Society: AGA

Background: Physiologic cholestasis is prevalent in preterm infants, occurring in ~15% of preterm births. Cholestasis causes poor neonatal growth and may progress to liver failure and death. Underlying causes of most cases are unknown. The potent choleretic ursodeoxycholic acid (UDCA) is used to treat cholestasis; however, its effects in preterm neonates are unclear. Microbial bile salt hydrolase (BSH) enzymes are among the numerous factors that mediate communication between the liver and the gut microbiome. BSH enzymes perform the first step in transforming primary bile salts into secondary bile salts through the deconjugation of glycine or taurine. In turn, bile salts shape gut microbiota composition and function. We sought to determine how the liver-gut-microbiome axis develops over time in preterm neonates, and whether cholestasis alters this development. Methods: We conducted a nested case-control study collecting 124 stool samples longitudinally from 24 preterm infants (mean 27.2 ± 1.8 weeks gestation), half of whom developed physiologic cholestasis. Samples were analyzed by whole metagenomic sequencing, quantitative mass spectrometry, and an in vitro BSH enzyme activity assay optimized for low microbial biomass stool samples. Results: Principal coordinate analysis revealed that gut microbiota from control preterm neonates develops over time in a predictable manner (Figure 1, P<0.0001). In control preterm neonates, acquisition of the secondary bile acid biosynthesis pathway is the most distinctive metagenomic feature of preterm microbiome development (P<0.00001). Control neonates also have increasing abundance over time of BSH genes carried by Clostridium perfringens (P<0.0001). Cholestasis interrupts this developmental pattern (P<0.05). BSH enzymatic activity is reduced in cholestatic neonates compared to controls (Figure 2, P<0.01), resulting in decreased unconjugated bile salts (46% vs 98%). Total fecal bile salt excretion was reduced in cholestasis (P<0.0001) and restored completely by enteral UDCA. However, samples from cholestatic neonates treated with UDCA had 522-fold higher quantities of fecal UDCA compared to untreated cholestatic neonates (P<0.0001), consistent with enteral UDCA being poorly absorbed in the neonatal small bowel. The majority of bile salts in early development are atypical 6-hydroxylated cholanoic acid isomers; the proportion of these isomers is reduced in cholestatic neonates (29.6% vs. 66.9%). Finally, we identified novel associations between metabolites within the BSH pathway, bile salt isomers, and neonatal growth. Conclusion: Acquisition of C. perfringens, the BSH gene, and the ability to produce secondary bile acids are key developmental features in preterm neonates that are absent in cholestasis. Metabolites within the BSH pathway warrant further study as biomarkers that may predict preterm infant growth.

<b>Figure 1</b>: Cholestasis disrupts gut microbiome maturation.

Figure 1: Cholestasis disrupts gut microbiome maturation.

<b>Figure 2</b>: Bile salt hydrolase activity is reduced in cholestatic neonates.

Figure 2: Bile salt hydrolase activity is reduced in cholestatic neonates.

Background: Adult patients with biliary acute pancreatitis (BAP) and/or choledocholithiasis who do not undergo cholecystectomy on index admission have worse outcomes, however, there is a paucity of data on the role of cholecystectomy during index hospitalization in the pediatric population. Our aim was to determine outcomes and readmission rates among pediatric patients with BAP and/or choledocholithiasis who underwent index cholecystectomy versus those who did not.

Methods: We performed a retrospective study of pediatric (<18 years old) admitted with BAP without infection or necrosis (ICD-10 K85.10) or choledocholithiasis (K80.3x-K80.7x) using the 2018 National Readmission Database (NRD). Exclusion criteria included age ≥18 years, necrotizing pancreatitis with or without infected necrosis, and death on index admission. Multivariable logistic regression was performed to identify factors associated with 30-day readmissions.

Results: We identified 1,122 unique index acute biliary pancreatitis (n=377, 33.6%) and choledocholithiasis (n=745, 66.4%) admissions. Mean age was 13 +/- 4.2 years. The majority of patients were female (n=792, 70.6%). Index cholecystectomy was performed in 663 (59.1%) cases and endoscopic retrograde cholangiography (ERCP) was performed in 363 (32.4%) cases. The thirty-day readmission rate for the entire cohort was 10.9% in patients who underwent index cholecystectomy and 48.8% in those who did not (p<0.001). Among patients with BAP, the 30-day readmission rate was 5.4% in patients who underwent index cholecystectomy and 20.7% in those who did not (p<0.001). Among patients with choledocholithiasis, the 30-day readmission rate was 14.4% in patients who underwent index cholecystectomy and 58.3% in those who did not (p<0.001). On multivariable analysis, patients who underwent index cholecystectomy or ERCP had lower odds of 30-day readmission than those who did not (OR 0.16, 95% CI 0.11-0.22, p<0.001) and (OR 0.26, 95% CI 0.18-0.40, p<0.001), respectively. Additionally, for every 1-year increase in age, the odds of readmission decreased by 6% (OR 0.94, 95% CI 0.91-0.98, p=0.001).

Conclusions: Index cholecystectomy was performed in only 59% of pediatric patients admitted with BAP or choledocholithiasis, but was associated with a staggering 84% decreased odds of readmission within 30 days. This is the largest study to date demonstrating the clinical benefit of index cholecystectomy in pediatric patients admitted with BAP or choledocholithiasis. Current guidelines should be updated to reflect these findings, and further work is needed to help translate these findings into clinical practice given the superior outcomes of index cholecystectomy.

Introduction: Biomarkers for acute pancreatitis (AP), limited to amylase or lipase, and chronic pancreatitis (CP) diagnosis based on imaging alone have poor specificity, limited application in children. There is a need for novel AP and CP biomarkers to ensure an early and accurate diagnosis. Here we aim to use urine proteomics to identify AP, and CP biomarkers that differentiate between AP and CP respectively, and from the control group including patients with pain from a bone fracture and healthy individuals.
Methods: We quantified urine proteins using the chromatogram library approach for data-independent acquisition mass spectrometry, where a sample pool was used to identify a complete ‘Library’ proteome. This library was used to quantify proteins in single-injection experiments. Differential biomarker candidates were identified with the false-discovery rate (FDR) corrected analysis of variance (ANOVA) analysis (FDR<0.05) and additional pairwise t-test comparisons to identify group-specific markers (FDR<0.05). Linear discriminant analysis (LDA) was used to derive a classifier from training data (50% of the cohort) that could differentiate between AP, CP, and control groups, which was validated using the remaining 50% of the cohort.
Results: We studied a cohort of 28 AP, 50 CP, 31 bone fracture controls, and 21 healthy controls with a mean age of 12 years, where 54% were female. The diagnosis of AP and CP was based on the consensus criteria and confirmed on imaging. We identified 2,137 urine proteins, from which we quantified 1,694 across all sub-groups. Of these, we detected 373 differential proteins for AP and 844 differential proteins for CP. The top AP-specific proteins included CELA2A, CRP, REG1A, POF1B, and AMY2A. Whereas the top CP-specific proteins included IDH1, S100P, CAPS, SNCG, and S100A6. Interestingly, in our cohort CELA2A (Elastase 2A) exhibited a 61-fold increase in AP compared to controls, while the current clinical biomarker AMY2A (Amylase) only had a 2.2-fold increase. Using LDA of combined proteins, we found that CELA2A+CRP+AMY2A performs better than any other combination with an AUROC of 97%, compared to 86% for AMY2A alone for the diagnosis of AP. At a low false positive rate (5%), the prediction accuracy for AMY2A was 59%, while CELA2A+CRP+AMY2A had 80% for AP diagnosis.
Conclusion: Our urinary proteomics results identify Elastase 2A as a novel biomarker that accurately diagnoses AP in children, and accurately discriminates from CP, healthy controls, and other causes of pain when used in combination with CRP and amylase. Additional validation studies are needed to evaluate the role of urine Elastase 2A as a biomarker for AP, in comparison to serum lipase among children and adults, and potentially develop a point-of-care test.

Figure 1: Volcano plot showing discovery cohort results for A: Acute Pancreatitis and B: Chronic Pancreatitis. The proteins in ‘red’ meet the FDR-corrected p-value threshold. Red panel shows Elastase 2A which stands out as a potential diagnostic biomarker for Acute Pancreatitis diagnosis.

Figure 1: Volcano plot showing discovery cohort results for A: Acute Pancreatitis and B: Chronic Pancreatitis. The proteins in ‘red’ meet the FDR-corrected p-value threshold. Red panel shows Elastase 2A which stands out as a potential diagnostic biomarker for Acute Pancreatitis diagnosis.

Figure 2: A - Box plots representing fold changes for key AP-specific protein biomarkers. B - ROC Plots for linear discriminant analysis demonstrating that combination of biomarkers including Elastase 2A has high accuracy in diagnosing AP.

Figure 2: A - Box plots representing fold changes for key AP-specific protein biomarkers. B - ROC Plots for linear discriminant analysis demonstrating that combination of biomarkers including Elastase 2A has high accuracy in diagnosing AP.

Background:
Non-alcoholic fatty liver disease (NAFLD) is considered the most common liver disease in the developed world with rapidly rising pediatric prevalence over the past 2 decades. The average age of onset is between ages 10-12 years with higher male predominance. Liver biopsy is the gold standard for diagnosis but is associated with significant risks. Laboratory assessment including hepatic function panels are commonly used for screening and diagnosis of NAFLD along with abdominal US. Ultrasound (US) is the most common imaging modality used as it is non-invasive, readily accessible and cost-effective. Liver span and echogenicity can serve as surrogates for hepatic abnormalities. US are commonly read in ambulatory radiology centers staffed by non-pediatric radiologists. There is a concern that these radiologists may not be as familiar with pediatric anatomy (including liver size that grows with age) which may negatively impact their interpretation.
Methods:
An IRB-approved retrospective chart review was performed of pediatric electronic medical records from Johns Hopkins (JHH), Mount Washington Pediatric Hospital, and Sinai Hospital of Baltimore. Inclusion criteria included patients between 2-18 years of age seen between 2015-2018 with complaints of upper abdominal pain, abnormal liver enzymes, or hepatomegaly with US performed. Abdominal US performed at JHH were reviewed by a single trained pediatric radiologist in the study for comparison. Liver and spleen dimensions and liver characteristics on repeat reads were compared to original results.
Results:
194 children were eligible. Mean age was 12.2 +/-4.1 years; 58.8% were male. Mean BMI was 28.48 +/- 9.48 kg/m²and mean Z-score was 1.52. Patients were most commonly Caucasian (41.2%), African American (24.7%), or Hispanic (21.1%). Figures 1, 2 and Table 1 show demographics and diagnosis.
85 of the 194 US were performed at JHH and reviewed by a single pediatric radiologist. 14 of the 85 US (16.5%) secondary reviews detected heterogenous echogenicity of the liver. 10 of 85 US (4.7%) secondary reviews detected enlarged liver dimensions for the child's age. Figures 3 and 4 compare location of ultrasound performed and highlight the difference in US diagnoses based on pediatric radiologist review.
Conclusions:
As NAFLD is becoming more prevalent, US remains a common and useful tool for screening. There is concern for alternate diagnosis with interpretation without attention to age-related standards. Determination of increased or heterogenous echogenicity and liver span is crucial for diagnosis of fatty liver disease or fibrosis.

Kenneth Ng

Appearances