Abstract

Background: Alagille syndrome (ALGS) is an inherited liver disorder dominated by high γ-glutamyltransferase (GGT) cholestasis. Previous studies have demonstrated limited efficacy of surgical interruption of the enterohepatic circulation in ALGS, with varying degrees of improvement in pruritus and xanthomas. Utilizing the GALA database, we sought to evaluate whether surgical biliary diversion (SBD) alters the natural history of liver disease.

Methods: Multicenter retrospective analysis of children with clinically and/or genetically confirmed ALGS. Laboratory data were collected preoperatively (−6 to 0 months) and postoperatively (3 to 12 months). Paired sample t-tests were used to compare continuous variables, and McNemar’s tests were used to compare binominal variables pre-and postoperatively. Receiver operating characteristic (ROC) curves were used to determine the optimal laboratory threshold for predicting native liver survival (NLS) following SBD. Cox proportional hazards models were constructed to determine NLS in ALGS-SBD patients.

Results: Of 1673 ALGS patients, 3.7% (n=62; 54.8% male) underwent SBD from 26 centers. The median age of SBD was 2.5 years (IQR 1.8 – 4.4). Most ALGS patients underwent a partial external biliary diversion (54.8%, n=34), followed by a partial internal biliary diversion in 19.4% (n=12) and ileal exclusion in 12.9% (n=8). 100% (n=62) of patients reported pruritus at the time of SBD, and 51.4% (n=18/35) reported xanthomas. ALGS-SBD patients had a 2.5-fold greater risk of liver transplantation (LT) or death (95% CI 1.6 – 3.9; p<0.01). Following SBD, there were no significant differences in total bilirubin (TB) (8.8 vs. 9.1 mg/dL, p=0.51), ALT (159.5 vs. 189.7 U/L, p=0.38), GGT (495.0 vs. 459.5 U/L, p=0.21) or cholesterol (493.0 vs. 414.6 mg/dL, p=0.21). Availability of serum bile acids (SBA) was limited; however, in 10 patients with SBAs pre- and postoperatively, there was a significant reduction following SBD (257.2 vs. 97.7 μmol, p=0.05). Among these patients, 80% achieved NLS. TB levels < 4.0 mg/dL following SBD were significantly associated with longer NLS (p=0.05; AUC TB, 0.784; sensitivity, 94%; specificity, 52%). There were no significant improvements in pruritus (100% vs. 91.7%, p=0.25) or xanthomas (51.4% vs. 45.1%, p=0.67) after SBD.

Conclusion: SBD in ALGS was associated with an increased risk of LT or death. SBD may be a marker for severe hepatic phenotype in ALGS. In contrast to PFIC, SBD does not appear to improve NLS in ALGS. However, in a subset of ALGS-SBD patients with SBA, higher rates of NLS were noted in those who experienced a substantial decrease in post-operative SBA levels. These findings also indicate that post-SBD TB levels can be used as a biomarker for NLS. In an era of ileal bile acid transporter (IBAT) inhibitors, SBD may become obsolete in ALGS.