Document qkDEwJEeDY1DMppDbwOQJzGyG

ICANCERRESEARCH54, 2386-2389,May1, 1994] Exposure to Phenoxyacetic Acids, Chiorophenols, or Organic Solvents in Relation to Histopathology, Stage, and Anatomical Localization of Non-Hodgkin's Lymphoma Lennart Hardell,' Mikael Eriksson, and Anna Degerman Department of Oncology, Orebro Medical Center, 5-701 85 Orebro [L. H.]; Department of Oncology, University Hospital, S-PAll 85 Umed [M. E.J; and Department of Occupational Medicine, University Hospital, 5-581 85 LinkOping [A. D.J, Sweden ABSTRACT MATERIALS AND METHODS Results on 105 cases with histopathologically confirmed non-Hodgkin's The study material consisted of all men aged 25"85years who were lymphoma (NHL) and 335 controls from a previously published case admiued to the Department of Oncology in Umea between 1974 and 1978 with control study on malignant lymphoma are presented together with some histopathologically verified NHL, 105 cases in total. The slides were previ extended analyses. No occupation was a risk factor for NHL Exposure to ously reexamined and are presented here with the Rappaport classification, phenoxyacetic acids yielded, in the univariate analysis, an odds ratio of 5.5 which was commonly used in Sweden during the 1970s. At the present time, with a 95% confldence interval of 2.7"11M. ost cases and controls were the Kid classification is usually used in Europe, and the Working Formulation exposed to a commercial mixture of 2,4-dichlorophenoxyacetic acid and is used in the United States. These three coding systems have been related to 2,4,5-tricifiorophenoxyacetic acid. Exposure to chlorophenols gave an each other and are comparable (9). odds ratio of4.8 (2.7"8.8w) ith pentachlorophenol being the most common The NHL cases in our study were classified according to the Ann Arbor type. Exposure to organic solvents yielded an odds ratio of 2.4 (1.4"3.9). staging system. In addition, the records were scrutinized in order to classify the These results were not significantly changed in the multivarlate analysis. disease at presentation as supradiaphragmatic, infradiaphragmatic, or general Dicifiorodiphenyltrichloroethane, asbestos, smoking, and oral snuff were ized. To evaluate prognosis, the records of all 105 NHL cases were scrutinized. not associated with an increased risk for NHL. The results regarding As controls, all 335 control subjects who had inititally answered the ques increased risk for NHL following exposure to phenoxyacetic acids, chlo tionnaire were used. They were matched for sex, age, place of residence, and rophenols, or organic solvents were not affected by histopathological type, vital status; deceased controls were also matched for year of death. The living disease stage, or anatomical site of disease presentation. Median survival controls were drawn from the National Population Registry and the deceased was somewhat longer in cases exposed to organic solvents than the rest. from the National Registry for Causes of Death. Details were presented in our This was explained by more prevalent exposure to organic solvents in the previous publication (1). group of cases with good prognosis NHL histopathology. Assessment of Exposure. A complete working history, various exposures, and leisure time activites covering the whole life were obtained by question INTRODUCTION In 1981, we reported a 6-fold risk of malignant lymphoma, both naire. If necessary, the answers were completed over the phone. Protective equipment against the chemicals associated with NHL in this study was not commonly used, and information on such protection and how it was used is Hodgkin's disease and NHL,2 among persons exposed to phenoxy uncertain. Details have been published in the earlier report (1). acetic acids or chlorophenols as the result of a case-control study (1). An association was also found for exposure to organic solvents. Since then, other studies have been performed within this field and an association between exposure to the phenoxy herbicide 2,4-D and NHL has been reported in studies from the United States (2, 3). NHL is one of the malignant diseases with the most rapidly in creasing incidence in the Western world. Pesticide exposure has been discussed as one contributing factor (4). This has prompted us to analyze and report data on NHL from our study (1), omitting All occupations were now classified according to the Nordic Working Classification system (10). The results are presented only for occupations with at least 10 subjects. Since UV radiation has been suggested to be a factor to be considered in the etiology of NHL (11), occupations were also classified as outdoor, indoor, or mixed. With regard to exposure to chlorophenols or organic solvents, exposure less than 1 week continously or less than 1 month in total was classified as low-grade and more than that as high-grade. Overall survival was analyzed for the cases. Calculations were performed for exposed and unexposed cases separately. Hodgkin's disease. Statistical Methods. The statistical analyses of the data were based on In this paper, we also present risk estimates for different histopatho logical subtypes, stage, and anatomical localization of NHL. The association between 2,4-D and NHL in a study from Eastern Nebraska did not appear to be specific to any subgroup, although the risk was somewhat higher in intermediate grade NHL (3). Exposure to organic solvents has been related to only supradiaphragmatic NHL in one study (5), whereas other studies have not provided information on tumor localization (6, 7). Mantel-Haenszel methods for calculation of OR by stratification by age and vital status (12). The 95% confidence intervals, given within parentheses, were calculated according to Cornfield as presented by Breslow and Day in 1980 (13). The analyses were performed with the Epilog Plus epidemio logical software, both the univariate analyses and the multivariate logistic regression. Potential confounding between exposures of interest was eval uated in the multivariate analysis. The included variables are presented under results. Some studies in this field have used occupation as a surrogate for exposure to phenoxy herbicides without showing any association (8). RESULTS In order to compare our results with the findings in such studies, we have calculated an OR for different occupations, including those with potential exposure to phenoxyacetic acids or chlorophenols, such as farmers, farm workers, afforestation workers, and lumberjacks. Occupations. No significantly increased or decreased risk for NHL was found in any occupation as presented in Table 1. Outdoor work yielded OR 0.9 (0.4"1.7),and for mixed outdoor/indoor work, OR = 0.9 (0.5"1.7)was obtained with indoor work used as a reference. Received 11/2/93; accepted 2/24/94. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Phenoxyacetic Acids. Exposure to phenoxyacetic acids or chioro phenols gave OR 4.6 (2.7"7.8;Table 2). Exposure to all types of phenoxyacetic acids yielded OR 5.5 (2.7"11).No case was exposed to I To whom requests for reprints should be addressed. 2 The abbrevations used are: NHI, non-Hodgkin's lymphoma; 2,4-D, 2,4-dichloro phenoxyacetic acid; 2,4,5-T, 2,4,5-trichlorophenoxyacetic acid; OR, odds ratio; TCDD, 2,3,7,8-tetrachlorodibenzo-p-dioxin; DDT, dichlorodiphenyltrichloroethane. only 2,4,5-T, whereas one control reported 2,4,5-T as the only phe noxyacetic acid. Mostly, a combination of 2,4-D and 2,4,5-T had been used in both occupational and leisure time exposure. The combination 2386 PESTICIDE EXPOSURE AND NHL Table 1 OR and 95% confidence interval in different occupations for non-Hodgkin lymphoma numbers of exposed controls. Exposures of 1"17days yielded OR 6.5 (2.1"18)1;8"43days, OR = 3.3 (1.0"11)a;nd >43 days, OR = exposed (CasesO/ccocnutproatliso)OnNRuCm1b9e5rAfforestation 7.3 (2.2"23). Median latency time period for cases exposed to phenoxyacetic wwoorrkkeerr71/330/04.7700..830.4"1.7"B1u.i7ldCinagrpenter13/361.10.6-2.4Driver3/140.70.2"2.6Farmer201740a.c7i0d.s4w"1a.s4F1ar8myears. Chlorophenols. High grade exposure to chlorophenols yielded OR 9.4 (3.6"25),whereas low grade exposure gave OR 3.3 (1.6"6.8). worker4/190.60.2"2.1Loader2/90.70.1"3.5Lumberjack32/861.20.7"M2.o1sMtelychapneinct5ac/h72lo.r5o0p.h7eno"l9.6hMoastobreen used in Sweden, and high grade exposure produced OR 8.8 (3.4"24). One case with high grade chiorophenol exposure could not specify the type used. Median la mechanic7/201.30.5"3.4Miner6/181.00.4"2.9Officer4/111.40.3"4.9Office tency period was 21 years. Organic Solvents. High grade exposure to organic solvents employee1/110.30.1"2.1Pulp workeinr6d/u1s3tr1y.50.5"43Salesman2/130.40.1"2.2Sawmill yielded OR 2.9 (1.6"5.6),whereas the risk for low grade exposure was lower, OR 1.8 (0.8"3.8).For different types of organic solvents, worker13/291.50.7"3.2Shop keeper1/90.30.2"2.7Smelting-house worker4/62.30.5"9.5Storeman5/131.30.4"4.3Truck significantly increased risk was found for benzine, thinner, trichloro ethylene, and white spirit. Exposure to degreasing agents gave OR 11 (2.9"72). The median latency period for exposure to organic solvents driver4/200.60.2"2.0a @ confidence interval. was 21 years. Other Agents. In the univariate analysis, exposure to DDT was associated with NHL, OR 2.4 (1.2"4.9).Smoking, use of oral snuff, or exposure to asbestos yielded no significantly increased ORs. Table 2 OR and 95% confidentcheeuinntievravraial tfeor exposure to different agents in analysisNumber Multivariate Analysis. In the multivariate analysis, exposure to phenoxy herbicides, high grade exposure to chlorophenols, and or exposedAgent C195aPhenoxyacetic (Cases/controls) OR ganic solvents gave signficantly increased ORs (Table 3). DDT cx posure yielded OR 1.5 (0.6"3.6).As in the univariate analysis, low 2.7"7.8Phenoxaycaicdesticor chlorophenols 47/51 4.6 grade exposure to organic solvents and asbestos gave no signficantly acidsall 2.7"112,4.5-1 1.9"8.i6n2c,l4u-dDing mixtures 1.2"o3n6l0yMCPA 4/0Chloropohnelynolsall 25/24 18/23 3/1 5.5 4.1 13 increased risks for NHL. Tableth3eAOgRenatnOdR9C5]%95caoPnhfeidneonxcyeacetic for exposure to multitvearvriaalte analysisdifferent agents in 2.7"8.8high 3.6"g2r5adloew 1.6gr"a6d.e8pentachlorophenol, 3.4"24Organic high grade solventsall 1.4"3.9high 1.6g"r5a.d6elow 0.8gr"a3d.e8benzine 1.8"730- t1h.4inne"r10trichloroethylene 1.3"42turpentine 0.9"17white 1.3"sp8i.3ridt egreaser 2.9"72DDT 35/35 16/9 19/26 15/9 45/88 31/50 14/38 3/1 11/14 4/4 5/7 12/20 7/4 4.8 9.4 3.3 8.8 2.4 2.9 1.8 28 3.4 7.2 3.3 3.2 11 acids5.21.6"17 Chlorophenols highgrade 1.2l-o6w.5Ogrgaadnei9c.0 solvents highgrade 0.5l-o2w.7DgDraTd0e3.6.5-3.6 2.82.9"28 1.11.7"7.1 1.5 Asbestos a C195, 95% confidence 1.3 0.5"3.6 interval. MCPA, 4-chloro-2-methylphenoxyacctic acid. 1.2"4.9Asbestos 0.8-4.0Smokingcurrent 17/26 11/21 2.4 1.7 Table 4 OR and 95% confidence intervalfor exposure to different agents related to different types of non-Hodgkin lymphoina according to the Rappaport classification. Number (n) of cases in different subtypes given within parentheses. 0.6"2.0previous 0.5"2.0Oral 37/115 25/80 1.1 1.0 0.9s"n2u.f5fa C195, 95% confidence 35/84 1.5 interval; MCPA, 4-chloro-2-methylphenoxyacetic acid. Histopathology (n) Lymphocytic, well-differentiated (32) Diffuse lymphocytic, poorly differentiated (5) Phenoxyacetic acids 3.3 0.9"11 10 1.0"120 Chlorophenols Organic solvents 2.8 2.9 1.1"7.6 1.3"6.6 4.4 0.2"76 of 2,4-D and 2,4,5-T was the type of exposure in 17 cases and 22 controls. One case was exposed to both 2,4,5-T and 4-chloro-2methylphenoxyacetic acid. Exposure to 2,4,5-T in any combination gave OR 4.1 (1.9"8.6). Three cases and one control reported 2,4-D as the only phenoxy acetic acid producing OR 13 (1.2"357).With regard to 4-chloro-2methylphenoxyacetic acid, four cases and no controls were exposed to that herbicide alone. Dose-response for exposure to phenoxy herbicides was analyzed by dividing the number of exposed controls in three groups with equal Diffuse mixed lymphocytic/histiocy(ti4c) Diffuse histiocytic (44) Nodular mixed lymphocytic/histiocyt(i6c) Nodular lymphocytic, 0.3we"ll1-d5i0fNfeordenutlaiarted (3)5.0 lymphocytic, 0.9poo"r4ly1Ndoifdfeurleanr tiated (7)15 histiocytic (4)52 5.0 0.2"83 6.2 2.5"17 2.3 0.1"28 8.7 7.6 0.8"98 0.7"200 5.5 1.8 2.4"13 0.8"3.8 2.0 2.2 0.1"25 0.2"17 0.2"835.4 2.0"15024 3.1"2106.0 0.6"99022 2.1"6703.5 0.3"34 2387 PESTICIDEE)CPO5UREAND NHL Table 5 OR and 95% confidence interval for exposure to different agents related to different stages of non-Hodgkin lymphoma. Number(n) of casesin differentstagesgivenwithinparentheses. Stage C19(n5)I Phenoxyacetic acids OR CI95 Chlorophenols OR C195 Organic solvents OR (23) A (21) 0.5B("29)90II 4.3 3.8 8.9 1.4"16 3.2 1.0"11 2.5 0.9"6.9 1.0-14 3.5 1.1"12 2.9 1.0-8.2 (11) A (8) 0.2B("33)00III 16 13 53 2.8"110 4.6 0.8"25 0.3 0.1"23 1.9"111 3.9 0.5"27 0.4 0.1"3.7 0.6"990 7.1 (27) A (21) 0.1B("67).1IV 5.3 5.7 4.3 1.7"19 7.2 3.0"20 2.5 1.0"5.9 1.7"24 83 3.3"27 3.2 1.1"83 0.2"68 2.9 0.1"39 1.0 (44) A (26) 0.7B("168.5) I-tV 5.3 6.0 4.1 1.9"15 4.6 2.0"11 1.8"21 5.7 2.1"16 0.7"20 3.4 0.9"12 3.1 3.8 2.1 1.5"6.5 1.5"9.6 A (76) 03"33aB (29) CI95, 95% confidienntecreval.Table 53 53 2.6"12 53 3.0"11 2.8 1.6"4.9 1.5"20 3.2 1.0"93 1.4 6 OR and 95% confidence interval for exposure to different agents related to anatomical site of non-Hodgkin lymphoma at diagnosis. Number (n) of cases given within parentheses.Phenoxyacetic solventsAnatomical acids Chiorophenols Organic C195Supra-diapshirteag(mn)atic 0.5"3.9Infra-diaphragm(2a4t)ic OR 5.7 CI95a OR C195 1.8"18 3.2 1.0"93 OR 1.4 0.6"4.7Generalized(21) 7.2 2.2"28 6.0 1.9"21 1.6 1.7"5.9a (60) CI95, 95% confidence interval. 4.7 1.9"11 5.2 2.6"11 3.2 Histopathology. ORs for exposure to phenoxyacetic acids, chlo significantly increased in three studies and significantly decreased in rophenols, or organic solvents for different types of NHL according to one study. In this investigation, OR for farmers was 0.7 and for farm the Rappaport classification are given in Table 4. Increased risk was found for all subtypes. Due to the low number of cases with some of the subtypes, broad confidence intervals were obtained in these calculations. Stage and Anatomical Site Tables 5 and 6 present ORs and 95% confidence interval for different stages and anatomical sites of NHL at the clinical presentation. Increased risks for exposure to phenoxy herbicides, chlorophenols, or organic solvents were found, regardless of stage or site. Prognosis. The median survival was 21 months in the unexposed cases, 27 months in cases exposed to phenoxyacetic acids or chloro phenols (both groups), and 36 months in patients exposed to organic solvents. workers 0.6. Jobs as afforestation workers or lumberjacks yielded OR 0.7 and 1.2, respectively. These results are in agreement with Swedish register studies where census data on occupations were linked with the Swedish Cancer Register (8, 17). Thus, it is necessary to obtain individual exposure data, and job title cannot be used as a surrogate for exposure. Experimental studies indicate that UV radiation may cause sys temic immunosuppression (18), and an increased risk for NHL has been found in immunosuppressed persons (19). Consequently, the risk might be increased in job categories classified as outdoor work. Such an association was not demonstrated in this study, nor was the risk increased in specific occupations of the outdoor type (Table 1). DISCUSSION Our data showed a significantly increased risk for NHL in persons exposed to phenoxyacetic acids or chlorophenols. In Sweden, mostly Exposures were self-reported and might have been influenced by recall bias. A similar questionnaire had been used in a previous case-control study on soft-tissue sarcoma in the same area (14). In a validation procedure, the exposure information in the questionnaires was verified by the employers in a high degree, i.e., 97% for chloro phenols, and regarding phenoxyacetic acids, even higher OR was a combination of 2,4-D and 2,4,5-T was used within the forestry occupation in order to combat hard woods. 2,4,5-T was contaminated by the dioxin congener TCDD (20) and was banned in 1977. Most cases and controls had been exposed to that combination and thus also to TCDD. For 2,4-D exposure, the risk was especially high, although based on a few subjects with exposure to only 2,4-D. It is noteworthy calculated in that study if data from employers were used. These that the studies done in the United States showed an association results indicated that the validity of self-reported exposures was good between 2,4-D and NHL (2, 3). Of interest is that dioxin contamina and would also be applicable to the present study (14). tion, occasionally including TCDD, has been found also in 2,4-D (21). The influence of potential recall or observational bias in these For chiorophenols, the risk increased with increasing exposure in studies has been previously evaluated and discussed (15). It was number of days, whereas no dose-response was seen regarding expo concluded that the results could not be explained by such bias. sure to phenoyxacetic acids. In a review of the risk for NHL among farmers in 21 descriptive The literature regarding the association between herbicides and studies, the risk ratios ranged from 0.6 to 2.6 (16). The risk was cancer has been discussed elsewhere (4, 22). An association between 2388 PESTICIDEEXPOSUREAND NHL DDT exposure and NHL has been suggested (6). A significantly increased risk was found in our study in the univariate analysis, but the risk did not remain significant in the logistic regression. Exposure to asbestos yielded a nonsignificantly increased OR, which decreased in the multivariate analysis. Asbestos has been discussed as a risk factor for NHL of the gastrointestinal tract and oral cavity (23). This was not verified in our study (24). Our finding of an increased risk for NHL following exposure to organic solvents is in agreement with other reports (5, 6, 25, 26). In contrast to a previous publication (5), the risk was not associated with any particular anatomical localization of NHL. In this study, we showed that the increased risk associated with exposure to phenoxyacetic acids, chlorophenols, or organic solvents applied to all histopathological types of NFIL. Exposure to 2,4-D was associated with an increased risk for all major types of NHL and for follicular large cell NHL in particular in one study (27). The follicular large cell NHL corresponds to the nodular histiocytic type in the Rappaport classification and yielded the highest OR for phenoxy herbicide exposure in this analysis. This fmding was based on low numbers, however. Use of tobacco products has been suggested as a risk factor for NHL (28). This was not confirmed in our study. Median survival was somewhat shorter in unexposed cases than in those exposed. Exposure to phenoxy herbicides, chiorophenols, or organic solvents was more often reported in cases alive at the time of interviews as compared to deceased cases, which explains the finding. Cases exposed to organic solvents seemed to have a better prognosis than others, even if only cases alive at the time of assessment of exposure were considered. This seems to be explained by the fact that exposure to solvents was most common in the subtype of well differentiated lymphocytic NHL, which belongs to the group of NHL of low malignancy with better prognosis than other types of NHL (29). REFERENCES 1. Hardell, L., Eriksson, M., Lenner, P., and Lundgren, E. Malignant lymphoma and exposureto chemicalse, speciallyorganicsolventsc, hlorophenolasndphenoxyacids: a case-control study. Br. J. Cancer, 43: 169"176,1981. 2. Hoar, S. K., Blalr, A., Holmes, F. F., Boysen, C. D., Robel, R. J., Hoover, R., and FraumeniJ, . F., Jr. Agriculturahl erbicideuseandriskof lymphomaandsoft-tissue sarcoma. JAMA1256: 1141"11471,986. 3. Hoar Zahm, S., Weisenburger, D. D., Babbitt, P. A., Saal, R. C., Vaught, J. B., Cantor, K. P., and Blair, A. A case-control study of non-Hodgkin's lymphoma and the herbicide 2,4-dichlorophenoxyacetic acid (2,4-D) in Eastern Nebraska. Epidemiol ogy, 1: 349"356,1990. 4. Hoar Zahm, S., and Blalr, A. Pesticides and non-Hodgkin lymphoma. Cancer Re search, 52 (Suppl.): 5485s"5488s, 1992. 5. Olsson, H., and Brandt, L Supradiaphragmatic presentation of non-Hodgkin's lym phoma in men occupationally exposed to organic solvents. Ada Med. Scand., 210: 415"4181,981. 6. Persson, B., Dahlander, A. M., Fredriksson, M., Noorlind Brage, H., Ohison, C-G., andAxelson,0. MalignanltymphomaandoccupationaelxposuresB. r.J. lnd.Med., 46: 516"5201,989. 7. Blair, A., Linos, A., Stewart, P. A., Burmeister, L F., Gibson, R., Everett, G., Schuman, L., and Cantor, K. P. Comments on occupational and environmental factors in the origin of non-Hodgkin's lymphoma. Cancer Res., 52 (Suppl.): 55O1s"55O2s, 1992. 8. Wildund, K@,Lindefors B-M., and Holm, L-E. Risk of malignant lymphoma in Swedish agricultural and forestry workers. Br. J. Ind. Med., 45: 19"241,988. 9. Weisenburger, D. D. Pathological classfication of non-Hodgkin lymphoma for epi demiological studies. Cancer Res., 52 (Suppl.): 5456s"5464s, 1992. 10. Nordisk Yrkesklassficiering. Arbetsmarknadsstyrelsen, Solna, Sweden 1983. 11. Zheng, T., Taylor Mayne, S., Boyle, P., Holford, T. R., Liang Liu, W., and Flannery, J. Epidemiology of non-Hodgkin lymphoma in Connecticut 1935"88. Cancer (Phila.), 70: 840"8491, 992. 12. Mantel, N., and Haenszel, W. Statistical aspects of analysis of data from retrospective studies of disease. J. Nail. Cancer Inst., 23: 719"748, 1959. 13. Breslow, N. E., and Day, N. E. In: Statistical Methods in Cancer Research, Vol. 1, The Analysis of Case-Control Studies. IARC Scientific Publications No. 32, pp. 133"134In. ternational Agency for Research on Cancer, Lyon, 1980. 14. Hardell, L, and Sandstrom, A. Case-control study: soft-tissue sarcomas and exposure to phenoxyacetic acids or chiorophenols. Br. J. Cancer, 39: 71 1"7171, 979. 15. Hardell, L. Relation of soft-tissue sarcoma, malignant lymphoma and colon cancer to phenoxy acids, chlorophenols and other agents. Scand. J. Work Environ. Health, 7: 119"13109,81. 16. Blair, A., and Zahm, S. H. Cancer among farmers. Occup. Med. State of the Art Rev., 6: 335"3541,991. 17. Eriksson, M., Hardell, L., Malker, H., and Weiner, J. Malignant lymphoproliferative diseases in occupations with potential exposure to phenoxyacetic acids or dioxins: a register-based study. Am. J. md. Med., 22: 305"312,1992. 18. Kripke, M. L, Effects of UV radiation on tumor immunity. J. NatI. Cancer Inst., 82: 1392"1396,1990. 19. Kinlen, L Immunosuppressive therapy and acquired immunological disorders. Can cer Res., 52 (Suppl.): 5474s"5476s, 1992. 20. Rappe, C., Buser, H. R., and Bosshardt, H. P. Identification and quantification of polychlorinatedibenzo-p-dioxin(sPCDDs)and dibenzofuran(sPCDFs)in 2,4,5-T ester formulations and herbicide Orange. Chemosphere, 7: 431"438,1978. 21. Hagenmeier, H. Determination of 2,3,7,8-tetrachlorodibcnzo-p-dioxin in commercial chlorophenols and related products. Fresenius Z. Anal. Chem., 325: 603"606,1986. 22. Morrison, H., Wilkins, K., Semenciw, R., Mao, Y., and Wigle, D. Herbicides and cancer. J. NatI. Cancer Inst., 84: 1866"1874, 1992. 23. Ross, R., Dworsky, R., Nichols, T., Paganini-Hill, A., Wright, W., Koss, M., Lukes, R., and Henderson, B. Asbestos exposure and lymphomas of the gastrointestinal tract and oral cavity. Lancet, 2: 1118"1120,1982. 24. Bengtsson, N. 0., Hardell, L., and Eriksson, M. Asbestos exposure and malignant lymphoma. Lancet, 2: 1463, 1982. 25. Vianna, N. J., and Polan, A. Lymphomas and occupational benzene exposure. Lancet, 2: 1394"1395,1979. 26. Olsson, H., and Brandt, L Risk of non-Hodgkin's lymphoma among men occupa tionally exposed to organic solvents. Scand. J. Work Environ. Health, 14: 246"251, 1988. 27. Weisenberg, D., Zahm, S., Ward, M., et aL Non-Hodgkin's lymphoma associated with the agricultural use of herbicides: analysis by histologic type. Third Meeting of the EuropeanAssociationof HematophatologyW, urzburgO, ctober7"119, 90. 28. Morris Brown, L, Evereu, G. D., Gibson, R., Burmeister, L F., Schuman, L M., and Blair, A. Smoking and risk of non-Hodgkin's lymphoma and multiple myeloma. Cancer Causes Control, 3: 49"551, 992. 29. Brittinger, G., Bartels, H., Common, H., et al. Clinical and prognostic relevance of the Kid classification of non-Hodgkin lymphomas. Results of a prospective multicenter study by the Kid lymphoma study group. Hematol. Oncol., 2: 269"306, 1984. 2389