Document pBxkJa6kZp9aLq2VNZkN189GB

A Retrospective Mortality Study of Polychlorlnited Biphenyl Manufacturing Workers* Judith A. Zick, H.P.H.Z Oivld C. Musch, M.P.H., Ph.0.3 ^Monsanto Company Department of Medicine and Environmental Health ^Present address: 1751 Legend Lane St. Louis, MO 63141 ^Present address: Department of Ophthalmology University of Michigan Medical School STZ512 Scott Turner Bldg., Box 004 1010 wall Street Ann Arbor, MI 48109 HONS 020967 Abstract Judith A. Zack/Page 2 A retrospective cohort mortality study of workers exposed to polychlori nated biphenyls (PCBs) during production of the compounds Is reported. A total of 89 workers exposed to PCBs for a minimum of six months during the period from 1945 to 1965 were followed through 1977, and their vital status was deter mined. Comparison of this PCB-exposed cohort's mortality experience with age-, race*, and cause-specific mortality rates of the U.S. male population was performed. Observed deaths were not statistically greater than expected for any of the cancer sites examined, ho deaths were observed for cancer of the liver or malignant melanoma. A statistically significant excess of circulatory disease deaths, exclusive of arteriosclerotic heart disease, was observed In white males. HONS 020968 Judith *. ZacJt/Pjje 3 The ten* "polychlorinated biphenyls" (PCBs) describes 1 group of mined chlorinated biphenyls which vary in consistency fro* light, free-flowing liquids to crystalline and non-crystal line solids, depending on the degree of chlorination, PCBs have been In commercial use since 1929. Their dielectric properties, resistance to oxidation, acids, bases, and other chemicals, and thermal stability have made them useful for many applications, particularly in the production of capacitors and transformers, and In the Insulation of elec, trie cables and wlres. Reports of adverse health effects in man Initiated from several incidents of accidental heavy exposure to PCBs. A process change in a chemical plant Introduced an unspecified PCB conpound (Arochlor) Into the work environment, resulting In exposure of 14 workers to estimated breathing zone PCB levels of 0.1 mg/m^.l Chloracne developed in one-half of the exposed workers, and liver function tests showed a borderline abnormality In one of the seven affected workers. After 13 months of follow-up, the chloracne had disappeared, but the liver function abnormality, although Improved, remained. Accidental Ingestion of rice oil contaminated with PCbs has resulted in several outbreaks of common symptomatoloty, reported from Japan ("Yusho" disease}? and Taiwan ("Yu-Gieng" disease).3 The most common acute synptoms observed were hyperplgmentatlon and acne-llke lesions, discharge from the eyes, neuroendocrine disturbances, emesis, and diarrhea. There was a dose-response relationship between the amount of oil ingested and the proportion of persons reporting symptoms. After six years, many of the patients still reported such symptoms as headache, stomach pain, nunbness of the extremities, arthralgia, and respiratory symptom.4 Although massive Ingestion of PCBs occurred In these incidents, analysis of the cooking oil In the Yusho episode showed that HONS 020969 Judith A. Zack/Page 4 polychlorinated dlbemofurans^ and polychlorinated quaterphenyls8 were cocontamlnants. Analysis of blood samples from patients in both Incidents, as well as liver and adipose tissue from Yusho patients, showed the presence of these compounds along with PCBs.7* 8 These co-contaminants obscure the rela tionship of PCBs to the symptoms resulting from Ingestion of the contaminated rice oils. The Yusho Incident served as a stlwlus to a large (umber of toxicologi cal studies dealing with the health effects of PCBs. Chronic toxicity studies of PCBs In mice and rats have shown hepatic effects ranging from hyperplasia to wel 1 -dlfferentlated hepatocellular carcinomas following the Ingestion of various PCB-contalnlng diets.9* 10 Studies of hunan exposure to PCBs have shown that boty burdens are directly related to the level of exposure to PCBs.33* 13 Morbidity studies have focused on the effects of PCBs on the skin, liver, blood chemistry, and other systems.13* 16 Chloracne, dermatitis, mild liver func tion abnormalities, and Increased triglycerides levels were reported with some consistency from these studies. Mortality studies of occupational'y-exposed Individuals have not yielded consistent results. A preliminary report37 addressed an Increase In deaths due to malignant melanoma (2 obs. vs. 0.04 exp.), which occurred In a group of Si research and development and 41 refinery workers at a Hew Jersey petrochemical plant. Another author coamented on problems In defining exposures In this study.38 in a study of 2567 capacitor plant workers exposed to PCBs,39 excesses of liver and rectal cancer deaths, which were not statistically sig nificant, were reported. Although few deaths were available for analysis by duration of exposure and latency, there was no trend by these measures for liver cancer deaths, and a slight Increase by latency but not by length of MQNS 020970 Judith A. Zack/Page 5 exposure for rectel cancer deaths. Another study of 1310 workers from a capac1 tor-manufacturing plant with at least six months enployment between 1946 and 1970 found statistically significant excess deaths from all cancers (In males) and all causes (In females).2 Suggestive, but not statistically significant, excesses were noted for deaths due to digestive system cancers (In males) and malignancies of lynphatlc and hematopoietic tissue (In males and females). Mo deaths were attributed to liver cancer or malignant melanoma. Oue to the rela tively young age of this worker cohort and the resultant small nunber of deaths (n * 27), site-specific excesses were based on two or three deaths, therefore, further studly of this cohort was proposed by the authors. At about the time of the Initial report dealing with the mortality experi ence of an occupationally-exposed worker cohort, a preliminary review of the mortality experience of Monsanto workers Involved In the production of PCBs was carried out. The present study represents a completed analysis of this worker cohort's mortality, with particular attention given to overall cancer, liver and pancreatic cancer, and malignant melanoma. Materials and Methods During the period from 1936 to 1977, PCBs were manufactured at Monsanto Company's Sauget, Illinois plant. Throughout this period, the process of manufacture remained relatively unchanged. This process. In brief, involved the batch chlorination of biphenyls In the presence of Iron and Iron chloride catalysts. The crude mixture produced In the chlorlnators was then aerated, distilled, earth treated, filtered, and transferred to storage tanks before shipping. While no Industrial hygiene monitoring data were available to quan tify the PCB-exposure levels during the stuqy time, Interviews with plant personnel established the fact that PCB exposure levels did not vary consider, ably within the PCB department. Therefore, all workers In this department HONS 020971 Judith A. Zaclc/Pagc 6 were considered to have a coninon exposure to PCBs, as well as other chemicals used In the department, Including hydrochloric acid, trl- and tetrachlorobenien, biphenyl, and chlorine gas. All male employees from the hourly roll who worked In the PCB department for at least six months between January 1, 1945 and December 31, 1965 were selected for the study- By use of the plant's co^uterlzed work history system, union records, and manual work history verification, 89 male, hourly employees met the criteria for Inclusion in the study- The vital status of 98.9t (88) of the cohort was determined as of December 31, 1977, using standard follow-up techniques. Table 1 smnmarlzes the results of the vital status tracing. A death certificate was obtained for each person found to be deceased, and coding of the underlying cause of death was performed by an experienced nosologlst, using the International Classification of Diseases, Adapted, Eighth Revision.21 Using a computer program developed for such an appllcatlon,22 the observed number of deaths among PCB workers was compared with those expected based on age-, race-, and cause-specif1c mortality rates of U.5. males. To compute ex pected numbers, person-years of observation were allocated to five-year age and calendar time categories and nultipi led by corresponding race- and causespecific death rates for U.$. males. The resulting quantities were summed over all ages and years to obtain the total expected ntnbers- Ratios of ob served to expected nwber of deaths were expressed as standardized mortality ratios ISMRs). Significance testing of the SHRs was performed at the 5i level, assuming that the observed ntaber of deaths follow a Poisson dlstrlbutlon.23 Results Of the 88 PCB workers for whom vital status tracing was successful, 58 were verified living and 30 were verified deceased. Death certificates were HONS 0209 72 ( Judith A. Zack/Page 7 obtained for each of the decedents. The person-years of observation contri buted by the entire worker cohort ire shown In Tible 2. A totil of 1800.1 person-ytirs were observed it ill 191s, with the majority observed In the mlddle-iqe range (iges 35-60 yrs. contributed 74*, or 1333.1 person-yeirs to the total) Cross-classification of PCB exposure duntlon with vltil stitus (Table 3) shows thit the majority of both living ind deceased workers were exposed less thin three years. The avenge length of exposure to PCBs for living and deceased workers was similar, with deceased workers having slightly more exposure time (3.7 yrs. vs. 3.2 yrs., respectively). Cause-specific SMRs are listed In Tables 4 and 5 for the 60 white and 29 nonwhite male members of the PCB study cohort. Deaths due to malignant neo plasms, as Indicated In Table 4, accounted for 271 (8) of the total nunber of deaths. No cancer deaths were attributed to malignant neoplasms of the liver or pancreas, or malignant melanoma. While many of the cancer-specific SMRs exceed 100, none are statistically significant. In white males, lung cancer accounted for 751 of the deaths due to malignant neoplasms, while no single cause predominates In nonwhite males. Table 5 presents a brief summary of the eight cancer deaths In the PCB worker cohort. Among the non-mallgnant causes of death (Table 6), diseases of the cir culatory system accounted for almost half of the deaths. For the category of circulatory system disease, exclusive of arteriosclerotic heart disease, the SMRs for white males and for the total cohort were significantly greater than 100. The excess In deaths due to this category Is accounted for solely by the experience of the white males, since nonwhite males shewed an SHR of 93. No other causes of death In Table 6 showed observed deaths which were statistically MQNS 020973 Judith A. Zack/Page 8 greater than expected. The overall SMR for malignant and non-malIgnant causes was 131 (30 observed, 22.BS expected deaths). Little difference was found In this overall SMR by race (for whites, 133; for nonwhltes, 12B). 01scusslon This Investigation dealt with the mortality experience of a cohort of 89 male workers involved for a least six months In the production of PCBs between January 1, 1945 and December 31, 1965. The only cause of mortality which showed a statistically significant excess In observed deaths was circulatory diseases exclusive of arteriosclerotic heart disease. This excess was apparent only for white males. Circulatory diseases In this category Included deaths from rhetmatlc heart disease, cerebrovascular disease, and other forms of heart disease. If this excess cause of death Is not an artifact, one would postulate that there may also be a relation between risk factors of these circulatory diseases and PCB exposure. Blood pressure, a risk factor (when elevated) of coronary heart disease (CHD) and cerebrovascular disease, has been related to blood PCB levels In two studies, with conflicting results. One study24 found that serum PCB levels made a statistically significant contribution to explaining the variability of dlastpl1c (but not systolic) blood pressure measurements in multiple regression analyses of data from a comounlty exposed to DDT and PCBs, while another studyfound no such association. Studies of lipids, which have been associated with CHD and (with less consistency) with cerebrovascular disease, have considered total cholesterol, high density lipoprotein (HDD, and triglyceride levels In relation to blood PCB levels. While no consistent association has been shown between blood PCBs and total cholesterol or HDl levels, most,13* 16 but not all,24 studies have shown increased triglyceride MQNS 020974 Judlth A. Zack/Page 9 levels with Increased blood PCB levels. Given the lick of support for an association between PCBs exposure and non-ASHD circulatory system deaths from previously reported mortality studies, and the Inconclusive relation between cardiovascular risk factors and blood PCBs, this excess cause of death xust be viewed as a preliminary finding, in need of support from other studies. While previous mortality studies as well as toxicological studies have Implicated a variety of cancers - digestive system, liver, pancreas, malignant melanoma - as possible results of exposure to PCBs, the present study provides no support to such findings. In the Monsanto worker cohort, no deaths were attributed to liver or pancreatic cancer, nor were any malignant melanomas found. The overall excess in cancer deaths, while not statistically signifi cant (8 observed, 4.46 expected, SMR 179), Is mainly a result of respiratory cancer deaths In workers for whom smoking habits were not ascertained. Although the present study's results do not support a carcinogenic effect associated with PCB exposure, there are significant limitations to the results. A principal limitation Is the site of the worker cohort, which prevents the detection of increases In risk to deaths from rare diseases. A such larger cohort followed for a longer period would be necessary to detect a two-fold increase, for example. In observed liver cancer deaths. In order to allov for a latency period, the study cohort was limited to those working at least six months until December, 1965. Since PCB production continued until 1977, further follow-up of the mortality of Monsanto's PCB-exposed workers will have larger numbers, as well as longer follow-up time. Another limitation common to retrospective studies of occupational cohorts is lack of knowledge of exposures from other work-related activities. While personal habits, such as smoking and drinking, can sometimes be roughly ascer tained if co-workers or family members are available for Interview, previous MONS 020975 Judith A. Zack/Page 10 exposures, or concurrent occupational exposures during follow-up, are often Inpossible to assess. The connection of an effect to a previous PCB exposure must first consider other potential effect-related exposures. A useful cooparlson group In occupational mortality studies Is an unex posed industrial cohort, fron the sane plant. If possible. Ascertainment of mortality of such a cohort allows for coaparlson of SHRs between exposed and unexposed cohorts, which reduces the possibility that an elevated cause of death In the exposed cohort Is a result of a generally greater than expected cause-specific mortality In that area or Industry. Such a conparlson group would have been useful In evaluating the elevated mortality due to non-ASHD circulatory system diseases found In this study* Given the size of the study cohort and the other limitations described above, the results of this analysis of PCB-exposed workers' mortality experi ence are not definitive, but rather contribute to the health-effects research on PCBs. A future study of this cohort could evaluate the mortality of workers Involved fron the Initial through the final year of PCB production, and thus have more power to detect any possible hazards of PCB exposure. MGNS 020976 Judith A, Zack/Pige jj References 1. Meigs. J.W., Albom, J.J., Kartln, B.L, Chloracne from an unusual expo sure to Arochlor. J.A.M.A. 154:1417-1418, 1954. 2. Kuratsune. M.. Yoshlnwra, T., Metsuzaka, J.. Yamaguchl, A. Epldemloloty study on Yusho, a poisoning caused by Ingestion of rice oil contaminated with a commercial brand of polychlorinated biphenyls. Environ, fealth Persp. 1:119-128, 1972. 3. Chang, K.J., Cheng, J.S. Huang, P.C., Tung, T.C. Study of patients with PCS poisoning. J. Formosan Med. Assoc. 79:304.313, 1980. 4. Uribe, H., Koda, H., Asahl, M. Present state of Yusho patients. Ann. N.Y. Acad. Scl. 320:273-276, 1979. 5. Nagayama, J., Kuratsune, M., Masuda, Y. Determination of chlorinated dlbenzofurans in Kanechlors and `Yusho oil.* Bull. Environ. Contain. Toxicol. 15:9-13, 1976. 6. Kamps, L.R., Trotter, W.J., Young, S.J., Carson, L.J., Roach, J.A.G., Sphon, J.A., Tanner, J.T., McMahon, B. Polychlorinated quaterphenyl s Identified In rice oil associated with Japanese *Yusho" poisoning. Bull, Environ. Contain. Toxicol. 20:569-591, 1976. 7. Kashlmoto, T., Mlyata. H., Kunlta, N. The presence of polychlorinated quaterphenyls In the tissues of Yusho victims. Food Cosmetics Toxicol. 19:335-340, 1981. 8. Kashlmoto, T., Mlyata, H., Kunlta, S., Tung, T.C., Hsu. S.T., Qiang, K.J.. Tang, S.Y., Ohl, 6., Nakagawa, J., Yamamoto, S. Role of polychlorinated dlbenzofuran In Yusho (PC6 poisoning). Arch. Environ. Health 36:321 326, 1981. HONS 02097? Judith A. Zack/Page 12 Kimbrough, R.O., Linder, R.E., Gaines, T.B. Morphological changes In livers of rets fed polychlorinated biphenyls. Arch. Environ, fealth 25:354-364, 1972. Kimbrough, R.D., Squire, R.A.. Linder. R.E., Strandberg, j.o,, H>ntal1, R.J., Burse, v.W. Induction of liver tumors In Sierman strain female rats by polychlorinated biphenyl Arochlor 1260. J. National Cancer lnst. 55:1453-1459, 1975. Ouw, H.K., Simpson, G.R., Slyall, O.S. Use and health effects of Arochlor 1242, a polychlorinated biphenyl, In an electrical Industry. Arch. Environ. Health 31:189-194, 1976. Maronl, H., Colombl, A., Cantonl, S., Ftrloll, E,, Foa, V. Occupational exposure to polychlorinated biphenyls In electrical workers. 1. Environ, mental and blood polychlorinated biphenyls concentrations. Brit. J. industr. Med. 38:49-54, 1981. Chase, K.H., Wong, 0., Thomas, 0., Barney, B.W., Simon, R.K. Clinical and metabolic abnormalities associated with occupational exposure to polychlor. Inated biphenyls (PCBs). J. Occup, Med. 24:109-114, 1982. Flschbeln, A., Wolff, M.S., Lllls, R.t Thornton, J., Sellkoff, l.J. Clinleal findings among PCB-exposed capacitor manufacturing workers. Ann. N.Y. Acad. Scl. 320:703-715, 1979. Maronl, N., Colombl, A., Arbostl, G., Cantonl, S., Foa, V, Occupational exposure to polychlorinated biphenyls In electrical wrkers. 11, Health effects. Brit. J. Industr. Med. 3B:55-60, 1981. Baker, E.L., Jr., Landrlgan, P.J., Glueck, C.L., Zack, M.M., Jr,, Liddle. j.A., Burse, Y.W., Houseworth, W.J., Needham, L.L. Metabolic consequences of exposure to polychlorinated biphenyls (PCB) In sewage sludge. Am. j. Epidemiol. 112:553-563, 1980. MONS 020978 Judith A. Zack/Pig* 13 17. Bahn, A.K., Rosenwalke, 1., Herrmann, N., Grover, P., Stellman, j., O'Leary. K. Melanoma after exposure to PCBs. N. Engl. j. Med. 295:450, 1976. 18. Lawrence, C. PCB7 and melanoma. N. Engl. J. Med. 296:108, 1977. 19. Brown, D.P., Jones, M. Mortality and Industrial hygiene stu<^ of workers exposed to polychlorinated biphenyls. Arch. Environ. Health 36:120*129, 1981. 20. Bertanl, P.A., Zocchettl, C., Guercllena, S., Foglla, M.O., Pesatorl, A., Rlboldl, L. Mortality study of male and female workers exposed to PCBs. lnt. Symposttm on Prev. of Occup. Cancer, Helsinki, Finland, April, 1981. 21. Eighth Revision, International Classification of Diseases, Adapted for Use In the United States. U.S. Department of Health, Education, and Welfare, Public Health Service, PHS Publication No. 1693. Washington: U.S. Govern* ment Printing Office, 1977. 22. Monson, R.R. Analysis of relative survival and proportional mortality. Comput. Blomed. Res. 7:325*332, 1974. 23. Ballar, J.C., Cderer, F. Significance factors for tne ratio of a Poisson variable to Its expectation. Biometrics 20:639-643, 1964. 24. Krelss. K., Zadk, M.M., Kimbrough, R.D., Needham, L.L., Smrek, A.L.. Jones, B.T. Association of blood pressure and polychlorinated biphenyl levels. J.A.M.A. 245:2505*2509, 1981. 25. Btmgamer, J.E., Hammer, D.I., Coluccl, A.V., treason, J.P., Flnklea, J.F. Polychlorinated biphenyl residues In refuse workers. Research Triangle Park, North Carolina, National Institute of Environmental Health and Safety, P'ibllc Health Service, Department of Health, Education, and Wel fare, 10 pp., 1973. MONS 020979 Judith A. Zack/Page 14 Tibi* 1 follow-up Status of Polychlorinated Biphenyl Workers as of December 31, 1977 Follow-up Status Nieaber Status known B8 Verified alive 58 Verified deceased 30 Death certificate found 30 Death certificate not found 0 Status Unknown 1 Total 89 MONS 020980 Judith A. ZicJc/P4ge is Table 2 Person-years of Observation Contributed by Polychlorinated Biphenyl Workers by Age *9* In Years Person-years of Observation 20-24 11.4 25-29 63.8 30-34 142.6 35-39 222.5 40-44 2B3.2 45-49 325.3 50-54 294.5 55-59 60-64 207.6 134.0 65+ 115.2 Total 1800.1 HONS 020981 Judith A. Zack/Page 15 Table 3 toratlon of Exposure to Polychlorinated Biphenyls by Vital Status Years of Exposure* Living (a 58) Vital Status Deceased -.30) Unknown (n * 1) <1 20 5 1 1-3 20 12 0 3-5 6 4 0 >5 12 9 0 Total {n - 89) 26 32 10 21 Average length of Exposure 3.2 yrs. 3.7 yrs. 0.7 yrs. 3.3 yrs. lifetime exposure prior to 1-1-66 MQNS 020982 Judith a. Uck/r*$* i? Teble 4 Observed end Expected Deeths Ouc to Neltgnent NooplOMM by Rice, PC# Morter Cohort I.C.O. No. (Eighth Rev.) 140-209 140-149 150-159 Ceuse of Booth All Nellgnent neopleees uctei cevlty 6 pherynx Olgestlve orgens * peritonei* Uhlte holes 0 E SNR* 4 2.70 14* 0 0.09 - 0 0.75 - Nonwhite Miles 0 E SNR* 4 1.76 227 0 0.07 - 1 0.5* 172 155, 155 150-154, 157-159 Liver Other digestlvo orgens 0 0.05 0 0.70 - 0 0.05 1 0.53 1*9 ICO-Id 152, 163 Ruplrotory systeo Lung 3 0.94 3 0.69 319 337 1 0.59 169 1 0.55 1*2 160, 161 Other resplretory ergons 0 0.05 - 0 0.04 - 1*5-1*9 Genltourinery orgens 0 0.2* - 1 0.23 435 200-209 164-164, 190-199 Lyophetlc 6 hoMtopoletlc tissue All other sites n 0.2* 1 0.36 270 1 0.12 *33 0 0.17 _ *No SNR Mt stettsttcelly slgnlfltent (Oil p-velues greeter then 0.05) lotel 0 E SNR* 4.46 179 0 0.16 - 1 1.33 0 0.10 1 1.23 75 *1 4 1.53 261 4 1.44 27* 0 0.09 1 0.51 196 1 0.40 1 0.53 250 1*9 Moms 020983 Judith A. lit*/?Kjt te Pace Ronuhlte ttonwhlta Noiwhlte White White White Nonwhlte White Year of Birth 1909 19)3 1909 1093 191? 1916 1907 1906 Table S Cate Suwoorie* of Cantor Deaths Anong Polychlorinated Biphenyl Workers Year of first Exposure Year of Oeoth 9CB Exposure Duration (*.) 194? 1959 13 1954 1965 65 1951 I960 1943 1963 195) I960 1961 1977 1944 1964 1940 1957 34 2B 33 B 79 Oeoth Certificate Cause of Oeoth (ICOO CortlnoM of colon (IS3.B) CartlnaaatotitPrloary site, kidney (1B9.0) Card now Of lung (16?.l) Cerdnnu of right lung (16?. 1) Corclnooi* of lung Cord none of lung (16Z.1) Multiple Myeloma (Z03.0) Metastatic corclnaeucarclnoaotosls (199.0) HONS 020984 Judith A. lack/Page 14 Table i i.e.n. no. (Eighth Rev.) 390-458 410-413 390-409. 414-458 4(0-519 570-577 800-998 - Observed tnd expected Deaths Due to Non.uallgnant Causes by bice, PC# Worker Cohort Cause of Oeath White Hales 0 E SMR ttomrhlte Hales 0 E SMR Diseases of the circula tory systen Arteriosclerotic heart disease It 6.87 161 4 4.99 80 5 4.35 115 1 7.70 116 All other circulatory systen 7 1.81 576* 7 7.15 91 Diseases of the respira tory systen 0 0,73 - 1 0.54 185 Diseases of the digestive systen 1 0.75 113 1 0.45 777 fitemal causes of death All otter noft-MMgAAnt cium 0 1.41 * 7 1.10 107 1 0.95 105 0 1.30 - Total 0 E 5HR 16 11. 17. 143 7 7.19 97 9 1.98 776* 1 1.77 79 7 1.70 1 7.38 161 47 7 7.40 83 T0l*t All causes, neltgnent 1 non-nallgnent 18 13.51 133 17 9.15 178 10 77.88 131 *p < 0.05 020`&5' *o*s