ABSTRACT
BACKGROUND: Biliary atresia (BA) is a rare disease of the newborn, resulting in liver cirrhosis due to obliterative cholangiopathy. Liver biopsies are commonly performed in order to confirm the diagnosis and in order to stage fibrosis. OBJECTIVES: The present study intended to analyze two established scores for evaluating liver fibrosis focusing on the interobserver variability as well as the prognostic reliability towards the time of liver transplantation. MATERIALS AND METHODS: Liver biopsies of BA patients between 2012 and 2015 were evaluated retrospectively by two pathologists at the Hannover Medical School (MHH) and the RWTH Aachen University Hospital. Fibrosis was measured using Ishak and Chevallier scores. Furthermore, a computerized automatically algorithm-based analyzation (ABAA) was performed. Results were evaluated towards the time point of liver transplantation and hepatoportoenterostomy (HPE). RESULTS: Overall, 34 liver biopsies were analyzed. The Ishak score showed a remarkable interobserver variability (ΚWâ¯= 0.68) while the Chevallier score was proven to have a poor interobserver variability (Fleiss' Κappaâ¯= -0.01). However, both scores were correlated positively, as was the ABAA (pâ¯< 0.001). Regarding prognostic reliability, ROC analyses of the Ishak score revealed the best validity towards an early liver transplantation within 12 months (AUC 0.813, pâ¯= 0.011). In addition, an increased Ishak score ≥4 reduced the survival time with the native liver (hazard ratio 6.6 [95% CI 1.9-23.3]). CONCLUSIONS: The Ishak score was revealed to have the best interobserver variability as well as prognostic validity towards an early liver transplantation in BA patients. Due to its easy applicability, the Ishak score was proven superior in comparison to the Chevallier score and ABAA. Therefore, use of the Ishak score is recommended in daily clinical routine for analyzing liver biopsies in BA patients.
