Document 6bk2ObEr0eyvd65xBvJ40e6ER

*IV 22 * JAMA. April 6, 1964 Asbestos Exposure and iSeopJ Irving J. Selikoff, MD, Jacob Churg, MU, and E. Cuyler Hammond, DSc.Nevo ^ m h *> Building trades insulation workers have relatively light, interm ittent, exposure to asbestos. Of 632 insulation workers, who entered tire trade before 1943 and were traced through 1962, forty-five died of cancer of the lung or pleura, whereas only 6.6 such deaths were expected. Three of the pleural tumors were mesotheliomas; there was also one peritoneal mesothelio ma. Four mesotheliomas in a total of 255 deaths is an exceedingly high incidence for such a rare tumor. In addition, an unex pectedly large number of men died of can cer of the stomach, colon, or rectum (29 compared with 9,4 expected). Other can cers were not increased; 20.5 were expected, 21 occurred. Twelve mert died of ashestosis. ALTHOUGH PULMONARY CARCINOMA had jf\_been observed in the earliest studies of asbestosis, association between the two conditions was first suggested by Lynch and Smith in 1935.* Ad ditional reports of such association followed. Per haps the most striking data was presented in the annual report of the Chief Inspector of Factories of Great Britain for 1955. Every death with asbestosis in the files of the Factory Department, from die first recognition of ashestosis as a disease entity, was studied. Altogether 365 such deaths were recorded (1924-1955). Sixty-five or 17.8% were found to be accompanied by cancer of.the lung or pleura. Doll,3 after reviewing die problem and add ing data of his own, concluded that lung cancer was a specific industrial hazard of heavily exposed asbestos workers. Nevertheless, some investigators have held that,, while these observations might be suggestive, they did not establish an increased incidence of carci noma of the lung in pulmonary ashestosis, and further, that the association was unproved. . The factor of selection was considered a poten tial weakness in evaluating reports of autopsy series. It was noted th at complicated and unusual cases would he more likely to crane to autopsy, thus raising the apparent frequency of associated lung neoplasms.4 Further, it was argued that au topsy statistics, which dealt with particular groups of those who died, do not reflect total populations of asbestos workers." Additional reservations were based on the frequent absence of data regarding From the Division of Thoracic Disease, Department of M edtane, and the Department of Pathology, Mount Sinai Hospital. Read before a joint m eeting of the sections on radiology and disease* of the chest and the American College of Chest Physicians a t Hie 112th Annual Meeting of the American Medical Association, Atlantic City, NJ, June 17, 1963. exposure, smoking habits, and personal history* a the size of series, and on inadequat histological verification in some cases. : Within the last few years a num berof additional problems connected with asbestos have appeared, making clarification die foregoing statistical uncertainty a jitatfW of con siderable concern. First, there has been greatly increased use of the various ty p eso f asbestos (a five-fold increase in world utilizatkm o f this group of minerals, from 500,000 tons to 3tJSOO,QOOtons per year m the last 30 years), as wvjpjar a ready in creased number and variety of hpustrial applica tions of asbestos (over 3,000 ujsnow re corded). Second, suspicion has heen growing that malignancy associated with asbestos exposure may include neoplasms other than carcinoma of the lung. Thus, a significant relationship has been claimed between diffuse mesothelioma of tfce^pleura mid peritoneum and asbestos exposure.* This communication is concerned with investi gations undertaken to study the following factors: (I) the incidence of deaths due tp pulmonary car cinoma among a group of workeas expqsed to as bestos under United States industrial conditions in the past several decades, (2) whether or not such individuals would also be found to have an increased risk of other neoplasms, and (3 ) whether such risks would be present in an industry other than the asbestos-producing o r asbestos-products industries, with which most reporte in the past have been concerned but which would not nec essarily represent the most important areas of as bestos exposure at this time. Further, it was hoped that study of an industry with asbestos exposure of limited extent and intensity would throw some light on the potential problems associated with minimal exposure to asbestos. Materials and Methods Our investigations have been concerned with 1,522 members of the Asbestos Workers Union in the New York metropolitan area, mem|>ers of New York Local 12 and Newark, h & L o caf3 2 of the International Association of Heat and Frost Insu lators and Asbestos Workers. As the full title im plies, these men are insulation workers. Althottgh the union is considered one o f th bfilding-trades unions, its membra* do. va riety of industries, including shipbitildihg. Called "laggers" in Great Britain, they are often desig nated "pipe coverers," "insulators," or "asbestos workers" in this country. Tlie union is one of the oldest in the s&<. tsi ' 142 V o l,i8 S , NO 1 SUESTGS EXPOSURE--SELIKOFF ET AE inder Associa __, was the t o t anion J p jS e United States, I t amal- *1., "V'|p- die current Asbestos Ifj!, W if stsM fty <rf1l4| naton In th e stability of Its member- w er, always a pipecovlogical importance to os p f men particularly suit- term effects of asbesworkers exposed in poorly paid industries, it records are mainhas also been con in ring _ and very few were the end of 1942 the of men with eonsidwho |oined in 1940, and 1962 onion _ ^ erf every. ! d either of the 1942, or ' union records, a list was who was a member locals on Dec 31, that date and Dec 31, , whatever bis subse1942 list included 632 . ^ . iw ere obtained from union records, and each man was detailed, ....... , m m , employment (war ser vice, outer employment, retirement, illness). These ' for calculation of the onset of exposure. For members who had i and Welfare Fund pro death. Copies of death on all but one of them. Autopsy protocols, histological specimens, and hos` * 5 " * * and reviewed in those f, in which die ters had otxiurred in a hospital. _ -Previous studies have suged diat neoplasia associated with asbestosis sel ; occurs until 20 years after t o t exposure to as bestos dust. T w e decided to limit the ' in p p n with such an exposure '" fp to p rd s eeYer aE members -jgpdve and retired memilSp) during m e 20-year 3, through Deo 31, 1902. $, the only men with a to ashCstos were of exceptions wore a -few men who foliiCd the .. after Jan 1, 1943 but who had been empl ' previously as asbestos workers men, 255 died before Jan 1, Table 1.--M an-Years dPtScperiem e o f 63 2 A sbestos W orkers Exposed to Asbestos O ust 2 0 Y ears o r lo n g er . Ag 35-39 ................ . 40-44 .................... 45-49 ................. . 50-54 .................... 55-59 .................... 60-54 ---- --- ------65-69 .................. 70-74 .................... 75-79 __ _______ 80-84 --------- -- . 35M- T otait - m a s# ms 339.5 ms 382.0 221# 139# IBM 31.5 if ' m. 1#*2# Tun so' fl^iSR.O' ; . 7 # . 486# 291.5 324.0 530# 308# 316# . 34I.S . 344.0 181# m# ., ttm - ' - 137# -m s 38.5 ' 8# 2 .3 3 6 # % 70# 314.5 - 502.S ' 268.5 255.0 280# . IS## ' 75# ./ 23# . Of these .. ,. . ,, to as- bestos dust prior to S LrT other words, as of Jan 1, 1943, 20 yeai " id elapsed since these 339 men were L The remaiitiiig 263 men read . ,. ,,_ at some time aft- Jan ti 1943, and before the end of 1962. The 339 men who ware first exposed prior to 1924 were counted in ead b of t o 20 y ea (or up to the dme of dead of those who died). The 293 who were first exposed in 1924 or later were counted only after they reached t o 20-years-since- first-exposure point (those who died being dropped a t the to m of death). When t o statistics were completed, we found th atw e had records covering a total of 8,737.5 man-years of experience of m m with a history o f 20 yea or longer since t o t exposure to asbestos dust. ' Of t o 8,737.5 man-years, 1912.0 were to.t o five- year period 1943-1947- 2,478.0 were to t o period 1948-1952; 2,336.5 were in t o period 1953-1957; and 2,011.0 were to t o period 1958-1962. Table 1 shows the age distribution of t o man-yea in each of these five-year periods. Table 1 shows (1) the average age-specific death rates erf aB US white males daring each of these periods, and (2) t o average age-specific death r a te from cancer of US white males during each period, as reported by t o US National Office of Vital Statistics. The man-yea were t o n multiplied by t o cor responding reported US death rates to ascertain the expected number of deaths under t o null do not. differ males ......... _ ... , eration). The resuits /r a te r of all US white ' ' .... into consid- in'T Shfe'3. as for aE US wi! ig those years (Ta ble 3). In other words, at start of t o study, t o asbestos workers had below average death rates. This is by no means 24 ASBESTOS EXPOSURE--SELUCOFF JAM A, April 6, 1964 Table 2 .--Total Deaths and D eaths From C ancer o f the Lung, Bronchus, Pleura, M ediastinum , and Trachea per 10,00Q W hite M ales per Year* . - , ______________ Age 35-39 -------40-44 ...._____ 45-49 .............. 50-54 _______ 55-59 _______ 60-64 _______ 65-69 --------- 75-79 .............. 80-84 _______ B5+ ----------- 1943-1947 1 "V*^ T o ta l Lung Concert 36 55 1 85 2 133 4 196 5 295 7 435 7 634 6 977 1.453 5 2.422 3 1948-1952 T o tal 30 48 78 124 189 281 417 605 915 1.353 2,162 Lung Cancert " .... 1 3 5 8 10 11 11 10 8 7 1953-1957 Total 26 43 72 115 178 276 416 586 878 1,346 2.017 Lung C ancert " .... 1 3 6 10 14 17 Ifi 14 12 9 , , 195SA-1962...................a T o ta l Lung C ancort 25 1 42 1 70 ` t tie - 7 178 11 272 17 21 sm 21 866 1/ H p8@ r~ rs 2,148 11 Death ra te s a s reported annually by th e US N ational Office-of Vital S tatistics. A five-year average is given h ere for each period. Average fa r 1958-1962 Is a D irectio n of 1958-1960. since 1961-1962 rates are not yet available. . . . . . . ... , . tD eath ra te s inclucta cancer of th e lung, bronchus, pleura, m ediastinum , and trachea, assigned to International list code No. 47b-f prior to 1949 and to No. 160-164 thereafter. is almost always found in the first few years of a prospective epidemiological study of this type. The explanation is almost certainly as follows: The 632 men in this analysis were actively employed as as bestos workers in 1942. Since disability from illness or other causes precludes employment in a trade of this type, these men were presumably well (or at least not disabled) at the start of the study pe riod. Almost any group so selected as to exclude the ill and disabled has a lower death rate during the ensuing few years than does the general popu lation, since ill and disabled persons have extreme ly high death rates. A selective effect of this type gradually wears off with time in d largely disap pears within five to ten years from the time of in itial selection. During the second five-year period (1948-1952) the death rate of the asbestos workers was slight ly higher than the death rate of all US white males, ie, 54 observed deaths compared with 50.8 ex pected deaths. In later periods, the death rate of the asbestos workers was proportionately higher. For the period 1953-1957*, there were 85 observed deaths compared with 56.6 expected deaths, and for the period 1958-1962, there were 88 observed deaths compared with only 54.4 expected deaths. Table 3.--O bserved and Expected Num ber of Deaths Among 63 2 A sbestos W orkers Exposed to Asbestos Dust 20 Years or Longer , r* Cause of 3 Total, a ........ ..................... 28 O bserved (asbestos w orkers) Expected (US w hite m ales) .... 39.7 Total c a n c e r,a ll site* ,, -- ........ 13 O bserved ( b a sto * workers) expected (US w hit* m atas) .... 5.7 Cancer o f h im and p le u ra ....... 6 I (asb eeto s workars) I (USW hite m a le s).... 0.8 Cancer of stom ach, colon, and rectum ................................. 4 Observed (asb esto s workers) Expected (US w hite m ales) .... 2.0 Cancer of all o ther site s com bined .................................... 3 Observed (asbestos workers) Expected (US w hite m ales) .... 2.9 A sbestosis ................................... 0 O bserved (asbestos workers) rears m 84 50.8 17 . 8.1 1.4 m m 56.6 26 13.0 13 2,0 47 2.5 2.6 56 4.2 8.4 14 ? a :m 88 255 54.4 203.5 39 95 9.7 36,5 18 45 2.4 6.6 14 29 2.3 9.4 7 21 5.0 20.5 7 12 Cancer of the Long, Pleura, and Trachea.--In each of the four five-year periods, far more deaths from cancer of the lung and pleura occurred among the asbestos workers than would have oc curred had their death rates from these diseases been the same as for all US white males (Table 3). Altogether 45 of thg 632 asbestos workers died of cancer of these sites, whereas only 6.6 such deaths would be expected from general US experience. Of these 45 deaths, 42 were recorded as due to bronchogenic carcinoma and 3 to neoplasms of the pleura. The pleural neoplasms were all recorded as mesotheliomas. Thus it was found that the death rate from can cer of the bronchus and pleura was 6.8 times as high among these asbestos workers as in the gen eral US white male population (both age and date being taken into consideration). It may be asked whether die high rate of lung cancer among these asbestos workers could pos sibly be attributed to an unusually large proportion of cigarette smokers among them. We cannot an swer this question directly, sbiee we have not yet been able to ascertain the smoking habits of die men who died. However, the following pieces of evidence indicate that unusual smoking habits can not account for the high death rate from lung can cer among these workers: We have interviewed 320 of the 377 surviving members of the 1942 group. Table 4 gives a sum mary of the smoking habits in this group compared with a sample of men drawn from the general population of 1,121 counties in 25 states.7 The un ion sample is somewhat inadequate since it does not include the men who died and does not in clude all of the present living members of the . union. Nevertheless, it shows that a substantial proportion of asbestos workers never smoked ciga rettes regularly. Certainly the 632 men in our analy sis of death rates were not all heavy cigarette smokers. In the general male population, lung-cancer death rates are about ten times as high among cigarette smokers as among nonsmokers; and the W ar rrr. if '%* Ar--; , 144 Vgt 188, No ASBEiSTOS EXPO SU RE-SELIK OFF ET AL 2S Table 4 .--Sm oking HabitsiIWP**I3200 Abestos W orkers Exposed to Asbestos W 2 0 Y ears or Longer Com pared With Sam ple of Men From the General Population*. ________ __ Age 4049 50-59 60*9 70+ Sm oked Newer Smoked Pipe, Cigar, Never Cigarette ' -, n SVfSwEIMfS?i --- ------- -- ---- General " Population, Asbestos ---- --------- ----------------------- Asbestos Workers. rnnS S m i nw lH lllr %% 9.3 18.8 4.6 SCI " 143 l 6.1 20.3 23.6 10.1 3" 16.2 - 25.5 37.1 11.8 5 1 3 ' 62.7 Sam ple of m en from general population as reported by Hammond and Garfinkel.'1 Sonerai P ep ttt|tiea, 73.7 70.2 S5 tS death rate from lung cancer increases greatly with die amount of cigarette smoking. However, a large proportion of all me in the United States have a history of regular cigarette smoking. From data in a prospective study on smoking, it may be esti mated that if all men smoked a pack or more of cigarettes a day (ie, if all the nonsmokers, cigar smokers, pipe smokers, and light cigarette smok ers had, instead, been heavy cigarette smokers) the lung-cancer death rate would be approximately 3.4 times as high as it is at this time. From this we may conclude that even if all our asbestos workers had smoked a pack or more of cigarettes a day (and, indeed, from our sample we know they did not), and if exposure to asbestos were of no significance, then their lung cancer death rate would have been about 3.4 times as high as the rate in the general US male popula tion. Clearly, the smoking -habits of the asbestos workers cannot account for the fact that their lungcancer death rate was 6.8 times as high as that of white males in the general population. Gastrointestinal Cancer.--Rather to our surprise, the death rate from cancer of the stomach and the death rate from cancer of the colon and rectum were higher among the asbestos workers than would be expected from the rates reported for the US white male population, calculated in the same way as for lung cancer. Twelve deaths from gastric cancer occurred among the asbestos workers, as compared w ith only 4,3 expected. Seventeen deaths from cancer of the colon and rectum occurred among the asbestos workers, as compared with 5.2 expected. Cancer of AU Other Sites.--The combined death rate from cancer of all sites other than lung and pleura, and stomach] 'colon, and rectum was not increased. Twenty-one such deaths occurred among asbestos workers, as compared with 20.5 expected. Asbestosis.--O f the 255 deaths, 12 were due to asbestosis ( pulmonary insufficiency, cor pulmo nale). The lapsed time from first asbestos exposure to death, from asbestosis averaged 45.8 years, with a range of 32 to 59 years. Comment Carcinoma of the Lung.--The results with regard to carcinoma of the lung are clear. Industrial ex posure to asbestos by insulation workers, as stud ied here, results in a marked increase in the inci dence of cancer of the lung, approximately six to seven times the expected incidence. Altogether, 45 (17.6%) of 255 men w ith more than 20 years elapsed since the onset erf exposure died of cancer of the lung or pleura. i These data do not give the "incidence of cancer of the lung in asbestosis." They relate to the spe cific conditions of our investigation: to a group of men with only intermittent exposure to materials containing limited amounts (often 2% to 20%) of asbestos under working conditions varying from very dusty, as in extracting old insulation in closed quarters, to those with little dust exposure, as in building constraction in open air. Moreover, they relate to the relatively recent past, in a trade with the shorter work week of the strong building trades unions, in an era when industry has been aware of potential asbestos hazard and the working popu lation has had some consciousness of potential risk associated with dust exposure. These data would not necessarily apply to asbestos exposure in other industries, such as the factory production of asbestos products, the asbestos textile industry, etc, where conditions of employment might be quite different. Our results do not contradict the even higher incidence of lung cancer suggested in other studies3; they are merely a shade less strik ing. Diffuse Pleural and Peritoneal Malignancy-- Mesothelioma.--Determining the incidence of dif fuse pleural mesothelioma is complicated by the insecurity of its histological verification. While some pathologists will so categorize a high propor tion of diffuse pleural tumors, in the experience of others it is a very rare tumor and may be mim icked by anaplastic peripheral carcinoma of the lung and diffuse fibrosarcoma of the pleura. It is difficult to evaluate completely the published re ports of diffuse pleural mesothelioma in asbestosis in the absence of complete details of each case. To the present time, there has been no informa tion concerning the incidence of diffuse pleural mesothelioma in asbestosis, since the published cases are reported without reference to a total pop ulation in which they occur. Nevertheless, the 145 ` .*<-7.' - i f i ' - 1* / . # 26 ASBESTOS EXPOSURE--SELIKOFF ET a z W JAMA, April 6, 1964 growing number of reports of individual cases suggests that these tumors are perhaps becoming relatively frequent complications of asbestos ex posure. . Our observations in this series are similarly sug gestive. In three of the 255 deaths among the men who had worked for 20 years or more, the exam ining pathologist considered the death due to dif fuse pleural mesothelioma, and in the two eases in which we have bear able to review the histological material, the histological appearance was that of ten so categorized, and asbestos bodies were pres ent. This incidence of more than 1% of deaths from pleural mesothelioma is strikingly high for a tumor which is generally considered to be extreme ly rare. In one ease in our series pathological ex amination suggested diffuse peritoneal mesotheli oma. This single experience is too fragmentary for evaluation. ^ Gastrointestinal Carcinoma.--Isolated instances of gastrointestinal carcinoma in the presence of as- bestosis have been known, but there have been no data to indicate that these were more than coinci dental findings. Among the asbestos workers studied here, cancer of the stomach, colon, and rectum was three times as frequent as expected. These data suggest that there may perhaps be an etiological relationship between industrial asbestos exposure and carcinoma of the gastrointestinal tract. Environmental Asbestos Exposure.--The recent demonstration, by South African and British * in vestigators of pleural and peritoneal neoplasms among individuals who had chance environmental exposure to asbestos many years before raises the very important question of possible widespread carcinogenic air pollution. The possibility of en vironmental exposure has long been known. Soon after the initial clarification of asbestosis as a clinic entity, H addow 10 demonstrated asbestos bodies in a man not employed in the industry but living next door to an asbestos factory. This finding was later mirrored in the finding of chronic? beryllium dis ease among rsidants of a community near a beryl lium factory." W hat is new, however, is an appre ciation of the potential extent of the problem. Thomson and associates ** have reported the fre quent findings of asbestos bodies in the lungs of urban dwellers. Among 6,312 individuals x-rayed in an area about an asbestos mine in Finland, Kiviluoto19 found 499 cases of pleural calcification of the type characteristically seen among asbestos workers, without obvious cause. In a comparable area without any asbestos mine, no cases were found among 7,101 persons x-rayed. It should be noted that these were not people who worked in the mine--none did--but, rather, were farmers, housewives, and others who lived in the general location. In one subject who came to autopsy, polarized-light microscopy deirmnsh-tod asbestos fibers in the lung. Similarly, the lung of a cow graz ing near the mine also showed the presence of asbestos. A particular variety of environmental exposure may be of even greater concern. Asbestos exposure in industry will not be limited to the particular craft that utilizes the material. The floating fibers do not respect job classifications. Thus, for ex ample, insulation workers undoubtedly share thenexposure with their workmates in other trades; intimate contact with asbestos is possible for elec tricians, plumbers, sheet-metal workers, steamfitters, laborers, carpenters, bailer makers, and fore- be m ^ d e T even the suPerv*sing architect should 1 E 100th St, New York 10029 ( Dr. Sefikoff). This study was supported by th e Health Research Coun cil of the City of New York. Cooperating in this investigation w ere th e executive of ficers of th e International Association of H eat and Frost Insulators and Asbestos W orkers, W ashington, DC, and th e officers and membership of the N ew York and Newark, NJ, locals of this Union. References 1. Lynch, K.M., and Smith, W.A.: Pulmonary Asbestosis: Carcinoma of Lung in Asbesto-Silicosis, Amer J Cancer 24:56-64 (May) 1935. 2. Annual Report of Chief Inspector of Factories for Year 1955, London: H er Majesty's Stationery Office, 1956, cmnd 8, p 206. 3. Doll, R.: M ortality From Lung Cancer in Asbestos W orkers, Brit J lndustr Med 12:81-86, 1955. 4. Bohlig, H ., and Jacob, G .: Neue G esichtspunkte liber den Lunge:ikrebs der A sbestarbeiter, Deutsch M ed Wschr 81:231-233 (Feb 17) 1956. 5. Braun, D.C., and Truan, T.D .: Epidemiological Study of Lung Cancer in Asbestos Miners, AM A Arch lndustr Health 17:634-653 (June) 1958. 6. W agner, J.C .; Sleggs, C.A.; and M archand, P.: Diffuse Pleural M esothelioma and Asbestos Exposure in N orth W estern Cape Province, Brit J lndustr Med 17:260-271, 1960. 7. Hammond, E.C ., and Carfinkel, L.: Smoking H abits of Men and Women, J Nat Cancer Inst 27:419-442 (Aug) 1961. 8. Hammond, E.C., and Horn, D .: Smoking and D eath Rates: Report on Forty-Four M onths of FoHow-up of 187, 783 Men. I. Total M ortality; II. D eath Rates by Cause, JAM A 166:1159-1172 (March 8); 1294-1308 (March 15) 1958. 9. McCaughey, W .T.E.; W ade, O .L.; and Ebnes, P C .: Exposure to Asbestos D ust and Diffuse Pleural Mesothelio mas, Brtt Med / 2:1397 (Nov 24) 1962. 10. Haddow, A .C , in Report of Annual M eeting of British M edical Association, M anchester, 1829, Lancet 2.236-231 (Aug 3) 1929. 11. Chesner. C .: Chronic Pulmonary Cranuloinatosis in Residents^ f Community N ear Beryuium Plant: Three Autopsied Cases, Ann Intern Med 32:1028-1048 (June) 1950. 12. Thomson, J.G.; Kaschula, R.O.C.; and M acDonald, R-R-: Asbestos, as M odem U rban H azard, S Afr Med J 37:77-81 (Jan 19) 1963. 13. Kiviluoto, R.: Pleural Calcification as Roentgenologic Sign of Non-Occupational Endem ic Anthnphyllite-Asbestosis, Acta Radial Suppl 194, pp 1-67, 1960. .