Letter to the editor in response to Finkelstein et al. (2012)

Dear Editor,

We are delighted to have the opportunity to respond to the comments from Dr. Murray Finkelstein regarding our recent publication in Inhalation Toxicology entitled “Asbestos fiber concentrations in the lungs of brake repair workers: commercial amphibole levels are predictive of chrysotile levels” (Marsh et al., 2011). In a most peculiar turn of events, Dr. Finkelstein was permitted to amend his original submission after we had responded and both letters were accepted for publication. For purposes of clarification, we will first respond to the comments in his original letter and then address the statements in his amended letter.

Regarding the alleged data errors, we have re-reviewed all our data and calculations and have verified them to be correct as published. Dr. Finkelstein’s allegations appear to stem from his misunderstanding of the values recorded on the counting sheets versus values reported in Table 1 of our manuscript. Below we address and clarify these areas of misunderstanding.

Regarding Cases 1 and 9, Dr. Finkelstein has omitted the coated fibers (asbestos bodies) in the count sheets which, when included, yield the values published in our article. In the original manuscript by Butnor et al. (2003) which is the source of most of the data for our current study, it is clearly stated in the Materials and Methods section that the calculations are based on total asbestos fiber concentrations, and the footnote to Table 2 clearly indicates that the columns represent the total of coated plus uncoated fibers. Similarly, in our paper, the footnote to Table 2 clearly indicates that the last column represents the total of coated plus uncoated fibers.

Regarding Cases 2, 3, 4, 8 and 9, Dr. Finkelstein has omitted the non-commercial amphibole fibers other than tremolite (actinolite and anthophyllite) in the count sheets, which, when included, yield the published values in our article. In Cases 2 and 4, the non-commercial amphiboles (TAA) identified were anthophyllite, in Case 3 actinolite, in Case 8 actinolite and tremolite, and in Case 9 anthophyllite and tremolite. In the original manuscript by Butnor et al. (2003) the Materials and Methods section clearly states that non-commercial amphiboles included actinolite, anthophyllite and tremolite and the Results section notes that non-commercial amphiboles, principally tremolite, with some actinolite and anthophyllite, predominated. Similarly, in Table 2, it is clearly indicated that the non-commercial amphiboles were combined for purposes of this analysis.

In our paper, this is again mentioned similarly in the Introduction section: “non-commercial asbestos fibers (principally tremolite, with some actinolite and anthophyllite, measured as a marker of chrysotile exposure)” (Marsh et al., 2011). Tremolite and actinolite are part of a single mineral series, with the distinction based on demonstrable iron content in the latter, and anthophyllite being more closely related to talc. All three non-commercial amphiboles have been mentioned as contaminants of chrysotile (Roggli & Coin, 2004; Churg, 1986).

With respect to Case 6, one asbestos body was counted but not analyzed. We were left with the option of ignoring this coated fiber or to assume that it was likely a commercial amphibole (as it is in 96% of coated fibers that we have analyzed) (Roggli & Sharma, 2004). We arbitrarily chose the latter option. Either way, this sample was not considered to be different from our background levels. With respect to Case 9, two samples were analyzed and the findings were nearly identical. The results reported in our paper were the weighted average of the two samples (i.e., the total number of fibers combined and the total weight of tissue combined). This was the only case that had two samples analyzed by electron microscopy.

After the submission of our original response, Dr. Finkelstein noted that he had learned from Dr. Roggli that the fiber counts reported were the sum of uncoated fibers 5 µm or greater in length plus asbestos bodies (there was no direct communication with Dr. Finkelstein regarding this information). He further states that this was nowhere reported in the paper (see the discussion above) and that this is an unusual procedure. He has thus further compounded his error. It is not at all uncommon for researchers to report their findings by electron microscopy as the total fiber count for any particular fiber type or group without distinction as to whether or not the fibers are coated or uncoated (Churg & Vedal, 1994; Gibbs et al., 1991). Indeed, we have done so on many occasions (Roggli & Sanders, 2000; Roggli et al., 2002a,b; Roggli & Vollmer, 2008; Schneider et al., 2010) (including the Butnor paper, 2 which Dr. Finkelstein apparently never read).

Dr. Finkelstein raises another issue in his amended letter regarding case 9, stating that this individual was exposed to friction products that contained crocidolite asbestos. No citation is provided for this statement. The company involved in this matter filed in the late 1980s answers to interrogatories stating that when Canadian mines were on strike, the company purchased ‘Russian blue’ as a substitute (Deposition of Carl Liggett, 2011). Russia is the location of the largest chrysotile mine in the world and was not a source for crocidolite asbestos (Roggli & Coin, 2004; Deposition of Carl Liggett, 2011). Thus, this company’s products never contained crocidolite, as implied by Dr. Finkelstein.

As noted in the Discussion section of our paper, two other laboratories using transmission electron microscopy (TEM) have reported similar findings to ours with respect to mesothelioma patients exposed to brake dust whose lung samples were submitted for analysis. Since the submission of our manuscript, an additional report from Dr. Bruce Case, again using TEM, reported findings similar to those of Butnor et al. (Case, 2011).

Declaration of interest

A conflict of interest form is available online.