Magnetic resonance imaging as a tool for diagnosis, evaluation, and prognosis in patients with primary sclerosing cholangitis

Abstract

Primary sclerosing cholangitis (PSC) is a chronic liver and bile duct disease of unknown etiology. It usually progresses to liver failure and need for liver transplantation. Additionally, patients with PSC have an increased risk of developing hepatobiliary malignancies, predominantly cholangiocarcinoma.

Radiology plays a central role in the diagnosis and follow-up of PSC patients. Diagnosis of the main variant of the disease, large duct PSC, is based on typical cholangiographic findings in magnetic resonance cholangiopancreatography (MRCP), in patients with relevant clinical and laboratory background. Patients that have normal cholangiogram but histological and clinical findings compatible with PSC are diagnosed with another, less frequent, variant with a better prognosis and milder course, called small duct PSC. Nevertheless, patients with small duct PSC may progress to large duct PSC. Magnetic resonance imaging (MRI)/MRCP is now recommended as a follow-up modality due to the lack of ionizing radiation and because it provides an overview and evaluation of both bile ducts and liver parenchyma. However, the radiological image of PSC is very complex making the interpretation of imaging findings difficult. Due to the high integration of imaging in the care of PSC patients, agreement of the interpretation of MRI/MRCP examinations of these patients must be secured.

It is hard to predict the main clinical events in PSC, namely which and when patients are going to develop liver failure, need for liver transplantation, and cholangiocarcinoma. The existing prognostic models are mostly for use in clinical studies and are not appropriate for application on individual patients. Imaging has been suggested as a potential prognostic biomarker, but its full potential has not been completely evaluated.

In this thesis we aimed to i) assess the agreement between radiologists of the interpretation of MRI/MRCP examinations of PSC patients, ii) evaluate the clinical and imaging features and the potential association of imaging and laboratory findings to progression of small duct PSC to large duct PSC, iii) evaluate the agreement and prognostic value of existing radiological prognostic scores for PSC, and lastly iv) to develop a novel prognostic radiological score for large duct PSC and assess its agreement and prognostic value.

In study 1 we evaluated the agreement between radiologists of the interpretation of the radiological course between consecutive MRI/MRCP examinations of PSC patients and found it to be low.

In study 2 we evaluated the radiological, clinical, and laboratory features of patients with small duct PSC and assessed the potential association between imaging and laboratory findings to progression to large duct PSC. More than half of the patients that had available follow-up studies progressed to large duct PSC and we observed no association between imaging/laboratory values and progression to large duct PSC.

In study 3 we evaluated the agreement and prognostic value of existing radiological scores for PSC called the ANALI scores. We found that the agreement was poor, however, after consensus between radiologists was reached, the scores were significantly associated with outcomes.

In study 4 we described the development and assessment of a new score for patients with large duct PSC, the DiStrict score, that was based on cholangiographic findings as seen on MRCP. We found that the score had a good agreement when applied by radiologists with experience in PSC and that it was significantly and independently associated with outcomes.

The results of our studies verify the insufficient agreement of the interpretation of MRI/MRCP findings in PSC, when performed in a non-standardized manner and when they are based on subjective parameters, and at the same time confirm the potential of imaging findings in the prognostication of patients with PSC.