Background and study aims: The role of endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) in the diagnosis and management of cystic pancreatic neuroendocrine tumors (PNETs) is unclear. We aimed to compare clinical/endosonographic characteristics of cystic with solid PNETs, determine diagnostic accuracy of preoperative EUS-FNA, and evaluate recurrence rates after resection.
Patients and methods: All patients with cystic or solid PNET confirmed by EUS-FNA between 2000 and 2014 were identified. A matched case-control study compared 50 consecutive patients with cystic PNETs with 50 consecutive patients with solid PNETs, matched by gender and age at diagnosis of index cystic PNET. We compared clinical/endosonographic characteristics, assessed diagnostic accuracy of preoperative EUS-FNA for identifying malignancy, and analyzed tumor-free survival of patients with cystic and solid PNETs.
Results: Cystic PNETs tended to be larger than solid PNETs (mean 26.8 vs. 20.1 mm, P = 0.05), more frequently nonfunctional (96 % vs. 80 %, P = 0.03), and less frequently associated with multiple endocrine neoplasia type 1 (10 % vs. 28 %, P = 0.04). With surgical pathology as reference standard, EUS-FNA accuracies for malignancy of cystic and solid PNETs were 89.3 % and 90 %, respectively; cystic PNETs were less associated with metastatic adenopathy (22 % vs. 42 %, P = 0.03) and liver metastasis (0 % vs. 26 %, P < 0.001). Cystic fluid analysis (n = 13), showed benign cystic PNETs had low carcinoembryonic antigen (CEA), Ki-67 ≤ 2 %, and no loss of heterozygosity. Patients with cystic and solid PNETs had similar recurrence risk up to 5 years after complete resection.
Conclusions: Cystic PNETs have distinct clinical and EUS characteristics, but were associated with less aggressive biological behavior compared with solid PNETs. EUS-FNA is accurate for determining malignant potential on preoperative evaluation. Despite complete resection, recurrence is observed up to 5 years following surgery.
