Biliary Atresia Biomarkers 2020

Two recent studies provide more information on biliary atresia (BA) biomarkers.

  • OG Behairy et al. JPGN 2020; 70: 344-9.
  • S Shamkar et al. JPGN 2020; 70: 350-5.

Behairy et al report on the use of serum IL-33 in a cohort of 90 infants, 30 with BA, 30 with cholestasis due to other causes, and 30 healthy infants.

  • Using a cut-off of 20.8 pg/mL, IL-33 had a specificity of 95% and sensitivity of 96.7% for identifying BA.
  • Interestingly, the test performed better in those with advanced fibrosis.  The mean value of IL-33 in those with grade 3/6 was 88.2 compared to 37.2 for 1/6 and 70.9 for 2/6. In comparison, the children with cholestasis due to other liver disease had a level of 18.5 for those with 3/6 fibrosis

The authors note in a prior study that IL-33 was higher in BA infants than those with a choledochal cyst.

While this is a small study, I disagree with the editorial (pg 278-9) which largely discounted the potential role of IL-33.  “IL-33 is elevated with many other diseases (bronchopulmonary dysplasia, asthma, allergy, and more) It, therefore, cannot easily be used as a highly specific marker for fibrosis. Furthermore, the use of IL-33 as a prognostic marker, is from a clinical point of view not of great importance, as follow-up clinical decisions are generally made based on patients’ clinical course.”

Shankar et al provide data on GGT values in BA (n=113 infants).

These infants underwent Kasai procedure at a median of 61 days

  • 12.3% had normal (<200) GGT values.
  • Those with normal GGT had worse outcomes: earlier need for liver transplantation (14 vs 20 months) and poorer transplant survival.
  • 9/14 (64%) with normal GGT and 53/99 (53.5%) of elevated GGT underwent liver transplantation

The authors note that decreased levels of GGT has been associated with reduced glutathione metabolism which could impari adaptive response to oxidative stress, leading to further hepatocyte injury.

My take: In my experience, I have had very few BA patients with GGT values <200 (lower than 10%).  The development of other biomarkers like MMP-7 and IL-33 increase the likelihood that BA will be recognized sooner and if elevated, could obviate the need for a liver biopsy prior to operative cholangiogram.  Nevertheless, practitioners cannot wholly rely on any the current biomarkers.

Related blog posts:

Fall on UNC Campus, Chapel Hill

Pediatric Liver Enzyme Reference Values

A recent study (S Bussler et al. Hepatology 2018; 1319-1330)  provides reference age-related reference values for ALT, AST, and GGT for children from 11 months of age through 16 years.  This study from Leipzig used the LIFE Child prospective longitudinal population-based cohort with 3,131 cases (normal weight).

  • Figure 2 provides age- and sex-related percentiles for ALT.  At all ages and in both genders, the 90% cutoff was ~30 IU (or less).
  • Figure 3 provides age- and sex-related percentiles for AST.  There was a negative sloping curve with both genders, such that the 90% cutoff was ~60 IU at ~1 yr and ~40 IU at ~13 yr. Females tended to have modestly higher values.
  • Figure 4 provides age- and sex-related percentiles for GGT.  At all ages and in both genders, the 90% cutoff was ~25 IU. Males tended to have modestly higher values.

These values overall are similar to previous studies. From NASPGHAN NAFLD Guidelines: For ALT: “In the United States, sex-specific biologically based cutoffs have been determined from nationally representative data and have been validated in a fairly diverse cohort. These cutoffs are 22 mg/dL for girls and 26 mg/dL for boys. A Canadian study found the upper limit of normal for ALT to be 30 mg/dL in children 1 to 12 years of age, and 24 mg/dL in those between 13 and 19 years”

Related blog post: Pediatric NAFLD Guidelines 2017

Near Banff


#NASPGHAN17 Presentations at Annual Meeting: GGT in PSC, Nutrition for Intestinal Failure

This blog entry has abbreviated/summarized this presentation. Though not intentional, some important material is likely to have been omitted; in addition, transcription errors are possible as well.

Improvement in GGT Predicts Event-free Survival in Primary Sclerosing Cholangitis Regardless of Ursodeoxycholic Acic Treatment. 

Mark Deneau et al. (Grand Watkins Prize).

Key points:

  • PSC is difficult to study due to its rarity and due to its slow progression; thus surrogate biomarkers are needed.
  • Alkaline phosphatase is not a good biomarker in children
  • GGT level at one year after diagnosis was predictive of prognosis
  • Ursodeoxycholic acid does not appear to be effective

Optimizing Nutrition in Intestinal Failure

Justine Turner, University of Alberta

Key points:

  • Human milk is an ideal “formula” for infants, including those with intestinal failure
  • Oral feedings are important
  • Combination of bolus feeds and continuous feeds is reasonable
  • SMOFlipid allows higher lipid dose administration without hepatoxicity; this may improve cognitive outcomes
  • Amino acid based formulas have higher osmolality which can contribute to diarrhea

Patients with >50% of small bowel and >50% of colon were most likely to achieve enteral autonomy (GIFT registry)



Predicting Future Liver Disease with GGT Levels in Biliary Atresia Patients

A recent study (AJ Freeman, VL Ng, S Harpavat, A Hrycko, Z Apted, P Bulut, T Leong, SJ Karpen. Clin Gastroenterol Hepatol 2017; 15: 1133-35) describes the predictive value of γ-glutamyltransferase (GGT) in predicting thrombocytopenia/portal hypertension among biliary atresia patients.

In this retrospective study from three centers who had followup for at least 4 years, GGT values at 2 years of age were examined among biliary atresia patients (n=46) who continued with their native liver.

Key findings:

  • GGT ≥100 U/L had a predictive positive relationship with thrombocytopenia at 4, 5, and 6 years of age.  Patients with elevated GGT had lower platelet count (160 vs. 211) and their values continued to decline. GGT ≥100 U/L at 2 yrs predicted thrombocytopenia (<150) at age 4 with a sensitivity of 0.88, specificity of 0.57.
  • Patients with normal GGT values had “essentially stable platelet counts over the next 4 years.” GGT <100 U/L at 2 yrs predicted a low risk of thrombocytopenia with negative predictive value of 0.89, 0.92, and 0.93 at age 4, 5, and 6 respectively.

My take: This study quantitates a useful point –patients with biliary atresia and elevated GGT values are likely to develop evidence of portal hypertension.

Brevard, NC

BRIC, PFIC, and nasobiliary drainage

Case reports, when effective, help clinicians understand meaningful differences in disease presentation; in addition, they highlight practical treatment approaches.  An excellent example of one such case report is the following:

  • Zellos A et al.  JPGN 2012; 55: 88-90

These authors present a case with unique features that highlight some of the clinical problems with benign recurrent intrahepatic cholestasis (BRIC) and progressive familial intrahepatic cholestasis (PFIC).  BRIC1 and PFIC1 are associated with mutations in ATP8B1; BRIC2 and PFIC2 are associated with mutations in ABCB11.  The primary difference between BRIC and PFIC is the phenotypic expression.  In BRIC, individuals have episodes of cholestasis; in PFIC, progressive chronic liver disease develops in the first months of life.  PFIC2/ABCB11 mutations cause defective bile salt export pump (BSEP) at the bile-canniculus membrane.

Both ATP8B1 and ABCB11 intrahepatic cholestasis conditions present in a similar fashion with low GGT values.  In this case report, a 5-year-old presented with jaundice, acholic stools and dark urine.  His laboratory values revealed an ALT of 60 U/L, direct bilirubin of 7.6 mg/dL and gamma-glutamyl transpeptidase (GGT) of 10 U/L.  Initially, after exclusion of other liver conditions (eg. NL MRCP, copper studies, α-1 antitrypsin, autoimmune serology, infectious etiologies), the authors suspected ‘a clinicopathologic intergrade between BRIC and PFIC’ likely due to ATP8B1 as there was BSEP expression on liver biopsy immunostaining.  After sequencing did not demonstrate any ATP8B1 mutations, the authors identified two heterozygote mutations in ABCB11.

From a treatment standpoint, once nasobiliary drainage (NBD) was in place, the patient quickly improved.  This occurred after >6 weeks of failure with urosdeoxycholic acid/conservative measures.  As a precaution, the authors cultured the bile once a week and instituted antibiotic treatment when positive cultures were identified.

One other point alluded to by the authors is that the natural history of BRIC2 is poorly described.  Whether this disorder is truly ‘benign’ as the name suggests is unclear.  In patients with similar mutations who develop PFIC2, there is a high risk of hepatocellular carcinoma (HCC).

Intrahepatic Cholestasis Genes/Disorder (Clin Liver Dis 2006; 10: 27-53.)

Gene: Disorder (protein)
ATP8B1: PFIC1 -Byler’s (FIC1), BRIC, GFC -Greenland Familial
CLDN1: NISCH (Claudin 1) -neonatal sclerosing cholangitis/icthyosis
VPS33B: ARC syndrome (Vascular protein sorting 33) -arthrogryposis-renal dysfn-cholestasis, low GGT
AKR1D1: BAS: Bile acid synthetic defect: neonatal cholestasis with giant cell hepatitis
HSD3B7: BAS (C27-3β-HSD)
TJP2: (ZO-2) FHC: Familial hypercholanemia (tight junction protein)
EPHX1: FHC (epoxide hydrolase)
JAG1: Alagille (JAG1) JAG1 is transmembrane cell-surface protein important in regulating cell fate during embryogenesis
PKHD1: ARPKD (fibrocystin -important in ciliary function and tubulogenesis)
PRKCSH: ADPLD (hepatocystin)
ABCC2: Dubin-Johnson syndrome (MRP2)
CIRH1A: NAIC -N Amer Indian childhood cirrhosis (Cirhin)

Additional references for BRIC/low GGT PFIC:

  • -JPGN 2010; 51: 494.  Use of biliary diversion –helpful in 18 PFIC2 cases with long-term f/u.
  • -Liver Transplantation 2010; 16: 856.  6 patients developed recurrent low gamma-glutamyl transpeptidase cholestasis, that mimics BSEP disease, following transplantation. All had documented genetic defects in ABCB11 that were predicted to lead to a congenital absence of BSEP protein.
  • -NEJM 2009; 361: 1359. Recurrence of BSEP deficiency p OLT due to antibodies against BSEP
  • -Hepatology 2010; 51: 1645. n=62 children & clinical course.
  • -Gastroenterol 2008; 134: 1203. Severe BSEP –82 different mutations in 109 families. (n=132 patients)
  • -JPGN 2008; 46: 241. Excellent review. FIC1 caused by mutations in ATP8B1, PFIC 2 caused by mutations in ABCB11 which encodes BSEP – bile salt export pump. Increased risk of HCC in PFIC2 especially.
  • -J Pediatr 2007; 150: 556.  Increase risk of HCC in PFIC2.
  • -Hepatology 2006; 44: 478-486. Cases of pediatric HCC in PFIC-2
  • -Gastroenterol 2006; 130: 908. Review of canalicular transport defects.
  • -Hepatology 2005; 42: 222. summary of cholestasis workshop
  • -Gastroenterol 2004; 126: 322. Review of bile salt transporters.
  • -JPGN 2002; 34: 7A. FTT, diarrhea persist p biliary diversion or transplant.


  • -Gastroenterol 2003; 124: 1037-42. MDR3 mutations causing cholelithiasis, cholestasis, biliary cirrhosis, & pregnancy cholestasis.
  • -Gastroenterol 2001; 120: 1448-1458. n=31 cases. MDR3 mutations. ABCB4 gene
  • -Gastroenterol 2001; 120: 1459-67. Gallbladder stones & chronic cholestasis in 6 MDR3+ pts. Avg age of presentation: 2.9yrs. Avg age of Tx: 7.5yrs.  Sx/S : high ggt cholestasis, pruritus, intrahepatic cholestasis of pregnancy in heterozygotes (& c contraception)
  • -Hepatology 1996; 23: 904-8. MDR3 gene assoc c PFIC