Resected pancreatic cancer (PC): Impact of adjuvant therapy (Rx) at a high-volume center (HVC) on overall survival (OS).

Abstract

191 Background: Surgical outcomes for resected PC are known to be superior at HVCs. However, the impact of adjuvant (Rx) performed at HVCs is less studied. We examined the impact of site of adjuvant Rx administration on our resected patients (pts). Methods: Eligible pts were diagnosed 2003-2014 and resected at HVC. Pts were excluded for neoadjuvant Rx, synchronous cancer, death/lost to follow-up within 3 months or contraindications (e.g. morbidity) to adjuvant Rx.. Pts were also excluded if they refused adjuvant treatment or if a community oncologist (CC) was not identified in the medical record or in the western Washington population-based cancer registry. Pt and tumor characteristics were compared in univariate analysis and survival was calculated from date of diagnosis to death or last follow-up. Five year OS was estimated by the Kaplan Meier method and compared using Cox proportional hazards modeling to evaluate the impact of HVC adjuvant Rx on OS while adjusting for potential confounding factors. Results: 245 pts were eligible for study: 139 (57%) treated at HVC, 106 (43%) treated at CC. HVC and CC pts were similar with respect to stage and tumor size, nodal status, resection margins and average distance travelled to HVC. They differed by age (HVC: 63.1, CC: 68.2 p < 0.01). Median and 5-yr OS was 36 mos and 33%. Median OS for HVC vs CC was 44 mos vs. 28 mos (p < 0.01), and 5yr OS was 38.6% vs. 24.8% (p < 0.01), adjustment for age did not alter our findings. Conclusions: 1) With respect to adjuvant Rx for resected PC, HVC and CC pts differed with respect to age only. 2) Both median and 5- yr OS was statistically superior at HVC vs CC. 3) Our study supports the use of HVCs for all Rx components for PC treated with curative intent. 4) Ongoing investigation of patterns of care and their impact on OS in PC is warranted.