Models based on the multistage theory of
carcinogenesis predict that the rate of
mesothelioma increases monotonically as a function of time since first exposure (TSFE) to
asbestos. Predictions of long-term mortality (TSFE >or= 40 years) are, however, still untested, because of the limited follow-up of most epidemiological studies. Some authors have suggested that the increase in
mesothelioma rate with TSFE might be attenuated by clearance of
asbestos from the lungs. We estimated mortality time trends from pleural and peritoneal
cancer in a cohort of 3,443
asbestos-cement workers, followed for more than 50 years. The functional relation between
mesothelioma rate and TSFE was evaluated with various regression models. The role of
asbestos clearance was explored using the traditional
mesothelioma multistage model, generalized to include a term representing elimination over time. We observed 139 deaths from pleural and 56 from peritoneal
cancer during the period 1950-2003. The rate of pleural
cancer increased during the first 40 years of TSFE and reached a plateau thereafter. In contrast, the rate of peritoneal
cancer increased monotonically with TSFE. The model allowing for
asbestos elimination fitted the data better than the traditional model for pleural (p = 0.02) but not for peritoneal
cancer (p = 0.22). The risk for pleural
cancer, rather than showing an indefinite increase, might reach a plateau when a sufficiently long time has elapsed since exposure. The different trends for pleural and peritoneal
cancer might be related to clearance of the
asbestos from the workers' lungs.