Combining Analysis of Per ERCP Bile Aspiration and Endobiliary Brushing Improve the Diagnostic Yield of Cytology in Malignant Strictures of the Biliary Tract. Mono-Centric Prospective Study

The cytological diagnosis obtained by brushing in malignant biliary strictures is considered to be highly specific but poorly sensitive. Improving diagnostic yield by bile exfoliate cytology has been suggested but is rarely performed. The aim of this prospective study was to assess the diagnostic performance bile aspiration associated with biliary brushing during an ERCP. Patients and Methods: From Jan 2007 to Sept 2008, 82 patients (51 M, 31 F), mean age 70 (32 - 93), among whom ERCP was performed for the treatment of a biliary stricture with suspicion of malignancy, were included. A final diagnosis of cancer was established in 57/82 patients by histology. A surgical resection was contra indicated for highly suspicious of malignant strictures because of the patient terrain or tumoral involvement. After sphincterotomy and crossing of the stricture with a guide wire, a single use brush protected by an 8 Fr catheter (ref. FS-CB-1.5 Cook ®) slipped on the wire was positioned above the stricture. Three to 10 ml of bile were aspirated and collected in a dry sterile tube before and after brushing. Brushing was performed with 10 passages in to-and-fro fashion across the stricture for classical cytological analysis (MGG, Papanicolaou) and in thin layer (Preservcyt ®). After centrifugation of the 2 samples of bile, an analysis was performed on a slide (MGG, Papanicolaou). Results: The biliary strictures were related to 34 pancreatic cancers, 16 cholangiocarcinoma and 7 other cancers including 4 metastatic tumors. Twenty-five were benign. One cholecystitis occurred (1.2%): . For the diagnosis of malignancy, the performance of procedures were as follows: brushing; Se = 65.5%, Sp = 100%, PPV = 100%, NPV = 54.5%, bile samples before brushing: Se = 75.9%, Sp = 100%, PPV = 100%, NPV= 63.2%. Bile samples after brushing did not provide additional diagnostic performance. The cellularity of samples was associated with better sensitivity for the cytological analysis of bile before brushing (p <0001) and brushing (p = 0.02). The best diagnostic performance was obtained by combining cytological analysis of bile before brushing and brushing: Se = 81%, Sp = 100%, PPV = 100%, NPV = 68.6%. The diagnostic sensitivity of combining cytological analysis was significantly higher for cholangiocarcinoma (100%) than for pancreatic cancer (82%) and metastases (66%) (p <0.01). Conclusion: In case of biliary stricture, the aspiration of bile before brushing during an ERCP is a simple and safe procedure. Analysis of bile before brushing in combination with brushing increase significantly the performance of cytological diagnosis of malignant strictures especially for cholangiocarcinoma.