Tobacco Documents Online

-6 I VVI OYl s ) Br. J. Cancer (1973) 37, 965 (15531 SQUANSOUS LESIONS IN LUNGS OF RATS EXPO5ED TO TOBACCO-SMOKE-CONDENSATE FRACTIONS BY REPE4TED INTRATRACHEAL INSTILLATION P. J. cnSCAS', P. N. LEEt exm F. J. C. HOE; Fron, rhe'Hnz(elon Ldmatorfea $ur~ Ltd, GE(ey R.w3, Har-,agale HG11P'. 16e td'abaceo %weurch Cm.rwtl, Glen House, Btag Piace, Lo nlon SiVSD 5AG . #4bi>gaRoarl, Wimbledmv,Landm+SWt9 iQV . . . $ecetvai 3 February 1978 AceepWd?9 Ms.ch 1973 ..-____sSummary.-Twice-weeldy intratracheal Instillations in rats of up to 24mg of squamous neoplasms (SqN). Fraction (R } P)G, which is a fractton of ctgarette- smoke condensate almost as tumoripenic for mouse skin as the nearly 30x larger mass of condensate from which it is derived, could be given in this way for up to 40 weeks icithout excessive mortality or any marked effect on the rate ot body-weight gain. Bv contrast, similar treatment with Fraction N(QG), a fraction having very low tumorigenic activity for mouse sldn, Induced no SqN and barely any excess of Sq]I over that induced by either vehicle alone. The effects of Fraction (R + P)G on the incidence of Sql$ and SqN were both time and dose related, the effect on SqM incidence being already evident after 10 weeks of treatment. No SqN seen were unequivocally malignant, though, due to the design of the experiment, only 5 rats exposed to Fraction (R } P)G were observed more than 60 weeks after the start of the experiment. Other changes in the lung, including aggregates of alveolar macrophages laden wlth golden-brown pigment (GB11) and foci of cuboidal,/columnar metaplasia of alveolar epithelium (CCSi), were frequently seen in response to both fractions. Fraction (R + P)G administered in I was more effective in causing SqM and SqN than the same fraction admimstered in BS/G. The implications of tlte findings are discussed, particularly the possibility that the intratracheal/instillation technique might be useful as a rapid bioassay for comparing the tumerigenicity of different cigarette-smoke condensates. . (BS/G) gave rise to foci of squamous metaplasia of alveo:ar epithe;imu qh'L) a`3'~ -~- pffcred satine/gelatine Fraction (R} P)G-su"spended in-either-infusine.-(I).-or~ DFSrrr2 several attempts, malignant tumours have not been produced in significant numbexs in the ]ungs of experimental animals by exposing them to cigarette smoke (Dontenwill et nl., 1973 : Bernfield eff at., 1974; Daviiet al., 197Btt). Lung tumoma have beeri readily induced bc carcinogens administered by intratracheal instillation (Shabad, 1962; Pylev, 1963; Schreiber et a1., 1972; Davis el a?., 1975b). Davis et al. (1975c) also sbowetl that the repeated fortnightly instillation of cigarette-smoke condensate (SRB) was associated with increased incidence of euboidal columnar meta- plasia (CCM) and squamous metaplasia (SqM), although such treatment failed to induce squamous neoplasms (SqN) in the lungs. Certain fractions of SWS known to be carcinogenic to mouse skin also pro- duced CC31 and SqM. In addition, Prac- tion P(SG), which oontains most of the polycyclic aromatic hydrocarbons of cigarette smoke, produced n Imv incidence of benign or doubtfully malignant Sq\, similar to tho.se induced by the known carcinogen, benz(a)pvrene. The object of the fitat experiment

968 P. 1. BIMONd, P. N. LEE AND F. 7. C. ROE described in the present report Svas to materials and subsequent ditribution of the .~ determine whether rats will tolerate more water-insoluble residue (Fraction C) between frequent intratraoheal instillations of 90% v)v aqueous methano[ ond cpelohesane SR'S or smoke-coriden'sate fractions and, (Fraction G). The oaffeine.complezing- if so, whether such treatment leads to the material (Praction (B. + P)G) was removed . early development of SqN, or a high by extracting fraction G dissolved in egclo- I hexane with a solution of caffeine in aqueous incidenre of SqN, avd/or to the develop- 91, formio aeid to leave a residual fraction ment of SqN of undoubted malignancy. The object of the second experiment was Fractions R{QGJ and L(QG) -Preparod by ' to compaxo the effects on rat lung of 2 distribution of Fraction Q(G) between cvclo- fractlons of tobacco-smoke condensate, haxene and dimetbylsniphozido (DJSfiO). ' ?r one of which (Fraction (R -; P)C#) is far F.actionHC(Qp).-Prepsredbyadsorption - j more tumoriganic for mouse akin than the of Fraction Q(G) on an alumina column, t f lL l - lie' et o owed b Ijpnsith pe ro enm other (Fraction V(QG))) nn . - naene. . Frudivas$(QC)andiFf(Q6):-Preparedby p e 1969; Lee et . ). cngms o al 19T fractions used in these espetlmenfs were~. agycegates of browmprooment-Ieden a(vealar ~ e f Fractdora(~,(R}~OaadQ(C).-Pcapared maerophages (GB,11); (ii) foet of eu0oidaV "S b I f 0 ~ t 1 ble columnar rnota lasia of alveolat e ithelium 4 ua ers u i ~ SW xeferredtoas staLewholesmokecondensate taken thmugh macrosmptcallp vustbe O (S{TS). . tumoura, nnless these were adequately The fractionation scheme has been de. represented in etandard sections. Sectiona O scribed in detail (Whitehead and Rothwell, were stained with haemarosvlin and eosin ~, ect of4 kinds of lesion: (i W) and scared in res f tb Th -29`C until userl. 9uch a condeu=_ate was 1e.vions at necropsp. Additional sections a'ere ~ - nctTERIeL9 n.4n StETSODS .. .-- - -shsking a solution of Q(G) in benzene nith Enta.-Non-inbred Wistar speeific-pstho- de-act(vated silica ge1. F7~actien N(QG) was gen-free (SPPJ rata obtained from Olac recoFered from the benzene and Froction (Southern) Ltd, of Blackthorn, near Bicester, M(QGl from the silicx by elutien mrh Osan, were used for the 2 main esperunents, methanol. - -- Non-inbred R%istar SPF rats bred in eur own CFxmricals.-Dotriacontane was obtained laboratories (strain TRCL) were used fer from %achd.ight Labs. Ltd, and hesadecane prakminary short-term toxicity trial9. from BRD Ltd. Both chemicals were re- In each experiment. female rats aged , 12 crystallized before use, weeks were allocated by a nonselective InfusineasdesoribedinDavisetaZ.(197aa) process to the various treatment groups. was used as the vehirle in most esperiments. Animals were housed in groups of 5 in ao)id- In the second main eaperiment phosphate. floor polypropylene cages with sawdust buffered saline (pH7N) containing 2°-a hedding, in natural daylight at 22 i 3°C. gelatine was used as one of the vehicles. This They were fed Ozoid 41B laboratory animal vehieie is referred ta as in the diet and water ad Nbittcrn. Tables. Preparat2n of condeneate and frastioras; A Treafneetta arzd ahaerautfon.-The method eingle batch of plain cigarettes (TRC code for intratraeheal insti)lation is described in TbiJ manufactured from a composite blend Davis et ad. (1975n). AnimaLv were inspected of 9uecuted tobacco was used as the source daily for state of general health, ehnicallr for all condensate. examined at the times of treatment, an3 Smeke candensate was prepared by smok- weighed wcekly. The techniques for post iug cigarettes in the automatic emoking mortem examination and microscopie esem- machine described by Day (1967). The ination of tissues are described ln Davis eE at. etandard smoking parameters used meret puff (1B75a). volume, 36 ml; puff duration, 2 s; puff Xislogatholtg:cal emlaadapn -All rsts that frequency, 1 min; bntt length, 20 mm. Smoke died or were killed during or at the end o4the was collected in a g)aas trap cooled by exf.eriments wers ezammed 1wst mortem. immeroion in acetone znd crushed solid CQ Four standard 6ym sections mere prepared (Davies and Dav 1969). It was stored at of the )ungs of rats shmring no localized p p y remo ~a o e c tom

RdT LL:Y0. E%POSED TO TCBACCO-BS[OliE FRACTIOFS 967, %Cflt); (iu) foci of squamous metaplesia of Analysis of variance, this time of untrane- alveolar epithelium (Sqll) and iv) squsmoua formed data, was al-=o ustut to compare neoplesms (SqPi). HEghtly different scoring srstems were used in the 3.esperiments, and details are given separately in the text describingeach esperiment.Histopathological emlustSons nere "blind", in the sense that the pathologist (FJCR) was unaware of the rreatmesst mta had received br of the meek of the experiment in which they died or were i`'.hd. .~mti.cicaI ecoluarimt-$Vonificance tests vere CaYried out to asse_ca betmeen-n oup . differences of 3 different parameterz: (i) pro- plr.ion af rats «ith a given lesion; (ii) mean numher of lesicns per rat and (ui) mean grade - oi:e,ianperrat.Ifrheresultubeingconsidered were ba9ed on rats dving in s aingle'cnterval, eb-eqnared analysis was used for the flrst parameieis aud one-way analy.is of variance for the second or thhd parameters. If a smnlraneous assesment of resultasrom sat® d,tiax in more than one interval n'aa being made, Peto's method for inciden:al tumcurs 11q4) was used for the first parameter, in order to corceot for passible differencas in survival betueen the groups. The other 2 parameters were analysed by unealanced tao.vac analysis ofvartanee, to give betceen- grvup differences "adjustedfortimecfdeath". A11 enalcres of var;ance were car.ied out after a snitable variancestaUilixing transformation of the date. For the second parametets, which are counta, the square•root trans- forma:ion y=yIz n'as used, while for the third parameters, which ere scores out of a masimum possible (i1) the trausformation g=loge[(x-F 1Ji(3f1- I-zJJ was used. groups ur terms of body a'eigftt and body- rveight gain. . . ERPEP-I~tE_\T~r. DESIG\ A?.-D R85ULT3 Exyeriraent I Praliminaq, toxicity studies showed that rats would toterate tdce•weekly intratacheal instHlatious of 12 mg of Fraction (R!,-P)G in infusine, but not similar exposure to whole condeneate (StVS) or Fraction G'.'The prupose of Exp. ~ I was to see whether even higher doses of -. (R } P)G wmild be tolernted and if so •• what effeets such treatment would have. - Seventy rate were allocated non-selae-=' tively to 3 treatment groups, each of 18- animals, and 3 controt groups, each of 4-e animais. Rats in the first 3 groups received trtice-meaklyinstillation of 22 mg, 24 ntg or 48 m, of Fraction (R + P)G in irrfusine (total volume per instfllation= 02 m1). One control o onp (6 rats) was similarly erposed to infusine only; another control group (6 rats) was anaesthetized twice-Reekly but received no treetment; and the remaining control group (4 rats) received neither anaestheticc nor tteatment. Treatment continued for 34 weeks, after which a proportion of the smvivors wera killed. The remaining rats .vere observed until the termination of the experiment at 84 weeks. The design of the study is Tesi-E L-Smrcisal of P,¢ts G4ven up to 69 Tmice-weeAiy Doses of Fraction (R =P)G Prepfred from QiauretteTRO%57 by InEraEr¢cRealInstYdln6ion IC~tt at Nae eunivan Icltled st tin. svrvivo+s st Ne¢k 34 vee;u foc at WeoY. 94 wseka IDr Treelment nietae tbolagieal ~ hiatapaoholegicel ryrcup .Yekl" 0, 10 20 34ea madan 34+ 60 e.eemmabon I 12n.-~&+P)G .19, 17 14 13 9 4 3 1 minfu:ine 2 2.1m, fRrP1G is is is 14 m infusme 3 28mf(R+ P)G 13 1g 8 5 v mfu 4 Iniu>ine anlv 6 6 9 8 in¢eathetic oNT u 5 6 4 6 None 4 4 4 4 R0 68 53 27 f l4;

P. J. SI5t0\S, P. T. LEE AFD F. J. C. ROE Tasns II.-Jlersn Inibial Bcdy IVeight oind Body-weight Ciadn \o. rate with Squamaus Neoplasms (SqN) in Rata Ezamieeed at Pas6 Jfo+tem.t Trestmcnt . Gmnp (2x weokw) 1 1-°mg(A}P)G ininasino 24mutt~ssineP)G 48mF (R}P)G tnmfusino 4 Infnsinocnly u Anaesthetinonly 6 Ncae Initisl . Weight gsin by meek . . tictlyweight 4. . .B 16 .:32 ~ ]feen 229-9 142 24-5 . 439 55-4 36 1-B 2-0 2.8 43 9 ean 22z,8 Q00 22,~`2 43a 1 54.5 JIesn 2231 42 193, 404 444 7-4 19 29 20 39 31eav 2395 110 282 51-5 056 64 2.9 611 ' ' 5.4 3.9 ]fesn 2#s - 3 100 35.0 , al-0 . 48-5 o. 4-3 2-4 5-4 - 3-2 1 '3B Mem 2292 4-8 28.2 360 f~g~ 0 ~ ~ '!•6 i TasLS III -.Vucroecopic¢ily yasible Znzng Tecn-wuro (T) 2nd Blierosrqpically Obsenad Exemined et ,--~--~ - ~ q-~~ Mean co. Postmcrtemj E 0 - E --- Bq\frsE- ~ 17 3 463 7 9.62 04] 15 5 445 lt 9~59 1•E0 15 4 3~02 7 580 2 07 8 0 - 0 - 0 • 0=S:mber obse-ed: E_\umber expeted if the effxta of the 3 dosa IoveLs hxd been the same. nfter apom - re11'Jtere es (see temt'). } dlt mecrtecop :allv observed tumous vvere found onm(eroscepic esamination to be squamous ncoplasm8. However severnl smeh squamous neeplaems not observed at necropsy wete d'ucovered on microscopie exeQ.v:rion of tbe lur.gs. $ a tmsred ste in groua 1-3 wern too sutolysed far aese.ment sott anly e eoatrol rcte (grcup 4-6) were e_ss;rL Otherwise ehe TnbL refers to nll rats in the experiment, irrespective of when they died or wete Silled- evident from Table I, wlvch also gives scopically at necropsy and GBbi, S0\I and survival data. SqN scores derived fiom a"blind" Treatment ~ Groop (2x weekly) I 1?mg(R}P)Gh,inNSine 2 24mg(R}P)Giaiafhslns 3 43mg(R-PJGinlnfusine 4, a, k 0 Combincd control groupz tracheal instillations of 12 mg or 24 mg, but not 48 mg, (R i P)G were well tolerated in terms of survival. As shown in Table II, the mean body weight of rats in all the groups increased during the study. ' Analy;is of variance showed that, though there were marked differences between the 8 groups in weight gain in the first 4 weeks (P'< 0001) due mainly to amarked dese-response trend, these were not sub- seqnently statistically significant. How- ever the (R = P)G-treated groups had consistently lower weight gains than the infilsine-only or anaest}mtic-only o oups. The behaviour of the untreated group wea surprisingly erratic. ' Tables III and IV summarize the fin& ings in terms of tumours observed macro- Table I shows that twice-weekly intra- histopathological evaluation off sections from 53 of the 70 rats. The smring system used in this study was as follows. The available sections of the lungs of each rat were given an arbitrary GBM score of 0-4 according to an overall assessment of the frequency of aggregates of pigment-laden macrophages. Arbitrary CC3I scores on a 0-4 scale and SqM on a 0-3 scalo were arrived at in a similar way. The squamous neoplasms encountered consisted of large masses of keratin surrounded by a narrow zone of squamous epithelium. Apart from their large size and escessive keratin component, the lesions lrere difficult to distinguish from squamous metaplasia. Undoubtedly benign lesions were graded "4" and lesions which showed posslble evidence of extension into surrounding

a I R4T LC*\GS EXPOSED TO TOBACCO-SpW%E FRACTIONS 969 Tssna IPc blieroseopic Findings im. Rats Killed at 34 Weeks .. . - -- nIeanpvloot T st e Group 1 - re men (2 x .reek:y) 12mg(RiY)Ginlatusino \u. rece . 9 OB]t ' 3-03 MI ' 303 6qli 1•58 24mg(R}P)Giniufuatca 9 3.36 3.53 . 226 3 aEmg(R}P)Gmidasine - 6 2•50 3-2] 2-49 A,sk6 Combinedcontmt8~uup 4 0 _ 030 0 number of SqN per rat, the mecn number twice to 3 x weekly }sitltout an adverse in Group 1 being very significantly (P < effect on survival of rats given 12 or 24 mg 0,001) less than that in the other 2 groups. of (R + P)G. The results showed that As all 3 lesions studied in Table IV are mortality .vith 3x weekly treatment was strongly time-related, the Table simplifies excessive and it was, therefore, decided to the comparison of the groups by consider- continue with 2 x weekly application in ing onlc those rats 611efl at Week 34. Experiment II. ' Though (P. } P)G itself had a marked The design of Experiment II, involving effect on all 3 types of lesion, there was a comparison of the effects of Fractions no clear evidence of a dose relationship in (R + P)G and N(QG) with I or BS!G as the effects of treatment on mean GBlI vehicle is evident from Table V.The and C'CSI scores. There was, however, a Table shows that 309.rats were allocated very highly siunifleant dose-related effect non-selectively to 15 0 oupe, of which 46 on mean SqJ7 srore (P <&01). died before ~Veek 1Q, 47 from 9 different• Full examination of other oigans at groups were killed at 10 weeks, 10 died necropsy revealed no evidencd of ineta- between 10 and 20 .ceeks, 42 from 7 stasis from anc of the lung tumours. - different groups were killed at 20 weeks, 7 died between 20 and 40 weeks, and the Esperiment II remaining 157 were ralled at 40 weeks. The purpose of this esperiment was to From these figures it can be seen that compare the response of the rat lung to a Fraction S(QG) at the dose levels 24 mg fraction of smoke condensate known to be or 12 mg twice weekly in I caused more tumorigenic for mouse-skin ((R + P)G) premature deaths than any other treat- with that to one known to have little or ment, and that Fraction (R + P)G at the no tumorigenic effect. At the same time it dose level of 24 mg twice weekly in BSf G Iung tissue were e aded "8". These grades were a direct extension of the grades 0-3 used for Sq3I lesions. ' The re?ults in Table III show clearly that tumours and SqS were related to the (R = P)G treatment, none being seen in any of the 6 control rats examined. Af'ter taking between-gronp survival differences into account, no statistically si_anificant side response was found in the numbers with either'tumour or SqN, though there was some indication of a positive trend, as indicated by the excess in the highc-st dose levels in che observed numbers (0) over those espected (13) under the assumption that dose does not influence incidence. There was, however, clear evidence of a significant dose-related effect on the mean was proposed to compare the influence on response of 2 different vehicles, infusine (I) and buffered saline;gelatine (BSjG). . Preliminary toxicity studies were under- ' taken to identify a suitable fraction of ~smoke condensate which was not tumori- 'genic for mouse-skin and which would be tolerated by rats exposed to it by intra- ~tracheal inseillation. Short-term studies ' inclicated that Fractions S(QG), L(QG), M(QG) and HC(QG), and also dotriacon- ._ tane, hemadecane and liquid paraffin (BPC), were too tosic, but that Fraction N(QG) might be suffmiently well-tolerated for the purposes of the proposed study. A_preliminary short-term study was also carried out to see whether the fre- quency of doaing could be increased from 9 !

970 P. J. SMIDNS, P. N. LEE AIdD F. J. C- ROE 29 23 NIQa)/n 20 infusino ~ 10 12 36 IR } PIGin 24 10 1G 2 O) SO Ames:hettoCbntcoL 22 309 Died -. Died No. betwecn befc:e 1d11eQ ao 10 md 20 SPcek 10 10 wecks - weeks 6 0 0 0 6 - 4 0 4 0 fi 3 3 Dfert ~ -~ - - No. bebwcKn. No.. - kiOeduE 40and40 AinedaE 40 weeks .veeks 40 rv ]s - 8 0 17 6 0 11 0 0 10 0 0 11 6 , .1 14 0 0 0. 2. 0 0 . 1 6 0 1 , 0 6 p- 12 .... - ~ 0.~ _. 9 0, 0 0 8 0 6 0' 46 4] 10 d2 ] 157 caused more deaths than when I was the clusters, 2= moderately frequent clus-. vehicle (10 as 5). - ters, 3=freqnent clusters, 4= vetc ~ There were no etriking differences numerous clustels, 5=extensive masses between groups in rate of bedy-weight of pigment-laden macrophages. Scores for gain during the experiment, except that CC31 were based on the total numbers of . ~ the anaesthetic control group (Group 15) lesions observed in complete scan of the put on more weight (80 g) between the 3th left and right lung section up to a maii- and 40th weeL-s of the experiment than any mum of 25 lesions per lung. Large foci of other group (range 32-50 g). CC1S were counted 2X, 3X, 4X etc., . Groups were compared in respect of (i) according to the number of lOw-power . incidence of macroscopically visible lung tumouia, (ii) severity of chronic respiratory diseases (CP.D), (iii) incidence of aggregates of GB~I, (iv) incidence of CC1I lesions, (v) incidence of SqT4 lesions and (vi) micro- scopie,illc observed SqN. Standard sections of the left and right lungs and post-caval lobe (ti.e. 3 sections in all) were evaluated "blind" in respect of parameters (ii)-(vi). CRD was assessed on the scale 0-4 (0 = . none, 1= minimal, 2= slight, 3= of moderate severity, 4 = severe.) Various kinds of spontaneous lung disease were taken into account in arriving at a score for eachanimal; e.g. aggregates of lym- phocytes and plasma cells around main airways and blood vessels, interstitial pneumonitis, focal granulomatous lesions, focal consolidation and bronchopneu- monia. The score reflected both the severity and extent of disease. Scores for aggregates of GBM were allocated as followa: 0 = none, 1= occasional TnsLE V. Experfinent 11: De,sign and SzrnnvnZ " ' TR94ment- -ts82xweeklvl N~.of G,vup fo~ up to 40 wcek» te fields upon which they encroached. Scorea for Sq3t were arrived at in a dLrectic comparable way to that used for CCilI: All the squamous neoplasms (8qS) seen were undoubtedly benign and were graded "4°.The mean diameters'of SqN were _ recorded. Only 4 rats had lung tumouls that were visible at necropsy. All were among groups - treated with (R } P)G in infusine that survived for the full 40 weeks of the experiment. Further Iung tumours lying deep in the lung tissue were seen when the ' lungs w'ere trimmed after fixation. In - addition, several small SqN were discov- ered for the first time during the micro- . scopic examination of the lungs. No animal that died or was lulled before the 40th week of the experiment was found to have an Sqr, though the 7 dying between 20 and 40 weeks were not examined micrascopie- ally. Table VI shows the incidence of Sq\ among the 157 rats that were killed at 40

A~T LUNGS EKPOSED TO TOB.4CC0-3ptOEE PIISCTiONS 971 'TaECS NZ-Ivacidenes of Lualg 4'antovas' ¢m.ong 167 Rats that were Killed at 40 Wee/rsj' No.vith 6lzesoflargo,t .. oquumeua 'Totelno. equsn.ousneonlewa 14eatmenE No. of rat neoplavms- . of squamous In eech rat Group (mg2xnecdy) killed oflun3 ceop!.,sms (mrnndlsmetuein v1' 241 2 16J (R - P)G in ; 3 wfusmo 0 6-s N(QG) au dose leaol3 m;nsslne I1 10 1l 14 10 2~ 13 ' 18 ~R+P1Gin -. :.1- 1- BS0 -4 15 AmestheticCaatroh t0 • Hetign squsmau ncopleass. 6 17 7,44,3,1,1,1 1 4 2 4 3.2 1 2 2 0 0 0 0 0 0 2 - 2 D U e 0 0 0 + No auimN that diad orwar kWed bHOre 40 wee:u had a macrowopfcally visible mmour, or rvas found > to have Sq] o oxoplc exomivsoien:8aweve:, the 7 aalmnl that dieu beheneu 20 and 40 weeke . (,o¢ Tsble V) were ot exammed mierexop eally. weeks. -4.Itegether, 12 rats treated svith Fraction (R + P)G developed a total of 29 neoplesms, whereas none of the rat= given Fraction S(QG) did so. Ten of the tumour-bearing rats were treated with Fraction (R + P)G in infusine, which was much more effective in giving rie to neoplasms than the same fraction in BS/G. 3Iean CRD scores tended to be slightly lower in anaesthetic control rats (Group 15) than in other groups, but the severity of CRD in the other groups was not associated with kind of treatment, dose, vehicle cr time. ' Zero or verc low mean GBM scores were a feattre of the anaesthetic and . vehicle-only control groups (Groups 5, 1-4 and 15). Esuoaure to Fraetion N(QG) in I was associated with higher GBb1 scores than eroesure to (R+ P)G in either I or BSG. (R+P)G in I and (B=P)G in )#SjG gave similar GB]I scores. In the case of each kind of treat- ment, GBjI scores were dose and time related. Table 4Ti summarizes the findings in respect of CC'-)I scores for animals killed at 10, 20 and 40 weeks. Treatment with Fraction (R } P)G in either vehicle or . with Fraction N(QG) increased CCM score in a dose-related and time-related fasluon. The response to Frnetion (R + P)G in I was signiHcantly higher than that to the same fraction in BSG (P < 001) and also significantly highor than that to N(QG) in the same vehicle (P < 0-001). Table VIII summarizes the findings in respect of Sq}I scores for rats killed at 10, 20 or 40 weeks. They are quite clear cut: ver-v low scores were a feature of the anaesthetie controls, rats given either vehicle only or Fraction N(QG) at any dose level. Bc contrast, exposure to Frac- tion (R -k- P)G was associatd with an increase in Sq}I score which was markedly time and dose related. 2.nalvsis of variance showed that response to (R + P)G in I was significantly (P = 0.01) greater t11an that to (R }P)G in BSJG. With either . vehicle; an effect of 24 mg (R + P)G twice weekly on Sq-NI score was already evident at 10 wecks. Where post mortem change did not obscure the picture, COM and 5q3I scores for animals dying before 10 weeks were generally lower than those for animals killed at 10 weeks, and scores for animals dying between 10 and 20 weeks were intermediate between those for animals killed at 10 weeks and 20 weeks. None of the 7 animals that died between 20 and 40 weeks was examined microscopically.

972 P. J. SIMONS, P. N. LEE .L\-D F. J. C. ROE TABLE VII. Effect of Treatmext on Incidence of 6Sabeidal/Colatmemr Dlpdap7asia of ' Alveolar Epithelium (C6'31 Leaions). blean Scores for Left and Right Lungs Com- T entment Group (mg2x rveeWy) 1 24 2 le fi(R+_P)Gin 4 3 ~nfusame Taar.z 4TII.-Effect of Treatment an Irwidence of Squanlmia hfel¢P7asia of Alveolar EpiEheSiunc (Sq]~I). D1ean Total Scores for Left and Rig7rt Lungs a>u,l Poat-cavai LNre CGTI s<ore st Gmcx ehown (No: of xats) 24' 125 (6) 26.5(6) 46-8 (17) 12 R + P)G In 5~3, (6) 19•2 (9) 43.2 (11) ( N~usme 3 224 (10) 24 7 (11) 0 l~0(6 F3(8) i-2 (I4) 241 2.5 (4) - 34 6 (5) 12 S(QG) in 80(4) - 6 infLSine - -- ' - fi•6 (6) 3 . 34- 0(ltl'-- 24 3-p(6) ' [6~0(6) - 43.6 (18). 6IB;GP)Gfn ( 4-3 (6) -. 4~3(6) - _ 37~4(12) "25.0 (0) 0 JJ 3 3-7 (3) 5.70 ) 18-2 (III ~ 3•9 (1>) Anms[heElc Control 1'5 (6) Y-6(0) -_ 2.0 (10) IP(qG) in infuslue 11 12 )R=P)6in 12 6 Rs/Q 13 3 14 UJ ib Anarsthetie orJy 0.0 (3) 0 0 (6) 20.weeks 40 weels 32(6) 20~5(19) U~3(6) 130(11) . 6O (10) 8-4 (11) 0.0 (6) 0-0 (14) - 40 (5) .0.3 (6) 0~1 (Il) 3-3 (6) 15~2 /13) 0-0 ()) 6-1 (12) 2-g (0) o U) 0~p(6) . 0'4(19) . 1 0•0 (6) " 0•2 (10) A"b&nd" evaluation of standard sec- tions of the larynz and trachea revealed no consistent differenees between the groups in terms of thickening of the epithelium or inflammatory infiltration of subepithelial layers. No examples of _ squamous Cnetaplaela of the laryngeal or tracheal epithelium were encountered, Macroscopic examination revealed no treatment-related differences in incidences of lesions outside the respiratory tract. nlscCSSrox . The results of the first experiment indicated that Fraction (R + P)G, which - bined for Rals Killed at 10, 20 and 40 Weeks for P.ats Killed at 10, 20 and 40 Neeks . " ' Bq)I zeore nt times sho,m (No. of mta) forms only 2-0% of whole-smoke conden- sate sate and is' almoet as tumorigenie for mouso skin as the much larger mass of whole-smoke condensate from which it-is - derived (Lee et aZ., 1977) when instilled twice-weekly into the lungs of rats, in- ereased the incidence of squamous meta- '- plasia of alveolar epithelium (Sq)[) and of squamous neoplasms (SqN), both.in- creases being dose-related, though in the case of SqN this was not atatistically sigmiGcant. Instillation of Fraction (R = - P)G also increased the incidence of 2 other lesions: aggregates of alveolar macrophages containlug golden-brown

i R AT LUSGS EXPOSED TO TOBACCO-SJIOliE FRACTIOSB 973 ' vL•ment (GBM) and foci of cuboidalJ . columnar metaplasia of alveolar epithelium (CC'M). This increase was time-related but apparently irrespective of dose, although the apparent lack of dose response may have been due to the response being masmal at the lowest dose studied. Some . ' f the SqN observed were doubtfully mah?nant, but no unequivocally malignant lung tumour was seen. However, only 5 E treated rats survived for longer than 60 _ - The second main experiment showed Devrse, R. F. & Dex, T. D. (19u9). A Study of the Compavstive Camnag~eNclt'a bf Clgamtte d ''3e.-5ereas b--fractina°vf=cigaret;a--e)~~nnt ~aa,ien te•. , ch.., an~e_?t_ r ~~ ' wee$F, InYa?atiov. J. nNn. Cnncer Inat, 3,1141. smoke condensate (Fraction (R T P)G) Cnnov,23-3s9 ' vhich is relatively atrongl_v tumorigenic for mouse s',dn gives rise to SaM and Sq'I when instilled into the Iungs of rats, another fraction (Fraction \(QG)`) which is more or less without tumorigenicitc for mouse skin, was almost entirely unaole to induce Sq\I or SqN when instilled into the Iungs of rats. By contrast, both fractions were active in causing GBM and CC\I, suggesting that these lesions are not speai5c to the tumour process. An important aspect of the 6ndin„gs is that the effect of Fraction (R -I- P)G on the incidence of Sq>I wes already evident - : at I O weeks. This may mean that the intra- tracheal instillation technique trtn SqM . a:: the end point could be Lsed as a shorb - term test for predicting tllmorigenicity of unequivocally malignant lung tumour was seen. Further studies will be needed to see how useftd'a predictor of tumorigenicit,v- the rat lung fnatil!ation assay is, and whether truly malignant lung neoplasms can be induced by this method. , REFERP.SCES Beasssm, P., Ho>rarr.ocn, F. 2 Rsss ienu, A. B. (1954) Semlu Diderences in Ibo Rcsponm of Inbred 6yeian Hamsters to Cigaeatte Smake mxE., J. K., Gns, 3f. E., Lcc, Devvs, B. P... R P. N., Bcrru••conre, A. D. 5 Roc, F. J. C. (19]Sn) Rescama of Rat Lung to Inhelad Tobacco- smake with o wirtaut Prior P•.rposum to Y,4- benzPymne (BP) Gh'en by Inr.otrachenl Inxtilla- tian. Bn J. Cm¢a, 31, .499. D3as. B. P.., yt ~re~nn, J. R:., Grr..., lL E., Lcc, P. N., Bc"rrssmoarz, A. D. & Roe, P. J. C. (19956) Responee of Rat Lung to 3,4ben.ryProno Admtnistered by Inoca:readx) InatilloHon in IrSu,ine .rith or nithaut Carbon Black. Brr. J. Canrz., 31, 443. De.v. B, R., wmrEVan, J. K., Gns, 7t. E., LES, P. N., Bc-rrc•roara, A, D. m Roe, F. J. C. (1975c) Reponx of Ret Lune to Tobacco Smoke Cmdensate on Fractions Der ived from ioAdminie- toml Repestxily by Intratrecheel Instillation. Br. J. Cmsea 31, 453. D.x, T. D. (196]) Cardnogenin Action of Cigawtte- moko Cndeu.rete on -llmo_n Skin. An Attempt at a Quantltarive Study. Br. J. Cm,«r, 21, 56. Do.-rE'wu' t{-., Caca,warz. H.S., Hesss, LerrEN2, Ii.. RECSg3, G. & Bc.mvEmEn. B. (1993) Inveati_estione on shc ERCOts of Crsonie Cigsrettesmoke Ichalacion in S,afian Golden smoke compocents. Hamtera. J. nofn. Canrer L>st, 61, 1781. Lea, P. !:., Roxa•.v.c, K. & Nmzanvwn, J. K. In the oasa of mouse skin, there is 11977) Fraadonation of xoase Skin Cnra(noven, evidence that the tumorigenic aetivity of ' Cgaretrn.socolpo Cendeaare. Br. J. Cnncc+, Fractions (R = P)G from di3erent smdoe P~ ' R. (lary) cn:detmes on Ihn .vnayats of condensates parallel the activities of the - T,tmonr Raleu andpeath R.,tes in Ezperunental ' smoke condensates from which they are de.riced (Lee e€ al., 1977). Lf this is as true for rat lung as for mouse skin, tests on rat lung of Fraction (R = P)G from different condensates might be useful for comparing the overall tumorigenicity of different condensates. In neither of the 2 studies reported was it possible to keep more than a few animals for long-term observation. It is perhaps not surprising, therefore, that no Avmsls. Br. J. Cnneen 29, 101. Prr.ev, L. ]. (1963) Inductron of L•mg Cancor In Rats by Intrsrrnchent fnsutTstioa of Cancerogenio HyrLocarbona..{rm On. int Cnncer,19, Eg6. SCxi,E[aER, H., ScrrESa_nr, P. L.lSenTrs, D. H. (1972) Rapid 4]ecelopmenv ofBronehioloalvootar EqumconaCeuTUmeure in Rets aFCerIntrat^aehe.el Injection af s~me:hylcholanlhreoe. J. nntn. Canttr Li.vt., 49, 551. S..rn, L. >t. (19c3) Exprlroentel Cancer of tho Lung. J. rmen. Cnnev Inse., 28, I506. wn[rescnn, J. K. & RorsrE~1., K. (1999) The Mouee 5kin Carclnegenicit5 af Cignrette Smoke Condenraee: Fractlonated b- Solvent Pa:cirlaa 34cthods. Br. J. Cannr, 23, 840.