Etiological Spectrum of Chronic Liver Disease in Infants and Children: A Histopathology-Based Clinicopathological Analysis.

disease, autoimmune hepatitis, galactosaemia, tyrosinaemia and extrahepatic biliary atresia among them—carry outcomes that depend critically on the interval between presentation and diagnosis.[3,4]

The single most important organising principle in this field is age at presentation. In the first year of life, chronic hepatic injury presents almost invariably as neonatal cholestasis, a syndrome affecting approximately one in every 2500 live births, within which extrahepatic biliary atresia and the monogenic intrahepatic cholestatic disorders account for the majority of cases.[5] Biliary atresia alone remains the commonest indication for paediatric liver transplantation worldwide, and the success of the Kasai portoenterostomy falls sharply when the procedure is undertaken beyond 60 to 90 days of life, so that the diagnostic window is measured in weeks.[6,7] Beyond infancy the spectrum changes character entirely: metabolic disorders, chief among them Wilson disease, autoimmune liver disease, chronic viral hepatitis and, increasingly, non-alcoholic fatty liver disease come to the fore.[8,9]

Reports from the Indian subcontinent have consistently documented a spectrum distinct from that of Western series. Wilson disease is disproportionately represented, being reported as the commonest cause of non-cirrhotic chronic liver disease in children in some Indian centres, at a frequency considerably higher than that described in Europe or North America.[10,11] Conversely, the historically important entity of Indian childhood cirrhosis has all but disappeared following the decline in the use of brass and copper vessels for infant feeding, illustrating how the local aetiological profile evolves with changing environmental exposures.[12] Even within India the adult spectrum documented by multicentric survey bears little resemblance to the paediatric one,[13] and hospital-based paediatric series from different regions differ appreciably among themselves.[14] Such regional variation means that spectrum data generated elsewhere cannot be transposed uncritically to Indian practice.

Against this background, liver biopsy occupies a pivotal position. Non-invasive serology, imaging and biochemistry will identify the cause in many children, but a considerable residue remains in which the diagnosis is either unresolved or misattributed. Histopathology permits the direct recognition of pattern-defining features—ductular reaction and bile plugs in large duct obstruction, giant cell transformation in neonatal hepatitis, interface hepatitis with plasma cells in autoimmune disease, ductopenia in Alagille syndrome, and periodic acid–Schiff-diastase-resistant globules in alpha-1 antitrypsin deficiency—and simultaneously provides the stage of fibrosis, which carries prognostic weight independent of aetiology.[15,16] Its role has been questioned by authors who argue that careful clinical assessment can obviate the procedure in a majority of children,[10] making the actual diagnostic yield of biopsy a question worth measuring rather than assuming.

The present study was therefore undertaken to define the aetiological spectrum of chronic liver disease in infants and children on a histopathological basis, to correlate the morphological findings with the clinical and biochemical profile, and to quantify the extent to which liver biopsy modified the pre-biopsy clinical diagnosis.

interpreted by the Ferenci score where Wilson disease was suspected,[17] autoantibody profile (ANA, anti-smooth muscle antibody, anti-LKM-1) with serum IgG, serum alpha-1 antitrypsin, metabolic screening where indicated, abdominal ultrasonography with Doppler, and upper gastrointestinal endoscopy where portal hypertension was suspected. A provisional clinical diagnosis was recorded by the treating paediatric gastroenterologist before biopsy and sealed for later comparison with the histological diagnosis.

Biopsy procedure and processing: Ultrasound-guided percutaneous liver biopsy was performed under sedation using a 16- or 18-gauge automated cutting needle, targeting a minimum core length of 15 mm. Specimens were fixed in 10% neutral buffered formalin, processed routinely and sectioned at 4 µm. All cases were stained with haematoxylin and eosin, Masson trichrome, reticulin and periodic acid–Schiff with and without diastase digestion; Perls Prussian blue, orcein and rhodanine were applied for iron and copper-associated protein. Immunohistochemistry for HBsAg was performed where viral serology was positive, and CK7 and CK19 to assess ductular reaction and enumerate interlobular bile ducts where paucity was suspected.

Histopathological assessment: Slides were reported independently by two pathologists blinded to the sealed clinical diagnosis, with discordance resolved by joint review. Portal and lobular inflammation, interface hepatitis, ductular reaction, bile duct proliferation or loss, cholestasis, giant cell transformation, steatosis, ballooning and storage material were recorded semiquantitatively, and fibrosis was staged by the METAVIR system (F0–F4),[18] with the Ishak system applied in parallel to the chronic hepatitis cases.[19] The final aetiological diagnosis integrated the histological pattern with clinical, biochemical and radiological data.

Statistical analysis: Data were analysed in IBM SPSS Version 21.0. Continuous variables are expressed as mean ± standard deviation and categorical variables as frequencies and percentages. The chi-square or Fisher exact test was used for categorical comparisons, the independent t-test for continuous variables, and Cohen’s kappa for clinicopathological agreement. A p-value <0.05 was considered significant.

diagnosis of Wilson disease; it should be noted that a negative copper stain does not exclude the diagnosis, particularly in the pre-cirrhotic stage where copper distribution is uneven. Cytokeratin immunohistochemistry proved decisive in distinguishing ductular reaction from true bile duct proliferation and in confirming the three cases of ductal paucity.

Table 9: Concordance between pre-biopsy clinical diagnosis and final histopathological diagnosis (n=100)

Overall agreement between the sealed pre-biopsy clinical diagnosis and the final histopathological diagnosis was substantial (Cohen’s κ = 0.61). Biopsy nevertheless refined or overturned the working diagnosis in almost a quarter of children. The commonest such reassignments were from a clinical label of idiopathic neonatal hepatitis to extrahepatic biliary atresia and vice versa (7 cases), from cryptogenic cirrhosis to Wilson disease or autoimmune hepatitis (6 cases), and from suspected chronic hepatitis to congenital hepatic fibrosis (3 cases). The eight non-contributory biopsies comprised six inadequate cores and two in which non-specific changes precluded a definite aetiological assignment.

This histopathology-based analysis of 100 children with chronic liver disease demonstrates an aetiological spectrum that is broad, regionally characteristic and, above all, sharply stratified by age. The male preponderance (1.38:1) and the predominance of jaundice and hepatomegaly at presentation accord closely with the Indian series of Dhole et al, who reported a 60% male preponderance and jaundice with abdominal distension as the leading complaints in 55 children.[20] The finding that infancy is dominated by obstructive and cholestatic disease, with extrahepatic biliary atresia accounting for 35.3% of infantile cases, is consistent with the joint NASPGHAN–ESPGHAN estimate that biliary atresia underlies 25–40% of neonatal cholestasis.[5] The clinical importance of this figure lies in its time-dependence: outcomes after Kasai portoenterostomy deteriorate sharply beyond the first two to three months of life, and the incidence in Asian populations is higher than in the West.[6,7] That 5 of our 34 infants already had established cirrhosis at first biopsy is therefore a sobering observation, and echoes the recurring theme of delayed referral in resource-constrained settings, and the wider difficulty of achieving timely recognition that continues to define this disease.[21]

Beyond infancy, Wilson disease emerged as the single commonest aetiology, at 14% of the whole cohort and 21.2% of children over one year. This closely mirrors the 21.6% reported by Samanta et al from Kolkata,[10] and the 16–22% range described in other South Asian series,[11,22] a frequency substantially higher than in European cohorts. Consanguinity, founder effects in the ATP7B gene and a low threshold for ceruloplasmin testing plausibly all contribute. Because effective chelation exists and untreated disease is fatal, this prevalence argues for screening every Indian child above three years presenting with unexplained chronic liver disease, as the ESPGHAN position paper recommends.[3] Our histology supports one caveat in that document: copper-associated protein was demonstrable in only 12 cases, so a negative rhodanine stain cannot exclude the diagnosis. Autoimmune hepatitis accounted for 13% of cases, in keeping with the 10.8% of Samanta et al[10] and with the recognition that juvenile autoimmune liver disease is diagnosed more often than in the past, through both greater awareness and the relative decline of viral disease.[8,23] Indian paediatric experience has documented the same trend.[24] The plasma cell-rich interface hepatitis we observed is supportive rather than pathognomonic and requires integration with autoantibody and IgG data, as current guidance emphasises.[25] Chronic hepatitis B, at 8%, was far less frequent than the 24% reported by Hanif et al from Pakistan,[22] plausibly reflecting the maturation of universal infant immunisation in the interval.

The central methodological question this study addresses is whether biopsy justified itself. It did so in 24% of cases, where histology refined or overturned the clinical impression, most often in the diagnostically treacherous separation of biliary atresia from idiopathic neonatal hepatitis and in the reassignment of cryptogenic cirrhosis to a treatable metabolic or autoimmune cause. This is a more favourable verdict than that of Samanta et al, who concluded that careful senior clinical assessment could obviate biopsy in 78.3% of children.[10] The two positions are reconcilable: biopsy is dispensable in the majority precisely because clinical evaluation performs well, yet the minority in whom it changes the diagnosis are disproportionately those with treatable disease, where the cost of error is high. A residual cryptogenic rate of 10% despite full evaluation, comparable to the 13.5% inconclusive rate reported in that series, marks the current limit of morphology-based diagnosis and identifies the group in whom next-generation sequencing panels are likely to yield most.[2].

disorders, whereas beyond infancy metabolic and autoimmune causes, led by Wilson disease and autoimmune hepatitis, predominate. Advanced fibrosis was already present in more than half of the children at first biopsy, indicating substantial diagnostic delay. Liver biopsy, supported by a targeted panel of special stains and immunohistochemistry, confirmed the clinical impression in about two-thirds of cases and refined or altered it in a further quarter, most often in favour of a treatable diagnosis. Histopathology therefore retains a decisive role in the evaluation of the child with chronic liver disease, and a low threshold for biopsy is justified wherever the aetiology remains unresolved after first-line investigation.

Limitations: This was a single-centre study with a modest sample size and an inherent referral bias, since only children considered to require biopsy were enrolled; the true population spectrum of paediatric chronic liver disease is therefore not fully represented. Genetic confirmation by sequencing was not available for all suspected monogenic disorders, and the cross-sectional design precluded assessment of outcome.

Recommendations: Multicentre studies incorporating molecular diagnostics are needed to define the national spectrum more precisely and to resolve the cryptogenic group. Given the frequency of Wilson disease observed here, screening for it should be routine in every child above three years presenting with unexplained chronic liver disease, and sustained efforts towards earlier recognition of neonatal cholestasis remain a public health priority.

19.    Ishak K, Baptista A, Bianchi L, Callea F, De Groote J, Gudat F, et al. Histological grading and staging of chronic hepatitis. J Hepatol. 1995;22(6):696–9.

20.    Dhole SD, Kher AS, Ghildiyal RG, Tambse MP. Chronic liver diseases in children: clinical profile and histology. J Clin Diagn Res. 2015;9(7):SC04–7.

21.    Bezerra JA, Wells RG, Mack CL, Karpen SJ, Hoofnagle JH, Doo E, et al. Biliary atresia: clinical and research challenges for the twenty-first century. Hepatology. 2018;68(3):1163–73.

22.    Hanif M, Raza J, Qureshi H, Issani Z. Etiology of chronic liver disease in children. J Pak Med Assoc. 2004;54(3):119–22.

23.    Maggiore G, Nastasio S, Sciveres M. Juvenile autoimmune hepatitis: spectrum of the disease. World J Hepatol. 2014;6(7):464–76.

24.    Yachha SK, Srivastava A, Chetri K, Saraswat VA, Krishnani N. Autoimmune liver disease in children. J Gastroenterol Hepatol. 2001;16(6):674–7.

25.    Mack CL, Adams D, Assis DN, Kerkar N, Manns MP, Mayo MJ, et al. Diagnosis and management of autoimmune hepatitis in adults and children: 2019 practice guidance and guidelines from the American Association for the Study of Liver Diseases. Hepatology. 2020;72(2):671–722.