AB080. Genetic findings provide insight of biliary atresia patient complexity

Background: Biliary atresia (BA) is a rare complex disease with unknown etiology. Current treatment of BA is Kasai portoenterostomy but is ineffective. BA is now the most common cause of pediatric liver transplantation worldwide. Characterize the disease complexity and stratify patients for personalized medicine is necessary. Genetic variants underlie BA pathogenesis and yet comprehensive genotype-phenotype correlations are yet to be investigated.

Methods: We first reviewed the disease course of 89 isolated BA patients with long term follow up (median =17.2 years), whose blood DNA was genotyped on Affymetrix5.0. Copy number variants (CNVs) and single nucleotide polymorphisms (SNPs) were called. Meanwhile 23 BA patients’ liver DNA was submitted to exome sequencing for discovery of de novo mutations. After wise we narrowed the genotype callings down to BA unique mutations, i.e., CNVs, SNV/INDELs, and BA-associated genes through gene-based association test of SNPs, then genotype-phenotype correlations were interrogated. Last, interconnectivity among the candidate genes were examined, topology of the molecular network was then interrogated correlating to the BA clinical complexity.

Results: Clinical revision revealed that 41.57% isolated BAs had chronic extra-hepatic diseases, with high prevalence of autoimmune-atopic diseases (22.47%) and glucose-6-phosphate dehydrogenase deficiency (14.29% in males). In genotype data, we shortlisted 29 CNVs ≥100 kb private to BA and related sted 29 (I) de novo BA-CNVs, perturbing genes known to hepato-biliary diseases, associated with BA liver pathology; (II) three BA-CNVs encompassing genes known to immunity defects, correlated with comorbidities of those immune disorders in three carriers, and overall BA-CNVs intersected ‘immunologically-important’ genes (P=0.017). Biologically BA-CNVs are anchored to other BA candidate genes as interactions were observed between genes encompassed by BA-CNVs (N=102) and BA-associated genes tagged by SNPs (N=103) (empirical P=0.039). Additional SNV/INDELs associated with intrahepatic biliary anomalies were uncovered in exome-sequencing. All together the BA candidate genes converge into a molecular network with inflammatory regulators as the signalling hub, moreover, the network fell into multiple function modules, which coincides with the BA patients’ clinical profile.

Conclusions: Genetic variants underpin BA clinical manifestations. The BA-associated common and rare genetic converge in a molecular network, which support the plausible associations of BA with a host of ‘non-BA’ diseases, as supported by observation in patients comorbidities with non-BA disease. We propose this ‘diseasome’ network approach that integrates clinical/epidemiological data and BA genetic findings to decode the phenotypic complexity of this rare disease.

Keywords: Rare complex disease; patient complexity; genetic screening; genotype-phenotype correlation; network