Impact of concurrent medications use on outcome of pancreatic cancer: SEER Medicare analysis.

Abstract

378 Background: Preclinical studies have suggested concurrent non-antineoplastic medication use may impact pancreatic cancer. We performed an analysis to examine these associations on survival in pancreatic cancer. Methods: Data from the Surveillance, Epidemiology and End Results (SEER)-Medicare with available part D data between 2006 and 2009 were analyzed. Cases with non-adenocarcinoma histology, autopsy only and death certificate only data were excluded. Drug use was defined as having two prescriptions filled within 12 months of pancreatic cancer diagnosis. The following medications were analyzed: beta-blocker, statin, insulin, metformin, thiazolidinedione, warfarin and heparin. The primary end point was overall survival. Stepwise Cox proportional hazard models were employed including each medication group as well as age, gender, stage, site, grade, diabetes, race and Charlson comorbidity score. Results: There were 13,702 cases which met inclusion criteria and had available Part D data. Median age was 76 years, There were 42.5% males, 77.1% were white and 34.0%had diabetes. Head of the pancreas tumors accounted for 49.9% cases and 49.6% had stage 4 disease and at diagnosis. The results of the stepwise Cox proportional hazard models are summarized in table 1. Beta blockers, heparin, insulin, warfarin were associated were significantly with improved survival (p<0.05). Conclusions: Concurrent medication use, particularly heparin, insulin warfarin and beta-blockers are associated with improved survival in patients with pancreatic cancer. Diabetes medications (metformin, TZD) did not have an impact in the multivariable model. Additional studies are needed to examine whether these medications may improve outcomes for patients with pancreatic cancer. [Table: see text]