Conditional survival probability of patients with pancreatic cancer.

Abstract

e14693 Background: Conditional survival gives more accurate estimate of survival probability for patients who have survived one or more years since initial diagnosis. The aim of this study was to analyze conditional survival probability for pancreatic cancer patients and impact of gender, age and extent of disease on conditional survival. Methods: Using the Surveillance, Epidemiology, and End Results database we analyzed 57,409 patients with pancreatic cancer diagnosed between 1990 and 2008. SEER*Stat: Version7.0.5 software was used to calculate conditional survival, defined as the calculated probability of survival having already survived a specified number of years from diagnosis, with sub-stratification by age(less than 65 yrs and above/equal 65yrs ), gender and stage summary (localized, regional and distant). Results: Conditional survival probability for patients with pancreatic cancer increased from 8.3% to 47.1% having survived one year after diagnosis. Conditional survival per gender increased from 8.5% to 49.8% in women and for men from 8.1% to 44.3%. Conditional survival probability per summary stage increased most for patients with localized disease (from 25.5% to 70.2%) as opposed to distant disease (3.2% to 37.0%). Age may have a slight impact on conditional survival probability, with increase from 12.5% to 50.7% in patient younger than 65 years, as opposed to from 5.7% to 42.3% in age group more than or equal to 65yrs. Conclusions: The expected 5-year conditional survival increases for pancreatic cancer patients who survive a number of years post diagnosis. The increase in conditional survival probability may be to a lesser extent than other cancers, but provides more useful longer term prognostic information.