Document qdXwYEVn66mVzrM96K44dREDR

CANCER RISK AMONG RUBBER WO Matti Hakama, lire Kilpikari Department of Public Health, University of Tampere, Tampere, ('inland Studies of the cancer risk timong rubber workers published sincc^4968 arc revJuy&j, tfa total material esummed lompnsed 80.000 workers and 4000 cuncerH^aths. An intceasdd risk ot stomach earner, tuna cancer, bladder cancer, and leukemia was founds The increased risk of cancer at a specific anatomic site was correlated with the mode of expose specific/ for the working area or division. On the other hand, no clear exposure-risk relationship cout/ be demonstrated. It is likely that there has been on increased risk of cancer since (he elimination of or- and frnuphtlnfammc h\ 1950, flnwewr, the ctiologir fraction is less than that due to prevalent exposures mtreadng the risk ot common cancers at specific anatomic sites. INTRODUCTION The rubber industry is a typical example of the chemical industries. Epidemiologic evidence lor an inti eased cancer tisk in the tubber industry was reported by the British Registrar General in 1934; an estimate of 10% was given for the excess risk of total deaths (General Register Office, 1934). This increased risk was attributed to cancer. The U.S. Department of Health, Education, and Welfare (1961-1963) reported an 8% excess of cancer deaths in male rubber workers dying in 1950. In their classical epidemiologic study, Case and Hoskcr (1954) investigated bladder cancer as an occupational disease in the rubber industry in England and Wales. The assumed carcinogen, (3-naphthyl,tminc, was largely eliminated from the production process by 1950. MATERIAL USED In this paper we review more recent studies. Few reports were published in the late 1950s and early 1%0s; the years of publication of the reports reviewed here range from 1968 to 1979. By restricting the discussion to relatively recent data, we attempted to examine present exposures within the rubber industry. However, some of the material originates from the 1940s, and the results may exaggerate the cancer risks in the rubber industry in the last few years. Nevertheless, there is a long latent period between the first carcinogenic exposure and the diagnosis of cancer, preventing the use of very recent data only. The studies reviewed here originate from England, the United States, and Finland. The published reports arc partly based on the same material. We have selected studies so that the same results are covered in this review only once. Altogether, 12 studies were accepted after this process or selection (Andjclkovich ct al., 1976, 1977; British Rubber Manufacturer's Association, 1976; Fox and Collier, 1976; Fox et al,, 1974; Kilpikari, 1979; McMichael ct al., 1974, 1976a, 1976b; Manson and Nakano, 1976a, 1976b; Monson and Fine, 1978) (Fig. 1); the total material was covered by six of these (Table 1). The total material comprised 1211(20t| CopyriRhi ( i 1080 by Hemisphere Publishing Corporation v I Hi 1212|292| FOLLOW UF in*r 1470- M, HAKAMA ANQ KIlPlKARl itt<M T ir<* vfiof <1) MofUMiy man <2) Modality white mi <3| loctdaoc* and mortality, procotimf and paan ii IMHwatwm 14) MuftaMy mimni FIGURE 1. Diagram of the materials reviewed. The numbers refer to individual studies: (1) Andjelkovich ct al. (1978); (2) Andjelkovich ct al. (1976); (3) Andjelkovich et al. (1977): (4) BRMA (1976); (6) Fox and Collier (1976); (7) Fox cl al. (1974); (9) Kilpikari (1979); (14) MeMichael et al. (1976a); (IS) McMichael ct al. (1976b); (16) MeMichael et al. (1974); (18) Monson and Nakano (1976a); (19) Monson and Nakano (1976b). 73,433 male and 7533 female rubber workers. Males only were included in a more detailed study. The time of follow-up ranged from 1940 to 1974 and the length of follow-up from 4 to 30 yr. As a rule, Ihc results were reported in terms of the observed and expected numbers of deaths transformed into standardized mortality ratios (SMRs). The total number of deaths TABLE 1. Standardized Mortality Ratios for All Deaths and All Cancers among Male Rubber Workers in Different Studies SMlt Study All deaths At! cancers Size of cohort Years of foliow*up Fox and Collier, 1976 Fox et al., 1974 BRMA, 1976 MeMichael et al., 1976a Monson and Nakano, 1976a Kilpikari, 1979 Total 97 IIS 40,867 i 102 111 40.867 3 88 98 37,221 20 94 100 18,903 9 88 94 13,571 31 70 100 1,331 22 90 97 73,433 "For the total SMR. OU 7122 I. KILPIKARI 'Iwlkovich ; (6) Fox (15} ) Monson Ictailed from 4 CANCER RISK AMONG RUBBER WORKERS 1213(293) among males was 12,786, 3156 of which were due to cancer. The number of cancer deaths among females was 1114. The observed and expected numbers of deaths in these 12 studies were pooled, and SMRs for the total material by determinants of risk are reported here. Consequently, the SMRs are age-study-adjusted mortality rates by the indirect method. The observed numbers are reported in the tables together with the SMRs. The total number of deaths is not equal to the sum of the deaths in individual categories because in some of the studies the results were not reported in each of the subclassificalion used in this review. RESULTS The SMR for the total material and for all causes of death was 90, which probably indicates a slight bias in the selection into work and in survival in the industry. In other words, the healthy worker effect in the total material is likely to be about 10%. The total risk of cancer deaths was very close to that expected (SMR =97), which indicates a lesser bias for factors related to cancer risk or an excess risk of cancer. The difference between the SMRs for all causes and for cancers is close to the 10% reported in 1934 for the period 1929-1931 by the British Registrar General (General Register Office, 1934). There were differences in the risk of death between the anatomic sites of cancer (Table 2). The SMRs for bladder cancer, stomach cancer, and leukemias were between 110 and 120, whereas the risk of brain tumors was only 70% of that expected. The only statistically significant difference was, in fact, the reduced risk ol lutmns of the central nervous system. The biological significance ol this result is questionable, however, because of the multitude TABLE 2. Standardized Mortality Ratios and Numbers of Deaths due to Site-specific Cancer among 80,966 Rubber Workers4 Male Female ICO no.6 Primary site SMR No. SMR No. 140-239 140-209 150-159 150-151 150-153 150-154 160-163 160-162 174 180-189 180-180 180-185 180-188 191-192 200-203 204-207 Ail neoplasms Malignant neoplasms Digestive organs and peritoneum Stomach Large intestine Rectum Respiratory organs Tfjclicj, bronchus, jmJ lung Breast Genitourinary organs Cervix uteri Prostate Bladder Central nervous system Lymphatic system Leukemia no 97 98 114 10S 91 102 - 102 117 69 101 116 762 3156 364 425 267 MS M0J 204 154 66 51 148 88 299 88 86 155 4 137 ii 103 4 122 24 76 57 107 39 84 2 -- 196 9 102 8 121 18 97 9 ihe data were obtained from tallowing studies: Andiclkovich et al. (1977, 1978); British Rubber Manufacturer's Association (19761; Fox and Collier (1976); Fox et al. (1974); MeMichael et al. (1976a, 1976b); Monson and Nakano (1976a). ^International Classification of Diseases, 8th rev. OLI 7123 1214(294] M. HAKAMA AND I. KILPIKARl of statistical tests carried out and because of the potential effects of selection. The risk of cancer among female rubber winkers was essentially the same as that among males (Table 2), The small excess risk in the selected primary sites should be correlated with the length of exposure. The original studies do not permit a mutually exclusive classification by length of exposure. Instead, we estimated the risk for those exposed for more lhan 1, 5, 10, and 25 yr (Table 3). Only lor stomach cancer was iliete an exposinc-iisk relationship. However, the different basic materials and lack of data for stomach cancer in workers with more than 25 yr of exposure may invalidate the comparability of the figures, even if the mortality was greater than expected for stomach cancer, lung cancer, bladder cancer, and leukemias. Indirect evidence for an exposure-risk relationship was provided by the classification of the subjects according to the year when the employee started work (Table 4). No differences were found, consistent with the hypothesis of an association between length of exposure and overall risk of cancer or risk of cancer at selected anatomic sites. On the contrary, there was a consistently increasing risk of cancer death with the year when the subject started work. The increased risk during recent years may he an indication of greater reliability of the basic material rather lhan ol an increase in carcinogenic exposure with lime. There was an SMR of 238 for bladder cancer lor the cohort starting work in 1940-1049 and a decrease for the more recent cohorts. Processing was the only division with a significantly increased risk of cancer (Table 5). In only one research project (Fox et al., 1974; Fox and Collier, 1976) was the risk of cancer reported for sectors of the mhber industry. The tire secim, however, was considered in four studies (Table 6), with data spcciiicd accoidiug lo division and to risk ol cancer by analomic site. In the total I itc sector there was a considerably ini leased risk ol stomach cancer (SMR = 145) and leukemia (SMR = 129). Mode of exposure varies between the different divisions within the tire sector. Processing workers arc exposed through gastrointestinal and respiratory organs by inhaling and ingesting dust, while building workers are exposed through the skin by handling the rubber. Curing workers are mainly exposed lo hot vapors, and this is likely to affect them through the respiratory organs. Increased chemical exposure of Finnish rubber workers in all the divisions was confirmed by Vainio et al. (1978), who found an increase in urinary thiocthcrs in workers in all these divisions. The mode ol exposure should he correlated with the anatomic site of the cancer. A classification of the data within the tire sector by division (Table 7) revealed that processing workers also had a high risk of cancer. High risks were found for stomach (SMR = 184) and bladder (SMR 166) cancers and for leukemia (SMR = 137). Similarly, high risks of leukemia, respiratory cancer, and stomach cancer were found within the curing sector. Because the numbers were small, only the iruieascd risk of respiralmy cancer (SMR 149) w.is stalislically signiluaiu. Iheellectsol exposure through the skin among building workers might result in an increased number of skin cancer eases. Skin cancer patients have high survival rates and few deaths from skin cancer are likely, even with an increased risk. The small number of cases results in a large random error, which may prevent reliable conclusions. There have been only two studies ol cancer incidence among rubber workers. The one reported from the United States (Monson and Fine. 1978) derived the new eases of cancer from hospital-based cancer registries; the only population-based cancer registry material was from Finland (Kilpikari et al., 1979). In die U S. vitidy, . SMR of 650 was reported for skin cancer among rubber workers in the tire assembly division. No eases of skin cancer were found in the Finnish material. The results by anatomic site of cancer are largely in accordance with the mode of exposure. -f- _ o < 3 S_ c. o2 rm> --^ --rf K o" <" o ^ O 3 E =r ? 2 5 3 O s 2. C. S. n-3-is3, o^o3a 3oq 3go ' o :< x 3 ^" 'S p* o O ^ o' af S o' f a-% *> n 5 S < Of 3 3 C - 3^ 2 X o . 5i 3 o. S * O -*t r s -* o Cj V KJ Q. (F"3i O fij Ui 3; < --o1 ' O s O^t w3 3^ 1O/1 ^ Ccr f'H 0 "< (U -- a* . 0. 3 2 3 r* w ^r KJ 99 k, 5. ^ 0 U 3 a * F > * Iso rliin TABLE 3. Standardized Mortalin Ratios and Numbers of Deaths by Length of Exposure and Cause ot Death in Male Rubber Workers Lengh of exposure fvr) Study >1 Fox nd Collier, 19?6 >5 \ BRMA, 1976 ) Monson and Nakano, 1976a >10 J BRMA, 1976 jMcMichael et al., 1976a >25 Monson and Nakano, 1976a Not available. All deaths SMR No. 102 2018 82 5079 94 5106 86 2890 All cancers SMR No* in 592 97 1955 100 1650 98 576 Stomach SMR No. 102 59 114 233 143 171 NAff NA Lung SMR No. 118 281 106 698 103 555 165 20 Bladder SMR No. 195 37 119 62 103 48 129 31 Leukemia SMR No. 128 14 105 83 101 64 163 37 TABLE 4. Standardized Mortality Ratios and Numbers of All Deaths and Deaths from Site-specific Cancers by Year Employee Started Working in Rubber Industry0 employee started working All deaths SMR No, All cancers SMR No. Lung SMR No. Bladder SMR No. Leukemia SMR No. 192S 86 2040 92 349 130 8 140 24 140 21 1925-1934 80 1740 100 378 200 14 140 19 160 25 1940-1949 103 485 104 133 117 67 239 11 NAb NA 1950-1959 102 622 105 174 120 88 175 10 107 16c 1960-1966 106 497 122 153 128 74 166 7 NA NA The data were obtained from the following studies: BRMA (1976); Fox and Collier (1976); Monson and Fine (1978), *Not available. c1951-1960. 1 AIII.L 5. Standardized Mortality Ratios and Numbers of Cancer Deaths among Male Rubber Workers by Division0 Division SMR No, Processing Produit Ubriuilion Curing Product Finishing Chemical production Services 118 96 101 106 82 95 285 764 105 304 28 645 The data were obtained from the following studies: Andjelkovich et al. ( 1977); BRMA (1976); Fox and Collier (1976); Monson and Nakano (1976a). TABLE 6. Standardized Mortality Ratios and Numbers of Deaths due to Site-specific Cancers among Male Tire Worker^1 ICD no. Primary site SMR No, 140-239 All neoplasms 107 1533 151 Stomach 14S 145 162 Lung 104 520 188 Bladder 112 61 191 Brain 90 28 204-207 Leukemia 129 57 The data were obtained from the following studies: Fox (1976); Fox et al. (1974): McMichael et al. (1976b); Monson and Nakano (1976a). i CANCER RISK AMONG RUBBER WORKERS 1217(2971 TABLE 7. Standardized Mortality Ratios and Numbers ol Deaths due to Site-specific Cancers among Male Tire Workers by Division0 Stomach Respiratory Bladder Leukemia All neoplasms Division SMR No. SMR No, SMR No. SMR No. SMR No. Processing 184 26 110 12 166 2 137 23 120 60 Building 103 II 56 13 135 17 122 25 85 70 Curing 140 7 149 64 7S 1 270 5 101 66 The figures are based on the following studies: Andielkovich et al. (1977); Fox and Collier (1976); Fox and al. (1974); McMichael et al. (1974); Monson and Fine (1978); Monson and Nakano (1976a). The total cancer risk among rubber workers in the studies published in the 1970s was not essentially different from that expected on the basis of the normal mortality of the total population that included the exposed cohort. There was an increased risk for two primary sites -stomach and bladder and for leukemias. A high risk of cancer at a specific anatomic site was likely lo be correlated wiih mode ol exposure; i.c., with the working division. The SMRs for the selected cancers varied between 140 and 180 among the high-risk groups by working area. Even within these high-risk groups, the relative risks were less than 2 when based on sufficient numbers of cases, indicating a relatively small ctiologic fraction due to exposure. For several of the leading primary sites of cancer there arc common risk factors with bigger relative risks. The relative risk of lung cancer due to smoking is about 10, that of breast cancer due to factors related to fertility is about 3, and that of stomach cancer related to nutrition or standard of living is more than 2. However, occupational exposures can often be prevented, and we do not wish to neglect the problem. Furthermore, epidemiologic studies of occupational exposure are subject to bias due to the healthy worker effect, and it is difficult to find reliable data on the exposed cohort and on cancer deaths. These biases can lead to underestimates of the true effects of exposure. On the basis of the most recent epidemiologic studies of cancer risk among rubber workers, we conclude that the risk is increased. However, only in exceptional circumstances is the excess risk of the same magnitude as that for the leading cancers and common exposures. REFERENCES Andjelkovich, D., Taulbee, I., and Symons, M. 1976. Mortality experience of a cohort of rubber workers, 1964-1973. /. Occup. Med. 8:387-394. Andielkovich, D., I aulhc-c, I., Symons, M., and Williams, T. 1977. Mortality of rubber workers with reference to work experience. /. Occup, Med 19:397-405. Andjelkovich, D., Taulbee, I., and Blum, S. 1978. Mortality of female workers in a rubber manufacturing plant./. Occup. Med. 20:409-413. British Rubber Manufacturer's Association (BRMA). 1976. Health Research Project Report. Birmingham: BRMA. Case, R. A. M. and Hosker, M. E. 1954. Tumor of the urinary bladder as an occupational disease in the rubber industry in England and Wales. Br. j. Prev. Soc. Med. 8:39-50. Fox, A. |. and Collier, P, E. 1976. A survey of occupational cancer in the rubber and cablemaking industries: Analysis of deaths occurring in 1972*74. Br. /, Ind. Med. 33:249-264. Fox, A, Lindars, D. C,, and Owen, R. 1974. A survey ol occupational cancer in the rubber and cablcmaking industries: Results or five year analysis, 1967-71. Br. /. Ind. Med. 31:140-151. General Register Office. 1934. Census ol England and Wales 1931, Occupations, table 2. London: Her Majesty's Stationery Office. 1218(298) M. HAKAMA AND I. KILPIKARI McMithael, A. Spirlas, K., and Kuppct, L. t~ 1974. An epidemiologic sludy ol mortality within a cohort of rubber workers. /. Occup. Med. 16*458-464. McMichacl. A. I., Amljelkovich, 0. A., and lyrolct, II. A. 1976,i. Cutter mortality among rubber workers, an epidemiologic study. Ann. A. Y. Acad. Sii. 271*125-137. McMichacl, A. Spinas, R., Gamble, I. F,, and Touscy, P. M, 1976b. Mortality among rubber workers; Relationship to specific jobs. /. Occup. Med. 18:178-185. Monson, R. R. and Fine, L, |. 1978. Cutter mnn.iliiy .iiirl moiludiiy among rulilier workers, /. Nad. Cancer inn. 61; 1047-1053. Monson, R. R. and Nakano, K. K. 1976a. Muri.tluy aiming luhhct workeis. I. While male union employees in Akron, Ohio. Am. /. Lpidemioi. 103:284-296. Monson, R. R, and Nakano, K. K. 1976b. Mortality among rubber workers. II. Other employees. Am. /. Epidemiol. 303.297-303. U.S. Department of Health, Education, and Welfare. 1961-1963. Mortality in 1950 by occupation and industry: Vital Statistics -Special Reports 53, numbers 1-5. Vainio, H,, Savolaincn, H, and Kilpikari, I. 1978. Urinary thioethers of employees of a chemical plant. Be. I. hid. Med. 35:232-234.