LUNG CANCER RELATIONSHIPS IN WOMEN Lung cancer death rates for women are presently much lower than the corresponding rates for men. In addition, it has been ob- served that among certain strains of mice exposed to carcinogenic agents, the male animals show a greater tendency to develop lung tumors than do the females (200, XV) although there are strains for which this is apparently not so. The extent of the influence of endocrine factors in the sex variation in the incidence of lung tumors is unknown. As of 1967 in the United States, women accounted for only about one-sixth of the total deaths from lung cancer (289). However, the lung cancer death rate in women has risen by over 400 percent in the past 40 years. From 1950 to 1967 alone, the rate per 100,000 population doubled, increasing from 4.5 to 8.9 (289,290). A number of retrospective studies concerning lung cancer and cigarette smoking among women have found that the difference in the prevalence of lung cancer between males and females is ac- counted for principally by those tumors classified as Kreyberg's Group I (154,311) . These, as was noted above, are the tumors, par- ticularly in males, which show the closest relationship with smok- ing. Haenszel, et al. (123), in a study of 158 women with lung cancer, observed that the sex differential for lung cancer death rates diminishes, but does not fully disappear when only non- smokers are considered, Hammond (118) found that the death rate for lung cancer in nonsmoking males was somewhat higher than for nonsmoking fe- males. However, the difference in male-female rates was much greater when smokers were compared. It appears that a substantial part of the difference in death rates between male smokers and fe- male smokers can be explained mainly by differences in their smok- ing habits. These differences in smoking habits between males and females are of two types. First, overall consumption among females is still significantly lower than that among males. In 1966 (,%?I), 30 per- cent of males reported that they had never smoked while for fe- males the corresponding figure was 59 percent. This study also noted that nearly three times as many males as females reported consuming more than 20 cigarettes per day. Second, it has been shown that women smoke differently than men (303) : They begin smoking later than men (114) and do not smoke cigarettes as close to the end, where proportionally more nicotine and "tar" are in- haled. Women smoke more filter-tip and "low tar and nicotine" cigarettes than men. Furthermore, cigarette smoking still tends to be heavily concentrated among women under the age at which lung cancer is most likely to occur. 251 Finally, analysis of the ratio of male and female lung cancer death rates (283, 284, 285, 286, 287, 288, 289, 2.90) reveals that since 1960 this ratio has shown a steady decline, reflecting the greater relative rise in mortality from lung cancer in the female population. LUNG CANCER, THE URBAN FACTOR, AND AIR POLLUTION A number of studies have been concerned with the relative influ- ences of smoking, urban residence, and air pollution in the etiology of lung cancer. Table 9 lists studies performed in the United States, Great Britain, and Japan which have dealt with this question. Kotin and Falk (149, 250) and more recently the Royal College of Physi- cians (228) have reviewed the literature concerning the influence of atmospheric and environmental factors in the pathogenesis of lung cancer. The studies listed in table 9 show a number of important trends. Lung cancer death rates are found to be higher among urban popu- lations than among rural populations. It is not known to what ex- tent this urban factor in the etiology of lung cancer is due to differences in the levels of air pollution. Other factors associated with urban residence which may influence the etiology of lung cancer are : differences in smoking habits between the two popula- tions, occupational differences, and possible differences in the re- porting of lung cancer deaths (228). The studies also uniformly show that within each urban/rural grouping, lung cancer death rates increase with increased smoking. Whether air pollution acts with cigarette smoking to influence lung cancer death rates in a combined manner is presently unclear (112, 126, 261, 265), and the evidence concerning a separate role of air pollution in the etiology of lung cancer is still inconclusive (228). The recent report of the Royal College of Physicians on air pollu- tion and health (228) concluded that "the study of time trends in the death rates of lung cancer in urban areas demonstrates the overwhelming effect of cigarette smoking on the distribution of the disease. Indeed, only the detailed surveys that have taken individual smoking histories into account have succeeded in separating the relatively very small influence of the `urban factor' on the over- riding effect of cigarette smoking in the development of cancer of the lung." 252 TABLE 9.-Epidemiologic investigations conwxing the wlationsl,ip of hng cancer to smoking, air polhtion, and urban or rural residence (Actual "umber of deaths shown in parentheses) Author, Population Ye*=. studied and country. method of Results reference data collection Comme"ts Doll, Estimated death rates Lung canecr mortality (1950) per 1,000 Authors noted that 1953, from lung cllncer M&.9 FC?Tl&S Nonsmokers estimates are based on England in English London Other urban Rural Londo?~ Other urban Rum4 All area8 very few deaths. (70). population and Age: among nonsmokers 25-44 0.126 0.095 0.070 0.028 0.028 0.012 0.020 obtained from 45-64 1.672 1.264 0.861 0.194 0.162 0.120 0 090 general register. 65-74 3.124 2.00G 1.164 0.440 0.32F 0.288 `219 Stocks and Death rates in Male lung cancer death rates 195161 (prr 100.000) ngctl 51-74 The authors noted the Campbell, England and upward gradient among 1955, Northern Wales. Rural (68) Mixed (118) Urban (5.79) nonsmokers. pipe England Review of patient Nonsmokers __.............._.......,......_ 14 131 (8.55). smokers and light chart or interview Pipe .._.........__... 41 25 143 c.igerette smokers and the with kin or Cigarettes: Light 87 153 297 lack of a similar physicians. Moderate ,,.,,........._....._,.......,... 183 132 287 gradient among Heavy ._....................,.......... 363 303 394 moderate and heavy cigarette smokers. Hammond 187,783 white males Age etandardized death rates dve to bronchogcnic carcinoma (males) - Data excluded and Horn, in 9 states. adcnocarcinoma. when 1958. Questionnaire Suburb City of City of standardized for age and U.S.A. and interview. Rural or town 10,300-50,000 (120). >SO.OOO smoking, rural rate was Nonsmokers . 4.7 (2) 9.3 (3) 14.7 (4) still noted to be 26 Cigarette smokers 66.2 (62) 71.7 (67) 70.9(69) 85.2(83) percent less than urban. .~ TABLE g.-Epidemiologic investigations concerning the relationship of lung Cancer to smoking, air pollution, and wban or rural residence (cont.) (Actual number of deaths shown in parentheses) Author, Population Ye-. studied snd country. method of Results comments reference data collection HW"SZ.4 10 percent of all Age- md emoking-standardized lung cancer mOrt&tU rsaQtii% Standardized Mortality et al.. white male lung (epidermoid and undifferentiated carcinontas mh) Ratio = IOO for U.S. 1962, canwr deaths in U.S.A. white males age 35 and U.S.A. for 1958 Metropolitan counties Nonmetropolitan counties (11.3). over in 1958. The authors for whom next of >50,000 ,119 2.600-50.000 . .QO kin or physicians *ISO noted ". . joint lO,OOO-50.000 .I61 Rural nonfarm .74 eff'Xt.s Of residence and supplied smoking 2.600-10.000 . .99 Farm . . . . . . . . . ...57 smoking histories in the data. 2.191 cases @chedule of hng-cancer with adequate =ateS far greater than information. those expected on the a.%umption of additivity of the separate effects. . :' DOU 41,000 male British Standardized death rates for lung cancer The authors noted that and Hill, physicians. 1964. Questionnaire and rural mortality data England Conurbation(49) Large Towns (34) Small Towns (ad) follow-up of death Rural (IS) Nonsmokers 0.03 were affected by a (74). 0.00 0.11 0.12 certificate. Cigarette smokers: significant number of 1-14 . . . . . . . . . . . . . .._ 0.48 city residents 0.32 0.87 0.52 16-24 .._.......... 1.31 retiring to the country. 1.66 1.06 1.16 >25 . . . . .._......... 1.90 4.43 2.20 1.17 Wicken, 1,908 male and 1966. Lung tamer death rate pe+ lOO,OOO--age- and smoking-standardized femalelu"gcs"eer Total number of deaths Northern deathsover 35 Inner outer Belfast urtmn noted under method of STlUdl Ireland data collection include Years of *ge from Belfaet Belfast Environs AT.X8 TOWM (308). RUTal register. Personal 964 contTob3. Males X7(241) 139(167) 136(46) 118(185) 137(26) 47(149) interviews with Females 22 (38) 17 (24) 12 (6) 23 (35) kin or physicians. 22 (6) 12 (43) TABLE 9.-Epidemiologic in.vestigations concerning the relationship of lung cancer to smoking, air pollution, and urban or rural residence (con&) (Actual number of deaths shown in parentheses) Author. Population Yea=, studied and COUIltrV. method of Results referenee Buell et al.. 1967. U.S.A. (49). Hitosugi, 1968, Japan (126). data collection 304 lung cancer Age-adjusted lung cancer death rates per 100.000 man years and mottalitg ratios The authors noted the lack deaths among of death-rate difference American San Francisco/ AU other between Los Angeles and Legionnaires Los Angeles San Diego California countiee San Francisco regions aged 25 and over. Rate Ratio Rate Ratio Rate Ratio and concluded that Questionnaires to Nonsmokers _,.,..._.__..... 28.1 2.6 43.9 3.9 11.2 1.0 photochemical smog ia next of kin. Smokers : not related to l . . 241.3 21.6 226.0 20.2 137.6 12.3 185 male and Lung caacer death rate per 100,000 The authors postulated a female lung cancer slight synergistic deaths and 4,191 PoUution region effect between smoking matched controls Males LOW Intermediate High and air pollution. aged 36-14. Data Nonsmokers .._.......,.,...,........... . . . . . 11.6 3.8 4.9 from Smokers : questionnaires 1-14 cigarettes/day _. . . 10.6 14.2 23.6 and interviews. >16 ,,,,,.................................... 21.3 18.6 31.4 Fern&a Nonsmokers ,,.,...__.._._._..._............... 4.6 6.9 3.8 Smokers: 1-14 cigarettes/day . 19.7 16.6 16.3 >16 . . . . . . 12.4 20.6 17.1 Age- and smoking-adjusted lung cancer death rate per 100,000 E Males . . ..___._................................. Females . . . . . . . . . . . . . . . . . _.._.._............... LOW Intermediate High 16.1 22.4 28.4 7.6 11.6 8.1 LUNG CANCER AND OCCUPATIONAL HAZARDS Uranium Mining The excess risk for the development of lung cancer among uran- ium and fluorspar miners has been known for more than 30 years. In a recent review, Bair (17) noted that radon and radon-decay products are the only inhaled radionuclides to be epidemiologically related to lung cancer. Lundin, et al. (178), in a continuation of the work initiated by Wagoner, et al. (299, 300, 301)) have re- cently reported on a l7-year follow-up of 3,414 white underground uranium miners. The authors estimated that smoking uranium miners experienced an excess of lung cancer ten times greater than did nonsmoking miners. Saccomanno (231)) in recent testimony, analyzed the data of the United States Public Health Service (USPHS) Study Group as presented by Lundin, et al. (178) above. He reported that cigar- ette smoking uranium miners incurred lung cancer rates four times greater than those of other cigarette smokers. Of the 62 lung cancer deaths in this population, 60 occurred in smokers. He also observed that among 100,000 uranium miners 700 lung cancer deaths per year would be expected to occur among cigarette smokers compared with only 4 among nonsmokers. Other Occupations Nelson (199) has recently reviewed certain environmental and occupational hazards as they relate to inhalation carcinogenesis. He observed that cancer of the respiratory tract has been linked epidemiologically and, in some cases, experimentally with occupa- tional exposure to the following materials: chromium, nickel, arsenic, and asbestos. Doll (72) and Goldblatt (100)) in earlier reviews, also noted an association with coal, natural gas, and graphite exposures. Nickel Morgan (194) noted that much of the nasal and lung cancer at- tributed to nickel exposure may have been due to arsenical impuri- ties found in processed nickel prior to 1925. Doll (69) found that the number of excess deaths among nickel workers under 50 years of age had declined following the change in nickel manufacturing processes. The experiments of Hueper (1.~1) and Sunderman, et al. ( 267,268,26.9) have shown that both guinea pigs and rats develop lung cancer following chronic exposure to nickel carbonyl or nickel dust. Sunderman and Sunderman (270) also reported that ciga- rette smoke contains nickel and that this concentration of nickel 256 may be capable of inhibiting the induction of lung aryl hydroxylase, an enzyme which is able to detoxify aromatic hydrocarbons includ- ing known carcinogens such as benzo[a]pyrene. Asbestos In 1955, Doll (71) found that lung cancer was a definite hazard among asbestos workers, In a more recent study, Selikoff, et al. (Zl,ZnZ) examined the relationship of smoking and asbestos ex- posure to lung cancer. These authors followed 370 people who had been asbestos workers during the years 1942-1962. Over a 5-year follow-up period, 94 deaths occurred in this group, of which 24 were due to bronchogenic carcinoma. The authors noted that according to data obtained from Hammond (118), only 3.16 deaths from lung cancer would have been expected among smokers, and calculated a 7.6 to 1.00 mortality ratio due to asbestos exposure. None of the 87 nonsmokers or pipe and cigar smokers died of lung cancer. When the expected number of nonsmoker deaths (0.26) is compared with the actual number (~4) which occurred among the smoking asbes- tos workers, an extremely high mortality ratio of 92 to 1 is obtained, thus reflecting the possible interaction of asbestos exposure and cigarette smoking. Exposure of mice (179) and rats (106) to asbestos dust or the intratracheal injection of chrysotile asbestos dust has resulted in the production of significant numbers of primary pulmonary car- cinomas. Miller, et al, (181) exposed hamsters to intractracheal injections of benzo[a]pyrene. These authors observed that the addi- tion of the chrysotile variety of asbestos to the injections appeared to promote benzo[a]pyrene carcinogenesis in the respiratory tract, as determined by the time of appearance and yields of papillomas and carcinomas. Auenic A recent epidemiologic study by Lee and Fraumeni (16.3) has indicated an excess of lung cancer deaths among smelter workers exposed to arsenic for more than one year. Cigarette smoking was not taken into account in their computations. Experimental work on the induction of cancer in animals using arsenic has yielded either negative or inconclusive results (133,135). Chromium Exposure to industrial bichromate compounds has been associ- ated with an excess of lung cancer deaths (22,25cJ). Laskin, et al. (159) have recently reported that intrabronchial pellet implanta- 257 tion of various chromium compounds in rats is associated with the development of squamous cell carcinomas and adenocarcinomas. However, Nettesheim, et al. (ZOO) exposed mice to chromium oxide dust and observed that it had no discernible effect on lung tumor incidence. PATHOLOGICALSTUDIES Investigators who have conducted detailed autopsy studies on patients who died of lung cancer have reported the increased pres- ence, when compared to noncancer patients, of bronchial epithelial changes which they considered to be precursors of bronchogenic carcinoma (7, 8, 23, 51,. 101, 208, 220, 279, 309). Such changes include squamous metaplasia, atypical squamous metaplasia (with acanthosis, dyskeratosis, and numerous mitotic figures), and car- cinoma in situ. Carnes (51) noted that carcinoma in situ was pres- ent in 119 cases of lung cancer but not in any of the 119 controls who were matched for age, sex, and race. Autopsy studies comparing the frequency of these cancer- related changes in the lungs of smokers and nonsmokers are pre- sented in table 10. Virtually all the studies noted an increased prevalence of these epithelial alterations among smokers as com- pared with nonsmokers. Definite dosage-dependent relationships were evident in the results of many of the reports. Also, Auerbach, et al. (14) observed that the number of cells with atypical nuclei decreases progressively in the bronchial mucosa of ex-cigarette smokers, depending upon the number of years between cessation of smoking and death, although it usually remains above that found in nonsmokers. The cytologic studies included in this table (182, 198, 222) all noted an increased percentage of sputum specimens showing meta- plasia among smokers as compared with nonsmokers. PULMONARYCARCINOGENESIS General Aspects of Carcinogenesis Agents found in cigarette smoke which have been identified as, or are suspected of being carcinogenic, are listed in table 11. The list includes certain compounds which most probably contribute to the pathogenesis of the various cancers discussed in the other sec- tions of this chapter. Many other agents have been identified in tobacco and tobacco smoke. At the present time, they do not appear to bear a direct relationship to carcinogenesis. Stedman (262) and Wynder and Hoffmann (319) provide detailed listings and discus- sions concerning these materials. 258 TABLE lO.-Pathologic and cytologic findings in the trachea-bronchial tree of smokers and nonsmokers (Actual number of cases show" in parentheses) Author. Y==, countrY, reference NuIu""`,";f method of selection R.?sllIts Comments Chang. 1967, U.S.A. and K0rell (55). 105 males and females 40-86 years of age. Percent of cases with brmachid basal cell hyperactivity Nonsmokers . . . . . . . . . . , . . . 23.6 (34) smokers . . . . . . . . . . . . . 43.7 (71) Heavy smokers . . . . . . . . . . . . . . . , . . . . . . . . . t61.3 (31) Smokers included pipe snd cigar smokers. t psO.01 in com- parison with nonsmokers. Hamilton et al.. 1957. U.S.A. (117). Selected autopsy material. Number Age range Smokers . . . . . . . . . . . . . . . . 16 39-71 Nonsmokers 20 26-83 Percent of cues with: Bad cell SQuamoua hyperplasia metaplaaia 86.6 20.0 40.0 15.0 Tsmaitional metaplaaia 40.0 36.0 No lung cancer patienta included. Sanderud, 100 males Percent of case6 with bronchial ag~arno~n epithelial netaplaia Nonsmokers in- 1956. autopsied at Nonsmokers . . . . . . . . . . . . . . . . . 64.0 (39) elude those Norway Gade Institute Pipe . . . . . . . . . . . . . . . . . . . . . . . . .._.........._.................. 80.6 (20) smoking less (240). on whom All cigarette . . . . . . . . . . . , . . . . . . . . . 79.0 (38) than or equal to smoking data Cigarettes per day: 6 grams per day. was available. E&14 . . . , . . . 70.0 (23) 16-26 . . . . . . . . . . , . . . . . 90.0 (10) >ZS . . . . . . . . . . . . . . . . , . . . . , . . . . . . , . . . 100.0 (6) Knudtaon, 100 persons Percent of cases with: Atupical Age, occupation, 1960, 23-86 years No. of NO Basal cell SQWtnOU8 proliferative and site of U.S.A. of age P+%WWd change hgperplasia metapIa.& ?"dWAlsia residence were (1.47). autopsied at Nonsmokers . . . . . . (21) 41.6 23.6 14.3 9.6 found to have no Seattle Cigarettes/day: appreciable Veterans 1-9 . . , . . . . . . . (9) 17.8 11.1 11.1 . . effect. Hospital on 10-16 . . . . . . . . . . . . . . . . . (11) 18.2 18.2 64.6 9.1 whom 16-20 . . .._........._.. (44) 20.4 29.6 29.6 29.6 smoking >21 . . . . . . . . . , . (9) 11.1 33.3 44.4 11.1 it data was Pipe or cigar , . , . , . (6) . . 100.0 . . available. rg TABLE lO.-Pathologic and cytologic findings in the trachea-bronchial tree of smokers and nonsmokevs (cont.) (Actual "umber of cases shown in parentheses) Author. year, country, refrrenee Results COlllllle"tS Auerbach 339 persons et al.. 22-m years 1961, of age U.S.A. autopsied at (f2). East Orange Veterans Hospital (excludes lung Wl"C.X) Nonsmokers: <40 years of age 40-59 . 60-69 . >70 . . . . Smokers 70 . . Smokers >l pack/day: <40 years of age 4rk69 6lL69 . . . . . . >70 . ........ ........ ........ ........ ........ ........ ........ ........ . . 8 11 28 18 14 . . . 24 36 22 17 ._.. 63 . . . . 84 . . . . 15 Number of sections of bronchial epithelium 333 560 1,463 918 727 1,240 1.772 1,101 880 3,027 4,136 756 Percent sectimla Percent sections with cilia absent with bone and sntirel~ atypical cells atypical cell8 and cilia absent 0.3 0.1 0.6 0.1 4.7 1.0 16.9 0.5 10.8 0.6 9.4 1.5 12.5 4.5 17.4 6.9 20.5 9.8 23.7 The suthors noted a dose-response re- lation of smoking to: a. loss of cilia, b. increase in number of atypical cells, c. carcinoma in situ. Average "umber of sections per case equaled 62.3. Cross et al.. 1961, U.S.A. (64). 140 persons Percent aectiom showing changes in bronchial epithcltum (number of sectiona) t The authors noted autopsied at Squamow Atypicd Carcinoma that the differ- Iowa City NW??tld Hyperplmia metapimia ?fWtUplLISG in situ Carcinoma ence between Veterans Nonsmokers (31) . . 61(562) 36(137) 8 (33) t15 (68) . . . . smokers and "on- Hospital Smokers ( 109 ) . . . . . . . . . . . . . 44(570) 43(662) 16(197) 20(263) l(12) 2.6(34) smokers was on whom statistically smoking significant. data was available. TABLE lO.-Pathologic and cytologic findings in the trachea-bronchial tree of smokers and nonsmokers (cont.) (Actual "umber of cases show" in parentheses) Author. ye.8=, c0u*t.rY, reference %:k;oaf method of selection Results comments Auerbach et al., 1962. U.S.A. (14). 72 autopsied former ciga- rette smokers who had been smoking for 210 yc*lg and had ceased 25 years ago. Number of Percent sections Percent s&ions Percent sectiona Each a-smoker sectimu of with cilia absent with mmc atypi- with 50 percent matched with a brachial and entirely cd celk and atypical eel& current smoker Number epithelium atypical cells cilia absent and cilia preeent plus "ever-smoker Nonsmokers 72 3,166 0.0 0.1 0.6 for age. OCC"PB- Rx-smokers . 12 3,436 0.2 0.9 2.5 tion. and resi- Current smokers . 72 3,537 8.0 19.0 80.8 dence. There was an average of 50.3 sections per subject and none had less than 18 sections. TABLE IO.--Pathologic and cytologic findings in the trachea-bronchial tree of smokers and nonsmokers (cont.) (Actual number of cases shown in parentheses) Author, Number of Year, cases and country, method of ResU1t.S COllUIl.Xlts reference selection Auerbach 456 male and Percent *ec- Percent sec- Percent *ec- Major findings et al.. 302 female Number of tima with timw with tiom with 50 noted: 1962. smokers and wctiona of cilia absent wme atypi- percmt atypical Urban nonsmokers U.S.A. nonsmokers bronchial and entirely cd cells and ceUs and showed more (13). autopsied and Number epithelium atypical cell.9 cilia absent cilia present lesion than rural. matched for Males : Both lesions and rage. oceu- Nonsmokers . . . . . 41 2,346 0.1 0.7 atypical nuclei pation. and Cigarette smokers . . . . . . 75 3,393 6.9 21.2 78.6 were much less residence. Females: . . frequent in non- Nonsmokers . . . . . . . . 47 2,379 . . 0.1 0.5 smokers and less Cigarette smokers . 76 3,607 2.5 13.3 62.6 frequent in pipe Males : and cigar smokers Nonsmokers . . . 36 1.106 0.3 0.2 0.6 than in cigarette Cigar smokers . . . . . . . 36 1,733 10.0 10.7 smokers. Cigarette smokers . . . . . . 35 1.626 12.8 27.3 83.1 57.1% of cases bad 60-55 sections 31.5% of eases had 40-49 sections 7.3% of cases had 30-b9 sections 4.6% of cases had 16-29 sections Robbins, 1966, U.S.A. (2.22). 103 students Percent in each cytologic clads Smokers defined as 17-24 years Slight& Moderat& strmg1y those having eon- of age who Normal atypicd &p&l atypical sumd 210 ciga- underwent Nonsmokers (45) . . . . . . . . 86.7 4.4 8.9 . rettes a day for WXOSOl Smokers (58) . . . . . 55.2 32.8 10.3 1.7 21 year. sputum induction. TARI E lO.---Pathologic md cytologic findings in the trachea-bronchial tree of smokers and nonsmokers (cont.) (Actual number of crises shown in parentheses) Author. Number of year. cases and country, method of reference selection Results Maltoni 1,000 healthy Number Percent ahowinp mctaplasia et al., males who Nonsmokers . . 294 41.16 1968, underwent Smokers: Italy sputum l-10 cigarettes/day _. . ". 189 47.09 (182). induction. 11-20 . 385 51.43 21-30 . . 93 61.29 >30 . . 39 69.23 -~~- __ Nasiell, 50 nonsmoking Sputum crtologic changes Percent with t Regarded by 1968. outpatients, Percent Percent with utupical author 89 "real SWdCn 39X smokers Number M&S Mean age metaplasia metap1naiat premalignant (198). wrtmpatmg Nonsmokers 50 42 57.1 18 4 change." in general Smokers _. 393 73 46.6 62 27 health exam- ination who underwent SPUtUm induction. Spain et nl., 1970, U.S.A. (258). 15'7 males and TX females nutopsied fol- lowing sudden or accidentnl death fol whom smok- ina data were available (ex- smokers ex- cluded from female data). Males : Nonsmokers . Ex-smokers l pack Females: Nonsmokers l pack - ......... ......... ......... ......... ......... ........ ....... . . ,... ........ ........ ........ Number 36 . 21 32 6.8 34 . 18 . 26 Percent with metaplasia The authors found no evidence af 50.0 carcinomn in situ 67.7 or pwneoplastic 62.5 atypicnl changes. 73.5 34.1 33.3 46.1 In order to facilitate understanding of the relationships of the various compounds to one another, the third column presents the presently underst.ood relative importance of each of the various groups of compounds. These compounds have been tested only in animals or tissue cultures, and it should be stressed that the rela- tive importance of one compound may not be the same in man as it is in animals. Table 11 is divided into two major sections. The first section details those compounds which are considered to be or are suspected of being cancer initiators. These are compounds which induce irreversible changes in responsive cells. In the second section are listed those compounds which are considered to be or are suspected of being tumor promoters. These compounds promote the malrg- nant reproduction of cells in which neoplastic changes have been initiated. A number of these initiators may also act as complete carcinogens in their own right. The evidence concerning the two stage initiation-promotion mechanism is still rather limited for respiratory tract carcinogenesis. The pol~nuclear aromatic hydrocarbons (PAH) listed are pres- ently considered to play a very significant role in pulmonary car- cinogenesis due to tobacco smoking. These compounds act as tumor initiators or complete carcinogens. The particular role of these agents in environmental and occupational carcinogenesis has been reviewed by Falk, et al. (93). That such hydrocarbons are pro- duced from tobacco during human smoking has been shown by Kiryu and Kuratsune (2-N). These authors reported the presence of benz[ a] anthracene, chrysene, benzo[ alpyrene, and benzo- [blfluoranthene in the "tar" produced by normal smoking and measured in either filters or stubs. Two hydrocarbons which have frequently appeared in the litera- ture on experimental tobacco carcinogenesis may not actually be present in tobacco smoke. They have been used as representatives of carcinogenic PAH, a class which includes many constituents that have been ident.ified in cigarette smoke condensate. They are `Y,?C!-dimethy!benz[a]anthracene and 3-methylcholanthrene and have been frequently used as tumor initiators or complete carcino- gens, particularly in skin painting and tracheal implantation experiments. The nitrosumine compounds listed are potent carcinogens affect- ing many organ systems, including the respiratory tract (188, 199). Nagee and Barnes (181) have presented a detailed account of experiments in this area. Xitrosamines have been identified in trace amounts in tobacco "tar" and the conditions required for their formation (the presence of secondary amines and nitric oxide) are 264 TABLE Il.-Identified or suspected tumorigenic agents in cigurette smoke' Estimated concentrs- tion in 100 cigarettes (85 mm. "onfilter) Presently understood relative importance in experimental tobacco carcinogen&s I. Complete carcinogens and tumor initiators: Polynuclear aromatic hydrocarbons ........ 1. Benzo (a) pyrene .................. 2. Dibe"z(a,hlanthracene .......... 3. Be"zo(b)fluora"thene ............. 4,Benzo(j)fluoranthene .............. 6. Dibenzo(a,iJ pyrene ............... 6. Benz(a)anthracene .............. 7. Cbrysene .......................... 8. Indeno(l,2,3-cd),,yrene ........... 9. Be"zo(ciphenanthrene' ............. 10. MethyIbenzo(a)pyre"es .......... 11. Methylchrysenes .................... N-heterocyclic hydrocarbons .......... l-2 1. Dibenz(a,h)acridine ............ 0.01 2,Dibenz(a,j)acridine ................ 1.0 3. `IH-dibenzo (c&z) carbazole ........... 0.07 N-nitrosamines" l-10 1. Dimethylnitrosamine ............. 0.4 2. Diethyinitrosamine ............... TKXe 3. Methyl-n-butylnitrosamine .......... TlYWe 4. Nitrosopyrrolidine .................. 0.4 5. Nitrosopiperidine .................. TP3C-Z Epoxides, peroxy compounds. and la&ones: 1. Epoxides ........................... 2. Peroxides ................... 3. LarAones ........................... a. a-Leva"re"olide ................ b. a-Levantenolide ................ No data PRSe"t 20.0 2.0 N-alkyl-heterocyclics: 1. I-methylindoie ..................... Pesticides and fungicides :< 1. TDE ............................. 2.o,l;-DDD ....................... 3. DDT ...................... 4. Maleic hydrazide ................. 10-100 10.?OO IO-IOF 10-100 2-3 1.5" Beta-naphthyiamine ..,.,,,,.. Polonium 210 10-30 "I3 3.9 0.4 0.3 0.6 TlYSe 0.3 2.0 0.5 TCDX 0.1 2.0 Tumor initiators Tumor initiators. suspected carcinogens of possible importance (presence in fresh smoke possible). Certain of these compounds 8~ know" carcinogens; presence in smoke condensate not established. Possib!e initiator. No essentiai contribution suspected. Suspected b!adder carcinogen: of doubtf"i significance at reported ievels. Of some importance on!y in the -;irocuries case of relatively high conten- tration. but not important at reported levels. Nickelcompounds . Present Suspected carcinogens of some importance. 265 TABLE Il.-Identified or suspected tumorigenic agents in cigarette smoke' (cont.) Estimated concentra- tion in 100 cigarettes (85 mm. nonfilter) Presently understood relative importance in experimental tobacco carcinogenesis II. Tumor promoting agents: Neutral promoters (polymers) (unknown structures.) No data Of possible importance. Volatile phenols 20-30 mg. Of possible importance. 1. Phenol 2. Cresol Nonvolatile fatty acids 20.100 mg. Of minor importance. 1. Stearic acid 2. Oleic acid N-alkyl heterocyelics: Of possible importance. 1. I)-methylcarbazole . . . . . . . . . . . Present 1 Modified and expanded from (319. 320) with reference to (52, 60, 89. Ill. 149. 402. 261. 299,494.295). 2 Has not been tested as an initiator, but is a known complete carcinogen. 3 See Neurath. (9X), `See (112,128). found in tobacco smoke (38). However, nitrosamines may be arti- facts dependent on the method of smoke collection (201). Neurath (202) considers the nitrosamines listed in table 11 as being present in fresh cigarette smoke (253, 254). However, con- clusive confirmation of their presence in fresh smoke is not available (38,138,15.5,319). Certain of the pesticides and fungicides presently in use on tobacco have been found to be carcinogenic (91,273,280). A num- ber of these, such as DDT, are now being phased out of regular domestic use. The compounds listed have been shown to be present in trace amounts in mainstream tobacco smoke (I 11,128). A recent, extensive review by Guthrie (2 12 ) provides more detailed informa- tion concerning these agents. Radioactive isotopes can be found in tobacco and tobacco smoke (105). Potassium-40, while present in tobacco leaf, is not trans- mitted in any substantial amount to mainstream smoke (230). Polonium-210 (PO,,,,), however, is transmitted into the mainstream smoke (94, 123, 142, 145, 215, 217). A number of autopsy studies (table A12) have shown that the bronchial epithelium of smokers contains significantly more Po~,~ than that of nonsmokers. LitBe, et al. (172, 173, 17s) have also noted that the concentration of polonium was markedly higher at sites of bronchial bifurcation. These authors stress the importance of this finding for pulmonary carcinogenesis by noting that bronchogenic carcinomas are fre- 266 quently located at bifurcations and that the polonium levels which they found in those regions probably have biologic significance (~16). Other investigators (11.1, 217') have not observed this excess at bifurcations, and in a recent discussion Wynder and Hoff- mann (8.20) concluded that it appears unlikely that PO?,,, in the amounts present in cigarette smoke plays a role in tobacco car- cinogenesis. Although not listed as a separate group, there are a number of agents in cigarette smoke which are potent inhibitors of ciliary movement. Their importance in cnrcinogenesis derives from the increased amount of time which they afford the kno1v-n carcinogens to be present on the surface of the bronchial epithelium. These inhibitors include volatile aldehydes, hydrogen cyanide, nitrogen oxides, volatile phenols, and certain volatile acids such as formic and acetic (1.29). In some respects, the animal and tissue culture studies detailed below apply to neoplastic transformations, not only in the lung but in ot,her tissues in which tobacco smoke, particularly cigarette smoke, is believed to play a role. These general experiments will be presented here, however, with the experiments which bear on lung tissue directly. Skin Painting a,nd Shxtaneous Injection Numerous animal studies on rats, mice, and rabbits, have been performed utilizing known carcinogens, whole tobacco "tar," and various tobacco condensate subfractions, or compounds known to be present in tobacco smoke. These experiments involve the single or repeated painting of shaved or unshaved animal skin. A selected number of these studies js presented in table A13. Numerous other studies, performed prior to and following 19.5:3, are reviewed by lvynder and Hoffmann (319). The skin painting method is still considered to be a valid pro- cedure for the identification of agents suspected of participating in Pulmonary carcinogenesis, as well as for the quantification of the reduction in tumorgenicity of specific agents, The exposure of tissue and organ cultures to cigarette smoke, its condensates, or its conhtituent compounds has been shown to sig- nificantly alter patterns of cell growth and reproduction. Table A14 Presents an outline of these experiments. Once again, less severe effects have been noted lvhen filtered smoke was used (165). 267 Tracheobronchial Implantation and Instillation. More complex experiments concerning the carcinogenicity of cigarette and tobacco smoke are represented by those which involve the direct implantation, instillation, or fixation of suspected ma- terials into the tracheobronchial tree of animals. Certain of these experiments are outlined in table AX. Recent reviews by Saffiotti (233,234) Laskin, et al. (159)) and Montesano, et al. (189) as well as that by Wynder and Hoffmann (319) provide more detailed and extensive accounts of these experiments. Of note among the results outlined in this table are the following : The enhanced carcinogenicity found when benzo[a]pyrene (B[a]P) is combined with a carrier such as hematite dust (235)) and the definite increase in bronchial epithelial preneoplastic and neo- plastic changes among dogs treated with smoke condensate as com- pared with those undergoing only physical bronchial stimulation (224). Inhalation Various species, including mice, rats, hamsters, and dogs, have been exposed to cigarette smoke or aerosols of its constituents. These inhalation experiments are outlined in table A16. It must be noted that the majority of the studies listed involve the passive inhalation of the material presented usually in a chamber. Active inhalation experiments, exemplified by the work of Rockey and Speer (223) and Auerbach and his colleagues (11, 129) involved animals which were trained to inhale voluntarily, thus more closely simulating human smoking. Results of note among these experiments include the following: Miihlbock (195) observed that cigarette smoke inhalation en- hances the already substantial rate of spontaneous alveolar cell carcinoma format.ion in hybrid mice, and various investigators in- duced adenomas in experimental animals (108, 168, 206). Harris and Negroni (I 21) found that exposure to cigarette smoke achieved some enhancement of adenocarcinoma formation in mice but did not observe proven squamous cell carcinoma. Some of their mice had also been exposed to Swine influenza virus aerosol. In a related study, Boren (32) exposed hamsters to cigarette smoke at set inter- vals over a 4%hour period. The author observed alterations in pul- monary cell kinetics (the pattern of DNA synthesis) as demon- strated by H"-thymidine autoradiography. The pattern of the label- ing response to cigarette smoke was significantly different from that of the response to high oxygen concentrations. Auerbach, et al. (22) have reported the development of early 268 invasive squamous cell bronchogenic carcinoma in dogs following a period of direct inhalation of cigarette smoke. These investiga- tors trained beagle dogs to inhale cigarette smoke through a tracheostoma (50) and divided the animals into groups according to dosage as detailed in table 17. A number of dogs died during the course of the experiment which ran for 875 days, or approximately 29 months. The causes of death are listed in table 18. All of the remaining dogs, with the exception of group "h" (high exposure, heavy weight), were sacrificed shortly after day 875 ; the survivors among the heavier dogs are continuing to smoke. Examination of the respiratory tree of the animals revealed a number of tumors (table 19). Most of these were similar to the type of tumor which in man is referred to as bronchiole-alveolar. This tumor arises in the bronchiolar and alveolar epithelium and tends to be multicentric. Two striking characteristics of these bronchiolo- alveolar tumors were the existence of a histologic spectrum (from a tumor resembling the benign condition of adenosis to frankly malignant tumors with invasion of the pleura and surrounding parenchyma) and the marked tendency to squamous change. Inva- sive bronchiole-alveolar tumors were found in 12 dogs in the group which had been exposed to the largest dosage of cigarette smoke. Several had tumors of more than one category. Ten of these dogs had invasive bronchiole-alveolar tumors which did not extend into the pleura, one dog had an invasive bronchiole-alveolar tumor which extended to the pleura, and four had invasive bronchiolo- alveolar tumors extending into the pleura beyond the pleural- pulmonary junctions. In addition, two bronchogenic squamous cell carcinomas were found in this group (table 19). The dosage de- pendence of tumor formation is shown in figures 2 and 3. Major findings of the study were twofold. First, that smoking filter-tip cigarettes was less harmful, both in terms of pulmonary parenchymal damage and lung tumors, than smoking identical cigarettes without filters. This supports the generally held view that total particulate matter is a meaningful indicator of the car- cinogenic potential of a cigarette, Second, lung cancer of two types found in man was produced by the inhalation of cigarette smoke. TWO of the dogs were found to have early invasive squamous cell carcinoma of the bronchus, and both belonged to the high-dosage group. These carcinomas were indistinguishable from early invasive squamous cell carcinomas found in the bronchial tubes of human beings who smoke cigarettes. The majority of tumors found in the dogs were of a bronchiole-alveolar type, which although not as common as squamous cell cancer in man, is not rare in humans. This type is often included in the category of adenocarcinoma. A number of studies have shown an excess of these tumors among 269 TABLE IT.-Datu on pedigreed m& beagle dogs of groups F, L, H, h, and N (Some of the figures apply only to dogs surviving 8'76 days or longer) Filter **O"p F NO filter gri?p NO filter grour, H fE* Nonsmokers group grO"P h N Number of doss on day No. 67' Weight at start (day No. 1) mean weight (pounds) Cigar&e s per dog II, 875 days _. t.. Mcarr numlwr of c~ynrettes per day _. _. Eauivalertt number of cigarettes per day for 150 pound man Type of cigarettes:' ___- 12 12 24 38 R 25.0 26.1 26.0 31.9 30.7 6,143 3,103 6,129 6.129 none 7.02 3.64 7.0 7.0 - 42.1 21.2 42.0 32.9 Milligrams of tar per cigarette . Milligrams of nicotine per cigarette Total dosage in 875 days: Grams of tar per dog _. Grams of niecrtinp per dog Dosage in Xi5 days relative to starting weight: Grams tar/pounds weight Grams nicotine/pounds weight _. _, 17.8 1.17 109.3 7.19 4.31 4.12 8.31 0.29 0.22 0.44 34.8 34.8 34.8 - 1.85 1.85 1.85 103.5 207.8 6.56 11.12 207.8 - 11.12 - 6.61 - 0.35 - -_I ____ ' The smoking dogs were divided into CLOUDS F. L. H. and h on dav No. 67. ? Dogs of groups L, H. and h smoked filter-tip cigarettes during a training period at the start of ,the experiment, but smoked nonfilter cigarettes thereafter. SOI'H~~~S: Adapted from Hammond, E. C. et al. (119). TABLE I%-Swmnary of principal cause of death (days No. 57 through No. 875) in dogs of groups F, L, H, h, atztl N (Each death classified according to most severe condition-some dogs mbinnti