Indian Pediatrics 2002; 39:421-425  

Neonatal Cholestasis Syndrome– Identifying the Disease from Liver Biopsy

Persistent jaundice in the newborn and young infant was first recognized as a major challenge in pediatric practice two and a half centuries ago(1). During these years clinical studies and application of emerging techno-logies have helped provide only some solutions to the problems of identifying and treating this not uncommon complex condition of Neonatal Cholestasis Syndrome (NCS). The global incidence of NCS is estimated at 1 in every 10,000 live births (2) and today NCS appears to account for approximately a fifth of all liver diseases encountered in tertiary pediatric health care centers(3). The average frequencies of diseases causing this syndrome are: biliary atresia (BA) with extra and/or intrahepatic biliary tract anomalies - 25 to 40%; idiopathic neonatal hepatitis (INH) of uncertain and variable, possibly infective etiology - 44%; metabolic diseases including alpha-1-antitrypsin deficiency (A1AD)- 8%; paucity of intrahepatic bile ducts, including Alagille’s syndrome - 4.5%; and a recently described entity, progressive familial intrahepatic cholestasis - 2%(1,4). The first two, namely BA and INH, are not only the most important ones causing NCS but also pose the greatest diagnostic dilemma. The differentiation between BA and INH is all the more crucial since the former needs urgent surgical intervention. In the USA alone, four to six hundred new cases of BA are encountered yearly (5).

Biliary atresia, at the time of clinical presentation is primarily an anatomical abnormality of the relatively larger intra-hepatic biliary channels and/or exrahepatic biliary passages. Advanced imaging proce- dures like endoscopic retrograde cholangio-pancreatography (ERCP), computerized tomography (CT), magnetic resonance cholangiopancreatography (MRCP) and endo-scopic ultrasonography (EUS) have, however as yet, failed to contribute significantly in the diagnosis of this disease because of the age of the subject and the small size of the passages. The need for an accurate diagnosis of BA soon after clinical manifestations is important as response to surgical results will be best if the operation is performed around 8 weeks of age (4,6). Of the infants with BA who undergo Kasai’s hepatic portoenterostomy, more than 80 percent become jaundice free if the surgery is done before 60 days of age as against only 25 to 35% of infants operated upon at a later age (6). Postoperative long-term survival with good quality life is also far better in infants with surgical intervention at an early age(7,8). In older infants with obstructive biliary tract disease of longer duration, irreversible changes in the liver occur and organ trans-plantation offers the only possible remedy. Since age is the single most important determinant in successful management of BA, recognition and definitive identification of the condition as the cause of NCS in a given case very early after onset of symptoms is of paramount importance. Unfortunately, at that young age diagnosis of anatomical abnormali-ties of biliary passages is extremely difficult to assess by imaging procedures. That leaves histological evaluation of liver biopsy as the principal modality of identifying biliary atresia. However, hepatic morphological alterations in BA are closely similar to and often indistinguishable from those in INH. In other diseases causing NCS, structural changes are more specific or even diagnostic, particularly in conjunction with biochemical, metabolic and genetic data. Thus, the biggest challenge in histological diagnosis of diseases causing NCS is the differentiation between BA and INH.

Making a definitive diagnosis of BA versus INH on the basis of hepatic histology has been variably reported from being almost impossible in early infancy (9), though being considerably difficult with many overlapping features (10,11), to having an accuracy as high as 96 percent(3). A recent two stage study(12) based on controlled statistical grounds first identified a set of histological parameters that served as powerful predictive variables in differentiating BA from INH and when applied to a further series of new samples of liver biopsy yielded an accuracy of 90.5 % and specificity of 75.9%. The authors identified 6 morphological discriminators whose cumula-tive scores helped distinguish BA from INH. These include bile ductal and ductular proliferation in portal tracts, bile plugs in portal ductules, portal to portal bridges, neutrophils, hepatocyte swelling and multi-nucleate giant hepatocytes(12). In a later report(13) the histopathological distinction between intra- and extra-hepatic neonatal cholestatic disease was based on the following discriminant features in decreasing order of importance: periportal ductular proliferation, portal ductal proliferation, portal expansion, neoductular cholestasis, foci of myeloid meta-plasia and portal to portal bridges. Myeloid metaplasia was considered an important feature of intrahepatic cholestasis namely INH and other non-BA diseases. One of us (NCN) has had the opportunity of studying during a 5 year period from 1983 to 1988 a little over 60 percutaneous or intraoperative liver biopsies from infants with NCS in whom the primary diagnostic objective was to decide between BA and INH. Over the last 3 years we have supplemented this experience on 15 additional identical material several of which are from young infants between the ages of 6 to 10 weeks. In all these earlier and current material unbiased observations of various morpho-logical features in the liver tissue were made on a consensus by two observers before knowing whether the case was finally categorized as BA or INH on the basis of peroperative cholangiographic and/or unequi-vocal HIDA scan data. The cumulated histo-logical informations were then compared to assess similarities and differences between cases of BA and INH with regard to portal changes namely, hyperplasia of bile ducts and ductules, ductular cholestasis, portal fibrosis, inflammation and cholangitis as well as lobular changes like hydropic change of liver cells, giant hepatocytes, cholestasis, inflamma-tion and hemopoiesis.

In general, three lesions in the portal tract emerge as far more discriminatory than most lesions except one in the lobular parenchyma. In the latter site the severity and extent of hepatocellular, canalicular and intramacro-phage cholestasis and hydropic change of hepatocytes are equally variable in both diseases. Multinucleated giant hepatocytes with intracellular cholestasis and frequent hydropic change are seen in both conditions and, in fact in our experience, these giant cells can be more prominent in cases of BA, particularly in younger infants. Thus, the term ‘giant cell hepatitis’ is best abandoned. Lobular inflammation with infiltration pre-dominantly by mononuclear cells and much less by neutrophils is minimal to moderate in BA and mild to marked in INH. The wide area of overlap does not allow this parameter to be used as a suitable discriminator for INH which is believed to be primarily an inflammatory disease. Presence of hemopoietic foci is mild to marked in INH but only none to minimal in BA. This feature, therefore, can act as one indicator in deciding between BA and INH. Portal tracts are widened and variably fibrosed in both diseases. In very young infants these lesions cannot distinguish INH from BA since in the latter disease it is too early for advanced fibrosis and significant bile duct hyperplasia to have occurred. Persistence of disease through advancing age of the infant in BA significantly augments the fibrosis in portal areas and soon irregular and wide portal to portal bridges develop, ultimately leading to a ‘jigsaw puzzle’ type of picture characteristic of secondary biliary cirrhosis. This is unusual in INH. Since extralobular biliary passages are the primary seat of disease in BA, compensatory hyperplasia of bile ducts in the portal tracts appear earlier and is more marked than in INH. Ductular hyperplasia at the portal tract- lobule junction occurs in both diseases but bile plug in the ductules appears to be characteristic of BA. In INH as well as in BA diffuse inflammatory cell infiltration of portal tracts comprising mostly of mononuclear cells is common. On the other hand cholangitis consisting of neutrophilic infiltration of bile ducts is relatively rare in both diseases. An ultrastructural study reports mostly bile duct damage in BA and mostly hepatic injury in INH with many instances of overlap of these lesions(14). Our observations largely support and complement findings reported in the two recent studies on histopathological diagnosis of BA and INH(12,13). Clearly, no single parameter or combination there of is absolutely diagnostic of either disease. In experienced hands, however, there is a high probability of arriving at a correct diagnosis on a careful semiquantitative assessment of the sum total of several important histological features. An approximate evaluation of these parameters in BA and INH is listed in Table I. It is apparent that while there is some or more overlap in most parameters, those constituting helpful discriminant are ductular bile plug, ductal and ductular proliferation, hemopoietic foci and fibrosis. In 90% of our cases ductular bile plugs and more than occasional hemopoietic foci appear to be relatively strong pointers to BA and INH respectively.

Being a complex condition with consider-able diagnostic difficulty, cases of NCS need to be evaluated only at specialized centers with attention from skilled and experienced professionals both at clinical and other investigational levels. It has been our experience and of some others with access to large material (Duttagupta, S., Personal communication) that even with the best of facilities, the accuracy in histological diagnosis between BA and INH in relatively young infants is between 85 to 90 percent. Obviously at less equipped centers this figure can be considerably lower. Since a large share of relevant histological information is obtained from portal tracts, it is important that the liver biopsy tissue should have enough of these elements before a logical interpretation is made. It has been shown that a length of approximately 1cm. needle biopsy from the adult shows 6 portal tracts(15), which is adequate for evaluation. In the adult there are on an average 3 bile ducts per portal tract(15). It is well known that in hepatocytic diseases, prolonged for some time, continuing damage to periportal parenchyma results in some ductular proliferation. Also in very young infants the morphological differentiation of newly formed bile ducts from proliferated ductules can be difficult. More accurate quantification of bile duct and ductular proliferation therefore may be needed to provide a better discriminator between BA and INH. Certainly, more studies on histopathological changes in well documented cases with proper qualitative and quantitative assessment are needed to decide on criteria that will help to raise the diagnostic accuracy level in these two conditions, particularly in infants younger than 10 weeks age who are most likely to benefit by a decision between surgical intervention and conservative management.

1. Bansal S, Dhawan A. Neonatal cholestasis syndrome - the saga continues. Indian Pediatr 2000; 37: 827-830.

2. Bates NO, Buccavalas JC, Rickman HC. Biliary atresia: pathogenesis and treatment. Semin Liver Dis 1998; 18: 281-293.

3. Yachha SK, Khanduri A, Kumar M, Sikora SS, Saxena R, Gupta RK, Kishore J. Neonatal cholestasis syndrome : an appraisal at a tertiary centre. Indian Pediatr 1996; 33: 729-734.

4. Yachha SK, Mohindra S. Neonatal cholestasis syndrome: Indian scene. Indian J Pediatr. 1999; 66: 894-896.

5. Lefkowitch JH. Biliary atresia. Mayo Clin Proc 1998; 73: 90-95.

6. Mieli-Vergani G, Howard ER, Portman B, Mowat AP. Late referral for biliary atresia -missed opportunities for effective surgery. Lancet 1989; I: 421-423.

7. Howard ER, Mowat AP. Extra hepatic biliary atresia: recent developments in management. Arch Dis Child 1977; 52: 825-827.

8. Ohkochi N, Chiba T, Endo N, Goto M, Ibrahim M. Long term follow up after surgery for patients with biliary atresia. J Pediatr Gastroenterol Nutr 1989; 9: 416-420.

9. Machnik G. Histological changes in liver tissue in cholestasis. J Gastroenterol 1993; 31 suppl 2: 7-10.

10. Vecchione R, Terracciano LM, D’Armiento M, D’Armiento FP. Neonatal cholestasis: the view point of the pathologist. Pediatr Med Chir 1993; 15: 229-237.

11. Agarwala S, Mitra DK. Biliary atresia- the current management. Indian J Pediatr 1996; 63: 719-724.

12. Zerbini MC, Gallacci SD, Maezono R, Ueno CM, Porta G, Maksood JG, Gayotto LC. Liver biopsy in neonatal cholestasis: a review on statistical grounds. Mod Pathol 1997; 10: 793-799.

13. Santos JL, Almeida H, Cerski CT, Silverira TR. Histopathological diagnosis of intra and extrahepatic neonatal cholestasis. Braz J Med Biol Res 1998; 31: 911-919.

14. Park WH, Kim SP, Park KK, Shoi SO, La J, Kaon KY. Electron microscopic study of the liver with biliary atresia and neonatal hepatitis. J Pediatr Surg 1996; 31: 367-374.

15. Crawford AR, Lin XZ, Crawford JM. The normal adult human liver biopsy: a quantitative reference standard. Hepatology 1998; 28: 323-331.