Document zQjyJERD5EM9nKvorbeQYE79R

76 Occupational Health LEUKlEMIA IN BENZENE WORKERS PETER F. INFANTE JOSEPH K. WAGONER ROBERT A. RINSKY RONALD J. YOUNG Industry-wide Studies Branch, Division ofSurveillance, Hazard Evaluations and Field Studies, National Institute for Occupational Safety and Health, Centerfor Disease Control, Cincinnati, Ohio 45202, U.SA. Summary Workers occupationally exposed to benzene in 1940-49 were followed for vital status up to 1975. In comparison with two control populations, a significant (~<0.002)excess of leukicmia was observed. A fivefold excessive risk of all leukicmias and a ten-fold excess of deaths from myeloid and monocytic leukemias combined are demonstrated in the study population compared with controls. These figures underestimate the true leukemia risk to benzeneexposed workers, because follow-up is only 75% complete and the untraced 25% of the study population were all regarded, in the statistical analysis, as being alive at the end of the study period. The environment of the workers in the study population was not contaminated with solvents other than benzene, and existing records indicate that the benzene levels themselves were generally Mow the limits recommended at the time of their measurement. INTRODUCTION ONE of the earliest reports of benzenerelated blood disorders consistent with what is now known to result from benzene poisoning was published in Paris in 1897.' In 1928, a clearly established case of leukemia was associated with benzene exposure in Italy.2 Since then, many cases of leukamia associated with exposure to benzene have been reported from various parts of the world.'-' A significant excess of chromosomal aberrations has also been observed in benzeneexposed workers.'0 Ishimaru et aL6demonstrated a significant association between leukamia and occupations with probable exposure to benzene, and Aksoy et al.' reported a significantly greater incidence of leukamia in shoe workers exposed to benzene than in the general population. In 1976 Vigliani` reported that 150 cases of leukamia attributed to benzene had been identified in Italy. McMichael et al." demonstrated an excess of leukamia among rubber workers previously exposed t o solvents, including benzene. However, in that study the leukemogenic effect could not be attributed directly to benzene because of concomitant exposure to other solvents used in the manufacture of rubber products. Before the 1914-18 war, benzene was used mainly as a solvent in the rubber industry, but during the war the demand for toluene in the manufacture of explosives greatly stimulated benzene production. Currently, 87% of benzene is used as a raw material in the synthesis of compounds such as styrene, phenol, and cyclohexane. Approximately 10.2 billion Ib (4-6 billion kg) -.'rere pro- %+! f 4)duced in in Japan the in U.S.in 1973, 1972, and 1.2 3b.i5lliboinlliIobn(I0b.5(1b.6illbiofi&n&k-t*, 1 the U.K. in 1972." In the 50 years since reported case of benzencassociated leukicmia, has been neither widely acknowledged nor unif controlled as a carcinogen in the U.S.I3 The failu-Yrr I control benzene as a carcinogen has resulted P from the reluctance of industry and Govern accept the accumulated case-reports and epidemiol- t. b observations as scientific evidence of the l e u k m properties of benzene. Today, in the U.S. alon0e-k, rs estimated two million workers are at risk of exposure +benzene." For these patterns in workers orcecausopnastiownealhlyaveexpstousdeidedtomo-% in the production of a natural rubber cast film ( r u b hydrochloride, marketed under the trade n- `Pliofilm')at two localities in Ohio. INDUSTRIAL-HYGIENE ASSESSMENT 1 I Production Methods In the production of `Pliofilm', natural rubber wps cated in an internal mixer, passed through a mill, and,- vcycd to mixing-tanks, where benzene was added and the + I ture was agitated. T h e resulting solution was pumped to, blending-tank, then transferred into a reactor vessel, w h a & was treated with hydrochloric acid to form rubber h y d r o c b i idc. At this stage, more benzene was added to adjust the pp portion of solids in the solution. T h e rubber hydrochloride rr, transferred to a neutralising-tank, where pliofilm scraps, ash, steam, plasticisers, and more benzene were added. i filtration, the rubber hydrochloride solution was pumped 10, casting-unit, then it was spread on a continuous copv-, where the benzene was evaporated, and the finished film taken up on a roll. T h e gas containing benzene was e x h a w from the casting-unit, passed through a wet scrubber, channelled into an activated-charcoal absorber unit, a h benzene was recovered and recycled. T h e rolls of pliofilm transferred to another room, where they were inspear4 trimmed, cut to customer specification, and packed (pfor food wrapping). In the production of pliofilm, the material known to be associated with blood disorders was & I 4"zene. Other materials used included hydrochloric acid, ash, and small amounts of antioxidants and plasticisen. T& ,itg-manufacturing process was essentially identical at both localities. z Environmental Monitoring ,:I. In 1942, Wilson, as a former member of corporate m a n a p ! ment, recommended a policy for the rubber industry, whicb would include pliofilm operations, of periodic air monitorin# and a complete biood-count for all new employm.lJ It was bir view that "a closed system of ventilation where the workrmn comes in contact with no f u m e is ideal. Any other type of v e b tilation is not entirely safe." Shortly thereafter, cxtendvc i 1 exhaust ventilation equipment was installed in the pliofilm department in one locality. T h e Industrial Commission d Ohio, after surveying that department in 1946, reported thU, "Tests were made with benzol detectors and the results in& cate that concentrations have been reduced to a safe level and in most instances range from zero to 10 or 15 parts per r d lion".I6 A further investigation of benzene exposure in thm I I I department in 1956 concluded that exhaust ventilation trols were of excellent design." In the early 1960s benzcnc point-source concentrations were measured by company p sonnel using detector tubcs. A total of 112 surveys were Qp ducted in the period 1963-74. These and earlier surveys haindicated that employees' benzene exposure was g c n d &low the recommended limit in effect at the time of each SlDL vey1J~~4~("ta-b*loe t). I ! I i THE LANCET, JULY 9,1977 77 TABLE I--CHRONOLOGY OF RECOMMENDED BENZENE CONCENTRATIONS IN WORKPLACES IN U.S.A. 1 17 - r Concentration (p.p.m.) Reference ~~ ~ 1941 1947 1948 1957 1963 1969 1971 ~ 100 MAC. 50 8h T.W.A. 35 8h T.W.A. 25 8h 1.W.A. 25 Ceiling 10 8h T.W.A. 10 8h T.W.A. 16, 17 14 14 14 14 18 13 II M.a.c.=Marimum allowable concentration. T.w.r.=Timcweightedaverage. Historical environmental data for the second locality in Ohio is less well defined. However, our own observations, dis- cussions with company personnel, and the meagre environ- mental data that have becn retained suggest that employees' benzene exposure at this location was generally well within the recommendedlimits. RETROSPECTIVE COHORT STUDY Merhoa's All White men who had had direct exposure to benzene at any time h e e n Jan. 1, 1940, and Dec. 31, 1949, were studied. The departments and jobs involving direct exposure to benzene were determined by a survey by N.I.O.S.H. personnel, which included a review of the pliofilm process, engineering controls, and air-sampling data. Company personnel also assisted in these determinations. Job classifications involving benzene exposure w-re determined without knowledge of the workers' vital s. F o M - u p of all study-cohort members was attempted for the time period horn first employment to June 30, 1975. To date, vital status has been determined for approximately 75% of the 748 cohort members. The m a i n i n g 25% were assumed to be alive to avoid overestimating the true risk of lymphatic and hemopoietic malignancies associated with benzene exposure. Causes of death were determined from death certificates and W e n coded according to the International Classification of Diseases in effect at the time of death; these codes WerC then converted to I.C.D. 7th revision numbers. A modified life-table technique was used to generate perWn-Years at risk of dying according to 5-year age-group and 5-yearcalendar periods. Pmn-years of observation and causes of death were damnincd for the period Jan. 1, 1950, to June 30, 1975. Person-years of observation and deaths occurring before Jan. 1, 1950, were excluded from analysis since statistics on lymphatic and hamopoietic malignan''a Were not published before that date. Two Populations were chosen as control groups for m m t i n g the numbers of expected deaths in the study mulation. The first group consisted of the U.S.White male general population standardised for age and time pried Over which the study cohort lived. The second group consisted of 1447 White men who had been Onpi0Yed in Ohio at a fibrous-glass construction-produas factory between Jan. 1, 1940, and Dec. 31, 1949, and who had achieved 5 or more years of employment by J"- 1, 1972 (which was the cut-off date for vital- tion).ar- mainment in the fibrous-glasscontrol popula- Results There were 140 observed deaths from all causa among benzennxposed workers compared with 187-6 expected deaths (table 11). This deficit reflects in part the incomplete follow-up of the study population. There was a significant excess of deaths from malignancy of the lymphatic and hamopoietic systems (I.C.D. 200-205) compared to that expected on the basis of death-rates of U.S. White males. This is due almost entirely to an excess of leukamia deaths (I.C.D. 204). To date, 7 leukamia deaths have been observed, compared with an expected 1.38 (~(0.002)based on rates for U.S. White males and an expected 1.48 (p<0.002)based on rates for fibrous-glass workers. Even more striking is the observation that all 7 leukamia deaths were from either the myelogenous or monocytic type. Data for cell t=, age at death, and interval since initial exposure to benzene are shown in table III. The period between initial exposure and death ranged from 2 to 21 years. These observations, indicating a specific type of leukamia, are consistent with the data of Vigliani,' which show a predominance of acute myelogenous l e u k m i a among workers exposed to benzene in the rotogravure and shoe manufacturing industries in Italy. Since data for specific TABLE 11-STATUS ON APRIL 13,1977, OF OBSERVED DEATHS FROM SPECIFIC CAUSES AMONG WHITE MALES BETWEEN JAN. 1, 1940, AND JUNE 30, 1975, WHO WE= EXPOSED TO BENZENE AND EMPLOYED SOMETIME BETWEEN JAN. 1,1940, AND DEC. 31, 1949, AS CONTRASTED WITH EXPECTED DEATHS BASED ON BOTH THE US. WHlTE MALE POPULATION (U.S.W.M.) AND AH INDUSTRIAL POPULATION EXPOSED TO FIBROUS GLASS (F.G.). Causes of death (I.C.D. codes)' Benzeneexposed workers Deaths observed Deaths expected Comparison All causes 140 187.5809 (200-205) 9 3.4497 (U.S.W.M.) (204) 7 1.3831 (200-205) 9 5.1020 (204) 7 1.4758 *?.C.D.-7th Revirion Coda. (200-205) = Total lymphatic and homopoieticcancer. (204) = Lcuk~miaonly. tp<o.os sp<0.002 ~ . ~ . ~ . = ~ b x r v c d / c ~ p e c 1t0e0d x Case no. Type Of leukomia Acute myelogenous Acute myelogenous Acute myelogenous Acute myelogenous Monocytic Chronic myelogenous Monocytic 8' Myelogenous Ageat death (yr) 60 65 62 57 57 29 36 28 Period from initial e x p u r c to death (yr) 13 10 21 19 1s 2 17 3 &&- A ~ 78 THE LANCET, JULY9, types of leukzmia mortality in the U.S. are not readily available for the study period, incidences for myeloid, monocytic, and total leukamia from the Connecticut Tumor Registry2' for the period 1960-62 were applied to the agedistribution of the benzencexposed cohort to generate the specific proportion of total-leukamia cases which would be of the myelomonocytic types. Of the total expected number of incidence cases of leukamia, 50.377 were calculated to be of the myelogenous and monocytic types. This proportion was applied to the total number of expected leukamia deaths in the study cohort which had been previously generated on the basis of rates for the U.S. White male population. The results indicated an expected 0.6967 (0.5037 x 1-3831 = 0.6967) deaths from myelogenous and monocytic leukzmia, compared with 7 observed, the standardised mortality-ratio being 1004. Thus, pliofilm workers exposed to benzene have an estimated 10-fold risk of dying from myelogenousand monocytic leukamia. As a result of past failure to control benzene as a carcinogen, millions of people, without knowledge of the hzmopoietic dangers, are continually being exposed to benzene at work. In addition, unknown numbers of consumers, including children, are exposed to benzene at home, from sources such as model-aeroplane glue and paint strippers (somecontaining ul to 508 benzene). We hope that our findings, which demonstrate overwhelmingly an increased risk of leukzmia in workers exposed to benzene, will stimulate efforts to control occupational and consumer exposure to benzene, an agent known for almost a century to be a powerful bonemarrow poison.' We thank M s Betsy Egan and .Mr David Brown and the entire sup port statl of the biometry section, National Institute for Occupational Safety and Health, for their dedication and assistance during the study. Rquats for reprintsshould be addressed to P. F. I. REFERENCES 1. IxNoir, C. Bull. Mem. SOCm. id. H6p. Paris 1897, 3, 1251. 2. Delore, P.. Rorgamano, C.7. Mid. L-von. 1928.9.227. 3. Gogucl. A,. Cavigncaux. A., Bernard, J. Nouv. Rerucfr. Hdmar. 1967, 4, 465. 4. Vigliani, E. C.Ann. N . Y .Acad.Sci. 1976,271.143. S. Ludwig, V. H..\X'enhemann,A.Schw~.med. Wschr. 1962,92.378. 6. IshimaN. T., Okada. H.. Torniyasu, T., Tsuchimoto, T., H a h i n a T., lchi- maru..M.Am.J. Epidem. 1971,93.157. 7. Akroy,,M.. ErdamS., DmCo1,G. Blood. 1974.44.837. 8. .Mallory, T. B., Gail, E. A.. Brickley. W.J. J . ind. H-vg. Toricol. 1939, 21, 355. 9.TarcetT, E. M., liontchalovskava, N. M.,Zonna, L. A. Acra. Un. inn. Cann, 1963.19.751. IO. Parni, A,, I'acifico. E.. Limonta. A. Archs enwir.Hlfh. 1971,22,373. 11. .McMichacl, A. J.. Spinas, R.. Kuppcr. L. L., Gamble, J. F. J . occup. Mcd. 1975,11,234. 12. 1A.R.C. Monopzphs on Ecduarion of Carcinogenic Risk of Chemicals IO .Man. Some Anti-thvroid and Related Substances, Nitrofurans and indus- trial Chemicals. 1974.7,203. 13, Code of Federal Regulations. Title 2 9 Subtitle B, Chapter XVII, &pa- tional Safety and Health Administration. Pan 1910.93, U.S. Washinpon, n c , 1971. 14. Critcna for a Recommended Standard. Occupational Exposure to Benzene. 1`5 Dcpt. Health, Education and Welfare, Public Health Service. Wash- ington. D.C.. 1974. 1 5 . Wilron. R. H.J . Lab. din. Mcd. 1942.21, 15 17. 16. IndustrialCommission ofOhio. Rcpon, Nov. 1946. 17. Ohio Depanment of Health. Rcpon, Jan. 1956. 18. Drinker. P. Cntena of lndustnal Health Hazards. In United Stain Dc~M- rnent of Labor: Protecting Plant .Manpower.Washinpon, D.C., 1941. 19. Sappington, C. 0.Esxntialrof Industrial Health. Philadelphia, 1943. 20. V.S.A. Standard Acceptable Concentrations of Benzene, 237.4. United ..Stain of Amenca Standards Institute, Inc. New York, 1969. 21. Ravliss. D. 1. Ikment. 1. M.. Wamner. 1. K.. Bleier. H. P. Ann. N.Y. Acad. I _ SCI1. 976,271,324. 22. Cancer in Connecticut, Incidence and Rates, 1935-1962. Conncnicut State k p t . of Health, Hanford, Connecticui, 1966. Before Our Time VIRGINIA BERRIDGE .8. e: Addiction Research Unir, Inciitute of Psychiatry, London sas The literature produced over the last ten years ana,, \ ing and describing the dom has almost totally "ndergulgecstecdenteh"eihnisthtoeriUcanliteexda`mciap, t ! ra,tion of English drug, and especially opiate, use. familiar examples are quoted and requoted; referto the mid-19th-century opium wars abound. But , 9did ordinary people use opium in the 1820s, when Quincey was writing his Confessions? How @cot, danger did the unrestricted sale of opium pose? How @ 4the medical profession, the chemists, and the public &' large react? Such questions have remained largely answered; and yet responses to them are needed if a pa- perly informed perspective is to be given to the worries, and a more rational appraisal made of con& porary attitudes and policies. 31% 4UNRESTRICTED OPIATE USE Narcotics have been under strict legal control as long as anyone can remember-there can be few *tors now in practice whose professional experience g& back to the years before the first Dangerous D r u g became law in 1920. It is therefore instructive to that, up to 1868, opium and cannabis (in fact rarely used) were available over the counter form of restriction. Imported opium, mostly origin, could be bought almost anywhere-n chemists, druggists, or pharmacists, but in vi1 from grocers, general stores, and corner back streets of the growing industrial cit coroner, for instance, noted in 1854 after an infant poisoned with laudanum sold i syrup of rhubarb, that the liquid opiate "was several other bottles of similar shape and appeara a shelf in the shop window of a Grocer in a .8thickly inhabited by factory operatives, and was given * a messenger between six and seven years old".' Acceptance of the drug went so far as to encourage enterprising farmers, doctors, and market gardenm to experiment with the cultivation of a British variety. Messrs Cowley and Staines, for instance, who cultivated severaf acres at Winslow in 1823, managed to produe 186 lbs of opium from 12 acres of land. Their profit, b cluding the yield from a crop of turnips prudently planted in addition, was over E96 an acre; 0th- reported similar successes, although "the precariousna of our climate" was universally recognised as a draw back.**' British opium was never generally used; bm opiata--chiefly laudanum, the camphorated tincture known as paregoric elixir, and raw opium, sold in p a and penny sticks-were widely used both on medid prescription and as self-medication by those who c o d not afford medical treatment. In one area of the country# the low-lying, marshy, ague-riddsn Fens, it was accepted that opiate use on a regular and sustained basis wm