Скачать 180.93 Kb.
Quantitative Human Health Risk Assessment for 1,3-Butadiene Based Upon Ovarian Effects in Rodents
Kirman CR1, Grant RL2
1Summit Toxicology LLP, Orange Village, OH; 2Texas Commission of Environmental Quality, Austin, TX
1,3-Butadiene (BD) is an important industrial chemical used in the production of styrene-butadiene rubber (SBR), polymers, and other chemicals. In addition to being carcinogenic to laboratory rodents (NTP, 1993; Owen et al., 1987) and associated with increased leukemia mortality in exposed workers (Cheng et al., 2007; Delzell et al., 2006; Graff et al., 2005; Macaluso et al., 2004), exposures to BD are associated with a number of noncancer endpoints in rodents including reproductive toxicity. The toxic effects of BD on mouse ovary have been identified as a particularly sensitive endpoint (chronic lowest-observed-adverse-effect-level or LOAEL = 6.25 ppm) (NTP, 1993). In sharp contrast, female rats do not exhibit ovarian atrophy at comparatively high concentrations (chronic no-observed-adverse-effect-level or NOAEL = 8,000 ppm) (Owen et al., 1987). The mode of action for ovarian toxicity by diepoxides has been well studied. Species differences in the metabolic activation of BD to a reactive diepoxide (diepoxybutane or DEB), which is believed to be the causative agent for ovarian effects, likely underlie the apparent species difference in sensitivity. Currently, the mouse ovarian toxicity data serves as the basis for noncancer risk assessment for BD in the U.S. (USEPA, 2002). Rat data for ovarian are not included in the quantitative dose-response assessment for BD, despite the fact that humans are more similar to rats than mice with respect to the metabolic activation of BD (Swenberg et al., 2007).
Recently the National Academy of Science (NAS, 2009) published a number of recommendations for unifying cancer and noncancer dose-response assessment. Recommendations included:
The text below provides a case study using the noncancer dose-response for BD ovarian toxicity in rodents. The goals of this work are:
The text below is organized into separate sections for mode of action (Section 2), dose-response assessment (Section 3), and discussion (Section 4). A flowchart depicting the noncancer assessment conducted for BD is provided in Figure 1.
Figure 1. Flowchart for BD Noncancer Assessment
2. Mode of Action Assessment
Metabolism is a key determinant of BD-induced ovarian toxicity. The metabolism of BD to reactive intermediates has been well studied, and has been reviewed (Kirman et al., 2010; Albertini et al., 2003; Himmelstein et al., 1997). The metabolism of BD is depicted in Figure 2 and summarized below, with emphasis on the metabolic steps for activation and detoxification with respect to ovarian toxicity.
Figure 2. BD Metabolism
There is strong evidence that ovarian atrophy is mediated by the formation of diepoxides, such as the BD metabolite DEB (Doerr et al., 1995; 1996). The ability to form diepoxides is necessary for the production of ovarian toxicity in rodents, as indicated by the presence of ovarian toxicity following exposure to diepoxides (DEB, vinylcylcohexane diepoxide) and diepoxide precursors (EB, BD dimer or vinylcyclohexene, vinylcyclohexene epoxide, isoprene), and the absence of ovarian toxicity in structural analogues that do not form diepoxides (ethylcyclohexene oxide, vinylcyclohexane oxide, cyclohexene oxide) (Doerr et al. 1995, 1996). There are marked species differences in effects observed between rats, which do not exhibit BD-induced ovarian atrophy following chronic exposures as high as 8,000 ppm (Owen et al., 1987), and mice, which exhibit BD-induced ovarian atrophy following chronic exposures as low as 6.25 ppm BD (NTP, 1993). Furthermore, EB has been shown to be toxic to mouse ovary but not to rat ovary, reflecting greater conversion of EB to DEB in mice, while DEB was reported to be toxic to the ovary of both species, albeit with a lower efficacy in rats than in mice (Doerr et al., 1996).
Species differences in ovarian effects (mouse > rat) correlate well with species differences in the formation of DEB (mouse > rat). In vitro studies on the Vmax and Km values for activation and detoxification pathways indicate that mice have a significantly higher ratio of EB activation to detoxification than either rats or humans (Csanady et al., 1992; Schmidt and Loeser, 1985; Krause and Elfarra, 1997; Bond et al., 1993; Kreuzer et al., 1991Seaton et al., 1995). In the effluent of mouse livers perfused with BD, all three epoxides (EB, DEB and EBD) and B-diol were observed while in effluents from rat livers perfused with BD only EB and B-diol were found (DEB was not detected, while EBD was not quantifiable because of an interfering peak). When the livers were perfused with EB, B-diol, EBD, and DEB were formed, with B-diol predominating in both species (Filser et al., 2001, 2010). DEB formation was greater in mouse than in rat livers (Filser et al., 2010). Following in vivo exposures of rats and mice to BD, differences in circulating DEB levels haven been reported to be over 100-fold greater in mice (Filser et al., 2007; Thornton-Manning et al., 1995a,b).
Quantitative differences in the in vivo production of BD metabolites are also reflected in their in vivo accumulations as hemoglobin adducts. Recently, a DEB-specific hemoglobin adduct, N,N-(2,3-dihydroxy-1,4-butadiyl)-valine (pyrVal), has been identified and measured, providing new insights into species and exposure differences in BD metabolism (Boysen et al., 2004). The formation of N,N-(2,3-dihydroxy-1,4-butadiyl) valine (pyr-Val) hemoglobin adducts has been studied in male and female mice and rats exposed to 1.0 ppm by inhalation for 6 hours/day for four weeks (Swenberg et al., 2007), in which adduct burdens in rats were more than 30-fold lower than the corresponding values in mice. More recently, the formation of DEB adducts in rats and mice of both sexes was assessed following 4-week exposures to either 1, 6.25, or 62.5 ppm BD for 6 hours/day (Georgieva et al., 2010). The difference between species was dose-dependent, with a larger difference observed at higher concentration compared to low concentrations (Figure 3). A less pronounced difference between species was also reported by these authors following 2-week exposures to BD, primarily because in the mouse the 2-week adduct burdens were appreciably lower than observed at 4 weeks, suggesting that steady-state had not been reached. Humans appear to form even less of the diepoxide than rats. Swenberg et al. (2007) compared results in occupationally-exposed workers in the Czech Republic to results in mice and rats for pyr-Val (Figure 3). Pyr-Val adducts were not detected in occupationally exposed human men and women. Female mice appear to produce more than 100-fold more pyr-Val adducts than do human females, which indicates more DEB is formed in female mice than female humans. Pyr-Val adducts for human females were not detected (LOD = 0.3 pmol/g Hb).
Figure 3. Pyr-Val Adduct Efficiencies in Rats and Mice as a Function of BD Exposure Concentrations (ppm 6 hours/day, 5 days/week for 4 weeks; Georgieva et al., 2010) Compared To Human Data (Swenberg et al., 2007)
LOD = 0.3 pmol/g
At the 2007 and 2008 Society of Toxicology meetings, Georgieva et al. (2007; 2008) presented results measuring pyr-Val adducts in the Czech Republic workers. Pyr-Val adducts were detected at low concentrations in Czech Republic workers. There was not a clear dose-response relationship between pyr-Val adducts and BD concentrations from the Georgieva et al. (2007) study using a LS-MS/MS analytical method with a limit of detection (LOD) and limit of quantitation (LOQ) of 5 and 10 fmol/g Hb, respectively, and it was hypothesized that the pyr-Val adducts may have been formed from other unknown sources. However, the Georgieva et al. study (2008) using a more sensitive nano-LC-MS/MS method with a LOD and LOQ of 1 and 4 fmol, respectively, showed the amount of pyr-Val was significantly higher in the higher-exposed polymerization workers than in the monomer workers and controls.
Based upon a review of the available information, there is sufficient information to establish an MOA for the ovarian effects of BD in rodents. The detection of low levels pyr-Val in humans, and reports of ovotoxicity in nonhuman primates exposed to a structurally similar diepoxide (vinylcyclohexene diepoxide; Appt et al., 2006), suggests that qualitatively the endpoint of rodent ovarian toxicity is relevant to human health. However, there are clear quantitative differences across species with respect to dosimetry that need to be addressed in using these data for human health risk assessment. A summary of the key steps in the proposed MOA for ovarian toxicity, along with potential implications to human health risk assessment is provided in Table 1.
Table 1. Key Events in the MOA for BD in Rodent Ovarian Atrophy
|An evaluation of risk to u. S. Consumers from methylmercury in commercial fish products, including a quantitative assessment of risk and beneficial health effects from fish||Does ratification of human-rights treaties have effects on population health?|
|Assessment Strategy for Evaluating the Environmental and Health Effects of Sanitary Sewer Overflows from Separate Sewer Systems||Aged Care Assessment Service (acas) and Office of the Public Advocate (opa) protocol. Incorporating health professionals undertaking assessment and placement within health networks|
|Dir 111 Risk Assessment and Risk Management Plan (February 2012) Office of the Gene Technology Regulator||Risk Assessment and Risk Management Plan Application for licence for dealings involving an intentional release into the environment dir 005/2001|
|Risk Assessment and Risk Management Plan||Human pth (1-34) Specific elisa kit Enzyme-Linked ImmunoSorbent Assay (elisa) for the Quantitative Determination of Human Parathyroid Hormone 1-34 Levels in Plasma|
|Risk assessment 4||Risk Assessment|