Background and study aims: Around 10% of patients with primary sclerosing cholangitis (PSC) develop cholangiocarcinoma, which is cholangiographically often indistinguishable from a benign dominant stricture. The aim of the present study was to assess the value of brush cytology in discriminating between benign and malignant dominant strictures in primary sclerosing cholangitis.
Patients and methods: The results of all brush cytology specimens from dominant strictures from patients with established primary sclerosing cholangitis, taken at endoscopic retrograde cholangiopancreatography between 1987 and 1996, were compared with the histological diagnosis or clinical status of the patients at least 2 years later.
Results: A total of 47 brush cytology samples, taken between 1987 and 1996, from 43 PSC patients could be included. Between 1993 and 1996, p53 immunocytochemical examination was done in 27 brush cytology specimens and K-ras mutation analysis in 25 patients. The sensitivity, specificity, positive predictive value, and negative predictive value of brush cytology for detection of malignancy were 60%, 89%, 59%, and 89%, respectively. These figures were not improved by adding the results of p53 and K-ras analysis. Logistic regression analysis did not reveal any additional benefit of p53 or K-ras analysis either. Prior stenting did not adversely affect specificity.
Conclusions: The sensitivity and positive predictive value of brush cytology for dominant strictures in PSC are rather poor. The specificity and negative predictive value are reasonably good. There was no additional value from p53 immunocytochemistry and K-ras mutation analysis. Prior stenting did not affect the results.