2 Seniori Costantini et al KEY WORDS:leukemia;multiple myeloma;benzene;solvents;case-control study INTRODUCTION aYaomepaac The association between leukemiaand benzene hasbeen 2006 exp to this oductiv and nsive Cancer classified benzene as carcinogenic to humans on the less information on the risk from exposure to other organic basis of case series reported in workers exposed to high levels solvents.An association between solvent exposure and of benzen hoe y as nd the [Aksoy et al197:Infante et al 1977:Rinsky eta19 reported in two case-control studies on AML and exposure 1987.ween 196y 2ofoganicoemsAbnuaL,2z0d,a Paxton etal 1994a.b:Seniori Costantini et al 20031 through air pollution:for smokers,the overwhelming source Hor of benzene 1 nstream cigarett ns o y moke and ong. a u暖ML et al 19:Hayes et al996.19972000]showed an a iation studies Kasim et al..2005:Pogoda and Preston-Martin. tries and AML 2006 tho in the Eu e present artic we present result elevated risks of AML for Study on Hematolymphopoietic Malignancies and Exposure 20031 to Solvents and Pesticides MATERIALS AND METHODS leukemia (CLLand multiple myeloma (MM)are still under debate.Few studieshave reportedarisk of CLLwith benzene Case and Control Recruitment ntrol studie We 11 malignancies.incident in males and females aged 20 in which an excess of leukemi years in the period were identified.A totalo 1996.1995 2003 ca In other studies of chemical and petroleum industries inUS subects randoml selected through the demographic fileso and United Kingdom.no incr ased risks of MM were the municipalities in each of the areas under study.stratified et al. 986 by sex and s-year age groups.Details on case also did sly Isam of an increased risk of MM Morris etal..1986:Flodin et al. Here we present results concerning 586 cases of 1987:Linet et al.,1987:Eriksson and Karlosson,1992; leukemia collected in 7 out off the 11 areas (Torino,Verona 1992 of .1992na which samexylene, and toluene and chronic lymphatic leukemia and between benzene and MM with longer latencies than have been observed for AML in other studies. Am. J. Ind. Med. 2008.
2008 Wiley-Liss, Inc. KEY WORDS: leukemia; multiple myeloma; benzene; solvents; case–control study INTRODUCTION The association between leukemia and benzene has been described among workers exposed to benzene in several productive sectors. The International Agency of Research on Cancer classified benzene as carcinogenic to humans on the basis of case series reported in workers exposed to high levels of benzene in shoe manufacturing and printing in Italy, France, and Turkey as well as some epidemiologic studies of workers in shoe production and the rubber industry [Aksoy et al., 1974; Infante et al., 1977; Rinsky et al., 1981; IARC, 1982]. Subsequently, new data from cohort studies have confirmed a strong association between leukemia and benzene exposure [Rinsky et al., 1987; Crump, 1994, 1996; Paxton et al., 1994a,b; Seniori Costantini et al., 2003]. The risk was particularly elevated for acute myeloid leukemia (AML)—for which a strong dose/response rela￾tionship was evidenced [Wong, 1995; Wong and Raabe, 1996]. Findings from studies conducted in China [Yin et al., 1996; Hayes et al., 1996, 1997, 2000] showed an association between benzene exposure in different industries and AML at levels lower than those described in the European or US studies. Recently two nested case–control studies in gas and electric utility workers and petroleum workers also found elevated risks of AML for people exposed to low levels of benzene [Guenel et al., 2002; Glass et al., 2003]. While there is a general consensus about the ability of benzene to induce AML, its effects on chronic lymphoid leukemia (CLL) and multiple myeloma (MM) are still under debate. Few studies have reported a risk of CLL with benzene exposure. In the two cited nested case–control studies, the risk of CLL was raised, but only for the highest exposure category [Guenel et al., 2002; Glass et al., 2003]. An in￾creased risk of MM was reported in the same rubber industry in which an excess of leukemia was seen [Rinsky et al., 1987; Wong, 1995], as well as in the Chinese cohort [Yin et al., 1996] and in the US chemical industry [Collins et al., 2003]. In other studies of chemical and petroleum industries in US, and United Kingdom, no increased risks of MM were observed [Bond et al., 1986; Wong, 1987a,b; Wong and Raabe, 1995, 1997, 2000]. Case–control studies on MM conducted in US and Europe also did not uncover evidence of an increased risk of MM [Morris et al., 1986; Flodin et al., 1987; Linet et al., 1987; Eriksson and Karlosson, 1992; Heinemann et al., 1992; Pottern et al., 1992]. In a recent meta-analysis of seven benzene cohort studies in which heavy exposures to benzene occurred, a significant excess of MM in relation to benzene exposure was reported [Infante, 2006]. In contrast to this rather extensive literature on hematolymphopoietic cancers and benzene, there is much less information on the risk from exposure to other organic solvents. An association between solvent exposure and mortality from hematolymphopoietic malignancies, and in particular lymphatic leukemia, was been reported for rubber workers [McMichael et al., 1975]. Increased risks were reported in two case–control studies on AML and exposure to mixtures of organic solvents [Albin et al., 2000; Lazarov et al., 2000]. Exposure to benzene is not confined to work environ￾ments. The major route of non occupational exposure is through air pollution; for smokers, the overwhelming source of benzene is mainstream cigarette smoke and suggestions of an increased risk of leukemia, and in particular AML have been reported for smokers in some recent case–control studies [Kasim et al., 2005; Pogoda and Preston-Martin, 2006]. In the present article, we present results concerning the association between solvent exposure and occurrence of leukemia and MM from the Italian Multicenter Case–control Study on Hematolymphopoietic Malignancies and Exposure to Solvents and Pesticides. MATERIALS AND METHODS Case and Control Recruitment We conducted a population-based case–control study in 11 areas in Italy, in which all cases of hematolymphopoietic malignancies, incident in males and females aged 20– 74 years in the period 1991–1993 were identified. A total of 2,737 cases of hematolymphopoietic malignancies were interviewed. The control group was formed from 1,779 subjects randomly selected through the demographic files of the municipalities in each of the areas under study, stratified by sex and 5-year age groups. Details on case ascertainment and recruitment and control sampling procedures have been described previously [Seniori Costantini et al., 2001]. Here we present results concerning 586 cases of leukemia collected in 7 out off the 11 areas (Torino, Verona, Imperia, Forlı`, Firenze, Ragusa, Latina), and 1,278 controls sampled in the same areas. We also report results for 2 Seniori Costantini et al