“Genetic contributors and modifiers of biliary atresia”

Abstract

To date, the etiology and pathogenic underpinning of the progression of the most prevalent serious neonatal liver disease, biliary atresia, remains elusive. This disease presents as an aggressive form of neonatal cholestasis characterized by the destruction and obliteration of the extrahepatic bile ducts within the first few weeks of life, and rapid progression of biliary fibrosis, likely due to unremitting cholestasis and retention of biliary constituents including bile acids. [1] In ~ 5% of patients, biliary atresia has is associated with laterality features, suggesting that a genetic underpinning to a disease that begins soon after birth. [2] However, biliary atresia does not occur within families and twins are discordant, indicating an absence of strict Mendelian inheritance. [3,4] Despite this, genes related to bile duct dysmorphogenesis/ciliopathies overlapping with features of biliary atresia in both humans and non-human model systems have been proposed. [5–10] Taken together, strict genetic etiologies leading to a common pathway of a neonatal cholangiopathy resulting in biliary atresia remain elusive. Contributions from fibrogenesis and inflammation-based studies suggest that early engagement of these pathways contributes to disease progression, [11–13] but a recent double-blind study did not suggest any benefit from early use of corticosteroids. [14] However, there are genetic contributions to the adaptation and response to cholangiopathies and cholestasis that may be present in certain populations that likely impact upon the response to HPE and subsequent biliary tract function. To explore these options, we performed whole exome sequencing of infants with biliary atresia and these studies have provided several candidate genes whose variants likely impact upon disease, including a disease-associated variant of the ABCB4 gene prevalent in African-Americans that correlates with worse outcomes. All together, combinations of exome sequencing and large population studies are expected to reveal causative and modifying genes relevant to patients with biliary atresia as a means to provide therapeutic targets and potential opportunities for genetic screening.

Introduction

Biliary atresia was first comprehensively described in an extraordinary compilation of cases in 1892 by the Scottish surgeon John Thompson. [15] Dr. Thompson reported varying appearances of the extrahepatic biliary tree in these infants, some patient’s biliary tracts had obstructions along with cystic components, while other patient’s extrahepatic biliary tracts were completely atretic without a discernibly patent lumen of the extrahepatic biliary tree through to the duodenum. Overall, there was clearly a variable presentation of this disease between patients, with a common feature being complete obstruction of bile flow.

Advances in treatment for this uniformly fatal neonatal liver disease were limited until the groundbreaking development of the hepatoportoenterostomy (HPE) by Dr. Morio Kasai, a Japanese pediatric surgeon, in 1959. [16] Although it took time for adoption of this technique by surgeons outside of Japan [17–19], it soon evolved to become the standard of care for BA and improved mortality among infants who received an early diagnosis, prior to the development of irreversible cirrhosis or the life-threatening sequelae of cholestatic liver disease. Despite this significant surgical intervention, only ~ 50% of infants who undergo the Kasai HPE prior to the development of cirrhosis are able to avoid liver transplantation by the age of 2, with approximately 70% of all patients eventually requiring liver transplant by adulthood (see below). [20]

A critical area of investigation into biliary atresia concerns the cause, or more likely, the multiplicity of causes responsible for the aberrant development of the non-patent biliary tree, which may inform the utility of the Kasai HPE and explain the marked distinction in outcomes of infants undergoing the procedure. This Chapter is focused upon highlighting the current state of discovery and knowledge regarding genes that may confer susceptibility to BA or modulate outcomes among infants after HPE. It is likely that a deeper understanding of the genetic profile of infants with biliary atresia will provide clinically substantive insights into the etiology and severity of the disease, and enable the development of more precise diagnostic methods and rational therapeutic targets to improve outcomes. [21,22]

Biliary atresia, epidemiology and salient clinical concepts

Among all causes of neonatal cholestasis, biliary atresia is the most prevalent, amounting to approximately 40% of all cases of neonatal cholestasis. [23] The overall incidence of neonatal cholestasis is ~ 1:2500 births [24], with varying incidences of biliary atresia reported around the world. Biliary atresia appears to be more common in Southeast Asia, with an incidence around 1:5000 live births, compared to North America and Western Europe, with an incidence from 1:10,000–15,000 births. {Wildhaber: him 2012jv} The reasons for the global discrepancy in the incidence of biliary atresia are unknown.

From a clinical standpoint, it is often challenging to differentiate the child with biliary atresia from one that has physiological jaundice or another form of cholestasis, thereby hampering prompt diagnosis. One of the few hallmarks of diagnosis that can be utilized for diagnosis or even for public health intervention includes the lack of pigment in the stool, arising from a frank impairment of bile flow. The clinical presence of acholic stools is sometimes inapparent to the (often new) parents of these children as well as to caregivers. However, once caregivers are appropriately educated, reporting of stool color can serve as a tremendously useful screening tool for the identification of biliary atresia in areas of high prevalence. The successful use of a stool color card provided to parents of all infants born in Taiwan led to the capture of nearly every child with biliary atresia within the second year of its implementation. [25] This simple population-wide screening tool is an exemplary intervention with demonstrable efficacy in promoting earlier detection and referral for Kasai HPE. At present, the analysis of dried blood spots derived from the standard newborn screening card has not proven useful for biliary atresia, primarily because of the lack of a diagnostic metabolite. However, recent data suggests that measurement of direct bilirubin levels within the first two days of life could be used as a screening tool. [21,26]

Clinically, infants with biliary atresia develop firm hepatomegaly and splenomegaly as young as 4–8 weeks of age and laboratory tests usually reveal elevations of ALT, AST, total and direct bilirubin, while GGT and alkaline phosphatase levels often rise to levels > 1000 U/dl. [27] Ultrasonographic imaging may reveal an absent or atretic gall bladder, as well as fibrotic regions in the porta hepatis (“triangular cord”) and intriguingly, increased subcapsular flow. [28–30] The finding of increased subcapsular flow may perhaps provide a readily utilized screening technique to help differentiate biliary atresia from other forms of neonatal cholestasis [27] Notably, infants with biliary atresia exhibit a spectrum of presentations, including some with the presence of a gall bladder, a normal serum GGT, and pigmented stools early in life, which may obscure the diagnosis. To confirm the presence of biliary tract obstruction, the utility of a percutaneous liver biopsy has substantial value, which can lead to prompt histologic recognition of peribiliary fibrosis and bile duct plugs, or conversely, a disorder unrelated to biliary atresia, thus abrogating the need for a cholangiogram or Kasai HPE. Once the possibility of biliary atresia is entertained, a surgical referral is warranted to perform an intraoperative cholangiogram, and if patency of the biliary tree is not visualized, to proceed to the Kasai HPE.

It is well-established that delay in diagnosis of biliary atresia negatively affects the outcome of the Kasai HPE, with worse outcomes reported by some studies if the infant is > 45 days of age [1,31,32] Often, children present after a routine pediatric visit at ~ 60 days of age, at which point some patients have already progressed to cirrhosis and end-stage liver disease, precluding any benefit from the intervention. Children with cystic presentations of biliary atresia may have a better response to the Kasai HPE, conversely, children with associated congenital malformations and laterality features typically have a worse prognosis (see below) [33–36]. The lack of a uniform screening tool, coupled with the generally normal appearance of the persistently jaundiced infant, often leads to these marked delays in diagnosis, a central gap in optimizing the care of these infants.

In general, outcomes after Kasai HPE are divided among patients who achieve restoration of bile flow and those who do not. Among those with restored biliary drainage, as evidenced by a total serum bilirubin 3 months after Kasai < 2 mg/dl, the majority of patients are able to postpone liver transplantation, while children whose jaundice does not resolve typically develop stigmata of cirrhosis and require evaluation for liver transplantation in the first year of life. [1,20,32,37–40] Overall, the stratification of patients into one group or the other is poorly understood.

Data from a large, population-wide cohort study suggests that if a child does not require liver transplantation by the age of 5, over 70% are likely to survive with their native liver into adulthood [20]. Despite recognition of important prognostic factors such as the anatomy of the biliary remnant, presence of laterality features, age at HPE, and experience of the surgical center [20,34] the outcome for an individual patient is largely unpredictable, which suggests that other factors may be responsible for determining disease progression.

Non-genetic etiologic theories of biliary atresia

The putative causes of biliary atresia can be separated into a variety of etiologic categories, including toxin-related, viral, immunological, maternal fetal factors, and genetic. [4,41–44] From Australia, there are intriguing data related to an episodic form of newborn lamb cholangiopathy associated with a change in diet due to harsh weather conditions, whereby ingestion of plant-derived toxin from the Dysphania species damages the developing biliary tree. [45]. However, an exogenous toxin has yet to be associated with the disease in humans. Investigations into a viral etiology have yielded mixed results, implicating various strains of rotavirus, reovirus, and cytomegalovirus, among others, but there have been inconsistencies in the detection of viruses in patient samples. Preclinical and animal models have identified intriguing contributions to disease progression from various immunological mediators, viruses, and immune cell types, several of which have been demonstrated in human studies. [1,4,13,34,42,43,46,47] Currently, there is a paucity of support for a single viral or immunological insult in the etiopathogenesis of biliary atresia, though these factors certainly may play modifying roles for disease progression.

Epidemiological data further diminishes support for a viral etiology for this disease, since biliary atresia does not occur outside of the immediate newborn period and the incidence is not epidemic among humans, whereas an immune-mediated or autoimmune process is less plausible due to the lack of recurrence after transplantation. These observations establish the need to consider alternative modalities of disease causation.

Proposed genetic etiologies of biliary atresia

Among the more salient aspects of biliary atresia is that approximately 10% of subjects present with anatomical features known as the biliary atresia splenic malformation syndrome (BASM). [31,48,49] These patients have anomalies that are usually associated with laterality malformations either above or below the diaphragm in the context of biliary atresia. Typical anomalies above the diaphragm include Tetralogy of Fallot, dextrocardia, anomalies of the pulmonary or cardiac vasculature, and atrial and ventricular septal defects. Below the diaphragm, in addition to biliary tract anomalies, patients may present with midline or left-sided liver, polysplenia or asplenia, and interrupted inferior vena cava. [2] The laterality defects exhibited in BASM are phenotypically similar to those observed in the ciliopathies, a heterogenous group of diseases cause by functional and structural defects in genes that encode cilia proteins. [7,9,50–55] Cilia are evolutionarily conserved, aid in establishing the left-right axis in vertebrates and are present on many cell types, including the apical surface of cholangiocytes, where cilium function is integral to bile flow and bile ductular formation. [56] Moreover, diseases such as autosomal recessive polycystic kidney disease (ARPKD; PKHD1 gene) are associated with biliary cystic malformations that are grouped together as Caroli’s disease, and expression of the PKHD1 gene product, polycystin, was found to be drastically reduced in the livers of BA patients with and without attendant renal cysts, in contrast to normal liver [7,57] Moreover, investigations into rhesus rotavirus murine models of BA have found substantial loss of primary cilia from extrahepatic cholangiocytes, which was corroborated by the finding of decreased ciliation in extrahepatic ducts in children with non-syndromic BA, although cilia appeared normal in neighboring peribiliary glands (Karjoo JPGN 2013). Ultimately, determining if abnormalities in the development or function of cholangiocyte cilia are the basis for biliary atresia requires more extensive genetic and immunohistochemical analyses. Finally, the high incidence of biliary atresia in areas of Southeast Asia as well as Polynesia is suggestive of an inherited predisposition, although a local dietary or viral factor cannot be excluded. [58]

Observations that argue against a Mendelian paradigm of inheritance of biliary atresia include the lack of familial penetrance, although there are rare cases of familial biliary atresia [59,60] and discordant presentation of biliary atresia among twins, including monozygotic twins. [3,61–63] However, a growing canon of developmental diseases display unconventional patterns of inheritance, including Bardet-Biedl syndrome, a clinically heterogeneous ciliopathy marked by oligogenic inheritance, whereby mutations in multiple genes interact to generate the phenotype. A similar phenomenon related to gene-gene interactions, or epistasis, may account for the diversity of presentations in BA. Non-Mendelian inheritance is also observed in Alagille syndrome, the result of a haploid insufficiency of JAG1 or its receptor NOTCH2, which has a varied clinical presentation within families as well as between patients. [64,65] Considerations for epigenetic modification (e.g., methylation) as a cause of biliary atresia requires further support and may necessitate direct studies of hepatocyte or cholangiocyte DNA in patients.

An intriguing body of work suggests that there may be a component of maternal environment or microchimerism that could be contributing to the phenotypic heterogeneity and nonclassical genetic inheritance of biliary atresia. [66,67]

More recent inquiries into the genetic underpinning of biliary atresia have taken the form of genome-wide association studies (GWAS) that investigate the association of individual, common genetic variants with disease, known as single-nucleotide polymorphisms (SNPs). A GWAS conducted in the Han Chinese population identified a potential susceptibility region for biliary atresia between the genes ADD3 and XPNPEP1, chromosome10q25.1, and the association of one SNP with disease was replicated in independent Chinese and Thai samples. {Garcia-Barcelona 2010, Kaewkiattiyot:2011ev, Zeng 2014} More in-depth sequencing of a Han Chinese sample identified that a 5-SNP risk haplotype was associated with BA and the genotype correlated with reduced levels of ADD3 expression. [68] An attempt to replicate the association in a Caucasian cohort discovered a stronger signal in the first intron of ADD3 following SNP imputation, although the exact genotype at this SNP was not predictive of the degree of ADD3 expression. The ADD3 gene encodes adducin-γ, a membrane skeletal protein that functions in regulating cell-cell adhesions in various epithelia and is more highly expressed in fetal liver relative to adult liver.

Another potential causative gene, glypican 1 (GPC1), has garnered attention through GWAS of DNA copy number variation that identified a potential region of susceptibility to biliary atresia on chromosome 2q37.3, which was replicated in an independent cohort of patients demonstrating heterozygous deletion of GPC1 as the sole gene in this region. [6] Glypican family members are involved in various signaling and developmental pathways in hepatocytes and cholangiocytes, though GPC1 had not been previously associated with liver development or function. In addition to reporting abnormal expression of GPC1 by cholangiocytes in patients with biliary atresia, the authors knocked down the GPC1 analog in zebrafish, which led to a markedly disordered biliary ductular network. [6]

Other genes involved in bile duct dysmorphogenesis and cholestasis, the immune response, vasculogenesis and left-right patterning have been proposed to contribute to biliary atresia on the basis of clinical investigations and non-human model systems, including CFTR, JAG1, ZIC3, CFC1, INV, MIF, VEGF, IFN-γ, and SOX17. [7,69] Arikan:2006tm, Cui:2013dk, Uemura:2013fa,Kohsaka:2002cq,Lee:2010gr,Lee:2010bn}

At this point in time, the application of genome-wide studies has led to the discoveries of two credible candidate genes that may underlie the development of biliary atresia, ADD3 and GPC1. We anticipate that larger genetic studies across a variety of ethnic and racial strata, especially with the advent of next-generation sequencing technologies, will uncover additional genes that may play a role in the etiology of biliary atresia.

Genetic modifiers of the course of biliary atresia

The genetic profile of an individual can greatly influence the clinical severity of a disease, and despite the phenotypic heterogeneity of BA, to date, there are no known genetic modifiers of outcomes after the HPE procedure. Several authors have suggested that genes with biological relevance to biliary function may be involved, including A1AT, JAG1, CFTR, but these remain to be explored in detail [4,6,7,70,71].

A substantive body of evidence has accrued over the last 10–15 years detailing the remarkable capacity of the liver to handle cholestatic loads, engaging pathways from the membrane to the cytoplasm to alter nuclear transcriptional programs, particularly through regulation of the members of the nuclear receptor family, centered upon the bile acid receptor FXR. [72,73] The liver can adapt remarkably well to bile acid retention, but this response is inadequate in patients who develop cholestatic liver disease, suggesting that compensatory mechanisms may fail in predisposed individuals when exposed to a non-physiologic stress, such as biliary atresia. Given the patient-to-patient variability in response to cholestatic and fibrotic liver diseases, it is imperative to consider the role of genetic variance in shaping the adaptive response to cholestasis.

Interrogating the role of genetic contributions to disease severity may be accomplished by leveraging new genomic technologies, an application which concordantly promotes the discovery of novel disease-causing genes in biliary atresia. With the support of the NIDDK-funded ChiLDReN network (childrennetwork.org), we initiated an unbiased whole exome sequencing study to elucidate which genes possessed more functional variants among children requiring an early liver transplant after the Kasai HPE relative to children who survived with their native liver. Surprisingly, the infants who received a liver transplant displayed a greater frequency of variants in the gene encoding the phospholipid floppase, ABCB4. This transporter is known, when mutated, to produce a variety of diseases manifesting in childhood and adulthood including progressive familial intrahepatic cholestasis, type 3, low-phospholipid associated cholelithiasis, intrahepatic cholestasis of pregnancy and has also been implicated in drug-induced hepatotoxicity. [74–77] Reduced ABCB4 function diminishes the phospholipid content of bile, thereby allowing for bile acid-mediated damage to cholangiocytes.

A validation study was performed to ascertain which variants in the ABCB4 gene might be associated with worse outcomes after the Kasai HPE. Sequence comparison of the protein-coding regions of ABCB4 in 98 subjects who underwent an early liver transplant and 97 subjects who survived with their native liver demonstrated that 1 non-synonymous polymorphism differentially segregated between the two groups. [78] This variant, p.A934T, has been described in several patients with peribiliary fibrosis and cholestatic disease. [77,79,80] Ongoing analyses will determine if this ABCB4 variant is more prevalent than expected among African-American biliary atresia patients with worse outcomes after HPE. If this is validated, then, to our knowledge, it will be the first known variant of any gene that impacts disease severity in patients with biliary atresia. Moreover, if a reduction in ABCB4 function is the mechanism underlying the predilection for early transplant in these patients, this finding generates opportunities for novel choleretic or chaperone-based therapies. [81]

Future considerations

Since its first descriptions, the underlying cause(s) and outcome contributor(s) to biliary atresia, the most common serious pediatric liver disease of infancy, remain largely unknown. Recent work indicates that this knowledge gap is likely to be shortened due to large-scale clinical consortia, application of modern whole exome and whole genome genetic screenings, and novel immunophenotyping. As these unknowns become conscribed to historical rather than active clinical issues, we expect to be able to understand bile duct morphogenesis, developmental cholangiopathies, roles for ciliary function, adaptations to bile acid retention, and means to effect choleresis. Along with identification of genes and small molecules that play central roles in pathogenesis and adaptation may be novel screening tools for disease identification and therapeutic targets. The dim past of our understanding of biliary atresia is beginning to appear brighter.