Observation Versus Resection for Small, Localized, and Nonfunctional Pancreatic Neuroendocrine Tumors: A Propensity Score Matching Study.

Abstract

The best treatment modalities for small (1-2 cm), localized, and nonfunctional pancreatic neuroendocrine tumors remain controversial. Therefore, we aimed to evaluate whether surgical resection provides survival benefit over observation in those patients. From 1973 to 2015, all eligible patients were retrieved from the Surveillance, Epidemiology, and End Results database. Propensity score matching (1:2) method was performed. The primary endpoints evaluated were overall survival (OS) and cancer-specific survival (CSS). We identified 681 patients, of which 122 and 559 patients received observation and resection, respectively. Propensity score-matched patients who underwent surgery (n = 183) had significantly improved OS (P = 0.008) compared with matched patients who underwent observation (n = 106), but there was no difference in CSS (P = 0.310). On multivariate analysis, resection could improve OS but not CSS. Besides, poorly differentiated/undifferentiated tumor had a worse OS and CSS. Subgroup analysis showed that patients 60 years and older who underwent resection could achieve a longer OS and CSS. This disease exhibits a very good prognosis. Patients undergoing resection were associated with comparable 5-year CSS but longer 5-year OS compared with those receiving observation. Elderly patients (≥60) may obtain benefit from surgery, whereas the treatment of younger patients should be individualized.