Benzene is considered a known human carcinogen by the NTP and a group one carcinogen by the International Agency for Research on Cancer (IARC, 1982; NTP, 1986). Exposure to benzene is associated with leukemia in humans (USEPA, 1997). In rodents, benzene-exposure produced lymphoma and increases in the number of tumors in other organs. In general, the rat is less responsive than the mouse to the induction of hematopoietic neoplasia.

In this study, the potential for haploinsufficient p16^Ink4a/p19^Arf mice to develop cancer in response to benzene was examined. Benzene was toxic to the hematologic system, particularly to males, resulting in treatment-related body weight effects and one early death in a high-dose (200 mg/kg) male. Mean body weights of all dosed groups of male mice were decreased compared to vehicle controls, while mean body weights of female mice were decreased only in the 200 mg/kg group. Mean erythrocyte, leukocyte, and lymphocyte counts were depressed at weeks 13 and 27 in dosed males and females. A treatment-related decrease in thymus weights was observed in male mice.

The incidence of malignant lymphoma was significantly increased in 200 mg/kg males compared to the vehicle controls and exceeded the incidence in historical controls (Table E1). All but one of the lymphomas observed in males were found at the scheduled sacrifice at 27 weeks. Treatment-related bone marrow atrophy, thymus atrophy, atrophy of the mandibular lymph node and other lymph nodes, and hemosiderin pigmentation in the bone marrow occurred in male mice. In addition, treatment-related pigmentation of the skin of the paw and hematopoietic cell proliferation in the spleen were observed in male and female mice.

Malignant lymphomas were not seen in female haploinsufficient p16^Ink4a/p19^Arf mice; other studies have also shown that male mice are more susceptible to benzene-induced cancer than female mice after inhalation exposure. This phenomenon suggests that benzene may be more clastogenic in males than in females. This hypothesis is supported by the micronucleus findings (Figure 5). Scoring for normochromatic erythrocytes (NCEs) reflects the induction of micronucleated erythrocytes during the preceding 5 to 6 weeks of dosing (MacGregor, 1990; Witt, 2000). The lowest concentration at which a significant increase in micronucleated NCEs occurred was 25 mg/kg in male and female mice. However, the number of micronuclei formed in males at the carcinogenic dose of 200 mg/kg at 27 weeks was approximately four times that in female mice. The ability of a chemical to induce micronuclei is generally considered to indicate a potential risk for cancer in humans (Shelby, 1988). Although benzene is a well-documented inducer of chromosomal damage in numerous in vivo models, it is not active in bacterial gene mutation assays (Zeiger and Haworth, 1985), likely due to the complex pathways required for biotransformation of benzene to its genetically active metabolites.

Figure 5

Effects of Benzene on the Micronuclei of Haploinsufficient p16^Ink4a/p19^ArfMice. (MN-NCE=micronucleated normochromate erythrocytes)

Hematopoietic stem cells are thought to be the target for benzene-induced DNA damage and are a small population of the cells in the bone marrow (approximately <0.05%; Morrison and Weissman, 1994); there is some evidence in a Swiss Webster mouse cell culture study that cells from males are more susceptible to benzene than cells from females (Corti and Snyder, 1998). Using cultured hematopoietic stem cells from male and female 129/SvJ mice, it has been shown that there are gender-specific gene expression patterns after benzene exposures (Faiola et al., 2004). Male mice (129/Sv background) deficient in epoxide hydrolase are not as susceptible to benzene induced toxicity as the wild type strain (129/Sv) (Bauer et al., 2003).

Studies have been done to compare gene expression patterns in bone marrow cells derived from p53^+/+ C57Bl/6 mice versus p53^+/− C57Bl/6 mice (Boley et al., 2002). These studies showed that there were significantly higher levels of p21, gadd45, and cyclin G, genes involved in restraining cancer cell growth and repairing DNA damage, after benzene exposure in the p53^+/+ mice than in the p53^+/− mice.

In the benzene study in the p53^+/− mouse (Recio et al., 2006), as in the current study, the thymus was a target organ. The p53^+/− male mice exposed to benzene by inhalation at 100 ppm 6 hours per day, 5 days a week for up to 52 weeks began to develop thymic lymphomas after 26 weeks of exposure. At the end of the 52-week exposure period, 71% of exposed p53^+/− male mice developed thymic lymphomas versus an incidence of less than 10% for these tumors in control p53^+/− male mice. When p53^+/− male mice were exposed to benzene by corn oil gavage at 0, 100, or 200 mg/kg for 26 weeks, treatment-related sarcomas and thymic lymphomas occurred (French et al., 2001).

Pigmentation consistent with melanin occurred in males and females at 50, 100, and 200 mg/kg. It is not certain how benzene causes this effect. However, in humans, the p16 (INK4) gene deficiency is often found in melanoma tissue (Sotillo et al., 2001; Soto et al., 2005). Mice with melanocyte-specific expression of H-ras on an INK4a deficient background develop spontaneous cutaneous melanomas (Chin et al., 1997), and INK4a/Arf deficiency promotes ultraviolet-radiation induced melanomagenesis (Recio et al., 2002). Thus, in these benzene studies, the deficiency in the p16 gene might account for the pigmentation observed in melanocytes. However, p16 gene deficiency does not appear to be sufficient for induction of melanomas in mice, but may require deficiency in other genes (e.g., ras or p53) (Chin et al., 1997; Bardeesy et al., 2001; Recio et al., 2002; Ackerman et al., 2005).

The haploinsufficient p16^Ink4a/p19^Arf mouse remains relatively free from neoplasms until 6 to 8 months of age. In the 6-month GMM studies of benzene and phenolphthalein, only one of 30 (3%) vehicle control haploinsufficient p16^Ink4a/p19^Arf male mice developed malignant lymphoma, and no neoplasms were found in any other male or female mice (approximate age of mice was 33 weeks) (NTP, 2008b). In 46-week-old mice in the aspartame and glycidol 9-month GMM studies, alveolar/bronchiolar neoplasms, histiocytic sarcoma, and malignant lymphoma occurred in 10% (3/30), 13% (4/30), and 7% (2/30) of vehicle control male mice, respectively, and in 3% (1/30), 47% (14/30), and 0% (0/30) of vehicle control female mice, respectively (NTP, 2005, 2008a).

The current study showed that the haploinsufficient p16^Ink4a/p19^Arf mouse had a benzene carcinogenic response after 27 weeks of benzene exposure and that the carcinogenic response was detected with fewer numbers of animals than in the traditional 2-year study in B6C3F₁ mice.

The current haploinsufficient p16^Ink4a/p19^Arf mouse study detected fewer sites for a benzene carcinogenic response than found in the 2-year B6C3F₁ mouse study or in the benzene p53^+/− mouse study (Table 1). In addition, the carcinogenic response in the haploinsufficient p16^Ink4a/p19^Arf mouse study was only seen in the 200 mg/kg males, while a carcinogenic response occurred in the 2-year B6C3F₁ mouse study at 25, 50, and 100 mg/kg and in the 26-week p53^+/− male mouse gavage study at 100 and 200 mg/kg. The total benzene dose administered in the 27-week haploinsufficient p16^Ink4a/p19^Arf mouse study at the carcinogenic dose of 200 mg/kg was 27,000 mg benzene/kg body weight, and the total benzene dose administered to mice in the 2-year study at the carcinogenic dose of 25 mg/kg was 12,875 mg benzene/kg body weight.

Footnotes

*

Explanation of Levels of Evidence of Carcinogenic Activity is on page 8. A summary of the Technical Reports Review Subcommittee comments and the public discussion on this Report appears on page 10.