UICC World Cancer Congress 2006

Methods: Data were obtained from the SEER program. Survival models tested included semi parametric and non parametric methods. The hazard function was also incorporated. Density frequency plots of age at diagnosis were used to determine if patient populations were homogeneous.

Results: The Top-down system is based on the availability of large cancer patient datasets, eg SEER. The top-down approach starts with survival and evaluates the association of all potential prognostic factors individually and in various combinations with outcome. These combinations can create an expanded staging system. This system takes advantage of all prognostic factors and clinical variables collected, eg co-morbidity. Prognostic factors are evaluated not only by the relation to survival but also by the hazard function. The hazard function can for instance be used to identify time limited prognostic factors. As a paradigmatic change, treatment is considered as a prognostic factor, which may influence clinical trial designs. The single system would apply to all oncological specialties, since factors used by each specialty are available. The system will be demonstrated for lung and breast cancers.