Biomarkers for Diagnosis and Post-Kasai Portoenterostomy Prognosis of Biliary Atresia: Systematic Review and Meta-Analysis

Abstract

Objectives To evaluate the accuracy of biomarkers for the early diagnosis of biliary atresia (BA) and prognostic stratification after Kasai portoenterostomy (KPE). Methods We conducted a systematic review of PubMed, Web of Science, Embase, Scopus and OVID for English literatures reporting BA biomarkers published before August, 2020. Biomarkers selected for study was based on a good evidence-level (better than Grade B, level 2b). Screening, data extraction, and quality assessment were done in duplicate. Results A total of 51 eligible studies were involved for systematic review, and data from 12 of them (4182 subjects) were extracted for meta-analysis in the following 2 domains: (1) serum matrix metallopeptidase-7 (MMP-7), interleukin33 (IL-33) and γ-glutamyl transferase (GGT) to differentiate BA from non-BA; and (2) aspartate aminotransferase to platelet ratio index (APRi) to predict post-KPE liver fibrosis/cirrhosis. Summary sensitivity, specificity and the area under the curve (AUC) of MMP-7 for diagnosing BA were 96%, 91% and 0.9847, respectively. Summary sensitivity and specificity of IL-33 for diagnosing BA were 77% and 85%, respectively. Summary sensitivity, specificity and AUC of GGT for diagnosing BA were 80%, 79% and 0.9645, respectively. Summary sensitivity and specificity of APRi for predicting post-KPE liver fibrosis were 61% and 80%, respectively. Summary sensitivity, specificity and AUC of APRi for predicting post-KPE cirrhosis were 78%, 83% and 0.8729, respectively. Moreover, good evidence-level was shown in the investigations of serum IL-18 and IL-33 in distinguishing BA from healthy control, serum IL-18 for prognosing post-KPE persistent jaundice, and serum hyaluronic acid and MMP-7 for prognosing post-KPE significant liver fibrosis. Conclusions MMP-7, IL-33 and GGT are useful biomarker to assist the diagnosis of BA. APRi could be used to predict post-KPE significant liver fibrosis and cirrhosis. These non-invasive biomarkers could be integrated into the management protocol of BA.