Portal hypertension predicts variceal hemorrhage in biliary atresia

WASHINGTON — Results of a prospective longitudinal analysis showed that the incidence of gastroesophageal variceal hemorrhage in children with biliary atresia was lower than previously described, and that gastroesophageal variceal hemorrhage was more common in patients with features of portal hypertension, according to a presentation at Digestive Disease Week 2018.

“The limitations of previous studies of variceal hemorrhage in biliary atresia include the fact that they were often single center, of retrospective design, and previous mortality rates do not account for improvements in care of patients with biliary atresia with variceal hemorrhage,” Lee Bass, MD, from the Ann & Robert H. Lurie Children’s Hospital of Chicago, said in his presentation. “[Our network] provides the opportunity for further characterization of natural history and risk factors impacting variceal hemorrhage.”

Bass and colleagues enrolled patients with their native liver from the Childhood Liver Disease Research Network (ChiLDReN). The researchers defined gastroesophageal variceal hemorrhage (GVH) as actively bleeding esophageal or gastric varices or bleeding in the presence of varices. They also assessed thrombocytopenia and splenomegaly at baseline.

The multi-center study comprised 636 patients with a mean age of 7.2 years (range, 1.9 months to 36 years). At enrollment, 30% of patients had both thrombocytopenia and splenomegaly while 30% had only one of these portal hypertension features. During a median follow-up of 1.7 years (range, 0-9.6 years), 59 patients had one or more episode of GVH.

The overall cumulative incidence of first GVH from baseline was 4.9% at 1 year, 9.3% at 3 years, and 12.5% at 5 years. Of those who experienced GVH, 42% had two features of portal hypertension at baseline, 37% had one feature, and 20% had none. Cumulative incidence of GVH in those with both features was 5.5% at 1 year, 11.1% at 3 years, and 16.8% at 5 years.

“While the highest incidence was in the youngest patients, there was otherwise an even spread across ages,” Bass noted.

The cumulative incidence of first hemorrhage correlated with the presence of portal hypertension features at baseline (P = .0126). Univariate analysis using models for time to first GVH showed that lower height z score (HR = 0.76; 95% CI, 0.61-0.95), weight z score (HR = 0.63; 95% CI, 0.51-0.78), platelet count (HR = 0.93; 95% CI, 0.89-0.97) and albumin (HR = 0.19; 95% CI, 0.13-0.29) correlated with a significantly higher risk for bleeding.

Platelet count (HR = 0.96; 95% CI, 0.92-1) and albumin (HR = 0.55; 95% CI, 0.37-0.81) remained significant after multivariate analysis.

“This multi-center study defines the overall natural history of variceal hemorrhage in biliary atresia,” Bass said. “We did not note any sign of mortality secondary to variceal hemorrhage. Hypoalbuminemia and thrombocytopenia significantly associated with future risk of variceal hemorrhage.” – by Talitha Bennett

Reference :

Bass L, et al. Abstract 502. Presented at: Digestive Disease Week; June 2-5, 2018; Washington, D.C.

Disclosure: Bass reports he received consulting, speaking or teaching fees from Mead Johnson. Please see the DDW faculty disclosure index for a list of all other authors’ relevant financial disclosures.