Prognosis of Biliary Atresia Associated With Cytomegalovirus: A Meta-Analysis

Objective: The etiology of biliary atresia is unclear, but viral infection has been implicated. The aim of the current meta-analysis was to investigate relationships between cytomegalovirus (CMV) and the prognosis of biliary atresia. Methods: PubMed, Embase, the Cochrane Library, the China National Knowledge Infrastructure database, and Wanfang Data electronic databases were searched for eligible studies. Each relevant text was thoroughly reviewed and examined, including related papers in their reference lists. Results: A total of nine studies including 784 patients were included in the analysis. Biliary atresia patients with CMV exhibited significantly lower jaundice clearance (odds ratio: 0.46, p < 0.0001; I2 = 15%, p = 0.31). There were no significant differences in the rates of cholangitis or native liver survival. CMV-pp65-positive biliary atresia patients had a significantly lower rate of jaundice clearance (odds ratio: 5.87, p = 0.003; I2 = 0%, p = 0.71) and a significantly higher rate of cholangitis (odds ratio: 0.21, p = 0.01; I2 = 0%, p = 0.43) than CMV antibody-positive biliary atresia patients. Conclusion: Biliary atresia patients who were also infected with CMV had a poorer prognosis, particularly with respect to jaundice clearance. CMV status may influence the prognosis of biliary atresia. Clinicians should be able to routinely identify the subset of biliary atresia patients who are also CMV-positive, in order to improve native liver survival.


INTRODUCTION
Biliary atresia (BA) is a destructive and obliterative cholangiopathy that occurs in infants. It involves both the extra-hepatic and the intra-hepatic bile ducts, and it is characterized by progressive inflammation and fibrosis of the biliary tree (1). The incidence of BA ranges from 1/5,000 to 1/20,000, and it is particularly high in Asian countries (2,3). The main treatment for BA is currently Kasai portoenterostomy (KPE), which re-establishes bile flow and prolongs the survival of the native liver (4,5). The etiology of BA may be associated with immune dysregulation, inflammation, and virus infection, but it is still unclear (6,7). Cytomegalovirus (CMV) affects BA. Evidence of CMV infection has been reported in 10-38% of infants with BA in series from England, Germany, Brazil, and Sweden, but up to 60% in China (4,6,8). In a series of studies, CMV-associated BA patients exhibited poor outcomes, but other studies have yielded inconsistent results. These inconsistences may be due to the relatively small sample sizes involved. The present study aimed to investigate outcomes in patients with CMV-associated BA after KPE.

MATERIALS AND METHODS
The present review and meta-analysis were conducted in accordance with the Assessing the Methodological Quality of

Study Selection and Data Extraction
The inclusion criteria were (1) studies that included CMV detection in serum or urine; (2) randomized clinical trials (RCTs) or prospective/retrospective studies comparing CMV + BA and CMV − BA; (3) studies that included at least one of the following outcomes: "clearance of jaundice, " "cholangitis, " "NLS in 2 years, " or "survival curve on NLS"; and (4) studies with scores ≥6. Case reports, meeting abstracts, review articles, letters, and opinions were excluded. Two investigators independently assessed the eligibility of the studies by reviewing titles, abstracts, and, when required, full texts. Differences were settled via discussion. The extracted data recorded included first author, year of publication, study type, number of patients, jaundice clearance and cholangitis data, rate of native liver survival (NLS) in 2 years, and NLS survival curves. CMV can be detected via several methods. In the present analysis, CMV − BA was defined as all negative indicators in both blood samples and urine samples, and any deviation from that was defined as CMV + BA. CMV-Ig-positive BA was defined as CMV antibody-positive but CMV-pp65-negative, and patients who were positive for CMV-pp65 were defined as CMV-pp65-positive.

Statistical Analysis and Exploration of Heterogeneity
Stata SE15.0 was used for the analysis. Odds ratios (ORs) with 95% confidence intervals (CIs) were calculated based on the reported numbers of patients and events. Hazard ratios (HRs) with 95% CIs were calculated according to calculate lnHR and its variance by NLS survival curves (9). Mean differences (MDs) with 95% CIs were used for continuous outcomes. Differences of p < 0.05 were considered statistically significant. Heterogeneity was assessed via the Cochrane Q test and I 2 values (10). p < 0.1   Frontiers in Pediatrics | www.frontiersin.org or I 2 < 50% was considered to indicate low heterogeneity and then a fixed effects model was used; otherwise, a random effects model was used (10).

Quality Assessment and Publication Bias
The quality of the studies included was evaluated via the Newcastle-Ottawa Scale (NOS) by two reviewers independently. Studies with NOS scores ≥6 were considered high quality and were included in analyses. Publication bias and jaundice clearance results were evaluated via Egger's test, but other parameters were not analyzed due to the small number of studies.

Search Results and Study Characteristics
A flowchart of the search results is shown in Figure 1. A total of 784 records were identified and evaluated by two investigators independently. After excluding 177 that were duplicates, a further 552 were excluded after screening the titles and abstracts. A further 42 studies were excluded due to no access to the full text or complete data. Nine studies published between 2007 and 2020 ultimately met the inclusion criteria and were included in the meta-analysis (6,(11)(12)(13)(14)(15)(16)(17)(18). Detailed characteristics of the selected studies are summarized in Table 1. Eight of the studies included were retrospective and one was prospective, and collectively they included a total of 694 BA patients. Of these, 294 were classified as CMV + BA patients. Characteristics of the reports included are shown in Tables 1, 2.

Quality Assessment and Publication Bias
NOS scores are shown in Table 1. An Egger's funnel plot derived from the studies describing jaundice clearance in the current meta-analysis is shown in Figure 7. Egger's test indicated no significant publication bias (t = −1.77, p = 0.127).

DISCUSSION
CMV is a hepatophilic double-stranded DNA virus (6). Domiati-Saad et al. (19) reported that CMV was related to BA in a PCR-based study conducted in the 1990's, and Accordingly, in the present study, any positive indicator in blood or urine was used to define CMV + BA, and a lack of any such indicators was used to define CMV − BA. CMV + BA is associated with a poor prognosis. In the present metaanalysis, patients with CMV + BA had a significantly lower rate of jaundice clearance. Some studies reported that CMV + BA patients presented symptoms later and were operated on later (6,11,16). CMV + BA was associated with more severe fibrosis and inflammation during surgery than CMV − BA (12,14). There were indications that the virus triggered proinflammatory mechanisms, and this may lead to an autoimmune response guided by Th1 cells (21). At the time of KPE, liver fibrosis and inflammation were more severe in CMV + BA patients than in Various test measurements were used to detect the presence and status of CMV, including serum CMV-IgG, CMV-IgM, and CMV-DNA. CMV-pp65 can be detected in serum and urine. Positive CMV-IgM indicates recent infection that has been cured. CMV-IgG can be transferred through the mother's placenta and then it disappears gradually over the subsequent 2 years. CMV-pp65 antigenemia indicates duplication of CMV. CMV-DNA is an indicator of early active infection. In the present analysis, CMV-pp65-positive BA was associated with a poor outcome with respect to jaundice clearance and cholangitis. In Chun et al. (12) and Luo et al. (18), CMV-pp65-positive BA was associated with more rapid development of liver fibrosis  than in other BA patients, including those who were CMV antibody-positive. CMV status may influence the outcome of BA. It is essential for clinicians to test for CMV status. Measurements derived from serum or urine do not represent the status of the liver. It remains a challenge to diagnose active CMV infection in the liver. Most pediatric surgeons agree that CMV is one of the etiologies of BA. CMV inclusion bodies have not been detected in the liver in any reported studies. Many problems remain to be solved with respect to CMV in BA. Antiviral therapy has been proposed to improve the prognosis of CMV + BA. In a study reported by Parolini et al. (24), antiviral drugs could negate the pathogenic effects of CMV, improve the rate of jaundice clearance, and reduce cholangitis. Notably however, there is not enough evidence to support the principle of adjuvant therapy, which currently depends on clinicians.
The present meta-analysis had several limitations. One was the lack of RCTs. Another was that the sample sizes of several of the studies included were relatively small. Some prognostic indicators were only included in a few of the studies analyzed, resulting in reduced statistical power and publication bias. No antiviral treatment was explicitly mentioned in any of the nine articles included. Lastly, there was a lack of comparisons of outcomes in CMV + BA patients undergoing antiviral therapy vs. CMV − BA patients.
In conclusion, the current meta-analysis indicates that CMV-IgM BA is a noteworthy subgroup of BA patients. BA patients with CMV exhibited poor prognosis, particularly with respect to jaundice clearance. CMV status may influence the prognosis of BA. Clinicians should be able to routinely identify CMV + BA patients, to facilitate attempts to improve NLS.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s.