Correspondence re: J.L. Pauly et al., glass fiber contamination of cigarette filters: an additional health risk to the smoker? Cancer Epidemiol. Biomark. Prev., 7: 967-979, 1998.

I read with great interest the paper entitled “Glass Fiber Contamination of Cigarette Filters: An Additional Health Risk to the Smoker?” by Pauly et al. (1). Dr. Pauly and his colleagues have drawn conclusions that are simply not justified. Following is a brief analysis of the scientific evidence that makes Reynolds Tobacco confident that the use of continuous glass filaments in Eclipse does not pose a risk to smokers of this cigarette. In Eclipse, smoke is formed by hot air generated by a carbon heat source that passes through a glycerin-rich substrate, volatilizing glycerin, nicotine, other natural tobacco flavors, and added flavorants. The heat source assembly used in Eclipse consists of carbon, glass mat insulator, tobacco sandwich paper, and an insulator paper overwrap. The glass mat insulator functions to insulate the heat source, reduce propensity for the ignition of other materials, and enhance cigarette performance by directing heat flow to the substrate materials of the cigarette. The glass mat insulator is composed of continuous filament glass fiber and a binder. It should be noted that glass filaments are ubiquitous and are commonly observed in ambient air in both occupational and nonoccupational settings. Continuous filament glass is commonly used in consumer products such as drapery fabrics, decorative ornaments, and air filters (2). Survey data demonstrates that consumers are exposed to literally thousands of glass filaments daily as a result of nonoccupational environmental exposure (2, 3). The potential transfer of continuous filament glass to Eclipse mainstream smoke has been evaluated. The results of these studies have been presented at scientific meetings (4). In brief, methodology was developed that allowed the capture of any filaments that might be transferred to Eclipse mainstream smoke. Subsequent to handling procedures that were designed to simulate commercial shipping conditions, cigarettes were machine smoked using exaggerated conditions (10 75-ml puffs, 2-s duration, taken every 35 s). These conditions were used specifically to maximize the probability that any potential transfer of continuous filament glass to Eclipse mainstream smoke would be detected. The results of these studies consistently demonstrate that, relative to background, biologically significant transfer of continuous filament glass to Eclipse mainstream smoke does not occur. The continuous filament glass used in Eclipse was designed to be nonrespirable. The continuous filament glass mat insulator is made from continuous glass filaments having an average diameter of 8.5 mm (range, 5.5–12 mm). These filaments are chopped to a length of 9500 mm and subsequently converted into a mat. The calculated aerodynamic diameter (parallel) of these filaments would be ;53 mm (range, 35–73 mm; Ref. 5). It is generally accepted that structures with aerodynamic diameters of .7 mm are unlikely to reach the pulmonary region of the respiratory tract (6, 7). In addition, there are epidemiological data that specifically focus on the potential impact of occupational exposure to continuous filament glass. In the studies in which the potential impact of occupational exposure to continuous filament glass has been specifically addressed, long-term health consequences have not been associated with occupational exposure to continuous filament glass (8, 9). Furthermore, chronic inhalation studies have been conducted with rat respirable fractions of representative glass fibers (10). In these studies, rats were exposed by nose only to a rat respirable (size-selected) fraction of two representative glass fibers (MMVF 10 and MMVF 11). Although crocidolite asbestos (the positive control) was demonstrably carcinogenic to rat lung, no statistically significant increase in fibrosis or lung tumors was observed subsequent to exposure to either MMVF 10 or MMVF 11 (10). It has been hypothesized that biopersistence plays a significant role in determining the potential bioactivity of a fiber (11). The relationship between in vivo biopersistence and chemical composition has been evaluated (12). The toxicological potential of fibers appears to be inversely correlated with the rate at which specific fibers disappear from the lung (11). For example, asbestos fibers and refractory ceramic fibers, which are relatively durable in the lung, do produce significant increases in lung tumors, mesothielioma, and/or fibrosis in the lung (10). Rock wool (MMVF 21), which is less durable than asbestos fibers or refractory ceramic fibers in the lung, produced fibrosis in rat lung but not lung tumors. Glass fibers such as MMVF 10 and MMVF 11 have been shown to be significantly more susceptible to chemical leaching processes within the lung. It has been hypothesized that this characteristic of the glass fibers accounts for their observed inactivity in animal models. In vivo biopersistence in the lung is frequently assessed by in vitro fiber dissolution studies using simulated biological solutions and flow past techniques (13, 14). Comparative studies have been conducted that include MMVF 10, MMVF 11, MMVF 21, and crocidolite asbestos as controls. Dissolution rates were calculated from weight loss data, assuming that the rate is proportional to the reaction surface area. The results of these studies indicate that the continuous filament glass used by RJR in Eclipse is significantly more soluble than either rock wool or crocidolite.