Screening for Biliary Dysplasia Using Brush Cytology in PSC : Prevalence, Risk Factors and Outcomes

BACKGROUND AND AIMS: Primary sclerosing cholangitis (PSC) is associated with significantly increased risk of cholangiocarcinoma (CCA) with a 5-year survival below 10%. Although CCA screening remains controversial, current guidelines recommend annual surveillance via ultrasound or MRI/MRCP. Brush cytology (BC) enables early detection of biliary dysplasia (BD), but its sensitivity varies widely (8%-89%). We assessed the accuracy of BC in detecting CCA, the prevalence and risk factors for BD, and the outcomes of dysplasia screening. METHODS: We analysed 4017 BC samples from 984 PSC patients who underwent ERC. The exclusion criteria (n = 114) were age at diagnosis < 18 or > 65 years, early CCA diagnosis (< 1 year from PSC diagnosis), and IgG4-related cholangitis. Demographics, laboratory data and ERC scores were compared between patients with repeated BD findings (≥ 3 times) or CCA (composite endpoint) and those without. Optimal predictors were assessed using time-weighted AUC and Cox regression. RESULTS: BD was detected at least once in 10% of the patients. Repeated suspicion of dysplasia in BC was also strongly associated with BD in explanted livers. The negative predictive value was 0.99, and the overall accuracy was 0.91. Only the ERC score (≥ 7, AUC 0.87) and S-IgG4-twAUC (≥ 1.06, AUC 0.76) predicted the composite endpoint. Combining both improved accuracy (ERC-IgG4-score, AUC 0.88). CONCLUSIONS: Brush cytology has a high negative predictive value for excluding CCA development. The ERC score and S-IgG4 are key predictors of biliary dysplasia, supporting their use in PSC surveillance. The ERC-IgG4-score may help guide individualized monitoring strategies.