Planimetry of the porta hepatis in biliary atresia.

Abstract

Orientation and digital analysis of the biliary remnants in the resected porta hepatis in infants with biliary atresia. Samples were orientated intra-operatively then stained with haematoxylin and eosin and immunostained for cytokeratin 7 (CK7). Sections were then digitized and analysed. Most proximal transected surface area was defined as the porta hepatis area (PHA) and the biliary epithelial area was defined as 'BEA'. Data are quoted as median (range). Non-parametric statistical comparisons were made as appropriate. P < 0.05 was regarded as significant. Thirty-eight infants underwent surgery [median age 53 (16-120) days]. Eight specimens were excluded from the study due to technical reasons, leaving 30 specimens as the study cohort. Median PHA was 70 (30-133) mm2 , median BEA 0.57 (0.07-5.5) mm2 (r = 0.51; P < 0.002). The median BEA/PHA ratio was 9.6 × 10-3 (1.9-104 × 10-3 ). There was a marked correlation of PHA with plasma γ-glutamyl transpeptidase (r = -0.51; P = 0.001). Both total BEA and the BEA/PHA ratio correlated with alkaline phosphatase (r = -0.35; P = 0.03 and r = -0.47; P = 0.005, respectively). Age at surgery correlated inversely with BEA (r = -0.44; P = 0.01) but not PHA (P = 0.1). Precise quantification of biliary remnants is possible and correlates with biochemical variables. Values for BEA were associated with and declined demonstrably with increasing age at surgery.