Brush cytology of ductal strictures during ERCP

Previous reports on endoscopic retrograde brush cytology (ERBC) of bile ducts and of the main pancreatic duct have reached widely varying sensitivity levels of 33 up to 85%.

To report our experience with ERBC in a series of biliary strictures (n = 98) and pancreatic duct strictures (n = 8). For the purpose of our study, that was mainly directed to the value of the cytologic examination as such, only those specimens that were considered satisfactory for cytological interpretation were studied.

From October 1988 until August 1994, 154 cytologic brushings were performed at ERCP in 132 patients. In 132 brushings obtained from 115 patients (86%), cell yield was satisfactory for cytologic interpretation. Nine patients lacked adequate follow-up. Hence, 123 brushings from 106 patients were included in this study. A final diagnosis of malignancy was obtained in 62 patients.

Cytological changes were described as 'benign', 'columnar cell intraepithelial neoplasia', 'inconclusive' by the presence of atypical cells, or 'malignant'.

For a positive diagnosis of the malignant nature of a stenosis, ERBC had an overall sensitivity of 63% with a specificity of 96%. One false positive result was obtained in a patient with a biliary infection by Fasciola Hepatica. Sensitivity was highest in malignant ampullary strictures (91%). Sensitivity was 60% for cholangiocarcinomas, and 65% for pancreatic cancer. The finding of 'columnar cell intra-epithelial neoplasia' in the ampullary region led to a Whipple resection and diagnosis of invasive carcinoma in one patient. Atypical cells were found in 4 brush specimens: in three of these 4 cases, a malignant lesion proved to be present.

Brush cytology is a simple technique with a high specificity and should be performed in biliary and pancreatic duct strictures of unknown etiology. Categorizing the smears according to cytomorphology may improve diagnostic accuracy.