O-253 Occupational asbestos exposure and gastrointestinal cancers: systematic review and meta-analyses

Introduction

To summarize the epidemiological evidence on occupational asbestos exposure and the risk of esophageal, stomach and colorectal cancer.

Methods

The search strategy was developed by investigators with occupational hygiene, exposure assessment, cancer epidemiology, and systematic review expertise; and applied to MEDLINE, Web of Science, Embase, CINAHL and Scopus databases with no limits on publication year, country/region or language. Studies that reported relevant effect estimates (e.g., ORs, HRs, SIRs, and SMRs, considered equivalent to RRs) were eligible upon independent review and consensus by two authors. Results are presented in forest plots with corresponding meta-relative risk (mRR) estimates generated from random effects models to account for study heterogeneity.

Results

The systematic review included 193 cohort (80%) and case-control (20%) studies. After the selection of preferred effect estimates as the most informative, 57 studies contributed unique effect estimates to meta-analyses for esophageal cancer; 100 for stomach cancer; and 88 for colorectal cancer. There were elevated mRRs for esophageal [1.17 [95% CI 1.07–1.29]], stomach [1.13 [1.06–1.21]] and colorectal [1.17 [1.09–1.26]] cancers associated with ever versus never occupational asbestos exposure. Unexplained heterogeneity was reduced, and the strength of association increased, in the analyses of studies with better exposure assessment and increased confidence of high asbestos exposures, including among workers in the highest versus lowest exposure-response categories [e.g., mRR=1.31 [1.10–1.57] for stomach cancer]; among workers with a history of significant occupational exposure (e.g., mRR=1.68 [1.19–2.36] for esophageal cancer among insulators/insulating manufacturing workers]; and among workers in cohorts with a two-fold or greater increased risk of asbestos-related lung cancer [e.g., mRR=1.51 [1.37–1.66] for colorectal cancer]. Sensitivity analyses indicate minimal influence from any single study on meta-estimates or from publication bias.

Conclusion

The evidence synthesis supports a causal link between occupational asbestos exposure and esophageal, stomach and colorectal cancer.