Tobacco Documents Online

C SlB-CO:u1ITTri: P.FiPO: T Imtroduction The Sub-Ccmmittee members, representing the Scientific Advisory Board of The Council for Tobacco Research - U.S.A., have met and considered the research needs and priorities in the field of smoking and health. In their consideration, they gave particular attientiom to the Report of the Surgeon Generayl's Committee on Smoking and health (1S,64) and to the 1c67 and 1c,,68 Supplemental Reports of the Secretary of Health, Education and Welfare, as well as to the views of their colleagues on~the Scientific Advisory Board, of consultants and cthers, and'to recent pertinent scientific literature. Particular attention was al!so focused onithe present and future pro;r<.ms of The Council , zrh~ch have bcen an3 are to plan for -nro support research onithe background and'development of the diseases with which snokino has been,consid'ered as associated. The cb,jecLiives are (1) identifying and evaluating the relative significance of the factors or influences involved in the etioloSy or development of these diseases, and (2) distinguishing between factors or influences which r.:2y be merely associated with these diseases and other factors or influences N:hich contribute i.m.oortantly to their oriCin or course of development,. It was decided'that the Sub-Ccrrmittee objective of statin the research needs and'priorities necessary to imnroved scientific understandin-, in the field of erokin~; and heclth involves, the follo-.:in7 eonsilderrctions concerning the diseases in question.

LUNG CA'TFC&R By 1~?68 a number of factors had been identified as associated with malignant enithelial lung tumors. These factors are (1) cigarette smoking, particularl,y heavy use; (2) urban residence; (3) country of'residence; (4) occupations with exposure to inhaled materials, including radioactive substances, oils, wood dust, printer's ink, etc.; (5) maleness; (6) various lung injuries, infections, pneumoconioses and other conditions that lead to pulr.:onary fibrosis; and, finally but of probably major importance, (7,) constitutional andi genetic characteristics. It should always be remembered that, scientifically, association, is not causation. To understand the relative importance and possible etiologieal effects and i2nplications of these associated factors, it is necessary to consider pertinent information and'.deficieneies in knowledge in several areas relatino to lung,cancer: (a), Clic ;r: L:: L= The data s'--ojst=---_soci=ti_r.s bet•. --•- various environmentall and constitutional factors and histologi'cal types of'lung lesions are inconsistent. The wide variations amony,,pathologists in; tumor classification and uneven application of Ivi1O-proposed criteria apparently affect the reported incidence of squamous or adenocarcinomatous tumors in different,series. Criteria for the establishment of the pat?iologicsl diagnosis of dysplasia and other non-neoplastie lesions,are unsettled, as is the ulti:mate question of'the relationship, if esiy, of these lesions, as well as rr.etaplasia, to lung neoplasia. A better understanding of these relationships may in turn shed light on the serious probler of oahy squw,:ous cell carcincma is so corr:mon in th= lun,,, v~here normall,v no squamous cells exist. (b) Eridc.maolo7l: F.pidemiolo„ical dat w haveprovided thebasi~sfor identii?yi:^t; the dsz= seven f!-.ctors above noted as a:soci^ted!rrith iun; ttu^ors. tinf:rtu ateiy, virtually C (.j ~ all the avciIoble dnta derive fres:i studi'es of sin; le or sr:al'1 r.u:,b=rs of f. ctors. ~. The growin~realizntion tli-:t ,_n ilnterIDl^y of r_+umProus 1.c-.ors t.ayweli_ be o2' r..;,'or ~ •, `_ etiolo;_cal :ilc-,rnificance sh~ulaencouy; r-e future i,r.tY i_-_L~el3. ~

- 2 - (c) Patho^enesis: At present, there are no unifbrmi or consistent data from which the roles of exoEenous and endo;;enous agents in l.un,=~ tumor patho_;enesis can be defined. There has not yet been a satisfactory derr.onstrc..tion of a pathof-;cnetic mechanism or mechanisms by which, lun., tiznors develop. These observations opply to animals as well as to man. (d)' Lung Ptlysiolo ,ry: Clearance mechanisms for irritants, both perticulate and nonparticulate, are incompletely understood, partilcul'arly the interrelationship of the action of alveolar macrophages with those of the cilia and the mucous bla.ti et., Specific experimental effects of various f2ctors on isolated individual elements of' the lung clearance mechanism have been rerorted. However, the existence, significLnce and relative eontribution,of t;e net effects of each such factor on the entire 1un~, clearance capability remain to be shown both in enir.:als and! humans. (e) Endocrinolo=f^r: The predominenee in males of such puIronary diseases as bronchi'.tlis, emphysema and lung cancer sua,,Iests a possible endocrinoloo7,ic<1 factor not as yet defi..ed. Part'her;<<ore, the reports of a relationship in both r.:en and: 1:or:en bet•,:e^n certain endocrine abnormalities and certain histological t5pes of lun;; tu._ors si::_„est the possibility of an as yet unidentified endocrine role in ea.rcino[7enesis. (f) Carcino^enesis: Distinctions bet*:reen: caz•cino;ens, co-cwrcino-~ens and pronotin3 substances need clarification. For exar-mle,,in tobacco smoke,, such potential substances have not been accurately defined, and problems such as the differentiation of the possible effects of the gas and pr.rticulate prases need extensive further .tud.y. The presence of more than a thousand substances in~ s:r.oke itself represents a l!arge field for investir-,ation;. There is virtually no information on pD,,sible int'erreletionships in ciu•cinc-~enesis between such fc.ctors as occ..nztinr.a1 c~t-ospheri.c _r.-)ll::tion and'tobaceo smoke. Studies of'r:ali ;nr_n tr;nsfor:r.ation of cell!s erpoced in •ri_tr~ to crreiir:_: ns have been reported,,but the subject in controvcr:iel„ r. -i the basis of the chr~°-es is as yet morly rnderstocd.

(g) Anir,i.1 Er.oerimentation: 0f salient importance is the fact that no satisfactory animal model for bioassay presently exists in the field~of lun, ; tumors. Skin painting ex.pcriments contribute little to understanding,lur.g cancer. Proper procedures for exposing, harvestinF, and processimi the target organ lungs in animals need further development. Once developed such methods need~wide application. Biochemical, histochemieal and morphologic studies should be better correlated. Reseerch1Iceds The above considerations su_,L;est the following reseerch needs: (a) Huma.n & Animal Problems: Studies of pathol,ogy related to lung tuc:ors should~ bee designed to answer auestions as to what,rr:eans of el'assification of lung cancers can~ be developed and rpplied to permit a more consistent evaluation of various epidemi~o- logical, clinicopathological and other studies. By what mechanism or mechanisms do varirnue l.un; inflammatory, degenerative, prene::nlastic and'neoplast'ic lesions develop? 4lhat is t!a patho,7enesis of lun~7, cancer? One would like to know whether the steps in its aavelop7nent c.,,n be sufficiently elucidated to permit creation of techni~oues for evaliuatin;-; the possible etiolcgicell si[;nifieance of' the various suspected factors at various tine intervals. The significance of individual non-neoplastic lesions, if any, in the development of lung tumors needs investigation. Recently, many viruses -,:hich, infect the respirator•,,, tract from birth to old age have been identii!ied. 'r,h;_t roles, if any, do these viruses pLa;f?I Does virus Smplicatilon differ in tunors of different histclogjeal types? It •rould be useful to know if there are si;nific:_-:t differences dependin,~ on whether the oril-in of the lesion, iisa vi:rall infection, a -neuroccniosis, sc:^e other lun; disease or other enviror-,°.enta1 or host ir_flucnces. C-rn any latent period or ti.!r.e period be deter:?Lned tr, cluc~ l~te the st; ~ es in lur.~; c<ncer cierelop=t? It may be thatdi1'ferent hi^tclo~ical trn°s of '_un~; cr_ncerit:'.hed ,r~^ dii'fcrent diser_se~ .rith differentetiolo,;icc:.

The reported correlations of various hormonal abnormalities with specifi.e histological types of lung tumors ney be indicative of hormonal influences in tumor develorsent or indicators of different host reactions to various disease processes. It would'be of value to,distinguish~between these possibilities. The correlations (or lack of correlations) bet"ween various environmental and host factors and the sites of origin of lung tumors -- more cent're11y in the lung or further toward the periphery -- are as yet poorly understood. Physiological studies of exposure of'humans and animals to tobacco smoke should be designed in order to answer such auestions as what are the sites of ex.nosure and the chemical reactions at such~sites of particulate phase constituents, or of gas phase constituents. ~ What are the effects, if any, of cigarette smoke on~total lun= clearance? This might repuire distint-uishing, between the particulate phase and the gas ph-rse. It is desirable to find out what happens to cigarette smoke constituents in the nujr!ar, body and what metabolic products result. Also, how if at all does ciFarette zmoke in the hu.,.an respiratory tract differ from,cigarette smoke produced by experimental smokin~, machines? The significance of any such differences is important for biloassay methods to be developed. The actual' d'cses of cigarette smoke and its constituents to which individual human smo ers are exnosed need! to be measured'in some way. How can chronic erperimenteli exposures of aniayals comparable to those of humans be accc:plished and their effects meFsured. Tissue andlorrjan culture techniaues mdFht prcvide either screenin;, or indicator information to permit r^~)re efficient develiopment ;:nd'utiliz[,ti:on of thc: lonZ-term bioassax systems. (b) Einn_rnil Probl:: E::1,eri^rentei studies sl:ouldi be desir,nedi to rceet, such needs as the devellc-,;ment of an adeq_uate model system, or sy.atems fc,r bioassoy. If such system would result i'ni the out~ro-th of hn«,.,n 1: ,rQ Pr;'7~,r l'l t,,r:-, t,,,, ~rs,„ t'-en

C -5- C enviroonmental factors (:e.g., cigarette smokin6, air pollution„ occupational exposures) and host factors (e.g., hormonah influences, liver d'anage, Cenetic strain characteristics) and'viruses can be introduced individually and in coa:bination, so that their effects on the model system can be determine&. Can a means be developed'of applying cigarette smoking in the model system, in, a manner reasonabl,y approximating t'hat of' hu;^an smokinl,? The role in experimental lung tumor development played by individual carcinogenic substances and the conditions of their activity is a matter of major i.*~:port'e,ice. Are there various possible cocarcinogenic or promoter ef~'fects?' Fire there inhibitors or anti-carcinoaens and how, dol these various substaneee interact. If a system can be develioped for reliably bioassaying individually the r,.ore than a thousand substances in cigarette smoke, as they occur in that smoke, it would assist in the identification of any whichimay play any role in disease prodt:ctioni. Whether viral factors: are necessary fore=,,-rimental lung t:um.or produetioni and whether viral! contaminants must (and can) be excluded is a,n em^:: inry, reajor field o.' research. Or are viruses necessary to provide a rea:^onsbJe of the hu-man condition? If so, controli of viruses for evalu ti;on purpcses would be import ant . (c) Huma,n Proble.^.s: Studies of defined populations should be designed to c:nswer such questions as the yield of information concerning possible interrelationships of various environr.:entPl and host factors obtained by the anpl'ication~of r.:ultiveriate statistical anaLysis tuchniques to data obtained in t:ultifactoricl studies. The relationships to lung cancer incidcuce of constituticnally influenced h~bits and st;yles, of life, as orell as built-in predispositions, ce•n thereby be clarified~. The trends of lun,~ c?ncer i_nci-cience in differentporul~-t`_ons,:,ith 'i~i'fori^,r environ:;.ental-end host,factor cheract'eristics m,~y be cquite diffcrent. D7f°crim,,;lung cancer incider.ceroatterns ih v^rious popul-ticros nny be co^rel:tec, vriitY6 one or r'ore environ ent^1 or host fcctors.r_,.nd :interrclationsiiiUs, of th-se _._„•,. ,..cv tiroveto be corre]atr_ci .;;-th inr_ir,,,!ice -a.ttcrns.

The correliations (or lack of correlations) in various popul~,tions between the incidence of lung cancer and of other diseases, both those st4tistically associated with tobacco use and others, need investigation. h'hat are the age trends of lu^.; cancer? What are the possible et'ioloL;ical implications of the constant or sligiitly increased'age of onset of lung ce.ncer? Lung,cancer incidence patterns in women, includinFr ~ the apparently still- widening,m2le-female sex ratio in the Uni~ted States, may be correlated!h'ith dilstinctive interrelaticnshirps of environmental andlhost factors. Factors or characteristics, other than disease incidence patterns, differentiate smokers fromlnon-smo;.ers, pipe ar.d' cigan sr.:o::ers from ci,l smokers or ex-smokers from either smokers or non-smokers. Do any such factors or characteristics have possible etiologi~cali significance?' Exposuse dP.tn shoiil.d be obtained superior to the presently available widely variable data concerning inhalation hebits and disease incidence. Criteria for meastu•ing exposure developed in lung physiology studies can be used in pcpulation studies once they are availc.ble. What factors Pre correlated with the observed declining rate of increase in lung cancer incidence? Postulated periods of latency develioked in studies of lun;; cancer pathc.-er.esis will likely permit confir:-tation or refl tation of the corrcl_'.tion of'the trends in one or more environ,ental factors with lunE; cancer incidence trends. The po: sible effects on lun.; cancer incidence trends of the ~;radual chs.r.~r;e in the population characteristics brou"ht about by the ir.crealsin.^l'y s;iccesrful! thar:,py of v<sious ch_•oni~c ar:d! acute 1unt~ diseases, 1-ni.ch thus l.__:-:its more "res.p.ir tor;,^ ticn. cripples ' to li-;e into the c: ncer I ~, «srant prospecti-,,e in tiI Tho eerrelr~tions (or lack of correlations):betr'ccn dilf'ferin::xend's of nd t} incidcr.,~e oi . sic s}iistolo;iccl tyr.es of lun:; cancer in .1Pferent areas a r:,.,,. cr..__ ._._.. c1 .... ho:;,, fr.",.,.., ,.. _ .

oTIER M;CTv.' s~ In cancers of other sites, many of the research sur~gestions made in the previous section are applieable. Ndtionally cnllected'data ominci!dence rates of cancer and other chronic diseases would:be more valuable and less subject to local discrepancies than the present information from state cancer registries. In both sexes, cancers of various sites in the oral cavity, upper respiratory tract and larynx, including,salivary gland tumors,,have not been adequately investigated'with reference to various,f2etors, as well as tobaeco use including rethods othe^• than by smokir.,;,. The tenuous relationship between tobacco use and cancers of the esophagus, pancreas;, kidney and urinary bladder is curious. Should further inuestigation of urinary tryptophan cet'abolites be pursued im smokers? CAP DIIOVASCU?,"SR DISEASES (a) Fri P n]_, , y; ?sultivariat.e analyses of „t:.~r:ez~~ou s:ez~~ous fdetars, including sex, n,re, blood pressure,,cholesterol, trifl'ycerides, lij~,prcteins, p},,ysical activity, soaio- environmental stress, personality type, hormonal factors, r..orphological characteristics, weiGht, smoking habits and family history should be undertaren. Large prospective studies including adolescents, females and special populations (Japan, Israeli and u.ndeveloped countries)i would be helpfuli in providing , rnore ade©uate d-;ta. b.ore sophisticated'statistical approeches to appropriate and reli4ble data are needed to work out the relative importance and interdependence of the multiple factors th^t have been associated with coronary heart disease. The reported statistical as.,oci'ati'c)n in rnal'es bet,:e-n clE:,~'rette srno.klrr, ::nd coronary 4rtery disease i s not cle rly of etiolo;;ic sir,nif ic nce. D^ta on fer,r.les ?_re scanty but pnir.t up an ic;portent "rialer ..,.. f: ^tor in c rdi .: cul y di's ~~.e -- st^tistica_lV sic_il,r tv that fcu:,d in other di:eas_,, c-.ncer of t'> 1"z, ... :':is. "c.>nf~tituticnuli" predilccticn providcspriori'ty for resec-rch;;b dpterr:ii,e if f-:ctnrs

C -8- C found associated with cardiovascular disease are bound up~in a complex interpley basically governed by genetic makeup. (b), Constitution - T`.rin S+u~nes: Fersonality cn] behavicr have been thouL;ht toibe more significant in predisposing to coronary disease than factors such as smoking. Twin studies seem to offer a unique opportunity to investigate the constitutional factors, and to date they support their probable iTlportance. After a certain age the sex differences in cardiovascular diseases seem to decline, V:hich may,indicate changes in the makeup of the population under observation perh!~ps attributable to the death of a constitutionally susceptible suopcYulatibn or the survival of those who are constitutionally resistant. Dc.terrilning the extent of the influence of various constitutional and other factors and ascertaining of the reasons for such influences should be f;iven higji priority. (c)Pethol.o-Y- t,'etabolisrr, - Her.=to1'o:rr -Fa,vsiol^rv: The questionofemer_nin;'Ful relationship bet,,:een cigarette srnoking and he,=rt d:scese m.ey wehl be resolvec3 by research tod'e teli:,lne Wi]etilel a R"ieciliii71p11'1 exists byY:tllc'n ei'eaLeSS1,n1I1Ccnr- CI7^.P.~Gs in the cardiovasculrr system. Tio date, no de2 init.e rela.ticr:silip betvreen r:c':in;; sr.d the development of coronary atherosclerosis has been de^onstratedl. Imd'eed',, acrtic ca.licification~ ner se,, one traditional radiologist's crit'erion, is not a good ir.dcx of ordinary atherosclerosis. Furt'her analyses of autopsy series rre called for srith precise measurer:mnt of various pathoplp,r,siolouie factors. The usually suEEe:-ted r.echenicrs of'utherozclerotic dLsetses are: (.L) diet, ('?), lipid n:etebolismy (3)vascul.sconnective'tis:;ue dz.r.l2ge, cmdl (L)endoarino dysfunctions. ':~hat is the effect, if a7y, of ;zokin~r„ on these ft.ctcrs?' G~rbon rr.onoxi:;e as a passi.bleccntnibutort^,: :°n^erobic r;yocardisl r. t: io9_l~r. ^ •r.Ln~=el, r;orlobin ur hemeslobin-oxy-en di:.sociation and r,yocardi,;li i ifc:.rction r_•-e<:smore .I lyri s.. Any chan^esihcoranar~.-flci,~ prrdrtzced by nicotincor rei ^t- 1shou=d be ,.,,_._ed'bf the ef,~ect, if my, ~jn r.u',_it. .al :]r.. e.na r]l __..... _ ir',t,i: i^rnce ©f eny acute cn :;^es noteeL, the extrn~; r:f , itii.n t1:^ cii tel_,' ,^

demonstrated in vitro, as weLl as platelet adhesiveness, both need study in vivo, ineliud 'nc~ evaluation of' the possiblie: role of catechola.mines in t'Yirombus forra.ticn. typerglycenia and hyperinsulinism have been reported'associated with smokir; and with coronary disease, and their possible relation to the reportedlincrease in corticosteroids found in smokers are alsolfields of interest. Men who are heavy smokers generally have loirer blood pressures than non-smokers, which points to a paradox iln, the claimed relationship between smoking and hypertension. Blocd'cholesterol is:either higher or lower in smokers, d'epending on the population studied. Inconsistent data between smo!:i'ng and angina pectoris need clarification and explanation using objective methods for demonstrc.ting coronary artery diisease. Fatty acid and lipid levelis,in serum are not consistently correlated with smoking. The association bett,,een smoking and strokes is not at all clear znd~ much future research, is needed, ccncernin,; z.rtiy possible: relationships bet,reen smoking, ~r.d~ cerebral vasculzr disease. The reported associatilon between, sudden death and s,T.ekir.-- needs considerable investigation, especially in reference to arrhyth.^ias. A better understandin~7 of distinctions between studies of preva,lence and incidence appears to be of considerable importance. Theories on smoking,and cardiovasculzs disease should be:further tested in aninnali models. AsuitabLear:irral model, invo)'vir ,~ oaac expcsuses, ; ould' i_dou t~di~ygreatliy assist in increasing our undexstun3in,~. CfiF.O:IIC' YFI DI^F.ASvS, E'SI'iiCIl,+.L`r rti::,Cal'TiS. ..;:D. Chronic bronehitis r:nd emph•ysema are simiiLto zxer-io<.clerosia, in th^t most p^rsons r.o,^,car to heve so-r.c d'e;ree of'thesc even ap;p-rentlv berrinnin: before or A birthy with <_ ten:?ency- .,._ n