2.0 Comments on Topic 1: Physiological Fate of Asbestos and SVF Fibers Less Than 5 Micrometers in Length
This section summarizes the panelists' discussions on the physiological fate of asbestos and SVF fibers less than 5 µm in length. Two panelists Dr. Lippmann and Dr. Oberdörsterwere designated discussion leaders for this part of the meeting, during which the panelists responded to the three specific charge questions regarding physiological fate of small fibers (Sections 2.1, 2.2, and 2.3) and addressed topics not identified in the charge (Section 2.4). Panelists also commented on the toxicity of asbestos and SVF fibers; these comments are summarized in Section 3. This section also summarizes observer comments made after the panelists completed their discussions (Section 2.5). Overall, this section presents a record of discussion of topics mentioned during the meeting, and it should not be viewed as a comprehensive literature review on the role of fiber length in the physiological fate of inhaled fibers. Dr. Lippmann's post-meeting comments (see Appendix E) also summarize these discussions.
Although the panelists focused their initial discussions on fiber length, several panelists stressed that length is not the only factor affecting fiber toxicity. These panelists noted that toxicity is rather a complex function of the fiber dose, dimensions, and durability, as has been widely documented in the scientific literature.
The first charge question asked the panelists: "What is the expected physiological depositional pattern for less-than-5-µm fibers in the lung?" When responding, the panelists provided relevant background information on lung physiology, reviewed what researchers have established for depositional patterns of particles, and then addressed what is currently known about depositional patterns for fibers:
Background on lung physiology. Before addressing the specific charge questions on how fibers deposit in the lung, one panelist first reviewed fundamentals of lung physiology, which largely dictate fiber dosimetry. He explained how air flows through the respiratory system: inhaled air enters the body at the nose or mouth, passes through the larynx and trachea, and eventually enters the lung in airways that branch numerous times before reaching terminal bronchioles. These airways are all conductive, meaning that they move air to the deeper portions of the lung where gas exchange occurs. The air flow velocity decreases as air moves into the more distant bronchi, because the cross-sectional area of the branched bronchi is greater than that of the parent airways. After passing through the terminal bronchioles, inhaled air enters into respiratory bronchioles, then alveolar ducts, and eventually alveolar sacs, where most gas exchange occurs. Movement of air in the respiratory bronchioles and alveolar sacs is dominated by diffusion, rather than by convective forces.
This panelist noted that clearance processes in the conductive airways differ from those in the airways distal to the terminal bronchioles. In the conductive airways, mucus is secreted onto the airways' surfaces, and ciliated cells on the bronchi and bronchioles gradually move the mucus up to the throat, where the mucus is swallowed. This mucus clearance mechanism efficiently removes particles that deposited on the conductive airways, typically within about 1 day following exposure. The clearance mechanisms for particles that deposit in the respiratory bronchioles, alveolar ducts, and alveolar sacs operate on a much longer time scale (see discussion on "phagocytosis" in Section 2.2).
Depositional patterns for particles. One panelist then reviewed the state of the science of how inhaled particles tend to deposit in the respiratory tract. For both fibrous and non-fibrous particles, the deposition pattern is dictated largely by the particles' aerodynamic diameter. The aerodynamic diameter, another panelist noted, is equivalent to the geometric diameter of a unit density sphere that has the same terminal settling velocity in still air as the particle in question.
The discussion leader then noted that researchers have long established that airborne particles with aerodynamic diameter larger than 10 µm typically do not pass the larynx, and the particles that enter the lungs deposit by one of three mechanismsimpaction, sedimentation, or diffusion (Brownian motion). The relative importance of these mechanisms is a function of the particle size. The largest particles that enter the lung, for example, have the most momentum, which causes them to have a greater tendency to deposit on airways by impaction as air flow changes direction at bronchial airway branches. Smaller particles1, on the other hand, are less likely to deposit by impaction and therefore are typically carried by convective forces further into the lung.
Any particle that enters the respiratory bronchiole will likely deposit either by sedimentation or Brownian motion; impaction is relatively unimportant in regions where the air flow velocity is low. Sedimentation and diffusion tend to be the more dominant mechanisms in the small lung airways for particles, which diffuse in air much slower than gases. One panelist noted that sedimentation is the dominant deposition mechanism for particles with aerodynamic diameters greater than roughly 0.8 µm, while smaller particles are increasingly subjected to diffusional deposition in the airways. Particles depositing in the respiratory bronchioles, alveolar ducts, and alveolar sacs will remain in these regions of the lung until cleared by other mechanisms (see Sections 2.2 and 2.3).
Depositional patterns for fibers. One panelist described depositional patterns of fibers, noting how their elongated shapes caused fibers to deposit differently in the lung than particles. The main difference between fiber and particle deposition is that fibers can be intercepted by airway surfaces, while particles generally cannot. For instance, as long fibers move through small airways, the end of a fiber might contact (and deposit on) an airway surface, even in cases when the fiber's center of mass is on a flow streamline in the center of the airway. Interception can therefore cause enhanced deposition of fibers, when compared to particles; and interception becomes an increasingly important deposition mechanism for longer fibers.
This panelist indicated that many researchers have evaluated the depositional patterns of fibers in the lung. He cited the following studies as examples:
Studies using hollow airway cast models from human lungs have demonstrated that the extent of fiber interception varies with fiber length. Specifically, interception has been shown to be relatively unimportant for fibers less than 10 µm in length (Sussman et al. 1991). This panelist indicated that these shorter fibers will likely act like particles in the lung, because the one deposition mechanism unique to fibers is unimportant.
Other studies using these models have reported that fibers with aspect ratios greater than 10 behave aerodynamically like unit density spheres with diameters three times the fiber width (Stöber et al. 1970; Timbrell 1972). Interception accounts for the fact that longer fibers have proportionally greater deposition in the conductive airways than shorter fibers.
For fibers less than 5 µm in length, Dr. Lippmann indicated, the information available on particle deposition and longer fibers suggests that fiber diameter likely has the greatest influence on deposition patterns. He noted that fibers less than 5 µm in length will have diameters less than 1.66 µm, assuming the aspect ratios are at least 3:1. This panelist estimated that 10% to 20% of short fibers with diameters between 0.1 and 1.6 µm will deposit in the lungs of healthy people.
Another panelist reviewed findings from multiple publications to illustrate how the four mechanismsimpaction, sedimentation, diffusion, and interception-affect fiber deposition patterns. First, this panelist summarized results of a lung modeling study (Asgharian and Yu 1988), which predicted the relative importance of the four deposition mechanisms as a function of fiber diameter. For all fiber dimensions considered, diffusion (Brownian motion) accounted for an increased amount of deposition as air traveled further into the lung. Further, impaction, interception, and sedimentation were relatively unimportant for the thinnest fibers (those with diameters of 0.01 µm), yet accounted for most of the predicted deposition pattern for the larger fibers (those with diameter of 10 µm). Second, he reviewed the extent to which fibers are filtered from inhaled air in the nose versus the mouth, as predicted by mathematical models. The model predicted that, for all fiber dimensions considered, nose breathing is considerably more effective at filtering airborne fibers than is mouth breathing. In fact, appreciable filtration for mouth breathing was predicted only for fibers at least 1 µm in diameter. Overall, these comments highlight that researchers have already predicted how fiber dimension (both length and diameter) affect depositional patterns in the lung (see Dr. Oberdörster's premeeting comments in Appendix B for references to relevant peer-reviewed publications).
Role of laboratory animal studies in evaluating depositional patterns in humans. The panelists acknowledged that laboratory animal studies have provided additional insights on how fibers deposit in the lung, but the panelists noted that inter-species differences in lung airway structure limit the utility of the animal data. One indicated, for instance, that laboratory animal studies have the advantage of being able to characterize lung fiber burdens at different time frames following highly controlled dosage conditions. On the other hand, he added, airway branching patterns in humans are nearly symmetrical, while rats (and most other mammals) have asymmetrical branching patterns. Such differences in branching patterns influence the cross-sectional air flow profiles, which in turn affect fiber deposition behavior. Consequently, lung deposition patterns in laboratory animals are expected to differ from those in humans.
Another panelist showed how modeling results of lung deposition patterns support this expectation. Based on predictions of a mathematical lung dosimetry model developed by the International Commission on Radiological Protection, this panelist illustrated differences between rats and humans in estimated deposition fractions of fibers in alveolar regions. His figure indicated that the predicted deposition fraction in humans was greater than that in rats for all fiber lengths considered, and this difference was most striking for longer fibers. Specifically, the model predicted that virtually no fibers with aerodynamic diameters of 3 µm and aspect ratios of 10:1 deposit in the alveolar region of rats, while more than 25% of these same fibers are predicted to deposit in the alveolar region of humans. Such predictions, this panelist noted, raise questions about whether rats are good models for humans in terms of fiber deposition in the lung.
The second charge question asked the panelists: "What is known about clearance/biopersistence
of less-than-5-µm fibers in the lung?" The panelists identified several
mechanisms by which asbestos and SVF are removed from lung tissue. As Section
2.1 explains, fibers depositing on the conductive airways are cleared, typically
within 1 day, by mucociliary transport; this clearance mechanism is not discussed
further here. The panelists' comments focused primarily on phagocytosis and
dissolution, but panelists considered several additional factors when discussing
lung clearance. All of the panelists' comments are summarized below; Section
2.3 addresses clearance of fibers by migration to other tissues.
Phagocytosis. Reviewing general lung clearance mechanisms, one panelist indicated that alveolar macrophages engulf and can eventually remove foreign materials (e.g., fibers, particles, bacteria) that reach the alveoli. Phagocytized material can then move to the ciliated airways, which would eventually clear the material up to the throat, or they can move into the pleura, lymphatics, or other tissues (see Section 2.3). Typical human macrophages have dimensions between 14 and 21 µm2. Consequently, alveolar macrophages can fully engulf fibers less than 5 µm long and remove them from the alveoli, but they are incapable of fully engulfing longer fibers. The extent of phagocytosis, therefore, clearly depends on fiber length, and may also depend on additional factors, such as surface properties of the inhaled fibers.
The panelists noted that removal of asbestos and SVF from the alveoli by phagocytosis generally takes much longer than removal of these materials from the conductive airways by mucociliary transportan observation that is supported by findings of lung clearance studies in rats (Coin et al. 1994). Specifically, the study reported how the half-life for lung clearance in rats varied with the length of chrysotile asbestos fibers. For fibers approximately 20 µm long, the estimated half-life for clearance (by all mechanisms combined) was 100 days3. One panelist also presented data on time frames for lung clearance of fibers in humans, noting that the estimated half-life for alveolar macrophage clearance was estimated to be between 400 and 700 days; the panelist noted that these estimates apply to poorly soluble spherical particles of low cytotoxicity and to short fibers which can be engulfed by alveolar macrophages. He added that long fibers that cannot be phagocytized and that do not dissolve or break will not be cleared from the lung.
As one exception to the previous observations, one panelist noted that phagocytosis is not an effective clearance mechanism in "overload" conditions, or when high exposure doses overwhelm the lung's clearance mechanisms. The panelists questioned whether the environmental exposures that ATSDR typically evaluates would ever cause overload conditions, though they noted that overload conditions may be observed in some occupational settings or in unexpected accidental or emergency situations.
Dissolution. Asbestos and SVF not only can be physically removed from the lung via phagocytosis, but can be chemically removed, or at least altered, by dissolution. A panelist indicated that the extent to which dissolution occurs depends largely on the fiber composition and the pH of the medium in which the fiber is located, and does not appear to depend on fiber length. Dissolution behavior can change when fibers are engulfed by macrophages, because pH varies considerably between the phagolysomes in the alveolar macrophages (pH = 4.5-5.0) and the extracellular fluid (pH = 7.4). Researchers already have documented the relative solubility of different fiber types (see Dr. Lockey's premeeting comments in Appendix B), which can be useful in characterizing the relative biopersistence of different fiber types.
Influence of co-exposure to other contaminants. A panelist reviewed results from a mixed-dust exposure study in rats (Davis et al. 1991) to illustrate how co-exposures to other contaminants affects fiber retention in the lung. In the study, groups of rats received different combinations of exposures: chrysotile asbestos and titanium dioxide, chrysotile asbestos and quartz, amosite asbestos and titanium dioxide, and amosite asbestos and quartz. Exposure concentrations for the chrysotile asbestos, amosite asbestos, and titanium dioxide were all 10 mg/m 3 , while the exposure concentration for quartz was 2 mg/m 3 . The animals were dosed for 1 year and lung tissues were analyzed for fiber retention after 2 years. The study found that co-exposure with titanium dioxide and quartz had no effect on lung retention of amosite fibers. For chrysotile fibers, on the other hand, co-exposure with titanium dioxide increased lung retention of the fibers (as compared to exposure to chrysotile alone) and co-exposure with quartz decreased lung retention of fibers. This panelist indicated that this study suggests that non-fibrous particles could affect fiber retention characteristics, though he acknowledged that the exposure concentrations used in the study are not relevant to typical environmental exposures.
Influence of physical structure: amorphous versus crystalline material. The panelists briefly discussed how the physical structure of fibers (amorphous or crystalline) affects biopersistence and toxicity. One panelist noted that a laboratory animal study examined this issue by comparing lung samples from rats exposed for 3 months to amorphous silica to samples from rats exposed for 3 months to crystalline silica (Johnston et al. 2000). The study found that significant amounts of crystalline silica remained in the rat lungs 3 months after exposure ceased, while the lung-retained amorphous silica was near background levels. The panelist indicated that this trend suggests that the amorphous silica is more soluble than crystalline silica in the lung.
Populations that may have impaired capacity to clear fibers in the lung. One panelist identified populations that may be susceptible to fiber-related health effects due to impaired capacity to clear fibers deposited in the lung. These populations included people with medical conditions (e.g., primary ciliary disorders, cystic fibrosis, asthma) that affect lung clearance mechanisms. Further, smokers with damaged cilia along the conductive airways may have impaired ability to clear fibers from the lung. Finally, some common pharmaceuticals are known to slow mucociliary transport (e.g., atropine), while others can enhance this transport (e.g., sympathomimetics).
Relevance of sputum samples. When discussing lung clearance, the panelists discussed the utility of analyzing sputum samples to characterize the distribution of retained fibers. One panelist explained that, in at least one study, concentrations of asbestos in sputum, when compared to cumulative exposure estimates, were more predictive of radiological changes in the lungs of workers at vermiculite mines and mills (Sebastien et al. 1988). Though these and other findings suggest that sputum samples can provide useful insight into asbestos exposures, the panelists indicated that implementing a sputum sample study has a potential drawback. While smokers can produce voluntary sputum samples relatively easily, non-smokers often cannot. Induced sputum samples can be collected from non-smokers to characterize past exposure, and bronchoalveolar lavage has also been used for this purpose. Both of these sampling techniques are invasive and require informed consent, and have a consistently better yield than simple sputum collection. More than 50 such studies conducted in North America, Europe, and Japan have already been published.
The third charge question asked the panelists: "What types of migration are expected within the body for less-than-5-µm fibers?" Both in their premeeting comments and during the expert panel review meeting, the panelists offered various perspectives on how fibers of different lengths migrate within the lung and from the lung to other organs. One panelist, for example, indicated that fibers with diameters less than 0.5 µm can penetrate through lung epithelia and be transported through lymph channels to lymph nodes, blood, and distant organs. However, most of the discussion focused on the extent to which small fibers translocate into the pleura. Three reviewers' perspectives on this matter follow:
First, one panelist indicated that several researchers have attempted to characterize the distribution of asbestos fibers in samples of human pleura. Although it has been reported that only short chrysotile fibers (average length <0.2 µm) translocate to the pleura, this panelist found these studies to be of questionable quality because they lacked matched controls or sampled tissue (such as tumors) other than the pleura. This panelist then reviewed two preliminary studies of fiber translocation, one in humans (Boutin et al. 1996) and the other in goats (Dumortier et al. 2002), which were based on more robust methods using controls. He noted that one study found that 22.5% of fibers detected in the pleura were longer than 5 µm and that the pleural samples had far greater amounts of amphibole asbestos fibers than chrysotile asbestos fibers (see Dr. Case's premeeting comments in Appendix B). The studies did not examine how fibers translocate to the pleura, though the findings suggest that lymphatic drainage paths may play an important role4. The authors of these studies hypothesized that the translocated fibers might contribute to formation of pleural plaques and mesothelioma.
Second, another panelist summarized the findings from a study of rats exposed via inhalation to kaolin-based refractory ceramic fibers with geometric mean length of 4.5 µm (Gelzeichter et al. 1996). The study reported that the fate of the fibers depended on fiber length: fibers in the pleural tissue 32 days5 after exposure had a geometric mean length of 1.5 µm and geometric mean diameter of 0.09 µm, while fibers in the parenchymal tissue were much larger with geometric mean length of 5 µm and geometric mean diameter of 0.3 µm. Thus, the study indicates that very thin fibers smaller than 5 µmfibers that would not be counted by conventional phase contrast microscopy (PCM) asbestos sampling methods-are capable of translocating to the pleural tissue (see Dr. Lockey's premeeting comments in Appendix B).
Third, a panelist reviewed findings of a rat inhalation study that investigated whether co-exposure to non-fibrous particles affects translocation of fibers to the pleura (Davis et al. 1991). The study found more amosite asbestos fibers translocated to the pleura in rats that were co-exposed to non-fibrous particles (quartz or titanium dioxide), as compared to rats that were exposed to amosite asbestos alone. The panelist noted, however, that the exposure doses of titanium dioxide (10 mg/m 3 ) might have overloaded the rat lungs and impaired alveolar macrophage clearance processes. If the observed fiber translocation to the pleura was caused by these overload conditions, the relevance of this study to environmental exposures is questionable.
The panelists noted that the extent to which fibers translocate to the pleura is not fully understood, but is likely an important consideration when evaluating pleural plaques, diffuse pleural thickening, and mesothelioma. For instance, if fibers must actually enter the pleura for these outcomes to occur (a hypothesis that has not been verified), then understanding fiber translocation into the pleura is critical. If, on the other hand, fibers localized toward the lung periphery beneath the pleura can cause disease, perhaps through chemical mediators that cross into the pleural space, then translocation of fibers is less important. Therefore, without a more detailed understanding of the mechanisms of toxicity for pleural reactions and other outcomes, the significance of fiber translocation into the pleura is not fully known. The panelists revisited fiber translocation issues when discussing the role of fiber length, if any, in causing pleural abnormalities.
After summarizing the panelists' responses to the three charge questions, the discussion leaders invited the panelists to provide comments on additional topics relevant to physiological fate of inhaled fibers. The panelists raised the following issues:
Terminology: fibers or particles? One panelist had reservations about calling structures with lengths less than 5 µm fibers. He explained that mineralogists, geologists, and health scientists generally do not consider such structures to be fibers, regardless of the aspect ratio; such structures would instead be considered particles. This panelist noted that regulators have established a precedent for distinguishing between fibers and particles: the Occupational Safety and Health Administration, for instance, regulates structures smaller than 5 µm as particles not otherwise regulated, rather than as fibers. For these and other reasons (see Dr. Case's premeeting comments in Appendix B), this panelist had concerns about the terminology ATSDR used to characterize the structures with dimensions less than 5 µm. As noted previously, this report refers to structures less than 5 µm as fibers, and the concern about using this term has been documented.
Importance of the distribution of fiber lengths. Noting that all mineral fiber exposures always involve inhalation of a wide distribution of fiber sizes, one panelist questioned the utility of focusing exclusively on fibers less than 5 µm in length. To illustrate this concern, he showed a graph depicting the fiber size distribution (in terms of length and diameter) in an ambient air sample collected at Libby. The graph showed that a clear majority of fibers were less than 5 µm, as is often observed in occupational and environmental exposure situations6. The sample also included many fibers approximately 15 µm long, though in considerably smaller amounts than the short fibers. In such cases, the panelist cautioned about focusing exclusively on fibers smaller than 5 µm, even if they account for the overwhelming majority of the dose, because the smaller amount of longer fibers contribute more to overall toxicity.
To illustrate this issue further, the panelist presented data on the distribution of fiber lengths measured in surgical lung tissue samples from six miners and four cement plant workers who were exposed to asbestos fibers (primarily chrysotile) and non-asbestos fibers (Case et al. 2002a). The men were hospitalized with various lung diseases, which were mostly not related to their asbestos exposures. In these individuals, the majority (71%, by fiber count) of lung-retained chrysotile asbestos fibers were shorter than 5 µm, with lesser amounts (25%) of chrysotile asbestos fibers between 5 and 20 µm, and even lesser amounts (4%) of chrysotile asbestos fibers longer than 20 µm (Case et al. 2002a). A similar pattern was observed for the lung-retained non-asbestos fibers, with an even greater number of fibers shorter than 5 µm (85%) and none longer than 20 µm. Based on these results, this panelist reiterated that characterizing how toxicity varies with fiber length is critical, because retained doses can vary considerably between different fiber length intervals. The panelists revisited this topic when discussing whether a critical fiber length exists below which adverse health effects from environmental exposures would be unlikely (see Section 3.4).
Comments on fibers detected in Libby. When evaluating the influence of fiber length on dosimetry, the panelists briefly discussed the significance of ambient air measurements in Libby, and asked Dr. Aubrey Miller (EPA) to summarize relevant data. Referring to trends among ambient air sampling data, Dr. Miller indicated that typically more than 60% of airborne fibers at the site are less than 5 µm long and therefore would not be counted by PCM testing for regulatory purposes. A panelist added that two asbestos amphibole minerals not currently regulated by the Occupational Safety and Health Administration (winchite and richterite) are included among the fibers in these samples. Dr. Miller noted that some Libby residents who were not occupationally exposed to asbestos and who had no household contacts with occupationally exposed individuals have developed pleural abnormalities, which raises questions about which fiber types are contributing to this disease. The panelists discussed this matter further when reviewing the current state of the science on human epidemiologic studies (see Section 3.1).
Dose metric issues. The panelists briefly discussed how the available dose metricsmass, number, and surface area of fiberscorrelate with toxicity. A panelist noted that one study (Timbrell et al. 1988) reported that surface area correlated best with pulmonary fibrosis scores and therefore might be the best dose metric for that endpoint. This panelist said this finding is consistent with toxicologic studies of non-fibrous particles, which also indicate that surface area correlates better with pulmonary fibrosis than do other dose metrics. Another panelist questioned whether surface area of retained fibers is an appropriate dose metric, noting that such a selection implies that short fibers (i.e., fibers less than 5 µm in length), if inhaled in substantial quantities, can be equally toxic as very long fibers. This issue was not resolved, but a panelist noted that surface area of fibers might be more predictive of certain endpoints (e.g., lung fibrosis) while other dose metrics may correlate better with carcinogenic endpoints.
Research needs. While discussing the physiological fate of fibers in the lung, the panelists identified several research needs. One panelist, for example, suggested that a laboratory study comparing dosimetry of fibers less than 5 µm to that of non-fibrous particles less than 5 µm could provide insights into lung deposition and clearance of shorter fibers. Another panelist advocated research that characterizes dosimetry for a series of fiber length intervals, rather than focusing entirely on fibers shorter than a given threshold length (i.e., 5 µm), because people are ultimately exposed to airborne fibers of varying lengths. One panelist suggested that studies consider the relevance of susceptible populations, but other panelists indicated that research on susceptible populations should be conducted after key studies on healthy populations have been completed. The panelists discussed additional research needs later in the meeting (see Section 3.5).
After the panelists finished addressing the first topic area, observers were invited to provide comments. The panelists were not required to respond to the observer comments. However, some comments led to further discussion among the panelists, as documented here. The observer comments are summarized in the order they were presented:
Comment 1: David Bernstein, consultant in toxicology
Dr. Bernstein presented findings from a chronic inhalation study that investigated the influence of fiber length and biopersistence on toxicity in rats. The study was conducted for the European Commission, but findings from the study have not been reported in the peer-reviewed literature and a written summary of the study was not provided to the expert panelists. Dr. Bernstein indicated that this study found that long fibers were more biopersistent than short fibers. He further noted that exposure to fibers up to 20 µm long were found to be uncorrelated with toxic response, and only those fibers longer than 20 µm were correlated with toxicity. These findings were reportedly derived by comparing a toxic endpoint at 24 months following exposure to the distribution of fiber lengths retained in the rats' lungs. The toxic endpoint considered was collagen deposition at bronchoalveolar junctionsa precursor to pulmonary fibrosis. Dr. Bernstein claimed that the panelists can draw from this study's findings to make definitive statements on the toxicity of fibers shorter than 5 µm.
Panelists' Discussion: When discussing this study, one panelist asked if preferential deposition of long fibers is expected to occur at the bronchial-alveolar junctions, and Dr. Bernstein said yes. This panelist noted that the apparent correlation between fiber size and toxicity might simply result from studying an endpoint where short fibers do not preferentially deposit. Another panelist encouraged Dr. Bernstein and his colleagues to publish these results.
Dr. Bernstein also presented data from an animal study on biopersistence of chrysotile fibers mined in Brazil. He explained that chrysotile fibers have a somewhat unique molecular structure, because more magnesium atoms are in the fiber surface; in amphibole fibers, on the other hand, these atoms are more concentrated internal to the fiber, away from the surface. Due to this unique structure, Dr. Bernstein argued, the chrysotile fibers are more readily dissolved in the lung. He reported that long chrysotile fibers (>20 µm) have a biopersistence half-life of only 1.3 days, while amphibole amosite fibers of similar length have a half-life of 466 days. He also showed a series of images depicting the fate of different length fibers in the lung as a function of days following exposure. Dr. Bernstein did not provide a reference for the data he presented.
Panelists' Discussion: One panelist took exception to these studies, noting that his colleagues have published a study (Finkelstein and Dufrense 1999) indicating that chrysotile fibers longer than 10 µm have an estimated half-life of 8 years in the lungs of Canadian miners. Further, he noted that a study of South Carolinian textile workers exposed to chrysotile fibers (Case et al. 2000) also supports a chrysotile half-life much longer than 1.3 days. That study found that the lung content of chrysotile fibers longer than 18 µm increased proportionally with the workers' cumulative exposure, suggesting that these longer fibers are more persistent in the lungs of occupationally exposed individuals than Dr. Bernstein's data imply.
Comment 2: Jay Turim, Sciences International, Inc.
Mr. Turim encouraged the panelists to consider the findings of two studies. First, he referred the panelists to a publication (Berman et al. 1995) that re-evaluated data from previous laboratory animal experiments in rats. This study reported that 99.7% of the potency for mesothelioma was due to asbestos fibers longer than 40 µm, with only 0.3% of the potency attributed to fibers shorter than 40 µm. Mr. Turim suggested that the panelists consider these findings when commenting on the carcinogenicity of short fibers.
Second, Mr. Turim reviewed a recent study (Brown et al. 2000) in which two groups of rats inhaled formulations of different refractory ceramic fibers (RCF1 and RCF1a). The fiber formulations were reported as having approximately the same number of long fibers, but the RCF1 formulation contains much more non-fibrous particles than does the RCF1a formulation. In the study, the rats were exposed for 3 weeks (6 hours per day, 5 days per week), and were followed up for 1 year after exposure ceased. Mr. Turim noted that the lung retention of long fibers did not differ between the two exposure groups, even though the study authors reported that macrophage clearance processes were severely impaired in the rats exposed to RCF1, due to lung overload conditions. Mr. Turim also indicated that the study provides evidence that RCF (and SVFs, in general) behave differently from asbestos fibers in the lung, because the short RCF fibers were largely removed despite the impaired macrophage activity. Finally, because the study found more persistent inflammatory response, as gauged by bronchoalveolar lavage (BAL) analysis, in the rats dosed with the RCF1 mixture, Mr. Turim argued that the study shows that the presence of non-fibrous particles must be considered when evaluating the toxicity of SVFs.
Panelists' Discussion: One panelist addressed this comment, noting that some aspects of the RCF study were not entirely clear to him. For instance, he did not think the publication adequately explained how lung clearance of short fibers could be similarly effective in the two groups, when macrophage activity was severely impaired only in the rats dosed with RCF1a. Further, he noted that the differences in toxicity between RCF1 and RCF1a were actually relatively minor, based on his interpretation of the BAL data and the histopathology results. Moreover, the panelist indicated that a follow-up study by the same group has found the non-fibrous FCF particles to be of high toxicity (Bellmann et al. 2002; Brown et al. 2002).
Comment 3: Jenna Orkin, 911 Environmental Action Concern
Ms. Orkin asked the panelists to comment on environmental contamination resulting from the WTC collapse, which blew contamination downwind toward downtown Brooklyn, where she lives. Concerned about ongoing exposure to WTC dust, Ms. Orkin indicated that she recently had a carpet sample from beneath a window in her house analyzed for fiber contamination using ultrasonication. She indicated that this analytical technique can detect about 100 times more asbestos fibers than can be found by ASTM MicroVac methods. Ms. Orkin noted that experts have reported that, for ASTM MicroVac samples, 1,000 structures per square centimeter is considered typical for rural homes and 10,000 structures per square centimeter typical for urban homes. However, she said that experts will not specify a safe level of structures measured by ultrasonication.
Ms. Orkin indicated that EMSL Analytical analyzed the carpet sample from her home and found "80,000 structures per square centimeter of asbestos." Seven chrysotile fibers were in the sample, including five long fibers. She indicated that the ultrasonication instrumentation eventually clogged, which she was told might mean that the contamination levels in the sample could not be measured because they were higher than the measurement sensitivity. Ms. Orkin asked if the panelists would comment on the data she presented, such as the exposure levels she and her family members might have experienced.
Panelists' Discussion: Three panelists and an EPA observer responded to the comment. One panelist noted that regulatory agencies have not established "safe limits" for measurements of asbestos fibers on fabrics. This panelist acknowledged that he was unfamiliar with the measurement method identified in the comment, but he did question why any sampling or analytical instrument would clog when analyzing a sample with only seven chrysotile fibers. Another panelist said the key issue for this scenario is characterizing the inhalation exposure, but he noted that no one has established how to estimate airborne exposure levels from asbestos levels in isolated carpet samples. Finally, noting that amphibole minerals make up 7% of the Earth's crust, a third reviewer suggested comparing the sampling results from the Brooklyn residence to measurements using identical methods in other locations that were not impacted by WTC dust.
Dr. Miller (EPA) indicated that EPA struggles with issues like those raised in the comment at many sites: What levels can be considered safe in homes? What fibers should one count when establishing these levels? When should regulatory agencies recommend abatement? He acknowledged that these decisions are beyond EPA's current regulatory guidelines.
Comment 4: Bertram Price, Price Associates, Inc.
Dr. Price's comment addressed asbestosis in Libby, Montanaa topic the panelists had questions about during their earlier discussions. Dr. Price indicated that ATSDR's recent study of Libby residents identified 12 cases of asbestosis: 11 among former mine workers, and 1 in a family member of a former mine worker. He said these findings illustrate the impact of dose on asbestosis, and he cautioned against attempting to distinguish environmental exposures from occupational exposures. Commenting on the influence of fiber length, Dr. Price noted that researchers have established a dose-response gradient between exposures to long asbestos fibers and asbestosis, though he acknowledged that the past studies used measurement techniques that did not count fibers shorter than 5 µm.
Panelists' Discussions: No panelists addressed this comment.
Comment 5: Suresh Moolgavkar, University of Washington
Dr. Moolgavkar's comments also addressed asbestos-related disease among residents in Libby, Montana. Dr. Moolgavkar noted that ATSDR has conducted two epidemiologic studies on Libby residentsthe second was necessary after the agency realized that some death certificate data were inadvertently omitted from the initial report. He indicated that the second study reported that lung cancer mortality in Libby was higher than expected when compared to the state of Montana and the United States, while the first study found no excess. Regarding asbestosis, Dr. Moolgavkar summarized the available data on asbestosis cases (see Dr. Price's comment, above), and noted that asbestosis is linked to the most highly exposed individuals, regardless of whether their exposures were environmental or occupational.
Dr. Moolgavkar then commented on results from multiple mortality studies published on occupational cohorts of Libby mine workers (Amandus and Wheeler 1987; McDonald et al. 1986, 2002). He found no indication that asbestos from the Libby mines is more toxic than is predicted from cancer risk calculations using asbestos unit risk data from EPA's Integrated Risk Information System. Dr. Moolgavkar mentioned this to question the suggestion among the panelists that Libby asbestos is more toxic than asbestos from other sites (see Section 3). In fact, Dr. Moolgavkar noted, radiological examinations documented in the previous mortality studies found no evidence (based on prevalence of lung abnormalities) that Libby asbestos poses a greater health risk than asbestos from other sites.
Panelists' Discussions: One panelist indicated that he agreed with the comment, in terms of lung cancer outcomes and lung parenchyma abnormalities, but he noted that mesothelioma cancer risks may in fact be uniquely higher at Libby. Specifically, the risk of developing mesothelioma among asbestos miners in Libby, as gauged by the proportional mortality ratio (PMR), is greater than that experienced by crocidolite asbestos miners in South Africa and Australia (see Section 3.1.1 for a more detailed summary of this argument).
Dr. Moolgavkar questioned this response, arguing that the PMR is not a good metric to use. He indicated that one would expect to see an elevated PMR if the Libby cohort had a strong "healthy worker effect."
Panelists' Discussions: The panelist who addressed this issue agreed with this response, but noted that there is no evidence of a "healthy worker effect" among Libby miners, as demonstrated by the large number of accidental deaths in the cohort. This panelist defended use of the PMR for mesothelioma because it is a rare disease, and use of other cancer risk metrics (e.g., the standardized mortality ratio) might not be appropriate.
1Two panelists had different opinions on the particle sizes that should be cited in this sentence. One panelist indicated at the meeting that particles with aerodynamic diameters less than roughly 2 µm would be expected to be carried by convective forces further into the lung. Another panelist, when reviewing a draft of this report, recommended that the size cut-off for this sentence be 0.8 µm.
2Noting that rat alveolar macrophages have dimensions roughly between 10.5 and 13 µm, a panelist indicated that phagocytosis in rats is less effective than in humans at clearing fibers between 13 and 20 µm.
3This half-life estimate likely understates the clearance half-life for amphibole fibers of the same length, one panelist noted, because more recent studies have shown that chrysotile fibers are cleared more readily from the lung than are amosite fibers of the same dimension.
4A panelist also noted that lymphatic transport has been demonstrated to occur in laboratory studies of dogs that were dosed with amosite asbestos by intrabronchial instillation (Oberdörster et al. 1988). Analyses of post-nodal lymph collected from the right lymph duct found fibers only of shorter dimensions: the maximum length of fiber detected was 9 µm, and the maximum diameter was 0.5 µm.
5When reviewing a draft of this report, one panelist noted that 32 days is a relatively short period of time to examine translocation of fibers into the pleura. He indicated that it may take longer for long fibers to reach the pleura, especially if direct penetration is required for the long fibers to enter the pleura (as compared to lymphatic transport for shorter fibers).
6During this discussion, one panelist cautioned about distinguishing environmental exposures from occupational exposures and instead encouraged scientists to focus on the exposure dose, regardless of whether it was experienced in an occupational or environmental setting. To illustrate this concern, he noted that some “environmental exposures,” such as those experienced by Libby residents, might exceed “occupational” exposures in well-regulated work places.