Egilman’s assessment regarding exposures of auto mechanics to amphiboles is not accurate

View correction statement:

Egilman’s assessment regarding exposures of auto mechanics to amphiboles is correct

Report of a recent “brake” through in the fiber burden-mesothelioma dialogue

This article refers to:

Egilman’s assessment regarding exposures of auto mechanics to amphiboles is correct

Dear Editor:

I would like to address, point by point, the underpinnings and shortcomings of Dr. David Egilman’s views in his recent letter regarding the Marsh, 2011 paper ([“Report of a recent ‘brake’ through in the fiber burden-mesothelioma dialogue” (Egilman, 2012)]).

a).	Dr. Egilman claims that “commercial amphiboles were used to manufacture many friction-related components” (Egilman, 2012, p. 136).  Asbestos is divided into two categories: the serpentine asbestos (e.g., chrysotile) and the amphiboles (there are five types). The commercial amphiboles which were sometimes used to make various products are amosite and crocidolite.
	Automotive friction products (e.g., brakes) manufactured in the United States contained approximately 20–60% chrysotile asbestos beginning in the early 1900s (Lynch, 1968; Paustenbach et al., 2003). There is no data or chemical testing showing that amphibole asbestos was ever used in the manufacture of brakes. In fact, virtually all of the scientific data demonstrate that chrysotile was the mineral form universally used for this purpose in the United States and Europe (Rosato, 1959; Lynch, 1968; Anderson et al., 1973; Langer and McCaughey, 1982; Weir et al., 2001; Blake et al., 2003; Madl et al., 2008; Sheehy et al., 1989; Jacko et al., 1973).  The rigid structure of amphibole fibers did not make them a desirable material for use in brakes, principally because the presence of amphiboles in these products could lead to “scoring” of the metal rotors or drums (Rosato, 1959; Sheehy et al., 1989). Amphiboles also do not blend as well as chrysotile when mixed with resins. I have also been told by other experts that at least one auto maker, and I assume others, requested the supplier of Canadian chrysotile to find veins that had as little tremolite as possible so as to minimize wear of the drum. Tremolite can be found naturally in two forms: asbestiform and non-asbestiform tremolite. Non-asbestiform lacks the toxicity of asbestiform tremolite. It is true that the Canadian mines from which the chrysotile used in friction products was obtained may have contained trace amounts of tremolite asbestos (an amphibole); however, samples collected from various Canadian chrysotile deposits used to make brakes found asbestiform tremolite present in concentrations of less than 1% (Finley et al., 2011), and often much less. Ford used chrysotile asbestos from Canadian mines in their brake linings, which were found to contain at the highest, “six parts in 10 millions, less than one part per million” percent tremolite (Anderson, 2003, p. 90). Furthermore, “[t]he Jeffery mine, Johns-Manville, had essentially no tremolite, it was not found in the tremolite survey, which means that it… weighed less than that. Essentially nothing” (Anderson, 2003, p. 90–91). Trace concentrations of tremolite have reportedly been measured in friction products (Millette, 2008a; b; c; d; e; f). However, there has been doubt cast on the accuracy of some of these tests. In at least one case, when the same brake pad analyzed by Millette (2008e) was reanalyzed using a more precise analysis, the resulting analysis determined that the tremolite was not asbestiform (Van Orden, 2009). With non-asbestiform comprising the vast majority of naturally occurring tremolite in Canadian mines, it is likely that the tremolite fibers found in the other Millette studies were also not asbestiform. In any event, all the available data indicates that tremolite, if measurable, would be found in de minimus quantities in bulk, and virtually none would be found in the brake wear debris. Beyond the fact that tremolite is sometimes just barely measureable, it is important to note that this amphibole has unique characteristics (much different than amosite or crocidolite). Regulatory and public health agencies and organizations, as well as scientific bodies, agree that the biological activity of asbestiform vs. non-asbestiform tremolite is markedly different, as the latter possesses reduced biologic activity (ATS, 1990; ATSDR, 2001; CPSC, 1988; OSHA, 1992; ACGIH, 1955; Addison and McConnell, 2008; Vu, 1993). Addison and McConnell (2008), in their review of carcinogenicity studies of asbestiform vs. non-asbestiform tremolite and other amphiboles, stated that the non-asbestiform materials “are no more hazardous than other silicate minerals widely considered nuisance dusts” (p. 1). Among other characteristics, non-asbestiform tremolite is cleared easily from the lung via macrophage engulfment and/or the mucociliary escalator (Ilgren, 2004). It is noteworthy that virtually all reliable studies employing sensitive analytical techniques have failed to detect amphibole fibers in asbestos-containing brakes and associated wear debris (Weir et al., 2001; Blake et al., 2003; Boelter and Spencer, 2003; RJ Lee Group Inc., 2004; Cohen and Van Orden, 2008; Jiang et al., 2008; Madl et al., 2008). Amphibole fibers, like tremolite, if present, are destroyed by the pressure and heat of braking, just as chrysotile is (Candela et al., 2007). Perhaps even more telling, though, is that TEM analysis fails to identify any measurable airborne concentration of amphibole fibers in air samples down to the 0.001 f/cc concentration (Madl et al., 2008). Thus, it is theoretically possible that mechanics could have been exposed to airborne amphiboles from brake repair, but, in a practical sense, such a claim is meaningless (especially for the biologically active fibers, such as those longer than 5 µm and smaller than 0.1 µm in diameter with an aspect ratio greater than 3).
b)	Dr. Egilman might claim that tremolite (an amphibole) may not be detectable in brake wear debris because of its low concentrations (or absence) or because of thermal degradation, but that doesn’t address the question of arcing brakes, a task that doesn’t involve such high temperatures. Arcing is a synonym for “grinding” brake shoes, whereby the diameter of the braking surface is altered to match the “arc” of the drum. Information is available to answer the question about whether mechanics are exposed to measurable concentrations of tremolite while arcing or grinding. First, as noted above, there are rarely measurable concentrations of tremolite fibers in brakes, and, if they are present, they would be found at ppm concentrations in the pad. Second, when air samples have been collected during arcing, no asbestiform tremolite fibers that qualify as biologically relevant have been measurable, even when TEM was used to analyze samples (Blake et al., 2003). With respect to clutches, Cohen and Van Orden (2008) reported finding only chrysotile asbestos in clutch debris, in amounts ranging from below the limit of detection (<0.01% by weight) to 0.31% by weight.
c)	Dr. Egilman writes that “Victor/Dana used crocidolite in gaskets until at least until [sic] 1963…” (Egilman, 2012, p. 136). I don’t know if Dr. Egilman is talking about gaskets used in some refineries or chemical plants, or autos (as he implies). However, in my review of catalogs and literature concerning gaskets, I have never come across any information indicating that crocidolite was ever used in automotive gaskets. The literature indicates that chrysotile-containing gaskets were always used in automobiles. Crocidolite-containing gaskets were occasionally utilized when acid-resistant materials were required, such as in some chemical and petroleum industries, but never in vehicles (Farrell, 1970). Regardless of fiber type, the asbestos fibers found in historic gasket products were encapsulated as a result of the processes by which they were manufactured. And just to be totally clear here, “encapsulated” does not mean that there cannot be asbestos measurable in air samples when these materials are manipulated. Rather, it means that the material is not friable, and, as such, it would be difficult for this material to produce airborne concentrations that would pose a significant hazard. Many researchers have assessed the exposure potential associated with all facets of gasket removal and replacement, and have consistently demonstrated that airborne fiber concentrations associated with handling these materials are well below contemporaneous and current occupational exposure limits unless atypical removal practices are used (Boelter et al., 2002; Longo et al., 2002; McKinnery and Moore, 1992; Spence and Rocchi, 1996; Liukonen and Weir, 2005; Blake et al., 2006; Mangold et al., 2006; Paustenbach et al., 2006; Madl et al., 2007; Yeung et al., 1999; Cheng and McDermott, 1991; Fowler, 2000). The only exceptions are some studies that used non-standard removal practices (like electric grinding wheels) and these often suffered from analytical shortcomings (Hatfield et al., 2003).
d)	Dr. Egilman states that “….Borg Warner admitted that it used crocidolite in brakes in one set of sworn interrogatories, and denied this in another set of interrogatory answers” (Egilman, 2012, p. 136). Sadly, when lawyers respond to interrogatories, chrysotile can get confused with crocidolite, especially in 1992, when persons (e.g., lawyers) were less sensitive to the importance of carefully documenting the differences. Indeed, such confusion is what apparently occurred in the Borg Warner submission of 1992 (Borg Warner, 1992). Once this error was identified, Borg Warner submitted amended responses to interrogatories (under oath), in which it corrected its prior submission, and noted that “Borg Warner disc brake pads incorporated chrysotile (serpentine) asbestos fibers” (Borg Warner, 1995, p. 4). Thus, Borg Warner has, for the past 17 years, consistently stated in its interrogatories that its brakes only contained chrysotile. Moreover, no automotive brakes, manufactured by any company, have ever been found in the past 50 years of the study to contain crocidolite, which seems to confirm the claims of the Borg Warner lawyers. Further, hundreds of airborne samples have been collected during brake work, and none have detected crocidolite (as discussed previously). In short, I am not aware of any information indicating that crocidolite has been used to manufacture automobile brakes.
e)	Dr. Egilman states that “…Borg Warner used crocidolite in transmission friction plates” (Egilman, 2012, p. 136). On the face of it, this statement is true, since Borg Warner did use crocidolite in some friction plates used in transmissions for a limited period of time (Borg Warner,1992; Borg Warner, 2006). However, it is important to point out that this crocidolite was used only in automatic transmissions, and that the potential for exposure of mechanics to dust from automatic transmission friction plates was very, very low (and probably not measurable on an 8-hour time weighted basis for nearly all, if not all, auto mechanics). Dr. Egilman neglects to mention that automatic transmissions were designed to operate in a wet, oil-based environment, and, as such, the release of respirable fibers during repairs would likely be virtually zero when removing the friction plates. It should also be noted that performing transmission work is actually uncommon for most general mechanics, since such work is typically sent to a shop specializing in transmissions. It should be noted that even mechanics who performed transmission or clutch work in the 1940s to 1980s era did not receive significant exposure (especially on an annual time weighted average basis). One simulation study conducted using modern day techniques, for example, demonstrated that the airborne concentrations of chrysotile during manual clutch work was 0.0016 fibers/cc on an 8-hr TWA basis (Cohen and Van Orden, 2008). If crocidolite had been used in these plates, the airborne concentrations would not be expected to be markedly different during the very brief period of time when intimate handling of these plates occurred.
f)	Dr. Egilman states that “….Ford hygienists were aware of the use of amosite in brakes” (Egilman, 2012, p. 136). This statement is deceptively misleading, and fundamentally an inaccurate characterization of the facts. While it is true that in a symposia proceeding, two Ford professionals, Hickish and Knight, noted that brake linings “may contain 40 to 60% asbestos when manufactured—the asbestos being normally in the chrysotile form, and occasionally in the amosite form” (Hickish,1968, p. 1), the published version of this symposia does not mention amosite (Hickish and Knight, 1970). The authors apparently recognized their initial error while they were preparing the formal manuscript for publication. It is possible that someone could assert that this change was made when the paper was written because the authors wanted to mislead their fellow industrial hygienists and misrepresent the hazard posed by brakes. However, as noted previously, no amosite was ever measured in literally dozens and dozens of bulk samples of brakes collected from vehicles of the 1940s–1980s, or in the perhaps more than two hundred air samples collected in which the presence of amphiboles was investigated (Sheehy et al., 1989; Blake et al., 2003; Madl et al., 2008). Obviously, then, the sentence in the symposia proceedings was not correct.
g)	Dr. Egilman claims that “…Maremont used crocidolite during the Canadian asbestos strike” (Egilman, 2012, p. 136). It is unclear to me if Dr. Egilman is claiming that crocidolite was used to manufacture automotive brakes, but that seems to be the implication. According to the deposition testimony of former Maremont employee Jonathan Sherr, “[c]hrysotile fiber was supplied by Asbestos Corporation, Ltd., Bell Asbestos, Cassiar, Atlas, Johns-Manville, and Vermont [and] [b]lue crocidolite fiber was supplied by Ural Asbestos” (Sherr, 2011, p. 9). According to various studies, however, pure chrysotile is found throughout the Ural Region (Uralbest) in Russia (WTO, 2001; IARC 2012; Virta, 2002). In a study of airborne dust levels from various quarries in the Ural, Russia, deposit, the authors reported that all fibers identified were chrysotile asbestos, and they did not identify any amphibole asbestos, such as tremolite, in any of the fibers (Kashansky et al., 2001).
h)	Dr. Egilman states that “automobile mechanics who worked on brakes would typically work with these products. Thus, it is incorrect to assume that brake workers acquired elevated lung amphibole levels from anywhere but automobiles absent evidence that a patient had exposure to amphiboles at some other occupation” (Egilman, 2012, p. 136). This statement is again misleading and lacking in support. Yes, mechanics who worked on brakes would be expected to handle gaskets, brakes, and transmission plates. However as noted above, all of the available studies indicate that automobile mechanics (the specific group I assume Dr. Egilman means when he uses the term “brake workers”) were not exposed to measurable concentrations of airborne amphiboles as a result of working with brakes (Weir et al., 2001; Blake et al., 2003; NIOSH, 1994). As a result, it would be entirely reasonable for scientists to conclude that amphiboles detected in the lungs of mechanics would, by necessity, be due to exposures that occurred in an occupation other than that of repairing cars or trucks (Langer and McCaughey, 1982). Indeed, this conclusion was also reported several years ago in the Butnor et al. (2002) study of brake mechanics.

Drs. Marsh or Roggli are best equipped to address Dr. Egilman’s criticisms of their laboratory work. Dr. Egilman’s statements regarding the implausibility of their conclusions, though, are completely unfounded.

Declaration of interest

Dr. Paustenbach has served as a consultant in asbestos litigation for a number of firms. He has published 18 papers over the past 10 years on asbestos. He has been frequently retained as an expert by automotive manufacturers, and has testified in cases in which they have been involved. He was not asked by anyone to submit this letter. He and Katherine White received no compensation from anyone for the more than 50 hours of time associated with conducting the associated research and drafting this letter.