Cancer in asbestos-exposed occupational cohorts: a meta-analysis

Our plain-English summary of Cancer in asbestos-exposed occupational cohorts: a meta-analysis, by M. Goodman, R.W. Morgan, R. Ray, et al (Cancer Causes & Control, 1999).

The study

• Authors: M. Goodman, R.W. Morgan, R. Ray, C.D. Malloy, K. Zhao
• Journal: Cancer Causes & Control
• Published: 1999;10(5):453-65
• Type of article: Meta-analysis
• Cited by: 87 articles, as counted by NIH iCite
• Read it: PubMed · Publisher

Abstract

Objective: To examine existing asbestos-exposed occupational cohorts and apply a meta-analytic technique to determine the magnitude of association between exposure and lung cancer and to investigate other cancer sites that may be related to such an exposure.

Methods: We summarized the data from 69 asbestos-exposed occupational cohorts reporting on cancer morbidity and mortality. Data were extracted regarding numbers of deaths for each cancer, numbers of mesotheliomas, occupations and latency for respiratory, gastrointestinal, urinary and lymphohematopoietic cancers. For each cancer, we calculated a meta-SMR and examined heterogeneity of results using a chi-square test and by calculating a Z-statistic for each study. To examine the dose-response effect, we divided the studies into tertiles according to the percentage of mesothelioma deaths that served as a proxy estimation of asbestos exposure.

Results: Lung cancer data demonstrated meta-SMRs of 163 and 148 with and without latency, respectively, with significant heterogeneity of results even after stratification according to occupational groups. Stratification of lung cancer studies according to percentage of mesothelioma deaths showed a dose-response effect. Z-scores ranged from -12.21 to + 29.49. Analysis for laryngeal cancer yielded meta-SMRs of 157 and 133 with and without latency, respectively, demonstrating homogeneous results across studies but accompanied by no evidence of a dose-response effect. Data for gastrointestinal cancers showed no evidence of a significant association and no dose-response effect. Kidney cancer demonstrated statistically non-significant meta-SMRs of 120 (95% CI 88-160) and 111 (95% CI 94-131) with and without latency respectively.

Conclusions: This meta-analysis demonstrates a wide variability of the association between occupational asbestos and lung cancer. There was a suggestion of an association between asbestos and laryngeal carcinoma and no clear association with other cancers.

In plain English

This meta-analysis pooled 69 occupational cohorts of asbestos-exposed workers from 68 articles, mostly from Europe and North America. The share of deaths from mesothelioma served as a stand-in for exposure. The pooled lung cancer standardized mortality ratio (100 means the expected rate) was 163 (95% CI 158 to 169) allowing at least 10 years of latency and 148 (95% CI 144 to 152) without latency; results varied widely. In the latency analysis, lung cancer ratios rose with the share of mesothelioma deaths, from 127 (95% CI 121 to 134) in the lowest third of studies to 285 (95% CI 271 to 299) in the highest third. Laryngeal cancer's pooled ratio was 157 (95% CI 95 to 245) in four cohorts with latency and 133 (95% CI 114 to 155) in 27 studies without, with no dose-response pattern. Gastrointestinal cancers showed no significant increase when latency was considered, and urinary cancers showed no significant increase. The authors concluded that, besides mesothelioma, only lung cancer was unequivocally associated with asbestos. They noted they could not adjust for smoking.

We wrote this plain-English summary from the full article with the help of AI, and a second AI model checked each sentence against the article. It is not the authors' wording; the abstract and the article are their own account.

What the study did and found

• Design: Meta-analysis (pooled analysis) of published occupational cohort studies reporting standardized mortality or incidence ratios
• Who was studied: 69 asbestos-exposed occupational cohorts from 68 articles. The text gives the publication range as 1967 to 1997, but Table 1 also lists a 1963 study (Mancuso). 42 cohorts (61%) were in Europe and 22 (32%) in North America; others were in Australia, Japan, China and South Africa. Cohort sizes ranged from 129 to 135,000 subjects.
• Exposure: Occupational asbestos exposure. The most common occupations were asbestos product manufacturers (22%), cement workers (20%), shipyard workers (12%), asbestos miners and millers (10%) and textile workers (10%). The percentage of all deaths due to mesothelioma was used as a substitute for exposure measurements.
• Main results:
  • Lung cancer, all cohorts: pooled meta-SMR 163 (95% CI 158 to 169; 37 cohorts) with latency of at least 10 years, and 148 (95% CI 144 to 152; 55 studies) without latency. There was significant heterogeneity between studies.
  • Lung cancer by occupation (with latency): highest among asbestos products manufacturers (meta-SMR 192, 95% CI 176 to 209), followed by cement workers (170, 95% CI 156 to 185) and asbestos miners and millers (135, 95% CI 124 to 146).
  • Lung cancer by share of mesothelioma deaths, with latency (Table 3): meta-SMR 127 (95% CI 121 to 134) in studies with 0.6% or less, 138 (95% CI 126 to 151) in the middle group, and 285 (95% CI 271 to 299) in studies with more than 2.4%. Without latency, the values were 118, 153 and 249. Table 4's dose-response analysis used latency data where available and gave 117 (95% CI 112 to 122), 144 (95% CI 135 to 154) and 255 (95% CI 244 to 267). The positive correlation was r = 0.591, p less than 0.001.
  • Laryngeal cancer, with latency (4 cohorts): meta-SMR 157 (95% CI 95 to 245), with homogeneous results. Without latency (27 studies), the meta-SMR was 133 (95% CI 114 to 155). There was no dose-response effect and no correlation with the share of mesothelioma deaths.
  • Gastrointestinal cancers: with latency of at least 10 years, none showed a significant increase compared with the general population. All gastrointestinal cancers combined had meta-SMRs of 103 (95% CI 95 to 111) with latency and 108 (95% CI 103 to 113) without. No dose-response effect was found.
  • Urinary cancers: no statistically significant increase. Kidney cancer had meta-SMRs of 120 (95% CI 88 to 160) with latency and 111 (95% CI 94 to 131) without. Prostate cancer data showed no evidence of increased rates.
  • A large cohort of insulation workers in the US and Canada had the highest positive Z-scores for lung cancer (+27.35 with latency, +29.49 without). Studies of Finnish locomotive drivers and Swedish construction workers had the most extreme negative Z-scores.
• The authors' conclusion: The authors concluded that, besides mesothelioma, lung cancer was the only cancer showing an unequivocal association with asbestos exposure. They wrote that the size of this association is frequently overestimated, and that most cohorts showed less than a two-fold increase in risk. They found a suggestion of an association with laryngeal cancer and no clear association with other cancers.
• Limitations the authors state:
  • Limited data on smoking meant the results for tobacco-related cancers could not be adjusted for smoking.
  • The analysis included both good and poor studies, may be affected by publication bias, and relied on published studies only.
  • Some pleural mesotheliomas may have been misdiagnosed as lung cancers, and some peritoneal mesotheliomas as gastrointestinal cancers, affecting individual study estimates.
  • The authors warned against relying on a single summary risk estimate when results vary significantly between studies.

Topics

The National Library of Medicine indexes this study under subjects our site covers: Lung cancer.

Medical terms in this study, explained in our Lexicon: Meta-Analysis, Lung Cancer, Mesothelioma, Latency, Dose-response relationship, Asbestos Exposure, Laryngeal Cancer, Carcinoma.

Related studies

• Kidney cancer and occupational exposure to asbestos: a meta-analysis of occupational cohort studies — D. Sali, P. Boffetta (2000)
• Occupational exposure to rock wool and glass wool and risk of cancers of the lung and the head and neck: a systematic review and meta-analysis — L. Lipworth, C. La Vecchia, et al. (2009)
• Cancer incidence among Swedish pulp and paper mill workers: a cohort study of sulphate and sulphite mills — E. Andersson, H. Westberg, et al. (2013)
• Cancer incidence in a chlorochemical plant in Isère, France: an occupational cohort study, 1979-2002 — V. Bonneterre, G. Mathern, et al. (2012)
• Cancer incidence within a cohort occupationally exposed to asbestos: a study of dose--response relationships — B. Clin, F. Morlais, et al. (2011)
• Finnish Institute of Occupational Health (FIOH): prevention and detection of asbestos-related diseases, 1987-2005 — M.S. Huuskonen, J. Rantanen (2006)

Browse more research in Cancer Causes & Control, or return to medical research.