Primary Tumor Resection in Patients with Metastatic Pancreatic Neuroendocrine Tumors - An Analysis of the National Cancer Database

Abstract

Introduction: Consensus on the benefit of primary tumor resection for patients with metastatic pancreatic neuroendocrine tumors (panNET) is lacking. We assessed trends in surgical treatment patterns and evaluated the survival impact of primary tumor resection in patients with metastatic panNET. Methods: Patients with synchronous metastatic non-functional pancreatic neuroendocrine tumors in the National Cancer Database (2004-2016) were categorized based on whether they underwent primary tumor resection. Logistic and Cox proportional hazard regressions identified associations between patient and disease characteristics and primary tumor resection and overall survival (OS), respectively. Survival was compared with Kaplan-Meier and log-rank test within a propensity score matched cohort. Results: In the overall cohort of 2,613 patients, median age was 61 years (IQR 52-70) and 43% were women. Approximately 90% of patients had metastasis to the liver. Most (68%, n=1,774) patients did not undergo primary tumor resection. The proportion of patients who underwent primary tumor resection decreased from 36% (2004) to 16% (2016, p<0.001). After propensity score matching on age, grade, tumor size, tumor location, and hospital type, primary tumor resection was associated with longer OS (69 vs 22 months; p<0.001). After adjusting for age, Charlson-Deyo score, and tumor grade, primary tumor resection remained independently associated with longer OS (HR 0.36, p<0.001). Post-surgical 30- and 90-day mortality was 1.9% and 5.4%, respectively. Conclusion: In this large cohort of patients with metastatic panNET, primary tumor resection was strongly and independently associated with improved overall survival, suggesting that well-selected patients with metastasis at presentation should be considered for surgery.