BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

Beskæftigelsesudvalget 2019-20
BEU Alm.del Bilag 101
Offentligt

Diesel exhaust

particles:

Scientific basis

for setting

a health-based

occupational

exposure limit

Anne Thoustrup Saber, Niels Hadrup, Sarah Søs Poulsen,

Nicklas Raun Jacobsen and Ulla Vogel

DIESEL EXHAUST PARTICLES: SCIENTIFIC

BASIS FOR SETTING A HEALTH-BASED

OCCUPATIONAL EXPOSURE LIMIT

Anne Thoustrup Saber

Niels Hadrup

Sarah Søs Poulsen

Nicklas Raun Jacobsen

Ulla Vogel

The National Research Centre for the Working Environment, Copenhagen 2018

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

NFA-report

Title

Diesel exhaust particles: Scientific basis for setting a health-based

occupational exposure limit

Anne Thoustrup Saber, Niels Hadrup, Sarah Søs Poulsen, Nicklas Raun

Jacobsen and Ulla Vogel

The National Research Centre for the Working Environment (NFA)

December 2018

nfa.dk

Authors

Institution

Publisher

Published

Internet version

The National Research Centre for the Working Environment (NFA)

Lersø Parkallé 105

DK-2100 Copenhagen

Phone: +45 39165200

Fax: +45 39165201

e-mail: [email protected]

Website: nfa.dk

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

OREWORD

In 2012, the International Agency for Research on Carcinogenicity (IARC) classified

diesel engine exhaust as carcinogenic to humans (Group 1) (IARC 2014). In 2014, the

Nordic Expert Group for Criteria Documentation of Health Risks from Chemicals (NEG)

and the Dutch Expert Committee on Occupational Safety (DECOS) co-produced a

criteria document on diesel engine exhaust (Taxell and Santonen 2016). NEG criteria

documents are used by the regulatory authorities of the Nordic countries as the scientific

basis for setting occupational exposure limits (OELs) for chemical substances. However,

NEG does not suggest OELs.

On this background and at request of the Danish Working Environment Authority, a

working group at the National Research Centre for the Working Environment (NFA) has

evaluated the possibility to establish a health-based OEL for diesel engine exhaust

particles.

Elizabeth Bengtsen and Karen Bo Frydendall, NFA, are gratefully acknowledged for

assistance with literature search.

The working group wishes to thank Chief Toxicologist Poul Bo Larsen, DHI, Denmark,

for reviewing the report.

Copenhagen, December 2018

iii

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

XECUTIVE SUMMARY

In this report, a working group at the National Research Centre for the Working

Environment reviewed data relevant to assessing the hazard of diesel exhaust particles

(DEPs), i.e. human studies (Chapter 2), toxicokinetics (Chapter 3), animal studies

(Chapter 4), mechanisms of toxicity (Chapter 5), previous risk assessments of DEPs

(Chapter 6), scientific basis for setting an occupational exposure limit (OEL) (Chapter 7)

and finally we summarize and suggest a health based OEL for DEPs (Chapter 8). The

focus of this report is only occupational exposure by inhalation.

Diesel is used as fuel for engines in vehicles for transport and power supply.

Occupational exposure to diesel exhaust occurs in many different sites, including

transportation, construction, railroad and mining industries (IARC 2014). Diesel exhaust

consists of a particulate phase (carbon particles with adsorbed organic matter) and a

gas/vapor phase which include volatile organic compounds, nitrogen oxides and carbon

oxides.

In 2012, the International Agency for Research on Carcinogenicity (IARC) classified

diesel engine exhaust as carcinogenic to humans (Group 1). IARC concluded that there is

sufficient evidence that diesel engine exhaust is causally related to lung cancer.

Additionally, a positive association has been observed between exposure to diesel engine

exhaust and increased risk of bladder cancer. Furthermore, IARC found sufficient

evidence for carcinogenicity of whole diesel engine exhaust (DEE), DEE particulate

matter and extracts of diesel exhaust particles (DEPs) in experimental animals. IARC

found inadequate evidence for the carcinogenicity of gas-phase DEE (i.e. particle free

DEE) (IARC 2014). This shows that the carcinogenic effect of DEE is driven by (DEPs).

Therefore, the present report focusses on the particulate phase of the DEE, the DEPs.

The present working group evaluated the relevant literature on DEPs from both

epidemiological and animal inhalation studies. However, since data regarding dose-

dependent effects (especially on cancer) following inhalation of DEE in humans are

available, data from animal studies are primarily included to strengthen the conclusions

drawn upon the human data and to allow comparison of health effects caused by other

types of particles.

Endpoints were evaluated based on reported adverse effects of diesel exhaust exposure

in reports and in the scientific literature. Especially the recent monography by IARC

(IARC 2014) and the criteria document by NEG/DECOS (Taxell and Santonen 2016) are

used as basis for the present report. The assessment by IARC evaluates cancer evidence

(IARC 2014). NEG/DECOS evaluates all health effects. Based on the evaluation,

NEG/DECOS regards lung cancer and pulmonary inflammation as the critical effects and

the present working group will therefore include both these endpoints.

A meta-regression of lung cancer mortality and cumulative exposure to elemental carbon

(EC)(as a proxy measure of DEPs) estimated the numbers of excess lung cancer deaths

for 45 years of occupational exposures of 1, 10, and 25 μg/m

EC to be 17, 200, and 689

per 10,000, respectively, by 80 years of age.

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

Pulmonary inflammation and carcinogenicity was observed in sub-chronic and chronic

inhalation studies in rats. Dose-response relationships were observed for inflammation

following inhalation of DEE and DEPs and instillation of DEPs in rodents. The working

group considers inflammation as a threshold effect.

The present working group found that there is evidence that both the extractable organic

fraction and the particulate fraction of DEP contribute to the carcinogenicity of DEP. The

molecular mechanisms include formation of bulky DNA adducts as well as oxidative

DNA damage likely induced by surface-dependent reactive oxygen species (ROS)

generation. In addition, inhalation of DEE and DEPs induced dose-dependent

pulmonary inflammation which could cause secondary genotoxicity. Thus, the available

data indicated induction of cancer through both direct and indirect genotoxic

mechanisms. Based on the observed mechanisms of genotoxicity, the present working

group concludes that DEP-induced mutagenicity and carcinogenicity occur by non-

threshold mechanisms.

The present working group identified a recent meta-analysis as suitable for risk

assessment (Vermeulen et al. 2014b). Five high quality chronic inhalation studies in rats

were identified, and the present working group decided also to select two of these for

calculation of excess cancer risk: A 2-year chronic cancer inhalation study in rats with

relatively low tumor incidence (0, 2.5 and 7 mg/m

) (Heinrich et al. 1995) and another 2-

year chronic inhalation study in rats with a relatively high tumor incidence (0.7, 2.2 and

6.6 mg/m

) (Brightwell et al. 1989) . In the table below excess lung cancer risk at 1 in 1

000, 1 in 10 000, and 1 in 100 000 using different approaches is presented.

Suggestion of an OEL for DEP calculated as elemental carbon

EC levels

Meta-

Rat inhalation

analysis of

study of DEE*

Human

Method I, Lung

Method II

Method II,

ECHA**

studies

burden (Heinrich)

ECHA**

(Vermeulen)

(Heinrich)

(Brightwell)

0.45 μg/m

5.6 μg/m

56 μg/m

15 μg/m

0.045 μg/m

0.56 μg/m

5.6 μg/m

1.5 μg/m

0.0045

0.056 μg/m

0.56 μg/m

0.15 μg/m

μg/m

Excess lung

cancer

incidence

1:1 000

1: 10 000

1: 100 000

*) For traditional diesel engine particles, it is assumed that 75% of the mass is elemental carbon

(Taxell and Santonen 2016)

**) European Chemicals Agency

Three different approaches were used for calculating excess lung cancer risk. First, lung

cancer risk was estimated based on the meta-analysis of epidemiological studies of the

association between exposure to DEE and lung cancer. Secondly, lung cancer risk was

estimated using two different approaches based on the same chronic inhalation study

(Heinrich et al. 1995). In the first approach, lung burden in rats after two years of

exposure was used to estimate the exposure limits for occupational exposure. In the

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

second approach, air concentrations were used directly. Thirdly, lung cancer incidence

was estimated based on a second chronic inhalation study in rats (Brightwell et al. 1989).

Independently of the applied method for risk assessment, the resulting exposure limits

were all very low.

The DEE exposure in the epidemiological studies was traditional DEE. Both of the

chronic inhalation studies were performed on traditional DEE. Typically, the proportion

of elemental carbon from a traditional heavy-duty diesel engine is 75% of the total

particle emission while this proportion is reduced to 13% when using “new technology”

diesel engines. Correspondingly, the proportions of sulfates are increased from 1% to

53% when exhaust after-treatment systems are applied (Taxell and Santonen 2016). The

present working group notes that there is limited available data on the biological effects

of DEP from “new technology” diesel engines and that the DEP concentrations in the

performed chronic inhalations studies with new technology engines in rats and mice

were likely too low to induce detectable levels of cancer.

The present working group notes that in chronic inhalation studies in rats, carbon black

nanoparticles and DEP have very similar carcinogenic potential (Heinrich et al. 1995).

Furthermore, the present working group notes that there is limited available data

regarding carcinogenicity of “new technology” DEE and there is no available evidence

suggesting that new technology DEP are less carcinogenic than DEP and carbon black

(CB).

The present working group notes that the risk estimates allowing 1: 10 000 excess lung

cancer cases or less are all close to the current ambient air concentrations of EC (ca. 0.4

μg/m

EC for rural measurements in Denmark (Massling et al. 2011) and 2.7 μg/m

levels on a major street in Copenhagen, Denmark (Palmgren et al. 2003).

The present working group recommends the approach using the epidemiological data to

derive OELs, since this approach relies on data from humans. Thus, the expected excess

lung cancer risk based on epidemiological data is 1: 1 000 at 0.45 μg/m

, 1: 10 000 at 0.05

μg/m

and 1: 100 000 at 0.005 μg/m

DEPs.

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

ANSK SAMMENFATNING

I denne rapport vurderer en arbejdsgruppe ved Det Nationale Forskningscenter for

Arbejdsmiljø data, der er relevante for at vurdere faren ved dieseludstødningspartikler,

dvs. humane studier (kapitel 2), toksikokinetik (kapitel 3), dyreforsøg (kapitel 4) ),

toksicitetsmekanismer (kapitel 5), tidligere risikovurderinger af

dieseludstødningspartikler (kapitel 6), det videnskabelige grundlag for fastsættelse af en

grænseværdi for dieseludstødningspartikler i arbejdsmiljøet (kapitel 7) og endelig

opsummeres og foreslås en helbredsbaseret grænseværdi for dieseludstødningspartikler

i arbejdsmiljøet (kapitel 8). Fokus i denne rapport er alene på erhvervsmæssig

eksponering ved indånding.

Diesel bruges som brændstof til motorer i køretøjer til transport og til strømforsyning.

Erhvervsmæssig eksponering for dieseludstødning forekommer i mange forskellige

brancher og erhverv, herunder ved transport, byggeri, jernbane og minedrift (IARC

2014). Dieseludstødning består af en partikelfase (carbonpartikler med adsorberet

organisk stof) og en gas/dampfase, der omfatter flygtige organiske forbindelser,

nitrogenoxider og carbonoxider.

I 2012 klassificerede WHO’s kræftagentur (IARC) udstødning fra dieselmotorer som

kræftfremkaldende for mennesker (Gruppe 1). IARC konkluderede, at der er

tilstrækkelig evidens for, at udstødning fra dieselmotorer forårsager lungekræft.

Derudover er der en positiv sammenhæng mellem eksponering for udstødning fra

dieselmotorer og øget risiko for blærekræft. Desuden fandt IARC tilstrækkeligt bevis for

kræftfremkaldende egenskaber for komplet udstødning fra dieselmotorer,

udstødningspartikler fra dieselmotorer og ekstrakter af udstødningspartikler fra

dieselmotorer i forsøgsdyr. IARC fandt utilstrækkelig evidens for kræftfremkaldende

egenskaber af gasfasen af udstødningen (dvs. partikelfri dieselmotorudstødning) (IARC

2014). Dette viser, at den kræftfremkaldende effekt af dieselmotorers udstødning er

drevet af dieselpartikler. Derfor er den nærværende rapport fokuseret på partikelfasen af

dieselmotorers udstødning, dieseludstødningspartiklerne.

Den nærværende arbejdsgruppe vurderede den relevante litteratur om

dieseludstødningspartikler fra både epidemiologiske studier og inhalationsstudier i dyr.

Da humane data om dosisafhængige virkninger (især på kræft) efter indånding er

tilgængelige, er data fra dyreforsøg primært medtaget for at styrke konklusionerne på

humane data og for at muliggøre sammenligning af helbredseffekter forårsaget af andre

typer af partikler.

Helbredseffekter forårsaget af udsættelse for dieseludstødning blev vurderet ud fra

rapporter og videnskabelig litteratur. Især anvendes den nylige monografi fra IARC

(IARC 2014) og et kriteriedokument om dieseludstødning fra Den Nordiske

Ekspertgruppe for dokumentation af helbredsrisiko ved kemikalier (NEG) og Den

hollandske komité for arbejdsmiljøsikkerhed (The Dutch Expert Committee on

Occupational Safety (DECOS)) (Taxell og Santonen 2016) som grundlag for denne

rapport. Bedømmelsen fra IARC vurderer evidens for kræft (IARC 2014). NEG/DECOS

vii

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

rapporten vurderer alle sundhedseffekter. På baggrund af deres evaluering vurderer

NEG/DECOS lungekræft og lungeinflammation som værende de kritiske effekter, og

forfatterne til den nærværende rapport vil derfor inkludere begge disse endepunkter.

En meta-regressionsanalyse af sammenhængen mellem lungekræftmortalitet og

kumulativ eksponering for elementært kulstof (anvendt som en proxy for

dieseludstødningspartikler) viste, at antallet af overskydende lungekræftdødsfald efter

45 års erhvervsmæssig eksponering for 1, 10 og 25 μg/m

elementært carbon til at være

henholdsvis 17, 200 og 689 pr. 10.000, ved en alder på 80 år.

Der blev observeret lungeinflammation og kræft i subkroniske og kroniske

inhalationsstudier af rotter. Dosis-respons-effekter blev observeret for inflammation efter

indånding af dieseludstødning og dieseludstødningspartikler samt instillation af

dieseludstødningspartikler i gnavere. Arbejdsgruppen betragter inflammation som en

tærskeleffekt.

Den nærværende arbejdsgruppe konstaterede, at både den ekstraherbare organiske

fraktion og partikelfraktionen af dieseludstødning bidrager til dieseludstødnings

kræftfremkaldende effekt. De molekylære mekanismer inkluderer dannelse af DNA-

addukter såvel som oxidative DNA-skader, der sandsynligvis induceres af

overfladeafhængig ROS-generering. Desuden inducerede indånding af dieselmotor

udstødning og dieseludstødningspartikler dosisafhængig lungeinflammation, som kan

forårsage sekundær genotoksicitet. De foreliggende data indikerede således induktion af

kræft gennem både direkte og indirekte genotoksiske mekanismer. Baseret på de

observerede genotoksicitetsmekanismer konkluderer den nærværende arbejdsgruppe, at

dieselpartikel-induceret mutagenicitet og kræft sker via ikke-tærskelmekanismer.

Den nærværende arbejdsgruppe identificerede en ny meta-analyse som egnet til

risikovurdering (Vermeulen et al. 2014b). Desuden blev der identificeret 5 kroniske

inhalationsundersøgelser hos rotter af høj kvalitet, og den nærværende arbejdsgruppe

besluttede at beregne overskydende kræftrisiko ud fra data fra to af disse studier: Et 2-

årigt kronisk kræftinhalationsstudie hos rotter med relativt lav tumorincidens (0, 2,5 og 7

mg/m

) (Heinrich et al. 1995) og et andet 2-års kronisk inhalationsstudie af rotter med en

relativt høj tumorincidens (0,7, 2,2 og 6,6 mg/m

) (Brightwell et al. 1989). I tabellen

præsenteres overskydende lungekræftrisiko ved 1 ud af 1 000, 1 ud af 10 000 og 1 ud af

100 000 ved anvendelse af forskellige fremgangsmåder.

viii

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

Forslag til grænseværdi for dieseludstødningspartikler beregnet som

elementært carbon

Overskydende Elementære

lungekræft-

carbon

Rotte-

incidens

niveauer

inhalationsstudie inhalationsstudie inhalationsstudie

Meta-

analyse af

dieseludstødning* dieseludstødning* dieseludstødning*

humane

Metode I, Lunge

Metode II

Metode II,

studier

burden (Heinrich) ECHA**

ECHA**

(Vermeulen)

(Heinrich)

(Brightwell)

1:1 000

0.45 μg/m

5.6 μg/m

56 μg/m

15 μg/m

1: 10 000

0.045 μg/m

0.56 μg/m

5.6 μg/m

1.5 μg/m

1: 100 000

0.0045

0.056 μg/m

0.56 μg/m

0.15 μg/m

μg/m

*) For partikler udledt fra traditionelle dieselmotorer antages det, at 75 % af massen er

elementært carbon (Taxell and Santonen 2016)

**) European Chemicals Agency

Der blev anvendt tre forskellige metoder til beregning af overskydende lungekræftrisiko.

Først blev risikoen for lungekræft vurderet ud fra meta-analysen af epidemiologiske

undersøgelser af sammenhængen mellem eksponering for udstødning fra dieselmotorer

og lungekræft. For det andet blev risikoen for lungekræft vurderet ved anvendelse af to

forskellige metoder baseret på den samme kroniske inhalationsundersøgelse (Heinrich et

al. 1995). Ved den første tilgang blev lungebyrden hos rotter efter to års eksponering

brugt til at estimere eksponeringsgrænserne for erhvervsmæssig eksponering. Ved den

anden tilgang blev luftkoncentrationerne anvendt direkte. For det tredje blev

lungekræftincidensen estimeret baseret på et andet kronisk inhalationsstudie i rotter

(Brightwell et al. 1989). Uafhængigt af den anvendte metode til risikovurdering, var de

resulterende eksponeringsgrænser alle meget lave.

I de epidemiologiske undersøgelser var eksponeringen fra dieselmotorer baseret på

”traditionel teknologi”. Begge de kroniske inhalationsundersøgelser blev udført på

udstødning fra dieselmaskiner baseret på ”traditionel teknologi”. Typisk er andelen af

elementært kulstof fra en traditionel dieselmotor 75 % af den totale partikelemission,

mens denne andel typisk reduceres til 13% ved anvendelse af dieselmotorer baseret på

"ny teknologi". Tilsvarende øges andelen af sulfater typisk fra 1% til 53%, når

udstødningsefterbehandlingssystemer anvendes (Taxell og Santonen 2016). Den

nærværende arbejdsgruppe bemærker, at der er begrænsede tilgængelige data om de

biologiske effekter af partikler fra dieselmotorer baseret på "ny teknologi", og at

partikelkoncentrationerne i de udførte kroniske inhalationsundersøgelser i mus og rotter

med ”nye teknologi”-motorer sandsynligvis var for lave til at fremkalde detekterbare

kræftniveauer.

Den nærværende arbejdsgruppe bemærker, at data fra kroniske

inhalationsundersøgelser i rotter viser, at carbon black nanopartikler og

dieseludstødningspartiklers kræftfremkaldende potentiale er sammenlignelige (Heinrich

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

et al. 1995). Den nærværende arbejdsgruppe bemærker endvidere, at data for kræft og

”ny teknologi” dieseludstødning er begrænset, og at der ikke foreligger nogen evidens

for, at ”ny teknologi” dieseludstødningspartikler er mindre kræftfremkaldende end

dieseludstødningspartikler fra ”traditionel teknologi” dieselmaskiner og carbon black.

Den nuværende arbejdsgruppe bemærker, at risikovurderingen, der tillader 1: 10 000

overskydende lungekræftsager eller mindre, ligger tæt på de nuværende omgivende

luftkoncentrationer af EC (ca. 0,4 μg / m

EC til landmålinger i Danmark (Massling et al.

2011 ) og 2,7 μg / m

EC niveauer på en hovedgade i København, Danmark (Palmgren et

al. 2003).

Den nærværende arbejdsgruppe anbefaler, at der til fastlæggelse af grænseværdi

anvendes tilgangen, som baserer sig på de epidemiologiske data, da denne tilgang er

baseret på data fra mennesker. Således er den forventede overskydende lungekræftrisiko

i forbindelse med erhvervsmæssig udsættelse for dieseludstødningspartikler 1: 1 000 ved

0,45 μg/m

, 1: 10 000 ved 0,05 μg/m

og 1: 100 000 ved 0,005 μg/m

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

ONTENTS

Foreword ....................................................................................................................................... iii

Executive summary...................................................................................................................... iv

Dansk sammenfatning ................................................................................................................ vii

Contents ......................................................................................................................................... xi

Abbreviations ............................................................................................................................... 12

Introduction.................................................................................................................................. 13

Human studies............................................................................................................................. 15

Human exposure ..................................................................................................................... 15

Epidemiological studies ......................................................................................................... 15

Toxicokinetics .............................................................................................................................. 20

Animal studies ............................................................................................................................. 21

Selection of studies and endpoints ....................................................................................... 21

Pulmonary inflammation ....................................................................................................... 21

Genotoxicity and cancer ......................................................................................................... 24

Mechanisms of toxicity ............................................................................................................... 31

Pulmonary inflammation, genotoxicity and cancer ........................................................... 31

Particle characteristics............................................................................................................. 35

Previous evaluations of diesel exhaust .................................................................................... 36

IARC (2012) .............................................................................................................................. 36

NEG/DECOS (2014)................................................................................................................. 36

Scientific basis for setting an occupational exposure limit .................................................... 38

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

BBREVIATIONS

ACES

BAL

DECOS

DEE

DEP

ECHA

HO-1

IARC

ICRP

LOAEC

LOAEL

NEG

NIST

NOAEC

NOAEL

NFA

OEL

PAH

ROS

SRM

TiO

TNF

VOC

Advanced collaborative emissions study

Assessment factor

Broncho alveolar lavage

Carbon black

Confidence interval

The Dutch Expert Committee on Occupational Safety

Diesel engine exhaust

Diesel exhaust particles

Elemental carbon

European Chemicals Agency

Heme oxygenase 1

Hazard ratio

The International Agency for Research on Cancer

International Commission on Radiological Protection

Interleukin

Intraperitoneal

Lowest observed adverse effect concentration

Lowest observed adverse effect level

The Nordic Expert Group for Criteria Documentation of Health Risks from

Chemicals

National Institute of Standards and Technology

Nitrogen monooxide

Nitrogen oxide

Nitrogen dioxide

No observed adverse effect concentration

No observed adverse effect level

National Research Centre for the Working Environment

Occupational exposure limit

Odd ratio

Polyaromatic hydrocarbons

Reactive oxygen species

Relative risk

Standard error

Standard reference materials

Titanium dioxide

Tumor necrosis factor

Volatile organic compounds

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

NTRODUCTION

Diesel is used as fuel for engines in vehicles for transport and power supply.

Occupational exposure to diesel exhaust occurs in many different sites, including

transportation, construction, railroad and mining industries (IARC 2014).

Diesel exhaust consists of a particulate phase (carbon particles with adsorbed organic

matter) and a gas/vapor phase which include volatile organic compounds, nitrogen

oxides and carbon oxides.

In 2012, the International Agency for Research on Carcinogenicity (IARC) classified

diesel engine exhaust (DEE) as carcinogenic to humans (Group 1). IARC concluded that

there is sufficient evidence in humans that DEE is causally related to lung cancer.

Additionally, a positive association has been observed between exposure to DEE and

increased risk of bladder cancer in humans. Furthermore, IARC found sufficient

evidence for carcinogenicity of whole DEE, DEE particulate matter and extracts of DEPs

in experimental animals. IARC found inadequate evidence for the carcinogenicity of gas-

phase DEE (i.e. particle free DEE) (IARC 2014). This shows that the carcinogenic effect of

DEE is driven by diesel exhaust particles (DEPs).

The present report focuses on the particulate phase of DEE and DEPs.

The first diesel-powered heavy duty diesel vehicles were introduced in the mid-19

century. These “traditional” diesel engines did not control the emission of particulate

matter. Increased concern for adverse health effects related to emissions from these

“traditional” diesel engines has led to the development of “new-technology” engines.

These “new-technology” engines are equipped with particulate filters that reduce the

emission of particulate matter considerably (more than 90% by mass)(IARC 2014;Taxell

and Santonen 2016). The composition of DEPs emitted from “traditional” and “new

technology” diesel engines is also different. Typically, the proportion of elemental

carbon from a traditional heavy-duty diesel engine is 75% of the particle emission while

this proportion is reduced to 13% when using new technology diesel engines. The

proportions of sulfates and organic carbon are increased from 1% to 53% and 19% to

30%, respectively, when exhaust after-treatment systems are applied (Taxell and

Santonen 2016).

Currently, “traditional” diesel engines are gradually being replaced by “new-

technology” diesel engines. However, it is expected to take a long time before this

transition has been completed (Taxell and Santonen 2016).

The Nordic Expert Group for Criteria Documentation of Health Risks from Chemicals

(NEG) and the Dutch expert Committee on Occupational Safety (DECOS) co-produced a

criteria document on DEE and have compiled occupational exposure limits (OELs) for

DEE levels in different countries (Taxell and Santonen 2016). These are presented in

Table 1.

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

Table 1. Occupational exposure limits (8-h TWAs) for diesel engine exhaust in

different countries (Table is adapted from (Taxell and Santonen 2016))

Country

Particles

(respirable

fraction)

(μg/m

)

100

(15 min STEL is:

400)

300

(15 min STEL is:

1200)

Elemental

carbon

(respirable

fraction) (μg/m

)

Total carbon,

Underground

mining

(μg/m

)

(ppm)

Austria

(Underground

mining)

Sweden

(general

occupational

exposure limit

for exhaust gas)

Switzerland

(underground

mining)

100

160

: Mine Safety and Health Administration, STEL: Short-term exposure limit,

Health-based occupational cancer risk values for DEE calculated by the DECOS has been

in public consultation.

The aim of the present report is to review the data and investigate if the present

knowledge allows for a suggestion of a health-based OEL for DEPs

In this document, we

review the relevant literature on the adverse effects of DEPs with the recent IARC

evaluation (IARC 2014) and the recent NEG/DECOS report (Taxell and Santonen 2016) as

central sources of information.

The risk assessment methodology of this report will follow the guidelines suggested by

REACH (ECHA 2012b). First, threshold or non-threshold effects are determined. For an

OEL based on threshold effects, the following traditional approach is made: 1)

identification of critical effect, 2) identification of the no observed adverse effect

concentration (NOAEC), 3) calculation of OEL using assessment factors (AFs) adjusting

for inter and intra species differences. For non-threshold effects, the current working

group will use two different approaches for calculating excess lung cancer risk. In the

first approach, lung burden will be used to estimate the exposure levels. In the second

approach, air concentrations will be used directly. Conclusively, the calculated OELs

will be compared, and lastly, a recommended OEL for DEP exposure will be proposed.

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

UMAN STUDIES

Human exposure

IARC recently reviewed occupational DEE exposure data:

“Exposure to diesel exhaust occurs in many different occupational settings, including the mining,

railroad, construction and transport industries. The main determinants of exposure are the size,

number and use of diesel engines indoors and outdoors, and the degree of ventilation. Several

different markers of exposure have been used, such as elemental carbon, nitrogen oxides and

polyaromatic hydrocarbons (PAHs). A generally accepted proxy for levels of exposure to diesel

engine exhaust is elemental carbon, although this is not specific to diesel engine alone. Miners (in

settings where diesel engines are used) and tunnel construction workers are the most highly

exposed occupational groups, with average levels of exposure to elemental carbon above 100

µg/m

. Dock workers, diesel mechanics and maintenance personnel are exposed on average to

levels between 20 and 40µg/m

; train crews, construction workers and workers involved in

loading and unloading ships are exposed to levels of elemental carbon of around 10 µg/m

; and

professional drivers are exposed on average to lower levels of around 2 µg/m

. Levels of exposure

to elemental carbon vary largely within job titles, and these relative rankings can therefore vary in

specific situations. Furthermore, the composition of diesel engine exhaust differs between

occupational settings due to variations in use scenarios, operating conditions and engine

technology”(IARC

2014).

The exposure to DEP (measured as elemental carbon) varies from 2 to more than 100

μg/m

. The present working group notes that exposure to DEP (measured as elemental

carbon) varies at least 50-fold between occupational settings.

Epidemiological studies

A recent evaluation by NEG on DEE (Taxell and Santonen 2016) concluded that:

‘The

critical health effects of DEE are pulmonary inflammation and lung cancer.’

The present

working group has therefore decided to focus on these end points as the critical effects.

Furthermore, the present report is focused on the particulate phase of DEE, because

IARC found sufficient evidence for carcinogenicity of whole DEE, DEE particulate

matter and extracts of DEPs in experimental animals while IARC found inadequate

evidence for the carcinogenicity of gas-phase DEE (IARC 2014).

Inflammation

NEG/DECOS has recently reviewed the human inhalation studies on the inflammatory

effects of DEE (Taxell and Santonen 2016):

“In inhalation studies on human volunteers, applying exhaust from older technology diesel

engines, slight increases in airway resistance and pulmonary inflammatory markers were

observed after single exposures at 100 µg DEP/m

(~ 75 µg EC/m

, 0.2–0.4 ppm NO

). This was

the lowest exposure level applied in these studies and represents the overall lowest observed

adverse effect level (LOAEL) for pulmonary inflammatory effects of older technology diesel engine

exhaust.”

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

No human studies on the pulmonary effects of exhaust from “new technology” diesel

engines were identified.

Controlled exposure of humans to NO

has been shown to result in pulmonary

inflammation (HEI, 2012).

The present working group notes that pulmonary inflammation was observed in

humans exposed to a single exposure of diesel exhaust containing 100 μg DEP/m

(~ 75

μg EC/m

). Since this was the lowest exposure level applied, it was not possible to

establish a no observed adverse effect concentration (NOAEC) for DEP-induced

inflammation.

Since exposure to NO

may cause pulmonary inflammation in humans, it is likely that

contributes to DEE-induced pulmonary inflammation. Therefore, the present

working group does not regard inflammation as a suitable DEP- induced critical effect.

Cancer

IARC evaluated the epidemiological studies on exposure to diesel exhaust and risk of

cancer (IARC 2014). IARC concluded

’There is sufficient evidence in humans for the

carcinogenicity of diesel engine exhaust. Diesel engine exhaust causes cancer of the lung. A

positive association has been observed between exposure to diesel engine exhaust and cancer of the

urinary bladder’.

IARC found inadequate evidence for the carcinogenicity of gas-phase DEE (i.e. particle

free DEE) (IARC 2014). This shows that the carcinogenic effect of diesel exhaust is driven

by DEPs. The present working group therefore considers lung cancer as the relevant

critical effect for diesel exhaust.

A meta-analysis of epidemiological studies was performed including studies with

information regarding dose-response relationship between exposure to diesel exhaust

quantified as elemental carbon and risk of lung cancer identified at the time of the IARC

evaluation (Vermeulen et al. 2014b).

The meta-analysis included all the epidemiological studies on diesel exhaust exposure

and lung cancer that fulfilled the following criteria: a) diesel engine exhaust exposure

was expressed as cumulative elemental carbon in the exposure response analysis, b) an

appropriate unexposed/low exposure reference group was used, c) no major

methodological shortcomings were identified. The identified studies which were

included in the IARC evaluation were one American study of non-metal miners (Attfield

et al. 2012;Silverman et al. 2012) and two American studies of trucking industry workers

(Garshick et al. 2012;Steenland et al. 1998). In addition, a fourth study of potash miners

(Möhner et al. 2013) that was published after the IARC evaluation was identified.

However, this study was not included in the meta-analysis. The reasons for this were: a)

the reference group had a relatively high exposure, b) the study contributed with only 68

lung cancer cases, and c) the derivation of the elemental carbon metric was not described

in detail.

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

The three studies were used as described in (Vermeulen et al. 2014b):

‘From the nested case–control study by Steenland et al. (1998) of trucking industry workers, we

used odds ratios (ORs) for cumulative elemental carbon (EC) exposure categories with a 5-year

lag. Steenland et al. (1998) included 994 lung cancer deaths and 1,085 controls. All cases and

controls had died in 1982–1983 and were long-term Teamsters enrolled in the pension system.

Cases and controls were divided by job categories based on the longest held job. In 1988–1989,

submicrometer elemental carbon (EC) was measured in 242 samples covering the major job

categories in the trucking industry (Steenland et al. 1998). Estimates of past exposure to

elemental carbon (EC), for participants in the epidemiologic study were made by assuming that a)

average 1990 levels for a job category could be assigned to all subjects in that job category, and b)

levels prior to 1990 were directly proportional to vehicle miles traveled by heavy duty trucks and

the estimated emission levels of diesel engines.’

‘From the cohort study of trucking industry workers by Garshick et al. (2012), we used hazard

rations (HRs) for cumulative elemental carbon (EC) exposure categories with a 5-year lag based

on analyses that excluded mechanics. In that study, work records were available for 31,135 male

workers employed in the unionized U.S. trucking industry in 1985. Mortality was ascertained

through the year 2000 and included 779 lung cancer deaths. From 2001 through 2006 a detailed

exposure assessment was conducted (> 4,000 measurements) that included personal and work-

area submicrometer elemental carbon (EC) measurements covering the major job categories in the

trucking industry. Exposure models based on terminal location in the United States were

developed. Historical trends in ambient terminal elemental carbon (EC) were modeled based on

historical trends in the coefficient of haze, a measurement of visibility interference in the

atmosphere. In addition to changes in ambient exposure, the historical model accounted for

changes in job-related exposures based on a comparison of elemental carbon (EC) measurement

data obtained in 1988 through 1989 with the newly collected elemental carbon (EC)

measurements.’

‘From the nested case–control miner study by Silverman et al. (2012), we used odd ratios (ORs)

for cumulative elemental carbon (EC) with a 15-year lag; we chose to use risk estimates from the

nested case–control study instead of estimates from the cohort analysis (Attfield et al. 2012)

because of their control for confounding, particularly from smoking, in the nested case–control

study. The case–control study was nested within a cohort of 12,315 workers in eight non-metal

mining facilities and included 198 lung cancer deaths and 562 incidence density–sampled

controls. Respirable elemental carbon (EC) was estimated for each surface and underground job

from the year of introduction of diesel-powered equipment in the facilities to 31 December 1997.

Between 1998 and 2001, a detailed exposure assessment was conducted measuring personal

respirable elemental carbon (EC) levels (> 700 measurements) covering the majority of job titles in

the facilities (Stewart et al. 2010). These estimates were back-extrapolated for underground jobs

per mine based on historical carbon monoxide measurement data and diesel engine exhaust

(DEE)-related determinants (e.g., diesel engine horsepower and ventilation rates).’

Study-specific categorical relative risk (RR) estimates for lung cancer mortality associated with

cumulative diesel exhaust particle levels (measured as elemental carbon) relative to the lowest

category of exposure for each study.”

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

The obtained exposure-response relationships for the three individual studies and for all

the studies combined are shown in table 2 (Vermeulen et al. 2014b):

Table 2. Exposure-response estimates (lnRR for a 1-µg/m

-increase in elemental

carbon from individual studies and the primary combined estimate based on a log-

linear model (reproduced from (Vermeulen et al. 2014b)).

Log-linear risk model (lnRR

= intercept + β x exposure. Exposure defined as elemental carbon in

μg/m

Thus, according to the meta-analysis, the risk of lung cancer from exposure to DEPs

measured as elemental carbon can be determined by:

lnRR (lung cancer)=0.088 +0.000982 μg/m

-years EC

The slopes for the three studies included in the meta-analysis (i.e., the lnRR estimated for

a 1-μg/m

-year increase in EC) were within a factor of two, and 95% confidence intervals

(CIs) overlapped (Table 1). The combined slope estimate was 0.000982 (95% CI: 0.00055,

0.00141). The exposure-response curve from the meta-analysis is shown in figure 1

(Vermeulen et al. 2014b).

Figure 1. Predicted exposure–response curve based on a log-linear regression model

using relative risk (RR) estimates from three cohort studies of DEE and lung cancer

mortality. Individual RR estimates [based on hazard ratios (HRs) reported by Garshick et

al. (2012) or odds ratios (ORs) reported by Silverman et al. (2012) and Steenland et al.

(1998)] are plotted with their 95% CI bounds indicated by the whiskers. The shaded area

indicates the 95% CI estimated based on the log-linear model. The insert presents the

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

estimates of the intercept and beta slope factor, the standard error (SE) of these estimates,

and the associated p-values. Reproduced from (Vermeulen et al. 2014b).

The meta-analysis by Vermeulen et al. has been extensively discussed in the literature

(Vermeulen and Portengen 2016;Vermeulen and Portengen 2017;Möhner and Wendt

2017;Möhner 2017a;Möhner 2017b;Vermeulen et al. 2014a) especially regarding the

choice of different time lags in the three included studies. In a sensitivity analysis,

different lags (0, 5, 10, 15 years) were used, the study by Möhner et al. (Möhner et al.

2013) was included or the highest dose from the study by Silverman (Silverman et al.

2012) was excluded. All these changes had limited effect on the slopes (β) of the risk

estimates, which varied from 0.000608 to 0.001021. The risk estimates were further

examined in another publication (Vermeulen and Portengen 2016), where the meta-

analyses from the above-mentioned sensitivity analysis were further elaborated and the

corresponding acceptable air concentrations of elemental carbon leading to acceptable

risk (4 x 10

-5

or one in 25 000 individuals) or maximum tolerated risk (4 x 10

-3

or one in

250 individuals) were estimated. The exposure levels for acceptable and maximum

tolerated risks were estimated to be 0.011 μg/m

and 1.03 μg/m

, respectively, for the

published meta-analysis, and the different estimates varied less than two-fold

(Vermeulen and Portengen 2016).

The present working group considers this study by Vermeulen et al. as an important and

relevant study that builds on the critical literature review performed by members of the

IARC working group. Moreover, it provides information on dose-response relationship

between exposure to diesel exhaust based on exposure measurements and risk of lung

cancer.

The present working group furthermore notes that the studies use cohorts of workers

and nested case-control designs thus minimizing the risk of potential confounding

caused by population-based comparison groups. Moreover, all three studies include

exposure assessment in terms of EC exposure measurements.

The present working group is of the opinion that the meta-analysis can be used for

quantitative risk assessment of DEPs even though the exposure was DEE and not only

DEPs. Animal exposure studies have clearly shown that DEE and DEPs are carcinogenic,

whereas filtered DEE is not (IARC 2014; HEI, 2012), thus showing that it is the

particulate fraction of diesel exhaust that causes lung cancer.

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

OXICOKINETICS

Exposures to DEPs occur in many occupational settings; primarily via inhalation but to a

lesser extent also dermal exposure and secondary ingestion occur. Focus in this section

will be on inhalation, the most critical exposure pathway.

Toxicokinetics has recently been reviewed and summarized by NEG as follows:

“Upon inhalation of diesel exhaust, DEP deposition will occur throughout the respiratory tract,

with a majority of the particles reaching the alveolar region (Oravisjärvi et al. 2014;EPA 2002).

In 9 healthy volunteers, the measured total deposited mass and number fraction of DEP

[generated during both idling (60 µg DEP/m

) and transient driving (300 µg DEP/m

)] in the

respiratory tract was ~ 30% and ~ 50–65%, respectively, at rest, with a high intra-individual

variation. The mean total deposited respiratory dose was calculated to be 0.14 µg per µg

DEP/m

/hour (Rissler et al. 2012). Applying measurement data on DEP number-size

distributions and the International Commission on Radiological Protection (ICRP) lung

deposition model, Oravisjärvi et al. estimated that ~ 60% of the deposited DEP particles are

retained in the alveolar region. Heavy exercise was estimated to increase the total deposition by 4–

5-fold, and the alveolar deposition by 5–6-fold (Oravisjärvi et al. 2014). From the

tracheobronchial region, DEP is cleared by mucociliary clearance and removed into the

gastrointestinal system within 24 hours (WHO 1996). The main clearance mechanism for

particles in the alveolar region is phagocytosis by alveolar macrophages, and subsequent

movement within alveolar and bronchial lumen into the conducting airways followed by

mucociliary clearance. There are also data suggesting that DEP, similarly to other types of fine

particles, may, in particular at high exposure levels, translocate through the alveolar epithelium

into the interstitium, lymph nodes and possibly end up into the systemic circulation (EPA 2002).

The clearance rate is substantially lower from the alveolar region than from the tracheobronchial

region; the alveolar retention half-time was 60–100 days in

rats with a lung burden of ≤ 1 mg

DEP/lung (WHO 1996). At higher lung burdens, the retention half-time increases linearly due to

an overwhelming of the alveolar macrophage mediated clearance (“lung overload”). In humans,

the alveolar clearance rate is even lower than in rats; retention half-times of several hundred days

have been reported for insoluble particles (EPA, 2002). The metabolism of PAHs and other DEP-

adsorbed organics in the lungs may lead to the formation of reactive metabolites (448). The

clearance rate of particle associated PAHs from the lungs is lower than the clearance of the

substances inhaled as such.”(Taxell

and Santonen 2016).

In the human experimental exposure study described in the NEG/DECOS report (Rissler

et al. 2012), the deposited fraction of DEPs was 0.27-0.28. Sixty percent of the deposited

DEPs are retained in the alveolar region. This equals an alveolar deposition of 16.8%

(0.6*28%=16.8%). The present working group also notes that exercise may increase the

alveolar deposition by 5–6-fold. Furthermore, the present working group notes that the

retention half-time for insoluble particles in humans is several hundred days and longer

than the clearance rate observed in rats (60-100 days).

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

NIMAL STUDIES

Selection of studies and endpoints

Since data on effects (especially on cancer) following inhalation of DEPs in humans are

available, data from animal studies are primarily included to strengthen the conclusions

drawn upon the human data and to allow comparison of health effects caused by other

types of particles. Rats are the preferred animal model in particle toxicology and are

more sensitive than mice to particle-induced lung cancer and fibrosis.

IARC found sufficient evidence for the carcinogenicity of whole DEE, DEE particulate

matter and extracts of DEPs in experimental animals while there was inadequate

evidence for the carcinogenicity of gas-phase DEE (i.e. particle free DEE) (IARC 2014).

This shows that DEPs rather than the gas-phase of the diesel exhaust are causative

agents for the carcinogenic effects observed. Therefore, in the present report, the focus is

on the particulate phase of the DEE.

In the present report, inhalation studies will be prioritised. For the description of

toxicological endpoints and mechanism of toxicity, studies using pulmonary deposition

such as intratracheal instillation will be included because intratracheal instillation

studies make it possible to evaluate the effects of the particulate phase of the diesel

exhaust alone. Dose-response assessments, however, are in the current report solely

conducted based on sub-chronic and chronic inhalation studies.

Endpoints were evaluated based on reported adverse effects of diesel exhaust exposure

in reports and in the scientific literature. Especially the recent monography by IARC

(IARC 2014) and the criteria document by NEG/DECOS (Taxell and Santonen 2016) are

used as basis for the present report. The assessment by IARC evaluates cancer evidence

(IARC 2014). NEG/DECOS evaluates all health effects. Based on their evaluation

NEG/DECOS considers lung cancer and pulmonary inflammation as the critical effects

and the authors of the present report will therefore include both these endpoints.

Pulmonary inflammation

NEG/DECOS has recently reviewed the studies on inflammatory effects in rats following

pulmonary deposition of DEPs (Taxell and Santonen 2016):

“In long-term animal inhalation studies, inflammatory and histopathological changes in the lungs

have been detected in rats at 210 µg DEP/m

or above (~ 160 µg EC/m

, 0.2 ppm NO

). Rats

exposed to filtered exhaust at 1.1 ppm NO

(10 µg DEP/m

) showed mild bronchial hyperplasia

and shortening of cilia”.

No NOAELs were identified, as effects were observed with lowest

tested dose.

“In a long-term (130 weeks) inhalation study in rats applying exhaust from a new technology

diesel engine, mild alveolar and bronchial epithelial hyperplasia, mild fibrotic lesions, and a mild

progressive decrease in pulmonary function mainly in the smallest airways consistent with the

morphological changes were observed at 4.2 ppm NO

(12 µg DEP/m

, ~ 3 µg EC/m

determined to be the LOAEL of this study. Corresponding but slightly milder effects were

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

reported in the same study for rats exposed at 3.6 ppm NO

(13 µg DEP/m

) for 13 weeks. The

findings were largely associated with NO

. No histopathological changes were detected after a

130-week

exposure at ≤ 0.9 ppm NO

(5 µg DEP/m

, ~ 1 µg EC/m

) or a 13-week

exposure at ≤

1.0 ppm NO

(≤ 4 μg DEP/m

) leading to a NOAEL of 0.9 ppm NO

.”

There are a number of relevant diesel engine exhaust inhalation studies and intratracheal

instillation studies of DEP addressing pulmonary inflammation. The studies considered

most relevant are described below.

In a chronic inhalation study by Mauderly and co-workers, male and female rats were

exposed to exhaust from “traditional diesel” engine by inhalation (Mauderly et al. 1994).

The diesel exhaust was generated by light-duty engines burning certification fuel and

operating on an urban-duty cycle .The dosage regimen was a mass concentration of 2.5

or of 6.5 mg/m

for 16 h/day, 5 days/week for 3, 6, 12, 18 or 24 months. The low and high

concentration diesel exhaust contained 0.7 and 3.8 ppm NO

, respectively, and 8.8 and 23

, respectively. Neutrophils were measured in bronchoalveolar lavage fluid (BAL) in

lungs after 12 months of exposure. In terms of increased neutrophil numbers in BAL, the

mass concentrations of 2.5 mg/m

and 6.5 mg/m

were determined to be the no observed

adverse effect concentration (NOAEC) and the lowest observed adverse effect

concentration (LOAEC), respectively. Other endpoints investigated in the study included

additional BAL fluid endpoints, organ weight, as well as neoplastic lesions. The

incidence of animals with neoplastic lesions is further described in the paragraph on

cancer. Concerning other BAL fluid endpoints, lactate dehydrogenase and beta

glucoronidase were increased at both dose levels. Concerning lung weight, the weight

was increased for both DEP dosed groups at 18 months and at later time points. The

authors of the report suggested that this reflected the inflammatory, proliferative and

fibrotic lesions resulting from the exposure. Notably the relative lung weights (lung

weight/body weight) were not increased.

Sub-chronic (16 h/day, 5 days/week, 13 weeks) exposure of rats to exhaust from “new

technology” diesel engines (12-13 μg/m

DEP ~ 3μg EC/m

, ~4 ppm NO

) resulted in

pulmonary inflammation (increase in inflammatory markers in broncheoalveolar

lavage). In mice, a similar exposure resulted in increased number of neutrophils in

bronchoalveolar lavage). No inflammatory effects were detected in rats exposed to a

lower concentration (4-5 μg DEP /m

~ 1 EC μg/m

, 0.9-1.0 ppm NO

) (McDonald et al.

2012).

There are also a number of relevant short-term inhalation studies with pulmonary

inflammation as endpoint. Of these we consider the following to be the most important.

Campen et al. investigated the effects of diesel exhaust in mice. Mice were exposed to

diesel exhaust (0.5 and 3.6 mg/m

) for 3 days (6 h/day) in whole-body inhalation

chambers with or without particulates filtered. Increased pulmonary inflammation

(measured as PMNs) was detected in mice exposed at the highest concentration of whole

exhaust. No effect on the number of PMNs was detected in mice exposed to filtered

exhaust (Campen et al. 2005).

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

McDonald et al. compared lung inflammation in mice exposed by inhalation 6 h/day for

7 days to exhaust generated from a diesel engine in two different cases. In the first case,

the diesel engine was operated with socalled “2003” fuel and no exhaust after-treatment

was applied (234 μg DEP/m

, 0.04 ppm NO

). In the other case, the engine was operated

with low-sulfur-fuel and a particle trap (7 μg DEP/m

, 0.04 ppm NO

). Otherwise, the

engine was operated identically in both cases. Mice exposed to full diesel exhaust

operated with “2003” fuel had increased levels of lung inflammatory markers (including

tumor necrosis factor (TNF), IL-6 and HO-1). No significant effects of inflammatory

markers were observed in mice exposed to filtered exhaust from engine operated with

low-sulfur-fuel (McDonald et al. 2004) .

The inflammatory effect of DEP alone has also been evaluated in studies where rodents

have been exposed to standard reference materials (SRM) from the US National Institute

of Standards and Technology (NIST). Of these, we consider the following to be among

the most important:

Short-term effects of SRM1650 DEP originating from heavy-duty diesel engines on

markers of inflammation were evaluated in mice. Mice were exposed by inhalation to

either a single 90 min exposure (20 or 80 mg/m

DEP) or as 4 repeated 90 min exposures

(5 or 20 mg/m

). Inhalation of DEP induced a dose-dependent inflammatory response

with infiltration of neutrophils and elevated gene expression of IL-6 in the lungs of mice

(Dybdahl et al. 2004).

The inflammatory response to the SRM1650b DEP was evaluated in mice 1, 3 and 28

days after a single intratracheal instillation of 18, 54 or 162 μg/mouse. A time- and dose-

dependent inflammatory response was observed. The response had returned to baseline

28 days post-exposure (Kyjovska et al. 2015).

Exposure of rats to NO

has been shown to induce pulmonary inflammation (HEI, 2012).

The present working group notes that pulmonary inflammation has been detected in rats

following chronic exposure to diesel exhaust at 210 μg DEP/m

or above (~ 160 μg

EC/m

, 0.2 ppm NO

). Low grade inflammation was observed in mice and rats exposed

to a new technology diesel engine, but the effects were ascribed to the relatively high

levels of NO

rather than to the DEPs. The present working group further notes that the

studies by Campen et al. and McDonald et al. demonstrate that the inflammatory effects

are reduced when DEPs are filtered away. Furthermore, based on the studies using DEP

standard reference materials, it is noted that DEP alone induces a dose-dependent

inflammatory response in the lungs of mice. Since exposure to NO

may cause

pulmonary inflammation in rats, it is likely that NO

contributes to DEE-induced

pulmonary inflammation.

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

Genotoxicity and cancer

Genotoxicity and cancer are well studied, possible adverse effects of exposure to DEPs.

Cancer

IARC recently reviewed chronic diesel exhaust inhalation studies in animals. IARC

found sufficient evidence for carcinogenicity of whole DEE, DEE particulate matter and

extracts of DEPs in experimental animals. In contrast, IARC found inadequate evidence

for the carcinogenicity of gas-phase DEE in experimental animals (IARC 2014). The

studies on whole DEE were generated from fuels and engines produced before the year

2000 (IARC 2014).

Inhalation studies

According to the IARC evaluation, whole diesel exhaust has been tested for

carcinogenicity by inhalation exposure in 19 studies in rats, 4 studies in mice, 3 studies in

hamsters and 1 study in monkeys. The study in monkeys was a 2-year inhalation study

and therefore too short for an evaluation of cancer. NEG/DECOS concluded that there

was no clear evidence of carcinogenicity of diesel exhaust in mice or hamsters even at

high particle loads (Taxell and Santonen 2016). Therefore, with regard to the endpoint

cancer the present working group has only focused on studies in rats, the most sensitive

species. In total, 11of the 19 rat studies identified by IARC showed an increased cancer

incidence in rats exposed to diesel exhaust. Of these 11 studies, 10 were on exhaust from

light-duty engines and 1 study was on exhaust from a heavy-duty engine. Based on the

IARC review (IARC 2014), 5 chronic dose-response studies in rats exposed to whole

diesel exhaust by inhalation were identified as well-designed and well-powered with

group size of more than 70 animals per sex (Mauderly et al. 1987;Brightwell et al.

1989;Mauderly et al. 1994;Heinrich et al. 1995;Stinn et al. 2005). These 5 studies are

described below and an overview of the studies is presented in Table 3.

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

Table 3. Diesel engine exhaust inhalation studies in rats with observed dose

carcinogenicity response

Reference

Mauderly et al., 1986

Strain (sex)

Group size

F344 (M/F)

N = 221-230

Exposure

Clean air and DEE (1980 5.7-L V8)

7 h/d, 5 d/w for up to 30 months

DEP

mg/m

0.35

3.5

7.0

Brightwell, 1989

F344 (M/F)

N = 72 or

144

Cond. air and DEE (VW Rabbit 1.5-L)

16 h/d, 5d/w for 24 + 6 months

0.7

2.2

6.6

Mauderly et al., 1994

F344 (M/F) Cond. air and DEE (Two ´88 LH6 GM 6.2L V8)

N = 100

16 h/d, 5d/w for 24 + 6 months

2.5

6.5

Heinrich et al., 1995

Wistar (F)

N = 100-220

Clean air and DEE (Two VW 40-kW 1.6-L)

18 h/d, 5d/w for 24 + 6 months

0.84

2.5

7.0

Stinn et al., 2005

Wistar (M/F)

N = 99

Clean air and DEE (VW 1.6-L)

6 h/d, 7 d/w for 24 + 6 months

0.3

1.2

3.8

4.7

4.0%

18.0%*

34.7%*

0.7

3.8

8.8

0.9-

2.8

0.01

0.3

0.7

1.5%

1,.4%

4.2%

22.5%*

3.0%

5.0%

9.0%

NO NOx

ppm

Lung tumor incidence

1.4%

0.7%

4.6%*

16.1%*

0.8%

0.0%

15.3%*

54.2%*

0.0%

8.0%

29.0%*

0.5%

0.0%

5.5%*

22.0%*

0.0%

28.0%*

56.9%*

The table is adapted from IARC Table 3.2 (IARC 2014). DEE, DEPs (measured particulate matter

in mg/m

). Cond.: Conditioned. Brightwell also included a filtered exhaust exposure with 99.7%

of the mass removed. No increased tumor incidence was observed.

Mauderly et al. 1986:

Groups of 221-230 male and female Fischer 344 rats were exposed by inhalation to 0.35,

3.5 and 7 mg/m

DEE for 7 h/day, 5 days/week for up to 30 months. Control rats were

exposed to filtered air. The diesel exhaust was generated by a 1980 model 5.7-L V8 diesel

engines (Volkswagen). The exhaust was diluted with clean air to reach the average diesel

particle concentrations of 0.35, 3.5 and 7 mg/m

. The diesel exhaust contained 0.1±0.1,

0.3±0.2 and 0.7±0.5 ppm NO

, respectively. Compared with controls, the incidence of rats

with lung tumors was significantly increased in rats at the high-dose exposure (16.1%).

These were combined data from both sexes. The article did not report the incidence

stratified by sex (Mauderly et al. 1986).

Brightwell et al. 1989:

Groups of 72 male and 72 female F344 rats were exposed by inhalation to 0.7, 2.2 and 6.6

mg/m

DEE or particle-filtered exhaust for 16 hrs/day, 5 days/week for 24 months

followed by 6 months in clean air. Control rats were exposed to conditioned air (n=144

per sex). The diesel exhaust was generated by a Volkswagen Rabbit 1.5-L diesel engine

and diluted with clean air to reach the mean diesel particle concentrations of 0.7, 2.2 and

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

6.6 mg/m

. The diesel exhaust contained 0.9±0.05, 2.7±0.08 and 8±1 ppm NO

respectively, and 0.7±0.5, 2.1±0.8 and 6±2 ppm NO, respectively. The medium dose

particle-filtered exhaust contained 2.8±0.5 ppm NO

and 2.2±0.7 ppm NO. The high dose

particle-filtered exhaust contained 8±2 ppm NO

and 7±2 ppm NO. Compared with

controls, the incidence of rats with lung tumors was significantly increased in both

female groups of both mid-dose exposure (15.3%) and high-dose exposure (54.2%); and

male group of high-dose exposure (22.5%). No increase in the number of rats with lung

tumors were observed in rats exposed to filtered diesel exhaust (Brightwell et al. 1989).

Mauderly et al.1994:

Groups of approximately 100 male and 100 female F344 rats were exposed by inhalation

to 2.5 and 6.5 mg/m

DEE for 16 h/day, 5 days/week for 24 months followed by 6 months

in clean air. Control rats were exposed to conditioned air (n=144). The diesel exhaust was

generated by two 1988 model LH6 General Motors 6.2-L V8 diesel engines and diluted

with clean air to reach the mean diesel particle concentrations of 2.5 and 6.5 mg/m

Compared with controls, the incidence of rats with lung tumors was significantly

increased in groups of female low-dose exposure (8%); and female and male high-dose

exposure (9% and 29%, respectively)(Mauderly et al. 1994).

Heinrich et al. 1995 :

Groups of 100-220 female Wistar rats were exposed by inhalation to 0.8, 2.5 and 7 mg/m

DEE for 18 hours/day, 5 days/week for 24 months followed by 6 months in clean air

(Heinrich et al. 1995). The diesel exhaust was generated by two 40-kW 1.6-L diesel

engines (Volkswagen). The exhaust was diluted with clean air (1:80, 1:27, 1:15 and 1:9) to

reach the average diesel particle concentrations of 0.8, 2.5 and 7 mg/m

. The diesel

exhaust contained 0.3-3.8 ppm NO

, 4.7-33.1 ppm NO

, 2.6-21.1 ppm CO and 0.2-0.7 %

. In addition, small amounts of SO

, CH

and C

(0.3-3.4 ppm) were measured in

the diesel exhaust. The number of rats with tumors after 30 months is given in Table 3.

Compared to rats exposed to clean air, exposure to diesel exhaust induced a significant

increase of cancer in the groups exposed to the two highest concentrations (2.5 and 7

mg/m

), while no rats developed cancer in the group exposed to the lowest concentration

(0.8 mg/m

Stinn et al. 2005:

Groups of 99 male and 99 female Wistar rats were exposed by nose-only inhalation to 3

and 10 mg/m

diesel engine exhaust for 6 h/day, 7 days/week for 24 months followed by

6 months in clean air (Stinn et al. 2005). The diesel exhaust was generated by a 40 kW 1.6-

L diesel engine (Volkswagen). The exhaust was diluted with air to reach the average

diesel particle concentrations of 3 and 10 mg/m

. The diesel exhaust contained 7 ppm NO

and 9 ppm NO

(low dose), and 23 ppm NO and 28 ppm NO

(high dose). Control

animals were exposed to clean air. Compared with controls, the incidence of rats with

lung tumors was significantly increased in both male and female groups of both low (9

out of 50 males (18%)); 14 out of 50 females (28%)) and high exposure (17 out of 49 males

(35%); 29 out of 51 males (57%)).

The study by Brightwell et al. showed no increase in lung tumor incidence when rats

exposed to filtered diesel exhaust (Brightwell et al. 1989). According to IARC, filtered

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

diesel exhaust without particles has been tested for carcinogenicity by inhalation

exposure in 7 studies in rats, 3 studies in mice, 3 studies in hamsters (IARC 2014). None

of the studies in rats and hamsters showed increased lung cancer incidences. One of the

studies in mice showed increased lung cancer incidence in mice exposed to filtered

exhaust. However, this study was considered inadequate by IARC because the incidence

of tumors in the control group was significant lower than historical controls in the same

laboratory. Furthermore, when the study was repeated performed in the same way, no

increase in the incidence of lung tumors was detected.

After the publication of the IARC monography, a chronic cancer study on “new

technology” DEE was published that met the above criteria for the selection of cancer

studies on exhaust from diesel engines (well designed and well powered with group size

of more than 70 animals per sex):

New technology diesel exhaust, ACES study:

Groups of 100 male and 100 female rats were exposed by inhalation to 3, 5 and 12 μg/m

DEE for 16 hours/day, 5 days/week for more than 24 months (McDonald et al. 2015). The

diesel exhaust was generated by US 2007 compliant heavy-duty diesel engine. The diesel

exhaust contained 0.1, 0.9 and 4.2 ppm NO

, respectively. Compared with controls, no

increase in the incidence of rats with lung tumors was observed.

Intratracheal instillation

In the review by IARC, 2 intratracheal instillation studies in rats were considered

adequate for an assessment of DEP- induced carcinogenicity. Both studies showed an

increased incidence of lung tumors (IARC 2014).

Summary

In summary, IARC found sufficient evidence for carcinogenicity of whole DEE, DEE

particulate matter and extracts of DEPs in experimental animals. IARC found inadequate

evidence for the carcinogenicity of the gas-phase of DEE (IARC 2014). The studies on

whole DEE were generated from fuels and engines produced before the year 2000 (IARC

2014).

When considering well-performed studies with multiple dose levels, 5 chronic 2-year

inhalation cancer studies on “traditional technology” whole diesel exhaust in rats were

identified based on the review performed by IARC (IARC 2014). In 4 of the studies,

increased lung cancer incidence occurred at exposure concentrations between 2.2 and 3.5

mg/m

DEP’s (Brightwell et al. 1989;Mauderly et al. 1987;Heinrich et al. 1995;Stinn et al.

2005). In the fifth study, increased cancer incidence was observed at 6.5 mg/m

DEPs

(Mauderly et al. 1994). No increase in the incidence of lung tumors was found in 7

studies of rats exposed to filtered exhaust (particles removed) (IARC 2014) or in rats

exposed to whole diesel exhaust

at ≤ 800 µg/m

for ≥ 2 years

(Taxell and Santonen

2016;Taxell and Santonen 2016). Intratracheal instillation of DEP showed an increased

incidence of lung tumors. This shows that the carcinogenic effect is caused by the

particulate phase of the diesel exhaust.

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

One chronic rat inhalation study of exhaust from “new-technology” diesel engines was

identified (ACES) (McDonald et al. 2015): No increase in cancer incidence was detected

in rats exposed in mass concentrations of DEP up to 12 μg/m

(4.2 ppm NO

) for 130

weeks.

The present working group notes that the particle concentration in this “new-

technology” diesel exhaust study is very low

(≤ 12 µ/m

) compared to the particle

concentrations in the cancer studies on “traditional diesel” exhaust

(effects at ≥ 2200

μg/m

). This may explain the lack of effect in the study. This is also in agreement with

the considerations by McDonald et al.: “It

is reasonable to speculate that the markedly lower,

and possibly nearly negligible, levels of exposure to diesel exhaust particles in the present study

may preclude effects attributable to the particles”

(McDonald et al. 2015).

NEG/DECOS (Taxell and Santonen 2016) concluded:

“A statistically significant increase in lung tumour incidence has been observed in several studies

in rats exposed to whole diesel

exhaust at concentrations of ≥ 2 200

µg DEP/m

for 104–130

weeks (Ishinishi et al. 1988;Brightwell et al. 1986;Mauderly et al. 1987;Stinn et al. 2005;Nikula

et al. 1995;Heinrich et al. 1995). No indication of carcinogenicity in other organs was detected in

the studies. The studies applied diesel engines from the mid-1990s or earlier. No effect on lung

tumour incidence was observed in rats exposed to filtered(particle-free) diesel exhaust or to whole

diesel exhaust at ≤ 800 μg DEP/m

for 104–152 weeks (Mohr et al. 1986;Brightwell et al.

1986;Heinrich et al. 1986;Mauderly et al. 1987;Heinrich et al. 1995). Correspondingly, no

indication of tumour development was detected in a 121–130-week inhalation study in rats

exposed to exhaust from a US 2007 compliant heavy-duty diesel engine at concentrations up to

4.2 ppm NO

(12 µg DEP/m

) (ACES programme) (McDonald et al. 2015). No clear evidence of

carcinogenicity of diesel exhaust in mice or hamsters has been observed even at high particle loads

(Brightwell et al. 1986;Heinrich et al. 1986;Heinrich et al. 1995).”

Conclusion

In conclusion, the present working group notes that chronic inhalation studies in rats

have shown elevated incidence of lung tumors in rats exposed to diesel exhaust

compared to control rats exposed to air. The importance of the particulate phase of DEE

for the carcinogenic effect is supported by studies showing 1) no increase in the

incidence of lung tumors in rats when the particulate phase is removed before exposure,

and 2) increase in incidence of lung tumors in rats intratracheally instilled with DEPs.

Genotoxicity

IARC recently summarized DEE-induced genotoxicity (IARC 2014):

“Diesel

engine exhausts and the mechanisms by which they induce cancer in

humans are complex in nature, and no single mechanism appears to predominate.

Organic solvent and physiological fluid extracts of diesel engine exhaust particles

and several of the individual components of these exhausts are genotoxic, and some

are carcinogenic, generally through a mechanism that involves DNA mutation.

These modifications include the formation of bulky DNA adducts and oxidized DNA

bases. Both the organic and particulate components of diesel engine exhaust

emissions can generate ROS, leading to oxidative stress. ROS can be generated from

washed particles, fresh particles, arene quinones formed by photochemical or

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

enzymatic processes, metals and the phagocytosis process, and as a result of the

inflammatory process. ROS can lead directly to the formation of oxidatively

modified DNA and DNA adducts from by-products of lipid peroxidation

(Voulgaridou et al. 2011), can cause lipid peroxidation, which generates cytotoxic

aldehydes (Barrera et al. 2008), and can also initiate a signalling cascade that leads to

inflammation, resulting in further induction of oxidative stress, which in turn leads to

cell prolif- eration and cancer (Milara and Cortijo 2012). In response to the

inflammatory insult, cyclooxygenase-2 is upregulated and is a potent mediator of

cell proliferation (Speed and Blair 2011).”

NEG/DECOS recently summarised DEE-induced genotoxicity (Taxell and Santonen

2016):

“DEP and DEP extracts have shown genotoxic responses in vitro. Bacterial mutagenicity studies

with the gaseous phase of diesel exhaust have also shown positive responses, although the data are

much more limited.

Inhalation studies with diesel exhaust in rodents have shown increases in the levels of DNA

strand breaks, DNA adduct levels, oxidative DNA damage and in gpt and lacI mutations in the

lungs of transgenic mice, whereas bone marrow and peripheral blood cell micronucleus, SCE and

chromosomal aberration tests have been mostly negative. Oral, intraperitoneal and intratracheal

administration of diesel exhaust particulates or DEP extracts have produced genotoxic responses

in several organs.

No in vitro genotoxicity studies on new technology diesel engines were located. However, recent

inhalation studies with diesel exhaust from a heavy-duty diesel engine fulfilling the US 2007

emission standards did not show local or systemic genotoxicity or oxidative DNA damage in

rodents. This suggests that new diesel engine and after-treatment technologies may decrease the

genotoxic potency of diesel exhaust when expressed per unit of engine work (per kWh). This

decrease can be mostly attributed to the significant reduction of particulate matter in the

exhaust.”(Taxell

and Santonen 2016).

There are a number of studies evaluating the genotoxic effects of DEP. Some of the

studies considered most relevant are described below.

Short-term effects of SRM 1650 DEP originating from heavy-duty diesel engines on

markers of genotoxicity were evaluated in mice. Mice exposed to DEP by inhalation to

either a single 90 min exposure (20 or 80 mg/m

) or as 4 repeated 90 min exposures (5 or

20 mg/m

) resulted in DNA strand breaks in BAL cells, oxidative DNA damage and

DNA adducts in lungs of mice (Dybdahl et al. 2004).

Genotoxicity was evaluated in mice and rats exposed to exhaust from “new technology”

diesel engines at three exposure-doses: Low: 3 μg DEP/m

/0.1 ppm NO

, Medium: 5 μg

DEP/m

/0.9 ppm NO

, High: 12 μg DEP/m

/4.2 ppm NO

). Mice and rats were exposed

16 h/day, 5 days/week for up to 3 months and 24 months, respectively. Compared to

controls, no increase in genotoxicity (DNA damage (comet assay) in the lungs, 8-OHdG

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

in serum and micronuclei in peripheral blood) was detected (Hallberg et al. 2015;Bemis

et al. 2015;Hallberg et al. 2012;Bemis et al. 2012).

The present working group concludes that diesel exhaust is genotoxic. This is based on

the fact that inhalation studies with DEE and intratracheal instillation studies of DEPs in

rodents have shown increased levels of genotoxicity. Therefore, the present working

group agrees with the evaluation by NEG/DECOS that the genotoxic potency of diesel

exhaust most likely is attributed to the particulate phase of the exhaust. A single study of

genotoxicity of exhaust from “new technology” diesel engines was identified. In that

study, no genotoxicity was observed when tested in rodents. The highest particle

concentration in that study was very low (12 μg DEP/m

) compared to the particle

concentrations in studies of exhaust from “traditional engines” resulting in effects. This

may explain the lack of effect in the study.

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

ECHANISMS OF TOXICITY

Pulmonary inflammation, genotoxicity and cancer

IARC concluded recently:

“there is strong mechanistic evidence that diesel engine exhaust, as well as many of

its components, can induce lung cancer in humans through genotoxic mechanisms

that include DNA damage, gene and chromosomal mutation, changes in relevant

gene expression, production of reactive oxygen species and inflammatory

responses. In addition, the co-carcinogenic, cell-proliferative, and/or tumour-

promoting effects of other known and suspected human carcinogens present in

diesel-engine exhaust probably contribute to its carcinogenicity in the human

lung.” (IARC 2014).

NEG/DECOS:

“DEP have been shown to induce genotoxicity (DNA strand breaks, DNA adducts, oxidative

DNA damage and mutations) in vivo and in vitro. In addition to the genotoxicity caused by

mutagens bound to DEP (e.g. PAHs and PAH derivatives) or present in the gas phase, DEP-

related chronic inflammation, oxidative stress and induction of ROS may contribute to the cancer

risk observed in humans. Although it can be hypothesized that the dose-response curve of diesel

exhaust related cancer may include a non-linear component, it is not possible to identify a

threshold level for the carcinogenicity of diesel exhaust”

(Taxell and Santonen 2016).

The mechanisms by which DEPs cause cancer are likely mediated by both primary

(particle driven) and secondary (cell driven) genotoxicity.

Chronic inhalation studies of DEE alongside filtered diesel exhaust showed that the

particulate fraction of diesel exhaust was required for diesel exhaust-mediated tumor

formation, as rats exposed to diesel exhaust developed tumors in contrast to rats

exposed to filtered diesel exhaust (Brightwell et al. 1989).

Both inhalation of diesel exhaust and pulmonary instillation of DEP and DEP extracts

increased the mutant frequency in lung tissue in

gpt

delta transgenic mice (Hashimoto et

al. 2007). Mice were exposed to 3 mg/m

for 12 weeks or to single instillations of 0.2 mg

DEP extract or 0.5 mg DEP. The mutant frequency potency of DEP extract and DEP was

and 2.7 x 10

-5

and 5.6 x 10

-5

, respectively. DEP extracts constituted 50% of DEP mass, and

the authors concluded that the mutagenic potential of DEP could be explained by the

extractable mutagens including PAH and nitrated PAH (Hashimoto et al. 2007).

However, chronic inhalation exposure to TiO

NPs, carbon black NPs and DEE in rats

performed in the same study showed that the cancer incidence from the three types of

exposure was very similar (table 4), and thus, the authors concluded that the study

supported the hypothesis that the carbon core of the diesel soot is the main causative

agent for DEE-related carcinogenicity (Heinrich et al. 1995). The CB and TiO

exposure

concentrations were changed during the experiment to obtain similar lung particle load

as in the diesel exposed animals

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

Similar conclusions were made based on a study comparing tumor formation in rats

following inhalation of DEE and CB nanoparticles (Nikula et al. 1995).

The present working group does not regard the change in exposure concentrations in the

Heinrich study as problematic.

Table 4. Lung cancer incidence in rats exposed to diesel exhaust, CB and TiO

after 30

months (24 months of exposure followed by 6 months in clean air) (Heinrich et al.

1995)

Average particle exposure (mg/m

)

Clean air,

control

Number of

rats with

tumors

a,b)

DEP

0.8

2.5

7.0

11.6

TiO

11/200** 22/100***

39/100

32/100

(4/200**) (9/100***)

(28/100)

(19/100)

Count without benign keratinising cystic squamous-cell tumors in parentheses.

*,**,***: Statistically significant compared

to clean air control at p≤0.05, 0.01 and

0.001, respectively.

1/217

0/198

Inhalation and pulmonary instillation of the standard DEP NIST1650 induced increased

levels of DNA strand breaks in lung tissue and bronchoalveolar lavage cells and bulky

DNA adducts in lung tissue (Kyjovska et al. 2015;Dybdahl et al. 2004) suggesting that

DNA damage is caused both by formation of bulky DNA adducts and by oxidative DNA

damage.

The SRM DEP NIST 1650 and carbon black Printex 90 also had similar mutagenic

potential in vitro in the murine lung fibroblast cell line FE-1 (Jacobsen et al.

2008;Jacobsen et al. 2007). The CB nanoparticles induced much more reactive oxygen

species compared to the DEP. On the other hand, the mutagenic potential of the PAH

content could not account for the observed mutagenicity. The authors concluded that the

observed mutagenicity was likely caused by both the organic fraction of DEP and by

particle-mediated ROS production (Jacobsen et al. 2008).

Thus, evidence suggests that both the extractable organic fraction and the particulate

fraction of DEP contribute to the carcinogenicity of DEP. The molecular mechanism

includes formation of bulky DNA adducts as well as oxidative DNA damage likely

induced by surface-dependent ROS generation. In addition, inhalation of DEE and DEP-

induced dose-dependent pulmonary inflammation which could cause secondary

genotoxicity. Thus, the available data indicated induction of cancer through both direct

and indirect genotoxic mechanisms. Based on the observed mechanisms of genotoxicity,

the present working group concludes that DEP-induced mutagenicity and

carcinogenicity occurs by non-threshold mechanisms.

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

OSE

RESPONSE RELATIONSHIPS

Inflammation

Human data

NEG/DECOS has recently reviewed studies on the inflammatory response in the lungs of

human volunteers exposed to whole diesel exhaust for 1-2 hours. Exposure levels at ~100

μg DEP/m

resulted in increased numbers of neutrophils and inflammatory cytokines in

BAL in healthy volunteers (Taxell and Santonen 2016).

Animal data

Dose-response relationships have been observed for inflammation following inhalation

of DEE (Mauderly et al. 1994) and DEP (Dybdahl et al. 2004) and instillation of DEP

(Kyjovska et al. 2015).

Cancer

Human data

Vermeulen et al. performed a meta-regression of lung cancer mortality and cumulative

exposure to EC (as a proxy measure of DEPs), based on RR estimates reported by two

studies of workers in the trucking industry and one study of miners (Vermeulen et al.

2014b). Based on a linear meta-regression model, a lnRR of 0.000982 (95% CI: 0.00055,

0.0014) for lung cancer mortality with each 1-μg/m

-year increase in cumulative EC was

estimated (Figure xx in the paragraph on epidemiological studies). The numbers of

excess lung cancer deaths for 45 years of occupational exposures of 1, 10, and 25 μg/m

EC were estimated to 17, 200, and 689 per 10,000, respectively, by 80 years of age.

Animal data

Strong dose-response relationships have been observed for lung cancer in rats following

inhalation of diesel exhaust (Figure 2). The present working group notes that female rats

are more sensitive than male rats. In figure 2, the cancer incidences in rats from 4 of the 5

identified chronic inhalation studies are shown (Brightwell et al. 1989;Heinrich et al.

1995;Mauderly et al. 1986;Mauderly et al. 1994;Stinn et al. 2005). The data from the fifth

study by Mauderly et al. is omitted on the figure because that study did not report the

incidence stratified by sex (Mauderly et al. 1986).

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

F e m a le s

M a le s

R a ts w ith tu m o u rs (% )

M a s s c o n c e n t r a tio n ( m g /m )

Figure 2.

Frequency of female and male rats with tumors as a function of DEP mass

concentrations in the chronic inhalation studies by (Brightwell et al. 1989;Heinrich et al.

1995;Mauderly et al. 1994;Stinn et al. 2005). Dotted lines represent 95% confidence interval for

the regression lines. Females: y = 5.6x – 0.088; Males: y = 2.8x +1.4

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

ARTICLE CHARACTERISTICS

Diesel exhaust consists of a particulate phase (carbon particles with adsorbed organic

matter) and a gas/vapor phase which includes volatile organic compounds (VOCs),

nitrogen oxides and carbon monoxide. The focus of the present report is on the

particulate phase of diesel exhaust.

“Traditional” diesel engines did not control the emission of particulate matter. In

contrast, “new-technology” engines are equipped with particulate filters that reduce the

emission of particulate matter considerable (more than 90% by mass)(IARC 2014;Taxell

and Santonen 2016). The composition of DEPs emitted from “traditional” and “new

technology” diesel engines is also different. Typically, the proportion of elemental

carbon from a traditional heavy-duty diesel engine is 75% while this proportion is

reduced to 13% when using new technology diesel engines. The proportions of sulfates

and organic carbon are increased from 1% to 53% and 19% to 30%, respectively, when

exhaust after-treatment systems are applied (Taxell and Santonen 2016).

Since the composition of DEPs is complex, different metrics to measure the exposure to

DEPs has been applied (Taxell and Santonen 2016): Measurement of EC is considered the

most specific and sensitive marker of DEP because in most occupational settings only

DEE contributes significantly to EC. In contrast, measurement of the total mass of

particles in the occupational setting does not allow a separation of DEP from other

particles in the environment. Another metric for DEP exposure is to quantify PAH or

other organic compounds adhered to the DEPs chemically.

These years, “traditional” diesel engines are gradually being replaced by “new-

technology” diesel engines. This will most likely result in lower concentrations of DEPs

in occupational settings. However, it is expected to take a long time before this transition

has been completed (Taxell and Santonen 2016). Furthermore, it should be emphasised

that the nanosized fraction of DEPs contributes very little to the DEP mass.

No studies were identified comparing the toxicity on a mass base of DEPs from

“traditional technology” and “new technology”.

The present working group notes that EC is considered the best marker of DEP. Further,

the present working group notes that there is a lack of knowledge regarding the toxicity

of DEPs from “new technology” compared to DEPs from “traditional technology”

engines.

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

REVIOUS EVALUATIONS OF DIESEL EXHAUST

The most recent evaluations of DEE are presented below.

IARC (2012)

In 2012, the IARC classified DEE as carcinogenic to humans (group 1). This classification

was based on sufficient evidence that DEE is causally related to lung cancer. In addition,

IARC found some evidence for a positive association between exposure to DEE and an

increased risk of bladder cancer. Furthermore, IARC found sufficient evidence for

carcinogenicity of whole DEE, DEE particulate matter and extracts of DEPs in

experimental animals. IARC found inadequate evidence for the carcinogenicity of gas-

phase DEE (i.e. particle free DEE) (IARC 2014). IARC does not differentiate between

exhaust from traditional and new technology diesel engines in their classification (IARC

2014). In Denmark, substances classified as group 1, 2A and 2B by IARC are considered

carcinogenic.

NEG/DECOS (2014)

In 2014, NEG and DECOS co-produced a criteria document on diesel exhaust (Taxell and

Santonen 2016).

NEG/DECOS summarised as follows:

“Diesel engine exhaust is a complex mixture of gaseous and particulate compounds

produced during the combustion of diesel fuels. The gas phase includes carbon dioxide,

nitrogen oxides (NO

), carbon monoxide and small amounts of sulphur dioxide and

various organic compounds. Diesel exhaust particles (DEP) contain elemental carbon

(EC), organic compounds, sulphates, nitrates and trace amounts of metals and other

elements. New technology diesel engines are characterised by a significant reduction of

the DEP mass emissions. Occupational exposure to diesel exhaust occurs in mining,

construction work, professional driving, agriculture and other activities where diesel-

powered vehicles and tools are applied.

The critical health effects of diesel exhaust are considered to be pulmonary inflammation

and lung cancer. For older technology diesel engines, pulmonary inflammatory responses

were observed in human volunteers after single exposure at 100 μg DEP/m

(~ 75 μg

EC/m

), and in rats after long-term

exposure at 210 μg DEP/m

(~ 160 μg EC/m

Development of lung tumours was seen in rats at 2 200 μg DEP/m

(~ 1 650 μg EC/m

For new technology diesel engines, pulmonary inflammatory changes were reported in

rats after 13 and 130 weeks of exposure at 3.6 and 4.2 ppm NO

(12–13

μg DEP/m

, ~ 3

μg EC/m

). The effect was absent at 0.9–1.0 ppm NO

(4–5

μg DEP/m3, ~ 1 μg EC/m

No indication of tumour development was detected.

Epidemiological studies associate occupational exposure to exhaust from older

technology diesel engines with increased lung cancer risk. Based on a log-linear meta-

regression model, 45 years of occupational exposure to

diesel exhaust at 1, 10 and 25 μg

EC/m

was estimated to result in 17, 200 and 689 extra lung cancer deaths per 10 000

individuals, respectively, by the age of 80 years. Although data allowing a direct

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

comparison of the carcinogenic potential of exhaust from new and older technology

diesel engines are not available, the significant reduction of the DEP mass concentration

in the new technology diesel engine exhaust is expected to reduce the lung cancer risk

(per kWh).

In addition to the critical effects, human and animal inhalation studies associate exposure

to older technology diesel engine exhaust with sensory irritation, increased airway

resistance, cardiovascular effects, genotoxicity and adjuvant allergenic effects. There are

also animal studies indicating neuroinflammatory effects, developmental effects and

effects on the male reproductive function.

When evaluating the health risk of diesel exhausts it is important to take into account that

the transition from “old” to “new” technology diesel engines is expected to take a long

time.”

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

CIENTIFIC BASIS FOR SETTING AN OCCUPATIONAL

EXPOSURE LIMIT

Different methods exist for calculating OELs. The choice of method depends on the

mode of action of the substance, and can fundamentally be split up in two approaches:

Threshold effects or non-threshold effects. The threshold effect approach relies on the

assumption that the organism can withstand a certain dose before adverse effects occur,

whereas for non-threshold effects it is assumed that any exposure to the substance can

result in adverse effects. The present working group considers cancer as the most severe

critical effect. Furthermore, the present working group considers that DEP-induced

mutagenesis and cancer occurs by non-treshold mechanisms. Therefore, in this report,

we will calculate proposed OELs based on non-threshold effects for lung cancer. The

calculations will be performed based on data from both human and animal studies.

Health-based exposure limit based on epidemiological data

DEE was recently classified as carcinogenic to humans by IARC (class 1). In a recent

meta-analysis of three studies of occupational exposure to diesel engine exhaust

(Vermeulen et al. 2014b), the association between lung cancer incidence and DEE

measured as EC was modelled.

lnRR for lung cancer= intercept + slope x exposure

Exposure was measured as EC in μg/m

-years. The intercept was set at 0, and the slope

was determined to be 0.000982 with a standard error of 0.000219:

lnRR for lung cancer = 0.000982 x exposure

In Denmark, the life time risk of developing lung cancer (0-74 years) is 4.9% for men and

4.5% for women. The relative risk caused by occupational exposure to a carcinogen,

which causes cancer at the different risk levels (1%, 0.1% and 0.01%) are given in table 5.

Table 5. Relative risk of lung cancer for carcinogens that cause 1%, 0.1% or 0.01%

excess lung cancer risk in a population with the current Danish lung cancer incidence

Men

Women

Life time risk (0-74 years)

4.9%

4.5%

2011-2015 in Denmark

Excess lung cancer risk

level

1:100

RR= (4.9+1)/ 4.9= 1.20

RR= (4.5+1)/4.5= 1.22

1:1 000

RR= (49+1)/49= 1.02

RR= (45+1)/45=1.02

1:10 000

RR= (490+1)/490= 1.002

RR= (450+1)/450= 1.002

1:100 000

RR= (4900+1)/4900= 1.000 2

RR= (4500+1)/4500= 1.000 2

1) http://www-dep.iarc.fr/NORDCAN/DK/StatsFact.asp?cancer=180&country=208

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

Assuming 1:1 000 excess lung cancer incidences among men, the calculation would be:

EC concentration in μg/m

-years = Ln (1.02)/ 0.000982= 20.16 μg/m

-years

For a 45-year work life this would correspond to 20.16 μg/m

-years/45 years = 0.45 μg/m

Assuming 1:100 000 excess lung cancer incidences among men, the calculation is:

EC concentration in μg/m

-years = Ln (1.0002)/ 0.000982= 0.2016 μg/m

-years

For a 45-year work life this would correspond to 0.0045 μg/m

Table 6. Estimated lung cancer risk based epidemiological data from (Vermeulen et al.

2014b)

Excess lung cancer risk

DEP air concentration

1: 1 000

0.45 μg/m

1: 10 000

0.045 μg/m

1: 100 000

0.0045 μg/m

Health-based exposure limit based on two chronic

inhalation studies in rats

The present working group notes that the five identified high quality chronic inhalation

studies in rats show that female rats are more sensitive to DEP-induced cancer than male

rats (Figure 2 in the Dose-response chapter of the present report). Furthermore, the

studies have somewhat different cancer incidences. The current working group has

therefore calculated health-based exposure limits based on two different studies,

representing low or high carcinogenic response, respectively, in female rats.

Low response (Heinrich study)

The chronic inhalation study by Heinrich (Heinrich et al. 1995) was identified as

representative of a relatively low response and was selected because the carcinogenicity

of CB and TiO

nanoparticles was assessed alongside DEP, thus allowing comparison of

the carcinogenic potency between the three types of particles.

Table 7. Observed cancer incidence following DEP exposure in (Heinrich et al. 1995)

DEP concentration

2.5 mg/m

7.0 mg/m

Cancer Incidence

1/217

11/200

22/100

Lung burden

23.7

63.9

(mg/lung)

Method I

A non-threshold effect is assumed. The genotoxic effects induced by DEP are probably

caused by a number of mechanisms including surface-associated PAH, carbon-core-

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

induced ROS and inflammation-induced ROS. DNA damage in terms of increased levels

of DNA strand breaks in the comet assay was observed following intratracheal

instillation of the standard DEP NIST1650b at dose levels that did not induce statistically

significant increases in neutrophil influx 1 and 28 days post-exposure (Kyjovska et al.

2015). This suggests that surface-associated PAH and carbon-core-induced ROS are

likely mechanisms of action. These are considered non-threshold effects.

The current working group has chosen to use the approach used by Kasai et al.(Kasai et

al. 2016) and Erdely et al. (Erdely et al. 2013), who use the measured lung burden in rats

exposed by inhalation and the alveolar surface area of rats and humans to estimate the

human equivalent lung burden:

Observed cancer incidence at 2.5 mg/m

(11/200 – 1/217)/(1-1/217) = 0.05 = 5%

Lung deposited dose in rats at 2.5 mg/m

: 23.7 mg/lung.

The human equivalent dose is:

(Rat deposited dose) x (human alveolar surface area)/(rat alveolar surface area) =

23.7 mg x 102 m

/0.4 m

= 6 043.5 mg DEP per human lung.

We assume the following standardised constants:

The standard value of human ventilation is 20 L/min during light work (1.2 m

/h).

An average work day is 8 h per day.

An average work week is 5 days.

In Denmark, an average employee work 45 weeks per year.

An average working life is 45 years.

Assuming 16.8% deposition as previously reported for humans by (NEG/DECOS)(Taxell

and Santonen 2016).

Using the values above, a lung burden of 6 043.5 mg in humans would require that

workers are exposed to:

Air concentration =

6 043.5 mg/(8h/day x 5 days/week x 45 weeks/year x 45 years x 1.2 m

/h x 0.168) =

0.37 mg/m

Thus, at an air concentration of 0.37 mg/m

during a 45-year work life, an excess lung

cancer incidence of 5% is expected. Assuming a linear dose-response relationship, then

1% excess lung cancer is expected at:

(0.37 mg/m

)/5 = 0.074 mg/m

Accordingly, assuming linear dose-response relationship, the excess lung cancer risk is

estimated:

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

Table 8. Excess cancer risk

Excess lung cancer risk

1: 1 000

1: 10 000

1: 100 000

Method II

DEP air concentration

7.4 μg/m

0.74 μg/m

0.074 μg/m

Risk estimates were calculated as recommended by ECHA (ECHA 2012a;

SCHER/SCCP/SCENIHR 2009), based on the 2 year DEE inhalation study in rats by

(Heinrich et al. 1995) (Table 4):

Excess cancer risk:

Observed excess cancer incidence at 2.5 mg/m

(5/200- 1/217)/(1-1/217)= 0.0506 = 5 %

Correction of dose metric for humans during occupational exposure (8h/d):

2.5 mg/m

x (18 h/day)/(8 h/day) x (6.7 m

/10 m

) = 3.769 mg/m

Calculation of unit risk for cancer:

Risk level = exposure level x unit risk

0.0506 = 3 769 μg/m

x unit risk

Unit risk = 1.34 x 10

-5

per μg/m

At a dose of 1 μg/m

, 1.34 x 10

-5

excess cancers are expected.

Calculation of dose levels corresponding to risk level of 10

-5

(and other risk levels)

-5

risk level = exposure level x unit risk (1.34 x 10

-5

per μg/m

)

Exposure level (10

-5

) = 0.74 μg/m

Thus, at 0.74 μg/m

, 1:100 000 excess lung cancer cases can be expected.

Table 9. Calculated excess lung cancer incidence at DEP mass concentrations based on

method II

Excess lung cancer incidence DEP Air concentration (μg/m

)

1: 1 000

1: 10 000

7.4

1: 100 000

0.74

High response (Brightwell et al.)

The study by Brightwell et al. was selected as an example of a high response study.

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

Here, rats were exposed 16 h/day, 5 days/week for 2 years to diesel engine exhaust at

particle concentrations of 0.7, 2.2 or 6.6 mg/m

particles (Brightwell et al. 1989). In

addition, filtered diesel engine exhaust was included at the middle and high doses. Lung

burden was not assessed. Females were identified as the most sensitive sex and the

middle dose (2.2 mg/m

) was selected as the lowest dose inducing lung tumors.

Table 10. Cancer incidence (Brightwell study)

Exposure

0 mg/m

0.7 mg/m

2.2 mg/m

Cancer

incidence in

female rats

Cancer

incidence in

male rats

1/126

0/71

11/72

6.6 mg/m

39/72

Filtered DE

(6.6 mg/m

)

0/72

2/134

1/72

3/72

16/71

0/71

For the study by Brightwell et al, excess lung cancer incidence at DEP mass

concentrations were calculated based on method II for (Brightwell et al. 1989).

Calculations based on Method I were not performed due to lack of information on

deposited dose in the study by Brightwell.

Excess cancer risk:

Observed excess cancer incidence at 2.2 mg/m

(11/72- 1/126)/(1-[1/126])= 0.146 =15 %

Correction of dose metric for humans during occupational exposure (8h/d):

2.2 mg/m

x (16 h/day)/(8 h/day) x (6.7 m

/10 m

) = 2.948 mg/m

Calculation of unit risk for cancer:

Risk level = exposure level x unit risk

0.146 = 2 948 μg/m

x unit risk

Unit risk = 5.0 x 10

-5

per μg/m

At a dose of 1 μg/m

, 4.4 x 10

-5

excess cancers are expected.

Calculation of dose levels corresponding to risk level of 10

-5

(and other risk levels)

-5

risk level = exposure level x unit risk (5.0 x 10

-5

per μg/m

)

Exposure level (10

-5

) = 0.20 μg/m

Thus, at 0.20 μg/m

, 1:100 000 excess lung cancer cases can be expected.

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

Table 11. Calculated excess lung cancer incidence at DEP mass concentrations based

on method II for (Brightwell et al. 1989)

Excess lung cancer incidence DEP Air concentration (μg/m

)

1: 1 000

1: 10 000

2.0

1: 100 000

0.20

Summary

In summary, excess lung cancer risks were estimated based on the meta-analysis of

epidemiological studies and for two different 2-year inhalation studies in rats:

Table 12. Summary of risk estimates for DEP-induced lung cancer

Excess lung

Human study Method I, μg/m

Method II, μg/m

cancer incidence μg/m

Heinrich

Brightwell

1: 1 000

0.45

7.4

1: 10 000

0.045

0.74

7.4

2.0

1: 100 000

0.0045

0.074

0.74

0.20

In the human study, DEE exposure was measured as elemental carbon (EC), whereas the

animal studies used total PM. For comparison, total PM was converted to EC assuming

that DEP contains 75% EC as found for traditional DEP (Taxell and Santonen 2016).

Table 13. Overview of exposure levels in terms of EC, resulting in extra cancer risk

levels at 1:1000, 1:10 000 and 1: 100 000 based on a non-threshold based mechanism of

action using different approaches

Suggestion of an OEL for DEP calculated as EC

Excess lung Human

Method I, μg/m

Method II, μg/m

cancer

studies

Rat inhalation study of

incidence

DEE*

Vermeulen

0.45 μg/m

0.045 μg/m

0.0045

μg/m

Heinrich

5.6 μg/m

0.56 μg/m

0.056 μg/m

Heinrich

5.6

0.56

Brightwell

1.5

0.15

1: 1 000

1: 10 000

1: 100 000

Method I is based on lung deposition. Method II is based on air concentrations and following

ECHA guidelines.*For traditional DEPs, it is assumed that 75% of the mass is EC (Taxell and

Santonen 2016).

The suggested health-based OELs for DEP can be compared to the unit risk values for

TiO

NPs and carbon black NPs based on chronic inhalation studies performed in

parallel (Heinrich et al. 1995).

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

Table 14. Unit risk values for TiO

NPs, CB and DEE in chronic inhalation studies

performed in parallel

Control

10 mg/m

2.5 mg/m

TiO

Carbon black

DEE

Total cancer

1/217

32/100

8/107

11/200

incidence

Lung burden (mg/g)

23.7

Unit risk pr ug/m

2.1 x 10

-5

1.97 x 10

-5

1.34 x 10

-5

Overall, the OEL suggestions derived from the meta-analysis of epidemiological studies

were 20 -100 fold lower than the OEL suggestions that were based on chronic inhalation

studies in rats. The present working group notes that one of the explanations may be that

the clearance rate of nanosized particles has been estimated to be ca. 60-100 days in rats

and mice, but substantially lower, several hundred days, in humans (Taxell and

Santonen 2016).

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

ONCLUSION

The present working group evaluated the relevant literature on diesel exhaust from both

epidemiological and animal inhalation studies.

A recent evaluation by NEG/DECOS on DEE (Taxell and Santonen 2016) concluded that:

‘The critical health effects of diesel engine exhaust are pulmonary inflammation and lung cancer.’

IARC evaluated the epidemiological studies on exposure to DEEs and risk of cancer

(IARC 2014) and concluded:

’There is sufficient evidence in humans for the carcinogenicity of

diesel engine exhaust. Diesel engine exhaust causes cancer of the lung. A positive association has

been observed between exposure to diesel engine exhaust and cancer of the urinary bladder’.

The present working group considers lung cancer as the most severe critical effect.

A meta-analysis of epidemiological studies was performed including only studies with

information regarding dose-response relationship between exposure to DEE and risk of

lung cancer identified at the time of the IARC evaluation (Vermeulen et al. 2014b).

Dose-dependent tumor formation was observed in lungs of rats in chronic inhalation

studies in rats. Chronic exposure to filtered DEE did not induce lung tumor formation in

rats. The present working group regards carcinogenicity as the critical adverse effect of

DEE exposure. Furthermore, the present working group regards the carcinogenic effect

as caused by the particulate fraction of DEE.

The present working group found that the mechanism of action of the carcinogenic effect

of DEP has not been fully clarified. Primary genotoxicity caused by PAH and surface-

dependent generation of reactive oxygen species has been demonstrated. Secondary

genotoxicity due to particle-induced inflammation is an important and well-documented

mechanism of action for the development of lung cancer. However, the available data

indicated induction of cancer through both direct and indirect genotoxic mechanisms.

Therefore, the present working group considers carcinogenicity as a non-threshold

effect. Consequently, the present working group decided to perform the risk assessment

based on a non-threshold mechanism of action.

The present working group identified a recent meta-analysis as suitable for risk

assessment (Vermeulen et al. 2014b). However, 5 highly quality chronic inhalation

studies in rats were identified, and the present working group decided also to select two

of these for calculation of excess cancer risk: A 2-year chronic cancer inhalation study in

rats with relatively low tumor incidence (0, 2.5 and 7 mg/m

) (Heinrich et al. 1995), and

another 2-year chronic inhalation study in rats with a relatively high tumor incidence

(0.7, 2.2 and 6.6 mg/m

) (Brightwell et al. 1989) . In Table 15, excess lung cancer risk at 1

in 1 000, 1 in 10 000, and 1 in 100 000 using different approaches is presented.

BEU, Alm.del - 2019-20 - Bilag 101: Orientering om NFA’s forslag til grænseværdier for fem kemiske stoffer, fra beskæftigelsesministeren

Table 15. Overview of exposure levels in terms of elemental carbon (EC), resulting in

extra cancer risk levels at 1:1000, 1:10 000 and 1: 100 000 based on a non-threshold

based mechanism of action using different approaches

Suggestion of an OEL for DEP calculated as elemental carbon

Excess lung EC levels

cancer

Meta-

Rat inhalation

incidence

analysis of

study of DEE*

Human

Method I, Lung

Method II

Method II,

ECHA

studies

burden (Heinrich)

ECHA

(Vermeulen)

(Heinrich)

(Brightwell)

1:1 000

0.45 μg/m

5.6 μg/m

56 μg/m

15 μg/m

1: 10 000

0.045 μg/m

0.56 μg/m

5.6 μg/m

1.5 μg/m

1: 100 000

0.0045

0.056 μg/m

0.56 μg/m

0.15 μg/m

μg/m

*) For traditional DEPs, it is assumed that 75% of the mass is EC (Taxell and Santonen 2016)

The DEE exposure in the epidemiological studies was traditional DEE. Both of the

chronic inhalation studies were performed on traditional DEE. The present working

group notes that there is limited available data on the biological effects of DEP from

“new technology” diesel engines. Typically, the proportion of EC from a traditional

heavy-duty diesel engine is 75% of the total particle emission while this proportion is

reduced to 13% when using “new technology” diesel engines. Correspondingly, the

proportions of sulfates are increased from 1% to 53% when exhaust after-treatment

systems are applied (Taxell and Santonen 2016).

The current working group notes that in chronic inhalation studies in rats, CB

nanoparticles and DEPs have very similar carcinogenic potential (Heinrich et al. 1995).

The present working group furthermore notes that there is no available evidence

suggesting that “new technology” DEPs are less carcinogenic than “traditional” DEP and

carbon black.