Document Bn4mzxMy6YOJ1aoZMmLjJzJo

Increased Risk of Developing Acute Leukemia After Employment as a Painter R. LINDQUIST, MD,* 6.NILSSON, B S , t G. EKLUND, PHD, ODONTD,DHc,t AND G. GAHRTON, MD' A case-control interview study of 125 adult patients with acute leukemia and 125 controls matched with respect to age (-+four years), sex, and residence was carried out in central Sweden during the period from September 1980 to May 1983. Their history of organic solvent exposure is described. A significant overrisk of developing acute leukemia was found when comparison between patients and controls revealed a difference in the solvent exposure rate, which was significantly higher in patients than in controls, with an estimated odds ratio (OR) of 4.9,95%CI (2.2 to 12.1). The most frequently exposed profession, painters, exhibit a relative risk of 13 (2.0 to 554). These results suggest that an etiologic relationship exists between organic solvent exposure and the development of acute leukemia in man. Cancer 60:1378-1384.1987. IN THE SEARCH FOR a causal relationship between the effect of chemical compounds on bone marrow cells and the development of leukemia, early studies on workers exposed to the aromatic solvent benzene revealed a possible relationship between benzene exposure and myelotoxicity (Santesson I897).' In 1928, Delore and Borgomano' described the first case of acute leukemia in heavily exposed workers and in 1932 Lignac3 found six leukemias and two lymphomas among 54 white mice exposed to benzene, whereas there was no leukemias among 1465 control mice. Later studies have confirmed that, in rodents, benzene inhalation produces myelogenous l e ~ k e m i a . ~ Although benzene may be the most important organic solvent associated with leukemia d e ~ e l o p m e n to,c~cupational exposure to organic solvents is usually complex. Therefore, results of epidemiological studies are usually based on information concerning the exposure to a mixture of organic solvents and not to benzene a1one.A few such studies have been carried out and an increased risk of developing acute leukemia6-' after organic solvent exposure has been suggested. The sources of information were register^,^.' questionnaires to the next of kin' and case records supplemented by interviews of patients or of their relatives.' From the *Division of Clinical Hematology and Oncology, Department of Medicine, Huddinge Hospital and Karolinska Institute, Huddinge: and the ?Department of Cancer Epidemiology, Radiumhemmet, Karolinska Hospital, Stockholm. Sweden. Supported by the Swedish Work Environment Fund and the Karolinska Institute. The authors thank Professor Lennart Enksson for his comments on the manuscript. Address for reprints: R. Lindquist, MD. Department of Medicine, Huddinge Hospital, S- 14 I 86 Huddinge, Sweden. Accepted for publication March 23. 1987. The aim of the present study was to compare the exposure of organic solvents in patients with acute leukemia and in matched controls to evaluate the significance of such exposure in the etiology of acute leukemia. Material and Methods Patients During the period from September I980 to May 1983, 125 patients with acute leukemia, who were treated in one of the five hospitals taking part in the collaborative treatment protocols of the Leukemia Group of Middle Sweden," were interviewed. Because of the way health care is organized, the majority of patients with leukemia in this region of Sweden are admitted to these hospitals for diagnosis and treatment. There are no private clinics for leukemia treatment in Sweden. If a control had developed leukemia, he/she would have been treated in the same hospital as the patient. Fourteen patients who were in too poor a condition to be interviewed or who had too poor a memory to recapitulate information were excluded. One further patient refused to take part. An alternative to interviewing the patients is to interview the next of kin. The information obtained was not comparable with the information obtained directly from the patients and controls and could not be used in this study. The patients were of rural and urban origin. Seventysix were men and 49 were women. The mean age was 49 years (range, 15 to 84 years). The mean age of the women was 5 I years (range, 16 to 84 years) and of the men 48 years (range, 15 to 84 years). FAB classification 'according to Bennett et uf. I was carried out with a slight 1378 -No. 6 RISKOF ACUTELEUKEMIAAFTERPAINTEXPOSURE Lindquisr et af. 1379 modification.'2 There were 14 patients with MI acute leukemia, 41 patients with M2, four patients with M3, 17 patients with M4, 2 1 patients with M5, 24 patients with acute lymphocytic leukemia and four patients with acute undifferentiated leukemia. Controls One control per patient was obtained from the population register of the taxation authorities, where all inhabitants are listed according to their addresses, district by district. Starting from the name of the patient in this list, the closest situated name of the same sex and age ( t 4 years), as the patient was taken as a control. One control refused to take part, one was not able to cooperate because of a recent cerebral hemorrhage and one could not be found. Instead of these three, the next closest persons in the population registers were interviewed. When possible, the patient and the corresponding control were interviewed at about the same time. Questionnaire and Gathering of Data The patients and controls were all interviewed in person by one of the authors (R.L.), according to a strictly standardized questionnaire. The questions were addressed to each subject in the same way and with the same words. Information was obtained concerning all occupations during the person's life-time, and of exposures to specific chemicals. A list of professions, in accordance with the principles of the Swedish Central Bureau of Statistics, was checked systematically with every subject to make it possible for everyone to recall former occupations. The questionnaire also contained questions about hobbies, tobacco use, use of medicines, exposure to x-rays, vaccines, former diseases, and living conditions. The educations and the professions of the subjects were given in chronological order to describe the total life-span up to the interview. They were asked to describe working instructions and environmental exposure concerning inhalation, skin exposure and ingestion. They were also asked about different tasks within occupations and the way they performed the work. Different activities within the profession of painters can easily be described and it is unlikely that the subjects did not recall their activities. Exposure to organic solvents, paints and solvents, are uniformly used all over the country and well known by their users. The Swedish National Board of Health and Safety has performed analyses of their content. As far as possible, quantitatively and qualitatively exact answers that could be directly coded for computer processing, were obtained. The chemical agents and the professions were listed and coded for computing purposes. The questionnaire dealt with professional and nonprofessional activities. These TABLEI . Definition of Odds Ratio* Patients Referents No. of pairs Exposed Exposed Nonexposed Nonexposed Exposed Nonexposed Exposed Nonexposed a b C d is calculated as the ratio ofthe numberof discordant patients and controls. two parts are not combined in the analyses or in the presentation of the results, except in table 6 concerning frequencies of skin exposure. Epidemiological and Statistical Methods The analyses of the material have been performed by comparing bound patient-control pairs. Odds ratios (OR)were estimated (Table 1) and tested for the discordant pairs according to Breslow and Day.'"I4 Concordant pairs, pairs with solvent exposure of both patient and control, are disregarded. Multivariate analysis with logistic regression on matched data was also p e r f ~ r m e d . ' ~ Dejnitions Constant daily occupational exposure to paints and/ or organic solvents and/or organic solvent-containing glue(s) during a defined period of time in life is termed professional exposure. Skin exposure is defined as skin cleaning with aromatic and aliphatic solvents. Therapeutic x-ray is defined as low-dose gamma irradiation forjoint or back pain, e.g., 2500 rad as surface dose four times as a course of treatment. Petroleum products is defined as fuel (diesel, gasoline, aircraft fuels) and its combustion products. Results Painters Thirteen patients and one control had been painters. All painters reported daily exposure to organic solvents by vapours. All of them cleaned their hands with solvents containing a mixture of aromatic and aliphatic hydrocarbons. The majority, 1 1 painters and the control, were exposed by direct skin application one or more times daily, while two painters were exposed at least by direct skin application every week and by other means of skin cleaning daily (Table 2). The estimated odds ratio was 13 with a 95% confidence interval (CI) (2.0 to 554). The median duration of exposure for painters was 16 years (range, 1 to 50 years) and the 1380 CANCERSeptember 15 1987 Vol. 60 ~~ Patient & control no. ~ Age*/sex TABLE2. Data Gathered by Questionnaire From Painter Patients Type of painters Occupation at the time of diagnosis Solvent exposure (Wi Interval from end of exposure to start of disease* Skin exposure to solvents Smoking duration years Type of leukemia (FAB) Patients 1 2 3 4 5 6 I 85 9 105 11 12 13 Control 1 41/F 72/M 32/M 20/M 53/M 70/M 19/M 58/M 59/M 66/M 45/M 73/M 71/M 39/M Spray painter Painter + hobby artist painter Car painter Painter (metal, warm machines) +Painter, rubber factory paint manufacturer Painter + hobby artist painter Painter Boat painter Car painter Boat and building painter Boat painter Spray painter (anticorrosive agents) Warm and cool asphalt painter Shoe factory (glue) Painter Car tester Railway, shunting yard worker Paint manufacturer Painter Painter Salesman Baker Pensioner Clerk Pensioner Pensioner Car painter Assistant nurse ALL: acute lymphocytic leukemia: FAB: French-American-British classification of the acute leukemias. * Median age, 58 years. 18 0 Daily 0 ALL 50 0 Daily 49 ALL 16 1 Daily 17 ALL 1 2 Daily 0 M 2 18 0 Daily 30 M4 50 0 Daily 46 M4 3 0 Daily 0 ALL 16 25 Daily 46 MI 3 7 Weekly 43 M2 10 28 Daily 0 ALL 4 22 Daily 19 M4 5 6 Weekly 40 M2 47 4 Daily 50 M5 6 12 Daily 0 *t Median exposure, 16 years (range, I to 50 years). Median interval. 2 years (range, 0 to 28 years). 15 Exposed to therapeutic X-ray treatment to shoulders. - median interval from the end of exposure to the development of disease was two years (range 0 to 28 years). Five were occupationally exposed at the time of leukemia diagnosis. Patients 8 and 10 had received x-ray treatment, Patient 8 twice, eight years (right shoulder) and six years (left shoulder and back), before the diagnosis of leukemia and Patient 10,eight years before (one shoulder) diagnosis. Excluding the two individuals exposed to x-rays, OR = 1 1 ( P = 0.012, sign test). The shortest exposure to solvents ( I year) was seen in one young painter (Patient 4) who painted warm machines with anticorrosive agents and lacquers. Two painters (Patients 2 and 6) also painted as a hobby (artists) with a long duration of exposure prior to the diagnosis of leukemia. Nonprofessional Painting There was no overall statistical differencebetween the nonprofessional painting activities of patients and controls. In fact 77 patients and 89 controls had been exposed to solvent-containing paints. The intensity of the nonprofessional exposure was ranked from single occasions in life to periods of once a year, once a month, once a week and daily painting. There was no significant statistical difference between the patients and controls as regards the low-grade exposure, but the more intense exposure, with frequencies of once a week to daily activity, was seen significantly more often among discordant pairs of patients compared to controls ( P = 0.03) (Table 3). Exposure by inhalation of solvent-containing air from restoration of the patients' and controls' homes during single brief periods during the previous 5 years was not significantlydifferent in the two groups. Organic Solvent Exposure in D@nnt 0ccicpulion.r Organic solvents are used as constituents of paints, glues, and cleaning fluids in many different professions (Table 4). If excluded all patient-control pairs of painters, 26 patients and seven controls were occupationally exposed to paints and/or solvents and/or glues during a certain period of life, disregarding four exposed -No. 6 RISKOF ACUTELEUKEMIAAFTER PAINTEXPOSURE Lindquist et al. 1381 TABL3E. Frequency of Nonprofessional Painting Activities Amone Patients and Controls Frequency of activity No. of discordant pairs of patients No. of discordant pairs of controls P value of comparison between discordant pairs of patients and controls Daily Weekly Monthly Yearly Once Never exposed 3 6 6 21 23 35 1l l 6 24 39 23 0.03* NS NS NS NS NS NS: not significant. * One-tailed significance test. concordant pairs. The ratio 26 to 7 is highly significant (P= 0.0013) when tested for the equality of proportions in matched samples (by sign test, 1947).15The OR of the overall exposure to solvents is estimated to be 3.7, with the 95%CI limit calculated by binomial distribution to be 1.6 to 10.1 (Table 5). The median time of exposure of discordant patients was longer than that of discordant controls (nine years compared to four years). The median time from cessation of exposure to the development of leukemia was 15 years (range, 0 to 40 years). Median time from cessation of exposure to interview for controls was 16 years (range 0 to 24). Skin Exposure to Organic Solvents The distribution of patients and controls regardingthe frequency of skin cleaning with organic solvents show that there was no difference in low frequency exposure, but daily exposure was noted in 18 patients and six controls (Table 6). They were using white spirit 17%to 22% aromatic hydrocarbons) and/or gasoline with a high aliphatic hydrocarbon content and about 5% aromatic hydrocarbons. The aromatic hydrocarbon fraction is composed of xylene, toluene, trimethylbenzene, and benzene.16 TABL4E. Distribution of Occupations Exposed to Organic Solvents According to the Interviews Occupation No. of discordant patients No. of discordant controls Painter Rubber industry worker Pharmaceutical industry worker Upholsterer Car wax cleaner Printing worker Dry cleaner Car repairman Woodworker Shoe factory worker Book binder Textile printing worker Train/bus wagon repairman Shipyard worker Sign painter New building cleaner Isolation worker Metal industry painting worker Dispensing chemist Nurse 13 1 I 1 1 I I 3 4 2 I I 2 1 1 I I 1 I I 1 0 0 1 0 I 2 0 0 1 0 0 1 0 0 0 1 0 0 0 Ether Anesthesia The exposure to ether anesthesia was the same in patients and controls (58 patients exposed compared with 57 controls; discordant pairs, 27 patients and 26 controls). The solvent-exposed patients were not significantly more often exposed to this form of anesthesia than other patients. Smoking Habits Thirty-one percent of the painters had not been smokers. Considering all patients and controls, the smoking habits were similar (Table 7). Forty-two percent of all the patients and 48%of the controls had not been smokers. Neither differences between discordant pairs nor strata of patients and controls proved to be of statistical significance when the duration of smoking, TABL5E. Exposure to Organic Solvents* Patients Controls Occupational organic solvent exposure Time of exposure (yr) Time from end of exposure to diagnosis (yr) Time of exposure (yr) Time from end of exposure to interview (yr) No. Median Range Median Range No. Median Range Median Range P OR (CI) Painters 13 16 1-50 2 0-28 1 Professions other than painters 26 9 0.1-60 15 0-40 7 Total 39 10 0.1-60 7 0-40 8 6 12 0.002 13 (2-554) 4 1.5-40 16 0-24 0.001 3.7 (1.6-10.1) 5 1.5-40 14 0-24 t0.001 4.9 (2.2-12.1) OR: odds ratio; CI: confidence interval. * Number of patients and controls exposed to organic solvents, me- dian time of exposure (years/range), median time from end of expo- sure to diagnosis/interview (years/range), OR with CI (binomial distribution) and P values. I382 CANCERSeptember 15 1987 Vol. 60 TABLE6. Frequency Skin Exposure of Professionals and Nonprofessionalsto Organic Solvents Frequency of exposure No. of No. of patients controls P OR CI (binomial distribution) Daily Weekly Some occasions/year Single occasions/year Single occasions Never exposed 18 12 34 12 18 31 6 <0.05 3.0 (1.14-9.24) I I NS 41 NS 10 NS 20 NS 37 NS OR: odds ratio; CI: confidence interval; NS: nonsignificant number of cigarettes consumed, pipe smoking, or snuff use were tested. Multivariate Analysis Multivariate analysis expressed as logistic regression on matched data showed no interaction between organic solvent exposure, smoking habits, x-ray exposure and petroleum products (Table 8). FAB Classes The distribution of French-American-British classification of the acute leukemias (FAB) classes'`,I2 among patients exposed to organic solvent-containing compounds showed that no single patient had the diagnosis of promyelocytic leukemia (M3) or erythroleukemia +(M6) (Table 9). Leukemias of the MI M2 types were found significantly more seldom among solventexposed than among non-exposed patients ( P = 0.01, chi-square test) and the M4-M5 leukemias were found significantly more often among exposed patients ( P = 0.02). Discussion Our results indicate that regular exposure to solvents is associated with an increased risk of developing acute leukemia. Heavily exposed persons, such as painters, exhibit the greatest OR ( 13), while solvent-exposed persons other than painters have an excess risk of 3.7 when compared with nonexposed persons. Multivariate analysis showed no interaction between solvent exposure, to- TABLE7. Tobacco Use (During any Period of Life) by Patients and Controls* Category No. of patients No. of controls Cigarrette smokers 63 62 Pipe smokers 20 20 Snuff takers 18 19 Nonsmokers 52 60 ~~ ~ * The same individual may have consumed tobacco in different forms. TABLE8. Multivariate Analysis ~~ Variable B P Therapeutic x-ray Organic solvents 1.51 0.008 I .96 <o.oo I Petroleum products 1.42 0.020 6:regression coefficient." * Multivariate analysis using logistic regression on matched data showed no interaction between exposure to therapeutic x-ray treatment, solvents, or petroleum products. All variables were coded 0- I :0, not exposed; I , exposed. Another considered (NS) variable was tobacco use. bacco use, therapeutic x-ray exposure, or petroleum products. The solvents used occupationally by painters consist of mixtures of aromatic and aliphatic hydrocarbons.16 Although knowledge of the toxic effects of benzene has resulted in its prohibition in Sweden as a pure solvent, it is impossible to prevent complete prohibition. It often occurs as an impurity in other solvents, in paints and is used as an additive in gasoline. Benzene is metabolized to benzene epoxide, a highly reactive compound, which in turn is metabolized to phenol, which is oxidized to alkylating agents.''.'*This could be one mechanism for cellular damage. For example, Snyder et al. have shown that a benzene metabolite is bound to bone marrow cells," and Karigaya et al. have noted stemcell injury leading to diminished cell replication and other cellular abnormalities in in vitro liquid bone marrow cell cultures exposed to benzene.20 Toluene (methylbenzene) and xylene (dimethylbenzene) can be metabolized by the same pathway but to a lesser extent, because they are also metabolized on methyl groups into carboxyl acids, which can be conjugated and excreted.21Human leukemia and aplasia of the bone marrow following inhalation of vapor from toluene-containing glue was observed in youngsters assembling model building s e t ~ . * ~It- *is~therefore proba- TABL9E. Distribution of FAB Classes Among Patients Exposed to Organic Solvents and Nonexposed Patients* Nonexposed Exposed All patients No. Percent No. Percent MI-M2 M3 M4+M5 ALL AUL 42 4 20 13 3 82 78 12 100 0 51 19 52 12 100 0 66 43 22 0 49 48 0 34 No. Pt 54 0.0I 4 0.36 39 0.02 25 0.12 3 0.55 125 ALL: acute lymphocytic leukemia; AUL: acute undifferentiated leukemia: FAB: French-American-British classification of the acute leukemias. * A comparison of the number of patients, exposed and non-exposed of the different FAB groups. t Chi-square test. -No. 6 RISKOF ACUTELEUKEMIAAFTER PAINTEXPOSURE Lindquist et ul. 1383 ble that the metabolism to epoxides and phenols is suf- ficient for the induction of bone marrow cell damage and leukemia. Cytogenetic analyses of human lymphocytes have in- dicated mutagenic properties of and tri- ~hloroethylene.A~l~so, increased chromatid exchange and chromatid breakages were found in laboratory and printing works employees who were exposed to solvents as compared to nonexposed controls.30Chromosome breakages are increased in blood lymphocytes of plastic factory workers exposed to ~tyrene.T~h' us, a substantial amount of data shows that many organic solvents and their metabolites are mutagenic. This may explain why exposure to them increases the risk of developing leuke- mia. However, the mechanism behind the mutagenic action is complex, and there is little knowledge about the relative risk of each specific organic solvent or its metabolite/s. For painters, the median time of solvent exposure was 16 years (range, 1 to 50 years), but the median interval from the end of exposure to onset of disease was two years (range, 0 to 28 years). Five painters acquired leu- kemia while still engaged in professional activity after 18, 18, 50, 50, and 3 years of exposure, respectively. Thus, the results suggest that there is an excess risk after a short period of exposure. Furthermore, we find that onset of the disease can occur in association with the exposure as well as after cessation of exposure. The same result was obtained by Rinsky et al. who studied leuke- mia in benzene workers.32 Organic solvent exposure has been associated with the development of both myelocytic and lymphocytic leu- kemias and lymphomas. In a report from T ~ r k e y 'o~n 34 patients chronically exposed to benzene, Aksoy et al. found both myelocytic and lymphocytic leukemias. McMichael et al. reported a sevenfold increase in the relative risk of developing acute lymphocytic leukemia in workers exposed to solvents.6Two case-control stud- ies, by Hardell ef and by Olsson and Brandt,35show an increased risk of developing lymphoma after expo- sure to organic solvents. This was also found by Ohlin and A h l b ~ mco~nc~erning cancer mortality among Swedish chemists exposed to solvents. Aksoy et al.33re- ported an incidence of leukemia of 13.5 per 100,000 individuals with chronic exposure to benzene, com- pared to the overall incidence of six per 100,000, a sig- nificantly greater risk (P < 0.001), when exposed. The OR was 13.5/6 = 2.25. When estimating the OR of malignant lymphoma, Hardell et al.34found that in pa- tients with high-grade exposure to organic solvents, the OR was 1 1.1, 95% CI (1.9 to 1 1.4). The high-grade ex- posed group is comparable with the group of painters in our study, whose OR = 13 (1.95 to 554). The results are consistent, suggesting the estimation of the overrisk to be in the lower span of the confidence interval. A comparison of the distribution of FAB classes (Table 9) among our patients exposed to organic solvents and nonexposed persons indicates that the FAB M4 and M5 subtypes of leukemias are more frequently observed among exposed persons (P = 0.02, chi-square test) and FAB M1 and M2 subtypes less frequently (P = 0.001, chi-square test). A Swedish death certificate cohort study of 4 16 paint- industry workers exposed to organic solvents, has been presented by Lundberg et al.37 Cause-specific mortality was studied. Among 96 workers who died, three died of multiple myeloma, compared with 0.6 expected deaths. Five deaths of disease of the lymphatic and hematopoietic systems were observed compared to 2.4 expected deaths with an estimated OR of 2.12, 95% CI (0.68 to 4.96). The alkylating agents of metabolized aromatic compounds, such as epoxids, have an effect most probably consistent with cancer initiator, inducing DNA damage and turning the normal cell into a malignant cell. An initiating agent, if sufficiently strong, can theoretically express its capacity after a single exposure, thus explaining the occurrence of leukemia in individuals with a short exposure time. The same DNA-damaging property can also induce chromosome abnormalities which may predispose second event/hit to induce the malignancy. The mechanisms of bone marrow carcinogenesisgiving rise to acute leukemias is still obscure. In the case of alkylating agents, the situation seems to have much in common with the experimental model for rat liver carcinogenesis, called the resistant hepatocyte m ~ d e I . ~ * * ~ ~ In this model, initiated cells are generated by the use of mutagens in the presence of cell proliferation. The initiated cells are selectively favored in their proliferation during the promotion phase because they are resistant to the mitoinhibitory effect of the p r o m ~ t o rT. ~hi~s differential sensitivity concentrates the mitogenic power to the initiated cells, permitting them to clonally expand to focal cell populations with a high rate of basal cell proliferation and a high risk of rare mutation-like events that pushes them another step forward in the process of cancer development. In the bone marrow, alkylators form DNA adducts that probably give rise to mutationlike events in rare cells. Alkylators have also been shown to be mitoinhibitors for bone marrow stem cells,4' depressing the production of leukocyte and thrombocytes. Thus, it is possible for resistant cells to develop (mutation/initiation) and the prerequisites (organic solvents), are at hand matching the analogy with experimental models obvious. In experimental carcinogenesisand in other examples of human cancer development, the process is slow and takes a considerable part of the lifespan to reach the terminal stage of malignant transformation. During that time, the subject does not have to be to the 1384 CANCERSeptember 15 1987 Vol. 60 carcinogen exposure constantly. In the optimized exper- imental liver models, dependent of both initiation events and promotion, initiation is a one-dose event and promotive treatment is performed for 2 weeks. Carci- nomas in this model are seen 10 months after initiation. Potent carcinogen exposure in young animals with a high rate of liver cell proliferation results in hepatoma development within 2 to 3 months. Our results point to a greatly increased risk of devel- oping acute leukemia after exposure to organic solvents. Furthermore, noxious effects of these compounds on the l u d n e y ~t,h~e ~nervous system43and the liver4 have been reported. Today, occupational exposure to organic solvents is widespread and complex, and it is important that further efforts be made to minimize such exposure. REFERENCES I . CG Santesson. Ueber chronische Vergiftungen mit Steinkohlenentheer benzene vier todesfalles. Arch Hyg Berl 1897; 3 1:336-376. 2. Delore P, Borgomano C. Leudmie aigue en course de I'intoxication benztnique sur I'origine toxique de certaines leu&mies aigues et leur relations avec les anemies graves. J M e d Lyon 1928; 9:227. 3. Lignac GOE. Die Benzol Leukemia bei Menschen and weissen Mausen. KIin Wochenschr 1932; 12:109-110. 4. Snyder CA. Goldstein BD, Sellakumar AR, Bromberg I, Laskin S, Albert RE. The inhalation toxicology of benzene: Incidence of hematopoietic neoplasms and hematotoxicity in AKR/J and C57BL/6J mice. Toxicol Appl Pharmacol 1980; 54:327-33 1. 5. Infante PF, Rinsky RA, Wagoner JK, Young RJ. Leukaemia in benzene workers. 1977;2:76-78. 6. McMichael AJ, Spirtas R, Kupper LL, Gamble JF. Solvent exposure and leukemia among rubber workers: An epidemiologic study. J Occup Med 1975; 17:234-239. 7. Morgan R, Claxton KW, Kaplan SD, Parsons JM, Wong D. Mortality of paint and coatings industry workers. J Occup Med 1985; 271377-378. 8. Flodin U, Andersson L, Anjou C-G, Palm U-B, Vikrot 0. A case-referent study on acute myeloid leukemia, background radiation and exposure to solvents and other agents. Scand J Work Environ Health 1981; 7:169-178. 9. Mitelman F, Brandt L, Nilsson PG. Relation among occupational exposure to potential mutagenic carcinogenic agents, clinical findings and bone marrow chromosomes in acute nonlymphocytic leukemia. Blood 1978; 52:1229- 1237. 10. Paul C, Killander A, Simonsson B et a/. Prolonged survival of patients with acute nonlymphoblastic leukemia (ANL) after intensive induction and consolidation treatment with combinations containing doxorubicin (DOX) or doxorubicin-DNA. A preliminary report (Abstr). XX Congress of the International Society of Hematology, Buenos Aim, 1984. 11. Bennet JM, Catovsky D, Daniel M-T et al. Proposals for the classification of the acute leukaemias. Br J Haematol 1976; 44:451458. 12. Ost A, Lagerlof B, Sundstrom C. A study of reproducibility of the diagnostic criteria for acute leukemia. Scand J Haematol 1983; 3 1:257-266. 13. Breslow NE, Day NL. Statistical Methods in Cancer Research, vol. 1. Lyon: IARC, 1980. 14. Lentner C. Geigy Scientific Tables, vol. 2. Bask: CIBA-GEIGY Ltd, 1982. 15. Siege1 S. Nonparametric Statistics for the Behavioral Sciences. New York: McGraw-Hill, 1956. 16. Bygghasan. Arbetshygieniska problem vid mileriarbete--l6sisningsmedelsingor frin olika fargprodukter. Byggforlaget, Stockholm, 1975. 17. Rusch GM, Lcong BKJ, Laskin S. Benzene metabolism. J Toxicol Environ Health 1977; 2(Suppl):23-36. 18. Snyder R, Kocsis JJ. Current concepts of chronic benzene toxicity. Crit Rev Toxicol 1975; 3:265-288. 19. Snyder R, Lee EW, Kocsis JJ. Binding of labeled benzene metabolites to mouse liver and bone marrow. Res Commun Chem Pathol Pharmacol 1978; 20:191-194. 20. Harigaya K, Miller M, Cronkite EP, Drew RT. The detection of in vivo hematotoxicity of benzene by in vitro liquid bone marrow cultures. Toxicol Appl Pharmacol 1981;60346-353. 21. Birgersson B, Sterner 0,Zimerson E. Kemiska Halsorisker. 194. 22. Storb R, Thomas ED, Buckner CD, Clift RA. Allogeneic marrow grafting for treatment ofaplastic anemia. Blood 1974;43:157- 180. 23. Isager H. Fatal aplasia of the bone-marrow following inhalation of the vapour from toluene-containing glue. Ugesk Laeger 1975; 137:2197-2198. 24. Caligiuri MA, Early AP, Marinello MJ, Preisler HD. Acute nonlymphocytic leukemia in a glue sniffer. Am J Hernatol 1985; 20~89-90. 25. Morimoto K. Combined cytogenetic effects of benzene and radiation on cultured human lymphocytes. Jpn J Ind Health 1975; 17:106-107. 26. Khan H, Khan MH. Cytogenetische Untersuchungen bei chronischer Benzol exposition. Arch Toxicol 1973; 31:39-49. 27. Pollini G, Colombi R. I1 danno cromosomico del linfociti nell 'emopatia benzenica. Med Lav 1964; 55:641-655. 28. Forni A, Cappellini A, Pacific0 E, Vigliani EC. Chromosomes changes and their evolution in subjects with past exposure to benzene. Arch Environ Health 1971; 23:385-391. 29. Konietzko H, Haberlandt W, Heilbronner H, Reill G, Weinhardt H. Cytogenetische Undersuchungen in Triclorathylen-Arbeitern. Arch Toxicol 1978;40:20 1-206. 30. Funes-Cravito F, Kolmodin-Hedman B, Lindsten J, Nor- denskjold M. Chromosome aberrations and sister-chromatid exchange in workers in chemical laboratories and a rotoprinting factory and in children of women laboratory workers. Lancet 1977;2:322-325. 31. Meretojn T, Jarventans H, Sossu M, Vainio H. Chromosome aberrations in lymphocytes of workers exposed to styrene. Scand J Work Environ Health 1978; 2(Suppl): 259-264. 32. Rinsky RA, Young RJ, Smith AB. Leukemia in benzene workers. A m J IndMed 1981;2:217-245. 33. Aksoy M, Erdem S, Dingo1 G. Types of leukemia in chronic benzene poisoning: A study in thirty-four patients. Acta Haematol 1976; 55~65-72. 34. Hardell L, Eriksson M, Lerner P, Lundgren E. Malignant lymphoma and exposure to chemicals, especially organic solvents, chlorophenols and phenoxy acids: A case-control study. Br J Cancer 1981; 42:169-176. 35. Olsson H, Brandt L. Occupational exposure to organic solvents and Hodgkin's disease in man. Scand J Work Environ Health 1980; 6:302-305. 36. Olin H, Ahlbom A. The cancer mortality among Swedishchemists graduated during three decades. Environ Res 1980; 22: 154- 161. 37. Lundberg L. Mortality and cancer incidence among Swedish paint industry workers with long-term exposure to organic solvents. Arbetarskydds-styrelsen 1985; 38:1-18. 38. Farber E. Cellular biochemistry of the stepwise development of cancer with chemicals: GHA Clowes Memorial Lecture. Cancer Res 1984; 44:5463-5474. 39. Farber E. The multistep nature of cancer development. Cancer Res 1984;44:4217-4223. 40. Eriksson LC. Aspects ofdrug metabolism in hepatocyte nodules in rats. In: Matern S, Bock KW, Gerok W, eds. Advances in Glucouronide Conjugation. Lancaster, England:MTP Press Ltd, 1985; 295-308. 41. Snyder CA, Green JD, Lobue J, Goldstein BD, Vaile CD, Albert RD. Protracted benzene exposure causes a proliferation of myeloblasts and/or promyelocytes in CD-I mice. Bull Environ Contam Toxicol 1981; 27~17-22. 42. Ravnskov U, Lundstrom S, NordCn A.Hydrocarbon exposure and glomerulonephritis: Evidence from patients' occupations. Lancet 1983;2: 1214-12 16. 43. Axelson 0, Hane M, Hogstedt C. A case-referent study on neuro-psychiatric disorders among workers exposed to solvent. Scand J Work Environ Health 1976; 2:14-20. 44. Dossing M, Arlien-Soborg P, Milling Petersen L, Ranek L. Liver damage associated with occupational exposure to organic solvents in house painters. Eur JClin Invest 1983; 13:151-157.