Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
1 Executive Summary ORIGIN OF THE STUDY l Nonoccu tiona1 health risks associated with exposure to airborne asbestifonm~ fibers have aroused concern because the adverse effects from occupational exposure to some types of these fibers have been well documented and because asbestos as well as other mineral and synthetic particles with similar properties are Widespread in the environment. Moreover, an excess occurrence of asbestos-related disease has been found among people who were not themselves occupationally exposed but who lived either near industrial facilities where asbestos was used or in households with asbestos workers. In this report, the Committee on Nonoccupational Health Risks of Asbestiform Fibers considers the health risks posed by nonoccupational airborne exposures to asbestos and other natural or synthetic asbestiform fibers. The issue is important because many people may be exposed to these materials, although at relatively low levels. To reach a better understanding of the relationship between charac- teristics of asbestiform fibers and possible adverse health effects from nonoccupational exposures, the U.S. Environmental Protection Agency asked the National Academy of Sciences to undertake a Study with two goals: · to evaluate the human health risks associated with nonoccupational exposure to asbestifo~" fibers, with emphasis on inhalation of outdoor arid indoor air, and . to determine the extent to which the physical- chemical properties of the fibers may be associated with the development of various human diseases and the extent to which such information may be incorporated into assessing health rinks resulting from exposure to the fibers. The committee found that much more information is available about asbestos than about the other materials of concern. Wherever possible, the committee compared data on the nonasbestos fibers with data on asbestos. This comparison required assessment of information on asbestos as well as on other materials. The term "asbestiform" in this report refers to fibers that share some specific physical properties with asbestos. These are described later in this summary and in Chapter 2. The term is a mineralogical one that has been used for more than a century. 1
2 MAJOR FINDINGS AND RECOMMENDATIONS Evaluation of Risk Nonoccupational exposure to asbestiform fibers in air presents a risk to human health. The extent of this risk i8 highly uncertain, depending on the nature and amount of exposure and other factors. Evidence for the existence of the risk includes the following: · Large excesses of lung cancer, mesothelioma, pulmonary fibrosis, and other pleural abnormal it lea have been found among workers occupa- tionally exposed to asbestos. Presumably, nonoccupational exposures would result in qualitatively similar effects. · Both a statistically excessive number of cases of mesothelioma and an excess frequency of pleural abnormalities have been observed among household contacts of asbestos workers. · Asbestiform fibers are distributed extensively outside the work- place, although usually in minute quantities. · The major pathological effects associated with human exposure to airborne asbestos have been duplicated experimentally in animals. · Increases in cell replication and other abnormalities have been seen in cultures of tracheal lining cells from humans and animals after the cells were exposed to synthetic or natural asbestiform materials. Estimating the extent of health risks from nonoccupational exposure to asbestiform fibers is fraught with uncertainty. Factors contributing to that uncertainty inc. Jude the fol lowing: · A great variety of asbestiform fibers has been found in the non- occupational environment. These fibers occur in a range of sizes and vary in physicochemical characteristics, such as flexibility and durability. · It is difficult to standardize methods for measuring amounts and characteristics of a~bestiform fibers. · A long time is required for health effects in humans to become detectable after exposure begins (often 20 to 40 years). ~ There is inadequate knowledge of the mechanisms by which asbestiform fibers lead to cancer and other health effects. ~ There are uncertainties in determining dose-response relationships from the occupational environment and then extrapolating them to the nonoccupational environment, where both exposure and population characteristics are usually very different and doses are typically much lower. 1 i
3 The committee made estimates of comparative risk of adverse health effects that might result from exposures to various a~bestiform subetances. It concluded that population risks from exposure to the other materials considered would generally be lower than the risks from exposure to chry~otile asbestos because the opportunities for airborne exposure to particles of respirable size were generally less for the other substances than for chrysotile. The nonasbestos materials considered were attapulgite (the trade name for the mineral palygor~kite, which exists in asbestiform and nonasbestiform varieties); several man-made mineral fibers, such as fibrous glass, mineral wool, and ceramic fibers; carbon fibers; and fibrous erionite. These were selected because they seemed likely to have, or are known to have, at least some of the properties of asbestiform fibers. The committee made a quantitative estimate of the risk of excess lung cancer and mesothelioma that might occur in persons breathing low levels of asbestos in the air. A concentration of 0.0004 fibera/cm was deemed reasonable to use in such calculations because a variety of measurements of indoor and outdoor air indicated that 0.0004 fibere/cm3 is the approximate average level that may be encountered. If a person inhaled air containing asbestos at that level throughout a 73-year lifetime, the committee's best judgment in that the lifetime risk of mesothelioma would be approximately nine in a million (range O to 350 per million, depending on assumptions regarding the relationship of dose to risk). Others have produced different estimates that are discussed in this report. Risks for continuous lifetime exposures to higher or lower levels would be proportionately higher or lower. Epidemiological data and the estimates derived from them indicate that the corresponding lifetime risk for lung cancer would be about 64 in a million for male smokers (range 0 to 290), 23 in a million for female smokers (0 to 110), and 6 and 3 in a million, respectively, for male and female nonsmokers. The risk to nonsmokers appears greater for mesothelioma than for lung cancer. Because of the great reliance on assumptions and on clearly deficient exposure and effects data, the committee views these risk estimates as guides to the qualitative assessment of nonoccupational health risks from asbestos and asbestiform fibers--not as definitive estimates of the amount of disease to be anticipated. These estimates and other considerations lead to the following five conclusions about risk: 0 Some deaths from mesothelioma and lung cancer will probably result from current and past levels of exposure to asbestos in ambient air. · Excess deaths from other diseases, such as asbestosis, and from exposures to other asbestifo.= fibers are also possible but are not likely to be as numerous as those from asbestos-induced mesotheliomas or lung cancer. 0 The numbers of annual or cumulative deaths expected to result from such exposures are very uncertain, but they are virtually certain to be lower, and probably much lower, than those resulting from past, heavier occupational exposures to asbestos.
4 o Deaths from past occupational exposures to asbestos can reasonably be estimated to total several thousand per year in the United States during the next few years. Among these, more deaths from lung cancer than from mesothelioma can be expected. O The greatest risks of continuous lifetime exposure to asbestos would be to smokers, who would be most at risk of lung cancer. However, the risks to nonsmokers might well be greater for mesothelioma than for lung cancer, because of the strong dependence of mesothelioma rates on time from first exposure. The time dependence factor also implies that restricting exposures of children to asbestos would be even more effective than a corresponding restriction for adults in reducing the lifetime risk of mesothelioma. Physicochemical Properties and Health Effects Some of the physical properties of asbestiform fibers appear to be important in causing adverse health effects, but the specific properties that are necessary and sufficient are not known. One clearly important characteristic is respirability. In addition, longer, thinner fibers appear to be more pathogenic than shorter, thicker fibers, but there is not a minimum size below which no effects would be expected. However, nonfibrous particles generally do not induce mesotheliomas in animals. Number, rather than mars, and durability of fibers also seem to be significant factors in pathogenicity of asbestiform fibers. All major commercial types of asbestos fibers used in the United States have been associated with lung cancer, mesothelioma, and asbestosis in humane. It is not known whether the physicochemical fiber properties responsible for fibrosis are similar to those involved in carcinogenesis. Recommenda t ions The committee's findings ant analysis let to the following recommends t ions: 1. Systematic monitoring and characterization of asbentiform fibers with standardized methods should be undertaken in nonoccupational environments, including urban, rural, indoor' and outdoor locations where exposure may be of special concern. 2. A program of systematic surveillance should be undertaken to determine the extent to which the occurrence of mesothelioma and lung cancer is associated with exposure to asbe~tifos~m fibers. 3. Cessation of cigarette smoking should be encouraged in view of the multiplicative effect of smoking and asbestos exposure in increasing the risk for lung cancer. 4. Steps should be taken to educate both the medical profession and the general public concerning possible exposures to asbestiform fibers and the resulting health effects.
. 5 The committee also made several recommendations concerning future research to resolve the many questions about the health risks of nonoccupational exposure to asbestifo~m f ibers. 1. A standardized terminology for asbestiform fibers should be adopted. The terminology should be based on mineralogical analysis and should distinguish these fibers from other types of particles. 2. The various characteristics of asbestiform fibers or ocher particles used in experiments should be described as completely as possible. 3. Standardized methods for measuring and characterizing asbestiform fibers should be improved. 4. In viva and in vitro laboratory ~ tud ie s wi th a she s t i f arm f ibe r s and non fib rous substances should be conducted to investigate the physico- chemical properties that are responsible for the biological effects. 5. Clinical studies of lung cancer, mesothelioma, and fibrosis should be continued with emphasis on the possible role of asbestiform fibers. 6. Epidemiological studies are needed to clarify further the rela- tionships between exposure to fibers and adverse health effects. These studies should include case-control studies for mesothelioma and lung cancer and prospective cohort studies among persons occupationally exposed to materials such as asbestos, attapulgite, and man-made mineral fibers. 7. To improve risk assessments, studies should be conducted to elucidate the relationships between amount of exposure and time factors and the development of adverse health effects. SUMMARY OF THE STUDY l Background Asbestos has been detected in both outdoor and indoor air, although almost always at concentrations far below the standard established by the Occupational Safety and Health Administration (OSHA) for the workplace. Since 1976 the workplace standard has been 2 fibere/cm3 for fibers longer than 5 Am seen in a phase contrast light microscope under specified conditions. The general population is also exposed to other fibrous materials with some of the same physical properties as asbestos but whose effects on health are not well known. These materials include man-made mineral fibers such as fibrous glans and mineral wool, which are sometimes used as substitutes for asbestos, as well as certain natural asbestiform varieties of minerals not marketed as asbestos. Sources of exposure to asbestiform fibers may be roughly divides into three broad categories:
6 · naturally occurring asbestiform fibers used commercially, such as asbe stop; · commercially used synthetic fibers with some properties similar to ! those of asbestos; and · naturally occurring types of asbestiform fibers that are not used commercially. There is a substantial amount of data on exposures in the workplace, but very little information on nonoccupational exposures. In the occupational setting, four diseases have been clearly associated with exposures to asbestos. These are (1) lung cancer; (2) mesothelioma, a rare but almost invariably fatal cancer of the tissues that line the chest cavity (pleural mesothelioma) or the abdominal cavity (peritoneal mesothelioma); (3) asbestosis, a nonmalignant, progressive fibrosis of the lung that may result in severe disability and death; and (4) nonmalignant pleural disease, including diffuse pleural thickening and effusions and the formation of fibrous and calcified plaques. The occurrence of these four diseases in various occupational settings and the presence of asbestiform fibers in the general environment led to current concern about potential health effects from nonoccupational exposures. During the course of its study, the committee was confronted with several difficulties: · Fibers in the general outdoor environment seem to differ in size and other physicochemical properties from those in the workplace; however, it is not easy to characterize these materials. Dif ferent types and samples of fibrous materials vary greatly in their physical properties, even when they are composed of the same mineral. Therefore, it is difficult to develop a consistent methodology for determining and expressing the characteristics and concentrations of fibers found in different environments or used in laboratory studies. · Because most health effects data are based on workplace exposures, it is necessary to extrapolate results from relatively high occupational concentrations to the much lower concentrations of fibers typically found outside the workplace. Although the health consequences are presumably similar among workers and nonworkers, incidence rates would be expected to be lower and the nonmalignant changes less severe among persons nonoccupationally exposed to lower levels of asbestos. Thus, the effects would be more difficult to detect. · Other factors associated with the diseases must be considered. For example, cigarette smoking multiplies the effect of asbestos in causing lung cancer. · The mechanisms by which the fibers produce disease are not well understood, nor is it clear how the fibers reach various parts of the body. i
7 ~ The length of time from initial exposure to the expression of certain health consequences is often several decades. Thus, current disease is the result of past exposures, whereas present exposures will produce disease only many years in the future. Exposure in childhood may increase the possibility of ultimate damage to health, because disease can occur long after external exposure has ceased and more years of life remain for children. Thus, great uncertainty is likely to attend any conclusions drawn about the relevant fiber characteristics and the health risks that may accompany exposure. The committee agreed that the major potential for future fiber- associated health problems is probably presented by inhalation exposures to airborne fibers rather than by the ingestion of there materials, for example, in water. Mont of the committee's attention was therefore devoted to airborne fibers of respirable size, that is, to fibers less than approximately 3 Em in diameter. The committee made quantitative risk assessments for lung cancer and mesothelioma from inhaled asbestos, but it did not attempt quantitative risk assessments for other cancers, from inhalation of other asbestiform fibers, or from ingestion of asbestifo~m fibers in water or food. Materials of Concern For purposes of this report, the term "asbestiform fibers" is used broadly to include both naturally occurring and certain synthetic inorganic and carbon fibers that share some specific physical properties with asbestos. Asbestos, the prime example of an asbestiform material, consists of the commercially marketed asbestiform varieties of several silicate minerals. They are primarily chrysotile, crocidolite, and the asbestiform variety of some amphibole minerals marketed as "amosite." Chrysotile accounts for approximately 95: of the asbestos currently sold in the United States. Because of its great strength, flexibility, and heat resistance, asbestos came into extensive use during the 20th century for textiles, thermal and electrical insulation, and high strength reinforce- ment in such products as vinyl-asbestos flooring and asbestos-cement sheet and pipe. In 1982, the United States used approximately 6: of the world production of asbestos. Five basic physical properties distinguish asbestifonm fibers from other materials. The presence of these properties generally depends on the physical and chemical conditions under which the fibers grow. Compared with a nonasbestiform variety of the same mineral, the properties are:
t 8 · microscopic, fiberlike dimensions and morphology, i.e., the fibers are much longer than wide; · enhanced strength and flexibility; · inverse relationship between diameter and strength, i.e., the smaller the diameter, the greater the strength per unit cross-sectional area; · enhanced physical and chemical durability; and · high quality, relatively defect-free surface structure. "High quality" fibers have all these properties to a great extent; "low quality" fibers possess them to a lesser extent. The presence of these properties does not necessarily indicate that a material is either carcinogenic or fibrogenic. Because these properties are interdependent and variable, naturally occurring asbestifo~m fibers, even those composed of the same mineral, have a range of physical characteristics. In contrast to fibers in the workplace, it is not currently possible to determine the sources of most fibers in the ambient environment or the extent to which these fibers have the above properties. Mineralogical terms pertaining to asbestiform fibers have sometimes been used inaccurately in scientific reports, including the literature on biological effects. As a result, it may be impossible to discern the composition of materials studied and extremely difficult to draw conclusions about their physical properties and biological effects. Relationship of Fiber Characteristics to Health Effects Various physical properties of asbestiform fibers appear to play a role in causing adverse health effects; however, the specific properties that are necessary and sufficient to produce such effects have not been postively identified. Furthermore, it is not known whether the properties associated with a given effect, for example, lung cancer, are the same or different from those associated with other effects, such as fibrosis or mesothelioma. Some characteristics that appear to be important are discussed below, in approximately descending order of the strength of the positive evidence. 1. Respirability. For significant health effects to result from inhalation of asbestiform fibers, the fibers must reach the lower portions of the respiratory tract where they cause the most damage. Although the , limiting upper diameter appears to be about 3 ~m, fibers that are much i longer than wide can penetrate deeply in the respiratory tract. l Length, Diameter, and Aspect Ratio (i.e., Ratio of Length to Diameter). ; Experiments inducing mesothelioma in rodents by injections of test material have indicated that long, thin fibers yield more tumors than do
9 short, thick fibers. Samples with an overwhelming majority of fibers shorter than 5Pm yielded mesotheliomas in rata when injected intraperitoneally, but the pathogenic role of abort fibers, especially those shorter than 3 Am, is unclear. Fibers longer than approximately 10 Pm cannot be completely engulfed and inactivated by macrophages, and they have tended to produce more disease in animal tests than have the shorter fibers. Other Properties. The number of fibers, which is also correlated with surface area, generally appears to be a more relevant measure than mass in determining pathogenicity. Durability also appears to be a factor. The more durable fibers appear to be more pathogenic in some studies than fibers that are less durable. The relevance of fiber surface charge to effects on human health remains to be demonstrated. Some experimental studies have indicated that surface charge appears to be involved in cytotoxicity. Although chemical composition is related to physical properties of asbestiform fibers, a direct role for chemical composition Per se in biological activity has not been demonstrated. Measurement and Extent of Exposure Measurement. Information about asbestiform fibers in the ambient environment, although scanty, indicates that they differ from those in the workplace. Different techniques for measuring the concentrations in the two environments have been used. The phase contrast light microscope has been adequate for counting fibers in the workplace. However, that technique has been less useful for the ambient environment, where fiber identity and character are usually unknown; almost all fibers are too small to be seen by light microscopy; and concentrations, expressed as mass, are usually hundreds or thousands of times lower than those in the workplace. Data on workplace fiber concentrations are generally given as numbers of fibers longer than 5 pm, whereas data on ambient concentrations obtained with transmission electron microscope techniques have usually been espresseed as mass per unit volume. Substantial uncertainty may be introduced in calculations that assume that ambient and workplace exposures differ only in fiber concentration. Furthermore, it is not usually possible to convert mass measurements to fiber concentrations accurately because the various conversion factors that are used assume particular fiber dimensions, and these vary greatly with different environments and sampling techniques. During the early 1970s, mass measurements of asbestos made in various U.S. cities ranged from 1 to 100 ng/m3. If we assume that 30 pg/m3 is equivalent to 1 fiber/cm3 (counting fibers longer than 5 pm through a light microscope), the mass measurements in those cities would feat to an expected concentration of 0.00003 to 0.003 asbestos fibers per cubic centimeter. Extent of Exposure. In assessing the likelihood that individuals would be exposed to various asbestiform fibers, the committee considered
10 patterns of use; production or consumption levels; fiber dimensions, i.e., whether the fibers are of respirable size; and potential for population exposure. In many situations, the fibers are tightly bound in a matrix during product manufacture and, therefore, might be expected to produce little subsequent exposure. Irk the United States, the annual use of asbestos peaked in 1973 at almost 800,000 metric tons, but decreased to approximately 250,000 metric tone, or about 6% of world production, in 1982. However, much of the more than 30 million metric tons of asbestos used in the United States since 1900 is still present in its original application and provides a potential for exposure. Attapulgite (palygorskite) is the only natural asbestiform material used in the United States in amounts greater than those of asbestos. Of the more than 700,000 metric tons used annually, most appears to be classifiable as anbestiform. Most attapulgite fibers are less than 5 Am long and have diameters of approximately 0.03 pm. Some uses of this material could result in the release of fibers, but the committee found no reported measurements of attapulgite in ambient air. Synthetic fibers with some physical properties similar to those of asbestos include man-made mineral fibers, of which more than 1 million metric tons are produced annually I the United States. Typical diameters of most of these fibers exceed the respirable size range, although diameters of fine grades of fibrous glass and some rock wool and slag wool are mostly below 3 ~m. Some fibrous erionite found in deposits in the western United States falls into the respirable size range. Mining and natural weathering of this material could lead to significant local air concentrations, but the committee did not find any measurements of such concentrations. Moreover, the population exposed is probably small. (;urrent U.S. consumption figures and use patterns indicate that future exposure of the general population to attapulgite and fibrous glass is likely to be somewhat greater than exposure to chrysotile, whereas exposure to mineral wool, ceramic fibers, other asbestos fibers, and carbon fibers would be less. However, material already in place would also contribute to total exposure. Health Effects Methodology To develop an understanding of the health risks associated with exposure to environmental agents such as asbestiform fibers, investigators usually evaluate data from clinical, epidemiological, and laboratory studies. Clinical observations often provide the first suggestion that exposure to a particular substance may cause an adverse health effect. Epidemiological studies are then undertaken to attempt to confirm the hypothesized association and to quantify it. Laboratory studies of the l !
1 11 response in animals (in viva) and in cells growing outside the body (in vitro) can provide further information. If a substance administered to animals produces pathological effects similar to those found in humans, the case for its being a causative agent in humans is strengthened. For asbestos, the major diseases observed in humans have been produced in animals by exposure to asbestos. In vitro and in viva studies do not necessarily adequately reflect the amounts or routes of exposures experienced by humans, nor do they take into account individual susceptibilities or other substances to which people might be exposed. Thus, such studies should be extrapolated to humans only with great caution. Except for smoking, however, no environ- mental or genetic factors have been unequivocally shown to influence the chance that a person will develop an asbestos-induced disease. Health Effects of Asbestos Appendix A of this report contains a chronological list of the major findings associating adverse health effects with exposure to asbestos. The first disease to be associated with asbestos exposure was asbestos~s, which was first noted in the early l900s. From 1938 to 1949, numerous autopsy reports indicated that a high proportion of persons dying of asbestosis also had lung cancer. In the 1950s, when the sharp increase in lung cancer attributable to smoking was occurring in the United States and other industrial nations, epidemiologists found that occupational exposure to asbestos also increased the risk of lung cancer, especially among cigarette smokers. In the early 1960s, the association with mesothelioma was established among asbestos miners in South Africa. Lung Cancer. Exposure to asbestos appears to increase a worker's underlying risk of getting lung cancer as much as fivefold. Since a smoker's risk of getting lung cancer is approximately 10 times greater than that of a nonsmoker, an asbestos worker who smokes has up to a 50-fold greater chance of dying from lung cancer than does a nonsmoker who does not work with asbestos. An increase in exposure, expressed as con- centration of asbestos and duration of exposure, appears to increase the lung cancer risk. Epidemiological data suggest that this relationship is linear; the data do not indicate the presence of an exposure threshold below which there is no increased risk. Mesotheliom~. Approximately 1,600 cases of mesothelioma occurred in the United states during 1980, according to projections from cases reported in the lOX of the U.S. population monitored by the National Cancer Institute's Surveil lance , Epidemiology , and End Results (SEER) program. Although exposure to asbestos has been strongly associated with most mesothelioma cases studied, some cases may occur without apparent asbestos exposure. The evidence does not exclude the possibility that ambient exposure to asbestiform fibers was associated with mesotheliomas for which exposure court not be documented. The percentage of workers
12 with mesothelioma has ranged from O to 2Z among chrysotile miners and chrysotile textile workers, but has been as high as 10: among workers who manufactured crocidolite-containing gsa masks. The disease seems to be independent of smoking but related to dose and to time from first exposure. Aabeatosis. All types of asbestos appear to be implicated in the development of asbesto~is. Data indicate that the incidence rate increases and the disease becomes more severe with increasing dust exposure, which is expressed as concentration of dust and duration of exposure. It is not clear whether an exposure threshold exists. Persons in the early stages of this condition may be free of symptoms, but beyond a certain stage the disease seems able to progress even in the absence of further exposure. Pleural Thickening. Another nonmalignant pathological effect of asbestos exposure is the formation of fibrous and sometimes calcified plaques and diffuse thickening of the pleural lining of the chest cavity. Effusion of fluid into the pleural cavity may also occur. Such pleural thickening is suggestive of asbestos exposure but is rarely a cause of significant, direct respiratory impairment. Gastrointestinal Cancer. Excess gastrointestinal (GI) cancers have been fount among some cohorts of asbestos workers, but the excesses were usually substantially less than for lung cancer. Dose-response data are not available. Recent animal feeding studies have failed to demonstrate asbestos induction of GI cancers. Moreover, because of inherent limitations in the epidemiological studies, including the limited sizes of the exposed populations and the lack of individual exposure data, it has not been possible to determine from these studies the extent to which there may be an association between GI cancers in humans and the presence of asbestiform fibers in drinking water. Wealth Effects of Nonasbestos Asbestiform Fibers . . Some natural asbestiform subatances other than asbestos seem to have biological effects similar to those of asbestos. For example, erionite, a fibrous zeolite, readily induces mesothelioma in animal tests, and populations living in central Turkey, where it is present in volcanic Luff, are reported to have an excess incidence of lung cancer, mesothelioma, and pulmonary fibrosis. As another example, epidemiological studisa are being conducted on workers exposes to attapulgite, but as yet there are essentially no data on humans indicating whether it is toxic when inhaled. - Exposure to man-mate mineral fibers is relatively recent, ant the occupational exposure levels apparently have not been as high as those for asbestos. Some epitemiological data do suggest, however, that diseases of the respiratory tract, such as pulmonary fibrosis and lung cancer, may result from long-term occupational exposure to these fibers. it, l .~ )
. 13 Evidence Associatin Fiber Pro erties with Adverse Health Effects Asbestos and other asbestiform fibers appear to react with Celia in a variety of ways. They may alter normal cell function, they may cause cell death, or they may directly or indirectly alter the genetic Information and the way cells replicate. Studies of cells lining the respiratory passages suggest that asbestos may act as a promoter (in the initiator/promoter model of carcinogenesis). The In vitro evidence that asbestos can damage DNA directly or be mutagenic to gene n or chromosomes is weak and inconclusive. Laboratory studies have not identified one type of asbestos as being more potent than others. In animal inhalation experiments, however, asbestos generally appears to be more pathogenic than most other asbestiform fiber e that have been tested. At present, none of the available in vitro motels can be uset to quantify the relative fibrogenic or carcinogenic potential of asbestiform fibers either in animals or in humans. Interpretation of results is hindered by the failure of most reported studies to define the test materials precisely, by the paucity of experiments showing dose-response information, and by the differences in response among species and cell types. Results of studies of various groups of workers indicate that it is extremely difficult to assess the role of fiber type (e.g., chrysotile or crocidolite) in determining the risk for developing either lung cancer or mesothelioma. Analysis of the epidemiological studies is complicated because of variations in type of industry, the diverse fiber characteristics within an industry, and the usual inadequacy of exposure data. Some of the apparent discrepancies may be explained by differences in physical properties of the fibers, their concentrations, and their characteristics in the different environments. These possibilities need further testing. Risk Assessments In general, three steps are necessary before one can assess health risk from environmental exposures: determination that a material is toxic and identification of adverse effects; determination of dose-response relationships; and determination of the extent of exposure. At least Rome types of asbestifonm fibers are toxic and have identifiable adverse health effects. However, few occupational studies have demonstrated tose- response relationships, ant there is great variability among those few studies. Entimates of exposure outside the workplace are particularly difficult to obtain, ant it is the risk from such exposure that is the focus of this report. Other factors that introduce uncertainty into risk assesaments for nonoccupational exposures include assumptions about the magnitude of effects at low doses; differences in the characteristics of fibers in the
14 occupational and nonoccupational environments, especially regarding size and composition; and differences in the populations exposed, age at onset of exposure, and duration of exposure. In this report, risk assessments are limited primarily to mesothelioma ant lung cancer as end points and to inhalation as the route of exposure. For asbestos, sufficient information was available for the committee to make quantitative estimates--albeit with great uncertainty--of the risk for lung cancer or mesothelioma after inhalation exposure. These risk assessments were conducted for a "generalized" asbestos exposure, rather than for exposure to a specific type of asbestos. However, the committee assumed that the rink estimates would also apply if chrysotile were the primary agent of exposure. For the other types of fibers considered, the committee made comparative, i.e., qualitative, risk assessments that were subject to yet greater uncertainty. For the quantitative risk assessment, the committee concluded that the epidemiological data supported the use of a linear, no-threshold model. Dose-response data from workplace studies were used in developing the equations. To estimate nonoccupational exposures, measurements of the mass of asbestos in the ambient environment were converted to the number of fibers longer than 5 Am that would have been found in the workplace at a similar mass concentration. These measured concentrations indicated to the committee that 0.0004 fibere/cm3 was a reasonable level to use in the risk assessment. However, there could be specific circumstances, such as schoolrooms with flaking asbestos, where persons are exposed to higher levels for limited periods. If a person were to inhale air containing asbestos at an average of 0.0004 fibere/cm3 throughout a 73-year lifetime, the committeets best estimate is that the lifetime risk of mesothelioma would be approximately nine in a million (range 0 to 350 per million, depending on assumptions regarding the relationship of dose to risk). The corresponding lifetime risk for lung cancer would be about 64 in a million for male smokers (range 0 to 290), 23 in a million for female smokers (range 0 to 110), and 6 and 3 in a million for male and female nonsmokers, respectively. The risk for mesothelioma is greater than that for lung cancer among nonsmokers because of the strong dependence of mesothelioma risk on time since first exposure. Occupational studies indicate the t me sothe 1 ioma usually fires appears about 20 years after onset of workplace exposures and that the incidence increases rapidly thereafter. The calculations suggest that a given exposure to asbestos in childhood markedly increases the lifetime risk of mesothelioma compared with an equivalent dose later. The risk estimates remain uncertain, especially because they are based on the assumption that the data on occupational exposures are transferable to the nonoccupational ~ ituation. Smal ler f iber s ize in the ambient environment would probably tend to lead to lower risk. The comparative or qualitative risk assessments for the other asbestiform fibers were based on chrysotile and lung cancer an the baseline case. Population risk for particular f ibers was compared with l 1 1
- 15 the population rink for lung cancer from chrysotile. In making comparative risk assessments, the committee considered such factors as respirsbility, biodisposition, and intrinsic toxicity as they are related to population exposures and to individual risk. The materials considered were crocidolite, other asbestos fibers as a group, attapulgite, fibrous glass, mineral wool, ceramic fibers, and carbon fibers. (Appendix H of this report presents the qualitative assessments for each substance.) The risks for developing lung cancer or mesothelioma as a result of exposure to the other materials considered by the committee were usually much lower than those for chrysotile, principally because of a lower potential for airborne exposure or because the fibers are less respirable--not because their intrinsic toxicity is necessarily less. For example, both ceramic and carbon fibers can be found in respirable size ranges and may have some biological properties similar to those of asbestos, but production and opportunities for exposure are low, although increasing. The materials with potentially greatest impact are fibrous glass and attapulgite because of their current large production volume and extensive use. .
