Document QE15dKQ7bOE0Xe6Mj4bqEbZ8

FILE NAME: Saranac 7th Symposium (SSY) DATE: 1952 DOC#: SSY018 DOCUM ENT DESCRIPTION: Transcript of Presentation by Dr. Levin with Discussion BI JPCKRj m miMdc you, Doctor I&cada. Ladies and g&atlsxon, I es supposed to talk to you on tiia pbilosppby of the teioatatistics of anvirctraaatal pulaormry c&noer, and I hs.ro nav-sr boon gore* up until tfala soaent, whether I ^aa supposed to concentrate on philosophy or Mostatistics or environmental pulmonary cancer. I shall probably try to do all throe, and succeed in neither of thsa or not one. I eartainly would like vary such to second the rsss&rka of Doctor Jv.-ads as sssde and perhaps I can do so and sake ocas ccsoosats oa ac~s of the things ha has said, by tailing you that ?sy m a point ox view is in this satter. It is primarily that of an pldaasiologlst hih is the field ia hieh I sraceivod scat of sy s&dieal training* In epidaoiolcaiet, a* you probably law, Is not a atat-ietieiaa Eieesaarily, although he teas to Zaiatr ccaathing about statistics and use then and knew hen h is ia trouble, scrae - eo tea can go to eoffiaouB t&o is a etatiistielan. 3k5 An epidemiologist is* primarily* a fellow who is interested in two questions* first of all*whether a certain group of people actually do or do not have more of a cer tain disease than other groups do have* He ! interested in knowing whether thats really so, and so he is rather crit ical of things which may look like that, and neverthelesa are not really good evidence, and he is interested in know ing, if that is so, what the reasons might be, and his reas-f i eng for those interests are quite practical, because be is the man who is usually called upon to make certain decis ions or to give certain dvice as to whether certain things should or should not be done in a given situation* Eov* it just happens that historically, that type of question usually arose with our epidemic diseases and that's why, I suppose, the term epidemiology* is usually thought of as referred to with people who deal with infec tious diseases* but precisely the same question can arise In almost any ,ype of disease, and certainly they do arise In .industry* " Eow, the eold;-uiologlata point of view on pul- I i aionu:*y cancer and the vidonee reapsctJng various etiolc-- j gical agents is, I would say, highly crit' 'al. That is, j :o ato aware from lung experience, that a lot of things j which look like etiological relationships are not nsces- ; sarily auch* The other point of view, I would say, perhaps would seem to you to be opposite, but I would offer to you ae the result of long reflection on the part of a good J ; many people* We epidemiologists suspect that, from the stand point of the causation of pulmonary cancer, our best exper imental animal may be man rather than any animal we can study in a laboratory. Now, ~an is not a very good ex perimental animals but it seems as if in cancer, he may be our only one in some forms, and the laboratory may or say not confirm what we find in man, but we have a strong feel ing that we are missing a pretty good bet in not studying much more intensively, possible relationships which may be etiological to man* We have about twenty-five thousand cases of pulmonary cancer each year in this country, anc it sesas very unlikely that the3a twenty-five thousand poople do not have a good many characteristics which distinguish them from their fellows, which may have some relationship to the fact it involves lung cancer, about which wo etill&o not know vary much* Now, Doctor F.hoad3 mentionod the relationship j to smoking tobacco, especially cig^re*'-as, I know of no ob servation which 'would .Imply or implicate any other type of > smoking, and I mentioned the work of one investigator or , rather two, W*uter and Graham, Uow, I happen to know Doctor: 1 Winter and I like him very much, and I agree with Doctor Rhoads* estimate of him as a very energetic, a capable and an enthusiastic chap. The kind of observation, however, which Winter Graham made about relationship between smoking and cancer is precisely the kind of observation which you will find in the literature for the last twenty years. If you look tip one of the earliest multigraphs on lung cancer published about 1930, you will find about four pages devoted to the relationship to smoking cigarettes. Hew, those observations were never taken serious ly, and they were never picked up, precisely because of tao uncritical, over-enthusiastic type of analysis en which I they were based, in other words, they were based on inade I II quate controls, they ware studied in sxich a vay thut oaa I could so that the person making the study has a pretty I strong bias in the first place, as to what he would find. Xt*s not the sort of evidence which really tells you auytaxh that you can rely on. How, it happens that the observation of Winter and Graham have been confirmed as to their general tenor. * I 2here is a relationship between smoking In tobacco, '"here ! is a great difference in the quantitative observations. If | we are to believe the observations of winter and Graham, ; pulmonary cancer is extremely rare in poople who do not smoke cigarettes, about throe p-.-rcent of lung cancers, A 8. according to their data, three to five percent, occur in people who do not smoke cigarettes* That is not in accordance with observations whioh we have made, and I, frankly* doubt whether it is in accord with what the facts are. I think that is one of the cru cial questions which remains to be proven or to be further tested* However, I did not intend to ro~lly go into the question of that particular observation, because that is not | our topic* What I thought I would do is to give you soese genorsl quantitative background for this whole question of lung cancer which would serve as sort of a backdrop for the papers which are to come, first of all, the importance of this topic, this fora of cancer as a human tumor# May I have the first slide? Thi3 first slide is simply the characteristic age curve of cancer, that is the incidence of cancer by age among uisles and among females aa usually occurs in Hew York State ahere, as you know, can cer is a reportable disease* Thsae are all cases rsportsd I ! from all sources for varlcnis periods of time, put on aa j annual basis* That is a very interesting curve, for one ! thing, because it does not follow tho type of curve that j you would expect in most forms of cane or which keeps going up with age* Most forms of cancer start in lower, young:-r /a groups and keep going straight up* In lung cancer* after about fifty-five, the rate drops# j Thera are a number of things which can be said about that observation, and I don't propose to go into all of the possible explanations, but it does mean, one thing, it does suggest one thing, that for some reason, people who were born, this is 19^4-5 roughly, and people aged sixty in j i 1914.5 , were born in 1885* as my arithmetic informs me. People born before 1885, for some reason, had less, have less j lung cancer than people born at an earlier age, en at ear lier period* That may not be self-evident to you, but it fol lows from a further analysis of these curves which show j that when you analyze these curve a by the year of birth, tney i keep on going straight up, and they keep on going up in a much higher fashion depending on when the poraen was b o m # fhat is not the sort of thing that you tjould expect in a carcinogen, if it is a carcinogen, which occurs generally, such as a carcinogen ic e air# It*3 apparently something,; j which pious out certain ages and not others# j Kow, in most - the important point that I wish to! 1 bring out in this slide is that you couldn't analyze any ! two groups of people, s&y the number of lung oancors in two, factories, or the number of lung cancel's in one factory as compered with -he rrxmber of lung cancers in the eurrounaing_ countryside, unless you took into account the age factor. 350. because obviously if you did hare a lot of people in this age group, you would have more or less cancer than if you had a lot of people in this age group or this age group. You would certainly have to adjust for age, and it's sur prising to find that some studies do not even take into cutcount such a syzapla and obvious v. table which has got to bo adjusted. Xiow, when you take this type of curve, that is the number of cases which occur by age in one year and ap ply to an ordinary life table, which ia the expectancy, the expectancy of living, at various ages in the deration of life, you will gat a tbblc such as the next slide, which will tell you about how many cases of cancer of various typsa one might expect to occur in the population today throughout the rest of their lives. These figures are all b-asii.i on an a base of one, that is this top figure gives you the probability of skin c-ncer occurring in males throughout life, end the fig w e .02 naans that two out of a hundred, two percent of all walos^alxscst two and a half perdent of all males, using the type of data which I have just described, th s is t-he actual incidence of ciacer as determined by reporting applied to the life table, two out of a hundred rales throughout the rest of th*lr lives will c /lop some form of skin cancer. Now, for lung, the figure is ju3t over one percen __ ________________________________ _________________________________ about one percent of all males according to Hew York State data, may be expected to develop lung cancer* That may be considered a sort of backdrop of this whole problem* It is a problem w hich, at a minimum, will effect one percent of the male population* I*d like to stress the point that that Is a minimum* First of all, these fig ures are baaed on tha data for '14-2 , `Ills-, since then, the rates for lung cancer have gone up and the death rates from all causes gone down, which means that the opportunity for developing lung csneer has increased an. i the rate at which it develops has also increased, ao that figure one percent is definitely minimum, but it is at least one pere&nt of the general population, Kay I have the next slide now? This -- oh, there i3 a p.'pi*r slide in between there -- may I have the first of the paper slidoa then? I don't knew whether you can s?e this* I think tills is another of the Important e n e r v a tions about lung cancer which we have to remember concern ing this whole problem, and that is the increase which haa occurred over the past two or three decades* These are the actual mortality rates in !Tw York State batv.-een 1931 arid lc,t8 in fatales and in mains, stan dardised for age, that Is tho a-: factor has been kept e&n- iuit in tho two periods, The si king thing here is the difference In the increase in the saxes. In females, the 352. increase has boon about from 2.5 to !(.*3 deaths pat* hundred 1 thousand per year, an increase of about seventy percent. In males, the increase has been from k7 to 19 per hundred thousand, an increase of over three hundred percent. Now, an increase In lung - in any fora of canoer, as you know, may be due and often is attributed to the faot that our diagnostic procedures are more effective. If that Is the explanation for iiiifl increase, then we would have to ask ourselves why, diagnostic procedures were throe hun- drad percent more effective now as compared with 1931 in salsa, bus only seventy percent more effective than in fe males. I would say that oertainly that causes or casts considerable doubt on the explanation that such an increase can be due to di;-nostic procedures alone ond jaakas It x- tremely probable that there has been a real increase in lung esnesr over Inis period, pi^ooably In both saxes. slide -- Another bit of evidence -- may I have the noxt j j which leads in the s.ue direction, is the differ- | ont way in which the ages have boon effected In 4ais in- j eroaso. New, thats the slide that you showed, fchs last | slide that you showed before this cna. Taat's It, I -- ; this la the mortality rate in the United States. I cell , your attention to tha upper left, your upper left-hand cor-. H8r here which gives you tha age specific mortality ratos. ; This is age down hera in *30, *39 and fc9. and j you will notice that the increaso has boon more marked fcdro. in the middle period than here in the older periods, which ! again conicides with the observation that there is something happening in these people around sixty, which apparently j is not happening nearly so markedly in people who are older | than that. That again is the sort of thing that one would j not expect tc see if this were purely a diagnostic phenoma-j non. There is no reason to suppose that diagnosis may - j in a disease with high case fatality suoh as censor ox the lungs, should make such a marked difference according to the age, sho'ld show Buch a marked different according to age alone, so that putting these data together, one would | lead to the conclusion that one could not safely assure th& this increase was purely artificial and due to the fact that we found more lung cancer, one would certainly huve assume that there was a very strong probability that this was something real, How, one of the points that one can make about roal changes in cancer incidence over time, is, as a natter of fact, in incidence in any disease over time, that if they are real, they are probably due to some environs.** factor. The reason for that goes bac.. one of the 1 of genetics, If you will remember your genet lea. You will recall that, according to the way the genes opar e, their proportion from ona generation to another remains constant under most circumstances, so that if genetic factors, if constitutional factors were the ones which obtained, we would not expect the incidence of the disease to change from year to year* If we find a real change, we should suspect soEt environmental cause rather than some gone tic oau3e as the reason for that change*, So I would feel safe, from the philosophic view point here, speaking as an epidemiologist that we have good reason to suppose, first, that there has been a real change in lung cancer and, second, that that change is probably due to some thing in the environment* Whatever that nay be, of course, remains to be seen* Hew, how youId we go about studying the possible relationship of 1; g cancer to various environmental fac tors? How, way I have the next slide* I have placed on this next slide, a hypothetical situation In a - I don't know liOw'GJ-l'w12T* you can see this or not* I have hare a state j i or a poptilation group in which there aro five million males! in whom nine hundred cases of lung canoor occur In a year, j j ftnd we have in that same population group an industry, we ; call Industry A, which her ten thousand males among vhea i nine c&aea of lung can -r occur in a year, so that in the = eatii-e grot , that la In the entire state, if this Is a j state K 3 *re talking about, there are nine hundred and nine cases of lung cancer In one year* How, suppose one wanted to study lung cancer as to whether It was related to, say, same particular indus try, According to -- we could tbso at least two methods which night be called the case nsethod or the retrospective incidsnoe method, H q could take all nine hundred and nine cases of lung cancer, as we do here, and we could take com parable number of controls, we take a thousand controls, and if we were Interested in Industry A, if we had reason to suspect Industry A, wed find out how many of these nine hundred and nine lung cancer cases were people who worked j in Industry A, If we got all our cases, we'd find that there u -re nine of the and in our control group, which should be representative >f the total rsale population, we should find the same proportion of industry A workers as ton thousand is to five Billion, that ia sale population to m l aa in industry A, the proportion of J( How, ten thousand to rive million ia two par j thousand, so w'd find two casco, two people who worked in j Indus try A in our controls. How, ssind you, I as muse that wo don't knew those facts, lie simply have these nine hun dred and nine cases and we're interested in whether Indui- ; try A has snything to do with lung cancer. If that is vh*t; we did, and we wore successful in getting all nine hundred ! and nJ.r.3 lung cancer cases and if your control group ia ; j__________________________________________________________ 356.___ i 1 I ! I properly selected, in other words, It wasn't biased In such! j; a way as to accidentally drop out these two Industry A j cases, or If our lung cancer cases were poorly selected so j i we might not - we may have missed these nine cases, in I other words, if our experiment were perfect, we'd find that ! there was one percent Industry A people In the lung cancer | cases and two per thousand Industry A people in the con ! trols, which amounts to a ratio of five to one, that is one ii i !l percent la five times as much as two tenthsof one percent, j | I | This gives you five to one. j ! How, if you go back to our p&stulatod population, j ' \ i this is a method, nine hundred and five million la a rate | II ! Ij of ninety par hundred thousand and nine par ten thousand j I .. ! ij ia a rate of eighteen, it's jtust - it's just tne otcar way [ around. This Is eighteen per hundred thousand and this ! nira should be here and the relation between eighteen and j | ninety is five to one, which is exactly what you would j | find in your ease method. j I. I " in other words, your case method, if ifc ware j r . oroperly performed would give you ao true situation with | ; r-ssc.Qct to the incidence of lung cancer in industry A as j co. reared to the population. If there was five tines as j : i'nuch lung cancer in Indus try A as in the population, us- cumin-, of course, that the age factor and other possible S variables which might effect lung cancer had been given consideration, I am assuming that that has been done. The other way in which you might do that would be the retro spective incidence method. You take ten years of Industry A and how many cases of lung cancer occurred in ten years. Now, if nine cases occurred in one year; in ten years, ninety cases could be expected to occur in the ton years, and they'd to - yo^'d have to ask yourself, how many cases do occur in that period, and one of the ways in i which you might answer that question la, why, you would say, I expect or would expect the sane number of cases that would occur if industry A had the same incidence as the general male population. That would give you ton years times ten thousand males, which is the same thing us a hun dred thousand and one, and an incidence of eighteen p-&i' hundred thousand would give you an expected case incidence of e :ghteoa oases, in this period, which again would be a j " i ratio of five to onsa. | Now, I*d like to call your attention to the fast that if thi .t particular case method wars dene on half tha number of cases, if Instead of getting all nine hundred and nine cases, you had gotten half of those, then those two ; -bservations would not be significantly Oifik>rnt from each I other from the standpoint of chance v - t * ons . That is ; Husy we- id be the kind of difforenco that would occur often j bv chance alone, so that one coul&a't be sure t :.-.t that was o -TV /"V i _ _______________________ ___ ____________________ ____ ..._____35.8_ j* a real difference, so that the actual ma. nituda, tha actualj 1 t number of cases that you'd study In an Industry or that you'd study of this kind* is as Important as tha selection ) of the cases themselves. Now, let us assume, however, that we have dona | this study in this way and wa do find a sigiificant dil'for- | one! in other words, we find that there are five ti:s9s as ! many cases in Industry A as in the general population, all factors being held constant. Does that mean necessarily that Industry A, thare Is some thing In Industry A which causes lung caneor? Now, in ;ay view that duos not necessarily moan that. I think that one of the important principals, phil osophical principals, to retnmber In s "'.idles of this kisd is that data do net tall you any more .nan what they actual* ly signify. They toll you that, for aesne reaeoa, peopla |; in Industry A have davaloped fiva times as much lung canceri j I it as people in the general population. That may he due to i | ! j, tha fact that there is something In Industry A which causes| !: lung c&ncor. It may ba duo to sou ithing else, for instance* | If we wars ^tudying a factor such as ths presence of aJx | 1 flngsrs on one hand rather than five, and found that one ' : Industry had fiva times as many people like that he expected In the general population, I think one would think twi : before one attributed that to comothing in bb? 359. industry. One right, normally, if that is a genetic trait, one might wonder whether tha owner of the industry had six fingers and whatnar he had many memb ors of M s family working for him. The point is that one can not automatically jump from association of t M s kind to causation. Causation may be the factor or it may not. Whether or not It's causa tion depends on other evidence, depends on what else we know about the disease. If we did not know that various dusts csn produce lung patholc y, that it is possible to produce lung cancer in animals by the application of cer tain chemical a, I think we would be lass apt to suspect that an observation like this actually indicated a cause of the effect. How, actually, there are only a fow ways In man that I know of, whether - or that I can think of, in one connacfcion, that can actually prove causation to one's satisfaction. Cne is by this second method lisre which I have celled experimental induction. I mean if one suspected j a certain substance in this industry and one applied it j t either to man, which, of course, is oxtr&maly improbably i although it has been suggested, and as a natter of fact, ; T,ja are making tentative plans along that direction, if one; fijonlied to siRn orsnin.^ls and actually got tne uioe&ss in . ouastion, I think one would have soma additional evidence 360* that that was an etiological agent* However, I 'd like to call your attention to the fact that if one didn't do that, one wouldn't have evidence that it wasn't an eti logical agent, because unfortunately, there ar many things or at least there are some tilings which will cause ex xar : man that will not cause it in animals, ao the negative evidence is not conslusive nega tively* Tho other way you can test it is by the protective ! test, by actual prevention, I naan if we actually mad ease element In Industry A which was a factor and then studied the incidence later on, the protective incidence and found out it had dropped, I think wa would have sesae additional evidence to the etiological relationship, but it ssesa to rsc that one or the other of those tests has got to be ap plied in 11 of our presumed data regarding otiologic - et iology in pulmonary e&neer, nvironsontal biology, boforo wo are absolutely certain that what wo think is etiology is or actually is* Kow, I don't kr _w whother I 'm running too much uvyi' my time, X hod come dais. which, from the philosophical point of view seemed to me to b interesting #-s to how la- i portaec oaviro ..mental, specially 'rupv.tlonal cancer migntj be la America, but might keep ''at, perhaps, for the ' discussion, Kr, Chairman* o 361. BY DOCTOR RHOADS: You have throe minutes* BY DOCTOR LEVINs I do have three minutes? Well, may I have the next slide? These are some - some of our data regarding the relationship between cigarette smoking and cancer, two ! forms of cancer, cancer of the lung and cancer of the larynx This is simply the proportion of cases of cancer of the lung i and cancer of the larynx who wore cigarette smokers, by a6<M compared with two types of controls, in the aarae hospital j l popiiln^iO' ^ j This Is our _ the Buffalo Hospital, the two oca* j j trols were patients vino had no malignant tusor and patients j who have other types of cancer than lung or larynx. You will no to one thing which apparently has not boon empha sized very mich, but which wo have c o n f i m s d by furtner j study of fci' ** ph*>noraen^r. That is apparently there is i r.'.'eh rslat.:. '.hip betwoe-n cigarette smoking p "i Iroaynx c*n-' i cer as bet*un cigarette s-aoklng and xurxg ct-iu^r* | You w* *1 notice, also, the Inter os ting thing ia , that the cigpreito smoker is ona of the ! ..i.ig3 w-Oj.ch !-iryij3 : 'off withincrasaing ng* You ware caused - one of tne . things we would probably have to find to explain tbs "t9 ! curve in near is scr.9 thing wnich dropped off with j-ucrrac ing age. For eome reaoon, at luust at present, ha olau. 2h2,_. jia man is, the leas likely ha Is to be a cigars tie smoker* I! i How, that*s a very interesting observation* That ! ig, again, as Doctor Rhoads indicated, and I think wo will | all agree, this does not necessarily mean that cigarette ! smoking causes lung cancer, though that may be simply an I added effect or factor* It simply means there is more lung j! cancer among cigarette 6mokors than there is among non- | | ! !j cigarette smokers* ! The interesting question is what does that actual" | ly naan from tho standpoint of numbers, quantitatively? vs I And there are various ways of doing that, end I won't try i ! to go into the arithmetic of it, but on the next slide, I HI have *jotted down seme conelusions which arise from our 1; Ij data* They do not arise, inci&ontally, from the data such ij as has been induced by Winter and Graham, because if their 1 data is right, practically all - practically only cigarette j ij smokers get lung cancer and that isn't truo according to , our lata at all* Thirty-five percent of lung cancer cases are not cigarette smokers according to our figures* If our data ! ara correct, throughou life, this is the Incidence of lung cancer* In non-smokers, five per thousand, this is through*? out life, nind you. In heavy srsc^rs. that Is not just : cigar-. le smokers, that's people wr. smoke oca to two packs . a day, it's six times as much or thirty par thousand, that'3 3634 three percent. You will recall that th lieftime expectancy j i,; tha general male population ia on percent and this ia J t,',,_ree times as much* j Now, according to a recent paper in Public Health Reports by Britain, Frasier and Coven, it's ia ay opinion that that agress at least in general with tha data of Hachla and Gregorius on chromate worker cancer) thaas are j not the figures given by Britain, Frasier and Coven, they are simply ssy interpretation of them, as to how much 1 g ooncer thoao worker# may be expected to develop in twonty-- five years if their data are correct, per thousand* j Among white males, there should be sizty-five | cases in twenty-five years In chrcaate workers and asong non-wM*.:e, two 1 'ndrod end t irty-fiv. How, unless my arithmetic is vo y, very far off, this mo i* one thing, possibly two, certainly that cigarette smoking, if it is carcinogen is a pretty mild ono, since it increases the general incider, ue throughout life only by a factor of threat &od if such an occupational factor as chromate dust is j carcinogen, it apparently, w ho-- It coos week, is a p -x evty j strong one, since it may increase it by a very m e t M g h e r , factor. That la, twenty-throe percent of all the wale ; workers, in twenty-five years, ray toe expected to develop lung cancer if the data published by these workers stand up and are representative of tha true picture. ___ I | So that wo have apparently, assuming that toe facts wo have; at our disposal, wo can roly on, two chief types of carcin ogen, which, are environmental, among others, because I hadn't considered general environmental dust, one which is hoe / smoking, which Is apparently very weak, and one of this type whose incidence we don't know since wo don't know j all the occupational carcinogens, but which when it does j i occur, is perhaps very strong* j I think I have more - one more elide, which por- j i haps puts this in numbers# Applying the one percon t torcusd- cnrt . fs figure to the seventy-five, approximately seventy-i ft.? illion male* in the XJn* ted states, throughout their j i lifetime, seven hundred and fifty thousand of thorn will develop lung cancer# Of those, a little less tnan half or j rather something like a third, two hundred twenty-five j thousand, will be non-smokers* She rest will be isacscorsj j but ' f.se there is no reason to suppose that smojisrs uro , ismuta to the same influences as the non-smokers, we c m i only attribute t 39 hundred and thirty-seven thossad of . ^"hose r :-3es to t>er se, so that what hni3 loads us ; in substance, to is l~'ls, that even if smoking gota ivu.1 credit, we have got to account for about four hundred thousand cases of lung cancer to otnar causes which, if theyre environmental ar.i if they occupational, means tnat we've got to find a lot mor9 occupational causes of lung 365, cancer than w have up to the present# The on thing which makes that likely is the fast that one, whan occupational causes do occur, they apparent ly are quit# strong as compared to such a mild agent as tobacco seams to be# (Applause)# BY DOCTOR RHOADS: I know Doctor Lavin'3 presentation will arouse & great deal of discussion# 1 would like to open th discus** sion by Just one question, liort# I assume that th drop off with age is Just for the number of individuals in that aga group population? . i BY DOCTOR tKYISf j Yes, that's thacase, J of 0 0 *0 -3 RHOADS I j Azid I also assume thatwhen you refer to the -igh j inci&oac in individuals brn in XC-35, that ia continuing j high as th years go a and similar studies ara being mad; it isn't that year, but that period aft it*? b y 3DOCTOK ?>!/1Y t Ye 3 pv ;>P F 'C CZ I 'm very nr oh intaros ted to c m r Doctor Levin in dicate his acceptance of smoking as a possible etiological or participating agent, because this, in view of the Impor tance of the problem, 1 '-a extent would Justify one perhaps H --- -------------------------- _____________ ____________ 266*.Lt i t i; i ii 1 In considering wuether that subject should lend itself to experimental work designed to prove or disprove the case more adequately, provide more adequate statistics on the subject* How, may we have- discussiout Doctor Larisa, would you care to make any coziunant at all? BY DOCTOR LANZAJ No, I have no cosaaent* E J J>C1*0R RHOADS S Is there no onr> who inclined to src-nka eigarofctas, who w&nta to ask quaation3? m DOCTOR FLDTSEaH: According to Doctor Levin, ho supposes that chro mate is a such more powerful - 'rclrogan than tobacco, but j how did he sseaaure the intake of chromate as coraparsd to j I tobacco in theao two groups! | BY BO gABAt ! Doctor Levjja, would you <: .*8 to answer that quoa-i tloa of Doctor Fletoh-rs? j b y d o c t q s j ;.5Yik s "ell, I think i ` a very good question. I In other words, it refers to the relative dosage, I wasn't trying j to ho ,is fine as that* I wee eimply reforriag to chromate : as the net axperienea of a chronate worker, wherever they're involved. The net experience of a chromate v- rker, if the r* /"V i________________________________________________________ 367i- i; data that we have available act or are correct, I have no j reason to suppose they're not, because thay scam to have been very carefully dons* The net experience, vtfiataver that is, is apparently much more powerful than the net exi perionc of haavy smoking* How, your question is what woulc. happen if wo took those chromate workers r d gave them one tenth as rruch dust as thay did got? I don't know* I'm v- simply referring to the actual facto as they were observed II I in actual lifs* j BY DOCTOR RK0A-D31 Toss hi. t answer the question? I'm surprised w i no cue has coms-aated on the possibility of air pollution* >; ii Doctor Smith, you must have come foelings on this point* I f.i saw Doctor Joe Smith in back here a few rT atea ago* ifould f; sj you care to co&ment on air pollution in view of Stott's iL, report, s a complimentary factor of environmental nature? ii B Y D O C T O R S H I T E : I have job to so r.'-otor Stott's report* It cer tainly opens up a very different problem* I think I'm in a fortunato cos- Ion hero in that most people who are really familiar with this subject and vho told me what little I'vei teen able to gather about it, are in this room fold I esn j r.as the problem on to then* > BY _DOC fOB RHQiPS * Well, there are o-'ar facie be ex*' .',3 on this question In one report, the incidence of lung cancer in -.dividuale in cities of over one hundred thousand# is muoh g,, ater than in rural coisannitlss# I don't know how valid those hints are, but they justify further ^Lscussion. Are there further coEtarents on Doctor Levin's paper? (No response) If not, we'll proceed to the next paper# Doctor Euaper needs no introduction# You know his long continued j contributions to tha ff. ?ld of environmental cancer# Ky j acquaintance with him dates ever many years, Xt has Alwaysj been most stimulating# Ke is one who has stuck to M s guns! in this Important problem, and I happen to anew, in some instances, under difficult circumstances, and .has givsn B Euch knowledge of this laport&at field# Doctor Euspar BY gQiJgCR KSSPBRt (Doctor Kuapar read a prepared paper, which Is on file at tha Csr-vnac Laboratory)# BY DGOTOB RH0&D3i lid like to congratulate both of tha npoahers on completing their prosont.&tions on time. I hops their iuecedents can be observed for tho .-st of tha day* I would like to raise one or two q w - tiens to initiate the further discussions# Doctor Kueyer, you raferred several times to the presence of benspyrar in sus pected material as supporting the view that these materials -^ r 369. were carcinogenic in the lung of man. I do $ot know of the evidence that pure benzpyrene, par se, will produce cancer in the private. Does aueh evidence exist? BY DOCTOR HtT15?SR No, that evidence does not exist. There is only some circumstantial evidence aval..-ble. There are two cr?e| on record among laboratory wori ors which, by accident, cane in contact with benzpyrene. One was the screen of the skin while later on a -sneer developed, but vinether that is con- | elusive evidence, I woxild - I am not quite certain, but wo j have to consider one fact, that benzpyrene is only one of j . j the carcinogenic chemicals .-.a these materials. For Instaacp9 . , __ _ _ j Falk and Steiner isolated three or four, and t-j^Sy are ail | shale oils available which do not contain the pyrene, but J i which are carcinogenic, so we chu'11.d >*ot - it n't ;-he j material as such, the crude paiterial as sucii^ wj.ich wo 3 loculi consider as earelnor aie, arid we should not twite the pros- ease of one car c i n e m a as authentic, but should take into account that ujh vat.-;rial3 are carcinc^-snico BY IhXMOP. ?.ZO'T-~: Title uAa o * of the things I hoped the doctor would bring cmt. If berspyrens is p. ooeot, it suggests the presence of possibly stronger carcinogeni.c agones, or a whole series. wy rnn -'OH H V C r t ___________________ _ __________________________ 170._ A whole spectre of carcinogens# BY DOCTOR RH0AD3: Doctor Huaper made also a very interesting point ir. Jo? that was the disposition of the point that the sun night be prophyllactlc# I think one can not wholly ignore the endocrine system in cancer, in view of 2IcCsnn, T&nnsnbaum's and other worker's surveys# There are surveys as yet unpublished, bu they cuggest very strongly indeed that hypcsectoaiy (?) will render inactive, inert compounds which, otherwise, are active, exceedingly carcinogenic, and it's not entirely inconceivable that the sun rai.r-t play o,- ..} role# b y sootor Various surveys have boon made by pallet, who studied the incidence of cancer of internal organs arsons Canadian - among the natives, and he suggests that persons j who develop cancer of the c` -n would bo come partially icyriuiie to cancer of the internal organs# I don't think the occti- I i patiora.l evidence shows that# If you t 'a exposed to an j occupational agent tba. * causes eaucer of the akin, you have, in fact, a higher l i d .iity to Co,-o', op caneor of the Inter-: nal organs# That *a the general ev:'.-.e.-.uo offered from a i general epidemiologic ?1 viewpoint# ' The pathologists of the University of Richmond - his nerne escapes as right now, ho rer>es very similar claims. 371. I don't think they are just"fled on the basis of the factual evidence wj have. Ono can speculate a lot out I think it's our feeling we have to stick quite strictly to what ve actual ly observe, the factual evidence, and not speculate too much. BY DOCTOR RHOADSt How, I'd like to ask Doctor Levin to cosiesat on the statistical validity of the evidence regarding occupa tion, where the incidence is comparatively low. Kany of those figures are frosu Eenilway series, of course; do you want to connent on the validity of those observations of j i the engineers, the tcbaccicnists, the barroom workers, other happy occupations? BY DOCTOR LSVIH: I don't think I have recently examined that data j i to - uuffieisntly to be at all critical* Gse of the things which isiprsssod r*_o ''.bout the type of information that Dostor Easier presented and which is, I think, good 3itorators j is that, in general, even at the with a fsw exceptions, the i lung cancers of Schnaborg, ueeais to be one of thos-o, c na j with the high dosage of exposure and +b.e majority of people ! exposed, apparently do not develop a disease,which strongly suggests that we are over-stressing the s ituaticn if wo cenv fine attention entirely to the ventilation and ignore the intrinsic constitutional agent* T think tr. pei-nps - I 21 don't moan to say that wo should do the other thing and throw overboard ths evidence regarding these things, but I think we should study just as carefully wh; it is that some people just as exposed to the concentration of the agjnt, do not develop the disease, I think we may get clues to the development of the disease in othora which may be more useful or certainly useful as the preventive aspects of tho occupational data* Somo of the published material that Doctor Rhoads reforrsd to is simply of a kind that's very difficult to evaluate because you don't have enough facts* You don't know how many people word in thf factory at a certain age for so many years* You doch i know whother all the cases ware studied. You don't know at vhat age they wore r-otlrod, You don't know the difference between the people that you did have information about, end the people that you did not hAva "nfomation about, but from the standpoint of a state official, which I am, r.nd from the standpoint of what all taxis sort of thing leads to, from ths standpoint or qus&tto. of actual action, official action, I would say - md since this is an in das trial group, lid like vary much your reaction to this propts&l - I would any tlat the evidence is sufficiently strong to at least justify one proposal and that is that industry be required to keep much 'cottar records than it dees regard`ng both its workers a> .what iI!.1 j! - j: t 1 ________________ ____________ ______________ -__________ 373*; | happens to them without any bias as to what conclusions j: ; will be reached and perhaps with ovary preservation of the confidentiality of that information, 1 I I think the information is there that will answer many of these questions beyond a shadow ox' a doubt; we sisa- : ply are not patting the information in such a way so that !! we e&n analyze it. and really be sure of some of the aa things. l i. I iiI think, however, that could very readily be done* I 1 i : by.docT m j m q A m t : !Thank you, Doctor Levin, I hope that pc-:nt will tj come out and I think I would Ilka to emphasise another |i i! point Doctor levin made previously, and that is it is per i haps the obligation. of industry to consider the setting up j j j i of the s o-odlod preventive experiments withcontrol groups* ! : ii j which, after all, will provide perhaps the moat final data, j I j Ij I take it, from your feelings about It, j f Like to have your further discussion, Yos, sir* j h ; I BY DOCTOR FRAYg: j j; in r^f^renco to carinoasa in asbeatosls cases, w*. j !; a s the by-pernlasi* of the bronohiul^epithelia and its ex- j ' tentlon into the alveolar &paces in areas ox' thelung that . . ,N*r' \ show fibrosis, but if we look at the epidemiology of bron- ; : chi, it aeobis to me it looks quite normal;yet in the cases that do have carcinoma together V t h thei.m aabestosis, the cases that live ssen, in the main give ovary evidence of being primary la the major bronchil It seems to raa that if there la a connection between asbestosis and cencor, it must be considerably more devious than simply hyperplasia, metaplasia and nsoplasiat# I'd like very much to hoar j Doctor Eusper's comments on that point# BY DOGTOD ?i03FBHs I'm glad you brought up that question, bac&us j It's boon a long time in nsy heart, but I'm glad you oreugf up that question because it's been a long bias on say heart,! cad I would like to shoot it off# j I Tea carcinogens are roally misnomers. If you j i'^vdjT 2?`?i of c&pcinog^nsy you ' ' " . ,, j Laparplasia, you see loirs of benign tumors and thou fIn&x&y| you ray see careiaoiaas, you may see that some of vise benign j tumora bacons malignant or you may soe. In the prepared j soil carcinomas, originating primarily# I personally place a groat deal of significance upon seeing the development or his tor- of carcinonas i n thoj slides I Ir:-.re studied in. the crass I lavs studied# 1 thank' they g i r o u s the l i f e 5, ?fcory o f tha carcinogen and tn a ! i < ^3 h i s t o r y o f th e c a n c e r * 7:o r e , t.ie efr cages j-n a s b ^ * : tosis are typical of that# I have seen them in the cancers Which we produce in dogs. I have seen thorn In cancers aicn o odace in nice v .i fch the various h y d ro ca rb o n s; it *3 always the aam atony; to me those progressive changes are ___________________________ _______________________ i significant* BY DOCTOR BH0AP3 : I think this is a very interesting point that comas up frequently In discus." ns of the hiato-pathologio development of naorlastic change, and always pusalad by aa | group of colls presumably equally In contact with others | bocc:.:a exposed to a rapid shift in mortality, nd I prosunoj tfaera are unknown factors regarding coneentratioa* 3#xt quo sfcionf v^r rp^.-Ty - f - v t j j I was the instigator of two of the stars**y<s .Scoterj 1 Huaper asniioaed, tea one of Lombardo's on arsenic and the ! on of Xngall*a on carbon black, so I have aa intarost, t ruora than a casual one, in Coeter Hnoper's remarks# 2c-l*y !l the big arsenic r-roduasr is S:radon, with Xtaxlco so end end j i ilia Uruitsd States third* *Jo will continue to produce t*r~ | conic for the indefinite fixture, b ocatxca of the fact that j the copper cones contaminated by r-'srni The Industrial experience of no smelters dates UaCv new in the United coat as i" r shout forty to fifty jour's, 30 that as experience goos in health matters. Obably havo as good as there ` i'x. H ***nan axper:3 a little briefer and so is th -.nilnli experience# VA3 *U. vey in the arsonlc works was instigated bjeau.e we wdi1* ft. little appareho valve after what was found in the chromate industry to which we may have some slight analogy* i That is, the chromate workers get fc.n9 perforated j scspta, and then they were shown to get lung cancers, ao we wondered if perhaps the same kind of thing did not occur in the arsenic bus in-, ;s* Doctor Hueper remarked that Lombard,and Lombard and Sehhager, from an interpretation of their statistics, the Interpretation was curious* I can't say I agree. I will admit that h*lf of something is - startles you, and - but half of four cr i If of ten Is not very startling, &nd I rather think that the explanation in *he smelter in the vest in which Ilusper takes tl m to taak fox* not, making more of the aroonical car :;er 13 due to the faot that the figures are too low, the amount of ru bers of workers are too low and they concluded that it rrd t just as well have been the other way, on a pure chance basis# They took for a control a smelter in the arsenic business that had no arsenical exposure and, in that case# the lung cancer, I regret to state, was a little more alarm ing than tv b of the community at 1urge, and there was no arsenic exposure whf ^er* Tho ^nelter handled no arsenic* ! i I think that the doe iota's plea for be titer statis- f! a in 5ndui> try is an OAtvanely good one end I 'mow that I ! the SE'Jlfein. business, fill of the big smeltsi-o now, I ha : f.miliar with what they're doing about thin, are no a thaTM anxious to gt their figures straight P'^d will keep recordsi I Thsir records, in the past, are very fi'&gmenfcary. Hus per did not mention the fact that there was a good arsenical survey in Utah by Heal and ofears, in which they found no j evidence of lung cancer* BY POC-fOa RHOADS % yell, here information comes from another angle, perhaps alluding to a point brought up before, and feat is fee tendency is to incriminate pure compounds of known activity when one is considering the possible carcinogenic action of a most confusing and complex me~s of m&torials j cosing out of a smelter or coming out of a complex organic j mixture, and perhaps the evidence, I take it the ovldtiise ia good for the problem involved, in contamination w d j pollution, but not so good for a singl pure conat.ity.fent j of the polluted material. Yes, sir? j . j BY BQCfOH -.AHUELfON * j On that point, in reference to carbon black, I 5d j like to correct the record* The ffcoirm* work showed that j 3sS|. benapyrene and 1*2 benzpyrene, a very raila car' fenogen I I was present in the extract of carbon black, and no evidence j of carcinogen to follow* j iav ^OC'^OH RHO ADS J 'Thank you. Will you giveyour name ploueo? ! 1'octor Samuelson i BY DOCTOR HHOADS: Very intoreating point* yes* B Q Q C TOa HEMPS S I would follow up .a8 question of Doctor Pratt* Did I hear it right that you were commenting on the f *t that in asbestos carcinoma or carcinoma with asbastosis, you did not find the so-called hyperpl -sia, metaplasia and so on* e* ha carcinoaat0 Was that what you said? BY DOCTOR PRA1T: That*s what I said* ' May I show ono slid? I come from Holland and there war only two cases up to now* of arbentos Holl and, ono with carcinoma* on without* It was by chance I got them} it was a chance I got the case with hyp&rpiasia anu metaplasia# I don't know how you us words, hyperplasia, metaplasia* I can only say that I tried to differentiate between the id of hyperplasia, as you seo it in tuberculosis* I tried to differentiate between the {..rnes a -ou sae it in til. -srculoaia, so-c?-lled istaplasia j and the type you aoe ordinarily In cu'clnaa, broncnisl j carcinoma of the liuv , as you can study it in apscions, and then I roust say that in the seme type of bronchi, the j middle lot, not only the small on, a, but the greatest ones,- ----------------------------------- ^79* bc0aflj3i* ytt w1 11 find a obtain type of metaplasia, tba* j certainly is relatad to carcinoma, ! ^ i I ao not *5ay that carcinoma does come from a typcj of metaplasia, but you do find thorn together, as you do j--,t! find the metaplasia in bronchiectasis or tubsr-culosis, and i this is the slide I got here. Our slides are not as boon- j ciful and our micrographs are not as beautiful as you nor- I Bially s9 ti: ,m hare, but I hope you will coa the point in ! `^hat I am toying t-o show, | H-arc you see the bronchi epithelia,' and you see j - the comment will cc>' t-v,,* ... .. i UOi> ontaro cro ss-sa e tio n j but yoxi will oae the tYTibal pa tini , .i . . , ' 3- of it, and you will sea entirely hew this bronchial api3.3 CUfc o f OS'50,^31*^ This is one of the greater bronchi; it is not a -- ..all o^a, and the next please, Kara you sea the contents j of -ho middle lobe bronchus, mucous colls, mueocytos. Harej you must imagine the basal wrfbwm. and here you see the j S Q --CO.j.ltjd tv t}iPa1 vr.:^ ** ` I don 11 know, n c.?, ._, ..u . . ^ .. I .uiaaiug to h y p e rp la sia , j Ai'K"'d ar-' ~ tiia greater oror,chi too, ana the next ples3S? Kdre yu ces " s that thick nucous that is -ghed up so diffi tly in asbestosis, the thickened basal membrane, and hare the very typical kind of epithelium