Portal venous pressure in biliary atresia.

UNLABELLED: Biliary atresia (BA) is characterized by a variable degree of fibrosis within the liver, causing portal hypertension sometimes evident at the time of presentation.

AIM: The aims of this study are to measure portal venous pressure (PVP) at time of Kasai portoenterostomy (KP) and to investigate the value of surrogate indices.

METHODS: At the time of KP and before any liver manipulation, an attempt was made to recanalize the umbilical vein, allowing a catheter to be sited. Preoperative noninvasive variables included maximum splenic diameter (on ultrasound); platelet count, aspartate aminotransferase, and bilirubin; and the aspartate aminotransferase/platelet index ratio (APRi). Clearance of jaundice was defined as achieving a bilirubin of less than 20 μmol/L. Data are quoted as median (range). Nonparametric statistical tests were used, and P < .05 was regarded as significant.

RESULTS: Portal venous pressure measurements were available in 61 infants, who underwent a KP during the period February 2007 to October 2010. Median age at KP was 52 (19-151) days. Median PVP was 9 (3-26) mm Hg and was significantly lower in those with isolated (n = 47) BA vs cytomegalovirus-associated BA (n = 6) (8 vs 17 mm Hg; P = .02). Overall, PVP correlated with spleen size (r(s) = 0.38; P < .0001), APRi (r(s) = 0.5; P < .0001), bilirubin (r(s) = 0.38; P = .001), and age at KP (r(s) = 0.33; P = .0004). Thirty-three infants (56%) cleared their jaundice (to <20 μmol/L). There was no significant difference in median PVP (7 vs 9.5 mm Hg; P = .1) between these groups. To date, 6 children have undergone endoscopy, with 2 requiring intervention for significant varices. Neither infant at time of KP had elevated PVP (3 and 9 mm Hg).

CONCLUSIONS: Portal venous pressure is a key variable in the assessment of the BA liver, and differences can be shown depending on underlying cause (eg, cytomegalovirus status). Furthermore, it has modest correlation with noninvasive indices (eg, spleen size and APRi) but appears of limited use when used as prognostic index of outcome and indeed in the later development of varices.