Abstract

Bile duct brush cytology has been employed as a diagnostic tool for the evaluation of pancreatic and biliary tract strictures. The specificity of this method is high however, its sensitivity is quite low. A recent study employing liquid based cytology (LBC) reported results comparable to those achieved via conventional cytology. Therefore, we have attempted to prospectively evaluate the diagnostic utility of bile duct brush cytology in pancreaticobiliary diseases. A total of 46 cases with bile duct stricture were enrolled including 11 cases of benign stricture, 29 cases of bile duct carcinoma, 3 cases of gallbladder cancer, and 3 cases of pancreatic cancer. Both conventional smear and LBC using MonoPrep2(TM) system were conducted in each case. The cytological diagnosis of each case was classed into the following categories; benign, suspicious for malignancy, and malignancy. The diagnostic accuracy of both cytologic methods was investigated. LBC evidenced a high rate of material insufficiency (13/46), which was attributed to low cellularity. The kappa index of both cytological methods was 0.508. Cytological and tissue diagnoses were correlated in 25 cases conducted from biopsy or operation. The sensitivity, specificity, positive predictive value, and negative predictive value were 41.2% (7/17), 100% (8/8), 100% (7/7), and 44.4% (10/18) in conventional smear; 58.8% (10/17), 87.5% (7/8), 90.9% (10/11), and 50.0% (7/14) in LBC; and 94.1% (16/17), 87.5% (7/8), 94.1% (16/17), and 87.5% (7/8) in any one of both cytological methods, respectively. Based on these results, the sensitivity of LBC was found to be superior to that of conventional smear and we were able to obtain higher positive predictive value upto 94.1% by simultaneously conducting both cytologic methods.

• Keywords: Bile duct;Brush cytology;Liquid based cytology;Kappa index