Cytologic criteria to reduce error in EUS-FNA of solid pseudopapillary neoplasms of the pancreas

Abstract

Standardization of error classification in pathology remains an important issue. This study assesses the extent of error in cytopathologic diagnosis of solid pseudopapillary neoplasms (SPN) of the pancreas. Because of morphologic overlap of SPN and pancreatic neuroendocrine neoplasms (NET), we compared cytologic characteristics to determine which best distinguishes these entities. We collected cases diagnosed as SPN either by endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) by cytology or surgical pathology from January 2000 to March 2013. An equal number of NET cases were randomly selected. Cytology and surgical pathology cases were evaluated for diagnostic errors and patient impact. Cytologic features in SPN and NET were scored based on presence of previously described characteristics. A total of 17 patients with EUS-FNA were diagnosed with SPN by cytology or surgical pathology. Of those, 14 had surgical follow-up and 13 had adequate cell blocks and immunohistochemistry. There were 5 discrepancies between cytology and surgical pathology (5 of 14, 36%). There were no false positives or false negatives, but 5 misclassifications: 4 diagnosed as NET on cytology, and 1 as NET versus SPN. All misclassification errors were associated with no harm. When compared with NET, fine chromatin, nuclear grooves, pseud papillae, pink stroma, and hyaline globules are statistically significantly associated with SPN. EUS-FNA of pancreatic SPN has excellent positive and negative predictive value, with no false positives or false negatives in this 12-year study. Only misclassification errors as pancreatic NET were made with minimal impact. We suggest that the presence of 3 of 5 major cytologic criteria offer accuracy in diagnosing SPN to prevent misclassification.