373 Background: Ampullary neuroendocrine tumors (NET) make up < 1% of all gastrointestinal NETs. Information about their behavior and prognosis is reliant on small case series. This study set out to describe the population of patients who are diagnosed with ampullary NETs and compare them to patients with duodenal and pancreatic head NETs. Methods: The National Cancer Database (2004 – 2016) was queried for patients with ampullary, duodenal, and pancreatic head NETs. Clinicopathologic and treatment characteristics were compared. Subset analysis was performed on patients who underwent surgery. Kaplan Meier (KM) analysis and Cox regression were used to analyze the survival of patients with ampullary NETs. Results: Overall, 872 patients were identified with ampullary NET, 9692 with duodenal NET, and 6562 with pancreatic head NET. Patients with ampullary NET had an average age of 60.9 +/- 14.5 years, were evenly split among men and women (N = 437, 50.1% vs N = 435, 49.9%, respectively), and primarily Caucasian (N = 663, 76.0%). 72.1% underwent local tumor destruction or surgery (N = 629). Most did not receive radiation (N = 832, 95.4%), chemotherapy (N = 627, 71.9%), or hormone therapy (N = 788, 90.4%). Patients with ampullary NETs had more poorly differentiated tumors (N = 119, 13.6%) than patients with duodenal (N = 159, 1.6%) or pancreatic head (N = 602, 9.2%) NETs. Patients with ampullary NETs had more positive lymph nodes (N = 288, 33%) than patients with duodenal (N = 915, 9.4%) or pancreatic head (N = 1381, 21%) NETs. At five years, the overall survival for patients with ampullary, duodenal, and pancreatic head NETs was 57%, 68%, and 46%, respectively. Within the surgical population, five-year survival for patients with ampullary (N = 367), duodenal (N = 991), and pancreatic head (N = 1961) NETs was 60%, 74%, and 72%, respectively. When compared, there was a statistically significant difference between the mean overall survival of patients with ampullary (98 +/- 4.7 months), duodenal (112 +/- 2.5 months), and pancreatic head (108 +/- 1.7 months) NETs (p < 0.001). In the cox regression analysis, sex, Charlson-Deyo score, lymph node positivity, lymph-vascular invasion, mitotic rate, chromogranin A level, 5-HIAA level, and tumor size did not correlate with survival. Increasing age (HR 1.04, CI 1.01 – 1.07, p = 0.008) and worse tumor differentiation (poorly differentiated HR 3.33, CI 1.38 – 8.04, p = 0.008 and undifferentiated HR 8.31, CI 2.77 – 24.92, p < 0.001 compared to well differentiated) were associated with increased mortality. Conclusions: This study sheds light on a rare tumor histology. When compared to patients who underwent surgical resection for duodenal or pancreatic head NETs, patients with ampullary NETs had a significantly worse prognosis. Identifying prognostic factors allows us to create more concrete treatment recommendations and provide patients with improved prognostic information.
DUODENUM-PRESERVING PANCREATIC HEAD RESECTION FOR CHRONIC PANCREATITIS, BENIGN AND PREMALIGNANT TUMORS OF THE PANCREAS
