Document nk38kkp3vw9LgOrYpyyV4MLqa

T @Arch Occup Ensiron Health (IW)62 13Y-IJ7 I l'hirty-four-year mortality follow-up of BASF employees sposed to 2,3,7,8-TCDD after the 1953 accident adreas Zober'. Peter Messerer', and Peter Hube? The influence on health tor the c.mpio!ee, i n \ d \ e d has been described comprehensively in the literature: BASF physicians gave firsr-hand accounts of the accident and presented touicolopcal investigations ( 5 .N. 1954: Hofmann 1957).The! published morbidit! studies IGoldmann 1972a. b. 197.3: T h i c s and Goldmann 1'476) and rnc,rtalit! ~ t u i i i cIT~hic<>and Frentzcl-Be!n\d IY7' \\ith IUIIOU-UI~5 OcrC>!?cr !U-<. Tliic.. c't lil IW with follow-up III 31 .tunr ;%\Ol. ,\ murf;litt! !tu~l! ni!h follou-up to 31 .August IY83 was published by Lehncrt and Szadlkouski ( 19x5). using documented information thai had been presented to the Brrufsgenossenschafr der Chcrnikchcn Indu\rric' i Emplo\ment .Accident Incurancc Fund ot the C'hemical InduNir! I - UG C'hcmi< - b! B.ASF Thc\c piil~lic.itinn\mnf,iin Je f ni l e d dcccrrption\ of the circumFtances surrounding rhc ,tccidenr. d> H C J I II:, the actibities in the pl.rnr in [lie !i.mfolla\\Ing the acci- dcnr. This reporr presents the findinrs ot a mortalit\ btudy with follow-up to 31 December IYX? of 2.17 employe<. who were divided into three cohorts on the hasis of exposure information. Limited morbidity data are also presented. along w i t h a di\cus\ion of the Iiteriirure on TCDD &At the time U i the .'Dioxin Inrrwpstiun Prugramme" all w n , in\ol\ed with the accident in 195? uho uerc known by and u h u Here ,till alive were iniurmed persunally ahour the and txtent of the planned clinical projrct I.?. company physi-f , To isr re$^ the riA of exposure. S\IR, u e r r cnlculared nr the n t i o oi oh*er\ed IO e\pccrsd death?, 2nd expre?,\cd3s perk'nrage \dues. Cdculatim\ u e r e c;irricd CIUI u i t h retrrence IO the time since h i expowre: Yll";b CII s e r e cslculared using I3trulaIed \:iI. ucs from thc Poi\son d i w h u t i o n for ohrrned n u m h c n IGardner and Alrman 14WI. The uidth <if the CII u a \ intended IO conye! rhc preci\ion o t the SMR. 3 u i d e confidcnce i n t c n a l indicating less precision t h m d narrou onc Siarirtical rem Herr c a m d i w i uith u = 5',, ime-*idcdI. The ak!-wJed test u a %i h w e n k c a u w oi pnor inform;irion concerning the arcinogenlcit! ot TCDD in Animal c s p r i m c n t s I E P A lW1. Table I. Penon-)ears h! cohort and !ears since first e\pclrure 0-4 Ill-lU 31- Results B.!- ( th 11 e end the n of the 4 jviwi fo il- l lou 111 i -up ,iIio p rt erio il d . 1.11 2- < I Decem t' t i i c S: ber iii . l'-)h71. ~~ht~ : t C 2 . and 24 of the 91 in cohort C? had died. There u e r c '1 death\ ~ l u ctci nialign.int ncoyi.r\m\ in C1. 1 I in C_'.ind 3 in C3. The results are described for the three cohort5 separately and for an additional subcohort containing dl members of cohorts C I . C2 and C.2 uith chloracne or "er!therna". The term "rr!thema" is used for find- i n g that ucre wgge\ti\e hut not cle;irly indic;iti\c of :hlor:icne T.,i)lc i . l i t l i \ . i i i c di.trii>uit+'ti $ 1 1 i i i i m r w i ? - , i b . i r ~ , h e r \ .ition rlrr tlic thrcc co1it)rtY grouped !.I i i n i r ,iiiic the fir\t cspmure. Fur the perwns nith chloracne o r "evthcrna". there was a total of 3.5SY pcrwn-!ears. , -. _ . ? - -c c c 7 .. 4 -C . I 4... -- ;E: ...., - m ( C r j - 10 149 and for persons from the subcohorts uith and without chloracnePerythenia.' (Fig.2). I1 can be seen that members of cohort C I show the highest median value of 24.5 ppt 2.3.7.S-TCDD in blood compared with those of cohort C2 (Y.5 ppt) and cohort C-3 (8.4 ppt). I t can also he seen that increased and "normal" concentration^ occur in all three cohorts. Looking at the distribution in the tuo g r o u p with and uithout skin manifestation. it becomes etident that persons with chloracne hate higher average blood concentrations of 2.3.7.8-TCDD (15 ppt) than persons uithout chloracne (5.8ppt,. Of special interest is the fact that in the group without chloracne some high concentrations were also ohserved. Discussion The rewlrx prc'ccnrcd in our .X!.e:ir mortalit!. f ~ l l ~ \ \ - ~ p of BASF uorkers who were esposed to TCDD require an estensive and detailed discussion. as currently the dioxin problem is of considerable importance in both German! and worlduide. The toxicity and cnrcinogenicit! of 2..;.'.P-TCDD in / m t t i m . T has heen summarized reccntl! (LB.-\ IW5: EP.4 lW5: \1,4KIW6: C'DI IUS-: C t r s L b n l c c L'Jh.'. T K b S lL),hh. h h c n e L.1 LiI. I < J h U i Tiir q u L - t i i m (11 rhc. c:irciiiogeiiic etfect I\ ot \pca:il iiiiportance. because in animal tests 2.j.i.S-TCDD has been found to be a potent carcinogen (Kociha et al. 197s). In the Intest evaluation with regard to human carcinogenicity or 2.3.7,s-TCDD. the International ,-\gent`! for Research on Cancer IXRC'J concluded that there itas "inadequ;tte s\idenc'c' for c:ircinogenicit! to hunirin\" ([,ARC I9S7). .\!t?!h otloloyicul l.~slreT The choice of an appropriate reference population for the calculation of an e.ypcctednumber of deaths is an important methodological issue. which has frequently been discussed (e.9. XlRC 1954: Gardner 1986). In spite of the xh3ntagec ot an intc'rniil control group. \\e h a \ e uwd ,in eYferii;il retcrcnce pcipuinrion. I e . . the n:itioii;il L!. till r.,::,, T::.!.c \ \ < r c : \ i L ~:_...\L -,fc>rI l 1 i . I Tlic rcii'iit rc.\tricri\ c' !cgi\l<i[i\iKconc-rniri; Jata p r o t ~ c ' t i u i iin [he Fcderal R ~ p b l i cot Cjeriiian!. and p i r titularly in Rhineland-Palatinate. uould hate required the time-consuming follo\\-up of another cohort. u hich uould have delayed the \tu$ uithuut ;in! guarantee of n more expre\si\e wientific a n w c r . 2. Narional death rates are based on larger numbers of c : i \ c ~t h x n thc r;iIc\ .iinoii; i i i i L . \ p w ~ I i + L ) r h c r j3opu!dm n h m d :ire rhercfoi-c \tati>[icall! mors \tat?lL,. .. '.:. Ciuse of death -Cohort CI C_' c3 I ? ,/ I , i I:.,.<.. ,411 m a l i g i a n i neoplasms C.:a!larnn of the stomach Carcinoma of the inteatine Carcinoma of the lung Lsuhcmia .k:ldcnt\ _S_u_t_c_idc - __ All carer I..!L., . ch..l.o..r~,a.cne LlICllid.. parable earlier data have been published. although they would have been preferable. During this period the local cancer rates for a11 malignant neoplasms uere about 4% higher than the national rates, which suggests that using the latter rates may lead to a slight overestimation rather than underestimation of the cancer risk in the exposed cohort. .A3 part ol' the e\Jluation of a stud!. i t is important IO calculate i t < \tatistical pouer to Jetect rignificiint eyce\s risks. The ability of our wid); to detect risk5 greater than two- or threefold at the job level (one-sided) for several causes of death is shown in Table 9. For all causes of death the size of the study was sufficient to detect any \isniiicaiit r i 4 greater than it\oiold. For ,111 i~i;~lign.~ni il ~ ~ ~ c~om~ h~inlcd~. [l hle *ltu~J\ had wtiicient pouer tu detect esccss r i A greater than three- fold and moderatel! good pouer to detect risks greater than twofold. even in the separate cohorts. For specific cancer sites. houever. the power of the study to detect these levels of risk is weak. .A serious methodological problem in a study of this niiture is that ut multiple ciimpari\oiib or h)puihe\is te3t- in? We have nnaly7ed the rewltz for tour cohort<. four pcriucls 01 time. diiJ 2 cdu\cz oi &,itti. Pur diiotlicr \\a!. there s e r e -352 &:ith rate coinpnriwnz. B! chance alvne J number of S \ I R t ma! lv ~ i y i f i c a n t l !C I C \ J I ~ ~ Thi>must he kept in mind u h e n t.\alunting o u r tindin?\. Confounding factors may ;itso disturb rezults. Be- sides occupational exposure such as to aromatic amines parrly used in the plants in question. smoking habits may also be a n important confounding factor when e\aIuat- ins exogenous carcinorens. Houe\er. there uas nu sig- nificant excess of smuking-related cau\es o f death (cir- culator); and respirator! diseaseb) in our \rud! Prci /om s r i i ~ i i z J The f i r h i mortalit! \tud! of the 1Y.i: BASF accident \\as reported in 1977 (Thiess and Frentzc.l-Be!me). uith up- dates in 19s' (Thiess et al.) and in 19S5 (Lehnert and Szadkouski ). There have been critical comments about the grouping of the cohorts as well as the statiaiQ, method.; ciwd. e<pccially in relation to the investig2:ici, in.1.Y8. 5 ( Le m f o r u m Dioxin 1986: Karmaus .:!-,.I I,.-. I: ~.IirL~c!L.!I.J t~!!icr 1986). Jiu such dent>. tur ewmplc the 5lonsanto accident in 1 ,',and Silberrrld 1W).The connection hetween this a i d ~ I I C i u u z l ~pLtb1ihc.d L3.ASF mortalit) icponn summarized in Table 10. Points of difficulty that might p~-it.l! ;iri\e from the different publications are :\. plained below. It was stated in various publications (Leserfo,,,,,, 19P6: Knrniaus 1986) that causes of death had been "manipulated" between the morbidity study by n i e s s and Goldmann (1976). in which 15 deaths were men- tioned (including 3 carcinomas. 1 lung and 2 stomachI, ~ n tdhe mL:rulit> \tuJ> b! Thxss a n d Frentzel-Be!Illc I Iq7-). Th!. .iecti\ntion is totall! unfounded for the f,i;. I c w in? rriiwns: I . In the I Y T publication. the causes of death Nerc coded by a nosologist according to the ICD. whereas in 1Y76 the); had been taken from entries in medical re. cords. Therefore. diagnoses that appear to be different in the 1976 and 1977 papers are in fact. of the same djsease process ( e . g . .haemorrhage of the oesophageal varices = cirrhosis of the liver). 2. The use of ICD-coded diagnoses from death cenific.~tc> iii [tic 19-7 pubtiidtion corrected t u o crronzou. diagnozes of death due to cardiocirculatory malfunction to Jcdth >I. J rchult OI ztoniiich cdnser and lung cancer. 3 , In the I Y 7 paper (Thiess and Frentzel-Beyme). the cohort was redefined to include only those workers who had been directly affected by the accident in 1953-1954. T h u c . the ! 4 F death due t o acute haemcwrhasic Fan- crc.lii[i\ , i f .I\\clJcr \\!io \\.I\ cxpo\cd J u r i n g thc dcmoii- tion 01 the ~ u t o c l a \ ein I Y 5 h nut included in thc cohort. but uas presented as a case report. In the 1976 paper. this case was included among the 15 deaths. During the years 1Y76 and 1977. there were 3 further deaths ( 2 accidents. I disease of the urogenital tract) resulting in the t i deaths. and these were correctly reported in the 10" p3pc'r. 4. The age at death of a case of bronchial carcinoma was :?i:.!.:hL.c!! a ! x ! i ~ c i t d I!! [ h e 1U-h ?.!Fer .:* 5 4 !e,.ar.. hct correctl! a\ Ch !ears in the 19x2 pulication. Cirrretir yriitfy .According to the derailed description of cohorts C1 to C.7 in the "Slethods" section. it can be assumed that as far as posiihle dII prrwns directl! Jffected b); the accidenr Jrc. in cohort C 1. This cohort of the most likely and m o s t hc:i\iI! c..iposed \\orkers thou\ only a moderately non.ignific3nt e/e\ation in risk for a11 malignant neoplasms. The onl! \tdtisticall\. significant death rate in cohon C1 uas for wicide in the 0-109 !ear period. It is difficult Table IO. BASF mortality studies after dioxin accident in 1953 acicrence Study t!'P End of Size of follow-up cohon Sumher of persons with chloracne or "en-thema" Thiessand Frentzel-Beyme I 1977) Thiesset al. (1982) Mortality \tonality 15-10-76 30-06-80 75 71 Noinfor- mation _<-< Sumhrr of deaths 17 21 Lehnen and \tonality 31-08-83 153 78 13 Szadkowski (1985) Present study '!OM\ I__._ __ _____ 151 Thereof Type of malignant neoplasms malignant neoplasms 6 2 bronchial carcinomas 3 carcinomas of the stomach 1 carcinoma of the colon 7 3 hronchial carcinomas 3 carcinomasof the stomach 1carcinoma of the colon 17 6 hronchial carcinomas 3 carcinomasof the stomach 2 carcinomasof the colon/recrum 2 carcinomas not localized 1 carcinoma of the laryx 1 carcinoma of the liver 1 carcinoma of the pleura I carcinoma of the tonsils v o hronchial carcinoma3 + I I 3 carcinomasoi the stomach -- > 4 i;iriiiiom.J>uf thi. .-olnn'r:crurn 4 carcinomas not king localized 23 ( 1 therenf gastrointe\tin;ll) 1 carcinoma of the larynx I carcinomaof the liver 1 carcinoma of the tonsils (triple tumor) 1 mesothelioma 1 leukosis I carcinoma of rhe Li~fne\ to interpret the importance of such a finding for only a Dart of the follow-up period w.ithout corresponding eviwnce lor the entire period. The two buicidc cdse?,\\?re Ziaftsmen Hho were both involved in clean-up work shortly after the accident and subsequently sutfered rrom chloracne. The two suicides were committed 6 and 8 years, respectively. after the accident. There is no reliable epidemiological knowledge about suicide in the Federal Republic of Germany. In a recent review from the United States ( M o n k 1987). psychiatric illness and especially depressive illness are given as strong risk fac:or< for cuicide. In the zpecial circumstances of cohort cl. a reacitve depreqcion due to the development of '.!I Ir.iinc c~iuldbe wggchtccl. lwt [hi\ i3 \pccul.iti\c. In cohort C-2. \\hich IS It.\\\\ell defincd \\ith regard t o in\ol\eiiient in the accident. the S h l R for 211 iiwlig"ant neoplasms in the total period is also not raised to a Ztatistically significant level. but is higher than in cohort c l . The 2Ot year period SRIR. however. %as signifiYntIy elevated, although there was no specific cancer *tfe that contributed to the elevation. onlv an excess 'mmg the other and uncpecitied cancers. Tahle 6 Iho\v< at least ceben different cancer sires ioxin exposure i n the individual cancer G I ~ ~ 1S3 i!c!c:ir, Thc S \ l R f o r other ,Ind tin\;ycciiicd malig'':'nr iieoplasms n :I\ signiricnnriy clc\ a i d toi t i l < icli'ti Dcrlr>f! (Sr\,lK - '2 1 ) Thic y w i p ,)f cancer\. h o \ \ d v e r . contains 3 death from pleural mecothelioma with knosn exposure t o asbestos. Exclusion of this death reduces the SSIR to 256. based on four deaths. uhich is no longer >ignific:intl! different from unit!. As in cohort C I . there was a sipnificant excess of suicides for a discrete pcriod of time. but not for the total period. One was a fireman who had cancer of the larynx and is neither mentioned in Table 6 , nor in Table 7 because he developed the disease in 1956 - well before his first exposure date during the burning of residue in 1961. The cancer was successfully operated on. He committed suicide nearly 20 years later. The two other persons were bricklayers and the suicides occurred 10 and 18 years. re- 5pectitcl~;.liter fhrir p o w h l c exposure. Di\cussion of ;1 ~.iti\;il rt.latit)nchip betu.een the nccident and the wicides \ \ ~ ) u l dbe < \ e n itiort' q x c u l a t i v e rhan in c:iw\ or w h o r t C 1. ;I> t!icre \\xin o inciJence of chloracne. Another case in Table 6 should be discussed furthtr. Case 14. uhich is correctly coded as carcinoma of the tonsils. was described in the death certificate as "threefold" carcinoma of the tonsils, lung and prostate. After a discussion with pathologists. we believe that the cancer of the pru\tatc ih ail entit! in anJ of itself. Tlie c i i w w that It'd to J w t h was most probably the carcinoma of the t o n A . uliilr .!,--t~.iIl! th'*rc w a z nn hronchi:il carcinoma but in\tead a iiict;i>id3iz ot tiic carciiiwiu o i tnc. w i i A ill; lt.n Colicirt C i is tlistinguished by a multitude of Fr . , , , I , , . L.i:i L3pLLi.i!i! A !*,\>, Y \ l l < ( i v : ff?: ::I! : : : , ~ l ~ ~ ! l ~ ! ! ? ! neoplazms. We cannot otter a n cuplanation fur this l o w SXfR except for unknown selection etiwh iil ~ t d i u iCt 3. q !I I I! i -. '3 '3 -' 153 There was no significantly excess mortality in the chloracne cohort for the total follow-up period. A significanr excess for all malignant neoplasms was observed for the 20- year period. which followed low risk levels for the first two time periods. It may be of interest to note that recent work (Beck et al. 1989) has demonstrated that some exposed individuals with elevated TCDD blood levels did not hare chloracne. Thus. while chloracne is usually a marker for dioxin exposure. its absence does not always imply nonesposure (see Fig. 2). i7 ;1 i ,Voiifaial iiimoius It was not possible to do an incidence study of the es- posed BASF workers. as there is no population-based cancer registry for the Rhineland-Palatinate area or the FRG. A limited evaluation is possible only by means of the cancer sites and the distribution of the turnours in the different cohorts. Four men who Here certified as having died from some nonmalignant disease are known to have had cancer - two in cohort C l and rwo in cohort Cl (Tnblc 7).For cases 2 and 3. causal factors other than possible dioxin exposure are probable (aromatic amines. X-ray<). The ten tuinors in persons still alive (Table 8) were also not concentrated in any one cohoit. The cases in- volve eight different cancer sites and different types of exposure. Cases 5.8, and 10were cancers of the bladder uith known exposure to aromatic amines. a recognized h u m a n zarcincyen. .A rctrospecti\e nnal!.;i. <)f the mate- ri.il\ in thc toriner hax tactor! - h c n t s t h a t wch ;irllillc.; had been handled there in the IYiOs. All 3 cases H e r e acknowledged by the BG Chemie as due to occupational 7 disease. Of special interest art' case5 1 and Y. The diagnosis in case 4 is, ciirciiioina ot the liter. \\ hich first tvcamr J V - parent due to acute bleeding in the abdomen at the age -* of 58 years. Surgery was carried o u t iiniiicdi'iiely m d histological evaluation ,hosed il liver cell carcinoma without cirrhosis o f the liver. This is a rare constellation (Thaler 1987). The concentration of 2.3.7.5-TCDD of S53 ppt in blood was the highest value found in all 28 persons examined. This unusual combination is il very striking one. .\5suming a half-life of 7 ! e m and ;I first-order kine- tics. this worker \r.ould have had n n r..itrernd~hish coil- -4 -a 'i -' ic.iiir,jticin ,,i I..;.-.k-TCDD i n [hi. blood < I f inore r h a i l 3 I . I H N I ppt nlic'ii e:\trapolatc.d tu the t i n i t of c.spo>tlt-c. .:!.I !ear\ bt.loi :. Conipred itli a t h c r c.srrapi)latc.d levels from indi\iduals with various exposures (Patter- son et al. 1YXY). this concentration i, one of the highest \ d u e s published to date among production worker%. The diagnosis of case 9 - a person without chloracne but with a rather high concentration of 2.3.7.5-TCDD In the blood - \\:I\ !ir\t Ji,iglici\c.ci i i i l0S-l ;iz rl i i o i l Hodskin'.; lymphoma ( S H L ] . In 1989. a rccvaiuation ot [!iL!l'i\(i,i,i;lcLil .lit!cL :I! rhL. Grrm::n I.!. !n!>!!c>t>1i\ R c ? I \ ter ( P r o f . Lenncrt. University ut k i e l ) - uhiilg in1i11uIIo- I ! i ~ t ~ i l o ~ i ~n.i:c1.r1hnciz - lctl 10 [he reviwtl di:igno\i\ o f ' I >lcrkct-ccij c;lrcllioiil,i. 'I i i i o l t + i i c l i i t tiiii\\bi Lii !iic hi:^. The. IlcrC~~I-icclalrcinoma or trubcculnr snrcincma cnn 15d occur - mostly as a solitary nodule - in many areas of the skin. Morphologically, it is a malignant argyrophilic carcinoid from the grcrip of mn!ignmt n p i i d o m ~ (.4PLID = amine precursor uptake and d e c a r b o y k t i o i l ) . I - t o t 1 1 ihe \IC\+ J i t l c r c i i t i a l di<igii<)\i\t.l ~ ctiiiiitti- mity of the tumor cells and their iininature aspect may sti<;c\t LI 1?13\tis t!pe o f l>i:iphonJ. Thi, :ax u n l s r scorcs the acknowledged difficulties in diagnosing a NHL (see Lever and Schaumburg-Lever 1983). Another person (not included in Table S ) hhould be mentioned because of the high concentration of 2.3,7.8T C D D (138 ppt) in his blood. The worker was diagnosed as having an intracerebral arteriovenous angioma (diagnosed by computed tomography at age 53). We did not include this case in Table 8 because this type of an,n' ioma is not a tumor. but a vascuhr malformation of "cr!'Flic" origin (Rubinstein 1972). Orlrrr caxposiires cuid srirdirs There are a multitude of case reports and epidemiological studies concerning human exposure to chlorinated phenols and contaminants like T C D D from routine occupational exposure or from industrial accidents. Table 11 summarizes these events and the corresponding literature. I t contains only data that could be verified in t h e original literature. Earlier published wr\c.!s ( Reggiani 1981: Youns and Reggiani 1988) contain additional iniormatioii about further industrial accident4 for u hish. however. no precise data were available. These involved three episodes of occupational exposures in plants of Hooker, Thompson-Hayward and Diamond Shamrock in the United StJtes bet\\t.cn 1956 and 1960 with an unknottn n u m b e r of atfectcd perwn5. :I' well ;I\ t o u r further occupational esposures from I952 to 1952 and from 1970 to 1974 with a total of 117 persons involved (Boehringer. FRG: unknown company. Japan: LinzChemie. Austria: and Bayer Uerdinpen. FRG). Case descriptions of diseases related to the occupational application of chlorinated hydrocarbons fe.p. Stingily 1940: Grimmer 1954: Oliver 1975) are not listed in this table. Tnble I 1 p r e w n b 20 epwdcq c ~ If ndu\triaI <-\po\flrc between 1899 and 1976 that are documented in detail and u here npprc>siinately 4.700 penon.; itere i n v o h e d . By also including accidents not as precisely devxibed. it can be estimated that there may be ahout 5.OW persons potentially exposed to dioxin or affected by dioxin as a result of occupational sources. The exact figure cannot be stated because in most publications there was no separation betwen exposed and affected persons. In addition to the BASF accident in 1953. sis further accidents have been documented in other plants ( h l o n wnto 194Y: Philips 1963: Dow IYhJ: Rhone-Poulenc 1966: Coalite 196s: lcmrsa IY7h). Up to no\\ only the four accidents betheen 1949 and IYW have been evaluated epidemiologically. X summary report ( E P A 19S5) points out various weak points of some of the studies (e.g.. small size. low sensitivity. lack of sufficient latency period). Apgrepation of the latest results of the studies bv Z x k and Suskind (IYSO). cook et al. (19SO). Dnldc.rup and Zellenrath (1983) and our present study (see Table 12) shows a total of 392 (probably heavily) exposed perw n * from indiiqtrial accident? with ohsenrstior, between 14 and 34 years. Altosether. there have been sF, i,)tdi c l i ; i i h z <t>iiijiiiiid \ i i t i i ilic iIG.5 < \ i l L L i c < i , ;Ind -, deaths from niahgnant neoplasms compared with the 24.5 e x y e c t d . The s t ~ ~ h t i c;iLlRI~~ t~ detest .I ,;2i,,,.. cantly excess risk greater than twofold for all malignent neoplasms from these asgregated data is ahnut 98% S u m special remarks are nece>siryabout the S C accident. which is under careful study by scientific corn. mittees (Reggani 1988). It appears that none of the plant workers was injured and their health has not been affected. There is ongoing health monitoring focused on the general population. A cancer registry was set up i n the Seveso area. and the cancer \tatistics betneen 19-1 and 1YP1 bhnu that the Incidence and frequency of tu. mours are coinparable to those of other populations. In 19S9. the first lO-!car mortality study of the population involved in the Seveso incident in 1976 was published (Bertazzi et al. 1989). Mortality from several cancers was elevated. based on very small numbers. No pattern related to exposure classification was evident. In the most heavily contaminated area (A), the mortality from all cancers was not increased and no deaths from leukemia. which might be expected to have had a short latent! period. sere ubsened: an increased relatike risk from cancer of the gallbladder and estrahepatic biliary tract in ieiiiales and irom cancer of ill-defined bites in males resulted from one death in each c a t e g o n . Other sources are under investigation. along with TCDD exposure after industrial accidents or through routine occupiitiunal actit itics. E s m p l e a are the milit~ r aynd c . i \ i l uze of tierhicicie~'ind the w ) - ~ i ! l d\ t a s t e disposal episodes. X recent summark of thesc topics has been presented by Young and Reggiani (19%). Because the reports of Hardell and colleagues (Hardell and Sandstrom 1979: Eriksson et al. 1981) on soft tissue sarcomas (STS) in users of chlorophenol and phenoxyacid herbicides possibly contaminated with dioxin in different rural industries are frequently cited. the conclusion of an , A u ~ t r : i l i a nRo!al Ctmmi-ion in rc.l;ition to the Hardell studies should be quoted: .... .the conimission does not accept the Hardell htudies a5 prminp. on the h l a n c e of probabilities. any causal association between soft tissue sarcoma and lymphoma and esposure to 2.4-D. 2.4.5-T and TCDD" (cited by Young and Repgiani 1988). This decision is mainly based on the presence of recall and observation bias. %lorcover. a recent case-control study (14 cases and 126 matched controls) among 37.000 chemical workers (Sobel et al. 1987) and a large cohort study on pesticide applicators (Wicklund et al. 198s) d o not support the thesis that STSs are caused by 2.3.T.d-TCDD. Finall!. a nen case-control study on STS by Hardell and Eriksson (198s) did not find any association nith chloro- phenols. Another rare malignant disease. non-Hodgkin's lym- phoma (NHL). has been associated nith esposure to phenoxyacetic acid herbicides and/or chlorophenols contaminants. Although often cited as an example of a T C D D qudy. the rescarch of Hoar et ai. (19Sh).which I55 rable It. Aggregation of published mortality data follow in$ industnal accidents a i t h dioxin 'Location - \tonsanto Yearof accident 1949 Endof folloa-up 31-12-1978 Observation penod in w a r s 29 Xumher of persons 121 Total deaths 0 E SMR 32 46.4 69 9 4SF 1953 3 1 - 1 2 - 1 w 34 6 9 ( C 1 ) 27 29 3 92 Philips 1963 1983 20 141 25 21 119 3 o w 1964 31-12-1978 I4 61 4 7.8 51 \lean value or sum 24.25 392 88 104.5 84 Deaths from malig- Reference nant neoplasms 0 E SMR 9 9.01 99 9 6.94 130 8 6 9 4 115 3 1.60 188 29 24.52 118 Zackand Suskind (1980) Presentstud! (1990) Dalderupand Zellenrath (198-7) Cooket al. (1980) .bowed an association between YHL and 2.4-D espou e s . imolies a different substance that does t i o f con.sin 2.3.7.S-TCDD. However. Cook et ai. (1986) found i b c case, of h H L as compared nith thc 2.1 cxpccted S M R = 2 3 : QC"b Cl: 77-556). in their cohort study of vorkers exposed to chlorinated dioxins. A n update of .he Cook et al. (1956) study by Bond et al. (1989) con.aim a separate analysis of their 323 chloracne cases. rhere were no NHL cases observed in that subgroup. 'ut one more N H L case was reported for the total co- iort. ;'rotnoror t:fjki Uthough animal evidence suggests a promotor effect for rCDD (Pitot et al. 1980: Poland et 31. 1982). our results iiow no support for such a n effect. Tuniors \ \ h e n the? lccurred tended to appear 20 or more year, iiiter initial \ p o w r e . Sloreover. i t may be an (~\er\iinpliiicntionto rrplain the toxicity spectrum of 2.3.73-TCDD solely via In isolated mechanism. I t is possible that multiple steps lr mechanisms act in an integrated way. depending on a nultitude of different influences. .-onclictfitig remurks \o strong conclusions can be drnu.n from the results of 111sstud!. The size of the investigatiori (247 a a r k e r b ) is . ~ ; 1 1 1 - tienit.. the ;ipp:rrenr lack of itiii\i\tc.nf etidcncc. ; C ' W C ~m~cbrtalit! doe5 not rsoneratc' TCDD. l f the c'f.xi of TCDD on cancer mortalit! i5 m a l l to moderate le>, than twofold). our study is insufficient in size to be -unfident that such an increase ~ o u l dbe detected. There \vas no significant increase in deaths from Inalignant neoplasms over the total 34-year follow-up neriod in any of the three cohorts, individually o r comined. There u ' i b . Iic)uc\er. ii 4gnit'ic.iiit c ~ c fosr all ;inters combined anlong thc chloracne kictims 20 or " o x \ ::::\ :!fter initial e \ p x u r c \\hen ; i n CYZCSC \v0~11d 'r rnog Iikcly to occur. I n addiiion. tlicre I > Ihc iiot'iblc inciin- I.! nne cage of liver cancyr \ + i l h o u t ciri-Iiosi\ iii <I -..*ahCi b i i i l '111 c \ ~ < l ~ t l < ~ iIli~isi:l >!c.;I I J ~ T C D L$)11 ti., '%oLl. I f T C D D were a highly potcnt carcinogcn in man. 3s Ins bcrn suggested. a substantial excess of cancer should by now have been o b s e n e d in our study population as well as in some other cohort5 of exposed workers. Our findings suggest that a small hazard may have been caused by the heavy exposure in the l Y 5 3 accident. bbt they d o not prove it. .-\cknow/edgentenrr. We are very grateful to Prof. Martin J. Cardner and Mr. Paul D . W'inter. ,MRC Environmental Epidemiology Unit. University of Southampton. England. for their epidemiologic insights and statistical advice and computation. and to Prof.Sir Rlchxrd Doll. Imperial Cancer Research Fund. Osford. Englmd. f<>rt.i. <ornrnciit< o n twir final ~lmic \!'e \vould like to thank Dr. Jc,.crli li. \kLJughhll ot :he \:irional C-:rncer Iiiatirurc. B c t h d d . \l.tr!ldiid. USA. lor 111sthuugh;iul ic\ic\\ oi thc papcr. Fur considerable secretanal assistance. we thank Mrs. Liane Hund and Mrs. Clrike Remlr. References Assennato G . Crrvino D. Emmett EA. Longo G. Merlo F (1989) Follow-up of subjccts who de\eloped chloracne followingT C D D esposure at Seveso. Am J Ind Med 16: 119-125 Baader EH'. Bauer HJ ( 1951) Industrial intoxication due io penta- chlorphenol. Ind Med Surg 20: 286-290 Baurr H. Schulz KH. Spiegelherg L' (1961) Berutliche Vergiftung bel der Herstellung von Chlorphrnol-Verhinduilgen. Arch G w e r b e p a t h o l Grwerhchyg 18:538-555 Bc:h H. E;h,irt h;. \fi1111ar \ \ \Yirrhmv\hi R ( 1'h')I Lr\c!, nf PCDD'\ Jnd P C D F s in J~IPUSCr i w i c oi ucc~upart~in:~ell\!- p i i ~\\~cirlhcr\ ~ ~ I i c n i ~ y l i c x Bcckrr \. Frenirel-Bc!me K. M \lort.ilit! 111the Fcdcrd Kcpuhlic u t Gcrrnun!. Springer. Ber- lin kleidelherg Sea York Be:rnui P.4.Zocchetti C. Pesatori .4C. Guercilena S. Sananci) M. Radice L 1989) Ten-year mortality xtudv of the population in- bolred in the Seveso incident in lY76. Am J Epidemiol 119: I lY7- I2M --Bettmann (Iu(111 "Chlor-Akne". cine besondere Form von profes- vuneilcr H.rurcr~l.;ii~huiDi~r.,iIi \led \\uchcibz.hr <.:-- un Rlcihcrs J . \\ nllcn M. Brodkiii R . .Applcluurn I I . 10hO In,lu\t- :I.,;;! . * . + f i ' L l ~ ~ ~ ~ l l d l ~!:<l hl . ,I>< B(ind GG. O t t \lG. Brcnncr FE. Cook R R (19x3) Medical and ;;I< ;:'i~Iit*, ~ i i ~ ~ ~ i 1 i Ii i. i i~ :l ;~: !~~i i~n~o i ~ ;~ ~ n l l + ~ !pL ~ ~ ~ * c ~ p i w 1r1n I c I j D . Br J Irw Irlcu + I . .: ih-.;:-r Boid GG. C w k RR. Urmncr FE. \lcl.arcn E.4 ( l W 1 Evnl!in- tion of iiiortality pattcrns 3inong chemical workcrs wlrh chlor- ncnc. Chcmosphrre 16:'I 17-2121 I I56 Bond G G , McLaren EA, Lipps TE,Cook RR (1989) L'pdate of mortality among chemical workers H i t h potential exposure to the higherchlnrinnred diouinc. I0cci:p VPC3!1 : !21-!2.: Butlcr LICi i1 Y 3 ) Xcncforin dcrmatobis produced h! ortho(2chlorophenyl,phenol sodium and tetra-chlorphenol sodium. Arch Durn ~ p :5 h: 2 5 ! - X Cook RR. Towmend JC. Ott MG.Silverstrin LG (1980) Mortality esperience of employees r.xpom.l to '1.3.75-TCDD. J Occup Med 22 :5 3 - 3 2 Cook RR. Bond GG. Olsoii RX, Ott Xi. Gondek MR (1956) Evaluation of the mortality experience of workers exposed to chlorinated dioxins. Chemosphere 15 : 1769-1776 Cook RR. Bond GG. Olson RA. Ott MG (1987) Update of the mortality esperience of workers exposed to chlorinated dioxins. Chernoyhere 16:2111-~116 Dalderup L\I. Zellenrath D ( I9S-T) Dioxin r\po>urc: 3 )\.ear I'dlow up. Lancet 11: 1134-113 Dugois P. Xlarichal J. Colomh L f 1958) AcnC chlorique 3u 2 . 4 5 trichlorophenol. Arch Mal Prof 19:626-627 Dugois P. Amblard P. Aimard M. Deshors G (1968) Acne chlorique collective et accidentelle d'un type nouveau. Bull Soc Fr Dermatol Syph 75 :260-261 EPA (1985)Report: health assessment document for polychlorinated dibenzo-p-dioxins. (EPA/600/8-84/014F). If'ashington. D.C. Eriksson M. Hardell L. Berg NO. hloller T. Axelson 0 (1981) Soft-tissue sarcomas and exposure to chemical substances: a cax-referent study. Br J Ind bled 28: 27-32 Fuss S (19.78)Chloracne. Zcntralhl Haut Geschl K r a n k h 57: 250 Gardner \lJ (1986)Considerations in the choice of cspccted iiumbers for appropriate comparisons in occupational cohort studies. bled Lav 77 :23-47 Gardner MJ.Altman DG (1959) Statistics with confidence. British Llc'dical Journal. London Gnlclmsnn P I I U 7 : i i Cchaerctc :tkurc C h h ~ r x h wJ u r < . h Trichlclr- phcnol-Zer~c'tzung~produkteA.rbcitsmed Sorialmed Xrheitshvg7:12-18 Goldmann P ( 1 9 7 3 ) Schwerste akute Chlorakne - sine ,Massen- intoxikation durch 2.3.6.7-Tetrachlordibenzodiosin. Proceedings of the International Conference Medichem. 27-2Y Apnl. Luduigshafen. pp 220-227 Goldmann P ( 1973) Schwcrite a k u t e Chlorakne. cii1e Mabsciiintoxikation durch 2..7.h.7-Tctr:ichl~irdihc~z(~iiic~H~aiunt.nrzr 24:14Y-I52 Greenlee W F ( IYXY) Diosin carcinogenesis - a pcrspecti\e. CIIT- .4ctnitie5 ,3:2-h Grimmer H I1954)Beruflich hedingtc .Aknc durch chloricrte aromatische Kohlenwasserstotfe. Zentralbl Arbeitsmed Arbeitsschutz 5 : 76-83 Hardell L. Sandstrom A (1079) A case-control study: soft tissue sarcomas and esposurc to phenoxyacetic acids or chlorophenols. B r J Cancer39:711-717 Hardell L. Enksson bl ( 1988) The association hetween soft tissue \arcomas and exposure to phenosyacetic acids - a new casereferent .itud!. Cancer b2 :652-656 Hay A. Silberfeld E (1985) Assessing the risk oi dioxin esposurc. Nature 315: 102 Herxheimer K ( I899) c b e r Chlorakne. bliinch \led N'ochenschr 46 :278 Hill ID (19'2) Computing man years at risk. Br J Prey Soc \led 26:132-134 Hoar SK. Blair A. Holmes FF. Boysen DC. Robel JR. Hoover R. Flammen JF I lW6) Agricultural herbicide uw and ri5k ot Iymphoma and soft ti~suesarcoma. JAblA 2 3 : 1151-1157 Hoffmann H ( IY57) Neue Erfahrungen mit hochtosischen Chlorkohlen\\-nssersroffen. Arch Esp Pathol Pharmacol 232 : 228230 Hofmann MF. Meneghini CL (1962) .L\ proposito delle follicolosi da idrocarburi clorosostituiti (acne clorica). G Ita1 Dermatol 103:427-450 IARC ( 19871IARC monographs on the c\alution oi carcinogenic risks to humans. supplement 7. IARC. Lyon. pp 331-354 --Jensen NE (1972) Chloracne: three cases. Proc R Soc Med a:,, 77 lira-ek L . liale!?.4v .I. Kiibec K . Panlerova J . Lukaz E t.7 , - . Chloraknc. Porphyria sutanca tarda und andcre Into.,ii;at,L nen durch Herbizide. Hautarzt 27:328-333 k.irmaus \\ ! fv\n) Der 1 ) I i ~ v c - L ' n r ~ki l i AASF I Y ~:-,;: , ~ - , Nachweis einer Krebsursache beim Fvlenschen. In: Elsner Karmaus u'.Lil3nc.r L (cds) Ut18. W e i t krank machenl ~ 5 ~ Hamburg, pp 158-167 Kirnrnig J, Schulz KH (1957) Chlorierte aromatische zykiische Ather ais Ursache der sogenannten Chlorakne. Natunvisren schafren 44: 337-338 Kociba RJ. Keyes DG. Beyer JE. Carreon RM. Wade CE. her DA. Kalnins RP. Francon LE. Park CN. Bernard s r ~ ~-Hummel R.A. Humiston CG ( I W S , Results of a t \ ~ ~ - ~ ; , ( chronic toxisit! dnd oncoyenicit! study oi '..;.:.S-TCDb rats. Toxicol . + ~ P I Pharnilrsol 46: 27Y-303 Lnuer H (1986) Dns ..Untercuchun_rspro_rramm Diosin- dr.r B: rufsgenossenschafr der chernischen lndusrrie. Arbelt,m,u Sorialmed Praventivmed 2 1:206-207 Lehnert G. Szadkowski D (1985) Zur Humankanzerogenitat 2.3.7.8-TCDD. Arbeitsmed Sozialmed Praventivmed 20: 225- 232 Leserforum Dioxin ( 1986) Arbeitsmed Sozialmed Pdventivmed 21: 153-175 Lever LVF. Schaumburp-Lewr G (1983) Histopathology of thc skin. 6th cdn. Lippincott. Philadelphia. pp 577-579 %!.AK-Wene. T,~xikoI[iei~h-arbritcmedizinizchBc ecnindunp ! O w . 2.3.7.8-TCDD. Vcrlag Chemic. IVcinhcim May G (1973) Chloracne from the accidental production of tetra- chlorodibenzodiosin. Br J Ind Med 30:276-253 M a y G (1982) Tetrachlorodihenzodiosin:a survey of subjects ten ! e x \ ;titer c\powrt. Br 1Ind \led 39: 13-1-35 \lorit, \ I \ + - I EpiJcmiciltig! ( i t wicide. Epid K c \ 'J:Cl-hQ \loses 41. Lilis R. Crow KD. Thornton J. Fischhcin ,A. Anderson HA. Selikoff IJ (1984) Health status of workers with past ex. posure to 2.3.7.8-tetrachlorodibenzo-p-dioxinin the manufacture of 2.4.5-trichlorophenoxyacetic acid: comparison of findings with and without chloracne. Am J Ind bled 5 : 161-182 MRC (Medical Recearch Council) ( 1984) Environmental Epide. miulog? Unit. E1pectc.d numbers i n cohort j1uJir.s. Scientific repon no 6 . SIRC. Swth:impton h h I 1954) Hnurmudigungcn durch unhekanntt: Lrrwzungsprodukte. BXSF. Ludtvig5hnfr.n..Arheitscchutz im LVerk 3: 11-12 Olitcr R J l i Iu-5 1 T o w L.tt'cct5 of :.-;.-.h-TCDD in laboratur! worker\. Br J Ind \lcd 32 :49-53 Ott blG, Holder BB. Olson RD (19SO) .4mortality analysis of emplo!ees c'ngaged in the manufacture of 2.4.5-T. J Occup bled 2 2 ;47-50 Ott SIG. Olson R.4. Cook RR. Bond GG (1987) Cohort mortality studv of chemical workers i b i t h potential exposure to the higher chlonnated dioxin\. J Occup bled 19:5.42-429 Patterson DG. Finserhut \I,+. Roberrs Dit'. Needham LL. Haring Sueens! \I. \farlo\\ D.4. .-\ndrews JS. Halperin WE I 19S9) Le\ els c>ipol!chlurinatrd dihenzo-p-diosins and dibenzofurcins in uorkrrs c'~po\edto '.3.;.8-tetrachlorodi~enzo-p diusin. A m J Ind \led 16: 135-146 Pazderova-Vejlupkova J. Z h c o v a M. Pickova J. Jirasek L (1981) The development and prognosis of chronic intosication bl; TCDD in men. Arch Env Health 36:s-11 Pitot HC. Goldsttorthy T. Campbell HA. Poland A (1980) Quantitativc' e\ aluarion oi thc promotion h! 2.j.i.P-TCDD oi hepatocarcii~ogenesis t'rom dieth?Initrosamine. Cancer Res 40: 36 I b-3620 Poland AP. Smlrh D. \letter G . Possick P (1971) .4 health survey of workers in a 2.4-D and 2.4.5-T plant. Arch Environ Health 22:316-327 Poland A . Palen D. Glover E (1982) Tumour prninotion TCDD in skin of HRS/J hairless mice. Nature 300:271-273 Rcggiani G ( IYSl)Tosicology of TCDD.Reg Tosicol Pharmacd I :1 11-2-13