Project 172738

The ability to assess an individual's risk of lung cancer can have important implications for public health, clinical medicine and for individual decision-making regarding lifestyle changes. We will develop and validate a series of inter-related statistical models that will enable accurate assessment of lung cancer risk on an individual and population level. We will use the data from two case-control studies that have been carried out in Montreal by Dr Jack Siemiatycki. These studies were characterized by development of a unique methodology for assessment of various exposures based on detailed interviews with over 4000 lung cancer cases and age- and sex-matched population-based controls. Base on these data, advanced statistical methodology for studying the roles of smoking history and occupational and chemical factors has been developed by our team. Based on the above-mentioned case-control data, we will develop multivariable statistical models. The set of candidate risk indicators (selected from the study database, comprising hundreds of variables) will include those representing tobacco smoking history (both first-hand and environmental), age, sex, occupational exposures (in particular, selected dusts and fumes), history of hay fever and respiratory illnesses (emphysema, chronic bronchitis), and some others. The models derived from this study can be applied to the scientific practice of lung cancer risk assessment, diagnosis, and prevention. At a community level they can be used to predict the numbers of cases of cancer that will arise, and thus be used for health services planning. For clinical purposes they can be used to triage people into screening programs. For the subjects themselves, these models can also be used to motivate lifestyle changes or at least to inform them of the potential consequences of lifestyle changes.