Grant: Operating Grant: Data Analysis Using Existing Databases and Cohorts

Diagnosis of esophageal and gastroesophageal junction cancer (EC) is rising in Canada. Accurate estimates of survival help patients and their families plan for the future. Currently, survival is predicted by looking at cancer size, lymph node involvement, and further spread. Many other factors influence survival, such as age, sex, and response to treatment; however, considering all of these factors together is challenging. Predictive models which include patient, cancer, and treatment information can estimate survival in a more precise and individualized way. International guidelines suggest the use of such tools in cancer risk prediction. We reviewed the literature for existing survival prediction tools in EC and found that none of the existing tools are appropriate or applicable for contemporary survival prediction in North America or Europe. We will develop and validate a post-operative survival prediction tool for EC patients. We will combine patient characteristics, cancer pathology, and treatment data with patient-reported symptom data, which has been shown to influence survival, but not been included in existing models. Following an operation, such a tool can help educate patients based on their heath and response to treatment. This tool can help select patients with poor estimated survival for earlier supportive therapy. The tool may also help find patients who may benefit from clinical trials of further chemotherapy to improve survival. We have partnered with www.projectbiglife.ca to make this tool available on the internet to help patients and providers make appropriate treatment decisions. A comprehensive survival prediction tool for resected EC generalizable to North American or European patients is lacking. We will develop and validate a more personalized, clinically relevant model which can be applied post-operatively to help patients and physicians assess prognosis more accurately and make treatment decisions.