Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
The paper by Gilham et al1 is important for a number of methodological and substantive reasons, and contributes novel information on the body of epidemiology evidence on the association between asbestos exposure and subsequent risk of pleural mesothelioma.
The use of lung fibre burden as a method to estimate past asbestos exposure has been advocated for a long time,2 and these authors have, for the first time, applied this approach to their large-scale population study, which has already provided important results on the association between asbestos and mesothelioma, based on conventional epidemiological indicators of exposure, that is, occupational categories and time-related factors.3 The method is not free from possible bias, in particular resulting from the difficulty in obtaining lung tissue samples from unbiased groups of mesothelioma cases and controls. In the study by Gilham et al, samples were analysed for 134 cases, which represent 21.5% of those included in the original case–control study,3 which in turn may be a somewhat biased sample of all mesothelioma cases occurring in Great Britain during the study period. The authors used resected lung cancer cases as controls for the lung burden analysis. Two important sources of bias can result from this choice: resected cases of lung cancer represent a subgroup of patients with less advanced/aggressive disease, and (as discussed by the authors) given the association between asbestos exposure and lung cancer, the lung fibre burden of such a comparison group is expected to be higher than that of the population at large. In addition, the sample retrieval rate in this group of controls was also suboptimal (262/420, or 62.4%). It would not be possible to collect an unbiased population-based series of lung samples; however, given the ample opportunity for selection bias, a quantitative assessment of its possible impact on the results4 would have been helpful.
Bringing these limitations into account, the study provides important evidence on several aspects of the role of asbestos as a cause of mesothelioma. The study confirms the predominant role of amphibole exposure in causing mesothelioma5 ,6; although lung fibre burden studies are inadequate to estimate the effect of chrysotile because of the short biopersistence of this type of fibre,2 the results on crocidolite and amosite exposure appear to adequately explain mesothelioma risk. The single most important result of this analysis is the linear dose–response between lung fibre (ie, amphibole) burden and mesothelioma risk, as predicted by the model of asbestos-related mesothelioma carcinogenesis proposed by Peto et al7 in 1982. In addition, this study provides the first estimate of the relative effect of exposure to crocidolite versus amosite, based on lung fibre burden, which is consistent (and refines) previous estimates based on data from environmental sampling.8
One of the most important pieces of evidence brought by the new study is the demonstration that a substantial number of cases of mesothelioma occurred among patients employed in low-risk occupations who were found to have relative high lung fibre burden. In other words, assessment of exposure based on job and industry titles would miss a sizable proportion of (occupation-related) cases of mesothelioma, and lung fibre burden analysis could be added to the standard approaches used for identifying cases deserving compensation for occupation-related mesothelioma. Alternatively, environmental exposures may have had a relevant contributory impact on fibre burden of selected populations in the past.
Gilham et al also discuss the possible implications of lung fibre burden in young people for future mesothelioma rates. This is important on a public health level, but—as in any prediction—it has to be considered with due caution.
The study by Gilham et al sets a new standard for epidemiological research on asbestos and mesothelioma, and in general stresses the importance of collecting tissue samples (both neoplastic and normal) in cancer epidemiology studies. In the case of asbestos-related mesothelioma, such samples would be uniquely important for exposure assessment, as shown in the study by Gilham et al. Tissue samples will also be important for outcome assessment (a relatively large proportion of cases of mesothelioma cannot be confirmed when stringent diagnostic criteria are applied9) and for mechanistic studies, given the recent increase in the understanding of genetic pathways leading to the disease.10
Competing interests PB and CLV acted as expert witnesses in asbestos-related litigations.
Provenance and peer review Commissioned; internally peer reviewed.