Early Human Development 111 (2017) 16–19

Contents lists available at ScienceDirect

Early Human Development

journal homepage: www.elsevier.com/locate/earlhumdev

Prenatal diagnosis of biliary atresia: A case series MARK

a,⁎ a b a c d c,e
O. Shen , H.Y. Sela , H. Nagar , R. Rabinowitz , E. Jacobovich , D. Chen , E. Granot
a
Department of Ob/Gyn, Shaare Zedek Medical Center, Jerusalem, Israel
b
Department of Pediatric Surgery, Shaare Zedek Medical Center, Jerusalem, Israel
c
Department of Pediatrics, Kaplan Medical Center, Rehovot, Israel
d
Tel-Aviv Sackler School of Medicine, Tel-Aviv University, Tel-Aviv, Israel
e
Department of Pediatrics, Shaare Zedek Medical Center, Jerusalem, Israel

A R T I C L E I N F O A B S T R A C T

Keywords: Background: Biliary atresia is a progressive disease presenting with jaundice, and is the most common indication
Biliary atresia for liver transplantation in the pediatric population. Prenatal series have yielded conflicting results concerning a
Prenatal diagnosis possible association between BA and prenatal nonvisualization of the gallbladder.
Ultrasound Aims: This retrospective case series was performed to assess the association between biliary atresia, prenatal
Gallbladder
nonvisualization of the gallbladder and other sonographic signs.
Triangular cord sign
Study design/subjects: We identified biliary atresia patients who underwent a Kasai procedure by a single
Prenatal nonvisualization of the gallbladder
Amniotic fluid pediatric surgeon and/or follow up by a single pediatric gastroenterologist. Axial plane images and/or video
Digestive enzymes recordings were scrutinized for sonographic signs of biliary atresia on the second trimester anomaly scan.
Biliary cystic malformation Outcome measures: Proportion of biliary atresia cases with prenatal sonographic signs.
Choledochal cyst Results: Twenty five charts of children with biliary and high quality prenatal images were retrieved. 6/25 (24%)
of cases analyzed had prenatal nonvisualization of the gallbladder or a small gallbladder on the prenatal scan.
Two cases had biliary atresia splenic malformation syndrome. None of the cases had additional sonographic
markers of biliary atresia.
Conclusions: Our study suggests that in addition to the well-established embryonic and cystic forms, an
additional type can be suspected prenatally, which is characterized by prenatal nonvisualization of the
gallbladder in the second trimester. This provides additional evidence that some cases of BA are of fetal rather
than perinatal onset and may have important implications for prenatal diagnosis, for counseling and for research
of the disease's etiology and pathophysiology.

1. Introduction 2. Patients and methods

Biliary atresia (BA) is a severe condition, with an incidence of
approximately 1:14,000–1:16,000 live births [1,2] frequently resulting
in severe morbidity or death. It is the leading indication for liver
transplantation in children [3,4]. Prenatal diagnosis is important, as it
allows parents to make informed decisions concerning continuation of
the pregnancy, as well as facilitating early neonatal diagnosis when
pregnancy is continued. Signs of BA on prenatal ultrasound are largely
unknown. The aim of this study was to identify possible signs of BA on
prenatal ultrasonography. This is the first attempt to assess the presence
of midtrimester sonographic markers, other than biliary cystic mal-
formations, in a cohort of children with BA.
This was a retrospective study of infants with biliary atresia.
Patients entered into the study were either operated by one of the
authors (HN) in the period between January 2004 and June 2014 at Tel
Aviv medical center or at Shaare Zedek medical center Jerusalem, or
had pediatric follow up by one of the authors (EG) during the same
period. All patients underwent hepatoportoenterostomy (Kasai proce-
dure). Some, subsequently, had liver transplantation.
Prenatal records were collected from the participating institutions'
databases. Electronic images and video recordings were collected from
the sonographers or from the patients. Postnatal charts and ultrasound
reports were reviewed in all cases. Second trimester fetal anomaly scan
reports and axial plane images/video recordings of the portal area were
reviewed by two experienced sonographers (OS and RR), who were
blinded to the reports. Cases lacking good quality prenatal electronic

⁎
Corresponding author at: Dep't. Ob/Gyn, Shaare Zedek Medical Center, P.O. Box 3235, Jerusalem 91031, Israel.
E-mail address: orishen@netvision.net.il (O. Shen).

http://dx.doi.org/10.1016/j.earlhumdev.2017.05.005
Received 9 December 2016; Received in revised form 8 May 2017; Accepted 13 May 2017
0378-3782/ © 2017 Published by Elsevier Ireland Ltd.
O. Shen et al. Early Human Development 111 (2017) 16–19

images and/or video recordings were excluded.

The following parameters were recorded from the prenatal ultra-
sound reports and electronic media: presence, size and shape of the
fetal gallbladder; gallbladder longitudinal and transverse diameter were
measured if it appeared subjectively to be small; presence of triangular
cord sign, a thickened peri portal echogenicity, as assessed subjectively;
hepatic hyperechogenicity; existence of biliary cystic malformations
(biliary cystic malformation); liver longitudinal length; presence of
echogenic bowel (EB); existence of associated anomalies.
Institutional review board approval was obtained and oral consent
given by the patients.

3. Results

We identified 53 cases of children with BA. Twenty eight cases were

excluded from the study; 13 were excluded because prenatal anomaly
scans were not performed, and in 15 cases the prenatal images were of
insufficient quality to be reliably assessed. Twenty five cases had good
quality electronic images and/or video recordings for review, and these
constituted the study group. Within the study group there was complete
agreement between the two reviewing sonographers concerning the
presence or absence of potential signs of BA.
Median gestational age at prenatal sonography was 22 weeks (range
19–24).
6/25 (24%) cases had an absent or abnormal gallbladder. Five of
these had PNVGB and one had an abnormally small gallbladder, < 2
standard deviations below the median longitudinal and transverse
diameters [5]. In all but one of these cases (PNVGB) the report issued Fig. 1. Features of BA cases
by the sonographer confirmed PNVGB. There were no cases with biliary BA, biliary atresia; PNVGB, prenatal nonvisualization of gallbladder; BASM (biliary
cystic malformation, defined as a cyst in the portal region assessed as atresia splenic malformation).
being part of the biliary system, and no cases of an abnormally thick
wall of the left or right portal vein branch, representing a triangular
cord sign [6]. All cases had normal liver echogenicity and length
between the 5th and 95th percentile [7]. None had echogenic bowel
(EB). Biliary atresia splenic malformation syndrome was present in 2/
25 (8%) cases in the study group. One case with an absent gallbladder
had polysplenia, centrally positioned liver, preduodenal portal vein,
annular pancreas, gut malrotation, horseshoe kidney and interrupted
inferior vena cava with azygous continuation. A second case with a
normally appearing gallbladder had gut malrotation and interrupted
inferior vena cava with azygous continuation. 3/28 (10.7%) of the
excluded cases had biliary atresia splenic malformation (Fig. 1) bring-
ing the total no. of cases with biliary atresia splenic malformation to 5/
53 (9.4%).
In all 6 cases with prenatal absent/small gallbladder (Fig. 2), the
gallbladder was not visualized on post-natal sonography. 10/19
(52.6%) of the cases with a normally visualized fetal gallbladder, had
non visualization of the gallbladder on the postnatal sonogram, 5/19
(26.3%) had a small gallbladder and 4/19 (21.1%) had normal
appearance and size of the gallbladder on postnatal sonography.
Mean age at surgery was 54 days (range 49–90) and 50 days (range
38–68) for infants with and without PNVGB respectively (p = 0.401).
In the excluded group a post-natal sonogram was not available in
Fig. 2. Abnormally small gallbladder in case of BA at mid trimester. Arrow pointing to
three cases, demonstrated a normally appearing gallbladder in 4/25 gallbladder.
(16%) cases and was absent or of abnormal shape or size in 21/25
(84%) cases. Average age at surgery in the excluded group was 53 days stools. The purpose of this study was to assess whether established
(range 44–78). postnatal sonographic signs are present on prenatal scans of infants
After completing data collection, a case of PNVGB associated with with BA.
presumably triangular cord sign (Fig. 3) was observed outside the Specific postnatal sonographic signs of BA include a small or absent
participating centers (courtesy CG). The postnatal diagnosis was gallbladder and triangular cord sign [9–13]. Triangular cord sign [6],
isolated gallbladder agenesis. present in most infants with BA is the sonographic depiction of a fibrous
thickening in the peri-portal area. An abnormal triangular or cone
4. Discussion shaped echogenicity is depicted as an abnormally thick (> 4 mm)
echogenic anterior wall of the right portal vein. It is a specific sign, and
There are few reports of prenatal diagnosis of BA [8]. Most cases are present in 73%–85% of infants [9–12]. Although not used by all
diagnosed post-natally following the appearance of jaundice or acholic

17
O. Shen et al. Early Human Development 111 (2017) 16–19

4.1. PNVGB

Second trimester PNVGB has a general population incidence of

1:875 [26]. In the majority of cases the gallbladder is visualized later in
pregnancy, or in the neonatal period. Between 15%–43% [8,27] of
cases with PNVGB have isolated gallbladder agenesis, which is a benign
condition. Up to 13% will have cystic fibrosis (CF) [8,27,29–30].
Several series that examined the outcome of PNVGB have shown
conflicting results concerning a possible association with BA. Our
literature search revealed 13 cases of BA associated with isolated
PNVGB in the midtrimester. This renders additional support to the
existence of a “fetal” rather than a “perinatal” timeframe for the insult,
whatever it may be, in a certain proportion of cases. In one study [26]
5/85 (5.8%) cases with isolated PNVGB, and 3/17 (17.6%) cases with
non-isolated PNVGB had BA (two of these were associated with EB). A
recent publication [27] reported that 1/28 (3.7%) fetuses with isolated
PNVGB, and 3/32 (9.3%) of all PNVGB, two of them associated with EB,
had BA. Several other series, including our own, failed to confirm such
an association [8,26,30,31]. By pooling the results from the available
series, we calculated a 3% probability of BA when PNVGB is confirmed
on repeat sonograms. When there is EB the risk may be higher, and in
the presence of left isomerism, the risk is much higher. In an attempt to
Fig. 3. False positive triangular cord sign 32 weeks gestation in a fetus with isolated
gallbladder agenesis. Arrow pointing to TCS.
resolve the discrepancy between the various reports concerning the
association between PNVGB and BA, we reviewed prenatal scans of
children born with BA. Our current results support the existence of an
sonographers, a recent meta-analysis [14] confirms its reasonable
association between second trimester PNVGB and BA, suggesting an
sensitivity and good specificity (85% and 92% respectively). Nonvisua-
insult occurring in the midtrimester or earlier.
lization of the gallbladder [10,12] and an abnormally small gallbladder
[10–12] have been reported in 23%–31%, and 25%–97% respectively
4.2. Amniotic fluid enzymes
of children with BA. Nonspecific signs include hepatomegaly and
abnormal liver echogenicity [13,15]. The only sign present in our
Further support for a prenatal insult in the pathogenesis of BA can
study at mid-trimester was PNVGB or small gallbladder.
be inferred from the association of low levels of amniotic fluid digestive
The pathogenesis of BA is, in all likelihood, multifactorial. Its
enzymes prior to 22 weeks gestation and BA [32,33]. Normal amniotic
etiology and nature are complex. Genetic, Infectious, immunological
fluid digestive enzymes associated with PNVGB effectively rules out BA
and teratogenic factors are likely culprits. Traditionally, two forms of
[27,28,30], while abnormal enzyme levels may predict BA in up to 50%
BA were recognized the embryonic form and the so called “perinatal
of cases of PNVGB.
form” [16,17]. The embryonic form, biliary atresia splenic malforma-
Our observation confirms that an additional “fetal” variant of BA
tion, is associated with heterotaxy syndrome, typically left isomerism
that manifests as PNVGB in the second trimester exists. This has
[18–20]. The proportion of cases in our series (8% in the study group
implications concerning BA research, and especially regarding its
and 9.4% in the entire series) with biliary atresia splenic malformation
etiology and pathogenesis which have been attributed in the past to
is similar to the reported figure in the literature of 10–15%. While
perinatal events. The proportion of BA cases with absent gallbladder on
prenatally diagnosed left isomerism is associated with BA in 4–10% of
postnatal sonography [10–12] is significantly higher than that of
cases [21,22], it would appear that PNVGB or a small gallbladder in the
PNVGB (24%) in our series. Possible explanations are that only a
presence of left isomerism, are highly suggestive of BA [20].
minority of cases are of the “fetal type”, or that changes in gallbladder
The non-embryonic form, comprising the majority of cases, has seen
appearance may be evident only after birth.
a paradigm shift concerning assessment of its time of onset, and
We chose to show an image of a false positive triangular cord sign as
possible etiological pathways. In the past, it was somewhat simplisti-
this sign has not been previously described on prenatal sonography. Its
cally and with scant evidence considered to be the result of a perinatal
significance in fetal life has yet to be determined. As its presence
insult, possibly related to viral infection. Current understanding [23]
implies fibrosis, it is less likely to be present early in the course of the
outlines several possible pathways leading to BA, with a growing body
disease, especially in utero.
of evidence pointing to an in utero insult in many cases. Cystic BA can
The strength of our study is our ability to review high quality images
be diagnosed in utero, usually in the midtrimester, by demonstrating a
of a relatively large group of BA cases. It is the first series to specifically
stable size cyst in the portal region. A mechanical factor may be at work
seek PNVGB or TCS in cases with proven BA. The weakness lies in its
in this anomaly, causing disruption of continuity between the intrahe-
retrospective design; signs for BA other than gallbladder visualization
patic and extrahepatic systems. Although none of the cases in our
were not specifically checked for during the scan. Only 45% of cases
population had this type, biliary cystic malformation, a cyst in the
had images of sufficient quality to be included in the study group.
region of the porta hepatis, present in 11–22% of cases of BA, is the
Nevertheless, any selection bias introduced due to the relatively low
most common prenatally detected abnormality in BA [24]. When
ascertainment rate is likely to be minimal, considering that the
biliary cystic malformation is associated with PNVGB or an abnormally
characteristics of the excluded cases were similar to those of the study
small gallbladder, the index of suspicion for BA should be very high
group, regarding postnatal ultrasound imaging, the prevalence of
[25]. Other possible mechanisms include immunological, genetic and
biliary atresia splenic malformation, and timing of the Kasai procedure.
environmental factors. Further support to the intrauterine origin of BA
Many sonographers in Israel include imaging the gallbladder in the
comes from observations tying ingestion of a naturally occurring
routine anomaly scan, and some cases of PNVGB and BA underwent
teratogen by pregnant sheep (Biliatresone) and BA in lambs. This has
therapeutical abortions [27,28,32]. We are aware of patients with
been reproduced in a murine model.
isolated PNVGB that elected to undergo termination of pregnancy, and
had a normal fetal autopsy (unpublished). Our approach, when PNVGB

18
O. Shen et al.
is diagnosed, is to rule out CF, search for additional sonographic
markers of BA, such as heterotaxy, biliary cystic malformation, EB
and triangular cord sign. If the patient seeks reassurance from a normal
result, or would consider a therapeutic abortion based on abnormal
amniotic fluid digestive enzymes, amniocentesis prior to 22 weeks
should be considered.
Further studies, employing multiple modalities, such as standar-
dized amniotic fluid digestive enzymes testing, fetal MRI, third
trimester ultrasound scanning and cordocentesis, are required to
establish appropriate strategies for prenatal diagnosis of BA. In the
current study no supporting sonographic signs, other than biliary
atresia splenic malformation and PNVGB were found. The role of
additional signs should be determined in future studies by carefully
scrutinizing cases at risk.
Acknowledgments
We would like to thank Dr. Catherine Garel for her thoughtful
remarks as well as for providing Fig. 3.
References
[1] A. Baker, N. Hadzic, A. Dhawan, G. Mieli-Vergani, Biliary atresia, Recent Adv.
Pediatr. 16 (1998) 25–40.
[2] P.W. Yoon, J.S. Bresee, R.S. Olney, L.M. James, M.J. Khoury, Epidemiology of
biliary atresia: a population-based study, Pediatrics 99 (3) (1997) 376–382.
[3] D. Alagille, Extrahepatic biliary atresia, Hepatology 4 (1984) 7–10.
[4] R.E. Kleinman, J.P. Vacanti, Liver transplantation, in: W.A. Walker, P.R. Durie,
J.R. Hamilton, J.A. Walker-Smith, J.B. Watkins (Eds.), Pediatric Gastrointestinal
Disease, II, 2 BD Decker, Philadelphia, PA, 1991, pp. 1108–1124.
[5] I. Goldstein, A. Tamir, A. Weisman, P. Jakobi, J.A. Copel, Growth of the fetal
gallbladder in normal pregnancies, Ultrasound Obstet. Gynecol. 4 (1994) 289–293.
[6] S.O. Choi, W.H. Park, H.J. Lee, S.K. Woo, ‘Triangular cord’: a sonographic finding
applicable in the diagnosis of biliary atresia, J. Pediatr. Surg. 31 (3) (1996)
363–366.
[7] A.M. Vintzileos, S. Neckles, W.A. Campbell, J.W. Andreoli Jr., B.M. Kaplan,
D.J. Nochimson, Fetal liver ultrasound measurements during normal pregnancy,
Obstet. Gynecol. 66 (4) (1985) 477–480.
[8] O. Shen, R. Rabinowitz, S. Yagel, M. Gal, Absent gallbladder on fetal ultrasound:
prenatal findings and postnatal outcome, Ultrasound Obstet. Gynecol. 37 (2011)
673–677.
[9] H.J. Lee, S.M. Lee, W.H. Park, S.O. Choi, Objective criteria of triangular cord sign in
biliary atresia on US scans, Radiology 229 (2) (2003) 395–400.
[10] S. Takamizawa, A. Zaima, T. Muraji, K. Kanegawa, Y. Akasaka, S. Satoh, Nishijima,
Can biliary atresia be diagnosed by ultrasonography alone? E.J. Pediatr. Surg. 42
(12) (2007) 2093–2096.
[11] A.P. Tan Kendrick, K.B. Phua, B.C. Ooi, C.E. Tan, Biliary atresia: making the
diagnosis by the gallbladder ghost triad, Pediatr. Radiol. 33 (5) (2003) 311–315.
[12] T.M. Humphrey, M.D. Stringer, Biliary atresia: US diagnosis, Radiology 244 (3)
(2007) 845–851.
19
Early Human Development 111 (2017) 16–19
[13] R. Iorio, D. Liccardo, F. Di Dato, M.G. Puoti, M.I. Spagnuolo, D. Alberti, et al.,
Ultrasound scanning in infants with biliary atresia: the different implications of
biliary tract features and liver echostructure, Ultraschall Med. 34 (5) (2013)
463–467.
[14] L. Zhou, Q. Shan, W. Tian, Z. Wang, J. Liang, X. Xie, Ultrasound for the diagnosis of
biliary atresia: a meta-analysis, AJR Am. J. Roentgenol. 206 (5) (2016) W73–W82.
[15] V. Mittal, A.K. Saxena, K.S. Sodhi, B.R. Thapa, K.L. Rao, A. Das, et al., Role of
abdominal sonography in the preoperative diagnosis of extrahepatic biliary atresia
in infants younger than 90 days, AJR Am. J. Roentgenol. 196 (4) (2011) 438–445.
[16] E. Kahn, Biliary atresia revisited, Pediatr. Dev. Pathol. 7 (2004) 109–124.
[17] C.L. Mack, R.J. Sokol, Unraveling the pathogenesis and etiology of biliary atresia,
Pediatr. Res. 57 (2005) 87–94.
[18] B. Fischler, B. Haglund, A. Hjern, A population based study on the incidence and
possible pre- and perinatal etiological risk factors of biliary atresia, J. Pediatr. 141
(2002) 217–222.
[19] M. Davenport, M. Savage, A.P. Mowat, E.R. Howard, Biliary atresia splenic
malformation syndrome: an etiologic and prognostic subgroup, Surgery 113 (1993)
662–668.
[20] O.R. Guttman, E.A. Roberts, R.A. Schreiber, C.C. Barker, Ng VL; Canadian Pediatric
Hepatology Research Group. Biliary atresia with associated structural malforma-
tions in Canadian infants, Liver Int. 31 (10) (2011) 1485–1493.
[21] I. Gottschalk, R. Stressig, J. Ritgen, U. Herberg, J. Breuer, A. Vorndamme, et al.,
Extracardiac anomalies in prenatally diagnosed heterotaxy syndromes, Ultrasound
Obstet. Gynecol. 47 (4) (2015) 443–449.
[22] M.C. Escobar-Diaz, K. Friedman, Y. Salem, G.R. Marx, B.T. Kalish, T. Lafranchi,
et al., Perinatal and infant outcomes of prenatal diagnosis of heterotaxy syndrome
(asplenia and polysplenia), Am. J. Cardiol. 114 (4) (2014) 612–617.
[23] M. Davenport, Biliary atresia: from Australia to the zebrafish, J. Pediatr. Surg. 51
(2) (2016) 200–205.
[24] E. Caponcelli, A.S. Knisely, M. Davenport, Cystic biliary atresia: an etiologic and
prognostic subgroup, J. Pediatr. Surg. 43 (2008) 1619–1624.
[25] L. Burc, E. Vuillard, J. Guibourdenche, M. Conti, C. Garel, D. Porquet, J.F. Oury,
D. Luton, Prenatal diagnosis and follow up of biliary atresia, Br. J. Obstet. Gynaecol.
108 (2001) 1108–1110.
[26] S. Blazer, M. Bronshtein, E. Zimmer, Nonvisualization of the fetal gallbladder in
early pregnancy: comparison with clinical outcome, Radiology 224 (2002)
379–382.
[27] S. Dreux, M. Boughanim, C. Lepinard, A. Guichet, J.M. Rival, A. de Becdelievre,
et al., Relationship of non-visualization of the fetal gallbladder and amniotic fluid
digestive enzymes analysis to outcome, Prenat. Diagn. 32 (5) (2012) 423–426.
[28] R. Bardin, E. Ashwal, B. Davidov, D. Danon, M. Shohat, I. Meizner, Nonvisualization
of the fetal gallbladder: can levels of γ-Glutamyl Transpeptidase in amniotic fluid
predict fetal prognosis? Fetal Diagn. Ther. 39 (2016) 50–55.
[29] I. Dugueperoux, V. Scotet, M.P. Audrezet, A.H. Saliou, M. Collet, M. Blayau, et al.,
Non-visualization of fetal gallbladder increases the risk of cystic fibrosis, Prenat.
Diagn. 32 (2012) 21–28.
[30] Y. Ochshorn, G. Rosner, D. Barel, M. Bronshtein, F. Muller, Y. Yaron, Clinical
evaluation of isolated nonvisualized fetal gallbladder, Prenat. Diagn. 27 (2007)
699–703.
[31] B.S. Hertzberg, M.A. Kliewer, C. Maynor, P.J. McNally, J.D. Bowie, H.H. Kay, et al.,
Nonvisualization of the fetal gallbladder: frequency and prognostic importance,
Radiology 199 (1996) 679–682.
[32] F. Muller, F. Gauthier, J. Laurent, M. Schmitt, J. Boué, Amniotic fluid GGT and
congenital extrahepatic biliary damage, Lancet 26 (337) (1991) 232–233.
[33] M. Ben-Ami, Y. Perlitz, S. Shalev, I. Shajrawi, F. Muller, Prenatal diagnosis of
extrahepatic biliary duct atresia, Prenat. Diagn. 22 (2002) 583–585.