INTRODUCTION: Various endoscopic tissue sampling techniques are utilized in the diagnosis of PBmalignancy. A number of small series suggest cytological examination of cellular debris from biliary stents affords sensitivity comparable to brush cytology. AIM: To assess the sensitivity and specificity of stent cytology in the diagnosis of PB malignancy. METHODS:All patients who underwent ERCP with stent removal for cytological examination were included. Nearly all subjects had prior negative intraductal tissue sampling but suspicion for malignancy persisted. Stents were transported in cytorich-red fixative and had debris scraped, placed on slides for PAP staining and were examined by an experienced cytopathologist. Stents removed from previously proven malignant strictures were not included.Final cancer diagnosis was based on cytologic results, surgery, other tissue sampling techniques, autopsy or clinical follow-up. RESULTS: Between Jan. 1994 and Aug. 1999, 68 biliary and 13 pancreatic stents in 68 patients were examined cytologically (M:F 38:31; mean age 61, range 19-87). Five patients were excluded because of lack of confirmation of the final diagnosis. All biliary stents were used in the treatment of biliary strictures. The indications for the 13 pancreatic stents were PD stricture (7), cystic lesion (4) and other (2). All cytology results recorded as atypical were considered benign for the purposes of this analysis. The sensitivity of stent cytology for the diagnosis of malignancy was 9.1%, specificity 100%, positive predictive value 100% and negative predictive value 58.9%. Three of the four cases with atypical cytology were from patients with malignant strictures: pancreatic adenocarcinoma, locally metastatic gallbladder carcinoma and mucinous cystadenocarcinoma. The fourth sample was from a patient with PSC. If these cases were included in the analysis, the sensitivity increases to 18.2% and the specificity is minimally decreased to 97.8%.SUMMARY: Cytological examination of cellular debris from biliary/pancreatic stents has a very low sensitivity for the detection of PB malignancy. This improves slightly when all grades of cellular atypia are considered indicative of malignancy. CONCLUSIONS: Stent cytology should not be used in isolation to aid in the diagnosis of pancreaticobiliary malignancy because of its low yield. It is an impractical test for the early confirmation of malignancy as cytological examination is delayed until stent removal.