Tobacco Documents Online

Page ;APTEA 4 A CLINICAL AEVIBN OF 493 QATIENT3 QF LdHG CANCEA (197G-:980) 34 .1 Materials 35 .2 Sex and age distribution 36 .3 Ristologic types 39 .4 Clinical features 43 .5 Chest radiological patterns 47 .R libreoptic bronchoecopic patterns 49 .7 Cigarette smoking (A) Cigarette smoking and lung cancer 53 (8) Cigarette smoking pattern in our patients 57 .8 Survivals in untreated, inoperable disesse 62 ,.9 Conclusion 68 :HAPTaIt 5 A P2.ANN$D RQTR08PBCTIVS CLiNICAL sTS3DY OP 603 pATISNTB OF LUIiG CAIVCER (1961- 19s4) a8 S.1 Yaterials - 3.2 8ex and Rge distribution 70 5.3 gistologic types 73 5.4 Clinical features 77 565 Chest radiological patterns 80 5.6 Pibreoptio bronchoacopio patterne sZ 5.7 Cigarette smoking pattern 86 6.8 Conclusion 83 sV----

CHAPTfiIt 0 BPPSC2+ OP CHBld0T1cBtiAPF ON SURVIVAL ~ A STUDY Or T13RZE COitBINATION CRS1dOTH8RAPY 3CK5u8S IN 139 PATIENTS XTTLi INOPERABLE LUNG CANCSR (1979- 1984) 6.1 iaa11 ce11 carcino+sa - ILACC chemotherany 97 (A) patienta and srethoda 97 (g) Results 102 (C) Discussion 202 6.2 •Ron-sraall cell lung ca.ncer 109 (A) KACC ohemotherapy 1. Patients and Methods 109 2. Results 112 (8) Two !AY ehemotherapy schemes ia bronchial adoaocarcinoma 119 1. Patients and ltethocls 120 2. Result^ 121 (C) Discussion 132 8.3 Conclusion 135 CHAPT6R 7 CABE-C4NTROL STU33Y OF'PA3SIVE SMOKING, uDaos2249 8TOV8 OSAOa XND HOtiE INCEN3E Bt1RKING IN RELATtON TO LUNG CANCER IN ItON-SMO1CTR PBKALES (1981-1994) 7.1 (A) Introduction Passive smoking (8) (C) Kerosene stove Incense burning ooorin8 at hosae 7.2 Fatients and methods 7.3 Reaults 7.4 Diecussion 7.5 Conclusion 95 136 137 137 138 130 140 143 143 156 CliAPTSR a D2RECTIONB POR FOTVRL* STtJDIES 157 8.1 tnidemiolosioal studies (A) Ya 8on8 KonQ (8) In collaboration with Guangzb,ow (Canton) zii

8.2 Pathologic studi.s: (A) Cliaico-Pathologlc study 161 (B) 8carring (tuberculoua) and lung cancer 161 8.3 1loat determinants t (A) ltistocomDatibility (HLA) antitea• and lung 184 cancer (8) Aryl Hydrocarbon HydroxYlae• (AFIlI) inducibility tod lung aa.sacar 1135 8.4 , Bpilo:gue 167 SSPS#tSNCfi3 169 ~ ~ _ r~xiii CJ

In HonB 7Con8, lung cancer is the commonest lethal malignant disease in both males s,ad females. This thesis represented the first sd.jor oliaiasl study of lung cancer (197fl- 19i4) in the local C'hinese poQulation. The patients were those sdmitted to the IIniversity De9artment of Medicine, Queen ldary 8ospital, Eon; B:onQ, and a11 bad histologically or cytolo8ica11y proven lung cancar. Ristolo8ical typin8 was based on the Torld Seaith Orsanization Classitication (19$1), wl.th 4 major types of lung canoer, naM41;y_. (1) squamous cell carcinoma (SQ), (Z) smLll cell carcinoma (9X), (3) adeaooaroinoms (iD), #.nd (4) large cell careinoma (U). prereQuisite Zor,a ollnicai s=uay oz iunQ oancer is aGOurate cell, typin=. ?~ Yy • pha0e-one stl2dr was to assess ; :. - • . - , .. ,-. !t . . .. _ . . . • coliaborately with the Departmsnt oi Qatholoty.the osll typing __- accuracy of cytodis8nosio (broncboscopio and sputum) in our , . . _ . X . . ._ • :~... ... . . - . _ .... . . . , hospital• In s"tive-yezr study period (1978-1983) in E73 patients. for both broachoscoptc and sDutum cytologic cell typins, accuracy wzs highest in $Q and 89 (7Q-100x), toilow.d by AD (so-aQ!.). That of LA was IDuoh lower (( 97x), bct the number OS patients Was amall. The aext phase is collection of clinical data base by a ell.nical review o2 493 patients admitted from 1978 to 1980. The atale to lemais sex ratio vas low (1.87:1), reflecting the ~ ~ ~

(,l) ..-.Cliaical •datn w.re coiiected from 503 pati.nts upon : 1981 to 1994 high incidenee of lung cancer In women In Hoag Xong. In s.n, SQ was the predominant cell type (44'1), followed by AD (23'l), 814 (13%) and LA (?%), but In women, the preponderaace of AD (44%; SQ 31%; 81C 10%; LA 2%) is notasorthy. Cigarstts smoking waa a nsjor tactor In •8Q and ali. The relative risk ot lung oancer in smokers was 6.4 to 10.7 for SQ and 8g, but was not sisaiiicant with AD or LA (( 1.e). SQ and 8g, b4ing smoking-related, showed features ot a centraliy located tumour. Our AD, contrary to classiosl tsaching, also showed clinical, radiological and broacriAScopic teatures of a oentrally situated tumour. A tbres-Vart study was thea carried out In parallel from . ..:diaQnosis from January 1981 to April 1884. The fiadings of the ~~ __ __ .. .. ..'-at^Yti: P...~... ~ . . ~ . . . .. c . - .. / ' , r.view study were confirmed. The male to temaie ratio`rras loN ; A history of cigarstte smoking was strongly associated with SQ and 8k. The relative risk of lung canoer In smokers was 5.S with SQ aad 21 with SM In men, and 10.5 with SQ and 33.9 with aK in women, but not excessive with AD aad LA (l to Z.1). In women, AD was the predominant ceil tTpe (56x), and 48% of all cases and 83% ot AD were life-long aon-smolcers. Again, AD showed features of a predominantly centrally situated tnmour. . . N C"7 W ('!J N ZV ~ tV

(2) That our AD, usuallr in tioa-smoker iemales and centrally eituated, was intriguing. A case-0optrol study ot 183 temale patients and 185 female controls was carried out to compare their exposure to three common environm.ntai, inhaled r wv1' nr U.rnrann ntrnrn tw..s .i home incense burning. Analysis for non-smokers showed that kerosene stove and incense burning were not aontributory iactors (p > 0.06). Passive siaoking was also not shown to be associated with AD~oi the central type, but may contribute to AD of the peripheral tFD+ (p. ( 0.05). ( 3) ' - The uedian survival ol our ' ~ • ~ ~' ' •„ paLisats.witli uatreaEed, •;_ " _ i ± _.inoperable diseaae was poor,"_beiAS i month f_or srm`~all:aeli cancsr au.`.~`# : ~tj ... and 3 5 taonths to aan ll ll ~ . r -ama ce canoerv~e studied tbe Y ..w t ~ _ ~s: . ~. - . . ~. .~ ,~.. _ ... - .._ ,....,. i .. . . ~.. ~. . .. e~ r~ . ...~~ _. 'ti : ettect ot tAree combination chemotherapy sohir4es...nisurvi'val o! ~~ these patieaLs. ; 2n,'-43 patients oo small cell:carciaoma, : YACC ` -..- . (metnotrexste, adriamycin, cyoiophosDriamide and CCNU)'-__ chemotherapy was eStective (21% oomDlete and 53% partial response), and sizniticnntly improved overall patients survival (aaedian survival 50 weoks). In no;n-sraall cell canoers, horever, YACC cbemotberapy (in 42 patients) and luAM/Fiik sch.mes (Futra2ulJS-Fluorouracil, adriamycin, tnitoraycin-C, in 44 patients of adeaocarcinoma), were inett.ctive. Although partial response occurred in 3-Z73 of patients, there Mas no overall ~7 . ;.;. survival benezit. - - rJ Sv CJ M W

4 With a oore nor solia data bese,. oollat>orative studios are being initieted, iaciuding CitY-ride eQidemSoloQica2 studier, clinico-Dstboloeio studies, and studies of host determinants.

families, and should theretore be rr-ezunined." C. Tnoen_, ee Burnins a 8em^ - snrain8 of Chinese inoense at 9 temDies in worshiy ot idols or sods, a ooimnoa scene in tourists books, is part of the traditional Chinese owtoans stiil practised in lton8 toot. Burning of iaasnse at home, either for ancestor worshiD (traditioaal Chinese ttlial yiet7) or Qeity worship, is also co=on s2ooaF the large aoa-Christiaa looal , poyulation. Chinese iacense smoks has bssn shown to contain oarcinorens, (dchoantal i Oibbard, 1807) but to-date, oo studi.s ha.e bsen uad.rtaken to examine its rslation to lua8 cancer. ' GiVen that (1) ia IIonB xona, home iocsnse burning is oommon, (2) . : ~. ~ ' . .N _ . _ . . . . _ . . . . . - . . . that ~aany adait women in long Zon` are honsswi.ss who spenQ most F~ ~:..x..~Yk ..+•'"i' . .!', . .. . .. . ' .. .. . , . .. . . . S tbeir tiw at hooae iahaii iac.nse amok~ whioh ooat i ouciao~eas,sh"'(S) that song Zoa= is o*ererowded with ~oaay ~ .`ramilits 1i.ina irl bousss/tlats of area 400 to aoo saQars seetlr ' ' ' ' oaiy,4.fiieh. sould taerease the inhaled doss os_ aay Poteatial ~ . c.. ~ 4 t' "inhaled caroiaosea"_.nrssent ia a sM&11 home ars; it is concei.aale that inoeaso smoke might well be importaat in the sen,.sis o!' lung cancer in our women who do not smoke. A study was thersfore carried out to esaraine wliether passive smoking, kerosene sto.e cooking and incsase burning at home are likely cattsatiwe ta0tors ia ltta= oanesr in aon-smokin= rhinese wonen. This Zorms Part C of the 19a2-188d lung caacer

study. (soe Chapter 5, pp at?)• 7.Z ?atiQnts E Methods This is a case-coatrol study. Tuo aases woru all of the Chinose temale patients who wore adimitted to the OalvMrsity t>epe,rtment of Medicine, Qu.oA Mary Ho4pital, Roag RoaQ, bet.oon Jsnuary 19191 and April 1984, with blstologicallr and/or oytoloSioally confirmed o+troinoma of the lung of the tour major cell types (Types 1-4, 1t.H.0. Classitication, 1la1). - Great care .. .. .ar taken to exclude seoondary carcinoma of the lttag (seo pp 15 j.:. '_ 19) but othorriso all Chtneso Zsnale patisnts were iaoludedwith ;; o otY~er 'Com-irison p-tieat (oontrols)'~~~ ; ~ seleotion oriterta a ,~ c ,, ' Ort6opaidic` rards ~"~ `; ~ s wdre Chir~oso temule pa tients adAitted ` to ! ths . F 1~'~. sriod 19a2-2064 ; ooro ~ th rable ;~ ri t it l d Y ~ ~ pa ~ , e p u at ary, bip a ln Quera - y i....>:^ .. ..:-aF.<". ;~_, too lung `aaooer pati.uts ino sgo and .ooial ciass - both oases ~an4'" controls .ere!patisnts ot the thir4 olass Qeneraf.ards aaA +veru ~ " .. . - . :.. mostly trom the lower ianouw group. patienta rith pathologioal ,- tractupes due to s~aokiat-related Aalignaroia, and periphoral; : vasoular disonse-retated orthopaedio aonditions woro excluded. It is oonsidered that onr ortbopaodio oo4trols should not be biasert towards smokins-assooLsted diseases. Aii sasa• a+re interviewed by mysell, and the aontrols ~ by myseli or xLss Ciudy Ling, our toahoiciaa aW1 =esearah assistaat, who .as tralned tor this iqvestigatloo and thoroughly - 140 - CID I'-7 . W GJ FJ ~ 0'a U~

=amiliar with 1ocw1 cult+tre. The que+ttions oovored dialect grouD, oocupation, smoking habito, pusttvo smokins, domsstic cooking including keroso4e stove, and home incsnsc burning, ic form of a staACtardiwad du.stionaaire (Fig. Z.1). gor very ili pativats, or for pr.tients e$o spoke a dialect otber than . Cantonose or Hahlarin, arraagemont would then be made for their next-ol-kia to be iaterviewed witlr the patients as lnterpretor. :Attempts at quantitatioa Of petssivo smoking Aas been ~~ . .. . . ..J.:~w,lf.~ .... .. ... .. _ . .. :.. -: recogaiasd as 4iftiault (Royai College of physiclans, 1983; .~ ._ . . ._ . . . . : ~ . i> Yeiss •t al. 1g83).: aido+ttream smoke. to which the passive ``r amoker is •xposod, .is dilated by roore air to a vsrkabie extent. . ~: ,. . . . . .. ,_{ ~y- The -room air itself also oont;ains smokr.{~which has been iahaiea L?7}s .cw~L~ ~ . aad tbvn exbs1sd into the rir. t tmouat aad duration of smoke . - F ` ~xthk'ki a bitidtili ` , eposurs,--e smoers•,smonga,'sso aa veaaton Of . w=s otc. are all impdrtUnt varlables, 'aad the amount of the _ various components of tobsoco smoke brexthod by the non-stmoker from a smoky atmosphere are therefore extremely variable aad napredictable, anut there aro ao agrood •tandards for expressiaa the o+ttd4t of pollution of indoor atmosDheres by tobacco smoko. The sams problem applt.v to Quantitatioo Of ezposure to kerosone sto'vu eookins luoas and burning of iacense at home. I bad the opportuqity Of disoussiag thin with Bir Xiobard Doll during hxs visit to the University Depwrtment of Uedictao, iiang xong, in ~ •14t- Cj ' ~•. V.l ~ ,~

fIG. 7.1. 1S!"G cMcEu NrtS) IwKn.RE Maew: SeUAge: cate ArWriss (vistrlct): 1100 toey4' Ox: Oom in ID llung Kong Q Chip• In Ibnq Konq for _yrs. Di+loct fo. _ Occulw lla,:. .- for . xrs. Scriooliny 0 4 6 yr. (~>iyr• lleriGT 5tatus: U Stnqle Q/lsrried []' Midowed r's. Ilusbaw': eccuystian - i!'DK11i0: non-eeokor M •s-swoker yrs. Cj awker L] elparaltas Q Mnd•ev1i.J Cj'l v.tOr•plip .....Jdlay a tASSIrE tKX1HL: ,. Ikwa: S 1 n o f Nowo t Nun•saukar ..._..._. r....r M.~. Othors (I 6ether cl m ICC IqCF1tSE HWeI1M4: CJ Yct ~ No fer _,r yrs. L 1 Datty V rastlrals only l:J Ms,Ur house j= Ovca+dr 1wusW HISIUKY tlk ptA,MIJMAtK 18: No. fo.SI y...icrrlI tlo m Yes C:] Cbest a-rsyt (No. ~ /4nther

riovomv.r 1863, abd ho adviava tt1rt it nrould'bu boat to pwt.gorisue •xDosure as porLtive and aslative only witM qo further kttsanyts at qunatlt.tioa. Thf #tiQnilioawoe levol tort the risk ratios are onloulat.d tor 2 1Lrts, Test A and Tout B. Tost A is whethor tt,w rirk•rstiu is r.a11q gr.attr than one, uainS the Aapealan weru too 111 to' LV intorvierroa, and wo failed to arraate 14eotinA tlioir relxtives, Roci thuy wsre •xcluded troe the stwdy.' Of thu 185 c:uatrols, 80 wura treot.d for frxatures, 17 for indactive bucd sntii joint disriusas (includtaat tuberculosis), 1s for ostwuarttYr+asia, d for rbow.aCoid arthritts, and 6s for other orthonaedic aoaditionr. D.mvsraphic cbxracEerletics of the cases and oontrols aru coaeDared ia Table 7.1. Twd groups aro siwilar iA as., as ~~ N CD

lndicuted by the distributioa In Tabl• 7.1, and wediwn• ot 07.5 years tor esasss and 60 yonrs Zor controls. aocioeoonomic atuttid (ae wawwurud by ocroupitlvn, year.t ot schooling) and recent reoidaricso are also sirnil1r in botli groupr. It is therutora tonatiilorud not auurksary to stratlty thssu variibles In the xnxlyaLs. The cases a.n4 controls' smoking habit ims obtain+rd f.u detail a.s duYaribwt oA pp 86-87, Cbuptor S, xasi the rnsuits .rsro prusuatsd In Tai)i• 5.8, pp 80, .hicb ia rvproducsd turrd as : -OriLhi- 7-2 tur nLL;rv'.'ratArands.`: ThuiruruiCrc wurik dLrnuAtad in •rihItpttlr~~3. w ~ dr~ ` ~ w* `.'tborir.+~r~are a total o! 78.Aon-smokor,i ir~ th• aases and .... ` t1w uuutroln,:. and tbdy_ torm ths 144 Aoa-amokwri - poi>ulation Zor the proront aAalrsis.. Ro atteAptY at quautitatioa (except for Test A) Mas aade ar described above In tSethods. Thon passivo smokin8 (r), ksrosoao (1C) and incensu (Y) were eoasidetroct together, three tatersn4tLnS oircles own be drawn Whowiur savsa possible combinatioas oI expwstzre, and onar isolated circle (N) indioating thcrse who had never boon lxposud to any ot thoso- aourca (?ig. 7.Z). Passive w.oking includes exposure Eo smoking husbands, aohabi tiait relativos, or wurkmatos. - 144 -

TaLE 7.2. t'ELL TYPE OF LUNG GNCEB ]WD S?lO1CM S)tBI?-IN 163'FP•ltNLB P.%TIp+TTS WITN TYP£S I-IV LONG'CANCEIt;*1981-1984 'c p; < 19 padc•-yr S (18) 20-39 pack-yr 8 (29) s 40 pack-yr 8 (29) A.,s.. J ~.:. 1.0 ...+..ti.~...~3~ JLtii. ,' i 14 (8) B (4) 41 I . c~ (22) a A ias

Tic. 7.2. EXl'OSURS GZ'LGORIES TO PASSIVffi sMOKINC,. KEi1OSEN: AN© INCENS6 I NO $XPOS41t3 CLAIPlED . - 147 - ~ zV N ~G~

Tbu au,ubar o! Aon-smc>kind aases with typu:l 1(squamoou,t eell), Z(sm.till asll) and 4(lar=o 4el1) lunar canoer are smail - baiag se.rua, Chreu and tive rra2ructively (Table 7.d ), and did not thurotora attord auaniaUful etati,rtiual anulyois. ?ur type 3 adunur.+crcinoiaa, tUu patho6onosis of wtslelr we kre s,ost intsrwetud in, tharo wor• 60 eioo-aqkrkere, wud the proportions of ditfuruut wxpoeurv uatvrorles are Yubutated In Tabin 7.4. Yhe oasns uro utratitivd lntv vsutral and psriphurxt tumoura to o=oine tliw uontunti*u tltxt our propoqd*ranaa of cantral sdoaouarcinoma midrit be related to inhalud carclnugons. - Tho`''. - •-_'- - ~=risk ratlos (aod tchair-~+44nlticsacu toval) for .:Nrp.; • : •. . JC- . .. , , ... . . . . . .. _~ ~.. . . •. r .-. ~, . :.K . puurivn ,rawkin6 hwroreno and iooonno In our non-smokvr :--.adenucrresiaoms~ team.lo patiet1ts are stiorn iA t able ToCal._;. pnsQtvu ymofcina and passive amokiag'dus to smoking husband alUae .. „- . ..... ~ .. . ._ .... .. _ ~ !-T - .. . . . . . _. . . .. ..._. .i ~,~. aru usamined soParataly. 7.4. DI scussion Tho problem pusod brrtoru us is the preponddraave ot adauocaroinoma o[ luag, usuaily In non-swvkvra, presIowinutftiy oM,itrally Niturtsrl, in uur IIsmais poPulwCioa. This cass-oontrcrl study wxs warried out to oo+npurod oxpusura of oaaoi-acd controls to thruo onviroqment,cl, ini,rled subst,ancas, namely passivQ smokLnd, kerugane atavu cvokind Zumva and home incensd burning tusus. The rasultr sl:orrod that (Tabie 7.5) kwrwwono anct incanso - 1Q3 ..

TABLE 7.3. DI?FSRENT lXPOSURE CATLGORZES !OR TYPCS 1, 2 i 4 LUNG CANCER IN NON-S!lOKIl1G MfOMEN • Ko, of cases Cat~egory squamous csll small ceYl large cell controls (1) (2) (4) * ploase roPer to Fiq.?.ti. I

9'AF3LE 7.4. DZPFERL'NT EXPOSUItE CATEGOfiIZS. F'OZ ADLNOCAJtCINOMA Op LUNG IN NON•SHOKINO WdMLN ExposurQ '~ No. of oasee No. o: CatdOorY C.ntral Periyhoral controls PI 1 3 7 5 32 * pioase rtFer to Ytq.7.2. 3 1 6 1 13 7 21 6 . 17 ; 3S 8 40 9 28 144 . i

eti aru oot ooutriLutcrry tactors tor adacouarcinoma, Central or periphural. Although th4 risk ratio ol pasRive vwul;iat in gruutrr thin oao tor ceutral adunocu.rc1noraa, the levot pi siguitlcanom ia only nbout 10 par cent by Test A. There La howevur sugl{owtion ot puyy4ve smolcind useocLxted •lth peripharul adonocxrclAomri, partLaularly Naesivs tiAokinr duu to emokiag uuavunds. Ths di2turunoae botwofln TeMts A and A ta Table 7.5 could Lw duu to a non-liaear LoKistLc dose-respotico curve or to rrpvrr Lu weunifi+Ilad th+ 1dv41 si oMpuuuro tluv to Lnoomplet• l ti t : oruu a un M ap lva uvlixhcI.A1 rs Therct Isaa beun onl ouw wrt wt r~ p y l , pn ~. . < ~a,uukinL ,iu Nmalo luug Catiuer patiento in fiong 1Cung` (Koo at at, . . " 'ji' . A¢1'F.`+k.~.~== "~1083). „Koo••tuund ttaat paabive ymvkurs &a a!Sroup ha4 a relative riuk or lueti than uuu. 40 vt tbo 56 non-smokur patients (71.4%) aad 63 ot thu 85 cou-amoktar c:crntrulv (74.1%) have boon oxuosoA -' to puteive swAiag, which iu aot istutLatically dltlsreat. The aktientn howervdr lnclueiucl all oell typea and were hotsroge4eoue in ttala rctaos. In additioa, the autbor did not distinsulsh csntrYl and lwrlipheral tumours. The assvolat[un ut pYSUivu rrouktn3 and tund cancsr ~ wuoulal bu further pareoed. Thure kr good tfseoretical support CD tue LUu neeociatioo. Aeccretly, it was rMl>ortvcl that, like W rctlvc auaokore, tlio pasaLvo smoker L* exposed to the snato ~ N

radioelsments in the tobacco, as 50 to 70 per cont of the 210po appeara In sidestreaca sraoke (lintsrs & DiFreaza, 1983). In addition, the exposure of ths pa.ssivs smoker to naturally oocurrins radon dau;hters is incres.sad In a smoky environment. It eas estimated that radon daughte:,exposure could account for 20 to 100 per cent of lun= CaAcars seen In non-smokers (1tarley i~ Paaternack, 1981; •inters L Dihren=a, 1983). The contlicting findinRr of tbe Japanese (H:rayama, 1981) and American (GarSiakle, 1981) studies mirnt be aue to diftarenoes in .methodolosy (xeizz et al, 1083). A potentially important factor is that the American study lacked saokin= data on 135 of the husbanda of nonscoking romen in comparison to only 28% in the =Japanese study, phich may have created biases in the data. A . ~ ~ .: _ : .. ., . - greater number of working women, larger bolnes and a higher :.__ _.. ---.-..,. .. , . divorce rate in tbe Onited 3tatos a.e other factors that could serve to'accosnt for the differences in results b.treon those studies. ia Hon; lCong, the Drobiem of overcrowding is notorious, with many families living In houses/Slats of area 400 to 800 square fset onSy, and this would increase the inhaied dose oZ any potential inhaled carcinogen prssent in home environment« previous estimates would have sstimated the attributable risk of 1un2 cancer due to passive smokias to be 30% greater in non-sraoksrs •xposed regularly to passive smoke -153-

compared to nonsmokers not exposed (Leeds, 1978). It has been tound that both the Japanese and the Ameriaan~ studies were in tact consistent with ruch an efloct (Meiss, 1083). ~ Tbe apparent association between passive sa,oking aad peripberal adenoesrcinoma (and oot central ttiuaours} in our Datieats is unexpected, and the reason uncieir. It is known however that there is,a ditference in chemical composition ol Msinstream and sidsstresm smoke (Stock, 1980; Corrsa e*_ al, 1l83; Weiss et a1 1983):. iiriustream smoke •rnerges into the environment a:ter tiavint been drawn through the citarette, tiltered by the smoksr"s own lungs, sad thsn sxhaled. 8ldestream smoke arises from the burning end of the cigarette -and •ataro; directly into the eavironmfnt, ::. A11• these lead to . : . _.~ ;. marked differencea in'the concentratioc oZ the coastitueats of mainstream aad sidestream smoke, and many potentially tosic gas - --.- • -.. . .. . -.. . . . • - . phase constituents,'SncludizQ nitrosaMins, are in higher. concentration in sidestream smoke thac in mainstream srsoke, aad nearly 85% of smoke in room reouits from sidestream smoke (19oise st al, 1983). It is true, of course, that sidestream smoke is tenerally diluted ia a oonsiderably larger voiucne. Thus, DassiVe smokere are ezposed to a quantitatively smallsr and qnaiitatively different smoke exposure than active smoksrs. lfhetber this might produce different proportion of histoloQical -154 - _ ~ . c~ ~.~ Ca . . co Gfl ~ N ~

types of t=our (preponderance of adeaoearcinooa), with peripheral location remains, at prese.nt, conjectural. T+o of the limitations of the present study are the relatively small sumber of subsects studied, and the inclusion of only one hospital, albeit a large, regiona'1 gensral hospital. are currently overcomioY the imroense logistic problems and Larte, city-+eide nulti-hospital studies are warranted, and we pursuing further collaborative studies in this area (vids infra). 7.5. Concluvion 1. Our reQUlts showed that kerosene stove fumes aad home passive smokine is also not sao.n to be associated with adenocarcinoma of lung, whether c.ntral or periDheral. -..incense burning are not contributory factors Sar adenocurcinoMa ot tha central tyqe. The reason for the preponderance of central adenocarciaoma In ou.- non- smoker Seraale patient population has therefore remained unanswered. Z0 xlher• is°however suggestion ot Passive ssaokinj assoaia:ed with periphera3 adenoaaroinoma, partiostarly ,passive smoking due to imoking ausbaads. ;The reason for the peripheral location o, the associated tumour is N Ca N W m. N N 4b O /.

~_C not ciear, although it is known that passive smokera are s=posed to a qualitatively ditter.r.t smoke as oospared to active smaker6. 3. These Siadings assd to be ccntirmed by lirge, city-wide, mu1ti-institutioaal studies.

---