Occupational Lung Cancer

Occupational Lung Cancer

PETER L. WHITESELL, M.D., CHARLES W. DRAGE, M.D., Division of Thoracic Diseases and Internal Medicine

Exposure to certain industrial agents has been thought to have carcinogenic potential, both for employees who work closely with such agents and for the general population that comes in contact with them. Although case reports, laboratory studies, and epidemiologic analyses help to determine the carcinogenicity of implicated agents, each of these types of investigation has limitations and deficiencies in distinguishing causal from noncausal associations. Asbestos has been linked with bronchogenic carcinoma, but several controversial factors-the degree of risk relative to exposure dose, the synergistic effect of cocarcinogens, and the question of existence of a threshold dose-complicate the understanding of the magnitude of the risk for exposed persons. Several other physical and chemical agents (such as chromium, nickel, and radon) have also been associated with an increased incidence of lung cancer in epidemiologic and animal studies, As with asbestos, the specific type of the agent and exposure conditions are important in determining the degree of carcinogenicity. In studies of exposure to man-made mineral fibers, formaldehyde, and silica, the findings have been inconsistent. Because the degree of health hazard attributable to asbestos and other known and suspected lung carcinogens is controversial, a wide range of opinions exists about the importance of occupational exposures to the overall incidence of lung cancers. Nevertheless, attempting to prevent lung cancers by minimizing or eliminating exposure to carcinogens is preferable to treating existent cases.

The etiologic association between exposure to industrial agents and bronchogenic carcinomas is a subject with considerable complexity, controversy, and, above all, consequence. Its potential effect extends not only to workers in occupations directly handling such agents but also to the population at large that comes in contact with such agentsusually, but not always, at lower levels of exposure. On review of the relatively extensive scientific and popular literature on the topic, one finds a diverse spectrum of opinions, ranging from complacent disregard to fanatic overemphasis. In this article, we discuss the difficulties inherent in delineating the carcinogenic potential of workplace exposures and present an overview of data on specific agents and occupations. For more detailed analyses of specific agents, readers can consult several published reviews!" and updated Individual reprints of this article are not available. The entire Symposium on Intrathoracic Neoplasms will be available for purchase as a bound booklet from the Proceedings Circulation Office at a later date. Mayo Clin Proc 1993; 68:183-188

information in monographs from the International Agency for Research on Cancer. EVIDENCE FOR AN ETIOLOGIC ASSOCIATION Data suggestive of an etiologic relationship between specific agents or occupations and lung cancer are available from case reports, laboratory studies, and epidemiologic research. Each of these resources has merits and disadvantages. Case reports by "astute clinicians" provided the first awareness of the potential carcinogenicity of chemical and physical agents. As early as the 16th century, an increased frequency of fatal lung disease, subsequently shown to be in part neoplastic, was noted among miners of uranium-bearing ore in the Erzgebirge Mountains of Europe." Case reports may be highly suggestive of a causal relationship when either the neoplasm is extremely uncommon, such as a mesothelioma, or the agent is an exceedingly potent carcinogen, such as chloromethyl methyl ether-which has been noted to increase bronchogenic cancers 2,OOO-fold in some settings" and first aroused suspicion when management per183

© 1993 Mayo Foundation for Medical Education and Research

184

OCCUPATIONALLUNGCANCER

sonnel noted an excessive number of cases of lung cancer among employees in one area of a chemical plant." Bronchogenic cancers, however, are relatively common in the general population, and no pathognomonic histologic features indicate an occupational cause. In addition, suspected carcinogens may cause only a minimal increase in the frequency of cancers. Nevertheless, even a slight increase would be important to recognize if numerous workers or the general population is exposed. The role of case reports is therefore limited. Laboratory testing is extensively used to evaluate the carcinogenicity of various agents. In vitro analysis, such as the demonstration of bacterial chromosomal mutation or the transformation of cell lines, allows relatively rapid and inexpensive screening of large numbers of chemical agents but fails to reflect the effects of in vivo metabolism and has uncertain relevance to the pathogenesis of neoplasms in humans. Animal studies have been more biologically relevant than in vitro investigations. All compounds currently established by the International Agency for Research on Cancer as carcinogenic in humans have also demonstrated carcinogenicity in at least one animal model." Time, expense, and ethical considerations limit the application of this approach. A prospective, placebo-controlled, randomized trial of the carcinogenicity of an agent is clearly impossible. Most epidemiologic data are generated from one of two approaches. Case-control studies identify patients with the disease (that is, lung cancer) and compare the frequency of their exposure to the investigated agent or occupation with that of a comparable population without lung cancer. Cohort studies compare the incidence of lung cancer in otherwise comparable exposed and unexposed populations. Both yield a relative risk for the development oflung cancer in comparison with that of a control population-a standard mortality ratio from cohort studies and an odds ratio from case-control studies. Although more directly relevant to humans than in vitro or animal models, epidemiologic studies also are associated with methodologic problems. The failure to select a control population that is comparable with respect to confounding variables may introduce biases that alter the relative risk associated with the agent under investigation. Early studies of exposure to asbestos included no data on smoking. Many more recent studies of occupational agents failed to recognize coexposure to asbestos or radon daughters. Seemingly minor variations in coexposures may have a major role whenever the increased risk from the agent under investigation is small. Bias may also be introduced at the time of collection of data. Patients and especially relatives of deceased patients with lung cancer tend to remember and report occupational exposures but underreport a history of

Mayo Clin Proc, February 1993, Vol 68

smoking, a potential problem known as recall bias." Finally, when many occupations are surveyed, investigators should know that an apparent increase in the standard mortality ratio may arise through chance occurrence, especially if the relative risk is low. Ideally, the associations found in an epidemiologic study should not be attributable to confounding variables or to biases in collection of data and should be replicated in several independent studies. A biologically appropriate doseresponse effect should be demonstrated, and the period of study should be sufficient to account for the long latency period noted with almost all lung cancers-up to 30 to 40 years for most agents. Hill'? proposed criteria that may be useful in distinguishing causal from noncausal associations. In addition to the aforementioned features, they include a large magnitude of association, biologic plausibility, experimental support, and analogy to other well-documented associations. Nonetheless, determining causality is frequently a difficult, tenuous process.

SPECIFIC AGENTS Asbestos.-After introduction of asbestos into widespread commercial use in the late 19th century, an awareness of the sequelae of pulmonary fibrosis first prompted restrictions of workplace exposure to this agent in the 1930s-before its carcinogenic potential was suspected. In 1955, a report by Sir Richard Doll provided the first conclusive epidemiologic data that linked asbestos with lung cancer; the occurrence of bronchogenic cancers among 113 male workers with more than a 20-year history of exposure to asbestos was 10 times that expected." The occurrence of mesothelioma (at that time a rare tumor) in asbestos-exposed patients was first described by Wagner and associates" in 1960. (Mesotheliomas are discussed elsewhere in this symposium.) Nevertheless, not until the 1970s did production of asbestos peak, and as of 1990, worldwide annual production remains at approximately 1 kg/yr per capita. 13 Although the carcinogenic properties of asbestos are established, several controversies complicate our understanding of the magnitude of the risk for exposed workers or the general population. One factor is the degree of risk relative to exposure dose. Although a response (that is, frequency of lung cancerj-versus-dose relationship has been found in almost all studies, the slope of this relationship has varied by as much as 50-fold in the various studies." The risk of exposure apparently depends not only on the amount of contact but also on the fiber type, the industrial use, and even the exact exposure conditions as well. For example, for the same cumulative exposure (number of fibers/em' of air per year), tremolite miners had a severalfold greater relative risk than did chrysotile miners. 15 The finding that textile workers seemed to have a greater relative risk than did miners with

Mayo Clio Proc, February 1993, Vol 68

similar cumulative exposures to asbestos from the same quarry suggests that cocarcinogenic compounds such as textile oils or fiber-processing effects may also be important in determining the overall risk." An additional factor in any consideration of risk is the synergistic effect of smoking on carcinogenesis. Several studies have indicated a multiplicative rather than an additive interaction-for example, a lO-fold risk from smoking and a 5-fold risk from asbestos would produce a 50fold (not a IS-fold) relative risk of lung cancer.P-'? Because of the small numbers and the low statistical power of studies that have addressed the risk of lung cancer in asbestos-exposed, nonsmoking populations, some investigators have questioned whether exposure to asbestos alone promotes lung cancers, but this contention has been fiercely debated."?" The question of the existence of a threshold dose is important in determining the extent of the population at risk and the overall effect of asbestos on public health. Proponents of a threshold dose emphasize that most epidemiologic studies identify a dose below which the exposed population has no demonstrated increased risk and that the same type of threshold dose has been noted in animal studies." Alternatively, other investigators argue that the lack of effect at lowdoses only reflects the difficulty in generating statistical significance because of the small numbers involved. As a consequence of these complexities and differences in interpretation of data, disagreement exists over the seriousness of the asbestos problem for both exposed workers and the general population and over the appropriateness of restrictions on further use of the product.18,19 A recent article maintains that the benefits of affordable potable water and disposable sewage through the use of asbestos cement pipes in Third World countries far exceed the potential risk of carcinogenesis." In summary, although asbestos is well established as a pulmonary carcinogen, our understanding of the overall magnitude of the problem and the variables involved in determining the risks associated with exposure remains incomplete, and ongoing research is necessary. Others.-Several chemical or physical agents other than asbestos have been associated with an increased incidence of lung cancer in epidemiologic and animal studies. Those agents established by the International Agency for Research on Cancer as having at least limited evidence for carcinogenicity in humans are listed in Table 1. "Limited" evidence was defined as observation of a positive association but inability to exclude chance, bias, or confounding variables with reasonable confidence." Agents categorized as "sufficient" were determined to have an established causal relationship. Of note, classification of an agent as having sufficient proof of carcinogenicity by the International Agency

OCCUPATIONAL LUNG CANCER

185

for Research on Cancer was not restricted solely to data on lung cancers, although some data support such an association for each agent listed. The standard mortality ratios compiled from the literature are an approximation of carcinogenic potential; they represent a wide range of studies with considerable variability in terms of history of exposure to the agent and quality and selection of control populations. As with asbestos, the specific type of the agent and exposure conditions are important in determining the degree of carcinogenicity. For example, only hexavalent chromium and nonmetallic nickel seem to be carcinogenic." In addition to the agents listed in Table 1, other chemicals and products have been suggested as potential lung carcinogens, but supportive data are limited or conflicting. Manmade mineral fibers, for example, which at one time were considered a safe alternative to asbestos, have now also aroused concerns about carcinogenicity. Some data have suggested a small (less than twofold) increased risk of lung cancers for workers in the production industry,26,27 but two recent reviews concluded that current evidence is insufficient to establish or exclude this potential association.F" More than 30 reports of epidemiologic studies have been published on exposure to formaldehyde and the associated risk of cancer; the conclusions have been inconsistent. The data from one large study (26,561 industrial workers) by the National Cancer Institute have been interpreted both in support of and against a causal relationship with lung cancers. 29-32 A recent review of the literature on formaldehyde found the evidence inconclusive, although a causal role for cancers of the nasopharynx and nasal cavities was thought probable." An equally large number of studies have been published on a relationship between exposure to silica and lung cancer. These investigations have been complicated by the presence of confounding exposures to smoking, asbestos, and radon and have been criticized because of bias in selection of cases and choice of control population. Consequently, no clear consensus has emerged on the subject. 34-37

OCCUPATIONAL SURVEYS In many occupations, employees are exposed to a varied milieu of potentially carcinogenic substances. For instance, combustion emissions often contain complex mixtures of thousands of organic and inorganic compounds." In such instances, individual assessment of the role of a single substance in the induction of lung cancer may be impossible. Consequently, the identification of occupational lung cancers at times relies on the determination of the incidence of cancer among various occupational groups without reference to specific exposures. Many published reports have described the incidence of lung cancers for various occupations in a variety of geo-

186

Mayo Clin Proc, February 1993, Vol 68

OCCUPATIONALLUNGCANCER

Table I.-Agents Associated With Occupational Lung Cancer* Agent

SMR

IARC category]

Acrylonitrile

120-150'

Limited

Arsenic

168-1,189'

Sufficient

Asbestos BCMEand CMME Beryllium Cadmium

104_l,400 Hl- 23

Sufficient

278-2,080' 120-180' 133-235'

Sufficient Limited Limited

Chromium

70-390'

Sufficient

Mustard gas Nickel

148-3,667' 60-531',23

Sufficient Sufficient

Radon Vinyl chloride

144_620'·5.24 112-17825

Sufficient Sufficient

Occupational examples Production of plastics, petrochemicals Copper smelting, manufacturing of pesticides Production, construction Production of chemicals Metal processing Smelting, manufacturing of batteries Manufacturing of pigments, leather products, plating Production, warfare Smelting, electrolysis, refining Mining Production of polyvinyl chloride

*BCME = bis(chloromethyl) ether; CMME = chloromethyl methyl ether; IARC International Agency for Research on Cancer; SMR = standard mortality ratio. tSee text for definition of categories.

graphic Iocations.v" In Sweden, hospital physicians must report all newly diagnosed lung and other cancers to the Swedish Cancer Registry, and these data have been linked to the national census, which includes occupational information. In one study, the incidence oflung cancer was assessed for 1.6 million men 30 to 64 years of age during the period 1961 to 1979. 40 Statistically significant, smoking-adjusted excess risks were found among assemblers and machine erectors, drivers, miners, longshoremen, and sheet-metal workers. Fewer than expected cases of lung cancer were noted among farmers, chemical and cellulose workers, and other groups. Although such studies are valuable, they are limited somewhat to factors specific to a certain geographic location. In another large population-based study that used data from the Shanghai Cancer Registry, smoking-adjusted lung cancer risks apparently were increased for workers in agricultural production and the chemical industry, in direct contrast to the Swedish data." Other studies have examined the relationship between histologic type of lung cancer and occupation. Although variations in histologic frequency have been noted, the importance of this finding is unclear, and an exclusive association of a single cell type with exposure to a specific physical or chemical agent has not been demonstrated." One interesting approach used by Rylander" has been to examine employees in particular occupations with a decreased frequency of lung cancer, in an attempt to identify agents or conditions that could be protective. Cotton textile

=

workers have been found to die of respiratory cancers at a significantly lower rate than control populations, a fact that apparently remains unexplained by differences in frequency of smoking. One theory is that long-term exposure to endotoxin, known in vitro to induce alveolar macrophage activation, somehow exerts a protective effect against the development of neoplasms.

IMPORTANCE OF EXPOSURE Because of the controversy associated with the degree of health hazard attributable to asbestos and many of the other known and suspected lung carcinogens, the wide spectrum of opinions about the importance of occupational exposures to the overall incidence of lung cancers should not be surprising. Reported values have ranged from I to 50%.47,48 Such estimates are based on various techniques, ranging from personal conjecture to statistical projections derived from estimates of the cumulative attributable risk from established carcinogens. Within a specific population, the incidence of occupational cancer obviously depends on the degree of industrialization and patterns of smoking. Some evidence suggests that the importance of occupational or at least environmentally related lung cancers is increasing. Among the Japanese female population, for whom the incidence of smoking is only 13% and only 26% of cases of lung cancer are attributed to smoking, the incidence oflung cancer has increased almost 700%, from 0.8 to 6.l/100,000 during the years 1950 to 1985. 49 In the United

Mayo Clio Proe, February 1993, Vol 68

States, the incidence of lung cancer among nonsmokers doubled between 1958 and 1969.50 Whether this trend will continue is unknown.v" Overall, smoking is by far the most important preventable cause of lung cancer. Among patients with lung cancer, approximately 98% of male patients worldwide and 70 to 90% of female patients in Europe and America report a prior history of smoking." Nevertheless, because the interaction of carcinogenic agents seems synergistic in some instances, such as with cigarette smoke and asbestos, the proportion of lung cancers preventable by the removal of a carcinogenic occupational exposure is potentially much greater than the absolute percentage attributed to specific agents." From both a cost-effectiveness and an ethical standpoint, preventing lung cancers through the recognition and elimination of exposure to carcinogens is clearly preferable to focusing on early diagnosis, cure, or palliation. ACKNOWLEDGMENT We are indebted to Dr. J. M. Samet for valuable editorial comments.

OCCUPATIONAL LUNG CANCER

13. 14. 15. 16. 17.

18. 19. 20. 21. 22. 23.

REFERENCES I. Cone JE. Occupational lung cancer. Occup Med 1987 Apr; 2:273-295 2. Morgan RW, Larson SR. Newer occupational lung carcinogens. In: Gee JBL, Morgan WKC, Brooks SM, editors. Occupational Lung Disease. New York: Raven Press, 1984: 69-77 3_ Seaton A. Occupational pulmonary neoplasms. In: Morgan WKC, Seaton A, editors. Occupational Lung Diseases. 2nd ed. Philadelphia: Saunders, 1984: 657-675 4. Coultas DB, Samet JM. Occupational lung cancer. Clin Chest Med 1992 Jun; 13:341-354 5. Axelson O. Cancer risks from exposure to radon progeny in mines and dwellings. Recent Results Cancer Res 1990; 120:146-165 6. Sakabe H. Lung cancer due to exposure to bis(chloromethyl) ether. Ind Health 1973; 11:145-148 7. Figueroa WG, Raszkowski R, Weiss W. Lung cancer in chloromethyl methyl ether workers. N Engl J Med 1973; 288: 1096-1097 8. World Health Organization. Overall evaluations of carcinogenicity: an updating of fARe Monographs volumes 1 to 42. IARC Monogr Eval Carcinog Risks Hum Suppl 1987; 7:1401 9. Rothman KJ. Modem Epidemiology. Boston: Little, Brown and Company, 1986: 77-97 10. Hill AB. The environment and disease: association or causation? Proc R Soc Med 1965; 58:295-300 I I. Doll R. Mortality from lung cancer in asbestos workers. Br J Ind Med 1955; 12:81-86 12. Wagner JC, Sleggs CA, Marchand P. Diffuse pleural mesothelioma and asbestos exposure in the north western Cape Province. Br J Ind Med 1960; 17:260-271

24. 25. 26.

27. 28. 29. 30.

31. 32.

33.

187

Kogan FM, Polzik EV. Prophylaxis of asbestos-induced diseases: studies and perspectives. Am J Ind Med 1990; 18:95-99 Cullen MR. Controversies in asbestos-related lung cancer. Occup Med 1987 Apr; 2:259-272 McDonald Je. Cancer risks due to asbestos and man-made fibres. Recent Results Cancer Res 1990; 120:123-131 Berry G, Newhouse ML, Turok M. Combined effect of asbestos exposure and smoking on mortality from lung cancer in factory workers. Lancet 1972; 2:476-479 Berry G, Newhouse ML, Antonis P. Combined effect of asbestos and smoking on mortality from lung cancer and mesothelioma in factory workers. Br JInd Med 1985; 42: 1218 Mossman BT, Gee JBL. Asbestos-related diseases. N Engl J Med 1989;320:1721-1730 Kern DG, Christiani DC, Kelsey KT, Hu H, Frumkin H, Kreiss K, et aI. Asbestos-related diseases [letter]. N Engl J Med 1991; 324:195-196 Hughes JM, Weill H. Asbestos exposure--quantitative assessment of risk. Am Rev Respir Dis 1986; 133:5-13 Murray R. Asbestos: a chronology of its origins and health effects. Br J Ind Med 1990; 47:361-365 Neuberger M, Kundi M. Individual asbestos exposure: smoking and mortality-a cohort study in the asbestos cement industry. Br J Ind Med 1990; 47:615-620 Report of the International Committee on Nickel Carcinogenesis in Man. Scand J Work Environ Health 1990; 16:1-82 Bowie C, Bowie SHU. Radon and health. Lancet 1991; 337:409-413 Storetvedt Heldaas S, Langard SL, Andersen A. Incidence of cancer among vinyl chloride and polyvinyl chloride workers. Br J Ind Med 1984; 41:25-30 McDonald J. Peer review: mortality of workers exposed to MMMF--current evidence and future research. In: Guthe T, editor. Biological Effects of Man-Made Mineral Fibres. Vol 1. Copenhagen: WHO Europe, 1984: 369-380 Doll R. Symposium on MMMF, Copenhagen, October 1986: overview and conclusions. Ann Occup Hyg 1987; 31:805819 Miettinen OS, Rossiter CEo Man-made mineral fibers and lung cancer: epidemiologic evidence regarding the causal hypothesis. Scand J Work Environ Health 1990; 16:221-231 Blair A, Stewart P, O'Berg M, Gaffey W, Walrath J, Ward J, et aI. Mortality among industrial workers exposed to formaldehyde. JNatlCancerInst 1986;76:1071-1084 Sterling TD, Weinkam JJ. Reanalysis of lung cancer mortality in a National Cancer Institute study on mortality among industrial workers exposed to formaldehyde. J Occup Med 1988; 30:895-901 Blair A, Stewart PA. Comments on the reanalysis of the National Cancer Institute study of workers exposed to formaldehyde. J Occup Med 1989; 31:881 Sterling TD, Weinkam 11. Reanalysis of lung cancer mortality in a National Cancer Institute study of "mortality among industrial workers exposed to formaldehyde": additional discussion. J Occup Med 1989; 31:881-884 Blair A, Saracci R, Stewart PA, Hayes RB, Shy C. Epidemiologic evidence on the relationship between formaldehyde exposure and cancer. Scand J Work Environ Health 1990; 16:381-393

188

34. 35. 36. 37. 38. 39. 40. 41. 42.

Mayo Clin Proc, February 1993,Vol 68

OCCUPATIONAL LUNG CANCER

McDonald Je. Silica, silicosis, and lung cancer [editorial]. Br J Ind Med 1989; 46:289-291 Ng TP, Chan SL, Lee J. Mortality of a cohort of men in a silicosis register: further evidence of an association with lung cancer. Am JInd Med 1990; 17:163-171 Hessel PA, Sluis-Cremer GK, Hnizdo E. Silica exposure, silicosis, and lung cancer: a necropsy study. Br J Ind Med 1990; 47:4-9 Hnizdo E, Sluis-Cremer GK. Silica exposure, silicosis, and lung cancer: a mortality study of South African gold miners. Br J Ind Med 1991; 48:53-60 Mumford JL, He XZ, Chapman RS. Human lung cancer risks due to complex organic mixtures of combustion emissions. Recent Results Cancer Res 1990; 120:181-189 Dave SK, Edling C, Jacobsson P, Axelson O. Occupation, smoking, and lung cancer. Br J Ind Med 1988; 45:790-792 Carstensen JM, Pershagen G, Eklund G. Smoking-adjusted incidence of lung cancer among Swedish men in different occupations. Int J Epidemiol 1988; 17:753-758 Levin LI, Zheng W, Blot WJ, Gao YT, Fraumeni JF Jr. Occupation and lung cancer in Shanghai: a case-control study. Br J Ind Med 1988; 45:450-458 Kjuus H, Skjerven R, Langard S, Lien JT, Aamodt T. A case-referent study of lung cancer, occupational exposures and smoking. 1. Comparison of title-based and exposurebased occupational information. Scand J Work Environ Health 1986; 12:193-202

43.

Sankila RJ, Sakari Karja1ainen E, Oksanen HM, Haku1inen TR, Teppo LHI. Relationship between occupation and lung cancer as analyzed by age and histologic type. Cancer 1990; 65:1651-1656 44. Zahm SH, Brownson RC, Chang JC, Davis JR. Study of lung cancer histologic types, occupation, and smoking in Missouri. Am J Ind Med 1989; 15:565-578 45. Ives JC, Buffler PA, Greenberg SD. Environmental associations and histopathologic patterns of carcinoma of the lung: the challenge and dilemma in epidemiologic studies. Am Rev RespirDis 1983; 128:195-209 46. Rylander R. Environmental exposures with decreased risks for lung cancer? Int J Epidemio1 1990; 19(5uppl 1):S67-S72 47. Samet JM, Lerchen ML. Proportion of lung cancer caused by occupation: a critical review. In: Gee JBL, Morgan WKC, Brooks SM, editors. Occupational Lung Disease. New York: Raven Press, 1984: 55-67 48. Epstein SS, Swartz JB. Fallacies of lifestyle cancer theories. Nature 1981; 289:127-130 49. Koo LC, Ho JH-e. Worldwide epidemiological patterns of lung cancer in nonsmokers. Int J Epidemiol 1990; 19(5uppl 1):S 14-S23 50. Enstrom JE. Rising lung cancer mortality among nonsmokers. J Natl Cancer Inst 1979; 62:755-760 51. Samet JM, Wiggins CL, Humble CG, Pathak DR. Cigarette smoking and lung cancer in New Mexico. Am Rev Respir Dis 1988; 137:1110-1113

End of Symposium on Intrathoracic Neoplasms, Part I.

Part II will appear in the March issue.