OSHA comments from the January 19, 1989 Final Rule on Air Contaminants Project extracted from 54FR2332 et. seq. This rule was remanded by the U.S. Circuit Court of Appeals and the limits are not currently in force.
CAS: 593-60-2; Chemical Formula: C2H3Br
OSHA had no former PEL for vinyl bromide. Based on the ACGIH recommendation, OSHA proposed a 5-ppm TWA PEL; this limit is established in the final rule. NIOSH has no REL for vinyl bromide. The ACGIH places vinyl bromide on its A2 list of industrial substances suspected of having carcinogenic potential in humans. Vinyl bromide is a colorless gas with a characteristic odor and is used as an intermediate in organic synthesis and in the manufacture of polymers, copolymers, and flame retardants. Its principal use is as a flame retardant.
Henschler and Hoos (1982/Ex. 1-818) believe that vinyl bromide undergoes the same mechanism of biotransformation as its structural analog, vinyl chloride, a recognized human carcinogen that has been regulated by OSHA in a Section 6(b) rulemaking. The microsomal oxidation of vinyl bromide leads to epoxide formation, which results, in turn, in the formation of a reactive intermediate. This intermediate has the potential to form covalent bonds with DNA to produce a mutagenic response. Vinyl bromide has been reported to be mutagenic in Salmonella typhimurium and tradescantia (IARC 1979a/Ex. 1-1125; NIOSH/OSHA 1978/Ex. 1-1119).
No epidemiological studies have been conducted on populations exposed to vinyl bromide. Benya, Busey, Dorato, and Berteau (1982/Ex. 1-244) reported a positive carcinogenic response in an inhalation study of rats exposed to vinyl bromide vapor; this study is important because inhalation is a major mode of occupational exposure. The results of the Van Duuren (1977/Ex. 1-284) study were equivocal (described below), in that female Swiss albino mice were exposed dermally or by subcutaneous injection either to vinyl bromide in acetone or to polymerized vinyl bromide in an aqueous latex solution.
Benya et al. (1982/Ex. 1-244) exposed male and female Sprague-Dawley rats to 0, 9.7, 52, 247, or 1235 ppm vinyl bromide by inhalation for six hours daily, five days per week, for two years. The incidence of angiosarcomas, primarily of the liver, was found to be statistically significant in all dose groups tested except controls. It should be noted that a closely related chemical analog, vinyl chloride, also causes liver angiosarcoma in humans and animals. The combined incidences of hepatic angiosarcomas in the treated male and female rats were 1/288, 17/240, 86/240, 122/240, and 84/240 for their respective dose levels. One female rat in the control group developed an hepatic angiosarcoma. Table C15-9 summarizes the incidence of angiosarcoma in control and treated rats.
Van Duuren (1977/Ex. 1-284) injected a group of female ICR/Ha Swiss mice once weekly for 48 weeks with 0.05 ml of commercial polymerized vinyl bromide aqueous latex suspension; the animals were observed for 420 days. Nineteen of the 30 mice developed sarcomas at the site of injection. Animals in a positive control group that had been injected with b-propriolactone (0.3 mg/.05 ml trioctanoin) developed 18 sarcomas and three squamous cell carcinomas (in 30 mice). No tumors developed in untreated controls or in controls injected with trioctanoin, an organic solvent, alone (Van Duuren 1977/Ex. 1-284).
In another injection study by the same author, a group of female IRC/Ha Swiss mice were treated with 25 mg vinyl bromide per animal in 0.05 ml trioctanoin once weekly for 48 weeks. The mice were observed for 420 days. One control group was given a weekly injection of trioctanoin alone and the other control group was untreated. No local tumors were seen in any of the test groups, although pathological examination of the animals appears to have been incomplete (Van Duuren 1977/Ex. 1-284).
Application of vinyl bromide to the skin of female ICR/Ha Swiss mice at a dose of 15 mg per animal administered in 0.1 ml of acetone three times weekly for 420 days resulted in no tumors. When this solution was applied once and was followed by an application of phorbol myristyl acetate (PMA) three times weekly, one of 30 mice developed a skin papilloma at 412 days, one control treated with PMA developed a tumor after 44 days, and no untreated controls developed tumors (Van Duuren 1977/Ex. 1-284).
In another dermal study, a dose of 0.1 ml of polymerized vinyl bromide in an aqueous latex suspension was applied three times weekly to the skin of female ICR/Ha Swiss mice for 420 days. No skin tumors developed. When this solution was applied once, followed by an application of PMA three times weekly, one of 30 mice developed a skin tumor at 175 days. No untreated controls developed skin tumors (Van Duuren 1977/Ex. 1-284).
The Benya et al. (1982/Ex. 1-244) study was a well-designed and -conducted study that yielded sufficient information for quantitative risk estimation. The route of administration used in the study, inhalation, is directly applicable to occupational exposure, and the incidence of hepatic angiosarcoma was significant. Angiosarcoma is a rare and malignant neoplasm that has a very low background incidence in animals and humans. Therefore, its appearance in the exposed rats supports the premise that vinyl bromide is potentially carcinogenic in humans. Also, it is the same tumor that is associated with the exposure of workers and animals to vinyl chloride, a recognized human carcinogen and a compound whose structure is similar to that of vinyl bromide.
To estimate excess cancer risk over background incidence for a chemical, experimental data (experimental doses and corresponding responses) are used to define various parameters of an assumed response model. At low doses, the slope of this dose-response curve is referred to as q1. The 95-percent upper-bound confidence limit for this slope is referred to as q 1* or the chemical’s potency. q1 and q1* are then used to determine the respective maximum likelihood estimate (MLE) of risk and the 95-percent upper-bound confidence limit (UCL) on risk associated with a given lifetime occupational exposure. A nonthreshold, linearized multistage model (GLOBAL83) was chosen to estimate the risk potentially associated with exposure to vinyl bromide because the scientific rationale for this model is biologically the most plausible. Additionally, the choice of a nonthreshold model is consistent with current methodologies when positive mutagenicity data are available (Guidelines for Carcinogenic Risk Assessment, EPA 1984d).
Since both male and female rats responded equally to vinyl bromide treatment, data from the two groups were combined by calculating the geometric means of the risk estimates derived from the male and female response data (Anderson 1983/Ex. 1-1009). The high-dose data for each test group were dropped, since their inclusion makes the dose-response curve nonmonotonic and precludes proper fitting of the linearized multistage risk model (EPA 1984d).
Since cancer risk modeling assumes lifetime exposure, adjustments were made to fit the animal data to this criterion. The adjustments made for the data in the Benya et al. (1982/Ex. 1-244) study were: multiplying dose by 5/7 to adjust for days of exposure per week and by 6/24 to adjust for hours of exposure per day. These adjusted doses were then changed to human equivalent doses.
Three hypothetical occupational exposure limits, 5 ppm, 20 ppm, and 250 ppm, were used to calculate the maximum likelihood estimates of risk of developing angiosarcoma of the liver. Five ppm has been the ACGIH limit since 1978. Twenty ppm was chosen as an intermediate exposure level, and 250 ppm was the ACGIH TLV before the ACGIH reduced it in 1978. These occupational dose levels were also adjusted for lifetime exposure. The adjustments made were: multiplying dose by 5/7 to adjust for days worked per week, by 50/52 to adjust for vacation time, by 8/24 to adjust for hours of exposure per day, and by 45/70 to adjust for work years per lifetime.
Because inhalation is the primary route of exposure to vinyl bromide in occupational settings, the occupational dose was calculated assuming that air intake in humans is 20 m3 per 24-hour day (Anderson 1983/Ex. 1-1009). The fraction of vinyl bromide absorbed was assumed to be 100 percent, because no absorption rate data were available for vinyl bromide. Because the log p (lipid solubility) value for vinyl bromide (1.52) is similar to that for vinyl chloride (1.38), OSHA assumed that the absorption rates of these two compounds would also be similar. The absorption rate for vinyl chloride used in risk estimations is assumed to be 100 percent (IRIS 1988).
The MLE shown in Table C15-10 for an occupational exposure to 250 ppm of vinyl bromide is 870 excess deaths per 1,000 workers. According to the linearized multistage risk model, 870 of 1,000 workers exposed over their working lifetimes to vinyl bromide at 250 ppm are at risk of developing angiosarcoma. The MLE for an occupational exposure to 5 ppm of vinyl bromide is 0.04; this indicates that, at the proposed PEL, 40 workers per 1,000 exposed to this substance over their occupational lifetimes are at risk of developing angiosarcoma.
NIOSH (Ex. 8-47, Table N6B) believes that the limit for vinyl bromide should be determined in a full Section 6(b) rulemaking. The International Chemical Workers Union (Tr. p. 216) and the Workers Institute for Safety and Health (Ex. 116) were both of the opinion that the residual cancer risk remaining at the proposed 5-ppm PEL is excessive, and that a further reduction in the PEL is warranted. OSHA agrees with these commenters that significant risk has not been eliminated at the 5-ppm level. However, as explained in Section III of this preamble, the broad scope of this rulemaking prevented the Agency from making detailed analyses of risk and feasibility for alternative PELs. As priorities indicate and resources permit in the future, OSHA may consider the need for a further reduction in the PEL for vinyl bromide.
Table C15-10 shows that workers exposed to this substance, which was formerly not regulated by OSHA, are clearly at significant risk of developing hepatic angiosarcomas, the same rare type of tumor associated with exposure to vinyl chloride, a structurally similar substance. OSHA determined in its prior rulemaking on vinyl chloride that this disease constitutes a material impairment of health and functional capacity. Promulgating a PEL of 5 ppm will not eliminate this significant risk, because, as Table C15-10 shows, the MLE estimate of residual risk at 5 ppm is 40 excess deaths per 1,000 exposed workers. Thus, residual risk at 5 ppm is clearly significant. At the present time, OSHA concludes that establishing a PEL of 5 ppm TWA will substantially reduce the significant risk of cancer potentially associated with exposure at the uncontrolled levels formerly permitted in the absence of an OSHA limit for this substance.