Review of the U.S. Navy's Exposure Standard for Manufactured Vitreous Fibers (2000)

IN 1995, the Navy's rationale for adopting an exposure standard of 2 fibers/cm³ for manufactured vitreous fibers (MVF) stemmed from its desire to establish a standard that was as strict as or stricter than occupational exposure levels proposed or used by other organizations (Navy Environmental Health Center 1997b; Krevonick 1998). The exposure standard of 2 f/cm³ was based on the systematic adoption or incorporation of existing laws and standards as defined in the Navy's 1994 Occupational Safety and Health Program Manual (CNO 1994). As stated in the manual, “...instructions based on these standards may simply refer to a specific Occupational Safety and Health Administration (OSHA) standard (e.g., 29CFR1910.95) or may paraphrase, transpose, or otherwise adopt the standard without altering the basic criteria” (CNO 1994).

When the Navy adopted as its standard for MVF the exposure level of 2 f/cm³ as measured by phase-contrast optical microscopy (PCOM), the National Institute for Occupational Safety and Health (NIOSH) had a recommended exposure limit (REL) of 3 f/cm³, which had been established in 1977, whereas OSHA had proposed permissible exposure limit (PEL) of 1 f/cm³ in 1992 (Krevonick 1998). The OSHA PEL was later withdrawn on the basis of the remanding of the entire air contaminants rule under which the PEL had been proposed. The NIOSH REL was based on exposure information from four epidemiological studies (Dement 1976; Hill et al. 1973; Konzen 1976; Fowler et al. 1971) that found an absence of health effects in workers exposed to glass fibers, at a mean concentration of about 3 f/cm³ (NIOSH 1977). OSHA's proposed PEL was based on the risk of nonmalignant respiratory disease (OSHA 1992) and was supported by results of several epidemiological studies, includ

ing Bayliss et al. (1976) and Enterline et al. (1987). OSHA was also concerned with potential carcinogenic effects of refractory ceramic fibers (RCF). In evaluating the existing exposure level, the Navy reviewed the scientific data on MVF but concluded that neither 3 f/cm³ nor 1 f/cm³ was necessarily scientifically justified. It selected a standard of 2 f/cm³ because this it was more stringent than the NIOSH REL but not as stringent as the OSHA PEL (Krevonick 1998). The Navy selected an exposure level that was the average of the two existing occupational exposure limits.

In January 1999, the Navy revised its Occupational Safety and Health Program Manual, changing the occupational exposure limit for MVF to the American Conference of Governmental Industrial Hygienists (ACGIH) threshold limit value (TLV) of 1 f/cm³ (CNO 1999). The Navy then requested that the subcommittee address several questions concerning the selection and adoption of the original 1995 standard of 2 f/cm³ and the appropriateness of the newly established standard of 1 f/cm ³. These questions were: Are there any studies in the scientific literature that the Navy should have considered that would have justified an exposure standard of more than or less than 2 f/cm³? Is the process that the Navy used to adopt the 2-f/cm³ standard appropriate for MVFs? Does the subcommittee feel that the reduction in the exposure standard from 2 to 1 f/cm³ is justified on the basis of current scientific evidence?

The subcommittee believes that the Navy did not fail to consider any specific scientific studies that would have justified selecting an exposure standard that was higher or lower than 2 f/cm³ when it adopted this standard in 1995. However, the subcommittee does not consider the method used by the Navy to adopt the 2-f/cm ³ standard to be justifiable on the basis of the exposure limits given by OSHA and NIOSH, nor was it scientifically defensible. The Navy could have chosen to adopt one of the existing exposure limits (that is, NIOSH or OSHA), as stated in the Navy Occupational Safety and Health Program Manual (CNO 1994). If the Navy had elected to use an existing limit, it should have considered selecting the more conservative exposure limit of 1 f/cm³. This exposure limit would be more protective of workers, particularly in light of the toxicological data that suggested the carcinogenic potential of RCF (Davis et al. 1983; Mast et al. 1995a; Mast et al. 1995b; McConnell et al. 1995). Alternatively, if the Navy intended to derive a new exposure standard, a risk assessment based on a rigorous scientific evaluation of the toxicological and epidemiological data should have been conducted to

develop an exposure limit at which potentially no adverse health effects would be observed. Although the Navy's document reviews much of the relevant literature on physical, chemical, toxicological, and epidemiological data on MVF, the Navy did not conduct the kind of systematic analysis, integration, and thorough evaluation of alternative arguments that are necessary for the kind of scientifically based risk analysis within which the Navy hopes to couch its exposure standards. Nor was the Navy able to find the scientific justification that it sought for its proposed standard through reference to a rigorous risk analysis conducted and documented by another organization.

It must be acknowledged that developing an exposure standard for MVF based on a quantitative analysis of potential health risks is a great challenge, given the scientific issues and the incomplete and inconsistent data available. Nonetheless, analyses by Maxim et al. (1999a), Fayerweather et al. (1997), Moolgavkar et al. (1999), and Wilson et al. (1999) show that animal studies can be examined quantitatively, with appropriate caveats, to develop a low-dose exposure standard that aims to protect humans from chronic health effects, such as chronic obstructive pulmonary disease (COPD), fibrosis, and cancer. Elements of those analyses bear discussion, but they generally show that an exposure standard of around 1 or 2 f/cm³ should result in relatively small, if any, chronic health risks. For example, Maxim et al. (1999a) found that exposure to two new MVF (Isofrax and Insulfrax) at 1 f/cm³ would result in a working-lifetime cancer risk of less than 10^-5. Moolgavkar et al. (1999) reported that exposure to kaolin-based RCF at 1 f/cm³ would be expected to result in an excess probability of lung cancer of 3.7 × 10^-5 for workers at the age of 70 years. Wilson et al. (1999) analyzed risks of lung cancer in blown-glass workers and found that risks ranged from 2.0 × 10^-6 for nonsmoking workers blowing glass fibers with a binder to 2.4 × 10^-4 for smokers blowing glass fibers without a binder; exposures were assumed to be at 1 f/cm³. The analyses also suggest that the lack of clear responses in the existing epidemiological studies is to be expected, given typical occupational exposure.

The Maxim et al. (1999a) and Fayerweather et al. (1997) analyses highlight some critical concerns:

• The lack of pronounced responses in rat and hamster inhalation bioassays is itself informative in limiting the tenable estimates of potency.

• It is possible to consider the implications of exposures to less durable fibers by examining risks posed by more durable fibers, as long as an allowance for biopersistence is made.

• Although critical assumptions must be made about dose-response curves and cross-species doses, the sensitivity of the outcome to the choice of risk assessment model can be illuminated with appropriate analyses.

Apart from those specific concerns, some key issues must be carefully examined:

• The biological mechanisms by which fibers might induce tumors must be examined for their impacts on the dose-response relationship. In particular, the relationship of particle overload and inflammation to high-dose effects observed in animals and the expected role of those processes at lower doses should be investigated. Existing models accommodate such processes poorly.

• The sensitivity of results to the selection of a deposition model must be considered, and the differential deposition patterns in rats and humans assessed.

• The considerable advances in understanding of fiber-clearance processes need to be incorporated into the analyses. These advances include better assessment of biopersistence, of the distribution of fiber dimensions over time, and of the impact of that distribution process on toxicity.

• Risk estimates are markedly influenced by the units that are used to express the dose of fibers when extrapolating from animals to humans (for example, fibers per unit surface area or fibers per unit volume of lung tissue).

• The presumption of tumorigenic equivalence across species for a specific achieved fiber concentration (after adjusting for biopersistence) needs to be examined, and potential alternatives need to be explored.

• Existing analyses focus on tumorigenicity, but fibrosis, bronchitis, and other potential chronic noncancer responses might be important in evaluating the health protectiveness of a fiber exposure standard.

In sum, even if the Navy were to choose to adopt an existing quantitative analysis, further discussions and exploration of assumptions would be in order. Accordingly, it is difficult for the subcommittee to provide a detailed review of the Navy's assessment and conclusions —in particu

lar, to address explicitly the question regarding the scientific validity of the Navy's proposed standard for exposure to MVF. The Navy's documentation does not articulate any specific risk analysis, nor does it attempt to justify its choice of exposure standards on the basis of risk-based goals.

The subcommittee does, however, support the Navy's recent adoption of the 1-f/cm³ standard in accordance with the Navy's Occupational Safety and Health Program Manual (CNO 1999). The ACGIH TLV is based on concern for irritation of the upper respiratory tract (Konzen 1980). Furthermore, the 1-f/cm ³ standard is also in accordance with OSHA's recommended PEL of 1 f/cm³ for glass fibers and rock and slag wools. An exposure standard of 1 f/cm³ has also been endorsed by the North American Insulation Manufacturers Association and the Building and Construction Trades Department of the American Federation of Labor-Congress of Industrial Organizations (ACGIH 1997), providing further support for this exposure limit. That endorsement, which stems from a formal agreement between manufacturers, fabricators, installers, and removers of MVF, is based on the feasibility of achieving the standard in the workplace by the manufacturers and users of MVFs, and not on an assessment of risk (NAIMA 1999). An exposure standard of 1 f/cm³ is also in agreement with standards recently adopted by New Zealand, Norway, Sweden, and the province of Alberta, Canada (ACGIH 1997; see Table 7-1). The subcommittee notes, however, that none of the existing MVF exposure limits is based on a quantitative analysis of risks of any chronic respiratory disease, including cancer.

Having reviewed the Navy's document, the subcommittee finds that the coverage of the relevant literature is, in some cases, outdated and incomplete. In particular, recent advances in understanding the basis and course of fiber dissolution and clearance rates in the respiratory tract are not fully covered, and recent quantitative analyses from animal bioassays are not addressed. Strengths and limitations of the epidemiological data are not sufficiently explored. For example, the Navy does not consider the impact of known or likely biases in the cited studies, such as exposure and diagnostic misclassification, confounding caused by lack of data on smokers, and, most important, the differences in exposures between workers who manufacture MVF and Navy personnel and contractors who are engaged primarily in removing MVF. However, even if those shortcomings in the review of the literature were remedied, the integration and analysis of information needed to establish a sound, risk-based scientific basis for the proposed standard has yet to be undertaken. The subcommittee believes that if the review of the toxicological, epidemiological, and risk-assessment literature were updated, the resulting data set would support a more extensive risk analysis than has so far been attempted.

The subcommittee notes that although a 1-f/cm³ standard is appropriate for MVF that include conventional glass fibers and rock and slag wools, it is unlikely to be sufficiently protective for MVF that are more biopersistent or have a high proportion of respirable fibers that are much smaller in diameter than the conventional glass fibers and rock and slag wools, which average at least 6 µm. For instance, some RCF and other durable MVF are sufficiently biopersistent to warrant a more restrictive exposure limit. Furthermore, some specialty fibers are both more biopersistent and smaller diameter than conventional glass fibers; many long glass fibers have average diameters less than about 0.3 µm and cannot be seen by PCOM analysis. Inhalation toxicity studies in rats and hamsters have demonstrated the carcinogenic effects of exposures to RCF and durable fiber glass that are currently regulated under the Navy's MVF standard (Gelzleichter et al. 1999; Mast et al. 1995a,b; McConnell et al. 1995, 1999).

The hazard associated with long, biopersistent fibers is related to the number concentration of all the long fibers, not just those measured by PCOM. For conventional glass fibers, rock and slag wools, and RCF, nearly all the fibers are resolvable by PCOM. For specialty glass fibers that are thin and biopersistent, the Navy should use scanning electron

microscopy analyses that record the number concentrations of fibers longer than 5, 10, and 20 µm.

Thus, in addition to supporting the Navy's adoption of the 1-f/cm³ standard for conventional glass fibers and rock and slag wools, the subcommittee believes that the Navy should consider establishing separate exposure standards for the more biopersistent MVF in recognition of their greater hazard potential. In 1997, ACGIH proposed an exposure level of 0.1 f/cm³ for RCF, which was amended to a proposed 0.2 f/cm³ in 2000; but ACGIH has not provided a scientific rationale for either of these levels. Furthermore, ACGIH has not moved this recommendation from the Notice of Intended Change to the adopted list. However, regulatory agencies in other countries have examined the issue and adopted standards for RCF that generally range from 0.5 to 1.0 f/cm ³ (Table 7-1). In 1997, the Refractory Ceramic Fiber Coalition (RCFC), a trade organization of U.S. RCF manufacturers, adopted a recommended exposure guideline of 0.5 f/cm³ (RCFC 1997). Given the extensive global review of this issue, it would be reasonable for the Navy to adopt a more stringent exposure level for the more biopersistent MVF.

In May 1993, the RCFC and the U.S. Environmental Protection Agency (EPA) entered into a voluntary exposure-monitoring consent agreement. Under the 5-year agreement, RCFC obtained and analyzed over 4,500 personal occupational-exposure monitoring samples of RCF from manufacturing process through user installation and removal. Across the lifecycle categories of RCF, average ambient concentrations ranged from 0.1 to 1.1 f/cm³. A proposed RCFC-EPA-OSHA-NIOSH agreement to continue the exposure-monitoring program is under discussion.

In light of the recently revised exposure standards for MVF that reflect improved emission-control measures and new scientific evidence of the adverse effects of MVF, particularly RCF, the subcommittee believes that the Navy should periodically reevaluate its exposure standard for MVF. The MVF exposure levels developed by OSHA, NIOSH, and ACGIH should be evaluated in an effort to maintain Navy exposure guidelines that are consistent with those adopted by these organizations.

One other issue that should be addressed by the Navy is the extent to which any standard developed by others on the basis of epidemiological studies of workers in MVF manufacturing is protective of Navy personnel engaged in the installation, maintenance, or removal of MVF in confined spaces. Navy personnel typically use specific glass fibers and wools for insulation purposes and continuous-glass fibers for reinforcement appli

cations. For these workers, exposures can be highly variable in terms of concentration and relatively acute, in contrast with the generally continuous and longer-term exposures that take place in the manufacturing industry. For RCF that have experienced repeated thermal stress, the resulting airborne fibers can have altered physical and chemical properties, including length distributions and crystalline forms that affect their biopersistence (TIMA 1993, p. 36). Non-RCF MVF, however, melt when exposed to heat so transformation to crystalline forms is of less concern (TIMA 1993, p. 28). The Navy should evaluate how appropriate the exposure standards are for protecting the health of the its personnel who are exposed to MVF that might have been altered by thermal stress.