RESUMEN
BACKGROUND: Nonsurgical pathologic diagnosis of malignant bile duct stricture with a high sensitivity and specificity is desirable for therapeutic scheme. Percutaneous transhepatic endobiliary brush cytology in detecting obstructive jaundice is evaluated. METHODS: Fifty-eight consecutive patients with obstructive jaundice underwent percutaneous transhepatic cholangiodrainage (PTCD). During the process, a brush was inserted into the bile duct through the preexisted percutaneous transhepatic 8-F sheath, then exfoliated cells were collected from the bile duct stenosis and sent for cytologic diagnosis. The suspicious results were considered of negative diagnosis. All patients had relevant clinical data and follow-up results (15 months to 3 years). Statistical analysis was performed with the chi-square test and Fisher's exact test of probabilities, and a P value=0.05 was considered significantly different. RESULTS: Of all patients, 42 were diagnosed as having malignant stenosis according to brush cytology. Atypical hyperplasia was found in 9 patients and was suspicious of cholangiocarcinoma. Seven patients were found to have benign strictures. The sensitivity, specificity, positive predictive value, negative predictive value and accuracy of brush cytology were 75%(42/56), 100%(2/2), 100%(42/42), 12.5%(2/16), and 75.9%(44/58), respectively. No complications were related to the procedures. In all patients with malignant stenosis, the sensitivity of brushing in diagnosis of cholangiocarcinoma was greater than that in those with non-cholangiocarcinoma (P<0.05). The site of stenosis did not influence the sensitivity and technical success (P>0.05). CONCLUSION: The results of this study indicate that cytological brushing of biliary stricture during PTCD is a useful method to establish a diagnosis of malignant biliary stenosis with a high sensitivity, but the negative predictive value is not satisfied.