Leukemia,Multiple Myeloma,and Solvent Exposure 7 data and previous and chemo-ther y),that are OR increased to 3.2 (95%CI:1.1-9.3).based on 7 cases totheAML risk and could berason ofreduced probability We have previously reported strong associations betweer of being part of and then of Is,we dic very low gave condlicting nested in a cohort of Australian petroleum workers showed supporting the association between occupational exposure t an excess the highest ben exposure d ns f CL -at no e nd Smith.20071. exposures less than 8 ppm-years [Glass et al.,2003]. An increased risk of MM was also suggested from these Evidence of increasing risk with with 10 to 19-yea e [Collins eta 2003].One cohort study on workers exposed to benzene logical evider of the association between benzene revealed a week for M w discu ot al etedin the agreemen and Canadian petroleum refinery and dist butic where exnosure declines in time while cil and mM risks are no benzene con ations were generally low failed to videpceorprodd evid Rinsky et al 200 Sch aniuk 1997-We t al 19901 The and tol conceming more than 300.in and risk f CLL and ber benzene and risk of MM.Benzene is toxic for blood marrow did not reveal any increased risk of leukemia or MM [Wong where A and MM originate from myelocyte: the lac evels of b y We found positive asocitions bet CLL and ex did not observe clear evidence of an association betwee posures to benzene,toluene,and xylene,albeit with wide benzene and AMl leukemia however the absence of confidence inter When those me sures and the small number of observa- sure to on ere subdi the interpretation duration.For medium/high intensity osure for more thar 15 vears.the OR were:for benzene 4.7(9%C:0.8 ACKNOWLEDGMENTS This work has be ed out with the n of s We c t think that a surveillar e bia Barcellini,G.Bari,LBellesini,V.Cacciarini.R.Carlin to the CLL findings,as suggested by Goldstein with regard to M.Casale,G.Castellino.G.Cremaschi,L.Davico,A.Fiorio the Gilass's cohor nested cas ontrol study Goldstein R.Gibilisco,LGuzzo,R.Hirvas,S.Legrotti.LM .U.Pa 20 C Pi place in our Tiberti,G.Tonini,P.Trada,T.Vescio,and M.Zanetta.We A recent review of the literature on benzene exposure thank the clinical department staffs involved in patient recruitmen were seen OE L se-contr REFERENCES [Schnatter et al9,2005).In our study a good assessment nshoe-workers exposed of exposure was done,even if based on experts'judgments and not onen ental me sures,so ou cem to b 000A exposure, may explain the lack of association in these data. When adjusting by possible confounders (Down syndrome and previous X-ray- and chemo-therapy), that are associated to the AML risk and could be reason of a reduced probability of being part of the workforce and then of being exposed to chemicals, we did not observe any difference in ORs. Results of studies concerning AML in industries where levels were very low gave conflicting results. A case–control study nested in a cohort of Australian petroleum workers showed an excess of leukemia for the highest benzene exposure category (>8 ppm-years), but lower than reported in previous studies. For CLL the risk was elevated also for cumulative exposures less than 8 ppm-years [Glass et al., 2003]. Evidence of increasing risk with increasing cumulative exposure was found for all leukemias and specifically for AML in the Monsanto cohort only for benzene peak exposures over 100 ppm for 40 days or more [Collins et al., 2003]. One cohort study on workers exposed to benzene revealed a week trend of increasing SMRs for leukemia with increasing low-level cumulative exposure [Bloemen et al., 2004]. Others studies conducted in the UK, US, and Canadian petroleum refinery and distributions where benzene concentrations were generally low failed to evidence or provided weak evidence of an association [Wong and Raabe, 1995; Schnatter et al., 1996; Rushton and Romaniuk, 1997; Wong et al., 1999]. The pooled analysis concerning more than 300,000 subjects in petroleum workers did not reveal any increased risk of leukemia or MM [Wong and Raabe, 2000]. The authors attributed the lack of an increased risk of AML to the low levels of benzene exposure. We found positive associations between CLL and ex￾posures to benzene, toluene, and xylene, albeit with wide confidence intervals. When those with medium/high expo￾sure to one of these agents were subdivided by duration of exposure there was a trend of increasing risk with increasing duration. For medium/high intensity exposure for more than 15 years, the ORs were: for benzene 4.7 (95% CI: 0.8–26.5); for xylene 3.3 (95% CI: 0.7–15.2); and for toluene 4.4 (95% CI: 1.1–18.8). We do not think that a surveillance bias could contribute to the CLL findings, as suggested by Goldstein with regard to the Glass’s cohort nested case–control study [Goldstein, 2004], as presumably a very small part of the CLL incidence could be explained by blood count examinations in work￾place in our data-set. A recent review of the literature on benzene exposure and leukemia sub-types suggested that elevated risks for CLL were seen in nested case–control studies that have advantages in terms of benzene exposure assessment [Schnatter et al., 1996, 2005]. In our study a good assessment of exposure was done, even if based on experts’ judgments and not on environmental measures, so our results seem to be strong enough and support this association. It is noteworthy that we also found evidence for the association with two other aromatic solvents, toluene and xylene. Interestingly, for subjects exposed to all three aromatic solvents, the OR increased to 3.2 (95% CI: 1.1–9.3), based on 7 cases. We have previously reported strong associations between exposure to benzene, xylene, and toluene and the occurrence of small cell non-Hodgkin’s lymphomas that include CLL as a subtype [Miligi et al., 2006]. In a recent review on benzene exposure and risk of NHL, epidemiological data strongly supporting the association between occupational exposure to benzene and risk of CLL—at now classified together with small lymphocytic lymphoma as a form of NHL—have been reported and discussed [Hartge and Smith, 2007]. An increased risk of MM was also suggested from these same solvents, and it was highest for people exposed for 15 or more years with 10- to 19-year latency period. It is note￾worthy that the association was higher for those exposed only to benzene. The biological plausibility and the epidemio￾logical evidence of the association between benzene exposure and MM was discussed by Goldstein [1990]. Our results, even if based on small numbers, are in agreement with the hypotheses that AML risk following benzene exposure declines in time while CLL and MM risks are not seen until a longer latency period has passed [Finkelstein, 2000; Rinsky et al., 2002]. Our results support the association between exposure to benzene, xylene, and toluene and risk of CLL and between benzene and risk of MM. Benzene is toxic for blood marrow where AML, CLL and MM originate from myelocytes, lymphocytes and plasma cells. Thus associations between benzene and CLL and MM seem biologically plausible. We did not observe clear evidence of an association between benzene and AML leukemia; however the absence of environmental measures and the small number of observa￾tions limited the interpretation of our results. ACKNOWLEDGMENTS This work has been carried out with the cooperation of S. Barcellini, G. Barni, L. Bellesini, V. Cacciarini, R. Carlini, M. Casale, G. Castellino, G. Cremaschi, L. Davico, A. Fiorio, R. Gibilisco, L. Guzzo, R. Hirvas, S. Legrotti, L. Migliaretti, R. Monteleone, G. Osella, T. Palma, G. Panizza, C. Picoco, G. Piergiovanni, G. Righetti, R. Sguanci, M. Tedeschi, D. Tiberti, G. Tonini, P. Trada, T. Vescio, and M. Zanetta. We thank the clinical department staffs involved in patient recruitment. REFERENCES Aksoy M, Erdem S, Dinol G. 1974. Leukaemia in shoe-workers exposed chronically to benzene. Blood 44(6):837–841. Albin M, Bjork J, Welinder H, Tinnerberg H, Mauritzson N, Johansson B, Billstrom R, Stromberg U, Mikoczy Z, Ahlgren T, Nilsson PG, Mitelman F, Hagmar L. 2000. Acute myeloid leukemia and clonal Leukemia, Multiple Myeloma, and Solvent Exposure 7