Early detection of biliary neoplasia/CCA

Interesting new doctoral dissertation. Anybody doing regular ERC and brush cytology (BC)?


Primary sclerosing cholangitis (PSC) is particularly common in the Nordic countries. Most (70-80%) PSC patients have concomitant inflammatory bowel disease (IBD). The majority of patients are male. Increased levels of serum cholestasis markers in IBD patients lead to PSC suspicion, and diagnosis is verified with magnetic resonance cholangiopancreatography or endoscopic retrograde cholangiography (ERC). PSC patients usually become symptomatic over time, and PSC may lead to liver fibrosis and finally to cirrhosis.

Cholangiocarcinoma (CCA) develops in 7-13% of PSC patients. The prognosis of CCA is poor. Dysplasia of the bile ducts is thought to precede the development of CCA. Liver transplantation (LT) may be performed in end-stage PSC cirrhosis, but premalignant biliary changes may also be a partial indication for LT even if the patient is asymptomatic.

At Helsinki University Hospital, PSC patients regularly undergo systematic dysplasia surveillance with ERC and brush cytology (BC) to detect dysplastic changes of the bile ducts and to determine disease progression. ERC changes are graded according to the modified Amsterdam ERC score. If suspicion of dysplasia is repeatedly seen in BC, patients are referred to liver surgeons for consideration of LT. Our aims were to evaluate the relevance of the screening protocol at PSC diagnosis and later during the disease course to identify risk factors for biliary neoplasia, and to evaluate the utility of ERC, BC and ploidy analysis of biliary epithelium in detecting biliary neoplasia.

In their first, diagnostic ERC (n=261), 81% of PSC patients were asymptomatic, but 43% already had advanced ERC changes. Seven percent had malignancy or malignancy suspicion in BC. Those who had elevated liver enzymes and advanced ERC changes were most likely diagnosed with biliary neoplasia during follow-up. Female patients had milder ERC changes and a lower cumulative risk of biliary neoplasia. During the follow-up period (median 4.7 years), 2.7% were diagnosed with CCA and 3.1% with biliary dysplasia. Independent risk factors for biliary neoplasia were elevated extrahepatic ERC score, suspicious/malignant BC and elevated serum alanine aminotransferase.

During surveillance (median 6 years) with ERC and BC, 3.8% out of 447 PSC patients were diagnosed with CCA, and 5.1% had dysplasia in the explanted liver. ERC score, male gender, elevated liver enzymes, serum tumour markers, suspicious BC and inflammation or aneuploidy in BC were risk factors for CCA. All patients with biliary neoplasia had intra- and extrahepatic ERC changes.

126 PSC patients underwent liver transplantation in Helsinki University Hospital during 1984-2012, and 28% of these LTs were due to neoplasia suspicion for asymptomatic patients. The combined sensitivity of BC and ploidy analysis to detect biliary neoplasia was 68% with 86% specificity. The ten-year survival of asymptomatic patients after LT was 91%.

Screening and surveillance of PSC patients with ERC and BC may help find patients that are most prone to develop CCA. Patients with advanced extrahepatic ERC changes and elevated liver enzymes are most likely to be later diagnosed with CCA. On the other hand, patients with intrahepatic ERC changes only have a negligible CCA risk. If BC is repeatedly suspicious, even asymptomatic patients should be referred to evaluation of LT.


Ted, thanks for posting this. Of the last two ERCP’s I had (Sept 2016 and March 2017, I think the doc did not do any brushing.