Epidemiology of Environmental and Occupational Cancer

Epidemiology of environmental and occupational cancer

Paolo Boffetta*,1,2

1International Agency for Research on Cancer, 150 cours Albert-Thomas, 69008 Lyon, France; 2German Cancer Research Centre, 69120 Heidelberg, Germany

Environmental carcinogens, in a strict sense, include this broad sense, the environment is implicated in the outdoor and indoor air pollutants, as well as soil and causation of the majority of human cancers (Tomatis drinking water contaminants. An increased risk of et al., 1990). In a more specific sense, however, mesothelioma has consistently been detected among environmental factors include only the (natural or individuals experiencing residential exposure to asbestos, man-made) agents encountered by humans in their while results for lung cancer are less consistent. Several daily life, upon which they have no or limited personal good-quality studies have investigated lung cancer risk control. The most important ‘environmental’ exposures, from outdoor air pollution based on measurement of defined in this strict sense, include outdoor and specific agents. Their results tend to show an increased indoor air pollution and soil and drinking water risk in the categories at highest exposure, with relative contamination. Most environmental carcinogens occur risks in the range 1.5. A causal association has been at higher concentrations in the workplace: since the established between exposure to environmental tobacco report of scrotal cancer among chimney sweeps by Pott smoke and lung cancer, with a relative risk in the order of (1775), occupational cancer research has played a 1.2. Radon is another carcinogen present in indoor air, pivotal role in the elucidation of environmental causes with a relative risk in the order of 1.06 for exposure at of human cancer. Furthermore, control of occupational 100 Bq/m3. In several Asian populations, an increased risk exposure circumstances has proven to be feasible and of lung cancer results among women from indoor pollution effective as compared to other cancer preventive from cooking and heating. There is strong evidence of an measures. increased risk of bladder, skin and lung cancers following In the following sections the evidence linking ex- consumption of water with high arsenic contamination; posure to selected environmental factors and risk of results for other drinking water contaminants, including cancer will be reviewed, followed by a summary of chlorination by-products, are inconclusive. A total of 29 current knowledge on occupational carcinogens. Agents occupational agents are established human carcinogens, whose exposure depends on lifestyle, such as solar and another 30 agents are suspected carcinogens. In radiation and food additives, will not be considered, nor addition, at least 12 exposure circumstances entail agents occurring in the environment as a consequence of exposure to carcinogens. Exposure is still widespread for accidents or warfare. many important occupational carcinogens, such as asbestos, coal tar, arsenic and silica, in particular in developing countries. Although estimates of the global burden of occupational and environmental cancer result in figures in Environmental cancer the order of 2% and less than 1%, respectively, these Cancer risk from environmental exposure to asbestos cancers concentrate in subgroups of the population; furthermore, exposure is involuntary and can, to a large Asbestos and asbestiform fibres are naturally occurring extent, be avoided. fibrous silicates with an important commercial use, Oncogene (2004) 23, 6392–6403. doi:10.1038/sj.onc.1207715 mainly in acoustical and thermal insulation. They can be divided into two groups: chrysotile and the group of Keywords: air pollution; arsenic; environment; environ- amphiboles, including amosite, crocidolite, anthophyl- mental tobacco smoke; neoplasms; occupation; radon lite, actinolite and tremolite fibres. Chrysotile is the most widely used type of asbestos. Although all types are carcinogenic to the lung and mesothelioma, the biological effects of amphiboles on the pleura and peritoneum seem to be stronger than those of chrysotile (INSERM, Introduction 1997). The use of asbestos has been restricted or banned in many countries. The concept of environment is often used with a broad In contrast to the many epidemiological studies scope in the medical literature, including all nongenetic available on asbestos-exposed workers, there are few factors such as diet, lifestyle and infectious agents. In studies on the health effects of nonoccupational (household and residential) exposure to asbestos. One type of *Correspondence: P Boffetta; E-mail: [email protected] household exposure concerns cohabitants of asbestos Environmental and occupational cancer P Boffetta 6393 a workers and arises from dust brought home on clothes. 1 Other household sources of asbestos exposure are o represented by the installation, degradation, removal a

and repair of asbestos-containing products. Residential a a a b exposure mainly results from outdoor pollution related a a (1979) (1984) a to asbestos mining or manufacturing, in addition to (1995) (2001) (2000) (1998) (1999) (1998) (1997) (1986)

natural exposure from the erosion of asbestos or (2000) et al. et al. (1998) et al. et al. et al. (1989) asbestiform rocks. The assessment of nonoccupational unspeciﬁed and mixed; et al. et al. et al. ¼ et al. Compared to residence et al. c

exposure to asbestos presents difficulties, since levels are et al., et al. generally low, and the duration and frequency of et al. exposure and the type of fibre are seldom known with precision. Table 1 summarizes the results of studies on risk of Women only. b pleural mesothelioma and lung cancer from environ- chrysotile; UM mental (residential) exposure to asbestos. Studies were ¼ available from various countries and, in most cases, exposure was defined as residence near a mine or another major source of asbestos exposure. A potential limitation of these studies, in particular those without amphiboles; C assessments of exposure at the individual level (‘ecolo- ¼ gical’ studies), is possible concomitant occupational or household exposure to asbestos. The risk of mesothelioma was greatly increased in all but one study among individuals with environmental exposure to asbestos. Results for lung cancer, however, are less consistent, with an increased risk detected in studies from South Africa and China, but not in studies from Europe and North America. Imperfect control of confounding by smoking and other lung carcinogens may explain the predominant type of fibre: A

lack of consistency. 14 6.7 2.0–22.2 Hansen Ca RR 95% CI Ca RR 95% CI ¼ Results derived from raw data reported in the publication. Cancer risk from outdoor air pollution a Air pollution is a complex mixture of different gaseous and particulate components, whose composition varies greatly by locality and time. In recent decades, emissions and air concentrations of traditional industrial air c pollutants, such as SO2 and smoke particles, have ecological study; TF conﬁdence interval. ¼

decreased, whereas there is an increasing or continued ¼ problem with air pollution from vehicles, with emissions 0.2 km from asbestos plant 47 1.9 0.5–6.4 Xu

of engine combustion products including volatile o organic compounds, nitrogen oxides and ﬁne particu- lates, as well as with secondarily increased ozone levels. Studies of risk of mesothelioma and lung cancer from environmental exposure to asbestos 20 years 5years in mining area 1 km from asbestos cement plant2 km from potential0.5 source miles from asbestos0.5km factory from potential source 36 6.6 11 17 4.1–11 5 11.53.5–38.2 5.4 6.6 1.8–17 0.9–50 Magnani Magnani Newhouse and Thompson (1965) Howel relative risk; CI

There is biological rationale for a carcinogenic potential cohort study; Ec o o o o 4 4 ¼ of numerous components of the air pollution mix, ¼ including benzo[a]pyrene, benzene, some metals, particles (especially fine particles) and possibly ozone. Table 1 Many definitions of outdoor air pollution exposure have been used in epidemiological studies. Earlier analytical studies generally compared residence in urban areas, where the air is considered more polluted, to residence in rural areas, sometimes providing limited data on the typical levels of some pollutants in the areas case–control study; Co ¼ studied. Other studies have attempted to address number of exposed cases; RR exposure to specific components of outdoor air, provid- ¼ ing risk estimates in relation to quantitative or semiquantitative air pollution exposure assessments (Hitosugi, 1968; Vena, 1982; Brownson et al., 1987;

Jedrychowski et al., 1990; Katsouyanni et al., 1991; residence; Ca study design: CC ¼

Mills et al., 1991; Jo¨ ckel et al., 1992; Dockery et al., ¼ New Caledonia CC A Use of contaminated building materials 14 40.9 5.1–325 56 0.9 0.6–1.3 Luce Country SD TF Source of exposure Mesothelioma Lung cancer Reference South Africa E A Res. in mining area 61 8.7 6.7–11.4 86 1.7 1.2–2.5Botha AustriaUKUKChina Ec A Res. in CC polluted town CC CC A A C Res. Res. Res. 36 0.8 0.4–1.6 Neuberger South AfricaCanadaCanadaUSAItalyItaly CCItaly, Spain, Switzerland A Co CC EcChina UM Res. C in mining areas CC CAustralia Res. Res. A Co in mining area Res. CC in mining area UM Res. near UM to asbestos plant Res. Res. in polluted Co city CC A A Res. in polluted Res. area 7 7 1.3 7.6 32 0.5–3.0 3.4–14.9 14.6 71 16 4.9–43.1 NA 1.1 3.6 182 41 0.9–1.4 1.4–9.3 0.9 NA Camus Mzileni 0.6–1.3 NA Hammond Theriault and Grand-Bois 5.7 (1978) Magnani NA Luo SD year 1993; Barbone et al., 1995; Pope et al., 1995; Zaridze Res.

Oncogene Environmental and occupational cancer P Boffetta 6394 et al., 1995; Biggeri et al., 1996; Pawlega et al., 1997; lung cancer risk (Blot and Fraumeni, 1975; Newman Abbey et al., 1999) or, in some cases, to more qualitative et al., 1976; Pershagen et al., 1977; Matanoski et al., exposure assessments (Pike et al., 1979; Xu et al., 1989). 1981; Cordier et al., 1983; Xiao and Xu, 1985), which Another type of study has addressed residence in the was confirmed by some (Brown et al., 1984; Pershagen, proximity of specific sources of pollution, such as major 1985; Frost et al., 1987) but not all (Lyon et al., 1977; industrial emission sources or heavy road traffic. Greaves et al., 1981; Rom et al., 1982; Marsh et al., Several cohort and case–control studies of outdoor air 1997, 1998) subsequent studies with individual informa- pollution have been reported (Stocks and Campbell, tion on exposure: arsenic exposure level in some of these 1955; Hammond and Horn, 1958a/b; Haenszel et al., studies, however, was likely to be low. 1962; Haenszel and Taeuber, 1964; Dean, 1966; Buell Limited results are available for cancers other than and Dunn, 1967; Buell et al., 1967; Hitosugi, 1968; the lung. In ecological studies, many individual cancers Hammond, 1972; Cederlo¨ f et al., 1975; Dean et al., have shown elevated urban/rural ratios, including 1977, 1978; Pike et al., 1979; Hammond and Garfinkel, cancers of the mouth and throat, nasopharynx, oeso- 1980; Doll and Peto, 1981; Vena, 1982; Tenkanen et al., phagus, stomach, colon, rectum, larynx, female breast, 1985; Brownson et al., 1987; Samet et al., 1987; bladder and prostate (Levin et al., 1960; Goldsmith, Tenkanen and Teppo, 1987; Xu et al., 1989; Jedry- 1980). Residual confounding by smoking or occupa- chowski et al., 1990; Holowaty et al., 1991; Katsouyanni tional exposures may be involved. Other ecological et al., 1991; Mills et al., 1991; Jo¨ ckel et al., 1992; studies have related cancer rates to air pollutant Dockery et al., 1993; Tenkanen, 1993; Barbone et al., measurements, or emission indexes or figures for fuel 1995; Pope et al., 1995; Zaridze et al., 1995; Biggeri et al., consumption. For example, an ecological study in 19 1996; Engholm et al., 1996; Pawlega et al., 1997; Beeson European countries found an inverse temporal associa- et al., 1998; Abbey et al., 1999; McDonnell et al., 2000; tion between gasoline use and leukaemia mortality or Nyberg et al., 2000; Hoek et al., 2002; Pope et al., 2002). morbidity, but a weak positive spatial association Overall, the studies suggest relative risks (RRs) of up to (Swaen and Slangen, 1995). about 1.5for urban vs rural residence or high vs low Overall, the evidence for an increased lung cancer risk estimated air pollution exposure. There is no clear from outdoor air pollution is quite strong, although indication if early or late exposure is more important, methodological problems in the available studies cau- and data on possible interaction with smoking or tion against a final conclusion, while the evidence occupational exposures are inadequate. Among these concerning cancers other than lung cancer, including studies, four cohort (Mills et al., 1991; Dockery et al., childhood cancer, is not consistent, and is insufficient to 1993; Pope et al., 1995, 2002; Abbey et al., 1999; draw any conclusion on the presence or absence of a McDonnell et al., 2000; Hoek et al., 2002) and 10 case– causal association. control (Hitosugi, 1968; Vena, 1982; Brownson et al., 1987; Jedrychowski et al., 1990; Katsouyanni et al., 1991; Jo¨ ckel et al., 1992; Barbone et al., 1995; Zaridze Cancer risk from exposure to environmental tobacco et al et al et al ., 1995; Pawlega ., 1997; Nyberg ., 2000) smoke studies were based on measurements of specific air components. Selected results from these studies, with the Environmental tobacco smoke is composed of side- corresponding air pollution differentials, are presented stream and mainstream smoke, in which known, in Table 2: they tend to show an increased risk of lung probable or possible human carcinogens are present. cancer in the categories at highest exposure, which does The International Agency for Research on Cancer not seem to be attributable to confounding factors. The (IARC) has evaluated the evidence of a carcinogenic studies of lung cancer and air pollution, however, suffer risk from exposure to environmental tobacco smoke, exposure misclassification, including dilution of the and has classified it as an established human carcinogen sufficiently exposed population by a considerable (IARC, 2004). Confounding by dietary, occupational number of minimally exposed persons. and social class-related factors can be reasonably To pinpoint possible industrial emissions responsible excluded, and bias from misclassification of smokers is for the suggested urban excess, populations living near not likely to explain the results. On that occasion, a point sources of air pollution have also been studied. meta-analysis of epidemiological studies of lung cancer Increased risks have been reported for living close to and adult exposure to environmental tobacco smoke industries such as smelters, foundries, chemical indus- was conducted, resulting in RRs of 1.22 (95% CI 1.12– tries, and others with various emissions (Lloyd, 1978; 1.32) in women and 1.36 (95% CI 1.02–1.82) in men Shear et al., 1980; Gailey and Lloyd, 1983; Lloyd et al., from spousal exposure, and of 1.15 (95% CI 1.05–1.26) 1985a/b, 1986; Gailey and Lloyd, 1986; Smith et al., in women and 1.28 (95% CI 0.88–1.84) in men from 1987; Williams and Lloyd, 1988; Xu et al., 1989; workplace exposure. Other meta-analyses have reached Barbone et al., 1995; Zaridze et al., 1995; Ko et al., very similar conclusions (Hackshaw et al., 1997; 1997; Bhopal et al., 1998; Petrauskaite et al., 2002). Boffetta, 2002). Results of these studies were inconsistent, and con- The evidence of a causal association between fidence intervals of risk estimates were mostly wide. environmental tobacco smoke exposure and cancers A number of studies concern sources of inorganic in organs other than the lung is inconclusive (IARC, arsenic in air. Ecological studies suggested an increased 2004).

Oncogene Table 2 Relative risk of lung cancer and outdoor air pollution measurements in some studies with quantitative or semiquantitative exposure assessment Location, study period, ref.Sex RR 95% CI Exposure contrast a Basis for exposure assessment for Comments individuals and/or areas

Cohort studies 3 3 USA, 1975–1991, Dockery M+F 1.37 0.81–2.31 Per 18.9 mg/m PM2.5 City of residence in 1975. Pollutant Study range 11–29.6 mg/m PM2.5 et al. (1993) average 1979–1985. across six studied cities 3 USA, 1982–1998, Pope et al. M+F 1.08 1.01–1.16 Per 10 mg/m PM2.5 (1979–1983) City of residence in 1982. Pollutant Study range roughly 10–30 or 3 3 (2002) 1.13 1.04–1.22 Per 10 mg/m PM2.5 (1999–2000) averages 1979–1983 and 1999–2000 5–20 mg/m PM2.5 (1979–1983 and 1999–2000) 3 USA, California, 1977–1992, M 3.56 1.4–9.4 Per 556h/year above 200 mg/m O3 Residential history 1973–1992 and Range 988 h over study subjects 3 Beeson et al. (1998) M 5.21 1.9–14.0 Per 24 mg/m PM10 local monthly pollutant levels 3 M 2.66 1.6–4.4 Per 11 mg/m SO2 (3.72 ppb) 1973–1992 used to calculate 3 F 2.14 1.4–3.4 Per 11 mg/m SO2 (3.72 ppb) individual subject averages over 1973–1992. Netherlands, 1986–1994, M+F 1.06 0.4–2.6 Per 10 mg/m3 black smoke Residential history of last 1–4 Estimated range 9.6–35.8 mg/m3 3 3 Hoek et al. (2002) 1.25 0.4–3.7 Per 30 mg/m NO2 residences up to 1986 matched by black smoke and 14.7–67.2 mg/m GIS to regional and urban NO2 over study subjects background estimates plus local exposure based on distance to major road

Case–control studies Boffetta P cancer occupational and Environmental Germany, ﬁve cities, 1984– M 1.16 0.6–2.1 High vs low time-weighted Lifetime residential history. BaP, 1988, Jo¨ ckel et al. (1992) semiquantitative index TSP, SO2 data + local energy, SO2 M 1.82 Same, using 20 years lag and coal use and degree of industrialization 1900–1980

Italy, Trieste, 1979–1981, M 1.4 1.1–1.8 4294 mg/m2/day vs o175of Last residence. Deposition 1985–1986, Barbone et al. particulate deposition measured 1972–1977 (1995) Russia, Moscow, 1991–1993, F 2.6 1.2–5.6 4200 SO2, 460 NO2, 42400 CO, Final ¼ 20 years of residential Never-smokers only Zaridze et al. (1995) 4300 BS vs o50, o48, o1300, o60 history. Pollutant average 1971–1975 Poland, Krakow, 1992–1994, M 0.24 0.1–0.5TSP 142 & SO 2 110 vs 78 & 71 Last residence. Pollutant average Residual confounding by Pawlega et al. (1997) 1973–1980 occupation possible despite some adjustment Sweden, Stockholm, 1985– M 1.6 1.1–2.4 NO2 X29.26 vs NO2o12.78 (Top 30-year residential history. NO2 1990, Nyberg et al. (2000) decile vs lowest quartile) estimates based on historical emission data and dispersion modelling

amg/m3 if not otherwise indicated Oncogene 6395 Environmental and occupational cancer P Boffetta 6396 Cancer risk from residential radon exposure 1996; Du et al., 1996; Koo and Ho, 1996; Wang et al., 1996; Ko et al., 1997; Zhong et al., 1999). If it appears The carcinogenicity of radon decay products has been plausible that indoor air pollution from combustion or widely studied in occupationally exposed populations, in cooking products (oil vapours in particular) could play a particular underground miners (see below). This agent role in the causation of lung cancer, the relevance of the causes lung cancer in humans, while the evidence for an risks estimated in China for present-day conditions in effect on other neoplasms is not conclusive (IARC, other countries is questionable. Frying is less common in 2001). Although exposure levels in the houses are much most other countries than in China, and kitchens are smaller than in underground mines, the duration of often generally larger, better ventilated and separated exposure and the number of exposed individuals stress from the living quarters. Central heating is increasingly the importance of residence as a source of exposure to common, and open combustion sources indoors are radon decay products. Several case–control studies of infrequent. lung cancer from residential radon exposure have been reported in the literature, and their results have been reviewed and summarized (Lubin and Boice, 1997; Cancer risk from inorganic arsenic in drinking water Darby et al., 2001; IARC, 2001). A pooled RR of Inorganic arsenic causes cancer at various sites in 1.06 (95% CI 1.01–1.10) has been calculated for 3 humans (IARC, 1987). The main source of environ- individuals exposed at 100 Bq/m vs unexposed (Darby mental exposure to arsenic for the general population is et al., 2001), which is in agreement with the extrapola- through ingestion of contaminated water. A high level tion from the results of occupationally exposed popula- of arsenic in groundwater (up to 2–5000 mg/l) is found in tions. Results of studies reported after these pooled many regions of the world, most significantly around the analyses confirm these conclusions (Pisa et al., 2001; Gulf of Bengal, in South America and in Taiwan. There Tomasek et al., 2001; Barros-Dios et al., 2002; Wang is strong evidence of an increased risk of bladder, skin et al., 2002). and lung cancers following consumption of water with Most studies of residential radon rely on the historical high arsenic contamination (IARC, 2004). The evidence reconstruction of exposure levels via household mea- for an increased risk of other cancers, such as those of surements. This approach is subject to substantial the liver, colon and kidney, are weaker but suggestive of misclassification, most likely resulting in an under- a systemic effect. Most of the available studies have been estimate of the risk. In a few studies, attempts were conducted in areas at elevated arsenic content (typically made to correct such biases, and the estimated RR above 200 mg/l). The results of studies of bladder cancer et al increased by about 50% (Lagarde ., 1997; Darby conducted in areas with low or intermediate contamina- et al., 1998). Furthermore, in one study, in which tion are suggestive of a possible increased risk (Bates cumulative radon exposure was estimated from surface et al., 1995; Kurttio et al., 1999). monitors rather than measurement in houses, the RR was higher (1.63, 95% CI 1.07–2.93 for exposure at 100 Bq/m3) (Alavanja et al., 1999). Cancer risk from water chlorination by-products Drinking water may contain a variety of potentially Cancer risk from other sources of indoor air pollution carcinogenic agents, including chlorination by-products, resulting from the interaction of chlorine with organic Based on the observation of very high lung cancer rates chemicals (IARC, 1991). Showering, bathing and in some regions of China and elsewhere among women drinking represent the main sources of exposure to who spend much of their time at home, exposure to chlorination by-products. Among the many halogenated indoor air pollution from combustion sources used for compounds that may be formed, trihalomethanes, heating and cooking, as well as high levels of cooking oil including chloroform, bromodichloromethane, chloro- vapours resulting from some cooking methods, have dibromomethane and bromoform, are those most been identified as risk factors for lung cancer. Three commonly found. Their concentrations of trihalo- main groups of factors influencing indoor air pollution methanes show a wide range, mainly due to the (‘smokiness’) have been studied: (i) heating fuel: type of occurrence of water contamination by organic chemicals. fuel, type of stove or central heating, ventilation, living A meta-analysis of seven studies of bladder cancer area, subjective smokiness; (ii) cooking fuel: type of fuel, risk from consumption of chlorinated water resulted in type of stove or open pit, ventilation of kitchen, location a RR of 1.21 (95% CI 1.09–1.34) (Morris et al., 1992). of cooking area in residence, frequency of cooking, This estimate was not modified after adjusting for smokiness; and (iii) fumes from frying oils: type of oil, smoking. A dose–response relationship was suggested in frequency of frying, eye irritation when cooking. Many three studies (Cantor et al., 1987; McGeehin et al., 1993; of the results are inconclusive, and the interpretation is King and Marrett, 1996). Results based on estimated difficult since the exposure measures used vary con- intake of trihalomethanes and on other disinfection by- siderably. Nonetheless, strong and significant increases products are too sparse to allow a conclusion. in risk have repeatedly been reported and merit Despite the good consistency of the available studies consideration (Gao et al., 1987; Xu et al., 1989, 1991, on bladder cancer, the uncertainties in exposure assess- 1996; Sobue, 1990; Wu-Williams et al., 1990; He et al., ment and possible residual confounding caution against 1991; Lan et al., 1993, 2002; Liu et al., 1993; Dai et al., the conclusion that a causal link has been established

Oncogene Environmental and occupational cancer P Boffetta 6397 between consumption of chlorinated drinking water and Table 3 Occupational agents, classified by the IARC Monographs increased risk of bladder cancer (Cantor, 1997). The programme as carcinogenic to humans (Group 1) evidence for an association between chlorination by- Agents, mixture, circumstance Main industry, use products and cancers in organs other than the bladder is inconclusive (Cantor, 1997). Agents, groups of agents 4-Aminobiphenyl Pigment Arsenic and arsenic compounds Glass, metal, pesticide Asbestos Insulation, filter, textile Occupational cancer Benzene Chemical, solvent Benzidine Pigment Beryllium and beryllium Aerospace A total of 29 occupational and environmental agents, compounds groups of agents and mixtures have been classified as Bis(chloromethyl)ether and Chemical intermediate carcinogenic by IARC (1972–2004), as well as 12 chloromethyl methyl ethera exposure circumstances (Table 3). While some (e.g. Cadmium and cadmium Dye/pigment compounds mustard gas) represent today a historic curiosity, Chromium[VI] compounds Metal plating, dye/pigment exposure is still widespread for carcinogens such as Dioxin Chemical asbestos, coal tar, arsenic and silica. Estimates of the Ethylene oxide Sterilant global burden of occupational cancer result in figures in Mustard gasa War gas 2-Naphthylamine Pigment the order of 2–4% (Doll and Peto, 1981; Peto, 2001). It Nickel compounds Metallurgy, alloy, catalyst should be stressed, however, that these cancers concen- Plutonium-239 and its decay Nuclear industry trate among exposed subjects (mainly male blue-collar products workers), among whom they may represent up to 20% Radium-226 and its decay Luminizing industry a of total cancers (Boffetta et al., 1995). Furthermore, products Radium-228 and its decay Luminizing industry unlike lifestyle factors, exposure is involuntary and can productsa be, to a large extent, avoided. In fact, reduction of Radon-222 and its decay products Mining exposure to occupational carcinogens has taken place in Silica, crystalline Stone cutting, mining, glass industrialized countries during recent decades. Efforts Solar radiation Agriculture Talc containing asbestiform fibres Paper, paints should be made to avoid exposure also in developing Vinyl chloride Plastics countries. Some of the most important known or X- and g-radiation Medical suspected occupational carcinogens are reviewed in detail below. Mixtures Coal-tar pitches b Construction, electrode Coal-tarsb Fuel, construction, chemical Asbestos Mineral oils, untreatedb Metal Shale-oilsb Fuel Asbestos is an important occupational lung carcinogen, Sootsb Pigment first documented following inhalation of asbestos fibres Wood dust Wood in the 1950s (Hughes and Weil, 1994). All different forms of asbestos – chrysotile and amphiboles, including Exposure circumstances Aluminum production crocidolite, amosite and tremolite – are carcinogenic to Auramine, manufacture ofa Pigment humans, causing mesothelioma and lung cancer, Boot and shoe manufacture and although the potency of chrysotile might be lower than repair that of other types, in particular with respect to Coal gasification Coke production mesothelioma risk (IPCS (International Programme on Furniture and cabinet making Chemical Safety), 1998). Haematite mining (underground) Studies of cancer risk from asbestos exposure have with exposure to radon been conducted among miners, manufacturers (asbes- Iron and steel founding a tos-cement, textile, friction materials) and applicators. Magenta, manufacture of Pigment Painter (occupational exposure as a) The interpretation of the results of these studies is Rubber industry complicated by several factors: (i) dose, geological type Strong inorganic-acid mists Metallurgy of fibres and industry are all important determinants of containing sulphuric acid risk and are strictly correlated; (ii) the biologically relevant exposures occurred 20 or more years before aAgent mainly of historical interest. bMixture of polycyclic aromatic appearance of the disease, and their quantitative hydrocarbons assessment is imprecise; (iii) the role of potential confounders, in particular tobacco smoking, can hardly be evaluated. In general, the risk of lung cancer is as compared to grouped, short and coarse fibres, as smaller in studies of miners and friction product those occurring in mining. manufacturers, is intermediate in studies of asbestos Tobacco smoking is the main cause of lung cancer, cement and asbestos product manufacturers, and is and this applies also to cohorts of asbestos exposed highest in studies of asbestos textile workers. This might workers. Despite the limitations of the available studies, reflect a stronger carcinogenic effect of individual, long which limit the precision of the estimate of the combined and thin fibres, as those occurring in the textile industry, effect of the two carcinogens, the relative risk of

Oncogene Environmental and occupational cancer P Boffetta 6398 asbestos exposure seems to be higher among nonsmo- inevitably as complex mixtures of variable composition: kers than among smokers (Liddell, 2001). This suggests an assessment of the risk from individual PAHs is less than multiplicative interaction, which is consistent therefore impossible. An increased risk of lung cancer with a mechanism of carcinogenesis in which the two has been demonstrated in several industries and agents act – at least in part – on the same stages of the occupations entailing exposure to PAHs such as process of cancer development. aluminium production, coal gasification, coke production, iron and steel founding, tar distillation, roofing Heavy metals and chimney sweeping (Boffetta et al., 1997). An increase has also been suggested in a few other Exposure to inorganic arsenic, known as a carcinogen industries, including shale oil extraction, wood impreg- since the late 1960s, occurs mainly among workers nation, road paving, carbon black production and employed in hot smelting; other groups at increased risk carbon electrode manufacture with an exposure–re- are fur handlers, manufacturers of sheep-dip com- sponse relationship in the studies with detailed exposure pounds and pesticides, and vineyard workers (IARC, information. Other cancers that have been definitely 1987; Hayes, 1997). Workers exposed to arsenic suffer associated with occupational exposure to PAHs are from an increased risk of cancer of the lung, skin, and those from the skin, in particular for coal-tar pitches, possibly urinary bladder and liver. Chromium[VI] coal-tars and untreated mineral oils, and the urinary compounds increase the risk for lung and sinonasal bladder, in particular following exposure to coal-tar cancer among chromate production workers, chromate pitches and in aluminium and coal gasification workers pigment manufacturers, chromium platers and ferro- (Boffetta et al., 1997). chromium producers (IARC, 1990a; Hayes, 1997). No Motor vehicle and other engine exhausts represent an such risk has been detected among workers exposed important group of mixtures of PAHs, since they only to chromium[III] compounds (IARC, 1990a). contribute significantly to air pollution. The available Studies of nickel miners, smelters, electrolysis workers epidemiological evidence strongly suggests 40–50% and high-nickel alloy manufacturers showed an in- excess of lung cancer among those occupationally creased risk of lung and sinonasal cancer (IARC, 1990b; exposed to diesel engine exhaust (Lipsett and Cample- Hayes, 1997). There is debate on whether all nickel man, 1999). The evidence for other cancers is inade- compounds are carcinogenic for humans: the available quate. Data on other types of exhausts, including evidence does not allow a clear separation between gasoline engines, do not allow any conclusions (IARC, different nickel salts to which workers are exposed. An 1989). increased risk of lung cancer has been demonstrated among workers in cadmium-based battery manufacture, Strongly reactive chemicals copper–cadmium alloy workers and cadmium smelters and recovery workers (IARC, 1993; Hayes, 1997). Occupational exposure to several strongly reactive chemicals has been shown to cause cancer in humans. Silica and other mineral dusts Workers exposed to chloromethyl methyl ether and bis(chloromethyl) ether are at increased risk of lung An increased risk of lung cancer has been consistently cancer, in particular of small cell type (IARC, 1987; reported in cohorts of silicotic patients (IARC, 1997; Blair and Kazerouni, 1997). An increased risk of Steenland and Stayner, 1997). Many authors investi- laryngeal cancer has been demonstrated in studies gated crystalline silica-exposed workers in foundries, of workers exposed to strong inorganic acids – in pottery, ceramics, diatomaceous earth mining, brick particular sulphuric acid – in metal plating, metal making and stone cutting, some of whom might have treatment, battery manufacture and in the chemical developed silicosis. An increased risk of lung cancer was industry (IARC, 1992; Blair and Kazerouni, 1997). An reported by some, but not all studies, and in the positive increased risk of liver angiosarcoma and possibly other studies the increase has been small (IARC, 1997; tumours has been reported in studies of workers Steenland and Stayner, 1997). Detailed dose–response exposed to vinyl chloride (IARC, 1987). analyses have suggested a linear relationship, with no et al evidence of a threshold (Steenland ., 2001). There is Ionizing radiation inadequate evidence of a carcinogenic effect of exposure to noncrystalline forms of silica (IARC, 1997). Exposure to X- and g-radiation causes leukaemia and solid tumours in humans (IARC, 2000). Occupational Polycyclic aromatic hydrocarbons exposure occurs in several circumstances, notably in medical professions and in the nuclear industry. An Polycyclic aromatic hydrocarbons (PAHs) are a com- increased risk of leukaemia, skin cancer and other solid plex and important group of chemicals formed during tumours has been described in early studies of radi- incomplete combustion of organic material. They are ologists and other medical workers: exposure levels have widespread in the human environment; diet and tobacco greatly decreased in recent decades and it is unclear smoke are two important sources of exposure of PAHs. whether a residual risk still exists. Cancer risk among A number of occupational settings entail exposure to nuclear power plant workers has been extensively high level of PAHs. These chemicals, however, occur studied in several countries: the most accurate estimates

Oncogene Environmental and occupational cancer P Boffetta 6399 suggest an excess RR of leukaemia in the order of 2.2% Table 4 Occupational agents, classified by the IARC Monographs (95% CI 0.1–3.7) per Sv of exposure (IARC Study programme as probably carcinogenic to humans (Group 2A) Group on Cancer Risk among Nuclear Industry Work- Agents, mixture, circumstance Main industry, use ers, 1994). No clear excess was found for solid tumours. Data on cancer among other workers exposed to X- and Agents, groups of agents Acrylamide Chemical, construction g-radiation (e.g. nuclear accident clean-up workers) are Acrylamide Chemical, construction also suggestive of an increased risk of leukaemia (IARC, Benz[a]anthracenea Combustion fumes 2000). Benzidine-based dyes Paper, leather, textile dyes Underground miners exposed to radioactive radon Benzo[a]pyrenea Combustion fumes and its decay products, which emit a-particles, have been 1,3-Butadiene Plastics, rubber Captafol Fungicide consistently found to be at increased risk for lung cancer a-Chlorinated toluenes Chemical intermediate (IARC, 2001). The risk increased with estimated 4-Chloro-ortho-toluidine Dye/pigment manufacture, cumulative exposure and decreased with attained age, textiles time since exposure, and time since cessation of Dibenz[a,h]anthracenea Combustion fumes Diethyl sulphate Chemical intermediate exposure (National Research Council, 1999). The excess Dimethylcarbamoyl chloride Chemical intermediate RR estimated from occupational cohorts, which in- Dimethyl sulphate Chemical intermediate cluded over 2500 cases of lung cancer occurring among Epichlorohydrin Plastics/resins monomer over 60 000 miners, has been estimated in the order of Ethylene dibromide Chemical intermediate, 0.0049 per working level month of exposure (Lubin fumigant Formaldehyde Plastics, textiles, laboratory et al., 1995). Further refinements of this estimates took agent into account age at exposure and time since first Glycidol Chemical intermediate exposure (National Research Council, 1999), as well as 4,40-Methylene bis (2- Rubber manufacture smoking status, with a stronger effect being shown chloroaniline) (MOCA)b N-Nitrosodimethylamineb Chemical intermediate among never-smokers than among smokers. An carci- Styrene-7,8-oxide Plastics, chemical nogenic effect of radon exposure on organs other than intermediate the lung is not substantiated. Tetrachloroethylene Solvent, dry cleaning Other radionuclides to which humans are, or have Ortho-Toluidine Dyestuff, rubber been, exposed in the workplace include radium-226 Trichloroethylene Solvent, dry cleaning, metal 1,2,3-Trichloropropane Solvent, chemical (increased risk of osteosarcoma among watch-dial intermediate painters) and plutonium-239 (increased risk of lung, Tris(2,3- Plastics, textiles, flame liver and bone neoplasms among heavily exposed dibromopropyl)phosphate retardant workers in plutonium production plants) (IARC, 2001). Vinyl bromide Plastics, textiles, monomer Vinyl fluoride Chemical intermediate

Exposure circumstances Mixtures Creosotesc Wood preservation An increased risk of cancer has been demonstrated Diesel engine exhaustc Transport among workers employed in several jobs and industries Nonarsenical insecticides Agriculture (spraying and application) (Table 3). While for some of these exposure circum- Polychlorinated biphenyls Electrical components stances the agents responsible for the excess cancer risk have been identified (e.g. wood dust in furniture making, Exposure circumstances mixtures of PAHs for aluminium, foundry, coke and Art glass, glass container and pressed ware (manuf. of) coal gasification workers), this has not been possible for Hairdresser and barber others, such as painters and rubber workers, which Petroleum refining limits the scope of preventive measures. aComponent of mixtures of polycyclic aromatic hydrocarbons. bAgent mainly of historical interest. cMixture of polycyclic aromatic hydro- Suspected carcinogens carbons Some 31 occupational agents and three exposure circumstances are classified as probable carcinogens (Table 4). They include important chemicals such as tri- 1998). The most informative cohort study of workers and tetrachloroethylene, formaldehyde and butadiene. exposed to butadiene was conducted among workers The reasons for the classification as suspected rather employed in the production of synthetic rubber based than established carcinogens lay mainly in the inade- on styrene and butadiene (Macaluso et al., 1996; quacy of the available evidence from epidemiological Sathiakumar et al., 1998). This study showed a clear studies, due to limited statistical power to detect a small increased risk of leukaemia, based on 58 deaths, with a excess of rare neoplasms, limitations in the available dose–response (RRs 1.0 (reference group), 1.5(95%CI exposure data and concomitant exposure to other 0.7–3.2) and 1.7 (95% CI 0.8–3.9) in the three categories potentially hazardous chemicals. of increasing butadiene exposure; P-value of test for The example of butadiene, an important occupational linear trend 0.03), which was not explained by exposure agent which under experimental conditions causes to styrene or other agents. These results receive a limited leukaemia in mice, is particularly illustrative (IARC, support by the only other large cohort study conducted

Oncogene Environmental and occupational cancer P Boffetta 6400 among butadiene production workers (Divine and order of 0.5–1% (Doll and Peto, 1981; Peto, 2001). Hartman, 1996), while no clear excess was present in However, this proportion is likely to be higher in the remaining, smaller studies (Bond et al., 1992; Cowles populations with special exposure circumstances, such et al., 1994; Ward et al., 1996). Overall, the presence of a as arsenic contamination of drinking water or proximity causal association between butadiene exposure and to an important source of asbestos exposure. Further- occurrence of cancer (leukaemia in particular) in more, the involuntary nature of the exposure makes its humans is plausible: this conclusion is supported by a elimination a particularly important target in terms of single large and very carefully conducted study and is health equity. compatible with the results of an additional valid, Knowledge on occupational causes of cancer is more although less informative, study. The inconsistency in advanced than in the case of other groups of carcino- the results of additional studies should be seen in the gens. However, current understanding of the relation- light of potential limitations, namely limited power, ship between occupational exposures and cancer risk is misclassification of leukaemia on death certificates and far from complete: for many experimental carcinogens nondifferential misclassification of exposure. that occur in the workplace, little or no human cancer data are available. Examples of such agents are ceramic fibres, acrylamide, dichloromethane and acetaldehyde: Conclusions they all cause tumours in experimental animals and no adequate data in exposed humans are available. One A number of circumstances of environmental exposure should also consider that the relatively low burden of to carcinogens has definitely been linked with an occupational cancer in industrialized countries is the increased risk of cancer in humans. For all of them, it successful result of strict regulations on recognized is not possible to quantify with precision the burden of carcinogens. As in the case of environmental cancer, the human cancer they cause. It is likely however that in facts that exposure is involuntary and can be eliminated most populations, environmental cancer is responsible give to prevention of occupational cancer a special for a relatively small proportion of total cancers, in the status among interventions towards cancer control.

References

Abbey DE, Nishino N, McDonnell WF, Burchette RJ, Buell P and Dunn JE. (1967). Arch. Environ. Health, 15, 291– Knutsen SF, Beeson WL and Yang JX. (1999). Am. J. 297. Respir. Crit. Care Med., 159, 373–382. Buell P, Dunn JE and Breslow L. (1967). Cancer, 20, 2139– Alavanja MC, Lubin JH, Mahaffey JA and Brownson RC. 2147. (1999). Am. J. Public Health, 89, 1042–1048. Camus M, Siemiatycki J and Meek B. (1998). New Engl. J. Barbone F, Bovenzi M, Cavallieri F and Stanta G. (1995). Am. Med., 338, 1565–1571. J. Epidemiol., 141, 1161–1169. Cantor KP. (1997). Cancer Causes Control, 8, 292–308. Barros-Dios JM, Barreiro MA, Ruano-Ravina A and Fig- Cantor K, Hoover R, Hartge P, Mason TJ, Silverman DT, ueiras A. (2002). Am. J. Epidemiol., 156, 548–555. Altman R, Austin DF, Child MA, Key CR, Marrett LD, Bates MN, Smith AH and Cantor KP. (1995). Am. J. Myers MH, Narayama AS, Levin LI, Sullivan JW, Swanson Epidemiol., 141, 523–530. GM, Thomas DB and West DW. (1987). J. Natl. Cancer Beeson WL, Abbey DE and Knutsen SF. (1998). Environ. Inst., 79, 1269–1279. Health Perspect., 106, 813–822. Cederlo¨ f R, Friberg L, Hrubek Z and Lorich U. (1975). The Bhopal RS, Moffatt S, Pless-Mulloli T, Phillimore PR, Foy C, Relationship of Smoking and Some Social Covariables to Dunn CE and Tate JA. (1998). Occup. Environ. Med., 55, Mortality and Cancer Morbidity: A Ten Year Follow-up in a 812–822. Probability Sample of 55000 Swedish Subjects Age 18 to 69,2 Biggeri A, Barbone F, Lagazio C, Bovenzi M and Stanta G. volumes. Department of Environmental Hygiene, Karolins- (1996). Environ. Health Perspect., 104, 750–754. ka Institute: Stockholm. Blair A and Kazerouni N. (1997). Cancer Causes Control, 8, Cordier S, Theriault G and Iturra H. (1983). Environ. Res., 31, 473–490. 311–322. Blot WJ and Fraumeni Jr JF. (1975). Lancet, 2, 142–144. Cowles SR, Tsai SP, Snyder PJ and Ross CE. (1994). Occup. Boffetta P. (2002). Scand. J. Work Environ. Health, 28, 30–40. Environ. Med., 51, 323–329. Boffetta P, Jourenkova N and Gustavsson P. (1997). Cancer Dai XD, Lin CY, Sun XW, Shi YB and Lin YJ. (1996). Lung Causes Control, 8, 444–472. Cancer, 14, 85–91. Boffetta P, Kogevinas M, Simonato L, Wilbourn J and Darby S, Hill D and Doll R. (2001). Ann. Oncol., 12, 1341– Saracci R. (1995). Int. J. Occup. Environ. Health, 1, 1351. 315–325. Darby S, Whitley E, Silcocks P, Thakrar B, Green M, Lomas Bond GG, Bodner KM, Olsen GW and Cook RR. (1992). P, Miles J, Reeves G, Fearn T and Doll R. (1998). Br. J. Scand. J. Work Environ. Health, 18, 145–154. Cancer, 78, 394–408. Botha JL, Irwig LM and Strebel P. (1986). Am. J. Epidemiol., Dean G. (1966). BMJ, 5502, 1506–1514. 123, 30–40. Dean G, Lee PN, Todd GF and Wicken AJ. (1977). Report on Brown LM, Pottern LM and Blot WJ. (1984). Environ. Res., a Second Retrospective Mortality Study in North-East 34, 250–261. England, Part I: Factors Related to Mortality from Lung Brownson RC, Reif JS, Keefe TJ, Ferguson SW and Pritzl JA. Cancer, Bronchitis, Heart Disease and Stroke in Cleveland (1987). Am. J. Epidemiol., 125, 25–34. County, with Particular Emphasis on the Relative Risks

Oncogene Environmental and occupational cancer P Boffetta 6401 Associated with Smoking Filter and Plain Cigarettes. Monographs on the Evaluation of the Carcinogenic Risk of Tobacco Research Council: London. Chemicals to Humans, Suppl 7 IARC: Lyon. Dean G, Lee PN, Todd GF and Wicken AJ. (1978). Report on IARC (1989). Diesel and Gasoline Engine Exhausts and Some a Second Retrospective Mortality Study in North-East Nitroarenes, IARC Monographs on the Evaluation of England, Part II: Changes in Lung Cancer and Bronchitis Carcinogenic Risks to Humans, Vol 46. IARC: Lyon, Mortality and in Other Relevant Factors Occurring in Areas pp. 41–185. of North-East England, 1963–72. Tobacco Research Council: IARC (1990a). Chromium, Nickel and Welding, IARC Mono- London. graphs on the Evaluation of Carcinogenic Risks to Humans, Divine BJ and Hartman CM. (1996). Toxicology, 113, Vol 49. IARC: Lyon, pp. 49–256. 169–181. IARC (1990b). Chromium, Nickel and Welding, IARC Mono- Dockery DW, Pope III CA, Xu X, Spengler JD, Ware JH, Fay graphs on the Evaluation of Carcinogenic Risks to Humans, ME, Ferris Jr BG and Speizer FE. (1993). New Engl. J. Vol 49. IARC: Lyon, pp. 257–445. Med., 329, 1753–1759. IARC (1991). Chlorinated Drinking Water; Chlorination By- Doll R and Peto R. (1981). J. Natl. Cancer Inst., 66, Products; Some Other Halogenated Compounds; Cobalt and 1191–1308. Cobalt Compounds, IARC Monographs on the Evaluation of Du YX, Cha Q, Chen XW, Chen YZ, Huang LF, Feng ZZ, Carcinogenic Risks to Humans, Vol 52. IARC: Lyon. Wu XF and Wu JM. (1996). Lung Cancer, 14, 9–37. IARC (1992). Occupational Exposures to Mists and Vapours Engholm G, Palmgren F and Lynge E. (1996). BMJ, 312, from Strong Inorganic Acids and Other Industrial Chemicals, 1259–1263. IARC Monographs on the Evaluation of Carcinogenic Frost F, Harter L, Milham S, Royce R, Smith AH, Hartley J Risks to Humans, Vol 54. IARC: Lyon, pp. 41–119. and Enterline P. (1987). Arch. Environ. Health, 42, IARC (1993). Beryllium, Cadmium, Mercury and Exposures in 148–152. the Glass Manufacturing Industry, IARC Monographs on the Gailey FA and Lloyd OL. (1983). Ecol. Dis., 2, 419–431. Evaluation of Carcinogenic Risks to Humans, Vol 58. Gailey FA and Lloyd OL. (1986). Environ. Health Perspect., IARC: Lyon, pp. 119–237. 68, 187–196. IARC (1997). Silica, Some Silicates, Coal Dust and Para- Gao YT, Blot WJ, Zheng W, Ershow AG, Hsu CW, Levin LI, Aramid Fibrils, IARC Monographs on the Evaluation of Zhang R and Fraumeni Jr JF. (1987). Int. J. Cancer, 40, Carcinogenic Risks to Humans, Vol 68. IARC: Lyon, 604–609. pp. 41–242. Goldsmith JR. (1980). J. Environ. Pathol. Toxicol., 3, 205–217. IARC (1998). Re-evaluation of Some Organic Chemicals, Greaves WW, Rom WN, Lyon JL, Varley G, Wright DD and Hydrazine and Hydrogen Peroxide (Part One), IARC Chiu G. (1981). Am. J. Ind. Med., 2, 15–23. Monographs on the Evaluation of Carcinogenic Risks to Hackshaw AK, Law MR and Wald NJ. (1997). BMJ, 315, Humans, Vol. 71. IARC: Lyon, pp. 109–225. 980–988. IARC (2000). Ionizing Radiation, Part 1: X- and Gamma- Haenszel W, Loveland DB and Sirken MG. (1962). J. Natl. Radiation, and Neutrons, IARC Monographs on the Cancer Inst., 28, 947–1001. Evaluation of Carcinogenic Risks to Humans, Vol 75. Haenszel W and Taeuber KE. (1964). J. Natl. Cancer Inst., 32, IARC: Lyon. 803–838. IARC (2001). Ionizing Radiation, Part 2: Some Internally Hammond EC. (1972). Environmental Factors in Respiratory Deposited Radionuclides, IARC Monographs on the Evalua- Disease, Lee DH (ed). Academic Press: New York, pp. tion of Carcinogenic Risks to Humans, Vol 78. IARC: Lyon. 177–198. IARC (2004). Tobacco Smoking and Involuntary Tobacco Hammond EC and Garfinkel L. (1980). Prev. Med., 9, Smoke, IARC Monographs on the Evaluation of Carcino- 206–211. genic Risks to Humans, Vol 83. IARC: Lyon, (in press). Hammond EC, Garfinkel L, Selikoff IJ and Nicholson WT. IARC Study Group on Cancer Risk among Nuclear Industry (1979). Ann. NY Acad. Sci., 330, 417–422. Workers (1994). Lancet, 344, 1039–1043. Hammond EC and Horn D. (1958a). JAMA, 166, 1159–1172. INSERM (1997). Effets sur la Sante´ des Principaux Types Hammond EC and Horn D. (1958b). JAMA, 166, 1294–1308. d’Exposition a` l’Amiante, (Expertise Collective INSERM). Hansen J, de Klerk NH, Musk AW and Hobbs MST. (1998). Institut National de la Santeét de la Recherche Me´dicale: Am. J. Respir. Crit. Care Med., 157, 69–75. Paris. Hayes RB. (1997). Cancer Causes Control, 8, 371–385. IPCS (International Programme on Chemical Safety) (1998). He XZ, Chen W, Liu ZY and Chapman RS. (1991). Environ. Chrysotile Asbestos, Environmental Health Criteria 203. Health Perspect., 94, 9–13. World Health Organization: Geneve. Hitosugi M. (1968). Bull. Inst. Public Health, 17, 237–256. Jedrychowski W, Becher H, Wahrendorf J and Basa-Cierpia- Hoek G, Brunekreef B, Goldbohm S, Fischer P and van den lek Z. (1990). J. Epidemiol. Community Health, 44, 114–120. Brandt PA. (2002). Lancet, 360, 1203–1209. Jo¨ ckel KH, Ahrens W, Wichmann HE, Becher H, Bolm- Holowaty EJ, Risch HA, Miller AB and Burch JD. (1991). Audorff U, Jahn I, Molik B, Greiser E and Timm J. (1992). Can. J. Public Health, 82, 304–309. Int. J. Epidemiol., 21, 202–213. Howel D, Arblaster L, Swinburne L, Schweiger M, Renvoize E Katsouyanni K, Trichopoulos D, Kalandidi A, Tomos P and and Hatton P. (1997). Occup. Environ. Med., 54, 403–409. Riboli E. (1991). Prev. Med., 20, 271–278. Hughes JM and Weill H. (1994). Epidemiology of Lung Cancer. King WD and Marrett LD. (1996). Cancer Causes Control, 7, Lung Biology in Health and Disease, Samet JM (ed). Vol 74. 596–604. Marcel Dekker: New York, pp. 185–205. Ko YC, Lee CH, Chen MJ, Huang CC, Chang WY, Lin HJ, IARC (1972–2004). IARC Monographs on the Evaluation of Wang HZ and Chang PY. (1997). Int. J. Epidemiol., 26, Carcinogenic Risks to Humans, Vols 1–80. International 24–31. Agency for Research on Cancer: Lyon. Koo LC and Ho JH. (1996). Lung Cancer, 14, 47–61. IARC (1987). Overall Evaluations of Carcinogenicity: An Kurttio P, Pukkala E, Kahelin H, Auvinen A and Pekkanen J. Updating of IARC Monographs Volumes 1 to 42, IARC (1999). Environ. Health Perspect., 107, 705–710.

Oncogene Environmental and occupational cancer P Boffetta 6402 Lagarde F, Pershagen G, Akerblom G, Axelson O, Baverstam National Research Council (1999). Committee on Health Risks U, Damber L, Enflo A, Svartengren M and Swedjemark of Exposure to Radon: BEIR VI, Health effects of exposure to GA. (1997). Health Phys., 72, 269–276. radon. National Academy Press: Washington, DC. Lan Q, Chapman RS, Schreinemachers DM, Tian L and He Neuberger M, Kundi M and Friedl HP. (1984). Arch. Environ. X. (2002). J. Natl. Cancer Inst., 94, 826–835. Health, 39, 261–265. Lan Q, Chen W, Chen H and He XZ. (1993). Biomed. Environ. Newhouse M and Thompson H. (1965). Br. J. Ind. Med., 2, Sci., 6, 112–118. 261–269. Levin ML, Haenszel W, Carroll BE, Gerhardt PR, Handy VH Newman JA, Archer VE, Saccomanno G, Kuschner M, and Ingraham SC. (1960). J. Natl. Cancer Inst., 24, Auerbach O, Grondahl RD and Wilson JC. (1976). Ann. 1243–1257. NY Acad. Sci., 271, 260–268. Liddell FD. (2001). Ann Occup. Hyg., 45, 341–356. Nyberg F, Gustavsson P, Jarup L, Bellander T, Berglind N, Lipsett M and Campleman S. (1999). Am. J. Public Health, 89, Jakobsson R and Pershagen G. (2000). Epidemiology, 11, 1009–1017. 487–495. Pawlega J, Rachtan J and Dyba T. (1997). Acta Oncol., Liu Q, Sasco AJ, Riboli E and Hu MX. (1993). Am. J. 36, 471–476. Epidemiol., 137, 145–154. Pershagen G. (1985). Am. J. Epidemiol., 122, 684–694. Lloyd OL. (1978). Lancet, 1, 318–320. Pershagen G, Elinder CG and Bolander AM. (1977). Environ. Lloyd OL, Ireland E, Tyrrell H and Williams F. (1986). J. Soc. Health Perspect., 19, 133–137. Occup. Med., 36, 2–8. Peto J. (2001). Nature, 411, 390–395. Lloyd OL, Smith G, Lloyd MM, Holland Y and Gailey F. Petrauskaite R, Pershagen G and Gurevicius R. (2002). Int. J. (1985a). Br. J. Ind. Med., 42, 475–480. Cancer, 99, 106–111. Lloyd OL, Williams FL and Gailey FA. (1985b). Br. J. Ind. Pike MC, Jing JS, Rosario IP, Henderson BE and Menck HR. Med., 42, 815–823. (1979). Energy and Health, Breslow N, Whitemore A (eds) Lubin JH and Boice Jr JD. (1997). J. Natl. Cancer Inst., 89, SIAM: Philadelphia, PA, pp. 3–16. 49–57. Pisa FE, Barbone F, Betta A, Bonomi M, Alessandrini B and Lubin JH, Boice Jr JD, Edling C, Hornung RW, Howe GR, Bovenzi M. (2001). Arch. Environ. Health, 56, 208–215. Kunz E, Kusiak RA, Morrison HI, Radford EP, Samet JM, Pope III CA, Burnett RT, Thun MJ, Calle EE, Krewski D, Ito Tirmarche M, Woodward A, Yao SX and Pierce DA. K and Thurston GD. (2002). JAMA, 287, 1132–1141. (1995). J. Natl. Cancer Inst., 87, 817–827. Pope III CA, Thun MJ, Namboodiri MM, Dockery DW, Luce D, Bugel I, Goldberg P, Goldberg M, Salomon C, Billon- Evans JS, Speizer FE and Heath Jr CW. (1995). Am. J. Galland MA, Nicolau J, Quenel P, Fevotte J and Brochard Respir. Crit. Care Med., 151, 669–674. P. (2000). Am. J. Epidemiol., 151, 259–265. Pott P. (1775). Chirurgical Observations. Hawes, Clarke and Luo S, Zhang Y, Mu S, Zhang C, Ma T and Liu X. (1998). Collins: London. Hua Xi Yi Ke Da Xue Xue Bao, 29, 63–65(in Chinese). Rom WN, Varley G, Lyon JL and Shopkow S. (1982). Br. J. Lyon JL, Fillmore JL and Klauber MR. (1977). Lancet, 2, 869. Ind. Med., 39, 269–272. Macaluso M, Larson R, Delzell E, Sathiakumar N, Hovinga Samet JM, Humble CG, Skipper BE and Pathak DR. (1987). M, Julian J, Muir D and Cole P. (1996). Toxicology, 113, Am. J. Epidemiol., 125, 800–811. 190–202. Sathiakumar N, Delzell E, Hovinga M, Macaluso M, Julian Magnani C, Agudo A, Gonzalez CA, Andrion A, Calleja A, JA, Larson R, Cole P and Muir DCF. (1998). Occup. Chellini E, Dalmasso P, Escolar A, Hernandez S, Ivaldi C, Environ. Med., 55, 230–235. Mirabelli D, Ramirez J, Turuguet D, Usel M and Terracini Shear CL, Seale DB and Gottlieb MS. (1980). Arch. Environ. B. (2000). Br. J. Cancer, 83, 104–111. Health, 35, 335–343. Magnani C, Dalmasso P, Biggeri A, Ivaldi C, Mirabelli D and Smith GH, Williams FL and Lloyd OL. (1987). Br. J. Ind. Terracini B. (2001). Environ. Health Perspect., 109, 915–919. Med., 44, 795–802. Magnani C, Terracini B, Ivaldi C, Botta M, Mancini A and Sobue T. (1990). Int. J. Epidemiol., 19, 62–66. Andrion A. (1995). Occup. Environ. Med., 52, 362–367. Steenland K, Mannetje A, Boffetta P, Stayner L, Attfield M, Marsh GM, Stone RA, Esmen NA, Gula MJ, Gause CK, Chen J, Dosemeci M, DeKlerk N, Hnizdo E, Koskela R and Petersen NJ, Meaney FJ, Rodney S and Prybylski D. (1997). Checkoway H. (2001). Cancer Causes Control, 12, 773–784. Environ. Res., 75, 56–72. Steenland K and Stayner L. (1997). Cancer Causes Control, 8, Marsh GM, Stone RA, Esmen NA, Gula MJ, Gause CK, 491–503. Petersen NJ, Meaney FJ, Rodney S and Prybylski D. (1998). Stocks P and Campbell JM. (1955). BMJ, 2, 923–929. Arch. Environ. Health, 53, 15–28. Swaen GM and Slangen JJ. (1995). Int. Arch. Occup. Environ. Health, 67, 85–93. Matanoski GM, Landau E, Tonascia J, Lazar C, Elliott EA, Tenkanen L. (1993). Cancer Causes Control, 4, 133–141. McEnroe W and King K. (1981). Environ. Res., 25, 8–28. Tenkanen L, Hakulinen T, Hakama M and Saxen E. (1985). McDonnell WF, Nishino-Ishikawa N, Petersen FF, Chen LH Int. J. Cancer, 35, 637–642. and Abbey DE. (2000). J. Expo. Anal. Environ. Epidemiol., Theriault GP and Grand-Bois L. (1978). Arch. Environ. 10, 427–436. Health, 33, 15–19. McGeehin MA, Reif JS, Becher JC and Mangione EJ. (1993). Tenkanen L and Teppo L. (1987). Scand. J. Soc. Med., 15, 67–72. Am. J. Epidemiol., 138, 492–501. Tomasek L, Kunz E, Mu¨ ller T, Hulka J, Heribanova A, Mills PK, Abbey DE, Beeson WL and Petersen F. (1991). Matzner J, Placek V, Burian I and Holecek J. (2001). Sci. Arch. Environ. Health, 46, 271–280. Total Environ., 272, 43–51. Morris RD, Audet AM, Angelillo IF, Chalmers TC and Tomatis L, Aitio A, Day NE, Heseltine E, Kaldor J, Miller Mosteller F. (1992). Am. J. Public Health, 82, 955–963. AB, Parkin DM and Riboli E. (1990). Cancer: Causes, Mzileni O, Sitas F, Steyn K, Carrara H and Bekker P. (1999). Occurrence and Control, IARC Scientific Publications No Tobacco Control, 8, 398–401. 100 IARC: Lyon.

Oncogene Environmental and occupational cancer P Boffetta 6403 Vena JE. (1982). Am. J. Epidemiol., 116, 42–56. Xu ZY, Blot WJ, Li G, Fraumeni Jr JF, Zhao DZ, Stone BJ, Wang SY, Hu YL, Wu YL, Li X, Chi GB, Chen Y and Dai Yin Q, Wu A, Henderson BE and Guan BP. (1991). WS. (1996). Lung Cancer, 14, 99–105. Relevance to Human Cancer of N-Nitroso Compounds, Wang Z, Lubin JH, Wang L, Zhang S, Boice Jr JD, Cui H, Tobacco Smoke and Mycotoxins, O’Neill IK, Chen J, Zhang S, Conrath S, Xia Y, Shang B, Brenner A, Lei S, Bartsch H (eds). IARC Scientiﬁc Publications No 105. Metayer C, Cao J, Chen KW, Lei S and Kleinerman RA. IARC: Lyon, pp. 460–465. (2002). Am. J. Epidemiol., 155, 554–564. Xu ZY, Blot WJ, Xiao HP, Wu A, Feng YP, Stone BJ, Sun J, Ward EM, Fajen JM, Ruder AM, Rinsky RA, Halperin WE Ershow AG, Henderson BE and Fraumeni Jr JF. (1989). and Fessler-Flesch CA. (1996). Toxicology, 113, 157–168. J. Natl. Cancer Inst., 81, 1800–1806. Williams FL and Lloyd OL. (1988). Public Health, 102, Xu ZY, Brown L, Pan GW, Li G, Feng YP, Guan DX, Liu 531–538. TF, Liu LM, Chao RM, Sheng JH and Gao GC. (1996). Wu-Williams AH, Dai XD, Blot W, Xu ZY, Sun XW, Xiao Lung Cancer, 14, 149–160. HP, Stone BJ, Yu SF, Feng YP, Ershow AG, Sun J, Zaridze DG, Zemlianaia GM and Aitakov ZN. (1995). Vestnik Fraumeni Jr JF and Henderson BE. (1990). Br. J. Cancer, Rossiiskoi Akademii Meditsinskikh Nauk, 4, 6–10 (in Russian). 62, 982–987. Zhong LJ, Goldberg MS, Parent ME and Hanley JA. (1999). Xiao HP and Xu ZY. (1985). NCI Monogr., 69, 53–58. Scand. J. Work Environ. Health, 25, 309–316.

Oncogene