Interpretation of follow-up 3D-MRCP/MRI in patients with primary sclerosing cholangitis

Background: In primary sclerosing cholangitis (PSC) many centers perform regular follow-up magnetic resonance cholangiopancreatography (MRCP) and magnetic resonance imaging (MRI), particularly to early detect bile duct pathologies. To date, no standardized MRI protocols for PSC have, however, been established, nor do MRI-based definitions (e.g. of dominant strictures) exist. Thus, the interpretation of follow-up MRCP/MRI and MRI-based recommendations for/against endoscopic retrograde cholangiopancreatography (ERCP) may significantly vary between physicians. We therefore aimed to evaluate the interpretation of follow-up MRCP/MRI for PSC among PSC experts.

Methods: Members/associates of the international PSC study group were invited to an online-survey (surveymonkey.com) consisting of 16 real-life PSC cases. Each case included essential clinical and biochemical information and video material of follow-up 3D-MRCP/MRI. Using a multiple-choice questionnaire, participants were asked to interpret 3D-MRCP/MRI and for recommendations, particularly with respect to ERCP. The agreement among the participants was calculated using Fleiss-kappa.

Results: 44 participants (19 hepatologists, 16 gastroenterologists, 9 radiologists) with a median PSC experience of 14 years completed the survey. With respect to the overall assessment, presence of dominant strictures and suspicion of cholangiocarcinoma the agreement reliability analysis revealed only a slight agreement among the group. The lowest agreement was found with respect to ERCP recommendation (yes/no; Fleiss-kappa = 0.12, 95% CI 0.11 – 0.14), with a relative agreement > 75% in only 4/16 cases.

Conclusion: Our study demonstrates that in PSC the interpretation of follow-up 3D-MRCP/MRI and MRI-based recommendations significantly vary even among experienced practitioners. Standardized MRI protocols/definitions and standards for ERCP indication are required for the management of PSC patients.