Document Yj3Ve5wMjd9ezgkggy5MEpDdE

FILE NAME: Smoking (SMOK) DATE: 1940 DOC#: SMOK002 DOCUMENT DESCRIPTION: Journal Article - Does Chronic Irritation Cause Primary Carcinoma of the Human Lung? | A rchives of Pathology I' I t \ $ K I) r i'O Jil A L .HO.M U) t LU D V IG H K K TO K N , Cliiri.Ko, (.1,ii'f K d lto r ; JAMES FAYING, N rw York OSCAIt T. M'.IHJI.TZ, I- vaii sl. m , 111. S, H. W O L H A C I I , l l os t on CKOItC.K II. V.T1UM>LK, Il oi lir. strr, N. V. FLANK H. VI NM', I'mlland, Ore. t s V o l u m e .10 1940 I'UHr.TSHFKS AMERICAN .MEDICAL ASSOCIATION CHICAGO, ILL. Critical Reviews IK.) liS CHRONIC IRRITA TIO N CAUSE PRIMARY C \RCi v^M A OF THE HUMAN LUNG? MADGE TI1U PLOW MACK LIN, M.U. AND CHARLES C. MACKLTN, M.D. I.ONDON, ONTAUJO, CAN AD A l.'riiuary cancer of (lie Inn" , in common willi other forms of malignant , g growth, is almost universally believed to originate through some form i.luotuc imitation, which usually sets up a chronic inflammatory process in the bronchial mucosa. J.he literature contains many examples of this view. Tints 1kividson * stated : `'Numerous theories have been advanced, A || musily relating to the factor of irritation in some form or another of the V.y|| respiratory tract, to account for this phenomenon." Simons," who listed tom teen causes, found chronic irritation of the bronchial mucosa a factor- common to all ol them and. concluded that this common factor is the cause ol pulmonary cancer. Ericd-1 said: . . it is beyond doubt that irritation is as necessary a predisposing factor in the causation of cancer of the lungs as it is in cancer of other organs." In the minds of the commentators it seems to make hut little difference what the irritant; is, so long as it is an irritant, and there arc various types, such as ; mechanical, chemical, thermal, bacterial, electrical and radioactive. Often more than one type is at work on a given region, as when an area of the :Af longues surlacc is persistently scratched by a sharp stump of a tooth cpi (mechanical) with erosion of the epithelium and exposure of the unpro- , > tec.ted tissues to irritants of bacterial and chemical types. The sine qua non seems to be that the irritant shall act constantly or at: short intervals 1,.',] over a lour/ period. There is no agreement as to the length of time that jfi is required in any given case to produce a cancer; it is conceded that thisT; factor is a variable one, and operates in association with other variables, . such as age, anatomic site and predisposition. The irritant, too, must From the. University of Western Ontario Medical School. 1. Davidson, M . : Lancer of the Lung and Other Infralhoracic Tumors, Bristol, England, John Wright & Sons., Ltd., 1930. 2. Simons, E. J.: Primary' Carcinoma of the Lung, Chicago, The Year Book: Publishers, Inc., 1937. 3. Fried, It. M.: Primary Carcinoma of the Lung, BaUigiorc, Williams & Wilkins Company, 1932, 924 min, u iim m nm m inm m aw w m w w i M A C K L I i i '- M A C K L I X -- P R I M A R Y C A R C I X O M . I OV I . U X i , act sufficiently strongly to produce an effect in the tissue but not so strongly as to cause total necrosis of the tissue. It seems certain also that in most if not all cases inflammation super venes in the region undergoing exposure to the irritant. The inflamma tion is a reaction on the part of the tissues to the injury occasioned by the. irritant. It may, theoretically at leas!, be regarded as protective in nature, though the means sometimes defeats the end, as in silicosis, Since chronic inflammation is a well nigh inevitable and universal concomitant of chronic irritation, the one term has often been used as synonymous with the other, and thus it: is that chronic inflammation is regarded as the ctiologic agent in primary cancer of the lung, without any distinc tion being mace between the fundamental irritating agent and the tissue responses following its chronic action. So widespread is this usage with regard to chronic inflammation, indeed, that we find it convenient, in the following pages, to discuss the alleged causes of primary cancer of the lung under headings of the pathologic conditions, generally inflam matory in nature, rather than of the specific irritants. Commentators have found it difficult, too, to distinguish between chronic inflammation per se and the. chronic reparative process, and it would seem that the two phenomena are so involved, one with the other, as to he practically insepa rable. Thus, if we were accurate, wc should ascribe the growth of the cancer to the hyperactive, repair process, with its urge to increased cell multiplication, reserving to chronic inflammation or to some other factor the carcinogenic influence which first sets a cell or cells on the cancerous path. it is our purpose in this review to bring forward dala indicating l/ial l)u? thesis that primary carcinoma of the hint/ is due to chronic injlammatory changes in the lung is unproved. We do not feel justified as yet in denying outright that chronic inflammation causes pulmonary cancer, but wc affirm that on the whole the evidence now available does not warrant such an assumption. Kvcn the carefully selected, so-called positive evidence that is commonly regarded as supporting the thesis is by no means unequivocal. That chroirc inflammation should have, been implicated as the cause of cancer is not surprising, for the two arc often present in the same region. Wc would point out, however, that because primary cancel' of the lung is not infrequently found associated with chronic inflammatory changes is no reason for assuming that it necessarily is dflie these changes. Indeed tin's association no more justifies the conclusion that the cancer is so cavsed than does the circumstance of a man being found near the scene of a crime establish beyond doubt that it was lie who com mitted the crime. 1.1is presence at the scene may have been merely fortuitous. It seems quite possible that chronic inflammation often stands m relation to p am.arv cancer of the lung as docs the innocent man a! the Cvi u i$ 9 A a r c h i r n s on r a t h o i.o c v site <>l the crime. Indeed, in many rases the inflammatory changes are neither the cause nor a chance associate of the cancer hut are its result. .Moreover, even ii at some luture time adequate proof should be forthcoming to show that in some cases chronic inflammation or irrita tion is the extrinsic cause of pulmonary cancer, such proof can he used only in mass data, not in individual eases. There are persons who present no evidence whatever ol chronic inllamniatorv disease of the lung.-, who nevertheless show development of cancer of the lung, and, on the other hand, there are persons who. despite having sultered hum inflammatory diseases of ihe lungs for vears. never have cancer of the lung. It is impossible to sav ol anv one ease that had the chronic inflammatory change not existed cancer ot the lung would not have develi iiicd. *;I'Wm'W /ft-M vv#f 'K'# i y K N V I K O X M K N T A M ) ) I I '.)< h i ) I T V I N I U K K T I o l . o o v OK P R I M A R Y ; CAXOKK OK T I I K l . K X U As with all other cancers, the problem of the origin of primary cancer ol ihe lung natnrallv divides ilscll into two aspects concerned with ., lil 'i^ P ( 1) supposed productive taelurs acting trom outside the bodv and (2) constitutional factors inherent in the bodv .and varying with the patient and particularly with the "strain" or line of descent. These agencies * .are characterized as environmental and hereditary, respectively. The environmental factors may he subdivided into two categories: (1) those that would he encountered in the course of ordinary living, which are the various chronic irritants commonly used to explain human cancer, and (2) those which have been discovered bv inducing cancers in animals through the. experimental use of specific chemical agents, called car cinogens. As man is not used in the ex]K'rimental production of cancer, all malignant growths arising in him must come under the heading of( 1) tumors induced by the chronic irritations of ordinary life or (2) T Ml f IVP tumors arising on account of constitution. These may he considered as primary poles, or extremes, useful in our mental processes in the study of the causes of cancer. .Actually, the constitutional factor enters into every produclion of tumor, and the only point which needs to lie deter mined with reference to it is the extent to which it is participating in any given case. So, il the environmental factor is admitted, all tumors are ol dual origin. \Ve shall now review briefly the sources of so-called chronic irritation to determine whal proof has been brought forward to sustain the thesis that chronic irritation causes pulmonary cancer in man. A large number k#$l Ci-' flVl of citations from the literature might He made in support of the idea' |i that chronic irritation, especially in the form of chronic intlammatory change, is responsible lor the genesis ol primary cancer of the lung. ` ' I ,,,iii-nTiitinniii-nitwmmttmm rmtntuniin M A C K /./.X- MACK U X - I 'RI M. I K V C A K C L X O M A O I ! /.('.Vo 'Ml Tlicsc would be, in practically all cases, merely opinions without any valid basis el proof. They might be encountered with an equally large number of statements questioning the role of chronic irritation in the causation ol primary cancer of the lung. These, again, would lie opinions, for the most part, with no proof to sustain them. 'I hey would show two tilings, however: tirst, that the theory of chronic irritation lias not been accepted without question and second, that the burden of proof should rest on those who assert that primary cancer ol the lung in man is caused by chronic irritation or inflammation. AI.I.EI KI> KNV1UO.NM K N T A F C A l ' S F S OK P R I M A R Y C A N C K R OF 'IMIF It I'MAX FONT. Simons : listed a number of supposed general causes of primal}' pulmonary tardnoma in man, and in them he finds a common factor, namely, clinnic irritation. These alleged causes should he divided into two sections: ( 1) those capable of being produced by pulmonary cancer (these are olten found associated with such cancer because they are sometimes the results of the presence of the cancer either alone or in combination with other influences) : (2) those which are not capable of being produced by cancer of the lung hut which might he found asso ciated with it in the same patient in a varying percentage of cases. Willi regard to gioup 1. it might well he that in some instances these conditions have been present before the cancer ever started, although tluw are not necessarily the cause ol the cancer, while in others these conditions have not been present until alter'the cancer started and are a direct result ol the growth of the tumor. We shall now discuss the. conditions which belong to group 1. '1 hcv arc chronic bronchitis, pulmonary abscess, asthma, bronchiectasis, pleurisy, emphysema and conditions simulating pneumonia and influenza. The conditions listed in group 2 by Simons * are trauma, tuberculosis, pneumonneoniosis, chalicosis, syphilis, roentgen in ablations, inhalation ol dust (especially from tarred roads), smoke i especially if bearing chemical irritants, as in certain of the industries), motor exhaust fumes and war gas, various industrial and occupational hazards involving the breathing of hot air ('especially if hearing; chemical fumes), tobacco smoke or the air ol certain mines, such as those of Schneeberg, .Bohemia, which contain not only irritant dusts hut radio active subs.antes as well. It will he found that as industry e l a n d s and new chemi.tals are encountered, this second list of conditions which are supposed either rightly or wrongly to cause cancer of the. lung will increase. l.et us review first the course of development of a cancer of the lung, so that we may see why the conditions in group 1 arc so often found associated with this maladv. 928 A R C H IV E S OE PATH O LO G Y COUKSK OF DEVELOPMENT OF TYPICAL CANCKR OF THE LUNG Although no one has followed the course of a cancer of the lung from its onset to its full expression, the following outline is regarded as probably accurate: Tbe epithelium in some local region of the bronchus undergoes a change which makes it "malignant" to the rest of the body. At this point we are not discussing what: the nature of the change is, or whether this, change is occasioned bv forces inherent in the cells or is jfak . . . r , . - ... brought about bv extrinsic influences. The cells whose ability to dilter- entiate into normal cells and to organize into normal tissue has been wholly or parti)- lost, still retain their ability to grow and do grow. The growth at first is probably slow. This is flic beginning of the tumor. In time tbe tumor causes a narrowing of the bronchus, so that an obstruction to airflow ensues, together with a handicap to the normal bronchia! eliminative mechanism. This impairment of function is followed by degenerativc changes in the part of the. lung served by the involved bronchus. The irritation produces a cough, often nonproductive at first, of chronic character, which simulates that, of chronic bronchitis. Inflammatory changes in the mucous membrane follow the damming back of the excretions normally wafted away by the cilia or removed by the perisuiltoid action of the bronchus or by coughing. The cough grows worse, and blood-streaked sputum is produced. When the obstruction to the bronchus is complete, either through blocking of the lumen by the growth within or by pressure on the. bronchus bv the encircling growth without, there is atelectasis in the lung beyond ihc point of obstruction. This is followed by bronchiectasis of the hVonchi in the collapse area, which dilate to make up for lost volume of lung. Another effect of the diminution in volume of lung is compensatory emphysema in the alveoli of the surrounding lung tissue. These become more dilated with cough ing. An asthmatic wheeze may develop in the narrowed bronchus before it is completely obstructed. If the center of the tumor becomes infected and undergoes necrosis, the patient may show an abscess of the lung with drainage of the pus through the bronchus. If the tumor grows extensively toward the periphery of the lung before it breaks down, the pleura becomes inflamed, and pleural effusion results. The pus may rupture into the pleural cavil)', and the patient may he operated on for etnpyema without the cancer of the lung being recognized. The onset of the symptoms may he acute and simulate influenza or pneumonia; or the onset may be gradual, and tuberculosis may be diagnosed. In any of these steps an erroneous diagnosis may he made, and the patient mav later, when the cancer is discovered, assume that the `'chronic bronchitis," ``influenza," "asthma" or "tuberculosis" which lie was supposed to have was the cause of the cancer. The physician sometimes accepts at face value statements t.0 iwmTMWTWWftWHWtltmmu' nA.diriiwitrU:;,ci' MA CKLIN-MA CKLl .V--PR/ /I1AKY CAh'CIXOMA Oh I.C.Yu 929 the efl.ec': that the patient had chronic bronchitis lor some time before the cancer mas recognized ami that the one was therefore the cause of the other. The course of bronchogenic cancer, as is probably true for all other cancers, can be divided into three periods: .First, there is a "silent'' period, which lasts from the moment of the initial change of normal cells into cancer cells up to the time when symptoms of such severity begin tint the patient associates them with the onset of bis illness. This stage is apparently the longest and is completely silent as far as the localized activities are concerned (though symptoms may he produced from a metastasis in some other region) hut may give rise to vague general symptoms. The second, or symptomatic, period is characterized by the onset of symptoms, which progress up to the time of the diagnosis of cancer of the lung, or until death if the diagnosis is not made. This period may he relatively brief, only a few months, or at the most it may be a few years. The third period is the "diagnosed'' stage and is usually ihe briefest; it is nonexistent, if the diagnosis is not made. In most cases it lasts from the time of diagnosis of cancer of the lung until death, for relatively few patients have been operated on and cured, even for a brief time. It is evident, then, that because of the initial silent period there, is trouble in ascertaining whether any particular chronic inflammatory disease ]'receded the onset of the cancer of the lung. Thus bronchitis, coming before the "diagnosed'' period, is thought to antedate the. cancer, although, the cancer existed in the silent period before the onset of the bronchitis. The symptomatic period, before cancer of the lung is recognized, during which the patient may have various pulmonary symptoms, helps to contuse cause and effect of cancer of the lung and to make patient and physician alike interpret the symptoms produced by the cancer as symptoms of conditions giving rise to the growth, . S t 'K C i n i C SO-O A 1.1,K I ) C A U S A T I V K KAOTOKS <) !' I I AM' I I K OF T1IF. I. UNO: C R O U P 1 Chronic Bronchitis.---Chronic bronchitis has been looked on by some writers as of importance in the etiology of cancer of the lung because (1) it produces chronic inflammation of the hronehia^muyysa and (2) it is found with some frequency in the histories of patients with cancer of the lung. What evidence is there to support or to deny the assertion that chronic bronchitis plavs a causative role in pulmonary cancer? One of the fr'st arguments that will occur to any one is that bronchitis may be generalized in distribution, while cancer of the lung is localized. The inflammatory changes of bronchitis are widespread' and peripheral as well as central, while cancer of the lung is loeali/.ed and is usually hilar. 9.50 A U C i i i i 'ES O f r . r n i o f . u t ,d 11 chronic iiiliammntory changes induce the cancer, why does not the cancer originate dilluselv over the lung? Chronic bronchitis has no special predilection lor the male sex, but cancer ot the lung occurs about tour times as often in males as in lemalcs. It would seem, if chronic bronchitis is a major cause of cancel ul the lung, that men should be much greater sufferers from bronchitis than are women, yet no such preponderance has been mentioned in the texts on diseases of the lungs, .and mmlalitv statistics do not show more deaths I'ronn'ironchitis in males than in females. Although it might he assumed thru the regeneration of the bronchial epithelium in chronic hroncliitis would lead to cancer of the lung, we must ask it there is proof that it does so. hirst, is bronchitis followed more oltcn hv cancer of the lung than not? Second, how inativ patients with cancer ol the lung give a historv of chronic bronchitis that: is of such long duration that it can legitimatelv he looked on as having ante dated even the silent period of the cancer's growth? There are no adequate answers to cither ol these questions because we do not know how many persons in the population ha\e chronic bronchitis and we have little idea, ol how long the silent period can he. Since most takers of cluneal histories content themselves with stating that the patient: had complained of chronic bronchitis, failing to give the time during which the bronchitis existed, we cannot place much reliance on reports in which it is merely stated that a certain percentage of patients with cancer of the lung have had hroncliitis. b'irst, we should have to exclude all those in whom bronchitis probably lollowed instead of antedated the silent period of the pulmonary cancer. ' Since we know that cancer in general is slow growing, and that in men who have worked in the Schnecberg mutes cancer of the lung may develop from ten to twenty-seven years after they have been pensioned oft from work in the mines, we should probably have to allow a minimum of at least ten vears bclorc the cancer was diagnosed as the period during which it was probably present in the silent stage. Therefore, all patients with cancer of the lung in whom chronic bronchitis started within the last ten years before the diagnosis of cancer of the lung was made should he excluded as most probably being patients in whom the cough was cither the result of the tumor or coexistent, with it though not actually caused by it. Those who had a history of cough for more than ten years might possibly be looked on as patients in whom the cough antedated the cancer, although this would not inevitably be so. Second, we should (ben compare the latter group of patients with the general population in regard to the incidence of chronic bronchitis. If 10 per cent of the population past I0 have a historv of chronic bronchitis, we need not be surprised if 10 per cent of patients with cancer of the lung M A CK 1.IX-MA C K L I X -- /`K IM A K 1' C AK C1X 0M A 01' JA JX i, 9.11 have a history of chronic bronchitis without there being anv cause and effect relationship between the two. If 60 per cent of the general popu lation have had chronic bronchitis, 60 per cent of patients with cancer of the lung may be exacted to have had chronic bronchitis antedating the earliest moment of their cancer's growth without its being in any way responsible for the tumor. Unfortunately, we do not have reliable data on the incidence of chronic bronchitis in the middle-aged population, hut Maxwell and Nicholson 45 remarked, ``The commonest antecedent condition was chronic bronchitis, which occurred in the. past history 21 times (100 cases), although when it is remembered that this condi tion is exceedingly common, the fact is deprived of any great signifi cance." Most of the writers whose observations are cited to show that chronic bronchitis is a causative agent in pulmonary cancer have for gotten two facts: (1) that' cancer of the lung causes chronic bronchitis in some of its victims and so explains a certain percentage of the cases in which the two conditions are found, and ( 2) that unless chronic bron chitis antedating the cancer is found more olten in patients with cancer of the lung than in the general population of the same age and sex as these patients, its frequency in the patients with cancer of the lung, how ever great, is not significant. Moreover, chronic bronchitis, even includ ing that caused by the cancer, does not occur in more than about 50 per cent of the patients if one strikes an average from the numerous series of cases reported in the literature. Thus we can say that chronic bronchitis is by no means universally found in the histories of patients with cancer of the lung, that when it is found it is because the presence of the tumor has caused the bronchitis in a large number of the cases, that there is no wav of showing that chronic bronchitis which might conceivable have antedated the. tumor of the lung is more commonly found in patients with cancer of the lung in the general population and therefore that the role of chronic bronchitis as a causatizr utjcnl in the production of pulmonary carcinoma has not been demonstrated. Jlronchiectasis.-- Like chronic bronchitis, bronchiectasis is Irequently a result of the presence of primary cancer of the lung, following on atelectasis caused by occlusion of a bronchus by the tumor. 1fetiee this condition might be found cither clinically or at autopsy in a large per centage of the eases of primary cancer of the lung, hut it-y cgtjpcd by. and is not the cause of, the cancer in the majority of eases (Zacherl r`). Because a benign tumor of the lung lasts over so much longer a period than docs pulmonary cancer, a benign tumor can almost always grow large enough to occlude, a bronchus before killing the patient. There- 4. Maxwell, J,, and Nicholson, W. A,: Quart. .1. Med. 24:29, 19.6. 5. Zacherl, S . : Wien. kiln. Wehnsclir. 44:997, 19.11. 932 A R C H IV E S OF I'ATHOLOGV fore, bronchiectasis is always an accompaniment of a bcnitjn tumor of the huuj of any, duration, according to Wcsslcr and Robin.0 It may be, of course, that a patient who has had bronchiectasis due to some other cause may subsequently have cancer of the lung. In such a case the bronchiectasis preceded but did not necessarily cause the cancer. Many patients have bronchiectasis who never have cancer of the. lung; indeed, Graham, Singer and llallon 67 stated that in patients in wbowf-lhija-primurv lesion is bronchiectasis "it is striking that there are but few examples of carcinoma which have developed in a hronchicctatic dilatation." Until the. proponents of the idea that the chronic inllammation caused by bronchiectasis induces cancer of the lung can show that bronchiectasis wlurh preceded the onset of the cancer is lollowed by cancer in the inflammatory zone' of the wall of the bronchiectatic cavity in a percent age of cases which is .significantly higher than the percentage of cases of cancer of the lung in the population of tint same age and sex dis tribution which is without bronchiectatic cavities, we may conclude that no relation has been demonstrated between the chronic inflammatory chant/cs occurrimj in bronchiectasis and pulmonary cancer. . Islhina.... Asthma, although not primarily inflammatory in nature, may induce chronic thickening of the mucosa as well as of the muscle coats. It has hecn advanced hv some authors as one of the causative agents in cancer of the lung. As in the case of the two conditions just discussed, asthmatic symptoms may result from the presence of a pul monary cancer, and so a history of asthma preceding the diagnosis of the cancer cannot he accepted as evidence that asthma caused the cancer. On the other hand, true asthma of long duration may be found in per sons in whom cancer of the lung later develops, and, again, this need not imply a causal relation between the two conditions unless cancer of the lung can he found in persons with chronic asthma in far greater proportion than in the general population. Again, we have no idea of the true, incidence of asthma which antedates cancer of the lung in patients with this form of cancer, since publications based on the clinical histories give for the most: part: merely a statement that asthma occurred in a certain percentage of cases of cancer of the lung without giving any idea, of the length of time during which the asthma had existed before the cancer was diagnosed. from eases reported in the literature it seems that on the average about 7 per cent of patients with cancer of the lung had a history of. asthma. Therefore, even if asthma, is looked on as contributing to the 6. Wcsslcr, .11., and Robin, C. 1!.: Am. T. M Sc. 183:164, 1932. 7. (graham, K. A.; Singer, T. J., and Gallon, II. C .: Surgical Diseases of the Chest, Pliiladelidiia, Lea 3: Febigcr, 19,15, p, 815. MACKUN-MACKUN--PMAUKV C.IIS 'I.XOM.I Oh U Wi, 93.S predisposition to pulmoiuuy cancer, it is by no means a universal cause of such cancer, and one has no evidence that it is a cause at all until one can prove that true asthma antedating beyond question the onset of a pulmonary tumor has occurred far more often in patients with cancer of the lung than it has in patients without cancer of the lung. Such proof at present is not forthcoming. Emphysema.-- Pulmonic alveolar emphysema of the degenerative, type, with a barrel-shaped chest, overly large alveolar spaces and other signs, is not primarily an inflammatory change associated with regen eration of the bronchial epithelium and cannot reasonably he looked on as productive of cancer of the lung. Emphysema of the compensators' type may be an accompaniment of cancer of the lung in the later stages of the latter, after atelectasis has ensued, and is then a result not a cause of the cancer. The incidence of this condition in the general population is not known, nor is its incidence as a condition truly antedating cancer of the lung known; therefore one can make no reliable deductions as to the role which alveolar emphysema plavs in causing cancer of the lung. Such a role has not: been demonstrated. Pulmonary Abscess.--This may be associated with cancer of the lung as a result of the breakdown of the center of the growth, so it is easily understood wiry such a history should he recorded frequently for patients with cancer of the lung. Funk 8 stated that a pulmonary abscess of obscure cause in a man past 30 should make one suspect that a malignant growth is present. Pleural Effusion.--An involvement of the pleura in the form of dry pleurisy, a pleural effusion or empyema maw he found in patients with cancer of the lung, and not infrequently it is the first symptom of which a patient complains. One of the methods of diagnosis of cancer of the lung is the centrifuging of fluid aspirated from a pleural effusion with a' view to finding cancer cells in the sediment. Maker-Tales and Ellis 0 stated that a patient with pulmonary cancer may lie treated surgically for empyema for some time without the underlying cause of the empyema being recognized. Summary.-- Summing up the evidence in this group of cases, we laid the following conclusions to be justifiable: Because chronic, bronchitis, bronchiectasis, asthma, emphysema, abscess11*of^TRe lung and pleural effusion may he found as a result of the presence of the tumor in the bronchus, it is inevitable that a certain degree of association between these conditions and cancer of the lung will he found in any senes of cases of pulmonary cancer. The crucial point to he proved to sustain the theory that these chronic inllnmmatory conditions actually . Funk, R. II.: J. A. M. A. 95:18/9. 1930. 9. ISaker-Ratcs, K, T., and Fllis. C. R. : I.ancet 1:676, 1633. tw!l,i;!i ' SO ISj predispose to rancor of flic lime js tliat their presence actually antedat ing the probable time of, onset of the, tumor is found in significantly higher percentages of patients with cancer of the lung than of those without cancer of the lung when the two groups are kept comparable as to age, sex, and industrial occupation where there is the slightest sus picion that such occupation might be a cause of pulmonary cancer. We have shown that such evidence has not been advanced for any one ol the afouijKnenjinned conditions because there has been little attempt to separate the patients with cancer of the lung into two groups, those in whom the pnlmonarv conditions were a result of the presence of the tumor or at least were initiated after the tumor had started to grow, and those in whom the conditions might legitimately he interpreted as having antedated the cancer. There also has been little attempt on the part of those who affirm tilth these conditions predispose to cancer of the lung to ascertain the incidence of these conditions in the general population of the same age and sex as the group witii cancer of the lung. The opinions advanced by the supporters of (lie idea that these chronic pulmonary conditions cause cancer of the lung have largely been based on the finding of these conditions in an appreciable number of patients with cancer of the lung, without regard to their incidence in the general population. We have shown that no single one of these, conditions has been j'nraed to he the cause of cancer of the lung. The addition of a number of uncertainties mav make a larger number, hut ft does not add to the accuracy of (he result. If none of these has been shown to cause cancer of the lung, the sum total of them has not been shown to cause cancer of the lung, nor has the common factor resident in them, chronic irritation, been shown to he conducive to the formation of tumors of the lung. nuot;i' n Y\ c now turn to the second group of so-called chronic irritations listed bv Simons ~ which arc looked on as causing cancer of the lung hut which arc not symptoms of the disease itself, although in some instances they may simulate some of these conditions. Influenza....In a search for previous infections of the lung which might explain the origin of primary pulmonary cancer and especially its apparently great increase since 1920, workers have suggested that influenza might he a major factor. Askanazy,10 Wiutcrnitz, Wasou and McNamara n and others found metaplasia in the bronchial epithelium following influenza which in some cases was almost indistinguishable from malignant growth. There is no proof that the epithelial metaplasia 10. Askanazy, M . : Centralbl. f. ally. Path. u. path. Anat. 30:443, 1019. 11. Wintcrnitz, M, C .; Wason, J. M., and McNamara, F. P . : The Pathology of Influenza, New Haven, Conn., Yale University Press, 1920. > j, k i -v.. M.-iCKI.lX~M.'IC1\ I.I.X-- I 'KIM.-IRY (A-l/i'( / A ( M / . - l ( / l.UNC 9.AS following influenza won on to cancer format ion, since carcinoma of the lung Inis developed in only a very small part of the population who had influenza in 1918 and recovered. In patients who had influenza and who have since died ot some other disease there has been no higher incidence of cancerous change in the lungs than in patients who have died without having had influenza. Thus, it 50 per cent of the popula tion now at the ages of 40 to 50 had influenza in the period from 1918 to 1920, one would expect 1o find that 50 per cent of the patients with cancer of the lung who are now between the ages of 40 and 50 would give a history of having had influenza. Itefore one may attribute to influenza a part in the causation of cancer of the lung it will have to he shown that influenza has occurred in a far greater number of persons with cancer of the lung than in persons not having cancer of the lung, and even then there would he no conclusive proof that a short attack of influe nza, lasting hut a few weeks, could produce such lasting changes as to .initiate a pulmonary cancer. Even with specific carcinogenic agents, whose efleet has been definitely demonstrated, a single applica tion or repeated applications for a short, time do not usually result in cancer; to he effective, they must he. applied frequently and over a long period or must lie introduced into the body in such a way that their effect is continuous. Other arguments against the role of influenza in causing cancer of the lung are as follows: Influenza is generalized; cancer of the lung is local zed. Influenza affects both sexes equally; cancel' of the lung predominates in the male sex. The increase in cancer of the lung began to he noted in some countries before the 1918 epidemic of influenza. Einally, it must be remembered that cancer of the lung may begin its synvitomatic period as an .attack simulating influenza from which the patient does not completely recover. There is fever, lassitude, cough and then a long-draggcd-uut partial recover)'. The patient says that he. had tin attack of influenza, but what actually happened was that his pulmonary cancer began to show ilself with respiratory symptoms that simulated an influenzal attack. Therefore, before deciding that, a his tory ot influenza in a patient with cancer of the lung is indicative that influema is a causative agent, one must he sure, as with the other dis eases discussed, that lhe disease was true influenza and that it ante dated or p1robably antedated the onset of the cancer-fti.. e.. there should have been a period of at least: ten years between the influenza and the diagnosis ol cancer. .15veil that might he too short a time, hut anything under that is almost certain to he ton brief an interval. An average derived from numerous scries of cases of cancer of the lung reported in the. literature shows that in 12.4 per cent the patient gave a history of previous attacks of influenza, some of which may have )M>- A R C inim s of p a t h o l o g y bccn results oE the presence of the cancer rather than the cause. Ke.rlcy '* found that although the intluenza epidemic had been particularly severe iit Iceland, there had been no great increase in the incidence of pul monary tumor there, lloyd,'-'1 Hill 14 and Maxwell and Nicholson,4 among others, have: felt that influenza probably does not play any role in cancer of the. lung. Until it can he shown that true intluenza ante dating the diagfiftsis*V)f cancer of the lung !>v ten rears or more occurs in the. history of patients with pulmonary cancer more often than tn the history of patients without pulmonary cancer one cannot accept influenza as having been proved to he an agent in producing primary pulmonary carcinoma. Tuberculosis. -If chronic inflammatory changes are looked on as a-possible, or even a probable source of cancer of the lung, it is natural 1hat tuberculosis should have been regarded as an etiologic .agent because of its chronicity. Ewing " felt that tuberculosis was associated with cancel- of the lung unduly frequently, hut practically all other workers have felt that there was a very infrequent association between the two conditions. Unless tuberculosis and cancer of the lung arc mutually exclusive, one should fmd about the same proportion of pulmonary can cers in persons dying ol tuberculosis (provided, of course, that the. two groups compared are kept constant in regard to age and sex) as one finds in a population not dying of tuberculosis. The necessity of the proviso is at once apparent, since the highest absolute number of deaths Iron: tuberculosis is usually among tile young, and among females. One finds, for example, in Canadian data for 1936 that 1 of every 8 men who died between the ages of 30 and 4-1- died of pulmonary tuberculosis. Let us assume that all of these had tuberculosis and that none had a cancer of the lung erroneously diagnosed. Also 1 of every 100 men who died at the same age died of recognized cancer of the lung. (One would expect, without, any significance being attached to the finding, that 1 ol every .8 men dying of pulmonary cancer between the ages of 40 and 4 1 would also have active tuberculosis, and that 1 of ever)- 100 men dying between the ages of 40 and 44 with tuberculosis would also have cancer ol the lung-. Ulnari J" stated that from 3 to 12 per cent of patients with cancer of the lung have active tuberculosis. These figures deal with men of all ages dying of tumor of the lung. Again illustrating from Canadian data and taking only men dying after the age of 40, since pulmonary tumor 12. Kerlcy, V. J.: H: it. J. Radiol. 30:333, 1923. 13. Boyd, W .: Cannd. M. A. J. 23:210, 1930. 14. Hill, R. M.: Edinburgh M. J. 41:320, 1934. 15. Jawing, J.: Neoplastic Diseases, ed. 3, Philadelphia, \Y. B. Saunders Com pany, 1928. Id. Chian, II.: Klin. Wchnschr. 17:183, 1938. MAA'K f.LX-M.-H 7\ IA X~~I' h'IM .1A T C A K C IX O M A 01- l.VSC, W ? is relatively uncommon before that age, we find that approximately 3,3 per cent of men dying in 1930 in Canada died of what was diagnosed as pulmonary tuberculosis. This probably represents the minimal esti mate of men dying with active tuberculosis in Canada in that year. On that bass we should ex|>ect approximately 3.3 per cent nt males dying of cance of 1lie' lung; after -10 to show, in .addition to the cancer, active tuberculosis, a figure .agreeing with Chian's"' data. In countries in which there is far more tuberculosis than in Canada the proportion ot patients dying of cancer of the lung who have active tuberculosis, in addition, will he far higher than 3.3 per cent ; in countries in which the tuberculosis rate is lower, the proportion will he under 3,3 per cent. It is found on examining the records o) those persons with pulmo nary cancer who showed active tuberculosis of the lungs at autopsy that the cancer did not grow in the wall of the tuberculous cavity, where the inflammatory changes were taking place, in the majority of instances. The tuberculosis was more often found in the lung opposite to the one in which the cancer developed than in the same lung, if one can judge from the reports in which the sites of the two diseases were given in the autopsy protocol. Or if the disease was present in the same lung, it was not in the same lobe. Fven if the two processes arc found in the same par: of the lung, which is apparently rarely the case, this still might he due to chance association. Most authors, including Boyd,"' Maxwell and Nicholson,'1 Chian,"' Funk," Hill," 1lusted and Beil in,min 17 and others, have expressed the opinion that there cannot he demonstrated any causative influence on the part of tuberculosis in pro ducing cancer of the lung. F rie d 1" said: "It is ol interest that the occurrence of tuberculosis and of cancer in the same person or in the same organ is rarely seen at necropsy." 'Phis observation taken together with the frequency of mberculosis in the general population leads one to conclude that tuberculosis docs not predispose to cancer ol the lung. J'natiiioiiui.- -Tile onset ol cancer of the lung may simulate pneu monia, so that the patient may give a history of pneumonia, which he thinks was the cause of the pulmonary tumor. ( )n the other hand, the patient mav have had true pneumonia, either antedating the silent period or antedating the symptomatic period of the cancel', without there being niiv association between the two. Since pneumonia is In' no means an uncommon disease, one would expect to Imd a great mafivlTftbe popu lation who have had pneumonia and recovered; therefore one would expect to laid an equal percentage of patients with enneer of the lung wlio have had pneumonia without there being anv significance in the 17. ihistel, IS, ami IViimann, fi.: Aria path, el mieruhinl. Srandmav. I'til tl. m7. O .iM arc h ives or n.-rrnoux, y fact. Tims in Uatiada, 5 per cent of the men living after the ape of 40 in 194(i died of pneumonia. For overv man that died there were prob able 4 men who recovered, since the mortality rate in pneumonia is or was at that time about 25 per cent. Therefore we can assume, that another 15 per cent of men dying at that ape had had the disease at some earlier date hut had recovered and died of something else. 4 lieiefure we should not he surprised to find that about 15 per cent of patients with puhiiTiTarTvcanccr give a history of haying had true pneumonia. Tile pneumonia may have occurred in the same lobe in which the cancer was to develop or had already developed, or it may have occurred m a dillcrcnt lobe or on the opposite side, as was the case with the patient reported bv (Taham and Sniper,111 As io the incidence of the history ol pneumonia in cases of cancel of the lull!;, we lind the following data: Parish"" muiul this history m 1 of ,12 cases; Mat/.,"1 in 15.2 per cent of ITS cases; Hill,'1 in 10 of 107 cases; Maxwell and Nicholson,' in 7 of 100 cases. These are the fu/uns Tee should expert if pneumonia has no relation whatever io the produc tion of eaueeiof the linitj. Il'ar (iassini/. -- It is not to he wondered at that war gassing has been suggested as a possible cause ol pulmonary cancer in some eases. I here would seem to he only two possible wavs m which war gas might cause cancer of the lung: (1) by a direct carcinogenic effect, in which case it would probably have to he applied repeatedly in small doses over a long period, .since a few exposures would probably not induce cancer; i 2) by the production of a chronic inflammatory condition in the lung that would he carcinogenic. The first; assumption carries little weight, because il has not been shown that war gases have a specific carcinogenic effect and because the re is no history of numerous repealed applications of them; the second reason is also without much value, since we have found th,at none of the chronic inflammatory or degenerative conditions which war gases caused, such as chronic bronchitis, emphysema, asthma, activated pulmonary tuberculosis and pulmonary hbrosis, can reasonably he linked with the development of cancer of the lung. , From data supplied by Mat/'-'1 and bv Stein ami |oslin~- on United States Veterans ol the World War we see that, ol 1(>4 men with cancer ol the lung, N, or 4.S per cent, gave a history of having been gassed m the war. .Maxwell and Nicholson ' lomid d men from among about 7.x to SO m Hu- age range ol World War veterans who had histories of having been gassed in the war and who also had pulmonary cancer. 1'). (iraliatn, 17. A., and Singer, | I,: 1, A. M. A. 501:1371, 1033. 20. Parish, T. N. : Practitioner 125:324, 1030. 21. Malz, P. I!.: J. A. M. A. 111:2020, 103S. ,9 t M A C K U X - M M ' K U X ~ r k l M . - U < y C . I I X ' I X O U . l O l 1 . 0 X 0 t>5<) *I They held that war gassing has no role in causing cancer oi the lung, in order to evaluate ade<|uatclv the possible role ol war gassing in the ilK production ol cancer of the lung we need to know (he following facts: What is the number of World War veterans living at the present time:' What percentage of these were gassed in the war? In whai number of World War veterans lias cancer of the lung developed up to the present ? What percentage of these were gassed in the w ar? Is the percentage, of gassed men significantly higher in (lie. group with cancel of the lung than in the group without cancer of the lung? To these questions we have no exact answers, hut we have the following data: Approximately 4.750.000 men were accepted by local draft hoards. Of these, 740,01)0 were rejected at camps because of physical and mental defects. About 87.000 cl ed of war injuries and disease, leaving approximately 3,023,000 men who survived the war. Of these, about 75,000 had been gassed and had survived. These figures are taken from the Military Suri/coii, of October 1936,23 and Irom Malz's 21 article, on war gases, 4 luis about 1.9 per cent of the enlisted men were gassed and survived, if the death rate in the ensuing years has been about equal in the gassed and nongassed groups, eve should expect that about 1.0 per cent of the surviving veterans have been gassed. If the death rate has been differ ential in the two groups, that figure will he higher if the gassed veterans have not died in such great numbers, and lower if they have died in larger numbers than the rest of the veterans. Stein and Josliu22 reported that of 16-1 veterans with cancer of the lung. 4.8 per cent gave a history of having been gassed. 'I his would appear io he two and a half times the proportion of gassed men in the veterans' group in general, lint if we calculate what departure from the average percentage of gassed men one might expect in so small a sample of veterans as a group of 164, we line! that, on the one hand, no gassed men might he found, and, on the other, as many as 1.1 per cent might have been gassed without there being any causal relation between the gassing and the cancer. Therefore, the 4,8 per cent found is not far removed from the 4.1 per cent expected, owing to variations in samples of veterans, and we can conclude that the difference is not enough to suggest that the gassing played any part in the production of cancer. Moreover, even if gassing were conceded to have induced cancer of the lung in some veterans, it would be obvious that most such cancers developed in veterans without previous gassing (^S .^per cent) and also that in most of the men who were gassed cancer of the lung has not developed--i. c., this lesion has appeared in only several hundred among a possible seventy-live, thousand. 23. `friflith, C. M. ; Mil .Surgeon 79:251, 193(>. y-ti) AKCIUCliS OP PATllOl.OC.Y Occupations in Relation to Primary Cancer of the I.hu<i.---Although man}' attempts have'becn made to link the occupation of the patient with his pulmonary cancer, these attempts have keen unsuccessful with the sole exception of the pulmonary cancer which occurs in the miners of Schnecberg and Joachimsthal, in Bohemia, the incidence of which is so high in comparison with that, of cancer of the lung in other places and so high iir^mnysirison with the. incidence among the neighbors who are not working in tTie mines that there seems to be little doubt that this occupation is strongly productive of cancer of the lung. It. is still a disputed point as to what the carcinogenic agent is in these mines. Arsenic, radioactive emanations and air-borne molds present in the mines have been implicated. Outdoor occupations in which dust is inhaled have been thought to be represented unduly .among the patients with pulmonary cancer ( Ihiguid ) ; indoor occupations have been found to be represented m undue numbers ( Houser 23). Industrial hazards ( Schachter 2tt) , labor ing jobs ( I'rockhank 27), chromium dust ( Alveus, llaukc .and Jonas28), iron oxide in silica, dust rather than the silica itself (Campbell20), silica dust (Anderson and Dible. :i"; Dibich" ; Klolz :l~), metal grinding, exposure to coal gas in the air or to coal tar on the skin, and the preparation and sale ol tobacco (the kennawavs 1:1) have all been brought lorward as possible causes ol cancer of the lung. In all such instances the. occupation given at the time the diagnosis of cancer is made, may not have been the occupation ol the patient at the. time his cancer developed, perhaps twenty or twenlv-tive rears before. More over, one must always take into consideration the district from which the data were drawn, rims, m an agricultural district practically all the pulmonary cancers will be found in farmers, just as will most ol the gastric cancers and most, of the chronic nephritis. In a district where mining is the sole occupation most of the pulmunarv cancers will be in miners. In districts where every one is working in tobacco, all the pulmonarv cancers will be in tobacco workers, and so on. For any one. district, the effect of occupation must be determined bv hiidmg the percentage of in.lies of certain ages working in that occupation who At. Dtiguid, J. 1 1, ! I.anect 2:111, 1927. AS. Houser, (1. M. : J. l h K. 28:340, 1929; 3-!:..!18, 1934. Jo. Schaditrr, \h : J. de mod. de Paris 52 :700, 1932 ; abst rack'd, Am. .1, Cancer10:165, 19SS. 27. Brockbank, \Y. ; Quart. J. Med. 1:32, 1932, 28. Ahvciis, \Y .; P.aukc, K, K., and Jonas, \V, : Klin. Wcbnsrhr. 15:1854, 1936. 29. Campbell, .1b A.: Bril. .M. J. 1:755, 1938. SO. Anderson. C. S,, and Dibit, J. H. : J. Ilyg. 38:185, 1938. SI. Dibit, .1. 11.: I. ancet 2:982, 1934. SJ. Klotz, M.: Am . J. Cancer 35:38, 1939. ss. Kcntiaway, M. f4., and Kennawav, Id 1..: J. 1Ivy. 3(5:236, 1036. M. t CK J IX-M.-H 'K l.l X - ~ l 'l\'l M. II\V C . I K C l X O M . t Oh' U X ` , `Ml have pulmonary cancers, as compared with the percentage of males of the same apes not in that industry who have such cancers. ( )ue must not compare the incidence of cancer of the lung in middle-aped males work ing in mines with the incidence of cancer of the limp in till the rest of the villagers, males and females, of all ages. One must try to keep the two populations compared as constant as possible with respect to till factors with one exception, namelv, the factor being investigated. Tims if one is attempting to ascertain the possible effect of dusty occupations on the incidence of pulmonary cancer, one must he sure that one deals with all those exposed to the same type of dust, on the one hand, and with those who do not meet that dust, on the other. Also, the distri butions of the groups compared must he the same as to age and sex. Third, those exposed to the dust must not he meeting some other factor which might also theoretically he productive of pulmonary cancer, for example tobacco smoke. Even these simple rules are seldom followed in ascertaining the role of any occupation in determining the onset of cancer of the lung. (a) Silicosis: Uecause of the iihrosis which occurs in the lungs ol persons alTccted with silicosis, it has been suggested that tins mav have a role to play in producing cancer of the lung. The changes induced by silicosis arc in the connective tissue rather than in the epithelium o; the bronchus, and so it is not probable that silica will prove to he carcinogenic. Silicosis was carefully looked for by Herhlinger :l1 in 5() cases of pulmonary cancer, and no evidence of it was found, although silicosis is very common in his district (Jena, (lennanv). That last clause changes the whole aspect of the finding. Had silicosis been rare in his district, his tailure to tmd it would have given no indication ol any possible lack ol relation between the two, H ill11 found no relation between silicosis .and cancer of the lung in Ins cases or in reviewing the literature. If the cases of pulmonary cancer are in a district where there is a high incidence of silicosis, we should find the two occurring together with some frequency; it" the eases are in a district where there is little silicosis, we should expect the two conditions to he associated seldom if thiv are not causally related. Thai the conclusions drawn from the incidence of these two diseases must he eareltiffv Tfffitrolled statistically is pointed out in an editorial in the British Medical Journal;'" in a discussion of the cases of Anderson and Dible."" These authors found a high silica content in 12 of their 70 cases of pulmonary cancer, as compared with 48 cases used .as controls, The editorial points out 34. ITrblimg-r, \Y.: Med. Klin. 27:1,137, 1931; abstracted, Atm J. Cancer 10: 839, 193.3. 35. Silicosis and Pulmonary Carcinoma, editorial, licit. M. J. 2:411, l'U8. 94 > . U\( 'HU '/'.'.S' oi: r .r n /o L o t ; ) that even if (heir conclusion is correct, that silica plays a role m causing pulmonary cancer, the conclusion is not justified lie' the material they had. It goes on 16 stay that such a conclusion would he justified only if there were a series of almost 3,(XX) controls, this figure being based on the incidence of primary cancer of the lung in the general population. Klotz :i" found pulmonary cancer in X per cent of his senes of 50 eases of silicosis, as contrasted with 1.17 per cent of 4,500 "unselected" autopsies wj^idi ^e used as controls. The latter group of autopsies were apparently on patients of 15 years and upward, as it did not include the autopsies of the pediatric service. The objection to conclusions based on such material is this: In using iinsclccled autopsies, perhaps to obtain a so-called random sample, the control material was most dissimilar from the eases of silicosis under review. Among the subjects of iinsclccled autopsies would he a large number of females, while the patients with silicosis would he entirely or almost entirely males. Cancer of the lung is most frequent in males; therefore one would expect to find more cases of cancer of the lung among patients with silicosis than among the subjects of autopsies in general merely on the basis of unequal sex distribution. Again, the patients with silicosis are tor the most part, in the middle age group or old age group, for it takes years for silicosis to develop, (lancer of the lung is most frequent alter 40. 4 he subjects ot unselected autopsies will include many jxiticnls in age groups well below 40, an age at which cancer of the lung is intrequenl. On the basis of age distribution we should expect to find more, cases of cancer of the lung among patients with silicosis than among imseiecfed subjects ot autopsies. Finally, there will he variations m the percentage of eases ol pulmonary cancer from group to group ol autopsv eases, even when these groups are standardized for age and sex; these variations arc attributable merely to the difference in samples of the population in each case. It may he that Klutz's conclusions are correct and that pulmonary caneer is more frequent among patients with silicosis than among a group similar in age and sex hut. without silicosis. Hut wc must reject his conclusions as unsound until he has compared his silicotic patients with a group of men of the same age distribution as the patients with silicosis and made due allowance for possible variations between different samples of the population. (b) Ashestosis: Few eases of asbestosis associated with primary cancer of the lung have been reported ( F.gbcrl and (leiger; Ifornig 3I). Here again we should expect to find the. two conditions associated in the same person in some cases, on chance distribution alone. Since .Vi. lrglin't. i). S., and twiner, A. J. : Am. Kev. Tulare. 34:14.?, 193(>. 37. Homin', F\ : /.tsclir. t. Krebstnrseh. 47:281, 1928. isin ihey. K k _ n the i 50 ted" were hide 'inns haps must Meets ' the i;l]cs, uuld with 1 the - are t akes Iu n i t limits ng is I find in on g n the i<>psv these II the unary mg a inject limits with Trent mi ary k 37) . ml in Since ,1/ A CK /. I ,\- M . I CK 1.1M--1'R l M A R Y C A R C I N O M A <>l; t . U X G `>43 $ A asliestnsis occurs in men of middle age, cancel' ot the lung may lie tound in these patients in a higher percentage than in the general population, but we have no proof that it is found in greater percentage when a group ot men of similar age from the general population is used as the standard. The criticism that was made with respect to Klutz's data on silicosis may he made also of the data of Lynch and Smiths'''1 Lynch and Smith have reported that in 2,.VIS consecutive necropsies over a period of twelve rears cancer of the long was found seven times, an incidence of 0.3 per cent. In the 35 cases of ashestosis, in some of which the lesions were very mild, cancer of the lung was found twice, an incidence of 5.7 per cent. Are these values significant, they asked, or is the sample of cases too small? before such figures can he ot any value, thev must select from their 2,3-13 consecutive necropsy records IaSpl orlv those on males, since their 35 cases of abestosis doubtless all relate to males. They must then sided from these records on males only those < on males in the age range of the males with ashestosis. When this is at done, it will he easier to decide whether the incidence ot pulmonary cancer is higher among patients with ashestosis than it is among those without ashestosis. ( (') J'neumonoeoniosis : Other dusts besides silica are said to cause cancer of the lung. There mav he much or little dust, depending on the occupations of the jiatients with cancer of the lung and on the district in which they live; rural jiatients will have little coal pigment, while residents of industrial cities like Pittsburgh will all have a great deal. South African miners have a high incidence of pneumonoconiosis hut a low incidence of cancer of the lung.- Yorwald and Karr examined the records of 15,587 persons who had long been exposed to dust}' atmospheres and found that only 3 of them had cancel' of the lung, lie stated that the dust must he carcinogenic before its presence will induce cancer of the lung. Arkin and Wagner and Berblinger :u both found pneumonoconiosis rare in their series of cases of pulmonary carcinoma. (</) 'barred Roads: Dusts and emanations from tarred roads have been said to he responsible for the increase of pulmonary cancer noted in the last two decades (Campbell'"), hut workers in Russia, where tarred roads arc few, have noted the same increase in cancer of the lung as has occurred in other countries, [lusted,17 in Denmark', found that jiatients with primary cancer of the lung'^ainf^in about the same pi ojiortion from areas where there were no tarred roads as from areas 3lA Lvncli, K. M., and Smith, \V. A.: Am, J. Cancer 3G:A>7, 1030, 30. Vorwald, A. J., and Karr, j. \ V . : Am. J. !\ith. 14:40, I'MS. 40. .Arkin, A., ant! Wagner, 1). II.: J. A. M. A. 10G:.W, 10,S. 41. Campbell, J. A.: I.ancet 1:333, 1034; J. Indnst. Mvsw A dnxieol. 19:440, 944 A K C H i n i S ()/' / '.4 7 7 /0 /.()(,) when.' tarring had been carried nut. lloyd" found no evidence that tarring h:ul played any role in his 14 cases. Valade *- pointed out that the conclusions which Campbell " reached as to the efficacy of this type of dust in causing cancel' of the lung were based on the reactions of mice, which produce a high number of spontaneous pulmonary cancers, and' that what might he carcinogenic for mice need not he so for men. We may conclude that there is no demonstrable proof at present that any form of dust lias any role in producing cancer of the lung, for it has not been demonstrated that pneumonoconiosis is present, in the lungs of patients with pulmonary cancer in any higher proportion than in a W4pj- t i ,, com'spoiti'lmggp'ou)) o) the genera! population. ( Hhcr agents, such as roentgen nil's, tobacco smoke and motor exhaust fumes, have, been suggested as possible causes. 'The suggestion that roentgen raws are a cause is almost absurd, for the probabilities are that only a small percentage of those in whom cancer of the lungs has developed ever had their chest under roentgen raws before they came lor diagnosis ot their cancer. .Motor mechanics working in garages where there is an abundance of exhaust fumes tire said to show no higher incidence ol pulmonary cancer than the general population, and Camp bell 1:1 staled that exhaust gases Irom internal combustion engines do not increase pulmonary cancer in mice. Tobacco smoke might be one factor in the increase of pulmonary cancer." Tile habit is one which permits long-continued application of the smoke to the bronchial mucosa; it is, or at least was, a habit more prevalent among men; there are two possible agents which might prove to be cancer producing, namely, beat and the derivatives of tobacco. There are those who point out that the increase in the consumption of tobacco parallels the increase in the incidence of pulmonary cancer. 'This may or may not have significance. It tobacco is a factor, it must be proved that deriva tives of tobacco are carcinogenic or that the repeated application of heated smoke is conducive to cancer. McXallv " stated that the tar in eigarct smoke contains enough chemical irritants to account for the increase in cancer of the lung. RotTo found among a group of 5,000 cancerous women 42 who exhibited cancer in what he called the "smoke stream," namely, the lips, tongue, jaws, larynx and pharynx. All of these women were heavy smokers, lie tound that the products of. burnt tobacco produced carcinoma when painted on rabbit ears, and lie teds that there are probably many aromatic substances in tobacco that are carcinogenic. 1lis hidings are suggestive. Hut the increase Id V a t . d r . \t. 1 1 : 521, 1929. o k a ! in in the 1.u i:u , 4 2 . C a m j i l i r l l , 1. A . : l>ril. J . 1*'x j h.t , P a t h , 1 7 : 1 4 6 , 1'Ufi, It. M c X a l l v , W . 1 ).. : A m , J . O u k -o - 1 6 : 1 ;;o_>, 1 922 4 5 . K o t i n , V . ,\. 1 1. : l ) ( ' i i K i ' h r m o l . W c l m Si'll !r. f.3 : 1A i 7 , A nnotatiniis lA v, i' ;i /,i c x u x -m x c x u x .. r h ' / M . i K V c x x a x o M . i o x i . r x c y-is in cancer of the lung should he nntch greater aim mg females than among males, since smoking lias increased among women far more rapidly in the last two decades than it has among m e n . Data p u b l i s h e d by the Metropolitan Life Insurance Company 10 .show that tint incidence is not greater among females. The rates among the industrial policyholders of 45 to 74 years of age show that in 1917 the rate for pulmonary cancer among females was 2.5 per hundred thousand. This rose to 8 per hundred thousand in ITCH, an increase of 220 per cent, 'i'he rate for cancer (if the lung in males rose in the same period from .4.2 to almost 24 per hundred thousand, an increase of more than 900 per cent. Their conclusion is that it appears doubtful that smoking is a factor m causing cancer of the lung. Controlled observations over a longer period are necessary lo settle this point. Investigators must not rest content with finding that a high percentage of patients with pulmonary cancer are tobacco smokers; they must find that the percentage of smokers in this group is sig nificantly higher than that in the general population of a similar age and sex distribution. When we review all the theories on the cause of pulmonary cancer, we find that not one supposed cause has been proved to lie a real cause and that none of (he alleged causes has been investigated with suffi cient statistical accuracy to enable the observer to pass an opinion carry ing any weight. No one cause, no sum of causes, no common factor in them, namely, chronic irritation through inflammation of the bron chial mucosa, has been demonstrated to occur with any greater frequency in patients with cancer of the lung than in a group from the general population of comparable age and sex distribution at a time sufficiently long before the onset of the cancer that it may he interpreted as ante dating this rather than following it. Naturally, any condition which is a sequela of cancer of the lung may be found more frequently in patients with pulmonary cancer than in persons not affected with this type of cancer. Such a condition when obviously the. result of cancer of the lung must he excluded from a discussion of the latter's causes. (r) Schticcberg Miners: The high frequency of cancer of the lung among the miners of the. Kehneebcrg and Joaehimsthal valleys has been recognized for manv years. It seems that this occupation is one which does enhance the incidence of pulmonary cancer beyoij^ that Known elsewhere and apparently beyond that of the neighbors of iTie vallcv.s who do not work in the mines. Thus I'irc.han and Sikl 467 stated that 12 of 19 Schnccbcrg miners dying in 1929-1930 had cancer of the lung at autopsy. Of 12 miners dying in 1932, 9 had a clinical diagnosis of 46. Cancer of the Lung a Crowing Health Problem, Statist. Hull. Atrlmp. Life Ins. Co. 20:7, 193T 47. Pirchan, A., and Sikl, It.: Am. J. Cancer 16:<>S1. l'l.H 94o . i K C i i w a s (>! r . v n i o i . n i . Y pulmonary cancer. Sikl reported similar figures for (lie minors of Joachimsi lial. ( )i 15 miners dying in eighteen months, autopsies were canoed out on 10; of these, 8 had carcinoma of the lung. The average age ot these men was 49, and the average period of exposure in the mines was seventeen years. '1'his incidence of cancer of the lung has not hern equaled anywhere else in the world. The cohalt, the arsenic, the dust, the radioactive emanations till have hern incriminated. Sikl ,s staled lhaCft-'vlattnical analysis of the lungs showed no trace of important lorcign elements. The. test for radioactivity was negative, although he. felt that radioactive emanations in the mines may have a part in producing pulmonary cancer. Maisin 1:1 quoted de J.act as saying that in I'clgium pulmonary cancer has not developed in workers in radium although they wear no masks and are covered with and breathe the dust of the works all the time. I lowever. the plant had been in operation only twelve years at the time of de I.act's report. ()( interest in this connection are 51artland's r,u findings in workers with radium paint. The girls who died of osteogenic sarcoma were exhal ing radioactive air, as shown bv tests, vet mine of them presented cancer of the lung. A whole series of questions to which we have as vet no answers, and may never have, arise. 11 the radioactive emanations in the air in the Sclmceberg mines are responsible for cancer of the lung, why did the radium pamt workers m the United Stall's die. of bone and not lung tumors, especially as it was shown thru the lungs were exposed to radioactive emanations.'' May it he that there is an inherited factor lor cancer oi the lung in the Schneeherg miners which has been increased through long-continued inbreeding in these rather isolated mountain 'alleys with a small population and that the. carcinogenic agent in the mines (whatever it may he) enhances an already present tendency toward pulmonary cancer? Is it possible that, taken in small doses, the radioactive substance induces pulmonary cancer, which may have a long latent period, whereas radium taken in presumably larger quantities, as is the ease when radium paint is swallowed, may induce hone tumors, with a shorter latent period of development ? Thus these, victims would die of osteogenic sarcoma before they had time to develop pulmonary, cancers. Certain it is that although the miners have a very high incidence of cancer of the lung they do not have a high incidence of those chronic pulmonary conditions which are suppised to predispose to cancer of the lung. Thus I'irchan and Sikl 17 said that of the 9 men with clin ically diagnosed cancer of the lung, not one. had a history of chronic pulmonary disease that was not directly related to the onset of the cancer.48950 48. Sikl, it.: Ztschr. 1. Krchsiorseh. 32:609, 1930. 49. Maisin, Tamer, Bruxelles 11:111, 1934; abstracted, Am. J, Cancer 8: 164, 1936. 50. Martlaml, II. S . : Am. J. Cancer 15:2435, 1931. ' of . vrc a.^C I he 1ins nic, ant ugh in hat aim lust ion UTS lalwer no in 1'r-' nul M'd tor wd air, 1lu ll'}' tw , ' a It'S, irs, .ild <ry U*(* nic of tu nic Ik; '.8: Unfit ; |l ill: >;f| , *V! 4 E wirff /If I'Killt: WfciSfT uhlM-a > u T>I? y\ h'i'&'.A ',sll p v^f ' & i $ M .-ICKLI N ~ M , !C K U X -- P R 1 M ; I K Y C A K C I X O M A OF J A ' X O 947 Throughout this discussion we have reixsitedly emphasized that it: makes no difference how great is the percent,age of patients with cancer of the lung who point to any particular disease in their past history, provided that a similarly great: percentage, of the general population has had tlie same disease. Thus if the cancerous patients come from a region in which the whole population has had measles, every patient with cancer of the lung will give a history of measles. If 90 per cent of the population past 40 have had tuberculosis, 90 per cent of the patients with cancer of the lung will probably give a history of tuberculosis. If the main occupation in the district is farming, most of the patients with cancer of the lung will he. fanners, hut if mining is the main industry, most of the patients with cancer of the lung will he miners. This fact ha been little appreciated by those who argue that because 25 or 50 or 75 per cent of patients with cancer of the lung give a history of a particular disease that disease must be carcinogenic; hence we shall try to present this matter more in detail at this point. In the accompanying chart the first rectangle represents a theoretic population. One fourteenth of them (black bar at the top') is charac terized by having a certain disease, X. The subsequent rectangles represent those, portions of the. population which have had some other eond lion. Tims' in rectangle 2 we might indicate those who had had, for example, chronic bronchitis; in 5 those who had had influenza; in 4, those with arteriosclerosis, and so on. We will assume that none of these, groups overlap. Now it will he seen that one fourteenth of the population in rectangle 2 also has disease X, the same proportion as was present in the, general population, 'fins is true of the groups represented in rectangles 2 to O. The group in 7 shows a much higher concentration of X patients than do the other groups, while the group in 8 has fewer X patients than has the general population, from such a diagram one would he justified in looking on the group in 7 as probably peculiarly susceptible to disease X, while those, in 8 would he. considered relatively immune. We. are assuming in this example that all the. groups are comparable as far as sex, age and other signi ficant factors are concerned. Now let us arrange these same data in rectangle 9, which represents [lie population affected with X, which is one fdffrtr$i?ih of the total population. .Mere we see that three groups each have a fourth of the patients with disease X. These are 2, 5 and 7. (iroup 4 has an eighth; group 5, one sixteenth, and groups 0 and 8, each a thirty second of all X patients. Jl we looked at this column without viewing the data in columns 1 to 8, we should he very likely to conclude that groups 2, 5 and 7 arc. equally susceptible to X and that groups 6 and 8 arc the most immune to this disease. The latter conclusions would, of course, lie wrong, as the data in the rest of the diagram clearly show. 948 A R C H irU S OF PATHOLOGY It is in tlic latter form that the data dealing with the cause uf pul monary cancer arev usually presented. We are told that a certain per centage ol patients with cancer of the lung give a history of chronic bronchitis or of intluenza or of some other condition, and if this per centage is high, it is assumed that the disease named is an important 1x -it*1* v z 3 4 567 BBSS SS3S] ill E 5 S 9 i-.-tu Ir A : ...7. RiUBS- * 1. TIk- general population is represented in me (an?; In 1. Those with a history of a disease which \vc will rail X an: shown by the Mack bar at tin: top. This k one fourteenth ni the total area of rectangle 1. All pcrmilrm'S in tins figure arc <4 course theoretic anti an: used n im 'lv for purposes of demonstration. 2. '1 his rectangle represents the proportion ol the total jHipulatioii, namely', 2a per eeut, whieh lias had chronic bronchitis. Xote that one fourteenth of this group also have X. .1. d his rectangle represents those in the population, namely, 2a per cent, who have, had influenza. One fourteenth of them have X. 4. 'This rectangle represents those in the population who have arteriosclerosis, namely, 12,5 per cent, One fourteenth of them have X. 5. Tins ivpnsents the part of the population which has asthma, namely, 6.25 per cent. One fourteenth of theta have X. (). This represents the part of the populated with a history of war gassing, namely, 2.12 per cent. One fourteenth of them have X. because. X has developed in the five groups 2 to 0 in the same propor'hm as in the general population, we cannot assign a specific role to any of these conditions in the production of disease X. 7, This is the part of the population who are miners of radioactive ores, namely, 3.12 per cent. Mere the incidence of X is eight times as great, as that found in the general population or in the live preceding groups of the population. It api>ears that some factor associated with mining of these ores is productive of X. But because not all miners have X, and because many who arc not miners do have X. other factors besides the mining of these ores enter into the cause of X. 8. 'ibis is tlie part of the population which has a history of tuberculosis, namely, 25 per cent. The incidence of X is much lower here than in the general population, as if some factor inhibiting' the development of X was present in this group. 6. This rectangle shows the part of the population in which X develops. One fourth of these leave had bronchitis, another fourth influenza, another fourth ex|M).sure to radioactive ores. One eighth have arteriosclerosis, one sixteenth have asthma, one thirty-second have had tuberculosis, and one thirty-second have been war-gassed. Mata in the latter form are usually vised to determine the importance ol any specific disease as a cause of cancer of the lung. It is assumed that all the groups in the rectangles 1 to 9 are of the same age and sex distribution and that no group in any rectangle overlaps that in any other. U'it-Uffi'UfTr.-'v'* S'TK'.i'at jlf A C X L I N - M . ICK 1.1A'-- P KIM A K Y CA h' CI XOMA O f W . X O 949 causative iact or in producing carcinoma of the lung. If the percentage is low, it is assumed that the disease has no significance in the causation of c a n c c i: of t h e l u n g . T h e d i a g r a m m f i g u r e 1 s h o w s t h a t it is o n ly when the disease in question is found in a much higher percentage of the group with pulmonary cancer than of the general population that one can consider that it has an cliologic role in producing pulmonary cancer. But even this is not enough. It is necessary that the group in the general population which we are using as a control for the pulmonary cancer population he comparable to the hitter in all respects except that the general group does not have cancer of the lung; otherwise one might find that the group with cancer of the lung had the condition in question in a higher percentage than had the general population without there being any significance in that fact. For example, some of the veterans of the World War are asking for compensation for cancer of the lung, which they claim was brought on by the chronic pulmonarv diseases which they contracted through exposure encountered during military service, Now how might one go about settling the justice of such a claim? One might say, "The incidence of pulmonary cancer among war veterans is much higher than among the general population, there fore it must have been the hardships of military service which caused this increase.'' ( We are assuming for the sake of the argument that it is more frequent among war veterans.) Hut the war veterans are all males, and the general population holds more lemalcs ihan males. Hence an adjustment for sex would have to he made so as to make both groups comparable. Next, the average age ot the war veterans lies between '-'-5 and if), while the age of the general population will range all the way from birth to 100 years. The two groups h u m he made comparable as far as age is concerned. Then one might argue that although the hardships of military service did not actually came pul monary. cancers they caused these lesions to develop much earlier, because the age at which the cancers were developing in the veterans was much younger than that of the patients with such cancers in the general population. But again the age distributions in the two groups are difterent, the ages of the veterans being concentrated in the range from 45 to 60, while the general population has those under 45, among whom pulmonary cancer is not frequent, and those from 60 on, among whom it is irequent. Hence only those among the vc^eiTf/fs who were destined :o get cancer of the lung at younger ages would he included in the veteran group. Those iti whom it will develop after <>() are not included, since the veterans have not yet reached 60 wars and over. Jn this way the average age at which pulmonary cancer devel oped in the veterans would he less than that at which it develops in the general population. 950 A R C l I I I 'l i S O R P A T U O I D C Y St'ProSKl) KXVI RON MKNTAL FACTORS : A C.KNKKAl. 1)1S( I'SSTON The introduction o i`^Consideration of experimental work into a dis cussion of carcinoma of the-human lung may seem irrelevant, but it is not really so. For a few decades the experimental production of car cinoma by so-called chronic irritants lent strong confirmation to the idea that chronic irritation was the sine qua non of all cancer. Thus when no evident chronic ir ri tant was found, it was as sumed that it was An* . . present anyway, as witness the citations at. the. begi nning of the pa[XT. The evident irritants were mechanical, thermal, chemical and bactcrio- logic agents. Sometimes it was stated that it was not the irritant but the chronic inllammatory changes set up bv it that produced the cancer. In recent times, however, we have come to look on carcinogens as involving more specific agents. Certain viruses, such as that of Rous, have been shown to he the cause of certain definite animal tumors. Specific chemical bodies, such as dihenr.authracene and the estrogens, have been shown to produce carcinoma tin repeated application. Much depends, apparentlv, on the chemical formula; comparatively minor alter ations in the structure of the compound mean wide alterations in the carcinogenic power. A large literature has grown up around these carcinogens (Cook/'1 A n d e r v o n ta n d others). It lias been found experimentally that not. all chronic irritants pro duce cancer and that pome of the. highly carcinogenic agents are not irritating, that what may he carcinogenic for one animal need not he for another animal or even in another strain of the same animal, and finally that what is carcinogenic for one organ in an animal need not he productive ol cancer in another organ oi the same animal. Let us illustrate these points briefly. Some marked irritants, such as mustard gas, are not cancer producing although the}- cause intense inflammatory reactions (Visser and ten Seldam r,:1) . Again some highly carcinogenic chemicals are nonirritating when applied to the skin f Andervont r,; ). 1ar painted on the skin of rats did not: produce cancer of the skin, while if painted on mice or rabbits it did cause such cancer. The same tar painted on mice differing in their genetic susceptibility produced cancer oi the skin in highly susceptible mice and failed to produce cancer in highly resistant mice ( Krevberg r`l). It is of interest 51. Cook, J. \ \ .: Proe. Hoy, Soc., London, s.lt 111:485, 1932. 52. Andervont, It. P . : Ufiect of bacterial Products on Growth of Malignant Tumors, in A Symposium on Cancer, Madison, Wis., University of Wisconsin Press, 19,18, p. 54. 53. Yisser, J., and ten Seldam, P . F. J . : Geuccsk. tijdsdir. v. Ncderl.-Indie 77:3092, 1937. 54. Kreybcrg, L .: Genetic and Constitutional Aspects of Spontaneous and Induced Tumors, in A Symposium on Cancer, Madison, Wis., University of Wis consin Press, 1938, p. 183. M A CK L I X - M A C K LI .V--/ '/ AM Ah'}' CA RC1XOMA (>/ L U S C 951 11ml cotton mule spinners suffer fmm cutaneous cancer due to tin: conslant spraying of the skin with liquid petrolatum, whereas they do not exhibit cancer of the lung in :mv increased percentage although they ennstantly inhale air in which the same oil is present in the hum ol line sprav {Simpson ). Chimney sweepers had a high percentage of cancer of the skin, vet the constant inhalation of an atmosphere con taining the same carbon particles as caused cutaneous cancer did not lead to cancer of tire lung. Injections of estrogens are productive ol cancer ol the breast and of the uterus in mice lint not ol cancer ol the gastrointestinal tract or of the lung. Therefore it can he stated that: before any condition, such as chronic pulmonary disease, can he regarded as producing cancer it must he proved beyond doubt, that (1) the condition is capable of producing cancer apart from any ability to cause chronic inflammatory change, ( 2) that it is callable of producing cancer in the animal under discussion and (, that it is capable of producing cancer in the sped lie organ in question. fMerclv to show that a condition induces chronic inflamma tory changes in an organ is no longer sufficient, for it is known that chronic inflammatory changes can exist for long periods without pro ducing cancer and also that cancer Tail occur with no reasonable evi dence of preceding inflammation----for instance, in retinal hlastomu shortly after birth. ckohaiu.k koi.e ok i i kkki hty ) n )'t;i..\i onary c ancf .k Although specific carcinogens can produce cancerous changes, are the conditions mentioned earlier in the paper shown to lie carcinogenic f It has been assumed that they may he because one has no other expla nation of pulmonary cancer if one insists on the presence of extrinsic 1actors as the sole causative agents of all cancers. lint cancer is a problem of failure ol differentiation, ol organization and 'of growth; differentiation, organization and growth are problems ol the inherent makeup of the cell itself and of its ability to utilize certain chemicals in its metabolic processes. Therefore one should not overlook the fact that "idiopathic" alterations in the inherent makeup of the cell may he responsible for cancerous changes without the intervention of any carcinogenic agent. In other words, the cancer mav arise through hereditary causes. It is known that under the inifuellf? of hereditary determiners descendant cells of the original zygote differentiate into cells with quite widely dithering potentialities. Absence or alteration of these determiners for normal differentiation may also he hereditary. Thus the alterations necessary for changing a normal cell into a cancer cell may he under the aegis of determiners located in the. germ plasm. 55. Simpson, S. T..: Quart. J. Mod. 2 2 : 413, !<>*>. 952 a r c h i v e s or r a t i i o l o c v This alteration sui generis may conic about through two dilTcrent mechanisms. (hup may he a somatic mutation, i. e., a sudden departure, from the normal line of cells from which the altered cell sprang, which departure then becomes hereditary for all cells descended from that original mutating cell. That such a mutation should occur would he wholly unexpected, since there was no factor in the original fertilized germ cell from which the individual came which played a vole in deter mining lingeritiy'ence of the mutation. The patient with cancer of the lung in this case would not have come from a strain in which factors in cancel' of the lung were inherited, nor would lie he liable to have more offspring in whom cancer of the lung would develop than a person without cancer of the lung. The second mechanism would he one in which the tendency to alter , into a cancerous cell would have been inherited, being due to factors resident in the. fertilized germ cell, just as then' arc inherited factors which guide the destiny of the cntodermal cells which form gut, then bile duct, then liver, so that this development proceeds normally in an orderly fashion at definite, times in the life of the embrvo, the times dillcring depending on whether the embryo conics from a mouse, a dog or a human ovum, and just as there are inherited factors which permit cartilaginous development at the. epiphysial lines for so many rears in the human humerus but for a different number of years in the femur helore the cartilage cells cease proliferating ami allow hone growth to cease, so there may he inherited factors which permit cells in a specific location at a specific time in the life history of the individual to alter and become, cancer cells. What proof have ivc that these last (wo mechanisms are possible and that not all cancer nor indeed all pulmonary cancer needs be inter preted as arising through the agency of extrinsic carcinogenic factors? hirst, with the single exception of the pulmonary cancer of the Schnecberg and Joaeliimsthal miners, we have no adequate proof that cancel of the lung in man is caused hv extrinsic factors. This is negative evidence, hut it should he of value in forcing us to hunt for other causes of human pulmonary cancer before we accept all pulmonsn cancer as being due to external agents, b'or the idea that sporadic somatic mutations or hereditary alterations are frequently the cause of cancer ot the human lung we have the following indirect additional evidence. '1 he vast majority of such cancers in man appear to rise from one point of origin, as would be expected were a somatic mutation the cause, whereas if external agents acting on the lung were responsible we should expect that the tumors would he multiple and over both lungs. Thus tumors induced in mice by hydrocarbons are not bron- . chiogenic as in man, but suhplcural, and may be multiple rather than .1/. U ' K J J S - M . I C K I J X --PRIM. IKY CA RCI NOM. I <>i I l ' S C <>\\ single, and may be bilateral. Second, we know that mutations do occur in the germ cells and that the cells with such mutations are then propagated as a new line of individuals. There is no reason to suppose that mutations are limited to germ cells alone and that such mutations could not occur in any somatic cell in the body. Have we any evidence that definite hereditary factors may he present in the germ ceil which cause cancer of the lung to appear at a definite time in the life cycle of the individual? As far as cancer of the lung is concerned we have no human pedigrees that would prove it to be inherited, and again the evidence must he indirect. Cancerous altera tions of cells have, been found to he inherited in strains of mammals in which experimentation is possible. The type of cancer produced by the strain is more or less constant. Thus there arc strains with a high incidence of pulmonary tumors (Lynch50), others with a high inci dence of mammary tumors and others with a high incidence of hepatic tumors. There are strains of mice in which several types of tumors may he capable of developing hut which tend to die of the tvpc devel oping tirst, thus obscuring the oilier cancerous potentialities. Since man biologically is much more closely identified with other mammals than he is differentiated jrom them, it would he reasonable to suppose that it cancer shows dependence on hereditary factors in mice the same will he true for man, although the actual mode of inheritance may differ widely in the two species. in rectal, gastric, mammary and uterine cancer in man wc see good evidence that related members ol a family tend to resemble each other more closely in the production of these tumors at a definhc time in their liie than do unrelated members of a community ( Macklin 07,1V This suggests that hereditary factors play a predominant vile m many if not in all of these tumors. Cancer of the lung was seldom recog nized, before 1900. Hence, even if it were inherited and transmitted according to a simple scheme so that one would frequently encounter familial cases, we should not have many records of familial instances, owing to the fact that pulmonary cancers in the preceding generations remained undiagnosed. Finally, wc may use the evidence furnished by tumors in twins ( .Macklin True we have no records of pulmonary cancers in twins, hut since cancer ot the lung is like all other canc$rs'r*Ka'pt that it occurs in the lungs instead of elsewhere, we may not he assuming too much when we use tumors in twins to show that cancer and hence pulmonary cancer may he inherited. If one examines records of can- 56. T.vncli, C. J . : J. F .x i x t . Med. 54:747, 10,11. 57. M addin, M. T . : (a) Familial 'Incidence of Cancer, in a Symposium on Cancer, Madison, Wis., University of Wisconsin Press, 1938, j>. 32; (1C 1. 1 [eredity a s :277, 1940. y 54 a r c h i v e s o v i ', v n i o i . o c ) uts in monozygotic and dizygotic twins, one finds the following facts: Monozygotic twins, with an identical inheritance, resemble each other lar more often-in botlgliaving a cancer and having it in the same organ .and at about the same -time of life than do dizygotic twins, who may have with respect to cancer either an identical or a different heredity. This is true even when the dizygotic twins studied are all of the same sex, thus eliminating differences in environment due to occupation which are present when the twins are of the opposite sex. This would indie,ate that $Wio3fch somatic imitations may be responsible for some cancers they ,are probably not responsible for all, for it would he very rare tor identical twins to have an identical somatic mutation in the same organ at about the same age were that mutation not under the guiding inthienee of heredity. To sum up, then, it would seem that the so-etdled chronic irritants .supposed to he responsible for cancer of the lung have not been proved to lie causes of such cancer, since (1) many have cancer of the lung who are not exposed to these chronic irritants, while ('2) man}- exposed to them never have cancer of the lung. I 'ntil such proof is forthcoming, investigators are closing their minds to other, perhaps fruitful possi bilities h\- insisting that because specific carcinogens are responsible for cancer ol (he lung in some instances thev must he in all. Such an atti tude is no more reasonable than to state that because diabetes develops in an animal from which the pancreas has been removed, all diabetics are without a pancreas. Sporadic mutations and hereditary influences may and most probably do plav a large role in causing pulmonary cancers as well as other types of malignant growth. SUMMARY In conclusion, then, we may say that none of the specific diseases or conditions causing chronic inflammation of the lungs can he said to have been proved to he causes of cancer of the lung. They have not been shown to he more common in patients with pulmonary cancer than in a group ol the general population of similar age and sex, and until a .wi/iiifit'unf dilierence can he found between their incidence in those with pulmonary cancer as compared with those without pulmo nary cancer their causal relation to this type of cancer will remain purely speculative. This, of course, refers to chronic inflammatory diseases which have antedated the cancer by a sufficient number of years so that they can he excluded from being interpreted as the result, not the cause, of the cancer. C.hrouic irritation, especially in the form of chronic inflammatory disease, as a cause of pulmonary cancer has been assumed: the idea has been copied from text to text, repeated from author to author, with in.?''- "T-7YreTtr `^Ty'7'fuyjY-'r M.H K U X - M . U ' K I IX O K IM .IK V ( I/V( 7,\ <),\l. I Oh 1.0.Xu little critical analysis ot ils possible role. Somatic mutations and heredi tary factors must be considered and their probable vole in causing cancer of the lung recognized, I rue carcinogenic agents for the pm duction oI e;mccr of the lung will 11<i cl<Ubt Ik - found as modern industry expands Us use of chemicals, lint, as some types of cancer will develop in one seventh of men oyer -10 and since about one fifth of all can cers in nudes are in the lungs, one thirty-fifth, or about d per cent, of all men oyer 40 will die of this type of cancer. 'There may be an increase ri the incidence of cancer of the lung in .any industry beyond the maximal rate for cancer of the lung in the general population, hut one can never state that any patient, would not have, had cancer of the lung had he not been in that occupation. 'Thus even with the presence of carcinogenic agents admitted, their role in causing cancer of the lung cannot he proved in any individual ease, nor can they he said to be 100 per cent effective, since a certain percentage of persons exposed to the carcinogenic agents would have had pulmonarv carcinoma even without this exposure. The effectiveness of any carcinogenic agent in producing carcinoma oi the lung can he judged only by the increment of increase of carcinoma ol the lung in large group's, not by ils loin! incidence. It is thcrctorc impossible to assign to anv extrinsic factor an absolute role in carcinogenesis in any individual case. uI i* A " & sy ' nti ' H .tt,y 0 4 / //, ./i/y