occupationally (Berlin et al., 1995: SIR=2.1, 95% CI=0.8–4.6), and leather or tannery workers (Costantini et al., 1989: SMR=1.64, 95% CI=0.53–3.82). Studies in which associations were not found (Anttila et al., 1998; Boice et al., 1999; Garabrant et al., 1988; Matanoski et al., 1986; Steenland and Palu, 1999; Walker et al., 1993; Wolf et al., 1981; Wong et al., 1993) generally had small numbers of exposed cases. Two studies (Anttila et al., 1995; Fu et al., 1996) provided evidence of a dose-response relationship with increasing relative risks as exposure increased.
IARC and the US Environmental Protection Agency (EPA) have determined that benzene is carcinogenic in humans on the basis of both animal and human studies (ATSDR, 1997a; IARC, 1987; NTP, 2001). That determination was based primarily on the findings on leukemia defined broadly. In addition, epidemiologic studies of occupations exposed to mixtures of solvents, including benzene have shown increased risks of developing cancer. Among those at risk are rubber workers, mechanics, and some groups of chemical workers, printers and paper-industry workers, and shoe and leather workers (IARC, 1987, 1989).
Based on its review of the literature on exposure to benzene, the committee found that the combination of consistently positive findings in the cohort of workers with known exposures to benzene and evidence of a dose-response relationship fulfilled the criteria for a conclusion of sufficient evidence of an association between exposure to benzene and adult leukemia. However, the committee decided that the evidence of an association between exposure to benzene and adult leukemia was not as strong as that for acute leukemia. Thus, it did not warrant a conclusion of causality. The findings, although mostly positive, are not as consistent and statistically precise as the findings on acute leukemia. Most likely, the positive studies on adult leukemia and exposure to benzene include cases of acute leukemia. However, they may also include cases of chronic leukemia, lymphatic leukemia, and hairy cell leukemia, for which the existence of associations is not as clear. On the basis of the studies reviewed, the committee believes that the evidence on exposure to benzene and adult leukemia, defined broadly, met the definition of sufficient evidence of an association but not sufficient evidence of a causal relationship.
The committee concludes, from its assessment of the epidemiologic literature, that there is sufficient evidence of an association between chronic exposure to benzene and adult leukemia.
For exposure to other solvents, such as trichloroethylene and toluene, the overall paucity of studies and the lack of consistently positive findings limits the evidence that the committee had to review.
The committee concludes, from its assessment of the epidemiologic literature, that there is inadequate/insufficient evidence to determine whether an association exists between chronic exposure to specific organic solvents under review, other than benzene, and adult leukemia.
In contrast, the findings for unspecified mixtures of organic solvents and adult leukemia showed increased relative risks, including two studies that provided evidence for a dose-response relationship with increasing levels of exposure. Table 6.40 identifies the key studies reviewed by the committee on adult leukemia. Unless indicated in the table, the study populations include both men and women.