·Publication Number 277
AMERICAN LECTURE SERIES
A Afonograph In
AMERICAN LECTURES IN CHEST DISEASES
Edited by
3 ARTHUR MYERS MD PhD FACP
Professor of Medicine and Public Health
University of Minnesota Medical and Graduate Schools
Alinneapolis Minnesota

·CAUSAL FACTORS IN
Cancer of the Lung
By
CARL V WELLER, MS, MD
Chairman of the Department of Pathology
and Professor of Pathology in the
Medical School, and Pathologut
to the Unilersity Hospital
Unicersity of Michigan
Ann Arbor Michigan
The Beaumont Lecture
Wayne County Medical Society
February 7, 1955
CHARLES C THOVIAS
Springfield Illinois
PUBLISHER
USA

·CHARLES C THOMAS PUBLISHER
BANNERSTOVE HOUSE
301-327 East Lawrence Avenue Springfield Illinois USA
Published simultaneously in the British Commonte calth of Nations by
BLACKWELL SCIENTIFIC PUBLICATIONS LTD OXFOND ENGLAND
Published simultaneously in Canada by
TIIE RYEISOY PRESS TORONTO
This monograph is protected by copyright No
part of it may be reproduced in any manner
without written permission from the publisher
Copyright 1956 by CHARLES C THOMAS
PUBLISHER
Library of Congress Catalog Card Number 55 11255
Printed in the United States of America

·ACKNOWLEDGMENTS TREDIT for the substance of this Beaumont Lecture must be given to the hundreds of clinical and experi mental investigators who for 500 years have sought to discover the etiology and pathogenesis of what is now known as bronchogenic carcinom? Reference will be made specifically to many published worhs and it is hoped that to some degree the usefulness of thus monograph will depend upon its reasonably complete bibliographic docu mentation It has been found impossible to present 1 complete surcy of the literature within the compass of one lecture cien though hberal erpansion has been per mitted in the printed version Those writers who fail to find specific reference to their papers may be assured that the omission does not imply that their contributions were considered of lesser importance than those to which ref crence could be made The author is responsible for the plan of organization and presentation lle rlone must bear the onus for er cerpts from the literature which may seem too brıcf when encountered out of context and also for the many omis sions The busic biologic philosophy of neoplasia presented herc is by no means onginal It his deeloped gradually over many years and cannot be attributed to any one source It has been a fundamental concept in the tezching of the Department of Pathology of the Unnersity of Michigan for more than forty years I one personal acknowledgment to the professional members of my Staff who live solunteered to assume didactic and practical responsibilities n luch ordinarily arc minc Without such rchef this lecture could not have been preparel Dorothy C Seifcricın departmental crocutive ·Vi CANCER OF THE LUNG CAUSAL FACTORS secretary and editorial assistant, has maintained thessential liaison with the Medical Library, has assisted 1 establishing the format, and has assumed personal respon sibility for the typescript CV W ·CONTENTS Poge V 3 5 9 03 ACKNOWLEDGVIENTS Chapter I INTRODUCTION N GENERAL CONSIDUTOns Incidence Etiologic Factors in Neoplasia Intrinsic and Extrinsic LIL TIC FIRST TUESE BERGKRANKHEIT IN TITE ErCZCEBAGE-SANEEBERG AND JOACHINIS TIUL (JACITYVOV) Chalcosis with Special Reference to Sil.cosas Cobalt, Nickel and Bismuth Arsenic Jonizing Radiations Schneeberg and Jáchymor on Retrospect IV Ture SECOND THEFTIRE SCANCY FOR THE CALSES OF INCREASING PULMOVARY CANCER IN TIL TUENTITH CENTURY Inlective Pneumonitis Increased Libility to Pulmonary Cancer an Occupational Groups Foundry Workers and Metal Gunders Chromate Hosiers Asbestosis Ihistorical Interlude Percull Po to V Mayneord Gas Workers Atmosphenc Pollution Arsenic Ionizing Radiation 37 40 41 43 46 48 49 50 52 52 53 56 62 63 Vat ·VII CANCER OF THE LUNG CAUSAL FACTORS Page 72 73 76 76 77 81 85 85 87 Chapter Benzpyrene Retrospect V THE TIMRD TIENE TOBACCO OY TRIAL Guit BY ASSOCIATION Casuistic Method Comparative Afethod Prospective Method Radioactivity of Tobacco Smoke and Ash Arsenic in Tobacco and Tobacco Smoke Tobacco Tar Difficulty in Determuning Carcinogens by Animal Experimentation Chronic Inflammatory or Dystrophic Changes Co carcinogenesis Importance of the Lag Period Lung Cancer Among Non smokers Comparative Risk of Developing Cancer of the Lung for Smokers and Non smokers Bronchogenic Carcinoma at the Mid Century Statistically Established Associations VI EPICRISIS VII REFERENCES INDEX 89 90 92 92 02 95 99 101 109 ·CAUSAL FACTORS IN CANCER OF THE LUNG

·No text on this page.

·I INTRODUCTION WISH TO EIPRESS my deep appreciation of the invitation of the Wayne County Medical Society to give the Beaumont Lecture for 1955 When I read the list of dis tinguished teachers clinicians and investagators who have had this privilege in the past thurty four years my sense of appreciation deepened but at the same time I ques tioned my worthiness and my ability to approach the pat tern of crcellence which has been established by my predecessors In a way my topic was chosen for me since the invita tion was accompanied by a suggestion that I speak on cancer of the lung Obviously some limitation was rem qured since only a restricted field could be crplored and developed in one lecture That field was determined by those current circumstances which have created an in tense interest in possible causal factors in bronchogenuc carcinoma we have a malignint neoplasm of which the incidence has inercased at an alarming rate within the lifetime of most of us The intrinsic qualities of human protoplasm cannot change that rapidly The evidence points to environmental factors I propose to crplore with you the causes intrinsic and extrinsic proved and sus pected of bronchogenic cancer In pursuing this purpose I shall draw upon medical history for clinical observations which have contributed to present knowledge Tłuis fusion of the old and the new is fully in heeping with the tradı tions of the Berumont Lectures and tomorrow today will be history I shull endeavor to establish a general biologic 3 ·CANCER OR THE LUNC CAUSAL FACTORS concept of causation in neoplasia and in particular of bronchogenic carcinoma, to assist in cultivating sound methods of evaluating causation, to present controversial matters impartially and to stress the fact that to the ex tent that an environmental factor can be established as the cause of any cancer to the same extent that cancer be comes a preventable disease That productive investigator of the environmental causes of cancer WC Hueper, now at the National Cancer Institute has reminded me that I wrote in 1929 ? No other form of neoplastic disease is more intriguing from the standpoint of incidence than primary carcinoma of the lungs for withm a generation it appears to have become one of the common forms of malignant disease instead of the rarsty which it was believed to be at the be ginning of the century In the twenty six years since 1929 the gross incidence of pulmonary cancer has continued to increase and so has interest in the disease itself but more particularly in sus pected causes The accumulated statistical information appears to implicate the smoking of tobacco as especially important This pleasant habit has long been under sus picion but in a rather detached way and at an acadernic level With the present emphasis this question has be come a matter of personal concern to many millions There is also the possibility of a dislocating impact upon a segment of the national economy which extends from the plantation to the retail outlet It is natural that, in der these stresses sound judgment yields to partisanship opponents and proponents use the same evidence to prove their respective views and lay editorial writers instruct the public ·II GENERAL CONSIDERATIONS OR ALL PRESENT PURPOSES cancer of the lung is lung are rare The small round celled sarcoma of earlier days is properly identified as carcinoma with cells of a low level of differentiation It is true that there is an Occasional example of a highly malignant neoplasm which has been variously designated as terminal bronchiolar carcinoma or alveolar cell cancer of the lung This highly malignant entity is represented in the Michigan necropsy matenal to an extent of less than one per cent It has cer tain structural resemblances to jagziekte a pulmonary disease of sheep in South Africa and Iceland This neo plasm is usually widely disseminated in both lungs in a manner wluch suggests acrogenous metastasis but which could be explained equally well if the causal agent is an air bome virus In my opinion this entity is an entirely different disease from bronchogenic carcinoma and it will not be included in further considerntion of the causes of pulmonary cancer Those new growths which are called bronchul adenomas likewise must be excluded from the general group of bronchogenic cancers This is done with some reluctance since bronchal "adenomas” may infiltrate locally and be found in adjacent lymph nodes However they differ so strikingly in age and ser incidence in potentialıtı for growth and in prognosis that only confusion would result of an attempt were made to include them in a discussion of etiologic factors for the highly malignant bronchogenic carcinomas 5 ·6 CANCER OF THE LUNG CAUSAL FACTORS Bronchogenic carcinoma can arise in any portion of a lung but large bronchi are more commonly insolved than the smaller branches Thus it comes about that there is a gradient of decreasing frequency from the hilar struc tures, as represented by the main stem bronchu and branches of the first order, toward the periphery of the kung This is reflected in the presenting signs and symp toms of the disease which are frequently those of a medi astinal tumor *cough pain in the chest, sputum, usually bloody or blaod streaked at some time dyspnea cyano sis, venous and lymphatic engorgement of the upper part of the body Laryngeal paralysis, hoarseness pupillary mequality, nequality of the radial pulses abdominal pain, dysphaga, and steytor in a second group the patent is led to seek medıcıl advice because of general systemic effects such as loss of weight weakness chills and fever, anorexia, or nausea and vomiting A third group is established for those patients in whom there is early involvement of the pleura with pleural pain and a hemorrhagic pleural effusion Unfortunately, in some patients the presenting signs and symptoms may be pro voked only by metastatic neoplasm and thus a fourth group must be established Many patients in this group will give evidence of a lesion of the central nervous system either enceplialıc or spinal The histologic exam nation of a tumor from the brain may give the first presumptive evidence of a bronchogenic carcinoma Farbere found that 34 per cent of the patients included in his very large study group first sought medical atten tion because of symptoms which were due wholly or u part to metastatic lesions Insghts* discovered that pul monary cancer was the most common source of cerebra) metastases in male patients (29 per cent) whereas mammary cancer took its place in female patients (47 per ·GENERAL CONSIDERATIONS 7 cent) These figures were derived from 94 cases found in 6500 necropsies at the Henry Ford Hospital Detroit Because there is some suggestive information by no means clear cut as yet that there may be significant corre latrons between environmental causal factors and type of cell in bronchogenic carcinoma brief reference must be made to histologic features Regardless of reputed origin in large medium or small bronchi or even in mucosal glands three hustologic types may be encountered These are designated as squamous celled gland celled and un differentiated celled Although these terms might seem to be mutually exclusive no sharp lines can be drawn between groups Not only my great variation in cell type be found in different portions of the same neoplasm but there is a large group in any collected series in which ncoplasm cells are in an intermediate level of differentia tion Usually the squamous celled carcinomas predomi nate Fırlicr så figures recomputed on the basis of 849 necropsied patients of known cell type give 373 per cent squrmous 280 per cent gland celled and 346 per cent with undifferentiated cells Sonic years ago I found that at facilitated presentation of the relationship between the various cell types of bron chogenic cancer if they are concened as arranged in a shaped pattem (Figure 1) At the intersection of the stem of the 1 and sts branches I locate the neoplasms retaining the level of poorly differentiated "basal cells of normal bronchial mucosa Along the left hand inchned Imb of the \ are distributed the arcinomas of gland celled type with the most fully differentiated columnat celled papillomatous mucın forming adenocarcinomas at the end of the line and less elaborately differentiated forms in descending pattem Tiose cancers with ccils resembling tlie normal columnar broncul cpithelium ·8 CANCER OF THE LUNG CAUSAL FACTORS occupy about the position marked with x Similarly, at the extremity of the right hand limb are placed the heav ily comfying squamous celled carcinomas, followed by those with little or no keratonyalın and finally, by medul lary carcinomas which have lost almost entirely the squamous celled character The stem of the Y receives the medullary carcinomas showing characteristics of Glond colod Squomous cod 3 -- Und tferentiated Firure 1 Scheme to show the mutual relationship of the cellular types of bronchogenic carcinoma neither gland cells nor squamous cells These are some times referred to as oat seed carcinomas On the upper portion of the stem belong such carcinomas with plump cells and some trace of alveolar pattern Arranged pro gressively downward are the less well differentiated forms with diffusely infiltrating cancers which have very small spherical cells at the bottom of the figure All investigators agree in classifying the carcinomas which would be placed near the extremities of the three branches but within the rectangular zone indicated by the broken line differences of opinion are inevitable Probably this is the chief reason for variance in statistical reports, but it may develop that certain study groups are differentially weighted in respect to age ser, or environ ·GENERAL COVSDERATIONS 9 mental risks It is because of this latter possibility that it has seemed necessary to testate this concept of the rela tionship of cell types in bronchogenic carcinoma There is even now some evidence that the proportion of squa mous celled carcinomas is increased under the influence of particular environmental factors and that the incidence of gland celled carcinomas more nearly represents a basic state zot significantly altered by the presently suspected extrinsic factors INCIDENCE General Incidence The importance of bronchogenic carcinoma in current medical espersence scarcely needs TABLE 1 MALES (ALL ACES) ANYUAL SPORTALSET FRONT CANCER OG TILE LUNG I ER KILLIOY LIVINC United States Neto York Store 1330 1045 Increase 2031-33 1948 50 Iacrcase 31 13: Per Cert! 326 47 208 Per Cert 355 Age distribution adjusted to that for New York in 19-40 (After Levin) emphasis for documentation of the picture as of recent scars such statistical analyses as that of Dorn' are avail able lle stated in 1953 that 8 / per cent of all deaths from malignant neoplasms can now be attributed to cancer of the lung and bronchus This is in rcference to all ages and both sercs The impact of thus statement is felt all the more forcibly when the concentration of bronchogenic ancer in males of late middle life is realized ror the Voted States as of 1915 Leun stated that the annual mortality from cancer of the lung wis 568 per million In ing for miles fifty five to sut; four years of 599 for those sext; five to seventy years of age age and ·30 CANCEN OF THE LUNG CAUSAT, FACTORS A striking feature has been the apprrent continuous increase for the past fifty years This was evident in 1929 when I reviewed earlier literature giving the incidence as found in over 600,000 necropsics The results appeared to show a five fold increase when the period 1872 to 1898 was compared with that of 1916 to 1924 For more recent periods we can abstract the material of Table I from a 1 BITE Perso-S BY SET AND ACL UNITED STATES 1950 1950 DEATH RATE VEA 100 000 WHITE MALES WHITE FEMALES 100 80 1950 60 1940 40 1950 1930 1940 20 1930 UMOEN 36 35 ng: ONCK DVER FICURE 2 Mortality from cancer of the respiratory system Repro duced by permission from the Statistical Bulletin of the Metropol tan Life Insurance Company detailed arrangement by Levin. For white persons the change between 1930 and 1950 is presented graphically in Figure 2 which shows an almost sig fold increase for white males of the age group sırty five to seventy four in that period. Thus the apparent rising incidence noted in earler years is not only continuing but has been accel erated There has always been serious questioning as to whether this increase is due to purely relative factors or ·CENERAL CONSIDERATIONS 11 to an actual greater propensity, whether natural or in duced on the part of human tissue to develop this form of cancer The two views are sometimes expressed by contrastıng "true" with "relative" incidence This is an unfor tunate designation for no one can deny that there is a CAUDE AND ACE ADJUSTED DEATH RATES BY SEX AND COLOR UNITED STATES 1930-1950 OLAT RATE OER 100 000 OLAT RATE ER 00 00 25 CAUDE * It -{0 20 RATE ADJUSTED 20 WHITE MALES NONWHITE MALES 15 15 10 10 WHITE FEMALES NONWHITE FEMALES 5 5 O 1930 1940 1950 1930 1940 1930 1930 1940 1950 1930 1940 1950 CRUDU ATT 40 11.B 241 2.1 3 5 5 4 13 51 145 07 17 34 ANTED - 10 3 194 - 3140 47 12 1 14 26 Detases for 1940 1950 dosted on the be of the *t* diuribution of the population for anch As lod color in 1930 FIGURE 3 Mortality from cancer of the respiratory system The stappled portions of the bars indicate the portion of the mortality which can be attributed to the increased age of the population Reproduced by permission from the Statistical Bulletin of the Metropolitan Life Insurance Company true increase "Real" and "relative are somewhat better adjectives for designating the difference Houcicr the mind has been reluctant to admit that cancer may be due to changing responses between the body and its environ ment and wherever possible has sought other explanations for the increase in bronchogenic carcinoma Such reasons are not hard to find ·12 CANCER OF TIDE LUNG CAUSAL FACTORS 1 To what degree is the increase due to the changing distribution of our population in respect to age? More persons are living in the age period in which broncho genic carcinoma occurs If there are significant environ mental causes, it may be that this is simply another way of stating that more people are exposed, and exposed longer to provocative agents That age distribution can TABLE II DISTRIBUTION Or 313 PULMONARY CARCINOMAS IN 14 000 NECROPSIES (UNIVERSITY OF MICILIGAN) 1695 1954 Necropsies by Thousands Vales Females Total 0 1000 1001 2000 2001 3000 3001 4000 4001 5000 5001 6000 6001 7000 7001 6000 8001 9000 9001 10000 10001 11000 11001 12000 12001 13000 13001 14000 Totals 2 8 9 10 16 17 18 17 21 21 30 34 31 37 271 2 8 10 14 18 21 22 12 SoccerwAANA 26 31 34 42 313 account for approximately one fifth of the increase be tween 1930 and 1950 is shown in Figure 3 The stippled area at the summit of each bar indicates the reduction in the crude rate which would have been experienced if the population had remained with the same distribution in respect to age sex and color in 1940 and 1950 as was true in 1930 Of the three variables only age proves to be of great importance in influencing the change in inci dence 2 It has been suggested that the increase is due to more ·GENERAL COSMETUATIONS 13 nccurate certification of the causes of death resulting from the use of such diagnostic methods as roentgenoscopy bronchoscopy tissue biopsy exploratory thoracotomy, and cytologic study Since thesc measures have been intro duced successively and employed with constantly increasing effectiveness, a gradual and not an abrupt increase in the apparent incidence might be expected The increase has followed such a pattern with a fairly smooth gradient It is impossible to assess the effect of better diagnoses with any degree of accuracy, and it must be remembercd that improved diagnostic methods will not necessarily yreld higher figures for incidence Some of the errors would have been in the opposite direction of certifying deaths as due to pulmonary cancer when such was not the case Moreover the early vrlues for incidence were basrd on necropsy experience and the continuity of methods and personnel in many institutes of Pathology has been such that there is no possibility that a signifi cant number of cases of bronchogenic cancer were missed years In Table II the destnbution of broncho genic carcinoma in 14000 necropsies at the University of Michigan is shown by thousands of necropsics and the same material is presented graphically in Figure 1 Di vided into numerical halcs the first 7000 includca 95 cases of bronchogenic carcinomi or 30 per cent of the total number Divided chronologially into Iulf penods the necropsies (two) on patients with bronchogenic car cinoma in the first twenty nine years were too fex to be statistically significant due in part to the fcucr necropsies of that purod Thev amounted to 023 per cent For the second half period 237 per cent of the necropsies have been upon patients with bronchogenic carcinomi Be crusc the older material is at hand for study and has been revicked from time to time there is no possibility that in earlier ·14 CANCER OF THE LUNG CAUSAL FACTORS bronchogenic carcinoma was missed to a significant ex tent in the earlier years 3 A third basis for an apparent rather than a real in crease in bronchogenic cancer concerns a compler of 45 40 KEY Males Females --- Combined 35 30 25 Number of Count 20 15 10 5 2 3 4 5 7 8 9 14 10 12 13 6 Necropsies by Thousands FICURE 4 Distribution of 313 cues of bronchogenic carcinoma in 14 000 necropsies in the University of Michigan series imponderable considerations such as growth of lay knowl edge and interest in cancer, lay insistence on securing better diagnostic aid more extensive referral of patients with diagnostic problems, and greater professional inter est and awareness in respect to bronchogenic carcinoma ·CENERAL CONSMERATIONS 15 Clearly, these considerations overlap the improved diag Dostic procedures discussed in the preceding paragraph and like them can operate in both a positive and a dega tre direction Probably some patients are non correctly diagnosed as having cancer of the lung ubo previously would have been certified as dying of chronic tubercu losis chronic poeumonitis asthma or even the totally in acceptable general debility In summary practically all who have investigated the problem of changing incidence of cancer of the lung agrec that some of the increase as "apparent" and depends chiefly on increasing longevity but that there is also a real increase which provides the immediate incentive for the further emploration of causal factors Age Incidence As is shown by the current study (Fig ure 2) From the Metropolitan Life Insurance Company the peak of mortality from pulmonary cancer for shto males in the United States occurs in the decade of suty five to scienty four years of age Incidence increases rapidly after age forty and the rate of sncreasc (indicated by stecpness of the cunc) is greatest for the inter al betueen fifty and sırty Similar results have been found in most countries where similar studies have been made Beyond age ses enty four mortality continues to show a ver maried increase over the expencnce for that age group in earlier periods but for the contemporary popu lation the ntc of change is rcucrsed and the cun edeclines Ict it remains tnily alarming that for the group seventy five years old and older the mortality from under of the lung was fire and one-half times greater in 1030 than in 1930 The down tism in thic mortality cure for ilute males after age scienty four has been evident in each of the periods indicated in Figure A similar reduction in mor irlty in advanced sears has been obser ed in most gene ·16 CANCER OF THE LUNG CAUSAL FACTORS graphic areas, but not always beginning at exactly the same age The significance of this decline is as yet un known When the answer is determined it may supply the hey for confirming the importance of environmental factors which are now under suspicion Among possible explanations are the following 1 The aged group may receive less sympathetic and skillful diagnostic attention in view of the improbability that curative treatment can be applied regurdless of the diagnosis It is surprising how often it is forgotten that the person who has lived beyond the life expectincy of his birth, still has the life expectancy of the age group of which he is now a member 2 The aged group may contain a hugher proportion of those whose cells are intrinsically less easily provoked, by environmental factors to that proliferative activity which constitutes cancer 3 It may be that in the older group are those who have not been exposed either in intensity or duration or in both intensity and duration to one or more provocative factors which have come more recently into human en vironment This was apparently the thought of Korteweglo when he predicted that the greater part of the future in crease in mortality from cancer of the lung will concern persons more than sixty years of age In presenting these possibilities it has been necessary to anticipate considerations which are developed more fully in the section on Etiologic Factors in Neoplasia Intrinsic and Extrinsic Only for white females for the year 1950 do we find an ascending mortality curve after age seventy four Ten years before this had been anticipated by a constant level This is probably significant in respect to a current change in incidence The tables of Kennaway and Waller" show ·GENERAL CONSIDERATIONS 17 that for England and Ivales since 1920 the lughest mor tality for cancer of the lung bronchus and pleura in females has occurred about ten years later than in males Sex Incidence For the lung as for the lower lip, mouth and esophagus there is marked inequality in the incidence of cancer in respect to scs As determined by mortality statistics, there las been a progressive widening of the gap between men and women Thirty ycars ago the ratio of deaths from this condition for men and for women A 4000 000 100 13 poest COD 31. 12 > > Ficune 5 Graphic comparison betwem () tatal mortality for males and females and (B) the annual percentage increase for each Sex with 1935 considered on ty stal statustics of bronchogenlo carcinoma in England and Wales" (Nedrawn by permission from Kennanay and kenaway Int J Cancer 1 260.208 1917 ) is about 31 Dorn' ins stated recents that for the United States the ratio is now about 51 Considerable kanunc from thus ratio has been noted in other countries in comparing rural and urban jwpullions and in the more or less peculiar crnenence of particular hospitals In tlic Unn ersity of Wiclngan bertopsy series (1893-1031) ·18 CANCER OF TIE LUNG CAUSAL FACTORS of 313 cases in 14,000 necropsies the ratio as to sex has been 651 Farber referred to a 9 1 ratio for his study group of 1,070 cases It is necessary to keep in mind the difference between the comparative mortality of the seves and the compara tive change in incidence Even with the same effective rate of increase for females as for males the spread between the sexes as shown by figures for gross morbidity or mor tality would continue to widen because of the larger base for males to which the rate is applied Disregard of this obvious consideration leads to a distorted concept of the relationship between ser and bronchogenic carcinoma This was explained by the Kennaways in 193612 and again in 194713 and Figure 5 is redrawn from material which they presented in the latter year The graph in the left half of the figure is based on the actual number of deaths from cancer of the lungs in England and Wales for the period 1935 to 1945 The same data are presented on the right but deaths per annum for each ser are indı cated as a percentage of those of 1935 considered as 100 The changes in rate of increase as applied to males and to females of successive age groups may aid in the identifica tion of significant environmental factors and in deter mining the required period (latent period) for the origin and development of a neoplasm Such analyses should prove increasingly helpful The sex of the prtient influences the cellular type of bronchogenic carcinoma A nt study by Steele't gives the distribution in a group of 201 cases of the patients with squimous celled carcinoma, 97 per cent were men of those with undifferentiated carcinoma 88 per cent were men, of those with adenocarcinoma, 65 per cent were men This relationship may be due to intrinsic differences between the sexes which are apart from environmental ·GENERAL COSSDERATIONS 19 14 N. IGLTE O Mortality rates for cancer of lung and bronchus by age color and sex United States 1930 1932 1939 1911 and 19199 (Ralraun by permission from Dom Acta Unio Intemationalis Conrra Cancum 9532-561 1933) factors to coposure to different couronmental factors to cuxosure to the same entironmental factors to diferent degrees or to different tissue response to the same en vironmental factors There is increasing evidence that with known environmental factors squamous-celled får cinoms is more often the result Therefore in considering ·18 CANCER OF THE LUNG CAUSAL FACTORS of 313 cases in 14 000 necropsies the ratio as to ser has been 65 1 Farber referred to a 9 1 ratio for his study group of 1,070 cases It is necessary to hcep in mind the difference between the comparative mortality of the seves and the compara tive change in incidence Even with the same effective rate of increase for females as for males the spread between the seres as shown by figures for gross morbidity or mor tality would continue to widen because of the larger base for males to which the rate is applied Disregırd of this obvious consideration leads to a distorted concept of the relationship between ser and bronchogenic carcinoma This was explained by the Kennaways in 193612 and again in 194713 and Figure 5 is redrawn from material which they presented in the latter year The graph in the left half of the figure is based on the actual number of deaths from cancer of the lungs in England and Wales for the period 1935 to 1945 The same data are presented on the right, but deaths per annum for each sex are indi cated as a percentage of those of 1935 considered as 100 The changes in rate of increase as applied to males and to females of successive age groups may aid in the identifica tion of significant environmental factors and in deter mining the required period (latent period) for the origin and development of a neoplasm Such analyses should prove increasingly helpful The ser of the patient influences the cellular type of bronchogenic carcinoma A recent study by Steele't gives the distribution in a group of 201 cases of the patients with squamous celled carcinoma, 97 per cent were men of those with undifferentiated carcinoma 88 per cent were men of those with adenocarcinoma 65 per cent were men This relationship may be due to intrinsic differences between the sexes which are apart from environmental

·CENERAL CONSIDERATIONS
19
mer
od
1
60
8
A
A
1
7
FIGURE 6 Mortakty rates for cancer of lung and bronchus by age
color and sex United States 1930 1932 1939 1931 and 1949 ?
(Redrawn by permission from Dom Acta Unio Internationalis
Contra Cancrum 9 532 561 1953)
factors to exposure to diferent environmental factors to
erposure to the same environmental factors to different
degrees or to different tissue response to the same en
vironmental factors There is increasing evidence that
with known environmental factors squamous-celled car
crnoma is more often the result Therefore in considering

·20 CANCER OF TIE LUNG CAUSAL FACTORS the causes of bronchogenic carcinoma, malencss must be uncluded There may be fundamental differences in the male which make him more vulnerable to the influence of his environment Sex hormones can produce both mor phologic and physiologic changes in squamous mucosa 2000 300) Rati of Roles 400) 1500) 1913 1925 1935 1943 1946 1941 1948 1949 1950 1951 1551 193 2331 4741 496.1 5.00 5161 5.301 3451 369 1 300 9 80 1000 200 500 40" 93 39 100 9) 35 200) 100 92 30 35 45 35 65 75 85 1900 10 20 30 40 50 HIGUTE 7 At the left mortality from lung cancer among men at different ages Britun 1911 1950 At the right death rates from lung cancer in England and Wales as to sex and ratio between sexes 1900 1950 19 (Redrawn by permission from Doll Brut MI 2 521 527 585 590 1953) Moreover, Moore and Barr15 have shown that even in the somatic cells there are recognizable differences in the nuclear apparatus of the sexes Perhaps maleness is only a conditioning or synergistic factor but its importance is increasingly evident As Figure 6 four graphs from Dorn' are redrawn to the same scale Together thev give mortality rates from can cer of the lung in the United States over a span of twenty ·CENERAL CONSIDERATIONS 23 years distributed by age, color, and ser When all of the facts of causation are fully known the striking differences in the resultiag twelve curves will find explanation The curves for white males in Britain (Dollie) have, in general a simlar form to those for the same group in the United States, and are shown in Figure 7 The accelerated 100 80 60 1950 DEATHS PE# 100 000 POPULATION 40 20 1940 40 $930 1914 0 20 60 80 $00 AGE TICURF 8 Death rates from cancer of the lung by age white males United States 1914 1930 1940 and 1950 17 (Reproduced by per mission from Cutler and Loveland 1 Nat Cancer Inst 15 201 211 1954 ) increase of the preceding decade is strikingly evident Increasing incidence in females, alsa, is shown in the night balf of this figure which is based on mortality re turns from England and Wales but the rate of accelera tion is lower so that the proportion of male female deaths sose from 155 1 in 1915 to 5 69 1 in 1951 The Risk of Developing Cancer of the Lung The probability of developing cancer of the lung between ·22 CANCER OS TIIE LUNG CAUSAL FACTORS specified ages depends upon the incidence rates of cancer of the lung at the specified and intervening ages and the probability of dying from any cause at the different ages Cutler and Loveland17 have determined this probability for white males by utılızıng the United States life tables with a method of cohort analysis which takes into account both the increasing incidence of cancer of the lung on the one hand and the increasing life expectancy of the mem TABLE III EXPECTED NUSIDER OF CASES OP LUNG CANCER NA COHORT OF 100 000 WITTE MALES BORN IN 1910 For Age intcrtal Number Alite at Beginning of Internal Person Years of Life During Interual Expected Number of New Cascs in Interval Cumulative Cases in This and All Subscquent Interi als 785 30-40 93 572 925 732 33 3 990 40-50 91 295 889 491 218 3 957 50 60 85 796 803 810 3739 60 70 73 501 835,533 1,502 2954 70 80 52 178 385 688 1 452 1 452 Reproduced by permission from Cutler and Loveland " J Nat Cancer Inst 15 201 211 1954 bers of each successive cohort on the other Death rates from cancer of the lung by age of white males of the United States are shown as of 1914 1930, 1940 and 1950 in Figure 8 These authors made use of the cancer surveys conducted in 1947 by the National Cancer Institute in ten urban areas for additional information on morbidity and mortality from cancer of the lung Naturally the values wall differ for each year of buth but our immediate interest is with those born in 1910 and in the decade before and after that year For this group Cutler and Love land developed a table of probabilities, which I have abridged as Table III It may be noted that 3 937 cases of cancer of the lung may be expected to develop by age ·CENERAL CONSIDERATIONS eighty among 91,295 white males of this cohort alive at age forty From this analysis of past experience and for ward projection the authors estimated that of every 1 000 white males born in 1910 and still kving in 1950 eleven may be expected to deselop cancer of the lung by age sixty, twenty seven by age seventy and forty three by age eighty This projection gives recognition to cvidence that while the incidence of cancer of the lung is still increasing the rate of increase is progressively decreasing Therefore the rates as forecast by Cutler and Loveland are not as high as have been computed by some others For those not trained in biostatistics such data may seem complex and difficult of interpretation What they mean for each of us or for the male infant bom in 1954 is not easily appreciated but there is crystal clarity in the following sentences which I quote from M L Levin 19 1954 In less than a decade the chance of developing lung cancer at some time during life has more than doubled among males in New York State Two per cent of all males may be expected to develop lung cancer at present rates of incidence ETIOLOGIC FACTORS IN NEOPLASIA INTRINSIC AND EXTRINSIC If causation be considered in its broadest aspects with due regard to both causal genesis and formal genesis it will be found that disease is never the result of a single factor Two causes or groups of causes must be considered One of these kas origin in the germ plasm of the individ ual the other is brought to bear upon the inherent patterns of living organism from outside the realm of chromosomes and genes The source of this latter cause may be found either within the body or in the extracorporal envuon ·24 CANCEN OF THE LUNG CAUSAL FACTONS ment, but it is quite aput from the germ plasm Consider ation of this relationship leads to the conclusion thrt such antithetical pairs of words as internal and extemal, endog cnous and exogenous, and even constitutional and envi ronmental fail to designato adequately the major groups of factors with definition, intrinsic and crtrinsic meet this need with complete satisfaction Intrinsic causes of disease lie within the germ plasm All other causes of disease are extrinsic It can usually be shown that if the ultimate cause (causal genesis) of a disease is intrinsic or extrinsic its mode of development and minner of expression will be conditioned by extrinsic or intrinsic factors (formal gencsis), respectively. For instance, truc hemophilia is entirely hereditary and there fore intrinsic as to its actuil cause but the expression of hemophulia in each of its victims will depend upon the vascular channels which are opened by extnnsıc factors Conversely, Treponema pallidum is entirely an extrinsic factor There is no hereditary lues, although the phrase is often carelessly used as a synonym for congenital syph ilis Yet the disease produced by this extrinsic factor vanes from patient to patient depending upon age, sex and race because intrinsic factors are conditioning the body s re sponse to the causal agent For each disease entity the proportion of these causal ingredients varies There is no reason for believing that neoplısia is unlike other diseases in having such dual elements in causation 20 The interplay of intrinsic and extrinsic factors is indicated schematically in Figure 9 For some neoplasms the intrinsic factor is very impor tant This has long been recognized for carcinoma of the large bowel developing upon the substratum of familial polyposisas, and the extrordinary concentration of retino blastoma in certain families ? far exceeds the most remote ·GENERAL CONSIDERATIOS 25 possibihty under free chance distribution There is sub stantial evidence that for such new growths somatic vana tions are inhented and that it is upon these structural deviations from the normal that cancer arises This thesis need not be developed here for there are other malignant neoplasms-and fortunately bronchogenic carcinoma is one of them-ın respect to which a specific intrinsic factor has never been proved to be important It is obvious that basic intrinsic factors are concerned Co Intel als FIGURE 9 Schematic representation of the relationship between intrinsic and extrinsic causal factors in neoplasia with the development of all neoplasms as they are with normal growth for behind every neoplasm is the intrinsic urge toward cell division and growth Such factors must be accepted as part of the total picture of the causation of cancer These are not the specific intrinsic factors which give certain neoplasms a hereditary pattem and which I contrast with extrinsic factors Although common to both normal and peoplastic growth the basic intrinsicurge must operate with different degrees of intensity under varying conditions and in different subjects To this biologic sub stratum are to be added specific intrinsic factors if any and whatever extrinsic factors may be operative I have just stated that it is fortunate that bronchogenic ·20 CANCEN OF THE LUC CAUSAL FACTORS carcinoma grics no evidence of any potent specific intrin sic factor If there is a hereditary pattern for this neo plasm, it is below the horizon This makes it all the more probable that there are significant extrinsic factors Once in cetrinsic factor is established, the way has been opened for attempting prevention In the opening paragraph of his artıcle on the socio-economic distribution of gastric circinom, VI Cohart' has written since the time of Percivall Pott almost without cxccption our all too-scant accomplishments in the pre icntion of cancer halo grown out of epidemiological obscrution It is my purpose to review, in which follows, cer tain of the hustonal aspects of the search for extrinsic factors causing cancer, with particular reference to bron chogenic carcinom. I will show that di crse threads of cuidcncc, beginning more than 500 years ago und weaving an intricate design, arc now giving promise of revealing enough of the truc pattern to suggest that to an as yet un determinable degree bronchogenic carcinoma is a pre ventable discasc ·III THE FIRST THEME BERGKRANKHEIT IN THE ERZGEBIRGE-SCHNEEBERG AND JOACHIMSTHAL (JÁCHYMOV*) MARCINOMA OF TIE LUNG as an endemic occupational disease in the miners in the Erzgebirge (Ore Moun tains) of Saxony and Bohemia is as well known to students of bronchogenic cancer as is the work of Beaumont to physiologists internists and medical historians Unlike the achievements of Beaumont which in the words of your lecturer of 195324 were the inception of the modem physiological approach to clinical problems of the gastro intestinal tract intensive investigation of this occupa tional disease has not yet led to results in keeping with the effort expended The miners of Schneeberg in Saxony and of Joachims thal (now Jáchymov) in Bohemia have been for at least 500 years the victims of a chronic pulmonary disease now known to be carcinoma From references in the literature beginning early in the sixteenth century from parish and other registries and from extensive investigations con ducted in the first quarter of the present century, it has been established that this particular Berghrankheit was the cause of death in from 41 to 75 per cent of the active and pensioned miners The clinical picture of the dis ease unvaried through the centuries was marked by cough mucoid mucopurulent or bloody sputum increas ing dyspnea loss of strength and weight increased sweat * Joachimsthal - Joachimstal - Jáchymov 27

·GEORGII AGRICOLAE
DERE METALLICA LIBRI XII QVI.
bus Ofiaa Inſtrumenta Machine acomnia denig ad Metallis
tamiputanua non modo luculenalume debuntur, fed & per
cfigies fuislous infestas, adiunctas Laanis Germaniasa, appels
Linocubusita ob oculosponuntur, ut canus tradi non pofling
E 1 V D & H
DŁ ANIMANTIBV: SYSTERRANETS Liber ab Autorero
cognitus cum Indiabus diucrfis quicquid in opere tractarum efts
pulchre demonſtrantibus
BASILEAE M D LVI>
Cum Pnuilegio Impeatonsinannas »
& Galliarum Regis ad Sexudum.
FIGURE 10 Title page of De Re Aſetallica Libri XII by George
Agricola 1556

·FIGURE 11 This illustration from De Re dletallica gives an indica
tion of the long ladders wluch the miners had to employ in going
to and from wosk.
29

·30 CANCER OF TIE LUNG CAUSAL FACTORS ing boring prin in the chest or back, and inevitably, death after a varying period of incapacity Miners thus afflicted were sud to be bergfertig that is so enfeebled in health by mining as to be unable to continue working Historical references to this unique occupational dis ease cmerge from the Middle Ages Pirchan and Šiklas wrote that it was mentioned by Agricola in 1500 If so the renowned father of mineralogy was a very precocious child for in that y car he would have attuned the nipe nge of six years ** However only the too well rounded date was in crror Who of that period would have been better qualified to record the miners diseases of Jorchusthal? About 1527 Agricola born Bauer was made town physi cian of this flourishing mining village from the silver coinage of which (Joachimsthaler) Our American word dollar was derived Alrendy the author of scholarly books Agricola continued his studies of mines and mining His principal work De Re Actallica long remained the standard technical trertise on mining and smelting 1 quote hun from the translation by the Honorable Herbert Clark Hoover and Lou Henry Hoover as follows Of the illnesses some nffect the joints others attach the lungs some the eyes and finally some are fatal to men Where water in shafts is abundant and vcry cold it fre quently injures the limbs On the other hand some mines are so dry that they are entirely devoid of water, and this dryness causes the workmen even greater harm for the dust which is stured and beaten up by digging penetrates into the windpipe and lungs and produces difficulty in breathing and the disease which the Greeks call ao@ua If the dust has corrosive qualities it eats away the lungs and implants consumption in the body, hence in the mines of the Carpathian Mountains women are found who have married seven husbands all of whom ·BERGKRANKHEIT IN THE ERZGEBIPGE 31 thus terrible consumption has carried off to a premature death Two hundred and fifteen years after the publication of Agricola s treatise Scheflery showed that this fatal ail FICURE 12 This Ilustration from De Re Metallica shows an early method by which the hud rock of the tunnel wall was cracked by buslding fises against at ment could not be pneumonia or peripneumonia because at lasted too long and could not be asthma montanum because asthma did not produce destructive lesions in the lung and after another 100 years Harting and Hesse*3 established in 1879 that the endemic occupational dis ·32 CANCER OF THE LUNG CAUSAL FACTORS ease of Schneeberg was primary cancer of the lung and that about 75 per cent of the deaths of miners were from this cause From observations made in the course of twenty necropsies, they concluded that the cancer was a lymphosarcoma, or, rarely, an endothelial carcinoma (The interpretation of a poorly differentiated round celled carcinoma as lymphosarcoma is an easily under stood error, and one from which we are not entirely free today ) Arsenic especially in combination with nichel and cobalt, was considered by Harting and Hesse to be the directly injurious, causative agent They were fully aware that miners cancer of the lung developed only after many years of labor They wrote that at the earliest, at appeared after twenty years usually after a longer period, and sometimes after as long as fifty years All subsequent investigators have recognized a long latent period although the average work period before the devel opment of the disease was found to be somewhat shorter than was stated by Harting and Hesse This feature was far more important than would at first appear The in cscapablc corollary was and is, that the results of modifi cation of environmental conditions cannot be known with certainty until at least two decades have passed In fact it seems not unreasonable to predict that for certain forms of cancer as related to particular extrinsic factors, it may be necessary to await the vital statistics of another gen eration before the results of preventive measures will be known beyond reasonable doubt It would be inconsistent with the limits which must be placed upon other phases of this discussion to review the many scientific contributions dealing with the miners cancer of Schneeberg and Joachimsthal but certain de velopments must receive further attention Harting s Iike Agricola was a local physician in these ·BERGKRANKHEIT IN TZE ERZGEBIGE 33 miang communities With Hesse he explored every phase of the mode of living and working of the miners from the rancd Galician butter which must be used durng a por tion of the year to the amount of oil which each miner burned in his lamp whule underground He was interested in the oil as a source of carbon dioxide As will be de veloped later had he possessed our present information he would have been more concerned with the oil as a source of the intermediate products of incomplete com bustion in their first illustrative clinical report which concerned thirty six year old mmer who had been em played for twenty and one half years it is evident that inquiry was made as to his smoking habits for he is quoted in translation "I have smoked tobacco but not to excess in the last year little and then only at home This was in 1878 During the first fifty years of medical and physico chemical investigation many environmental factors were suggested as of causal significance for Schneeberg cancer The miners were compelled to descend and ascend ladders for considerable distances sometimes as much as 1 000 feet The working shifts were long and the labor heavy In the harder rock where drilling was necessary fine stone dust was produced in quantity Arsenic present in a content of 05 per cent in some of the rock dust was suspect as were also the abundant moulds growing luxurantly on and in the timbers of damp portions of the workings Car bon dioxide not only from the miners oil lamps but also from fires bult against the face of the rock and smoke and other fumes from these sources and from the use of blasting powder and dynamite were given consideration Thus by no means exhausts the list of hazards which were explored Alargarete Uhlig” in 1991 and while still a candidate ·34 CANCER OF THE LUNG CAUSAL FACTORS for her Doctorate in Medicine, produced an admirable study of cancer of the lungs in the Schneeberg miners Lihe Agricola and Hartıng she wrote from on the scene experience The opening sentence of her report in the Virchow Archio can be translated In my native town, the old village of Schneeberg in the Ore Mountains, the expression 'bergfertig often is heard After discussing the various causal factors proposed by her predecessors, she wrote that she had seen a published article "by a layman, to be sure in which the Schneeberg lung cancer was ex plained as an occupational disease of mines in which the rock contained radium and in which the ur contained radium emanations Both of these conditions were met to varying degrees in the Schneeberg mines but even more so at Joachimsthal In fact it was from Joachimsthal that the Austrian government had obtained the ton or so of residual material (after the recovery of uranium from pitchblende) from which the Curies in 1898 isolated radium chloride It was natural therefore for Margarete Uhlig whose reference to the unproved crusative action of radium and radium emanation in respect to cancer of the lung is the first that I have found in medical literature to turn to Joachimsthal for further evidence If this was the explanation of the miners cancer of Schneeberg she reisoned the incidence of the disease should be greater in Joachimsthal where the rocks and ore possessed greater radioactivity Her inquiry directed to Langhans brought the reply that at Joachimsthal there were no proved cases of cancer of the lungs known to him It subsequently de veloped that only the lack of necropsies had prevented recognition of the fact that pulmonary cancer was an occupational disease in Joachimsthal as well as in Schnee berg Somewhat later when this truth was known Peller80 computed that the liability of the miners of Joa chimsthal to cancer of the lungs was twenty eight times ·BERGRILANKIIEIT IN THE ERZGEBIRGE 35 in greater than that of males of corresponding age in Vienna Amstein had shown in 1913 that the cancer of one Schneeberg miner was clearly squamous celled Beginning in 1921 thorough clinical roentgenologic and pathologic studies of lung cancer in the Schneeberg miners were car ned on by Thiele Rostoska Srupe and Schmor} * ** the course of which it was established that all of the neo plasms encountered were carcinomas that pulmonary cancer was still endemic in this group and that it occurred only in those who worked underground or were in untı mate contact with the products of the mines As to Joachimsthal the first case of cancer of the lung to be recognized on the Boliemian side of the Erz Mour tains was observed by Pirchan in 1926 according to Pur chan and $i ? who mentioned it in their revien This occurred in a worher in the radium factory In 1929 Lowy referred to the many deaths which had been ob served in miners and radium factory workers between thirty and forty years of age described the signs and symptoms and proved that two patients had cancer of the lung, in one case by bronchoscopic examination, and in the other by necropsy With more frequent necropsies at was established by S.k137 38 and his associates that cancer of the lung was highly prevalent and that the time spent in the mines by those who developed cancer ranged from thirteen to twenty three years Whereas Schmorls had emphasized pneumocomosis as of probable prime impor tance at Schneeberg no notable degree of silicosis or pneumoconiosis y as found at Jáchymos By Sikl radium emanation, found in the air of some of the pits in a concen tration up to 50 Mache units was considered the more im portant cause of cancer of the lung As has been stated several environmental factors which attracted early attention as possible causes of cancer of the lung at Schneeberg and Joachimsthal can be elim ·36 CANCER OF TIDE LUNG CAUSAL FACTORS inated for one or both of two reasons (n) they were not restricted in their application to those of the local popu lation who went underground or were in close association with the products of the mines above ground, (b) they were not peculiar to mincs and mining in this district, as compared to other geogrphic areas After these elimina tions there remained four possible causes each of which will be discussed in its appropriate relationship to other than the local problem These are 1 Pneumoconiosis including chalicosis in general, and more particularly, sılıcosis 2 Compounds of cobalt and nichel 3 Arsenic 4 Ridioactivity No one has suggested that the miners of the Erzgebirge used more tobacco than their fellow townsmen who did not go underground Moreover Agricola was born only two years after the discovery of America and the first use of tobacco in Europe is said to have been in England in 1586 What is the situation in Schneeberg and Joachimsthal today? Perhaps some of the answers are known by ap propriate governmental agencies but they are not to be found in medical Lterature for the Ore Mountains are well behind the iron curtain " The flow of information ceased about 1939 and has not been reopened This is all the more understandable in view of the present ne cessity for the production of urinium Successively, this region has been exploited for silver for nickel cobalt bismuth and arsenic for radium and now for uranium A generation has passed since corrective measures were being introduced to mitigate some of the hazards Hoists were provided to make climbing the long ladders un necessary hours of labor were to be reduced and more ·BERGKRANKHEIT IN THE ERZCEBIRGE 37 mechanical equipment provided more effective forced ventilation was recommended to reduce the atmospheric content of both rock dust and radium emanation respi rators were provided and wet drilling encouraged in the dusty areas Let us hope that in years to come interna tional tensions will be relaxed to the degree that we may learn whether the incidence of cancer of the lung in Schneeberg Jachymov and neighboring mining areas has been lowered by such measures, and thus we may gain insight in respect to causation Or will it be revealed that in the competitive struggle for uranium hundreds or thousands of new workers have been exposed perhaps unwillingly to environmental hazards which are almost inexorable causes of cancer of the lungs? Fortified with basic information in respect to Schnee berg cancer it is now possible to proceed to consideration of selected extrinsic factors which were postulated for Bergkrankheit It has been suggested that there was an intrinsic (hereditary) factor operating in Schneeberg and Joachimsthal through many generations of intermarriage as a result of social isolation There is no evidence to sup port this view CHALICOSIS, WITH SPECIAL REFERENCE TO SILICOSIS By the early investigators of cancer of the lung in the Schneeberg miners stone dust was considered important but usually for other reasons than its content of sılıca Šil139 found that an association of carcinoma of the lung with silicosis was erceptional in the necropsies upon miners at Joachimsthal He was convinced that no essen tral role could be attributed to silicosis However many authors from various geographic areas have been im pressed by the coincidence of silicosis and pulmonary cancer in a single case or a small group of cases It must be remembered that when two conditions are each rela ·38 CANCER OF THE LUNG CAUSAL FACTORS tively common, chance distnbution will lead to their simultaneous occurrence to an extent which is predictable if the incidence of each is known The comparative study of pulmonary cancer in relation to silicosis by Klotz89 often is cited as in support of a causative relation, although Klotz was very guarded in his own conclusions He found an incidence of lung cancer of 8 per cent among 50 silicotic patients as compared with an incidence of 1 17 per cent in 4,500 unselected necropsies Such a comparson is statistically unsound The four examples of combined silicosis and carcinoma were in males forty five to fifty four years of age The control group should have matched them in sex and age instead of being unselected Quite apart from this comparative inadequacy the occurrence of four cases of pulmonary cancer in a group of fifty middle aged males is of dubious statistical significance in view of the general incidence of the disease during the period when these cases were be ing collected Kennaway and hennaw ay,12 after tabulatıng occupa tional hazards from the death certificates of England and Wales for the period 1920 to 1932 concluded that some excess of cancers of the respiratory tract was found in as sociation with certain dusty occupations but that, in gen eral, activities leading to silicosis played no important role in the production of cancer Extensive studies of South African gold miners, based on necropsies in Johannesburg have been interpreted as demonstrating that silicosis was neither a causal nor pre disposing factor for pulmonary cancer Table IV arranged by Hueperfrom Reports of the Miners Phthisis Medical Bureau of South Africa, gives the percentage incidence and coincidence of silicosis and pulmonary cancer in 3 215 European males for 1932 and 4510 for 1935 ·DERGKRANKHEIT IN THE ERZGEBIRGE 39 Vorwald and harro reviewed the literature compiled their own observations, and exposed 3 338 animals to various mineral dusts for long periods They reached the conclusion that rock dusts in general and silica cannot be considered etiologic factors in the development of pul monary cancer Also Schulteti found among the miners of the Ruhr that combined roentgenologic and necropsy studies established a complete dissociation between sila cosis and cancer of the lung TABLE IV ConcØENCE OF PULMONARY CANCER AND SILICOSES IN SOUTH AFRICA European Mon misters European liners Without Slikosts European Miners With Suicosts 1109 054 1932 Necropsies 1023 1083 Lung cancer per cent 0 58 037 1935 Necropsies 1393 1679 1438 Lung cancer per cent 093 071 070 From Reports of the Aliners Phthisze Medical Bureau of South Afrca (Cited by Hueper) There is a corollary to the apparently well established fact that silicosis has no etiologic significance for bron chogenic carcinoma-a corollary which has not been suf ficiently emphasized It is that chronic inflammatory disease per se likewise must have little or no etiologic significance for it is difficult to imagine more intense or protracted fibrosing pneumonitis than that which accom papies the more severe forms of silicosis It is becoming increasingly clear that the vague concept of chronic imtation" as a cause of cancer so frequently applied in respect to malignant disease of various organs and regions in past years was greatly Overstressed the extrinsic fac ·40 CANCER OF THE LUNG CAUSAL FACTORS tors are now being sought with increasing success among physical and chemical entities COBALT, NICKEL, AND BISMUTTI In the Schneeberg mines, bismuth was not suspected as a cause of miners cancer and there have been no subse quent observations which indicate that it is important in respect to bronchogenic carcinoma Cobalt was given serious consideration for mıny years, but during the period when the Schneeberg mines were being worked especially for cobalt, it was found that can cer of the lung did not occur in those employed in the co balt pigment factory as it did in the miners Surveys in 1925 by a special committee of the Labour Office of the League of Nations covering cobalt miners of many geographic areas failed to show increased pulmonary cancer in this group (Hueper') With nichel developments have been somewhat differ ent Barnett " in the Annual Report of the Chief Inspector of Factories for the Ycar 1947, stated that forty seven cases of cancer of the nose and eighty two of cancer of the lung had been reported from a nickel works in Britain Amor3 had dealt with this group in detail These neoplasms occurred in workers engaged in the refining of nichel by the carbonyl process and the incidental produc tion of copper sulfate Although nichel carbonyl is itself a highly toxic substance there was little evidence that it was the responsible carcinogenic agent The sulfuric acid used in the process contained arsenic In the Canadian refineries where the concentrated ore has the same nuchel copper, and sulfur content, sulfuric acid is not used and therefore arsenic is not introduced Amor concluded that the evidence is strongly suggestive that arsenic, in com bination with a metal-probably copper-was responsible Morgan, in discussing a paper by William E Smith" in ·BERGKRANKHEIT IN TILE ERZGEBIRGE 41 1953 was able to bring this account up to the present The sulfuric acid used in this particular refinery has been free from arsenic since 1923 dust preventive measures were improved in 1924 and the wearing of masks was instituted The incidence of bronchogenic cancer appears to have been greatly reduced in men engaged since 1924 Morgan considered that the results offered support to the view that arsenic and not nichel carbonyl was the causal factor If this is true it strengthens the case for arsenic under other circumstances as in the Schneeberg mines Hueper' has considered that the evidence for nickel as provocative of lung cancer is well worth further considera tion It cannot be summarily dismissed from the list of possible causative factors ARSENIC The high arsenical content of the ore in the Schneeberg mines put arsenic under suspicion as a causal factor as soon as the cancerous nature of the miners disease had been established Hartıng and Hesse,28 in 1879 as has been mentioned considered the inhalation of arsenical dust to be the chief cause Others thought that the abundantly growing moulds in the damp portions of the mines formed arsines which were inhaled as well as the dust Currie45 has stated that a typical analysis of Schneeberg mine dust indicated the presence of cobalt arsenide to the extent of 0 19 per cent and that hyperheratosis of the palms of the hands occurred in the men washing and selecting the arsenical cobalt ores This latter point has not been emphasized by others in discussing Berghrank heit It has long been known that arsenic when ingested over a protracted period leads to a characteristic hyper keratotic pigmented dermatosis upon which carcinoma may develop Neubauer" reviewed the entire field of ·42 CANCER OF THE LUNG CAUSAL FACTORS IS a arsenical cancer and reached the same conclusion as Currie, 13 namely that there is little evidence to indı cate that arsenic has been the cause of cancer of the lung at Schneeberg and Jáchymoy These miners did not have an unusual incidence of skin cancers Aliners and smelters in other regions where arsenic is abun dant are not known to hive a high incidence of bron chogenic cancer and bronchogenic carcinoma rare complication of typical arsenical cancer However Kennaway and hennau ay's have collected several ap parently isolated cxamples of concomitant arsenical der matosis and bronchogenic circinoma, and of bronchogenic carcinoma after prolonged use of Fowler s solution with out skin lesions due to arsenic While such obscrvations by themselves offer no proof of a causal relationship it is important to have them on record in the hope that similar cases will be noted and reported by others As to the allegation that the carcinogenic effect of cer tain tars is actually due to arsenic Hieger47 has stated that the minute arsenic content of various tars bears no relation to their individual carcinogenic potency Hill and those associated with himº8 found an appar ently significant increase in cancer deaths among the production workers in a factory handling inorganic com pounds of arsenic Dust from sheep dup powder was the partıcular hazard and skin and lungs were the affected organs However, no case of pulmonary cancer was de tected in the working personnel during the investigation Even with these comparatively negative observations I im not prepared to drop arsenic entirely from the list of possible environmental factors at Schneeberg Perhaps when inhaled in even small amounts over a long period arsenic may have a conditioning or synergistic effect upon bronchial epithelium Further reference to arsenic must ·BERGKRANKITEIT IN THE ERZGEBIRGE 43 be made under the topics of general atmospheric contam ination and of possible carcinogens in tobacco smoke JOVZINC RADIATIONS It will be recalled that the pitchblende from which the Curies isolated sadium came from Joachimsthal and that Uhlig 29 in 1921 recorded in medical Literature a sug gestion from a pon medical source that Bergkrankheit depended upon the radioactivity of the miners environment After radim became commercially important about 2 gin was produced each year at Joachimsthal Investiga tron showed that radioactivity in the mines depended upon radium emanation (radon) in the air and water and upon the radium content of the ore Both of these varied greatly in different portions of the mines and fre quently radon was in excess of that which would have been in equilibrium with the local content of radium indicating the probability that other ore bodies were con siderably richer in radium Radon entered the body by the inspired air and by ingested water Radium was taken i as a constituent of roch and ore dust In environmental radioactivity there seemed to be a highly probable explanation for the prevalence of bron chogenic carcinoma in Schneeberg and Joachimsthal This appeared all the more impressive when physical in vestigations were published and is was the custom the results were expressed in Mache units * Ludewig and Lorenser'' found values up to 50 Mache units in the air in the Schneeberg mines The average was considerably low • 1 000 Maché unita denotes the amount of emanation in equi 1 librium with 2000 mg of radium i microctite 2 670 Maché units (Stedmans Medical Dictionary 13th edition Witam Wood & Com pany ) ·CANCEN OF DIE LUNG CAUSAL FACTONS arsenical cancer and reached the same conclusion as Currie," namely, that there is litlc evidence to indi cate that arsenic has been the cause of cancer of the lung at Schnceberg and Jáchymov Thesc miners did not have an unusual incidence of skin cancers Vincrs and smelters in other regions whicre arsenic is abun dant are not boun to base a hugh incidence of bron chogenic cancer, and bronchogenic carcioma is a rire complication of typical arsenical cancer Honevet, hennyway and hennaway' have collected several ap parently isolated examples of concomitant arsenical der matosis and bronchogenic carcinoma and of bronchogenic carcinoma after prolonged use of Fowlers solution with out shin lesions due to arsenic While such observations by themsches offer no proof of 1 causal relationship it is important to have them on record in the hope that similar cises will be noted and reported by others As to the allegation that the carcinogenic effect of cer tain tars is actually due to arsenic, Hieger" has stated that the minute arsenic content of various tars bears no relation to their individur) carcinogenjc potency Hill and those associated with lum** found an appar ently significant increase in cancer deaths among the production workers in a fuctor handling inorgznic com pounds of arsenic Dust from sheep dip powder was the particular hazard and skin and lungs were the affected organs However, no case of pulmonary cancer was de tected in the working personnel during the investigation Even with these comparatively negative observations, am not prepared to drop arsenic entirely from the list of possible environmental factors at Schneeberg Perhaps When inhaled in even small amounts over a long period, arsenic may have a conditioning or synergistic effect upon bronchial epithelium Further reference to arsenic must ·BERGKRANKIERT IN THE ERZGEBIRGE 43 and upon be made under the topics of general atmospheric contam mnation and of possible carcinogens in tobacco smoke IOVIZINC RADIATIONS It will be recalled that the pitchblende from which the Cunes isolated radium came from Joachimsthal and that Uhlig " in 1921 recorded in medical literature a sus gestion from a non medical source that Bergkrankheit depended upon the radioactivity of the miners environment After radiun became commercially important about 2 gn was produced each year at Joachimsthal Investiga tion showed that radioactivity in the mines depended upon radium emanation (Tadon) in the air and water the radnim content of the ore Both of these vaned greatly in different portions of the mines and fre quently radon was in ercess of that which would have been in equkbrium with the local content of radnum indicating the probability that other ore bodies were con siderably richer in radium Radon entered the body by the inspired ar and by ingested water Radium was taken in as a consutuent of rock and ore dust In environmental radioactivity there seemed to be highly probable explanation for the prevalence of bron chogenic carcinoma in Schneeberg and Joachimsthal This appeared all the more impressive when physical in vestigations were published and as was the custom the results were expressed in Mache ants Ludewig and Lorenser" found values up to 50 Niache units in the air in the Schneeberg mines The average was considerably low 3000 Mlaché units denotes the amount of emuagation 19 egut 1 Ubrtup with mg of radíura I microcusie 2670 Maché units (Stedmans Medical Dictionary 1384 edition Walbam Wool & Com pany ) ·CANCER OF THE LUNG CAUSAL FACTORS 60 er, but a high value was found in the mine known locally as the death mine beciuse of the high mortality of its workers Other biopliysical studies followed, and numer ous articles appeared in most of which the authors ac cepted the importance of radium emanation Rajewsky, in 1939, found valucs in air in the mines up to 227 Maché units and in mine water up to 183 units He not only meas ured environmental radioactivity but also that of the breath, urine and sweat of the miners before during and after working He concluded that it could no longer be denied that injury through breathing the emanation laden air of the mines sas at the very least one of the causes of Schneeberg cancer Two years later Behounek and Fořti reported upon similar biophysical investigations of the Joachimsthal area They found that the emanation content of the air in the mines had been improved over previous findings by dilu tion with outside air through better ventilation High val ues were found for the water in the mines and for the rock A systematic investigation of the radioactivity of the lungs and vertebrae of necropsied miners was included The time element of exposure appeared to be extremely important and they suggested 10,000 hours of exposure as the critical limit Among their conclusions are found the statements that radiation from inspired emanation was determinative for the ongin and development of the dis ease, and that radiation from the deposit in the body of more permanent sources of radiation was not significant in promoting cancer They believed that the mechanical effect of the dust increased the sensitivity of lung tissue to radiation In 1944 a somewhat different conclusion was reached by Lorenzo after a critical review of all available clinical, biophysical, and experimental data In his opinion, the ·BEACKRANKHEIT IN TIBE ERZGEBIRGE 45 claim that radon is the sole cause of lung cancer of the miners of the Erzgebirge cannot be sustained He closed hus discussion with a suggestion of hereditary susceptibility as a result of inbreeding but admitted the lack of statisti cal evidence Obviously hereditary susceptibility lihewise could not be the sole factor since the local inhabitants not connected with mining apparently have no more pul monary cancer than the general population The significance of irradiation in respect to pulmonary cancer has not been determined by experimental methods, notwithstanding many attempts and it is probable that the answer will not be found by animal experimentation Apparently the latent penod" so significant for man cannot be duplicated in small animals Ionizing radiation cannot be dismissed lightly Under other circumstances and for other systems and organs its cancer producing effect kas been proved and accepted Sarcomas of bone and other supporting tissues carcinomas of skin and mucous membranes and the leukemias are particularly the forms of cancer which can be attributed to radiation That similar endemic carcinoma has not been found in other geographic arcas za wkich radioactive minerals are present may be due at least in respect to recently opened uranium projects to the fact that insuffi cient time has elapsed Martland s's dial painters must have inspired radon as well as ingested radium salts but so far as I know bronchogenic carcinoma has not been observed in that group Hueper' has invited attentioa to the many and varied applications of sources of radiation in modern living Continued vigilance will be needed to detect environmental hazards It is significant that re sponsible biophysicists have repeatedly lowered the level of permissible total body irradiation per day As to the micrs of Schoceberg and Jáchymov I must ·46 CANCER OF THE LUNG CAUSAL FACTONS continue to include ionizing radiation whether from radon or ridium, among the limited group of factors still under suspicion as inducing bronchogenic carcinoma SCHNEEBENC AND JÁCHYMOV IN RETROSPECT The light from the diligent investigation of causal fac tors in neoplasia which characterizes our era can now Bronchogenic Carcinoma Schneeberg ond Joachimsthal We Fort til Fetara c Us WD pa # 3 M ma FIGURE 13 Schematic summary of possible causal factors of car cinoma in the Schneeberg miners The device below indicates the bronchial mucosa The interrupted arrows come from factors which have been eliminated as causes The other factors named must tentar ender consideration be turned on the Berghrankheit of Schneeberg and Jáchy moi, but neither the experience of 500 years nor the methodology of modern exploration of industrial hazards has provided a fully satisfying answer The evidence may be summarized as follows Intrinsic Factors The basic intrinsic factors promoting cellular reproduction and growth are ubiquitous A spe.

·BERGKRANKIET IN THE ERZGEBIRGE 47 cific intrinsic factor which has been segregated or empha sized by inbreeding can be excluded The possibility of an unknown intrinsic factor particularly of a sex specific character, must be recognized, Extrinsic Factors Excluded are the general factors of heavy labor climate infective pneumonitis and pneumo coniosis in the general sense Compounds of cobalt and mchel remun as remote possibilities Radioactity and arsenic are of primary interest Because no one agent has satisfied critical studies, the way is open for the discovery of causal factors not yet recognized The sequential, ad ditive or synergistic effect of multiple factors is certain to receive increased attention ·IV THE SECOND THEME THE SEARCH FOR THE CAUSES OF INCREASING PULMO NARY CANCER IN THE TWENTIETH CENTURY INERS AND PENSIONERS continued to die at Schnee M berg and Joachiumstear and conteduea secente. Se mere est in this unique occupational cancer was evident from the numerous published studies of suspected causal fac tors but in the medical world at large not too much con cern was felt over this apparently local problemn However, in the second quarter of the present century it was reluctantly admitted that there was an alarming increase in bronchogenic carcinoma in almost all countries in which vital statistics were collected The local problem had be come universal throughout the industrialzed portions of the world The time at which tlus change began is somewhat in definite and differed regionally In a general way it can be seen in retrospect that the increase was well under way before 1925 and was beginning by 1910 or even earlier The then current biologic concept was that the proclivity of an organ to develop a neoplasm was a rela tively stable attribute which of changeable at all would change only in the course of biologic time and over sev eral or many generations This was the reason for the reluctance to accept a fact which was becoming increas ingly obvious--that within one hfetime a form of cancer which had been considered very rise was becoming very common When that conclusion could no longer be de 48 ·THE SEARCH FON CAUSES 49 nied the necessity for searching for the cause became all the more urgent That search still continues and certain social occupational and endemiologic factors have been investigated with varying results In the following sec tions these factors will be presented INFFCTIVE PNEUMONITIS Because the increased incidence of pulmonary cancer became evident after the great pandemic of influenza and post influenzal pneumonia of 1919 and 1920 the possibili ty of a causal relationship was a natural assumption Nu merous papers discussed this matter and particularly in Europe post influenzal pneumonitis was accepted as the cause of the increase Patients were questioned as in sistently in regard to previous pulmonary disease and the elapsed interval since such illness as they are now ques tioned about their use of tobacco It is unnecessary to review the medical literature on this question for our pres ent purposes Reflection then as now should have served as a reminder that there had been other great pandemics of influenza after which cancer of the lung remained a rare disease that pneumonitis as the sole cause left the riddle of male preponderance unexplawed and that the increase in cancer of the lung was becoming evident even before 1919 1920 Although pandemic influenza and pneumonia must be discarded as a cause of the increase in bronchogenic car cinoma there are occasional instances in which chronic pneumonitis seems to be important When a squamous celled carcinoma de elops in the wall of a bronchiectatic abscess following metaplasia of the orginal columnar epithelium to squamous it is reasonable to think of the inflammatory disease as a cause of the cancer Actually at may not have been the ultimate cause but may have ·50 CINCER OF TIIE LUNG CAUSAL FACTONS been causal to the extent of inducing the proliferation of cells which were so conditioned as to be vulnerable to the factor which originated irreversible proliferation INCREASED LIABILITY TO PULAIONARY CANCER IN OCCUPATIONAL GROUPS Investigation of the liability of those engaged in par ticular occupations should be helpful in indicating pre sumptive causes of cancer of the lung In view of the high total incidence no one occupation, as such can be of great importance in determining the total picture, but it might be of extreme importance if it furnishes the clue to a significant cause of wider application Conclusions drawn from apparent occupational concen tration of cancer of the lung are subject to many statistical hazards Fundamental is the fact that pulmonary cancer is widespread and the more the population is subdivided the greater becomes the risk of distortion of statistical results by chance distribution Within a particular occu pational group there are usually functional subdivisions for which the hazards are entirely different For extrinsic factors whuch are of accepted importance there is a long latent or lag period between the first exposure and the first signs of cancer Therefore the patient may have changed occupations repeatedly during that time The occupation recorded at the time of dignosis or of death may not be the one which entailed a special hazard These are some of the reasons why occupational studies of can cer are difficult and why conclusions derived from them must be accepted with caution There are several occupations in respect to which there is evidence of special liability to pulmonary cancer Those with particularly strong documentation are discussed briefly in the paragraphs which follow ·THE SCARCH FOR CAUSES 51 Foundry Workers and Metal Grinders In view of the concentration of the cutlery industry in Sheffield the results of a study by Turner and Graces assume special significance However this dealt with cancer deaths between 1926 and 1935 when the incidence of cancer of the lung was far below its present level During that period there were in the Sheffield area 3 861 deaths from cancer in an average population of 178 559 males over age four teen Of the cancer deaths 96 per cent were from cancer of the lungs and bronch The working male population and also the deaths from cancer of the lungs were tabu lated under thirteen occupational categories The lowest incidence was found in the group of professional com mercial financial and clerking males This was considered as the control or base figure and given a value of 100 рег cent Warehousemen and pachers showed no significant variation (102 per cent) from this base but metal grind ers were 133 per cent and foundry workers and smiths 145 per cent above the basal group The authors con cluded that in Sheffield a significantly ercessive mortality from cancer of the respiratory tract was found in en gineers foundry worhers and metal grinders and in no other occupational groups Production and Processing of Chromates That environ mental factors in the chromate industry are productive of bronchogenic carcinoma is now well established in 1936 Alwens Bauke and Jonas noted the excessive frequency of thus condition in workers in the heavy chemical indus try of Ludwigshafen and Griesheim They considered chromate dust to be the chief cause In the occupational group which they studied there appeared to be a latent period of from twenty two to forty years between begin ning work with chromates and the development of car cinoma The 3'erage period was thirty-one years ·CUYON OF TE LLC CATSIL FACTORS O'rsonshons similar to those in Germany were subsequen'li mide in the United States Michle and Grega Til s' in 1915, remed the emperence of seven plants posgaged in the extraction and processing of chromates for oss Thics found that in this industrial group, lung cancry composed 0 per cent of all cancers as compared with an expected ratio of 9 per cent The crude death rate for cancer of the lung was trents fire times greater than 1 27 which was then considered sommal. These authors in med sorte esidence suggesting that monochromates urte responsible Imprescia" fourd seven cases of pa min lung cancer in a chromate plant with 400 employees Til maulmer was all the more significant because the annual labor tumoser 40 to 50 per cent in this plaat, shule the average crposure of those who had carcinoma *25 sixteca scurs Mancuso and Flueper," in a thorough appraisil ol the hmith kizard in one chromate plant, slmned that the ratio of lung cancer deaths to a1 destks ol diramalo usorkers is approximately fifteen tires goriler than that of the general population in the same counlı The gcomcinc mein latent period for respiratory cincer in these workers wys 106 years somerhat shorter than had Ixca indicated by carlier sun cys These authors mmisidered that insoluble chromium compounds, such as the dust of tlic cliromite ore itself, might be cusative In an industry sdc study of chromate workers for an cleven your period based on the morbidity and mortakty of those enrolled in sich benefit associations, Braton Frasier, and Aaron found about twenty nine times as mins deaths due to respintory cincer as would have been expected in those of the same age in the general popula tron Another comprehrsse study of the abnormally hugh prevalence of pulmonary cancer in the chromate industry in the United States is included in a monograph by Gafa ·THE SEANCH FOR CAUSES 53 fer and associates 61 In this report the suggestion is made that the failure of Bidstrups to find a significant inci dence of bronchogenic carcinoma in 724 workmen in the British chromate industry may have depended upon a specific difference in technical procedures in Britain as compared to Germany and the United States If so this affords an excellent example of the possibility of prevent ing bronchogenic carcinoma by suppressing an environ mental factor Processing of Asbestos Asbestosis a special form of chalıcosis results from the inhalation of finely divided asbestos fibers However asbestos is not a single sub stance 69 The term can be applied to any fibrous silicate of which there are many, and these differ both chemically and physically Srce pneumoconiosis due to asbestos can be diagnosed microscopically by the characteristic asbes tos bodies morphologic recogrution of the coincidence of asbestosis and pulmonary cancer in necropsy and surgical material is on more secure ground that would be true for most dust diseases The earler literature on asbestos as a causative factor for cancer of the lung was reviewed by Hueper in 1942 and again in 1952 - In the later report he collected about sixty eight examples of the concomitance of these two conditions These were described 2 suyle cases or as small groups by many authors However Barnett' in an analysis of deaths in England and Wales which were at tnbuted to industrial hazards found that of 235 in which asbestosis had bcen established there was concomitant cancer of lungs or pleura in thirty-one (132 per cent) This appears to be significant in the overall picture and, from the mortality figures two additional items can be derived which indicate the importance of asbestosis Of 128 malc deaths, in tu enty tho (17.2 per cent) asbestosis ·CANCEN OF THE LUNG CAUSAL FACTORS was complicated by carcinoma of the lungs or pleura, and of 107 female derthis ninc (84 per cent) were similarly affected The ratio of the percentages is almost exactly 2 1, whereas a much greater male preponderance is usu ally found in necropsy material (Sec Sex Incidence, P 17 ) Agnin, the mean age at denth of the males with asbestosis and carcinoma of the lung was 55 2 ycars and of the females 44 6 years The mcan duration of exposure was 20 1 years for males with carcinomi and 76 years for females It seems probable that the women had more hazardous assignments Moreover, asbestos, or some factor associated with asbestosts give an incidence of pulmo nary cancer by percentage ten times that found in fatal cases of silicosis (ninety one pulmonary cancers in 6884 cases, or 132 per cent ) and in so doing lung cancer was caused to appear more often and earlier in females than is true of bronchogenic carcinoma in general Thus indicates that maleness as a cause of pulmonary cancer may be found to be of little signsficance intrinsically, but only an attribute associated with particular extrinsic factors In 1943 Wedler's had anticiprted the conclusions of Barnett** in respect to the proportionately increased inci dence of bronchogenic carcinoma in women with asbes tosis and also in regard to the earlier average age for the recognition of cancer in that sex He found cancers of squamous celled type most abundant Between beginning work with asbestos and the recognition of carcinoma, in, tervals of twelve to forty two years were noted He found also that the first chnical evidence of carcinoma might appear many years after exposure to asbestos dust had ceased He was convinced that cancer of the lung in an asbestos worker should be considered a compensable occupational disease ·TIE SEARCH FOR CAUSES 55 There has not been unanimity of opinion in respect to the crusal role of asbestos On both statistical and er perimental grounds, Vorwald and harr concluded that mhaled dusts except those containing recognized carcino genic substances such as radium and tar cannot be considered of significance in the causation of pulmonary cancer Their experimental evidence wcluded 235 guiner Pigs which had been made to inhale asbestos dust, and in which no pulmonary tumors were found Such experi ments on laboratory animals necessarily of short duration compared to the working period of man do not create comparable conditions Hueperst has recently pointed out that clinical observations made on living asbestos workers in general, bave little evidential value and should not be used to discredit other evidence of a causal relationship Such mass statistics do not take into account a probable effective threshold of exposure and are diluted by the inclusion of workers without asbestosts It is also possible that the chemical nature of particular kinds of asbestos or the prevailing size and leagth of inhaled fibers may cause a real difference in the frequency with wbuch car cinoma develops In connection with a case report and survey of the It erature Isselbacher Haus and Hardy prepared Table V They stated somer hat guardedly Our conclusion at present is in favor of the concept that the association of bronchogenic carcinom.. with asbestosis is more than coincidence They also referred to the fact that in contrast, bron chogenic carcinoma is associated with silicosis only ap proumately to the degree that it occurs in the general population The weight of evidence secms heavily in favor of ·56 CANCER ON TIIE LUNG CAUSAL FACTORS assigning ctrologic significance to asbestosis when it is found in association with pulmonary cancer TABLE V INCIDENCE OF ASBESTOSTS AND CARCINOMA OF LUNG Author Number of Deaths Number of Deaths with Due to Asbestos's Cancer of Lung Incidence 31 15 17 3 17 83 % 132 163 14 8 7.5 14 1 138 Merewether 235 Wedler 92 Wyers 115 Lynch Cannon" 40 Cloyne 121 Total 603 Reproduced by permission from Isselbacher Klaus and Flardy * Am ) Aled 25 721 732 1953 Compounds of Nickel, Arsenic, Ionizing Radiation Since salts of nickel arsenic containing ores and ionizing radiation were all in the physical chemical environment of Schneeberg and Jáchymnov, these factors have been dis cussed in an earlier section (pp 40-46) All continue to play an important part in modern life-radioactivity to a rapidly increasing extent The latter two, particularly, remain under suspicion as possibly having a share in creating the current problem of bronchogenic carcinoma A HISTORICAL INTERLUDE Again we turn to medical history to seek the Origins of what is now a vast store of knowledge about more than 700 condensed ring hydrocarbons as proved or suspected causal factors in cancer The hazards therefrom have been found in the domestic and social habits of man, in occu pations in accidental exposures, and in the inevitable at mospheric pollution which comes with the concentration of human beings in a highly industrialızed and highly mo torized manner of living ·THE SEARCH FOR CAUSES 57 Ideally the combustion of organic matter results in complete oudation with water (hydrogen oxide) carbon dioxide mineral ash (chiefly oudes) and a small amount of other oudes as the products of its carbohydrate and hydrocarbon complexes It has been the good fortune of man that such complete ondation is rarely achieved under natural conditions and he has had as his servants a series of resources ranging from natural gas through intermedi ate petroleum products paraffin tars and patches to coal and the diamond These servants and their alles are also sources of danger Because complete combustion is un economical or impossible in releasing energy as heat or power man adds smole smog soot volatile products and carbonaceous dusts to his environment Among these sub stances or incorporated in them or adsorbed on their surfaces or arising as secondary products of their further oxidation are proved chemical carcinogens Much of the proof has been secured by cutaneous application or sub cutaneous snjection with mice as the subjects It should be clearly understood that is of the present no chemical carcinogen has been proved to be the cause of pulmonary cancer in man It is difficult to conceive how such proof could be adduced Perhaps it can be approximated by statistical analysis of association with elimination of other suspected causes Since bronchial epithelium can undergo metaplasia to a squamous type and squamous celled carcinoma can arise from bronchial mucosa the clinical and experimental obsen tions of squamous celled carcinoma of the skin induced by environmental factors are not too far removed from our fcid of interest Percn all Port and Chimney Sweeps Cancer Percn all Portio wrote in 1775 after referring to occupations precipitating lead colic

·CHIRURGICAL WORKS
PERCIVALL POTT, FRS
AND
SURGEON TO ST BARTHOLOMEW -
HOSPITAL
A NEW IDITION WITH ADDITIONS
IN
TRREE VOLVMLS
A center port to explores i pol
dars is de gra
Expo kasi # # u decu", os m.burda
**** darkeleg de mue Cadenasto on Destani w, /10) ferl
A CORK Cscot
LONDON
Prlated for T Lowxa5 forntox Roscator TCA??!!
Tlvany, w for J Bay, and > HATI
8pcc1xxx!!!
FIGURE 14 Title page of the third volume of The Chirucal
(Vorhs of Percit all Pott in which reference to clumpey sweeps
cancer appears
58

·с A N с ER ѕ c R от т. AMAZINI has written a book de morbis artificum, the Colic of Poictou is a well-known diſtemper , and every body is acquainted with the diforders to which painters, plummers, glaziers, and the workers in white lead, are liable: but there is a diſeaſe as peculiar to a certain ſet of people, which has not, at Icaſt to my knowledge, been publicly noticed, I mean the chimney-fwecper's cancer It is a diſcafe which always makes its firſt attack on, and its firſt appearance in the inferior part of the ſcrotum, where it produces Ficure 15 The page from The Chirurgical II orks on which Perci vall Pott introduced the topic of chimney sweeps cancer of the scrotum 59 ·60 CANCEN OF THE LUG CAUSAL FACTORS but there is a discase as peculiar to a certain set of pcoplc which has not, at least to my knowledge been publicly noticed I mean the chimney sweepers cancer After describing the appearance, course, and surgical treatment of cancer scrotı, he continued The discase, in these people seems to derve its origin from a lodgment of soot in the rugae of the scrotum This observation of an occupational environmental can cer was basic for present day knowledge of chemical carcinogens Soot is not finely divided carbon alone, but combines other products of incomplete combustion of organic matter The broadening line of descent of knowl edge of extrinsic factors cusative for cancer can be traced from Percivall Pott and his chimney sweeps cancer, through the occupational tar cancers to dibenzanthra cenc mcthylcholanthrene and other chemical agents so potent for the production of cancer that the experimen tilist now can produce a cincer to order in a suitable laboratory animal Neither knowledge that many scrotal cancers were occupational nor regulatory legislation based thereon brought about 1 dramatic relief from chimney sweeps cancer Regulations were poorly cnforced and the long Intent period before the appearance of cancer gave a false sense of security In an analysis of deaths attributed to scrotal cancer Henry and Irvine reported, 160 years after Percivall Pott that in England and Wales between 1911 and 1930, there were eighty to deaths from this disease in chimney sweeps ranging in age from thirty five to eighty four years (average 61 6 years) hingri Carcinoma By 1910 Dr Ernest F Neve, ta then surgeon to the Kashmir Mission Hospital, had recognized and interpreted correctly as due to an environmental fac- ·THE SEARCH FOR CAUSES 61 tor a squamous celled carcinoma which came under his observation Dunng twenty five years 1720 malignant neoplasms were removed in that hospital of which no less than 848 were epitheliomas” of the thighs or abdomen Those who developed the lesion were all users of the hangri a portable fire basket carried under the clothing In a squatting position it might rest against the inner sur face of the thighs as well as the abdomen Epithelial proliferation hyperkeratotic areas mottled pigmentation and finally, carcinoma developed This malignant lesion was twice as common in males as in females and the average age at which it was manifest was fifty five Me tastases to regional lymph nodes were common Neve thought that heat must be the provocative agent but in ? later report?he wrote "Volatile substances products of combustion auxetics may possibly play a secondary part In the light of present know ledge heat as such would be relegated to a minor rôle In the use of the hangri con dations were ideal for incomplete combustion a confined spice beneath heavy clothing with restriction of orygen and retention of soot and tary products a fire which of necessity must smoulder or glow and never burst into flame Thus it has been recognized (Hueper') that the resulting cancers are grouped more appropriately with the soot or tar cancers The hangri carcinoma of Kashmır provides a remarkable and instructive example of the production of squamous celled carcinoma of the skin by an cxtrinsic agent This example is all the more impressive because of its lugh incidence in the group exposed and because it involved an anatomical site in which such cancer is otherwise of rare occurrence Tar Cancers and Identification of Cancerogenic IIydrocarbons Cven a cursory sun ey of the historical pathway from Percivall Pott to the current investigation of carcino- ·62 CANCEN OF TIDE LUNG CAUSAL FACTORS genic agents produced by incomplete combustion must include reference to the clinical recognition of tar and pitch cancers and of the successful identification of some of the causative agents Brilliant achievements have made possible the synthesis of hundreds of organic compounds of proved cancerogenic action in laboratory animals As a result the investigator of today can, practically speaking, create and propagate a malignant neoplasm to order This availability of biologic tools for investigation goes far to refute the careless phrase, "the cause of cancer is not known The first description of occupational cancer of the skin from tar or pitch is attributed to Volkmann" (1875) but at was not until the early years of the present century that case reports became frequent (Literature by Hueper) Yamagawa and Itchikawa? hnd succeeded in producing cancer of rabbits errs by inunction with tar in 1914 and 1915 but this information was not generally available untıl several years later 76 By fractionation of carcinogenic tar by systematic investigation of pure substances known to be present in tar or capable of derivation therefrom, and by application of fuorescent spectroscopy because the carcinogenic fraction of tar was fluorescent, Kennaway," Mayneord 78 and Cook, Hewett, and Hieger 19 with others established the importance of 1 2 benzanthracene and of 3 4 benzpyrene among the cancerogenic components of particular English coal tars and pitches These brief references to the historical background of the clinical recognition and physicochemical identification of the cancerogenic products of incomplete combustion make possible the discussion of another occupational haz ard in respect to cancer of the lung Cancer of the Lung in Gas Workers in England and Wales From an analysis of death certificates in England ·TJIE SEARCH FOR CAUSES and Wales upon which the cause of death had been certz fied as cancer of the lung hennaway and Kennau ayla found that the number of deaths attributed to gas workers appeared to be well in excess of the general incidence for employed males Since the actual number employed or pensioned had to be determined by deduction further in vestigation was indicated This is a difficult field because of the varied activities in the artificial gas industry some of which require no special contact with carcinogenic components of tar and other wastes Such a study has been completed by Dolls0 (1952) who found that mortality from all causes among the pensioners of a large London gas company was approximately that which would have been expected in a comparable fraction of the general population However the mortality from all cancer was 181 per cent of that expected and the excess was due to increased deaths from cancer of the lung Of these there were 25 whereas the computed expectation was 13 8 The excess mortality occurred principally among employees engaged in the production of illuminating gas and in the treatment of waste products of gas plants While no large body of evidence has been accumulated such figures sug gest that products of incomplete combustion as found some gas and tar operations may be causal for broncho genic carcinoma just as such products were shown to produce cancer of the skin If fully confirmed such infor mation will be of even greater value in its broad im plications as to the etiology of cancer of the lung than in indicating the means by which the incidence of that con ditron can be reduced in the limited sas vorhers ATHOSENERIC POLLUTION Important as the knowledge may be for the particular groups concerned it is evident that the special liabhty to cancer of the lung on the part of asbestos and chromate group of ·64 CANCER OF THE LUNG CAUSAL FACTORS workers, some gas plant employees, metal processors, and other relatively small occupational groups is an inade quate explanation for the increase in this disease in the 40 3B 36 Grooter Copenhagen Provincial Towns Rural Arcos 34 32 28 26 24 22 100 000 20 18 16 Deoths Per 12 10 6 4 2 1931 33 35 37 39 41 43 45 47 49 FICURE 16 Comparative mortality from cancer of the lung in greater Copenhagen provincial towns and rural areas of Den mark 83 (Redrawn by permission from Clemmesen Nielsen and Jensen Acta Unio Internationales Contra Cancrum 9 603 636 1953 ) last forty years The magnitude of the increase suggests that one or more cancerogenic substances may have en tered the human environment as atmospheric contaminants within the last half century There are numerous avenues by which exploration of such a hazard has been at ·11E SEARCH FOR CAUSES 65 tempted and comprehensive surveys of current opinion can be found in the reviews of Huepers and of Ham wrond Effect of Density of Population and Industrialızation There is statistical evidence of a significant positive cor relation between density of population or, purticularly, density of occupied duellings and of industrialization and the incidence of cancer of the lungs Mortality rates from this cause are invariably higher in urban thaa in rural communsties It is recognized that there may be TABLE VI RELATIVE MORTALITY FROM CANCER OF THE LUNG ANO LARYNX IN ENCLAND AND WALES FOR 1936 1949 BY CLASSE CED AREAS (LONDON - 100) Cancer of Lung Cancer of Larynt Male Female Vale Female London Other large towns Ou er urban districts Rural districts 100 78 83 43 100 73 64 54 100 BO 68 59 100 1&2 170 239 Reproduced by permission from Lerngway and Water " Acta Unlo Internationalis Contra Cancrum 9 485-494 1953 other vanables influencing accuracy or completeness of reporting but the differences are too great to be espluned entirely on that basis From various available reports I select first that of Clemmesen Nielsen and Jensen83 from whuch Figure 18 is redrawn This graphic presentation of the marked difference in mortality from cancer of the lung in Greater Copenhagen as compared to the provin cial towns and rural areas of Denmark is self-esplanatory In all faimess the authors state that conditions in Den mar, are such that it seems impossible to crplain how atmospheric pollution could cause the differences Similar results were obtuned by Lennau ay and Wal ·66 CANCER OF THE LUNG CAUSAL FACTORS ler' for England and Wales As is shown in Table VI, the relative mortality from cancer of the lung in rural districts was less than one half for males and only slightly over one half for females of that for each sex in London Other 3500 Mon Women 000 2500 Geoter London 2000 1500 collons tow > 100 000 1000 500 Towns < 100 poo Rrot disi icts Landon conubio o tow &>JOO 000 Tow < 100,000 Rol diet let 25 45 55 65 75 83 Ficune 17 Death rate from cancer of the lung by ages and by sex in different areas of England and Wales 1950 This illustrates the higher incidence associated with increased density of population 1 (Redrawn by permission from Doll Brit ar 1 2 521 527 585 590 1953 ) large towns as a group and other urban districts yielded intermediate figures For as yet unexplained reasons, a reverse relationship was found for cancer of the laryny in women This apparently discordant note may prove significant as extrinsic factors are determined more pre cisely It at once points up the fallacy of grouping dis eases of the respiratory tract in one category ·TIDE SEARCH FOR CAUSES 67 Dollte apportioned the death rates from cancer of the lung according to four areas as to density of population rural districts towns of less than 100 000 conurbations and towns of more than 100 000 and greater London With further division according to age and sex, the re sulting curves are presented in Figure 17 For the United States Hammond has reported that death rates for males from cancer of the respiratory sys tem were in the ratio of 9 4 for urban and rural districts and for females in the ratio of 5 4 These ritios were based on 1944 data supplied by the Bureau of the Census and do not necessarily reflect the ratios which exist today The following quotation from Hammond indicates his cautious evaluation There is still considerable doubt whether the urban rural difference is renl or whether it is due to errors in diagnosis and artifacts resulting from the methods used in reporting place of residence at time of death Until this matter is clenred up as it should be within a year or two one should be cautious in drawing conclusions based upon this evidence Reasons which have been adduced for the higher death rate in urban districts include (1) Better diagnostic fr cilities in more populous districts (2) tabulation by place of death rather than by actual residence (3) differences in smoking habits and (4) atmospheric differences The qualitative and quantitative differences in respect to med ical care between urban and rural areas are rapidly dis appearing in those regions to which reference has been made The crror introduced by death certificates giving place of death can be corrected by adequate investiga tion Differences in smoking habits as of today do not secm vcry important but as will be brought out it may be necessary to consider what differences may have cust ·68 CANCEN OF THE LUNG CAUSAL FACTORS cd twenty or more years ago Our present concern is with atmospheric pollution Domestic and Industrial Smoke Soot is the chief com ponent of smoke and clinical observations since Percivall Pott have confirmed the cutaneous carcinogenic nature of soot By the use of the fluorescence spectrum, Goulden and Tipler84 determined the presence of 3 4 benzpyrene in mixed domestic soot from a chimney sweeps stoch, with an indicated content of about 300 mg of benzpy renc per kg of soot There is less visible smoke and soot in the United States and Britain than there was thirty or forty years ago, There have been marked changes in the patter of fuel consumption Coal and wood are being replaced in urban districts by oul and gas for domestic heating and also to a considerable degree in the production of power Elec tricity supplies light and power When electrical energy is produced in steam plants from coil the fuel is now burned more efficiently and new plants are located in relation to economical transportation of coal rather than to central distribution of electricity In this country the railroads are rapidly shifting from soft coal burning en gines to oil burners or to electricity Most of these changes, it would appear are in the direction of reducing atmos pheric pollution by smoke However that may not mean an equal reduction in carcinogenic components, and changes for the better if such there are cannot affect the incidence of cancer of the lung for years to come The incomplete combustion in public and private incinerators is a newer factor in augmenting atmospheric pollution Stocks86 was impressed by the smoke hazard and stated that there appeared to be i general tendency for mortality from cancer of the lung to incrcase with the total number of chimneys in a commumty until the dwellings exceeded 100 000, after which other factors appeared to be additive ·THE SEARCH FOP CAUSES 69 Soot and other products of incomplete combustion have not been proved to be causes of pulmonary cancer This possibility is still in the stage of hypothetic exploration, as is true of others which must be mentioned Dust and Fumes from Bituminous Surfaced Roads It can be shown that the rise in pulmonary cancer has been contemporaneous with the growth of highway systems with oiled, tarred or asphalted surfaces Bituminous products used for this purpose are known to contun car cinogenic agents and particulate matter is abraded from the road surface to be disseminated as dust To this dust is added carbon black from automobile tires In spite of continued interest in this matter for many yeurs no im portant evidence has been advanced that bituminous roads have contributed specifically to the incidence of bronchogenic cancer This also is a field for exploration of hypotheses Exhaust Products from Internal Combustion Engines Therc is no more conspicuous change in the pattern of human living during the period of the rise of broncho genic carcinoma than that afforded by the ubiquity of self propelled vehicles Circumstantial evidence of the same character as that advanced in respect to cigarette smoking can be used to cast suspicion on the automobile (1) There is remarkable chronologic parallelism between the production and use of automobiles and the increase in cancer of the lung (2) The concentration of automo biles in urban and industrialized areas is in accord with the higher incidence of bronchogenic cancer in such com munities (3) Sufficient time has clapsed as of today, since the automobile became of common usc to satisfy the requirement of a latent period kotine 87 and his associates hise included chcular exhausts from both gasoline and diesel engines in a $35 tematic search for the sources of carcinogenic au pollut ·70 CANCER OF THE LUNG CAUSAL FACTORS ants in the Los Angeles area Their interest was particu larly in the aromatic polycyclic hydrocarbons and from the exhaust of a gasoline engine they obtained pyrene 3 4 benzpyrene, 1 2 benzpyrenc, 1 12 benzperylene capo nene, anthranthrene and an unidentified yellow crystallıne compound While the output of individual hydrocarbons varied somewhat with load and speed, it was found that the worst conditions for a gasoline-propelled vehicle from the standpoint of emission of hydrocarbons occurred during acceleration from low speed In descending order of emission this was followed by idling at a low engine rate, and then by deceleration It will be noted that these determinations have an important bearing on urban driving Under conditions of maximum hydrocarbon production 75 mg of pyrene was formed per munute by the engine under test Benzpyrene was found in 10 per cent of this amount Benzene extracts of the particu late phase of the exhaust products produced cutaneous papillomas and squamous celled carcinomas of the skin in C57 black mice in long term painting experiments Three agents known to be cancerogenic are present in the atmosphere of densely inhabited areas Reference to each of these has run like a thread through the preceding discussion They are arsenic, ionizing radiation, and benz Pyrene Arsenic Presumably atmospheric arsenic comes chiefly from the burning of fuel the smelting of ores and the use of insecticides Goulden, Kennaway and Urquhartos determined the arsenic in suspended matter in the air at eight localities in England and in the winter and summer Their results are presented in Figure 18 redrawn from their report The amount found averaging about 007 ug per cu m computed as As2O3 was often only twice that found in the blank control so no high degree of accuracy ·TIDE SEARCH FOR CAUSES 71 Was claimed It will be seen that the values presented are highest for London and umformly higher in winter than lo summer reflecting fuel consumption The authors computed that a man breathing 20 cu m of air a day would 02 Winter Summer $ *9&s03/m? W . LI Shfd COU 1 MI ) FIGURE 18 Mead seasonal variations in arsenic content of air at eight stations in England While the content of arscruc is very small, it is avanably hugher in winter than in summer 88 (Redrawn by permission from Coulden kennaw.y and Urquburt Bnt ; Can 6 17 1952) inhale but 05 mg of As Os per year Considering known industrial hazards this amount would secm insignificant unless it plays an additive or synergistic role in association with other carcinogens Ionužng Radiation Apart from special local environ ·72 CANCER OF THE LUNG CAUSAL FACTONS mental conditions which might lead to unusual exposure to radioactivity, there is measurable atmospheric radio activity Presumably this is derived from the decay of radioactive materials in the earth s crust If so, it can be assumed to have been fairly constant in hind and quantity over long periods and this would seem to exclude it from participation as a cause of the current increase in lung cancer Dawsons found no great difference in the open air in cities and in the country, but there was marked day to day viriation out of doors which depended upon wind The inore stationary the air, the greater was the radioac tivity In closed spaces such as an underground air raid shelter the amount might be several hundred times that found in the open street For instance, the average value found in an air raid shelter in central London was 11796 10 12 curies/cu m and at a nearby control point in the open, 96 5 10 12 curies/cu m Even the higher of these values is well below the limit of tolerance for mın Dawson' has brought together the determinations of other observers and discussed the dosages deduced there from He very properly called attention to special risks in hospitals laboratories and isotope installations of all types but so far as general atmospheric radioactivity is concerned there is no basis for considering it of signifi cance in causing pulmonary cancer Benzpyrene in the Atmosphere Since benzpyrene was found in soot (Goulden and Tipler84), its presence in air without visible smoke was to be expected Wallero esti mated the benzpyrene content of filters through which air had been drawn at ten stations in Great Britain The mean annual concentration was of the general order of 27 yg per 100 cu m of air The seasonal variation was greater than the difference between stations The concen tration rose sharply during the winter indicating the con ·TID SEARCEZ FOR CAUSES 73 tribution from space heating sources and there was also a tendency for the mead annual values to locrease with the size of the communty Kotin Falk Afader and Thomas found 3 4 benzpy rene present in the air of Los Angeles in an amount of only 900 pg per 1000000 cu ft (28 300 cu m ) The high tumor yield from the painting of extracts of atmos pheric residue on C57 black mice in spite of the low benzpyrene content was thought to indicate either the presence of unknown substances of hugh carcinogenic po tency or the production of other carcinogens during the oxidation of known atmospheric hydrocarbons The emphasis which has been put upon the benzpyrene content of smoke soot road dust and air depends in part upon the extensive knowledge of its fluorescence spectrum which has made estimation of very minute amounts possible As a possible cause of cancer of man benzpyrene thus far suffers only from guilt by associa tion and by inference from results of animal crperiments I know of no proof that any of the benzpyrenes has caused cancer m man It must be remembered that other car cinogens have been proved to be present in some tars (Berenblum and Schoental") Still more potent carcino gens possibly water soluble may awast discovery REMOSPECT In retrospect it has been shown in the serch for the causes of the increasing incidence of pulmonan cancer in the twentieth century that certain occupations are or were particularly hazardous Thus was true of some foun dry workers and metal processors of those producing and packaging chromates processing asbestos refining nickel and working with arsenical dust Most of these hazards were preventable with impro, ed industrial practices In ·74 CANCER OF THE LUNG CAUSAL FACTORS another industrial group, namely, worhers with certain tars in gas production and distribution, the greater lia bility to bronchogenic carcinoma was assumed to be due to exposure to cancerogenic liy drocarbons known to be products of incomplete combustion The forty years more or less, in which the incidence Bronchogenic Carcinoma in the Twent eth Cantury the I Fctors til Factor unt Her drey Mlanet Know G CNI Di tal DHT on 11 dialo Artan AD C*m of Influs Carel le Pr Out of Pont farve Incomplm Cm IN Preno sitlen La 2 Kpl 1 in whe FIGURE 19 Schematic summary of possible causal factors for bron chogenic carcinoma in respect to the increase in the first half of the twentieth century The broken arrows at night and left lead from factors which are believed to have been eliminated from further consideration To be compared with Figure 13 of bronchogenic carcinoma has been rising to an alarmig degree has been a period of increased industrialization of changes in commonly used fuels, and of the ascendancy not to say dominance, of the automobile It has been the era of the internal combustion engine These violent changes in the pattern of life (all of which are incorpo rated in our higher scale of living') have accentuated some long suspected enyironmental hazards and have in ·THE SEARCH FOR CAUSES 75 troduced new ones But the cause so far as the general population is concerned has not been clearly established Another widespread change during the last forty or fifty years has been in the use of tobacco Not only has the con sumption of tobacco increased in those regions in which pulmonary cancer has increased but a proportionately greater amount of tobacco is now being smoked and this in turn with a shift from cigars and the pipe to cigarettes During this period also the women of Europe and the Americas have become cigarette smokers An association between heavy cigarette smoking and pul monary cancer is now being claimed The evidence upon which this is based and possible causal factors provided the relationship should be more than association form the subject matter of the Thurd Theme ·V THE THIRD THEME TOBACCO SMOKING AND BRONCHOGENIC CARCINOMA TH TELENE IIAS BEEN INTEREST in the possible relationship between smoking and cancer of the lung almost from the time that this neoplasm was recognized Even at Schneeberg we found Harting28 making inquiry in 1878 as to the smoking habits of an afflicted miner It is easily appreciated that the presence in the airway of the prod ucts of slowly burning tobacco should arouse suspicion when cancer develops in that tract It was not long be fore the possibility of association of the two was under statistical investigation and there have been three meth ods applied in succession, by which this association has been studied These can be designated, respectively, as the Casuistic, Comparative and Prospective Methods THE CASUISTIC METHOD The casuistic method, which depends entirely upon the history of the afflicted patient, is the original approach when 1 causal factor is suspected but the earlier reporters of collected cases of cancer of the lung were more interest ed in industrial hazards than in personal habits Some like Perret en commented on the excessive use of tobacco in their patients Such observations proved nothing be cause there was no basis for comprison with the smoking habits of a comparable group from the general population, but they did indicate the need for more critical investiga tion They seemed to be in accord with earuer observa tions of the frequency of cancer of the lung in cigar makers and tobacconists generally 76 ·SMOKING AND BHOVCHOGENIC CARCINOVA THE COMPARATIVE METHOD The comparative method be the casuistic method requires the collection of personal bustories as to smoking from those with pulmonary cances, but it then takes the next step and compares the smoking habits of the can cerous group with those of a non pulmonary-cancerous population as a control Statistscal value depends on the size of the respective samples and on the precision with wluch the control group is selected It must be made truly similar in age sex, and color The value will be further enhanced if the groups to be compared are of similar geographic extraction equally urbin and/or rural of similar social status and similarly distributed in diverse occupations la short they should differ so far as can be achieved only in that one group has cancer of the lung and the other does not By personal interrogation the smoking habits of each member of each group are learned as to amount manner and duration of smoking. There are two general methods by which the control group can be established One is a case by case matching plan utilizing for the control group patients with non cancerous disease, or those with cancer of other areas than the respiratory tract or selected persons from the apparently healthy population This has die adiantage of giving opportunity for detailed appraisal of the controls as individuals The second method is to compare the smoking habits of those with pulmonar cancer with the smoking habits of an assumed standard population nbuch in turn las been established by a large scale census The second plan offers the adı antage of deriving the control data from a large group who might thereby be morc repre sentatic of the population at nsk Scicre cities will insist that one group of human bengs can never be truly reprc sentative of another group ·78 CANCER OF TICE LUNG CAUSAL FACTORS The first use of the comparative method in connection with smoking appears to have been by Muller 93 who, in 1937, studied the histories of eighty six male patients of the University Hospital of Cologne who had, or had died of, cancer of the lung He divided them into five categories as extreme smokers, very heavy smokers heavy smokers, moderate smokers and non smokers (In the last group TABLE VII COMPARISON OF TIE USE OF TOBACCO PER DAY BY 86 MALES WIT PULMONARY CANCER AND OP 88 HEALTHY MALES Actual Numbers Percentage Smoking Habits With Healthy Of 86 Of 86 Cancer Controls With Cancer Healthy Men Extreme smokers Very heavy smokers Heavy smokers Moderate smokerst Non smokers Totals 25 18 13 27 3 88 4 5 22 41 14 86 29 07 2093 15 12 31 39 3 49 100 00 4 65 581 25 58 47 68 16 28 100 00 * Extreme smoker = 10 to 15 cigars or more than 35 cigarettes or more than 60 gm pipe tobacco daily + Moderate smoker = 1 to 3 cigars or 1 to 15 cigarettes or 1 to 20 gm pipe tobacca daily (From Muller) there were but three ) When compared with eighty six healthy men of similar ages there was a marked difference in the trend toward heavy smoking in those with cancer of the lung Muller s summary appears as Table VII It falls far short of proof because it would be extremely difficult to pick a control group of only eighty six without some bias in the manner of selection In the same report Muller stated that the use of cigarettes in Germany had increased more than five fold between 1907 and 1937 Studies employing similar comparative methods have appeared with increasing frequency until there are now about fifteen available with each author striving to make ·SMOKING AND BRONCHIOGENIC CARCINOVIA 79 cigars and his control group more truly representative than those of his predecessors With but a few minor erceptions these studies are in agreement that there is an association be tween heavy smoking and pulmonary cancer The actual proportions of those in the various categones has varied considerably in these reports This is to be expected be cause of geographic locations differences in definition of smoking categones and differences in expressing the cigarette equivalent of Pipe tobacca Especial ly important is it to secure the smoking history of each person over a period of many years The non smoker of today may consider his heavy smoking of ten or twenty years ago of no importance and msistent questioning may be needed to bring out the truth Because of the many minor differences in method and definition it is impossible to bring together in one table the results of all the important comparative studies Par tial compulations can be found in articles by Levin " and by Hammond and Horn ** As a more modern example of the use of the compara tive method the work of Wynder and Graham (1950) can be cited The smoking habits of 605 patients with cancer of the lungs were compared with those of 780 male patients with other diseases Whereas there were 146 per cont non smokers among the general male group there were 13 per cent pon smokers among male patients with pulmonary cancer There were 517 per cent heavy to chain smokers and 191 per cent excessive and chain smokers in the non pulmonary-cancer group Correspond mg figures for the pulmonary cancerous group u ere 864 and 51 2 per cent Dol and Hill" (1932) have utilized companson with both non pulmonary cancerous hospital patents and a sample of the general public in a study of the smoking habits of the respectne groups Theu comprehensisc un ·80 CANCER OF THE LUNG CAUSAL FACTORS vestigation included inquiry into many aspects of the use of tobacco and extended also into cnvironmental factors other than smoking For our present purposes, their Table V has been rearranged and appears as Table VIII An important feature of the report by Doll and Hill is the segregation and also inclusion of 202 patients incorrectly TABLE VIII AVERACE AMOUNT OF TOBACCO SAOKED DAILY OVER TILE 10 YEARS PRECEDEC THE PRESENT ILLNESS Men Women Smoling Habits 1357 1357 108 108 With Lung With Other With Lung With Other Carcinoma Diseases Carcinoma Discases No % No No % No % Non smokers <5 cigarettes daily 5 15 cigarettes 15 25 cigarettes 25 50 cigarettes > 50 cigarettes 7 05 55 40 489 360 475 350 293 216 38 28 61 45 109 95 570 420 431318 154 113 12 09 40370 16 | 148 24 | 222 14 130 14 130 0 00 59 S16 25 231 18 18 7 B 58 0 00 0 00 • 1 oz of pipe tobacco a week = 4 cigarettes a day Rearranged from Table V of Doll and Hill Brit MJ 2 1271 1286 1952. Reproduced by permission. thought to have carcinoma of the lung This group gave histories of smoking habits proportioned as were those of patients with other respiratory diseases with cancer in other sites, and with other diseases in general, but unlike those of patients proved to have pulmonary cancer Thus shows that the results from the major group were not fallacious because of special selective factors associated with hospitalization for bronchogenic carcinoma or of bras on the part of those interviewing the group having bronchogenic carcinoma In view of the decisive results quoted from Wynder and Graham 5 and from Doll and Halle and the general agree ·SMOKING AND BRONCHOGENIC CARCINOMA 81 ment of their findings with those of Levin and many others it must be concluded that, by the comparative method of statistical study, an association has been dem onstrated between tobacco smoking and carcinoma of the lung The assertion of a relation of association does not necessarily imply a causal relation In addition to proof of an association between smoking and pulmonary cancer certain miscellaneous data have been established with varying degrees of certainty It is cvident that intensity and duration of smoking are very important In the series of Wynder and Graham "s of 605 patients with cancer of the lung 96 I per cent had smoked for twenty years or more For an equal weight of tobacco consumed cigarette smoking appeared more significant than pipe smoking A far larger percentage of cigarette smokers than of cigar or pipe smokers admitted that they consciously inhaled smoke Proportionately more non smokers were found among males with bronchial adeno carcinoma than among those having squamous celled and undifferentiated neoplasms Adenocarcinoma of the lung is proportionately more common in women and surveys show that there are far fewer women than men who have been heavy smokers for twenty years or more The infer ence is that extnnsic factors play a much smaller part in the causation of adenocarcinoma than of other forms There is insufficient evidence as yet in respect to a possi ble lessened effect in those who use cigarette holders or filters There is no evidence that lighter fluid in mechanı cil lighters is of any importance TTIE PROSPECTIVE MIFTTIOD The comparative method is largely historical It is also retrospective and is sometimes referred to by that term The prospective method of recent introduction is histon ·82 CANCER OF TIIE LUG CAUSAL FACTORS cal so far as smoking habits are concerned, but looks only to the future for the advent of disease and death Under the auspices of the American Cancer Society and with the aid of 22 000 volunteer workers, usable question naires were secured which gave in adequate detail the smoking characteristics of 187,766 white males between the ages of fifty and sixty-nine The death of any man in her group was to be reported by the responsible worler, after which the cause of derth would be learned through official channels The questionnaires were largely com pleted in the first five months of 1952 An apparent objection to the prospective method was the period of delay before significant results could be known but with such a large initial group the delay was surprisingly short, as will appear After about twenty months, 4 854 deaths had occurred These form the subject of a preliminary report by Ham mond and Homer Of these deaths, 167 were attributed to primary pulmonary cancer The distnbution in respect to smoking habits at the time of questioning is shown in Table IX It will be noted that even in this early report the difference in Jung cancer deaths between non smokers and regular smokers proved to be statistically significant (P-0002 or less) I quote from Hammond and Hom The best estimate that can be made at the present tune (at the 5% level of confidence) is that lung cancer deaths are from 3 to 9 times as common among men with a his tory of cigarette smoking as among men who have never smoled regularly and that lung cancer deaths are from 5 to 16 times as common among men who smoke one pack or more per day (as among non smokers) The prospective method permits the simultaneous in vestigation of any cause or of all causes of death in rela tion to the historical attribute (in this instance, smoking) ·SVOKNE AND BRONCHOGENIC CARCLYOVLA The preliminary report of Hammond and Hornºt includes by products" which may be of even greater importance than demonstration of the association of smoking and cancer of the lung Smoking was found to be associated with a considerably higher total death rate, due largely TABLE 1X LUNG CANCER DEATIT RATES BY TYPE OF SMOKINC Type of Smoking at Time of Questioning Population of Risk Desths Attributed to Pulmonary Cancer Death Rate Per 100,000 44 091 12 27 2 Never smoked or occasionally only Cigar and/or pipe cigarettes never smoked regularly Regular egarette smoking 35 853 12 33.5 107 822 143 1326 Totals 187 786 167 889 Regular cigarette smoking <one pack a day Regular cigarette smoking one + packs a day 54 789 80 113? 25 497 81 239.2 Abridged from Sammond and Hom").A.BLA 155 1316-1398 1854 Reproduced by permission to a higher death rate from coronary occlusivc disease and from cancer in general as well as cancer of the lung More complete reports will follow Doll and Hills were already applying the prospective method to British physicians Usable questionnaires were obtained in 1951 from 24 359 male doctors over thirty fisc years old Duning the next twenty nine months, the deaths of 789 doctors were reported and of these thirty six werc attnbuted to pulmonary cancer The distnbution of these deaths in respect to smoling habits is shown in ·84 CANCER OF TIIE LUNG CAUSAL FACTORS Table X The number is small and does not reflect the true death rate for doctors, because it was evident that those who were seriously ill in 1951 did not return question naires Nevertheless, the progressive trend toward higher derth rates from pulmonary cancer with increased smok TABLE MORTALITY OF BRITSI PITYSICIANS AS RELATED TO SMOKING HABITS A PRELLVITNANY REPORT Standardized Deatlu From Death Rate From Cancer Cancer of Lung of Lung Per 1000 Per (29 Vonths) Annum Number at Risk Ocer Age 35 Smoking Habits Non smokers 1 14 gm per day 15 24 gm per day 25+ gm per day Total 3093 8431 7 602 5 203 12 11 13 000 048 067 114 21 389 36 066 (all groups) • I cigarette - I gm I oz pipe tobacco a weck - 4 gma a day Rearranged from Doll and flui" BHE AI I 1451 1455 1954 Re. produced by permission ing is clearly evident Much more information will be available in the next few years If there were hidden fallacies in the comparative or historical method it might be expected that they would be revealed by the entirely different approach of the pros pective method However all methods agree in indicating a statistically significant association between protracted cigarette smoking and pulmonary cancer, with the degree of association increasing with the amount of tobacco used While proving association does not prove causal relation ship, no one has succeeded in explaining how or why such an association exists of causal factors are not implicated Possible clusal factors in tobacco and tobacco smoke are considered in the following sections ·SUOKING AND BROYCHOGENIC CARCINOMA 85 Radioactivity of Tobacco Smoke and Tobacco Ash A recent report from the Vaiversity of Leeds** indicates that the average radioactivity of one agarette approxi mates that of 25 mg of potassium and that the radioactiv ity of the ash is of the same order The radioactivity of the smoke from one cigarette is about that of 6 ug of potassium These values indicate that such radioactivity as is present in tobacco remains in the ash Little if any is transferred to the smoke If these determinations are sus tained httle importance can be attached to the radio activity of tobacco Of course the possibility of an additive cffect is always open no matter how small the increment may be Arsenuc in Tobacco and Tobacco Smoke There has been interest in the arsenical content of tobacco for rca sons other than a possible causal relationship to cancer of the lung In 1940 Barksdalelas expressed conoern that while pure food lass permitted only 1 43 parts per million of arsenic tobacco might have fifty times that amount He recognized that the arsenic in all American tobacco was due to the use of lasechcides and contrued a me chanical smoker to collect the arsenic for assay He found 00365 to 00195 mg of As in each cigarette By subtract ing the arsenic which remained in the ash and in the butt he found values for cigarette smoke of 0 002 to 0001 mg of As per ten cigarettes Barksdale considered tobacco to be the source of the arsenic in scveral cases of characters tic arsenical dermatitis of sluch he give case histories In the following year Carratalá and Vucetichi te ported on the arsenic content of Argentınıın tobacco They also recognized that arsenical dusts and spray's used in controlling plant diseases were the source They found from 00 to 0035 mg of arsenic per lg of tobacco Thus wide variance from the results of Barksdale my been due to differences in methods of assay or to actual hase ·86 CANCER OF TIDE LUNG CAUSAL FACTORS differences in the tobacco Carratalá and Vucetich found one sixth of the arsenic to be retained in the ash The earlier literature was reviewed in 1945 by Thomas and Collier" who reported their own results with apparatus which simulated the varying air currents of actual smoking Their interest was directed primarily toward the arsenic in the smoke They found that from 11 to 23 per cent of the arsenic in tobacco was contained in the smoke the remainder being distributed between the ash and the butt The concentration of arsenious oxide in puffed smole ranged from 0 2 to 30 mg per cu m with cigars, from 17 to 57 mg per cu m with pipe tobacco, and from 33 to 105 mg per cu m with cigarettes of the volatılızed arsenic in the smoke, some must leave the body by exhalation, as has been recognized by Daff and Kennaway 103 Presumably the amount retained is small in their report of 1950, they gave detailed infor mation on the arsenic content of fifteen brands of ciga rettes which was found to range from 0 to 106 ug per gm (ug per gm - parts per million) of As Os The lowest values were found in nine brands classed as Turkish and in French and Rhodesian tobacco Even with the same brand the amounts in the upper ranges were very varia ble Relatively high values were found for three American brands, with 25 3 to 470 ug per gm of As Os By sub tracting the As content of butt and ash from that of the entire cigarette, it appeared that from 76 to 18 2 per cent of the arsenic was lost in smoking Because there are differences in the arsenic content of tobacco Daff, Doll, and kennaway104 attempted a com parison between the types of tobacco smoked and the inci dence of cancer of the lung in various countries Their paper of 1951 is a rich collection of data on this and re lated subjects Information in respect to Turkey seems particularly important Having in mind that Turkish to ·SVOKING AND BRONCHOGENIC CARCINOMA 87 acco is low in arsenic and that Turkey produces an er ss of tobacco so that Turkish tobacco is smoked almost sclusively the high incidence of cancer of the lung rebrted from Istanbul shows that the arsenic content of tobacco has not provided any simple and exclusive explanation of the association between cigarette smoking and this form of cancer I in the other hand an affirmative argument might be ade for Iceland where the amount of tobacco used as garettes is small and a very low incidence of pulmonary trcinoma is reported (Dungal105) Further studies along his line are indicated *Tars from Tobacco The amount and character of the which is produced when tobacco is burned will de ’nd upon such physical conditions as the temperature of Imbustion the supply of oxygen and the compactness nd dampness of the tobacco Smoking methods are erefore important as is also the opportunity for conden {tion and sublimation of volatile products through cool g and filtration In well conducted experiments with moking machines” the apparatus is so contrived as to ve alternating periods of active draft and of no draft 1 as to simulate the conditions of smoking Attention is ven also to the length of the butt when it is discarded nce tars and particulate matter will be retained to some tent in the unburned portion of a cigarette the more mpletely it is smoked the greater the proportion of ch products which will be inspired There can be no bubt that under suitable conditions of incomplete com astion carcinogenic products are included in the tar and Sot from tobacco and also from cigarette paper cNally 100 in 1932 collected tar from cigarette stole hd determined some of its major constituents This was ·88 CANCER OF THE LUNG CAUSAL FACTORS done with only slight reference to pulmonary cancer and with greater concern over chronic bronchitis and leukopla kia His animal expenments were not conclusive since most of the animals died too soon to expect the appear ance of cancers Of particular interest is the statement that cigarettes should not be smoked too short, as the last two centimeters retain most of the tar and other products of incomplete combustion Bogen and Loomis107 also failed to produce neoplasms of the shin of mice by the application of tobacco tar, al though they were uniformly successful in producing papil lomas and finally squamous celled carcinomas of the skin of control mice to which gas house tar was applied Pro duction of cutaneous cancer by tar from burning tobacco has been reported for rabbits by Lamb und Sanders108 and for mice by Flory, 109 and by Graham Wynder, and Croninger, 110 and others Flory reviewed the work on tobacco tars before 1940 It is of interest that Wacker and Schminckell were injecting pipe tar into rabbits in 1911, [translation) because carcinoma of the lung occurs in tobacco handlers and cancer of the lip in pipe smokers There is also an extensive group of reports on the at tempted production of pulmonary neoplasms in mice by exposure to tobacco smoke Results have been highly var iable Lorenz Stewart Daniel and Nelson" exposed strain A mice to tobacco smoke for several hours each day for from 25 to 250 days No increase in lung tumors was observed Stewart113 has commented recently on these results to the effect that polycyclic hydrocarbons of known carcinogenic activity are not formed in the combustion of tobacco in the presence of Orygen at certain levels His recent comprehensive survey of pulmonary tumors in ·SMOLC AND BRONCHOCENIC CARCINOYA 89 animals with particular reference to mice is documented with seventy two references which will make it clear hy a complete review of this subject would be out of scale in thus discussion However this is an approprate point at which to comment on mice and other laboratory animals as test objects for suspected pulmonary carcinogens Difficulty in Determining Pulmonary Carcinogens by Anmal Experimentation In view of the comparative ease with which alıphatic and polycyclic hydrocarbons can be tested for carcinogenicity in laboratory animals particu larly in mice it may well be asked why there cannot be a direct experimental approach to testing the suspected causes of pulmonary cancer Such attempts have been made but they have not given definite answers as yet The mouse would seem to be the animal of choicc be cause of the frequency wsth which pulmonary neoplasms develop appirently spontancously in that species But thes very frequency introduces a statistical hazard It must be shown by a sufficiently large series of animals that the number of neoplasms of the lung developing under experi mental conditions is significantly greater than in a control group Icre the strain of mice used is of utmost impor tance of strass A mice 80 to 90 per cent may des clop "spontaneouis“ pulmonary tumors while strin L mice rirely dev clop such neoplasms Vforcover those strains wluch have a relatively high incidence of "spontaneous" tumors nre precisely the ones ufuch are particularly sus ccptible to environmental factors (Shumkin") In the second place nether the spontaneously occur ring neoplasms nor the induced neoplasms of the lungs of mice bear any close resemblance to bronchogenic can ces of men In mice tumors are usually peripheral and multiple appear to be of bronchiolar or alcolar ongin arc composed of low columnas or cuboklal cells and ·90 CANCER OF THE LUNG CAUSAL, FACTORS metastasis is rare In its usual forms, bronchogenic car cinoma of man is more or less diametrically opposite for each of the features mentioned In the third place, it is probably impossible to simulate in the mouse the prolonged Ing" or latent” period wluch is so conspicuous in min A mouse at eighteen months may be as old as 1 min at fifty years as judged by the life pattern of mice That docs not mean that there have been as many cycles of cellular replacement in the bronchial tissues of the mouse as of the min There is considerable evidence, which cannot be displayed here that the cap cerogenic effect of some extrinsic factors depends upon cell selection applied to many successive generations of cells These are three important reasons why the problems of causation of cancer of the lung bv environmental agents cannot be solved within a fcw months or, at the most in a few years of animal experimentation Chronic Inflammatory or Trophic Changes, Non infective and Associated with Smoking "Cigarette Cough” In an earlier section it has been possible to exclude for the greater part infective bronchitis and pneumonitis, and particularly pandemic influenza and the post influenzal complications as causes of the general increase in pul monary cancer The same cannot be said for those chronic inflammatory, trophic, and mctaplastic processes which ure known to accompany protracted and more or less heavy smoking In the intensive search for specific car cinogens in tobacco, tobacco tar, and tobacco smoke these more general effects have received less attention than they deserve Coughing itself is known to aggravate inflammatory conditions of the bronchial and tracheal mucosa and mucosal glands, and behind the actual mech anism of coughing are those functional and structural al ·SVOKINC AND BRONCHIOGENIC CARCLOMA 91 terations which activate the cough refler Thus search for specific carcinogens may not give the answer to the in creasing incidence of bronchogenic carcinoma It may be found that physical and chemical properties of tobacco smoke which would never be carcinogenic in short term experiments but which induce dryness mild chronic in flammation and metaplasia and which frequently incite coughing are responsible This supposition is in complete accord with what is known about the effect of an extended period of risk and of intensified nsk from increased amount of tobacco smoked Co carcinogenesis In the sense that extrinsic factors act upon tissues possessing an intrinsic facility for reproductron may be said that all cancer results from co carcino gens More precisely the terms co carcinogenesis and syn carcinogenesis refer to the effective action of two or more extrinsic agents These may be synergistic in that cach augments the essential activity of the other or they may be simply additive or more frequently one may produce a preliminary alteration in the cells which makes them vulnerable to a second agent There is both clinical and experimental evidence of co carcinogenesis in a varie ty of combinations but as yet thus principle has not been used effectively in explaining the origin of bronchial car cinoma If no single carcinogen can be proved responsible the possibility of co carcinogenic action must be enter tained The Latent or Lag Period as a Bar to Immediate Deter mination of the Mode of Action of Cigarctic Smoking Throughout this discussion of the causes of primary can cer of the lung reference has been made to the essential element of protracted exposure to risk and also to the fact that cancer may not appear until soinc years after exposure has ceased This is truc of vanous ngents sus ·CHON OF TIIT LUTC CATSAL FACTORS poclei of Inducng occupitional cancer of the lung Table V is alindged from a more general tabulatıon by llue- As has been demonstrated repeatedly by the tables and fraplis which have lxen presented, the latent penod for agarctic smoling is usually not less than ten years and the intensity of the effect increases throughout life For TABLE LATTYY l'essory or OCCITATIOYAL Casam Arerage Loton Park Range of Latent Period Orol Arena Icon I un 1321 37 Commate hled cashx.t731 I rart 18 13 21 16 55-35 8-03 Igmrin radisilon Alunan from kleryki 15 Public Niralth Report's B773-779 1852. Reprobeerd dry palsson some persons the latent period may appear to be forty or fifty years This shows the frustrating futility of attempting cvaluation of varying methods of smoking or of cigarette holders or of filters, by establishing human erpenmental groups In the $3) in which most crperimental ins estiga tions are conducted [sen iſ dependable human subjects Herc at hand in sufficient numbers, the results might not kxc known for a generation To some extent the prospec tic method ulusch starts with the smoking habits of ten or morc y curs 190, avoids this delay, but as yet thuis bas gisch a definite anser only to the main question LUYC CANCER AMOLG NOY SVOKERS Primary pulmonary cancer occurs among those who are clused as non smokers The definition of a non smoker has not been uniform If it is taken to mean a person who ·SVOKING AND BRONCHOGENIC CARCINOVA 93 has never smoked one cigarette one pipeful of tobacco or one cigar in his lifetime practically all males and most females would be excluded Doll and H.11° considered a practical definition of a non smoker to be a person who has never consistently smoked for as long as one year at the rate of as much as one cigarette or one gramme of tobacco a day The effect on a non smoker of exposure to the tobacco smoke of others has, as yet received very little serious consideration Statistical investigation will cncounter many difficulties Some years ago the rhetorical question was posed by Lickint116 whether the non smoking wife who had lived many years with a heavy smoker was still a non smoker for both clinical and statistical purposes At the symposium on the endemiology of cancer of the lung at Louvain in 1952 the following dialogue'it oc curred Dr Hammond Are you suggesting that the control group should never have been crposed to tobacco smoke? Dr Massin This would not be possible in a civilized community I have never smoked but look how I am er posed to tobacco smoke just now When detailed answers are known the passive expo sure of the non smoker may be found to be at least a co factor in the production of bronchogenic carcinoma but any suggestion now to that effect is highly speculatie Using the definition of a non smoker previously cited from Doll and II:]]Doll'18 attempted a statistical invest: gation of cancer of the lung in non smokers The diffi culties in such a project are obvious in view of the proportionately small class of non smokers and the er tremely small class of non smokers with cancer of the lung The author commented on the "bold assumptions which were necessary and suggested that the incidence ·CANCER OF THE LUNG CAUSAL FACTORS of lung cancer in non smokers may be the same in men and in women, and in residents in areas of difcring density of population Environmental factors other than those per tuning to smoking should bear upon non smokers and smokors like Doll concluded, further, that one in five of lung cancer deaths in persons aged twenty five to seventy four years in 1950, may be attributible to causes other than those associated with smoking CourAKTIVE Risk OF DEVELOPAG CANCER OF THE LONG FOR SYOKERS Avy NO SVOKENS Under the appropriate topic I have included reference to the work of Cutler and Loveland" who found that of 2,000 white males born in 1910 and still living in 1950 43 may be expected to have developed cancer of the lung by age cighty Hon will thus incidence be apportioned between smokers and non smokers? As Cutler and Love land pointed out if the risk of developing cancer of the lung were 1 in 100 000 for non smokers and 1 1 10,000 for smokers, the entire question might be regarded as trivial However, when the sum of these risks must ap proximate 13 in 1 6o determination of whatever differ ence may exist becomes very important By an extension of the method used by Comfeld,118 and with his assist roce Cutler and Loveland have developed an answer which is embodied in Table XII The data on smokers and non smokers were obtained from tliree comparativ studies to which reference has been made previously These were the work of Sndowsky, Gilliam and Com field,' of \Vynder and Graham, as ind of Doll and Hills Together, they provided information on 2 434 cases of can cer of the lung and 2 729 non cancerous controls When separately analyzed the three studies were in agreement as to general trends but differed somewhat in the predicted number of cases in the non smokers Two assumptions ·SMOKLIC AND BRONCHIOCENIC CARCLOVA 95 were necessary to complete the analysis One was that the control group was representative of the smoking charac teristics of the general population from which the lung cancer patients were derived The other had to do with establishing cigarette equivalents for cigars and pipe smoling Neither assumption would appear to result in TABLE VII CASES OF LUNG CANCER EXPECTED TO DEVELOP PER 1000 11 TE VIALES 40 YEARS OF AGE Smoking Class By Age 60 By Age 70 By Age 60 B 4 Non smokers Smokers, daily of pack or less to I rack More than 1 pack 18 0 05 49 80 AP persons 53 27 12 (After Cutler and Loveland') more than minor alteration in the number of cases pre dicted even if it were not entirely correct The results of this investigation not only showed that in the respectnc age groups the liability to lung cancer of smokers in general was from four and one half to six times that of non smolers but it indicated very clearlt the cf fect of wount and of duration of smoking The added ycars at risk are very important and the effect is cumula tise even in the eighth decade of life Thuis is the answer to the fatalistic attitude that any carcinogenic cffect has been exerted in the first decade or two of smoking BRONCHIOCESSC CARCINOVA AT TIE VD-CENTURY STATISTICALLY ESTABLISTIED ASSOCIATIONS Apart from the greater lizbility to bronchogenic cancer in ccrtain occupiuonal groups which accounts for but a small part of the total incidence strong statistical cui dencc has been presented for association between costain ·96 CA\CER OF THE LUVC CAUSAL FACTORS human inbutes and primary cincer of the lung These attributes rc maleness, urban living industrialization and the smoking of tobacco, particully as cigarettes Alalencss Mileness is a biologic quality, presumably has not clunged in the last half century Yet the propor tron of males developing cancer of the Jung has increased faster than the proportion of females de cloping cancer of the lung although both ure increasing This alone cstablishes that there must be an extrinsic agent If maleness is significant it must be as ? co circinogen In 1929, I wrote with complete confidence that if smoking was a cause of bronchogenic carcinoma there would be a trend toward sex equality in incidence in another generation As yet such cquality has not developed In fact the gap has wadencd However more is known now about the im portance of the latent period and of prolonged vents of risk than in 1929 It is still evident that a much greater proportion of miles have been heavy smokers for twenty or more years than of females As appears in Table VIII Doll and Hilfae found only 14 women who smoked more than 25 cigarettes a dry in the combmed cancerous and control groups, which torilled 216 If the smoking habits of the women had been like those of the 2,714 men 40 would have been crected In other words, the Prarent cffect of maleness cannot be separated even today from the presumptive effect of smoking Urbin living and industrialization include the concen tration of motor vehicles and of increased sources of air pollution The many carcinogenic possibikties therein cin not be cast aside completely Some of the known carcino gens, even in minute amounts are potent producers of malignant neoplasms elsewhere than in the lungs There is evidence for circinogens not yet identified in the prod ucts of incomplete oxidation as found in the exhaust of internal combustion engines It is an arresting fact that ·SMOKING AND BRONCHOGENIC CANCIOVA 97 arguments for the importance of tobacco which are based on economic data such as increasing national consump tion apply almost equally well to the automobile But it is true also that the manner and degree of use of tobacco vary between urban and rural living and that long term Smoking and Pulmonary Cancer Assoc ot on Proved-Ceusal Relation Presumptne Causal Cut I I dont NI MU tello Different Romerd do 10 Annie Chr Wenty R. Um Trop Cheng Cardin Motocletiyerto Copo of com Cambrities 1777 Ficune 20 If the association between smoking and pulmonary cancer should prove to be causal the relationship of intrinsic and extrinsic factors might prove to be that shich is shown in this schematic diagram Symbols have the same significance as in Figures 13 and 19 crcessive cigarette smokers are in proportionately greater numbers in luger communities Tobacco chewing and pipe smoking in the villages and on the farm hai c not been replaced by cigarette smoking as they have in the city It is concen able that the association with urban living and industrialization mat lse rcsoked on the basis of cigarette smoking although evidence on that score falls short of proof Smoking of tobacco particularly the excessive smoking ·98 CANCER OF TTIE LUNG CAUSAL FACTORS of cigarettes over 1 long period, is associated definitely with the occurrence of bronchogenic carcinoma accord ing to the results of several different methods of statistical investigation Association does not necessarily imply a causal relation, but there has been no other satisfactory explanation why this association should prevail Non smokers develop pulmonary cancer and many heavy smokers escape pulmonary cancer, but for a long term heavy smoker the chance that he will develop cancer of the lung by age cighty is of the general Order of fifteen times greater than for a non smoker (Cutler and Love landa?) In about four fifths of the cases of primary can cer of the lung association with tobacco smoking appears to be a significant factor (Dolf118) If the association between tobacco smoking and bron chogenic cancer is accepted as of causal significance what the specific cruse may be remains unknown Arsenic the known and unknown carcinogens resulting from incom plete combustion of organic matter, and the complex of factors which produce those dystrophic changes which may find expression in cigarette cough” are the chief suspects One of these may act alone or with others through a relationship of co carcinogenesis or of syn car cinogenesis We can look to the future with assurance that, if the relationship is causal, the nature of the clusal agent eventually will be determined Its identification or perchance the discovery of new evidence which will point the finger of suspicion away from smoking is a challeng ing possibility for each and every physician The solution may come through a combination of the prospective sta tistical method and intensive case studies With such knowledge prevention of four fifths or more of the pres ent mortality from bronchogenic carcinoma would be a reasonable expectation ·VI EPICRISIS IN THIS DISCUSSION of the historical and practical ap I have made every effort to gue as complete and for a presentation of the evidence as the scale of treatment made possible I have endeavored to present the evidencc in an unbiased manner so that each reader could form his own conclusions as the evidence unfolded I have not been an advocate for any particular view and certainly I have no intention of being a crusader My own views have undergonc change through the years I early subscribed to the then prevailing opinion that protoplasm was relatively stable and the chromo somes particularly so The rate of incidence of cancer in any organ was considered to be a fairly constant function not readily influenced by environment although numer ous occupational carcinomas of the shin had been demon strated At first it scemed improbable to me that thic increase in cancer of the lung was something new and I joined others in attempting to explain it by the aging population by the advent bf radiography by clinical awareness and by better diagnostic methods in general Thcsc cxplanations eventually prosed inadequate and it was necessary to admit that some recently acquired fea turc of our way of life was very rapidly changing the inci dence of pulmonary cancer Similurly, I have been resistant to 1 ready acceptance of the association between cigarette smoking and bronchogenic carcinoma I have searched the literature for other 99 ·100 CANCUN OF THE LUNG CAUSAL FACTORS reasonable explinations or for recognizablc fallacies I hue found none of importance As of today, I must agree with many of the specialists in statistical analysis and in the endemiology of cancer, that this association has been established What is the next step? When will it be not only proper but rcqusite that the medical profession take cognizance? Those ure questions which I cannot and should not at tempt to answer at the moment I can houerer, gc you the answer of J H Maisin and ] Clemmesen, who were editors' i for the report of the Louvain Symposium on the Endemiology of Cancer of the Lung In the final para griplis of their editorzıl they commented that it was about 150 years after Perci all Pott cxplained the etiology of chimney suceps cancer and prescribed methods for its prevention before carcinogens were demonstrated in soot and continued May we show the same practical sense as our forefath crs and not look for direct proofs which are out of reach before we transmit esperience into practical meas tires ·VII REFERENCES 1 Ilveper WC Occupational Tumors and Allied Discases Spring field m Charles Thomas Publisher 1942, 890 pp 2. Weller CV The pathology of pronary carcinoma of the lungs Arch Path 7 478-519 1929 3 eller CV The pathology of certain signs and symptoms in primary carcinoma of the lung ilustrative cases Ann Ini Med 725-740 1028 29 4 Farber S M Lung Concer American Lecture Senes no 187 Springâield 1 Charles C Thomas Publisher 1954 157 PP 5 Knights EM Jr Aletastatic tumors of the brain and their rela tion to primary and sccondary pulmonary cancer Cancer 7 259 285 1954 6 Weller CV Entdifferentiation in primary carcinoma of the bron chu and lungs I Cancer Research 13 218-238 1929 Dorn IF Morbidity and mortality from cancer of the lung in the United States Acta Unto Internat Contra Cancrum 9552 56L 1953 8 Levin ML The occurrence of lung cancer is man Acta Unlo Internat Contra Cancrum 9 531-511 1053 0 Lew E A. Trends in respiratory cancer mortality Statist Bull sictror Life Insur Co 35 3.8 1954 10 Lorteweg 1. Some remaris on the age cunc lo lung cancer Acta Unlo Internal Contra Concum 9 529-530 1933 11 Lendaway E L and Waller RE Studies on cancer of the lung Acta Unto Interat Contra Cancrum 0:485-491 1950 12 Lennaway N Veod Lennaway E L A study of the incidence of cancer of the luns and larynx Ilye 36 236-267 1938 13 Lennanay EL. and hennaway A further study of the Incidence of cancer of the lung and Liryar Dal Cancer I 260- 238 1947 14 Stecle C Bronchogenic carctooma analysis of 201 prosed cascs Ann Otol Rhin & Laryng, Q75-21 194 15 Moore A. L, and Barr ML\uclear morphology according to $ck, in buman tissues Arta anat. 21 191 08 1934 10 Doll, Bronchial carcinoma incidence and setiology BNP ). 3.1.527 585 500 19 * Cutler S )and Loveland D B The risk of developing fung 101 ·102 CANCER OF TIIE LUNG CAUSAL FACTONS cancer and its relationship to smoklog | Nat Cancer Indt 35 201 211 1954 18 Public Health Sen ice US Department of Health Education and Wellare Cancer Illness is Ten Urban areas of the United States Cancer Morbidity Ser 1950-52 Reports 1 30 19 Levin M L Etiology of lung cancer present status New York State Med 54 789 777 3951 20 Weller CV Jatrinsic factors in the etiology of Beoplasms Am I Cancer 30 39-40 1937 21 Barker LF Polyposis of the colon M Clin North America 14 77-85 1930 2 Weller CV The inheritance of retinoblastoma and its relation ship to practical eugenics Cancer Research : 517-535 1911 23 Cohart EM Socioeconomic distribution of stomach cancer in New Haven Cancer 7 455-481 105-1 24 Middleton US Physiological Methods in Chmical Practice The Beaumont Lecture Wayne County Aledical Society February 2 1953 Springfield Ill Charles Thomrs Publisher 1954 40 FP 25 Purchan A and Six I Cancer of the lung in the miners of Jáchymov (Joachimstal) report of cases observed in 1929 1930 Am J Cancer 16 631 72 1932 20 Floover HC and Hoover L II Translators of Agricola Ceot gius De Re Metallica 2558 638 pp New York City Dover Publications Inc 1950 (Reprinted The Mining Magazine London 1912) 27 Scheher CL Die Gesundheit der Bergleute Chemnitz 1770 (Cred by Uhlig) 28 Harting T H and Hesse W Der Lungenkrebs die Berghsaak hert in der Schaccberger Gruben ferteljahrschs gerichti Aled u offentl Sanllatsız esen 30 298-309 (NF) 1879 Ibid 31 102 132 313-337 1879 29 Uhlig Al Ober den Schneeberger Lungenkrebs VirchouArcs paih Anat 230 70 38 1921 30 Peller S Lung cancer among mine workers ia Joachimsthal Iluman Blol 11 130 143 1939 31 Arastein A Veber den sogenannten "Schneeberger Lungenkrebs * Vien klin #chnisch 26 748 752 1913 32 Schmori G Uber des Schneeberger Lungenkrebs erhand? deutsch path Gesellsch 19 193 196 1923 33 Thiele Rostos Saupe and Schmorl Veber den Schneeberger Lungenksebs lunchen med il chosch 71.24 25 1924 34 Saupe E Der Lungenkrebs der Schneeberger Bergleute Bed Welt 2 1800 1830 1928 35 Hostoski o Saupe E and Schmor] G Die Bergkrankheit der ·REFERENCES 103 Erzberglente in Schadeberg in Sachsea ("Schneeberger Lringen krebs*) Ztschr krebsforsch 3 360-384 1928 36 Lawy ) Über die Joachimstaler Bergkrankheit. Med Run 5 141 142 3929 37 Srld 1 Ober den Lungenkrets der Bergleute in Joachimstal (Tschechoslowakei) Ztech Asebsforsch 39 609 633 1930 38 Sk! IL Le cancer pulsionalre chez les mineurs de Jáchymov (Joachimstal) Pressemed (48} 673-674 1938 39 Kotx M O The association of Blucosis and carcinoma of the lung Am ) Cancer 35 38-49 1939 40 Vorwald A. and hart J V Poeumoconuosts and pulmonary carcinoma Ann Path 14 49-57 1938 41 Schulte G Pneumokoniosen des Ruhubergleute und Lungenkarzi xom Fortschr Geb ADnigenstrahlen 41 444-445 1930 42 Pamett C P Annual Report of the Chief Inspector of Factones for the Year 1947 London, His afajestys Stat Of 1918-49 PP 79-81 43 Amor A Cross of the Respiratory Tract Bericht aber den VIII Internationalen hongress fur Unfallmedizin und Berufs krankhestes Fraphhut & i Segteruber 28-30 1938 Laipzig G Tbienne 29441 382 1939 41 Smuth WE An evaluation of claims for Occupational factors in cance of the lungs Acta Unio Internat Contra Cancrum D 476. 484 1953 45 Currie A The role of arsenic in carcinogenesis Br? Y Bull 4 402-405 1916-7 46 Neubauer 0 Arsenical ancer a review Bn Cancer I 193 351 18:17 47 (heger 1 Chemical carcinogenesis a review Brit 1 Indust Bleu 6 123 19:19 48 Mill AB Fanning E L Perry A Bowler R G. Buckel, 11 M Druett II & And Schulling RSF Studies la the lockdienos of cancer in a factory handling inorganic compounds of arsenic Brit Indus! Ved 5115 1913 49 Ludewig P and Lorenses E Untersuchung der Grubenloft is den Scharelerger Gruben auf den Gehalt Ar Radiomeranation Strahlentherapke 17 428-435 1924 50 Rajewsky B Bericht über die Schneeberger Untersuchungen Zeschr Krebsforsch 19 315--340 1939 51 Bhounek F And Fort i loschimstaler Berginannskrankbelt Si ahlentherapie "O 497-49S 1911 52. LOTEZ, E Radioactssity and lung cancrt acrilical sestaw of lang cancer in the minert af Schneeberg and Joachimsthal lat Cancer Inst. 3 1 35 19 143.

·104 CANCEN OF TIIE LUNG CAUSAL FACTORS 53 Martland HS The perirrence of malignancy in radioactive persons am Cance 15 2 135.2510 1831 54 Hueper HC Neront des clopments in environmental cancer AVA Arch Path 58 300-399 475-521 1951 55 Turner 1 and Grace C An Incstigation into KNOCE mortality among males in certain Sheffield trades ) Ilye 38 90 103 1938 50 Alxens W., Bauke EE and Jonas Auffallende Igufung von Bronchialkrebs bei Arbeitern der chemischen Industrie Vünchen med Mchnickar $ 455-487 2030 57 Machle 11 and Gregorius F Cancer of the respiratory system in the United States chromate-producing industry Pub Health Rep 63 1114 1127 1918 58 Imprescia S Bronchogenic creinoma and chromates Dá Chest 213-47-355 3052 59 Mancuso T F and Ilueper x C Occupational cancer and other health hazards in a cliromałe plant a medical appraisal 1 LAOS cancers in chromato workers Indust Vod 20 358-303 1951 60 Brinton XP Frasier E $ and hones A L. Vorbidity and mortality experience among chromate workers Pub Hicalth Rep 07 835-817 1952 OL Cafaler Mand associates Health of workers in chromate producing industry Public Ilealth Service Publications MO 192 Federal Security Agency Public Health Senice 1933 62 Bidstrup P L Carcinoma of the lung in chromate workers But 1 Indust Med 8 302-305 1951 63 lyers il Asbestosis Post Grad 1125 831-838 1919 64 Hueper ! C Occupations} ssd Environmental Pulmonary Can cer with Special Reference to Pneumoconiosus The Proceedings of the Seventh Saranac Symposium on l'neumoconiosis Sep tember 22 26 1932 as Medier 11 Asbestose und Lungenkrebs Deutsche med Wchnschr 69 575-576 1913 66 Isselbacher I Klaus Hand Hardy HL Asbestosis and bronchogenic carcinoma Am I Med 15 721 732 1953 67 Merewether ERA See Barnett B8 Lynch X and Cannon 1 M Asbestosis VI Analysis of 40 necropsied cases Dis Chest 14 874 885 1919 89 Clayne SR Pneumoconiosis A histologic survey of necropsy material za 1205 cases Lancet 1 810 814 1951 TO Pott, P Chirurgical Observacions Relate to the Cataract the Polypus of the Nose the Cascer of the Scrotum the Different Ninds of Ruptures and the Stortification of the Toes and Feet Printed by T ) Carnegle 1775 Hawes Clarke and Collins London ·REFERENCES 105 71 Henry SA and Irvine E D Cancer of the scrotum in the Blackburn Registration District 1837 1929 I llyg 36 310-310 1936 72 Neve E F One cause of cancer as illustrated by epithelioma in Kashmir Brut MI J 589.591 1910 73 Neve CF The causation of cancer Practitioner 109 355-359 1929 74 Volkmann R. Theer Paraffin Russkrebs (Schornsteinfegerkrebs) in Beitrage zur Chirurgie usw Leipzig Breitlopf u Hartel 1875 370 PR 75 Yamagiwa k and Itchikava k Erperimentelle Studie uber die Pathogenese der Epithelialgeschwulste Afitt med Fakult Unio Tokyo 15 295-344 1916 76 Yamagiwa . and Itchikawa Evp rimental study of the pathogenesis of carcinoma } Cancer Research 31 29 1918 77 kennaway E L Experiments on cancer producing substances Brit MJ 2:1-4 1925 78 Mayneord w V (1927) unpublished cited by Cook Ilewett and Hieger ** 79 Cook J W Hewett CL and Flicger I The isolation of a cancer producing hydrocarbon from coal tar Parts I II and III J Chem Soc 1 395-405 1933 80 Doll R The causes of death among gus workers with special reference to cancer of the lung Brit I Indust Med 9 180 183 1902 81 Hueper WC Air pollution and cancer of the lung Rhode Island MI 3G 21 et seq 1953 82 Hammond EC Lung cancer and common inhalants Cancer 7 1100 1108 1957 83 Clemmesen) Nielsen A and Jensen E Mortality and ina dence of cancer of the lung in Denmark and some other countries Acta Unto Intemat Contra Cancum 9 603-636 19.3 81 Goulden F and Tipler M M Experiments on the identication of 3 4 benzpytene in domestic soot by means of the fluorescence spectrum Brit ) Cancer 3 157 160 1919 85 Stocks Þ Endemiology of cancer of the lung in England and Wales Brit J Cancer 6 99 111 1952 80 Kotin P Fall L. Hader P. and Thomas V Aromatic hydrocarbons I Presence in de Los Angeles atmosphere and thic carcinogenicity of atmospheric extracts A VA Arch Indust Ilus 9 153-163 1934 87 Lotin P. Fall, I L and Thomas An matic hydrocarbons II Presence in particulite phase of gasoline engine exhausts and the carcinogenicity of exhaust extracts AJI A. Arch Indust llys 9 164 177 1931 ·CANCER OF THE LUNC CAUSAL FACTONS 89 Goulden F Kennaway E L and Urquhart ME Arsenic in the stespended matter of town air Brit Cancer 617 1958 89 Dawson AB Radioactive material in the atmosphere But ) Cancer 6 20 31 1932 90 Walles RE The benzpyrene content of town air Brit Cancer 682 1952 91 Brrenblum 1 and Schocatal R Carcinogruc constituents of coal tar Brit Cancer 1 157 165 1917 92. Perret } M Primary intrathoracic malignancy eight cases Neu Orleans NS 80 213 228 1027 93 Muler F11 Tabakmissbrauch und Lungencarclaors Ztsch Arcbsfarsch 49 57-85 1939 94 llammond EC and lom D Tobacco and lung cancer Acta Unio Internat Contra Canctum g 507 511 1953 05 Wynder E L and Graham E A Tobacco smoking as a possible etiologic factor in bronchioge cuscinoma A study of 884 proved cases JAVA 143 329-338 1930 00 Doll Rand Hill A B A study of the aetiology of carcinoma of the lung Brit 1 1271 1280 1952 97 flammond EC and Horn D The relationship between human sinoking habits and death rates & follow up study of 187 778 men JAVA 155 1318 3328 1934 99 Doll Rand Full AD The mortahty of doctors in relation to their smoles, habits Brte Al 1 i 1451 1455 1954 99 Anon "orcigo letter / SMA 356 66 67 2954 200 Barksdale E E Cutaneous manifestations from tobacco with special reference to arsenical exfoliatrc dermatitus JAMA 115 672-676 19-20 101 Carratalá Rand Vucetich, DC Presencia de Arsénico en el tabaco Investigacióa esperimental Prensa med argent 28 2181 2185 19-11 202 Thomas MD and Collser TR The concentration of arsenic in tobacco smoke determined by a rapid titrimetnc method ] Indust Hyg o Toxicol 27 201 206 1945 103 Dalf ML and hennaway E L The arsenic content of tobarco And of tabaco smoke Boll / Cancer 4 173 162 1950 104 Daff ME Doll Rand Lenngway EL Cancer of the Jung in relation to tobacco Bot Concer 51 20 1951 105 Dungal N Lung carcinoma in Iceland Lancet 2 245 217 1950 106 AIL Nally SO The tar in cigarette smoke and its possible of fects Am Cancer 18 1502 1514 1932 207 Bogen E and Loomis R N Tobacco tur an experimental 10 vestigation of its alleged carcinogenic action Am J Cancer 16 1515 1521 1932 108 Lamb FW and Sanders E The role of tobacco smoking in ·REFERENCES 107 the production of cancer Ul Animal experiments ) Hus 32 298-300 1932 109 Flory CA The production of tuniors by tobacco tars Cancer Research 1 282 276 1911 110 Grahas EA Tłynder E Laod Croninger A B Cigarette smoking and cancer of the lung (Abstract) Science 116 521 522 1952 21 Vacher L end Schmucke a Experimentelle Vatersuchungen zur kausalen Cenese atypescher Epithelwucherungen fünchen med 1 chaschr 59 1807 1813 1830-1882 1911 119 Lorenz E Stewart, 1 L. Daniel and Nelson C9 The effects of breathing tobacco smoke on straia A mice (Abstract. Cancer Flescarch 3 103 19-13 213 Stewart HL Pulmonary tumors to animals with particular reference to mice Acto Onio Intemat Contra Concrum 9 512 528 1953 114 Shimkın MB Induced pulmonary tumors in mice Suscepta bility of seven strains of mice to the actson of intravenously inject ed methylcholaathree Arch Path 29 2.9 238 19:10 Induced pulmonary tumors in mice II Reaction of lungs of strain A mke to carcinogenic hydrocarbons Ibid 29 239 055 1910 115 llueper WC. Age aspects of egyironmental and cupational cancer Pub Ilealth Rep 67 773 770 1932. 316 Lickint, F Der Bronchialkrebs der Raucher Munchen. med TVchnschr 83 123 1234 1933 117 llammond E C. and Valsun, IL Acta Unlo Iatemat Contre Cancrum 8 510 19-3 118 Doll R Mortality from Jang cancer among non-smokers. BH! } Cancer 7 303 312, 1953 119 Comfeld ) A method of estimating comparative rates from chini cal data Applications to cuocer of the lung brcast and cervix Mat Cancer Inst 11 1269-323 1950-51 190 Sadowsky D A. Gillum AG and Comfield J The statistical association between sraoking and carcinomi of the lung 1 lat Cancer last 13 1297 10.9 190_53 221 Alaisin, J II and Clemmesen, Editors The eademiology of cancer of the long Ada Unio Internof Contra Canctum 47 439 193 ·ܪ ܕ ·INDEX British physicians prospective Adenocarcinoma bronchogenic 81 method of assayrag stoking 83 Adenoma bronchial, 5 Bronchial adenom1 5 Agricola C 30 Sronchitis chronle non infective Alwens 151 90 American Cancer Society (zvesti Bronchogenic caruinomn gastos by prospective method 8.2 age Incidence 15 21 Amor AJ 40 associations statistically estab Animal experimentation lished 95 screbral metastases 6 culty in determining : monary carcinogens by 89 clinical types 6 Latent period 90 general leatuses 6 Arnstein A 35 general incidence 9 Arsene histologic features 7 ar causal factor 41 58 99 histologic types and sex 81 Atmosphers: D histologic types and smoking 81 in cigvettes 88 incidencc affected by age dus La tobacco 85 inbution 12 in tobacm smoke 85 incidence age 15 21 in Turkish tobaczyt 88 incidence Dennusth BI 65 Asbestosis 53 faciümce Great Britain 17 20 incidence New lork State 9 B incidence ractal, 10 Baskes F. incidence set, 37 Barksdale EE 85 incidence Sheffield SI Barnett, CP 40 53 54 focidence tu enticth century 19 Bart Vt Ln 20 Incidence United States, o 19 Baoke EE 51 Bthounck F 41 incidence increasing 10 Denzpyrene in non sinokets 02 In atmosphere 72 Ludwigshafen 51 in engine Exhaust, 70 reasons for a parent increase 12 in root, 88 risk of developlog 21 Berenutom 1.3 Search for artet in twrptirth Bismuth, a causal fastor 40 century In Artrospect, 73 Bituminous surfaced roads 69 signs and symptoms # Bogen, L. 85 tobacco straling 6-99 Duinton, 11 P., 52 Bronchogenic carcinoma ant to 109 ·110 CANCER OF THE LUNG CAUSAL FACTORS bacco smoking association es tablished 81 84 Dorn HF 9 17 10 20 Dungal N 87 E Engine erhrust 69 Falk n L 73 Farbcs S M 7 18 Flory CM 88 Fott M 44 Foundry workers 51 Trasier ES 52 G с Cancer of the lung equivalent to bronchogenic carcinoma 5 Cancer of the scrotum 57 Cancerogenic hydrocarbons iden tification of 61 Carbon black from tires 69 Carcinoma of the stomach, 20 Carratall R 85 Casuistic method of evaluating smoking 70 Cell types in hronchogenic carct noma influcnced by set 18 Chimney su ceps cancer 57 Chromate workers Germany 51 Great Britain 53 Vatted States 52 Cigarette cough 00 98 Clemmesen) 01 05 100 Cobalt, As causal factor 40 Co-carcinogenesis 01 Cohart CX20 Collier TR 86 Colonic polyposis 4 Comparative method of evaluating smoking 77 Cook W 62 Cornfeld } 34 Croninger A B 88 Currie AN 42 Cutler Sy 21 22 944 99 Gafaler (ct of I 52 Gas workers BR Cilliam AG 04 Coulden F 68 70 72 Grace HC 51 Gmham E A 79 80 81 88 94 Great Britain density of population 65 incidence of bronchogenic car cinoma 17 18 Gregorius F 52 H Harting F H 31 41 76 Ilammand EC 67 79 82 93 Hardy II L 55 Henry S A 80 Jesse W 31 41 Ilewett CL 62 Hieger 1 42 62 Hull AB 79 80 83 93 94 96 FA B (ct all 42 Histology of bronchogenic carci noma 7 Floover HC And Hoover LR 30 Hom D 79 82 Hueper WC 4 38 39 40 41 45 52 53 55 61 92 D Daff ME 80 Daniel J H 88 Dawson RB 79 Denmark 65 Density of population 65 Doll R 20 63 86 87 79 80 83 88 93 94 96 98 ·INDEX 111 I Imprescia S 52 Incidence density of population 65 industnalization, 65 urbanization 65 Incidence of bronchogenic carci doma see bronchogenic car cinoma Industrialization 65 96 Influenza as causal factor 49 Internal combustion engines 69 Intrinsic and extrinsic causal fac tors definition, 23 Ionizating radiation atinospheric 71 Ionizing radiations 34 35 43 56 Irvine E D 60 Isselbacher LJ 55 Itchikawa K 62 Latent period 50 92 prevents prompt identification of causal factors 91 Leeds University of report on radioactivity of tobacco 85 Levin ML 9 10 23 79 81 Lew E A 10 15 Lickint, F 93 Lowy J 35 London 66 Loomis RN 88 Lorenser E 43 Lorenz, E 44 88 Los Angeles study of air pollutants 70 3 4 benzpyrene in air 73 Loveland D B 21 02 01 99 Ludewig P 43 Ludwigshafen chromate workers 51 M Machic 1152 Mlader F 73 Maisin J H 93 100 Maleness in incidence and caus1 tion of bronchogenic carci noma 20 54 96 Mancuso T F 52 Martland, HS 45 Mayncord 1 V 62 VcWally W D 87 Vietal grinders 51 Metropolitan Life Insurance Corn pany 10 11 15 Middleton, WS 2 Miners cancer 27 Voore KL. 20 Vorçın 40 Muller F II 78 J Jagzichte 5 Jensen E 64 65 Joachimsthal 27 Jonas W 51 K Kangri carcinoma 60 Larr J W 39 55 Kennaway E L. 18 17 18 38 42, 62, 63 65 70 86 Kennaway NM 17 18 38 42, Klaus 11 55 Notz, M O 39 Knights EM Jr 6 Lorteweg 10 Lotin, P., 69 Loven, A L 52 L Lamb r VI 83 Larynx carcinoma of 68 National Cancer Institute 22 Nelson, C. V. 89 leubauer 0., 41 ·112 CANCER OF THE LUNG CAUSAL FACTORS spect, 46 Neve IT60 Schmor! G 35 New York incidence of broncho Schneeberg 27 genic carcinoma 9 Schneeberg and Jáchymov possi Nichel as causal factor 40 36 ble causal factors in retro Nielsen A 64 65 Non smokers passively exposed 93 Schocntal R 73 o Sheffield metal workers 51 Occupation 50 Shimkin VB 89 chromate workers 51 Signs and symptoms of broncho foundry workers 51 genic carcinomi 6 Sikl H 30 35 37 gas workers 62 latent period 50 Silicosis 37 metal grinders 51 Smith W E 40 Smoke domestic and industrial 69 miners in the Erzgebirge 27 Organic matter incomplete com Smokers comparative risk of de bustion of 57 veloping cancer of lung 94 98 Р Smoking habits and bronchogenic Peller S 31 Curcinoma 82 97 Perret J 176 Smoking machines 87 Parchan A 30 35 Smoking of tobacco 76 98 Pneumonitis as causal factor 39 if causal presumpure factors 97 49 Soot contains 34 benzpyrene 68 Population density of 65 Steele CH 18 Pott P 57 Stewart HT 88 Prospertive method of evaluating Stocks P 68 smoking 81 T R Tar cancer 61 Radioactivity 34 35 43 58 Thiele 35 of tobacco 85 Thomas M 73 of tobacco smoke and ash 85 Thomas MD 86 Rajewsky B 44 Tipler M M 69 72 Retinoblastoma 24 Tobacco smoke esperimental use Risk of developing cancer of lung 88 smokers and non smokers 94 Tobacco smoking 78 98 Rostoski o 35 duration of 81 Intensity of 81 s Tobarro tar esperimental applica Sadowsky DA 94 tion 88 Sanders E88 Turner HM 51 Saupe F 35 U Scheller C L 31 Schmincke A 88 Uhlig M 33 43

·INDEX
113
Vorwald A J 39 55
Vucetich D C 85
14
United States
chromate workers 52
effect of urbanization 67
incidence of bronchogenic car
ciom 9 19 20
University of Michigan Decrop
$ies on bronchogenic carcı
noma 12 13
Urbanization 65 96
Urquhart, ME 70
Wacker L 88
Waller RE 16 65 72
Vedler HV 54
Weller CT 4 7 10 24 96
Wyers H 53
Wynder EL 79 80 81 88 91
Y
Yamagiwa 82
Volkmann R 62

·一

·This Book
CAUSAL FACTORS IN
CANCER OF THE LUNG
By
CARL V WELLER, MS, MD
ulas set and printed by The Cold Bell Press Inc, of
Fulton Alissours It tas bound by the Pantagraph
Printing & Stationery Company of Bloomington Ili
nous The engravings u ere made by G R Grubb and
Company of Champaign, Ninows The page trim size
is 5' 8 inches The type page is 23 x 39 picas The
type face is Caledonia sct II point on 13 point The
text paper is 70 lb Carfax Eggshell The cover 25
Palco Leride No 25 Embossing 32 Finish TT Black
BOKS
Thomas
YMBOLS
With THOVAS BOOXS careful attention is given to
all detasis of manufacturing and sign It is the Pub.
lishers desire to present books that are satisfactory as
to their physical qualities and artistic possibilities
and appropriate for their porticular usc THOVIAS
BOOKS will be truc to those laus of quality that
SSSUTE a good name and good will

·No text on this page.

·No text on this page.