Biliary Atresia - emerging diagnostic and therapy opportunities.

Biliary Atresia is a devastating pediatric cholangiopathy affecting the bile ducts of the liver. In this review, we describe recent progress in the understanding of liver development with a focus on cholangiocyte differentiation and how use of technical platforms, including rodent, zebrafish and organoid models, advances our understanding of Biliary Atresia. This is followed by a description of potential pathomechanisms, such as autoimmune responses, inflammation, disturbed apical-basal cell polarity, primary cilia dysfunction as well as beta-amyloid accumulation. Finally, we describe current and emerging diagnostic opportunities and recent translation breakthroughs for Biliary Atresia in the area of emerging therapy development, including immunomodulation and organoid-based systems for liver and bile duct repair.

Introduction

Cholangiopathies are diseases that primarily affect the bile ducts in the liver and are responsible for nearly 80% of all pediatric liver transplantations [1]. The biliary system takes the bile from the liver to the gall bladder and eventually into the small intestine. Biliary Atresia (BA) is a cholangiopathy which affects both the extra- and intrahepatic bile ducts of the liver (Figure 1). The incidence is particularly high in Asia (100-500 per 100,000 live births in Taiwan and Japan [2] as compared to Europe (5-25 per 100,000 live births) [3], for review see [4]. The current treatment options are limited to a surgical procedure called Kasai portoenterostomy (KPE), but KPE fails to improve the condition in nearly 50% of the patients and leaves the intrahepatic cholangiopathy unresolved. Furthermore, BA leads to fibrosis, portal hypertension and liver failure in many cases. Ultimately, liver transplantation is often needed, requiring life-long immunosuppression, which affects the quality of life for BA patients.

Fig. 1

Schematic depiction of Biliary Atresia. Potential causes for BA are described, as well as the major pathological features of BA, including obstruction of the extrahepatic bile ducts and deterioration of intrahepatic bile ducts. The gallbladder (green) pancreas (beige) and duodenum (light pink) are also shown. Hematoxylin and eosin (H&E) stained and cytokeratin 19 (CK19)-immunostained liver sections from a BA patient and a non-BA control are displayed. At the bottom of the figure, the main topics of the review article are described.

Genetics in BA is complex, and most cases are nonsyndromic. PKD1L1 mutations have been linked to syndromic form of BA [5], while common variants of a small number of genes (ADD3, CRIPTO, NODAL, LEFTY, GPC, EFEMP1 and ARF6) have been shown to be associated with increased risk of nonsyndromic BA [6], [7], [8], [9], [10]. Most recently, our trio-based whole exome sequencing study identified mutations of ciliary genes including PCNT, KIF3B and TTC17 in 31.5% of non-syndromic BA patients [11] (Figure 1). The etiology of BA is complex with both virus and hepatotoxin exposure appearing to be important factors, and the fact that elevated bilirubin levels can be detected at birth in children that will proceed to develop BA may suggest that exposure occurs at the fetal stage [12]. A number of viruses, including Cytomegalovirus (CMV), Epstein-Barr virus (EBV) and Human papillomavirus (HPV) have been linked to BA [13], and hepatotoxins such as the plant toxin biliatresone, can induce BA-like pathology in various species [14] (Figure 1). BA can be classified into different subcategories. One classification scheme divides BA into isolated BA (70-80%), syndromic BA (10-15% and including Biliary Atresia Splenic Malformation (BASM)), cystic BA (5-10%) and CMV IgM+ BA (approximately 10%) [15]. An alternative classification scheme reports approximately 80% isolated BA cases, a variable incidence of CMV BA, a congenital malformation form (10-20%), the BASM variant (1-10%), BA with major malformation variant (5-10%) and cystic variant (8%) [16]. Proposed aetiologies for BA and their potential relationships are also discussed by Tam et al., (2017) [17].

Molecular and cellular mechanisms regulating normal development of the extra- and intrahepatic bile ducts and aberrations in Biliary Atresia

BA affects both extra- and intrahepatic bile ducts, which have different developmental origins. The intrahepatic bile ducts, along with the cranial part of the common hepatic duct which represents a part of the extrahepatic bile ducts, originally stem from the cranial (proximal) part of the liver bud (pars hepatica) and the cholangiocytes forming the intrahepatic bile ducts are derived from hepatoblasts along the portal vein (Figure 2). The other major parts of the extrahepatic system, i.e., the caudal (distal) part of the common hepatic duct along with the gallbladder and cystic duct, instead originate from the caudal part of the liver bud (pars cystica) (Figure 2). The architecture of the intrahepatic ducts is complex, with probably at least 18 branchings. The branching process and the reciprocal interactions between the portal vein and intrahepatic bile ducts are yet poorly understood, but Notch signalling regulates branching after biliary injury [18]. A novel microCT method allows the simultaneous visualization of both the intrahepatic bile duct system and the vasculature [19], allowing the reciprocal interactions to be studied in further detail. The extrahepatic bile ducts appear earlier [15], pouching out from the foregut around 20 days of gestation, while the intrahepatic bile ducts appear around 45 days of gestation (see [20] for review) (Figure 2).

Fig. 2

Overview of the liver, with the extra- and intrahepatic bile ducts and their different developmental origins. The intrahepatic bile ducts and the cranial (proximal) common bile duct (shown in red) are derived from one portion of the liver bud (pars hepatica). The extrahepatic bile ducts, including the caudal (distal) common bile duct, as well as the gallbladder and cystic duct connecting the gallbladder to the common bile duct (shown in yellow), are derived from the pars cystica in the liver bud.

Cholangiocytes in the extra- and intrahepatic bile ducts follow different developmental trajectories (for review see [18]). For the intrahepatic bile ducts, cholangiocytes and hepatocytes differentiate from a common progenitor, the hepatoblast. Whether all hepatoblasts are bipotent and can give rise to both hepatocytes and cholangiocytes is not clear, but recent observations indicate that an Lgr5-positive cell pool remains bipotential throughout life [21]. We are gaining insights into the factors regulating the hepatoblast's choice to become a cholangiocyte or a hepatocyte, and this information is relevant for BA, as an instability in the cholangiocyte fate has been observed in BA organoids (see below). We are also beginning to identify transcriptional and epigenetic factors controlling hepatoblast fate choices, including TGF-beta, Wnt, Notch and Hippo/YAP signalling [20]. Cholangiocytes can furthermore be generated by transdifferentiation from hepatocytes in a TGF-beta-dependent manner [22], indicating that the differentiated hepatocyte state is not carved in stone but may be to some extent plastic.

During the early steps of differentiation of the extrahepatic bile ducts, i.e., when they appear as a diverticulum from the foregut, the transcription factor Sox17 plays an important role, and its absence leads to conversion to pancreatic tissue [23]. Downstream of Sox17 are other transcription factors, including Hfn6, Hnf1b and Hhex [24,25] and Fgf signalling contributes to the patterning of the extrahepatic biliary tree [26]. Collectively, these observations point to the existence of distinct molecular programs for cholangiocyte differentiation in the extra- and intrahepatic bile ducts.

Decoding the molecular landscapes of extra- and intrahepatic cholangiocytes by single cell transcriptomics reinforces the notion that the two types of cholangiocytes are different. Molecular profiles from human intrahepatic cholangiocytes have been obtained [27], and recently a comparison between intra- and extrahepatic cholangiocyte transcriptomes was conducted [28]. While there is a core transcriptional profile across cholangiocytes from different locations, different parts of the biliary tree show transcriptomic heterogeneity at the single cell level [28]. Interestingly, specific gene expression patterns for intra- and extrahepatic cholangiocytes were revealed, where for example the Notch ligand JAGGED1, which is mutated in Alagille Syndrome (ALGS), was found to be expressed in intrahepatic cholangiocytes, which may explain the intrahepatic bile duct paucity in ALGS. Furthermore, the transcriptomic profiling provided evidence for zonation along the biliary tree [28], and zonation has also been observed along the porto-central axis for hepatocytes and endothelial cells [29,30] as well as for hepatic stellate cells [31].

Technology platforms underpinning Biliary Atresia research

Considerable progress has been made in our understanding of BA through the use of different technology platforms for disease modeling, including rodents, zebrafish and liver organoids (see Table 1 for a list of studies using these systems).

Table 1

List of publications using model systems (rodents, zebrafish, organoids) to study Biliary Atresia.

Rodent (mouse & rat)	Main findings	References
Virus-induced.Inoculation of Balb/c mouse pups (1-2 days) with rhesus rotavirus (RRV)	Intraperitoneally RRV-infected neonatal mice developed jaundice and liver fibrosis, with a role for the RRV Spike Protein VP4	Riebentoff-Talty et al., (1993). REF 32Wang et al., (2011). REF 64Mohanty et al., (2020). REF 35
Surgery-induced.Ligation of the common bile duct in 21-30 days old Sprague-Dawley rats	Operated rats developed jaundice, choluria and acholia, progressive cholestatic injury, leading to massive liver fibrosis	Garrido et al., (2017). REF 37
Toxin-induced.Biliatresone treatment of Balb/c mouse pups (0-3 days)	Neonatal mice injected with biliatresone developed biliary obstruction and dysplasia or absence of extrahepatic bile ducts	Yang et al., (2020). REF 38
Morpholino-gene knockdown or CRISPR/Cas9 genome editing	Knockdown or knockout of BA candidate genes resulted in biliary developmental defects in zebrafish larvae	Cui et al., (2013). REF 6Tang et al., (2016). REF 43Cheung et al., (2021). REF 44
Biliatresone treatment	Zebrafish larvae treated with biliatresone displayed morphological defects of the gallbladder and extrahepatic bile ducts	Lorent et al., (2015). REF 14Zhao et al., (2016). REF 42Yang et al., (2020). REF 38
Epigenetics	Analysis of DNA methylation status using zebrafish	Matthews et al., (2011). REF 46Cofer et al., (2016). REF 45
Liver organoids in BA research	BA-liver tissue-derived and IPSCs-derived organoids from BA patients displayed biliary developmental defects resembling those observed in BA	Tian et al., (2019). REF 58Babu et al., (2020). REF 54Amarachintha et al., (2021). REF 55
CRISPR/Cas9 gene knockouts in IPSC-derived organoids	Knockout of ADD3 and GPC1 in organoids resulted in reduced formation of ductal structures	Tian et al., (2019). REF 58
Biliatresone treatment	Biliatresone treatment led to a disruption of apical-basal polarity in liver spheroids/spheroids	Waisbourd-Zinman et al., (2016). REF 56Fried et al., (2020). REF 57

Rodents

An initial breakthrough in experimental BA research was the development of a mouse model, where mouse pups are infected with rhesus rotavirus (RRV) [32]. This results in phenotypes which in many ways resemble the human pathology [32]. Although it is still debated whether rotavirus infection leads to BA in humans [33,34], work in the RRV mouse model has been very important to shed light on BA disease mechanisms (Table 1). A modified version of the RRV mouse model, based on a reengineered rotavirus was recently described and yields improved survival rates [35]. Success of RRV mouse experiments relies on genetic background, time of infection, virus dosage and the immunological status of the mouse. Balb/c mice are the most susceptible mouse strain, high dosage of RRV is still the only known agent that induces definite extrahepatic bile duct atresia, and the first 24 hours postpartum is the optimal time point for intraperitoneal injection of the virus. Modulation of the mouse immune system, for example by depletion of granulocyte receptor-1 (Gr-1) expressing myeloid cells attenuated the BA symptoms [36], further underscoring the importance of the immunological status. A surgical rat BA model was recently established, based on ligation of the common bile duct [37], and biliatresone induces BA-like effects in standard laboratory mice [38].

Zebrafish

Zebrafish (Danio rerio) is gaining ground as a model system for analysis of various organs because genetic and chemical manipulations can readily be conducted, transparent larval development allows organ development to be visualized in vivo, maintenance cost is low, and fecundity is high. Zebrafish is increasingly used also in early stages of cholestasis research, and liver development and function are in many aspects similar to that of humans, with zebrafish hepatocytes secreting bile salts through a bile canaliculus on the apical side, and the bile fluid is drained to a network of intrahepatic bile ducts that is connected to an extrahepatic bile duct. There are however some differences, including a different organization of the liver vasculature and the portal triads [39,40], and fibrosis and inflammation has been difficult to assess, possibly linked to that the adaptive immune system does not mature until after 4–6 weeks post-fertilization [41].

These differences however do not detract from the importance of zebrafish as a model system for BA, and zebrafish has significantly contributed to our understanding of BA, notably in terms of working out the effects of biliatresone (see above) [14,42] and validating candidate genes from genome wide association studies (GWAS) [6,43,44] (Table 1). Furthermore, pharmacological and genetic manipulations of DNA methylation in zebrafish revealed the role of DNA hypomethylation in BA [45,46].

Organoids

Liver organoids, i.e., 3D-cultured mini organs, are increasingly used to study different aspects of disease processes, including liver diseases [47], [48], [49]. Liver organoids can be differentiated towards both cholangiocyte and hepatocyte fates and have been used to study an increasing number of liver diseases, including a1-antitrypsin (A1AT) deficiency and ALGS [49,50], for review see [51]. Organoids can be generated from induced pluripotency cells (iPS) or embryonic stem cells (ES cells), hepatic progenitor cells, as well as directly from tissue-derived cells [51]. In one study, differences between intra- and extrahepatic organoids, notably with regard to their response to Wnt signaling, were observed [52]. Primary human extrahepatic cholangiocyte organoids were used to construct the mouse extrahepatic biliary tree [53] and repair human intrahepatic ducts [28].

In terms of BA research, organoids from BA patients or RRV-infected mice showed a breakdown in apical-basal polarity [54], a phenotype which was corroborated in a recent study [55], which also showed delayed epithelial differentiation in BA organoids. Furthermore, biliatresone led to a similar apical-basal disruption in control liver spheroids, accompanied by decreased glutathione and Sox17 levels combined with elevated Hey2 levels, indicating activation of the Notch pathway [56,57] (Table 1). Tian et al., (2019) used CRISPR/Cas9 genome editing to remove ADD3 and GPC1 in organoids [58], which resulted in reduced formation of ductal structures. Human cholangiocyte organoids (ICOs) were recently shown to be susceptible to rotavirus infection and mirror BA development, with a disrupted organoid morphology, combined with the induction of an immune response [59].

The recapitulation of at least some BA phenotypes in organoids opens up new vistas to use organoids both to screen for novel drugs that may ameliorate the disease process and to identify potential cholangiotoxic agents. While this may warrant cautious optimism, it should however also be recognized that organoids from different locations in the extrahepatic tree take on a similar transcriptomic profile after in vitro culturing [28], suggesting that some of the in vivo properties may be lost upon in vitro culturing. To identify niche conditions that keep their “originality” would be important and shed light on niche factors in general. As an alternative strategy, Du et al., (2020) reported the construction of a cholangiocyte epithelial layer with barrier function and apical-basal polarization on a chip [60].

Novel insights into Biliary Atresia pathomechanisms

Virus, inflammation and immune responses as pathomechanisms for BA

Virus infections have long been suspected to play a role as a cause for BA, and several viruses, such as EBV, CMV and HPV have indeed been found in BA livers [13]. Notably, 60% of BA cases in China contained CMV [61], and Davenport (2012) proposed CMV-associated BA as a BA subgroup, defined by liver biopsies staining positive for immunoglobulin M (IgM) antibodies against CMV [62]. CMV has also been used to recapitulate BA in mice depleted for regulatory T cells [63]. Progress has been made regarding the underpinning mechanisms for viral infections. In the RRV mouse model, the SRL peptide in the VP4 protein has been shown to be responsible for pathogenesis and interacts with heat shock protein 70 (Hsp70) on cholangiocytes [64]. Interestingly, the SRL peptide is not unique to proteins in RRV, but is also found on CMV proteins, and it may therefore be asked whether the SRL-Hsp70 axis is a common node for all viral cholangiocyte infections. If true, this may facilitate more generic therapy development covering different types of viral infections, but additional research is required to settle this question.

Viral infections may lead to autoimmune responses, which indeed are observed in BA patients, with accumulation of circulating macrophages and CD4/CD8 positive T-cells and autoantibodies [65]. Furthermore, autoantibodies targeting cholangiocytes were observed in BA patients [66]. Interestingly, in a recent report, Wang et al., (2020) conducted an analysis of the immune profiles of BA patients, and revealed an imbalance in the immune response, with expansion of cytotoxic T-cells and a reduction of CXC3RA T effector and natural killer cells [67]. These changes were accompanied by a continued hepatic B lymphocyte production in BA patients, along with accumulation of autoantibodies. In the same study, the use of rituximab to reduce B cell numbers was successful and in line with the findings in humans, the RRV mouse model exhibited a decrease in the CXC3RA T cell effector and NK cells [67]. Transfer of T-reg cells to newborn mice prior to RRV exposure blocked the extrahepatic bile duct obstruction [68,69], see [70] for review).

Inflammation is intimately associated with BA. In a recent report, RRV infection in mice was shown to cause increased expression and release of the protein HMGB1, which has been implicated in mediating inflammation [71]. HMGB1 induction was mediated via the p38 MAPK signalling pathway, which offers potential therapeutic opportunities. An inflammatory response, with elevated Toll like- and cytokine receptor expression was also observed in RRV-infected liver organoids [59]. Elevation of IL-17, produced by γδ T cells, was suggested to contribute to hepatic inflammation in the RRV mouse model of BA and in the livers of BA infants [72]. Treatment by AM80, a synthetic retinoid with superior pharmacological properties, suppressed the IL-17 production of γδ T cells and ameliorated hepatic inflammation in RRV mice [73]. In conclusion, there are a number of recent exciting findings that link viral infection to autoimmune responses and inflammation, which provide interesting novel ideas for future therapy development.

Disruption of apical-basal polarization in BA cholangiocytes

Acquisition of apical-basal polarity is fundamental for proper development and function of bile ducts [20]. Cholangiocytes have an apical and basal (basolateral) side, and are joined together by tight junctions, with expression of different claudins [74,75]. There are several cholangiocyte proteins that exhibit apical-basal polarization including CFTR, TGRS and AE2 at the apical side (facing the lumen) and the secretin receptor (SCTR) at the basal side [76]. Lysosomes and multivesicular bodies are present close to the apical side [77]. In BA livers and organoids, the apical-basal localization of a number of proteins including ZO-1, CFTR and SCTR, is disrupted [54], which corroborates earlier reports on apical-basal disruption and lumen obstruction in liver spheroids in response to exposure to biliatresone [56,57]. The altered distribution of CFTR may impact bile flow, as it plays a role in bile duct secretion. A recent report extends these findings by demonstrating that ZO-1, β-catenin, E-cadherin and claudin-3 are mislocalized in BA livers [78]. The same study also revealed a downregulation of Cdc42, a regulator of cell polarity, in BA livers, and knocking out Cdc42 in mouse cholangiocytes, which resulted in extra- and intrahepatic bile duct obstruction, further underscoring the importance of apical-basal polarization in cholangiocytes. Perturbation of the apical-basal axis is not restricted only to BA but also observed in ALGS [50] and may thus be common to a number of diseases affecting cholangiocytes [20,79]. To further explore the underpinning molecular principles for apical-basal polarity, including the role of Rho/Rac signalling and its dysregulation in BA, remains an important topic for future studies.

Primary cilia dysfunction

A special aspect of apical-basal dysregulation in BA is the effect on localization and abundance of primary cilia. Primary cilia, localized at the apical side of cholangiocytes [80], [81], [82], are important for cholangiocyte polarization and for sensing flow in the bile ducts through the mechanoreceptors PC-1 and PC-2. A reduced number of primary cilia was observed on BA cholangiocytes [83]. The importance of primary cilia as a BA pathomechanism is further corroborated by the fact a number of genes important for ciliogenesis, including PDK1L1, are mutated in BA (see below), and that laterality abnormalities are observed in BA [84,85]. Importantly, our recent whole genome sequencing study detected rare, deleterious de novo or biallelic variants of liver-expressed ciliary genes, including PCNT, KIF3B and TTC17, in 28 of 89 nonsyndromic BA patients [11]. Primary cilia play an important role for Hedgehog signalling [86] and it will be interesting to explore to what extent Hedgehog signalling is perturbed in BA. This may also shed light on cilia dysfunction in other liver diseases, such as neonatal sclerosing cholangitis and polycystic liver disease [16,87,88].

Beta-amyloid accumulation in BA

Beta-amyloid deposits in the form of neural plaques is a hallmark of Alzheimer's disease, but recently, Babu et al. (2020) showed that beta-amyloid also accumulates in livers in BA patients as well as in the RRV mouse model and in BA organoids (see above) [54]. BA thus joins the ranks of diseases with beta-amyloid deposits, such as Alzheimer's disease and cerebral amyloid angiopathy (CAA), but with the notable difference that beta-amyloid deposition appears to be intracellular in BA, in contrast to the extracellular deposits in other amyloid diseases. To explore the consequences of the intracellular beta-amyloid aggregates in BA, and what type of cleavage products of amyloid precursor protein (APP) they represent, remains to be explored. It also remains to be elucidated whether BA beta-amyloid deposition is caused by other viruses than RRV. We also do not yet know whether strategies to abrogate beta-amyloid accumulation, which have been explored in Alzheimer research, will prove beneficial, or whether beta-amyloid accumulation in BA is more of a bystander effect. It is of note that the γ-secretase inhibitor DAPT, which blocks cleavage of APP to beta-amyloid, led to amelioration of the BA phenotype in the RRV mouse model [89], but as DAPT also blocks Notch signalling [90], improvement may also be related to perturbing Notch signaling. Additional work is required to specifically address the effects of APP processing and beta-amyloid deposition and its role in the BA disease process.

Alterations in extracellular matrix and the cholangiocyte niche

Extracellular matrix (ECM) is critical for providing a niche for several cell types, such as intestinal stem cells [91], and there is an increasing interest in exploring the role of ECM also for cholangiocytes in BA, as progressive fibrosis is frequently observed in BA patients. Pathway enrichment analysis of BA patients with low survival after KPE revealed that the top 10 most dysregulated pathways were ECM-related [92] and Kyrönlahti et al., (2021) observed an increase in ECM gene expression after KPE [93]. Furthermore, elevated expression of matrix metalloproteinase 7 (MMP7), which modulates ECM by breaking down for example fibronectin and proteoglycans, is seen in extrahepatic cholangiocytes in BA patients and serves as a biomarker for the disease (see below) [94,95]. Much however remains to be learned about the ECM status in the vicinity of extra- and intrahepatic bile ducts and how ECM may function as a niche for cholangiocytes. This can be addressed by proteomics approaches, but as discussed above, the increasing sophistication of cholangiocyte organoid systems may also allow various ECM combinations to be directly tested for their effects on cholangiocytes.

Destabilization of the cholangiocyte cell fate in BA

As discussed above, intrahepatic cholangiocytes and hepatocytes emerge from a common hepatoblast progenitor, and it is therefore of interest to explore whether cell fate decisions or stability are affected in BA. Analysis of organoids from BA patients revealed a transcriptomic shift from a cholangiocyte towards a hepatocyte fate, i.e., cholangiocyte markers such as SOX1, FOXA1 and the Notch ligand JAGGED1 were downregulated, while hepatocyte markers, including HNF4A and TTR, were upregulated [54]. This may be indicative of a destabilized cholangiocyte cell fate in the BA disease process, and in line with this reasoning it is interesting to note that hepatic progenitor cells accumulated in BA livers [96]. Furthermore, cells with a marked intermediary biliary-hepatocyte phenotype were increased in ALGS [96], suggesting that the cholangiocyte-hepatocyte differentiation process may be affected also in other cholangiopathies. To decode which differentiation or fate stability steps are affected in BA remains a priority for future research.

Emerging diagnostic opportunities for Biliary Atresia

The importance of improving early diagnosis of BA is well recognized [97], as the negative impact of late surgery has been demonstrated in previous studies [98], and novel biomarkers that can stratify patients with regard to the need for liver transplants are warranted. In an ideal world, the investigation for BA should be non-invasive and able to provide accurate and sensitive diagnosis during the early phase of the disease, and although not yet a reality, it is an important aim to work towards this goal. The examination of bile-stained stool in infants with suspected BA is completely non-invasive. Some countries adopted the policy of universal screening of infant stool colour and reported promising results [97], but the cost-effectiveness remains controversial due to its low specificity. Except the minor discomfort associated with venipuncture, blood tests are nearly non-invasive. MMP7 was recently identified as a biomarker with promising potential, in particular as MMP7 levels can be established from a blood test [95]. A recently published meta-analysis has reported high sensitivity and specificity for the use of MMP7 in differentiating BA from other cholestatic diseases in newborns [99]. Genetic mutations can also be identified from DNA acquired through blood tests, but while genes with specific mutations (PKD1L1) [5] or carrying common variants in BA patients, such as ADD3, CRIPTO, NODAL, LEFTY, GPC, EFEMP1, ARF6, PCNT, KIF3B and TTC17 [[6], [7], [8],11], continue to be identified, genetic screening will likely only capture a small percentage of BA patients, as most cases are nonsyndromic. Traditionally, histological examination of liver biopsy sample by an experienced pathologist is considered as a relatively accurate diagnostic test [100], but acquisition of a liver biopsy is an invasive procedure, and a provisional diagnosis can only be established 6 weeks after birth [16]. The recent discovery of beta-amyloid deposits in the liver of BA patients [54] may become an adjuvant test to overcome some of the limitations in human histological analysis, but currently also requires a liver biopsy. It may however be hoped that in the near future, blood test for beta-amyloid can replace existing investigations to provide an early and the most precise diagnosis for infants with BA, similar to blood test-based analysis of beta-amyloid in Alzheimer patients.

Towards therapy development for Biliary Atresia

While KPE offers relief of jaundice in some BA patients, residual problems ranging from minor quality of life impairment to severe life-threatening events are common among the survivors [17]. Despite numerous attempts to improve the operative outcomes by modifications of surgical techniques, the result of KPE remains unsatisfactory and less than 60% of BA patients could remain transplant-free after 10 years from the KPE. Moreover, the native liver survivors are not ‘cured’ as more than half of them would suffer from complications including portal hypertension, recurrent cholangitis and progressive liver failure [101]. Even though liver transplant is generally regarded as the salvage treatment for a failed KPE, it is a major operation with significant morbidities and necessitates life-long immunosuppression [102]. In some countries, this treatment option is furthermore handicapped by the shortage of organ donors.

Novel treatments are thus needed, and there are at least two principal avenues towards therapy development for BA: pharmacological treatments that block or ameliorate the disease, or directly repairing or replacing the ailing bile duct tissue. In the first category, there are a number of clinical trials testing repurposed therapeutics (Table 2), but the efficacy of these trials is not yet known. An additional recent example is the outcome of characterizing the immunological responses in BA patients and the RRV mouse model [67]. As discussed above, B cell lymphogenesis persisted after birth in BA patients causing autoimmune responses, and in a small-scale clinical trial, a B-cell modifying therapy using rituximab resulted in a reduction of B cells combined with restoration of macrophage function.

Table 2

Registered clinical trials of therapeutics for Biliary Atresia.

Treatment tested	Masking	No of participating centres	Locations	Study period	Phase of study	ClinicalTrials.gov Identifier
Prednisolone	Double	1	Europe	1/2000 – 9/2008	3	NCT00539565
Corticosteroid	Triple	14	North America	11/2005 – 1/2013	3	NCT00294684
Intravenous immunoglobulin	Open label	7	North America	10/2013 – 7/2016	1,2	NCT01854827
Pentoxifylline	Open label	1	North America	1/2013 – 1/2022	2	NCT01774487
Granulocyte Colony-Stimulating Factor	Open label	2	North America, Asia	1/2018 – 1/2020	1	NCT03395028
N-Acetyl cysteine	Open label	1	North America	5/2018 – 12/2024	2	NCT03499249
Odevixibat	Triple	70+	North America, Europe, Middle East, Asia Pacific	7/2020 – 6/2024	3	NCT04336722
Maralixibat	Triple	7	North America, Europe	7/2021 – 4/2023	2	NCT04524390

On the tissue repair side, the technology for engineering endodermal tissues in vitro on decellularized matrices has improved during the last few years [103], and may be applied to engineering bile ducts. Engineered neo-bile ducts were successfully transplanted into pigs and well tolerated [104]. Recently, Sampaziodis et al., (2021) showed that extrahepatic organoids could successfully be transplanted to repair intrahepatic biliary ducts in human livers ex vivo and organoid-derived cells regenerated 40-85% of the injected ducts [28]. In immunocompromised mice where biliary duct damage was induced, only cholangiocyte organoids, but not MSC, grafted and similarly led to an amelioration of the duct damage [28]. Interestingly, primary cholangiocyte were not successful as donor cells in the restoration process, indicating that cells may need to be passed through the organoid state prior to transplantation. In conclusion, although there is still some way to go before therapies for BA are in routine clinical use, data from pharmacological modulations as well as cell transplantation experiments show considerable promise.

Conclusions

Biliary atresia is a devastating disease affecting newborns, and there is thus a dire need for therapy development. There is considerable progress on several fronts, ranging from recent insights into the underlying pathomechanisms to direct small-scale clinical trials using immunomodulatory therapies. There is also progress in advancing BA diagnosis, which is important to better stratify patients, in particular to decide if and when liver transplants are required. Innovative research using different model systems, including organoids, is also likely to further accelerate the discovery process for new pharmacological intervention principles and cell therapy approaches. There is thus reason to be cautiously optimistic that in a not-too-distant future, we will be able to develop therapies to ameliorate, or even cure, biliary atresia, for the benefit of the patients.

Contributors

Urban Lendahl: Literature search, concepts development, manuscript drafts

Vincent C. H. Lui: Literature search, concepts development, manuscript drafts

Patrick H. Y. Chung: Literature search, concepts development, manuscript drafts

Paul K. H. Tam: Literature search, concepts development, manuscript drafts

All authors have read and approved the final version of the manuscript.

Outstanding questions

BA remains a devastating inflammatory disease, primarily affecting newborns, and some outstanding questions remain before the disease can be fully combated. Firstly, it is important to further analyze the underlying pathomechanisms, and in detail learn at which steps during intra- and extra-hepatic differentiation steps the various disease-causing agents act. Secondly, there is room for improving regarding BA diagnostics, striving towards accurate and minimally invasive procedures. Finally, future research will hopefully provide pharmacological or cell-based therapies that are effective in routine clinical use.

Search strategy and selection criteria

Data for this review have been collected from Pubmed using search themes including: Biliary Atresia, liver disease, cholangiocyte, hepatoblast, therapy, diagnosis, mouse, rotavirus, zebrafish and organoid. Articles published in the timeframe 1990-2021 have been used.

Declaration of Competing Interest

The authors declare that they have no competing interests.