Under the auspices of our currently funded CISNET grant we have been developing models based on ideas of multistage carcinogenesis for prediction of lung cancer incidence and mortality rates. In this renewal CISNET application we propose to use these models to predict lung cancer risk in the US under diverse smoking scenarios. We will also develop user-friendly software to implement our models. This software will be made freely available to interested scientists. Our models can explicitly accommodate detailed smoking histories on individuals including age at initiation, number of cigarettes smoked per day, changes in levels of smoking, and age at quitting if an ex-smoker. Moreover the models can be used to predict risks both in individuals and populations. Thus the models can be used to predict both individual and population risks under various intervention scenarios for smoking cessation. Since the models are based on the biological paradigm of initiation, promotion and progression in carcinogenesis, they can be used to generate biological hypotheses regarding the mechanism of tobacco induced lung cancer and to explore the extent to which projected risks depend on specific mechanistic aspects of smoking-induced lung cancer. We propose to explore collaboration with other investigators supported by CISNET, particularly those interested in using our model as the 'natural history'component of screening models.
