This prematurely-born infant presented with a cholestatic jaundice that transiently improved after discontinuation of parenteral nutrition but then recurred and was ultimately found to have BA. At the time of his initial liver biopsy, the patient had prolonged exposure to TPN and had only recently begun to tolerate full feeds. These factors, combined with an improving conjugated bilirubin, led to the decision that close clinical observation was appropriate since these findings were consistent with PNAC. Patients on TPN frequently have a conjugated hyperbilirubinemia and can have concurrent biochemical markers of liver injury as well as histologic changes ranging from steatosis to obstructive features with portal fibrosis and cirrhosis [
8]. This patient’s conjugated hyperbilirubinemia was initially felt to be related to TPN initially, especially since work-up for additional causes of cholestasis was negative. In general, there is an expectation that after discontinuation of TPN, cholestasis should improve, although the time to resolution is not well characterized in the literature. Some patients have a rise in their conjugated bilirubin in the first month after TPN discontinuation prior to resolution. [
9]
Although this patient initially followed an expected course for PNAC, his recurrent biochemical evidence of cholestasis and liver injury and continued jaundice prompted repeat evaluation. Particularly challenging in this case is the lack of histologic distinction between BA and PNAC on liver biopsy [
10]. Since BA and PNAC share histologic findings of obstructive cholestasis, additional tests are needed to help distinguish the diagnoses. HIDA scan is commonly used to help distinguish BA from non-BA etiologies and has been shown to have high sensitivity (98.7%) but low specificity (70.4%) [
11], meaning that 30% of patients with non-excreting HIDA scans will ultimately have a normal extrahepatic biliary tree. Both PNAC and BA cause obstructive cholestasis, and recent studies have shown no pathologic basis for distinguishing the two diseases [
10,
12]. Given the limits of imaging, biomarker, and pathology studies, intraoperative cholangiogram with the option for HPE remains the gold standard for the diagnosis of biliary atresia, especially given the time-sensitive nature of diagnosis and intervention [
4], and was the modality that provided the definitive diagnosis in this case. A question that naturally emerges for the patient in question was whether he should have undergone interoperative cholangiogram with his initial cholestasis evaluation. Although it seems that an earlier cholangiogram could have identified his BA sooner, it is also possible that the full obliteration of his biliary tree had not yet occurred, and he could have had a falsely negative intraoperative cholangiogram. A second question that emerges is when and how to evaluate patients with long-term TPN requirements for biliary atresia, which was especially challenging in our patient. Finally, it remains unclear from the literature whether the timeline for premature infants should include their corrected or chronological age [
7].
This case highlights the ongoing barriers that exist for timely diagnosis of biliary atresia despite the availability of multiple diagnostic tests. Clinicians who care for patients who may have multifactorial causes of their cholestasis should be aware that liver biopsy by itself cannot distinguish biliary atresia from PNAC and know the appropriate places to refer patients who may need invasive diagnostic testing (e.g. intraoperative cholangiogram). This study also reinforces the need for more non-invasive diagnostic tests that are specific to biliary atresia.