Pathogenesis and Outcome of Biliary Atresia: …...(18), overlap in values precludes the...

Invited Review

Pathogenesis and Outcome of Biliary Atresia: Current Concepts

*†‡Ronald J. Sokol, *†Cara Mack, *†Michael R. Narkewicz, and *§Frederick M. Karrer

*Pediatric Liver Center and Liver Transplantation Program, Section of Pediatric Gastroenterology, Hepatology and Nutrition,Departments of †Pediatrics and §Surgery, and the ‡Pediatric General Clinical Research Center, University of Colorado Health

Sciences Center and The Children’s Hospital, Denver, Colorado, U.S.A.

Neonatal cholestatic disorders are a group of hepato-biliary diseases occurring within the first 3 months oflife. Bile flow is impaired, and patients have conjugatedhyperbilirubinemia, acholic stools, and hepatomegaly.Overall, 1 in 2,500 live births is affected with a neonatalcholestatic disorder (1). The two most common causes ofneonatal cholestasis are biliary atresia and idiopathicneonatal hepatitis, accounting for up to 50% to 70% ofcases. Other causes include a variety of neonatal infec-tions (viral, toxoplasmosis, syphilis, bacterial), metabolicand genetic diseases, progressive familial intrahepaticcholestatic disorders (PFIC), paucity of interlobular bileduct disorders (e.g., Alagille syndrome), choledochalcyst, ischemia–reperfusion injury, association with par-enteral nutrition administration, and other conditions(Table 1). Despite clinical improvement after the porto-enterostomy procedure, approximately 70% to 80% ofchildren with biliary atresia will eventually require livertransplantation; thus, biliary atresia alone accounts foralmost 50% of all liver transplants performed in children(1). It should be noted that $77 million is spent each yearin the United States on liver transplantation for childrenand the ensuing hospitalizations (2). This sum of moneycovers 0.2% of total health care expenditures related tochildren, even though these children represent 0.0006%of the total pediatric population. Importantly, this dispro-portionate expenditure for liver transplantation in chil-dren could be cut in half if improved therapies for biliaryatresia were developed that could abrogate or furtherdelay the need for liver transplantation. Remarkably,little is known about the etiopathogenesis of biliary atre-sia; consequently, there has been slow progress in devel-oping improved therapies or preventative strategies dur-ing the past decade. The purpose of this review is to

summarize recent advances in the diagnosis and manage-ment of biliary atresia, examine the clinical outcome,describe the evolving theories of the etiology and patho-genesis of this disorder, and highlight gaps in our currentknowledge.

DEFINITIONS AND CLINICAL FEATURES

Biliary atresia is the most common neonatal cholestat-ic disorder, occurring in approximately 1 of 8,000 (Asiancountries) to 1 of 18,000 (European countries) livebirths, with a female preponderance, characterized bycomplete fibrotic obliteration of the lumen of all or partof the extrahepatic biliary tree within 3 months of life(1). Fibrous obliteration may involve the entire extrahe-patic biliary system or any part of the system, with con-comitant injury and fibrosis of intrahepatic bile ducts(hence, the term extrahepatic has been dropped from thename of this disorder in recent years). In a prospectivestudy conducted between 1968 through 1993 in Atlanta,Georgia, the calculated incidence of biliary atresia was0.73 cases per 10,000 live births, with a higher preva-lence in African-American children than in white chil-dren (3). There was also a considerable seasonal cluster-ing of cases, with the incidence three times higher ininfants born between December and March. In contrast,a recent study using the Swedish National Health Data-base found no seasonal occurrence of biliary atresia be-tween 1987 and 1997, with an incidence of 1 in 14,000live births (4).

Biliary atresia is most likely a clinical phenotype re-sulting from a number of prenatal or perinatal insults tothe hepatobiliary tree, although the etiologic factors andpathogenesis of the obliteration of the biliary tree arepoorly understood (5). In approximately 20% of patientswith biliary atresia, the presence of at least one othermajor congenital anomaly suggests that defective devel-opment of the bile duct apparatus played a role in thesecases (5,6). In particular, polysplenia syndrome (poly-splenia, midline liver, interrupted inferior vena cava, si-tus inversus, preduodenal portal vein, and malrotation ofthe intestine) is present to some degree in 8% to 12% of

Supported in part by NIH grants 5 MO1 RR00069, General ClinicalResearch Centers Program, National Center for Research Resources;P30 DK34914; RO1 DK38446; UO1 DK62453; K08 DK060710; TheMadigan Foundation; and the Abby Bennett Liver Research Fund.

Address correspondence and reprint requests to Dr. Ronald J. Sokol,Professor and Vice Chair of Pediatrics, Box B290, The Children’sHospital, 1056 East 19th Avenue, Denver, Colorado 80218, U.S.A.(e-mail: [email protected]).

Journal of Pediatric Gastroenterology and Nutrition37:4–21 © July 2003 Lippincott Williams & Wilkins, Inc., Philadelphia

4

all children with biliary atresia (7,8). Biliary atresia thatis associated with these and other congenital anomalieshas been termed the fetal or embryonic form, although itmay be a common phenotype of multiple prenatal etiol-ogies (9,10). For some patients, it has been proposed thatthese anomalies are caused by abnormal expression ofgenes (somatic or inherited mutations) that regulate bileduct development, such as those that determine lateralityof thoracic and abdominal organ development (associa-tion with polysplenia syndrome) (11). Alternatively, anintrauterine insult may interrupt normal development ofmultiple organs, including the biliary tree.

The more common form (found in 80% of patients)of biliary atresia is not associated with other congenitalanomalies and has been termed the perinatal or acquiredform, in which it is believed that various perinatal orpostnatal events trigger progressive injury and fibrosisof a normally developed biliary tree during a criticalperiod in the first 3 months of life (5,9,12). Clinically,the fetal form of biliary atresia is associated withearly-onset jaundice and acholic stools (within thefirst 3 weeks of life without a jaundice-free period),whereas the acquired form of biliary atresia generally hasonset of jaundice and acholic stools in the 2nd to 4thweeks of life after a period of normally pigmented stools.Despite these potential disparate etiologies, the clinicalphenotype of these two forms of biliary atresia mayappear identical unless other congenital anomalies

(suggesting the fetal form) are discovered on clinicalevaluation.

Idiopathic neonatal hepatitis is a descriptive term usedfor neonatal intrahepatic cholestasis in which the char-acteristic “giant cell hepatitis” lesion is present on liverbiopsy, and for which no other infectious, genetic, meta-bolic, or anatomic cause is identified (13). In older series,idiopathic neonatal hepatitis comprised up to 30% to40% of all neonatal cholestasis cases. However, duringthe past two decades, infants believed to have idiopathicneonatal hepatitis were later found to have newly dis-covered metabolic and genetic diseases, such as �-1 an-titrypsin deficiency, PFIC, neonatal iron storage disease,inborn errors of bile acid synthesis (14), type 2 citrul-linemia (15), Niemann–Pick type C disease, and infec-tions with newly described viral agents (e.g., parvovirusand human herpes type 6 [HHV-6]). Up to 20% of casesof idiopathic neonatal hepatitis are progressive, appear tobe familial, and hold a worse prognosis. These cases maybe caused by novel genetic or metabolic disorders, someof which are now classified as PFIC, and others yet to bedefined. Thus, the term idiopathic neonatal hepatitis nowdefines an ever-shrinking percentage of cases of intrahe-patic cholestasis (perhaps 10–20%) for which an etiologyhas not been discovered.

The clinical presentation of idiopathic neonatal hepa-titis and biliary atresia is similar, although patients withbiliary atresia tend to be female and appear well nour-ished, whereas those with idiopathic neonatal hepatitisare frequently male, small for gestational age, and failingto thrive. It has been stated that the well-nourished ap-pearance of infants with biliary atresia may result in adelay of diagnosis. However, when careful anthropomet-ric evaluation is performed, infants with biliary atresiahave significantly decreased fat stores and lean bodymass (16). The added mass of an enlarged liver andspleen and the occasional finding of subclinical ascitesmay account for the appearance of a well-nourished in-fant (a relatively normal weight for age and weight forlength on standardized growth curves) (16). In both con-ditions, jaundice, hypopigmented or frankly acholic(white or gray) stools, dark urine, and hepatosplenomeg-aly develop within the first 3 months of life. Althoughthere are no other defining physical features of biliaryatresia or neonatal hepatitis, findings that may lead todiagnosis of other causes of neonatal cholestasis includeheart murmur and dysmorphic facial features (Alagillesyndrome), hypotonia and long forehead (Zellweger syn-drome), and lymphedema (Aagenaes syndrome). Labo-ratory evaluation typically shows conjugated hyperbili-rubinemia (>20% of total serum bilirubin) with elevationof serum concentrations of hepatocellular (aspartate ami-notransferase [AST], alanine aminotransferase [ALT])and canalicular (alkaline phosphatase, �-glutamyl-transpeptidase [GGT]) enzymes (17). Although patientswith biliary atresia tend to have higher GGT concentra-tions than those with intrahepatic forms of cholestasis

TABLE 1. Differential diagnosis of neonatal cholestasis

Intrahepatic causesCongenital infections—viral, protozoan, spirochetal, bacterial

sepsisMetabolic disorders—galactosemia, tyrosinemia, hereditary

fructose intolerance, alpha-1-antitrypsin deficiency, cysticfibrosis, hypopituitarism, bile acid synthesis defects, citrindeficiency, respiratory chain disorders.

Storage diseases—neonatal iron storage disease, Niemann-Picktype C, Gaucher’s disease, Wolman’s disease, glycogenstorage disease type 4.

Genetic syndromes—Alagille syndrome, Turner syndrome, DownSyndrome, Aagenaes syndrome, Zellweger syndrome,arthrogryposis/cholestasis syndrome

Progressive familial intrahepatic cholestasis—FIC1 deficiency,BSEP deficiency, MDR3 deficiency, Byler syndrome

Idiopathic disorders—idiopathic neonatal hepatitis,non-syndromic paucity of interlobular bile ducts

Toxins and drugs—endotoxemia, total parenteralnutrition-associated cholestasis, chloral hydrate, antibiotics,other drugs

Miscellaneous—ischemia-reperfusion injury, neonatal lupus,congenital hepatic fibrosis, Caroli’s syndrome, inspissated bilesyndrome, histiocytosis X

Extrahepatic causesBiliary atresiaCholechochal cystSpontaneous perforation of the common bile ductCholedocholithiasisNeonatal sclerosing cholangitisBile duct stenosis, compression by tumors or masses

BILIARY ATRESIA PATHOGENESIS AND OUTCOME 5

J Pediatr Gastroenterol Nutr, Vol. 37, No. 1, July 2003

(18), overlap in values precludes the differential value ofthis test. Surprisingly, in patients with biliary atresia,serum total bilirubin is rarely more than 12 mg/dL (andmay be as low as 5–8 mg/dL), and conjugated bilirubinis usually less than 8 mg/dL despite complete bile ductobstruction, whereas bilirubin levels may exceed 20mg/dL in patients with idiopathic neonatal hepatitis.Consequently, infants with biliary atresia may appear toonly have mild scleral icterus without cutaneous jaun-dice, particularly if they have dark skin. Elevated serumbile acid concentrations are universal for patients withthese disorders and do not differentiate between intrahe-patic and extrahepatic obstruction. However, unexpect-edly low or normal serum bile acid concentrations sug-gest a defect in synthesis of bile acids and should promptevaluation for these disorders (14).

Although biliary atresia and idiopathic neonatal hepa-titis are the two most common forms of cholestatic liverdisease that occur in neonates, there are many other dis-orders that must be considered in the differential diag-nosis (Table 1) (17,19). Treatable causes (primarily in-fectious and metabolic) must be identified promptly toprevent progressive liver injury and irreversible damageto other organs. Recent discovery of the genetic and mo-lecular causes of several forms of PFIC (e.g., mutationsin genes coding for BSEP, FIC1, and MDR3) (20) hasnot only provided an explanation for the etiology of theserare but important disorders but also has led to the dis-covery of new bile acid and phospholipid membranetransporters. This new knowledge has revolutionized ourunderstanding of mechanisms of bile formation, pre-disposition to gallstone disease, and the role of heterozy-gote carriage of these and other genes in modifying orcausing other hepatobiliary diseases (21,22). The discov-ery of the genetic basis of Alagille syndrome, mutationsin the JAGGED-1 gene that codes for a ligand to theNotch receptor (23), has invigorated investigation intothe molecular control of differentiation and morphogen-esis of the intrahepatic bile ducts and the cardiovascularsystem.

CURRENT THEORIES OF ETIOLOGY OFBILIARY ATRESIA

Our understanding of the etiology and pathogenesis ofliver and bile duct injury in patients with biliary atresiaand idiopathic neonatal hepatitis has remained essen-tially unchanged until recent years. Continued investiga-tion in this area is essential to provide a rational scientificbasis for the development of novel therapeutic and pre-ventative strategies. Currently, biliary atresia is believedto be a common phenotypic response of the neonatalliver and bile ducts to a variety of prenatal and perinatalfactors, including genetic and environmental insults (vi-ral, metabolic, vascular) that perturb the normal devel-opment or maturation of the biliary tree and that occurduring a specific period (prenatally to before 3 months ofage) amid the milieu of a genetic or immunologic sus-ceptibility to this disease (Fig. 1). Biliary atresia andidiopathic neonatal hepatitis are not believed to be in-herited disorders (except for the 10–20% of familialcases of idiopathic neonatal hepatitis and some cases ofpolysplenia syndrome-associated biliary atresia) becausehuman leukocyte antigen (HLA)-identical twins discor-dant for biliary atresia have been described, and recur-rence of biliary atresia within the same family is exceed-ingly rare (24,25). Nevertheless, this does not excludethe possibility that during fetal development somatic mu-tations of key genes regulating morphogenesis of thesestructures may be involved, or that there is a geneticpredisposition to an aberrant immune response that isonly expressed on exposure to an exogenous agent dur-ing a critical period.

An additional factor leading to progressive sclerosis ofthe extrahepatic bile ducts, regardless of the etiology ofthe initial injury and disruption to the ductal epithelium,may be the extravasation of bile with its detergent con-stituents into the submucosa, eliciting a secondary in-flammatory sclerosing process. The potential role of thismechanism was first described in 1901 by Rolleston andHayne (26), who coined the term descending cholangitis.

FIG. 1. Proposed interaction of four factors that result in the phenotypic expression of biliary atresia.

R. J. SOKOL ET AL.6


Tan et al. (27) have demonstrated the presence of ex-travasated bile in bile duct remnants of patients withbiliary atresia, perhaps originating from breaks in thebiliary mucosa at the porta hepatitis. This group furtherproposed that there was a vulnerable stage of humanbiliary development between 11 and 13 weeks of gesta-tion, when failure of remodeling of the ductal plate struc-tures could lead to disturbances in the normal formationof an adequate mesenchymal cuff around developing hi-lar bile ducts, which could potentially be prone to ruptureat the initiation of bile flow at 12 to 13 weeks of gesta-tion. Extravasated bile in adjacent periductal tissueswould then lead to protracted inflammation and fibrosis,causing secondary obliteration and obstruction of themore distal extrahepatic bile duct.

The pathogenesis of ongoing injury and sclerosis ofintrahepatic bile ducts after portoenterostomy is also notwell understood. Is it continued ductal obstruction to bileflow even after a “successful” portoenterostomy thateventually leads to biliary cirrhosis, albeit at a slowerpace if initial bile flow is established? Is it an ongoingimmune or autoimmune process that targets intrahepaticbile ducts? Is it asymptomatic bacterial cholangitis orabsorption of bacterial cell wall products and lipopoly-saccharides that continue to drive the intrahepatic in-flammation? Clearly, bile duct epithelial cells are fullycapable of secreting a number of cytokines and growthfactors that can recruit inflammatory and hepatic stellatecells and activate synthesis of extracellular matrix (28).In addition, cholestasis leads to the retention of hydro-phobic bile acids in hepatocytes, which stimulate thegeneration of considerable oxidative stress, mitochon-drial dysfunction, and cell pathways involved in hepato-cellular injury and cell death (29). In this way, hepato-cytes may participate in the induction of fibrosis in pa-tients with biliary atresia through secretion of cytokinesand lipid peroxide products that regulate collagen syn-thesis in hepatic stellate cells (30).

The following summarizes the six current theories ofthe etiology of biliary atresia (Table 2).

Viral infection

Epidemiologic studies support a possible infectiousetiology to biliary atresia and idiopathic neonatal hepa-titis. There has been continued demonstration of seasonalclustering of cases, suggesting environmental exposureto an infectious agent (3). In addition, several models ofviral infection in newborn mice produce lesions similarto biliary atresia (5), as described below. In 1974, Ben-jamin Landing, a pediatric pathologist, proposed that bil-iary atresia, idiopathic neonatal hepatitis, and chole-dochal cyst represented the result of different primarysites of injury to the hepatobiliary tree by a commoninsult, and coined the term infantile obstructive cholan-giopathies (31). Although Landing proposed involve-ment of hepatitis B virus, subsequent studies have shownno association between the common hepatotropic viruses(hepatitis A, B, and C) and biliary atresia. More recentattention has focused on the possible role of five viruses.

For many years, cytomegalovirus (CMV) has beenproposed as a possible etiologic agent because a modestproportion of infants with biliary atresia and idiopathicneonatal hepatitis have been infected with CMV, as arehealthy infants (32). Although a recent study from Swe-den (33) showed a higher prevalence of CMV antibodiesin mothers of patients with biliary atresia, and CMVDNA was present in livers from 50% of infants withbiliary atresia, a Canadian group (34) could not demon-strate CMV in bile duct remnants from 12 children withbiliary atresia. The role of CMV has not been explored ina large prospective multicentered study with proper con-trols.

The two viruses most commonly implicated are reo-virus and rotavirus. Interest in reovirus stemmed fromthe observation that infection in weanling mice causedpathologic features of the intrahepatic and extrahepaticbile ducts and the liver similar to those of biliary atresia(35). These lesions persisted even after infectious virusor viral antigens could no longer be detected. One groupdetected reovirus antigens in bile duct remnants frominfants with biliary atresia (36,37) and in an infant rhesusmonkey with biliary atresia (38); however, others did notreplicate these findings in a study of infants with biliaryatresia (39). Serologic studies of reovirus antibodies ininfants with biliary atresia have likewise been inconclu-sive (36,39–41). The high incidence of passively trans-ferred maternal antireovirus immunoglobulin (Ig) G mayhave confounded these studies. Two groups of investi-gators have examined hepatobiliary tissues removedfrom infants with biliary atresia for reovirus RNA. Steeleet al. (42) failed to detect reovirus RNA in archived,formalin-fixed preserved hepatic tissues from 14 biliaryatresia patients, 20 idiopathic neonatal hepatitis patients,and 16 control subjects using a nested reverse transcrip-tase–polymerase chain reaction (RT-PCR) assay. In con-trast, Tyler et al. (43) reported nested RT-PCR evidenceof reovirus infection in snap frozen liver or bile duct

TABLE 2. Proposed etiologies of biliary atresia

Infectious—reovirus, rotavirus, retrovirus, cytomegalovirus, humanpapilloma virus, other agents

Immune dysregulationAutoimmune mechanismVascular lesion/arteriopathyDefective morphogenesis

Inherited mutationsLaterality genes (presumed), associated with polysplenia and

asplenia syndromes (e.g., CFC1)Ductal plate malformation (e.g., HNF6)Jagged 1Other genes

Somatic mutationsModifier genes

Toxin exposure



from 55% of patients with the acquired/perinatal form ofbiliary atresia and in only 8% to 15% of autopsy controlsamples and infants aged less than 1 year with other liverdiseases. Reovirus RNA was not detected in three pa-tients with the fetal form of biliary atresia. The discrep-ancies between these two studies may lie in the methodsof preparation of the tissue, different methods of RNAisolation, and the use of PCR primers for different reo-virus genes. If reovirus is shown to be involved, potentialantiviral strategies may be entertained. In the future, thepotential role of this virus can only be definitively evalu-ated using large numbers of well-characterized patientsand appropriate disease and healthy control subjects.

Recent interest has also focused on group C rotavirus(another virus of the Reoviridae family) in the etiologyof biliary atresia. Group A rotavirus infection was shownto produce extrahepatic bile duct obstruction in newbornmice with hepatic histology similar to biliary atresia (44).Czech-Schmidt et al. (45) further characterized thismodel and demonstrated that the optimal intraperitonealinfecting dose of rhesus rotavirus (RRV) serogroup 3 innewborn BALB/c mice was 105 to 106 plaque-formingU/mL administered 12 hours after birth, resulting in ap-pearance of cholestasis and biliary atresia in 38% to 86%of infected mice. Intrauterine infection did not cause cho-lestasis or biliary atresia despite transplacental infectionof the fetal mice. Immunity induced by previous infec-tion with RRV in dams appeared to protect subsequentnewborn mice from developing hepatobiliary diseasewhen infected postnatally with RRV, whereas maternalmilk was not protective. Petersen et al. (46) reported thatthe administration of interferon-� before RRV infectionprevented the biliary disease, and Qiao et al. (47) re-ported an increase in the incidence of bile duct obstruc-tion in normal newborn BALB/c mice compared withSCID (immunodeficient) mice infected with rotavirus,indicating that the role of the immune system is criticalin this mouse model. Riepenhoff-Talty et al. (48) exam-ined hepatobiliary tissues from human patients for RT-PCR evidence of group C rotavirus infection. Ten of 18patients with biliary atresia and 0 of 12 liver diseasecontrol subjects showed evidence of rotavirus RNA. Incontrast, Bobo et al. (49) failed to detect RNA evidencefor rotavirus groups A, B, or C in tissues from 10 patientswith biliary atresia and 14 liver disease control patientsusing a nested RT-PCR enzyme immunoassay; however,almost half the patients were aged more than 12 monthswhen tissues were obtained. Thus, there is suggestive,but inconclusive, evidence that rotavirus infection maybe involved in up to 50% of cases of biliary atresia,similar to the prevalence of reovirus infection in thestudy of Tyler et al. (43).

The possible role of other viruses has recently beeninvestigated. Human papilloma virus (HPV) was de-tected using PCR in archived liver tissue from 16 of 18patients with biliary atresia and from 0 control patientsfrom Argentina (50,51). However, Domiati-Saad et al.

(52) failed to demonstrate evidence of HPV DNA in 19patients with biliary atresia or idiopathic neonatal hepa-titis from the United States, although they did detectHHV-6 DNA in several patients with neonatal hepatitisand biliary atresia. The possible role of HPV and HHV-6in biliary atresia is unsettled and requires further inves-tigation.

Finally, Mason et al. (53,54) recently described im-munoreactivity and PCR evidence of retroviral infectionin the liver of patients with primary biliary cirrhosis, andserum immunoreactivity against a retrovirus in a limitednumber of patients with biliary atresia. They attributedthis finding to an autoimmune response to antigenicallyrelated cellular proteins or to an immune response touncharacterized viral proteins. Further work in this po-tentially important area of adult and pediatric biliary dis-orders is warranted.

Immune injury in biliary atresia

Schreiber and Kleinman (55) proposed that biliaryatresia was the result of a “multihit” pathologic process,in which a viral or toxic insult to biliary epithelium leadsto newly expressed or altered antigens on the surface ofbile duct epithelia, which, in the proper genetically de-termined immunologic milieu, are presented by macro-phages to T lymphocytes. Cytotoxic T cells then elicit aTH1 cellular response causing bile duct epithelial injury,eventually resulting in fibrosis and occlusion of the ex-trahepatic bile duct (Fig. 2). Unique aspects of innate andacquired immunity that are present in neonates may alsoplay an important role in determining why this disorderonly occurs within the first 3 months of life and in apresumably small percentage of infants infected with theputative agent. In addition, passively acquired maternalfactors could potentially affect presentation and immunerecognition of antigens and T-cell activation in neonates,causing liver injury as it does in the neonatal lupus syn-drome (56).

Genes that encode a variety of immune regulatory pro-teins, in part, control the susceptibility of immune orautoimmune injury to biliary epithelia. A number ofimmune-mediated liver diseases, including autoimmunehepatitis, primary sclerosing cholangitis, and primarybiliary cirrhosis, are associated with specific HLA anti-gens (57). For this reason, several investigators havesought HLA associations in biliary atresia. Silviera et al.(58) initially reported that HLA-B12(49% biliary atresiapatients vs. 23% control subjects) and haplotypes A9-B5and A28-B35 were associated with biliary atresia in aEuropean population. Other groups could not replicatethese findings but reported a relationship of biliary atre-sia with HLACw4/7 (59) and, in Japan, with A33, B44,and DR6 (60). However, a Spanish group of investigatorscould not detect any HLA association in 48 patients withbiliary atresia (61). More recently, Donaldson et al. (62)

R. J. SOKOL ET AL.8


reported molecular genotyping of 101 European childrenwith biliary atresia for selected HLA-A, HLA-B, DRB1,DQA1, DQB1, and DPB1 alleles that had previouslybeen implicated in the pathogenesis of biliary atresia(58). No relationship could be demonstrated betweenthese alleles and biliary atresia, with a statistical powerof 75% at a type 1 error rate of 5%. Thus, a weak asso-ciation may not have been detected by this study. Itshould also be noted that the complete MHC I and IIgenomes were not investigated in this study (there aremore than 100 HLA genes and 1,000 sequenced alleles inthe human MHC genome); therefore, it is possible thatother HLA relationships may be present. Most recently,A-Kader et al. (63) reported a significantly increasedfrequency of HLA-B8 (83% patients vs. 6.5% of generalpopulation) and DR3 (94% patients vs. 15%) in 18 Egyp-

tian children with biliary atresia, 10 of whom harboredthe B8/DR3 haplotype. These findings are of particularinterest because the HLA-B8 and DR3 haplotypes arefrequently found in patients with primary sclerosingcholangitis and inflammatory bowel disease (64). Thehypothesis that MHC class I and class II genotypes maypredispose persons to biliary atresia or idiopathic neona-tal hepatitis needs to be investigated in larger multiethniccohorts of patients.

There is substantial evidence to support the role of acytotoxic T-cell response in the pathogenesis of biliaryatresia. In 1977, Gosseye et al. (65) demonstrated lym-phocytes in the connective tissue of the portahepatis inpatients with biliary atresia, and Bill et al. (66) noted therelationship of intramural mononuclear inflammatorycells with epithelial cell necrosis in bile duct remnants. In

FIG. 2. Proposed model of viral-induced autoreactivity leading to T-cell–mediated destruction of bile duct epithelium in patients withbiliary atresia. The initial injury to the bile duct epithelial cells (BDEC) from a virus (or other potential insult) leads to new appearance ofpreviously sequestered self-antigens or self-antigens altered by caspases induced during BDEC apoptosis, inciting further ductal damagethrough autoreactive T-lymphocyte–mediated inflammation. Initial T-cell activation specific to the viral antigen leads to interferon-�stimulation of macrophages with release of nitric oxide, reactive oxygen species, and tumor necrosis factor and subsequent epithelial celldeath through apoptotic or necrotic pathways. Previously sequestered or altered bile duct epithelial antigens released from this initialinjury are now presented to autoreactive T cells, causing further activation of this immune cascade and progressive destruction of bile ductepithelium. T = T cell; M� = macrophage; NO = nitric oxide; O2 = reactive oxygen species; TNF = tumor necrosis factor-�. Used withpermission: Sokol RJ, Mack C. Etiopathogenesis of biliary atresia. Semin Liver Dis 2001;21:517–24.



1995, Ohya et al. (67) further showed that degenerationof intrahepatic bile ducts was associated with lympho-cytic infiltration into bile duct epithelial cells of patientswith biliary atresia at the time of diagnosis.

For T-cells to effectively mediate inflammation, theymust encounter antigen presented by a competentantigen-presenting cell (APC). Two signals are requiredfor full T-cell activation from the APCs, including sur-face expression of self-MHC molecules bearing the an-tigenic peptide, which interacts with the T-cell receptor,and also costimulatory molecules (B7–1, B7–2) that in-teract with CD28 on the T cell (68). Adhesion of APCswith T cells also requires the expression of intracellularadhesion molecules (ICAMs). Helper T cells (CD4+) rec-ognize antigenic peptides in the context of self-MHCclass II expression, and cytotoxic T cells (CD8+) recog-nize antigen in the context of self-MHC class I mol-ecules. Based on this paradigm, several investigatorshave proposed that bile duct epithelial cells may functionas APCs in patients with biliary atresia. Normally, MHCclass I antigens, but not class II antigens, are expressedby bile duct epithelium. However, several groups (60,61,69) showed that HLA-DR (MHC class II molecules) wasaberrantly expressed by bile duct epithelium in liverspecimens from patients with biliary atresia. Davenportet al. (70) further demonstrated that CD4+ lymphocytesand natural killer (CD56+) cells predominated in the liverand extrahepatic bile duct of patients with biliary atresia,that the cellular infiltrate was activated and proliferating,and that ICAM-1 was expressed in sinusoidal endothe-lium. These data are consistent with the hypothesis thatlymphocyte adhesion and T-cell activation and cytotox-icity, at least in part, mediate the extrahepatic bile ductdamage and obliteration in patients with biliary atresia.

The Kupffer cell (resident liver macrophage) alsofunctions as an APC in the liver. A recent study fromJapan demonstrated increased numbers and size ofKupffer cells in liver tissue from patients with biliaryatresia at the time of diagnosis (71). Davenport (70) alsoshowed that an increase in CD68+ macrophage infiltra-tion in portal tracts and biliary remnant tissue from pa-tients with biliary atresia predicted a poor outcome afterthe portoenterostomy procedure, consistent with thefunction of activated macrophages to release cytokines,reactive oxygen intermediates, and growth factors thatmay signal hepatic stellate cells to synthesize and secretecollagen, thereby promoting fibrogenesis and cirrhosis.One other important feature of macrophages is the ca-pacity to secrete tumor necrosis factor-� (TNF-�), reac-tive oxygen species, and nitric oxide, which may be in-volved in the induction of apoptotic and necrotic intra-cellular pathways. Along these lines, Funaki et al. (72)have shown that apoptosis of intrahepatic bile duct epi-thelial cells is highly prevalent in the liver of patientswith biliary atresia compared with normal liver or thatfrom patients with choledochal cyst. Moreover, Liu et al.(73) reported a relationship between Fas ligand (FasL)

mRNA expression in bile duct epithelia and the presenceof apoptosis in patients with biliary atresia. BecauseFasL is not normally expressed in bile duct epithelialcells, this finding appeared to be specific to biliary atre-sia. Bile drainage after the portoenterostomy procedurewas significantly better in patients with negative signalsfor FasL on bile duct epithelium than in patients withpositive signals (73). This observation suggests that up-regulation of FasL may result in apoptotic fratricide inwhich bile duct epithelial cells actually injure other simi-lar cells, or perhaps bile duct epithelia are resisting attackfrom infiltrating lymphocytes by posing a counterattackagainst Fas-expressing lymphocytes (68).

Bezerra et al. (74) took a different approach to inves-tigating the immune pathogenesis of biliary atresia. DNAmicroarrays, using the Affymetrix U95Av2 gene micro-chip (Santa Clara, CA) containing 12,651 gene products,were assessed on liver biopsies obtained from 14 infantswith biliary atresia and compared with 6 taken from pa-tients with intrahepatic neonatal cholestasis. A TH1 im-mune response in the patients with biliary atresia wassuggested by the upregulation of a number of genes en-coding products involved in lymphocyte differentiation,and several that regulate the TH1 response (osteopontin,�-interferon) combined with downregulation of genesencoding immunoglobulin domains, consistent with sup-pression of a TH2 response. These provocative resultsemphasize complex interactions among macrophages, Tlymphocytes, bile duct epithelial cells, hepatocytes, andother cells in the liver that should form the basis forfuture investigations.

Autoimmunity in biliary atresia

Biliary atresia shares features with several autoim-mune diseases, such as the female predominance, appar-ent triggering by viral infection, and aberrant MHC ex-pression in bile duct epithelium. Consequently, it hasbeen proposed that tissue injury in patients with biliaryatresia may represent an autoimmune-mediated process.Two studies have examined circulating autoantibodies inpatients with biliary atresia. Vasiliauskas et al. (75) havereported that 10 of 11 patients with biliary atresia werepositive for serum IgG and IgM antineutrophil cytoplas-mic antibodies (ANCA), with higher levels of the IgM-ANCA in patients with biliary atresia compared withchildren and adults with other liver diseases. At our cen-ter, Burch et al. (56) studied autoantibodies in mothers ofchildren with biliary atresia and idiopathic neonatalhepatitis to test the hypothesis that maternal transfer ofautoantibodies might be involved in liver and bile ductinjury. Low titer anti-Rho antibodies were more preva-lent in mothers of infants with biliary atresia and idio-pathic neonatal hepatitis than in control subjects, and lowtiter antinuclear antibodies were more common in moth-ers of infants with liver disease. The possible influence

R. J. SOKOL ET AL.10


of maternal factors on immune-mediated bile duct injuryis a provocative, yet unconfirmed, observation.

An exciting advance in understanding risk factors forautoimmunity has been the demonstration of polymor-phisms in genes that predict susceptibility of persons toautoimmune disorders. Recent reports of Bernal et al.(76) and Mitchell et al. (77) have shown that TNF-� genepolymorphisms are associated with susceptibility to pri-mary sclerosing cholangitis (PSC; 58% of PSC patientsvs. 29% of control subjects in Bernal’s study). Polymor-phisms in the IL-1 gene family have recently been asso-ciated with susceptibility and disease progression in pa-tients with primary biliary cirrhosis (78). Similarly,IL-10 polymorphisms have been linked to disease pro-gression in patients with hepatitis C viral and alcoholichepatitis (79,80). These studies raised the possibility thatpolymorphisms in genes that regulate immune functionand the inflammatory response could potentially predis-pose persons to biliary atresia. In this regard, Davidson etal. (62) have recently examined the prevalence of patho-logic polymorphisms in the promoter region of theTNF-� gene, two IL-1 genes (ILIB and IL1RN), andthree sites in the promoter region of the immunoregula-tory gene, IL-10 gene, in genomic DNA of 101 childrenwith biliary atresia in the United Kingdom comparedwith 96 geographically and racially matched adult con-trol subjects. There were no statistical differences in thedistribution of alleles, genotypes, or haplotypes for anyof these genes between the patients with biliary atresiaand control subjects, making these unlikely candidategenes for predisposing persons to biliary atresia. How-ever, there are certain to be other immunoregulatorygenes that will be found to play a role in biliary tractinjury and which will be candidates for future investiga-tion. Extending beyond immunoregulatory genes, it maywell be that genes that regulate metabolism and canalic-ular transport of bile acids and phospholipids, apoptosis,or cell survival could play a potential role in protectionand recovery from bile duct injury in patients with biliaryatresia.

Vascular etiology

A vascular/ischemia etiology for biliary atresia hasbeen proposed based on direct experimental evidence(81). Intrahepatic and extrahepatic bile ducts receivetheir blood supply exclusively from the hepatic arterialcirculation; thus, hepatic arterial ischemia causes bileduct strictures, as observed after liver transplantation.Pickett et al. (82) demonstrated the development of bil-iary obstruction after ligation of hepatic vessels in fetalsheep. Several investigators have demonstrated an arte-riopathy in branches of the hepatic artery in the extrahe-patic biliary tree of patients with biliary atresia (83). Ithas been proposed that the vasculopathy may be the pri-mary lesion in patients with biliary atresia; however, the

causative or consequential nature of these lesions re-mains unclear.

Defective morphogenesis

Several lines of evidence suggest that certain cases ofbiliary atresia (fetal form) are caused by defective mor-phogenesis of the biliary tree. Because anomalies of vis-ceral organ symmetry (the polysplenia syndrome) areassociated with biliary atresia, it is of interest that a re-cessive insertional mutation in the proximal region ofmouse chromosome 4 or complete deletion of the inver-sion (inv) gene in the mouse leads to anomalous devel-opment of the hepatobiliary system in this model (11,84).Schon et al. (85) recently mapped the human INV geneand examined 64 patients with heteroaxia. No consistentmutations in this gene were identified; specifically, therewere no mutations in seven patients with biliary atresiaand various congenital laterality defects, making it un-likely that the INV gene is involved in the majority of“fetal” cases of biliary atresia.

Other human genes that determine laterality (ZIC3,LEFTB, and ACVR2B) have been associated with a smallpercentage of human situs defects. Bamford et al. (86)recently searched for mutations of the CFC1 gene (whichencodes the human CRYPTIC protein) in genomic DNAfrom 144 patients with familial and sporadic cases oflaterality defects. Nine of the patients had heterozygousmutations of this gene, including one with biliary atresiaand the polysplenia syndrome. Jacquemin et al. (87) ex-tended this observation by identifying heterozygous genemutations in two brothers with laterality defects, onewho had biliary atresia, which were inherited from theirhealthy mother. It appears that these heterozygous mu-tations of CFC1 may not be solely responsible for thisphenotype but may provide for a predisposition that thenrequires a second genetic or environmental factor to pro-duce the disease phenotype. Although the precise func-tion of CRYPTIC protein is not known, it appears to actas a cofactor in the Nodal pathway that determines left–right axis development. Further investigations of this andother related genes that determine laterality may shedlight in the coming years on whether inherited or somaticmutations or deletions are responsible for individualcases of biliary atresia.

Intrahepatic bile duct development depends on inter-actions between mesenchyme and portal venous radicals.Primitive hepatic precursor cells differentiate into asingle layer of cells that soon form a double layer (theductal plate) as the primitive intrahepatic bile ductuleanlage. Cells then scatter and remodel as a single layeraround the lumen to form the portal tract bile ducts (12).Abnormal remodeling leads to the ductal plate malfor-mation that is believed to be responsible for the liverlesion of congenital hepatic fibrosis and other bile ductdysplasias. However, a number of infants with biliary



atresia appear to show evidence of the ductal plate mal-formation on liver biopsy (12), suggesting that interac-tions between hepatocyte growth factor or scatter factorand receptors such as the c-met oncogene may be defec-tive in patients with biliary atresia and ductal platemalformation (1,88). Abnormalities in induction ofhepatocyte growth factor during a critical period formesenchymal/epithelial signaling or other defects in in-tracellular adhesion systems could account for defectivebile duct development in patients with biliary atresia,although there is no direct proof of these mechanisms.

The roles of JAGGED1, which is expressed in ductalplate epithelia (89), and HNF6, a nuclear transcriptionfactor expressed in the developing extrahepatic bile ductand gall bladder (90), are being explored in patients withbiliary atresia and paucity of interlobular bile duct dis-orders. Recently, Kohsaka et al. (89) identifiedJAGGED1 missense mutations in 9 of 102 patients withbiliary atresia. These patients had no phenotypic featuresof Alagille syndrome. Remarkably, prognosis was worsefor the group with mutations, suggesting that JAGGED1could be a modifying factor in patients with biliary atre-sia. Alternatively, it is possible that certain patients withbiliary atresia have a unique new presentation of Alagillesyndrome limited to the biliary tree. The hnf6 knockoutmouse or lack of hnf1ß in the mouse causes a ductal platemalformation associated with anomalies of the hepaticartery and its branches (90), demonstrating the intricateinteractions between vascular structures and intrahepaticbile duct development.

Toxin exposure

Time and space clustering of cases of biliary atresiahave led to the proposal that an environmental toxincould be involved in its pathogenesis. Currently, otherthan infectious agents, no environmental agent has beenclearly associated with biliary atresia or idiopathic neo-natal hepatitis in humans. Two outbreaks of biliary atre-sia in lambs and calves in Australia may have been re-lated to a fungal or other environmental toxin exposure(91). There are also possible toxic and inflammatory ef-fects of bile that has extravasated through the injured bileduct epithelial layer into the submucosa of the damagedextrahepatic bile ducts.

DIAGNOSIS OF NEONATALCHOLESTATIC DISORDERS

It is important to identify the small subgroup of infantswith jaundice who have direct (conjugated) hyperbiliru-binemia and to establish the underlying diagnosis as soonas possible to afford the benefit of specific medical thera-pies and to allow for optimal outcome after portoenter-ostomy for biliary atresia. For this reason, it is recom-mended that a fractionated bilirubin be obtained for all

infants who remain visibly jaundiced beyond 2 to 3weeks of age, or who develop acholic stools or hepato-megaly. Clinical differentiation among the commoncauses of cholestasis is limited. In both biliary atresia andidiopathic neonatal hepatitis, infants present with jaun-dice in the first 8 weeks of life, progressive loss of pig-mentation in their stools, and development of hepato-megaly and splenomegaly (17). Biliary atresia is morecommonly found in infant girls who were appropriate forgestational age at birth and appear to be thriving,whereas idiopathic neonatal hepatitis is more common ininfant boys who were small for gestational age and arefailing to thrive. The most accurate diagnostic test fordifferentiating biliary atresia from other conditions ispercutaneous liver biopsy. When obtained before lapa-rotomy, liver biopsy has a diagnostic accuracy for biliaryatresia when examined by experienced pathologists of90% to 95% if an adequate biopsy size is obtained(92,93), and will prompt urgent surgical exploration ifcharacteristic findings are present. Liver biopsy samplesfrom patients with biliary atresia generally show bileductular proliferation, canalicular and cellular bile stasis,portal or periportal inflammation, and fibrosis with thepresence of bile plugs in portal tract bile ducts (Fig. 3)(9,92). Hepatocyte giant cell transformation is found inat least 25% of patients with biliary atresia, particularlyif the biopsy is obtained during the first 6 weeks of life.It is important to note that liver histology in patients with�-1-antitrypsin deficiency, and occasionally in thosewith Alagille syndrome (94), cystic fibrosis, and totalparenteral nutrition (TPN)-related cholestasis, can re-semble all the features of biliary atresia, necessitating theuse of other diagnostic studies to differentiate these dis-orders. Liver biopsy specimens from patients withidiopathic neonatal hepatitis show lobular disarray, avariable inflammatory infiltrate with marked giant celltransformation of individual hepatocytes, individual he-patocyte necrosis and apoptosis, increased extramedul-lary hematopoiesis, and cellular bile stasis. However,bile plugs in portal tract bile ducts are absent, and bileductular proliferation is usually minimal or absent. Portaltract fibrosis is occasionally found but is not extensive(9). These histologic features, although characteristic, arenot specific to idiopathic neonatal hepatitis and havebeen observed in patients with PFIC, bile acid synthesisdisorders, �-1-antitrypsin deficiency, and other condi-tions.

Other diagnostic studies are less accurate in differen-tiating biliary atresia from intrahepatic causes of chole-stasis. No serum or urine biochemical tests differentiatebetween these disorders. Hepatobiliary scintigraphy is oflimited value. Imaging studies are inconclusive. Failureof isotope excretion into the small intestine during scin-tigraphy (iminodiacetic acid derivatives) has only 50% to75% specificity for biliary atresia despite more than 95%sensitivity (95). Severe intrahepatic cholestasis and pau-city of interlobular bile duct disorders may yield scinti-



graphic findings indistinguishable from biliary atresia.However, clear demonstration of secretion of the radio-isotope into the small bowel essentially excludes biliaryatresia. Likewise, ultrasonography of the liver and biliarytree is useful but not diagnostic for biliary atresia. Forexample, ultrasonography may identify a small, nondis-tended gallbladder (suggesting biliary atresia) if severeintrahepatic cholestasis is present and, conversely, mayshow a clear fluid-filled gallbladder remnant in a patientwith biliary atresia that is indistinguishable from normal(96). This modality is also not sensitive enough to de-termine presence or absence of the common hepatic andcommon bile ducts in small infants. However, ultraso-nography is absolutely essential for the diagnostic evalu-ation of infants with cholestasis to exclude the presenceof a choledochal cyst and to determine the presence offeatures of polysplenia syndrome, which is strongly as-sociated with biliary atresia.

Newer ultrasonographic techniques have led to the de-scription of two novel findings that may prove to bemore diagnostic. Choi et al. (97) reported that a unique

triangular or tubular echogenic density (the “triangularcord” sign), representing the fibrous cone of the bile ductremnant at the hepatic porta, is a specific ultrasonograph-ic finding for biliary atresia. Another recent study sug-gests that using a higher frequency (13 MHz rather than7 MHz) ultrasound transducer may identify abnormali-ties in gall bladder shape, wall thickness, and morphol-ogy that are characteristic of biliary atresia (96). Theusefulness of these sonographic findings for evaluatinginfants with cholestasis will require further study at othercenters. Early studies suggested that magnetic resonancecholangiopancreatography (MRCP) using T2-weightedturbo spin-echo sequences might hold promise as a non-invasive method for diagnosis of biliary atresia (98).However, Norton et al. (99) recently demonstrated limi-tations of MRCP for differentiation of severe intrahe-patic cholestasis from biliary atresia because the abilityto identify extrahepatic bile ducts depends on bile flow.The use of endoscopic retrograde cholangiopancreatog-raphy (ERCP) has been proposed for identification of theextrahepatic biliary tree, although it requires consider-

FIG. 3. Histology of liver and bile duct in biliary atresia. (A) Hepatocellular and canalicular cholestasis, swollen hepatocytes, multinu-cleated giant cells (arrow), and portal tract inflammation (hematoxylin and eosin, ×400) (B) Expanded portal tracts with increased fibrosis,bile duct proliferation (line arrows), and bile plug in portal tract bile duct (block arrow) Trichrome, ×250). (C) Proximal common hepaticduct in portal hepatis with apoptosis and disruption of biliary epithelium lining a narrowed but patent lumen, lymphocytic infiltration aroundduct, and increased concentric sclerosis (hematoxylin and eosin, ×100). (D) Remnant of common bile duct with absent lumen (arrow) andconcentric sclerosis (hematoxylin and eosin, ×40).



able technical expertise and general anesthesia, and theproper sized side-viewing endoscope is not widely avail-able (100,101). Finally, several groups have reported theusefulness of analyzing a duodenal bile specimen forpresence of bilirubin or bile acids (102,103) as a test todemonstrate patency of the extrahepatic biliary tree.

Delay in diagnosis of biliary atresia is a considerableproblem in the United States because neonatal jaundicemay be incorrectly ascribed to “breast milk jaundice;”infants with biliary atresia may have only mild jaundiceand not be thoroughly investigated; infants are generallyseen by health care providers only at age 2 and 8 weeks;and physician extenders unaware of these rare diseasesmay be providing the patient care. Thus, an infant aged2 weeks with jaundice may not be seen again by a healthcare provider until after age 8 weeks, already the agelimit for optimal results of portoenterostomy. To helpreduce the average age for diagnosis of biliary atresia,and hence improve outcome from the portoenterostomyprocedure, several groups in Japan and Taiwan have de-veloped pilot programs in which stool color cards aregiven to mothers of newborns (104). These cards haveseven or eight numbered color photographs of age-matched infant stools of varying colors, including threeacholic stools. When infants are aged 1 month, the moth-ers are instructed to compare their infant’s stool colorwith those printed on the card, to fill in a correspondingnumber, and to take the card to their 1-month physicianoffice visit. If the mother and the physician identifiedthe child’s stool color as one of the acholic exampleson the card, the physician calls or sends a fax to thescreening center. The mother and an informed pediat-ric specialist are contacted, and the child is evaluatedfor cholestasis and biliary atresia. The results of large-scale screening using this method appear promising(104).

The diagnosis of biliary atresia is established aftertimely exclusion of intrahepatic (infectious, metabolic,genetic, and toxic) causes of cholestasis and choledochalcyst (by ultrasonography) (Table 1). In North America,biliary atresia is then generally diagnosed at minilapa-rotomy by intraoperative cholangiography that fails todemonstrate a lumen in some portion of the extrahepaticbiliary tree, by surgical findings of a fibrotic, nonpatentextrahepatic bile duct, and by characteristic findings onliver and bile duct histology, in the absence of otherknown etiologies (17). In some countries, percutaneoustranshepatic cholecystography (105) or ERCP (101) arealternative tests to define the anatomy of the extrahepaticbiliary tree. Caution must be entertained in assigning thediagnosis of biliary atresia and performing the portoen-terostomy if the only findings are failure to visualize thecommon hepatic duct and intrahepatic ducts on cholan-giography because these findings may represent hypo-plastic but patent proximal biliary structures in thosewith Alagille syndrome (106).

The diagnosis of idiopathic neonatal hepatitis is as-signed only after infectious, metabolic, genetic, andstructural causes of “giant cell hepatitis” are excluded(17). Therefore, as newer etiologies for intrahepatic cho-lestasis are discovered, infants thought previously tohave “idiopathic” neonatal hepatitis may have their di-agnosis reassigned.

TREATMENT OF BILIARY ATRESIA ANDNEONATAL HEPATITIS

Optimal therapy for patients with biliary atresiadiagnosed before age 12 weeks is the surgical por-toenterostomy, in which a Roux-en-Y loop of jejunum isconnected by anastomosis to the porta of the liver aftercareful surgical dissection to locate patent bile ductremnants in the porta (6,107). If performed within thefirst 60 days of life by experienced pediatric surgeons,the portoenterostomy should yield bile drainage from theliver into the intestinal tract in at least 70% to 80% ofpatients, resulting in increased pigmentation of the stoolsand resolution of jaundice (6,92,107,108). If performedbetween 60 and 90 days of life, approximately 40% to50% of patients show bile drainage; if performed after 90days of life, up to 25% show drainage (108); and ifperformed after 120 days of life, only 10% to 20% ofpatients, at best, show evidence of bile drainage. Thus,many surgeons will not perform the portoenterostomy ininfants with biliary atresia diagnosed after age 3 to 4months (109). Consequently, it is absolutely essentialthat infants who remain jaundiced after age 2 to 3 weeksbe evaluated expediently for cholestasis and, if pres-ent, for biliary atresia. Prompt surgical exploration andintraoperative cholangiogram must be performed if bili-ary atresia cannot be excluded using other diagnostictests.

There is no standardized protocol for postoperativetreatment of patients with biliary atresia in the UnitedStates. Antibiotic prophylaxis of cholangitis (trimethap-rin–sulfamethoxasole or neomycin (110)), use of shortcourses of corticosteroid pulse therapy to manage refrac-tory cholangitis (111), empiric use of oral ursodeoxycho-lic acid (UDCA) to stimulate bile flow and as a cytopro-tective agent (112), optimization of nutrition (use of amedium chain triglyceride-containing infant formula),and prevention of fat-soluble vitamin deficiencies (113)are frequently used; however, there is no uniformity inpractice (19,92). Supplementation with infant formulahigh in branched-chain amino acids, available in Europeand Australia, was well tolerated and associated withimproved growth in one study (113). Patients with failureto thrive awaiting liver transplantation may requiresupplemental nasogastric tube feedings or parenteral ali-mentation.



Postoperatively, ascending cholangitis and sclerosis ofpatent intrahepatic bile ducts occur in 40% to 60% ofpatients and may lead to progressive biliary cirrhosis andliver failure (114). Because of the inflammatory compo-nent of cholangitis and the possible immune mechanismsinvolved in the pathogenesis of biliary atresia, anti-inflammatory therapy (e.g., corticosteroids) after porto-enterostomy to prevent cholangitis and reduce intra-hepatic bile duct injury and fibrosis could theoreticallybe of potential benefit. Unfortunately, there are no pub-lished randomized controlled trials of corticosteroidtreatment for patients with biliary atresia. Short-term(1–2 weeks) postoperative therapy has been used inmany centers (115), and an 8- to 10-week course ofcorticosteroids after portoenterostomy appeared to im-prove clinical outcome compared with historical controlsubjects in a retrospective review (116). Many centers inJapan use intravenous and oral corticosteroid therapyalong with antibiotic and bile acid (dehydrocholic acidintravenously and UDCA orally) therapy for up to sev-eral months after portoenterostomy (6). However, with-out a prospective randomized controlled trial of cortico-steroid therapy after portoenterostomy, it is not possibleto recommend routine treatment with this agent at thistime. One additional therapy used in some Asian coun-tries is the traditional Chinese medicine, kanzou (licoriceroot, glycyrrhizic acid), a hepatoprotective and cell-proliferative agent (117).

Liver transplantation can be life saving and is indi-cated for patients with biliary atresia who do not undergoan attempt at portoenterostomy because of delayed diag-nosis, those in whom portoenterostomy has failed to re-establish hepatic-to-intestinal bile flow, and those withdecompensated cirrhosis and end-stage liver disease de-spite initial “success” of portoenterostomy. Long-termsurvival after liver transplantation for biliary atresia ap-proaches 80% to 90% (92).

The treatment for patients with idiopathic neonatalhepatitis is largely supportive, with optimization of nu-trition, prevention of vitamin deficiencies, and use ofcholeretic agents and antipruritic agents (19) as needed.Therefore, infant formulas containing medium-chain tri-glyceride oil for improved fat absorption and adequateamounts of long-chain fat to provide for essential fattyacids are preferred; fat-soluble vitamin supplements aregiven; and oral UDCA or cholestyramine is used to in-duce choleresis. In up to 20% of those with idiopathicneonatal hepatitis, patients will show progression to cir-rhosis and chronic liver failure and may require livertransplantation. In recent years, many of these patientshave been found to harbor forms of PFIC (Byler diseaseand syndrome; PFIC types 1 and 2), MDR3 deficiency(PFIC type 3), bile acid synthesis defects, or neonataliron storage disease (118). Patients with PFIC types 1and 2 may benefit from partial external biliary diversionwith improvement in pruritus, liver function, growth, andliver histology (119–121).

OUTCOME AND COMPLICATIONS

Outcome

If the portoenterostomy is not performed in patientswith biliary atresia, 50% to 80% of children will die(without liver transplantation) from biliary cirrhosis byage 1 year, and 90% to 100% will die by age 3 years(109,122). Complications of portal hypertension, malnu-trition, fat-soluble vitamin deficiencies, pruritus, andcholestasis add to the misery of these children beforedeath. Successful portoenterostomy, when performed be-fore 60 to 90 days of age, is associated with approxi-mately a 30% to 40% 10-year survival at North Ameri-can and European centers with the most experience(122–124), whereas 10-year survival after portoenteros-tomy performed before 60 days of age in Japan mayapproach 60% (Table 3) (6). Thus, the age at referral ofthe patient for evaluation is one of the most importantfactors determining surgical outcome. If the portoenter-ostomy is not successful in establishing bile flow, sur-vival without transplantation is similar to or worse thanthat of patients not undergoing surgery. Postoperativecare after portoenterostomy differs in Japan (6) from thatused in the United States, as described previously. It isnot clear if these differences in management or otherfactors (e.g., genetic or surgical technique) are respon-sible for the generally improved prognosis in Japan.

A large multicenter review of the outcome of all chil-dren diagnosed with biliary atresia in France between1986 and 1996 was recently published (108). Combiningthe results of portoenterostomy with liver transplanta-tion, 10-year overall survival for 472 patients with biliaryatresia was 68%. The 10-year actuarial survival with thenative liver after portoenterostomy was 29%, and 5-yearsurvival after liver transplantation was 71%. Prognosticfactors predictive of overall 10-year survival were theperformance of the portoenterostomy, age at portoenter-ostomy (survival of 80.4% with surgery at age <45 daysvs. 68.5% at >45 days), anatomic pattern of atresia(100% for atresia of the common bile duct only vs.65.4% for complete extrahepatic atresia), presence ofpolysplenia syndrome (48.3% for yes vs. 69.9% for no),

TABLE 3. Transplant-free survival vs. age atportoenterostomy for biliary atresia at Tohoku

University Hospital

Age at operation(days) # Cases

10 yr. survival# (%)

1–60 63 36 (57%)61–70 49 20 (41%)71–90 61 18 (30%)>90 78 10 (13%)

Total 251 84 (33%)

Used with permission: Ohi R. Biliary Atresia: a surgical perspective.Clin Liv Dis 2000;4:779–804.



and experience of the center performing the portoenter-ostomy (54% for �2 new patients per year, 59.8% for3–5 per year, and 77.8% for >20 per year). The samefactors predicted 5- and 10-year survival with the nativeliver after portoenterostomy. Improved survival with thenative liver when portoenterostomy was performed atcenters caring for more than five new patients per yearwas also reported from the United Kingdom and Ireland(125), resulting in a change in national policy that re-stricted performance of the portoenterostomy to the threemost-experienced centers in the United Kingdom. Thus,the experience of the center performing the portoenter-ostomy is one of the most important factors determiningsurgical outcome.

The suggestion of delayed neurodevelopmental func-tion in children with biliary atresia was put forth byStewart et al. (126), who showed that children with end-stage liver disease with onset during the first year of lifeshowed significant delays in verbal performance andfull-scale IQs at a mean age of 8.4 years. They relatedthese changes to nutritional status, severity of liver dys-function, and duration of liver disease, and there waslittle improvement in these parameters 1 year after livertransplantation (127). However, because the patients re-ported included children with other liver diseases asso-ciated with delayed development (e.g., Alagille syn-drome), its relevance to biliary atresia has been ques-tioned. Wayman et al. (128) subsequently reported thatchildren aged less than 1 year with biliary atresia whowere evaluated for liver transplantation had a mean men-tal development score in the low average range and amean psychomotor developmental score 1 SD belownormal (Bayley Scales of Infant Development). Al-though many patients showed improvement 1 year aftertransplantation, 35% continued to be developmentallydelayed. Developmental domains particularly affectedincluded expressive language and independent walking.Factors associated with this persisting delay were de-creased weight before transplantation, low serum albu-min, length of hospitalization for transplantation, andyounger age at transplantation. These data suggest thatan improved neurodevelopmental outcome may beachieved using a more aggressive approach to nutritionalrehabilitation before liver transplantation, attention tomicronutrient supplementation, and transplantation be-fore the onset of significant malnutrition. However, thedegree to which the developmental delays may be attrib-uted to other factors related to chronic liver failure hasnot been determined.

Early complications

The most important early complication after the por-toenterostomy procedure is the development of ascend-ing bacterial cholangitis. Breach of the barrier to pen-etration of intestinal bacteria into the biliary system is

one of the unavoidable consequences of the portoenter-ostomy because the ampulla of Vater no longer providesthis protection. Early diagnosis and management of as-cending cholangitis is essential to prevent sclerosis andloss of remaining intrahepatic bile ducts, and to preservefunction of the native liver. Cholangitis is suspected aftera successful portoenterostomy if any of the followingdevelop: fever, hypopigmented stools, elevated liverfunction tests (particularly the bilirubin concentration),generalized or right upper quadrant abdominal pain, rightshoulder pain, or sudden increased pruritus. Ascendingcholangitis can develop during or after viral infectionsbecause of poor oral intake causing decreased stimula-tion of bile flow and bile stasis, and because of the in-hibitory effects of circulating cytokines on bile flow.Thus, it may be difficult to differentiate viral causes offever and abnormal liver function tests versus ascendingcholangitis. Particularly during the first 6 to 12 monthsafter portoenterostomy, most fevers with any evidence ofincreased liver dysfunction or reduction in stool pigmen-tation are treated as if cholangitis is present. In our ex-perience, imipenem/cilastin (or meropenem) has been themost effective broad-spectrum antibiotic for manage-ment of cholangitis in the absence of a positive bloodculture. Others have used third-generation cephalospor-ins and aminoglycosides. If fever persists beyond 24 to48 hours of therapy or if liver function tests do not showimprovement, we initiate intravenous corticosteroidpulse therapy for 5 days in addition to the antibiotics.Intravenous methylprednisolone is administered in 1hour as a single dose of 10 mg/kg on day 1, followed by5 mg/kg, 2.5 mg/kg, 1 mg/kg, and 0.5 mg/kg on succes-sive days. If cholangitis does not fully respond to thisregimen, the patient may also be treated with 1 mg/kgoral prednisone for 2 to 4 weeks. Persistent or recurringcholangitis should also prompt evaluation with hepaticscintigraphy and ultrasonography of the Roux-en-Y loopfor kinking or obstruction, which may require surgicalrevision. As a means of reducing ascending cholangitis,some have advocated the surgical construction of an in-tussuscepted valve in the Roux-n-Y loop, although itsbenefit is debatable.

Late complications

The majority of surviving patients with biliary atresiawill develop portal hypertension. Portal hypertensionmay become clinically manifest simply by the finding ofprogressive hepatosplenomegaly with a very firm to hardliver texture, the development of gastrointestinal hemor-rhage from esophageal or gastric varices or portal hyper-tensive gastropathy, hypersplenism, ascites, spontaneousbacterial peritonitis, portosystemic encephalopathy, orportopulmonary syndrome. Significant variceal hemor-rhage has been reported in 20% to 60% of patients withbiliary atresia. In a study from the University of Colo-



rado, 29% of 134 patients developed esophageal varicealhemorrhage (129). By age 5 years, 40% of the patientswho were alive without a liver transplant had at least 1episode of gastrointestinal hemorrhage, increasing to60% by age 10 years. Recent interest has centered onattempts to prevent variceal hemorrhage by pharmaco-logic or endoscopic methods. Prophylactic sclerotherapyof esophageal varices before onset of hemorrhage hasbeen evaluated in one randomized controlled trial (130).Although sclerotherapy significantly reduced the fre-quency of gastrointestinal hemorrhage from esophagealvarices, there was a coincident increase in bleeding fromgastric varices and portal hypertensive gastropathy, suchthat prophylactic sclerotherapy had no effect on the over-all incidence of gastrointestinal hemorrhage or on sur-vival. Another proposed approach to prevent varicealhemorrhage is the prophylactic use of pharmacologicagents (selective �-blockers and vasodilators) to reduceportal venous pressure. This therapy was effective inreducing the frequency of variceal hemorrhage in adultswith cirrhosis (131). However, this has not been studiedin a randomized controlled fashion in children. Therehave been two open-label studies of propranolol for pri-mary and secondary prophylaxis of variceal hemorrhagein children (132,133), showing reasonable safety, butbleeding rates on propranolol ranged from 15% to 53%.Thus, the efficacy and safety of �-blocker therapy inchildren remain to be determined.

Once gastrointestinal hemorrhage occurs, the initialmanagement focuses on establishing hemodynamic sta-bility. Subsequent treatment options center on medicalmanagement with intravenous somatostatin (or itslonger-acting analogs, such as octreotide), vasopressin,or terlipressin and endoscopic management with varicealsclerotherapy or variceal band ligation (134–137). Ifthese methods are not successful in controlling hemor-rhage, a surgical or radiologic portosystemic shunt (vs.liver transplantation) needs to be considered. Since thedevelopment of the technique of radiologic placement ofa transjugular intrahepatic portosystemic shunt (TIPS),this has been the preferred procedure in some centers forpatients weighing more than 12 kg to 15 kg (138,139). Inmost patients with biliary atresia, liver transplantationmay be more appropriate than a surgical portosystemicshunt.

Ascites, portosystemic encephalopathy, portopulmo-nary syndrome, and severe hypersplenism are less com-mon complications resulting from portal hypertension.Their incidence and prevalence have not been well stud-ied in patients with biliary atresia. Management of theascites is directed at salt restriction and diuretic treatmentwith large-volume paracentesis occasionally required.Encephalopathy is best managed with ammonia-bindingagents such as lactulose, protein restriction, and oral an-tibiotics. Rarely, partial splenic embolization (140,141)or TIPS may be indicated for hypersplenism.

Pruritus caused by progressive cholestasis is managedwith UDCA, antihistamines, rifampicin, and occasion-ally opiate antagonists or ultraviolet light exposure, andlocal measures to prevent scratching or skin dryness (19).We generally avoid bile acid-binding resins (e.g., chole-styramine) for pruritus after a portoenterostomy becauseof concern that the resin will enter and potentially ob-struct the Roux-en-Y loop or intrahepatic bile ducts.

Prognosis

Factors shown to predict outcome after portoenteros-tomy include 1) age at operation, 2) experience of thecenter (surgeon), 3) site of atresia of the extrahepatic bileduct, and 4) number and severity of episodes of postop-erative ascending cholangitis. Patients with a patentproximal common hepatic duct but distal atresia farebetter than patients with atresia of the proximal biliarytree extending into the liver (6). Several groups haveattempted to develop prognostic indices to predict sur-vival and timing of liver transplantation for children withbiliary atresia. The proposed indices include serum con-centrations of hyaluronic acid (142) and procollagen IIIpeptide and type IV collagen (143), the hepatic arteryresistance index (144), urinary excretion of D-glucaricacid (145), postoperative bile bilirubin excretion (146),and a predictive model combining serum zinc, copper,zinc sulfate turbidity test, total bilirubin, GGT, and cho-linesterase (147). These indices have not been validatedprospectively in large numbers of patients, and many arenot routinely available; thus, these indices have notproved valuable in clinical practice. Malatack et al. (148)developed a prognostic score for risk of death in pediat-ric patients at referral for liver transplantation that in-cluded serum concentrations of cholesterol, indirect bil-irubin, partial thromboplastin time (PTT), and history ofascites. Although this prognostic index is simple andwell characterized, it has not been validated earlier in thecourse of liver disease for patients with biliary atresiabefore referral for liver transplantation. Recently, the Pe-diatric End-Stage Liver Disease (PELD) scoring systemhas been developed using data accumulated by the Stud-ies for Pediatric Liver Transplantation Research Group(149). Three-month survival of children listed for livertransplantation was predicted by five variables: serumtotal bilirubin, international normalized ratio (INR), al-bumin, growth failure, and age less than 1 year. Althoughthe PELD score was derived in large part from infantsand children with biliary atresia, it has not been validatedfor infants and children before reaching the stage of dis-ease at which they would be evaluated for liver trans-plantation and, thus, may be of limited prognostic valueearlier in the course of biliary atresia. In another study,serum total bilirubin concentration was predictive of sur-vival after the first episode of esophageal variceal hem-orrhage (128).



In the context of current therapeutic options, 70% to80% of patients with biliary atresia in North Americawill require liver transplantation during the first two de-cades of life, despite initial success with portoenteros-tomy (92,122). Consequently, biliary atresia accounts for40% to 50% of all liver transplants performed in children(152). It should be noted that there is no single liverdisease in adults that accounts for as large a proportion ofliver transplants. Factors that determine long-term sur-vival without transplantation have not been carefullyevaluated in North America since the advent of improvednutritional therapies, newer broad-spectrum antibiotics,and corticosteroid regimens. Moreover, quality of lifeoutcome (QOL) measures for patients with biliary atresiaand other cholestatic disorders have not been prospec-tively analyzed in a large enough cohort using age-specific tools that are now available (6,150).

RECOMMENDATIONS

Based on the data reviewed and the overall poor sur-vival without liver transplantation of children with bili-ary atresia in North America, we propose the followingsteps that could be taken to improve the outcome ofpatients with biliary atresia.

1. A major public health campaign to educate primarycare providers to follow all infants with jaundice atthe 2-week well-baby examination (even if breast-fed) and to obtain a total and direct (conjugated) bil-irubin if the jaundice persists when infant is agedmore than 2 to 3 weeks or if the infant has pale stools,dark urine, or hepatomegaly (151).

2. Development of a national, Internet-based registry totrack age at diagnosis of patients with biliary atresiaand patient outcome, and to evaluate factors that in-fluence outcome.

3. Explore the feasibility of piloting a stool color cardprogram for infants at age 1 month (similar to theJapanese and Taiwanese programs) to determine if itwill improve identification of patients with biliaryatresia before age 6 to 8 weeks in a cost-effectivemanner.

4. Referral of patients with suspected biliary atresia, be-fore surgical exploration, to centers with a definedcare team and pediatric surgeons who perform at leastfour or five portoenterostomies per year to improveoutcome after surgery.

5. Prospective controlled trial of corticosteroid therapyafter portoenterostomy in newly diagnosed patientswith biliary atresia to determine the efficacy andsafety of such therapy.

6. Multicentered cooperative studies to investigate pos-sible genetic, infectious, and immune/autoimmunemechanisms involved in the etiology and pathogen-esis of biliary atresia.

7. Continued investigation of the pathogenesis of bileduct injury in infectious and genetic animal models ofbiliary atresia.

The new Biliary Atresia Clinical Research Consortium(BARC), funded for 5 years by the National Institute ofDiabetes, Digestive and Kidney Diseases (NIDDK) ofthe National Institutes of Health, is a network of 9 clini-cal centers and a data coordinating center in the UnitedStates, in cooperation with the Division of Digestive Dis-eases and Nutrition at NIDDK, that will embark on anumber of these initiatives. The goals of BARC are 1) todevelop and implement a clinical database and the col-lection of tissue, serum, and DNA specimens as a meansof investigating the etiology and pathogenesis of biliaryatresia; 2) to develop and initiate clinical trials and out-come studies; and 3) to stimulate others to investigate theetiology, pathogenesis, and management of biliary atre-sia. Through this and other multicentered collaborativegroups, it is anticipated that major advances in our un-derstanding of biliary atresia and related diseases willtake place over the next 5 years, resulting in improvedhealth and outcomes for children with biliary atresia.

Acknowledgments: The authors thank Dr. Gail Deutsch forproviding the photomicrographs and Cynthia Wyman for sec-retarial assistance.

REFERENCES

1. Balistreri WF, Grand R, Hoofnagle JH, et al. Biliary atresia:current concepts and research directions. Summary of a sympo-sium. Hepatology 1996;23:1682–92.

2. Bucuvalas JC, Ryckman FC, Atherton H, et al. Predictors of costof liver transplantation in children: a single center study. J Pedi-atr 2001;139:66–74.

3. Yoon PW, Bresee JS, Olney RS, et al. Epidemiology of biliaryatresia: a population-based study. Pediatrics 1997;99:376–82.

4. Fischler B, Haglund B, Hjern A. A population-based study on theincidence and possible pre- and perinatal etiologic risk factors ofbiliary atresia. J Pediatr 2002;141:217–22.

5. Sokol RJ, Mack C. Etiopathogenesis of biliary atresia. SeminLiver Dis 2001;21:517–24.

6. Ohi R. Biliary atresia. A surgical perspective. Clin Liver Dis2000;4:779–804.

7. Davenport M, Savage M, Mowat AP, et al. Biliary atresia splenicmalformation syndrome: an etiologic and prognostic subgroup.Surgery 1993;113:662–8.

8. Karrer FM, Hall RJ, Lilly JR. Biliary atresia and the polyspleniasyndrome. J Pediatr Surg 1991;26:524–7.

9. Desmet V. The cholangiopathies. In: Suchy FJ, ed. Liver Diseasein Children. Philadelphia: Lippincott, Williams & Wilkins;2001:39–62.

10. Schweizer P. Treatment of extrahepatic biliary atresia: results andlong-term prognosis after hepatic portoenterostomy. Pediatr Surg1986;1:30–6.

11. Mazziotti MV, Willis LK, Heuckeroth RO, et al. Anomalousdevelopment of the hepatobiliary system in the Inv mouse. Hep-atology 1999;30:372–8.

12. Desmet VJ. Congenital diseases of intrahepatic bile ducts: varia-tions on the theme “ductal plate malformation.” Hepatology1992;16:1069–83.

13. Balistreri WF. Liver disease in infancy and childhood. In: Schiff



ER, Sorrell MF, Maddrey WC, eds. Schiff’s Diseases of the Liver.Philadelphia: Lippincott-Raven Publishers; 1999:1357–512.

14. Setchell K, O’Connell N. Disorders of bile acid synthesis andmetabolism: a metabolic basis for liver disease. In: Suchy FJ,Sokol RJ, Balistreri WF, eds. Liver Disease in Children. Phila-delphia: Lippincott, Williams & Wilkins; 2001:701–34.

15. Tazawa Y, Kobayashi K, Ohura T, et al. Infantile cholestaticjaundice associated with adult-onset type II citrullinemia. J Pe-diatr 2001;138:735–40.

16. Sokol RJ, Stall C. Anthropometric evaluation of children withchronic liver disease. Am J Clin Nutr 1990;52:203–8.

17. Suchy FJ. Approach to the infant with cholestasis. In: Suchy FJ,Sokol RJ, Balistreri WF, eds. Liver Disease in Children. Phila-delphia: Lippincott, Williams & Wilkins; 2001:187–94.

18. Fung KP, Lau SP. Differentiation between extrahepatic and in-trahepatic cholestasis by discriminant analysis. J Paediatr ChildHealth 1990;26:132–5.

19. Feranchak A, Ramirez R, Sokol RJ. Medical and nutritional man-agement of cholestasis. In: Suchy FJ, Sokol RJ, Balistreri WF,eds. Liver Disease in Children. Philadelphia: Lippincott, Willi-ams & Wilkins; 2001:195–238.

20. Jacquemin E. Progressive familial intrahepatic cholestasis. Ge-netic basis and treatment. Clin Liver Dis 2000;4:753–63.

21. Jacquemin E, Cresteil D, Manouvrier S, et al. Heterozygous non-sense mutation of the MDR3 gene in familial intrahepatic chole-stasis of pregnancy. Lancet 1999;353:210–11.

22. Jacquemin E, De Vree JM, Cresteil D, et al. The wide spectrumof multidrug resistance 3 deficiency: from neonatal cholestasis tocirrhosis of adulthood. Gastroenterology 2001;120:1448–58.

23. Colliton RP, Bason L, Lu FM, et al. Mutation analysis of Jagged1(JAG1) in Alagille syndrome patients. Hum Mutat 2001;17:151–2.

24. Schweizer P, Kerremans J. Discordant findings in extrahepaticbile duct atresia in 6 sets of twins. Z Kinderchir 1988;43:72–5.

25. Hyams JS, Glaser JH, Leichtner AM, et al. Discordance for bil-iary atresia in two sets of monozygotic twins. J Pediatr 1985;107:420–2.

26. Rolleston HD, Hayne LB. A case of congenital hepatic cirrhosiswith obliterative cholangitis (congenital obliteration of the bileducts). BMJ 1901;1:758–60.

27. Tan CE, Driver M, Howard ER, et al. Extrahepatic biliary atresia:a first-trimester event? Clues from light microscopy and immu-nohistochemistry. J Pediatr Surg 1994;29:808–14.

28. Reynoso-Paz S, Coppel RL, Mackay IR, et al. The immunobiol-ogy of bile and biliary epithelium. Hepatology 1999;30:351–7.

29. Yerushalmi B, Dahl R, Devereaux MW, et al. Bile acid-inducedrat hepatocyte apoptosis is inhibited by antioxidants and blockersof the mitochondrial permeability transition. Hepatology 2001;33:616–26.

30. Friedman SL, Maher JJ, Bissell DM. Mechanisms and therapy ofhepatic fibrosis: report of the AASLD Single Topic Basic Re-search Conference. Hepatology 2000;32:1403–8.

31. Landing BH. Considerations of the pathogenesis of neonatal hepa-titis, biliary atresia and choledochal cyst—the concept of infantileobstructive cholangiopathy. Prog Pediatr Surg 1974;6:113–39.

32. Tarr PI, Haas JE, Christie DL. Biliary atresia, cytomegalovirus,and age at referral. Pediatrics 1996;97(6 Pt 1):828–31.

33. Fischler B, Ehrnst A, Forsgren M, et al. The viral association ofneonatal cholestasis in Sweden: a possible link between cytomeg-alovirus infection and extrahepatic biliary atresia. J Pediatr Gas-troenterol Nutr 1998;27:57–64.

34. Jevon GP, Dimmick JE. Biliary atresia and cytomegalovirus in-fection: a DNA study. Pediatr Dev Pathol 1999;2:11–4.

35. Bangaru B, Morecki R, Glaser JH, et al. Comparative studies ofbiliary atresia in the human newborn and reovirus-induced chol-angitis in weanling mice. Lab Invest 1980;43:456–62.

36. Morecki R, Glaser JH, Cho S, et al. Biliary atresia and reovirustype 3 infection. N Engl J Med 1982;307:481–4.

37. Morecki R, Glaser JH, Johnson AB, et al. Detection of reovirustype 3 in the porta hepatis of an infant with extrahepatic biliary

atresia: ultrastructural and immunocytochemical study. Hepatol-ogy 1984;4:1137–42.

38. Rosenberg DP, Morecki R, Lollini LO, et al. Extrahepatic biliaryatresia in a rhesus monkey (Macaca mulatta). Hepatology 1983;3:577–80.

39. Brown WR, Sokol RJ, Levin MJ, et al. Lack of correlation be-tween infection with reovirus 3 and extrahepatic biliary atresia orneonatal hepatitis. J Pediatr 1988;113:670–6.

40. Dussaix E, Hadchouel M, Tardieu M, et al. Biliary atresia andreovirus type 3 infection. N Engl J Med 1984;310:658.

41. Glaser JH, Balistreri WF, Morecki R. Role of reovirus type 3 inpersistent infantile cholestasis. J Pediatr 1984;105:912–5.

42. Steele MI, Marshall CM, Lloyd RE, et al. Reovirus 3 not detectedby reverse transcriptase-mediated polymerase chain reactionanalysis of preserved tissue from infants with cholestatic liverdisease. Hepatology 1995;21:697–702.

43. Tyler KL, Sokol RJ, Oberhaus SM, et al. Detection of reovirusRNA in hepatobiliary tissues from patients with extrahepatic bil-iary atresia and choledochal cysts. Hepatology 1998;27:1475–82.

44. Riepenhoff-Talty M, Schaekel K, Clark HF, et al. Group A rota-viruses produce extrahepatic biliary obstruction in orally inocu-lated newborn mice. Pediatr Res 1993;33(4 Pt 1):394–9.

45. Czech-Schmidt G, Verhagen W, Szavay P, et al. Immunologicalgap in the infectious animal model for biliary atresia. J Surg Res2001;101:62–7.

46. Petersen C, Bruns E, Kuske M, et al. Treatment of extrahepaticbiliary atresia with interferon-alpha in a murine infectious model.Pediatr Res 1997;42:623–8.

47. Qiao H, De Vincentes A, Alashari M, et al. Pathogenesis ofrotavirus-induced bile duct obstruction (a model for human bili-ary atresia) in normal Balb/c and CB17 scid mice with severecombined immunodeficiency (SCID). Pediatr Res 1999;45:116A.

48. Riepenhoff-Talty M, Gouvea V, Evans MJ, et al. Detection ofgroup C rotavirus in infants with extrahepatic biliary atresia. JInfect Dis 1996;174:8–15.

49. Bobo L, Ojeh C, Chiu D, et al. Lack of evidence for rotavirus bypolymerase chain reaction/enzyme immunoassay of hepatobiliarysamples from children with biliary atresia. Pediatr Res 1997;41:229–34.

50. Drut R, Drut RM, Gomez MA, et al. Presence of human papil-lomavirus in extrahepatic biliary atresia. J Pediatr GastroenterolNutr 1998;27:530–5.

51. Drut R, Gomez MA, Drut RM, et al. [Human papillomavirus,neonatal giant cell hepatitis and biliary duct atresia]. Acta Gas-troenterol Latinoam 1998;28:27–31.

52. Domiati-Saad R, Dawson DB, Margraf LR, et al. Cytomegalovi-rus and human herpesvirus 6, but not human papillomavirus, arepresent in neonatal giant cell hepatitis and extrahepatic biliaryatresia. Pediatr Dev Pathol 2000;3:367–73.

53. Mason AL, Xu L, Guo L, et al. Detection of retroviral antibodiesin primary biliary cirrhosis and other idiopathic biliary disorders.Lancet 1998;351:1620–4.

54. Mason A, Nair S. Primary biliary cirrhosis: new thoughts on patho-physiology and treatment. Curr Gastroenterol Rep 2002;4:45–51.

55. Schreiber RA, Kleinman RE. Genetics, immunology, and biliaryatresia: an opening or a diversion? J Pediatr Gastroenterol Nutr1993;16:111–3.

56. Burch J, Sokol RJ, Narkewicz MR, et al. The role of maternalantibodies in cholestatic liver disease in infants: a new isolatedfinding in neonatal lupus? J Pediatr Gastroenterol Nutr2000;31(suppl 2):S108.

57. Donaldson PT, Manns MP. Immunogenetics of liver disease. In:Benhamou J-P, McIntyre N, Rizetto M, et al., eds. Oxford Text-book of Clinical Hepatology, 2nd ed. Oxford, New York: OxfordUniversity Press; 1999:173–88.

58. Silveira TR, Salzano FM, Donaldson PT, et al. Association be-tween HLA and extrahepatic biliary atresia. J Pediatr Gastroen-terol Nutr 1993;16:114–7.



59. Broome U, Nemeth A, Hultcrantz R, et al. Different expression ofHLA-DR and ICAM-1 in livers from patients with biliary atresiaand Byler’s disease. J Hepatol 1997;26:857–62.

60. Kobayashi H, Puri P, O’Briain DS, et al. Hepatic overexpressionof MHC class II antigens and macrophage-associated antigens(CD68) in patients with biliary atresia of poor prognosis. J Pe-diatr Surg 1997;32:590–3.

61. Jurado A, Jara P, Camarena C, et al. Is extrahepatic biliary atresiaan HLA-associated disease? J Pediatr Gastroenterol Nutr 1997;25:557–8.

62. Donaldson PT, Clare M, Constantini PK, et al. HLA and cytokinegene polymorphisms in biliary atresia. Liver 2002;22:213–9.

63. A Kader HH, El-Ayyouti M, Hawas S, et al. HLA in Egyptianchildren with biliary atresia. J Pediatr 2002;141:432–3.

64. Chapman RW. Role of immune factors in the pathogenesis ofprimary sclerosing cholangitis. Semin Liver Dis 1991;11:1–4.

65. Gosseye S, Otte JB, De Meyer R, et al. A histological study ofextrahepatic biliary atresia. Acta Paediatr Belg 1977;30:85–90.

66. Bill AH, Haas JE, Foster GL. Biliary atresia: histopathologicobservations and reflections upon its natural history. J PediatrSurg 1977;12:977–82.

67. Ohya T, Fujimoto T, Shimomura H, et al. Degeneration of intra-hepatic bile duct with lymphocyte infiltration into biliary epithe-lial cells in biliary atresia. J Pediatr Surg 1995;30:515–8.

68. Abbas AK, Lichtman AH, Pober JS. Cellular and MolecularImmunology, 4th ed. Philadelphia: WB Saunders; 2000.

69. Nakada M, Nakada K, Kawaguchi F, et al. Immunologic reactionand genetic factors in biliary atresia. Tohoku J Exp Med 1997;181:41–7.

70. Davenport M, Gonde C, Redkar R, et al. Immunohistochemistryof the liver and biliary tree in extrahepatic biliary atresia. J Pe-diatr Surg 2001;36:1017–25.

71. Urushihara N, Iwagaki H, Yagi T, et al. Elevation of seruminterleukin-18 levels and activation of Kupffer cells in biliaryatresia. J Pediatr Surg 2000;35:446–9.

72. Funaki N, Sasano H, Shizawa S, et al. Apoptosis and cell prolif-eration in biliary atresia. J Pathol 1998;186:429–33.

73. Liu C, Chiu JH, Chin T, et al. Expression of fas ligand on bileductule epithelium in biliary atresia—a poor prognostic factor. JPediatr Surg 2000;35:1591–6.

74. Bezerra JA, Tiao G, Ryckman FC, et al. Genetic induction ofproinflammatory immunity in children with biliary atresia. Lancet2002;360:1653–9.

75. Vasiliauskas E, Targan S, Cobb L, et al. Biliary atresia—an au-toimmune disorder? Hepatology 1995;22(4 Pt 2):87.

76. Bernal W, Moloney M, Underhill J, et al. Association of tumornecrosis factor polymorphism with primary sclerosing cholangi-tis. J Hepatol 1999;30:237–41.

77. Mitchell SA, Grove J, Spurkland A, et al. Association of thetumour necrosis factor alpha-308 but not the interleukin 10-627promoter polymorphism with genetic susceptibility to primarysclerosing cholangitis. Gut 2001;49:288–94.

78. Donaldson P, Agarwal K, Craggs A, et al. HLA and interleukin 1gene polymorphisms in primary biliary cirrhosis: associationswith disease progression and disease susceptibility. Gut 2001;48:397–402.

79. Grove J, Daly AK, Bassendine MF, et al. Interleukin 10 promoterregion polymorphisms and susceptibility to advanced alcoholicliver disease. Gut 2000;46:540–5.

80. Yee LJ, Tang J, Gibson AW, et al. Interleukin 10 polymorphismsas predictors of sustained response in antiviral therapy for chronichepatitis C infection. Hepatology 2001;33:708–12.

81. Klippel CH. A new theory of biliary atresia. J Pediatr Surg 1972;7:651–4.

82. Pickett LK, Briggs HC. Biliary obstruction secondary to hepaticvascular ligation in fetal sheep. J Pediatr Surg 1969;4:95–101.

83. Ho CW, Shioda K, Shirasaki K, et al. The pathogenesis of biliaryatresia: a morphological study of the hepatobiliary system and thehepatic artery. J Pediatr Gastroenterol Nutr 1993;16:53–60.

84. Yokoyama T, Copeland NG, Jenkins NA, et al. Reversal of left-rightasymmetry: a situs inversus mutation. Science 1993;260:679–82.

85. Schon P, Tsuchiya K, Lenoir D, et al. Identification, genomicorganization, chromosomal mapping and mutation analysis of thehuman INV gene, the ortholog of a murine gene implicated inleft-right axis development and biliary atresia. Hum Genet 2002;110:157–65.

86. Bamford RN, Roessler E, Burdine RD, et al. Loss-of-functionmutations in the EGF-CFC gene CFC1 are associated with humanleft-right laterality defects. Nat Genet 2000;26:365–9.

87. Jacquemin E, Cresteil D, Raynaud N, et al. CFCI gene mutationand biliary atresia with polysplenia syndrome. J Pediatr Gastro-enterol Nutr 2002;34:326–7.

88. Schmidt C, Bladt F, Goedecke S, et al. Scatter factor/hepatocytegrowth factor is essential for liver development. Nature 1995;373:699–702.

89. Kohsaka T, Yuan ZR, Guo SX, et al. The significance of humanjagged 1 mutations detected in severe cases of extrahepatic biliaryatresia. Hepatology 2002;36:904–12.

90. Clotman F, Lannoy VJ, Reber M, et al. The onecut transcriptionfactor HNF6 is required for normal development of the biliarytract. Development 2002;129:1819–28.

91. Harper P, Plant JW, Unger DB. Congenital biliary atresia andjaundice in lambs and calves. Aust Vet J 1990;67:18–22.

92. Balistreri WF, Bove K, Ryckman FC. Biliary atresia and otherdisorders of the extrahepatic bile ducts. In: Suchy FJ, Sokol RJ,Balistreri WF, eds. Liver Disease in Children. Philadelphia: Lip-pincott, Williams & Wilkins; 2001:253–74.

93. Zerbini MC, Gallucci SD, Maezono R, et al. Liver biopsy inneonatal cholestasis: a review on statistical grounds. Mod Pathol1997;10:793–9.

94. Deutsch GH, Sokol RJ, Stathos TH, et al. Proliferation to paucity:evolution of bile duct abnormalities in a case of Alagille syn-drome. Pediatr Dev Pathol 2001;4:559–63.

95. Gilmour SM, Hershkop M, Reifen R, et al. Outcome of hepato-biliary scanning in neonatal hepatitis syndrome. J Nucl Med1997;38:1279–82.

96. Farrant P, Meire HB, Mieli-Vergani G. Improved diagnosis ofextraheptic biliary atresia by high frequency ultrasound of the gallbladder. Br J Radiol 2001;74:952–4.

97. Choi SO, Park WH, Lee HJ, et al. “Triangular cord:” a sono-graphic finding applicable in the diagnosis of biliary atresia. JPediatr Surg 1996;31:363–6.

98. Guibaud L, Lachaud A, Touraine R, et al. MR cholangiography inneonates and infants: feasibility and preliminary applications.AJR Am J Roentgenol 1998;170:27–31.

99. Norton KI, Glass RB, Kogan D, et al. MR cholangiography in theevaluation of neonatal cholestasis: initial results. Radiology 2002;222:687–91.

100. Wilkinson ML, Mieli-Vergani G, Ball C, et al. Endoscopic ret-rograde cholangiopancreatography in infantile cholestasis. ArchDis Child 1991;66:121–3.

101. Iinuma Y, Narisawa R, Iwafuchi M, et al. The role of endoscopicretrograde cholangiopancreatography in infants with cholestasis.J Pediatr Surg 2000;35:545–9.

102. Meisheri IV, Kasat LS, Kumar A, et al. Duodenal intubation andtest for bile—a reliable method to rule out biliary atresia. PediatrSurg Int 2002;18:392–5.

103. Larrosa-Haro A, Caro-Lopez AM, Coello-Ramirez P, et al. Duo-denal tube test in the diagnosis of biliary atresia. J Pediatr Gas-troenterol Nutr 2001;32:311–5.

104. Matsui A, Dodoriki M. Screening for biliary atresia. Lancet 1995;345:1181.

105. Garel LA, Belli D, Grignon A, et al. Percutaneous cholecystog-raphy in children. Radiology 1987;165:639–41.

106. Markowitz J, Daum F, Kahn EI, et al. Arteriohepatic dysplasia. I.Pitfalls in diagnosis and management. Hepatology 1983;3:74–6.

107. Karrer FM, Bensard DD. Neonatal cholestasis. Semin PediatrSurg 2000;9:166–9.



108. Chardot C, Carton M, Spire-Bendelac N, et al. Epidemiology ofbiliary atresia in France: a national study 1986–96. J Hepatol1999;31:1006–13.

109. Chardot C, Carton M, Spire-Bendelac N, et al. Is the Kasai op-eration still indicated in children older than 3 months diagnosedwith biliary atresia? J Pediatr 2001;138:224–8.

110. Mones RL, DeFelice AR, Preud’Homme D. Use of neomycin asthe prophylaxis against recurrent cholangitis after Kasai portoen-terostomy. J Pediatr Surg 1994;29:422–4.

111. Karrer FM, Lilly JR. Corticosteroid therapy in biliary atresia. JPediatr Surg 1985;20:693–5.

112. Balistreri WF. Bile acid therapy in pediatric hepatobiliary dis-ease: the role of ursodeoxycholic acid. J Pediatr GastroenterolNutr 1997;24:573–89.

113. Chin SE, Shepherd RW, Cleghorn GJ, et al. Pre-operative nutri-tional support in children with end-stage liver disease acceptedfor liver transplantation: an approach to management. J Gastro-enterol Hepatol 1990;5:566–72.

114. Lilly JR. Experimental investigations into the etiology of chol-angitis following operation for biliary atresia. J Pediatr Surg1978;13:559–60.

115. Muraji T, Higashimoto Y. The improved outlook for biliary atre-sia with corticosteroid therapy. J Pediatr Surg 1997;32:1103–6;discussion 1106–7.

116. Dillon PW, Owings E, Cilley R, et al. Immunosuppression asadjuvant therapy for biliary atresia. J Pediatr Surg 2001;36:80–5.

117. van Rossum TG, Vulto AG, Hop WC, et al. Glycyrrhizin-inducedreduction of ALT in European patients with chronic hepatitis C.Am J Gastroenterol 2001;96:2432–7.

118. Vohra P, Haller C, Emre S, et al. Neonatal hemochromatosis: theimportance of early recognition of liver failure. J Pediatr 2000;136:537–41.

119. Melter M, Rodeck B, Kardorff R, et al. Progressive familial in-trahepatic cholestasis: partial biliary diversion normalizes serumlipids and improves growth in noncirrhotic patients. Am J Gas-troenterol 2000;95:3522–8.

120. Ng VL, Ryckman FC, Porta G, et al. Long-term outcome afterpartial external biliary diversion for intractable pruritus in patientswith intrahepatic cholestasis. J Pediatr Gastroenterol Nutr 2000;30:152–6.

121. Emerick KM, Whitington PF. Partial external biliary diversion forintractable pruritus and xanthomas in Alagille syndrome. Hepa-tology 2002;35:1501–6.

122. Karrer FM, Price MR, Bensard DD, et al. Long-term results with theKasai operation for biliary atresia. Arch Surg 1996;131:493–6.

123. Laurent J, Gauthier F, Bernard O, et al. Long-term outcome aftersurgery for biliary atresia. Study of 40 patients surviving for morethan 10 years. Gastroenterology 1990;99:1793–7.

124. Howard ER, MacLean G, Nio M, et al. Survival patterns in biliaryatresia and comparison of quality of life of long-term survivors inJapan and England. J Pediatr Surg 2001;36:892–7.

125. McKiernan PJ, Baker AJ, Kelly DA. The frequency and outcomeof biliary atresia in the UK and Ireland. Lancet 2000;355:25–9.

126. Stewart SM, Uauy R, Kennard BD, et al. Mental development andgrowth in children with chronic liver disease of early and lateonset. Pediatrics 1988;82:167–72.

127. Stewart SM, Uauy R, Waller DA, et al. Mental and motor devel-opment, social competence, and growth one year after successfulpediatric liver transplantation. J Pediatr 1989;114(4 Pt 1):574–81.

128. Wayman KI, Cox KL, Esquivel CO. Neurodevelopmental out-come of young children with extrahepatic biliary atresia 1 yearafter liver transplantation. J Pediatr 1997;131:894–8.

129. Miga D, Sokol RJ, Mackenzie T, et al. Survival after first esoph-ageal variceal hemorrhage in patients with biliary atresia. J Pe-diatr 2001;139:291–6.

130. Goncalves ME, Cardoso SR, Maksoud JG. Prophylactic sclero-therapy in children with esophageal varices: long-term results of

a controlled prospective randomized trial. J Pediatr Surg 2000;35:401–5.

131. Lui HF, Stanley AJ, Forrest EH, et al. Primary prophylaxis ofvariceal hemorrhage: a randomized controlled trial comparingband ligation, propranolol, and isosorbide mononitrate. Gastro-enterology 2002;123:735–44.

132. Ozsoylu S, Kocak N, Demir H, et al. Propranolol for primary andsecondary prophylaxis of variceal bleeding in children with cir-rhosis. Turk J Pediatr 2000;42:31–3.

133. Shashidhar H, Langhans N, Grand RJ. Propranolol in preventionof portal hypertensive hemorrhage in children: a pilot study. JPediatr Gastroenterol Nutr 1999;29:12–7.

134. Karrer FM, Narkewicz MR. Esophageal varices: current manage-ment in children. Semin Pediatr Surg 1999;8:193–201.

135. Sasaki T, Hasegawa T, Nakajima K, et al. Endoscopic varicealligation in the management of gastroesophageal varices in post-operative biliary atresia. J Pediatr Surg 1998;33:1628–32.

136. McKiernan PJ, Beath SV, Davison SM. A prospective study ofendoscopic esophageal variceal ligation using a multiband ligator.J Pediatr Gastroenterol Nutr 2002;34:207–11.

137. Price MR, Sartorelli KH, Karrer FM, et al. Management of esoph-ageal varices in children by endoscopic variceal ligation. J Pedi-atr Surg 1996;31:1056–9.

138. Hackworth CA, Leef JA, Rosenblum JD, et al. Transjugular in-trahepatic portosystemic shunt creation in children: initial clinicalexperience. Radiology 1998;206:109–14.

139. Heyman MB, LaBerge JM, Somberg KA, et al. Transjugularintrahepatic portosystemic shunts (TIPS) in children. J Pediatr1997;131:914–9.

140. Brandt CT, Rothbarth LJ, Kumpe D, et al. Splenic embolizationin children: long-term efficacy. J Pediatr Surg 1989;24:642–4;discussion 644–5.

141. Harned RK 2nd, Thompson HR, Kumpe DA, et al. Partial splenicembolization in five children with hypersplenism: effects ofreduced-volume embolization on efficacy and morbidity. Radiol-ogy 1998;209:803–6.

142. Trivedi P, Dhawan A, Risteli J, et al. Prognostic value of serumhyaluronic acid and type I and III procollagen propeptides inextrahepatic biliary atresia. Pediatr Res 1995;38:568–73.

143. Kobayashi H, Miyano T, Horikoshi K, et al. Prognostic value ofserum procollagen III peptide and type IV collagen in patientswith biliary atresia. J Pediatr Surg 1998;33:112–4.

144. Broide E, Farrant P, Reid F, et al. Hepatic artery resistance indexcan predict early death in children with biliary atresia. LiverTranspl Surg 1997;3:604–10.

145. Fujimoto T, Ohya T, Miyano T. A new clinical prognostic predictorfor patients with biliary atresia. J Pediatr Surg 1994;29:757–60.

146. Vazquez-Estevez J, Stewart B, Shikes RH, et al. Biliary atresia:early determination of prognosis. J Pediatr Surg 1989;24:48–50;discussion 50–1.

147. Endo M, Masuyama H, Watanabe K, et al. Calculation of biliaryatresia prognostic index using a multivariate linear model. J Pe-diatr Surg 1995;30:1575–9.

148. Malatack JJ, Schaid DJ, Urbach AH, et al. Choosing a pediatricrecipient for orthotopic liver transplantation. J Pediatr 1987;111:479–89.

149. McDiarmid SV, Anand R, Lindblad AS. Development of a pediatricend-stage liver disease score to predict poor outcome in childrenawaiting liver transplantation. Transplantation 2002;74:173–81.

150. Schweizer P, Lunzmann K. Extrahepatic bile duct atresia: howefficient is the hepatoporto-enterostomy? Eur J Pediatr Surg1998;8:150–4.

151. Mowat AP, Davidson LL, Dick MC. Earlier identification ofbiliary atresia and hepatobiliary disease: selective screening in thethird week of life. Arch Dis Child 1995;72:90–2.

152. 2000 Annual Report. The US Scientific Registry of TransplantRecipients and The Organ Procurement and Transplantation Net-work. HHS/HRS/OSP/DOT;United Network for Organ Sharing,pp. 191–236.



Pathogenesis and Outcome of Biliary Atresia: …...(18), overlap in values precludes the...

Documents

Transcript of Pathogenesis and Outcome of Biliary Atresia: …...(18), overlap in values precludes the...