Diagnostic Tools for Early Detection of Biliary Atresia: Is a Newborn Screen Attainable?

Short abstract

Content available: Audio Recording

Abbreviations

BA

biliary atresia

HPE

hepatic portoenterostomy

MMP

matrix metalloproteinase

NBS

newborn screening

SCC

stool color card

Listen to an audio presentation of this article.

Biliary Atresia

Biliary atresia (BA) is a devastating disease of infants characterized by bile duct injury and extrahepatic biliary obstruction, resulting in cirrhosis in the majority of patients. Although BA is a rare disease, occurring in ~1 in 5600 to 1 in 18,000 infants worldwide, it is the most common indication for liver transplantation in children in the United States. Two distinct phenotypes have been described, an acquired form theorized to be due to a perinatal insult associated with autoimmune/autoinflammatory responses targeting the biliary tree (“isolated BA”; ~85% of cases), or an embryonic fetal form theorized to be due to defective genetic regulation of bile duct and spleen development (“BA splenic malformation syndrome”; ~15% of cases). ¹ At diagnosis, the Kasai hepatic portoenterostomy (HPE) is performed in an attempt to reestablish bile flow. Prompt recognition and treatment with HPE by 30 to 45 days of life has been associated with higher rates of clearance of jaundice and transplant‐free survival (Fig. 1). ² The objective of this review is to focus on current strategies that may lead to an earlier diagnosis of BA, as well as outline plans for a future public health initiative aimed at diagnosing BA within the first few weeks of life.

FIG 1.

BA survival with the native liver based on age at time of HPE. Shown are the long‐term results of survival with the native liver from patients with BA in France, 1986‐2015. Adapted with permission from Journal of Pediatric Gastroenterology and Nutrition. ² Copyright 2019, Lippincott Williams & Williams.

Diagnostic Tools in BA

Direct Bilirubin

A landmark study by Harpavat et al. ³ revealed that infants who were later diagnosed with BA had significant elevation of the direct bilirubin in the newborn nursery. A subsequent strategy to identify infants with BA at an earlier age examined a two‐stage neonatal direct bilirubin screening study. ⁴ In stage 1, newborn direct bilirubin levels were tested within 60 hours of life, and a positive screen was considered a direct bilirubin greater than the 95th percentile. Infants with a positive stage 1 screen underwent stage 2 screening, which was a retest at the infant’s 2‐week well child visit. At stage 2, a positive screen was defined as a direct bilirubin greater than the stage 1 result or greater than 1 mg/dL. Of the 124,385 newborns in this study, direct bilirubin testing had a sensitivity of 100%, specificity of 99.9%, positive predictive value of 5.9%, and negative predictive value of 100.0%. This screening mechanism resulted in earlier diagnoses of BA, with a mean age at HPE of 36 days of life compared with 56 days of life at time of HPE for those not involved in the study, suggesting that direct bilirubin is a promising diagnostic biomarker for BA.

Matrix Metallopeptidase‐7

Matrix metallopeptidases (MMPs) are a family of endopeptidases involved in remodeling of the extracellular matrix, liver regeneration, and both antifibrotic and profibrotic responses. A large‐scale proteomics approach led to the ground‐breaking finding of serum MMP‐7 as a specific biomarker for BA diagnosis. ⁵ In that study, the age at time of MMP‐7 analyses was ~60 days. The median concentration of MMP‐7 was 2.86 ng/mL in healthy infants, 11.47 ng/mL in non‐BA cholestatic subjects, and 121.1 ng/mL for patients with BA (diagnostic sensitivity, 98.67%; specificity, 95.00%; negative predictive value, 98.28%). A validation study of MMP‐7 in healthy control subjects and infants with cholestasis across three pediatric centers worldwide found that an MMP‐7 cutoff value of >52.8 ng/mL accurately predicted BA. ⁶ Given that serum MMP‐7 has a high sensitivity and specificity to differentiate BA from non‐BA neonatal cholestasis, it can be considered as a reliable biomarker for the detection of BA. What is not known, however, is how accurate MMP‐7 would be in differentiating BA from non‐BA in the newborn nursery setting, and future research should focus on MMP‐7 detection in newborns as a screening tool for BA.

Serum Bile Acids

Newborns have a transient physiological cholestasis as a result of immature bile acid circulation, but newborns with cholestatic liver disease have a more significant and persistent elevation in serum bile acids. ⁷ Hence measuring total serum bile acids has been a proposed screening tool for BA. Zhou et al. ⁸ found that taurocholate levels had a sensitivity of 79.1% and a specificity of 62.5% in the diagnosis of BA. A follow‐up study on the use of bile acids in the diagnosis of BA included a scoring system composed of glycochenodeoxycholic acid/chenodeoxycholic acid levels, acholic stools, and gamma‐glutamyl transferase. A cutoff value of 15 identified BA with a receiver operating characteristic of 0.87, sensitivity of 85.3%, and specificity of 81.3%. ⁹ Both studies analyzed bile acids at the time of diagnosis (i.e., ~60 days of age), and it will be important to determine whether bile acids are disease specific within the newborn screening (NBS) time frame.

Stool Color Card

Clinical features of BA include pale‐colored stools within the first few weeks of life. Japan became the first country to leverage this presentation by incorporating a stool color card (SCC), where families could identify and report pale‐colored stools in their infants. The SCC is given to families in the newborn nursery and now consists of nine representative colors of either normal or acholic stools (Fig. 2). ¹⁰ The families are instructed to monitor the color of the stool in the first 21 days of life and to contact the health care facility if their infant has one of the abnormal stool colors depicted on the card. Other countries, including Taiwan and Canada, have also analyzed the effectiveness of SCC in the earlier detection of BA. ¹¹ In the Taiwanese cohort, the SCC had a 97% sensitivity and 99% specificity for the accurate diagnosis of BA. ¹² Use of the SCC resulted in younger age at time of HPE, reduced liver transplant/mortality, and reduced hospitalization rate, making this a promising screening tool. ¹³

FIG 2.

SCC shows various stool colors from infants at time of diagnosis of BA (1‐6) compared with healthy infants (7‐9). Parents are advised to monitor stool color in the first month of life and to call the health care provider if infant has stool colors 1‐6. Adapted with permission from Perinatal Services BC. ¹⁰ Copyright 2019, Dr. Richard A. Schreiber, University of British Colombia/BC Children’s Hospital and Perinatal Services.

Public Health Initiative: A NBS for BA?

The development of a NBS for BA is critically important to have a positive impact on this disease. Many studies worldwide have shown that early detection and intervention with HPE improves outcomes. The purpose of a NBS is to detect fatal or disabling diseases in newborns before they show symptoms, so that patients may receive prompt treatment and minimize adverse outcomes. The criteria for creation of a NBS are summarized in Table 1, and candidate diagnostic tools are summarized in Table 2. Currently, the best candidate for a NBS initiative would be the direct bilirubin level in the newborn nursery, with follow‐up testing at 2 weeks of age. Limitations to this approach include the fact that most newborn nurseries screen for only total bilirubin, and most use a transcutaneous measurement. ¹⁴ Furthermore, the majority of conditions detected by the NBS use dried blood spots for tandem mass spectrometry; this technique is unable to accurately measure conjugated/direct bilirubin.

TABLE 1.

Considerations in Developing a NBS

1. Condition sought should be an important health problem.

2. There should be an acceptable treatment for patients diagnosed early on in the course of disease.

3. There should be a suitable screening test.

4. The screening test should be acceptable to the population.

5. There should be an adequate understanding of the condition’s natural history.

6. The cost of the case finding should be economically balanced in relation to possible expenditure on medical care as a whole.

7. Case finding should be a continuing process.

Adapted with permission from Wilson and Jungner. Copyright 1968, World Health Organization. ¹⁶

TABLE 2.

Candidate Diagnostic Biomarkers in BA

	Intervention Age	Method	Sensitivity (%)	Specificity (%)
Direct bilirubin	Birth and weeks of life	Blood draw	100	99.9
MMP‐7	At time of diagnosis	Blood draw	98.67	95
SCC	14‐21 days of life	Stool	97	99
Serum bile acids	Birth	Dried blood spot	79.1	62.5

With regard to the cost‐effectiveness of creating a NBS, the conjugated/direct bilirubin screening may be more expensive compared with the SCC. One study calculated that conjugated/direct bilirubin screening would cost $2.4 million more compared with SCC screening, with a total annual cost for performing the SCC of only ~$200,000. ¹⁵ SCC screening resulting in earlier BA diagnosis was estimated to save approximately $8 million over 20 years. ¹² Limitations for use of the SCC specific to the United States include the fact that there is no standard 1‐month well child visit to collect the stool cards and no national call center to triage calls. The SCC may be the most cost‐effective strategy; however, the conjugated/direct bilirubin screening is likely more accurate and would lead to the diagnosis of BA as early as 2 weeks of age.

Serum bile acids and MMP‐7 show promise for diagnostic accuracy; however, future studies are warranted to assess these as screening tools for BA in the newborn nursery. Serum bile acids can be detected in dried blood spots, which would use testing infrastructure already in place for routine NBS. ⁹ Ongoing advocacy efforts are necessary to facilitate the optimal screening strategy for an earlier BA diagnosis and subsequent improved outcomes.