Objective: To determine the diagnostic accuracy and pitfalls of endoscopic ultrasound (EUS)-guided fine needle aspiration (FNA) cytology of pancreatic neuroendocrine tumour (PanNET). Methods: A search of our laboratory information system was performed from July 1992 to June 2010 to identify all FNA cytology and corresponding surgical specimens in which the diagnosis of PanNET was rendered or considered. Results: One hundred and thirty-two cases diagnosed by EUS-guided FNA were collected. Histological correlation was available for 77 (58%) of FNAs; 55 patients may have been treated elsewhere or had no surgery because of advanced disease or co-morbidity. Among 56 cases diagnosed as PanNET on FNA, 54 (96%) were confirmed histologically; the remaining two were poorly differentiated adenocarcinoma with focal neuroendocrine features in one case and no tumour was found in the other. Follow-up histology of nine patients diagnosed as suspicious for PanNET on FNA showed four PanNETs, two pancreatic ductal adenocarcinomas (PDA), one solid pseudopapillary tumour (SPT) and two cases of chronic pancreatitis. Nine cases rendered by FNA as atypical (n = 3), no atypical cells identified (n = 4) or unsatisfactory (n = 2) were PanNETs on histology. Lastly, three cases of oncocytic variant of PanNET were misdiagnosed on FNA as either adenocarcinoma (n = 2) or as suspicious for carcinoma (n = 1). Conclusions: Overall, 54 of the 70 histologically confirmed PanNET cases (77%) were correctly diagnosed by preoperative FNA as PanNET. FNA cases designated as no atypical cells identified and unsatisfactory (7/132, 5%) were attributable to sampling error. Diagnostic pitfalls in our study mainly included PDA, SPT and chronic pancreatitis.
- Endoscopic ultrasound
- EUS-guided FNA cytology
- Fine needle aspiration
- Pancreatic ductal adenocarcinoma
- Pancreatic neuroendocrine tumour
ASJC Scopus subject areas
- Pathology and Forensic Medicine