Role of Hepatobiliary Scintigraphy and Preoperative Liver Biopsy for Exclusion of Biliary Atresia in Neonatal Cholestasis Syndrome
All diagnostic algorithms for Neonatal Cholestasis Syndrome (NCS) focus on differentiating numerous medical causes from Biliary Atresia (BA). No preoperative diagnostic algorithm has 100% diagnostic accuracy for BA and yet, timely diagnosis is crucial to optimize surgical outcome. Markers for high index of clinical suspicion for BA are: a “usually” well thriving infant with conjugated hyperbilirubinemia, raised gamma glutamyl transpeptidase, persistently “acholic” stools, firm hepatomegaly with dysmorphic, hypoplastic gall bladder. In the presence of above ‘red flag’ signs, there has been much debate on diagnostic accuracy of percutaneous liver biopsy (PLB) vs. hepatobiliary scintigraphy (HBS) to substantiate or exclude BA. Recent guidelines suggest a shift towards PLB (91.6% overall diagnostic accuracy) as the diagnostic cornerstone with key differentiating feature being ‘bile ductular proliferation’. HBS has a high (98.7%) sensitivity but low specificity (37–74%) with an overall diagnostic accuracy of 67% for BA. Severe hepatocellular disease without anatomic obstruction would also have a non-excretory scan. Thus, while excretory HBS excludes BA, non-excretion does not confirm BA. Hence, diagnostic algorithms relying on non-excretory HBS as the primary standalone benchmark for surgical exploration would be mired by a high negative laparotomy rate revealing a normal peroperative cholangiogram (POC). However, an excretory HBS obviates need for laparotomy in case of equivocal stool color or PLB. A POC continues to be the ultimate gold standard. Hence, with high index of clinical suspicion but equivocal ultrasonography or PLB and a non-excretory HBS, the baby should not be denied a POC within time frame crucial for successful hepatoportoenterostomy.
KeywordsNeonatal cholestasis Biliary atresia Hepatoportoenterostomy Liver biopsy Radionuclide scan
Cholestatic jaundice is always pathological because it signifies hepatocellular dysfunction. Recent (2017) joint recommendations of the North American Society for Pediatric Gastroenterology, Hepatology and Nutrition (NASPGHAN) and European Society for Pediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) state- any infant noted to be jaundiced after 2 wk of age be evaluated for cholestasis with measurement of total and direct serum bilirubin and that an elevated direct serum bilirubin (direct bilirubin levels >1.0 mg%, >20% of total bilirubin) warrants specialist attention .
The incidence of Neonatal Cholestasis Syndrome (NCS) is approximately 1 in 2500 live births . A consensus report on NCS prepared under the aegis of Pediatric Gastroenterology Subspeciality chapter of Indian Academy of Pediatrics analysed the distribution of the causes of NCS in the Indian scenario as follows : i) Hepatocellular causes: 53% [neonatal hepatitis (NH) 47%, metabolic 4%, varied etiology 2%]; ii) obstructive causes: 38% [biliary atresia (BA) 34%; choledochal cyst 4%]; iii) ductal paucity: 3%; iv) idiopathic: 6%. NH group comprised the following: idiopathic giant cell hepatitis (64%), TORCH infections (22%), sepsis (8%), urinary tract infections and other infections (6%). Metabolic group comprised of the following: galactosemia (35%), α-1 antitrypsin deficiency (33%), parenteral nutrition (PN) related cholestasis (19%), tyrosinemia (7%), storage disorders (4%), hemochromatosis (2%). The group of varied etiology due to hepatocellular cause were constituted by inspissated bile syndrome, recurrent intrahepatic cholestasis, progressive familial intrahepatic cholestasis and hypothyroidism. Thus, BA and NH together account for nearly 70% of NCS [3, 4].
The focus of all diagnostic algorithms for NCS is to differentiate extensive list of medical causes from BA which is a surgical semi-emergency.
The challenges in the diagnosis of BA are related to: i) well thriving babies and hence draw medical/parental attention late; ii) difficulty in distinguishing from the more common physiological jaundice of infancy, breast milk jaundice and a long list of medical causes of NCS; iii) no preoperative diagnostic algorithm has 100% diagnostic accuracy; iv) timely diagnosis is crucial to optimize response to hepatoportoenterostomy (HPE) aimed at establishing bile flow; v) a negative laparotomy wherein a peroperative cholangiogram (POC) reveals normal biliary anatomy adds to unnecessary morbidity and hence, the significance of an accurate preoperative diagnosis.
The diagnostic algorithm used to differentiate medical causes of NCS from BA include a detailed history and physical examination, daily inspection of stool color, liver function tests including gamma glutamyl transpeptidase (GGTP), fasting abdominal ultrasonography (USG), percutaneous liver biopsy (PLB) and hepatobiliary scintigraphy (HBS). The diagnostic algorithm has varied from centre to centre depending on the expertise available and no algorithm has a 100% preoperative diagnostic accuracy. The caveat, though, is that no BA cases are missed while a few non-BA cases with equivocal preoperative investigations may be subjected to laparotomy/laparoscopy wherein, a surgical cause is excluded or confirmed by a peroperative cholangiogram (POC).
This article reviews relevant literature on the role of percutaneous liver biopsy (PLB) and HBS on excluding the diagnosis of BA preoperatively which is the most crucial question to be addressed by a pediatric gastroenterologist in the work up of a baby with NCS.
The markers for high index of clinical suspicion for BA are: typically a well thriving full term infant with conjugated hyperbilirubinemia starting from the first few weeks of life with ‘persistently’ pale stools and firm hepatomegaly. Parents may often misidentify abnormally pale stools as normal. Direct visualization of at least 3 stool samples for stool pigment by an experienced health care professional is a crucial key aspect of complete evaluation of a jaundiced infant. Early identification of acholic stools has been reported to be facilitated by use of ‘stool color card’ with 95.2% sensitivity [5, 6, 7].
On the other hand, clinical features which may tilt the balance towards heterogenous medical causes of NCS are: a) a family history of consanguinity, NCS in siblings, prior repeated fetal loss, b) maternal infections (TORCH); c) an infant with history of prematurity, small for gestational age, neonatal sepsis or parenteral nutrition; d) a ‘sick baby’ with failure to thrive or neonatal liver decompensation; e) associated anomalies like facial dysmorphism, pulmonary stenosis. Splenomegaly appears in BA usually after 4 wk; early appearance of splenomegaly around 2–4 wk of age may point towards storage or hemolytic disorder.
BA is characterized by conjugated hyperbilirubinemia (direct bilirubin fraction of >1 mg% or >20% of total bilirubin). Of the accompanying liver function tests, BA babies have been consistently reported to have a higher GGTP levels than those without BA (902.7 mmol/l vs. 263.2 mmol/l) . An elevated GGTP more than 300 IU/L has been reported to have a specificity of 98% and sensitivity of 38% in differentiating BA from NH . However, raised GGTP would also be seen in infantile choledochal cysts (CDC), inspissated bile syndrome, PN related cholestasis, paucity of intralobular bile ducts, neonatal sclerosing cholangitis and α-1 antitrypsin deficiency. On the other hand, NCS causes characterized with normal or low GGTP are: progressive familial intrahepatic cholestasis (PFIC) type 1 and 2 and bile acid synthesis disorders . While no component of serum biochemistry is truly pathognomonic, the pointers towards BA are: conjugated hyperbilirubinemia with significantly elevated GGTP and initially well preserved synthetic liver function tests.
A fasting abdominal ultrasonography performed by an experienced ultrasonologist using a high frequency transducer probe is a useful non-invasive initial screening investigation for NCS. It excludes other structural anomalies like CDC, inspissated bile syndrome and spontaneous biliary perforation. In addition, infants with cystic variant of BA would show a ‘cyst at porta’ which may be differentiated from cystic choledochal malformation on the basis of a hypoplastic gall bladder and absence of intrahepatic biliary dilatation in BA . It needs to be emphasized that BA is characterized by absence of intrahepatic biliary dilatation unlike other causes of surgical jaundice. Also, a ‘cyst at porta’ on USG is a definite pointer towards a surgical cause of NCS (a cystic BA variant or CDC) and obviates need for a liver biopsy.
The overall accuracy of USG in differentiating BA from non-surgical causes of NCS is 75.0% [63.3% sensitive, 86.7% specific, positive predictive value (PPV) of 82.6% and negative predictive value (NPV) of 70.3%] .
Hypoplastic or undetectable gall bladder defined as gall bladder size of <19 mm in a baby fasting for >4 h (sensitivity 83.3%, specificity 82.6%, PPV 67.6%, NPV 91.9%, accuracy 82.8%),
‘Triangular cord sign’ defined as echogenicity anterior to the right portal vein >4 mm (sensitivity 23.3%, specificity 97.1%, PPV 77.8%, NPV 74.4%, accuracy 74.7%),
Abnormal or dysmorphic gall bladder, also described as a “ghost gall bladder” (sensitivity 83.3%, specificity 82.6%, PPV 67.6%, NPV 91.9%, accuracy 82.8%),
Lack of gall bladder contraction after oral feeding (gall bladder contractility <86% ± 18% (mean ± SD) in patients younger than 12 wk or <67% ± 42% in patients 12 wk and older) (sensitivity 87%, specificity 72.5%, PPV 51.3%, NPV 94.3%, accuracy 76.1%),
Non visualization of the common bile duct (sensitivity 93.3%, specificity 47.8%, PPV 43.8%, NPV 94.3%, accuracy 61.6%),
Right hepatic artery diameter > 1.5 mm (sensitivity 92%, specificity 87%, accuracy 89%),
Right hepatic artery diameter to right portal vein diameter ratio > 0.45 (sensitivity 76%, specificity 79%, accuracy 78%),
Hepatic subcapsular blood flow defined as hepatic arterial flow extending to the hepatic surface (sensitivity 100%, specificity 86%, PPV 85%, NPV 100%).
In addition, preoperative USG may diagnose features of Biliary Aresia Splenic Malformation (BASM). These are: situs inversus, polysplenia, asplenia, and preduodenal portal vein.
The pitfalls of an ultrasonography however are: i) a hypoplastic gall bladder could be a consequence of severe intrahepatic cholestasis with poor bile flow; ii) about 20% of type 3 BA babies who have a patent gall bladder and distal bile duct would have well distended gall bladders (a gall bladder mucocele) with post feed contractility.
The NASPGHAN and ESPGHAN guidelines (2017) on the role of abdominal USG in the work-up for BA have been summarized as – abdominal ultrasonography is useful in excluding choledochal cyst or gallstone disease causing extrahepatic bile duct obstruction. It may demonstrate an absent or abnormal gall bladder or other features suggestive, but not diagnostic of BA .
In the presence of clinical features, serum biochemistry and an abdominal USG suggestive of BA, the next investigation to substantiate or exclude the diagnosis of BA has been a percutaneous liver biopsy (PLB) or a hepatobiliary scintigraphy (HBS) or both.
A PLB be performed in most infants with undiagnosed cholestasis. If the results are equivocal and biopsy was performed when the infant was <6 wk of age, a repeat biopsy may be necessary.
In the hands of an experienced pediatric pathologist, histopathological findings of bile duct proliferation, bile plugs, and fibrosis in an appropriately timed liver biopsy is the most supportive test in evaluation of an infant with protracted conjugated hyperbilirubinemia. Diseases other than BA that cause cholestasis can be determined via histological examination of the liver.
The other specific etiologies of NCS which would be diagnosed on PLB are: paucity of intralobular bile ducts, α-1 antitrypsin deficiency and storage disorders like Niemann Pick deficiency.
Pathologists participating in the National Institutes of Health-sponsored Biliary Atresia Research Consortium (BARC) developed and evaluated a standardized system for histologic reporting of liver biopsies for NCS . Histologic features that best predicted BA, on the basis of logistic regression, included bile duct proliferation, portal fibrosis, and absence of sinusoidal fibrosis. Thus while there may be an overlap of histopathological features between BA and NH, there is consensus that ‘bile ductular proliferation’ as the histopathological feature has the highest overall sensitivity, specificity, NPV, PPV and accuracy in predicting BA [17, 18, 19]. Other supportive features are: bile plugs, widening of portal tracts because of portal or perilobular fibrosis, and edema, with preservation of the basic hepatic lobular architecture. On the other hand, idiopathic neonatal hepatitis is characterized by more prominent multinucleated giant cell transformation, lobular disarray, inflammatory cells within the portal areas, while bile ductules show little or no alteration .
Adequacy of biopsy specimen, which should ideally have 5–7 portal tracts because features differentiating BA from other causes of NCS are best appreciated in the portal tracts.
Age at obtaining liver biopsy: ‘Bile ductular proliferation’ as the key histological feature predicting BA may not manifest in a biopsy performed too early (<4 wk of age) and may be a reason for missed diagnosis. Hence, recent studies recommend performing biopsy between the critical window of 6 to 8 wk of age to obtain accurate and prompt diagnosis and also the best chance of establishing biliary flow post portoenterostomy [16, 19, 21, 22, 23, 24].
Overlap in histopathological features between NH and BA: Overlap features as giant cell transformation, bile plugs, hepatocellular necrosis, portal and lobular inflammation may confound histolopathological differentiation between BA and NH and hence the significance of an appropriately timed liver biopsy with adequate number of portal tracts interpreted by an experienced pathologist.
Complications: The reported complications are pain, hemoperitoneum, hemothorax pneumothorax, and pleural effusion with rates ranging from <1% to 2% with a mortality rate of 0.016% [25, 26]. There is no statistically significant difference in complications between blind vs. ultrasound guided biopsies. This has a positive practical relevance for developing countries with large patient load and constrained interventional radiology support [25, 27]. While preoperative liver biopsy is an invasive diagnostic method, the actual risks in experienced hands are minimal and the benefits (such as avoiding a negative laparotomy or achieving a specific diagnosis other than BA) outweigh the risks.
Overlap of histological features in the portal tracts in inspissated bile syndrome, cystic fibrosis, α-1 antitrypsin deficiency, infantile choledochal cyst and other causes of biliary obstruction: Hence, the biopsy has to be interpreted in the light of clinical picture, biochemistry and imaging.
In a recent meta-analysis on the value of preoperative liver biopsy in the diagnosis of BA , a sensitivity of 91.2%, specificity of 93.0% (n = 1231), PPV of 91.2%, NPV of 92.5% (n = 1182), and accuracy of 91.6% (n = 1106) was reported.
However, if there is a high index clinical suspicion of BA but a non-suggestive/equivocal liver biopsy, the infant should not be denied the opportunity of a peroperative cholangiogram followed by a HPE (if indicated) within the crucial time frame wherein it is expected to result in successful bile flow.
Hepato-Biliary Scintigraphy (HBS)
Protocol for HBS in NCS entails pre-treatment with phenobarbital (5 mg/kg/d in 2 divided doses) and ursodeoxycholic acid (10–20 mg/kg/d in 2 divided doses) for 5 d. The most popular radiopharmaceutical is 99m–Technetium trimethylbromo-iminodiacetic acid (Mebrofenin) that is transported to the liver bound to albumin, actively extracted by hepatocytes and then excreted into the biliary system . After intravenous administration of 1 mCi of radiopharmaceutical, sequential 1-min anterior images of the abdomen are obtained for 1 h. Thereafter, static anterior and right lateral images of the abdomen are obtained at 2 h, 4 h, 6 h and 8 h till biliary excretion is demonstrated or up to a maximum of 24 h post-injection. Presence of biliary excretion excludes BA. Absence of biliary excretion supports the diagnosis of BA but may also be present in severe neonatal hepatitis, interlobular bile duct paucity, alpha I antitrypsin deficiency, inspissated bile duct syndrome, cystic fibrosis and hence, does not confirm the diagnosis of BA [28, 29, 30, 31, 32]. Though the sensitivity of HBS for the diagnosis of BA is high, the specificity of HBS for BA is low (37%–74%) with an overall diagnostic accuracy of 67% [28, 29, 30, 31, 32].
A recent meta-analysis addressing the utility of HBS yielded a pooled sensitivity of 98.7% (98.1–99.2%) and a specificity of 70.4% (range 68.5%–72.2%) of a non-draining HBS for excluding BA . The false negative results (excretion of the tracer into the bowel despite BA) were reported to be extremely rare.
NASPGHAN and ESPGHAN (2017) recently concluded that its limited specificity precludes the use of the HBS scan as a standalone test in making a definitive diagnosis of BA; definitively demonstrated bile flow by selective use of HBS maybe of value in excluding BA .
The clinical implication of the above is that diagnostic algorithms relying on a non excretory HBS as the primary or standalone benchmark for surgical exploration would be mired by a high negative laparotomy rate wherein a peroperative cholangiogram may reveal a normal biliary tree. HBS is of practical significance wherein an excretory scan would obviate a need for a laparotomy in a baby with equivocal stool color/ liver biopsy.
There has been much debate on the overall diagnostic accuracy of a PLB and HBS to substantiate or exclude the diagnosis of BA. Recent guidelines suggest a shift towards a PLB as the cornerstone for preoperative diagnosis; the key differentiating histopathological feature in BA is ‘bile ductular proliferation’. HBS has a low specificity and overall diagnostic accuracy for BA; many patients without anatomic obstruction but with severe hepatocellular disease would have a non excretory scan, thereby implying that while excretion excluded BA, non excretion does not confirm BA. Diagnostic algorithms based on a non-excretory HBS as a standalone benchmark for surgical exploration would be mired by a high negative laparotomy rate. However, HBS has practical significance wherein an excretory scan obviates need for a laparotomy in a baby with equivocal stool color or liver biopsy. However, these investigations are not a substitute for the clinical judgement. The peroperative cholangiogram continues to be the ultimate gold standard. Hence, in the presence of high index clinical suspicion but a non supportive/equivocal PLB and a non excretory HBS, the baby should not be denied the opportunity of a peroperative cholangiogram within the crucial time frame wherein a HPE is expected to result in successful bile flow.
The authors deeply acknowledge the contribution of Prof. S K Yachha, Prof Rakesh Pandey, Prof Ujjal Poddar, Dr. Anshu Srivastava, and the entire Department of Pediatric Gastroenterology & Pathology at SGPGIMS for their tremendous clinical and academic contributions to the arena of NCS at our centre.
RL: Conception & design, interpretation & analysis of literature, revised the manuscript critically for important intellectual content; AM: Drafting the manuscript; NM: Assisted in drafting the manuscript. RL will act as guarantor for the paper.
Compliance with Ethical Standards
Conflict of Interest
Source of Funding
- 1.Fawaz R, Baumann U, Ekong U, et al. Guideline for the evaluation of cholestatic jaundice in infants: joint recommendations of the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition and the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition. J Pediatr Gastroenterol Nutr. 2017;64:154–68.Google Scholar
- 3.Indian Academy of Pediatrics. Consensus report on neonatal cholestasis syndrome (NCS). Pediatric Gastroenterology subspecialty chapter of Indian Academy of Pediatrics. Indian Pediatr. 2000;37:845–51.Google Scholar
- 5.Chen SM, Chang MH, Du JC, et al; Taiwan Infant Stool Color Card Study Group. Screening for biliary atresia by infant stool color card in Taiwan. Pediatrics. 2006;117:1147–54.Google Scholar
- 15.Moyer V, Freese DK, Whitington PF, et al. Guideline for the evaluation of cholestatic jaundice in infants: recommendations of the North American Society for Pediatric Gastroenterology, Hepatology and Nutrition. J Pediatr Gastroenterol Nutr. 2004;39:115–28.Google Scholar
- 16.Russo P, Magee JC, Boitnott J, et al; Biliary Atresia Research Consortium. Design and validation of the biliary atresia research consortium histologic assessment system for cholestasis in infancy. Clin Gastroenterol Hepatol. 2011;9:357–62. e2.Google Scholar
- 22.Park WH, Choi SO, Lee HJ, Kim SP, Zeon SK, Lee SL. A new diagnostic approach to biliary atresia with emphasis on the ultrasonographic triangular cord sign: comparison of ultrasonography, hepatobiliary scintigraphy, and liver needle biopsy in the evaluation of infantile cholestasis. J Pediatr Surg. 1997;32:1555–9.CrossRefPubMedGoogle Scholar
- 32.Yang JG, Ma DQ, Peng Y, Song L, Li CL. Comparison of different diagnostic methods for differentiating biliary atresia from idiopathic neonatal hepatitis. Clin Imaging. 2009;33:439–46.Google Scholar