Back

Lymphatic Dysplasia Shapes Liver Epithelial Outcomes in Biliary Atresia

Schumm, L.; Chua, R. L.; Kuzinska, M. Z.; Meindl, M.; Quach, S.; Weihs, J.; Gogol, D.; Zschiesche, M.; Duennebacke, R.; Uecker, M.; Behncke, R. Y.; Sudy, A.; Liebig, J.; Kamp, J.-C.; Kucklick, M.; Globke, B.; Oellinger, R.; Kuebler, J.; Henning, S.; Slavetinsky, C.; Haegerling, R.; Madadi-Sanjani, O.; Eils, R.; Guillot, A.; Bufler, P.; Conrad, C.; Rezvani, M.

2026-10-13 pediatrics
10.64898/2026.10.12.26365253 medRxiv
Show abstract

Biliary Atresia (BA) is the leading cause of pediatric liver transplantation. Given the heterogeneity of the disease, outcomes after diagnosis remain unpredictable. A multiscale atlas of BA hepatic tissue established early lymphatic dysplasia in infants that later proceed toward liver failure and transplantation (severe BA). An architecturally distorted, immature and CCL21-high lymphatic vasculature marked severe BA, along with an expanding cholangiocyte compartment pivoting toward immune-reactive MMP7/SOX9-high states. Hepatocyte-to-biliary transdifferentiation emerged broadly across BA- and non-BA-cholestatic injuries, while dedifferentiation and acute-phase-reactions were specific to severe BA. Functionally, lymphatic endothelial tissue, isolated from severe BA livers, induced an analogous LGR5- and SOX9-positive epithelial progenitor response when assembled with cholangiocytes as organoids. Further experiments revealed that lymphatic endothelium signals ectopically on liver epithelium via CCL21-CCR7-axis after BA-specific inflammatory priming. CCL21-depletion strategies mitigated the BA-specific MMP7-program. In sum, a maladaptive lymphatic endothelial dysplasia perturbs intrahepatic epithelium in severe BA and offers CCL21 as a possible target for epithelial stability.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.