Asbestos, Pleural Plaques, and Lung Cancer: Untangling the Relationships

Asbestos exposure remains an important public health and clinical problem; industrial use has been significantly reduced but not eliminated (1). Significant risks of asbestosis, lung cancer, malignant mesothelioma, and other effects continue for many former exposed workers and family members with prior paraoccupational exposure. Pleural plaques, or localized thickening of the parietal pleura, are the most common consequence of asbestos exposure (2). Assessing the relationship between pleural plaques and lung cancer risk is particularly timely now that low-dose computed tomography (LDCT) screening of high-risk tobacco smoking–exposed populations has been demonstrated to reduce mortality. High-resolution computed tomographic screening policy largely focuses on tobacco smokers and inadequately addresses persons with significant risk from asbestos or other occupational carcinogens.

Brims and colleagues (3) in this issue of the Journal (pp. 57–62) provide very important data on the relationship of pleural plaques to lung cancer risk. Asbestos exposure causes lung cancer and produces pleural plaques. The paper by Brims and colleagues addresses the important question, “Among persons with known moderate-heavy asbestos exposure, do those with pleural plaques have increased lung cancer risk relative to similarly exposed persons without plaques?” All participants were known to have significant asbestos exposure. The investigators used Cox regression analyses to assess whether pleural plaques were associated with an elevated hazard ratio (HR) for lung cancer. The analyses were adjusted for asbestos exposure, sex, tobacco smoking, and the presence of asbestosis. Pleural plaque status was determined from the most recent radiographic imaging (either CT or chest radiography) or the most recent imaging at least 1 year before cancer diagnosis.

The authors conclude that plaques per se, when adjusted for the extent of asbestos exposure and other risk factors, do not enhance the risk of lung cancer. This has important implications for patient counseling and selecting participants to optimize the benefit–risk relationship for individuals and programmatic cost-effectiveness of LDCT screening.

This study has unique strengths. The results were consistent in two distinct, well-defined cohorts. Western Australian crocidolite asbestos miners and community members with extensive residential exposure comprise the first cohort. The second cohort is a nationwide collection of workers in occupations well known to have extensive exposure. The Australian surveillance program is particularly effective at accurately assessing each participant’s individual cumulative exposure (4–7). The long latency between initial asbestos exposure and development of malignancy requires long-term studies; the Australian program includes annual follow-up since 1990 and well-standardized radiographic methods. Specificity of plaques for asbestos exposure was increased by limiting the definition to bilateral plaques to reduce the likelihood of observed thickening due to chest trauma or a prior infection.

Several technical limitations do not seriously reduce the impact of the data. Asbestos fiber–type exposures were heterogeneous; the Western Australian cohort was exposed largely to crocidolite, whereas the national worker group had mixed fiber types, including multiple amphiboles and chrysotile types. In the national worker group, the association of cumulative exposure with lung cancer risk did not reach traditional statistical significance (HR, 1.81; 95% confidence interval, 0.94–3.50). The optimal exposure covariate metric may not be log-linear or dichotomous as used in the regression analyses. The potentially complex causal interactions among exposure, pulmonary fibrosis/asbestosis, plaque, cancer, age, and smoking constrain drawing mechanistic conclusions (8), but they have less impact on the practical implications for counseling and screening.

Pairon and colleagues previously reported results for 5,400 participants in a 6-year follow-up study of asbestos-exposed workers in a CT screening program in France (9). In contrast to the study by Brims and colleagues, they found a significantly elevated HR (HR, 2.41; 95% confidence interval, 1.21–4.85) for pleural plaques and lung cancer when adjusting for smoking and asbestos cumulative exposure. The difference might be due to lower overall cumulative exposure in the French study or to not adjusting for asbestosis. Both the Pairon and Brims studies are much more powerful than earlier approaches to this important topic (10).

The strengths of the study by Brims and colleagues limit its policy implications for screening groups with less–well-characterized exposure. Their finding that plaques do not in themselves confer risk of lung cancer depends on accurate and precise estimates of cumulative exposure (4–7). Both the Western Australian crocidolite miners/residents and the cohort of workers in occupations with well-known asbestos exposure had clear a priori indication of significant exposure and reasonably good estimates of cumulative exposure. Therefore, finding bilateral plaques did not add significant new information about each subject’s exposure. However, more general populations have less precise information about asbestos exposure. Many workers may be unaware of prior exposure or may have forgotten owing to the long latency. In a study of male LDCT participants selected for smoking rather than asbestos exposure, most with plaques were unaware of prior exposure (11).

Hence, when exposure misclassification or inaccurate quantification is likely, the presence of bilateral pleural plaques is likely to confer useful exposure information. Plaques are a biomarker of exposure, albeit imperfect; plaques increase the likelihood of sufficient exposure to seriously consider preventive interventions such as LDCT screening or tobacco control program enrollment.

Several reports show that LDCT screening of exposed workers detects lung cancers with yields similar to those of high-risk smokers (12–14). A meta-analysis (12) showed that the baseline prevalence of cancers was similar in asbestos workers and in heavy smokers (about 1%). Many of the malignancies were in an early stage and therefore potentially curable. The studies are relatively small and generally limited to initial tests, and they have not examined impact on population mortality.

LDCT screening of high-risk individuals appears advisable even if not yet empirically proved by large prospective trials. The study by Brims and colleagues (3) shows that exposure rather than plaque determines risk, and therefore persons with known moderate to heavy exposure should not be denied screening if they do not have plaques. Conversely, for a general population in which exposure classification and quantification are more ambiguous, detecting bilateral plaques should prompt the clinician to assess exposure history in detail, consulting industrial hygienists or occupational medicine specialists as appropriate. Although the study by Brims and colleagues focuses on plaques per se, additional analyses may develop integrated individual risk–benefit profiles considering multiple personal factors such as smoking, asbestosis, age, and comorbidities.