Significance and indications for reoperative portoenterostomy in biliary atresia in light of long-term outcome.

Background: Among biliary atresia (BA) patients undergoing revision following failed portoenterostomy, long-term native liver survival (NLS), physical condition, and indications for revision were explored.

Methods: From 1977-2015, 33 of 95 BA patients (35%) at our institution underwent revision. Ten- and 20-year NLS rates (NLSRs) and biochemical statuses of 20-year-old native liver survivors were compared between post-portoenterostomy BA patients with and without history of revision. Factors associated with 10-year NLS following revision and optimal cut-offs for potential predictors were analyzed.

Results: Overall 10- and 20-year NLSRs were 57% and 54%, respectively. Ten- and 20-year NLSRs were 49% and 45% in the revision group and 63% and 61% in the non-revision group, respectively. Among 20-year-old native liver survivors, differences in admission rates between ages 10-20 years were not significant for cholangitis (revision, 46%; non-revision, 40%; P = 0.30) or portal hypertension (revision, 38%; non-revision, 15%; P = 0.21). Differences in hepatobiliary function-associated blood tests between these groups at 20 years old were only significant for aspartate aminotransferase (P = 0.02) and gamma-glutamyl transpeptidase (P = 0.047). Among potential predictors of 10-year NLS that we investigated, rate of change in total bilirubin (TB) over the first month post-portoenterostomy was best (P = 0.0019), and the receiver operating characteristic curve revealed an optimal cut-off for this first-month change in TB of -3.7 mg/dl (area under the curve, 0.85; sensitivity, 0.79; specificity, 0.83).

Conclusions: In approximately half of the patients, revision provided 10- and 20-year NLS, and biochemical status at 20 years old was comparable across revision and non-revision patients. Rate of change in TB during the first month post-portoenterostomy offers a sensitive predictor of revision.