Tobacco Documents Online

""~ ---t ~ 70 00 .e 1958 REPORT . of the ~ SCIENTIFIC DIRECTOR CLAREiYCE COOK LITit.E, Se.D. I TOBACCO INDU3TRY ItESEARCR DOMIHITT

Organization and Policy The Tobocco bdut" Re,anm+cfi Cormnlnet is the sponsoring agency of a rawrb program into qnestions of tobacco use and health. It was org.niaed in early 1954 by repreaentati.es of tobacco manufacturers, growets and ..ardtousenKe. The Committee itself neither operates any rt.drth facility nor has aeir determination in the scientific program other than making available nooaaarp funds. ~ The ScFenN/ic AarJ"y Board to the Tobacco Industry Research Com- mittee has full respomiblity for reaearch policy and programming. The Board comists of independent acientiats, doctor's, and educators who main- tain their respective institutional affiliations. The Scientific Advisory Board does not itself engage in raearch for the T.I.R.C. In its continuing effort to evolve a well-rounded research program, the Board this year invited Dr. Richard Bing bf Washington University. St. Lonis, Mo., to become a Board memba. Dr. Bing's acceptance brought tbe Board's membership to ten. Also, as a step toward setting regular terms of tnnne, the Board elected Dr. Lynch to serve as its Chairman for 1958- 59. He replaoes in this capacity Dr. Little who was 5nt elected chairman In June, 1954. Dr. Littte continues as a Board member and as Scientific Dkector of the T.i.R.C. n -y All Tobacco Industry Research Committee grants are made upon M recommendation of the Scientific Advisory Board. Recipients of grants are a' an~ned oompktrt scientific freedom in conducting their investigations and repoctimg the results of thdr research in the accepted scientific manner t10tough medical and seknN6c Journals and societies. The investigators re- cei.htg grants from de Committee ate alone responsible for publishing or reporti% their tesea:eb tesults. 1958 REPORT of the SCIENTIFIC DIRECTOR (7,ARENCF, C(1OK LITR,E, S,.D. TOBACCO INDUSTRY RFSEARCH COMMITTEF 150 Fsat •t2nd Street, New York 17, N. Y.

n ~ . . 0 0 I+& 0 w F-+ ..~L-•i CJT SCIENTIFIC ADVISORY BOARD to the Tobacco Industry Research Committee KENNETH MERRILL LYNCH. M.D., Sc.D.. LL.D., CMirman Preardent, Dean of Faculty and Professor of Pathology Medical College of South Carolina Charleston, South Carolina RICHARD 1. BINO. M.D. Professor of Medicine. Washington Uni.ersity Director. Washington University Medical Service, V. A. Hospital St. Louis, Missouri McKEEN CATTELL, Ptt.D., M.D. Pro%sror of Pharntoculory Cornell University Medical College New York, N. Y. JULIUS H. COMROE, 1R., M.D. Dbecto., Cardiovascular Research Institute University of California Medical Center San Francisco, California LEON O. JACOBSON, M.D. Professor of Medrdne. University of Chicago Director, Argonne Cancer Research Hospi"l Chicago, Illinois PAUL KOTIN, M.D. Associate Professor of Pathology University of Southern CalNornia, School of Medicine Los Angeles, California CLARENCE COOK LITTLE, Sc.D., LL.D., Lrrr. D. Scientific Dirrctor, Tobacco Industry Research Committee Director Emerftut, Roscoe B. Jackson Memorial Laboratory Bar Harbor, M:me STANLEY P. REIMANN, M.D., Sc.D. Scientific Director Eneerltas, The Institute for Cancer Research Director Emertnrs, The Lankenau Hospital Research Institute Philadelphia, Pennayhrania WILLIAM F. RIENHOFF, JR., M.D. Associate Professor of SaraM Johns Hopkins University School of Medicine Baltimore, Maryland EDWIN B. WILSON, Px.D., L.L..D. Professor Emeritus o/ Yital StaHsNcs Harvard University Cambridge, Maaadtttyetts ROBPRT C. HOCK&f', Pa.D. A.ssocfate Seknrt/Fc Dlrector O

Observations On Research Preface Although this report is listed as that of the Scientific Director of the Tobacco Industry Research Committee, it really represents a cooperative activity. In the rdarch program, the Scientific Advisory Board, the executive officers and staB, and the various independent research scientists who have attended the informal conferences, have shared importantly. So has the Associate Scientific Director, Dr. Robert C. Hockett, who has with great wisdom and untiring energy established and maintained an unusually close and constructive contact with the multitude of actual oper- ative phases of the T.I.R.C.'s program. 4.3 It is perhaps impossibk to re0oct ie the subject matter of a report the --jesprit de corps of the rottp which is working together throughout the years. 00 It is, howerer, the happy dutl of the Scientific Director to record the continuation and gmwth of that spirit of unity to a degree that makes the time pan rapidly and abaorbingly ana that makes our joint adventures as ~ a team exciting and spiritually rewardins. INTRODUlTION Gratifying progress in the development of the rcsearch program Into tobacco use and health can alpin be reported this year. As of September 30. 1959. the Scientific Advisory Board to the To- bacco Industry Research Committee hau since its establishment in June, 1954, recommended grants..M-aid to 75 scientists in 50 institutions. Thea grants have all been approved by the T.LR.C. As explained in detail in prerious teports, the Scientific Advisory Board does not merely review grant app/'iqtiotts. Much of its time, efforts, and thought are devoted to exploring promising avenues of research and encouraging grant applications from scientists to carry out needed investi- gations. Of the $3.200.000 in funds made available by the Tobacco Industry Research Committee, including $500,000 for the coming year, the Board has actually granted some $2,410,000 to applicants. A complete list of current or recently completed projects being suppotted is given In another section of this report. About 535,000 a year is granted in research fellow- ships to medical school students who devote their off-school terms to re- search on problems selected and approved by their own institutions. Elsewhere in this report are abstracts of 31 research papers completed for publication in scientific journals during the past year by scientists who acknowledge support of their work, in whole or in part, by the Tobacco Industry Research Committee. PRINt:IPLFS AND RBSPONSIBILJTiF.t When the Tcbacco lndnstry Research Committee was organized In 1954 and a Scientific Advisory Board of independent scientists was ap- pointed, all ooncerned, including the Scientific Diredor, accepted certain definite principles and responsibilities. It was reoognited that: (1) The origin and nature of lung attcer and, in fact all types of anoer, praent one of the most complex and difficult scientific problems that mankind hai ever faced. (2) The oompkxit~r aed subletr of cardio.ascular disease are also evident and ate admitted by the vast nu of those responsible for rcsearch, therapy and education In that Ae . (3) The twn-accepbmoe of rt:potted. tttatistical associations as proof of a cause and effect relatiotesfdp between tobacco use and these two groups of diseases is justified not only by accepted scientif4c 5

procedure but also br the complexity of the diseases and the need for more direct expetsmental research into these diseases. (4) This research task reqdves ertative skill, patience, time, hard work and money. The Tobacco Industry Research Committee's stated purpose is to advance such an effort to the best of its ability. The Tobacco Industry Research Committee has been and will con- tinue to be faithful to those principles and responsibilities. It is not considered enough to sponsor investigations solely into to- bacco use in various fotms, even though we are finding that existing knowl- edge in this area is far from accurate or sufficient. Research is needed also into the basic problems of constitutional diseases. Many factors of modern living, of which tobacco use is onlyy one, are being studied for clues to a better understanding and control of cancer and heart disease. IIVFLUE/1iCE OF TRE "TOBACCO THEORY" There are those who insist that failure to accept the "tobacco theory" of lung cancer may tesnlt in delaying progresa toward eventual conquest of this disease. It may be noted in passing that there has been no similar insistence by those who are most concerned with advancing research into the complexities of heart disease. Progress can scarcely be hampered by the recognition that eatt knowledge is essential to the solution of a probkm, particularly if it is accompanied by a determination and a program to de- velop that knowledge. On the other hand a problem may wen be obscured, and its solution delared, by the soothing acceptance of an over-simplified and immature hypotheisis. The attitude is fraught with the danger of possible disillusion- ment and forced modification or abandonment of dogmatic generalizations. The proponents of the tobacco theory have generated iecreasingly intettsive and extensive propaganda by selecting evidence for and assert- ing an acceptance of this theory. As a result, a non-scientific atmosphere, conducive to prematurity, unbalance, and inadequacy of public judge- ment has pervaded the whole field. Tbis has been abetted by the emotional anti-tobacco adherents, both lay and professional, whose un- compromising position admits of no other judgments, scientific or moral- istic, than their own. While the tesnhing situation is t! gtettable, It Is not final or conclusive. It must be faced with confidence th.t, in the field of science. the public wift demand, and A, -1 - will pt'o.ide, oontinued research; both basic and cliniial. until quesdoea are aaswered and doubts allayed beyond the dangets of selective ietetpet`tatioe. Almost everyone agtees that the practical evaluation of possible effects of tobacco use on an individual is a matter of determination by the indi vidual facing a decision. T!x same is true of decisions regarding automobile driving, sun bathing, dietary procedurc, amount of exercise, and a number of general and widespread human customs that may conceivably affect health and survival. The public can rest assured that those engaged in performance or support of research are just as impatient for the answers as anyone can be. They are also deadly in earnest and are determined to work without undue haste or avoidable waste until the goal is attained. In order to give the public a more complete picture of the scientific knowledge on which it may base its judgment of tobacco and health questions, this report presents some of the evidence that has not been widely publicized and may have been largely overlooked in the face of emotional attacks that have been made on tobacco use. LIM OF EViDENCE Three types of evidence offered by advocates of the tobacco theory in connection with cancer are statistical studies, pathological investigations, and animal experimentation. Statistical 3tudle. The statistical evidence was the first to appear. Various studies of selected populations indicated a positive mathematical association in adult white males of certain age groups between excessive cigarette smoking and death from lung cancer. The association was weak and much nare uncer- tain between pipe, cigar or light cigarette smoking. It also varied for these types of smoking in the results obtained in studies by different investigators. There appeared early in the evaluation of the statistical data certain major differences of opinion concerning several basic factors in the situa- tion. These diRerences still exist. (1) Certain statisticians believe that 1irAoforlon of smoke Is an essential or vitally important factor. Others produced evidence that inhalation was a negligible factor. (2) Certain statisticians believe that the claimed harmful effects of tobacco use are directly and proportbnately related to the amount and duration of exposure. Others believe that there is a threslrold up to which effects are negligibk. (3) Certain statisticians thearize tMt a htent period of de8nite duration is necessary before carcinogenic e~ech appear. Estimates of the duration of such a period vary contlderably as do those of the relative importance of tobacco as otte ot a tromber of possible car- cinogenic factors. , 6 7

Until differences of opinion of these typa' can be settled and their initial appearance explained by statisticiana thenaelves, quantitative inter- pretations of the effect of tobacco based upon statistical studies will remain of little value. Cow jUeths Readt, CUert A more reeent extension of the statistical association of tobacco use with deaths from a long liat of diseases complicates the matter still fur- ther at a time when the situatitxt, even as regards lung cancer alone, is not well integrated and reconciled. An example of the type of conflicting results which appears in the statistical data and which needs explanation and/or reconciliation is seen in deaths from cancer of the stomach in white males. An Amerian report shows the mortality rate from stomach cancer of smokers is 61 % greater than for non-smokers. If the association were one of cause and eftect, one would expect the death rate from cancer of the stomach to be on the rise with Increased cigarette use. Actually U. S. Gov- ernment figures show a drop from 16.7 deaths per 100,000 in 1949 to 13.2 deaths per 100,000 in 1956 from this cause. A report published in 1957 in the American lournol of P:rbGc Health reported that "this study failed to indicate any appreciable association be- tween smoking and gastric cancer." On top of this a British statistical study showed a death rate from stomach eancer of 28 per 100,000 among smokers, while that for non- smokers was 41, or nearly 50% greater-almost exactly the reverse of the American finding. The confused and contradictory nature of these findings is evident. ~~'1T+eo P.eb Stand Ort i= An examination of the reported statistical association between cig- arette smoking and lung cancer reweab two facts clearly. First: Non- smoken develop bronchogenic carcinoma. Second: The overwhelming majority of smokers do not. (Out of 1,564,476 deaths in the U. S. in 1957, U. S. vital statistics show that 13,995 deaths or leu than 1 per ant (.899b ) were from lung cancer reported as primary; that Is, believed to have orig- inatt:d In the lungs and not to have spread from elsewhere. ) '11K Mt obaervatbn shows that smoking Is not an essential condition for p;;ducing the disease. The second obaervation constitutes strong and conr,ncing evidence that the chain of causation of the disease is complex. This inference is supported by the vast body of experience In medical resarch showing that a number of conditions must occur simultaneously, t or some considerable series of factors must interact in a particular manner in order to produce any particular disease. All Indications are that the condition known as bronchogenic eardnotna or hmg cancer is no excep- tion to this genenl medical experknce. Much philosophical and semantic discussion has been devoted to statistical correlations and their relation to the concept of "causation" of a disease. The practical issues are actually rather simple. Medical experience has demonstrated that the etiology or development of most diseases, particulary constitutional diseases, is invariably complex. Many factors enter into the long chain of causation. Better undetstanding of the respective roles of the several factors involved has usually re+mlted in improved methods of prevention and oontrd by showing new ways in which the chain of causation can be broken more effectively. For instance, malaria was associated from early times with the 'bad air" of swamps and marshes. This erroneous "miasma"-hence, "bad air"-theory long delayed the evolution of efficient cormtive measures, although a degree of control was achieved by draining swamps or moving people away from stagnant waters. When the role of the female mosquito (Anopheles) was learned and its life cycle descn'bed, a more effective and less expensive method of control became available. Spraying stagnant water with oil and later with insecticides killed the mosquito larvae. Full knowl- edge of the role of Plasmodium In malaria provided understanding of the action of quinine and helped develop more effective treatment for those who became infected. Still more recently, study of mosquito behavior has reveakd their habit of taking a siesta on walls immediately after stinging a victim. Treat- ing with DDT the bedroom walls of malaria patients has become an effec- tive method of blocking the spread of malaria at its source. Thus, the vision of complete eradication of this disease, which still kills more people in the world than any other, has actually become practical for the first time in history. It would be possible to debate phikosophieally the degree to which it is proper to call swamps, mosquitos or Plasmodia, respec tively, the "causes" of malaria, but this is obviously quite unimportant once the practical result has been obtained. Stadlea oJ T.oin. Pose Atet. Coweepts From the outset the adherents of the tobaaoo theory have chal- knged research workers to preseet any other plausible alternativr ecphna- tion. They may have admitted but minimind the possibility that the as- sociation between increased deaths from certain disea.us and excrsive 8 9

L7 ~ --i ~ ~ ~ ~ .~ 0 0 ~ t~-a ~ l'-& CJi 00b smoking may reflect the common influence of a third factor as yet un- eccognined. However, two events of the past year-one specific and one general- have provided interesting evidence that the simple cause and etiect relation- ship so eagerly grasped by the tobacco theory proponents may not be the correct explanation. The first came from Sir Ronald Fisher, one of the world's great sta- tisticians, and a renowned biologist as well. In the July 12 and August 30, 1958, issues of Nature, he published two sets of interesting data on the smoking habits of twins. These papers show de6nitely that genetic charac- teristics may bear on future smoking habits. Sir Ronald obtained data on the smoking habits of twins-both "iden- ticar and "non-identicar sets. "Ibe identical twins, arising from a single fertilized mrom, are really a single individual divided during early develop- ment into two separate entities. The non-identical twins, arising from two fertilized ova, arc on the other hand two distinct individuals. Fisher's first paper showed that among the pairs of identical twins, only 24% had clearly unlike smoking habits. However, among the non- identical twins, S1 % had differences in smoking habits. When the question was raised that the identical twins might exert greater mutual social influence on the smoking habits of each other than did non-identical twins, Friher published additional data. These confirmed his earlier findings by showing only 17% of identical twins with unlike smoking habits compared with 50% of non-identical twins. In addition, these data showed that the similarity in smoking habits among identical twins was consistent whether the twins were separated from each other at birth (as about half of the sets were) or not separated. Tfie question of common home environment was also considered by Fisher, and it happened that the periods of a common home environment averagod longer in the non-identical twins than in the identical; • a situation which gave a greater opportunity for non-genetic influences to express themselves. • Thus, it appears that a certain type of person is going to be a certain type of smoker or a non-smoker. Hence, if the smokers and non-smokers have different health risks as a part of their basic biological nature, a ,atitistical association of high mathematical significance between their smok- ing habits and probable risks of death would exist without smoking having anT causal effect as such. It is not nootssary at this trtage to argue the quantitative significance of Fisher's data. As they stand they show beyond ttny doubt that a possible explanation other than the cause and effect hypothesis glibly accepted by 10 I the tobacco theory proponenta does exist and must be evaluated. Sir Ronald himself recognizea that his data are of a prtliminary nature. They serve to establish a suggestire and stimulating line for research to determine the nature and extent of the bio-genetic factocs obviously involved. His results as they stand are not intended to replace or eiiminate the cause and effect hypothesis. They do, however, make it neces.acy for that hypothesis to °move over" to make room for another possible explanation. It will be intensely interesting to folknw the future de.elopment or research which may add evidence bearing on bb.gtnetic characteristits as determiners of human habits and human survival risks. Doe. Smoking S6ose Di,eerencea in ProplaT T1+e second category of evidence bears on the hypothesis that some heavy smokers are types of people who smoke as a reflection of a certain psychosomatic relationship and tempo of living that 'are characteristic of their biologic make-up. Such people would differ from non-smokers in one or more psycho-physiological characteristics which should, on the basis of research, be recognirabk. We would not, of course, expect to find that alf heavy smokers were of one or more different types from all non-smokers, but that in the smok- ing group one might 6nd a predominance of one or more types different from those predominating in the non-smoking group. This interesting and complex problem was discussed at sonv length in the 1957 Report of the Scientific Director. Data obtained since then from a number of independent investigations still do not identify specific factors. They do, however, strongly support the hypothesis that smoking habits may be a concomitant of active, rapid, inteese and venturesome tempo of living. Existing evidence itdicates: (1) that different biological trpea of peraoo ceact differently to psycho-physiologial stras and etrain, (2) that such differences in rtxc$oe are able to influence phys~. iological functional balance and ttnbafanoe within the body; nda (3) that such internal unbalaneea when out of control of the organism may lead to dbaaes of . tttatq types, Including totnor formation. There is no conflict betwoett this hypoMts bf petsonalitr-trpe dif- ference and the results of 3ir Ronald Fishen tl(udy bn twfns. In fact, `tn- etic differtnces in potential neuro-physbkokical tdmpo would be one of 11

the basic contributing factors in the origin of psYchosomatic differences being discussed. Continued research in this field, including informal conferences of the type already being held under the auspices of the Tobacco Industry Re- search Committee, is essential for progress towards complete analysis of the relation of smoking habits to health and is being supported by the T.I.R.C. Pathologk.l InvedIRationa Pathological evidence, based on clinical observations, first came into prominence in an effort to obtain support for the tobacco tR+eory. In 1956 data were published on microscopic studies of lung tissue obtained at autopsy from a comparatively few smokers and fewer non-smokers. In a variabk number of areas in the lungs of both,'structural changes. in the direction of disorganization and disarrangement had occurred. These were reported to be more numerous in the lungs of smokers than of non-smokers. These changes were considered by the investigator as being precanctrons. This point of view was eagerly accepted by those who believcd in tobacco guilt. From the very start, however, lively disagxcement developed among pathologists themselves as to the screntific accuracy and significance of the description of these lesions as being precancerous. This was the crux of the whole matter, and it was the interpeetation of these lesions as "pre- cancerous" on which the whole edifice of pathological evidence was built. -~1 ~ .., lYr.e Report. Rrd.e Qne.dows At the Seventh Interrational Cancer Congress in London, July 1959, scveral research reports were psesented which have thrown new light upon this problem. M An independent group of American workers reported that the inci- dence of basal cell overgrowth was observed in lungs of both smokers and non-smokers. So also were the various degrees of structural disorganiza- tion known as metaplasia and the advanced form of this process which had beg c'al led "precaneerous." These changes occurted approximately as frequently in the trachea, which is direcrfr eaposed to smoke and where cancer incidence Is fow, ar ~ they did In areas of the MoncMu wAtre catcn lnddence is much higher. G This clearly e•stsblishes that such changes are not as such diagnostic for i "pre-canoer" and that to interpret them as such is to exceed the scientific evidence. it was also rcported that "The oeevrrenct of pneumonia leads to epithelial patterns (tissue changes) similar to those found in heavy emokers" This important fad proves that these changes are entirely non- specific for tobacco smoke. These facts, coupled with the previously mentioned strong doubt ex- pressed by many pathobgists concerning the pte-caneeroen significance of the observed pulmonary tissue changes, shake the pathological foundation of the tobacco hypothesis. Aninaa Experimentstlon Workers in the field of caneer research know that the results of animal experimentation, while important, should be extrapolated to man with ex- treme caution. This is especially so (1) when the animal tissue ditfen from the human in type or location or both and (2) when the substances found to be active in animals differ in constitution, concentration, or in methods of application from those experienced naturally by man. The clas- sic exampk of a case in which most, if not afl, of these variables are in- volved is the series of much publicized animal-painting experiments. On the basis of the extremely insecure foundation of activity of tobacco smoke concentrates painted on mouse and rabbit skin, there has been built an extraordinary superstructure of surmise and prediction, even to quantitative "exactness" applied to man. Lung !~',rperiment. 3hote Neaetlee Rtadb How ill-advised and insecune this procedure of over-interpretation is was shown by results described by a grantee of the Tobacco Industry Re- search Committee at the 1958 International Canoer Congress. In this work, the rnnas of mice were heavily and repeatedly exposed to whok cigarettr smoke obtained by methods closely akin to human smoking. Contrast this with the animal-painting work which used the sJEln, csually shaved, of mice, and the vlsoors cnncenlrotes of great quantities of smoke condensed from a smoking machine. There is no doubt as•to the greater resemblance • of the inhalation work to the actual conditions of human smoking. In the inhalation work there wetr observed In the lungs of mice tissue changes dircctly oompanbk to those described by the pathologists in smokers and non-amokers. Then, however, two vitally impotfi+nt thUp ttuppened: (1) When exposure to cigarette smoke /ros stoppcd the changed 12 13

arras in the luna disoppeared and were replaced by normal tissue. This •.as one proof that they were not in fact precancerous lesions. (2) When exposure was continued into old age and throughout the normal life span of the animak, the ksiav persisted but no cancer occurred. This was a second proof that they .rere not precancerous. 'llaere had been reported previously negative resuhs in animab after exposnre to eigarette smoke. In no previous case, however, has there been a paralkl series of pathological observations so extensively and accurately Made. The reoent work Mported above definitely leaves it up to the adherents of the "tar and skin caneer" hypothesis to show why and how that highly developed and propagandined theory provides as direct, significant and extrapolatabk evidence as do the uwre recent negative results of smoke in the Mngs. 'The simple facts are that (1) Inng tissue in mice is more like lung tissoe in man than is skin tissue in mice and, (2) exposure to cigarette smoke in mice is more like exposure to cigacette smoke in man than is painting with smoke concentrates. .) i ~ . . I Sa.nd Progress Poreaetw The tobacco theory proponatta therefore now find themselves with the "patlalogical" theory shaken and unconvincing. Furthermore, the indirect animal "experimentaP" supporting evidence used as a atrong argument is no.w largely vitiated by more direct and sig- nificant negative experimental results. As a eonsequence, the wfiok situation lapses back to reliance on the sole line of statistical evidence in which, as before stated, there exist as many loopholes and contradictions ttdq as there did when it was first advanoed. When scientists teeogniae and can evaluate experimentally the various aompotteM ekmenb of hmg eanar ansatioe, sarod progress towards solution of the problem wip ieevitably follow. Until then, the scientific or lay ovaemphasis on any one suspected facta .ritl contlnne to be a dangaont threat to the public's appreciation ~ of the oompkxity of the skuation. It will, in itself, create a false sense of .chie.ement and beoome a most unfortunate delaying action. We thus find ounehes in 1959 wwith ditletsnces, doubts and eoesicting eWmt of statistical evidenoe still eoreoondled; with the pathological "e.i- dnroe" shaken. selective atd hadequate, and with the most direct and -Aultkal evidence from animal experimenhtion showing negative results. Stated in another war, the expottents of the' tobaoco theory have ~ 14 ~ failed since the first public propaganda for that theory in 1953 to prvduce any nea evidence that provides any qualitatively different support beyond the original type of statistical association. Extended efforts to provide soand evidence have failed. Evidence of the importance of facton other than smoking In the etiology of eoestitu- tional disease has greatly increased. Both jnstify eontiaued autioo and skepticism to..ards acceptance of aggressive antkobaeco propaganda of genenl and conclusive nature. ObkeNee. /or Resoareii To summarize briefly the general backgtctund of causation factors in constitutional disease, an analogy may be drawn with the operation of a key in a tumbler lock. The key will turn the Ieck only when all ekvatbns and dcpreasions are aligned in ppvper order and have the right dimensions. If a single depnxsion is 611ed or a single bump Bkd away, the key may fail to operate, although sometimes minor alterations may be made in the key's pattern without destroying its ability to unlock. At prosent the nature and identity of most elevations and depressions in the etiologic key to lung cancer are unknown. Thus, eRorts to assign a primary causal role to tobaoeo use on the basis of statistical associations ignore all the unknowns and focuses undue attention on tobacco use. The statistical relationship might exist under any variety of conditions -whether smoking exerted an incidental, aecidental, or non-specific effect or was associated with physiological or psychological characteristics that were independently connected with susceptibility to the disease. This situation emphasizes the fallacy of attempting, from prcaeM knowiedge, to assign pereentage figures to any possibk etiotogic agents involved in hmg cancer. Such efforts neglect enticelT the still unidentifled factors that are certainly in.ol.ed and Ignore abo the posdbiUtr of syner- gism or intenctioo among facton. It Is cattoelvable and probable that any one of several conditions and factors, aeting alone, will fail to prodnce the disease but that a number of these, wben acting In ooncert, might create conditions favoring the appearatroe of tfte diowe. The proper objective of raearcb Into luttg aecw etiobgr mnst, there- fore, be to Identify tnd deterntine the ditnetr:toes of as many factors as possible that may constituEe the ekqtkotta and deprtaaions in the key to this disease. As undastatdg d lung attotx ptoN esses, Mcceasingly efe fective and practical measurd fat pcl.lMiott and eotttrd may be expected to evolve. It b reasonable to ho(6 t11rt aottle eltes'We otulftol may develop bng before every phase of the etloloo Mb b6W t!(ntlpletely elucidated. The necessary alteration in the key may wefl pw.e to bf ta simple one. 15

The task is to find and piece together the numerous elements involved in producing lung cancer and to locate the link or links most vulnerable to interruption. In this task, the interests of the public, the medical profession. the voluntary health agencies, the government departments and the tobacco industry, as well as other industries that may be concerned, are essentially identical. In developing a research program on this rational foundation, it has been recognized, of course, that there is a vociferous minority that advocates total elimination of tobacco use as a possible control measure. As a practical as well as scientific matter, it is highly questionable whether prohibition of tobacco use is feasible or pocsible, whether it is desinbte, and whether the sought-for objective.would he accomplished. The prohibition concept discounts or ignores all consideration of smoking benefits in terms of pleasure, relaxation, relief of tension or other functions. These are admittedly ill-defined at present because they appcar to lie largely within the soeiological, psychological or border-line psycho- physiological realm, where techniques of measurement and evaluation are relatively under-developed. Neverthekss, they do exist. An unemotional and objective consideration of a practice adopted by a substantial majority of an adult population suggests that the functions of tobacco use in modern life have some positive significanee and value. Efforts to describe and evaluate these functions in positive terms by long-time research should. and are being, made. They will be continued. Recognition that the search for factors involved in lung cancer still has far to go was clear in November. 1959, at the National Conference on Air Pollution. Public health authorities and scientific investigators ac- knowkdged then that evidence implicating air pollutants as factors in hmg cancer was as great as, or greater than, that pointing the suspicious Bnger at smoking. ~~ Who knows what other factors need to be identified? Many possibil- itks have been suggested by clinical observations and by animal experi- a- ments. These suggestions have to be supplemented and explored. The histories of lung eancer victims need to be investigated minutely for common ekments at every level of personal and medical experience. Sev- Ukral efforts already made along such lines have demonstrated the promise that this approach may yield vital information. Among these may be men- tioned the studies of Walter Finke, Sheldon Sommers, and of Norman Trieger and his eollaboratoa. , The constant danger (s that this highly necessary kind of hunt for dut?s will be handicapped by the unjustified presumption that a primary ausative factor has been idt:ntitkd. F~+ i Progress in Research This section is devoted to a discnssion of the major lina of research effort being supported by the T.LR.C. The tesaftvh ptogram has a two- fold purpose. One is the advancement of teedical and scientific undentand- ing of cancer and heart disease, particularty. The other is oriented to questions that have been raised concetning the possible role of tobacco In these and other human ailtnents. I. Basic ,Bxperirnentarl Resean;h The Scientific Advisory Board has contimted its active Interest in three broad lines of research that need develoQrnent and extension. It is hoped that such procedure will contribute steadily and eonstnec- tively to.rards the establishment of a sound foundation of experimentai research on the etioiogy of consdtutional diseaaw. Such a foundation is an essential element In the final solution of the problems of control or causation of such diseases. Its development must precede and will generate the type of quantitative atudies which are an absolute prrrequisite to exact knowledge. The three lines of research mentioned are (1) Carcinogenicity and Bioassay: (2) Tissue Culture; (3) Psycho-physioiogical factors. 'i?tese will be briefly discussed In order. 1. CARQlYOGErRQTY AND BIOAg3AT STUDIE3 When the statistical association studies on hmg cancer incidence and cigarette smoking were first published, the simpkst and most obvious tlKOty pursued was that tobacco smoke might contain one or more substances known to science to be carcirwgeeic to animals. Among such substancea were some polynuckar hydrocarbons that previously had been separated from coal tar and Identified by chemists. It was logicat, therefore, for investigatots to undertake chemical anal- ysis of tobacco smoke.vith the very sensitive modern techniques of chnoma- tographic separation and specttometric methods of identification. Such investigations have been under.ray in sereral laboratories for a number of years and a certain amount of atodc-hking ia now possible. The use of these modern ttxhniquea has retohed in ttiporta by ittde- pendent Investigators of the detection of aeatal'poiytmclat hydroearboty in cigarette smoke. None of these has tatteaffy *W 1btsEed f[otn tvbaeoo smoke in a visible quantity. They have only belft ldqtti}kd spectograph- 17