Document 2RY6xd360nRqJ9OY2GNoKoxwp

FILE NAME: Smoking (SMOK) DATE: 1958 DOC#: SMOK024 DOCUMENT DESCRIPTION: Book Excerpt - Carcinoma of the Lung - Asbestos lall allace Dukes ;usson 1 dthers hyroid, NEOPLASTIC DISEASE AT VARIOUS SITES - ^ f f General Editor 'MQ% D? W/SMITHERS, M.D., F.R.C.P., F.F.R. VOLUME I CARCINOMA OF THE LUNG Edited by J. R. ^IGNALL, M.D., M.R.C.P. Assistant Physician, Brompton Hospital; Senior Lecturer, Institute of Diseases of the Chest, Brompton Hospital; Research Assistant, Radio therapy Department, Institute of Cancer Research, Royal Cancer Hospital E. & S. LIVINGSTONE LTD. EDINBURGH AND LONDON 1958 CARCINOMA OF THE LUNG A sbestos The first case of lung cancer to be described in association with asbestosis^ was reported by Lynch and Smith (1935) in the U.S.A. Since then over 60 casesMli have been reported from Britain, Canada, Germany and the U.S.A.--the' majority from Britain. That the association was not just coincidental was indicated by comparison of the incidence o f lung cancer at autopsy in cases of asbestosis and cases of silicosis. Merewether (1949) found that lung cancer was reported in 13*2 per cent o f subjects with asbestosis (31 out o f 235) and in.! 1*3 per cent o f subjects with silicosis (91 out o f 6,884) and Gloyne (1951), on j | personal examination, found corresponding incidences o f 14*1 and 6'9 per cent) (17 out o f 121 and 55 out o f 796). It should, moreover, be noted that wome: formed an appreciable proportion of the subjects with asbestosis and that thej incidence among men with the disease was even higher-- 17-2 per cent in* Merewether's series and 19-6 per cent in Gloyne's. Since all--or almost allsilicotics are likely to have been men, the contrast between subjects o f the| same sex is likely to have been somewhat greater than that actually report for both sexes together. These results are most readily explained if asbestosis predisposes to the development of lung cancer and this explanation was subse-Jt quently borne out by D oll's (1955) study o f the mortality recorded among# asbestos workers. He studied 113 men who had worked for 20 or more years' in places where they were liable to be exposed to asbestos dust and found thal between 1922 and 1953 39 died, whereas on the basis o f the mortality rate: suffered by men of the same ages in the whole population only 154 deaths] would have been expected. Of the 39 deaths 11 were due to lung cancer, whereas less than 1 (0-8) would have been expected. All the cases o f lun; cancer were confirmed histologically and all were associated with asbestosis Precautions to prevent the dissemination o f dust greatly reduced the amount to] which the men were exposed from 1933 onwards and the incidence o f the disease has become progressively less as the number o f years during which the men] were exposed to the pre-1933 conditions has diminished. It is evident, thereforeJ| that lung cancer is a specific industrial hazard o f asbestos workers, but tha with scrupulous attention to the control of dust the risk may be greatly red u ce^ and may, perhaps, be eliminated. The fibrils of asbestos consist o f giant molecules o f polymerised silicooxygen tetrahedra arranged in chains or bands. In the case o f CanadiaqB asbestos, the silica is present as hydrated magnesium silicate containing 6 p e n cent of iron oxide; in other, industrially less important, types the silicates m am be calcium and magnesium or sodium and iron. Animal experiments have, inf general, failed to show any significant carcinogenic activity from asbestos fibred and the mode o f production of the disease is uncertain. In Hueper's vieW (1957a) the activity may result from the occurrence o f the material in the fbrnd o f a linear polymer and the mechanism by which cancer is produced may similar to that with various polymerised carbon compounds, which have beei shown to be carcinogenic experimentally. 50 SPECII The combustion and dist carcinogens and have, for m. cancer of the skin among m with the material. It would, the fumes and dust from tar ' in fact, such a risk is shown 1 The Japanese observati Kawahata 1936; Kuroda 19.' cancer were recognised amoi Yawata Steel Works betweer doubt that this indicates an logical data are not available in Japan and it is striking thi cancer were observed amon observed among the other ' accounted for 80 per cent of appear to have constituted at 500 out of 20,000) and the n dissimilar from that among 49 out o f 19,000). The men they stirred the coal, yellow-b In Britain a high mortal was noted by Kennaway and tions given on the death cert and 1938. Of the 56 occupat might be supposed to have estimated mortality from lu male population and 4 o f the mortalities (Table VIII). Sul Occupations with t, (56 sel Occupatio! Labourers, patent fuel Gas stokers and coke < Gas producer men Metal grinders Gas works foremen a * Extracted from a Ta CHAPTER VI TH E SM O K ING OF TO BACCO R ichard D oll THE evidence linking tobacco with the production of lung cancer is derived from two principal sources. The great mass of evidence has been obtained from the retrospective study o f patients with lung cancer and the comparison of their past smoking habits with the past habits of patients with other diseases, or with the habits o f healthy persons. More recently, evidence of another type has been obtained by comparing, over a period o f years, the mortality in groups of persons whose smoking habits had previously been defined. R etrospective Studies The first serious study o f the subject was reported by Muller in Germany in 1939. He obtained the smoking histories o f 86 men with bronchial carcinoma from hospital notes, by personal interview or from a questionnaire sent to the relatives of those who had died and he compared the histories with those given ! by 86 healthy men o f the same ages. The results showed gross differences ,j between the groups in the proportion o f non-smokers and of heavy smokers. By 1957, 20 studies o f this general type had been reported from 8 countries.! The principal results are summarised in Table X. All agree in showing that there are more heavy smokers and fewer non-smokers among patients with lung cancer than among patients with other diseases, and, with one exception (the difference between the proportions of non-smokers found by McConnell, ) Gordon and Jones 1952), the differences are large enough to merit serious) attention. More detailed results of two o f the investigations are shown in) Tables XI and XII. From these it is seen (1) that there is a steady increase in the relative proportions o f lung cancer to control patients as the amount smoked daily increases, and (2) that the difference in smoking habits holds for both sexes, but is more marked for men than for women. The method o f smoking was not studied separately in all the investigations, r When it was the results agreed in showing that the difference between the)1 lung cancer and the control groups was principally due to a difference in the:' proportion o f cigarette smokers. The results with regard to pipe smoking and, cigar smoking were inconstant. In some investigations it appeared that the. disease was equally associated with all types of smoking (McConnell, Gordon and Jones 1952; Randig 1954); in the majority the disease was most closely associated with the smoking of cigarettes. Few o f the investigators made specific inquiries about inhaling. In one study it was found that the proportion o f patients who said they inhaled was similar among the lung cancer patients 60 THE SMOKING OF TOBACCO T able X principal Characteristics o f the Smoking Histories o f Men with and without Lung i Cancer, Reported by Various Authors SACCO on o f lung cancer is derived ' evidence has been obtained cancer and the comparison 'patients with other diseases, ;cently, evidence of another riod of years, the mortality reviously been defined. Author Date Country Muller 1939 Germany <ve Schairer and Schoniger 1943 Germany Wassink 1948 Holland Shrek et al. 1950 U.S.A. Number o f men Percentage o f `non-smokers' among men Percentage o f `heavy smokers' among men with without with without with without lung lung lung lung lung lung cancer cancer cancer cancer cancer cancer 86 86 3-5 16-3 65 36 93 270 3-2 15-9 52 27 134 100 4-5 19-0 55 19 82 522 14-6 23-9 18 9 IES i orted by Muller in Germany : ten with bronchial carcinoma i a a questionnaire sent to the S ;he histories with those given Its showed gross differences ;< )kers and o f heavy smokers, j 1 reported from 8 countries, j . All agree in showing that * ers among patients with lung | md, with one exception (the >kers found by McConnell, ge enough to merit serious ; investigations are shown in j at there is a steady increase j itrol patients as the amount J ; in smoking habits holds for 1 women. ; ately in all the investigations, j : the difference between the j ly due to a difference in the * . regard to pipe smoking and ] gations it appeared that the ! noking (McConnell, Gordon , the disease was most closely v o f the investigators made 'as found that the proportion ong the lung cancer patients 1 Mills and Porter Levin et al. 1950 U.S.A. 1950 U.S.A. 444 430 7-0 31-0 -- -- 236 481 15-3 21-7 -- -- Wynder and Graham e Doll and Hill 1950 U.S.A. 1950 Britain 605 780 1-3 14-6 51 19 649 649 0-3 4-2 26 13 McConnell et al. 1952 Britain 93 186 5-4 6-5 35 22 Doll and Hill* 1952 Britain 708 708 0-7 4-8 24 14 Sadowsky et al. }i f * Wynder and Cornfield 1953 U.S.A. 1953 U.S.A. 477 615 3-8 13-2 -- -- 63 133 4-1 20-6 68 29 Koulumies 1953 Finland 812 300 0-6 180 66 31 Lickint i Breslow et al. 1953 Germany 224 1,000 1-8 160 74 29 1954 U.S.A. 518 518 3-7 10-8 74 42 if Watson and Conte 1954 U.S.A. 265 277 1-9 9-7 73 57 cWSR. Randig Gsell 1954 Germany 415 381 1-2 5-8 34 18 1956 Switzerland 150 150 1-3 19-3 67 15 Kreyberg 1956 Norway 213 4,158 1-4 13-2 18 7 gl jgfHB -Schwartz and Denoix 1957 France 602 1,204 31 15-8 13 7 Ipo/e--It has not been possible to make all the figures completely comparable. Some series, for example, include a few women. The individual papers should be referred to before any. TM detailed use of the figures is made. 5Doll and Hill's 1952 paper gave results for 1,357 men with lung cancer: the results shown here 'elate only to those which were not included in the 1950 paper. CARCINOMA OF THE LUNG (1956), who cultured human foetal lung in vitro, with and without the addition of benzpyrene. The addition o f the benzpyrene was found to result in hyper.; plasia o f the bronchiolar epithelium with 3 or 4 layers replacing the normal one and ii%gular enlargement and abnormal mitosis in the hyperplastic cells, similar to that described for the initial stages o f carcinogenesis in mouse skin after painting with benzpyrene. The possibility must also be considered that cigarette smoke does not itself initiate carcinogenesis, but that it acts as a co-carcinogen or activator to substances already present in the air from other sources. For example, pyridine,; pyrrole and other solvents are found in the smoke and these might be supposed ] to elute the benzpyrene adsorbed on to carbon particles in town air and so render the town air more active. Such an effect might contribute to the1 difference in lung cancer mortality between urban and rural areas, but it is: unlikely to account for the whole effect o f cigarette smoking since there is a well marked relationship between smoking and the disease in areas o f minimal ; atmospheric pollution (pp. 86 and 87). Evidence that cigarette tar might act as a co-carcinogen or as a promoting agent in conjunction with other environmental carcinogens has been obtained; by Gellhom, Klausner and Hibbert (1956). It is conceivable also that it might, act in conjunction with weak initiating agents produced in pulmonary scars. C onclusio n It was concluded earlier in this Chapter that cigarette smoking should be ; regarded as one o f the causes o f lung cancer, so long as this conclusion did not1 conflict with any o f the established facts which were relevant to it. In the author's view, the relevant facts are not inconsistent and the hypothesis should, therefore, be accepted. The credibility of the conclusion is strengthened by (1) the observation that histological changes which commonly precede the development o f cancer are also related to smoking, (2) the discovery o f several carcinogenic substances in tobacco smoke--including one which is believed , to be widely effective in man and (3) the demonstration that extracts o f tobacco ; tar can, under suitable conditions, cause cancer o f the skin in animals. The failure to reproduce the disease in the bronchi o f animals cannot justify the rejection o f the conclusion, nor would it do so, even if it were possible; to reproduce the conditions o f human smoking. Species differences in suscepti bility are a commonplace in cancer research and positive results in one animal} can never be overthrown by negative results in others. Moreover, as Kennaway; (1957) says: `Negative results o f smoke inhalation experiments on rodents seem} to me to be o f no significance, because the animals, presumably, unlike smokers, keep their mouths shut and pass the smoke over their turbinates'. Positive} results, however, would undoubtedly reinforce the conclusion and would be invaluable in helping to isolate the responsible agent. In the author's view the conclusion that cigarette smoking is a cause--and an important cause--of the disease is inescapable, but the picture is not yet complete. It is not known whether the polycyclic hydrocarbons in tobacco smokei 78 i without the addition ad to result in hyper replacing the normal the hyperplastic cells, genesis in mouse skin ette smoke does not nogen or activator to or example, pyridine, ;se might be supposed s in town air and so ffit contribute to the rural areas, but it is Dicing since there is a >e in areas o f minimal en or as a promoting ;ns has been obtained ible also that it might n pulmonary scars. te smoking should be lis conclusion did not elevant to it. In the he hypothesis should, >n is strengthened by mmonly precede the e discovery o f several >ne which is believed at extracts o f tobacco in in animals, ai o f animals cannot iven if it were possible lifferences in susceptiresults in one animal jreover, as Kennaway lents on rodents seem lably, unlike smokers, turbinates'. Positive lusion and would be Dicing is a cause--and j the picture is not yet] Dons in tobacco smoke [i THE SMOKING OF TOBACCO je sufficient to account for the production of the disease, or whether some other rcinogenic agent has yet to be detected. It is not clear whether the explanation jiat has been offered is adequate to account for the whole o f the difference in jortality between Britain and other countries (particularly the U.S.A.) nor iwhat extent the temporal changes in mortality can be attributed to changes tobacco consumption. These questions cannot be answered until more is jown about the mechanism o f carcinogenesis and the part played in the production of the disease by atmospheric pollution. The few data on the effect o f inhaling are conflicting and there is also some conflict in the evidence regarding the relative risks of smoking tobacco in pipes and in the form of agars and cigarettes. The reason for these differences in risk is also unknown, be fields o f uncertainty are, however, small in relation to the extent of stablished knowledge and do not justify throwing doubt on the main conclusion. This conclusion is not accepted by all who have studied the subject. |Kotin (1956) and Hueper (1951b) believe that cigarette smoking is a relatively unimportant cause of the disease and Rigdon (1957) and Fisher (1957a) doubt 'it is o f any causal significance at all. These views, however, do not represent be generality o f scientific opinion; for example, they are not supported by |any o f the independent commissions which have been appointed to review the fevidence. In Holland a commission was set up at the request o f the Director of ublic Health, and it concluded that there was a connection between cigarette Ismoking and lung cancer. In the commission's view it was not possible to be pure that the connection was causal, but the other possible factors, which Iwere also studied, `all proved to be less probable in this respect than smoking' [(Netherlands Ministry of Social Affairs and Public Health, 1957). In the U.S.A. a study group was organised by the American Cancer fSociety, the American Heart Association, the National Cancer Institute and fthe National Heart Institute. Their report was published in extenso in the national press and provides a valuable survey o f the principal data. They concluded that, `The sum total o f scientific evidence establishes beyond reasonfable doubt that cigarette smoking is a causative factor in the rapidly increasing ^incidence o f human epidermoid carcinoma of the lung'. (Study Group on | Smoking and Health, 1957.) In Britain the Ministry o f Health constituted a panel under the chairmanIship o f the Government Actuary to advise'on the statistical aspects of the I subject and, as a result o f their report, the Standing Advisory Committee on jiCancer and Radiotherapy of the National Health Service advised the Minister | of Health that `(1) It must be regarded as established that there is a relationship [between smoking and cancer o f the lung. (2) Though there is a strong presump|tion that the relationship is causal, there is evidence that the relationship is not j a simple one . . . ' (Minister of Health, 1954.) Three years later, more evidence IIw as available and the Medical Research Council (1957) advised the Minister that: `In scientific work, as in the practical affairs of everyday life, conclusions Ihave often to be founded on the most reasonable and probable explanation | o f the observed facts and, so far, no adequate explanation for the large increase 79 .k k CARCINOMA OF THE LUNG in the incidence o f lung cancer has been advanced save that cigarette smokinj is indeed the principal factor in the causation o f the disease. The epidemiological evidence is now extensive and very detailed, and it follows the classical patte: upon which many advances in preventive medicine have been made in the pasi It is clearly impossible to add to the evidence by means o f an experiment ii man. . . . Evidence from many investigations in different countries indicate: that a major part o f the increase is associated with tobacco smoking, particular! in the form o f cigarettes. In the opinion o f the Council, the most reasonabli interpretation o f this evidence is that the relationship is one of direct cause? and effect.' fi ATMO D ifference in M op rpH E principal reason ft | 1 responsible for some case Ibeen recorded as being hi; 11952; Clemmesen, Nielser i Kennaway and Kennaway Kreyberg 1956). The greate: ; Denmark, where the ratios in the rural districts are ap U.S.A. the differences have i f the ratio between the male | districts is slightly less thai | for which detailed data are i extent o f urbanisation. D | shown in Table III o f Chap l Similar data for the period l who showed that for 89 lar f cancer mortality in men am Kennaway and Kennaway i i mortality increased in pro] Correlation between Mortal Rural Distru Standardised mortality, Sex 39 38 36 M 69 63 76 F (22) (77) 96 Coefficient of correlation bet' * After Cur Figures in 80