Prognostic factors in metastatic gastric cancer: results of a population-based, retrospective cohort study in Ontario.

Abstract

Stage IV gastric cancer is lethal, and little population-based research on prognostic factors has been performed in low-incidence countries. Therefore, we investigated the consistency of the associations of patient, disease and healthcare system factors identified in previous population-based research to understand their generalizability to other low-incidence populations. A population-based, retrospective cohort study of patients diagnosed with Stage IV gastric cancer in Ontario between 1 April 2005 and 31 March 2008 was performed. Kaplan-Meier methodology and the log-rank test were used for bivariate analysis. Multivariate Cox proportional hazard regression was performed. Hazard ratios (HRs) and 95% confidence intervals (CIs) are presented. On multivariate analysis, patient, disease and healthcare system factors were independent predictors of survival. Increasing age per 10 years (HR 1.07; 95% CI 1.02-1.10), a tumor located in the gastroesophageal junction (HR 1.09; 95% CI 0.94-1.27) or middle of the stomach (HR 1.14; 95% CI 0.97-1.35), presence of carcinomatosis (HR 1.61; 95% CI 1.42-1.83) and a larger burden of metastatic disease (2-3 sites of metastatic disease: HR 1.17; 95% CI 1.03-1.32; ≥ 4 sites: HR 1.69; 95% CI 1.30-2.20) were associated with worse prognosis. Female gender, receipt of surgery, chemotherapy and radiotherapy and treatment from a high-volume, gastric cancer specialist were all associated with significantly better prognosis. In addition, there was evidence of significant geographic variation in survival. This study provides supporting evidence for patient, disease and healthcare system prognostic factors in metastatic gastric cancer. Future work investigating the role of emerging molecular and biologic information will need to take these established prognostic factors into consideration.