The outcomes of "atypical" and "suspicious" bile duct brushings in the identification of pancreaticobiliary tumors: Follow-up analysis of surgical resection specimens.

Abstract

Bile duct brushing cytology is frequently used to investigate suspected pancreaticobiliary tumors, but it is associated with low diagnostic sensitivity. As recently proposed by the Papanicolaou Society of Cytopathology in its new classification scheme for pancreaticobiliary cytology, the use of "atypical" and "suspicious (for malignancy)" categories may give a better prediction of malignancy and improve patient management. A series of 65 "atypical" and "suspicious" results initially diagnosed at the University of Washington and Harborview Medical Centers between 2005 and 2012 were analyzed to determine outcomes based on the review of subsequent surgical resection specimens. Seven of 15 (47%) "atypical" cases demonstrated malignant outcomes, whereas 49 of 50 (98%) "suspicious" cases showed documented malignancy within a mean follow-up of 4 months (P < 0.0001 from Pearson's chi-squared test). The analysis of surgical resection specimens demonstrates that 59% of "atypical" and "suspicious" bile duct brushings with malignant outcomes were pancreatic adenocarcinoma (95% confidence interval (CI): 46-71%), 97% of pancreatic adenocarcinomas caused strictures in the head of the pancreas (95% CI: 86-100%), and that 86% of pancreatic adenocarcinomas were moderate to poorly differentiated (95% CI: 73-99%). Fifty-three of the 60 (88%) patients in the presence of prior stenting had a histologically confirmed malignancy. The use of "atypical" and "suspicious" categories may allow more accurate surveillance and/or more rapid surgical exploration of pancreaticobiliary tumors.