Predictors of outcome after surgery for gastric cancer in a Western cohort

Background: Gastric cancer (GC) is a common cause of cancer mortality. There are well-documented prognostic factors for GC but these have not been rigorously examined in an Australian context. This study examines the clinical, surgical and histopathological variables associated with survival in a GC cohort from a predominantly Caucasian-based population. Methods: A multi-centre cohort of patients undergoing curative resection for GC enrolled in an ongoing tissue bank study from 1999 to 2009 was retrospectively analysed. Prospectively collected demographic, surgical and pathological variables were available for this cohort. The primary endpoints investigated were cancer-specific survival and recurrence-free survival using multivariate Cox proportional hazard modelling. Results: Five-year cancer-specific survival was 45.9%, 5-year relapse-free survival was 44.7% and 30-day mortality was 2.2%. Variables showing significance on multivariate analysis for cancer-specific and relapse-free survival were AJCC stage, Lauren classification and age at surgery. Conclusion: This study demonstrates that the prognostic variables for a predominantly Caucasian GC population are congruent with published prognostic features. These findings emphasize the importance of the pathological review in allocating prognosis in GC.