Modeling the Probability of Second Cancer in Controlled Clinical Trials

Due to the advancement of medical technologies and cancer care, the long-term survival of cancer patients has been substantially increased. For some patients, increases in survival have been offset by the long-term late effects of cancer and its treatments. One of the most life-threatening sequelae is the diagnosis of a new malignant cancer. The number of patients with multiple primary cancers is growing with second new cancer now representing approximately 16% of all cancers reported to the SEER Program of National Cancer Institute. Second malignant cancers reflect not only the late effects of the original cancers and therapies but also the influence of shared etiologic factors, genetic susceptibility, environmental exposures, cancer drug exposures, and older age, etc. In this research, we attempt to outline a framework to model the patient-level probability of the occurrence of new cancer malignancies using patient's demographics, medical history and concomitant medications, clinical laboratory tests, drug efficacy, and adverse event data-which are all commonly collected in clinical trials.