© 2008 American Thoracic Society doi: 10.1164/rccm.200806-859ED
Airborne Environmental Exposure to Asbestos
Imperial College London In this issue of the Journal (pp. 624–629), Kurumatani and Kumagai (1) present an investigation into cases of mesothelioma in Amagasaki City, Japan, linked to a former asbestos cement pipe factory. The Amagasaki plant was known to have used crocidolite and chrysotile asbestos between 1957 and 1975 and chrysotile into the late 1980s. Because amphibole fibers such as crocidolite are considered more potent carcinogens, the authors reasonably considered the important exposure period as 1957–1975 and further simplified exposure to its spatial components during this time period. Investigation of clusters of mesothelioma is more straightforward in some respects than, for example, clusters of lung cancer or chronic obstructive pulmonary disease in nonsmokers, in which a variety of risk factors are recognized. At its simplest, patients with mesothelioma not exposed at work must have had another source of exposure to asbestos and, as in Amagasaki city, a putative source is often readily identifiable. Spatial epidemiologic studies can be very helpful in situations where an environmental source is suspected (2) to provide corroborative evidence of a causal link and inform appropriate public health measures. Key features of such studies are careful assessment of exposure (information about the type of asbestos in use, the amounts emitted, the exposure pathways), careful case ascertainment conducted independently of exposure assessment, and careful investigation of dose–response function; plausibility of the latter can often be strengthened by using two or more methodologic approaches. A variety of environmental exposure pathways for asbestiform materials have been identified. Mesothelioma cases in families of asbestos workers have long been recognized, as in women washing work clothes of asbestos miners or mill workers (3). Asbestos-containing material from local mines and asbestos mills may be used in construction of pavements, roads, and school playgrounds. In Wittenoom, Australia (4), crocidolite was widely used in this way and the resulting community environmental exposures have been estimated to double the background risk of mesothelioma (5). Mesothelioma cases in rural communities with naturally occurring asbestos may occur where houses are constructed (6) or painted with asbestiform-containing material (7); other housing exposures may occur through use of asbestos for loft insulation. Transport of asbestos materials in nonsealed containers may result in asbestos contamination (8). Multiple exposure pathways may be identified, such as in Libby, Montana, where asbestos-contaminated vermiculite (a silicate mineral) was used in activities ranging from construction to being played with by children (9). More recently, a cancer registry–based spatial study in California was strongly suggestive of windborne exposure, showing a dose–response relationship between mesothelioma risk and residential distance from naturally occurring asbestos (10), although the study did not fully account for occupational exposures or construction usage of asbestos-containing material. The main exposure route from the Amagasaki plant was almost certainly airborne (1). First, there was a plausible emission source: fluffed asbestos entering the atmosphere from pneumatic ducts in site workshops. Second, a detailed questionnaire was used to exclude alternate exposures through occupation, domestic exposure via occupation of household members, or use of asbestos-containing materials in housing. Third, there was no evidence for local use of asbestos in pavements, gardening, or roads (apart from a small stretch of road from the site found to have asbestos residuals), nor was there naturally occurring asbestos in the area. Fourth, estimation of both distance and dispersion of asbestos from the factory site considering meteorologic information (wind direction and air stability) suggested an excess mesothelioma risk that declined with distance. If excess risks had not tailed off in this way, alternative exposure routes would have had to be considered. Retrospective epidemiologic studies of this type are difficult. There were problems estimating the historical population at risk (underrecognition would increase the standardized mortality ratios [SMRs]) and likely underrecognition of cases (which would decrease the SMRs). However, serious bias seems unlikely because population estimates and case identification are unlikely to vary with the exposure measure, essentially a function of distance from plant. Also, the fact that cases were identified through a generous compensation program should not influence the findings; this is unlikely to vary by distance as there was intense local publicity and a relatively small area was involved. So how does this study help in other settings? It substantiates the airborne exposure route as a cause of community asbestos-related disease many years later and, consistent with other studies (4), suggests that risks of environmental exposures are higher in women. The study indicates that windborne asbestos exposure could result in cancer cases at some distance from a site. However, the observed cutoff of excess risk at 2.2 km is not generalizable without quantitative measures of population asbestos exposure; the authors are working on methods to retrospectively estimate these. Risk assessment could be further refined, if information were available, by taking person-years of exposure into consideration, and considering risks from chrysotile exposure including assessment of contamination of chrysotile by amphibole fibers. The groupings of asbestos concentration presented in the study are nonintuitive—quintiles are more commonly used—and alternative groupings would affect appearance of exposure maps and, potentially, the exposure response function. A more sophisticated small area analysis could potentially shed light on the true shape of the exposure–response function. A further consideration is that asbestos exposure will also cause lung cancer and act synergistically with smoking. A U.K. study (11) suggested up to one asbestos-related lung cancer death for every mesothelioma death in males in 1980–2000 using information on occupational and smoking, but this may not be valid for Japan with different smoking trends. Asbestos is still in widespread use in many countries. Although well-managed industry with minimal asbestos release is unlike to result in appreciable community levels of disease (12), any exposure seems unnecessary because adequate substitutes to asbestos exist. Unfortunately, the asbestos industry has moved to less well regulated areas, such as Mexico and India (13, 14), with potential for resultant community as well as occupational exposures. We should also not forget the ghastly legacy of poorly regulated asbestos mining and milling in South Africa where the industry continued until the end of the 20th century (15) and where continuing widespread environmental contamination persists (16). The Amagasaki (1) and California (10) studies suggest that windborne exposures may, sadly, result in far wider geographic impacts than anticipated. FOOTNOTES Conflict of Interest Statement: A.H. has no financial relationship with a commercial entity that has an interest in the subject of this manuscript. REFERENCES
Related articles in AJRCCM:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
