Indian childhood cirrhosis – down but not out: Report of a rare case with a practical clinicopathological diagnostic approachK Gaur1, P Sakhuja1, RN Mandal2, S Kapoor2
1 Department of Pathology, GB Pant Institute of Postgraduate Medical Education and Research, New Delhi, India
2 Department of Pediatrics, Maulana Azad Medical College, New Delhi, India
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/jpgm.JPGM_359_17
Source of Support: None, Conflict of Interest: None
Keywords: Biopsy, copper, rhodanine, Wilson disease
Indian childhood cirrhosis (ICC), a mysterious disease often presenting with rapidly progressive liver failure had a peak incidence in the 1970s and 80s. Until the 1980s it was the most common cause of chronic liver disease in Indian children. Currently, ICC cases are infrequent, some suggesting the condition to have vanished. This may be due to an epidemiological decline of the disease. The next generation of pathologists may also be unaware of the nuances of diagnosing this entity. The present case emphasizes that, in the absence of copper exposure, ICC cases may exist and must be considered. We have also highlighted a practical clinicopathological approach to diagnose this entity in modern day practice.
A 13-month-old, first born female child, born of nonconsanguineous mating presented with jaundice, high-grade, fever and passage of black-colored stools for three days. There was no history of hematemesis, decreased urinary output, seizures, pruritus, or indigenous drug intake. Antenatal, developmental, past, and family histories were unremarkable. The child was irritable, and had tachycardia (148/min) and pedal edema. Anthropometric indices of the child were normal for age [Weight, 9.4 kg (expected, 9.2 kg); height, 77 cm (expected, 75 cm; head circumference, 45 cm (expected, 45cm)], as tabulated in the relevant World Health Organization-Indian Academy of Pediatricians (WHO-IAP) combined growth chart for girls aged 0–18 years. The abdomen was distended, liver enlarged, firm-to-hard (7 cm below the costal margin), with a nodular, irregular surface, and the spleen was enlarged (3 cm below the costal margin) and firm. Signs of liver cell failure were absent. Other systems were unremarkable on examination.
Hematological workup revealed anemia (hemoglobin, 78 g/L; reference range, 115–135 g/L); elevated white blood cell count (WBC) 39 × 109/L (reference range, 5–17 × 109/L); neutrophilia and deranged prothrombin time (31 s; control, 11 s). Other findings were a normocytic normochromic peripheral blood with few schistocytes, a positive direct coombs test (DCT) (1+), and elevated reticulocyte count (7%, corrected, 3.5%; reference range, 0.8–2.2%). Serum haptoglobin, osmotic fragility/sickling/G6PD workup were unremarkable. Serum ferritin was elevated (194.9 μg/L; reference range, 5–25 μg/L). [Table 1] summarizes the results of major hematological and biochemical investigations displaying deranged parameters.
Viral (Hepatitis A, B, C, human Immunodeficiency virus, cytomegalovirus), autoimmune (antinuclear antibody, anti smooth muscle, anti liver kidney microsomal, anti liver cytosol type 1, anti soluble liver antigen antibodies), metabolic disease workup [urine gas chromatography-mass spectrometry (GC-MS), tandem mass spectrometry (TMS) for acyl carnitine and plasma amino acid profile, blood gas analysis] and thyroid profile (T3, T4, TSH) were unremarkable. Wilson's workup (Serum ceruloplasmin, urinary copper excretion, and slit lamp examination) were all within normal limits. Hepatic copper estimation could not be performed due to financial constraints.
Ultrasonography showed coarse liver echotexture, hepatosplenomegaly, increased periportal echogenicity, and moderate free fluid suggesting chronic liver disease (CLD). The gall bladder was visualized and the biliary system appeared patent. No cystic/mass lesion was apparent. Upper gastrointestinal endoscopy showed grade 2 esophageal varices.
An ascitic tap was performed and peritoneal fluid WBC count was 0.25 × 109/L (<0.1 × 109/L) with 90% polymorphs and albumin level of 10 g/L. Fluid culture showed pure growth of Staphylococcus aureus.
A working diagnosis of CLD with portal hypertension, spontaneous bacterial peritonitis, and possible autoimmune hemolytic anemia was made.
On admission, the child was put on a regimen of intravenous (IV) cefotaxime (450 mg) TDS for 5 days, IV furosemide (10 mg) BD, and oral aldactone (50 mg) BD (both 2 mg/kg/day). In addition, 225 ml N/4 saline in 5% dextrose with potassium chloride (1:100) was administered every 6 hours. Vitamin (A, B, C, D, E, K) and mineral (zinc, calcium) supplements were also given. Antibiotics were upscaled after 5 days to include IV amikacin (65 mg) BD and IV tazobactam (900 mg) every 8 hours as the patient showed little improvement. IV albumin and fluid restriction was initiated to manage increasing anasarca. The patient rapidly deteriorated in the next two days failing to respond to upscaled antibiotics and fluid restriction. A trial of empirical steroid (oral prednisolone) 10 mg BD produced transient improvement in ascites at which time a liver biopsy was done ensuring that the coagulation profile was normal at the time. However, after three days, the child succumbed to the disease. Consent for postmortem evaluation was declined.
Biopsied liver core tissue revealed loss of lobular architecture, hepatocyte ballooning, rosetting, lobular inflammation (6–8 foci/10 hpf), intrahepatocytic and canalicular cholestasis, and abundant Mallory hyaline. The portal tracts showed mild lymphocytic inflammation. Reticulin condensation with marked pericellular fibrosis, bridging fibrosis, and occasional nodule formation was noted [Figure 1]a and [Figure 1]b. Diffuse panlobular orcein positivity for copper-associated protein was noted. Marked copper deposition was confirmed on rhodanine stain [Figure 1]c and [Figure 1]d. Steatotic change, intranuclear inclusions, and giant cell transformation of hepatocytes were absent. Bile duct injury, inflammation, or periductal fibrosis was not seen. Perl's stain was negative for iron and Periodic acid Schiff (PAS) with diastase failed to reveal globular inclusions.
A young child with a negative viral, autoimmune, vascular, and metabolic workup as described here presented a major diagnostic dilemma. Female gender, transaminases in the range of 200–400 U/L, Coomb's positive hemolysis, and transient amelioration of symptoms after initiating steroids warranted the initial consideration of autoimmune hepatitis (AIH) despite the patient's age. Autoimmune profile, however, was negative.
Wilson's disease (WD) was also considered due to persistent hyperbilirubinemia. However, normal copper and serum ceruloplasmin did not support this diagnosis. No anatomical defects were noted on imaging. After three weeks, the diagnosis remained elusive. The liver biopsy had been withheld on account of the patient's acute symptomatology which was performed after a trial of steroids and antibiotic cover. Histologically, a cholestatic picture was observed. Marked orcein and rhodanine positive deposits with hepatocyte ballooning, marked pericellular fibrosis, Mallory hyaline, and nodule formation narrowed down the possibilities to WD and ICC.
ICC commonly occurs in the age group of 4 months to 5 years, as seen in this case. Cases of WD below 3 years are very rare and are generally asymptomatic. Features favoring ICC were normal serum ceruloplasmin levels, absence of neurological symptoms/KF rings, sharp liver margin on palpation, and rapidly progressive clinical course., Progressive liver failure though seen in both conditions is more characteristic of ICC. Aminoacidemia reflecting renal tubular dysfunction and hemolytic anemia may also be found in both. The histological picture of panlobular copper deposits, pericellular fibrosis, sparse regenerative activity, and diffuse Mallory hyaline favored ICC., Steatosis, glycogenated nuclei, irregular copper distribution, macronodules, and associated hemosiderin deposition, i.e., features more characteristic of WD, were absent in this case. Cholestasis is variable in both conditions and though supportive is not a sine qua non for histological diagnosis. The diffuse copper distribution pattern in ICC, however, is very characteristic and can be a useful pointer.
The needle biopsy with its limited sampling may be inappropriate to judge copper deposition, especially in cases of WD. However, the panlobular distribution seen here coupled with other diagnostic and clinical clues overwhelmingly shifted the diagnosis in favor of ICC. In view of the diffuse histological copper deposition seen in this case, a history of feeding via copper utensils was retrospectively enquired from the parents. No such history could be elicited. The child was also not exposed to copper through drinking water as the family dwelled in an urban colony where galvanization of water pipes was the norm. Genetic testing for ATP7B mutation was not performed in this case due to financial constraints.
Clinicopathological correlation: Role of pathological analysis
In the present case, a panel of serum-based diagnostic tests and imaging modalities failed to reveal the etiology of CLD. The liver biopsy eventually proved to be the most informative diagnostic test. This reiterates the fact that, in an era where clinicians are inclined to employ noninvasive testing modalities, liver biopsy is still an essential important component of the CLD workup. However, practical problems affecting the timing and technique of the biopsy may adversely affect the final outcome.
In this case, ascites, deranged prothrombin time, and peritonitis prevented a timely percutaneous liver biopsy. The biopsy was eventually performed when the patient showed transient improvement on steroid therapy. This may have been prevented by performing a transjugular liver biopsy (TJLB). However, pediatric TJLB requires a high level of expertise, niche specialization, and a well-equipped angiography suite infrastructure that is not available at many centres in the country. Histologically, despite the lack of any history suggesting copper ingestion, classical features suggesting ICC, i.e., creeping pericellular fibrosis, abundant Mallory hyaline, extensive copper deposition, and parenchymal nodules were present. Because the duration of steroid therapy given to the patient before the liver biopsy was less than a week (three days), it is unlikely that a masking of histological features on the liver biopsy may have occurred.
In a complex scenario, suspecting ICC as a possible etiology of CLD by the clinician is difficult. More so, with the prevalent impression that ICC is an “extinct” entity. A PUBMED search performed using the term “Indian childhood cirrhosis” in March 2017 yielded 329 results, of which 25 were recorded after 2000. This suggests that, though on the decline, it is certainly not extinct. Its exact pathogenesis remains elusive; the role of copper ingestion being the most debated. The 2006 Multicentre National Collaborative study (MCNS) study  found no credible evidence for the same, as seen in our case. Some authors, however, have maintained that copper plays a definite role in disease causation. Muller et al. investigated 138 cases of Tyrolean infantile cirrhosis (TIC), another non-Wilsonian “copper” toxicoses similar to ICC, and concluded that the disease was due to copper toxicosis and an inherited predisposition to cirrhosis. Interestingly, TIC has not been observed after 1974. Nayak and Chitale  recently emphasized that ebbing of ICC was more due to time-related sociocultural and environmental change and the inability to diagnose the entity. This viewpoint was corroborated by Patra et al.
Pathologists must be aware that, apart from the classical picture of established ICC, pre-cirrhotic disease may exist. This is characterized by extensive hepatocyte swelling and focal reticulin condensation in a hepatitic setting. Occasional pseudolobule formation and fibrous bands joining the portal tracts are other indicators of early disease. Ramakrishna et al. described an “atypical” copper cirrhosis presenting in older children and having a slower progression. Hence, classical ICC may be evolving to a more insidious variant which may escape diagnosis. This makes the timely study of biopsied liver tissue essential for instituting correct management. After studying the clinicopathological aspects of the present case, we suggest a relevant diagnostic algorithm to avoid underdiagnosing potential cases of ICC [Figure 2]. We acknowledge that though the nonperformance of postmortem analysis and molecular tests is a limitation of our work, such a scenario may be encountered frequently in routine hospital practice. The key point the present case highlights is that, while specialized tests for autoantibodies, WD, and metabolic disease are being done, a simultaneous liver biopsy must be performed at the earliest possible clinical opportunity.
ICC is uncommon but definitely not extinct. Cases occurring in the post copper utensil era assume significance. The alert clinician must be aware that such cases still exist and the pathologist has a key role to play in recognizing ICC and its variations. Timely biopsy diagnosis, continuous honing of pediatric biopsy skills, infrastructure establishment, and clinician–pathologist communication is essential to improve clinical outcome possibly preventing mortality.
The present work was performed after taking informed consent from the patient's parents and a sincere effort has been made to uphold patient confidentiality.
The authors would like to thank Eve A. Roberts, The Hospital for Sick children, Toronto for useful insights and guidance in preparing the manuscript.
[Figure 1], [Figure 2]