Document byGXpOO3pgDpM20rx0B6XNo16

*-. Brief Communication Excess Leukemia in a R e f i n e r y . . - Population ' , " I -- ' - Dorothy S. MCCraw, M.S.; Roy E. joyner, M.D.; and Philip Cole, M.D. A retrospective follow-up mortality study of white male emdovees of an oil refinery located in southern Illinois &e identified; eight of the CancerZ were of the AML cell type. The work histories of the decedents were examined for the potential for benzene exposure. Although specificity of the work histories was limited, the investigation has shown that the subjects did not work in jobs identified as having the highest benzene exposures. At this time, the excess leukemia remains unexplained. Exposure to benzene at high levels is associated with an increased risk of acute myeloid leukemia (AML) developing.'-z Although many petroleum refinery workers have had some exposure to benzene, mortality studies of these workers have not shown a pattern of excess leukemia.This report presents the results of a study of r e f i n e r bb Methods The Wood River Refinery located in southern Illinois began operations in 1918and currently produces a full line of refinery products. A research laboratory operated o n the site from 1933 to 1975. The study population consisted of all white male employees (salaried and hourly) who had worked at the Wood River Refinery for at least one day during the 10-year period, Jan. l, 1973,through Dec. 31,1982.and retirees whose last A r k location was the study refinery and who were alive as of Jan 1.1973. Vital status as of Jan. 1,1983, was determined for all employees and retirees in the study group by the companfs pension and life insurance departments. For employees who terminated employment for reasons other than retirement, vital status information was not obtained. These individuals thus are treated as lost to follow-up. Because a death certificate is required as proof of death before the p a p e n t of benefits to a surviving spouse, most certificates are received by the company. Those not received were requested from the appropriate state vital statistics departments. For each death certificate, t h e underlying a u s e of death was coded to the eighth revi- sion of the International Classification of Diseases by a trained nosologist.* A person entered the study on Ian. 1,1973, if he began employment on or before that date. Otherwise, he entered the study the day employment started. Person-years were accumulated from the entry date to the date of death, the date of termination, or the study end date, whichever came first. Standardized mortality ratios (SMRs) adjusted for age using the US. white male death rates as the standard were calcu1ated.w (Leukemia rates for the county in which the refinery i s located were virtually the same as US. rates.) An additional analysis of cell-typeqwcificSMRs was done using the Surveillance, Epidemiology and End Results mortality rates for white males, all regions combined, 1973 to 1977." The 95% confidence intervals (CIS) of the WRs were also estimated.= The employment history of each person who died of leukemia UN also investigated. Employment history before working at Wood River Refinery was identified from pre-employment application forms. A work history r e cord, kept forpayroll purposes, listed department and job title by time period. From Shell Oil Company, Houston (MS.McCraw, Epidemiologist. ana 0 . . Wer. Corporate Medical Dircclori; and the university of Alabarw in Birmingham(Or. b l e . Associate Director for Epidemiology ot the Comprehensive Gncer Center and Professor of the De- . partment of Epidemiology). Address conepodmce to Shdl Oil, P.O. Box 2463. Hourton, Teus 77OOl (Ms.Mdlraw). 220 Results The study population included 3,976 white men. About half of the cohort began employment before 1948, and 68% of the cohort had become pensioners before the end of the study period. The mean age of the population o n the day of entry into the cohort was 49 years. Eight percent of employees left er..ployment for reasons I-7 - ,i I I I i I T8ble 1 4 u m b e n o! Observed rad Expected Oa8tfts rad SMRt Afcwdlng to Cause' Cause of Death ~~ All causes All malignant neoplasms Leukemialaleukemia 'Lymphosarcoma and reticulosarcorna Hodgkin's disease Cancer of other lymphatic tissue Other malignant neo- plasms All diseases of the blood and blood-formingor- gans AJI diseases of the arculatory system All respiratory tract diseases All diseases of the digestive system Observed 640 161 14 0 2 6 139 2 361 28 15 Expeded 637.6 177.0 6.6 2.8 1.o 4.8 161.8 1.8 470.0 64.2 30.6 IMR Et 91- 213t - . 0 200 126 86 111 nt ut 4% U.S. white males is the comparison population t An SMR is statistically significant at p c -05 if the 95% C1 does not indude 100 ma ?I -83 n-io6 ii7-35a <130 - -- 22-722 466-274 72-101 1342 69-85 29-63 27-81 - T8ble 2--Laukemia Cell-Specific SMRs for were observed but are based on numbers too small to be Specific Foms of Leukemia. interpretable (Table 2).No lymphosarcoma or reticulosar- 1 coma was seen even though 2.8 cases were expected. * Form Observed Expected SMR %%CI Three of the leukemia patients had held laboratory jobs. W e lymphatic 1 ~ 0.3 340 4-1.854 five spent most of their careers as maintenance workers, i Chronic lymphatic 1 1.3 74 1-428 but it i s not known if they were ever assignedto benzene Aarte myeloid 8 2.0 3941 172-788 units. None of the case subjects was an operator in the Chronic myeloid '1 0.8 121 2-696 benzene units or a loader in the dispatching department. Awte monocytic Other and 1 .0.2 661 82.m Six subjects (twodraftsmen, a patrolman, and three com- pounders) probably had no benzene exposure above re- Nonspecified 2 2.0 100 11-361 finery background level. According to the companfs ! employment application forms, two had worked at other Surveillance. Epidemiology and End Results white males, all re- oil companies before being employed at the Wood River ! gions combined. is the comparison population Refinery. , t An SMR is statistically significant at the p e -05 level if the 95% i CI does not include 100 Discussion iI &e identification of excess (fyk9Fia deaths in po it* other than retirement and were therefore lost to follow- up. At the time of their leaving, 94% of this lost to followup group was under 40years of age and 73% had five years or less of employment. study. A twGGoWincreasein IeukGZGms observed. Vir- ~ my ah of the excess n-er of leukemia cases were of. %e M L type. There were 640 deaths from all causes with 838 ex- U.S. males seem an appropriate comparison population -i pected, an overall SMR of 76. Table 1lists the SMRs for since the proportion of deaths from leukemia is 0.80% for major auses of death and for leukemia. The SMRs for all the United States, 0.80% for Illinois, and 0.84% for Mad- I causes, all diseases of the circulatory system, all diseases of the respiratory system, and all diseases o f the digestive ison County, the county in which the refinery is louted." Leukemia probably is not overdiagnosed in this region system are lower than expected, with the differences sta- and overdiagnosis of AML i s unlikely. tistically signifiant. At the Wood River Refinery, the concentration of ben- For leukemia, there were 14 deaths observed with 6.6 me in refinerystreams did not exceed 5% until 1942when expected, an SMR of 213 with a 95% CI o f 117 to 358. In the toluene unitswere built with one stream of 5% to YO% 12 cases cell type was specified on the death certificate benzene. Subsequentty, seved unitswith benzene streams and, of these, eight were of the acute myelogenous cell from 5% to 100% were constructed. These includeda cu- tvpe (AML). Two deaths from A M L were expected; thus, a,the SMR for AML i s 394 (95% 172 to 788). Similar in- mene unit (operating from 1943 to lWS),a dodecyl benzene unit (1950to 1954),benzene extraction units (1960to creases in acute lymphatic and acute monocytic leukemia present), and catalytic reforming units (1960to present). Joumal of Occupational MedianeNol. 27. No. march 1985 221 Benzene, as well os other solvents, was used both in the qualitytontrol laboratory (1941 to present) and In the re*arch laboratory (1933to 1975) in several tests and iiitermittently in the cleaning o f glassware. Benzene and benzenecontaining compounds were transferred to barges and tank cars and to and from trucks in the dispatching department. Purchases of coal-derived benzene used in refinery processes were made periodically from 1943 to the present. What might account for the differences between this study and other studies of refinery workers? Although refineries are basically similar, we do not have the specific information necessary to compare the operations at the Wood River Refinery with the operations in other refineries that have been studied. In addition, the Wood River Refinery purchased both finished and crude coalderived benzene. A recent study of rubber workers found a stronger association between coalderived solvents and lymphocytic leukemia than between petroleumderived solvents and lymphocytic leukemia." Another difference between the present study and others i s the cell-specific analysis. The SMR for leukemia as a whole is modest when compared with that for AML. It is possible that other refinery studies could have an elevation in A M L that i s diluted by the summary SMR for all cell types of leukemia combined. Since A M 1 has been associated with benzene exposure, work histories of the case subjects were reviewed with company industrial hygienists to assess potential for ex- posure. While individual exposures cannot be defined completely by this preliminary study, it was determined that the subjects had not been assigned to jobs deemed to have the highest potential for exposure, i.e., processes involving high concentrations of benzene. Efforts are under way t o provide a more definitive assessment of the effects of past exposures. At this time, the excess of leukemia remaim plained. Acknowledgments Nancy Anderson performed the computer program. ming that made this analysis possible. References 1. Atsoy M: Different types of malignancies due l o occuptiorui exposure to benzene: A review of reCmt observations in Turk-. fnvrron Res 23:181-190, 1360. 2. Vigiiani Ec,k i t a C:&Nene and kukemia. NEnglIMed271 :m- 676.1964. 3. Decoufle P. Elanner WA. Blair A: Mortality among c h e e r workers exposed t o benzene and other agents. Environ Res u):i&z. 1983. 4. Hanis NM, Stavraky KM. Fowler 11: Cancer moflality in oil r e finery worken. j Occup Med 21 :167-174,1979. 5. Hanis NM, Holmes TM. Shallenberger LG. e( ai: Epidemidogtc study of refinery and chemical plant workers. j Occup M e d 2 4 : m s n2.1982. 6. Rushton1,Alderson MR: An epidemiological sumy of eight oil refineries in Britain. 81/ Ind Med 38:225-234,1981. 7. Theriault G. Coulet 1:A mortality study of oil refinery woken. I Occup Med 21367-370.1979. 8. Wen CP,Tsai SP, McCkllan WA. et d: long-term mortalii stuciy of oil refinery workers. Am / pidemio/118:526-542,1983. . 9. International Classification of Diseases. Mappled for Use in thc UnhedStates. eighth revision. M S publication1693. Washington, D.C: US. Dept. of Health, Education, and Welfare, 1%7. 10. Monson RR: Analysis of relative survival and proportional mortality. Cornput Biomed Res 7:32!%332,1974. 11. Surveillance, Epidemiology and End Results: Incidence and Moflality Data 1973-1977. NIH publication 81-2330. Mhesda. Md.: National Institutes of Health, 1061. pp 110111. 12. Rothman KI. Eoice ID: Epidemiologic Analysis Wcth a Rogrammable Glculator. Boston: Epidemiology Resources. I=. pp 31. 13. Vital Statistics of the United States 1978Volume II Mortality htl A. publication 83-1101. Washington, D.C.: U.S. Dept. oi Health and Human Services. 1982, pp 524.563. 14. Arp LW, Wolf PH, Chcckoway H: Lymphocytic kukemio and exposures to benzene and other s o t v m ~in the rubber industr).. Occup Med 25:598-602.1983. Medicine: The Icy Science? The other day, my doctor sat at my bedside pst to talk. He assured me my physical complaints will be eased and that he will be in regular attendance. We talked frankly of the dying process and the need of living as I am dying: living to fully appreciate every moment of life. I liked our conversation, but it i s hard to come by. Most physicians have lost the pearl that was once an intimate part of medicine, and this i s humanism. Medicine i s now an icy science; its charm belongs to another age. The dying man can get little comfort from the mechanical doctor. 4 r o m 'Thoughts of a w n g Physician" by F. Srenn in f m m on Medicine. Vd. 3:?lb719,1929. 222 Leukemia in Refinery WorkerYMcGraw d al