Document V3X1NaXJak0wr9M0npXyon9Eo

An 2 ierdc 1ogj cal Stall Lung Cancer in Asbestos Mine L * ,<2'vXUkl^ und T. David Troan, AKA Arcliives, J- io 1956, Yol. 17, do. 6 31. flew by Thosiaa ?. l-'uncuoo, M.D. Chief, Divisic i of Industrial Hygiene ffovtrser 13, I960 PLAINTIFF'S EXHIBIT Af firmtivc Association: Reviews Negative i rrrsCL' JCTCHY MATERIAL 1 Early literature rsfers to cancer cases which occurred giruitaneoualy with csbe3to3i9. In other words, reviewed tfros cases of n3beatooi~. to find vfcat percent of cancer (lung) occurred in thiG group. 2. Example: Merevetber (1954) l6.C$ Gloyne (1951) l4.1$ 1 Huepe(1952) status 7"17^ of eases of esbeatosis, lifter 1^ to 20 yeuro latent period, develop lung cancer* 2. IfOidaraim and Sor-re - ordinal e^periaento- produced lung cancer in nice which were exposed to asbestos dust. 1. Snuoe (1939), found no cases of lung cancar among 630 cases of asbestosis he examined* 2. Ijolleb (1942), found the number of cases of asbestos is with lung cancer too sirs 11 for statistical evaluation* 3 We.cc 11 ns (19^7), in 126 x-rayed cases of asbestos is among 476, found no lung cancer associated in the gi-oup* Ibncuno: Lung cancer occurs In asbestos workers with and without aabestoais* All studies were related to cases of lung cancer associated with aaLeatoais and whan thi3 sirrultan ecus requirement occurs, it in usually ti t-miller percentage. The Wegeliua study cakes the assurgotioa i Lung Cancer in Asbestos Miners Toll: that for lung cancer 1 occur, this would he only vhen asbestoeis is present y x-ray evidence. Thi3 cay or nay not be true. lung cancer an occur in these vorisers as a cause of death, without asbeato.-is as a secondary cause. By -considering, only-the-siwuItaneous < <;curr^ce^fLJfungjjc^ncex^and ._asbgstoglsA ; the-investigator ia au -ocatical I^excluding, a .large pQt.er*tial of-deaths-'With no diagnosis of usbestosis, even though -exposure. and-duratlon- of' exposure haVe"'USen' 'established- Doll'^1955) studied records of men vbo worked for at least 20 years in exposed situations] found that in workers eX%r^_20- ' Kaneuso: Braun: r' Kancuao: n j. (Found 6l cases of lung cancer-studied 113 cen who had worked for 20 years or core.) .. In Doll's study, he excluded those who died from other causes, after shorter employment. It is evident that lung cancer cases could have occurred anong employees with le38 than 20 years employmer exposure, depending, of course, on what 13 found to be the pi nlimn number of years exposure required and the latent period. Cite3 selective factor of Dr. Gloyne's series of l^.l of autopsy cases saying these cane as part of unusual group of lung cases to autopsy. Also cites percent of lung cancer at Iondoa Cheat Hospital was 21.37$* Tills could very well relate to all the lung cancer cases which come to the hospital, like coming to Dr. Cchsner's Clinic, so this part of the criticism i6 valid. However, association of asbestos la with lung cancer is not a bias that exists by virtue of the Clinic-so observation of Gloyne is still an unusual and valid suggestive lead. 3, Dun..-' Cancer in. Asbestos Miners Braun: J.'ancr.LSO: Refers to Klotz ' x> found only 4 cu3ea of lung cancer la 473 cases of aabestof' Is. Initial observation in regard to the llaitatioaa of thla interpretation atill 'olds. Other cases could have occurred but not simultaneous 1/. yJTha simultaneous requirement places a restrictive bias in c :tual situation.* DBMIOLCGICAL APPP-CACTI - by Braun :>C'.u-ca of Data; 1. Braun used through, thc~ ia Quebec. .2 Established cohort, with/total exposure of 5 or jrorc year who vers on ecgloycent. rolls in 1950. Tie office force a: 2'ancuso: non-exposed vs re not included. ^for late^ peri^ja^uiared for lmja-'cancaar If the cutoff date in 1955, then tha cohort Intent period is only being tested against a period of 11 years, 1945-1955; question whether there is enough of this population in the cohort with exposure of 10 yean or core. For example, if they (the aunber of workers with 10 or more years exposure) represented only 20^ of the cohort, the expectancy of lung cancer would definitely be diluted by the other o. The cohort followed through 1955>,Jhether alive or dead. It he states-ell but a email number of these vero accounted for os either living or dead." Thin was 133 out of 6,091- If the I33 were found anrl only 1 lung cancer recorded had occurred, the entire conclusion of the study would have been different. Certainly a vary slight thread to draw 3uch definite conclusions. F 1 4. Long Cancer In Aat'-sron Miners Braun: J-tmCUSOI Kancuso: I-raun' rlancuso: 1. Through Office of /ital Statistic? tabulated lung cancer rates by oeic for 2ars 1950-1955/ inclusive* 2. Croup policy cove igo of one company studied for all cancers of lung and Braun states: "Very fev not covered by this policy- and those who leavs industry and are not covered, are a snail nu Bov irony? This nay offoct results. Braun states "fev cases in which lung cancer 11 strongly suspected but not proved os the cause of death vere considered separately*" Bentos from lung cancer vere daterained from clinical records in. the medical service of industry and checked by means of death certificates and the insurance company records. Bowver, he refers to only the group insurance plan coverage of industry. Cohort-5,953, placed in three (3) categories according to increasing degrees of exposure based on weighted average of years, & Table l~6qjh.of- cohort imputation weighted tev)tbd^ not-fthoMuuausl inc IdeaMEBffiPtT.xmg ^aSeer-"ficStS3t>''ia thqLOhdg?&e/.groups , therefore, rate? of this group as a >total for.C&dC*r^Vtllbelovn^^ a^^dilwtion 5factor of considerable;proportion. Eovever, the age specific rates for 20-44, if higher than the control, would bo significant. Again, however, one Ince this is probably the latent period, otherwise the rate becomes markedly diluted. The number of this group by the very nature of the ages, 20-44, would indicate thal the number of workers of 20 or more years would be very nmll, too Bmll to use* statistically since the cohort was only followed fo: 11 years. Bov may vere there with 11 or more years exposure in the age group of 20-44 comprising 6&f> of the total cohort? 5 lung Cancer in Asbestos Minora Mnucuso: rtincuso: icuao: {Aoncuso: Table 2. The r.uri' and percent distribution by length of erploynent shoved On3y 30$ vith 20 years hr core exposure; 7C$ hi less than 20 years. If it takes 20 or nore years to develop lung then the dilution factor is 7^;jJ the lung cancer rates should be t on the 3^ Even should ve cooaider that 10 or nore years arc req then there is still 3# of the cohort vhioh would not bo the population vben the rates are calculated. Materially, vould increase the rates. Notice, however, in addition thatthg-length < e;nplo>7aent in- not - given by age grout* Could be as high as 50^ in tlie age ^rcru? vith less than 10 years of e=ploynent. Don't know actual percent hurt represents a high dilutioa factor. Ihblo 3 Exposure categories. Again, in categories not related to the age groups like 20-44. Table k. Smoking Habits. ^ot relatod to age groupo. -Definition of SX&$rar--CigarsttS a day. This is not tha definition usually used which refers to a pack a day, etc* Table p, ;rurher of Cases. Defined proved cases of lung cancer as "diagnosis suggested by autopsy, surgical section. Microscopic examination." "Suspected casco-3, diagnosis in doubt but evidence points to cancer*** * Basic error hero. In comparing this type of ffpanradfapqay , vnlch the rates for lung cancer in the control are derived from luag cancorr as listed on the death certificates. There la a^-rrtfrtv*- u Actually, this would bo ^lijpoesible without reviewing all ths lung cancer case* of the count IVm'icso: y.oncuno j "ancuso: / / i Train: w"Establishing thei entire clinical recordo Secondly, tha goncrn.1 population nay not :v *3 o-aina opportunity of podi-cnl care. This coal wri berth ways, axel'- .ca other cases without such confirmation. Controls include both confirmed and nonconfirmed, usually nonconfixr* group in larger becau; a of medical or available nodical care. In discussion sta'-ee "If ho uses the 3 suspected cases. Table 12 would be otatlatica.il;.- oignlfleant. " I*e goes on to analyze the 3 questionable cases because this cokes the study so clcac to the significant level. Tobla U. Lung Cancer Deaths by Aga Groups. There vac only 1 case of the 9 proved cases in-the 20-hh ago grou Recall that this age group hod 66s of tha population. of ths cohort. The rata for 20-hb, per 100, COO jncre years of exposure, vas only h; and rate for bfj-jb vac l2_. (Cccjprised 17$ of cohort). Rates era based on very craa.ll numbers, one each in 2Q-bb and ^p-51** Cannot use total average rate of 25 because of mrkod dilutio factor of a3 20-hh iota cf b which comprised 66$ of population; czsst raake ago cpoclfic corptiriocna with the control. Table 9* Lung Cancer Deaths by Length of Employment. Sbovs'that- cny~l"piTT7*5a~e,S^ in^tha iealTTMrr'20-- yeJup^Length -of enploynerffc-*35~^^ co^cflS^dJ'l'0$'m6il~ tEsTCbbort. Tko rates under 31 years length of employment vsre ^ per 100, 000. This again dilutes total rate and lovere average rate to 25* Ilote that tha rates were high in 20-29 yr. length of employment Group which was 3b, and the 30-39 yro. length of employment vos 4 13/100,000 ron employment years. Discussion of 2'i0 nen over bo years of age, followed for 6 years, did not develop lung cancer. ; - 7 Broun: I Sane us o: Mancuso: Table 10 Iran.-; Cane; Deaths by Exposure Category . States rates do r. t increase by category of exposure. "Thu* tliia is proof that as' estes is not a carcinogenic agent." Categories,, hovev: r, are not related to age distribution of the population) for example, 66of population in the 20*44 age group. Do not know hov much e contribution is cade to these various categories, I, II and III. Thi3 cay be the reason vhy the rates did not rise with category increase froa I to III. 1/ Secondly, workers.^re so small,-^ caseB In category III; the .audition-of 1 or 2 cases rakes sizeable difference. Table 11. Categories I, II find III were comprised of 1,795 persons with employment only of 5~? years and comprised 30$ of cohort population; adding the 10-19 yrs. length of employment group gives 70$ of the population. However, of those groups at least 66^ must have been in age group 20-44; 1300 out of 1772 were in category III-from employment group less than 20 years;- l4X) out'of 2,1^0 were in' category II; 1300 out of 2,000 in Category I. It is obvious that these categories are biased by contributions of different age groups and higher numbers under 20 years in these categories. Therefore, one cannot use thin approach because the categories are weighted by a larger number in the younger age group; even in category III which would be less likely to have cases of lung "cancer occurring. Braun states "Average number of years of employment in each exposure category is almost identical.'1 He does not take into consideration distribution by age group. luar Cancer States net a elrH .a caaa of ling cancer occurred in 1,265 noa-s.vckera of both " >roved" and "suspected" cases* ' Tables 13,l^i15. Broun: " i'nncuao: and greater age (data does cot show this) cae would expect the non-sacking group to i hov a higher rate if lung cancer was due to asbestos." Actually hs does :>ot ohcnrln vfcat ages these coses occur in the rstackers *" Table 16, Cocparieon of Co*ix>rt Experience with Province of Quebec. Gives - Annual rate per 101,000 Province ' 22.5 Cc'sort ' (Excluding Asbestos workers) Ttrtal 33*<3 Proved 25.3 He gives no ;flqt,Jag included. This would be statistically significant since same exclusion factors are not applied to the controls. Secondly, he fee] that the general population is cot studied oa diligently as the miner and, tea refore, the rate cany be low in tha control, implying that the difference, therefore, is not real. Tha fact that the cohort did cot necessarily show a progressive increase in these years la not a valid argument because variations tho re would be a rl3e in lung cancer after the latent period is fulfilled. > 9. Lung Cancer in Aab33toa "in<r i rVvr.cu.toi Braua: Tabic 17* Braun uses -a-ccla-ss.if'iscartiosn -o`f "prlsary" and "pxonrm'Ua** datcrr-ining vhotlir death certificate carried tha inclusion of autopay or Burger/, etc. This lsjgot a valid cocporlsocu There not the saca degree of accuracy in recording data on death certi Tabiel3. Given corg3ar3 >n of actual and expected nuntber of lung cancer deaths by age wrong asbestos sinew. * 'r-ncuso! First, Braun incs the 195^ cad 1955 province rates rather than 1 years of actual study froa vhich the population vaa studied, tm? had a higher rate, anti vith the control higher, it becoaee less likely to pich up the difference in the cohort because of the Basil nuabow involved* Ih ths 65 years and over group, there were 5 obder deaths compared to the expected of 2. Be, hovever, points out and emphasizes that this includes 1 of the "suspected** case* to indicate vhat appears to he a suggestion of statistical evidence. Be that "it i3 rather likely that tbs rate for the general population is understated in this age group for the obvious reason that the exact cause of death in ths very old is cot a ratter of the saoa intensity of interoat oa it ie in tta younger person. ** This is not true# Ha expects to icply that the rate for this group would be higher in the control and, therefore, tho differences observed by the cohort in this age group would not neon anything. In T-blo 18, says nenbow of cohort have not died frora lung cancer at an age earlier than general population. Karvcusoj that the cohort consists of 6S of age j \y Croups of 20-hh, and ` ,1 10. king Ccs.ccr in Asbestos Mir.*rs ;ci:so: expsrieneej so, if the Zj-hh group did not have a cufficlent latent purled, they vould not nave lung cancers. Hocall too that the cohort vas followed onfor 6 ;eaw? As to the corosat about tho vith core than to ; rs of employment, this may be a selective fa with the hardy mirrivor i. There 1b rtiua: }b reuse: Discussion-page 6^7 Broun rjakeo his cos a again on the total rotee-oll ages, 20.3 per ICO,000 for Canada. Compares to 253 per 100,000 proven cases. The 33*3 for total cases included the suspected. Feol that,.Table 16, ago Buapectod, ratbor thaa| disprove, particularly vbect rceor;aizin$ that Cfft of the cohort la 20-- age group were in tha 7^> under 20 years exposure. d&r^cPCPul&tlon < cohort without suf^cieit follov-uphna autocaticslly diluted the, ^results. /actually by ago apacific rates, the controls are lover than the cohort study as it now stands. If cohort vas followed longer thE vould shew up rore strlXlrygly that limitation of time period of folio* influences absence of cases. To not agroa vith statement ba mbca that "tin re are no important differences between tho rates for asbestos nInara and general population of Quebec and Canada." Then proceeds to catg^q^yith S^t^jafaMji^teh nra^ Bruun: y/mcn^oo: (tSmsvraa&xtsam&a Page 643 of discussion. 1 rr ! r 1 1 . _ _. . _ -1 --- Wovf Speculates that about 50,000 persona exposed throughout tho world have 1,000,000 ma years of exposure. Since only 150 cases reported in the literature, gives rate 15 per 100,000. I s~ .^ .Ki.ieva.v3: (continued) ' Eilo is ridiculous ccause.icttuy ^caaeo of Lung cancer never are reported. In tha literato* nirply because studies vero not mde in the onbostoa plants, -hxuap'e: United States report* la literature axe not an accurate gauge a- to vbat baa or bos not occurred as far as lung cancer in concern;.. Naneuso j Table 22. This table sboxa tfcrvt lung cancer rate of tbe county la vtiicb tha . onbotes plants were lomted bus rate 2 tinea tha adjacent oountiaa, 13.9 to 9.4. This is a valid ccspirlsoa, certainly cole* a better control. He has been using the Province of Quebec as a control vhich. . I believe includes ttontreal vhich raises the overall rate of the "* control* r.raua: Refers to additional cases not included in tbo cohort. ?or example the 9 deaths prior to the beginning of ctudy due to lung cancer. !-h.r.crjjJO: i Braua refers again to "at least 7 of 19 proved lung cancers were ' not accocixialed .by 0.5besreals. " This is not tha point about dealing 1 only uith the biatologically proven cases. Ifo should have uaod as a i1 . control, the age specific rates of tbo surrounding counties \ ^ of tha total Province which van weighted with other industrial areas, j vhich raised tbs total rate of tbo control. If ho hod coaparod with the adjoining counties on an ago specific 1 basis, it would bava been statistically significant beceune the ooe county van elreody 2 tisrta tha rata of the adjoining county. - \