The dilemma of "indeterminate" interpretations of pancreatic neuroendocrine tumors on fine needle aspiration

Background Indeterminate (atypical/suspicious) fine needle aspiration (FNA) interpretation of Pancreatic Neuroendocrine Tumor (PanNET) can be frustrating for both pathologists and clinicians. Although published literature has often addressed the significance of indeterminate cytologic diagnoses in non-neuroendocrine epithelial lesions, the significance of indeterminate FNA results for PanNET is less established. Methods The pathology records of The Johns Hopkins hospital were searched for histologic diagnoses of PanNET that had corresponding pre-operative FNA results. Cytopathologic diagnoses were categorized according to the index of suspicion described in the report (atypical, suspicious for PanNET and PanNET). Additionally, the exact reasons for the indeterminate nature of the diagnoses - such as lack of adequate cytologic features, unusual findings, or lack of confirmatory immunocytochemistry (ICC) were investigated. Results One hundred and twenty (120) cases of PanNET were identified. Of these, the FNA diagnoses were as follows: 78% (94/120) were PanNET, 9% (11/120) were suspicious for PanNET, 6% (7/120) were atypical, 4% (5/120) were benign, 2% (2/120) were non-diagnostic and 1% (1/120) was called a mucinous neoplasm. Of the 94 cases that were definitively PanNET, 74 (78%) required ICC for the diagnosis. Of the 26 cases that were not diagnosed definitively as PanNET, 73% (19/26) were morphologically suggestive of PanNET but did not have sufficient material for confirmatory ICC. Conclusions ICC is commonly employed to confirm the cytologic suspicion for PanNET. The most common reasons for indeterminate diagnoses were a quantitative lack of lesional cells and lack of material for confirmatory ICC.