Standard lung cancer screening protocols may miss high-risk cases among the millions of workers previously exposed to asbestos, potentially leaving a vulnerable population inadequately monitored despite their elevated cancer risk. The validation of five widely-used risk prediction models in Western Australia's asbestos-exposed population reveals concerning gaps in current screening approaches that could affect occupational health strategies worldwide.
Researchers analyzed 2,126 participants in the Western Australia Asbestos Review Program, where 51 individuals developed lung cancer over the study period. When standard models including PLCO variants, Liverpool Lung Project, and Bach algorithms were applied, all showed modest predictive accuracy with area-under-curve values between 0.602-0.675. Most critically, nearly all models systematically underestimated cancer risk in this asbestos-exposed cohort, with only the Liverpool model overestimating risk. The population included 55% former smokers and 36% never-smokers, with median smoking duration of 24 years.
This validation study highlights a significant blind spot in precision medicine approaches to cancer screening. Current models were primarily developed using general population data, potentially missing the unique risk profile created by asbestos exposure's synergistic interaction with tobacco use. The findings suggest that occupational exposure history may require distinct risk stratification algorithms rather than relying on models designed for broader populations. Given that asbestos exposure affects millions globally through construction, shipbuilding, and manufacturing industries, these results indicate an urgent need for exposure-specific screening protocols. The modest performance metrics suggest that enhanced models incorporating occupational history, exposure duration, and fiber types could substantially improve early detection rates in this high-risk population.