Traffic Air Pollution and Cancer Annie J Sasco

Traffic Air Pollution and Cancer Annie J. Sasco, MD, Dr PH 1, 2 Ann Olsson, MPH 1 Unit of Epidemiology for Cancer Prevention International Agency for Research on Cancer - World Health Organization 2 Director of research, Institut National de la Santé et la Recherche Médicale 1 Mireille Chiron, MD & Patrice Reungoat Institut National de Recherche sur les Transports et leur Sécurité Lyon, France The opinions expressed in this talk only represent those of the speaker (AJS) and should not be considered as official views of either IARC, WHO, INSERM or INRETS

Why me on this topic? • Cancer epidemiologist – Chief of the Unit of Epidemiology for Cancer Prevention at IARC – Deeply concerned about environmental issues – Member of ISEE, SER, SPER, IEA, ADELF • Having been previously involved at the European (and international) level as expert on issues such as – – tobacco use of hormones as growth promoters in animal husbandry hormonal treatments (OC, HRT, tamoxifene) screening for cancer

But. . . • Not currently working on traffic air pollution and cancer. • Concentrating on two main domains: – tobacco, cannabis and cancer – breast cancer and pesticides

Measurement from the national ATMO index, built from 3 pollutant levels: sulphur dioxide, nitrogen dioxide, ozone, without taking into account other important pollutants such as particulate matters Source: Agence de l’Environnement et de la Maîtrise de l’Energie, 1998

Expected death rate potentially avoidable by a reduction of 50% of the levels of indicators for atmospheric pollution in the 9 agglomerations studied* Annual rate for 100 000 inhabitants * The results for Bordeaux do not take into account pollution by photo oxidant as the indicators were not available

How did I prepare for this talk? • Usual way: Medline – not much • Call the experts (in the French setting) – Dr. Mireille Chiron, INRETS – Prof. Denis Zmirou, AFSSE

• Consult important reports: – EPA: Health Assessment Document for Diesel Exhaust (2002) – HEI: Health Effects of Acute Exposure to Air Pollution (2002) – HEI: Research on Diesel Exhaust (1999) – SFSP: La pollution atmosphérique d’origine automobile et la santé publique (1996) – IARC: IARC Monographs on the evaluation of carcinogenic risks to humans. Volume 46. Diesel and Gasoline Engine Exhausts and Some Nitroarenes (1989) – WHO: Transport, Environment and Health (2000) – WHO-IARC: World Cancer Report (2003) • Rely on the IARC Monographs Programme List of IARC Evaluations http: //www-cie/monoeval/grlist. html

How did I really get into it? • Enlisted the active participation of one of my trainees, Ann Olsson • Went back to some historical work I did in 1979 • Drew a parallel between smoking // air pollution and cancer from the 1930´s to post 2000

Back to history • Early studies on air pollution and mortality/morbidity – Firket (1931): Sur les causes des accidents survenus dans la vallée de la Meuse, lors des brouillards de Décembre 1930 – Schenk et al. (1949): Air pollution, Donora, Pennsylvania. Epidemiology of the unusual smog episode of October 1948 – Ministry of Health, UK (1954): Mortality and morbidity during the London fog of December 1952

At the same time • First major studies on tobacco and lung cancer – Wynder and Grahams (1950). Tobacco smoking as a possible etiologic factor in bronchiogenic carcinoma – Doll and Hill (1950). Smoking and carcinoma of the lung – Schwartz et al. (1961). Results of a French survey on the role of tobacco, particularly inhalation, in different cancer sites • Setting-up of prospective studies – British doctors in the UK – American Cancer Society volunteers in the USA

Since then • Hundreds of studies on tobacco • Few studies on air pollution Why such difference? • Difficult to study validly the long term effects of air pollution

Knowledge about specific compounds based on the IARC Monographs on the Evaluation of Carcinogenic Risks to Humans

Main pollutants resulting from engine exhausts • • • Carbon dioxide (CO 2) Carbon monoxide (CO) Nitrous oxides (NOX), in particular NO and NO 2 Particles Organic volatile compounds: hydrocarbons (alcanes, alcenes, aromatic monocyclic, in particular benzene and toluene), oxygenated compounds (aldehydes, acids, ketones, ethers…) • Aromatic polycyclic hydrocarbons(benzo[a]pyrene, benzo[k]fluoranthene, benzo[b]fluoranthene, benzo[g, h, i]perylene, benz[a]anthracene • Sulphur dioxide (SO 2) • Metals, lead in particular

Diesel and Gasoline Engine Exhausts • Vol. 46 (1989) • Diesel engine exhaust Group 2 A (probably carcinogenic) • Engine exhaust, gasoline Group 2 B (possibly carcinogenic) • Contain thousands of gaseous and particulate substances ( 1 is individually classified in Group 1, 6 in Group 2 A and 16 in Group 2 B)

Benzene • Vol. 29 (1982) & Suppl. 7 (1987) • Group 1 (Carcinogenic to humans) • Increased incidence of various types of leukemia among workers exposed to benzene

Group 2 A (probably carcinogenic to humans) • 1, 3 - Butadiene (106 -99 -0) Vol. 71; 1999 • Benz[a]anthracene (56 -55 -3) Suppl. 7 ; 1987 • Benzo[a]pyrene (50 -32 -8) Suppl. 7 ; 1987 • Dibenz[a, h]anthracene (53 -70 -3) Suppl. 7 ; 1987 • Ethylene dibromide (106 -93 -4) Vol. 71 ; 1999 • Formaldehyde (50 -00 -0) Vol. 62 ; 1995

Group 2 B (possibly carcinogenic to humans) • Acetaldehyde (75 -07 -0) Vol. 71 ; 1999 • Dibenz[a, h]acridine (226 -36 -8) Suppl. 7 ; 1987 • Dibenz[a, j]acridine (224 -42 -0) Suppl. 7 ; 1987 • 1, 2 -Dichloroethane (107 -06 -02) Vol. 71 ; 1999 • Lead (7439 -92 -1) and lead compounds, inorganic Suppl. 7 ; 1987 • 1, 6 -Dinitropyrene (42397 -64 -8) Vol. 46 ; 1989 • 1, 8 -Dinitropyrene (42397 -65 -9) Vol. 46 ; 1989

Group 2 B (possibly carcinogenic to humans) • 2 -Nitrofluorene (607 -57 -8) Vol. 46 ; 1989 • 1 -Nitropyrene (5522 -43 -0) Vol. 46 ; 1989 • Polycyclic aromatic compounds – Benzo[b]fluoranthene (205 -99 -2) Suppl. 7 ; 1987 – Benzo[j]fluoranthene (205 -82 -3) Suppl. 7 ; 1987 – Benzo[k]fluoranthene (207 -08 -9) Suppl. 7 ; 1987 – Dibenzo[a, e]pyrene (192 -65 -4) Suppl. 7 ; 1987 – Dibenzo[a, h]pyrene (189 -64 -0) Suppl. 7 ; 1987 – Indeno[1, 2, 3 -cd]pyrene (193 -39 -5] Suppl. 7 ; 1987 – 5 -Methylchrysene (3697 -24 -3) Suppl. 7 ; 1987

Methyl tert-Butyl Ether (MTBE) • • Vol. 73 (1999) Volatile synthetic chemical CAS no 1634 -04 -4 Fuel additive in motor gasoline Produced in very large quantities since 1979 to replace lead as an octane enhancer • Group 3 (not classifiable as to its carcinogenicity to humans)

Agents or Exposures proposed for Evaluation or Re-evaluation in future IARC Monographs • • Priority Diesel engine exhaust (2 A) High Gasoline engine exhaust (2 B) High Ozone High Air pollution (some air pollutants) High

International Trends. . . • Concentrations of sulphur dioxide (SO 2)and suspended particulate matter are decreasing in developed countries, while those of Nitrogen oxides (NOX)and Ozone (O 3) are either constant or increasing. • Effective legislation • Improved technology • Increasing traffic

. . . International Trends • In developing countries, concentrations of SO 2, NOX and O 3 and suspended particulate matter are raising. • Increasing traffic and industrial emissions • Weak legislation • Poor technology

Pollutants of current interest • Ground-Level Ozone the prime ingredient of smog – cause acute respiratory problems – impair the body’s immune system • Particulate Matter (PM) is the term used for a mixture of solid particles and liquid droplets in the air – The size varies, from a few nm to tens of µm (PM 10, PM 2. 5) – Health concern because they easily reach the deepest recesses of the lungs and other tissues • Nitrogen dioxide serves (in most circumstances) as a surrogate for all traffic-related combustion products

In brief. . . • Pollution of air, water and soil is estimated to account for 1 -4% of all cancers (WHO-IARC, 2003) • However, uncertainties are many, notably because cancer takes decades to develop. . .

Epidemiological approaches

1. Cross sectional studies Comparison of morbidity/mortality between “exposed” regions and “non-exposed” regions Difficulties: – Adequate reference population? – Population movements? – Confounding factors? Interpretation: – “Negative” results: Not possible to exclude increased risk – “Positive “ results: Chance?

2. Case-control studies Comparison of exposures for cases and controls Difficulties: – Choice of pathology? – Does pertinent exposure data exist?

3. Cohort studies An exposed population followed over time Difficulties: – Low risk – Rare diseases – Logistic difficulties? Follow a specific population over a long period – Comparisons?

Difficult to estimate health and ecological consequences: – Lack of information on type and level of current exposures – Lack of information on past exposures – Complex models of estimating and predicting “uncertain risks”

Lack of power in epidemiological surveillance • Small study sizes • Population “at risk” not well identified • Pertinent exposures – Not well identified – Multiple – Unknown levels

Lack of power in epidemiological surveillance • Risk level : Low • Pathology outcome – Vague – Non specific – Rare …difficult to interpret the results … correct parallel: passive smoking // air pollution

Is air pollution dangerous to health? • The answer is: YES – Contains well documented toxic compounds • How dangerous is it? It depends… – Characteristics of exposure • Type • Amount • Distance to population – Individual characteristics

Individual characteristics GENETIC SUCEPTIBILITY • Heredity LIFE STYLE • Smoking • Nutrition • Physical Activity OCCUPATION • Work title • Specific exposures ENVIRONMENTAL • The concentration of specific components vary greatly with locality and time Cumulative long-term effects of exposure to multiple compounds at varying levels remain to be evaluated

Occupational exposure to diesel exhaust and lung cancer risk • 14 cohort mortality studies from 1981 onwards UK, Canada, USA, Sweden, Denmark Mostly occupational but also general cohorts • 13 case-control studies from 1984 onwards USA, Sweden, France, UK, Denmark, Germany. Mostly population based • Most (but not all) studies are slightly positive, with in several evidence for a dose response relationship

General population exposure to air pollution Potential target cancer sites • Lung cancer in adults • Childhood cancers • Other: breast, melanoma

Lung Cancer • A small proportion of lung cancer is attributable to outdoor air pollution by industrial effluent, engine exhaust products and other toxins • Several studies have provided evidence for an increased risk of lung cancer among residents in areas with higher levels of air pollution • Has been studied more extensively than other cancer types because of an a priori biologic hypothesis • Other cancer types have a partly unknown etiology and therefore more research is necessary to refute or strengthen causal relationships with urban air pollution

Overall mortality and lung cancer in the USA Reference: Dockary DW et al. 1993 Type of study: Prospective cohort study, 14 -16 year follow-up, 8111 adults from six U. S. cities Exposure: Fine Particles (FP) <2. 5 µm, inhalable particles, SO 2, O 3, suspended sulphates Comparison: most/least polluted city Overall mortality: 1. 26 (1. 08 -1. 47) Lung cancer: 1. 37 (0. 81 -2. 31) Results: Comment: All cause mortality is increased in various models adjusting for smoking, education, BMI, occupation. Mortality most strongly associated with FP, including sulphates

Air pollution and lung cancer in Trieste, Italy Reference: Biggeri A et al. 1996 Type of study: Case-control study of deceased men 755 cases, 755 controls from local autopsy registry Exposure: Distance from sources and air particulates Results: The risk of lung cancer was highly related to city center (p=0. 0243), with an excess relative risk at zero distance of 2. 2 and a smooth decrease moving away from the source (-0. 015) 1. 4 (1. 1 -1. 8) for air particulates >0. 298 g/m 2/day Model adjust for subject-specific confounders Comment:

Lung cancer incidence in the USA Reference: Beeson LW et al. 1998 Type of study: Prospective cohort study, followed 1977 -1992 6338 non-smoking, non-Hispanic white adults California, U. S. A. Exposure: Monthly air pollution data O 3 (Interquartile range increase in 100 ppb) PM 10 (Interquartile range increase <10 µm) SO 2 (Interquartile range increase <10 µm) Results: Men Women Comment: O 3 RR 3. 56 (1. 35 -9. 42) PM 10 RR 5. 21 (1. 94 -13. 99) SO 2 RR 2. 66 (1. 62 -4. 39) PM 10 RR 1. 21 (0. 55 -2. 66) (>50 µm /m 3) SO 2 RR 2. 14 (1. 36 -3. 37) Sex differences partially due to differences in exposures

Lung cancer in Sweden Reference: Nyberg F et al. 2000 Type of study: Population based case-control study Men 40 -75 years, stable residents of Stockholm county 1042 cases, 2364 population controls Exposure: Retrospective models of estimating NOX/NO 2 and SO 2, Results: 1. 2 (0. 8 -1. 6) for top NO 2 decile 1. 4 (1. 0 -2. 0) (for 20 years previously) Comment: Controlled confounding for smoking, radon, socioeconomic grouping, work in risky occupations and occupational exposure to diesel exhaust , other combustion products and asbestos

Overall, cardiopulmonary and lung cancer mortality in the USA Reference: Pope CA et al. 2002 Type of study: Prospective cohort study 500 000 adults among 1. 2 million American Cancer Society volunteers (CPS II) National data sources related to address Particules (PM 10; PM 2. 5), SO 2, NO 2, . . . Exposure: Results: Adjusted mortality relative risk associated with a 10 µm /m 3 change in PM 2. 5 Overall mortality: 1. 06 (1. 02 -1. 11) Lung cancer: 1. 14 (1. 04 -1. 23) Comment: Controlled confounding for smoking, education. marital status, BMI and alcohol consumption

Overall mortality and lung cancer in the Netherlands Reference: Hoek G et al. 2002 Type of study: Prospective cohort study followed from 1986 to 1994 5000 adults 55 -69 years Exposure: Estimated from home address (black smoke and NO 2) Results: Overall mortality: 1. 41 (0. 94 -2. 12) for living near a major road Lung cancer: 1. 06 (0. 43 -2. 63) for black smoke 1. 25 (0. 42 -3. 72) for NO 2 Results obtained after adjustment for potential confounders Comment:

Leukemia • Occupational exposure to benzene is associated with acute myeloid leukemia in adults • Studies suggest an association between proximal high traffic streets and leukemia among children. • Note: results are not unanimous

Childhood cancer in the U. S. A. Reference: Type of study: Savitz DA & Feingold L 1989 Exposure: Traffic density of street of residence Results: Reference: < 500 vehicles /day Population based case-control study Children 0 -14 years, U. S. A. 328 cases, 262 controls All cancers: Leukaemias: Brain cancer: Soft tissue: Comment: 1. 7 (1. 0 -2. 8) 2. 1 (1. 1 -4. 0) 1. 7 (0. 8 -3. 9) 1. 4 (0. 5 -4. 4) with dose response : All cancers Leukaemias 500 -4999 v/d > 5000 v/d > 10000 v/d 1. 6 (0. 7 -3. 5) 1. 8 (0. 9 -3. 3) 3. 1 (1. 2 -8. 0) 1. 2 (0. 4 -3. 9) 2. 7 (1. 3 -5. 9) 4. 7 (1. 6 -13. 5) Adjustments for age, sex, year of diagnosis, type of residence and geographic zone do not change results

Childhood cancer in Sweden Reference: Feychting M et al. 1998 Type of study: Case-control study nested in a population of children having lived for at least a year within 300 m of 220 and 440 k lines in Sweden during 1960 -1985 142 cases, 568 controls Estimated NO 2 concentration based on home address Exposure: Results: <39µm/m 3 40 -49 >50 Comment: All cancers 1 1. 3 (0. 4 -4. 3) 2. 7 (0. 9 -8. 5) Leukaemia 1 1. 7 (0. 2 -14. 6) 2. 7 (0. 3 -20. 6) CNS 1 1. 0 (0. 1 -12. 7) 5. 1 (0. 4 -61. 2) Adjustments are made for EMF and confounders

Childhood leukaemia in the UK Reference: Harrisson RM et al. 1999 Type of study: Case control and incidence ratio study West Midlands, UK Children 0 -15 years old with leukemia (130 cases) or solid tumors (251 controls) diagnosed between 19901994 Exposure: Distance of home from main road and petrol station Results: Comment: case-control analysis < 100 m main road 1. 61 (0. 90 -2. 87) < 100 m petrol station 1. 99 (0. 73 -5. 43) both 5. 91 (0. 61 -57. 3) IR analysis 1. 16 (0. 74 -1. 72) 1. 48 (0. 65 -2. 93) 0. 81 (0. 16 -2. 38) Adjustments are made for age and sex

Leukaemia and childhood cancer in the U. S. A. Reference: Pearson RL et al. 2000 Type of study: Case-control study Children 0 -14 years living in Denver, U. S. A. 320 cases, 259 controls Exposure: Weighted traffic density at home address Results: All cancers 5. 90 (1. 69 -20. 56) Leukemia 8. 28 (2. 09 -32. 80) ( for > 20000 vehicles/day)

Childhood cancer in Denmark Reference: Raaschou-Nielsen O et al. 2001 Type of study: Case-control study Children from the Danish Cancer Registry diagnosed with cancer (leukemia, tumor of the central nervous system, or malignant lymphoma) before 15 years of age between 1968 -1991 1989 cases, 5506 population controls Exposure: Advanced model of estimated NO 2 and benzene exposure from traffic intensity in utero and during childhood

Results: All cancers Leukaemias CNS Lymphomas Pregnancy 1. 0 (0. 9 -1. 1) 0. 9 (0. 7 -1. 2) 0. 7 (0. 5 -1. 1) 0. 9 (0. 8 -1. 0) 0. 8 (0. 6 -1. 2) 0. 8 (0. 5 -1. 3) 1. 0 (0. 9 -1. 2) 0. 7 (0. 5 -1. 2) 0. 6 (0. 3 -1. 1) 1. 0 (0. 8 -1. 4) 1. 7 (1. 0 -2. 8) 1. 2 (0. 5 -3. 0) 5000 -9999 >10000 0. 9 (0. 8 -1. 0) 0. 8 (0. 6 -1. 1) 1. 0 (0. 7 -1. 6) 0. 9 (0. 8 -1. 1) 0. 8 (0. 5 -1. 2) 1. 1 (0. 6 -2. 2) 0. 9 (0. 7 -1. 0) 0. 6 (0. 4 -1. 1) 0. 9 (0. 4 -1. 8) 0. 9 (0. 7 -1. 2) 1. 5 (0. 8 -3. 0) 1. 3 (0. 4 -4. 8) Benzene 0. 5 -1. 2* >1. 3* Hodgkin’s disease NO 2 1. 7 (0. 8 -3. 8)1. 5 -2. 9* 1. 5 (0. 7 -3. 2) 4. 3 (1. 5 -12. 4) >3. 0* 6. 7 (1. 7 -26. 0) 500 -4999 veh/d 5000 -9999 >10000 Childhood 500 -4999 veh/d *in 1000 ppb - days

Breast cancer in the U. S. A. Reference: Lewis-Michl EL et al. 1996 Type of study: Case-control study Women 20 -79 years old, Nassau and Suffolk counties, Long Island, U. S. A. 1420 cases, 1420 controls (derived from NY driving license registry) Exposure: Results: Residential proximity to industrial facilities and traffic High traffic intensity and post menopausal breast cancer Nassau Suffolk Adjusted OR 1. 29 (0. 77 -2. 15) 0. 89 (0. 40 -1. 99) Comment: Adjustments are made for age, occupation, education

Melanoma • Chloroflourocarbons cause destruction of the ozone layer and enhance the risk of skin cancer through increased ultraviolet radiation

For increased power in epidemiological surveillance. . . • Epidemiology – – – Identification of compounds and “true” exposure levels Identification of “exposed” people Precise choice of pathology Test dose-response relationships Biological markers of exposure and effects • …for surveillance • …for epidemiology • Prevention – Exposure surveillance – Intervention

Conclusion Overall, there is an association between lung cancer in adults and some childhood tumours and air pollution, including but not limited to the one coming from traffic. Even if the relative risk is of limited magnitude, the extent of the population exposed is large and therefore prevention is warranted.