The association of adjuvant therapy with survival at the population level following pancreatic adenocarcinoma resection

Background: Using a retrospective observational cohort approach, the overall survival (OS) following curative-intent resection of pancreatic adenocarcinoma (PC) was defined at the population level according to adjuvant treatment, and predictors of OS were identified. Methods: Patients undergoing resection of PC in the province of Ontario between 2005 and 2010 were identified using the provincial cancer registry, and linked to databases that include all treatments received and outcomes experienced in the province. Pathology reports were abstracted for staging and margin status. Patients were identified as having received chemotherapy (CT), chemoradiation therapy (CRT), or no adjuvant treatment (NAT). Kaplan-Meier survival analysis of patients surviving >= 6 months was performed, and predictors of OS identified by log-rank test. Cox multivariable analysis was used to define independent predictors of OS. Results: Among the 473 patients undergoing PC resection, the median survival was 17.8 months; for the 397 who survived >= 6 months following surgery, the 5-year OS for the CT, CRT, and NAT groups was 21%, 16%, and 17%, respectively (p = 0.584). Lymph node-negative patients demonstrated improved OS associated with chemotherapy on multivariable analysis (HR = 2.20, 95% CI = 1.25-3.83 for NAT vs. CT). Conclusions: Following PC resection, only patients with negative lymph nodes demonstrated improved OS associated with adjuvant chemotherapy.