystals of cholesterol, and hyperplastic
28 and metaplastic lesions. Biochemical studies of lung lavage fluids of hamsters and mice indicated
29 that exposure to filtered exhaust caused fewer changes than did exposure to whole exhaust. The
30 latter produced significant increases in lactate dehydrogenase, alkaline phosphatase, glucose-6-
31 phosphate dehydrogenase, total protein, protease (pH 5.1), and collagen. The filtered exhaust had a
32 slight but nonsignificant effect on G6P-DH, total protein, and collagen. Similarly, cytological
33 studies showed that while the filtered exhaust had no effect on differential cell counts, the whole
34 exhaust resulted in an increase in leukocytes (161 ± 43.3/yuL versus 55.7 ± 12.8//uL in the controls),
35 a decrease in AMs (30.0 ± 12.5 versus 51.3 ± 12.5/yuL in the controls), and an increase in
1175/99 5-80 DRAFT—DO NOT CITE OR QUOTE
-------
1 granulocytes (125 ± 39.7 versus 1.23 ± 1.147/uL in the controls). All values presented for this study
2 are the mean with its standard deviation. The differences were significant for each cell type. There
3 was also a small increase in lymphocytes (5.81 ± 4.72 versus 3.01 ± 1.23//^L in the controls).
*4 Iwai et al. (1986) exposed rats (24 per group) to whole or filtered diesel exhaust 8 h/day,
5 7 days/week for 24 mo. The whole exhaust was diluted to achieve a concentration of
6 4.9 ± 1.6 mg/m3 DPM. Body weights in the whole exhaust group began to decrease after 6 mo and
7 in both exposed groups began to decrease after 18 mo, when compared with controls. Lung-to-body
8 weight ratios of the rats exposed to the whole exhaust showed a significant increase (p<0.01) after
9 12 mo in comparison with control values. Spleen-to-body weight ratios of both exposed groups
10 were higher than control values after 24 mo. After 6 mo of exposure to whole exhaust, DPM
11 accumulated in AMs, and Type II cell hyperplasia was observed. After 2 years of exposure, the
12 alveolar walls had become fibrotic with mast cell infiltration and epithelial hyperplasia. In rats
13 exposed to filtered exhaust, after 2 years there were only minimal histologic changes in the lungs,
14 with slight hyperplasia and stratification of bronchiolar epithelium and infiltration of atypical
15 lymphocytic cells in the spleen.
16 Brightwell et al. (1986) evaluated the toxic effects of whole and filtered diesel exhaust on rats
17 and hamsters. Three exhaust dilutions were tested, producing concentrations of 0.7, 2.2, and 6.6
18 mg/m3 DPM. The test animals (144 rats and 312 hamsters per exposure group) were exposed for
19 five 16-h periods per week for 2 years. The four exposure types were gasoline, gasoline catalyst,
Bl diesel, and filtered diesel. The results presented were limited to statistically significant differences
21 between exhaust-exposed and control animals. The inference from the discussion section of the
22 paper was that there was a minimum of toxicity in the animals exposed to filtered diesel exhaust:
23 "It is clear from the results presented that statistically significant differences between exhaust-
24 exposed and control animals are almost exclusively limited to animals exposed to either gasoline or
25 unfiltered diesel exhaust." Additional results are described in Section 5.1.2.3.
26 Heinrich et al. (1995) exposed female NMRI and C57BL/6N mice to a diesel exhaust dilution
27 that resulted in a DPM concentration of 4.5 mg/m3 and to the same dilution after filtering to remove
28 the particles. This study is focused on the carcinogenic effects of DPM exposure, and inadequate
29 information was presented to compare noncancer effects in filtered versus unfiltered exhaust.
30 A comparison of the toxic responses in laboratory animals exposed to whole exhaust or
31 filtered exhaust containing no particles demonstrates across studies that when the exhaust is
32 sufficiently diluted to limit the concentrations of gaseous irritants (NO2 and SO2), irritant vapors
33 (aldehydes), CO, or other systemic toxicants, the diesel particles are the prime etiologic agents of
34 noncancer health effects, although additivity or synergism with the gases cannot be ruled out.
35 These toxic responses are both functional and pathological and represent cascading sequelae of lung
11 /5/99 5-81 DRAFT—DO NOT CITE OR QUOTE
-------
1 pathology based on concentration and species. The diesel particles plus gas exposures produced
2 biochemical and cytological changes in the lung that are much more prominent than those evoked
3 by the gas phase alone. Such marked differences between whole and filtered diesel exhaust are also
4 evident from general lexicological indices, such as decreases in body weight and increases in lung
5 weights, pulmonary function measurements, and pulmonary histopathology (e.g., proliferative
6 changes in Type II cells and respiratory bronchiolar epithelium, fibrosis). Hamsters, under
7 equivalent exposure regimens, have lower levels of retained DPM in their lungs than rats and mice
8 do and, consequently, less pulmonary function impairment and pulmonary pathology. These
9 differences may result from lower DPM inspiration and deposition during exposure, greater DPM
10 clearance, or lung tissue less susceptible to the cytotoxicity of deposited DPM.
11
12 5.3. INTERACTIVE EFFECTS OF DIESEL EXHAUST
13 A multitude of factors may influence the susceptibility to exposure to diesel exhaust as well as
14 the resulting response. Some of these have already been discussed in detail (e.g., the composition
15 of diesel exhaust and concentration-response data); others will be addressed in this section (e.g., the
16 interaction of diesel exhaust with factors particular to the exposed individual and the interaction of
17 diesel exhaust components with other airborne contaminants).
18 Mauderly et al. (1990a) compared the susceptibility of normal rats and rats with preexisting
19 laboratory-induced pulmonary emphysema exposed for 7 h/day, 5 days/week for 24 mo to diesel
20 exhaust containing 3.5 mg/m3 DPM or to clean air (controls). Emphysema was induced in one-half
21 of the rats by intratracheal instillation of elastase 6 weeks before exhaust exposure. Measurements
22 included lung burdens of DPM, respiratory function, bronchoalveolar lavage, clearance of
23 radiolabeled particles, pulmonary immune responses, lung collagen, excised lung weight and
24 volume, histopathology, and mean linear intercept of terminal air spaces. None of the data for the
25 63 parameters measured suggest that rats with emphysematous lungs were more susceptible than
26 rats with normal lungs to the effects of diesel exhaust exposure. In fact, each of the 14 emphysema-
27 exhaust interactions detected by statistical analysis of variance indicated that emphysema acted to
2O «.~*J..^~~ 4-U^r ,^.CPW,*+.« .~-P *4«*-k«**t ,-v«rl»rn »*•••*• rtv*%^n-*»^a T^opol mo*^v^>TiYo+ja r**of+*»i- o/^^ifwmlci'f'/arl mi»r»ll loco
W l^UkAWW C11W W.1..1.WWIO V/A UAW|3WJL *w f\ I 144 ti«>Tl W^Y|^V k?l*A w. A~rAWb»WA f./«AA *A W •AAltfcW A**l*bk^A V*+f*f hM.**«u£»*w*» ^A^Wfef^A *^wh*
29 rapidly in the lungs of emphysematous rats than in those of normal rats. The mean lung burdens of
30 DPM in the emphysematous rats were 39%, 36%, and 37% of the lung burdens of normal rats at 12,
31 18, and 24 mo, respectively. No significant interactions were observed among lung morphometric
32 parameters. Emphysema prevented the exhaust-induced increase for three respiratory indices of
33 expiratory flow rate at low lung volumes, reduced the exhaust-induced increase in nine lavage fluid
34 indicators ol lung damage, prevented the expression of an exhaust-induced increase in iung
35 collagen, and reduced the exhaust-induced delay in DPM clearance.
11/5/99 5-82 DRAFT—DO NOT CITE OR QUOTE
-------
1 Mauderly et al. (1987b) evaluated the relative susceptibility of developing and adult rat lungs
2 to damage by exposure to diesel exhaust. Rats (48 per test group) were exposed to diesel exhaust
containing 3.5 mg/m3 DPM and about 0.8 ppm NO2. Exposures were for 7 h/day,
5 days/week through gestation to the age of 6 mo, or from the age of 6 to 12 mo. Comparative
5 studies were conducted on respiratory function, immune response, lung clearance, airway fluid
6 enzymes, protein and cytology, lung tissue collagen, and proteinases in both age groups. After the
7 6-month exposure, adult rats, compared with controls, exhibited (1) more focal aggregates of
8 particle-containing AMs in the alveolar ducts near the terminal bronchioles, (2) a sixfold increase in
9 the neutrophils (as a percentage of total leukocytes) in the airway fluids, (3) a significantly higher
10 number of total lymphoid cells in the pulmonary lymph nodes, (4) delayed clearance of DPM and
11 radiolabeled particles (tI/2 = 90 days versus 47 days for controls), and (5) increased lung weights.
12 These effects were not seen in the developing rats. On a weight for weight (milligrams of DPM per
13 gram of lung) basis, DPM accumulation in the lungs was similar in developing and adult rats
14 immediately after the exposure. During the 6-month postexposure period, DPM clearance was
15 much more rapid in the developing rats, approximately 2.5-fold. During postexposure, diesel
16 particle-laden macrophages became aggregated in the developing rats, but these aggregations were
17 located primarily in a subpleural position. The authors concluded that exposure to diesel exhaust,
18 using pulmonary function, structural (qualitative or quantitative) biochemistry as the indices, did
19 not affect the developing rat lung more severely than the adult rat lung.
Vi As a result of the increasing trend of using diesel-powered equipment in coal mining
21 operations and the concern for adverse health effects hi coal miners exposed to both coal dust or
22 coal mine dust and diesel exhaust, Lewis et al. (1989) and Karagianes et al. (1981) investigated the
23 interaction of coal dust and diesel exhaust. Lewis et al. (1989) exposed rats, mice, and cynomolgus
24 monkeys to (1) filtered ambient air, (2) 2 mg/m3 DPM, (3) 2 mg/m3 respirable coal dust, and (4) 1
25 mg/m3 of both DPM and respirable coal dust. Gaseous and vapor concentrations were identical in
26 both diesel exhaust exposures. Exposures were for 7 h/day, 5 days/week for up to 24 mo.
27 Synergistic effects between diesel exhaust and coal dust were not demonstrated; additive toxic
28 effects were the predominant effects noted.
29 Karagianes et al. (1981) exposed rats (24 per group) to diesel exhaust containing 8.3 mg/m3 of
30 DPM alone or in combination with about 6 mg/m3 of coal dust. No synergistic effects were found
31 between diesel exhaust and coal dust; additive effects in terms of visual dust burdens in
32 necropsied lungs were related to dose (i.e., length of exposure and airborne paniculate
33 concentrations).
34 The health effects of airborne contaminants from sources other than diesel engines may be
35 altered in the presence of DPM by their adsorption onto the diesel particles. When adsorbed onto
11/5/99 5-83 DRAFT—DO NOT CITE OR QUOTE
-------
1 diesel particles, the gases and vapors can be transported and deposited deeper into the lungs, and
2 because they are more concentrated on the particle surface, the resultant cytotoxic effects or
3 physiological responses may be enhanced. Nitrogen dioxide adsorbed onto carbon particles caused
4 pulmonary parenchymal lesions in mice, whereas NO2 alone produced edema and inflammation but
5 no lesions (Boren, 1964). Exposure to formaldehyde and acrolein adsorbed onto carbon particles (1
6 to 4 //m) resulted in the recruitment of PMNs to tracheal and intrapulmonary epithelial tissues but
7 not when the aldehydes were tested alone (Kilburn and McKenzie, 1978).
8 There is no direct evidence that diesel exhaust, at concentrations found in the ambient
9 environment, interacts with other substances in the exposure environment or the physiological
10 status of the exposed subject other than impaired resistance to respiratory tract infections. Although
11 there is experimental evidence that gases and vapors can be adsorbed onto carbonaceous particles,
12 enhancing the toxicity of these particles when deposited in the lung, there is no evidence for an
13 increased health risk from such interactions with DPM under urban atmospheric conditions.
14 Likewise, there is no experimental evidence in laboratory animals that the youth or preexisting
15 emphysema of an exposed individual enhances the risk of exposure to diesel exhaust.
16
17 5.4. COMPARATIVE RESPONSIVENESS AMONG SPECIES TO THE PULMONARY
18 EFFECTS OF DIESEL EXHAUST
19 There is some evidence indicating that species may differ in pulmonary responses to diesel
20 exhaust (DE). Mauderly (1994) compared the pulmonary histopathology of rats and mice after 18
21 mo of exposure to DE. There was less aggregation of macrophages in rats. Diffuse septal
22 thickening was noted in the mice, but there were few inflammatory cells, no focal fibrosis, little
23 epithelial hyperplasia, and no epithelial metaplasia, as was observed in rats. Heinrich et al. (1986a)
24 reported that wet lung weight of hamsters increased only 1.8-fold following chronic exposure to
25 DE, compared with an increase of 3.4-fold in rats. Smaller increases in neutrophils, lactic acid
26 dehydrogenase, collagen, and protein supported the conclusion of a lesser inflammatory response in
27 Syrian hamsters. The histopathologic changes in the lungs of Chinese hamsters after 6 rao exposure
28 to DE. on the other hand, was similar to that of rats (Fepeiko and reirano, I9S3). Guinea pigs
29 respond to chronic DE exposure with a well-defined epithelial proliferation, but it is based on an
30 eosinophilic response in contrast to the,neutrophil-based responses in other species. Epithelial
31 hyperplasia and metaplasia were quite striking in the terminal and respiratory bronchioles of cats
32 exposed for 27 mo to DE (Plopper et al., 1983). This study is of particular interest because the
33 terminal airways of cats are more similar to those of humans than rodent species are. It should be
34 noted, however, that exposure concentrations were very high (12 mg/m3) for most of the period.
35 Lewis et al. (1989) exposed rats and Cynomolgus monkeys 8 hours per day, 5 days per week for 2
11/5/99 5-84 DRAFT—DO NOT CITE OR QUOTE
-------
1 years to DE at a particle concentration of 2 mg/m3. Unfortunately, this exposure rate was
2 sufficiently low that few effects were noted in either species other than focal accumulations of
particles, primarily in the alveolar macrophages, interstitium, and lymphoid tissue. It is apparent
that species do vary in their pulmonary responses to DE exposure, despite the difficulty in making
5 direct comparisons because of differences in exposure regimes, lifespans, and pulmonary anatomy.
6 Most species do respond, however, suggesting that humans are likely to be susceptible to induction
7 of pulmonary pathology during chronic exposure to DE.
8
9 5.5. DOSE-RATE AND PARTICULATE CAUSATIVE ISSUES
10 The purpose of animal toxicological experimentation is to identify the hazards and
11 dose-response effects posed by a chemical substance or complex mixture and to extrapolate these
12 effects to humans for subsequent health assessments. The cardinal principle in such a process is
13 that the intensity and character of the toxic action are a function of the dose of the toxic agent(s) that
14 reaches the critical site of action. The considerable body of evidence reviewed clearly denotes that
15 major noncancerous health hazards may be presented to the lung following the inhalation of diesel
16 exhaust. Based on pulmonary function and histopathological and histochemical effects, a
17 determination can be made concerning which dose/exposure rates of diesel exhaust (expressed in
18 terms of the DPM concentration) result in injury to the lung and which appear to elicit no effect.
*The inhalation of poorly soluble particles, such as those found in diesel exhaust, increases the
pulmonary particulate burden. When the dosing rate exceeds the ability of the pulmonary defense
21 mechanisms to achieve a steady-state lung burden of particles, there is a slowing of clearance and
22 the progressive retention of particles in the lung that can ultimately approach a complete cessation
23 of lung clearance (Morrow, 1988). This phenomenon, which is reviewed in Section 3.4, particle
24 overload, has practical significance both for the interpretation of experimental inhalation data and
25 for the prevention of disease in humans exposed to airborne particles.
26 The data for exposure intensities that cause adverse pulmonary effects demonstrate that they
27 are less than the exposure intensities reported to be necessary to induce lung tumors. Using the
28 most widely studied laboratory animal species and the one reported to be the most sensitive to
29 tumor induction, the laboratory rat, the no-adverse-effect exposure intensity for adverse pulmonary
30 effects was 56 mg-h-m'3/week (Brightwell etal., 1986). The lowest-observed-effect level for
31 adverse pulmonary effects (noncancer) in rats was 70 mg-h-nr3/week (Lewis et al., 1989), and for
32 pulmonary tumors, 122.5 mg-h-m'Vweek (Mauderly et al., 1987a). The results clearly show that
33 noncancerous pulmonary effects are produced at lower exposure intensities than are pulmonary
34 tumors. Such data support the position that inflammatory and proliferative changes in the lung may
35 play a key role in the etiology of pulmonary tumors in exposed rats (Mauderly et al., 1990b).
1175/99 5-85 DRAFT—DO NOT CITE OR QUOTE
-------
1 Adults who have a preexisting condition that may predispose their lungs to increased particle
2 retention (e.g., smoking or high paniculate burdens from nondiesel sources), inflammation (e.g.,
3 repeated respiratory infections), epithelial proliferation (e.g., chronic bronchitis), and fibrosis (e.g.,
4 silica exposure) and infants and children, due to their developing pulmonary and immunologic
5 systems, may have a greater susceptibility to the toxic actions of diesel exhaust. It should be noted
6 that both the developing lung and a model of a preexisting disease state have been studied with
7 regard to their effect on the lungs' response to diesel exhaust (Mauderly et al., 1990a, 1987b).
8 Mauderly et al. (1987b) showed that diesel did not affect the developing lung more severely than
9 the adult rat lung, and in fact, that clearance was faster in the younger lung. Mauderly et al. (1 990a)
1 0 compared the pulmonary response to inhalation of diesel exhaust in rats with elastase-induced
1 1 emphysema with normal rats. They found that respiratory tract effects were not more severe in
1 2 emphysematous rats and that the lung burden of particles was less in the compromised rat. These
1 3 studies provide limited evidence that some factors that are often considered to result in a wider
1 4 distribution of sensitivity among members of the population may not have this effect with diesel
1 5 exposure. However, these studies have no counterpart in human studies and extrapolation to
1 6 humans remains uncertain.
1 7 There is also the issue of whether the noncancerous health effects related to exposure to diesel
1 8 exhaust are caused by the carbonaceous core of the particle or substances adsorbed onto the core, or
19 both.
20 Current understanding suggests that much of the toxicity resulting from the inhalation of
2 1 diesel exhaust relates to the carbonaceous core of the particles. Several studies on inhaled aerosols
22 demonstrate that lung reactions characterized by an appearance of particle-laden AMs and their
23 infiltration into the alveolar ducts, adjoining alveoli and tracheobronchial lymph nodes, hyperplasia
24 of Type II cells, and the impairment of pulmonary clearance mechanisms are not limited to
25 exposure to diesel particles. Such responses have also been observed in rats following the
26 inhalation of coal dust (Lewis et al., 1989; Karagianes et al.9 198 1), titanium dioxide (Heinrich et
27 ai., 1995; LeeeiaL, 1985), carbon black (Nikula et al., 1995; Heinrich et ai., 1995), titanium
20 «,vtOT.,a.i. — : A~. \,,,A-~\,,~:~ ~..n/4,,»4.~ (\ ~^ n* „!
U l^ tA Cl^/111 WA AV^W 11 T V1AV& J OAO JL/i Wh*Wt*3 yj— IWW Wl C4A., A-SOWy, V^1.4t«AI^> ^.Lt^Wk)lWA AA/lbWA ClliVA ' .... •.--. • n". | I I j l
-------
1 Dungworth et al. (1994) reported moderate to severe inflammation characterized by multifocal
2 bronchoalveolar hyperplasia, alveolar histiocytosis, and focal segmental fibrosis in rats exposed to
carbon black for up to 20 mo at exposure rates of 510 to 540 mg-h-nT3/week. The observed lung
pathology reflects notable dose-response relationships and usually evolves in a similar manner.
5 With increasing dose, there is an increased accumulation and aggregation of particle-laden AMs,
6 Type II cell hyperplasia, a foamy (degenerative) macrophage response, alveolar proteinosis,
7 alveolar bronchiolization, cholesterol granulomas, and often squamous cell carcinomas and
8 bronchioalveolar adenomas derived from metaplastic squamous cells in the areas of alveolar
9 bronchiolization.
10 Heinrich et al. (1995) compared effects of diesel exposure in rats and mice with exposure to
11 titanium dioxide or carbon black. Exposures to TiO2 and carbon black were adjusted during the
12 exposure to result in a similar lung burden for the three types of particles. At similar lung burdens
13 in the rat, DPM, TiO2, and carbon black had nearly identical effects on lung weights and on the
14 incidence of lesions, both noncancer and cancer. Also, a similar effect on clearance of a labeled test
15 aerosol was measured for the different particles. A comparison of the effect of DPM, TiO2, and
16 carbon black exposures in mice also showed a similar effect on lung weight, but noncancer effects
17 were not reported and no significant increase in tumors was observed.
18 Murphy et al. (1998) compared the toxicological effects of DPM with three other particles
« chosen for their differing morphology and surface chemistry. One mg each of well-characterized
crystalline quartz, amorphous silica, CB, and DPM was administered to laboratory rats by a single
21 intratracheal instillation. The laboratory rats were sacrificed at 48 h, and 1, 6, and 12 weeks after
22 instillation. Crystalline quartz produced significant increases hi lung permeability, persistent surface
23 inflammation, progressive increases hi pulmonary surfactant and activities of epithelial marker
24 enzymes up to 12 wk after primary exposure. Amorphous silica did not cause progressive effects
25 but did produce initial epithelial damage with permeability changes that regressed with time after
26 exposure. By contrast, CB had little if any effect on lung permeability, epithelial markers, or
27 inflammation. Similarly, DPM produced only minimal changes, although the individual particles
28 were smaller and differed hi surface chemistry from CB. The authors concluded that DPM is less
29 damaging to the respiratory epithelium than silicon dioxide, and that the surface chemistry of the
30 particle is more important than ultrafine size in explaining biological activity.
31 These experiments provide strong support for the idea that diesel exhaust toxicity results from
32 a mechanism that is analogous to that of other relatively inert particles in the lung. This qualitative
33 similarity exists along with some apparent quantitative differences in the potency of various
34 particles for producing effects on the lung or on particle clearance.
35
11/5/99 5-87 DRAFT—DO NOT CITE OR QUOTE
-------
1 The exact relationship between toxicity and particle size within the ultrafine particle mode,
2 including DPM (BeruBe et al., 1999), remains unresolved. Studies reviewed in the PM CD (U.S.
3 Environmental Protection Agency, 1996) suggest a greater inherent potential toxicity of inhaled
4 ultrafine particles. Exposure to ultrafine particles may increase the release of promflammatory
5 mediators that could be involved in lung disease. For example, Driscoll and Maurer (1991)
6 compared the effects of fine (0.3 um) and ultrafine (0.02 um) TiO2 particles instilled into the lungs
7 of laboratory rats. Although both size modes caused an increase in the numbers of AMs and PMNs
8 in the lungs, and release of TNF and fibronectin by AMs the responses were greater and more
9 persistent with the ultrafine particles. While fine particle exposure resulted in a minimally
10 increased prominence of particle-laden macrophages associated with alveolar ducts, ultrafine
11 particle exposure produced a somewhat greater prominence of macrophages, some necrosis of
12 macrophages and slight interstitial inflammation of the alveolar duct region. Moreover, collagen
13 increased only with exposure to ultrafine particles.
14 Oberdorster et al. (1992) compared the effects of fine (0.25 um) and ultrafine (0.02 um) TiO2
15 particles instilled into the lungs of laboratory rats on various indicators of inflammation. Instillation
16 of ultrafine particles increased the number of total cells recovered by lavage, decreased the
17 percentage of AMs, and increased the percentage of PMNs and increased protein. Instillation with
18 fine particles did not cause statistically significant effects. Thus, the ultrafine particles had greater
19 pulmonary inflammatory potency than did larger sizes of this material. The investigators attributed
20 the enhanced toxicity to greater interaction of the ultrafine particles with their large surface area,
21 with alveolar and interstitial macrophages, which resulted in enhanced release of inflammatory
22 mediators. They suggested that ultrafine particles of low in vitro solubility appear to enter the
23 interstitium more readily than do larger sizes of the same material, which accounted for the
24 increased contact with macrophages in this compartment of the lung. Driscoll and Maurer (1991)
25 noted that the pulmonary retention of ultrafine TiO2 particles instilled into rat lungs was greater
26 than for the same mass of fine mode TiO2 particles. Thus, the available evidence tends to suggest a
27 potentially greater toxicity for inhaled ultrafine particles.
28 Particle size, volume, surface area, and composition may be the critical elements in the
29 overload phenomenon following exposure to particles, which could explain those quantitative
30 differences. The overloaded AMs secrete a variety of cytokines, oxidants, and proteolytic enzymes
31 that are responsible for inducing particle aggregation and damaging adjacent epithelial tissue
32 (Oberdorster, 1994). For a more detailed discussion of mechanism, see Chapter 3.
33 The principal noncancerous health hazard to humans posed by exposure to diesel exhaust is a
34 structural or functional injury to the lung based on the laboratory animal data. Such effects are
35 demonstrable at dose rates or cumulative doses of DPM lower than those reported to be necessary to
1175/99 5-88 DRAFT—DO NOT CITE OR QUOTE
-------
1 induce lung tumors. An emerging human health issue concerning short-term exposure to ambient
2 DE/DPM is the potential for allergenic responses in several studies. Heightened allergenic
responses including increased cytokine production as well as increased numbers of inflammatory
cells have been detected in nasal lavage from humans exposed to inhaled or instilled DE/DPM. In
5 individuals already allergic to ragweed, exposure to DE/DPM with the allergen was observed to
6 result in an enhanced allergenic response, particularly IgE production. Current knowledge indicates
7 that the carbonaceous core of diesel particles is the major causative factor in the injury to the lung
8 and that other factors such as the cytotoxicity of adsorbed substances on the particles also may play
9 a role. The lung injury appears to be mediated through effects on pulmonary AMs. Because
10 noncancerous pulmonary effects occur at lower doses than tumor induction does in the rat, and
11 because these effects may be cofactors in the etiology of diesel exhaust-induced tumors,
12 noncancerous pulmonary effects must be considered in the total evaluation of diesel exhaust,
13 notably the particulate component.
14
15 5.6. SUMMARY AND DISCUSSION
16 5.6.1. Effects of Diesel Exhaust on Humans
17 The most readily identified acute noncancer health effect of diesel exhaust on humans is its
18 ability to elicit subjective complaints of eye, throat, and bronchial irritation and neurophysiological
« symptoms such as headache, lightheadedness, nausea, vomiting, and numbness and tingling of the
extremities. Studies of the perception and offensiveness of the odor of diesel exhaust and a human
21 volunteer study in an exposure chamber have demonstrated that the time of onset of the human
22 subjective symptoms is inversely related to increasing concentrations of diesel exhaust and the
23 severity is directly related to increasing concentrations of diesel exhaust. In one study in which a
24 diesel engine was operated under varying load conditions, a dilution factor of 140 to 475 was
25 needed to reduce the exhaust level to an odor-detection threshold level.
26 A public health issue is whether short-term exposure to diesel exhaust might result in an acute
27 decrement in ventilatory function and whether the frequent repetition of such acute respiratory
28 effects could result in chronic lung function impairment. One convenient means of studying acute
29 decrements in ventilatory function is to monitor differences in pulmonary function in occupationally
30 exposed workers at the beginning and end of a workshift. In studies of underground miners, bus
31 garage workers, dock workers, and locomotive repairmen, increases in respiratory symptoms
32 (cough, phlegm, and dyspnea) and decreases in lung function (FVC, FEV,, PEFR, and FEF25.75)
33 over the course of a workshift were generally found to be minimal and not statistically significant.
34 In a study of acute respiratory responses in diesel bus garage workers, there was an increased
35 reporting of cough, labored breathing, chest tightness, and wheezing, but no reductions in
11/5/99 5-89 DRAFT—DO NOT CITE OR QUOTE
-------
1 pulmonary function were associated with exposure to diesel exhaust. Pulmonary function was
2 affected in stevedores over a workshift exposure to diesel exhaust but normalized after a few days
3 without exposure to diesel exhaust fumes. In a third study, there was a trend toward greater
4 ventilatory function changes during a workshift among coal miners, but the decrements were similar
5 in miners exposed and not exposed to diesel exhaust.
6
7 Smokers appeared to demonstrate larger workshift respiratory function decrements and
8 increased incidents of respiratory symptoms. Acute sensory and respiratory symptoms were earlier
9 and more sensitive indicators of potential health risks from diesel exposure than were decrements in
1 0 pulmonary function. Studies on the acute health effects of exposure to diesel exhaust in humans,
1 1 experimental and epidemiologic, have failed to demonstrate a consistent pattern of adverse effects
1 2 on respiratory morbidity; the majority of studies offer, at best, equivocal evidence for an exposure-
1 3 response relationship. The environmental contaminants have frequently been below permissible
1 4 workplace exposure limits; in those few cases where health effects have been reported, the authors
1 5 have failed to identify conclusively the individual or collective causative agents in the diesel
1 6 exhaust.
1 7 Chronic effects of diesel exhaust exposure have been evaluated in epidemiologic studies of
1 8 occupationally exposed workers (metal and nonmetal miners, railroad yard workers, stevedores, and
1 9 bus garage mechanics). Most of the epidemiologic data indicate an absence of an excess risk of
20 chronic respiratory disease associated with exposure to diesel exhaust. In a few studies, a higher
2 1 prevalence of respiratory symptoms, primarily cough, phlegm, or chronic bronchitis, was observed
22 among the exposed. These increased symptoms, however, were usually not accompanied by
23 significant changes in pulmonary function. Reductions in FEV, and FVC and, to a lesser extent,
24 FEF50 and FEF75, also have been reported. Two studies detected statistically significant decrements
25 in baseline pulmonary function consistent with obstructive airway disease. One study of stevedores
26 had a limited sample size of 17 exposed and 1 1 controls. The second study in coal miners showed
27 that boih underground and surface workers at diesei-use mines had somewhat lower pulmonary
, . .
l'i; UACU.1 U1WA1 lAXCLLtsAAWU. t/U>ll!l.V/13. JL JL1W Jjl tj'jjiu'l CJ.UJL1 Ol VVOllV^ia iil
29 however, showed equivalent evidence of obstructive airway disease and for this reason the authors
30 of the second paper felt that factors other than diesel exposure might have been responsible. A
3 1 doubling of minor restrictive airway disease was also observed in workers in or at diesel-use mines.
32 These two studies, coupled with other reported nonsignificant trends in respiratory flow- volume
33 measurements, suggest that exposure to diesel exhaust may impair pulmonary function among
34 occupational populations. Jtpidemiologic studies of the effects of diesel exhaust on organ systems
35 other than the pulmonary system are scant. Whereas a preliminary study of the association of
11/5/99 5-90 DRAFT— DO NOT CITE OR QUOTE
-------
1 cardiovascular mortality and exposure to diesel exhaust found a fourfold higher risk ratio, a more
2 comprehensive epidemiologic study by the same investigators found no significant difference
between the observed and expected number of deaths caused by cardiovascular disease.
Caution is warranted in the interpretation of results from the epidemiologic studies that have
5 addressed noncarcinogenic health effects from exposure to diesel exhaust. These investigations
6 suffer from myriad methodological problems, including (1) incomplete information on the extent of
7 exposure to diesel exhaust, necessitating in some studies estimations of exposures from job titles
8 and resultant misclassification; (2) the presence of confounding variables such as smoking or
9 occupational exposures to other toxic substances (e.g., mine dusts); and (3) the short duration and
i
10 low intensity of exposure. These limitations restrict drawing definitive conclusions as to the cause
11 of any noncarcinogenic diesel exhaust effect, observed or reported.
.12
13 5.6.2. Effects of Diesel Exhaust on Laboratory Animals
14 Laboratory animal studies of the toxic effects of diesel exhaust have involved acute,
15 subchronic, and chronic exposure regimens. In acute exposure studies, toxic effects appear to have
16 been associated primarily with high concentrations of carbon monoxide, nitrogen dioxide, and
17 aliphatic aldehydes. In short- and long-term studies, toxic effects have been associated with
18 exposure to the complex exhaust mixture. Effects of diesel exhaust hi various animal species are
« summarized in Tables 5-2 to 5-15. In short-term studies, health effects are not readily apparent, and
when found, are mild and result from concentrations of about 6 mg/m3 DPM and durations of
21 exposure approximating 20 h/day. There is ample evidence, however, that short-term exposures at
22 lower levels of diesel exhaust affect the lung, as indicated by an accumulation of DPM, evidence of
23 inflammatory response, AM aggregation and accumulation near the terminal bronchioles, Type II
24 cell proliferation, and the thickening of alveolar walls adjacent to AM aggregation. Little evidence
25 exists, however, from short-term studies that exposure to diesel exhaust impairs lung function.
26 Chronic exposures cause lung pathology that results in altered pulmonary function and increased
27 DPM retention in the lung. Exposures to diesel exhaust have also been associated with increased
28 susceptibility to respiratory tract infection, neurological or behavioral changes, an increase in
29 banded neutrophils, and morphological alterations in the liver.
30
31 5.6.2.1. Effects on Survival and Growth
32 The data presented in Table 5-3 show limited effects on survival in mice and rats and some
33 evidence of reduced body weight in rats following chronic exposures to concentrations of
34 1.5 mg/m3 DPM or higher and exposure durations of 16 to 20 h/day, 5 days/week for 104 to
35 130 weeks. Increased lung weights and lung to body weight ratios in rats, mice, and hamsters;
11/5/99 5-91 DRAFT—DO NOT CITE OR QUOTE
-------
1 an increased heart to body weight ratio in rats; and decreased lung and kidney weights in cats have
2 been reported following chronic exposure to diesel exhaust. No evidence was found of an effect of
3 diesel exhaust on other body organs (Table 5-4). The lowest-observed-effect level in rats
4 approximated 1 to 2 mg/m3 DPM for 7 h/day, 5 days/week for 104 weeks.
5
6
7 5.6.2.2. Effects on Pulmonary Function
8 Pulmonary function impairment has been reported in rats, hamsters, cats, and monkeys
9 exposed to diesel exhaust and included lung mechanical properties (compliance and resistance),
10 diffusing capacity, lung volumes, and ventilatory performance (Table 5-5). The effects generally
11 appeared only after prolonged exposures. The lowest exposure levels (expressed in terms of DPM
12 concentrations) that resulted in impairment of pulmonary function occurred at 2 mg/m3 in
13 cynomolgus monkeys (the only level tested), 1.5 and 3.5 mg/m3 in rats, 4.24 and 6 mg/m3 in
14 hamsters, and 11.7 mg/m3 in cats. Exposures in monkeys, cats, and rats (3.5 mg/m3) were for 7 to 8
15 h/day, 5 days/week for 104 to 130 weeks. While this duration is considered to constitute a lifetime
16 study in rodents, it is a small part of the lifetime of a monkey or cat. Exposures in hamsters and rats
17 (1.5 mg/m3) varied in hours per day (8 to 20) and weeks of exposure (26 to 130). In all species but
18 the monkey, the testing results were consistent with restrictive lung disease; alteration in expiratory
19 flow rates indicated that 1.5 mg/m3 DPM was a LOAEL for a chronic exposure (Gross, 1981).
20 Monkeys demonstrated evidence of obstructive airway disease. The nature of the pulmonary
21 impairment is dependent on the dose of toxicants delivered to and retained in the lung, the site of
22 deposition and effective clearance or repair, and the anatomy and physiology of the affected
23 species; these variables appear to be factors in the disparity of the airway disease in monkey versus
24 the other species tested.
25
26 5.6.2.3. Histopathological and Histochemical Effects
27 Histoiogicai studies have demonstrated that chronic exposure to diesei exhaust can result in
**/•» ./3" .. .•... , .: .-*_,..A __A^.: /-^--vt-- /r r\ in :__ixr I.:.._._.: '.-,",, „« * >..•„-.• •._-.:_ A n JT
<£o tijuccu> Oil icapiiauuiy uaci tiaiuc ( lauifc J-\J>. i_ypicai luiu-uigi inciuCic aivcOidi liisuOcy lusia, .i-iivx
29 aggregation, tissue inflammation, increase in PMNs, hyperplasia of bronchiolar and alveolar Type II
30 cells, thickened alveolar septa, edema, fibrosis, and emphysema. Lesions in the trachea and bronchi
31 were observed in some studies. Associated with these histopathological findings were various
32 histochemical changes in the lung, including increases hi lung DNA, total protein, alkaline and acid
33 phosphatase, glucose-6-phosphate dehydrogenase; increased synthesis of collagen; and release of
34 inflammatory mediators such as leukotriene LTB and prostaglandin PGF2a. Although the overall
35 laboratory evidence is that prolonged exposure to DPM results in histopathological and
1175/99 5-92 DRAFT—DO NOT CITE OR QUOTE
-------
1 histochemical changes in the lungs of exposed animals, some studies have also demonstrated that
2 there may be a threshold of exposure to DPM below which pathologic changes do not occur. These
3 no-observed-adverse-effect levels for histopathological effects were reported to be 2 mg/m3 for
4 cynomolgus monkeys (the only concentration tested), 0.11 to 0.35 mg/m3 for rats, and 0.25 mg/m3
5 DPM for guinea pigs exposed for 7 to 20 h/day, 5 to 5.5 days/week for 104 to 130 weeks.
6
7 5.6.2.4. Effects on Airway Clearance
8 The pathological effects of DPM appear to be strongly dependent on the relative rates of
9 pulmonary deposition and clearance (Table 5-7). Clearance of particles from the alveolar region of
10 the lungs is a multiphasic process involving phagocytosis by AMs. Chronic exposure to DPM
11 concentrations of about 1 mg/m3 or above, under varying exposure durations, causes pulmonary
12 clearance to be reduced with concomitant focal aggregations of particle-laden AMs, particularly in
13 the peribronchiolar and alveolar regions, as well as in the hilar and mediastinal lymph nodes. The
14 exposure concentration at which focal aggregates of particle-laden AMs occur may vary from
15 species to species, depending on rate of uptake and pulmonary deposition, pulmonary clearance
16 rates, the relative size of the AM population per unit of lung tissue, the rate of recruitment of AMs
17 and leukocytes, and the relative efficiencies for removal of particles by the mucociliary and
18 lymphatic transport system. The principal mechanism of reduced particle clearance appears to be
19 an effect on pulmonary AMs. Impairment of particle clearance seems to be nonspecific and applies
!0 primarily to dusts that are persistently retained in the lungs. Lung dust levels of approximately 0.1
21 to 1 mg/g lung tissue appear to produce this effect in the Fischer 344 rat (Health Effects Institute,
22 1995). Morrow (1988) suggested that the inability of particle-laden AMs to translocate to the
23 mucociliary escalator is correlated to an average composite particle volume per AM in the lung.
24 When this particle volume exceeds approximately 60 Aim3 per AM in the Fischer 344 rat,
25 impairment of clearance appears to be initiated. When the particulate volume exceeds
26 approximately 600 //m3 per cell, evidence suggests that AM-mediated particulate clearance virtually
27 ceases and agglomerated particle-laden macrophages remain in the alveolar region and increasingly
28 nonphagocytized dust particles translocate to the pulmonary interstitium. Data for other laboratory
29 animal species and humans are, unfortunately, limited.
30 Several laboratory animal studies have indicated that exposure to DPM can reduce an animal's
31 resistance to respiratory infections. This effect, which can occur even after only 2 or 6 h of
32 exposure to diesel exhaust containing 5 to 8 mg/m3 DPM, does not appear to be caused by direct
33 impairment of the lymphoid or splenic immune systems; however, in one study of influenza virus
34 infection, interferon levels and hemagglutinin antibody levels were adversely affected in the
35 exposed mice. Studies on the effects of exposure to diesel exhaust or DPM on the immune system
11/5/99 5-93 DRAFT—DO NOT CITE OR QUOTE
-------
1 of laboratory animals have produced equivocal results (Table 5-8).
2
3 5.6.2.5. Neurological and Behavioral Effects
4 Behavioral effects have been observed in rats exposed to diesel exhaust from birth to 28 days
5 of age (Table 5-14). Exposure caused a decreased level of spontaneous locomotor activity and a
6 detrimental effect on learning in adulthood. In agreement with the behavioral changes was
7 physiological evidence for delayed neuronal maturation. Exposures were to 6 mg/m3 DPM for 8
8 h/day, 7 days/week from birth to about 7,14,21, or 28 days cf age.
9
10 5.6.2.6. Effects on Immunity and Allergenicity
11 Several laboratory animal studies have indicated that exposure to DPM can reduce an animal's
12 resistance to respiratory infection. This effect, which can occur even after only 2 or 6 hrs of
13 exposure to DE containing 5 to 8 mg/m3 DPM, does not appear to be caused by direct impairment
14 of the lymphoid or splenic immune systems; however, in one study of influenza virus infection,
15 interferon levels and hemaglutinin antibody levels were adversely affected in the exposed mice.
16 Studies on the effects of exposure to diesel exhaust or DPM on the immune system of laboratory
17 animals have produced equivocal results (Table 5-8).
18 As with humans, there are animal data suggesting that DPM is a possible factor in the
19 increasing incidence of allergic hypersensitivity. The effects have been demonstrated primarily in
20 acute human and laboratory animal studies and appear to be associated mainly with the organic
21 fraction of DPM. It also appears that synergies with DPM may increase the efficacy of known
22 allergens. Both animal and human cell culture studies suggest that DPM also has the potential to
23 act as an adjuvant.
24
25 5.6.2.7. Other Noncancerous Effects
26 Essentially no effects (based on the weight of evidence of a number of studies) were noted for
27 reproductive and teratogenic effects in mice. rats, rabbits, and monkeys: clinical chemistry and
28 hematology in the rat, cat, hamster, and monkeys; and enzyme induction in the rat and mouse
29 (Tables 5-11 through 5-13 and 5-15).
30
31 5.6.3. Comparison of Filtered and Unfiltered Diesel Exhaust
32 The comparison of the toxic responses in laboratory animals exposed to whole diesel exhaust
33 or filtered exhaust containing no particles demonstrates across laboratories that diesei particles are
*34, tVie «ri««"«ir«al <=ti/-»1r»«ri/> fttr(*nt rvF nrmoanr^rrmc Vi*»a1tK <=>fffvtc in 1aV»r»ratr»rv animals f»Y«r»c*»rl to Hif»c*»l
— - M_v f~& &.»*»..£*•« v«~._v^.._ »^_-._. v_ -_~_~.-__-~ —_ ~~~ „._ ___vv_ __w..^ „_„„-p.,, — — _ .
35 exhaust (Table 5-16). Whether the particles act additively or synergistically with the gases cannot
36 be determined from the designs of the studies. Under equivalent exposure regimens, hamsters have
1175/99 5-94 DRAFT—DO NOT CITE OR QUOTE
-------
1 lower levels of retained DPM in their lungs than rats and mice do and consequently less pulmonary
2 function impairment and pulmonary pathology. These differences may result from a lower intake
3._ rate of DPM, lower deposition rate and/or more rapid clearance rate, or lung tissue that is less
4 susceptible to the cytotoxicity of DPM. Observations of a decreased respiration in hamsters when
5 exposed by inhalation favor lower intake and deposition rates.
6
7 5.6.4. Interactive Effects of Diesel Exhaust
8 There is no direct evidence that diesel exhaust interacts with other substances in an exposure
9 environment, other than an impaired resistance to respiratory tract infections. Young animals were
10 not more susceptible. In several ways, animals with laboratory-induced emphysema were more
11 resistant. There is experimental evidence that both inorganic and organic compounds can be
12 adsorbed onto carbonaceous particles. When such substances become affiliated with particles, these
13 substances can be carried deeper into the lungs where they might have a more direct and potent
14 effect on epithelial cells or on AM ingesting the particles. Few specific studies to test interactive
15 effects of diesel exhaust with atmospheric contaminants, other than coal dust, have been conducted.
16 Coal dust and DPM had an additive effect only.
17
18 5.6.5. Conclusions
19 Conclusions concerning the principal human hazard from exposure to diesel emissions are as
follows:
21
22 • The primary acute (high-concentration, short-term) effects of DE in humans include
23 irritation, mild airway inflammation, and indicators of mild inflammation in lung lavage
24 fluids. Allergenic effects also have been demonstrated under short-term exposure
25 scenarios to either DE or DPM; the toxicological significance of these effects has yet to
26 be resolved.
27 • Noncancer effects in humans from long-term chronic exposure to DPM are not evident.
28 Noncancer effects from long-term exposure to DPM of several laboratory animal species
29 include pulmonary histopathology and inflammation.
30
31 Although the mode of action of DE/DPM is not clearly evident for any of the effects
32 documented in this chapter, the respiratory tract effects observed under acute scenarios are
33 suggestive of an irritant mechanism, while lung effects observed in chronic scenarios indicate an
34 underlying inflammatory response. Current knowledge indicates that the carbonaceous core of the
35 diesel particle is the causative agent of the lung effects, with the extent of the injury being mediated
1175/99 5-95 DRAFT—DO NOT CITE OR QUOTE
-------
1 at least in part by a progressive impairment of alveolar macrophages. It is noted that lung effects
2 occur in response to DPM exposure in several species and occur in rats at doses lower than those
3 inducing particle overload and a tumorigenic response (see above); it follows that lung effects such
4 as inflammation and fibrosis are relevant in the development of risk assessments for DPM.
5
6 5.7. REFERENCES
7 Ames, RG; Attfield, MD; Hankinson, JL; et al. (1982) Acute respiratory effects of exposure to diesel emissions in coal
8 miners. Am Rev Respir Dis 125:39-42.
9
10 Ames, RG; Reger, RB; Hall, DS. (1984) Chronic respiratory effects of exposure to diesel emissions in coal mines. Arch
11 Environ Health 39:389-394.
12 Attfield, MD. (1978) The effect of exposure to silica and diesel exhaust in underground metal and nonmetal miners. In:
13 Kelley, W. D., ed. Industrial hygiene for mining and tunneling: proceedings of a topical symposium; November;
14 Denver, CO. Cincinnati, OH: The American Conference of Governmental Industrial Hygienists, Inc.; pp. 129-135.
15 Attfield, MD; Trabant, GD; Wheeler, RW. (1982) Exposure to diesel fumes and dust at six potash mines. Ann Occup
16 Hyg 26:817-831.
17 Bamhart, MI; Chen, S-T; Salley, SO; et al. (1981) Ultrastructure and morphometry of the alveolar lung of guinea pigs
18 chronically exposed to diesel engine exhaust: six month's experience. J Appl Toxicol 1:88-103.
19 Barnhart, MI; Salley, SO; Chen, S-T; et al. (1982) Morphometric ultrastructural analysis of alveolar lungs of guinea
20 pigs chronically exposed by inhalation to diesel exhaust (DE). In: Lewtas, J., ed. Toxicological effects of emissions
21 from diesel engines: proceedings of the Environmental Protection Agency diesel emissions symposium; October, 1981;
22 Raleigh, NC. New York: Elsevier Biomedical; pp. 183-200. (Developments in toxicology and environmental science:
23 v. 10).
24 Battigelli, MC. (1965) Effects of diesel exhaust. Arch Environ Health 10:165-167.
25 Battigelli, MC; Mannella, RJ; Hatch, TF. (1964) Environmental and clinical investigation of workmen exposed to
26 diesel exhaust in railroad engine houses. Ind Med Surg 33:121-124.
27 Battigelli, MC; Hengstenberg, F: Mannela, RJ; et al. (1966) Mucociliary activity. Arch Environ Health 12:460-466.
28 Bayram, H; Devalia, JL; Sapsford, RJ; et al. (1998) The effect of diesel exhaust particles on cell function and release of
29 inflammatory mediators from human bronchial epithelial cells in vitro. Am J Respir Cell Mol Biol 18:441-448.
30 BeiuBe, KA; Jones, TP; Williamson, BJ; et al. (1999) Physicochemical characterisation of diesel exhaust particles:
31 factors for assessing biological activity. Atmos Environ 33:1599-1614.
32 Bhatnagar, RS; Hussain, MZ; Sorensen, KR; et al. (1980) Biochemical alterations in lung connective tissue in rats and
33 mice exposed to diesel emissions. In: Pepelko, WE; Danner, RM; Clarke, NA, eds. Health effects of diesel engine
34 emissions: proceedings of an international symposium, v. 1; December 1979; Cincinnati, OH. Cincinnati,
35 OH: U.S. Environmental Protection Agency, Health Effects Research Laboratory; pp. 557-570; EPA report no.
36 EPA-600/9-80-057a. Available from: NTIS, Springfield, VA; PB81-173809.
37 Bice, DE; Mauderly, JL; Jones, RK; et al. (1985) Effects of inhaled diesel exhaust on immune responses after lung
38 immunization. Fundam Appl Toxicol 5:1075-1086.
1175/99 5-96 DRAFT—DO NOT CITE OR QUOTE
-------
1 Blomberg, A; Sainsbury, C; Rudell, B; et al. (1998) Nasal cavity lining fluid ascorbic acid concentration increases in
2 healthy human volunteers following short term exposure to diesel exhaust. Free Radical Res 28:59-67.
3 Boland, S; Baeza-Squiban, A; Fournier, T; et al. (1999) Diesel exhaust particles are taken up by human airway
4 epithelial cells in vitro and alter cytokine production. Am J Physiol 276:L604-L613.
5 Bolton, RE; Vincent, JH; Jones, AD; et al. (1983) An overload hypothesis for pulmonary clearance of UICC amosite
6 fibres inhaled by rats. Br J Ind Med 40:264-272.
7 Boren, HG. (1964) Carbon as a carrier mechanism for irritant gases. Arch Environ Health 8:119-124.
8 Brightwell, J; Fouillet, X; Cassano-Zoppi, AL; et al. (1986) Neoplastic and functional changes in rodents after chronic
9 inhalation of engine exhaust emissions. In: Ishinishi, N; Koizumi, A; McClellan, RO; et al., eds. Carcinogenic and
10 mutagenic effects of diesel engine exhaust: proceedings of the international satellite symposium on toxicological effects
11 of emissions from diesel engines; July; Tsukuba Science City, Japan. Amsterdam, Holland: Elsevier Science Publishers
12 B. V.; pp. 471-485. (Developments in toxicology and environmental science: v. 13).
13 Campbell, KI; George, EL; Washington, IS, Jr. (1980) Enhanced susceptibility to infection in mice after exposure to
14 dilute exhaust from light duty diesel engines. In: Pepelko, WE; Danner, RM; Clarke, NA, eds. Health effects of diesel
15 engine emissions: proceedings of an international symposium, v. 2; December 1979; Cincinnati, OH. Cincinnati, OH:
16 U.S. Environmental Protection Agency, Health Effects Research Laboratory; pp. 772-785; EPA report no.
17 EPA/600/9-80-057b. Available from: NTIS, Springfield, VA; PB81-173817.
18 Campbell, KI; George, EL; Washington, IS, Jr. (1981) Enhanced susceptibility to infection in mice after exposure to
19 dilute exhaust from light duty diesel engines. Environ Int 5:377-382.
20 Castranova, V; Bowman, L; Reasor, MJ; et al. (1985) The response of rat alveolar macrophages to chronic inhalation of
11 coal dust and/or diesel exhaust. Environ Res 36:405-419.
J
22 Chan, TL; Lee, PS; Hering, WE. (1981) Deposition and clearance of inhaled diesel exhaust particles in the respiratory
23 tract of Fischer rats. J Appl Toxicol 1:77-82.
24 Chaudhari, A; Dutta, S. (1982) Alterations in tissue glutathione and angiotensin converting enzyme due to inhalation of
25 diesel engine exhaust. J Toxicol Environ Health 9:327-337.
26 Chaudhari, A; Farrer, RG; Dutta, S. (1980) Effect of exposure to diesel exhaust on pulmonary prostaglandin
27 dehydrogenase (PGDH) activity. In: Pepelko, WE; Danner, RM; Clarke, NA, eds. Health effects of diesel engine
28 emissions: proceedings of an international symposium, v. 1; December 1979; Cincinnati, OH. Cincinnati, OH: U.S.
29 Environmental Protection Agency, Health Effects Research Laboratory; pp. 532-537; EPA report no.
30 EPA/600/9-80-057a. Available from: NTIS, Springfield, VA; PB81-173809.
31 Chaudhari, A; Farrer, RG; Dutta, S. (1981) Effect of exposure to diesel exhaust on pulmonary prostaglandin
32 dehydrogenase (PGDH) activity. J Appl Toxicol 1:132-134.
33 Chen, KC. (1986) Induction of aryl hydrocarbon hydroxylase in rat tissue following intratracheal instillation of diesel
34 particulate extract and benzo[a]pyrehe. J Appl Toxicol 6:259-262.
35 Chen, KC; Vostal, JJ. (1981) Aryl hydrocarbon hydroxylase activity induced by injected diesel particulate extract vs
36 inhalation of diluted diesel exhaust. J Appl Toxicol 1:127-131.
37 Chen, S; Weller, MA; Barnhart, MI. (1980) Effects of diesel engine exhaust on pulmonary alveolar macrophages.
38 Scanning Electron Microsc 3:327-338.
11/5/99 5-97 DRAFT—DO NOT CITE OR QUOTE
-------
1 Devalia, JL; Bayram, H; Abdelaziz, MM; et al. (1999) Differences between cytokine release from bronchial epithelial
2 cells of asthmatic patients and non-asthmatic subjects: effect of exposure to diesel exhaust particles. Int Arch Allergy
3 Immunol 118:437-439.
4 Diaz-Sanchez, D. (1997) The role of diesel exhaust particles and their associated polyaromatic hydrocarbons in the
5 induction of allergic airway disease. Allergy (Copenhagen) 52(suppl. 38):52-56.
6 Diaz-Sanchez, D; Dotson, AR; Takenaka, H; et al. (1994) Diesel exhaust particles induce local IgE production in vivo
7 and alter the pattern of IgE messenger RNA isoforms. J Clin Invest 94:1417-1425.
8 Diaz-Sanchez, D; Tsien, A; Casillas, A; et al. (1996) Enhanced nasal cytokine production in human beings after in vivo
9 challenge with diesel exhaust particles. J Allergy Clin Immunol 98:114-123.
10 Diaz-Sanchez, D; Tsien, A; Fleming, J; et al. (1997) Combined diesel exhaust paniculate and ragweed allergen
11 challenge markedly enhances human in vivo nasal ragweed-specific IgE and skews cytokine production to a T helper
12 cell 2-type pattern. J Immunol 158:2406-2413.
13 Driscoll, KE; Maurer, JK. (1991) Cytokine and growth factor release by alveolar macrophages: potential biomarkers of
14 pulmonary toxicity. Toxicol Pathol 19:398-405.
15 Dungworth, DL; Mohr, U; Heinrich, U; et al. (1994) Pathologic effects of inhaled particles in rat lungs: associations
16 between inflammatory and neoplastic processes. In: Mohr, U; Dungworth, DL; Mauderly, JL, et al., eds. Toxic and
17 carcinogenic effects of solid particles in the respiratory tract: [proceedings of the 4th international inhalation
18 symposium]; March 1993; Hannover, Germany. Washington, DC: International Life Sciences Institute Press; pp.
19 75-98.
20 Dziedzic, D, (1981) Differential counts of B and T lymphocytes in the lymph nodes, circulating blood and spleen after
21 inhalation of high concentrations of diesel exhaust. J Appl Toxicol 1:111-115.
22 Edling, C; Axelson, O. (1984) Risk factors of coronary heart disease among personnel in a bus company. Int Arch
23 Occup Environ Health 54:181-183.
24 Edling, C; Anjou, CG; Axelson, O; et al. (1987) Mortality among personnel exposed to diesel exhaust. Int Arch Occup
25 Environ Health 59:559-565.
26 El Batawi, MA; Noweir, MH. (1966) Health problems resulting from prolonged exposure to air pollution in diesel bus
27 garages. Ind Health 4:1-10.
28 Fedan, JS; Frazier, DG; Moorman, WJ; etal. (1985) Effects of a two-year inhalation exposure of rats to coal dust
29 and/or diesei exhaust on tension responses of isolated airway smooth muscle. Am Rev Respir Dis 131:651-655.
30 Fujimaki, H; Saneyoshi, K; Shiraishi, F; et al. (1997) Inhalation of diesel exhaust enhances antigen-specific IgE
31 antibody production in mice. Toxicology 116:227-233.
32 Gamble, JF;Jones, WG. (1983) Respiratory effects of diesel exhaust in salt miners. Am Rev Respir Dis 128:389-394;
33 Gamble, J; Jones, W; Hudak, J; et al. (1979) Acute changes in pulmonary function in salt miners. Presented at:
34 Industrial hygiene for mining and tunneling: proceedings of a topical symposium; November 1978; Denver, CO.
35 Cincinnati, OK: American Conference of Governmental Industrial Hygienists, Inc.; pp. i 19-12S.
36 Gamble, J; Jones, W; Hudak, J. (1983) An epidemiological study of salt miners in diesel and nondiesel mines. Am J Ind
37 Med 4:435-458.
11/5/99 5-98 DRAFT—DO NOT CITE OR QUOTE
-------
1 Gamble, J; Jones, W; Minshall, S. (1987a) Epidemiological-environmental study of diesel bus garage workers: acute
2 effects of NO2 and respirable paniculate on the respiratory system. Environ Res 42:201-214.
3 Gamble, J; Jones, W; Minshall, S. (1987b) Epidemiological-environmental study of diesel bus garage workers: chronic
4 effects of diesel exhaust on the respiratory system. Environ Res 44:6-17.
5 Green, FHY; Boyd, RL; Danner-Rabovsky, J; et al. (1983) Inhalation studies of diesel exhaust and coal dust in rats.
6 Scand J Work Environ Health 9:181-188.
7 Griffis, LC; Wolff, RK; Henderson, RF; et al. (1983) Clearance of diesel soot particles from rat lung after a subchronic
8 diesel exhaust exposure. Fundam Appl Toxicol 3:99-103.
9 Gross, KB. (1981) Pulmonary function testing of animals chronically exposed to diluted diesel exhaust. J Appl Toxicol
10 1:116-123.
11 Hahon, N; Booth, JA; Green, F; et al. (1985) Influenza virus infection in mice after exposure to coal dust and diesel
12 engine emissions. Environ Res 37:44-60.
13 Hare, CT; Springer, KJ. (1971) Public response to diesel engine exhaust odors [final report]. San Antonio, TX:
14 Southwest Research Institute; report no. AR-804. Available from: NTIS, Springfield, VA; PB-204012.
15 Hare, CT; Springer, KJ; Somers, JH; et al. (1974) Public opinion of diesel odor. In: Automotive engineering congress;
16 February-March; Detroit, MI. New York: Society of Automotive Engineers; SAE technical paper no. 740214.
17 Hatch, GE; Boykin, E; Graham, JA; et al. (1985) Inhalable particles and pulmonary host defense: in vivo and in vitro
18 effects of ambient air and combustion particles. Environ Res 36:67-80.
19 Health Effects Institute. (1995) Diesel exhaust: a critical analysis of emissions, exposure, and health effects: a special
20 report of the Institute's Diesel Working Group. Cambridge, MA: Health Effects Institute.
M Heinrich, U; Peters, L; Funcke, W; et al. (1982) Investigation of toxic and carcinogenic effects of diesel exhaust in
22 long-term inhalation exposure of rodents. In: Lewtas, J., ed. Toxicological effects of emissions from diesel engines:
23 proceedings of the Environmental Protection Agency diesel emissions symposium; October 1981; Raleigh, NC. New
24 York: Elsevier Biomedical; pp. 225-242. (Developments in toxicology and environmental science: v. 10).
25 Heinrich, U; Muhle, H; Takenaka, S; et al. (1986a) Chronic effects on the respiratory tract of hamsters, mice, and rats
26 after long-term inhalation of high concentrations of filtered and unfiltered diesel engine emissions. J Appl Toxicol
27 6:383-395.
28 Heinrich, U; Pott, F; Rittinghausen, S. (1986b) Comparison of chronic inhalation effects in rodents after long-term
29 exposure to either coal oven flue gas mixed with pyrolized pitch or diesel engine exhaust. In: Ishinishi, N; Koizumi, A;
30 McClellan, RO; et al., eds. Carcinogenic and mutagenic effects of diesel engine exhaust: proceedings of the
31 international satellite syposium on toxicological effects of emissions from diesel engines; July; Tsukuba Science City,
32 Japan. Amsterdam, Holland: Elsevier Science Publishers BV; pp. 441-457. (Developments in toxicology and
33 environmental science: v. 13).
34 Heinrich, U; Fuhst, R; Rittinghausen, S; et al. (1995) Chronic inhalation exposure of Wistar rats and two different
35 strains of mice to diesel engine exhaust, carbon black, and titanium dioxide. Inhalation Toxicol 7:533-556.
36 Henderson, RF; Pickrell, JA; Jones, RK; et al. (1988) Response of rodents to inhaled diluted diesel exhaust:
37 biochemical and cytological changes in bronchoalveolar lavage fluid and in lung tissue. Fundam Appl Toxicol
38 11:546-567.
1175/99 5-99 DRAFT—DO NOT CITE OR QUOTE
-------
1 Hyde, DM; Plopper, CG; Weir, AJ; et al. (1985) Peribronchiolar fibrosis in lungs of cats chronically exposed to diesel
2 exhaust. Lab Invest 52:195-206.
3 Ikeda, M; Suzuki, M; Watarai, K; et al. (1995) Impairment of endothelium-dependent relaxation by diesel exhaust
4 particles in rat thoracic aorta. Jpn J Pharmacol 68:183-189.
5 Ikeda, M; Watarai, K; Suzuki, M; et al. (1998) Mechanism of pathophysiological effects of diesel exhaust particles on
6 endothelial cells. Environ Toxicol Pharmacol 6:117-123.
7 Iwai, K; Udagawa, T; Yamagishi, M; et al. (1986) Long-term inhalation studies of diesel exhaust on F344 SPF rats.
8 Incidence of lung cancer and lymphoma. In: Ishinishi, N; Koizumi, A; McClellan, RO; et al., eds. Carcinogenic and
9 mutagenic effects of diesel engine exhaust: proceedings of the international satellite symposium on toxicological effects
10 of emissions from diesel engines; July; Tsukuba Science City, Japan. Amsterdam, Holland: Elsevier Science Publishers
11 B. V.; pp. 349-360. (Developments in toxicology and environmental science: v. 13).
12 Jorgensen, H; Svensson, A. (1970) Studies on pulmonary function and respiratory tract symptoms of workers in an iron
13 ore mine where diesel trucks are used underground. J Occup Med 12:348-354.
14 Kahn, G; Orris, P; Weeks, J. (1988) Acute overexposure to diesel exhaust: report of 13 cases. Am J Ind Med
15 13:405-406.
16 Kaplan, HL; MacKenzie, WF; Springer, KJ; et al. (1982) A subchronic study of the effects of exposure of three species
17 of rodents to diesel exhaust. In: Lewtas, J., ed. Toxicological effects of emissions from diesel engines: proceedings of
18 the Environmental Protection Agency diesel emission symposium; October, 1981; Raleigh, NC. New York: Elsevier
19 Biomedical; pp. 161 -182. (Developments in toxicology and environmental science: v. 10).
20 Karagianes, MT; Palmer, RF; Busch, RH. (1981) Effects of inhaled diesel emissions and coal dust in rats. Am Ind Hyg
21 AssocJ 42:382-391.
22 Katz, M; Rennie, RP; Jegier, Z. (1960) Air pollution and associated health hazards from diesel locomotive traffic in a
23 railroad tunnel. Occup Health Rev 11:2-15.
24 Kilbum, KH; McKenzie, WN. (1978) Leukocyte recruitment to airways by aldehyde-carbon combinations that mimic
25 cigarette smoke. Lab Invest 38:134-142.
26 Klosterkotter, W; Bunemann, G. (1961) Animal experiments on the elimination of inhaled dust. In: Davies, C. N., ed.
27 Inhaled particles and vapours: proceedings of an international symposium; March-April 1960; Oxford, United
28 Kingdom. New York, NY: Pergamon Press; pp. 327-341.
29 Knox, RB; Suphioglu, C; Taylor, P; et al. (1997) Major grass pollen allergen Lol p 1 binds to diesel exhaust particles:
30 implications for asthma and air pollution. Clin Exp Allergy 27:246-251.
31 Kobayashi, T; Ito, T. (1995) Diesel exhaust particulates induce nasal mucosal hyperresponsiveness to inhaled histamine
32 aerosol. Fundam Appl Toxicol 27:195-202.
33 Kobayashi, T; Ikeue, T; Ito, T; et al. (1997) Short-term exposure to diesel exhaust induces nasal mucosal
34 hyperresponsiveness to histamine in guinea pigs. Fundam Appl Toxicol 38:166-172.
35 Laurie, RD; Boyes, WK. (1980) Neurophysiological alterations due to diesel exhaust exposure during the neonatal life
36 of the rat. In: Pepelko, WE; Danner, RM; Clarke, NA, eds. Health effects of diesel engine emissions: proceedings of an
37 international symposium, v. 2; December 1979; Cincinnati, OH. Cincinnati, OH: U.S. Environmental Protection
38 Agency, Health Effects Research Laboratory; pp. 713-727; EPA report no. EPA-600/9-80-057b. Available from: NTIS,
39 Springfield, VA;PB81-173817.
40 Laurie, RD; Boyes, WK. (1981) Neurophysiological alterations due to diesel exhaust exposure during the neonatal life
11/5/99 5-100 DRAFT—DO NOT CITE OR QUOTE
-------
1 of the rat. Environ Int 5:363-368.
2 Laurie, RD; Lewkowski, JP; Cooper, GP; et al. (1978) Effects of diesel exhaust on behavior of the rat. Presented at:
3 71st annual meeting of the Air Pollution Control Asscociation; June; Houston, TX. Pittsburgh, PA: Air Pollution
4 Control Association.
5 Laurie, RD; Boyes, WK; Wessendarp, T. (1980) Behavioral alterations due to diesel exhaust exposure. In: Pepelko,
6 WE; Danner, RM; Clarke, NA, eds. Health effects of diesel engine emissions: proceedings of an international
7 symposium, v. 2; December 1979; Cincinnati, OH. Cincinnati, OH: U.S. Environmental Protection Agency, Health
8 Effects Research Laboratory; pp. 698-712; EPA report no. EPA-600/9-80-057b. Available from: NTIS, Springfield,
9 VA;PB81-173817.
10 Lee, IP; Suzuki, K; Lee, SD; et al. (1980) Aryl hydrocarbon hydroxylase induction in rat lung, liver, and male
11 reproductive organs following inhalation exposure to diesel emission. Toxicol Appl Pharmacol 52:181-184.
12 Lee, KP; Trochimowicz, HJ; Reinhardt, CF. (1985) Pulmonary response of rats exposed to titanium dioxide (TiO2) by
13 inhalation for two years. Toxicol Appl Pharmacol 79:179-192.
14 Lee, KP; Kelly, DP; Schneider, PW; et al. (1986) Inhalation toxicity study on rats exposed to titanium tetrachloride
1 5 atmospheric hydrolysis products for two years. Toxicol Appl Pharmacol 83:30-45.
16 Lee, KP; Kelly, DP; O'Neal, FO; et al. (1988) Lung response to ultrafine Kevlar aramid synthetic fibrils following
17 2-year inhalation exposure in rats. Fundam Appl Toxicol 11:1-20.
18 Levins, PL. (1981) Review of diesel odor and toxic vapor emissions. Washington, DC: U.S. Department of
19 Transportation, National Highway Traffic Safety Administration; report no. DOT-TSC-NHTSA-81-9.
20 Lewis, TR; Green, FHY; Moorman, WJ; et al. (1989) A chronic inhalation toxicity study of diesel engine emissions and
,21 coal dust, alone and combined. J Am Coll Toxicol 8:345-375.
22 Linnell, RH; Scott, WE. (1962) Diesel exhaust composition and odor studies. J Air Pollut Control Assoc 12:510-515,
23 545.
24 L0vik, M; Hegseth, AK; Gaarder, PI; et al. (1997) Diesel exhaust particles and carbon black have adjuvant activity on
25 the local lymph node response and systemic IgE production to ovalbumin. Toxicology 121:165-178.
26 Maejima, K; Tamura, K; Taniguchi, Y; et al. (1997) Comparison of the effects of various fine particles on IgE antibody
27 production in mice inhaling Japanese cedar pollen allergens. J Toxicol Environ Health 52:231-248.
28 Mauderly, JL. (1994) Noncancer pulmonary effects of chronic inhalation exposure of animals to solid particles. In:
29 Mohr, U; Dungworth, DL; Mauderly, JL; et al., eds. Toxic and carcinogenic effects of solid particles in the respiratory
30 tract: [proceedings of the 4th international inhalation symposium]; March 1993; Hannover, Germany. Washington, DC:
31 International Life Sciences Institute Press; pp. 43-55.
32 Mauderly, JL; Benson, JM; Bice, DE; et al. (1981) Observations on rodents exposed for 19 weeks to diluted diesel
33 exhaust. In: Inhalation Toxicology Research Institute annual report 1980-1981. Albuquerque, NM: Lovelace
34 Biomedical and Environmental Research Institute; pp. 305-311; report no. LMF-91.
35 Mauderly, JL; Benson, JM; Rice, DE; et al. (1984) Life span study of rodents inhaling diesel exhaust: effects on body
36 weight and survival. In: Guilmette, RA; Medinsky, MA, eds. Inhalation Toxicology Research Institute annual report.
37 Albuquerque, NM: Inhalation Toxicology Research Institute; pp. 287-291; report no. LMF-113. Available from: NTIS,
38 Springfield, VA; DE85-009283.
11 /5/99 5-101 DRAFT—DO NOT CITE OR QUOTE
-------
1 Mauderly, JL; Bice, DE; Carpenter, RL; et al. (1987a) Effects of inhaled nitrogen dioxide and diesel exhaust on
2 developing lung. Cambridge, MA: Health Effects Institute; research report no. 8.
3 Mauderly, JL; Jones, RK; Griffith, WC; et al. (1987b) Diesel exhaust is a pulmonary carcinogen in rats exposed
4 chronically by inhalation. Fundam Appl Toxicol 9:208-221.
5 Mauderly, JL; Gillett, NA; Henderson, RF; et al. (1988) Relationships of lung structural and functional changes to
6 accumulation of diesel exhaust particles. In: Dodgson, J; McCallum, RI; Bailey, MR; et al., eds. Inhaled particles VI:
7 proceedings of an international symposium and workshop on lung dosimetry; September 1985; Cambridge, United
8 Kingdom. Ann Occup Hyg 32(suppl. l):659-669.
9 Mauderly, JL; Bice, DE; Cheng, YS; et al. (1990a) Influence of preexisting pulmonary emphysema on susceptibility of
10 rats to diesel exhaust. Am Rev Respir Dis 141:1333-1341.
11 Mauderly, JL; Cheng, YS; Snipes, MB. (1990b) Particle overload in toxicology studies: friend or foe? In: Proceedings
12 of the particle lung interaction symposium; May; Rochester, NY. J Aerosol Med 3(suppl. 1): s-169-s-187.
13
14 Mauderly, JL; Banas, DA; Griffith, WC; et al. (1996) Diesel exhaust is not a pulmonary carcinogen in CD-I mice
15 exposed under conditions carcinogenic to F344 rats. Fundam Appl Toxicol 30:233-242.
16 McClellan, RO; Bice, DE; Cuddihy, RG; et al. (1986) Health effects of diesel exhaust. In: Lee, SD; Schneider, T;
17 Grant, LD; et al., eds. Aerosols: research, risk assessment and control strategies: proceedings of the second U.S.-Dutch
18 international symposium; May 1985; Williamsburg, VA. Chelsea, MI: Lewis Publishers, Inc.; pp. 597-615.
19 Meiss, R; Robenek, H; Schubert, M; et al. (1981) Ultrastructural alterations in the livers of golden hamsters following
20 experimental chronic inhalation of diluted diesel exhaust emission. Int Arch Occup Environ Health 48:147-157.
21 Mentnech, MS; Lewis, DM; Olenchock, SA; et al. (1984) Effects of coal dust and diesel exhaust on immune
22 competence in rats. J Toxicol Environ Health 13:31-41.
23 Misiorowski, RL; Strom, KA; Vostal, JJ; et al. (1980) Lung biochemistry of rats chronically exposed to diesel
24 particulates. In: Pepelko, WE; Danner, RM; Clarke, NA, eds. Health effects of diesel engine emissions: proceedings of
25 an international symposium, v. 1; December 1979. Cincinnati, OH. Cincinnati, OH: U.S. Environmental Protection
26 Agency, Health Effects Research Laboratory; pp. 465-480; EPA report no. EPA/600/9-80-057a. Available from: NTIS,
27 Springfield, VA;PB81-173809.
28 Miyabara, Y; Ichinose, T; Takano, H; et al. (1998a) Diesel exhaust inhalation enhances airway hyperresponsiveness in
29 mice. Int Arch Allergy Immunol 116:124-131.
30 Miyabara, Y; Takano, H; Ichinose, T; et ai. (i998b) Diesel exhaust enhances allergic airway inflammation and
31 hyperresponsiveness in mice. Am J Resp Crit Care Med 157:1138-1144.
32 Moorman, WJ; Clark, JC; Pepelko, WE; et al. (1985) Pulmonary fuction responses in cats following long-term
33 exposure to diesel exhaust. J Appl Toxicol 5:301-305.
34 Morrow, PE. (1988) Possible mechanisms to explain dust overloading of the lungs. Fundam Appl Toxicol 10:369-384.
35 Murphy, SA; BeruBe", KA; Pooley, FD; et al. (1998) The response of lung epithelium to well characterised fine
38 particles. Life Sci 62.1789-1799.
37 Navarro, C; Charboneau, J; McCauley, R. (1981) The effect of in vivo exposure to diesel exhaust on rat hepatic and
38 pulmonary microsomal activities. J Appl Toxicol 1:124-126.
11/5/99 5-102 DRAFT—DO NOT CITE OR QUOTE
-------
1 Nikula, K.J; Snipes, MB; Barr, EB; et al. (1995) Comparative pulmonary toxicities and carcinogenicities of chronically
2 inhaled diesel exhaust and carbon black in F344 rats. Fundam Appl Toxicol 25:80-94.
3 Nikula, K.J; Avila, KJ; Griffith, WC; et al. (1997) Sites of particle retention and lung tissue responses to chronically
4 inhaled diesel exhaust and coal dust in rats and cynomolgus monkeys. In: Driscoll, KE; Oberdorster, G, eds.
5 Proceedings of the sixth international meeting on the toxicology of natural and man-made fibrous and non-fibrous
6 particles; September 1996; Lake Placid, NY. Environ Health Perspect Suppl 105(5): 1231-1234.
7 Nilsen, A; Hagemann, R; Eide, I. (1997) The adjunct activity of diesel exhaust particles and carbon black on systemic
8 IgE production to ovalbumin in mice after intranasal instillation. Toxicology 124:225-232.
9 Oberdorster, G. (1994) Extrapolation of results from animal inhalation studies with particles to humans? In: Mohr, U;
10 Dungworth, DL; Mauderly, JL; et al., eds. Toxic and carcinogenic effects of solid particles in the respiratory tract:
11 [proceedings of the 4th international inhalation symposium]; March 1993; Hannover, Germany. Washington, DC:
12 International Life Sciences Institute Press; pp. 335-353.
13 Oberdorster, G; Ferin, J; Gelein, R; et al. (1992) Role of the alveolar macrophage in lung injury: studies with ultrafine
14 particles. Environ Health Perspect 97:193-199.
15 Odor Panel of the CRC-APRAC Program Group on Composition of Diesel Exhaust. (1979) Development and
16 evaluation of a method for measurement of diesel exhaust odor using a diesel odor analysis system (DOAS). New
17 York, NY: Coordinating Research Council, Inc., Air Pollution Research Advisory Committee; CRC-APRAC project
18 no. CAPI-1-64.
19 Ohtoshi, T; Takizawa, H; Okazaki, H; et al. (1998) Diesel exhaust particles stimulate human airway epithelial cells to
20 produce cytokines relevant to airway inflammation in vitro. J Allergy Clin Immunol 101:778-785.
21 Ormstad, H; Johansen, BV; Gaarder, PI. (1998) Airborne house dust particles and diesel exhaust particles as allergen
'2 carriers. Clin Exp Allergy 28:702-708.
23 Pattle, RE; Stretch, H; Burgess, F; et al. (1957) The toxicity of fumes from a diesel engine under four different running
24 conditions. Br J Ind Med 14:47-55.
25 Penney, DG; Baylerian, MS; Fanning, KE; et al. (1981) A study of heart and blood of rodents inhaling diesel engine
26 exhaust particulates. Environ Res 26:453-462.
27 Pepelko, WE. (1982a) Effects of 28 days exposure to diesel engine emissions in rats. Environ Res 27:16-23.
28 Pepelko, WE. (1982b) EPA studies on the toxicological effects of inhaled diesel engine emissions. In: Lewtas, J, ed.
29 Toxicological effects of emissions from diesel engines: proceedings of the Environmental Protection Agency diesel
30 emissions symposium; October 1981; Raleigh, NC. New York, NY: Elsevier Biomedical; pp. 121-142. (Developments
31 in toxicology and environmental science: v. 10).
32 Pepelko, WE; Peirano, WB. (1983) Health effects of exposure to diesel engine emissions: a summary of animal studies
33 conducted by the U.S. Environmental Protection Agency's Health Effects Research Laboratories at Cincinnati, Ohio. J
34 Am Coll Toxicol 2:253-306.
35 Pepelko, WE; Mattox, JK; Yang, YY; et al. (1980a) Pulmonary function and pathology in cats exposed 28 days to
36 diesel exhaust. J Environ Pathol Toxicol 4:449-458.
37 Pepelko, WE; Mattox, J; Moorman, WJ; et al. (1980b) Pulmonary function evaluation of cats after one year of exposure
38 to diesel exhaust. In: Pepelko, WE; Danner, RM; Clarke, NA, eds. Health effects of diesel engine emissions:
39 proceedings of an international symposium, v. 2; December 1979; Cincinnati, OH. Cincinnati, OH: U.S. Environmental
1175/99 5-103 DRAFT—DO NOT CITE OR QUOTE
-------
1 Protection Agency, Health Effects Research Laboratory; pp. 757-765; EPA report no. EPA/600/9-80-057b. Available
2 from: NTIS, Springfield, VA; PB81-173817.
3 Pepelko, WE; Mattox, J; Moorman, WJ;etal. (1981) Pulmonary function evaluation of cats after one year of exposure
4 to diesel exhaust. Environ Int 5:373-376.
5 Pereira, MA; Sabharwal, PS; Gordon, L; et al. (1981) The effect of diesel exhaust on sperm-shape abnormalities in
6 mice. Environ Int 5:459-460.
7 Plopper, CG; Hyde, DM; Weir, AJ. (1983) Centriacinar alterations in lungs of cats chronically exposed to diesel
8 exhaust. Lab Invest 49:391-399.
9 Purdham, JT; Holness, DL; Pilger, CW. (1987) Environmental and medical assessment of stevedores employed in ferry
10 operations. Appl Ind Hyg 2:133-139.
11 Quinto, I; De Marinis, E. (1984) Sperm abnormalities in mice exposed to diesel paniculate. Mutat Res 130:242.
12 Rabovsky, J; Petersen, MR; Lewis, TR; et al. (1984) Chronic inhalation of diesel exhaust and coal dust: effect of age
13 and exposure on selected enzyme activities associated with microsomal cytochrome P-450 in rat lung and liver. J
14 Toxicol Environ Health 14:655-666.
15 Rabovsky, J; Judy, DJ; Rodak, DJ; et al. (1986) Influenza virus-induced alterations of cytochrome P-450 enzyme
16 activities following exposure of mice to coal and diesel particulates. Environ Res 40:13 6-144.
17 Reger, R. (1979) Ventilatory function changes over a work shift for coal miners exposed to diesel emissions. In:
18 Bridbord, K; French, J, eds. Toxicological and carcinogenic health hazards in the workplace: proceedings of the first
19 annual NIOSH scientific symposium; 1978; Cincinnati, OH. Park Forest South, IL: Pathotox Publishers, Inc.; pp.
20 346-347.
21 Reger, R; Hancock, J; Hankinson, J; et al. (1982) Coal miners exposed to diesel exhaust emissions. Ann Occup Hyg
22 26:799-815.
23 Research Committee for HERP Studies. (1988) Diesel exhaust and health risks: results of the HERP studies. Tsukuba,
24 Ibaraki, Japan: Japan Automobile Research Institute, Inc.
25 Rudell, B; Sandstrom, T; Stjemberg, N; et al. (1990) Controlled diesel exhaust exposure in an exposure chamber:
26 pulmonary effects investigated with bronchoalveolar lavage. J Aerosol Sci 21(suppl. I):s411-S414.
27 Rudell, B; Sandstrom, T; Hammarstrom, U; et al. (1994) Evaluation of an exposure setup for studying effects of diesel
28 exhaust in humans. Int Arch Occup Environ Health 66:77-83.
29 Rudell, B; Ledin, MC; Hammarstrom, U; et al. (1996) Effects on symptoms and lung function in humans
30 experimentally exposed to diesel exhaust. Occup Environ Med 53:658-662.
31 Sagai, M; Furuyama, A; Ichinose, T. (1996) Biological effects of diesel exhaust particles (DEP). III. Pathogenesis of
32 asthma like symptoms in mice. Free RadicBiol Med 21:199-209.
33 Salvi, S; Blomberg, A; Rudell, B; et al. (1999) Acute inflammatory responses in the airways and peripheral blood after
34 short-term exposure to diesel exhaust in healthy human volunteers. Am J Respir Crit Care Med 159:702-709.
35 Schneider, DR; Felt, BT. (1981) Effect of diesel particulate exposure on adenylate and guanylate cyclase of rat and
36 guinea pig liver and lung. J Appl Toxicol 1:135-139.
37 Schreck, RM; Soderholm, SC; Chan, TL; et al. (1981) Experimental conditions in GMR chronic inhalation studies of
1175/99 5-104 DRAFT—DO NOT CITE OR QUOTE
-------
1 diesel exhaust. J Appl Toxicol 1:67-76.
2 Steerenberg, PA; Zonnenberg, JAJ; Dormans, JAMA; et al. (1998) Diesel exhaust particles induced release of
3 interleukin 6 and 8 by (primed) human bronchial epithelial cells (BEAS 2B) in vitro. Exp Lung Res 24:85-100.
4 Stober, W. (1986) Experimental induction of tumors in hamsters, mice and rats after long-term inhalation of filtered
5 and unfiltered diesel engine exhaust. In: Ishinishi, N; Koizumi, A; McClellan, RO; et al. Carcinogenic and mutagenic
6 effects of diesel engine exhaust: proceedings of the international satellite symposium on toxicological effects of
7 emissions from diesel engines; July; Tsukuba Science City, Japan. Amsterdam, Elsevier Science Publishers B. V.; pp.
8 421-439. (Developments in toxicology and environmental science: v. 13).
9 Strom, KA. (1984) Response of pulmonary cellular defenses to the inhalation of high concentrations of diesel exhaust.
10 J Toxicol Environ Health 13:919-944.
11 Suzuki, T; Kanoh, T; Ishimori, M; et al. (1996) Adjuvant activity of diesel exhaust particulates (DEP) in production of
12 anti-IgE and anti-IgG 1 antibodies to mite allergen in mice. J Clin Lab Immunol 48:187-199.
13 Takafuji, S; Suzuki, S; Koizumi, K; et al. (1987) Diesel-exhaust particulates inoculated by the intranasal route have an
14 adjuvant activity for IgE production in mice. J Allergy Clin Immunol 79:639-645.
15 Takano, H; Yoshikawa, T; Ichinose, T; et al. (1997) Diesel exhaust particles enhance antigen-induced airway
16 inflammation and local cytokine expression in mice. Am J Respir Crit Care Med 156:36-42.
17 Takano, H; Ichinose, T; Miyabara, Y; et al. (1998a) Inhalation of diesel exhaust enhances allergen-related eosinophil
18 recruitment and airway hyperresponsiveness in mice. Toxicol Appl Pharmacol 150:328-337.
19 Takano, H; Ichinose, T; Miyabara, Y; et al. (1998b) Diesel exhaust particles enhance airway responsiveness following
20 allergen exposure in mice. Immunopharmacol Immunotoxicol 20:329-336.
}l Takenaka, H; Zhang, K; Diaz-Sanchez, D; et al. (1995) Enhanced human IgE production results from exposure to the
"22 aromatic hydrocarbons from diesel exhaust: direct effects on B-cell IgE production. J Allergy Clin Immunol
23 95:103-115.
24 Turk, A. (1967) Selection and training of judges for sensory evaluation of the intensity and character of diesel exhaust
25 odors. Cincinnati, OH: U.S. Department of Health, Education, and Welfare, National Center for Air Pollution Control;
26 Public Health Service publication no. 999-AP-32. Available from: NTIS, Springfield, VA; PB-174707.
27 U.S. Environmental Protection Agency. (1996) Air quality criteria for paniculate matter. Research Triangle Park,'NC:
28 National Center for Environmental Assessment-RTP Office; report nos. EPA/600/P-95/001aF-cF. 3v. Available from:
29 NTIS, Springfield, VA; PB96-168224.
30 UlfVarson, U; Alexandersson, R; Aringer, L; et al. (1987) Effects of exposure to vehicle exhaust on health. Scand J
31 Work Environ Health 13:505-512.
32 Vallyathan, V; Virmani, R; Rochlani, S; et al. (1986) Effect of diesel emissions and coal dust inhalation on heart and
33 pulmonary arr four different running conditions. Br J Ind Med 14:47-55.
34 Vinegar, A; Carson, Al; Pepelko, W E. (1980) Pulmonary function changes in Chinese hamsters exposed six months to
35 diesel exhaust. In: Pepelko, W E; Danner, R M; Clarke, N A, eds. Health effects of diesel engine emissions:
36 proceedings of an international symposium, v. 2; December, 1979; Cincinnati, OH. Cincinnati, OH: U.S.
37 Environmental Protection Agency, Health Effects Research Laboratory; pp. 749-756; EPA report no.
38 EPA-600/9-80-057b. Available from: NTIS, Springfield, VA; PB81-173817.
Vinegar, A; Carson, A; Pepelko, WE; et al. (198 la) Effect of six months of exposure to two levels of diesel exhaust on
11/5/99 5-105 DRAFT—DO NOT CITE OR QUOTE
-------
1 pulmonary function of Chinese hamsters. Fed Proc 40:593.
2 Vinegar, A; Carson, A; Pepelko, WE. (1981b) Pulmonary function changes in Chinese hamsters exposed six months to
3 diesel exhaust. Environ Int 5:369-371.
4 Vostal, JJ; Chan, TL; Garg, BD; et al. (1981) Lymphatic transport of inhaled diesel particles in the lungs of rats and
5 guinea pigs exposed to diluted diesel exhaust. Environ Int 5:339-347.
6 Vostal, JJ; White, HJ; Strom, KA; et al. (1982) Response of the pulmonary defense system to diesel particulars
7 exposure. In: Lewtas, J., ed. Toxicological effects of emissions from diesel engines: proceedings of the Environmental
8 Protection Agency diesel emissions symposium; October 1981; Raleigh, NC. New York, NY: Elsevier Biomedical; pp.
9 201 -221. (Developments in toxicology and environmental science: v. 10).
10 Wade, JF, III; Newman, LS. (1993) Diesel asthma1: reactive airways disease following overexposure to locomotive
11 exhaust. J Occup Med 35:149-154.
12 Wallace, MA; Salley, SO; Barnhart, MI. (1987) Analysis of the effects of inhaled diesel exhaust on the alveolar
13 intravascular and interstitial cellular components of rodent lungs. Scanning Microsc 1:1387-1395.
14 Watanabe, N; Oonuki, Y. (1999) Inhalation of diesel engine engine exhaust affects spermatogenesis in growing male
1 5 rats. Environ Health Perspect 107:539-544.
16 Wehner, AP; Dagle, GE; Clark, ML; et al. (1986) Lung changes in rats following inhalation exposure to volcanic ash
17 for two years. Environ Res 40:499-517.
18 Werchowski, KM; Chaffee, VW; Briggs, GB. (1980a) Teratologic effects of long-term exposure to diesel exhaust
19 emissions (rats). Cincinnati, OH: U.S. Environmental Protection Agency, Health Effects Research Laboratory; EPA
20 report no. EPA-600/1-80-010. Available from: NTIS, Springfield, VA; PB80-159965.
21 Werchowski, KM; Henne, SP; Briggs, GB. (1980b) Teratologic effects of long-term exposure to diesel exhaust
22 emissions (rabbits). Cincinnati, OH: U.S. Environmental Protection Agency, Health Effects Research Laboratory; EPA
23 report no. EPA-600/1-80-011. Available from: NTIS, Springfield, VA; PB80-168529.
24 White, HJ; Garg, BD. (1981) Early pulmonary response of the rat lung to inhalation of high concentration of diesel
25 particles. J Appl Toxicol 1:104-110.
26 Wiester, MJ; Iltis, R; Moore, W. (1980) Altered function and histology in guinea pigs after inhalation of diesel exhaust.
27 Environ Res 22:285-297.
28 Wolff, RK; Gray, RL. (1980) Tracheal clearance of particles. In: Diel, JH; Bice, DE; Martinez, BS, eds. Inhalation
23 Toxicology Research Institute annual report: 1979-1980. Albuquerque, MM: Lcvehce Bicmedica! and Environmental
30 Research Institute; p. 252; report no. LMF-84.
31 Wolff, RK; Henderson, RF; Snipes, MB; et al. (1987) Alterations in particle accumulation and clearance in lungs of rats
32 chronically exposed to diesel exhaust. Fundam Appl Toxicol 9:154-166.
33 Yang, HM; Ma, JY; Castranova, V; et al. (1997) Effects of diesel exhaust particles on the release of interleukin-1 and
34 tumor necrosis factor-aipha from rat alveolar macrophages. Exp Lung Res 23:269-234.
1175/99 5-106 DRAFT—DO NOT CITE OR QUOTE
-------
6. NONCANCER DOSE-RESPONSE EVALUATION: RfC DERIVATION
1 6.1. INTRODUCTION—BACKGROUND OF THE INHALATION RfC AND ORAL
2 RfD
3 Construction of a risk assessment for a toxicant requires several steps, including synthesis
4 of information into a coherent reasonable evaluation of the hazard it presents to humans and
5 definition of the relationship between dose of the substance and the resultant biological response.
6 The EPA's vehicle for construction of these vital portions of a risk assessment, hazard
7 identification and dose-response, is the inhalation reference dose (RfD) for an orally ingested
8 toxicant or the inhalation reference concentration (RfC) for an inhaled airborne toxicant.
9 This chapter explains the concept and structure of the RfC as the Agency's estimate of a
10 "safe" level, and utilizes the information documented in Chapter 5 to synthesize this estimate for
11 diesel.
12
13 6.1.1. The Acceptable Daily Intake
14 Since its inception, EPA has advocated critical evaluation of data related to noncancer
15 toxicity of compounds. When possible, quantitative estimates were calculated from combining
16 effect levels, such as a no-observed-adverse-effect-level (NOAEL) or a lowest-observed-adverse-
17 effect-level (LOAEL), with certain "safety factors" into an Acceptable Daily Intake (ADI). Such
\ 8 procedures have a wide and historical basis; the National Research Council (NRC) recommended
19 the ADI approach in 1977 to characterize levels of pollutants in drinking water with respect to
20 human health (NRC, 1977, 1980). These approaches, as well as the oral reference dose (RfD)
21 and inhalation reference concentration (RfC) discussed below, are based on the assumption that a
22 threshold exists for the human population below which no effect will occur. Basically, all of
23 these approaches attempt to identify an estimate of a likely subthreshold concentration.
24
25 6.1.2. Oral RfD and Inhalation RfC—Dose-Response Assessments Inclusive of
26 Uncertainty Factors
27 The National Academy of Sciences (NAS) report entitled "Risk Assessment in the
28 Federal Government: Managing the Process" was issued in 1983 (NRC, 1983). Among the many
29 fundamental concepts and principles put forth in this report was the recommendation that
30 scientific aspects be explicitly separated from policy issues in the risk assessment process.
31 EPA's response included development of the RfD and guidelines on its derivation
32 (Barnes and Dourson, 1988) and subsequent development of the parallel inhalation RfC and its
33 formal methodology (U.S. EPA, 1994). The definition of the inhalation RfC is:
11/5/99 6-1 DRAFT—DO NOT CITE OR QUOTE
-------
1 An estimate (with uncertainty spanning perhaps an order of magnitude) of a
2 continuous inhalation exposure to the human population (including sensitive
3 subgroups) that is likely to be without an appreciable risk of deleterious noncancer
4 effects during a lifetime.
5
6 Similar to ADIs in intent, RfC/Ds are dose-response assessments for noncancer effects based
7 upon a more rigorous methodology adhering to the principles set forth in the 1983 NRC report.
8 The RfC methodology includes guidance on the consistent application to effect levels of
9 "uncertainty factors" (UFs) rather than the ADI "safety factors " for extrapolations.
10 The basic quantitative formula for derivation of an RfC, given in Equation 6-1, has as its
11 basic components an effect level and UFs. The units of an RfC are mg/m3.
12
13 RfC= NOAEL (6-1)
14 UF
15
16 The concept of an effect level, such as the NOAEL or LOAEL, is consistent with the ADI
17 construct. Alternatively, the benchmark dose/concentration (BMC) approach may be used as the
18 effect level in Equation 6-1. The BMC approach applies a line-fitting model to the key data and
19 then uses the dose-response relationship to interpolate an exposure concentration that is predicted
20 to result in a predefined level of response (BMR), such as a 10% incidence of a lesion. The
21 lower confidence limit on the concentration predicted to result in the BMR is designated the
22 BMC and would be the numerator in Equation 6-1.
23
24 6.1.3. UFs—Designation and Application
25 The UFs, their components, and their intended usage in the RfC methodology are given in
26 Table 6-1. As can be seen, they are fitted to the RfC definition providing consideration for
27 lifetime exposure (subchronic-to-chronic duration factor) for sensitive subgroups (human-to-
28 sensitive human factor) within the human population (animal-to-human extrapolation factor).
29 Consideration for effect levels (a LOAEL to a NOAEL extrapolation factor) and a database factor
30 are also part of the RfC methodology. The default values for the A and H UF are also shown
31 with their pharmacokinetic (PK) and pharmacodynamic (PD) components, each at 10° 53 which is
32 rounded to 3 when applied singly. The pharmacokinetic adjustments tc dose provided for in
33 derivation of RfCs (EPA, 1994) allow for application of only the PD component of this UF.
34 As with the safety factors for the ADI, UFs for the RfD/C are applied in a multiplicative
35 manner. Unlike safety factors, which are almost always applied to effect levels as even factors of
1175/99 6-2 DRAFT-DO NOT CITE OR QUOTE
-------
1 10, UFs may be applied to effect levels as partial values of 10, e.g., 10°5 (rounded to 3) or 1,
2 based on the circumstances. An example of application of a partial UF is for animal-to-human
3 extrapolation with dosimetry adjustments as explained below in Section 6.1.4.
4
5 6.1.4. Animal-to-Human Extrapolation Factor in the RfC—A Human Equivalent
6 Concentration
7 A major difference exists between the oral RfD and inhalation RfC in the animal-to-
8 human (A) extrapolation procedure. Table 6-1 indicates that the A UF may have the default
9 value of 10 and, furthermore, that this factor may be differentiated into pharmacokinetic (PK;
10 dose to tissue) and pharmacodynamic (PD; tissue response) components. Adjustments to the
11 externally applied factors may be made to address the PK component of this UF. The RfC
12 methodology (U.S. EPA, 1994) provides models and procedures for adjustments with both
13 particles and gases. In this assessment, several pharmacokinetic models, some capable of
14 adjusting for all aspects of the PK component such as absorption, uptake, and clearance, are
15 reviewed and evaluated. The goal of these adjustments is to derive an external concentration that
16 would produce the same internal tissue dose in humans as in animals, i.e., to produce a Human
17 Equivalent Concentration (HEC) from the animal effect level. When this adjustment is made,
18 the quantitative pharmacokinetics are considered the same and the PK component of this UF is
19 addressed. This adjustment for dosimetry is accommodated by application of a partial UF for
interspecies extrapolation of 10°5 for the remaining uncertainty about the PD component. When
21 applied singly this factor, by policy, is rounded to 3.
Table 6-1. UFs and their default values used in EPA's noncancer RfD and RfC
methodology
UF—Area of extrapolation Default values
A—animal-to-human 10 (10°5 PK x 10°5 PD)
H—human-to-sensitive human 10 (10°5 PK x 10°5 PD)
S—subchronic-to-chronic 10
L—LOAEL-to-NOAEL 10
D—incomplete-to-complete data 10
11/5/99 6-3 DRAFT—DO NOT CITE OR QUOTE
-------
1 6.1.5. Basic Procedures for Derivation of an RfC—Identification of the Critical Effect, the
2 Principal Study, Application of UF, and Assignment of Confidence Level
3 The goal of the RfC/D methodologies is to provide rationale and guidance on a
4 quantitative approach in evaluating toxicity data to derive a dose-response assessment.
5 Equation 6-1 is a condensation of the RfC process and serves as a basis for discussing the
6 procedures for its derivation. Having a NOAEL for this equation implies that a specific adverse
7 effect has been identified and that there is documentation that this effect does not occur at this
8 particular concentration, i.e., the NOAEL.
9 RfC derivation provides for evaluation of the toxicity database to identify a "critical
10 effect," which is defined as "the first adverse effect, or its known precursor, that occurs as the
11 dose rate increases." Analysis of the database also allows for choice of a "principal study," "the
12 study that contributes most significantly to the qualitative and quantitative risk," in characterizing
13 the dose-response of the critical effect. To fulfill the definition of the RfC, the critical effect
14 would have to be consonant with the definition of the RfC given above, e.g., relevant to humans
15 and observed under chronic, long-term conditions. Other studies that are pertinent to identifying
16 the dose-response or threshold for the effect are included in the derivation as supporting studies.
17 Thus, the NOAEL in Equation 6-1 would be based on the absence of the critical effect as
18 documented in the principal study.
19 Assignment of an appropriate UF would be accomplished in consideration of the
20 information available on the specific chemical as per Table 6-1. General guidelines were
21 discussed briefly in this introductory section and are discussed at length in the RfC Methods
22 (U.S. EPA, 1994). As explained above, assignment of specific values of UF may have both
23 policy and science implications. General policy is to provide clear explanatory text with each UF
24 assignment. Composite UF values vary widely. In cases where information on the NOAEL is
25 well defined in a known sensitive subgroup of humans, the UF may be 1. With sparse
26 information, UF values have ranged up to 3000. If none of the areas of extrapolation in Table
27 6-1 are addressed (i.e., all areas of uncertainty are applicable), then no RfC is derived.
28 Confidence statements are synthesized for each RfC. They are meant to serve as a
29 repository for statements that clearly communicate associated uncertainties, establish and
30 dichotomize policy from scientific bases, make clear specific limitations and strengths, and
31 express any other concerns reflecting on the overall quality of the assessment (U.S. EPA, 1994;
32 Ohanian et al., 1997). The RfC/D methodologies allow for high, medium, and low levels of
33 confidence, with the level being assigned subsequent to an analysis as above. Levels are
34 surmised for both the overall database and the principal study/ies, with the database confidence
35 taking precedence over that assigned to the study. In general, the level of confidence is inversely
11/5/99 6-4 DRAFT-DO NOT CITE OR QUOTE
-------
1 related to both the composite UF and the likelihood that the RfC would change with the
2 availability of new information; an RfC based on a sensitive effect in a sensitive human subgroup
3 as reported in a exemplary study with a composite UF of < 30 would more than likely be one of
4 high confidence.
5
6 6.2. ISSUES IN DERIVATION OF THE DIESEL RfC
7 Information available on diesel particulate matter (DPM) is that in other databases and
8 therefore includes several areas of controversy and uncertainly. This section introduces issues
9 concerning DPM. Subsequent sections will then more fully examine and consider these issues.
10
11 6.2.1. Chronic Noncancer Effects in Humans—Relevancy of Rodent Data
12 Current information shows that humans and rodents share some noncancer responses to
13 poorly soluble particles such as DPM that are qualitatively similar. These analogous responses
14 suggest that a potential commonality exists between humans and rodents in the underlying
15 mode(s) of action of DPM. These analogous responses and shared steps in the mode of action do
16 not appear to extend to the tumorigenic response seen in one particular rodent species, the rat.
17 As discussed in other sections of this document, the relevance to humans of the tumorigenic
18 response in rat lungs occurring under particle overload conditions is problematic.
J9
6.2.2. Pulmonary Pathology and Immunologic Effects as Critical Effects
21 Recent investigations in both laboratory animals and humans in clinical settings have
22 associated exposure to DPM with immunologic effects, especially enhanced allergenicity. The
23 relationship between pulmonary histopathology and allergenic effects is compared and contrasted
24 in the choice of pulmonary histopathology as a scientifically defensible critical effect upon which
25 to base this assessment.
26
27 6.2.3. Application of UFs
28 As discussed above, applications of UF consider both science and policy. Because of the
29 extensive database of well-conducted long-term chronic studies in several species, much is
30 known about the effects of DPM on the lung as target organ. Relatively few areas of uncertainty
31 are applicable to the diesel database. Moreover, the application of a pharmacokinetic model in
32 this assessment obviates a portion of the animal-to-human UF as explained above. Questions
33 concerning the application of uncertainty for consideration of the enhanced allergenic effects are
34 presented and discussed.
35
1175/99 6-5 DRAFT—DO NOT CITE OR QUOTE
-------
1 6.2.4. Relationship of DPM to Ambient Levels of PM2.S
2 DPM is acknowledged as a component of the fine paniculate matter (PM25) present in
3 ambient air, especially in urban areas. It is known that compared with PM2 5, DPM has a higher
4 proportion of fine and ultrafine particles and a higher content of organic compounds absorbed
5 onto the carbon core. DPM could thus be considered a subcategory of PM2 5 with greater
6 toxicologic potential from the higher organic compound content, which would penetrate more
7 efficiently into the alveolar compartment because of the preponderance of small particles in
8 DPM.
9
10 6.3. APPROACH FOR DERIVATION OF THE RfC FOR DIESEL ENGINE
11 EMISSIONS
12 6.3.1. Consideration of Long-Term Inhalation Studies
13 Twelve long-term (>1 year) laboratory animal inhalation studies of diesel engine
14 emissions have been conducted. These studies focused on effects in the pulmonary region.
15 Studies at the Inhalation Toxicology Research Institute (ITRI) and the Japanese Health Effects
16 Research Program (HERP) consisted of large-scale chronic exposures, with exposed animals
17 being designated for the study of various endpoints and at various time points (Ishinishi et al.,
18 1986,1988; Mauderly et al., 1987a,b, 1988; Henderson et al., 1988; Wolff et al., 1987). Each
19 research program is represented by multiple published accounts of results. These programs were
20 selected as the principal basis for deriving the RfC because each contains studies that identify an
21 LOAEL and an NOAEL for respiratory effects after chronic exposure (see Section 6.2) as well as
22 pulmonary histopathology. Effects in the upper respiratory tract and other organs were not found
23 consistently in chronic animal exposures.
24
25 6.3.2. Derivation of a HEC—Application of a Pharmacokinetic Model
26 PK models may be used to project across species concentrations of a toxicant that would
27 result in equivalent internal doses. When used for these purposes, PK models may be termed
28 dosimetric models. Chapter 3 reviewed and evaluated a number of dosimetric models applicable
29 to DPM. The model developed by Yu and Yoon (1990) that accounts for species differences in
30 deposition efficiency, normal and particle overload lung clearance rates, respiratory exchange
31 rates and particle transport to lung-associated lymph nodes was selected for use in this
32 assessment. A major assumption in this model is that the particle overload phenomenon occurs
33 in humans and. it) rats at equivalent lung burdens expressed as mass per unit surface area (Yu and
34 Yr»OH 1 OQO^ TVli« aQQiimntioti al1r»wc fnr th*» H«»^7*»1r»r»morit nf a /lioool nor+ir-lo^orio^ifi/^ lin»i-io»->
' S X" " ' ~ "~ "~ " ~*--'f——"- *-— —• ~« — «*~» f «M. »««r««r W£* WAAAW A»hU.k.»UA
35 retention model and therefore allows extrapolation from rat studies to human exposures. See
1175/99 6-6 DRAFT-DO NOT CITE OR QUOTE
-------
1 Chapter 3 for further discussion of the model and Appendix B for complete specifics on the use
2 of the model.
^fc A principal and critical decision in utilizing any dosimetric model is the measure of dose.
4 DPM is composed of an insoluble carbon core with a surface coating of relatively soluble organic
5 constituents. Because macrophage accumulation, epithelial histopathology, and reduced
6 clearance have been observed in rodents exposed to high concentrations of chemically inert
7 particles (Morrow, 1992), the toxicity of DPM may result from the carbon core rather than from
8 the associated organics. However, the organic component of diesel particles, consisting of a
9 large number of polycyclic aromatic hydrocarbons and heterocyclic compounds and their
10 derivatives (Chapter 2), may also play a role. It is not possible to separate the carbon core from
11 the adsorbed organics to compare the toxicity. Therefore, the whole particle was used as the
12 measure of dose. See Chapters 6 and 9 for further details.
13 The input data required to run the dosimetric model include the particle size
14 characterization expressed as mass median aerodynamic diameter (MMAD) and the geometric
15 standard deviation (og). In the principal and supporting studies used for the RfC derivation,
16 these parameters are measured using different methods and reported in different levels of detail.
17 Simulation data presented by Yu and Xu (1986) show that across a range of MMAD and og
18 inclusive of the values reported in these studies, the pulmonary deposition fraction differs by no
more than 20%. The minimal effect of even a large distribution of particle size on deposition
probably results because the particles are still mostly in the submicron range and deposition is
21 influenced primarily by diffusion. It has also been shown, however, that the particle
22 characteristics in a diesel exhaust exposure study depend very much on the procedures used to
23 generate the chamber atmosphere. Because of the rapid coagulation of particles, the volume and
24 temperature of the dilution gas are especially important. The differences reported in particle
25 sizes and distributions in various studies likely reflected real differences in the exposure
26 chambers as well as different analytical methods. Because the particle diameter and size
27 distribution were not reported in the two lowest exposure concentrations in the HERP studies, it
28 was decided to use a representative DPM particle size of MMAD = 0.2 urn and ag = 2.3 (values
29 typically reported for DPM) for modeling of lung burden. For consistency, the lung burdens for
30 the other studies were also calculated using this assumption. The difference in the HEC using
31 the default particle size compared with the actual reported particle size is no more than 4% in the
32 HERP study and 19% in the ITRI study.
33
1175/99 6-7 DRAFT-DO NOT CITE OR QUOTE
-------
1 6.4. CHOICE OF THE CRITICAL EFFECT—RATIONALE AND JUSTIFICATION
2 6.4.1. Mode-of-Action and Candidate Effects
3 Mode-of-action information about respiratory effects from diesel exposure indicates that
4 the pathogenic sequence following the inhalation of diesel exhaust begins with the phagocytosis
5 of diesel particles by alveolar macrophages (AMs). These activated AMs release chemotactic
6 factors that attract neutrophils and additional AMs. As the lung burden of DPM increases, there
7 are aggregations of particle-laden AMs in alveoli adjacent to terminal bronchioles, increases in
8 the number of Type II cells lining particle-laden alveoli, and the presence of particles within
9 alveolar and peribronchial interstitial tissues and associated lymph nodes. The neutrophils and
10 AMs release mediators of inflammation and oxygen radicals, and particle-laden macrophages are
11 functionally altered, resulting in decreased viability and impaired phagocytosis and clearance of
12 particles. The latter series of events may result in pulmonary inflammatory, fibrotic, or
13 emphysematous lesions like those described in the studies reviewed in Chapter 7. Epidemiologic
14 studies of occupationally exposed people provide suggestive evidence for a respiratory effect.
15 Although detailed information describing the pathogenesis of respiratory effects in humans is
16 lacking, the effects reported in studies of humans exposed to diesel exhaust lend qualitative
17 support to the findings in controlled animal studies and therefore to this basic mode of action.
18 Evidence from the available lexicological data on diesel exhaust consistently indicates
19 that inhalation of diesel exhaust can be a respiratory hazard, based on findings in multiple
20 controlled laboratory animal studies in several species with suggestive evidence from human
21 occupational studies, most of which are described and evaluated hi Chapter 7. The endpoints of
22 concern include biochemical, histopathological, and functional changes in the pulmonary and
23 tracheobronchial regions.
24 The occurrence of a lung cancer response in rats under conditions of "clearance overload"
25 from diesel exhaust/DPM has been discussed elsewhere hi this document as being possibly
26 unique to the rat and of problematic relevance to human lung responses. Yet effects in the rat
27 lung are being proposed as the basis for the RfC. There are several reasons why these effects are
28 considered valid and relevant for RfC derivation. First, the effects considered, inflammation
29 (inflammatory cell infiltration) and fibrosis, are noncancer effects. Second, similar noncancer
30 effects are seen in other species (mouse, hamster), albeit under conditions of higher exposure
31 than rats, and these species do not manifest a cancer response. Third, rats and humans do exhibit
32 similar noncancer responses (macrophage response and interstitial fibrosis) to less toxic particles
33 (i.e., coal dust) and to lower concentrations of poorly soluble particles such as DPM. Thus, when
34 viewed across species the pulmonary effects of inflammation and fibrosis are considered
35 dissociable from the cancer response and of likely relevance to humans.
1175/99 6-8 DRAFT-DO NOT CITE OR QUOTE
-------
1 Some evidence suggests liver and kidney changes in animals exposed to diesel exhaust.
2 There have also been some indications of neurotoxicity at high concentrations of diesel exhaust.
These data, however, are inadequate to indicate that a hazard exists for these endpoints.
4 Studies of other endpoints, including reproductive and developmental toxicity, in
5 controlled animal exposures have shown no potential hazard.
6 Recent evidence has accumulated for effects of diesel exhaust and DPM on respiratory
7 system-related immune function, especially enhanced or exacerbated allergenicity. Chapter 5
8 describes studies of human cells in vitro as well as human nasal instillation and inhalation studies
9 that have demonstrated the potential for PPM to enhance allergic inflammatory responses. This
10 effect included observations wherein increases of IgE were produced in nasal lavage, especially
11 when DPM was instilled concomitantly with allergen in atopic rhinitic subjects. DPM has also
12 been shown to enhance histamine-induced increase of certain inflammatory mediators such as
13 IL-8 and GM-CSF. Exposure of healthy human subjects to dilute diesel exhaust (300 ug) for 1
14 hour with intermittent exercise led to an acute mediator and cellular inflammatory response in the
15 airways and peripheral blood.
16
17 6.4.2. Rationale and Justification
18 The choice of critical effect for DPM must be consonant with the definition given above
fand made in consideration of the purposes of the RfC, e.g., a lifetime continuous exposure that is
without adverse effects. From the discussion above, the principal candidate critical effects are
21 the pulmonary histopathological changes in rats and enhanced allergenic effects in the upper
22 airways of animals and humans. The following points compare and contrast these effects:
23
24 • Pulmonary histopathology is shown consistently in several species with
25 clear dose-response under long-term realistic exposure scenarios. Allergenic
26 effects are shown consistently hi both animal and clinical human studies but,
27 dose-response and concentration * times (C * t) relationships are not
28 available under any exposure scenario.
29
30 • The relevance of these candidate effects to humans is each subject to
31 qualifications. Enhanced allergenic effects have been demonstrated in
32 humans. However, the observations were mostly in sensitized individuals
33 exposed via nasal instillation, a questionable route, and to relatively high
34 bolus doses. The pulmonary histopathology observed in rat studies is only
marginally supported by effects that may occur in humans.
i
11/5/99 6-9 DRAFT-DO NOT CITE OR QUOTE
-------
1
2 • Events that stimulate inflammatory processes may underlie both these
3 effects. Fibrogenesis is necessarily preceded and accompanied by
4 . inflammation. Events such as enhancement of inflammatory cytokines have
5 been associated with allergenic enhancement.
6
7 As the RfC is a dose-response assessment for effects encountered under conditions of
8 chronic exposure, pulmonary histopathology would therefore be the most robust and defensible
9 choice for the critical effect. Long-term, dose-response, and mode-of-action information could
10 warrant reconsideration of allergenic effects as being critical or possibly co-critical.
11
12 6.5. PRINCIPAL STUDIES FOR INHALATION RfC DERIVATION
13 The experimental protocol and results for the principal studies demonstrating and
14 characterizing the critical effect are discussed in Chapter 7 and Appendix A and are briefly
15 reviewed here. In studies conducted at ITRI, rats and mice were exposed to target DPM
16 concentrations of 0, 0.35, 3.5, or 7 mg/m3 for 7 h/day, 5 days/week for up to 30 mo (rats) or 24
17 mo (mice) (Mauderly et al., 1988). A total of 364 to 367 rats per exposure level were exposed
18 and used for studies examining different endpoints such as carcinogenicity, respiratory tract
19 histopathology and morphometric analysis, particle clearance, lung burden of DPM, pulmonary
20 function testing, lung biochemistry, lung lavage biochemistry and cytology, immune function,
21 and lung cell labeling index. Subsets of animals were examined at 6, 12, 18, and 24 mo of
22 exposure and surviving rats were examined at 30 mo. Diesel emissions from a 5.7-L engine
23 operated on a Federal Test Procedure urban driving cycle were diluted and fed into the exposure
24 chambers. Particle concentrations were measured daily using a filter sample, and weekly grab
25 samples were taken to measure gaseous components including carbon monoxide, carbon dioxide,
26 nitrogen oxides, ammonia, and hydrocarbons. The actual DPM concentrations for the low-,
27 medium-, and high-exposure levels were 0.353, 3.47, and 7.08 mg/m3, respectively. Mass
23 median diameters (geouicuic standard deviations) determined using an impactor/parallel flow
29 diffusion battery were 0.262 (4.2), 0.249 (4.5), and 0.234 (4.4) for the low-, medium-, and high-
30 exposure groups, respectively.
31 Lung wet weight to dry weight ratio was increased significantly in the two highest
32 exposure groups. Qualitative descriptions of the histopathological results in the respiratory tract
33 are found in Mauderly et al. (1987a, 1988), Henderson et al. (1988), and McClellan et al. (1986).
34 Aggregates of particle-laden AMs were seen after 6 mo in rats exposed to 7 mg/m3 DPM target
35 concentrations, and after 1 year of exposure histopathological changes were seen, including focal
36 areas of epithelial metaplasia. Fibrosis and metaplasia increased with duration of exposure and
11 /5/99 6-10 DRAFT-DO NOT CITE OR QUOTE
-------
1 were observable in the 3.5 and 7 mg/m3 groups of rats at 24 mo. Changes in the epithelium
2 included extension of bronchiolar cell types into the alveoli. Focal thickening of the alveolar
septa was also observed. Histopathological effects were seen in areas near aggregations of
particle-laden AMs. The severity of inflammatory responses and fibrosis was directly related to
5 the exposure level. In the 0.35 mg/m3 group of rats, there was no inflammation or fibrosis.
6 Although the mouse lungs contained higher lung burdens of DPM per gram of lung weight at
7 each equivalent exposure concentration, there was substantially less inflammatory reaction and
8 fibrosis than was the case in rats. Fibrosis was observed only in the lungs of mice exposed at 7
9 mg/m3 DPM and consisted of fine fibrillar thickening of occasional alveolar septa.
10 Groups of 16 rats and mice (8/sex) were subjected to bronchoalveolar lavage after 6, 12,
11 18, and 24 (rats only) mo of exposure (Henderson et al., 1988). Lung wet weights were increased
12 at 7 mg/m3 in mice and rats at all time points and in mice at 3.5 mg/m3 at all time points after 6
13 mo. An increase in lavagable neutrophils, indicating an inflammatory response in the lung, was
14 seen at 3.5 and 7 mg/m3 in rats and mice at most time points. An increase in protein content of
15 the bronchoalveolar lavage fluid was observed in rats exposed to 3.5 or 7 mg/m3 at 12 and 18
16 mo but not at 24 mo. Increased protein content was also seen in mice at the two higher
17 concentrations at all time points. Increases in lavage fluid content of lactate dehydrogenase,
18 glutathione reductase, p-glucuronidase, glutathione, and hydroxyproline were observed in rats
«and mice exposed to 3.5 or 7 mg/m3 at various time points. At the lowest exposure level, no
biochemical or cytological changes occurred in the lavage fluid or in lung tissue in either Fischer
21 344 rats or CD-I mice.
22 Mauderly et al. (1988; see also McClellan et al., 1986) examined the impairment of
23 respiratory function in rats exposed according to the protocol described above. After 24 mo of
24 exposure to 7 mg/m3 DPM, mean TLC, C^, quasi-static chord compliance, and CO diffusing
25 capacity were significantly lower than control values, and nitrogen washout and percentage of
26 forced vital capacity expired in 0.1 s were significantly greater than control values. There was no
27 evidence of airflow obstruction. Similar functional alterations were observed in the rats exposed
28 to 3.5 mg/m3 DPM, but such changes usually occurred later in the exposure period and were
29 generally less pronounced. There were no significant decrements in pulmonary function for the
30 0.35 mg/m3 group at any time during the study.
31 Wolff et al. (1987) investigated alterations in particle clearance from the lungs of rats in
32 the ITRI study. Progressive increases in lung burdens were observed over time in the 3.5 and 7.0
33 mg/m3 exposure groups. There were significant increases in 16-day clearance half-times of
34 inhaled radiolabeled particles of gallium oxide (0.1 um MMAD) as early as 6 mo at the 7.0
35 mg/m3 level and 18 mo at the 3.5 mg/m3 level; no significant changes were seen at the 0.35
mg/m3 level. Rats that inhaled fused aluminosilicate particles (2 (am MMAD) radiolabeled with
11/5/99 6-11 DRAFT-DO NOT CITE OR QUOTE
-------
1 cesium after 24 mo of diesel exhaust exposure showed increased clearance half-times in the 3.5
2 and 7.0 mg/m3 groups.
3 In the HERP studies, histopathological effects of diesel exhaust on the lungs of rats were
4 investigated (Ishinishi et al., 1986,1988). In this study, both light-duty (LD, 1.8-L) and heavy-
5 duty (HD, 11-L) diesel engines were operated under constant velocity and load conditions. The
6 exhaust was diluted to achieve target concentrations of 0.1 (LD only), 0.4 (LD and HD), 1 (LD
7 and HD), 2 (LD and HD), and 4 (HD only) mg/m3 DPM. Particle concentrations were
8 determined by filter samples. Actual concentrations were 0.11, 0.41, 1.18, and 2.32 mg/m3 for
9 the light-duty engine and 0.46, 0.96,1.84, and 3.72 mg/m3 for the heavy-duty engine. Fischer
i
10 344 rats (120 males and 95 females per exposure level for each engine type) were exposed for 16
11 h/day, 6 days/week for 30 mo. Particle size distributions were determined using an Andersen
12 cascade impactor and an electrical aerosol analyzer. At the 24-mo sampling, the MMAD and
13 distribution (ag) were 0.22 (2.93) and 0.19 (2.71) for the light-duty engine groups at 2.32 and
14 1.18 mg/m3, respectively, and 0.27 (3.18) and 0.22 (2.93) for the heavy-duty engine groups at
15 3.72 and 1.84 mg/m3, respectively (Ishinishi et al., 1988). The number and tuning of the samples
16 are not clear from the published reports, nor is it clear which method was used for the results
17 reported above. Particle size data were not reported for the other exposure groups, although
18 measurements for all groups, including those of ITRI, are quite similar to one another.
19 Hematology, clinical chemistry, urinalysis, and light and electron microscopic examinations were
20 performed. The body weight of females exposed to 4 mg/m3 DPM was 15% to 20% less than
21 that of controls throughout the study. No histopathological changes were observed in the lungs
22 of rats exposed to 0.4 mg/m3 DPM or less. At concentrations above 0.4 mg/m3 DPM,
23 accumulation of particle-laden AMs was observed. In areas of AM accumulation, there was
24 bronchiolization of the alveolar ducts, with bronchiolar epithelium replacing alveolar epithelium.
25 Proliferation of brochiolar epithelium and Type II cells was observed. In these areas, edematous
26 thickening and fibrosis of the alveolar septum were seen. Fibrosis of the alveolar septum
27 developed into small fibrotic lesions. These are collectively referred to as hyperplastic lesions by
28 the authors and their incidence is reported.
29 From a total of 123 to 125 animals examined (approximately equal numbers of males and
30 females), hyperplastic lesions were reported in 4, 4,6,12, and 87 animals in the light-duty engine
31 groups exposed to 0, 0.11, 0.41, 1.18, and 2.32 mg/m3 DPM, respectively, and in 1, 3, 7, 14, and
32 25 animals in the heavy-duty engine groups exposed to 0,0.46, 0.96,1.84, and 3.72 mg/m3 DPM,
33 respectively. Statistical analysis of these results was not reported, but there was no difference in
34 the severity ascribed to changes in pulmonary pathology at similar exposure concentrations
35 between the LD and the HD series.
11/5/99 6-12 DRAFT-DO NOT CITE OR QUOTE
-------
1 The ITRI and HERP studies are complementary for identifying the critical effect and its
2 LOAEL and NOAEL. The ITRI study provides results on many different endpoints reflecting
f pulmonary toxicity, and the effect levels are the same, but the LOAEL and NOAEL are different
by a factor of 10. In the HERP study, the concentrations differ by a factor of 2-4, but only
5 histopathology .is reported. Taken together, these two studies (including several published
6 reports for the ITRI study) provide good definition of the low-concentration effects of diesel
7 emissions.
8 The HERP study identifies LOAELs for rats exposed chronically at 1.18 and 0.96 mg/m3
9 (actual exposure) for the LD and HD series, respectively, and NOAELs at 0.41 and 0.46 mg/m3
10 (actual) for the LD and HD series. The ITRI studies identify a NOAEL for biochemical,
11 histopathological, and functional changes in the pulmonary region at 0.35 mg/m3 (LOAEL = 3.5
12 mg/m3). The HECs for the principal studies were obtained using the deposition and retention
13 model of Yu and Yoon (1990), as discussed previously. The HEC calculation is based on the
14 assumption that the estimate for the human exposure scenario (a 70-year continuous exposure)
15 should result in an equivalent dose metric, expressed as mass of diesel particle carbon core per
16 unit of pulmonary region surface area, to that associated with no effect at the end of the 2-year rat
17 study. To obtain the HEC, the lung burden in the rat study is calculated using the exposure
18 regimen (concentration, number of hours per day, and days per week) and values for rat tidal
19 volume, functional residual capacity, and breathing frequency. A continuous human exposure
^P resulting in the same final lung burden is calculated and is the HEC. The HEC values
21 corresponding to the animals' exposure levels in the principal studies are shown in Table 6-2,
22 along with a designation of the concentrations as AEL (adverse-effects level) or NOAEL; the
23 LOAELs (HEC) are 0.30, 0.36, and 0.36 mg/m3. These values, along with the LOAELs from
24 other studies (discussed below), show strong support for an experimental threshold in rats in the
25 range of 0.15 to 0.3 mg/m3 DPM. The highest NOAEL (HEC), which is below all LOAELs
26 (HEC), is 0.155 mg/m3 DPM from the HERP heavy-duty diesel study. This NOAEL (HEC) is
27 selected as the basis for the RfC calculation.
28
29 6.6. SUPPORTING STUDIES FOR INHALATION RfC DERIVATION
30 Chronic inhalation studies using male F344 rats and male Hartley guinea pigs were
31 carried out at the General Motors (GM) Research Laboratories (Barnhart et al., 1981, 1982).
32 Exposures to target concentrations of 0.25, 0.75, and 1..5 mg/m3 DPM were generated 20 h/day,
33 5.5 days/week for up to 2 years. Exposures at 0.75 and 1.5 mg/m3 for 2 weeks to 6 mo were
34 reported by Barnhart et al. (1981,1982). The focus of these studies is on electron micrographic
35 morphometry, and very little descriptive light microscopic histology is reported. These data
show that no appreciable changes in morphometric parameters occurred after a 2-year exposure
1115/99 6-13 DRAFT-DO NOT CITE OR QUOTE
-------
Table 6-2. Human equivalent continuous concentrations from the principal studies
Study
HERP-light duty
HERP-heavy duty
ITRI
Exposure
concentration
(mg/m3)
0.11
0.41
1.18
2.32
0.46
0.96
1.84
3.72
0.353
3.47
7.08
AEL/NOAEL"
NOAEL
NOAEL
AEL
AEL
NOAEL
AEL
AEL
AEL
NOAEL
AEL
AEL
HEC"
(mg/m3)
0.038
0.139
0.359
0.571
0.155
0.303
0.493
0.911
0.042
0.360
0.582
"AEL: adverse-effects level; NOAEL: no-observed-adverse-effect level.
bHEC: human equivalent concentration obtained from applying the dosimetric model of Yu and Yoon (1990).
1 to 0.25 mg/m3, while exposure to 0.75 or 1.5 mg/m3 DPM resulted in increased thickness of
2 alveolar septa and increased number of various types of alveolar cells. Increased numbers of
3 PMNs and monocytes were lavaged from rats exposed to 0.75 or 1.5 mg/m3, and biochemical
4 changes occurred in lung tissue at these concentrations (Misiorowski et al., 1980; Eskelson et al.,
5 1981; Strom, 1984). These studies demonstrate a LOAEL of 0.796 mg/m3 DPM and a NOAEL
6 of 0.258 mg/m3 DPM for male guinea pigs in a chronic study for respiratory endpoints, mclndina
7 light and electron microscopy, lavage cytology, and lung tissue biochemistry.
8 A 15-mo inhalation study was performed by Southwest Research Institute for General
9 Motors (Kaplan et al., 1983). Male F344 rats, Syrian golden hamsters, and A/J mice were
10 exposed to diluted diesel exhaust at target concentrations of 0.25, 0.75, and 1.5 mg/m3 for 20
11 h/day and 7 days/week. Focal accumulation of particle-laden AMs was associated with minimal
12 to mild fibrosis of the alveolar wall. Based on accumulation of particle-laden macrophages, this
13 study identifies a LOAEL at 0.735 mg/m3 and a NOAEL at 0.242 mg/m3.
11 /5/99 6-14 DRAFT-DO NOT CITE OR QUOTE
-------
1 In a study performed by NIOSH (Lewis et al., 1986, 1989; Green et al., 1983), male and
2 female F344 rats and male Cynomolgus monkeys were exposed to target levels of 2 mg/m3 diesel
f particles. Accumulations of black-pigmented alveolar macrophages were seen in the alveolar
ducts of rats adjacent to terminal bronchioles, and epithelial lining cells adjacent to collections of
5 pigmented macrophages showed marked Type II cell hyperplasia. No evidence of impaired
6 pulmonary function as a result of the exposure to diesel exhaust was found in rats. Histological
7 examination of lung tissue from monkeys exposed for 24 mo in the same regimen used for rats
8 revealed aggregates of black particles, principally in the distal airways of the lung. No fibrosis,
9 focal emphysema, or inflammation was observed. The monkeys exposed to diesel exhaust
10 demonstrated small-airway obstructive disease. This study demonstrates a LOAEL for rats and
11 monkeys at a diesel particle concentration of 2 mg/m3. Although the data suggest that the
12 pulmonary function effect in primates more closely resembles that in humans, this study had only
13 one exposed group, making evaluation of dose response impossible. Thus, it was not considered
14 sufficient to eliminate consideration of the strong rodent database.
15 Heinrich et al. (1986; see also Stober, 1986) exposed male and female Syrian golden
16 hamsters, female NMRI mice, and female Wistar rats to diesel engine emissions with a
17 4.2 mg/m3 particulate concentration. Lung weights were increased by a factor of 2 or 3 in rats
18 and mice after 2 years of exposure, and in hamsters the lung weights were increased by 50% to
70%. Although histopathological examination revealed different levels of response among the
three species, histopathological effects were seen in all species and effects on pulmonary function
21 were observed in rats and hamsters. This study demonstrates a LOAEL of 4.2 mg/m3 in rats for
22 respiratory system effects.
23 The effects of diesel exhaust on the lungs of 18-week-old male Wistar rats exposed to 8.3
24 ± 2.0 mg/m3 particulate matter were investigated by Karagianes et al. (1981). Histological
25 examinations of lung tissue noted focal aggregation of particle-laden alveolar macrophages,
26 alveolar histiocytosis, interstitial fibrosis, and alveolar emphysema. Lesion severity was related
27 to length of exposure. No exposure-related effects were seen in the nose, larynx, or trachea.
28 This study demonstrates a LOAEL of 8.3 mg/m3 DPM for respiratory effects after chronic
29 exposure of rats to diesel emissions.
30 Lung function was studied in adult cats chronically exposed to diesel exhaust
31 concentrations of 6.34 mg/m3 for the first 61 weeks and 6.7 mg/m3 from weeks 62 to 124. No
32 definitive pattern of pulmonary function changes was observed following 61 weeks of exposure;
33 however, a classic pattern of restrictive lung disease was found at 124 weeks (Pepelko et al.,
34 1980).
35 Heinrich et al. (1995) exposed Wistar rats to diesel exhaust at DPM concentrations of 0.8,
2.5, and 7 mg/m3,18 h/day, 5 days/week for 24 mo. Body weights were significantly decreased
11/5/99 6-15 DRAFT-DO NOT CITE OR QUOTE
-------
1 in the two higher exposure groups. Bronchoalveolar hyperplasia and interstitial fibrosis of
2 increasing incidence and severity at greater concentrations were seen in all exposure groups.
3 This study demonstrates a LOAEL of 0.8 mg/m3.
4 Nikula et al. (1995) exposed Fischer 344 rats to diesel exhaust at DPM concentrations of
5 2.4 and 6.3 mg/m3 16 h/day, 5 days/week for 23 mo. Survival was decreased in the high-
6 exposure males, while body weights were reduced in both males and females in the high-
7 exposure group. Pulmonary hyperplasia, inflammation, and fibrosis were seen in a high
8 percentage of rats in both exposure groups. The high exposure concentrations precluded use of
9 this study for development of an RfC.
10 Werchowski et al. (1980a) reported a developmental study in rabbits exposed on days 6
11 through 18 of gestation to a l-in-10 dilution of diesel exhaust (DPM concentration =12 mg/m3).
12 Exposure to diesel emissions had no effect on maternal toxicity or the developing fetuses. In a
13 companion study (Werchowski et al., 1980b), 20 SD rats were exposed for 8 h/day during days 5
14 to 16 to a target concentration of 12 mg/m3 of DPM. Fetuses were examined for external,
15 internal, and skeletal malformations, and the numbers of live and dead fetuses, resorptions,
16 implants, corpora lutea, fetal weight, litter weight, sex ratio, and maternal toxicity were recorded.
17 No conclusive evidence of developmental effects was observed in this study.
18 In an EPA-sponsored reproductive study summarized by Pepelko and Peraino (1983),
19 CD-I mice were exposed to a target concentration of 12 mg/m3 DPM for 8 h/day and
20 7 days/week. The F0 and F, animals were exposed for 100 days prior to breeding, and
21 100 mating pairs were randomly assigned to four exposure groups of 25 each. Viability counts
22 and pup weights were recorded at 4, 7, and 14 days after birth and at weaning. No treatment-
23 related effects on body weight in F0 mice or in F, animals through weaning or in mating animals
24 through gestation were found. No treatment-related effects on gestation length, percent fertile,
25 litter size, or pup survival were observed. The only organ weight difference was an increase in
26 lung weight in exposed F0 and F, mice (lung weight and lung weight/body weight) and in F2
27 males (lung weight/body weight). Based on this study, a NOAEL for reproductive effects in rats
28 is identified at 19 mg/m3 DPM.
29 The reproductive and developmental studies described in Chapter 5 show that effects in
30 the respiratory system are the most sensitive effects that result from diesel exhaust exposures.
31 These studies add to the confidence that a variety of noncancer effects have been studied and are
32 required for a designation of high confidence in the database and the RfC (discussed further
33 below).
34 Several cpidemiulugic studies have evaluated the effects of chronic exposure to diesel
35 exhaust on occupationally exposed workers. The human studies, taken together, are suggestive
36 but inconclusive of an effect on pulmonary function, as described in Chapter 7. The studies are
11/5/99 6-16 DRAFT-DO NOT CITE OR QUOTE
-------
1 not directly useful for deriving the RfC because of inadequate ability to directly relate the
2 observed effects to known concentrations of DPM. The studies are confounded by coexposures
to other particles or by a lack of measurement of particle exposure.
5 6.6.1. Respiratory Tract Effects in Species Other Than the Rat
6 In several of the chronic inhalation studies described in Chapter 7, one or more species
7 other than the rat were also exposed and examined for toxic effects. These provide a basis for
8 comparison of the effects in rats with the effects in other species. In the study performed at ITRI
9 (Henderson et al., 1988; Mauderly et al., 1988), male and female CD-I mice were exposed
10 similarly to the rats. The LOAEL and NOAEL in rats and mice from this study would be the
11 same, with the NOAEL for respiratory tract effects being 0.35 mg/m3 DPM (duration adjusted
12 NOAEL is 0.074 mg/m3), although some differences in the severity of the effect were apparent.
13 In the study conducted by the GM Biomedical Science Department (Bamhart et al., 1981,
14 1982; Strom, 1984; Gross, 1981), male Hartley guinea pigs as well as F344 rats were chronically
15 exposed to 0.258,0.796, and 1.53 mg/m3 DPM. The evidence from this study leads to the
16 conclusion that the LOAEL and NOAEL for rats and guinea pigs are the same, although
17 important differences in the endpoints were reported in the two species. The NOAEL is 0.258
18 mg/m3 (duration-adjusted NOAEL is 0.17 mg/m3).
Kaplan et al. (1982) reported a subchronic study in F344 rats, A/J mice, and Syrian
golden hamsters exposed to 1.5 mg/m3 DPM. The histopathological observations, including AM
21 accumulation and associated thickening of the alveolar wall, were described together, with no
22 distinction between species, suggesting that the observed effects were similar in the species
23 examined. Kaplan et al. (1983) reported a 15-mo study in which F344 rats, A/J mice, and Syrian
24 golden hamsters were exposed to 0.25, 0.75, or 1.5 mg/m3 DPM. No exposure-related lesions
25 were found in tissues other than the respiratory tract. Based on particle-laden AM accumulation,
26 this study identifies a LOAEL at 0.735 mg/m3 and a NOAEL at 0.242 mg/m3. The descriptions
27 provided suggest that the pulmonary effects were similar across the three species examined, but
28 this conclusion is compromised by the lack of detailed reporting and the possibility of infection
29 in rats and poor animal health (as evidenced by poor growth) in hamsters. The duration-adjusted
30 NOAEL is 0.202 mg/m3.
31 Lewis et al. (1986,1989) exposed rats and monkeys to 2 mg/m3 DPM for 2 years and
32 reported pulmonary function and histopathology. Pulmonary function was affected in both
33 species, although with a different pattern of response, as discussed in Chapter 5. Significant
34 differences were observed in the histopathological response. In monkeys, slight particle
35 accumulation was observed, but no fibrosis, focal emphysema, or inflammation was present. Rat
11/5/99 6-17 DRAFT-DO NOT CITE OR QUOTE
-------
1 lungs in this experiment showed AM accumulation, multifocal histiocytosis, and associated
2 fibrosis and inflammatory cells in the interstitium.
3 Heinrich et al. (1986) exposed Wistar rats, Syrian golden hamsters, and NMRJ mice
4 chronically to 4 mg/m3 DPM. Lung weight was increased 1.5-fold in hamsters, twofold in mice,
5 and threefold in rats. The activity of enzymes recovered in bronchoalveolar lavage was increased
6 to roughly the same extent in rats, mice, and hamsters. Hamsters showed thickened alveolar
7 septa and slight epithelial hyperplasia, with no AM accumulation. Mice also showed epithelial
8 hyperplasia and interstitial fibrosis. Rat lungs had severe inflammatory changes, thickened
9 alveolar septa, hyperplasia, and metaplasia. This study presents the clearest indication of a
10 possibly greater severity of noncancer effects in rats compared with other rodent species. It also
11 suggests that the effect in rats may be qualitatively different, with AM accumulation playing a
12 greater role in pathogenesis in rats than in other rodent species.
13 Heinrich et al. (1995) also compared effects of chronic diesel exposure on rats and two
14 strains of mice exposed to fairly high concentrations of diesel particles. Similar lung burdens
15 were reported in rats and mice on the basis of particle mass per unit lung wet weight. Lung
16 weight was increased to about the same extent in rats and mice. However, the study is focused
17 on cancer effects, and insufficient information is provided to make a detailed comparison of
18 noncancer histopathology in rats and mice.
19 Several of the studies described above and in Chapter 7 suggest a significant difference in
2O the carcinogenic response of rats and other experimental animal species. It is less clear whether
21 such a difference holds for noncancer effects at lower exposure levels. The studies described
22 above show similar effect levels for different species for effects that occur earlier or at lower
23 exposure concentration, including accumulation of particles, bronchoalveolar lavage
24 measurements, lung weight, and minor epithelial thickening and hyperplasia. At higher diesel
25 concentrations there are clear differences between rats and the other species tested, especially in
26 the progression to more severe histopathologically observed endpoints, such as hyperplasia,
27 metaplasia, and inflammatory response. Thus the NOAEL for chronic effects of diesei does not
28 appear to be substantially different among species, although, there is seme suggestion in the
29 literature of a more sensitive as well as a qualitatively different response in rats. This
30 comparison is weakened because the published reports often give less emphasis to noncancer
31 responses and because the effects in rats and other species are not always measured or reported in
32 the same way. The pathogenesis of diesel exhaust effects has not been studied as thoroughly in
33 any other species as it has in the rat. For example, no specific measurement of particle clearance
34 from the lung has been reported in any species ether than the rat. \Vitliiu ilic resolving power of
35 the available studies, it is concluded that there is limited evidence for a difference in the NOAEL
36 for noncancer effects across species, but the evidence is not adequate to quantitatively define the
11/5/99 6-18 DRAFT-DO NOT CITE OR QUOTE
-------
1 difference, especially at low exposure concentrations. Hence there is no clearly more appropriate
2 species on which the RfC derivation for nbncancer effects should be based.
A3 Mice were included in the ITRI, Kaplan et al. (1982), and Heinrich et al. (1986, 1995)
4 studies. The Heinrich studies used a single exposure to high concentrations and are supportive of
5 the other results in mice but are not appropriate to define a LOAEL for mice. The Kaplan study
6 defines an LOAEL and NOAEL of 0.735 and 0.242 mg/m3 DPM, respectively. The duration-
7 adjusted LOAEL and NOAEL are 0.613 and 0.202 mg/m3, respectively. The ITRI study defined
8 the adjusted LOAEL and NOAEL at 0.723 and 0.074 mg/m2, respectively. Because the dose
9 spacing is so wide in the ITRI study, the Kaplan study is more appropriate for defining a
10 NOAEL. Likewise, the Kaplan et al. study is the only multiple-dose study in hamsters, and it
11 defines the same LOAEL and NOAEL for hamsters as for mice. The GM study is the only
12 chronic study in guinea pigs, and it defines the LOAEL and NOAEL for this species at 0.796 and
13 0.258 mg/m3, respectively. The adjusted LOAEL and NOAEL for guinea pigs from the GM
14 study are 0.52 and 0.17 mg/m3, respectively. The effects levels for mice, hamsters, and guinea
15 pigs are similar to the duration-adjusted LOAEL and NOAEL for rats, which are 0.723 mg/m3
16 (ITRI study) and 0.26 mg/m3 (from Ishinishi et al., 1988), respectively. If the RfC were to be
17 derived based on the duration-adjusted NOAEL, the rat data would be preferred because of the
18 more complete database of chronic rat studies and the more complete presentation of the
9 noncancer endpoints in the rat studies.
The method for deriving inhalation RfCs (U.S. EPA, 1994) includes dosimetric
21 adjustments of animal exposure to arrive at a human equivalent concentration. The default
22 calculation of an HEC for a particle exposure uses the ratio of animal-to-human regional
23 deposited dose (RDDR) to a specific region of the respiratory tract. The methods also allow
24 replacement of the default approach when a better model is available. The derivation of the RfC
25 in this case makes use of the Yu and Yoon (1990) model to calculate the HEC from the rat
26 studies. Since the Yu and Yoon model has been developed only for the rat-to-human
27 extrapolation, the chosen approach assumes that dosimetric differences between rats and other
28 small-animal species would not result in a substantially lower HEC. The LOAEL (HEC) and
29 NOAEL (HEC) from the rat studies based on the Yu and Yoon model are 0.36 and 0.155 mg/m3,
30 respectively.
31
32 6.6.2. Application of the Benchmark Dose Approach to Derivation of the RfC
33 An alternative to deriving the RfC based on the NOAEL identified in the animal studies
34 is application of the BMC approach. The BMC was described by Crump (1984) and recently
35 discussed by EPA (1995b). The BMC approach involves fitting a dose-response function to dose
and effect information from a single study and using the dose-response curve to predict the dose
11 /5/99 6-19 DRAFT-DO NOT CITE OR QUOTE
-------
1 that will result in a response that is defined a priori as the benchmark response. For example, a
2 10% increase in incidence of epithelial hyperplasia might be defined as the benchmark response,
3 and a dose-response curve relating inhaled DPM to hyperplasia in rats chronically exposed to
4 diesel exhaust would be used to estimate the exposure concentration resulting in a 10% increase.
5 The lower confidence limit of that concentration is the BMC, and it is used as the representative
6 value for the dose-response assessment.
7 Several key issues concerning the derivation and interpretation of BMCs, especially in a
8 comparative manner over a variety of studies with a myriad of endpoints with differing types of
9 data such as with diesel, are yet to be resolved by the Agency. Several principal limitations are
10 the following:
11 • Some key studies in rats have inadequate quantitative data for BMC.
12 • The scientific criteria for selecting BMC from many endpoints remains to be
13 established.
14 • A deposition model is available only for rats (it is not clear how to compare BMCs
15 based on deposition/retention models with BMCs based on default duration-adjusted
16 concentrations).
17 Because of the issues and questions raised by these aspects of the BMC approach, the BMC will
18 not be used to derive the RfC at this time.
19
20 6.7. DERIVATION OF THE INHALATION RfC
21 6.7.1. The Effect Level—A NOAEL From a Chronic Inhalation Study
22 Based on the analysis above, the studies of chronic exposures to diesel emissions
23 performed at ITRI and HERP (Ishinishi et al., 1988; Mauderly et al., 1988) were selected as the
24 basis of the RfC, because they identify both a NOAEL and a LOAEL for rats exposed
25 chronically, because they identified the highest NOAEL (Table 6-2), and because they are
26 thoroughly reported. The only other study identifying both a NOAEL and a LOAEL was the GM
27 study, which was not used because information characterizing the pulmonary lesions in rats was
28 limited. The availability of the dosimetric model for rats and not for other species, along with the
29 apparent comparability between the rat and other rodent species in response, are also contributory
30 to choosing the rat as the basis for developing the RfC. Although the data from the monkey in
31 the Lewis et al. (1989) study suggest that the pulmonary function effect hi primates more closely
32 resembles that in humans, this study had only one exposed group, making evaluation of dose
33 response impossible. Thus, these data are not sufficiently robust for derivation of an RfC but
34 may be used as supporting information. The pulmonary effects, including histopathological
35 lesions, biochemical changes, pulmonary function impairment, and impaired particle clearance,
1175/99 6-20 DRAFT—DO NOT CITE OR QUOTE
-------
1 were determined to be the critical noncancer effect. Sufficient documentation from other studies
2 showed that there is no effect in the extrathoracic (nasopharyngeal) region of the respiratory
system or in other organs at the lowest levels that produces pulmonary effects in chronic
4 exposures. The exposure concentration of 0.46 mg/m3 from the study of Ishinishi et al. (1988) is
5 the NOAEL. Application of the dosimetric model of Yu and Yoon (1990) to this value resulted
6 in a NOAEL(HEC) of 0.155 mg/m3.
7
8 6.7.2. Application of UFs—Animal-to-Human and Sensitive Subgroups
9 Principal areas of uncertainty for this assessment are the human-to-sensitive human and
10 animal-to-human extrapolations (Table 6-1). Because the RfC is based on a NOAEL from a
11 chronic animal study, neither LOAEL-to-NOAEL nor subchronic-to-chronic extrapolations are
12 .needed. Also, the database for diesel is robust, with numerous well-conducted chronic studies in
13 addition to information showing no adverse effects on development in two species or on
14 reproduction in a two-generational study, all of which serve to eliminate the need for a UF for
15 database deficiencies.
16 No quantitative information exists regarding subgroups that may be sensitive to the
17 effects of diesel exhaust or DPM. The information available on enhanced allergenic effects
18 discussed above and in Chapter 7 suggests that individuals already sensitized by various antigens
are more sensitive to exposure to DPM than are those who are not, especially when undergoing
an allergenic inflammatory episode. However, no quantitation of the relative sensitivity is
21 available. Nor is there information indicating that children or male or female neonates are
22 especially more or less sensitive. Therefore the default value of 10 is used to accommodate
23 human-to-sensitive human extrapolation (Table 6-1).
24 Several issues reside in applying the UF for animal-to-human extrapolation to the diesel
25 database. First, the PK component of this UF (see Table 6-1) has been addressed by the
26 application of a dosimetric model to obtain a HEC, thereby decreasing the UF to 3 (or 1005) for
27 the residual PD component. Second, information discussed above and in Chapter 6 indicates that
28 for certain endpoints such as chronic inflammation, the rat appears to have a more sensitive
29 response than other species, including humans. That rats are more sensitive to the effects of
30 inhaled DPM than are humans could be considered evidence sufficient to eliminate the remaining
31 PD component of this UF. However, mode-of-action evidence for the various effects observed
32 with diesel, especially pulmonary histopathology and immunologic effects such as enhanced
33 allergenicity, indicate that events stimulatory to inflammatory processes underlie these effects,
34 i.e., neutrophilic inflammation preceding fibrogenesis and such events as increased cytokine
35 production preceding immunologic effects. Although indications are that humans are less
sensitive than are rats to the inflammatory-mediated endpoint of fibrogenesis, it is problematic to
11/5/99 6-21 DRAFT-DO NOT CITE OR QUOTE
-------
1 presume that humans would also be less sensitive to other inflammatory-mediated endpoints such
2 as enhanced allergenicity that are now documented in the literature. In consideration of this
3 missing specific mode-of-action information on inflammation, the PD component is retained at
4 the value of 3.
5 The total composite UF is therefore 10x3 = 30.
6
7 The resultant RfC = NOAEL(HEC) = 0.155mg/m3 = 5E-3 mg/m3 (5 ug/m3)
8 UF 30
9
10 6.7.3. Designation of Confidence Level
11 The studies used as the basis for the RfC were well-conducted chronic studies with
12 adequate numbers of animals, in which the target tissues (i.e., the respiratory tract) were
13 thoroughly examined and in which the LOAELs and NOAELs were consistent across studies.
14 The database contains several chronic studies, including multiple species, that support the
15 LOAEL observed in the principal study. The availability of multiple chronic studies all having
16 consistent effect levels imparts a high confidence to the principal study. Developmental and
17 multigeneration reproductive studies also exist, resulting in a high-confidence database. The
18 endpoints chosen have relevancy to the human response to other poorly soluble particulates.
19 The modeling employed in this assessment to derive FEECs includes both deposition and
20 clearance mechanisms, although assumptions have been made with certain of the clearance
21 parameters. Current mode-of-action information indicates that events stimulatory to
22 inflammatory processes underlie the effects reported in the pulmonary (target) tissues. Continued
23 investigation in this area may clarify the status of other effects (e.g., immunologic) reported from
24 diesel exposure.
25 The application of this RfC to general ambient paniculate matter such as PM2 5 must be
26 limited. Compared with PM,,, DPM has a relatively high organic content and a preponderance
27 of small particles capable of penetrating to the lung. As a consequence, DPM may be consiueied
28 a subcategory of PM2 5, with perhaps a greater potential for eliciting toxicity.
29 High confidence hi both the studies and database leads to high confidence hi the RfC
30 itself.
31
32 6.8. SUMMARY
33 Table 6-3 summarizes the principal decision points in dcrivatioii of the diesei Rfu, the
34 Agency's estimate of a continuous inhalation exposure that is considered to be without an
35 appreciable risk of deleterious noncancer effects during a lifetime.
11/5/99 6-22 DRAFT—DO NOT CITE OR QUOTE
-------
Table 6-3. Decision summary for the derivation of the RfC for diesel engine emissions
Critical effect
Principal study
NOAEL
Model adjusted NOAEL =
NOAEL(HEC)
UFs
Composite UF
NOAEL(HEC) / UF = RfC
Confidence in the RfC
Pulmonary histopathology in rats
Ishinishi et al., 1988; Mauderly et al.,
1988
0.46 mg/m3
0.155mg/m3
1 0 — Human-to-sensitive human
3 — Animal-to-human (pharmacodynamics)
30
0.155 mg/m3 / 30 = 5E-3 mg/m3
High
1 The derivation of this RfC was made in consideration of several candidate critical effects
2 (including immunologic endpoints), in consideration of the relevancy of the critical effect chosen
3 to the human response, and in recognition of the strengths and limitations of the modeling
^fc applied to obtain a human equivalent concentration (HEC).
6 6.9. REFERENCES
7 Barnes, DG; Daston, GP; Evans, JS; et al. (1995) Benchmark dose workshop: criteria for use of a benchmark
8 dose to estimate a reference dose. Regul Toxicol Pharmacol 21:296-306.
9 Barnes, DG; Dourson, ML. (1988) Reference dose (RfD): description and use in health risk assessments. Regul
10 Toxicol Pharmacol 8:471-486.
11 Barnhart, MI; Chen, S-T; Salley, SO; et al. (1981) Ultrastructure and morphometry of the alveolar lung of guinea
12 pigs chronically exposed to diesel engine exhaust: six months' experience. J Appl Toxicol 1:88-103.
13 Barnhart, MI; Salley, SO; Chen, S-T; et al. (1982) Morphometric ultrastructural analysis of alveolar lungs of
14 guinea pigs chronically exposed by inhalation to diesel exhaust (DE). In: Lewtas.J, ed. Toxicological effects of
15 emissions from diesel engines: proceedings of the Environmental Protection Agency diesel emissions symposium;
16 October 1981; Raleigh, NC. (Developments in toxicology and environmental science: v. 10.) New York: Elsevier
17 Biomedical; pp. 183-200.
18 Creutzenberg, O; Bellmann, B; Heinrich, U; et al. (1990) Clearance and retention of inhaled diesel exhaust
19 particles, carbon black, and titanium dioxide in rats at lung overload conditions. J Aerosol Sci 21(Suppl):S455-
20 S458.
21 Crump, KS. (1984) A new method for determining allowable daily intakes. Fundam Appl Toxicol 4:854-871.
Crump, KS. (1995) Calculation of benchmark doses from continuous data. Risk Anal 15:79-89.
11/5/99 6-23 DRAFT—DO NOT CITE OR QUOTE
-------
1 Eskelson, CD; Strom, KA; Vostal, JJ; et al. (1981) Lipids in the lung and lung gavage fluid of animals exposed to
2 . diesel particulates. Toxicologist 1:74-75.
3 Gaylor, DW; Slikker, W, Jr. (1990) Risk assessment for neurotoxic effects. Neurotoxicology 11:211-218.
4 Green, FHY; Boyd, RL; Danner-Rabovsky, J; et al. (1983) Inhalation studies of diesel exhaust and coal dust in
5 rats. Scand J Work Environ Health 9: 181-188.
6 Gross, KB. (1981) Pulmonary function testing of animals chronically exposed to diluted diesel exhaust. J Appl
7 Toxicol 1:116-123.
8 Heinrich, U; MuhJe, H; Takenaka, S; et al. (1986) Chronic effects on the respiratory tract of hamsters, mice, and
9 rats after long-term inhalation of high concentrations of filtered and unfiltered diesel engine emissions. J Appl
10 Toxicol 6:383-395.
1 1 Heinrich, U; Fuhst, R; Rittinghausen, S; et al. (1995) Chronic inhalation exposure of Wistar rats and two strains
12 of mice to diesel engine exhaust, carbon black, and titanium dioxide. Inhal Toxicol 7:553-556.
1 3 Henderson, RF; Pickrell, JA; Jones, RK; et al. (1988) Response of rodents to inhaled diluted diesel exhaust:
14' biochemical and cytological changes in bronchoalveolar lavage fluid and in lung tissue. Fundam Appl Toxicol
15 11:546-567.
1 6 Howe, RB. (1990a) THRESH: a computer program to compute a reference dose from quantal animal toxicity data
1 7 using the benchmark dose method. Clement International Corporation, K.S. Crump Division, Ruston, LA.
1 8 Howe, RB. (1990b) THC: a computer program to compute a reference dose from continuous animal toxicity data
1 9 using the benchmark dose method. Clement International Corporation, K.S. Crump Division, Ruston, LA.
20 Howe, RB. (1990c) THRESHW: a computer program to compute reference doses from quantal animal toxicity
21 data using the benchmark dose method. Clement International Corporation, K.S. Crump Division, Ruston, LA.
22 Howe, RB. (1990d) THWC: a computer program to compute a reference dose from continuous animal toxicity
23 data using the benchmark dose method. Clement International Corporation, K.S. Crump Division, Ruston, LA.
24 Ishinishi, N; Kuwabara, N; Nagase, S; et al. (1986) Long-term inhalation studies on effects of exhaust from
25 heavy and light duty diesel engines on F344 rats. In: Ishinishi, N; Koizumi, A; McClellan, RO; et al., eds.
26 Carcinogenic and mutagenic effects of diesel engine exhaust: proceedings of the international satellite symposium
27 on lexicological effects of emissions from diesel engines; July; Tsukuba Science City, Japan. (Developments in
28 toxicology and environmental science: v. 13.) Amsterdam: Elsevier Science Publishers BV; pp. 329-348
29 Ishinishi, N; Kuwabara. N; TaVaH, Y; st cl. (1288) Loiig-ienn inhalation experiments oa diesel exhaust. In:
3O Diesel exhaust and health risks: results of the HERP studies. Tsukuba, Ibaraki, Japan: Japan Automobile Research
31 Institute, Inc., Research Committee for HERP Studies; pp. 11-84.
32 Iwai, K; Udagawa, T; Yamagishi, M; et al. (1986) Long-term inhalation studies of diesel exhaust on F344 SPF
33 rats. Incidence of lung cancer and lymphoma. In: Ishinishi, N; Koizumi, A; McCiellan, RO; et al., eds.
34 Carcinogenic and mutagenic effects of diesel engine exhaust: proceedings of the international satellite symposium
35 on lexicological effects of emissions from diesel engines; July; Tsukuba Science City, Japan. (Developments in
36 toxicology and environmental science: v. 13.) Amsterdam: Elsevier Science Publishers BV: pp. 349-^60
37 Kaplan, HL; MacKenzie, WF; Springer, KJ; et al. (1982) A subchronic study of the effects of exposure of three
38 species of rodents to diesel exhaust. In: Lewtas, J, ed. Toxicological effects of emissions from diesel engines:
39 proceedings of the Environmental Protection Agency diesel symposium; October 1981; Raleigh, NC. New York:
40 Elsevier Biomedical; pp. 161-182.
1 1 75/99 6-24 DRAFT— DO NOT CITE OR QUOTE
-------
1 Kaplan, HL; Springer, KJ; MacKenzie, WF. (1983) Studies of potential health effects of long-term exposure to
2 diesel exhaust emissions. San Antonio, TX: Southwest Research Institute; SwRI project no. 01-0750-103.
Karagianes, MT; Palmer, RF; Busch, RH. (1981) Effects of inhaled diesel emissions and coal dust in rats. Am
Ind Hyg Assoc J 42:382-391.
5 Lewis, TR; Green, FHY; Moorman, WJ; et al. (1986) A chronic inhalation toxicity study of diesel engine
6 emissions and coal dust, alone and combined. In: Ishinishi, N; Koizumi, A; McClellan, RO; et al., eds.
7 Carcinogenic and mutagenic effects of diesel engine exhaust: proceedings of the international satellite symposium
8 on lexicological effects of emissions from diesel engines; July; Tsukuba Science City, Japan. (Developments in
9 toxicology and environmental science: v. 13.) Amsterdam: Elsevier Science Publishers BV; pp. 361-380
10 Lewis, TR; Green, FHY; Moorman, WJ; et al. (1989) A chronic inhalation toxicity study of diesel engine
11 emissions and coal dust, alone and combined. J Am Coll Toxicol 8:345-375.
12 Mauderly, JL; Jones, RK; Griffith, WC; et al. (1987a) Diesel exhaust is a pulmonary carcinogen in rats exposed
13 chronically by inhalation. Fundam Appl Toxicol 9:208-221.
14 Mauderly, JL; Bice, DE; Carpenter, RL; et al. (1987b) Effects of inhaled nitrogen dioxide and diesel exhaust on
15 developing lung. Cambridge, MA: Health Effects Institute; research report no. 8.
16 Mauderly, JL; Gillett, NA; Henderson, RF; et al. (1988) Relationships of lung structural and functional changes
17 to accumulation of diesel exhaust particles. In: Dodgson, J; McCallum, RI; Bailey, MR; et al., eds. Inhaled
18 particles VI: proceedings of an international symposium and workshop on lung dosimetry; September 1985;
19 Cambridge, United Kingdom. Ann Occup Hyg 32(suppl. l):659-669.
20 McClellan, RO; Bice, DE; Cuddihy, RG; et al. (1986) Health effects of diesel exhaust. In: Lee, SD; Schneider,
2J T; Grant, LD; et al., eds. Aerosols: research, risk assessment and control strategies: proceedings of the second
U.S.-Dutch international symposium; May 1985; Williamsburg, VA. Chelsea, MI: Lewis Publishers; pp.
597-615.
24 Misiorowski, RL; Strom, KA; Vostal, JJ; et al. (1980) Lung biochemistry of rats chronically exposed to diesel
25 particulates. In: Pepelko, WE; Danner, RM; Clarke, NA, eds. Health effects of diesel engine emissions:
26 proceedings of an international symposium; December 1979. Cincinnati, OH: U.S. Environmental Protection
27 Agency, Health Effects Research Laboratory; pp. 465-480; EPA report no. EPA-600/9-80-057a. Available from:
28 NTIS, Springfield, VA; PB81-173809.
29 Morrow, PE. (1992) Dust overloading of the lungs: update and appraisal. Toxicol Appl Pharmacol 113:1-12.
30 National Research Council. (1977) Drinking water and health. Washington, DC: National Academy of Sciences,
31 National Academy Press; pp. 801-804.
32 National Research Council. (1980) Drinking water and health, Vol. 2. Washington, DC: National Academy of
33 Sciences, National Academy Press.
34 National Research Council. (1983) Risk assessment in the federal government: managing the process.
35 Washington, DC: National Academy of Sciences, National Academy Press.
36 Nikula, KJ; Snipes, MB; Barr, EB; et al. (1995) Comparative pulmonary toxicities and carcinogenicities of
37 chronically inhaled diesel exhaust and carbon black in F344 rats. Fundam Appl Toxicol 25:80-94.
38 Ohanian, EV; Moore, JA; Fowle, JR; et al. (1997) Risk characterization: a bridge to informed decision making.
39 Fundam Appl Toxicol 39:81-88.
11/5/99 6-25 DRAFT-DO NOT CITE OR QUOTE
-------
1 Pepelko, WE; Peirano, WB. (1983) Health effects of exposure to diesel engine emissions: a summary of animal
2 studies conducted by the U.S. Environmental Protection Agency's Health Effects Research Laboratories at
3 Cincinnati, Ohio. J Am Coll Toxicol 2:253-306.
4- Pepelko, WE; Mattox, J; Moorman, WJ; et al. (1980) Pulmonary function evaluation of cats after one year of
5 exposure to diesel exhaust. In: Pepelko, WE; Danner, RM; Clarke, NA., eds. Health effects of diesel engine
6 emissions: proceedings of an international symposium, v. 2; December 1979; Cincinnati, OH. Cincinnati, OH:
7 U.S. Environmental Protection Agency, Health Effects Research Laboratory; pp. 757-765; EPA/600/9-80-057b.
8 Available from: NTIS, Springfield, VA; PB81-173817.
9 Stober, W. (1986) Experimental induction of tumors in hamsters, mice ?nd rats after long-term inhalation of
10 filtered and unfiltered diesel engine exhaust. In: Ishinishi, N; Koizumi, A; McClellan, RO; et al., eds.
11 Carcinogenic and mutagenic effects of diesel engine exhaust: proceedings of the international satellite symposium
12 on lexicological effects of emissions from diesel engines; July; Tsukuba Science City, Japan. (Developments in
13 toxicology and environmental science: v. 13.) Amsterdam: Elsevier Science Publishers B.V.; pp. 421-439.
14 Strom, KA. (1984) Response of pulmonary cellular defenses to the inhalation of high concentrations of diesel
15 exhaust. J Toxicol Environ Health 13:919-944.
16 U.S. Environmental Protection Agency. (1994) Methods for derivation of inhalation reference concentrations and
17 application of inhalation dosimetry. Research Triangle Park, NC: Office of Research and Development, National
18 Center for Environmental Assessment; EPA/600/8-90/066F.
19 U.S. Environmental Protection Agency. (1995a) Integrated Risk Information System (IRIS). Online. Cincinnati,
20 OH: National Center for Environmental Assessment.
21 U.S. Environmental Protection Agency. (1995b) The use of the benchmark dose approach in health risk
22 assessment. Washington, DC: Office of Research and Development, Risk Assessment Forum.
23 Werchowski, KM; Chaffee, VW; Briggs, GB. (1980a) Teratologic effects of long-term exposure to diesel exhaust
24 emissions (rats). Cincinnati, OH: U.S. Environmental Protection Agency, Health Effects Research Laboratory;
25 EPA/600/1-80-010. Available from: NTIS, Springfield, VA; PB80-159965.
26 Werchowski, KM; Henne, SP; Briggs, GB. (1980b) Teratologic effects of long-term exposure to diesel exhaust
27 emissions (rabbits). Cincinnati, OH: U.S. Environmental Protection Agency, Health Effects Research Laboratory;
28 EPA/600/1-80-011. Available from: NTIS, Springfield, VA; PB80-168529.
29 Wolff, RK; Henderson, RF; Snipes, MB; et al. (1987) Alterations in particle accumulation and clearance in lungs
30 of rats chronically exposed to diesel exhaust. Fundam Appl Toxicol 9-154-166.
31 Yu, CP; Xu, GB. (1986) Predictive mod*!? for deposition cf dicsel exiiausc particuiates in human and rat lungs.
32 Aerosol Sci Teehaol 5:337-347.
33 Yu, CP; Yoon, KJ. (1990) Retention modeling of diesel exhaust particles in rais and humans. Amherst, NY: State
34 University of New York at Buffalo (Health Effects Institute research report no. 40).
11/5/99 6-26 DRAFT—DO NOT CITE OR QUOTE
-------
7. CARCINOGENICITY OF DIESEL EXHAUST
7.1. INTRODUCTION
2 Initial health hazard concerns regarding the potential carcinogenicity of diesel exhaust were
3 based on the reported induction of skin papillomas by diesel particle extracts (Kotin et al., 1955),
4 evidence for mutagenicity of extracts (Huisingh et al., 1978), evidence that components of diesel
5 extract act as weak tumor promoters (Zamora et al., 1983), and the knowledge that diesel
6 particles and their associated organics are respirable. During the 1980s, both human
7 epidemiology studies and long-term animal cancer bioassays were initiated. In 1981, Waller
8 published the first epidemiologic investigation, a retrospective mortality study of London
9 transport workers. Since then a large number of cohort and case-control studies have been
10 carried out with railroad workers, dockworkers, truck drivers, construction workers, and bus
11 garage employees. During 1986 and 1987, several chronic animal cancer bioassays were
12 published. These and numerous laboratory investigations carried out since then have been
13 directed toward assessing the carcinogenic potential of whole exhaust, evaluating the importance
14 of various components of exhaust in the induction of cancer, and understanding the mode of
15 action and implications of deposition, retention, and clearance of diesel exhaust particles.
16 The purpose of this chapter is to evaluate the carcinogenic potential of diesel exhaust in
^P both animals (Section 7.3) and humans (Sections 7.1 and 7.2), determine likely mode/s of action
18 (Section 7.4), and provide an overall weight-of-evidence (Section 7.5) for carcinogenicity in
19 humans. This assessment focuses on diesel exhaust, although diesel particles comprise a portion
20 of all ambient particulate matter (PM). Although PM, notably PMIO (PM <.\im in diameter), has
21 been identified for many years as potentially impacting human health, these effects have been
22 evaluated in a separate document (EPA, 1996). This document is also undergoing revision.
23 In this section, various mortality and morbidity studies of the health effects of exposure to
24 diesel engine emissions are reviewed. Although an attempt was made to cover all the relevant
25 studies, a number of studies are not included for several reasons. First, the change from steam to
26 diesel engines in locomotives began in 1935 and was about 95% complete by 1959 (Garshick et
27 al., 1988). Diesel buses also were introduced about the same time. Therefore, exposure to diesel
28 exhaust was less common, and the followup period for studies conducted prior to 1959 (Raffle,
29 1957; Kaplan, 1959) was not long enough to cover the long latency period of lung cancer. The
30 usefulness of these studies in evaluating the carcinogenicity of diesel exhaust is greatly reduced;
31 thus, they are not considered here.
11/5/99 7-1 DRAFT-DO NOT CITE OR QUOTE
-------
1 Second, hypothesis-generating studies were excluded from this review because their
2 findings need subsequent confirmation by definitive studies (Silverman et al., 1983; Schenker et
3 al., 1984; Buiatti et al., 1985; Flodin et al., 1987; Siemiatycki et al., 1988; Swanson et al., 1993;
4 Cordier et al., 1993; Notani et al., 1993).
5 Third, studies in which exposure to diesel exhaust was uncertain or was defined as motor
6 exhaust (which includes both gasoline and diesel exhaust) were excluded because they would
7 have contributed little to the evaluation of the carcinogenicity of diesel exhaust (Waxweiler et al.,
8 1973; Ahlberg et al., 1981; Stern et al., 1981; Vineis and Magnani, 1985; Gustafsson et al., 1986;
9 Silverman et al., 1986; Jensen et al., 1987; Garland et al., 1988; Risch et al., 1988; Guberan et
10 al., 1992).
11 Fourth, a study by Coggon et al. (1984) was not included because the occupational
12 information abstracted from death certificates had not been validated; this would have resulted in
13 limited information.
14 Three types of studies of the health effects of exposure to diesel engine emissions are
15 reviewed in this chapter: (1) cohort studies, (2) case-control studies of lung cancer, and (3) case-
16 control studies of bladder cancer. In the cohort studies, the cohorts of heavy construction
17 equipment operators, railroad and locomotive workers, and bus garage employees were studied
18 retrospectively to determine increased mortality and morbidity resulting from exposures to
1 9 varying levels of diesel emissions in the workplace. A total of 9 cohort mortality studies (one of
20 the mortality studies also included a nested lung cancer case-control study), 10 lung cancer case-
21 control studies, and 7 bladder cancer case-control studies are considered in this section.
22
23 7.2. EPIDEMIOLOGIC STUDIES OF THE CARCINOGENICITY OF EXPOSURE TO
24 DIESEL EXHAUST
25 7.2.1. Cohort Studies
26 7.2.1.1. Waller (1981): Trends in Lung Cancer in London in Relation to Exposure to Diesel
27 Fumes
28 A retrospective mortality study of a cohort of London transport workers was conducted to
29 determine if there was an excess of deaths from lung cancer that could be attributed to diesel
30 exhaust exposure. Nearly 20,000 male employees aged 45 to 64 were followed for the 25-year
31 period between 1950 and 1974, constituting a total of 420,700 man-years at risk. These were
32 distributed among five job categories: drivers, garage engineers, conductors, motomien or
33 guards, and engineers (works). Most employees lived in the greater London area. Lung cancer
34 cases occurring in this cohort were ascertained only from death certificates of individuals who
11/5/99 7-2 DRAFT-DO NOT CITE OR QUOTE
-------
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-year periods and 5-year age ranges, finally combining the
4 results to obtain the total expected deaths in the required age range for the calendar period. A
5 total of 667 cases of lung cancer was reported, compared with 849 expected, to give a mortality
6 ratio of 79%. In each of the five job categories, the observed numbers were below those
7 expected. Engineers in garages had the highest mortality ratio (90%), but this did not differ
8 significantly from the other job categories. Environmental sampling was done at one garage, on
9 1 day in 1979, for benzo[a]pyrene concentrations and was compared with corresponding values
10 recorded in 1957. Concentrations of benzo[a]pyrene recorded in 1957 were at least 10 times
11 greater than those measured in 1979.
12 This study has several methodologic limitations. The lung cancer deaths ascertained for
13 the study occurred while the worker was employed (the worker either died of lung cancer or
14 retired after lung cancer was diagnosed). Although man-years at risk were based on the entire
1 5 cohort, no attempt was made to trace or evaluate the individuals who had resigned from the
16 London transport company for any other reason. Hence, information on resignees who may have
17 had significant exposure to diesel exhaust, and lung cancer deaths among them, was not available
«for analysis. This fact may have led to a dilution effect, resulting in underascertainment of
observed lung cancer deaths and underestimation of mortality ratios. Eligibility criteria for
20 inclusion in the cohort, such as starting date and length of service with the company, were not
21 specified. Because an external comparison group was used to obtain expected number of deaths,
22 the resulting mortality ratios were less than 1; this may be a reflection of the "healthy worker
23 effect." Investigators also did not categorize the five job categories by levels of diesel exhaust
24 exposure, nor did they use an internal comparison group to derive risk estimates.
25 The age range considered for this study was limited (45 to 64 years of age) for the period
26 between 1950 and 1964. It is not clear whether this age range was applied to calendar year 1950
27 or 1964 or at the midpoint of the 25-year followup period. No analyses were presented either by
28 latency or by duration of employment (surrogate for exposure). The environmental survey based
29 on benzo[a]pyrene concentrations suggests that the cohort in its earlier years was exposed to
30 much higher concentrations of environmental contaminants than currently exist. It is not clear
31 when the reduction in benzo[a]pyrene concentration occurred because there are no environmental
32 readings available between 1957 and 1979. It is also important to note that the concentrations of
33 benzo[a]pyrene inside the garage in 1957 were not very different from those outside the garage,
11/5/99 7-3 DRAFT-DO NOT CITE OR QUOTE
-------
1 thus indicating that exposure for garage workers was not much different from that of the general
2 population. Last, no data were collected on smoking habits.
3
4 7.2.1.2. Howe et al (1983): Cancer Mortality (1965 to 1977) in Relation to Diesel Fumes and
5 Coal Exposure in a Cohort of Retired Railroad Workers
6 This is a retrospective cohort study of the mortality experience of 43,826 male pensioners
7 of the Canadian National Railroad (CNR) between 1965 and 1977. Members of this cohort
8 consisted of male CNR pensioners who had retired before 1965 and who were known to be alive
9 at the start of that year, as well as those who retired between 1965 and 1977. The records were
10 obtained from a computer file that is regularly updated and used by the company for payment of
11 pensions. To receive a pension, each pensioner must provide, on a yearly basis, evidence that he
12 is alive. Specific cause of death among members of this cohort was ascertained by linking these
13 records to the Canadian Mortality Data Base, which contains records of all deaths registered in
14 Canada since 1950. Of the 17,838 deaths among members of the cohort between 1965 and 1977,
15 16,812 (94.4%) were successfully linked to a record in the mortality file. A random sample
16 manual check on unlinked data revealed that failure to link was due mainly to some missing
17 information on the death records.
18 Occupation at time of retirement was used by the Department of Industrial Relations to
19 classify workers into three diesel fume and coal dust exposure categories: (1) nonexposed, (2)
20 possibly exposed, and (3) probably exposed. Person-years of observation were calculated and
21 classified by age at observation in 5-year age groups (35 to 39,40 to 44,..., 80 to 84, and 2:85
22 years). The observed deaths were classified by age at death for different cancers, for all cancers
23 combined, and for all causes of death combined. Standard mortality ratios (SMRs) were then
24 calculated using rates of the Canadian population for the period between 1965 and 1977.
25 Both total mortality (SMR = 95, /?<0.001) and all cancer deaths (SMR = 99. o>0.05)
26 were close 10 that expected for the entire cohort. Analysis by exposure to diesel fume levels in
27 the three categories (nonexposed, possibly exposed, and probably exposed) revealed an increased
28 relative risk for lung cancer among workers with increasing exposure to diesel fumes. The
29 relative risk for nonexposed workers was presumed to be 1.0; for those possibly exposed, the
30 relative risk was elevated to 1.2, which was statistically significant (p=0.013); and, for those
31 probably exposed, it was elevated to 1.35, which was statistically highly significant (p=0.001).
32 The corresponding rates for exposure to varying levels of coal dust were very similar st 1.00,
33 1.21 (p=0.012), and 1.35 (p=0.001), respectively. The trend tests were highly significant for both
34 exposures (p<0.001). Analysis performed after the exclusion of individuals who worked in the
3 5 maintenance of steam engines, and hence were exposed to high levels of asbestos, yielded the
11/5/99 7-4 DRAFT—DO NOT CITE OR QUOTE
-------
1 risk of lung cancer to be 1.00, 1.21, and 1.33 for those nonexposed, possibly exposed, and
2 probably exposed to diesel exhaust, respectively, with a highly significant trend (pO.OO1).
^P An analysis done on individuals who retired prior to 1950 showed the relative risk of lung
4 cancer among nonexposed, possibly exposed, and probably exposed to be 1.00, 0.70, and 0.44,
5 respectively, based on fewer than 15 deaths in each category. A similar analysis of individuals
6 who retired after 1950 found the results in the same categories to be 1.00, 1.23, and 1.40,
7 respectively. Although retirement prior to 1950 indicated exposure to coal dust alone, retirement
8 after 1950 shows the results of mixed exposure to coal dust and diesel fumes. As there was
9 considerable overlap between occupations involving probable exposure to diesel fumes and
10 probable exposure to coal dust, and as most members of the cohort were employed during the
11 years in which the transition from coal to diesel occurred, it was difficult to distinguish whether
12 lung cancer was associated with exposure to coal dust or diesel fumes or a mixture of both.
13 Although this study showed a highly significant dose-response relationship between
14 diesel fumes and lung cancer, it has some methodological limitations. There were concurrent
15 exposures to both diesel fumes and coal dust during the transition period; therefore,
16 misclassification of exposure may have occurred, because only occupation at retirement was
17 available for analysis. It is possible that the elevated response observed for lung cancer was due
18 to the combined effects of exposure to both coal dust and diesel' fumes and not just one or the
"^^ other. However, it should be noted that so far coal dust has not been demonstrated to be a
20 pulmonary carcinogen in studies of coal miners. No information was provided on duration of
21 employment in either diesel work or the coal dust-related jobs for other than those jobs held at
22 retirement. Therefore, it was not possible to evaluate whether this omission would have led to an
23 under- or overestimate of the true relative risk. Furthermore, a lack of information on potential
24 confounders such as smoking makes interpretation of the excess risk of lung cancer even more
25 difficult. Information on cause of death was acquired from the mortality data linkage. There is a
26 possibility that the cause of death may have been misclassified because of miscoding of the
27 underlying cause of death.
28
29 7.2.1.3. Rushtonet aL (1983): Epidemiological Survey of Maintenance Workers in the
30 London Transport Executive Bus Garages and Chiswick Works
31 This is a retrospective mortality cohort study of male maintenance workers employed for
32 at least 1 continuous year between January 1,1967, and December 31,1975, at 71 London
33 transport bus garages (also known as rolling stock) and at Chiswick Works. For all men, the
34 folio wing-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
11/5/99 7-5 DRAFT-DO NOT CITE OR QUOTE
-------
1 left their job, date of and reason for leaving were also obtained. For those who died in service or
2 after retirement and for men who had resigned, full name and last known address were obtained
3 from an alphabetical card index in the personnel department. Additional tracing of individuals
4 who had left was carried out through social security records. The area of their residence was
5 assumed to be close to their work; therefore their place of work was coded as their residence.
6 One hundred different job titles were coded into 20 broader groups. These 20 groups were not
7 ranked for diesel exhaust exposure, however. The reason for leaving was coded as died in
8 service, retired, or other. The underlying cause of death was coded using the eighth revision of.
9 the International Classification of Diseases (ICD). Person-years were calculated from date of
10 birth and dates of entry to and exit from the- study using the man-years computer language
11 program. These were then subdivided into 5-year age and calendar period groups. The expected
12 number of deaths was calculated by applying the 5-year age and calendar period death rates of
13 the comparison population with the person-years of corresponding groups. The mortality
14 experience of the male population in England and Wales was used as the comparison population.
1 5 Significance values were calculated for the difference between the observed and expected deaths,
16 assuming a Poisson distribution.
17 The number of person-years of observation totaled 50,008 and was contributed by 8,490
18 individuals in the study with a mean followup of 5.9 years. Only 2.2% (194) of the men were
19 not traced. Observed deaths from all causes were significantly lower than expected (observed =
20 495, p<0.001). The observed deaths from all neoplasms and cancer of the lung were
21 approximately the same as those expected. The only significant excess observed for cancer of
22 the liver and gall bladder at Chiswick Works was based on four deaths (p<0.05). A few job
23 groups showed a significant excess of risks for various cancers. All the excess deaths observed
24 for the various job groups, except for the general hand category, were based on very small
25 numbers (usually smaller than five) and merited cautious interpretation. Only a notable excess in
26 the general hand category for lung cancer was based on 48 cases (SMR = 133- /?
-------
inherent problems of inaccuracy, misdiagnosis, and errors in coding, and it was not known
whether a trained nosologist coded the death certificates. No adjustments were made for the
confounding effects of smoking and socioeconomic factors.
4
5 7.2.1.4. Wong et aL (1985): Mortality Among Members of a Heavy Construction Equipment
6 Operators Union With Potential Exposure to Diesel Exhaust Emissions
7 This is a retrospective mortality study conducted on a cohort of 34,156 male members of
8 a heavy construction equipment operators union with potential exposure to diesel exhaust
9 emissions. Study cohort members were identified from records maintained at Operating
o
10 Engineers' Local Union No. 3-3 A in San Francisco, CA. This union has maintained both work
11 and death records on all its members since 1964. Individuals with at least 1 year of membership
12 in this union between January 1, 1964, and December 31, 1978, were included in the study.
13 Work histories of the cohort were obtained from job dispatch computer tapes. The study
14 followup period was January 1964 to December 1978. Death information was obtained from a
15 trust fund, which provided information on retirement dates, vital status, and date of death for
16 those who were entitled to retirement and death benefits. Approximately 50% of the cohort had
17 been union members for less than 15 years, whereas the other 50% had been union members for
15 years or more. The average duration of membership was 15 years. As of December 31, 1978,
29,046 (85%) cohort members were alive, 3,345 (9.8%) were dead, and 1,765 (5.2%) remained
20 untraced. Vital status of 10,505 members who had left the union as of December 31,1978, was
21 ascertained from the Social Security Administration. Death certificates were obtained from
22 appropriate State health departments. Altogether, 3,243 deaths (for whom death certificates were
23 available) in the cohort were coded using the seventh revision of the ICD. For 102 individuals,
24 death certificates could not be obtained, only the date of death; these individuals were included in
25 the calculation of the SMR for all causes of death but were deleted from the cause-specific SMR
26 analyses. Expected deaths and SMRs were calculated using the U.S. national age-sex-race
27 cause-specific mortality rates for 5-year time periods between 1964 and 1978. The entire cohort
28 population contributed to 372,525.6 person-years hi this 5-year study period.
29 A total of 3,345 deaths was observed, compared with 4,109 expected. The corresponding
30 SMR for all causes was 81.4 (p=0.01), which confirmed the "healthy worker effect." A total of
31 817 deaths was attributed to malignant neoplasms, slightly fewer than the 878.34 expected based
32 on U.S. white male cancer mortality rates (SMR = 93.0,p=0.05). Mostly there were SMR
33 deficits for cause-specific cancers, including lung cancer for the entire cohort (SMR = 98.6,
34 observed = 309). The only significant excess SMR was observed for cancer of the liver (SMR =
166.7, observed = 23./K0.05).
11 /5/99 7-7 DRAFT-DO NOT CITE OR QUOTE
-------
1 Analysis by length of union membership as a surrogate of duration for potential exposure
2 showed statistically significant increases in SMRs of cancer of the liver (SMR = 424, p
-------
The cause-specific mortality analysis in the low-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,/?<0.05) among those with 15 to 19
4 years of membership and length of followup of at least 20 years, and cancer of the liver (SMR =
5 871.9, observed = 3, /K0.05) among those with 10 to 14 years of membership and length of
6 followup of 10 to 19 years. There were 7,032 individuals who contributed to 78,402.9 person-
7 years of observation in the low-exposure group.
8 For the unknown exposure group, a statistically significant SMR was observed for motor
9 vehicle accidents only (SMR = 173.3, observed = 21,/?<0.05). There were 3,656 individuals
10 who contributed to 33,388.1 person-years of observation in this category.
11 No work histories were available for those who started their jobs before 1967 and for
12 those who held the same job prior to and after 1967. This constituted 9,707 individuals (28% of
13 the cohort) contributing to 104,447.5 person-years. Statistically significant SMR excesses were
14 observed for all cancers (SMR =112, observed = 339, /?<0.05) and cancer of the lung (SMR =
15 119.3, observed = 141,/7<0.01). A significant SMR elevation was also observed for cancer of
16 the stomach (SMR =199.1, observed = 30,/KO.Ol).
17 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
20 from the large intestine and/or rectum, since tumors of these sites will frequently metastasize to
21 the liver. The excess in liver cancer mortality and the deficits in mortality that are due to cancer
22 of the large intestine and rectum could also, as the authors indicate, be due to misclassification.
23 Both possibilities have been considered by the investigators in their discussion.
24 Cancer of the lung showed a positive trend with length of membership as well as with
25 latency, although none of the SMRs were statistically significant except for the workers without
26 any work histories. The individuals without any work histories may have been the ones who
27 were in their jobs for the longest period of tune, because workers without job histories included
28 those who had the same job before and after 1967 and thus may have worked 12 to 14 years or
29 longer. If they had belonged to the category hi which heavy exposure to diesel exhaust
30 emissions was very common for this prolonged time, then the increase hi lung cancer, as well as
31 stomach cancer, might be linked to diesel exhaust. Further information on those without work
32 histories should be obtained if possible because such information may be quite informative with
33 regard to the evaluation of the carcinogenicity of diesel exhaust.
34 The study design is adequate, covers about a 15-year observation period, has a large
enough population, and is appropriately analyzed; however, it has too many limitations to permit
11/5/99 7-9 DRAFT—DO NOT CITE OR QUOTE
-------
1 any conclusions. First, no exposure histories are available. One has to make do with job
2 histories, which provide limited information on exposure level. Any person who ever worked at
3 the job or any person working at the same job over any period of time is included in the same
4 category; this would have a dilution effect, since extremely variable exposures were considered
5 in the study. Second, the length of time worked in any particular job is not available. Third,
6 work histories were not available for 9,707 individuals, who contributed 104,447.5 person-years,
7 a large proportion of the study cohort (28%). These individuals happen to show the most
8. evidence of a carcinogenic effect. Confounding by alcohol consumption for cancer of the liver
9 and smoking for emphysema and cancer of the lung was not ruled out. Last, although 34,156
10 members were eligible for the study, the vital status of 1,765 individuals was unknown.
11 Nevertheless, they were still considered in the denominator of all the analyses. The investigators
12 fail to mention how the person-year calculation for these individuals was handled. Also, some of
13 the person-years might have been overestimated, as people may have paid the dues for a
14 particular year and then left work. These two causes of overestimation of the denominator may
1 5 have resulted in some or all the SMRs being underestimated.
16 As for the smoking survey, the investigators took a very small sample (133 out of 34,156,
17 which was not even 1%). Of 133, only 107 (80%) participated. It was a systematic sample, but
18 the authors neglected to mention how the list was prepared. Hence, the sample may not be
1 9 representative of the study population and, with a small sample size, the results are not
20 generalizable. The questionnaire asked only for current smoking history. No detailed history
21 was obtained for the amount smoked or length of smoking history, both of which have a bearing
22 on emphysema as well as lung carcinoma.
23
24 7.2.1.5. Edling et aL (1987): Mortality Among Personnel Exposed to Diesel Exhaust
25 This is a retrospective cohort mortality study of bus company employees, which
26 investigated a possible increased mortality in cardiovascular diseases and cancers from diesel
27 exhaust exposure. The cohort comprised all males employed at five different bus companies in
28 southeastern Sweden between 1950 and 1959. Based on infonnation from, personnel registers,
29 individuals were classified into one or more categories and could have contributed person-years
30 at risk hi more than one exposure category. The study period was from 1951 to 1983;
31 information was collected from the National Death Registry, and copies of death certificates
32 were obtained from the National Bureau of Statistics. Workers who died after age 79 were
33 excluded from the study because diagnostic procedures were likely to be more uncertain at
34 higher ages (according to investigators). The cause-, sex-, and age-specific national death rates
35 in Sweden were applied to the 5-year age categories of person-years of observation to determine
11/5/99 7-10 DRAFT-DO NOT CITE OR QUOTE
-------
expected deaths for all causes, malignant diseases, and cardiovascular diseases. A Poisson
distribution was used to calculate p-values and confidence limits for the ratio of observed to
expected deaths. The total cohort of 694 men (after loss of 5 men to followup) was divided into
4 three exposure categories: (1) clerks with lowest exposure, (2) bus drivers with moderate
5 exposure, and (3) bus garage workers with highest exposure.
6 The 694 men provided 20,304 person-years of observation, with 195 deaths compared
7 with 237 expected. A deficit in cancer deaths largely accounted for this lower-than-expected
8 mortality in the total cohort. Among subcohorts, no difference between observed and expected
9 deaths for total mortality, total cancers, or cardiovascular causes was observed for clerks (lowest
10 diesel exposure), bus drivers (moderate diesel exposure), and garage workers (high diesel
11 exposure). The risk ratios for all three categories were less than 1 except for cardiovascular
12 diseases among bus drivers, which was 1.1.
13 When the analysis was restricted to members who had at least a 10-year latency period
14 and either any exposure or an exposure exceeding 10 years, similar results were obtained, with
1 5 fewer neoplasms than expected, whereas cardiovascular diseases showed risk around or slightly
16 above unity.
17 Five lung cancer deaths were observed among bus drivers who had moderate diesel
exhaust exposure, whereas 7.2 were expected. The only other lung cancer death was observed
among bus garage workers who had the highest diesel exhaust exposure. The small size of the
20 cohort and poor data on diesel exhaust exposure are among the major limitations of this study.
21 Although lifetime occupational histories were available, no industrial hygiene data were
22 presented to validate the classification of workers into low, moderate, and high exposure to diesel
23 exhaust based on job title. The power of the present study was estimated to be 80% to detect a
24 relative risk of 1.2 for cardiovascular diseases and 1.4 for cancers, but for specific cancer sites,
25 the power was much lower than this. No information was available on confounding effects of
26 smoking and asbestos exposure at the work sites.
27
28 7.2.1.6. Boffetta and Stellman (1988): Diesel Exhaust Exposure and Mortality Among Males
29 in the American Cancer Society Prospective Study
30 Boffetta and Stellman conducted a mortality analysis of 46,981 males whose vital status
31 was known at the end of the first 2 years of followup. The analysis was restricted to males aged
32 40 to 79 years hi 1982 who enrolled in the American Cancer Society's prospective mortality
33 study of cancer. Mortality was analyzed hi relation to exposure to diesel exhaust and to
34 employment in selected occupations related to diesel exhaust exposure. In 1982, more than
77,000 American Cancer Society volunteers enrolled over 1.2 million men and women from all
11/5/99 7-11 DRAFT-DO NOT CITE OR QUOTE
-------
1 50 states, the District of Columbia, and Puerto Rico in a long-term cohort study, the Cancer
2 Prevention Study II (CPS-II). Enrollees were usually friends, neighbors, or relatives of the
3 volunteers; enrollment was by family groups with at least one person in the household 45 years
4 of age or older. Subjects were asked to fill out a four-page confidential questionnaire and return
5 it in a sealed envelope. The questionnaire included history of cancer and other diseases; use of
6 medications and vitamins; menstrual and reproductive history; occupational history; and
7 information on diet, drinking, smoking, and other habits. The questionnaire also included three
8 questions on occupation: (1) current occupation, (2) last occupation, if retired, and (3) job held
9 for the longest period of time, if different from the other two. Occupations were coded to an ad
10 hoc two-digit classification in 70 categories. Exposures at work or in daily life to any of the 12
11 groups of substances were also ascertained. These included diesel engine exhausts, asbestos,
12 chemicals/acids/solvents, dyes, formaldehyde, coal or stone dusts, and gasoline exhausts.
13 Volunteers checked whether their enrollees were alive or dead and recorded the date and place of
14 all deaths every other year during the study. Death certificates were then obtained from State
1 5 health departments and coded according to a system based on the ninth revision of the ICD by a
16 trained nosologist.
17 The data were analyzed to determine the mortality for all causes and lung cancer in
18 relation to diesel exhaust exposure, mortality for all causes and lung cancer in relation to
1 9 employment hi selected occupations with high diesel exhaust exposure, and mortality from other
20 causes in relation to diesel exhaust exposure. The incidence-density ratio was used as a measure
21 of association, and test-based confidence limits were calculated by the Miettinen method. For
22 stratified analysis, the Mantel-Haenszel method was used for testing linear trends. Data on
23 476,648 subjects comprising 939,817 person-years of risk were available for analysis. Three
24 percent of the subjects (14,667) had not given any smoking history, and 20% (98,026) of them
25 did not give information on diesel exhaust exposure and were therefore excluded from the main
26 diesel exhaust analysis. Among individuals who had provided diesel exhaust exposure history,
27 62,800 were exposed and 307,143 were not exposed. Comparison of the population with known
28 information on diesel exhaust exposure with the excluded population with no information en
29 diesel exhaust exposure showed that the mean ages were 54.7 and 57.7 years, the nonsmokers
30 were 72.4% and 73.2%, and the total mortality rates per 1,000 per year were 23.0% and 28.8%,
31 respectively.
32 The all-cause mortality was elevated among railroad workers (relative risk [RR] = 1.43,
33 95% confidence interval [CI] = 1.2, 1.72), heavy equipment operators (RR = 1.7, 95% CI = 1.19,
34 2.44), miners (RR = 1.34, 95% CI = 1.06,1.68), and truck drivers (RR =1.19, 95% CI = 1.07,
35 1.31). For lung cancer mortality the risks were significantly elevated for miners (RR = 2.67,
11/5/99 7-12 DRAFT-DO NOT CITE OR QUOTE
-------
1 95% CI = 1.63, 4.37) and heavy equipment operators (RR = 2.60, 95% CI = 1.12, 6.06). Risks
4 were also elevated but not significantly for railroad workers (RR = 1.59, 95% CI = 0.94, 2.69)
and truck drivers (RR = 1.24, 95% CI = 0.93, 1.66). These risks were calculated according to the
4 Mantel-Haenszel method, controlling for age and smoking. Although the relative risk was
5 nonsignificant for truck drivers, a small dose-response effect was observed when duration of
6 diesel exhaust exposure for them was examined. For drivers who worked for 1 to 15 years, the
7 relative risk was 0.87, while for drivers who worked for more than 16 years, the relative risk was
8 1.33 (95% CI = 0.64,2.75). Relative risks for lung cancer were not presented for other
9 occupations. Mortality analysis for other causes and diesel exhaust exposure showed a
10 significant excess of deaths (p<0.05) in the following categories: cerebrovascular disease,
11 arteriosclerosis, pneumonia, influenza, cirrhosis of the liver, and accidents.
12 The two main methodologic concerns in this study are the representativeness of the study
13 population and the quality of information on exposure. The sample, though very large, was
14 composed of volunteers. Thus, the cohort was healthier and less frequently exposed to important
1 5 risk factors such as smoking and alcohol. Self-administered questionnaires were used to obtain
16 data on occupation and diesel exhaust exposure. None of this information was validated. Nearly
17 20% of the individuals had an unknown exposure status to diesel exhaust, and they experienced a
18 higher mortality for all causes and lung cancer than both the diesel exhaust exposed and
19 unexposed groups. This could have introduced a substantial bias in the estimate of the
20 association. Although only 0.8% of the subjects were lost to followup, the use of death
21 certificates alone as a source of medical information poses problems in accuracy and coding. But
22 the authors report that cancer deaths are routinely checked by histological confirmation from
23 physicians or cancer registries. Given the fact that all diesel exhaust exposure occupations, such
24 as heavy equipment operators, truck drivers, and railroad workers, showed elevated lung cancer
25 risk, this study is suggestive of a causal association.
26
27 7.2.1.7. Garshick et al (1988): A Retrospective Cohort Study of Lung Cancer and Diesel
28 Exhaust Exposure in Railroad Workers
29 An earlier case-control study of lung cancer and diesel exhaust exposure in U.S. railroad
30 workers by these investigators had demonstrated a relative odds of 1.41 (95% CI = 1.06, 1.88)
31 for lung cancer with 20 years of work in jobs with diesel exhaust exposure. To confirm these
32 results, a large retrospective cohort mortality study was conducted by the same investigators.
33 Data sources for the study were the work records of the U.S. Railroad Retirement Board (RRB).
34 The cohort was selected based on job titles in 1959, which was the year by which 95% of the
locomotives in the United States were diesel powered. Diesel exhaust exposure was considered
11/5/99 7-13 DRAFT-DO NOT CITE OR QUOTE
-------
1 to be a dichotomous variable depending on yearly job codes between 1959 and death or
2 retirement through 1980. Industrial hygiene evaluations and descriptions of job activities were
3 used to classify jobs as exposed or unexposed to diesel emissions. A questionnaire survey of 534
4 workers at one of the railroads where workers were asked to indicate the amount of time spent in
5 railroad locations, either near or away from sources of diesel exhaust, was used to validate this
6 classification. Workers selected for this survey were actively employed at the time of the survey,
7 40 to 64 years of age, who started work between 1939 and 1949, in the job codes sampled in
8 1959, and were eligible for railroad benefits. To qualify for benefits, a worker must have had 10
9 years or more of service with the railroad and should not have worked for more than 2 years in a
10 nonrailroad job after leaving railroad work. Workers with recognized asbestos exposure, such as
11 repair of asbestos-insulated steam locomotive boilers, passenger cars, and steam pipes, or
12 railroad building construction and repairs, were excluded from the job categories selected for
13 study. However, a few jobs with some potential for asbestos exposure were included in the
14 cohort, and the analysis was done both ways, with and without them.
1 5 The death certificates for all subjects identified in 1959 and reported by the RRB to have
1 6 died through 1980 were searched. Twenty-five percent of them were obtained from the RRB and
1 7 the remainder from the appropriate State departments of health. Coding of cause of death was
18 done without knowledge of exposure history, according to the eighth revision of the ICD. If the
19 underlying cause of death was not lung cancer, but was mentioned on the death certificate, it was
20 assigned as a secondary cause of death, so that the ascertainment of all cases was complete.
21 Workers not reported by the RRB to have died by December 31,1980, were considered to be
22 alive. Deceased workers for whom death certificates had not been obtained or, if obtained, did
23 not indicate cause of death, were assumed to have died of unknown causes.
24 Proportional hazard models were fitted that provided estimates of relative risk for death
25 caused by lung cancer using the partial likelihood method described by Cox, and 95% confidence
26 intervals were constructed using the asymptotic normality of the estimated regression
27 coefficients of the proportional hazards model. Exposure was analyzed by diesel exhaust-
28 exposed jobs hi 1959 and by cumulative number of years of diesel exhaust exposure through
29 1980. Directly standardized rate ratios for deaths from lung cancer were calculated for diesel
30 exhaust exposed compared with unexposed for each 5-year age group in 1959. The standardized
31 rates were based on the overall 5-year person-year time distribution of individuals in each age
32 group starting in 1959. The only exception tc this was between 1979 and 19805 when a 2-year
33 person-year distribution was used. The Mantel-Haenszel analogue for person-year data was used
34 to calculate 95% confidence intervals for the standardized rate ratios.
11/5/99 7-14 DRAFT-DO NOT CITE OR QUOTE
-------
The cohort consisted of 55,407 workers, 19,396 of whom had died by the end of 1980.
Death certificates were not available for 11.7% of all deaths. Of the 17,120 deaths for whom
death certificates were obtained, 48.4% were attributable to diseases of the circulatory system,
4 whereas 21% were attributable to all neoplasms. Of all neoplasms, 8.7% (1,694 deaths) were due
5 to lung cancer. A higher proportion of workers in the younger age groups, mainly brakemen and
6 conductors, were exposed to diesel exhaust, while a higher proportion of workers in the older age
7 groups were potentially exposed to asbestos. In a proportional hazards model, analyses by age in
8 1959 found a relative risk of 1.45 (95% CI = 1.11, 1.89) among the age group 40 to 44 years and
9 a relative risk of 1.33 (95% CI = 1.03, 1.73) for the age group 45 to 49 years. Risk estimates in
10 the older age groups 50 to 54, 55 to 59, and 60 to 64 years were 1.2, 1.18, and 0.99, respectively,
11 and were not statistically significant. The two youngest age groups in 1959 had workers with the
12 highest prevalence and longest duration of diesel exhaust exposure and lowest exposure to
13 asbestos. When potential asbestos exposure was considered as a confounding variable in a
14 proportional hazards model, the estimates of relative risk for asbestos exposure were all near null
15 value and not significant. Analysis of workers exposed to diesel exhaust in 1959 (n = 42,535),
16 excluding the workers with potential past exposure to asbestos, yielded relative risks of 1.57
17 (95% CI = 1.19,2.06) and 1.34 (95% CI = 1.02,1.76) in the 1959 age groups 40 to 44 years and
45 to 49 years. Directly standardized rate ratios were also calculated for each 1959 age group
based on diesel exhaust exposure in 1959. The results obtained confirmed those obtained by
20 using the proportional hazards model.
21 Relative risk estimates were then obtained using duration of diesel exhaust exposure as a
22 surrogate for dose. In a model that used years of exposure up to and including exposure in the
23 year of death, no exposure duration-response relationship was obtained. When analysis was done
24 by disregarding exposure in the year of death and 4 years prior to death, the risk of dying from
25 lung cancer increased with the number of years worked in a diesel-exhaust-exposed job. In this
26 analysis, exposure to diesel exhaust was analyzed by exposure duration groups and in a model
27 entering age in 1959 as a continuous variable. The workers with greater than 15 years of
28 exposure had a relative risk of lung cancer of 1.72 (95% CI = 1.27,2.33). The risk for 1 to 4
29 years of cumulative exposure was 1.20 (95% CI = 1.01,1.44); for 5 to 9 years of cumulative
30 exposure, it was 1.24 (95% CI = 1.06,1.44); and for 10 to 14 years of cumulative exposure, it
31 was 1.32 (95% CI = 1.13,1.56). Directly standardized rate ratios were also calculated for each
32 1959 age group based on diesel exposure in 1959. The results obtained confirmed those obtained
33 by using the proportional hazards model.
34 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
11/5/99 7-15 DRAFT-DO NOT CITE OR QUOTE
-------
1 conducted by the same investigators in railroad workers dying of lung cancer from March 1981
2 through February 1982. This cohort study has addressed many of the weaknesses of the other
3 epidemiologic studies. The large sample size (60,000) allowed sufficient power to detect small
4 risks and also permitted the exclusion of workers with potential past exposure to asbestos. The
5 stability of job career paths in the cohort ensured that of the workers 40 to 44 years of age in
6 1959 classified as diesel exhaust-exposed, 94% of the cases were still in diesel exhaust-exposed
7 jobs 20 years later.
8 The main limitation of the study is the lack of quantitative data on exposure to diesel
9 exhaust. This is one of the few studies in which industrial hygiene measurements of diesel
10 exhaust were done. These measurements were correlated with job titles to divide the cohort in
11 dichotomous exposure groups of exposed and nonexposed. This may have led to an
12 underestimation of the risk of lung cancer since exposed groups included individuals with low to
13 high exposure. The number of years exposed to diesel exhaust was used as a surrogate for dose.
14 The dose, based on duration of employment, may have been inaccurate because individuals were
1 5 working on steam or diesel locomotives during the transition period. If the categories of
16 exposure to diesel exhaust had been set up as no, low, moderate, and high exposure, the results
17 would have been more meaningful and so would have been the dose-response relationship.
18 Another limitation of this study was the inability to examine the effect of years of exposure and
19 latency. No adjustment for smoking was made in this study. However, an earlier case-control
20 study done in the same cohort (Garshick et al., 1987) showed no significant difference in the risk
21 estimate after adjusting for smoking. Despite these limitations, the results of this study
22 demonstrate that occupational exposure to diesel exhaust is associated with a modest risk (1.5) of
23 lung cancer.
24
25 7.2.1.8. Gustavsson et aL (1990): Lung Cancer and Exposure to Diesel Exhaust Among Bus
26 Garage Workers
27 A retrospective mortality study (from 1952 to 1986), cancer incidence study (from 1958
28 to 1984), and nested case-control study were conducted among a cohort of 708 male workers
29 from five bus garages in Stockholm, Sweden, who had worked tor at least 6 months between
30 1945 and 1970. Thirteen individuals were lost to followup, reducing the cohort to 695.
31 Information was available on location of workplace, job type, and beginning and ending
32 of work periods. Workers were traced using a computerized register of the living population,
33 death and burial books, and data from the Stockholm city archives.
34 For the cohort mortality analyses, death rates of the general population of greater
35 Stockholm were used. Death rates of occupationally active individuals, a subset of the general
1175/99 7-16 DRAFT-DO NOT CITE OR QUOTE
-------
1 population of greater Stockholm, were used as a second comparison group to reduce the bias
4 from "healthy worker effect." Mortality analysis was conducted using the "occupational
mortality analysis program" (OCMAP-PC). For cancer incidence analysis, the "epidemiology in
4 Linkoping" (EPILIN) program was used, with the incidence rates obtained from the cancer
5 registry.
6 For the nested case-control study, both dead and incident primary lung cancers, identified
7 in the register of cause of deaths and the cancer register, were selected as cases (20). Six controls
8 matched on age ±2 years, selected from the noncases at the time of the diagnosis of cases, were
9 drawn at random without replacements. Matched analyses were done to calculate odds ratios
10 using conditional logistic regression. The EGRET and Epilog programs were used for these
11 analyses.
12 Diesel exhaust and asbestos exposure assessments were performed by industrial
13 hygienists based on the intensity of exposure to diesel exhaust and asbestos, specific for
14 workplace, work task, and calendar time period. A diesel exhaust exposure assessment was
1 5 based on (1) amount of emission (number of buses, engine size, running time, and type of fuel),
16 (2) ventilatory equipment and air volume of the garages, and (3) job types and work practices.
17 Based on detailed historical data and very few actual measurements, relative exposures were
estimated (these were not absolute exposure levels). The scale was set to 0 for unexposed and 1
for lowest exposure, with each additional unit increase corresponding to a 50% increase in
20 successive intensity (i.e., 1.5,2.25, 3.38, and 5.06).
21 Based on personal sampling of asbestos during 1987, exposures were estimated and time-
22 weighted annual mean exposures were classified on a scale of three degrees (0, 1, and 2).
23 Cumulative exposures for both diesel exhaust and asbestos were calculated by multiplying the
24 level of exposure by the duration of every work period. An exposure index was calculated by
25 adding for every individual contributions from all work periods for both diesel exhaust and
26 asbestos. Four diesel exhaust index classes were created: 0 to 10,10 to 20, 20 to 30, and >30.
27 The four asbestos index classes were 0 to 20,20 to 40,40 to 60, and >60. The cumulative
28 exposure indices were used for the nested case-control study.
29 Excesses were observed for all cancers and some other site-specific cancers using both
30 comparison populations for the cohort mortality study, but none of them was statistically
31 significant. Based on 17 cases, SMR for lung cancer were 122 and 115 using Stockholm
32 occupationally active and general population, respectively. No dose-response was observed with
33 increasing cumulative exposure. The cancer incidence study reportedly confirmed the mortality
34 results (results not given).
11/5/99 . 7-17 DRAFT-DO NOT CITE OR QUOTE
-------
1 The nested case-control study showed increasing risk of lung cancer with increasing
2 exposure. Weighted linear regression gave RRs of 1.34 (95% CI = 1.09 to 1.64), 1.81 (95% CI =
3 1.20 to 2.71), and 2.43 (95% CI = 1.32 to 4.47) for the diesel exhaust indices 10 to 20, 20 to 30,
4 and >30, respectively, using 0 to 10 as the comparison group. The study was based on 17 cases
5 and six controls for each case matched on age ± 2 years. The results from conditional logistic
6 regression were similar to those obtained by weighted linear regression, but none was statistically
7 significant. Adjustment for asbestos exposure did not change the lung cancer risk for diesel
8 exhaust.
9 The main strength of this study is the detailed exposure matrices constructed for both
10 diesel exhaust and asbestos exposure, although they were based primarily on job tasks and very
11 few actual measurements. There are a few methodological limitations to this study. The cohort
12 is small and there were only 17 lung cancer deaths; thus the power is low. Exposure or outcome
13 may be misclassified, although any resulting bias in the relative risk estimates is likely to be
14 toward unity, because exposure classification was done independently of the outcome. Although
1 5 the analysis by dose indices was done, no latency analysis was performed. Finally, data on
1 6 smoking were missing, thus potentially confounding the lung cancer results. The authors suggest
17 that even the heaviest smoking among individuals who were heavily exposed to diesel exhaust
18 will be unable to explain the excess relative risk of 2.4 observed in this group. This may be an
19 overstatement, however, as cigarette smoking is a very strong risk factor for lung cancer.
20 Overall, this study provides some support to the excess lung cancer results found earlier among
21 populations exposed to diesel exhaust.
22
23 7.2.1.9. Hansen (1993): A Followup Study on the Mortality of Truck Drivers
24 This is a retrospective cohort mortality study of unskilled male laborers, ages 15 to 74
25 years, in Denmark, identified from a nationwide census file of November 9, 1970. The exposed
26 group included all truck drivers employed in the road delivery or long-haul business (14,225).
27 The unexposed group included all laborers in certain selected occupational groups considered to
28 be unexposed to fossil fuel combustion products and to resemble truck drivers in terms of work-
29 related physical demands and various personal background characteristics (43,024)
30 Through automatic record linkage between the 1970 census register (the Central
31 Population Register 1970 to 1980) and the Death Certificate Register (1970 to 1980), the
32 population was followed for cause-specific mortality or emigration up to November 9, 1980.
33 Expected number of deaths among truck drivers was calculated by using the 5-year age group
34 and 5-year tune period death rates of the unexposed group and applying them to the person-years
35 accumulated by truck drivers. ICD Revision 8 was used to code the underlying cause of death.
11/5/99 7-18 DRAFT-DO NOT CITE OR QUOTE
-------
1 Test-based CIs were calculated using Miettinen's method. A Poisson distribution was assumed
4 for the smaller numbers, and CI was calculated based on exact Poisson distribution (Ciba-Geigy).
Total person-years accrued by truck drivers were 138,302, whereas for the unexposed population,
4 they were 407,780. There were 627 deaths among truck drivers and 3,811 deaths in the
5 unexposed group. Statistically significant excesses were observed for all cancer mortality (SMR
6 = 121, 95% CI = 104 to 140); cancer of respiratory organs (SMR = 160, 95% CI = 128 to 198),
7 which mainly was due to cancer of bronchus and lung (SMR = 160, 95% CI = 126 to 200); and
8 multiple myeloma (SMR = 439, 95% CI = 142 to 1,024). When lung cancer mortality was
9 further explored by age groups, excesses were observed in most of the age groups (30 to 39, 45
10 to 49, 50 to 54, 55 to 59, 60 to 64, and 65 to 74), but there were small numbers of deaths in each
11 group when stratified by age, and the excesses were statistically significant for the 55 to 59
12 (SMR = 229, 95% CI = 138 to 358) and 60 to 64 (SMR = 227, 95% CI = 142 to 344) age groups
13 only.
14 As acknowledged by the author, the study has quite a few methodologic limitations. The
1 5 exposure to diesel exhaust is assumed in truck drivers based on diesel-powered trucks, but no
16 validation of qualitative or quantitative exposure is attempted. It is also not known whether any
17 of these truck drivers or any other laborers had changed jobs after the census of November 9,
«1970, thus creating potential misclassification bias in exposure to diesel exhaust. The lack of
smoking data and a 36% rural population (usually consuming less tobacco) in the unexposed
20 group further confound the lung cancer results. The followup period is relatively short, and a
21 latency analysis was not attempted. At best, the findings of this study are consistent with the
22 findings of other truck driver studies.
23 Table 7-1 summarizes the foregoing cohort studies.
24
25 7.2.2. Case-Control Studies of Lung Cancer
26 7.2.2.1. Williams et al (1977): Associations of Cancer Site and Type With Occupation and
27 Industry From the Third National Cancer Survey Interview
28 This paper reports findings of the analysis of the Third National Cancer Survey (TNCS).
29 The lifetime histories, occupations, and industries were studied for associations with specific
30 cancer sites and types after controlling for age, sex, race, education, use of cigarettes or alcohol,
31 and geographic location. Of 13,179 cancer patients, a 10% random sample of all incident
32 invasive cancers in eight areas, a total of 7,518 were successfully interviewed in the 3 years
33 surveyed by the TNCS. These comprised 57% of those eligible to participate. The interview
34 included items on use of tobacco and alcohol (by type, amount, and duration), family income,
patient education, and employment history. Actual descriptions of the occupation and industry
1175/99 7-19 DRAFT-DO NOT CITE OR QUOTE
-------
Table 7-1. Epidemiologic studies of the health effects of exposure to diesel exhaust: cohort mortality studies
Authors
Population studied
Diesel exhaust exposure
Results
Limitations
Five job categories used to SMR = 79 for lung cancer for the Exposure measurement of
Waller
(1981)
Approximately 20,000 male
London transportation workers
Aged 45 to 64 years
25 years followup (1950-1974)
define exposure
total cohort
Environmental SMRs for all five job categories
benzo[a]pyrene were less than 100 for lung
concentrations measured in cancer
1957 and 1979
benzo[a]pyrene showed very little .
difference between inside and outside
the garage
Incomplete information on cohort
members
N>
o
No adjustment for confounding such
as other exposures, cigarette smoking,
etc.
No latency analysis
O
O
O
Howe et al. 43,826 male pensioners of the
(1983) Canadian National Railway
Company
Mortality between 1965 and
1977 among these pensioners
was compared with mortality
of general Canadian population.
Exposure groups
classified by a group
of experts based on
occupation at the time
of retirement
Three exposure groups:
Nonexposed
Possibly exposed
Probably exposed
RR= 1.2(^=0.013) and
RR= 1.3^=0.001) for lung
cancer for possible and probable
exposure, respectively
A highly significant
dose-response relationship
demonstrated by trend
test(p
-------
VO
Table 7-1. Epidemiologic studies of the health effects ot exposure to diesel exhaust: cohort mortality studies
(continued)
Authors
Population studied
Diesel exhaust exposure
Results
Limitations
Rushton 8,490 male London transport
etal. (1983) maintenance workers
100 different job titles were
grouped in
20 broad categories
Mortality of workers employed for
1 continuous year between January The categories were not
1, 1967, and December 31, 1975, ranked for diesel exhaust
was compared with mortality of exposure
general population of England and
Wales
SMR = 133 (p<0.03) for lung
cancer in the general hand job
group
Several other job
categories showed SS increased
SMRs for several other sites
based on fewer than five cases
Ill-defined diesel exhaust exposure
without any ranking
Average 6-year followup (i.e., not
enough time for lung cancer latency)
No adjustment for confounders such
Wong et al. 34,156 male heavy construction
(1985) equipment operators
Members of the local union for
at least 1 year between
January 1, 1964, and December 1,
1978
20 functional job titles
grouped into three job
categories for potential
exposure
SMR =166 (p<0.05) for liver
cancer for total cohort
No validation of exposure categories,
which were based on surrogate
information
SMR = 343 (observed = 5,
/?<0.05) for lung cancer for high- Incomplete employment records
Exposure groups (high, low, exposure bulldozer operators with
and unknown) based on job 15-19 years of membership, 20+ Employment history other than .from
description and proximity to years of followup the union not available
source of diesel exhaust
emissions SMR =119 (observed =141, No data on confounders such as other
p<0.01) for workers with no work exposures, smoking, etc.
histories
-------
Table 7-1. Epidemiolcgic studies of the health effects of exposure to diesel exhaust: cohort mortality studies
^! (continued;
Auihori Population studied Diesel exhaust exposure Results Limitations
Edlingetiil. 694 male bus garage employees Three exposure groups No SS differences were observed Small sample size
(1981) based on job titles: between observed and expected
Followup from 19 il through 1983 High exposure, bus for any cancers by different No validation of exposure
garage workers exposure groups
Mortality of thes; men was Intermediate exposure, No data on confounders such as other
compared with mortality of bus drivers exposures, smoking, etc.
general populatiot. of Sweden Low exposure, clerks
M Doff ;tta and 46,981 male volunteers enrolled in Self-reported occupations Total mortality (SS) elevated for Exposure information based on self-
Stelhian the American Cancer Society=s were coded into 70 job railroad workers, heavy reported occupation for which no
(1983) Prospective Mortality Study of categories equipment operators, miners, and validation was done
Cancer in 1982 truck drivers
Employment in high diesel Volunteer population, probably
Aged 40 to 79 ysars at enrollment exhaust exposure jobs were Lung cancer mortality (SS) healthy population
compared with nonexposed elevated for miners and heavy
First 2-year followup jobs equipment operators
Lung cancer mortality (SNS)
O • elevated among railroad workers
-------
/O
c
Table 7-1. Epidemiologic studies of the health effects of exposure to diesel exhaust: cohort mortality studies
(continued)
Authors
Population studied
Diesel exhaust exposure
Results
Limitations
-j
to
OJ
Garshick 55,407 white male railroad
etal. (1988) workers
Aged 40 to 64 years in 1959
Started work 10-20 years earlier
than 1959
Industrial hygiene data RR = 1.45 (40-44 year age group) Years of exposure used as surrogate
correlated with job titles to RR = 1.33 (45-49 year age group) for dose
dichotomize the jobs as Both SS
"exposed" or "not exposed" Not possible to separate the effect of
After exclusion of workers time since first exposure and duration
exposed to asbestos of exposure
RR = 1.57 (40-44 year age group)
RR = 1.34 (45-49 year age group)
Both SS
Dose response indicated by
increasing lung cancer risk with
increasing cumulative exposure
D
I
O
2
O
H
O
Gustavsson 695 male workers from 5 bus
et al. (1990) garages in Stockholm, Sweden,
who had worked for 6 months
between 1945 and 1970
34 years followup( 1952-1986)
Nested case-control study
17 cases, six controls for each case
matched on age " 2 years
Four diesel exhaust indices SMRs of 122 and 115 (OF and
were created: GP), respectively, SNS
Oto 10, 10 to 20,20-30, and
>30 based on job tasks and Case-control study results
duration of work RR = 1.34 (10 to 20)
RR=1.81 (20 to 30)
RR = 2.43 (>30)
All SS with 0-10 as comparison
group
Exposure matrix based on job tasks
(not on actual measurements)
Small cohort, hence low power
Lack of smoking data
-------
^1
K)
O
s
o
z
3
n
»— t
H
rr)
/O
G
O
Table 7-1. Epidemiologic studies of the health effects of exposure to diesel exhaust: cohort mortality studies
(continued)
Authors Population studied Diesel exhaust exposure Results Limitations
Hansen Cohort of 57,249 unskilled Diesel exhaust exposure SS SMR = 160 for bronchus and No actual exposure data available
(1993) laborers, ages 15 to 74, in assumed based on diesel- lung for total population
Denmark (nationwide census file) powered trucks Lack of smoking data
November 9, 1970
Job changes may have occurred from
Followup through November 9, laborer to driver
1980
Short follow-up period
Abbreviations: RR = relative risk; SMR = standardized mortality ratio; SNS = statistically nonsignificant; SS = statistically significant;
OF = occupationally active; GP = general population.
-------
1 were recorded by interviewers and were coded separately for main lifetime employment, recent
2 employment, and other jobs held according to the 1970 Census Coding Scheme. Occupations or
^P industries were combined to form larger groups. Coding of occupational and industrial labels in
4 meaningful job categories was done by one of the authors. Of the 3,539 interviewed males and
5 3,937 interviewed females, 95% and 84%, respectively, listed some main employment. The
6 basic analysis consisted of an intercancer comparison and involved comparing the proportions of
7 specific main lifetime industries and occupations among patients with cancer at one site with
8 those of patients having cancer at other sites combined as a control group; this was done using a
9 series of Mantel-Haenszel stratified contingency table analyses to yield odds ratios and chi-
10 square values. Odds ratios were computed separately for males and females, controlling for age,
11 race, education, tobacco, alcohol, and geographic location.
12 A total of 432 and 128 lung cancers were present in males and females, respectively. For
13 males, an excess risk of lung cancer was observed for the following main industrial groups:
14 mines (odds ratio [OR]= 1.21), construction (OR = 1.24), transportation (OR =1.17), utility and
1 5 sanitary services (OR = 2.79, /K0.05), and professional (OR = 1.41). An excess of bladder
16 cancer was reported for the mining industry (OR = 1.61). For females, an excess of lung cancer
17 was detected for the transportation industry (OR = 1.96); finance and retail industry (OR = 1.73);
18 and the business, car repair, and miscellaneous service industry (OR = 2.29). None of these
^P excesses were statistically significant. All of these odds ratios were adjusted for age, race,
20 education, tobacco, alcohol, and geographic location. The transportation industry for males and
21 females also showed a nonsignificant excess risk for cancers of the liver and gall bladder ducts.
22 When the analysis was done for specific lifetime industries, the odds ratio for lung cancer in
23 males was 1.40 for railroad workers and 1.34 for truck drivers. Both of these excesses were
24 statistically nonsignificant.
25 The strengths of the TNCS interview data set are its large size, histological confirmation
26 of nearly 95% of diagnoses, availability of information on occupation, and details of
27 confounding variables obtained by personal interview and ability to control for them. Among its
28 weaknesses are a 47% nonresponse rate and the fact that the population surveyed came from
29 predominantly urban areas and did not represent many industries. Also, most of the associations
30 observed did not achieve statistical significance because they were based on small numbers of
31 patients who had both specific cancers and specific types of employment. The control group was
32 the combined "other cancers," which may have diluted the association because diesel exhaust is
33 also suspected of being associated with bladder cancer, and this category was included in the
34 control group when the comparison was made with lung cancer. The study presented several
tables, but the total population in each table was different and never added up to the initial
11/5/99 7-25 DRAFT-DO NOT CITE OR QUOTE
-------
1 number interviewed. The authors failed to explain these omissions. Furthermore, when multiple
2 comparisons are made, some significant associations arise by chance. This analysis suggests an
3 association with lung cancer for three industries with potential diesel exhaust exposure: trucking,
4 railroading, and mining.
5
6 7.2.2.2. Hall and Wynder (1984): A Case-Control Study of Diesel Exhaust Exposure and
7 Lung Cancer
8 Hall and Wynder (1984) conducted a case-control study of 502 male lung cancer cases
9 and 502 controls without tobacco-related diseases that examined an association between
10 occupational diesel exhaust exposure and lung cancer. Histologically confirmed primary lung
11 cancer patients who were 20 to 80 years old were ascertained from 18 participating hospitals in
12 six U.S. cities, 12 months prior to the interview. Eligible controls, patients at the same hospitals
13 without tobacco-related diseases, were matched to cases by age (±5 years), race, hospital, and
14 hospital room status. The number of male lung cancer cases interviewed totaled 502, which was
1 5 64% of those who met the study criteria for eligibility. Of the remaining 36%, 8% refused, 21%
1 6 were too ill or had died, and 7% were unreliable. Seventy-five percent of eligible controls
17 completed interviews. Of these interviewed controls, 49.9% were from the all-cancers category,
18 whereas 50.1 % were from the all-noncancers .category. All interviews were obtained in hospitals
19 to gather detailed information on smoking history, coffee consumption, artificial sweetener use,
20 residential history, and abbreviated medical history as well as standard demographic variables.
21 Occupational information was elicited by a question on the usual lifetime occupation and was
22 coded by the abbreviated list of the U.S. Bureau of Census Codes. The odds ratios were
23 calculated to evaluate the association between diesel exhaust exposure and risk of lung cancer
24 incidence. Summary odds ratios were computed by the Mantel-Haenszel method after adjusting
25 for potential confounding by age, smoking, and socioeconomic class. Two-sided, 95%
26 confidence intervals were computed by Woolf s method. Occupational exposure to diesel
27 exhaust was defined by two criteria. .First, occupational titles were coded "probably high
23 exposure" as defined by the industrial hygiene standards established for ilic various jobs. The
29 job titles included under tms category were warebouacmcii, bus and truck drivers, railroad
30 workers, and heavy equipment operators and repairmen. The second method used the National
31 Institute for Occupational Safely and Health (NIOSH) criteria to analyze occupations by diesel
32 exposure. In this method, the estimated proportion of exposed workers was computed for each
33 occupational category by using the NIOSH estimates of the exposed population as the numerator
34 and the estimates of individuals employed hi each occupational category from the 1970 census as
35 the denominator. Occupations estimated to have at least 20% of their employees exposed to
11/5/99 7-26 DRAFT-DO NOT CITE OR QUOTE
-------
diesel exhaust were defined as "high exposure," those with 10% to 19% of their employees
exposed were defined as "moderate exposure," and those with less than 10% of their employees
exposed were defined as "low exposure."
4 Cases and controls were compared with respect to exposure. The relative risk was 2.0
5 (95% CI = 1.2, 3.2) for those workers who were exposed to diesel exhaust versus those who were
6 not. The risk, however, decreased to a nonsignificant 1.4 when the data were adjusted for
7 smoking. Analysis by NIOSH criteria found a nonsignificant relative risk of 1.7 in the high-
8 exposure group. There were no significantly increased cancer risks by occupation either by the
9 first method or by the NIOSH method. To assess any possible synergism between diesel exhaust
10 exposure and smoking, the lung cancer risks were calculated for different smoking categories.
11 The relative risks were 1.46 among nonsmokers and ex-smokers, 0.82 among current smokers of
12 <20 cigarettes/day, and 1.3 among current smokers of 20+ cigarettes/day, indicating a lack of
13 synergistic effects.
14 The major strength of this study is the availability of a detailed smoking history for all the
1 5 study subjects. However, this is offset by lack of diesel exhaust exposure measurements, use of a
16 poor surrogate for exposure, and lack of consideration of latency period. Information was
17 collected on only one major lifetime occupation, and it is likely that those workers who had more
than one major job may not have reported the occupation with the heaviest diesel exhaust
exposures. Furthermore, occupational histories were obtained from self-reports and were not
20 validated with work records. This could have resulted in recall bias and misclassification of
21 exposure status.
22
23 7.2.23. Dumber andLarsson (1987): Occupation and Male Lung Cancer, a Case-Control
24 Study in Northern Sweden
25 A case-control study of lung cancer was conducted hi northern Sweden to determine the
26 occupational risk factors that could explain the large geographic variations of lung cancer
27 incidence in that country. The study region comprised the three northernmost counties of
28 Sweden, with a total male population of about 390,000. The rural municipalities with 15% to
29 20% of the total population have forestry and agriculture as dominating industries, and the urban
30 areas have a variety of industrial activities (mines, smelters, steel factories, paper mills, and
31 mechanical workshops). All male cases of lung cancer reported to the Swedish Cancer Registry
32 during the 6-year period between 1972 and 1977 who had died before the start of the study were
33 selected. Of 604 eligible cases, 5 did not have microscopic confirmation and in another 5 the
34 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
11/5/99 7-27 DRAFT-DO NOT CITE OR QUOTE
-------
1 dead control was drawn from the National Death Registry matched by sex, year of death, age,
2 and municipality. Deaths hi controls classified as lung cancer and suicides were excluded. A
3 living control matched to the case by sex, year of birth, and municipality was also drawn from
4 the National Population Registry. Postal questionnaires were sent to close relatives of cases and
5 dead controls, and to living controls themselves to collect data on occupation, employment, and
6 smoking habits. Replies were received from 589 cases (98%), 582 surrogates of dead controls
7 (96%), and 453 living controls (97%).
8 Occupational data were collected on occupations or employment held for at least 1 year
9 and included type of industry, company name, task, and duration of employment.
10 Supplementary telephone interviews were performed if occupational data were lacking for any
11 period between age 20 and time of diagnosis. Data analysis involved calculation of the odds
12 ratios by the exact method based on the hypergeometric distribution and the use of a linear
13 logistic regression model to adjust for the potential confounding effects of smoking. Separate
14 analyses were performed with dead and living controls, and on the whole there was good
15 agreement between the two control groups. A person who had been active for at least 1 year in a
16 specific occupation was hi the analysis assigned to that occupation.
17 Using dead controls, the odds ratios adjusted for smoking were 1.0 (95% CI = 0.7,1.5)
18 and 2.7 (95% CI = 1.0, 8.1) for professional drivers (s 1 year of employment) and underground
1 9 miners (si year of employment), respectively. For 20 or more years of employment in those
20 occupations, the odds ratios adjusted for smoking were 1.2 (95% CI = 0.6,2.2) and 9.8 (95% CI
21 =1.5,414). These were the only two occupations listed with potential diesel exhaust exposure.
22 An excess significant risk was detected for copper smelter workers, plumbers, and electricians, as
23 well as concrete and asphalt workers. Occupational asbestos exposure was also associated with
24 an elevated odds ratio of 2.6 (95% CI = 1.6, 3.6) for * 1 year of employment and 3.6 (95% CI =
25 1.9, 7.2) for ^20 years of employment. All the odds ratios were calculated by adjusting for age,
26 smoking, and municipality. After comparison with the live controls, the odds ratios were found
27 to be lower than those observed with dead controls. None of the odds ratios were statistically
28 significant in this comparison.
29 This study did not detect any excess risk of lung cancer for professional drivers, who,
30 among all the occupations listed, had the most potential for exposure to motor vehicle exhaust.
31 However, it is not known whether these drivers were exposed exclusively to gasoline exhaust,
32 diesel exhaust, or varying degrees of both. An excess risk was detected for underground miners,
33 but it is not known if this was due to diesel emissions from engines or from radon daughters in
34 poorly ventilated mines. Although a high response rate (98%) was obtained by the postal
11/5/99 7-28 DRAFT-DO NOT CITE OR QUOTE
-------
1 questionnaires, the use of surrogate respondents is known to lead to misclassification errors that
2 can bias the odds ratio to 1.
4 7.2.2.4. Lerchen et ai (1987): Lung Cancer and Occupation in New Mexico
5 This is a population-based case-control study conducted in New Mexico that examined
6 the association between occupation and occurrence of lung cancer in Hispanic and non-Hispanic
7 whites. Cases involved residents of New Mexico, 25 through 84 years of age and diagnosed
8 between January 1,1980, and December 31, 1982, with primary lung cancer, excluding
9 bronchioalveolar carcinoma. Cases were ascertained through the New Mexico Tumor Registry,
10 which is a member of the Surveillance Epidemiology and End Results (SEER) Program of the
11 National Cancer Institute. Controls were chosen by randomly selecting residential telephone
12 numbers and, for those over 65 years of age, from the Health Care Financing Administration's
13 • roster of Medicare participants. They were frequency-matched to cases for sex, ethnicity, and
14 10-year age category with a ratio of 1.5 controls per case. The 506 cases (333 males and 173
15 females) and 771 controls (499 males and 272 females) were interviewed, with a nonresponse
16 rate of 11% for cases. Next of kin provided interviews for 50% and 43% of male and female
17 cases, respectively. Among controls, only 2% of the interviews were provided by next of kin for
18 each sex. Data were collected by personal interviews conducted by bilingual interviewers in the
1^P participants' homes. A lifetime occupational history and a self-reported history of exposure to
20 specific agents were obtained for each job held for at least 6 months since age 12. Questions
21 were asked about the title of the position, duties performed, location and nature of industry, and
22 time at each job title. A detailed smoking history was also obtained. The variables on
23 occupational exposures were coded according to the Standard Industrial Classification scheme by
24 a single person and reviewed by another. To test the hypothesis about the high-risk jobs for lung
25 cancer, an a priori listing of suspected occupations and industries was created by a two-step
26 process involving a literature review for implicated industries and occupations by the principal
27 investigator. The appropriate Standard Industrial Classification and Standard Occupational
28 Codes associated with job titles were also determined by the principal investigator. For four
29 agents—asbestos, wood dust, diesel exhaust, and formaldehyde—the industries and occupations
30 determined to have exposure were identified, and linking of specific industries and occupations
31 was based on literature review and consultation with local industrial hygienists.
32 The relative odds were calculated for suspect occupations and industries, classifying
33 individuals as ever employed for at least 1 year in an industry or occupation and defining the
34 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
11 /5/99 7-29 DRAFT-DO NOT CITE OR QUOTE
-------
1 smoking status. For occupations with potential diesel exhaust exposure, the analysis showed no
2 excess risks for diesel engine mechanics and auto mechanics. Similarly, when analyzed by
3 exposure to specific agents, the odds ratio adjusted for age, smoking, and ethnicity was not
4 elevated for diesel exhaust fumes (OR = 0.6, 95% CI = 0.2, 1.6). Elevated odds ratios were
5 found for uranium miners (OR = 2.8, 95% CI = 1.0, 7.7), underground miners (OR = 2.4, 95% CI
6 = 1.2,4.4), construction painters (OR = 2.4, 95% CI = 0.6, 9.6), and welders (OR = 4.3, 95% CI
7 = 1.6, 11.0). No excess risks were detected for the following industries: shipbuilding, petroleum
8 refining, construction, printing, blast furnace, and steel mills. No excess risks were detected for
9 the following occupations: construction workers, painters, plumbers, paving equipment
10 operators, roofers, engineers and firemen, woodworkers, and shipyard workers. Females were
11 excluded from detailed analysis because none of the Hispanic female controls had been
12 employed in high-risk jobs; among the non-Hispanic white controls, employment in a high-risk
13 job was recorded for at least five controls for only two industries, construction and painting, for
14 which the odds ratios were not significantly elevated. Therefore, the analyses were presented for
1 5 males only.
1 6 Among the many strengths of this study are its population-based design, high
17 participation rate, detailed smoking history, and the separate analysis done for the two ethnic
18 groups, southwestern Hispanic and non-Hispanic white males. The major limitations pertain to
19 the occupational exposure date. Job titles obtained from occupational histories were used as
20 proxy for exposure status, but these were not validated. Further, for nearly half the cases, next of
21 kin provided occupational histories. The authors acknowledge the above sources of bias but state
22 without substantiation that these biases would not strongly affect their results. They also did not
23 use a job exposure matrix to link occupations to exposures and did not provide details on the
24 method they used to classify individuals as diesel exhaust exposed based on reported
25 occupations. The observed absence of an association for exposure to asbestos, a well-established
26 lung carcinogen, may be explained by the misclassification errors in exposure status or by
27 sample size constraints (not enough power). Likewise, the association for diesel exhaust
28 reported by only 7 cases and 17 controls also may have gone undetected because of low power.
29 In conclusion, there is insufficient evidence from this study to confirm or refute an association
30 between lung cancer and diesel exhaust exposure.
31
3 2 7.2.2.5. Garshick et aL (1987): A Case-Control Study of Lung Cancer and Diesel Exhaust
33 Exposure in Railroad Workers
34 An earlier pilot study of the mortality of railroad workers by the same investigators
35 (Schenker et al., 1984) found a moderately high risk of lung cancer among the workers who were
11/5/99 7-30 DRAFT-DO NOT CITE OR QUOTE
-------
exposed to diesel exhaust compared with 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.
4 The population base for this case-control study was approximately 650,000 active and retired
5 male U.S. railroad workers with 10 years or more of railroad service who were bom in 1900 or
6 later. The U.S. Railroad Retirement Board (RRB), which operates the retirement system, is
7 separate from the Social Security System, and to qualify for the retirement or survivor benefits
8 the workers had to acquire 10 years or more of service. Information on deaths that occurred
9 between March 1, 1981, and February 28,1982, was obtained from the RRB. For 75% of the
10 deceased population, death certificates were obtained from the RRB, and, for the remaining 25%,
11 they were obtained from the appropriate state departments of health. Cause of death was coded
12 according to the eighth revision of the ICD. The cases were selected from deaths with primary
13 lung cancer, which was the underlying cause of death in most cases. Each case was matched to
14 two deceased controls whose dates of birth were within 2.5 years of the date of birth of the case
1 5 and whose dates of death were within 31 days of the date of death noted in the case. Controls
16 were then selected randomly from workers who did not have cancer noted anywhere on their
17 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 the RRB from 1959 until death or retirement. The year 1959 was chosen as the
20 effective start of diesel exhaust exposure for this study, since by this time 95% of the
21 locomotives in the United States were diesel powered. Investigators acknowledge that because
22 the transition to diesel-powered engines took place hi the early 1950s, some workers had
23 additional exposure prior to 1959; however, if a worker had died or retired prior to 1959, he was
24 considered unexposed. Exposure to diesel exhaust was considered to be dichotomous for this
25 study, which was assigned based on an industrial hygiene evaluation of jobs and work areas.
26 Selected jobs with and without regular diesel exhaust exposure were identified by a review of job
27 title and duties. Personal exposure was assessed hi 39 job categories representative of workers
28 with and without diesel exhaust exposure. Those jobs for which no personal sampling was done
29 were considered exposed or unexposed on the basis of similarities in job activities and work
30 locations and by degree of contact with diesel equipment. Asbestos exposure was categorized on
31 the basis of jobs held hi 1959, or on the last job held if the subject retired before 1959. Asbestos
32 exposure hi railroads occurred primarily during the steam engine era and was related mostly to
33 the repair of locomotive steam boilers that were insulated with asbestos. Smoking history
34 information was obtained from the next of kin.
11/5/99 7-31 DRAFT-DO NOT CITE OR QUOTE
-------
1 Death certificates were obtained for approximately 87% of the 15,059 deaths reported by
2 the RRB, from which 1,374 cases of lung cancer were identified. Fifty-five cases of lung cancer
3 were excluded from the study for either incomplete data (20) or refusal by two States to use
4 information on death certificates to contact the next of kin. Successful matching to at least one
5 control with work histories was achieved for 335 (96%) cases s64 years of age at death and 921
6 (95%) cases ^65 years of age at death. In both age groups, 90% of the cases were matched with
7 two controls. There were 2,385 controls in the study, 98% were matched within ±31 days of the
8 date of death, whereas the remaining 2% were matched within 100 days. Deaths from diseases
9 of the circulatory system predominated among controls. Among the younger workers,
10 approximately 60% had exposure to diesel exhaust, whereas among older workers, only 47%
11 were exposed to diesel exhaust.
1 2 Analysis by a regression model, in which years of diesel exhaust exposure were the sum
1 3 total of the number of years in diesel-exposed jobs, used as a continuous exposure variable,
14 yielded an odds ratio of lung cancer of 1.39 (95% CI = 1.05,1.83) for over 20 years of diesel
1 5 exhaust exposure in the ^64 years of age group. After adjustment for asbestos exposure and
1 6 lifetime smoking (pack-years), the odds ratio was 1.41 (95% CI = 1.06, 1.88). Both crude odds
17 ratio and asbestos exposure as well as lifetime smoking adjusted odds ratio for the ^65 years of
18 age group were not significant. Increasing years of diesel exhaust exposure, categorized as ^20
19 diesel years and 5 to 19 diesel years, with 0 to 4 years as the referent group, showed significantly
20 increased risk hi the £ 64 years of age group after adjusting for asbestos exposure and pack-year
21 category of smoking. For individuals who had ^20 years of diesel exhaust exposure, the odds
22 ratio was 1.64 (95% CI = 1.18, 2.29), whereas among individuals who had 5 to 19 years of diesel
23 exhaust exposure, the odds ratio was 1.02 (95% CI = 0.72, 1.45). In the ^65 years of age group,
24 only 3% of the workers were exposed to diesel'exhaust for more than 20 years. Relative odds for
25 5 to 19 years and s20 years of diesel exposure were less than 1 (p>0.01) after adjusting for
26 smoking and asbestos exposure.
27 Alternate models to explain post-asbestos exposure were tested. These were variables for
28 regular and intermittent exposure groups and an estimate of years of exposure based on estimated
29 years worked prior to 1959. No diiierences in results were seen Tne interactions between uicsci
30 exhaust exposure and the three pack-year categories (<50, >50, and missing pack-years) were
31 explored. The cross-product terms were not significant. A model was also tested that excluded
32 recent diesel exhai^t exposure occurring within the 5 years before death and gave an odds ratio
33 of 1.43 (95% CI = 1.06, 1.94) adjusted for cigarette smoking and asbestos exposure, for workers
34 with 15 years of cumulative exposure. For workers with 5 to 14 years of cumulative exposure,
35 the relative odds were not significant.
11/5/99 7-32 DRAFT-DO NOT CITE OR QUOTE
-------
1 The many strengths of the study are consideration of confounding factors such as
2 asbestos exposure and smoking; classification of diesel exhaust exposures by job titles and
^P industrial hygiene sampling; exploration of interactions between smoking, asbestos exposure,
4 and diesel exhaust exposure; and good ascertainment (87%) of death certificates from the 15,059
5 deaths reported by the RRB.
6 The investigators also recognized and reported the following limitations: overestimation
7 of cigarette consumption by surrogate respondents, which may have exaggerated the contribution
8 of smoking to lung cancer risk, and use of the Interstate Commerce Commission (ICC) job
9 classification as a surrogate for exposure,'which may have led to misclassification of diesel
10 exhaust exposure jobs with low intensity and intermittent exposure, such as railroad police and
11 bus drivers, as unexposed. These two limitations would result in the underestimation of the lung
12 cancer risk. This source of error could have been avoided if diesel exhaust exposures were
13 categorized by a specific dose range associated with a job title that could have been classified as
14 heavy, medium, low, and zero exposure instead of a dichotomous variable. The use of death
15 certificates to identify cases and controls may have resulted hi misclassification. Controls may
16 have had undiagnosed primary lung cancer, and lung cancer cases might have been secondary
17 lesions misdiagnosed as primary lung cancer. However, the investigators quote a third National
18 Cancer Survey report in which the death certificates for lung cancer were coded appropriately in
1^P 95% of the cases. Last, as in all previous studies, there is a lack of data on the contribution of
20 unknown occupational or environmental exposures and passive smoking. In conclusion, this
21 study, compared with previous studies (on diesel exposure and lung cancer risk), provides the
22 most valid evidence that occupational diesel exhaust emission exposure increases the risk of lung
23 cancer.
24
25 7.2.2.6. Benhamou et aL (1988): Occupational Risk Factors of Lung Cancer in a French
26 Case-Control Study
27 This is a case-control study of 1,625 histologically confirmed cases of lung cancer and
28 3,091 matched controls, conducted in France between 1976 and 1980. This study was part of an
29 international study to investigate the role of smoking and lung cancer. Each case was matched
30 with one or two controls whose diseases were not related to tobacco use, sex, age at diagnosis
31 (±5 years), hospital of admission, or interviewer. Information was obtained from both cases and
32 controls on place of residence since birth, educational level, smoking, and drinking habits. A
33 complete lifetime occupational history was obtained by asking participants to give their
34 occupations from the most recent to the first. Women were excluded because most of them had
listed no occupation. Men who smoked cigars and pipes were excluded because there were very
11/5/99 7-33 DRAFT-DO NOT CITE OR QUOTE
-------
1 few in this category. Thus, the study was restricted to nonsmokers and cigarette smokers.
2 Cigarette smoking exposure was defined by age at the first cigarette (nonsmokers, ^20 years, or
3 >20 years), daily consumption of cigarettes (nonsmokers, <20 cigarettes a day, and £20
4 cigarettes a day), and duration of cigarette smoking (nonsmokers, <35 years, and ^35 years).
5 The data on occupations were coded by a panel of experts according to their own chemical or
6 physical exposure determinations. Occupations were recorded blindly using the International
7 Standard Classification of Occupations. Data on 1,260 cases and 2,084 controls were available
8 for analysis. The remaining 365 cases and 1,007 controls were excluded because they did not
9 satisfy the required smoking status criteria.
10 A matched logistic regression analysis was performed to estimate the effect of each
11 occupational exposure after adjusting for cigarette status. Matched relative risk ratios were
12 calculated for each occupation with the baseline category, which consisted of patients who had
13' never been engaged in that particular occupation. The matched relative risk ratios adjusted for
14 cigarette smoking for the major groups of occupations showed that the risks were significantly
1 5 higher for production and related workers, transport equipment operators, and laborers (RR =
16 1.24, 95% CI = 1.04, 1.47). On further analysis of this group, for occupations with potential
17 diesel emission exposure, significant excess risks were found for motor vehicle drivers (RR =
18 1.42, 95% CI = 1.07, 1.89) and transport equipment operators (RR = 1.35, 95% CI = 1.05, 1.75).
19 No interaction with smoking status was found in any of the occupations. The only other
20 significant excess was observed for miners and quarrymen (RR = 2.14, 95% CI = 1.07, 4.31).
21 None of the significant associations showed a dose-response relationship with duration of
22 exposure.
23 This study was designed primarily to investigate the relationship between smoking (not
24 occupations or environmental exposures) and lung cancer. Although an attempt was made to
25 obtain complete occupational histories, the authors did not clarify whether, in the logistic
26 regression analysis, they used the subjects' first occupation, predominant occupation, last
27 occupation, or ever worked in that occupation as the risk factor of interest. The most important
23 limitation of this study is that the occupations were not coded into exposures for different
29 chemical and physical agents, thus precluding the calculation of relative risks for riiesei
30 exposure. Using occupations as surrogate measures of diesel exposure, an excess significant risk
31 was obtained for motor vehicle drivers and transport equipment operators, but not for motor
32 mechanics. However, it is net knov/a if subjects in these occupations worked with diesel engines
33 or nondiesel engines.
34
11/5/99 7-34 DRAFT-DO NOT CITE OR QUOTE
-------
7.2.2.7. Hayes et al (1989): Lung Cancer in Motor Exhaust-Related 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
4 exhaust-related (MER) occupations and lung cancer risk adjusted for confounding by smoking
5 and other risk factors. The three studies were carried out by the National Cancer Institute in
6 Florida (1976 to 1979), New Jersey (1980 to 1981), and Louisiana (1979 to 1983). These three
7 studies were selected because the combined group would provide a sufficient sample to detect a
8 risk of lung cancer in excess of 50% among workers in MER occupations. The analyses were
9 restricted to males who had given occupational history. The Florida study was hospital based,
10 with cases ascertained through death certificates. Controls were randomly selected from hospital
11 records and death certificates, excluding psychiatric diseases, matched by age and county. The
12 New Jersey study was population based, with cases ascertained through hospital records, cancer
13 registry, and death certificates. Controls were selected from among the pool of New Jersey
14 licensed drivers and death certificates. The Louisiana study was hospital based (it is not
15 specified how the cases were ascertained), and controls were randomly selected from hospital
16 patients, excluding those with lung diseases and tobacco-related cancers.
17 A total of 2,291 cases of male lung cancers and 2,570 controls were eligible, and the data
on occupations were collected by next-of-kin interviews for all jobs held for 6 months or more,
including the industry, occupation, and number of years employed. The proportion of next-of-
20 kin interviews varied by site from 50% in Louisiana to 85% in Florida. The coding schemes
21 were reviewed to identify MER occupations, which included truck drivers and heavy equipment
22 operators (cranes, bulldozers, and graders); bus drivers, taxi drivers, chauffeurs, and other motor
23 vehicle drivers; and automobile and truck mechanics. Truck drivers were classified as routemen
24 and delivery men and other truck drivers. All jobs were also classified with respect to potential
25 exposure to known and suspected lung carcinogens. Odds ratios were calculated by the
26 maximum likelihood method adjusting for age by birth year, usual amount smoked, and study
27 area. Logistic regression models were used to examine the interrelationship of multiple
28 variables.
29 A statistically significant excess risk was detected for employment of 10 years or more
30 for all MER occupations (except truck drivers) adjusted for birth cohort, usual daily cigarette use,
31 and study area. The odds ratio for lung cancer using data gathered by direct interviews was 1.4
32 (95% CI = 1.1,2.0), allowing for multiple MER employment, and 2.0 (95%. CI = 1.3, 3.0),
33 excluding individuals with multiple MER employment. Odds ratios for all MER employment,
34 except truck drivers who were employed for less than 10 years, were 1.3 (95% CI = 1.0, 1.7) and
1.3 (95% CI = 0.9, 1.8) including and excluding multiple MER employment, respectively. Odds
11/5/99 7-35 DRAFT-DO NOT CITE OR QUOTE
-------
1 ratios were then derived for specific MER occupations and, to avoid the confounding effects of
2 multiple MER job classifications, analyses were also done excluding subjects with multiple
3 MER job exposures. Truck drivers employed for more than 10 years had an odds ratio of 1.5
4 (95% CI = 1.1, 1.9). A similar figure was obtained excluding subjects with multiple MER
5 employment. An excess risk was not detected for truck drivers employed less than 10 years.
6 The only other job category that showed a statistically significant excess for lung cancer was the
7 one that included taxi drivers and chauffeurs who worked multiple MER jobs for less than 10
8 years (OR = 2.5, 95% CI = 1.4,4.8). For the same category, the risk for individuals working in
9 that job for more than 10 years was 1.2 (95% CI = 0.5,2.6). A statistical significant positive
10 trend (p<0.05) with increasing employment of <2 years, 2 to 9 years, 10 to 19 years, and 20+
11 years was observed for truck drivers but not for other MER occupations. A statistically
1 2 nonsignificant excess risk was also observed for heavy equipment operators, bus drivers, taxi
1 3 drivers and chauffeurs, and mechanics employed for 10 years or more. All of the above-
14 mentioned odds ratios were derived, adjusted for birth cohort, usual daily cigarette use, and State
1 5 of residence. Exposure to other occupational suspect lung carcinogens did not account for the
1 6 excess risks detected.
17 Results of this large study provide evidence that workers in MER jobs are at an excess
18 risk of lung cancer that is not explained by their smoking habits or exposures to other lung
19 cancers. Because no information on type of engine had been collected, it was not possible to
20 determine if the excess risk was due to exposure to diesel exhaust or gasoline exhaust or the
21 mixture of the two. Among the study's limitations are possible bias due to misclassification of
22 jobs reported by the large proportion of next-of-kin interviews and the problems in classifying
23 individuals into uniform occupational groups based on the pooled data in the three studies that
24 used different occupational classification schemes.
25
26 7.2.2.8. Steenland et aL (1990): A Case-Control Study of Lung Cancer and Truck Driving in
27 the Teamsters Union
28 Steeniand et al. conducted a case-control study of lung cancer deaths in the Teamsters
29 Union to determine tne risk of lung cancer among dinerem occupations. Deaiii ctii.incai.ca wcie
30 obtained from the Teamsters Union files in the central States for 10,485 (98%) male decedents
31 who had filed claims for pension benefits and who had disci in 1982 and 1983. Individuals were
32 required to have 20 years' tenure m the union to be eligible to claim benefits. Cases comprised
33 all deaths (n = 1,288) from lung cancer, coded as ICD 162 or 163 for underlying or contributory
34 cause on the death certificate. The 1,452 controls comprised every sixth death from the entire
35 file, excluding deaths from lung cancer, bladder cancer, and motor vehicle accidents. Detailed
11/5/99 7-36 DRAFT-DO NOT CITE OR QUOTE
-------
1 information on work history and potential confounders such as smoking, diet, and asbestos
2 exposure was obtained by questionnaire. Seventy-six percent of the interviews were provided by
^P spouses and the remainder by some other next of kin. The response rate was 82% for cases and
4 80% for controls. Using these interview data and the 1980 census occupation and industry
5 codes, subjects were classified either as nonexposed or as having held other jobs with potential
6 diesel exhaust exposure. Data on job categories were missing for 12% of the study subjects. A
7 second work history file was also created based on the Teamsters Union pension application that
8 lists occupation, employer, and dates of employment. A three-digit U.S. census code for
9 occupation and industry was assigned to each job for each individual. This Teamsters Union
10 work history file did not have information on whether men drove diesel or gasoline trucks, and
11 the four principal occupations were long-haul drivers, short-haul or city drivers, truck mechanics,
12 and dockworkers. Subjects were assigned the job category in which they had worked the
13 longest.
14 The case-control analysis was done using unconditional logistic regression. Separate
15 analyses were conducted for work histories from the Teamsters Union pension file and from
16 next-of-kin interviews. Covariate data were obtained from next-of-kin interviews. Analyses
17 were also performed for two time periods: employment after 1959 and employment after 1964.
18 These two cut-off years reflect years of presumed dieselization; 1960 for most trucking
I^P companies and 1965 for independent driver and nontrucking firms. Data for analysis could be
20 obtained for 994 cases and 1,085 controls using Teamsters Union work history and for 872 cases
21 and 957 controls using next-of-kin work history. When exposure was considered as a
22 dichotomous variable, for both Teamsters Union and next-of-kin work history, no single job
23 category had an elevated risk. From the next-of-kin data, diesel truck drivers had an odds ratio of
24 1.42 (95% CI = 0.74, 2.47) and diesel truck mechanics had an odds ratio of 1.35 (95% CI = 0.74,
25 2.47). Odds ratios by duration of employment as a categorical variable were then estimated. For
26 the Teamsters Union work history data and when only employment after 1959 was considered,
27 both long-haul (p<0.04) and short-haul drivers (not significant) showed an increase in risk with
28 increased years of exposure. The length of employment categories for which the trends were
29 analyzed were 1 to 11 years, 12 to 17 years, and 18 years or more. Using 1964 as the cutoff date,
30 long-haul drivers continued to show a significant positive trend (p=0.04), with an odds ratio of
31 1.64 (95% CI = 1.05, 2.57) for those who worked for 13+ years, the highest category. Short-haul
32 drivers, however, did not show a positive trend when 1964 was used as the cutoff date. Similar
33 trend analysis was done for most next-of-kin data, A marginal increase in risk with increasing
34 duration of employment as a truck driver (p=0.12) was observed. For truck drivers who
primarily drove diesel trucks for 35 years or longer, the odds ratio for lung cancer was 1.89 (95%
11/5/99 7-37 DRAFT-DO NOT CITE OR QUOTE
-------
1 CI = 1.04, 3.42). The odds ratio was 1.34 (95% CI = 0.81, 2.22) for gasoline truck drivers and
2 1.09 (95% CI = 0.44, 2.66) for truck mechanics. No significant interactions between age and
3 diesel exhaust exposure or smoking and diesel exhaust exposure were observed. All the odds
4 ratios were adjusted for age, smoking, and asbestos in addition to various exposure categories.
5 The authors acknowledge several limitations of this study, which include possible
6 misclassifications of exposure and smoking habits, as information was provided by next of kin;
7 lack of sufficient latency to observe lung cancer excess; and a small nonexposed group (n = 120).
8 Also, concordance between Teamsters Union and next-of-kin job categories could not be easily
9 evaluated because job categories were defined differently in each data set. No data were
10 available on levels of diesel exposure for the different job categories. Given these limitations,
11 the positive findings of this study are probably underestimated.
12
13 7.2.2.9. Steenland et al (1998): Diesel Exhaust and Lung Cancer in the Trucking Industry:
14 Exposure-Response Analyses and Risk Assessment
1 5 Steenland et al. (1998) conducted an exposure-response analysis by supplementing the
16 data from their earlier case-control study of lung cancer and truck drivers in the Teamsters Union
17 (Steenland et al., 1990) with exposure estimates based on a 1990 industrial hygiene survey of
18 elemental carbon exposures a surrogate for diesel exhaust hi the trucking industry.
19 Study subjects were long-term Teamsters enrolled in the pension system who died during
20 the period 1982-1983. Using death certificate information, the researchers identified 994 cases
21 of lung cancer for the study period, and 1,085 non lung cancer deaths served as controls.
22 Subjects were divided into job categories based on the job each held the longest. Most had held
23 only one type of job. The job categories were short-haul driver, long-haul driver, mechanic,
24 dockworker, other jobs with potential diesel exposure, and jobs outside the trucking industry
25 without occupational diesel exposure. Smoking histories were obtained from next of kin. Odds
26 ratios were calculated for work in an exposed job category at any time and after 1959 (an
27 estimated date when the majority of heavy duty trucks had converted to diesel) compared with
28 work in nonexposed jobs. Odds ratios were adjusted for age, smoking; and potential asbestos
29 exposure. Trends in effect estimates for duration of worse in an exposed job were also calculated.
30 An industrial hygiene survey by Zaebst et al. (1991) of elemental carbon exposures in the
31 trucking industry provided exposure estimates for each job category in 1990. The elemental
32 carbon measurements were generally consistent with the epidemiologic results, in that mechanics
33 are found to have the highest exposures and relative risk, followed by long-haul and then
34 short-haul drivers, although dockworkers have the highest exposures and the lowest relative
35 risks.
11 /5/99 7-38 DRAFT-DO NOT CITE OR QUOTH
-------
Past exposures were estimated assuming that they were a function of (1) the number of
heavy-duty trucks on the road, (2) the particulate emissions (grams/mile) of diesel engines over
time, and (3) leaks from truck exhaust systems for long-haul drivers. Estimates of past exposure
4 to elemental carbon, as a marker for diesel exhaust exposure, for subjects in the case-control
5 study were made by assuming that average 1990 levels for a job category could be assigned to all
6 subjects in that category, and that levels prior to 1990 were directly proportional to vehicle miles
7 traveled by heavy-duty trucks and the estimated emission levels of diesel engines. A 1975
8 exposure level of elemental carbon in term of micrograms per cubic meter was estimated by the
9 following equation: 1975 level = 1990 level* (vehicle miles 1975/vehicle miles 1990) (emissions
10 1975/emissions 1990). Once estimates of exposure for each year of work history were derived
11 for each subject, analyses were conducted by cumulative level of estimated carbon exposure.
12 Estimates were made for long-haul drivers (n = 1,237), short-haul drivers (n = 297),
13 dockworkers (n = 164), mechanics (n = 88), and those outside the trucking industry (n = 150).
14 Logistic regression was used to estimate odds ratios adjusted for five categories of age, race,
15 smoking (never, former-quitting before 1963, former-quitting in 1963 or later, current-with <1
16 pack per day, and current-with 1 or more packs per day), diet, and reported asbestos exposure. A
17 variety of models for cumulative exposure were considered, including a log-linear model with
cumulative exposure, a model adding a quadratic term for cumulative exposure, a log transform
of cumulative exposure, dummy variables for quartile of cumulative exposure, and smoothing
20 splines of cumulative exposure. The estimates of rate ratios from logistic regression for specific
21 levels of exposure to elemental carbon were then used to derive excess risk estimates for lung
22 cancer after lifetime exposure to elemental carbon.
23 The log of cumulative exposure was found to be the best fitting model and was a
24 significant predictor (p = 0.01). Odds ratios for quartile of cumulative exposure show a pattern
25 of significantly increasing trends in risk with increasing exposure, ranging between 1.08 and
26 1.72, depending on the exposure level and lag structure used. The lifetime excess risk of lung
27 cancer death (through age 75) for a male truck driver was estimated to be in the range of 1.4-
28 2.3% (95% confidence limits ranged from 0.3% to 4.6%) above the background risk, depending
29 on the emissions scenarios assumed. The authors conclude that the data suggest a positive and
30 significant increase hi lung cancer risk with increasing estimated cumulative exposure to diesel
31 exhaust among workers in the trucking industry. They assert that these estimates suggest that the
32 lifetime excess risk for lung cancer is 10 times higher than the OSHA standards, but caution that
33 the results should be viewed as exploratory.
34 The authors acknowledge that the increasing trend in risk with increasing estimates of
cumulative exposure is partly due to the fact that a component of cumulative dose is simple
11/5/99 7-39 DRAFT-DO NOT CITE OR QUOTE
-------
1 duration of exposure, and that analyses by simple duration also exhibit a positive trend with
2 duration. This analysis essentially weights the duration by contrived estimates of exposure
3 intensity, and they acknowledge that this weighting depends on very broad assumptions.
4 This is not an analysis of new data that provides independent estimates of relative risk for
5 diesel exhaust and lung cancer incidence. Instead, it is an attempt to convert the data from
6 Steenland's earlier study of lung cancer for the purpose of estimating a different risk metric,
7 "lifetime excess risk of lung cancer," by augmenting these data with limited industrial hygiene
8 data and rationalizations about plausible models for cumulative exposure.
9 The Health Effects Institute (HEI, 1999) and others have raised some questions about the
10 exposure estimations and control for confounding variables. EPA and NIOSH will address these
11 concerns in the year 2000. It should be noted that these concerns are about the use of these data
1 2 for quantitative risk assessment. As far as qualitative risk assessment is concerned, this study is
13 still considered to be positive and strong.
14
1 5 7.2.2.10. Boffetta et aL (1990): Case-Control Study on Occupational Exposure to Diesel
16 Exhaust and Lung Cancer Risk
17 This is an ongoing (since 1969) case-control study of tobacco-related diseases in 18
18 hospitals (six U.S. cities). Cases comprise 2,584 males with histologically confirmed primary
19 lung cancers. Sixty-nine cases were matched to 1 control, whereas 2,515 were matched to 2
20 controls. Controls were individuals who were diagnosed with non-tobacco-related diseases. The
21 matching was done for sex, age (±2 years), hospital, and year of interview. The interviews were
22 conducted at the hospitals at the time of diagnosis; In 1985, the occupational section of the
23 questionnaire was modified to include the usual occupation and up to five other jobs as well as
24 duration (in years) worked in those jobs. After 1985, information was also obtained on exposure
25 to 45 groups of chemicals, including diesel exhaust at the workplace or during hobby activities.
26 A priori aggregation of occupations was categorized into low probability of diesel exhaust
27 exposure (reference group), possible exposure (19 occupations), and probable exposure (13
28 occupations). Analysis was conducted based on "usual occupation" on all study subjects, and
23 any occupation with sufficient cases was eligible for lurther analysis. In addition, cases enrolled
30 after 1985 for which there were self-reported diesel exhaust exposure and detailed work histories
31 were also analyzed separately.
32 Both matched and unmatched analyses were done by calculating the adjusted (for
33 smoking and education) relative odds using the Mantel-Haenzael method and calculating the test-
34 based 95% confidence interval using the Miettinen method. Unconditional logistic regression
11/5/99 7-40 DRAFT-DO NOT CITE OR QUOTE
-------
was used to adjust for potential confounders (the PROC LOGIST of SAS). Linear trends for risk
were also tested according to Mantel.
Adjusted relative odds for possible and probable exposure groups as well as the truck
4 drivers were slightly below unity, none being statistically significant for the entire study
5 population. Although slight excesses were observed for the self-reported diesel exhaust exposure
6 group and the subset of post-198 5 enrollees for highest duration of exposure (for self-reported
7 exposure, occupations with probable exposure and for truck drivers), none was statistically
8 significant. Trend tests for the risk of lung cancer among self-reported diesel exhaust exposure,
9 probable exposure, and truck drivers with increasing exposure (duration of exposure used as
10 surrogate for increasing dose) were nonsignificant too. Statistically significant lung cancer
11 excesses were observed for cigarette smoking only.
12 The major strength of this study is availability of detailed smoking history. Even though
13 detailed information was obtained for the usual and five other occupations (1985), no effort was
14 made to estimate or verify the actual exposure to diesel exhaust; instead, duration of employment
15 was used as a surrogate for dose. The numbers of cases and controls were large; however, the
16 number of individuals exposed to diesel exhaust was relatively few, thus reducing the power of
17 the study. This study did not attempt latency analysis either. Given these limitations, the
findings of this study are unable to provide either positive or negative evidence for a causal
association between diesel exhaust and occurrence of lung cancer.
20
21 7.2.2.11. Etnmelinetal. (1993): Diesel Exhaust Exposure and Smoking: A Case-Referent
22 Study of Lung Cancer Among Swedish Dock Workers
23 This is a case-control study of lung cancer drawn from the cohort defined as all-male
24 workers who had been employed as dockworkers for at least 6 months between 1950 and 1974.
25 In the population of 6,573 from 20 ports, there were 90 lung cancer deaths (cases), identified
26 through Swedish death and cancer registers, during the period 1960 to 1982. Of these 90 deaths,
27 the 54 who were workers at the 15 ports for which exposure surrogate information was available
28 were chosen for the case-control study. Four controls, matched on port and age, were chosen for
29 each case from the remaining cohort who had survived to the time of diagnosis of the case. Both
30 live and deceased controls were included. The final analyses were done on 50 cases and 154
31 controls who had complete information on employment dates and smoking data. The smoking
32 strata were created by classifying ex-smokers as nonsmokers if they had not smoked for at least 5
33 years prior to the date of diagnosis of the case; otherwise they were classified as smokers.
34 Relative odds and regression coefficients were calculated using conditional logistic
regression models. Comparisons were made both with and without smoking included as a
11/5/99 7-41 DRAFT-DO NOT CITE OR QUOTE
-------
1 variable, and the possible interaction between smoking and diesel exhaust was tested. Both the
2 weighted linear regressions of the adjusted relative odds and the regression coefficients were
3 used to test mortality trends with all three exposure variables.
4 Exposure to diesel exhaust was assessed indirectly by initially measuring (1) exposure
5 intensity based on exhaust emission, (2) characteristics of the environment in terms of
6 ventilation, and (3) measures of proportion of time in higher exposed jobs. For exhaust
7 emissions, annual diesel fuel consumption at a port was used as the surrogate. For ventilation,
8 the annual proportion of ships with closed or semiclosed holds was used as the surrogate. The
9 proportion of time spent below decks was used as the surrogate for more exposed jobs. Although
10 data were collected for all three measures, only the annual fuel consumption was used for
11 analysis. Because every man was likely to rotate through the various jobs, the authors thought
12 using annual consumption of diesel fuel was the appropriate measure of exposure.
13 Consequently, in a second analysis, the annual fuel consumption was divided by the number of
14 employees in the same port that year to come up with the fuel-per-person measure, which was
15 further used to create a second measure, "exposed tune." The "annual fuel" and exposed-time
16 data were entered in a calendar time-exposure matrix for each port, from which individual
17 exposure measures were created. A third measure, "machine tune" (years of employment from
18 first exposure), was also used to compare the results with other studies. All exposure measures
19 ' were accumulated from the first year of employment or first year of diesel machine use,
20 whichever came later. The last year of exposure was fixed at 1979. All exposures up to 2 years
21 before the date of lung cancer diagnosis were omitted from both cases and matched controls. A
22 priori classification into three categories of low, medium, and high exposure was done for all
23 three exposure variables: machine time, fuel, and exposed time.
24 Conditional logistic regression models, adjusting for smoking status and using low
25 exposures and/or nonsmokers as a comparison group, yielded positive trends for all exposure
26. measures, but no trend test results were reported, and only the relative odds for the exposed-time
27 exposure measure in the high-exposure group (OR = 6.8, 90% CI = 1.3 to 54.9) was reported as
28 statistically significant. For smokers, adjusting for diesel exhaust exposure level, the relative
23 odds were statistically significant and about equal for all three exposure variables: machine time,
30 OR = 5.7 (90% CI = 2.4 to 13.3); fuel, OR = 5.5 (90% CI = 2.4 to 12.7); and exposed tune, OR =
31 6.2 (90% CI = 2.6 to 14.6). Interaction between diesel exhaust and smoking was tested by
32 conditional logistic regression in the exposed-time variable. Although there were positive trends
33 for both smokers and nonsmokers, the trend for smokers was much steeper: low, OR = 3.7 (90%
34 CI = 0.9 to 14.6); medium, OR = 10.7 (90% CI = 1.5 to 78.4); and high, OR = 28.9 (90% CI =
35 3.5 to 240) indicating more than additive interaction between these two variables.
11 /5/99 7-42 DRAFT-DO NOT CITE OR QUOTE
-------
1 In the weighted linear regression model with the exposed-time variable, the results were
2 similar to those using the logistic regression model. The authors also explored the smoking
^P variable further in various analyses, some of which suggested a strong interaction between diesel
4 exhaust and smoking. However, with just six nonsmokers and no further categorization of
5 smoking amount or duration, these results are of limited value.
6 The diesel exhaust exposure matrices created using three different variables are intricate.
7 Analyses by any of these variables essentially yield the same positive results and positive trends,
8 providing consistent support for a real effect of diesel exhaust exposure, at least in smokers.
9 However, methodological limitations to this study prevent a more definitive conclusion. The
10 numbers of cases and controls are small. There are very few nonsmokers; thus testing the effects
11 of diesel exhaust exposure in them is futile. Lack of information on asbestos exposure, to which
12 dockworkers are usually exposed, may also confound the results. Also, no latency analyses are
13 presented. Overall, despite these limitations, this study supports the earlier findings of excess
14 lung cancer mortality among individuals exposed to diesel exhaust.
1 5 Table 7-2 summarizes the above lung cancer case-control studies.
16
17 7.2.3. Case-Control Study of Prostate Cancer
18 7.2.3.1. Aronsen et aL (1996): Occupational Risk Factors for Prostate Cancer: Results from
i^P a Case-Control Study in Montreal, Quebec, Canada
20 A population-based case-control study was undertaken in 1979 to explore possible
21 associations between many types of cancer and hundreds of occupational exposures. The current
22 report provides a more refined analysis focusing only on prostate cancer and those exposures that
23 showed associations with this site in the original analysis.
24 A total of 557 cases of incident, histologically confirmed prostate cancer were identified
25 among males aged 35 to 70 years resident in the Montreal area. The timeframe for eligibility of
26 incident cancers was not provided. Of these, 449 (81%) subjects were interviewed. Two sets of
27 controls were used. Out of 740 population controls, 533 (72%) persons identified by random
28 digit dialing or electoral lists provided interview data. Additionally, 1,550 controls were selected
29 from the non-prostate cancer cases identified hi the original study. Both control groups were
30 pooled for the analysis after determining that the estimates did not depend on the control group.
31 The exposure data were obtained by interview questionnaire, with a structured section
32 requesting information on potential confounders and a semistructured probing section designed
33 to obtain a detailed description of each job the subject had in his working lifetime. A team of
34 chemists and hygienists translated each job into a list of potential exposures by means of a
checklist of 294 substances. The current analysis focused on 17 occupations, 11 industries, and
11/5/99 7-43 DRAFT-DO NOT CITE OR QUOTE
-------
Table 7-2. Epidemiologic studies of the health effects of exposure to diesel exhaust; case-control studies of lung cancer
Authors Population studied Diesel exhaust exposure Results Limitations
Williams 7,518 (3,539 maies and
et al. (1977) 3,979 females) inc ident invasive
cancers from the Third National
Cancer Survey
Lung cancer cases
32 in males
28 in females
Combined other oncer sites
were used as cor.t 'ols
Main lifetime, recent,
and other employment
information obtained at
the time of survey
1970 Census Coding
Scheme for Employment
was used to code the
occupations by one of
the authors
SNS elevated relative odds were
observed among occupations of
trucking, railroading, and mining
Exposure estimation based on self-
report that was not validated
47% nonresponse
Control group consisted of other
cancers, probably diluting the risk
estimation
Small numbers in cause-specific
cancers and individual occupations
Hall and 502 historically confirmed
Wynder lung cancers
(1984) Cases diagnosed ! 2 mo prior to
interviews
5G2 matched host ital controls
without tobacco-related diseases,
matched for age .sex, rice, and
geographical area
Population from 18 hospitals in
controls
U
O
2
O
H
n
Based on previous
Industrial Hygiene
Standards for a
particular occupation,
usual lifetime occupation
coded as "probably high
exposure" and "no
exposure"
NIOSH standards used
to classify exposures:
High
Moderate
Low
SNS excess risk after adjustment
for smoking for lung cancer:
RR= 1.4 (1st criteria)
and
RR= 1.7 (NIOSH criteria)
Complete lifetime employment
history not available
Self-reported occupation history not
validated
No analysis by dose, latency, or
duration of exposure
No information on nonoccupational
diesel exposure
-------
Table 7-2. Epidemiologic studies of the health effects of exposure to diesel exhaust: case-control studies of lung cancer
Authors Population studied Diesel exhaust exposure Results Limitations
Damber and 589 lung cancer cases who had
Larsson died prior to 1979 reported to
(1987) Swedish registry between 1972
and 1977
582 matched dead controls (sex,
age, year of death, municipality)
drawn from National Registry
of Cause of Death
453 matched living controls
(sex, year of birth, municipality)
drawn from National
Population Registry
Occupations held for at
least 1 year or more
Using a 5-digit code the
occupations were
classified according to
Nordic Classification of
Occupations
SS OR = 2.7(^1 year
of employment)
SS OR = 9.8 (*20 years of
employment)
Uncertain diesel exhaust exposure
No validation of exposure done
Underground miners data not adjusted
for other confounders such as radon,
Adjustment for smoking was done etc.
SNS OR = 1.2 for professional
drivers (*20 years of employment)
with dead controls
SNS OR = 1.1 z 20 years of
employment) with living controls
Lerchen 506 lung cancer cases from
et al. (1987) New Mexico tumor registry
(333 males and 173 females)
Aged 25-84 years
Diagnosed between January 1,
1980, and December 31, 1982
771 (499 males and 272 females)
frequency matched with cases,
selected from telephone directory
Lifetime occupational
history and self-reported
exposure history were
obtained
Coded according to
Standard Industrial
Classification Scheme
No excess of relative odds was
observed for diesel exhaust
exposure
Exposure based on occupational
history and self-report, which was not
validated
50% occupational history provided by
next of kin
Absence of lung cuicer association
with asbestos suggests
misclassification of exposure
Garshick 1,319 lung cancer cases who died
et al. (1987) between March 1, 1981,
and February 28, 1982
Personal exposure
assessed for 39 job
categories
2,385 matched controls (two each, This was corrected with
age and date of death)
Both cases and controls drawn
from railroad worker cohort
who had worked for 10 or
more years
job titles to dichotomize
the exposure into:
Exposed
Not exposed
SS OR = 1.41 (s64 year age group) Probable misclassification of diesel
exhaust exposure jobs
SS OR = 1.64 (s64 year age group)
for ^20 years diesel exhaust Years of exposure used as surrogate
exposure group when compared to for dose
0- to 4-year exposure group
13% of death certificates not
All ORs adjusted for lifetime ascertained
smoking and asbestos exposure
Overestimation of smoking history
-------
O^
Table 7-2. Epidemiologic studies of the health effects of exposure to diesel exhaust; case-control studies of lung cancer
Authors Population studied Diesel exhaust exposure Results • Limitations
Benhamou 1,260 histologically confirmed
et al. (1988) lung cancer cases
2,084 non-tobacco-related
disease matched controls
(sex, age at diagnosis,
hospital admission, and
interviewer)
Occurring between 1976 and
1980 in France
Based on exposures Significant excess risks were found Exposure based on occupational
determined by panel of
experts
The occupations were
recorded blindly using
International Standard
Classification of
Occupations as chemical
or physical exposures
in motor vehicle drivers
(RR= 1.42) and
transport equipment operators
(RR = 1.35) (smoking adjusted)
histories not validated
Exposures classified as chemical and
physical exposure, not specific to
diesel exhaust
Hayes et al. Pooled data from three different Occupational information SS, OR = 1.5 for truck drivers (> 10 Exposure data based on job
(1989) studies consisting of 2,291 male from next of kin for all years of employment)
lung cancer cases
2,570 controls
jobs held
Jobs classified with
respect to potential
exposure to known and
suspected pulmonary
carcinogens
SS positive trend with increasing
employment as truck driver
description given by next of kin,
which was not validated
Could have been mixed exposure to
both diesel and gasoline exhausts
Job description could have led to
misclassification
-------
Ul
Table 7-2. Epidemiologic studies of the health effects of exposure to diesel exhaust: case-control studies of lung cancer
Authors
Population studied
Diesel exhaust exposure
Results
Limitations
Steenland 1,058 male lung cancer deaths
et al. (1990) between 1982 and 1983
Longest job held: diesel
truck driver, gasoline
truck driver, both types
1,160 every sixth death from entire of trucks, truck
mortality file sorted by social mechanic, and
security number (excluding lung dockworkers
cancer, bladder cancer, and motor
vehicle accidents)
Cases and controls were from
Central State Teamsters who
had filed claims (requiring 20-year
tenure).
As 1964 cut-off point: Exposure based on job titles not
validated
SS OR= 1.64 for long-haul drivers
with 13+ years of employment Possible misclassification of exposure
and smoking, based on next-of-kin
Positive trend test for long-haul information
drivers (p=0.04)
Lack of sufficient latency
SS OR = 1.89 for diesel truck
drivers of 35+ years of employment
Bqffetta et From 18 hospitals (since 1969) A priori aggregation of
al. (1990) 2,584 male lung cancer cases occupations categorized
matched to either one control (69) into low probability,
or two controls (2,515) were possible exposure (19
drawn. Matched on age, hospital, occupations), and
and year of interview probable exposure (13
occupations) to diesel
exhaust
OR slightly below unity SNS
No verification of exposure
Duration of employment used as
surrogate for dose
Number of individuals exposed to
diesel exhaust was small
Emmelin et 50 male lung cancer cases from Indirect diesel exhaust SS OR for high-exposure group =
al. (1993) 15 ports (worked for at least exposure assessment done 6.8
6 months between 1950 and 1974), based on (I) exposure
154 controls matched on age and intensity, (2)
port characteristics of
ventilation, (3) measure
of proportion of time in
higher exposure jobs
Numbers of cases and controls are
small
Very few nonsmokers
Lack of exposure information on
asbestos
No latency analysis
Abbreviations: OR = odds ratio; RR = relative risk; SNS = statistically nonsignificant; SS = statistically significant.
-------
1 27 substances as exposures. Unconditional logistic regression models were used to provide
2 effect estimates adjusted for potential confounding by nonoccupational variables such as age,
3 family income, ethnicity, Quetelet index, and respondent status. Diesel exhaust exposure was
4 identified by reporting a history of work as a truck driver or as a heavy machinery operator.
5 The odds ratio for "possible exposure" to diesel exhaust is 1.47 (95% CI = 1.01, 2.13).
6 When the criteria for exposure were "substantial" (defined by rating both Concentration and
7 frequency of exposure as medium or high), the odds ratio is 1.10 (95% CI= 0.72, 1.67). Diesel
8 exhaust showed an increase in risk with duration of exposure (1-10 vs. 11+ years). The odds
9 ratio for 11 or more years of exposure is 1.5 (95% CI = 1.1, 2.1). Risk is not positively
10 correlated with concentration (low vs. medium/high) or frequency of exposure (low/medium vs.
11 high).
12 This study identifies associations between diesel exhaust exposure as inferred by
13 occupational history and histologically confirmed incident prostate cancers. The crude exposure
14 assessment and lack of a substantive a priori hypothesis for the relation require that the results be
15 considered preliminary or exploratory. The lack of control for smoking and diminished effect
16 with increasing certainty of exposure also undermine the credibility of the observed association.
17
18 7.2.4. Summaries of Studies and Meta-Analyses of Lung Cancer
1 9 7.2.4.1. Cohen and Higgins (1995): Health Effects of Diesel Exhaust: Epidemiology
20 The Health Effects Institute (HEI) reviewed all published epidemiologic studies on the
21 health effects of exposure to diesel exhaust available through June 1993 identified by a
22 MEDLINE search and by reviewing the reference sections of published research and earlier
23 reviews. HEI identified 35 reports of epidemiologic studies (16 cohort and 19 case-control) of
24 the relation of occupational exposure to diesel emissions and lung cancer published between
25 1957 and 1993,
2 6 HEI reviewed the 3 5 reports for epidemiologic evidence of health effects of exposure to
27 diesel exhaust for lung cancer, other cancers, and nonmalignant respiratory disease. They found
*>r> ..i.... .1 i-- .. ...JL. . •. JC—i TI__ ;j „ >._ j .c>_-.i . ,. ±1 i
^.o uiai me uaia wcie suuugc&i lui iuug coiicci. me cviucucc auggcaicu uuu Gc^ujjauuuoi cApiouic
OO *>^>. A*r*r***l ^-**>t«r»»«-*-* ^^/-*«-»-» A-^rr*i*f*f*, f**^iTw*r*£*r* i»*/%*pa*»ci«=»f *V*^ «-r»+*a r*£ 1~TtTT /*o*t f^at" ^*t' O^0£ +r% A AO^ tT"
30 exposed workers generally, and to a greater extent among workers with prolonged exposure.
31 They also found that the results are not explicable by confounding caused by cigarette smoking
32 or other known sources of bias.
33 Control for smoking was identified in 15 studies. Six studies (17%) reported relative risk
34 estimates less than one; 29 studies (83%) reported at least relative risk indicating positive
3b association. Twelve studies indicating a relative nsk greater than 1 had ysu/o confidence
36 intervals, which excluded unity.
1175/99 7-48 DRAFT-DO NOT CITE OR QUOTE
-------
The authors conclude that epidemiologic data consistently show weak associations
between exposure to diesel exhaust and lung cancer. They find that the evidence suggests that
long-term exposure to diesel exhaust in a variety of occupational circumstances is associated
4 with a 1.2- to 1.5-fold increase in the relative risk of lung cancer compared with workers
5 classified as unexposed. Most of the studies that controlled for smoking found that the
6 association between increased risk of lung cancer and exposure to diesel exhaust persisted after
7 such controls were applied, although in some cases the excess risk was lower. None of the
8 studies measured exposure to diesel emissions or characterized the actual emissions from the
9 source of exposure for the time period most relevant to the development of lung cancer. Most
10 investigators classified exposure on the basis of work histories reported by subjects or their next
11 of kin, or by retirement records. Although these data provide relative rankings of exposure, the
12 absence of concurrent exposure information is the key factor that limits interpretation of the
13 epidemiologic findings and subsequently their utility in making quantitative estimates of cancer
14 risks.
15 This is a comprehensive and thorough narrative review of studies of the health effects of
16 diesel exhaust. It does not undertake formal estimation of summary measures of effect or
17 evaluation of heterogeneity in the results. The conclusion drawn about the consistency of the
results is based on the author's assessment of the failure of potential biases and alternative
explanations for the increase in risk to account for the observed consistency. In many if not most
20 studies, the quality of the data used to control confounding was relatively crude. Although the
21 studies do include qualitative assessment of whether control for smoking is taken into account,
22 careful scrutiny of the quality of the control or adjustment for smoking among the studies is
23 absent. This leaves open the possibility that prevalent residual confounding by inadequate
24 control for smoking in many or most studies may account for the consistent associations seen.
25
26 7.2.4.2. Bhatia et aL (1998): Diesel Exhaust Exposure and Lung Cancer
2 7 Bhatia et al. (1998) report a meta-analysis of 29 published cohort and case-control studies
28 of the relation between occupational exposure to diesel exhaust and lung cancer. A search of the
29 epidemiologic literature was conducted for all studies concerning lung cancer and diesel exhaust
30 exposure. Occupational studies involving mining were excluded because of concern about the
31 possible influence of radon and silica exposures. Studies in which the minimum interval from
32 time of first exposure to end of followup was less than 10 years, and studies in which work with
33 diesel equipment or engines could not be confirmed or reliably inferred, were excluded. When
34 studies presented risk estimates for more than one specific occupational category of diesel
11/5/99 7-49 DRAFT-DO NOT CITE OR QUOTE
-------
1 exhaust-exposed workers, the subgroup risk estimates were used in the meta-analysis.
2 Smoking-adjusted effect measures were used when present.
3 Thirty-five studies were identified in the literature search, of which 23 met the criteria for
4 inclusion in the meta-analysis. The observed relative risk estimates were greater than 1 in 21 of
5 these studies; this result is unlikely to be due to chance. The pooled relative risk weighted by
6 study precision was 1.33 (95% CI = 1.24, 1.44) indicated increased relative risk for lung cancer
7 from occupational exposure to diesel exhaust. Subanalyses by study design (case-control and
8 cohort studies) and by control for smoking produced results that did not differ from those of the
9 overall pooled analysis. Cohort studies using internal comparisons showed higher relative risks
10 than those using external comparisons. (See Figure 7-1.)
11 Bhatia and colleagues conclude that the analysis shows a small but consistent increase in
12 the risk for lung cancer among workers with exposure to diesel exhaust. The authors evaluate the
13 dependence of the relative risk estimate on the presence of control for smoking among studies,
14 and provide a table that allows assessment of whether the quality of the data contributing to
1 5 control for smoking is related to the relative risk estimates (albeit in a limited number of studies).
16 Bhatia et al. assert that residual confounding is not affecting the summary estimates or
17 conclusions for the following reasons: (1) the pooled relative risks for studies adjusted for
1 8 smoking were the same as those for studies not adjusting for smoking; (2) in those studies giving
1 9 risk estimates adjusted for smoking and risk estimates not adjusted for smoking, there was only a
20 small reduction in the pooled relative risk from diesel exhaust exposure; and (3) in studies with
21 internal comparison populations, in which confounding is less likely, the pooled relative risk
22 estimate was 1.43.
23 The validity of this assessment depends on the adequacy of control for smoking in the
24 individual studies. If inadequate adjustment for smoking is employed and residual confounding
25 by cigarette smoking pertains in the result of the individual studies, then the comparisons and
26 contrasts of the pooled estimates they cite as reasons for dismissing the effect of residual
27 confounding by smoking will remain contaminated by residual confounding in the individual
28 studies. In fact, Bhatia et al. erroneously identify the treatment of the smoking data in the.main
29 analysis for the 1987 report by Garshick et ai. as a continuous variable representing pack-years of
30 smoking, whereas the analysis actually dichotomized the pack-years data into two crude dose
31 categories (above and below the 50 pack-years level). This clearly reduced the quality of the
32 adjustment for smoking, which already suffered from 'die fact that information on cumulative
33 cigarette consumption was missing for more than 20% of the lung cancer cases. In this instance,
34 the consistency between the adjusted and unadjusted estimates of the relative risk for diesel
3 5 exhaust exposure may be attributable to failure of adj ustment rather than lack of confounding by
11/5/99 7-50 DRAFT-DO NOT CITE OR QUOTE
-------
0.5
RR •stlmates * 95% Cl
1 1.S
All Studies
CaseControl Studies
Cohort Studies
Internal Comparison
Population
External Comparison
Population
Smoking Adjusted
Smoking Not Adjusted
Sub-analysis by
Occupation
Railroad Workers
Truck Drivers
Bus Workers
1 0
1 C
1
1 1
Figure 7-1. Pooled relative risk estimates and heterogeneity-adjusted 95%
confidence intervals for all studies and subgroups of studies included in the
meta-analysis.
Source: Bhatiaetal., 1998.
11/5/99
7-51
DRAFT-DO NOT CITE OR QUOTE
-------
1 cigarette smoking; and pooled estimates of association of diesel exhaust with lung cancer derived
2 in the meta-analysis would remain confounded. A similar problem exists for the Bhatia et al.
3 representation of the control for confounding in the study by Boffetta and Stellman (1988). Such
4 mischaracterizations may indicate an overstatement by Bhatia et al. that the association of DE
5 and lung cancer is insensitive to adjustment.
6 An evaluation of the potential for publication bias is presented that provides reassurance
7 that the magnitude of published effects is not a function of the precision or study power;
8 however, this assessment cannot rule out the possibility for publication bias.
9
10 . 7.2.4.3. Lipsett and Campleman (1999): Occupational Exposure to Diesel Exhaust and Lung
11 Cancer: A Meta-Analysis
12 Lipsett and Campleman (1999) conducted electronic searches to identify epidemiologic
1 3 studies published between 1975 and 1995 of the relationship of occupational exposure to diesel
14 exhaust and lung cancer. Studies were selected based on the following criteria. (1) Estimates of
1 5 relative risks and their standard errors must be reported or derivable from the information
16 presented. (2) Studies must have allowed for a latency period of 10 or more years for
1 7 development of lung cancer after onset of exposure. (3) No obvious bias resulted from
18 incomplete case ascertainment in followup studies. (4) Studies must be independent: that is, a
1 9 single representative study selected from any set of multiple analyses of data from the same
20 population. Studies focusing on occupations involving mining were excluded because of
21 potential confounding by radon, arsenic, and silica, as well as possible interactions between
22 cigarette smoking and exposure to these substances in lung cancer induction.
23 Thirty of the 47 studies initially identified as relevant met the specified inclusion criteria.
24 Several risk estimates were extracted from six studies reporting results from multiple mutually
25 exclusive diesel-related occupational subgroups. If a study reported effects associated with
26 several levels or durations of exposure, the effect reported for the highest level or longest
27 duration of exposure was used. If estimates for several occupational subsets were reported, the
28 most diesel-specific occupation or exposure was selected. Adjusted risk estimates were used
29 when available.
30 Thirty-nine independent estimates of relative risk and standard errors were extracted.
31 Pooled estimates of relative risk were calculated using a random-effects model. Among study
32 populations most likely to have had substantial exposure to diesel exhaust, the pooled smoking
33 adjusted relative risk was 1.47 (95% CI = 1.29, 1.67). (See Figure 7-2.)
34 The between-study variance of the relative risks indicated the presence of significant
35 heterogeneity in the individual estimates. The authors evaluated the potential sources of
11/5/99 7-52 DRAFT-DO NOT CITE OR QUOTE
-------
1.8
U
5?
Si 1.6
•3
c
7» 1.4
1.2
o
o
a.
0.8
•* ,/ / / / /
Categories of Epidemiologies! Studies Included
Note. Cl = confidence interval; HWE = healthy worker effect.
Figure 7-2. Pooled estimates of relative risk of lung cancer in epidemiological
studies involving occupational exposure to diesel exhaust (random-effects
models).
Source: Lipsett and Campleman, 1999.
11/5/99
7-53
DRAFT-DO NOT CITE OR QUOTE
-------
1 heterogeneity by subset analysis and linear metaregressions. Major sources of heterogeneity
2 included control for confounding by smoking, selection bias (a healthy worker effect), and
3 exposure patterns characteristic of different occupational categories. A modestly higher, pooled
4 relative risk was derived for the subset of case-control studies, which, unlike the cohort studies,
5 showed little evidence of heterogeneity.
6 An evaluation of the potential for publication bias is presented that provides reassurance
7 that the magnitude of published effects is not a function of the precision or study power;
8 however, this assessment cannot rule out the possibility of publication bias.
9 Although a relatively technical approach was used in deriving summary estimates of
10 relative risk and the evaluation of possible sources of variation in the relative risks in this meta-
11 analysis, this approach should not be confused with rigorous evaluation of the potential
12 weaknesses among the studies included in the analysis. The heterogeneity attributable to
13 ' statistical adjustment for smoking was evaluated based on a dichotomous assessment of whether
14 control for smoking could be identified in the studies considered. This does not reflect the
1 5 adequacy of the adjustment for smoking employed in the individual studies considered. The
1 6 potential for residual confounding by inadequate adjustment for the influence of smoking
17 remains in the summary estimate of the relative risk.
18
19 7.2.5. Case-Control Studies of Bladder Cancer
20 7.2.5.1. Howe etaL (1980): Tobacco Use, Occupation, Coffee, Various Nutrients, and
21 Bladder Cancer
22 This is a Canadian population-based case-control study conducted in the provinces of
23 British Columbia, Newfoundland, and Nova Scotia. These areas were selected because they had
24 cancer registries and were believed not to have concentrations of high-risk industries. All
25 patients with newly diagnosed bladder cancer occurring in the three provinces between April
26 1974 and June 1976 were identified, and 77% of them were interviewed at home. A total of 480
27 male and 152 female case-control pairs were available for analysis. For each case, one
28 neighborhood control, matched by age (±5 years) and sex, was also interviewed at home to
29 obtain data on smoking, occupation, dietary sources of nitrites and nitrates that convert to
30 nitrosamines (nonpublic water supply and preserved meat products), and beverage consumption,
31 including a detailed history of coffee consumption. A detailed smoking history was obtained.
32 The occupational history included a chronological account of all jobs and the number of years
33 and months during which the respondent had worked in each job, experience in industries that
34 were suspected a priori to increase the risk of bladder cancer, and exposure to any jobs that
1175/99 7-54 DRAFT-DO NOT CITE OR QUOTE
-------
involved exposure to dust and fumes at the workplace. Relative risk estimates were computed
using the linear logistic model applied to individually matched case-control pairs.
A baseline comparison of cases and controls showed that, whereas male patients were
4 similar to controls on income and education, there was an excess of female cases with low family
5 incomes and low levels of educational attainment. For both sexes, the mean ages for cases and
6 controls did not differ, and the times required for the interview were similar. An analysis by the
7 a priori suspect industries showed elevated risks for a number of industries for males. These
8 included the chemical (RR = 7.5, 95% CI = 1.7, 67.6), rubber (RR = 5.0, 95% CI = 0.6, 236.5),
9 petroleum (RR = 5.3, 95% CI = 1.5,28.6), medicine (RR = 2.6, 95% CI = 0.9, 9.3), and spray
10 painting (RR = 1.8, 95% CI = 0.7, 4.6) industries. The excess risks were statistically significant
11 only for the petroleum and chemical industries. The estimates did not change when the analysis
12 was done separately for subjects who reported only one exposure and for those who reported
13 exposure to more than one suspect industry. The estimates also remained unchanged after
14 controlling for smoking. Too few females reported working in the a priori suspect industries to
15 make any meaningful contribution to the analysis. Among males, statistically nonsignificant
16 excess risks were observed for tanning, electric cable, photographic, commercial paint, tailoring,
17 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% CI = 0.8,11.8). Statistically significant excess risks
20 were observed for railroad workers (RR = 9.0, 95% CI = 1.2, 394.5) and welders (RR = 2.8, 95%
21 CI = 1.1, 8.8). For occupations other than those on the a priori list for males and females,
22 statistically significant excesses were detected for metal machinists (RR = 2.7, 95% CI = 1.1,
23 7.6), metal recorders (RR = 2.6, 95% CI = 1.0, 7.3), and nursery men (RR =5.5, 95% CI = 1.2,
24 51.1). Statistically nonsignificant excesses were also detected for exposure to two chemicals:
25 benzidine and its salts, RR = 1.3, and 6w-chloromethyl ether, RR = 5.0. A detailed analysis was
26 done for cigarette smoking, which demonstrated statistically significant increasing bladder
27 cancer risk with increasing duration of smoking, total lifetime consumption of packs of
28 cigarettes, and average frequency of cigarettes per day. In males the highest significant risk was
29 observed for latency of less than 35 years; after that time the risk reduced slightly with increasing
30 latency. In females the highest significant risk was for more than 35 years of latency. Risks
31 were elevated for males consuming all types of coffee and for females consuming instant coffee.
32 Hair dye usage in females and phenacetin usage in males and females carried no risk. Significant
33 risks for use of artificial sweeteners and use of nonpublic water supplies (nitrates and nitrites)
were found among males only.
11/5/99 7-55 DRAFT-DO NOT CITE OR QUOTE
-------
1 This study was mainly designed to evaluate the various risk factors for bladder cancer
2 such as smoking, coffee consumption, nitrates and nitrites in diet, and so on. The major
3 limitation of this study, as the authors noted, was that the three selected provinces did not have
4 high concentrations of industries suspected to be linked to bladder cancer. An excess risk was,
5 however, detected for railroad workers and for those in the "exposed to diesel and traffic fumes
6 category." Risks for those exposed to "diesel fumes only" were not available, nor do we know
7 the exact job title of the railroad workers and the type of engines they were operating. The
8 authors also did not detail the method by which they coded the information given by respondents
9 in response to questions on exposure to dust and fumes into the various categories they used in
10 the analysis. These analyses were done for subjects who reported having "ever been exposed"
11 versus "never been exposed" to these fumes, and although detailed chronological work histories
12 were obtained, no attempt was made to develop a lifetime cumulative exposure index to diesel
13 fumes. In multiple logistic regression models, the authors used the a priori high-risk
14 occupations; hence, nothing can be concluded about exposure to diesel exhaust for occupations
1 5 that were not part of that list. The authors provided no explanation on possible selection bias, as
16 only 77% of the eligible population was included in the study.
17
18 7.2.5.2. Wynder et al (1985): A Case-Control Study of Diesel Exhaust Exposure and Bladder
19 Cancer
20 A case-control study of diesel exhaust exposure and bladder cancer risk was conducted by
21 Wynder et al. (1985). Cases and controls were obtained from 18 hospitals located in six U.S.
22 cities between January 1981 and May 1983. Cases were individuals with histologically
23 confirmed primary cancer of the bladder, diagnosed within 12 months before the interview.
24 Controls were individuals with non-tobacco-related diseases who were matched to the cases by
25 age (within 8 years), race, year of interview, and hospital of admission. Women were excluded
26 from the study because the focus was on male-dominated occupations. A structured
27 questionnaire was administered in the hospital to cases and controls to elicit information on usual
28 occupation, smoking history, alcohol and coffee consumption, as well as other demographic
29 factors.
30 Two methods were used to define occupational exposure to diesel exhaust. First,
31 occupational titles defined by the industrial hygiene standards as probable high exposure were
32 classified as exposed or not exposed to diesel exhaust. The probable high-exposure category
33 consisted of bus and truck drivers, heavy equipment operators and repairmen, railroad workers,
34 and warehousemen. In the second method, guidelines set by NIOSH were used to classify
35 occupations based on exposure to diesel exhaust. In this method, the estimated proportion of
11/5/99 7-56 DRAFT-DO NOT CITE OR QUOTE
-------
1 exposed workers was computed for each occupational category by using the NIOSH estimates of
fthe 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
4 at least 20% of their employees exposed to diesel exhaust were defined as "high exposure," those
5 with 10% to 19% of their employees exposed as "moderate exposure," and those with less than
6 10% of their employees exposed as "low exposure." The odds ratio was used as a measure of
7 association to assess the relationship between bladder cancer and diesel exhaust exposure. The
8 overall participation among those eligible and available for interview was 75% and 72% in cases
9 and controls, respectively.
10 A total of 194 bladder cancer cases and 582 controls were examined, and the two groups
11 were found to be comparable by age and education. Except for railroad workers, who had
12 relative odds of 2.0 based on two cases and three controls (95% CI = 0.34, 11.61), the relative
13 odds were less than 1 for other diesel exhaust exposure occupations such as bus and truck
14 drivers, warehousemen, material handlers, and heavy equipment workers. When the risk was
1 5 examined using the NIOSH criteria for high, moderate, and low exposure, relative odds were
16 1.68 and 0.16 for high and moderate, respectively, with low as the referent group; neither was
17 statistically significant. Cases and controls were compared by smoking status. Cases were more
f likely to be current cigarette smokers than were controls. Current smokers of 1 to 20
cigarettes/day had relative odds of 3.64 (95% CI = 2.04, 6.49), current smokers of 21+
20 cigarettes/day had relative odds of 3.51 (95% CI = 2.00, 6.19), while ex-smokers had relative
21 odds of 1.72 (95% CI = 1.01,2.92). After controlling for smoking, there was no significant
22 increase in the risk of bladder cancer for occupations with diesel exhaust exposure compared
23 with occupations without diesel exhaust exposure. A synergistic effect between the two also was
24 not detected.
25 This study has two major methodologic limitations, both pertaining to exposure
26 classification. First, the use of "usual" occupation may have led to misclassification of those
27 individuals who had held a previous job with diesel exhaust exposure that was not their usual
28 occupation; this may have resulted in reduced power to detect weak associations. Second, since
29 there was no information on amount or duration of diesel exhaust exposure, no analysis of dose-
30 response relationships could be done. Also, no information was available on other confounding
31 risk factors of bladder cancer such as urinary retention, amphetamine abuse, and smoking within
32 the confined space of a truck cab, all of which are lifestyle factors specific to the truck-driving
33 occupation.
11/5/99 7-57 DRAFT—DO NOT CITE OR QUOTE
-------
1 7.2.5.3. Hoar and Hoover (1985): Truck Driving and Bladder Cancer Mortality in Rural New
2 England
3 This study investigated the relationship between the occupation of truck driving and
4 bladder cancer mortality in a case-control study in New Hampshire and Vermont. Cases
5 included all white residents of New Hampshire and Vermont who died from bladder cancer
6 (eighth revision of the ICD) between 1975 and 1979. Death certificates were provided by the
7 vital records and health statistics office of the two States, and the next of kin were traced and
8 interviewed in person. Two types of controls were selected for each case. One control was
9 randomly selected from all dther deaths, excluding suicides, and matched on State, sex, race, age
10 (=2 years), and year of death. The second control was selected with the additional matching
11 criterion of county of residence. Completed interviews were obtained from 325 (out of 410) next
1 2 of kin for cases and 673 (out of 923) for controls. Information on demographic characteristics,
13 lifetime occupational and residential histories, tobacco use, diet, and medical history was
14 obtained on each subject. The odds ratio was calculated to ascertain a measure of association
1 5 between truck driving and bladder cancer. Because separate analyses of the two control series
1 6 gave similar results, the two control series were combined. Also, because matched analyses
1 7 yielded results similar to those provided by the unmatched analyses, results of the unmatched
1 8 analyses were presented.
19 Sixteen percent (35) of the cases and 12% (53) of the controls had been employed as
20 truck drivers, yielding an odds ratio of 1.5 (95% CI = 0.9, 2.6) after adjustment for county of
21 residence and age at death. For New Hampshire, the odds ratio was 1.3 (95% CI = 0.7, 2.3), and
22 for Vermont, the odds ratio was 1.7 (95% CI = 0.8, 3.4). For a large number of subjects, the next
23 of kin were unable to give the durations of truck driving, and there was an inconsistent positive
24 association with years of truck driving. Crude relative odds were not altered after adjustment for
25 coffee drinking, cigarette smoking, and education as a surrogate for social class. Little variation
26 hi risks was seen when the data were analyzed by the industry in which the men had driven
27 trucks. No relationship was seen between age at which employment as a truck driver started and
28 occurrence of bladder cancer. Analysis by duration of employment as a truck driver and bladder
29 cancer showed a positive trend of increasing relative odds with increasing duration of
30 employment. The trend test was statistically significant (p=0.006). The odds ratio was
31 statistically significant for the 5 to 9 years of employment category only (OR = 2.9, 95% CI =
32 1.2, 6.7). Similarly, analysis by calendar year first employed showed a statistically significant
33 odds ratio for 1930 to 1949 (OR = 2 6,95% CJ = 1.3, 5,1.), whereas relative odds were not
"7 A cirrr»ifir»or»t tf*onVii*»r»tc \\rf*rf* prrmlnvpH nrinr tn 1 Q90 r>r 5»-fi"Pr 1QSO
V . V.^AJU.«.«. V MU.. *.-. _—•—J »— ~ - • — - — v.._gj |. - - _ _ - - .
11/5/99 7-58 DRAFT-BO NOT CITE OR QUOTE
-------
The effects of reported diesel exhaust exposure from fuel or engines in truck driving or
other occupations were then analyzed. An odds ratio of 1.8 (95% CI = 0.5, 7.0) was derived for
those who were exposed to diesel exhaust during their truck-driving jobs as compared to an odds
4 ratio of 1.5 (95% CI = 0.8, 2.7) for those not reporting diesel exhaust exposure. Analysis by
5 duration of exposure (0, 1 to 19 years, 20 to 29 years, 30 to 39 years, and 40+ years) to diesel
6 fuel or engines in other occupations, which were "self-reported" by participants, showed a
7 statistically significant positive trend (p=0.024) for bladder cancer, although none of the
8 individual odds ratios in these duration categories were statistically significant.
9 This study investigated an association between truck driving and bladder cancer in an
10 attempt to understand the reasons for the high rates of bladder cancer in rural areas of New
11 Hampshire and Vermont. Although an elevated odds ratio for bladder cancer (not statistically
12 significant) was observed for reported truck-driving occupations, there was insufficient evidence
13 to conclude that the excess risk of bladder cancer was due to exposure to diesel emissions. This
14 is because the excess bladder cancer risk was observed for all truck drivers irrespective of their
15 exposure to diesel emissions. Also, no information was provided on the confounding effects of
16 other aspects of the road environment such as urinary retention, amphetamine abuse, and
17 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.
20
21 7.2.5.4. Steenland et al (1987): A Case-Control Study of Bladder Cancer Using City
2 2 Directories as a Source of Occupational Data
23 The primary objective of the study was to test the usefulness of city directories as a
24 source of occupational data in epidemiologic studies of illness and occupational exposure.
25 Commercial city directories provide data on occupations and employers and are compiled from a
26 door-to-door yearly census of all residents 18 years old and older. The directories are available
27 in most medium-size cities in the United States. A unique feature of city directory data is that
28 they identify specific employers, and as the authors suggest, this information may be better than
29 death certificates for rapid, inexpensive, record-based epidemiologic studies.
30 A case-control study was conducted of male bladder cancer deaths in Hamilton County
31 (including Cincinnati), OH. This county was selected because it is in an industrialized area with
32 high bladder cancer rates and also because city directories cover approximately 85% of the.
33 people living in the county. A computerized list of all male bladder cancer deaths (n = 731) and
34 all other male deaths (n = 95,057), with the exclusion of deaths from urinary tract tumors and
pneumonia, that occurred between 1960 and 1982 was obtained from the Ohio Department of
11/5/99 7-59 DRAFT—DO NOT CITE OR QUOTE
-------
1 Vital Statistics. Death certificates had been coded by a nosologist according to the ICD code in
2 use at the time of death. A pool of six controls was created for each case matched on sex,
3 residence in Hamilton County at time of death, year of death, age at death (±5 years), and race.
4 Two types of analysis were performed, one based on city directory data and the other
5 based on death certificate data. In the former, cases and controls were restricted to individuals
6 who had at least one yearly directory listing with some occupational data. The first two controls
7 from the pool of six who met the requirements were selected. This analysis involved 648 cases
8 (627 cases had 2 controls and 21 cases had only 1 control) and 1,275 controls.
i
9 The death certificate analysis involved all 731 cancer deaths, with two controls per case.
10 In most cases, the same two controls were used in this analysis too. The usual lifetime
11 occupation and industry on the death certificate was abstracted from them. Data on occupation
12 and industry were coded with a three-digit U.S. census code using the method adopted by the
13 U.S. Bureau of the Census. Five of the occupational data were recorded for occupation and
14 industry by a second coder, with a high degree of reproducibility. Odds ratios were calculated
15 for bladder cancer using a Mantel-Haenszel procedure.
16 The city directory data identified four employers significantly associated with bladder
17 cancer deaths; only one of them was identified by the death certificate data, which provided only
18 lifetime type of industry rather than the name of a specific employer. The industries represented
19 by the four employers were a chemical plant, printing company, valve company, and machinery
20 plant. The city directory data analysis demonstrated significant positive associations for quite a
21 few occupations. The occupations that had at least 10 cases or more were engineers (OR = 3.00,
22 p=0.01), carpenters (OR = 2.36,/K0.01), tailors (OR = 2.56,p<0.01), and furnace operators (OR
23 = 2.5, p=0.03). Relative odds were increased significantly with increased duration of
24 employment (^20 years) for truck drivers (OR =12, p=0.01) and furnace operators (based on
25 four cases and no controls, p=0.05). For occupations hi which subjects had ever been employed,
26 a significant increase in the relative odds with increased duration of employment was observed
27 for the railroad industry (^20 years of employment, OR = 2.21, p<0.05). Both truck driving and
28 railroad industry occupations involve ciiescl emission exposures.
29 The analysis of death certificate data yielded associations in the same direction for most
30 of the occupations. A check of the validity of city directory data indicated that 77% of the
31 listings agreed with the Social Security earnings report for the employer in any given year. A
32 comparison of city directory and death certificate information on occupations indicated a match
33 for occupation between at least one city directory listing and occupation on death certificates for
34 68.3% of the study subjects.
11/5/99 7-60 DRAFT-DO NOT CITE OR QUOTE
-------
This study demonstrated that city directories are a relatively inexpensive and accessible
source of occupational data for epidemiologic studies. Limitations of this study include the
problem in tracing women because of the change from maiden to married name and the
4 availability of data for only the year of residence in the city. They are superior to death
5 certificates in being able to identify high-risk employers in specific geographic sites. Although
6 death certificate data reflect usual lifetime occupation, city directories yield data on short-term
7 jobs, some of which may involve critical exposure. Thus, a combination of the two approaches
8 may be most productive in record-based hypothesis-generating studies. The occupational data
9 were missing for 15%, whereas employer data were missing for 36% in the city directory. In the
10 context of the mentioned pros and cons of using city directories, this study found an excess risk
11 for bladder cancer among two occupations with potential diesel exposure: truck drivers and
1 2 railroad workers. Two sources of bias that may have influenced these findings are selection bias
13 arising from the use of deaths instead of incident cases, because survival for bladder cancer is
14 high, and the absence of data on confounding factors such as smoking, beverage consumption,
1 5 and medication use.
16
17 7.2.5.5. Iscovich etaL (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,
20 Argentina, to estimate the risk of bladder cancer associated with different types of tobacco,
21 beverages, and occupational exposures. Bladder cancer is one of the most common cancers
22 among males in the La Plata area.
23 Cases were selected from patients with a histologically confirmed diagnosis of bladder
24 cancer (transitional, squamous-cell, or nonspecific cell type) admitted to the 10 general hospitals
25 in the greater La Plata area (population in 1980 = 580,000) between March 1983 and December
26 1985. Cases with true bladder papilloma and individuals who were residents of greater La Plata
27 for less than 5 years were excluded. Of the 120 cases eligible to participate, 1 died prior to the
28 interview, 2 refused to participate, and the remaining 117 cases, representing 60% of the incident
29 cases registered in the registry, were interviewed. Two control groups (117 neighborhood and
30 117 hospital controls) were matched by sex and age (±5 years). Of the 117 cases, 99 were males
31 and 18 were females. Hospital controls, selected from the same hospital as the cases, were
32 hospitalized for the first time within 3 months of diagnosis of the illness of the cases. Twelve
33 percent of the hospital controls had illnesses known to be associated with tobacco smoking.
34^ Neighborhood controls were sampled from among persons living in the same block. The
interviewer proceeded north in the block where the case resided and selected the first control who
11/5/99 7-61 DRAFT—DO NOT CITE OR QUOTE
-------
1 met the matching criteria. Seven hospital controls could not be interviewed because of their poor
2 physical health and were replaced. Three neighborhood controls refused to participate and were
3 replaced. Cases and hospital controls were interviewed at the hospital and the neighborhood
4 controls at their homes to collect data on demographic, socioeconomic, and medical variables,
5 detailed smoking habits, and alcoholic and other beverages consumed.
6 The interviews were done by trained interviewers, two physicians and a social worker.
7 The cases and hospital controls were interviewed in the hospital by the physicians; hence, the
8 interviews could not be conducted "blind." A detailed occupational history was obtained for the
9 three occupations of longest duration and the most recent one. For each job title, the activity of
10 the plant and type of production were also ascertained. Job titles were coded according to the
11 International Labor Union (ILO) 1970 classification. Plant activity and type of production were
12 coded according to the United Nations 1975 classification categories. Information was also
13 collected on exposure to 33 chemical and physical agents, which included confirmed or
14 suspected bladder carcinogens. A detailed history of smoking habits was also obtained, and
1 5 individuals were categorized as current smokers if they were smoking at present or if they had
1 6 stopped smoking less than 2 years previously. Ex-smokers were those who ceased smoking at
17 least for 2 years or more than 2 years previously. For each subject a cumulative lifelong
18 consumption of cigarettes by type was estimated, and an average consumption of cigarettes/day
19 was computed.
20 Relative risks were computed for occupational factors using the unconditional logistic
21 regression method, adjusting for age and tobacco smoking. These risks were derived for those
22 who were ever employed in that occupation versus those who were never employed in that
23 occupation. Socioeconomic status of cases and neighborhood controls was similar, but there
24 were fewer professionals and more manual workers among hospital controls compared with
25 cases. Occupational variables included job title and type of activity of the plant. Significant
26 excess risks were observed for truck and railroad drivers (RR = 4.31, p
-------
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.
4 This case-control study was conducted primarily to determine the reasons for the high
5 rates of bladder cancer in the La. Plata region of Argentina. Only 60% of the cases registered in
6 the cancer registry were interviewed, and no information was provided for the remaining 40%
7 eligible nonrespondents to determine if the study sample was selectively biased in any way. The
8 sample size of 117 was small, and the analysis of males reduced it to 99. Although the use of
B two different types of control groups is a strength of this study, none of the interviews were done
10 blind, and it appears that the hospital interviews were done by the physicians and the
11 neighborhood interviews were done by the social worker. Job titles were used as surrogates of
12 exposure, but the authors state that although they attempted to analyze by an exposure index
13 derived from a job exposure matrix (details not provided), they found no difference in exposure
14 between cases and controls. This explanation is ambiguous. The authors also grouped truck and
1 5 railroad drivers together for reasons not mentioned and did not present separate risk estimates. A
16 table showing the distribution of cases and controls for selected activities or professions did not
17 indicate 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
20 provided for the data collected in the interviews on exposures to the 33 chemical and physical
21 agents, and it was not known if the truck and railroad drivers were operating diesel engines.
22 Although rare in the La. Plata area, the occupations known to be associated with bladder cancer
23 (dye, rubber, leather, and textile workers) are acknowledged by the authors.
24
2 5 7.2.5.6. Iyer et al (1990): Diesel Exhaust Exposure and Bladder Cancer Risk
26 This study is a hospital-based case-control study of bladder cancer and potential exposure
27 to diesel exhaust using data from a large ongoing case-control study of tobacco-related
28 neoplasms conducted by the American Health Foundation. An earlier study by Wynder et al.
29 (1985) looked at the relationship between occupational exposure to diesel exhaust and the risk of
30 bladder cancer. For this study, the objective was to evaluate the relationship between the
31 different measures of exposure to diesel exhaust, occupational and self-reported, and the risk of
0
32 bladder cancer. Cases comprised 136 patients with histologically confirmed primary cancer of
33 the urinary bladder interviewed at 18 hospitals in six U.S. cities. Two controls were selected per
34^ case, matched for sex, age (within 2 years), race, hospital, and year of interview. A total of 160
controls had non-tobacco-related malignancies distributed as follows: stomach cancer (6%),
11/5/99 7-63 DRAFT-DO NOT CITE OR QUOTE
-------
1 colorectal cancer (20%), prostate cancer (6%), and leukemia or lymphoma (8%). Among the 112
2 controls with nonmalignant diseases, 3% had benign neoplasms, 6% had hyperplasia of the
3 prostate, and 6% had dorsopathies. Distribution of the other nonmalignant illnesses was not
4 provided. Occupational history included information on usual occupation and up to five other
5 jobs. Exposure to diesel exhaust in hobby activities also was collected. For the purpose of this
6 analysis, occupations were aggregated a priori into three categories: low probability of exposure
7 (reference group), possible exposure, and probable exposure. Analyses were also done for self-
8 reported exposure to diesel exhaust. Risk estimates were obtained by unconditional logistic
9 regression using PROC LOGIST of SAS. Cases and controls were first compared by age, race,
10 education, and smoking habit. Cases were found to be less educated than controls (p<0.05).'
11 Crude odds ratios for diesel exhaust exposure, based on occupational or self-reported exposure,
12 were not significantly elevated after controlling for smoking and educational status (OR = 1.2,
13 95% CI = 0.8,2.0). When individual occupations were analyzed separately, truck drivers
14 showed no excess risk (OR = 0.48, 95% CI = 0.15,1.56).
1 5 The authors concluded that their study does not support the hypothesis of an association
16 between exposure to diesel exhaust and bladder cancer. Several significant limitations of
17 exposure assessment and analysis are evident in this study. In the introduction, the authors stated
18 that they refined the definition of exposure to diesel exhaust by obtaining a lifetime occupational
19 history, but in the methods section they stated that they restricted analysis to usual occupational
20 history and five other jobs, which was not that different from their earlier study (Wynder et al.,
21 1985). The terms, low probability of exposure, possible exposure, and probable exposure, also
22 were not clearly defined. Information on duration of employment or exposure was not presented,
23 and no attempts were made to validate occupational history. No information was available on
24 calendar years of employment in the truck-driving industry or the locomotive occupations.
25 Because diesel trucks and locomotives were introduced in the mid-1950s and the dieselization
26 was completed by 1960, it would be important to use 1960 as a cutoff date and to restrict analysis
27 to subjects who worked in these industries after that date. No information on nonrespondent
28 cases and controls was provided. The authors indicated in the methods section that cases were
29 individually matched to controls, but they performed an unmatched analysis to calculate the odds
30 ratios and did not address why they did not preserve the matching in the analysis, especially
31 because such an analysis could bias the risk estimates to unity.
1175/99 7-64 DRAFT-DO NOT CITE OR QUOTE
-------
7.2.5.7. Steineck et al. (1990): Increased Risk of Urothelial Cancer in Stockholm From 1985
to 1987, After Exposure to Benzene and Exhausts
This study was conducted to investigate the association between benzene, diesel, and
4 petrol exhausts as well as some other industry-related chemicals and the risk of urothelial cancer.
5 Designed as a population-based case-control study, it was conducted among all men born
6 between 1911 and 1945 and living in the County of Stockholm for all or part of the observation
7 period (September 15, 1985, to November 30, 1987). All incident cases of urothelial cancer and
8 squamous-cell carcinoma of the lower urinary tract were contacted for inclusion hi the study.
9 Controls were selected by stratified random sampling during the observation period from a
10 computerized register for the population of Stockholm. A postal questionnaire was sent to study
11 subjects at their homes to collect information on occupational history. The questions on
12 occupation included exposure to certain specified occupations/industries/chemicals and lists of
13 all jobs held and duration in each job. An industrial hygienist, unaware of case-control status,
14 classified subjects as being exposed or unexposed to 38 agents and groups of substances,
1 5 including 17 exposure categories with aromatic amines. Using all the exposure information, the
16 exposure period was defined and the annual dose was rated as low, moderate, or high based on
17 the accumulated dose (exposure duration multiplied by intensity of exposure) during the course
tof 1 average year for the defined exposure period. Swedish and international data were used to
classify subjects as exposed, based on air concentrations in the work environment that were
20 higher than for the general public, or skin contact with liquids of low volatility. To allow for
21 latency, the authors ignored exposures after 1981. Data were gathered from 256 cases and 287
22 controls. Controls were selected by stratified random sampling four times from the computerized
23 register during the observation period of the population of the County of Stockholm. These
24 subjects comprised 80% of eligible cases and 79% of eligible controls. Nine of the cases and
25 16% of the controls refused to participate in the study.
26 • The distribution of urothelial cancers was as follows: 5 tumors in the renal pelvis, 243
27 in the urinary bladder, 5 in the ureter, none in the urethra, and 3 at multiple sites. Two cases who
28 were exposed to a high annual dose of aromatic amines were omitted from all further analysis to
29 eliminate then- confounding effects. Crude relative risks were calculated for men classified as
30 exposed or not exposed to several substances. Twenty-five cases and 19 controls reported having
31 been exposed to diesel exhaust, yielding an odds ratio of 1.7 (95% CI = 0.9, 3.3). The
32 corresponding relative odds for petrol exhausts, based on 24 cases and 24 controls, were 1.0
33 (95% CI = 0.5,1.9). Odds ratios were then calculated for low, moderate, and high levels of the
34 annual dose adjusted for smoking and year of birth. For diesel exhausts, the odds ratio was 1.3
9 (95% CI = 0.6, 3.1) for low levels, 2.2 (95% CI = 0.7, 6.6) for moderate levels, and 2.9 (95% CI
11/5/99 7-65 DRAFT-DO NOT CITE OR QUOTE
-------
1 = 0.3, 30.0) for high levels, indicating a dose response. The corresponding odds ratios for petrol
2 exhausts were 0.6 (95% CI = 0.3, 1.3), 1.4 (95% CI = 0.5, 3.7), and 3.9 (95% CI = 0.4, 35.5).
3 Restricting the analysis to only moderate or high annual doses of exposure adjusted for
4 year of birth and smoking showed a sevenfold increased risk for subjects exposed to both diesel
5 and petrol exhausts (OR = 7.1, 95% CI = 0.9, 58.8). For exposure to diesel (OR = 1.1) and petrol
6 (OR = 1.0) exhausts alone, no excess risk was detected in this analysis. Odds ratios were
7 calculated for low, moderate, and high exposure to benzene, with rates of 1.7 (95% CI = 0.6, 5.1)
8 for low annual doses, 1.1 (95% CI = 0.3, 4.5) for moderate annual doses, and 3.0 (95% CI = 1.0,
9 8.7) for high annual doses.
10 The authors discuss misclassification and confounding as sources of bias in this study.
11 To examine misclassification they compared hygienist-assessed exposure and self-reported
12 exposure for printing ink and found a higher relative risk and fewer exposed subjects for
13 hygienist-assessed exposure, indicating that specificity was a problem for self-reported exposure.
14 It is not known to what extent this may have affected the risk estimates for diesel exhausts since
15 data on self-reported exposure to diesel are not presented. They also mention the possibility of
16 exposure misclassification from using an average annual dose in which a person exposed to an
17 agent at a high level for a few working days and a person exposed to a low level for many days
18 are both rated as exposed to low annual doses. Although statistically nonsignificant elevated
19' odds ratios of 1.3, 2.3, and 2.9 were derived for low, moderate, and high levels of diesel
20 exposure, the authors state that some of their subjects may have later worked in jobs with
21 benzene exposure, and because an elevated risk was detected for benzene exposure, this
22 confounding effect may explain some of the excess risk. An almost statistically significant
23 interaction was observed for exposure to combined diesel and petrol exhausts (OR = 7.1, 95% CI
24 = 0.9, 58.8), which changed to 5.1 (95% CI = 0.6,43.3) after adjustment for benzene exposure,
25 again providing evidence for the confounding role of benzene exposure in explaining some of the
26 observed results.
27 Table 7-3 summarizes the bladder cancer case-control studies.
28
29 7.2.6. Discussion and Summary
30 Certain extracts of diesel exhaust have been demonstrated as both mutagenic and
31 carcinogenic in animals and hi humans. Animal data suggest that diesel exhaust is a pulmonary
32 carcinogen among rodents exposed by inhalation to high doses over long periods of time.
33 Because large working populations are currently exposed to diesel exhaust and because
34 nonoccupational ambient exposures currently are of concern as well, the possibility that exposure
35
11/5/99 7-66 DRAFT-DO NOT CITE OR QUOTE
-------
Table 7-3. Epidemiologic studies of the health effects of exposure to diesel exhaust: case-control studies of bladder
cancer
Authors
Population studied
Diesel exhaust exposure
Results
Limitations
Howe et al. 480 male case-control pairs
(1980)
152 female case-control pairs
Cases diagnosed between April
1974 and June 1976 in three
Canadian provinces
Based on occupational SNS RR = 2.8 for diesel and traffic Exposure based on occupational
history of jobs involving fumes history, which was not validated
exposure to dust and SS RR = 9.00 for railroad workers
fumes
A priori suspect
industries
Diesel exhaust and traffic fumes were
combined
Only 77% of eligible population
included in the study
ON
-J
Wynder et 194 histologically confirmed male
al. (1985) cases between the ages of 20 and
80 years
582 matched controls (age,
race, year of interview, and
hospital of admission); diseases
not related to tobacco use
From 18 hospitals located in
six U.S. cities between
January 1981 and May 1983
Occupational titles were SNS ORs were 1.68 and 0.16 for
defined by Industrial
Hygiene Standard into
dichotomous "exposed"
and "not exposed"
Also defined by NIOSH
standards into "high
exposure," "moderate
exposure," and "low
exposure"
high and moderate exposure,
respectively, as compared to low
exposure
Exposure based on usual occupation
may have led to misclassification
Dichotomous classification made
dose-response analysis unattainable
No data on other confounders such as
smoking
s
o
z:
o
H
O
o
&
/o
c;
o
-------
Table 7-3. Epidemiologic studies of the health effects of exposure to diesel exhaust: case-control studies of bladder
cancer (continued)
Hoar and
Hoover
(1985)
•^J
00
Population-based, case-control
study
325 cases from the residents of
New Hampshire and Vermont who
died of bladder cancer between
1975 and 1979
A total of 673 controls were
chosen from other deaths during
the same time period
Two matched controls (age, sex,
race, state, year of death, and
Lifetime occupational
history obtained from
next of kin
SS OR = 2.9 for 5 to 9 years of
employment as truck driver but not
for *. 10 years of employment
Positive trend (p=0.006) observed
with increasing duration
of employment as truck driver
Exposure defined as occupation of
"truck driver" (i.e., it could have been
diesel or gasoline or both)
No histological confirmation of
bladder cancer diagnosis
No data on other confounders such as
other exposures, smoking, etc.
Steenland 648 male bladder cancer deaths Occupation or industry OR = 12 (/>=0.01) for truck drivers Exposure based on city directory or
et al. (1987) from Hamilton County, OH
1,275 matched controls from other
deaths (pool of six controls for
eacli case, excluding urinary tract
tumors and pneumonias matched
on sex, age at death, year of death,
race)
listed in city directory
and on death certificates
with £20 years of employment
OR = 2.21 (psO.05) for railroad
workers with >20 years of
employment
0
O
death certificate listing that was not
validated
Lack of controlling for confounders
City directory usually has short-term
job listing
Missing data on 15% of occupations
and 36% for employers in the
n
h-4
a
o
ja
o
G
O
H
W
-------
o
Table 7-3. Epidemiologic studies of the health effects of exposure to diesel exhaust: case-control studies of bladder
cancer (continued)
Iscovich 117 histologically confirmed
et al. (1987) bladder cancer cases (60% of
all incident cases)
117 hospital controls and
117 neighborhood controls
(matched on age and sex)
Cases and hospital controls from
10 general hospitals in greater
La Plata between March 1983
and December 1985
Past and present
occupational data
were collected by
questionnaire
An exposure index based
on a job exposure matrix
was generated
SS OR = 4.3 for truck and railway Exposure based on job held that was
drivers not validated
SS RR = 6.2 for oil refinery workers 40% of eligible cases were
nonrespondent
Small sample size
Interviewers were not "blind" to the
status of an individual, and this could
have biased the findings
Truck and railroad drivers were
grouped together
Iyer et al. 136 histologically confirmed
(1990) bladder cancer cases
Lifetime occupational
history
No excess found
272 controls, two each matched on Self-reported diesel
sex, age, race, hospital, and year of exhaust exposure
interview (160 malignant, 112
nonmalignant) Exposure aggregated a
priori into:
From 18 hospitals in six U.S. cities Low probability
Possible
Exposure based on self-report, which
was not validated
Although lifetime occupational
history was obtained, analysis was
restricted to usual occupation
A priori classification was ambiguous
3
n
-------
s
o
Z
O
H
n
Table 7-3. Epidemiologic studies of the health effects of exposure to diesel exhaust: case-control studies of
bladder cancer (continued)
Steineck Population-based study from
etal. (1990) County of Stockholm
Men born between 1911 and 1945
256 (243 bladder) urinary tract
cancer incident cases (80% of
eligibles)
287 controls (79% of eligibles)
from population of Stockholm
Observation period September 15,
1985, to November 30,1987
Occupational history
classified into exposed
and nonexposed by
industrial hygienist
"blind" toward case or
control status
Using all exposure
information, annual
dose rated as "low",
"moderate," and "high"
SNS OR = 1.3 for low, OR = 2.2 for Elaborate exposure history
moderate, and OR = 2.9 for high classification not used to advantage
exposure were observed for diesel by simultaneous adjustment
exposure
Misclassification in exposure may
have occurred
SNS OR = 7.1 observed for diesel
and gasoline exhaust combined
exposure
Small sample size of only 25 cases
and 19 controls were exposed to
diesel exhaust
Confounding by other exposures not
accounted for, except benzene
Abbreviations: OR = odds ratio; RR = relative risks; SNS = statistically nonsignificant; SS = statistically significant.
-------
1 to this complex mixture may be carcinogenic to humans has become an important public health
2 issue.
^P Because diesel emissions become diluted in the ambient air, it is difficult to study the
4 health effects in the general population. Nonoccupational exposure to diesel exhaust is
5 worldwide in urban areas. Thus, "unexposed" reference populations used in occupational cohort
6 studies are likely to contain a substantial number of individuals who are nonoccupationally
7 exposed to diesel exhaust. Furthermore, the "exposed" group in these studies is based on job
8 titles, which in most instances are not verified or correlated with environmental hygiene
9 measurement. The issue of health effect measurement is further complicated by the fact that
10 occupational cohorts tend to be healthy and have below-average mortality, usually referred to as
11 the "healthy worker effect." Hence, the usual standard mortality ratios observed in cohort
12 mortality studies are underestimations of real risk.
13 A major difficulty with the occupational studies considered here was measurement of
14 actual diesel exhaust exposure. Because all the cohort mortality studies were retrospective,
15 assessment of health effects from exposure to diesel exhaust was naturally indirect. In these
16 occupational settings, no systematic quantitative records of ambient air were available. Most
17 studies compared men in job categories with presumably some exposure to diesel exhaust with
18 either standard populations (presumably no exposure to diesel exhaust) or men in other job
^B categories from industries with little or no potential for diesel exhaust exposure. A few studies
20 have included measurements of diesel fumes, but there is no standard method for the
21 measurement. No attempt is made to correlate these exposures with the cancers observed in any
22 of these studies, nor is it clear exactly which extract should have been measured to assess the
23 occupational exposure to diesel exhaust. All studies have relied on the job categories or self-
24 report of exposure to diesel exhaust. In the studies by Garsbick et al. (1987, 1988), the diesel
25 exhaust exposed job categories were verified on the basis of an industrial hygiene survey done by
26 Woskie et al. (1988a,b). The investigators found that in most cases the job titles were good
27 surrogates for diesel exhaust exposure. Also, in the railroad industry, where only persons who
28 had at least 10 years of work experience were included in the study, the workers tended not to
29 change job categories over the years. Thus, a job known only at one point in time was a
30 reasonable marker of past diesel exhaust exposure. Unfortunately, the exposure was only
31 qualitatively verified. Quantitative use of this information would have been much more
32 meaningful. Occupations involving potential exposure to diesel exhaust are miners, truck
33 drivers, transportation workers, railroad workers, and heavy equipment operators.
34 With the exception of the study by Waxweiler et.al. (1973), no known studies of miners
have assessed whether diesel exhaust is associated with lung cancer. Currently, there are about
11/5/99 7-71 DRAFT-DO NOT CITE OR QUOTE
-------
1 385 underground metal mines in the United States. Of these, 250 have been permanently
2 operating and 135 have been intermittently operating (Steenland, 1986). Approximately 20,000
3 miners are employed, but not all of them are currently working in the mines. Diesel engines
4 were introduced in the metal mines in the early to mid-1960s. Although all these mines use
5 diesel equipment, it is difficult to estimate how many of these miners were actually exposed to
6 diesel fumes.
7 Diesel engines were introduced in coal mines at an even later date, and their use is still
8 quite limited. In 1983, approximately 1,000 diesel units were in place in underground coal
9 mines, up from about 200 units in 1977 (Daniel, 1984). The number of units per mine varies
10 greatly; one mine may account for more than 100 units.
11 Even if it were possible to estimate how many miners (metal and coal) were exposed to
12 diesel exhaust, it would be very difficult to separate out the confounding effects of other potential
13 pulmonary carcinogens, such as radon decay products or heavy metals (e.g., arsenic, chromium).
14 Furthermore, the relatively short latency period limits the usefulness of these cohorts of miners.
15
16 7.2.6.1. The Cohort Mortality Studies
17 The cohort studies mainly demonstrated an increase in lung cancer. Studies of bus
18 company workers by Waller (1981), Rushton et al. (1983), and Edling et al. (1987) failed to
19 demonstrate any statistically significant excess risk of lung cancer, but these studies have certain
20 methodological problems, such as small sample sizes, short followup periods (just 6 years in the
21 Rushton et al. study), lack of information on confounding variables, and lack of analysis by
22 duration of exposure, duration of employment, or latency that preclude their use in determining
23 the carcinogenicity of diesel exhaust. Although the Waller (1981) study had a 25-year followup
24 period, the cohort was restricted to employees (ages 45 to 64) currently in service. Employees
25 who left the job earlier, as well as those who were still employed after age 64 and who may have
26 died from cancer, were excluded.
27 Wong et al. (1985) conducted a mortality study of heavy equipment operators that
28 demonstrated a significantly increased risk of liver cancer in total and in various subcohorts. The
29 same analysis also showed statistically significant deficits in cancers of the large intestine and
30 rectum. Metastases from the cancers of the large intestine and rectum in the liver probably were
31 misclassified as primary liver cancer, which led to an observed excess risk. This study did
32 demonstrate a nonsignificant positive trend for cancer of the lung with length of membership and
33 latency. Analysis of deceased retirees showed a significant excess of lung cancer. Individuals
34 without work historic who started work prior to 1967. when records were not kept, mav have
11/5/99 7-72 DRAFT-DO NOT CITE OR QUOTE
-------
been in the same jobs for the longest period of time. Workers without job histories included
those who had the same job before and after 1967 and thus may have worked about 12 to 14
3 years longer; these workers exhibited significant excess risks of lung cancer and stomach cancer.
4 If this assumption about duration of jobs is correct, then these site-specific causes can be linked
5 to diesel exhaust exposure. One of the methodological limitations of this study is that most of
6 these men worked outdoors; thus, this cohort might have had relatively low exposure to diesel
7 exhaust. The authors did not present any environmental measurement data either. Because of
8 the absence of detailed work histories for 30% of the cohort and the availability of only partial
9 work histories for the remaining 70%, jobs were classified and ranked according to presumed
10 diesel exposure. Information is lacking regarding duration of employment in the job categories
11 (used for surrogate of exposure) and other confounding factors (alcohol consumption, cigarette
1 2 smoking, etc.). Thus, this study cannot be used to support a causal association or to refute the
13 same between exposure to diesel exhaust and lung cancer.
14 A 2-year mortality analysis by Boffetta and Stellman (1988) of the American Cancer
1 5 Society's prospective study, after controlling for age and smoking, demonstrated an excess risk
16 of lung cancer in certain occupations with potential exposure to diesel exhaust. These excesses
17 were statistically significant among miners (RR = 2.67, 95% CI = 1.63,4.37) and heavy
^fc equipment operators (RR = 2.6, 95% CI = 1.12, 6.06). The elevated risks were nonsignificant in
19 railroad workers (RR = 1.59) and truck drivers (RR = 1.24). A dose response was also observed
20 for truck drivers. With the exception of miners, exposure to diesel exhaust occurred in the three
21 other occupations showing an increase in the risk of lung cancer. Despite methodologic
22 limitations, such as the lack of representiveness of the study population (composed of volunteers
23 only, who were probably healthier than the general population), leading to an underestimation of
24 the risk and the questionable reliability of exposure data based on self-administered
25 questionnaires that were not validated, this study is suggestive of a causal association between
26 exposure to diesel exhaust and excess risk of lung cancer.
27 Two mortality studies were conducted by Gustavsson et al. (1990) and Hansen (1993)
28 among bus garage workers (Stockholm, Sweden) and truck drivers, respectively. An SMR of
29 122 was found among bus garage workers based on 17 cases. A nested case-control study was
30 also conducted hi this cohort. Detailed exposure matrices based on job tasks were assembled for
31 both diesel exhaust and asbestos exposures. Statistically significant increasing lung cancer
32 relative risks of 1.34,1.81, and 2.43 were observed for diesel exhaust indices of 10 to 20, 20 to
33 30, and >30, respectively, using 0 to 10 as a comparison group. Adjustment for asbestos
11/5/99 7-73 DRAFT—DO NOT CITE OR QUOTE
-------
1 exposure did not change the results. The main strength of this study is the detailed exposure
2 matrices; some of the limitations are lack of smoking histories and low power (small cohort).
3 Hansen ( 1 993), on the other hand, found statistically significant SMR of 1 60 due to
4 cancer of bronchus and lung. No dose response was observed, although the excesses were
5 observed in most of the age groups (30 to 39, 45 to 49, 50 to 54, 55 to 59, 60 to 64, and 65 to
6 74). There are quite a few methodologic limitations to this study. Exposure to diesel exhaust
7 was assumed in truck drivers for diesel-powered trucks, but no validation of exposure was
8 attempted. Smoking data were lacking, fpllowup period was short, and no latency analysis was
9 done. The findings of both these studies are consistent with the findings of other truck driver
1 0 studies.
1 1 Two mortality studies of railroad workers were conducted, by Howe et al. (1983) in
1 2 Canada and Garshick et al. (1988) in the United States. The Canadian study found relative risks
13 of 1.2 (pO.Ol) and 1.35 (pO.OOl) among "possibly" and "probably" exposed groups,
1 4 respectively. The trend test showed a highly significant dose-response relationship with
1 5 exposure to diesel exhaust and the risk of lung cancer. The main limitation of the study was the
1 6 inability to separate overlapping exposures of coal dust and diesel fumes. Information on jobs
1 7 was available at retirement only. There was also insufficient detail on the classification of jobs
18 by diesel exhaust exposure. The exposures could have been nonconcurrent or concurrent, but
1 9 because the data are lacking, it is possible that the observed excess could be due to the effect of
20 both coal dust and diesel fumes and not due to just one or the other. However, it should be noted
2 1 that, so far, coal dust has not been demonstrated to be a pulmonary carcinogen in studies of coal
22 miners, but lack of data on confounders such as asbestos and smoking makes interpretation of
23 this study difficult. When three diesel exhaust exposure categories were examined for smoking-
24 related diseases such as emphysema, laryngeal cancer, esophageal cancer, and buccal cancer,
25 positive trends were observed, raising a possibility that the dose-response demonstrated for diesel
26 exposure may have been due to smoking. The findings of this study are at best suggestive of
27 diesel exhaust bein0 » ''Jng carcinogen.
78 The roost defmitive evidence for linking diesel exhaust exposure to lung cancer comes
29 from the Garshick et al. (1988) railroad worker study conducted in the United States. Relative
30 risks of 1 .57 (95% CI = 1 . 1 9, 2.06) and 1 .34 (95% CI = 1 .02, 1 .76) were found for ages 40 to 44
3 1 and 45 to 49, respectively, after the exclusion of workers exposed to asbestos. This study also
3 2 found that the risk of lung cancer increased with increasing duration of employment. As mis was
33 a large cohort study with lengthy fcllcwup and adequate analysis, including dose response (based
34 uu Juiatiuii of employment as a surrogate) as v/ell 25 adjustment for other confounding f?
1 1 15199 7-74 DRAFT-DO NOT CITE OR QUOTE
-------
such as asbestos, the observed association between increased lung cancer and exposure to diesel
exhaust is more meaningful.
3
4 7.2.6.2. Case-Control Studies of Lung Cancer
5 Among the 10 lung cancer case-control studies reviewed in this chapter, only 2 studies
6 did not find any increased risk of lung cancer. Lerchen et al. (1987) did not find any excess risk
7 of lung cancer, after adjusting for age and smoking, for diesel fume exposure. The major
8 limitation of this study was a lack of adequate exposure data derived from the job titles obtained
9 from occupational histories. Next of kin provided the occupational histories for 50% of the cases
10 that were not validated. The power of the study was small (analysis done on males only, 333
11 cases). Similarly, Boffeta et al. (1990) did not find any excess of lung cancer after adjusting for
12 smoking and education. This study had a few methodological limitations. The lung cancer cases
13 and controls were drawn from the ongoing study of tobacco-related diseases. It is interesting to
14 note that the leading risk factor for lung cancer is cigarette smoking. The exposure was not
1 5 measured. Instead, occupations were used as surrogates for exposure. Furthermore, there were
16 very few individuals in the study who were exposed to diesel exhaust. On the other hand,
17 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
1 9 (1984) in workers who were exposed to diesel exhaust versus those who were not (OR = 1.4 and
20 1.7 with two different criteria), and by Damber and Larsson (1987) in professional drivers (OR =
21 1.2). These rates were adjusted for age and smoking. Both Williams et al. (1977) and Hall and
22 Wynder (1984) had high nonparticipation rates of 47% and 36%, respectively. Therefore, the
23 positive results found in these studies are underestimated at best. In addition, the self-reported
24 exposures used in the study by Hall and Wynder (1984) were not validated. This study also had
25 low power to detect excess risk of lung cancer for specific occupations.
26 The study by Benhamou et al. (1988), after adjusting for smoking, found significantly
27 increased risks of lung cancer among French motor vehicle drivers (RR = 1.42) and transport
28 equipment operators (RR = 1.35). The main limitation of the study was the inability to separate
29 the exposures to diesel exhaust from those of gasoline exhaust because both motor vehicle
30 drivers and transport equipment operators probably were exposed to the exhausts of both types of
31 vehicles.
32 Hayes et al. (1989) combined data from three studies (conducted in three different states)
33 to increase the power to detect an association between lung cancer and occupations with a high
potential for exposure to diesel exhaust. They found that truck drivers employed for more than
1175/99 7-75 DRAFT-DO NOT CITE OR QUOTE
-------
1 1 0 years had a significantly increased risk of lung cancer (OR =1.5, 95% CI = 1 . 1 , 1 .9). This
2 study also found a significant trend of increasing risk of lung cancer with increasing duration of
3 employment among truck drivers. The relative odds were computed by adjusting for birth
4 cohort, smoking, and State of residence. The main limitation of this study is again the mixed
5 exposures to diesel and gasoline exhausts, because information on type of engine was lacking.
6 Also, potential bias may have been introduced because the way in which the cause of death was
7 ascertained for the selection of cases varied in the three studies. Furthermore, the methods used
8 in these studies to classify occupational categories were different, probably leading to
9 incompatibility of occupational categories.
1 0 The most convincing evidence comes from the Garshick et al. (1987) case-control study
1 1 of railroad workers and the Steenland et al. (1990) case-control study of truck drivers hi the
1 2 Teamsters Union. Garshick et al. found that after adjustment for asbestos and smoking, the
1 3 relative odds for continuous exposure were 1 .39 (95% CI = 1 .05, 1 .83). Among the younger
1 4 workers with longer diesel exhaust exposure, the risk of lung cancer increased with the duration
15 of exposure after adjusting for asbestos and smoking. Even after the exclusion of recent diesel
1 6 exhaust exposure (5 years before death), the relative odds increased to 1.43 (95% CI = 1.06,
17 1 .94). This study appears to be a well-conducted and well-analyzed case-control study with
1 8 reasonably good power. Potential confounders were controlled adequately, and interactions
1 9 between diesel exhaust and other lung cancer risk factors were tested.
20 Steenland et al. (1990), on the other hand, created two separate work history files, one
2 1 from Teamsters Union pension files arid the other from next-of-kin interviews. Using duration of
22 employment as a categorical variable and considering employment after 1959 (when presumed
23 dieselization occurred) for long-haul drivers, the risk of lung cancer increased with increasing
24 years of exposure. Using 1964 as the cutoff, a similar trend was observed for long-haul drivers.
25 For short-haul drivers, the trend was positive with a 1959 cutoff but not when 1964 was used as
26 the cutoff. For truck drivers who primarily drove diesel trucks and worked for 35 years, the
27 relative odds were i.89. Tlic limitations of this study include possible rnisclassificciticns cf
20 e^pOSuic and snicking, lack cf Isvelc cf disse! sxpcr^rs, sinaller ncnexpesed. group, ?«^
29 insufficient latency period. Given these limitations, the findings of this study are probably
30 underestimated.
3 1 Emmelin et al. (1993) hi then- Swedish dockworkers from 15 ports found increased
32 relative odds of 6.8 (90% CI = 1.3 to 34.9). Intricate exposure matrices were created using three
33 different variables, but no direct exposure measurement was done. Of 50 cases and i 54 controls,
11/5/99 7-76 DRAFT-DO NOT CITE OR QUOTE
-------
only 6 individuals were nonsmokers. A strong interaction between smoking and diesel exhaust
was observed in this study.
3
4 7.2.6.3. Reviews and Meta-analyses of Lung Cancer
5 Three summaries of studies concerned with the relationship of diesel exhaust exposure
6 and lung cancer risk are reviewed. The HEI report is a narrative study of more than 35
7 epidemic logic studies (16 cohort and 19 case-control) of occupational exposure to diesel
8 emissions published between 1957 and 1993. Control for smoking was identified in 15 studies.
9 Six of the studies (17%) reported relative risk estimates less than 1, whereas 29 (83%) reported at
10 least 1 relative risk, indicating a positive association. Twelve studies indicating a relative risk
11 greater than 1 had 95% confidence intervals that excluded unity. These studies found that the
12 evidence suggests that occupational exposure to diesel exhaust from diverse sources increases the
13 rate of lung cancer by 20% to 40% in exposed workers generally, and to a greater extent among
14 workers with prolonged exposure. They also found that the results are not explicable by
15 confounding due to cigarette smoking of other known sources of bias.
16 Bhatia et al. (1998) identified 23 studies that met criteria for inclusion in the meta-
17 analysis. The observed relative risk estimates were greater than 1 in 21 of these studies. The
1^P pooled relative risk weighted by study precision was 1.33 (95% CI= 1.24,1.44), which indicated
19 increased relative risk for lung cancer from occupational exposure to diesel exhaust.
20 Subanalyses by study design (case-control and cohort studies) and by control for smoking
21 produced results that did not differ from those of the overall pooled analysis. Cohort studies
22 using internal comparisons showed higher relative risks than those using external comparisons.
23 Lipsett and Campleman (1999) identify 39 independent estimates of relative risk among
24 30 eligible studies of diesel exhaust and lung cancer published between 1975 and 1995. Pooled
25 relative risks for all studies and for study subsets were estimated using a random effect model.
26 Interstudy heterogeneity was also modeled and evaluated. A pooled smoking-adjusted relative
27 risk was 1.47 (95% CI = 1.29,1.67). Substantial heterogeneity was found in the pooled-risk
28 estimates. Adjustment for confounding by smoking, having a lower likelihood of selection bias,
29 and increased study power were all found to contribute to lower heterogeneity and increased
30 pooled estimates of relative risk.
31 There is some variability in the conclusions of these summaries of the association of
32 diesel exhaust and lung cancer. The three analyses find that smoking is unlikely to account for
33 the observed effects, and all conclude that the data support a causal association between lung
cancer and diesel exhaust exposure. On the other hand, Stober and Abel (1996), Muscat and
11/5/99 7-77 DRAFT-DO NOT CITE OR QUOTE
-------
1 Wynder (1995) and Cox (1997) call into question the assertions by Cohen and Higgins (1995),
2 Bhatia et al. (1997), and Lipsett and Campleman (1999) that the associations seen for diesel
3 exhaust and lung cancer are unlikely to be due to bias. They argue that methodologic problems
4 are prevalent among the studies, especially in evaluation of diesel engine exposure and control of
5 confounding by cigarette smoking. The conclusions of the two meta-analyses are based on
6 magnitude of pooled relative risk estimates and evaluation of potential sources of heterogeneity
7 in the estimates. Despite the statistical sophistication of the meta-analyses, the statistical models
8 used cannot compensate for deficiencies in the original studies and will remain biased to the
9 extent that bias exists in the original studies.
10 It should be noted that a recent publication by Bruske-Hohlfeld et al. (1999) found a
11 strong association between DE exposure and the occurrence of lung cancer. This pooled analysis
12 of two case-control studies has a large sample size, is adjusted for smoking and asbestos
13 exposures, and exposure to DE was estimated on the basis of job codes. This study is not
14 critiqued hi detail here but will be included when the document is finalized.
15
16 7.2.6.4. Case-Control Studies of Bladder Cancer
17 Of the seven bladder cancer case-control studies, four studies found increased risk in •
18. occupations with a high potential diesel exhaust exposure. A significantly increased risk of
19 bladder cancer was found in Canadian railroad workers (RR = 9.0; 95% CI = 1.2, 349.5; Howe et
20 al., 1980), truck drivers from New Hampshire and Vermont (OR = 2.9, j?<0.05; Hoar and
21 Hoover, 1985), and in Argentinean truck and railroad drivers (RR = 4.31, /K0.002; Iscovich et
22 al., 1987). A positive trend with increasing employment as truck driver (p=0.006) was observed
23 by Hoar and Hoover (1985) in their study of truck drivers from New Hampshire and Vermont.
24 Significantly increased risks also were observed with increasing duration of employment of ^20
25 years in truck drivers (OR = 12, /?=0.01) and railroad workers (OR = 2.21,/?<0.05; Steenland et
26 al., 1987). No significant increased risk was found for any diesel-related occupations in studies
27 by Wynder et al. (1985), Iyer et al. (1990), or Steineck et al. (1990). All these studies had several
28 limitations, including inadequate characterization of diesel exhaust exposure, lack of validation
29 of surrogate measures of exposure, and presence of other confounding factors (cigarette smoking,
30 urinary retention, concentrated smoke within the truck cab, etc.); most of them had small sample
31 sizes and none presented any latency analysis.
32
11/5/99 7-78 DRAFT-DO NOT CITE OR QUOTE
-------
7.2.6.5. Relevant Methodologic Issues
Throughout this chapter, various methodologic limitations of individual studies have
3 been discussed, such as small sample size, short followup period, lack of latency analysis, and
4 lack of data on confounding variables. However, two of the major methodologic concerns in
5 these studies are use of death certificates to determine cause of death and lack of data on cigarette
6 smoking, which is a strong risk factor for both lung cancer and bladder cancer. Death certificates
7 were used by all of the cohort mortality studies and some of the case-control studies of lung
8 cancer and case-control studies of bladder cancer to determine cause of death. Use of death
9 certificates could lead to misclassification bias. Studies of autopsies done between 1960 and
10 1971 demonstrated that lung cancer was overdiagnosed when compared with hospital discharge,
11 with no incidental cases found at autopsy (Rosenblatt et al., 1971). Schottenfeld et al. (1982)
12 also found an overdiagnosis of lung cancer among autopsies conducted in 1977 and 1978. On
13 the other hand, Percy et al. (1981) noted 95% concordance when comparing 10,000 lung cancer
14 deaths observed in the Third National Cancer Survey from 1969 to 1971 (more than 90% were
15 confirmed histologically) to death certificate coded cause of death. For bladder cancer, the
16 concordance rate was 91%. These more recent findings suggest that the diagnosis of lung cancer
17 as well as bladder cancer on death certificates is better than anticipated. Furthermore, an
1^B overdiagnosis of lung cancer or bladder cancer on death certificates would reduce the ability of
19 the study to detect an effect of diesel exhaust exposure.
20 A persistent association of risk for lung cancer and diesel exhaust exposure is observed
21 in more than 30 epidemiologic studies published over the past 40 years. Evaluation of whether
22 this association can be attributed to a causal relation between diesel exhaust exposure and lung
23 cancer requires careful consideration of whether chance, bias, or confounding might be likely
24 alternative explanations.
2 5 Many of the studies provide confidence intervals for their estimates of excess risk or
26 statistical tests, which indicate that it is unlikely that the individual study findings were due to
27 random variation. The persistence of this association between diesel exhaust and lung cancer
28 risk in so many studies indicates that the possibility is remote that the observed association in
29 aggregate is due to chance. It is unlikely that chance alone accounts for the observed relation
30 between diesel exhaust and lung cancer.
31 The excess risk is observed in both cohort and case-control designs, which contradicts
32 the concern that a methodologic bias specifically characteristic of either design (e.g., recall bias)
33 might account for the observed effect. Selection bias is certainly present in some of the
occupational cohort studies that use external population data in estimating relative risks, but this
11/5/99 7-79 DRAFT-DO NOT CITE OR QUOTE
-------
1 form of selection bias (a healthy worker effect) would only obscure, rather than spuriously
2 produce, an association between diesel exhaust and lung cancer. Several occupational
3 epidemiologic studies that use more appropriate data for their estimates are available. Selection
4 biases may be operating in some case-control studies, but it is not obvious how such a bias could
5 be sufficiently uniform in effect, prevalent, and strong enough to lead to the persistent
6 association seen in the aggregate data. Given the variety of designs used in studying the diesel
7 exhaust and lung cancer association and the number of studies in different populations, it is
8 unlikely that routinely studying noncomparable groups is an explanation for the persistent
9 association seen. Exposure information bias is certainly a problem for almost all of the studies
10 concerned. Detailed and reliable individual-level data on diesel exhaust exposure for the period
11 of time relevant to the induction of lung cancer are not available and are difficult to obtain.
12 Generally, the only information from which diesel exposure can be inferred is occupational data,
13 which is a poor surrogate for the true underlying exposure distribution. Study endpoints are
14 frequently mortality data taken from death certificate information, which is frequently inaccurate
1 5 and often does not fully characterize the lung cancer incidence experience of the population in
16 question. Using inaccurate surrogates for lung cancer incidence and for diesel exposure can lead
17 to substantial bias, and these shortcomings are endemic in the field. In most cases these
18 shortcomings will lead to misclassification of exposure and of outcome, which is nondifferential.
19 Nondifferential misclassification of exposure and/or outcome can bias estimates of a diesel
20 exhaust-lung cancer association, if one exists, toward the null; but it is unlikely that such
21 misclassification would produce a spurious estimate in any one study. It is even more unlikely
22 that it would bias a sufficient number of studies in a uniform direction to account for the
23 persistent aggregate association observed.
24 All the cohort studies considered for this report are retrospective mortality studies.
25 Smoking history is usually difficult to obtain in such instances. The smoking histories obtained
26 from surrogates (next of kin, either spouse or offspring) were found to be accurate by Lerchen
27 and Sainct (1986) and McLaughlin si a!. (1987). Lerchen and Samet did not detect any
28 coiisistent bias in the repcr* of nioarpttp; consnrnptinn. Tn contrast. oveTeuorting of cigarette
29 smoking by surrogates was observed by Rogot and Reid (1975), Kolonel et al., (1977), and
30 Humble et al. (1984). Kolonel et al. found that the age at which an individual started smoking
31 was reported within 4 years of actual age 84% of the time. These studies indicate that surrogates
32 were able to provide fairly credible information on the smoking habits of the study subjects. If
33 the surrogates of the cases were more likely to overtreport cigarette smoking compared with the
34 controls, then it might be harder to find an effect uf diesel exiia-usi because uiost of the increase
11/5/99 7-80 DRAFT-DO NOT CITE OR QUOTE
-------
in lung cancer would be attributed to smoking rather than to the effect of exposure to diesel
exhaust.
3~~ Many studies do not adjust for tobacco smoke exposure. These studies are correctly
4 dismissed as not contributing to the body of information suitable for causal inference. Several
5 studies do attempt to adjust for smoking. Sometimes the data are aggregate data and the methods
6 used for adjustment are indirect and rely on critical and unverifiable assumptions for effective
7 adjustment (Pfluger, 1994). Frequently, individual-level data are used to adjust estimates of
8 effect by conventional methods. Usually, these data are not a careful, detailed, and thorough
9 assessment of smoking behavior. Generally the classification of smoking behavior is crude
10 (smoker vs. nonsmoker) and cannot be considered to fully characterize actual exposure. Given
11 these shortcomings, a possibility remains that the statistical adjustment for smoking is not
12 completely effective, and residual confounding by smoking may persist to bias the measure of
13 the diesel exhaust-lung cancer association.
14
15 7.2.6.6. Criteria of Causal Inference
16 In most situations, epidemiologic data are used to delineate the causality of certain
17 health effects. Several cancers have been causally associated with exposure to agents for which
1|^ there is no direct biological evidence. Insufficient knowledge about the biological basis for
19 diseases hi humans makes it difficult to identify exposure to an agent as causal, particularly for
20 malignant diseases when the exposure was in the distant past. Consequently, epidemiologists and
21 biologists have provided a set of criteria that define a causal relationship between exposure and
22 the health outcome. A causal interpretation is enhanced for studies that meet these criteria.
23 None of these criteria actually proves causality; actual proof is rarely attainable when dealing
24 with environmental carcinogens. None of these criteria should be considered either necessary
25 (except temporality of exposure) or sufficient in itself. The absence of any one or even several of
26 these criteria does not prevent a causal interpretation. However, if more criteria apply, this
27 provides more credible evidence for causality.
28 Thus, applying the criteria of causal inference to the seven cohort mortality and eight
29 case-control studies hi which risk of lung cancer was assessed resulted in the following:
30
31 • Temporality: There is a temporality of exposure to diesel exhaust prior to the
32 occurrence of lung cancer hi every cohort and case-control study.
33
11/5/99 7-81 DRAFT-DO NOT CITE OR QUOTE
-------
1 • Strength of association: The strength of association between exposure and the
2 occurrence of lung cancer in the cohort studies showed a 30% to 57% higher risk
3 among exposed persons as compared with nonexposed (Howe et al., 1983; Wong
4 et al., 1985; Boffetta and Stellman, 1988; Garshick et al., 1988). In case-control
5 studies, the risk varied from 20% to 89% higher among exposed compared with
6 nonexposed (Williams et al., 1977; Hall and Wynder, 1984; Damber and Larsson,
7 1987; Garshick et al., 1987; Benhamou et al., 1988; Hayes et al., 1989; Steenland
8 et al., 1990; Gustavsson et al., 1990; Emmelin et al., 1993). Some of these studies
9 did adjust for the confounding effects of smoking, asbestos, and other exposures.
10 Furthermore, a recent publication by HEI (1995) demonstrates this strength of
11 association in graphic presentation (Figures 7-3 and 7-4). Meta-analyses by
1 2 Bhatia et al. and Lipsett et al. also show the pooled estimated RR of 1.33 and
13 1.47, respectively. Although the studies had smaller increases in lung cancer risk
14 and only some of the studies considered by HEI (1995) are considered in this
1 5 chapter, it demonstrates the lung cancer excesses consistently all across the
1 6 various populations.
Howe-j
Hall-j
Garshick (88) -j
Boffetta (90) -I
I
Schenker-]
Garshick (87) -j
Bums-i
]
Siemiatycki -j
Williams-j
Boffetta (88) -j
0.1 1 10
Relative Risk (95% Cl)
Figure 7-3. Lung cancer and exposure to diesel exhaust in railroad workers. • = Relative
risk adjusted for cigarette smoking; O = relative risk not adjusted for cigarette
smoking. For the two studies by Howe and Williams, confidence intervals were
not reported and could not be calculated.
Source: HEI, 1995.
11/5/99 7-82 DRAFT-DO NOT CITE OR QUOTE
-------
MendH
Hall-j
Leupker-
Ahlberg-
Siemiatycki -
Boffetta (88) -j
Hayes -j
Boffetta (90) -j
Steenland -j
I
Bumsi
Williams H
0.1 1 10
Relative Risk (95% Cl)
Figure 7-4. Lung cancer and exposure to diesel exhaust in truck drivers. • = Relative risk
adjusted for cigarette smoking; O = relative risk not adjusted for cigarette
smoking. For the study by Williams, confidence intervals were not reported and
could not be calulated. For the Steenland study, the data were gathered from
union reports of long-haul truck drivers; for the Boffetta (1988) study, the data
were self-reported by diesel truck drivers; and for the Siemiatycki study, they
were self-reported by heavy-duty truck drivers (personal communication).
urce: HEI, 1995.
1 • Consistency: Several cohort and case-control (including one nested case-control)
2 studies of lung cancer conducted in several populations in the United States and
3 Europe consistently found the same effect (i.e., lung cancer).
4
5 • Specificity: All of the above-mentioned studies found the same effect (i.e., lung
6 cancer).
7
8 • Biological gradient: The biological gradient, which refers to the dose-response
9 relationship, was observed in the cohorts of Canadian railway workers (Howe et
10 al., 1983), heavy bulldozer operators (Wong et al., 1985), and truck drivers who
11 had enrolled in the American Cancer Society's prospective mortality study
12 (Boffetta and Stellman, 1988). In the case-control studies, a dose response was
13 observed in railroad workers (Garshick et al., 1988; Hayes et al., 1989; Steenland
14 et al., 1990). Although other studies failed to observe a dose response, these
1 5 studies were methodologically limited due to confounding by other exposures and
lack of either quantitative data on exposure or surrogate data on dose.
11/5/99 7-83 DRAFT-DO NOT CITE OR QUOTE
-------
1 • Biological plausibility: Because diesel exhaust consists of a carbon core particle
2 with surface layers of organics and gases, the tumorigenic activity may reside in
3 one, some, or all of these components. As explained in Chapter 9, there is clear
4 evidence that the organic constituents have the capacity to interact with DNA and
5 give rise to mutations, chromosomal aberrations, and cell transformations, all
6 well-established steps in the process of carcinogenesis. Furthermore, these
7 organic chemicals include a variety of polycyclic aromatic hydrocarbons and
8 nitroaromatics, many of which are known to be pulmonary carcinogens.
9 Alternatively, Vostal (1986) suggests that "diesel" particles themselves induce
10 lung cancer, most likely via an epigenetic mechanism, if they are present at
11 sufficiently high-doses. This makes a convincing argument for biological
12 plausibility of lung cancer occurrence under some condition of exposure.
13
14 When the same causal inference criteria were applied to the seven case-control studies in
1 5 which risk of bladder cancer was assessed, the results were:
16
17 • Temporality: There is temporality of exposure to diesel exhaust prior to the
18 occurrence of bladder cancer.
19
20 • Strength of association: The relative odds of getting bladder cancer among
21 exposed compared with nonexposed ranged from 2 to 12 times higher (Howe et
22 al., 1980; Hoar and Hoover, 1985; Iscovich et al., 1987; Steenland et al., 1987).
23 None of these studies adjusted for other confounding effects such as cigarette
24 smoking, exposures to other chemicals, or urinary retention.
25
26 • Consistency: Four out of seven bladder case-control studies conducted in the
27 United States and abroad found increased relative odds of bladder cancer in the
28 exposed population. None of the cohort studies showed increased bladder cancer
29 mortality; however, people rarely die from bladder cancer, so bladder cancer
30 excess is unlikely to be detected hi mortality studies.
31
32 • Specificity: Four out of seven case-control studies found an excess of bladder
33 cancer. The specificity criterion, per se, does not apply in this particular instance
34 because these are case-control studies.
11/5/99 7-84 DRAFT-DO NOT CITE OR QUOTE
-------
• Biological gradient: Dose response was observed in two out of four studies
3 . showing increasing relative odds with increasing length of employment (Hoar and
4 Hoover, 1985; Steenland et al., 1987).
5
6 • Biological plausibility: It has been demonstrated that motor exhaust emissions
7 contain PAHs and nitro-PAHs (Stenberg et al., 1983; Rosenkranz and
8 Mermelstein, 1983). There is some evidence that nitro-PAHs may be responsible
9 for the induction of human bladder cancer. Nitro-PAHs can be metabolized to
10 aromatic amine derivatives, and some of these agents are known to be capable of
11 inducing urinary bladder cancer (Clayson and Garner, 1976). Furthermore, 1-
12 nitropyrene (1-NP) has been reported to be carcinogenic in the rat mammary
13 gland (Hirose et al., 1984); the structurally related 4-aminobiphenyl, which
14 induces bladder cancer hi humans, also induces mammary gland tumors in rats
1 5 (Hirose et al., 1984). Although the applicability of these experimental results to
16 humans is unknown, the laboratory evidence certainly suggests the biological
17 plausibility of diesel exhaust to be a urinary bladder carcinogen.
f
19 In summary, although some of the causality inference criteria do apply to bladder cancer,
20 the evidence for bladder cancer in populations exposed to diesel exhaust is inadequate. On the
21 other hand, all the causality inference criteria apply well to lung cancer. An excess risk of lung
22 cancer was observed in several cohort and case-control studies. A recent meta-analysis shows
23 the consistency of elevated risks in 23 of 29 diesel exposure epidemiologic studies, with
24 statistically significant relative risks of 1.33 (Bhatia et al., 1998). Lipsett et al. (1999) also found
25 a pooled estimate RR of 1.47 after adjusting for smoking. However, because of lack of actual
26 data on exposure to diesel exhaust in these studies and other subtle methodologic limitations, the
27 human evidence falls just short of being sufficient to call diesel exhaust a human carcinogen.
28
29 7.3. CARCINOGENICITY OF DIESEL EMISSIONS IN LABORATORY ANIMALS
30 This chapter summarizes studies that assess the carcinogenic potential of diesel exhaust
31 in laboratory animals. The first portion of this chapter summarizes results of inhalation studies.
32 Experimental protocols for the inhalation studies typically consisted of exposure (usually
33 chronic) to diluted exhaust in whole-body exposure chambers using rats, mice, and hamsters as
model species. Some of these studies used both filtered (free of particulate matter) diesel exhaust
11/5/99 7-85 DRAFT-DO NOT CITE OR QUOTE
-------
1 and unfiltered (whole) diesel exhaust to differentiate gaseous-phase effects from effects induced
2 by DPM and its adsorbed components. Other studies were designed to evaluate the relative
3 importance of the carbon core of the diesel particle versus that of particle-adsorbed compounds.
4 Finally, a number of exposures were carried out to determine the combined effect of inhaled
5 diesei exhaust and tumor initiators, tumor promoters, or co-carcinogens.
6 Particulate matter concentrations in the diesel exhaust used in these studies ranged from
7 0.1 to 12 mg/m3. In this chapter, any indication of statistical significance implies thatp^O.05
8 was reported in the reviewed publications. The experimental protocols and exposure atmosphere
9 characterizations are not described in detail here but may be found in Appendix A. A summary
10 of the animal inhalation carcinogenicity studies and their results is presented in Table 7-4.
11 . Results of lung implantation and intratracheal instillation studies of whole diesel
12 particles, extracted diesel particles, and particle extracts are reported in Section 7.3.3 and in
13 Tables 7-6 and 7-7. Studies destined to assess the carcinogenic effects of DPM as well as solvent
14 extracts of DPM following subcutaneous (s.c.) injection, intraperitoneal (i.p.) injection, or
15 intratracheal (itr.) instillation in rodents are summarized in Section 7.3.5. Individual chemicals
16 present in the gaseous phase or adsorbed to the particle surface were not included in this review
17 because assessments of those of likely concern (i.e., formaldehyde, acetaldehyde, benzene,
18 PAHs) have been published elsewhere (U.S. EPA, 1993).
19
20 7.3.1. Inhalation Studies (Whole Diesel Exhaust)
21 7.3.1.1. Rat Studies
22 The potential carcinogenicity of inhaled diesel exhaust was first evaluated by Karagianes
23 et al. (1981). Male Wistar rats (40 per group) were exposed to room air or diesel engine exhaust
24 diluted to a DPM concentration of 8.3 (± 2.0) mg/m3, 6 hr/day, 5 days/week for up to 20 mo. The
25 animals were exposed in 3,000 liter plexiglass chambers. Airflow was equal to 50 liters per
26 minute. Chamber temperatures were maintained between 25° and 26.5 °C. Relative humidity
27 ranged form 45% to 80%. Exposures were carried out during the daytime. The exhaust-
28 generating system and exposure atmosphere characteristics are presented in Appendix A. The
29 type of engine used (3-cylinder, 43 bhp diesel) is normally used in mining situations and was
11/5/99 7-86 DRAFT-DO NOT CITE OR QUOTE
-------
Table 7-4. Summary of animal inhalation V^inogenicity studies
23
Species/
Study strain
Karagianes et al. Rat/
(1981) Wistar
Kaplan et al. Rat/F344
(1983)
While et al.
(1983)
Heinrich et al. Rat/
(I986a,b) Wistar
Mohretal.
(1986)
Iwai et al. Rat/F344
(1986)
Takemoto et al. Rat/F344
(1986)
Mauderly el al. Rat/F344
(1987)
Sex/total
number
M.40
M.40
M.30
M.30
M, 30
M, 30
F.96
F.92
F.95
F,24
F.24
F.24
F. 12
F.2I
F, 15
F, 18
M + F, 230"
M + F, 223
M + F.22I
M + F.227
Exposure
atmosphere
Clean air
Whole exhaust
Clean air
Whole exhaust
Whole exhaust
Whole exhaust
Clean air
Filtered
exhaust
Whole exhaust
Clean air
Filtered
exhaust
Whole exhaust
Clean air
Clean air
Whole exhaust
Whole exhaust
Clean air
Whole exhaust
Whole exhaust
Whole exhaust
Particle
concentration
(mg/m1)
0
8.3
0
0.25
0.75
1.5
0
0
4.0
0
0
4.9
0
0
2-4
2-4
0
0.35
3.5
7.1
Other
treatment
None
None
None
None
None
None
• None
None
None
. None
None
None
None
DIPN"
None
DIPNk
None
None
None
None
Post-
Exposure exposure
protocol observation
6 hr/day, NA
5 days/week,
Tor up to
20 mo
20 hr/day, 8 mo
7 days/week, 8 mo
for up to 8 mo
15 mo 8 mo
19 hr/day, NA
5 days/week
for up to
35 mo
8 hr/day, NA
7 days/week,
for 24 mo
4 hr/day, NA
4 days/week,
18-24 mo
7 hr/day. NA
5 days/week
up to 30 mo
Tumor type and incidence (%)'
Adenomas
0/6 (0)
1/6(16.6)
Broncho-alveolar carcinoma
0/30 (0)
1/30(3.3)
3/30(10.0)
1/30(3.3)
Squamous cell
Adenomas Carcinomas tumors
0/96(0) 0/96(0) 0/96(0)
0/92 (0) 0/92 (0) 0/92 (0)
8/95 (8.4) 0/95 (0) 9/95 (9.4)
Adenocarcinoma
and Large cell and
adenosquamous squamous cell
Adenomas carcinoma carcinomas
1/22(4.5) 0/22(0) 0/22(0)
0/16(0) 0/16(0) 0/16(0)
3/19(0) 3/19(15.8) 2/1'; (10.5)
Adenoma Carcinoma
0/12(0) 0/12(0)
10/21 (47.6) 4/21 (19)
0/15(0) 0/15(0)
12/18(66.7) 7/18(38.9)
Adenocarcinoma
+ squamous cell Squamous
Adenomas carcinoma cysts
(0) (0.9) (0)
(0) (1.3) (0)
(2.3) (0.5) (0.9)
(0.4) (7.5) (4.9)
Comments
All tumors
0/96 (0)
0/92 (0)
17/95(17.8)'
All tumors
1/22 (4.5)'
0/16(0)
8/19
lA-y tvt
V**- * )
All
tumors
(0.9)
(1.3)
(3.6)'
(12.8)'
-------
Table 7-4. Summary of animal inhalation carcinogenicity studies (continued)
Lft
•j-4
oo
00
Species/
Study strain
Ishiimhi el al. Rut/F344
(I988a)
Heavy-duty
engine
Ishinishi et al. Ral/F344
(I988a)
Light duty
Heavy duty
Sen/tola!
number
M -f F, 123
M+F, 123
M + F. 125
M + F, 123
M + F, 124
NS,
NS.
NS,
NS,
NS,
NS,
NS,
NS,
NS
NS,
NS,
NS,
5
8
II
5
9
11
5
9
II
5
6
13
Particle
Exposure concentration Other Exposure
atmosphere (mg/mj) treatment protocol
Clean air
Whole exhaust
Whole exhaust
Whole exhaust
Whole exhaust
Whole exhaust
Whole exhaust
Whole exhaust
Whole exhaust
Whole exhaust
Whole exhaust
Whole exhaust
Whole exhaust
Whole exhaust
Whole exhaust
Whole exhaust
Whole exhaust
0
0.5
1.0
1.8
3.7
O.I
O.I
0.1
1.1
1.1
I.I
0.5
0.5
0.5
1.8
1.8
1.8
None I6hr/day,
None 6 days/week,
None for up to
None 30 mo
None
None I6hr/day,
None 6 days/week,
None for 12 mo
None
None
None
None I6hr/day,
None 6 days/week,
None for 12 mo
None
None
None
Post-
exposure
observation
NA
6 mo
12 mo
18 mo
6 mo
12 mo
18 mo
6 mo
12 mo
18 mo
6 mo
12 mo
18 mo
Tumor type and Incidence (%)'
Adenomas
0/123 (0)
0/123(0)
0/125(0)
0/123(0)
0/124(0)
Adenomas
0/5 (0)
0/8 (0)
0/1 1 (0)
0/5 (0)
0/9 (0)
0/1 1 (0)
0/5 (0)
0/9 (0)
0/1 1 (0)
0/5 (0)
0/6 (0)
0/13(0)
Comments
Adenosquamous Squamous cell
carcinomas carcinomas All tumors
1/123(0.8) 0/123(0) 1/123(0.8)
0/123(0) 1/123(0.8) 1/123(0.8)
0/125(0) 0/125(0) 0/125(0)
4/123(3.3) 0/123(0) 4/123(3.3)
6/1 24 (4.8) 2/1 24 ( 1 .6) 8/1 24 (6.5)'
Carcinomas All tumors
0/5 (0)
0/8 (0)
0/1 1 (0)
0/5 (0)
0/9 (0)
0/1 1 (0)
0/5 (0)
0/9 (0)
0/1 1 (0)
0/5 (0)
0/6 (0)
1/13(0)
0/5 (0)
0/8 (0)
0/1 1 (0)
0/5 (0)
0/9 (0)
0/1 1 (0)
0/5 (0)
0/9 (0)
0/1 1 (0)
0/1 1 (0)
0/6 (0)
1/13(0)
Primary lung tumors
Brightwell et al. Rat/344
(1989)
M + F.
M + F,
M + F,
M + F.
M + F,
M + F,
260
144
143
143
144
143
Clean air
Filtered
exhaust
(medium
exposure)
Filtered
exhaust (high
exposure)
Whole exhaust
Whole exhaust
Whole exhaust
0
0
0
0.7
2.2
6.6
None I6hr/day,
None 5 days/week,
Tor 24 mo
None
None
None
None
NA
3/260(1.2)
0/144 (0)
0/143(0)
1/143 (0.7)
1 4/144 (9.7)c
55/143(38.5)'
Tumor
incidence for
all rats dying
or sacrificed
» 24/25
(96%) after
24 mo
-------
Table 7-4. Summary of animal inhalation carcinoge^lty studies (continued)
00
VO
Study
llenrichelal.
(I989a)
Lewis et al.
(1989)
Takaki et al.
(.1989)
Light-duty
engine
Heinrich et al.
(1995)
Nikulaetal.
(1995)
Iwai et al.
(1997)
Species/ Sex/total
strain number
Rat/ F, NS
Wistar F, NS
F.NS
F.NS
F.NS
F.NS
Rat/F344 M + F.288"
Rat/F344 M + F, 123
M + F, 123
M + F.I25
M + F, 123
M + F, 124
Rat/ F, 220
Wistar F, 200
F.200
F. 100
F, 100
F, 100
Rat/F344 M + F.2I4"
M + F.2IO
M + F.212
M + F.2I3
M + F.2I1
F/344 121, F
108, K
I53.F
Exposure
atmosphere
Clean air
Whole exhaust
Filtered
exhaust
Clean air
Whole exhaust
Filtered
exhaust
Clean air
Whole exhaust
Clean air
Whole exhaust
Whole exhaust
Whole exhaust
Whole exhaust
Clean air
Whole exhaust
Whole exhaust
Whole exhaust
Carbon black
TiO,
Clean air
Whole exhaust
Whole exhaust
Carbon black
Carbon black
Clean air
Filtered air
Whole exhaust
Particle
concentration
(mg/m1)
0
4.2
0
0
4.2
0
0
2.0
0
O.I
0.4
I.I
2.3
0.
0.8
2.5
7.0
11.6
10.0
0
2.5
6.5
2.5
6.5
0
0
3.2-9.4
Other Exposure
treatment protocol
DPNJ
DPN11
DPNd
DPN'
DPN'
DPN'
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
19 hr/day,
5 days/week
Tor 24 to
30 mo
7 hr/day,
5 days/week.
24 mo
16 hr/day.
6 days/week.
for up to
30 mo
18 hr/day,
5 days/week,
for up to
24 mo
16 hr/day,
5 days/week
for up to
24 mo
NA
48-56 hr/day
48-56 hr/day
Post-
exposure
observation
NA
NA No tumors
Adenosquamous
carcinomas
NA 1/23 (0.8)
1/23 (0.8)
1/25(0.8)
0/23 (0)
1/24(8.1)
Adenomas
6 mo 0/217(0)
0/198(0)
2/200(1)
4/100(4)
13/100(13)
4/100(4)
Adenomas
6 weeks 1/214 (
-------
Table 7-4. Summary of animal inhalation carcinogenicity studies (continued)
vo
-J
I
Study
Orlhoeferetal.
(1981)
(Pepelko and
Peirano, 1983)
Kaplan et al.
(1982)
Species/ Sex/total Exposure
strain number atmosphere
Mouse/ M, 25 Clean air
Strong A
Whole exhaust
Whole exhaust
Mouse/ M + F, 40 Clean air
Jackson A
M + F.40 Whole exhaust
Mouse/
Jackson A F, 60 Clean air
F. 60 Clean air
F, 60 Whole exhaust
F, 60 Whole exhaust
M, 429 Clean air
M, 430 Whole exhaust
Mouse M, 4S8 Clean air
A/J M, 18 Clean air
M. 485 Whole exhaust
Particle
concentration
(mg/m1)
0
6.4
6.4
0
6.4
0
0
6.4
6.4
0
6.4
0
0
1.5
Other
treatment
None
None
UV
irradiated
None
None
None
Urethan1
None
Urethan*
None
None
None
Urethank
None
Post-
Exposure exposure
protocol observation
20 hr/day,
7 days/week,
for 7 weeks
26 weeks
26 weeks
20 hr/day, 8 weeks
7 days/week,
for 8 weeks
8 weeks
20 hr/day,
7 days/week.
for approx.
7 mo.
20 hr/day, 6 mo
7 days/week,
for 3 mo
Tumor type and incidence (%)'
3/22(13.6)
7/19(36.8)
6/22 (27.3)
Lung tumors
16/36(44.4)
1 1/34 (32.3)
4/58 (6.9)
9/52(17.3)
14/56(25.0)
22/59(37.3)
73/403(18.0)
66/368(17.9)
Pulmonary adenomas
144/458(31.4)
18/18(100)
165/485(34.2)
Pulmonary adenoma
Comments
O.I 3 tumors/
mouse
0.63 tumors/
mouse
0.27 tumors/
mouse
0.5 tumors/
mouse
0.4 tumors/
mouse
0.09 tumors/
mouse
0.25 tumors/
mouse
0.32 tumors/
mouse
0.39 tumors/
mouse
0.23 tumors/
mouse
0.20 tumors/
mouse
-------
Table 7-4. Summary of animal inhalation carcinogemHfy studies (continued)
VO
Study
Kaplan et al.
(1983)
White et al.
(1983)
Pepelko and
Peirano(l983)
Pepelko and
Peirano(l983)
lleinrich et al.
(I986a,b)
Species/ Sex/total Eiposure
strain number atmosphere
Mouse/ A/ M, 388 Clean air
J M. 388 Whole exhaust
M.399 Whole exhaust
M, 396 Whole exhaust
Mouse/ M + F, 260 Clean air
Sencar Clean air
Clean air
Whole exhaust
Whole exhaust
Whole exhaust
Mouse/ M + F, 90 Clean air
Strain A
Clean air
Whole exhaust
Whole exhaust
Clean air
Whole exhaust
Mouse/ M + F. 84 Clean air
NMRI M + F, 93 Filtered
exhaust
M + F, 76 Whole exhaust
Particle
concentration
(nig/m1)
0
0.25
0.75
1.5
0
0
0
12
12
12
0
0
12
12
0
12
0
0
4.0
Other
treatment
None
None
None
None
None
BHT1
Urethan*
None
BUT1
Urethan1
None
Exposure
(darkness)
Exposure
(darkness)
Urethan1"
Urethan"1
None
None
None
Post-
Exposure exposure
protocol observation
20 hr/day, NA
7 days/week,
Tor up to
8 mo
Continuous NA
for 1 5 mo
NA
19 hr/day, NA
5 days/week
for up to
30 mo
Adenomas
(5.1)
(12.2)
(8.1)
(10.2)'
(5.4)
(8.7)
Adenomas
9/84(11)
11/93(12)
11/76(15)
Tumor type and Incidence (%)'
130/388(33.5)
131/388(33.8)
109/399(27.3)
99/396 (25.0)
Carcinomas All tumors
(0.5) (5.6)
(1.7) (2.8)
(0.9) (9.0)
(1.0) (11.2)'
(2.7) (8.1)
(2.6) • (11.2)
All tumors
21/87(24)
59/237 (24.9)
10/80 12.5)
22/250(0.10)
66/75 (88)
42/75 (0.95)
Squamous cell
Adenocarcinoma tumors All tumors
2/84(2) - 11/84(13)
18/93(19)' — 29/93(31)'
13/76(17)' — 24/76(32)'
Comments
0.29 tumors/
mouse
0.27 tumors/
mouse
0.14
0.10
2.80
0.95
Takemoto et al. Mouse/ M + F, 45 Clean air 0 None
(1986) IRC M + F.69 Whole exhaust 2-4 None
Mouse/ M + F, 12 Clean air' 0 None
C57BL M + F, 38 Whole exhaust 2-4 None
4 hr/day,
4 days/week,
for 19-28 mo
4 hr/day,
4 days/week
for 19-28 mo
NA
NA
Adenoma Adenocarcinoma
3/45 (6.7) 1/45 (2.2)
6/69(8.7) 3/69(4.3)
-------
Table 7-4. Summary of animal inhalation carcinogenicity studies (continued)
Study
Hiinrichetal.
(1995)
Mauderly el al.
(1996)
Species/ Sex/total
strain number
Mouse/
C57BL/
6N
Mouse/
NMRI
Mouse/
NMRI
Mouse/
CD-I
F,
F.
F,
F,
F,
F.
. F,
F.
M + F
M + F
M + F
M + F
120
120
120
120
120
20
20
i20
157"
171
, 155
186
Exposure
atmosphere
Clean air
Whole exhaust
Particle-free
exhaust
Clean air
Whole exhaust
Carbon black
TiO,
Clean air
Whole exhaust
Particle-free
exhaust
Clean air
Whole exhaust
Whole exhaust
Whole exhaust
Particle Post-
concentration Other Exposure exposure
(rng/m1) treatment protocol observation
0
4.5
0
0
4.5
11.6
10
0
4.5
0
0
. 0.35
3.5
7.0
None
None
None
None
None
None
None
None
None
None
None
None
None
None
I8hr/day, 6 mo
5 days/week,
for up to 21
mo
I8hr/day. 9.5 mo
5 days/week
for up to
13.5 mo
I8hr/day, None
5 days/week,
23 mo
7 hr/day, 5 None
days/ week,
for up to 24
mo
Tumor type and incidence (%)"
1/12(8.3)
8/38(21.1)
Adenomas
(25)
\*"f t
(21.8)
(11.3)
(11.3)
(25)
(18.3)
(31.7)
Multiple Multiple
adenomas carcinomas
1/157(0.6) 2/157(1.3)
2/171(1.2) 1/171(0.6)
0/155(0) 1/155(0.6)
0/186(0) 0/186(0)
0/12(0)
3/38 (7.9)
Adenocarcinomas
(154)
\ * ^'^ /
(15.4)
(10)
(2.5)
(8.8)
(5.0)
(15)
Alveolar/
Adenomas/ bronchiolar
carcinoma adenoma
1/157(0.6) 10/157(6.4)
1/171 (0.6) 16/171 (9.4)
0/155(0) 8/155(5.2)
0/186(0) 10/186(5.4)
Comments
5.1% tumor
rate
8.5% tumor
rate
3. 5% tumor
rate
Alveolar/
bronchiolar
carcinoma
7/157(4.5)
5/171 (2.9)
6/155(3.9)
4/186(2.2)
-------
Table 7-4. Summary of animal inhalation carcinog^Pfty studies (continued)
in
-J
Jo
Species/
Study strain
Heinrich et al. Hamster/
(I986a,b) Syrian
Brightwell et al. Hamster/
(1989) Syrian
Golden
Sex/total
number
M + F, 96
M + F, 96
M + f, 96
M + F,
M + F. 202
M + F. 104
M + F. 104
M + F. 101
M + F. 102
M + F. 101
M + F. 204
M + F. 203
Particle Post-
Exposure concentration Other Exposure exposure
atmosphere (mg/m1) treatment protocol observation
Clean air
Filtered exhaust
Whole exhaust
Clean air
Clean air
Filtered
exhaust
(medium
dose)
Filtered
exhaust
(high dose)
Whole exhaust
Whole exhaust
Whole exhaust
Filtered
exhaust
(high dose)
Whole exhaust
0
0
4.0
0
0
0
0
0.7
2.2
6.6
0
6.6
None 19hr/day
None 5 days/week
for up to
None 30 mo NA
None 16hr/day, NA
DEN' 5 days/week,
DEN' for 24 mo
DEN'
DEN'
DEN'
DEN'
None
None
Tumor type and incidence (%)'
Squamous cell
Adenomas Adenocarcinoma tumors
0/96(0) 0/96(0) 0/96
0/96(0) 0/96(0) 0/96
0/96(0) 0/96(0) 0/96
Primary lung
tumors
7/202 (3.5)
4/104 (3.8)
9/104(8.7)
2/101 (2.0)
6/102(5.9)
4/101 (3.9)
1/204(0.5)
0/203 (0)
Comments
All tumors
0/96(0)
0/96(0)
0/96(0)
Respiratory
tract tumors
not related to
exhaust
exposure for
any of the
groups
•Table values indicate number with tumors/number examined (% animals with tumors).
""Number of animals examined for tumors.
'Significantly different from clean air controls.
JDipenlylnilrosamine; 6.25 me/kg/week s.c. during first 25 weeks of exposure.
'Dipentylnitrasamine; 12.5 mg/kg/week s.c. during first 25 weeks of exposure.
'Splenic lymphomas also detected in controls (8.3%), filtered exhaust group (37.5%) and whole exhaust group (25%).
•5.3% incidence of large cell carcinomas.
*l g/kg, i.p. I/week for 3 weeks starting 1 mo into exposure.
'Includes adenomas, squamous cell carcinomas, adenocarcinomas, adenosquamous cell carcinoma, and mesotheliomas.
'4.5 mg/dielhylnitrosamine (DENVkg, s.c., 3 days prior to start of inhalation exposure.
'Single i.p. dose I mg/kg at start of exposure.
'Butylated hydroxytoluene 300 mg/kg, i.p. for week 1, 83 mg/kg for week 2, and 150 mg/kg for weeks 3 to 52.
m!2 mg/m'from 12 weeks of age to termination of exposure. Prior exposure (in utero) and of parents was 6 mg/m'.
"120-121 males and 71-72 females examined histologically.
"Not all animals were exposed for full term, at least 10 males were killed at 3,6, and 12 mo of exposure.
NS = Not specified.
NA = Not applicable.
-------
1 connected to an electric generator and operated at varying loads and speeds to simulate operating
2 conditions in an occupational situation. To control the CO concentration at 50 ppm, the exhaust
3 was diluted 35:1 with clean air. Six rats per group were sacrificed after 4, 8, 1 6, and 20 mo
4 exposure for gross necropsy and histopathological examination.
5 The only tumor detected was a bronchiolar adenoma in the group exposed over 16 mo to
6 diesel exhaust. No lung tumors were reported in controls. The equivocal response may have
7 been caused by the relatively short exposure durations (20 mo) and small numbers of animals
8 examined. In more recent studies, for example, Mauderly et al. (1987), most of the tumors were
9 detected in rats exposed for more than 24 mo.
1 0 General Motors Research Laboratories sponsored chronic inhalation studies at the
1 1 Southwest Research Institute using male Fischer 344 rats, 30 per group, exposed to DPM
1 2 concentrations of 0.25, 0.75, or 1.5 mg/m3 (Kaplan' et al., 1983; White et al., 1983). The animals
1 3 were exposed in 12.6 m3 exposure chambers. Airflow was adjusted to provide 13 changes per
1 4' hour. Temperature was maintained at 22 ± 2 °C. The exposure protocol was 20 hr/day, 7
1 5 days/week for 9 to 1 5 mo. Exposures were halted during normal working hours for servicing.
1 6 Some animals were sacrificed following completion of exposure, while others were returned to
1 7 clean air atmospheres for an additional 8 mo. Control animals received clean air. Exhaust was
1 8 generated by 5.7-L Oldsmobile engines (four different engines used throughout the experiment)
1 9 operated at a steady speed and load simulating a 40-mph driving speed of a full-size passenger
20 car. Details of the exhaust-generating system and exposure atmosphere are presented in
21 Appendix A.
22 Although five instances of bronchoalveolar carcinoma were observed in 90 rats exposed
23 to diesel exhaust for 1 5 mo and held an additional 8 mo hi clean air, compared with none among
24 controls, statistical significance was not achieved in any of the exposure groups. These included
25 one tumor in the 0.25 mg/m3 group, three in the 0.75 mg/m3 group, and one hi the 1 .5 mg/m3
26 group. Rats kept in clean-air chambers for 23 mo did not exhibit any carcinomas. No tumors
27 were observed hi any of the 180 rats exposed to diesel exhaust for 9 or 15 mo without a recovery
28 period, or hi the respective controls for these groups. Equivocal results may again have been due
79 to less-tban-lifetime duration of the study as well as insufficient exposure concentrations.
UU Although the increases in rumor incidences in the groups cApobcJ for 1 5 ino and held an
3 1 additional 8 mo in clean air were not statistically significant, relative to controls, they were
32 slightly greater than the historic background incidence of 3.7% for this specific lesion hi this
33 strain of rat (Ward, 1983). The first definitive studies linking inhaled diesel exhaust to
34 induction of lung cancer in rats were reported by researchers hi Germany, Switzerland, Japan,
35 and the United States hi the mid-to-late 1980s. In a study conducted at the Fraunhofer institute
36 of Toxicology and Aerosol Research, female Wistar rats were exposed tor iy hr/day, 5
11/5/99 7-94 DRAFT— DO NOT CITE OR QUOTE
-------
1 days/week to both filtered and unfiltered (total) diesel exhaust at an average particulate matter
2 concentration of 4.24 mg/m3. Animals were exposed for a maximum of 2.5 years. The exposure
^fc system as described by Heinrich et al. (1986a) used a 40 kilowatt 1.6-L diesel engine operated
4 continuously under the U.S. 72 FTP driving cycle. The engines used European Reference Fuel
5 with a sulfur content of 0.36%. Filtered exhaust was obtained by passing engine exhaust through
6 a Luwa FP-65 HT 610 particle filter heated to 80 °C and a secondary series of filters (Luwa FP-
7 85, Luwa NS-30, and Drager CH 63302) at room temperature. The filtered and unfiltered
8 exhausts were diluted 1:17 with filtered air and passed through respective 12m3 exposure
9 chambers. Mass median aerodynamic diameter of DPM was 0.35 ± 0.10 jim (mean ± SD). The
10 gas-phase components of the diesel exhaust atmospheres are presented in Appendix A.
11 The effects of exposure to either filtered or unfiltered exhaust were described by
12 Heinrich et al. (1986b) and Stober (1986). Exposure to unfiltered exhaust resulted in 8
13 bronchoalveolar adenomas and 9 squamous cell tumors in 15 of 95 female Wistar rats examined,
14 for a 15.8% tumor incidence. Although statistical analysis was not provided, the increase
15 appears to be highly significant. In addition to the bronchioalveolar adenomas and squamous cell
16 tumors, there was a high incidence of bronchioalveolar hyperplasia (99%) and metaplasia of the
17 bronchioalveolar epithelium (65%). No tumors were reported among rats exposed to filtered
18 exhaust (n = 92) or clean air (n = 96).
Mohr et al. (1986) provided a more detailed description of the lung lesions and tumors
identified by Heinrich et al. (1986a,b) and Stober (1986). Substantial alveolar deposition of
21 carbonaceous particles was noted for rats exposed to the unfiltered diesel exhaust. Squamous
22 metaplasia was observed in 65.3% of the rats breathing unfiltered diesel exhaust, but not in the
23 control rats. Of nine squamous cell tumors, one was characterized as a Grade I carcinoma
24 (borderline atypia, few to moderate mitoses, and slight evidence of stromal invasion), and the
25 remaining eight were classified as benign keratinizing cystic tumors.
26 Iwai et al. (1986) examined the long-term effects of diesel exhaust inhalation on female
27 F344 rats. The exhaust was generated by a 2.4-L displacement truck engine. The exhaust was
28 diluted 10:1 with clean air at 20 °C to 25 °C and 50% relative humidity. The engines were
29 operated at 1,000 rpm with an 80% engine load. These operating conditions were found to
30 produce exhaust with the highest particle concentration and lowest NO2 and SO2 content. For
31 those chambers using filtered exhaust, proximally installed high-efficiency particulate air
32 (HEP A) filters were used. Three groups of 24 rats each were exposed to unfiltered diesel
33 exhaust, filtered diesel exhaust, or filtered room air for 8 hr/day, 7 days/week for 24 mo. Particle
34 concentration was 4.9 mg/m3 for unfiltered exhaust. Concentrations of gas-phase exhaust
35 components were 30.9 ppm NOX, 1.8 ppm NO2,13.1 ppm SO,, and 7.0 ppm CO.
1175/99 7-95 DRAFT—DO NOT CITE OR QUOTE
-------
1 No lung tumors were found in the 2-year control (filtered room air) rats, although one
2 adenoma was noted in a 30-mo control rat, providing a spontaneous tumor incidence of 4.5%.
3 No lung tumors were observed in rats exposed to filtered diesel exhaust. Nineteen of the 24
4 exposed to unfiltered exhaust survived for 2 years. Of these, 14 were randomly selected for
5 sacrifice at this time. Four of the rats developed lung tumors; two of these were malignant. Five
6 rats of this 2-year exposure group were subsequently placed in clean room air for 3 to 6 mo and
7 four eventually (time not specified) exhibited lung tumors (three malignancies). Thus, the lung
8 tumor incidence for total tumors was 42.1% (8/19) and 26.3% (5/19) for malignant tumors in rats
9 exposed to whole diesel exhaust. The tumor types identified were adenoma (3/19),
10 adenocarcinoma (1/19), adenosquamous carcinoma (2/19), squamous carcinoma (1/19), and
11 large-cell carcinoma (1/19). The lung tumor incidence in rats exposed to whole diesel exhaust
12 was significantly greater than that of controls (p<. 0.01). Tumor data are summarized in Table
13 7-4. Malignant splenic lymphomas were detected in 37.5% of the rats in the filtered exhaust
14 group and in 25.0% of the rats in the unfiltered exhaust group; these values were significantly
15 (psO.05) greater than the 8.2% incidence noted in the control rats. The study demonstrates
16 production of lung cancer in rats following 2-year exposure to unfiltered diesel exhaust. In
17 addition, splenic malignant lymphomas occurred during exposure to both filtered and unfiltered
18 diesel exhaust. This is the only report to date of tumor induction at an extrarespiratory site by
19 inhaled diesel exhaust in animals.
20 A chronic (up to 24 mo) inhalation exposure study was conducted by Takemoto et al.
21 (1986), hi which female Fischer 344 rats were exposed to diesel exhaust generated by a 269-cc
22 YANMAR-40CE NSA engine operated at an idle state (1,600 rpm). Exposures were 4
23 hours/day, 4 days/week. The animals were exposed in a 376-L exposure chamber. Air flow was
24 maintained at 120 L/min. Exhaust was diluted to produce a particle concentration of 2-4 mg/m3
25 Concentrations of the gas-phase components of the exhaust are presented in Appendix A. When
26 not exposed the animals were maintained in an air-conditioned room at a temperature of 24 ±
27 2°C and a relative humidity of 55 ± 5% with 12 hr of light and darkness. Temperature and
28 humidity in the exposure chambers was not noted. The particle concentration of the diesel
29 exhaust in the exposure chamber was 2 to 4 rng/rn3. B[a]P and 1-rdtrcpyrene concentrations were
30 C.oj anu 93 ug/'g of particles, respectively. No lung tur".™^ were reported, in the diesel-exposed
31 animals. It was also noted that the diesel engine employed in this study was originally used as an
32 electrical generator and that its operating characteristics (not specified) were different from those
33 for a diesel-powered automobile. However, the investigators deemed it suitable for assessing the
34 effects of diesel emissions.
11/5/99 7-96 DRAFT—DO NOT CITE OR QUOTE
-------
1 Mauderly et al. (1987) provided data affirming the carcinogenicity of automotive diesel
2 engine exhaust in F344/Crl rats following chronic inhalation exposure. Male and female rats
^^ were exposed to diesel engine exhaust at nominal DPM concentrations of 0.35 (n = 366), 3.5
4 (n = 367), or 7.1 (n = 364) mg/m3 for 7 hr/day, 5 days/week for up to 30 mo. Sham-exposed
5 (n = 365) controls breathed filtered room air. A total of 230,223,221, and 227 of these rats
6 (sham-exposed, low-, medium-, and high-exposure groups, respectively) were examined for lung
7 rumors. These numbers include those animals that died or were euthanized during exposure and
8 those that were terminated following 30 mo of exposure. The exhaust was generated by 1980
9 model 5.7-L Oldsmobile V-8 engines operated through continuously repeating U.S. Federal Test
10 Procedure (FTP) urban certification cycles. The engines were equipped with automatic
11 transmissions connected to eddy-current dynamometers and flywheels simulating resistive and
12 inertial loads of a midsize passenger car. The D-2 diesel control fuel (Phillips Chemical Co.) met
13 U.S. EPA certification standards and contained approximately 30% aromatic hydrocarbons and
14 0.3% sulfur. Following passage through a standard automotive muffler and tail pipe, the exhaust
15 was diluted 10:1 with filtered air in a dilution tunnel and serially diluted to the final
16 concentrations. The primary dilution process was such that particle coagulation was retarded.
17 Mokler et al. (1984) provided a detailed description of the exposure system. The gas-phase
18 components of the diesel exhaust atmospheres are presented in Appendix A. No exposure-
1Mt related changes in body weight or life span were noted for any of the exposed animals, nor were
^j there any signs of overt toxicity. Collective lung tumor incidence was greater (z statistic,
21 psO.05) in the high (7.1 mg/m3) and medium (3.5 mg/m3) exposure groups (12.8% and 3.6%,
22 respectively) versus the control and low (0.35 mg/m3) exposure groups (0.9% and 1.3%,
23 respectively). In the high-dose group the incidences of tumor types reported were adenoma
24 (0.4%), adenocarcinomas plus squamous cell carcinomas (7.5%), and squamous cysts (4.9%). In
25 the medium-dose group adenomas were reported in 2.3% of animals, adenocarcinomas plus
26 squamous cell carcinomas in 0.5%, and squamous cysts in 0.9%. In the low-exposure group
27 adenocarcinomas plus squamous cell carcinomas were detected in 1.3% of the rats. Using the
28 same statistical analysis of specific tumor types, adenocarcinoma plus squamous cell carcinoma
29 and squamous cyst incidence was significantly greater in the high-exposure group, and the
30 incidence of adenomas was significantly greater in the medium-exposure group. A significant
31 (p<0.001) exposure-response relationship was obtained for tumor incidence relative to exposure
32 concentration and lung burden of DPM. These data are summarized in Table 7-4. A logistic
33 regression model estimating tumor prevalence as a function of time, dose (lung burden of DPM),
34 and sex indicated a sharp increase in tumor prevalence for the high dose level at about 800 days
35 after the commencement of exposure. A less pronounced, but definite, increase in prevalence
with time was predicted for the medium-dose level. Significant effects were not detected at the
11/5/99 7-97 DRAFT—DO NOT CITE OR QUOTE
-------
1 low concentration. DPM (mg per lung) of rats exposed to 0.35, 3.5, or 7.1 mg of DPM/m3 for 24
2 mo were 0.6, 11.5, and 20.8, respectively, and affirmed the greater-than-predicted accumulation
3 that was the result of decreased particle clearance following high-exposure conditions.
4 In summary, this study demonstrated the pulmonary carcinogenicity of high
5 concentrations of whole, diluted diesel exhaust in rats following chronic inhalation exposure. In
6 addition, increasing lung particle burden resulting from this high-level exposure and decreased
7 clearance was demonstrated. A logistic regression model presented by Mauderly et al. (1987)
8 indicated that both lung DPM burden and exposure concentration may be useful for expressing
9 exposure-effect relationships.
10 A long-term inhalation study (Ishinishi et al., 1988a; Takaki et al., 1989) examined the
11 effects of emissions from a light-duty (LD) and a heavy-duty (HD) diesel engine on male and
12 female Fischer 344/Jcl rats. The LD engines were 1.8-L, 4-cyUnder, swirl-chamber-type power
13 plants, and the HD engines were 11-L, 6-cylinder, direct-injection-type power plants. The
14 engines were connected to eddy-current dynamometers and operated at 1,200 rpm (LD engines)
15 and 1,700 rpm (HD engines). Nippon Oil Co. JIS No. 1 or No. 2 diesel fuel was used. The 30-
16 mo whole-body exposure protocol (16 h/day, 6 days/week) used DPM concentrations of 0, 0.5, 1,
17 1.8, or 3.7 mg/m3 from HD engines and 0, 0.1, 0.4, 1.1, or 2.3 mg/m3 from LD engines. An
18 analysis of gas-phase components is presented in Appendix A. The animals inhaled the exhaust
19 emissions from 1700 to 0900 h. Sixty-four male rats and 59 to 61 female rats from each
20 exposure group were evaluated for carcinogenicity.
21 For the experiments using the LD series engines, the highest incidence of hyperplastic
22 lesions plus tumors (72.6%) was seen in the highest exposure (2.3 mg/m3) group. However, this
23 high value was the result of the 70% incidence of hyperplastic lesions; the incidence of adenomas
24 was only 0.8% and that of carcinomas 1.6%. Hyperplastic lesion incidence was considerably
25 lower for the lower exposure groups (9.7%, 4.8%, 3.3%, and 3.3% for the 1.1, 0.4, and 0.1
26 mg/m3 and control groups, respectively). The incidence of adenomas and carcinomas, combining
27 males and females, was not significantly different among exposure groups (2.4%, 4.0%, 0.8%,
28 2.4%, and 3.3% for the 2.3,1.1, 0.4, and 0.1 mg/m3 groups and the controls, respectively).
29 For the experiments using the HD series engines, the total incidence of hyperplastic
30 lesions, adenomas, and carcinomas was highest (26.6%) in the 3.7 mg/m3 exposure group. The
31 incidence of adenomas plus carcinomas for males and females combined equaled 6.5%, 3.3%,
32 0%, 0.8%, and 0.8% at 3.7, 1.8,1, and 0.4 mg/m3 and for controls, respectively. A statistically
33 significant difference was reported between the 3.7 mg/m3 and the control groups for the HD
34 series engines. The carcinomas were identified as adenomas, adenosquamous carcinomas, and
35 squamous cell carcinomas. Although the number of each was not reported, it was noted that the
11 /5/99 7-98 DRAFT—DO NOT CITE OR QUOTE
-------
1 majority were squamous cell carcinomas. A progressive dose-response relationship was not
2 demonstrated. Tumor incidence data for this experiment are presented in Table 7-4.
^^ The Ishinishi et al. (1988a) study also included recovery tests in which rats exposed to
4 whole diesel exhaust (DPM concentration of 0.1 or 1.1 mg/m3 for the LD engine and 0.5 or
5 1.8 mg/m3 for the HD engine) for 12 mo were examined for lung tumors following 6-, 12-, or 18-
6 mo recovery periods in clean air. The incidences of neoplastic lesions were low, and pulmonary
7 DPM burden was lower than for animals continuously exposed to whole diesel exhaust and not
8 provided a recovery period. The only carcinoma observed was in a rat examined 12 mo
9 following exposure to exhaust (1.8 mg/m3) from the HD engine.
10 Brightwell et al. (1986,1989) studied the effects of diesel exhaust on male and female
11 F344 rats. The diesel exhaust was generated by a 1.5-L Volkswagen engine that was computer-
12 operated according to the U.S. 72 FTP driving cycle. The engine was replaced after 15 mo. The
13 engine emissions were diluted by conditioned air delivered at 800 m3/h to produce the high-
14 exposure (6.6 mg/m3) diesel exhaust atmosphere. Further dilutions of 1:3 and 1:9 produced the
15 medium- (2.2 mg/m3) and low- (0.7 mg/m3) exposure atmospheres. The CO and NOX
16 concentrations (mean ± SD) were 32 ± 11 ppm and 8 ± 1 ppm in the high-exposure concentration
17 chamber. The inhalation exposures were conducted overnight to provide five 16-h periods per
18 week for 2 years; surviving animals were maintained for an additional 6 mo.
1^ For males and females combined, a 1.2% (3/260), 0.7% (1/144), 9.7% (14/144), and
2^^ 38.5% (55/143) incidence of primary lung tumors occurred in F344 rats following exposure to
21 clean air or 0.7,2.2, and 6.6 mg of DPM/m3, respectively (Table 7-4). Diesel exhaust-induced
22 tumor incidence in rats was dose-related and higher in females than in males (Table 7-4). These
23 data included animals sacrificed at the interim periods (6,12,18, and 24 mo); therefore, the
24 tumor incidence does not accurately reflect the effects of long-term exposure to the diesel
25 exhaust atmospheres. When tumor incidence is expressed relative to the specific intervals, a lung
26 tumor incidence of 96% (24/25), 76% (19/25) of which were malignant, was reported for female
27 rats in the high-dose group exposed for 24 mo and held in clean air for the remainder of their
28 lives. For male rats in the same group, the tumor incidence equaled 44% (12/27), of which 37%
29 (10/27) were malignant. It was also noted that many of the animals exhibiting tumors had more
30 than one tumor, often representing multiple histological types. The numbers and types of tumors
31 identified in the rats exposed to diesel exhaust included adenomas (40), squamous cell
32 carcinomas (35), adenocarcinomas (19), mixed adenoma/adenocarcinomas (9), and
33 mesothelioma (1). It should be noted that exposure during darkness (when increased activity
34 would result in greater respiratory exchange and greater inhaled dose) could account, in part, for
35 the high response reported for the rats. .
11 /5/99 7-99 DRAFT—DO NOT CITE OR QUOTE
-------
1 Lewis et al. (1989) also examined the effects of inhalation exposure of diesel exhaust
2 and/or coal dust on tumorigenesis on F344 rats. Groups of 216 male and 72 female rats were
3 exposed to clean air, whole diesel exhaust (2 mg soot/m3), coal dust (2 mg/m3 respirable
4 concentration; 5 to 6 mg/m3 total concentration), or diesel exhaust plus coal dust (1 mg/m3 of
5 each respirable concentration; 3.2 mg/m3 total concentration) for 7 h/day, 5 days/week during
Q daylight hours for up to 24 mo. Groups of 10 or more males were sacrificed at intermediate
7 intervals (3, 6, and 12 mo). The diesel exhaust was produced by a 7.0-L, 4-cycle, water-cooled
8 Caterpillar Model 3304 engine using No. 2 diesel fuel (<0.5% sulfur by mass). The exhaust was
9 passed through a Wagner water scrubber, which lowered the exhaust temperature and quenched
10 engine backfire. The animals were exposed in 100-cubic-foot chambers. Temperature was
11 controlled at 22±2 °C and relative humidity at 50±10%. The exhaust was diluted 27-fold with
12 chemically and biologically filtered clean air to achieve the desired particle concentration. An
13 analysis of the exposure atmospheres is presented in Appendix A.
14 Histological examination was performed on 120 to 121 male and 71 to 72 female rats
15 terminated after 24 mo of exposure. The exhaust exposure did not significantly affect the tumor
16 incidence beyond what would be expected for aging F344 rats. There was no postexposure
17 period, which may explain, in part, the lack of significant tumor induction. The paniculate
18 matter concentration was also less than the effective dose hi several other studies.
19 In a more recent study reported by Heinrich et al. (1995), female Wistar rats were
20 exposed to whole diesel exhaust (0.8, 2.5, or 7.0 mg/m3) 18 h/day, 5 days/week for up to 24 mo,
21 then held in clean air an additional 6 mo. The animals were exposed hi either 6 or 12 m3
22 exposure chambers. Temperature and relative humidity were mainlined at 23-25 °C and 50%-
23 70%, repectively. Diesel exhaust was generated by two 40-kw 1.6-L diesel engines
24 (Volkswagen). One of them was operated according to the U.S. 72 cycle. The other was operated
25 under constant load conditions. The first engine did not supply sufficient exhaust, which was
26 filled by the second engine. Cumulative exposures for the rats hi the various treatment groups
27 were 61.7,21.8, and 7.4 g/m3 * h for the high, medium, and low whole-exhaust exposures.
28 Significant increases hi tumor incidences were observed hi the high (22/100; /KO.OO 1) and mid
29 (11/200; p<0.01) exposure groups relative to clean-air controls (Table 7-4). Only one tumor
30 (i/217;, an adenocarcinoma, was observed in clean-air controls. Relative to clean-air controls,
31 significantly increased incidences were observed hi the high-exposure rats for benign squamous
32 cell tumors (14/100;/K0.001), adenomas (4/100;/K0.01), and adenocarcinomas (5/100;/K0.05).
33 Only the incidence of benign squamous cell tumors (7/200; /KO.01) was significantly increased
34 in the mid-exposure group relative to the clean-air controls.
35 Particle lung burden and alveolar clearance also were determined in the Heinrich et al.
36 (1995) study. Relative to clean air controls, alveolar clearance was significantly compromised b>
11/5/99 7-100 DRAFT—DO NOT CITE OR QUOTE
-------
1 exposure to mid and high diesel exhaust. For the high-diesel-exhaust group, 3-mo recovery time
2 in clean air failed to reverse the compromised alveolar clearance.
2^ In a study conducted at the Inhalation Toxicology Research Institute (Nikula et al., 1995)
4 F344 rats (114-115 per sex per group) were exposed 16 hr/day, 5 days/week during daylight
5 hours to diesel exhaust diluted to achieve particle concentrations of 2.5 or 6.5 mg/m3 for up to 24
6 mo. Controls (118 males, 114 females) were exposed to clean air. Surviving rats were
7 maintained an additional 6 weeks in clean air, at which time mortality reached 90%. Diesel
8 exhaust was generated using two 1988 Model LH6 General Motors 6.2-L V-8 engines burning D-
9 2 fuel that met EPA certification standards. Chamber air flow was sufficient to provide about 15
10 exchanges per hour. Relative humidity was 40% to 70% and temperature ranged from 23 to 25
11 °C.
12 Following low and high diesel exhaust exposure, the lung burdens were 36.7 and 80.7
13 mg, respectively, for females and 45.1 and 90.1 mg, respectively, for males. The percentages of
14 susceptible rats (males and females combined) with malignant neoplasms were 0.9 (control), 3.3
15 (low diesel exhaust), and 12.3 (high diesel exhaust). The percentages of rats (males and females
16 combined) with malignant or benign neoplasms were 1.4 (control), 6.2 (low diesel exhaust), and
17 17.9 (high diesel exhaust). All primary neoplasms were associated with the parenchyma rather
18 than the conducting airways of the lungs. The first lung neoplasm was observed at 15 mo.
Among 212 males and females examined in the high-dose group, adenomas were detected in 23
animals, adenocarcinomas in 22 animals, squamous cell carcinomas in 3 animals, and an
21 adenosquamous carcinoma in 1 animal. For further details see Table 7-4. Analysis of the
22 histopathologic data suggested a progressive process from alveolar epithelial hyperplasia to
23 adenomas and adenocarcinomas.
24 Iwai et al. (1997) carried out a series of exposures to both filtered and whole exhaust
25 using a light-duty (2,369 mL) diesel engine. The protocol for engine operation was not stated.
26 Groups of female SPF F344 Fischer rats were exposed for 2 years for 8 hr/day, 7 days/week, 8
27 hr/day, 6 days/week, or 18 hr/day, 3 days/week to either filtered exhaust or exhaust diluted to a
28 particle concentration of 9.4,3.2, and 5.1 mg/m3, respectively. Cumulative exposure (mg/m3 *
29 hrs of exposure) equaled 274.4,153.6, and 258.1 mg/m3. The animals were then held for an
30 additional 6 mo in clean air. Lung tumors were reported in 5/121 (4%) of controls, 4/108 (4%)
31 of those exposed to filtered exhaust, and 50/153 (35%) among those exposed to whole exhaust.
32 Among rats exposed to whole diesel exhaust the following number of tumors were detected; 57
33 adenomas, 24 adenocarcinomas, 2 benign squamous cell tumors, 7 squamous cell carcinomas,
34 and 3 adenosquamous carcinomas. The authors stated that benign squamous cell tumors
35 probably corresponded to squamous cysts in another classification.
1115199 7-101 DRAFT—DO NOT CITE OR QUOTE
-------
1 7.3.1.2. Mouse Studies
2 A series of inhalation studies using strain A mice was conducted by Orthoefer et al.
3 (1981). Strain A mice are usually given a series of intraperitoneal injections with the test agent;
4 they are then sacrificed at about 9 mo and examined for lung tumors. In the present series,
5 inhalation exposure was substituted. Diesel exhaust was provided by one of two Nissan CN6-33
6 diesel engines having a displacement of 3244 cc and run on a Federal Short Cycle. Flow through
7 the exposure chambers was sufficient to provide 15 air changes per hour. Temperature was
8 maintained at 24 °C and relative humidity at 75%. In the first study, groups of 25 male Strong A
9 strain (A/S) mice were exposed to irradiated diesel exhaust (to simulate chemical reactions
10 induced by sunlight) or nonirradiated diesel exhaust (6 mg/m3) for 20 h/day, 7 days/week.
11 Additional groups of 40 Jackson A strain (S/J) mice (20 of each sex) were exposed similarly to
12 either clean air or diesel exhaust, then held in clean air until sacrificed at 9 mo of age. No
13 tumorigenic effects were detected at 9 mo of age. Further studies were conducted in which male
14 A/S mice were exposed 8 hr/day, 7 days/week until sacrifice (approximately 300 at 9 mo of age
15 and approximately 100 at 12 mo of age). With the exception of those treated with urethan, the
16 number of tumors per mouse did not exceed historical control levels in any of the studies.
17 Exposure to diesel exhaust, however, significantly inhibited the tumorigenic effects of the 5-mg
18 urethan treatment. Results are listed in Table 7-4.
19 Kaplan et al. (1982) also reported the effects of diesel exposure in strain A mice. Groups
20 of male strain A/J mice were exposed for 20 h/day, 7 days/week for 90 days and held until 9 mo
21 of age. Experimental conditions are described hi Appendix A. Briefly, the animals were
22 exposed in inhalation chambers to diesel exhaust generated by a 5.7-L Oldsmobile engine
23 operated continuously at 40 mph at DPM concentrations of 0, 0.25, 0.75, or 1.5 mg/m3. Controls
24 were exposed to clean air. Temperture was maintained at 22 ± 2 °C and relative humidity at 50
25 ± 10% within the chambers. Among 458 controls and 485 exposed animals, tumors were
26 detected in 31.4% of those breathing clean air versus 34.2% of those exposed to diesel exhaust.
27 The mean number of tumors per mouse also failed to show significant differences.
28 In a follow-up study, strain A mice were exposed to diesel exhaust for 8 mo (Kaplan et
29 al., 1983; White et al., 1983). After exposure to the highest exhaust concentration (1.5 mg/m3),
30 the percentage of mice with pulmonary adenomas and the mean number of tumors per mouse
31 were significantly less (p<0.05) than those for controls (25.0% vs. 33.5% and 0.30 ± 0.02 [S.E.]
32 vs. 0.42 ± 0.03 [S.E.]) (Table 7-4).
33 Pepelko and Peirano (1983) summarized a series of studies on the health effects of diesel
34 emissions hi mice. Exhaust was provided by two Nissan CN 6-33, 6-cylinder, 3.24-L diesel
35 engines coupled to a Chrysler A-272 automatic transmission and Eaton model 758-DG
36 dynamometer. Details of the exposure atmosphere are presented in Appendix A. Sixty-day pilot
11/5/99 7-102 DRAFT—DO NOT CITE OR QUOTE
-------
studies were conducted at a 1:14 dilution, providing DPM concentrations of 6 mg/m3 The
engines were operated using the Modified California Cycle. These 20-hr/day, 7-days/week pilot
studies using rats, cats, guinea pigs, and mice produced decreases in weight gain and food
4 consumption. Therefore, at the beginning of the long-term studies, exposure tune was reduced to
5 8 h/day, 7 days/week at an exhaust DPM concentration of 6 mg/m3. During the final 12 mo of
6 exposure, however, the DPM concentration was increased to 12 mg/m3. For the chronic studies,
7 the engines were operated using the Federal Short Cycle. Chamber temperature was maintained
8 at 24 °C and relative humidity at 50%. Airflow was sufficient for 15 changes per hour.
9 Pepelko and Peirano (1983) described a two-generation study using Sencar mice exposed
10 to diesel exhaust. Male and female parent-generation mice were exposed to diesel exhaust at a
11 DPM concentration of 6 mg/m3 prior to (from weaning to sexual maturity) and throughout
12 mating. The dams continued exposure through gestation, birth, and weaning. Groups of
13 offspring (130 males and 130 females) were exposed to either diesel exhaust or clean air. The
14 exhaust exposure was increased to a DPM concentration of 12 mg/m3 when the offspring were 12
15 weeks of age and was maintained until termination of the experiment when the mice were 15 mo
16 old.
17 The incidence of pulmonary adenomas (16.3%) was significantly increased in the mice
18 exposed to diesel exhaust compared with 6.3% hi clean-air controls. The incidence in males and
1 ^B females combined was 10.2% in 205 animals examined compared with 5.1% hi 205 clean-air
20 controls. This difference was also significant. The incidence of carcinomas was not affected by
21 exhaust exposure hi either sex. These results provided the earliest evidence for cancer induction
22 following inhalation exposure to diesel exhaust. The increase in the sensitivity of the study,
23 allowing detection of tumors at 15 mo, may have been the result of exposure from conception. It
24 is likely that Sencar mice are sensitive to induction of lung rumors since they are also sensitive to
25 induction of skin tumors. These data are summarized hi Table 7-4.
26 Takemoto et al. (1986) reported the effects of inhaled diesel exhaust (2 to 4 mg/m3,4
27 h/day, 4 days/week, for up to 28 mo) in ICR and C57BL mice exposed from birth. Details of the
28 exposure conditions are presented in Section 7.3.2.1 and Appendix A. All numbers reported are
29 for males and females combined. Four adenomas and 1 adenocarcinoma were detected in 34
30 diesel exhaust-exposed ICR mice autopsied at 13 to 18 mo, compared with 3 adenomas among
31 38 controls. Six adenomas and 3 adenocarcinomas were reported hi 22 diesel-exposed ICR mice
32 autopsied at 19 to 28 mo, compared with 3 adenomas and 1 adenocarcinoma in 22 controls. Four
33 adenomas and 2 adenocarcinomas were detected in 79 C57BL mice autopsied at 13 to 18 mo,
34 compared with none hi 19 unexposed animals. Among males and females autopsied at 19 to 28
mo, 8 adenomas and 3 adenocarcinomas were detected in 71 exposed animals, compared with 1
adenoma among 32 controls. No significant increases hi either adenoma or adenocarcinoma
1175/99 7-103 DRAFT—DO NOT CITE OR QUOTE
-------
1 were reported for either strain of exposed mice. However, the significance of the increase in the
2 combined incidence of adenomas and carcinomas was not evaluated statistically. A statistical
3 analysis by Pott and Heinrich (1990) indicated that the difference in combined benign and
4 malignant tumors between whole diesel exhaust-exposed C57BL/6N mice and corresponding
5 controls was significant aip<05. See Table 7-4 for details of tumor incidence.
6 Heinrich et al. (1986b) and Stober (1986), as part of a larger study, also evaluated the
7 effects of diesel exhaust in mice. Details of the exposure conditions reported by Heinrich et al.
8 (1986a) are given in Section 7.3.1.1 and Appendix A. Following lifetime (19 h/day, 5
9 days/week, for a maximum of 120 weeks) exposure to diesel exhaust diluted to achieve a particle
10 concentration of 4.2 mg/m3, 76 female NMRI mice exhibited a total lung tumor incidence of
11 adenomas and adenocarcinomas combined of 32%. Tumor incidences reported for control mice
12 (n = 84) equaled 11% for adenomas and adenocarcinomas combined. While the incidence of
13 adenomas showed little change, adenocarcinomas increased significantly from 2.4% for controls
14 to 17% for exhaust-exposed mice. In a follow up study, however, Heinrich et al. (1995) reported
15 a lack of tumorigenic response in either female NMRI or C57BL/6N mice exposed 17 h/day, 5
16 days/week for 13.5 to 23 mo to whole diesel exhaust diluted to produce a particle concentration
17 of 4.5 mg/m3. These data are summarized in Table 7-4.
18 The lack of a carcinogenic response in mice was reported by Mauderly et al. (1996). In •
19 this study, groups of 540 to 600 CD-I male and female mice were exposed to whole diesel
20 exhaust (7.1, 3.5, or 0.35 mg DPM/m3) for 7 hr/day, 5 days/week for up to 24 mo. Controls were
21 exposed to filtered air. Diesel exhaust was provided by 5.7-L Oldsmobile V-8 engines operated
22 continuously on the U.S. Federal Test Procedure urban certification cycle. The chambers were
23 maintained at 25-28 °C, relative humidity at 40%-60%, and a flow rate sufficient for 15 air
24 exchanges per hour. Animals were exposed during the light cycle, which ran from 6:00 AM to
25 6:00 PM. DPM accumulation in the lungs of exposed mice was assessed at 6,12, and 18 mo of
26 exposure and was shown to be progressive; DPM burdens were 0.2 ± 0.02,3.7 ± 0.16, and 5.6 ±
27 0.39 mg for the low-, medium-, and high-exposure groups, respectively. The lung burdens in
28 both the medium- and high-exposure groups exceeded that predicted by exposure concentration
29 ratio for the low-exposure group. Contrary to what was observed in rats (Keinrich et al., 1986b;
30 Stober, 1986; Nikula et al., 1995; Mauderly et ai., 1987), an exposure-related increase in primary
31 lung neoplasms was not observed in the CD-1 mice, supporting the contention of a species
32 difference in the pulmonary carcinogenic response to poorly soluble particles. The percentage
33 incidence of mice (males and femaies combined) with one or more malignant or benign •
34 neoplasms was 13.4. 14.6. 9.7, and 7.5 for controls and low-, medium-, and high-exposure
35 frroiim resnectivelv.
W X •" A V
1175/99 7-104 DRAFT—DO NOT CITE OR QUOTE
-------
1 While earlier studies provided some evidence for tumorigenic responses in diesel-
2 exposed mice, no increases were reported in the two most recent studies by Mauderly et al.
^fc (1996) and Heinrich et al. (1995), which utilized large group sizes and were well designed and
4 conducted. Overall, the results in mice must therefore be considered to be equivocal.
5
6 7.3.1.3. Hamster Studies
7 Heinrich et al. (1982) examined the effects of diesel exhaust exposure on tumor
8 frequency in female Syrian golden hamsters. Groups of 48 to 72 animals were exposed to clean
9 air or whole diesel exhaust at a mean DPM concentration of 3.9 mg/m3. Inhalation exposures
10 were conducted 7 to 8 hr/day, 5 days/week for 2 years. The exhaust was produced by a 2.4-L
11 Daimler-Benz engine operated under a constant load and a constant speed of 2,400 rpm. Flow
12 rate was sufficient for about 20 exchanges per hour in the 250-L chambers. No lung tumors were
13 reported in either exposure group.
14 In a subsequent study, Syrian hamsters were exposed 19 hr/day, 5 days/week for a
15 lifetime to diesel exhaust diluted to a DPM concentration of 4.24 mg/m3 (Heinrich et al., 1986b;
16 Stober, 1986). Details of the exposure conditions are reported in Appendix A. Ninety-six
17 animals per group were exposed to clean air or exhaust. No lung tumors were seen in either the
18 clean-air group or in the diesel exhaust-exposed group.
In a third study (Heinrich et al., 1989b), hamsters were exposed to exhaust from a
Daimler-Benz 2.4-L engine operated at a constant load of about 15 kW and at a uniform speed of
21 2,000 rpm. The exhaust was diluted to an exhaust-clean air ratio of about 1:13, resulting in a
22 mean particle concentration of 3.75 mg/m3. Exposures were conducted in chambers maintained
23 at 22 to 24 °C and 40% to 60% relative humidity for up to 18 mo. Surviving hamsters were
24 maintained in clean air for up to an additional 6 mo. The animals were exposed 19 hr/day, 5
25 days/week beginning at noon each day, under a 12-hr light cycle starting at 7 AM. Forty animals
26 per group were exposed to whole diesel exhaust or clean air. No lung tumors were detected in
27 either the clean-air or diesel-exposed hamsters.
28 Brightwell et al. (1986,1989) studied the effects of diesel exhaust on male and female
29 Syrian golden hamsters. Groups of 52 males and 52 females, 6 to 8 weeks old, were exposed to
30 diesel exhaust at DPM concentrations of 0.7,2.2, or 6.6 mg/m3. They were exposed 16 hr/day, 5
31 days/week for a total of 2 years and then sacrificed. Exposure conditions are described in Section
32 7.3.1.1 and in Appendix A. No statistically significant (t test) relationship between tumor
33 incidence and exhaust exposure was reported.
34 In summary, diesel exhaust alone did not induce an increase in lung tumors in hamsters of
35^ either sex in several studies of chronic duration at high exposure concentrations.
11/5/99 7-105 DRAFT—DO NOT CITE OR QUOTE
-------
1 7.3.1.4. Monkey Studies
2 . Fifteen male cynomolgus monkeys were exposed to diesel exhaust (2 mg/m3) for 7
3 hr/day, 5 days/week for 24 mo (Lewis et al., 1989). The same numbers of animals were also
4 exposed to coal dust (2 mg/m3 respirable concentration; 5 to 6 mg/m3 total concentration), diesel
5 exhaust plus coal dust (1 mg/m3 respirable concentration for each component; 3.2 mg/m3 total
6 concentration), or filtered air. Details of exposure conditions were listed previously in the
7 description of the Lewis et al. (1989) study with rats (Section 7.3.1.1) and are listed in Appendix
8 A.
9 None of the monkeys exposed to diesel exhaust exhibited a significantly increased
10 incidence of preneoplastic or neoplastic lesions. It should be noted, however, that the 24-mo
11 timeframe employed in this study may not have allowed the manifestation of tumors hi primates,
12 because this duration is only a small fraction of the monkeys' expected lifespan. In fact, there
13 have been no near-lifetime exposure studies hi nonrodent species.
14
15 7.3.2. Inhalation Studies (Filtered Diesel Exhaust)
16 Several studies have been conducted in which animals were exposed to diesel exhaust
17 filtered to remove PM. Since these studies also included groups exposed to whole exhaust,
18 details can be found in Sections 7.3.1.1 for rats, 7.3.1.2 for mice, and 7.3.1.3 for hamsters, and hi
19 Appendix A. Heinrich et al. (1986b) and Stober (1986) reported negative results for lung tumor
20 induction in female Wistar rats exposed to filtered exhaust diluted to produce an unfiltered
21 particle concentration of 4.24 mg/m3. Negative results were also reported hi female Fischer 344
22 rats exposed to filtered exhaust diluted to produce an unfiltered particle concentration of 4.9
23 mg/m3 (Iwai et al., 1986), hi Fischer 344 rats of either sex exposed to filtered exhaust diluted to
24 produce an unfiltered particle concentration of 6.6 mg/m3 (Brightwell et al., 1989), in female
25 Wistar rats exposed to filtered exhaust diluted to produce an unfiltered particle concentration of
26 7.0 mg/m3 (Heinrich et al., 1995), and hi female Fischer 344 rats exposed to filtered exhaust
27 diluted to produce unfiltered particle concentrations of 5.1, 3.2, or 9.4 mg/m3 (Iwai et al., 1997).
28 In the Iwai et al. (1986) study, splenic lymphomas were detected in 37.5% of the exposed rats
29 compared with 8.2% in controls.
30 In the study reported by Heinrich at al. (1986a) and Stober (1986), primary lung tumors
31 were seen in 29/93 NMRI mice (males and females combined) exposed to filtered exhaust,
32 compared with 11/84 in clean-air controls, a statistically significant increase. In a repeat study by
33 Heinrich et al. (1995), however, significant lung tumor increases were not detected hi either
34 female NMRI or C57BL/6N mice exposed to filtered exhaust diluted to produce an unfiltered
3 5 particle concentration of 4.5 mg/m3.
11 /S/99 7-106 DRAFT—DO NOT CITE OR QUOTE
-------
1 Filtered exhaust also failed to induce lung tumor induction in Syrian Golden hamsters
2 (Heinrich et al., 1986a; Brightwell et al., 1989).
^ft Although lung tumor increases were reported in one study and lymphomas in another,
4 these results could not be confirmed in subsequent investigations. It is therefore concluded that
5 little direct evidence exists for carcinogenicity of the vapor phase of diesel exhaust in laboratory
6 animals at concentrations tested.
7 '
8 7.3.3. Inhalation Studies (Diesel Exhaust Plus Co-Carcinogens)
9 Details of the studies reported here have been described earlier and in Table 7-4. Tumor
10 initiation with urethan (1 mg/kg body weight i.p. at the start of exposure) or promotion with
11 butylated hydroxytolulene (300 mg/kg body weight i.p. week 1, 83 mg/kg week 2, and 150 mg/kg
12 for weeks 3-52) did not influence tumorigenic responses in Sencar mice of both sexes exposed to
13 concentrations of diesel exhaust up to 12 mg/m3 (Pepelko and Peirano, 1983).
14 Heinrich et al. (1986b) exposed Syrian hamsters of both sexes to diesel exhaust diluted to
15 a particle concentration of 4 mg/m3. See Section 7.3.1.1 for details of the exposure conditions.
16 At the start of exposure the hamsters received either one dose of 4.5 mg diethyhiitrosamine
17 (DEN) subcutaneously per kg body weight or 20 weekly intratracheal instillations of 250 ug BaP.
18 Female NMRI mice received weekly intratracheal instillations of 50 or 100 figBaP for 10 or 20
1Jfe weeks, respectively, or 50 ug dibenz[ah]anthracene (DBA) for 10 weeks. Additional groups of
"2u 96 newborn mice received one s.c. injection of 5 or 10 ug DBA between 24 and 48 hr after birth.
21 Female Wistar rats received weekly subcutaneous injections of dipentylnitrosamine (DPN) at
22 doses of 500 and 250 mg/kg body weight, respectively, .during the first 25 weeks of exhaust
23 inhalation exposure. Neither DEN, DBA, or DPN treatment enhanced any tumorigenic responses
24 to diesel exhaust. Although response to BaP did not differ from that of BaP alone in hamsters,
25 results were inconsistent in mice. Although 20 BaP instillations induced a 71% tumor incidence
26 in mice, concomitant diesel exposure resulted in only a 41% incidence. However, neither 10 BaP
27 instillations nor DBA instillations induced significant effects.
•
28 Takemoto et al. (1986) exposed Fischer 344 rats for 2 years to diesel exhaust at particle
29 concentrations of 2 to 4 mg/m3. One month after start of inhalation exposure one group of rats
30 received di-isopropyl-nitrosamine (DIPN) administered i.p. at 1 mg/kg weekly for 3 weeks.
31 Among injected animals autopsied at 18 to 24 mo, 10 adenomas and 4 adenocarcinomas were
32 reported in 21 animals exposed to clean air, compared with 12 adenomas and 7 adenocarcinomas
33 hi 18 diesel-exposed rats. According to the authors, the incidence of adenocarcinomas was not
34 significantly increased by exposure to diesel exhaust.
Brightwell et al. (1989) investigated the concomitant effects of diesel exhaust and DEN in
Syrian hamsters exposed to diesel exhaust diluted to produce particle concentrations of 0.7,2.2,
11/5/99 7-107 DRAFT—DO NOT CITE OR QUOTE
-------
1 or 6.6 mg/m3 for 2 years. The animals received a single dose of 4.5 mg DEN s.c. 3 days prior to
2 start of inhalation exposure. DEN did not affect the lack of responsiveness to diesel exhaust
3 alone. Heinrich et al. (1989b) also exposed Syrian hamsters of both sexes to diesel exhaust
4 diluted to a particle concentration of 3.75 mg/m3 for up to 18 mo. After 2 weeks of exposure,
5 groups were treated with either 3 or 6 mg DEN/kg body weight, respectively. Again, DEN did
6 not significantly influence the lack of tumorigenic responses to diesel exhaust.
7 Heinrich et al. (1989a) investigated the effects of DPN hi female Wistar rats exposed to
8 diesel exhaust diluted to achieve a particle concentration of 4.24 mg/m3 for 2-2.5 years. DPN at
9 doses of 250 and 500 mg/kg body weight was injected subcutaneously once a week for the first
10 25 weeks of exposure. The tumorigenic responses to DPN were not affected by exposure to
11 diesel exhaust. For details of exposure conditions of the hamster studies see Section 7.3.1.3.
12 Heinrich et al. (1986a) and Mohr et al. (1986) compared the effects of exposure to
13 particles having only a minimal carbon core but a much greater concentration of PAHs than
14 DPM does. The desired exposure conditions were achieved by mixing coal oven flue gas with
15 pyrolyzed pitch. The concentration of B[a]P and other PAHs per milligram of DPM was about
16 three orders of magnitude greater than that of diesel exhaust. Female rats were exposed to the
17 flue gas-pyrolyzed pitch for 16 hr/day, 5 days/week at particle concentrations of 3 to 7 mg/m3 for
18 22 mo, then held in clean air for up to an additional 12 mo. Among 116 animals exposed, 22
19 tumors were reported hi 21 animals, for an incidence of 18.1%. One was a bronchioloalveolar
20 adenoma, one was a bronchioloalveolar carcinoma, and 20 were squamous cell tumors. Among
21 the latter, 16 were classified as benign keratinizing cystic tumors and 4 were classified as
22 carcinomas. No tumors were reported in 115 controls. The tumor incidence hi this study was
23 comparable to that reported previously for the diesel exhaust-exposed animals.
24 In analyzing the studies of Heinrich et al. (1986a,b), Heinrich (1990), Mohr et al. (1986),
25 and Sto'ber (1986), it must be noted that the incidence of lung tumors occurring following
26 exposure to whole diesel exhaust, coal oven flue gas, or carbon black (15.8%, 18.1%, and 8% to
27 17%, respectively) was very similar. This occurred despite the fact that the PAH content of the
28 PAH-enriched pyrolyzed pitch was more than three orders of magnitude greater than that of
29 diesei exhaust; carbon black, on the other hand, had only traces of PAHs. Based on these
30 foldings, particle-associated, effects appear to be the primary cause of uiesel-cxliaust-iiiduccd
31 lung cancer in rats exposed at high concentrations. This issue is discussed further in Chapter 7.
32
33 73.4. Lung Implantation or Intratracheal Instillation Studies
34 73.4.1= Rat Studies
35 Grimmer et al. (1987), using female Osbome Mendel rats (35 per treatment group),
36 provided evidence that PAHs in diesel exhaust that consist of tour or more rings have
1175/99 7-108 DRAFT—DO NOT CITE OR QUOTE
-------
1 carcinogenic potential. Condensate was obtained from the whole exhaust of a 3.0-L passenger-
2 car diesel engine connected to a dynamometer operated under simulated city traffic driving
^P conditions. This condensate was separated by liquid-liquid distribution into hydrophilic and
4 hydrophobic fractions representing 25% and 75% of the total condensate, respectively. The
5 hydrophilic, hydrophobic, or reconstituted hydrophobic fractions were surgically implanted into
6 the lungs of the rats. Untreated controls, vehicle (beeswax/trioctanoin) controls, and positive
7 (B[a]P) controls were also included in the protocol (Table 7-5). Fraction lib (made up of PAHs
8 with four to seven rings), which accounted for only 0.8% of the total weight of DPM condensate,
9 produced the highest incidence of carcinomas following implantation into rat lungs. A
10 carcinoma incidence of 17.1% was observed following implantation of 0.21 mg lib/rat, whereas
11 the nitro-PAH fraction (lid) at 0.18 mg/rat accounted for only a 2.8% carcinoma incidence.
12 Hydrophilic fractions of the DPM extracts, vehicle (beeswax/trioctanoin) controls, and untreated
13 controls failed to exhibit carcinoma formation. Administration of all hydrophobic fractions (Ila-
14 d) produced a carcinoma incidence (20%) similar to the summed incidence of fraction lib
15 (17.1%) and lid (2.8%). The B[a]P positive controls (0.03, 0.1,0.3 mg/rat) yielded a carcinoma
16 incidence of 8.6%, 31.4%, and 77.1 %, respectively. The study showed that the tumorigenic
17 agents were primarily four- to seven-ring PAHs and, to a lesser extent, nitroaromatics. However,
18 these studies demonstrated that simultaneous administration of various PAH compounds resulted
^fc in a varying of the tumorigenic effect, thereby implying that the tumorigenic potency of PAH
20 mixtures may not depend on any one individual PAH. This study did not provide any
21 information regarding the bioavailability of the particle-associated PAHs that might be
22 responsible for carcinogenicity.
23 Kawabata et al. (1986) compared the effects of activated carbon and diesel exhaust on
24 lung tumor formation. One group of 59 F344 rats was intratracheally instilled with DPM (1
25 mg/week for 10 weeks). A second group of 31 rats was instilled with activated carbon using the °
26 same dosing regime. Twenty-seven rats received only the solvent (buffered saline with 0.05%
27 Tween 80), and 53 rats were uninjected. Rats dying after 18 mo were autopsied. All animals
28 surviving 30 mo or more postinstillation were sacrificed and evaluated for histopathology.
29 Among 42 animals exposed to DPM surviving 18 mo or more, tumors were reported in 31,
30 including 20 malignancies. In the subgroup surviving for 30 mo, tumors were detected in 19 of
31 20 animals, including 10 malignancies. Among the rats exposed to activated carbon, the
32 incidence of lung tumors equaled 11 of 23 autopsied, with 7 cases of malignancy. Data for those
33 dying between 18 and 30 mo and those sacrificed at 30 mo were not reported separately.
11/5/99 7-109 DRAFT—DO NOT CITE OR QUOTE
-------
Ol
s
o
z:
3
o
t— I
a
o
50
Table 7-5. Tumor incidence and survival time of rats treated by surgical lung implantation with
fractions from diesel exhaust condensate (35 rats/group)
Material portion by weight (%)
Hydrophilic fraction (I) (25)
Hydrophobia fraction (II) (75)
Nonaromatics +
PACC 2 + 3 rings (Ha) (72)
PAH" 4 to 7 rings (lib) (0.8)
Polar PAC (He) (1.1)
Nitro-PAH (Hd) (0.7)
Reconstituted hydrophobics
(la, b, c, d) (74.5)
Control, unrelated
Control (beeswax/trioctanoin)
Benzo[
-------
1 Statistical analysis indicated that activated carbon induced a significant increase in lung tumor
2 incidence compared with no tumors in 50 uninjected controls and 1 tumor in 23 solvent-injected
^fc controls. The tumor incidence was significantly greater in the DPM-instilled group and was
4 significantly greater than the increase in the carbon-instilled group.
5 A study reported by Rittinghausen et al. (1997) suggested that organic constituents of
6 diesel particles play a role in the induction of lung tumors in rats. An incidence of 16.7%
7 pulmonary cystic keratinizing squamous cell lesions was noted in rats intratracheally instilled
8 with 15 mg whole diesel exhaust particles, compared with 2.1% in rats instilled with 15 mg
9 particles extracted to remove all organic constituents, and none among controls. Instillation of
10 30 mg of extracted particles induced a 14.6% incidence of squamous lesions, indicating the
11 greater effectiveness of particles alone as lung particle overload increased.
12 Iwai et al. (1997) instilled 2,4, 8, and 10 mg of whole diesel particles over a 2 to 10 week
13 period into female F/344 rats, 50 or more per group. Tumors were reported in 6%, 20%, 43%,
14 and 74% of the rats, with incidence of malignant tumors equal to 2%, 13%, 34%, and 48%,
15 respectively. In a second experiment comparing whole with extracted diesel particles, tumor
16 incidence equaled 1/48 (2%) in uninjected controls, 3/55 (5%) in solvent controls, 12/56 (21%)
17 in extracted diesel particles, and 13/106 (12%) in animals injected with unextracted particles.
18 Although the extracted particles appeared to be more potent, when converted to a lung burden
basis (mg/100 mg dry lung) the incidence was only 14% among those exposed to extracted
exhaust compared with 31% in those exposed to whole particles.
21 Dasenbrock et al. (1996) conducted a study to determine the relative importance of the
22 organic constituents of diesel particles and particle surface area in the induction of lung cancer in
23 rats. Fifty-two female Wistar rats were intratracheally instilled with 16-17 doses of DPM,
24 extracted DPM, printex carbon black (PR), lampblack (LB), benzo|>]pyrene (BaP), DPM + BaP,
25 or PR + BflP. The animals were held for a lifetime or sacrificed when moribund. The lungs were
26 necropsied and examined for tumors. Diesel particles were collected from a Volkswagen 1.6-L
27 engine operating on a US FTP-72 driving cycle. The mass median aerodynamic diameter
28 (MMAD) of the diesel particles was 0.25 um and the specific surface area was 12 m2/gm.
29 Following extraction with toluene, specific surface area increased to 138 m2/gm. The MMAD
30 for extracted PR was equal to 14 nm, while the specific surface area equaled 271 m2/gm. The
31 MMAD for extracted lampblack was equal to 95 nm, with a specific surface area equal to 20
32 mVgrn. The BaP content of the treated particles was 11.3 mg per gm diesel particles and 29.5
33 mg BaP per gm PR. Significant increases in lung tumors were detected in rats instilled with 15
34 mg unextracted DPM and 30 mg extracted DPM, but not 15 mg extracted DPM. Printex CB was
35
11/5/99 7-111 DRAFT—DO NOT CITE OR QUOTE
-------
1 more potent than lampblack CB for induction of lung tumors, while BaP was effective only at
2 high doses. Total dose and tumor responses are shown in Table 7-6.
3 A number of conclusions can be drawn from these results. First of all, particles devoid of
4 organics are capable of inducing lung tumor formation, as indicated by positive results in the
5 groups treated with high-dose extracted diesel particles and printex. Nevertheless, toluene
6 extraction of organics from diesel particles results in a decrease in potency, indicating that the
7 organic fraction does play a role in cancer induction. A relationship between cancer potency and
8 particle surface area was also suggested by the finding that printex with a large specific surface
9 area was more potent than either extracted DPM or lampblack, which have smaller specific areas.
10 Finally, while very large doses of BaP are very effective in the induction of lung tumors, smaller
11 doses adsorbed to particle surfaces had little detectable effect, suggesting that other organic
12 components of diesel exhaust may be of greater importance in the induction of lung tumors at
13 low doses pf BaP (0.2-0.4 mg).
14
15 7.3.4.2. Syrian Hamster Studies
16 Kunitake et al. (1986) and Ishinishi et al. (1988b) conducted a study in which total doses
17 of 1.5, 7.5, or 15 mg of a dichloromethane extract of DPM were instilled intratracheally over 15
18 weeks into male Syrian hamsters that were then held for their lifetimes. The tumor incidences of
19 2.3% (1744), 0% (0/56), and 1.7% (1/59) for the high-, medium-, and low-dose groups,
20 respectively, did not differ significantly from the 1.7% (1/56) reported for controls. Addition of
21 7.5 mg of B[a]P to a DPM extract dose of 1.5 mg resulted in a total tumor incidence of 91.2%
22 and malignant tumor incidence of 88%. B[a]P (7.5 mg-over 15 weeks) alone produced a tumor
23 incidence rate of 88.2% (85% of these being malignant), which was not significantly different
24 from the DPM extract + B[a]P group. Intratracheal administration of 0.03 ug B[a]P, the
25 equivalent content in 15 mg of DPM extract, failed to cause a significant increase in tumors in
26 rats. This study demonstrated a lack of detectable interaction between DPM extract and B[a]P,
27 the failure of DPM extract to induce carcinogenesis, and the propensity for respiratory tract
28 carcinogenesis following intratracheal instillation of high doses of B[a]P. For studies using the
29 DPM extract, some concern must be registered regarding the known differences in chemical
30 composition between DPM extract and DPM. As with all intratracheal instillation protocols,
31 DPM extract lacks the complement of volatile chemicals found in whole diesel exhaust.
32 The effects on hamsters of intratracheally instilled DPM suspension, DPM with Fe,O3, or
33 DPM extract with Fe^ as the carrier were studied by Shefrier et al. (1982). The DPM
34 component in each of the treatments was administered at concentrations of 1.25, 2.5, or 5.0
11/5/99 7-112 DRAFT—DO NOT CITE OR QUOTE
-------
Table 7-6. Tumor incidences in rats following intratracheal instillation of diesel
exhaust particles (DPM), extracted DPM, carbon black (CB), benzo[a]pyrene
(BaP), or particles plus BaP
Experimental
group
Control
DPM (original)
DPM (extracted)
DPM (extracted)
CB (printex)
CB (lampblack)
BaP
BaP
DEP + BaP
CB (printex) + BaP
Number
of
animals
47
48
48
48
48
48
47 .
48
48
48
Total dose
4.5 mL
15 mg
30 mg
15 mg
15 mg
14 mg
30 mg
15 mg
15mg+ 170 ng
BaP
15mg-f-443 ug
BaP
Animals with
tumors
(percent)
0 (0)
8 (17)
10 (21)
2 (4)
10 (21)
4 (8)
43 (90)
12 (25)
4 (8)
13 (27)
Statistical
significance'
-
<0.01
< 0.001
NS
< 0.001
NS
< 0.001
< 0.001
NS
< 0.001
"Fischer's exact test.
Source: Dasenbrock et al., 1996.
1 mg/week for 15 weeks to groups of 50 male Syrian golden hamsters. The total volume instilled
2 was 3.0 mL (0.2 mL/week for 15 weeks). The DPM and dichloromethane extracts were
3 suspended hi physiological saline with gelatin (0.5% w/v), gum arabic (0.5% w/v), and propylene
4 glycol (10% by volume). The FejC^ concentration, when used, was 1.25 mg/0.2 mL of
5 suspension. Controls received vehicle and, where appropriate, carrier particles (Fe2O3) without
6 the DPM component. Two replicates of the experiments were performed. Adenomatous
7 hyperplasia was reported to be most severe hi those animals treated with DPM or DPM plus
8 Fe2O3 particles and least severe in those animals receiving DPM plus FejOj. Of the two lung
9 adenomas detected microscopically, one was hi an animal treated with a high dose of DPM and
10 the other was hi an animal receiving a high dose of DPM extract. Although lung damage was
11 increased by instillation of DPM, there was no evidence of tumorigenicity.
1;
11/5/99
7-113 DRAFT—DO NOT CITE OR QUOTE
-------
1 7.3.4.3. Mouse Studies
2 Ichinose et al. (1997a) intratracheally instilled 36 four-week-old male ICR mice per group
3 weekly for 10 weeks with sterile saline or 0.05, 0.1, or 0.2 mg DPM. Particles were collected
4 from a 2.74-L four-cylinder Isuzu engine run at a steady speed of 1,500 rpm under a load of 10
5 torque (kg/m). Twenty-four hours after the last instillation, six animals per group were sacrificed
6 for measurement of lung 8-hydroxydeoxyguanosine (8-OHdG). The remaining animals were
7 sacrificed after 12 mo for histopathological analysis. Lung tumor incidence varied from 4/30
8 (13.3%) for controls to 9/30 (30%), 9/29 (31%), and 7/29 (24.1%) for mice instilled with 0.05,
9 0.1, and 0.2 mg/week, respectively. The increase in animals with lung tumors compared with
10 controls was statistically significant for the 0.1 mg dose group, the only group analyzed
11 statistically. Increases in 8-OHdG, an indicator of oxidative DNA damage, correlated well with
12 the increase in tumor incidence in the 0.05 mg dose group, although less so with the other two.
13 The correlation coefficients r = 0.916, 0.765, and 0.677 for the 0.05, 0.10, and 0.20 mg DPM
14 groups, respectively.
15 In a similar study, 33 four-week-old male ICR mice per group were intratracheally
16 instilled weekly for 10 weeks with sterile saline, 0.1 mg DPM, or 0.1 mg DPM from which the
17 organic constituents were extracted with hexane (Ichinose et al., 1997b). Exhaust was collected
18 from a 2.74-L four-cylinder Izuzu engine run at a steady speed of 2,000 rpm under a load of 6
19 torque (kg/m). Twenty-four hours after the last instillation, six animals per group were sacrificed
20 for measurement of 8-OHdG. Surviving animals were sacrificed after 12 mo. The incidence of
21 lung tumors increased from 3/27 (11.1%) among controls to 7/27 (25.9%) among those instilled
22 with extracted diesel particles and 9/26. (34.6%) among those instilled with unextracted particles.
23 The increase in number of tumor-bearing animals was statistically significant compared with
24 controls (p<0.05) for the group treated with unextracted particles. The increase in 8-OHdG was
25 highly correlated with lung rumor incidence, r = 0.99.
26
27 7.3.5. Subcutaneous and Intraperitoneal Injection Studies
28 7.3.5.1. Mouse Studies
29 In addition to inhalation studies, Oitlioefer et al. (1981) also tested the effects of i.p.
30 injections of DPM on male (A/'S) strain mice. Three groups of 30 mice were injected with 0.1
31 mL of a suspension (particles in distilled water) containing 47,117, or 235 ug of DPM collected
32 from Fluoropore filters in the inhalation exposure chambers. The exposure system and exposure
33 atmosphere are described in Appendix A. Vehicle controls received injections cf particle
34 suspension made up of particulate matter from control exposure filters, positive controls received
35 20 mg of urethan, and negative controls received no injections. Injections were made three times
36 weekly for 8 weeks, resulting in a total DPM dose of 1.1, 2.'8, and 5.6 mg for the low-, medium-,
11/5/99 7-114 DRAFT—DO NOT CITE OR QUOTE
-------
1 and high-dose groups and 20 mg of urethan for the positive control group. These animals were
2 sacrificed after 26 weeks and examined for lung tumors. For the low-, medium-, and high-dose
^^ DPM groups, the tumor incidence was 2/30, 10/30, and 8/30, respectively. The incidence among
4 urethan-treated animals (positive controls) was 100% (29/29), with multiple tumors per animal.
5 The tumor incidence for the DPM-treated animals did not differ significantly from that of vehicle
6 controls (8/30) or negative controls (7/28). The number of tumors per mouse was also unaffected
7 by treatment.
8 In further studies conducted by Orthoefer et al. (1981), an attempt was made to compare
9 the potency of DPM with that of other environmental pollutants. Male and female Strain A mice
10 were injected i.p. three times weekly for 8 weeks with DPM, DPM extracts, or various
11 environmental mixtures of known carcinogenicity, including cigarette smoke condensate, coke
12 oven emissions, and roofing tar emissions. Injection of urethan or dimethylsulfoxide (DMSO)
13 served as positive or vehicle controls, respectively. In addition to DPM from the Nissan diesel
14 previously described, an eight-cylinder Oldsmobile engine operated at the equivalent of 40 mph
15 was also used to compare emission effects from different makes and models of diesel engine.
16 The mice were sacrificed at 9 mo of age and their lungs examined for histopathological changes.
17 The only significant findings, other than for positive controls, were small increases in numbers of
18 lung adenomas per mouse in male mice injected with Nissan DPM and in female mice injected
with coke oven extract. Furthermore, the increase in the extract-treated mice was significant only
in comparison with uninjected controls (not injected ones) and did not occur when the
21 experiment was repeated. Despite the use of a strain of mouse known to be sensitive to tumor
22 induction, the overall findings of this study were negative. The authors provided several possible
23 explanations for these findings, the most likely of which were (1) the carcinogens that were
24 present were very weak, or (2) the concentrations of the active components reaching the lungs
25 were insufficient to produce positive results.
26 Kunitake et al. (1986) conducted studies using DPM extract obtained from a 1983 HD
27 MMC—6D22P 11-L V-6 engine. Five s.c. injections of DPM extract (500 mg/kg per injection)
28 resulted in a significant (p<0.01) increase in subcutaneous tumors for female C57BL mice (5/22
29 [22.7%] vs. 0/38 among controls). Five s.c. doses of DPM extract of 10,25, 30, 100, or 200
30 mg/kg failed to produce a significant increase in tumor incidence. One of 12 female ICR mice
31 (8.3%) and 4 of 12 male ICR mice (33.3%) developed malignant lymphomas following neonatal
32 s.c. administration of 10 mg of DPM extract per mouse. The increase in malignant lymphoma
33 incidence for the male mice was statistically significant at/K0.05 compared with an incidence of
34 2/14 (14.3%) among controls. Treatment of either sex with 2.5 or 5 mg of DPM extract per
35 mouse did not result in statistically significant increases in tumor incidence.
1115/99 7-115 DRAFT—DO NOT CITE OR QUOTE
-------
1 Additional studies using DPM extract from LD (1.8-L, 4-cylinder) as well as HD engines
2 with female ICR and nude mice (BALB/c/cA/JCL-nu) were also reported (Kunitake et al., 1988).
3 Groups of 30 ICR and nude mice each were given a single s.c. injection of 10 mg HD extract, 10
4 mg HD + 50 ug 12-O-tetradecanoylphorbol 13-acetate (TPA), 10 mg LD extract + 50 fag TPA, or
5 50 ug TPA. No malignant tumors or papillomas were observed. One papillomatous lesion was
6 observed in an ICR mouse receiving LD extract + TPA, and acanthosis was observed in one nude
7 mouse receiving only TPA.
8 In what appears to be an extension of the Kunitake et al. (1986) s.c. injection studies,
9 Takemoto et al. (1988) presented additional data for subcutaneously administered DPM extract
10 from HD and LD diesel engines. In this report, the extracts were administered to 5-week-old and
11 neonatal (<24 hr old) C57BL mice of both sexes. DPM extract from HD or LD engines was
12 administered weekly to the 5-week-old mice for 5 weeks at doses of 10,25, 50,100,200, or 500
13 mg/kg, with group sizes ranging from 15 to 54 animals. After 20 weeks, comparison with a
14 control group indicated a significant increase in the incidence of subcutaneous tumors for the 500
15 mg/kg HD group (5 of 22 mice [22.7%],/K0.01), the 100 mg/kg LD group (6 of 32 [18.8%],
16 /K0.01), and the 500 mg/kg LD group (7 of 32 [21.9%],/K0.01) in the adult mouse experiments.
17 The tumors were characterized as malignant fibrous histiocytomas. No tumors were observed in
18 other organs. Theneonates were given single doses of 2.5, 5, or 10 mg DPM extract
19 subcutaneously within 24 hr of birth. There was a significantly higher incidence of malignant
20 lymphomas in males receiving 10 mg of HD extract and of lung tumors for males given 2.5 mg
21 HD extract and for males given 5 mg and females given 10 mg LD extract. A dose-related trend
22 that was not significant was observed for the incidences of liver tumors for both the HD extract-
23 and LD extract-treated neonatal mice. The incidence of mammary tumors in female mice and
24 multiple-organ tumors in male mice was also greater for some extract-treated mice, but was not
25 dose related. The report concluded that LD DPM extract showed greater carcinogenicity than did
26 HD DPM extract
27
28 7.3.6. Dermal Studies
29 7.3.6.1. Mouse Studies
30 In one of the earliest studies of diesel emissions, the effects of dermal application oi
31 extract from DPM were examined by Kotin et al. (1955). Acetone extracts were prepared from
32 the DPM of a diesel engine (type and size not provided) operated at warmup mode and under
33 load. These extiacts were applied dennally three times weekly to male and female C57BL and
34 strain A mice. Results of these experiments are summarized in Table 7-7. In the initial
11/5/99 7-116 DRAFT—DO NOT CITE OR QUOTE
-------
Table 7-7. Tumorigenic effects of dermal application of acetone extracts of diesel particulate
matter (DPM)
\
Number of
animals
52
50
25
Strain/sex
C57BL/40 F
C57BL/12 M
Strain A/M
Strain A/F
Sample material
Extract of DPM obtained during
warmup
Extract of DPM obtained during
full load
Extract of DPM obtained during
full load
Time to first
tumor (mo)
13
15
13
Survivors at
time of first
tumor Total tumors
33 2
8 4
20 17
Duration of
experiment
(mo)
22
23
17
Source: Kotinetal. (1955).
-------
1 experiments using 52 (12 male, 40 female) C57BL mice treated with DPM extract from an
2 engine operated in warmup mode, two papillomas were detected after 13 mo. Four tumors were
3 detected 16 mo after the start of treatment in 8 surviving of 50 exposed male strain A mice
4 treated with DPM extract from an engine operated under full load. Among female strain A mice
5 treated with DPM extract from an engine operated under full load, 17 tumors were detected in 20
6 of 25 mice surviving longer than 13 mo. This provided a significantly increased tumor incidence
7 of 85%. Carcinomas as well as papillomas were seen, but the numbers were not reported.
8 Depass et al. (1982) examined the potential of DPM and dichloromethane extracts of
9 DPM to act as complete carcinogens, carcinogen initiators, or carcinogen promoters. In skin-
10 painting studies, the DPM was obtained from an Oldsmobile 5.7-L diesel engine operated under
11 constant load at 65 km/h. The DPM was collected at a temperature of 100 °C. Groups of 40
12 C3H/HeJ mice were used because of their low spontaneous tumor incidence. For the complete
13 carcinogenesis experiments, DPM was applied as a 5% or 10% suspension in acetone.
14 Dichloromethane extract was applied as 5%, 10%, 25%, or 50% suspensions. Negative controls
15 received acetone, and positive controls received 0.2% B[a]P. For tumor-promotion experiments,
16 a single application of 1.5% B[a]P was followed by repeated applications of 10% DPM
17 suspension, 50% DPM extract, acetone only (vehicle control), 0.0001% phorbol 12-myristate 13-
18 acetate (PMA) as a positive promoter control, or no treatment (negative control). For the tumor-
19 initiation studies, a single initiating dose of 10% diesel particle suspension, 50% diesel particle
20 extract, acetone, or PMA was followed by repeated applications of 0.0001% PMA. Following 8
21 mo of treatment, the PMA dose in the initiation and promotion studies was increased to 0.01%.
22 Animals were treated three times per week hi the complete carcinogenesis and initiation
23 experiments and five times per week in promotion experiments. All test compounds were
24 applied to a shaved area on the back of the mouse.
25 In the complete carcinogenesis experiments, one mouse receiving the high-dose (50%)
26 suspension of extract developed a squamous cell carcinoma after 714 days of treatment. Tumor
27 incidence in the B[a]P group was 100%, and no tumors were observed hi any of the other groups.
28 For the promotion studies, squamous cell carcinomas with pulmonary metastases were identified
29 in one mouse of the 50% DPM extract group and in one in the 25% extract group. Another
30 mouse in the 25% extract gioup developed a grossly diagnosed papilloina. Nineteen positive
31 control mice had tumors (11 papillomas, 8 carcinomas). No tumors were observed for any of the
32 other treatment groups. For the initiation studies, three tumors (two papillomas and one
33 carcinoma) were identified hi the group receiving DPM suspension and tbree tumors (two
34 papillomas and one fibrosarcoma) were found hi the DPM extract group. These findings were
35 reported to be statistically insignificant using the Breslow and Mantel-Cox tests.
11 /5/99 7-118 DRAFT—DO NOT CITE OR QUOTE
-------
1 Although these findings were not consistent with those of Kotin et al. (1955), the
2 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
5 assumptions.
6 Nesnow et al. (1982) studied the formation of dermal papillomas and carcinomas
7 following dermal application of dichloromethane extracts from coke oven emissions, roofing tar,
8 DPM, and gasoline engine exhaust. DPM from five different engines, including a preproduction
9 Nissan 220C, a 5.7-L Oldsmobile, a prototype Volkswagen Turbo Rabbit, a Mercedes 300D, and
10 a HD Caterpillar 3304, was used for various phases of the study. Male and female Sencar mice
11 (40 per group) were used for tumor initiation, tumor promotion, and complete carcinogenesis
12 studies. For the tumor-initiation experiments, the DPM extracts were topically applied in single
13 doses of 100, 500, 1,000, or 2,000 ng/mouse. The high dose (10,000 ug/mouse) was applied in
14 five daily doses of 2,000 ug. One week later, 2 ug of the tumor promoter TPA was applied
15 topically twice weekly. The tumor-promotion experiments used mice treated with 50.5 ug of
16 B[a]P followed by weekly (twice weekly for high dose) topical applications (at the
17 aforementioned doses) of the extracts. For the complete carcinogenesis experiments, the test
18 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.
21 In the tumor-initiation studies, both B[a]P alone and the Nissan engine DPM extract
22 followed by TPA treatment produced a significant increase in tumor (dermal papillomas)
23 incidence at 7 to 8 weeks postapplication. By 15 weeks, the tumor incidence was greater than
24 90% for both groups. No significant carcinoma formation was noted for mice in the tumor-
25 initiation experiments following exposure to DPM extracts of the other diesel engines, although
26 the Oldsmobile engine DPM extract at 2.0 mg/mouse did produce a 40% papilloma incidence in
27 male mice at 6 mo. This effect, however, was not dose dependent.
28 B[a]P (50.5 ug/week), coke oven extract (at 1.0,2.0, or 4.0 mg/week), and the highest
29 dose of roofing tar extract (4.0 mg/week) all tested positive for complete carcinogenesis activity.
30 DPM extracts from only the Nissan, Oldsmobile, and Caterpillar engines were tested for
31 complete carcinogenic potential, and all three proved to be negative using the Sencar mouse
32 assay.
33 The results of the dermal application experiments by Nesnow et al. (1982) are presented
34 in Table 7-8. The tumor initiation-promotion assay was considered positive if a dose-dependent
35 response was obtained and if at least two doses provided a papilloma-per-mouse value that was
three times or greater than that of the background value. Based on these criteria, only emissions
11/5/99 7-119 DRAFT—DO NOT CITE OR QUOTE
-------
Table 7-8. Dermal tumorigenic and carcinogenic effects of various emission extracts
-J
^^
K)
O
O
O
2
O
H
O
o
&
o
c
1
Sample
Benzo[#]pyrene
Topside coke oven
Coke oven main
Roofing tar
Nissan
Oldsmobile
VW Rabbit
Mercedes
Caterpillar
Residential furnace
Mustang
Tumor initiation Complete carcinogenesis
Papillomas' Carcinomas'* Carcinomas'*
+/+• +/+ ' +/+
+/+ -/+ NDd
+/+ +/+ +/+
+/+ +/+ +/+
+/+ +/+ -/-
+/+ -/- -/-
+/+ -/- r
+/- -/- ND
-/- . -/- -/-
-/- -/- ND
+/+ -/+ ND
Tumor promotion
Papillomas'
+/+
ND
+/+
+/ +
ND
ND
ND
ND
ND
ND
ND
'Scored at 6 mo.
bCumulative score at 1 year.
'Male/female.
dND = Not determined.
el •-- Incomplete.
Source: Nesnow et al. (1982).
-------
1 from the Nissan were considered positive. Tumor initiation and complete carcinogenesis assays
2 required that at least one dose produce a tumor incidence of at least 20%. None of the DPM
^P samples yielded positive results based on this criterion.
4 Kunitake et al. (1986, 1988) evaluated the effects of a dichloromethane extract of DPM
5 obtained from a 1983 MMC M-6D22P 11-L V-6 engine. An acetone solution was applied in 10
6 doses every other day, followed by promotion with 2.5 ug of TPA three times weekly for 25
7 weeks. Exposure groups received a total dose of 0.5, 5, 15, or 45 mg of extract. Papillomas
8 were reported in 2 of 50 animals examined in the 45 mg exposure group and in 1 of 48 in the 15
9 mg group compared with 0 of 50 among Controls. Differences, however, were not statistically
10 significant.
11
12 7.3.7. Summary and Conclusions of Laboratory Animal Carcinogenicity Studies
13 As early as 1955, Kotin et al. (1955) provided evidence for tumorigenicity and
14 carcinogenicity of acetone extracts of DPM following dermal application and also provided data
15 suggesting a difference in this potential depending on engine operating mode. Until the early
16 1980s, no chronic studies assessing inhalation of diesel exhaust, the relevant mode for human
17 exposure, had been reported. Since then long-term inhalation bioassays with diesel exhaust have
18 been carried out in the United States, Germany, Switzerland, and Japan, testing responses of rats,
^fc mice, and Syrian hamsters, and to a limited extent cats and monkeys.
20 It can be reasonably concluded that with adequate exposure, inhalation of diesel exhaust
21 is capable of inducing lung cancer in rats. Responses best fit cumulative exposure (concentration
22 * daily exposure duration * days of exposure). Examination of rat data shown in Table 7-9
23 indicates a trend of increasing tumor incidence at exposures exceeding 1 * 104 mg-hr/m3.
24 Exposures greater than approximately this value result in lung particle overload, characterized by
25 slowed particle clearance and lung pathology, as discussed in Chapters 3 and 5, respectively.
26 Tumor induction at high doses may therefore be primarily the result of lung particle overload
27 with associated inflammatory responses. Although tumorigenic responses could not be detected
28 under non-particle-overload conditions, the animal experiments lack sensitivity to determine if a
29 threshold exists. If low-dose effects do occur, it can be hypothesized that the organic
30 constituents are playing a role. See Chapter 7 for a discussion of this issue.
31 While rats develop adenomas, adenocarcinomas, and adenosquamous cell carcinomas,
32 they also develop squamous keratinizing lesions. This latter spectrum appears for the most part
33 to be peculiar to the rat. In a recent workshop aimed at classifying these tumors (Boorman et al.,
34 1996), it was concluded that when these lesions occur in rats as part of a carcinogenicity study,
11/5/99 7-121 DRAFT—DO NOT CITE OR QUOTE
-------
Table 7-9. Cumulative (concentration x time) exposure data for rats exposed to whole diesel exhaust
D;
V^5
^J
h— »
to
to
O
j>
^
0
O
H
O
(— i
a
o
o
c
o
H
W
Study
Mauderly et al.
(1987)
Nikulaetal. (1995)
Heinrich et al.
(1986a)
Heinrich et al.
(1995)
Ishinishi et al.
(I988a)
(Light-duty engine)
(Heavy-duty engine)
•
Exposure
rate/duration
(hr/week, mo)
35.30
35,30
35,30
35,30
80,23
80,23
80,23
95,35
95,35
90,24
90,24
90,24
90,24
96,30
96,30
96,30
96,30
96,30
96,30
96,30
96,30
96,30
96,30
Total exposure
time (hr)
4,200
4,200
4,200
4,200
7,360
7,360
7,360
13,300
13,300
8,640
8,640
8.640
8.640
11,520
11,520
11,520
11,520
11,520
11,520
11,520
11,520
11,520
11,520
Particle
concentration
(mg/m3)
0
0.35
3.5
7.1
0
2.5
6.5
0
4.24
0
0.8
2.5
7.0
0
0.1
0.4
1.1
2.3
0
0.5
1.0
1.8
3.7
Cumulative exposure
(mghr/m3)
Per week
0
12.25
122.5
248.5
0
200.0
520.0
0
402.8
0
72.0
225.0
630.0
0
9.6
38.4
105.6
220.8
0
48.0
96.0
172.8
355.2
Total
0
1,470
14,700
29,820
0
18,400
47,840
0
56,392
0
7,400
21,800
61,700
0
1,152
4,608
12,672
26,496
0
5,760
11,520
20,736
42,624
Tumor incidence
(%)"
0.9
1.3
3.6
12.8
1.0
7.0
18.0
0
17.8
0
0
5.5
22.0
3.3
2.4
0.8
4.1
2.4
0.8
0.8
0
3.3
6.5
7-122
-------
Table 7-9. Cumulative (concentration x time) exposure data for rats exposed to whole diesel exhaust (continued)
t— *
^g Exposure
rate/duration
Study (hr/week, mo)
Tj
to
O
j>
r
i
o
2!
3
O
a
o
/o
c
o
Brightwell et al.
(1989)
Kaplan etal. (1983)
Iwaietal. (1986)
Takemoto et al.
(1986)
' Karagianes et al.
(1981)
Iwaietal. (1997)
"Combined data for males
80,24
80,24
80,24
80,24
140, 15
140, 15
140, 15
140, 15
56,24
56,24
16, 18-24
16, 18-24
30,20
30,20
56,24
48,24
54,24
and females.
Total exposure
time (hr)
7,680
7,680
7,680
7,680
8,400
8,400
8,400
8,400
5,376
5,376
1,152-1,536
1,152-1,536
2,400
2,400
5,376
4,992
5,616
Particle
concentration
(mg/m3)
0
0.7
2.2
6.6
0
0.25
0.75
1.5
0
4.9
0
2-4
0
8.3
9.4
3.2
5.1
Cumulative exposure
(mg-hr/m3)
Per week
0
56.0
176.0
528.0
0
35.6
105.0
210.0
0
274.4
0
32-64
0
249
526
154
275
Total
0
5,376
16,896
50,688
0
2,100
6,300
12,600
0
26,342
0
3,456-4,608
0
19,920
54,704
15,974
28,642
Tumor incidence
(%)'
1.2
0.7
9.7
38.5
0
3.3
10.0
3.3
0
36.8
0
0
0
16.6
42
12
42
-------
1 they must be evaluated on a case-by-case basis and regarded as a part of the total biologic profile
2 of the test article. If the only evidence of tumorigenicity is the presence of cystic keratinizing
3 epitheliomas, it may not have relevance to human safety evaluation of a substance or particle.
4 Their use in quantifying cancer potency is even more questionable.
5 The evidence for response of common strains of laboratory mice exposed under standard
6 inhalation protocols is equivocal. Inhalation of diesel exhaust induced significant increases in
7 lung tumors in female NMRJ mice (Heinrich et al., 1986b; Stober, 1986) and in female Sencar
8 mice (Pepelko and Peirano, 1983). An apparent increase was also seen hi female C57BL mice
9 (Takemoto et al., 1986). However, hi a repeat of their earlier study, Heinrich et al. (1995) failed
10 to detect lung tumor induction in either NMRI or C57BL/6N mice. No increases in lung tumor
11 rates were reported in a series of inhalation studies using strain A mice (Orthoefer et al., 1981;
12 Kaplan et al., 1982; Kaplan et al., 1983; White et al., 1983). Finally, Mauderly et al. (1996)
13 reported no tumorigenic responses in CD-I mice exposed under conditions resulting in positive '
14 responses in rats. The successful induction of lung tumors in mice by Ichinose et al. (1997a,b)
15 via intratracheal instillation may have been the result of focal deposition of larger doses. Positive
16 effects in Sencar mice may be due to use of a strain sensitive to tumor induction in epidermal
17 tissue by organic agents, as well as exposure from conception, although proof for such a
18 hypothesis is lacking.
19 Attempts to induce significant increases hi lung tumors hi Syrian hamsters by inhalation
20 of whole diesel exhaust were unsuccessful (Heinrich et al., 1982, 1986b, 1989b; Brightwell et al.,
21 1986). However, hamsters are considered to be relatively insensitive to lung tumor induction.
22 For example, while cigarette smoke, a known human carcinogen, was shown to induce laryngeal
23 cancer in hamsters, the lungs were relatively unaffected (Dontenwill et al., 1973).
24 Neither cats (Pepelko and Peirano, 1983 [see Chapter 7]) nor monkeys (Lewis et al.,
25 1986) developed tumors following 2-year exposure to diesel exhaust. The duration of these
26 exposures, however, was likely to be inadequate for these two longer-lived species, and group
27 sizes were quite small. Exposure levels were also below the maximum tolerated dose (IvlTD) in
29 Long-term exposure to diesel exhaust filtered to remove particulate matter failed to
30 induce lung tumors hi rats (Heinrich et al., 1986b; Iwai et al., 1986; Brightwell et al., 1989), or in
31 Syrian hamsters (Heinrich et al., 1986b; Brightwell, 1989). A significant increase hi lung
32 carcinomas was reported by Heinrich et al. (1986b) hi NMRI mice exposed to filtered exhaust.
33 However, hi a more recent study the authors were unable to confirm earlier results in either
34- NMJKl or C57BL/6N mice (Heinrich et al., 1995). Although filtered exhaust appeared to
35 potentiate the carcinogenic effects of DEN (Heinrich et al., 1982), nevertheless, because of the
11/5/99 7-124 DRAFT—DO NOT CITE OR QUOTE
-------
1 lack of positive data in rats and equivocal or negative data in mice, it can be concluded that
f filtered exhaust is either not carcinogenic or has a low cancer potency.
Kawabata et al. (1986) demonstrated the induction of lung tumors in Fischer 344 rats
4 following intratracheal instillation of DPM. Rittinghausen et al. (1997) reported an increase in
5 cystic keratinizing epitheliomas following intratracheal instillation of rats with either original
6 DPM or DPM extracted to remove the organic fraction, with the unextracted particles inducing a
7 slightly greater effect. Grimmer et al. (1987) showed not only that an extract of DPM was
8 carcinogenic when instilled in the lungs of rats, but also that most of the carcinogenicity resided
9 in the portion containing PAHs with four to seven rings. Intratracheal instillation did not induce
10 lung tumors in Syrian hamsters (Kunitake et al., 1986; Ishinishi et al., 1988b).
11 Dermal exposure and s.c. injection in mice provided additional evidence for tumorigenic
12 effects of DPM. Particle extracts applied dermally to mice have been shown to induce
13 significant skin tumor increases in two studies (Kotin et al., 1955; Nesnow et al., 1982).
14 Kunitake et al. (1986) also reported a marginally significant increase in skin papillomas in ICR
15 mice treated with an organic extract from an HD diesel engine. Negative results were reported
16 by Depass et al. (1982) for skin-painting studies using mice and acetone extracts of DPM
17 suspensions. However, in this study the exhaust particles were collected at temperatures of
P100 °C, which would minimize the condensation of vapor-phase organics and, therefore, reduce
the availability of potentially carcinogenic compounds that might normally be present on diesel
20 exhaust particles. A significant increase in the incidence of sarcomas in female C57B1 mice was
21 reported by Kunitake et al. (1986) following s.c. admini'stration of LD DPM extract at doses of
22 500 mg/kg. Takemoto et al. (1988) provided additional data for this study and reported an
23 increased tumor incidence in the mice folio whig injection of LD engine DPM extract at doses of
24 100 and 500 mg/kg. Results of i.p. injection of DPM or DPM extracts in strain A mice were
25 generally negative (Orthoefer et al., 1981; Pepelko and Peirano, 1983), suggesting that the strain
26 A mouse may not be a good model for testing diesel emissions.
27 Results of experiments using tumor initiators such as DEN, B[a]P, DPN, or DBA
28 (Brightwell et al., 1986; Heinrich et al., 1986b; Takemoto et al., 1986) were generally
29 inconclusive regarding the tumor-promoting potential of either filtered or whole diesel exhaust.
30 A report by Heinrich et al. (1982), however, indicated that filtered exhaust may promote the
31 tumor-initiating effects of DEN in hamsters.
32 Several reports (Wong et al., 1986; Bond et al., 1990) affirm observations of the potential
.33 carcinogenicity of diesel exhaust by providing evidence for DNA damage in rats. These findings
are discussed in more detail in Section 3.6. Evidence for the mutagenicity of organic agents
present in diesel engine emissions is also provided in Chapter 4.
11/5/99 7-125 DRAFT—DO NOT CITE OR QUOTE
-------
1 Evidence for the importance of the carbon core was initially provided by studies of
2 Kawabata et al. (1986), which showed induction of lung tumors following intratracheal
3 instillation of carbon black that contained no more than traces of organics, and studies of
4 Heinrich (1990) that indicated that exposure via inhalation to carbon black (Printex 90) particles
5 induced lung tumors at concentrations similar to those effective in DPM studies. Additional
6 studies by Heinrich et al. (1995) and Nikula et al. (1995) confirmed the capability of carbon
7 particles to induce lung tumors. Induction of lung tumors by other particles of low solubility
8 such as titanium dioxide (Lee et al., 1986) confirmed the capability of particles to induce lung
9 rumors. Pyrolyzed pitch, on the other hand, essentially lacking a carbon core but having much
1 0 higher PAH concentrations than DPM, also was effective in tumor induction (Heinrich et al.,
11 1986a, 1994).
1 2 The relative importance of the adsorbed organics, however, remains to be elucidated and
13 is of some concern because of the known carcinogenic capacity of some of these chemicals.
1 4 These include polycyclic aromatics as well as nitroaromatics, as described in Chapter 2. Organic
1 5 extracts of particles also have been shown to induce tumors in a variety of injection, intratracheal
1 6 instillation, and skin-painting studies, and Grimmer et al. (1987) have, in fact, shown that the
1 7 great majority of the carcinogenic potential following instillation resided in the fraction
1 8 containing four- to seven-ring PAHs.
19 In summary, based on positive inhalation and intratracheal instillation data in rats and on
20 i.p. injection or skin painting in mice, and supported by positive mutagenicity studies, the
21 evidence for carcinogenicity of diesel exhaust is considered to be adequate. The contribution of
22 the various fractions of diesel exhaust to the carcinogenic response is less certain. Exposure to
23 filtered exhaust generally failed to induce lung tumors. The presence of known carcinogens
24 adsorbed to diesel particles and the demonstrated tumorigenicity of particle extracts hi a variety
25 of injection, instillation, and skin-painting studies indicate a carcinogenic potential for the
26 organic fraction. Studies showing that insoluble particles (e.g., carbon black, TiO2) can also
27 induce tumors, on tuC other hanu, nave proviuCu uCjufiitiYc cviucncc that tuc caruon core 01
29 sufficient exposure conditions. The ability of diesel exhaust to induce lung tumors at non-
30 particle-overload conditions, and the relative contribution of the particles' core versus the
3 1 particle-associated organics (if effects do occur at low doses) remains to be determined.
32
33 7.4. MODE OF ACTION OF DIESEL EMISSION-INDUCED CARCINOGENESIS
34 As noied in Chapter 2, diesei exhaust (DIE) is a complex mixture that includes a vapor
35 phase and a particle phase. The particle phase consists of an insoluble carbon core with a large
1 1 75/99 7-126 DRAFT— DO NOT CITE OR QUOTE
-------
number of organic compounds, as well as inorganic compounds such as sulfates, adsorbed to the
particle surface. Some of the semivolatile and particle-associated compounds, in particular
3 PAHs, nitro-PAHs, oxy-PAHs, and oxy-nitro-PAHs (Scheepers and Bos, 1992), are considered
4 likely to be carcinogenic in humans. The vapor phase also contains a large number of organic
5 compounds, including several known or probable carcinogens such as benzene and 1,3-
6 butadiene. Since exposure to the vapor phase alone, even at high concentrations, failed to induce
7 lung cancer in laboratory animals (Heinrich et al., 1986), discussion will focus on the particulate
8 matter phase. Additive or synergistic effects of vapor-phase components, however, cannot be
9 totally discounted, since chronic inhalation bioassays involving exposure to diesel particles alone
10 have not been carried out.
11 Several hypotheses regarding the primary mode of action of diesel exhaust have been
12 proposed. Initially it was generally believed that cancer was induced by particle-associated
13 organics acting via a genotoxic mechanism. By the late 1980s, however, studies indicated that
14 carbon particles virtually devoid of organics could also induce lung cancer at sufficient inhaled
15 concentrations (Heinrich, 1990). This finding provided support for a hypothesis originally
16 proposed by Vostal (1986) that induction of lung tumors arising in rats exposed to high
17 concentrations of diesel exhaust is related to overloading of normal lung clearance mechanisms,
1^) accumulation of particles, and cell damage followed by regenerative cell proliferation. The
19 action of particles is therefore mediated by epigenetic mechanisms that can be characterized
20 more by promotional than initiation stages of the carcinogenic process. More recently several
21 studies have focused upon the production of reactive oxygen species generated from particle-
22 associated organics, which may induce oxidative DNA damage at exposure concentrations lower
23 than those required to produce lung particle overload. Since it is likely that more than one of
24 these factors is involved in the carcinogenic process, a key consideration is their likely relative
25 contribution at different exposure levels. The following discussion will therefore consider the
26 possible relationship of the organic components of exhaust, inflammatory responses associated
27 with lung particle overload, reactive oxygen species, and physical characteristics of diesel
28 particles to cancer induction, followed by a hypothesized mode of action, taking into account the
29 likely contribution of the factors discussed.
30
31 7.4.1. Potential Role of Organic Exhaust Components in Lung Cancer Induction
32 More than 100 carcinogenic or potentially carcinogenic components have been
33 specifically identified in diesel emissions, including various PAHs and nitroarenes such as
PI-nitropyrene (1-NP) and dinitropyrenes (DNPs). The majority of these compounds are adsorbed
to the carbon core of the particulate phase of the exhaust and, if desorbed, may become available
1175/99 7-127 DRAFT—DO NOT CITE OR QUOTE
-------
1 for biological processes such as metabolic activation to mutagens. Among such compounds
2 identified from diesel exhaust are benzo(a)pyrene (B[a]P), dibenz[a,/z]anthracene, pyrene,
3 chrysene, and nitroarenes such as 1-NP, 1,3-DNP, 1,6-DNP, and 1,8-DNP, all of which are
4 mutagenic, carcinogenic, or implicated as procarcinogens or cocarcinogens (Stenback et al.,
5 1976; Weinstein and Troll, 1977; Thyssen et al., 1981; Pott and Stober, 1983; Howard et al.,
6 1983; Hirose et al., 1984; Nesnow et al., 1984; El-Bayoumy et al., 1988). More recently Enya et
7 al. (1997) reported isolation of 3-nitrobenzanthrone, one of the most powerful direct-acting
8 mutagens known to date, from the organic extracts of diesel exhaust.
9 Grimmer et al. (1987) separated diesel exhaust particle extract into a water- and a lipid-
10 soluble fraction, and the latter was further separated into a PAH-free, a PAH-containing, and a
11 polar fraction by column chromatography. These fractions were then tested in Osborne-Mendel
12 rats by pulmonary implantation at doses corresponding to the composition of the original diesel
13 exhaust. The water-soluble fraction did not induce tumors; the incidences induced by the lipid-
14 soluble fractions were 0% with the PAH-free fraction, 25% with the PAH and nitro-PAH-
15 containing fractions, and 0% with the polar fraction. The PAH and nitro-PAH-containing
16 fraction, comprising only 1% by weight of the total extract, was thus shown to be responsible for
17 most, if not all, of the carcinogenic activity.
18. Exposure of rats by inhalation to 2.6 mg/m3 of an aerosol of tar-pitch condensate with no
19 carbon core but containing 50 (ig/m3 benzo[a]pyrene along with other PAHs for 10 months
20 induced lung tumors in 39% of the animals. The same amount of tar-pitch vapor condensed onto
21 the surface of carbon black particles at 2 and 6 mg/m3 resulted in tumor rates that were roughly
22 two times higher (89% and 72%). Since exposure to 6 mg/m3 carbon black almost devoid of
23 extractable organic material induced a lung tumor rate of 18%, the combination of PAHs and
24 particles increases their effectiveness (Heinrich et al., 1994). While this study shows the tumor-
25 inducing capability of PAHs resulting from combustion, it should be noted that the
26 benzo[a]pyrene content in the coal-tar pitch was about three orders of magnitude greater than in
27 diesel soot, moreover, because organics are present on diesel particles in a thinner layer and the
?8 psrticles ara quits convoluted, they may be mere tightly Ix/uiul ami less bluavailable.
29 Nevertheless, these studies provide evidence supporting the involvement of organic constituents
30 of diesel particles in the carcinogenic process.
31 Exposure of humans to related combustion emissions provides some evidence for the
32 involvement of organic components. Mumford et al. (1989) reported greatly increased human
33 lung cancer mortality in Chinese communes burning so-called smoky coal, but not wood, in
34 uiivcmed open-pit fires used for heating and cooleing. Although particle concentrations were
35 similar, PAH levels were five to six times greater in the air of communes burning smoky coal.
11/5/99 7-128 DRAFT—DO NOT CITE OR QUOTE
-------
Coke oven emissions, containing high concentrations of PAHs but lacking an insoluble carbon
core, have also been shown to be carcinogenic in humans (Lloyd, 1971).
3 Adsorption of PAHs to a carrier particle such as hematite, CB, aluminum, or titanium
4 dioxide enhances their carcinogenic potency (Farrell and Davis, 1974). As already noted,
5 adsorption to carbon particles greatly enhanced the tumorigenicity of pyrolyzed pitch condensate
6 containing B[a]P and other aromatic carcinogens (Heinrich et al., 1995). The increased
7 effectiveness can be partly explained by more efficient transport to the deep lung. Slow release
8 also enhances residence time in the lungs and prevents overwhelming of activating pathways. As
9 discussed in Chapter 3, free organics are likely to be rapidly absorbed into the bloodstream,
10 which may explain why the vapor-phase component of exhaust is relatively ineffective in the
11 induction of pathologic or carcinogenic effects.
12 Even though the organic constituents may be tightly bound to the particle surface,
13 significant elution is still likely because particle clearance half-times are nearly 1 year in humans
14 (Bohning et al., 1982). Furthermore, Gerde et al. (1991) presented a model demonstrating that
15 large aggregates of inert dust containing crystalline PAHs are unlikely to form at doses typical of
16 human exposure. This allows the particles to deposit and react with the surrounding lung
17 medium, without interference from other particles. Particle-associated PAHs can then be
expected to be released more rapidly from the particles. Bond et al. (1984) provided evidence
that alveolar macrophages from beagle dogs metabolized B[a]P coated on diesel particles to
20 proximate carcinogenic forms. Unless present on the particle surface, B[a]P is more likely to
21 pass directly into the bloodstream and escape activation by phagocytic cells.
22 The importance of DE-associated PAHs in the induction of lung cancer in humans may
23 be enhanced because of the possibility that the human lung is more sensitive to these compounds
24 than are rat lungs. Rosenkranz( 1996) summarized information indicating that in humans and
25 mice, large proportions of lung cancers contain both mutated p53 suppressor genes and K-ray
26 genes. Induction of mutations in these genes by genotoxins, however, is much lower in rats than
27 in humans or mice.
28 B[a]P, although only one of many PAHs present in diesel exhaust, is the one most
29 extensively studied. Bond et al. (1983,1984) demonstrated metabolism of particle-associated
30 B[a]P and free B[a]P by alveolar macrophages (AM) and by type II alveolar cells. The
31 respiratory tract cytochrome P-450 systems have an even greater concentration in the nonciliated
32 bronchiolar cells (Boyd, 1984). It is worthy to note that bronchiolar adenomas that develop
33 following diesel exposure have been found to resemble both Type II and nonciliated bronchiolar
34 cells. It should also be noted that any metabolism of procarcinogens by these latter two cell types
probably involves the preextraction of carcinogens in the extracellular lining fluid and/or other
11/5/99 7-129 DRAFT—DO NOT CITE OR QUOTE
-------
1 endocytotic cells, since they are not especially important in phagocytosis of particles. Thus,
2 bioavailability is an important issue in assessing the relative importance of PAHs.
3 Additionally, a report by Borrn et al. (1997) indicates that incubating rat lung epithelial-
4 derived cells with human PMNs (either unactivated or activated by preexposure to phorbol
5 myristate acetate) increases DNA adduct formation caused by exposure to benzo[a] pyrene;
6 addition of more activated PMN in relation to the number of lung cells further increased adduct
7 formation in a dose-dependent manner. The authors suggest that "an inflammatory response in
8 the lung may increase the biologically effective dose of poly cyclic aromatic hydrocarbons
9 (PAHs), and may be relevant to data interpretation and risk assessment of PAH-containing
10 particles." These data raise the possibility that DE exposure at low concentrations may result in
11 levels of neutrophil influx that would not necessarily be detectable via histopathological
12 examination as acute inflammation, but which might be effective at amplifying any potential
13 diesel exhaust genotoxic effect.
14 Nitro-PAHs have also been implicated as potentially involved in diesel-exhaust-induced
15 lung cancer. Although the nitro-PAH fraction of diesel was less effective than PAHs in the
16 induction of lung cancer when implanted into the lungs of rats (Grimmer et al., 1987), in a study
17 of various extracts of diesel exhaust particles, 30%-40% of the total mutageniciry could be
18 attributed to a group of six nitroarenes (Salmeen et al., 1984). Moreover, Gallagher et al. (1994)
19 reported results suggesting that DNA adducts are formed from nitro-PAHs present in DNA and
20 may play a role in the carcinogenic process. Nitroarenes, however, quantitatively represent a
21 very small percentage of diesel particle extract (Grimmer et al., 1987), making their role in the
22 tumorigenic response uncertain.
23 The induction of DNA adducts in humans occupationally exposed to diesel exhaust
24 indicates the likelihood that PAHs are participating in the tumorigenic response, and that these
25 effects can occur at exposure levels less than those required to induce lung particle overload.
26 Distinct adduct patterns were found among garage workers occupationally exposed to diesel
27 exhaust when compared to nonexposed controls (Nielsen and Autrup. 1994). Furthermore, the
28 findings were concordant with the adduct patterns observed in groups exposed to low
29 concentrations of PAHs from combustion processes. Hemminki et al. (1994) also reported
30 significantly elevated levels of DNA adducts in lymphocytes from garage workers with known
31 diesel exhaust exposure compared to unexposed mechanics. Hou et al. (1995) found elevated
32 adduct levels in bus maintenance workers exposed to diesel exhaust. Although no difference in
33 mutant frequency was observed between the groups, the adduct levels were significantly different
34 (3.2 vs. 2.3 x 10-8). Nielsen et al. (1996) measured three biomarkers in DE-exposed bus garage
35 workers: lymphocyte DNA adducts, hydroxyethylvaline adducts in hemoglobin, and
1175/99 7-130 DRAFT—DO NOT CITE OR QUOTE
-------
1-hydroxypyrene in urine. Significantly increased levels were reported for all three. Qu et al.
(1996) detected increased adduct levels, as well as increases in some individual adducts, in the
3 blood of underground coal miners exposed to DE.
4
5 7.4.2. Role of Inflammatory Cytokines and Proteolytic Enzymes in the Induction of Lung
6 Cancer by Diesel Exhaust
7 It is well recognized that the deposition of particles in the lung can result in the efflux of
8 polymorphonuclear leucocytes (PMNs) from the vascular compartment into the alveolar space
9 compartment in addition to expanding the AM population size. Following acute exposures, the
10 influx of the PMNs is transient, lasting only a few days (Adamson and Bowden, 1978; Bowden
11 and Adamson, 1978; Lehnert et al., 1988). During chronic exposure the numbers of PMNs
12 lavaged from the lungs of diesel-exposed rats generally increased with increasing exposure
13 duration and inhaled diesel particulate matter (DPM) concentration (Strom, 1984). Strom (1984)
14 also found that PMNs in diesel-exposed lungs remained persistently elevated for at least 4
15 months after cessation of exposure, a potential mechanism that may be related to an ongoing
16 release of phagocytized particles. Evidence in support of this possibility was reported by
17 Lehnert et al. (1989) in a study in which rats were intratracheally instilled with 0.85, 1.06, or 3.6
mg of polystyrene particles. The PMNs were not found to be abnormally abundant during the
clearance of the two lower lung burdens, but they became progressively elevated in the lungs of
20 the animals in which alveolar-phase clearance was inhibited. Moreover, the particle burdens in
21 the PMNs became progressively greater over time. Such findings are consistent with an ongoing
22 particle relapse process, in which particles released by dying phagocytes are ingested by new
23 ones.
24 The inflammatory response, characterized by efflux of PMNs from the vascular
25 compartment, is mediated by inflammatory chemokines. Driscoll et al. (1996) reported that
26 inhalation of high concentrations of carbon black stimulated the release of macrophage
27 inflammatory protein 2 (MIP-2) and monocyte chemotactic protein 1 (MCP-1). They also
28 reported a concomitant increase in hprt mutants. In a following study it was shown that particle
29 exposure stimulates production of tumor necrosis factor TNF-cc, an agent capable of activating
30 expression of several proteins that promote both adhesion of leucocytes and chemotaxis (Driscoll
31 et al., 1997). In addition, alveolar macrophages also have the ability to release several other
32 effector molecules or cytokines that can regulate numerous functions of other lung cells,
33 including their rates of proliferation (Bitterman et al., 1983; Jordana et al., 1988; Driscoll et al.,
34 1996).
11/5/99 7-131 DRAFT—DO NOT CITE OR QUOTE
-------
1 Another characteristic of AMs and PMNs under particle overload conditions is the release
2 of a variety of potentially destructive hydrolytic enzymes, a process known to occur
3 simultaneously with the phagocytosis of particles (Sandusky et al., 1977). The essentially
4 continual release of such enzymes during chronic particle deposition and phagocytosis in the lung
5 may be detrimental to the alveolar epithelium, especially to Type I cells. Evans et al. (1986)
6 showed that injury to Type I cells is followed shortly thereafter by a proliferation of Type II cells.
7 Type II cell hyperplasia is a common feature observed hi animals that have received high lung
8 burdens of various types of particles, including unreactive polystyrene microspheres.
9 Exaggerated proliferation as a repair or defensive response to DPM deposition may have the
10 effect of amplifying the likelihood of neoplastic transformation in the presence of carcinogens
11 beyond that which would normally occur with lower rates of proliferation, assuming an increase
12 in the cycling of target cells and the probability of a neoplastic-associated genomic disturbance.
13'
14 7.4.3. Role of Reactive Oxygen Species in Lung Cancer Induction by Diesel Exhaust
15 Phagocytes from a variety of species produce elevated levels of oxidant reactants in
16 response to challenges, with the phy siochemical characteristics of a phagocytized particle being a
17 major factor hi deterrnining the magnitude of the oxidant-producing response. Active oxygen
18 species released by the macrophages and lymphatic cells can cause lipid peroxidation in the
19 membrane of lung epithelial cells. These lipid peroxidation products can initiate a cascade of
20 oxygen free radicals that progress through the cell to the nucleus, where they damage DNA. If
21 this damage occurs during the epithelial cell's period of DNA synthesis, there is some probability
22 that the DNA will be replicated unrepaired (Lechner and Mauderly, 1994). The generation of
23 reactive oxygen species by both AMs and PMNs should therefore be considered as one potential
24 factor of what probably is a multistep process that culminates hi the development of lung tumors
25 in response to chronic deposition of DPM.
26 Even though products of phagocytic oxidative metabolism, including superoxide anions,
27 hydrogen peroxide, and hydroxyl radicals, can kill tumor cells (Klebanoff and Clark, 1978), and
28 the reactive oxygen species can peroxidize lipids to produce cytotoxic metabolites such as
29 malonyldialdehyde, some products of oxidative metabolism apparently can also interact with
30 DNA to produce mutations. Cellular DNA is damaged by oxygen free radicals generated from a
31 variety of sources (Ames, 1983; Trotter, 1980). Along this line, Weitzman and Stossel (1981)
32 found that human peripheral leukocytes are mutagenic in the Ames assay. This mutagenic
33 activity was related to PMNs and blood monocytes; blood lymphocytes alone were not
34 mutagenic. These investigators speculated that the mutagenic activity of the phagocytes was a
35 result of their ability to produce reactive oxygen metabolites, inasmuch as blood leukocytes from
11/5/99 7-132 DRAFT—DO NOT CITE OR QUOTE
-------
a patient with chronic granulomatous diseases, in which neutrophils have a defect in the
NADPH oxidase generating system (Klebanoff and Clark, 1978), were less effective in
3 producing mutations than were normal leukocytes. Of related significance, Phillips et al. (1984)
4 demonstrated that the incubation of Chinese hamster ovary cells with xanthine plus xanthine
5 oxidase (a system for enzymatically generating active oxygen species) resulted hi genetic damage
6 hallmarked by extensive chromosomal breakage and sister chromatid exchange and produced an
7 increase in the frequency of thioguanidine-resistant cells (HGPRT test). Aside from interactions
8 of oxygen species with DNA, increasing evidence also points to an important role of phagocyte-
9 derived oxidants and/or oxidant products in the metabolic activation of procarcinogens to their
10 ultimate carcinogenic form (Kensler et al., 1987).
11 Hatch and co-workers (1980) have demonstrated that interactions of guinea pig AMs with
12 a wide variety of particles, such as silica, metal oxide-coated fly ash, polymethylmethacrylate
13 beads, chrysotile asbestos, fugitive dusts, polybead carboxylate microspheres, glass and latex
14 beads, uncoated fly ash, and fiberglass increase the production of reactive oxygen species.
15 Similar findings have been reported by numerous investigators for human, rabbit, mouse, and
16 guinea pig AMs (Drath and Karnovsky, 1975; Allen and Loose, 1976; Beall et al., 1977; Lowrie
17 and Aber, 1977; Miles et al., 1977; Rister and Baehner, 1977; Hoidal et al., 1978). PMNs are
also known to increase production of superoxide radicals, hydrogen peroxide, and hydroxyl
radicals in response to membrane-reactive agents and particles (Goldstein et al., 1975; Weiss et
20 al., 1978; Root and Metcalf, 1977). Although these responses may occur at any concentration,
21 they are likely to be greatly enhanced at high exposure concentrations with slowed clearance and
22 lung particle overload.
23 Reactive oxygen species can also be generated from particle-associated organics. Sagai et
24 al. (1993) reported that DPM can nonenzymatically generate active oxygen species (such as
25 superoxide [O2'] and hydroxyl radical [OH] hi vitro without any biologically activating systems)
26 such as microsomes, macrophages, hydrogen peroxide, or cysteine. Because DPM washed with
27 methanol could no longer produce these radicals, it was concluded that the active components
28 were compounds extractable with organic solvents. However, the nonen2ymatic contribution to
29 the DPM-promoted active oxygen production was negligible compared to that generated via an
30 enzymatic route (Ichinose et al., 1997a). They reported that O2~ and OH can be enzymatically
31 generated from DPM by the following process. Soot-associated quinone-like compounds are
32 reduced to the semiquinone radical by cytochrome P-450 reductase. These semiquinone radicals
33 then reduce O2 to O2~, and the produced superoxide reduces ferric ions to ferrous ions, which
34 catalyzes the homobiotic cleavage of H2O2 dismutated from O, by superoxide dismutase or
spontaneous reactions to produce OH. According to Kumagai et al. (1997), while quinones are
11/5/99 7-133 DRAFT—DO NOT CITE OR QUOTE
-------
1 likely to be the favored substrates for this reaction, the participation of nitroaromatics cannot be
2 ruled out.
3 One of the critical lesions to DNA bases generated by oxygen free radicals is
4 8-hydroxydeoxyguanosine (8-OHdG). The accumulation of 8-OhdG as a marker of oxidative
5 DNA damage could be an important factor in enhancing the mutation rate leading to lung cancer
6 (Ichinose et al., 1997a). For example, formation of 8-OHdG adducts leads to G:C to T:A
7 transversions unless repaired prior to replication. Nagashima et al. (1995) demonstrated that the
8 production of (8-OHdG) is induced in mouse lungs by intratracheal instillation of DPM.
9 Ichinose et al. (1997b) reported further that while intratracheal instillation of DPM in mice
10 induced a significant increase in lung tumor incidence, comparable increases were not reported
11 when mice were instilled with extracted DPM (to remove organics). Lung injury was also less in
12 the mice instilled with extracted DPM. Moreover, increases in 8-OHdG in the mice instilled
13 with unextracted DPM correlated very well with increases in tumor rates. In a related study,
14 Ichinose et al. (1997a) intratracheally instilled small doses of DPM, 0.05, 0.1, or 0.2 mg weekly
15 for 3 weeks, in mice fed standard or high-fat diets, either with or without p-carotene. High
16 dietary fat enhanced DPM-induced lung tumor incidence, while P-carotene, which may act as a
17 free radical scavenger, partially reduced the tumorigenic response. Formation of 8-OHdG was
18 again significantly correlated with lung tumor incidence in these studies, except at the highest
19 dose. Dasenbrock et al. (1996) reported that extracted DPM, intratracheally instilled into rats (15
20 mg total dose) induced only marginal increases in lung tumor induction, while unextracted DPM
21 was considerably more effective. While adducts were not measured in this study, it nevertheless
22 provides support for the likelihood that either activation of organic metabolites and/or generation
23 of oxygen free radicals from organics are involved in the carcinogenic process.
24 Additional support for the involvement of particle-associated radicals in tissue damage
25 was provided by the finding that pretreatment with superoxide dismutase (SOD), an antioxidant,
26 markedly reduced lung injury and death due to instillation of DPM. Similarly Hirafuji et al.
27 (1995) found that the antioxidants catalase, deferoxamine, and MK-447 inhibited the toxic
28 effects of DPM on guinea pig trachea! cells and tissues in vitro.
29 Although the data presented supported the hypothesis that generation of reactive oxygen
30 species resulting from exposure to DPM is involved in the carcinogenic process, it should be
31 noted that 8-OHdG is efficiently repaired and that definitive proof of a causal relationship in
32 humans is stiii lacking. It is also uncertain whether superoxide or hydroxyl radicals chemically
33 generated by DPM alone promote 8-OHdG production in vivo and induce lung toxicity, because
34 SOD is extensively located in mammalian tissues. Nevertheless, demonstration that oxygen free
35 radicals can be generated from particle-associated organics, that their presence will induce adduct
11/5/99 7-134 DRAFT—DO NOT CITE OR QUOTE
-------
1 formation and DNA damage unless repaired, that tumor induction in experimental animals
2A correlated with OhdG adducts, and that treatment with antioxidant limits lung damage, provides
3 strong support for the involvement of oxygen free radicals in the toxicologic and carcinogenic
4 response to diesel exhaust.
5
6 7.4.4. Relationship of Physical Characteristics of Particles to Cancer Induction
7 The biological potential of inhaled particles is strongly influenced by surface chemistry
8 and character. For example, the presence of trace metal compounds such as aluminum and iron,
9 as well as ionized or protonated sites, is important hi this regard (Langer and Nolan, 1994). A
10 major factor is specific surface area (surface area/mg). PMNs characteristically are increased
11 abnormally in the lung by DE exposure, but their presence in the lungs does not appear to be
12 excessive following the pulmonary deposition of even high lung burdens of spherical TiO2
13 particles in the 1-2 um diameter range (Strom, 1984; Lee et al., 1986). In these studies lung
14 tumors were detected only at an inhaled concentration of 250 ug/m3. In a more recent study in
15 which rats were exposed to TiO, hi the 15-40 nm size range, inhibition of particle clearance and
16 tumorigenesis were induced at concentrations of 10 mg/m3 (Heinrich et al., 1995). Oberdorster
17 and Yu (1990) compared the results of several chronic inhalation studies and found that
carcinogenic potency related to specific particle surface area. Heinrich et al. (1995) also found
that lung tumor rates increased with specific particle surface area following exposure to diesel
20 exhaust, carbon black, or titanium dioxide, irrespective of particle type. Langer and Nolan
21 (1994) reported that the hemolytic potential of Min-lI-Sil 15, a silica flour, increased in direct
22 relationship to specific surface area at nominal particle diameters ranging from 0.5 to 20 um.
23 Ultrafine particles appear to be more likely to be taken up by lung epithelial cells. Riebe-
24 Imre et al. (1994) reported that CB is taken up by lung epithelial cells in vitro, inducing
25 chromosomal damage and disruption of the cytoskeleton, lesions that closely resemble those
26 present hi tumor cells. Johnson et al. (1993) reported that 20-nm polytetrafluoroethylene
27 particles are taken up by pulmonary epithelial cells as well as polymorphonuclear leucocytes,
28 inducing an approximate 4-, 8-, and 40-fold increase in the release of interleukin-1 alpha and
29 beta, inducible nitric oxide synthetase, and macrophage inflammatory protein, respectively.
30 The carcinogenic potency of diesel particles, therefore, appears to be related, at least to
31 some extent, to then* small size and convoluted shape, which results in a large specific particle
32 surface area. While toxicity and carcinogenicity increased with decreasing particle size into the
33 submicron range, it is uncertain if toxic and carcinogenic potential continues to increase as
34 particle size decreases even further. The relationship between particle size and toxicity is of
concern because, as noted in Chapter 2, newer engines equipped with more advanced emission
11/5/99 7-135 DRAFT—DO NOT CITE OR QUOTE
-------
1 controls emit greater numbers of particles in the nanometer size range. Other than disruption of
2 the cytoskeleton of epithelial cells, there is little information regarding the means by which
3 particle size influences carcinogenicity as well as noncancer toxicity.
4
5 7.4.5. Integrative Hypothesis For Diesel-induced Lung Cancer
6 The induction of lung cancer by large doses of carbon black via inhalation (Heinrich et
7 al., 1995; Mauderly et al., 1991; Nikula et al., 1995) or intratracheal instillation (Kawabata et al.,
8 1994; Pott et al., 1994; Dasenbrock et al., 1996) led to the development of the lung particle
9 overload hypothesis. According to this hypothesis the induction of neoplasia by insoluble,
10 biochemically inert particles is associated with an inhibition of lung particle clearance and the
11 involvement of persistent alveolar epithelial hyperplasia. Driscoll (1995), Driscoll et al. (1996),
12 and Oberdorster and Yu (1990) outlined a proposed mechanism for the carcinogenicity of DE at
13 high doses that emphasizes the role of phagocytic cells. Following exposure, phagocytosis of
14 particles acts as a stimulant for oxidant production and inflammatory cytokine release by lung
15 phagocytes. It was hypothesized that at high particle exposure concentrations the quantity of
16 mediators released by particle-stimulated phagocytes exceeds the inflammatory defenses of the
17 lung (e.g., antioxidants, oxidant-metabolizing enzymes, protease inhibitors, cytokine inhibitors),
18 resulting in tissue injury and inflammation. With continued particle exposure and/or the
19 persistence of excessive particle burdens, there then develops an environment of phagocytic
20 activation, excessive mediator release-tissue injury and, consequently, more tissue injury,
21 inflammation, and tissue release. This is accompanied by cell proliferation. As discussed in a
22 review by Cohen and Ellwein (1991), conceptually, cell proliferation can increase the likelihood
23 that any oxidant-induced or spontaneously occurring genetic damage becomes fixed in a dividing
24 cell and is clonally expanded. The net result of chronic particle exposures sufficient to elicit
25 inflammation and cell proliferation in the rat lung is an increased probability that the genetic
26 changes necessary for neoplastic transformation will occur. A schematic of this hypothesis has
27 been outlined by McClellan (1997) (see Figure 7-5). In support of this hypothesis, it was
28 reported that concentrations of inhaled CB resulted in increased cytokine expression and
29 inflammatory influx of neutrophils (Oberdorster et al., 1995), increased formation of 8-OhdG
30 (Ichinose et al., 1997b), and increase in the yield of hprt mutants, an effect ameliorated by
31 treatment with antioxidants (Driscoll, 1995; Driscoll et al., 1996). Metabolism of carcinogenic
11/5/99 7-136 DRAFT—DO NOT CITE OR QUOTE
-------
Carbon black
Exposure
Clearance
Macrophage v^ESSi
I
Cytokines
Growth factors
Proteases
Reactive
oxygen
species
Diesel exhaust
Particulate matter
I Exposure
I Deposition
Desorption
b
1 Organic chemicals
Unique
to
diesel
Activation
Inflammation
Cell injury
Cell proliferation
Hyperplasia
DNA adducts
Mutations
/
Initiation
Activation of
protooncogenes
Inactivation of tumor
suppressor genes
Promotion
Progression I
Fibrosis
Initiated cell
0 ° (~\ Preneoplastic
n _a^ lacinn
lesion
Malignant
tumor
Figure 7-5. Pathogenesis of lung disease in rats with chronic, high-level exposures to particles.
Source: McClellan, 1997.
11/5/99
7-137 DRAFT—DO NOT CITE OR QUOTE
-------
1 organics to active forms as well as the generation of reactive oxygen species from certain organic
2 species are likely to contribute to the toxic and carcinogenic process.
3 At low concentrations, inflammatory effects associated with lung particle overload are
4 generally absent. However, activation of organic carcinogens and generation of oxidants from
5 the organic fraction can still be expected. Actual contribution depends upon elution and the
6 effectiveness of antioxidants. Direct effects of ultrafine diesel particles taken up by epithelial
7 cells are also likely to play a role.
8 While high-dose induction of cancer is logically explained by this hypothesis, particle
9 overload has not been clearly shown to induce lung cancer in other species. As noted in the
10 quantitative chapter, the relevance of the rat pulmonary response is therefore problematic. The
11 rat pulmonary noncancer responses to DPM, however, have fairly clear interspecies and human
12 parallels. In response to poorly soluble particles such as DPM, human and rats both develop an
13 alveolar macrophage response, accumulate particles in the interstitium, and show mild interstitial
14 fibrosis (ILSI, 1999). Other species (mice, hamsters) also have shown similar noncancer
15 pulmonary responses to DPM, but without accompanying cancer response. The rat response for
16 noncancer pulmonary histopathology, however, seems to be more pronounced compared to
17 humans or other species, i.e., rats appear to be more sensitive. Although many critical elements
18 of interspecies comparison such as the role of airway geometry and patterns of particle deposition
19 need further elucidation, this basic interspecies similarity and greater sensitivity of pulmonary
20 response make pulmonary histopathology in rats a valid *basis for noncancer dose-response
21 assessment.
22
23 7.4.6. Summary
24 Recent studies have shown tumor rates resulting from exposures to nearly organic-free
25 CB particles at high concentrations to be similar to those observed for DE exposures, thus
26 providing strong evidence for a particle overload mechanism for DE-induced pulmonary
27 carcinogenesis in rats. Such a mechanism is also supported by the fact that carbon particles per
28 se cause inflammatory responses and increased epithelial cell proliferation and that AM function
29 may be compromised under conditions of particle overload.
30 The particle overload hypothesis appears sufficient to account for DE-induced lung
31 cancer in rats. However, there is increasing evidence for lung cancer induction in humaas at
32 concentrations insufficient to induce lung particle overload (see Chapter 7). Uptake of particles
33 by epithelial cells at ambient or occupational exposure levels. DNA damage resulting from
34 oxygen-free radicals generated from organic molecules, and the gradual in siru extraction and
35 activation of procarcinogens associated with the diesel particles are likely to play a role in this
11/5/99 7-138 DRAFT—DO NOT CITE OR QUOTE
-------
1 response. The slower particle clearance rates in humans (up to a year or more) increase the
t likelihood of significant extraction of organics. This is supported by reports of increased DNA
adducts in humans occupationally exposed to diesel exhaust at concentrations unlikely to induce
4 lung particle overload. While these modes of action can be expected to function at lung overload
5 conditions also, they are likely to be overwhelmed by inflammatory associated effects.
6 The evidence to date indicates that caution must be exercised in extrapolating
7 observations made in animal models to humans when assessing the potential for DE-induced
8 pulmonary carcinogenesis. The carcinogenic response and the formation of DNA adducts in rats
9 exposed to diesel exhaust and other particles at high exposure concentrations may be species-
10 specific and not particle-specific. The likelihood that different modes of action predominate at
11 high and low doses also renders low-dose extrapolation to ambient concentrations uncertain.
12
13 7.5. CANCER WEIGHT-OF-EVIDENCE: HAZARD EVALUATION
14 7.5.1. Cancer Hazard Summary
15 Diesel engine exhaust is "highly likely" to be carcinogenic by the inhalation route of
16 exposure, according to EPA's 1996 Proposed Guidelines for Carcinogen Risk Assessment. The
17 hazard is viewed as being applicable to ambient-environmental exposures. There is no available
evidence to evaluate the hazard from other routes of exposure. The "likely" classification
generally compares with other agents designated as "B-l probable human carcinogen" under the
20 EPA's 1986 Guidelines for Carcinogen Risk Assessment, though the overall weight-of-evidence
21 for diesel exhaust (DE) places it at the upper end of this grouping and hence gives it a "highly
22 likely" designation under the proposed Guidelines. The carcinogenic potential of DE is indicated
23 by (1) a consistent statistically increased association between observed lung cancer and DE
24 exposure in certain occupationally exposed workers, (2) the induction of lung cancer in some but
25 not all animal experiments, (3) mutagenic and carcinogenic activity of the particle organic
26 extracts, (4) the presence of individual organic compounds having known mutagenic and/or
27 carcinogenic properties (e.g., PAH derivatives and nitro-PAHs), and (5) limited evidence for the
28 bioavailability of the organics. The mode of action for carcinogenicity in humans is unknown; it
29 is suspected that either the organics, the elemental carbon diesel particle, or both contribute to the
30 carcinogenic activity.
31 Increases in relative risk for lung cancer have been consistently noted hi a number of
32 epidemiologic studies, and causality considerations for this observed association are consistent
33 with DE exposure being causally related to lung cancer. An inability to satisfactorily minimize
34 all confounding, bias, and exposure uncertainties, coupled with the magnitude of the relative
11/5/99 7-139 DRAFT—DO NOT CITE OR QUOTE
-------
1 risks, limits the human evidence from being considered sufficient to characterize DE as a
2 "known" human carcinogen.
3 While lung cancer has been induced experimentally in rats via inhalation of DE at high
4 exposure concentrations, and in rats and mice via intratracheal instillation of diesel particles and
5 particle extracts, these responses appear to be mediated primarily by inflammation and
6 subsequent pathology related to lung particle overload. Because the persistent-chronic overload
7 inflammatory responses in the rat are not seen at lower test exposures (or at ambient DE
8 concentrations), and uncertainty remains whether induction of inflammatory responses in humans
9 will lead to lung cancer, rat bioassay data are not completely irrelevant for human hazard
10 characterization. However, the rat lung cancer response data are unsuitable for estimating human
11 risk at environmental levels of exposure.
12 The plausibility of an environmental hazard is supported by (1) considering that
13 mutagenic compounds and tumor-initiating carcinogens (e.g., PAH derivatives and nitro-PAHs)
14 are present in small quantities hi the DE organic mixture, which qualitatively implies a
15 nonthreshold mode of action for these agents; and (2) noting that there could be little difference
16 between higher-end environmental exposures and some occupational levels where increased
17 relative risks of about 1.4 are seen (e.g., exposure estimates for some truck drivers could be
18 overlapping some environmental estimates, they also may have somewhat higher relative risks).
19 For these reasons, the extrapolation of the occupational hazard to ambient environmental
20 exposures is judged plausible and prudent. In the absence of evidence to the contrary, and
21 recognizing the mutagenic potential of the organics, it would also be feasible to evaluate dose
22 response using linear models, at least at low exposure levels.
23 Overall, the evidence for a likely human carcinogenic hazard by inhalation is strong,
24 even though inferences are involved. Uncertainties remain, including (1) methodologic
25 limitations in the epidemiologic studies as well as a lack of assured historical exposure data for
26 occupationally exposed cohorts, (2) uncertainties regarding the extent of bioavailability of
27 organic compounds present on diesel particles, and (3) uncertainties regarding the mode of action
28 in humans.
29
30 7.5.2. Supporting Information
31 7.5.2.1. Human Data
32 An increased relative risk for lung cancer and DE exposure has been observed in more
33 than 30 epidemiologic studies. The excess risk is observed in both cohort and case-control study
34 designs. Additionally, consistent and statistically significant elevated pooled relative risks
35 ranging from 1.33 to 1.47 were derived in several meta-analyses. In some studies, the effects of
11/5/99 7-140 DRAFT—DO NOT CITE OR QUOTE
-------
smoking were accounted for and the increased relative risks prevailed. When the meta- analysis
focused only on the smoking-controlled studies, the relative risks tended to increase. A few
3~" individual studies had smoking-adjusted relative risks exceeding 1.5 (e.g., Steenland et al., 1990
4 [RR 1.64, 1.89]). The uncertainties with the epidemiologic data are typical ones including the
5 possibility that chance, bias, or confounding are influencing the observed lung cancer increases.
6 The persistence of this association in so many studies indicates that chance alone is unlikely to
7 account for the observed relation between DE and lung cancer. A causal interpretation for DE is
8 enhanced when the "Hill" causality criteria are evaluated, noting that a weakness or absence in
9 one or several of the criteria does not prevent a causal interpretation. A weakness in the
10 epidemiologic studies is due to the fact that the information from which diesel exposure can be
11 inferred is based on job codes, area descriptions, etc., which are surrogates for the true underlying
12 exposure. This can lead to "nondifferential" misclassification of exposure, and while unlikely,
13 might result in a spurious risk estimate in any one study. It is even more unlikely, however, that
14 it would bias a sufficient number of studies in a uniform direction to account for the persistent
15 aggregate association observed. Moreover, any bias would likely be toward a lower risk
16 estimate. In those studies where the confounding effect of smoking was controlled, there remains
17 the suspicion that the statistical adjustment for smoking may not be completely effective, and
1^k residual confounding by smoking may persist to bias the correlation of DE exposure with lung
^^ cancer occurrence.
20
21 7.5.2.2. Animal Data
22 Numerous studies have shown that inhalation of DE and intratracheal instillation of diesel
23 particles or particle extract result in the induction of lung cancer in rats. Although evidence of
24 lung cancer induction from DE in mice via inhalation exposure is equivocal, positive results have
25 been obtained by intratracheal instillation of diesel particles. Attempts to induce lung cancer in
26 Syrian hamsters by inhalation of DE have been unsuccessful, but this species is known to be
27 resistant to the induction of lung cancer. Although cats and monkeys have been exposed to DE,
28 the durations of exposure were inadequate to evaluate carcinogenicity. As supported by an
29 expert panel (ILSI, 1998), the high-dose rat data are unsuitable for predicting a low-exposure
30 human risk. Because it is unknown whether high lung burdens of poorly soluble particles (e.g.,
31 diesel particles) can lead to lung cancer in humans via mechanisms similar to those of the rat,
32 there are insufficient data to conclude that the rat response is completely irrelevant for human
33 hazard identification. Intratracheal instillation studies in rats and mice reveal that diesel particles
34 as well as the particle organic extracts can elicit a lung cancer response.
11/5/99 7-141 DRAFT—DO NOT CITE OR QUOTE
-------
1 7.5.2.3. Other Key Data
2 Organic extracts of DE particles have been shown to induce tumors in mice, both by skin
3 painting and subcutaneous injection, and to be mutagenic in several test systems. Additionally, a
4 number of PAHs and nitro-PAHs present on diesel particles as well as in the vapor phase are
5 loKJwn to be mutagenic and/or carcinogenic. Further evidence for the presence of carcinogenic
6 agents in DE is provided by the reported induction of dermal tumors following skin painting of
7 diesel particle extracts.
8
9 7.5.2.4. Mode of Action
10 The mode of action for DE carcinogenicity, especially at nonparticle overload exposure
11 conditions, remains to be established. There is some evidence and thus plausibility that diesel
12 particles as well as particle-associated organics are involved in the carcinogenic process. The rat
13 model shows that at high-exposure concentrations, particle-overload-induced inflammatory
14 responses, associated DNA damage, and rapid cell turnover are likely the primary factors
15 responsible for lung cancer induction. It is not known whether humans have a similar response
16 pattern at high exposures, though it has not been historically observed. At low exposure levels,
17 cancer induction is more likely to be due to organic compounds, although there is some evidence
18 that ultrafine diesel particles at low concentrations are ingested by epithelial cells and induce
19 DNA damage. DNA damage in blood cells of occupationally exposed workers indicates at least
20 some degree of elution of organic compounds from the particle and subsequent entry into the
21 bloodstream. Studies have also suggested that bioavailability may be greater at low-exposure
22 concentrations because the particles are not aggregated. A significant percentage of particles in
23 humans are deposited at the branchings of small airways rather than alveoli, and the residence
24 time of organic compounds eluted at those locations is greater, increasing the likelihood of
25 metabolism to an activated state. A variety of carcinogenic compounds (e.g., PAH derivatives
26 and nitro-PAHs), a number of which are mutagens and carcinogens, are present on the diesel
27 particle. It has also been shown that reactive oxygen species capable of damaging DNA are
28 generated by the metabolism of DE organics with quinone-like structures.
29
30 7.6. DISCUSSION OF THE ROLE OF DIESEL EXHAUST IN THE OVERALL
31 PICTURE OF PM,0
32 It is very difficult to assess exposure to diesel emissions because they are highly complex
33 mixtures and constitute only a small portion of a broader mix of air pollutants. For example,
34 combustion of other materials, such as fossil fuel and tobacco, produces many of the same
3 5 chemical components present in diesel emissions; furthermore, both natural and manmade
11 /5/99 7-142 DRAFT—DO NOT CITE OR QUOTE
-------
sources of respirable particles are common. No single constituent of diesel exhaust serves as a
unique marker of exposure.
3 Whether air pollution contributes to the occurrence of lung cancer is a matter of wide
4 debate. Ambient air in urban or industrialized areas can be contaminated by chemicals, some of
5 which are definitely known to be carcinogenic. Known carcinogens that occur in ambient air
6 include arsenic, asbestos, benzene, cadmium, and polycyclic aromatic hydrocarbons. However,
7 the information on ambient exposure is sparse. Detailed measurements of such substances over
8 long exposure periods and for large geographic areas are rarely available. Many descriptive
9 epidemiologic studies demonstrate increased lung cancer risk in urban and industrialized areas
10 (Hemminki, 1994). Frequently, those differences have been partially explained by differences in
11 air pollution; however, such correlations might have other explanations and do not represent final
12 conclusive evidence. The coincidence of diesel exhaust exposure and air pollution in urban
13 airsheds poses important questions about whether the observed association of an increased lung
14 cancer risk in urban and industrialized areas can be attributed to diesel exhaust exposure.
1 5 The contribution of diesel particles to PM10 (particles <, 10 um diameter) are difficult to
16 determine, although most estimates indicate they constitute only a small fraction. For example,
17 in an analysis conducted in the Los Angeles basin in the early 1980s, diesel emissions accounted
ffor approximately 3% of the mass of PM10. Because 90% of diesel particles are less than 1 um
diameter they make up a larger percentage of fine and ultrafine ambient PM. For example, the
20 EPA has estimated that diesel particles accounts for 5.7% of all PM2.5 and 21% of PM2.5
21 excluding natural and fugitive dust sources (see Chapter 2 for details). Since smaller particles
22 appear to be more toxic/carcinogenic, the size distribution as well as mass of diesel particles
23 relative to other PM are important considerations in any attempt to estimate the contribution of
24 DE to PM induced toxicity and/or carcinogenicity.
25
26 7.7. REFERENCES
27
28 Adamson, IYR; Bowden, DH. (1978) Adaptive response of the pulmonary macrophagic system to carbon. II.
29 Morphologic studies. Lab Invest 38:430-438.
30
31 Ahlberg, J; Ahlbom, A; Lipping, H; et al. (1981) [Cancer among professional drivers—a problem-oriented register-
32 based study]. Lakartidningen 78:1545-1546.
33
34 Allen, RC; Loose, LD. (1976) Phagocytic activation of a luminol-dependent chemiluminescence in rabbit alveolar
35 and peritoneal macrophages. Biochem Biophys Commun 69:245-252.
36
37 Ames, BN. (1983) Dietary carcinogens and anticarcinogens. Science 221:1256-1264.
Jfe
^p Aronsen, KJ; Siemiatycki, J; Dewar, R; et al. (1996) Occupational risk factors for prostate cancer: results from a
40 case-control study in Montreal, Quebec, Canada. Am J Epidemiol 143(4):363-373.
11/5/99 7-143 DRAFT—DO NOT CITE OR QUOTE
-------
1 Beall, GD; Repine, JE; Hoidal, JR; et al. (1977) Chemiluminscence by human alveolar macrophages: stimulation
2 with heat killed bacteria or phorbol myristate acetate. Infect Immunol 1 7: 1 1 7- 1 20.
3
4 Benhamou, S; Benhamou, E; Flamant, R. (1988) Occupational risk factors of lung cancer in a French case-control
5 study. Br J Ind Med 45:23 1-233.
6
7 Bhatia, R; Lopipero, P; Smith, A. (1998) Diesel exhaust exposure and lung cancer. Epidemiol 9(1):84-91.
8
9 Bitterman, PB; Aselberg, S; Crystal, RG. (1983) Mechanism of pulmonary fibrosis: spontaneous release of the
1 0 alveolar macrophage-derived growth factor in interstitial lung disorders. J Clin Invest 72: 1 80 1 - 1 8 1 3 .
11
1 2 Boffetta, P; Harris, RE; Wynder, EL. (1990) Case-control study on occupational exposure to diesel exhaust and lung
1 3 cancer risk. Am J Ind Med 1 7: 577-59 1 .
14
1 5 Boffetta, P; Stellman, SD. (1988) Association between diesel exhaust exposure and multiple myeloma: an example
1 6 of confounding. Prev Med 17:236-237.
17
1 8 Bohning, DE; Atkins, HL; Cohn, SH. (1982) Long-term particle clearance in man: normal and impaired. Ann
19 OccupHyg 26:259-271.
20
2 1 Bond, JA; Johnson, NF; Snipes, MB; et al. (1990) DNA adduct formation in rat alveolar type II cells: cells
22 potentially at risk for inhaled diesel exhaust. Environ Mol Mutagen 16:64-69.
23
24 Bond, JA; Mitchell, CE; Li, AP. (1983) Metabolism and macromolecular covalent binding of benzo[a]pyrene in
25 cultured Fischer-344 rat lung type II epithelial cells. Biochem Pharmacol 32:3771-3776.
26
27 Bond, JA; Butler, MM; Medinsky, MA; et al. (1984) Dog pulmonary macrophage metabolism of free and particle-
28 associated [14C]benzo[a]pyrene. J Toxicol Environ Health 14:181-189.
29
30 Boorman, GA; Brockman, M; Carlton, WW; et al. (1996) Classification of cystic keratinizing squamous lesions of
31 the rat lung:report of a workshop. Toxicologic Pathol. 24:564-572.
32
33 Borm, PJA; Knaapen, AM; Schins, RFP; et al. (1997) Neutrophils amplify the formation of DNA adducts by
34 benzo[a]pyrene in lung target cells. Environ Health Perspect 105 (Suppl 5):1089-1093.
35
36 Bowden, DH; Adamson, IYR. (1978) Adaptive responses of the pulmonary macrophagic system to carbon. I.
37 Kinetic studies. Lab Invest 3 8 :422-43 8.
38
39 Boyd, MR. (1984) Metabolic activation and lung toxicity: a basis for cell-selective pulmonary damage by foreign
40 chemicals. Environ Health Perspect 55:47-5 1 .
41
42 Brightwell, J; Fouillct, X; Cassano-Zoppi, A-L; ct al. (1986) Nccplastic and functional changes in rodents after
"a» -•* •- -• - •. _»._*,..•.; — „.£•_,„ _•„, „ — i.. ---- k ,— _,.: — :„„.„ *•__. »,,ui_:_*.: >». Tr^:*^.— ; A. A/f^r^i^**^*, *a.
-------
Coggon, D; Pannett, B; Acheson, ED. (1984) Use of job-exposure matrix in an occupational analysis of lung and
bladder cancers on the basis of death certificates. J Natl Cancer Inst 72:61-65.
4 Cohen, SM; Ellwein, LB. (1991) Genetic error, cell proliferation and carcinogenesis. Cancer Res 51:6493-6505.
5
6 Cordier, S; Clavel, J; Limasset, JC; et al. (1993) Occupational risks of bladder cancer in France: a multicenter case-
7 control study. Int J Epidemiol 22:403-411.
8
9 Cox, LA. (1997) Does diesel exhaust cause human lung cancer? J Risk Anal 17(6):807-829.
10
11 Crump, KS.- (1999) Lung cancer mortality and diesel exhaust: reanalysis of a retrospective cohort study of U.S.
12 railroad workers. Inhal Toxicol. 11:1-17. .
13
14 Damber, LA; Larsson, LG. (1987) Occupation and male lung cancer: a case-control study in northern Sweden. Br J
15 IndMed 44:446-453.
16
17 Daniel, JH. (1984) The use of diesel-powered equipment in U.S. underground coal operations. Presented at:
18 American Mining Congress international coal show; May, 1984; Chicago, IL. Washington, DC: Department of the
19 Interior, Bureau of Mines.
20
21 Dasenbrock, C; Peters, L; Creutzenberg, O; et al. (1996) The carcinogenic potency of carbon particles with and
22 without PAH after repeated intratracheal administration in the rat. Toxicol Lett 88:15-21.
23
24 Depass, LR; Chen, KC; Peterson, LG. (1982) Dermal carcinogenesis bioassays of diesel particulates and
25 dichloromethane extract of diesel particulates in C3H mice. In: Lewtas, J, ed. Toxicological effects of emissions
26 from diesel engines. Amsterdam: Elsevier; pp. 321-326.
Dontenwill, W; Chevalier, H-J; Harke, H-P; et al. (1973) Investigations on the effects of chronic cigarette smoke
29 inhalation in Syrian golden hamsters. J Natl Cancer Inst 51:1681 -1832.
30
31 Drath, DB; Karnovsky, ML. (1975) Superoxide production by phagocytic leukocytes. J Exp Med 141:257-262.
32
33 Driscoli, K. (1995) Role of inflammation in the development of rat lung tumors in response to chronic particle
34 exposure. In: Proceedings of the MIT toxicology symposium: particle overload in the rat lung and lung cancer:
35 relevance for human risk assessment Cambridge, MA: Massachusetts Institute of Technology; 29 pp.
36
37 Driscoli, KE; Carter, JM; Hassenbein, DG; et al. (1997) Cytokines and particle-induced inflammatory cell
3 8 recruitment Environ Health Perspect 105 (Suppl 5): 1159-1164.
39
40 Driscoli, KE; Carter, JM; Howard, BW; et al. (1996) Pulmonary inflammatory, chemokine, and mutagenic
41 responses in rats after subchronic inhalation of carbon black. Toxicol Appl Pharmacol 136:372-380.
42
43 Edling, C; Anjou, C-G; Axelson, O; et al. (1987) Mortality among personnel exposed to diesel exhaust. Int Arch
44 Occup Environ Health 59:559-565.
45
46 El-Bayoumy, K; Rivenson, A; Johnson, B; et al. (1988) Comparative tumorigenicity of 1-nitropyrene, 1-
47 nitrosopyrene, and 1-aminopyrene administered by gavage to Sprague-Dawley rats. Cancer Res 48:4256-4260.
48
49 Emmelin, A; NystrQm, L; Wall, S. (1993) Diesel exhaust exposure and smoking: a case-referent study of lung
50 cancer among Swedish dock workers. Epidemiology 4:237-244.
51
Enya, T; Suzuki, H; Watanabe, T; et al. (1997) 3-Nitrobenzanthrone, a powerful bacterial mutagen and suspected
human carcinogen found in diesel exhaust and airborne particulates. Environ Sci Technol 31:2772-2776.
1175/99 7-145 DRAFT—DO NOT CITE OR QUOTE
-------
1 Evans, M; Shami, SG; Martinez, LA. (1986) Enhanced proliferation of pulmonary alveolar macrophages after
2 carbon instillation in mice depleted of blood monocytes by strontium-89. Lab Invest 54:154-159.
3
4 Farrell, RL; Davis, GW. (1974) Effect of participate benzo(a)pyrene carrier on carcinogenesis in the respiratory
5 tract of hamsters. In: Karbe, E; Davis, GW, eds. Experimental lung cancer. Berlin: Springer; pp. 186-198.
6
7 Flodin, U; Fredriksson, M; Persson, B. (1987) Multiple myeloma and engine exhausts, fresh wood, and creosote: a
8 case-referent study. Am J Ind Med 12:519-529.
9
10 Gallagher, J; Heinrich, U; George, M; et al. (1994) Formation of DNA adducts in rat lung following chronic
11 inhalation of diesel emissions, carbon black, and titanium dioxide particles. Carcinogenesis 15:1291-1299.
12
13 Garland, FC; Gorham, ED; Garland, CF; et al. (1988) Testicular cancer in US Navy personnel. Am J Epidemiol
14 127:411-414.
15
16 Garshick, E; Schenker, MB; Munoz, A; et al. (1988) A retrospective cohort study of lung cancer and diesel exhaust
17 exposure in railroad workers. Am Rev Respir Dis 137:820-825.
18
19 Garshick, E; Schenker, MB; Munoz, A; et al. (1987) A case-control study of lung cancer and diesel exhaust
20 exposure in railroad workers. Am Rev Respir Dis 135:1242-1248.
21
22 Gerde, P; Medinsky, MA; Bond, JA. (1991) Particle-associated polycyclic aromatic hydrocarbons-a reappraisal of
23 their possible role in pulmonary carcinogenesis. Toxicol Appl Pharmacol 188:1-13.
24
25 Goldstein, IM; Roos, D; Kaplan, HB; et al. (1975) Complement and immunoglobulins stimulate superoxide
26 production by human leukocytes independently of phagocytosis. J Clin Invest 56:1155-1163.
27
28 Grimmer, G; Brune, H; Deutsch-Wenzel, R; et al. (1987) Contribution of polycyclic aromatic hydrocarbons and
29 nitro-derivatives to the carcinogenic impact of diesel engine exhaust condensate evaluated by implantation into the
30 lungs of rats. Cancer Lett 37:173-180.
31
32 Guberan, E; Usel, M; Raymond, L; et al. (1992) Increased risk for lung cancer and for cancer of the gastrointestinal
33 tract among Geneva professional drivers. Br J Ind Med 49:337-344.
34
35 Gustafsson, L; Wall, S; Larsson, LG; et al. (1986) Mortality and cancer incidence among Swedish dock workers—a
36 retrospective cohort study. Scand J Work Environ Health 12:22-26.
37
38 Gustavsson, P; Plato, N; Lidstr5m, EB; et al. (1990) Lung cancer and exposure to diesel exhaust among bus garage
39 workers. Scand J Work Environ Health 16:348-354.
40
41 Hall, NEL; Wynder, EL. (1984) Diesel exhaust exposure and lung cancer: a case-control study. Environ Res 34:77-
42 86.
43
44 Hansen, ES. (1993) A foilow-up study on the mortality of truck drivers. Am J Ind Ivied 23:Si i-821.
45
46 Hatch, GE; Gardner, DE; Menzel, DB. (1980) Stimulation of oxidant production in alveolar macrophages by
47 pollutant and latex particles. Environ Res 32:121-136.
48
49 Hayes, RB; Thomas, T; Silverman, DT; Vineis, P; Blot, WJ; Mason, TJ; Pickle, LW; Correa, P; Fontham, ETH;
50 Schoenberg, JB. (1989) Lung cancer in motor exhaust-related occupations. Am J Ind Med 16:685-695.
51
52 PI?™™ Hf??CtS Institute. flQQ^ Oioc^l gvViqnct* a C"tlC^l 5*Ha^vS*S o^ATT1'cc'rtTllC PYnnQiirp /»nH health pflPftct*?
53 Cambridge, MA: Related HEI Publications; p. 27.
54
11/5/99 7-146 DRAFT—DO NOT CITE OR QUOTE
-------
Heinrich, U. (1990) Presentation at workshop on health effects of diesel emissions. U.S. EPA, Research Triangle
Park, July 1990.
4~~ Heinrich, U. (1990) Carcinogenicity of carbon black "Printex 90" in female Wistar rats. Presented at: U.S.
5 Environmental Protection Agency Diesel Emissions Peer Review Workshop; July 18-19; Chapel Hill, NC.
6
7 Heinrich, U; Muhle, H; Takenaka, S; et al. (1986) Chronic effects on the respiratory tract of hamsters, mice and rats
8 after long-term inhalation of high concentrations of filtered and unfiltered diesel engine emissions. J Appl Toxicol
9 6:383-395.
10
11 Heinrich, U; Peters, L; Fuhst, R; et al. (1989a) The effect of automotive exhaust exposure on the carcinogenicity of
12 diphenylnitrosamine (DPN) in the respiratory tract of rats. Exp Pathol 37:51-55.
13
14 Heinrich, U; Dungworth, DL; Pott, F; et al. (1994) The carcinogenic effects of carbon black particles and tar-pitch
15 condensation aerosol after inhalation exposure of rats. Ann Occup Hyg 48 (suppl. 4):351-356.
16
17 Heinrich, U; Fuhst, R; Rittinghausen, S; et al. (1995) Chronic inhalation exposure of Wistar rats and two different
18 strains of mice to diesel engine exhaust, carbon black, and titanium dioxide. Inhal Toxicol 7:533-556.
19'
20 Heinrich, U; Mohr, U; Fuhst, R; et al. (1989b) Investigation of a potential cotumorigenic effect of the dioxides of
21 nitrogen and sulfur, and of diesel-engine exhaust, on the respiratory tract of Syrian golden hamsters. Research
22 Report No. 26. Cambridge, MA: Health Effects Institute; pp. 27.
23
24 Heinrich, U; Pott, F; Rittinghausen, S. (1986a) Comparison of chronic inhalation effects in rodents after long-term
25 exposure to either coal oven flue gas mixed with pyrolized pitch or diesel engine exhaust. In: Ishinishi, N; Koizumi,
26 A; McClellan, R; Stober, W, eds. Carcinogenic and mutagenic effects of diesel engine exhaust. Amsterdam:
Elsevier; pp. 441-457.
29 Heinrich, U; Peters, L; Funcke, W; et al. (1982) Investigation of toxic and carcinogenic effects of diesel exhaust in
30 long-term inhalation exposure of rodents. In: Lewtas, J, ed. Toxicological effects of emissions from diesel
31 emissions. Amsterdam: Elsevier; pp. 225-242.
32
33 Hemminki, K; Soderling, J; Ericson, P; et al. (1994) DNA adducts among personnel servicing and loading diesel
34 vehicles. Carcinogenesis 15:767-769.
35
36 Hemminki, K; Pershagen, G. (1994) Cancer risk of air pollution: epidemiological evidence. Environ Health Perspect
37 102(Suppl4):187-192.
38
39 Hirafuji, M; Sakakibara, M; Endo, T; et al. (1995) Biological effects of diesel exhaust particles (DEP) on tissues and
40 cells isolated from respiratory tracts of guinea pigs. Res Com Mol Pathol Pharmacol 90:221-233.
41
42 Hirose, M; Lee, M-S; Wang, CY; et al. (1984) Induction of rat mammary gland tumors by 1-nitropyrene, a recently
43 recognized environmental mutagen. Cancer Res 44:1158-1162.
44
45 Hoar, SK; Hoover, R. (1985) Truck driving and bladder cancer mortality in rural New England. J Natl Cancer Inst
46 74:771-774.
47
48 Hoidal, JR; Repine, JE; Beall, GD; et al. (1978) The effect of phorbol myristate acetate on metabolism and
49 ultrastructure of human alveolar macrophages. Am J Pathol 91:469~482.
50
51 Hou, S-M; Lambert, B; Hemminki, K. (1995) Relationship between hprt mutant frequency, aromatic DNA adducts
and genotypes for GSTM1 and NAT2 in bus maintenance workers. Carcinogenesis 16:1913-1917.
11/5/99 7-147 DRAFT—DO NOT CITE OR QUOTE
-------
1 Howard, PC; Gerrard, JA; Milo, GE; et al. (1983) Transformation of normal human skin fibroblasts by 1-
2 nitropyrene and 6-nitrobenzo[a]pyrene. Carcinogenesis 4:353-355.
3
4 Howe, GR; Burch, JD; Miller, AB; et al. (1980) Tobacco use, occupation, coffee, various nutrients, and bladder
5 cancer. J Natl Cancer Inst 64: 701 -713.
6
7 Howe, GR; Fraser, D; Lindsay, J; et al. (1983) Cancer mortality (1965-77) in relation to diesel fume and coal
8 exposure in a cohort of retired railway workers. J Natl Cancer Inst 70:1015-1019.
9
10 Huisingh, J; Bradow, R; Jungers, R; et al. (1978) Application of bioassay to the characterization of diesel particle
11 emissions. In1. Waters, MD; Nesnow, S; Huisingh, JL; Sandhu, SS; Claxton, L, eds. Application of short-term
12 bioassay in the fractionation and analysis of comptex mixtures. New York: Plenum; pp. 3 8 i -418.
13
14 Humble, CG; Samet, JM; Skipper, BE. (1984) Comparison of self- and surrogate-reported dietary information. Am
15 JEpidemiol 119:86-98.
16
17 Ichinose, T; Yajima, Y; Nagashima, M; et al. (1997a) Lung carcinogenesis and formation of 8-hydroxy-
18 deoxyguanosine in mice by diesel exhaust particles. Carcinogenesis 18:185-192.
19
20 Ichinose, T; Yamanushi, T; Seto, H; et al. (1997b) Oxygen radicals in lung carcinogenesis accompanying
21 phagocytosis of diesel exhaust particles. Int J Oncol 11:571-575.
22
23 International Life Sciences Institute (ILSI). (1999) The relevance of the rat lung response to particle overload for
24 human risk assessment: workshop consensus report. Inhal Toxicol: In press.
25
26 IPCS. (1996) International Program on Chemical Safety Environmental Health Criteria 171. Diesel fuel and exhaust
27 emissions. Geneva: World Health Organization, pp. 172-176.
28
29 Iscovich, J; Castelleto, R; Esteve, J; et al. (1987) Tobacco smoking, occupational exposure and bladder cancer in
30 Argentina. Int J Cancer 40:734-740.
31
32 Ishinishi, N; Inamasu, T; Hisanaga, A; et al. (1988b) Intratracheal instillation study of diesel particulate extracts in
33 hamsters. In: Diesel exhaust and health risk. Ibaraki, Japan: Research Committee for HERP Studies; pp. 209-216.
34
35 Ishinishi, N; Kuwabara, N; Takaki, Y; et al. (1988a) Long-term inhalation experiments on diesel exhaust. In: Diesel
36 exhaust and health risks. Results of the HERP studies. Ibaraki, Japan: Research Committee for HERP Studies; pp.
37 11-84.
38
39 Iwai, K; Udagawa, T; Yamagishi, M; et al. (1986) Long-term inhalation studies of diesel exhaust on F344 SPF rats.
40 Incidence of lung cancer and lymphoma. In: Ishinishi, N; Koizumi, A; McClellan, R; StQber, W, eds. Carcinogenic
41 and mutagenic effects of diesel engine exhaust. Amsterdam: Elsevier; pp. 349-360.
42
43 Iwai, K.; Higuchi, K; Udagawa, T; et al. (1997) Lung tumor induced by long-term inhalation or intratracheal
44 instillation of diesel exhaust particles. Exp Pathol Toxicol 49:393-401.
45
46 Iyer, V; Harris, RE; Wynder, EL. (1990) Diesel exhaust exposure and bladder cancer risk; Eur J Epidemiol 6:49-54.
47
48 Jensen, OM; Wahrendorf, J; Knudsen, JB; et al. (1987) The Copenhagen case-referent study on bladder cancer
49 risks among drivers, painters, and certain other occupations. Scand J Work Environ Health 13:129-134.
5r*
W
51 Johnson, C; Finkelstein, J; Gelein, R; et al. (1993) Early alterations in the mRNA abundance of IL-lcc. IL-lB. iNOS:
52 MIP-2, TGFP, and VEGF associated with ultrafine particle exposure. Fundam Appl Toxicol 50:138 (abstr).
53
11/5/99 7-148 DRAFT—DO NOT CITE OR QUOTE
-------
Jordana, M; Richards, C; Irving, LB; et al. (1988) Spontaneous in vitro release of alveolar macrophage cytokines
after the intratracheal instillation of bleomycin in rats: characterization and kinetic studies. Am Rev Respir Dis
137:1135-1140.
4
5 Kaplan, HL; Mackenzie, WF; Springer, KJ; et al. (1982) A subchronic study of the effects of exposure of three
6 species of rodents to diesel exhaust. In: Lewtas, J, ed. Toxicological effects of emissions from diesel engines. New
7 York: Elsevier Biomedical; pp. 161-220.
8
9 Kaplan, HL; Springer, KJ; MacKenzie, WF. (1983) Studies of potential health effects of long-term exposure to
10 diesel exhaust emissions. Final Report No. 01-0750-103 (SWRI) and No. 1239 (SFRE). San Antonio, TX:
11 Southwest Research Institute.
12
13 Kaplan, I. (1959) Relationship of noxious gases to carcinoma of the lung in railroad workers. J Am Med Assoc
14 171:2039-2043.
15
16 Karagianes, RF; Palmer, RF; Busch, RH. (1981) Effects of inhaled diesel emissions and coal dust in rats. Am Ind
17 Hyg Assoc J. 42:382-391.
18
19 Kawabata, Y; Iwai, K; Udagawa, T; et al. (1986) Effects of diesel soot on unscheduled DNA synthesis of tracheal
20 epithelium and lung tumor formation. In: Ishinishi, N; Koizumi, A; McClellan, R; StOber, W, eds. Carcinogenic and
21 mutagenic effects of diesel engine exhaust. Amsterdam: Elsevier; pp. 213-232.
22
23 Kawabata, Y; Udagawa, T; Higuchi, K; et al. (1994) Early one year exposure to diesel engine exhaust causes lung
24 tumors. Toxic and carcinogenic effects of solid particles on the respiratory tract. In: Mohr, U; Dungworth, DL;
25 Mauderly, JL; et al., eds. Toxic and carcinogenic effects of solid particles in the respiratory tract. Washington, DC:
26 International Life Sciences Institute Press, pp. 429-431.
Kensler, TW; Egner, PA; Moore, KG; et al. (1987) Role of inflammatory cells in the metabolic activity of
29 polycyclic aromatic hydrocarbons in mouse skin. Toxicol Appl Pharm 90:337-346.
30
31 Klebanoff, SJ; Clark, RA. (1978) The metabolic burst. In: The neutrophil: function and clinical disorders.
32 Amsterdam: North-Holland; pp. 283-368.
33
34 Kolonel, LN; Hirohata, T; Nomura, AMY. (1977) Adequacy of survey data collected from substitute respondents.
35 Am J Epidemiol 106:476-484.
36
37 Kotin, P; Falk, H; Thomas, M. (1955) Aromatic hydrocarbons. III. Presence in the particulate phase of diesel engine
38 exhausts and the carcinogenicity of exhaust extracts. Arch Ind Health 11:113-120.
39
40 Kumagai, Y; Arimoto, T; Shinyashiki, M; et al. (1997) Generation of reactive oxygen species during interaction of
41 diesel exhaust particle components with NADPH cytochrome P450 reductase and involvement of the bioactivation
42 in the DNA damage. Free Radical Biol Med 22:479-487.
43
44 Kunitake, E; Shimamura, K; Katayama, H; et al. (1986) Studies concerning carcinogenesis of diesel particulate
45 extracts following intratracheal instillation, subcutaneous injection, or skin application. In: Ishinishi, N; Koizumi, A;
46 McClellan, R; StSber, W, eds. Carcinogenic and mutagenic effects of diesel engine exhaust. Amsterdam: Elsevier;
47 pp. 235-252.
48
49 Kunitake, E; Imase, A; Shimamura, K; et al. (1988) Skin application and subcutaneous injection experiments of
50 diesel particulate extracts using ICR mice and nude mice. In: Diesel exhaust and health risks. Results of the HERP
51 studies. Ibaraki, Japan: Research Committee for HERP Studies; pp. 217-225.
11/5/99 7-149 DRAFT—DO NOT CITE OR QUOTE
-------
1 Langer, AM; Nolan, RP. (1994) Factors controlling the biological potential of inorganic dusts: surface area,
2 chemistry and character. In: Mohr, U; Dungworth, DL; Mauderly, JL; et al., eds. Toxic and carcinogenic effects of
3 solid particles in the respiratory tract. Washington, DC: International Life Sciences Institute Press, pp. 147-156.
4
5 Lechner, JF; Mauderly, JL. (1994) Sequence of events in lung carcinogenesis: initiation and promotion,
6 protooncogenes and tumor suppressor genes and particulates. In: Mohr, U; Dungworth, DL; Mauderly, JL; et al.,
7 eds. Toxic and carcinogenic effects of solid particles in the respiratory tract. Washington, DC: International Life
8 Sciences Institute Press, pp. 235-251.
9
1 0 Lee, KP; Henry, NW, III; Trochimowicz, HJ; et al. (1986) Pulmonary response to impaired lung clearance in rats
1 1 following excessive TiO2 dust deposition. Environ Res 4 1 : 1 44- 1 67.
12
1 3 Lehnert, BE; Valdez, YE; Bomalaski, SH. (1988) Analyses of particles in the lung free cell, tracheobronchial lymph
1 4 nodal, and pleural space compartments following their deposition in the lung as related to lung clearance
1 5 mechanisms. Ann Occup Hyg 32:125-140.
16
1 7 Lehnert, BE; Ortiz, JB; London, JE; et al. (1989) Migratory behavior of alveolar macrophages during clearance of
1 8 light to heavy lung burdens of particles. 1989 Annual Meeting of the American Thoracic Society, Cincinnati, OH,
19 May 14-17.
20
21 Lerchen, ML; Wiggins, CL; Samet, JM. (1987) Lung cancer and occupation in New Mexico. J Natl Cancer Inst
22 79:639-645.
23
24 Lerchen, ML; Samet, JM. (1986) An assessment of the validity of questionnaire responses provided by a surviving
25 spouse. Am J Epidemiol 123:481-489.
26
27 Lewis, TR; Green, FHY; Moorman, WJ; et al. (1989) A chronic inhalation toxicity study of diesel engine emissions
28 and coal dust, alone and combined. J Am Coll Toxicol 8:345-375.
29
30 Lipsett, M; Campleman, S. (1999) Occupational exposure to diesel exhaust and lung cancer: a meta-analysis. Am
31 J Public Health 80(7):1009-1017.
32
33 Lloyd, JW. (1971) Long-term mortality of steelworkers. V. Respiratory cancer in coke plant workers. J Occup Med
34 13:53-68.
35
36 Lowrie, DB; Aber, VR. (1977) Superoxide production by rabbit alveolar macrophages. Life Sci 21:1575-1584.
37
38 Mauderly, JL; Banas, DA; Griffith, WC; et al. (1996) Diesel exhaust is not a pulmonary carcinogen in CD-I mice
39 exposed under conditions carcinogenic to F344 rats. Fundam Appl Toxicol 30:233-242.
40
42 carcinogenicity of diesel exhaust. Eighth Health Effects Institute Annual Conference, April 21-24, Colorado .
43 Springs, CO.
44
45 Mauderly, JL; Jones, RK; Griffith, WC; et al. (1987) Diesel exhaust is a pulmonary carcinogen in rats exposed
46 chronically by inhalation. Fundam Appl Toxicol 9:208-221.
47
48 McClellan, R. (1997) Use of mechanistic data in assessing human risks from exposure to particles. Environ Health
49 Perspect 1 05(Suppl 5): 1-17.
50
5 1 McLaughlin, JK; Dietz, MS; Mehl, ES; et al. (1987) Reliability of surrogate information on cigarette smoking by
52 type of informant. Am J Epidemiol 126: 144-146.
53
1 1/5/99 7-150 DRAFT— DO NOT CITE OR QUOTE
-------
Miles, PR; Lee, P; Thrush, MA; et al. (1977) Chemiluminescence associated with phagocytosis of foreign particles
in rabbit alveolar macrophages. Life Sci 20:165-170.
4 Mohr, U; Takenaka, S; Dungworth, DL. (1986) Morphologic effects of inhaled diesel engine exhaust on lungs of
5 rats: comparison with effects of coal oven flue gas mixed with pyrolized pitch. In: Ishinishi, N; Koizumi, A;
6 McClellan, R; StOber, W, eds. Carcinogenic and mutagenic effects of diesel engine exhaust. Amsterdam: Elsevier;
7 pp. 459-470.
8
9 Mokler, BV; Archibeque, FA; Seethe, RJL; et al. (1984) Diesel exhaust exposure system for animal studies. Fundam
10 Appl Toxicol 4:270-277.
11
12 Mumford, JL; Chapman, RS; Harris, DB. (1989) Indoor air exposure to coal and wood combustion emissions
13 associated with a high lung cancer rate in Xuan Wei, China. Environ Int 15:315-320.
14
15 Muscat, JE. (1996) Carcinogenic effects of diesel emissions and lung cancer: the epidemiologic evidence is not
16 causal. JClin Epidemiol 49(8):891-892.
17
18 Muscat, JE; Wynder, EL. (1995) Diesel engine exhaust and lung cancer: an unproven association. Environ Health
19 Perspect 103:812-818.
20
21 Nagashima, M; Kasai, H; Yokota, J; et al. (1995) Formation of an oxidative DNA damage, 8-
22 hydroxydeoxyguanosine, in mouse lung DNA after intratracheal instillation of diesel exhaust particles and effects of
2 3 high-dietary fat and beta-carotene on this process. Carcinogenesis 16:1441 -1445.
24
25 Nesnow, S; Evans, C; Stead, A; Creason, J; Slaga, TJ; Triplett, LL. (1982) Skin carcinogenesis studies of emission
26 extracts. In: Lewtas, J, ed. Toxicological effects of emissions from diesel engines. Amsterdam: Elsevier; pp. 295-
320.
m
29 Nesnow, S; Triplett, LL; Slaga, TJ. (1984) Tumor initiating activities of 1-nitropyrene and its nitrated products in
30 Sencar mice. Cancer Lett 23:1 -8.
31
32 Nielsen, PS; Autrup, H. (1994) Diesel exhaust-related DNA adducts in garage workers. Clin Chem 40:1456-1458.
33
34 Nielsen, PS; Andreassen, A; Farmer, PB.; et al. (1996) Biomonitoring of diesel-exhaust exposed workers. DNA and
35 hemoglobin adducts and urinary 1-hydroxyproline as markers of exposure. Toxicol Lett 86:27-37.
36
37 Nikula, KJ; Snipes, MB; Barr, EB; et al. (1995) Comparative pulmonary toxicities and carcinogenicities of
38 chronically inhaled diesel exhaust and carbon black in F344 rats. Fundam Appl Toxicol 25:80-94.
39
40 Notani, PN; Shah, P; Jayant, K; et al. (1993) Occupation and cancers of the lung and bladder a case-control study
41 in Bombay. Int J Epidemiol 22:185-191.
42
43 OberdQrster, G; Baggs, R; Gelein, R; et al. (1995) Pulmonary effects of inhaled carbon black in rats. Toxicologist
44 15:46.
45
46 OberdSrster, G; Yu, CP. (1990) The carcinogenic potential of inhaled diesel exhaust: a particle effect? J Aerosol Sci
47 21 (suppl. 1):S397-S401.
48
49 Orthoefer, JG; Moore, W; Kraemer, D. (1981) Carcinogenicity of diesel exhaust as tested in strain A mice. Environ
50 Int 5:461-471.
51
Pepelko, WE; Peirano, WB. (1983) Health effects of exposure to diesel engine emissions. J Am Coll Toxicol 2:253-
306.
54
11/5/99 7-151 DRAFT—DO NOT CITE OR QUOTE
-------
1 Percy, C; Stanek, E, III; Gloeckler, L. (1981) Accuracy of cancer death certificates and its effect on cancer mortality
2 statistics. Am J Public Health 71:242-250.
3
4 Pfluger, DH; Minder, CE. (1994) A mortality study of lung cancer among Swiss professional drivers: accounting
5 for the smoking related fraction by a multivariate approach. Soz Prav 39:372-378.
6
7 Phillips, BJ; James, TEB; Anderson, D. (1984) Genetic damage in CHO cells exposed to enzymatically generated
8 active oxygen species. Mutat Res 126:265-271.
9
10 Pott, F; Steber, W. (1983) Carcinogeniciry of airborne combustion products observed in subcutaneous tissue and
11 lungs of laboratory rodents. Environ Health Perspect 47:293-303.
12
1 3 Pott, F; Dungworth, DL; Heinrich, U; et al. (1994) Lung tumours in rats after intratracheal instillation of dusts. Ann
14 Occup Hyg 38:357-363, Suppl. 1.
15
16 Pott, F; Heinrich, U. (1990) Relative significance of different hydrocarbons for the carcinogenic potency of
17 emissions from various incomplete combustion processes. In: Vaino, A; Sorsa, M; McMichael, AJ; eds. Complex
1 8 Mixtures and Scientific Risk. Scientific publication 104. Lyon, France: IARC; pp. 288-297.
19"
20 Qu, SX; Leigh, J; Koelmeyer, H; et al. (1996) Detection of DNA adducts in coal miners exposed to diesel emissions
21 in Australia. Fundam Appl Toxicol 50:237 (abstr).
22
23 Raffle, PAB. (1957) The health of the worker. Br J Ind Med 14:73-80.
24
25 Riebe-Imre, M; Aufderheide, M; Gartner-Hubsch, S; et al. (1994) Cytotoxic and genotoxic effects of insoluble
26 poarticles in vitro. In: Mohr, U; Dungworth, DL; Mauderly, DL; et al., eds. Toxic and carcinogenic effects of solid
27 particles in the respiratory tract. Washington, DC: International Life Sciences Institute Press, pp. 519-523.
28
29 Risen, HA; Burch, JD; Miller, AB; et al. (1988) Occupational factors and the incidence of cancer of the bladder in
30 Canada. Br J Ind Med 45:361-367.
31
32 Rister, M; Baehner, RL. (1977) Effect of hyperoxia on superoxide anion and hydrogen peroxide production of
33 polymorphonuclear leukocytes and alveolar macrophages. Br J Haematol 36:241-248.
34
35 Rittinghausen, S; Mohr, U; Dungworth, DL. (1997) Pulmonary cystic keratinizing squamous cell lesions after
3 6 inhalation/instillation of different particles. Exper Toxicol Pathol 49:433-446.
37
38 Rogot, E; Reid, DD. (1975) The validity of data from next-of-kin in studies of mortality among migrants. Int J
39 Epidemicl 4:51-54.
40
41 R.cct, RK; Metcalf, JA. (1977) H2O2 Release from human granulocytes during phagocytosis. J Clin Invest 60:1266-
42 1279.
43
44 Rosenblatt, MB; Teng, PK; Kerpe, S; et al. (1971) Causes of death in 1,000 consecutive autopsies. NY State J Med
45 71:2189-2193.
46
47 Rosenkranz, HS. (1996) Mutagenic nitroarenes, diesel emissions, particle-induced mutations and cancer an essay
48 on canc37 causation by a moving target Mutel Res 367:65-72.
49
50 Rosenkraoz, HS; Mermelstem; R. (1983) Mutaeenicity and genotoxicity of nitroarenes: all nitro-containing
51 chemicals were not created equal. Mutat Res 114:217-267.
52
11/5/99 7-152 DRAFT—DO NOT CITE OR QUOTE
-------
Rushton, L; Alderson, MR; Nagarajah, CR. (1983) Epidemiological survey of maintenance workers in London
Transport Executive bus garages at Chiswick Works. Br J Ind Med 40:340-345.
4 Sagai, M; Saito, H; Ichinose, T; et al. (1993) Biological effects of diesel exhaust particles. I. In vitro production of
5 superoxide and in vivo toxicity in mouse. Free Radical Biol Med 14:37-47.
6
7 Salmeen, IT; Pero, AM; Zator, R; et al. (1984) Ames assay chromatograms and the identification of mutagens in
8 diesel exhaust particle extracts. Environ Sci Technol 18:375-382.
9
10 Sandusky, CB; Cowden, MW; Schwartz, SL. (1977) Effect of particle size on regurgitive exocytosis by rabbit
11 alveolar macrophages. In: Pulmonary macrophages and epithelial cefls. CONF-760927, National Technical
12 Information Service, U.S. Dept. of Commerce, pp. 85-105.
13
14 Scheepers, PIT; Bos, RP. (1992) Combustion of diesel fuel from a toxicological perspective. II. Toxicity. .Int Arch
15 Occup Environ Health 64:163-177.
16
17 Schenker, MB; Smith, T; Munoz, A; et al. (1984) Diesel exposure and mortality among railway workers: results of a
18 pilot study. Br J Ind Med 41:320-327.
19
20 Schottenfeld, D; Eaton, M; Sommers, SC; et al. (1982) The autopsy as a measure of accuracy of the death
21 certificate. Bull NY Acad Med 58:778-794.
22
23 Shefher, AM; Collins, SC; Dooley, L; et al. (1982) Respiratory carcinogenicity of diesel fuel emissions interim
24 results. In: Lewtas, J, ed. Toxicological effects of emissions from diesel engines. Amsterdam: Elsevier; pp. 329-350.
25
26 Siemiatycki, J; Gerin, M; Stewart, P; et al. (1988) Associations between several sites of cancer and ten types of
exhaust and combustion products: results from a case-referent study in Montreal. Scand J Work Environ Health
14:79-90.
29
30 Silver-man, DT; Hoover, RN; Albert, S; et al. (1983) Occupation and cancer of the lower urinary tract in Detroit. J
31 Natl Cancer Inst 70:237-245.
32
33 Silverman, DT; Hoover, RN; Mason, TJ; et al. (1986) Motor exhaust-related occupations and bladder cancer.
34 Cancer Res 46:2113-2116.
35
36 Steenland, K; Burnett, C; Osoria, AM. (1987) A case-control study of bladder cancer using city directories as a
37 source of occupational data. Am J Epidemiol 126:247-257.
38
39 Steenland, K. (1986) Lung cancer and diesel exhaust: a review. Am J Ind Med 10:177-189.
40
41 Steenland, K; Deddens, J; Stayner, L. (1998) Diesel exhaust and lung cancer in the trucking industry: exposure-
42 response analyses and risk assessment. Amer J Ind Med 34:220-228.
43
44 Steenland, NK; Silverman, DT; Hornung, RW. (1990) Case-control study of lung cancer and truck driving in the
45 Teamsters Union. Am J Public Health 80:670-674.
46
47 Steineck, G; Plato, N; Gerhardsson, M; et al. (1990) Increased risk of urothelial cancer in Stockholm during 1985-
48 87 after exposure to benzene and exhausts. Int J Cancer 45:1012-1017.
49
50 Stenback, F; Rowland, J; Sellakumar, A. (1976) Carcinogenicity of benzo[a]pyrene and dusts in the hamster lung
51 (instilled intratracheally with titanium oxide, aluminum oxide, carbon and ferric oxide). Oncology 33:29-34.
Stenberg, U; Alsberg, T; Westerholm, R. (1983) Emission of carcinogenic components with automobile exhausts.
54 Environ Health Perspect 47:53-63.
11/5/99 7-153 DRAFT—DO NOT CITE OR QUOTE
-------
1 Stern, FB; Lemen, RA; Curtis, RA. (1981) Exposure of motor vehicle examiners to carbon monoxide: a historical
2 prospective mortality study. Arch Environ Health 36:59-66.
3
4 Stober, W; Abel UR. (1996) Lung cancer due to diesel soot particles in ambient air? A critical appraisal of
5 epidemiological studies addressing this question. Int Arch Occup Environ Health 68(Suppl):S3-S6 1 .
6
7 StSber, W. (1986) Experimental induction of tumors in hamsters, mice and rats after long-term inhalation of filtered
8 and unfiltered diesel engine exhaust. In: Ishinishi, N; Koizumi, A; McClellan, R; Steber, W, eds. Carcinogenic and
9 mutagenic effects of diesel engine exhaust. Amsterdam: Elsevier; pp. 421-429.
10
1 1 Strom, KA. (1984) Response of pulmonary cellular defenses to the inhalation of high concentrations of diesel
1 2 exhaust. J Toxicol Environ Health 1 3 :9 1 9-944.
13
1 4 Swanson, GM; Lin, CS; Bums, PB. (1993) Diversity in the association between occupation and lung cancer among
1 5 black and white men. Cancer Epidemiol Biomark Prev 2:3 13-320.
16
1 7 Takaki, Y; Kitamura, S; Kuwabara, N; et al. (1989) Long-term inhalation studies of exhaust from diesel engine in F-
1 8 344 rats: the quantitative relationship between pulmonary hyperplasia and anthracosis. Exp Pathol 37:56-6 1 .
19
20 Takemoto, K; Yoshimura, H; Katayama, H. (1986) Effects of chronic inhalation exposure to diesel exhaust on the
2 1 development of lung tumors in di-isopropanol-nitrosamine-treated F344 rats and newborn C57BL and ICR mice. In:
22 Ishinishi, N; Koizumi, A; McClellan, R; StSber, W, eds. Carcinogenic and mutagenic effects of diesel engine
23 exhaust. Amsterdam: Elsevier; pp. 3 1 1-327.
24
25 Takemoto, K; Katayama, H; Kuwabara, T; et al. (1988) Carcinogenicity by subcutaneous administration of diesel
26 paniculate extracts in mice. In: Diesel exhaust and health risks. Ibaraki, Japan: Research Committee for HERP
27 Studies; pp. 227-234.
28
29 Thyssen, J; Althoff, J; Kimmerle, G; et al. (1981) Inhalation studies with benzo[a]pyrene in Syrian golden hamsters.
30 J Natl Cancer Inst 66:575-577.
31
32 Trotter, JR. (1980) Spontaneous cancer and its possible relationship to oxygen metabolism. Proc Natl Acad Sci
33 77:1763-1767.
34
35 U.S. Environmental Protection Agency. (1993) Motor vehicle-related air toxics study. Ann Arbor, MI: Office of
36 Mobile Sources. EPA/420/R-93/005. Available from National Technical Information Service, Springfield, VA,
37 PB93-182590/XAB.
38
39 U.S. Environmental Protection Agency.. (1996) Air Quality Criteria for Particulate Matter. Prepared by the National
40 Center for Environmental Assessment, U.S. Environmental Protection Agency, Research Triangle Park, NC.
41 EPA/600/P-95/001bF.
42
44 606.
45
46 Vostal, JJ. (1986) Factors limiting the evidence for chemical carcinogenicity of diesel emissions in long-term
47 inhalation experiments. In: Ishinishi, N; Koizumi, A; McCleiian, RO; StSber, W, eds. Carcinogenic and niutaganic
48 effects of diesel engine exhaust: proceedings of the international satellite symposium on lexicological effects of
49 emissions from diesel engines; July 1986; Tsukuba Science City, Japan. (Developments in toxicology and
50 environmental science: v. 13.) Amsterdam: Elsevier Science Publishers BV; pp. 381-396.
51
R ? \W»inct*»in TT%- TVrtll ^V f 1 Q*77A XTotiori-al /"'on/* AT- TnotitiitA tir/M-UeUrw* r»r» timi<-t«* rnvrtmn+irtii onH /•
-------
Waller, RE. (1981) Trends in lung cancer in London in relation to exposure to diesel fumes. Environ Int 5:479-483.
Ward, JM. (1983) Background data and variations in tumor rates of control rats and mice. Prog Exper Tumor Res
26:241-258.
5
6 Waxweiler, RJ; Wagoner, JK; Archer, VE. (1973) Mortality of potash workers. J Occup Med 15:486-489.
7
8 Weiss, SJ; Rustagei, PK; LoBuglio, AF. (1978) Human granulocyte generation of hydroxyl radical. J Exp Med
9 147:316-323.
10
11 Weitzman, SA; Stossel, TP. (1981) Mutation caused by human phagocytosis. Science 212:546-547.
12
13 White, HJ; Vostal, JJ; Kaplan, HL; et al. (1983) A long-term inhalation study evaluates the pulmonary effects of
14 diesel emissions. J Appl Toxicol 1:332.
15
16 Williams, RR; Stegens, NL; Goldsmith, JR. (1977) Associations of cancer site and type with occupation and
17 industry from the Third National Cancer Survey interview. J Natl Cancer Inst 59:1147-1185.
18
19 Wong, O; Morgan, RW; Kheifets, L; et al. (1985) Mortality among members of a heavy construction equipment
20 operators union with potential exposure to diesel exhaust emissions. Br J Ind Med 42:435-448.
21
22 Wong, D; Mitchell, CE; Wolff, RK; et al. (1986) Identification of DNA damage as a result of exposure of rats to
23 diesel engine exhaust. Carcinogenesis 7:1595-1597.
24
25 Woskie, SR; Smith, TJ; Hammond, SK; et al. (1988a) Estimation of the diesel exhaust exposures of railroad
26 workers: I. Current exposures. Am J Ind Med 13:381-394.
Woskie, SR; Smith, TJ; Hammond, SK; et al. (1988b) Estimation of the diesel exhaust exposures of railroad
workers: II. National and historical exposures. Am J Ind Med 13:395-404.
30
31 Wynder, LE; Dieck, G; Hall, NE. (1985) A case control study of diesel exhaust exposure and bladder cancer.
32 Environ Res 34:475-489.
33
34 Zamora, PO; Gregory, RE; Brooks, AL. (1983) In vitro evaluation of the tumor-promoting potential of diesel-
35 exhaust-particle extracts. J Toxicol Environ Health 11:188-197.
11/5/99 7-155 DRAFT—DO NOT CITE OR QUOTE
-------
8. CANCER DOSE-RESPONSE EVALUATION
8.1. INTRODUCTION
2 This chapter discusses the dose-response aspects of the key carcinogenicity data. The
3 identification of a dose-response can enhance the understanding of the cancer hazard and lead to
4 an estimation of possible disease impact. One measure of impact used by EPA is the cancer unit
5 risk. Unit risk is the estimated cancer risk at 1 ng/m3 of exposure, in this case ug/m3 of diesel
6 exhaust (DE) particulate matter, from a continuous 70-year exposure. Unit risk derivation
7 procedures and specifications are defined in EPA's risk assessment guidance (U.S. EPA, 1986,
8 1996).
9 Evidence shows that DE is likely to pose a human hazard for lung cancer by the
10 inhalation route of exposure. The mode of action (MoA) for humans has not been determined,
11. and the presumed MoA for rats does not justify using these data to estimate cancer unit risk for
12 humans. EPA believes that a role for both mutagenic effects and particle-specific effects is
13 plausible. According to EPA Cancer Guidelines, the mutagenic effects would allow the
14 modeling of dose-response using models with a linear term at low doses. With lexicologically
15 suspect organics thought to be in proportion to the mass of particulates, the use of |ig/m3 of DE
16 particles as the dosimeter is feasible. With no clear indication that key organic components have
^B changed disproportionately to the particle mass over the years (note that the overall particle mass
18 has been decreasing), the use of older toxicological results based on older engine exposures to
19 predict current-day hazards is also feasible, though uncertainty exists in this assumption. The
20 overall challenge with DE is to judge the uncertainties going into the dose-response analysis and
21 decide whether to proceed and, if so, what certainties/uncertainties to ascribe to any resulting
22 analysis and follow-on unit risk derivation. For DE, the cascade of desirable to less desirable
23 approaches clearly starts with human data, then to comparative potency approaches that use
24 various surrogate exposure-response relationships.
25 For a variety of reasons EPA is not, at this time, adopting or recommending a cancer unit
26 risk or risk range for DE. EPA will monitor ongoing research and reanalysis of epidemiology- •
27 exposure studies and may revisit dose-response and unit risk derivation.
28
29 8.2. REVIEW OF PREVIOUS QUANTITATIVE RISK ESTIMATES
30 Early attempts to quantitatively assess the carcinogenicity of diesel engine emissions were
31 hindered by a lack of positive epidemic logic studies and long-term animal studies. One means of
32 overcoming these obstacles was the use of the so-called comparative potency method (Albert et
tal., 1983). An attempt to estimate risk based on human data was also made at this tune by Harris
(1983), although it was based upon equivocal evidence. By the late 1980s the availability of data
11/5/99 8-1 DRAFT—DO NOT CITE OR QUOTE
-------
1 from animal bioassays and epidemiologic studies provided an opportunity to the derivation of
2 both animal and human data-based estimates. See Table 8-1 for a historical overview.
3
4 8.2.1. Comparative Potency Method
5 In the comparative potency method, a combustion or pyrolysis product is selected that has
6 a previously determined cancer potency estimate based on epidemiologic data. The ratios of the
7 potency of this agent (e.g., coke oven emissions) to diesel particulate matter (DPM) extract in a
8 variety of in vivo and in vitro tests are then multiplied by the epidemiology-based potency
9 estimate for coke oven emissions and averaged. If epidemiology-based estimates from more than
10 one pollutant are used, the derived potencies are generally averaged to obtain an overall mean.
11 The comparative potency estimate of Albert et al. (1983) is probably the best known.
12 Their results were obtained using epidemiology-based unit cancer risk estimates for coke oven
13 emissions, cigarette smoke condensate, and roofing tar. Samples of particulate matter were
14 collected from three light-duty engines (a Nissan 220 C, an Oldsmobile 350, and a Volkswagen
15 turbocharged Rabbit), all run on a highway fuel economy test cycle, and from a heavy-duty
16 engine (Caterpillar 3304) run under steady-state, low-load conditions. The particulate matter was
17 extracted with dichloromethane and tested in a variety of assays. Dose-dependent increases in
18 response were obtained for the four assays listed below:
19 • Ames Salmonella typhimurium (TA98) reverse mutation,
20 • Gene mutation in L5178Y mouse lymphoma cells,
21 • Sencar mouse skin tumor initiation test, and
22 • Viral enhancement of chemical transformation in Syrian hamster embryo cells.
23 Only the first three assays were used to develop comparative potency estimates because
24 of variability of responses in the enhancement of the viral transformation assay. The in vitro
25 studies were carried out both in the presence and absence of metabolic activators. The potency,
26 defined as the slope of the dose-response curve, was measured for each sample in each short-term
27 assay.
28 . The skin tumor initiation test was positive for all the engines tested except the Caterpillar
29 engine. Only the Nissan engine, however, gave strong dose-response data. Because skin tumor
30 initiation was considered to be the most biologically relevant test, it was used to derive potency
31 estimates for the Nissan engine. An estimate for the Nissan engine was then derived by
32 multiplying the epidemiology-based potency estimates for each of the three agents (coke oven
33 emissions, roofing tar, and cigarette smoke condensate) by the ratios of their potencies in the skin
34 tumor initiation test to that of the Nissan diesel engine. .According to this method, three 95%
35 urmer-bnunH e<5timatp«s of lifrtrme cancer risk per microgrsm per cubic mster cf sxtrsctsbls
11/5/99 8-2 DRAFT—DO NOT CITE OR QUOTE
-------
Table 8-1. Estimated 95% upper confidence limits of the lifetime risk of
cancer from inhalation of 1 ug/m3 diesel participate matter (DPM)
Method
Potency
Comments
Reference
Comparative potency
Comparative potency
Comparative potency
Comparative potency
Multistage model
Straight-line
extrapolation
Time-to-tumor model
Logistic regression
Multistage model
Armitage-Doll model
Multistage model
Multistage model
Biological model
Biological model
Epidemiologic analysis
Epidemiologic analysis
Epidemiologic analysis
Epidemiologic analysis
3.5 x 10'5
2.6 x ID"5
7.0 x lO'5
6.8 x 1Q-4
1.6x ID'5
Nissan engine
Average of 3 engines
Light-duty engines
Average of 3 engines
Lung cancer rats"
6-12 x 10'5 Lung cancer ratsb
2-3 x 10'5 Lung cancer ratsa
8 x 10"5 Lung cancer rats0
1.4 x 10~5 Lung cancer ratsd
5.2 x 10'5 Lung cancer rats8-6
8.9 x 10~5 Lung cancer ratsf
3.4 x 10'5 Lung cancer ratsd
1.7 x 10'5 Lung cancer ratsd
3.5 x lO'6
Assuming particle
threshold
1.4 x 10~3 London transport study
6 x 10"4 Railroad workers
0.6-2 x 1Q-3 Railroad workers
1.6 x 10'2 Truck drivers8
Albert etal., 1983
Albert etal., 1983
Cuddihy et al., 1984
Harris, 1983
Albert and Chen, 1986
Pott and Heinrich, 1987
Smith and Stayner, 1990
McClellan etal., 1989
Pepelko and Chen, 1993
Hattis and Silver, 1994
CAL-EPA, 1998
WHO, 1996
Chen and Oberdorster,
1996
Chen and Oberdorster,
1996
Harris, 1983
CAL-EPA, 1998
U.S. EPA, 1998
Steenland et al., 1998
"Used data from studies by Mauderly et al., 1987.
"Used data from studies by Brightwell et al., 1989; Heinrich et al., 1986a; and Mauderly et al., 1987.
GUsed data from studies by Brightwell et al., 1989; Ishinishi et al., 1986; Iwai et al., 1986; and Mauderly et al., 1987.
dUsed data from studies by Brightwell et al., 1989; Ishinishi et al., 1986; and Mauderly et al., 1987.
"Maximum likelihood estimate based on 53 years of exposure, 8 hours/day, 240 days/year.
tJsed data from studies by Brightwell, et al., 1989; Heinrich et al., 1995; Ishinishi et al., 1986; Mauderly, 1987; and
Nikula et al., 1995.
Estimated risk of 45 years occupational exposure to 5 ug/m3.
11/5/99
8-3
DRAFT—DO NOT CITE OR QUOTE
-------
1 organic matter were derived for the Nissan diesel, based on potency comparisons with each of
2 the three agents. These values are: coke oven emissions, 2.6 * 10"4; roofing tar, 5.2 * 10"4; and
3 cigarette smoke condensate, 5.4 x 10"*. The average of the three equals 4.4 x 10"4.
4 The potency of the other diesel emission samples was not estimated directly because of
5 the weak response in the skin tumor initiation test. Instead, their potency relative to the Nissan
6 engine was estimated as the arithmetic mean of their potency relative to the Nissan in the
7 Salmonella assay in strain TA98, the sister chromatid exchange assay in Chinese hamster ovary
8 cells, and the mutation assay in mouse lymphoma cells. The estimated lifetime cancer risk per
9 microgram per cubic meter of extractable organic matter for extracts from these engines are as
10 follows: Volkswagen, 1.3 x 1Q-4; Oldsmobile, 1.2 x 1Q-4; and Caterpillar, 6.6 x 10'6.
11 To convert these values to a lifetime risk per microgram per cubic meter of total DPM,
12 the results were multiplied by the fraction of extractable organic matter in the particles. This
13 conversion was based on the assumption that the carcinogenic effects were caused solely by the
14 organic fraction. These fractions were as follows: Nissan, 0.08; Volkswagen, 0.18; Oldsmobile,
15 0.17; and Caterpillar, 0.27. After this adjustment, the resulting estimated potencies per
16 microgram per cubic meter of inhaled DPM varied from 3.5 x l Q'5 for the Nissan to 1.8 x 10"6 for
17 the Caterpillar.
18 Harris (1983) developed comparative potency estimates for the same four engines used by
19 Albert et al. (1983) but used only two epidemiology-based potency estimates: those for coke
20 oven emissions and for roofing tar. He employed preliminary data from three of the same assays
21 used by Albert et al. (1983): the Sencar mouse skin tumor initiation assay, enhancement of viral
22 transformation in Syrian hamster embryo cells, and the L5178 mouse lymphoma test. The mouse
23 lymphoma test was used both with and without metabolic activation, whereas the Salmonella
24 assay was not used.
25 The diesel cancer potency estimates by Harris (1983) were then derived by multiplying
26 the epidemiology-based cancer potency estimates for both coke oven emissions and roofing tar
27 by the ratio of their potencies compared with DE particles in each of the four bioassays. For
28 example, the epidemiology-based relative risk of exposure to 1 ng/m3 of coke oven emissions
29 was estimated to equal 4.4 x 10"4. In the skin tumor initiation test, 2.1 papillomas per mouse
30 were reported for the coke oven sample, compared with 0.53 for the Nissan engine extract. The
31 benzene-extractable fraction was assumed to equal 0.06 (slightly less than that in the Albert et al.
32 [1983] studies). The diesel potency estimate using this comparison is then equal to (0.53/2.1) x
33 0.06 x 4.4 x 10"*/ug/m3, or 6.6 x 10^/ug/m3 DPM. A total of eight comparisons were made for
34 each engine, four bioassays times two epidemiology-based potency estimates.
35 The Harris (1983) estimates are not comparable to those of Albert et al, (1983) without
36 adjustment. The unit risk estimates of Albert and co-workers are based on absolute risk during
11/5/99 8-4 DRAFT—DO NOT CITE OR QUOTE
-------
1 lifetime exposure, whereas Harris reported his values in terms of relative risk per year of
2 exposure. To adjust this to lifetime risk for continuous exposure, it is necessary to multiply
4fc Hams' values by a factor of 2.7 = (70 x 0.039), where 70 reflects the lifetime exposure (70 years)
4 and 0.039 is the lifetime lung cancer mortality rate in the U.S. population.
5 The range of potencies varied from 0.2 x 10'5 to 0.6 x 10"s for the Nissan sample,
6 0.1 x 10'5 to 2.4 x 10'5 for the Oldsmobile 350, 0.2 x lQ-5to 27.8 x 10'5 for the Volkswagen
7 Rabbit, and 0.1 x IQ'5 to 2.5 x lO'V^g/m3 DPM for the Caterpillar sample. Harris (1983) derived
8 an overall mean relative risk value of 3.5 x 10~5/ug/m3 for the three light-duty engines with a 95%
9 upper confidence limit of 2.5 x 10"4. Individual mean values for each engine were not reported.
10 After multiplying by 2.7 to convert to a unit risk, the upper-bound estimate of potency from the
11 tliree light-duty engines was equal to 6.8 x lO^/ug/m3 DPM. McClellan (1986), Cuddihy et al.
12 (1981,1984), and Cuddihy and McClellan (1983) estimated a risk of about 7.0x1O'Vug/m3
13 DPM using a comparative potency method similar to those reported in the preceding paragraph.
14 The database was similar to that used by Albert et al. (1983) and Harris (1983). This estimate
15 agrees quite well with those reported by Albert et al. (1983). Although the Harris (1983)
16 estimate is somewhat greater, it should be remembered that it was based on preliminary data.
17
18 8.2.2. Suitability of Comparative Potency Approach
As noted earlier, in this method the potency of DPM extract is compared with other
combustion or pyrolysis products, for which epidemiology-based unit risk estimates have been
21 developed. Comparisons are made using short-term tests such as skin painting, mutations, and
22 mammalian cell transformation. The ratio of the potency of DPM extracts to each of these agents
23 is then multiplied by their unit risk estimates to obtain the unit risk for DE.
24 Although this test was based originally on the belief that cancer induction at low doses is
25 due to the organic fraction present on the diesel particles, it is possible to argue, through a
26 biologically based dose-response modeling concept, that the relative cancer risk of two
27 compounds is approximately equal to the ratio of initiation rate of the two compounds at low
28 doses even though particles may assert other effects at higher doses; thus, making it a reasonable
29 approach for risk derivation. A major strength of this approach is avoidance of lung particle
30 overload effects. Furthermore, independent tests have shown that the organic fraction of DE may
31 damage DNA and thus may induce cancer (see Chapter 7). Finally, the carcinogenic potency of
32 the organic fraction can be compared with related emissions for which cancer potency is
33 reasonably well defined.
34 A major uncertainty of this approach is the assumption that cancer potency can be
35 determined on the basis of the effectiveness of the organic fraction alone. Under lung particle
overload conditions, particles are considered to play a primary role in lung cancer induction. As
1175/99 8-5 DRAFT—DO NOT CITE OR QUOTE
-------
1 noted in Chapter 4, ultrafine diesel particles may be ingested by epithelial cells even at low
2 concentrations, inducing damage to the genetic material and possible carcinogenic effects. The
3 potency estimates vising this approach may therefore underestimate risk by not accounting for
4 possible effects of particles or even reactive oxygen species. Association of organics with
5 particles may also influence their potency depending on relative elution rates, efficiency of
6 activation, etc. A final uncertainty involves the assumption that potency in short-term tests is an
7 accurate predictor of lung cancer potency.
8 The uncertainties of this approach preclude its unilateral adoption for predicting upper-
9 bound estimates.
10
11 8.2.3. Animal Bioassay-Based Cancer Potency Estimates
12 With the availability of chronic cancer bioassays, a considerable number of potency
13 estimates were derived using lung tumor induction in rats. Albert and Chen (1986) reported a
14 risk estimate based on the chronic rat bioassay conducted by Mauderly et al. (1987). Using a
15 multistage model and assuming equivalent deposition efficiency in humans and rats, they derived
16 a 95% upper confidence limit of 1.6 * 10"s for lifetime risk of exposure to 1 jag/m3. Pott and
17 Heinrich (1987) used a linear extrapolation for data reported by Brightwell et al. (1989), Heinrich
18 et al. (1986a), and Mauderly et al. (1987). They reported risk estimates of 6 x lO'5 to 12 x 10' .
19 5/u.g/m3. More recently, Smith and Stayner (1990), using time-to-tumor models based on the data
20 of Mauderly et al. (1987), derived 95% upper confidence limits ranging from 1.5 x 10"5 to 3 * 10"
21 5/u.g/m3. Pepelko and Chen (1993) developed unit risk estimates based on the data of Brightwell
22 et al. (1989), Ishinishi et al. (1986), and Mauderly et al. (1987) using a detailed dosimetry model
23 to extrapolate dose to humans and a linearized multistage (LMS) model. Taking the geometric
24 mean of individual estimates from the three bioassays, they derived unit risk estimates of 1.4 x
25 10~5/ug/m3 when dose was based on carbon particulate matter per unit lung surface area rather
26 than whole DPM, and 1.2 x 10'4/|ig/m3 when based on lung burden per unit body weight. Hattis
27 and Silver (1994) derived a maximum likelihood estimate for occupational exposure of 5.2 x
28 10"5/jag/m3 based on lung burden and bioassay data reported by Mauderly et al. (1987) and use of
29 a five-stage Armitage-Doll low-dose extrapolation model. The EPA (1998) derived a unit risk
30 estimate of 3.4 X 10"5/ug/m3, based on lung burden of DPM per unit lung surface area, using an
31 LMS model and calculating the geometric mean from results of bioassay data reported by
32 Mauderly et al. (1987), Ishinishi et al. (1986), and Brightwell et al. (1989). California EPA
33 (OEHHA, 1998) derived a geometric mean estimate of 6 x 10"5/ng/m3 from five bioassays .using
34 an LMS model.
35 In an attempt to demonstrate the possible influence cf particle effects as well as particle-
36 associated organics, an additional modeling approach was attempted by Chen and Oberdorster
1175/99 8-6 DRAFT—DO NOT CITE OR QUOTE
-------
1 (1996). Employing a biologically based two-stage model and using malignant tumor data from
2 Mauderly et al. (1997), the upper bound risk estimate for exposure to 1 ug/m3 was estimated to
^P be 1.7 x 10"5. This estimate is virtually identical to that using the LMS model, assuming
4 nonthreshold effect of particles. If a threshold of particle effect is assumed, however, the
5 estimated risk decreases about fivefold. The results also show that the mechanism of diesel-
Q induced lung tumor at high exposure concentrations may differ from that at low exposure
7 concentrations, with organics and particles playing primary roles of tumorigenesis respectively at
8 low and high concentrations.
9
10 8.2.4. Suitability of Laboratory Animal Bioassay Approach
11 Cancer risk assessment from exposure to DE, based on available animal bioassays,
12 traditionally has strengths and uncertainties. For DE the best studies are adequately designed,
13 eliminating confounding factors often present in epidemiology studies. Exposure duration and
14 exposure levels can be precisely controlled and monitored. The presence or absence of tumors
15 can be verified by pathological evaluation. Although animal-to-human extrapolation of dose is
16 required and has uncertainty, the development of dosimetry models has eliminated much of the
17 uncertainty in this area. Nevertheless, two important uncertainties remain: the adequacy of the
18 rat as a model for evaluating human risk of cancer from exposure to DE and the shape of the
^fc dose-response curve.
20 It is believed by a consensus of experts that the rat seems to be unique in its response to
21 particulate matter, and therefore its use for assessing human lung cancer risk is problematic
22 (ILSI, 1998; Mauderly, 1994). As noted in Chapter 7, the rat is the only species that has
23 unequivocally been shown to develop lung cancer in response to inhaled DE. However, what is
24 happening in the human lung is uncertain. It has also been argued that humans are more resistant
25 to particle-induced lung cancer; although coal miners develop pneumoconiosis, lung cancer
26 seldom occurs. Rats, on the other hand, were reported to develop lung cancer in response to coal
27 dust (Martin et al., 1977), though this study was poorly described and the number of animals
28 exposed was small (4/36 developed lung cancer). Moreover, exposure levels were very high and
29 lung burdens were greater than those generally encountered in coal miners (Mauderly, 1994).
30 Although lung cancer has not been reported in most epidemiology studies of coal miners, Zhong
31 and Dehong (1995) reported that Chinese workers suffering coal miners' pneumoconiosis did
32 have an increased risk of lung cancer.
33 Although rat data may still have limited value for hazard identification they are much less
34 suitable for quantitating human environmental risk. For example, particle deposition patterns are
35 different in the rat and human. Because of the absence of respiratory bronchioles in the rat, a
^P greater fraction of inhaled particles deposit in the alveolar regions; in primates, deposition occurs
11/5/99 8-7 DRAFT—DO NOT CITE OR QUOTE
-------
1 to a larger extent at the bifurcation of the small bronchi. Differing deposition patterns are likely
2 to result in different pathologic responses, as reported by Nikula et al. (1997) for rats and
3 monkeys.
4 Another major uncertainty in the use of rat bioassay data concerns extrapolation of lung
5 cancer to ambient concentrations. Significant lung rumor increases hi experimental animals have
6 generally been obtained only at concentrations resulting in lung particle overload with
7 concomitant pathological effects. As discussed in Section 7.4, it has been hypothesized that lung
8 cancer induction results from a secondary effect associated with release of various inflammatory
9 mediators by particle-overloaded phagocytic cells. The resultant inflammatory response, with
10 accompanying cell division, can increase the likelihood that any oxidant-induced or
11 spontaneously occurring genetic damage becomes fixed hi a dividing cell and is clonally
12 expanded (Driscoll, 1995).
13 If the primary means of lung cancer induction in rats is via particle-overload mechanisms,
14 then it can be surmised that different factors are plausibly responsible for induction of lung
15 cancer in humans exposed at occupational or ambient concentrations. Experimental evidence
16 provides some support for the existence of low-dose mechanisms. Riebe-Imre et al. (1994)
17 reported that carbon black is taken up by lung epithelial cells in vitro, inducing chromosomal
18 damage and disruption of the cytoskeleton (lesions that closely resemble those in tumor cells) at
19 concentrations that did not induce measurable toxicity. Ichinose (1997a, b) reported that not
20 only are reactive oxygen species generated from organics present on the surface of diesel
21 particles, but the production of these radicals is well correlated with increased in 8-
22 hydroxydeoxyguanine adducts. Finally, Dasenbrock et al. (1996) reported that extraction of the
23 organic fraction from diesel particles decreased their carcinogenic potency, suggesting a role for
24 organic constituents. Because the primary modes of cancer induction are likely to differ as
25 exposure concentrations decrease below those required to induce lung particle overload, the slope
26 of the dose-response curve is also likely to change. Since a change in slope at low doses cannot
27 be determined from available bioassay data, low-dose extrapolation results hi a considerable
2 8 degree of uncertainly.
29 In summary, the use of rat data to quantitate human cancer risk at environmental
30 exposures is not recommended.
31
32 8.2.5. Epidemiology-Based Estimation of Cancer Potency
33 The first lung cancer risk estimates based on epidemiologic data were derived by Harris
34 (1983). He assessed the risk of exposure to diesel engine emissions using data from the London
35 TVflTKSnnrt Wrvrkpr £tuHv rf»TWrt<»H Kv Wa11*>r (\ Q81 "V Pi\/<» rrrrmr»o nf o«i»i1/v«/ooo £rr>m «-Ko T /-n-.^/-.»i
4 j —x — j • —• v — ~ -/• ---— o~ ~ —f " *"" --*•»-£•'*vy —»— «-~- * fc*-^- *wv».f*»v
36 Transport Authority (LTA) were used: bus garage engineers, bus drivers, bus conductors,
11/5/99 8-8 DRAFT—DO NOT CITE OR QUOTE
-------
1 engineers in central works, and motormen and guards. The first group was considered to have
2 received the highest exposure; the next two, intermediate; and the last two, none. When cancer
^P death rates for the high-exposure group were compared with those of London males, there was no
4 increase in the observed-to-expected (O/E) ratios. The author, in fact, considered the results to
5 be negative. However, because the low rate of lung cancer in all the LTA exposure groups may
6 have been the result of a "healthy worker" effect, Harris (1983) compared the exposed groups
7 with internal controls. He merged the three exposed groups and compared them with the two
8 groups considered to be unexposed. An adjustment was made for the estimated greater exposure
9 levels of garage engineers compared with bus drivers and conductors. Using this method, the
10 relative risk of the exposed groups was greater than 1 but was statistically significant only for
11 garage engineers exposed from 1950 to 1960. In this case, the O/E ratio was 29% greater than
12 the presumed unexposed controls.
13 Harris (1983) identified a variety of uncertainties relative to potency assessment based on
14 this study. These included:
15 • Small unobserved differences in smoking incidences among groups, which could have
16 a significant effect on lung cancer rates;
17 • Uncertainty about the magnitude of exposure in the exposed groups;
18 • Uncertainty regarding the extent of change in exposure conditions over time;
^fc • Random effects arising from the stochastic nature of the cancer incidence; and
20 • Uncertainty in the mathematical specification of the model.
21 Taking the uncertainties into account, he derived a maximum likelihood excess relative
22 risk estimate of 1.23 x 10^ with a 95% upper confidence limit of 5 x 1 O^/ug/m3 DPM per year.
23 These estimates are equal to 5 x 10"4 and 2 x 10"3, respectively, when converted to an absolute
24 risk for lifetime exposure to 1 |ag/m3 particulate matter. It should be noted that, because of the
25 high degree of uncertainty, the 95% lower confidence limit would predict no risk.
26 McClellan et al. (1989) reported risk estimates based on the Garshick et al. (1987) case-
27 control study in which lung cancer in railroad workers was evaluated. Using a logistic
28 regression, the expected relative risk of lung cancer death was estimated to rise 0.016 per year of
29 exposure to DE. Adjustments were made to convert to continuous exposure (168 vs. 40 hours)
30 for 70 years. Because exposure levels could not be defined exactly, two sets of calculations were
31 made, assuming inhaled DPM concentrations of either 500 or 125 |ig/m3 DPM. Using a 95%
32 upper confidence limit, the number of excess cancer deaths per year in the United States was
33 estimated to range from 1900 to 7400. Employing these values, the lifetime 95% upper
34 confidence limits of the risk from exposure to 1 ug/m3 DE can be derived, by dividing the
estimated excess number of annual cancer deaths in the United States by the total population and
11 /5/99 8-9 DRAFT—DO NOT CITE OR QUOTE
-------
1 multiplying by 70, the estimated mean lifespan. Based upon the exposure estimates of 500 or
2 125 ug/m3 DPM, unit risks of 0.6 x 10'3 and 2 * 10'3/ug/m3 were derived. Even using the 95%
3 lower confidence limits, an excess of 1 00 to 400 deaths is predicted, unlike the Harris (1983)
4 study in which no excess deaths could be predicted based on the lower confidence limit.
5 California EPA (1 998) derived unit risk estimates for lung cancer, based upon the
6 Garshick et al. (1987) case-control study and the Garshick et al. (1988) cohort study of U.S.
7 railroad workers. A variety of exposure patterns were considered, characterized by two
8 components: the average exposure concentration for the workers as measured by Woskie et al.
9 (1988) and the extent of change from 1959 to 1980. The lowest lifetime risk estimate derived
10 was 1.3 x lO^/ug/m3 and the highest was 2.4 * 10~3/ug/m3. The geometric mean was 6 x 10'
1 1 4/ug/m3.
1 2 Steenland et al. ( 1 998) estimated lung cancer risk of truck drivers on the basis of a case-
13 control study of decedents in the Teamsters Union (Steenland et al., 1990). Retrospective
14 exposure estimates were made starting with a set of 1991 exposure measurements for different
1 5 job categories and then retrospectively estimating from 1982 to about 1950 using various factors,
16 including diesel vehicle miles traveled and DE engine emission rates per mile. The 1991 job
17 category estimates came from an extensive industrial hygiene survey of elemental carbon (EC)
18 exposures in the trucking industry by Zaebst et al. (1991). Lifetime (through age 75) excess risk
19 of lung cancer death for male truck drivers was calculated with the aid of a cumulative exposure
20 model. Assuming a most likely emissions scenario of 4.5 gm/mile in 1970, and a 45-year
21 exposure to 5 ug/m3 of EC beginning at age 20 and ending at age 65, the estimated excess lung
22 cancer risk was determined to be 1 .6% (95% CI 0.4%-3 . 1 %).
23
24 $.2.6. Suitability of Using Epidemiologic Data
25 A major advantage in the use of human data is the elimination of uncertainty due to
26 possible differences in sensitivity to cancer induction by DE among species. Second,
27 epidemiology studies are based on occupational exposures, which generally occur at
2S concentrations insufficient to result in lunCT "article overload. Thus, lung cancer in the human
29 studies is likely to be induced by non-particle-overload mechanisms (at least as defined in the rat
30 studies) under either occupational or ambient exposure levels. Uncertainty hi extrapolating risk
31 from occupational studies is therefore decreased, not only because low-dose extrapolation occurs
32 over a smaller range, but because mechanisms of cancer induction are less likely to vary within
33 this range with accompanying changes in the dose-response curve.
34 There is considerable evidence for nonoverload mechanisms of cancer induction by
11/5/99 8-10 DRAFT— DO NOT CITE OR QUOTE
-------
1 concentration of P AHs but lacks an insoluble carbon core. Increased levels of aromatic DN A
2 mortality in Chinese communes burning so-called smoky coal containing high concentrations of
^P polycyclic aromatic hydrocarbons (PAHs). Demonstration of the carcinogenicity of coke oven
4 emissions in humans (Lloyd, 1971) also provided evidence for a role by organics because coke
5 oven PM contains a high adducts were reported in bus maintenance and terminal workers by
6 Hemminki et al. (1994) and in garage workers and mechanics exposed to DE (Nielsen and
7 Autrup, 1994). Studies by Sagai et al. (1993) have indicated that DPM could produce superoxide
8 and hydroxyl radicals in vitro without any biologically activating systems. On the basis of these
9 findings, they suggested that most DE toxicity in lungs is due to active oxygen radicals. In a
10 more recent study, these investigators reported that instillation of only 0.1 mg of DPM into
11 mouse lungs resulted in the production of 8-hydroxyguanosine in lung cell DNA. The critical
12 lesion may thus be induced by oxygen free-radicals generated from DPM (Nagashima et al.,
13 1995).
14 An uncertainty associated with most of the diesel epidemiology studies was the inability
15 to eliminate all confounding factors, resulting in possible errors in estimating relative risk ratios.
16 Small errors in adjustment for smoking, for example, can result in considerable error because
17 smoking has a much larger effect on relative cancer risk than is likely for DE. The likelihood of
18 significant confounding errors hi the Garshick et al. (1987, 1988) studies is decreased by the
l^fc considerable effort exerted to eliminate or reduce such factors, especially smoking. Moreover,
20 meta-analyses by Bhatia et al. (1998) and Lipsett et al. (1999) using a number of diesel
21 epidemiology studies resulted in relative risk ratios quite similar to the one reported by Garshick
22 et al. (1987). Although exposure levels are likely to have differed somewhat among studies, the
23 agreement still suggests that a relative risk near 1.4 is a reasonable approximation.
24 The greatest uncertainty in estimating DE-induced cancer risk from epidemiology studies
25 is determination of exposure levels. Even though DPM concentrations were often measured near
26 the end of the studies, historic exposure data are generally lacking. Such information is critical,
27 since there is indirect evidence, based on other pollutant .measurements such as nitrogen oxides,
28 that exposure levels have decreased considerably in recent years, especially in the railroad
29 industry (Woskie et al., 1988a). In the only historic study found in which DPM was measured,
30 Heino et al. (1978) reported average concentrations of 2 mg/m3 in Finnish roundhouses.. Woskie
31 et al. (1988b), by contrast, reported a mean of 134 ug/m3 for roundhouse workers near the end of
32 the Garshick et al. (1987,1988) studies. While the relationship between DPM concentrations in
33 Finnish and U.S. railroad roundhouses during the 1970s is uncertain, it does point to the •
34 likelihood that exposure levels have decreased over time.
11/5/99 8-11 DRAFT—DO NOT CITE OR QUOTE
-------
1 With some of the uncertainties about the dose-response for both railroad workers and
2 truck drivers being actively investigated by EPA, NIOSH, and others, EPA feels that additional
3 in-depth dose-response analysis should await the newer data that are expected during 2000.
4
5 8.2.6.1. Railroad Worker Data
6 Although there have been previous efforts to use Garshick's cohort data to conduct a
7 quantitative risk assessment, a consistent dose-response between DE exposure and lung cancer
8 has not been found. Both positive and negative dose-responses have been reported, depending
9 on how the data are analyzed. These conflicting analyses have become a source of continuing
10 debate about estimated cancer risk to humans through DE exposure. Presently, a known
11 mortality under count in the Garshick data is being funded by NIOSH and updated by Garshick.
12 In the near future this updating will likely be reanalyzed by several risk assessing institutions to
13 see if some of the uncertainties have been resolved. For these reasons, EPA will not conduct its
14 own quantitative risk assessment until more information becomes available. This decision
15 should not be construed to imply that the railroad worker studies contain no useful information
16 on lung cancer risk from exposure to DE.
17 Crump et al. (1991) reported that the relative risk can be positively or negatively related
18 to the duration of exposure depending on how age was controlled in a model. Garshick et al.,
19 (1988) reported a positive relationship of relative risk and duration of exposure by modeling age
20 in 1959 as a covariant in an exposure-response model. The positive relationship disappeared
21 when attained age was used instead of age in 1959. This negative dose-response continues to be
22 upheld and further clarified in Crump (1999). California EPA (Cal-EPA, 1998) also found a
23 positive dose-response by using age in 1959 but allowing for an interaction term of age and
24 calendar year in the model. Cal-EPA produced a unit risk estimate in the range of 10° to 10"4.
25 A recent special report by HEI (1999) also suggested there was no positive dose-response
26 if a model similar to that used by Cal-EPA included a variable to stratify data into three job
27 categories: clerks, shop workers, and train workers. This observation suggests that one should
28 be cautious in estimating relative risk by taking the ratio of two relative risks calculated from two
29 different job groups (e.g., train workers vs. clerks) because there appear to be some unknown
30 job-category-specific effects operating among these groups.
31
32 8.2.6.2. Teamster Truck Driver Data
33 Steenland et al. (1998) is a case-control study of members of the Teamsters Union who
34 died in 1982-1983. Smoking histories were obtained from next of kin. Available data indicate
35 that exposure to workers in the trucking industry in 1990 averaged 2-27 fig/in3 of elemental
36 carbon (EC). The exposure information in 1990 was used as a baseline exposure measurement to
I i/5/99 8-12 DRAFT—DO NOT CITE OR QUOTE
-------
1 reconstruct past exposure (in the period of 1949 to 1983) by assuming that the exposure for
2 workers in different job categories is a function of highway mileages traveled by heavy-duty
^^ vehicles, and efficiency of the engine over the years.
4 Steenland et al. (1998) provide a potentially valuable database for calculating unit risk for
5 DE emissions. The strength of this data set is that the smoking history of workers were obtained
6 to the extent possible. Smoking is especially important in assessing the lung cancer risk due to
7 DE exposure, because smoking has much higher relative risk (or odds ratio) of lung cancer in
8 comparison to that of DEP. For the Steenland et al. (1998) study, the overall (ever-smokers vs.
9 nonsmokers) odds ratio for smoking is about 7.2, which is about fivefold larger than the 1.4 odds
10 ratio of diesel exposure. It is possible that a moderate change of information on smoking and
11 diesel exposure might alter the conclusion and risk estimate.
12 • EPA has noted that Steenland's Teamsters Union truck driver case control study workers
13 had cumulative exposure ranging from 19 to 2440, with the median and 95th percentile
14 respectively of 358 and 754 jag/m3-years of EC. These EC levels correspond respectively to 197
15 and 415 ug/m3-years of DE in ambient air, or approximately 3 and 6 ug/m3 of DE in ambient air.
16 This information is useful for comparing the exposure of the trucking personnel to animal
17 bioassay exposures and to estimated environmental exposures.
18 Steenland et al. (1998) indicate that their risk assessment is exploratory because it
^fc depends on estimates about unknown past exposures. With the Steenland risk assessment being
20 a recent publication, independent evaluation of the uncertainties have been limited to HEI (1999)
21 and a few interactions with stakeholders. HEI raised questions about the exposure estimates and
22 the selection of controls; EPA has also noted that it may have databases to facilitate the
23 development of improved exposure estimates. EPA and NIOSH are jointly pursuing some of the
24 questions, including the exposure aspects. This work will be ongoing well into 2000.
25 Give the ongoing review and reanalysis, EPA will not use the Steenland occupational risk
26 assessment to derive equivalent environmental parameters and cancer unit risk estimates until the
27 additional investigation and reanalysis is completed and can be evaluated.
28
29 8.3. OBSERVATIONS ABOUT RISK
30 8.3.1. Perspectives
31 A decision has been made for this report that, despite the finding that DE is best
32 characterized as highly likely to be a lung cancer hazard, the available data are currently
33 unsuitable to make a confident, quantitative statement about the magnitude of the lung cancer
34 risk attributable to DE at ambient exposure levels. However, the following information is
provided to put DE cancer hazard in perspective and to help decision makers and the public make
11/5/99 8-13 DRAFT—DO NOT CITE OR QUOTE
-------
1 prudent public health judgments in the absence of a definitive estimate of the upper bound on
2 cancer risk.
3 The characterization of lung cancer hazard is based on over 20 studies that demonstrate a
4 consistent, positive association between exposure to DE in occupational settings and lung cancer
5 risk. Several meta-analyses have also reported this consistent but relatively low increase in
6 relative risk of about 40%. Notwithstanding the discussion that was previously presented as to
7 why these relative risk estimates and their attendant exposure assessments are uncertain, these
8 positive associations suggest the potential for lung cancer risk at typical, historical occupational
9 exposures. Since epidemiology is a relatively crude tool, we have come to expect that increased
10 risks that are discemable over background cancer mortality in such studies will generally exceed
11 1 x 10'5 (1 in 100,000) and are often as high as 1 * 10° (1 in 1000). Unless the magnitude of the
12 risk is in this range, given typical population sizes and competing biases that tend toward the
13 null, results would be expected to be nonpositive (no statistically significant association). If
14 actual risk were much higher than this (> 10"3 or 1 in 1000), relative risks would probably be
15 much higher. While ambient average exposures are surely less than workplace exposures today,
16 the margin of exposure for some individuals at the high end (>90 percentile) of the exposed
17 population compared to the lower end of occupational exposures appears to be less than a factor
18 of 10. This means that, when occupational exposures are converted to full day rather than
19 workday and proximity to urban sources is considered, some individuals in the population may
20 be experiencing exposures that are close to or even overlapping the exposures characterized in
21 the Steenland et al. (1998) truckers study. Exposure estimates reported in that study are the
22 focus of additional scrutiny but, based on EPA insights about the exposure review, it seems
23 unlikely that exposure estimates would change by orders of magnitude and significantly alter the
24 perspective just presented. While this perspective falls short of providing a definitive estimate of
25 risk, it should be useful in providing a perspective on potential risk.
26 In addition, approaches to characterize risk to DE using comparative potency
27 methodologies suggest that upper-bound risks under estimated ambient exposure situations could
28 be in the range of 10"4 to 10"s with upper-bound unit risk estimates clustering around 10~5per
29 K?/ni3 of DEP. While this approach is not amenable to deriving a reliable point estimate of
30 upper-bound lung cancer risk for reasons described in Section 8.2., it does help to put pptential
31 risk in perspective because it relies on comparisons with other combustion products (coke oven
32 emissions or cigarette smoke condensate) or from pyrolysis products (roofing tar) for which
33 epidemiologic-based unit risk estimates have been developed.
34 EPA (1998) also developed a cancer risk estimate using benzo(a)pyrene (B[a]P) as a
35 dosimeter Pilf and Rpnrterion (1981) foiinH £ood agreement when relating the concentration of
36 (B[a]P) to lung cancer risk in smokers, British gas workers, U.S. coke oven workers, and U.S.
11/5/99 8-14 DRAFT—DO NOT CITE OR QUOTE
-------
1 hot pitch workers and when comparing residents of rural and urban locations. They concluded
2 that while B[a]P is unlikely to be the only carcinogen and perhaps not even the most important
^B one present in combustion emissions, nevertheless it serves as a reasonably accurate dosimeter.
4 Based on an estimated cancer risk of 1/1500 per ng/m3 B[a]P and a reported B[a]P concentration
5 of 3.9 ng/ug DEP in exhaust from a Volkswagen engine (Heinrich et al., 1995), a maximum
6 likelihood estimate of cancer risk from lifetime exposure to 1 ug/m3 can be calculated to be 3 x
7 10"6. The 95% upper bound was not derived, but was estimated to be near 1 x 1Q'5. The use of
8 B(a)P as a dosimeter provided reasonably good estimates of lung cancer risk for combustion and
9 pyrolysis products of coke ovens, hot pitch, gas production, refining, etc., in spite of the fact that
10 B(a)P may constitute a relatively small fraction of the carcinogens present in these emissions.
11 Risk estimates were based on well-documented lung cancer rates in the occupationally exposed
12 groups. On the other hand, while predictions are good for the pollutants tested, the particles
13 present from those combustion sources, unlike diesel particles, generally lack an insoluble carbon
14 core. As noted in Chapters 3 and 7, adsorption to an insoluble particle core is likely to influence
15 potency of individual organic components because of possible differing elution or activation
16 rates. Estimates of cancer risk will also vary based on B(a)P concentration on the particle. The
17 variability in B(a)P concentration among different DE sources and its effect on cancer potency
18 have not been evaluated in detail. While this approach appears to be useful for estimating risk
ffrom a variety of related combustion emissions, because these emissions lacked an insoluble
core, there is uncertainty.
21 In the absence of definitive risk estimates in this assessment, the three perspectives just
22 discussed are not to be taken as absolutes, but rather as an outlook. With ongoing investigations
23 regarding the existing key epidemiologic studies, as well as newly started epidemiologic studies
24 focused On diesel exhaust, there are likely to be new future opportunities to consider risk
25 estimation for diesel exhaust.
26
27 8.4. SUMMARY OF CANCER DOSE-RESPONSE CONSIDERATIONS
28 A number of attempts have been made to estimate cancer risk from exposure to DE. The
29 present scientific consensus, however, suggests that animal- and other nonhuman-based risk
30 estimates are too uncertain to extrapolate to humans. Therefore, EPA will focus on the use of
31 epidemiological data to develop quantitative risk assessment. Because some of the uncertainties
32 about the dose-response for both railroad workers and truck drivers can be reduced or better
33 characterized by obtaining additional data within a reasonable time frame, EPA sees no value at
34 this juncture in further teasing the existing data with additional dose-response analysis to unravel
»some of the uncertainties. With additional investigations underway in two of the richest
epidemiology data sets, the railroad worker data base and the Teamster truck driver database,
11/5/99 8-15 DRAFT—DO NOT CITE OR QUOTE
-------
1 EPA will await developments before taking additional steps to derive a cancer unit risk for
2 exposure to DE. EPA expects this newer information to be incrementally available throughout
3 2000.
4
5 At this time EPA is not adopting or recommending any cancer unit risk estimate for DE.
6
7 While there are uncertainties and controversy with the DE risk estimates that are available
8 from various investigators, this uncertainty can not be resolved with currently available scientific
9 information. The uncertainty about estimating unit cancer risk should not be confused with the
10 inference that DE is "highly likely" to be a human carcinogen and that most would agree that it
11 has produced risk (e.g., a relative risk increase of about 1.4) in some occupational settings. This
12 elevation of risk is of public health concern even though the exposure-lung cancer risk
13. relationship (e.g., unit risk) is uncertain. This concern is further demonstrated when comparing
14 some high end environmental exposure estimates to lower end occupational exposure estimates
15 and realizing that the exposure margin seems to be relatively small. Having information to put
16 the DE cancer hazard into perspective may be useful in the absence of definite risk estimates.
17 Risk estimates derived from epidemiology studies are preferred for DE and are the pursuit
18 of additional research and analysis.
19
20 8.5. REFERENCES
21
22 Albert, RE; Chen, C. (1986) U.S. EPA diesel studies on inhalation hazards. In: Ishinishi, N; Koizumi, A;
23 McClellan, R; et al., eds. Carcinogenicity and mutagenicity of diesel engine exhaust. Amsterdam: Elsevier, pp. 411-
24 419.
25
26 Albert, RE; Lewtas, J; Nesnow, S; et al. (1983) Comparative potency method for cancer risk assessment: application
27 to diesel particulate emissions. Risk Anal 3:101-117.
28
29 Bhatia, R; Lopipero, P; Smith, AH. (1998) Diesel exhaust exposure and lung cancer. Epidemiology 9:84-91.
30
31 Brightwell, J; Fouillet, X; Cassano-Zoppi, AL; et al. (1989) Tumours of the respiratory tract in rats and hamsters
32 following chronic inhalation of engine exhaust emissions. J Appl Toxicol 9:23-31,
33
34 California ET?vfrn"mff«tal Protean*, Agency. (CAL-EPA, OEHIiA} (Tcb. 1998) Health risk a=»e->suieiH for (iiesei
35 exhaust. Public and Scientific Review Draft.
36
37 Chen, CW; OberdOrster, G. (1996) Selection of models for assessing dose-response relationships for particle-
38 induced lung cancer. Inha! Toxicol 8(suppl):259-278.
39 -
40 Crump, KS; Lambert, T; Chen, C. (1991) Assessment of risk from exposure to diesel engine emissions. Clement
41 International Corporation. Prepared for U.S. EPA under contract no. 68-02-4601; 56 pp.
42
43 Crump. KS (IQQQ) Lung cancer mortality ~d die::! exhaust: rcanaiysis of a icuOaycwtlvc cuhuil sluuy of U.S.
44 railroad workers. Inhal Toxicol 11:1-17.
11/5/99 8-16 DRAFT—DO NOT CITE OR QUOTE
-------
Cuddihy, RG; McClellan, RO. (1983) Evaluating lung cancer risks from exposures to diesel engine exhaust. Risk
Anal 3:119-124.
Cuddihy, RG; Griffith, WC; Clark, CR; et al. (1981) Potential health and environmental effects of light-duty diesel
6 vehicles. Report No. LMF-89. Albuquerque, NM: Inhalation Toxicology Research Institute.
7
8 Cuddihy, RG; Griffith, WC; McClellan, RO. (1984) Health risks from light-duty diesel vehicles. Environ Sci
9 Technol 18:14-2 la.
10
11 Dasenbrock, C; Peters, L; Creutzenberg, O; et al. (1996) The carcinogenic potency of carbon particles with and
12 without PAH after repeated intratracheal administration in the rat. Toxicol Lett 88:15-21.
13
14 Driscoll, K. (1995) Role of inflammation in the development of rat lung tumors in response to chronic particle
15 exposure. Inhal Toxicol 8(Suppl):139-153.
16
17 Garshick, E; Schenker, MB; Munoz, A; et al. (1987) A case-control study of lung cancer and diesel exhaust
18 exposure in railroad workers. Am Rev RespirDis 135:1242-1248.
19
20 Garshick, E; Schenker, MB; Munoz, A; et al. (1988) A retrospective cohort study of lung cancer and diesel exhaust
21 exposure in railroad workers. Am Rev Respir Dis 137:820-825.
22
23 Harris, JE. (1983) Diesel emissions and lung cancer. Risk Anal 3:83-100.
24
25 Hattis, D; Silver, K. (1994) Use of mechanistic data in occupational health risk assessment: the example of diesel
26 particulates. In: Smith, MC; Christiani, DC; Kelsey, KT, eds. Chemical risk assessment and occupational health.
27 Westport, CT: Auburn House, pp. 167-178.
28
Health Effects Institute (HEI) (1999) Diesel emissions and lung cancer: epidemiology and quantitative risk
assessment. A Special Report of the Institute's Diesel Epidemiology Expert Panel. Cambridge, MA.
31
32 Heinrich, U; Muhle, H; Takenaka, S; et al. (1986) Chronic effects on the respiratory tract of hamsters, mice and rats
33 after long-term inhalation of high concentration of filtered and unfiltered diesel engine emissions. J Appl Physiol
34 6:383-395.
35
36 Heinrich, U; Fuhst, R; Rittinghausen, S; et al. (1995) Chronic inhalation exposure of Wistar rats and two different
37 strains of mice to diesel engine exhaust, carbon black, and titanium dioxide. Inhal Toxicol 7:553-556.
38
39 Hemminki, K; Sdderling, J; Ericson, P; et al. (1994) DNA adducts among personnel servicing and loading diesel
40 vehicles. Carcinogenesis 15:767-769.
41
42 Ichinose, T; Yajima, Y; Nagashima, M; et al. (1997a) Lung carcinogenesis and formation of 8-hydroxyguanosine in
43 mice by diesel exhaust particles. Carcinogenesis 18:185-192.
44
45 Ichinose, T; Yamanushi, T; Seto, H; et al. (1997b) Oxygen radicals in lung carcinogenesis accompanying
46 phagocytosis of diesel exhaust particles. Int J Oncol 11:571-575.
47
48 International Life Sciences Institute (ILSI) (1999) The relevance of the rat lung response to particle overload for
49 human risk assessment a workshop consensus report Inhal Toxicol: In press.
50
51 Ishinishi, N; Kuwabara, N; Nagase, S; et al. (1986) Long-term inhalation studies on effects of exhaust from heavy
52 and light duty diesel engines on F344 rats. In: Ishinishi, N; Koizumi, A; McClellan, RO; et al., eds. Carcinogenic
53 and mutagenic effects of diesel engine exhaust. Amsterdam: Elsevier, pp. 329-348.
11/5/99 8-17 DRAFT—DO NOT CITE OR QUOTE
-------
1 Lipsett, M; Campleman, S. (1999) Occupational exposure to diesel exhaust and lung cancer: a meta-analysis. Am J
2 Pub Health 89(7): 1 009- 1017.
3
4 Lloyd, JW. (1971) Long-term mortality study of steelworkers V. Respiratory cancer in coke plant workers. J Occcup
5 Med 13:53-68.
6
7 Martin, JC; Daniel, H; LeBouffant, L. (1977) Short and long-term experimental study of the toxicity of coal mine
8 dust and some of its components. In: Walton, WH, ed. Inhaled particles IV. Vol 1. Oxford: Pergamon, pp. 361-370.
9
1 0 Mauderly, JL. ( 1 994) Contribution of inhalation bioassays to the assessment of human health risks from solid
1 1 airborne panicles. In: Mohr, U; Dungworth, DL; Mauderly, JL; et al., eds. Carcinogenic effects of solid particles in
12 the respiratory tract. ILS I: Washington, DC; pp. 355-365.
13
14 Mauderly, JL (1997)
15
16 Mauderly, JL; Jones, RK; Griffith, WC; et al. (1987) Diesel exhaust is a pulmonary carcinogen in rats exposed
17 chronically by inhalation. Fundam Appl Toxicol 9:208-221.
18
19 McClellan, RO. (1986) Health effects of diesel exhaust: a case study in risk assessment. Am Ind Hyg Assoc J 47: 1-
20 13.
21
22 McClellan, RO; Cuddihy, RG; Griffith, WC; et al. (1989) Integrating diverse data sets to assess the risks of airborne
23 pollutants. In: Mohr, U, ed. Assessment of inhalation hazards. New York: Springer Verlag, pp. 3-22.
24
25 Mumford, JL; Chapman, RS; Harris, DB. (1989) Indoor air exposure to coal and wood combustion emissions
26 associated with a high lung cancer rate in Xuan Wei, China. Environ Int 15:3 15-320.
27
28 Nagashima, M; Kasai, H; Yokata, J; et al. (1995) Formation of an oxidative DNA damage, 8-hydroxyguanosine, in
29 mouse lung DNA after intratracheal instillation of diesel exhaust particles and effects of high dietary fat and beta
30 carotene on this process. Carcinogenesis 6: 1441-1445.
31
32 Nielsen, PS; Autrup, H. (1994) Diesel exhaust-related DNA adducts in garage workers. Clin Chem 40:1456-1458.
33
34 Nikula, KJ; Snipes, MB; Barr, EB; et al. (1995) Comparative pulmonary toxicities and carcinogenicities of
35 chronically inhaled diesel exhaust and carbon black in F344 rats. Fundam Appl Toxicol 25:80-94.
36
37 Nikula, KJ; Avila, KJ; Griffith, WC; et al. (1997) Lung tissue responses and sites of particle retention differ
38 between rats and Cynomolgous monkeys exposed chronically to diesel exhaust and coal dust. Fundam Appl Toxicol
39 37:37-53.
40
41 Pepelko, WE; Chen, C. (1993) Quantitative assessment of cancer risk from exposure to diesel engine emissions.
42 Regul Toxicol Pharmacol 17:52-65.
43
44 Pepelko, WE; Peirano, WB. (1983) Health effects of exposure to diese! eng'f.e emhsiors, J Am Coll Toxicol 3:253-
K.A
46
47 Pott, F; Heinrich, U. (1987) Dieselmotorabgas und Lungenk auf die GefMhrdung des Menschen. In: Umwelthygiene,
48 vol. 19. Med. Institut. f. Umwelthygiene, Annual Report 1986/87. Dusseldorf, F.R.G., pp. 130-167.
49
50 Riebe-Imre (1994)
51
52 Sagai, M; Saito, H; Ichinose, T; et al. (1993) Biological effects of diesel exhaust particles. I. In vitro production of
53 superoxide and in vivo toxicity in mouse. Free Radic Biol Med 14:37-47.
1175/99 8-18 DRAFT DO NOT CITE OR QUOTE
-------
1 Smith, RA; Stayner, L. (1990) An exploratory assessment of the risk of lung cancer associated with exposure to
2 diesel exhaust based on a study with rats. Final report. Division of Standards Development and Technology
3 Transfer; Cincinnati, OH: NIOSH.
Steenland, NK; Silverman, DT; Homung, RW. (1990) Case-control study of lung cancer and truck driving in the
6 Teamsters Union. Am J Pub Health 80:670-674.
7
8 Steenland, K; Deddens, J; Stayner, L. (1998) Diesel exhaust and lung cancer in the trucking industry: exposure-
9 response analysis and risk assessment. Am J Ind Med 34:220-228.
10
11 U.S. Environmental Protection Agency. (1986) Guidelines for carcinogenic risk assessment. Federal Register
12 51(185):33992-43003.
13
14 U.S. Environmental Protection Agency. (1996) Proposed guidelines for carcinogen risk assessment. Federal
15 Register 61(79): 17960-18011.
16
17 U.S. Environmental Protection Agency. (1998) Health assessment document for diesel emissions-review draft.
18 EPA/600/8-90/057C.
19
20 Waller, RE. (1981) Trends in lung cancer in London in relation to exposure to diesel fumes. Environ Int 5:479-483.
21
22 World Health Organization (WHO). (1996) Diesel fuel and exhaust emissions. International Program on Chemical
23 Safety, EHC 171. Geneva, Switzerland.
24
25 Woskie, SR; Smith, TJ; Hammond, SK; et al. (1988a) Estimation of the diesel exhaust exposures of railroad
26 workers: II. National and historical exposures. Am J Ind Med 13:395-404.
27
28 Woskie, SR; Smith, TJ; Hammond, SK; et al. (1988b) Estimation of the diesel exhaust exposures of railroad
workers: I. Current exposures. Am J Ind Med 13:381-394.
31 Zaebst, D; Clapp, D; Blade, L; et al. (1991) Quantitative determination of trucking industry workers' exposures to
32 diesel particles. Am Ind Hyg Assoc J 52:529-541.
33
34 Zhong, Y; Dehong, L. (1995) Potential years of lifework and tenure lost when silicosis is compared with
35 pneumoconiosis. Scand J Work Environ Health 21 (Suppl 2):91 -94.
11/5/99 8-19 DRAFT—DO NOT CITE OR QUOTE
-------
9. CHARACTERIZATION OF HEALTH HAZARD AND DOSE-RESPONSE FOR
DIESEL ENGINE EXHAUST
9.1. INTRODUCTION
2 Earlier chapters focused on specific health assessment topics and developed key findings
3 for these topics or provided an overview of relevant background information. This chapter will
4 integrate the key findings about health hazards and dose-response analysis for humans exposed to
5 diesel exhaust (DE). Health hazard characterization and dose-response analysis are two of the
6 four components of risk assessment. A third component, exposure assessment, is not within the
7 scope of this report, though an environmental exposure perspective is included in Section 2.4 to
8 assist in evaluating some aspects of the available lexicological information. The fourth
9 component, a population-based risk characterization for environmental exposures to diesel engine
10 exhaust, is beyond the scope of this document.
11 For introductory purposes, an overview of themes from the key assessment areas will
12 help put the remainder of this chapter into perspective.
13
14 • The DE particle and its coating of organics, as well as the accompanying gases and
15 semivolatiles, have biochemical and lexicological properties that raise suspicions
f about adverse health effects for DE given a sufficient dose, dose-rate, or
cumulative exposure.
18 • Because DE is a mixture, the choice of a dosimeter for measuring exposure is
19 important; ug/m5 of diesel particulate matter (PM) is used as Ihe dosimeter for the
20 entire DE mixture.
21 • Ambient exposures to DE vary widely depending on the proximity to sources of
22 diesel engine emissions, including on-road vehicles, off-road machinery, railroad
23 locomotives, and ships. Generally speaking rural locations have lower
24 concentrations of DE than do urban areas, and proximity to occupational settings
25 where diesels are in frequent use provides opportunities for even higher exposures.
26 The margin between high end environmental exposures and occupational
27 exposures is of interest.
28 • Noncancer toxicity: For chronic exposure, there is scanty human but much animal
29 evidence for adverse respiratory effects, such as airway restriction, inflammation,
30 and related measures of pulmonary histopathology. Acute exposure in humans
31 may elicit symptoms of irritation, ranging from annoying or temporarily
debilitating symptoms reflecting tissue irritation. An emerging concern is the
11/5/99 9-1 DRAFT-DO NOT CITE OR QUOTE
-------
possible role of DE in exacerbating or initiating allergenic effects following acute
2 or chronic exposure. The similarity or difference in these DE effects compared
3 with ambient fine particulate matter is of interest.
4 Carcinogenicity: Occupational epidemiologic studies, using surrogates for DE
5 exposure, show a pattern of increased cancer risk for the lung. Most rat and some
6 mouse inhalation studies show a carcinogenic response in the lung at high test
7 exposures; in the rat these responses occur under conditions of particle overload.
8 Organic components of DE have known or suspected mutagenic/genotoxic and
9 carcinogenic properties. Mode-of-action information provides a framework to
10 evaluate the observed lung cancer responses and judge the confidence in
11 establishing the human hazard potential as well as suggesting the best approach for
12 conducting dose-response analysis and estimation of cancer unit risk.
13
14 9.2. WHAT IS DIESEL EXHAUST IN A HEALTH HAZARD ASSESSMENT
15 CONTEXT?
16 DE is a complex mixture of literally hundreds of components. As reviewed in Chapter 2,
17 the mixture consists of particles and gases. The particulate matter consists of an elemental carbon
18 core particle with hundreds of organic and some inorganic compounds adsorbed to the particle
19 surface. The gaseous fraction is also made up of many organic and multiple inorganic
20 compounds. Some organics and inorganics also exist in a semivolatile state. The elemental
21 carbon core, the particle coating of adsorbed compounds, and the gaseous and semivolatile
22 elements each have constituents with known toxicological properties, and in addition there is a
23 possible aggregate toxicological potential for the whole mixture.
24 The DE particle fraction is made up of a distribution of particle sizes (e.g., nano/ultrafine
25 particles of 0.005-0.05 um mean mass aerodynamic diameter), as well as clusters of aggregated
26 particles (e.g., fine particles of 0.05-0.7 um MMAD) and a small number of larger particles (e.g.,
27 coarse size of 1.0-10.0 //m MMAD) (Section 2.6.5). Typically the particles average about 0.2 um
28 MMAD and have a very large surface area (50-200 m2 /g). Most of the particle mass is in the fine
23 size range, while Lhe majority of the panicles are in lie nauo/uircarjne range. The vast majority of
30 DE particles will be present in a PM15 fraction of total PM. In any given ambient PM sample,
31 diesel particles may or may not be present, depending on the proximity to a diesel emission
32 source. The diesel particle is crudely distinguishable from other PM by virtue of its elemental
33 carbon core and possibly certain qualitative or quantitative differences in the adsorbed organics.
34 DE may contribute significantly to total ambient PM: for instance, Schauer et al. (1996) reported
35 nationwide diesel contributions to total PM2 5 mass of 12.8%-35.7% in several urban California
11/5/99 9-2 DRAFT-DO NOT CITE OR QUOTE
-------
1 regions in 1982, whereas the more current Denver area NFRAQS (1998) study showed diesel
2 PM2.S to be 9.7%-10.2% of total PM 2.5 mass. The U.S. EPA Air Quality Planning and Standards
report on air pollutant trends indicates that annual emissions of diesel PM2.5 nationwide are 5.7%
of the total PM 2.s inventory and 21% of the inventory excluding natural and fugitive dust sources.
5 The diesel particle size distribution is significant for exposure-response purposes because
6 smaller particles have a greater likelihood of being deposited more deeply in the lung than do
7 larger diameter particles. Additionally, smaller particles have a larger surface area per unit of
8 mass and therefore may adsorb and transport more organic compounds into the respiratory system
9 than the same mass of larger particles, and may elicit more of an inflammatory response
10 characteristic or poorly soluble particles (Section 7.4.1). From these circumstances, it would be
11 suspected that DE particles may have a different (e.g., increased) potential for lexicological
12 consequences compared to larger particles of other than DE origin.
13 The main constituent by weight of the diesel particle is elemental carbon (Section
14 2.2.6.1). Various studies show the DE particle composition to vary considerably, with the
15 elemental carbon content ranging from 30% to 90% of total mass, with 80% being typical. (For
16 reference, PM from gasoline engine exhaust typically has a much smaller fraction of elemental
17 carbon and a large organic fraction.) The DE inorganics include nitrates, compounds of sulfur,
18 and some carbon monoxide. The DE particle organics include many compounds, a number of
which are considered to have a mutagenic and carcinogenic hazard potential for humans (see
Table 2-9 for classes of compounds), though the concentrations of the organics are generally low.
21 Many PAHs and PAH derivatives are toxic, especially the nitro-PAHs. Many of the compounds
22 emitted as gases are also potentially carcinogenic or otherwise toxic at some dose, though not
23 necessarily known to be toxic to the lung. These include benzene, 1,3-butadiene, various
24 aldehydes, ethylene dibromide, nitroaromatics, oxides of nitrogen, and sulfur compounds.
25 Additionally, there is evidence that the mixture of organics emitted and as altered in atmospheric
26 transformation provides the chemical species necessary for the formation of free radicals (e.g.,
27 reactive oxygen or hydroxyl species formed from certain organics with or without mammalian
28 metabolism); free radicals are known to cause DNA damage in biological systems (Section 7.4.3).
29 The quantitative physical-chemical composition of any discrete diesel exhaust depends on
30 numerous factors, including operating conditions, heavy-duty versus light-duty engines, engine
31 design, engine age, fuel used, exhaust control technology, and the sampling and measurement
32 system used. Diesel particle measurement in the laboratory under controlled conditions versus
33 sampling hi the ambient environment is likely to produce varied results, because the formation of
34 particles is influenced by dilution ratios and conditions of temperature and humidity. These
factors mostly affect particle size but may also affect particle composition. The available human
i
11/5/99 9-3 DRAFT-DO NOT CITE OR QUOTE
-------
1 and animal studies were based on engine exhaust representative of engines and conditions at
2 various times since 1980, while some of the epidemiology studies cover exposures from the 1950s
3 through the mid 1980s. This leads to two questions: how the physical-chemical nature of the past
4 exposures compares to present-day exposures, and how applicable the toxicological results
5 generated from the older exposures are to current-day DE exposure-related hazards. These
6 questions frame a risk assessment uncertainty for which there are no definitive answers.
7 The overwhelming majority of the emission, exposure, and toxicological data uses
8 particle emission mass expressed in units of ug/m3 for DE measurement. This was assumed early
9 on by researchers to be a useful dosimeter, At first glance this approach seems to ignore the
10 gaseous component and it does not distinguish between elemental carbon and the accompanying
11 organics.
12 . In Sections 2.2.6 and 2.2.7 an attempt is made to characterize the changes in engine
13 emissions over the years, taking into consideration the lack of consistent and reliable data and the
14 variability of in-use engines. What can be crudely inferred from the available data is that trends in
15 the emission composition over the years have not changed much, qualitatively, though some
16 quantitative changes are discernible in the past 20 years. By analysis, on-road diesel engine
17 particulate emissions were reduced about sixfold, at most 10-fold, on a g/mile basis from 1977 to
18 1997. Both the elemental carbon and organic content are decreasing. The decrease in organics is
19 mostly a consequence of engine designs that seek to reduce oil consumption. Available research
20 suggests that while most PAH emissions, including nitro-PAHs show a declining trend on a g/mi
21 basis, the overall PAH composition profile has not changed significantly. There is no available
22 evidence that toxicologically significant organic components of DE (e.g., PAHs, PAH derivatives,
23 and nitro-PAHs) have significantly increased or decreased out of proportion to the change in
24 organic mass. Limited particle size measurements suggest that current engine emissions may
25 have higher concentrations of nano/ultrafine particles; however, the methods for measuring these
26 particles are in an early stage of development, and at the moment there is little concrete evidence
27 that modem engines produce greater amounts of O.05 urn particles than older engines. Given
28 ihis inforniaiion 'dad iccogfii/ing the extensive use of jj.g/nr in published research results, ug/m: is
TO ,,~~j „„»'_„ j,,,-;_„*.,..;.. *v;-.-.-.-.-..•.—^-~.+ "rt.~ u—,+ -,u^:^^ _rj~^: * ? -..*___ ... >. j ..,- .
t±\J IJOWU UO kUW \J.WOXJ.iiWt^i lii UXL.7 UOOWJJXXl.t.U.1.. i iiW t/WOk t-liWiVO, WJL VJ.WOliU.v-l.l-l. CUIU OUCOCl^UCUL iCU.UCU.Uil
30 of uncertainty will only be discernible when there is a better understanding of diesel's
31 toxicological mode of action,
3?. The second question, the applicability of past exposure-toxicological results to present-
33 day exposure scenarios, is not fully answerable and thus remains an area of uncertainty. The
34 observation that there is no particular evidence for a major qualitative change in organic
35 composition, especially for PAHs, and that organics can be viewed as proportional to the particle
1175/99 9-4 DRAFT-DO NOT CITE OR QUOTE
-------
1 mass provides a rationale for the applicability of prior-year assessment findings to more current
2 exposures when ug/m3 is used as the dosimeter.
Once diesel emissions are released in the air, they are subject to dispersal, dilution, and
chemical and physical transformations (Section 2.3.3). Newly emitted exhaust is termed "fresh"
5 while exhaust more than 1 or 2 days old is referred to as "aged" because of alterations caused by
6 sunlight and other chemical physical conditions of the ambient atmosphere. It is not clear what
7 the overall lexicological consequence of exhaust aging is, because some compounds are altered to
8 more toxic forms while others are made less toxic. For example, PAHs present in fresh emissions
9 may be nitrated by atmospheric NO3 to form nitro-PAHs, thus adding to the existing burden of
10 nitro-PAHs present in fresh exhaust; alkanes and alkenes may be converted to aldehydes, and
11 oxides of nitrogen to nitric acid. The atmospheric lifetime for some of the transformed
12 compounds ranges from hours to days (Chapter 2, Table 2-9). On the other hand, PAHs present
13 in the gas phase react with hydroxyl radicals present in the ambient air, leading to reduced
14 atmospheric halftimes of the original PAHs. In general, secondary pollutants formed in an aged
15 aerosol mass are more oxidized, and therefore have increased polarity and water solubility.
16 Comprehensive assessment of the health hazards posed by DE would also consider the hazards
17 posed by the atmospheric reaction products, a task that is not directly addressed in this
18 assessment. In terms of environmental and occupational concentrations of DE, most people
exposed to DE receive a mixture of both fresh and aged exhaust, with the proportion of fresh
exhaust likely related to their proximity to the source of emissions. On the other hand, the DE
21 used in animal bioassays had a high percentage of fresh exhaust.
22 '
23 9.3. NONOCCUPATIONAL AND OCCUPATIONAL EXPOSURE
24 While a rigorous and comprehensive exposure analysis for DE has not been conducted as
25 part of this assessment, some exposure information from EPA's Office of Mobile Sources has
26 been included in Section 2.4.3 to provide a context for the hazard assessment and dose-response
27 analysis. Nonoccupational exposure to DE occurs worldwide in urban areas, with lesser exposure
28 in rural areas. The concentration of DE constituents in the air will vary within any geographic
29 area based on the number and types of diesel engines (on-road and off-road) in the area, the
30 atmospheric patterns of dispersal, and the proximity of the exposed individual to the diesel
31 emission source. Certain occupational populations can be exposed to much higher levels of DE
32 compared with a majority of the population.
33 In developing a perspective on human exposure one has to distinguish between airborne
34 concentrations present at any given time versus actual human exposure. Estimates of annually
averaged DEP (diesel exhaust participates) at fixed sites In urban and suburban areas in the 1980s
11/5/99 9-5 DRAFT-DO NOT CITE OR QUOTE
-------
1 ranged from approximately 4.4 ug/m3 to 11.6 ug/m3 from chemical mass balance (CMB)
2 modeling which covers all (on-road and off-road) sources of emission (Section 2.2.4). Modeling
3 shows that an above-average day, representing a high concentration day, may be in the 10 to 22
4 ug/m3 range, and that "hotspots" (near highways, bus depots, or other transportation facilities)
5 may range up to 47 ug/m3. In a broader sense, DEP concentrations assessed by CMB for both on-
6 road and off- road at fixed sites in suburban and urban areas range from approximately 1.2 to 3.6
7 ug/m3.
8 For exposure estimation EPA relies on a Hazardous Air Pollutant Exposure Model which
9 deals with on-road sources only (Section 2.4.4). This model indicates that on an annual basis, the
10 urban population is exposed to levels of DEP from 0.6 to 1.7 ug/m3. For more highly exposed
11 individuals in urban areas the range is 0.9 to 4.1 ug/m3. These estimates include projections into
12 the 1990s. Those in the population that have outdoor time in proximity to diesel exhaust sources
13 such as highway truck routes are likely to have a higher exposure during the outdoor time, and
14 thus their annual average exposure is somewhat higher than those with lesser outdoor time.
15 Recent studies, including a study of the Baltimore Harbor Tunnel (conducted by the
16 Desert Research Laboratory for the American Petroleum Institute) and an ORD measurement
17 study of tailpipe emissions from a moving heavy-duty diesel truck, have confirmed that dioxins
18 are formed and emitted from heavy-duty diesel trucks (Section 2.2.6.4). ORD's dioxin source
19 emission inventory estimates that 60 g TEQ were emitted from heavy-duty U.S. trucks in 1995.
20 This does not account for other vehicular diesel emissions (e.g., diesel automobiles and other
21 truck categories) or any off-road emissions from the many diesel-powered engines. When the
22 heavy duty truck estimate is compared with total estimated U.S. emissions of 3000 g TEQ for
23 1995, it appears that the heavy-duty diesel trucks are not a major dioxin source. The human
24 dioxin exposures of concern have been primarily noninhalation exposures associated with human
25 ingestion of certain foods, e.g., beef, vegetables, and dairy products contaminated by dioxin. It is
26 unknown whether heavy-duty truck DE deposition has a local food chain impact.
27
28 9.4. HAZARD CHARACTERIZATION
29 9.4.1. Health Effects Other Than Cancer: Acute Exposures
30 As reviewed in Chapter 5, the most readily identified acute (e.g., usually single-exposure)
31 noncancer health effect of DE on humans is its ability to elicit complaints of eye, throat, and
32 bronchial irritation as well as physiological symptoms such as headache, lightheadedness, nausea,
33 vomiting, and numbness or tingling of the extremities. Such symptoms have been reported by
34 individuals exposed to DE on busy city streets or in bus stations, most of which are case reports
35 without an understanding about the possibility of confounding exposures. Recent human and
11/5/99 9-6 DRAFT-DO NOT CITE OR QUOTE
-------
1 animal studies also suggest that acute DE exposure episodes may play a role in the development
2 of immunological allergic reactions, possibly resulting in prolonged hypersensitivity to DE and
^^ perhaps other ambient contaminants. It is premature to further characterize DE's allergenicity
^^ effects until additional information is available.
5
6 9.4.2. Effects Other Than Cancer: Chronic Exposure
7 Based on limited evidence in human occupational studies, but combined with multiple
8 controlled laboratory animal studies in several species, a high level of confidence exists that
9 chronic exposure to DE constitutes a noncancer respiratory hazard for humans. As DE exposure
10 levels and duration increase, the onset of respiratory symptoms in humans is observable, with
11 limited evidence of long-term consequences, whereas in animal studies the onset of symptoms
12 and adverse consequences is more clear and replicable. Current data also identify possible
13 neurological and behavioral effects, though these occur at higher exposure levels than the
14 respiratory effects. Animal studies show a possible high-exposure reproductive effect, but no
15 other reproductive or developmental consequence is identified. Section 5.6 summaries discuss
16 this topic in more depth.
17 A few human studies in various diesel occupational settings suggest that diesel exposure
18 may impair pulmonary function, as evidenced by increases in respiratory symptoms and some
reductions in baseline pulmonary function consistent with restrictive airway disease. Other
studies found no particular effects. The methodologic limitations in these studies limit their
21 usefulness in drawing any firmer conclusions (Sections 5.6.1, 5.6.9).
22 There is a considerable body of animal evidence that clearly correlates DE exposure with
23 pulmonary injury. Short-term animal exposures of high concentrations of diesel PM resulted in
24 histological and cytological changes in the lungs, but only minimal effects on pulmonary
25 function. A number of long-term laboratory studies with rats, mice, Chinese hamsters, Syrian
26 golden hamsters, cats, and Cynomolgus monkeys found varying degrees of adverse lung
27 pathology. Histological studies show a variety of changes hi respiratory tract tissue, including
28 focal thickening of the alveolar walls, replacement of Type I alveolar cells by type II cells, and
29 fibrosis. Exposures for several months or longer to levels markedly above environmental ambient
30 concentrations resulted in accumulation of particles in the animal lungs and an unpaired ability to
31 clear particulate matter from the lungs. While the applicability of rat lung cancer responses to
32 possible human hazard has been questioned, the noncancer rodent responses are thought to be
33 relevant for humans, though the rat is more sensitive than other rodent species and is also
34 suspected to be more sensitive than humans for a number of toxic effects (ILSI, 1998). Because
these effects were seen in a wide range of animal species, there is a qualitative basis to believe
11/5/99 9-7 DRAFT-DO NOT CITE OR QUOTE
-------
1 that humans could also experience hazard for these effects and may be at risk under some
2 condition of exposure.
3 Available data limit current efforts to develop hypotheses regarding specific mechanisms
4 or mode of action for DE's respiratory disease impact on humans. The MoA information comes
5 almost entirely from observing rodents, which demonstrate the following: (1) the particle fraction
6 of DE is involved in the etiology of toxicity, though a constituent role for the particle organics and
7 the DE gases cannot be dismissed; (2) similar particle-driven effects occur in different animal
8 species, although the observable onset varies by species; (3) lung injury appears to be mediated by
9 a progressive impairment of normal lung function by invading alveolar macrophages; and (4) it is
1 0 believed that the adverse effects have a biological threshold, there being no available evidence to
1 1 the contrary.
1 2 Animal studies have also suggested that liver and kidney changes may be occurring at
1 3 high concentrations, along with some indication of neuro toxic effects and impacts on
1 4 spermatogenesis. Impaired growth rates have also been observed in animals chronically exposed
15 to DE. However, these effects are seen at exposures higher than the respiratory effects. An
1 6 assessment focused on determining levels that are likely to be protective for respiratory hazards
1 7 will be protective for all effects observed to date.
1 8 Respirable particles in general have been implicated as etiologic factors in various types
1 9 of chronic human lung diseases (U.S. EPA, 1996). Ambient PM is associated with increased
20 morbidity and mortality, aggravation of respiratory and cardiovascular disease, changes in lung
21 function and increased respiratory symptoms, changes to lung tissues and structure, and altered
22 respiratory defense mechanisms. The majority of DE particle mass is in the low end of the "fine"
23 particle range, and thus contributes to ambient levels of PM2.5.
24
25 9.4.3. Health Effects Other Than Cancer: Derivation of Inhalation Reference
26 Concentration
27 A considerable body of evidence provides a basis to infer a noncancer respiratory health
28 hazard folio wmg uuiuiiauGii of DE. On the basis Oi pulmonary function and histopathological and
V " rr+f-. -»t-^--iw» ~ ~«^"Y -.^CW^t*^ 'v^ ™*»*fl f* r»<-k-~*<» *-»O«*
30 dose/exposure rates of DE (measured in terms of the concentration of diesel PM) cause an adverse
31 effect and which exposures do not; this then is a starting point for estimating protective margins
32 for human exposure. The available human studies, while qualitatively suggestive of possible
33 adverse effects, were inadequate for RfC determination. A reliable experimental database and
34 established EPA dose-response evaluation methods have been used to derive an inhalation
35 reference concentration (RfC) for chronic exposure to DE. •
1 1/5/99 9-8 DRAFT-DO NOT CITE OR QUOTE
-------
1 The derivation of an RfC for DE is a dose-response approach used by EPA for chronic
2 noncarcinogenic effects. An RfC is defined as an estimate of a continuous inhalation exposure to
the human population, including sensitive subgroups, with uncertainty spanning perhaps an order
of magnitude, that is likely to be without appreciable risks of deleterious noncancer effects during
5 a lifetime. The RfC approach is based on the assumption that a threshold exists for the human
6 population below which no effect will occur. The approach identifies a "critical" effect and
7 related NOAEL; "critical" is defined as the first effect, or its known precursor, that occurs as the
8 dose rate increases. There may be various uncertainties associated with this selection. Second,
9 depending on the critical study, any of several types of uncertainty factors are used to reduce the
10 NOAEL to a level that is thought to be without appreciable hazard to humans. The selection of
11 uncertainty factors is driven by both science and policy considerations focused on uncertainties in
12 the available data or, in some cases, reflecting the absence of data. The resulting RfC is not a
13 bright line (i.e., just above which hazard can be expected); rather, as the human exposure
14 increases beyond the RfC, the margin of protection decreases and the likelihood of hazard is
15 considered to increase.
16 The DE RfC evaluation closely examined 10 long-term (greater than 1 year) DE
17 inhalation studies in laboratory rats. This is beneficial to the process of RfC determination
18 because the data base on the critical effect has an unusually large number of relevant studies
^^ (Chapter 6). The available human studies, as discussed earlier, were qualitatively suggestive of
2u adverse effects but were inadequate for RfC determination. Two key rat studies (Mauderly, 1988;
21 Ishinishi, 1988) were selected because each identified respiratory effects after chronic exposure
22 and provided good information about pulmonary histopathology. The selected studies also
23 spanned a wide range of exposures from 350 to 7000 ug/m3, with three exposures in the 350-960
24 |ig/m3 range. Human equivalent concentrations (HEC) were calculated from the animal exposure
25 information using a dosimetry model developed by Yu et al. (1991) that accounted for species
26 differences in respiratory exchange rates, particle deposition efficiency, differences in particle
27 clearance rates at high and low doses, and transport of particles to lymph nodes. The adopted RfC
28 evolved from a NOAEL of 460 ug/m3 (HEC = 155 ug/m3) that was related to a LOAEL of 960
29 ug/m3 (HEC = 300 u.g/m3) (Table 6-2). Although particle overload conditions are thought to
30 occur above 1000 ug/m3, the likelihood of lung overload conditions is thought to be minimal at
31 460 ug/m3.
32 Two principal areas of uncertainly are present in the RfC derivation (Section 6.1.3). As
33 the RfC is based on a chronic animal study, an uncertainty factor of 10 is usually applied for the
34 animal-to-human extrapolation of an effect to account for the possibility that humans may be a
more sensitive species than the rodent. This uncertainty is equally parceled (10°-5 each) between a
11/5/99 9-9 DRAFT-DO NOT CITE OR QUOTE
-------
1 pharmacokinetic (PK) component and a pharmacodynamic (PD) component. As a PK model was
2 used in this assessment to derive the HEC, the uncertainty about the PK component was
3 considered resolved. Application of uncertainty for the PD component was more complex.
4 Although the rat appeared to be clearly more sensitive than humans for the inflammatory
5 responses underlying the observed pulmonary pathology, it was not clear if rats were also more
6 sensitive than humans to those inflammatory processes underlying the observed enhanced
7 allergenicity. In light of this uncertainty, the uncertainty for the PD component is maintained at
8 10°5, which is rounded to 3. A second uncertainty factor of 10 is generally used to account for
9 possible inter-individual variability in sensitivity unless mechanistic or other data suggest
10 otherwise. This uncertainty factor is considered appropriate for the current assessment. The total
11 uncertainty factor is 3 * 10 = 30. With 155 ug/m3 divided by 30, the resulting RfC is
12
13 RfC = 5 jug/m3 ofdiesel paniculate matter (DEP).
14
15 A comparison of the DE RfC and the PM2 5 regulatory standard is not a straightforward
16 endeavor, and caution should be exercised in comparing the output of a health hazard assessment
17 RfC and the product of a science-based regulatory process. Nonetheless, conclusions reached in
18 each of these processes are remarkably similar. EPA's 1997 PM2.3 standard is 15 ug/m3, as a 3- -
19. year average, based on human studies. The noricancer respiratory effects from DE are
20 qualitatively similar to some of those for PM2.5. The DE particulates can be a component of
21 ambient PM2.5. Compared to ambient PM2 5 with no DE component, DE is likely to have a higher
22 proportion of fine and ultrafine particulates and is likely to have a higher or at least a varied
23 content of lexicologically active organic compounds. Although some similarities exist between
24 DE and ambient PM, the differences are potentially significant. A comparison of the DE RfC and
25 the PM2.5 standard has considerable complexity.
26
27 9.5. CARCINOGENICITY HAZARD CHARACTERIZATION
28 For inhalation expos arc, uotii uuuiaii studies and annual bioassays arc available for
^£5? U^)^tC^^i.XXCX.ll OX *--* M-!J- AAA XdW b* OW Mk^ kXAW I t ^>t t *^** t b^A^* ^r^>^ I '.'• • I I \-m |_- -v "•' >i_r -*^ * <«**•« f ",• t > m * BI .'. 1 -u-i^ i_-itn UK •^ \ ^^
30 evidence that exposure to DE has the potential to be carcinogenic to humans under some
31 conditions of exposrzre. Chapter 7 reviews the cancer data in detail. A finding about the hazard
32 potential does not specify the magnitude of the possible impact on an exposed population; this i
33 an issue for dose-response assessment, which is discussed in Chapter 8.
11/5/99 9-10 DRAFT-DO NOT CITE OR QUOTE
-------
1 9.5.1. Cancer Hazard
2 Diesel engine exhaust is "highly likely" to be carcinogenic by the inhalation route of
exposure, according to EPA's 1996 Proposed Guidelines for Carcinogen Risk Assessment. This
hazard is viewed as being applicable to ambient (i.e., environmental) exposures. There is no
5 available evidence to evaluate the hazard from other routes of exposure. The "likely"
6 characterization generally compares with the weight-of-evidence designation "B-l, probable
7 human carcinogen" from the EPA's 1986 Guidelines for Carcinogen Risk Assessment. The
8 overall weight of evidence for DE places it at the upper end of ihe grouping and hence gives the
9 "highly likely" designation (Section 7.5). The carcinogenic potential of DE is indicated by: (1)
10 consistent association between observed increased lung cancer and DE exposure in certain
11 occupationally exposed workers; (2) induction of lung cancer by whole DE and DE particles in
12 some, but not all, inhalation animal bioassays: (3) induction of cancer from various fractions of
13 the DE mixture, as shown in skin painting, intratracheal, and other noninhalation animal test
14 • systems; and (4) the presence of organics on the diesel particles and in the DE gases, some of
15 which have potent mutagenic and carcinogenic properties in their own right, as well as some
16 evidence for the bioavailability of the organics. The mode of action for carcinogenicity in humans
17 is unknown, though it could be suggested that either or both the organics in the DE and the
18 elemental carbon diesel particle contribute to the carcinogenic activity.
f Increases in relative risk for lung cancer have been consistently noted in a number of
epidemiologic studies, and causality considerations for this observed association are very
21 consistent with DE exposure being causally related to lung cancer (Section 7.5.1) Aggregate
22 estimates from meta-analysis of the statistically increased relative risks for smoking-adjusted
23 studies are 1.33 in one analysis and 1.47 in another (33% or 47% increase in lung cancer above
24 background), though individual studies, such as Steenland et al. (1990), had higher relative risks
25 (e-g-, 1.64 and 1.89) for specific groups of workers. Meta-analyses are a tool to evaluate relative
26 risk estimates from multiple compatible studies. Although the approach weights the influence of
27 individual study results in the overall outcome, the analysis does not override uncertainties or
28 limitations in the individual studies. A very recent publication provides yet another pooling of
29 diesel occupational exposure-lung cancer data from two large case-control studies in Germany
30 (Briiske-Holfeld et al., 1999). The aggregate relative risk results were similar to those previously
31 mentioned, with some specific job categories having relative risks greater than 2. This paper will
32 be evaluated further before this assessment is finalized.
33 The uncertainties with the DE epidemiology data are the typical ones including the
34 possibility that chance, bias or confounding are influencing the observed lung cancer increases
(Section 7.2.6.5). The persistence of the lung cancer association in multiple studies and statistical
1175/99 9-11 DRAFT-DO NOT CITE OR QUOTE
-------
1 confidence limits in key studies indicates that chance alone is unlikely to account for the observed
2 relation between DE and lung cancer. A causal interpretation for DE is enhanced when the "Hill"
3 causality criteria are evaulated, noting that an absence or weakness in one or several of the criteria
4 does not prevent a causal finding, though it could be a basis to limit one (Section 7.2.6.6). A
5 weakness in the epidemiology studies showing a positive association is that diesel exposure is
6 inferred from job codes, area descriptions, and the like, which are surrogates for the true
7 underlying exposure. This can lead to nondifferential misclassification of exposure, and while
8 unlikely this might result in a spurious risk estimate in any one study. It is even more unlikely,
9 however, that it would bias a sufficient number of studies in a uniform direction to account for the
10 persistent association observed. Moreover, any bias would likely be toward a lower risk
11 estimate. Not all studies controlled for a tobacco smoke effect. In those studies that did adjust
12 for smoking, there remains a possibility that the adjustment may not be completely effective, and
13 residual confounding by smoking may persist to bias the correlation of DE exposure with lung
14 cancer occurrence. This uncertainty is currently unresolvable.
15 An inability to satisfactorily minimize all confounding, bias, and exposure uncertainties
16 has resulted in the human evidence being judged not quite adequate to support a finding of
17 causality and characterization of DE as a "known" human carcinogen. Others looking at the same
18 evidence may reach slightly different conclusions as scientific judgment is involved. Cal-EP A,
19 for instance, has judged the epidemiologic evidence to be sufficient to support a causality finding
20 under its criteria. Others, HEI (1995) for example, have argued that human data are consistent in
21 showing weak associations between DE exposure and lung cancer, but that there is insufficient
22 evidence to conclude whether confounding and exposure uncertainties have influenced the
23 association.
24 While lung cancer has been induced experimentally in rats via inhalation of DE at high
25 exposure concentrations, the data show that the primary factors that are likely to be responsible
26 for lung cancer are high particle concentrations producing a particle overload in the lung, and
27 subsequent induction of persistent inflammatory responses, followed by DNA damage, rapid cell
28 Turnover, and eventual luiig cancer (Section 7.4.2). This rncdc of action for lung carcincgenicity
23 lii 'die iat under overload conditions is thought to be unique. It is net known wiwh*»r b.vr?.s>p
-------
1 high-exposure-related rat lung cancer responses, however, are unsuitable for estimating risk at
2 lower environmental levels of exposure in humans.
Generally, rats showed significant increases in lung tumors beginning at exposures of
>2200 ug/m3 (HEC is approximately 700-900 ug/m3). These exposure levels clearly represent
5 lung overload conditions for the rat. In addition, these human equivalent exposure concentrations
6 are significantly higher than those found in the human occupational studies discussed in Chapter
7 7. These range from about 3 ug/m3 as an environmental equivalent calculated from the
8 Teamster's Union truck study (Steenland et al., 1998, Section 8.3.8.2) to 141-192 ug/m3 (and
9 possibly up to 500 ug/m3), which is reported as an occupational level in the railroad worker study
10 (Woskie et al., 1988b and others). These reported levels overlap with, but range significantly
11 higher than, nationwide ambient continuous exposure estimates for humans of 0.6-4.1 jag/m3, not
12 counting hotspots (Section 2.4.3).
13 Organic extracts of DE particles have been shown to induce tumors in mice, both by skin
14 painting, and subcutaneous injection, and to be mutagenic in several test systems. Additionally, a
15 number of PAHs and nitro-PAHs present on diesel particles as well as in the vapor phase are
16 known to be mutagenic and/or carcinogenic. As discussed in Section 7.3.2, filtered DE (i.e.,
17 exposure to DE gases) does not produce a lung tumor response in rats. Intratracheal studies
18 (Section 7.3.4) show that DE particles with and without organics elicit a lung cancer response, as
fdoes a pure elemental carbon particle, carbon black, with a modestly higher response for the
whole DE particle. Also, four- to seven-ring PAHs are shown to be a particularly potent fraction
21 of the organic extracts.
22 The plausibility of an environmental lung cancer hazard from DE by inhalation exposure
23 is supported by findings contained in this assessment: (1) that mutagenic and tumor initiating
24 carcinogens are present in small quantities in the DE organic mixture; (2) that some
25 bioavailability of the organics is expected and that deposited particulates seem to have much
26 longer residence times in humans than in animal species. This provides an extended opportunity
27 for elution, metabolism if needed, and uptake of the organics. These organics include many well
28 characterized mutagens and carcinogens; and (3) that there may be a relatively small margin of
29 exposure between higher end environmental exposures and some occupational levels in studies
30 where statistically increased aggregate relative risks in the range of 1.33 to 1.47 are seen (e.g.,
31 exposure estimates for some truck drivers could be overlapping some environmental estimates).
32 Overall, the evidence for a likely human lung cancer hazard by inhalation is persuasive,
33 even though, in the absence of complete data, inferences and thus uncertainties are involved.
34 Some of the key uncertainties include: (1) methodologic limitations inherent in epidemiologic
studies, as well as a lack of reliable historical exposure data for occupationally exposed cohorts,
11/5/99 9-13 DRAFT-DO NOT CITE OR QUOTE
-------
1 (2) uncertainties regarding the extent of bioavailability of organic compounds present on diesel
2 particles and their impact on the carcinogenic process, and (3) other uncertainties regarding the
3 mode of action of DE on lung cancer in humans.
4 The epidemiologic evidence for DE being associated with other forms of cancer is
5 inconclusive.
6
7 9.6. CANCER DOSE-RESPONSE ASSESSMENT
8 Cancer dose-response assessment describes what is known about the relationship of
9 exposure/dose to a cancer response (e.g., lung cancer) and how the response might change with
1 0 dose within the range of empirical observations. It also evaluates the applicability of this
1 1 relationship to human low-exposure circumstances. The low-exposure aspects are approached by
12 extrapolation, if appropriate, from an observable response range to lower exposure/dose levels,
1 3 such as ambient levels of interest. Key choices in dose-response assessment are influenced by
1 4 epidemiologic and toxicologic data and informed by reasoning about the possible mode(s) of
1 5 action. In the absence of such information, standard assumptions (i.e., defaults) are used, many of
1 6 which are conservative toward public health protection. Chapter 8 contains a more detailed
1 7 review of dose-response issues.
1 8 Human data are preferred as a starting point for DE dose-response assessment, one
1 9 purpose of which would be to estimate cancer potency (i.e., cancer unit risk). Unit risk is the
20 estimated cancer risk at 1 ug/m3 of exposure for a lifetime; in this case, jig/m3 of DE particulate
21 matter from a continuous 70-year exposure. Unit risk derivation procedures and specifications are
22 defined in EPA's risk assessment guidance.
23 The overall challenge with DE is to judge the uncertainties in the dose-response analysis,
24 given available data, and to decide whether to proceed. If the analysis is carried out, it is
25 important to decide what certainties/uncertainties to ascribe to any resulting output of the analysis
26 and follow-on unit risk derivation.
27 The mode of action (MoA) for humans is unknown, and the presumed MoA for rats does
28 iiui justify using rat lung tumui. data to estimate low-exposure cancer risk for humans. This report
<•„- — r~,-:« ,»..,.,+,,~*,..:,,/~,»,-~*~,,:~ - --- .*:+, — *-. ,,.CT>T- „-, — 1-> --- ...i_ .c_.
A.V/JL WA £{ • I I I W u^fcAkltg^WAAAW gWAAVJ bW^XXW WWUO iO.bUiWjLX tO WX LJ 1-1 OO W dl Od d 1O1C 1U1
30 particles is plausible, recognizing that the relative contributions of each may vary with dose-
31 exposure. With organics thought to be in relative proportion to the mass of particulates, the use of
32 |J.g/m3 of DE particles as the dosimeter is feasible. With no clear indication that key organic
33 components have changed disproportionately to total organics over the years (Section 2.5). the use
34 of lexicological results based on older engine exposures to. predict current-day hazards is also
35 feasible, though uncertainty exists.
1 1/5/99 9-14 DRAf T-DO NOT CITE OR QUOTE
-------
1 Section 8.2 reviews a number of past attempts to estimate diesel cancer potency (i.e., unit
2 cancer risk) using epidemiology data, rat data, and comparative potency approaches. With the rat
estimates now being thought unsuitable, the comparative potency-based estimates having
limitations and thus being uncertain, and the epidemiology-based estimates having outstanding
5 issues and questions to be resolved, these historical risk estimates lack a consensus of support.
6 With ongoing investigations to update mortality in the Garshick railroad worker study and
7 additional review and analysis of the Steenland et al. (1998) study underway, the Agency has
8 determined that there is no scientific support for further analysis of the existing epidemiologic
9 data until some newer information is available. Additional information is expected over the next
10 few years.
11 A decision has been made in this assessment that, despite the finding that DE is best
12 characterized as highly likely to be a lung cancer hazard, the available data are currently
13 unsuitable to make a confident quantitative statement about the magnitude of the lung cancer risk
14 attributable to DE at ambient exposure levels. Therefore, this assessment does not adopt or
15 recommend a specific cancer unit risk estimate for DE. However, information is provided in
16 Section 8.3 to put DE cancer hazard in perspective and to assist decisionmakers and the public to
17 make prudent public health judgments in the absence of a definitive estimate of the upper bound
18 on cancer risk. This perspective is based on the consistent observations of a relatively low
3^ (-40%) increase in relative risk and the power of epidemiologic studies to detect low levels of
^T absolute risk. In addition, Section 8.2 describes the use of historic approaches that consider
21 comparative potency to inform the perspective. This discussion leads to the conclusion that the
22 available science can support a position that, if one accepts the conclusion that DE has human
23 carcinogenic potential, risks may be in the range of regulatory interest ^lO"6 or 1 in 1 million),
24 but that they are not likely to exceed levels that often result in immediate regulatory action (>10°
25 or 1 in 1000). The Agency does not believe that the current data support a more precise
26 perspective.
27
28 9.7. SUSCEPTIBLE SUBGROUPS
29 The hazards previously characterized, i.e., acute and chronic effects, are assumed to be
30 possible consequences in individuals of average health and in their adult years. There is no DE-
31 specific information that provides direct insight to the question of variable susceptibility within
32 the population. Default approaches to account for uncertainty in inter-individual susceptibility
33 have been included in the derivation of the RfC. Individuals with preexisting lung burdens of
34 particulates may have less of a margin of safety from DE particulate-driven hazards than might be
inferred from incremental DE exposure analysis, although this cannot be quantified. DE exposure
11/5/99 9-15 DRAFT-DO NOT CITE OR QUOTE
-------
1 could be additive to many other daily or lifetime exposures to organics and PM. For example,
2 adults who predispose their lungs to increased particle retention (e.g., smoking or high particulate
3 burdens from nondiesel sources), have existing respiratory or lung inflammation or repeated
4 respiratory infections, or have chronic bronchitis, asthma, or fibrosis could be more susceptible to
5 adverse impacts from DE exposure. Although there is no information from studies of DE, infants
6 and children could have a greater susceptibility to the acute/chronic toxicity of DE because they
7 have greater ventilatory frequency, resulting in greater respiratory tract particle deposition (U.S.
8 EPA, 1996b). The issue of DE impacts on allergenicity and potential onset and exacerbation of
9 childhood asthma is being actively investigated, but firm conclusions await peer review and
10 publication of ongoing work.
11 Another aspect of differential susceptibility involves subgroups that may-receive
12 additional exposure to DE because of their proximity to DE sources. Earlier it was mentioned that
13 those having outside time in their daily routine and being near a diesel emission source would
14 likely receive more exposure than others in the population. The highest exposed are most likely
15 the occupational subgroups whose job brings them very close to diesel emission sources (e.g.,
16 trucking industry, machinery operations, engine mechanics, some types of transit operations,
17 railroads, etc.).
18
19 9.8. REFERENCES
20
21 Bhatia, R; Lopipero, P; Smith, A. (1998) Diesel exhaust exposure and lung cancer. Epidemiol 9(l):84-9l.
22
23 California Environmental Protection Agency-OEHHA (Cal-EPA). (1998) Part B: Health risk assessment for Diesel
24 Exhaust, Public and Scientific Review Draft.
25
26 Garshick, E; Schenker, MB; Munoz, A; Segal, M; Smith, TJ; Woskie, SR; Hammond, SK; SpeizerJFE. (1987) A
27 case-control study, of lung cancer and diesel exhaust exposure in railroad workers. Am Rev Respir Dis 135:1242-
28 1248
29
30 Garshick, E; Schenker, MB; Munoz, A; Segal, M; Smith, TJ; Woskie, SR;Hammond, SK; Speizer, FE. (1988) A
31 retrospective cohort study of lung cancer and diesel exhaust exposure in railroad workers. Am Rev Respir Dis
32 137:820-825.
33
34 Health Effects Institute. (1995) Diesel Exhaust: A Critical Analysis of Emissions, Exposure, and Health Effects.
35 Cambridge, MA:
36 . . .
37 Health Effects Institute. (1999) Diesel Emissions and Lung Cancer Epidemiology and Quantitative Risk Assessment,
38 A Special report of the Institute's Diesel Epidemiology Expert Panel. Cambridge, MA
39
40 Lipsett, M; Campleman,S. (1999) Occupational exposure to diesel exhaust and lung cancer: a meta-analysis. Am J
41 Pub. Health 80(7) 1009-1017.
i 1/5/99 9-16 DRAFT-DO NOT CITE OR QUOTE
-------
1 NFRAQS-Northem Front Range Air Quality Study, Colorado. January 1998, Volume I.
2
3 Schauer, JJ; Rogge,WF; Hildemann, LM; Mazureik, MA; Cass, GR; B.R.T. Simoneit (1996)
Source apportionment of airborne particulate matter using organic compounds as tracers. Atmos. Environ. 30(22):
3837-3855.
Q
7 WHO -World Health Organization (1996). Diesel Fuel and Exhaust Emissions, Environmental Health Criteria 171,
8 International Program on Chemical Safety, Geneva Switzerland.
9
10 Steenland, K; Silverman, DT; Homung, RW. (1990) Case-control study of lung cancer and truck driving in the
11 Teamsters Union. Am J Public Health 80:670-674
12
13 Steenland, K; Deddens, J; Stayner, L (1998) Diesel exhaust and lung cancer in the trucking industry: exposure-
14 response analyses and risk assessment. Am J. Ind. Med. 34:220-228.
15
16 Woskie SR, Smith TJ, Hammond SK.Schenker MB, Garshick E, Speizer FE (1988a). Estimation of the Diesel
17 Exhaust Exposures of Railroad Workers: I. Current Exposures. Am Journal of Ind Med 13:381-394, 1988
18
19 Woskie SR, Smith TJ, Hammond SK,Schenker MB, Garshick E, Speizer FE (1988b). Estimation of the Diesel
20 Exhaust Exposures of Railroad Workers: II. National and Historical Exposures. Am Journal of Ind Med 13:395-404.
21 U.S. Environmental Protection Agency. (1986, Sept. 24) Guidelines for carcinogenic risk assessment. Federal
22 Register 51(185):33992-43003.
23
24 U.S. Environmental Protection Agency. (1996a) Proposed guidelines for carcinogen risk assessment. EPA/600/P-
25 92/003C.
U.S. Environmental Protection Agency (1996b) Air Quality Criteria Document for Particulate Matter. EPA/600/P-
28 95/00 laF.
11/5/99 9-17 DRAFT-DO NOT CITE OR QUOTE
-------
Appendix A
Experimental Protocol and Composition
of Exposure Atmospheres
11/5/99 A-l DRAFT-DO NOT CITE OR QUOTE
-------
APPENDIX A. EXPERIMENTAL PROTOCOL AND COMPOSITION OF
EXPOSURE ATMOSPHERES3
Facility/Sponsor
Reference
Engine type
Operating mode
Fuel type
Fuel sulfur
Exposure regime
Exposure conditions
Particle cone, (mg/m3)
Particle size
CO, (%)
CO (ppm)
NO, (ppm)
NO (ppm)
SO, (ppm)
SO/2 (MS/m3)
Ozone (ppm)
Aliphatic aldehydes
(ppm)
Formaldehyde (ppm)
Acrolein (ppm)
NH/
THC (ppm)
PAHs
Benzo(a)pyrene
Benzo(e)pyrene
Benzo(a)anthracene
Beazo(k)fluoranthene
T71» »*>••*» «+V»*»T*^
1. JH*WA fc****J.*W**W
Pyrene
Phenafittireiie
Chrysene
Perylene
Indeno(l,2,3-Cd)
fbioranthene
Indeno(l,2,3-Cd)
oyrene
Benzo(ghi)perylene
U.S. Environmental Protection Agency
Bhatnager et al., 1980; Campbell et al., 1980, 1981;
Hyde et al., 1985; Moorman et al., 1985; Pepelko et
al., 1980b, 1981; Pepelko, 1982b; Pepelko and
Peirano, 1983; Plopperet al., 1983
Nissan CN 6-33, 3.24 L, 6-cylinder
Federal short cycle
No. 2 diesel
0.15%
8 h/d, 1 d/week, 124 weeks
Control
0.00
0.04 ± 0.002
2.20 ± 0.50
0.03 ± 0.03
0.05 ± 0.04
0.03 ± 0.02
-
0.00
0.00
0.00
-
2.82 ± 0.50
Exhaust -
weeks 1-61
6.34 ± 0.81
Exhaust -
weeks 62-124
11.70 ± 0.99
90% < I jim by mass;
50% z 0.3 urn by mass
0.30 ± 0.04
20.17 ± 3.01
2.68 ± 0.80
11.64 ± 2.34
2.12 ± 0.58
-
0.177 + 0.043
0.106 ± 0.029
0.025 ± 0.003
-
7.93 ± 1.42
0.52 ± 0.04
33.30 ± 2.94
4.37 ±1.19
19.39 ± 3.80
5.03 ± 1.03
-
0.338 ± 0.057
0.251 ± 0.059
0.034 ± 0.009
-
11.02 ± 1.04
15.9 /xg/g extract
28.6 /ig/g extract
53.8 jig/g extract
77.8 Aig/g extract (k+b)
1^ B >ft
*~~.« r-&' p ~- -— •*— -
198 /ig/g extract
145.2 jig/g extract
71.6 /tg/g extract
3.5 figlg extract
10.9 /ig/g extract
14.8 ng/g extract
21.1 /ig/g extract
Laurie et al., 1980; Laurie and
Boyes, 1980, 1981
3.24 L, 6 cylinder
Federal short cycle
No. 2 diesel
0.15%
8 h/d, 7 d/week, 16 weeks
Control
0.01
0.05 ± 0.00"
1.86 ±0.06b
0.03 ± 0.00"
0.08 ± 0.01"
0.46 ± 0.02"
3.22 ± 0.08"
Exhaust
5.97 ±0.17b
0.28 ± 0.01"
19.20 ± 0.35b
2.51 +0.10"
11.14 ± 0.43"
1.82 ±0.07"
7.29 + 0.11"
a All ± are S.D., unless specified otherwise.
h Standard error of mean values.
11/5/99
A-2
DRAFT-DO NOT CITE OR QUOTE
-------
APPENDIX A. EXPERIMENTAL PROTOCOL AND COMPOSITION OF
EXPOSURE ATMOSPHERES3
Facility/Sponsor
Reference
Engine type
Operating mode
Fuel type
Fuel sulfur
Exposure regime
Exposure conditions
Particle cone, (mg/m3)
Particle size
C0,(%)
CO (ppm)
NO, (ppm)
NO (ppm)
SO, (ppm)
SO/2 (fig/m3)
Ozone (ppm)
Aliphatic aldehydes
(ppm)
Formaldehyde (ppm)
Acrolein (ppm)
NH/
THC (ppm)
PAHs
Benzo(a)pyrene
Nitropyrene
U.S. Environmental Protection Agency
Wiesteretal., 1980
Nissan CN6-33, 3.24 L, 6 cylinder
California cycle, modified
No. 2 diesel
0.15%
20 h/d, 7 d/week, 4 weeks
Control
0.00
0.04
2.0
0.07
0.11
0.0
0.00
0.0
0.00
Exhaust
6.32 ± 1.31
Exhaust - irradiated
6.83 ± 1.44
0.1-1.0 nm
0.261 ± 0.01
17.4 ± 2.5
2.3 ± 0.4
5.9 ± 0.6
2.1 +0.8
0.57 ±0.12
0.0
31.6 ± 2.3
0.25 ± 0.03
16.7 ± 4.0
2.9 ± 0.7
5.0 ± 1.2
1.9 + 0.8
0.57 ±0.13
<0.01
26.1 ± 1.6
Pepelkoetal., 1980a
3.24 L, 6 cylinder
California cycle, modified
No. 2 diesel
20 h/d, 7 d/week, 4 weeks
Exhaust
6.40 ± 0.36b
0.26 ± 0.008b
14.61 ± 0.90b
2.13 ±0.09b
6.13 + 0.18"
2.10 + 0.21"
0.577 + 0.019b
31.56 ± 1.25"
1 All ± are S.D., unless specified otherwise.
b Standard error of mean values.
11/5/99
A-3
DRAFT—DO NOT CITE OR QUOTE
-------
APPENDIX A. EXPERIMENTAL PROTOCOL AND COMPOSITION OF EXPOSURE
ATMOSPHERES'
Facility/Sponsor
Reference
Engine type
Operating mode
Fuel type
Fuel sulfur
Exposure regime
Exposure conditions
Panicle cone, (mg/m3)
Particle size (fan)
MMD" (GSD)C
C02 (%)
CO (ppm)
NO, (ppm)
NO (ppm)
NO, (ppm)
SO, (ppm)
S04'2 (Mg/m3)
0,(%)
Ozone (ppm)
Aliphatic aldehydes
Formaldehyde (ppm)
Acrolein (ppm)
NH4:
TT J _ T. . . t . - \
i i% ui in*
-------
APPENDIX A. EXPERIMENTAL PROTOCOL AND COMPOSITION OF EXPOSURE ATMOSPHERES"
Facility/Sponsor
Reference
Engine type
Operating mode
Fuel type
Fuel sulfur
Exposure regime
Exposure conditions
Particle cone, (mg/m1)
Respirable particles"
(mg/m3)
Particle size Otm)
MMDC (GSD)"
CO, (%)
CO (ppm)
NO, (ppm)
NO (ppm)
SO, (ppm)
SO/2 (Mg/m3)
Aliphat. aldehydes (ppm)
Formaldehyde (ppm)
Acetaldehyde (ppm)
Acrolein (ppm)
NH, (ppm)
NH/
-------
APPENDIX A. EXPERIMENTAL PROTOCOL AND COMPOSITION OF EXPOSURE ATMOSPHERES"
Facility/Sponsor
Reference
Engine type
Operating mode
Fuel type
Fuel sulfur
Exposure regime
Exposure conditions
Particle cone, (mg/m3)
Respirable particles''
(mg/m3)
Particle size (jitn)
MMDC (GSD)d
C02(%)
CO (ppm)
NO, (ppm)
NO (ppm)
SO2 (ppm)
SO.'2 (/.g/m3)
Aliphatic aldehydes
Formaldehyde (ppm)
Acetaldehyde (ppm)
Acrolein (ppm)
NH3 (ppm)
NH + fnnm-V
THC (ppm)
JPAH (ug/m3)
Benzo(a)pyrene
Xitropyrane
National Institute for Occupational Safety and Health
Green etal.. 1983; Rabovsky et al., 1986
Caterpillar, 7 L, 4 cylinder, 4-cycle
8-mode mining cycle, 60% idling
No. 2 diesel
< 0.5%
7 h/d, 5 d/week, 12 mo.
Control
0.08
2.3
0.04
0.07
0.01
0.63
Exhaust
2
2.01
0.21
12.7
1.6
9.7
0.83
1.13
Coal dust
5
1.97
0.09
2.4
0.04
0.08
0.83
Exhaust +
coal dust
3
2.08
0.20
11.1
1.3
1.3
0.56
0.54
Rabovsky etal.. 1984
Caterpillar, 7 L, 4 cylinder, 4-cycle
8-mode mining cycle, 60% idling
No. 2 diesel
7 h/d, 5 d/week. 24 mo.
Control
0.07 ±
0.02
2.0 ±
0.9
0.06 ±
0.04
0.08 ±
0.13
0.5 ±
0.6
Exhaust
1.9 ± 0.3
0.16 ±
0.04
10.5 ±
2.3
1.5 ± 0.5
7.8 ±3.1
0.6 ± 0.4
0.6 ± 0.8
- - --
i
Coal dust
2.1 ± 0.4
0.08 ±
0.04
2.1 ±0.8
0.07 ±
0.05
0.08 ±
0.28
0.003 ±
0.05
0.6 ± 0.7
-- --
Exhaust +•
coal dust
2.0 ± 0.3
0.17 ±0.06
10.3 ± 2.0
1.5 ± 0.05
7.6 ± 2.8
0.5 ± 0.3
0.4 ± 0.3
* All ± are S.U. unless specified otherwise.
" < 7/xm.
c Mass median diameter
J Geometric standard deviation.
11/5/99
A-6 DRAFT-DO NOT CITE OR QUOTE
-------
APPENDIX A. EXPERIMENTAL PROTOCOL AND COMPOSITION OF EXPOSURE
ATMOSPHERES8
Facility/Sponsor
Reference
Engine type
Operating mode
Fuel type
Fuel sulfur
Exposure regime
Exposure conditions
Particle cone, (mg/m3)
Particle size Otm)
MMD" (GSD)
CO, (%)
CO (mg/m3)
NO, (ppm)
NO (ppm)
NO, (mg/m3)
Sulfur (mg/m3)
SO, (ppm)
Aliphatic aldehydes
Formaldehyde (ppm)
Acrolein (ppm)
NHd+
THC (ppm)
PAHs
Benzo(a)pyrene
Nitropyrene
General Motors Research Lab
Barnhart et al., 1981, 1982; Chaudhari etal., 1980,
1981; Chaudhari and Dutta, 1982; Chen and Vostal,
1981; Dziedzic, 1981; Eskelsonetal., 1981; Penney et
al., 1981: Misiorowski et al., 1980, 1981; Navarro et al.,
1981; Schneider and Felt, 1981; Schreck et al., 1980,
1981; Strom, 1984; Vostal et al., 1981; Wallace et al.,
1987; White and Garg, 1981
1978 350D Oldsmobile, 5.7 L, 4-cycle
1350 rpm, 96 N-m
Amoco type 2D
0.27%
20 h/d, 5'/4 d/week, 104 weeks
Control
0.007 ±
0.009
2.2 ± 0.6
0.05
Exhaust
0.258 ±
0.087
Exhaust
0.796 ±
0.228
Exhaust
1.533 ±
0.346
0.19
3.4 ± 0.8
2.1 ±0.6
5.3 ± 0.9
5.0 ± 1.2
7.9 ± 2.1
9.2 ± 1.6
Gross, 1981
5.7 L
1350 rpm, 96 N-m
Amoco type 2D
0.27%
20 h/d, 5'/i d/week, 87 weeks
Control
0.007
±0.009
1.9
<0.04
Exhaust
1.533
± 0.346
0.2
7
0.5
6.7
7.2
1.4
•
a All ± are S.D., unless specified otherwise.
h Mass median diameter.
11/5/99
A-7
DRAFT-DO NOT CITE OR QUOTE
-------
APPENDIX A. EXPERIMENTAL PROTOCOL AND COMPOSITION OF EXPOSURE
ATMOSPHERES'
Facility/Sponsor
Reference
Engine type
Operating mode
Fuel type
Fuel sulfur
Exposure regime
Exposure conditions
Particle cone, (mg/m3)
Particle size (/tm)
MMD" (GSD)C
C02 (%)
CO (ppm)
NO, (ppm)
NO (ppm)
SO, (ppm)
SO/2 tog/m3)
Aliphatic aldehydes
(ppm)
Formaldehyde (ppm)
Acrolein (ppm)
Ammonia (ppm)
Hydrocarbons (ppm)
PAHs
Benzo(a)pyrene
Nitropyrene
Inhalation Toxicology Research Institute
Bice et al., 1985; Cheng et al., 1984; Henderson et al., 1983, 1985, 1988; Mauderly et al.,
1983, 1984, 1987a, b, 1988; McClellan etal., 1986; Wolfetal., 1987
1980 Oldsmobile V8, 5.7 L
Federal Test Procedure, urban driving cycle
Phillips No. 2 diesel
0.34%
7 h/d, 5 d/week, 130 weeks
Control
0.013 ± 0.006
0.2005 ±
0.0390
1.0 ± 0.7
0
0
1.1 ±3.0
2.6 ± 0.6
Exhaust
0.353 ± 0.071
0.183 ±0.04(4.8 ±
0.28)d
0.262 ± 0.06 (4.2 ±
0.24)'
0.2284 ± 0.0371
2.9 ± 1.0
0.05 ± 0.09
0.7 ± 0.3
1.4 ± 1.3
3.8 ± 0.9
Exhaust
3.469 + 0.447
0.184 ±0.02(5.3 ±
0.64)d
0.249 ± 0.03 (4.5 ±
0.54)c
0.4355 ± 0.0590
16.5 ± 7.1
0.34 ± 0.22
5.7 ± 1.5
0.9 ± 0.9
8.7 ± 5.2
Exhaust
7.082 ± 0.808
0.213 ± 0.06 (4.7 ±
0.94)" 0.234 ± 0.06
(4.4 ± 0.88)'
0.6643 ± 0.1320
29.7 ± 12.9
0.68 ± 0.48
10.0 ± 2.6
0.7 ± 0.6
13.4 ± 8.3
' All ± are S.D. unless specified otherwise; data for particles through 30 mo.; data for gases from 35th week through 30 mo.
h Mass median diameter.
c Geometric standard deviation.
d Lovelace multiple jet impactor, mass median aerodynamic diameter.
e Impactor/parallel flow diffusion battery, mass median diameter.
11/5/99
A-8
DRAFT-DO NOT CITE OR QUOT1
-------
APPENDIX A. EXPERIMENTAL PROTOCOL AND COMPOSITION OF EXPOSURE ATMOSPHERES'
Facility/Sponsor
Reference
Engine type
Operating mode
Fuel type
Fuel sulfur
Exposure regime
Exposure conditions
Particle cone, (mg/m3)
Particle size (tan)
MMDC (GSD)"
C0,(%)
CO (ppm)
NO, (ppm)
NO (ppm)
NO, (ppm)
SO, (ppm)
SO;2 (Mg/m3)
0, (%)
Ozone (ppb)
Aliphatic aldehydes
Formaldehyde (ppm)
Acrolein (ppm)
Ammonia
Hydrocarbons (ppm)
HTHC (ppm)
PAHs
Benzo(a)pyrene
Nitropyrene
Inhalation Toxicology Research Institute
Inhalation Toxicology Research Institute
- Annual Report, 1980
1980 GM, 5.7 L
California 7-mode urban cycle
Phillips No. 2 diesel
7 h/d, 5 d/week. 12 weeks
Control
0.039
± 0.020
1.1 ±
0.6
2.8 ±
0.7
Exhaust
0.230
± 0.073
1.5 ±
0.6
3.2 ±
0.8
Exhaust
1.030
± 0.340
3.7 ±
1.1
2.9 ±
0.9
Exhaust
4.260
± 1.110
0.2080 ±
0.04
11.5 ±2.6
0.4 ± 0.4
0.80 ±
0.25
14.6 ± 3.1
2.5 ± 0.7
4.0 ± 0,8
Mauderly eta].,
1981"
1980 GM, 5.7 L
• California 7-mode urban cycle
Phillips No. 2 diesel
7 h/d, 5 d/week, 19 weeks
Control
0.050
± 0.024
Exhaust
0.210
± 0.070
Exhaust
1.020
± 0.350
Exhaust
4.380
± 1.160
' All ± are S.D. unless specified otherwise.
h Concentrations of gaseous components reported to be proportional to these in 12-week study.
c Mass median diameter.
11 Geometric standard deviation.
11/5/99
A-9
DRAFT-DO NOT CITE OR QUOTE
-------
APPENDIX A. EXPERIMENTAL PROTOCOL AND COMPOSITION OF EXPOSURE ATMOSPHERES' ||
Facility /Sponsor
Reference
Engine type
Operating mode
Fuel type
Fuel sulfur
Exposure regime
Exposure
conditions
Particle core.
(mg/m3)
Particle size (jim)
MMD"(GSD)C
CO, (%)
CO (ppm)
NO, (ppm)
NO (ppm)
NO, (ppm)
SO, (pprn)
SO/2 fcg/m')
O, (%)
Aliphatic
aldehydes
Formaldehyde
(ppm)
Acrolein (ppm)
NH/
LTHC (ppm)
HTHC (ppm)
PAHs (n«/m3)
Benzo(a)pyrene
Benzo(kjfiuorantnene
iBenzoi'ghi)perylene
1-Nitropyrene
Japan Automobile Research Institute Inc. (Health Effects Research Program - HERP) Jl
J
HERP 1988: Ishinishi et al., 1986; Ishinishi et al.. 1989 ^|
Light duty, 1 .8 L, 4-cylinder, swirl chamber
1700 rpm, eddy current dynamometer
Nippon Oil Co JIS No. 1 or 2 diesel
0.41%
16 h/d, 6 d/week, 30 mo.
Control
0.003
0.026
0.80
0.011
0.033
0.044
0.06
0.41
20.8
0.002
2.17
2.20
Exhaust
0.11
0.050
1.23
0.08
1.16
1.24
0.38
18.8
20.8
0.01
2.27
2.44
Exhaust
0.41
0.105
2.12
0.26
3.81
4.06
1.06
62.4
20.7
0.03
2.51
2.93
f
Exhaust
1.08
0.19
(2.37-
2.71)
0.219
3.96
0.70
9.44
10.14
2.42
151
20.5
0.07
2.87
3.82
Exhaust
2.32
0.21-0.22
(2.23-2.93)
0.418
7.10
1.41
18.93
20.34
4.70
315
20.3
0.13
3.57
5.49
5.3 ± 10.6
5.4 ± 7.7
2.7 ± 3.9
46.6 ±
44.0
Heavy duty. 1 1 L, 6-cylinder. direct injection
1200 rpm. eddy current dynamometer
Nippon Oil Co JIS No. 1 or 2 diesel
0.41%
16 h/d, 6 d/week, 30 mo.
Control
0.002
0.035
0.63
0.021
0.042
0.061
0.06
0.49
20.8
0.003
3.50
2.43
Exhaust
0.46
0.084
2.65
0.46
5.71
6.17
0.98
62.9
20.8
0.05 '
4.27
4.63
, 1
Exhaust
0.96
0.140
4.85
1.02
12.11
13.13
1.79
111
20.7
0.11
5.16
7.15
Exhaust
1.84
0.20-0.23
(2.73-3.07)
0.215
7.75
1.68
19.99
21.67
2.82
198
20.6
0.18
5.90
9.94
Exhaust
3.72
0.25-0.28
(2.75-3.18)
0.360
12.91
3.00
34.45
37.45
4.57
361 I
20.4
0.29
7.62
15.65
7.5 ••- 3.2
1 1
6.0 ± 3.0
8.9 ± 2.5
43.4 ±
9.8
1 All ± are S.D., unless specified otherwise.
" Mass median diameter.
0 Geometric standard deviation.
11/5/99
A-10
DRAFT—DO NOT CITE OR QUOTE
-------
APPENDIX A. EXPERIMENTAL PROTOCOL AND COMPOSITION OF EXPOSURE
ATMOSPHERES'
Facility/Sponsor
Reference
Engine type
Operating mode
Fuel type
Fuel sulfur
Exposure regime
Exposure conditions
Panicle cone, (mg/m3)
Particle size 0*m)
MMD" (GSD)C
CO, (%)
CO (ppm)
NO, (ppm)
NO (ppm)
NO, (ppm)
SO, (ppm)
SO/2 (/zg/m3)
0, (%)
Aliphatic aldehydes
Formaldehyde (ppm)
Acrolein (ppm)
NH/
LTHC (ppm)
HTHC (ppm)
PAHs
Benzo(a)pyrene
Nitropyrene
Japan Automobile Research Institute Inc. (Health Effects Research Program - HERP)
HERP, 1988; Ishinishi et al., 1986; Ishinishi et al., 1989
Heavy duty, 11 L, 6-cylinder, direct injection
1200 rpm, eddy current dynamometer
Nippon Oil Co JIS No. 1 or 2
0.41%
16 h/d, 6 d/week, 30 mo.
Control
0.004
0.068
0.06
0.024
0.040
0.062
0.03
0.35
20.8
0.003
3.62
2.38
Exhaust, filtered
0.005
0.083
2.54
0.42
5.16
5.58
0.96
1.43
20.7
0.04
4.43
3.74
Exhaust
0.39
0.084
2.50
0.44
5.37
5.81
0.98
57.7
20.7
0.04
4.41
4.53
Exhaust, filtered
0.019
0.391
13.00
3.96
32.81
36.76
4.50
1.61
20.4
0.24
7.79
12.68
Exhaust
2.99
0.31-0.35
(2.58-2.83)
0.412
12.90
4.95
31.50
36.45
4.03
358
20.3
0.20
7.68
13.79
1 All ± are S.D. unless specified otherwise.
" Mass median diameter.
c Geometric standard deviation.
11/5/99
A-ll
DRAFT—DO NOT CITE OR QUOTE
-------
APPENDIX A. EXPERIMENTAL PROTOCOL AND COMPOSITION OF EXPOSURE ATMOSPHERES'
Facility/Sponsor
Reference
Engine type
Operating mode
Fuel type
Fuel sulfur
Exposure regime
Exposure conditions
Particle cone, (mg/m3)
Particle size (pm) MMDb
CO, (%)
CO (ppm)
NO, (ppm)
NO(ppm)
NO, (ppm)
SO, (ppm)
so;2 (Mg/m3)
O, (vol%)
Aliphatic aldehydes
Formaldehyde (ppm)
Acrolein (ppm)
NIV
THC (ppm)
CH, (ppm)
PAHs (nzls. part.):
Benzo(a)pyrene
Benzo(e)pyrene
Benz(a)anthracene
Fluoranthene
Pyrene
Benzo(a)fluoramhene
Benzo(b)fluoranthene
Benzo(ghi)perylene
Chrysene
Fraunhofer Institut fur Toxikologie und Aerosolforschung
Heinrich et al. 1982
2.4 L
Constant load of 16kW, 2400 rpm
European reference fuel
0.36%
7-8 h/d, 5 d/week, 104 weeks
Exhaust
3.9 ± 0.5
0.1
0.54 ±0.15
18.5 ± 4.9
1.2 ± 1.7
16.5 ± 5.8
18.6 ± 5.8
3.1 ± 1.8
19.5 ± 0.6
9.3 ± 4.6
3.0 ± 2.2
7.0
14.1
9.8
134.6
65.8
5.4
5.3
21.4
25.7
Exhaust, filtered
0.52 ±0.1 3
18.0 ± 4.4
1.0 ± 1.5
17.2 ± 5.9
19.2 ± 6.1
2.8 ± 1.7
20.0 ± 0.7
7.9 ± 3.3
2.6 ± 1.8
Heinrich etal., 1986; Stober, 1986
1.6 L
FTP (1972)
European reference fuel
0.36%
19 h/d, 6 i/week, 120-140 weeks
Control
0.10 ± 0.01
0.16 ± 0.27
.
_
.
.
3.5 ± 0.29
2.3 ± 0.17
Exhaust
4.24 ± 1.42
0.35 ±0.10
0.38 ± 0.05
12.5 ± 2.18
1.5 ± 0.33
10.0 ± 2.09
11.4 ± 2.09
1.12 ± 0.89
5.5 ± 0.69
2.6 ± 0.19
3 (13 ng/m3)
- (21 ng/m3)
- (51 ng/m3)
Exhaust, filtered
0.35 ± 0.05
11.1 ± 1.92
1.2 ± 0.26
8.7 ± 1.84
9.9 ± 1,80
1.02 ± 0.62
.
5.2 ± 0.65
2.4 ± 0.20
' All ± are S.D. unless specified otherwise.
h Mass median diameter.
11/5/99
A-12 DRAFT-DO NOT CITE OR QUOTE
-------
APPENDIX A. EXPERIMENTAL PROTOCOL AND COMPOSITION OF EXPOSURE ATMOSPHERES'
Facility/Sponsor
Reference
Engine type
Operating mode .
Fuel type
Fuel sulfur
Exposure regime
Exposure conditions
Particle cone, (mg/m3)
Particle size (/im)"
CO, (%)
CO (ppm)
NO, (ppm)
NO (ppm)
NO, (ppm)
SO, (ppm)
SO/2 (M8/m3)
O, (vol%)
Aliphatic aldehydes
Formaldehyde (ppm)
Acrolein (ppm)
NH/
THC (ppm)
CH« (ppm)
PAHs
Benzo(a)pyrene
Nitropyrene
Fraunhofer Institut fur Toxikologie und Aerosolforschung
Heinrich et al., 1979: Meiss et al., 1981
2.4 L . .
Constant load of 16 kW, 2400 rpm
European reference fuel
0.36%
7-8 h/d, 5 d/week, 5 mo.
Control
0.1
<1
<1
6
Exhaust
4
0.1
0.5
11
0.6
25
26
3
8
5
Exhaust,
filtered
0.5
11
0.5
22
23
4
8
5
Exhaust
11
0.1
0.9
25
1.5
43
45
8
11
5
Exhaust,
filtered
0.95
27
1.3
43
44
8
12
5
Exhaust
17
0.1
1.4
42
2.6
75
78
13
13
5
Exhaust,
filtered
1.6
45
2.7
68
71
12
13.
5
* Values estimated from graphically depicted data.
h Aerodynamic diameter of the modal peak of the particle mass distribution.
11/5/99
A-13
DRAFT-DO NOT CITE OR QUOTE
-------
APPENDIX A. EXPERIMENTAL PROTOCOL AND COMPOSITION OF EXPOSURE
ATMOSPHERES"
Facility/Sponsor
Reference
Engine type
Operating mode
Fuel type
Fuel sulfur
Exposure regime
Exposure conditions
Particle cone, (mg/m3)
Particle size (jon)
C02 (%)
CO (ppm)
NO, (ppm)
NO (ppm)
NO, (ppm)
SO, (ppm)
SO/2 (nE/m3)
0, (%)
Aliphatic aldehydes
Formaldehyde (ppm)
Acrolein (ppm)
1 iNriV
Hydrocarbons (ppm)
PAHs
Benzo(a)pyrene
Nitropyrene
Southwest Research Institute
Kaplan et al., 1983; White et al., 1983
5.7 L
Steady state, 1347 rpm, equivalent to constant 40 mph
Emissions 2D
0.23-0.24%
20 h/d, 7 d/week, 65 weeks
Control
0.01 ± 0.009
0.0649 ±
0.0020
5.81 ±0.2
0
0.05 ± 0.0
3.43 ± 0.2
Exhaust
0.242 ±
0.049
88-93% <
1.0
79-85% <
0.5
0.0781 ±
0.0028
6.39 ± 0.3
0.56
0.65 ±0.1
3.76 ± 0.3
Exhaust
0.735 ± 0.084
88-94% <1.0
76-84% <0.5
0.1026 ±
0.0043
7.43 ± 0.3
1.69
1.85 ±0,2
'
4.31 ± 0.3
•
Exhaust
1.500 ±
0.136
91-94% <1.0
81-85% <0.5
0.1355 ±
0.0062
9.40 ± 0.5
3.42
3.73 ± 0.4
4.99 ± 0.3
Kaplan et al.,
1982
5.7 L
Steady state,
40 mph
20 h/d, 7
d/week,
12-13 weeks
Exhaust
1.500
t
~ • 1
* AH ± are S.D. unless specified otherwise.
11/5/99
A-14 DRAFT—DO NOT CITE OR QUOTE
-------
APPENDIX A. EXPERIMENTAL PROTOCOL AND COMPOSITION OF EXPOSURE ATMOSPHERES"
Facility/Sponsor
Reference
Engine type
Operating mode
Fuel type
Fuel sulfur
Exposure regime
Exposure conditions
Particle cone, (mg/m3)
Particle size (ion)
MMD" (GSD)C
CO, (%)
CO (ppm)
NO, (ppm)
NO (ppm)
NOX (ppm)
SO, (ppm)
SO/2 teg/m3)
0,(%) •
Aliphatic aldehydes
Formaldehyde (ppm)
Acrolein (ppm)
NH/
Hydrocarbons (ppm)
PAHs
Benzo(a)pyrene
Nitropyrene
Battelle-Geneva Research Center
Brightwell et al., 1986; Bernstein et al., 1 984
1.5L
FTP- 1972
16 h/d, 5 d/week, 104 weeks
Control
1 ±3
0.1 ±
Q.I
Exhaust
0.7
Exhaust
2.2
Exhaust
6.6
0.46 ±
0.03C
32 ± 11
5.8 ± 2.0°
8± 1
18.9 ± 4.1°
Exhaust,
filtered
0.47 ±
0.03'
32 ± 11
6.0 ± 2.(f
8±2
18.8 ± 4.1e
Japan Anti-Tuberculosis Association
Iwai et al., 1986
2.37 L
Steady state, 1000 tpm
8 h/d, 7 d/week, 96 weeks
Control
Exhaust,
filtered
7.0 ± 1.4"
1.8 + 1.8"
30.9 ± 10.9"
13.1 ±3.6"
Exhaust
4.9 ± 1.6
7.0 ± 1.4"
1.8 ± 1.8"
30.9 ±
10.9"
13.1 ± 3.6"
1 All ± are S.D. unless specified otherwise.
h Mass median diameter.
0 Geometric standard deviation.
d Samples from dilution tunnel, exposure chamber reported to have approximately the same concentrations.
e Data from first year of study (Bernstein et al., 1984).
11/5/99
A-15
DRAFT-DO NOT CITE OR QUOTE
-------
APPENDIX A. EXPERIMENTAL PROTOCOL AND COMPOSITION OF EXPOSURE
ATMOSPHERES'
Facility/Sponsor
Reference
Engine type
Operating mode
Fuel type
Fuel sulfur
Exposure regime
Exposure conditions
Particle cone, (mg/m3)
Resp. particles (mg/m3)
Particle size 0*m)
MMDb (GSD)C
CO, (%)
CO (ppm)
NO, (ppm)
NO (ppm)
NO, (ppm)
SO, (ppm)
SO/2 (w?/m3)
0, (%)
Aliphatic aldehydes (ppm)
Formaldehyde (ppm)
Acrolein (ppm)
Ammonia (ppm)
Hydrocarbons (ppm)
PAHs
Benzo(a)pyrene
Nitrqpyrene
Battelle, Pacific Northwest Laboratory
Karagianesetal., 1981
43 bhp, 3 cylinder
Simulated mining cycle
Equivalent to VV-F-800 A grade DF-2
_
6 h/d, 5 d/week, 87 weeks
Control
_
Exhaust
8.3 ± 2.0
95% respirable
0.71 (2.3)
50 ±3
4-6
<1
<1
26-40
Exhaust -I- coal
dust
13.5 ±4.0
a All ± are S.D. unless specified otherwise.
b Mass median diameter.
c Geometric standard deviation.
11/5/99
A-16
DRAFT-DO NOT CITE OR QUOTE
-------
APPENDIX A. EXPERIMENTAL PROTOCOL AND COMPOSITION OF EXPOSURE ATMOSPHERES'
Facility/Sponsor
Reference
Engine type
Operating mode
Fuel type
Fuel sulfur
Exposure regime
Exposure conditions
Particle cone, (mg/m3)
Particle size fron)
CO, (%)
CO (ppm)
NO, (ppm)
NO (ppm)
NO, (ppm)
SO, (ppm)
SO/2 (pg/m3)
0, (%)
Aliphatic aldehydes
Formaldehyde (ppm)
Acetaldehyde
Acrolein (ppm)
NH4+
Hydrocarbons (ppm)
Benzene (ppm)
Toluene (ppm)
PAHs («g/m3):
Benzo(a)pyrene
Nitropyrene
University of Pittsburgh
Battigelli, 1965
7 hp, four cycle, single cylinder
15-60 rain
Dilution A
0.1
<20
1.3
0.2
20.5
<1.0
<0.1
<0.05
<2.0
Dilution B
0.9
30
2.8
0.5
20.0
-------
APPENDIX A. EXPERIMENTAL PROTOCOL AND COMPOSITION OF EXPOSURE ATMOSPHERES'
Facility/Sponsor
Reference
Engine type
Operating mode
Fuel type
Fuel sulfur
Exposure regime
Exposure conditions
Particle cone.
(mg/nr1)
Particle size (/tin)
CO, (%)
CO (ppm)
NO2 (ppm)
NO (ppm)
NO, (ppm)
SO, (ppm)
H,SO« (ppm)
Oxidants
(ppm as O,)
Aliphatic aldehydes
Formaldehyde •
(ppm)
Acrolein (ppm)
NH,*
Jl Hydrocarbons
II r,,,. T- \
V^-n <~J ^*+tl
PAHs
Benzo(a)pyrene
Nitropyrene
U.S. Environmental Protection Agency
Gillespie. 1980; Hydeetal.. 1980; Malanchuk, 1980; Orthoefer, 1980: Staraetal., 1980
Automobile gasoline engine
Urban cycle
16 h/d, 7 d/week, 68 mo.
Control
4.9
0.04
0.04
0.03
-
0.02
•
2.7
Non-irradiated
gasoline exhaust
(R)
97.5 ± 10.0
0.05 ± 0.02
1.45 ±0.42
-
-
-
?l.S± 4.4
Irradiated
gasoline
exhaust (I)
94.5 ± 19.6
0.94 ± 0.36
0.19 ± 0.29
-
-
0.20 ± 0.09
SO2 +
HjSO4
-
-
-
0.42 ±
0.22
0.02 ±
0.01
-
... J
23.9 ± 6.1
"
R +
so, +
H,S"O4
98.4 ±
13.8
0.05 ±
0.03
1.51 ±
0.44
0.48 ±
0.23
0.02 ±
0.01
-
27.4 +
4.j
I +
SO, +
H,SO4
-
0.89 ±
0.36
0.19 ±
0.29
0.42 ±
0.21
0.03 ±
0.01
0.20 ±
0.08
23.9 ±
6.0
Nitrogen
oxides
- '
0.64 ±
0.12
0.25 ±
0.06
-
-
-
-
Nitrogen
oxides
-
0.15 ±
0.33
1.67 ±
0.21
-
" AH ± are S.D. unless specified otherwise.
11/5/99
A-18
DRAFT-DO NOT CITE OR QUOTE
-------
Appendix B
Models for Calculating Lung Burdens
11/5/99 B-1 DRAFT—DO NOT CITE OR QUOTE
-------
1 B.I. INTRODUCTION
2 As discussed in Chapter 4, the lung burden of diesel exhaust particles (DEPs) during
3 exposure is determined by both the amount and site of particle deposition in the lung and,
4 subsequently, by rates of translocation and clearance from the deposition sites. Mathematical
5 models have often been used to complement experimental studies in estimating the lung burdens
6 of inhaled particles in different species under different exposure conditions. This section
7 presents a mathematical model that simulates the deposition and clearance of DEPs in the lungs
8 of rats and humans.
9 Diesel particles are aggregates formed from primary spheres 15-30 nm in diameter. The
10 aggregates are irregularly shaped and range in size from a few molecular diameters to tens of
11 microns. The mass median aerodynamic diameter (MMAD) of the aggregates.is approximately
12 0.2 urn. The primary sphere consists of a carbonaceous core (soot) on which numerous kinds of
13 organic compounds are adsorbed. The organics normally account for 10% to 30% of the particle
14 mass. However, the exact size distribution of DEPs and the specific composition of the adsorbed
15 organics depend upon many factors, including engine design, fuels used, engine operating
16 conditions, and the thermodynamic process of exhaust. The physical and chemical
17 characteristics of DEPs have been reviewed extensively by Amann and Siegla (1982) and
18 Schuetzle(1983).
19 Four mechanisms deposit diesel particles within the respiratory tract during exposure:
20 impaction, sedimentation, interception, and diffusion. The contribution from each mechanism to
21 deposition, however, depends upon lung structure and size, the breathing condition of the
22 subject, and particle size distribution. Under normal breathing conditions, diffusion is the most
23 dominant mechanism. The other three mechanisms play only a minor role.
24 Once DEPs are deposited in the respiratory tract, both the carbonaceous cores and the
25 adsorbed organics of the particles will be removed from the deposition sites as described in
26 Chapter 4. There are two mechanisms that facilitate this removal: (a) mechanical clearance,
27 provided by mucociliary transport in the ciliated conducting airways as well as rnacrophage
28 phagocytosis and migration in the nonciliated airways; and (b) clearance by dissolution. As the
29 carbonaceous scot of DEPs is insoluble, it is removed from the lung primarily by mechanical
iSU clearance, whereas the adsorbed organics are removed principally by dissolution.
31
32 B.2. PARTICLE MODEL
33 To develop a mathematical model that simulates the deposition and clearance cf DEPs in
34 the lung, an appropriate model for diesel particles must be introduced. For the deposition study,
35 we employed an equivalent sphere model developed by Yu and Xu (1987) to simulate the
36 dynamics and deposition of DEPs in the respiratory tract by various mechanisms. For the
1175/99 B-2 DRAFT—DO NOT CITE OR QUOTE
-------
1 clearance study, we assume that a diesel particle is composed of three different material
2 components according to their characteristic clearance rates: (1) a carbonaceous core of
approximately 80% of the particle mass; (2) absorbed organics of about 10% of particle mass,
which are slowly cleared from the lung; and (3) adsorbed organics quickly cleared from the lung,
5 accounting for the remaining 10% of particle mass. The presence of two discrete organic phases
6 in the particle model is suggested by observations that the removal of particle-associated organics
7 from the lung exhibits a biphasic clearance curve (Sun et al., 1984; Bond et al., 1986), as
8 discussed in Chapter 4. This curve represents two major kinetic clearance phenomena: a fast-
9 phase organic washout with a half-time of a few hours, and a slow phase with a half-time that is a
10 few hundred tunes longer. The detailed components involved in each phase are not known. It is
11 possible that the fast phase consists of organics that are leached out primarily by diffusion
12 mechanisms while the slow phase might include any or all of the following components: (a)
13 organics that are "loosened" before they are released, (b) organics that have become intercalated
14 in the carbon core and whose release is thus impeded, (c) organics that are associated for longer
15 periods of time because of hydrophobic interaction with other organic-phase materials, (d)
16 organics that have been ingested by macrophages and as a result effectively remain in the lung
17 for a longer period of time because of metabolism by the macrophage (metabolites formed may
18 interact with other cellular components), and (e) organics that have directly acted on cellular
components, such as the formation of covalent bonds with DNA and other biological
macromolecules to form adducts.
21 The above distinction of the organic components is largely mechanistic and does not
22 specifically imply the actual nature of the organics adsorbed on the carbonaceous core; the
23 distinction is made to account for the biphasic clearance of DEPs. However, this distinction is
24 necessary in appreciating the dual-phase nature of DEPs. For aerosols made of pure organics,
25 such as benzo(a)pyrene (BaP) and nitropyrene (NP) in the same size range of DEPs, Sun et al.
26 (1984) and Bond et al. (1986) observed a nearly monophasic clearance curve. This might be
27 explained by the absence of intercalative phenomena (a) and of hydrophobic interaction imposed
28 by a heterogeneous mixture of organics (b). The measurement of a pure organic might also
29 neglect that quantity which has become intracellular (c) or covalently bound (d).
30
31 B.3. COMPARTMENTAL LUNG MODEL
32 To study the transport and removal of DEPs from the lungs, we used a compartmental
33 model consisting of four anatomical compartments: the nasopharyngeal or head (H),
34 tracheobronchial (T), alveolar (A), and lung-associated lymph node (L), as shown in Figure B-l.
35 In addition, we used two outside compartments B and G representing, respectively, the blood and
gastrointestinal (GI> tract. The alveolar compartment in the model is obviously the most
11/5/99 B-3 DRAFT—DO NOT CITE OR QUOTE
-------
B
(1)
rH
0
H
•*-{
(0
0
X
TG
(0
(i)
0)
o
0)
L
Figure B-l. Compartmental model of DEP retention.
\ important for long-term retention studies. However, for short-term consideration, retentions in
2 other lung compartments may also be significant. The presence of these lung compartments and
3 the two outside compartments in the model therefore provides a complete description of all
4 clearance processes involved.
5 In Figure B-1, r $ r$ and r ^are, respectively, the mass deposition rates of DEP material
6 component i (i=l [core], 2 [slowly cleared organics], and 3 [rapidly cleared organics]) in the
7 head, tracheobronchial, and alveolar compartments; and XJk represents the transport rate of
11/5/99
B-4
DRAFT—DO NOT CITE OR QUOTE
-------
1 material component i from any compartment X to any compartment Y. Let the mass fraction of
2 material component i of a diesel particle be//. Then
rn = fi rn ' (B-l)
rr = fi rr ' (B-2)
4
^-f^A . (B-3)
5
6 where rH, rT, and rA are, respectively, the total mass deposition rates of DEPs in the H, T, and A
7 compartments, determined from the equations:
8
rH = c(TV)(RF)(DF)H ,
9
rT = c(TV)(RF)(DF)T ,
10
rA = c(TV)(RF)(DF) .
ft fi
In Equations B-4 to B-6, c is the mass concentration of DEPs in the air, TV is the tidal
13 volume, RF is the respiratory frequency, and (DF)H, (DF)T, and (DF)A are, respectively, the
14 deposition fractions of DEPs in the H, T, and A compartments over a respiratory cycle. The
15 values of (DF)H, (DF)T, and (DF)A, which vary with the particle size, breathing conditions, and
16 lung architecture, were determined from our deposition model (Yu and Xu, 1987).
17 The differential equations for mfy, the mass of material component i in compartment X as
18 a function of exposure time t, can be written as
19
20 Head (H)
21
dmH _ (0 (0 (0 (0
fit rH HG^H ~ HBmH ' ^ '
22 Tracheobronchial (T)
dm(i)
= "+ *• ~ -
1175/99 B-5 DRAFT—DO NOT CITE OR QUOTE
-------
Alveolar (A)
Jmw
dt A
2 Lymph nodes (L)
dm(L
~~d~ '"AC'A "LEf'L
3 Equation B-9 may also be written as
dmA
dt A
4 where
n — n "4" A "^ n /T^ 1 ^\
/4 /4 7* >4£ y4^ " \^~ ^ ~)
5 is the total clearance rate of material component i from the alveolar compartment. In Equations
6 B-7 to B-10, we have assumed vanishing material concentration in the blood compartment to
7 calculate diffusion transport.
8 The total mass of the particle-associated organics in compartment X is the sum of m ^
9 and m ^the total mass of DEPs in compartment X is equal to
mx = mx* + m? + m? (B-13)
10 The lung burdens of diesel soot (core) and organics are defined, respectively, as
mLung = mrl) + <} ' (B-14)
11 and
(2H3) = 0) + m(2) + w(3) + m(3)
Lung r /4 r X \" LJt
Lung
1 2 Because the clearance of diesel scot from compartment T is much faster than from comDartment
1 3 A, m '/'< m ^ a short time after exposure, Equation B-14 leads to
^ • 0-16)
14 Solution to Equations B-7 tc B-10 can be obtained once all the transport rates ^ are
1 5 known. When >?$• are constant, which is the case in linear kinetics, Equations B-7 to B-l 0 will
1 6 have a solution that increases with time at the beginning of exposure but eventually saturates
1 7 reaches a steady-state value. This is the classical retention model developed by the International
1 1/5/99 B-6 DRAFT— DO NOT CITE OR QUOTE
-------
1 Commission of Radiological Protection (ICRP, 1979). However, as discussed in Chapter 4, data
2 have shown that when rats are exposed to DEPs at high concentration for a prolonged period, the
3 diesel soot accumulates in various peribronchial and subpleural regions in the lung and the long-
f termed clearance is impaired. This is the so-called overload effect, observed also for other
insoluble particles. The overload effect cannot be predicted by the classical ICRP model.
6 Soderholm (1981) and Strom et al. (1987, 1988) have proposed a model to simulate this effect by
7 adding a separate sequestrum compartment in the alveolar region. In the present approach, a
8 single compartment for the alveolar region of the lung is used and the overload effect is
9 accounted for by a set of variable transport rates Xj^ ^ and AA} which are functions of mA. The
10 transport rates /if and AA'}L in Equations B-7 to B-l0 can be determined directly from experimental
11 data on lung and lymph node burdens, and /$• and A.fy from Equation B-l2.
12
13 B.4. SOLUTIONS TO KINETIC EQUATIONS
14 Equation B-11 is a nonlinear differential equation of m Ai} with known function of AA}. For
15 diesel soot, this equation becomes
d^ ,„ '
_ (l) iCV™ \— (1) /TJ 1 T\
- rA ~ A-A \mA)mA • (D-if)
dt
A
Because clearance of the particle-associated organics is much faster than diesel soot, m(A}and
/^constitute only a very small fraction of the total particle mass (less than 1%) after a long
exposure, and we may consider JfA}as a function of m^alone. Equation B-17 is then reduced to
i» differential equation with mfjfihe only dependent variable.
20 The general solution to Equation B-17 for constant r^at any time, t, can be obtained by
21 the separation of variables to give
/i \
f
Jo
r(D .
A A A
If r^is an arbitrary function oft, Equation B-17 needs to be solved numerically such as b
a Runge-Kutta method. Once m^is found, the other kinetic equations B-7 to B-l 0 for both diesel
soot and the particle-associated organics can be solved readily, as they are linear equations. The
solutions to these equations for constant r$, r$ and rf are given below:
-
where M1? = Ml + Ml (B-20)
H HG HB
11/5/99 B-7 DRAFT—DO NOT CITE OR QUOTE
-------
1 Tracheobronchial (T)
m« = exp <-*« / ) / ( r« * X» »« ) exp (J» / ) * + „«
"•T rc "re U
2
3 Lymph nodes (L)
w^ = exp(-A.^ t) f' X®m®ex.p(k®) dt + mfy (B-23)
4 In Equations B-19 to B-23, m$0 represents the value of m^ at t = 0.
5 In the sections to follow, the methods of determining r§{ r% and r% or (DF)H, (DF)T, and
6 (DF)A r(DH\ T^DT> and ^D1A as well as the values of /$ in the compartmental lung model are
7 presented.
8
9 B.5. DETERMINATION OF DEPOSITION FRACTIONS
10 The mathematical models for determining the deposition fractions of DEPs in various
11 regions of the respiratory tract have been developed by Yu and Xu (1986, 1987) and are adopted
12 in this report. Yu and Xu consider DEPs as a polydisperse aerosol with a specified mass median
13 aerodynamic diameter (MMAD) and geometrical standard deviation og. Each diesel particle is
14 represented by a cluster-shaped aggregate within a spherical envelope of diameter de. The
15 envelope diameter de is related to the aerodynamic diameter of the particle by the relation
16
de _m ca £ ]/2
•j = (•£-) (^-) 0
a e o
17
18 where £ is the bulk density of the particle in g/cm3, C0 = 1 g/cm3; (j) is the packing density, which
19 is the ratio of the space actually occupied by primary particles in the envelope to the overall
20 envelope volume; and Cx is the slip factor given by the expression:
A .
Cx = 1 + 2A [L257 + 0.4 exp -(—r-* )] (B-25)
"x A
21 in which X = 8 x 10'6cm3 is the mean free path of air molecules at standard conditions. In the
22 diesel particle model of Yu and Xu (1986), C has a value of 1.5 g/cm3 and a d> value of 0.3 is
23 chosen based upon the best experimental estimates. As a result, Equation B-24 gives djd^ =
11/5/99 B-8 DRAFT—DO NOT CITE OR QUOTE
-------
1 1.35. In determining the deposition fraction of DEPs, de is used for diffusion and interception
2 according to the particle model.
B.5.1. Determination of (DF)H
5 Particle deposition in the naso- or oropharyngeal region is referred to as head or
6 extrathoracic deposition. The amount of particles that enters the lung depends upon the
7 breathing mode. Normally, more particles are collected via the nasal route than by the oral route
8 because of the nasal hairs and the more complex air passages of the nose. Since the residence
9 time of diesel particles in the head region during inhalation is very small (about 0.1 s for human
10 adults at normal breathing)-, diffusional deposition is insignificant and the major deposition
11 mechanism is impaction. The following empirical formulas derived by Yu et al. (1981) for
12 human adults are adopted for deposition prediction of DEPs:
13 For mouth breathing:
(DF)H .n = 0, for d^ 3000 (B-26)
(DF)H .n = -1.117 + 0.324 log(rf0, for dQ > 3000 (B-27)
H. ex = 0, (B-28)
and for nose breathing:
(DF)H .n = -0.014 + 0.023 Iog(rf20, for d*Q 337 (B-30)
(DF)Hgx = 0.003 H- 0.033 Iog(rf20, for cPQ s 215 (B-31)
+ °'399 \°Z(dl®> f°r dlQ > 215 (B-32)
15
16 where (DF)H is the deposition efficiency hi the head, the subscripts in and ex denote inspiration
17 and expiration, respectively, da is the particle aerodynamic diameter hi um, and Q is the air
18 flowrate in cmVsec.
19 Formulas to calculate deposition of diesel particles hi the head region of children are
20 derived from those for adults using the theory of similarity, which assumes that the air passage in
21 the head region is geometrically similar for all ages and that the deposition process is
11/5/99 B-9 DRAFT—DO NOT CITE OR QUOTE
-------
1 characterized by the Stokes number of the particle. Thus, the set of empirical equations from B-
2 26 through B-32 are transformed into the following form:
3 For mouth breathing:
(DF)H . = 0, for d2Q <> 3000 fB-3^
/T, in a ' V1-* -*-J/
(DF)H .n = -1.117 + 0.972 logK +
0.324 log( 3000 (B~34)
ex ~ v- (B-35)
and for nose breathing:
(DF)H .n = - 0.014 + 0.690 log K + 0.023 log(^20,
for d2Q z 337 " (B'36)
(DF)H .n = -0.959 f 1.191 log K +
0.397 log ( 337 C8'3^
(DF)^ ex = 0.003 + 0.099 log K +
0.033 log(J20, for d2Q '™* (B"38)
(DF)H ex = 0.851 + 1.197 log K
0.399 Iog(cf0, for dQ >215
2 (B"39)
a"
5 where K is the ratio of the linear dimension of the air passages in the head region of adults to that
6 of children, which is assumed to be the same as the ratio of adult/child tracheal diameters.
7 For rats, the following empirical equations are used for deposition prediction of DEPs in
8 the nose:
(DF)H .n = (DF)H ex = 0.046 +
0.009 \og(d2Q), for d2Q i 13.33
(DFL = (DF)U = -0.522 +
/7, Ifl rtt SX _-—.
0.514 log(W20, for d2Q > 13.33 (B"41)
9 B.5.2. Determination of (DF)T and (DF)A
10 The deposition model adopted for DEPs is the one previously developed for
11 monodisperse (Yu, 1978) and polydisperse spherical aerosols (Diu and Yu, 1983). In the model,
12 the branching airways are viewed as a chamber model shaped like a trumpet (Figure B-2). The
13 cross-sectional area of the chamber varies with airway depth, x, measured from the beginning of
14 the trachea. At the last portion of the trumpet, additional cross-sectional area is present to
15 account for the alveolar volume per unit length of the airways.
1175/99 B-10 DRAFT—DO NOT CITE OR QUOTE
-------
Summed Aheolar Croa Sectional Are*
Tr»d>ct
Ainrey Lcaftkx
Sectional ATM A»(x)
Figure B-2. Trumpet model of lung airways.
1 Inhaled diesel particles that escape capture in the head during inspiration will enter the trachea
2 and subsequently the bronchial airways (compartment T) and alveolar spaces (compartment A).
3 Assuming that the airways expand and contract uniformly during breathing, the equation
4 for the conservation of particles takes the form:
* fi - -
(B-42)
11/5/99
B-ll
DRAFT—DO NOT CITE OR QUOTE
-------
1 where c is the mean particle concentration at a given x and time t; A, and A2 are, respectively,
2 the summed cross-sectional area (or volume per unit length) of the airways and alveoli at rest; r)
3 is the particle uptake efficiency per unit length of the airway; p is an expansion factor, given by:
vt
P = 1 + y (B-43)
/
4 and Q is the air flow rate, varying with x and t according to the relation
Q v*
0- = l - Y * (B-44)
*^o I
5 where Q0 is the air flow rate at x = 0. In Equations B-43 and B-44, V, is the volume of new air in
6 the lungs and Vx and Vf are, respectively, the accumulated airway volume from x = 0 to x, and
7 total airway volume at rest.
8 Equation B-42 is solved using the method of characteristics with appropriate initial and
9 boundary conditions. The amount of particles deposited between location x, and x2 from time t,
10 to t2 can then be found from the expression
'2*2
DF = f JQc^dxdt (B-45)
'1*1
1 1 For diesel particles, t| is the sum of those due to the individual deposition mechanisms
12 described above, i.e.,
^ = "H/ + ^s + Up + ^D (B-46)
1 3 where r^, %, %, and T|D are, respectively, the deposition efficiencies per unit length of the airway
14 due to impaction, sedimentation, interception, and diffusion. On the basis of the particle model
1 5 described above, the expressions for T],, t|s, r|p, and T)D are obtained in the following form:
(B-47)
'1 1/3
sin' e] (B-48)
1b
11/5/99 B-12 DRAFT—DO NOT CITE OR QUOTE
-------
r]D = l[l-0.819exp(-14.63A) -
•Lt
0.0976 exp(-89.22A) - . (fi-5°)
0.0325 exp(-228A) - 0.0509 exp(-125A2/3)]
1 sfor Reynolds numbers of the flow smaller than 2000, and
nD= ^A1/2(l -0.444A1/2) (B-51)
2 for Reynolds numbers greater than or equal to 2000, where ST=cfaii/(18fjiR) is the particle Stokes
3 number, 0 = L/(8R), e = 3nu±/(32uR), P= dJR, and A = DL/(4R2u). In the above definitions u
i
4 is the air velocity in the airway; u is the air viscosity; L and R are, respectively, the length and
5 radius of the airway; us = C^f^JSju) is the particle settling velocity; and D = CJcTfi x/tdj is the
6 diffusion coefficient with k denoting the Boltzmann constant and T the absolute temperature. In
7 the deposition model, it is also assumed that r|, and rjp = 0 for expiration, while T|D and TIS have
8 the same expressions for both inspiration and expiration.
9 During the pause, only diffusion and sedimentation are present. The combined deposition
10 efficiency in the airway, E, is equal to:
E = 1 -(1 - £s) d-EJ . (B.52)
1J_ where ED and Es are, respectively, the deposition efficiencies due to the individual mechanisms
of diffusion and sedimentation over the pause period. The expression for E^ and Es are given by
3 A 3 ,
1 Y^ 4 / 2 w, V^ 4 \
1 - 2^ — exP(~ "fCnX1 ~ 2^ -?) exP
I = 1tt, X = 1 0?
13 where TD = Dt/R2 in which T is the pause time and a,, a2, and cc3 are the first three roots of the
14 equation:
J0W = ° ' (B-54)
15 in which J0 is the Bessel function of the zeroth order, and:
Es = 1.1094rs - 0.1604TJ, for 0 < T5 < 1. (B-55)
16 and
Es = 1 - 0.0061
for Tc > 1,
= 1 - 0.0069T"1 -0.0859TZ2 - O.C^»-.C ,
O O O
where TS = usT/2R.
11/5/99 B-13 DRAFT—DO NOT CITE OR QUOTE
-------
1 The values of (DF)T and (DF)A over a breathing cycle are calculated by superimposing DF
2 for inspiration, deposition efficiency E during pause, and DF for expiration in the
3 tracheobronchial airways and alveolar space. It is assumed that the breathing cycle consists of a
4 constant flow inspiration, a pause, and a constant flow expiration, each with a respective duration
5 fraction of 0.435, 0.05, and 0.515 of a breathing period.
6
7 B.5.3. Lung Models
8 Lung architecture affects particle deposition in several ways: the linear dimension of the
9 airway is related to the distance the particle travels before it contacts the airway surface; the air
10 flow velocity by which the particles are transported is determined by the cross-section of the
11 airway for a given volumetric flowrate; and flow characteristics in the airways are influenced by
12 the airway diameter and branching patterns. Thus, theoretical prediction of particle deposition
13 depends, to a large extent, on the lung model chosen.
14
15 B.5.3.1. Lung Model for Rats
16 Morphometric data on the lung airways of rats were reported by Schum and Yeh (1979).
17 Table B-1 shows the lung model data for Long Evans rats with a total lung capacity of 13.784
18 cm3. Application of this model to Fischer rats is accomplished by assuming that the rat has the
19 same lung structure regardless of its strain and that the total lung capacity is proportional to the
20 body weight. In addition, it is also assumed that the lung volume at rest is about 40% of the total
21 lung capacity and that any linear dimension of the lung is proportional to the cubic root of the
22 lung volume.
23
24 B.5.3.2. Lung Model for Human Adults
25 The lung model of mature human adults used in the deposition calculation of DEPs is the
26 symmetric lung model developed by Weibel (1963). In Weibel's model, the airways are assumed
27 to be a dichotomous branching system with 24 generations. Beginning with the 18th generation,
28 increasing numbers of alveoli are present on the wall of the airways, and the last three
29 generations are completely aleveolated. Thus, the alveolar region in this model consists of all the
3C airways in me iast seven generations, labie B-2 presents the morphometrie data of the airways
31 of Weibel's model adjusted to a total lung volume of 3000 cm3.
11/5/99 B-14 DRAFT—DO NOT CITE OR QUOTE
-------
Table B-l. Lung model for rats at total lung capacity
Generation
number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16"
17
18
19
21
22
25
24
Number of
airways
1
2
3
5
8
14
23
38
65
109
184
309
521
877
1,477
2,487
4,974
9,948
19,896
39,792
79,584
318,336
636,672
Length (cm)
2.680
0.715
0.400
0.176
0.208
0.117
0.114
0.130
0.099
0.091
0.096
0.073
0.075
0.060
0.055
0.035
0.029
0.025
0.022
0.020
0.019
0.017
0.017
Diameter (cm)
0.340
0.290
0.263
0.203
0.163
0.134
0.123
0.112
0.095
0.087
0.078
0.070
0.058
0.049
0.036
0.020
0.017
0.016
0.015
0.014
0.014
0.014
0.014
Accumulative
volume8 (cm)
0.243
0.338
0.403
0.431
0.466
0.486
0.520
0.569
0.615
0.674
0.758
0.845
0.948
1.047
1.414
1.185
1.254
1.375
1.595
2.003
2.607
7.554
13.784
"Including the attached alveoli volume (number of alveoli = 3 x 107, alveolar diameter = 0.0086 cm).
"Terminal bronchioles.
11/5/99
B-15 DRAFT—DO NOT CITE OR QUOTE
-------
Table B-2. Lung model by Weibel (1963) adjusted to
3000 cm3 lung volume
Generation
number
0
2
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16"
17
18
19
20
21
22
23
Number of
airways
1
2
4
8
16
32
64
128
256
512
1,024
2,048
4,096
8,192
16,384
32,768
65,536
131,072
262,144
524,283
1,048,579
2,097,152
4,194,304
8,388,608
Length (cm)
10.260
4.070
1.624
. 0.650
1.086
0.915
0.769
0.650
0.547
0.462
0.393
0.333
0.282
0.231
0.197
0.171
0.141
0.121
0.100
0.085
0.071
0.060
0.050
0.043
Diameter (cm)
1.539
1.043
0.710
0.479
0.385
0.299
0.239
0.197
0.159
0.132
0.111
0.093
0.081
0.070
0.063
0.056
0.051
0.046
0.043
0.040
0.038
0.037
0.035
0.035
Accumulative
volume2 (cm)
19.06
25.63
28.63
29.50
31.69
33.75
35.94
38.38
41.13
44.38
48.25
53.00
59.13
66.25
77.13
90.69
109.25
139.31
190.60
288.16
512.94
925.04-
1,694.16
3,000.00
'Including the attached alveoli volume (number of alveoli = 3 x 101, alveolar diameter = 0.0288 cm).
"Terminal bronchioles.
11/5/99
B-16
DRAFT—DO NOT CITE OR QUOTE
-------
1 B.5.3.3. Lung Model for Children
2 The lung model for children in the diesel study was developed by Yu and Xu (1987) on
the basis of available morphometric measurements. The model assumes a lung structure with
dichptomous branching of airways, and it matches Weibel's model for a subject when evaluated
5 at the age of 25 years, the age at which the lung is considered to be mature. The number and size
6 of airways as functions of age t (years) are determined by the following equations.
7
8 B.5.3.3.1. Number of airways and alveoli. The number of ttirways N((t) at generation i for age t
9 is given by
N.(f) = 2', for 0 < i <20 (B-57)
N..(t)=N(f),
{ ^(0=^3(0=0. for "# * 2 (B-58)
#210) = 221,
( N22(t) = Nr(t) -221, for 221 < Nf(t) i 222 (B-59)
JV23(0 = 0,
yo - 22>,
for Nr(f) > 221 + 222, (B-60)
o21 _922
^> <£
10 where Nr(t) is the total number of airways in the last three airway generations. The empirical
11 equation for Nr which best fits the available data is
, 2.036 x 107(1 -0.926e 'QA5t\ t < 8
1' ~ 1.468 x io7, / > 8
12
13 Thus, Nr(t) increases from approximately 1.5 million at birth to 15 million at 8 years of age and
14 remains nearly constant thereafter. Equations B-58 to B-60 also imply that in the last three
15 generations, the airways in the subsequent generation begin to appear only when those in the
16 preceding generation have completed development.
17 The number of alveoli as a function of age can be represented by the following equation
18 according to the observed data:
NA(f) = 2.985 x 108(1 -0.919e-°45r) (B-62)
19
11/5/99 B-17 DRAFT—DO NOT CITE OR QUOTE
-------
1 The number of alveoli distributed in the unciliated airways at the airway generation level
2 is determined by assuming that alveolization of airways takes place sequentially in a proximal
3 direction. For each generation, alveolization is considered to be complete when the number of
4 alveoli in that generation reaches the number determined by Weibel's model.
5
6 B.5.3.3.2. Airway size. Four sets of data are used to determine airway size during postnatal
7 growth: (a) total lung volume as a function of age; (b) airway size as given by Weibel's model;
8 (c) the growth pattern of the bronchial airways; and (d) variation in alveolar size with age. From
9 these data, it is found that the lung volume, LV(t) at age t, normalized to Weibel's model at 4800
1 0 cm3 for an adult (25 years old), follows the equation
LV(f) = 0.959 x 105(1 - 0.998e -°-0020 (cm3). (B-63)
11
1 2 The growth patterns of the bronchial airways are determined by the following equations
0/0 - D^ = a.[H(f) - #(25)], (B.64)
1 3 where D;(t) and Lf(t) are, respectively, the airway diameter and length at generation i and age t,
14 Diw and Liw the corresponding values for Weibel's model, a; and pf are coefficients given by
a. = 3.26 x I0-2exp[-1.183 (i+l)as] (B-66)
P. = 1.05 x lO'6 exp [10.1] (i+.l)-0-2] (B-67)
1 5 and H(t) is the body height, which varies with age t in the form
H(f) = 1.82 x 1Q2(1 - 0.725e'°140 (cm). (B-68)
16
1 7 For the growth patterns of the airways in the alveolar region, it is assumed that
D. L D
for 17 *i i 23 (B-69)
D L. D
r,v r,v {
18 where Da is the diameter of an alveolus at age t, Daw = 0.0288 cm is the alveolar diameter for
19 adults in accordance with Weibel's model, and f(t) is a function determined from
20
11/5/99 B-l8 DRAFT—DO NOT CITE OR QUOTE
-------
16
(LV(t) -
23
ETl •)
—D2L
i = 174 nv "' • Jo —- "
1
2 B.6. TRANSPORT RATES
3 The values of transport rates /i$ for rats have been derived from the experimental data of
4 clearance for diesel soot (Chan et al., 1981; Strom et al., 1987, 1988) and for the particle-
5 associated organics (Sun et al., 1984; Bond et al., 1986; Yu et al., 1991). These values are used
6 in the present model of lung burden calculation and are listed below:
A®, = 1.73 (/ = 1,2,3) (B-71)
= 0.00018 (B-72)
= 0.0129 (B-73).
= 12-55 (B-74)
A.W = 0.693 (i = 1,2,3) (B-75)
,AL = 0.00068 [1 - exp(-0.046mj62)] (B-76)
: 2,3) (B-77)
,AT = 0.012 exp(-0.11mj76) +
0.00068 exp(-0.046m!62) (/=.!,2,3) (B"78)
A
0.012 exp(-0.11m!76) + 0.00086 (B"79)
A
= A2 + ^AT + A?i = °-012 exp(-0.11mj-76) +
0.00068 exp(-0.046mo162) +0.0161 (B"80)
•AL ^S^2 = 0.012exp(-O.llm;76) +
0.00068 exp(-0.046m!62) + 15.7 (
where /^ is the unit of day"1, and mA = m ^ is the particle burden (in mg) in the alveolar
compartment.
11/5/99 B-l9 DRAFT—DO NOT CITE OR QUOTE
-------
1 Experimental data on the deposition and clearance of DEPs in humans are not available.
2 To estimate the lung burden of DEPs for human exposure, it is necessary to extrapolate the
3 transport rates /l$ from rats to humans. For organics, we assume that the transport rates are the
4 same for rats and humans. This assumption is based upon the observation of Schanker et al.
5 (1986) that the lung clearance of inhaled lipophilic compounds appears to depend only on their
6 lipid/water partition coefficients and is independent of species. In contrast, the transport rates of
7 diesel soot in humans should be different from those of rats, since the alveolar clearance rate, XA,
8 of insoluble particles at low lung burdens for human adults is approximately seven times that of
9 rats (Bailey et al., 1982), as previously discussed in Chapter 4.
10 No data are available on the change of the alveolar clearance rate of insoluble particles in
11 humans due to excessive lung burdens. It is seen from Equation B-79 that A ^for rats can be
12 written in the form
Xjj0 = a exp(-bm^) + d (B-82)
13 where a, b, c, and d are constants. The right-hand side of Equation B-82 consists of two terms,
14 representing, respectively, macrophage-mediated mechanical clearance and clearance by
15 dissolution. The first term depends upon the lung burden, whereas the second term does not. To
16 extrapolate this relationship to humans, we assume that the dissolution clearance term is
17 independent of species and that the mechanical clearance term for humans varies in the same
18 proportion as in rats under the same unit surface particulate dose. This assumption results in the
19 following expression for>2 ^/;in humans
X
-------
Table B-3. Ratio of pulmonary surface areas between humans
and rats as a function of human age
Age (year)
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
27
28
19
20
21
22
23
24
25
Surface area
4.99
17.3
27.6
36.7
44.7
51.9
58.5
64.6
70.4
76.0
81.4
86.6
91.6
96.4
101
106
110
115
119
123
128
132
136
140
144
148
11/5/99
B-21
DRAFT—DO NOT CITE OR QUOTE
-------
1 The equations for other transport rates that have a lung-burden-dependent component are
2 extrapolated from rats to humans in a similar manner. The following lists the values of 2& (in
3 day"1) for humans used in the present model calculation:
= 1.73 (i = 1,2,3) (B.84)
= °-00018 (B-85)
= °-0129 (B-86)
= 12'55 (B-87)
= 0.693 (/ = 1,2,3) (B-88)
A.J] = 0.00068 {1 - 0.0694 exp[-0.046(wy5)162]} (B-89)
*2. = \ *& (' = 2, 3) (B-90)
^. = 0.0694 (0.012 exp[-0.11(>zyS)176] +
.
0.00068 exp[-0.046(mS) ]} (i = 1, 2, 3)
1 76 (B~91)
0.0694 {0.012 exp[-0.11(7w7,S)176]} + 0.00086
A
A,(2) =
AL AT AB
0.0694{0.012 exp[-0.11(wy^)176] + (B-93)
0.00068 exp[-0 .046^/5)' 76]} + 0.016
(B-94)
4 B.7. RESULTS
5 B.7.1. Simulation of Rat Experiments
6 To test the accuracy of the model, simulation results are obtained on the retention of
7 diesel soot hi the rat lung and compared with the data of lung burden and lymph node burden
8 obtained by Strom et ai. (1988). A particle size of 0. 19 jam MMAD and a standard geometric
G deviation, aa, of 2.3 (as used iii Strom's expeiiriifciu) are uscu m uie calculation.
1 0 The respiratory parameters for rats are based on their weight and calculated using the
11 following correlations of minute volume, respiratory frequency,, and growth curve data.
12 Minute volume = 0.9 W (cmVmin) (B-95)
1 3 Respiratory frequency = 475 W03 (1/min) (B-96)
14
15
1 1 75/99 B-22 DRAFT— DO NOT CITE OR QUOTE
-------
1 where W is the body weight (in grams) as determined from the equation
2
W = 5+537T/(100+T), for T^56 days (B-97)
in which T is the age of the rat measured in days.
5 Equation B-95 was obtained from the data of Mauderly (1986) for rats ranging in age
6 from 3 mo to 2 years old; Equation B-96 was obtained from the data of Strom et al. (1988); and
7 Equation B-97 was determined from the best fit of the experimental deposition data. Figures B-3
8 and B-4 show the calculated lung burden of diesel soot (m ^-i- m ($ and lymph node burden,
9 respectively, for the experiment by Strom et al. (1988) using animals exposed to DEPs at 6
10 mg/m3 for 1, 3, 6, and 12 weeks; exposure in all cases was 7 days/week and 20 h daily. The solid
11 lines represent the calculated accumulation of particles during the continuous exposure phase and
12 the dashed lines indicate calculated post-exposure retention. The agreement between the
13 calculated and the experimental data for both lung and lymph node burdens during and after the
14 exposure periods was very good.
15 Comparison of the model calculation and the retention data of particle-associated BaP in
16 rats obtained by Sun et al. (1984) is shown in Figure B-5. The calculated retention is shown by
17 the solid line. The experiment of Sun et al. consisted of a 30-min exposure to diesel particles
18 coated with [3H\ benzo[a]pyrene (fH] - BaP) at a concentration of 4 to 6 ug/m3 of air and
19 followed by a post-exposure period of over 25 days. The fast and slow phase of (fH] - BaP)
^^ clearance half-times were found to be 0.03 day and 18 days, respectively. These correspond to A
21 A20 = 0.03 85 day'1 and A % - 23.1 day1 in our model, where A
-------
26 39
Time, week
CO
Figure B-3. The experimental and predicted lung burdens of rats to DEPs at a solid and
dashed concentration of 0.6 mg/m3 for different exposure spans. Lines are,
respectively, the predicted burdens during exposure and post-exposure.
Particle characteristics and exposure pattern are explained in the text. The
symbols represent the experimental data from Strom et al. (1988).
11/5/99
B-24
DRAFT—DO NOT CITE OR QUOTE
-------
en
o>
T3
w.
CD 3
o
T3
O
Q.
0 ****«cSwi=c
S^wk
4wk
, — —.—o-
U
«
26 39
Time, week
52
65
Figure B-4. Experimental and predicted lymph node burdens of rats exposed to CEPs at
a concentration of 6.0 mg/m3 for different exposure spans. The solid and
dashed lines are, respectively, the predicted burdens during exposure and
post-exposure. Particle characteristics and exposure pattern are explained
in the text. The symbols represent the experimental data from Strom et al.
(1988).
11/5/99
B-25 DRAFT—DO NOT CITE OR QUOTE
-------
0.8
Figure B-5. Comparison between the calculated lung retention (solid line) and the
• experimental data obtained by Sun et al. (1984) for the particle-associated
BaP in rats.
11/5/99
B-26 DRAFT—DO NOT CITE OR QUOTE
-------
Figure B-6. Calculated lung burdens of diesel soot per unit exposure concentration in
human adults exposed continuously to DEPs at two different concentrations
of 0.1 and 1.0 mg/m3. Exposure patterns are (a) 24 h/day and 7 days/week,
(b) 12 h/day and 7 days/week, and (c) 8 h/day and 5 days/week.
11/5/99
B-27
DRAFT—DO NOT CITE OR QUOTE
-------
Figure B-7. Calculated lung burdens of the particle-associated organics per unit
exposure concentration in human adults exposed continuously to DEPs at
two different concentrations of 0.1 and 1.0 mg/m3. Exposure patterns are
(a) 24 h/day and 7 days/week, (b) 12 h/day and 7 days/week, and (c) 8 h/day
and 5 davs/week.
11/5/99
B-28
DRAFT—DO NOT CITE OR QUOTE
-------
1 mg/cu. m
10 15
Age, year
20
25
Figure B-8. Calculated lung burdens of diesel soot per gram of lung per unit exposure
concentration in humans of different ages exposed continuously for 1 year to
DEPs of two different concentrations of 0.1 and 1.0 mg/m3 for 7 days/week
and 24 h daily.
11/5/99
B-29
DRAFT—DO NOT CITE OR QUOTE
-------
0.01
mg/cu. m
03
o
3
0.002
10 15
Age, year
20
25
Figure B-9. Calculated burdens of the particle-associated organics per gram of lung per
unit exposure concentration in humans of different ages exposed
continuously for 1 year to DEPs of two different concentrations of 0.1 and
1.0 mg/m3 for 7 days/week and 24 h daily.
11/5/99
B-30
DRAFT—DO NOT CITE OR QUOTE
-------
1 the amount of particle intake, the steady-state lung burdens per unit concentration were highest
2 for exposure pattern (a) and lowest for exposure pattern (b). Also, increasing soot concentration
from 0.1 to 1 mg/m3 increased the lung burden per unit concentration. However, the increase
was not noticeable for exposure pattern (c). The dependence of lung burden on the soot
5 concentration is caused by the reduction of the alveolar clearance rate at high lung burdens
6 discussed above.
7 Figures B-8 and B-9 show the effect of age on lung burden, where the lung burdens per
8 unit concentration per unit weight are plotted versus age. The data of lung weight at different
9 ages are those reported by Snyder (1975). The exposure pattern used in the calculation is 24
10 h/day and 7 days/week for a period of 1 year at the two soot concentrations, 0.1 and 1 mg/m3.
11 The results show that, on a unit lung weight basis, the lung burdens of both soot and organics are
12 functions of age, and the maximum lung burdens occur at approximately 5 years of age. Again,
13 for any given age, the lung burden per unit concentration is slightly higher at 1 mg/m3 than at 0.1
14 mg/m3.
15
16 B>8. PARAMETRIC STUDY OF THE MODEL
17 The deposition and clearance model of DEPs in humans, presented above, consists of a
18 large number of parameters that characterize the size and composition of diesel particles, the
19 structure and dimension of the respiratory tract, the ventilation conditions of the subject, and the
^P clearance half-times of the diesel soot and the particle-associated organics. Any single or
21 combined changes of these parameters from their normal values in the model would result in a
22 change in the predicted lung burden. A parametric study has been conducted to investigate the
23 effects of each individual parameter on calculated lung burden in human adults. The exposure
24 pattern chosen for this study is 24 h/day and 7 days/week for a period of 10 years at a constant
25 soot concentration of 0.1 mg/m3. The following presents two important results from the
26 parametric study.
27
28 B.8.1. Effect of Ventilation Conditions
29 The changes in lung burden due to variations in tidal volume and respiratory frequency
30 are depicted in Figures B-10 and B-l 1. Increasing any one of these ventilation parameters
31 increased the lung burden, but the increase was much smaller with respect to respiratory
32 frequency than to tidal volume. This small increase in lung burden was a result of the decrease in
33 deposition efficiency as respiratory frequency increased, despite a higher total amount of DEPs
34 inhaled.
35
11/5/99 B-31 DRAFT—DO NOT CITE OR QUOTE
-------
100
o
o
en
re
6
0.3
0.4 0.5 0.6
Tidal Volume, liter
Figure B-10.
Calculated lung burdens in human adults versus tidal volume in liters for
exposure to DEPs at 0.1 mg/m3 for 10 years at 7 days/week and 24 h daily.
Parameters used in the calculation are: (a) MMAD=0.2 pm, og=2.3,
/2=0.1, /3=0.1; (b) respiratory frequency = 14 min'1; and (c) lung volume
= 3000 cm3.
11/5/99
B-32
DRAFT—DO NOT CITE OR QUOTE
-------
60
50
40
O>
—-30
o
o
en
20
10
Soot
Organics
12 14 16
Respiratory Frequency, 1/min
1.4
1.2
O)
0.8
0.4
0.2
c
re
18
Figure B-ll. Calculated lung burdens in human adults versus respiratory frequency in
bpm for exposure to DEPs at 0.1 mg/m3 for 10 years at 7 days/week and 24
h daily. Parameters used in the calculation are: (a) MMAD=0.2 urn,
og=2.3, /2=0.1, /3=0.1; (b) tidal volume = 500 cm3, and (c) lung volume
= 3200 cm3.
11/5/99
B-33
DRAFT—DO NOT CITE OR QUOTE
-------
1 The mode of breathing has only a minor effect on Itmg burden because switching from
2 nose breathing does not produce any appreciable change in the amount of particle intake into the
3 lung (Yu and Xu, 1987). All lung burden results presented in this report are for nose breathing.
4
5 B.8.2. Effect of Transport Rates
6 Transport rates have an obvious effect on the retention of DEPs in the lung after
7 deposition. Because we are mainly concerned with the long-term clearance of diesel soot and the
8 associated organics, only the effects of two transport rates, /i ^and /ZJ^, are studied. Experimental
9 data of 2 ^ from various diesel studies in rats have shown that A % can vary by a factor of two or
10 higher. We use a multiple of 0.5 to 2 for the uncertainty in /I % and X (2A> to examine the effect
11 on lung burden. Figures B-12 and B-13 show respectively, the lung burden results for diesel soot
12 and the associated organics versus the multiples of X % and 2 (%} used in the calculation. As
13 expected, increasing the multiple of X % reduced the lung burden of diesel soot with practically
14 no change in the organics burden (Figure B-12), while just the opposite occurred when the
t
15 multiple of A % was increased (Figure B-13).
11/5/99 B-34 DRAFT—DO NOT CITE OR QUOTE
-------
Multiple of XX
Figure B-12. Calculated lung burdens in human adults versus multiple of A % for
exposure to DEPs at 0.1 mg/m3 for 10 years at 7 days/week and 24 h daily.
Parameters used in the calculation are: (a) MMAD=0.2 pm, og=2.3,
/2=0.1, /3=0.1; (b) tidal volume = 500 cm3, respiratory frequency = 14
min"1; and (c) lung volume = 3200 cm3.
11/5/99'
B-35 DRAFT—DO NOT CITE OR QUOTE
-------
-i.il/
0.8
1 1.2 1.4 1.6 1.8
Multiple of XA(2)
f C-' t . ..-;',"'
Figure B-13. Calculated lung burdens in human adults versus multiple of /2 ^ for
exposure to DEPs at 0.1 mg/m3 for 10 years at 7 days/week and 24'ft' daily.
Parameters used in the calculation are: (a) MMAD=0.2 fim
-------
1 B.9. REFERENCES
2
Amann, CA; Siegla, DC. (1982) Diesel particles - what are they and why. Aerosol Sci Technol 1:73-101.
Bailey, MR; Fry, FA; James, AC. (1982) The long-term clearance kinetics of insoluble particles from the human
6 lung. Ann Occup Hyg 26:273-289.
7 ~ "
8 Bond, JA; Sun, JD; Medinsky, MA; et al. (1986) Deposition, metabolism and excretion of l-[MC]nitropyrene and 1-
9 [MC]nitropyrene coated on diesel exhaust particles as influenced by exposure concentration. Toxicol Appl
10 Pharmacol 85:102-117.
11
12 Chan, TL; Lee, PS; Hering, WE. (1981) Deposition and clearance of inhaled diesel exhaust particles in the
13 respiratory tract of Fisher rats. J Appl Toxicol 1:77-82.
14
15 Diu, CK; Yu, CP. (1983) Respiratory tract deposition of polydisperse aerosols in humans. Am Ind Hyg Assoc J
16 44:62-65.
17
18 ICRP. (1979) Limits for intakes of radionuclides by workers. Ann ICRP 2. Publication 30, part 1.
19
20 Mauderly, JL. 1'986. Respiration of F344 rats in nose-only inhalation exposure tubes. J Appl Toxicol 6:25-30.
21 J .
22 Sclianker, LS; Mitchell, EW; Brown, RA. (1986) Species comparison of drug absorption from the lung after aerosol
23 inhalation or-intratracheal injection. Drug Metab Dispos 14(l):79-88.
24
25 Scheutzle, D. (1983) Sampling of vehicle emissions for chemical analysis and biological testing. Environ Health
26 Perspect 47:65-80.
27
28 Schum, M; Yeh, HC. (1979) Theoretical evaluation of aerosol deposition in anatomical models of mammalian lung
^fc airways. Bull. Math Biol 42:1-15.
31 Snyder, WS. (1975) Report of task group on reference man. Oxford, London: Pergamon Press, pp. 151-173.
32
33 Solderholm, SC. (1981) Compartmental analysis of diesel particle kinetics in the respiratory system of exposed
34 animals. Oral presentation at EPA Diesel Emissions Symposium, Raleigh, NC, October 5-7. In: Toxicological
35 effects of emissions from diesel engines (Lewtas J, ed.). New York: Elsevier, pp. 143-159.
36 , L
37 Strom, KA; Chan, TL; Johnson, JT. (1987) Pulmonary retention of inhaled submicron particles in rats: diesel
38 exhaust exposures and lung retention model. Research Publication GMR-5718. Warren, MI: General Motors
39 Research Laboratories. • <*-•
40
41 Strom, KA; Chan, TL; Johnson, JT. (1988) Inhaled particles VI. Dodgson, J; McCallum, RI; Bailey, MR; et al., eds.
42 Londorn-Pergamon Press, pp. 645-658.
43
44 Sun, JD; Woff, RK; Kanapilly, GM; et al. (1984) Lung retention and metabolic fate of inhaled benzo(a)pyrene
45 associated with diesel exhaust particles. Toxicol Appl Pharmacol 73:48-59.
46
47 Weibel, ER. (1963) Morphometry of the human lung. Berlin: Springer-Verlag.
48
49 Yu, CP. (1978) Exact analysis of aerosol deposition during steady breathing. Powder Technol 21:55-62.
50 * ' :
51 Yu, CP.; Diu7<£K;>Sbong, TT. (1981) Statistical analysis of aerosol deposition in nose and mouth. Am Ind Hyg
52 Assoe J 42:726,733^'
53
Yu, CP; Xu, GB. (1986) Predictive models for deposition of diesel exhaust participates in human and rat lungs.
Aerosol Sci Technol 5:337-347.
11/5/99 ..._., B-37 DRAFT—DO NOT CITE OR QUOTE
-------
1 Yu, CP; Xu, GB. (1987) Predicted deposition of diesel particles in young humans. J Aerosol Sci r8'4l9-429
2
3 Yu, CP; Yoon, KJ, Chen, (1991) Retention modeling of diesel exhaust particles in rats and humans Res Rep
4 Health Efflnst 40:1-24.
11/5/99 B-38 DRAFT—DO NOT CITE OR 00OTE
------- |