Tobacco Documents Online

Environmental Tobacco Smoke and Cancer Jonathan M. Samet, M.D. Professor of Medicine Chief, Pulmonary Division Department of Medicine University of New Mexico Albuquerque, NM 87131 40

grcrruEa_ u7p T"LEB. CSAPTER 4 ~ 52 ~ m O m N ~ N

2:677- 8. Trichopoulos D, Kalandidi A, Sparros L, MacMahon B. Lung cancer and passive smoking. Int J Cancer 1981; 27:1-4. United States Department of Health and Human Services, Public Health Service. The health consequences of smoking. A report of the Surgeon General. Washington, D.C.: U.S. Government Printing Office, 1982. DHHS (PHS) publication no. 82-50179. United States Department of Health and Human Services, Public Health Service. The Health consequences of involuntary smoking. Washington, D.C.: U.S. Government Printing Office, 1986. DHHS (PHS) publication no. (CDC) 87-8398. United States Public Health service. Smoking and health. Report of the Advisory Committee to the Surgeon General. Washington D.C.: U.S. Government Printing Office, 1964. PHS publication no. 1103. Wells AJ. An estimate of adult mortality in the United States from passive smoking. Environ Int 1988; 14:249-65. World Health Organization. IARC monographs an the evaluation of the carcinogenic risk of chemicals to humans: Tobacco smoking, Vol. 38. Lyon, France, World Health Organization, IARC, 1986. Wu AH, Henderson BE, Pike MC, Yu MC. Smoking and other risk factors for lung cancer in women. J Natl Cancer Inst 1985; 4:747- 51. Go 51 Q OD 0 07 N .P N

Most of the case-control and the cohort studies indicate in- creased lung cancer risk in nonsmokers married to smokers, but these studies do not uniformly show increased risk for sources of exposure other than smoking by the spouse (Tables 1 and 2). The first major study on involuntary smoking and lung cancer was reported by Hirayama in 1981 (Table 1). Hirayama conducted a cohort study of 91,540 nonsmoking women in Japan. Mortality in these women was assessed over a 14-year follow-up period. The ratio of the observed to expected numbers of lung cancer deaths increased in a statistically significant pattern with the amount smoked by the husbands. The findings could not be explained by other factors, such as age and occupation of the husband, and were unchanged when the follow-up was extended by several years (Hirayama 1984). After its publication, this article received intensive scrutiny, and correspondence in the British Medical Journal, which had published it, raised concern about various aspects of the study's methods and findings. In his responses to the correspondence, Hirayama satisfactorily answered most of these criticisms, although he could not eliminate the possibility of unreported smoking by women classified as nonsmokers. If self- reported nonsmokers married to smokers were more likely to actually be smokers, than the resulting bias would tend to indi- cate an increased risk from marriage to a smoker. Based on the same population, Hirayama has also reported significantly in- creased risk of lung cancer for nonsmoking married men whose wives smoke (Hirayama 1984). In 1981,'Trichopoulos and coworkers (1981) also reported in- creased lung cancer risk in nonsmoking women married to cigarette smokers (Table 2). These investigators conducted a case-control study in Athens, Greece, that included selected histological types of lung cancer and control subjects ascertained at a hospi- tal for orthopedic disorders. The finding of increased risk was unchanged when the case and control series were enlarged (Trichopoulos et al. 1983). The results of subsequently reported case-control studies have also demonstrated significantly increased risk of lung cancer in nonsmokers exposed to environmental tobacco smoke (Table 2). The findings from the more recent reports based on studies throughout the world greatly strengthen the evidence from the earlier studies. Several of the newer studies included relatively large numbers Dalager of nonsmokers (Garfinkel et al. 19857 Akiba et al. et al. 1986; Lam et al. 1987; Gao et al. 1986; 1987). Furthermore, in most of the newer studies, involuntary smoking reports. was assessed in greater detail than in the earlier The results of two other investigations have also been in- terpreted as showing an increased lung cancer risk associated 44

TABLE 1 Cohort Studies of Involuntary Smoking and Lung Cancer Study Findings Comments 91,540 nonsmoking females, 1966-1981, Japan (Hirayama 1981). 176,139 nonsmoking females, 1960-1972, U.S. (Garfinkel 1981). 8,128 males and females, 1972-1982, Scotland (Gillis et al. 1984). Age-oicupation adjust- ed RR by husbands' Trend statistically significant. All smoking: Nonsmokers Exsmokers - 1.00t - 1.36 histological types of lung cancer. Current smokers < 20/day.- 1.45 _> 20/day - 1.91 Age-adjusted RR by husbands' smoking: Nonsmokers - 1.00t Current smokers < 20/day - 1.27 _> 20/day - 1.10 Age-adjusted RR for exposure to a tobacco smoker in the home: Males - 3.25 Females - 1.00 All histologies. Effect of husbands' smoking not stat- istically signifi- cant. Preliminary, small numbers of cases. *RR = relative risk, as estimated by the ratio of observed to expected number of lung cancer deaths. tReference category, risk arbitrarily set to unity as the reference point for comparison. 524,

l available on the carcinogenicity of active smoking, on the qualitative similarities between environmental tobacco smoke and mainstream smoke, and on the epidemiolgic data on involuntary smoking. The extent of the lung cancer hazard associated with in- voluntary smoking in the United States remains uncertain, however (U.S. DHHS 1986; Weiss 1986). The epidemiological studies provide varying and imprecise measures of the risk (Tables 1 and 2), and exposures to environmental tobacco smoke have not been characterized for large and representative population samples. Thus, any risk assessments for involuntary smoking and lung can- cer are subject to substantial uncertainty. Nevertheless, risk assessment can provide insight -into the magnitude of the lung cancer problem posed by involuntary smoking. Repace and Lowrey (1985) used data on lung cancer mortality in Seventh Day Adventists, a nonsmoking group, to estimate the effect of exposure to environmental tobacco smoke in increasing lung cancer risk. Their analysis lead to an estimate of 4,666 lung cancer deaths per year attributable to environmental tobacco smoke exposure. An appendix to the National Research Council's 1986 report provides estimates of the numbers of attributable lung deaths. For the year 1985, the risk assessment projects ap- proximately 1,000 lung cancer deaths in males and 2,000 to 3,000 lung cancer deaths in females attributable to environmental tobacco smoke. Wells (1988) attributed 3,000 lung cancer cases annually in the U.S. to involuntary smoking. Further epidemiological studies of involuntary smoking and lung cancer are in progress. These studies should refine our un- derstanding of exposure-response relationships for lung cancer and exposure to environmental tobacco smoke. Other investiga- tions are addressing the characteristics and toxicity of en- vironmetal tobacco smoke and patterns of exposure to environmen- tal tobacco smoke. While the results of these new studies will provide needed information for scientific purposes, the available data and the conclusions of the scientific community already provide a compelling rationale for reducing involuntary exposure to environmental tobacco smoke. Involuntary Smoking and Cancer at Sites Other Than the Lung Several reports have suggested that exposure to environmental tobacco smoke may increase risk of cancer at sites other than the lung. One study found that in children, maternal exposure to environmental tobacco smoke during pregnancy was associated with increased risk of brain tumors (Preston-Martin et al. 1982), and in another study paternal but not maternal smoking increased the risk of childhood rhabdomyosarcoma, a cancer of the soft tissues (Grufferman et al. 1982). In adults, involuntary smoking has been linked to a generally increased risk of malignancy (Miller 47

Scientists draw on a wide range of evidence in judging whether an agent, such as environmental tobacco smoke, causes disease. In addition to epidemiological data, the findings of laboratory studies involving in vitro systems and of animal studies involving exposure to the agent are often relevant. Criteria have been developed for guidance in making judgments on the causality of exposure-disease relationships, but these criteria only provide guidelines, not strict rules of evidence (U.S. PHS 1964; Rothman 1986). Interpretation of the evidence on particular exposure-disease relationships often requires review by multidisciplinary panels of scientists who are instructed to reach a consensus, often in the setting of substantial uncer- tainty. For example, the World Health Organization regularly convenes panels of scientists to address the carcinogenicity of environmental agents. For environmental tobacco smoke and lung cancer, the evidence has been considered by scientists convened by the International Agency for Research on Cancer of the World Health Organization, the National Research Council, and the U.S. Surgeon General (Table 3). All three groups concluded that environmental tobacco smoke causes lung cancer among nonsmokers, although the approach used by each group was unique. Consensus among the three groups, in spite of differing methodology, strengthens the determination that involuntary smoking causes lung cancer. The International Agency for Research on Cancer of the World Health organization (1986) reviewed the evidence available through the end of-1984. It reached its conclusion concerning involuntary smoking and lung cancer largely on the basis of biological plausibility. The agency cited the characteristics of sidestream and mainstream smoke, the absorption of tobacco smoke materials during involuntary smoking, and the nature of dose- response relationships for carcinogenesis, which project some risk for any level of exposure. In reaching its conclusion, the National Research Council committee considered the biological plausibility of an associa- tion between environmental tobacco smoke exposure and lung cancer and the supporting epidemiological evidence, available through mid-1986. The committee carefully considered the sources of bias that may have affected the epidemiological studies and concluded that the association documented in the studies could not be at- tributed solely to bias. Based on a pooled analysis of the epidemiological data and adjustment for bias, the report's authors concluded that the best estimate for the excess risk of lung cancer in nonsmokers married to smokers was 25%. The 1986 report of the U.S. Surgeon General also characterized involuntary smoking as a cause of lung cancer in nonsmokers. This conclusion was based on the extensive information already 46

exposures to factors of interest are assessed, often by inter- view. For example, a case-control study of lung cancer and in- voluntary smoking might be conducted by identifying nonsmokers with lung cancer and a suitable control group, and then inter- viewing the subjects concerning the smoking habits of their spouses, other household members, and colleagues at work. Each type of study has advantages and disadvantages, and the results of both types may be distorted by bias. Misclassifica- tion of exposure is of particular concern in studying lung cancer and involuntary smoking. Misclassification of exposure refers to the incorrect categorization of actually exposed subjects as non- exposed and of nonexposed as exposed. When misclassification oc- curs randomly, it tends to bias studies towards showing negative results; if nonrandom, it may exaggerate or reduce the apparent effect of an exposure. With regard to involuntary smoking and lung cancer, two types of misclassification are of concern. Sub- jects classified as nonsmokers may have actually been active smokers and the degree of exposure of nonsmokers to the smoking of others may not be accurately classified. Misclassification of both types is discussed below in relationship to specific studies. The diagnosis of lung cancer is also subject to misclas- sification; a cancer that originated at another primary site and then spread to the lung may be incorrectly diagnosed as a primary cancer of the lung. For example, in the case-control study reported by Garfinkel and colleagues (Garfinkel et al. 1985), 13 percent of cases originally diagnosed as lung cancer were reclas- sified to other sites after histological review. With regard to exposure misclassification in this study, 40 percent of the cases initially classified as nonsmokers on the basis of information in medical charts were found to be smokers on interview. Epidemiological Evidence on Involuntary Smoking and Lung Cancer Evidence concerning involuntary smoking and lung cancer has been sought indirectly in descriptive data on mortality rates and directly with case-control and cohort studies. Time trends of lung cancer mortality across this century in nonsmokers have been examined with the rationale that temporally increasing exposure to environmental tobacco smoke should be paralleled by increasing mortality rates (Enstrom 1979; Garfinkel 1981). These data can only provide indirect evidence on the lung cancer risk associated with involuntary exposure to tobacco smoke. Enstrom (1979) cal- culated lung cancer mortality rates from various nationwide sources for the period 1914-1968 and concluded that a real in- crease had occurred among males after 1935. In contrast, Gar- finkel (1981) did not identify time trends of lung cancer mor- tality in nonsmoking participants in two cohort studies, the Dorn Study of U.S. veterans, 1954 to 1969, and the American Cancer society study, 1960-1972. 43

TABLE 2 (continued) Case-control Studies of Involuntary Smoking and Lung Cancer I Study 199 never smoking female cases and 335 controls (Lam et al. 1987). 246 nonsmoking female cases and 375 controls, Shanghai (Gao et al. 1987). 90 nonsmoking female cases and 163 controls, Japan (Shimizu et al. 1988). 28 smoking female cases and 62 controls, Japan (Inoue and Hirayama 1988). 54 nonsmoking female cases and 93 controls, Tianjin, China (Geng et al. 1988). '°RR - relative risk as estimated by the odds ratio. Findings Overall odds ratio ~ 1.7, significantly increased for marriage to a smoker. Odds ratio - 2.1 for adenocarcinoma. Overall odds ratio - 0.9 for ever living with a smoker. Risk increased with duration of living with a smoking husband. Odds ratio for husbands' smoking was 1.1. No effect of exposure at work. overall odds ratio ~ 2.3 for marriage to a smoker. Overall odds ratio ~ 2.2 for marriage to a smoker. Comments All histologies. No evidence for exposure-response. All histologies, but majority adenocar- cinoma. No effect of childhood exposure. A11 histologies. Increased risk from other household members' smoking. 52 J Risk increased with the number of cigarettes smoked by the husband. All histologies.

TABLE 3 Conclusions of the World Health Organization, National Research Council and U.S. Surgeon General on Involuntary Smoking and Lung Cancer World Health Organization "Knowledge of the nature of sidestream and mainstream smoke, of the materials absorbed during "passive" smoking, and of the quantitative relationships between dose and effect that are commonly observed from exposure to carcinogens, however, leads to the conclusion that passive smoking gives rise to some risk of cancer." - National Research Council "The weight of evidence derived from epidemiologic studies shows an association between ETS exposure of nonsmokers and lung cancer that, taken as a whole, is unlikely to be due to chance or systematic bias. The observed estimate of increased risk is 34%, largely for spouses of smokers compared with spouses of nonsmokers." U.S. Surgeon General "Involuntary smoking can cause lung cancer in nonsmokers." "The absence of a threshold for respiratory carcinogenesis in active smoking, the presence of the same carcinogens in mainstream and sidestream smoke, the demonstrated uptake of tobacco smoke constituents by involuntary smokers, and the demonstration of an increased lung cancer risk in some populations with exposures to ETS leads to the conclusion that involuntary smoking is a cause of lung cancer." :Zc